id
stringlengths 20
20
| year
int64 2.01k
2.02k
| title
stringlengths 21
249
| sections
stringlengths 4.88k
32.5k
| headings
stringclasses 217
values | abstract
stringlengths 32
4.12k
| summary
stringlengths 227
2.45k
| keywords
stringlengths 7
1.31k
⌀ | toc
stringclasses 267
values |
|---|---|---|---|---|---|---|---|---|
journal.pcbi.1006593 | 2,018 | A regularity index for dendrites - local statistics of a neurons input space | The primary function of dendritic trees is to collect inputs from other neurons in the nervous tissue 1 , 2 ., Different cell types play distinct roles in wiring up the brain and are typically visually identifiable by the particular shape of their dendrites 3 ., However , so far no branching statistic exists that reliably associates individual morphologies to their specific cell class 4 , 5 , indicating that we have not yet identified the morphological features that are characteristic for the differences in how neurons connect to one another ., Theoretical considerations have provided systematic qualitative insight into the question of how dendrite shape relates to specific connectivity ., Dendrites are thought to collect their inputs using the shortest amount of cable and minimizing conduction times in the circuit 3 , 6–9 and they have been proposed to maximize the possible connection repertoire 10 ., Of the possible connections that a neuron could make by anatomical proximity only a small , relatively invariable number become functional synapses 11 ., But it has generally been proposed that the connection probability between a dendrite and an axon can be determined by the amount of anatomical overlap between the two 12–14 ., Furthermore , dendrite shape has been linked to the number of synapses based on the optimal wiring assumptions described above , linking total dendrite length and the number of synapses that determine the morphology 15 ., This leads to the question whether specific axonal arrangements or synapse distribution patterns may coincide with specific typical dendritic morphological characteristics 16 ., A useful concept to relate dendritic trees with their underlying connectivity comes from extended minimum spanning trees ( MSTs ) that connect a set of target points while minimizing total cable length and path lengths in the tree toward the root where signals get integrated 6 , 7 ., Such MSTs were shown to produce accurate dendritic morphologies when the corresponding target points were selected adequately 7 , 15 , 17 , 18 ., This approach has previously linked the distribution of target points to actual synapse locations , as well as to the number of branch points ( BPs ) and termination points ( TPs ) 15 ., Here , we study the relationship between local statistics of spatial input distributions and the respective dendritic morphology using MSTs generated on different target point distributions ., In order to do this we use a regularity index R that measures the degree of clustering of points in a given volume based on the average nearest neighbor ( NN ) distance ., Specifically , the statistic R is defined as the ratio between the observed mean NN distance of a set of points in a given volume and the expected mean NN distance of the same number of points distributed uniformly in the same volume , i . e . in a setting of complete spatial randomness ., Randomly distributed points from a uniform distribution ( i . e . samples from a Poisson process ) thereby yield a value of R = 1 in the limit ., When R > 1 , NN distances are on average larger than in a uniform random distribution , meaning that the points are distributed more regularly ., When R < 1 , NN distances are on average shorter , indicating that the points are more clustered than expected by chance ., The measure R has been used in a wide variety of scientific disciplines , such as physics , biology , geography and astronomy 19–22 ., In neuroscience , a variant of R was used to describe the regular spacing of ganglion cells in the retina 23–25 ., In particular , Cook performed a comparison of measures based on NN distances to study retinal mosaics 26 , including R , which they refer to as the dispersion index ., The statistic R has also been applied to graph theoretical problems such as minimum spanning trees 27 , but to the best of our knowledge it has not yet been considered to characterize neuronal morphology ., In the following , we first estimate R for BPs and TPs in real dendrites and then use MST-based morphological models to compare distributions of BPs and TPs with the underlying distribution of target points as a proxy for their corresponding synaptic input distributions ., The value of the statistic R , defined as the quotient of the average nearest neighbor ( NN ) distance to the one expected from a matching uniform distribution can be obtained for any given set of points in Euclidean space ., We first calculated R for the set of BPs and TPs in real dendrites to estimate how regularly the dendrites spread in the circuit ( Fig 1 ) ., One issue when computing R on finite point clouds that has been given little attention in the literature so far is that a naïve calculation yields a biased result due to boundary effects ., This bias is due to the fact that the enclosing volume of the point cloud is usually not known ., In the special case of planar convex volumes a bias correction can be performed analytically 28 , but these methods are not available for higher-dimensional and non-convex carrier volumes such as the ones occurring in our case ., We addressed this issue by developing a Monte Carlo ( MC ) based approach to estimate for a given volume the reference value for a Poisson process numerically and obtain the value of R without edge effects ., This method also provided us with specific confidence intervals for our estimates ( see Methods and S1–S4 Figs ) ., We then estimated R for BPs and TPs using our MC based approach , resulting in separate statistics RBP and RTP for the sets of branch and termination points , respectively ( Fig 2 ) ., The mean estimated values of R in four three-dimensional ( 3D ) and four two-dimensional ( 2D ) cell types varied widely ., For almost all cases we observed a tendency of RTP being slightly larger than RBP ., To study whether there were significant differences between the RBP and RTP values of different cell types , and because not all of the necessary assumptions for ANOVA were satisfied , we used the Kruskal-Wallis test and then applied the Mann-Whitney test with the Bonferroni method to adjust the p-values for pair-wise comparisons ., Using the Kruskal-Wallis test we found significant differences ( p-value < 2 . 2 × 10−16 ) in the four tests performed , namely between the RBP values in 3D cells , RTP in 3D cells and the two analogous cases in 2D ., In 3D dendrites , the spatial distribution of BPs and TPs was most clustered in motoneurons , followed by hippocampal pyramidal cells , neocortical pyramidal cells , and finally dentate granule cells ( Fig 2A ) ., Pair-wise comparisons revealed that there was no significant difference between the RBP values of neocortical pyramidal cells and hippocampal pyramidal cells or between RTP values of neocortical pyramidal cells and granule cells ., In the case of the four planar cell types ( Fig 2B ) , dendritic arborization ( da ) neurons in the fly larva were well characterized by the clustering of their BPs ., We found no significant difference between the RTP values of Lobula Plate tangential cells ( TCs ) in the fly , cerebellar Purkinje cells and Retinal ganglion cells , a large inhomogeneous group of cell types ., Pair-wise comparison also showed no significant difference between the RBP values of Purkinje cells and Retinal ganglion cells ., In view of the above results , the only cell types where no differences were detected either in the R values for their branch or termination points were Purkinje cells and Retinal ganglion cells ., However , it should be noted that our data set only contained 15 reconstructions of Purkinje cells ., The results for Purkinje cells could therefore be dependent on the small number of data in that group ., It is important to note that all eight populations in Fig 2 were composed of subgroups with strong differences in their functional role in the nervous system ., Moreover , morphologies within the separate subgroups were partly obtained in different species , preparations and developmental ages ., To illustrate the effect this can have on the analysis , we dissected fly da neurons and TCs into their respective characteristic subgroups ( Fig 3 ) ., Da neurons are known to subdivide into morphologically distinct classes ( I–IV ) and using our statistic R these can be separated into clusters ( Fig 3A ) , corresponding to their specific R values ., In particular , class III da neurons with their large number of small terminal segments ( STSs ) exhibited small R values consistent with the clustering of BPs and TPs due to these STSs ., On the other hand , sub-classes of TCs ( two types of horizontal system cells—HSN and HSE , and three types of vertical system cells—VS2 , VS3 and VS4 ) did not separate into different clusters according to their R values ( Fig 3B ) ., This was not surprising since TCs were previously characterized in detail using morphological models and shown to have similar inner branch rules even though their spanning areas are easy to distinguish 17 ., In all TC classes , TPs were more regularly distributed than BPs and all RTP values were close to 1 , indicating configurations close to complete spatial randomness ., We also tested if the average distance to the NN of BPs and TPs of each individual cell corresponds to that of a uniform random pattern ., The p-values for the tests were computed using the simulations of Poisson point cloud instances with the observed number of points , generated with our MC based approach ., Considering 2D cells , da neurons showed a strong tendency to clustering in both their BPs and TPs: in 97 . 06% and 79 . 41% of the cases we rejected that the distribution of BPs and TPs , respectively , was uniform random in favor of a clustered distribution ., All 3D cell types , except for dentate granule cells , showed rejection of a random distribution in their BPs with high confidence , in favor of a clustered distribution ., In general , only few 2D or 3D cell types hinted to regular BP or TP distributions ( see S1 Table for detailed results of all analyzed cell types ) ., The statistic R for BPs and TPs is therefore a useful measure to distinguish between cell classes and characterize the relationship between dendritic tree structure and input architecture ., However , it remains to be shown that the use of this local statistic in dendritic morphology is not simply an altered version of another traditional branching statistic ., In order to test this and to check whether the input architecture as measured by R is reflected in other branching statistics , we computed the pairwise correlations between R and other commonly used statistics in 3D ( Fig 4A ) and 2D ( Fig 4B ) dendrites ., We found that RBP and RTP do not have strong correlations with other typical branching statistics of dendritic trees in both cases ., Since RBP and RTP were different in distinct cell types and were weakly correlated with other branching statistics , we postulate that these measures are a useful addition to the collection of branching statistics used to classify dendritic morphology ., In order to estimate how clustered or regular distributions of synapse locations co-depend with the clustering of BPs and TPs , we generated morphological models targeting different sets of input points with specified values RInput of their statistic R . This process required a target point cloud generator for specified R values ., To obtain a wide range of sets of target points with specific R values and number of target points , we started with a set of points obtained from a uniform random distribution; we then iteratively moved the input points until the point cloud reached a set target R value ( see Methods and Fig 5 ) ., Dendrites were considered as tree structures connecting these target points 7 , 15 ., We computed synthetic branching structures connecting the target points with minimal resources using the extended minimum spanning tree ( MST ) algorithm 7 ., The MST connecting the target points minimizes at the same time both total cable length and the path length from any point along the tree to the root , using a parameter which weighs the two costs via a balancing factor ( bf; see Methods ) ., Fig 6A shows planar and 3D sample trees obtained from connecting 100 target points in a 200 μm x 200 μm square and 200 μm x 200 μm x 200 μm cube with different values of RInput , using the extended minimum spanning tree for different values of bf ., Tree structures on sets of target points connected with the MST algorithm were previously studied for periglomerular neurons in the olfactory bulb and they were shown to approximate numbers and features of actual synapses 15 ., Studying the relationship between target points and MSTs will therefore be informative about the relationship between synapses and dendritic trees ., We generated morphological models on different sets of 2D and 3D target points with specific values of RInput and found that with higher RInput the trees became denser and the branches were more regularly distributed ., This can be clearly observed in the 3D sample trees of Fig 6 as RInput increases ., As might be expected , more regularly distributed inputs generally resulted in more regular branching structures ( Fig 6B for 3D cases in Fig 6A ) ., Compared with the input point distributions , BPs and TPs describing the dendritic geometry were more regularly distributed in all cases where RInput < 1 and were slightly less regular than the input point distribution in cases where RInput > 1 ., Furthermore , the spatial organization of BPs and TPs was clearly different: in line with reconstructions of real dendritic trees , RTP values were consistently closer to 1 while RBP followed RInput more faithfully ., We obtained similar relations between RInput , RBP , and RTP for different numbers of points ., RBP is therefore most likely a better estimate for the regularity of a neuron’s underlying synaptic input organization ., To show that this was also the case in more detailed morphologies we used a morphological model from layer 5 cortical pyramidal cells 7 and obtained similar results in a realistic range of RInput = 0 . 7 ± 0 . 1 ( Fig 7 ) ., We compared this result with one biological dataset where synapse locations and dendritic reconstructions were available ( Fig 8A ) , specifically for the aforementioned adult-born periglomerular neurons of the olfactory bulb 15 , 29 , 30 ., Interestingly , the results for both actual reconstructions and MSTs constructed on the actual synapse locations were in line with our simulations from this study ., These results indicate that RInput could be estimated using RBP whereas termination points in dendrites and MSTs are more randomly distributed , characterized by a value of RTP close to 1 ., We then compared RInput , RBP , and RTP in two connectome datasets from the fly ( Fig 8B and 8C ) 31 , 32 ., Using serial electron microscopy it has increasingly become possible to obtain data of morphological reconstructions and the exact position of synapses between the same cells ., Interestingly , we found the predictions from our model to be validated by these two datasets with RBP and RInput values being more similar and RTP being closer to 1 ., Overall , we believe that constructing synthetic versions of real dendritic trees with more detailed morphological models would be useful for inferring the underlying spatial organization of the synaptic inputs ., This will provide invaluable predictions for connectome and large-scale neural circuit analyses ., Apart from the relations between RInput , RBP , and RTP , it is interesting to study the relation of RInput with branching statistics typically used to characterize dendritic trees ( Fig 9 ) ., As was the case for RBP and RTP in real dendritic tree reconstructions , RInput was weakly correlated with other branching statistics , suggesting that input architecture is not well captured by traditional branching statistics whereas RBP and RTP would be useful measures for this and to classify dendritic morphology accordingly ., However , both total length and number of branch points increased reliably with RInput , requiring the minimum spanning tree to use more cable to connect the points that are more widely spread and more branches to reach out to all distributed inputs in space ., This correlation clearly affected the scaling behavior that was previously observed between number of inputs and total length as well as between number of inputs and number of branch points 15 ., Here , the previously reported 2/3 power between these measures was not affected by RInput , but a clear increase in total length was observed as an offset in the relationship ( Fig 10 ) ., MST-based dendrites connecting target points with an increased RInput required much more cable length ., We have presented here a new branching statistic for dendrites , the regularity index R , which is based on the average nearest neighbor ( NN ) distance between branch , termination or input points of a given dendritic tree , capturing the regularity of their respective distributions ., Specifically , R is defined as the ratio of the observed average NN distance to the one expected in a matching random point cloud ., This makes R independent of the absolute scale of the dendritic arbor , but rather captures the clustering characteristic of the branch and termination points and allows for comparison of cells of different sizes ., We found that the measure allowed to distinguish dendritic trees from different cell classes for which the local statistics of the spatial input distribution differed ., The values of R computed for the sets of branch points ( RBP ) and termination points ( RTP ) of reconstructions of real dendritic trees correlated little with most other commonly considered branching statistics , indicating that these measures provide new descriptive power for dendritic trees that was not captured by existing measures ., Using morphological models , we then found that , in the range of observed RBP and RTP values in real cells , overall RBP values were good predictors of the input distribution ( RInput ) while RTP values were generally closer to 1 , implying more randomly distributed termination points in dendritic trees ., We also showed that more regular input distributions with higher values of RInput showed increased dendritic total length in the model ., An analogous increase in total length was observed in real dendrites with increasing RBP ., Input targets can loosely be compared to synaptic inputs for certain cell types such as the periglomerular neurons of the olfactory bulb in Fig 8A 15 ., While this indicates a specific link between the input organization and dendritic morphology it by no means implies a causal relation between input locations and resulting dendrite growth ., Rather , dendrites and axons organize to implement possible connectivity patterns between neurons and specific dendritic morphology could constrain the space of potential synapses ., However , the precise biological processes are not yet known ., On the one hand , the correlation between the synapse distribution and RBP can be a direct consequence of optimal wiring since minimum spanning trees shows the same correlation as real dendrites ., But it is not known how the biological growth process leads to such optimal wiring ., On the other hand , the precise distribution of branches within the range of optimally wired dendrites can vary substantially and the details of the growth process can lead to these differences in morphology ., Increased self-avoidance of branches during dendrite development through molecular mechanisms such as Dscam would for example most likely lead to larger values of R 33 ., Increased localized branching due to some accumulation of molecular cues could in the other extreme lead to clustered branches in mature dendrites leading to R values lower than, 1 . Overall , we expect the proposed measure R to be able to better predict these type of local features of the input organization of given dendritic tree types compared to other existing branching statistics ., As more realistic morphological models based on minimum spanning trees become available , as is the case for example for TCs 17 and dentate gyrus granule cells 18 , this information can be further refined ., The expected NN distance of a uniform random distribution used for calculating R has traditionally been obtained analytically , assuming infinitely many points in an unbounded volume ., Yet , in practice all of our volumes V containing a given point cloud C were finite and bounded resulting in strong biases and edge effects if R is estimated in a naïve way ( see Methods ) ., Our approach to remedy these adverse effects was to use Monte Carlo ( MC ) simulations to predict the expected NN distances of uniform distributions numerically , using instances of Poisson point clouds ., This procedure furthermore allowed for the specification of confidence intervals for the estimated values of R ( S2–S4 Figs ) ., We found that confidence intervals were mostly dependent on the number of points in the point cloud ( decreasing as the number of points increases ) and only to a lesser extent on the shape of its supporting volume ., For small numbers of points ( N < 20 ) , the confidence intervals were large , making the estimated values of R less reliable ( S3 Fig ) ., Our MC based approach will be useful in further studies beyond the scope of dendritic morphology since point processes in any type of finite volumes will necessarily exhibit similar important boundary effects ., There are several ways in which the measure R could be generalized ., First of all , for simplicity we assumed point clouds with uniform homogeneous densities when computing R . This can be extended to non-homogenous cases by studying local estimates of point densities ., This would lead to a localized version of measure R . Secondly , we only considered one nearest neighbor per point ., This allows for point clouds with different global clustering behavior to get assigned the same value of R ( e . g . a point cloud where all points have the same coordinates and a point cloud where pairs of points have the same coordinates both get assigned a value of R = 0 ) ., But the statistic R can easily be extended to consider neighborhoods of higher order , containing the k nearest neighbors for each point , k ≥ 2 . Furthermore , other quantities commonly used for the analysis of spatial point patterns such as the cumulative distribution function of the nearest-neighbor distances , known as G function , Ripley’s K function 34 and related quantities can be used ., These extensions are subject of future studies ., If neighbor distances are used to define simplicial complexes on a given point cloud , and the resulting complexes are examined using methods from algebraic topology , this leads to techniques used in topological data analysis 35 ., Related techniques most recently also were proposed as methods to examine dendritic structures 36 ., Overall , we presented a new statistic , the regularity index R , for dendrites that allows to relate the morphology of a neuron with the specific connectivity that it implements ., It has low correlation with most commonly used statistics of dendritic branching and is extendable in several ways , providing a useful new statistic for the classification of dendritic trees ., The average nearest neighbor ( NN ) ratio R=r¯0/r¯E compares the observed average NN distance r¯0 between a set of N points with the expected average distance r¯E between nearest neighbors under the assumption of a uniform random distribution ( with the same number of points covering the same total area or volume ) ., This approach was first described in 20 , 37 ., R provides a measure of the clustering of the points in a point cloud C . Concretely , the closer the points are to a random ( Poisson ) distribution , the closer to 1 the value of R becomes ( as the values of r¯0 and r¯E are more similar ) ., Values of R less than 1 correspond to clustering ( r¯0<r¯E ) ., When all points overlap ( R = 0 ) the most clustered condition is reached ., For values of R greater than 1 , nearest neighbors are further apart than it would be expected for a random distribution ( r¯0>r¯E ) ., In 2D arrangements , the most dispersed situation is the one in which the points are spaced on a triangular lattice , yielding a value of R = 2 . 1491 20 ., The measure R has the advantage of being easily interpretable ., For example , R = 0 . 5 indicates that nearest neighbors are , on average , half as distant as expected under random conditions ( c . f . Fig 5D ) ., Note though that R only considers pairwise NN distances and e . g . cannot distinguish the case in which all pairwise NN distances are 0 ( pairs of points have the same coordinates ) from a fully clustered situation ( all points of the point cloud have the same coordinates ) ., Formally , for a finite point cloud C , i . e . , a set of N points , the average NN distance is, r¯0=1N∑i=1i≠jNmin{di , j} ,, where di , j denotes the Euclidean distance between the i-th and the j-th point in C . This is the numerator in the definition of R=r¯0/r¯E ., The denominator in R is the expected NN distance r¯E for a Poisson process that can be analytically computed as r¯E=1/2ρ in the 2D case and as r¯E=Γ ( 4/3 ) /4πρ/33 in the 3D case , where Γ ( ∙ ) is the gamma function and ρ is the point density , i . e . , the mean number of points per unit area or volume V . For a uniform random distribution , an unbiased estimator of ρ is ρ^=N/V ., Thus , to obtain the point density , an accurate estimate of the supporting volume V of the point cloud C is required ., In order to estimate R , a volume V supporting a given point cloud C needs to be estimated ., The most common way to do this is to use the convex hull of C . Yet , with this choice the supporting volume is overestimated if it is non-convex , which results in incorrect values of R ( see for examples S1 Fig ) ., Better estimates of R were obtained using α-shapes ., α-shapes were devised to characterize the shapes of point clouds and can be seen as an extension to the notion of a convex hull 38 , 39 ., Formally , to any given finite point cloud C in 2D or 3D Euclidean space a one parameter family of curves or surfaces Sα called α-shapes can be constructed , with α ϵ 0 , ∞ ., By construction , S∞ corresponds to the convex hull and S0 to the point cloud itself ., For any finite C , Sα is a finite set and a smallest value α0 exists ( called critical value of α ) such that Sα0 is connected and contains all points of C . Furthermore a smallest value αk < ∞ exists for which Sαk=S∞ ., The α-spectrum of C is defined as the monotonically increasing , finite sequence of values ( αi ) 0≤i≤k , 0 ≤ αi < ∞ , αi < αi+1 for which each Sαi is a distinct α-shape and the shapes do not change between two consecutive values αi , αi+1 ., To compute what we call a “tight hull” around a point cloud C we selected the center point αk/2 of the α-spectrum , for which we rounded the index k/2 to the next integer value ., Especially for point clouds with non-convex supporting volumes , this yielded much better estimates of the true volume and thus less biased estimates of R . See S1B and S1C Fig for an example of the convex hull compared to the tight hull of point clouds with non-convex support , and the resulting values of R . After estimating a supporting volume V using α-shapes as described in the previous section , naïve calculations of R still yielded biased results due to boundary effects ., This is due to points in C for which the distance to the boundary of V was smaller than the average NN distance r¯0 in C . As the calculation of r¯E assumes that balls of radius r¯E surrounding all points are always completely contained in V , R was overestimated ., Two proposed techniques correcting for such boundary induced biases are the so-called toroidal edge correction and the border area edge correction ., The first one removes boundaries by transforming a ( bounded ) rectangular study region into a torus by identifying opposing edges ., The second one specifies a buffer zone around the boundary of the study region and uses the part remaining in the middle as the new study region ., Points in the buffer zone are used only to take measurements of points ( NN distances in our case ) that are within the new study region and are further discarded ., Yet , both techniques have their drawbacks: the toroidal edge correction cannot be used for non-rectangular regions , as is the case of dendrites , and the border area edge correction discards a large number of available points , which makes it inappropriate for many dendrites for which the number of points was not large enough ., Moreover , analytical bias corrections were also derived 28 , but those require convex planar surface areas as supporting volumes ., Since most dendritic arbors form non-convex areas and volumes ( S1A Fig ) and many of them do not have a high number of BPs and TPs , instead of computing r¯E analytically from an estimate of the point density and using an edge correction technique , we used a Monte Carlo ( MC ) simulation approach to estimate r¯E ., For a point cloud C consisting of N points contained in a volume V , we first computed r¯0 as the observed mean NN distance in C . We then sampled M = 100 uniform random point clouds within V , each containing N points ., Importantly , we scaled each sampled point cloud so that its supporting volume ( again computed using α-shapes ) matched V . When this process , which we call volume correction , was not performed the estimates of R values were positively biased , especially for small point clouds ( S2 Fig ) ., For each of those point clouds we then computed the average NN distance , r¯Ei , i=1 , … , M , and obtained an estimate of r¯E as the mean of the r¯Ei , i=1 , … , M , values of the simulations ., No edge corrections were necessary since all the average NN distances were biased by the same edge effects ., To check the correctness and convergence properties of this approach , we generated point clouds with known R values and compared them to the R values estimated from our MC based method ( S4 Fig ) ., In each MC iteration we additionally estimated confidence intervals ci− , ci+ for r¯Ei , i=1 , … , M with confidence level 1 − α using 1 , 000 bootstrap samples ., We then obtained the corresponding confidence intervals c− , c+ for r¯E by computing the sample mean of the set of ci− and ci+ , i=1 , … , M , to obtain c− and c+ , respectively ., The confidence interval c− , c+ for r¯E in turn leads to the statistic R as r¯0c+ , r¯0c− ., Throughout this study we used α = 0 . 05 ., We observed that the confidence intervals were mainly influenced by the number N of points in the point cloud and to a much lesser extent by the shape of the supporting volume V . We assessed this by computing confidence intervals both for planar sample configurations with known R value ( S2 Fig ) in a square area and for the 3D cell classes considered in this work ( S3 Fig ) ., To evaluate the measure R on real cells , we obtained a number of reconstructions of dendritic trees from NeuroMorpho . org 40 , Version 7 . 0 ( released on 09/01/2016 ) using the TREES toolbox 7 ., Specifically , we chose reconstructions belonging to eight well-known cell classes for our investigations , namely cortical pyramidal cells , hippocampal pyramidal cells , dentate granule cells , motoneurons , retinal ganglion cells , cerebellar Purkinje cells , fly larva dendritic arborization ( da ) neurons and fly Lobula Plate tangential cells ( TCs ) ., The first four classes were 3D cells and the last four classes were 2D ., For selecting the reconstructions , we obtained all reconstructions from NeuroMorpho . org that were classified as either having moderate or complete reconstructions of their dendritic trees and belonged to the control group ( to exclude mutant cells ) ., We then grouped all reconstructions by archive and sorted out archives that contained poor reconstructions by manual visual inspection a | Introduction, Results, Discussion, Methods | Neurons collect their inputs from other neurons by sending out arborized dendritic structures ., However , the relationship between the shape of dendrites and the precise organization of synaptic inputs in the neural tissue remains unclear ., Inputs could be distributed in tight clusters , entirely randomly or else in a regular grid-like manner ., Here , we analyze dendritic branching structures using a regularity index R , based on average nearest neighbor distances between branch and termination points , characterizing their spatial distribution ., We find that the distributions of these points depend strongly on cell types , indicating possible fundamental differences in synaptic input organization ., Moreover , R is independent of cell size and we find that it is only weakly correlated with other branching statistics , suggesting that it might reflect features of dendritic morphology that are not captured by commonly studied branching statistics ., We then use morphological models based on optimal wiring principles to study the relation between input distributions and dendritic branching structures ., Using our models , we find that branch point distributions correlate more closely with the input distributions while termination points in dendrites are generally spread out more randomly with a close to uniform distribution ., We validate these model predictions with connectome data ., Finally , we find that in spatial input distributions with increasing regularity , characteristic scaling relationships between branching features are altered significantly ., In summary , we conclude that local statistics of input distributions and dendrite morphology depend on each other leading to potentially cell type specific branching features . | Dendritic tree structures of nerve cells are built to optimally collect inputs from other cells in the circuit ., By looking at how regularly the branch and termination points of dendrites are distributed , we find characteristic differences between cell types that correlate little with other traditional branching statistics and affect their scaling properties ., Using computational models based on optimal wiring principles , we then show that termination points of dendrites generally spread more randomly than the inputs that they receive while branch points follow more closely the underlying input organization ., Existing connectome data validate these predictions indicating the importance of our findings for large scale neural circuit analysis . | medicine and health sciences, dendritic structure, nervous system, statistics, electrophysiology, neuroscience, mathematics, ganglion cells, brain mapping, neuronal dendrites, research and analysis methods, statistical distributions, animal cells, mathematical and statistical techniques, statistical methods, monte carlo method, probability theory, connectomics, cellular neuroscience, neuroanatomy, cell biology, anatomy, synapses, pyramidal cells, physiology, neurons, biology and life sciences, cellular types, physical sciences, neurophysiology | null |
journal.pgen.1002256 | 2,011 | Genetic Effects at Pleiotropic Loci Are Context-Dependent with Consequences for the Maintenance of Genetic Variation in Populations | Metabolic syndrome ( MetS ) is an array of co-occurring disorders including dyslipidemia , high blood pressure , impaired glucose tolerance , and obesity ., Individuals diagnosed with MetS have increased risk of developing cardiovascular disease ( CVD ) and type-2 diabetes ( T2D ) 1 ., MetS prevalence currently exceeds 20% in the United States and is increasing in developing countries 2 ., This increase is hypothesized to be the result of over-consumption of high-caloric foods in conjunction with sedentary lifestyles 3 ., There is also a genetic component as individual responses to dietary environment and to lifestyle modifications vary 1 , 4 ., Understanding MetS etiology is challenging because phenotypic variation is caused by complex interactions of many genes of small effects , by environmental factors , and by gene-by-environment interactions 5–9 ., Thus animal models are valuable because genetic and environmental influences can be controlled for and monitored in populations of known genetic structure 10 ., Mouse models have made major contributions to our understanding of complex disease etiology , including hypertension , obesity , and T2D 11–15 ., However , MetS per se is not well defined in mice because the physiological features of individual components vary between mice and humans , reflecting 65–85 million years of divergent evolution 16–19 ., Nevertheless , mouse models have increased our understanding of the pathophysiology of metabolic disorders and genes with robust effects have been identified using both spontaneous ( e . g . ob/ob mice ) and transgenic models ., Further understanding of MetS will come from interrogating genes with small allelic effects on physiological processes , the source of variation in complex traits relevant to evolution and biomedicine , rather than from single genes with large defects ., We present results of a study of loci associated with normal variation in multiple MetS components: obesity ( fatpad and organ weights ) , serum lipid levels ( cholesterol , triglycerides and free-fatty acids levels ) , and diabetes ( serum insulin and glucose levels , and response to a glucose challenge ) in an F16 generation of an Advanced Intercross Line ( AIL ) formed from the LG/J and SM/J inbred mouse strains ( Wustl:LG , SM-G16 ) ., Variation in complex traits in LG/J x SM/J is due to many genes of small effect interacting with each other and with the environment ., Quantitative trait loci ( QTL ) have previously been mapped for obesity , serum chemistries and growth-related phenotypes in crosses of these strains 20–25 ., This study is the first to look at variation in multiple MetS components mapping to the same locus in a very advanced generation of the LG/J x SM/J AIL ., Here we examine these MetS QTL under a systems biology framework , incorporating both biomedical and evolutionary perspectives ., We report additive and dominance genotypic effects in addition to parent-of-origin genomic imprinting effects ., Parent-of-origin imprinting is defined as the unequal expression of maternally and paternally derived copies of an allele , and has been shown to affect variation in metabolic traits 26– ., We examine the context-dependency of these genetic effects – additive , dominance and imprinting – by examining response to high- and low-fat dietary treatments ., Context-dependency , defined as genotype-by-environment and genotype-by-sex interactions 29 is a proposed mechanism by which genetic variation is maintained in populations 30–33 ., We examine whether additive genotypic values for a given trait or trait combination change rank across different environments , which is consistent with a so-called ecological cross-over 34 ., When different alleles are favored in different environments , selection can maintain genetic variation at the locus ., Another mechanism that can maintain genetic variation in a population is balancing selection at pleiotropic loci , those associated with variation in multiple phenotypes , with different dominance relations for the different traits , so-called differential dominance 22 , 35 ., When differential dominance is present , some linear combination of traits will display over- or underdominance , even when no single trait does ., If directional selection occurs along these linear combinations , there is balancing selection on the locus and genetic variation will be maintained ., We examine context-dependent genetic effects and differential dominance at pleiotropic loci associated with MetS components ., Patterns of pleiotropy are thought to reflect functional and developmental relationships among traits 36 , and have been hypothesized to serve as potential constraints on adaptive evolution 37 as well as underlie correlated phenotypic responses to selection 38 ., Although pleiotropy has long been proposed to be ubiquitous , few studies have measured enough traits in a focal population to analyze this aspect of genetic architecture 39 ., Our results show that additive , dominance and parent-of-origin genomic imprinting genetic effects vary among diet , sex and diet-by-sex environments among metabolic traits mapping to the same locus ., This indicates that context-dependency is an important aspect of pleiotropic connections among components of MetS , a result supported by recent work on the foraging gene in Drosophila melanogaster 40 , 41 ., Understanding these connections and their evolutionary implications is important for understanding disease etiology and is relevant to personalized medicine ., We identify 23 pleiotropic QTL associated with normal variation in two or more MetS components ., Of these 23 loci , 12 pass genome-wide significance while 11 pass chromosome-wise significance ., The average locus is associated with variation in 4 traits ., The traits examined here show moderate to high genetic correlations among each other and are reported with their respective heritabilities in Ehrich et al . 2005 22 ., Fourteen loci ( 61% ) are associated with both diabetes ( glucose levels , glucose tolerance , and serum insulin ) and obesity ( fatpad and organ weights ) ., Six loci ( 26% ) are associated with both dyslipidemia ( serum cholesterol , free-fatty acid and triglycerides levels ) and obesity ., Three loci ( 13% ) are associated with adiposity ( fatpad weight ) and liver weight ., Liver weight is moderately correlated with percent liver fat ( r\u200a=\u200a0 . 61 ) 42 , and nonalcoholic fatty liver disease is strongly associated with MetS 43 , 44 ., Additive effects are found at 20 loci ( 87% ) , and dominance and imprinting effects are found at 21 loci ( 91% ) ., On average , in cohorts showing additive effects , LL homozygotes have higher serum lipid levels ( cholesterol , triglyceride , free-fatty acid ) and heavier weights ( fatpad and/or organ weights ) but respond better to a glucose challenge ( intra-peritoneal glucose tolerance test ) than SS homozygotes ., In cohorts showing dominance effects , the L allele is dominant to the S allele 52% of the time ., In cohorts with dominance effects and no additivity , we find overdominance ( heterozygotes have significantly higher genotypic values ) 60% of the time and underdominance ( heterozygotes have significantly lower genotypic values ) 40% of the time ., In cohorts showing parent-of-origin imprinting effects , 15% show maternal expression imprinting , 10% show paternal expression imprinting , 21% show polar dominance imprinting ( no additive effects ) , and 54% show bipolar dominance imprinting ( no additive or dominance effects ) ( Table S1 ) ., Description of the various parent-of-origin imprinting patterns is found in Wolf et al . ( 2008 ) 45 ., High-fat fed males are the most commonly affected cohort for the organ weights and diabetes-related traits , and high-fat fed females are the most commonly affected cohort for the serum lipid levels and fatpad weights 24 , 25 , 46 ., Table S1 breaks down the context-dependency of the QTL reported here and lists candidate genes found in the intervals ., The mean QTL support interval is ≈4 . 0 Mb and contains 39 genes , many previously associated with metabolic disorders ., Some of these positional candidates show expression differences between LG/J and SM/J in liver and white-fat tissues ( Tables S1 , S2 and S3 ) , and we have annotated SNPs between the two strains in both coding and noncoding DNA in these intervals ( Table S4 ) ., For example , we find a highly significant QTL on chromosome 1 , DMetS1b , associated with variation in both serum lipid levels and obesity ., This region overlaps QTL previously associated with high-density lipoprotein cholesterol ( HDL ) levels in studies using multiple crosses of mouse 15 , 25 , 47 , 48 ., Additionally , this region was recently reported as associated with both cholesterol and free-fatty acid levels in LG/J x SM/J 25 ., The current analysis reveals this region is also associated with variation in gonadal and total fat-pad weights ., The genotypic effects at this QTL are complex ( Figure 1a–1d ) ., For cholesterol , there is an additive effect in the full population whereby individuals homozygous for the L allele have higher cholesterol ., For free-fatty acid levels , in addition to this additive effect , high-fat fed females have maternal expression imprinting and low-fat fed females have paternal expression imprinting ., High-fat fed males have polar dominance imprinting and low-fat fed males have underdominance effects with no significant additive or imprinting effects ., For gonadal fatpad weight , high-fat fed females have bipolar dominance imprinting ., For total fatpad weight , high-fat fed females have bipolar dominance imprinting and high-fat fed males have an additive effect ., This QTL spans 2 . 2Mb and contains 47 genes , 10 of which are candidates previously associated with metabolic disorders ., Expression analysis of genes in this QTL show that in white-fat , 43 of these 47 genes are expressed in LG/J and SM/J , and 9 ( 21% ) are significantly differentially expressed between the two strains ., Three of these 9 genes , F11r , Fcgr2b and Nr1i3 , are associated with variation in MetS components ( Figure S1a-S1c and Table S2 ) 49–51 ., In liver , 39 genes in the interval are expressed in LG/J and SM/J , and 10 ( 26% ) of these genes are significantly differentially expressed between the two strains ., Five of these 10 genes , Apoa2 , F11r , Hsd17b7 , Nr1i3 , and Usf1 , are MetS candidates ( Figure S2a-S2e and Table S3 ) 47 , 49–53 ., There are 4 , 933 SNPs between LG/J and SM/J in DMetS1b ( Table S4 ) ., Twenty–four of these SNPs are non-synonymous , and two of these non-synonymous SNPs ( rs8258232 and rs8258226 ) fall in Apoa2 ., One of these SNPs , rs8258226 , is the location of a mutation previously found to affect HDL cholesterol levels in multiple strains of mice 47 ., The same Ala61 -to- Val61 substitution first identified by Wang et al . ( 2004 ) as the potential causal change underlying HDL variation is the same substitution found in LG/J ., Many other DMetS1b SNPs , both in and around MetS candidates , fall within noncoding DNA having high regulatory potential 54 ., Differential dominance is a property of pleiotropy that occurs when different traits mapping to the locus vary in the magnitude of their dominance ratios ( d/a ) ., Because the dominance ratios vary , the additive and dominance vectors are not colinear and some combination of traits will display over- or underdominance 35 , 55 , 56 ., An example of differential dominance is found at a QTL on chromosome 6 , DMetS6c ., This locus is associated with variation in diabetes-related traits and liver weight ( Table S1 ) ., The dominance ratios at DmetS6c differ between glucose levels in low-fat fed females ( d/a = −0 . 9 ) and insulin levels in high-fat fed males ( d/a = 1 . 2 ) ., These two traits also display antagonistic pleiotropy , where glucose in the low-fat fed females has a significant positive additive genotypic value ( LL homozygotes have higher levels ) and insulin in the high-fat fed males has a significant negative additive genotypic value ( SS homozygotes have higher levels ) ( Figure 2a-2b ) ., Six loci ( 26%; DMets2c , DMetS6b , DMetS6c , DMetS7c , DMetS10b , DMetS16a ) show differential dominance ( Table S1 ) ., Statistically significant interactions are of two types: ‘spreading’ , where there is no difference between the genotypes in one sex and/or environment but a substantial difference in the alternate sex and/or environment , and ‘crossing’ , where the rank order of allelic effects changes between sexes and/or environments 57 ., Only crossing interactions can act to maintain allelic variation at a locus ., We find 4 loci ( 17% of loci ) having traits ( 6% of traits mapping to these loci ) showing significant crossing interactions ( DMetS6c , DMetS8b , DMetS15a , DMetS16a , Figure 3 , Tables S1 and S7 ) , indicating that with a few exceptions , the rank order of the homozygote genotypes remains the same in multiple environments for individual traits mapping to these loci ., Three of the four crossing interactions occur in diabetes-related traits ( glucose and insulin levels and area under the curve at 10wks ) , which have relatively lower genetic correlations among the traits mapped here 22 ., This supports theoretical predictions that crossing genotype-by-environmental interactions would give rise to lower genetic correlations among traits 38 ., Pleiotropic effects underlie genetic correlations among traits ., Variation in pleiotropy is essential for selection to shape patterns of phenotypic covariation 58 ., We identify 23 pleiotropic loci affecting normal variation in multiple components of MetS , including obesity- , T2D- , and CVD-related traits ., We find additive , dominance and parent-of-origin imprinting effects are equally prevalent , highlighting the complex genetic architecture underlying these common/ complex traits that together characterize a syndrome ., This result has implications for human genome-wide association studies ( GWAS ) , which generally assume just additivity despite growing evidence for non-additive genetic effects on complex traits 59 , 60 ., Indeed a recent study identified parent-of-origin effects associated with T2D in the Icelandic genealogy database , finding the effects are different between males and females 61 and illustrating the complex connections among genomic sequence , genetic effects , environment , and disease risk ., Identifying these connections is a challenge in human population studies because recording or controlling an individuals environment over time is not possible , although some studies have successfully examined gene-by-environment interactions 5 , 62–64 ., We show that genotype interacts with environment in significant ways , and these interactions are not always consistent among genotypes , across environments , or across traits within the same population ., This is seen clearly at DMetS1b discussed above 25 , 47 , 65 ., While multiple traits individually map to this locus ( cholesterol , free-fatty acids , gonadal fatpad and total fatpad weight ) , the underlying genetic effects vary among the traits and are highly context dependent ( Figure 1a–1d ) ., We find the additivity is consistent among cohorts showing significant additive effects ., However , when considering the heterozygotes , we find complex interactions between the L and S alleles in the different patterns of imprinting and dominance ., For free-fatty acids , the females have opposite parent-of-origin imprinting effects depending on whether they were fed a high-fat ( maternal expression imprinting ) or low-fat ( paternal expression imprinting ) diet ., If females were pooled together for analysis without considering dietary environment , no parent-of-origin effects would be detected because the two sex-by-diet cohorts effects negate each other when combined ., Further , for high-fat fed females , paternal inheritance of the L allele is protective for free-fatty acids , gonadal fatpad weight , and total fatpad ., For gonadal fatpad and total fatpad weight , maternal inheritance of the L allele results in higher weight ., Thus the same allele , in the same cohort , confers both protection and risk depending on parent-of-origin ., We find differential dominance among traits at some of these loci , for example at DMets6c discussed above ., We acknowledge that our QTL contain multiple genes that may be tightly linked and may individually influence each trait mapped to the locus 66 ., While our results are consistent with differential dominance occurring at some QTL , we do not have the resolution to test if this is due to multiple tightly linked genes ., However , Keightley and Kascer showed that differential dominance is expected in systems in which nonlinearities are present , for example in the saturation and feedback inhibition systems of metabolic networks 67 ., In this situation , some combination of traits will display under- or overdominance at the same locus , even in the absence of trait-specific under- or overdominance , a phenomenon called multivariate overdominance 55 ., Directional selection on the synthetic phenotype could result in balancing selection at the locus and the maintenance of genetic variability 33 ., At DMetS6c , directional selection for fitness could favor a genotype associated with lower levels of both insulin and glucose levels ( Figure 4 ) ., It is tempting to speculate that maintenance of variability through interactions between alleles at a locus associated with metabolic traits could facilitate short-term adaptations to rapidly changing environments ., While this issue is outside the scope of the current study , the results presented here can inform testing of evolutionary hypotheses such as The Thrifty Gene Hypothesis 68 , under a systems biology framework , in an attempt to understand the determinants of the increasing prevalence of MetS and other disorders affecting metabolic homeostasis ., Genetic variation can also be maintained when the rank order of homozygous genotypes changes between environments ., While most loci examined here do not show significant crossing interactions within the range of the experiment , we do see a few loci consistent with an ecological cross-over between sex and dietary environments for single traits ( Figure 3 ) ., It is important to note , however , that an interaction between sexes is different than an interaction across environments ., The evolutionary outcome of an ecological cross-over will depend on the frequency with which each environment is experienced ., For sex it is generally ≈50∶50 ., Thus even in the absence of differential selection between the sexes , allelic variation will be maintained in a crossing scenario ., For environmental interactions such as diet , allelic variation will be a function of the relative frequency of the environment experienced ., Empirical evidence supporting the theory that heterogeneous environments produce crossing interactions is inconsistent 69 , 70 ., In this study , the genotypes for most of our single traits at these loci differ in magnitude between environments without significant crossing , a so-called spreading interaction ., However , the QTL examined here are pleiotropic and multivariate combinations of traits may exhibit rank order changes in heterogeneous environments ., Consider , for example , the antagonistic pleiotropy and hence rank order change of homozygous genotypes seen between glucose levels in low-fat fed females and insulin levels in high-fat fed males at DMetS6c ( Figure 2a–2b ) ., Such a multidimensional synthetic interaction is consistent with the complex nonlinear connection of the traits comprising the MetS , and remains an open question for further exploration 14 ., Once a genomic association is made , examination of the QTL can lead to identification of the quantitative trait gene ( QTG ) and eventually the quantitative trait nucleotide ( QTN ) affecting variation in the trait ., We acknowledge that our QTL may contain multiple QTG and QTN , even when we fail to reject pleiotropy ., We have identified candidate QTGs in our QTL regions , both from the literature and by examining differential expression between LG/J and SM/J in relevant tissues ., We further identified QTN for experimental follow-up by examining SNPs between the strains both in and around QTGs ., We present many fruitful regions for follow-up , including some novel positional candidate genes , for example Cacna7a located in DMetS8b , which is associated with normal variation in obesity and cholesterol levels ., This gene encodes the pre-forming A1A subunit of voltage-gated calcium channels and has been found to influence the functionality of cholesterol-rich microdomains 71 ., It is differentially expressed between LG/J and SM/J in both liver and white-fat and contains many SNPs in non-coding flanking and intronic regions having high-regulatory potential ., This gene is associated with chronodisruption , the desynchronization of circadian rhythms 72 , and with migraine headaches 73 ., Recent research demonstrates an association among chronodisruption , migraine , and MetS components 74–76 ., Another attractive locus for follow-up is the Apoa2 gene , which falls in DMetS1b ., Variations in the homologous human APOA2 sequence have been well studied for association with MetS components in humans 77 , 78 ., We not only find Apoa2 is differentially expressed between LG/J and SM/J , but also identify a non-synonomous coding SNP ( rs8258226 ) that has been independently associated with elevated cholesterol levels in multiple other strains of mice 47 ., While this result is encouraging proof-of-principle in going from QTL → QTG → QTN , further experimentation is required to know if this mutation , let along this gene , is associated with variation in the other traits mapping to this locus ., Indeed it has been found that , when a single locus is associated with multiple traits , different polymorphisms within the locus are independently associated with the various traits 29 , 79 ., So while the QTL is pleiotropic , the pleiotropy breaks down at the nucleotide level ., Overall , we find the genetic effects at these 23 QTL are highly context-dependent and are not consistent among the individual traits mapped ., Our results indicate that if context such as sex and/or diet are not considered , not only can genetic signals in specific cohorts be masked and/or cancelled out in an aggregate study population , but also genetic effects can be erroneously assigned to specific cohorts within a population if the effects are pooled over all its members ., We have shown that the genetic architecture underlying the individual traits mapping to these QTL is complicated , and so are the relationships among the traits themselves ., Further , some patterns are consistent with evolutionary theory with respect to the maintenance of genetic variation in populations , even when specific variants are deleterious in particular environments or in particular combinations ., While over- or underdominance and crossing interactions for single phenotypes may not be common , multidimensional synthetic phenotypes at QTL with pleiotropic effects can produce situations that favor the maintenance of genetic variation in populations ., As Lewontin ( 80; p318 ) noted , “Context and interaction are of the essence” ., Gluckman et al . 81 recently discussed the challenges associated with understanding human biology in light of the current epidemic of metabolic disorders , and Sing et al . 6 proposed a series of steps a researcher should take to address issues of complex disease etiology ., As the era of personalized medicine and individual whole-genome sequencing looms , it is important to keep in mind the ultimate goal of developing treatments and prevention strategies for individuals ., For MetS , this goal may be attained through understanding the underlying genetic architecture of its disease components , of how these components relate to each other evolutionarily , and in what context ., Mouse models may be especially appropriate for bridging the divide between evolutionary and biomedical research because they allow the study of the effects of natural alleles on normal variation , and human-mouse homology is well defined ., Our results are important because they can be used to elucidate gene-by-environmental effects that could inform large-scale genomic study design in humans ., Our study involved mice and all animal care and handling procedures conformed to IACUC guidelines ., The LG/J x SM/J Advanced Intercross Line ( AIL ) is managed as a pseudo-randomly mated line starting from the F2 generation ., The LG/J strain originated from a selection experiment for large body size at 60 days and the SM/J strain originated from a selection experiment for small body size at 60 days 82 ., Animals from each strain have been inbred by brother-sister mating for over 150 generations making them genetically homozygous with the exception of spontaneous mutations and the agouti locus in SM/J which is maintained heterozygous ( a/Aw ) for breeding purposes 83 , 84 ., The AIL was generated from an initial cross of 10 male SM/J mice and 10 female LG/J mice ., Animals are randomly mated but brother-sister mating is not allowed ., Only one male and one female are chosen from each family as breeders for the next generation , thereby eliminating variation in familial contributions to the next generation ., This is an effective method of reducing inbreeding and doubling the effective population size of a colony relative to its census size 85 ., The average number of breeding pairs in the AIL is 75 , giving a census size of 150 and an effective population size of approximately 300 individuals ., This study used an experimental F16 population of 1 , 002 animals in 76 sibships , each averaging 6 . 8 animals ., Animal husbandry details can be found in Ehrich et al . 2005 86 ., At weaning , males and females from each litter were partitioned into cohorts fed either high-fat ( 253 males; 248 females ) or low-fat ( 247 males; 254 females ) diets ., The diets were isocaloric with the exception of calories from fat ( Harlan Teklad cat . No . TD88137 , 42% energy from fat; and Research Diets cat . No . D12284 , 15% energy from fat , specially formulated; Table 1 ) ., Animals were weighed weekly for 20 weeks ., A subset of animals ( 217 females , 113 fed the low-fat diet and 104 fed the high-fat diet; 213 males , 103 fed the low-fat diet and 110 fed the high-fat diet ) were subject to an intra-peritoneal glucose tolerance test ( IPGTT ) at 10 and 20 weeks of age as described in Ehrich et al . 2005 86 ., Readings taken over the course of 2 hours were used to calculate the area under the curve ( AUC ) , a measure of glucose tolerance ., Animals were necropsied at 20 weeks of age ( also described in Ehrich et al . 2005 ) and fasting ( 4 hr ) serum cholesterol , free-fatty acid , triglyceride , glucose , and insulin were obtained from blood via cardiac puncture ., Serum was frozen at −20°C until assayed by the Nutrition Obesity Research Center – Animal Model Research Core at Washington University ., Additionally , fat pads ( inguinal , mesenteric , renal and gonadal ) and internal organs ( heart , kidneys , liver , and spleen ) were removed and weighed ., DNA was extracted from liver tissue using the QIAGEN kit ., 1 , 536 single nucleotide polymorphisms ( SNPs ) were chosen from the CTC/Oxford SNP survey ( www . well . ox . ac . uk/mouse/INBREDS/ ) and scored with the Illumina Golden Gate Bead Array ., Genotyping was performed at the Washington University Genome Sequencing and Analysis Center ., 1 , 402 autosomal SNPs were reliably scored and used in this study ( Table S5 ) ., A genetic map was created based on physical order of the SNPs along the autosomes ( mm9; NCBI build 37 ) ., Recombination fractions were estimated using R/qtl 87 ., Ordered genotypes were reconstructed at each marker from familial SNP data ( F15 parents and their F16 offspring ) using the Integer Linear Programming algorithm as implemented in PedPhase 2 . 1 88 ., Due to the computational intensity of the algorithm , it was necessary to partition the larger chromosomes before running the program ., Additive ( Xa ) and dominance ( Xd ) genotypic scores were assigned at each marker: Xa = 1 , 0 , −1 and Xd = 0 , 1 , 0 for the LL , LS and SL , and SS genotypes , respectively ., ‘L’ refers to an allele derived from the LG/J strain and ‘S’ refers to an allele derived from the SM/J strain ., Further , we assigned parent-of-origin imprinting genotypic scores ( Xi ) to distinguish between reciprocal heterozygotes , LS and SL ., By convention the first allele refers to that inherited from the father and the second from the mother ., Imprinting genotypic scores for LL , LS , SL , and SS are Xi = 0 , +1 , −1 , 0 , respectively 89 ., Additional genotypes were imputed at 1cM intervals between the markers using the equations of Haley and Knott 90 with the inclusion of equations derived for imputing imprinting genotypic scores 91 ., Single locus analyses were performed using maximum likelihood in the Mixed Procedure in SAS 9 . 2 ., Our full mapping included: sex , diet , sex-by-diet interaction , the direct effects of the genomic locations ( Xa , Xd , Xi ) , and their two- and three-way interactions with sex and diet as fixed effects ., The full model explains variation in trait ( Y ) using the linear equation:where µ is the population mean and e is the residual ., The regression coefficients are the additive a = ( GLL ) − ( GSS ) ) /2 , dominance d = ( ( GLS+GSL ) − ( GLL-GSS ) ) /2 and imprinting i = ( GLS−GSL ) /2 genotypic scores , where G refers to the mean phenotype of all individuals sharing the subscripted genotype , and their interactions with sex ( s ) and/or with diet ( d ) ., Family and its interactions with sex and diet , including the three-way interaction , were included as random effects in the model ., The −2 ln ( likelihood ) of this model was compared to a null model including only sex , diet and sex-by-diet interaction terms using a chi-square test with 12 df ., Probabilities were transformed into LOD = −log10 ( Pr ) ., The number of independent tests was calculated using the Li and Ji method based on the eigenvalues of the correlation matrix of marker additive genotype scores 92 ., This was used to calculate Bonferroni adjusted significance thresholds , 1− ( 1−α ) 1/M , where M is the number of independent tests ., A significance threshold was calculated at the genome-wide level ( LOD ≥3 . 90 ) as well as separately for each autosome ( Table S6 ) ., With chromosome-wise significance , we expect 1 false positive result per trait ., Our results overwhelm this in that there are 6–10 times the number of significant results for each trait as expected by chance under a null model of no QTL ., Further , QTL with chromosome-wise significance have a history of replication across different mapping populations of this cross 24 ., QTL for separately analyzed traits related to two or more metabolic syndrome ( MetS ) components mapping to the same cM position are considered pleiotropic QTL ., When QTL support intervals for separately analyzed MetS component traits overlapped , but the separate trait peaks did not map to the same cM position , a formal test of pleiotropy was performed as described by Cheverud 36 ., First , the most likely peak QTL positions for each single trait were identified , e . g . the position with the highest LOD score , and then the most likely combined position of the all the traits mapping to the region , weighted by their LOD scores , was identified ., A X2 for model fit was obtained at each single trait peak and at the combined-trait position ., The differences in X2 values between the separate and the combined-trait models were added together to generate a X2 test for pleiotropy 34 ., The | Introduction, Results, Discussion, Materials and Methods | Context-dependent genetic effects , including genotype-by-environment and genotype-by-sex interactions , are a potential mechanism by which genetic variation of complex traits is maintained in populations ., Pleiotropic genetic effects are also thought to play an important role in evolution , reflecting functional and developmental relationships among traits ., We examine context-dependent genetic effects at pleiotropic loci associated with normal variation in multiple metabolic syndrome ( MetS ) components ( obesity , dyslipidemia , and diabetes-related traits ) ., MetS prevalence is increasing in Western societies and , while environmental in origin , presents substantial variation in individual response ., We identify 23 pleiotropic MetS quantitative trait loci ( QTL ) in an F16 advanced intercross between the LG/J and SM/J inbred mouse strains ( Wustl:LG , SM-G16; n\u200a=\u200a1002 ) ., Half of each family was fed a high-fat diet and half fed a low-fat diet; and additive , dominance , and parent-of-origin imprinting genotypic effects were examined in animals partitioned into sex , diet , and sex-by-diet cohorts ., We examine the context-dependency of the underlying additive , dominance , and imprinting genetic effects of the traits associated with these pleiotropic QTL ., Further , we examine sequence polymorphisms ( SNPs ) between LG/J and SM/J as well as differential expression of positional candidate genes in these regions ., We show that genetic associations are different in different sex , diet , and sex-by-diet settings ., We also show that over- or underdominance and ecological cross-over interactions for single phenotypes may not be common , however multidimensional synthetic phenotypes at loci with pleiotropic effects can produce situations that favor the maintenance of genetic variation in populations ., Our findings have important implications for evolution and the notion of personalized medicine . | We look at gene-by-diet and gene-by-sex interactions underlying natural variation in multiple metabolic traits mapping to the same regions of the genome in a mouse model ., We find that the underlying genetic architecture of these traits is different in different sex and diet contexts ., We further use expression data and whole-genome polymorphism data to identify compelling candidates for experimental follow-up ., We use these results to examine theoretical evolutionary predictions about how variation in populations can be maintained ., There has been much discussion of late on how to use evolutionary theory to inform medical genomics ., Mouse models may be especially appropriate for bridging the divide between evolutionary and biomedical research , because they allow the study of the effects of natural alleles on normal variation and because human-mouse homology is well defined ., Our study is unique in examining quantitative trait loci from both evolutionary and biomedical perspectives , and we highlight the complex connections of the traits comprising the metabolic syndrome and the evolutionary implications of their underlying genetic architecture ., This is important for understanding disease etiology and is relevant to personalized medicine . | public health and epidemiology, population genetics, quantitative traits, population biology, genetic epidemiology, genetic polymorphism, environmental epidemiology, epidemiology, biology, evolutionary theory, evolutionary genetics, trait locus, phenotypes, heredity, genetics, evolutionary biology, genetics and genomics, complex traits | null |
journal.pcbi.1000440 | 2,009 | PoreWalker: A Novel Tool for the Identification and Characterization of Channels in Transmembrane Proteins from Their Three-Dimensional Structure | Transmembrane channel proteins play pivotal roles in maintaining the homeostasis and responsiveness of cells and the cross-membrane electrochemical gradient by mediating the transport of ions and molecules through biological membranes 1 ., For instance , aquaporins facilitate the flux of water and small uncharged solutes across cellular membranes and , in humans , are involved in several diverse functions , like concentrating urine in kidneys and participating in forming biological fluids 2–5 ., In contrast , potassium channels are fundamental regulators of cell membrane potential and control the action potential waveform and the secretion of hormones and neurotransmitters 6–8 ., Moreover , a family of transmembrane proteins , known as translocons , have been found to mediate protein transfers between different cellular compartments and consequently to be involved in the folding of membrane and secretory proteins 9 ., Understanding the structure and function of transmembrane channel proteins and studying their properties and biochemical mechanisms is therefore a very important task in biological and pharmaceutical research 10 , 11 ., Transmembrane channel proteins usually show a cavity spanning the whole protein , herein designated as the pore , which forms the path used by ions and/or molecules to cross the membrane ., The pore has two openings , one on each side of the membrane , and it has been hypothesized ( and in some cases shown ) that the specificity and selectivity to different solutes is strongly dependent on the particular structural or amino acid composition features of the channel 5 , 8 , 12 ., Consequently , computational methods for the identification and description of pores in transmembrane protein 3D-structures represent key tools to gain insights into how these proteins function ., To our knowledge , several methods for the analysis of protein surface and cavities have been developed 13–19 but the only currently available method for the structural analysis and visualisation of transmembrane channels is HOLE , developed in 1993 and still widely used 20 , 21 ., This elegant algorithm implements a Monte Carlo simulated annealing approach to find the path that a sphere of variable radius can use to go through a channel and also provides pore anisotropy analysis and conductance prediction tools ., The path is optimised so that it can be considered as the route of a plastic sphere squeezing through the channel , i . e . at each point of the path the channel can accommodate the largest possible sphere ., Three more recent methods , developed for the detection of internal cavities and tunnels in any protein structure , CAVER 22 , its improved version MOLE 23 and MolAxis 24 can be applied to identify pores in transmembrane proteins ., CAVER explores the protein inner space using a grid-based approach , while MOLE implements an algorithm based on Voronoi polyhedra ., Both approaches use an optimality criterion based on the minimization of a cost function , which depends on reciprocal atomic distances , and calculates the optimal way out from a user-specified starting point inside the protein to the outside environment ., MolAxis exploits computational geometry techniques , in particular the alpha shapes theory and the medial axis concept , to detect possible routes that small molecules or ions can take to pass through channels and cavities ., It is worth highlighting here that all the four programs , to be applied to transmembrane proteins , require user-defined specific information about the geometry of the channel that necessitate a fairly good knowledge of the location of the pore and/or of key residues lining the pore walls , like a starting point for the path search through the channel or a vector approximating the location of the pore within the protein 3D-structure ., Moreover , they provide only a limited description of the channel geometry mainly consisting of diameter values and some of the residues lining the pore walls ., Herein we present PoreWalker , a method to provide a detailed description of the three dimensional geometry of a channel ( or pore ) through a transmembrane protein , given the coordinates of the protein structure ., These 3D pore descriptors provide a quantitative description , including the size , shape and regularity of the pore , which we hope will help to explain pore specificity , the critical biological function of these molecules ., PoreWalker is fully automated , requiring only the 3D protein coordinates from the PDB file , and so can be applied to any new structure or across all transmembrane proteins in the PDB ., The method was applied to several structures of transmembrane channel proteins and was able to identify shape/size/residue features representative of specific channel families ., The software is implemented as a web-based resource at http://www . ebi . ac . uk/thornton-srv/software/PoreWalker/ and its source codes will soon be available upon request to the authors ., In transmembrane proteins , the channel runs approximately perpendicular to the membrane plane and parallel to the bundle or barrel that makes up the transmembrane portion of the pore ., The first step of the program consists in re-orienting the protein structure so that the origin lies at the centre of gravity of the transmembrane portion of the protein and the bundle/barrel lies perpendicular to the membrane plane ., The main axis of the transmembrane bundle/barrel is calculated according to the position of the secondary structure elements that putatively form it ., Each secondary structure element in the protein is identified from the separation of sequential C-alphas as described in Supplementary Text S1 and , if the helix or the strand is longer than 15 or 10 amino acids , respectively , it is approximated by a vector , which starts at its centre of mass and points toward the centre of mass of the terminal four and two amino acids of the helix or strand , respectively ., The length threshold was applied because , on average , transmembrane helices and strands used for this calculation need to be sufficiently long to cross the membrane ., This excludes small helices which often do not lie perpendicular to the membrane plane ., The sign of all the vectors is selected so that they point in approximately the same direction and the averaged vector is calculated ., However , outlying secondary structures found to be more perpendicular than parallel to the bundle/barrel axis are excluded from the averaging at this stage so that the transmembrane portion of the structure is orientated as parallel to the membrane axis as possible ., The whole protein 3D structure is then re-oriented so that its calculated main axis overlaps with the x-axis of the current 3D system and the centre of gravity of its transmembrane portion lies at the origin ., In this way , the structure is moved into a new reference frame that approximately aligns the transmembrane secondary structure elements perpendicular to the membrane ., The pore axis is then approximated as coinciding with the protein main axis ( see Figure 2 , step 2 ) ., This starting assumption , despite its crudeness , simplifies and speeds up the following steps of the method ., The centre of the pore is defined by iteratively maximising the number of detected putative pore-lining residues , i . e . water-accessible amino acids pointing towards the pore axis ., At the beginning of the process , the centre of the pore and the pore axis , i . e . the linear vector going through the middle of the pore , are assumed to correspond to the centre of mass of the protein and to the x-axis , respectively ., Putative pore-lining amino acids around the pore axis are then selected to satisfy three criteria: ( 1 ) the relative sidechain solvent accessibility calculated by NACCESS ( 25 , downloadable at http://www . bioinf . manchester . ac . uk/naccess/ ) must be higher than 5%; ( 2 ) the vector defined by the C-alpha-C-beta bond must point towards the pore axis; and ( 3 ) the distance of the C-alpha atom from the pore axis must be below a given threshold ., The distance threshold is calculated at each iteration as the smallest distance between any pore-lining residue C-alpha and the current pore axis plus 6 Å ., This prevents the inclusion of “second shell” residues in the selection of putative pore-lining residues and in the calculation of the final centre of the pore ., Glycines lack C-betas and are therefore treated differently ., For each Gly , a dummy atom is defined as the average of 3D-coordinates of its backbone carbonyl carbon and amide nitrogen ., This atom can be considered a mirror image of the C-betas of a virtual side chain located between the two hydrogen atoms bound to its C-alphas and can therefore be used to evaluate the orientation of Gly backbone atoms ., Glycines with a total relative accessibility higher than 5% and with the dummy atom pointing away from the pore axis are defined as pore-lining ., A new centre of the pore is then calculated from the selected putative pore-lining amino acids and the protein structure is translated so that the new pore centre and the x-axis corresponds to the origin of the 3D-system and to the new pore vector , respectively ., The above procedure is performed iteratively and stops when the number of newly selected putative pore-lining residues converges to its maximum , indicating that the pore centre has reached its optimal position ., As a result of this first process , the protein structure is translated in space so that the x-axis goes through the current best-guess of the centre of the pore and a preliminary list of putative pore-lining residues is generated ( see Figure 2 , step 3 ) ., The effectiveness of this step of the method was assessed by monitoring the distance of the selected pore-lining residues from the pore centre , as described in Supplementary Text S1 and shown in Figure S1 ., To derive the best possible axis and cavity of the pore an iterative slice-based approach is used , in which the centre of the pore is systematically optimised for each slice and therefore eventual irregularities in the cavity can be detected ., At each iteration , the protein structure is mapped onto a 3D-grid of 1Å steps and then sliced along the x-axis ( i . e . the current pore axis ) in 1Å thick layers ., The pore centre of each slice is then identified by a grid-based approach so that it lies at the centre of the sphere with the maximum radius that the slice can accommodate ., The maximum sphere and its centre are derived by expanding the sphere from the current centre until it clashes with a pore-lining atom , and systematically shifting the centre on the vertices of a 2D-grid so that the centre of the sphere of maximum volume for that slice can be identified ., The pore centre of the slice is initially set as the average of C-alpha and C-beta atoms of the putative pore-lining amino acids belonging to the slice selected in the previous step of the program , and the corresponding maximum sphere is calculated ., A square 2D-grid perpendicular to the current pore axis ( x-axis ) is then built and used to optimize the location of the pore centre ., The grid has 0 . 1 Å squares , it is centred at the pore centre , and its size depends on the sequence length of the protein and on the size of the pore ., Grid vertices not surrounded by atoms in all the possible y and z directions are taken as located outside the pore and excluded from the optimization process ., The sphere of maximum volume at a given centre is calculated by increasing its radius by 0 . 1 Å until it hits a vertex of the 3D-grid occupied by a backbone or C-beta atom ., The current sphere radius is adjusted by subtraction of the atomic van der Waals radius ( 1 . 8 Å , corresponding to the average radius of all types of heavy atoms found in protein structures as in the AMBER united force field 26 ) or approximate residue side chain radius ( as in Levitts amino acid ‘lollypop model’ 27 ) if a backbone atom or a C-beta is hit , respectively ., If the radius value is above any previously calculated radius , the current radius and corresponding sphere centre are taken as the maximum radius and pore centre for that slice ., At the end of the iteration , coordinates of the last four consecutive sphere centres at each end of the pore , that represent the two pore openings , are averaged to generate two points , which define the new pore axis ., The structure is then re-oriented to align with the new vector ( see Figure 2 , steps 4–8 ) ., The last four consecutive spheres are used because the ends of the channels can be very irregular in term of shapes and therefore pore axes derived from the two very last sphere centres ( one per end ) often do not cross correctly one or both the pore entrances ( the value 4 was derived on a trial-and-error basis in the range of values from 1 to 5 ) ., The refinement process stops when the new pore vector “overlaps” to the old pore axis ( i . e . when their angle is lower than 0 . 5 degrees ) and the current pore axis and maximum sphere radii ( i . e . those calculated in the previous iteration ) are retained as optimal and used in the further analysis of the pore shape ., The last step of the method is the analysis and calculation of three main pore descriptors: the pore-lining atoms and residues ( Section 4 . 1 ) , and the shape of the pore cavity ( Section 4 . 2 ) and its regularity ( Section 4 . 3 ) ., Pore descriptors calculated by PoreWalker for a submitted structure are summarised in the corresponding output webpage , which shows the features of the channel cavity and several visualizations of the pore based on the identified pore-lining residues ., As an example , the output of the bovine aquaporin-1 ( PDB code 1j4n ) is summarised in Figure 3 ., The 3D shape of the pore is simplified in 2D as a stack of building blocks shaped as trapezia for funnel-like shapes ( Figure 3B ) going from the most negative to the most positive coordinate along the pore axis ., In addition , the pore cavity is represented as a series of consecutive straight and wiggly lines representing channel areas where pore centres can ( straight ) or cannot ( wiggly ) be fitted to a line , respectively ( Figure 3E ) ., It is worth highlighting here that the approach does not take into account any chemistry ( e . g . H-bonds ) but just calculates the path of the pore centres ., In practice , ions/molecules may well hop between low energy off-centre sites , within the channel , that optimize their interactions with pore residues during their passage through the channel ., Vertical and horizontal visualizations of the pore help to provide a better understanding of the channel features ., Vertical sections ( Figure 3A , D ) are generated halving the protein structure along the pore axis , while horizontal sections ( Figure 3G , I ) are produced as 5Å slices of the protein structure perpendicular to the pore axis ., Amino acids are coloured according to whether they are classified as pore-lining and red spheres represent optimal pore centres ., PoreWalker was tested on the 19 structures from the “Membrane Proteins of Known 3D Structure” resource ( http://blanco . biomol . uci . edu/Membrane_Proteins_xtal . html ) listed in Table 1 , that include both ion and small molecule channels with straight and curve pores ., Results are shown in Table 1 , Figure 4 and Supplementary Figure S2 ., Although there is no fully comprehensive experimental data to assign with certainty the location and residue composition of channels in transmembrane protein 3D-structures , the position of the pore axis and of the pore centres , visually analysed in relation to the protein structure , and the minimum diameter value give a hint of the effectiveness of the method ., From visual inspection , PoreWalker seems able to locate correctly the pore axis and the pore centres in most of the cases and therefore to identify fairly correctly the amino acids that line the pore walls with one or more atoms ., In fact , the pore axis seems wrongly located only for the Amt-B ( Figure 4K ) , Amt-1 ( Supplementary Figure S2E ) and the SecYE-beta translocon ( Figure 4H ) channels ( PDB codes 1xqf , 2b2f and 2yxr , respectively ) ., Both Amt-B and Amt-1 channels share a common hour-glassed shape with multiple exits at one of the pore gates and can therefore be thought to include more than one transmembrane tunnel of different length ( Figure 5B ) ., Likewise , the SecY-beta translocon shows two flexible loops at both sides of the channel that make a further narrower but longer cavity crossing the protein structure ., Despite the misassignments of pore axis and pore centres , in these three examples most of the pore-lining residues still seem to be identified correctly because the calculated optimal cavities , indicated by red spheres , partially overlap with the “true” cavities , indicated by the black arrows ., In terms of pore shape , PoreWalker seems to recognise common sub-shapes across channel families ., For instance , all aquaglyceroporins show a DU-like string shape ( where D and U represent funnel-like shape of decreasing and increasing diameter , respectively ) , which represents a hour-glasssed shape confirmed by a few published data 5 , 28 , 29 ., Likewise , potassium channels present a shared sub-shape , a DUD sub-string shape at the cytoplasmic side of the channel , that is in agreement with the channel features reported by Mackinnon et al . , i . e . a constriction at the cytoplasmic side , the internal pore , widening into a larger water-filled void , the internal cavity , which leads towards the narrow selectivity filter located at the periplasmic side of the channel 12 ., In addition , the linearity of the cavity seems to give some insights on the pore selectivity to different types of solutes ( Table 1 ) ., In fact , 10 of the 13 channels for inorganic ions in the set showed a very regular cavity , with average percentage of co-linear pore centres of 91 . 9% ( SD\u200a=\u200a7 . 0% ) and organic small molecule/ion channels had less regular pores , with percentage of co-linear centres lower than 60% ., For completeness , PoreWalker output was also compared with results obtained using HOLE and MolAxis on the same set of structures ., A systematic comparison with MOLE results could not be performed because , probably due to the intrinsic looseness of some structures , like the MthK and the ASIC1 channels , many of the tunnels identified by MOLE lie parallel and not perpendicular to the membrane axis and could not be considered as transmembrane ., Within the set of pore features produced by PoreWalker and HOLE , the only comparable quantitative measure is the diameter , calculated along the pore at given heights ., Diameter profiles obtained at 1Å steps for the 19 transmembrane proteins in the set were compared using the R2 correlation coefficient ( see Supplementary Text S1 , Table 1 , and Figures 6 , S3 , S4 and S5 ) ., Pore diameter analyses performed with the two methods showed good agreement for 12 of the 19 diameter profiles , with R2 higher than 0 . 75 ., However , the remaining 7 profiles showed very poor correlation coefficients , with R2 very close or equal to zero ., This behaviour seemed to be strongly affected by the regularity of the cavity ., In fact , R2 values showed a good correlation with the number of co-linear pore centres ( Supplementary Figure S5 ) with a R2 of 0 . 70 and only one strong outlier , the sodium-potassium channel ( PDB code 2ahy ) ., The disagreement between the two profiles in this case was due to a completely different pore exit at the top channel side identified by HOLE that seems visually incorrect and makes the diameter trend in that area very peculiar ., As for MolAxis , the program does not calculate diameter values at given heights along the channel axis but provide a partial list of the amino acids that contribute to the pore surface ., Therefore , minimum diameters and pore lining residues were used to compare PoreWalker and MolAxis results ., MolAxis could not identify a channel for 9 of the 19 test protein structures ( Table 1 ) , the water , glycerol and ammonia channels and three potassium channels ., For the remaining 10 proteins , minimum diameter values derived from the two methods gave poor correlation ( R2\u200a=\u200a0 . 46 ) ., The exclusion of the SecYE-beta translocon , incorrectly characterised by PoreWalker , lead to an R2 of 0 . 69 ( corresponding MolAxis-HOLE R2 were 0 . 60 and 0 . 57 , respectively ) ., Minimum diameters calculated by HOLE and PoreWalker gave a better correlation , with R2 of 0 . 54 and 0 . 90 , respectively ( the overall R2 on the 19 structure set was 0 . 67 ) ., In term of pore-lining residues , MolAxis provides a list of the amino acids responsible for the calculated diameters , i . e . a subset of the amino acids that make the surface pore ., MolAxis pore-lining residues were fully included in PoreWalker pore-lining residue list in all the compared proteins but the SecYE-beta translocon ., In this case , 23 of the 24 pore-lining residues detected by MolAxis were included in the list generated by PoreWalker , showing that the method can reliably identify amino acids which build a channel despite mis-placements of its pore vector ., Finally , transmembrane pores identified by PoreWalker were found to coincide well with molecules of solute found in the 3D structure ., Figure 5C–F shows the SoPIP2;1 plant aquaporin ( 1z98 ) and the sodium-potassium channel ( 2ahy ) filled with water molecules and sodium and calcium cations , respectively ., In both cases the cavities generated by PoreWalker completely surround and include water molecules and ions , which provide good evidence for the location and shape of the pore ., Interestingly , PoreWalker is also able to identify the two main choke points in the water channel of the SoPIP2;1 reported to be in a closed state -the canonical Ar/R constriction site near the top of the pore and a narrower restriction close to the bottom of the channel ( Figure 5D ) ., The method can therefore analyse and characterise both “open” and “closed” transmembrane protein channels and transmembrane transporters ., The KcsA potassium channel is a homotetrameric integral membrane protein with high sequence similarity to all the potassium channels , particularly in the pore region ., Its channel includes three elements:, 1 ) a narrow entrance , known as the internal pore , starting at the intracellular side of the membrane;, 2 ) an internal cavity , about 10Å in diameter , at the middle of the membrane;, 3 ) a further narrowing , the selectivity filter , which leads to the extracellular environment 30 ., The KcsA channel is therefore a good target to assess the ability of PoreWalker to detect constrictions , gates and internal cavities in the 3D-structure of a channel protein ., The 3D structures of the Kcsa potassium channel in the presence of low ( 3 mM , Figure 7A ) and high ( 200 mM , Figure 7C ) K+ concentrations are available at the wwPDB ( codes 1k4c 30 and 1bl8 8 , respectively ) and their pore features were derived and analysed using PoreWalker ( Figures 7–9 ) ., The diameter profile of the low-K+ channel ( Figure 7B , solid line ) shows that PoreWalker can neatly identify the three main features of the channel: first a ∼3Å narrowing corresponding to the internal pore , the internal ∼9 . 0Å bigger cavity and a second narrower ( ∼1Å ) constriction corresponding to the selectivity filter , highlighted in the Figure in orange , blue and red , respectively ., It is interesting to notice here that diameter values calculated at 1Å steps by both HOLE ( dotted line ) and PoreWalker ( dashed line ) at the maximum width of the internal cavity ( ∼4Å ) were significantly smaller than those reported in the description of the 3D-structure 30 ( ∼10Å ) and found using the standard PoreWalker protocol at 3Å steps ( ∼9Å ) ., The calculated diameters of the internal pore and cavity also strongly agree with the proposed mechanism of ion conductance through the pore ., In fact , potassium cations are thought to move through the internal pore and cavity in a hydrated form and to be dehydrated at the selectivity filter ., The internal pore detected by PoreWalker is ∼3Å in diameter and could allow through one water molecule per time ( the average diameter of a water molecule is usually taken as 2 . 8Å ) ., Therefore , K+ ion could move through it alternating with water molecules ., On the other side , the selectivity filter has a predicted diameter of ∼1Å and could therefore let through only dehydrated K+ cations ., The comparison of the diameter profiles of the channel in presence of low and high quantity of ions ( Figure 7D , solid and dotted line , respectively ) showed that besides expected differences at the cytoplasmic side of the pore , where a gate mechanism is known to operate , the entrance of the selectivity filter is ∼2 . 5Å wider at high concentrations of K+ ., According to PoreWalker , the pore lining residues , which define access to the selectivity filter , are the Thr75s from the four chains making up the pore ., The difference in pore diameters at this point seems mainly to be due to different Thr sidechain conformations ( Figure 7E–F ) ., A significant difference in the two conformations of the KcsA selectivity filter had been previously highlighted at the level of residues Val76 and Gly77 ., A deeper analysis of the whole selectivity filter ( Figure 8A ) showed that the periplasmic side of the filter ( at the top of the Figure ) varies very slightly , while a major change is hinged at Gly77 and extends through Val76 to Thr75 , where a pincher-like shutting mechanism could reasonably be hypothesized ( RMSDs of all-atom superpositions were 0 . 33Å , 0 . 58Å and 0 . 99Å for Gly77 ( Figure 8B ) , Val76 ( Figure 8C ) and Thr75 ( Figure 8D ) , respectively ) ., Besides , the internal cavity accommodates K+ ions as hydrated by eight water molecules ., The 3D-structure of the low-K+ channel cavity ( Figure 9 ) shows that the four water molecules facing the filter are aligned to the sidechain oxygens of Thr75s and can make hydrogen bonds with them ( inter-oxygen distances are 3 . 9Å ) ., Moreover , their distances from the corresponding K+ ion are close to optimal ( 3 . 4Å versus 2 . 8Å 31 ) ., Therefore , it might be reasonably thought that the pinching mechanism could be aimed at weakening the water-K+ hydration complex by increasing the distance between the water molecules and the ion to facilitate its way into the pore ., We developed PoreWalker , a novel web-available method for the detection and characterisation of channels in transmembrane proteins from their three-dimensional structure ., PoreWalker is fully automated and very user-friendly , requiring as input only the 3D coordinates of a transmembrane protein structure ., A key prerequisite of the submitted structure is the presence of a transmembrane helix bundle or beta-barrel creating the pore , which is needed for the geometrical identification of the main protein axis ., If this condition is not met , the detection/description cannot be performed with the current version of the software ., In term of outputs , in addition to diameter profiles , PoreWalker describes several specific pore features , in particular the shape and the regularity of the channel cavity , the atoms and corresponding amino acids lining the pore wall , and the position of pore centres along the channel ., These features can be very helpful to gain further insights into the functional and structural properties of transmembrane protein channels by triggering specific in silico or experimental analyses , as shown from the recent characterization of the bacterial TolC channel 32 ., PoreWalker is based on the assumption that , in a transmembrane channel protein , the pore is made by the longest cavity crossing the protein along the main axis of its transmembrane portion and therefore detects the longest widest cavity in a transmembrane protein structure ., However , there are cases , as in the Amt-B and the SecYE-beta translocon , where the longest widest cavity does not correspond to the most likely “true” channel and in such cases the method assigns incorrectly one or both the pore gates ., Interestingly , for these examples , calculated optimal cavities partially overlapped with the “true” cavities and most of the pore-lining residues were anyway identified properly ., In summary , PoreWalker provides a robust and automated resource to interpret , coordinate data and derive quantitative descriptors , which help to provide a deeper understanding and classification of membrane protein structures . | Introduction, Materials and Methods, Results/Discussion | Transmembrane channel proteins play pivotal roles in maintaining the homeostasis and responsiveness of cells and the cross-membrane electrochemical gradient by mediating the transport of ions and molecules through biological membranes ., Therefore , computational methods which , given a set of 3D coordinates , can automatically identify and describe channels in transmembrane proteins are key tools to provide insights into how they function ., Herein we present PoreWalker , a fully automated method , which detects and fully characterises channels in transmembrane proteins from their 3D structures ., A stepwise procedure is followed in which the pore centre and pore axis are first identified and optimised using geometric criteria , and then the biggest and longest cavity through the channel is detected ., Finally , pore features , including diameter profiles , pore-lining residues , size , shape and regularity of the pore are calculated , providing a quantitative and visual characterization of the channel ., To illustrate the use of this tool , the method was applied to several structures of transmembrane channel proteins and was able to identify shape/size/residue features representative of specific channel families ., The software is available as a web-based resource at http://www . ebi . ac . uk/thornton-srv/software/PoreWalker/ . | Transmembrane channel proteins are responsible for the transport of ions and molecules through biological membranes and are pivotal for the physiology of the cell ., In fact , their incorrect functioning is involved or related to several diseases ( diabetes , myotonia , Parkinsons disease , etc . ) ., Moreover , their specificity and selectivity to different ions or molecules have been hypothesized and sometimes shown to strongly depend on the shape and size or amino acid composition of the channel ., Therefore , computational methods to identify and quantitatively characterise channel geometry in transmembrane protein structures are key tools to better understand how they function ., We have developed PoreWalker , a new method to detect and describe the geometry of these channels in transmembrane proteins from their 3D structures ., The method is fully automated , very user-friendly , identifies the location of the channel and derives a number of channel features: diameter profiles at given heights along the channel , all the residues lining the channel walls , size , shape and regularity of the channel ., These features can be very helpful in the study of how these channels might function ., We have applied PoreWalker to several channel protein structures and were able to identify shape/size/residue features that were representative of specific channel families . | computational biology/macromolecular structure analysis | null |
journal.pcbi.1000663 | 2,010 | In Silico Analysis of the Apolipoprotein E and the Amyloid β Peptide Interaction: Misfolding Induced by Frustration of the Salt Bridge Network | Alzheimers disease ( AD ) is one of the most common neurodegenerative diseases at the present time ., The disease is characterized by the formation of neurofibrillary tangles and plaques in the brain , leading to neuronal dysfunction , neuronal loss and finally death ., The main component of the plaques is the amyloid-β peptide ( Aβ ) , a 39–43 amino acids long hydrophobic peptide generated by the cleavage of the amyloid precursor , which accumulates in the form of soluble and non-soluble aggregates ., The connection between Apolipoprotein E ( ApoE ) and AD is well established 1 , 2 ., Structurally , ApoE is a 299 residues protein with an N-terminal domain involved in binding to heparin , low density lipoprotein receptors ( LDLR ) and LDLR-related proteins 3 , 4 ., The C-terminal domain has been related to heparin and lipid binding 5 , 6 ., Three main isoforms have been described for human ApoE , i . e . ApoE2 , ApoE3 and ApoE4 ., The standard variant is ApoE3 , while ApoE2 is defective for receptor binding , causing APOE ε2/ε2 homozygotic individuals to have a higher predisposition to diseases related to high amounts of cholesterol and triglycerides 3 , 7 ., For ApoE4 , the receptor binding affinity remains unaffected , but APOE ε4/ε4 homozygotic individuals have higher risk for coronary heart disease and a significantly greater risk for developing AD ., 1 , 8 Around 80% of all AD cases are related to the genetic variance at the ApoE locus 9 , 10 ., The only difference between the ApoE isoforms is found in residues 112 and 158 , where Cys112/Cys158 corresponds to ApoE2 , Cys112/Arg158 to ApoE3 , and Arg112/Arg158 to ApoE4 ., The presence of cysteines at these positions confers oligomerization properties to ApoE ., Indeed , ApoE2 and ApoE3 are able to form disulfide-linked homo- and hetero-oligomers due to the presence of “respectively” two and one Cys residue ., ApoE4 lacks the possibility of strong disulfide linking; however , it is unclear whether weaker interactions could promote the oligomerization of ApoE4 ., The Cys/Arg substitution in ApoE4 also has molecular impact in terms of intra-protein polar contacts: the orientation of Arg61 is different in ApoE4 compared to ApoE3; the orientation of Arg61 towards the C-terminal domain ( See Figure 1A ) facilitates a salt bridge between Arg61 and Glu255 ., The electrostatic interaction between Arg61 and Glu255 promotes an N- and C-domain interaction that packs the structure tighter , which seems crucial for the interaction of ApoE4 with triglyceride-rich lipoproteins ., The interaction between Arg61 and Glu255 is absent in ApoE3 leading to a more open structure and a preferential binding of phospholipid-rich high-density lipoproteins 11 , 12 ., Chemical and thermal denaturation experiments have shown that the most unstable structure belongs to ApoE4 , which displays a partially unfolded intermediate ( molten globule ) containing some β structure that may be related to the fact that ApoE4 enhances the deposition of Aβ 13 , 14 ., Although different mechanisms have been proposed to explain the physiological and pathological relationship between ApoE and the Aβ peptide , the details of the interaction between ApoE and Aβ at a molecular level are unknown ., Such detailed knowledge is however important for the understanding of the pathological mechanisms of AD , and may also help to identify potential therapeutic target sites where the interaction between ApoE4 and Aβ can be blocked ., In the present study we are using molecular docking simulations based on global minimum energy to investigate the interaction process of Aβ with the N-terminal domain of the different ApoE isoforms in order to determine potential Aβ peptide binding sites in ApoE ., In the next step , molecular dynamics ( MD ) calculations are undertaken to explore the conformational dynamics of ApoE under Aβ interaction and evaluate the stability of each of the ApoE-Aβ complexes ., From the analysis and the statistics of the electrostatic interactions of the three ApoE isoforms , we present a model explaining the role of the Aβ-ApoE interaction and its relevance for AD ., Molecular dockings followed by MD simulations were used to study the interaction of Aβ with the different isoforms of ApoE ., In order to study the Aβ peptide binding site on the N-terminal domains of the three ApoE truncated isoforms we used the Aβ ( 1–40 ) peptide as ligand , employing an SDS-induced α-helix solution structure previously determined by NMR spectroscopy 15 ., Indeed , such helical fold in the Aβ monomeric state ( non-aggregated ) has been shown to be the most populated one in highly hydrophobic environments 16 ., On the other hand , the structures of the three ApoE truncated isoforms were taken from lipid-free structure determinations by X-ray crystallography 11 , 17 , 18 , which correspond only to the N-terminal domain ( 144 residues including the LDLR domanin of ApoE ) ., Water molecules in the pdb files were removed prior to docking and energy minimizations were carried out to refine the structures ., All 3D models of the ApoE-Aβ complexes were found to be quite different ., Although the Aβ ( 1–40 ) peptide assembles between the first and fourth ApoE helix for all ApoE isoforms , the orientation of the peptide was found to depend on the ApoE variant ( Figure 1B; see Figure S1 for comparison of the 10 lowest energy solutions for each isoform ) ., For ApoE2 and ApoE4 , the C-terminus of the peptide faces the N-terminus of the protein , though the assembly is different ., For ApoE3 , the peptide is turned around , and the N-terminus of the peptide faces the N-terminus of the protein ., Early studies indicated that ApoE interaction with Aβ fibrils is partially dependent on ionic interactions 19 ., Thus , the single change of Cys158 in ApoE2 to Arg158 in ApoE3 changes the distribution of ionic residues influencing the assembly of Aβ ( 1–40 ) , while the double change of Cys112 and Cys158 to Arg112 and Arg158 in ApoE4 distributes the ionic residues in an ApoE2-like way ., A 10 ns classical MD simulation including explicit water of the three ApoE isoforms together with the Aβ peptide was carried out on each of the lowest energy ApoE-Aβ models obtained by docking calculations as well as on each isolated species ., Figure 1C shows the root-mean-square deviation ( RMSD ) of the MD simulation for the three ApoE isoforms in the presence and absence of the peptide ., In their unbound form , no conformational transitions were detected for the ApoE isoforms , in agreement with previous results 20 ., However , in presence of the peptide , different behaviors were observed between the isoforms ., Despite the existence of interaction , no conformational transitions were detected for the ApoE2-Aβ or the ApoE3-Aβ complexes ., However , the ApoE4-Aβ complex showed a large conformational transition indicated by a significant RMSD change of about 10 Å in the 10 ns timescale ( Figure 1C ) ., In Figure 1D , four snapshots of the 10 ns MD simulation for the ApoE4-Aβ complex are presented ., Focusing on ApoE4 , during the first 0 . 3 ns , the third helix of ApoE4 started to unfold and a loop appeared between residues 112 and 92 which affected the whole third helix ., This structural disturbance was caused by the onset of new electrostatic interactions rising from the interaction with the peptide ., For the Aβ peptide , the first conformational change appeared in the Glu22-Asp23 region ., At 1ns the second helix of ApoE4 showed a conformational change ., In the snapshots of 5 ns , the first and fourth helices of ApoE4 were still stable , but at 10 ns a large conformational change had occurred , coinciding with a fully extended Aβ ( 1–40 ) peptide ., At 10 ns , the hydrophobic groups inside the ApoE4 helices had become exposed to the solvent ., The interruption of the stable salt bridge network by external electrostatic interactions ( coming from the peptide ) was thus transmitted from the dense helix region to the whole protein , causing a severe loss of α-helical structure ., Further investigation on the conformational change induced in ApoE4 by the complexation with Aβ was carried out through the analysis of the distances between charged residues ., For this analysis , direct salt bridges have been assumed to be around 4 . 3 Å , whereas indirect or water-mediated salt bridges have been assumed to have a distance between 4 . 3 and 7 . 0 Å as reported by Dzubiella et al . 21 ., In the most stable ApoE4-Aβ complex , the peptide interacted with helices I and IV of ApoE4 ., The Aβ residues responsible for these interactions were the negatively charged Asp1 and Asp23 , which interacted with positively charged arginines in ApoE4 ( Arg38 in helix I and Arg142 in helix IV respectively ) ., The direct salt bridge between AβAsp23 and ApoE4Arg38 was very strong ( Figure 2A ) , while the salt bridge between AβAsp1 and ApoE4Arg142 did not exist during most of the MD simulation , and only became more plausible at the end of the MD simulation ( the distance for an indirect salt bridge being reached after circa 8 ns , Figure 2A ) ., Focusing on helices I and II of the N-terminal domain of ApoE4 , the distance between Arg38 and Asp35 changed during the 10 ns time window ( see Figure 2B ) ., A transition occurred from 10 to 2 . 5 Å in the 2 ns time window , which then went back to 10 Å ( indicating the breaking of the Arg38-Asp35 salt bridge ) , and became stable at 7 ns ., For comparison , the same distance is shown for the MD simulation of ApoE4 alone , where no change at all can be seen , as the distance was within the 4 . 3 and 7 . 0 Å range during the whole 10 ns ( Figure 2B ) ., The salt bridge between Asp35 and Arg32 was stable below 4 . 4 Å before 2 ns ( Figure 2C ) ., For ApoE4 in the absence of Aβ , the distance remained constant around the 7 . 0 Å threshold , making it difficult to determine the existence of an indirect salt bridge ., For the ApoE4-Aβ complex , the direct salt bridge involving Arg32 and Glu66 ( in helices I and II , respectively ) was affected and showed a maximal fluctuation from 2 . 5 to 7 . 5 Å and then back to 2 . 5 Å in the 10 ns time window ( Figure 2D ) ., In the ApoE4 alone MD , this Arg32-Glu66 pair did not show any propensity to interact ( the distance was over 7 . 0 Å during the whole 10 ns ) ., For helices II and III of the N-terminal domain of ApoE4 , the transitions of the Arg61-Glu66 , Arg61-Glu109 and Glu109-Arg112 salt bridges were monitored in the ApoE4-Aβ complex ( see Figure 3 ) ., At 5 ns the distance between Glu66 and Arg61 from helix II dropped from about 10 to 3 Å , becoming stable and forming a direct salt bridge ( see Figure 3A ) ., However , for ApoE4 alone , this salt bridge was never formed ., For the ApoE4-Aβ complex in the 5 ns interval , the direct salt bridge between Arg61 and Glu109 ( helix III ) broke down ( the distance increased from about 3 to 12 . 5 Å , Figure 3B ) ., In ApoE4 alone the distance for this pair was out of range during most of the MD simulation ., However , the distance between Glu109 and Arg112 ( both in helix III ) remained relatively stable and below the salt bridge distance threshold ( Figure 3C ) ., In the ApoE4-Aβ complex , the Glu109-Arg112 salt bridge was direct ( below 4 . 3 Å ) , whereas for ApoE4 alone , the salt bridge was more indirect or water mediated ., The MD results for the Arg112-Asp110 pair ( Figure 3D ) were similar to those for the Arg112-Glu109 pair ., In the complex , the distance for the electrostatic pair indicated a direct salt bridge , whereas for ApoE4 alone , this distance was closer to an indirect salt bridge ( if any ) ., Electrostatic interactions between helix III and helix IV were more complex and insensitive to the interaction with the peptide , and the bridge network involving helices III and IV remained stable during the simulation ( data not shown ) ., In the ApoE4-Aβ complex , the interaction between Arg112 with Asp110 and Glu109 in helix III is connected to helix IV via the Asp110-Arg147 and Asp107-Arg151 ion pairs ( see Figure 4 ) ., Also Asp107 in helix III and Asp151 in helix IV interacted with Arg147 ., Another inter-helical ion pair network existed between Arg103 , Glu96 and Arg92 in helix III and Arg150 , Arg153 , Arg154 and Arg158 in helix IV ( see Figure 4 ) ., Arg158 acted as a bridge for extending the electrostatic interaction between Glu96 and Arg92 ., Our computational approach assumes a direct interaction between ApoE and Aβ ., Although the docking was plausible for ApoE2 and ApoE3 , the interactions did not generate any conformational transition in the 10 ns time window while for ApoE4 , the interaction promoted unfolding of the ApoE4 , as shown by the MD simulations ., This result is compatible with earlier thermal and chemical denaturation studies using circular dichroism and scanning calorimetry , which have indicated stability differences ( ApoE4<ApoE3<ApoE2 ) among the three isofoms ( experiments were carried out on the 22 KDa truncated protein , corresponding to the N-terminal domain ) 13 , 14 ., The present results also agree with the existence of a partially unfolded intermediate for ApoE4 22 ., However , a direct comparison of the present results with the previous experimental results is not possible ., The MD results for the ApoE isoforms alone do not indicate any of the trends shown experimentally , probably because of the time scale ( nanoseconds vs . seconds/minutes ) ., But in the case of ApoE4 , it is likely the Aβ peptide behaves as an unfolding catalyzer ., Thus , effects on the stability of ApoE2 and ApoE3 exerted by Aβ peptide at longer time scale cannot be discarded ., The proposed ApoE4-Aβ complex forms between helices I and IV of ApoE4 ( proposed model in Figures 1B and 4 ) ., As seen from the docking procedure , the complex formation does not directly affect the salt bridges involving Arg61 , but the cascade of events generated by the interaction leads to the stabilization and destabilization of the Arg61-Glu66 and Arg61-Glu109 salt bridges , respectively ., Arg112 in ApoE4 causes the side chain of Arg61 to extend away from the four–helix bundle which will allow electrostatic interaction with Asp65 , Glu66 and Glu59 ( see Figure 4 ) ., In ApoE2 and ApoE3 , Arg61 shows a different orientation ( due to Cys122 , see Figure 1A ) , hindering the interaction with the charged residues from helix III ., The fluctuation of the salt bridges in helices I and II could be explained by the interruption of the Arg38-Asp25 salt bridge in ApoE4 ., This effect is most likely induced by Asp23 of Aβ , which will affect the neighboring salt bridge between Asp35 and Arg32 ., Another affected interaction would be the inter-helix salt bridge between Arg32 ( helix I ) and Glu66 ( helix II ) ., The MD simulations show that this initial chain of events induced by the presence of the Aβ peptide and occurring in helices I and II of ApoE4 ( but not ApoE3 ) would soon be transmitted to helix III stabilizing the Arg61-Glu66 and breaking the Arg61-Glu109 salt bridges in this N-terminal domain , and probably affecting also the Arg61-Glu255 salt bridge in the full protein form ., Disruption of this domain interaction by the ApoE4 R61T mutation has been shown to reduce Aβ production 23 ., In the same study , an ApoE4 docking site involving residues 109 , 112 and 61 , was defined as a binding site for blocking agents capable to disrupt the domain interaction leading to a decrease in Aβ production 23 ., The other contact point comprising AβAsp1 and ApoE4Arg142 appears less relevant for the destabilization of the salt bridge network; however , Arg142 is within the heparin and receptor binding region ( localized around residues 141–150 of ApoE ) ., This direct interaction may shield the ApoE4 binding region , affecting the cell membrane recognition of ApoE4 interacting with Aβ ., As shown by in vitro studies , both ApoE3 and ApoE4 interact with Aβ and form SDS stable complexes ., ApoE-Aβ complexes have been isolated from AD brain extracts and shown to be stable and as tightly packed as Aβ fibrils 24 , 25 ., Our results indicate the possibility that both ApoE3 and ApoE4 bind to the peptide with different orientations ., Assuming the protective role of ApoE3 compared to the detrimental role of ApoE4 in AD ( for an extensive review see Huang et al . 26 ) , we can speculate the following: the binding of the peptide with ApoE3 does not affect the stability of the protein nor the complex , leading to the peptide clearance ., On the other hand , the lower stability of ApoE4 is even more emphasized by the interaction with Aβ: the interaction triggers the partial unfolding of ApoE4 into a misfolded intermediate which we suggest is incapable of performing the clearance of Aβ , and leading to pathogenic effects such as the promotion of amyloid forming processes ., In our results , mostly the N-terminus of the peptide is involved in the ApoE4-Aβ complex formation ( residues 1 and 23 ) ., Previous studies with Aβ peptide have shown that electrostatic interactions are the main cause for the formation of larger oligomers and that the C-terminus region is important for the formation of such oligomers 27 ., Discrete MD simulations have shown that the Gly37-Gly38 turn plays an important role in the formation of Aβ ( 1–42 ) pentamers 28 ., Thus , we can speculate that the non-involvement of the C-terminus in the complex formation could favor the interaction of free Aβ C-termini , thus provoking the aggregation of the ApoE4-Aβ complexes ., This Aβ effect could probably be overcome by the usage of agents ( such as GIND-25 and GIND-105 ) 23 binding to the Arg61/Glu109/Arg112 ApoE4 binding site , which would stabilize the protein by disrupting the Arg61-Glu255 salt bridge , generating an ApoE3-like variant ., In the same way , Aβ and ApoE derived peptides have also been used as blocking therapeutic agents of both the protein and the peptide 29 , 30 ., We propose that the interaction of Aβ with ApoE4 induces a partially unfolded intermediate by the frustration of the existent network of salt bridges ., The four-helix bundle of ApoE4 opens up and the hydrophobic core becomes exposed due to the ApoE4-Aβ complex formation , presumably rendering the protein incapable of performing Aβ clearance ., The interaction with Aβ affects the proposed binding site formed by Arg61/Glu109/Arg112 in ApoE4 , a binding site that has been shown to be relevant for substances capable of reducing the Aβ production ., The model here presented has implications for therapeutic drug design for AD , as it defines on a molecular level the ApoE-Aβ complex as a potential drug target ., Crystal structures of the three ApoE truncated isoforms ( containing only the N-terminal domain ) were downloaded from the PDB database ( ApoE2 , E3 , E4 , respective IDs: 1LE2 , 1LPE and 1LE4 ) , together with the Aβ peptide solution structure , determined by NMR in 10% SDS/Water ( ID:1BA4 ) and used as the docking model ., Crystallographic waters were removed and the structures were fully solvated before energy minimization ., Energy minimization was performed for the macromolecules using the GROMACS3 . 3 . 2 software with GROMOS96 as the force field 31 ., The RMSD between the initial and the energy minimized structures was lower than 0 . 01 Å for the ApoE isoforms ., For the Aβ peptide , due to the flexibility of the N-terminus , the RMSD was 4 . 7 Å ( RMSD of 0 . 8 Å for the α-helix Aβ residues 13 to 40 ) ., The structures obtained after energy minimization were used in PatchDock ( http://bioinfo3d . cs . tau . ac . il/ ) , where candidate solutions were generated by rigid-body docking methods 32 , 33 ., PatchDock determined the best starting candidate solutions based on shape complementarity of soft molecular surfaces ., The Clustering RMSD was 4 . 0 Å for analysis and the complex type was set to default ., The PatchDock algorithm divides the Connolly dot surface representation of the molecules into concave , convex and flat patches ., Then , complementary patches are matched in order to generate candidate transformations 32 , 33 ., Each candidate transformation is further evaluated by a scoring function that considers both geometric fit and atomic desolvation energy ., The 1000 best docked candidate transforms from PatchDock , based on global energy , aVdW , rVdW , atomic contact energy , and insideness measurements , were then used in FireDock ( http://bioinfo3d . cs . tau . ac . il/ ) 34 ., FireDock optimized , refined and rescored the 10 top candidate solutions by restricting the flexibility to the side-chains of the interacting surface and allowing small rigid-body movements ., For this study , we selected the first best candidate solution from FireDock for the ApoE2- , ApoE3- , and ApoE4-Aβ complex ., Energy minimization , equilibration and molecular dynamics simulations were carried out at neutral pH using the GROMACS3 . 3 . 2 software with GROMOS96 as the force field 31 ., The complexes of each of the three ApoE isoforms with Aβ peptide from the above-mentioned docking were used as the starting points for the simulations ., Bond lengths were constrained using the LINCS algorithm and the SETTLE algorithm was used for hydrogen bonding of water ., First , macromolecules from the docking model were solvated in a cubic box of 8Å cutoff with TIP3P water ., Each complex was minimized with 2000 steps using the steepest descent algorithm in order to relieve bad interactions between ApoE and Aβ peptide ., The system was equilibrated by first running 10 ps of position-restrained molecular dynamics; then the temperature of the system was gradually increased to 300 K . Berendsens temperature coupling method ( time constant of 0 . 1 ps ) was used in an unrestrained simulation ., Water molecules were equilibrated in the presence of the protein complex for 10 ps before running an unrestrained molecular dynamics simulation for 10 ns ., For unrestrained molecular dynamics simulation , the temperature coupling and pressure coupling were conducted in the NpT ensemble by using a Berendsen thermostat of 300 K and 0 . 1 ps relaxation time ., The pressure was 0 . 5 bar with 0 . 000045 compressibility and 1ps relaxation time , respectively ., The simulations with 300 K were applied by 173529 seeds ., Isotropic pressure coupling and Berendsens temperature coupling were then used during a 10 ns molecular dynamics simulation ., In addition , two MD simulations were run involving the three ApoE isoforms alone , following the above-mentioned process ., All molecular representations in this study were generated using Chimera v1 . 4 ( http://www . cgl . ucsf . edu/chimera/ ) 35 ., The g_rms and g_dist of GROMACS3 . 3 . 2 were used to analyze the MD results . | Introduction, Results, Discussion, Methods | The relationship between Apolipoprotein E ( ApoE ) and the aggregation processes of the amyloid β ( Aβ ) peptide has been shown to be crucial for Alzheimers disease ( AD ) ., The presence of the ApoE4 isoform is considered to be a contributing risk factor for AD ., However , the detailed molecular properties of ApoE4 interacting with the Aβ peptide are unknown , although various mechanisms have been proposed to explain the physiological and pathological role of this relationship ., Here , computer simulations have been used to investigate the process of Aβ interaction with the N-terminal domain of the human ApoE isoforms ( ApoE2 , ApoE3 and ApoE4 ) ., Molecular docking combined with molecular dynamics simulations have been undertaken to determine the Aβ peptide binding sites and the relative stability of binding to each of the ApoE isoforms ., Our results show that from the several ApoE isoforms investigated , only ApoE4 presents a misfolded intermediate when bound to Aβ ., Moreover , the initial α-helix used as the Aβ peptide model structure also becomes unstructured due to the interaction with ApoE4 ., These structural changes appear to be related to a rearrangement of the salt bridge network in ApoE4 , for which we propose a model ., It seems plausible that ApoE4 in its partially unfolded state is incapable of performing the clearance of Aβ , thereby promoting amyloid forming processes ., Hence , the proposed model can be used to identify potential drug binding sites in the ApoE4-Aβ complex , where the interaction between the two molecules can be inhibited . | Unraveling the molecular details of the interaction between apolipoprotein E and the amyloid β peptide will yield insights into the relationship between Alzheimers disease and lipid transport and metabolism ., The isoform E4 of apolipoprotein E has been shown to be closely related to Alzheimers disease ., We have therefore used a computational approach to depict a detailed interaction map for this peptide-lipoprotein interaction ., The simulation shows that the specific formation of the lipoprotein isoform E4 and the peptide complex affects the structure of the lipoprotein and the peptide ., We suggest that this is related to some of the pathogenic effects in Alzheimers disease ., Our results provide a molecular model to work with for the design of potential therapeutic agents capable of modulating this interaction . | neurological disorders/alzheimer disease, computational biology/molecular dynamics | null |
journal.pcbi.1007252 | 2,019 | On the optimal design of metabolic RNA labeling experiments | Changes in gene expression are frequently observed in pathological conditions ., In the simplest model 1 , steady state RNA levels are governed by synthesis ( transcription ) and degradation rates ( RNA stability ) ., A paradigm is the generation of the hypoxic response in pathological conditions such as heart insufficiency 2 and fast growing tumors 3 ., Hypoxia ( <2% O2 ) results in a global decrease of total transcription 4 ., However , the transcription of specific target genes is induced under hypoxic conditions by hypoxia inducible factor 1 ( HIF1 ) 5 , which is composed of a stable β-subunit and an oxygen labile α-subunit 6 ., Furthermore , different RNA binding proteins such as HuR and TTP as well as miRNAs regulate the stability of their cognate target mRNAs dependent on oxygen availability 7 and contribute to changes in gene expression profiles ., Metabolic labeling experiments are a versatile tool to discern dynamic aspects in physiological and pathological processes ., These experiments drive our understanding of key processes in molecular systems , such as synthesis and decay of metabolites , DNA , RNA and proteins ., Pulse-chase experiments help to determine the kinetic parameters of synthesis and decay in various contexts ., In the pulse phase of an experiment , the label is introduced to newly synthesized compounds and unlabeled or pre-existing molecules are only subjected to degradation or some other form of processing ., In contrast , during the chase phase , the label in the system is gradually replaced by unlabeled compounds ., A typical metabolic labeling experiment may include a pulse , a chase or both phases ., The first transcriptome-wide studies by 8 and 9 used 4-thiouridine ( 4sU ) labeling in cell culture experiments to infer kinetic parameters ., This approach has become quite popular in RNA biology , which is shown by a vastly increasing number of studies ( see 10 for review ) ., Massively parallel RNA sequencing ( RNA-seq ) in combination with metabolic labeling has become the de facto standard approach to study alterations in RNA transcription , processing or decay at the transcriptome-wide level ., At the time of writing , the most widely used approach involves metabolic labeling with thiol-labeled nucleoside analogs such as 4sU ( 4sU-tagging ) 11 ., Briefly , total cellular RNA is isolated and thiol groups are biotinylated ., Subsequently , total cellular RNA can be efficiently separated into newly transcribed ( labeled ) and pre-existing ( unlabeled ) RNA ., Very recent innovations are new methods involving the chemical conversion of 4sU residues into cytosine analogs , which is observed as point mutations in RNA-seq data ( T-to-C transitions ) , ( see 12 , 13 and 14 ) ., The absence of any biochemical separation method makes metabolic labeling more accessible due to lower input amounts and less laborious protocols ., Regardless of all advances in the experimental protocols and techniques , a few important questions remain to be answered by any experimentalist , namely the specific characteristics of experimental design: what should be measured ( i . e . sequenced ) and when ?, For example , which approach should I take ( e . g . biochemical separation vs . nucleotide conversion ) , when should I collect my samples ( e . g . time points in a pulse experiment ) and how could this affect my estimates on kinetic parameters ., In 15 , the authors proposed guidelines for the design of metabolic labeling experiments , however they provide no kinetic or statistical models for the optimization of such experiments ., Within this manuscript , we use kinetic and statistical models to infer the degradation rates from a pulse experiment ( see Fig 1 and Eqs 1 and 2 ) , and derive several aspects on the optimal design of metabolic RNA labeling experiments ., We illustrate these implications on a pulse-chase SLAMseq data set 12 and an example for a pulse labeling experiment with biochemical separation ., MCF-7 cells ( ACC-115 ) were obtained from the Leibniz Institute DSMZ German Collection of Microorganisms and Cell Cultures ., Cells were routinely tested for mycoplasma contamination with Venor GeM Classic ( Minerva Biolabs ) ., MCF-7 cells were cultured at 37°C and 5% CO2 and maintained in DMEM ( Thermo Fisher Scientific ) supplemented with 10% fetal calf serum ( Merck ) , 1xMEM non-essential amino acids ( Thermo Fisher Scientific ) and 1xPenicillin/Streptomycin ( Thermo Fisher Scientific ) ., MCF-7 cells were seeded 48 hrs prior to the experiment at a cell density of 0 . 3 × 105cells/cm2 ., Cells were labeled with 4-thiouridine ( 4sU ) ( Sigma-Aldrich ) at a final concentration of 200 μM for 2 , 4 or 8 hrs ., Cells were scraped in DPBS and the pellet resuspended in Trizol ( Thermo Fisher Scientific ) ., Total RNA was isolated using the Trizol method ., Briefly , the cell pellet was resuspended in 750 μl Trizol , and incubated 5 min at room temperature before addition of 200 μl chloroform ., Samples were centrifuged ( 20 min , 10 . 000g , room temperature ) and the aqueous phase re-extracted with one volume chloroform: isoamylalkohol ( 24:1 ) ( 5 min , 10 . 000g , room temperature ) ., The RNA in the aqueous phase was precipitated with one volume isopropanol ( 30 min , 20 . 8000g , 4°C ) , washed twice with 1 ml 80% ethanol in DEPC-H2O and dissolved in 25 μl DEPC-H2O ( 10 min , 55°C , shaking ) ., For in vitro transcription of linearized plasmids ( pBSIIKS-Luc-pA-NB 16 and pBSIIKS-Renilla-pA 17 ) , the MEGAscript T7 Transcription Kit ( Thermo Fisher Scientific ) was used according to the manufacturers instructions ., Briefly , the reaction was set up in a total volume of 20 μl containing 1 μg linearized plasmid and 2 μl 10x reaction buffer , 3 μl 40 mM m7GppG-cap analogon ( KEDAR ) , 2 μl 15 mM GTP , 2 μl 75 mM CTP , 2 μl 75 mM ATP , 2 μl enzyme mix and 2 μl 75 mM UTP ( for RLuc ) or 2 μl 75 mM 4-S-UTP:UTP in a 1:10 ratio ( for FLuc ) ., Reactions were incubated 3 hrs at 37°C ., Plasmid-DNA was removed by addition of 1 μl Turbo-DNase ( 15 min , 37°C ) ., In vitro transcribed RNA was purified by phenol extraction and Chromaspin-100 ( Clontech ) purification ., RNA was precipitated over night after addition of sodium acetate to a final concentration of 0 . 3 M and 2 . 5 volumes 100% ethanol ., After centrifugation ( 30 min , 20 . 800g , 4°C ) the pellet was washed with 1 ml 80% ethanol and dissolved in 40 μl DEPC-H2O ., Concentration was determined by Nanodrop ( Thermo Fisher Scientific ) measurement and integrity checked by agarose gel electrophoresis ., Total RNA was spiked with in vitro transcribed 4sU-labeled FLuc and non-labeled RLuc RNAs and biotinylated using MTSEA biotin-XX ( Biotium ) as described by 18 ., Briefly 80 μg total RNA was incubated with 8 ng FLuc and 4 . 8 ng RLuc ( equimolar amounts , 130 amol ) , 10 mM HEPES pH 7 . 5 , 1 mM EDTA and 5 μg MTSEA biotin-XX ( freshly dissolved in DMF ) in a total volume of 250 μl ., Reactions were incubated 30 min in the dark at room temperature ., Biotinylated RNA was recovered by extraction with one volume phenol: chloroform: isoamylalkohol ( 24:24:1 ) and separated using Phase-Lock-tubes ( 5Prime ) by centrifugation ( 5 min , 20 . 800g , room temperature ) ., RNA was precipitated by addition of 350 μl isopropanol , 25 μl 5 M sodium chloride and 1 μl glycogen ( Roche Diagnostics , 20 μg/μl ) to assist precipitation ( 30 min , 20 . 800g , 4°C ) ., RNA was washed twice with 500 μl 80% ethanol in DEPC-H2O and dissolved in 25 μl DEPC-H2O ( 10 min , 55°C , shaking ) ., For purification of biotinylated RNAs the method described by 1 was adapted ., 25 μg biotinylated total RNA was adjusted to 100 μl with DEPC-H2O and filled up with Streptavidin binding buffer ( Strep-BB ) ( 20 mM Tris , pH 7 . 4 , 0 . 5 M sodium chloride , 1 mM EDTA ) to 200 μl ., RNA was denatured 10 min at 65°C and subsequently placed on ice ., 100 μl magnetic streptavidin beads ( New England Biolabs ) were washed once with 200 μl Strep-BB and resuspended in 100 μl Strep-BB ., RNA and beads were incubated 15 min at room temperature on a rotating wheel ., Beads were washed three times with 500 μl Strep washing buffer ( 100 mM Tris pH 7 . 4 , 1 M sodium chloride , 10 mM EDTA , 0 . 1% Tween 20 ) prewarmed to 55°C ., RNA was eluted three times by de-biotinylation with 100 μl freshly prepared 100 mM DTT and elution fractions pooled for further analysis ., RNA was recovered from total RNA , flow through and eluate by phenol: chloroform: isoamylalkohol ( 24:24:1 ) extraction using Phase-Lock-tubes and isopropanol precipitation as described above ., The amount of recovered RNA was determined by Nanodrop measurement ., 1 μg biotinylated RNA was applied to nylon membrane ( Hybond-N , GE Healthcare ) using a dot blot device ( Carl Roth ) ., RNA was crosslinked twice at 254 nm using the “Optimal Crosslink” mode of the Spectroline Select XLE-1000 crosslinker ., The membrane was blocked 20 min with PBS + 10% SDS and incubated 2 hrs with Streptavidin-HRP ( Thermo Fisher Scientific , 1:5000 in PBS + 10% SDS ) ., Prior to detection with SuperSignal West Pico ( Thermo Fisher Scientific ) the membrane was washed each three times 10 min with PBS + 10% SDS , PBS + 1% SDS and PBS + 0 . 1% SDS ., Images were acquired with the LAS4000 system ( GE Healthcare ) ., 1 μl RNA from streptavidin purification was reverse transcribed using the Maxima H Minus First Strand cDNA Synthesis Kit ( Thermo Fisher Scientific ) with Random Primers according to the manufacturers protocol ., For absolute quantification reverse transcription reactions were set up with different amounts of spike in RNAs , ranging from 1600% to 1 . 56% for FLuc and 400 to 3 . 12% for RLuc in 1:2 dilutions ., Briefly , RNA was mixed in a total volume of 15 μl with 1 μl Random Primer and 1 μl dNTP solution and denatured ( 5 min , 65°C ) ., Reaction was completed by addition of 4 μl 5xRT buffer and 1 μl Maxima enzyme and incubated 10 min at room temperature followed by 30 min , 50°C and denaturation ( 5 min , 85°C ) ., Reverse transcription reactions were diluted 1:10 and used for qPCR analysis on a StepOnePlus instrument ( ThermoFisherScientific ) with Power SYBR Green PCR Master Mix ( Thermo Fisher Scientific ) and primers directed against FLuc ( forward: CCTTCCGCATAGAACTGCCT , reverse: GGTTGGTACTAGCAACGCAC 19 ) and RLuc ( forward: GTTGTGCCACATATTGAGCC , reverse: CCAAACAAGCACCCCAATCATG 20 ) ., Total and enriched samples were depleted for ribosomal RNA ( rRNA ) contamination using RiboZeroGold , which is based on the removal of rRNA with biotinylated oligos using streptavidin beads ., Thus , also the biotinylated 4sU-labeled molecules were removed from the total samples by the RiboZeroGold procedure and were treated as flow through ., Libraries of 2 biological replicate 4sU pulse experiment were sequenced 1x 50bp on an Illumina HiSeq4000 ., All relevant details on sequencing depth and mapping rates are listed in S1 Table ., Sequencing adapters and low-quality reads were removed from the raw sequencing data with flexbar v3 . 0 . 3 21 using standard filtering parameters ., We excluded all reads with more than 1 uncalled base from the output ., All remaining reads ( >18bp ) were then aligned to a custom sequence index including rRNA , tRNA and snoRNA gene loci using bowtie2 with the –very-fast option 22 ., Only reads that did not align to any of the contaminant sequences were considered for further analysis ., Reads were then aligned to the human genome ( EnsEMBL 85 ) and splice sites from the reference annotation with a splice-aware aligner ( STAR , v2 . 5 . 3a; 23 ) ., The BAM files were analyzed with StringTie 1 . 3 . 3b 24 and the final read count matrix was prepared with the supplemented python script prepDE . py ., We describe RNA-seq read counts with the negative binomial distribution , which is widely used in this setting and accounts for overdispersion 25 ., For a given gene , the read count follows X ∼ NB ( m ( μ , δ , t ) , k ) , where m is the mean read count , which depends on the time of labeling t , the degradation rate δ and the expression level in the steady-state μ , and k is the overdispersion parameter of the negative binomial distribution NB ., In this case , the variance is var ( X ) = m ( m + k ) /k , where low k values correspond to high overdispersion in the data ., We describe the RNA amount m in metabolic labeling experiments using simple first order kinetics:, d m d t = s - δ m , ( 1 ), where s is the synthesis rate and δ is the degradation rate ., In a steady-state , the expression level of a gene is μ = s/δ ., The expression level μ can be derived from the total fraction , which ensures identifiability of at least this parameter ., For that reason , we use μ and δ to parametrize the model ., In this section , we only discuss the case of pulse labeling experiments throughout ., However , our considerations extend to chase labeling experiments , where the equations are the same , except that the labeled fraction behaves as the unlabeled one in the pulse experiment and vice versa ., For simplicity , we assume that fraction cross-contamination is negligible , in which case , RNA amounts for a given gene are proportional to the means mL , mU and mT derived from the kinetics for labeled , unlabeled and total fractions scaled by sample-specific factors xi ( see Eq 4 in section 2 of Extended Methods ) :, m T ( t ) = 1 · μ m L ( t ) = x L μ ( 1 - e - δ t ) m U ( t ) = x U μ e - δ t ( 2 ) Here we treat the mean read count in the total sample as a reference ( coefficient is 1 ) , to make the system identifiable ., In the case of labeled and unlabeled fractions , expected read numbers must be scaled by additional coefficients , xU and xL , because the RNA material can be normalized by different degrees during library preparation from chemically separated fractions ., A preservation of the ratio of labeled to unlabeled fractions ( see Fig 1 ) yields xU = xL ., If the sequencing depth is approximately the same for all samples , we may assume for simplicity xU = xL = 1 , and in this case , mT ( t ) = mL ( t ) + mU ( t ) = μ ., In the conventional approach , where labeled and unlabeled molecules are separated , xU ≠ xL , the fraction ratio must be inferred from the data itself or by using an external normalization by spiking in labeled and unlabeled known molecules 26 ., In the presence of cross-contamination , the estimations for the rates are biased depending on the relation of the labeling time and the degradation rate: if δt ≪ 1 ( slow rate ) , the bias is towards faster rate values , and , if δt ≫ 1 ( fast rate ) , it is towards slower rate values , for more details see Eqs 13 and 14 , section 2 . 1 in Extended methods ., Efficiency of separation procedure may vary between species due to different uridine content , which can be another source of bias , see section 2 . 2 in Extended methods ., This phenomenon can be modeled by introducing an additional coefficient to the model , see , for example , 27 and 28 ., Although both sources of a bias may potentially affect estimates of certain RNA species , they are beyond the scope of our current work ., Here , we concentrate on theoretical results , which are derived from statistical properties of our outlined model ., In the following , we discuss pulse labeling experiments with different labeling times t ., On the one hand , subtle changes in the RNA level are masked by the measurement noise for short labeling times ., On the other hand , estimations at long labeling times are also less informative , because the difference between the steady state level and the RNA levels at time t is negligible and will be masked by the noise as well ., To estimate the degradation rate δ from the RNAseq read counts , we use the method of maximum likelihood estimation ( MLE ) ., This estimator δ ^ varies from experiment to experiment , and one is interested to minimize its variance , as a large variance results in large confidence intervals and , hence , poor estimates of the true δ ., In this paper , we use the asymptotic properties of the MLE , when the number of experiment repetitions n → ∞ , in which case the system can be treated analytically 29 , 30 ., Under regularity conditions , the MLE θ ^ is asymptotically normally distributed:, n ( θ ^ - θ ) ∼ N ( 0 , I 1 - 1 ( θ ) ) , ( 3 ), where I 1 ( θ ) is the Fisher information matrix ( FIM ) for a single experiment repetition 29 , 30 ., The FIM characterizes the curvature of the log-likelihood function L ( θ , X ) near the true parameter values θ and is defined as, I ij ( θ ) = - E ∂ 2 log L ( θ , X ) ∂ θ i ∂ θ j ., ( 4 ) We assume that the overdispersion parameter k is shared between all genes and neglect the uncertainty in δ propagating from k , i . e . only two parameters , δ and μ , are used to construct the FIM:, I ( θ ) = ( I δ δ ( θ ) I δ μ ( θ ) I δ μ ( θ ) I μ μ ( θ ) ., ) ( 5 ) The FIM is additive , i . e . if I U ( θ ) and I L ( θ ) correspond to the labeled and unlabeled fractions , the total FIM for the experiment is I ( θ ) = I U ( θ ) + I L ( θ ) , and for n such repetitions , I ( θ ) = n ( I U ( θ ) + I L ( θ ) ) ., The diagonal terms of the inverse FIM estimate the variance of θ i ^, var ( θ i ^ ) = ( I - 1 ( θ ) ) ii ., ( 6 ) In some cases we use 1 / I ii ( θ ) as a lower bound for ( I - 1 ( θ ) ) ii ., Since, ( I - 1 ( θ ) ) δ δ = ( I δ δ ( θ ) - I δ μ ( θ ) I μ δ ( θ ) / I μ μ ( θ ) ) - 1 , ( 7 ), and using the fact that I δ μ ( θ ) = I μ δ ( θ ) and I μ μ ( θ ) > 0 , the diagonal term of the inverse matrix is bounded as, ( I - 1 ( θ ) ) δ δ ⩾ 1 / I δ δ ( θ ) ., ( 8 ) ( I - 1 ( θ ) ) δ δ = 1 / I δ δ ( θ ) if there is no uncertainty , propagating from other parameters , i . e . I δ μ ( θ ) = 0 ., Since the FIM I ( θ ) depends on the experiment parameters , such as the labeling time t and the sequencing depth , it is our main interest to reduce the variance of the MLE by selecting the optimal conditions accordingly ., Due to additive property of the FIM , it suffices to optimize the FIM of a single experiment repetition ., In the case of multiple parameters , it may be not possible to achieve the minimal variance for all parameters at the same time ., Different criteria can be constructed as a combination of the elements of the inverse FIM 29 , 31 ., We are interested to optimize the estimation of δ only and do not consider variance of the expression level estimator μ ^ in the design criteria ., Let us consider first a simpler experimental setup , which preserves the fraction ratio ( e . g . SLAMseq ) ., Here we first discuss the case of the Poisson model , which corresponds to the case of no overdispersion ( k → ∞ ) ., The derivations for the Poissonian and for more general cases are left to section 3 of the Extended Methods , see Eqs 25 and 26 ., Let XL and XU be the read counts corresponding to the labeled and unlabeled molecules for a given gene in a SLAMseq sample , and let t be the time of labeling ., In this case , the inverse FIM is diagonal:, I slam - 1 ( θ ) = ( I L ( θ ) + I U ( θ ) ) - 1 = ( e δ t - 1 μ t 2 00μ ) ( 9 ) The parameters δ and μ are information orthogonal , because I δ μ ( θ ) = 0 and inference about δ can be done as μ were known exactly ., Indeed , for XL ∼ Pois ( mL ( t ) ) , XU ∼ Pois ( mU ( t ) ) , the conditional distributions P ( XL|XU + XL ) and P ( XU|XU + XL ) are binomial with the rates mU ( t ) / ( mU ( t ) + mL ( t ) ) = e−δt and mL ( t ) / ( mU ( t ) + mL ( t ) ) = 1 − e−δt and do not depend on μ ., This model was recently discussed in a Bayesian framework for SLAMseq experiments by 32 ., For a diagonal I ( θ ) , the inverse term ( I slam − 1 ( θ ) ) δ δ = ( ( I slam ( θ ) ) δ δ ) − 1 = ( ( I U ( θ ) ) δ δ + ( I L ( θ ) ) δ δ ) − 1 ., The maximum of the term ( I slam ( θ ) ) δ δ corresponds to the minimal asymptotic variance of δ ^ due to Eq 3 ., By optimizing ( I slam ( θ ) ) δ δ with respect to t , we get, t slam = 1 ., 59 τ , ( 10 ), where τ = 1/δ is the characteristic time of degradation ., That means , if one optimizes the SLAMseq experiment and targets the gene with the characteristic time of degradation τ , the measurement at time point 1 . 59τ corresponds to the asymptotically optimal design ., For example , if one is interested in an RNA species with half-life time of λ = 1 hr ( i . e . the characteristic time τ = λ/log ( 2 ) ≈ 1 . 44 hr ) , a pulse phase of 1 . 59 × 1 . 44 ≈ 2 . 3 hr corresponds to the asymptotically optimal design ., In Fig 2A , we depicted the dependency of ( I slam ( θ ) ) δ δ and corresponding values of ( I U ( θ ) ) δ δ and ( I L ( θ ) ) δ δ as functions of normalized time t/τ for the degradation rate δ = 1 . Interestingly , ( I U ( θ ) ) δ δ and ( I L ( θ ) ) δ δ achieve maximum at tU = 2τ and tL ≈ 0 . 64τ , and the main contribution to the sum ( I slam ( θ ) ) δ δ = ( I U ( θ ) ) δ δ + ( I L ( θ ) ) δ δ comes from the term corresponding to labeled counts at shorter labeling times , and from the term for unlabeled counts at times longer than τ , see Fig 2A ., Usually one is interested to measure a rate with a certain relative precision ., To reflect this , we normalize the variance of the degradation rate estimator by δ2:, var ( δ ^ ) δ 2 ≈ 1 I δ δ ( θ ) δ 2 ., ( 11 ), Using a non-dimensional substitute α = t/τ , the corresponding denominator terms are, ( I L ( θ ) ) δ δ δ 2 = α 2 μ e 2 α - e α ( I U ( θ ) ) δ δ δ 2 = α 2 e - α μ ( I slam ( θ ) ) δ δ δ 2 = α 2 μ e α - 1 , ( 12 ), see Eqs 50 , section 3 . 5 in Extended Methods ., For labeling times much shorter than the characteristic degradation time of a given gene , α ≪ 1 , the normalized FIM terms behave as a power function:, ( I slam ( θ ) ) δ δ δ 2 , ( I L ( θ ) ) δ δ δ 2 ∼ α , ( I U ( θ ) ) δ δ δ 2 ∼ α 2 ., ( 13 ), However , for labeling times much longer than the characteristic time of degradation τ , α ≫ 1 , the normalized FIM terms vanish exponentially:, ( I L ( θ ) ) δ δ δ 2 ∼ e - 2 α , ( I slam ( θ ) ) δ δ δ 2 , ( I U ( θ ) ) δ δ δ 2 ∼ e - α , ( 14 ), see derivations in Extended Methods , section 3 . 5 , Eqs 51 and 52 ., In a typical high-throughput experiment , the kinetic parameters are monitored for a large set of genes ( in the order of thousands ) , which may have different degradation rates ., In this case , every time point in the experiment will be only optimal for a subset of these genes ., To illustrate this effect , we simulated read counts for an ideal SLAMseq experiment ( with no overdispersion ) and fitted the model using various sets of samples ., In our in silico experiment , we always included the total fraction ( t = 0 hr ) , and either one additional time point ( labeled and unlabeled fractions ) or all time points ( 2 , 4 , and 8 hr ) ., The normalization coefficients were set to 1 to mimic an ideal SLAMseq scheme , as discussed earlier , Eq 2 . We fitted the model using the pulseR package and computed the 95% confidence intervals ( CI ) for δ using the profile likelihood approach 33 ., Since we assume no overdispersion ( Poisson distribution ) , for high read counts ( μ = 10000 ) the quadratic approximation of the log-likelihood function applies , and the confidence intervals for the rate estimations may be approximated by the Wald intervals , i . e . ( δ ^ − 1 . 96 ( I − 1 ( θ ) ) δ δ , δ ^ + 1 . 96 ( I − 1 ( θ ) ) δ δ ) , and hence , they reflect the behavior of the FIM term for δ ., As expected , the relative CI width is minimal only for a certain subset of the rates , depending on the set of measurements included , see ( Fig 2B ) ., If the degradation rate is very fast in comparison to the experiment time scale , the CI width for these fast genes is defined by the earliest time point in the experiment ( see Fig 2B ) ., Since every labeling time is optimal only for a single degradation rate , it might be beneficial to focus the design on genes with faster rates δ , if sample size is limited and no other criteria of optimality are given ., The justification follows from the faster decay of the FIM term for α ≫ 1 ( i . e . genes with faster kinetics ) , Eqs 13 and 14 ., Read count data from RNA-seq experiments exhibit overdispersion ( variance > mean ) , and the negative binomial distribution ( NB ) is the model of choice to account for that 25 ., In this section , we explore how overdispersion would affect MLE of δ ., The overdispersion parameter k of the NB distribution describes the level of overdispersion in the data , in which case the variance is defined as var ( X ) = m + m2/k for counts X ∼ NB ( m , k ) with mean m ., Smaller values of k correspond to higher overdispersion level , and , for k → ∞ , the NB distribution converges to the Poisson distribution , for which var ( X ) = m ., For simplicity , we assume that distributions of read counts in all samples share the same value of k ., In addition , we do not consider uncertainty in the overdispersion parameter k when we make inference about δ for individual genes , in a way as it is implemented in some packages for differential expression analysis , for example , in DESeq , 25 ., A more advanced quasi-likelihood approach , which accounts for uncertainty in the overdispersion parameter , is discussed in 34 ., In the case of NB distribution , the FIM is not diagonal for the SLAMseq experiment , see Eqs 29 and 30 in section 3 of the Extended Methods ., Hence we need to work with the inverse FIM , and the diagonal term for the SLAMseq design is, ( I slam - 1 ( θ ) ) δ δ = e δ t - 1 μ t 2 + 2 ( 1 - e - δ t ) 2 k t 2 ., ( 15 ), The presence of overdispersion shifts the optimal time to higher values ., But the most important change is that the profile of I - 1 ( θ ) δ δ is more sensitive to the labeling time t near the optimal point ., For higher overdispersion values , the variance of the rate estimator δ ^ increases faster in the vicinity of the optimum ( see Fig 2C ) ., This imposes stricter conditions on the experimental design ., The second term in the Eq 15 vanishes for times t ≫ 1 , and the equation coincides with the case of no overdispersion ., The contribution of the second term is higher for smaller values of k ( higher overdispersion ) and for shorter labeling times t , with the maximal value at t → 0:, lim t → 0 2 ( 1 - e - δ t ) 2 k t 2 = 2 δ 2 k ., ( 16 ), Another limitation , which arises in the over-dispersed model is that an increase of the sequencing depth has a limited effect on the variance ., Indeed , only the first term in Eq 15 can be eliminated by an increase of sequencing depth:, lim μ → ∞ ( I slam - 1 ( θ ) ) δ δ = 2 ( 1 - e - δ t ) 2 k t 2 ., ( 17 ), In contrast , repeating the experiment n times affects both terms in I δ δ - 1 ( θ ) , since for n repetitions ,, I - 1 ( θ ) = 1 n I 1 - 1 ( θ ) , ( 18 ), where I 1 - 1 ( θ ) is the inverse FIM for one repetition ., In the Poissonian case , when k → ∞ and the second term is absent ( see Eq 9 ) , doubling the number of samples or increasing the sequencing depth by two fold results to the same FIM and , consequently , the same approximation of the variance var ( δ ^ ) ., Standard deviation of the rate estimate is a linear function of the depth μ on the logarithmic scale and is not bounded below ( Fig 2D , dashed line ) ., In contrast , due to Eq 17 , presence of overdispersion imposes a limit , which can not be overcome by arbitrary high sequencing depth ( Fig 2D , solid line with the horizontal asymptote ) ., In essence , spreading the sequencing capacity between several biological replicates can be more beneficial than increasing the sequencing depth on a smaller number of samples ., A similar phenomenon is discussed by 35 in the context of differential gene expression analysis by RNA-seq ., If one is interested in estimating the rates of extreme values by using very short ( e . g . TT-seq , 36 ) or long labeling times , it may be less efficient to use the protocols , which preserve the ratio of labeled and unlabeled molecules ( e . g . SLAMseq ) ., Let us consider a study of fast gene kinetics , where very short labeling times are used ., In this case , δt ≪ 1 for the majority of the genes , the labeled fraction constitutes only a minor proportion of the input SLAMseq sample , because mL ( t ) = μ ( 1 − e−δt ) ≈ μδt ≪ 1 . After a short labeling time , any SLAMseq sample mainly consists of unlabeled molecules from genes with slower synthesis , which leads to spending sequencing resources on mostly non-informative material ., The same idea holds for very long times , when δt ≫ 1 and when most of the unlabeled molecules were already degraded , mU ( t ) = μe−δt ≪ 1 . In contrast , conventional experimental setups with a separation step can be used to focus sequencing capacity on the relevant molecules ., However , the conventional approach suffers from the need to normalize sequencing results from different fractions as it does not preserve the ratio of labeled and unlabeled molecules as defined by the input sample ., In typical RNA-seq experiments , the normalization coefficients are assumed to be shared between all the genes in a given sample 25 , but nevertheless , it introduces additional uncertainty into rate estimations ., As previously mentioned , a whole range of normalization approaches has been discussed in literature 26 ., In the following derivations , we neglect the uncertainty in estimating the fraction normalization coefficients xi from Eq 2 . To illustrate the benefit of the conventional approach , let us consider a set of fast turned over genes F , such that there exists labeling time t , when the majority of genes i∉F do not contribute to the labeled fractions , i . e . μ ( 1 − e−δi t ) ≪ 1 for i∉F , but μ ( 1 − e−δit ) ≈ 1 for i ∈ F . If the sequencing depth of the labeled fraction is approximately the same as for the total sample , then the normalization factor is, x L = ∑ i μ i ∑ i μ i ( 1 - e - δ i t ) ≈ ∑ i μ i ∑ i ∈ F μ i , ( 19 ), which can be high at short times ., Such “zooming” effect can be considered as corresponding increase of the sequencing depth in SLAMseq experiments by the factor of xL for the labeled fraction ., The same idea can be applied to the unlabeled fraction and long labeling times , when the sequencing depth is shared out between the most stable set of genes ., Since the normalization factor depends on the rate distribution and the expression level in a given system , it is not possible to derive the optimal design criteria analytically without imposing additional assumptions ., As in the case of SLAMseq , inference can be improved to a limited extent by increase of sequencing depth , if overdispersion is present in the data , compare to Eq 17:, lim μ → ∞ ( I L ( θ ) ) δ δ = t 2 e - 2 δ t k ( 1 - e - δ t ) 2 ⩽ k δ 2 lim μ → ∞ ( I U ( θ ) ) δ δ = t 2 k ( 20 ) For derivations , see Eqs 58 and 59 in section 4 of Extended Methods ., It is interesting to note , that for the case of the unlabeled fraction , the bound can be improved by use of longer labeling times ( provided very high sequencing depth ) , which is not the case for the labeled fraction ( with the upper bound I L ( θ ) → k / δ 2 at t → 0 ) ., In summary , biochemical separation should be considered for estimation of degradation rates of RNA species with extreme values ., Another design choice is to reduce the number of sequencing reactions by using external spike-ins ., For slowly turned over RNA species , one may sequence total and unlabeled fractions , and , for fast turned over RNA species , the total and the labeled fractions ., The use of external spike-ins ensures identifiability of the normalizing coefficient from only two fractions ., In this section , we consider a published SLAMseq pulse-chase experiment from 12 ., Here , mESCs were treated for 24 hrs with 100 μM 4sU ( pulse phase ) with samples being collected after 0 , 0 . 5 , 1 , 3 , 6 , 12 and 24 hr of label chase , and subjected to QuantSeq mRNA 3’ end sequencing ., While inspecting the data , we noticed that not all the molecules were fully labeled ( i . e . not all reads show T → C conversions ) after a 24hr pulse phase ., In this case , the labeled fraction does not reach the | Introduction, Materials and methods, Results, Discussion | Massively parallel RNA sequencing ( RNA-seq ) in combination with metabolic labeling has become the de facto standard approach to study alterations in RNA transcription , processing or decay ., Regardless of advances in the experimental protocols and techniques , every experimentalist needs to specify the key aspects of experimental design: For example , which protocol should be used ( biochemical separation vs . nucleotide conversion ) and what is the optimal labeling time ?, In this work , we provide approximate answers to these questions using the asymptotic theory of optimal design ., Specifically , we investigate , how the variance of degradation rate estimates depends on the time and derive the optimal time for any given degradation rate ., Subsequently , we show that an increase in sample numbers should be preferred over an increase in sequencing depth ., Lastly , we provide some guidance on use cases when laborious biochemical separation outcompetes recent nucleotide conversion based methods ( such as SLAMseq ) and show , how inefficient conversion influences the precision of estimates ., Code and documentation can be found at https://github . com/dieterich-lab/DesignMetabolicRNAlabeling . | Massively parallel RNA sequencing ( RNA-seq ) in combination with metabolic labeling has become the de facto standard approach to study alterations in RNA transcription , processing or decay ., In our manuscript , we address several key aspects of experimental design:, 1 ) The optimal labeling time ,, 2 ) the number of replicate samples over sequencing depth and, 3 ) the choice of experimental protocol ., We provide approximate answers to these questions using asymptotic theory of optimal design . | sequencing techniques, nucleic acid synthesis, experimental design, rna extraction, nucleotides, research design, molecular biology techniques, rna synthesis, rna sequencing, extraction techniques, rna transcription labeling, chemical synthesis, research and analysis methods, cell labeling, proteins, molecular biology, biotinylation, biosynthetic techniques, metabolic labeling, biochemistry, rna, post-translational modification, nucleic acids, biology and life sciences, nucleic acid labeling | null |
journal.pgen.0040012 | 2,008 | Dominant-Negative CK2α Induces Potent Effects on Circadian Rhythmicity | Circadian rhythms that orchestrate daily fluctuations in biochemistry , physiology and behavior are observed across distinct phylogenetic kingdoms ., Underscoring the evolutionary importance of these clocks , the molecular processes that drive circadian rhythms are also highly conserved ., At the core of the circadian pathway is a transcriptional feedback loop ., In Drosophila melanogaster , CLOCK ( CLK ) and CYCLE ( CYC ) activate expression of target genes such as period ( per ) and timeless ( tim ) 1–3 ., PER and TIM ultimately translocate to the nucleus and inhibit CLK/CYC transcription 3–6 ., Notably , the overall architecture of this feedback loop , as well as some of the molecular players , are observed in organisms as diverse as fungi , plants , and mammals 3 ., In addition to transcriptional influence in circadian rhythms , posttranslational modification , particularly for PER , has been shown to play a critical role in normal and disordered circadian timing 7–10 ., The most well studied kinase CK1/DOUBLETIME ( DBT ) is hypothesized to regulate PER nuclear entry , repression , and degradation 7 , 11 , 12 ., A second enzyme , glycogen synthase kinase ( GSK3β ) /SHAGGY ( SGG ) , regulates phosphorylation of TIM protein , levels , and nuclear entry 13 ., These rhythmic phosphorylation cycles also necessarily include the activity of a phosphatase , and protein phosphatase 2A ( PP2A ) has been implicated in the rhythmic dephosphorylation of PER 14 , while protein phosphatase 1 ( PP1 ) has been implicated in the dephosphorylation of both TIM and PER 15 ., Our laboratory has been investigating the function of the protein kinase CK2 in circadian clock function 16 , 17 ., The CK2 holoenzyme is a heterotetramer consisting of two alpha catalytic and two beta regulatory subunits 18 , 19 ., Mutant CK2α and CK2β alleles result in period lengthening phenotypes ( <3 h long ) , consistent with their proposed clock role 16 , 20 ., The manner in which CK2 is important for setting circadian period remains unclear ., CK2 also functions in various developmental processes 18 , consistent with the pre-adult lethality observed in CK2α and CK2β mutants 16 , 21 ., This developmental function raises the question that CK2 phenotypes may derive from its activity during maturation rather than in adults ., While both CK2 subunits are expressed in pacemaker neurons 16 , 20 , it is uncertain if CK2 functions in these neurons to regulate circadian rhythms ., RNAi studies in S2 cells suggest that the role of CK2 phosphorylation is to promote transcriptional repression by PER 22; however , it is not clear if this is true in vivo ., To better address these questions , we expressed a dominant negative CK2α Timekeeper ( Tik ) mutant 16 in a spatially and temporally controlled manner and queried the effects on behavior , PER protein levels , phosphorylation , repression , and nuclear entry in core pacemaker neurons of adult D . melanogaster ., Taken together , these findings reveal remarkably potent effects of manipulating CK2 activity in adult circadian neurons and uncover a role consistent with the regulation of PER nuclear localization and feedback repression ., Prior studies implicate CK2 in the control of circadian function in Drosophila , Arabidopsis , and Neurospora 16 , 20 , 23 , 24 ., Testing of the strongest homozygous mutants alleles is limited by developmental lethality 16 , 21 ., More modest period phenotypes raised questions as to the functional significance of CK2 action in the circadian clocks ., To determine the consequences of suppressing CK2 activity , we used the GAL4/UAS system to drive expression of CK2α bearing the dominant Tik mutation ( CK2αTik ) 25 ., The CK2αTik allele contains two missense mutations , one of which introduces a charged residue into the putative hydrophobic binding pocket for the phosphodonor nucleotide 16 , 19 ., In vitro analysis indicates that these mutations eliminate most catalytic activity 26 ., The molecular lesion , the loss of biochemical activity and the dominant behavioral phenotype suggest that Tik encodes a dominant negative form of CK2α ., To examine the behavioral consequences of CK2αTik expression , we crossed flies bearing UAS-driven CK2αTik ( UASTik ) with timGal4–62 driver flies 27 and assayed circadian behavior in the progeny ( timGal4/+; UASTikT1/+ , “timTik” ) ., The Drosophila circadian network consists of six bilateral groups of cells: large and small ventral lateral neurons ( lg- and sm- LNv ) , dorsal lateral neurons ( LNd ) , and three clusters of dorsal neurons ( DN1–3 ) 28 ., The tim promoter induces GAL4 expression in all of these key neuronal clusters that coordinate circadian behavior 29 ., To our surprise , these timTik flies display extraordinarily long periods averaging ∼33 h relative to control periods of ∼24 h ( Figure 1 , compare Figure 1A and Figure 1B; Table 1 ) ., Moreover , the influence on period is dose-dependent; by increasing Gal4 dosage in timTik flies with a second circadian driver , cry16Gal4 30 , the period is further lengthened to ∼37 h ( Table 1 ) ., Confirming the circadian specificity of this result , expression of UASTik only in photoreceptor neurons with the GMRGal4 driver 31 does not result in period lengthening ( data not shown ) ., Heterozygous Tik/+ mutant flies display periods 2–3 h longer than wild-type controls with a reduction of ∼50% in CK2 activity 16 ., The magnitude of the period effects strongly argues that CK2 activity is more gravely inhibited in timTik flies ., The fact that the magnitude of period effects exceeds that of nearly all circadian mutant alleles suggests that CK2 activity is critically important for setting circadian period ., By increasing dosage of the dominant allele with double copies of both the broad circadian timGal4 driver and the UASTik transgene ( timGal4; UASTikT1 , “timTik2x” ) , rhythmicity is undetectable in constant darkness ( Figure 1C; Table 1 ) ., The above results suggest that CK2α function in central pacemaker neurons is essential for wild type behavioral rhythms ., Thus , CK2α and DBT appear to be the only core circadian kinases demonstrated to be obligatory for free-running behavioral rhythms 11 , 32 ., Mutations in a catalytic subunit of the cAMP-dependent protein kinase ( PKA ) also result in behavioral arrhythmicity 33; however , as this lesion leaves core molecular cycling of the clock intact , it is likely to function in an output capacity ., The neuropeptide Pigment-Dispersing Factor ( PDF ) mediates transmission of timing information from core LNv pacemaker neurons to downstream neural circuits 34 ., The CK2α and β subunits are strongly expressed in the pacemaker LNv 16 , 20 ., To test the hypothesis that CK2α functions in pacemaker neurons , CK2αTik was induced in the LNv using a pdfGal4 driver 34 ., Similar to timTik flies , CK2αTik expression in PDF+ neurons ( pdfGal4/+; UASTikT1/+ , “pdfTik” ) also results in dramatically long periods ( ∼32 h; Figure 1D; Table 1 ) ., Again , these effects are dose-dependent , as adding an additional Gal4 driver , cry16Gal4 , increased the period length to ∼37 h ( Table 1 ) ., We previously identified a spontaneous revertant allele , TikR , which deletes a portion of the Tik coding region , largely reverts the dominant circadian phenotype but still lacks catalytic activity , consistent with its characterization as a recessive loss-of-function allele 16 ., Supporting this hypothesis , pdfGal4 expression of independent UASTikR lines had no significant effect on circadian rhythms ( Table 1 ) ., These results confirm the hypothesis that the Tik mutation acts as a dominant-negative to inhibit function of endogenous wild type CK2α ., PDF neurons communicate with and reset the clocks in non-PDF pacemaker neurons to synchronize different clusters in the network 35 ., The CK2αTik period effects were blocked in a pdf null 34 background or by coexpressing an inwardly rectifying potassium channel that hyperpolarizes the LNv ( UASKir2 . 1 , 36 , Table 1 ) , indicating that CK2αTik period effects are transmitted by LNv activity and PDF output ., These manipulations alone ( pdf01 null mutants or expression of UASKIR with pdfGal4 ) result in short , weak periods ( Figure S1; Table 1 ) ( 34 , 36 ) ., These data provide functional evidence that CK2α operates in pacemaker LNv to regulate circadian period , consistent with published expression data ., While pdfTik flies show a long period phenotype , we also noted variability in the period measurement and reduction of the strength of the rhythm in these flies ( Table 1 ) ., It was hypothesized that “wild-type” non-PDF clock neurons were unable to entrain to the long period program in PDF+ cells , and were expressing a secondary rhythm ., To see if flies were exhibiting more than one period , we performed periodogram analysis using the Lomb-Scargle method 37 , 38 ., This approach eliminates misidentification of periods that are simply multiples of a true period ., This analysis reveals that approximately 45% of pdfTik animals display two significant periods ( Figures 1E and 2A ) ., The dominant period is 35 . 3 ( +/− 0 . 3 ) h while a secondary peak indicates an average period of 23 . 2 ( +/− 0 . 1 ) h ( Figures 2B and S2 ) ., When UASTik is expressed as a heterozygote with the broader expressing timGal4 driver , reduced rhythm strength and splitting is not detectable ( Figure 2A; Table 1 ) , suggesting that hyper-elongating period only in PDF-positive LNv causes uncoupling of clock cell groups ., We propose that non-PDF neurons are unable to maintain synchrony with PDF clocks with extreme periods ., To our knowledge , this is the first example of complex rhythmicity due to altering period length in a subset of pacemaker neurons ., Given that CK2 acts in multiple pathways throughout the life cycle of the fly , we queried if the CK2αTik phenotype is due to developmental/compensatory effects or whether loss of CK2α function during adulthood would still result in lengthening of period ., In order to address this concern , we utilized a temperature sensitive Gal80 inhibitor of Gal4 expressed in all cells under the tubulin promoter ( tubGal80ts , 39 ) ., This conditional , temporal and regional gene expression targeting ( TARGET ) system has been previously used to examine the genetic basis of complex behaviors such as memory in Drosophila 39 ., tubGal80ts represses GAL4 at the permissive temperature of 18 °C , but is inactivated and fails to repress GAL4 at the restrictive 29 °C temperature ., We generated pdfGal4/tubGal80ts; UASTikT1/+ flies and raised them at the permissive temperature ( 18 °C ) to prevent expression of UASTik so that CK2α function would be largely intact during development ., Flies were then tested at either permissive ( 18 °C ) or restrictive ( 29 °C ) temperatures and period was calculated during constant conditions ., A cardinal feature of circadian clocks is their temperature compensation , i . e . , period is roughly invariant over a broad temperature range 40 , 41 ., Consistent with this idea , the control strain here ( tubGal80ts/+; UASTikT1/+ ) shows little period change between 18 °C and 29 °C ( Figure 3A and 3B , top panels ) ., Constitutively inhibiting CK2α in pdfTik flies again demonstrates the severe period lengthening effect at both temperatures ( Figure 3A and 3B , middle panels ) ; splitting of rhythms in these flies is observed at levels similar to those described above , but only at 29 °C ( unpublished data ) ., Interestingly , when dominant-negative UASTik is selectively activated at 29 °C during testing of adult flies , the extreme long period phenotype ( >30 h ) is still manifested ( Figure 3A and 3B , bottom panels ) ., Slight period lengthening is detectable at 18 °C; however , this effect is much smaller than observed at 29 °C , and is likely due to incomplete Gal80 inhibition of Gal4 ., Additionally , two split periods are again observed at 29 °C in conditionally inhibited circadian CK2α flies , but not at 18 °C ( data not shown ) ., These results indicate that CK2α plays a direct role in adult circadian rhythms , and its loss of function in Tik and UASTik animals is not likely due to some developmental artifact ., Consistent with this idea , inspection of LNv structure and PDF labeling in UASTik-expressing brains reveals no gross abnormalities of circadian pacemaker anatomy ( unpublished data ) ., To our knowledge , this is one of the few temporal investigations of clock gene function demonstrating an acute role of a circadian gene during adulthood 42 , 43 ., To determine the effects of CK2α loss of function on core molecular clock rhythms , we tested whether expression of UASTik in PDF-positive LNv altered cycling of the core clock protein PER ., Levels and cellular distribution of PER protein in smLNv were examined quantitatively on the first day of DD in pdfTik or control Gal4 flies ., Although we do observe splitting in these flies , behavior remains largely synchronous on the first day of DD ( Figure 4A ) ., Control flies show the typical evening peak of activity at ∼CT12 while the long-period pdfTik flies have a delayed evening activity peak , regardless of whether they exhibit split periods or not ( Figure 4A , pdfTikL v . pdfTikS ) ., Measurements of pixel intensity indirectly report the amount of PER protein in smLNv 44; as seen in Figure 4B and 4C , PER levels are elevated in smLNv of pdfTik during the subjective day relative to controls ., Wild type PER levels wane from CT4–8 and begin accumulating again in the subjective evening ( CT16–20 ) ; in contrast , a prolonged decline in PER throughout the day ( CT4–12 ) is evident in pdfTik flies , and levels only disappear during subjective evening ( CT12–20 ) , consistent with a long period phenotype ., Peak and trough PER levels are also elevated in pdfTik flies relative to controls ( p < 0 . 001 comparing pdfGal4/+ CT0 to pdfTik CT4 for peak and pdfGal4/+ CT12 to pdfTik CT16 for trough , Figure 4C ) ., PER typically transitions from the cytoplasm to a predominantly nuclear distribution during the middle of the night , and such a pattern is observed in pdfGal4/+ control flies ( Figure 4B and 4D ) ., However , the amplitude of the localization rhythm ( as quantified by the nuclear:cytoplasmic ratio ) is seriously reduced in pdfTik flies ( Figure 4D , p < 0 . 001 at CT0 , CT4 , CT8 , and CT20 pdfGal4/+ v . pdfTik ) ., The timing of nuclear localization is also delayed in pdfTik flies; while PER never becomes predominantly nuclear , the time at which the most PER is localized to the nucleus occurs later from CT4–12 in pdfTik smLNv , rather than CT0–4 for the GAL4 control ( Figure 4D and 4E ) ., This finding is supported by analysis of nuclear PER levels in pdfGal4/+ and pdfTik smLNv ., Nuclear PER levels accumulate to a similar degree in pdfTik as in the Gal4 control; however , nuclear levels do not rise until later in the subjective day relative to control ( Figure 4E ) ., The overall fraction of nuclear PER is lower ( Figure 4D ) , as more of the PER protein in pdfTik neurons remains sequestered in the cytoplasm ( Figure 4F ) ., Indeed , the reduced nuclear PER levels in the face of elevated cytoplasmic PER levels at CT0 provide the most compelling evidence that CK2α is important for nuclear PER localization independent of regulating its cytoplasmic abundance ., These results are consistent with prior reports that reduction of CK2 activity inhibits nuclear entry 16 ., To quantitatively examine the effect of CK2αTik on PER cycling and phosphorylation , we used western blots of whole head extracts on the first day of DD ., The far majority of PER in whole heads is expressed in the eye 45 ., To examine CK2αTik effects we used the timGAL4 driver that includes strong expression in the eye ., In wild-type flies , PER phosphorylation ( evident as reduced mobility ) peaks in the early subjective morning ( CT1 , Figure 5Aa ) ., PER levels are subsequently reduced , presumably reflecting phosphorylation-induced degradation ., PER begins to appear early in the subjective night ( CT13 , Figure 5Aa ) , and levels accumulate during the night as PER becomes progressively more phosphorylated ., In Tik/+ and timTik flies , both level and mobility rhythms are delayed , consistent with a lengthened period in these flies ( Figure 5Ac , 5Ad , and 5B ) , while expression of UASTikR was not detectably different than wild type ( Figure 5Ab ) ., We then examined flies with two copies of the timGAL4 and UASTik transgenes ( timTik2x ) ., These flies did not exhibit any significant behavioral rhythms ( Table 1 ) ., Severe reduction of CK2α activity in homozygous timTik2x flies exacerbates PER metabolism during constant conditions , causing constitutive elevations in PER protein and minimizing the amplitude of PER cycling ( Figure 5Ae and 5B ) , consistent with reduced behavioral rhythmicity ., These effects are most evident at wild type trough times for PER ( p < 0 . 01 , significant effect of genotype at CT9 ) ., In addition , PER fails to achieve a hyperphosphorylated state in timTik2x flies ( Figure 5Ae ) ., This consequence is most evident at CT1 , time of peak phosphorylation in wild type ., These findings support the notion that CK2α ensures the proper timing of PER cycling and function ., While we cannot exclude the possibility that CK2 indirectly regulates the post-translational modification of PER , the strong effects on PER mobility in CK2 loss-of-function flies argue that PER is an in vivo CK2 substrate , consistent with previous studies 17 ., Previous studies have implicated CK2 in promoting PER repression of CLK activation 22 ., However , these studies were performed in cultured Drosophila S2 cells which do not harbor functioning circadian clocks ., To test the hypothesis that CK2 promotes PER repression in vivo , we examined circadian transcription in UASTik expressing flies ., Levels of two CLK-activated transcripts , per and vrille ( vri ) 46 were analyzed using quantitative real-time reverse-transcriptase polymerase chain reaction ( qRT-PCR ) ., We hypothesized that if negative feedback is unaffected in UASTik expressing flies , then elevated PER levels would strongly repress CLK , reducing per and vri transcript levels ., If negative feedback is disrupted , then elevated PER levels would fail to appropriately repress per and vri transcription ., Expression of dominant-negative UASTik in circadian neurons in timTik flies postpones the decline in per transcript until early subjective night ( Figure 6A ) , consistent with the effect of CK2α loss of function in the heterozygous Tik/+ mutant ., Whereas wild type per transcription peaks around CT9–13 , per levels do not achieve maximum until CT13–17 in timTik flies , and a similar pattern emerges from analysis of the vri transcript ( Figure 6B ) ., The most informative result becomes apparent in timTik2x flies ., Further reductions in CK2α activity in timTik2x result in per and vri transcript levels with a severely reduced amplitude rhythm ( Figure 6A and 6B ) ., Importantly , per and vri never reach wild-type trough levels ( per: p < 0 . 001 for y w at CT1 relative to timTik and timTik2x at CT5 and p < 0 . 01 for vri at the same time points ) , consistent with the hypothesis that elevated PER protein levels are unable to fully repress CLK target genes in UASTik-expressing flies ., Taken together , the magnitude of the observed effects suggests that CK2 not only promotes PER repression activity in vivo , but that it has a sizable impact on transcriptional repression ., The role of posttranslational modification in regulating precise circadian timing is well established 9 , and indeed may be principally responsible for molecular cycling 8 ., CK2 has been implicated in regulating circadian rhythms , PER modification and metabolism 16 , 17 ., The present study sought to determine if CK2α activity is required in adult core pacemaker neurons for molecular and behavioral rhythmicity ., Broad spatial expression of UASTik in pacemaker neurons with the timGal4 driver causes severe lengthening of circadian period to ∼33 h , a degree even greater than that of the heterozygous Tik mutant ., Radical reductions in CK2α activity by increasing copy number of the transgenes in timTik2x flies ultimately result in behavioral arrhythmicity , demonstrating that CK2α is an obligatory component of circadian rhythms ., Previous work demonstrated that overexpression of wild type CK2 mildly lengthens period 17; taken together , these data indicate that period is highly sensitive to CK2 activity ., Expression of UASTik in PDF+ LNv is also sufficient to lengthen period; indeed , the effect requires LNv activity and output of the PDF neuropeptide ., That the period length is not exacerbated by additional clock neuron expression in timTik versus pdfTik flies implies that the phenotype originates largely from the LNv; however , the possibility that CK2α additionally functions in other circadian cells cannot be excluded ., Given that splitting is eliminated when the genetic programs of both LNv and downstream circadian neurons are identical with respect to UASTik expression , the data imply that this manipulation affects CK2 function in other clock cells ., Indeed , if CK2 is a true component of the core transcriptional pacemaker , it would be expected to regulate feedback in all cells that have a functional molecular clock ., As CK2 functions to dictate period in LNv cells during constant conditions , it is possible that CK2 activity may also regulate morning behavior , as these cells drive morning activity , while downstream neurons dictate evening activity 47 , 48 ., However , as CK2 may also function in cells responsible for evening behavior , some balance of CK2 between LNv and non-LNv neurons could favor a strong morning or evening activity phase ., The contribution of CK2 activity to morning and evening behavior is currently under investigation ., Further evidence that CK2α activity is important in LNv derives from the finding that inhibition by CK2αTik causes delayed nuclear entry of PER in these core pacemaker cells ., An unanticipated consequence of pdfTik expression is splitting of the behavioral rhythm into long ( ∼35 h ) and short ( ∼23 h ) components ., All of the above behavioral effects are due to acute CK2α activity as adult-specific inhibition is able to induce the rhythm phenotypes ., At the molecular level , elevated levels and diminished phosphorylation of PER protein is associated with reduced CK2α function; this effect on PER protein is further correlated with elevated and delayed transcription of per and vri clock genes ., The severity of the observed behavioral phenotype places CK2 as a critical regulator of circadian rhythms ., Of known circadian kinases , only mutants of doubletime ( dbt ) and PKA are also capable of completely eliminating rhythmicity as is observed in timTik2x flies 32 , 33 ., As the core molecular clock is unperturbed by PKA mutations , this kinase is proposed to function in circadian locomotor output 32 , 33 , leaving DBT and CK2 as the only critical core circadian kinases ., Originally , DBT was found to regulate PER stability and electrophoretic mobility; this initial study concluded that DBT-mediated phosphorylation led to PER degradation 11 , 49 ., Subsequent studies suggested that DBT may retard the ability of PER to enter the nucleus and repress transcription 12 , 22 , 50 ., Many other gene mutations that result in arrhythmicity affect either input or output of the circadian system ., As the core molecular feedback loop is disrupted in UASTik-expressing flies , CK2 appears to regulate timing of the core clock and shows phenotypes similar to mutants of other core circadian genes such as per , tim , and Clk ., However , the magnitude of the period phenotype in both pdfTik and timTik flies is greater than nearly all circadian mutants ., The only other alleles which produce a similar degree of period lengthening include the timUL mutation 51 and a novel dominant-negative kinase dead dbt allele whose expression also results in period lengthening or arrythmicity 52 ., We present numerous pieces of evidence to support the hypothesis that CK2α acutely functions in the PDF+ LNv neurons ., The long period phenotype observed when CK2αTik is expressed in PDF-positive LNv is associated with splitting of the behavioral rhythm into two components: a predominant , long , ∼35 h period and a weak shorter period of approximately 23 h ., The splitting is reflected in the low strength of behavioral rhythms observed in pdfTik flies ., Splitting was originally observed in Syrian hamsters maintained under constant light; this finding was the foundation for a two-oscillator model whose coordinated output is manifested as an overt circadian rhythm 53 ., It has similarly been shown that non-conventional entrainment conditions can induce multi-period splitting and desynchronization of circadian neurons in mammals 54 , 55 ., Early reports indicated splitting of the Drosophila circadian period in sine oculis mutants that have disrupted optic development , suggesting that dual periods may arise from entrainment through different input pathways 56 ., Similarly , both wild type flies under low light and mutants of cryptochrome , the major circadian photoreceptor , exhibit split rhythms under constant light 57 , 58 ., These periods include short ( ∼22 h ) and long ( ∼25 h ) components that alternatively decrease or increase with light intensity , respectively , again implicating variation of the oscillator system input pathway ., Ectopic misexpression of the PDF output neuropeptide induced multiple periods during DD ( of ∼22 h and ∼25 h ) 59 ., Nitabach et al . 38 further identified complex rhythmicity by activating LNv neurons; at least two periods of ∼22 or ∼25–26 h lengths are observed ( with an occasional 3rd , shorter ∼20–21 h peak ) ., Elevated PDF levels and desynchronization of circadian neurons are detected in these flies 38 ., Both of the above cases suggest that split periods arise from misregulation of neuronal output from the core pacemaker neurons ., The result presented here is the first demonstration of splitting as a consequence of altering a core clock component ., It is hypothesized that driving the oscillator period to such an extreme only in the LNv uncouples them from non-UASTik expressing , PDF-negative “wild type” circadian neurons ( i . e . , LNd/DNs ) that then contribute the shorter , weaker behavioral rhythm ( ∼23 h ) , such as that seen in pdf01 mutants 34 ., This notion is further supported by the behavior of timTik flies that express UASTik in all circadian neurons; in this case , when the genetic programs of all clock cells are identical , no such splitting is observed and rhythm strength returns to normal levels ., These results begin to examine the limits of entrainment of one oscillator by a coupled oscillator in a circadian pacemaker network ., CK2 has a number of roles in cellular biology 18 ., It is required at multiple transitions during the cell cycle including mitosis and functions to regulate caspase-mediated apoptosis and cell survival 18 ., Developmentally , CK2 regulates proliferation and cell fate decisions 25 , 60 ., Not surprisingly , it is an essential gene , as homozygous Tik mutants are not viable as adults 16 ., We were able to utilize the TARGET system 39 to conditionally induce dominant-negative CK2α in adult flies ., Interestingly , when CK2α activity was inhibited in LNv solely during adulthood , the behavioral phenotypes are still manifested ., Thus , an acute CK2α loss of function impacts rhythmicity in the adult circadian system , presenting it as a critical and direct regulator of the circadian clock ., While it has been shown that such adult-specific rescue of per is able to restore rhythmicity in Drosophila 43 , it will be important to investigate the life-stage properties of other circadian genes; for example , Clk is also known to have developmental roles 61 ., Thus , CK2α is an acute , direct , and essential component of circadian rhythms; we propose that CK2 regulates the core oscillator by phosphorylating PER to promote nuclear entry and repression ., There is abundant evidence for PER as a bona fide CK2α substrate ., CK2α can phosphorylate PER in vitro at specific predicted CK2α sites 16 , 17; moreover , mutation of these CK2α target residues causes period phenotypes similar to the Tik mutation when expressed in vivo , demonstrating the functional relevance of this modification on PER activity 17 ., Finally , we present here the clear defects in PER mobility observed with a deficiency in CK2α activity ., Expressing UASTik singly or in double dosage with the timGal4 driver results in increased , hypophosphorylated PER ., Indeed , the amplitude of PER cycling appears completely diminished when CK2α is severely inhibited in timTik2x flies ., Again , the molecular PER phenotype mirrors the behavioral effect of these manipulations; timTik2x flies are arrhythmic under constant conditions , supporting the idea that CK2α activity is critical for the maintenance of a molecular and behavioral clock ., A further consequence of CK2α loss of function in core pacemaker neurons is a pattern of delayed PER decline , consistent with the long period phenotype observed in these flies ., Despite increases in overall and cytoplasmic PER levels , nuclear PER levels are lower relative to wild type during the early subjective day , providing further evidence that nuclear translocation is not strictly driven by protein accumulation 50 , 62 ., The dampened and delayed nuclear entry of PER protein of CK2αTik-expressing smLNv provides support that CK2α normally functions to promote PER nuclear translocation ., A second possibility is that the high levels of PER protein saturate the nuclear entry pathway , preventing the majority of PER from localizing to the nucleus in pdfTik flies ., Yet , the delay in nuclear accumulation is consistent with the hypothesis that CK2α activity typically functions to permit timely PER nuclear entry ., Previous evidence indicated that knock-down of CK2 levels in cultured Drosophila S2 cells limits the ability of PER to repress a Clk-driven luciferase reporter 22 ., It is critical to validate such studies in vivo to determine the true function of the kinase in the circadian system ., While one may expect that the increased levels of PER associated with CK2αTik expression ( particularly at trough time points ) would lead to enhanced clock gene repression , we do not see such an effect ., Conversely , CK2α inhibition results in delayed per and vri transcription , and elevated trough transcript levels , confirming that CK2α normally operates to promote repression of clock gene transcription ., The features of CK2α function are both in opposition with and complementary to those put forth for the DBT kinase ., DBT is thought to retard PER nuclear entry 12 and signal its degradation 11; in contrast , CK2α appears to promote nuclear entry of PER ( and hence repression ) , but may also influence its turnover ., We have outlined a model of the way in which CK2 promotes repression of circadian transcription developed from existing and currently presented data ., We speculate that effects of CK2 on PER nuclear localization may operate through the proposed interval timer described in S2 cells ., Based on the interval timer model , PER and TIM heterodimerize in the cytoplasm in a time-insensitive manner 63; after some lag or upon some signal , they dissociate and enter the nucleus independently 44 , 63 where PER mediates transcriptional repression ., The role of nuclear TIM is yet unclear ., Recent work indicates that repression is not achieved merely by physical association of PER with CLK , but perhaps by PER acting as a scaffold to bridge CLK and DBT 64 ., It is hypothesized that phosphorylation of CLK by DBT diminishes its transactivating capabilities 64 , similar to the model proposed in Neurospora 23 ., Nawathe | Introduction, Results, Discussion, Materials and Methods, Supporting Information | Circadian clocks organize the precise timing of cellular and behavioral events ., In Drosophila , circadian clocks consist of negative feedback loops in which the clock component PERIOD ( PER ) represses its own transcription ., PER phosphorylation is a critical step in timing the onset and termination of this feedback ., The protein kinase CK2 has been linked to circadian timing , but the importance of this contribution is unclear; it is not certain where and when CK2 acts to regulate circadian rhythms ., To determine its temporal and spatial functions , a dominant negative mutant of the catalytic alpha subunit , CK2αTik , was targeted to circadian neurons ., Behaviorally , CK2αTik induces severe period lengthening ( ∼33 h ) , greater than nearly all known circadian mutant alleles , and abolishes detectable free-running behavioral rhythmicity at high levels of expression ., CK2αTik , when targeted to a subset of pacemaker neurons , generates period splitting , resulting in flies exhibiting both long and near 24-h periods ., These behavioral effects are evident even when CK2αTik expression is induced only during adulthood , implicating an acute role for CK2α function in circadian rhythms ., CK2αTik expression results in reduced PER phosphorylation , delayed nuclear entry , and dampened cycling with elevated trough levels of PER ., Heightened trough levels of per transcript accompany increased protein levels , suggesting that CK2αTik disturbs negative feedback of PER on its own transcription ., Taken together , these in vivo data implicate a central role of CK2α function in timing PER negative feedback in adult circadian neurons . | The molecular mechanism that governs organization of physiology and behavior into 24-h rhythms is a conserved transcriptional feedback process that is strikingly similar across distinct phyla ., Notably , cyclic phosphorylation of negative feedback regulators is critical to time molecular rhythms ., Indeed , mutation of a putative phosphoacceptor site in the human PERIOD2 gene , a key negative regulator , is associated with Advanced Sleep Phase Syndrome ., This study reveals a critical role for the protein kinase CK2 for setting the period of behavioral and molecular oscillations in Drosophila ., Circadian phenotypes due to CK2 disruption are due to a direct requirement in adult circadian pacemakers ., These findings further demonstrate that CK2 modification of the negative feedback regulator PERIOD alters its cyclical phosphorylation , protein abundance , nuclear translocation , and transcriptional repression activity ., These studies place CK2 as a central kinase in circadian timing . | neuroscience, drosophila, genetics and genomics | null |
journal.pcbi.1000842 | 2,010 | Reliability of Transcriptional Cycles and the Yeast Cell-Cycle Oscillator | Cells have to operate reliably under internal and external noise in order to survive ., Their robustness is partially a result of various signal-processing sub-networks called “motifs , ” embedded in the transcriptional network of the cell that controls gene expression 1–5 ., Such motifs are employed by the cell to produce reliable responses to internal and external signals: a negative auto-regulation motif decreases response time and increases robustness to noise 3 , 6 , 7; a positive feedback generates bistability and thus can act as a switch 8–10; a coherent feed-forward loop with OR logic acts like a capacitor , sustaining a high output when the input signal is transiently lost 11; and an incoherent feed-forward loop allows adaptation to a sustained input signal 12 ., It is known that combinations of some motifs such as positive and negative feedback loops , can generate stable cyclic behavior 1 , 10 , 13–19 ., The exact mechanism underlying the oscillations may vary 20–22 ., Two examples have been particularly well studied ., In a negative feedback oscillator ( ) , a sufficiently long time delay in the negative feedback loop makes the system repeatedly overshoot an unstable steady state 10 , 13 ., In an activator-inhibitor oscillator ( ) , a positive feedback loop creates bistability and a negative feedback loop causes oscillations due to hysteresis 10 , 13 , 15 , 16 ., An important feature in these examples is the spontaneous activation of , which is required to avoid collapse to a quiescent state ., In a transcriptional oscillator , this corresponds to a constant input signal ( due , for example , to a constitutive promoter ) or positive auto-regulation sufficiently strong to cause levels of to rise to an active state as long as the inhibitor is not present ., To our knowledge , all models of biological oscillatory networks described in the literature , such as cyclin-cdc2 oscillations 23 , 24 , or circadian oscillations in Drosophila 25 , require spontaneous activation to sustain the oscillations 1 , 13 , 20 , 21 ., This is also true for synthetic examples such as the repressilator 26 , ( in which all three genes have constitutive but repressible promoters ) , the E . coli predator-prey system 27 , and the synthetic gene-metabolic oscillator 28 ., The recently published transcriptional yeast ( Saccharomyces cerevisiae ) cell-cycle oscillator 29 , however , does not seem to share this feature ., The gene expression data suggest that this oscillator relies mainly on a sequence of activations on a long , slow positive feedback loop 29–31 ., There does not appear to be an element in this transcriptional network that is activated spontaneously ., Expression profiles also indicate that the period of the oscillator is very close , if not identical , to the time it takes for the wave of activations to cycle around the long positive feedback loop ., Here , we show how it is possible to maintain stable oscillations within this architecture ., We demonstrate that a slow positive feedback loop coupled to certain stabilizing motifs can sustain oscillations , and that a model of the transcriptional oscillator associated with the yeast cell-cycle works in this fashion ., Oscillator stability is conventionally studied in the context of a differential equation model 20 , 21 ., On the other hand , the essential organizing logic of regulatory networks can be studied much more easily using Boolean models 29 , 32–40 ., A drawback of the standard synchronous Boolean approach is that it does not permit the implementation of small perturbations , i . e . , noise , of the type that would result from stochastic fluctuations of the number of molecules of a given species or the rates of production of the various species involved ., Indeed , synchronous Boolean models are known to produce many cyclic attractors that represent only marginally stable behavior , which disappear in the presence of noise 41 , 42 ., Here we take an intermediate approach that emphasizes the essential Boolean logic of the system within a continuous-time updating scheme that allows the modeling of small perturbations 41 , 43–46 ., We associate a time delay with each link in the network of regulatory interactions that determines the timing of activation and deactivation events ., The stochastic fluctuations thus appear in our model as deviations of the delay times from their nominal values ., Such models have been termed autonomous Boolean networks 47 , 48 to distinguish them both from models based on synchronous or random asynchronous timing of updates and from Boolean Delay Equations 49 , 50 that do not account for finite response times ., The results presented here apply as well to appropriately constructed ordinary differential equation ( ODE ) models 46 ., Regulatory networks based on the cyclin/CDK-centered view of the cell cycle 51 in S . cerevisiae 38 and Schizosaccharomyces pombe 52 have been studied previously using a synchronous Boolean framework ., In those models , the intrinsic dynamics is not cyclic and the transition sequence corresponding to the cell cycle must be triggered by an external signal ., We emphasize that the network we study is based on the recent experiments 29 , 53 suggesting the existence of a self-sustaining transcriptional oscillator in yeast ., The rest of the paper is organized as follows ., We first define the autonomous Boolean formalism and discuss the necessity for it ., We then demonstrate that it is possible to construct a stable autonomous Boolean oscillator consisting of a long positive feedback loop with two stabilizing motifs added ., This toy oscillator has topological features resembling the yeast cell-cycle oscillator ., We then describe numerical experiments demonstrating that these features are the source of stability in the autonomous Boolean version of the network of Orlando et al . 29 ., We close with a discussion of the implications of these findings ., The details of the computer simulations are provided in the Methods section ., In an autonomous Boolean network ( ABN ) , each node takes one of only two values at any given time: or ., Updates are executed in continuous time as follows ., When a node , , changes its state , it signals all the downstream nodes directly connected to its outputs ., Each downstream node , , receives the signal after a time delay , , which is a real ( not necessarily integer ) value ., When the signal is received , reevaluates its state according to its assigned Boolean function and adopts the resulting value , ., If the new value is different from its present value , a new signal is sent to its own downstream targets ., Nodes do not update at externally dictated times , as in the synchronous model or various asynchronous versions ., The update dynamics is determined by the timing of events , delays , and the topology of the network ., In principle , delays associated with activation ( switch-on events ) , , can be different from the ones associated with deactivation ( switch-off events ) , , because of the different physical processes involved ., The former characterizes multiple processes , including transcription , and translation , folding , post-translational modification , and spatial transport , while the latter can be attributed to degradation of mRNAs and transcription factors ., The difference between and can cause a change in the duration of a pulse of transcriptional activity as it propagates down a chain of nodes 46 ., Consider , for example , a simple cascade with two nodes , where output of regulates ., Suppose we turn on manually at and turn it off at , forming a pulse of width , as shown in Figure 1 ., The rising edge of this pulse arrives at at and the falling edge arrives at ., When , the initial pulse grows as it propagates ( Figure 1 ) and if , it shrinks ., Small perturbations due to stochastic fluctuations , or noise , can significantly alter the dynamics of a network and can be used as a mathematical tool for analyzing the stability of cycles ., Noise is incorporated by taking the time delay associated with a switching event to be , where the noise term , for each propagating signal is drawn at random from a uniform distribution on with ., For present purposes , we take the intrinsic delays and to be equal , allowing the noise to play a dominant role in determining which cycles are stable ., The choice of corresponds to the regime in which the asymmetry in propagation times is small compared to , so that pulses grow or shrink according to the relative values of chosen for the leading and trailing edges ., In certain cases , the noisy dynamics can generate a pulse of negligibly small width , which we call a spike 41 , 47 , 48 ., In the present context , a spike would correspond to arbitrarily fast build-up and degradation of transcripts and therefore is not realistic ., We employ a short-pulse rejection mechanism in the simulations , discarding both pulses and dips with widths less than time unit ., The Methods section below provides details of our computer simulation of ABNs ., As mentioned above , the backbone of the oscillator in the network of interest is a positive feedback loop , also known as a loop of copiers or a simple loop because each node simply assumes the value of its input after some specified time delay ., To demonstrate the need for a stabilization mechanism , we consider first the simple case of a loop of two copiers ., We can assume without loss of generality that the two links have identical delays , ., The network cycles between the 01 and 10 states when one node is initialized with a pulse of sufficiently large width ., Setting that width equal to and setting the noise level to zero reproduces the dynamics of the synchronous Boolean case ., To test the stability of the cycle , we apply arbitrarily small random perturbations: each time a signal propagates across a link , the delay is taken to be , where is a random number drawn from a distribution that is symmetric around zero ., Each perturbation causes the pulse width to grow or shrink as explained above , so that the oscillation eventually collapses to either the or fixed point ( stationary state ) ., Thus the cycle is only marginally stable in the autonomous model and its apparent stability under synchronous updating is an artifact of that scheme ., We identify two classes of motifs , 1 , 2 , 4 , which we call rectifiers and growers , that can correct small perturbations to the timing of the updates and stabilize cycles on an autonomous loop of copiers ., A rectifier imposes an upper limit on the width of the pulse traveling on the positive feedback loop ., The simplest example of a rectifier is auto-repression ( Figure 2A ) , which cuts long pulses down to a width equal to the delay on the auto-repressive link , , and lets short pulses pass through unaffected 46 ., Small perturbations that cause the pulse width to exceed will be filtered by this motif as seen in Figure 2C ., An incoherent feed-forward loop of type 1 ( I1-FFL in the notation of 1 ) , and a negative feedback containing more than one node can also function as rectifiers ., Grower motifs increase the duration of a pulse by a constant amount , but do not adjust them to a particular value ., One example is the coherent feed-forward loop with OR logic ( C1-FFL-OR 1 , 2 ) shown in Figure 2B ., This motif grows pulses by transmitting the input pulse of width from to through two paths with time delays that differ by ., The slower path sustains the output , producing a pulse of width , assuming ., ( If the condition is not met , two pulses will be generated . ), A diamond motif 1 with OR logic , in which both paths connecting the input to the output contain an intermediate node , functions in the same manner ., We also note that both C1-FFL and the diamond motifs function as shrinkers when their output is an AND gate , shrinking the input pulse by or destroying it completely ., A rectifier cannot prevent the collapse to the all-OFF state and a grower alone inserted in a loop will keep growing the pulse until the all-ON attractor is reached ., The two motifs working in tandem ( Figure 2B ) , however , can act as a stabilizing module for cyclic attractors , as seen in Figure 2D: both pulse-growing and pulse-shrinking perturbations are filtered because the grower-rectifier combination resets the pulse width to after each cycle ., Such a network can sustain stable oscillations that have been started with an external signal ., The two motifs will be incompatible if because the grower will generate two pulses from each rectified pulse ., We note that there is no simple motif that acts as a low-pass rectifier , allowing long pulses to pass unaffected while boosting short pulse widths up to a specified value ., Thus the shrinker motif is of limited use for stabilizing oscillations ., Furthermore , a grower-shrinker combination cannot be a stabilizer as it simply acts either as an overall grower or an overall shrinker ., If one allows and to be different , a pulse may grow or shrink as it travels around a simple loop ., When for the links in the loop , we have a source of “intrinsic growth” that may render a grower motif unnecessary , or just assist the grower in restoring pulse widths more rapidly ., In fact , it has been shown using an ODE model with time delays that when switch-on events propagate faster than switch-off events , an auto-repressive link can by itself create a stable cycle on a loop of copiers 46 ., Similarly , when along the loop , a pulse will shrink as it propagates ., Stabilization in the presence of intrinsic shrinkage requires a grower regardless of the noise level ., We do not consider intrinsic growth or shrinkage here , focusing instead on cases where stochastic effects ( noise ) dominate over the intrinsic effects ., Also , we consider only the stabilization of single-pulse cycles , in which each node along the loop ( through ) turns on and off exactly once per cycle time , which we define as the time required for a single signal to propagate around the loop once ., For a simple loop , the cycle time is equal to the sum of the delays , but for more complex circuits , it can depend on the pulse width ., A crucial feature of the oscillator architecture under consideration here ( Figure 2B ) is that it does not rely on any constitutive input or positive auto-regulation ., Consider , for example , the model of circadian oscillations in Drosophila 13 , 21 , 25 , which contains one protein , PER , whose biphosphorylated form represses its own transcription ., It is assumed that Per mRNA is transcribed at the maximum rate in the absence of biphosphorylated nuclear PER , thereby building up spontaneously ., Such an oscillator can be represented as a simple negative feedback loop , PerPER Per with a long time delay on the repressive link ., A Boolean model of the oscillator can be constructed by assigning a NOT function to Per indicating that it builds up spontaneously , but only in the absence of PER; and a COPY function to PER as it is produced only in the presence of Per ., This model has a cycle containing all four states of the circuit , ., From the Boolean perspective , the underlying principle for these oscillations is the impossibility of satisfying all the Boolean functions simultaneously , as the combination of an inverter and a copier creates frustration 42 ., For this reason , we refer to the Boolean versions of such oscillators , which have no fixed points , as frustration oscillators ., The oscillator we propose in Figure 2B , however , has the all-OFF fixed point attractor; there is no frustration in its logic ., It therefore belongs to a different class that involves a stable transmission of a pulse on a loop of copiers , i . e . , a positive feedback loop ., We refer to these as transmission oscillators ., The recently published cell-cycle oscillator network in yeast consists of nineteen interactions between eight transcription factors and one cyclin , CLN3 , which was used as a proxy for currently unidentified transcription factors that complete the circuit ( Figure 3A ) 29 , 30 , 53 ., The regulatory logic functions of the multi-input nodes are not known ., This oscillator was studied using a synchronous Boolean model with eight different “biologically interpretable” logic configurations for the network given in Figure 3A and Table 1 29 ., Each logic configuration was found to support at most two out of the three possible cycles in addition to the all-OFF fixed point ., All three cycles match the sequential order of the expression of the transcription factors ., We emphasize here , however , that these features may only be artifacts of the synchronous update scheme and their stability requires further investigation ., This version of the yeast cell-cycle oscillator is a complex network that does not seem to be a frustration oscillator ., Expression profiles of transcription factors suggest that sequential activations are triggered by immediate upstream regulators in the network 29 ., Therefore , the oscillations are unlikely to be driven by a frustration oscillator that is either a part of or coupled to the circuit ., Several intertwined feed-forward and negative feedback motifs in the network suggest that a grower-rectifier combination may be at play in stabilizing the oscillations ., Specifically , we hypothesize that this network is a simple loop consisting of CLN3 , SBF , SFF , and ACE2 or SWI5 ( since this is the loop of copiers with the least number of links ) , and all other nodes conspire to provide stabilizing motifs ., We use computer simulations to test this hypothesis ., Briefly , we assign random delays to each link and start the network by manually turning CLN3 on then off ., The distribution we choose for the delays roughly captures the variation in delays seen in the experiments 29 ., A broader distribution would not qualitatively change the results ., The details of the simulations are described in Methods ., We have shown using an autonomous Boolean model , that a long positive feedback loop can be turned into a stable oscillator with the addition of two stabilizing motifs that can correct fluctuations in the pulse width ( the duration of activity of each node in the network ) : a rectifier involving a repressor that limits the width of the traveling pulse , and a grower that lengthens the duration of a pulse so that it cannot shrink and disappear ., In combination , a grower and a rectifier ensure that the pulse width returns to the same value after each cycle ., The recently published yeast cell-cycle oscillator 29 has a structure built around a long positive feedback loop , on which waves of activation events propagate ., Numerical simulations of eight different logic configurations and multiple realizations of randomly assigned time delays revealed the presence of grower and rectifier functions in this network ., To our knowledge , there is no other biological oscillator model described in the literature that relies on a long , slow positive feedback loop ., We note that a proposed cell cycle network for Caulobacter crescentus 54 has a structure reminiscent of that of yeast , but no dynamical model of it has yet been reported ., Previous synchronous Boolean models of Drosophila segmentation network 39 , or cyclin/CDK-based cell-cycle networks of S . pombe 52 and S . cerevisiae 38 predicted essential features of the robust dynamics of these networks 55 ., We have demonstrated that the autonomous Boolean framework can be used to further study such problems , since it addresses important elements of the regulatory dynamics associated with the timing of updates and the effects of stochastic fluctuations ., We note that ABNs have also been used recently for analyzing chaos and the stability of periodic orbits in digital electronic oscillators 47 , 48 ., Our results also point to a drawback of fully asynchronous Boolean models: a stable cycle in a continuous-time system such as that of Figure 2D would not be observed in an asynchronous model ., In asynchronous models , cycles generated by loops containing an even number of inverters cannot be sustained 42 because there always exists a sequence of updates that leads to the fixed point state ., We have shown , however , that when appropriate motifs are present , the autonomous rules for determining the order in which nodes are updated never permit evolution to the fixed point even in the presence of a substantial level of noise ., In analyzing the dynamics of gene networks containing feedback loops , it is therefore important to take into account timing information associated with signal propagation ., For gene networks containing feedback loops , results from discrete-time Boolean models ( both synchronous and asynchronous ) should be interpreted with care ., The stability of the oscillations we have observed is not an artifact of the autonomous Boolean model ., The presented results are qualitatively compatible with ODE analogs involving explicit time delays 56 ., An ODE model of a similar system with explicit time delays has already been shown to exhibit stable oscillations very similar to our Boolean idealization when synthesis rates , Hill coefficients , and time delays are large enough 46 ., Our own preliminary studies indicate that it is also possible to construct an ODE model of a transmission oscillator without explicit time delays by selecting appropriate parameters for the stabilizing motifs ., To simulate the dynamics of an autonomous Boolean network , we use an event-driven code ., A time-ordered event queue is established , in which each event represents the switching of an input at a specified node ., Each time an event is processed that results in the switching of a node , events are added to the queue according to the time delay associated with each output link from that node ., After each update of a node , we check to see whether it creates a short pulse that should be rejected ., If so , the queue is purged of all events derived from the leading and trailing edges of the pulse ., To avoid causality problems coming from propagation of a switching event that is later rejected , we choose the maximum noise amplitude , , to be less than half of the short-pulse rejection time ( time unit ) ., To reveal the structure of the yeast cell-cycle oscillator , we study numerical simulations of autonomous Boolean versions of the network with the logic choices in Reference 29 and different randomly selected sets of time delays ., For each logic configuration , we generate an ensemble of 10000 networks with quenched random delays on each link ., Delays were chosen from a uniform distribution between and 2 time units ., The system was initialized by turning CLN3 on at and turning it off at , while other nodes were OFF ., All nodes were assumed to be OFF for ., To simulate noise , a random value selected from a uniform distribution on the interval , was added to the delay associated with each update ., We are interested in the stability of a particular cycle , so we have chosen an initial condition that is very likely to lie in the basin of attraction of that cycle ( if the cycle exists ) ., A different initial condition , turning CLN3 on at and letting a repressor turn it off , yields roughly the same statistics reported in Table 1 ., We do not test the networks robustness to general changes in the initial condition 38 , 39 , 57 , 58 ., We run simulations up to 125000 updates with noise turned on between the 800th and 90000th updates in order to eliminate marginally stable oscillations ., For single-pulse oscillations , this typically translates into a runtime of time units under noise ., Periodic single-pulse oscillations that survive this long with noise present are highly likely to be stable attractors ., An oscillation is considered to be PSP if pulse widths on two consecutive cycles differ by less than time unit on each node ., We do not check whether all nodes turn on and off once per cycle time , i . e . , whether the cycle is a single-pulse or a dual-pulse with identical pulse widths ., However , we never observed the latter in the inspected realizations and believe that it is very unlikely to occur in this circuit ., The numbers of oscillating realizations differ in the different logic configurations for two reasons ., First , an FFL or diamond motif operates as a grower with OR logic and as a shrinker with AND logic ., When two logic configurations differ only by the selection of or , the one with OR logic always has a larger number of oscillating networks ., In configurations 1 and 5 , both SFF and CLN3 are AND gates , so the motifs they belong to will act as shrinkers ., The existence of two shrinkers in the network should make oscillations very unlikely , and indeed we find that all realizations in both configurations collapse on the all-OFF attractor ., On the other hand , configurations that contain a larger number of AND logic for and generate mostly periodic single-pulse oscillations and fewer complex ones ., The second reason for the difference in the number of oscillating networks is the logic , , of the repressors ., The case gives a smaller total number of oscillating realizations than . | Introduction, Results, Discussion, Methods | A recently published transcriptional oscillator associated with the yeast cell cycle provides clues and raises questions about the mechanisms underlying autonomous cyclic processes in cells ., Unlike other biological and synthetic oscillatory networks in the literature , this one does not seem to rely on a constitutive signal or positive auto-regulation , but rather to operate through stable transmission of a pulse on a slow positive feedback loop that determines its period ., We construct a continuous-time Boolean model of this network , which permits the modeling of noise through small fluctuations in the timing of events , and show that it can sustain stable oscillations ., Analysis of simpler network models shows how a few building blocks can be arranged to provide stability against fluctuations ., Our findings suggest that the transcriptional oscillator in yeast belongs to a new class of biological oscillators . | Technologies such as gene arrays enable acquisition of large amounts of data on gene expression variations , which reveal the structures of gene regulatory networks that govern the metabolic and developmental machinery in the cell ., We study a model of an oscillatory gene regulatory network that has been recently suggested to play an integral role in maintaining the cell cycle in yeast ., The oscillator differs from other known biological and synthetic oscillatory networks in that it seems to rely on a long positive feedback loop ., We show that the presence of certain stabilizing sub-networks can account for the robustness and the unusual architecture of this oscillator ., Our modeling approach elucidates both the logical structure of the system and the importance of the timing of update events . | computational biology/systems biology, physics/interdisciplinary physics, computational biology/transcriptional regulation | null |
journal.pntd.0001161 | 2,011 | Genetic Reconstruction of Protozoan rRNA Decoding Sites Provides a Rationale for Paromomycin Activity against Leishmania and Trypanosoma | Aminoglycoside antibiotics show broad-spectrum antibacterial activity and are a common choice for treatment of serious infections due to gram-negative bacilli , including endocarditis , sepsis , pneumonia , and pyelonephritis 1 ., Among the aminoglycoside antibiotics , paromomycin has also been shown to be effective against some protozoa and cestodes ., The cost of paromomycin is low , making it a particular good drug candidate in countries that carry a burden of high parasitic infection rates ., While paromomycin is out of use as an antibacterial , it is marketed as an oral treatment for amoebiasis and giardiasis ., Paromomycin is also used in combination therapy as a topical treatment for cutaneous leishmaniasis 2 ., Recently , paromomycin was licensed as a treatment for visceral leishmaniasis , the most severe form of leishmaniasis ( reviewed in 3 ) ., Aminoglycosides exert their antibacterial activity by binding to a highly conserved region in helix 44 of bacterial 16S-rRNA 4 ., We have previously reconstructed the drug target site of protozoan cytosolic ribosomes in chimeric bacterial ribosomes to demonstrate that the decoding site of cytosolic Leishmania ribosomes is susceptible to paromomycin but not to various other aminoglycosides 5 ., These results have been recently confirmed by studies that showed specific paromomycin binding to the decoding site of Leishmania cytosolic ribosomes by surface plasmon resonance analysis 6 ., While these studies have collectively provided a molecular rationale for the antileishmanial activity of paromomycin , these findings did not address whether in addition the Leishmania mitochondrial ribosome is targeted by aminoglycosides ., Recent data have demonstrated that mitochondrial translation is essential for both the procyclic and the bloodstream form of Trypanosoma brucei and that consequently mitochondrial protein synthesis may represent an important drug target throughout the life cycle of trypanosomes 7 ., There is , however , some inconsistency in the literature with regards to the effect of paromomycin on mitochondrial protein synthesis in Leishmania ., For instance Maarouf et al . 8 have reported that paromomycin interferes with mitochondrial protein synthesis in Leishmania , whereas Horvath et al . 9 found no effect of paromomycin on mitochondrial translation ., Structural analysis of the Leishmania mitochondrial ribosome has revealed a remarkable morphologic similarity to the eubacterial ribosome 10 ., However , the homolog of bacterial 16S rRNA helix 44 in trypanosome mitochondria is truncated in comparison to its bacterial counterpart , although the proximal part constituting the decoding site and the aminoglycoside-binding site is fully retained ( Fig . 1 ) ., This is not surprising as the ribosomal decoding site is one of the most important catalytic domains within the ribosome , and it is universally conserved across all phylogenetic domains of life including organelles ., At the same time , the mitochondrial rRNA of trypanosomes carries unique signatures within its decoding site sequence ., Not only is this rRNA motif significantly different from bacterial 16S rRNA , it also shows a considerable sequence difference between the two closely related genera Leishmania and Trypanosoma ( Fig . 1 ) ., The distinctive structural features of the mitochondrial decoding site make it difficult to predict its functional susceptibility to compounds that bind to the bacterial decoding site ., Here we reconstructed the mitochondrial decoding sites of Leishmania and Trypanosoma in bacterial ribosomes to analyze their susceptibility to aminoglycoside antibiotics and to allow for a comprehensive evaluation of the therapeutic potential of this class of drugs against trypanosome parasites ., Based on the susceptibility pattern of chimeric ribosomes mimicking the ribosomal decoding sites of Trypanosoma , we assessed the antiprotozoal activity of paromomycin in cultures of T . brucei and in a mouse model of infection ., Animal experiments were carried out in compliance with Swiss federal law ( TSchG ) and cantonal by-laws ( TSchV Basel-Stadt ) ., All protocols and procedures were reviewed and approved by the local veterinary authorities of the Kanton Basel-Stadt ( Permit Number: 739 ) ., All rRNA nucleotides discussed in this study are numbered according to their homologous position in E . coli 16S rRNA ., 9S rRNA genes of L . donovani , L . major , L . amazonensis , L . tarentolae , T . brucei , and T . cruzi ( Genbank accession numbers FJ416603 , EU140338 , HM439238 , M10126 , M94286 , and DQ343645 , respectively ) were used for rRNA sequence alignments of the mitochondrial decoding sites ., 18S rRNA genes of L . donovani , L . major , L . mexicana , L . amazonensis , L . braziliensis , L . tarentolae , T . brucei , and T . cruzi ( accession numbers FR799614 , FR796423 , GQ332360 , GQ332354 , GQ332355 , M84225 AL929603 , and FJ001665 , respectively ) were used for rRNA sequence alignments of the cytosolic decoding sites ( Fig . S1 ) ., Recombinant Mycobacterium smegmatis strains with chimeric ribosomes were constructed by previously described gene replacement procedures ., 1 ) rRNA fragments coding for mutant rRNA were generated by PCR mutagenesis and cloned into a suitable vector to result in chromosomal gene replacement via homologous recombination with donor DNA by selection of the mutant genotype 11 ., 2 ) Plasmid replacement in a Δrrn knockout strain of M . smegmatis mc2-155 ., DNA sequences coding for chimeric rRNA were generated by PCR , cloned into a plasmid coding for a full rRNA operon , and used to replace the wild-type rRNA sequence in M . smegmatis by means of plasmid exchange 5 ., Strains and plasmids used in this study are listed in Tables S1 and S2 ., Successful rRNA replacement was controlled by sequence analysis ., Recombinant M . smegmatis strains were studied for susceptibility to paromomycin , neomycin , gentamicin , and tobramycin ( Sigma Aldrich ) ., Minimal inhibitory concentrations ( MIC ) were determined by broth microdilution tests as described previously 12 ., The gentamicin used in this study is a mixture of gentamicin C1 , gentamicin C1a , and gentamicin C2 in a 45∶35∶30 ratio ., The chemical structures of aminoglycoside antibiotics are provided as Fig . S2 ., Bloodstream forms of Trypanosoma brucei rhodesiense STIB900 and axenic amastigote forms of L . donovani ( MHOM-ET-67/L82 ) were used for the in vitro assays ., Cytotoxicity was assessed with rat skeletal myoblasts ( L6 cells ) ., IC50 values were obtained using a resazurin-based assay ( alamarBlue ) as described earlier 13 ., The following compounds were used as standards: melarsoprol ( T . b . rhodesiense ) , miltefosine ( L . donovani ) , and podophyllotoxin ( L6 cells ) ., Test compounds were dissolved in distilled water ., Following initial experiments the assay duration time was extended to 96 hrs , 120 hrs and 144 hours for L . donovani to demonstrate in vitro activity of paromomycin ., For the assays with extended duration a subculture was prepared after 72 hrs by transferring 10 µl of the initial assay culture to 100 µl fresh medium containing the corresponding compound concentration ., Groups of 4 female NMRI mice ( Harlan Netherlands ) of 22–25 g were infected by the intraperitoneal route with 104 bloodstream form of T . brucei brucei STIB795 per mouse ., Paromomycin was formulated in physiological saline solution and administered i . p . in a volume of 10 ml kg−1 . Treatment was initiated on day 3 post-infection ., The mice were monitored for parasitaemia 24 hrs after the last treatment ( day 7 ) and again on day 10 and 14 post-infection ., Once parasitaemia exceeded 107/ml the animals were euthanized ., The day of relapse was used as endpoint ., The parasitic trypanosomes ( Kinetoplastida: Trypanosomatidae ) include the genera Trypanosoma and Leishmania ., We have previously reconstructed the aminoglycoside drug binding pocket , i . e . the decoding A-site , of trypanosome cytosolic ribosomes in bacteria 5 ., To do so , we have replaced the proximal stem of helix 44 in bacterial 16S rRNA , i . e . nucleotides homologous to positions 1408–1416 and 1484–1491 in bacterial 16S rRNA , with the corresponding cytosolic rRNA sequences present in Leishmania and Trypanosoma species ( Fig . 1 and 2 ) ., In the same study , we demonstrated the specific activity of paromomycin against Leishmania and Trypanosoma chimeric cytosolic ribosomes 5 ., For a comprehensive analysis of trypanosomal susceptibility to 2-deoxystreptamines , we reconstructed the mitochondrial decoding A-sites of Leishmania and Trypanosoma in bacterial ribosomes ., The mitochondrial rRNA sequence , and in particular the decoding A-site is considerably different between Leishmania and Trypanosoma ( Fig . S1 ) ., Base pair 1409—1491 , which is considered key in aminoglycoside susceptibility , is characterized by an U•C interaction in Leishmania versus an A—U base pairing in Trypanosoma ( Fig . 1 ) ., Furthermore , the mitochondrial A-site of Leishmania and Trypanosoma exhibits a unique 1406C—1495G base pairing ., This is in contrast to the cytosolic decoding A-site , which shows the typical 1406U—1495U interaction that is universally conserved across the three domains of life including organelles ., Attempts to reconstruct the entire mitochondrial trypanosome A-site in bacteria as a 27-nucleotide helix comprising residues 1404–1416 and 1484–1497 were unsuccessful ., We thus reconstructed the trypanosome homologue of helix 44 as two mutant chimeric ribosomes: one with a 19-nucleotide helix of Leishmania and Trypanosoma mitochondrial rRNA corresponding to M . smegmatis positions 1408–1416 and 1484–1493 , and one mutant recombinant carrying the characteristic trypanosome mitoribosomal 1406C—1495G pair ( Fig . 2 ) ., Studying the bacterial recombinants revealed that the mitochondrial chimeras with the 19-nucleotide replacement are susceptible to compounds with a 6′-amino substituent , such as neomycin , gentamicin , and tobramycin , and less susceptible to paromomycin , which carries a 6′-hydroxy substituent ( Fig . 2; the chemical structures of the tested 2-deoxystreptamines are provided in Fig . S2 ) ., In addition , the Trypanosoma 19-nucleotide chimera was found to be generally more susceptible to aminoglycosides than its Leishmania counterpart reflecting the canonical base pair interaction between residues 1409 and 1491 ( Fig . 2 ) ., Recombinants carrying the 1406C—1495G pair were susceptible to the 4 , 5-disubstituted aminoglycosides neomycin and paromomycin , but highly resistant to the 4 , 6-disubstituted aminoglycosides gentamicin and tobramycin ( Fig . 2 ) ., From the combined results we conclude that among the aminoglycosides tested , neomycin has the highest activity against the protozoan mitoribosomes , while gentamicin and tobramycin are virtually inactive ., The finding that the Trypanosoma ribosomal decoding sites are at least as susceptible to paromomycin as those of Leishmania ribosomes prompted us to determine the efficacy of aminoglycosides against Trypanosoma in a cell culture assay ., We found that paromomycin suppressed growth of Trypanosoma brucei rhodesiense in vitro ( Table 1 ) ., Compared to T . brucei rhodesiense , IC50 determinations of paromomycin in L . donovani required extended assay durations ., Notably , the 50% inhibitory concentration ( IC50 ) of paromomycin was more than tenfold lower with T . brucei than with L . donovani ( Table 1 ) ., To assess whether mitochondrial protein synthesis is involved in paromomycins potent antiprotozoal activity , we also studied the efficacy of neomycin against T . brucei ., Neomycin shows significant activity against Trypanosoma mitochondrial recombinant ribosomes but little activity against Trypanosoma cytosolic chimeras ( Fig . 2 ) ., Thus , if trypanosome mitochondria were to be targeted by aminoglycosides , we would expect neomycin to show activity against T . brucei ., However , compared to paromomycin , neomycin had no effect on the growth of T . brucei ( Table 1 ) ., Tobramycin , an aminoglycoside that is inactive against both cytosolic and mitochondrial chimeric ribosomes ( Fig . 2 ) , was used as a control and had no effect on trypanosome growth in culture ( Table 1 ) ., Based on the above findings we wished to study the efficacy of paromomycin in the T . brucei brucei STIB795 mouse model of infection ., Paromomycin suppressed the growth of T . brucei in vivo and increased the mean survival time even at the lowest dose regimen of 100 mg per kg of body weight given for 4 days ( Table 2 ) ., However , high concentrations of paromomycin did not fully eradicate the parasite in the mouse infection model , as parasitemia relapsed once treatment was stopped ., Protein synthesis is an established drug target in antibacterial chemotherapy and has been considered as target for antiprotozoal drugs 14 ., Although the mechanisms of action of drugs targeting the ribosome have still to be studied in more detail , it has become clear that both cytosolic and organelle ribosomes represent potential drug targets 15 , 16 ., The plastidal ribosome is an established drug target in Apicomplexa 17 , and serves as target for drugs such as clindamycin in the treatment of Toxoplasma 18 or plasmodia infections 19 ., Among the various aminoglycosides paromomycin has been repeatedly found to be the most potent antiprotozoal compound ., In early studies , paromomycin showed good antiamoebic and some antitrichomonal activity 20 whereas the related compound neomycin , which differs from paromomycin by a single substituent ( amino instead of hydroxy group at position 6′; Figure S2 ) , was poorly active against Entamoeba histolytica 21 ., Likewise , paromomycin was shown to be more potent against amoebiasis than kanamycin , neomycin , and gentamicin , respectively , and in a study on Giardia lamblia paromomycin was the only 2-deoxystreptamine that showed activity 22 ., Until recently , leishmaniasis was exclusively treated with sodium stibogluconate , pentamidine , or amphotericin B , but such treatments are expensive and potentially toxic ., In contrast , paromomycin has been recently proposed as a well-tolerated and affordable treatment for visceral leishmaniasis ( reviewed in 3 ) , which is still considered the second biggest parasitic killer after malaria ., Aminoglycoside antibiotics are cationic compounds with a 2-deoxystreptamine core that is glycosidically linked at position 4 to a glucopyranosyl ( Fig . S2 ) ., Additional amino sugars are attached to either position 5 or 6 of the 2-deoxystreptamine moiety ., Both the 4 , 5- and 4 , 6-disubstituted aminoglycosides target the ribosome by direct interaction with ribosomal RNA , affecting protein synthesis by inducing codon misreading and by inhibiting translocation of the tRNA-mRNA complex 23–25 ., The binding site is defined by a number of nucleotides in the proximal stem of helix 44 in bacterial 16S rRNA ( reviewed in 26 ) ., Genetic and biochemical studies showed that the following bases are particularly relevant for aminoglycoside binding: A1408 , C1409—G1491 , and U1406—U1495 12 , 27–30 ., We have previously emphasized the role of nucleotide 1408 in aminoglycoside susceptibility by studies of a cytosolic rRNA chimera of the trypanosome Blastocrithidia ., This parasite carries a ribosomal decoding site that is identical to that of Leishmania and Trypanosoma with the exception of a 1408 adenosine instead of a guanosine 5 ., In line with structural predictions , the cytosolic A-site of Blastocrithidia was found to be highly susceptible to all aminoglycosides tested ., In contrast , the aminoglycoside target site in chimeric cytosolic ribosomes of Leishmania and Trypanosoma showed drug susceptibility merely to paromomycin but not to other aminoglycosides 5 ( Fig . 2 ) ., These findings are in agreement with previous observations that Leishmania cultured in vitro is susceptible to paromomycin 31 , 32 ., Maarouf et al 8 have reported that paromomycin interferes with mitochondrial protein synthesis in Leishmania , whereas Horvath et al . 9 found no effect of paromomycin on mitochondrial protein synthesis ., As a result , the precise role of mitochondria in trypanosome susceptibility to aminoglycosides has remained elusive ., The uracil-uracil opposition at position 1406—1495 is universally conserved across the three phylogenetic domains of life including their organelle ribosomes ., The trypanosome mitochondrial ribosomes are exceptional in that they are characterized by a C—G base pair at 1406—1495 ., Any alteration of the U—U pair , particularly in U1406 , results in decreased susceptibility to 4 , 6-substituted aminoglycosides such as gentamicin and tobramycin 26 ., This is most likely due to a distortion of the amino sugar attached to position 6 of the aminocyclitol ring , thereby disrupting its hydrogen bonds to G1405 ., The conformation of aminoglycosides with an amino sugar attached to position 5 of the aminocyclitol ring is such that no contacts are made with G1405 , and thus binding of 4 , 5-aminoglycsides is generally less affected by base alterations in 1406—1495 ., Hence , with regards to 4 , 6-substituted compounds , the 1406C—1495G pair confers high-level drug resistance ( Fig . 2 ) ., It appears that the C–G pair alone renders the mitochondrial ribosome of trypanosomes resistant to a large number of aminoglycoside antibiotics , except the 4 , 5-disubstituted aminoglycosides paromomycin and neomycin ., Neomycin has significant activity against the trypanosome mitochondrial A-site in recombinant chimeric ribosomes ., Among the aminoglycosides tested , which include paromomycin , gentamicin , and tobramycin , neomycin has the most potent antimitoribosomal activity ( Fig . 2 ) ., However , in accordance with previous studies , we find that neomycin has no anti-trypanosomal activity in cell culture in-vitro ( Table 1 ) ., Together with recent results demonstrating the essentiality of mitochondrial translation in the life cycle of T . brucei , our findings indicate that mitochondrial translation is not accessible to aminoglycosides ., The organelles resistance to neomycin action apparently is not due to an intrinsic resistance of the drug target structure , but most likely reflects limited mitochondrial permeability ., The observation that the cytosolic ribosomes of Trypanosoma are identical to that of Leishmania with respect to both rRNA sequence of the drug binding site and aminoglycoside susceptibility prompted us to test the efficacy of paromomycin against Trypanosoma in vitro and in vivo ., Paromomycin was the most active antibiotic when compared to other aminoglycosides targeting protein synthesis ., Intriguingly , in the cell culture model T . brucei showed significantly higher susceptibility to paromomycin than Leishmania donovani ., In the T . brucei STIB795 mouse model , paromomycin treatment was able to suppress parasitemia and thus significantly increase the mean survival time of infected mice , suggesting that paromomycin inhibits the growth of T . brucei ., However , paromomycin treatment did not eradicate the pathogen and relapse occurred in all tested animals after termination of treatment ., Apparently , the anti-trypanosomal activity of paromomycin is sufficient to prevent Trypanosoma propagation and spread , but by itself is insufficient to cure the disease ., Together , the findings reported here point to additional potential for antiprotozoal aminoglycosides beyond the currently established treatment options . | Introduction, Materials and Methods, Results, Discussion | Aminoglycoside antibiotics target the ribosomal decoding A-site and are active against a broad spectrum of bacteria ., These compounds bind to a highly conserved stem-loop-stem structure in helix 44 of bacterial 16S rRNA ., One particular aminoglycoside , paromomycin , also shows potent antiprotozoal activity and is used for the treatment of parasitic infections , e . g . by Leishmania spp ., The precise drug target is , however , unclear; in particular whether aminoglycoside antibiotics target the cytosolic and/or the mitochondrial protozoan ribosome ., To establish an experimental model for the study of protozoan decoding-site function , we constructed bacterial chimeric ribosomes where the central part of bacterial 16S rRNA helix 44 has been replaced by the corresponding Leishmania and Trypanosoma rRNA sequences ., Relating the results from in-vitro ribosomal assays to that of in-vivo aminoglycoside activity against Trypanosoma brucei , as assessed in cell cultures and in a mouse model of infection , we conclude that aminoglycosides affect cytosolic translation while the mitochondrial ribosome of trypanosomes is not a target for aminoglycoside antibiotics . | Rational design of novel therapeutics relies on the knowledge and understanding of potential drug targets ., Historically , the majority of therapeutics have not been rationally designed , but empirically discovered ., Paromomycin , an aminoglycoside with antibacterial activity , has been found to show considerable activity against leishmaniasis , a disease caused by the protozoan parasite Leishmania ., However , the mechanisms of aminoglycoside action against protozoan parasites have in part remained unclear ., In this study we demonstrate that the cytosolic ribosome is the preferred drug target , and that the mitochondrial ribosome does not contribute to the antiprotozoal activity of aminoglycosides ., As the cytosolic ribosome of Trypanosoma , the causative agent of sleeping sickness and Chagas disease , resembles that of Leishmania , we tested the efficacy of paromomycin against Trypanosoma ., We found that paromomycin not only inhibits the growth of Trypanosoma in culture , but also suppresses trypanosomiasis in a mouse infection model ., Our results point to the cytosolic ribosome as a promising drug target for antiprotozoal drug development . | medicine, parasite groups, african trypanosomiasis, microbiology, parasitic diseases, parasitology, prokaryotic models, model organisms, neglected tropical diseases, infectious diseases, protozoan models, biology, chagas disease, leishmaniasis, genetics, molecular cell biology, genetics and genomics | null |
journal.pcbi.1002069 | 2,011 | Mutual Inactivation of Notch Receptors and Ligands Facilitates Developmental Patterning | Notch signaling is the canonical metazoan juxtacrine signaling pathway ., It is involved in many developmental processes in which neighboring cells adopt distinct fates ., Examples of such processes include the delineation of sharp boundaries during the formation of Drosophila wing veins 1 , 2 and the formation of checkerboard-like patterns of differentiation , as occurs during Drosophila microchaete bristle patterning 3 ., Notch signaling occurs through contact between a Notch receptor on one cell and a Delta/Serrate/LAG-2 ( DSL ) ligand such as Delta or Serrate ( Jagged in mammalian cells ) on a neighboring cell ., This interaction leads to cleavage of Notch , releasing its intracellular domain , which translocates to the nucleus and serves as a co-transcription factor to activate target genes 4 ., In addition to this activating trans interaction between Notch and DSL on neighboring cells , inhibitory cis interactions between Notch and DSL in the same cell suppress Notch signaling 5 , 6 , 7 , 8 , 9 , 10 ., Recent work indicates that this cis-interaction between Notch and DSL is symmetric: Notch inhibits its ligand , and the ligand inhibits Notch 9 , 11 , 12 ., The molecular mechanism of this mutual inactivation between Notch and DSL , and whether or not it occurs at the cell surface , is still unclear 9 , 12 , 13 , 14 ., In an individual cell , mutual inactivation of Notch and DSL results in an ultrasensitive switch between ‘sending’ ( low Notch/high DSL ) and ‘receiving’ ( low DSL/high Notch ) cellular states ( see Fig ., 1 ) 11 ., A cell with more total Notch than DSL ( i . e . with a higher production rate of Notch than DSL given equal first order degradation rates ) has an excess of free Notch but very little free DSL , making it a receiver ( Fig . 1A , left ) ., Conversely , a cell with more total DSL than Notch would have an excess of DSL and very little Notch , thus becoming a sender ( Fig . 1A , right ) ., In either state , both ligand-mediated inhibition of receptor and receptor-mediated inhibition of ligand contribute to the nonlinearity of the system ., For a sufficiently strong cis interaction , the transition between these two states becomes very sharp , or ultrasensitive ( Fig . 1A ) ., This switch generates strongly-biased signaling if a sender cell interacts with a receiver cell ( Fig . 1B , bottom ) , but if both interacting cells are in the same signaling state ( Fig . 1B , top and middle panels ) much less signal is transduced ., Given that the Notch signaling system is involved in many developmental processes , it is important to determine how this cis-dependent send/receive signaling switch impacts pattern formation in developing tissues ., A well-studied class of biological patterning systems is local self-activation with long-range inhibition 15 ., Our model of Notch signaling-driven lateral inhibition patterning may be discussed in similar terms , with the mutual cis inhibition contributing to both the local and long-range effects ., However , in this case the coupling required for “long-range” inhibition occurs via short-range nonlinear juxtracrine interaction between neighboring cells , instead of via linear diffusion of a signaling molecule across long distances 16 ., Moreover , the mutual inactivation of Notch and DSL discussed above provides an improved source of intra-cellular self-activation 17 leading to the effects on pattern formation described here ., In order to understand the implications of the Notch-DSL signaling switch for developmental patterning , we analyzed mathematical models of two canonical developmental patterning processes: ( 1 ) morphogen gradient-driven boundary formation and ( 2 ) lateral inhibition ., We compared models incorporating mutual inactivation in cis to alternative models lacking this interaction ., The results show how mutual inactivation provides several key advantages for patterning circuits: it can allow sharp boundary formation without intracellular feedback , maintain it across a broad range of morphogen gradient slopes , and make patterning insensitive to correlated fluctuations ( ‘extrinsic noise’ ) in Notch and ligand expression ., In lateral inhibition circuits , mutual inactivation speeds up patterning and relaxes parametric requirements on the regulatory interactions ., Finally , it permits a surprisingly simple , and counter-intuitive , lateral inhibition circuit architecture , in which Notch activates its own expression , and no additional feedback or involvement of other components is required ., Wing vein formation in the developing fly is a classic model system for studying the generation of sharp boundaries ., In the Drosophila wing , there are four longitudinal veins that include several rows of cells that are more compact and have darker pigmentation than intervein cells ., The position of the wing veins in the wing imaginal disk is initiated by EGF signaling during the early stages of larva development 18 ., The final form ( position and width ) of the wing veins is refined by several subsequent processes ., Notch signaling has been shown to specifically control the sharpening of the boundary between pro-vein ( the region competent to produce vein fates ) and intervein regions in the wing disc 1 , 2 ., In this system , the Delta production rate is controlled by a gradient of veinless expression diminishing outward from the center of the pro-vein region ( Fig . 2A , left ) ., Notch signaling is observed in two sharply defined side-bands , which restrict further vein development to the region between them ( Fig . 2A , right ) ., We analyzed two simplified models of boundary formation , with or without mutual inactivation ( Figs . 2BC , Eqns . 1–6 , and Supporting Information Text S1 . 2 ) ., In both models , we assume constant Notch production ( at a rate denoted ) throughout the field of cells ( blue line in Fig . 2D , top ) ., We also assume that a linear gradient from the center of the vein , , controls the rate of ligand production , denoted ( red lines in Fig . 2D , top ) ., Alternative models with other gradient shapes lead to the same results shown below ., In the mutual inactivation ( MI ) model ( Fig . 2C ) , mutually exclusive signaling states generate sharp side-bands ( as observed experimentally ) where ‘sender’ cells contact ‘receiver’ cells near the crossing of the Notch and DSL production rate profiles ., This model does not consider any feedback of Notch signaling on either DSL or Notch itself in the same cell , and thus lateral inhibition does not arise in this case ( in contrast with the lateral inhibition models below ) ., Alternatively , in the ‘bandpass’ ( BP ) model a similar Notch activity profile can be generated in the absence of mutual inactivation , but this requires a bandpass filter of Notch activity level which we represent phenomenologically as the product of increasing and decreasing Hill functions ( Figs . 2B , S1A ) ., Such a bandpass filter represents the effective action of diverse regulatory processes downstream of Notch signaling , which could exist in different signaling architecture alternatives to the MI mechanism ., We note here that while transcriptional feedbacks on Notch and DSL have been described in vein formation 1 , 2 , we do not explicitly consider them in these models in order to focus on the main effects of the mutual inactivation process ., Our qualitative conclusions are insensitive to their inclusion ., The equations representing these models are derived in the Supporting Information Text S1 . 2 and summarized in Eqns ., 1–6 ., The slope of the morphogen gradient is expected to vary in natural systems from fluctuations and/or genetic variability , and thus may be an important factor in determining boundary features ., To investigate the effect of such variability on boundary formation , we systematically analyzed the responses of the two models to different morphogen gradient slopes ., For both models , we maintained the position of the threshold at a constant distance from the center of the vein ( Fig . 2D , top ) ., In the MI model , the width of the signaling bands remained nearly constant across a wide range of morphogen gradient slopes ( Fig . 2D , middle ) ., This resulted from the sharp switch from a sending to a receiving state at the intersection ., In contrast , the amplitude of the signaling bands changed systematically with the magnitude of the slope ., This can be understood by considering how much free Notch and free DSL is available at the sender-receiver interface ., The concentration of free DSL or Notch in the sending or receiving cell , respectively , is approximately proportional to the difference in Notch and DSL production rates , which in turn is proportional to the slope of the gradient ., In contrast , the BP model shows substantial broadening of the bands at lower values of the gradient slope ( Fig . 2D , bottom ) ., Unlike in the MI model , here Notch signaling occurs throughout the field of cells and is simply filtered by the downstream band-pass ., As a result , the width of the Notch signaling bands is approximately proportional to the width of the bandpass divided by the slope of the Notch signaling profile ( Fig . S1B ) ., The key parameters controlling the reporter expression profiles are the strength of the cis-interaction , for the MI model ( decreasing leads to increasing cis-interaction strength ) , and cooperativity , for the BP model ., Interestingly , the BP model supports a sharp boundary only for sufficiently large and sufficiently high slopes ( Fig . 2E , bottom ) ., In contrast , with the MI model , band sharpness is preserved across a broad range of values and morphogen slopes ( Fig . 2E , top ) ., Thus , mutual inactivation enables a more robust patterning mechanism ., A striking aspect of the Drosophila wing vein system is observed in the heterozygous mutants of Notch and Delta ( e . g . single copies of the Notch and Delta genes ) ., While heterozygous mutants of Notch ( Notch+/− ) or Delta ( Delta+/− ) alone exhibit mutant phenotypes ( causing thicker veins ) , the Notch+/− Delta+/− double mutant restores the wild-type phenotype 19 , 20 , 21 ., More generally , several mutant phenotypes seem to depend on the ratio between the copy numbers of the Notch and DSL genes 19 ., This ratiometric dependence of the vein width cannot be derived from the several known feedbacks operating in the Drosophila wing vein , but emerges automatically from the MI model ., This is because the position of the Notch signaling band occurs where Notch and DSL production rates are equal ., This position remains unchanged when both rates are multiplied by the same factor ., By the same reasoning , the vein width ( distance between side bands ) increases with increasing ratios between the effective copy numbers of DSL and Notch , as shown in Figs ., 3A , S2 ., Interestingly , however , this picture breaks down when the maximum DSL production rate , , becomes smaller than the Notch production rate , ., What phenotype would we expect in this case ?, Here , since all cells are essentially ‘receivers’ we expect negligible levels of Notch signaling , leading to a phenotype of an unsharpened , diffusely-defined vein , that defaults to the pre-patterned vein-competent region ., Indeed , the Delta+/− phenotype exhibits broad veins with diffuse boundaries , similar to Delta null mutant clones 19 ., This result makes a quantitative prediction: the maximal DSL ( Delta ligand in the case of the wing vein ) production rate should be less than twice the constitutive Notch production rate in this system ., In the fly larva , the width of the vein remains quite constant over length-scales of many cells ., This occurs despite the possibility of substantial fluctuations , or ‘noise’ , in the expression of Notch , Delta , and other components 22 ., In order to understand how gene expression noise affects the MI wing vein model , we considered the response of the system between two limiting cases 23 ., At one extreme , noise can be completely ‘intrinsic’ , meaning that Notch and DSL production rates fluctuate in an uncorrelated manner ., At the opposite extreme , ‘extrinsic’ noise could dominate , generating correlated fluctuations in Notch and DSL production ., As shown in Fig . 3B , intrinsic noise causes the width of the vein to become irregular ( Fig . 3B , bottom ) , while extrinsic noise of the same magnitude has significantly less effect on width ( Fig . 3B , top ) ., To show the generality of this effect , we performed simulations of boundary formation patterning for a range of different noise amplitudes and correlations ( Fig . 3C ) ., These simulations show that the standard deviation of peak position ( which is a measure of pattern robustness ) decreases as the noise becomes more extrinsic ., This behavior emerges from the ratiometric sensitivity of the MI model to the levels of Notch and DSL ., In the MI model , the signaling state of a cell ( sending or receiving ) is determined by the ratio of Notch to DSL – in ‘sender’ cells this ratio is smaller than one , and in ‘receivers’ it is greater than one ., As the vein edge is defined by Notch signaling , it is restricted to the area where sender cells are in direct contact with receiver cells , at which Notch and DSL production rates are comparable ( Fig . 2D ) ., Extrinsic noise tends to maintain constant relative expression of Notch and DSL ., Therefore , it does not disturb the segregation of cells into senders and receivers , and preserves the band of Notch signaling activity ., This effect is maintained across a broad range of noise amplitudes and correlation levels ., Lateral inhibition models have been used to describe the formation of checkerboard-like patterns in which high DSL cells are surrounded by low DSL neighbors ., This type of structure occurs in bristle patterning in Drosophila 3and hair cell patterning in the vertebrate inner ear 24 ., Standard lateral inhibition ( LI ) models assume that neighboring cells inhibit each others differentiation through Notch signaling , which indirectly down-regulates DSL expression to form an intercellular positive feedback loop ( Fig . 4A , Supporting Information Text S1 . 3 ) ., Under the right conditions , this feedback loop can amplify small initial differences between cells and generate patterns in which neighboring cells exhibit alternating expression levels ., A lateral inhibition model of this type was analyzed previously 16 , 25 ., How does mutual inactivation affect the lateral inhibition patterning process ?, To address this question we systematically compared the standard LI model ( Fig . 4A ) to a lateral inhibition with mutual inactivation ( LIMI ) model ( Fig . 4B , equations are summarized in Eqns . 10–12 , and derived in Supporting Information Text S1 . 3 ) ., Because the MI interaction constitutes an additional , rapid intracellular feedback , we intuitively expected an effect on both the patterning speed and accessibility ., To test this hypothesis , we performed dynamical simulations to determine patterning speed , and linear stability analysis about the systems homogeneous steady state ( HSS ) to determine pattern accessibility ., The HSS is defined as the steady state in which all cells have identical concentrations of signaling system components 16 , 26 ., Using dynamical simulations , we first compared how rapidly the LI and LIMI models are able to reach the patterned state from an initially non-patterned state ., Fig . 4CDE shows the dynamics of DSL concentration in single cells for both models with one set of parameters ( black dot in Fig . 5 ) ., The LI model initially spends a considerable time in a nearly homogeneous state ( left of the dashed line in Fig . 4D ) before DSL concentrations diverge ( red and blue curves , right of the dashed line ) ., In contrast , in the LIMI model , DSL concentrations diverge much earlier ( Fig . 4E ) ., The LIMI process approaches the final patterned state more rapidly than the LI process , largely due to the difference in the rate of deviation from homogeneity ., A similar difference in the patterning speed is observed over a large region in parameter space as shown in Fig . S3CDEF ., Why are the dynamics accelerated in the LIMI model ?, A key difference in the LIMI model is the inactivation terms , which are equivalent to effective degradation terms ( e . g . ) ., Because protein degradation is assumed to be the slowest timescale in the system , increasing the degradation rate speeds up the overall response time ., In principle such acceleration could be achieved in the LI model as well , just by increasing the magnitude of the constitutive degradation terms ., Note , however , that in the LIMI model the additional degradation only occurs when both Notch and DSL are simultaneously present on the same cell ., This causes an acceleration specifically during patterning , while avoiding unnecessary protein turnover that would result from increased constitutive degradation ., The potential for lateral inhibition pattern formation in a given system is strongly controlled by its dynamical behavior near the HSS ., For some parameter sets , the HSS is stable and no patterning occurs ., For other parameter sets , the HSS is unstable ., In this case , although components concentrations may initially approach their HSS values , in the presence of even arbitrarily small heterogeneous fluctuations they must subsequently diverge , generating the patterned state ( Fig . 4C ) ., We next set out to systematically compare the patterning ability of the LI and LIMI models ., We performed linear stability analysis of the HSS 16 , 26 across a broad range of parameter values , and determined the subset of parameter sets for which the systems HSS is unstable to perturbations ( Fig . 5A–D ) ., Formally , this is done by calculating the maximal escape rate from the non-patterned HSS ( Supplementary Information S4 ) ., If this rate , termed the Maximal Lyapunov Exponent ( MLE ) , is positive , the HSS becomes unstable and patterning occurs ., In Fig . 5A–D we plot the MLE as a function of the production rates and for two different effective cooperativities , for both the LI and LIMI models ., At high cooperativity ( ) , both models show a large region of parameter space in which the system patterns ( ) ( Fig . 5AB ) , although quantitatively the LIMI MLE is generally greater than the LI MLE ., In contrast , when , only the LIMI model supports patterning anywhere in the parameter space ( Fig . 5CD ) ., Thus , the mutual inactivation model circumvents the requirement for cooperative regulatory feedback in the standard lateral inhibition model ., The qualitative behavior of Fig . 5A–D is maintained as long as the cis interaction is strong enough ( ) ., Mutual inactivation can have a more dramatic effect on patterning: Besides improving the performance of standard patterning circuits , it can enable an altogether different , and simpler , lateral inhibition circuit architecture ., The essential requirement for lateral inhibition is that increased Notch activity in one cell reduces its ability to signal to its neighbors ., In the presence of mutual inactivation , one way to achieve this is for Notch activity to directly up-regulate Notch expression ( Fig . 6A ) ., Increased levels of Notch result in more rapid removal of DSL through the mutual inactivation interaction , effectively down-regulating it ., Thus , a circuit in which Notch activates its own expression implements lateral inhibition with only a single level of transcriptional feedback , i . e . instead of Notch activating a repressor of DSL , there is direct downregulation of DSL through the mutual inactivation interaction ., This type of autoregulation has been observed in some cases , such as the C . elegans AC/VU fate determination system 27 ., We term this circuit architecture ‘Simplest Lateral Inhibition with Mutual Inactivation’ ( SLIMI ) ., Linear stability analysis of this SLIMI circuit ( Fig . 6B ) shows that patterning can occur across a broad range of parameter values ., Moreover , as with the LIMI model , SLIMI does not require explicit cooperativity for patterning ., Thus , lateral inhibition can be achieved with a startlingly simple circuit architecture ., In the wing vein boundary , graded expression of Delta is converted to two sharply defined ‘side bands’ of Notch activity 1 , 2 ., The mutual inactivation mechanism achieves this conversion without requiring additional circuit components ., Furthermore , unlike a broad class of alternative models based on transcriptional cooperativity ( e . g . the BP model ) , the MI model can generate sharp boundaries over a wide range of gradient profiles and biochemical parameters ( Fig . 2DE ) ., The MI model has a unique property that can experimentally distinguish it from other models: The pattern of expression of Notch target genes depends on the relative expression levels of Notch and DSL rather than on their absolute concentrations ( Figs . 3A , S2 ) ., This property can explain the ratiometric behavior observed in Notch and Delta heterozygous mutants 19 , 20 ( Fig . S2 ) ., Interestingly , when DSL expression in our model is reduced below Notch expression level everywhere , very little signaling occurs ( below the blue line in Fig . 3A ) ., In this condition Notch signaling is no longer expected to restrict vein width , resulting in a broader vein with diffuse boundaries 2 ., This leads to the following experimental prediction: by reducing Delta production continuously , the width of the veins should first decrease as the crossing points between Notch and Delta production rates move toward the center of the vein ., However , this thinning should be followed by an abrupt switch to the unrestricted ( wider ) vein regime once ( Fig . 3A ) ., The same ratiometric behavior also underlies the dependence of the pattern on noise ( Fig . 3BC ) : while the width of the boundary is sensitive to intrinsic noise ( uncorrelated between Notch and DSL ) it is robust to extrinsic noise ( correlated between Notch and DSL ) ., Experimental measurements of the correlations between Notch and Delta expression in wing discs ( or other systems ) would help to determine which noise regime is most relevant in vivo ., We note that transcriptional feedback of Notch signaling on Notch and Delta expression has been shown to occur in the Drosophila wing vein boundary 1 , 2 ., Here we have omitted these feedbacks in order to focus specifically on the effects of mutual inactivation ., However , it is important to note that these feedbacks are not sufficient to explain the experimentally observed ratiometric behavior ( Fig . S10 in ref 11 ) ., Experimental disruption of these feedbacks could help to determine what role they play in patterning , e . g . whether they function to control the pattern itself , to increase its amplitude , or to provide some other functionality in normal development ., Mutual inactivation facilitates lateral inhibition patterning in several ways ., First , mutual inactivation accelerates patterning dynamics compared to an equivalent model without it ( Fig . 4DE ) ., The LIMI model accelerates dynamics by increasing protein turnover , but does so selectively only when both proteins are present on the same cell ., Thus , once patterning is complete , there is no additional protein turnover cost ., Notch has been shown to exhibit relatively fast response times in some systems , and the lifetime of the cleaved intracellular domain of Notch is short and highly regulated 40 , suggesting that the acceleration provided by mutual inactivation could be important in development ., Furthermore , recent work has attributed minimization of errors in patterns of the sensory organ precursors to faster dynamics due to cis-inhibition 21 ., A second advantage is that mutual inactivation removes the requirement that would otherwise exist for an explicitly cooperative step in the lateral inhibition feedback loop ( Fig . 5 ) ., This requirement on the LI model was previously proven analytically both for a 1D chain 16 and a 2D 11 hexagonal lattice ., In fact , mutual inactivation plays a dual role here: in addition to providing the non-linearity required for the amplification of small differences between neighboring cells , it also introduces an additional intracellular feedback reinforcing the intercellular feedback loop ., When Notch signaling down-regulates DSL , this also reduces the rate of Notch inactivation , effectively freeing additional Notch receptors and leading to an additional increase in Notch signaling ., Finally , mutual inactivation allows a new , alternative circuit architecture for lateral inhibition: Instead of transcriptionally down-regulating DSL , Notch can up-regulate its own expression ( Fig . 6A ) ., This architecture is sufficient for lateral inhibition patterning across a broad range of parameters ( Fig . 6B ) ., This alternative architecture is intriguing because in some natural lateral inhibition circuits the regulatory pathway for Notch-dependent down-regulation of DSL remains unclear 41 , 42 ( we note that in other systems downregulation of DSL by Notch has been observed ) ., At the same time , Notch up-regulation by Notch signaling has been shown in several lateral inhibition patterning examples , such as the AC/VU system in C . elegans 27 ., This mechanism may provide the main feedback in lateral inhibition circuits , or may work in combination with the classical lateral inhibition feedback mechanisms on DSL ( LIMI model ) ., It will be interesting to determine to what extent this mechanism participates in various lateral inhibition systems ., In general , mutual inactivation of Notch and DSL in cis may be conceived as a direct , rapid , and sharp replacement for an additional level of intracellular feedback that would otherwise have been required to drive neighboring cells to distinct fates in a fine-grained spatial pattern ., In this sense we may say that an intrinsic difference-promoting logic is encoded in the signaling system itself by the mutual inactivation phenomenon ., Because of this , regulatory circuit architecture that achieves fine-grained patterns without MI can be made less complicated ( i . e . with fewer regulatory levels ) by including MI ., Both examples analyzed here demonstrate this feature ., Together the results above provide a theoretical framework as well as testable hypotheses for the role of mutual inactivation between Notch and DSL in the generation of fine-grained developmental patterns ., In the future , this analysis can be expanded to include additional circuit details such as further regulatory feedbacks , multiple Notch ligands and receptors , and modifiers of Notch signaling , and extended to additional Notch-dependent patterning systems ., ( 1 ), ( 2 ), ( 3 ) ( 4 ), ( 5 ), ( 6 ), Compared to the MI model , these equations remove the cis-inhibition terms from the rates of change in Notch and DSL , and the production rate of the reporter is now the product of two Hill functions , one decreasing and one increasing , with affinity and cooperativity ., ( 7 ), ( 8 ), ( 9 ), The parameters are defined consistently with the above ., In these equations there is no cis-inhibition ., The lateral inhibition is implemented by decreasing the production rate of DSL as a function of signaling Reporter levels , by the factor ., ( 10 ), ( 11 ), ( 12 ), These differ from the LI model only by the inclusion of an additional cis-inhibition degradation term ( ) to the dynamics of both Notch and DSL ., ( 13 ), ( 14 ), Because of the mutual cis-inhibition , upregulation of Notch expression in response to Notch signaling ( represented as an increasing Hill function with strength , affinity , and cooperativity ) can implement lateral inhibition ., Dynamical simulations were performed using Matlabs ode15s solver ( ver . 7 . 6 . 0 , The Mathworks ) ., Figs ., 2DE were generated by solving Eqs ., 1–3 for the MI model and Eqns ., 4–6 in the BP model ., Simulations were performed on a 12x48 hexagonal cell array assuming periodic boundary conditions ., The DSL production profiles used were for the MI model and for the BP model , where are the indicated slopes ., Fig . 3A was generated using Eqns ., 1–3 with DSL production rate profiles given by , where is as indicated in the figure ., Figs ., 3BC were generated using Eqns ., 1–3 with multiplicative ( static ) noise terms for and ., Generation of noise is described in Supporting Information Text S1 . 5 ., Figs ., 4CDE were generated by solving Eqns ., 7–12 ., These simulations were performed on a 12x12 hexagonal cell array assuming periodic boundary conditions ., The MLE values in Fig . 5A–D were calculated by performing linear stability analysis on Eqns ., 7–12 using previously described techniques ( 16 , Supporting Information Text S1 . 4 ) ., Parameters used throughout the analysis are provided in Table S1 . | Introduction, Results, Discussion, Materials and Methods | Developmental patterning requires juxtacrine signaling in order to tightly coordinate the fates of neighboring cells ., Recent work has shown that Notch and Delta , the canonical metazoan juxtacrine signaling receptor and ligand , mutually inactivate each other in the same cell ., This cis-interaction generates mutually exclusive sending and receiving states in individual cells ., It generally remains unclear , however , how this mutual inactivation and the resulting switching behavior can impact developmental patterning circuits ., Here we address this question using mathematical modeling in the context of two canonical pattern formation processes: boundary formation and lateral inhibition ., For boundary formation , in a model motivated by Drosophila wing vein patterning , we find that mutual inactivation allows sharp boundary formation across a broader range of parameters than models lacking mutual inactivation ., This model with mutual inactivation also exhibits robustness to correlated gene expression perturbations ., For lateral inhibition , we find that mutual inactivation speeds up patterning dynamics , relieves the need for cooperative regulatory interactions , and expands the range of parameter values that permit pattern formation , compared to canonical models ., Furthermore , mutual inactivation enables a simple lateral inhibition circuit architecture which requires only a single downstream regulatory step ., Both model systems show how mutual inactivation can facilitate robust fine-grained patterning processes that would be difficult to implement without it , by encoding a difference-promoting feedback within the signaling system itself ., Together , these results provide a framework for analysis of more complex Notch-dependent developmental systems . | Multicellular development requires tightly regulated spatial pattern formation , frequently including the generation of sharp differences over short length scales ., Classic examples include boundary formation in the Drosophila wing veins and lateral inhibition patterning in the differentiation of sensory cells ., These processes and a diverse variety of others are mediated by the Notch signaling system which allows neighboring cells to exchange information , via interaction between the Notch receptor on one cell and its ligands such as Delta , on another ., Interestingly , recent evidence has shown that Notch and Delta within the same cell ( in cis ) also interact , mutually inactivating each other ., However , the significance of this interaction for pattern formation has remained unclear ., Here we show , by analytical and computational modeling , how this cis interaction intrinsically generates a difference-promoting logic that optimizes the system for use in fine-grained pattern formation ., Specifically , boundary formation and lateral inhibition patterning both operate more effectively and with simpler circuit architectures than they could without this interaction ., Our results provide a foundation for understanding these and other Notch-dependent pattern formation processes . | systems biology, developmental biology, biology, computational biology, signaling networks, pattern formation, genetics and genomics | null |
journal.pcbi.1000745 | 2,010 | A Multi-Variant, Viral Dynamic Model of Genotype 1 HCV to Assess the in vivo Evolution of Protease-Inhibitor Resistant Variants | Hepatitis C virus ( HCV ) is estimated to infect 170 million people worldwide 1 ., Current HCV treatment with pegylated interferon ( Peg-IFN ) and ribavirin ( RBV ) for the most common genotype 1 strain requires 48 weeks and only 42% to 50% of patients naïve from treatment achieve sustained viral response ( SVR ) 2 , 3 ., Several specifically-targeted antiviral therapies for HCV ( STAT-C ) are under development 4 ., Telaprevir ( also known as VX-950 ) , is a STAT-C that targets the HCV NS3•4A protease and has demonstrated antiviral activity in an HCV replicon assay 5 and in clinical trials 6 , 7 ., Previously published models of HCV viral dynamics in subjects treated with interferon ( IFN ) , Peg-IFN and RBV have assumed the HCV population within a subject to be relatively homogeneous with respect to sensitivity to these antiviral agents 8 , 9 , 10 , 11 ., However , as a consequence of its high replication rate and error-prone polymerase , HCV exists as a quasispecies ., In fact , recent data from clinical trials evaluating HCV protease inhibitors have revealed the presence of viral variants with varying levels of sensitivity to these agents 12 , 13 , 14 , 15 ., Viral variants have also been detected at levels around 10−3 of wild-type NS3•4A HCV ( WT ) level prior to dosing in treatment-naïve subjects 12 , 13 ., Upon exposure to protease inhibitors , the composition of the HCV quasispecies was altered , as revealed by sequencing of plasma HCV RNA and isolated viral clones obtained from subjects dosed with telaprevir 14 , 15 and boceprevir 13 ., These variants have also been reported to exhibit reduced fitness 16 , 17 and reduced susceptibility to other protease inhibitors in vitro 18 ., Models of viral dynamics and emergence of resistance have been developed for viruses like HIV that exhibit a high degree of genetic variability and are capable of establishing chronic infections 19 , 20 , 21 ., In these models , variants were assigned different replicative rates , based either on their infection rate constants , production rate constants , or both ., Typically , these models were parameterized using on-treatment HIV RNA levels and CD4+ counts for a small number of resistant variants; however , many of the models did not include sufficient data , in particular the prevalence of variants , to allow estimation of model parameters with good precision ., Here , a multi-variant model was developed to represent HCV viral dynamics in subjects dosed with telaprevir monotherapy , to estimate the fitness of variants resistant to telaprevir , and to investigate the importance of replication space dynamics , mutations during treatment , and pre-existing variants on the overall response ., The study protocol and informed consent form ( ICF ) were reviewed and approved by an Independent Ethics Committee ( IEC ) at each of the 3 study centers before initiation of the study ., The sites are: Pharma Bio-Research Group BV Medisch Ethische Toetsings Commissie METC Stichting Beoordeling Ethiek Bio-Medisch Onnderzoek P . O . Box 1004 9400 BA Assen , Amsterdam Medical Center Medisch Ethische Toetsings Commissie METC Stichting Beoordeling Ethiek Bio-Medisch Onderzoek P . O . Box 1004 9400 BA ASSEN , The Netherlands Medisch Ethics Toetsings Commissie Meibergdreef 9 P . O . Box 22660 NL 1100 DD Amsterdam , Saarland University Hospital Ärztekammer des Saarlandes Ethikkommission Faktoreistraβe 4 66111 Saarbrücken Germany ., Written informed consent was obtained in accordance with the Declaration of Helsinski ., Thirty-four subjects with HCV genotype 1 were enrolled in Study VX04-950-101 , a randomized , double-blind , placebo-controlled , 14-day , multi-dose , Phase 1b trial ., Subjects received placebo ( n\u200a=\u200a6 ) or one of the following dosages of telaprevir administered as a suspension: 450 mg every 8 hours ( n\u200a=\u200a10 ) , 750 mg every 8 hours ( n\u200a=\u200a8 ) , or 1250 mg every 12 hours ( n\u200a=\u200a10 ) ., Subjects baseline characteristics are provided in Supplementary Table S1 ., Variants were detected using clonal sequencing ( details provided in 14 ) ., For the model parameterization described here , data from 26 of the 28 subjects dosed with telaprevir were used ., No variants were detected in one subject , and therefore this subject was excluded from further analysis ., Estimation results in another subject with 8 variants did not converge to a global optimum---a standard requirement for computationally rigorous estimation; this subject was also excluded ., For each subject , we examined only variants identified by clonal sequencing that were present either at ≥5% of the HCV population at 2 measurement points or ≥10% of the HCV population at 1 time point ( 5% is the detection limit of the clonal sequencing measurement performed here ) ., The number of variants per subject ranged from 2 to 6; the number of variants for each subject is provided in Supplementary Table S2 ., These clonal sequencing results identified amino acid differences in HCV NS3•4A that correlated with changes in telaprevir resistance in vitro ., A larger network representation of quasispecies containing an even greater number of variants could have provided a more complete picture , but was not examined here because no in-subject kinetic data were available to estimate their fitness , and/or no in vitro data were available on their resistance to telaprevir ., The basic evolutionary dynamic among HCV resistant variants in subjects dosed with telaprevir follows Equations 1–5 , with variable and parameter descriptions provided in Table 1 . ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) Variant Vi represents a virion with characterized amino-acid substitution ( s ) and in vitro defined telaprevir sensitivity ., Variant Vi infects target cells T to form variant-i-infected cells Ii at rate βTVi ., It is assumed that each infected cell Ii is infected by only one variant , and each variant competes for the same target cells T . The assumed single-variant infection is consistent with the fact that recombination in HCV appears rarely 22 ., Target cells T also represent limited replication “space” shared by all variants ., Target cells T ranges from their baseline level T0 to their maximum level Tmax ., Each infected cell Ii produces a population of variants at production rate pfi , with fraction mi , j mutating to produce variant j ., The mi , i were normalized to follow mi , i + ∑j , j≠ i mi , j\u200a=\u200a1 ., Different production rate constants pfi , but the same infection rate constants ( β ) and clearance rate constants ( c ) are assumed for different variants ., The assumption of different production pfi is consistent with the function of the NS3•4A protease in cleaving a precursor polyprotein 23 , and with variants having been observed with reduced catalytic activity in vitro ( data not shown ) ., The production rate ratio fi quantifies variant i replication disadvantage in the absence of telaprevir ., In the presence of telaprevir , the viral production was further reduced by a factor ( 1- εi ) ., The assumed same infection β is consistent with lack of interference between HCV protease and HCV envelope proteins ., The assumed same plasma clearance c is consistent with the data in interferon-based and telaprevir-based treatments 24 ., Despite large differences in the antiviral blockages between both treatment groups , similar c values were observed ., This suggests that c is independent of antiviral blockage and therefore , may have the same value among variants ., Alternative model formulations with different variants fitness assigned to infection rates or to plasma virion clearance rates produced similar dynamics ( Supplementary Figure S3 ) ., Antiviral activities of telaprevir were implemented by assuming a dual role ., Telaprevir blocks the production of HCV by inhibiting the activity of the NS3•4A protease with blockage factors εi calculated using Equation, 4 . The blockage factors for all variants within a subject were calculated using a single effective telaprevir concentration TVR , with its value estimated from the HCV RNA , variant prevalence dynamics within each subject , and in vitro susceptibility of variants to telaprevir ., The susceptibility factor IC50 , i and Hill coefficient hi were estimated from in vitro susceptibility of variant i to telaprevir 16 , 17 and were provided in Supplementary Table S3 ., The second role of telaprevir is to enhance infected cell clearance δ , a parameter contributing to the second-phase decline ., WT δWT values were up to 10-times higher in subjects dosed with telaprevir than in subjects treated with Peg-IFN/RBV 24; only a 0 . 2-fold increase in the second-phase decline is explained by increased telaprevir blockage alone ( detail calculations are provided in the Supplementary Text S1 ) ., On the other hand , in the limit when a subject is not dosed with telaprevir , δi should converge to the clearance without drugs δnodrug ., These observations were incorporated into the model by assuming that δi increased proportionally to the logarithmic of blockage factor ( 1-εi ) , given in Equation, 5 . We also examined alternative models to Equation 5 , given by Equations 6 or 7 . ( 6 ) ( 7 ) Prior to dosing , the differential equations were initialized at steady-state ., The steady-state initialization is consistent with years of chronic HCV infection ., This steady-state solution was used to predict the pre-dosing variant prevalence ., During dosing with telaprevir , replication rates of WT and variants were reduced by factors proportional to their sensitivity to telaprevir ( blockage factors ) ., Following completion of telaprevir dosing , these blockages were removed ., Consequently , the WT and variants present would compete for available replication space with competitive advantages governed by fitness of WT and variants in the absence of any drug ., The majority of the results were reported with replication space T described by Equation 1 ., We also examined another representation of T given by Equation 8 10 ., The values of T0 and I0 , i were fixed prior to estimation to the steady-state values of T and Ii obtained from models with Equation 1 ., To obtain a similar rate of T increase , the regeneration rate γ were related to parameters in Equation 1 by γ\u200a=\u200as/ ( T0+∑i I0 , i ) ., ( 8 ) Previously reported HCV mutation rates range from 1 . 5×10−3 nucleotide changes/site/y 25 to 5×10−3 nucleotide changes/site/y 26 ., These values were converted to per nucleotide position per replication cycle by assuming an average duration of the HCV replication cycle of 9 . 5 days calculated as ( 1/c+1/δ ) with typical values for c and δ assumed to be the same as those from Peg-IFN/RBV treatments 8 , 9 ., These calculations resulted in a mutation rate m of 1 . 2×10−4 nucleotide changes/site/cycle ., The estimations were repeated for different mutation rates of 1 . 2×10−2 , 1 . 2×10−3 and 1 . 2×10−5 nucleotide changes/site/cycle ., The mutation rates were computed prior to each estimation by assuming a rate of 1 . 2×10−4 per nucleotide position per replication cycle ., The specific mutation rates between two variants were computed by exponentiating the mutation rate for a single mutation by the number of nucleotide mutations between these variants ., These rates were genotype specific ., For example , to produce NS3•4A protease mutation at position 36 V36M , genotype 1a requires a single nucleotide mutation ( from codon GTG to ATG ) , while genotype 1b requires two mutations ( from GTT to ATG ) ., During-dosing and post-dosing HCV RNA levels and post-dosing variant prevalence data from the clinical study described above ( previously published in 6 , 14 ) were used simultaneously to estimate model parameters ., This simultaneous estimation allowed fitness estimation in subjects with HCV RNA levels below the clonal sequencing detection limit ( 100 IU/mL ) at the end of telaprevir dosing but with detectable HCV RNA within a week after completion of dosing ., The estimation minimized the maximum likelihood objective function ., Parameters estimated for each subject included c , δ1 , p , fi , and TVR ., Parameter bounds are provided in Supplementary Table S4 ., Fitness parameter fi was estimated for each variant; the number of assessed variants for each subject varies between 2 to 6 ( Supplementary Table S2 ) ., The remaining parameters were pre-computed prior to estimation runs , including mi , j , β\u200a=\u200a0 . 05 h−1 , and δnodrug\u200a=\u200a0 . 12 d−1 , assuming that the clearance without drugs δnodrug is the same as the clearance on Peg-IFN/RBV treatment ., Because we do not have direct measurement of target cells nor productively infected cells , we were able to estimate only the overall viral replication rates , or the basic reproductive ratio R0 , WT\u200a= pβTmax/ ( cδ ) ., If measurements of infected cells and target cells were available , parameters p , β , s , and d may be adjusted to match the measured infected and target cells while maintaining the constraint that R0 , WT remains constant ., For a given R0 , WT value , some degree of freedoms exist in choosing p , β , s , and d , while clearance parameters c and δ would be constrained by the decline kinetics ., Numerically , this is implemented by fixing parameters β and the s/d ratio , and normalized HCV RNA levels to the baseline value similar to the implementation in 19 ., In the base run , we chose β\u200a=\u200a0 . 05 h−1 and Tmax\u200a= s/d\u200a=\u200a10 , and obtained R0 , WT from estimates of p ., These choices of β and Tmax prior to estimation did not change the estimates of other parameters estimated from these data , including clearance rates c , δ , R0 , WT , and fitness fi ., For example , re-estimation with different β values , shown in Supplementary Figure S2 , resulted in similar fitness and R0 , WT values ., The results also demonstrated an inverse relationship between estimated p and assumed β , supporting the fact that we were only able to estimate R0 , WT , but not p nor β individually ., The breakdown of replication parameters constituting R0 , WT ( p , β and Tmax ) may be refined in future studies when direct measurements of target and infected cells become available ., Susceptibility factors IC50 , i and hi , were fixed during each estimation ., The robustness of the estimates as a function of the dynamics of target cells T synthesis rate s was also examined by comparing the case when s was estimated to the case when s was fixed to 1 h−1 ., In addition to the modeling approach described in details , we also computed relative fitness ( RF ) ., For a variant i at two consecutive times t1 and t2 with with viral loads Vit1 and Vit2 , the RF was computed from data from the equation below ., If the prevalence was below the detection limit , the value was assumed to be at the limit ( 5% in this study ) ., Model-derived RF was computed similarly , except that rather than evaluating viral load changes at two consecutive times , the change was evaluated at a specified time t using time-derivatives ., The simulations were implemented by normalizing the plasma virion value with the baseline values obtained after solving the steady-state initial condition ., The clearance and replication rates , the balance of which is implicit in the baseline viral load , were estimated directly from HCV RNA decline ( during dosing ) and rebound ( after dosing ) ., The simulation and estimation were implemented using Jacobian Software ( R ) ( RES Group Inc . ) , using methods described in 27 , 28 , 29 , 30 , 31 ., Additional information is provided in Supplementary Text S1 ., A parameterized multi-variant viral dynamic model was developed to represent the antiviral responses of subjects to telaprevir and to estimate the fitness of variants resistant to telaprevir ., Descriptions and schematic of the model is shown in Figure 1 ., The list of major variants and the nucleotide distances between them are shown in Figure 1a ., A schematic of the model is shown in Figure 1b and described by Equations 1–5 ., Replicative fitness of variants was represented by their different production rate constants pfi ., The basic reproductive ratio of WT HCV R0 , WT , relative fitness fi , and clearances c , δ were estimated from time series of both plasma HCV RNA and variant prevalences , with details provided in the Methods section ., The results of the best-fit model corresponded well with observed data ., Results in Figure 1c and Figure 1d show an example of a subject who received 450 mg telaprevir q8h for 2 weeks , and whose plasma HCV RNA levels rebounded on dosing ., The correspondence between data and model for additional subjects is provided in Supplementary Figure S1 ., Assuming a pre-dosing steady-state , the best-fit model predicted that HCV with WT NS3•4A protease dominated the HCV quasispecies population as the most fit variant in the absence of any drug , and variant prevalences were near the levels predicted from HCV mutation rates ., Upon dosing with telaprevir , WT declined rapidly ., Resistant variants also declined initially because of reduced influx of new mutations from WT and , in variants with low-level resistance , because of blockage by telaprevir ( the assessment of relative contribution is provided in later section ) ., As the total viral load declined , available replication space was predicted to increase ., Such an increase , along with sufficient on-dosing fitness of variants , is necessary for emergence of variants ., Had T remained constant , none of the variants would have rebounded because each variant would have had a reduced replication flux due to the reduced number of WT available to generate the variant ., For the subject shown in Figure 1c and Figure 1d , variants with a single mutation ( V36M or R155K ) or a double mutation ( V36M/R155K ) within their NS3•4A protease were responsible for the increase in HCV RNA levels detected initially on Day 6 ., WT levels were predicted to increase again around Day 8 because of back mutations from variants ., When telaprevir dosing was stopped , WT , V36M , and R155K variants out-competed the V36M/R155K variant , and WT eventually regained dominance of the HCV quasispecies population to reach a predicted level of ≥95% of the viral population in 300 days , although V36M persisted for up to 200 days in this subject ., The model predicted that immediately after dosing was stopped , V36M initially out-competed WT for available replication space because it was relatively fit and it was 104-times more prevalent ., V36M persisted because infected-cell clearance was relatively slow ., The estimated fitness obtained from 26 subjects suggests reduced replicative capacity of all telaprevir-resistant variants analyzed compared to WT ., Table 2 summarizes estimated production rate ratio ( f ) for all variants , which ranges from 1% to 68% of WT replication ., Estimation errors associated with the fitness values were reported as the standard deviation of the estimation error; the errors ranged between 0 . 03 and 0 . 12 for variants detected in ≥4 patients , and between 0 . 4 and 3 . 26 for variants detected in ≤2 patients ., The estimate errors were large especially for variants detected in ≤2 patients that the fitness estimates must be interpreted with cautions ., The variants were sorted based on their resistance to telaprevir as measured in replicon cells ., The first 7 variants ( R155M to A156S ) are low-level resistant variants ( defined as variants with IC50 ≤ the mean estimated effective telaprevir concentration in vivo when telaprevir is dosed orally at 750 mg q8h ) ., Among all variants , V36M and R155K single mutant variants with low-level telaprevir resistance have the highest f values of 0 . 68 and 0 . 66 , respectively ., Among the high-level resistant variants , the double mutant V36M/R155K had the highest f of 0 . 51 ., Previously , we reported fitness estimates based on variants growth in a subset of subjects 14 , using a method similar to the relative fitness ( RF ) approach 32 , 33 , normalized to 0–100 scale ., In contrast to our earlier RF estimates 14 , the fitness estimates herein included data of on-dosing HCV RNA and 3–7 month clonal sequencing , and was expanded to include more subjects ( n\u200a=\u200a8 for previous estimates , n\u200a=\u200a26 for current estimates ) ., The correspondence between these previously reported in vivo fitness estimates , the current in vivo fitness estimates presented herein , and the in vitro fitness estimates reported by others 16 , 17 are provided in Figure 2 ., Figure 2a shows the correspondence of the current in vivo estimates to the in vitro fitness estimates ., With the exception of estimates for variant V36A ( which appeared to be an outlier in the in vitro estimates ) , all these fitness estimates were in good agreement ., In contrast , Figure 2b shows correspondence of our previous in vivo RF estimates 14 with the in vitro estimates: the in vivo fitness estimates of many of the variants were higher than those of in vitro fitness ., Thus , compared to the RF method , the fitness estimates presented herein were more consistent with both clinical data 14 and the fitness estimates as measured in vitro ., For example , current estimates suggest that variant V36M/R155K is less fit than variant R155K , but RF estimates suggested otherwise ., The reduced fitness of V36M/R155K is consistent with the data that show increased prevalence at later times ( Day 21–23 vs . Day 14; Month 3–7 vs . Day 21–23 ) of WT , V36M and R155K as compared to the decreased prevalence of V36M/R155K ., An example from a subject ( Subject 2 , Supplementary Figure S1a and b ) demonstrated why fitness estimated from the current model corresponded better with all data than the fitness estimated using the RF method 14 ., For this subject , the clonal sequencing data show that although both WT and V36A comprised <5% of the population at Day 14 , both were detectable at Day 21 , with V36A prevalence relative to that of WT increasing between Days 14 and 21 ., The current model predicted that at Day 14 , when telaprevir dosing was discontinued , V36A HCV RNA levels were about 2-log higher than those of WT ., Thus , despite the reduced fitness of V36A over WT , V36A would continue to infect the majority of the target cells T between Days 14 and 21 ., However , by Day 150 , WT dominated the quasispecies because of its higher fitness ., If we computed RF of V36A ( vs . WT ) based only on prevalence data at Day 14 and 21 ( RFD14 vs . D21 , V36A vs . WT ) and , assuming the same Day 14 prevalence levels of 5% , then RFD14 vs . D21 , V36A vs . WT =\u200a0 . 337 , a value >0 that misleadingly implies that V36A is more fit than WT ., However , this conclusion is inconsistent with the RF calculated between Day 21 and Day 150 ( RFD21 vs . D150 , V36A vs . WT ) of −0 . 101 , a value <0 which implies that WT is more fit than V36A ., On the other hand , the modeling herein estimated fV36A\u200a=\u200a0 . 578 , and by using the on-dosing data , correctly accounts for higher V36A levels than WT levels at Day 14 , higher V36A levels at Day 21 , and reduced levels of V36A at Day 150 ., Based on the current modeling results , we also can calculate RF values at specific timepoints using a time-derivative of the HCV RNA levels ( model-derived RF , Table 3 ) ., For the V36A variant , the Day 15 model-derived RF was much higher than the RF at Day 100 ( −0 . 671 vs . 0 . 023 ) , demonstrating dependency of RF values on the specific timing of sample collection ., This example demonstrates the advantages of the modeling approach for estimating variants fitness ., To examine the likelihood of these resistant variants pre-existing before dosing , the time necessary to generate these variants , if they did not pre-exist , was estimated ., The best-fit model for Subject 1 above was reinitialized with an HCV population consisting only of WT before dosing started ., The results are provided in Figure 3 ., Had the HCV population consisted only of WT before dosing , the predicted HCV RNA rebound on dosing would be delayed compared to the observed rebound ., The poorer ( delayed ) fit of this modified simulation compared to the one started with a steady-state level of variants before dosing further highlights the likelihood that these variants pre-exist at a steady-state level ., Because most subjects in this study have been infected with HCV for years , plenty of time was available for the variants to reach their steady-state levels prior to dosing ., To understand the contribution of replication space dynamics to the rebound dynamics of resistant variants , we examined three cases:, 1 ) target cells T followed Equation 1 with synthesis rate s estimated ,, 2 ) T followed Equation 1 with s fixed to its upper bound ( 1 h−1 ) , and, 3 ) T followed Equation 8 ., For these three cases applied to Subject 1 , all models corresponded well with observed data ( Figure 4 ) , suggesting robustness of the models to these assumptions of T dynamics ., For T dynamics represented by Equation 1 , increasing s implies faster dynamics for target cells to reach their maximum levels ( Tmax ) , resulting in an earlier HCV RNA rebounds ., The objective values for the first two cases ( s estimated and s fixed ) for all subjects are shown in Figure 5a ., The third case was not applied to all subjects because this case has different underlying equations compared to the first two cases --- making a comparison of objective functions difficult ., Both cases of s produced similar objective values , suggesting robustness of the model fits to the s values in the range examined ., The estimated s values ( Figure 5b ) had a logarithmic median of 10−0 . 88 h−1 , a value comparable to the regeneration rate of liver tissues ( 10−0 . 3–10−0 . 6 h−1 ) 34 ., The optimal estimates of R0 , WT varied with s , with median ( range ) log10 values of 1 . 66 ( 0 . 93 , 3 . 43 ) and 1 . 49 ( 0 . 85 , 3 . 43 ) for estimated s and fixed s cases respectively , and higher s correlated to lower estimates of R0 , WT ( Figure 5c ) ., However , the estimates of the production rate ratio fi were more robust to the s values ( Figure 5d–f ) ., In the base runs , the mutation rates were assumed to be random with no effect of evolutionary selection , using a value reported from data including evolutionary selection 26 ., The inclusion of evolutionary selection may result in mutation rates that underestimated actual rates ., Because of this difference in the inclusion of selection to the assumed base mutation rate , we further examined the sensitivity of the estimation results to the assumed rates by repeating the estimations with 10-fold lower , 10-fold and 100-fold higher rates ., The results were provided in Figure 6 ., The objective function values were the lowest for the base case with m of 1 . 2×10−4 changes/site/cycle , suggesting that this rate produced the best correspondence between data and model ., The ranking of estimated fitness fi was qualitatively similar in the three lowest m values explored , suggesting that the fitness estimates were robust to the assumption of m values in the range of 1 . 2×10−5 , 1 . 2×10−3 changes/site/cycle ( fi for the highest m were not reported because of poor model fits ) ., The fitness estimates of the double mutant variants ( V36M/R155K , V36M/T54S ) were affected the most by m; lower m produced higher estimates of production rate ratio f for these variants ., This relationship could be explained by the fact that lower m corresponded to lower pre-dosing levels of these double mutants , and to correspond to the levels measured at Day 14 , the estimation converged to faster estimates of replication rates ( or higher f values ) of these mutants ., One feature distinguishing the model proposed here from that previously proposed in HIV 19 is the increase of infected-cell clearance δ as a function of on-dosing blockage , given by Equation 5 ., The higher δ was motivated by the observation that WT δ values were up to 10-times higher in subjects dosed with telaprevir than in subjects treated with Peg-IFN/RBV 24 ., This second-phase decline is much more rapid than the 0 . 2-fold increase in the decline predicted by increased telaprevir blockage alone , a lower bound value calculated when T is held constant ( similar differences in the estimates were also obtained when T was allowed to vary 24 ) ., On the other hand , in the limit of no telaprevir , δi should converge to δ0 ., These two limits constrain alternative relationships between δ and on-dosing blockage ., For the base case , we chose a model in which δi decreases linearly with log10 ( 1-εi ) as given in Equation 5 ., To examine the contribution of δ enhancement , we compared the correspondence between data and model fits with and without δ enhancement ., The model with δ enhancement ( δdrug estimated , δnodrug was assumed to be the same as that from Peg-IFN/RBV treatments , and was fixed to the mean value of 0 . 005 h−1 ) was compared to that without δ enhancement ( δdrug\u200a=\u200a0 , δnodrug estimated ) ; both models have the same number of estimated parameters ., The results are shown in Figure 7 ., Figure 7a shows the objective values for both models ., With the same degree of freedom in both models , the model with δ enhancement tends to have lower objective values than that without δ enhancement , suggesting better correspondence of the δ enhancement model to the data ., Figure 7b and c show the example of the correspondence of model without enhancement applied to Subject 1 ., The model without enhancement missed the much more rapid on-dosing HCV RNA declines while maintaining the fit to variant prevalence data ., These observations suggest that δ enhancement affects both WT and variants ., To represent the δ enhancement in WT and variants , while satisfying the two limits of the observed second slope described above , we also examined alternative equations to Equation 5 ., In particular , we examined models with δi as a linear function of ( 1-εi ) ( Equation 6 ) and models with δi as a step-function to the presence of telaprevir ( Equation 7 ) ., Because the number of parameters estimated in each patient are the same , one may compare the objective functions directly to represent the goodness of fit ., The results are provided in Figure 8 ., Compared to the base model with Equation 5 , the model including Equation 6 produced similar quality of fit ., However , the model with Equation 7 produced inferior fits ., These suggest that the increase in the second slope is blockage-dependent ., Upon dosing with telaprevir , variants RNA levels were predicted to decline initially because of two factors: blockage of replication by telaprevir and reduced influx mutations from WT due to rapid WT clearance ., To examine the relative contributions of these two factors , the model components were examined at the beginning of dosing ., At this timepoint , the reduction of variant is replication flux by telaprevir blockage can be approximated as -εi fi Ii , and the reduction of influx mutation by WT clearance can be approximated as mWT , i fWT IWT ., Thus , assuming that prior to dosing Vi/VWT = mWT , i/ ( 1-fi ) , the ratio of reduced replication of variant i by blockage to that by mutation can be approximated by εi fi/ ( 1-fi ) ., If we assumed an effective telaprevir concentration TVR of 4 µM ( the average TVR in the cohorts studied here ) and calculated ε from Equation 5 , then for the V36M variant with intermediate susceptibility ( IC50\u200a=\u200a4 . 73 µM , Hill coefficient h\u200a=\u200a3 . 5 , fV36M\u200a=\u200a0 . 68 ) , this ratio was 0 . 97 ., For the A156T variant with high resistance and low fitness ( IC50 =\u200a103 µM , h\u200a=\u200a1 , fA156T\u200a=\u200a0 . 1 ) , the ratio was 4×10−4 ., For the V36M/R155K variant with high resistance and high fitness ( IC50\u200a=\u200a142 µM , h\u200a=\u200a3 . 5 , fV36M/R155K\u200a=\u200a0 . 5 ) , this ratio is 4×10−6 ., The fact that these | Introduction, Methods, Results, Discussion | Variants resistant to compounds specifically targeting HCV are observed in clinical trials ., A multi-variant viral dynamic model was developed to quantify the evolution and in vivo fitness of variants in subjects dosed with monotherapy of an HCV protease inhibitor , telaprevir ., Variant fitness was estimated using a model in which variants were selected by competition for shared limited replication space ., Fitness was represented in the absence of telaprevir by different variant production rate constants and in the presence of telaprevir by additional antiviral blockage by telaprevir ., Model parameters , including rate constants for viral production , clearance , and effective telaprevir concentration , were estimated from, 1 ) plasma HCV RNA levels of subjects before , during , and after dosing ,, 2 ) post-dosing prevalence of plasma variants from subjects , and, 3 ) sensitivity of variants to telaprevir in the HCV replicon ., The model provided a good fit to plasma HCV RNA levels observed both during and after telaprevir dosing , as well as to variant prevalence observed after telaprevir dosing ., After an initial sharp decline in HCV RNA levels during dosing with telaprevir , HCV RNA levels increased in some subjects ., The model predicted this increase to be caused by pre-existing variants with sufficient fitness to expand once available replication space increased due to rapid clearance of wild-type ( WT ) virus ., The average replicative fitness estimates in the absence of telaprevir ranged from 1% to 68% of WT fitness ., Compared to the relative fitness method , the in vivo estimates from the viral dynamic model corresponded more closely to in vitro replicon data , as well as to qualitative behaviors observed in both on-dosing and long-term post-dosing clinical data ., The modeling fitness estimates were robust in sensitivity analyses in which the restoration dynamics of replication space and assumptions of HCV mutation rates were varied . | Hepatitis C virus ( HCV ) infects an estimated 170 million people worldwide ., Current treatment for HCV is 48 weeks of peginterferon and ribavirin of which patient response has large variability ., Recently , specifically targeted antiviral therapies for HCV ( STAT-C ) are under clinical development and have shown potentials to improve response ., Within a patient , HCV exists as quasispecies consisting of multiple variants ., Models of HCV dynamics in response to peginterferon and ribavirin treatment have been proposed elsewhere , with HCV quasispecies assumed to respond homogenously to treatment ., However , some of the HCV variants possess different degrees of sensitivity to a STAT-C compound , and therefore , selections and competitions among variants have been observed in patients treated with STAT-C ., We have developed a viral dynamic model that quantifies the evolution of multiple variants in patients dosed in monotherapy with telaprevir , a compound specifically designed to inhibit HCV NS3 . 4A protease ., Our novel modeling approach integrated data from both in vitro and in patients , both during and after dosing with telaprevir ., Our model quantified the antiviral response to telaprevir and the in vivo fitness of variants ., The model provides a useful framework for the designs of STAT-C during clinical development and for understanding the consequences of failure to STAT-C . | virology/persistence and latency, pharmacology/drug resistance, virology/new therapies, including antivirals and immunotherapy, pharmacology/drug development, virology, computational biology/evolutionary modeling, virology/antivirals, including modes of action and resistance, pharmacology/personalized medicine, computational biology/systems biology | null |
journal.pntd.0003608 | 2,015 | Prevalence and Diversity of Small Mammal-Associated Bartonella Species in Rural and Urban Kenya | Bartonella species are Gram-negative haemotrophic bacteria that infect mammalian erythrocytes and are transmitted between hosts by blood-sucking arthropods ., Over 30 species of Bartonella have been described and members of this genus infect a broad range of mammalian hosts including rodents , bats , carnivores and ruminants 1 ., Arthropod vectors including fleas , sandflies , lice , ticks , bat flies and ked flies are implicated in the transmission of these pathogens 2–4 ., The genus Bartonella has a global distribution ., The Bartonella elizabethae complex includes several Bartonella genotypes and strains ( including B . elizabethae , B . tribocorum , B . rattimassiliensis and B . queenslandensis ) that have been isolated from Rattus and Bandicota species around the world 1 ., Recent analyses indicate that this complex has south-east Asian origins and has been globally dispersed by Rattus species 5 ., Several Bartonella species are recognized as human pathogens that cause diverse clinical presentations 6 ., Among rodent-associated Bartonella species , B . elizabethae is a known cause of human endocarditis 7 ., Other rodent-associated species including B . tribocorum , B . vinsonii subsp ., arupensis , B . washoensis and B . alsatica have been associated with a range of symptoms in humans including fatigue , muscle and joint pain , and serious complications , such as endocarditis and neurological signs , particularly in immunocompromised patients 8 , 9 ., Bartonella species have been identified as important causes of febrile illness in some settings ., In two studies conducted in Thailand , 15% of febrile patients were diagnosed with confirmed Bartonella infection based on a four-fold rise in antibody titres , and six different Bartonella species were identified by culture from blood clots collected from febrile patients 10 , 11 ., Non-specific clinical signs and difficulties in culturing the organism present substantial challenges to the diagnosis of bartonellosis ., Consequently , Bartonella species may well be under-recognized as a cause of human disease 12 ., This is particularly true for Africa , where very few data on the etiology of febrile illness are currently available 13 ., In the Democratic Republic of Congo ( DRC ) , a seroprevalence study identified IgG antibodies against Bartonella ( B . henselae , B . quintana or B . clarridgeiae ) in 4 . 5% of febrile patients 14 ., Bartonella bacteraemia was detected by PCR in 10% of HIV-positive patients in South Africa 15 ., Apart from these studies however , there is little information on the impact of Bartonella on human health on the African continent ., A variety of Bartonella species have been detected in animal and ectoparasite populations in Africa ., Considering rodents and small mammals specifically , B . elizabethae and two other Bartonella lineages were detected in Namaqua rock mice sampled in South Africa , where 44% of the 100 individuals sampled were positive by PCR for Bartonella species 16 ., B . elizabethae , B . tribocorum and a Bartonella species with intermediate species classification based on sequence data were detected in 28% of rodents and hedgehogs ( n = 75 ) sampled in Algeria 17 ., B . elizabethae , B . tribocorum and novel Bartonella species were also detected in rodents sampled in the Democratic Republic of Congo ( DRC ) and Tanzania 18 ., Small mammals trapped in Ethiopia , had an overall Bartonella infection prevalence of 34% ( n = 529 ) and were infected with multiple genotypes including genotypes very closely related to B . elizabethae 19 ., B . elizabethae has also been detected in invasive and indigenous rodents sampled in Uganda 20 and genotypes related to B . rochalimae , B . grahamii and B . elizabethae have been detected in Mearn’s pouched mice studied in Kenya 21 ., Bartonella have also been detected in fleas collected in Egypt , Morocco , DRC and Uganda 20 , 22–24 ., The first objective of this study was to determine the presence and prevalence of Bartonella infections in small mammals trapped at rural and urban locations in Kenya ., We also aimed to characterize the Bartonella isolates obtained using partial sequences of the citrate synthase ( gltA ) gene and to compare the Bartonella genotypes detected in these distinct Kenyan populations with each other and with Bartonella detected in small mammals in other parts of the world ., Cross-sectional rodent trapping surveys were conducted within two locations: Asembo , a rural area on the northern shore of Lake Victoria in Nyanza Province western Kenya ( Latitude-0 . 1443 , longitude 34 . 3468 ) and Kibera , an urban informal settlement in Nairobi City ( Latitude-1 . 3156 , longitude 36 . 7820 , Fig . 1 ) ., These locations are the study sites for ongoing population-based human health surveillance 25 ., In Asembo , subsistence farming is the primary occupation for 65% of household heads , 13% work in the informal economy and 5% are salaried 25 ., Households are clustered into compounds of closely related family units ., Livestock ownership is common: 44% of Asembo households own cattle and 43% own at least one sheep or goat ., In contrast , in urban Kibera , 53% of heads of household are salaried and 43% work in the informal sector 25 ., Ownership of large livestock species in Kibera is very rare and prohibited by City Council law ., In Asembo , trapping was conducted over the period of July—August 2009 ., Traps were placed at 50 compounds that were a randomly selected subset of livestock-owning compounds enrolled in a larger study of zoonoses epidemiology 26 ., Within each selected compound , five or six medium-sized foldable Sherman traps ( H . B . Sherman Traps Inc . , Tallahassee , FL ) were placed for three or four nights ., Traps were placed in three categories of habitat: within occupied dwellings; within outbuildings , which included unoccupied dwellings , stores , latrines or kitchens separate from the main dwelling; and outside , in areas within the compound yard ., In Kibera , trapping was conducted over the period of September—November 2008 ., The overall study site was divided for this study into five trapping zones of similar area and within each zone a 50m x 50m trapping area was defined ( Fig . 2 ) ., Within each of the five trapping zones , medium-sized foldable Sherman traps were placed for a minimum of two consecutive nights and a maximum of six nights with the aim of trapping approximately 50 rodents per zone ., In Kibera , all traps were placed indoors at 270 occupied dwellings ., All trapped animals from both locations were euthanized by overdose of the inhalant anesthetic halothane and whole blood was collected by cardiocentesis using aseptic technique ., Blood samples were processed to remove serum and the remaining blood clots frozen at -80°C prior to testing ., Blood clots were shipped on dry ice to the Bartonella laboratory at the Centers for Disease Control and Prevention , Fort Collins , Colorado for laboratory testing ., Morphometric data were collected from each trapped animal for species identification at the National Museums of Kenya ., The Asembo small mammals were submitted for archiving under accession numbers NMK 171860—NMK 171922 ., The Kibera rodent population included in this study is as described previously 27 ., Culture was performed using previously described techniques 28 ., Briefly , blood clots were re-suspended 1:4 in brain heart infusion broth supplemented with 5% amphotericin B , then plated onto agar supplemented with 5% sterile rabbit blood and incubated at 35°C in an aerobic atmosphere of 5% carbon dioxide for up to 30 days ., Bacterial colonies were presumptively identified as Bartonella based on their morphology ., Subcultures of Bartonella colonies from the original agar plate were streaked onto secondary agar plates and incubated at the same conditions until sufficient growth was observed , usually between 5 and 7 days ., Pure cultures were harvested and stored in 10% ethanol ., The identity of presumptive Bartonella isolates was confirmed by PCR amplification and sequencing of a specific fragment of the Bartonella citrate synthase ( gltA ) gene ., Crude DNA extracts were obtained from bacterial cultures by heating a heavy suspension of the microorganisms ., Two oligonucleotides ( BhCS . 781 . p and BhCS . 1137 . n ) were used as PCR primers to generate a 379-bp amplicon of the Bartonella gltA gene 29 ., PCR products were separated by 1 . 5% agarose gel electrophoresis and visualized by ethidium bromide staining ., Sequencing reactions were carried out in a PTC 200 Peltier Thermal cycler ( Applied Biosystems; Foster City , California ) using the same primers as the initial PCR assay ., Sequences were analysed using Lasergene 12 Core Suite ( DNASTAR , Madison , WI ) to determine sequence consensus for the gltA amplicons ., Unique gltA sequences generated through this study were submitted to GenBank ( accession numbers KM233484—KM233492 ) ., The Clustal V program within the MegAlign module of Lasergene was used to compare homologous Bartonella gltA sequences generated in this study with others available from the GenBank database ., Phylogenetic trees were constructed using the neighbor-joining method with the Kimura’s 2-parameter distance model and bootstrap calculations were carried out with 1000 replicates ., B . tamiae was used as the outgroup ., A criterion of >96% homology was used to define similarity of study sequences to known Bartonella species 30 ., Generalized linear models were used to examine associations between individual Bartonella infection status ( culture positive or negative for Bartonella ) and host and environmental variables in R ( Version 3 . 0 . 3 ) 31 ., Binomial family models with a logit link function were used and p values ≤ 0 . 05 were considered statistically significant ., Variables examined included host species , sex , mass and trapping location ., Data from the Asembo ( S1 Table ) and Kibera ( S2 Table ) sites were analysed separately ., Written informed consent for trapping was obtained from representatives of the study households ., The protocols and consent forms were reviewed and approved by the Animal Care and Use and Ethical Review Boards of the Kenya Medical Research Institute ( #1191 ) ., The study protocols were also approved by the Institutional Animal Care and Use Committee and Institutional Review Board of the U . S . Centers for Disease Control and Prevention ( #5410 ) and complied with the Public Health Service Policy on Humane Care and Use of Laboratory Animals ., A total of 49 small mammals trapped at 29 compounds in Asembo and 220 rodents trapped at 143 households in Kibera were included in this study ., The small mammals trapped in Asembo included Crocidura olivieri ( n = 16 , African giant shrew ) , and rodents of the species Lemniscomys striatus ( n = 2 , striped grass mouse ) , Mastomys natalensis ( n = 14 , Natal multimammate mouse ) , Mus minutoides ( n = 1 , pygmy mouse ) , and Rattus rattus ( n = 16 , black rat ) ., All of the rodents trapped in Kibera were Mus musculus ( n = 178 , house mouse ) , Rattus norvegicus ( n = 10 , brown rat ) or Rattus rattus ( n = 32 ) ( Table 1 ) ., Ten of the 49 ( 21% ) animals trapped in Asembo were culture-positive for Bartonella , including: Crocidura olivieri ( n = 1 , 7% ) ; Lemniscomys striatus ( n = 1 , 50% ) ; Mastomys natalensis ( n = 6 , 43% ) ; and Rattus rattus ( n = 2 , 13% ) ., Overall , 24 of the 220 ( 11% ) animals trapped in Kibera were culture positive: including Rattus norvegicus ( n = 5 , 50% ) and R . rattus ( n = 19 , 60% ) ., None of the 178 samples collected from Mus musculus in Kibera were positive ( Table 2 ) ., Culture-positive Rattus species were trapped in four of the five trapping grids established at the Kibera site ( Fig . 2 ) ., Information on gltA sequences was obtained from all 34 culture-positive animals ., The phylogenetic relationships between the isolates obtained in this study and previously described Bartonella species are shown in Figs 3 and 4 ., Bartonella detected in one M . natalensis and two R . rattus trapped in Asembo belong to the B . elizabethae species complex based on the similarity of the gltA sequences ( Fig . 3 & Table 2 ) ., Five additional sequences detected in M . natalensis and one detected in L . striatus were most closely related to B . tribocorum ., None of the strains obtained from M . natalensis or L . striatus were identical ( ≥ 96% sequence identity ) to reference strains of the Bartonella species described previously in Rattus species trapped elsewhere ., The strain of Bartonella cultured from a C . olivieri was not very similar to any previously described Bartonella reference species but has 93 . 8% similarity B . birtlesii ., Three pathogenic Bartonella species ( B . elizabethae , B . tribocorum and B . queenslandensis ) were detected in the two Rattus species sampled at the Kibera site ., The gltA sequences for all Bartonella strains from Kibera rodents were identical ( ≥ 96% sequence identity ) to reference isolates that are typical of Bartonella detected in Rattus populations globally ( Table 2 & Fig . 4 ) ., At the Asembo location , there was a weak association between individual infection status and host species ( likelihood ratio test p = 0 . 053 ) where infection probability was higher in M . natalensis individuals than in the reference species C . olivieri ( OR = 11 . 25 , 95% CI = 1 . 15–110 . 47 , p = 0 . 038 ) ., Approximately half ( 24/49 ) of the small mammals trapped in Asembo were trapped outside ( Table 1 ) ., None of the Bartonella positive animals trapped in Asembo were trapped within occupied dwellings ., Two positive R . rattus were trapped in outbuildings but all other positive animals ( one C . olivieri , one Lemniscomys striatus and six M . natalensis ) were trapped outside ., There were no statistically significant associations between the probability of Bartonella infection and small mammal sex or mass within the Asembo population ., At the Kibera location , there was a clear influence of genus upon infection probability ., None of the 178 Mus trapped were Bartonella positive but 24/42 Rattus were positive indicating much higher infection probability in Rattus ( OR = Infinite ) ., Considering the data for Rattus individuals only , there were no statistically significant associations between the probability of Bartonella infection and rodent species , sex , or mass ., The proportion of infected Rattus and proportion of infected rodents overall varied by trapping zone in Kibera ( Fig . 2 ) ., There was no statistically significant difference in the probability of Bartonella infection in Rattus from different trapping zones ., However the sample size for this analysis was small and the existence of zones where no positive individuals were trapped complicate this analyses and its interpretation ., Descriptively , the trapping data from Kibera fall into two groups ., In zones A , B and D few Rattus were trapped ( Fig . 2 ) , the rodent populations in these zones were dominated by Mus musculus and only four Bartonella isolates were identified in the combined rodent populations from these three zones ( Fig . 2 ) ., In contrast , in trapping zones C and E , Rattus made up larger proportions of the total trapped population ( 51% in zone C and 40% in zone E ) and more Bartonella isolates of several species were identified in these populations ., This study reports isolation of Bartonella strains from rodent and shrew species in Asembo and Kibera , Kenya ., Bartonella strains were found in several small mammal species with variation observed in the infection prevalence and in the strains of Bartonella detected between host species and study sites ., The majority of Bartonella isolates obtained from these Kenyan mammals are genetically similar to reference strains of known human pathogens ., Several recent studies indicate that the prevalence of Bartonella infection in Rattus in Africa may be low in contrast to the frequently high prevalences observed in Asian Rattus populations 18–20 , 32 ., It has been argued that this pattern of lower prevalence in African Rattus populations could be attributed to host escape during colonization 19 , 20 , a phenomenon where relatively small founding populations of invading species can leave their parasites behind when colonizing new areas 33 ., Consistent with this , a relatively low prevalence was seen in R . rattus from Asembo ( 13% ) ., However , the high infection prevalence observed in Rattus trapped at the Kibera site ( e . g . 50% R . norvegicus and 60% R . rattus ) is more similar to prevalence values observed in studies of Asian Rattus populations than to other African populations 19 ., There are multiple possible explanations for the difference in the prevalence observed in Rattus at these two sites ., Phylogenetic analyses indicate that B . elizabethae complex strains originated in Southeast Asia and have been disseminated throughout Asia , Europe , Africa , Australia and the Americas through multiple dispersal events of commensal Rattus species 5 ., The Kibera study site is near the centre of Nairobi , the Kenyan capital , and is likely to have greater international connectivity ( in terms of international rodent movement through trade etc . ) than the Asembo site , which is more rural ., The higher prevalence observed at the Kibera site could therefore be explained by repeated introduction of Rattus and their associated Bartonella species to this site 19 , 33 ., Further analyses would however be needed to elucidate the colonization history of Rattus and their associated Bartonella at these sites specifically ., The number of species trapped in Kibera was smaller than the number trapped in Asembo , indicating a simpler species composition at this site and these data could also suggest a possible dilution effect of the increased community complexity in Asembo on Bartonella prevalence 34 ., Finally , temporal dynamics in host and ectoparasite population structure are known to affect Bartonella infection prevalence 21 , 35 ., This study involved cross-sectional trapping surveys conducted at different times of year in the two study locations ., There are few data on the seasonal variation in the abundance or diversity of the rodent populations at these sites but it is likely that there are seasonal influences upon rodent abundance and diversity with differences between the urban Kibera site and the more rural Asembo site in the seasonal population dynamics observed 36 ., It is therefore possible that differences in the sampling time may have contributed to the differences in infection prevalence seen between these two Rattus populations ., Notably , no bartonellae were detected in Mus musculus trapped in Kibera despite a large number of tested animals and high infection prevalence observed in Rattus trapped in the same locations ., Low-level Bartonella infection has been reported from Mus trapped in Ethiopia but the absence of Bartonella in Mus was also reported in a small-scale study from Nigeria 19 , 37 ., All of the Kibera rodents were trapped within residents’ homes ., In contrast , although nearly half of the animals trapped in Asembo were trapped indoors or in outbuildings , none of the positive animals at this site were trapped indoors , and only two positive animals were trapped in outbuildings ., Approximately one third of the animals trapped outside in Asembo however were culture positive for Bartonella ., The two culture positive animals trapped in outbuildings were R . rattus and they were carrying Bartonella similar to the B . elizabethae reference strain ( Fig . 2 ) ., This species was most commonly found indoors or in outbuildings ( 14/15 records ) and therefore may pose a risk due to closer human contact , even though only 2/16 were positive ., Many of the Bartonella detected at in this study ( except the birtlesii-like isolates from Crocidura ) belong to the Bartonella elizabethae complex and many of the strains identified in invasive Rattus hosts particularly are closely related to known human pathogens ., All of the Bartonella strains isolated from Kibera rodents have ≥ 96% sequence identity with strains that are common in Rattus species sampled in Asia and on several other continents 5 ., In contrast , several strains isolated from Asembo rodents and shrews were less similar to the international reference strains from Rattus and were more similar to isolates gathered previously from Ugandan rodents , suggesting a longer history of circulation of these strains within these species at the Asembo site ., The identification of similar B . tribocorum sequences in Mastomys and Lemniscomys individuals trapped in Asembo suggests an absence of strong host-pathogen associations in these populations ., There is a high incidence of acute febrile illness in people in both Asembo and Kibera 25 ., A variety of pathogens are known to account for a proportion of febrile illness in Asembo and Kibera but considerable proportions remain unexplained 38–41 ., Bartonella species have been identified as important causes of human febrile illness in several global settings but there has been little investigation of the impact of bartonellosis upon human health in Africa particularly and it is conceivable that Bartonella may be an important cause of febrile illness in these study populations ., The data presented from the Asembo and Kibera sites indicate clear differences in: the prevalence of Bartonella infection in the same host ( Rattus species ) at the two sites; the prevalence of infection in different hosts trapped at the same sites; the abundance of different infected hosts between the two locations and also between trapping zones in Kibera; the strains of Bartonella detected and finally in the locations within communities where rodents overall and Bartonella infected rodents were trapped ., The impact of this variation in rodent host community composition , infection prevalence , ectoparasite vector preferences , and other ecological factors need to be understood to evaluate human Bartonella infection risks at these sites ., The data presented here suggest that investigations of the multi-host infection dynamics of Bartonella and the public health significance of Bartonella infections at these Kenyan locations and others where there are close associations between people and small mammals are warranted . | Introduction, Materials and Methods, Results, Discussion | Several rodent-associated Bartonella species are human pathogens but little is known about their epidemiology ., We trapped rodents and shrews around human habitations at two sites in Kenya ( rural Asembo and urban Kibera ) to determine the prevalence of Bartonella infection ., Bartonella were detected by culture in five of seven host species ., In Kibera , 60% of Rattus rattus were positive , as compared to 13% in Asembo ., Bartonella were also detected in C . olivieri ( 7% ) , Lemniscomys striatus ( 50% ) , Mastomys natalensis ( 43% ) and R . norvegicus ( 50% ) ., Partial sequencing of the citrate synthase ( gltA ) gene of isolates showed that Kibera strains were similar to reference isolates from Rattus trapped in Asia , America , and Europe , but that most strains from Asembo were less similar ., Host species and trapping location were associated with differences in infection status but there was no evidence of associations between host age or sex and infection status ., Acute febrile illness occurs at high incidence in both Asembo and Kibera but the etiology of many of these illnesses is unknown ., Bartonella similar to known human pathogens were detected in small mammals at both sites and investigation of the ecological determinants of host infection status and of the public health significance of Bartonella infections at these locations is warranted . | Bartonella are bacteria that infect many different mammal species and can cause illness in people ., Several Bartonella species carried by rodents cause disease in humans but little is known about their distribution or the importance of bartonellosis as a cause of human illness ., Data from Africa are particularly scarce ., This study involved trapping of rodents and other small mammals at two sites in Kenya: Asembo , a rural area in Western Kenya , and Kibera , an informal urban settlement in Nairobi ., Blood samples were collected from trapped animals to detect and characterize the types of Bartonella carried ., At the Kibera site over half of the trapped rats were infected with Bartonella very similar to human pathogenic strains isolated from rats from other global regions ., In Asembo , Bartonella were detected in four of the five animal species trapped and these Bartonella were less similar to previously identified isolates ., All of the small mammals included in this study were trapped in or around human habitations ., The data from this study show that Bartonella that can cause human illness are carried by the small mammals at these two sites and indicate that the public health impacts of human bartonellosis should be investigated . | null | null |
journal.pntd.0005265 | 2,017 | Predicting Ebola Severity: A Clinical Prioritization Score for Ebola Virus Disease | Ebola virus disease ( EVD ) caused by the virulent Zaire ebolavirus strain is described by the WHO as one of the world’s most deadly infections , with case fatality rates exceeding 80% in past epidemics 1 , 2 ., Supportive care in the 2013–2015 outbreak in West Africa was shown to reduce the EVD mortality rate to around 50% 3 , and overall , the WHO has reported 40% fatalities among the 28 , 603 people affected by EVD 4 ., Despite its notoriety as a fatal disease , over 80% of patients survived when treated in resource-rich environments of Europe and the USA 5 ., Further , asymptomatic infections are not only possible but could constitute up to a third of all transmissions 6–9 ., Overlooking these infected ( but minimally contagious ) individuals was proposed to result in the overestimation of EVD epidemic modelling , and revealed the heterogeneous range of EVD symptomology 10 ., Improved prognostic tools that objectively stratify mortality risk among EVD patients could better allocate limited resources by identifying those most in need of intensive treatment and to aid clinical decision-making ., Further , the clinical trials undertaken in the Ebola response have been criticised for the potential bias introduced via the lack of randomisation and contemporaneous controls 11 , 12; thus , a method of objectively controlling differences in mortality risk among participants may aid analysis ., Existing EVD staging models used in Sierra Leone , were based on a WHO protocol adapted from the clinical presentation of Lassa fever 13 , where 3 symptomatic stages were described:, 1 ) Early/non-specific ,, 2 ) gastrointestinal and, 3 ) late/complicated , featuring haemorrhage and organ failure ., While it has since been shown that these three stages of the disease are broadly correlated to EVD outcome 14 , the system could be greatly improved by using statistically weighted symptoms that better stratify the risk of mortality ., Several studies have already identified single symptoms statistically predictive for EVD mortality , such as confusion 15–17 , diarrhoea , 16 , 18 asthenia 15 , 18 , hiccups 14 , haemorrhagic signs 14 , 16 , 19 , dizziness 18 , extreme fatigue 15 , and high viral load 14 , 17 , 18 , 20 ., However , the various permutations in which symptoms occur in each individual , necessitates a multivariate approach to more accurately predict mortality ., The Ebola virus has been hypothesized to exercise its diverse range of virulence through the mammalian immune system ., Here , it causes a pathologic overstimulation of innate immune receptors , triggering a flood of inflammation that causes collateral damage to multiple organ systems 21 , 22 , and results in a wide range of symptomatic presentations 8 , 23 ., It is then easy to imagine the additive detrimental effect of an inflammatory co-infection such as the malaria parasite , Plasmodium falciparum ., The annual incidence of malaria in Sierra Leone is 350 cases per 1000 population and it has been reported to be more prevalent in EVD triaged patients than EVD itself 24 ., However , despite these statistics , little is known about EVD/malaria co-infection or its effect on patient prognosis ., In this retrospective cohort study , we analyse the clinical and epidemiological data from 158 EVD ( + ) patients admitted to the GOAL-Mathaska ETC in Port Loko , Sierra Leone ., We investigate the role of malaria in EVD pathogenesis and evaluate the potential of the clinical characteristics in predicting EVD mortality at triage as well as in on each day of patient care ., Further , we use these results to construct a statistically weighted disease scoring and staging system , which identifies the most prevalent factors that are predictive of mortality ., This retrospective cohort study uses anonymized patient data collected between December 14 , 2014 and November 15 , 2015 at the GOAL ETC in Port Loko , Sierra Leone ., Data comprised patient demographics , geographic location , clinical signs and symptoms , and laboratory results ( for malaria infection and semi-quantitative Ebola viremia ) , as well as the final patient outcome of death or survival ., We evaluate the potential of clinical characteristics in predicting EVD mortality and use these results to construct a symptom-based disease staging system , which corresponds to the prognostic power of the most prevalent symptoms ., The ETC was run by the humanitarian organization GOAL Global in cooperation with the Sierra Leonean Ministry of Health and Sanitation ( MoHS ) ., The ETC opened in December 2014 and accepted 600 patients from a catchment area spanning 200km ., EVD surveillance in Sierra Leone was implemented through District Ebola Response Centres ( DERCs ) ., Individuals who were sick were encouraged to report their illness ( or the suspected illness of others ) via the national or district Ebola call-lines ., Individuals that met the WHO case definition for EVD 25 , as well as those with confirmed EVD infection , were referred to the ETC from surrounding communities , holding centres , health facilities , and quarantine houses ., All EVD ( + ) patients were treated according to standard treatment protocols developed by WHO and Médecins Sans Frontières 13 , 26 ., This included empiric antimalarial treatment ( Artesunate and Amodiaquine ) , broad-spectrum antibiotics , and nutritional supplementation for all patients , as well as oral or intravenous fluid rehydration ., Signs and symptoms were recorded daily , on admission and throughout the patients’ length of stay at the ETC ., Once triaged , blood was drawn and tested for EVD in on-site laboratories managed by Public Heath England ., EVD diagnosis was determined by semi-quantitative reverse transcriptase-PCR ( qRT-PCR ) as previously described 27 ., Briefly , the cycle threshold ( Ct ) value was used as an inverse proxy for viral load and a cut-off of 40 was used to discriminate between positive and negative values ., Patients qualified as EVD ( - ) and were discharged from the ETC after returning two negative Ebola-specific qRT-PCR tests ., Histidine-rich protein-II ( HRP-II ) antigen rapid diagnostic kits were used for testing of malaria infections , which were performed on admission at the ETC ., Symptoms were reported by the patient during a comprehensive questionnaire by trained staff ., Haemorrhaging , pyrexia , and disorientation were recorded by clinicians after examination ., Haemorrhagic signs included visible blood loss such as hematochezia , hematemesis , haematuria , epistaxis , haemoptysis or persistent haemorrhage from an IV catheter site as well as subcutaneous haemorrhage such as purpura and petechiae ., Pyrexia was defined as a body temperature over 38°C , measured using an infrared thermal sensor ., Disorientation was measured by trained ETC clinicians as per the AVPU alertness scale ( where pain and unconsciousness were considered “disorientated” ) ., Additionally , any specific mention of “confusion” or “disorientation” in the medical notes was also considered a positive for this variable ., Of the 600 patients admitted to the ETC , 10 were declared dead on arrival and 24 were classified as late transfers from other ETCs or holding centres ( treated elsewhere and thus convalescent on arrival ) or had incomplete data ., Thus , a total of 34 patients were excluded from this analysis ., Of the 566 patients involved in the study , 100% had diagnostic test results for EVD , where 27 . 9% tested EVD ( + ) ( n = 158 ) ., 543/566 patients had malaria test results , of which 34 . 6% were malaria ( + ) ( n = 188 ) ., The cohort was evaluated for missing values in each variable ., Referral time ( the time in days from symptom onset to admission at the ETC ) had 20 cases of missing data ., Further analysis was undertaken to evaluate the aetiology of missingness , which included demographic variables ( such as age and sex ) , clinical severity variables ( such as EVD viral load ) as well as the covariates used in the final scoring model ., Here , we found that subjects with missing data did not differ systematically from those with observed referral time , which is in favour of the hypothesis that the data were missing completely at random ., In addition , we performed a sensitivity analysis using the “Hotdeck” imputation technique , which showed that the model coefficients did not change when using complete data 28 ., The patient catchment area and mortality rates can be visualised in S1 Fig . Ethical approval for this research was granted by the Sierra Leone Ethics and Scientific Review Committee ( SLESRC ) ., To maximize data fidelity , patient files were entered into a secure Microsoft Excel database and crosschecked by 3 independent and trained analysts ., Entry of clinical data was overseen by members of the clinical ETC staff ., Graphs were constructed using Graphpad Prism , version 6 ., Univariate and multivariate analysis was conducted using STATA software , version 14 ( StataCorp ) ., Score validation was performed using “RMS” R-Package ( R Development Core Team . ISBN 3-900051-07-0 , URL: http://www . R-project . org ) ., Results were deemed statistically significant at a p-value of less than 0 . 05 ., Epidemiological data and symptoms were summarized by their frequencies and percentages ., Univariate logistic regression was performed to assess the association between each predictor and the outcome of death ( reported as Odds-Ratios ( OR ) and p-values ) ., Potential interactions were tested ( such as the effect of sex , age , referral time and Ebola contact ) ., The functional relation between the outcome of death and continuous variables ( age , days admitted , referral time and EVD viral load ) were checked using a fractional polynomial model ., The linearity assumption was confirmed for days admitted and referral time but not for age ( S2 Fig ) or Ct Value ( S3 Fig ) ., To simplify the prognostic score , age was coded into three categories: ( 1 ) <5 years + 25–45 years , ( 2 ) 5–25 years , and ( 3 ) >45 years ., S2 Fig shows the rationale for the chosen categories on their polynomial curve ., The 5–25 group is used as a reference , being the lowest risk group ., Comparing the <5 years category to the reference , we obtained an OR of 5 . 35 ( p = 0 . 006 ) , while the 25–45 category returned an OR of 2 . 61 ( p = 0 . 002 ) ., Comparisons between the 0–5 and 25–45 groups , however , showed that they were not significantly different ( p = 0 . 2 ) and they could thus be grouped to simplify the user interface of the score ., The area under the ROC curves for the scoring systems presented in this study were not statistically different when comparing these age categorisations with the polynomial function of age as a continuous variable ., However , a continuous function would undoubtedly be more accurate on a larger sample size ., For PCR results , a Ct value lower than 20 cycles was categorized as “high viral load” and correlated to the natural threshold for the probability of fatal outcome ( S3 Fig , dotted line ) ., As there was an insufficient number of patients in the survival group ( death = 96 , survival = 62 ) compared to the number of 31 potential predictors , only the predictors associated to the outcome at level of p<0 . 20 were considered into a Stepwise Backward selection procedure to fit a multivariable logistic regression model 29 ., The “daily” score for calculating risk after triage would ideally be handled with a time-dependent model in order to limit immortal time bias ., However , these were not a good fit for our data , as the proportionality-hazards assumption was violated by non-parallel lines between categorical variables on log-log plots ., Thus , a logistic multivariate model was privileged ., Model diagnostics were performed to check for influential observations that impact coefficient estimates and a Hosmer-Lemeshow goodness-of-fit test was performed to assess calibration ., Discriminative performance of the final model was assessed by calculating the Area Under the Receiver Operating Characteristics ( ROC ) Curve ( AUC ) and its 95% confidence interval ., The β-coefficient = log ( OR ) of each covariate of the final model was converted into an integer-based point-scoring system ., The score was then derived as the sum of the covariates’ weighted scores ., Internal validation using the bootstrap method ( repeated 1000 times ) as described in Harrell et al 30 was used to provide a more accurate estimate of the performance of the original model ( model based score: AUCoriginal ) ., The algorithm calculates the optimism of the predictive discrimination in the original model ., The difference ( AUCoriginal−optimism ) gives the bootstrap-corrected ( i . e . internally validated ) performance of the original model ., As described in Steyerberg et al 31 , bootstrapping has unavoidable limitations in small cohorts with a large number of predictors ., The funder had no role in study design , data collection , data analysis , data interpretation , or writing of the report ., The corresponding author had full access to the data in the study and had final responsibility for the decision to submit for publication ., Of the 566 patients included in this study , 27 . 9% tested positive for EVD ( n = 158 ) ., The crude mortality rate among EVD ( + ) patients was 60 . 8% ( Fig 1A ) ., Mortality rates were slighter higher in males ( 68 . 4% vs . 53 . 7% ) , with a statistically insignificant 1 . 9 fold increase in odds of death ( p = 0 . 06 ) ( Fig 1B ) ., EVD survivors were on average 10 years younger than those who died ( 24 . 9 years vs . 34 . 3 years , p = 0 . 014 ) ( Fig 1C ) ., However , mean age of death did not differ among EVD ( - ) patients who died before being transferred out of the ETC ( Fig 1C ) or among genders ( p>0 . 05 ) ., In general , case fatality rates for EVD were higher at the youngest and oldest extremes of age ., The patient group aged between 5 and 24 years had the lowest mortality rate of 42 . 5% , which was significantly lower than other age groups ., The over-45’s and under-5’s were particularly vulnerable , being 11 . 6 and 5 . 4 fold more likely to die , respectively ( Fig 1D ) ., Age groups were selected in order to ensure the mathematic simplicity of the final score ., The polynomial curve of this continuous variable is shown in S2 Fig . Categorisation did not significantly alter the accuracy of the final scores ( p>0 . 05 ) ., In an effort to better predict the risk of EVD death , we analysed the prevalence and prognostic potential of the major clinical characteristics among EVD ( + ) patients ., Triage symptoms reported by over 50% of fatal EVD ( + ) patients were asthenia , myalgia , diarrhoea , anorexia , vomiting , pyrexia , and headache ( Fig 2A and Table 1 ) ., The prevalence of several of triage symptoms was notably different between fatal and non-fatal outcomes , as can be seen by comparing their ranking ( Fig 2A ) or their differential prevalence ( Fig 2B ) ., As expected , high viral load ( Ct value <20 ) was approximately 50% more common among fatal outcomes and univariate analysis revealed it as a major correlate of mortality with 11 . 8 fold odds of death ( p>0 . 0001 ) ( Table 1 ) ., While disorientation on admission was not common in EVD ( + ) patients ( 11 . 4% ) , when present , it was associated with 94 . 4% of fatalities in EVD ( + ) patients , and was therefore the strongest indicator of fatal outcome ( OR 13 . 1 , p = 0 . 014 ) ( Table 1 ) ., Other factors showing a statistically significant association with death were diarrhoea , hiccups , myalgia , dyspnoea and conjunctivitis ( all p<0 . 05 ) ( Table 1 ) ., While haemorrhagic signs were infrequent on admission ( 14 . 6% ) , developing haemorrhage at any point during admission at the ETC was associated with a 6-fold higher odds of mortality ( p>0 . 0001 ) ( Table 1 ) ., Finally , malaria infection was more prevalent in fatal outcomes ( Fig 2B ) and will be discussed further below ., EVD diagnoses were routinely confirmed by qRT-PCR , where the cycle threshold ( Ct ) value is inversely proportional to the Ebola virus copy number ., We used similar parameters presented in other studies to delineate high and low Ct values 14 , 32 , where a Ct value lower than a threshold of 20 cycles was categorized as “high viral load” ., This also correlated to the natural threshold for the probability of fatal outcome in our cohort ( S3 Fig ) ., Ct values were available for 144/158 of the EVD ( + ) cohort ( 91 . 1% ) and ranged from 13 . 5 to 37 . 9 ., Of these , 39% ( n = 57 ) were classified as having with high viral loads ( Fig 3A ) ., The mean Ct value ( 22 . 0 ) did not vary by gender ( p>0 . 05 ) but was differentially distributed across ages , where each 10 years corresponded to a decrease of 0 . 4 Ct points ( i . e . an increase of viral load ) ( p = 0 . 035 ) ( Fig 3B ) ., Finally , the average Ct value for survivors was significantly higher than those who had a fatal outcome ( 24 . 9 vs . 20 . 6 , p<0 . 01 ) ( Fig 3C and 3D ) , where odds of death were 12 . 6 times higher for patients with Ct values of less than or equal to 20 ( p<0 . 0001 ) ., An EVD staging model developed by the UK Defense Medical Services in reference to WHO guidelines on the pathogenesis of haemorrhagic fever , divides the temporal evolution of EVD into three symptomatic stages 13 , 14 ., Here , the “early” stage is comprised of non-specific symptoms lasting three days ., Considering this timeline , we investigated the impact of early referral on mortality ., Crude analysis showed that a fatal outcome was not associated to a later referral time ., Oppositely , EVD survivors presented at the ETC one day later than those who died ( 4 . 6 vs . 3 . 6 days ) albeit a statistically insignificant trend ( p = 0 . 12 ) ., Indeed , those presenting within 3 days of symptom onset had a 15% higher mortality than those presenting later ( p = 0 . 09 ) ( Fig 4A ) ., This counter-intuitive trend of earlier healthcare seeking behaviour resulting in higher death risk could be theoretically explained as the confounding effect of disease severity ., Here , we propose that patients presenting earlier are doing so as they have a more severe acute disease and thus represent a population predisposed to mortality risk ., To investigate this hypothesis , we compared viral loads as a proxy for disease severity and found that those presenting earlier had similar viral loads to those presenting later ( Fig 4B ) ., This result indicates that early presenters have more severe acute disease ., Correcting for viral load as a confounding factor ( where comparisons are only made between patients with equal viral loads ) , we found the more intuitive result that delayed treatment was significantly associated with mortality ., Here , the probability of death increased by an average of 12% for each day of delayed treatment during the 1st week of symptoms ( p = 0 . 012 ) ( Fig 4C ) ( Table 1 ) ., This quantifies the benefits of early health care seeking behaviour ., Oppositely , each day spent within the ETC increased the odds of survival by 1 . 4 fold ( p<0 . 0001 ) irrespective of viral load ( Fig 4D ) ( Table 1 ) ., The average admission duration for EVD ( + ) survivors was 10 . 4 days while deaths occurred , on average , within the first 4 . 2 days of admission ( Fig 4D ) ., Finally , quarantine status upon admission was available for 96% of ETC admissions ( n = 551 ) , 25% of which were referred from quarantine houses ., EVD ( + ) patients referred from quarantined homes had an earlier referral time than those not referred from quarantine facilities ( 3 . 8 days vs . 4 . 7 days , p = 0 . 03 ) ( Fig 4E ) ., This quantifies the potential patient benefit of quarantine in the region , however , we found no difference in mortality by quarantine status ( OR = 1 . 8 , p = 0 . 08 ) ., Of the 543 EVD ( + ) and EVD ( - ) patients with a known malaria test , 34 . 8% tested malaria ( + ) ., Among EVD ( + ) patients , 24% were co-infected with malaria compared to 38% EVD ( - ) ( OR = 2 , p = 0 . 005 ) ., The prevalence for malaria infection varied drastically across age categories , where 5 year olds had an over 50% probability of being malaria ( + ) in both EVD ( - ) and EVD ( + ) cohorts ( Fig 5A ) ., Despite the WHO standard of care to treat all ETC admissions with Arteminisin Combination Therapy ( ACT ) upon admission ( irrespective of malaria status ) 33 , EVD ( + ) /malaria ( + ) co-infected patients suffered a significantly higher mortality rate compared to EVD alone ( 74 . 3% vs . 53 . 6% , OR = 3 . 9 , p = 0 . 03 after controlling for age and gender ) ( Fig 5B ) ., Ebola viral load was differentially distributed across EVD ( + ) patients who were co-infected or not with malaria ., Co-infected EVD ( + ) /malaria ( + ) patients had significantly higher viral loads compared to patients infected with EVD alone ( mean Ct = 20 . 8 vs . 22 . 3 , p<0 . 01 ) ( Fig 5C ) ., Controlling for viral load , the increased mortality in malaria co-infected EVD ( + ) patients was abrogated ( p = 0 . 107 ) ( Fig 5D ) ., Taken together , these results reveal a potential pathogenic synergy between the malaria parasite and Ebola virus ., Performing multivariate analysis of the above data , we selected the clinical characteristics most predictive for EVD mortality using data collected at triage or at any time during the patients’ stay in the ETC ( Table 1 ) ., By stepwise backwards elimination , and prioritizing the most prevalent symptoms , we identified several characteristics which yielded significant predictive values at triage and during admission ., Characteristics that were statistically significant predictors of mortality at admission were vulnerable age groups ( <5 , 25–45 and >45 years ) , myalgia , disorientation and referral-time ( normalised to viral load ) ( Table 1 ) ., Characteristics that were statistically significant predictors of mortality after admission were vulnerable age groups ( <5 and >25 years ) , disorientation and haemorrhage ., Oppositely , days spent in the ETC was a significant predictor of survival ( OR 1 . 5-fold for each day , p<0 . 0001 ) ( Table 1 ) ., Despite the strong association of malaria co-infection with fatality in our univariate analysis above , malaria infection was rendered insignificant in our multivariate analysis ., We then calculated weightings for both scores from the predictive coefficients with the aim to find a simplified scoring model using whole integers and calculations limited to subtraction or addition ( Table 1 ) ., Testing the sensitivity and specificity of these weightings for the prediction of EVD infection , we found that the characteristics yielded an area under the ROC curve ( AUC ) of 91 . 4% ( CI95%: 87–96% ) for discriminating mortality at triage ( Fig 6A ) and 97 . 5% ( CI95%: 95–99% ) for calculations after admission ( Fig 6B ) ., The risk category cut-offs for each score are illustrated in Fig 6C and 6D ., Each category contains at least 10% of the cohort ., The 3 risk cut-offs ( Low , Medium and High ) were selected based on the linear risk curve ( Fig 7 ) , where “Low” and “High” categories represent risk plateaus on the extremes of the risk statistic ( Low <7% , and High >98% ) ., Examining the accuracy of the triage mortality score , we found that the “high” risk classification was composed of 99% correctly classified fatal outcomes while the “low” risk category had a less than 10% mortality rate ( Fig 6E ) ., After triage , the “high” risk category of the daily score was composed of 98% fatalities compared to less than 1% in the “low” risk category ( Fig 6F ) ., An internal validation of the triage and daily scores yielded a final discriminative power of 89 . 12% and 97 . 04% respectively ( Table 2 ) ., As described in Steyerberg et al 31 , bootstrapping has unavoidable limitations in small cohorts with a large number of predictors and thus the optimism may be over-estimated ., External validation is needed to best test these associations ., All prognostic tools carry the risk of becoming self-fulfilling prophecies if incorrectly used as an indicator for palliation: dooming severely ill patients to death when the score is not reflective of clinical advances ., This score is specifically adapted to an Ebola response in resource-constrained settings , where clinical resources achieved a 40% survival rate ., As an 80% survival rate was possible among patients in resource rich environments 5 , it is clear that the interpretation of the score would need to evolve with anticipated clinical advancements ., However , the major asset of this score is not limited to prediction of the binary outcome of death , but rather its use as a proxy for “disease severity” in resource limited environments ., Thus , while the outcome of “death” may change with improved treatment options , patients scoring highly on this tool can still be shortlisted for intensive intervention ., Additionally , with the exciting potential of machine-learning predictive tools 43 , scoring systems such as these can become more durable and evolve with their developing environments , where a future of accurate EVD diagnosis and prognosis is a realistic possibility ., As we found for malaria , EVD is certainly not the only contributing factor to mortality within an ETC , and patients who have lived a lifetime within a poorly resourced health care system very probably have diverse and complex competing risks ., This is an unavoidable bias , as accurate secondary diagnostics for co-morbidities were primitive at best for the bulk of the patients ., We await retrospective analyses on patient samples that may reveal the presence of other co-infections or confounding genetic/immunologic anomalies ., Finally , external validation is an essential step before the endorsement of any clinical tool ., Without external validation , the level of inaccuracy within this cohort cannot be estimated and thus these scoring systems must be used with this caution in mind ., This study identifies several epidemiological and clinical features , which are significantly predictive for the outcome of EVD infection and proposes several highly accurate statistical tools to predict the clinical severity of EVD and aid objective clinical prioritization ., External validation and systematic meta-analyses of the clinical features of EVD are needed to fine-tune the statistical weightings of this score to further improve its accuracy and geographical relevance . | Introduction, Methods, Results, Discussion | Despite the notoriety of Ebola virus disease ( EVD ) as one of the world’s most deadly infections , EVD has a wide range of outcomes , where asymptomatic infection may be almost as common as fatality ., With increasingly sensitive EVD diagnosis , there is a need for more accurate prognostic tools that objectively stratify clinical severity to better allocate limited resources and identify those most in need of intensive treatment ., This retrospective cohort study analyses the clinical characteristics of 158 EVD ( + ) patients at the GOAL-Mathaska Ebola Treatment Centre , Sierra Leone ., The prognostic potential of each characteristic was assessed and incorporated into a statistically weighted disease score ., The mortality rate among EVD ( + ) patients was 60 . 8% and highest in those aged <5 or >25 years ( p<0 . 05 ) ., Death was significantly associated with malaria co-infection ( OR = 2 . 5 , p = 0 . 01 ) ., However , this observation was abrogated after adjustment to Ebola viral load ( p = 0 . 1 ) , potentially indicating a pathologic synergy between the infections ., Similarly , referral-time interacted with viral load , and adjustment revealed referral-time as a significant determinant of mortality , thus quantifying the benefits of early reporting as a 12% mortality risk reduction per day ( p = 0 . 012 ) ., Disorientation was the strongest unadjusted predictor of death ( OR = 13 . 1 , p = 0 . 014 ) followed by hiccups , diarrhoea , conjunctivitis , dyspnoea and myalgia ., Including these characteristics in multivariate prognostic scores , we obtained a 91% and 97% ability to discriminate death at or after triage respectively ( area under ROC curve ) ., This study proposes highly predictive and easy-to-use prognostic tools , which stratify the risk of EVD mortality at or after EVD triage . | The unprecedented spread of EVD across the fragile healthcare systems of West Africa during the 2013–2015 outbreak infected over 28 , 600 patients and established it as a disease for which low-income countries are at disproportionate risk ., In order to improve the standard of patient care , it is essential to better allocate scarce resources amongst the heterogeneous symptomatic presentations of EVD ., This retrospective cohort study on 158 EVD ( + ) patients in Sierra Leone constructs 2 easy-to-use scoring systems that accurately stratify EVD severity and thus objectively identify the patients in most need of intensive therapy ., Using statistically weighted symptoms and demographic characteristics , we obtained scores with a 91% and 97% ability to discriminate death at or after triage respectively ., These scores included Ebola viral load , patient age and referral time as well as the symptoms of disorientation , haemorrhage and myalgia ., Further univariate analysis revealed several independent predictors of mortality , where patients aged between 5 and 25 years were most likely to survive , while malaria co-infection increased the risk of death by 2 . 5-fold ( p = 0 . 01 ) ., Correcting referral-time for viral load , we also quantify the benefits of early reporting as a 12% mortality risk reduction per day ( p = 0 . 012 ) ., Mortality in this cohort was 3-fold more than patients treated in resource-rich settings ( 60 . 8% vs . 18% ) and we propose that focused patient care is a feasible and low-cost effort that may begin to close this gap . | death rates, medicine and health sciences, pathology and laboratory medicine, viral transmission and infection, demography, pathogens, tropical diseases, microbiology, parasitic diseases, parasitic protozoans, viruses, filoviruses, protozoans, rna viruses, signs and symptoms, forecasting, mathematics, statistics (mathematics), viral load, research and analysis methods, infectious diseases, malarial parasites, medical microbiology, mathematical and statistical techniques, microbial pathogens, people and places, diagnostic medicine, ebola virus, virology, viral pathogens, hemorrhage, co-infections, biology and life sciences, malaria, physical sciences, vascular medicine, statistical methods, hemorrhagic fever viruses, organisms | null |
journal.pcbi.1002456 | 2,012 | Warm Body Temperature Facilitates Energy Efficient Cortical Action Potentials | Brain signaling is metabolically expensive ., Energy expenditure not only constrains the size and architecture of the brain , which limits its computational power , but is critical to the interpretation of functional brain imaging signals through related metabolic mechanisms ( e . g . oxygen consumption and blood flow ) 1 ., Comprising only about 2% of the bodys mass , the mammalian brain consumes about 20% of its energy 2 , 3 , 4 ., Another unique feature of mammals is their warm body temperature ( about 35–39°C ) ., How a warm body temperature affects signaling and energy budget in the brain is largely unknown ., Here we address this critical and interesting question through simple Hodgkin-Huxley models as well as recordings from cortical neurons during changes in temperature ., Operating neurons is expensive , in part owing to the need to maintain a significantly higher concentration of Na+ ions outside , versus inside , nerve cells 5 , 6 , 7 , 8 ., Na+ entry into neurons , which must be returned to its extracellular location through the operation of the Na+/K+ ion pump by the expenditure of energy via hydrolysis of ATP , occurs through generation of action potentials ( particularly along long intracortical , unmyelinated axons ) and synaptic potentials during active signaling ., These influxes of Na+ into neurons occur in addition to a background leak of Na+ ions through the neuronal membrane ., Thus , to understand the energy costs of neuronal signaling in the cortex , it is essential to understand the entry of Na+ into neurons and neuronal processes during neuronal activity ., A maximally energy efficient action potential would entail no overlap of the inward Na+ current that generates the upstroke and the outward K+ current that facilitates the downstroke ., Any overlap of these two opposing currents would merely result in an electrically neutral exchange of positive ions ., Classic investigations by Hodgkin of squid giant axon revealed an excess entry of approximately 4 times as much Na+ as minimally required to generate the action potential 9 ., This value of 4 times excess Na+ entry has figured prominently in estimates of the distribution of the sources of energy consumption in the mammalian brain 1 , 5 , 10 , 11 , as well as in the calculation of the average firing rate of cortical neurons 6 ., For example , one classic modeling study of energy consumption in mammalian brains stated “A realistic estimate of the Na+ entry needed is obtained for action potential generation by quadrupling the minimal Na+ entry to take account of simultaneous activation of Na+ and K+ channels” 5 ., Calculations such as these have predicted that up to 50% or more of the energy consumption in mammalian brains is devoted to the reversal of ion exchanges owing to Na+ entry ( and K+ exit ) during action potentials 5 , 10 ., Based upon this calculation , it has been proposed that the brain can support an average firing rate of less than 0 . 2 spikes/second , suggesting that the nervous system operates through a very sparse code 6 ., Reconstruction of the inward Na+ and outward K+ currents occurring during action potential generation in mammalian cortical axons revealed , in contrast to the results predicted , an excess ratio of Na+ entry of only 1 . 3 12 , 13 , indicating that axons in the mammalian brain are far more energy efficient than previously appreciated ., This high energy efficiency in action potential generation is achieved through a relatively complete Na+ channel inactivation prior to substantial activation of the outward K+ current 12 , 13 ., These results prompted studies based on the hypothesis that the kinetics of ion channels may be optimized through evolution from invertebrates ( e . g . squid giant axon ) to mammals ( e . g . rodent cortical axons ) 12 , 13 ., This increased efficiency of mammalian axons has important implications not only for the average firing rate of cortical neurons , but also the practical limitations on the size and morphology of the brain ., Here , we reexamine this issue through a set of computational models and experimental study ., We hypothesize that the achievement of the highly energy efficient action potentials in cortical neurons is achieved in large part through the development of a warm body temperature ., Prior recording and modeling studies have demonstrated that the efficiency of action potential generation is highly sensitive to the kinetics of underlying ionic currents 13 , 14 , 15 ., Indeed , changing temperature has a strong influence on the kinetics of the Na+ and K+ currents underlying action potential generation 16 , 17 , 18 ., Homeothermic animals ( e . g . mammals and birds ) have the ability to maintain a relatively constant brain temperature in the range of 34–42°C 19 , 20 , 21 , 22 , 23 , 24 , while poikilothermic animals experience much wider variations in body temperature ., The species of squid most studied ( Loligo ) lives in an ocean environment varying in temperature from approximately 10–23°C , which is far colder than naturally occurring in most mammals ., Since temperature is a strong determinant of ion channel kinetics , which in turn can dramatically change action potential efficiency , we explored here the possibility that mammalian neurons may generate action potentials with maximal energy efficiency at normal body temperatures ., The examination of spike efficiency through simulation studies of squid giant axon action potential generation by Hodgkin 9 were performed with a temperature of 18°C , while the recordings of mammalian axons were performed at 37°C 12 ., We explored whether or not these variations in temperature may help to explain the marked difference in excess Na+ entry between these two species by performing simple Hodgkin-Huxley style simulations of action potentials in either the traditional HH single compartment model or in a simple uniform cylindrical model of a cortical axon 25 and varying temperature ( Figure 1A ) ., We used simple , single compartment models with only INa , IK , and ILeak so as to clearly demonstrate the principles of the effects of temperature on excess Na+ entry during action potentials ., Simulations with more complete computational models of layer 5 pyramidal neurons , including back propagating action potentials from the axon 25 , yielded similar results as those reported here ., For the present simulations , we assumed a Q10 for both Na+ and K+ currents of 2 . 3 17 , 18 , 26 , 27 and varied reversal potential with temperature according to the Nernst equation ., Varying the Q10 used from 1 . 5 to 3 yielded qualitatively similar results as those shown in Figure 1 ., Using either the traditional HH or cortical axon single compartment models , increases in temperature caused a marked decrease in the excess Na+ entry occurring during action potential generation , such that at 18°C a value of approximately 4X excess is obtained , while at 37°C , a value of 1 . 41 is observed for our model of cortical axons ( Figure 1A ) ., The traditional HH model of the squid axon fails to generate action potentials at temperatures above approximately 28°C ., At this temperature , the excess Na+ ratio is 2 . 5 ( Figure 1A ) ., In our model of a cortical axon spike , examination of action potentials at 18°C reveal an inward Na+ current that exhibits a prominent inward shoulder during the repolarizing phase of the spike , resulting in a strong overlap of inward Na+ and outward K+ currents ( Figure 1B ) , as in squid giant axons at this temperature ( Figure 1C; 9 ) ., The same simulation , but at 37°C , however reveals an inward Na+ current that overlaps very little with the outward K+ current , as seen in mammalian axons at 37°C ( Figure 1B; 12 , 13 ) ., For the squid axon HH model , the overlap of inward Na+ and outward K+ also decreases when temperature increases ( Figure 1C ) ., In addition , we also notice a dramatic decrease in spike duration as a function of temperature for both models ( see Figure 1D ) ., There is a nearly linear correlation between spike duration and excess Na+ entry ratio ( see Figure 1D , inset ) ., This relationship in our models arises from the fact that lowering temperature results in kinetically slower ionic currents , resulting in both an increase in overlap of the inward Na+ and outward K+ currents and a longer duration action potential ( see Figure 1B , D ) ., Examining the relationship between action potential duration and excess Na+ entry ratio ( inset in Figure 1D ) may lead one to hypothesize that it is the shortening of the duration of the action potential that is the primary effect in the reduction of Na+ entry with each spike at higher temperatures ( e . g . Figure 1A , B ) ., To test this hypothesis , we fixed the action potential waveform to either that occurring in the cortical axon model at 18°C , or to that occurring in the model at 37°C ( Figure 2 ) ., We then “injected” this waveform into the model and examined the amplitude –time course of the resulting Na+ and K+ currents , when their kinetics were set to temperatures varying from 6 to 37°C ( Figure 2; supplemental Figure S1 ) ., When fixing the action potential waveform to that obtained at 18°C , we found that making the kinetics faster ( i . e . raising temperature ) dramatically reduced the overlap between the Na+ and K+ currents ( Figure 2A–D ) and reduced the excess Na+ entry ratio ( supplemental Figure S1 ) ., Interestingly , fixing the action potential waveform at that occurring at 37°C , but using the ion channel kinetics occurring at 18°C , resulted in a large increase in Na+ entry and Na+/K+ current overlap , despite the fact that the action potential was much shorter in duration than that which normally occurs at 18°C ( cf . Figure 2B , F ) ., As with the long duration action potential ( Figure 2A–D ) , making the kinetics of the underlying Na+/K+ currents faster ( e . g . increasing temperature ) while injecting the fixed 37°C action potential waveform , resulted in a marked decrease of the total Na+ current and overlap of Na+ and K+ currents ( Figure 2E–H ) ., Our analysis of these results reveals that the faster channel kinetics associated with increased temperature result in a marked decrease in the total Na+ current and Na+/K+ current overlap ., This is primarily the result of an increase in Na+ channel inactivation , especially during the falling phase of the action potential , when the driving force on Na+ is especially large ( see below ) ., Thus , increasing temperature does not result in a decrease in total Na+ current or a decrease in Na+/K+ current overlap by decreasing the action potential duration ., Rather , the increase in temperature results in both a decrease in spike duration as well as a decrease in excess Na+ entry owing largely to the increased rate of Na+ channel inactivation ( see below ) ., Changes in the overlap of inward Na+ and outward K+ currents results in systematic changes in dV/dt of the action potential , and the ratio of the maximal falling to rising dV/dt values , with temperature ( Figure 3 ) ., We define γ as ( IdV/dtImin ) / ( IdV/dtImax ) ., This variable is strongly influenced by the level of separation of the inward Na+ and outward K+ currents , and by other factors such as the peak amplitude of INa and IK ., We include it here because it is a readily measureable variable in real neurons , and therefore useful for comparison with results of our model ., Figure 3B shows that the value of γ increases nonlinearly with an increase in temperature , for both the classical Hodgkin-Huxley model as well as our simple model of a cortical action potential ., Plotting the excess Na+ entry ratio as a function of γ revealed that as γ increases ( representing in part decreased overlap of INa and IK ) , the excess Na+ entry ratio decreases ( Figure 3C ) ., The higher values of γ for the traditional HH model reveals a higher rate of spike repolarization ( relative to spike depolarization ) than is present in our simple cortical model ., We reasoned that the marked changes in overlap of inward Na+ and outward K+ currents during action potential generation with increases in temperature were due to changes in the kinetics of Na+ activation and inactivation , and K+ activation ., Plots of the peak values of the time constants for activation of INa ( τm ) and IK ( τn ) and inactivation ( τh ) of INa , revealed an exponential and strong decrease in all three with increases in temperature ( Figure 4A–C ) ., Phase plots of the INa activation, ( m ) and inactivation ( h ) and IK activation ( n ) variables versus membrane potential during generation of an action potential at 18 and 36°C revealed significant and important effects of temperature ( Figure 4D–F ) ., Since the currents vary over a very wide range of values , a logarithmic scale was used to monitor the smaller values during spike repolarization ( Figure 4D–F ) ., Interestingly increases in temperature from 18 to 36°C resulted in an increase in Na+ channel inactivation during nearly all phases of the action potential , with peak inactivation increasing from 0 . 5% of channels available at 18°C to only 0 . 1% available at 36°C ( Figure 4E ) ., In addition , increasing temperature from 18 to 36°C also results in a significant reduction in IK activation during the rising phase of the action potential ( Figure 4F ) ., Even though increasing temperature increases ionic current kinetics substantially , the plot of m , h , and n versus membrane potential during action potential generation were substantially different from the steady state values of the currents ( m∞; h∞ , n∞; supplement Figure S2 ) ., Comparison of the results in Figures 1B and 4D–F suggest that the main contribution to the increase in efficiency of spike generation with increasing temperature is the strong increase in Na+ channel inactivation ( Figure 4E ) , especially during the first half of spike repolarization , when Na+ activation is still high ( Figure 4D ) , resulting in little excess Na+ entry during the falling phase of the action potential , as well as a decrease in spike duration ., However , changes in IK activation with increases in temperature may also contribute , by decreasing spike duration ., These changes in INa activation/inactivation and IK activation result in an exponential decrease in the total amount of Na+ that enters during each action potential ( Figure 5B , black circles ) , thus decreasing the metabolic demand of spiking ., The critical role of temperature dependent increases in the rate of INa inactivation was confirmed by keeping this rate constant ( τh ) constant while allowing τm and τn to vary ( Figure 5A , red circles ) ., In this circumstance , the strong decrease in Na+ entry per action potential with increases in temperature was reversed , such that the Na+ entry ratio actually increased with temperature ( Figures 5B , D , 6E ) ., Keeping either INa activation rate ( τm ) or IK activation rate ( τn ) constant individually did not abolish the strong decrease in Na+ entry/spike with temperature ( Figures 5B , 6E ) ., The decrease in spike duration with increase in temperature still occurred during constant τh , τn or τm , although this effect was greatly reduced when Na+ channel inactivation ( τh ) was invariant ( Figure 5C ) ., This result indicates that the decreases in spike duration and excess Na+ entry ratio with temperature ( effects that are inter-related; see discussion ) result largely from changes in the kinetics of Na+ channel inactivation ., One complicating factor is that temperature affects the intrinsic excitability and spiking rate of neuronal elements 19 , 21 , 28 , 29 , 30 , 31 , 32 , 33 , 34 ., Indeed , in our HH cortical model , increases in temperature result in an increase in firing rate in response to constant current pulses ( 0 . 005–0 . 02 pA/µm2 , 500 ms ) ., Interestingly , at higher temperatures ( approximately 38–40°C ) , the firing rate in our models increases rapidly with increases in temperature ( Figures 6A ) ., Thus , for a constant current input , increases in temperature result in a marked decrease in the amount of Na+ that enters with each action potential ( Figure 6B ) , but an increase in the number of action potentials generated in response to a constant input such as a square pulse of current ., The total Na+ load ( the number of spikes generated times the Na+ entry per spike ) to a constant square pulse input decreases with temperature to a minimum at approximately 38–40°C ., At temperatures above this minimum , the rapid increase in firing rate results in an increasing Na+ load on the neuronal process ( Figure 6C ) ., We suspected that this non-linear increase in firing rate with temperature in the simple HH model ( Figure 6A ) may result from a non-linear effect on spike afterhyperpolarization ( AHP ) since the amplitude and duration of the AHP largely determines neuronal discharge rate 31 , 33 and it is known that increases in temperature decrease the duration of single spike afterhyperpolarizations in cortical neurons 31 ., Plotting the duration of the model AHP ( measured as the time to return to baseline following the generation of an action potential ) as a function of temperature revealed a highly non-linear relationship , with AHP duration decreasing rapidly with increases in temperature above approximately 37°C ( Figure 7A ) ., Plotting the amplitude-time course of IK ( Figure 7C ) as well as the ratio of the K+ current to Na+ current ( Figure 7B ) indicated that even though the K+ currents during the late phases of the AHP are small , lowering temperature results in a significant increase in their amplitude during the late phases of spike repolarization ., At cold temperatures ( e . g . 18°C ) , IK became especially large even 25–60 msec after the spike ., These changes resulted in a larger , more prolonged spike AHP ( Figure 7 ) ., In confirmation of the important role of changes in IK on these effects on the HH model , we found that keeping the activation rate ( τn ) of IK constant , while allowing the activation ( τm ) and inactivate rates ( τh ) of INa to vary with temperature , nearly abolished the ability of changes in temperature to cause non-linear changes in discharge rate ( Figure 6A , D ) as well as the non-linear decrease in AHP duration with increases in temperature ( Figure 7D ) ., Keeping either the activation rate ( τm ) or inactivation rate ( τh ) of INa constant , while allowing the other kinetic time constants to vary with temperature , did not change the presence of this non-linear relationship , although it did alter the range of temperatures over which it occurred ( Figures 6A , D; 7D ) ., As shown above , keeping the inactivation rate ( τh ) of INa invariant inverted the relationship between the total Na+/spike and temperature , while keeping τm or τn invariant does not fundamentally alter this relationship ( see Figure 6E ) ., Consequently , the relationship between total Na+ entry per direct current pulse ( Na+/signal ) is strongly affected by keeping INa inactivate rate ( τh ) invariant ( Figure 6F ) ., Keeping the activation rate of INa ( τm ) constant has relatively little effect , while keeping the activation rate of IK ( τn ) constant exaggerates the decrease in total Na+/current pulse ( Figure 6F ) ., Our HH-style simulation results suggest that increases in temperature may result in several important changes in neuronal action potential generation:, 1 ) action potentials will become shorter in duration and smaller in amplitude , with a marked decrease in overlap of the inward Na+ and outward K+ currents , resulting in a marked reduction in Na+ load/spike;, 2 ) The firing rate of the neuronal process to a constant increases will increase non-linearly with changes in temperature , particularly at temperatures above approximately 37°C ., Next we tested whether or not these predictions would be confirmed in somatosensory layer 5 and entorhinal layer 2/3 cortical pyramidal cells recorded in vitro in response to constant current pulse injection ( Figures 8 , 9 ) , or during the spontaneous generation of the cortical slow oscillation ( supplemental Figure S3 ) ., As predicted by HH simulations , increases in temperature resulted in a significant decrease in spike duration ( Figure 8A , B ) , spike height ( Figure 8C ) , and a steady increase in firing rate in layer 5 somatosensory cortical pyramidal cells between approximately 20 and 35°C in response to the intrasomatic injection of a constant current pulse ( 500 msec , 100 pA; Figure 8D; n\u200a=\u200a6 cells ) ., At temperatures above approximately 35°C , the firing rate increased markedly to the constant current pulse , such that the slope of the frequency-temperature ( f-T ) relationship increased dramatically ( Figure 8D ) ., Interestingly , in layer 2/3 cortical pyramidal neurons spontaneously generating the recurrent network-driven slow oscillation 35 , increases in temperature also resulted in a marked increase in neuronal spiking during Up states ( supplemental figure S3; n\u200a=\u200a10 cells ) , as reported previously 36 ., Similarly , layer 2/3 cortical pyramidal cells also increased their responsiveness to the intracellular injection of a current pulse ( 150 pA; 500 msec ) with increases in temperature from 23 to 42°C ( Figure 9A–C , E; n\u200a=\u200a10 cells ) ., Increases in temperature resulted in a small depolarization of layer 2/3 entorhinal cortical pyramidal cells ( −76 . 7+/−3 . 2 mV 23°C; −74 . 9+/−3 . 1 mV 36°C; −68 . 9+/−3 . 7 mV 42°C; p<0 . 01 , t-test between Vm at 23 and 42°C ) ( Figure 9D ) ., The increase in responsiveness to a constant current pulse was only partially due to the small depolarization of the resting membrane potential with increases in temperature ., Compensation for the change in membrane potential with the intracellular injection of current did not abolish the increase in responsiveness with temperature ( Figure 9E , compare orange and green bars; n\u200a=\u200a10 ) ., Since the number of action potentials and the discharge rate changed with temperature , we were unable to measure the effects of temperature on single spike or spike train induced afterhyperpolarizations ., Previous results have demonstrated that decreases in temperature slow the kinetics of fast , medium , and slow afterhyperpolarizations , although to a differential degree , presumably owing to the properties of intracellular Ca2+ signaling 31 ., In both layer 5 and layer 2/3 pyramidal cells , as predicted by the HH model , the action potential duration ( as measured at half amplitude ) decreased exponentially with increases in temperature ( Figures 8B , 9F ) , action potential amplitude decreased with temperature ( Figures 8A , B; 9A–C ) , and the ratio of the minimum dV/dt to maximum dV/dt during the spike increased with temperature ( Figures 8E , 9G ) ., These results confirm the validity of this model as a basic representation of the effects of temperature on cortical action potential generation ., Increases in temperature from 23 to 42°C also resulted in a significant decrease in apparent input resistance in both layer 5 ( 105+/−15 MOhms 23°C; 88+/−11 MOhms 42°C; p<0 . 01 ) and layer 2/3 ( 246+/−25 MOhms 23°C; 164+/−31 MOhms 42°C; p<0 . 01 ) pyramidal neurons ., Increases in temperature from 23 to 37°C also resulted in a small decrease ( −1 . 6+/−1 . 7 mV; n\u200a=\u200a10; p<0 . 05 ) in spike threshold for layer 2/3 pyramidal neurons ( −47 . 2+/−3 . 4 mV 23°C; −48 . 8+/−2 . 9 mV at 37°C ) ., Simulation of changes in apparent input resistance and resting membrane potential reveal that , as expected , both decreasing input resistance and hyperpolarization decrease firing rate at nearly all temperatures , without changing strongly the shape of the f-T relationship ( supplemental Figure S4A ) or the relationship between temperature and total Na+ entry per current pulse ( supplemental Figure S4B ) ., These results suggest that decreases in apparent input resistance with temperature will at least partially offset the depolarization of resting membrane potential ( Figure 9D ) ., Simulations also indicate that decreases in apparent input resistance and membrane time constant with increases in temperature will result in only small increases in excess Na+ entry per action potential ( see supplement Figure S5 ) , although these may increase the energetic costs of the network as a whole , owing to the requirement for more action potentials per unit of time in the presynaptic neuronal network in order to reach firing threshold in the postsynaptic cell ., One possible confounding factor in our in vitro recordings is that increasing temperature indirectly increased neuronal excitability through decreasing the oxygen content of the bathing solution ., To examine this possibility , we measured the oxygen content of the ACSF at the upper interface of layer 2/3 of the entorhinal cortical slice and the bath solution while varying temperature ( Figure 10A ) ., We then independently reduced the oxygen content of the ACSF while maintaining a constant temperature ( Figure 10; n\u200a=\u200a8 ) ., Increasing temperature from 30 to 41°C resulted in a marked decrease in oxygen content of the bathing solution , from an average of 96 . 5 ( +/−3 . 4; n\u200a=\u200a5 ) to 15 . 4 ( +/−4 . 5 ) mm Hg ( Figure 10A; n\u200a=\u200a5 ) ., As observed previously , increasing temperature resulted in an increase in pyramidal cell action potential response rate to a constant current pulse ( Figure 10B , C ) ., In contrast to these effects , acutely decreasing ACSF oxygen content from 98 . 5 ( +/−3 . 2 ) down to 6 . 4 ( +/−4 . 3 ) mm Hg did not significantly affect action potential duration ( Figure 9D ) , and resulted in a small , but statistically significant ( p<0 . 01; t-test between lowest and highest mm Hg ) decrease in action potential response rate ( from 35 . 1+/1 . 4 Hz at 98 . 5 mm Hg to 33 . 4+/−0 . 95 Hz at 6 . 4 mm Hg ) to the intracellular injection of a depolarizing current pulse ( Figure 10E , F ) ., These results indicate that decreases in ACSF oxygen content do not explain the increase in neuronal excitability associated with increases in temperature , and if anything , result in the under-estimation of the magnitude of this effect ., Increases in temperature have marked and strong effects on action potential generation , and these in turn impact upon the energy efficiency of neuronal activity ., Here we demonstrate that increases in temperature result in a large increase in the energy efficiency of single action potential generation owing to the increased rate of Na+ channel inactivation and subsequent decreased spike amplitude , duration and overlap between inward Na+ and outward K+ currents ., Thus , increases in temperature naturally result in a marked decrease in excess Na+ entry during spike generation , reducing the need for activation of the Na+/K+ ion pump ( ATPase ) , and thus reducing energy expenditure ., This result suggests that the higher body temperature of endotherms such as mammals ( versus ectotherms such as squid ) has the advantage of resulting in a marked increase in energy efficiency of single action potential generation ., However , increases in temperature also resulted in an increase in spike discharge rate to a constant amplitude input or during spontaneous network activity , owing in part to decreases in the amplitude and duration of K+ currents initiated by action potentials 31 , 33 ., Interestingly , Hodgkin-Huxley style models , and whole cell recordings from cortical pyramidal cells , reveal increases in firing rate to be particularly pronounced at temperatures above approximately 37°C ., Thus , even though individual spike efficiency increases with increasing temperature , the enhanced firing rate raises energy requirements ., Maximal overall energy efficiency in neuronal responsiveness is observed near 37°C ., Energy expenditure in the brain is divided among requirements for action potentials , synaptic potentials , maintenance of resting membrane potential , axonal and dendritic transport , and other metabolic functions 1 , 5 , 6 , 7 , 10 , 11 , 21 , 37 ., Estimates of the relative contribution of energy demands related to action potential generation to the overall energy needs of the mammalian brain have varied from approximately 25 to more than 50% 5 , 10 ., The estimates of high energy demands related to action potentials have led to the speculation that average firing rates in the brain may be very low ( <0 . 2 Hz ) 6 , a hypothesis that has some experimental support 38 ., However , the estimates of unusually high energy demands of action potential generation have been based largely upon the observation by Hodgkin 9 of four times excess Na+ entry during spike generation in the squid giant axon 5 , 13 ., This observation and calculation was performed at 18°C , and the results must be corrected to 36–39°C in order to be applied to the mammalian brain ., Unfortunately , this correction has not been systematically applied ., Our calculations predict that at 37°C , there should be relatively little excess Na+ entry during action potential generation ( excess ratio of around 1 . 3 ) ( Figure 1A ) ., Recent observations in cortical pyramidal cells and axons at 37°C confirm the high energy efficiency of action potential generation in these neurons , owing in part to a markedly reduced overlap in inward Na+ and outward K+ currents 12 ., Taking these observations into account suggests that the energy load of action potential generation in endothermic animals may be as much as three times lower than previously calculated , allowing for a significantly higher average discharge rate ., Increases in temperature increase the kinetics of conformational state changes in all ionic channel types involved in action potential generation , with a Q10 ranging from 1 . 5 to 4 17 , 18 , 26 , 27 ., Particularly important for action potential energy efficiency is the temperature dependent increase in rate of Na+ channel inactivation 13 , 18 , which markedly reduces the duration of action potentials , with a smaller effect on action potential amplitude , as well as decreasing the overlap between inward Na+ and outward K+ currents owing to nearly complete Na+ channel inactivation during the falling phase of the spike at 37°C ( Figures 1–3 ) ., Normal mammalian brain temperature varies from approximately 36–39°C , depending upon state of the animal ( e . g . resting , exercise ) and location within the brain , although some mammals ( and some species of birds ) can exhibit brain temperatures as high as 45°C 19 , 20 , 21 , 22 , 23 , 24 ., Under stress , such as during infection or environmental conditions that reduce the effectiveness of body cooling mechanisms ( e . g . high humidity and temperature ) , human brain temperature can reach levels in excess of 40°C 21 ., Rapid rises in temperature to high levels , especially in children and adolescents , can result in the initiation of a febrile seizure 39 , suggesting that the operational balance of excitation and inhibition is temperature dependent ., Even small changes ( e . g . 1–2°C ) in brain temperature can have significant effects on network function ., Prior investigations of the effect of increased temperature on neurons reveal consistent changes in action potentials including decreased duration and amplitude , increased rate of rise and fall , and decreased spike afterhyperpolarization , as partially predicted by HH equations for changes in the kinetics of the underlying ionic channels 19 , 21 , 28 , 29 , 30 , 31 , 32 , 33 , 34 ., In addition , increases in temperature typically result in a decrease in membrane input resistance , and can , in some cell types , depolarize the membrane potential through the activation of TRPV channels which conduct cations and have a reversal potential well above rest 32 , 40 , 41 , 42 ., Interestingly , even without changes in membrane potential , simple HH equations predict that increases in temperature will result in enhanced neuronal res | Introduction, Results, Discussion | The energy efficiency of neural signal transmission is important not only as a limiting factor in brain architecture , but it also influences the interpretation of functional brain imaging signals ., Action potential generation in mammalian , versus invertebrate , axons is remarkably energy efficient ., Here we demonstrate that this increase in energy efficiency is due largely to a warmer body temperature ., Increases in temperature result in an exponential increase in energy efficiency for single action potentials by increasing the rate of Na+ channel inactivation , resulting in a marked reduction in overlap of the inward Na+ , and outward K+ , currents and a shortening of action potential duration ., This increase in single spike efficiency is , however , counterbalanced by a temperature-dependent decrease in the amplitude and duration of the spike afterhyperpolarization , resulting in a nonlinear increase in the spike firing rate , particularly at temperatures above approximately 35°C ., Interestingly , the total energy cost , as measured by the multiplication of total Na+ entry per spike and average firing rate in response to a constant input , reaches a global minimum between 37–42°C ., Our results indicate that increases in temperature result in an unexpected increase in energy efficiency , especially near normal body temperature , thus allowing the brain to utilize an energy efficient neural code . | Conserving energy is essential to life ., The brain , while only 2% of the body mass , uses an astounding 20% of its energy ., It has long been assumed that this large energy consumption was due to the need to generate the electrical signals through which brain cells communicate: the action potentials ., However , recent results reveal that the wires of the mammalian brain – the axons – are remarkably energy efficient ., How is this energy efficiency obtained ?, Here we addressed this question by performing recordings and computational models of mammalian brain cells ., We found that the increase in body temperature associated with the evolution of warm-blooded animals had an energetic benefit ., The action potentials of warm-blooded animals became remarkably energy efficient , owing simply to the increase in body temperature ., These results indicate that mammalian brains , although requiring a great deal of energy to operate , are actually more efficient than expected . | cellular neuroscience, ion channels, computational neuroscience, single neuron function, biology, neuroscience | null |
journal.pntd.0002969 | 2,014 | A Novel Live-Attenuated Vaccine Candidate for Mayaro Fever | Mayaro virus ( MAYV ) is an important and growing human health concern in the neotropics ., First isolated in Mayaro county , Trinidad in 1954 , cases of Mayaro fever ( MAY ) have since been reported in 9 different countries in northern South America 1 ., In addition , serological surveys suggest that MAYV has expanded into the Central American countries of Costa Rica , Guatemala , and Panama 2 ., Typical presentations of MAY consist of an acute febrile illness accompanied by headache , retro-orbital pain , myalgia , vomiting , diarrhea , and rash 3 ., However , the hallmark manifestation of MAY is arthralgia 4 , which is often severe and debilitating , and can persist for up to a year , with recurring relapses possible ., The high incidence of dengue fever in the same areas in which MAYV circulates , and the similarity of the initial signs and symptoms , leads to the misdiagnosis and underreporting of MAY cases 5 , 6; therefore , MAYV is typically neglected as an important cause of tropical diseases ., For example , in several areas of northern South America approximately 1% of all febrile illness that is clinically similar to dengue is caused by MAYV 7 ., MAYV is a zoonotic pathogen that circulates in an enzootic cycle involving Haemagogus spp ., mosquitoes and as yet unidentified vertebrate hosts 3 ., Although seropositivity has been detected in birds and rodents , non-human primates have consistently demonstrated the highest rates of antibodies , suggesting that they are the principal reservoir hosts ., Infection of humans typically occurs in communities near humid tropical forests , and is often associated with logging or other forest activities 1 , 8–10 ., However , as land use and demographic changes in South America lead to human populations expanding within regions of tropical forest , an increasingly higher percentage of the population may be at risk 11 ., In addition , the demonstration that the urban mosquito , Aedes aegypti , can transmit MAYV after exposure to bloodmeals with titers approximating human viremia levels 5 , 12 raises the concern that the virus could emerge into an urban transmission cycle similar to that of its close relative , chikungunya virus ( CHIKV ) ., MAYV belongs in the family Togaviridae , genus Alphavirus ., Despite circulating exclusively in the New World , MAYV belongs genetically , antigenically 13 , 14 ., The genome of MAYV is a single-stranded , positive sense RNA , approximately 11 . 45 Kb in length that encodes 4 nonstructural proteins ( nsP1-4 ) on the 5′ end and 3 structural proteins on the 3′ end , including the capsid and envelope glycoproteins , E1 and E2 ( Fig . 1 ) 13 , 15 ., Genomic RNA includes 2 open reading frames ( ORFs ) ; the nonstructural polyprotein ORF is translated in a cap-dependent manner from genomic RNA , while the structural polyprotein ORF is translated from a subgenomic RNA transcript , which is also capped 16 , 17 ., There is no licensed vaccine available for MAY , and current control strategies rely on reducing human exposure to potentially infected mosquito vectors ., Only one attempt to generate a vaccine for MAYV infection is described in the literature 18 ., Formalin inactivation of wild-type ( wt ) MAYV strain TRVL15537 was tested in immunocompetent CD-1 mice using a single vaccination ., This vaccine was immunogenic , and some efficacy was demonstrated via passive transfer of immune mouse sera to infant mice , followed by lethal challenge ., The ideal MAYV vaccine would produce rapid , long-term immunity after a single dose to rapidly control outbreaks , with a low risk of adverse side effects ., The vaccine would also need to be cost effective for use in resource-poor parts of Latin America , and easy to produce ., For a live-attenuated vaccine , which typically meets most of these criteria , mosquito-transmission incompetent would also be highly desirable for use in non-endemic locations ., To produce such a vaccine , we employed an attenuation strategy involving an encephalomyocarditis virus ( EMCV ) internal ribosome entry site ( IRES ) , which has been successfully used for other alphavirus vaccines 19–24 ., Replacement of the subgenomic promoter reduces expression of the structural proteins , which are now translated via the IRES from genomic RNA , and the inefficient recognition of the IRES by insect ribosomes results in a phenotype that is also incapable of replicating in mosquito cells 25 ., For this study , we tested the efficacy of an IRES-based vaccine candidate for MAYV ( henceforth called MAYV/IRES ) , which was highly attenuated , efficacious , and safe when tested in murine models ., A full-length genomic cDNA clone was generated from MAYV strain CH using RT-PCR and standard cloning methods as described previously 20 ., The virus strain , a 2001 human isolate from Iquitos , Peru , was obtained from the World Reference Center for Emerging Viruses and Arboviruses at the University of Texas Medical Branch ., It was passaged once on Vero cells before RNA extraction ., Details on primers and restriction sites are available upon request ., To produce an attenuated MAYV that was capable of replicating in vertebrate cells , but not in invertebrate cells , the translation of viral structural proteins was placed under control of the EMCV IRES , directly downstream from the subgenomic promoter ., The subgenomic promoter was also inactivated with 14 synonymous mutations using standard PCR-based mutagenesis methods ( Fig . 1 ) ., These mutations were chosen to inactivate the promoter while preserving the amino acid sequence of the nsP4 C-terminus ., A single PCR-derived amplicon containing mutated subgenomic promoter and IRES sequence was cloned into wt MAYV plasmid at SanDI – NcoI sites ., The complete cDNA clone was sequenced to ensure that no errors occurred during PCR amplifications or cloning ., Plasmid DNA was linearized with PacI prior to in vitro transcription , semi-quantified by gel electrophoresis , and recombinant viral RNA was electroporated into Vero cells using conditions described previously 20 ., Titers of rescued wt MAYV and MAYV/IRES were both 4 . 0×107 PFU/mL at 28 h post electroporation ., Cell culture supernatants were harvested 28 h post electroporation , centrifuged to pellet cell debris , and stored at −80°C ., All mice were purchased from Charles River Laboratories ( Wilmington , MA ) ., Animal studies were approved by the University of Texas Medical Branch Institutional Animal Care and Use Committee ., African green monkey kidney ( Vero ) and human fetal lung fibroblast cells ( MRC-5 ) cells were purchased from the American Type Culture Collection ( ATCC , Manassas , VA ) and maintained in culture with Dulbeccos Modified Eagles Medium ( DMEM ) supplemented with 5% fetal bovine serum ( FBS ) and gentamicin sulfate and incubated at 37°C with 5% CO2 ., Aedes albopictus-derived C6/36 cells were maintained in DMEM supplemented with 10% FBS , 1% tryptose phosphate broth ( TPB ) solution , and an antibiotic mixture of penicillin/streptomycin at 29°C and 5% CO2 ., Vero and MRC-5 cells were used to assess the replication kinetics of the MAYV/IRES vaccine candidate and wt MAYV ., Cells were grown to 95% confluency in 6-well plates ., Virus was added to each well at a multiplicity of infection ( MOI ) of 0 . 1 plaque forming units ( PFU ) /cell in triplicate and incubated with the cells for 1 h ., The cells were then washed twice with phosphate buffered saline ( PBS ) to remove residual virus , and 2 mL of medium were added to each well ., At designated timepoints ( 6 , 12 , 24 , 36 and 48 hours post infection ( hpi ) for Vero cells , and 24 , 48 , 72 , and 96 hpi for MRC-5 cells ) , the culture supernatant was harvested for virus titration by plaque assay , then fresh medium ( 2 mL ) was added to replace the volume ., To assess the stability of the MAYV/IRES vaccine candidate , 5 passages were performed in duplicate on both Vero and C6/36 A . albopictus cells in T25 flasks , with the cells at 95% confluency before infection at a MOI of 0 . 1 PFU/cell ., As a control , wt MAYV was also passaged ., Vero cells were incubated at 37°C and 5% CO2 for 48 h , while the C6/36 Ae ., albopictus cells were incubated at 29°C and 5% CO2 for 72 h ., Culture supernatants were then collected and used to infect a new flask at the same MOI ., Virus titers from each passage were measured by plaque assay ., To evaluate the genetic stability of the MAYV/IRES vaccine candidate , viral genomes from Vero passages 3 and 5 of both MAYV/IRES and wt MAYV were fully sequenced ., Viral RNA was extracted using a QIAamp Viral RNA Mini Kit ( Qiagen , Valencia , CA ) ., This was followed by RT-PCR which was performed in a two-step reaction process involving SuperScript III One-Step RT-PCR System ( Invitrogen , Grand Island , NY ) in conjunction with Phusion High-Fidelity DNA Polymerase ( New England Biolabs , Ipswich , MA ) ., PCR amplicon sizes were confirmed by gel electrophoresis and then purified by a QIAquick PCR Purification Kit ( Qiagen ) ., A BigDye kit ( Applied Biosystems , Foster City , CA ) was then used to prepare the samples for Sanger sequencing ., Thirty-nine overlapping amplicons were used to cover the entire genome; primer sequences are available from the authors ., Infant outbred CD1 mice have been shown to develop disease similar to humans for the arthralgic alphavirus CHIKV 26 , and were therefore chosen as a model to evaluate the MAYV/IRES attenuation ., Cohorts of six-day-old outbred CD1 mice were infected over the dorsum subcutaneously ( SC ) with 104 PFU , a dose used previously 26 , and were subsequently monitored daily for 10 days for survival and body weight ., To evaluate immunogenicity , cohorts of adult 28-day-old CD1 mice were also infected SC with 105 PFU , and survival and body weights were monitored daily until day 28 post infection ., Mice were bled on days 1–3 after infection , and serum was tested for viremia by plaque assay 27 to assess attenuation ., On day 28 post infection , the animals were bled and a plaque reduction neutralization test ( PRNT ) was performed on the sera to measure antibodies as described previously 27 ., MAYV produces no detectable disease in adult , immunocompetent mice ., Therefore , to assess attenuation , cohorts of ca ., 5–8-week-old interferon type I receptor-deficient A129 mice were infected intradermally ( ID ) on the left footpad ( FP ) with 104 PFU ., The animals were monitored for survival , body weight changes , and viremia ., Footpad swelling was also measured using a caliper at the site of inoculation ., At day 28 post infection , sera were collected and PRNTs were performed ., On day 29 post infection , the mice were challenged SC with 104 PFU of wt MAYV strain CH ., The mice were monitored the following 7 days for survival , change in body weight , and viremia ., The University of Texas Medical Branch ( UTMB ) Institutional Animal Care and Use Committee approved the animal experiments described in this paper under protocol 02-09-068 ., UTMB complies with all applicable regulatory provisions of the U . S . Department of Agriculture ( USDA ) - Animal Welfare Act; the National Institutes of Health ( NIH ) , Office of Laboratory Animal Welfare - Public Health Service ( PHS ) Policy on Humane Care and Use of Laboratory Animals; the U . S Government Principles for the Utilization and Care of Vertebrate Animals Used in Research , Teaching , and Testing developed by the Interagency Research Animal Committee ( IRAC ) , and other federal statutes and state regulations relating to animal research ., The animal care and use program at UTMB conducts reviews involving animals in accordance with the Guide for the Care and Use of Laboratory Animals ( 2011 ) published by the National Research Council ., Analysis of variance ( ANOVA ) followed by a Tukeys post-hoc test , Kruskall-Wallis with Bonferroni correction for multiple comparisons , Kaplan-Meier , and Mann-Whitney test were performed using Prism 5 ( GraphPad Software , La Jolla , CA ) ., P-values<0 . 05 were considered significant ., To assess the replication kinetics , virus derived from electroporated Vero cells was compared to wt MAYV after infection of Vero cells ( Fig . 2A ) ., Infections were performed in triplicate ( n\u200a=\u200a3 ) at a MOI of 0 . 1 PFU/cell ., Both MAYV/IRES and wt MAYV titers peaked 36 hpi , but wt MAYV had a slightly higher titer of 1 . 1×108 PFU/mL while MAYV/IRES had a peak titer of 7 . 8×107 PFU/mL ., Significant differences were seen only at the 48 hpi timepoint ( ANOVA , p<0 . 05 ) ., Plaque morphology was consistent throughout the experiment , with wt MAYV having a slightly larger ( 0 . 5–3 mm ) and more diffuse plaque morphology than MAYV/IRES ( 0 . 5–2 mm ) under 0 . 4% agarose in MEM ( 48 h incubation ) ., MRC-5 cells are well characterized and widely used in cell culture-based vaccine production ., Therefore , we also measured the replication kinetics of the MAYV/IRES vaccine candidate , as well as wt MAYV on this cell line in triplicate wells ( n\u200a=\u200a3 ) at a MOI of 0 . 1 PFU/cell ( Fig . 2B ) ., The MAYV/IRES virus reached a peak titer of 106 . 7 PFU/ml at 72 hpi , which was much later and at a lower titer than wt MAYV ., Plaque morphology measured on Vero cells of MAYV/IRES virus derived from MRC-5 or Vero cells was comparable ., The stability of MAYV/IRES was tested in vitro by 5 serial passages in Vero cells , in duplicate at an MOI of 0 . 1 PFU/cell ., MAYV/IRES maintained a slightly lower titer than wt MAYV throughout the passages , with a range of 4 . 2×107 PFU/mL after passage 2 , to a peak of 1 . 9×108 PFU/mL after passage 3; wt MAYV titers remained between 108 and 109 PFU/mL ( data not shown ) ., To evaluate the genetic stability of the MAYV/IRES vaccine candidate , the complete consensus sequences of passages 3 and 5 were determined using overlapping amplicons generated by RT-PCR , and no mutations were detected ., MAYV/IRES was also serially , blind passaged 5 times in C6/36 A . albopictus mosquito cells to confirm its lack of mosquito host range ., As expected , the virus was not detected during any passage , while wt MAYV replicated to high titers ( data not shown ) ., Cohorts of 6-day-old CD1 mice were infected SC with 104 PFU of either MAYV/IRES ( n\u200a=\u200a14 ) , wt MAYV ( n\u200a=\u200a15 ) , or were sham-infected with PBS ( n\u200a=\u200a15 ) ., Mice infected with wt MAYV began to die starting 3 dpi and complete mortality was seen by day 8 ( data not shown ) ., All MAYV/IRES- and sham-infected mice survived until the study was terminated 10 days after inoculation ., As expected , the wt MAYV-infected cohort did not gain weight as quickly as the MAYV/IRES- or sham-infected animals , and the average weight of wt-infected animals declined rapidly beginning 4 days post-infection ., There was no significant difference in weight change between MAYV/IRES- and sham-infected animals ( Kruskall-Wallis with Bonferroni correction for multiple comparisons ) ., Due to the high mortality in infant CD1 mice infected with wt MAYV , adult CD1 mice ( 28 days-old ) were also tested as a potential virulence model ., Mice were infected SC with 105 PFU of either MAYV/IRES ( n\u200a=\u200a10 ) or wt MAYV ( n\u200a=\u200a10 ) , and negative controls were sham ( PBS ) -infected ( n\u200a=\u200a6 ) ., Unlike the infant 6-day-old CD1 mice , the 28-day-old mice all survived infection with wt MAYV until the study was terminated 28 days after infection ., To assess with greater sensitivity signs of disease , the animals were weighed post-vaccination ( Fig . 3A ) ., The MAYV/IRES- and sham-infected cohorts gained weight steadily throughout the experiment , while the wt MAYV-infected mice lost some weight initially , but recovered by day 5 post-infection , then proceeded to gain weight in a manner similar to the other cohorts ., However , these differences in weight change were not significant ( p≥0 . 07 , Kruskall-Wallis with Bonferroni correction for multiple comparisons ) ., To quantify viral loads of the MAYV/IRES vaccine candidate , viremia was assessed post-vaccination ( Fig . 3B ) ., Both MAYV/IRES and wt MAYV produced a peak viremia titer at day 2 post-infection , but MAYV/IRES viremia was of shorter duration and of significantly lower mean peak titer , just over 103 PFU/mL , compared to 107 PFU/mL for wt MAYV ., Serum neutralizing antibody titers were measured at 28 days post-infection using an 80% PRNT ., MAYV/IRES titers ranged from 160 to ≥640 ( mean\u200a=\u200a≥304 ) , and were not significantly different from those of wt MAYV-infected animals ( Kruskall-Wallis with Bonferroni correction for multiple comparisons ) ( Fig . 3C ) ., A129 mice lack functional type 1 interferon receptors and are therefore a very sensitive model for human arthritic alphavirus infection 28 ., They have been used as a lethal model for alphavirus vaccine safety and challenge studies 20 ., Cohorts of adult A129 mice ( n\u200a=\u200a8 ) were infected with MAYV/IRES or wt MAYV , or sham-infected with PBS ., Injections were performed intradermally on the left footpad with 104 PFU ., All MAYV/IRES- and sham-infected mice survived until the experiment was terminated on day 28 , while all wt MAYV-infected mice died by day 5 ( Fig . 4A ) ., Both the MAYV/IRES and wt MAYV cohorts lost weight initially , but wt MAYV-induced loss was more dramatic and significantly greater than that of the MAYV/IRES-infected animals ( Fig . 4B ) ( p<0 . 01 , Kruskall-Wallis with Bonferroni correction for multiple comparisons ) ., There was no significant difference in footpad swelling among cohorts until 3 days after infection , when wt MAYV-infected mice showed a large increase in footpad diameter , which was significantly greater than mean swelling of both MAYV/IRES- and sham-infected cohorts ( Fig . 4C ) ( p<0 . 01 , Kruskall-Wallis with Bonferroni correction for multiple comparisons ) ., Viremia was measured post-vaccination to quantify the viral load ( Fig . 4D ) ., Both MAYV/IRES and wt MAYV cohorts reached high titers in the peripheral blood , with MAYV/IRES peaking at day 3 post-infection with a titer of 5 . 5×108 PFU/mL and wt MAYV reaching a slightly higher titer of 1 . 4×109 PFU/mL ., Differences were significant only on day one post-infection ( p<0 . 001 , Kruskall-Wallis with Bonferroni correction for multiple comparisons ) ., At day 28 post-infection , 7 of the 8 MAYV/IRES-vaccinated A129 mice had neutralizing antibody titers ≥640 , while the remaining mouse had a titer of 320 ( mean\u200a=\u200a≥604 ) ., The mean PRNT antibody titer for A129 mice was significantly higher than that for CD1 immunocompetent mice ( Students T-test , p<0 . 01 ) , possibly reflecting greater vaccine replication in the former ( although the ages were not exactly matched ) ., The sham-vaccinated A129 mice ( n\u200a=\u200a3 ) did not have detectable antibodies ( <20 ) ., Mice were then challenged SC with 104 PFU of wt MAYV to assess the efficacy of the MAYV/IRES vaccine ., All vaccinated mice survived , while all of the sham-vaccinated mice were dead by day 7 , representing a significant difference in mortality ( p<0 . 01 , Kaplan-Meier; see Fig . 5A ) ., To monitor disease in a more sensitive manner , weight was measured post-vaccination ( Fig . 5B ) ., The sham-vaccinated , challenged cohort lost weight more quickly and dramatically than the MAYV/IRES-vaccinated group ( p<0 . 01 , Mann-Whitney ) ., To assess viral load , viremia post-challenge was also measured ( Fig . 5C ) ., The MAYV/IRES-vaccinated group showed a decreased viremic response upon challenge compared to the sham-vaccinated animals , only reaching a mean titer of 2 . 0×102 PFU/mL at day 3 post-challenge , while the control group reached a much higher titer of 4 . 8×108 PFU/mL 3 days post-challenge ( p<0 . 05 , Mann-Whitney ) ., It has been over 60 years since the discovery of MAYV in Trinidad , and there is still no licensed vaccine available despite continued outbreaks , and the potential for urban transmission in a dengue-like cycle 5 , 12 that could expose millions of people ., Our MAYV/IRES vaccine was designed to offer single-dose , rapid protection to protect people both in endemic regions and in the event of an urban outbreak ., Previous attempts to generate a vaccine to protect against MAY focused on inactivated wt virus 18 ., A single vaccination proved immunogenic in adult CD1 mice , and efficacy was demonstrated indirectly via passive transfer of the immune mouse sera to infant mice , followed by lethal challenge ., However , no further testing of this vaccine has been reported ., To capitalize on the advantages of live-attenuated vaccines , including rapid and long-lasting immunity as well as ease of manufacture , we used the IRES-based attenuation approach that has been demonstrated to offer highly stable and predictable attenuation for alphaviruses 19–24 ., Unlike traditional alphavirus attenuation derived from cell culture passages that typically relies on unstable point mutations , resulting in reactogenicity and the potential for reversion to wt virulence and transmissibility 29–32 the IRES-based rationale approach suppresses structural viral protein expression by elimination of the subgenomic promoter using multiple inactivating mutations ., Thus , reversion is highly unlikely because the promoter sequence is very specific and intolerant of change 33 , resulting in superior attenuation stability following serial mouse passages compared to traditional point mutation-dependent attenuation 22 ., Further safety is achieved through the use of the encephalomyocarditis virus IRES , which inefficiently mediates translation in insect cells 25 , and thus eliminates the possibility for mosquito transmission ., Finally , the titers of nearly 108 PFU/cell of MAYV/IRES produced by vaccine substrate-approved Vero cells should be adequate for large-scale manufacture , and the stability we demonstrated following Vero cell passages will be critical for licensure ., Like previous studies using the IRES-based alphavirus attenuation approach , our results showed that MAYV/IRES is stable in cells of mammalian origin ( Vero ) , but incapable of efficient replication in a C6/36 A . albopictus cell line ., Previous studies have showed that other IRES-based attenuated alphaviruses are also incapable of replication after intrathoracic inoculation into A . albopictus mosquitos 20 , 22 ., In every murine model we tested , MAYV/IRES was highly attenuated , only producing minimal signs of disease in the highly stringent A129 model that cannot mount an effective interferon response ., This vaccine candidate was also highly immunogenic , inducing high levels of neutralizing antibody titers in both adult CD1 and A129 mice at 28 days post-vaccination ., Challenge of A129 vaccinated mice at 29 days post-infection with a high dose of wt MAYV showed complete protection from detectable disease , despite the high virulence and complete lethality of MAYV in unvaccinated animals ., These murine studies indicate that MAYV/IRES is highly attenuated , highly immunogenic , and provides strong protection against MAYV challenge ., Further studies in another animal model are needed ., Typically , nonhuman primates such as macaques reproduce human-like disease after alphavirus infection 24 , 34–39 ., These animals should be evaluated as models for human MAYV to determine if they will be useful for the next steps in preclinical evaluation of MAYV/IRES ., A variety of alternative vaccine development approaches are available for alphaviruses including inactivated virus , subunit protein , DNA and virus-like particles ( VLP ) as well as traditionally attenuated and chimeric vaccines 40 , 41 ., All of these approaches emphasize safety but have significant drawbacks including a multiple dose requirement for efficacy , short-lived immunogenicity necessitating boosters , challenging delivery ( DNA via electroporation ) and complex , expensive manufacture ( VLPs ) and the risk of residual live virus after inactivation , which was shown to result in the death of an eastern equine encephalitis-vaccinated horse in California 42 ., Our MAYV/IRES candidate overcomes all of these shortcomings to generate rapid immunity following a single dose , and should have greatly reduced reactogenicity due to its robust , highly stable attenuation design ., Although further testing should be done to evaluate the duration of protective immunity , other IRES-based alphavirus vaccines have generated completely protective immunity in macaques for over one year ( C . Roy , S . C . W . , unpublished ) ., MAYV/IRES therefore should be ideal for inducing rapid , long-lived immunity after a single dose for use in developing countries where MAYV is endemic , as well as for a travelers vaccine for persons visiting South America ., In summary , our MAYV/IRES vaccine candidate is highly attenuated and immunogenic , unable to infect mosquito cells , and provides protection from lethal challenge in murine models ., These results indicate that further preclinical development of MAYV/IRES is justified for its evaluation as a potential human vaccine that could protect people from MAY in South America , but also on other locations if the virus spreads and urbanizes like the closely related CHIKV 5 , 43–46 ., Furthermore , MAYV/IRES should be evaluated for its ability to protect against CHIKV and Ross River viruses , other closely related alphaviruses that cause epidemics in Africa and Asia , or Australia and Oceania , respectively ., CHIKV is of particular concern because in December of 2013 it invaded the Caribbean , representing the first autochthonous transmission in the Western Hemisphere 47–49 ., This event could portend a major epidemic throughout the Americas if spread to the mainland occurs into dengue-endemic regions where both A . aegypti and A . albopictus mosquito vectors are present along with a nearly naïve human population ., The latter vector is highly susceptible to Asian CHIKV strains with adaptive mutations that dramatically enhance its vectorial capacity 50–55 , and it is unknown if similar mutations could enhance MAYV urbanization in a similar manner ., An effective vaccine could greatly mitigate these risks and have a major impact on public health in South America . | Introduction, Materials and Methods, Results, Discussion | Mayaro virus ( MAYV ) is an emerging , mosquito-borne alphavirus that causes a dengue-like illness in many regions of South America , and which has the potential to urbanize ., Because no specific treatment or vaccine is available for MAYV infection , we capitalized on an IRES-based approach to develop a live-attenuated MAYV vaccine candidate ., Testing in infant , immunocompetent as well as interferon receptor-deficient mice demonstrated a high degree of attenuation , strong induction of neutralizing antibodies , and efficacy against lethal challenge ., This vaccine strain was also unable to infect mosquito cells , a major safety feature for a live vaccine derived from a mosquito-borne virus ., Further preclinical development of this vaccine candidate is warranted to protect against this important emerging disease . | Mayaro virus ( MAYV ) is a mosquito-borne alphavirus that causes severe and sometimes chronic arthralgia in persons in South America , where it circulates in forest habitats ., It is widely neglected because it is typically mistaken for dengue due to the overlap in the clinical signs and symptoms , and the lack of laboratory diagnostics in most endemic locations ., Furthermore , MAYV has the potential to initiate an urban transmission cycle like that of dengue , which could result in a dramatic increase in human exposure ., Because there is no effective vaccine or specific treatment , we developed a candidate vaccine to protect against MAYV infection ., We used an attenuation approach based on the elimination of the MAYV subgenomic promoter and insertion of a picornavirus internal ribosome entry site to mediate translation of the structural proteins ., This vaccine was well attenuated in mouse models , highly immunogenic , and protected against fatal MAYV infection ., Our results indicate that this MAYV strain is promising for further development as a potential human vaccine . | public and occupational health, medicine and health sciences, global health, biology and life sciences, microbiology | null |
journal.pcbi.1006773 | 2,019 | In vitro and in silico multidimensional modeling of oncolytic tumor virotherapy dynamics | Tumor therapy with replication competent viruses ( oncolytic virotherapy ) is an exciting new field of therapeutics ., In principle , amplification of the virus in target cancer cells could allow ongoing spread of the infection within the tumor and its eventual elimination 1 , 2 ., The advantages of recombinant viruses for cancer therapy include, ( i ) specific engineering for infection , replication and killing of tumor cells 1 ,, ( ii ) amplification of the therapy itself by the tumor ,, ( iii ) stimulation of an anti-tumor immune response by breakdown of tumor immune tolerance 3 ,, ( iv ) a bystander effect especially if the virus is armed with specific genes such as the sodium iodide symporter ( NIS ) 4 ., With the exception of cancer therapy with recombinant chimeric antigen receptor ( CAR-T ) T cells , tumor virotherapy is an exercise in population dynamics in which the interactions between the virus , the tumor and the immune system determine the outcome of therapy 5–13 ., Many mathematical models have been developed to describe the outcome of such interactions 5 , 6 , 8–13 ., Most models are based on the Lotka-Volterra approach and assume mass action kinetics with well-mixed populations ., As a result , the models are helpful in illustrating general principles but lack important features , in particular the spatial geometry of the cells in a tumor , to be of predictive value if applied to in vivo scenarios ., This is a critical deficiency especially if we are to attempt optimization of therapy 9 ., Durrett and Levin and many others have addressed the problem of spatial constraints on the interactions between populations in ecological systems 14–16 and reference therein ., More recently , Paiva et al described a three-dimensional computational simulator of tumor and virus interactions and concluded that complex dynamics are in place with the spatial arrangements between cells being important determinants of outcome 17 ., Reis et al reported on a 3D computational model of cancer therapy that illustrated the important differences when considering dynamics in 2 versus 3 dimensions and how restricted the parameter space may be to achieve tumor eradication 18 ., Wodarz and colleagues have reported on their work with agent based modeling of tumor virotherapy where space is explicitly considered 7 , 19 ., Using experimental data on the spread of adenovirus in a monolayer ( 2D ) of 293T cells as a guide , they showed that various patterns of virus spread such as ‘hollow ring structure’ , ‘filled ring structure’ and a ‘dispersed pattern’ are possible and how space and virus/tumor cell parameters can interact to determine the outcome of therapy 7 ., Ring structure formation is associated with a quadratic growth of the virus population which subsequently becomes linear ., The dispersed pattern of spread is invariably associated with therapeutic failure while the ring structures may be associated with a cure especially if the center of the ring is associated with elimination of the target population and the virus continues to expand radially and catch up with all the target population ( since a boundary will be reached ) ., Wodarz and colleagues found that the local dynamics on a smaller scale can predict the outcome of the spatially explicit system 7 ., Interestingly , the experiments also showed two patterns of infection–limited spread versus robust expansion of the infected cell population ., Which pattern the infection followed was established early on ., Subsequently , they showed that in part this dichotomy in outcomes was due to interferon induction in infected cells that inhibited virus spread 19 ., This suggests that there is a local race between spread of the virus against the development of an interferon response which limits viral replication ., Modeling suggests that multiple infections by the virus are also necessary to explain the dynamics , especially when the populations are small ., However , many modeling approaches described to date have generally lacked any experimental data to validate them ., To address this problem , we have developed an in silico computational model that captures the dynamics between the tumor and virus populations in a spatially explicit manner ( two and three dimensions ) ., We use in vitro 2D and 3D data to inform the model parameters and then use the computational model to explore various critical properties of oncolytic viruses ., We show that the introduction of a third dimension alters the dynamics significantly and that this has important implications for the outcome of therapy ., To quantitate cell populations based on fluorescent imaging , we initially determined the pixel area that represents a cell ., Two independent observers quantitated the number of cells ( n = 375 cells per time point per observer ) in a given area based on phase contrast images and the pixel area of the cells in the corresponding fluorescent images ., The median number of pixels was 326 . 9 versus 323 ( p = 0 . 8723 , Mann Whitney ) for each observer ., There was excellent agreement between the two observers ( Fig 1A ) ., This means that the inter observer variability in cell area was 1 . 1% ., The number of cells present in a growing population was measured at 6 different time points ., We determined that the cell populations grew exponentially in 2D culture ( Fig 1B ) ., The estimated doubling time for the population was 28 hours based on an exponential fit to the data ., In contrast , the rate of replication of individual cells based on serial tracking was estimated to be 21 . 9 hours ( Fig 1C ) and varied from 20 hours for the first replication ( n = 127 events ) to 22 . 5 hours for the second ( n = 97 events ) ., The difference between these two observations was not statistically significant ( Wilcoxon sign rank test , p = 0 . 6563 ) ., There was a strong positive correlation between the two cell cycle times observed ( Spearman’s rho = 0 . 795 , p = 0 . 0072 ) ., We measured the growth of tumor cells in 3 dimensions serially by imaging multiple spheroids at specific time points ( Fig 2A ) ., Each spheroid was monochromatic , implying that each spheroid arose from one founder cell even though a mixture of HT1080 cells with all 4 colors ( blue , yellow , green and red ) were plated ., We found a linear increase in diameter of the spheroids as a function of time although there was considerable variability as the spheroids grew ( Fig 2D ) ., The median increase in spheroid diameter was 15μm/day or 0 . 63 μm/hr ., In addition , we also determined the radius of gyration of representative spheroids ( n = 12 ) across the 3 axes of growth as a function of time 20 ., As can be seen from Fig 2E , the tumor cells growing in 3D generally retained a spherical shape with a median radius of gyration of 97 . 5μm in the XY plane , 114μm in the XZ plane and 101 . 7 μm in the YZ plane ., Given that the average diameter of a cell is ≈10μm , our observations suggest that the variability in the radius of gyration was of approximately 1 cell in any axis and therefore growth of the spheroids was generally uniform in all directions ., Since oncolytic measles viruses ( as well as other viruses ) generally spread from cell to cell , we hypothesized that the number of cells surrounding any given cell is of critical importance ., Therefore , we wanted to determine the number of nearest cell neighbors based on whether cells are growing in the 2D plane versus in 3 dimensions ., This data informed the development of the computational model to realistically simulate the in vitro dynamics ., We studied cell populations by Voronoi tessellation analysis to determine the distribution of nearest neighbors for cells growing in the 2D ( Fig 3A and 3B ) plane as well as in spheroids ( Fig 3C and 3D ) ., As expected , the number of nearest neighbors was significantly different in 2D versus 3D with a median of 6 ( range: 3–10 ) versus 16 ( range: 4–30 ) neighbors respectively ., We utilized serial imaging studies to determine the rate of growth of the tumor and virus infected cell populations both in the 2D plane and in 3 dimensions ., A total of 14 independent experiments were studied in 2D ., Fig 4 presents snapshots of the spread of a single focus of infection ( green ) due to syncytium formation where cells fuse together to form a multicellular object ., In Fig 5A–5C , we provide a representative case of data capture , digitalization and then analysis of cell population size by the Voronoi tessellation method ( C ) ., Fitting of serial imaging data to the mean field solution ( see methods ) , enabled us to determine the best parameter estimates for cell replication and virus spread ( Fig 5D ) ., Although the rates of tumor cell and virus spread varied , the median rate of replication for tumor cells was 4 . 39 per hour while the virus infection was spreading at a median rate of 18 . 94 cells/hour which implies that the virus was spreading 4–5 times as quickly as the tumor cell population was growing ., A faster rate of spread of the virus compared to tumor cell growth is a necessary condition for any plausible scenario where the virus can eliminate the tumor cell population leading to a potential cure–something that is consistently observed in vitro 21 , 22 and also predicted by others 7 , 19 ., We used the best estimate of the parameter set obtained from the data fitting ( the black dot in Fig 5D ) to determine cell population size and compare that to the actual measurements ., As can be seen from Fig 5E , the computational output mirrored the experimental results with a high degree of accuracy ., In virtually all of our experiments with cells growing in the 2D plane , the virus consistently eliminated the tumor cell population within 48 hours ., The dynamics of virus spread in 3D tumor spheroids were surprisingly different with the virus spreading more slowly in the 3D environment ., Although various independent foci of infection occurred in each spheroid ( Fig 6 ) , with the formation of multinucleated syncytia ( red ) , many infected cells remained viable for the duration of the experiment ( ~7 days ) ., We also observed that many cells in the spheroids never become infected despite being in close proximity to virus-infected cells ., More recently we documented syncytium formation in vivo in a mouse dorsal skin fold chamber model of cancer growth ( Kemler et al–submitted ) where again we observed cells in close proximity to highly infected foci that did not become infected for the duration of the experiment ., We utilized these observations to perform in silico simulations of cell dynamics either in the 2D plane or in 3 dimensions each under two scenarios: growth on a regular lattice or growth on a Voronoi lattice ., We studied the dynamics across a wide range of parameter estimates ( Figs 7 and 8 ) ., All four networks studied had 1 × 106 nodes with the 2D networks having a dimension of 1000 × 1000 while the 3D networks had 100 × 100 × 100 dimensions ., At the start of each simulation , 90% of the nodes were occupied by normal cells , 9% were occupied by cancer cells and the initial viral inoculum infects 1% of the tumor cell population ., If these simulations were allowed to run on an infinitely large and complete network ( appropriately defined ) , the simulations would be stochastically identical to the mean field equations ( see Methods ) ., Starting with simulations in 2D , for the set of parameters chosen , simulations led to equilibria with the three cell populations present ., The time to reach an equilibrium in the 2D regular lattice architecture was ~6000 time units ( average number of neighbors: 4 ) , while in the 2D Voronoi lattice ( average number of neighbors: 6 ) , the time to equilibrium was 5000 time units ., In the case of 3D simulations , the time to equilibrium on the regular lattice ( average number of neighbors:, 6 ) was 250 time units , while in the case of the 3D Voronoi lattice ( average number of neighbors: 16 ) , the average time to equilibrium was 150 time units ., Therefore , the main determinant of the speed to reach equilibrium is the dimensionality of the network more than the number of neighbors , although the latter is also important ., In Figs 7 and 8 , we illustrate specific examples of such simulations in 2D ( Fig, 7 ) and 3D ( Fig 8 ) ., In parallel , we also determined the results of the mean field solutions given by the mathematical model ., It is clear that the mean field solution overestimates the effect of therapy with a larger population of infected tumor cells at equilibrium both in the 2D and 3D simulations ., The mean field solution also overestimates the speed at which equilibrium is reached ., Spread of the virus in 3D leads to a larger fraction of tumor cells infected at equilibrium compared to the 2D scenario but overall the tumor cell population is larger at equilibrium in the 3D network and illustrates the difficulty of controlling the 3D tumor compared to the tumor cells growing in vitro ., There are also striking differences in the pattern of infection in 2D versus 3D that again illustrates the role of connectivity between cells ., There are five outcomes of tumor virotherapy regardless of the model and number of dimensions considered ., ( i ) The tumor population will go extinct and the virus infected tumor population will soon follow , leading to permanent cure of the tumor ., ( ii ) The virus infected cell population goes extinct and the result will be the eventual takeover of the simulation space by the tumor cells since they grow faster than normal cells ., This will mean that therapy has failed ., ( iii ) The three populations of cells co-exist and have a ( non zero ) stable size ., This will imply partial success of therapy ., ( iv ) Normal cells are eliminated and at equilibrium only tumor cells and infected tumor cells coexist ., ( v ) All populations die out ., We do not consider the last scenario in our simulations ., We were mainly interested in the range of virus specific parameters that maximize the chance of tumor elimination ., Using data from prior work on in vivo tumor control with the same virus 8–11 , we fixed the replication and natural death rates of normal and cancer cells and varied the parameters for virus replication and virus induced cell death rates across a wide range of values ( λ3: 0 − 100; δ3: 0 − 15 ) ., A total of 14 , 000 simulations were performed with each simulation continuing until either the tumor cell population was eliminated or 1000 days had passed , whichever came first ., We report the cumulative results of these simulations in Fig 9 ., As can be seen , the results are qualitatively different ., The mean field solution predicts that most of the time , the 3 population equilibrium is the most likely outcome ., In 2 dimensions , the parameter range where cure of the tumor is possible is wider compared both to the mean field estimate and the 3D simulations ., Moreover , in 2D there is very little difference in output between the regular grid lattice and the Voronoi lattice likely due to the fact that the number of nearest neighbors is similar ( 4 versus 6 respectively ) ., However , the probability of a cure is less for a Voronoi type network in 3D compared to a regular lattice , although the Voronoi lattice increases the chances for the co-existence of all three populations with less chance of the tumor taking ( failed therapy ) over compared to the 3D regular lattice ., This is likely due to the higher number of neighbors that each cell possesses which increases the chance for infection ., All simulations agree that the ideal virus should replicate rapidly ( high λ ) but kill cells slowly ( low δ ) ., Indeed , the model shows that there is a wider tolerance for replication rates and less so for the death rates of infected cells ., Tumor eradication is more likely on a 2D surface compared to a 3D object for any set of parameters even though in 3D the number of cell neighbors is higher and equilibrium is reached faster , implying faster dynamics of virus spread ., This is compatible with our in vitro observations and illustrates some of the intrinsic barriers to virus spread imposed by a 3D architecture versus a surface ., Tumor therapy with replication-competent viruses is an exciting novel approach to cancer therapy in which the target to be eliminated amplifies the agent responsible for its own death ., Perhaps the only other member of this paradigm is cancer immunotherapy with chimeric antigen receptor T cells that are stimulated to replicate by engagement of cell surface antigens expressed by tumor cells ., However , for successful tumor control with viruses , the latter have to establish foci of infection within the tumor , replicate to amplify the virus population and spread across the tumor ., At the same time , the virus has to evade as much as possible the immune response that can neutralize the virus population or eliminate infected cells which would halt virus propagation ., The outcome of such therapy is highly dependent on the dynamic interactions between the various populations of cells 5–7 , 23–26 ., However , as our results show , the outcome is also quite sensitive to the architecture of the tumor since there may be several barriers to the spread of the oncolytic virus ., These barriers may be physical or chemical in nature 13 , 27 ., It has been argued that modeling with differential equations that provide a mean field approximation may be good enough to optimize therapy with these viruses and that such equations can capture well the dynamics without the need to consider space explicitly 26 ., We have addressed the problems with this postulate in our work ., Initially we provided a detailed analysis of tumor cell growth and virus spread in vitro both in 2D and 3D coupled with an analysis of the rate of replication of cells as well as the number of neighboring cells in a given environment ., Our modeling approach differs from other publications 17 since we used this data to generate realistic computational models of tumor cell growth and virus spread ., Subsequently , we analyzed through simulations many potential scenarios that consider two critical virus parameters: its rate of replication and the rate at which it kills cells ., These two parameters have been repeatedly shown to be important for the outcome of virotherapy 6 , 8–10 , 12 , 13 , 17 , 18 , 28 ., The wide spectrum of viruses available for oncolytic therapy have different kinetics of spread or can be engineered to alter their kinetics of replication or cell killing 1 ., Our in vitro studies show that the same virus will kill cells more slowly in a 3D environment compared to the 2D setting , despite the fact that in 3D the average number of cells in a neighborhood is higher ., As a result , in 3D , for the same set of parameters , the probability of tumor elimination is lower compared to the mean field approximation and also less than in a 2D environment despite the system reaching equilibrium by at least an order of magnitude faster ., This fits well with all in vitro studies where the virus is highly efficient in killing tumor cells growing in the 2D plane but less so in a 3D environment whether in vitro ( spheroids ) or in vivo as tumor xenografts 8–12 , 29 ., There are several possible explanations for the difference in outcome for tumor control in 2D versus 3D ., While in 2D virtually all the tumor population is likely accessible to the virus that spreads at a fast rate , the same cannot be said for the 3D scenario where geometry not only makes some areas of the tumor quite distant from the infected foci but the virus also physically appears to spread at a slower rate even between adjacent cells ., Moreover , one can envisage scenarios in 3D where a part of the tumor loses contact with the main tumor that is being infected ., This will impose even greater spatial restrictions on the spread of the virus and reduces the probability of tumor control even further ., However , in 3D the equilibrium is reached faster due to the higher number of contacts between cells ., It is not difficult to see why the mean field model will overestimate the effect of therapy , since this approach assumes the presence of a well-mixed population based on mass action kinetics , thus rendering tumor cells accessible to viruses at all times ., However , in a 3D structure such as a tumor , not all cells are at the same risk of being infected due to their spatial proximity , or lack thereof , to infected foci ., Any part of the tumor that loses contact with infected foci will result in tumor regrowth unless virus can diffuse and establish a new infection there–something that appears to be unlikely with the current scenarios 27 ., Moreover , we have observed from in vivo studies that many cells in proximity to a highly infected focus never become infected and the size of infected foci can be quite variable ( Kemler et al , submitted ) ., The biological and physical bases for these observations require further analysis ., Our work complements that of Wodarz et al 7 , 19 who studied spread of an adenovirus in a monolayer and showed the importance of local interactions on the spread of the virus ., Their work expanded on the importance of initial conditions and the potential impact of an antiviral state due to interferon production ., It is important to note that 293T cells used in their experiments are not derived from a tumor and so are likely to respond to interferon production ., In contrast , we used cell lines that are derived from tumors that generally do not mount a robust immune response against measles virus ., We also extended our in vitro studies and computational modeling to 3D where the number of neighbors and the spatial structures become more complex ., Our results show that the number of neighbors surrounding a cell can facilitate spread of the virus and leads to the three population equilibrium more often and more quickly ( compare Voronoi network with grid lattice in Fig 9 ) ., However , in the presence of an immune response , we hypothesize that the outcome could be worse for a Voronoi type lattice compared to a regular lattice since the latter has a higher probability of a cure ., The simulations also show that the mean field solution generally provides a more optimistic view of the outcome compared to the Voronoi type architecture that seems to exist in spheroids ., However , the higher number of neighbors in a Voronoi network is associated with failure of therapy less often compared to a regular grid lattice at least in theory ., Our results highlight the need for spatially explicit modeling to accurately capture the dynamics of tumor virotherapy ., We also show the problems that arise from the introduction of a third dimension into such a model–the probability of a cure decreases significantly when the virus is used in an attempt to cure a 3D tumor compared to cells in the 2D plane ., The human fibrosarcoma cell line HT1080 was obtained from ATCC and grown in Dulbecco’s modified Eagle’s medium ( DMEM ) supplemented with 10% fetal bovine serum ( FBS ) and maintained at 37°C with 5% CO2 ., 293T cells were maintained in DMEM with 10% FBS ., The human cell line 293-3-46 was maintained in DMEM with 10% FBS and geneticin ( 1 . 2mg/ml ) while Vero cells were maintained in DMEM with 5% FBS ., PCR products containing the yellow fluorescent protein ( YFP ) , enhanced blue fluorescent protein ( eBFP ) , and the tdTomato genes were generated using Roche Fast-Start High Fidelity PCR kit using pCAG-YFP , CFP , pCSCMV:tdTomato ( Addgene ) as template DNA ., The primers ( Forward: GGGATCCACGCCACCATGGTGAGCAAGGGCG and reverse: GAGGCGGCCGCAGTTTACTTGTACAGCTCGTCCATGCC ) had restriction sites for BamHI and NotI to facilitate cloning in the lentiviral vector backbone 30 ., PCR products were cloned into pHR`CMV-eGFP-SIN ( a gift of Dr Y . Ikeda , Mayo Clinic , Rochester ) after the excision of GFP via digestion with BamHI and NotI ., The resulting constructs were verified by restriction digestion and sequencing ., Lentiviral particles containing the respective reporter genes were generated via transfection of 293T cells with pMD . G , pΔCMV . 8 . 91 , and the plasmid encoding the vector genome with the fluorophore as previously described 30 ., Vector containing supernatants were harvested after 72 hours , filtered ( 0 . 42μm ) and used to transduce HT1080 cells ., Cells expressing the reporter gene were sorted by flow cytometry and plated as single cells into 96 well plates in DMEM with 10% FBS and expanded as clones for further studies ., In order to generate recombinant replication competent measles viruses expressing different fluorescent reporter genes , PCR products containing the fluorophores eBFP , YFP , and tdTomato were generated using pCAG template DNA ( Addgene ) ., PCR primers had the MluI and AatII restriction sites ( underlined ) in their flanking region to facilitate cloning ( Forward: GACGCGTACGCCACCATG GTGAGCAAGGGCG and reverse: GAGACGTCAGTTTACTTGTACAGCTCGTC CATGCC ., The PCR products were gel purified , digested and subsequently ligated into pCR2 . 1Topo , expanded in TOP10 cells ( Invitrogen ) , and inserts excised with MluI and AatII digestion followed by ligation into p ( + ) MVeGFP ( N ) that was digested with the same enzymes to remove the eGFP gene ., Viruses were rescued by transfection of 293-3-46 cells together with pEMC-La followed by overlay on Vero cells as previously described 4 , 31 ., Rescue of the recombinant viruses was inferred from the presence of syncytia and fluorophore detection under ultraviolet light ., The recombinant viruses were expanded by infection of Vero cells ., Cell associated viruses were freed by freeze thawing of the cells three times in liquid nitrogen followed by filtration ., The viral titers ( 50% infectious virus dose , TCID50/ml ) were determined using the Spearman and Karber method as previously described 4 , 31 ., All viruses were stored at -80°C until they were used ., HT1080 spheroids were grown on either Matrigel coated glass bottom 6 well plates or poly-HEMA coated round bottom plates to prevent cell attachment ., For Matrigel coated wells , media , tips , and glass bottom plates were cooled at 4°C and 200μl Matrigel added ., Matrigel was allowed to solidify for 15 minutes ., HT1080 cells were washed 3 times with phosphate buffered saline ( PBS ) and dislodged by trypsin , counted and overlaid at various concentrations into 2ml DMEM with 10%FBS and 2% Matrigel ., The media were freshly replaced every 3–4 days and cells were imaged with a multiphoton microscope ( Olympus ) ., Spheroids were infected using measles viruses encoding various fluorophores in 2 mL Opti-MEM for 2 hours at 37°C at an MOI of 1 . 0 ., HT1080-tdTomato cells were plated in 6 well plates and 24 hours later infected with MVeGFP at an MOI of 1 . 0 Tumor spheroids from the same cell line expressing eCFP were produced as above and infected with MV-tdTomato ., Starting twenty-four hours post infection , the cells were imaged in Z stacks every 30 minutes over the course of a week using an Olympus multiphoton microscope ., Digital images were captured for subsequent analysis ., Average pixel area per cell: The average pixel area of a cell was determined by having two independent scientists counting the number of cells in a given field of view with a fixed color pixel threshold and correlated this with the phase contrast images of the cells ., The total pixel area was divided by the number of cells to determine the average pixel area per cell ., Serial images of in vitro cell growth and virus spread were digitally analyzed using the established cell area parameters and the output was converted back into number of cells ., Voronoi tessellation: Digital images from the in vitro experiments with cells growing in the 2D-plane or in 3D spheroids were analyzed using MatLab to generate Voronoi tessellation analysis of nearest neighbors in 2D and 3D ., We developed a ‘mean field’ mathematical model for tumor growth and viral infection of tumor cells as follows:, duNdt=λNuN ( 1-uN-uC-uV ) -δNuN, duCdt=λCuC ( 1-uN-uC-uV ) -δCuC-λVuCuV, duVdt=-δVuV+λVuCuV, In the model , ui represents the various cellular fractions with N representing normal cells , C cancer cells and V the infected cancer cells ., λi represents the proliferation and δi the death rates of the respective populations ., The model assumes mass action kinetics ., The model was fitted to data from in vitro studies using the purpose built simplex induction hybrid ( SIH ) program 32 and the results of the fits displayed as a heat plot ( see results ) ., The goodness of fit was determined using the chi square method ., The relative parameters for virus infected cells and tumor cells were used to inform the computer simulations ., The model we developed can run simulations of cell populations and infection both in 2 and 3 dimensions and population growth can be either on a lattice structure ( regular ) or a Voronoi network with a variable number of neighboring cells 18 ., The input for the network is described by a list of adjacencies for each node obtained by analysis of the imaging obtained from in vitro experiments ., The simulator itself has no dimensionality and therefore , it can run simulations in two , three or higher dimensions ., The number of neighbors for each node and their location can vary as observed in vitro ., Routines embedded in the program enable modification of the state of the network to simulate the addition of the virus that will spread within the cell population ., Each node in the network is occupied at most by one cell ., Three types of cells are possible: normal cells , cancer cells and infected cancer cells ., A node without a cell is empty ., When a normal or cancer cell proliferates , the new cell generated has to occupy a neighboring empty node while infected cells can target only nodes occupied by a cancer cell since the virus is tumor cell specific 33 and the virus spreads from cell to cell 18 , 21 , 31 ., Each cell type has a node type that it considers as a replication target ., Empty nodes are proliferation targets for normal and cancer cells while a cancer cell node is a proliferation target for infected cells ., The growth and death rates of normal and cancer cells as well as infected cells can be varied as well as the time when the virus is administered ., The virus infection parameters can also be specified ., Event arrivals follow a Poisson process with the time to the next event being exponentially distributed ., Furthermore , the model assumes that cell proliferation and death are independent events and do not depend on the prior state of the network ., To initiate the infection , the program determines the coordinates of all cancer cells and identifies the cancer cell that is closest to the center ., The simulator has 4 infection routines that can be used to specific which individual nodes are infected at the start of the simulation ., ( i ) Random selects cells at random within the tumor population until the pre-specified fraction of cells are infected ., ( ii ) Center selects the cancer cell closest to the center of the tumor and this cell is infected followed by its neighbors and neighbors’ neighbors are infected until the required fraction of infected cells is reached ., ( iii ) Multinode selects the tumor cells closest to the center for infection and the infection spreads from this focus by generating a line in a random direction that passes through the infected node ., Nodes along the line are visited and if they are cancer cells infected with a determined probability ., The process continues until normal cells are reached which cannot be infected ., ( iv ) Perimeter determines the center of the tumor and the distance of all nodes from the center ., Cancer cells occupying the nodes furthest from the center are infected until the pre-specified fraction of cancer cells are infected ., The dynamics of normal and uninfected cancer cells continue with the same parameters once the in | Introduction, Results, Discussion, Materials and models | Tumor therapy with replication competent viruses is an exciting approach to cancer eradication where viruses are engineered to specifically infect , replicate , spread and kill tumor cells ., The outcome of tumor virotherapy is complex due to the variable interactions between the cancer cell and virus populations as well as the immune response ., Oncolytic viruses are highly efficient in killing tumor cells in vitro , especially in a 2D monolayer of tumor cells , their efficiency is significantly lower in a 3D environment , both in vitro and in vivo ., This indicates that the spatial dimension may have a major influence on the dynamics of virus spread ., We study the dynamic behavior of a spatially explicit computational model of tumor and virus interactions using a combination of in vitro 2D and 3D experimental studies to inform the models ., We determine the number of nearest neighbor tumor cells in 2D ( median = 6 ) and 3D tumor spheroids ( median = 16 ) and how this influences virus spread and the outcome of therapy ., The parameter range leading to tumor eradication is small and even harder to achieve in 3D ., The lower efficiency in 3D exists despite the presence of many more adjacent cells in the 3D environment that results in a shorter time to reach equilibrium ., The mean field mathematical models generally used to describe tumor virotherapy appear to provide an overoptimistic view of the outcomes of therapy ., Three dimensional space provides a significant barrier to efficient and complete virus spread within tumors and needs to be explicitly taken into account for virus optimization to achieve the desired outcome of therapy . | Tumor therapy with replicating oncolytic viruses is based on the premise that if the tumor specific virus infects and is amplified by the tumor population and spreads sufficiently within the tumor , it will lead to eradication of the cancer ., The outcome of this approach is an exercise in population dynamics , and , as in ecology , depends on the detailed interactions between the various players involved ., Mathematical models have been used to capture these dynamics , but space is often explicitly excluded from these models ., We combine in vitro experiments studying tumor growth and virus spread in two and three dimensions to inform the development of a spatially explicit computational model of tumor virotherapy and compare the outcome with non-spatial , mean-field models ., Viruses generally spread from cell to cell , and therefore the number of nearest neighbors close to an infected cell is important ., Experimental data show that in three dimensions , the median number of nearest neighbors is 16 compared to 6 in the 2D plane ., However , while simulations in 3D reach equilibrium faster than in 2D , tumor eradication is much less common in 3D than in 2D ., Three dimensional space plays a critical role in the outcome of tumor virotherapy and this additional spatial dimension cannot be ignored in modeling . | cell death, medicine and health sciences, pathology and laboratory medicine, 293t cells, pathogens, cancer treatment, cell processes, biological cultures, microbiology, oncolytic viruses, viral structure, simulation and modeling, oncology, viruses, ht1080 cells, rna viruses, measles virus, research and analysis methods, paramyxoviruses, medical microbiology, microbial pathogens, viral replication, cell lines, cell biology, virology, viral pathogens, biology and life sciences, organisms | null |
journal.pgen.1003093 | 2,012 | Genetic Adaptation Associated with Genome-Doubling in Autotetraploid Arabidopsis arenosa | The duplication of an entire set of chromosomes is a game-changing mutation ., Whole-genome duplication ( WGD ) may create challenges for basic biological functions ., For example , the regulation of gene expression , chromosome segregation , chromatin structure , and the maintenance of cellular homeostasis with altered cell size may be perturbed by duplicating an entire set of chromosomes 1–8 ., That WGD can be challenging to organisms across kingdoms is evidenced by observations of dysfunction in very different contexts , such as reduced fertility observed in many newly formed plant autopolyploids , and mitotic instability in polyploid cancer cells 1 , 5 , 9 ., Despite potential roadblocks , polyploid species are abundant in nature and genome doubling has been implicated in speciation and adaptive radiations 10 ., Polyploids are especially well known among plants , but also occur in a diverse array of animals , including vertebrates 11 ., The short-term consequences of WGD have been extensively studied in both natural and synthetic polyploids , especially in plants ., These studies indicate that chromosome structural changes and rearrangements are common following WGD , as are abnormalities in mitosis and meiosis; in some cases changes in gene expression have also been observed ( e . g . see 1–8 ) ., These observations support the idea that polyploidy can pose challenges to aspects of gene regulation , chromosome organization and chromosome segregation ., A yeast mutant screen indicates that some of these challenges are common across kingdoms ., Genes encoding proteins implicated in the maintenance of genome integrity , including homologous recombination , DNA repair , sister chromatid cohesion and mitotic spindle function were identified as essential genes specifically in tetraploids 12 ., The existence of stable , fertile polyploid species in different kingdoms demonstrates that the challenges that genome-doubled organisms may face at their inception are not insurmountable , and suggests that genome-doubled lineages should experience a period of compensatory genetic adaptation to their genome-doubled state ., In sharp contrast to our understanding of the early transcriptional or genomic responses of organisms to WGD 1–8 , very little is known about what molecular mechanisms might contribute to longer-term stabilization of polyploids or adaptation to a genome-doubled state ., In plants , a single gene important for polyploid stabilization has been molecularly characterized: the homologous pairing suppressor ( Ph1 ) from allohexaploid wheat ., Allopolyploids like wheat have hybrid origins and carry already somewhat divergent sets of chromosomes ., Ph1 enhances meiotic pairing preferences of chromosomes for more similar ( homologous ) chromosomes over less similar ( homeologous ) ones , resulting in bivalent pairing and stable meiosis 13 ., This work provides an important molecular insight into the process of meiotic stabilization in allopolyploids ., However , not all polyploids stabilize meiosis by developing pairing preferences ., Autopolyploids arise from within-species genome duplications and thus carry four homologs of each chromosome 1 , 3–5 , 14 ., Established autopolyploids often have cytologically diploidized meiosis ( forming primarily bivalent associations ) , but show polysomic inheritance at genetic markers , which is possible if the chromosomes lack pairing preferences and partner randomly at meiosis 4 , 5 , 14 ., Thus there must be at least two mechanisms by which polyploids can stabilize meiosis , one that involves enhancing pairing preferences ( as is common in allopolyploids like wheat ) and one that ensures bivalent formation without affecting pairing preference ., The molecular mechanisms that underlie long-term polyploid stabilization and evolution remain largely mysterious ., To help fill this gap , we undertook a population genomic analysis of an established autotetraploid plant , Arabidopsis arenosa ., This species is closely related to two sequenced Arabidopsis diploids: its sister taxa A . lyrata and the model system A . thaliana 15–17 ., Like A . lyrata , A . arenosa is obligately outcrossing , and abundant throughout Europe 16 , 17 ., Tetraploid A . arenosa is cytologically diploidized , with primarily bivalent chromosome associations at meiosis 18 ., We sequenced the genomes of twelve tetraploid A . arenosa individuals from four populations in Germany and Austria and tested for allele frequency patterns suggestive of selective sweeps ., We identified 192 genes in the A . arenosa genome with patterns of polymorphism indicative of recent or ongoing selective sweeps ., Several functional classes represented among these genes are consistent with adaptation to WGD ., We provide candidate genes that will help boost our mechanistic understanding of these processes , while also suggesting new hypotheses ., Similarities of the functional classes we identified with those identified in a yeast mutant study 12 indicate that at least some challenges are shared across kingdoms , and suggests that the functions targeted by selection in A . arenosa are especially critical in tetraploids ., We selected 12 A . arenosa individuals grown from seeds collected at four sites in Austria and Germany ( Figure 1 ) for genome sequencing ., Cytological and flow cytometric analyses demonstrated that A . arenosa populations throughout these regions are tetraploid 19 , 20 ., We confirmed ploidy for at least one individual from each population by flow cytometric analysis of nuclear DNA content ( Figure S1 ) , and performed testcrosses for the remainder ., We aligned DNA sequence data to the publicly available reference genome of A . lyrata 15 ., After filtering for sequence and mapping quality , overall genome coverage per sequenced individual averaged 25× across the eight A . lyrata chromosome scaffolds ( Figure S2 ) ., We focused subsequent analyses on coding regions ., We used a maximum likelihood method to infer tetraploid genotypes for each single nucleotide polymorphism ( SNP ) in each individual ., We generated three-species alignments with consensus sequences from all sites in the A . arenosa sample that had at least 4× coverage per individual , with homologs from both A . thaliana and A . lyrata ., In total 26 , 655 , 179 bp were aligned , representing 20 , 889 homologous genes ., The final dataset contains 3 , 148 , 695 segregating sites ( Table 1 ) ., The average divergence of A . arenosa from A . lyrata per site was 8 . 7×10−4 , 2 . 7×10−4 , and 9 . 6×10−4 for synonymous , non-synonymous , and intronic positions , respectively ., In addition , there were 13 , 634 fixed differences in A . arenosa consensus sequences relative to both A . thaliana and A . lyrata , distributed among 5 , 855 protein-coding genes , 2 , 147 of which contained at least one non-synonymous fixed difference relative to the A . lyrata reference ., Other studies have previously found that polymorphism in A . arenosa is higher than in A . lyrata 21 , 22 ., Consistent with this , we found high levels of segregating variation genome-wide in A . arenosa ( Table 1 ) , and synonymous site diversity approximately double that estimated for diploid A . lyrata 21 ., This is consistent with the prediction that equilibrium genetic variation in an outcrossing autotetraploid population with tetrasomic inheritance should be approximately double that of a diploid population of similar size 23 ., The site frequency spectrum ( SFS ) of non-synonymous SNPs showed a significant skew toward low-frequency mutations compared to the synonymous SFS ( Mann-Whitney U Test p<7×10−8 ) , consistent with widespread purifying selection ( Figure 2 ) ., Importantly , the sequencing error rate we estimated from the data ( 0 . 1–0 . 2%; see Methods ) was an order of magnitude below our estimates of Theta for all classes of sites , and the likelihood function in our genotyping algorithm explicitly accounted for errors ., Thus , sequencing errors were unlikely to have contributed significantly to our estimates of diversity ., Inheritance can vary in tetraploids from disomic to tetrasomic ., Disomic inheritance results when chromosomes have pairing partner preferences ( genes thus behave as duplicates segregating two alleles each ) ., Tetrasomic inheritance occurs in species that lack pairing preferences among the four homologous copies of each chromosome , in which case each locus segregates four alleles ., Whether populations have tetrasomic or disomic inheritance has significant implications for population genetic analyses of tetraploids 3 ., Therefore we investigated historic and ongoing modes of inheritance in A . arenosa by comparing our sequence data to simulated datasets ., We used coalescent simulations to generate expected neutral SFS and genotype frequencies under different historical scenarios and inheritance models ., Our observed data did not differ significantly from simulated SFS for the tetrasomic model , but did differ from both disomic models ( p<0 . 01 Mann-Whitney U test; Table S1 ) ., Similar results were obtained for inferred genotypic classes ( Figure S3; Table S1 ) ., Importantly , we do not observe an excess of duplex ( AAaa ) genotypes , or a high number of SNPs with frequency ∼50% in the data , both of which are expected if the A . arenosa sample had been evolving under disomic inheritance for a significant amount of time ( Figure S3 ) ., These results strongly support the hypothesis that A . arenosa has tetrasomic inheritance ., Together with prior findings that this species has bivalent chromosome associations at meiosis 18 , this places A . arenosa on a growing list of established tetraploids with cytologically , but not genetically diploidized meiosis 14 ., Importantly , tetraploid A . arenosa will display patterns of polymorphism typical of a population of diploids with twice the effective size 23 , 24 ., Therefore , signatures of adaptive evolution are detectable using methods developed for diploids ., We used diploid A . lyrata and A . thaliana reference genomes 15 , 25 to identify 20 , 265 genes that had >80% sequence identity among all three species ., These genes comprise the dataset used in all analyses described below ., The sampled individuals originate from four populations with distinct habitats ( Figure 1 ) ., We tested for population structure or habitat-associated differentiation by pairwise FST comparisons across the genome 26 ., Overall there was low differentiation among populations ., Genome wide pairwise FST at synonymous sites ranged from 0 . 047 to 0 . 063 ( Table S2 ) , which is an order of magnitude lower than average pairwise FST measured between populations of A . lyrata 22 ., This suggests that A . arenosa lacks strong local population differentiation in this geographic region ., During the formation and early establishment of an autotetraploid , alleles that contribute to tetraploid formation or are important for the success of the tetraploid lineage should experience strong selection ., To perform genome-wide tests for selection in tetraploid A . arenosa we identifed genes for which SFS were skewed toward high frequency derived haplotypes 27 and genes in which polymorphism was low ., The two measures were uncorrelated genome-wide ( R2\u200a=\u200a0 . 014 ) and together provide evidence of past selective sweeps ., There were 192 genes that were both within the 5% most skewed SFS and the 5% lowest polymorphism ( Table S3 ) ., In most cases , candidate selected genes were unlinked ., There were only eight instances where genes separated by less than 10 kb both showed signatures of selection ., As a result , almost all potential selective sweep signatures in A . arenosa are sufficiently narrow to identify single candidate genes based on homology to A . thaliana ( www . arabidopsis . org ) ., Several gene ontology categories are over-represented among these genes ( Fishers Exact Test p<0 . 005 for each category ) compared to their representation within the entire genome ., These include functions related to the regulation of basal transcription , epigenetic regulation , sister chromatid cohesion , homologous recombination , DNA repair , cell cycle , cell morphogenesis and cell growth ., The genes representing the most enriched categories are summarized in Table S4 ., We focus below on two general categories in more detail: transcriptional regulation and meiosis ., A “retuning” of basal transcription in response to increased cell size may be important in polyploids for maintaining a balance between expression from additional chromosome copies and altered cell size and/or nuclear membrane surface to volume ratio 1 , 3 ., In this light , it is intriguing that numerous genes showing indications of selection in A . arenosa encode proteins implicated in basal transcription , including the large subunits of two of the core DNA-dependent RNA Polymerases ( Pol ) II and III ( Tables S3 , S4 ) ., The gene encoding the large subunit of Pol II ( NRPB1 ) has numerous high frequency SNP differences in A . arenosa relative to A . lyrata and A . thaliana ., These include two fixed amino acid differences flanking either side of the highly conserved long C-terminal tail ( CTD; Figure 3A ) ., The CTD consists of a series of heptad repeats whose phosphorylation state regulates the activity of the Pol II complex 28 ., In yeast , phosphorylation of the CTD is orchestrated by three cyclin dependent kinases , CDK 7 , 8 and 9 29 ., A homolog of CDK8 , HUA ENHANCER 3 ( HEN3 ) 29 , also shows evidence of having undergone a selective sweep in A . arenosa ., Two other CTD-interactors , PRE-MRNA PROCESSING PROTEIN 40A ( PRP40A ) and GENERAL TRANSCRIPTION FACTOR B1 ( GTB1 ) also show evidence of selective sweeps ( Table S4 ) ., In addition to the CTD-interactors , other genes encoding regulators of Pol II activity or recruitment also show signatures of selection in A . arenosa ( Table S4 ) ., These include genes encoding core transciption factors such as two TRANSCRIPTION FACTOR IIS ( TFIIS ) family genes and TBP-ASSOCIATED FACTOR 5 ( TAF5 ) , which encodes a subunit of TFIID ., TFIID and TFIIS are general transcription factors that associate with Pol II and promote its movement during transcription 28 ., We also find evidence of selection on STRUWWELPETER and CENTER CITY , which encode subunits of RNA Pol II-recruiting mediator complexes 30 , 31 ., Together , the signatures in these genes , as well as epigenetic regulators including genes implicated in RNA-mediated silencing , histone modification and chromatin remodeling ( Table S4 ) , suggest that a global re-tuning of transcription may have been very important in the history of A . arenosa ., Autopolyploids also face an important handicap in meiosis: They are equipped with meiotic machinery inherited from diploid ancestors optimized over evolutionary time to segregate pairs of homologous chromosomes ., That an increase to four homologs presents an obstacle is evident in newly formed tetraploids , which often show high rates of sterility due to failures of chromosome segregation in meiosis 1–5 ., In A . arenosa , eight loci homologous to genes essential for meiosis fit our selective sweep criteria ., These have predicted roles in chromosome synapsis , cohesion and homologous recombination ( Tables S3 , S4 ) ., These genes include SISTER CHROMATID COHESION2 ( SCC2 ) , which encodes an adherin that loads cohesins during meiosis 32 , and one of its substrates , the cohesin subunit STRUCTURAL MAINTENANCE OF CHROMOSOMES 3 ( SMC3 ) 33 , 34 ., SMC5 and SMC6a are also among the eight meiosis-related genes that show signatures of selective sweeps ., These encode proteins that function together in sister chromatid alignment , cohesion , DNA repair and homologous recombination during mitosis 35 ., Recently the SMC5/6 complex was also shown to play an essential role in meiosis 36 ., While sister chromatid cohesion has not previously been specifically discussed as a possible challenge for tetraploid plants , genes involved in sister chromatid cohesion were also shown to be crucial for survival of tetraploid yeast 12 ., We compiled a list of 59 genes annotated in TAIR10 ( www . arabidopsis . org ) as playing a role in meiosis that also had clear homologs in A . lyrata as well as in our A . arenosa sample ( Table S5 ) ., This set of genes showed enrichment for the signatures of positive selection ., Among the 59 genes , 17 ( 29% ) showed a significantly skewed SFS and nine showed low polymorphism in A . arenosa ( Table S5 ) ., Eight of these genes ( 13 . 5% ) were among the 192 that were in both the upper 5% tail of the CLR distribution as well as the lower 5% tail of the π/site distribution ( Table S3 ) , which is a 10-fold enrichment ( Fishers exact test p≪0 . 001 ) ., Six meiosis-related genes with skewed SFS in A . arenosa ( top 5% genome-wide ) are homologous to genes that were also identified as critical for survival in tetraploid yeast 12 ., These are RAD54 , MEIOTIC RECOMBINATION 11 ( MRE11 ) , RECQ4A , TOPOISOMERASE3 ( TOP3 ) , SMC1 and SEPARASE ( ESP ) ( Fishers Exact Test p<0 . 001 ) ., This indicates again that fundamental aspects of chromosome biology present challenges upon genome doubling in very different species and that sister chromatid cohesion , homologous recombination and DNA repair are key shared processes ., In A . arenosa , the chromosome synapsis gene ASYNAPSIS1 ( ASY1 ) 37 has a strongly skewed SFS , low polymorphism and an abundance of high frequency derived SNPs relative to A . lyrata and A . thaliana ( Figure 3B ) ., A high-frequency derived SNP in the tetraploid A . arenosa population sample of ASY1 causes an amino acid change in the conserved HORMA domain ., This alters an ancestral positively charged lysine ( K ) to a negatively charged glutamic acid ( E ) in the derived allele ., We examined other ASY1 sequences reported to date in Genbank and found that this amino acid position is conserved in a wide range of vascular plants ( Figure 4 ) ., Only two other plant species have amino acid changes at this residue ., Both replaced the lysine with a polar uncharged asparagine ( N ) ., We tested whether this polymorphism is differentiated between diploid and tetraploid cytotypes within A . arenosa using a PCR marker ., We genotyped 38 plants from two diploid populations collected from the Carpathian Mountains in Slovakia ( SN and CA in Figure 1B ) ., We found that the derived allele is present , but rare in the diploids ( at a frequency of ∼4% ) ., In sharp contrast , in tetraploid A . arenosa , the derived allele represents 41 of the 48 assayed sequences in our genome resequencing data ( 85% ) and in a wider sample of 75 tetraploids from five additional populations , the derived allele has a frequency ∼90% ., We next asked whether any of the selected genes in A . arenosa are predicted to interact using the AtPIN database 38 ., Forty-six ( ∼24% ) of the 192 candidate selected proteins are known or predicted to interact with at least one other on the list ( ) ., Twelve genes encode products indicated in pairwise interactions ., A set of four forms a small network associated with TARGET OF RAPAMYCIN ( TOR ) and RAPTOR , which regulate a variety of processes associated with cell proliferation 39–41 ., A set of three is associated with a ubiquitin protein ligase , UPL4 42 ( Table S6 ) ., All of the remaining 27 genes are linked in a single network of predicted interactions , many with multiple connections per node ( Figure 5 ) ., The two most connected are NRPB1 ( 9 connections ) and HEN3 ( 6 connections ) ., Many of the additional genes linked to these encode regulators of basal transcription , chromatin structure and cell cycle ., This includes several additional interactors of the CTD tail of NRPB1 , core transcription factor components such as TAF5 28 , 43 , and histone modifiers implicated in the regulation of transcription , including HISTONE ACETYLTRANSFERASE 5 and TAF1 28 , 43 , 44 ( Figure 5 ) ., Shared links through nuclear-cytoplasmic trafficking via EXPORTIN1B connect the network surrounding NRPB1 and HEN3 to a small group of genes involved in regulation of chromatin structure and cohesion in meiosis , including SMC3 and SCC2 ., None of these 27 genes are closely linked in the genome , suggesting that multiple components of this interaction network have been under selection ., Here we report results from a population genomic analysis in autotetraploid A . arenosa ., We show that A . arenosa has high genetic diversity , little population structure , and allele and genotype frequencies consistent with a history of tetrasomic inheritance , in which four alleles segregate at each genomic locus ., We identified 192 genes that exhibit two signatures of selective sweeps: reduced diversity and a SFS skewed toward high frequency derived alleles ., It is important to note that our analysis could not identify loci contributing to polyploid stabilization strictly via adaptive changes in gene expression pattern , unless accompanied by a signature of selection that extended into coding regions ., Identification of such loci would require comparative analysis of gene expression patterns among diploids and tetraploids , and/or analysis of sequence evolution in intergenic regions ., Nevertheless , our focus on adaptive evolution within protein-coding regions allowed identification of putatively selected genes that have clear orthologs in A . thaliana , and for which functional information is therefore available ., This work suggests candidate genes and processes that may have been important for compensatory adaptation of A . arenosa to its genome-doubled state ., The functional annotations of the A . thaliana homologs of these genes point to the modulation of fundamental biological processes , including the regulation of core transcription , epigenetic regulation , DNA repair , cell division and morphogenesis , chromosome synapsis and cohesion , homologous recombination , and chromosome segregation ., Several of these categories represent functions that have been previously demonstrated or hypothesized to be problematic for neo-polyploids , but for which the mechanisms of longer-term stabilization have not been studied 1–5 ., Several functional classes represented among candidate selected genes in A . arenosa , particularly chromosome cohesion , segregation and repair , show considerable overlap with genes necessary for survival specifically in polyploid yeast 12 ., Moreover , six genes with SFS indicative of selection are the closest ( or only ) Arabidopsis homologs of the genes identified in the yeast screen ., These are RAD54 , MRE11 , SMC1 , TOP3 , RECQ4A , and ESP ., That these genes are truly fundamental in genome maintenance is also underlined by the fact that all of them have been implicated in numerous human diseases associated with genome instability , including cancer , Ataxia-Telangiectasia-like disorders , Bloom Syndrome and others , e . g . 45–50 ., This indicates that at least some of the fundamental challenges to the maintenance of genome integrity that organisms face after genome perturbations , including whole genome duplication , are broadly shared across kingdoms ., It also provides corroborative evidence that at least some of the signatures of selection in A . arenosa are indeed attributable to adaptation to a doubled genome ., There have been numerous studies of gene expression in response to whole genome duplication ( see e . g . 2 , 6–8 ) ., Though most have focused on allopolyploids , several have directly compared gene expression in diploids and their autotetraploid derivates ( e . g . 51–57 ) ., In most cases , there is little or no overlap with the functional classes or specific genes identified in expression studies and those we identified in our study ., This suggests that the genes and functional classes involved in short-term responses to genome duplication are largely distinct from those that may be under selection during longer-term polyploid evolution ., There are some exceptions: In Paspalum notatum , gene expression changes in new polyploids occur in some of the same gene classes as those we identified here , including transcription , DNA repair and chromatin structure regulation 51 ., Thus in some cases early gene expression responses do occur in genes or functional classes that may be under selection in longer-term polyploid evolution , suggesting that some of the selection acting on polyploid genomes may be a compensatory response to early shifts in gene expression ., One of the genes we identified as putatively under selection in A . arenosa , RAD54 , which is involved in DNA repair as well as homologous recombination 58 , 59 , has also been reported to be upregulated in response to genome duplication in autotetraploid A . thaliana 39 ( though see 54 ) ., Another feature of the putatively selected genes in A . arenosa is that many are known or predicted to interact ., This is especially true of genes implicated in the regulation of basal transcription ., That multiple functionally connected , but unlinked genes may have experienced selective sweeps suggests that these loci either contribute incrementally to fitness through modifications of a common process or have been selected together as a functional module ., Entire networks can experience selection effectively as units if epistatic interactions are synergistic and alter the selective environment for mutations at functionally related loci , allowing a larger coordinated response to selection 60 , 61 ., Indeed , findings in other species support the idea that genetic modules encoding networks of interacting proteins can in some circumstances respond to selection as units 60–65 ., Whether interaction effects have driven selection on a functional module surrounding basal transcription in A . arenosa , or whether the polymorphisms contribute additively to a selected phenotype merits further exploration ., Interestingly , in yeast it has also been noted that genes important in tetraploid survival are predicted to interact extensively 12 , suggesting that this , too , may be a shared feature of polyploids across kingdoms ., Processes such as core transcription are interlinked with other cellular functions ., For some genes we have identified it will be possible to clearly hypothesize what the selected function is ., However , for other genes , it is less clear what function selection has acted to modulate , or if there are pleiotropic effects ., For example , GTB1 , which shows evidence of selection in A . arenosa , binds the C-terminal extension of Pol II and participates in regulation of Pol II processivity 28 ., Thus it is reasonable to suppose it might have been under selection for its contribution to the regulation of basal transcription ., However , GTB1 has also been predicted to interact with ARGONAUTE ( AGO ) proteins which function in the processing of small RNAs 66 ., AGO1 also shows evidence of a selective sweep in A . arenosa ( Tables S3 , S4 ) , and AGO4 also shows evidence of adaptive protein evolution ( not shown ) ., This however , may not be due to polyploidy per se , since AGO genes show evidence of selective sweeps in diploid species as well ., For example , successive selective sweeps in an Argonaute gene in Drosophila species have been suggested to be associated with host-pathogen co-evolution 67 ., The picture may be even more complex , since small RNAs have also been implicated in DNA double-strand break repair 68 , 69 , meiotic chromosome pairing 70 , and mitotic and meiotic chromosome structure and segregation 71–73 ., Indeed , AGOs have themselves been directly implicated in maintaining chromatin silencing during meiosis 71 , 73 ., These are fundamental genome maintenance processes strongly implicated in polyploid stabilization ., Thus the true causes of selection on genes like GTB1 or AGO1 that are implicated in multiple distinct but interlocked processes provide extensive opportunities for follow-up studies to unravel the complexities of selection acting on interconnected pleiotropic genes , more than one of which may be under selection for different reasons ., For the chromosome synapsis gene ASY1 , we confirmed differentiation among A . arenosa cytotypes of an amino acid substitution at a conserved position ., ASY1 is related to the Hop1 gene in yeast , which plays important roles in the assembly of the synaptonemal complex and the regulation of homologous recombination 74 ., In plants , these functions are conserved , e . g . 37 , 75 ., Synapsis is a process that has been hypothesized to play a role in meiotic stabilization of tetraploids 1 , 4 , and ASY1 itself has been functionally implicated in polyploid meiosis ., Expression of wheat TaASY1 is affected by Ph1 , and transgenic downregulation of TaASY1 results in reduced synapsis but strengthened associations of homeologs at metaphase I 76 ., If the derived ASY1 allele in A . arenosa was important in polyploid evolution , as the signature of selection suggests , this implies that this gene may play a role in promoting meiotic stability in both allo- and autopolyploids ., The presence of the derived ASY1 allele at low frequency in the diploid gene pool suggests that standing variation for ASY1 , rather than de novo mutation , may have been important for a rapid response to selection during tetraploid stabilization ., This is consistent with findings in other species that genetic variation in diploids can affect meiotic stability after artificial genome doubling , e . g . 77 ., Overall our data indicate that selection has acted on numerous genes in the tetraploid A . arenosa genome , providing specific candidate genes and mutations for mechanistic follow-up work ., Some of this selection may have been on standing genetic variation in diploid A . arenosa that contributes to polyploid formation , for example by promoting unreduced ( diploid ) gamete formation ., However , many of these selected alleles are likely to have been involved in the stabilization of fundamental biological processes after whole genome duplication ., Our analysis implicates several fundamental processes and functions in adaptation to polyploidy , both supporting previous hypotheses about polyploid stabilization , such as modulation of meiosis , and suggesting new ones , such as involvement of a network associated with the regulation of core transcription ., Finally , our analysis reveals an overlap of putatively selected genes and functions in A . arenosa with genes identified as essential in tetraploid yeast 12 and implicated in disease-associated failures of genome maintenance in humans ., This suggests that key challenges faced by polyploids are shared across kingdoms and understanding how natural selection can circumvent these problems in a variety of species will provide important insights ., Plants were grown directly from seeds collected from wild populations in the summers of 2009 and 2010 ., Seeds were collected in late June 2009 from the railway station in Triberg ( TBG ) in the Black Forest of southwestern Germany , and from a limestone outcrop near the Upfinger Steige ( US ) , between Upfingen and Bad Urach in the Swabian Alb region of southwestern Germany ., Seeds were collected in June 2010 from Kasparstein castle , in southern Austria ( KA ) and Berchtesgaden railway station ( BGS ) in southeastern Germany ., Seeds were surface sterilized with 70% ethanol/0 . 05% Triton X-100 , and then stratified at 4°C in the dark for six to eight days on 1/2×MS plates with 8% agar ., Seeds were germinated in a tissue culture incubator at 16°C with 16 hour long days , and then transferred to soil ( 50% Sunshine Mix #4/50% fine vermiculite ) and grown in a growth chamber with 16-hour long-day light cycles ., Ploidy was verified by flow cytometry on at least one individual per population , and by testcrosses to known diploid and tetraploid individuals ( the Streçno castle | Introduction, Results, Discussion, Materials and Methods | Genome duplication , which results in polyploidy , is disruptive to fundamental biological processes ., Genome duplications occur spontaneously in a range of taxa and problems such as sterility , aneuploidy , and gene expression aberrations are common in newly formed polyploids ., In mammals , genome duplication is associated with cancer and spontaneous abortion of embryos ., Nevertheless , stable polyploid species occur in both plants and animals ., Understanding how natural selection enabled these species to overcome early challenges can provide important insights into the mechanisms by which core cellular functions can adapt to perturbations of the genomic environment ., Arabidopsis arenosa includes stable tetraploid populations and is related to well-characterized diploids A . lyrata and A . thaliana ., It thus provides a rare opportunity to leverage genomic tools to investigate the genetic basis of polyploid stabilization ., We sequenced the genomes of twelve A . arenosa individuals and found signatures suggestive of recent and ongoing selective sweeps throughout the genome ., Many of these are at genes implicated in genome maintenance functions , including chromosome cohesion and segregation , DNA repair , homologous recombination , transcriptional regulation , and chromatin structure ., Numerous encoded proteins are predicted to interact with one another ., For a critical meiosis gene , ASYNAPSIS1 , we identified a non-synonymous mutation that is highly differentiated by cytotype , but present as a rare variant in diploid A . arenosa , indicating selection may have acted on standing variation already present in the diploid ., Several genes we identified that are implicated in sister chromatid cohesion and segregation are homologous to genes identified in a yeast mutant screen as necessary for survival of polyploid cells , and also implicated in genome instability in human diseases including cancer ., This points to commonalities across kingdoms and supports the hypothesis that selection has acted on genes controlling genome integrity in A . arenosa as an adaptive response to genome doubling . | Duplication of an entire set of chromosomes is a dramatic mutation disruptive to core cellular functions ., Genome duplication and the genomic instability that generally follows can cause problems with fertility and viability , and in mammals is associated with cancer and spontaneous abortion ., Yet , established polyploids occur naturally in both plants and animals ., How do these organisms overcome these early problems and ultimately stabilize ?, The genetic basis of the adaptive response to polyploidy has remained almost completely unknown ., We took advantage of modern genomic approaches to gain insight into this using a stable polyploid plant , Arabidopsis arenosa ., We found evidence of selection in genes that control core genome maintenance processes ., These overlap with genes or functions shown in yeast to be necessary for survival of polyploid cells and in humans implicated in cancer ., Our results identify genes controlling core genome maintenance functions that may have undergone compensatory adaptation after genome doubling . | plant science, plant biology, genetics, biology, genomics, evolutionary biology, genetics and genomics | null |
journal.pcbi.1000798 | 2,010 | Expansion of the Protein Repertoire in Newly Explored Environments: Human Gut Microbiome Specific Protein Families | Every ecological niche presents specific challenges that face the population of organisms that inhabit them ., When analyzing species that thrive in any particular environment , we can expect that certain key functional characteristics would correlate with success and differentiate those species from others that fail in colonizing that environment ., This is especially obvious for microbes , and detailed analysis of almost every sequenced microbial genome provides examples of adaptation , mostly in terms of the presence of genes that code for specific functions required for that microbe to succeed in a given environment ., However , studying microbes one genome at a time does not generally provide enough data and meaningful statistics to explore fully the relationships between individual gene families and their environments ., This has now changed with the advent of metagenomics , which can investigate entire microbial communities associated with single environments ., In metagenomics shotgun sequencing , which identifies genes present in a given environment , the associations between gene families and specific environments can be analyzed directly ., All such studies carried out so far have identified unique distributions of functional classes of protein families that are strongly correlated with the specific features of the given environment , be it presence of specific nutrients , acidity , high temperature , etc ., For instance , Gill et al . have shown that the human gut microbiome is enriched in proteins associated with amino acid and vitamin production 1 ., Another study has confirmed these observations and found additional functional groups of proteins overrepresented in the human gut , such as for carbohydrate and lipid transport and metabolism 2 ., Similar observations have been made during analysis of the genomes of several human gut–associated microbes , such as Bacteroides fragilis 3 and Bacteriodes thetaiotaomicron 4 ., However , these analyses have focused exclusively on already recognized and functionally characterized protein families—all of which were previously identified and characterized by resources such as PFAM 5 , COG 6 or Interpro 7 ., As a result , two important groups of protein families were not included in such analyses; namely , families already discovered but not yet characterized , and novel families specific to a newly studied environment but rarely or never found in microbes or in the environments previously studied ., ., Both sets represent a possible wealth of information about the processes necessary for microbes to survive in the human gut ., Their importance for further study was exemplified by a recent metaproteomics study 8 , in which almost 20% of all recognized proteins , including several of the most abundant ones , were classified as “hypothetical proteins” and did not belong to well-characterized protein families ., Thousands of such environment-specific protein families have also been identified in other environments , such as the ocean 9 , 10 ., In this study , we address this important issue by an ab initio search for protein families in datasets that represent the environment we are studying , and a subsequent abundance/conservation analysis of all protein families , including new examples and those not covered by any functional category ., An important issue in interpreting results of such large-scale studies involves widespread inconsistencies in use of the term “protein family” ., While the general definition of a protein family as a group of proteins that evolved from a common ancestor seems very clear , in practical applications , this term can mean anything from a group of very close homologs to an extensive , very divergent group of proteins that shared a common ancestor billions of years ago , but have now evolved into a multitude of sub-families with different functions ., Automated procedures for indentifying protein families typically indentify closely related families composed of highly similar proteins , which , upon further analysis , could be included in an already known family or combined with others to form a larger family ., Therefore , estimates of the numbers of new protein families provided in large-scale automated project are typically too high ., In the context of this paper , we address this problem with detailed analysis of some of the families found in the automated analysis ., The human gut is a very specific environment , rich in diverse nutrients , but also full of challenges for its microbial inhabitants ., Because of its richness , the microbes inhabiting human gut form one of the densest microbial communities on Earth , reaching 1011 cells per gram 11 ., Species that inhabit that environment have to be able to extract energy from diverse and rapidly changing sources , reflecting the diverse human diet that can vary significantly in content and quantity over time in both daily and seasonal cycles ., Species forming the human gut microbiome also need to survive encounters with the human immune system and to coexist with other microbes ., Sets of specific microbial proteins must carry out the essential tasks of recognizing new nutrients , transporting them into the cytosol and metabolizing them , neutralizing or suppressing human immunity , and signaling to other bacteria and host cells ., The presence of genes coding for such proteins in a genome would provide a distinct competitive advantage to a human gut symbiont or commensal microbe ., In this paper , we seek to identify such environmentally specific protein families , focusing on the human gut as a target environment ., Because of the obvious importance of this environment for human health , several groups have performed large-scale , random , shotgun sequencing experiments on representative samples providing a direct view of the gene content of this environment 1–2 ., At the same time , a major sequencing effort , the NIH Human Microbiome Project ( HMP ) , is specifically targeting genomes of human gut microbes 12 as identified , for instance , by 16S rRNA studies ., Genomic sequencing provides information for individual species but , with a coordinated effort to sequence the genomes of hundreds of microbes from a single environment , the resultant data can also be translated into an overall gene content ., Thus , two sets of independent data can be obtained that describe the gene content of the same environment ., Both approaches have their advantages and shortcomings: metagenomic shotgun sequencing provides a relatively unbiased , but small sample of genes that can be found in a given environment ., On the other hand , genomic sequencing provides a full set of proteins from a genome , but its success depends on our ability to culture specific species and , thus , might leave large groups of microbes without any representation ., Arguably , both of these approaches provide only a very crude approximation of the actual gene content of an environment ., However , as we will show , data from both methods present a surprisingly coherent view of the gene content of the human gut , at least on the level of protein families , which encourages us that the data are robust enough for a survey analysis , such as presented here ., We hypothesize that genes coding for proteins that are necessary and beneficial for survival of microbes in the human gut environment will be found abundantly both in the genomes of the species found in that environment and in metagenomic data sampling of the same environment ., Hence , we can verify observations made on one set of data by using the other as a reference ., At the same time , since an extensive study of the human gut environment and its microbiome was only started very recently , protein family databases and annotation resources , which typically work with significant time lag in recognizing novel protein families , simply havent had enough time to include data for new families found only in this environment ., In this manuscript , by automated clustering in metagenomics samples from the human gut we identify about 1 , 800 novel protein families and curate and analyze in detail about 180 of them ., Some of these families have been confirmed and characterized by structural studies , since the PSI large-scale Structural Genomics Centers have used a preliminary version of our analysis to select some of the most abundant protein families in the human gut as targets for structural determination 13 ., We also present a comprehensive analysis of the distribution of protein families in the human gut environment , including both those previously known , as well as the new families identified in this study ., While many of the ORFs identified in metagenomics shotgun sequencing projects can be classified into already known and defined protein families , many—often over 50% ( see Figure 1 ) —cannot ., About 6% are singletons ( sometimes called ORFans ) 14 , i . e . , proteins that dont have any homologs in current protein databases ., Nevertheless , most of the unclassified proteins do form families of varying sizes and such new families may play very important roles in specific environments , but , by default , were omitted from all previous analyses ., In our study , we aim to get a complete picture of protein family distributions in the new environment ., To this end , we optimized a previously introduced 10 clustering technique ( see the Methods section for details ) and used it on the set of over 600 , 000 ORFs from two large human gut metagenomics projects 1 , 2 ., We identified almost 1 , 800 protein families fulfilling our size criteria , of which 926 could be matched to PfamB , the uncurated section of the PFAM database , while the other 835 were found de novo in the metagenomic data ., We now describe results of various types of analyses applied to these data , including manual curation and experimental verification ., In Figure 2 , we compare the distribution of sizes of the new protein families identified here to that of PfamA families that were represented in the metagenomics samples , as sorted by the approximate number of members present in the metagenomic dataset ., Both sets have similar size distributions , with PFAM families being somewhat larger ., It is interesting to note that only about 2 , 300 ( from over 10 , 000 ) PFAM families pass the size threshold ( i . e . have ten or more members in the gut-related genomes and metagenomic samples ) to be included in this histogram ., In the next step , we study coverage of the metagenomics datasets , as well as both reference genome sets ( HGR and HGU ) by the expanded set of families that includes characterized domains from the PFAM database ( PfamA ) 5 , as well as the families newly found in this work ( see the previous section ) ., The level of coverage of HGR and HGU genomes by PfamA families is 51% and 52% , respectively ., However , the level of coverage drops dramatically to 39% for metagenomic samples ., Clearly , while both HGR and metagenomics samples represent the same environment , the metagenomic datasets contain a larger portion of previously uncharacterized genes , most likely from genomes of as-yet-uncharacterized species ., Adding new families identified in this work increases coverage of the metagenomic dataset by approximately 8 . 9% and increases coverage of reference genome sets by 8 . 4% and 3 . 5% for HGR and HGU genomes , respectively ., However , in all sets , a large percentage , 40–45% of all ORFs , still cannot be assigned to either an already known or a new family ., This group of ORFs can be broadly divided into two groups: a majority ( ∼88% of the unclassified proteins , i . e . 45% of the total ) are proteins that form small families ( <10 members ) , which were not included in the analysis because of the size thresholds used in this work ., These “microfamilies” may be an important source of information , but the computational complexity of applying detailed analysis to each of these possible families must await future research ., It is very likely that these microfamilies will expand to full-sized families with the addition of new metagenomics datasets , or will be found to be included in already defined families as the sensitivity of their profile description improves with addition of further homologs ., The remaining 12% of the unclassified proteins , i . e . 6% of the total , have no BLAST matches internal to the human gut metagenomics samples and , thus , cannot be grouped into clusters of metagenomic sequences ., Truly unique protein sequences may be specific to uncharacterized , rare organisms , but it is also possible that they represent failures of sequencing technologies , bad ORF calls , etc ., The validity of ORF calls can be monitored; in the analysis of the GOS metagenomics samples , the number of similar sequences has been shown to be strongly correlated with the validity of an ORF call 9 , other criteria can be used as well 10 ., Once a complete set of protein families is identified , the next step is to determine the extent to which these families are specific to our target environment ( the human distal gut ) ., To this end , we calculate an “essentiality coefficient” ( Es ) for every family ( see Methods section for a formal definition of Es , as well as for definitions of other measures of environmental specificity of protein families ) ., An essentiality coefficient equal to 1 means that at least one member of a given family was found in the genome of each of the human gut–associated microbes , but no members were found in any of the reference set of genomes—thus , this family is considered as essential for the gut environment ., An Es close to 0 indicates lack of preference , and an Es close to −1 indicates an “anti-preference . ”, Figure 3 presents the distribution of essentiality coefficients for protein families from the PfamA database and for new protein families found in this work ., PfamA protein families show an almost symmetric distribution of preference and anti-preference for genomes of human gut–related microbes ., At the same time , the new families found in this work are very specific for human gut–related microbes ., This outcome is , of course , expected as these families were identified by clustering from the metagenomics datasets with the aim of identifying environment-specific families ., Interestingly , some families that were not specific for the human gut environment were notably found by clustering metagenomics ORFs ( lower-right region of graphs in Figure 3b ) ., These are the protein families , found by clustering the metagenomic datasets that turn out to be more frequently conserved in random genomes not connected to that environment ., One example is the family HGC00614 , composed of 18 proteins found in the metagenomic data ., Upon constructing an appropriate HMM , we found that this family is a likely new family in the PFAM PLP_aminotran ( CL0061 ) clan , with many homologs across multiple species ., It is also worth noting , that some families found in metagenomic data have not been found in any fully sequenced genomes of microbes from the same environment , clearly showing that complete genome sequencing still hasnt fully explored the diversity of genes present in this environment ., Several different measures can be proposed to compare distributions of a protein family between two datasets ., For instance , the comparative overrepresentation ( Ov ) in a specific dataset details the number of members a family has in one dataset as compared to another reference set ., Another metric is the expansion ( Ex ) of a protein family when the relative counts of protein families are compared but , rather than normalizing by the total number of proteins in the genomic set , counts are normalized by the number of genomes that contain at least one match ., This metric highlights families that may not have the largest counts , but when found , have multiple copies in the same genome ., Yet another measure is the essentiality coefficient ( Es ) used above in Figure 3 , which compares the percentage of genomes in each group that contain at least one member of a family ., So far , we have only used the latter specificity measure ( Es ) ., In the following analysis , we will use and compare all three measures as each captures some of the intuitive notion of specificity ., Each measure corresponds to a different biological mechanism of “specificity” ., Having multiple paralogs of proteins from families with high Es , but low Ov or Ex , clearly does not provide an advantage to a microbe , therefore protein families that score well with Es likely perform highly specific , but essential functions ., On the other hand , large number of members of overrepresented or expanded families provides such advantage , but may represent only one of many possibilities of solving a given problem; hence , they are not present in all microbes in a given environment ., For instance , metabolic enzymes would likely belong to the latter category , while defense and host signaling proteins would likely belong to the former ., As discussed extensively in the papers that study the human gut microbiome directly through metagenomics sequencing 1 , 2 or indirectly through genome sequencing of specific microbes representative of this environment 3 , 4 , certain protein families involved in specific types of function were observed to be strongly expanded in the human gut microbiome compared to families found in “average” microbes ., However , these studies did not cover protein families of unknown function , and focused only on one measure of specificity that is related to our overrepresentation measure Ov , in Table 1 ., In our analyses , we use and compare three different specificity measures: Ov in Table 1 , Ex in Table 2 , and Es in Table 3 ., Our research also focused on complete family coverage , including families of known and unknown function , as well as new families specific for the gut environment ., Novel families were ranked by the three different ranking methods , with the top 10 hits listed; Ov in Table 4 , Ex in Table 5 , and Es in Table 6 ., The full list of 180 annotated protein families is detailed in Table S1 , in the supplemental material ., Domains of unknown function ( DUF ) dominate the overrepresented group with four such families in the top 10 when sorted by overrepresentation ( Ov ) , but the DUFs are also present in other forms of ranking ., The presence of so many weakly characterized protein families in all specificity categories clearly illustrates the inadequacy of our knowledge about this important environment ., Similarly , all previous analyses focused mostly on metabolic proteins and interpreted the specificity of the human gut environment predominantly in the view of its unique metabolic content ., We show here that protein families involved in regulatory and DNA exchange functions are also strongly present among the most overrepresented families ., It is possible and , indeed , very likely that , by using more sensitive sequence analysis tools , many of the families identified here would be eventually grouped into larger entities , such as clans in PFAM 5 ( or superfamilies in other protein classification systems ) , that represent more distant evolutionary relationships ., However , for the purpose of this analysis , we will focus on the family level as practically defined by major community resources , such as PFAM 5 or Interpro 7 ., Upon further analysis of the families identified in an automated , ab initio clustering of protein sets we realized that many may not fit such definitions ., For instance , proteins that form distant branches of already existing families may form well-defined clusters in the automated analysis , but careful optimization of HMMs that define old versus new families would be necessary to decide if they would form a new family or if they could be included in the old family by readjusting its definition ., For instance , we found several potential families that belong to the SusC and SusD mega-families ., SusC and SusD are part of the sus ( starch utilization system ) operon in B . thetaiotamicron , an archetype of polysaccharide utilization loci found in multiple copies in all Bacteroides and related species 15 , 16 ., Both families are extremely divergent; only a small number of their members are covered by PFAM HMMs that define Ton_B–like and SusD families , respectively ., The complex evolution of the SusD protein family is the subject of a separate paper 17 ., Families that define new domains in proteins with already recognized PFAM domains form the second group ., Again , without detailed analysis , it would be difficult to decide if such families should be defined as new or covered by readjustment of the boundaries of already defined families ., We used several filters to identify and remove the group of new families that would be most likely to overlap with already existing PFAM families ( see Methods ) , undoubtedly eliminating some genuine , novel families ., Next , we analyzed the remaining ones by hand to identify those that are most likely to conform to the “PFAM standard” , i . e . , families that represent functional domains that do not overlap with protein families described in the PFAM database ., At this point , the hand-curated set of PFAM-quality families exceeds 180 and would undoubtedly expand further as the curation and analysis continue ., We provide the current list of curated families as a Table S3 in the Supplemental Materials ., Tables 4–6 present the top families from this group in three different “specificity” categories ., ( An analogous table for the Pfam families was presented in the previous section . ), As mentioned earlier , a preliminary version of this analysis was used to select structure determination targets for the four large NIH Protein Structure Initiative production centers in two “target drafts” 13 in mid- and late 2008 ., As of May 2009 , representatives of almost 800 of the 1 , 761 protein families identified here had been successfully expressed and purified in vitro , supporting the conjecture that the new families represent real proteins and not “shadow ORFs” or other sequencing artifacts ., The last column in Tables 4–6 provides information about the status of the representative of a given family that is most advanced in the PSI production pipeline ., The structures of representatives of several protein families described here have been successfully solved , and their coordinates deposited into the Protein Data Base ., For instance , Thermotoga maritima proteins TM1486 ( 1VPV ) and TM841 ( 1MGP ) represent DegV ( PF02645 ) , a large family of proteins , shown by structure analysis to be involved in fatty acid binding ., The Lactobacillus acidophilus NCFM protein LBA1001 , PDB entry 3EDO , incorrectly described in the literature as a TRP repressor , has 142 homologs in metagenomic datasets , and at genomic levels of conservation goes from 12% of species in the HGU sample to 84% in the HGR set ., A third example of a protein prevalent in the human gut environment is the protein family represented by PDB entry 2PC1 ., This acetyltransferase/GNAT family protein has 47 metagenomic homologs and is present in 73% of HGR species , while it is rare in the HGU list ( 5% of species ) ., Other protein families determined to be important to the human gut environment and found independently by this study include PfamA family PF08842 ( DUF1812 ) , represented by PDB entry 3GF8 ., Only 3 homologs are found in genomes of free-living bacteria ( HGU set ) , compared to 47 in the of human gut-related microbes ., The proteins matching this family in the HGU genomes were found to be hypothetical proteins in Porphyromonas gingivitis , an human oral pathogen , which was included in the HGU set because of its specific definition ( see Methods ) , but should probably be reclassified to the HGR set ., The Protein Structure Initiative has also solved several proteins from family PB002962 ( PDB entries 3DB7 and 3DUE ) ., This family was found in eight of the thirteen of the metagenomic samples , with a total of 34 homologs and present in only 1 . 6% of HGU genomes as compared to 21 . 5% of HGR genomes ., The gastrointestinal tract is extremely important for overall human health ., Numerous diseases , from digestive disorders and immune diseases to numerous types of cancer , notably involve the GI system ., At the same time , the human GI system , and especially , the distal gut , is a surprisingly complex and little understood environment , inhabited by a complicated bacterial community that carry enzymes for processing byproducts and downstream products of metabolism in the stomach and proximal gut ., Rich in nutrients , the gut harbors one of the densest microbial populations known ., These microbes and their metabolism play a critical role both in health and in diseases of the GI system ., While the culturable microbes living in the human gut have been studied for decades ( for instance , E . coli ) , the development of new technologies and the concept of metagenomics provided a decisive , paradigm-changing shift in studies of this environment , in which the diversity and the communal nature of the human gut microbiome could be uncovered ., We thus now have access to several synergistic , but independent , lines of investigation into the surprisingly unknown world of microbes inhabiting human cavities ., Here , we investigated what types and number of novel , previously uncharacterized , protein families can be found in this environment ., In our analysis , we have shown that many protein families , most completely uncharacterized , show strong specificity for this environment ., Undoubtedly , the functions of these proteins play an important role in the maintenance and operation of the human gut microbiome ., Approximate function predictions based on distant homology recognition identified many proteins that are involved not only in metabolism , but also in signaling , regulation , and phage activity , and are obviously very important in such dense bacterial communities ., We have identified not only a few thousand known protein families as strongly overrepresented in the human gut environment , but also , many potentially new protein families ., Many of these assignments have now been confirmed by structural determination by the PSI centers , and many of their functions have been predicted due to fold recognition techniques ., However , many yet uncharacterized or completely novel families have been shown to be specific to the human gut environment ., This observation , in turn , suggests that many unknown and uncharacterized processes are yet to be discovered in this environment ., Apart from these interesting insights about this specific environment , our observations suggest this approach is applicable to analyzing other environments ., Historically , genomic analysis has focused on individual species , but it is important to remember that an organism does not exist in a vacuum ., Organisms evolved their specific traits in the context of their environment ., By sampling the gene pools in a given environment , we can learn about the protein families that are key for survival in those environments ., The methods presented here should aid in organizing and streamlining such analyses ., Our analysis is derived from several different sources: metagenomic sequencing , 16S rRNA sampling , fully sequenced cultured genomes from NCBI , and draft genomes published by the Human Gut Microbiome Initiative ( HGMI ) 12 ., Each of these data sources is publicly available ., We used a human gut metagenomic dataset derived from the Kurokawa 2 study ., This dataset contains 350 , 000 assembled contigs from 13 individuals , both male and female , with ages ranging from 3 months to 45 years ., Although these genomic data come from 13 separate individuals , we have treated them as a single set to improve the odds of finding human gut–related proteins ., Preparation of the sequence metagenomic data begins with Open Reading Frame ( ORF ) prediction done by Metagene 18 ., Metagene analysis produced a set of 665 , 559 ORFs ., From this initial set , incomplete ORFs that ran off the edge of the sequence read were removed ., A total of 303 , 314 complete ORFs were left ., This set was then used to identify protein families ( see the section Clustering and identification of uncharacterized and new families ) ., The HGMI sequenced genomes provide an ideal reference set of human gut–related microbial genomes ., In addition to the human gut–related reference genomes ( HGR ) , we also needed a set of genomes not related to the human gut environment for comparative analysis ., The set of selected fully sequenced genomes was derived from the collection of bacterial genomes available from NCBI ., As of July 2008 , this library included 765 bacterial genomes ., We utilized data from 16S rRNA sampling to eliminate genomes linked to the human gut environment by targeted metagenomic sampling ., The 16S rRNA data was derived from two sources: Greengenes 19 and David Relmanns published human gut sample 16S RNA set 20 ., Using data available in the Green Genes , we searched for 16S rRNA sequences associated with keywords “human” and one of the following: “fecal , ” “faecal , ” “colon , ” “intestine , ” “stool , ” “rectum , ” “cecum , ” “feces , ” “intestinal , ” “colitis , ” “stomach , ” or “gut . ”, This search produced a set of 38 , 839 16S rRNA sequences ., This set was added to the 11 , 831 sequences from the Relman dataset ., Using a broad Operational Taxonomic Unit ( OTU ) of 90% sequence identity , we ran BLAST against the set of NCBI bacterial genomes and selected 493 species not linked to the human gut microbiome ( i . e . those which did not match any 16S RNA sequences from species related to human gut ) ., We refer to this latter set as the Human Gut–Unrelated ( HGU ) set ., To create the set of Human Gut–Related genomes , we started with 45 genomes from the HGMI project , each currently in the draft stage ., In addition to that base set , we added 20 finalized NCBI bacterial genomes tagged with matching 16S rRNA sequences that were manually confirmed by examining NCBI genome project annotations ., This provided us with a set of 65 genomes referred to as the Human Gut–Related ( HGR ) set ., Detailed information about both sets is available in the Table S2 in Supplemental Materials ., One of the important aspects of analysis of metagenomic sequences is the identification of novel sequences ., These sequences with no known homolog in existing sequence databases are referred to as orphan sequences ., In the study by Kurokawa et al . 2 , orphan analysis was carried out by taking over 600 , 000 predicted ORFs and looking for genes previously seen with BlastP with a threshold of 1 . 0e-5 against a custom , extended , non-redundant ( NR ) , sequence database ., Of the original set , 162 , 647 genes were determined to be orphan sequences ., This set was combined with 503 , 115 other orphan genes from other metagenomic environments ., The total set of orphans was calculated by producing an all-to-all BlastP 21 comparison ., Connections were drawn between proteins with alignments that had a Blast score of 60 or greater and were marked as a match and the connection graph was then clustered with TribeMCL 22 ., The main difference between our analysis and that of Kurokawa et al . is that they augment their human gut metagenomic ORF orphan set with orphans from other metagenomic environments ., We believe the main benefit of metagenomic sequencing is that protein families related to specific environments can be targeted ., These environmentally specific signals may have been lost by adding sequences from other environments ., In our study focused on identification of novel and uncharacterized protein families we used the procedure described below ( outline of the procedure is also give in a separate table ( T1 ) in the supplement materials ) ., We used the set of metagenomi | Introduction, Results, Discussion, Methods | The microbes that inhabit particular environments must be able to perform molecular functions that provide them with a competitive advantage to thrive in those environments ., As most molecular functions are performed by proteins and are conserved between related proteins , we can expect that organisms successful in a given environmental niche would contain protein families that are specific for functions that are important in that environment ., For instance , the human gut is rich in polysaccharides from the diet or secreted by the host , and is dominated by Bacteroides , whose genomes contain highly expanded repertoire of protein families involved in carbohydrate metabolism ., To identify other protein families that are specific to this environment , we investigated the distribution of protein families in the currently available human gut genomic and metagenomic data ., Using an automated procedure , we identified a group of protein families strongly overrepresented in the human gut ., These not only include many families described previously but also , interestingly , a large group of previously unrecognized protein families , which suggests that we still have much to discover about this environment ., The identification and analysis of these families could provide us with new information about an environment critical to our health and well being . | Metagenomics provides a unique opportunity to sample the gene content of microbial communities adapted to specific environments and for the study of the correlations between the presence or absence of gene families that occur in organisms within that environment ., Such studies provide detailed information about the adaptation of microbes to a given environment and , indirectly , provide clues about the most important molecular processes that are specific for that environment ., Having performed such an analysis for the community of the human distal gut , we report many new protein families and identify many others that are highly specific for this particular environment ., The function of most of these proteins is unknown , which illustrates the extent of our ignorance about the organisms within this environment that are so important for human health and well being . | computational biology/macromolecular sequence analysis, computational biology/metagenomics, computational biology/ecosystem modeling, biochemistry/bioinformatics | null |
journal.pcbi.1005725 | 2,017 | Deciphering HLA-I motifs across HLA peptidomes improves neo-antigen predictions and identifies allostery regulating HLA specificity | HLA-I molecules play a central role in defence mechanisms against pathogens and immune recognition of cancer cells ., Their main functionality is to bind short peptides ( mainly 9- to 12-mers ) coming from degradation products of endogenous or viral proteins ., The peptides are cleaved in the proteasome , transported by the transporter associated with antigen processing ( TAP ) complex , loaded onto the HLA-I molecules in the ER and presented at the cell surface 1 ., Non-self peptides presented on HLA-I molecules , such as those derived from degradation of viral proteins , mutated proteins ( referred to as neo-antigens ) , and other cancer specific and abnormally expressed proteins can then be recognized by CD8 T cells and elicit cytolytic activity ., Neo-antigens have recently emerged as promising targets for development of personalized cancer immunotherapy 2 ., Human cells express three HLA-I genes ( HLA-A/B/C ) ., These genes are the most polymorphic of the human genome and currently more than 12 , 000 different alleles have been observed 3 ., Such a high polymorphism makes it challenging to model the different binding specificities of each allele and predict antigens presented at the cell surface ., Information about binding motifs ( mathematically defined here as Position Weight Matrices and graphically represented as sequence logos ) of HLA-I molecules has been mainly obtained from biochemical assays where chemically synthesized peptides are tested in vitro for binding ., This in vitro approach is experimentally laborious , time consuming and expensive ., Currently , the most frequent HLA-I alleles have thousands of known ligands that provide a detailed description of their binding specificity ., Many of these ligands are stored in very important resources such as IEDB 4 , 5 and have been used to train machine learning algorithms for HLA-I ligand predictions 6–11 ., As a result , for frequent alleles in Caucasian populations , much is known about their binding specificity ., However , the vast majority ( >95% ) of HLA-I alleles still lack documented ligands and despite very valuable algorithmic developments to generalize prediction methods to any allele 12 , it remains more challenging to make accurate predictions for alleles without known ligands ., Importantly , although these alleles are only found in a small fraction of the Caucasian population , they are more frequently encountered in other ethnic groups and gaining more understanding of their binding specificity would be desirable for expanding the scope of therapeutic strategies relying on HLA-I ligand predictions ., Moreover , many alleles with known motifs are supported by only a few tens of peptides and some of these ligands have been selected based on a priori expectations of the binding specificity rather than unbiased screening of random peptide libraries ., Such potentially biased datasets can be sub-optimal for training HLA-I ligand predictors ., Mass-spectrometry ( MS ) analysis of HLA-I binding peptides eluted from cell lines or tissue samples is a promising alternative to the use of HLA-I ligand interaction predictions 13 and MS is increasingly used to directly identify viral 14 , 15 or cancer-specific ( neo- ) antigens 13 , 16–20 ., For neo-antigen discovery , tumor exome sequencing is first performed ., Patient-specific non-synonymous somatic alterations are included in a customized database ., MS-MS spectra of HLA ligands eluted from the patient’s tissue samples are searched against this expanded database permitting either the wild-type or the mutated peptide sequences to be identified ., Neo-antigens directly identified this way , or alternatively predicted in-silico , may be further validated using targeted mass-spectrometry approaches in which isotopically heavy-labeled synthetic peptides are spiked into the HLA ligands eluted from the patient’s tumor tissue ., Identification of co-eluting pairs of heavy ( standard ) and light ( endogenous ) peptides validate the presentation of the neo-antigen in the investigated tissue 19 ., However , this technique is only applicable to a small fraction of samples due to the large amount of material that is required for MS analysis ( typically 1cm3 of tissue material , or 1x108 cells in culture ) and the complexity of these experiments ., In addition to potentially immunologically relevant ( neo- ) antigens , tens of thousands of endogenous peptides naturally presented on HLA-I molecules are identified in such HLA peptidomics studies , providing a unique opportunity to collect very large numbers of HLA-I ligands that can be used to better understand the binding properties of HLA-I molecules ., The challenge in studying HLA-I motifs based on such pooled peptidomics data from unmodified cell lines or tissue samples is to determine the allele on which each peptide was displayed ., The most widely used approach is to predict binding affinity of each peptide to each allele present in a sample 21 ., Recent studies by ourselves and others have shown that HLA-I motifs can be identified in HLA peptidomics datasets in an unsupervised way by grouping peptides based on their sequence similarity 17 , 22–25 ., However , this strategy still relies on previous information about HLA-I binding specificity when associating predicted motifs with HLA-I alleles and is therefore restricted to alleles whose motifs have been already characterized ., Here , we describe a computational framework for direct identification and annotation of dozens of HLA-I motifs without any a priori information about HLA-I binding specificity by taking advantage of co-occurrence of HLA-I alleles across both newly generated and publicly available HLA peptidomics datasets ., Our approach recapitulates and refines motifs for many common alleles and uncovers new motifs for eight alleles for which , until this study , no ligand had been documented ., Importantly , this approach is highly scalable and will enable continuous refinement of motifs for known alleles and determination of novel motifs for uncharacterized alleles as more HLA peptidomics data will be acquired in the future ., Training HLA-I ligand predictors based on the refined motifs significantly improves neo-antigen predictions in tumor samples with experimentally determined neo-antigens ., Our large collection of HLA-I ligands further allowed us to unravel some of the molecular determinants of HLA-I binding motifs and revealed allosteric modulation of HLA-I binding specificity ., To elucidate the underlying molecular mechanisms , we show how a single point mutation ( W97R ) in HLA-B14:02 outside of the B pocket significantly changes the amino acid preferences at P2 in the ligands ., To study the binding properties of HLA-I alleles without relying on a priori assumption on their binding specificity and investigate whether this unbiased approach could improve neo-antigen predictions from exome sequencing data , we reasoned that HLA-I binding motifs might be identified across samples with in-depth and accurate HLA peptidomics data by taking advantage of co-occurrence of HLA-I alleles ., To this end , we measured the HLA peptidome eluted from six B cell lines , two in vitro expanded tumor-infiltrating lymphocytes ( TILs ) samples and two leukapheresis samples ( peripheral blood mononuclear cells ) selected based on their high diversity of HLA-I alleles ( see Methods and S1 Dataset ) ., By applying a stringent false discovery rate for peptides identification of 1% , we accurately identified 47 , 023 unique peptides displayed on 32 HLA-I molecules ., To expand the coverage of HLA-I alleles , we further collected 40 publicly available high-quality HLA peptidomics datasets 17 , 18 , 22 , 23 , 26–28 ( see Methods and S2 Dataset ) ., Our final data consists of a total of 50 HLA peptidomics datasets covering 66 different HLA-I alleles ( 18 HLA-A , 32 HLA-B and 16 HLA-C alleles , see Table A in S1 Supporting Information ) ., The number of unique HLA-I ligand interactions across all samples reaches 252 , 165 for a total of 119 , 035 unique peptides ( 9- to 14-mers ) , which makes it , to our knowledge , the largest currently available collection of HLA peptidomics datasets both in terms of number of peptides and diversity of HLA-I molecules ., Binding motifs in each HLA peptidomics dataset were identified for 9- and 10-mers ( see Fig A in S1 Supporting Information ) using a motif discovery algorithm initially developed for multiple specificity analysis 29 , 30 and recently applied to the analysis of a small dataset of seven HLA peptidomics studies 24 ., Importantly , this method does not rely on HLA-I peptide interaction predictions ( see Methods ) ., To assign each motif to its allele even in the absence of a priori information about the alleles’ binding specificity , we developed a novel computational strategy illustrated in Fig 1 ., In this example , one allele ( HLA-A24:02 ) was shared between all three samples ., Remarkably , exactly one identical motif was shared between the three samples ., As such , one can predict that this motif corresponds to the shared allele ., Similarly , two alleles ( HLA-A01:01 and HLA-C06:02 ) were shared between exactly two samples and here again two motifs were shared among the corresponding samples , and could therefore be annotated to their corresponding alleles ., Moreover , if one sample shares all but one allele with another sample , it can be inferred that the motif that is not shared corresponds to the unshared allele ( see example in Fig B in S1 Supporting Information ) , even if some of the shared motifs have not been annotated yet ., Finally , if all but one motif had been annotated in a sample to all but one allele , one can infer that the remaining motif corresponds to the remaining allele ., These three ideas can then be recursively applied to identify HLA-I motifs across our large collection HLA peptidomics datasets ( see Methods ) ., Of note , motifs identified in distinct samples that have some alleles in common show very high similarity ( Fig 1 and Fig B in S1 Supporting Information ) and our new approach builds upon this remarkable inherent reproducibility of in-depth and accurate HLA peptidomics data ., We applied our algorithm to the 50 HLA peptidomics datasets considered in this study ., In total , motifs could be found for 44 different alleles without relying on any a priori assumption of HLA-I binding specificity ( Fig 2A ) ., These include seven alleles ( HLA-B13:02 , HLA-B14:01 , HLA-B15:11 , HLA-B15:18 , HLA-B18:03 , HLA-B39:24 and HLA-C07:04 ) that did not have known ligands in IEDB , and 5 additional ones ( HLA-B38:01 , HLA-B39:06 , HLA-B41:01 , HLA-B56:01 , HLA-C07:01 ) that had less than 50 known ligands ., To validate our predictions , we compared the motifs predicted by our fully unsupervised method with known motifs derived from IEDB 4 , when available ., Despite some differences ( e . g . HLA-A25:01 motif at P9 ) affecting especially alleles with low number of ligands in IEDB ( Fig C in S1 Supporting Information ) , we observed an overall high similarity confirming the reliability of our predicted motifs ( Fig 2A ) ., However , it is important to realize that even small differences in the motifs can have important effects on the performance of predictors that are trained on such data when ranking very large lists of potential epitopes ., When comparing with data recently obtained by HLA peptidomics analysis of mono-allelic cell lines 31 , a very high similarity was also observed ( Fig 2B , stars in Fig C in S1 Supporting Information ) , which further validates our computational approach for HLA-I motif identification and annotation from in-depth pooled HLA peptidomics data ., As expected from many previous studies , alleles with the same two first digits code showed high similarity in their binding specificity , apart from HLA-B15 alleles which are known to be more diverse 32 ., This includes many of the new motifs ( e . g . , HLA-B14:01 vs HLA-B14:02; HLA-B18:03 vs HLA-B18:01; HLA-B39:24 vs HLA-B39:06 ) , which provides further evidences of the accuracy of our predictions for these uncharacterized alleles ., For the most frequent HLA-I alleles , including several shown in Fig 2A , a good description of their binding motifs can be already obtained from existing databases 4 ., To further expand our collection of HLA-I binding motifs , we used similarity to the binding motifs derived from IEDB ligands to annotate motifs that could not be assigned to their corresponding allele by the fully unsupervised algorithm , following the approach previously introduced by ourselves in ref 24 ( see Methods ) ., This enabled us to determine the binding motifs of 8 additional alleles ( Fig 2C ) ., Of note , the new motif of HLA-A02:20 was predicted by observing that it was the only motif not annotated in one sample ( RA957 ) and could only be annotated to this allele based on the motifs identified in other samples for all the other alleles ( see Fig A in S1 Supporting Information ) ., The final list of motifs for the 52 alleles and detailed comparison with IEDB derived data , when available , is shown in Fig D in S1 Supporting Information ., Importantly , for the majority of alleles considered in this study , the motifs are supported by significantly more ligands than what is available in existing databases ( Fig E in S1 Supporting Information ) , and in total our approach enabled us to collect 88’051 unique 9-mer peptide HLA-I interactions for all alleles annotated in this work , compared to the 57’651 interactions available in IEDB for the same set of alleles ., 14 out of a total of 195 motifs ( corresponding to 5’703 9-mer peptide-HLA interactions ) could not be annotated by our approach ( see Fig A in S1 Supporting Information ) ., To investigate how our approach depends on the number of samples in which a motif is found , we show in Fig F in S1 Supporting Information the distance between our predicted motifs and those derived from mono-allelic cell lines ( Fig F, ( a ) in S1 Supporting Information ) or those pooled from all samples ( Fig F, ( b ) in S1 Supporting Information ) as a function of the number of samples ( i . e . sub-sampling ) ., As expected , higher similarity could be observed by integrating several samples , justifying our idea of collecting as much data as possible from different HLA peptidomics datasets to refine our motifs ., But overall , all distances are very small ( D2 < 0 . 03 ) , highlighting the excellent reproducibility of the motifs deconvoluted from HLA peptidomics datasets ., To explore the statistical significance of the motifs associated to the same alleles , we followed the approach of ref ., 33 and compared the similarity ( both Euclidean distance and BLiC score ) between each pair of motifs ( mi , mj ) annotated to the same allele h ( i = 1 , …Nh , j = 1…Nh , i≠j with Nh the number of motif annotated to allele h ) to the distribution of similarity values between motif mi and all known HLA-I motifs 33 ( see Methods ) ., Fig G in S1 Supporting Information shows that more than 99% of the pairs of motifs associated to the same alleles have a statistically significant similarity ( P<0 . 05 ) , confirming the few examples shown in Fig 1 ., Exceptions consist mainly of motifs annotated to HLA-C alleles , which are more degenerate and therefore more difficult to deconvolute 24 ., We therefore recognize that our motifs are likely less accurate for HLA-C alleles , but emphasize that these alleles are also poorly described in existing databases or literature ., We further explored the effect of the threshold T = 0 . 078 on Euclidean distance manually defined in this work ( see Methods ) ., As expected lower values of T still results in highly similar motifs annotated to the same alleles , but in fewer alleles to which motifs can be annotated ( Fig H in S1 Supporting Information ) ., Reversely , larger values tend to increase the number of alleles with annotated motifs , but at some point ( T>0 . 09 ) more distinct motifs ( P>0 . 05 , based on Euclidean distance ) become associated to the same alleles ( Fig H in S1 Supporting Information ) ., The full pipeline was also applied on 10-mers identified by MS across the 50 HLA peptidomics studies and revealed six new motifs for poorly characterized alleles in IEDB ( Fig I in S1 Supporting Information ) ., Different technical biases may affect MS data , which could undermine their use for training HLA-I ligand predictors ., To investigate this potential issue , we computed amino acid frequencies at non-anchor positions ( P4 to P7 ) in our HLA peptidomics data , excluding alleles displaying anchor residues at these positions ( see Methods and Table B in S1 Supporting Information ) ., The reason for focusing on middle positions is that they display low specificity ( especially in 9-mers , see discussion in 24 , 34 , 35 for longer peptides ) and therefore could provide a global view of potential MS biases on amino acid frequencies that is not affected by the constraints of binding to HLA-I molecules ., As expected , we observed a good correlation between amino acid frequencies at non-anchor positions in our HLA peptidomics data and in the human proteome ( r = 0 . 85 ) ( Fig 3 and Fig J in S1 Supporting Information ) ., The most important difference that could strongly affect predictors trained on such data was found for cysteine , which is prone to post-translational modifications that are typically not included in database searches and was observed at very low frequency in the HLA peptidomics data ( the same observation was recently made in mono-allelic cell lines 31 ) ., Moreover , IEDB data clearly indicate that cysteine can be found at non-anchor positions , including for immunogenic epitopes , and therefore the low frequency observed in MS data very likely corresponds to a technical bias ., Other amino acids were less under- or over-represented and the differences observed with the human proteome may also reflect some residual specificity at non-anchor positions ., Moreover , no clear pattern emerged from these data with respect to amino acid biophysical properties ( e . g . , charge , hydrophobicity , size ) ., Overall , our results suggest that HLA peptidomics data do not show strong technical biases , apart from under-representation of cysteine ( see next section for a proposed method on how to compensate this bias ) , and therefore could provide ideal data for training HLA-I peptide interaction predictors , especially for ligands coming from human cells like neo-antigens ., To test whether our unique dataset of naturally presented peptides could help predicting HLA-I ligands , including neo-antigens in tumors based on exome sequencing data , we trained a predictor of HLA-I ligands ( referred to as MixMHCpred ) ., As MS only includes positive examples and HLA-I ligands in general do not show strong amino acid correlations between different positions ( see discussion in 24 for some exceptions ) , we built Position Weight Matrices ( PWMs ) for each of the 52 alleles ., These PWMs were built by pooling together all peptides assigned to each allele across all our HLA peptidomics datasets ( see Methods and Fig 1 ) ., We further included MS data from mono-allelic cell lines for 6 rare alleles that were not present in our dataset , resulting in a total of 58 alleles available in our predictor ., To correct for the low detection of cysteine observed in HLA peptidomics data we further renormalized our predictions by amino acid frequencies at non-anchor positions ( see Methods and Table B in S1 Supporting Information ) ., As a first validation , we attempted to re-predict naturally presented peptides experimentally identified in ten mono-allelic cell lines whose alleles overlapped with our dataset 31 ., For this analysis , we did not include data from these mono-allelic cell lines in our training set ., To assess our ability to predict naturally presented peptides , we added 99-fold excess of decoy peptides randomly selected from the human proteome to each mono-allelic cell line dataset and measured the fraction of Positives among the top 1% Predictions ( PP1% , i . e . , True Positive Rate among the top 1% ) , which in the case of 99-fold excess of decoy is equivalent to both the Precision and the Recall , since the number of predictions ( top 1% ) is equal to the number of actual positives ( 1% ) ., For all but one allele , our algorithm showed higher or equal predictive power compared to standard HLA-I ligand predictors 8 , 12 , 36 ( Fig 4A ) ., We further measured the average Area Under the Curve ( AUC ) for these alleles and obtained quite similar values ( 0 . 978 for MixMHCpred , 0 . 976 for NetMHC , 0 . 979 for NetMHCpan and 0 . 977 for NetMHCstabpan ) ., However , we emphasize that most random peptides used as negatives are quite distinct from the positives , which can explain the very high AUC values and we suggest that precision values for top 1% of the predictions shown in Fig 4A are more representative of the actual performance of the algorithms ., We then collected currently available datasets that included direct identification of neo-antigens displayed on cancer cells as well as exome sequencing data ( Mel5 , Mel8 , Mel15 from 17 and 12T from 20 ) for a total of ten 9- and 10-mers mutated peptides experimentally found to be presented on cancer cells ( see Table 1 ) ., This dataset has the unique advantage of not being restricted to peptides selected based on in silico predictions , and is therefore an ideal testing set for benchmarking our predictor ., Moreover , as these studies are quite recent , the neo-antigens used here as testing set are not part of the training set of any existing algorithm ., In particular , they are not part of the large training set used in this study since we only included wild-type human peptides in our pipeline ., We then retrieved all possible 9- and 10-mer peptides that encompassed each missense mutation ( S3 Dataset ) and ranked separately for each patient these potential neo-antigens based on the score of our predictor ( see Methods and Table 1 ) ., Remarkably , six of the ten neo-antigens fell among the top 25 predicted peptides , suggesting that by testing as few as 25 mutated peptides per sample , we could identify more than half of the neo-antigens identified by MS ( Table 1 ) ., Considering that the total number of potential neo-antigens ( i . e . 9- and 10-mers containing a missense mutation ) can be as large as 25 , 000 for tumors with high mutational load , our predictor trained on naturally presented human HLA-I ligands clearly enabled us to significantly reduce the number of peptides that would need to be experimentally tested to identify neo-antigens from exome sequencing data ., We further added two datasets of neo-antigens identified in lung cancer patients ( L011 and L013 ) 37 , although only peptides pre-selected based on binding affinities predicted with existing tools 12 were tested in this study ., Here again , our predictor ranked one neo-antigen in the top 25 predicted peptides in both samples ( Table C in S1 Supporting Information ) ., When comparing with standard tools that are widely used to narrow-down the list of potential neo-antigens predicted from exome sequencing data 8 , 12 , 36 , our method trained on HLA peptidomics data showed clear improvement ( Table 1 ) with a mean AUC value of 0 . 979 , compared to 0 . 932 for NetMHC 8 , 0 . 942 for NetMHCpan 12 and 0 . 945 for NetMHCstabpan 36 ( Fig 4B ) and increased number of neo-antigens in the top 1% of the predictions ( i . e . , typically what is experimentally tested for immunogenicity ) across all six samples ( Fig 4C ) ., This is especially clear for the 12T sample , where the single neo-antigen was very well predicted by our model and poorly predicted by existing tools ( >6’000nM with HLA-B51:01 , see also Table 1 and 20 ) ., We still note that , due to the low number of neo-antigens publicly available together with exome sequencing data , performance metrics in Fig 4B and 4C can be sensitive to one neo-antigen being better or less well predicted and we stress that the values shown in Fig 4B and 4C are simply a graphical way of looking at data shown in Table 1 and Table C in S1 Supporting Information ., Importantly , even if we did not include in the training of our predictor MS data ( i . e . wild-type peptides ) from the samples in which the neo-antigens were identified , neo-antigens were still more accurately predicted compared to other tools ( see Fig K in S1 Supporting Information ) ., This demonstrates that our approach for neo-antigen predictions from the list of somatic mutations identified by exome sequencing of tumors does not require HLA peptidomics data from the same sample where neo-antigens had been identified ., Nevertheless , both our predictor and standard prediction tools failed to identify some neo-antigens ( e . g . , KLILWRGLK from NCAPG2 P333L mutation , see Table 1 ) ., This suggests that , when enough tumor material is available for HLA peptidomics analyses , direct identification of neo-antigens with MS should still be performed to optimally enrich in true positives the list of ligands to be experimentally tested for immunogenicity 16 , 19 ., The number of studies reporting both neo-antigens and exome sequencing results is still limited ., To benchmark our algorithm with larger datasets of immunologically relevant tumor antigens , we tested our ability to predict epitopes from cancer testis antigens ., We retrieved all epitopes listed in the CT database 38 ( see Methods and Table D in S1 Supporting Information ) ., We then assessed how our predictor could prioritize these epitopes from all possible peptides encoded by these cancer testis antigens ., Although we cannot exclude that some of these epitopes had been selected for experimental testing after prediction by older versions of HLA-I ligand predictors , we still observed improvement using our predictor trained only on naturally presented HLA-I ligands , both in terms of AUC and fraction of true positives that fall in the top 1% of the predictions ( Fig 4D and 4E ) ., This indicates that improvement in prediction accuracy is not restricted to elution data ( see similar results in 24 ) ., MS data can contain false positives for many different reasons , such as co-eluting peptide contaminations or errors in the computational identification of peptides from the spectra ., Therefore , despite the high quality of HLA peptidomics datasets generated in this study ( <1% FDR ) , we do expect our data to contain some noise ., To test the robustness towards contaminations of our motif discovery and annotation pipeline , and our HLA-I ligand predictor , we incorporated 5% of random peptides from the human proteome into all HLA peptidomics datasets considered in this work and rerun the whole motif annotation pipeline and training of the predictor ., Remarkably , the accuracy of the predictions was only very modestly affected by this noise and predictions were still better than with other existing tools ( Fig L in S1 Supporting Information ) ., To explore the effect of wrong peptide identification , we reprocessed with MaxQuant 39 the three MS samples shown in Fig 1 and chose the second best hit for 1% of the peptides ., Overall the motifs predicted by our approach remained almost unchanged ( Fig M in S1 Supporting Information ) ., This suggests that our pipeline is robust and indicates that the wealth of unbiased and accurate data provided by MS can compensate the inherent contaminations , when using these data for training HLA-I ligand predictors ., An important step in our predictor is the renormalization by amino acid frequencies observed at non-anchor positions , which was designed to correct for biases in MS data ., As expected , doing this renormalization step with amino acid frequencies observed in the human proteome ( or no renormalization at all ) results in very low frequency of cysteine-containing peptides among the top predicted ligands ., As such , it improves the predictions of MS data ( see especially Fig N, ( a ) in S1 Supporting Information ) , but decreases the performance in other datasets ( e . g . , Fig N ( b-c ) in S1 Supporting Information for L011 and L013 ) ., These results highlight the importance of carefully considering MS biases when including such data to train predictors in order to avoid over-fitting elution data ., We anticipate that additional work may further improve this step , such as inclusion of cysteine modifications in spectral searches 31 or better estimations of the expected baseline amino acid frequencies in HLA peptidomics datasets ., One of our novel HLA-I motifs describes the binding specificity of HLA-A02:20 ( Fig 2C ) ., HLA-A02 binding motifs have been widely studied ., However , HLA-A02:20 motif differs from standard HLA-A02 motifs at P1 , with a clear preference for charged residues ( Fig 5A ) ., Interestingly , HLA-A02:20 is among the very few ( <2% ) HLA-A02 alleles that do not have a conserved lysine pointing towards P1 at position 66 ( residue numbering follows 2BNQ X-ray structure hereafter ) ., Instead an asparagine is found there ( Fig 5A ) , and this residue is the only difference with the sequence of the very common A02:01 allele ., To explore how the absence of lysine at position 66 could explain the observed difference in binding specificity , we collected all HLA-I alleles showing preference for charged amino acids at P1 ( see Fig O in S1 Supporting Information ) ., All of them had either asparagine or isoleucine at position 66 ., We then explored available crystal structures of HLA-I peptide complexes with charged residues at P1 ., HLA-B57:03 was crystalized with such a ligand ( KAFSPEVI ) 40 ., Superposing the crystal structure of this complex with the structure of HLA-A02:01 provides a possible mechanism for understanding the change in binding specificity at P1 ., In HLA-A02:01 , lysine at position 66 interacts with the hydroxyl group of serine at P1 ( Fig 5A , green sidechains ) ., Such a conformation would not be compatible with a longer residue ., Reversely , when asparagine was found at position 66 , it did not point towards P1 ( Fig 5A , pink sidechains ) , thereby freeing space for larger sidechains like lysine or arginine at P1 ., Overall , our analysis indicates that the presence of asparagine at residue 66 may be responsible for the change in binding specificity between HLA-A02:01 and HLA-A02:20 ., More generally , our results suggest that lysine at residue 66 in HLA-I alleles strongly disfavours charged residues at P1 ., The new motif identified for HLA-B15:18 ( Fig 2A ) displayed strong preference for histidine at P2 , which is not often observed in HLA-I ligands ., To gain insights into the mechanisms underlying this less common binding motif , we surveyed all alleles that show preference for histidine at P2 ( Fig P, ( a ) in S1 Supporting Information ) ., Sequence and structure analysis showed that all of them have a conserved P2 binding site , commonly referred to as the B pocket ( see Fig 5B ) ., However , several HLA-B14 alleles have exactly the same B pocket but show specificity for arginine at P2 ( Fig P, ( b ) in S1 Supporting Information ) ., Among them , HLA-B14:02 had the highest sequence similarity to HLA-B15:18 , with only 8 different residues in the peptide binding domain , none of them making any contact with arginine at P2 in the crystal structure of HLA-B14:02 ( orange residues in Fig 5C ) ., This suggests that the difference in binding specificity at P2 between HLA-B14:02 and HLA-B15:18 is likely explained by allosteric mechanisms ., Of particular interest is residue 97 ( W in HLA-B14:02 and R in HLA-B15:18 ) , which is in the HLA-I binding site and contacts the peptide ( mainly P3 to P6 , Fig 5C ) but is more than 7Å away from the arginine sidechain at P2 ., This residue is part of a network of aligned aromatic residues ( Y9 , W97 and F116 ) in HLA-B14:02 ( Fig 5C ) compatible with pi-pi interactions ., Interestingly , X-ray structures with Arg at position 97 ( e . g . , 4O2C ) show a flip in the | Introduction, Results, Discussion, Methods | The precise identification of Human Leukocyte Antigen class I ( HLA-I ) binding motifs plays a central role in our ability to understand and predict ( neo- ) antigen presentation in infectious diseases and cancer ., Here , by exploiting co-occurrence of HLA-I alleles across ten newly generated as well as forty public HLA peptidomics datasets comprising more than 115 , 000 unique peptides , we show that we can rapidly and accurately identify many HLA-I binding motifs and map them to their corresponding alleles without any a priori knowledge of HLA-I binding specificity ., Our approach recapitulates and refines known motifs for 43 of the most frequent alleles , uncovers new motifs for 9 alleles that up to now had less than five known ligands and provides a scalable framework to incorporate additional HLA peptidomics studies in the future ., The refined motifs improve neo-antigen and cancer testis antigen predictions , indicating that unbiased HLA peptidomics data are ideal for in silico predictions of neo-antigens from tumor exome sequencing data ., The new motifs further reveal distant modulation of the binding specificity at P2 for some HLA-I alleles by residues in the HLA-I binding site but outside of the B-pocket and we unravel the underlying mechanisms by protein structure analysis , mutagenesis and in vitro binding assays . | Predicting the differences between cancer and normal cells that are visible to the immune system is of central importance for cancer immunotherapy ., Here we introduce a novel computational framework to harness the wealth of data from in-depth HLA peptidomics studies , including ten novel high quality ( <1% FDR ) datasets generated for this work , to improve predictions of peptides displayed on HLA-I molecules ., These high-throughput and unbiased data enable us to refine models of HLA-I binding specificity for many alleles ( including some that had no ligand until this study ) and improve predictions of neo-antigens from exome sequencing data in melanoma and lung cancer samples ., Moreover , the refined description of HLA-I binding specificity reveals cases of allosteric modulation of HLA-I binding specificity at the second amino acid position ( P2 ) of their ligands by residues that are part of the HLA-I binding site but outside of the B pocket . | chemical characterization, medicine and health sciences, chemical compounds, cancers and neoplasms, organic compounds, oncology, basic amino acids, amino acids, sequence motif analysis, chemical synthesis, research and analysis methods, sequence analysis, peptide synthesis, bioinformatics, proteins, melanomas, chemistry, binding analysis, biosynthetic techniques, biochemistry, histidine, peptides, arginine, organic chemistry, proteomes, database and informatics methods, biology and life sciences, physical sciences | null |
journal.pbio.0050167 | 2,007 | Comparison of C. elegans and C. briggsae Genome Sequences Reveals Extensive Conservation of Chromosome Organization and Synteny | The comparative analysis of the related nematodes Caenorhabditis elegans and C . briggsae offers a powerful approach toward understanding the genetic basis for the form and function of these simple animals ., Studies to date have already yielded valuable insights into the evolution and role of particular sequences , genes , and pathways 1 , 2 ., Morphologically , the two species are almost indistinguishable , despite the fact that their most recent common ancestor ( MRCA ) existed about 100 million years ago ( Mya ) ., Both are soil-dwelling , self-fertilizing hermaphrodites , with facultative males ., Both have a ~100-megabase ( Mb ) genome apportioned into six chromosomes ., Genes isolated in one species will frequently rescue mutants in the other 3 , 4 ., Despite these similarities , nucleotide alignments ( using the wobble-aware bulk aligner WABA algorithm 5 ) of the complete genome sequence of C . elegans 6 , 7 with the draft sequence of C . briggsae strain AF16 reveals that 52 . 3% of the C . elegans genome and 50 . 1% of the C . briggsae genome aligns between the two species with the bulk of this in coding sequence 8 ., The substantial body of knowledge accrued about C . elegans over the past few decades will help interpret the sequence similarities and differences ., Much less is known about C . briggsae ., To facilitate the molecular genetic study of C . briggsae and thus enhance its utility for further comparative analysis , we sought to convert the whole genome sequence assembly into a genome map , in which the genome sequence and genetic maps are linked to each other through common markers across the chromosomes ., Before our present work , the draft whole genome assembly contained 102 Mb of sequence in 142 physical map–based contigs ( fpc contigs ) , with the remaining 6 Mb in 463 supercontigs ( see Materials and Methods ) ., The classical genetic map ( Bhagwati Gupta , personal communication ) has fewer than 40 mutants placed on the six linkage groups and only ten of these have a molecular assignment ., The large number of contigs and the paucity of genetic mapping data did not allow meaningful merging of the two maps ., We undertook the construction of a genome map by first generating a genetic map using molecularly based single nucleotide polymorphism ( SNP ) markers ., This more detailed genetic map based on SNPs would be of use in its own right , for example , simplifying positional cloning of genetically defined genes ., But it would also provide long-range continuity , which would in turn allow the placement of much of the assembled sequence along the chromosomes ., This long-range map of the genome would in turn allow a direct comparison of chromosomal organization in C . briggsae to the distinctive features of C . elegans organization 6 , 9 ., Using other wild isolates of C . briggsae , we discovered thousands of SNPs ., By genotyping selected SNPs across recombinant inbred ( RI ) lines between the sequenced strain ( AF16 ) and the SNP source strains , we generated a genetic map ., We then combined the resultant genetic map and the sequence assembly information to place 91 . 2 Mb of sequence onto the six linkage groups , with another 9 . 9 Mb tentatively associated ( but not ordered ) with chromosomes ., The integrated map allowed us to correct several misassemblies in the initial C . briggsae sequence ., Of broader interest , we were also able to explore chromosomal scale phenomena ., Like in C . elegans , rates of recombination appear much higher on arms than in central regions for the autosomes ., Autosome arms and centers also differ in their repeat content , coding density , and fraction of highly conserved genes , as is seen in C . elegans ., Unexpectedly , the comparison also revealed an extensive conservation of synteny between the two organisms , with the vast majority of genes with 1:1 orthologs that reside on one chromosome in one species lying on a single chromosome in the other ., Long-range gene order within the chromosomes has been substantially altered in the 100 million years ( Myr ) since their MRCA , but despite these rearrangements , sequences tend to remain in their respective domains of arm or center ., Our findings support the emerging recognition of the importance of overall chromosomal organization in metazoans ., To find a suitable strain for SNP discovery , we investigated four independent wild isolates that grow well in culture , are interfertile with the sequenced AF16 strain , and represent both tropical and temperate groups 10 , 11 ( Table 1 ) ., We initially aligned a small number of random genomic sequences against the AF16 assembled sequence to ascertain the approximate incidence of single nucleotide variation ., Two temperate strains of Japanese origin ( HK104 and HK105 ) both had relatively high rates of difference ( ~1 SNP/110 bases ) while the Hawaiian ( VT847 ) and the Ohio ( PB800 ) strains ( tropical and temperate respectively ) had apparently lower rates ( see Materials and Methods for details of SNP detection ) ., We selected one strain of each level ( HK104 and VT847 ) for more extensive SNP discovery ., From 8 , 405 and 9 , 970 aligned sequence reads from whole genome shotgun libraries from each strain we identified respectively 32 , 246 and 14 , 183 substitutions with Phred 12 quality scores of greater than 35 , giving overall rates of 8 . 7 and 3 . 2 per kilobase ., We also identified a number of candidate small insertion/deletion differences ( 7 , 118 events affecting 18 , 196 bases and 3 , 575 events affecting 8 , 315 bases , respectively ) ., To construct a genetic map , 390 SNPs distributed across the sequence were tested against 93 AF16 × HK104 RI lines and the parental strains 10 using the fluorescent polarization detection ( FP-TDI ) assay ( Vieux et al . 2002; see Materials and Methods for details of SNP assay ) ., To maximize the amount of sequence mapped and to provide an independent check of the assembly , the 390 SNPs were selected from the larger supercontigs , thus ensuring that the larger physical map–based contigs ( called fpc contigs for simplicity , after the program used to assemble the physical map 13 ) would contain multiple markers and thus serve to check the assembly ., In about a quarter of the cases , a second SNP was selected within a single supercontig to test the assembly at this level ., A SNP was declared as mapped when the assays were successful on between 80% and 100% of the 95 strains tested , with a total of 248 SNPs ( 64% ) meeting this criterion ., Some 84 SNPs ( 22% ) had success rates between 0% and 40% and were deemed failures ., The high rate of failures was likely caused by PCR problems due to unaccounted SNPs in primer sites , a problem faced by all investigated genotyping platforms 14 ., Some five SNPs ( 1 . 3% ) were monomorphic and likely due to false SNP calls ., Other SNPs failed quality control tests or had success rates of 40–80% ., These same SNP assays were also tested against the VT847 strain , for which RI lines were also available ., Relatively few of the AF16/HK104 SNPs were polymorphic between AF16 and VT847 , suggesting that the overlap in variation between the HK104 and VT847 is very limited ., This meant that genotyping of these additional RI lines with these markers added little new map information ., We tested several different parameters for map construction , using the program Map Manager QTXb20 ( http://www . mapmanager . org/ 15 ) ., The different versions varied in map length per chromosome , total map length , and in the local order of markers within a chromosome , but assignment of markers to common linkage groups was a robust feature of the maps ., The latter was due in part to the large number of nonrecombinant chromosomes in the RI lines ( 35–60% per chromosome ) , which allowed ready assignment to linkage groups ., Based on these experiences , we adopted the following strategy to build version 3 . 0 of the genetic map: we used an initial set of 115 very high quality markers ( >95% call rates ) and a second set of slightly lower quality ( 80–95% call rates ) ., We used the Haldane function and an initial probability of incorporation of a SNP into the map of 10−6 ., Seven linkage groups were formed , one with only two SNPs ( cb23233 and cb23314 ) ., We reduced the probability required for incorporation to 10−3 , and the latter group was incorporated into the end of chromosome CbIV ., Thus the number of linkage groups matched the observed number of chromosomes ( Table 2 ) ., The program provided map positions in centimorgans ( cM ) for each of the incorporated markers , with each of the chromosomes approximately 50 cM in length ., Inspection of the raw data in the version 3 . 0 map in conjunction with the marker order in the sequence assembly highlighted places where markers of equivalent or nearly equivalent position in the genetic map could be shuffled to reconcile their order with that in the assembly ., In addition the initial genetic map of the X chromosome ( CbX; see below for chromosome assignments ) showed a number of inconsistencies with the sequence assembly that could all be reconciled by a single inversion of the central segment of the genetic map for CbX ., Additional recombinant data obtained for CbX from an experimental cross ( see Materials and Methods ) supported the revised genetic marker order ., These changes were incorporated into version 3 . 1 ., Finally , inspection of the raw data in conjunction with the known groupings of markers based on the assembly suggested alternate orders of markers on chromosomes CbI , CbIV , and CbV that reduced the number of multiply recombinant chromosomes ., These changes reduced overall map length by over 16 cM and did not reduce logarithm of the odds scores ( logarithm of the odds score is a statistical estimate of whether two loci are likely to be near each other on a chromosome and therefore likely to be inherited together ) of any markers below the threshold; they were incorporated into the genetic map to produce version 3 . 2 ., Using this framework map , inspection of the remaining markers with lower call rates indicated that 44 of them could be readily linked to chromosomes and tentatively positioned within the chromosome ., These added markers sometimes helped in orienting contigs and in five cases , positioned previously unplaced contigs ., However , the lower overall call rates of these markers make their placement less certain ., With the genetic map in place , we examined the frequency of parental alleles within the RI lines across the chromosomes ., For chromosomes CbII and CbX , there was little variation from the expected value of 50% for each marker ., But for other chromosomes , there were regions of biased representation of the AF16 and HK104 alleles ., For example , the AF16 allele was consistently underrepresented for most of CbIII , whereas it was overrepresented for much of chromosome CbIV ( Figure 1 ) ., Chromosomes CbI and CbV also showed biased representation , but over more limited regions ( see Datasets S1 and S2 ) ., The biased representation of alleles presumably reflects some selective advantage for offspring with these regions , either singularly or in combination ., The selection of the first progeny at each generation in establishing the RI lines may have contributed to this bias ., The relatively small number of recombinant events in these lines however precludes finer localization of such factors ., The sequence-based markers used in the construction of the genetic map allowed for ready integration of the genetic and sequence maps into a genome map ., The association of a genetic marker with a supercontig and , in turn , an fpc contig positioned that sequence on a specific chromosome , and when multiple , genetically separated markers were assigned to a single sequence assembly , that sequence could be oriented ., Generally , multiple markers from the same supercontig or fpc contig had adjacent positions in the genetic map , confirming the assembly in these instances ., However , markers assigned to 21 sequence assemblies were derived from more than one linkage group , indicating an error in either the genetic linkage assignment or in the sequence assembly ., Because marker assignment to linkage groups was a robust feature of the genetic map and inspection of the raw data revealed no problems with the assignment in these discordant cases , the sequence map was probed for possible errors ., Only one discrepancy was noted among 68 supercontigs with more than one marker , suggesting that misassemblies within supercontigs ( constructed by using read-pair information to link sequence contigs ) were unlikely to account for the bulk of the observed discrepancies ., On the other hand , we noted that most markers with discordant linkage fell on fpc contigs ( in which supercontigs were linked based on the physical clone map information ) ., Detailed inspection showed that in these cases , a join based on the physical clone map information fell between discordant markers ., Once the conservation of synteny between C . elegans and C . briggsae chromosomes was established ( see below ) , the 1:1 orthology landmarks were used to delimit the region with the assembly problem , making it clear that the discrepancies arose because of false joins based on the lower resolution physical clone map ( Figure 2 ) ., Inspection of the physical map in a sample of these regions revealed questionable clone overlaps often accompanied by an editors comment to that effect , consistent with a misassembly at that point ., As a result , 27 breaks were made in the fpc contigs at the site defined by the orthology landmarks ( renamed as segments a , b , etc . of the parent contig ) ., The single discordant supercontig was also broken at a site bounded by the ortholog landmarks ., These breaks in the sequence assembly eliminated the inconsistencies between assignment of the markers to sequence assemblies and linkage groups ( Table 3 ) ., Obviously , other misassemblies may remain undetected , because misassembled regions failed to have a genetic marker ., Investigation of the entire sequence for clusters of discordant orthologs suggests five regions of more than 50 kilobases ( kb ) that are likely candidates for misassembly ., Further , our analysis is less sensitive to misassemblies within the same chromosome , because precise order within linkage groups is less robust , making detection harder ., Nonetheless , with one exception , markers in a single sequence assembly lie adjacent to one another in the current map ., In the exception ( cb25 . fpc4010 ) , a high-quality marker maps to the right end of chromosome CbIII , whereas two low-confidence markers suggest positions near the opposite end ., Further , with one exception , multiple markers in a single sequence assembly fall in an order consistent with the genetic map order ., In the single exception , a simple inversion of a pair of SNP markers in cb25 . fpc3752 would reconcile the maps ., However , we only altered the sequence assembly where there were compelling genetic data that the assembly was in error ., The integrated genetic and sequence map provide an initial genome map ., The confidently placed genetic markers position 141 sequence assemblies , accounting for 89 . 4 Mb of the sequence along the chromosomes , with 42 of these oriented , accounting for 47 . 7 Mb ., Inclusion of the lower-confidence markers provides tentative positions for an additional five assemblies , containing 1 . 8 Mb ., By using read-pair information for assemblies adjacent in the genetic map , we were able to orient an additional 45 contigs , bringing the total oriented sequence to 67 . 3 Mb ., In addition , by considering local order of 1:1 orthologs in both species ( see below ) , we could tentatively order an additional 4 . 4 Mb ., This reconciled genome map is reflected in version 3 . 3 of the genetic map ., In C . elegans , a distinctive feature of the genetic map is the reduced recombination per Mb of the centers of the autosomes compared with the arms 16 ., We looked at the recombination rates across the C . briggsae autosomes and the putative X chromosome ( see below ) to see if the same features existed ., Similar to that of C . elegans , each of the C . briggsae autosomes shows reduced recombination in the centers compared to the arms ( Figure 3A , Datasets S3 and S4 , and Figures S1–S4 ) ., Indeed , no recombinant events were observed in the RI lines over several megabases of the centers of several chromosomes , even though 60–70 recombinant events were observed on the 11–16-Mb autosomes ., In contrast , recombination rates are more uniform on the presumptive X chromosome ( Figure 3B ) ., These observations must be interpreted with some caution , because the C . briggsae genome map contains only 85% of the sequence , and assembly biases could mean that much of the unassigned sequence belongs on the arms ., Further , some biases were introduced in the recovery of the RI lines , as noted above , which might also distort recombination rates ., Finally the sequence differences and perhaps even inversions between the two strains could reduce recombination rates in local regions ., Nonetheless , the general features seen here seem likely to be reflected in a more comprehensive map ., In addition to the marked variation in recombination rates along the autosomes in C . elegans , repeat density and gene density were found to vary by region 6 ., We observed similar variation in density of these features in the C . briggsae autosomes , with the repeat density higher and intron length greater on the arms and exon density greater in the centers ( Figure 4 , Datasets S3 and S4 , and Figures S1 and S2 ) ., Again , as seen in C . elegans , telomere related repeats ( TTAGGC ) show a particularly marked difference in distribution ., Strikingly , 1:1 orthologs , even after accounting for the greater exon density in the centers , are more common in the centers ., With the bulk of the C . briggsae genome placed along chromosomes , the conservation of synteny ( using synteny here in the originally defined sense of genes on the same linkage group or chromosome ) and colinearity ( meaning the order of genes along the chromosome ) between C . elegans and C . briggsae could be investigated directly across the whole genome ., Early analyses of colinearity , using clone-based datasets of limited sequence continuity , estimated median tract lengths of <10 kb in one study 5 and 17 kb for autosomes and 41 kb for the sex chromosome in a second study 17 ., The initial analysis of the C . briggsae whole-genome assembly observed a mean of 37 , 472 base pairs ( bp ) and a median 5 , 557 bp with a maximum block of 1 . 68 Mb 8 ., This initial analysis used genome-wide alignment data and allowed regions to match as many as five segments in the reciprocal genome ., Inspection of the junctions between the 4 , 837 candidate colinear blocks ( minimum length 1 . 8 kb ) suggested the breakpoints represented 1 , 384 inversions , 244 translocations , and 2 , 735 transpositions ., To make the present analysis less sensitive to repeated sequences and to small blocks of similarity that may have arisen by the large number of transposition events , we began by identifying 9 , 767 1:1 gene pairs ( where each gene was represented only once in its genome and matched only one gene in the other genome ) using the previously defined gene set 8 ., These data provide an unambiguous orthologous landmark on average about every 10 kb ., For those sequence assemblies that had only one genetic marker or that had genetic markers all on a single linkage group in the initial map , we found that the 1:1 orthologs on that assembly overwhelmingly derived from a single C . elegans chromosome ., The same observations held for the corrected assemblies ., More remarkably , we noted for sequence assemblies assigned unambiguously to the same C . briggsae linkage group that the 1:1 orthologs assignments were consistently from a single C . elegans chromosome ( Table 4 ) ., Exceptional orthologs were often isolated , singular events ., This remarkable conservation of synteny between the two species allowed us to assign not just regions but each of the entire C . briggsae linkage groups to its corresponding C . elegans chromosome ., To look at the colinearity of the orthologs within chromosomes , we plotted their location in each of the pairs of syntenic chromosomes ( Figure 5 , Dataset S5 , and Figure S3 ) ., There have been extensive intrachromosomal rearrangements , but large colinear blocks remain , especially in the centers ., More interestingly , sequences that are in the central , low-recombination segment of one species tend to be in the corresponding region in the other species ., By contrast , there is mixing between the two arms ., To quantify this , we established blocks of sequence with the same order of genes in the two genomes allowing minor exceptions ( see Materials and Methods ) ., Our methods yielded only 851 blocks using a minimum block size of one ortholog , with only a third of these more than 50 kb long ., Because our analysis excludes repeated sequences , these numbers do not reflect most transposition events , which formed the bulk of the rearrangements detected in 8 ., Nonetheless , 351 of the 1:1 ortholog blocks are small enough ( <20 kb ) to be consistent with transposition events ., Only 12 blocks greater than 20 kb involve nonsyntenic orthologs and might represent translocations; none of these have confirmatory genetic markers and could all represent assembly problems ., Thus the only confirmed rearrangements represent intrachromosomal events ., Their distribution across the chromosomes is distinctly nonrandom ., As seen in Table 5 , the block size of the X chromosomes is considerably larger than for the autosomes , and similarly within the autosomes , the block size in the centers is much larger than the arms ., The median for the autosomes is similar to that obtained in 17 , whereas the median for the X is considerably larger , perhaps because of the greater continuity of the sequence in our study ., Given the overwhelming tendency of orthologs to remain on the same chromosome , we investigated the nonsyntenic ortholog pairs to see what features might distinguish them from syntenic pairs ., To minimize the likely contamination of the nonsyntenic set with misassemblies , we excluded 12 larger clusters of nonsyntenic orthologs ( see Materials and Methods ) ., The most distinctive difference between the two sets was the lower percent identity of the aligned nonsyntenic pairs ( Figure 6 ) ., These differences existed among pairs regardless of whether the members of the pair lay both on chromosome arms , both in chromosome centers , or one on an arm and one in the center ., One explanation for the greater divergence of the nonsyntenic ortholog pairs might be that the true ortholog is missing in the draft C . briggsae sequence ., We looked for evidence of this by finding the 1:1 orthologs ( e . g . , A and C ) flanking the C . elegans member of a nonsyntenic ortholog pair , ( ABC , where B is from the nonsyntenic pair ) and then searching the region between the C . briggsae orthologs of A and C for evidence of large gaps or partial genes ., Of 175 nonsyntenic ortholog pairs , we detected homology in the interval defined by the flanking orthologs for only 19 cases , and only 29 regions had 4% or more of the interval as uncalled bases ( Ns ) ., Almost half the intervals had less than 1% of the sequence as Ns ., Thus , while the draft nature of the C . briggsae sequence may result in incorrect assignment of 1:1 orthology , producing an apparent increased divergence , it seems unlikely to account for the bulk of our observations ., The comparison of random clone sequences from whole genome shotgun libraries from the Japanese ( HK104 ) and Hawaiian ( VT847 ) isolates with the genome assembly of AF16 provided in each case adequate numbers of widely distributed SNPs to develop markers across the genome assembly ., The sequence generated also provides the opportunity for more in-depth studies of patterns of variation among the different isolates ., In this study we have confined our analysis to the overall rates of differences , determined by the simple method of scoring base differences between aligned sequences with quality scores >35 ., With this quality score cutoff , errors should contribute a false SNP no more than one per 3 , 200 bases , and given that most bases have quality scores well above this , the contribution is likely to be much smaller ., Since the observed rates of difference considerably exceed this , errors will only slightly inflate the observed rates ., Indeed , of the more than 320 SNP assays that provided data , only five ( 1 . 5% ) were monomorphic ., The SNP rates we observed between these C . briggsae strains are higher than those observed between the most divergent C . elegans strains tested to date , with the HK104/AF16 differences about 8-fold higher and the VT847/AF16 differences about 3-fold higher than rates observed in similar experiments between N2 , the standard strain of C . elegans , and CB4856 , a strain from Hawaii that is among the most divergent strains yet isolated 18 , 19 ., The SNP rates we observed for both VT847 and HK104 compared with AF16 are similar to those reported by studies focused on a few genes 20 , 21 ., We also looked at regions of overlap of VT847 and HK104 sequences ( total 129 kb ) and noted that few differences were shared between the strains ., Similarly we observed that the HK104/AF16 SNP assays were predominantly monomorphic when assayed against VT847/AF16 RI lines ., These results are consistent with those of 21 , on studies of 4 . 2 kb of sequence from six genes ., The authors of 21 noted that strains from temperate regions across the globe , including HK104 , HK105 , and PB800 , had little diversity among themselves , but were more variant as groups from tropical strains , which include both AF16 ( India ) and VT847 ., In contrast to the temperate strains , the tropical strains contained considerable diversity ., These results suggest that the effective population size of C . briggsae may be several-fold larger than that observed for C . elegans ., Initial analysis suggests that the overall SNP rates may be greater on chromosome arms than in the centers ., However , the differences in gene density and other features between the chromosomal regions may contribute to the apparent rate differences ., A more careful parsing of the sequence reads among the features of the genome , a process now underway ( LW Hillier and RH Waterston , unpublished data ) , will be required to evaluate the different regions ., The placement of 248 markers onto six linkage groups is in accord with cytogenetic estimates of chromosome number 22 ., The observed length in centimorgans of the autosomes is consistent with the hypothesis that each chromosome undergoes one recombinant event per meiosis , as is thought to be the case for C . elegans ., However for CbX , the total length was only 34 cM ., Of course the present markers may not extend to the ends of the chromosome , although the X , at more than 20 Mb , is the largest of the chromosomes and had no additional assemblies assigned to it based on ortholog assignments ., Also , the two strains used to generate the RI lines might differ significantly in some regions in the genome , reducing recombination , e . g . , through an inversion ., If the X length is not artifactually short for one of the reasons given above , the genetic length of 34 cM would suggest that other mechanisms exist to ensure normal segregation of the X chromosome ., Such mechanisms must exist in males , which are XO , and might be operative in XX animals in C . briggsae ., Although the RI lines served adequately to generate the map , they had shortcomings that might be improved in future studies ., There was clearly biased recovery of some markers , with markers from the AF16 strain underrepresented on chromosome CbII and overrepresented on CbIV ., This bias might be readily corrected by a more-random selection of progeny to establish each line ., In addition , the RI lines had relatively few recombinant events ., As a result , central regions of low recombination often contain several successive markers at the same distance ., Strategies to establish lines that allowed several rounds of interbreeding would capture more events ., The long-range continuity of the genetic map served to order and in many cases orient more than 90 Mb of the sequence assemblies along the chromosomes ., Combining this with linking information from read-pairs and ortholog local colinearity , additional order and orientation of the contigs was provisionally imposed on the map ., By exploiting the conservation of synteny , another 9 Mb could be tentatively assigned to chromosomes , although not ordered along them ., The conflicts between the genetic map and sequence assembly exposed misassemblies in the whole genome assembly ., By carefully defining a set of 1:1 orthologous genes between the two species , the extensive conservation of synteny between the two species became more apparent and made clear that the problems lay in the assembly ., The analysis also suggests at least another five regions of potential misassembly , each spanning more than 79 kb with a cluster of ten or more orthologs matching to a nonsyntenic chromosome ., Smaller clusters of genes from nonsyntenic chromosomes also exist , but the fraction of these ( or indeed the larger clusters ) that represent assembly errors is uncertain ., Positioning markers within these regions and testing them against the RI lines should distinguish misassembly from rearrangements ., The integrated map revealed that organization into arms and centers for a number of features found in C . elegans is also present in C . briggsae ., These include the rates of recombination as a function of physical distance ( Marey maps ) , the distribution of repeats and exons and the size of introns ., Comparative analysis also shows a relative paucity of 1:1 orthologs in the arms as opposed to the centers , beyond that expected from the difference in exon density alone ., The maintenance of this distinctive organization over approximately 200 My of evolution , and despite numerous intrachromosomal inversion events , strongly supports the selective advantage this organization confers ., The enrichment for strongly conserved genes with yeast and for 1:1 orthologs in the centers suggests that genes are protected in this environment from the mutagenic effects of the high recombination and associated transposable element ( TE ) activity that is prevalent on the arms ., By contrast , the arms are enriched for rapidly evolving gene families , where recombination , higher mutation rates , and TEs may facilitate family expansion and rapid gene adaptation 23 ., The association between regions of higher recombination and more rapidly evolving genes has been reported in other species as well , including yeast 24 and Drosophila 25 , 26 ., The genome map revealed a striking degree of synteny conservation ., More than 95% of 1:1 orthologs remain on the same autosome despite the extensive evolutionary time since the MRCA ., For the X chromosome , the conservation is even greater , with about 97% of orthologs remaining syntenic in accord with theory 27 ., Even this may underestimate the extent of conservation , since misassemblies may still contribute to some of the nonsyntenic regions ., The conservation of synteny in worms does not reflect a lack of overall rearrangements , however , since hundreds of rearrangements have occurred intrachromosomally ., But within chromosomes , the observed breakpoints are not randomly distributed , with the block size much greater in the centers ., Nor is there substantial mixing of the centers with the arms ., The extensive conservation of synteny between C . elegans and C . briggsae may extend beyond the genus to more | Introduction, Results, Discussion, Materials and Methods, Supporting Information | To determine whether the distinctive features of Caenorhabditis elegans chromosomal organization are shared with the C . briggsae genome , we constructed a single nucleotide polymorphism–based genetic map to order and orient the whole genome shotgun assembly along the six C . briggsae chromosomes ., Although these species are of the same genus , their most recent common ancestor existed 80–110 million years ago , and thus they are more evolutionarily distant than , for example , human and mouse ., We found that , like C . elegans chromosomes , C . briggsae chromosomes exhibit high levels of recombination on the arms along with higher repeat density , a higher fraction of intronic sequence , and a lower fraction of exonic sequence compared with chromosome centers ., Despite extensive intrachromosomal rearrangements , 1:1 orthologs tend to remain in the same region of the chromosome , and colinear blocks of orthologs tend to be longer in chromosome centers compared with arms ., More strikingly , the two species show an almost complete conservation of synteny , with 1:1 orthologs present on a single chromosome in one species also found on a single chromosome in the other ., The conservation of both chromosomal organization and synteny between these two distantly related species suggests roles for chromosome organization in the fitness of an organism that are only poorly understood presently . | The importance of chromosomal organization in the fitness of a species is only poorly understood ., The publication of the C . elegans genome sequence in 1998 revealed features of higher level organization that suggested its chromosomes were organized into distinct domains ., Chromosome arms were accumulating changes more rapidly than the centers of chromosomes ., In this paper , we have compared the organization of the nematode C . briggsae genome with that of C . elegans ., By building a genetic map based on DNA variations between two strains of C . briggsae , and by using that map to organize the draft genome sequence of C . briggsae published in 2003 , we found the following: ( 1 ) Intrachromosomal rearrangements are frequent within and even between arms but are less common within central regions and between arms and centers ., ( 2 ) Genes have remained overwhelmingly on the same chromosomes ., ( 3 ) The distinctive features that distinguish C . elegans arms from centers also are seen in C . briggsae chromosomes ., The conservation of these features between these two species , despite the approximately 100 million years since their most recent common ancestor , provides clear evidence of the selective advantages of the domain architecture of chromosomes ., The continuing association of genes on the same chromosomes suggests that this may also be advantageous . | caenorhabditis, computational biology, molecular biology, genetics and genomics | The conservation of both chromosomal organization and synteny between two distantly related species suggests roles for chromosome organization in the fitness of an organism. |
journal.ppat.1005691 | 2,016 | Cooperation between Monocyte-Derived Cells and Lymphoid Cells in the Acute Response to a Bacterial Lung Pathogen | Innate immune responses in infected peripheral tissues are essential for controlling invading pathogens in the early phases of infection to prevent rapid pathogen replication and widespread dissemination ., Despite this vital role , the main cells and factors that control innate immune responses in tissues are poorly defined ., In particular , the innate functions of dendritic cells ( DC ) in peripheral tissues are not well understood compared to their role as antigen presenting cells in lymphoid organs and the significance of tissue-borne lymphoid cells in peripheral innate immunity has been recognized only recently ., Components of the innate immune response to pathogens have mostly been studied in isolation and there are few examples where the interplay between distinct innate components that mediate pathogen clearance in vivo is well understood ., Ly6Chi or “classical” monocytes are circulating mononuclear cells that rapidly enter inflamed tissues upon insult or infection ., Here , the cells can mediate effector function whilst maintaining an undifferentiated phenotype 1 , or undergo terminal differentiation upon which a proportion lose expression of Ly6C 2 ., Monocyte derivatives can contribute functions that are otherwise associated with either macrophages or DC 3–5 , which has led to monocyte-derived cells being referred to as monocyte-derived DC 2 , 5 , 6 or inflammatory monocytes/macrophages 7 , 8 ., Since the exact developmental origins and functions of differentiated monocytes in inflammatory sites is usually unclear a recent proposal suggests the term monocyte-derived cell ( MC ) 5 , which we have adopted here ., To gain an integrated understanding of the in vivo innate immune network in lung tissue , here we investigated the acute response to respiratory infection with the intracellular bacterial pathogen Legionella pneumophila ., L . pneumophila is an opportunistic human pathogen and the causative agent of Legionnaires’ Disease , an acute form of pneumonia associated with high rates of morbidity and mortality 9 ., Following inhalation into the lung , L . pneumophila replicates in alveolar macrophages within an intracellular vacuole that evades fusion with the endocytic pathway 10 , 11 ., Host resistance to L . pneumophila in mice requires a rapid inflammatory response in tissue that combats bacterial replication and is stimulated by innate pattern-recognition receptors 12–14 ., This is followed by an adaptive immune response mediated by T and B cells that begins ~5 days after infection 12 , 13 , 15 ., The innate response to L . pneumophila is greatly compromised in the absence of effector cytokines such as IFNγ 16 , 17 , although the cellular targets of IFNγ have not been defined ., Many studies have focussed on the role of macrophages in early immune responses to L . pneumophila , while the function of conventional DC ( cDC ) and MC have not been studied in detail ., While facets of the cellular immune response have been investigated previously 12 , 13 , 15 , 16 , 18–22 , only recently has an analysis of the temporal kinetics of lung phagocytes and lymphocytes present in the acute stages of L . pneumophila infection been made 20 ., Here we applied the methodology of Lambrecht and colleagues 2 and specifically separated neutrophils , AM , cDC and the monocytic compartment , and found that neutrophils and MC were the major phagocytic cell types that interact with L . pneumophila early in infection ., While MC are widely recognised as infiltrating inflamed tissue , the significance of their role during lung infection is only partially understood in part due to the difficulty of delineating MC from other DC types 2 , 23 ., We observed that MC were recruited rapidly during lung infection ., MC were required for optimal bacterial clearance by instructing lymphoid cells to produce the key cytokine IFNγ , which in turn activated the bactericidal activity of MC ., This work demonstrates that MC play a key immunoregulatory and protective role during pulmonary bacterial infection and also helps to define a specific role for IFNγ ., Recent work 2 , 23 showing that expression of Fcε receptor I and CD64 can be used to differentiate MC and cDC has allowed an accurate appraisal of the monocytic cells that invade the lung ., We used these approaches to characterise the time course of phagocyte recruitment during the early immune response to L . pneumophila ., Using a gating strategy to identify neutrophils , alveolar macrophages ( AM ) and DC types in lung ( Fig 1A ) , the total numbers of phagocytic cells per lung before and after infection were determined ( Fig 1B ) ., As expected , lung AM represented the major monocytic/phagocytic cell type in steady state ., Both CD11b+ and CD103+ cDC were detected but very few MC were present ., However , within 24 h of infection with L . pneumophila , neutrophils and MC became the dominant phagocytic cell types in the lung and by day 3 the numbers of MC were comparable to neutrophils ( Fig 1B ) ., In contrast , the number of neutrophils rapidly waned after day, 2 . Interestingly , the number of AM significantly decreased over the first 3 days of infection but then began to rebound at day 4 as the bacterial load lessened ., ( Note that bacterial number was not directly ascertained in these experiments . However in similar experiments we found that bacterial number peaks at day 1–2 and we have previously published that the number of Legionella+ neutrophils , which peaks at day 2 , is a reliable indication of bacterial load 24 . ) cDC increased in number and rivalled AM in abundance by day 3 ( Fig 1B ) ., Thus , contrasting with the steady state levels , MC , neutrophils and cDC outnumbered AM in the lung during acute L . pneumophila infection ., Using an anti-L ., pneumophila antibody 24 , we observed that large proportions of AM , neutrophils and MC had phagocytosed bacteria or material derived from the bacteria ( Fig 1C shows flow cytometric plots at 2 days after infection , Fig 1D shows enumeration of the number of antibody stained cells per lung ) ., At 24 h the vast majority of L . pneumophila staining occurred in neutrophils ., However , by day 2 MC staining for L . pneumophila was comparable to neutrophils and remained at high levels through days 3–7 ., The number of L . pneumophila-containing MC and neutrophils was >10-fold higher than AM at days 2 and, 3 . cDC also stained with the anti-L ., pneumophila antibody although they represented less than 1% of total L . pneumophila-positive cells at all time points ., This comprehensive analysis of phagocyte recruitment showed that neutrophils and MC were the dominant phagocytic cells present after L . pneumophila infection and were associated with bacterial material ., In contrast , AM numbers rapidly decreased in the first 3 days of infection and became a minor L . pneumophila-associated cell type ., To gauge the functional importance of MC in combatting L . pneumophila infection , mice deficient in the C-C chemokine receptor 2 ( CCR2-/- mice 6 , 25 ) were infected and bacterial load was assessed ( Fig 2 ) ., Monocytes in CCR2-/- mice have an impaired ability to exit the bone marrow and , hence , to infiltrate tissue and convert to MC 25 ., We observed that MC were significantly reduced in the lungs of CCR2-/- mice infected with L . pneumophila at days 3 and 5 ( Fig 2A and 2B ) , and this was associated with a ~10-20-fold increase in bacterial burden 3 and 5 days after infection ( Fig 2C ) ., Of the other inflammatory cell types , neutrophils were significantly increased on day 3 and 5 in CCR2-/- mice , which would not be expected to contribute to an increase in bacteria , and CD11b+ cDC showed a small decrease at day, 3 . No other significant differences were found ( S1 Fig ) ., To understand the impact of reduced MC numbers during infection with L . pneumophila , we analysed cytokine profiles in the bronchoalveolar lavage fluid ( BALF ) of CCR2-/- mice ., The levels of IFNγ , a cytokine known to be important for resistance to L . pneumophila infection 16 , 17 were greatly reduced in the BALF of CCR2-/- mice ( Fig 3A ) ., To resolve the factors that drive the secretion of IFNγ , we focussed on IL12 , which is a known inducer of IFNγ 26We found that expression of mRNAs for the two subunits of IL12p70 27 IL12p35 and IL12p40 , were induced in the lung 24 h after L . pneumophila infection and peaked at day 2 ( Fig 3B and 3C ) ., In CCR2-/- mice , we found that IL12p70 was not detectable in the BALF upon L . pneumophila infection ( Fig 3D ) , suggesting that MC were a major source of IL12p70 ., To further characterise IL12 expression , we infected IL12p40-enhanced yellow fluorescent protein ( IL12p40-YFP ) reporter mice 28 with L . pneumophila ., By staining in vitro stimulated DC with anti-IL12p40 antibodies , we confirmed that these mice faithfully report IL12p40 expression ( S2A Fig ) ., Although CD103+ cDC , CD11b+ cDC and MC produced some IL12p40 ( Fig 3E and 3F ) , MC predominated in the response ( Fig 3F ) , supporting the conclusion that MC were a major source of IL12 ., Surprisingly , AM did not produce detectable IL12p40 ( Fig 3E and 3F ) ., Consistent with IL12 playing a major role in the induction of IFNγ , IL12p35-deficient mice ( IL12p35-/- ) and IL12p40-deficient mice ( IL12p40-/- ) infected with L . pneumophila had greatly reduced levels of IFNγ in the BALF ( Fig 3G ) ., Hence , MC play a major role in clearance of L . pneumophila and are an important source of IL12 , which is largely responsible for inducing IFNγ after infection ( Figs 2 and 3 ) ., As IL18 also induces IFNγ secretion , we assessed the role of IL18 in the response to L . pneumophila ., IL18-/- mice contained reduced levels of IFNγ in BALF compared to C57BL/6 mice ( S3A Fig ) ., However , bacterial clearance in IL18-/- mice was not significantly altered ( S3B Fig ) suggesting that although reduced , the level of IFNγ produced in the absence of IL18 was still sufficient to stimulate optimal clearance of bacteria ., These data are in agreement with previous findings 16 , 21 , 29 ., To determine which phagocyte types support L . pneumophila replication in vivo and gauge their bactericidal activity , phagocytes were purified from the lungs of wild type mice 2 days after infection , lysed and plated on bacteriological plates for enumeration of L . pneumophila colonies ( CFU ) ( Fig 4A ) ., The number of viable bacteria recovered on a per cell basis was relatively high for AM but low in neutrophils , MC and cDCs , suggesting that AM are poorly bactericidal and constitute the major site for replication of L . pneumophila , which consistent with previous work 20 ., We also examined if bacterial survival in AM , neutrophils and MC was influenced by IFNγ by repeating the above experiment with cells isolated from infected lungs of wild type and IFNγ-/- mice ( Fig 4B ) ., While IFNγ deficiency made no difference to the number of viable bacteria isolated per cell from AM or neutrophils , 26-fold more viable bacteria were recovered from MC from IFNγ-/- mice compared to wild type mice ( Fig 4B ) ., Note that these differences were not due to alterations in bacterial load in lung tissue , because the number of viable bacteria in lungs of C57BL/6 mice and IFNγ-/- mice at 2 days after infection was not significantly different ( C57BL/6 , 1 . 22 x 106 ± 3 . 45 x 105; IFNγ-/- , 1 . 39 x 106 ± 2 . 6 x 105; n = 5 , NS ) ., These data suggest that AM are efficient replicative hosts of L . pneumophila in vivo even in the presence of IFNγ , a finding that differs from in vitro studies in which macrophage cell lines efficiently kill L . pneumophila following activation by IFNγ 30 The anti-bacterial activity of neutrophils does not require IFNγ ., In contrast to both AM and neutrophils , optimal control of bacterial numbers by MC required IFNγ stimulation thus defining a significant cellular target for this cytokine ., These data also suggest that MC contributed to L . pneumophila clearance both through their direct IFNγ-dependent bactericidal activity and their ability to stimulate IFNγ production via IL12 ., To identify the source of IFNγ during L . pneumophila infection , IFNγ-YFP reporter mice ( GREAT mice 31 ) were used to detect IFNγ-producing cells in lung ., We verified that these mice faithfully report IFNγ production using antibody staining ( S2B Fig ) ., At steady state , YFP expression was not detected in any cell type ( Fig 5 ) ., In mice infected with L . pneumophila 2 days earlier , IFNγ-YFP expression was found in NK cells and T cells and not any other cell type in lung ., While there have been previous reports of individual sub-types of T cells contributing to the innate response 32–39 , we examined the totality of the innate T cell response to L . pneumophila ., For T cells , CD8+ , NKT , CD4+ , CD4-CD8- ( DN ) and γδ cells all contributed to IFNγ production ( Fig 5A , upper panels ) ., The proportions of cells that produced IFNγ varied between cell types with ~30–80% of NKT and DN T cells producing IFNγ while ~ 5–15% of CD8+ , CD4+ and γδ T cells were IFNγ-YFP+ ., Less than 10% of DN T cells stained with an MR1-tetramer and thus were mucosal-associated invariant T cells , however less than 4% of these MR1-tetramer stained cells expressed IFNγ-YFP ., Quantitation of the number of IFNγ producing cells is shown in Fig 5B and 5C ., At 2 days after infection , IFNγ+ NK cells predominated but by day 3 the number of IFNγ producing NK and T cells were similar ., CD8+ T cells appeared to be the predominant T cell subset producing IFNγ particularly at day 3 ( Fig 5C ) ., Consistent with the findings discussed above , this IFNγ production depended on IL12 , as demonstrated by crossing the IFNγ-YFP mice onto an IL12p35-/- background ( IFNγ-YFP . IL12p35-/- , Fig 5A , 5D and 5E ) ., For all cell types IFNγ production was largely IL12-dependent , in agreement with results from IL12-/- and CCR2-/- mice ( Figs 2 and 3 ) ., We next examined if IFNγ-producing T cells found in the lung were naïve or had previously been activated by antigen ., Almost all T cells that were IFNγ+ also expressed CD44 ( Fig 5F ) , suggesting they were antigen-experienced and most likely memory T cells ., ( Note that NKT cells and γδ T cells constitutively express CD44 ) ., It seems likely that these CD44+ T cells were infiltrating circulating T effector/memory cells rather than T resident memory ( Trm ) cells as less than 3% of the IFNγ-secreting T cells expressed the typical Trm markers CD103 and CD69 40 ., The conclusion that the memory T cells infiltrated the lung subsequent to infection rather than being tissue resident was supported by analysis of the number CD44+ T cells in tissue over time ., CD44+TCRβ+ T cells were found at very low levels in steady-state lung but after infection infiltrated the tissue ( Fig 5G ) ., The same was true for NK cells and γδ T cells ., The high proportion of T cells expressing IFNγ and the rapidity of this response led us to suspect that T cells were being stimulated independently of TCR engagement ( non-cognate stimulation ) ., To determine if expression of IFNγ required TCR stimulation , CD8+ T cells from a mouse strain expressing a TCR specific for herpes simplex virus antigen gB and also containing the IFNγ-YFP reporter ( gBT-I . IFNγ-YFP mice ) were activated in vitro and seeded into wild type mice and allowed to convert to memory T cells over 30 days ., Approximately 90% of these cells initially produced IFNγ 4 days after activation with antigen in vitro , but after ~20 days in the wild type host , expression of the IFNγ reporter gene was undetectable and ~99% expressed CD44 indicating a memory T cell phenotype ., At 30 days after transfer the mice were infected with L . pneumophila ., The number of transferred gBT-I . IFNγ-YFP T cells that infiltrated the lungs of mice was comparable to that of endogenous T cells at 2–3 days after L . pneumophila infection ( Fig 6A ) and ~ 50–60% of these cells expressed YFP ( Fig 6B and 6C ) indicating that cognate MHC-peptide stimulation of the TCR was not required for activation of the memory T cells in the lung to drive IFNγ secretion ., Also of note , very few IFNγ-YFP+ cells were detectable in spleen indicating that stimulation was localized to the lung and not a systemic phenomenon ( Fig 6B ) ., To determine if non-cognate production of IFNγ could contribute to pathogen clearance , T cells from gBT-I mice or gBT-I mice on a IFNγ-/- background ( gBT-I . IFNγ-/- mice ) were activated with gB498-505 in vitro , transferred into IFNγ-/- mice and allowed to develop into memory T cells over 30 days ( Fig 6D ) ., Accordingly , the wild type gBT-I T cells were the only cells capable of producing IFNγ in these mice ., IFNγ-/- mice had a significantly higher bacterial load than wild type mice ( Fig 6E ) ., Three days after L . pneumophila infection , mice that had received IFNγ-sufficient gBT-I T cells had a bacterial load comparable to wild type mice while mice that received the gBT-I . IFNγ-/- T cells had a bacterial burden similar to that of IFNγ-/- mice without T cell transfer ( Fig 6E ) ., These data showed that non-cognate stimulation of T cells was sufficient to control bacterial load in the lung and that the production of IFNγ was required for this action ., The lung is a major site for potential infection by pathogens ., It is therefore important to gain an understanding of the inflammatory milieu in the lung following infection ., In this work we examined the cellular interplay following lung infection with L . pneumophila ., L . pneumophila is the causative agent of Legionnaires’ disease , a potentially fatal pneumonia that results from environmental exposure to the bacteria ., Upon entering the lung , L . pneumophila replicates inside alveolar macrophages ., Intracellular replication requires the bacterial Dot/Icm type IV secretion system to inject bacterial effector proteins into the macrophage cytosol , which establishes a ‘Legionella containing vacuole’ permissive for bacterial replication 10 , 11 , 41 ., Until recently , the inability to reliably identify inflammatory cell types , particularly myeloid cells , in lung has made it difficult to analyse cellular responses during infection ., However , newly identified marker sets allow accurate differentiation of lung macrophages , MC and DC subtypes 2 ., Here we applied these techniques to study responses to L . pneumophila lung infection in mice and quantitated the immune cell types recruited during acute L . pneumophila infection ., At 1 day after infection , and presumably hours before that time , AM represented a minor proportion of inflammatory phagocytic cells in the lung and by day 2 represented only ~ 1% of inflammatory phagocytes ., To identify which phagocytes had internalised L . pneumophila , cells were stained with an anti-L ., pneumophila antibody ., At day 2 and 3 after infection over 97% of stained cells were neutrophils or MC ., Similar to a recent study 20 , we found that much higher numbers of live bacteria could be recovered from lysed AM than other phagocytic cells in the lung , on a per cell basis ., This suggested that in vivo AM have a relatively poor ability to kill ingested L . pneumophila , even when stimulated by inflammatory cytokines found in the infected lung ., IFNγ did not influence the bactericidal activity of AM as the recovery of live bacteria was equivalent from AM isolated from wild type and IFNγ-/- mice ., This contrasts with in vitro studies in which macrophage cell lines efficiently kill L . pneumophila following activation by IFNγ 30 ., We also found that in contrast to other phagocytes , the number of AM rapidly decreased after infection ., At the peak of infection ( days 2 and 3 ) , AM levels were ~1/3 of those found prior to infection and numbers recovered as the bacterial burden waned ., AM arise from tissue-resident precursor cells seeded during embryogenesis and are not replenished by the ingress of myeloid cells 42 , 43 , so presumably this decrease is due to an increased rate of AM death that can not be compensated by an increased rate of in situ generation ., Regardless of the mechanism , a decrease in AM numbers during infection may contribute to control of L . pneumophila infection by limiting the replicative niches , as proposed for Salmonella enterica , serovar Typhimurium 44 ., Nogueira et al . 45 suggested that L . pneumophila infection may induce rapid apoptosis of DC to limit bacterial replication ., However , it is not clear to what extent such a mechanism occurs in vivo given the relative rarity of cDC and their poor ability to support L . pneumophila replication ., Additionally , death of AM may also contribute inflammatory mediators such as IL1β and death associated molecular patterns that initiate the inflammatory response ., Therefore , our work suggests that AM are probably amongst the first cell types to phagocytose L . pneumophila after lung infection and may well be pivotal in initiating the inflammatory response they appear to play a more minor role in the mass clearance of bacteria at the height of the acute infection ., MC develop in inflamed tissues from bone marrow-derived monocytes that flood into tissues in response to inflammatory signals 6 , 46 , 47 ., We found that MC accumulated in the L . pneumophila-infected lung more slowly than neutrophils , but by day 3 were present in numbers comparable to neutrophils ., MC have been shown to be involved in immunity to lung infections by Klebsiella pneumoniae 48 and Mycobacterium tuberculosis 8 , 49 as well as in the spleen after infection with Listeria monocytogenes 6 and Brucella melitensis 50 ., MC were highly phagocytic and by day 3 were the dominant population associated with L . pneumophila ., MC played a significant role in L . pneumophila clearance because mice with reduced infiltration of MC in lung ( CCR2-/- ) suffered significantly higher bacterial burden ., One mechanism by which MC contributed to L . pneumophila clearance is by their direct bactericidal activity , which has been documented for other bacterial species 6 , 46 ., In this work we found that optimal bacterial killing by MC required stimulation by IFNγ as lower levels of viable bacteria were recovered from MC isolated from infected wild type mice compared to infected IFNγ-/- mice ., In fact , numbers of viable bacteria in MC from IFNγ-/- mice were comparable to that isolated from AM ., This was in contrast to neutrophils where low numbers of viable bacteria were recovered from both wild type and IFNγ-/- neutrophils , indicating that the bactericidal pathways used by neutrophils were not dependent upon IFNγ stimulation ., MC appear to persist in lung for a longer time after infection than neutrophils with L . pneumophila+ MC present in lung beyond 7 days after infection ., The continued presence of MC may be important in maintaining immunity during the transition from the innate to the critical adaptive response ., Additionally , the immunoregulatory and antigen presentation activities of MC in the infected lung may play a role in the adaptive response for example by reactivating antigen specific T cells and inducing cytokine secretion 51 ., MC also played an important role in stimulating early production of the effector cytokine IFNγ in response to L . pneumophila ., CCR2-/- mice had very low levels of IFNγ and IL12 in lung and subsequent work showed that MC were major producers of IL12 in infected lung tissue ., While it was previously shown that NK cells produce IFNγ after L . pneumophila infection 16 , here we show that various T cell lineages including memory T cells as well as NKT and γδ T cells , also made significant contributions to IFNγ production ., Approximately 65% of IFNγ-producing T cells were ‘conventional’ TCRαβ memory T cells , the majority of which were CD8+ T cells ., These memory T cells made IFNγ very rapidly , and up to 80% of some T cell sub-types produced IFNγ in mice that had not previously seen L . pneumophila antigens ., This led us to conclude that the T cell stimulation did not require classical TCR-MHC-peptide engagement , in other words was the result of non-cognate stimulation ., To support this we found that mice seeded with memory T cells specific for an irrelevant antigen could produce a robust IFNγ response after L . pneumophila infection , thus confirming that TCR stimulation was not required for IFNγ production ., Furthermore , we demonstrated that IFNγ produced by non-cognate stimulation of T cells could effectively substitute for other IFNγ sources to enable optimal control of L . pneumophila infection ., Therefore , it appears that memory T cells can be considered bona fide members of the lymphoid armamentarium during acute responses ., In other studies non-cognate production of IFNγ by CD8+ T cells was shown to be induced by IL12 and IL18 in combination 36 , or IL18 alone 32 , by CD4+ T cells in response to IL18 and IL33 37 , 38 ., In our system IFNγ production was largely IL12-dependent but we cannot rule out the possibility that IL12 acts in concert with other molecules ., Indeed , we observed ~ 3-fold less IFNγ in lung after L . pneumophila infection of IL18-/- mice although the clearance of the bacteria was not influenced by a lack of IL18 , a result consistent with previous studies 21 , 29 ., We did not investigate if IL18 stimulated particular lymphoid cells to secrete IFNγ ., While there have been previous reports of sub-types of T cells contributing to the innate response 32–39 , this study examined the totality of the innate T cells response to L . pneumophila ., One novel finding here is the contribution of DN T cells to innate immunity ., While these cells were present at low total number in infected tissue , a very large proportion , ~45–80% , produced IFNγ ., The origin of these DN cells is uncertain ., A small proportion of the DN T cells stained with MR1 tetramer and are thus likely to be MAIT cells , but only very few of these cells expressed YFP ., DN T cells have also been shown to arise from self-reactive CD8+ T cells 52 and can rapidly produce inflammatory cytokines 53 and our findings here may indicate a hitherto unsuspected role for these cells in protective innate immunity ., Our work indicates that the sources of IFNγ are more numerous than previously thought and it is likely that there is some level of redundancy ., The relative contribution of each cell type probably depends on the relative abundance in the lung after infection that would be influenced by a number of circumstances such as previous immunological experience , the environment and the microbiota ., Based upon these and other studies , we propose the following model for the role and interactions of phagocytic and lymphoid cells in the acute phase of L . pneumophila infection ( Fig 7 ) ., Tissue resident phagocytic cells , namely , AM and conventional DC first engulf bacteria and produce inflammatory mediators such as cytokines and chemokines ., AM are rapidly depleted , and may release inflammatory death-associated signals to potentiate the immune response ., The decrease in AM may also act as a mechanism to limit L . pneumophila replication ., Neutrophils and monocytes infiltrate the lung early in the response to inflammatory stimuli whereupon neutrophils effectively engulf and kill bacteria without requiring activation by IFNγ ., Monocytes develop into mature MC in situ and become the dominant persistent phagocytic cell type ., MC contribute to bacterial clearance by production of IL12 , which in turn stimulates NK cells and various populations of memory T cells , NKT cells and γδ T cells locally to produce IFNγ ., IFNγ stimulates the bactericidal activity of MC , an activity that appears critical for optimal bacterial clearance ., Overall , our findings contribute to a growing body of knowledge on the events in infected tissues that contribute to immunity to pathogenic organisms ., A greater understanding of the cell types in infected tissues and their interplay may lead to an appreciation of the basis for sensitivity and resistance to pathogens and lead to more directed and effective therapies ., All mice were bred under specific pathogen-free conditions ., C57BL/6 were used as the wild type strain and all other strains were either created on a C57BL/6 background or had been backcrossed to C57BL/6 for at least 10 generations ., B6 . 129S4-Ccr2tm1Ifc ( CCR2-/- ) 25 , B6 . 129-Il12btm1Lky ( IL12p40-YFP ) 28 , B6 . 129S7-Ifngtm1Ts ( IFNγ-/- ) , B6 . 129S1-Il12atm1Jm ( IL12p35-/- ) , B6 . 129S1-Il12btm1Jm ( IL12p40-/- ) 54 , C . 129S4 ( B6 ) -Ifngtm3 . 1Lky ( GREAT , IFNγ-YFP ) 31 , Tg ( TcraHsv2 . 3 , TcrbHsv2 . 3 ) L118-1Cbn ( gBT-I ) 55 , B6 . 129P2-Il18tm1Aki 56 mice were used in this study ., L . pneumophila JR32 ΔflaA 19 was used for all experimental procedures in this study ., For animal infection , L . pneumophila was cultured under optimal conditions on selective buffered charcoal yeast extract ( BCYE ) agar ., Bacterial inoculum was generated by collecting colonies in PBS and adjusting via UV-spectroscopy ., In all experiments mice were administered 2 . 5x106 CFU in PBS via the intranasal route under controlled isoflurane induced anaesthesia ., To quantitate L . pneumophila in lung samples the right lobes were collected and homogenised in PBS , followed by lysis with 0 . 1% w/v saponin for 30 minutes at 37°C and L . pneumophila were enumerated by serially diluting the homogenate in PBS and plating onto selective BCYE ., Lungs were prepared for flow cytometry analysis as previous 24 ., Briefly , lung tissue was minced and digested by resuspension and gentle pipetting in RPMI-1640 ( Gibco ) with 3% v/v FCS ( Gibco ) , 1 mg/mL DNAseI ( Sigma Aldrich ) and 1 mg/mL Collagenase-III ( Worthington Biochemical ) ., Undigested material was filtered with 70 μm filters ( Corning ) to produce single cell suspensions ., Single cells were stained using antibodies and tetramers described in S1 Table ., Intracellular L . pneumophila staining was as described 24 ., Briefly , lung cells were fixed and permeabilised using the Fixation/Permeabilisation Kit ( eBioscience ) as per the manufacturers instructions , and cells stained using a polyclonal FITC-anti-Legionella antibody ( ViroStat ) ., Total numbers for each cell type were enumerated from the lung by addition of a known quantity of APC-labelled microspheres ( BD Calibrite ) to each sample prior to flow cytometry analysis ., BALF samples were obtained by injecting and recovering 1 . 5 mL chilled PBS into lungs and pelleting cells and debris ., The resulting supernatant was used to analyse cytokines and chemokines via a BD cytometric bead array flex kit as per the manufacturer’s instructions ., For qRT-PCR analysis , right lung tissue was collected into RNAlater ( Sigma ) , homogenised in TRIsure TRI-reagent ( Bioline ) and mRNA extracted via phase separation and precipitation using chloroform and isopronanol , respectively ., mRNA ( 4 μg ) was used for DNAse treatment and 1 μg pure mRNA was used for cDNA synthesis using an iScript cDNA synthesis kit ( Biorad ) as per the manufacturer’s instructions ., Primers for Il12a and IL12b were used in conjunction with SSOAdvanced Universal SYBR Green Supermix ( Biorad ) to quantitate relative levels of these genes in the lung ( See S1 Table ) ., qRT-PCR analyses was performed using a Quantstudio 7 Flex Real Time PCR System ( Applied Biosystems ) ., Cells were prepared and pooled from whole lungs of L . pneumophila infected C57BL/6 or IFNγ-/- mice as described ., Lung CD11c+ cells were enriched via positive selection with an automated magnetic bead separation device using anti-PE microbeads ( Miltenyi ) against CD11c-PE ., Neutrophils were obtained from the negative flow through fraction ., Cells were stained for flow sorting using a Beckton Dickinson MoFlo Astrios ., Sorted cells were lysed with 0 . 05% w/v digitonin ( Sigma ) and lysate plated on selective BCYE agar ., For adoptive transfer studies , CD8+ T cells were isolated from spleens of gBT-I . IFNγ-YFP or g | Introduction, Results, Discussion, Materials and Methods | Legionella pneumophila is the causative agent of Legionnaires’ disease , a potentially fatal lung infection ., Alveolar macrophages support intracellular replication of L . pneumophila , however the contributions of other immune cell types to bacterial killing during infection are unclear ., Here , we used recently described methods to characterise the major inflammatory cells in lung after acute respiratory infection of mice with L . pneumophila ., We observed that the numbers of alveolar macrophages rapidly decreased after infection coincident with a rapid infiltration of the lung by monocyte-derived cells ( MC ) , which , together with neutrophils , became the dominant inflammatory cells associated with the bacteria ., Using mice in which the ability of MC to infiltrate tissues is impaired it was found that MC were required for bacterial clearance and were the major source of IL12 ., IL12 was needed to induce IFNγ production by lymphoid cells including NK cells , memory T cells , NKT cells and γδ T cells ., Memory T cells that produced IFNγ appeared to be circulating effector/memory T cells that infiltrated the lung after infection ., IFNγ production by memory T cells was stimulated in an antigen-independent fashion and could effectively clear bacteria from the lung indicating that memory T cells are an important contributor to innate bacterial defence ., We also determined that a major function of IFNγ was to stimulate bactericidal activity of MC ., On the other hand , neutrophils did not require IFNγ to kill bacteria and alveolar macrophages remained poorly bactericidal even in the presence of IFNγ ., This work has revealed a cooperative innate immune circuit between lymphoid cells and MC that combats acute L . pneumophila infection and defines a specific role for IFNγ in anti-bacterial immunity . | Legionnaires’ Disease , a leading cause of community-acquired pneumonia resulting in significant morbidity and death , develops after infection with Legionella bacteria that replicate inside specialised sentinel cells of the lung ., Although some factors that help combat Legionella infection are known , an overall view of the early immune events that are triggered by infection were unclear and we have addressed this issue here using recently developed methods ., Our study implicates a number of new cells in the defence against Legionella infection and identifies key molecules that participate in a feedback circuit required for eradication of bacteria ., In particular , we find that specific immune cells derived from blood monocytes invade the infected lung and trigger other blood-derived cells to produce the potent inflammatory mediator IFNγ ., In turn IFNγ stimulates monocyte-derived cells to destroy bacteria ., Surprisingly , IFNγ did not influence the behaviour of other abundant immune cells ., The reported mechanism provides a basis for future investigation into the host response to combat intracellular bacteria , particularly in lung , and for assessing the risk to individuals infected with lung pathogens . | blood cells, flow cytometry, medicine and health sciences, immune cells, pathology and laboratory medicine, pathogens, immunology, microbiology, signs and symptoms, phagocytes, bacteria, neutrophils, bacterial pathogens, research and analysis methods, specimen preparation and treatment, staining, legionella pneumophila, white blood cells, inflammation, spectrum analysis techniques, memory t cells, animal cells, legionella, medical microbiology, microbial pathogens, t cells, immune response, spectrophotometry, cytophotometry, cell staining, diagnostic medicine, cell biology, biology and life sciences, cellular types, organisms | null |
journal.pgen.1003058 | 2,012 | RHOA Is a Modulator of the Cholesterol-Lowering Effects of Statin | Genome-wide association studies ( GWAS ) have been used to identify genetic contributors to a number of common diseases and traits 1 ., However , a major problem with this approach is that very large sample sizes are generally required to detect statistically significant associations 2 ., This is especially the case for pharmacogenomics , where identification of gene variants associated with drug response may require larger sample sizes than are generally available ., Consequently , GWAS has had limited success in the identification of pharmacogenetically relevant single nucleotide polymorphisms ( SNPs ) that survive the stringency of genome-wide multiple testing 3 , 4 ., In the largest single statin clinical trial GWAS published to date ( the JUPITER trial of ∼7000 individuals ) only three loci ( ABCG2 , APOE and LPA ) achieved genome-wide significance for association with the magnitude of LDL cholesterol reduction , and in total accounted for only a minor fraction of the overall variation in response 5 ., Moreover , GWAS studies are limited by their ability to probe only common genetic variation and thus the limited findings suggest that association studies alone are unlikely to yield the basis for all or even the majority of the genetic variance associated with drug response ., In the present report , we describe the use of transcriptome-wide profiling to identify and prioritize genes that may contribute to inter-individual variation in statin-induced plasma LDL-cholesterol lowering ., Statins inhibit HMG-CoA reductase ( HMGCR ) , the enzyme that catalyzes the rate limiting step of cholesterol biosynthesis , thus lowering intracellular cholesterol levels 6 ., This in turn elicits an increase in expression of cellular LDL receptors that mediate plasma LDL clearance 7 ., Since the HMGCR gene is transcriptionally regulated by intracellular sterol content 8 , the magnitude of induction of this gene is a cellular marker of in vitro statin response ., We used expression array data from in vitro statin-exposed immortalized human hepatoma cell lines and lymphoblastoid cell lines established from participants of the Cholesterol and Pharmacogenetics ( CAP ) clinical trial of simvastatin treatment 9 to establish a set of “biological rules” for identifying genes whose expression characteristics qualified them as having biologically plausible effects on cholesterol metabolism ., RHOA emerged from this analysis , and subsequent functional and genetic studies , as a novel candidate gene contributing to variation in LDL response to statins ., We used a series of filters applied to genome-wide gene expression data from 480 human lymphoblastoid cell lines ( LCLs ) derived from participants in the Cholesterol and Pharmacogenetics study to identify genes that appeared to be biologically plausible candidates for modulating the effects of statins on cholesterol metabolism ., The following filter criteria were used ( Table 1 ) :, 1 ) expression in normal human liver;, 2 ) change in transcript levels in HepG2 ( n\u200a=\u200a4 ) and Hep3B ( n\u200a=\u200a3 ) human hepatoma cell lines incubated with 2 . 0 µM activated simvastatin versus sham buffer for 24 hr , FDR<0 . 01 ,, 3 ) change in transcript levels in CAP LCLs incubated with 2 . 0 µM activated simvastatin versus sham buffer for 24 hr ( Q<0 . 05 ) ;, 4 ) consistent directionality of statin-induced change in transcripts in hepatoma cell lines and LCLs;, 5 ) correlation of statin-induced gene expression change in CAP LCLs with change in expression of HMGCR ., After Bonferroni correction for multiple testing ( p<1 . 17e-04 ) we identified 45 genes which passed all filter criteria ( Table 2 ) ., When ranked in order of correlation , only two of the top thirteen genes did not encode enzymes in the cholesterol biosynthesis pathway: transmembrane protein 97 ( TMEM97 ) and ras homolog gene family member A ( RHOA ) ., Although both had been previously implicated in lipid metabolism 10 , 11 , 12 , neither had been shown to play a role in the cholesterol lowering effects of statin ., However , RHOA was particularly intriguing since inhibition of RHOA signaling is thought to be a major mechanism by which statins exert pleiotropic ( or non-lipid lowering ) actions , such as anti-inflammatory effects ., Figure 1A demonstrates the strong correlation between statin-induced change in RHOA and HMGCR transcript levels ( p\u200a=\u200a7 . 64E-16 , r2\u200a=\u200a0 . 13 ) ., To determine if RHOA has a direct effect on markers of intracellular cholesterol homeostasis , we transfected HepG2 cells ( n\u200a=\u200a10 ) with siRNAs specific for RHOA or a non-targeting negative control and tested for changes in expression of HMGCR , low-density lipoprotein receptor ( LDLR ) and sterol response element binding transcription factor ( SREBF2 aka SREBP2 ) gene expression ., Knock-down reduced RHOA transcript levels by 60-98% ( Figure 1B ) , with no remaining detectable RHOA protein ( Figure 1C ) , and also generated statistically significant reductions in expression of HMGCR ( 0 . 76±0 . 04 fold , p\u200a=\u200a0 . 002 ) , SREBF2 ( 0 . 58±0 . 03 fold , p\u200a=\u200a0 . 0003 ) , and LDLR ( 0 . 73±0 . 13 fold , p\u200a=\u200a0 . 03 ) Figure 1D ., RHOA knockdown-mediated reductions in expression of HMGCR , LDLR , and SREBF2 were confirmed in a second hepatoma cell line , Huh7 ( n\u200a=\u200a6 ) ; however , the magnitude of the effect was less dramatic than that observed in the HepG2 transfections ., To further test the functional role of RHOA , we also measured levels of secreted APOB and APOA1 , the major proteins on LDL and HDL particles respectively , in the culture media 48 hours after knock-down ., APOB accumulation in the cell culture media was increased in HepG2 cells after RHOA knock-down ( 1 . 28±0 . 08 fold , p\u200a=\u200a0 . 03 , n\u200a=\u200a12 ) , while a similar but non-statistically significant trend was observed in Huh7 cells ( 1 . 08±0 . 06 fold , p\u200a=\u200a0 . 10 , n\u200a=\u200a8 ) , ( Figure 1D ) ., No significant changes in secreted levels of APOA1 were observed in either hepatoma cell line ., Reduced HMGCR , LDLR , and SREBF2 transcript levels together with increased APOB secretion with RHOA knock-down are all consistent with higher intracellular cholesterol levels , which was documented in the case of cholesterol esters ( 1 . 56±0 . 18 fold vs . controls , p\u200a=\u200a0 . 004 , n\u200a=\u200a16 ) , ( Figure 1E ) ., Although we also detected a trend for elevated free cholesterol after knock-down , this was not statistically significant ., A trend of increased intracellular cholesterol ester and free cholesterol was also observed in Huh7 cells after RHOA knock-down ( 1 . 15±0 . 14 fold , p\u200a=\u200a0 . 05 and 1 . 06±0 . 15 fold , p\u200a=\u200a0 . 27 , n\u200a=\u200a8 ) ., Lastly , since many genes involved in the maintenance of intracellular cholesterol are transcriptionally regulated in response to changes in intracellular sterol content through SREBF2 , a transcription factor , we sought to test if RHOA was also subject to SREBF2 regulation ., Sterol depletion activates SREBF2 , thus stimulating expression of SREBF2 target genes ., We confirmed that RHOA mRNA and protein levels were substantially increased by extreme sterol depletion in HepG2 cells with 2 µM simvastatin +10% lipoprotein deficient serum for 24 hr ( Figure 1F and 1G ) ., Induction of HMGCR mRNA and protein levels served as a positive control for the effects of cholesterol depletion ., Finally , we found small but statistically significant reductions in RHOA transcript levels after SREBF1 knock-down in HepG2 ( 0 . 83±0 . 07 fold , p\u200a=\u200a0 . 05 ) and Huh7 ( 0 . 86±0 . 02 fold , p\u200a=\u200a0 . 001 ) cell lines ( Figure 1H ) ., Although statin-induced changes of RHOA and LDLR mRNA were positively correlated in the LCL panel ( Figure 2A ) , change of the RHOA transcript was inversely correlated with level of LDLR cell surface protein ( Figure 2B ) ., Consistent with this relationship , we also identified an inverse correlation of RHOA transcript levels in statin- treated CAP LCLs with absolute changes in plasma total cholesterol ( p\u200a=\u200a0 . 02 , r2\u200a=\u200a0 . 01 ) , LDL cholesterol ( p\u200a=\u200a0 . 04 , r2\u200a=\u200a0 . 01 ) and APOB ( p\u200a=\u200a0 . 007 , r2\u200a=\u200a0 . 01 ) , measured in vivo before and after simvastatin treatment of the individuals from whom these cell lines were derived ( Table S1 ) ., In contrast , levels of RHOA in sham-treated LCLs were not significantly correlated with these measures at baseline ( Table S1 ) ., Moreover , RHOA transcript levels in statin-treated LCLs were not significantly associated with statin-induced changes in plasma HDL cholesterol levels ( data not shown ) ., We next investigated the association of common genetic variation near RHOA with in vivo statin response ., Analysis of HapMap3 CEU data 13 with Haploview 14 , revealed that RHOA fell within a large block of linkage disequilibrium spanning almost 500 kb and that there were four major haplotypes when considering markers within 10 kb of the gene , all with frequencies greater than 10% in the CEU population ( Figure S1; Table 3 ) ., Haplotypes were inferred based on directly genotyped SNPs ( Table S2 ) or imputed genotypes ( rs11716445 for H3B ) , and the number of copies of each haplotype were tested for association with change in LDL-cholesterol ( delta log ) in response to statin treatment of Caucasian participants in CAP ( n\u200a=\u200a580 ) and in the Pravastatin Inflammation CRP Evaluation ( PRINCE: pravastatin 40 mg/day , 24 weeks , n\u200a=\u200a1306 ) clinical trial , with adjustment for sex , age , BMI , smoking status , and study population ., Of the four haplotypes , H3B showed the strongest association with statin response ( p\u200a=\u200a0 . 01 ) , with homozygous H3B carriers having a 29% smaller reduction in the unadjusted percent change of LDL-cholesterol compared to non-carrier controls ( −21 . 8±4 . 5% versus −30 . 7±0 . 4% , Figure 3 ) ., When the CAP and PRINCE cohorts were analyzed independently , the directionality of this association was consistent between the two populations ( Figure S2 ) ., Haplotype H2 also demonstrated a modest association , with carriers having greater statin-induced changes in LDL-C ( p\u200a=\u200a0 . 04 , n\u200a=\u200a1886 , Figure 3 ) ., There were no significant associations of H3B or H2 carrier status with baseline LDL-cholesterol ( p\u200a=\u200a0 . 3 for both ) ., We found no association of either H3B or H2 with RHOA transcript levels in CAP LCLs after treatment with 2 µM statin or sham buffer ( n\u200a=\u200a115 ) ( Figure S3A ) ., However , rs11716445 , the SNP that defines the H3B haplotype , is located in a rare 45 bp cryptic exon ( referred to as RHOA exon 2 . 5 ) that we identified in multiple unique sequences during RNA-Seq analysis of three human hepatoma cell lines ( HepG2 , Hep3B and Huh7 ) , and CAP LCLs ( n\u200a=\u200a3 ) , Figure 4A ., Expression of the RHOA 2 . 5 exon was validated by Sanger sequencing ., Notably , we found that the H3B haplotype showed a very strong association with RHOA exon 2 . 5 levels under both sham ( p\u200a=\u200a2 . 7×10−7 , n\u200a=\u200a119 ) and statin ( p\u200a=\u200a9 . 1×10−13 , n\u200a=\u200a115 ) conditions , with carriers exhibiting the highest levels of exon 2 . 5 inclusion ( Figure 4B and Figure S3B ) ., The H2 haplotype also exhibited a more modest association with RHOA exon 2 . 5 levels in the opposite direction from H3B ( p<0 . 01 ) , consistent with their in vivo relationships ., Using Sanger sequencing , we found evidence of allele-specific expression ( ASE ) at rs11716445 with over 90% of the exon 2 . 5-containing transcripts originating from the H3B chromosome ( Figure 4C and 4D ) ., We also observed evidence of ASE at rs2878298 , another SNP found within exon 2 . 5 ( Figure 4D and Figure S3C ) ., There was no significant difference in the relative amount of ASE between the statin- and sham-treated states ., Finally , to determine if rs11716445 was a general expression quantitative trait locus ( eQTL ) or a splicing QTL , we tested for ASE at rs3448 , a SNP in the 3′ UTR of RHOA , in eight heterozygous carriers ( or H2/H3B ) and found no evidence that the ASE extended beyond exon 2 . 5 ( Figure 4D and Figure S3C ) ., These results strongly suggest that rs11716445 is a cis-acting splicing QTL ., We here present results of applying a set of biologically meaningful filters to identify and rank candidate genes associated with inter-individual variation in statin effects on cholesterol metabolism based on their gene expression characteristics ., Using unbiased genome-wide screens , we identified genes that were normally expressed in the human liver and changed in response to statin treatment in a manner that was correlated with statin-induced change in HMGCR quantified as an in vitro marker of statin response ., From these analyses we identified a number of genes not previously implicated in the lipid-lowering response to statin as potential candidates for future study ., We selected RHOA since many of the non-lipid lowering benefits , or pleiotropic effects , of statin treatment have been attributed to its ability to inhibit RHOA activity ., Our validation of RHOA as a modulator of cellular cholesterol metabolism , as well as the discovery that genetic variation within RHOA is associated with the magnitude of LDL-cholesterol response to statin treatment , support the continued studies of other novel candidate genes identified through this integrative genomics strategy ., RHOA has been previously implicated in cholesterol metabolism through the modulation of ABCA1-mediated cholesterol efflux via two distinct and opposing mechanisms ., RHOA inhibition stimulates ABCA1 gene expression via PPARγ and LXR activation 15 , while RHOA activation increases ABCA1 protein stability 11 ., Although excess intracellular levels of free cholesterol have been shown to increase RHOA activity 16 , here we demonstrate that RHOA knock-down results in increased levels of secreted APOB , suggesting that RHOA may influence the pool of intracellular cholesterol available for lipoprotein production ., Consistent with this hypothesis , we found that knock-down of RHOA in hepatoma cell lines resulted in increased intracellular content of cholesterol esters , the storage form of cellular cholesterol that can be mobilized for lipoprotein secretion ., This occurred in conjunction with reduced expression of HMGCR and LDLR , presumably due to cholesterol-induced down-regulation of SREBF2 ., Very recently a novel protein , LAMTOR1 ( also known as Pdro/p27RF-Rho ) , was found to both activate RHOA 17 , and to regulate LDL-C uptake and intracellular cholesterol egress from the late endosome/lysosome 10 , further supporting a link between RHOA and cholesterol metabolism ., Additional evidence for such a link is provided by the strong correlations that we observed between statin-induced changes in RHOA mRNA levels and both HMGCR and LDLR transcripts ., On the other hand , there was an inverse correlation between change in RHOA mRNA and cell surface LDLR protein ., While this may appear to be at odds with the change in LDLR transcript level , it is consistent with our finding that greater statin-induced RHOA gene expression was associated with reduced in vivo response lipid response to statin treatment ., It is possible that increased RHOA expression directly or indirectly reduces functional LDLR at the cell surface by altering post-translational processing or cellular trafficking , hypotheses that will be tested in future studies ., Increased magnitude of this effect may contribute to attenuation of statin-induced plasma LDL lowering ., Based on its role in mediating the pleiotropic effects of statin response , RHOA has been proposed as a candidate gene for the study of statin pharmacogenetics; however , genetic variation within RHOA associated with statin response has not been previously identified 18 ., Here , we report that a common RHOA haplotype , H3B , is associated with reduced LDL-cholesterol lowering in response to statin treatment in data derived from two independent clinical trials ., Within RHOA , this haplotype was defined by a single SNP , rs11716445; however , since rs11716445 is in strong linkage disequilibrium with many SNPs in other genes up to 500 kb away from RHOA , it is possible that its association with statin response may also be due to genetic variation affecting other genes ., rs11716445 explained less than 1% of the overall variation in LDL cholesterol response to statin , so neither the H3B haplotype or rs11716445 genotype alone would be a clinically useful diagnostic , but it could be included with other known markers of statin response to improve prediction algorithms ., Here we demonstrate that rs11716445 is a cis-acting splicing QTL also associated with allele-specific expression of RHOA exon 2 . 5 , a rare exon found within RHOA intron 2 ., The presence of this exon does not disrupt the open reading frame and is predicted to cause a 14 amino acid inclusion in the B3 domain of the RHOA protein , a region with no known interactions 19 , 20 ., Although the functional impact of exon 2 . 5 inclusion is unknown , the fact that the two RHOA haplotypes associated with its expression levels , H3B and H2 , are also the only two RHOA haplotypes found to be associated with in vivo variation in statin-induced change in LDL-cholesterol , strongly supports the likelihood that RHOA alternative splicing is functionally relevant ., In silico analysis with ESEfinder 3 . 0 identified SRSF2 ( aka SC35 ) , SRSF5 ( aka SRp40 ) , and SRSF1 ( aka SF2/ASF ) binding sites within 20 bp of the exon 2 . 5 splice donor 21 ., Notably , the rs11716445 “T” allele is predicted to disrupt an SRSF5 binding motif ( TAGAT/CC ) ( Figure S4 ) ., This finding is consistent with previous reports demonstrating that SRSF5 and SRSF2 antagonize SRSF1 to promote exon exclusion , as the loss of the SRSF5 binding with the “T” allele would be predicted to result in exon 2 . 5 inclusion 22 ., Thus , these results strongly suggest that the rs11716445 “T” ( minor ) allele enhances expression of the RHOA 2 . 5 exon ., We also found that the proportion of the expressed RHOA 2 . 5 exon containing the “T” allele was reduced in H1/H3B compared to H2/H3B and H3A/H3B heterozygotes ( Figure S3C ) ., Since the H1 haplotype contains the minor allele of the second common SNP within the 2 . 5 exon , rs2878298 , which is predicted to generate a SRSF1 binding site , these findings suggest that there are multiple gene variants that regulate expression of this novel exon; however the functional effects of these SNPs ( rs11716445 and rs2878298 ) as well as the expression of the cryptic RHOA exon remain to be tested ., In summary , we here report using a combination of expression array data , functional studies , and genetic analyses that RHOA is a novel candidate gene associated with variation in both in vitro and in vivo response to statin ., Although additional studies of statin effects will be required to corroborate these findings , they demonstrate the value of using data from a variety of molecular techniques , including the combination of in vivo and in vitro genetically-regulated phenotypes , as a novel approach for identifying genes involved in drug response ., Lymphoblastoid cell lines ( LCLs ) from 480 Caucasian participants from the Cholesterol and Pharmacogeneomics ( CAP ) clinical trial 9 and HepG2 and Hep3B cell lines were grown under standard conditions and exposed to 2 µM simvastatin or sham buffer for 24 hours as previously described 23 ., Although much higher than normal circulating levels of plasma simvastatin , 2–40 nM 24 , this concentration of simvastatin was selected based on previous dose response experiments that were used to determine the amount that elicits a consistent and significant induction of both HMGCR and LDLR mRNA ( Figure S5 ) 23 ., Simvastatin was provided by Merck Inc . ( Whitehouse Station , NJ ) and activated to the β-hydroxyacid prior to use 25 ., Cell surface LDLR protein was measured in statin and sham treated CAP LCLs as previously described 26 ., To confirm statin regulation , HepG2 cells were grown in 6-well plates and incubated with 2 µM activated simvastatin +10% lipoprotein deficient serum ( Hyclone ) for 24 hours ., Genome-wide gene expression was measured in RNA from CAP samples and statin and sham treated HepG2 and Hep3B cells ., RNA was converted to biotin-labeled cRNA using the Illumina TotalPrep-96 RNA amplification kit ( Applied Biosystems , Foster City , CA ) ., cRNA was hybridized to Illumina HumanRef8v3 expression beadchips ( Illumina , San Diego , CA ) ., Data were analyzed using GenomeStudio ( Illumina ) ., All beadchips had a signal P95/P05>10 ., Significance analysis of microarrays ( SAM ) 27 was performed on the 10 , 291 of 18 , 630 probed genes that were expressed in LCLs ( FDR<0 . 05 ) ., Expression traits were adjusted for known covariates ( age , gender , exposure batch , cell growth rate as determined by cell count on exposure day , and RNA labeling batch ) and also for unknown sources of variation through adjustment for those principal components that described greater than 5% variance across the dataset 28 ., Adjusted data were quantile normalized across each gene to ensure normality ., Gene expression in human liver was determined using mean detection p-value as determined by GenomeStudio ( Illumina , San Diego , CA ) from expression profiles measured by Illumina Ref8v2 beadarray on 120 human liver samples ( 2 technical replicates each of 60 samples , GEO accession number: GSE28893 29 ) ., Mean detection p-values across all 120 samples was assessed , and genes with a p<0 . 05 were called expressed ., HMGCR , LDLR , SREBF1 , SREBF2 , RHOA ( total ) , and RHOAexon2 . 5 transcript levels were quantified by qPCR with gene expression normalized to CLPTM ( TaqMan Assay number: Hs00171300_m1 , Life Technologies ) as previously described 30 ., Primers used for qPCR of total RHOA were F: CGGAATGATGAGCACACAAG and R: TGCCTTCTTCAGGTTTCACC and those used for qPCR of RHOA exon 2 . 5 were F: TATCGAGGTGGATGGAAAGC and R: GCCAACTCTACCATAGTACATTGAAA ., RHOA , SREBF1 and SREBF2 knock-down was achieved by 48 hour transfection of 80 , 000 HepG2 or Huh7 cells/well in 12-well plates using either the Ambion Silence Select siRNA ( s759 ) or non-targeting control according to the manufacturers protocol ., Cell culture media was collected from all samples at time of harvest , and APOB and APOAI were quantified in triplicate by sandwich-style ELISA ., Samples with a coefficient of variation greater than 15% were subject to repeat measurement ., Cholesterol was extracted from the cell pellets with hexane-isopropanol ( 3∶2 , v/v ) and dried under nitrogen ., The extracted cholesterol was reconstituted with reaction buffer ( 0 . 5 M potassium phosphate , pH 7 . 4 , 0 . 25 M NaCl , 25 mM cholic acid , 0 . 5% Triton X-100 ) ., Total cholesterol content was determined with the Amplex Red Cholesterol Assay Kit ( Invitrogen ) and normalized to total cellular protein quantified by the Pierce BCA Protein Assay Kit ( Thermo Scientific ) ., To quantify RHOA protein levels , cells were lysed in Cell Lytic lysis buffer ( Sigma ) , loaded on a 4–12% Tris-Glycine Gel ( Invitrogen ) , and proteins were transferred onto a PDVF membrane using the iBLOT gel transfer system ( Invitrogen ) ., The blot was then probed with antibodies diluted 1∶200 to RHOA ( SC26C4 ) , HMGCR ( SCH300 ) and β-actin ( SC ACTBD11B7 ) , all purchased from Santa Cruz Biotechnology ., Band densities were analyzed using the Mulitplex Band Analysis tool in Alphaview SA version 3 . 4 . 0 ., Haplotypes H1 , H2 , and H3A were assigned using genotype data from tag SNPs ( Table S2 ) , while haplotype H3B was inferred using imputed rs11716445 genotypes ., Imputation was performed in BIMBAM using 317K or 610K genotypes in a similar manner as previously described 31 except for use of the HapMap3 and 1KGP CEU pilot data as a reference population ., LDL-cholesterol was quantified in self-reported Caucasian American participants of the Cholesterol and Pharmacogenetics ( CAP ) clinical trial twice at baseline and after both 4 weeks and 6 weeks of simvastatin 40 mg/day and in the participants of the Pravastatin Inflammation and CRP Evaluation ( PRINCE ) clinical trial after 12 and 24 weeks of pravastatin 20 mg/day as previously described 9 , 32 ., Delta log LDL-cholesterol was calculated as the log average value of LDL-cholesterol on treatment minus the log average of the two baseline measurements , and percent change was the average on-statin value minus the average baseline value over the average baseline value ., The CAP trial is registered at ClinicalTrials . gov ( NCT00451828 ) ., Informed consent was obtained and approved by the institutional review boards of the sites of recruitment , University of California Los Angeles and San Francisco General Hospital ., In addition , all research involving human participants was approved by the Childrens Hospital Oakland Research Institute IRB ., All haplotypes with a minor allele frequency greater than 5% were identified using Haploview 14 with HapMap3 CEU data ., Using an additive genetic model , haplotypes were tested for association with change in delta log LDL-cholesterol using combined results of both clinical trials with adjustment for age , sex , BMI , smoking status , and study population as well as for each trial separately with adjustment for age , sex , BMI , and smoking status ., Hep3B , HepG2 , and Huh7 cells were incubated in duplicate under either standard growth conditions ( MEM supplemented with 10% FBS , 1% nonessential amino acids and 1% sodium pyruvate ) or sterol depleted conditions ( MEM supplemented with 1% nonessential amino acids , 1% sodium pyruvate , 2 . 0 µM simvastatin and 10% lipoprotein deficient serum ) for 24 hours ., RNA was extracted as previously described and samples from the duplicate experiments were pooled ., Sequencing libraries were prepared by isolating mRNA from 7–10 µg total RNA using two rounds of the MicroPoly ( A ) Purist kit ( Ambion ) , fragmenting the mRNA for 20 seconds , synthesizing cDNA using random primers , repairing ends , dA-tailing , ligating adapters , gel purifying fragments , amplifying libraries using indexed primers for 15 PCR cycles , and performing another round of gel purification ., Libraries were sequenced to an average depth of 60 million 100 bp reads ( 30 million paired-end fragments ) ., Expression of the novel RHOA exon was verified in independent samples through RT-PCR and Sanger sequencing ., DNA and RNA was isolated from CAP LCLs after 24 hours of exposure to sham buffer or 2 µM simvastatin ., The DNA sequences of exon 2 . 5 and exon 5 were amplified using F: CAAGGCAGGAGAATGGTGTG and R: CCACTGACGATGATTGCTTC and F: GGCCATATTACCCCTTTTCG and R: CCAGAGGGATCTAGGCTTCC , respectively ., RT-PCR was performed to amplify the transcript sequences of exon 2 . 5 and exon 5 ( 3′UTR ) using F: TCGTTAGTCCACGGTCTGGT and R: GCCAACTCTACCATAGTACATTGAAA and F: CGGAATGATGAGCACACAAG and R: TTGGAAAAATTAACTGGTACAGAAA , respectively ., PCR products were then subject to Sanger sequencing . | Introduction, Results, Discussion, Materials and Methods | Although statin drugs are generally efficacious for lowering plasma LDL-cholesterol levels , there is considerable variability in response ., To identify candidate genes that may contribute to this variation , we used an unbiased genome-wide filter approach that was applied to 10 , 149 genes expressed in immortalized lymphoblastoid cell lines ( LCLs ) derived from 480 participants of the Cholesterol and Pharmacogenomics ( CAP ) clinical trial of simvastatin ., The criteria for identification of candidates included genes whose statin-induced changes in expression were correlated with change in expression of HMGCR , a key regulator of cellular cholesterol metabolism and the target of statin inhibition ., This analysis yielded 45 genes , from which RHOA was selected for follow-up because it has been found to participate in mediating the pleiotropic but not the lipid-lowering effects of statin treatment ., RHOA knock-down in hepatoma cell lines reduced HMGCR , LDLR , and SREBF2 mRNA expression and increased intracellular cholesterol ester content as well as apolipoprotein B ( APOB ) concentrations in the conditioned media ., Furthermore , inter-individual variation in statin-induced RHOA mRNA expression measured in vitro in CAP LCLs was correlated with the changes in plasma total cholesterol , LDL-cholesterol , and APOB induced by simvastatin treatment ( 40 mg/d for 6 wk ) of the individuals from whom these cell lines were derived ., Moreover , the minor allele of rs11716445 , a SNP located in a novel cryptic RHOA exon , dramatically increased inclusion of the exon in RHOA transcripts during splicing and was associated with a smaller LDL-cholesterol reduction in response to statin treatment in 1 , 886 participants from the CAP and Pravastatin Inflamation and CRP Evaluation ( PRINCE; pravastatin 40 mg/d ) statin clinical trials ., Thus , an unbiased filter approach based on transcriptome-wide profiling identified RHOA as a gene contributing to variation in LDL-cholesterol response to statin , illustrating the power of this approach for identifying candidate genes involved in drug response phenotypes . | Statins , or HMG CoA reductase inhibitors , are widely used to lower plasma LDL-cholesterol levels as a means of reducing risk for cardiovascular disease ., We performed an unbiased genome-wide survey to identify novel candidate genes that may be involved in statin response using genome-wide mRNA expression analysis in a sequential filtering strategy to identify those most likely to be relevant to cholesterol metabolism based on their gene expression characteristics ., Among these , RHOA was selected for further functional study ., A role for this gene in the maintenance of intracellular cholesterol homeostasis was confirmed by knock-down in hepatoma cell lines ., In addition , statin-induced RHOA transcript levels measured in a panel of lymphoblastoid cell lines was correlated with statin-induced change in plasma LDL-cholesterol measured in individuals from whom the cell lines were derived ., Lastly , a cis-acting splicing QTL associated with expression of a rare cryptic RHOA exon was also associated with statin-induced changes in plasma LDLC levels ., This result exemplifies the power of applying biological information of well understood molecular pathways with genome-wide expression data for the identification of candidate genes that influence drug response . | rna interference, gene regulation, dna transcription, gene function, molecular genetics, personalized medicine, gene expression, gene splicing, biology, molecular biology, genotypes, phenotypes, heredity, gene identification and analysis, genetics, human genetics, molecular cell biology, genetics and genomics, complex traits | null |
journal.ppat.1003495 | 2,013 | Mutated and Bacteriophage T4 Nanoparticle Arrayed F1-V Immunogens from Yersinia pestis as Next Generation Plague Vaccines | Plague , also known as Black Death , is one of the deadliest infectious diseases known to mankind ., Yersinia pestis , the etiologic agent of plague , is a Gram-negative bacterium transmitted from rodents to humans via fleas 1 ., The bite of an infected flea results in bubonic plague which can then develop into secondary pneumonic plague , resulting in person-to-person transmission of the pathogen through infectious respiratory droplets 2 ., Pneumonic plague can also be caused by direct inhalation of the aerosolized Y . pestis , leading to near 100% death of infected individuals within 3–6 days 2 , 3 ., Due to its exceptional virulence and relative ease of cultivation , aerosolized Y . pestis poses one of the greatest threats for deliberate use as a biological weapon 4 ., Since the disease spreads rapidly , the window of time available for post-exposure therapeutics is very limited , usually 20–24 h after the appearance of symptoms 3 ., Although levofloxacin has recently been approved by the Food and Drug Administration ( FDA ) for all forms of plague ( http://www . fda . gov/NewsEvents/Newsroom/PressAnnouncements/ucm302220 . htm ) , prophylactic vaccination is one of the most effective means to reduce the risk of plague ., Stockpiling of an efficacious plague vaccine has been a national priority since the 2001 anthrax attacks but no vaccine has yet been licensed ., Previously , a killed whole cell ( KWC ) vaccine was in use in the United States , and a live attenuated plague vaccine ( EV76 ) is still in use in the states of former Soviet Union 5 ., However , the need for multiple immunizations , high reactogenicity , and insufficient protection made the KWC vaccine undesirable for mass vaccination , and , consequently , it was discontinued in the United States 6 ., In fact , the live-attenuated vaccine may not meet FDA approval because of the highly infectious nature of the plague bacterium and the virulence mechanisms of vaccine strains have not been fully understood 6 , 7 ., A cautionary tale related to this is the recent fatality of a researcher as a result of exposure to the attenuated pigmentation-minus Y . pestis strain , KIM/D27 ( http://en . wikipedia . org/wiki/Malcolm_Casadaban ., The focus in the past two decades , thus , has shifted to the development of recombinant subunit vaccines 3 , 6 , 8 , 9 containing two Y . pestis virulence factors , the capsular protein ( Caf1 or F1; 15 . 6 kDa , Figure 1A and B ) and the low calcium response V antigen ( LcrV or V; 37 . 2 kDa , Figure 1A and C ) , which is a component of the type 3 secretion system ( T3SS ) ., F1 assembles into flexible linear fibers via a chaperone/usher mechanism 10 , forming a capsular layer that allows Y . pestis to adhere to the host cell and escape phagocytosis 11 ( Figure 1A and 1B ) ., The V antigen forms a “pore” at the tip of the “injectisome” structure of the T3SS needle , creating a channel that delivers a range of virulence factors , also known as the Yersinia outer membrane proteins ( Yops ) , into the host cytosol ( Figure 1A ) 12 ., The V antigen is also critical for impairment of hosts phagocytic responses 13 ., Abrogation of these functions by F1 and V antibodies appears to be one of the mechanisms leading to protection of the host against lethal Y . pestis infection ., Two types of F1/V recombinant vaccines have been under investigation , one containing a mixture of F1 and V antigens 14 , and another , a single F1-V fusion protein 15 , 16 ., Although both induce protective immunity against Y . pestis challenge in rodent and cynomolgus macaque models , protection of African Green monkeys was insufficient and highly variable 6 , 17 ., A phase I clinical trial in humans showed that a vaccine consisting of a mixture of F1 and V proteins was immunogenic , however , the antibody titers varied over a wide range leading to concerns about the consistency of vaccine efficacy 18 ., One of the problems associated with the current plague vaccines is that the naturally fibrous F1 polymerizes into heterodisperse aggregates , compromising the quality and overall efficacy of the vaccines 15 , 19 , 20 , 21 , 22 ., Second , the subunit vaccines do not induce adequate cell-mediated immune responses that also appear to be essential for optimal protection against plague 23 ., Third , it is unclear if inclusion of other Y . pestis antigens such as the YscF , the structural unit of the injectisome needle ( Figure 1A and 1D ) , can boost the potency of the F1/V vaccines ., This is particularly important as F1-minus strains of Y . pestis exist in nature which are as virulent as the wild-type strains 24 , 25 and significant diversity in the LcrV sequence of these strains might render the current F1/V vaccines ineffective 26 , 27 ., Finally , the reported immunosuppressive property of V antigen 13 , 28 and whether it could compromise the innate immunity of humans , is a significant concern ., These questions must be addressed to generate a next generation plague vaccine that could pass licensing requirements , as well as be manufactured relatively easily for stockpiling ., Recently , we have developed a novel vaccine delivery system using the bacteriophage T4 nanoparticle 29 , 30 , 31 , 32 ., The T4 capsid ( head ) is an elongated icosahedron , 120 nm long and 86 nm wide , composed of three essential capsid proteins: a major capsid protein , gp23*; vertex protein , gp24*; and a portal protein , gp20 ( Figure 1E ) ., It is decorated with two non-essential proteins , Soc , the small outer capsid protein , and Hoc , the highly antigenic outer capsid protein ., Binding sites for these proteins appear following head “expansion , ” a major conformational change that increases the outer dimensions of the capsid by ∼15% and inner volume by ∼50% 33 ., Approximately 870 molecules of the tadpole- shaped Soc protein ( 9 kDa ) assemble into trimers at the quasi three-fold axes , clamping to adjacent capsomers and forming a reinforced cage around the shell ( Figure 1E ) 34 ., This stabilizes an already stable head that can withstand harsh extracellular environment ( e . g . , pH 11 ) 34 ., Hoc , on the other hand , is a linear “fiber” containing a string of four domains , three of which are immunoglobulin ( Ig ) -like 35 ., One hundred and fifty five copies of Hoc fibers , with their NH2-termini projected at ∼160 Å distance from the capsid assemble at the center of each capsomer ( Figure 1E ) ., Hoc binds to bacterial surfaces , apparently enriching the phage near its host for infection 36 ., Although Soc and Hoc provide survival advantages , they are completely dispensable under laboratory conditions showing no significant effect on phage productivity or infectivity 37 ., Purified Soc ( or Hoc ) protein binds to Hoc− Soc− capsid with high specificity and nanomolar affinity , properties that are not compromised by attachment of a pathogen antigen at the NH2- and COOH-termini 29 , 30 , 31 , 32 ., Individual domains , or full-length proteins as large as 90 kDa , or multilayered oligomeric complexes of >500 kDa fused to Soc can be arrayed on T4 capsid , making it a robust antigen delivery platform 29 , 30 ., Here , we describe two basic approaches to generate next generation plague vaccines , structure-based immunogen design and T4 nanoparticle delivery ( Figure 1 ) ., We designed an F1 mutant that retained the T cell epitopes but folded into a soluble monomer rather than into an insoluble fiber ( Figure 1B ) ., The mutated F1 was fused to V antigen to produce a bivalent F1mut-V immunogen that was also expressed as a soluble monomer ., We then constructed an oligomerization deficient YscF mutant ( Figure 1D ) as well as a V mutant without the putative immunomodulatory sequence ( Figure 1C ) ., The mutated antigens were fused to Soc and arrayed on phage T4 nanoparticle ( Figure 1F ) ., The F1mut-V monomer induced robust immunogenicity , and the T4-decorated F1mut-V without any adjuvant , in addition , induced balanced TH1 and TH2 responses ., Both the soluble and T4 decorated F1mut-V provided 100% protection to mice and rats against intranasal challenge with high doses of Y . pestis CO92 ., Inclusion of YscF showed a slight enhancement in the potency of F1-V plague vaccine , whereas replacement of V with V10 mutant , which lacks the putative immunosuppressive sequence , did not significantly alter vaccine efficacy ., These results provided new insights into plague vaccine design and produced next generation plague vaccine candidates by overcoming some of the concerns associated with the current subunit vaccines ., The X-ray structure and biochemical studies established that F1 polymerizes into a linear fiber by head to tail interlocking of F1 subunits through a donor strand complementation mechanism 10 ( Figure 1B ) ., Each subunit has an Ig-like domain consisting of a four-stranded anti-parallel β-sheet ., Of the four β-strands , three belong to one subunit forming a cleft into which the NH2-terminal β-strand of the “n+1” subunit locks in , resulting in a bridge connecting adjacent subunits ( inter-molecular complementation ) ( Figure 1B ) ., Stringing of subunits in this fashion leads to assembly of linear F1 fibers of varying lengths ., Caf1M chaperone is required for this process because prior to filling the cleft , a “spare” β-strand of Caf1M temporarily occupies the cleft until it is replaced by the β-strand of the incoming subunit with the assistance of an outer membrane usher protein , Caf1A ., Over-expression of the F1 gene in a heterologous system such as E . coli ( Figure 2 ) exposes the unfilled hydrophobic cleft , resulting in uncontrolled aggregation of F1 subunits into insoluble inclusion bodies ., This is demonstrated in Figure 2B in which all of the over-produced F1 protein partitioned into the pellet ( lane 8 ) and none was detected in the supernatant ( lane 7 ) ., Denaturation of the insoluble protein recovered some of the F1 protein into the soluble fraction but it still aggregated rapidly leading to precipitation ( in the Histrap column ) upon removal of the denaturant ., Similar aggregation behavior of F1 was observed in previously published studies 38 , 39 ., We hypothesized that shifting of the NH2-terminal β-strand of F1 to the COOH-terminus should reorient the β-strand such that it fills its own cleft ( intra-molecular complementation ( Figure 1B ) , and furthermore , it should no longer require the assistance of chaperone or usher proteins ., To test this hypothesis , we constructed an F1 mutant ( F1mut1 ) by deleting the NH2-terminal donor strand amino acid ( aa ) residues 1–14 and fusing it to the COOH-terminus with a short ( Serine-Alanine ) linker in between ( Figure 2A ) ., The recombinant F1mut1 , as predicted , folded into a soluble protein in the absence of Caf1M or Caf1A , and approximately 70% of the protein partitioned into the cell-free lysate ( Figure 2B , lanes 9 and 10 ) ., In addition , for reasons unknown , the mutated F1 protein was expressed at significantly higher levels than that of the native F1 protein after IPTG induction ( Figure 2B , compare lane 5 with lane 3 ) ., The gel filtration profile showed that the F1mut1 eluted as a symmetrical peak corresponding to a molecular mass of ∼19 kDa ( Figure 2C ) , a monomer , suggesting that the interlocking mechanism had shifted from inter- to intra-molecular interactions ., A bioinformatics approach was used to determine if the strand shifting might have disrupted the NH2-terminal epitopes of F1 ., The aa residues 7 to 20 are reported to contain a mouse H-2-IAd restricted CD4+ T cell epitope 40 ., Of the fifty-three predicted 9-mer CD8+ T cell epitopes that encompassed 46 human MHC-I alleles ( Table S1 in Text S1 ) , four peptides ( aa residues: 9–17 , 10–18 , 11–19 and 13–21 ) fell in this region , and of the 9 peptides predicted to contain CD4+ T cell epitopes ( Table S2 in Text S1 ) , only one ( aa residues 1–18 ) belonged to this region ., We determined that the integrity of these potential linear epitopes could be restored by extending the sequence of the switched strand by up to the aa residue 21 , which would duplicate the residues 15 to 21 at the COOH-terminus ., Thus , the F1mut2 was constructed ( Figure 2A ) and tested ., The F1mut2 behaved in a similar manner as the F1mut1 with respect to over-production and solubility ( Figure 2B , lanes 12–15 ) and was also purified as a monomer ( data not shown ) ., Fusion of F1mut2 to V would generate a bivalent plague vaccine ., Consequently , a mutated F1-V fusion protein ( F1mut-V ) was produced by fusing F1mut2 to the NH2-terminus of V with a two aa linker in between ( Figure 3A ) , and its solubility was compared to that of the native polymeric F1-V ., The native F1-V protein , as reported previously 20 , 22 , was insoluble and partitioned into inclusion bodies ( Figure 3B; lanes 5 and 6 ) ., Denaturation and refolding solubilized some of the protein but it also eluted , as was reported previously 20 , over a wide range of high molecular weight sizes in a gel filtration column ( Figure 3C , red profile ) ., F1mut-V protein , on the other hand , was nearly 100% soluble ( Figure 3B; lanes 7 and 8 ) and eluted as a symmetrical peak corresponding to a molecular weight of ∼64 kDa , equivalent to the mass of monomeric F1mut-V fusion protein ( Figure 3C , blue profile ) ., The yield of F1mut-V was quite high , ∼20 mg pure protein per liter of the E . coli culture ., Furthermore , its stability to trypsin digestion was similar to that of the native F1-V ( Figure 3D ) ., The Y . pestis V antigen has been reported to induce interleukin ( IL ) -10 and suppress the production of pro-inflammatory cytokines such as interferon ( IFN ) -γ and tumor necrosis factor ( TNF ) -α , which could lead to lowering of innate immunity in vaccinated individuals 41 ., A truncated V in which the COOH-terminal 30 aa residues ( 271–300 ) were deleted ( referred to as “V10” mutation ) was reported to lack this immunomodulatory function 41 ., A mutated F1mut-V10 recombinant was therefore constructed by deleting these residues ( Figure 3A ) ., This mutant protein was also over-produced in E . coli , which was also highly soluble and could be purified as a monomer ( Figure 3C , green profile ) ., Inclusion of YscF might expand the breadth of efficacy of F1-V plague vaccine formulation to Y . pestis strains containing variant V antigens 26 , or of those strains devoid of capsule but highly virulent in nature 24 , 25 ., Since YscF is a structural component of the injectisome , over-production of this protein caused aggregation 42 ., A mutant YscF was constructed by mutating the aa residues Asn35 and Ile67 , that are involved in oligomerization ( Asn35 changed to Ser , and Ile67 changed to Thr ) ( Figure 4A ) 43 ., The resultant YscF35/67 mutant protein was soluble and the gel filtration profile showed two peaks , a high molecular weight aggregate near the void volume , and a second peak corresponding to a molecular mass of ∼22 kDa , which is equivalent to a dimer ( Figure 4B , blue profile; C ) ., The native YscF , on the other hand , eluted over a wide range of high molecular weight sizes consistent with the formation of heterodisperse aggregates ( Figure 4B , red profile ) ., The mutant dimer did , however , show slow aggregation during concentration and storage , as evident by the appearance of small amounts of precipitates ., A large number of F1 , V , F1-V , and YscF recombinant proteins , both in native and mutated forms , were fused to the NH2- and/or the COOH-termini of either phage T4 Soc or the T4-related phage RB69 Soc and screened for their solubility as well as ability to bind to T4 phage ( Figure 5A , and data not shown ) ., Our previous studies showed that the RB69 Soc binds to T4 capsid at nearly the same affinity as T4 Soc 34 ., The RB69 Soc-fused plague antigens , with the exception of the native F1-Soc , produced soluble proteins whereas the T4 Soc-fused antigens were insoluble ., Several of the RB69 immunogens were purified ( Figure 5B ) and tested for binding to T4 using our previously established in vitro assembly system ., A typical result is shown in Figure 5C and D , which also exemplifies the versatility of the T4 nanoparticle display ., Consistent with the crystal structure of Soc , which showed that both the NH2- and COOH-termini are exposed on the capsid surface , the plague immunogens F1mut and V could be efficiently displayed as an F1mut-V fusion protein that in turn was fused to the NH2-terminus of Soc ( Figure 5C ) ., At the same time , its COOH-terminus could be fused to YscF35/67 , and the resultant F1mut-V-Soc-YscF35/67 fusion protein containing all three plague immunogens could be displayed on T4 capsid ( Figure 5G , lane 4 ) ., The 66 kDa F1mut-V-Soc bound to T4 even at a relatively low 1∶1 ratio of F1mut-V-Soc molecules to Soc binding sites ( Figure 5C , red arrow ) ., Binding increased with increasing ratio and reached saturation at 20–30∶1 ., The copy number of bound F1mut-V-Soc per capsid ( Bmax ) was 663 , which meant that ∼76% of the Soc binding sites were occupied , and its apparent binding affinity ( Kd ) was 292 nM , which was ∼4-fold lower than that of Soc binding ( Kd\u200a=\u200a75 nM ) 34 ( Figure 5D ) ., This is consistent with the expectation that the 66 kDa F1mut-V-Soc , unlike the 10 kDa Soc , would encounter steric constraints to occupy all the binding sites on the capsid exterior ., Given this limitation , the observed copy number was remarkably high , with the capsid surface presumably tightly packed with the F1mut-V molecules ( model shown in Figure 1F ) and exposing , consequently , the plague antigen epitopes for presentation to the immune system ., Indeed , cryo-electron microscopy showed that these T4 capsids , unlike the wild-type capsids ( Figure 5E ) , were decorated with a layer of F1mut-V molecules , seen as fuzzy protrusions around the perimeter of the capsid wall ( Figure 5F ) ., A series of nanoparticle decorated plague immunogens were prepared , including all three plague immunogens displayed on the same capsid using the F1mut-V-Soc-YscF35/67 fusion protein ( Figure 5G , lane 4 ) ., The immunogenicity of mutated F1 immunogen was tested in a mouse model ., The animals ( Balb/c ) were immunized according to the scheme shown in Figure 6 A and B and antibody titers in the sera were determined by ELISA ., The data showed that all the three plague antigens adjuvanted with alhydrogel induced antigen-specific antibodies ( Figure 6C ) ., The V antigen induced the highest titers with the end point titer reaching as high as 7×106 ., The YscF antigen was the least immunogenic ( Figure 6C , panel III ) , with the endpoint titers about 1–2 orders of magnitude lower than that of F1 and V antigens ( Figure 6C , panels I and II ) ., No significant differences in F1-specific antibody titers were observed among the various groups ( i . e . , F1-V versus F1mut-V versus F1-V+YscF; panel II ) ., Importantly , the monomeric F1mut-V induced comparable antibody titers as the native polymeric F1-V , suggesting that the capsular structure of F1 per se did not afford a significant advantage to induction of antibodies ., However , unexpectedly , the V-specific IgG titers were at least an order of magnitude higher when YscF was also included in the vaccine ( p<0 . 001 ) ( Figure 6C , panel I; compare F1-V to F1-V+YscF ) ., Intranasal challenge of animals with 90 LD50 of Y . pestis CO92 1 LD50\u200a=\u200a100 colony forming units ( CFU ) in Balb/c mice , one of the most lethal strains , showed that all the control mice died by day, 3 . However , the mice immunized with native V immunogen showed 83% survival ( two of twelve mice died ) , whereas the mice immunized with F1-V , F1mut-V , or F1-V plus YscF were 100% protected ( Figure 6D ) ., The surviving mice were then re-challenged with a much higher dose , 9 , 800 LD50 , of Y . pestis CO92 on day-48 post-first challenge ., The purpose of re-challenge was to determine if a strong adaptive immunity was generated after first infection with Y . pestis , which should in turn confer a much higher level of protection against subsequent challenges ., Indeed , our data showed that all of the mice survived the re-challenge except two mice in the native F1-V group that succumbed to infection ( 83% protection ) ( Figure 6D ) ., All of the naïve animals of same age which were used as a re-challenge control died as expected ., These efficacy results showed that the monomeric F1mut-V was as efficacious as or even slightly better than the native F1-V polymer ., The immunogenicity of nanoparticle decorated plague antigens was tested by vaccinating mice with phage T4 particles ( Figure 7A ) ., The amount of the antigen was kept the same as that of the soluble preparations ( Figure 6 ) ; however , the T4 formulations contained no adjuvant ., The data showed that the T4 displayed plague antigens induced comparable antibody titers as the adjuvanted soluble antigens ( Figure 7B ) ., The challenge data showed that all the T4 decorated plague antigens , including the V alone group , provided 100% protection to mice against intranasal challenge with 90 LD50 of Y . pestis CO92; all the control animals died by day, 4 . Upon re-challenge on day 48 post-first challenge with 9 , 800 LD50 ( Figure 7C ) , all of the mice were completely protected ., As expected , the control re-challenge group of mice succumbed to infection ., Overall , these data suggested that the T4 nanoparticle arrayed plague antigens might be more potent than the soluble antigens , as two deaths in each of the V and F1-V groups of mice occurred with the soluble vaccines ( Figure 6D ) but not with the T4 vaccines ., Stimulation of both arms of the immune system , humoral ( TH2 ) and cellular ( TH1 ) , is probably essential for protection against Y . pestis infection 6 , 23 , 44 , 45 ., In mice , the TH1 profile involves induction of antibodies belonging to IgG2a subclass whereas the TH2 profile primarily involves the induction of IgG1 subclass ., To determine the specificity of antibodies induced by soluble vs T4 displayed antigens , the subclass IgG titers were determined by ELISA ( Figure 8 ) ., These data showed that the soluble antigens and the T4 displayed antigens induced comparable IgG1 titers ( TH2 response ) ( Figure 8A ) whereas the T4 antigens evoked 1–2 orders of magnitude higher IgG2a titers than the soluble antigens ( TH1 response ) ( Figure 8B ) ., These results suggested that the T4 decorated plague immunogens stimulated stronger cellular responses as well as humoral responses , whereas the soluble antigens showed a bias towards the humoral responses as was observed in the previous studies 46 ., The immunogenicity and protective efficacy of F1mut-V vs F1mut-V10 was evaluated by three criteria: F1- and V-specific antibody titers , cytokine responses , and protection against Y . pestis CO92 challenge ., Both the F1- and V-specific IgG antibodies ( Figure 9B ) and subclass IgG titers ( Figure 8A and B ) were not significantly different between the F1mut-V and F1mut-V10 immunized groups of mice when the immunogen used was soluble and alhydrogel-adjuvanted ., However , when decorated on phage T4 nanoparticle with no adjuvant , F1mutV elicited higher total IgG ( Figure 9B ) and IgG1 titers ( Figure 8A ) than F1mutV10 ( p<0 . 05 ) ., These trends were also reflected in the production of the TH2 cytokines , IL-4 and IL-5 , by splenocytes of immunized mice stimulated ex vivo with F1-V ., Similar levels of IL-4 and IL-5 were produced by the soluble F1mut-V and F1mut-V10 antigens or the T4-displayed F1mut-V , whereas the T4 displayed F1mut-V10 showed slightly reduced levels ( Figure 10 ) ., The induction of proinflammatory cytokines , such as IL-1α and IL-1β was also similar , irrespective of whether the antigens were soluble or T4 displayed ( Figure 10 ) ., However the levels of TNF-α , an inflammatory mediator that synergistically acts with IFN-γ to help bridge the gap between innate and cell-mediated immune responses , were significantly higher in mice immunized with soluble F1mut-V10 than those immunized with F1mut-V ( Figure 10 ) ., However , the trend was opposite when F1mut-V and F1mut-V10 immunogens were T4 displayed , although the data did not reach statistical significance ., In fact , the T4 displayed F1mut-V10 induced overall weaker IFN-γ and cytokine responses when compared to its F1mut-V counterpart ., With respect to animal survival , both the F1mut-V and F1mut-V10 immunogens , either soluble or T4 displayed , provided 100% protection to mice upon intranasal challenge with 5 , 350 LD50 of Y . pestis CO92 ( Figure 9C ) , with the control animals dying by day, 3 . When the mice were re-challenged with an extremely high LD50 ( 20 , 000 ) on day 88 post-first challenge , all the groups showed 100% protection except the T4-displayed F1mut-V10 group in which one mouse died ( 92% protection ) ( Figure 9C ) ., All of the naïve re-challenge control animals died by day, 4 . To further test the efficacy of the mutated inmunogens , a rat study was conducted ., Rats 47 , the natural host of Y . pestis , were vaccinated with alhydrogel adjuvanted F1mut-V , and F1mut-V10 as well as the T4 nanoparticle displayed F1mut-V and F1mut-V10 ( Figure 11A ) ., The same immunization scheme as shown in Figure 6B was used and the animals were challenged with a 5 , 000 LD50 of Y . pestis CO92 ., The data showed that all the control animals died by day 4 whereas all the F1mut-V and F1mut-V10 immunized animals were 100% protected ( Figure 11B ) ., Since the deadly anthrax attacks in 2001 , stockpiling of recombinant anthrax and plague vaccines to protect masses against a potential bioterror attack became a national priority ., However , no plague vaccine has yet been licensed ., The reasons include poor stability , insufficient immunogenicity , and/or manufacturing difficulties associated with the current formulations ., New immunogen designs and vaccine platforms that could overcome some of these problems would be of great interest not only to stockpile efficacious biodefense vaccines but also to develop vaccines against a series of infectious diseases of public health importance ., Here , by using structure-based immunogen design and T4 nanoparticle delivery approaches , we have engineered new and efficacious plague vaccines that could be manufactured relatively easily and provide complete protection against pneumonic plague in two rodent models ., The surface-exposed Y . pestis antigens F1 and V have been the leading candidates for formulating a subunit plague vaccine for nearly two decades 14 , 15 , 16 , 17 ., Although poorly immunogenic by themselves , their immunogenicity could be enhanced by adjuvantation with Alum 15 or by fusion with a molecular adjuvant such as flagellin 19 ., While complete protection was observed in rodent models 17 , these vaccines impart partial and varied protection in African Green monkeys 6 , 17 ., Another concern has been that the naturally polymeric F1 has high propensity to aggregate ( Figure 2 ) ., When produced in a heterologous system such as E . coli , the recombinant F1-V protein partitions into insoluble inclusion bodies 15 , 19 , 20 , 21 , 22 ( Figure 3 ) ., Although it can be partially recovered in soluble form by denaturation and re-folding , the preparation still consists of a mixture of heterogenous aggregates and varying amounts of the misfolded protein 20 ., These might also trap contaminants , compromising the overall purity , stability , and efficacy of the vaccine ., Attempts to produce a monomeric vaccine by mutating the lone cysteine residue in V have not been successful 20 ., We proposed three hypotheses to design a soluble monomeric plague vaccine , yet retaining its structural and epitope integrity ., First , we hypothesized that the β-strand that connects the adjacent F1 subunits requires repositioning ., This was achieved by transplanting the NH2-terminal β-strand to the COOH-terminus in such a way that the reoriented β-strand fitted into its own β-sheet cleft rather than that of the adjacent F1 subunit ., It also eliminated the need for chaperone and usher mediated oligomerization as there would no longer be an unfilled β-sheet pocket exposed in the F1 subunit ., Second , by using epitope predictions , the NH2-terminal aa residues 15–21 of F1 flanking the β-strand were duplicated at the COOH-terminal end to restore any potential T-cell epitopes that might have been lost during the switch ., This is important because in a previous study , a simple β-strand switch produced a less stable monomer with diminished immunogenicity 48 ., Third , the mutated F1 was fused to the NH2-terminus of V with a flexible linker in between to minimize interference between the F1 and V domains ., The bivalent F1mut-V immunogen thus produced showed a remarkable shift in solubility , from an insoluble F1-V polymer to a completely soluble monomer ( Figure 3 ) ., The monomer could be purified from cell-free lysates at high yields , ∼20 mg of pure protein from a liter of E . coli culture , which we believe could be substantially increased under optimized conditions in a fermentor ., Several lines of evidence demonstrated that the F1mut-V monomer was as efficacious as , if not better than , the native F1-V polymer ., In four separate immunization studies and two animal models ( Figures 6 , 7 , 9 , and 11 ) , F1mut-V induced robust immunogenicity and protective efficacy ., It showed similar levels of F1- and V-specific antibody titers as the native F1-V , and no significant differences were observed in TH1 vs TH2 specific IgG subclass titers ., Furthermore , F1mut-V overall showed stronger cytokine responses and conferred 100% protection in vaccinated mice and rats , including when very high doses of Y . pestis CO92 , ∼5 , 350 LD50 for first challenge and ∼20 , 000 LD50 for re-challenge , were administered by the intranasal route ( Figure 9 ) ., The native F1-V , on the other hand , showed slightly lower protection ( ∼83% ) upon re-challenge ( Figure 6 ) ., The possibility of increasing the breadth and potency of F1-V vaccine by inclusion of YscF was tested by constructing an oligomerization deficient YscF35/67 mutant 43 ., Such a vaccine might be effective even against those Y . pestis strains that contain variant V antigens or lack the capsule , but are highly virulent 26 ., The mutated protein , purified as a soluble dimer , elicited YscF-specific antibodies on its own , and , when it was mixed with F1-V , it enhanced the induction of V-specific antibody titers as well as survival rate in mice ( Figure 6 ) ., While these results indicated enhanced potency of F1-V vaccine in the presence of YscF , more studies are needed to determine if the cost of an additional protein can be justified for vaccine manufacture ., On the other hand , the T4 displayed trivalent vaccine , F1mut-V-Soc-YscF ( Figures 5 and 7 ) , might offer an alternative to incorporate YscF into the plague vaccine formulation ., Y . pestis infection stimulates IL-10 production which in turn suppresses the production of proinflammatory cytokines IFN-γ and TNF-α ., Both IFN-γ and TNF-α are important for innate immunity , as well as to elicit TH1 immune responses that might be essential for protection against pneumonic plague 49 , 50 , 51 ., These immunomodulatory functions , in part , were attributed to the V antigen , specifically to the NH2-terminal aa residues 31–49 49 ., Deletion of these residues , or of the COOH-terminal aa residues 271–300 ( V10 mutation ) , have been reported to abrogate the suppression of IFN-γ and TNF-α 41 , presumably by preventing the interaction of V with toll like receptor 2 ( TLR2 ) and CD14 , the receptors of the innate immune system 49 , 52 ., Our studies showed that both the F1mut-V and F1mut-V10 immunogens produced similar levels of IFN-γ and other proinflammatory cytokines , such as IL-1α and IL-1β , upon stimulation ex-vivo of splenocytes from immunized mice with F1mut-V | Introduction, Results, Discussion, Materials and Methods | Pneumonic plague is a highly virulent infectious disease with 100% mortality rate , and its causative organism Yersinia pestis poses a serious threat for deliberate use as a bioterror agent ., Currently , there is no FDA approved vaccine against plague ., The polymeric bacterial capsular protein F1 , a key component of the currently tested bivalent subunit vaccine consisting , in addition , of low calcium response V antigen , has high propensity to aggregate , thus affecting its purification and vaccine efficacy ., We used two basic approaches , structure-based immunogen design and phage T4 nanoparticle delivery , to construct new plague vaccines that provided complete protection against pneumonic plague ., The NH2-terminal β-strand of F1 was transplanted to the COOH-terminus and the sequence flanking the β-strand was duplicated to eliminate polymerization but to retain the T cell epitopes ., The mutated F1 was fused to the V antigen , a key virulence factor that forms the tip of the type three secretion system ( T3SS ) ., The F1mut-V protein showed a dramatic switch in solubility , producing a completely soluble monomer ., The F1mut-V was then arrayed on phage T4 nanoparticle via the small outer capsid protein , Soc ., The F1mut-V monomer was robustly immunogenic and the T4-decorated F1mut-V without any adjuvant induced balanced TH1 and TH2 responses in mice ., Inclusion of an oligomerization-deficient YscF , another component of the T3SS , showed a slight enhancement in the potency of F1-V vaccine , while deletion of the putative immunomodulatory sequence of the V antigen did not improve the vaccine efficacy ., Both the soluble ( purified F1mut-V mixed with alhydrogel ) and T4 decorated F1mut-V ( no adjuvant ) provided 100% protection to mice and rats against pneumonic plague evoked by high doses of Y . pestis CO92 ., These novel platforms might lead to efficacious and easily manufacturable next generation plague vaccines . | Plague caused by Yersinia pestis is a deadly disease that wiped out one-third of Europes population in the 14th century ., The organism is listed by the CDC as Tier-1 biothreat agent , and currently , there is no FDA-approved vaccine against this pathogen ., Stockpiling of an efficacious plague vaccine that could protect people against a potential bioterror attack has been a national priority ., The current vaccines based on the capsular antigen ( F1 ) and the low calcium response V antigen , are promising against both bubonic and pneumonic plague ., However , the polymeric nature of F1 with its propensity to aggregate affects vaccine efficacy and generates varied immune responses in humans ., We have addressed a series of concerns and generated mutants of F1 and V , which are completely soluble and produced in high yields ., We then engineered the vaccine into a novel delivery platform using the bacteriophage T4 nanoparticle ., The nanoparticle vaccines induced robust immunogenicity and provided 100% protection to mice and rats against pneumonic plague ., These highly efficacious new generation plague vaccines are easily manufactured , and the potent T4 platform which can simultaneously incorporate antigens from other biothreat or emerging infectious agents provides a convenient way for mass vaccination of humans against multiple pathogens . | medicine, biology | null |
journal.pgen.1000106 | 2,008 | Sepsid even-skipped Enhancers Are Functionally Conserved in Drosophila Despite Lack of Sequence Conservation | Recent studies revealing how the gain , loss and repositioning of transcription factor binding sites within regulatory sequences can alter gene expression with observable phenotypic consequences 1 have focused efforts to understand the molecular basis for organismal diversity on the evolution of regulatory DNA ., However , a growing body of work has demonstrated that alterations of binding-site composition and organization often leave regulatory sequence function unchanged 2–9 ., The potential for significant changes in regulatory sequences to have no functional consequences complicates efforts to identify sequence changes that are likely to affect gene expression and phenotype ., But precisely because many of these changes do not affect regulatory output , they provide a powerful opportunity to understand how the arrangement of transcription factor binding sites in a regulatory sequence determines its output ., We believe that identifying divergent enhancers that drive similar patterns of expression , and distilling the common principles that unite them , will allow us to decipher the molecular logic of gene regulation ., We began to explore the effectiveness of this approach with the extensively studied regulatory systems of the early D . melanogaster embryo 10 , using the recently sequenced genomes of 12 Drosophila species to document the evolutionary fate of transcription factor binding sites in early embryonic enhancers ( Peterson , Hare , Iyer , Eisen , unpublished ) ., A consistent pattern emerged: while binding site turnover is common , a large fraction of the binding sites in most enhancers are conserved across the genus ( see Figure 1 ) ., The extent to which variation in enhancers from sequenced Drosophila species represented all of the possible variation in these sequences was unclear ., Perhaps the conserved sites were an imperturbable core essential for each enhancers function ., Or , perhaps , there had simply not been enough time since the divergence of the genus for mutation to have generated alternative configurations that would produce identical expression patterns ., To resolve this ambiguity it was necessary to reconstruct binding site turnover events that occurred over longer evolutionary timescales by comparing Drosophila enhancers to their counterparts in species from outside the genus ., The appropriate species for such comparisons would share basic patterning mechanisms with Drosophila species , but be sufficiently diverged from Drosophila to provide significant additional data on the constraints on binding site turnover ., Ideally , these species would be amenable to experimental analysis and have fully sequenced genomes ., Unfortunately , the closest available genome sequences were from several very distantly related mosquito species 11 , whose most recent common ancestor with Drosophila lived approximately 220 million years ago ., These sequences were unlikely to be informative because of several important differences between early-embryonic patterning in Drosophila and mosquitoes ., Mosquitoes , for example , lack the primary anterior morphogen in Drosophila , the modified Hox gene Bicoid , which is found only in higher cyclorrhaphan Diptera ( the “true flies” ) 12 ., With essentially no information on non-coding sequences and regulatory networks from flies outside the Drosophilidae , we reasoned that other groups within the Acalyptratae , the speciose 100 million year-old division of Diptera that includes Drosophila , represented the best compromise between our aims to maximize sequence divergence and minimize regulatory network divergence ., We selected three families , Sepsidae , Diopsidae and Tephritidae , that span acalyptrate diversity , have well-characterized phylogenies , and contain multiple species whose specimens could be readily obtained ., In this paper we present results on gene regulation in sepsids , which , due to their small genomes , were the most amenable to genome analysis ., Specifically , we report the sequence and experimental characterization of the even-skipped locus from six sepsid species ., The particular species were selected to include the major sepsid lineages , and , in several cases , because of the amenability of the species for embryological study ., We chose to characterize multiple sepsid species to facilitate the identification of sepsid enhancers by intra-family comparisons 13 , 14 and to enable comparisons of enhancer evolution between sepsids and drosophilids ., The six sepsid species we selected for this study , Sepsis punctum , Sepsis cynipsea , Dicranosepsis sp ., , Themira superba , Themira putris and Themira minor , have genome sizes that range from 134 Mb to 285 Mb ( Table 1 ) ., We generated a whole-genome fosmid library for each species , identified eve-containing clones by hybridization with a species-specific eve probe generated by degenerate PCR , and shotgun sequenced the clones to an average 13× coverage ( Table S1 ) ., We annotated the assembled sequences ( Table S2 ) to identify all protein-coding genes with homologs in D . melanogaster ( Figure S1 ) ., All of the sequenced clones contained clear eve orthologs , and the organization of the eve locus is very similar in sepsids and drosophilids ( Figure 2 ) ., The sepsid loci are slightly larger ( Table 1 ) , consistent with their overall larger genome sizes ., The genes flanking eve , however , are different between the families ., A maximum likelihood tree calculated using seven protein-coding gene sequences in all six sepsids and a subset of Drosophila species demonstrates that the sepsid species are about twice as diverged from D . melanogaster than D . melanogaster is from the most distantly related Drosophila species ( Figure 3A ) ., Examination of the eve locus from sequenced Drosophila species shows that there is readily detectable non-coding sequence conservation spanning the entire locus , even between the most distantly related species ( Figure 2 ) ., The average pairwise noncoding match score ( a BLASTZ 15 based measure of sequence similarity; see Materials and Methods ) between D . melanogaster and members of the virilis-repleta clade is 20% ( Table S3 ) ., We observe a similar pattern in the sepsid eve loci ., The average pairwise noncoding match score between S . cynipsea and Themira species is 17% ( Table S3 ) ., However , there is minimal non-coding sequence conservation between families outside of a few small ( approximately 20–30 bp ) blocks of extremely high conservation scattered across the locus ( Figure 2 ) ., The average pairwise noncoding match score between D . melanogaster and the sepsids is 4% ( Table S3 ) ., Maximum likelihood non-coding trees from the eve locus in sepsids and Drosophila reveal that the two families span roughly the same amount of non-coding divergence ( Figure 3B ) ., We established a colony of the T . minor from adults captured in Sacramento , CA , and developed protocols to recover and fix T . minor embryos ., The overall morphology and pattern of embryonic development is very similar in sepsids and Drosophila ( Figure S2 ) ., As expected from studies of other dipterans , T . minor eve is expressed in a characteristic set of seven stripes in blastoderm embryos ( Figure 4D , H ) ., We were additionally interested in comparing the trans-regulatory network of this sepsid to that of drosophilids ., In D . melanogaster , eve expression in the blastoderm is regulated by the transcription factors Bicoid ( BCD ) , Caudal ( CAD ) , Hunchback ( HB ) , Giant ( GT ) , Krüppel ( KR ) Knirps ( KNI ) and Sloppy-paired 1 ( SLP1 ) ., hb , gt and Kr are expressed in T . minor in patterns that mimic those of their orthologs in D . melanogaster embryos ( Figure 4A–C , E–G ) ., This is in contrast to AP patterning factors in the mosquito , in which there have been shifts in expression domains , and presumably changes in regulation of the downstream genes 16 ., We were unable to clone the kni , slp1 and cad genes from T . minor ., In D . melanogaster , bcd RNAs are tethered to the anterior pole of the embryo , with BCD protein diffusing away from the pole to create a strong anterior to posterior gradient ., BCD antibodies were not cross-reactive in T . minor , and we were unable to characterize the T . minor BCD gradient ., Key elements of the heart regulatory network are conserved between flies and vertebrates 17 ., As we therefore expect this network to be conserved between the sepsid and Drosophila species , and our supply of T . minor embryos was limited , we did not examine the expression of heart regulators ., Since the sepsid and Drosophila trans-regulatory networks regulating eve expression appear to be similar , we reasoned that sepsid enhancers would contain similar collections of transcription factor binding sites as their Drosophila counterparts ., In D . melanogaster , clusters of HB , CAD , KNI , KR , and BCD binding sites in the eve locus have been shown to correspond to known stripe enhancers 18 ., We therefore examined the density of predicted HB , CAD , KNI , KR , GT and BCD binding sites across each fosmid sequence ( Figure S3 ) and identified 18 candidate sepsid stripe enhancers ( Table S4 ) ( We recently generated GT in vitro binding data which was not available when the initial D . melanogaster work was carried out ) ., Each of these predicted enhancers contained a small number of short ( 20–30 bp ) sequences conserved between sepsids and drosophilids , which established presumptive orthology with specific regions of the D . melanogaster genome ., In essence , the binding site plots showed us where sepsid enhancers could be found , and the small islands of sequence conservation suggested their likely function ., We also identified putative eve muscle-heart enhancers ( MHE ) ( Table S4 ) in the sepsid species by looking for short blocks ( 20–30 bp ) of high similarity ( >90% ) that overlap functionally verified transcription binding sites from the D . melanogaster MHE in pairwise alignments between the D . melanogaster MHE and each of the sepsid intergenic regions ., We chose to test whether candidate enhancers from one species in each of the two sepsid clades were capable of driving expression in D . melanogaster embryos ., Enhancer-reporter cassettes for each of these 8 constructs were introduced into the D . melanogaster genome via Phi-C31 phage-mediated targeted integration 19 , 20 ., Remarkably , despite their extensive sequence differences , all of the tested sepsid sequences drive very similar expression patterns to those driven by their orthologous D . melanogaster enhancers ( Figure 5 ) , although there are some small and intriguing differences ., This confirms that these sepsid sequences are functional eve enhancers that , with their high degree of sequence divergence , represent markedly different examples of how to construct an eve enhancer ., The D . melanogaster minimal stripe 2 element drives expression in a single stripe in the stage 5 blastoderm from 63–57% egg-length through activation by broad anterior gradients of BCD and HB and localized repression by GT and SLP1 in the anterior and KR in the posterior 21 ( Figure 5A; Table 2 ) ., The sepsid stripe 2 enhancers in the transgenics similarly drive expression from 62–55% egg-length ( Figure 5B , C; Table 2 ) ., In 78% of embryos containing the S . cynipsea enhancer and 55% of embryos containing the T . putris enhancer , we observe expression in stripe 7 from the sepsid stripe 2 enhancers; similar behavior has also been observed for D . melanogaster stripe 2 constructs 21 ., The D . melanogaster stripe 3+7 enhancer ( Figure 5D ) is broadly activated by dStat and Tailless ( TLL ) ( stripe 7 only ) , and the two stripes of expression at 53–47% and 21–12% egg-length ( Table, 2 ) are carved out by domains of HB , KNI , and SLP1 repression 22 ., Stripe 3 expression in the transgenics containing sepsid stripe 3+7 enhancers agrees well with D . melanogaster ( Figure 5E , F; Table 2 ) ., The anterior border of stripe 7 corresponds to that in D . melanogaster , but in embryos containing either the S . cynipsea or T . putris stripe 3+7 element , stripe 7 expression extends posteriorly ., Significantly , the stripe 3+7 enhancer has been inverted in the Sepsis species relative to the other sepsids and Drosophila ., This strongly suggests that these enhancers are orientation-independent in their native genomic context ., The D . melanogaster stripe 4+6 enhancer drives expression in 2 stripes from 47–40% and 30–22% ( Figure 5G ) ., There is some evidence that stripe 4+6 expression is activated broadly by Dichaete and restricted to 2 stripes by HB and KNI repression , but the precise details of its regulation are less well understood 23 , 24 ., This pattern is reproduced in our transgenics , with expression from 46–40% and 31–25% egg-length ( Figure 5H , I; Table 2 ) ., In stage 11 D . melanogaster embryos , eve is expressed in laterally-symmetric , metameric pairs of pericardial cells in the dorsal mesoderm ( Figure 5J ) 25 ., The eve MHE integrates activation and repression from multiple signaling pathways , including DPP and WG from the dorsal ectoderm and RAS in the dorsal mesoderm 26 ., In addition , broad domains of TIN and TWI in the dorsal mesoderm activate expression ., This metameric pattern is faithfully reproduced by the sepsid MHE enhancers ( Figure 5K , L ) ., That enhancers with minimal sequence conservation have conserved function suggests that they share some common features beyond primary sequence ., In order to examine what these shared properties might be , we examined and compared the composition and organization of predicted transcription factor binding sites in all of the characterized eve enhancers ., We restricted our analysis of each enhancer to those factors known to be involved in the activity of the particular enhancer ., We aligned enhancer sequences from within each family , and plotted predicted transcription factor binding sites on these alignments ( Figure 6 ) ., 92% of D . melanogaster binding sites are found in the same location in enhancers from other species within the closely related melanogaster subgroup ( Table S3 ) , 29% of sites are similarly conserved between D . melanogaster and the species of the virilis-repleta clade ( Table S3 ) ., An average of 22% of sites are conserved between S . cynipsea and Themira species ., The non-coding divergence between these two sepsid clades is similar to that between D . melanogaster and the virilis-repleta clade ( Figure 3 ) ., This is likely an underestimate of the conserved sites within the sepsids as these are not minimal enhancers and thus should contain a larger portion of non-conserved background sites ., The lack of sequence similarity between families made nucleotide level alignment of sepsid enhancers to their Drosophila orthologs impossible ., However , the previously described small blocks of high sequence conservation allowed us to orient and crudely align the sepsid and drosophilid enhancers to each other ., In examining plots like this for all four enhancers , it was clear that few of the binding sites conserved within each family were conserved between families ( Figure 6; Figure S4 ) ., Only 5% of D . melanogaster binding sites are conserved in pairwise comparisons with sepsid species , representing an additional 84% reduction in conserved sites compared to the virilis-repleta clade ( Table S3 ) ., However , we note that all of the highly conserved blocks contained at least one , and often several , highly conserved binding sites , and that most of these sites correspond to known in vitro footprints for the corresponding factor in D . melanogaster 27 ( Figure S4 ) ., Most early embryonic enhancers in D . melanogaster contain unusually large numbers – compared to random non-coding sequence – of predicted binding sites for the factors involved in their regulation 18 , 28 , although the exact relationship between binding site density and function remains to be elucidated ., Binding site density is conserved between enhancers in D . melanogaster and D . pseudoobscura 13 , 14 , but it is not clear how much of this conservation is due to selection to maintain binding sites , and how much is due to the overall high level of sequence conservation between D . melanogaster and D . pseudoobscura ., Given the overall lack of sequence and binding site conservation between sepsid and Drosophila enhancers , we were particularly interested in the characteristics of the small sequence blocks that are conserved between the families ., We noticed that all of these blocks contained overlapping or tightly spaced binding sites ., To analyze this more rigorously , we classified predicted D . melanogaster binding sites for footprinted factors in the eve MHE , stripe 2 and stripe 3+7 enhancers into four categories ranging from non-conserved ( present only in D . melanogaster and its immediate sister taxa ) to extremely highly conserved ( present in Drosophila and sepsids ) ., We then classified sites based on their proximity to other predicted binding sites: overlapping sites that share one or more bases with another binding site , neighboring sites that are within 10 bases of another site but do not overlap , and isolated sites ., Overlapping sites are more often extremely conserved , close sites are more often highly conserved and isolated sites are more often minimally or non-conserved than expected by chance ( Figure 7; p<0 . 007 , p<0 . 01 , p< . 049 , Chi-squared test ) ., However , the number of sites is too small to detect relationships between conservation and the spacing of pairs of sites for specific factors ., Our work extends in both the extent of divergence and number of enhancers examined the pioneering work on binding site turnover of Ludwig and Kreitman , who showed in a series of papers that the eve stripe 2 enhancer from other Drosophila species drives a stripe 2 pattern in transgenic D . melanogaster embryos despite the imperfect conservation of functional binding sites 5 , 6 , 8 ., Although several examples of Drosophila regulatory sequence conservation over long evolutionary distances had been reported prior to Ludwig and Kreitmans work on eve stripe 2 29 , 30 , eve regulation has become the preeminent model for the study of binding site turnover ., It remains one of the few cases where observations of expression pattern conservation have been followed up with studies of functional complementation 7 ., We have nearly doubled the evolutionary distance analyzed by Ludwig and Kreitman ., Furthermore , in their comparisons the majority of binding sites were conserved , while our species sample has very few conserved binding sites ., We have also generalized their observation to include additional enhancers responding to a different suite of transcription factors , including one ( the MHE ) active following gastrulation ., Previous reports of the functional equivalence of divergent enhancers in Drosophila have involved blastoderm enhancers , leaving open the possibility that the observed binding site turnover was a byproduct of the syncitial nature of the early Drosophila embryo ., Our data on the MHE demonstrates that extreme binding site turnover with functional conservation occurs in enhancers active in a cellular context ., A handful of isolated case studies support our findings ., For example , the tailless enhancer from the house fly Musca domestica 31 and the single-minded enhancer from the mosquito Anopheles gambiae 32 drive similar patterns as their endogenous orthologs in D . melanogaster embryos despite having different organization of binding sites , and non-coding sequences from the human RET locus drive ret-specific expression in zebrafish despite the absence of detectable sequence similarity between human and zebrafish RET non-coding DNA 33 ., Nonetheless , in each of these cases simple transcription factor “grammars” were conserved , offering a ready molecular explanation for the conserved function ., No such grammar is as of yet apparent in the eve enhancers ., Such remarkable flexibility in the organization of enhancers suggests that the protein-protein and protein-DNA interactions that mediate the activity of developmental enhancers are not highly structured as , for example , is seen in enhanceosomes 34 ., If they were , it is hard to imagine how such wildly different sequences could produce identical expression patterns in the same trans-regulatory context ., The extent of binding site turnover is consistent instead with the recently proposed “billboard” model of enhancer activity in which enhancers contain multiple sub-elements that independently interact with cofactors and the basal machinery to dictate transcriptional output 35–37 ., In proposing the billboard model , Kulkarni and Arnosti proposed that billboard enhancers would be more evolutionarily pliable than enhanceosomes , and suggested that the eve stripe 2 results from Ludwig and Kreitman were understandable if eve stripe 2 were a billboard enhancer 35 ., Their model does not , however , predict how evolutionarily flexible billboard enhancers should be ., Our discovery of extreme sequence and binding site divergence between functionally equivalent sepsid and Drosophila enhancers shows that they are extremely flexible , a fact that must be accounted for in future models of enhancer activity ., However even billboard enhancers are not infinitely flexible ., One remarkable aspect of enhancer evolution is that despite the clearly frequent repositioning or replacement of transcription factor binding sites within enhancers , the enhancers themselves remain fairly compact ., There must , therefore , be selection to keep the different sub-elements that contribute to an enhancers output within the one to two kilobase span of a typical enhancer ., This spatial constraint implies some functional interaction between enhancer sub-elements not currently captured by the billboard model ., Given the extent of non-coding divergence between Drosophila and sepsids across most non-coding DNA , we were surprised to observe small islands of very strong sequence conservation ., Our finding that there is a significant enrichment of overlapping or adjacent binding sites within conserved blocks lends evolutionary support to long-standing suggestions of the importance of direct competitive and cooperative interactions between bound transcription factors ., Numerous studies have demonstrated that appropriate regulation of the eve stripe enhancers ( and other enhancers ) relies on the close proximity of multiple binding sites for both activators and repressors 21 , 36 , 38–41 ., Of the 12 footprinted BCD , HB , KR , and GT sites in the minimal stripe 2 element , 8 fall into 2 clusters of about 50 base pairs each containing overlapping activator ( HB or BCD ) and repressor ( KR or GT ) sites ., In transient transfection experiments using these binding site clusters , BCD and HB dependent activation was repressed by DNA binding of GT or KR , consistent with the short-range repression mechanisms of quenching or competition 40 ., Knirps also mediates short-range repression in a range of 50–100 bp through quenching or direct repression of the transcriptional machinery when bound near a promoter 42 ., Similarly , HB and BCD co-expression in transient transfection experiments results in multiplicative activation of a reporter construct containing a subset of the eve minimal stripe 2 element 40 ., Mutation of single activator sites in the minimal stripe 2 element results in a significant reduction in expression , again suggesting that HB and BCD bind cooperatively to this enhancer 21 ., The local quenching and cooperativity models predict that binding sites in close proximity to each other should be under strong purifying selection to remain close to each other ., Under the generally accepted model of binding site turnover , sites are lost in one region of an enhancer when new mutations create a complementary site elsewhere in the same enhancer ., The appearance of new sites is the rate-limiting step as there are more mutational steps required to create a new site from random sequence than to destroy an existing site ., Since random mutations are far less likely to produce pairs of adjacent sites than single sites , we expect functionally linked pairs of sites to be subject to far lower rates of binding site turnover ., In contrast , if binding site turnover is driven by base substitutions , we expect functionally independent sites that are adjacent or even partially overlapping to have essentially the same rates of binding site turnover as isolated sites ., The conserved blocks we observed between sepsids and Drosophila were generally larger than individual sites , as has been previously reported within Drosophila 43 , consistent with the former model ., Our observation that proximal sites are preferentially conserved additionally supports their direct functional linkage ., However , we note that insertions and deletions are a major source of sequence variation in Drosophila , with D . melanogaster having a strong deletion bias 44 and deletion is thought to contribute significantly to binding site turnover 45 ., Taking this into account , we expect to observe reduced turnover in even functionally independent binding sites if they are overlapping or adjacent , as some fraction of the deletions that would remove a binding site with a complementary site elsewhere would also affect adjacent , and presumably uncompensated sites ., These deletions would be subject to purifying selection , and the rate of turnover for the proximal sites would be reduced ., Assessing whether such an effect could explain our observation requires more data on relative rates of nucleotide substitution and insertion and deletions of different sizes in sepsids , which will be accomplished with the sequencing of sepsid genomes ., We can , however , test the significance of our observation directly ., The linked function model predicts that the paired binding sites we observe to be conserved between families should be more sensitive to manipulations that alter the spacing between the sites than paired binding sites that are not conserved ., Though expression of the sepsid eve enhancers in D . melanogaster embryos is qualitatively very similar to the patterns driven by the D . melanogaster enhancers , there are subtle and interesting differences ., Expression of stripe 7 exhibits the most variability across all enhancers in transgenics , including those enhancers from D . melanogaster ., It was previously observed that stripe 7 is weakly expressed in D . melanogaster stripe 2 transgenics , and stripe 7 expression is weaker than the endogenous stripe in stripe 3+7 transgenics 21 , 22 , 40 ., We frequently observed stripe 7 expression in all our non-Drosophila stripe 2 transgenics , and stripe 7 expression did not perfectly recapitulate endogenous expression , suggesting that regulatory information specifying this stripe is distributed across the upstream region , thus challenging the model of enhancer modularity in agreement with 46 ., Information may be more diffusely spread across the locus in sepsids , resulting in missing information in our discrete cloned enhancers , in which case the native D . melanogaster pattern should be more accurately reproduced by cloning a larger regulatory region ., Alternately , there could be changes within the non-Drosophila enhancers which result in expression differences in D . melanogaster despite conserved native eve expression , suggesting co-adaptation of each enhancer and its native trans environment ., We began this study seeking taxa that were significantly more diverged from D . melanogaster than any Drosophila species , but which had sufficiently conserved cis-regulatory networks that their enhancers would have similar function to their D . melanogaster counterparts ., Our choice of sepsids was guided by their relatively close – but not too close – position to Drosophila on published trees of Diptera 47 , by their relatively similar morphology suggestive of similar developmental mechanisms , and by practical considerations such as genome size and availability ., We have now shown that the extensive sequence divergence between sepsids and Drosophila was not accompanied by extensive differentiation of early embryonic patterning mechanisms ., Thus sepsids provide a valuable model for comparative analysis of Drosophila embryology and developmental cis-regulation ., We were also able to establish a colony of sepsids ( T . minor ) in the lab from flies caught locally , and collect embryos for the developmental gene expression and morphology data presented here ., Based on our experience , we believe that more extensive embryological and molecular work with sepsids is very feasible , although some may find the need to provide the colonies with fresh cow dung objectionable ., The additional sequence divergence has enabled us to reach two important conclusions that could not be obtained in analyses of the 12 sequenced Drosophila genomes ., Previous analyses of binding site turnover in Drosophila revealed substantial numbers of conserved binding sites within the genus , leaving open the question of whether these sites represented an imperturbable core necessary for enhancer function , or if there had simply not been sufficient divergence time for mutation to generate alternative configurations ., We have now largely answered this question , at least for the eve enhancers – there does not appear to be an imperturbable core of sites at the level of overall enhancer organization ., Although binding site conservation in Drosophila has been extensively studied , our observations about the relationship between conservation and binding site proximity were never described because this pattern was simply not evident in examinations of the multitude of conserved binding sites across the Drosophila genome ., This relationship only became apparent when we observed just how striking the conservation of a small subset of sites was ., More generally , this study highlights the value of the infrequently studied ( at least by molecular biologists ) Dipteran species outside of the genus Drosophila ., It also points to a general strategy for dissecting the still elusive molecular mechanisms of enhancer function in which genome sequencing and functional studies are combined to catalog the diverse ways in which regulatory sequences with common function can be generated ., Our initial foray into this domain has yielded exciting and unanticipated results ., With the cost of genome sequencing plummeting , and with great improvements in Drosophila transgenesis , we expect this approach to be even more productive in the years to come ., Sepsis punctum , Sepsis cynipsea , Themira superba , Themira putris and Dicranosepsis sp ., stocks were maintained in the Evolutionary Biology Laboratory at the National University of Singapore ., Themira minor cultures were established at LBNL from specimens collected at McKinley Park in Sacramento , CA ., Samples for genome sizing and genomic DNA isolation were flash-frozen adult flies ., Genome sizing methods were adapted from 48 ., Five adult heads for each species were dissected into 1 . 5 mL of Galbraith buffer on ice , homogenized with 15 strokes of an A pestle in a 15 mL Kontes Dounce tissue homogenizer , and filtered through 30 um nylon mesh ., T . superba heads were combined with 5 D . virilis heads before homogenization ., 7 uL of 1∶10 chicken red blood cells ( diluted in PBS ) and 50 uL of 1 mg/mL propidium iodide were added and samples were stained for 4 hours rocking at 4 degrees in the dark ., Mean fluorescence of co-stained nuclei was quantified on a Beckman-Coulter EPICS XL-MCL flow cytometer with an argon laser ( emission at 488 nm/15 mW power ) ., The propidium iodide fluorescence and genome size of Gallus domesticus ( red blood cell standard , 1 , 225 Mb ) were used to calculate the unknown genome sizes ., For T . superba , D . virilis at 328 Mb , was used as a second internal sta | Introduction, Results, Discussion, Methods | The gene expression pattern specified by an animal regulatory sequence is generally viewed as arising from the particular arrangement of transcription factor binding sites it contains ., However , we demonstrate here that regulatory sequences whose binding sites have been almost completely rearranged can still produce identical outputs ., We sequenced the even-skipped locus from six species of scavenger flies ( Sepsidae ) that are highly diverged from the model species Drosophila melanogaster , but share its basic patterns of developmental gene expression ., Although there is little sequence similarity between the sepsid eve enhancers and their well-characterized D . melanogaster counterparts , the sepsid and Drosophila enhancers drive nearly identical expression patterns in transgenic D . melanogaster embryos ., We conclude that the molecular machinery that connects regulatory sequences to the transcription apparatus is more flexible than previously appreciated ., In exploring this diverse collection of sequences to identify the shared features that account for their similar functions , we found a small number of short ( 20–30 bp ) sequences nearly perfectly conserved among the species ., These highly conserved sequences are strongly enriched for pairs of overlapping or adjacent binding sites ., Together , these observations suggest that the local arrangement of binding sites relative to each other is more important than their overall arrangement into larger units of cis-regulatory function . | The transformation of a fertilized egg into a complex , multicellular organism is a carefully choreographed process in which thousands of genes are turned on and off in specific spatial and temporal patterns that confer distinct physical properties and behaviors on emerging cells and tissues ., To understand how an organisms genome specifies its form and function , it is therefore necessary to understand how patterns of gene expression are encoded in DNA ., Decades of analysis of the fruit fly Drosophila melanogaster have identified numerous regulatory sequences , but have not fully illuminated how they work ., Here we harness the record of natural selection to probe the function of these sequences ., We identified regulatory sequences from scavenger fly species that diverged from Drosophila over 100 million years ago ., While these regulatory sequences are almost completely different from their Drosophila counterparts , they drive identical expression patterns in Drosophila embryos , demonstrating extreme flexibility in the molecular machines that interpret regulatory DNA ., Yet , the identical outputs produced by these sequences mean they must have something in common , and we describe one shared feature of regulatory sequence organization and function that has emerged from these comparisons ., Our approach can be generalized to any regulatory system and species , and we believe that a growing collection of regulatory sequences with dissimilar sequences but similar outputs will reveal the molecular logic of gene regulation . | developmental biology/embryology, genetics and genomics/comparative genomics, computational biology/transcriptional regulation, evolutionary biology/evolutionary and comparative genetics, computational biology/comparative sequence analysis, developmental biology/developmental evolution, evolutionary biology/genomics, evolutionary biology/pattern formation, developmental biology/molecular development, evolutionary biology/developmental evolution | null |
journal.pcbi.1002639 | 2,012 | Prediction of Mutational Tolerance in HIV-1 Protease and Reverse Transcriptase Using Flexible Backbone Protein Design | The relationship between protein sequence and structure is fundamental for protein function , evolution and design 1 , 2 ., Many sequences are compatible with a given structure and function and thus proteins are often robust to point mutation 3 , 4 , 5 ., The concept of “tolerated sequence space - the set of sequences that accommodate a given structure and function - has been applied to characterize the emergence of protein families 6 , to describe protein interaction specificity 7 and to explain the evolution of new protein functions 8 , 9 ., Tolerated sequence variability ( robustness to mutation ) should be an advantage if proteins need to satisfy multiple functional constraints simultaneously ., If each constraint can be accommodated by many sequences , it should be easier to find a subset of sequences that satisfy multiple requirements 10 ., Moreover , a protein that has many tolerated sequences may be able to accommodate new constraints without abandoning some existing function 8 , 11 , 12 ., An example of this ability of proteins to rapidly adapt to new pressures is the emergence of drug-resistance mutations in pathogens ., In many cases , variants of pathogenic proteins that are resistant to inhibitors appear quickly , while still preserving their essential functions for the pathogen ., It is likely that some of these mutations are already present in the population as part of naturally occurring nearly neutral sequence variation 13 and are then selected by inhibitor treatment ., Thus , the a priori prediction of the tolerated sequence variation of pathogenic proteins would have implications for development of inhibitors against which resistance is less likely to arise quickly 14 ., Here we develop and assess a computational approach to predict the tolerated space of single mutations around a given protein sequence ., As model systems for validating our approach , we use the protease and reverse transcriptase from HIV-1 ., With more than 50 , 000 known sequences and several hundred experimentally determined structures , these two viral proteins are among the best-characterized systems available of tolerated variants around a native sequence ., Because protein sequences have been collected before and after viral inhibitor treatment 15 , predictions of mutational tolerance can be assessed in both a nearly neutral setting and under selective pressure to evolve resistance mutations ., In testing our model for HIV-1 protease mutational tolerance , we also make use of a large-scale mutagenesis experiment which evaluated the in vivo function of roughly 50% of all mis-sense mutations reachable by a single-nucleotide change from a starting consensus sequence 16 ., We find that our approach , which employs computational protein design methods in Rosetta 17 , recapitulates a substantial fraction of mutations experimentally observed to be tolerated by HIV protease and reverse transcriptase ., For accurate predictions , we show that it is critical to treat the protein not as a rigid single structure , but to allow conformational variation to accommodate sequence changes 18 , 19 , 20 ., We show that essentially the same prediction accuracy is achieved when obtaining conformational variation from an ensemble of experimentally determined structures of HIV protease 21 or reverse transcriptase , or from computationally generated conformational ensembles 18 , 19 , 20 , 22 ., We thus expect our approach to also be applicable to systems for which there is only one structure known ., Computational models of accessible mutational space , such as the one presented here , may prove generally useful for describing the evolvability of proteins by forecasting the emergence of mutations that can enable new protein functions 8 ., To predict a proteins tolerance to mutation , ideally all constraints acting on that protein should be modeled explicitly ., In addition , accurate predictions of mutational tolerance may require that conformational adjustments in response to mutation be considered ., Here we present a methodology that incorporates multiple functional constraints as well as backbone flexibility into RosettaDesign 17 and apply it to the prediction of mutational tolerance ., We first consider the viral protein HIV-1 protease , and later extend our results to HIV-1 reverse transcriptase ., HIV-1 protease is an ideal test system for several reasons ., First , the mutational tolerance of HIV-1 protease is well characterized: mutations of HIV-1 protease , including those causing resistance to protease inhibitors in HIV treatment , have been extensively documented and are available in the Stanford HIV-1 Drug Resistance Database 15 ., Second , HIV-1 protease is under at least three structural and functional constraints that are straightforward to model: ( 1 ) the 99-residue protease sequence must adopt a stable fold; ( 2 ) the active enzyme is a homodimer , and ( 3 ) the dimeric form must bind at least 10 endogenous peptides ., Finally , HIV-1 protease is structurally well characterized , with hundreds of crystal structures of native and mutated forms in the apo state or with peptide or inhibitors bound ., Figure 1A outlines the computational strategy for predicting mutational tolerance , starting from three-dimensional structural information on the protein of interest ., Figure 1B gives an example of the calculations for one sequence position in HIV-1 protease ., We started from the consensus sequence for HIV-1 protease ( see Methods ) , and considered all individual point mutations independently ( the simplest model of mutational space around a given sequence ) ., We used RosettaDesign 10 , 17 , 18 to mutate , in silico , each sequence position to 19 naturally occurring amino acid types ( mutations to and from cysteine were excluded; see Methods ) ., For each residue change , the side-chain conformations were optimized around the site of mutation ., We then calculated the per-residue energy contribution ( termed ERES ) of each point mutation using the Rosetta all-atom force field ( see Methods ) ., ERES scores were computed with respect to the three functional pressures described above: ( 1 ) the stability of the protease fold ( ERESFold , Figure 1B , left ) ; ( 2 ) the stability of the protease dimer interface ( ERESDimer , Figure 1B , middle ) ; and ( 3 ) the stability of the binding interactions with endogenous substrate peptides ( ERESPeptide , Figure 1B , right ) ., The model has the following key steps and components ( Figure 1A ) : Evaluating the robustness of a protein to mutation requires accurate distinction between sites that display amino acid variation and ones that do not ., Some protein sites are mutation intolerant under neutral conditions but become more tolerant under selective pressure; other sites are intolerant to mutation under both neutral and selective conditions ., Approximately 2/3 of protease sites ( 63 out of 96 ) within the Stanford HIV-1 Database sequences 15 appeared largely intolerant to mutation prior to inhibitor treatment ( Figure 2A; intolerance to mutation defined as a mutation frequency of <1% ) ., Further , about half of protease sites within the database ( 43 out of 96 ) were largely intolerant to mutations under inhibitor treatment ( Figure 2B ) ., The neutral and selective models correctly identified the majority of these intolerant protease sites ( Figure 2A–B; 45/63 and 31/43 , respectively ) ., Within the database sequences , only a few protease sites displayed high mutational tolerance ( Figure 2A–B; 8 and 14 sites , in the absence and presence of inhibitors , respectively; high mutational tolerance defined as a mutation frequency >20% ) ., The neutral and selective models correctly identified over half of these frequently mutated protease sites ( Figure 2A–B; 5/8 and 8/14 sites , respectively ) , including five sites that displayed high mutational tolerance in both a neutral setting and under selective pressure ( Figure 2C; 35E , 37N , 62I , 63L , and 77V ) ., Importantly , the individual mutations observed in the Stanford database were also correctly predicted for many of the frequently mutated sites ( Figure 2C; bold residues in 4th and 7th columns ) ., Similar results were observed at sites within the database with moderate mutational tolerance; these sites were often correctly predicted by both the neutral and selective models ( Figure 2C; 12T , 14K , 18Q , 19L , 20K , 39P , 60D , 61Q , 70K , and 92Q; moderate mutational tolerance is defined as amino-acid variation between 1–20% ) ., Therefore we conclude that the models can , in many cases , recapitulate both protease sites and individual protease mutations that are functionally tolerated ( the results for the neutral model are shown in Figure S1 ) ., To quantify the overall ability of the neutral and selective models to recapitulate individual mutations observed in the Stanford database , we used two standard metrics: ( 1 ) We computed a Receiver Operating Characteristic ( ROC ) curve by calculating the true positive rate ( TPR ) and false positive rate ( FPR ) of identifying protease mutations observed within the Stanford database above a threshold frequency of 1% and ( 2 ) we calculated an Area Under the Curve ( AUC ) for each ROC plot ( Figure 3 ) ., Both the neutral and selective models recapitulated many HIV-1 protease database mutations without incorrectly predicting a large number of false positives ( Figure 3A and 3E; black curve and black bar ) ., Commonly , a model with no predictive power will have a ROC curve that is a diagonal line and an AUC value of 50% ., We chose two naïve mutation tolerance prediction models as additional references ., In control model 1 , each site can tolerate all mutations that are accessible by a single nucleotide change from the consensus sequence ., In control model 2 , each site can tolerate amino acid types chemically similar to the native amino acid ( see Methods ) ., Control model 1 predicted the majority of the experimentally observed mutations ( TPR ∼90% , red triangle in Figure 3A ) ., However , a large number of non-observed mutations were incorrectly predicted as tolerated ( ∼37% FPR ) ., The computational models had a lower FPR at the same TPR ., Control model 2 rarely predicted tolerance to mutations that were not observed within the database ( ∼11% FPR ) , but did not capture tolerance to many database mutations ( ∼60% TPR , blue square in Figure 3A ) ., The computational models ranked more mutations correctly at the same FPRs ., In addition to recapitulating database mutations found in either neutral or selective conditions , our prediction scheme was also successful in recovering literature-documented drug resistance mutations ( DRMs ) for protease ., The comparison between predictions of the neutral and selective models ( Figure 2C ) yielded 18 sites that showed increase in mutation frequency ( rare/moderate to moderate/high ) , 9 out of which contain previously characterized DRMs ( as listed in 23 , see circles in Figure 2C ) ., Thus comparing predictions from the neutral and selective models may , in some cases , allow for identification of sites that contain drug resistance mutations ., Overall , the agreement between the individual mutations appearing within the database and the mutations predicted as tolerated by the models was strong ( Figures 2C , 3A and Figure S1 ) ., Nevertheless , several notable under- and over-predictions were observed ., Under-predictions of mutational tolerance by the neutral model were most notable at 10 sites ( Figure 2C , 2nd and 5th columns; 13I , 15I , 16G , 33L , 36M , 41R , 57R , 64I , 89L , and 93I ) ., The same 10 sites were also under-predicted for the selective model , with additional under-predictions occurring at 8 sites ( Figure 2C , 3rd and 6th columns; 10L , 20K , 48G , 54I , 73G , 82V , 84I , and 90L ) ., At most of these sites , the specific mutations observed in the Stanford database were correctly identified , but the predicted frequencies of mutation were significantly less than experimentally observed ( Figure 2C; 4th and 7th columns ) ., Notably , almost all under-predicted sites contained DRMs ( see circles in Figure 2C; exceptions are 15I , 41R and 57R ) ., Under-predictions may result from errors in the Rosetta energy model or from the inability to correctly capture structural changes in response to sequence changes ., Over-predictions of mutational tolerance occurred primarily within the beta-sheet pairing of the dimer interface ( 1P , 3I , 6W , 98N ) , three sites in the dimer flaps ( 45K , 46M and 47I ) , and several surface sites ( 21E , 35E , 43K , 55K , 58Q , 65E , 69H , and 72I , Figure 2C ) ., DRMs were relatively rare within sites that were over-predicted , although they did occur at two sites within the protease flaps ( 46M , 47I ) and at surface sites ( 35E , 43K , 58Q , and 69H; circles in Figure 2C ) ., As with under-predictions , model over-predictions could be due either to inaccuracies of the Rosetta model or additional functional pressures not captured ., The high predicted frequency of mutation at sites 46 and 47 likely occurred due to the presence of a clash with one of the modeled substrate peptides at these sites ., Thus , predictions at these two sites might be improved if a crystallographic structure of protease bound to this modeled peptide was available ., In addition , Rosetta often performed poorly at predicting mutation frequencies at polar exposed sites ., This poorer performance highlights known difficulties in accurately modeling the energetics of polar interactions ., Furthermore , despite the inclusion of two terms to disfavor mutations away from polar residues , we may not correctly capture other pressures acting particularly on surface residues , such as selection against aggregation ., As described above , we noted several instances where the selective model predictions did not agree with the mutations observed in the HIV-1 protease database sequences ., However , we found that some predictions instead agreed with mutations shown to be tolerated in an experimental study of single mis-sense mutations 16 ( Figure 2C , bold residues in 8th column ) ., This finding suggests that the selective model might capture protease mutational tolerance not yet observed at high frequency within the database sequences ., In support of this idea , we note that three mutations recently identified in the presence of inhibitors ( M46V , F53Y , and N83D ) 24 , 25 were predicted as tolerated by the selective computational model ( Figure 2C , 4th column ) ., All three newly identified mutations were not yet found within the protease database sequences at appreciable frequencies ., As described above , differences observed between the selective and neutral models can be used to recapitulate and predict DRMs ., In this section we examine in detail the ability of the model to recapitulate tolerance for 71 previously characterized DRMs ., We used a list of mutations from 23 and their grouping into major and minor DRMs ., Both groups show an increased frequency of mutation after inhibitor treatment , but only major DRMs have been directly implicated in causing resistance to inhibitors ., The selective model permits mutations near the protease inhibitor binding-site by weakening constraints on the protease dimer and substrate-binding interface ., We first analyzed whether the selective model predicts tolerance to DRMs located within the inhibitor-binding site ., Of the 18 DRMs near the substrate-binding site , 12 were predicted as tolerated by the selective model ( Figure 4A; 3 DRMs were disfavored by the model as they required more than a single nucleotide change from the consensus sequence ) ., Not surprisingly , most DRMs within the substrate-binding site were predicted to have mild-to-moderate destabilizing effects on binding of at least one of the 10 endogenous peptide substrates ( Figure 4A , red coloring ) ., The three DRMs not identified by the selective model were predicted to highly destabilize binding of at least one peptide ( Figure 4A , red boxes; 82L/F , 48V ) ., In contrast , effects on fold and dimer stability of the DRMs within the inhibitor-binding site were predicted as mostly energetically favorable or neutral ( Figure 4A , blue and beige coloring; 47A , 48V and 53L are notable exceptions ) ., At least one mechanism to compensate for substrate binding destabilization is known ., Peptide sequences cleaved by HIV protease can co-evolve with the appearance of DRMs such that mutations within the cleavage sequences counteract the predicted losses in substrate binding affinity 26 , 27 , 28 ., Although the selective model does not directly mimic this mechanism of co-evolution , it correctly predicted tolerance to most documented DRMs within the protease inhibitor-binding site ., We next examined DRMs known to occur outside of the protease substrate-binding site ., Here , the selective model correctly predicted mutational tolerance towards almost all major DRMs and towards the majority of minor DRMs ( Figure 4B; 7/8 and 31/45 , respectively; note 6 minor DRMs were disfavored by the model ) ., In the cases where the model did not predict a DRM to be tolerated , it was because the mutation was calculated to strongly destabilize the protease fold ( Figure 4B , red coloring ) ., These predicted destabilizing effects of some mutations may need to be compensated for by other co-occurring mutations ., Consistent with this hypothesis , 4 out of the 12 predicted destabilizing DRMs ( close and far from the substrate binding site ) occurred in the 53 most statistically significant correlated pairs of mutations observed after protease inhibitor treatment 29 ., Even though the selective model currently cannot account for correlated mutations , it nevertheless correctly predicts tolerance towards a considerable number of DRMs outside of the protease-binding site ., We next examined the contribution of stabilizing mutations to DRMs in HIV protease ., This analysis was based on a set of 62 out of the 71 documented DRMs , which had a frequency of >0 . 5% in the Stanford HIV database ., In total , 11 of the DRMs were predicted to stabilize the protease fold , both within ( 30N , 32I , 46I/L , 50L Figure 4A ) and outside ( 35G , 43T , 63P and 71V/I/T , Figure 4B ) the binding site ., Interestingly , DRMs at sites 30 , 32 and 50 are predicted to have a favorable effect on fold stability , and a destabilizing effect on peptide binding ., We asked whether DRMs that are predicted to have a fold-stabilizing effect ( out of all 62 DRMs that are both documented and predicted ) are over-represented relative to any documented protease mutation predicted to have a fold-stabilizing effect ( out of all possible protease mutations reachable by a single nucleotide change from the consensus sequence ) ., We found that there is a significant overrepresentation of DRMs that are predicted to be stabilizing ( ΔERESFold<0 ) : 17 . 7% ( 11/62 ) , in contrast to only 10% ( 72/705 ) of all protease mutations observed in the HIV-1 database reachable by a single nucleotide change ( p value\u200a=\u200a1 . 43E-7 , Mann-Whitney test ) ., One possible reason for the overrepresentation of stabilizing DRMs is that these sites reside in special locations ( such as buried sites that generally contribute more to stability ) in the protein structure ., We thus calculated the percentage of buried and exposed DRMs and compared these values to the percentage of buried and exposed residues of all documented protease mutations ( Figure S2 ) ., We found no significant difference in the burial of positions at which DRMs appear ., In addition , we studied a list of 33 frequent DRMs that often occur in combination ( extracted from the Stanford HIV database , Table S5 ) ., Assigning our calculated ERESFold scores for these mutations , we found that 22/33 of the co-occurring mutations included a combination of at least one destabilizing and one stabilizing mutation ., These analyses suggest that the modeled stabilizing DRMs may play a role in drug resistance by compensating for the destabilizing effects of other mutations ., We next analyzed whether two key features of the model – incorporating multiple constraints and using backbone ensembles – contributed to prediction performance , using the ROC and AUC metrics introduced above ., We first asked whether the model we present , which incorporates fold , dimer and peptide constraints for HIV-1 protease , would outperform a simpler model that considers only fold stability ., To do so , we recalculated mutational tolerance at every protease site , but this time we used only the scores for each point mutation and we set all the and terms to zero ( “single constraint model ) . The predictions of mutational tolerance from this single constraint model were less accurate than the original multiple constraint model , at least under selective conditions ( Figure 3B and 3E; cyan curves and bars ) . Thus , incorporating multiple constraints may be particularly useful for modeling selective pressure , because it allows weakening of certain constraints ( such as dimer stability and substrate binding ) over others . We next tested how accurately protease mutational tolerance would be predicted if only a single protease structure was used . To do so , we tested a “single structure model , in which we made 263 independent calculations of HIV-1 protease mutational tolerance ., In each set of predictions , we used the ERESFold and ERESDimer scores calculated from a single backbone structure rather than finding the minimum and scores calculated over the entire ensemble of structures ( identical scores were used in all cases , see Methods ) ., When we compared ROC curves and AUC values obtained from predictions made using single protease structures ( Figure 3C grey curves shown for 11 structures; Figure 3E grey bars ) to model predictions made using the ensemble of crystal structures ( black curves and black bars ) , we again observed consistently poorer model performance ., This suggests incorporating backbone variability by making predictions over an ensemble of backbone structures can be important for correctly predicting protease mutational tolerance ., HIV protease has been particularly well characterized and hundreds of solved crystal structures exist within the Protein Data Bank ., Many of these protease crystal structures originally contained point mutations ., Thus the improvement seen in predicting mutational tolerance using the ensemble of protease crystal structures could have been influenced by the original presence of these point mutations ( while all mutations were computationally reverted to the consensus sequence at the start of our simulations , any backbone structural changes present in the mutated structure remained , see Methods ) ., Furthermore , other proteins may not have comparably large ensembles of experimental structures and thus the method we describe here could , for this reason , be less applicable ., To address both these issues , we next tested whether accurate predictions of mutational tolerance could be made using a computationally generated , rather than an experimentally determined , ensemble of protease backbones ., To ensure that the computational ensemble did not contain “structural memory of point mutations present in the original crystallographic ensemble , we selected as templates 11 protease crystal structures that did not contain mutations from the consensus sequence . From each of the 11 templates , we used a computational method termed “backrub to generate an ensemble of 400 protease structures 20 , 30 with “near-native backbone conformations ( ensemble members had Cα RMSDs of 0 . 2 to 0 . 6 Å to the original starting template structure ) . We then repeated the calculations of mutational tolerance using each computational ensemble as described for the ensemble of experimentally determined structures . Remarkably , the same crystal structures that had resulted in poorer ROC curves and AUC values when considered as single structures ( 11 grey curves and grey bar , Figure 3C and 3E ) now showed improved results when the structures were used as starting templates for a computationally generated ensemble ( 11 orange curves and orange bar , Figure 3D and 3E ) . Furthermore , ROC curves and AUC values for predictions made using computationally generated ensembles were almost identical to those originally made using the ensemble of experimentally determined crystal structures ( compare black and orange curves and bars , Figure 3D and 3E ) . Therefore , while increasing the computational cost linearly with the number of backbones ( see Text S1 for estimates on computational time ) , backbone ensemble calculations can result in considerably better prediction than when using only a single backbone . To gain insight into how structural flexibility might have resulted in improved predictions of mutational tolerance , we examined the model predictions in more detail . Figure 5 shows two mutations as examples where backbone flexibility appeared to be crucial for correctly predicting tolerance to mutations observed in the Stanford database . When mutations A71V and I93L were individually modeled onto HIV-1 protease fixed backbone structures crystallized in the absence of any mutation , large to moderate clashes resulted ( Figure 5 , left ) . In each case , the clashes could be resolved when the same mutation was modeled onto a backbone computationally generated from an unmutated starting structure using the backrub method ( Figure 5 , middle ) . The mutations modeled onto the computationally generated backbones had structures and ERESFold scores close to those seen in experimentally determined structures that had originally contained the mutation ( Figure 5 , right ) . These results suggest that backrub ensembles , even though they were generated in the absence of mutations , can capture sufficient protein conformational variability to accommodate amino acid changes 20 , 30 . Figure S3 confirms that the mutations 71V and 93L were predicted as tolerated when modeled onto either experimental or backrub ensembles , but never when modeled onto a single fixed backbone of the consensus sequence ( similar behavior was also observed for mutations 24I and 77I , Figure S3 ) . We note that a few mutations within the Stanford database sequences that had been poorly predicted when using the ensemble of experimentally determined structures were found to be tolerated when using computationally generated ensembles ( e . g . 33F and 12S , see Figure S3 ) . To test the applicability of our model , we chose another protein system , the HIV-1 reverse transcriptase . This RNA-dependent DNA polymerase transcribes the single-stranded retroviral RNA genome into a double-stranded proviral DNA . Reverse transcriptase is a heterodimer built of the p66 subunit ( 560 residues ) and the p51 subunit , which has an identical sequence to the first 440 residues of p66 . The unique C-terminal part of the p66 subunit comprises an RNaseH domain . Similar to the protease system , many reverse transcriptase structures have been determined , many pre/post drug treatment mutations are catalogued in the Stanford database 15 and mutational tolerance prediction can be made using both fold and dimer stability as functional constraints . Nonetheless , the reverse transcriptase model has several limitations . There are fewer crystal structures than for protease ( see Table S1 ) and there are stretches of sequence with missing density in these structures . The substrates of reverse transcriptase are DNA/RNA hybrid molecules , for which interaction energy calculations are less established than for protein-protein interactions . We therefore did not consider reverse transcriptase residues in the interface with nucleic acids . In addition , model predictions could not be verified for the RNaseH domain , since mutational data are too sparse in this protein segment ( see Methods ) . In sum , we evaluated our analysis based on an ensemble of 91 structures and 656 of the 1 , 000 residues in the reverse transcriptase heterodimer ( still a much larger number of residues than in HIV protease; note that while some residues are excluded from the analysis , all protein residues present in the structures were used in the calculations ) . We repeated all mutational tolerance calculations as described for protease , calculating ERESFold and ERESDimer scores for every structure within the ensemble ( ERESPeptide could not be computed for reverse transcriptase that does not have peptide substrates ) . Otherwise the model parameters determined for protease were used unchanged for reverse transcriptase . Mutations were made simultaneously for every shared sequence position in the p51 and p66 subunits , while the p66-specific RNAseH domain sites were mutated only on the p66 subunit . The detailed results for the modeled versus observed mutational tolerance for reverse transcriptase are given in Figure S4 ( neutral model ) and Figure S5 ( selective model ) . As was observed for protease , a sizeable number of reverse transcriptase sites have low mutational tolerance , and a rather small number of sites were frequently mutated ( 224 and 17 sites , see Figure 6A ) . The neutral model correctly identified the majority of these sites ( 70% and 53% for the rarely mutated and frequently mutated sites , respectively ) . In contrast , the performance of the selective model for reverse transcriptase was poorer: 130/201 sites that rarely mutate were correctly predicted , and 14/31 sites that frequently mutate ( see Figure 6B ) . Under-predictions were seen at five out of 17 sites ( 177D , 211R , 329I , 334Q and 376A ) , while over-predictions were seen for 42 out of 224 ( 19% ) sites . Thirteen out of these 42 sites are exposed polar residues ( as for protease , Rosetta performed poorly at predicting at polar exposed sites ) . Many sites with over-predicted mutational tolerance are in protein segments that rarely mutate due to constraints likely not captured in our prediction scheme . For example , 17 of the over-predicted sites are located in the Palm domain ( positions 86–119 and 151–244 ) . Within it , sites 88W , 111V , 113D , 116F , 182Q and 233E were shown to be involved in primer loading 31 . Another over-predicted stretch of residues spans positions 216 to 243 ( the “primer grip ) that is involved in positioning the primer\s terminus 32 ., This region is almost invariant in the neutral data and is known to mutate after drug treatment ( as shown in the selective settings – both in the database and the modeled data ) ., An additional over-predicted segment spans position 251 to 271 ( the ‘helix clamp’ ) that is conserved among other nucleic acid polymerases 33 ., Several residues within these regions were not directly in contact with nucleic acid in any of the available structures but were previously shown to be important for the catalytic cycle of reverse transcription , providing a possible explanation for the over-predictions ., As with HIV-1 protease , we calculated ROC curves and AUC values for the reverse transcriptase model predictions to quantify overall performance ( Figure 7 ) ., The ROC curves show that the computational model correctly ranked many mutations tolerated by HIV-1 reverse transcriptase ., AUC values are generally slightly lower for reverse transcriptase than for protease , but exceed 80% ( black bars , Figure 7C ) ., In accordance with the results obtained for protease , predictions of mutational tolerance made using any single reverse transcriptase structure were worse than using an ensemble of experimentally determined structures or backrub ensembles computationally generated from a single template structure ( grey , black and orange curves and bars in Figure 7A–C ) ., In conclusion , although mutational tolerance predictions for reverse transcriptase were less accurate than for protease , the results still demonstrate reasonable agreement with mutations observed in the database ., The application of the model to reverse transcriptase also confirms the notion that using either ensembles of experimental | Introduction, Results, Discussion, Methods | Predicting which mutations proteins tolerate while maintaining their structure and function has important applications for modeling fundamental properties of proteins and their evolution; it also drives progress in protein design ., Here we develop a computational model to predict the tolerated sequence space of HIV-1 protease reachable by single mutations ., We assess the model by comparison to the observed variability in more than 50 , 000 HIV-1 protease sequences , one of the most comprehensive datasets on tolerated sequence space ., We then extend the model to a second protein , reverse transcriptase ., The model integrates multiple structural and functional constraints acting on a protein and uses ensembles of protein conformations ., We find the model correctly captures a considerable fraction of protease and reverse-transcriptase mutational tolerance and shows comparable accuracy using either experimentally determined or computationally generated structural ensembles ., Predictions of tolerated sequence space afforded by the model provide insights into stability-function tradeoffs in the emergence of resistance mutations and into strengths and limitations of the computational model . | Many related protein sequences can be consistent with the structure and function of a given protein , suggesting that proteins may be quite robust to mutations ., This tolerance to mutations is frequently exploited by pathogens ., In particular , pathogens can rapidly evolve mutated proteins that have a new function - resistance against a therapeutic inhibitor - without abandoning other functions essential for the pathogen ., This principle may also hold more generally: Proteins tolerant to mutational changes can more easily acquire new functions while maintaining their existing properties ., The ability to predict the tolerance of proteins to mutation could thus help both to analyze the emergence of resistance mutations in pathogens and to engineer proteins with new functions ., Here we develop a computational model to predict protein mutational tolerance towards point mutations accessible by single nucleotide changes , and validate it using two important pathogenic proteins and therapeutic targets: the protease and reverse transcriptase from HIV-1 ., The model provides insights into how resistance emerges and makes testable predictions on mutations that have not been seen yet ., Similar models of mutational tolerance should be useful for characterizing and reengineering the functions of other proteins for which a three-dimensional structure is available . | sequence analysis, biophysic al simulations, evolutionary modeling, biology, computational biology, macromolecular structure analysis | null |
journal.pcbi.1003584 | 2,014 | How the Brain Decides When to Work and When to Rest: Dissociation of Implicit-Reactive from Explicit-Predictive Computational Processes | Suppose that you are given a job whose payoff is proportional to the effort made within a limited time , say for instance the number of Christmas cards sold at the end of the day ., Maximizing your payoff would require running from house to house , but this effort would induce such fatigue that you decide to walk from time to time ., This sort of situation can be examined through economic decision theory , which would suggest you to write down the expected costs and benefits , and try to figure out whether the effort is worthy ., If the cost of a given effort is anticipated to increase with fatigue 1 , 2 , then you will find an optimal duration that can be determined before engaging any action ., Yet the literature on exercise performance has developed a different perspective on this issue 3 , 4 , which would suggest that you start by running , and only stop when some physiological variable , for instance in cardiovascular function ( such as heart beat rate ) or in muscular metabolism ( such as lactate concentration ) , attains a given limit 5 , 6 ., In other words , effort cessation would be a reaction to homeostatic failure , and would not require any explicit anticipation of effort cost ., These two extreme perspectives have obvious limitations ., The physiological view does not account for the effect of expectations that might pre-configure behavioral performance 4 , 7 , 8 ., The economic view does not integrate the constraints imposed by physiological reactions , which might be difficult to anticipate 9 ., Here , we intend to overcome these limitations by integrating the two perspectives into the same computational model ., Furthermore , we have built this model so as to explain the duration not only of effort exertion but also of rest ( recovery from fatigue ) ., Let us assume that a single waning and waxing variable triggers decisions to stop and restart effort exertion when reaching bounds ( see Figure 1A for a graphical presentation ) ., As this variable linearly accumulates during effort and dissipates at rest , it can be seen as a simple reflection of physiological reactions that predict the proximity of homeostatic failure ., Alternatively , it can be interpreted as tracking cost increase with fatigue , by integrating past effort over time ., Thus , the basic architecture of the model ( the accumulation-to-bound principle ) can account for implicit , online adaptation to actual effort costs , complying with physiological constraints ., On this basis , the modulation of the model latent parameters ( slopes and bounds ) could allow for anticipatory adjustments , depending on explicit costs and benefits ( see Figure 1B for a graphical illustration ) ., To dissociate the effects of actual and expected effort costs , we developed seven variants of a paradigm that was employed in a previous paper 10 to identify the neural underpinnings of the modeled variable , which we termed cost evidence ( see Figure 2 for an overview ) ., The task involved participants squeezing a handgrip with a given force , knowing that their payoff will be proportional to their effort duration ., Cost evidence can be manipulated by varying either an imposed duration or an imposed force ( task difficulty ) ., In a first study , we used three tasks that impose variable durations in order to verify that the behavior is adapted on the fly due to internal constraints ( bounds ) ., In a second study , we demonstrate that explicit ratings of subjective exhaustion do not follow the cost-evidence variable that accounts for the decision to stop effort exertion ., In a third study , we used three other tasks that vary the difficulty in order to dissociate the effects of expected and actual costs ., In our previous paper 10 , we suggested that the alternation of effort and rest periods observed in the Effort Allocation Task was well explained by a waning and waxing accumulation signal ., However , this cost-evidence signal that we localized in the brain could be epiphenomenal , in the sense that it would not reflect any causal mechanism triggering the decisions to stop and restart effort ., In this first study , we wished to verify that the level of cost evidence imposes actual constraints on subsequent behavior , as predicted by the accumulation-to-bound principle ., We therefore tested the predictions of the accumulation model on the behavior that followed an effort whose duration was imposed ., The difficulty was not manipulated in this study , for two reasons: firstly , the effect of difficulty was already shown in the previous paper 10 and will be further investigated in the following studies , and secondly , manipulation of difficulty only applies to effort periods , whereas manipulation of duration can be equally applied to both effort and rest periods ., Predictions of the accumulation model are that, 1 ) prolonging effort should decrease the next effort period ( if compensatory resting is not allowed ) ,, 2 ) prolonging rest should increase the next effort period ( up to a maximum corresponding to full recovery ) , and, 3 ) prolonging effort should increase the next rest period ( if compensatory resting is allowed ) ., These three predictions were tested in different groups of participants ( n\u200a=\u200a36 in total ) , using three variants of the Effort Allocation Task ., These three Adaptation Tasks had the same structure , with first an imposed effort ( between start and stop signals ) , second a rest period ( either fixed or free ) and third a free effort exertion ., Difficulty of both efforts was fixed at 60% of the maximal force , and payoff was proportional to the duration of the last effort , which was the main dependent measure ., Data were regressed at the individual level against a linear model that included the factor of interest ( the imposed duration ) and several potential confounds ( see methods ) ., The statistical significance of regressors was estimated at the group level using two-sided one-sample t-tests ., Results are given as standardized effect size ( beta ) ± inter-subject standard error of the mean ., In this task , cost evidence was increased by prolonging the first effort period ( from 1 to 10s ) , then the second effort duration was observed after a fixed 2-s rest ( Figure 3A ) ., To ensure that the rest duration was well controlled , we checked that initiation delay of the second effort after the go signal was not significantly impacted by the duration of the first effort ( 6 . 0 10−2±3 . 2 10−2 , df\u200a=\u200a11 , p\u200a=\u200a0 . 09 ) , by cumulated duration of efforts produced in the current session ( 1 . 1 10−2±3 . 4 10−2 , df\u200a=\u200a11 , p\u200a=\u200a0 . 76 ) , and by the session number ( −2 . 8 10−2±2 . 3 10−2 , df\u200a=\u200a11 , p\u200a=\u200a0 . 25 ) ., Critically , the second effort was significantly shortened by prolonging the first effort ( −8 . 29 10−1±2 . 3 10−1 , df\u200a=\u200a11 , p\u200a=\u200a0 . 0037 ) ., Next we examined the shape of the transfer function from imposed to observed effort duration ., The model predicts that this link should be negative , except if resting is long enough to fully dissipate the accumulated cost ., We therefore compared a model with pure negative correlation ( no saturation , #1 ) to models with an upper plateau ( over shortest efforts ) , followed by a decrease ., We tried two possibilities for this saturation effect: first a constant followed by a linear decrease ( #2 ) and second a negative exponential ( #3 ) ., The latter was implemented because it provides a better fit of plateau effects when data are noisy ( see methods ) ., Bayesian model selection revealed that the pure linear model was far better than the two saturation models in the family comparison ( model 1 versus models 2 & 3 ) , with an expected frequency ef\u200a=\u200a0 . 81 ( which is much higher than chance level - 1/2 ) and an exceedance probability xp\u200a=\u200a0 . 96 ( confidence that the model is more frequently followed than the others ) ., Thus , the result supports linear accumulation of cost evidence , which limits subsequent effort production due to the existence of an upper bound ., However , we found no evidence for the existence of a lower bound in cost dissipation , probably because our rest period was not long enough ., This limitation was overcome in the next task , where rest period was systematically varied ., This task ( Figure 3B ) was very similar to Task 1 , except that effort duration was now fixed ( to 7 s ) and rest duration was systematically varied ( from 1 to 12 s ) ., We checked again that subjects were not delaying effort initiation to compensate for variations in the imposed rest duration ( −3 . 7 10−2±2 . 1 10−2 , df\u200a=\u200a11 , p\u200a=\u200a0 . 10 ) ., In addition we found that the initiation delay was slightly affected by the cumulated duration of past efforts ( 1 . 2 10−2±5 . 0 10 −3 , df\u200a=\u200a11 , p\u200a=\u200a0 . 03 ) , but not by the session number ( −1 . 1 10−2±1 . 4 10−2 , df\u200a=\u200a11 , p\u200a=\u200a0 . 46 ) ., Critically , observed effort was significantly prolonged by longer rest ( 6 . 9 10−1±1 . 9 10−1 , df\u200a=\u200a11 , p\u200a=\u200a0 . 0035 ) ., Next we tested the existence of a saturation , meaning that beyond a certain rest duration , cost evidence is entirely dissipated and subsequent effort cannot be further prolonged ., As was done for the previous task , we compared three models for the link between rest and effort duration:, 1 ) a linear effect ( no saturation ) ,, 2 ) a linear effect bounded by an upper plateau ( over longest rests ) ,, 3 ) an exponential asymptotic plateau ., Bayesian model selection showed that the saturation family was now more plausible ( models 2 and 3 versus model 1 , chance level is 1/2 , ef\u200a=\u200a0 . 79 , xp\u200a=\u200a0 . 94 ) ., Direct comparison between models 2 and 3 revealed that the asymptotic saturation was more likely than the linear plateau ( xp\u200a=\u200a0 . 98 ) ., Thus , the results confirmed that prolonging rest after a first effort augments the capacity to produce a second effort , as if cost evidence was dissipated ., Moreover , the saturation effect suggests the existence of a threshold after which prolonging rest is useless , which would correspond to a lower bound for cost-evidence dissipation ., This task ( Figure 3C ) was quite similar to Task 2 , except that participants were not asked to resume their effort immediately at the go signal , but only when they felt ready to do so ., There were therefore two dependent variables of interest: rest duration and subsequent effort duration ., Critically , rest duration was significantly increased by prolonging the imposed effort duration ( 6 . 5 10−1±1 . 3 10−1 , df\u200a=\u200a11 , p\u200a=\u200a0 . 0005 ) ., We expected that participants would rest long enough to fully dissipate the first effort cost , which hence would have no impact on the second effort duration ., This was not the case: prolonging the first effort significantly shortened the second effort ( −4 . 7 10−1±1 . 4 10−1 , df\u200a=\u200a11 , p\u200a=\u200a0 . 006 ) ., Thus , subjects did not wait long enough to compensate for the imposed effort cost ., This partial recovery might be related to the fact that the total time allowed for rest and effort was limited to 20s , so that participants may have shortened rest to make sure there would be enough time for effort ( even if in reality , 20s was largely enough to fully dissipate and accumulate cost again ) ., So far , our results suggest that effort duration is not entirely planned in advance but adapted on the fly so as to keep cost evidence within pre-defined bounds ., The next study was designed to assess whether our participants could explicitly report the cost evidence that was monitored by their brain in order to regulate their behavior ., The first study only manipulated the duration of effort or rest periods ., Yet our model posits that cost evidence accumulation during effort depends on task difficulty ., Therefore , cost-evidence level should reflect the interaction of task difficulty and effort duration ., The logic of this second study was first to examine whether introspective reports would reflect the interaction of difficulty and duration , and then to verify that behavioral choices were indeed driven by this interaction , For introspective reports we asked a new group of 18 participants to perform a Cost Rating Task , in which they had to rate their degree of exhaustion after effort exertion ., Note that we could have directly inserted cost ratings within the Effort Allocation Task , but subjects in this case might have artificially aligned their behavior to their explicit reports ( or vice-versa ) ., Another issue with this possibility was that effort duration would not have been sufficiently varied , at least not orthogonally to effort difficulty , since subjects would have stopped their effort when cost evidence ( difficulty times duration ) reached a pre-defined bound ., We chose to frame the question in terms of exhaustion because debriefing of previous studies revealed that exhaustion is the intuitive term that subjects spontaneously use to describe the sensation that makes them cease their effort ., The precise question was ‘Have you exhausted your resources ? ’ and the response scale was ranging from ‘not at all’ to ‘completely’ ., In this Cost Rating Task , both effort duration ( from 3 to 7 s ) and task difficulty ( from 40 to 60% of maximal force ) were imposed and varied experimentally ( Figure 4A ) ., To keep similarity with the Effort Allocation Task , we also manipulated the incentive level ., Yet we acknowledge that the comparison between tasks has limitations , first because they implement different range of forces and durations , second because they are performed by different subjects , who might have different sensitivity to effort cost ., On each trial , the payoff was calculated as the incentive multiplied by the fraction of the imposed duration that subjects spent squeezing at the required target force level or higher ., As participants were asked to be as accurate as possible , this fraction was almost 100% ( mean over subjects: 98 . 7% , extreme subjects: 94 . 6% and 99 . 9% ) ., The difference between required and produced force levels did not vary significantly across conditions ( multiple regression analysis and two-sided t-test with df\u200a=\u200a17; incentive: 4 . 1 10−3±3 . 7 10−3 , p\u200a=\u200a0 . 28; duration: −3 . 4 10−6±4 . 6 10−3 , p\u200a=\u200a0 . 99; difficulty: −1 . 8 10−3±3 . 3 10−3 , p\u200a=\u200a0 . 59; interactions between these factors: all p>0 . 21 ) , suggesting that effort production was well controlled by the experimental design ., Cost ratings were not significantly impacted by incentives ( 1 . 4±0 . 86 , df\u200a=\u200a17 , p\u200a=\u200a0 . 1 ) , and marginally by the initial position of the cursor on the scale ( 1 . 8±0 . 9 , df\u200a=\u200a17 , p\u200a=\u200a0 . 056 ) ., Critically , cost ratings increased with both duration ( 1 . 9±0 . 79 , df\u200a=\u200a17 , p\u200a=\u200a0 . 028 ) and difficulty ( 3 . 2±0 . 49 , df\u200a=\u200a17 , p\u200a=\u200a5 10−6 ) , without significant interaction between these factors ( p\u200a=\u200a0 . 96 ) ., We then fitted cost ratings with a linear combination of regressors meant to capture the impact of duration and difficulty ., We considered three possibilities: main effects of duration and difficulty , non-linear effects ( power functions ) of duration and difficulty , and interaction between duration and difficulty ., Including or not each possibility in the linear combination made a total of eight models , which we compared using Bayesian model selection ( Figure 4C ) ., This analysis confirmed the absence of significant interaction between duration and difficulty , since the best model was simply additive ( chance level is 1/8 , ef\u200a=\u200a0 . 48; xp\u200a=\u200a0 . 93 ) ., In principle , this additive effect could arise from half the subjects reporting duration and the other half reporting difficulty ., This would imply that the effect sizes of these factors are anti-correlated across subjects ., We found the opposite result ( Pearson rho: 0 . 82 , df\u200a=\u200a16 , p\u200a=\u200a3 10−5 ) , suggesting that subjects who were good at perceiving duration were also good at perceiving difficulty ., Yet they reported the addition of the two dimensions , and not their product , as should be the case if they were simply introspecting cost evidence ., We next re-analyzed the behavioral choices observed in our Effort Allocation Task ( Figure 4B ) that involved subjects ( n\u200a=\u200a38 ) squeezing a handgrip in order to accumulate as much money as possible 10 ., The payoff was calculated as the monetary incentive multiplied by the time spent above a target force level ( which indexed task difficulty ) ., Both the incentive ( 10 , 20 or 50 cents ) and difficulty levels ( 70 , 80 or 90% of maximal force ) were varied across trials such that we could assess their effects on effort allocation ., Incentive levels were sufficient for subjects to initiate the effort and to reach the target , but difficulty levels were too demanding for subjects to sustain their effort throughout trials , which lasted 30 seconds ., Instead , they freely alternated effort and rest periods within trials ( as can be seen in Figure 1A ) ., We used the normalized cumulative distribution of effort durations to calculate the probability of stopping the effort after a given duration at a given difficulty level ., This probability was fitted with a sigmoid function of cost-evidence level , which accounts for higher cost evidence making effort cessation more likely ., Cost evidence was then modeled with the same linear combinations as used for fitting cost ratings ., Results of Bayesian model selection ( Figure 4D ) showed that the most plausible model was pure interaction ( chance level is 1/8 , ef\u200a=\u200a0 . 62 , xp\u200a=\u200a0 . 988 ) ., The Cost Rating and Effort Allocation tasks thus elicited distinct forms of cost evidence , with additive versus multiplicative effect of effort difficulty and duration ., The critical difference is the shape of iso-value lines of cost evidence in the duration by difficulty space , with straight lines for explicit report and convex lines for effort cessation ( compare Figures 4E and 4F ) ., To directly compare the curvature of cost evidence inferred from introspective reports and behavioral choices , we fitted a model with constant elasticity of substitution ( CES ) between duration and difficulty ( see methods ) ., This model has a free parameter that captures the curvature of cost in the duration by difficulty space , which should be equal to one in the absence of interaction , and below one in the case of a convex interaction ., We found that the curvature parameter was significantly below one in the Effort Allocation Task ( median: 0 . 52 , SEM: 0 . 06; two-sided sign-test of the median against 1: p\u200a=\u200a6 . 7 10−8 ) but not in the Cost Rating Task ( median: 1 . 01 , SEM: 0 . 12; sign-test of the median against 1: p\u200a=\u200a1 ) , with a significant difference between tasks ( p\u200a=\u200a3 10−6 , two-sided Wilcoxon rank sum test for equal medians ) ., When debriefing the Cost Rating Task , participants unambiguously reported having noticed variations in both difficulty and duration ., When asked whether one of these two factors had a greater impact on their ratings , 13 subjects favored the duration , 3 favored the difficulty , and 2 could not favor one or the other , describing something like an interaction ., However , comparison of standardized effect size revealed a greater impact of difficulty on ratings ( paired t-test on duration minus difficulty effect size: −1 . 3±0 . 48 , df\u200a=\u200a17 , p\u200a=\u200a0 . 016 ) ., Among the 16 subjects who favored a main effect , 12 got it wrong ( the other factor had a higher impact on their ratings ) , which is more than expected by chance ( binomial test , p\u200a=\u200a0 . 028 ) ., To summarize , the costs reported in subjective ratings do not have the same shape as the costs inferred from behavioral choices ., What subjects report is an addition of duration and difficulty , whereas what drives their behavior is an interaction between the two ., Furthermore , at a meta-cognitive level , subjects have poor insight into the factors that modulate their sensation of exhaustion ., The two studies presented so far are compatible with a completely implicit and automatic model , in which decisions to cease and resume effort production are controlled by an internal variable fluctuating between bounds that might be determined by physiological constraints ., In this last study , we explored whether explicit information about cost could impact the mechanics driving decisions to start and stop effort exertion ., In our previous paper 10 , we had observed that task difficulty shortened effort duration , which could reflect cost evidence ( difficulty times duration ) reaching the upper bound , but did not affect rest duration ., We hypothesized that the last observation could arise from task difficulty not being made explicit to participants ., Indeed , monetary incentives , contrary to difficulty levels , were explicitly presented with coin images at trial start and affected both effort and rest durations ( with longer effort and shorter rest for higher incentive ) ., We therefore tested whether providing explicit information about difficulty level would change the way participants process cost evidence ., We constructed three variants of the Effort Allocation Task , which were administered to three different groups of participants ( n\u200a=\u200a67 in total ) ., In all tasks , incentives ( coin images ) were explicitly displayed before and during trials , which had a fixed duration ( 30s ) that was specified to participants prior to the experiment ., The Implicit Task ( Figure 4B ) is the task used in our previous paper 10 , with no visual cue for difficulty level ., In the Explicit Task , the only change is that difficulty level ( percentage of maximal force: 70 , 80 or 90% ) was announced before the beginning of trials , on the same screen as incentive level ., In the Dissociation Task , we kept the explicit cues , but they were no longer predictive of the actual task difficulty ., To maintain sufficient statistical power , only two difficulty levels were used ( 75 and 85% ) , in a full factorial design ( two cued difficulties crossed by two actual difficulties ) ., This design was meant to disentangle the effects of implicit versus explicit cost processing ., Monetary incentives were also manipulated in all tasks and crossed with the three ( Implicit and Explicit Tasks ) or four ( Dissociation task ) cells corresponding to variations in difficulty ., We only used two incentive levels ( 10 versus 20c ) in the Dissociation task to avoid combinatorial inflation ., In every task , the effect of experimental factors ( incentive , actual and cued difficulty ) on the duration of effort and rest epochs were estimated in separate multiple linear regressions followed by two-sided one-sample t-tests ., Note that because they must add up to 30s , the cumulative durations of effort and rest are anti-correlated ., However , this dependency was broken first because the last rest epochs were discarded from the analysis , since they are interrupted by trial ending , and second because we considered the single epoch durations , which are not predictable from the cumulative durations , since they depend on the number of alternations between effort and rest ., The remaining correlation was rather low ( Pearson rho: −0 . 15±0 . 026 in the main Implicit Effort Allocation Task ) and probably due to opposite effects of experimental factors ( see below ) ., As previously shown 10 , in the Implicit Task ( Figure 5 , left ) , effort duration was both longer for higher incentive ( 1 . 5±0 . 26 , df\u200a=\u200a37 , p\u200a=\u200a8 . 1 10−7 ) and shorter for higher difficulty ( −1 . 1±0 . 13 , df\u200a=\u200a37 , p\u200a=\u200a1 . 6 10−10 ) ., In contrast , rest duration was shorter for higher incentive ( −0 . 37±0 . 08 , df\u200a=\u200a37 , p\u200a=\u200a2 . 0 10−5 ) but was not modulated by the difficulty ( 0 . 03 , ±0 . 03 , df\u200a=\u200a37 , p\u200a=\u200a0 . 32 ) ., Interactions were included in the regression model , but the incentive x difficulty interaction was not significant , neither for effort or for rest duration ( all p>0 . 084 ) ., All significant results were replicated in the Explicit Task ( Figure 5 , middle ) : effort duration was both longer for higher incentive ( 2 . 2±0 . 53 , df\u200a=\u200a13 , p\u200a=\u200a1 . 1 10−3 ) and shorter for higher difficulty ( −1 . 8±0 . 24 , df\u200a=\u200a13 , p\u200a=\u200a6 . 0 10−6 ) , and rest duration was shorter for higher incentive ( −0 . 4±0 . 09 , df\u200a=\u200a13 , p\u200a=\u200a9 . 7 10−4 ) ., The novel result is that rest duration was now increased by higher difficulty ( 0 . 31±0 . 08 , df\u200a=\u200a13 , p\u200a=\u200a1 . 6 10−3 ) , which was correctly cued at trial start ., The difference in standardized effect sizes between Implicit and Explicit Tasks was also significant ( p\u200a=\u200a1 . 2 10−4 , unpaired t-test , df1: 37 , df2: 13 ) ., All interactions remained non-significant , neither for effort or rest duration ( all p>0 . 1 ) ., Thus , the difficulty in the Explicit Task , which was both expected and experienced during effort exertion , affected both effort and rest durations ., The results obtained with the Implicit and Explicit Tasks are compatible with the actual difficulty affecting effort duration , and the expected difficulty affecting rest duration ., In the Implicit Task , there was no explicit cue , so subjects did not expect any particular difficulty level , and consequently only effort duration ( not rest duration ) was affected by task difficulty ., In the Explicit Task , both effort and rest durations were modulated because the actual difficulty was fully expected ., However , as the explicit cues were perfectly valid , we could not formally demonstrate with this task that rest duration is not concerned with actual difficulty , or that effort duration is not concerned with expected difficulty ., To complete our demonstration , we intended to dissociate the two effects within the same task ., In the Dissociation Task ( Figure 5 , right ) , the levels of actual and cued difficulty were manipulated independently ., As in the Implicit and Explicit tasks , higher incentive increased effort duration ( 0 . 42±0 . 16 , df\u200a=\u200a14 , p\u200a=\u200a0 . 022 ) and shortened rest duration ( −0 . 22±0 . 06 , df\u200a=\u200a14 , p\u200a=\u200a1 . 5 10−3 ) ., Effort duration was affected by the actual ( −0 . 47±0 . 18 , df\u200a=\u200a14 , p\u200a=\u200a0 . 021 ) but not by the cued difficulty ( 0 . 07±0 . 15 , df\u200a=\u200a14 , p\u200a=\u200a0 . 64 ) ., The difference in standardized effect size was at significance limit ( −0 . 54±0 . 25 , df\u200a=\u200a14 , p\u200a=\u200a0 . 050 , paired t-test ) ., We also verified that the effect of cued difficulty on effort duration in the Dissociation Task was significantly lower than the ( actual ) difficulty effects observed in the Implicit ( p\u200a=\u200a4 . 3 10−7 , unpaired t-test , df1: 37 , df2: 14 ) and Explicit ( p\u200a=\u200a2 . 3 10−6 , unpaired t-test , df1: 14 , df2: 13 ) tasks ., Conversely , rest duration was affected by the cued ( 0 . 22±0 . 06 , df\u200a=\u200a14 , p\u200a=\u200a1 . 7 10−3 ) but not by the actual ( 0 . 03±0 . 06 , df\u200a=\u200a14 , p\u200a=\u200a0 . 63 ) difficulty ., The difference in standardized effect size was as well significant ( −0 . 19±0 . 08 , df\u200a=\u200a14 , p\u200a=\u200a0 . 045 , paired t-test ) ., We also verified that the effect of cued difficulty on rest duration was higher in the Dissociation Task than the ( actual ) difficulty effect observed in the Implicit Task ( p\u200a=\u200a0 . 002 , unpaired t-test , df1: 37 , df2: 14 ) , and that the effect of actual difficulty in the Dissociation Task was lower than the ( cued ) difficulty effect observed in the Explicit Task ( p\u200a=\u200a0 . 008 , unpaired t-test , df1: 14 , df2: 13 ) ., Thus , within- and between-task comparisons both support a double dissociation between the actual and cued difficulty effects on effort and rest durations ., As some critical p-values were near 0 . 05 type I error rate , we conducted a permutation test to ensure the reliability of the parametric t-distribution in our small sample ., This permutation-based t-distribution yielded the same exact p-values up to the 3rd decimal ., Second and third order interaction terms between incentive , cued and actual difficulty were included in the model , but none of them was significant neither for rest or effort duration ( all p>0 . 18 ) ., We also checked that there was no interaction of cued difficulty with time , which could potentially reflect a progressive discount of the cue effect ( as subjects would learn that cues are not predictive of actual difficulty ) ., Time was modeled at three nested scales ( rest or effort period position within a trial , trial position within a session , and session number ) ., Two-way interactions with cued difficulty were estimated for each time scale: none of them was significant ( all p>0 . 25 ) ., We compared different versions of our accumulation model to identify how the latent parameters ( A: amplitude between bounds , SE: accumulation slope during effort , and SR: dissipation slope during rest ) were affected by the experimental factors ( I: Incentive , Da: actual difficulty , Dc: cued difficulty ) ., We started with the formalization that we proposed in our previous publication 10 to account for the behavior observed in the Implicit Task ., All models were built as a set of three equations that defines each latent parameter as a linear combination of the different factors ( see methods ) ., Only models that can produce the behavioral results ( significant effect on effort or rest duration ) were included in the space covered by Bayesian Model Selection ., In the Implicit Task , this left 24 possible models ( see Figure 6A ) with one that was much more plausible than the others ( chance level is 1/24 , ef\u200a=\u200a0 . 30 , xp\u200a=\u200a0 . 90 ) ., For the novel tasks ( Explicit and Dissociation ) , we explored two possibilities for integrating the additional factor ( cued difficulty ) ., The first possibility was to integrate it as an additive term , just as was done with actual difficulty ( see Figure 6B and 6C ) ., Note that these purely linear models do not enable dissociating the effects of actual and expected difficulty in the Explicit Task ., The second possibility was to integrate cued difficulty as a hyperbolic discounter of incentives , which is quite standard in the literature for capturing temporal discounting 11–13 ., Thus , for the novel tasks that manipulate expected difficulty , we included the hyperbolic equivalent of our linear models ( see Figure 6D ) ., With this hyperbolic version , we can dissociate the effect of actual and expected difficulty ( the former is linear , the latter hyperbolic ) even in the Explicit Task where the two factors are confounded ., Family comparison revealed that there was far more evidence in favor of a hyperbolic rather than linear discount of incentives by cued difficulty , in both the Explicit and Dissociation tasks ( chance level is 1/2 , ef>0 . 91 , xp>0 . 999 ) ., Among the 78 possible hyperbolic models , a best model was identified with xp\u200a=\u200a0 . 90 ( chance level is 1/78 , ef\u200a=\u200a0 . 13 ) in the Dissociation Task and with xp\u200a=\u200a0 . 82 ( chance level is 1/78 , ef\u200a=\u200a0 . 14 , ) in the Explicit Task ., Crucially , the best hyperbolic model identified in the Explicit and Dissociation tasks was the same model , which also corresponded to the best model identified in the Implicit Task ( where modulation by cued difficulty is necessarily absent ) ., This best model is written as follows ( Te and Tr being effort and rest duration , α , β | Introduction, Results, Discussion, Methods | A pervasive case of cost-benefit problem is how to allocate effort over time , i . e . deciding when to work and when to rest ., An economic decision perspective would suggest that duration of effort is determined beforehand , depending on expected costs and benefits ., However , the literature on exercise performance emphasizes that decisions are made on the fly , depending on physiological variables ., Here , we propose and validate a general model of effort allocation that integrates these two views ., In this model , a single variable , termed cost evidence , accumulates during effort and dissipates during rest , triggering effort cessation and resumption when reaching bounds ., We assumed that such a basic mechanism could explain implicit adaptation , whereas the latent parameters ( slopes and bounds ) could be amenable to explicit anticipation ., A series of behavioral experiments manipulating effort duration and difficulty was conducted in a total of 121 healthy humans to dissociate implicit-reactive from explicit-predictive computations ., Results show, 1 ) that effort and rest durations are adapted on the fly to variations in cost-evidence level ,, 2 ) that the cost-evidence fluctuations driving the behavior do not match explicit ratings of exhaustion , and, 3 ) that actual difficulty impacts effort duration whereas expected difficulty impacts rest duration ., Taken together , our findings suggest that cost evidence is implicitly monitored online , with an accumulation rate proportional to actual task difficulty ., In contrast , cost-evidence bounds and dissipation rate might be adjusted in anticipation , depending on explicit task difficulty . | Imagine that ahead of you is a long time of work: when will you take a break ?, This sort of issue – how to allocate effort over time – has been addressed by distinct theoretical fields , with different emphasis on reactive and predictive processes ., An intuitive view is that you start working , stop when you are tired , and start again when fatigue goes away ., Biologically , this means that decisions are taken when some physiological variable reaches a given bound on the risk of homeostatic failure ., In a more economic perspective , fatigue translates into effort cost , which must be anticipated and compared to expected benefit before engaging an action ., We proposed a computational model that bridges these perspectives from sport physiology and decision theory ., Decisions are made in reaction to bounds being reached by an implicit cost variable that accumulates during effort , at a rate proportional to task difficulty , and dissipates during rest ., However , some latent parameters ( bounds and dissipation rate ) are adjusted in anticipation , depending on explicit costs and benefits ., This model was supported by behavioral data obtained using a paradigm where participants squeeze a handgrip to win a monetary payoff proportional to effort duration . | medicine and health sciences, decision making, social sciences, neuroscience, cognitive neuroscience, cognitive psychology, human performance, cognition, behavior, consciousness, pain, psychology, pain management, motivation, psychophysics, biology and life sciences, sensory perception, cognitive science, motor reactions | null |
journal.pcbi.1006059 | 2,018 | Population dynamics of engineered underdominance and killer-rescue gene drives in the control of disease vectors | Mosquito-borne diseases represent one of the most severe public health burdens worldwide ., For example , dengue viruses that are primarily transmitted by Aedes aegypti mosquitoes have rapidly increased in prevalence in recent years 1 ., One recent study estimated that around 3 . 9 billion people , in over 100 countries , live in regions ‘at risk’ for dengue infections 2 , with ∼390 million dengue infections per year 3 of which perhaps 50-100 million cases are symptomatic 4 ., Of these cases ∼3 . 9 million are classified as severe and 9 , 000 are fatal 4 ., The methods currently used to control dengue do not appear sufficient to eliminate the problem and this is exacerbated by the lack of drug treatments presently available 5 ., A first dengue vaccine has recently been licensed , however it is only recommended for use in individuals over nine years of age that live in areas of high dengue burden 6 , 7 ., Thus , while it may prove a useful tool in some situations , it appears unlikely to be sufficient for eliminating the threat of dengue ., As such , a number of additional methods for the control of dengue and other vector-borne diseases are currently being investigated ., Genetic control methods are one such alternative and , with advances in tools available to molecular biologists , have become a realistic prospect in recent years ., In particular , a range of gene drive systems have been proposed that could , in theory , be used to spread desirable genetic traits through a mosquito population 10 ., Each of these systems would be implemented by introducing individuals carrying the drive system into the wild where they mate with existing mosquito populations ., This potentially gives them an edge over traditional control methods since they exploit the natural behaviour of mosquitoes to seek mates and find breeding/resting sites which can be extremely difficult for humans to locate and reach ., These genetic systems can be classified in a number of ways that include their intended effect , persistence and invasiveness , each of which affect how they may be viewed both by the public and regulators 10 ., Two such classes of gene drive that are currently under development in our research group for Aedes aegypti mosquitoes are two-locus engineered underdominance ( UD ) and killer-rescue ( KR ) ., These have both attracted attention due to their potential to mitigate a number of key regulatory concerns ., Underdominance is a phenomenon most commonly thought of in the context of two or more alleles at a single genetic locus ., It refers to the scenario whereby individuals heterozygous at a given genetic locus are less fit than either of the homozygote states and is the inverse of the better known hybrid vigour ., UD systems , as proposed by Davis et al . 8 , seek to attain a similar effect through a reduction in the fitness of hybrids between parental strains ., This is achieved via the introduction of two independently inherited transgenic constructs inserted at unlinked loci ( Fig 1 ) ., Each of these transgenic constructs carries a lethal genetic element , a suppressor for the lethal at the other locus and a “cargo” gene conferring a desirable phenotype ., The requirement for lethal effectors within UD systems leads to a number of challenges for design and construction of such systems ., It is theoretically possible to engineer these systems sequentially ., However , their development and testing will be much simpler if lethal genes are either conditionally lethal , repressible or incompletely lethal ., Within this study we restrict our attention to cargo genes that render individuals refractory to one or more viruses ( i . e . a reduced vector competence trait ) , significantly reducing their capacity for infecting humans ., Further , this trait is assumed to be dominant meaning it will be fully effective in a single copy ., Such refractory genes have already been developed , for example those of Franz and colleagues 11 , 12 , with further examples likely in development ., Under such a system 8 , individuals ( other than wild-type ) that do not inherit at least one copy of each transgenic construct are non-viable because they carry one of the lethal genes but lack the relevant suppressor ., This creates a selection pressure for individuals to carry both constructs or neither ., Under certain conditions this can theoretically allow the transgenic constructs , and crucially , the cargo ( refractory ) gene ( s ) to spread toward fixation within a population if they are introduced at a sufficiently high frequency ., Due to the ongoing selection pressure for individuals to carry both transgenic constructs , it is expected that , in absence of resistance or mutation , the introduced transgenes will persist indefinitely ., Thus , it is anticipated that in such situations the prevalence of infections within a targeted area could be significantly reduced , or even eliminated ., The UD system is essentially composed of two orthogonal KR systems 9 , and a key developmental milestone will be the construction and testing of one/both KR system ( s ) ., Hence , we consider here the behaviour of this class of gene drive system also ., In this case one lethal ( killer ) and one suppressor ( rescue ) gene are independently inherited and inserted at unlinked loci ( Fig 1 ) ., As long as it is common within the population , the deleterious effect of the lethal transgene creates a selective pressure for individuals to also carry copies of the suppressor gene ., This causes the suppressor ( rescue ) construct to increase in frequency within the population ., However , this selective pressure is diminished as the lethal becomes more rare ( due to fitness costs ) ., Thus the suppressor too begins to reduce in frequency ( assuming it confers some fitness cost ) ., In theory this eventually leads to elimination of both transgenic constructs and thus a return to the pre-release state ., Such systems are interesting for the control of dengue vectors since they may reduce the number of mosquitoes capable of transmitting viruses sufficiently to disrupt disease transmission ., If so , the pathogen could be eliminated without removing the mosquito species from its ecological niche ., The power of such systems to alter natural populations has led to some concerns about ways in which they should be regulated to ensure they are used safely 13 , 14 ., Thus it is important to consider how such systems may be viewed in terms of a number of key issues ., Questions considered important in this context include the following ., Will a gene drive spread into regions that did not authorise the use of genetically modified mosquitoes ?, If so , what effects will they have in the non-target area ?, Can the gene drive be reversed/recalled in the event of unanticipated/undesirable effects ?, These questions will be addressed in the work presented here ., Previous modelling work has demonstrated that—in absence of resistance or mutation—UD systems should persist indefinitely provided that they are introduced above some threshold transgene frequency 8 , 15 ., Below this threshold the transgene will simply be eliminated from the population due to both the associated fitness costs and genetic drift ., For realistic rates of dispersal between two neighbouring mosquito populations , it has been shown that the flow of transgenes from a targeted population is not likely to be sufficient to exceed the threshold in the neighbouring population 16 , 17 ., This implies that the UD gene drive system is unlikely to spread , at any significant frequency , into neighbouring regions ., These results have been considered within a discrete population structure , however isolation-by-distance may be a more appropriate consideration for some populations ., A number of previous studies have considered such a population structure in the context of an advantageous gene or chromosome rearrangements ( e . g . 60–64 ) but , to our knowledge , not in the context of two-locus gene drive systems such as those studied here ., The existence of the threshold transgene frequency should also allow for the UD system to be recalled by introducing sufficient numbers of wild-type mosquitoes back into the population such that the transgene frequency falls below the threshold , thus causing the effect of the fitness cost to exceed the strength of the genetic drive mechanism ., Due to their self-limiting nature , KR systems may be viewed as an ideal system with which to test the effects of a given cargo gene before they are incorporated into self-sustaining systems ., However , in spite of the fact that KR is inherently self-limiting , the same regulatory questions are still likely to need addressing ., Modelling work has demonstrated that KR transgenes can spread into neighbouring populations at appreciable frequencies and then be eliminated due to their self-limiting nature ., For example , the modelling work of Marshall & Hay 16 considered the case of a KR system with heterozygous and homozygous fitness costs of 2 . 5% and 5% , respectively for each allele and 1% migration between two neighbouring populations ., When considering an initial transgene frequency of 0 . 5 , maximum transgene frequencies of ∼0 . 95 and ∼0 . 3 were reached in the target and neighbouring populations , respectively ., To our knowledge there have been no formal recall mechanisms detailed for KR systems as they should naturally diminish in time ., However , releasing wild-type individuals into the population would lower the transgene frequency , and should result in more rapid transgene elimination ., Much of the previous modelling work on UD and KR systems has assumed a discrete ( non-overlapping ) generation framework in order to study the population genetics of these systems 8 , 9 , 15 , 16 , 18 ., The works of Huang et al . 19 , 20 considered the effects of age and spatial structure in a mosquito population in the context of the UD system ., A large number of ecological factors have been included in studies of KR systems using the simulation model Skeeter Buster 21 , 22 , however the individual effects of each are not easily separated in such a complex model ., Here we aim to further the understanding of how ecological factors affect the efficacy of UD gene drive systems and disentangle their effects from one another in the case of KR systems ., Here we begin by formulating population dynamics models of UD and KR gene drive systems that include a number of ecological factors such as the number of breeding sites; strength of density dependent larval competition; population growth rate; adult mortality rate; and rates of migration between two populations ., These models also capture a number of characteristics of the control measures including the amount of mosquitoes released; the sexes released ( release of males and females in 1:1 ratio ( “bisex release” ) or male-only release ) ; fitness costs of introduced transgenes; sexes affected by lethal effects ( bisex lethality or female-specific ) ; whether one copy of a suppressor construct is sufficient to counter two copies of the related lethal construct; and the timing of the fitness/lethal effects ( early- or late-acting ) ., These models are first used to investigate the effects of the ecological factors considered here on the equilibrium size of a mosquito population in absence of control ., We then go on to explore how such variation in mosquito population size ( due to numbers of breeding sites ) impacts upon UD or KR gene drive systems in terms of threshold transgene frequencies , relative degrees of population suppression and the time-scales of action ., The strength of density dependent larval competition is then considered in terms of its effects on the observed dynamics of UD and KR systems as well as the thresholds for transgene introgression ( UD ) or increases in rescue transgene frequency ( KR ) ., Finally , using adjusted versions of the models we consider how rates of migration between a target and a non-target mosquito population can impact upon the efficacy of UD and KR gene drives and discuss results in terms of the desirability of observed outcomes ., Whilst we present results for Ae ., aegypti mosquito populations , we expect the models presented to be equally applicable for other species given a suitable parameterisation of the models ., Here we present a population dynamics model that may be used flexibly to describe both the UD and KR systems shown in Fig 1 ., With appropriate choices of lethality parameters , we anticipate this model could also be used to represent some alternative classes of gene drive such as Medusa 58 and reciprocal chromosome translocations 59 ., We also believe this model to be both simple and general enough as to be equally applicable to other species ., However , we restrict our attention to Ae ., aegypti due to both the ongoing development of these systems and the availability of parameter values in the previous literature ., This model is based on those of Yakob & Bonsall 23 and Alphey & Bonsall 24 in that it is a deterministic representation of a panmictic ( randomly mating ) population with continuous reproduction and a 1:1 male to female ratio both in the initial population and in the eggs laid in subsequent generations ., Further , this model assumes that transgenic constructs do not display any sex linkage , will not mutate and the individual genetic components at a given locus will not separate ., We also assume here that resistance alleles will not emerge within the population ., In such systems , the introduction of two independently inherited transgenic constructs at unlinked loci results in a total of nine possible genotypes ., Within this work these are referred to by a collection of four letters depending on the presence ( A and B ) or absence ( a and b ) of the two transgenic constructs and assigned a number ( i = 1 , … , 9 ) in order to simplify notation within the mathematical model ( see Table 1 ) ., For example , an individual that is heterozygous for both transgenic constructs will be denoted AaBb and assigned a genotype number i = 5 ., We consider the overall fitness ( Ωi ) of each genotype ( relative to wild-type individuals ) to be composed of the lethal effects from transgenic constructs ( γi ) and the fitness costs associated with the insertion of transgenes into the mosquito genome ., Within the mathematical model these effects are captured by, Ω i M , F = ϵ A η i A ϵ B η i B ( 1 - γ i M , F ) , ( 1 ), where ϵA and ϵB denote the relative fitness of individuals carrying one copy of transgenic constructs A and B , respectively ., These fitness effects manifest themselves as differences in the rates of survival of transgenic individuals relative to their wild-type counterparts ., Note that the fitness cost associated with a particular choice of relative fitness parameter may be calculated as 1 − ϵA or 1 − ϵB ., These are applied multiplicatively such that , for example , an individual carrying two copies of construct A has relative fitness ϵ A 2 while a double heterozygote has relative fitness ϵ A 1 ϵ B 1 ., For details on the number of copies of each transgenic construct carried by a given genotype see Table 1 ., Here a value of ϵ = 1 means an individual is as fit as a wild-type individual whereas ϵ = 0 represents a completely non-viable individual ., The other component of an individual’s overall fitness is the lethal effect of the carried transgenes ( γi ) ., In this case γi = 0 means that individuals of genotype i are 100% viable whereas individuals with γi = 1 receive a 100% lethal effect ., Note that some systems may exhibit incomplete penetrance of lethal transgenes ( i . e . 0 < γi < 1 ) , however this is beyond the scope of this study ., Parameter values describing the lethal effects of transgenes on each genotype in UD and KR systems are summarised in Table 1 ., Note that as in a previous population genetics study of UD 18 we here consider the possibility that different numbers of suppressor copies may be required to nullify the effect of the associated lethals ., These are termed a “strongly suppressed” system when one suppressor copy is sufficient to nullify two copies of a given lethal or a “weakly suppressed” system where two suppressor copies are required to nullify the effect of two lethal copies ., It is now possible to outline a set of equations describing the dynamics of each mosquito genotype in time ., As with the work of Alphey & Bonsall 24 , the models presented here are adaptations of Kostitzin’s 33 work that used Lotka-Volterra type equations to represent competition between different genotypes ., These models also consider a nonlinear representation of intraspecific competition due to Maynard Smith & Slatkin 34 that can describe a wide range of density dependence scenarios 35 ., We also consider two sets of equations representing distinct scenarios in terms of the developmental stage at which fitness/lethal effects of transgenic constructs act ., In particular , we consider the cases where these effects are either “early-acting”; taking effect before density-dependent competition ( e . g . in eggs/early instar larvae ) or “late-acting”; having an impact after density-dependent competition and before mating ( eg . in pupae or pharate adults ) ., This results in the following equations Early-acting fitness/lethal effects:, d M i ( t ) d t = ρ v i ( t - τ ) Ω i M 1 + ( ∑ i ∈ σ α v i ( t - τ ) Ω i M + w i ( t - τ ) Ω i F ) β - μ M i ( t ) , ( 2 ), d F i ( t ) d t = ρ w i ( t - τ ) Ω i F 1 + ( ∑ i ∈ σ α v i ( t - τ ) Ω i M + w i ( t - τ ) Ω i F ) β - μ F i ( t ) , ( 3 ) Late-acting fitness/lethal effects:, d M i ( t ) d t = ρ v i ( t - τ ) Ω i M 1 + ( ∑ i ∈ σ α v i ( t - τ ) + w i ( t - τ ) ) β - μ M i ( t ) , ( 4 ), d F i ( t ) d t = ρ w i ( t - τ ) Ω i F 1 + ( ∑ i ∈ σ α v i ( t - τ ) + w i ( t - τ ) ) β - μ F i ( t ) ., ( 5 ) Note that these two models are identical except for the removal of the fitness/lethal effect parameter ( Ω ) from the denominator of the late-acting model since these effects do not act until after the density dependent phase ., A limitation of such models is that density dependent competition may only occur between larvae born at the same time ., Definitions of parameters and variables are given in Table 2 alongside a base set of parameter values used throughout this study ., Expressions for vi and wi are outlined in S1 Appendix Section 1 and S1 Table alongside details regarding the implementation of release ratios ( and therefore initial conditions ) and transgene frequency calculations ., As discussed above , one of the key concerns associated with gene drive systems is the degree to which they might affect non-target , neighbouring populations ., In order to assess how population dynamics can impact upon the ability of a system to spread into a neighbouring population , we formulate here a two-deme population dynamics model ( Fig 2 ) ., This is used to investigate both the spread of transgenes into the neighbouring population and changes ( lasting or transient ) in the overall size of a neighbouring population ., Two-deme models such as this are well established in the literature for population genetics models of gene drive systems ( for example 16 , 36 , 37 ) but these models rarely incorporate population dynamics in addition to genetics ., We considered that for gene drive systems predicted to cause significant excess mortality , consideration of population dynamics might be important ., The consideration of two demes requires some modifications to be made to the equations stated in the previous section ., Firstly , since we are now modelling the dynamics of two populations , we must consider an extra set of delay-differential equations representing the second ( non-target ) population ., These take the same form as those given in the previous section except for the incorporation of migration between the two demes ., This is achieved through the addition of extra terms into each equation which are of the form ψ ( M i O ( t ) - M i T ( t ) ) for the target population and ψ ( M i T ( t ) - M i O ( t ) ) for the non-target population ., Here the superscripts T and O denote the target and non-target populations , respectively ., Similar terms are also included for female populations but with M replaced by F . This gives models of the form Early-acting fitness/lethal effects:, d M i T ( t ) d t = ρ v i T ( t - τ ) Ω i M 1 + ( ∑ i ∈ σ α v i T ( t - τ ) Ω i M + w i T ( t - τ ) Ω i F ) β - μ M i T ( t ) + ψ ( M i O ( t ) - M i T ( t ) ) , ( 6 ), d F i T ( t ) d t = ρ w i T ( t - τ ) Ω i F 1 + ( ∑ i ∈ σ α v i T ( t - τ ) Ω i M + w i T ( t - τ ) Ω i F ) β - μ F i T ( t ) + ψ ( F i O ( t ) - F i T ( t ) ) , ( 7 ), d M i O ( t ) d t = ρ v i O ( t - τ ) Ω i M 1 + ( ∑ i ∈ σ α v i O ( t - τ ) Ω i M + w i O ( t - τ ) Ω i F ) β - μ M i O ( t ) + ψ ( M i T ( t ) - M i O ( t ) ) , ( 8 ), d F i O ( t ) d t = ρ w i O ( t - τ ) Ω i F 1 + ( ∑ i ∈ σ α v i O ( t - τ ) Ω i M + w i O ( t - τ ) Ω i F ) β - μ F i O ( t ) + ψ ( F i T ( t ) - F i O ( t ) ) , ( 9 ) Late-acting fitness/lethal effects:, d M i T ( t ) d t = ρ v i T ( t - τ ) Ω i M 1 + ( ∑ i ∈ σ α v i T ( t - τ ) + w i T ( t - τ ) ) β - μ M i T ( t ) + ψ ( M i O ( t ) - M i T ( t ) ) , ( 10 ), d F i T ( t ) d t = ρ w i T ( t - τ ) Ω i F 1 + ( ∑ i ∈ σ α v i T ( t - τ ) + w i T ( t - τ ) ) β - μ F i T ( t ) + ψ ( F i O ( t ) - F i T ( t ) ) , ( 11 ), d M i O ( t ) d t = ρ v i O ( t - τ ) Ω i M 1 + ( ∑ i ∈ σ α v i O ( t - τ ) + w i O ( t - τ ) ) β - μ M i O ( t ) + ψ ( M i T ( t ) - M i O ( t ) ) , ( 12 ), d F i O ( t ) d t = ρ w i O ( t - τ ) Ω i F 1 + ( ∑ i ∈ σ α v i O ( t - τ ) + w i O ( t - τ ) ) β - μ F i T ( t ) + ψ ( F i O ( t ) - F i O ( t ) ) , ( 13 ), within which ψ represents the rate of migration while the definitions of other parameters and variables remain unchanged ., Note that migration terms do not include any developmental time delays since this is assumed to comprise adult mosquitoes moving from one deme to the other prior to mating on a given day ., Here we have excluded from our model the possibility of human-mediated movement of mosquito eggs as this is likely to be negligible relative to adult movement on the spatial scale considered here ., It is also worth noting that models not including migration are a special case of those presented here with the rate of migration set to zero ( i . e . ψ = 0 ) ., All numerical simulations within this study are created using MATLAB ( The MathWorks Inc . , Natick , MA ) delay differential equation solver dde23 38 ., The mathematical models outlined in the previous section contain a number of ecological parameters , namely the carrying capacity , primarily based on the number and quality of breeding sites ( i . e . carrying capacity is proportional to 1/α ) ; the strength of density-dependent larval competition ( β ) ; the intrinsic per capita rate of population growth ( ρ ) ; and the adult mortality rate ( μ ) ., Before investigating the impact of such ecological parameters on the efficacy of UD and KR gene drive systems , we begin by studying their effects on the overall size of an isolated mosquito population in absence of genetic control ., This can be investigated by considering the steady-state equation defined by Alphey & Bonsall 24 , i . e ., N * = 1 α ( ρ μ - 1 ) 1 / β , ( 14 ), where N* denotes the equilibrium population size in absence of control ., We now investigate the effects of each parameter on the overall equilibrium population size by considering variations in each while the other three are held constant—at the base values detailed in Table 2 . Since little is known about these ecological parameters in wild populations we vary each over a range wide enough that it is likely to span both realistic and biologically unfeasible scenarios ., This gives results as shown in Fig 3 . These results clearly show that , in a mathematical model of this type , each of the four ecological parameters are able to significantly affect the equilibrium population size ., We see from Fig 3 ( a ) and 3, ( c ) that parameters related to density-dependent larval competition ( i . e . α and β ) have the largest impact on the overall population size ., The intrinsic per capita rate of population growth ( ρ ) and the adult mortality rate ( μ ) are also capable of altering the overall equilibrium population size , but over a much smaller range than is possible with α and β ., It is clearly possible here for variations in these parameters to result in very large or small population sizes ., It is expected that these results span both realistic and biologically non-feasible population sizes ., Combining mathematical models and data on population sizes ( e . g . 39 ) can help to narrow down the realistic parameter range , however further experimental data is required to refine these yet further ., The previous section explored the effects of various ecological parameters on the overall size of a mosquito population ., We now examine whether such changes in the overall size of a population before the release of transgenic mosquitoes will impact upon the outcome or dynamics of UD and KR gene drive releases ., We explore this in the context of a number of important performance metrics ., In order to explore these effects we simulate a 1:1 ( introduced:wild ) release of each gene drive system ., This release is modelled as the introduction of AABB genotype adults at a ratio equal to the number of wild-type adult mosquitoes ( at equilibrium ) at the time of release ( see Supplementary Information Section 1 for further details ) ., We explore this release strategy over the full range of relative fitness parameters ( i . e . 0 ≤ ϵA/B ≤ 1 ) and for a number of different initial mosquito population sizes ., For changes in the initial population size , we consider only changes to the availability of larval habitats within an environment ( proportional to 1/α ) ., Specifically , we consider eight different initial population sizes spanning five orders of magnitude ( 100 to 104 ) , corresponding to α values of 0 . 7 , 0 . 2 , 0 . 07 , 0 . 02 , 0 . 007 , 0 . 002 , 0 . 0007 and 0 . 0002 ( see Fig 3 ( a ) ) ., To explore whether these changes in population size alter the behaviour of the systems studied here , we consider a number of metrics that represent different aspects of their performance ., For the UD system we consider three performance metrics , namely the change in equilibrium population size; the maximal degree of transient population suppression; and the time taken for the system to reach maximum population suppression ., For the KR system we consider the same metrics except there will be no equilibrium change in the population size since the system is self-limiting ., In each case we run numerical simulations for each of the eight α values and repeat this for values of ϵ ( relative fitness ) spanning the entire feasible range ( i . e . 0 ≤ ϵ ≤ 1 ) ., Results obtained here indicate that no behavioural differences result from variation in the initial population size , so long as the release ratio is held constant ( i . e . the absolute release size must be adjusted relative to the overall population ) ., From the results of Fig 4 we identify a number of interesting features of the UD gene drive system ., Firstly , similar to previous results for other classes of gene drive ( e . g . 24 , 30 ) we find significant differences between systems that use early-acting as opposed to late-acting transgenes ., Late-acting transgenes cause greater amounts of both maximal and equilibrium population suppression ( ∼65% and ∼40% , respectively ) than the early-acting system that produces just ∼5% and ∼4% , respectively when considering a 1:1 ( introduced:wild ) release ratio ., However , we do not see any large differences in either the time taken to reach the maximal level of population suppression or the threshold fitness costs that produce lasting transgene introgression between the early and late acting systems ., The region with no equilibrium population suppression below ϵA = 0 . 88 = ϵB is due to the elimination of the transgenes , but above this we see an amount of suppression that decreases as transgenic individuals become more fit ( i . e . as ϵA = ϵB → 1 ) ., For the KR system , the results in Fig 5 show that late-acting systems cause much greater levels of population suppression than early-acting ones ., In particular , for the release ratio considered here ( 1:1 , introduced:wild ) we see that a late-acting system causes at most ∼65% population suppression whereas the early-acting system produces just ∼2 . 5% ., We also observe a significantly different pattern of suppression as relative fitness varies ., Unlike the UD system , here we observe sizeable differences in the time taken to reach the maximum level of population suppression ., The early-acting system has a greatest time to maximal population suppression of ∼80 days whilst the late-acting system can take up to ∼70 days ., We also observe a difference between early and late acting systems in terms of the relationship between relative fitness parameters and the time taken to reach maximal population suppression ., The differences observed here in levels of population suppression between early- and late-acting systems have been explained previously for other gene drive systems 24 , 30 ., For early-acting systems , fitness/lethal effects act to lower larval density , thereby reducing the density dependent competition experienced later and partially compensating for the effects of introduced transgenes ., For late-acting systems , fitness/lethal effects act after density-dependent competition , providing an extra round of suppression that cannot be compensated for ., Thus , late-acting systems produce a greater degree of population suppression ., Whilst we observe significant differences in the degrees of population suppression caused by early- and late-acting UD and KR systems , we notice only tiny variations due to changing the initial population size ., In particular , for each panel in Figs 4 and 5 , some coloured symbols ( representing results with different initial population sizes ) display extremely small deviations from the grey lines ( randomly chosen sample simulations ) ., These deviations are likely to represent differences in numerical error between the eight examples , as explained in S1 Appendix Section 2 , S1 and S2 Figs ., This suggests that each system will produce consistent behaviour in terms of time scale , genotype frequencies and population suppression for any initial population size ( see S3 Fig for time courses showing population size effects for the full range of relative fitness parameters studied here ) ., We would expect this result to hold true so long as the population size is large enough to avoid stochastic and Allee effects ( which are not included in this model ) ., Laboratory and field studies have suggested that the larval stages of mosquito development may be subject to some degree of density-dependent competition 27 , 28 , 40–48 ., This often manifests itself as a reduction in the proportion of larvae that survive to adulthood in more densely packed environments 27 , 42 , 48 ., A number of mechanisms have been proposed to explain why this may occur 28 , 40 , 49 yet , to our knowledge , none of these has yet been conclusively shown ., Also subject to some degree of doubt is the strength of this density dependent effect ., Some studies have suggested that mosquito populations are subject to only a very small density dependent effect 27 , while others suggest that their dynamics may be strongly overcompensatory 46 ., Since Alphey & Bonsall 24 demonstrated the potential for density dependent effects to result in oscillatory dynamics in the context of endonuclease-based gene drive systems , here we explore the potential range of behaviour that may occur as a result of varying the strength of density dependent larval competition in mosquito populations when either a UD or a KR system is introduced ., While r | Introduction, Materials and methods, Results, Discussion | A number of different genetics-based vector control methods have been proposed ., Two approaches currently under development in Aedes aegypti mosquitoes are the two-locus engineered underdominance and killer-rescue gene drive systems ., Each of these is theoretically capable of increasing in frequency within a population , thus spreading associated desirable genetic traits ., Thus they have gained attention for their potential to aid in the fight against various mosquito-vectored diseases ., In the case of engineered underdominance , introduced transgenes are theoretically capable of persisting indefinitely ( i . e . it is self-sustaining ) whilst in the killer-rescue system the rescue component should initially increase in frequency ( while the lethal component ( killer ) is common ) before eventually declining ( when the killer is rare ) and being eliminated ( i . e . it is temporally self-limiting ) ., The population genetics of both systems have been explored using discrete generation mathematical models ., The effects of various ecological factors on these two systems have also been considered using alternative modelling methodologies ., Here we formulate and analyse new mathematical models combining the population dynamics and population genetics of these two classes of gene drive that incorporate ecological factors not previously studied and are simple enough to allow the effects of each to be disentangled ., In particular , we focus on the potential effects that may be obtained as a result of differing ecological factors such as strengths of larval competition; numbers of breeding sites; and the relative fitness of transgenic mosquitoes compared with their wild-type counterparts ., We also extend our models to consider population dynamics in two demes in order to explore the effects of dispersal between neighbouring populations on the outcome of UD and KR gene drive systems . | Vector-borne diseases represent a severe burden to both human and animal health worldwide ., The methods currently being used to control a range of these diseases do not appear sufficient to address the issues at hand ., As such , alternate methods for the control of vector-borne diseases are currently being investigated ., Among the promising techniques currently being considered are a range of genetic control methods known as gene drive systems ., These allow desirable genetic traits ( such as a much reduced capacity for vectors to transmit viruses ) to be spread through a target population; taking advantage of natural mate seeking behaviour to locate vector sub-populations that can be extremely difficult for humans to locate and reach ., Here we use mathematical models ( parameterised to consider mosquito populations ) to demonstrate the robustness of the engineered underdominance and killer-rescue classes of gene drive to different ecological factors including birth and death rates; the number and quality of breeding sites ( i . e . carrying capacity ) ; and the strength of density-dependent competition during the larval development phase ., We then go on to explore the range of potential outcomes that may result from the migration of individuals between two neighbouring populations . | biotechnology, invertebrates, medicine and health sciences, ecology and environmental sciences, population dynamics, population genetics, animals, theoretical ecology, population biology, insect vectors, genetic engineering, infectious diseases, population ecology, disease vectors, insects, arthropoda, population metrics, mosquitoes, population size, eukaryota, ecology, genetics, biology and life sciences, species interactions, introgression, evolutionary biology, evolutionary processes, organisms | null |
journal.pntd.0005470 | 2,017 | Using simulation to aid trial design: Ring-vaccination trials | The West African Ebola epidemic highlighted the need to identify a range of trial designs to evaluate vaccine effects rapidly , efficiently and rigorously during emerging disease outbreaks ., The ring-vaccination trial approach employed in the Ebola ça suffit trial in Guinea is one innovative approach 1 , which produced valuable evidence that the vaccine could prevent Ebola infection 2 ., Other approaches considered include individual randomization and a stepped-wedge design 3 , 4 ., In such trials it is difficult to estimate the likely effect of an infectious disease intervention because of indirect effects , and this issue is compounded by complex trial design ., Sample size calculations are based on group-level quantities such as intervention effect and are therefore potentially inaccurate ., By creating a transmission dynamic model for a ring vaccination trial , we show that we can make sample size calculations based on disease characteristics and individual intervention efficacy ., With this framework in place we are then able to examine the estimated vaccine effect and sample size under a range of assumptions about the properties of the vaccine , the trial , and the study population ., Although the only implementation of the ring trial design has been in Guinea during the Ebola epidemic , lessons can be learned and extended to other diseases and contexts ., Here , we examine the tail end of an epidemic of a disease with a latent and asymptomatic phase with effective contact tracing to illustrate a more widely-applicable set of findings ., In particular , we use baseline parameters values consistent with Ebola in West Africa in 2014–6 , but we vary several assumptions over broader ranges than those occurring in the Ebola ça suffit trial , with the aim of being relevant to a range of potential future situations ., The simulation is based on a stochastic , susceptible-exposed-infectious-detected-removed-vaccinated ( SEIDRV ) model for individual disease events , and it represents progression of the disease in a small cluster ( henceforth ‘ring’ ) with homogeneous mixing ., The ring represents both contacts and contacts of contacts so the assumption of homogeneous mixing is a simplifying assumption , which we can relax by modelling ‘contacts’ and ‘contacts of contacts’ as separate compartments with the highest transmission among the contacts ., New cases arise through direct contact between an infectious individual and a susceptible individual within the ring , and through external infectious pressure , denoted by F , which is constant and fixed for all members of the ring ., Members of the ring undergo surveillance by the study team , meaning that infectious individuals are detected and isolated with a daily probability pH , ending their infectious period ., We assume in the baseline scenario that detection rate in the trial is equivalent to routine surveillance , reflecting the fact that the trial doesn’t interrupt or enhance disease control efforts ., If infectiousness ends naturally , individuals can no longer be detected ., A ring is enrolled into the trial when a case is detected through routine surveillance ., This first detected case is defined as the index case for the purposes of the trial , but may or may not be the true index case of the outbreak in the ring ., Once a ring enters the trial all its members are randomly assigned to immediate vaccination ( on day 1 ) or delayed vaccination ( on day 22 ) ., In the baseline scenario we assume no ineligibility or non-consent , so that all susceptible and exposed individuals in the ring are vaccinated , and that there is no heterogeneity or administrative delay affecting the day of vaccination ., The mechanism of the vaccine in an individual is as follows: multiplicative leaky efficacy 5 increases linearly from 0 to VE ( set at baseline to be 0 . 7 ) over a period of Dramp days following vaccination , after which there is no change in efficacy over the study period 6 ., Statistical analysis of the trial is based on cumulative incidence in the rings by end of follow-up and a 95% confidence interval is calculated and reported 7 ., The required sample size to test a vaccine effect with 80% power is based on a difference in cumulative incidence 8 , using parameters output by a simulated trial with 15 , 000 rings ., We chose this analysis method because of the existence of simple closed-form sample size and vaccine efficacy formulae ., Because both arms receive the vaccine , cases that contribute towards the cumulative incidence in each arm are only counted during a window in which the immediate arm is presumed to be protected by the vaccine , and the delayed arm is not protected ., The window length is set to 21 days , equal to the vaccination delay between the arms ., Because the disease has an asymptomatic phase and the vaccine has a ramp-up period during which it is not fully efficacious , the window starts at 16 days , the sum of the average asymptomatic period length and Dramp , in an attempt to exclude cases in the immediate arm who were infected before they were fully protected by the vaccine ., We did not explicitly implement clustering in the simulation , instead assuming that transmission dynamics in all rings are independent ., However , clustering of cases within rings arises naturally due to dependent happenings ., We measure this clustering using the intracluster correlation coefficient ( ICC ) , calculated as per Shoukri et al 9 , adjusting for the covariate of trial arm and accounting for variable ring size where appropriate ., In conducting the statistical analysis we assume full knowledge of the vaccine mechanism , and that cases are only included if they are detected before their infectious period ends , and their symptoms appeared during the window ., For additional details on the disease transmission model , ring initiation , and analysis of the trial see the supplementary appendix ., Table 1 shows the parameters used in the model , their meanings , values under baseline assumptions , and references or justifications ., In order to align this model with the presumed context of the Ebola ça suffit trial , we modelled an entirely susceptible study population at the end of an epidemic , so that Reff has fallen to below one due to behaviour change ., To calibrate the model , we set Reff to reproduce a monthly detected attack rate of 2% when starting from one infected individual in a ring of 50 unvaccinated susceptible individuals , in the presence of case detection at a rate pBH ., Under baseline parameters in this model , the median total vaccine effect calculated from performing 100 trials with 89 rings in each arm was 70% ., This value should include direct and indirect effects , so we would expect it to exceed the direct effect of 70% ., However , while direct effects begin immediately , indirect effects are only important in the second generation of preventable cases onwards ., There are cases in this generation that occur in the case-counting window because Reff is small and the window duration is not much longer than a typical disease generation ( 17 days ) , so the indirect effects are small ., Fig 1 shows the effect of six variables on the point estimate of vaccine effect: daily probability of detection , true individual vaccine efficacy , proportion of infections from outside the ring , baseline attack rate in the unvaccinated population , administrative delay in vaccination , and start day of case-counting window ., Firstly , if there is enhanced surveillance in both arms of the trial leading to more rapid isolation of infectious cases ( pH>pBH ) , this will modestly reduce effectiveness estimates ( Fig 1A ) ., Secondly , as individual vaccine efficacy properties increase the estimated vaccine effect increases ( Figs 1B and S1 ) ., Thirdly , the percentage of infections from within the ring shows a weak negative association with the estimate of vaccine effect ( Fig 1C ) ., While the magnitude of indirect effects is modest as discussed above , they are almost negligible when most infections are from outside the ring , because preventing infections within the ring does not confer as much protection to susceptible individuals ., The increase in vaccine effect with higher attack rate seen in Fig 1D is driven by the increase in indirect vaccine effects in the immediate arm ., Finally , delay between ring formation and vaccination means that by the beginning of the time window the vaccine has had less time to prevent cases in the immediate arm ., Thus the reduction in incidence in the immediate arm does not reflect the true effect of the vaccine and the vaccine effect estimate is reduced ( Fig 1E ) ., A major determinant of the effect estimate is the choice of time window in which to count cases , as seen in Fig 1F ., Not surprisingly , starting the window too early reduces the estimated effects because it includes a period of time during which the vaccine cannot affect the incidence of cases becoming symptomatic–many cases becoming symptomatic on day 8 , for example , will have been infected by the index case prior to isolation , or will have been infected by a contact on ( say ) day 3 , before the vaccine had time to induce protection ., Starting the window later than the baseline of 16 days allows the trial to capture later generations in the chain of transmission , from a vaccinated person to another vaccinated person ., This increases the vaccine effect estimate as it includes indirect effects ., One might expect to see that starting the window too late would reduce effect estimates because it would include a period when the delayed group was also protected by the vaccine ., This does not appear to be the case , at least up to a start time of 35 days ( Fig 1F ) –see the supplementary material for an explanation of this phenomenon ., Fig 2 shows the effect of the same six variables on the required sample size: baseline attack rate in unvaccinated population , start day of case-counting window , daily probability of detection , true individual vaccine efficacy , administrative delay in vaccination , and force of external infection ., The effect of each parameter on the sample size can be understood through its effect on one or more of the three factors that determine the power of this trial: the number of events , how they are distributed between the two arms , and the level of clustering of cases within rings ., Respectively these factors are represented by the attack rate in the controls , the cumulative incidence difference between the arms , and the intracluster correlation coefficient ( ICC ) 8 ., Variables that decrease the incidence rate in the controls and cases will decrease the power because for the same sample size the trial will observe fewer events ., The baseline detected attack rate among unvaccinated individuals is a simple example of such a parameter ( Fig 2A ) ., Two other parameters act on the overall incidence in the trial ., Firstly , making the start of the case-counting window later decreases incidence in both arms because with Reff<1 the incidence is on average declining , so across all rings in the trial the number of cases decreases over the follow-up period ( Fig 2B ) ., Secondly , the case detection decreases detected incidence rate at both extremes ( Fig 2C ) ., When case detection is high , transmission chains are interrupted by case isolation and the true incidence decreases ., When case detection is low , many cases die or recover before they can be detected and consequently the detected incidence decreases ., Variables that make the two arms of the trial appear more different will increase the power of the trial as the ability to differentiate between them is increased , and Fig 1 identifies such variables ., Vaccine characteristics , in particular vaccine efficacy ( Fig 2D ) , are simple examples of such a parameter , since the immediate arm receives greater protection against disease compared to the delayed arm ., Changes to two other parameters increase the incidence difference in this way , as explained above: reducing the delay between ring formation and vaccination ( Fig 2E ) and starting the case-counting window earlier ( Fig 2B ) ., The effect of the timing of starting to count cases thus reflects two opposing forces on the sample size: it decreases sample size by increasing the incidence difference , and it increases sample size by decreasing the overall incidence ., When the window is early , the former of these effects dominates as seen by the increase in sample size for early time windows in Fig 2B ., When the window is late , the latter effect dominates , as seen by the increase in sample size for late time windows in the same figure ., Finally , the level of clustering within rings inflates the sample size , because more clustering means that each individual case provides less information ., It is often not intuitive to predict the direction in which a parameter will cause the ICC to change , and in many cases the ICC is not sensitive to the parameter ., One exception is the infection from outside the ring ( Fig 2F ) ., The most significant effect of introducing external infection and reducing within-ring transmission is to make infection probability for one individual within a ring independent from the infection prevalence within the same ring ., This reduces clustering in incidence ( making it more Poisson-like ) , thus reducing the ICC and the necessary sample size ., The width of the confidence intervals is affected in the same way by the three variables described above ., In particular , low incidence in either arm , high ICC and a small incidence difference between the arms all lead to a wider confidence interval ., The formula for the confidence interval is different from the formula used to make the power calculation , so the trends do not completely align because the size of the effect of each of the three factors is different for the confidence interval and the sample size ., For an investigation of the sensitivity of the total vaccine effect estimate and sample size to other parameters in the model , see the supplementary material ., For an interactive tool to explore the sensitivity of the trial parameters , see https://matthitchings . shinyapps . io/ShinyApps/ ., The ring-vaccination , cluster-randomized design has two key strengths that make it a good candidate when disease transmission exhibits spatiotemporal variation ., Firstly , by including members of the study population who are contacts of cases , the trial preferentially selects those at higher risk of disease acquisition , leading to an increase in efficiency while preserving false-positive rate through randomization ., Indeed , when a vaccine with 0% efficacy was tested in our simulations the false positive rate was maintained at 5% ., Secondly , even those study subjects who are randomized to delayed vaccination are theoretically in close contact with the study team meaning that individuals from the source population who are at the highest risk are followed closely and benefit from the trial even in the absence of vaccination 12 ., In addition , vaccination of clusters when they arise allows for gradual inclusion , meaning that this design is appropriate when logistical constraints make immediate vaccination of all participants impossible or inappropriate ., In this respect it is similar to a stepped-wedge cluster trial , in which prespecified clusters within the study population are vaccinated in a random order ., Although we have not made a direct comparison in this study , Bellan et al 13 showed that the stepped-wedge design is underpowered when the incidence is declining because it cannot prioritize the vaccine for those at highest risk ., The ring vaccination design , on the other hand , is inherently risk-prioritized because all study participants should be at higher risk than the general population ., All trials should be correctly powered in order to avoid erroneous rejection of an efficacious vaccine ., For a trial design with several complexities such as the one presented here , a sophisticated approach to sample size calculation is merited ., A standard approach to sample size calculation for this trial would involve specifying the attack rate among the controls , the desired effect of the vaccine on the population level , and the ICC ., In the context of a serious epidemic , these parameters are unlikely to be estimated with certainty; for example , the ICC requires cluster-level data to be estimated accurately ., The ICC is an important parameter in designing cluster-randomized trials , yet in the absence of data it is often assumed to be 0 . 05 ., In our simulations the range of ICCs observed was 0 . 01–0 . 04 , suggesting that the value of this uncertain parameter should not always be assumed to be fixed at 0 . 05 ., Therefore , the modelling approach replaces assumptions about these cluster-level quantities with assumptions about population-level parameters and disease characteristics , which are more likely to be available through analysis of data from the outbreak ., A second advantage of the modelling approach is that , based as it is on a simulating the transmission of disease within a trial , it is possible to explore the impact of parameters describing the design of the trial and the properties of the disease ., The added detail gained from specifying the disease model allowed us in this study to identify some key issues with the design that are worth considering ., Firstly , as seen in Fig 2C , increasing case-finding efficiency above background rate has a negative impact on power , as fast isolation of cases in both arms leads to an overall decrease in cases observed by the trial ., In future trials it is worth considering if there are alternative or composite endpoints , if the disease in question permits , that can be used to allow for efficacy estimates while maintaining close follow-up ., Secondly , a key design consideration in the delayed-arm ring-vaccination trial is when to count cases ., An intuitively appealing approach is to place the window so that the immediate arm is receiving full protection and the delayed arm none ., This should in theory minimize bias caused by misclassification of unvaccinated individuals as vaccinated and vice-versa ., While this placement achieves nearly maximal power , it does not maximize the VE estimate ., Indirect effects that are important later in time increase the VE estimate for later time windows , while at the same time declining incidence within each ring decreases power for later time windows ., Finally , the above point draws attention to the fact that caution is required when interpreting the VE estimate produced by the trial ., As seen in Fig 1 , many parameters that are not characteristics of the vaccine can influence the estimated effect ., Whether this is due to misclassification ( for example , when the time window is too early ) or due to indirect effects ( for example , when the attack rate is high enough to cause long transmission chains ) , the context of the trial should be taken into account when interpreting the VE estimate ., While in the baseline scenario the trial appears to correctly estimate the individual efficacy , this is the result of misclassification and indirect effects cancelling each other out ., This claim is supported by the fact that the median VE estimate falls below the individual-level vaccine efficacy when most or all infections are from outside the ring ( Fig 1C ) and indirect effects are negligible ., The focus of this model was to explore parameters within each ring and understand how they affect the quality of data coming from the trial ., As a result , we did not consider the wider context of the population disease dynamics , and in particular how and when the rings arise ., For example , we calibrated Reff to a secondary attack rate in a cluster was 2% , which is not necessarily comparable to the monthly cumulative incidence in the population ., If transmission takes place mainly in clusters then population cumulative incidence could be somewhat lower than cluster secondary attack rate , increasing the efficiency of a ring-vaccination trial relative to a stepped-wedge cluster trial or individual RCT ., Linking this model to a model of disease within the general population would allow us to make direct comparisons to other trial designs such as the stepped-wedge cluster trial and the individually-randomized trial investigated elsewhere 13 , 14 , but it would require detailed information about the nature of clustering of the disease in this context , and for simplicity we focused on the within-ring dynamics only ., As with every model , there are limitations to these simulation results ., The strength of the modelling approach compared with a standard approach is that it better estimates the parameters on which the sample size depends ., However , some of the model parameters might still be uncertain in a situation in which such a model might be useful ., For example , we may have limited information about the characteristics of a disease , in particular its latent and incubation period , and its Reff ., The simulation results are dependent on these assumptions , and so they cannot be used at the very outset of epidemic , or else they risk being highly inaccurate ., Even at the end of the West African Ebola epidemic , there were no more than four or five reliable estimates of the latent and infectious periods of EVD , and indeed there is perhaps evidence that our understanding of the natural history of the disease remains limited 15 ., In addition , we have considered only the simplest method of analysis for the trial–a comparison of attack rates between the two arms after correction for clustering of cases within rings ., More sophisticated methods , including time-to-event analyses incorporating ring-level random effects , as performed in the Ebola ça suffit trial , would have somewhat different sample size requirements ., However , we believe that the trends seen here would be similar for other methods , because the VE estimates returned by various methods will be similar for a rare outcome 5 ., In building the model we made some simplifying assumptions , and although we tested the robustness of the results to these assumptions ( see supplementary material ) it is possible that a more sophisticated model would provide more accurate results , particularly if superspreading events are not rare in this study population ., For a vaccine trial in an epidemic , when the level of indirect effects is hard to predict , power calculations can be sensitive to parameters about which very little is known ., Simulations such as these can be important aids in understanding a range of values for these parameters before a trial is carried out , and thus ensuring that the trial has sufficient power to detect an efficacious vaccine ., In this trial , a finding significantly different from the null likely indicates one or more types of vaccine efficacy at the individual level , but the magnitude of the effect and the power to detect the effect will vary across settings . | Introduction, Methods, Results, Discussion | The 2014–6 West African Ebola epidemic highlights the need for rigorous , rapid clinical trial methods for vaccines ., A challenge for trial design is making sample size calculations based on incidence within the trial , total vaccine effect , and intracluster correlation , when these parameters are uncertain in the presence of indirect effects of vaccination ., We present a stochastic , compartmental model for a ring vaccination trial ., After identification of an index case , a ring of contacts is recruited and either vaccinated immediately or after 21 days ., The primary outcome of the trial is total vaccine effect , counting cases only from a pre-specified window in which the immediate arm is assumed to be fully protected and the delayed arm is not protected ., Simulation results are used to calculate necessary sample size and estimated vaccine effect ., Under baseline assumptions about vaccine properties , monthly incidence in unvaccinated rings and trial design , a standard sample-size calculation neglecting dynamic effects estimated that 7 , 100 participants would be needed to achieve 80% power to detect a difference in attack rate between arms , while incorporating dynamic considerations in the model increased the estimate to 8 , 900 ., This approach replaces assumptions about parameters at the ring level with assumptions about disease dynamics and vaccine characteristics at the individual level , so within this framework we were able to describe the sensitivity of the trial power and estimated effect to various parameters ., We found that both of these quantities are sensitive to properties of the vaccine , to setting-specific parameters over which investigators have little control , and to parameters that are determined by the study design ., Incorporating simulation into the trial design process can improve robustness of sample size calculations ., For this specific trial design , vaccine effectiveness depends on properties of the ring vaccination design and on the measurement window , as well as the epidemiologic setting . | The urgency , as well as the logistical and sometimes ethical challenges of clinical trials for interventions during epidemics of emerging diseases prompts the need for novel designs and analytic strategies ., The successful use of a novel cluster-randomized ring-vaccination trial to test an Ebola vaccine in Guinea raises the general question of what circumstances would favour the use of trials of similar design and how the properties of the population , the vaccine and the trial would influence the necessary sample size and the expected results ., We present a generalized transmission dynamic model for a ring vaccination trial to address these questions ., This work is an example of the general phenomenon that mechanistic , transmission-dynamic simulations can aid in the design and interpretation of intervention trials for infectious diseases , when the trial itself can have non-obvious effects on transmission dynamics that may not be fully captured by effect- and sample-size calculations for noncommunicable diseases . | medicine and health sciences, infectious disease epidemiology, immunology, tropical diseases, ebola hemorrhagic fever, simulation and modeling, vaccines, preventive medicine, clinical medicine, mathematics, statistics (mathematics), neglected tropical diseases, pharmacology, infectious disease control, vaccination and immunization, research and analysis methods, public and occupational health, viral hemorrhagic fevers, infectious diseases, epidemiology, cluster trials, drug research and development, confidence intervals, disease dynamics, clinical trials, biology and life sciences, viral diseases, physical sciences | null |
journal.pgen.1006279 | 2,016 | Gain- and Loss-of-Function Mutations in the Breast Cancer Gene GATA3 Result in Differential Drug Sensitivity | High-throughput genome sequencing has allowed the systematic analysis of the complex mutational landscape of tumours and has provided key insights into tumour evolution and cancer etiology 1–3 ., Mutation patterns in individual genes also reveal important insights into their role in tumourigenesis and can assist in distinguishing driver from passenger mutations 1–4 ., Mutation rates are elevated in protein domains or regulatory sites , indicating their functional importance for cancer development 5 , 6 ., It is typically assumed that all mutations within an individual gene have the same downstream consequences for tumourigenesis ., However , at least one notable example challenges this paradigm ., Distinct mutations in the TP53 gene ( encoding p53 ) lead to both loss-of-function and gain-of-function , impinging on multiple different pathways 7–10 ., Yet , it is unclear if this type of dual activity of mutant p53 represents an exceptional case or is more common ., We hypothesised that mutations in different positions in a cancer gene may result in different downstream consequences ., To investigate this , we developed an unbiased computational approach and applied it to breast cancer , as large publicly available data sets are available for this cancer type ., Breast cancer has been studied extensively in terms of its molecular and genetic markers ., Its classification into subtypes according to expression of receptors and gene expression profiles is used for diagnostic and prognostic purposes and forms the basis for treatment decisions 11–17 ., Breast cancer is genetically heterogeneous and only four driver genes are mutated in more than 10% of patients 18–25: PIK3CA ( encoding the catalytic subunit of PI3K ) , CDH1 ( encoding E-cadherin ) , TP53 , and GATA3 ( encoding GATA-binding protein 3 ) ., While the roles of the pro-survival PI3K pathway , cell adhesion , and p53 as the guardian of the genome in tumourigenesis are well studied , comparatively little is known about the role of the equally commonly mutated gene GATA3 ., To some extent this is due to the relatively recent discovery of the high prevalence of GATA3 mutations 19–22 , 26 ., In addition , model systems ( e . g . , cell lines , animal models ) to study GATA3 in breast cancer are lacking , hampering functional studies ., GATA3 is a member of the GATA family of transcription factors and forms homodimers that bind conserved hexanucleotide sequences containing the central GATA motif 27–29 ., It is a master regulator of helper T cell specification 30 and plays a critical role in development and differentiation of various tissues , including the mammary gland 31–33 ., During normal mammary development , GATA3 , together with the estrogen receptor ( ER ) 34–37 , controls differentiation of the luminal epithelium in the terminal end buds in the breast ., In adult tissues , GATA3 helps to maintain the luminal identity 38–41 ., The contribution of GATA3 to cancer is , in contrast , poorly understood ., Most of our current knowledge regarding GATA3’s potential function in breast cancer has been revealed from genomic studies highlighting an ER/FOXA1/GATA3 co-operating network of transcription factors in luminal tumours 14 and ER-positive cell line models 34 , 35 , 37 , 42 , 43 ., Yet , the observation of GATA3 downregulation during tumour progression and predominant frameshift mutations have led to the view that GATA3 acts primarily as a tumour suppressor 44 , 45 ., In this study , we identify differential functional consequences of mutation types in GATA3 ., We present evidence that the most common mutation type results in a protein with elongated C-terminus that displays effects consistent with gain-of-function activity in a cell line model ., This is highly surprising , as frameshift mutations are generally believed to yield inactive proteins due to premature termination of translation ., In addition , we describe a synthetic lethal interaction between this GATA3 mutant and drugs targeting the histone methyltransferases G9A and GLP , providing a first putative therapeutic opportunity for patients carrying GATA3 mutations ., Together , our findings demonstrate that different mutations in the same gene can result in differential drug sensitivities and contest the view that GATA3 acts only as a tumour suppressor ., To study mutation patterns in breast cancer , we used publicly available data from The Cancer Genome Atlas ( TCGA ) 23 and from the Molecular Taxonomy of Breast Cancer International Consortium ( METABRIC ) 25 ., Fig 1A shows the most commonly mutated genes in breast cancer ., Somatic mutations in these recurrently mutated breast cancer genes are often mutually exclusive 46 , 47 ( Fig 1A , S1 Table ) and distributed in a non-uniform fashion along the gene body ( Fig 1B ) ., The observed patterns are largely consistent between the TCGA and METABRIC datasets ., For instance , PIK3CA mutations chiefly occur at just two positions corresponding to different protein domains: E545 in a helical regulatory domain and H1047 in the kinase domain 48 ., Clear hotspot mutations at single amino acid residues or within narrow regions are also present in TP53 , and to some extent in GATA3 49 ., Mutations in GATA3 ( Entrez Gene ID: 2625 ) have not yet been extensively characterised , but the non-uniform distribution and mutual exclusivity with mutations in other cancer genes are strong indicators that GATA3 is a cancer driver gene 25 , 50 ( Fig 1B , S1 Table ) ., In order to assess potential functional consequences of regional mutation patterns , we devised an unbiased , systematic approach for linking the position of a mutation within a cancer gene with gene expression data ., We reasoned that such an analysis could highlight domains in cancer genes that–when mutated–would result in differential downstream effects ., First , we extracted from TCGA the genomic position for each mutation found in a patient in the seven selected driver genes , and the non-driver control gene TTN 51 , along with the gene expression profiles from the same patients ., Next , we used a segmentation approach to identify regions within a driver gene that led to a change in expression levels of another gene ( see Materials and Methods ) ., The identification of such patterns would suggest that the mutations in a particular region of the gene are functionally distinct ., We termed genes that displayed altered expression along distinguishable segments of driver mutations “response genes” and refer to the border between two segments as a “segmentation breakpoint” ( Fig 1C ) ., Strikingly , we found that the highest number of response genes was associated with GATA3 mutations ( Fig 1D , S2 Table ) , where more than 900 genes displayed a segmented pattern ., In comparison , around 200 response genes were linked with PIK3CA mutations , and fewer than 100 response genes were identified for the non-driver control gene TTN ., The observation that the majority of response genes displayed a single breakpoint ( Fig 1D ) suggests that patient-derived GATA3 mutations can be divided into two functionally distinct regions and that mutations in these regions are associated with differential gene expression in tumours ., Most GATA3 mutations ( 66/99; 67% ) in the TCGA dataset are heterozygous frameshift mutations in exon 5 and exon 6 ( Fig 2A ) ., Frameshifts in general lead to premature stop codons , which can substantially disrupt protein function ., Indeed , approximately 41% ( 27/66 ) of the frameshift mutations in GATA3 are predicted to result in an early stop codon ( Fig 2B ) ., These truncated proteins ( hereafter referred to as GATA3-trunc ) are stable and expressed in tumours 52 , and , as GATA3 likely forms a homodimer 29 , it is probable that they may act in a dominant negative manner 53 , 54 ., These mutations would thus be consistent with a haplo-insufficient tumour suppressor function of GATA3 ., Interestingly , most ( 39/66 , 59% ) frameshift mutations in GATA3 result in a protein with an extended C-terminus ., These extension mutations occur predominantly ( 33/39 , 85% ) in exon 6 and affect the resulting mutant proteins starting from different residues between alanine 395 and glycine 444 , with a hotspot ( 11/39 ) at proline 409 ( Fig 2A and 2B ) 49 ., The mutations are strongly biased toward the +1 frame ( Fig 2A , bottom ) ., This is surprising , as -1 frameshifts in this position would result in a shortened and aberrant C-terminus ., The alternative +1 frame alters up to 49 amino acids of the original C-terminus and extends the protein by 63 novel amino acids ( hereafter GATA3-ext , Fig 2B ) ., Because frameshift mutations in the TCGA dataset as a whole do not display a frame preference , the bias toward the +1 frame in GATA3 is suggestive of positive selection ., One potential explanation for this could be that the GATA3-ext mutation is functionally distinct from other ( truncating ) mutations , for instance by providing a gain-of-function ., Together , this demonstrates that our analysis can identify functional distributions of mutations as well as novel candidate tumourigenic mechanisms ., We next revisited our segmentation analysis to investigate the positional distributions of mutations and segmentation breakpoints ( S1A–S1H Fig ) ., In GATA3 , the distribution of breakpoints present in single-breakpoint response genes was distinct from the distribution of all mutations ( Fig 2C ) ., It sharply peaked at a position that separated GATA3-ext from GATA3-trunc mutations ., This is illustrated with BCAS1 , the strongest GATA3 response gene ( Fig 2D , S3 Table ) ., This indicates that genes like BCAS1 are differentially expressed in tumours that contain the GATA3-ext mutation but not in tumours harbouring the GATA3-trunc mutation ., Other response genes such as SYT17 displayed more complex patterns , but one of the breakpoints often tended to separate the extension and truncation mutants ( Fig 2D ) ., Thus , the differential expression of response genes can functionally define the type of mutation ., Together with the observation that the +1 frameshift mutations are under positive selection , this suggests that these GATA3-ext mutations are mechanistically distinct ., To investigate what functional aspects of cancer cells specifically correlate with GATA3-ext mutations , we calculated the association of this mutation class with gene expression levels without performing segmentation ., We divided patients carrying GATA3 mutations into two groups: those with GATA3-ext mutations and those with all other types ., Corroborating the segmentation analysis , we obtained differentially expressed genes between these groups ( S4 Table ) , which matched many of the previously identified response genes ., This confirms that cellular processes are indeed differentially affected in GATA3-ext tumours in comparison to the other GATA3 mutant tumours ., Differential gene expression in GATA3-ext tumours may indicate distinct tumour characteristics and this could affect disease progression and therapeutic response ., To address this , we used the previously defined patient groups and performed survival analysis of the TCGA patients ., Only GATA3-ext patients progressed during the follow-up period of the cohort but no patients with an alternative GATA3 mutation did ., Accordingly , patients with GATA3-ext mutations displayed significantly ( p = 0 . 0029 ) shortened disease-free survival in the TCGA cohort ( Fig 2E ) ., This indicates that GATA3-ext is a putative biomarker for disease progression and is consistent with the notion that extension mutants have important mechanistic properties that are distinct from other GATA3 mutations ., To assess whether GATA3-ext mutations are associated with similar expression changes in an independent patient cohort , we analysed the METABRIC dataset ., The METABRIC cohort carries relatively fewer GATA3-ext mutations and displays a moderately different mutation distribution in GATA3 ( compare lower panels of S1D with S1I Fig ) ., Interestingly , substantial differences in GATA3 mutation patterns have also been noted in a Chinese cohort 46 ( see Discussion ) , suggesting that genetic background and environmental factors play an important role in GATA3-driven breast cancer ., We further noted that in the METABRIC cohort disease-free survival was similar for patients harbouring GATA3-ext or other GATA mutations , further indicating considerable differences in cohort composition ( S1J Fig ) ., Despite these dissimilarities , we repeated the segmentation analysis in both the METABRIC and TCGA datasets using the GATA3-ext gene signature derived from TCGA ( S4 Table ) ., The analysis was limited to 46/50 genes , as expression data for four genes were not present in the METABRIC cohort ., As expected , all 46 genes showed segmentation for GATA3 in the TCGA dataset ( S1K Fig ) ., Notably , 34/46 signature genes qualified as GATA3-specific response genes in the independent METABRIC dataset ( Fig 2F , S1I Fig ) ., This implies that these genes are differentially expressed in GATA3-ext tumours ., Next , we calculated the fold change of the 46 genes in GATA3-ext samples relative to all other GATA3 mutant tumours ., Strikingly , the changes in both datasets occurred in consistent directions for all 46 genes ( Fig 2G ) , indicating qualitative agreement between TCGA and METABRIC for the GATA3-ext gene signature ., Hence , the distinctive effects of GATA3-ext are recapitulated in an independent breast cancer patient cohort ., Following investigations of GATA3 mutation types in human patient data , we wished to study these mutations types in vitro in order to understand their functional implications in more detail ., An endogenous heterozygous truncating mutation in exon 5 ( cDNA:1006insG ) in the commonly used MCF7 breast cancer cell line has been previously reported and was shown to decrease DNA binding and increase protein half-life 53 , 55 ., However , we did not find cancer cell lines with GATA3-ext mutations by mining the Cancer Cell Line Encyclopaedia ( CCLE ) 56 and the Catalogue of Somatic Mutations in Cancer ( COSMIC ) 57 databases or by analysing 45 breast cancer cell lines by Sanger sequencing and Western blot ., The high frequency of GATA3 mutations in breast tumours is thus not well represented in cell lines ., We also tested tumour tissue from 10 luminal patient-derived xenograft ( PDX ) mouse models and did not detect any GATA3-ext mutations either ., Although the small sample numbers preclude a strong conclusion , together these results suggest that GATA3-ext mutant cells may not adapt well to ex vivo culture conditions ., The lack of cell lines with naturally occurring GATA3-ext mutations impelled us to search for an alternative model system ., To distinguish putative gain-of-function from dominant negative effects , we wished to study GATA3-ext in the absence of wild-type GATA3 ., We attempted to inactivate the endogenous locus by CRISPR/Cas9 gene editing in several ER-positive breast cancer cell lines ( MCF7 , T47D and CAMA1 ) , but this did not yield viable homozygous null clones ( ~150 clones analysed ) ., CRISPR/Cas9-directed replacement of endogenous GATA3 by GATA3-ext was equally unsuccessful ( >100 clones analysed ) ., This suggests that at least one copy of wild-type GATA3 is required for viability in these cell lines , which is in accordance with the findings from human cancer samples but complicates the introduction of a mutated version for in vitro models ., To establish an alternative model , we used the non-tumourigenic MCF10A breast epithelial cell line that naturally expresses very low protein levels of endogenous GATA3 ( Fig 3A and 3B ) ., We stably expressed wild-type GATA3 ( GATA3-wt ) , GATA3-ext ( cDNA:1224insG; p:P409fs ) and GATA3-trunc ( cDNA:1006insG; p:G336fs ) through retroviral transduction and puromycin selection ( Fig 3A ) ., The GATA3-ext protein was stable , albeit expressed at slightly lower levels than GATA3-wt ., Importantly , the expression levels of the GATA3 proteins were in the physiological range of endogenous GATA3 observed in various other breast cancer cell lines ( Fig 3B ) ., Furthermore , confocal microscopy showed nuclear localisation for both mutants as well as for the wild-type protein ( S2A Fig ) ., We noted a slight but significant decrease in growth rate of MCF10A GATA3-ext cells as compared to GATA3-wt ( Fig 3C ) ., This was consistent between independent infections of the parental MCF10A cells with titrated virus , excluding an effect of the viral transduction itself ., The specific effect of GATA3-ext shows that this mutation affects MCF10A cells’ ability to proliferate in standard tissue culture medium conditions ., We next performed RNA sequencing on MCF10A GATA3-wt , GATA3-ext , GATA3-trunc or control vector expressing cells to characterise the effects of GATA3 mutations at the cellular level ., The expression of GATA3-wt and GATA3-ext resulted in up- or downregulation of 725 and 853 genes , respectively , with respect to the control ( p <0 . 05 , FC >1 . 5 ) , indicating widespread transcriptional changes ., In contrast , expression of GATA3-trunc yielded a considerably smaller signature ( 134 genes ) , which could be indicative of loss-of-function ( Fig 3D , S5 Table ) ., The majority of the GATA3-wt and GATA3-ext signatures consisted of uniquely regulated , non-overlapping genes ( Fig 3D , S2B Fig ) ., Accordingly , gene ontology ( GO ) analysis revealed significant enrichment of gene sets relating to unique terms for each of the GATA3 constructs ( S5 Table ) ., For instance , the GATA3-wt gene set is enriched for cytokine-linked processes , whereas the GATA3-ext signature shows a significant enrichment for peptidyl-tyrosine modification processes ., These results indicate that expression of GATA3-ext and GATA3-trunc invoke starkly distinct changes in gene expression , and the large number of uniquely regulated genes in GATA3-ext cells supports a gain-of-function of this mutant ., We found a small 4-gene overlap between the TCGA and MCF10A GATA3-ext signatures ( S2B and S2C Fig ) ., We validated one of these , the triglyceride metabolism gene PNPLA3 , in an independent set of experiments by qRT-PCR and observed consistent downregulation in cells expressing GATA3-ext ., ( S2D and S2E Fig ) ., This is in agreement with patient tumour data ( S2F Fig ) ., Yet , the observation that most signature genes derived from the ER-negative MCF10A cell line model do not overlap with the patient data reflects the well-known biological differences between patient tumour samples and cell culture model systems ., Together , these data indicate that GATA3-ext is functionally active upon overexpression and that GATA3-ext and GATA3-trunc mutants are mechanistically different from each other and from the wild-type protein ., Chemical genetic interactions can reveal therapeutic vulnerabilities and pinpoint cellular processes that are affected by mutant proteins 58 , 59 ., Therefore , we performed a chemical genetic screen to identify compounds that specifically affect GATA3-ext cells ., We assembled a small-molecule library containing ~100 approved and experimental anti-cancer drugs , and a number of tool compounds ., We used MCF10A cells expressing the GATA3-ext protein or a control vector and exposed them to compounds for 6 days before measuring viability ., To mimic limited nutrient supply in a tumour and to render the cells more responsive to drugs , cells were treated under reduced media supplement conditions ( S3A and S3B Fig , S6 Table ) ., The compound that most strongly reduced GATA3-ext viability was BIX101294 60 , 61 , a specific inhibitor of the histone 3 lysine 9 methyltransferases G9A ( EHMT2 ) and GLP ( G9A-like protein; EHMT1 ) ( Fig 3E , S6 Table ) ., As expected , BIX01294 reduced global histone 3 lysine 9 dimethylation ( H3K9me2 ) in MCF10A-GATA3-ext and control cells over time ( Fig 3F ) ., We validated this unexpected interaction with short- and long-term treatment and in both full and reduced media supplement conditions ( Fig 3G and 3H and S3C Fig ) ., The effective concentration resulting in a 50% inhibition of viability ( EC50 ) for cells expressing the GATA3-ext mutant was consistently 5- to 10-fold lower than in control cells ., To determine if this sensitivity was specific for GATA3-ext , we tested the compound on MCF10A cells expressing GATA3-wt or GATA3-trunc ., The sensitivity of these cells to the compound was identical to control cells infected with an empty vector ( Fig 3I and 3J ) ., Accordingly , MCF7 cells , which heterozygously express GATA3-trunc , display average sensitivity to BIX101294 when compared to a panel of 25 other breast ( cancer ) cell lines ( S3D Fig ) ., Next , we wished to rule out that the observed effects of GATA3-ext overexpression were due to a dominant negative effect ., To address this , we depleted endogenous GATA3 by shRNAs and tested if this could phenocopy GATA3-ext expression ., The knockdown did not result in enhanced sensitivity to BIX101294 ( S4A Fig ) ., We thus conclude that the sensitivity arises from a specific interaction between the drug and the extended GATA3 protein ., GATA3-ext mutations in patients are predominantly heterozygous , and as endogenous GATA3 protein levels in MCF10A cells are very low , we co-expressed GATA3-ext and GATA3-wt to assess whether the presence of GATA3-wt alters the differential toxicity of BIX0124 ., GATA3-ext+wt cells were equally sensitive as GATA3-ext cells ( S4B Fig ) ., This experiment suggests that the GATA3-ext-induced BIX01294 sensitivity is independent of the presence of a wild-type GATA3 allele ., Together , these data further highlight functional differences between GATA3 truncation and extension mutants and imply that extension mutants act by a mechanism that is different from typical loss-of-function or dominant negative effects ., G9A and GLP are histone methyltransferases ( HMTs ) that form a heterodimer and catalyse specific mono- and di-methylation at histone 3 lysine 9 ( H3K9 ) 62 ., Di-methylation of this residue is associated with transcriptional repression and has been demonstrated to occur aberrantly at tumour suppressor genes , often coinciding with upregulation of G9A 63 ., In the TCGA dataset , EHMT1 and EHMT2 are not differentially expressed in GATA3-ext tumours and do not show a segmentation pattern ( S5 Fig ) ., To assess the specificity of the synthetic interaction between GATA3-ext and G9A/GLP , we tested a second G9A/GLP inhibitor ( UNC0638 64 ) ., Although this compound did not score as a hit in the screen , possibly due to a suboptimal screening concentration , repeated validation showed a similar degree of hypersensitivity of GATA3-ext cells ( Fig 4A and 4B and S3C Fig , S4 Fig ) ., Next , we tested a set of inhibitors of various other HMTs and did not detect differential sensitivity ( Fig 4C–4F ) , suggesting that the interaction with GATA3-ext does not occur with histone methyltransferase activity in general ., Further , GATA3-ext and control cells were equally responsive to other structurally similar quinazoline compounds not targeting G9A/GLP , consistent with a specific and on-target effect of BIX01294 and UNC0638 ( Fig 4G , S6A Fig ) ., In order to verify the involvement of G9A and GLP more directly , we depleted them by shRNA in GATA3-wt , GATA3-ext and control cells ., Only the viability of GATA3-ext cells was significantly affected ( Fig 4H , S6B Fig ) , suggesting that both enzymes contribute to the sensitivity to BIX01294 and UNC0638 ., To characterise the mechanisms underlying the sensitivity of GATA3-ext cells to G9A/GLP inhibition , we first analysed potential cell cycle effects upon BIX01294 treatment ., We did not observe a difference in cell cycle progression between GATA3-ext and control cells as assessed by BrdU incorporation or DNA content ( S7 Fig ) ., However , GATA3-ext cells were more prone to undergo apoptosis upon drug treatment than control cells ( Fig 5A ) ., As GATA3 is functionally linked with ER expression and activity 34–37 , we also assessed the impact of ER signalling on sensitivity to G9A/GLP inhibition in GATA3-ext cells ., We expressed ERα in our MCF10A model and confirmed that ER target genes were induced upon ER expression and/or treatment with the ER agonist β-estradiol ( E2 ) ( Fig 5B and 5C ) ., The sensitivity of GATA3-ext cells to G9A/GLP inhibition was not significantly influenced by the level of ER signalling ( Fig 5D ) , suggesting a mechanism that is independent from previously described functional interactions of GATA3 ., Recent studies have begun to address the role of GATA3 in breast cancer ., GATA3 has been suggested as a negative regulator of epithelial-to-mesenchymal transition and metastasis but putative tumour promoting effects have also been reported 26 , 45 , 53–55 , 65–76 ., Critically , these studies have almost exclusively focussed on wild-type GATA3 and only a few have studied GATA3 truncating mutations ., To our knowledge , our study is the first that highlights and addresses the most frequent GATA3 mutation type , i . e . mutations resulting in an extended C-terminal protein ., Protein-extending mutations in cancer genes are unusual but not unprecedented ., Recently , frameshift extension mutations in CALR ( encoding calreticulin ) were identified in myeloproliferative neoplasms 77 and WT1 extension mutants have been described in Wilms kidney tumours 78 ., Cancer driver mutations are often divided into gain-of-function and loss-of-function mutations ., Loss-of-function mutations result in an inactive or less active protein , whereas gain-of-function mutations lead to a more active protein or acquisition of a different function ., Several observations in our study indicate that GATA3-ext proteins are mechanistically distinct from other GATA3 mutants and GATA3 wild-type , hinting toward a gain-of-function: First , GATA3-trunc mutants lack a larger part of the normal GATA3 protein sequence than GATA3-ext ., This makes it rather unlikely that GATA3-ext is more perturbed in its normal physiological function than other GATA3 mutants ., Second , in patients , GATA3-ext is associated with the differential expression of a distinct group of response genes that is not affected by other GATA3 mutants ., Differential effects on gene expression were also observed in the MCF10A cell line model expressing GATA3-ext and GATA3-trunc ., Third , we have found differences in outcome for patients harbouring GATA3-ext mutations , at least in the TCGA cohort ., There , GATA3-ext is associated with reduced disease-free survival compared to other GATA3 mutations , suggesting that these tumours display a different pathology with respect to recurrence ., Of note , all GATA3 mutations together correlated with improved disease-free and overall survival in a Chinese patient cohort 46 ., GATA3 mutations as a whole displayed a marginally significant trend to improved overall survival only in ER-positive patients in the TCGA and METABRIC cohorts 25 , 46 but not in a smaller Dutch study 76 ., Interestingly , GATA3 frameshift mutations were strongly underrepresented in the Chinese cohort ( 22% vs . 78% missense mutations ) as compared to TCGA ( 93% vs . 7% ) ., The authors suggest different mutational evolution of luminal breast cancer in different populations as an explanation for these discrepancies , with few Asian patients being included in the TCGA cohort 46 ., However , these studies do not discriminate between GATA3-ext , GATA3-trunc or other GATA3 mutations ., Our survival analysis indicates that indeed this separation is important , as only GATA3-ext mutations are associated with reduced disease-free survival ., Fourth , we observe strong genetic selection for +1 frameshift mutations , leading to one specific C-terminal extension ., Fifth , GATA3-ext is stable in cells and displays functional characteristics ( e . g . , drug sensitivities ) that are not observed in cells expressing other GATA3 proteins or cells in which GATA3 is depleted ., Taken together , these lines of evidence provide substantial support for the hypothesis that GATA3-ext adopts certain neomorphic functions that might replace or act in addition to its wild-type properties ., Importantly , our findings challenge the view that GATA3 only acts as a tumour suppressor that is downregulated or inactivated in breast cancer 14–16 , 19–26 , 79 ., This GATA3-ext gain-of-function hypothesis parallels TP53 mutations in certain aspects , including gain- and loss-of-function in the same gene ., For this reason , we have adopted the gain-of-function terminology in analogy to p53 and propose to label GATA3-truncation mutations as primarily loss-of-function and GATA3-extension mutations as gain-of-function ., Like GATA3 , p53 is a transcription factor that acts as a homo-oligomer , and hence , gain-of-function mutations do not necessarily imply a constitutively active form of the protein , as it is observed for many kinase gain-of-function mutants ., Instead , a plethora of different functions for oncogenic p53 have been described , including altered subcellular localization , changed DNA-binding affinities and a different spectrum of binding partners and target genes ., Ultimately , these activities can lead to enhanced proliferation , inhibition of apoptosis , chemoresistance , or increased invasiveness 8–10 ., It remains unclear how GATA3-ext exerts its specific activity ., It has been postulated 80 that the GATA3 C-terminus is essential for maintaining protein stability but we did not observe strong differences upon ectopic expression in MCF10A cells ., Therefore , an alternative mechanism is likely to underpin GATA3-ext function ., For instance , GATA3-ext may display differential binding partners or altered DNA binding sites ., The GATA3-ext protein rendered cells sensitive to inhibition of the G9A and GLP histone methyltransferases ., G9A and GLP are upregulated in a number of cancers , correlating with higher H3K9me2 levels and silencing of tumour suppressor genes 63 ., Intriguingly , wild-type GATA3 and G9A have been recently found to physically interact 65 ., The biochemical and functional interaction of GATA3 with histone methyltransferases may explain the changes of active histone modifications and altered enhancer accessibility in breast cancer cells depleted of GATA3 37 ., Yet , if and how this relates to drug sensitivity specifically in GATA3-ext expressing cells remains unclear ., Our MCF10A cell line model does not fully recapitulate the context of GATA3 mutations in tumours in several ways , among them the heterozygous mutation state and the ER status ., Due to lack of a more appropriate model system , we addressed these concerns separately by co-expression and knock-down experiments of mutant and wild-type GATA3 and ESR1 ( encoding ER ) ., Even though RNA sequencing data from the MCF10A model only show a marginal overlap with the TCGA patient-derived GATA3-ext signature on an individual gene level , we believe that the MCF10A cell line model provides a valid context to study basic mechanistic differences between GATA3-wt , GATA3-ext and GATA3-trunc ., The patient data and MCF10A model agree in that GATA3-ext and GATA3-trunc mutants act in a mechanistically different manner from each other and from the wild-type protein ., We believe that this finding is biologically and potentially clinically relevant despite the exact mechanisms not yet being understood ., In this regard , the identified synthetic lethal interaction between GATA3-ext and G9A/GLP inhibition provides the first clinically testable hypothesis for application of these drugs and the first lead for a treatment of this major subgroup of breast cancer patients ., Thus , further pre-clinical study of the uncovered gene-drug interaction is warranted ., Together , our study provides important insights into the function and potential druggability of one of the most frequent breast cancer mutants and a striking example of how different mutations in the same cancer driver c | Introduction, Results, Discussion, Materials and Methods | Patterns of somatic mutations in cancer genes provide information about their functional role in tumourigenesis , and thus indicate their potential for therapeutic exploitation ., Yet , the classical distinction between oncogene and tumour suppressor may not always apply ., For instance , TP53 has been simultaneously associated with tumour suppressing and promoting activities ., Here , we uncover a similar phenomenon for GATA3 , a frequently mutated , yet poorly understood , breast cancer gene ., We identify two functional classes of frameshift mutations that are associated with distinct expression profiles in tumours , differential disease-free patient survival and gain- and loss-of-function activities in a cell line model ., Furthermore , we find an estrogen receptor-independent synthetic lethal interaction between a GATA3 frameshift mutant with an extended C-terminus and the histone methyltransferases G9A and GLP , indicating perturbed epigenetic regulation ., Our findings reveal important insights into mutant GATA3 function and breast cancer , provide the first potential therapeutic strategy and suggest that dual tumour suppressive and oncogenic activities are more widespread than previously appreciated . | Cancer is a disease caused by genetic mutations ., Mutation patterns are often indicative of a gene’s function as either tumour promoting or tumour suppressive ., Here we describe the frequently mutated , but poorly studied , breast cancer gene GATA3 as a rare exception: We discover that two different functional classes of mutations in this gene can lead to either gain- or loss-of-function activities ., The most common type of mutations , resulting in an unusually extended protein , is associated with differential gene expression and decreased disease-free survival ., This mutant , in contrast to other mutations or the non-mutated protein , renders cells specifically vulnerable to inhibitors of two chromatin-modifying enzymes , the histone methyltransferases G9A and GLP ., Our findings shed light on the functional consequences of frequent GATA3 mutations in breast cancer and represent a first lead toward personalised therapy for a large subgroup of breast cancer patients . | medicine and health sciences, breast tumors, morphogenic segmentation, cancers and neoplasms, dna-binding proteins, oncology, mutation, developmental biology, gene types, pharmacology, frameshift mutation, morphogenesis, tumor suppressor genes, proteins, gene expression, breast cancer, histones, biochemistry, genetics, biology and life sciences, drug interactions, suppressor genes | null |
journal.pgen.1004403 | 2,014 | Rapid Evolution of PARP Genes Suggests a Broad Role for ADP-Ribosylation in Host-Virus Conflicts | Post-translational modifications ( PTMs ) of proteins regulate a wide variety of cellular processes , including several aspects of innate immunity against pathogens ., As a result , pathogens have evolved mechanisms to block , reverse or usurp this machinery in order to successfully replicate within their hosts 1 ., For example , numerous viruses subvert the dynamics of phosphorylation , employing kinases , substrate mimics and phosphatases to disrupt host signaling 1 ., Likewise , addition and removal of acetyl groups by histone acetyltransferases ( HATs ) and deacetylases ( HDACs ) can have a dramatic effect on viruses such as HIV , herpesviruses , polyomaviruses and papillomaviruses ., In response , several viral classes encode proteins to specifically disrupt host phosphorylation and acetylation 2 ., Beyond small-molecule PTMs , conjugation and cleavage of ubiquitin and ubiquitin-like molecules has emerged as an important point of cellular regulation that several viruses target or subvert in order to replicate 3 ., In contrast , ADP-ribosylation is still poorly characterized for its role in innate immunity , despite being one of the first identified PTMs ., Transfer of ADP-ribose ( ADPr ) from NAD+ ( nicotinamide adenine dinucleotide ) to proteins is catalyzed within eukaryotic cells by members of the PARP ( poly-ADP-ribose polymerase ) , or ARTD ( ADP-ribosyltransferase , diphtheria toxin-like ) protein family ( Figure 1A ) 4 , 5 ., The best-studied PARPs , including the founding member PARP1 , catalyze the formation of long , branched chains of ADP-ribose known as poly-ADP-ribose ( PAR ) 4 , 6 , 7 , 8 ., These PAR-forming enzymes perform critical housekeeping functions , such as nucleation of DNA-damage foci ( PARP1 and 2 ) and proper chromosome segregation during mitosis ( PARP5a ) 7 , 8 ., In contrast to these well-described functions , most human PARP proteins are poorly understood , in part due to their lack of conservation in model organisms such as C . elegans and D . melanogaster 4 , 9 , 10 ., In total , 17 genes in the human genome contain PARP domains , with each gene containing a variety of other functional domains that likely endow each PARP with their individual functions ( Figure 1A ) 4 , 10 ., Many of the poorly-characterized human PARP proteins are found in the cytoplasm 11 and are predicted to only catalyze addition of a single ADPr , rather than PAR , to proteins 4 , 9 , 10 ., Several recent descriptions of PARP functions in cellular signaling , miRNA regulation and stress granule formation 12 , 13 , 14 suggest that many functions for cytoplasmic ADP-ribosylation , especially mono-ADP-ribosylation , likely remain uncharacterized ., Moreover , the discovery that a subset of macrodomain containing proteins can , in addition to binding mono-ADP-ribosylated proteins , also remove mono-ADP-ribose from proteins 15 , 16 , sheds further light on the regulation and function of this dynamic PTM ., One function of ADP-ribosylation may be to regulate viral infectivity and pathogenesis , consistent with the role of other PTMs in immunity ., For example , both vaccinia virus 17 and herpes simplex virus 18 require ADP-ribosylation activity for viral replication ., Moreover , diverse RNA viruses , such as alphaviruses , hepatitis E virus , rubella virus and SARS coronavirus encode one or more macrodomains , potentially conferring the ability to specifically recognize , and possibly reverse , ADP-ribosylation upon these viruses 19 ., Mutations in the macrodomain of Sindbis virus led to reduced virulence in mice 20 ., Similarly , mutations in the SARS coronavirus macrodomain sensitized the virus to the antiviral effects of the signaling cytokine , interferon ( IFN ) 21 ., As IFN functions as one of the primary mediators of the innate immune system against viruses 22 , these results indicate that macrodomains , and therefore ADP-ribosylation , could be important viral regulators of host immunity ., Moreover , host PARP genes can play a direct role in antiviral immunity ., For example , overexpression of PARP13 , also known as ZAP or ZC3HAV1 ( Zinc-finger CCCH-type antiviral protein 1 ) , is sufficient to restrict replication of several different families of viruses , including a retrovirus ( murine leukemia virus 23 ) , filoviruses ( Ebola and Marburg 24 ) , a togavirus ( Sindbis 25 ) and a hepadnavirus ( Hepatitis B virus 26 ) ., This antiviral activity is mediated through direct binding of viral RNA by PARP13 , followed by recruitment of the exosome and specific degradation of viral RNA 27 , 28 , although more recently , additional signaling roles for PARP13 have been proposed 14 , 29 ., Beyond the well-described PARP13-mediated antiviral functions , PARP1 , 7 , 10 and 12 have been shown to play roles in repressing viral replication 30 , 31 , 32 , 33 , although the mechanisms of these antiviral actions are unknown ., While these results indicate that there may be a role for individual PARPs in regulating viral infectivity or pathogenesis , there has been no cohesive model for how ADP-ribosylation may influence host-viral interactions ., We reasoned that if ADP-ribosylation is the focus of a host-virus conflict , we might see evolutionary signatures of positive ( diversifying ) selection acting on the specific host genes involved ., Positive selection is a hallmark of host genes locked in genetic conflict with viruses that counter-evolve to evade the host antiviral defenses , and has been seen in both antiviral kinases and antiviral ubiquitin ligases 34 ., Positive selection is characterized by the accumulation of amino acid-altering , nonsynonymous changes in the DNA at a rate that is greater than the accumulation of neutral , synonymous changes ., When such protein changes are recurrently selected for ( due to their adaptive advantage ) , the ratio of nonsynonymous to synonymous substitution rates exceeds one ( dN/dS > 1 , where dN is the nonsynonymous substitution rate and dS is the synonymous substitution rate ) ., Such analyses can not only identify a gene that has evolved under positive selection but can also pinpoint domains and even individual codons within that gene located at the direct interface between host and viral factors 35 , 36 ., We previously analyzed primate PARP13 orthologs to determine if the direct antiviral activity of PARP13 has led to a genetic conflict with viruses ., Indeed , consistent with its antiviral function , we found a robust signature of positive selection in PARP13 in primates 37 ., Interestingly , despite the fact that the zinc-finger domains of PARP13 directly bind viral RNA 27 , we found no signature of positive selection in these domains ., Instead , we found sites of positive selection in the PARP catalytic domain , implying that this domain is a target for genetic conflict with viruses 37 ., Although this domain in PARP13 appears to lack catalytic activity 4 , we nevertheless found that its removal from PARP13 decreased the level of viral restriction 37 , arguing that some function of the PARP domain remains intact ., Thus , using an evolutionary signature of positive selection as a guide , we were able to identify a domain important for the antiviral activity of PARP13 ., To address whether ADP-ribosylation plays a broad role in viral immunity , we wished to take a comprehensive evolutionary approach to look for evidence of rapid evolution in all of the human PARP genes ., We reasoned that evolutionary signatures of recurrent adaptation , such as those previously observed in PARP13 , might reveal other uncharacterized PARP proteins that are involved in host-virus interactions ., We therefore screened all 17 human PARP genes and their primate orthologs for signatures of recurrent positive selection ., Contrary to expectations that most PARP genes are involved in ‘housekeeping’ functions , we found that nearly one third of human PARP genes bore signatures of recurrent genetic conflicts ., In addition to PARP13 , our evolutionary screen revealed four other PARP genes that have evolved under very strong positive selection in primates: PARP4 , 9 , 14 and 15 ., Two of these genes ( PARP14 and 15 ) have also undergone dramatic gene turnover ( gain and loss ) during vertebrate evolution , an additional hallmark of gene innovation also seen in innate immunity genes such as APOBEC3 and TRIM5 38 , 39 ., Based on their rapid evolution , we hypothesize that these four additional PARP genes are involved in as-yet-undescribed host-virus conflicts ., Importantly , we anticipate that the identification of these rapidly evolving PARP genes and domains will enable future experiments to elucidate the role ADP-ribosylation plays in viral replication and host immunity ., Motivated by our hypothesis that ADP-ribosylation may be an important PTM in host-virus conflicts , and our prior use of positive selection analyses to identify an important antiviral domain in PARP13 , we investigated whether any of the other 16 human PARP genes also show signatures of recurrent positive selection ., We searched publicly available primate genome sequences and identified orthologs of all 17 human PARP family members from a minimum of four hominoids , two Old World monkeys and one New World monkey ., We performed a series of maximum likelihood analyses to detect recurrent positive selection for each PARP gene alignment ., These analyses determine whether a model allowing positive selection at a subset of amino acid residues is a statistically better fit to the sequence data than a model that does not allow for positive selection ., Using PAML software 40 , we found that five PARP genes showed highly statistically significant ( p-values <0 . 0001 ) signatures of positive selection ( Figure 1B ) ., In addition to confirming our earlier findings on PARP13 , we found that PARP4 ( also known as vPARP ) and the three macrodomain-containing PARP genes ( PARP9/BAL1 , PARP14/BAL2 and PARP15/BAL3 ) all show signatures of positive selection ., We followed up our PAML analyses with the more conservative PARRIS software implemented in the HyPhy package 41 , which takes into account recombination and variation in synonymous substitution rates across codons ., Using PARRIS , we again found these five PARP genes to be clearly distinct from the remaining 12 as judged by likelihood ratio tests ( LRT ) allowing or disallowing positive selection ( Figure 1B ) ., While our limited screen of seven orthologs in PARRIS only gave a statistically significant p-value ( <0 . 01 ) for PARP4 and PARP13 , analysis of additional sequences of PARP9 , PARP14 and PARP15 met statistical significance ( see below ) ., Finally , we performed branch-site analyses 42 to look for episodic signatures of positive selection on all 17 primate PARPs ., We found that only PARP4 , PARP9 and PARP13 demonstrated statistically significant signatures of episodic positive selection ( Figure S1 ) ., This initial screen might underestimate the total number of PARP genes evolving under positive selection , firstly because our search is restricted to the primate lineage ( selection might have operated only in other mammalian lineages ) and secondly because we use only seven orthologs ., Although such small alignments may lack power to detect weak selection , previous simulation studies have shown that strong selection on a subset of residues can be detected using PAML even with rather limited species surveys 43 ., Given the signatures of positive selection we observed for PARP4 , 9 , 14 and 15 in this initial screen , we characterized these four genes in further detail , collecting additional orthologous sequences to examine which domains contain positively selected residues in order to create a model for how viral conflict may have driven their evolution ., PARP4 , also known as vPARP ( vault PARP ) is a catalytically active poly-ADP-ribosyltransferase that is a component of widely conserved , large cytoplasmic ribonucleoprotein structures known as “vaults” ., Vaults are barrel-shaped particles composed of three proteins , MVP ( major vault protein ) , PARP4 , and TEP1 ( telomerase associated protein ) , as well as vRNA ( vault RNA ) 44 ., The function of vaults is unknown , but they have been implicated in drug resistance , cancer and immunity ., In support of a role in immunity , MVP , the core structural component of the >10 mDa mass of vaults , is upregulated by IFN , and vaults are most highly expressed in immune cell types such as dendritic cells and macrophages 45 ., From the alignment of seven PARP4 orthologs , we noted a ∼360 amino acid region that was much more divergent than the rest of the protein ( Figure 2A , Alignment S1 ) ., This protein segment is completely encoded by the largest exon of PARP4 ( exon 30 in humans ) ., To illustrate the unusual selective pressures on exon 30 , we performed a pairwise dN/dS comparison of human and rhesus PARP4s ., We found that whereas the overall dN/dS ratio over PARP4 is 0 . 63 , the dN/dS ratio for exon 30 alone is 1 . 75 ( >95% confidence for dN/dS > 1 ) ( Table S3 ) ., This striking discrepancy between the evolution of exon 30 and the rest of the protein raised the possibility that this exon alone was responsible for the signature of positive selection in PARP4 ., We therefore repeated our positive selection analyses with exon 30 alone and found a robust signature of positive selection ., In contrast , the remainder of PARP4 showed no signature of positive selection upon removal of exon 30 ( Figure 2B ) ., Although we cannot formally rule out the possibility of weak selection acting outside exon 30 in PARP4 , our analysis strongly suggests that exon 30 of PARP4 has uniquely evolved under strong recurrent positive selection in primates ., Because this evolutionary signature is isolated to a single exon , we next asked whether exon 30 is ever excluded from the PARP4 transcript ., We searched human and rhesus expressed sequence tag ( EST ) databases and found that all isoforms of PARP4 include exon 30 , suggesting that exon 30 is important for PARP4 function ., Next , we searched the region encoded by exon 30 for sequence or structural homology to other protein domains ., Surprisingly , secondary structure prediction software ( JPRED 46 ) indicated that the region encoded by exon 30 in human PARP4 is almost entirely disordered ., Taken together , we conclude that PARP4 has evolved under recurrent positive selection in primates , but that positive selection is focused on the disordered region encoded by exon 30 alone ., We explored the signature of adaptive evolution in exon 30 of PARP4 in more detail by sequencing exon 30 from genomic DNA from additional primates ( Table S1 ) ., Analysis of a total of 15 primate PARP4 exon 30 sequences confirmed our initial screening results that this region has evolved under positive selection ( PAML p-value <0 . 0001 , PARRIS p-value <0 . 01 ) ( Figure S2A and Alignment S1 ) ., These analyses also identified several codons within exon 30 that display dramatic signatures of recurrent positive selection ( Table S4 ) ., For instance , despite being in close proximity in the primary sequence to codons that are strictly conserved across primates , codon 1517 has undergone at least six amino acid changes during approximately 45 million years of simian primate evolution , with a calculated dN/dS ratio >3 ( Figure 2C ) ., We also found that this pattern of rapid evolution in exon 30 extends to other vertebrate lineages ., Despite high conservation in the rest of the PARP4 protein , the sequence and length of the largest exon ( corresponding to human exon 30 ) in PARP4 is highly variable among vertebrates ., Consistent with our results in primates , all closely related pairs of vertebrate PARP4 orthologs analyzed demonstrated a signature of purifying selection throughout much of PARP4 contrasting with evidence for positive selection in the region corresponding to exon 30 of human PARP4 ( Figure 2D and Table S3 ) ., To gain further insight into PARP4 evolution outside of primates , we asked whether other mammalian lineages show evidence for recurrent positive selection as we observed in primates ., To do this , we took advantage of publicly available bat genome sequences , which , like primates , are divergent enough to provide sufficient evolutionary divergence , but not so divergent that the rate of synonymous mutation ( dS ) is saturated ., Using sequences from 10 bat species ( Alignment S2 ) , we again found that PARP4 has evolved under recurrent positive selection in its largest exon ( PAML p-value <0 . 0001 , PARRIS p-value <0 . 01 ) ( Figure S2B-C ) ., PAML identified six positively sites with high confidence ( Figure S2B-C , Table S5 ) ., Although there is no overlap between positively selected sites identified in primates and bats , we found nine residues to be absolutely conserved across all 25 primate and bat species we analyzed ( Figure S2B-C ) , suggesting substantial constraint even within this rapidly evolving disordered protein domain ., Combined , these broader phylogenetic analyses indicate that a single PARP4 region has been subject to positive selection throughout mammalian and bird evolution , suggestive of an ancient conflict with intracellular pathogens ., Our evolutionary screen also revealed strong signatures of positive selection in PARP9 , PARP14 and PARP15 ., Strikingly , these three genes encode the only three human proteins that contain both a PARP catalytic domain and macrodomains , and are the only human genes to encode more than one macrodomain ., The macrodomain is unique among protein domains in its ability to recognize mono-ADP-ribosylated proteins 47 ., Furthermore , some macrodomains have recently been shown to catalyze the removal of mono-ADPr 15 , 16 ., Although the molecular functions of macro-PARPs are unclear , the presence of both PARP domains and macrodomains may conceivably allow them to both add and specifically recognize and/or reverse protein ADP-ribosylation ., This , combined with the presence of macrodomains in viruses , prompted us to explore in more depth the evolution of other human macrodomain-containing proteins and ADP-ribosylhydrolases ., Apart from the macro-PARPs , we found no evidence for positive selection in any other human gene encoding a macrodomain or ADP-ribosylhydrolase ( Figure S3 ) , suggesting that the combination of the macro- and PARP domains is important for their rapid evolution and , consequently , for their putative antiviral roles ., In order to further pinpoint which domains and codons in the macro-PARP genes have evolved under positive selection , we sequenced additional macro-PARP orthologs from a diverse panel of primates ., Combining these with publically available sequences , we aligned and analyzed 15 or more orthologs for each macro-PARP gene ( Table S1 , Figure S4 ) ., Based on these expanded alignments , we confirmed the results of our initial screen; all macro-PARP genes have evolved under positive selection in simian primates ( PAML p-value <0 . 0001 , PARRIS p-value <0 . 01 ) ., In contrast to the recurrent positive selection on only a single exon of PARP4 , we found that positively selected sites were broadly distributed throughout the macro-PARP genes ( Figure 3A ) ., For all three macro-PARPs , we observed strong evidence of positive selection acting on the macrodomains ., However , removal of the macrodomain-containing segments did not result in a loss of positive selection signatures , indicating that both macrodomains as well as other domains have evolved under positive selection ( Figure 3B ) ., For instance , we found significant evidence for positive selection in the PARP domain of PARP14 ( Figure 3B ) , similar to PARP13 37 ., In contrast , our analyses did not reveal evidence of positive selection acting on the PARP domains of PARP9 and PARP15 ( Figure 3B ) , although it is possible that sequencing of additional orthologs might reveal more subtle signatures of selection ., Thus , we conclude that macro-PARPs are evolving very rapidly , including substantial positive selection in the macrodomains of all three macro-PARPs ., Our finding that macrodomains encoded by macro-PARP genes have evolved under positive selection motivated us to investigate whether equivalent residues were rapidly evolving in each macrodomain ., Such a conserved pattern could suggest that related genetic conflicts ( for example , similar viral pathogens ) drove their evolution ., Instead , we observed that a different set of residues is rapidly evolving in each macro domain at a primary sequence level ( Figure 3C , Tables S6-S8 and Alignment S3 ) ., While equivalent amino acids are not evolving in all macro-PARPs , it is possible that positive selection has acted on a single three-dimensional protein surface ., We therefore modeled the positively selected residues from PARP9 and PARP14 macrodomains onto a structure that has been determined for the first macrodomain of PARP14 48 ., We found that positively selected residues map to a single surface of each macrodomain , but that each macrodomain shows positive selection on a distinct surface ( Figure 3D ) ., As each positively-selected surface is distinct relative to the site of ADP-ribose binding , these results suggest that ADP-ribose binding is not being altered or optimized by positive selection of the macrodomains ., Instead , our findings suggest that each macrodomain has engaged in its own evolutionary arms race with as-yet-unidentified pathogen factors ( see Discussion ) ., Because most antiviral genes do not serve essential housekeeping functions , they can be lost during periods when selective pressures are relieved , for example during periods when fewer relevant viral pathogens are prevalent in the population ., In contrast , selection to increase the breadth of antiviral specificities could also lead to increase in gene copy number 34 ., As a result of these repeated rounds of innovations , many organisms undergo dramatic changes in their antiviral gene repertoires over evolutionary timeframes , as has been observed with APOBEC and TRIM genes in mammals 38 , 39 ., In our initial evolutionary screen , we had observed that most of PARP15 is missing from the white-cheeked gibbon genome ., Coupled with previous findings of PARP15 absence in the mouse genome 10 , we therefore investigated PARP genes in general , with an emphasis on the macro-PARP genes , for signatures of rapid gene turnover ., From our investigation of all seventeen PARP genes across a wide range of vertebrates , we found that PARP15 is unique in its pattern of recurrent loss ( Figure 4A ) ., In contrast , other PARP genes are present in all genomes we examined , with the exception of PARP10 , which has been lost in the carnivore lineage ., To explore the dynamics of PARP15 birth and loss , we conducted a more in-depth survey of PARP15 genes in vertebrate genomes ( Figure S5 ) ., We found that PARP15 was born early in mammalian evolution via a partial duplication of PARP14 , consisting of the second and third macrodomains and the PARP domain ., We found that PARP15 has been independently lost via deletion or inactivating mutations in five different mammalian lineages; PARP15 is therefore absent from gibbons , all glires ( rodents and lagomorphs ) , the cow/sheep/dolphin clade , alpaca/camel , and armadillo ( Figure 4B ) ., Elephant and manatee have a conserved but shorter form of PARP15 , missing the first of the two macrodomains ., In contrast to these losses in PARP15 , we identified several PARP14 duplications that occurred both within and outside the lineage that contains PARP15 ., For instance , although fish and birds lack PARP15 orthologs , many fish and bird genomes have one or more additional copies of PARP14 that could possibly serve PARP15-analogous functions ., Guinea pig and bushbaby each appear to have at least one extra intact copy of PARP14 , with the caveat that in each case a single exon is within a genome assembly gap ., The microbat ( Myotis lucifigus ) genome contains at least eight PARP14/15 genes , of which at least two copies are intact ( two additional genes are incomplete in the assembly but are uninterrupted in available sequence by stop codons or frameshifts , suggesting they are also intact ) ( Figure 4B ) ., Moreover , pairwise comparisons of duplicated PARP14 genes in microbat and bushbaby suggest that these paralogs may have regions that have rapidly diverged under positive selection ( Figure S7 ) , although additional sequences will be required to strengthen such a conclusion ., Coupled with our findings that both PARP14 and PARP15 are evolving under positive selection in primates ( Figure 3 ) , the gene turnover we describe for PARP14 and PARP15 supports the idea that these genes have been selected for functional innovation , perhaps in response to a recurrent genetic conflict with pathogens ., Post-translational protein modifications are a common regulatory mechanism for modulating protein activity , stability and localization ., As such , numerous viruses manipulate host PTM machinery to regulate their own replication or evade host antiviral immunity ., Research aimed at understanding these viral strategies has provided critical insight into the host processes mediated by PTMs , including tyrosine phosphorylation and regulation of histone acetylation 1 , 2 ., Inspired by the fact that signatures of positive selection can be used to highlight important genes and PTMs in host-virus conflicts , we performed an evolutionary screen on all of the primate PARP genes to ask if ADP-ribosylation is an important player in host-virus dynamics ., Contrary to what would be expected of a PTM that is solely dedicated to housekeeping functions , we found strong evidence for rapid evolution in five of seventeen primate PARP genes , suggesting a broad involvement for PARPs , and ADP-ribosylation , in genetic conflicts ., Moreover , we observed evolutionary signatures that suggested an ancient history of conflict for these PARP genes ., For example , we see positive selection on PARP4 in diverse mammalian clades and recurrent gain and loss of PARP14 and PARP15 across vertebrates ., Our findings suggest that PARP4 , 9 , 13 , 14 and 15 are each locked in a genetic conflict , likely with one or more pathogenic agents ., Our data do not exclude the possibility that other genetic conflicts , perhaps in addition to viral conflicts , drove PARP positive selection ., Indeed , the first discovery of manipulation of host processes by ADP-ribosylation emerged from the study of bacterial toxins ( e . g . , diphtheria , cholera toxins ) 49 , leaving open the possibility that bacterial or eukaryotic pathogens drove the evolution of PARP genes ., However , we hypothesize that viruses may be significant or even the primary pathogens in these evolutionary arms races for several reasons ., First , numerous viruses replicate poorly when ADP-ribosylation is inhibited , including viruses that replicate in the nucleus ( HSV ) 18 and cytoplasm ( vaccinia ) 17 ., Second , several families of mammalian RNA viruses , including corona- and togaviruses , have non-structural proteins that contain macrodomains ., In both corona- and togaviruses , disruption of viral macrodomains has been shown to reduce virulence 20 , 21 , and in the case of coronaviruses , this reduced virulence is due to increased sensitivity to the antiviral activity of interferon ( IFN ) 21 ., This suggests a simple model in which the macrodomains ( at least in coronaviruses ) are required to counteract some IFN-stimulated host gene product ., Although the identity of that IFN-stimulated factor is unknown , we note that several of the rapidly evolving PARP genes we identify here , including PARP9 , PARP13 and PARP14 , are upregulated by IFN 50 , 51 ., Furthermore , overexpression of PARP9 , independent of IFN , is sufficient to upregulate several known antiviral effectors 50 ., Finally , overexpression of several PARP genes has been shown to inhibit replication of viruses , the most well-described example being PARP13 ., Taken together , we favor a model in which PARP gene evolution has been driven primarily by genetic conflicts with viruses ., The patterns of evolution of the PARP genes allow us to make several inferences about the role of these proteins in genetic conflicts ., First , the fact that we observe a robust evolutionary signature of positive selection in PARP4 , 9 , 13 , 14 and 15 argues strongly that these genes are important for organismal fitness ., Similar to strong evolutionary conservation , signatures of positive selection indicate that fixation of a particular allele , in this case , a novel allele , results in a strong enhancement of fitness ., While rapid evolution may seem antithetical to functional constraint , in fact positive selection is a common hallmark of critical host immunity genes 34 ., Thus , we infer that the functions of the rapidly evolving PARP genes we have identified are important for fitness in the face of rapidly-evolving pathogens ., Second , we also find that PARP14 and PARP15 show recurrent gene duplication and loss ., This form of genetic innovation is another common hallmark of immunity genes ., Gene losses occur during periods of relaxed selection due to non-exposure or extinction of relevant pathogen ( s ) , whereas gene duplications often provide additional genetic substrates for diversifying selection to increase anti-pathogen repertoires 34 , 52 ., While other PARP proteins , such as PARP1 , PARP7 , PARP10 and PARP12 30 , 31 , 32 , 33 , have been identified as having antiviral functions , our initial screen suggests they have not been subject to strong recurrent antagonistic evolution with viral factors in primates , perhaps because their encoded proteins do not directly interact with virus-encoded factors ., Instead , our analyses lead to our novel hypotheses that PARP4 , PARP9 , PARP14 and PARP15 , as well as the molecular complexes they reside in , possess antiviral activity ., For instance , PARP4 is a component of large cytoplasmic structures known as vaults , whose functions are poorly understood ., Although vaults are extremely ancient , dating back to the origin of eukaryotes , they have been lost in multiple lineages 9 , suggesting that they are not universally necessary to perform an essential , housekeeping function ., Instead , there are several tantalizing pieces of evidence that vaults may be involved in immunity ., These include an increased number of vaults in immune cell types , IFN-upregulation of MVP , the major component of vaults , and upregulation of noncoding vault RNAs ( vRNAs ) on infection with pathogens such as Epstein-Barr virus 45 ., PARP4 itself is present at ∼10 molecules per vault , but its functional role there is unknown 44 ., However , our observation that the positively selected residues we find in PARP4 are localized to a single disordered region in PARP4 suggests a model for its role in vault-mediated immunity ., Such a localized pattern of positively selected sites is reminiscent of two well-characterized rapidly evolving antiviral factors , TRIM5a and MxA , shown to be on the offensive ( i . e . directly binding to viral proteins ) side of the host-virus conflict 34 ., TRIM5a and MxA both use their rapidly evolving regions , also in the context of multimeric complexes , to directly recognize and target viral proteins , lentiviral capsids in the case of TRIM5a and orthomyxovirus nucleoproteins in the case of MxA 35 , 36 ., Thus , we infer that the positively selected region of PARP4 ( exon 30 in humans ) has evolved to maintain recognition of a factor encoded by pathogens that can infect many diverse mammalian lineages , or is a common means to counteract independent unrelated pathogens ., This interaction may be used to directly ADP-ribosylate viral components , which could affect their activity and impede infection ., Alternatively , independent of ADP-ribosylation , PARP4 interaction may recruit viral proteins to the vault struct | Introduction, Results, Discussion, Materials and Methods | Post-translational protein modifications such as phosphorylation and ubiquitinylation are common molecular targets of conflict between viruses and their hosts ., However , the role of other post-translational modifications , such as ADP-ribosylation , in host-virus interactions is less well characterized ., ADP-ribosylation is carried out by proteins encoded by the PARP ( also called ARTD ) gene family ., The majority of the 17 human PARP genes are poorly characterized ., However , one PARP protein , PARP13/ZAP , has broad antiviral activity and has evolved under positive ( diversifying ) selection in primates ., Such evolution is typical of domains that are locked in antagonistic ‘arms races’ with viral factors ., To identify additional PARP genes that may be involved in host-virus interactions , we performed evolutionary analyses on all primate PARP genes to search for signatures of rapid evolution ., Contrary to expectations that most PARP genes are involved in ‘housekeeping’ functions , we found that nearly one-third of PARP genes are evolving under strong recurrent positive selection ., We identified a >300 amino acid disordered region of PARP4 , a component of cytoplasmic vault structures , to be rapidly evolving in several mammalian lineages , suggesting this region serves as an important host-pathogen specificity interface ., We also found positive selection of PARP9 , 14 and 15 , the only three human genes that contain both PARP domains and macrodomains ., Macrodomains uniquely recognize , and in some cases can reverse , protein mono-ADP-ribosylation , and we observed strong signatures of recurrent positive selection throughout the macro-PARP macrodomains ., Furthermore , PARP14 and PARP15 have undergone repeated rounds of gene birth and loss during vertebrate evolution , consistent with recurrent gene innovation ., Together with previous studies that implicated several PARPs in immunity , as well as those that demonstrated a role for virally encoded macrodomains in host immune evasion , our evolutionary analyses suggest that addition , recognition and removal of ADP-ribosylation is a critical , underappreciated currency in host-virus conflicts . | The outcome of viral infections is determined by the repertoire and specificity of the antiviral genes in a particular animal species ., The identification of candidate immunity genes and mechanisms is a key step in describing this repertoire ., Despite advances in genome sequencing , identification of antiviral genes has largely remained dependent on demonstration of their activity against candidate viruses ., However , antiviral proteins that directly interact with viral targets or antagonists also bear signatures of recurrent evolutionary adaptation , which can be used to identify candidate antivirals ., Here , we find that five out of seventeen genes that contain a domain that can catalyze the post-translational addition ADP-ribose to proteins bear such signatures of recurrent genetic innovation ., In particular , we find that all the genes that encode both ADP-ribose addition ( via PARP domains ) as well as recognition and/or removal ( via macro domains ) activities have evolved under extremely strong diversifying selection in mammals ., Furthermore , such genes have undergone multiple episodes of gene duplications and losses throughout mammalian evolution ., Combined with the knowledge that some viruses also encode macro domains to counteract host immunity , our evolutionary analyses therefore implicate ADP-ribosylation as an underappreciated key step in antiviral defense in mammalian genomes . | innate immune system, genetics of the immune system, clinical immunology, immunity, biology and life sciences, comparative genomics, immunology, computational biology, evolutionary biology, evolutionary genetics, immune system | null |
journal.pgen.1000756 | 2,009 | New Evidence Confirms That the Mitochondrial Bottleneck Is Generated without Reduction of Mitochondrial DNA Content in Early Primordial Germ Cells of Mice | Mammalian mitochondrial genome shows a 5 to 10 times greater mutation rate than the nuclear genome 1 , 2 ., This elevated mutation rate coupled with clonal maternal transmission leads to the high mtDNA polymorphism in populations ., However , despite the prevalence of genetic variance within a species , most individuals possess only a single mtDNA variant ., Pedigree analyses of heteroplasmic individuals in cattle , mice and humans revealed that mtDNA genotypes shift rapidly among offspring and return to homoplasmy in some progeny within a few generations 3–8 , suggesting that a mtDNA bottleneck accounts for the rapid segregation ., Early studies have proposed that the bottleneck occurs in embryonic development in consequence of a drastic reduction of mtDNA content in PGCs 9 , 10 ., In mice , the size of the bottleneck is estimated as to be ∼200 mtDNA segregation units 11 ., To test these hypotheses , three independent research groups have attempted to quantify mtDNA copy number in single germ cells at different developmental stages in mice ., Cao et al . 12 made the first direct measurements of mtDNA copy number in single PGCs ( identified by endogenous alkaline phosphatase ( ALP ) activity ) in wild-type mice using quantitative real-time PCR ( qrt-PCR ) and found that PGCs contained consistent amounts of mtDNA with a mean of ∼1350–3600 copies per cell between 7 . 5 days post coitum ( dpc ) and 13 . 5 dpc , indicating that the bottleneck occurs without a marked reduction of mtDNA copies in PGCs ., Recently using Stella-GFP transgenic mice to isolate PGCs , a study determined a mean of ∼450 mtDNA copies per PGC at 7 . 5 dpc ( median ∼200 ) and a mean of ∼1100–2200 copies between 8 . 5 dpc and 14 . 5 dpc ., The drastic reduction in PGC mtDNA content at 7 . 5 dpc was suggested to be the cause of the bottleneck 13 ., Taking advantage of using Oct4ΔPE-EGFP mice heteroplasmic for two mtDNA sequence variants , Wai et al . 14 measured both mtDNA copy number and heteroplasmy in single germ cells ., They detected that PGCs possessed a mean of ∼280 mtDNA copies ( median 145 ) per cell at 8 . 5 dpc , the earliest stage at which PGCs could be isolated in their mouse strain , and a mean of ∼2000–6000 copies per germ cell between 9 . 5 dpc and 16 . 5 dpc ., Interestingly , despite the remarkable low mtDNA content in 8 . 5 dpc PGCs , an increase in genotypic variance at any point during embryonic oogenesis was not found ., Instead an inequality of genotypic variance in germ cells between postnatal day 8 and day 11 was discovered ., It was concluded that the mitochondrial genetic bottleneck occurs not during embryonic oogenesis but during postnatal oocyte maturation through replicating a subpopulation of genomes 14 ., The two reports by Cao et al . 12 and Wai et al . 14 , therefore , reached concordant conclusions that the genetic bottleneck is not attributed to the decline in PGC mtDNA content ., However , it remains unknown why three studies aimed to determine mtDNA copy number in single PGCs produced distinct results at early developmental stages , that is , while Cao et al . 12 detected no severe decrease of mtDNA copies in PGCs , the other two studies found a significant reduction in PGC mtDNA content at 7 . 5 and 8 . 5 dpc , respectively 13 , 14 ., One proposed explanation is that the ALP histochemical staining used for PGC isolation may confound mtDNA estimation made by qrt-PCR , while the PGC-specific markers , Stella-GFP and Oct4ΔPE-EGFP , omit the staining step and may interfere less with qrt-PCR amplification 13 , 14 ., The other possible explanation is that the isolation of PGCs using flow cytometry may introduce PGC sample contamination 14 , 15 ., Given the important implications of a mitochondrial bottleneck in mtDNA disease inheritance , it is vital to determine the true germline mtDNA copy number in order to fully understand the underlying mechanisms of the bottleneck ., In this study we manually isolated PGCs from mice expressing fluorescent proteins specifically in PGCs to by-pass ALP histochemistry and flow cytometry sorting procedures , and examined mtDNA copy number in single PGCs at different embryonic development stages ., The present results confirm that no severe reduction of mtDNA content occurs in PGCs ., Blimp1 expression marks nascent PGCs as well as precursors of PGCs in early developing mouse embryos ., In the restricted posterior region ( after removal of visceral endoderm ) of 7 . 5 dpc embryos , Blimp1 has been proved to express specifically in PGCs 19 ., To facilitate the isolation of PGCs without any staining steps at 7 . 5 dpc , we used bacterial artificial chromosome ( BAC ) transgenic mice in which monomeric red fluorescent protein gene ( mRFP ) was inserted into the Blimp1 locus 16 ., First we determined the Blimp1-mRFP expression profile ., At both EB and LB stages mRFP expression was observed at the posterior end of the embryonic ectoderm and visceral endoderm in embryos ( Figure 1B and 1J ) , consistent with the endogenous Blimp1 expression 19 ., To further characterize Blimp1-mRFP positive cells , embryos were immuostained for Stella ( PGC7 ) , a PGC-specific marker , with the anti-Stella antibody whose specificity for PGCs has been proved 21 ., Stella was detected exclusively in PGCs located at the posterior region of the embryos ., All Stella positive cells were Blimp1-mRFP positive ., At EB and LB stages Stella protein was expressed in 62 . 5% and 90% of Blimp1-mRFP positive cells in the posterior region , respectively ( embryo number\u200a=\u200a2 at both stages ) ( Figure 2 ) , which is comparable with the results of Seki et al . 22 ., As a second assay , we stained cells isolated from the posterior fragments ( visceral endoderm removed ) of Blimp1-mRFP embryos for alkaline phosphatase , another classical marker of PGCs ., Cells were divided into Blimp1-mRFP positive and negative two groups prior to staining ., At the EB stage , Blimp1-mRFP positive and negative cells were 79 . 4% ( n\u200a=\u200a34 ) and 0% ( n\u200a=\u200a28 ) positive for ALP staining , while at the LB stage , 92 . 9% ( n\u200a=\u200a28 ) and 0% ( n\u200a=\u200a40 ) , respectively ( Figure 3 ) ., The results agree with that Blimp1 expression precedes that of Stella and alkaline phosphatase 19 ., The cells positive for Blimp1-mRFP but negative for Stella are either precursors of PGC or PGCs with weak Stella expression beyond the detection limit of anti-Stella antibody ., Taken together , the Blimp1-mRFP expression profile of our Blimp1-mRFP line is highly similar to that of Blimp1 transgenic mice reported 19 , 22 ., Our Blimp1-mRFP , therefore , can be used as a reliable marker to identify PGCs in the posterior region of the embryo at EB and LB stages ., The mean numbers of mtDNA molecules in single 7 . 5 dpc EB and LB , 13 . 5 dpc female and 13 . 5 dpc male PGCs were 1396 , 1479 , 1747 and 2039 , respectively ., Of note , no extremely low mtDNA copy number was detected in any PGCs , and the minimum number of mtDNA copies determined in single PGCs was 767 ., The variation in mtDNA copy number per cell was similar for each group of PGCs ( CV\u200a=\u200a0 . 25–0 . 45 , Table 1 , Figure 4 ) ., At 13 . 5 dpc no significant difference in the average mtDNA copy number between female and male PGCs was observed ( t-test , P\u200a=\u200a0 . 06 ) ., In contrast , somatic cells from the gonad were found to contain less than half the number of mtDNA copies found in PGCs at 13 . 5 dpc ( mean mtDNA copy number in somatic cells\u200a=\u200a702 ) ( Table 1 , Figure 4 ) ., All these results are , therefore , in very good agreement with previous estimates obtained from PGCs identified by alkaline phosphatase activity 12 , taking into account that the cell samples in the present study were collected randomly without cell size classification ., The present data confirm our previous findings 12:, i ) There is no occurrence of remarkable reduction of mtDNA copies in early PGCs;, ii ) The amount of mtDNA molecules in PGCs is moderate ( mean >1000 copies , comparable with that in adult somatic cells 23 ) and consistent across stages; and, iii ) Embryonic somatic cells possess much lower mtDNA amount than PGCs ., How can we explain the discrepant findings among Cree et al . 13 , Wai et al . 14 and ours ?, Cree et al . 13 detected median ∼200 mtDNA copies ( mean 451 ) in PGCs at 7 . 5 dpc and more than 1000 copies at later stages ., This result may be associated with the aspects of their methodology for PGC sorting ., Cree et al . 13 identified and sorted PGCs by flow cytometry ., Isolation of PGCs using flow cytometry was shown to be inaccurate for early developmental stage embryos 14 , 15 ., Szabo et al . 15 carried out PGC sorting from Oct4ΔPE-EGFP transgenic mice that were generated using the construct identical to that in Yoshimizu et al . 18 ., This Oct4ΔPE-EGFP were reported to be specifically expressed in PGCs from 8 . 5 dpc onwards 17 , 18 ., At 8 . 5 dpc ( five somites ) only 62% EGFP+ cells sorted by flow cytometry were positive for the PGC marker , whereas 96% or more were positive at 9 . 5 dpc and later stages 15 , indicating the inaccuracy of flow cytometry in identifying GFP cells expressing a PGC marker at early developmental stages when PGCs are small in number relative to non-PGC cells in the sample ., The number of PGCs in single embryos at 7 . 5 dpc is even smaller than at 8 . 5 dpc ., Therefore it is possible that more than 38% of GFP+ cells sorted and analyzed by Cree et al . 13 at 7 . 5 dpc were not PGCs but somatic cells ., Embryonic somatic cells have been shown to contain significantly lower amounts of mtDNA than PGCs 12 , 24 ., By studying whole embryos , Aiken et al . 24 reported that the mean of mtDNA copies per somatic cell was ∼300 between 6 . 5 dpc and 18 . 5 dpc , comparable with the 451 copies in the 7 . 5 dpc PGC sample of Cree et al . 13 ., In contrast , our isolation of 7 . 5 dpc PGCs was performed manually under a fluorescent microscope using micromanipulators ., Hence the purity of PGCs was ensured ( Figure 1 ) ., Wai et al . 14 found a median of 145 mtDNA copies ( mean ∼280 ) per PGC at 8 . 5 dpc and more than 1000 copies at other stages ., It is currently unclear why the study of Wai et al . 14 gave the result significantly different from that of Cree et al . 13 and ours for 8 . 5 dpc PGCs ., However , the fact that both Cree et al . 13 and Cao et al . 12 detected a mean of mtDNA copies more than 1000 per 8 . 5 dpc PGC weakens the possibility of extremely lower mtDNA copy number ( mean ∼280 ) in 8 . 5 dpc PGCs ., Why do PGCs contain more than 1000 mtDNA copies in sharp contrast with consistent ∼300 per embryonic somatic cell across stages between 6 . 5 dpc and 18 . 5 dpc ?, One explanation is that the moderate mtDNA copy number in PGCs allows the cell to have an elevated tolerance for less deleterious mtDNA mutations and serves as a device to maintain the adaptive potential of mtDNA genome , which for a nuclear genome is achieved via diploidy by means of sexual reproduction ., On the other hand , a tight physical bottleneck in somatic lineages during embryonic development enables pathogenic mtDNA mutations ( both severe and less deleterious variants ) to rapidly segregate and be eliminated more efficiently ., The mtDNA copy number in gonadal somatic cells at 13 . 5 dpc was much higher ( mean 702 ) than the average mtDNA level per cell of the whole embryo ( mean ∼300 ) 24 , suggesting that the physical bottleneck in embryonic somatic cells appears lineage specific which might be associated with tissue-specific cell function and bioenergic demands ., Alternatively , the moderate mtDNA copy number may be entailed to meet the energetic demand for PGC migration , mitotic replication and cell function ., Recent findings from two research groups have provided new insight into the mtDNA segregation between generations ., Wai et al . 14 showed that the mtDNA genetic bottleneck occurs not during embryonic oogenesis but during postnatal oocyte maturation as a result of replication of a subgroup of mtDNAs ., Fan et al . 25 demonstrated a strong and rapid purification selection of mtDNA in the germline , eliminating severe mutations within three generations in mice ., If the purifying filter takes effect only before the cease of PGC proliferation by 13 . 5 dpc , females carrying severe mtDNA mutations would produce progeny with a similar frequency of mutant variants in any given litter ., However the female mice harbouring either ND6 frameshift mtDNA or mtDNA with a 4696-bp deletion produced offspring with a declining proportion of mutant mtDNA in their successive litters 25 , 26 , indicating that the purifying filter acts after 13 . 5 dpc ., Data from ND6 frameshift mtDNA mice have shown that the selection appeared not to operate at embryonic stage through selectively eliminating foetuses with the highest percentages of mutation 25 ., One possibility is that the purifying selection acts at the cell level through a competition between germ cells with different proportions of severe mutant variants ., When germ cells with the highest ratio of severe mutations are held in the ovary for a long time they might gradually lose their competence in competing to develop into mature oocytes , as opposed to those with lower ratio of mutations , due to the cumulative detrimental effect of mutation on cells over time ., Consequently the proportion of severe mutant mtDNA would be lower in younger siblings and the severe mutation would progressively be lost and eventually disappear in following generations , which are the patterns revealed in the mouse pedigrees harbouring either a ND6 frameshift mutation 25 or a mtDNA large deletion mutation 26 ., In human pedigrees , a decline in the mtDNA mutation load in younger siblings has not been found 27–30 ., Pedigree studies of pathogenic mutations in human mtDNA have mainly focused on large deletion mutations and point mutations ( either missense mutations in protein genes or base substitution mutations in tRNA and rRNA genes ) ., These point mutations are apparently milder than the ND6 frameshift mutation resulting in a premature termination of gene transcription 25 and the mtDNA large deletion mutation where a number of genes were missing from the genome 26 , and hence , may not undergo a strong purification selection in the germline ., Deletion mtDNAs in humans were suggested not to be maternally transmitted , rather to arise spontaneously , since mothers and siblings of affected individuals rarely harboured deletion mtDNAs 31 ., It is possible that the purification selection of germline mtDNA is much stronger in humans than in mice , and virtually no large deletion mtDNAs in human oocytes would survive this selection ., In rare cases where a deleted genome slips through , clonal expansion of such a deleted molecule ( s ) in embryogenesis could lead to mitochondrial disease in the child ., Taken together , these may explain the failure to find a decrease in mtDNA mutation load in offspring as maternal age increases in human pedigrees ., Alternatively , severe mtDNA mutations could be selected against at the organelle level 32 ., Mitochondria with a higher proportion of mutation diminish with time , leading to a decline of mutant mtDNA in oocytes ., The selection at the organelle level could be oocyte-specific , since it is not supported by measurements of deletion mutation loads in non-dividing somatic tissues in which mutant mtDNAs accumulate with age 33 , 34 ., It is known that the number of mtDNA copies per mitochondrion decreases during oogenesis , and eventually reaches approximately a mtDNA per mitochondrion in mature oocytes 35 ., This decline of mtDNA copy number in single mitochondria may render the mitochondrion more sensitive to severe mutations and facilitate the elimination of mutant mitochondria in oocytes ., In summary , our new data show solid evidence that there is no drastic reduction of mtDNA copy number in PGCs ., The amount of mtDNA in single PGCs is moderate and comparable at 7 . 5 and 13 . 5 dpc stages , while somatic cells in 13 . 5 dpc gonads contain far fewer mtDNA copies than PGCs ., The results from this and our previous study 12 are highly consistent despite using different mouse lines ( wild-type mice vs . transgenic mice ) and different PGC isolation methods ., These results reinforce our conclusion that the mitochondrial bottleneck occurs without reduction of mtDNA copy number in germline cells , which is also supported by the study showing the mtDNA genetic bottleneck occurrence as a result of replication of a subpopulation of mtDNA genome and not a remarkable reduction in PGC mtDNA content 14 ., The rapid segregation of mtDNA variants between generations is likely achieved through a genetic bottleneck resulting from replication of a subgroup of mitochondrial genome during oocyte maturation 14 in combination with a strong purifying selection against severe mutations over a long time period in the germ line 25 ., The reconfirmation of mtDNA copy numbers in PGCs in the present study is of importance because it clarifies the confusions and the information of germline mtDNA content will facilitate the development of therapeutic strategies blocking the mitochondrial disease transmission from mother to progeny ., Blimp1-mRFP transgenic mice ( BDF1 background ) were generated using a 203-kb bacterial artificial chromosome ( BAC ) expressing mRFP under the control of Blimp1 regulatory elements as previously described 16 ., Oct4ΔPE-GFP mice ( BDF1 background ) were generated using the construct identical to that of Yoshimizu et al . s 18 , carrying the GFP sequence driven by an 18-kb Oct4 genomic fragment with deletion of the proximal enhancer 16 ., The Blimp1-mRFP mouse strain ( RBRC01830 ) and Oct4ΔPE-GFP mouse strain ( RBRC00821 ) were provided by the RIKEN BioResource Center ( http://www . brc . riken . jp/inf/en/ ) ., All animal experiments were approved by the Institutional Animal Experiment Committee of RIKEN BioResource Center and of the Tokyo Metropolitan Institute of Medical Science ., 7 . 5 days post coitum ( dpc ) embryos of Blimp1-mRFP mice were collected in Dulbeccos Modified Eagles Medium ( DMEM ) ( Invitrogen ) supplemented with 10% FCS , and were further precisely classified as at either early bud ( EB ) or late bud ( LB ) stage according to the morphological landmarks 20 ., Immunohistochemistry assay was carried out as in Sugimoto and Abe 16 ., Briefly , isolated embryos were incubated in PBS containing 0 . 5% Triton X-100 for 3–5 min on ice and fixed with 4% paraformaldehyde in PBS for 10 min at room temperature ., Following stringent washing , the embryos were incubated with anti-Stella primary antibody ( 1∶1 , 000 ) ( kindly provided by Dr . Toru Nakano ) diluted with blocking buffer ( PBS containing 1% BSA and 0 . 1% Triton X-100 ) for 1 h at room temperature , washed with washing buffer ( PBS in 0 . 1% Triton X-100 ) , and incubated with Alexa-Fluor-488 conjugated goat-anti-rabbit secondary antibody diluted in blocking buffer for 45 min at room temperature ., Fluorescent images were acquired in Z series every 0 . 82 µm using Zeiss LSM510 Meta confocual microscope ., Posterior fragments after removal of visceral endoderm from Blimp1-mRFP embryos of EB and LB stages were dissected out ., The fragments were disaggregated with Trypsin ., Cells were divided into Blimp1-mRFP negative and positive groups , and underwent alkaline phosphatase ( ALP ) staining as in Cao et al . 12 ., ( C57BL/6N×DBA/2N ) F1 female mice were mated with Blimp1-mRFP and Oct4ΔPE-GFP males ., 7 . 5 and 13 . 5 dpc embryos were collected in DMEM supplemented with 10% FCS ., The restricted posterior parts of 7 . 5 dpc EB and LB Blimp1-mRFP embryos bearing PGCs ( visceral endoderm removed , Figure, 1 ) and the gonads of 13 . 5 dpc Oct4ΔPE-GFP embryos were isolated using fine needles , and incubated in Trypsin-EDTA solution ( Sigma ) at 37°C for 10 min ., Cells were dissociated in DMEM supplemented with 10% FCS by pipetting , were pelleted and washed twice in PBS and were resuspended in modified Dulbeccos phosphate-buffered medium ( PBI ) ., Single PGCs expressing fluorescent proteins and the gonad somatic cells were randomly collected under a fluorescent microscope using micromanipulators ( Leica ) ., PGCs from female and male gonads were isolated separately at 13 . 5 dpc , at which time the gender of the fetus can be identified microscopically ., Single cells were extracted and used directly for qrt-PCR analysis as previously described 12 , 36 ., Quantification of absolute mtDNA copy number per cell was carried out by the standard curve method ., Details of the probe , primers and standard DNA are as in Cao et al . 12 ., To evaluate whether pure DNA of plasmids containing mouse mtDNA fragment is suitable to serve as standard DNA for the measurement of mtDNA copy number in single cells , we performed qrt-PCR on samples containing the plasmid DNA in the presence of a single mtDNA-less ρ0B82 cell using pure plasmid DNA as standard DNA as described previously 12 , and compared the sample plasmid copy numbers estimated by PCR with their corresponding true copy numbers ., ρ0B82 line was derived from L cells 37 , 38 whose mtDNAs encompass the same mtDNA fragment integrated into the plasmid ., Mixing the plasmid DNA with a ρ0B82 cell creates sample conditions closer to that of single cell DNA for PCR amplification ., It showed that the estimated copy values highly correlated with their true copy values ( r2>0 . 99 ) over a range of mtDNA concentrations that comprise the values detected in our single cell studies ( Figure S1 ) , indicating that a systematic underestimation or overestimation of actual plasmid copies did not occur ., Therefore the pure plasmid DNA is suitable for using as standard DNA for our determination of mtDNA copies in single cells ., This qrt-PCR system could detect as few as 10 copies of standard DNA template reliably ., The linear regression analysis of all standard curves for samples with copy numbers between 10 and 106 showed a high correlation ( r2>0 . 99 ) ., Samples from both 7 . 5 and 13 . 5 dpc were measured in at least duplicate plates to avoid systematic errors that may be caused by transfer of standard DNA or other factors ., The means of mtDNA copy number determined from the plates were assessed by t-test ., No significant differences were found between the plates carrying the samples of matched categories . | Introduction, Results/Discussion, Materials and Methods | In mammals , observations of rapid shifts in mitochondrial DNA ( mtDNA ) variants between generations have led to the creation of the bottleneck theory for the transmission of mtDNA ., The bottleneck could be attributed to a marked decline of mtDNA content in germ cells giving rise to the next generation , to a small effective number of mtDNA segregation units resulting from homoplasmic nucleoids rather than the single mtDNA molecule serving as the units of segregation , or to the selective transmission of a subgroup of the mtDNA population to the progeny ., We have previously determined mtDNA copy number in single germ cells and shown that the bottleneck occurs without the reduction in germline mtDNA content ., Recently one study suggested that the bottleneck is driven by a remarkable decline of mtDNA copies in early primordial germ cells ( PGCs ) , while another study reported that the mtDNA genetic bottleneck results from replication of a subpopulation of the mtDNA genome during postnatal oocyte maturation and not during embryonic oogenesis , despite a detected a reduction in mtDNA content in early PGCs ., To clarify these contradictory results , we examined the mtDNA copy number in PGCs isolated from transgenic mice expressing fluorescent proteins specifically in PGCs as in the aforementioned two other studies ., We provide clear evidence to confirm that no remarkable reduction in mtDNA content occurs in PGCs and reinforce that the bottleneck is generated without reduction of mtDNA content in germ cells . | Mutations of mtDNA are responsible for many types of mitochondrial diseases in humans , including myopathy and neurological disorders ., Females carrying a mixture of mutant and wild-type mtDNA variants transmit a variable amount of mutant mtDNA to each offspring ., The proportion of mutated mtDNA inherited from the mother determines the onset and severity of diseases ., Studies have suggested that the mtDNA genome is transmitted through a bottleneck , but the underlying mechanism remains controversial ., By detecting mtDNA copy number in single cells , we previously showed that the bottleneck occurs without reduction of mtDNA content in germline cells ., However , recently a study reported a marked decline of mtDNA copies in embryonic germ cells and attributed this reduction to the creation of the bottleneck ., Yet another study concluded that the bottleneck occurs during postnatal oocyte maturation and not during embryonic oogenesis ., To resolve these controversies , we examined mtDNA copies in embryonic germ cells identified using the same methodology as in the other two studies ., We show solid evidence to confirm our previous findings ., This confirmation is important because the understanding of mtDNA content in female germ cells will facilitate the development of therapeutic strategies preventing the transmission of mitochondrial diseases from mother to offspring . | genetics and genomics/animal genetics | null |
journal.pgen.1000071 | 2,008 | The Impact of Recombination on Nucleotide Substitutions in the Human Genome | Genomic landscapes are not uniform across vertebrate chromosomes ., Notably , the genomes of amniotes ( mammals , birds and reptiles ) show a very strong heterogeneity of base composition along chromosomes ( the so-called isochores ) ( for review , 1 ) ., These Mb-scale variations in GC-content result from variations of substitution patterns that have affected both coding and non-coding regions ., These genomic landscapes are correlated with many other important features ( gene density , intron size , distribution of transposable elements , replication timing ) ., Thus , isochores clearly reflect some fundamental aspects of genome organization ., Although isochores have been discovered more than 30 years ago 2 , the reason for their origin is still highly debated: are they the result of selection 3–8 , or do they simply reflect variations in neutral substitution patterns 9–15 ?, Unraveling the origin of isochores ( neutral evolution or selection ) is essential to understand the functional significance ( if any ) of this peculiar genomic organization ., Moreover , a better knowledge of genome-wide variations in neutral evolutionary processes is also important for practical reasons ., Indeed , comparative sequence analysis is commonly used to identify genes or regulatory elements within genomes ., The basic principle of this approach is that functional elements are subject to the action of natural selection and therefore , their pattern of sequence variation ( within populations or between different species ) differs from what would be expected under the null hypothesis of neutral evolution ., Hence , to be able to detect functional elements within genomes it is crucial to understand the parameters that affect the neutral processes of sequence evolution ., Recently , different lines of evidence have suggested that isochores might be a consequence of the process of recombination ( for review , 16 ) ., Notably , analyses of the pattern of substitution in primate non-coding sequences have shown that recombination affects the relative rate of AT→GC and GC→AT substitutions 15 , 17 , 18 ., We and others have proposed that this effect might result from the neutral process of biased gene conversion ( BGC ) 11 , 14 , 19 , 20 ., According to this model , gene conversion ( i . e . the copy/paste during meiotic recombination of one allele onto the other one at heterozygous loci ) is biased in favor of GC-alleles , which leads to an increase of probability of fixation of GC-alleles compared to AT-alleles ., Thus , BGC should lead to an enrichment in GC-content in genomic regions of high recombination compared to regions of low recombination ., Understanding the impact of BGC on genome evolution is of fundamental importance ., Indeed , the effect of BGC is very similar to that of directional selection 21 , and hence BGC can confound the tests that have been developed to detect selection in genomic sequences 22 ., Although many lines of evidence support the BGC hypothesis 16 , there remain several important theoretical problems with this model , pointed out by Spencer and colleagues 23 ., First , it is now clearly established that in humans , recombination occurs predominantly in hotspots ( typically 2 kb long ) that cover about 3% of the genome 24 ., If recombination affects only very short regions , how can it drive the evolution of GC-content in Mb-long genomic fragments ?, Second , the analysis of human SNPs has shown that there is a fixation bias in favor of GC-alleles ( in agreement with the BGC model ) , but that this bias is relatively weak 23 ., Furthermore , the location of recombination hotspots is not conserved between human and chimpanzee , which indicates that hotspots have a short lifespan 25 , 26 ., Given these spatial and temporal fluctuations in recombination rate , is it possible that the BGC process affects the evolution of base composition ?, Some other authors have proposed that it is the base composition of sequences ( and not recombination ) that is the major determinant of substitution patterns 13 ., Indeed , the rate of cytosine mutation depends directly on the DNA melting ( and hence on the GC-content of sequences ) ., Therefore , the GC-content is expected to affect the relative rate of AT→GC and GC→AT substitutions ., Given that GC-content and recombination rate are positively correlated , this effect could contribute to the correlations between recombination rate and substitution patterns that were previously reported 15 , 17 , 18 ., To address these issues we performed two complementary analyses ., First , we took advantage of newly available data ( fine scale crossover map in humans and complete genome sequences of human , chimpanzee and macaque ) to re-assess the genome-wide relationship between patterns of substitution and recombination , controlling for the impact of GC-content ., For this purpose , we developed a new method to compute substitution rates for individual nucleotides , taking into account the hypermutability of CpG dinucleotides and the non-stationarity of base composition ., This method is based on a maximum-likelihood ( ML ) approach , and hence is more reliable than the parsimony approach used previously ., Second , we modeled the process of BGC , taking into account recombination hotspots , to theoretically assess the potential impact of this molecular drive on the evolution of genome landscapes ., Our analyses confirm that recombination is the major determinant of the evolution of GC-content and allows us to definitively reject selectionist models of isochore evolution ., Moreover , these analyses shed light on the evolution of recombination rate since the divergence between human and chimpanzee , on the distribution of non-crossover recombination events and on the differences in patterns of recombination between males and females ., Finally , theoretical calculations demonstrate that despite the short lifespan of recombination hotspots , BGC can have a strong impact on genome evolution ., In previous works , we had used parsimony to infer substitutions 15 , 18 ., While this concept is very simple and powerful for closely related sequences , it fails as divergence among sequences increases 30 , 31 ., Notably , because of CpG mutation hotspots , parsimony may fail at reconstructing sequences of the human/chimp last common ancestor 16 ., Hence , we had to exclude from our analyses many sites for which the ancestral state was ambiguous 15 , 18 ., One can avoid such problems using the maximum likelihood approach , which was pioneered by Felsenstein 32 ., In this framework one searches the parameters of the substitution rate matrix that maximizes the likelihood of sequence data given a stochastic model of nucleotide substitutions ., However the various ML methods to phylogeny reconstruction that have been proposed previously , make at least one of the following assumptions:, ( i ) the substitution model is time-reversible and the same in all branches of a given tree ( only the branch length might vary from one branch to another , not all substitution processes are considered independently ) ,, ( ii ) the genomes under considerations are in the stationary state with respect to this model , and, ( iii ) neighbor dependent nucleotide substitutions can be neglected ., These assumption are thought to be necessary to efficiently compute the likelihood for a given substitution model and tree topology 32 ., However all these simplifying assumptions are not necessarily granted: notably , we know that the base composition is by far not constant and stationary for mammalian species 15 , 16 , 33–38 ., Moreover , the neighbor dependent and irreversible CpG methylation deamination process ( CpG→CpA/TpG ) is the predominant nucleotide substitution process in vertebrates 35 , 39 , 40 ., We introduce here a new ML method , that takes into account non-stationary and non-reversible processes ( as already proposed 41 , 42 ) and furthermore includes neighbor dependent substitutions processes , like the CpG methylation deamination process ., This approach is described in detail in the methods section ., We measured 7 substitution rates ( pooling together complementary rates ) : the 4 transversion rates ( A→C+T→G; A→T+T→A; C→A+G→T; C→G+G→C ) , the 2 transition rates at non-CpG sites ( A→G+T→C; G→A+C→T ) , and the transition rate at CpG sites ( G→A+C→T ) ., We will hereafter use the notation to indicate complementary substitutions ( e . g . A:T→G:C\u200a=\u200aA→G+T→C ) ., When convenient , we will use the notation W ( weak ) for A or T and S ( strong ) for C or G . Thus , the notation W→S indicates all substitutions ( transitions or transversions ) from A or T to G or C . Note that the total substitution rate ( K ) in a given genomic regions depends on its base composition and on the base-specific substitution rates ., In the model considered here ( with 7 base-specific substitution rates ) K is given by the following equation: ( 1 ) where FGC , FAT and FCpG denote the frequencies of the different categories of sites and the parameters denote the base-specific substitution rates ., We measured base-specific substitution rates independently in the human and chimpanzee lineages ., From these substitution rates , we inferred for each lineage the stationary GC-content of sequences ( hereafter noted GC* ) , using a method that accounts for CpG hypermutability 43 ., GC* corresponds to the GC-content that sequences would reach at equilibrium if patterns of substitution remained constant over time ., GC* therefore provides information about the recent trend of evolution of GC-content ., In fact , GC* can be considered as a summary statistics of the average substitution matrix during the last 6 Myrs ., It should be noticed that GC* is a measure of substitution patterns that is independent of the total substitution rate; it simply reflects the relative contribution of W→S and S→W substitutions to the total number of substitutions ., We first investigated the relationship between GC* , recombination rate and the regional base composition ( GC-content ) ., As an estimator of recombination rate , we took the rate of crossover from the HAPMAP genetic map 44 and from the deCODE genetic map 45 ., The HAPMAP genetic map is based on patterns of allelic associations , and hence reflects the sex-averaged crossover rate that occurred in human populations ( i . e . the historical crossover rate ) ., The deCODE genetic map is based on pedigree studies and provides both sex-averaged and sex-specific crossover rates ., In agreement with our previous results 15 , we found at the 1 Mb scale a strong correlation between GC* and the sex-averaged rate of crossover on autosomes , both with the HAPMAP data ( Pearson correlation R2\u200a=\u200a0 . 36 , Figure 1b ) and with the deCODE data ( R2\u200a=\u200a0 . 31 ) ., GC* is also strongly correlated with the local GC-content ( R2\u200a=\u200a0 . 25 , Figure 1a ) , but this correlation is weaker than with the crossover rate ., We observed that the pattern of substitution tends to decrease the GC-content of our genome: GC* is lower than the present GC , particularly in GC-rich regions ( Figure 1a ) ., However note that this process is extremely slow: since the divergence between human and chimpanzee ( about 6 Myrs ago ) , regions with more the 50% GC lost about 0 . 2% GC ., If these substitution patterns would not change in time , we can extrapolate that it would take at least 500 Myrs for such a region to reach a GC-content of 40% ., Thus , the human genome appears to be evolving toward a more homogenous and less GC-rich base composition , in agreement with previous findings 15 , 16 , 33–38 ., It should be noted that the correlation between GC* and the current GC is far from perfect ( 75% of the variance in GC* is not predicted by the current GC-content ) ., In other words , the GC-content toward which sequences are evolving is largely independent from the current GC-content ., Thus , the forces that have driven the evolution of isochores in mammalian genomes have changed both in intensity ( these forces are not strong enough to maintain GC-rich isochores ) and in localization along chromosomes ., GC* correlates strongly both with crossover rate and GC-content ., We have previously proposed that recombination was the major determinant of GC* 15 ., However , other authors also suggested that the GC-content was a strong direct determinant of GC* , because the rate of cytosine mutation depends directly on the DNA melting ( and hence on the GC-content of sequences ) 13 ., Given that GC-content and crossover rate are also positively correlated ( R2\u200a=\u200a0 . 15 , Figure 1c ) , this raises the question of which variables ( GC , recombination or both ) are truly involved in determining GC* , and which happen to covary simply because they are influenced by another , causal variable ., It has been proposed that a higher GC-content might promote recombination 46–48 ., Indeed , in human , recombination hotspots occur preferentially in locally GC-rich regions 23 ., Thus , if GC-content determines both the recombination rate and GC* , this could explain the correlation between the rate of crossover and GC* ., However , in agreement with our previous analyses 15 , we found that the rate of crossover correlates much more strongly with the stationary GC-content ( GC* ) than with the present GC-content ( GC ) ( compare Figure 1b and 1c ) : the crossover rate explains 36% of the variance in GC* , compared to only 15% of the variance in GC ., The same pattern is observed on the X chromosome ( Table 1 ) ., If the correlation between GC* and crossover rate was due to the impact of base composition on recombination , then we would have expected a much stronger correlation of the rate of crossover with the present GC-content than with the stationary GC-content ( i . e . the future GC-content of sequences ) ., Our observations therefore definitively demonstrate that at the genomic scale considered here ( 1 Mb ) , recombination drives the evolution of GC-content ., This does not exclude however that the GC-content might also affect GC* ., Indeed , multivariate regression indicate that both GC-content and crossover rate are significant predictors of GC* ( p<10−10 ) ., Thus , the correlation between GC* and GC is not simply an indirect consequence of the correlation between GC and crossover rate ., Taken together , GC and crossover rate explain 44% of the variance of GC* ., We investigated the correlation between crossover rate and GC* separately in introns and intergenic regions ., We found similar correlations for all kinds of non-coding sequences ( Table 1 ) , which indicates that recombination affects the evolution of base composition in all genomic compartments , transcribed or not ., HAPMAP and deCODE sex-averaged crossover rates are not perfectly correlated ( R2\u200a=\u200a0 . 53 at the 1 Mb scale ) , which indicates that these data are noisy ., It is presently not known to which extent this noise is due to the imprecision of the methods used to estimate crossover rates or to real variations in crossover rates during the evolution of human populations ( given that recombination rates evolve rapidly , crossover rates estimated from pedigree-based genetic maps may differ from historical crossover rates ) ., But in any case , this indicates that HAPMAP and deCODE crossover rates are not perfect predictors of the average recombination rate in the human lineage during the last 6 Myrs ., Thus , even if recombination was the unique determinant of GC* , we would not expect a perfect correlation between GC* ( which is inferred from the pattern of substitutions in the human lineage during the last 6 Myrs ) and the HAPMAP or deCODE crossover rates ., Taken together , HAPMAP and deCODE sex-averaged crossover rates explain 39% of the variance in GC* ( i . e . significantly more than each variable taken separately , p<10−10 ) ., However , this is certainly still an underestimate of the true correlation between GC* and recombination rate ., To try to better characterize the impact of recombination on sequence evolution , we searched for additional predictors of recombination rate ., It is known that in humans , the rate of recombination increases near telomeres 45 , 49 ., Indeed , there is a negative correlation between HAPMAP crossover rates and the distance to telomere ( in log scale , hereafter noted LDT ) ( R2\u200a=\u200a0 . 27 , p<10−10 ) ., We observed a strong negative correlation between GC* and LDT ( R2\u200a=\u200a0 . 35 , p<10−10 ) ( Figure 2a ) ., As shown above for crossover rates , LDT correlates much more strongly with GC* than with the current GC-content ( R2\u200a=\u200a0 . 19 , Figure 2b ) ., Again , this demonstrates that the correlation between LDT and GC* is not an indirect consequence of the correlation between LDT and GC ., To try to disentangle the contribution of the different variables ( crossover rate , GC-content and LDT ) to the variation of GC* , we performed a multivariate regression analysis ., By using a stepwise procedure , we found that the best two predictors of GC* are the HAPMAP crossover rates and LDT ( Table 2 , Supplementary Text S1 ) ., Taken together , HAPMAP crossover rate and LDT explain 47% of the variance in GC* at the 1 Mb scale ., The GC-content significantly improves the model , but the gain in accuracy of prediction is relatively modest ( R2\u200a=\u200a0 . 51 , Table 2 ) ., The addition of other variables ( deCODE sex-averaged , male or female recombination rates ) does not further improve the model ., To get a clearer picture of the dependencies of the stationary GC-content on the recombination rate and GC-content , we analyzed the base-specific substitution rates ( which are the underlying determinants of GC* ) according to crossover rate , LDT and the current GC-content ., Partial correlation analyses indicate that all base-specific substitution rates are affected negatively by the current GC-content and positively by recombination rate ( i . e . positively by crossover rate and negatively by LDT ) , but the strength of correlations with each variable varies greatly among base-specific substitution rates ( Table 2 ) ., Note that the effect of LDT on base-specific substitution rates is always parallel to that of crossover rate , which supports our assumption that LDT and crossover rate are two complementary predictors of the recombination rate ., Interestingly , S→W and W→W substitution rates show a very weak dependency on recombination rate , but a strong dependency on GC-content ( compare in Table 2 the R2 of the model including only recombination predictors – i . e . LDT and crossover rate - to the R2 of the full model ) ., Conversely , W→S substitution rates show a much stronger dependency on recombination rate than on GC-content ., This dependency of W→S substitution frequencies on the recombination rates is in the end responsible for the correlation of GC* on the recombination rate ., S→S substitution rates appear to be affected by both variables ., The fact that base-specific substitution rates are differently affected by GC-content and by recombination rate is clearly seen in pairwise correlation analyses ( Table 3; compare Figures 3 and 4 ) ., It should be noticed that the total substitution rate ( K ) is positively correlated to GC-content ( Figure 3a ) ., This might seem a priori unexpected given that base-specific substitution rates show either a negative correlation ( Figure 3b , c ) or no correlation with GC-content ( Figure 3d ) ., However , K depends not only on base-specific substitution rates but also on the base composition ( see equation ( 1 ) ) ., Thus , given that , S→W substitution rates are on average higher than their respective W→S back substitutions ( Table 3 ) , K tends to increase with the GC-content ( FGC in equation ( 1 ) ) ., In other words , the positive correlation between the total substitution rate and GC-content does not reflect a higher exposure of GC-rich regions to mutagenic factors , but simply a higher proportion of GC bases that are more prone to substitutions than AT bases ., Given the strong correlation between GC* and recombination rate , GC* can be used as an indicator to investigate the evolution of patterns of recombination ., Notably , it is presently not clear what is the time scale and genomic scale of evolution of recombination rate ., It has been recently shown that recombination hotspots evolve very rapidly ., Indeed , the locations of recombination hotspots in human and chimpanzee are totally uncorrelated , despite considerable sequence identity 25 , 26 , and it has been demonstrated that hotspot activity may vary strongly among individuals in human populations 50 ., Given our previous results , these rapid changes in fine scale recombination maps are expected to lead to variations in substitution patterns during time ., In apparent contradiction with that prediction , at the genomic scale considered here ( 1 Mb ) , we found a strong conservation of substitution patterns between human and chimpanzee lineages: the correlation between GC* measured in human and chimpanzee orthologous regions is R2\u200a=\u200a0 . 70 ( p<10−10 ) ., Notably , GC* measured in the chimpanzee lineage is more strongly correlated to the rate of crossover measured in human populations ( R2\u200a=\u200a0 . 36 , i . e . as strong as the correlation observed with human GC* ) , than to the current GC-content in chimpanzee ( R2\u200a=\u200a0 . 24 ) ., The only possible interpretation for this correlation is that at the Mb scale , rates of recombination are highly conserved between human and chimpanzee ., This conclusion is in agreement with the hypothesis proposed by Myers et al . ( 2005 ) 24 that , at the Mb scale , the regional hotspot density and activity remains fairly constant over relatively long evolutionary time , despite fine-scale changes in hotspot location ., This conclusion ( rapid local fluctuation of hotspot location , but conservation of regional hotspot density ) may explain the first paradox raised by Spencer and colleagues 23: although at a given time , hotspots occupy only 3% of the genome , on the long term , a large fraction of the genome may be affected by hotspot activity ., The conservation of recombination rate at the Mb scale probably reflects some constraints on the distribution of crossover events ., Indeed it is known that in mammals ( as in many other taxa ) , there is a requirement of one crossover per chromosome arm to ensure a proper segregation of chromosomes during meiosis ( for review , see 51 ) ., This constraint leads to a higher crossover rate in smaller chromosome arms 15 , 51–53 ., The resolution of the HAPMAP genetic map allowed us to investigate the correlation between GC* and recombination at finer scale ., The strength of correlations decreases with smaller window size ( Table 1 ) , and becomes very weak below 200 kb , possibly because at this scale , other factors contribute to variations in substitution patterns ., Interestingly , the correlation between GC* measured in human and chimpanzee orthologous regions remains high ( R2>40% ) , up to 200 kb ( Table 1 ) ( NB: this is an underestimate because the accuracy of the measure of GC* decreases with smaller window size 54 ) ., Moreover , GC* measured in the chimpanzee lineage shows the exactly same correlation to the rate crossover measured in human populations as GC* measured in the human lineage ( Table 1 ) ., This suggests that the regional hotspot density remains conserved between human and chimp at least up to the 200 kb scale ., The rate of meiotic recombination differs between males and females: the rate of crossover in autosomes is on average 65% higher in females than in males , and the genetic maps are poorly correlated between the two sexes ( crossover rates in females are higher around the centromeres , whereas those in males tend to be higher towards the telomeres ) 45 ., In a previous work , we had found that GC* correlated more strongly with female than with male recombination rate 15 ., However , this result was based on the analysis of 33 loci only , and the difference became non-significant after excluding only one data point 15 ., Moreover , the analysis of substitution patterns in Alu repeats lead to the opposite conclusion 17 ., To resolve that issue , we analyzed in our whole-genome data set , the correlation between GC* and sex-specific crossover rates provided by the deCODE genetic map ., We found that on autosomes , GC* is much more strongly correlated to male crossover rate ( R2\u200a=\u200a0 . 27 ) than to female crossover rate ( R2\u200a=\u200a0 . 15 ) ., On the X chromosome , that recombines only in females ( we excluded pseudo-autosomal regions from our analyses ) , we found a correlation between GC* and crossover rate that is weaker than that observed in autosomes ( deCODE: R2\u200a=\u200a0 . 22 , HAPMAP: R2\u200a=\u200a0 . 17 ) ., Thus , we confirm the observation of Websters and colleagues 17 , that male crossover rate is a much stronger predictor of GC* than female crossover rate ., We have previously reported different observations that support , qualitatively , the BGC model for the evolution of isochores 16 ., However , it is important to quantify more precisely the prediction of the BGC model: given that recombination occurs essentially in hotspots that cover only 3% of the genome , that the BGC effect in hotspots is weak , and that hotspots have a short lifespan , is it possible that BGC drive the long term evolution of the base composition of Mb-long sequences ?, To address that issue , we performed theoretical calculations to quantify the potential impact of BGC on genome evolution ., We considered a model of genome evolution , where sequences are only subject to mutations and to BGC ( i . e . no selection ) ., Advancing a model by Lipatov and colleagues 55 , we assume here a model in which BGC only occurs in hotspots , with all other DNA undergoing neutral evolution ., Let the fraction of the genomic region that is involved in a hotspot be f ., We assume that the mutation process is the same both in and out of hotspots and that the mutations rate from W→S is μw→s and the rate from S→W is μs→w ., Then the rate of substitution from W→S in a given genomic region is: ( 2 ) and the rate from S→W is ( 3 ) where N is the effective population size and P ( s ) is the probability that a mutation subject to BGC of strength s will be fixed ., BGC behaves just like selection of a semi-dominant mutation 21 so: ( 4 ) P ( 0 ) is the probability that a mutation , which is not subject to BGC , is fixed under random drift: i . e . P ( 0 ) =\u200a1/2N ., The rate of recombination varies along chromosomes , as a consequence of variations in density and intensity of recombination hotspots 24 ., Thus , the impact of BGC in a given genomic fragment depends on the local density and intensity of recombination hotspots ., We considered genomic fragments of 1 Mb ., We assume that at this genomic scale , and for the period of time considered here ( i . e . corresponding to the human/chimpanzee divergence ) , the hotspot density and average intensity remain constant during time ., However , we do not assume that hotspots remain at the same position within the fragment ., To investigate independently the impact of hotspot density and intensity on genome evolution we considered two models: in the first one ( M1 ) , we consider that the rate of recombination in a given genomic fragment varies only through the density in recombination hotspots , which are assumed to have all the same intensity; in the second one ( M2 ) , we keep the density of hotspots constant over across the chromosome but vary the intensity of hotspots in the genomic fragments ., The distribution of densities ( for M1 ) and intensities ( for M2 ) are chosen to mimic the actually observed genome wide distributions of recombination rates in the human genome ., The BGC coefficient ( s ) depends on the intensity of the hotspot, ( i ) ( i . e . its rate of recombination ) , the length of the heteroduplex, ( h ) and the bias in the repair of W:S mismatches ( b ) ., It is known that i varies among hotspots 56 ., There is presently no evidence for variations of b and h along chromosomes ., Hence we will simply assume here that variations in s reflect variations in i , so: ( 5 ) where i is the rate of recombination and k a constant factor ., We used equations ( 2 ) and ( 3 ) to compute S→W and W→S substitution rates predicted by the BGC model , independently for transversions , non-CpG transitions and CpG transitions ., S→S and W→W substitution rates are not affected by BGC , and hence were assumed to be identical to their mutation rates and constant across the genome ., For our calculations , we chose parameters as realistic as possible ., We considered a sequence with a base composition typical of the human genome ( i . e . GC-content\u200a=\u200a40 . 6% , CpG density\u200a=\u200a1% ) ( NB: we do not assume that the base composition of the sequence is at equilibrium ) ., We calculated substitution rates predicted by the model ( at CpG and non-CpG sites ) for a period of time corresponding to the human/chimpanzee divergence ., To estimate mutation rates , we took from our above analyses the average substitution rates measured in fragments of low recombination of human autosomes ( <0 . 44 cM/Mb , i . e . corresponding to the first 10% of the dataset ) ., Recombination rates in 1 Mb-long fragments of human autosomes were taken from HAPMAP data , and range from 0 . 02 cM/Mb to 4 . 71 cM/Mb ( 1 . 33 cM/Mb on average ) ., Recombination hotspots are typically 2 kb long , and cover 3% of our genome 24 ., Thus , the average intensity of recombination hotspots, ( i ) is 44 . 4 cM/Mb ., In model M1 , we consider that f varies from 0 . 05% to 10 . 7% ( with i\u200a=\u200a44 cM/Mb ) , whereas in model M2 , i varies from 0 . 66 cM/Mb to 157 cM/Mb ( with f\u200a=\u200a3% ) ., We considered an effective population size N\u200a=\u200a104 ., We presently have no direct measure of the BGC parameter within recombination hotspots , but the order of magnitude of this parameter can be estimated from the analyses of Spencer and colleagues 23 ., These authors computed the average BGC parameter ( 4Ns ) in large genomic regions by fitting a population genetics model to the frequency distribution of SNPs in human populations 23 ., They divided their genome-wide data set into quintiles of recombination rate and found that the average BGC parameter increases 2 . 6 fold from 4Ns\u200a=\u200a0 . 5 in genomic regions of low recombination ( i . e . the first 20% , average crossover rate\u200a=\u200a0 . 42 cM/Mb ) to 4Ns\u200a=\u200a1 . 3 in regions of high recombination ( i . e . the top 20% , average crossover rate\u200a=\u200a2 . 54 cM/Mb ) 23 ., Thus , in these highly recombining regions , the average value of k is kref\u200a=\u200a7 . 25 10−7 ( see equation ( 5 ) ) ., We computed GC* according to the substitution rates predicted by models M1 and M2 for several values of k ( from k\u200a=\u200akref to k\u200a=\u200a10 kref ) ., The values of k for which the correlation between GC* and crossover rate was the closest to the one observed in the data were k\u200a=\u200a4kref and k\u200a=\u200a5kref ( i . e . on average , within recombination hotspots , 4Ns\u200a=\u200a5 . 2 to 6 . 5 ) ., The hypothesis that k might be 4 to 5 times higher in recombination hotspots than in the set of highly recombining regions analyzed by Spencer and colleagues is perfectly plausible , given that the average crossover rate within recombination hotspots is 17 times higher ( 44 . 4 cM/Mb ) ., The correlation between GC* predicted by model M1 ( with k\u200a=\u200a4kref\u200a=\u200a2 . 9 10−6 ) and the rate of crossover in the human genome is presented in Figure 1b ( green dots ) ., The slope of the correlation is very close to that observed in real data ( blue dots ) ., Note that for the range of recombi | Introduction, Results, Discussion, Material and Methods | Unraveling the evolutionary forces responsible for variations of neutral substitution patterns among taxa or along genomes is a major issue for detecting selection within sequences ., Mammalian genomes show large-scale regional variations of GC-content ( the isochores ) , but the substitution processes at the origin of this structure are poorly understood ., We analyzed the pattern of neutral substitutions in 1 Gb of primate non-coding regions ., We show that the GC-content toward which sequences are evolving is strongly negatively correlated to the distance to telomeres and positively correlated to the rate of crossovers ( R2\u200a=\u200a47% ) ., This demonstrates that recombination has a major impact on substitution patterns in human , driving the evolution of GC-content ., The evolution of GC-content correlates much more strongly with male than with female crossover rate , which rules out selectionist models for the evolution of isochores ., This effect of recombination is most probably a consequence of the neutral process of biased gene conversion ( BGC ) occurring within recombination hotspots ., We show that the predictions of this model fit very well with the observed substitution patterns in the human genome ., This model notably explains the positive correlation between substitution rate and recombination rate ., Theoretical calculations indicate that variations in population size or density in recombination hotspots can have a very strong impact on the evolution of base composition ., Furthermore , recombination hotspots can create strong substitution hotspots ., This molecular drive affects both coding and non-coding regions ., We therefore conclude that along with mutation , selection and drift , BGC is one of the major factors driving genome evolution ., Our results also shed light on variations in the rate of crossover relative to non-crossover events , along chromosomes and according to sex , and also on the conservation of hotspot density between human and chimp . | Mammalian genomes show a very strong heterogeneity of base composition along chromosomes ( the so-called isochores ) ., The functional significance of these peculiar genomic landscapes is highly debated: do isochores confer some selective advantage , or are they simply the by-product of neutral evolutionary processes ?, To resolve this issue , we analyzed the pattern of substitution in the human genome by comparison with chimpanzee and macaque ., We show that the evolution of base composition ( GC-content ) is essentially determined by the rate of recombination ., This effect appears to be much stronger in male than in female germline , which rules out selective explanations for the evolution of isochores ., We show that this impact of recombination is most probably a consequence of the process of biased gene conversion ( BGC ) ., This neutral process mimics the action of selection and can induce strong substitution hotspots within recombination hotspots , sometimes leading to the fixation of deleterious mutations ., BGC appears to be one of the major factors driving genome evolution ., It is therefore essential to take this process into account if we want to be able to interpret genome sequences . | genetics and genomics/comparative genomics, evolutionary biology/human evolution, computational biology/comparative sequence analysis, evolutionary biology/genomics, computational biology/evolutionary modeling, genetics and genomics/population genetics | null |
journal.pbio.1000446 | 2,010 | Foxp1 and Lhx1 Coordinate Motor Neuron Migration with Axon Trajectory Choice by Gating Reelin Signalling | Neural circuits are frequently organised in a topographic manner such that the position of a neuronal cell body is correlated with the location of the post-synaptic target and therefore its axon trajectory ., Since the inference of such organisational principles 1 , the molecular identity of many neuronal migration and axon guidance cues has been uncovered 2 , 3 ., Recent studies have also begun to identify the transcription factors that control neuronal identity and deploy the repertoire of neuronal migration and axon guidance receptors and signals employed in neural circuit assembly 4 , 5 , 6 ., These observations raise the possibility that correlated neuronal soma localisation and axon trajectory of topographically ordered neural circuits arise as a consequence of specific transcription factors directing both axon guidance and cell body migration effector expression ., Vertebrate spinal motor neurons are organised myotopically in longitudinal columns such that the location of their soma in the ventral spinal cord corresponds to the position of their muscle targets in the periphery 7 ., In mouse and chick , motor neurons innervating axial and body wall muscles are located in medially positioned columns , whereas motor neurons innervating limb muscles are located in the lateral motor column ( LMC ) present only at spinal cord levels in register with limbs ., LMC neurons are further subdivided according to their axon trajectory within the limb: lateral LMC ( LMCl ) neurons innervate dorsal limb muscles , whereas medial LMC ( LMCm ) neurons innervate ventral limb muscles 8 , 9 , 10 ., Motor pools are also organised myotopically such that , in general , the anterio-posterior location of a pool within the LMC correlates with the proximo-distal location of its limb muscle target 7 , 9 , 11 , 12 ., A motor axon guidance decision point is at the base of the limb where LMC axons interact with mesenchymal cells resulting in the selection of a dorsal or a ventral limb nerve trajectory 10 , 13 ., Concomitant with this process , LMC somata migrate from the progenitor-rich ventricular zone to the ventral horn of the spinal cord 14 , 15 , with the later-born LMCl neurons migrating past the earlier-born LMCm neurons in a manner reminiscent of the inside-out lamination of the developing cerebral cortex 16 , 17 , 18 ., Recent studies also describe a topographic relationship between motor neuron soma and dendrite localisation in Drosophila and the patterns of motor neuron recruitment during swimming in fish 19 , 20 ., The molecular signals controlling the trajectory of LMC axons are characterised , but those controlling LMC soma position in the spinal cord are poorly understood ., The LIM homeodomain proteins Isl1 and Lhx1 , expressed by LMCm and LMCl neurons respectively , act in conjunction with the pan-LMC forkhead domain transcription factor Foxp1 to specify the dorsoventral axon trajectory in the limb by regulating the expression of axonal Eph tyrosine kinase receptors that enable LMC growth cones to respond to ephrin ligands in the limb mesenchyme ., Genetic evidence argues that ephrin-A ligands in the ventral limb repulse EphA-expressing LMCl axons into the dorsal limb nerve , while ephrin-B ligands in the dorsal limb repulse EphB-expressing LMCm axons into the ventral limb nerve 21 , 22 , 23 , 24 , 25 , 26 ., The clustering of some motor pools relies on EphA4 , type II cadherins , and the ETS transcription factor Pea3 27 , 28 , 29 , while migration of LMCl and LMCm neurons into their appropriate columnar location can be biased by Lhx1 and Isl1 and requires Foxp1 21 , 22 , 23 ., These observations raise the possibility that Foxp1 , Lhx1 , and Isl1 control the migration of LMC cell bodies within the ventral horn by restricting the expression of specific effectors of neuronal migration ., The extracellular matrix protein Reelin is a crucial neuronal migration signal that acts through the lipoprotein receptors VLDLR or ApoER2 to induce the phosphorylation of the intracellular adaptor protein Dab1 leading to remodelling of the actin cytoskeleton 30 ., Loss of Reelin or its signalling effectors disrupts the layering of the neuronal somata within the cerebral cortex 31 , 32 , 33 but the role of Reelin in neuronal migration remains controversial ., Reelin has been proposed to act as a neuronal migration stop signal 34; however , since Reelin expression in the ventricular zone can partially rescue the pre-plate splitting defects in Reelin-deficient mice , Reelin has also been proposed to act as a permissive signal enabling neurons to interpret distinct migration cues 35 ., Similar to cortical neurons , spinal neuron progenitor clones migrate away from the ventricular zone in radial spoke-like trajectories 14 and the migration of preganglionic ( PG ) motor neurons and the layering of the dorsal horn laminae is controlled by Reelin 36 , 37 ., These studies raise the possibility that Reelin may also regulate the localisation of LMC neurons and is thus a general migration cue specifying the position of many different classes of spinal neurons including LMC motor neurons ., Using gain and loss of function experiments in chick and mouse , we provide evidence that Reelin directs LMC neuron migration but not the selection of limb axon trajectory ., We also show that Foxp1 and Lhx1 , the transcription factors specifying LMC axon trajectory choice , gate Reelin signalling through the restriction of Dab1 , a key signalling intermediate ., Thus , the same transcription factors are directing neuronal soma migration and axon trajectory selection revealing the molecular hierarchy controlling the establishment of a somatotopic map ., To explore the possibility that Reelin signalling might control LMC soma migration , we monitored the expression of Reelin , its receptors , and their adaptor protein Dab1 in mouse embryos between embryonic day of development ( e ) 11 . 5 and e12 . 5 and in chick embryos between Hamburger and Hamilton ( HH ) stages ( St ) 23 and 30 38 in limb-level spinal cord ., These stages correspond to the times at which LMCl neurons are migrating out of the ventricular zone and reach their final position lateral to LMCm neurons 17 , 22 ., We used the transcription factor Foxp1 as a pan-LMC marker and subdivided the LMC based on the presence of Isl1 and Lhx1 transcription factors 21 , 23 , 25 ., Reelin has previously been detected in the thoracic spinal cord adjacent to PG neurons 36 ., At limb levels Reelin is expressed from e10 . 5 ( Figure S1 ) and in e11 . 5 mouse embryos we observed Reelin expression in cells medio-dorsal to LMC neurons , and by e12 . 5 this domain expanded ventrally , resulting in a Reelin-rich band intercalated between the ventricular zone and the LMC ( Figure 1A–H ) ., We also observed a similar Reelin mRNA and protein distribution in chick embryos ( Figure S1 ) ., We next monitored the expression of Reelin receptors VLDLR and ApoER2 and their intracellular adaptor protein Dab1 in mouse and chick spinal cords ., In e11 . 5 mouse embryos at both limb levels , VLDLR protein and mRNA were apparently expressed in all LMC neurons ( Figure 1I–L; unpublished data ) ., However , VLDLR protein levels appeared higher in LMCl neurons relative to LMCm neurons ( Figure 1K ) ., By e12 . 5 VLDLR mRNA and protein levels appeared uniform throughout the LMC ( Figure 1M–P; unpublished data ) ., In chick embryos , VLDLR mRNA was present in apparently all lumbar LMC neurons at both HH St 24 and HH St 30 ( Figure S1 ) ., At the stages examined , ApoER2 mRNA was expressed in the ventricular zone adjacent to the floor plate of both mouse and chick embryos; however , its expression in LMC neurons was only apparent in mouse embryos ( Figure 1Q–T; Figure S1; unpublished data ) ., In mouse , Dab1 mRNA and protein were present throughout the LMC from e10 . 5 , at both limb levels; however , at later ages examined , an LMC subpopulation expressed Dab1 mRNA and protein at noticeably higher levels ( Figure 1U–AF; Figure S1 , Figure S4; unpublished data ) ., At e11 . 5 , this expression domain ( Dab1high ) was confined to the medio-ventral aspect of the LMC corresponding to Foxp1+Isl1− LMCl neurons while the low-level Dab1 expression domain ( Dab1low ) was confined to the dorsally positioned Isl1+Foxp1+ LMCm neurons ( Figure 1U–X ) ., By e12 . 5 , Dab1high and Dab1low LMC neurons were found in , respectively , lateral and medial aspect of the LMC , and corresponded to LMCl and LMCm neurons ( Figure 1Y–AB ) ., Similar Dab1 mRNA distribution was observed in chick embryos ( Figure S1 ) ., Together , our expression data raise the possibility that Reelin signalling directs LMC soma migration and the disparate Dab1 expression levels in LMCl and LMCm neurons suggest that these neuronal populations may differ in their responsiveness to Reelin ., To determine whether Reelin signalling influences LMC neuron migration , we examined the spinal cord of Dab1 and Reelin ( Reln ) mutant mice ( Figure, 2 ) 31 , 32 ., Since Reelin signalling is required for the appropriate positioning of PG neurons which share a part of their migration trajectory with LMC neurons 36 , 39 , we focused our analysis on caudal lumbar-sacral ( LS ) levels , which contain no PG neurons , as assessed by phospho-Smad1 expression 23 ., During LMC migration , the total number of LMC neurons , LMCl and LMCm subtype specification , and radial glia development was unaffected by Dab1 and Reln loss of function ( Figure S2 , Figure S3; unpublished data ) ., Additionally , most likely because of its impaired degradation 40 , Dab1 protein levels in LMC neurons were increased in Reln mutants , suggesting that all LMC neurons are responsive to Reelin ( Figure S4 ) ., We next analysed the localisation of lumbar LMC neurons in Dab1 and Reln mutants at e12 . 5 , the time at which , in control embryos , the majority of wild type LMCl neurons have terminated their migration and are positioned lateral to LMCm neurons ( Figure 2A–D ) ., In Dab1 mutants , LMCl neurons settled ventral to LMCm neurons , which were abnormally shifted to a lateral position in the ventral horn , and many LMCl and LMCm neurons were intermingled ( Figure 2E–H ) ., This neuronal displacement was more evident when we superimposed the position of LMCl and LMCm neurons in images of adjacent wild type ( wt ) and Dab1 mutant spinal cords sections ( Figure 2D , H ) ., To assess the expressivity of this phenotype and to account for LMC neuron displacement along mediolateral ( ML ) and dorsoventral ( DV ) axes simultaneously , we performed a two-dimensional position analysis of LMC neuron position using the bivariate statistical Hotellings T2 test ., We measured the mean ML and DV coordinates of wild type and Dab1 mutant LMC neurons within the ventral spinal cord ., To compensate for sectioning artefacts , we normalised the ML coordinates to the distance from the ventricular zone to the lateral edge of the Foxp1+ expression domain and the DV coordinates to the dorsoventral extent of the Foxp1+ expression domain , two standard measurements that are not different between Dab1 mutants and wild type littermates ( see Experimental procedures for details; unpublished data ) ., Thus , with the lateral-most edge of the LMC defined as ML: 100% , and with the dorsal-most domain of the LMC defined as DV: 100% , in wild type embryos , the mean position of LMCm neurons was not changed significantly by Dab1 mutation; however , these neurons were spread over a larger mediolateral zone compared to wild type littermates ( Figure 2I; Table S2 ) ., In contrast , by visual inspection of at least six spinal cord sections per embryo , we noted that in six out of six embryos analysed , LMCl neurons were positioned aberrantly ., Quantification revealed that LMCl neuron position was significantly shifted in a medio-ventral direction in Dab1 mutants relative to wild type littermates ( ( ML: 73%; DV: 33% ) versus ( ML: 79%; DV: 39% ) ; p<0 . 0035 , Hotellings T2 test; Table S2 ) , which could be observed at least until e15 . 5 ( Figure 2S–U , W–Y; unpublished data ) ., A similar LMC migration phenotype was also observed in the cervical spinal cord as well ( unpublished data ) , and in chick LMC neurons expressing a Dab1 protein in which the five tyrosines essential for Reelin signalling have been mutated ( Dab15YF; Figure S5 , Table S3; 41 ) ., We also noted that in four out of four embryos , the position within the ventral spinal cord of a Pea3-expressing motor neuron pool was shifted medio-ventrally at e15 . 5 ( Figure 2V , Z ) ., Together , these results demonstrate that in the limb-level spinal cord , Dab1 is essential for the normal migration of LMC neurons and motor pool position ., We next examined the position of lumbar LMC neurons in Reln mutant embryos at e12 . 5: Reln mutation did not alter the mean position of LMCm neurons ( Figure 2J–Q; Table S2 ) , although as in Dab1−/− embryos , these neurons were spread over a larger area of the LMC when compared to controls ( Figure 2R ) ., In contrast , in three out of four embryos , we observed that LMCl neurons were positioned abnormally , with quantification revealing that the mean LMCl neuron position in Reln mutants was significantly shifted in the medio-ventral direction relative to wild type , with many LMCl neurons found intermingled with LMCm neurons ( ( ML: 75%; DV: 35% ) versus ( ML: 80%; DV: 41% ) ; p<0 . 0473 , Hotellings T2 test; Figure 2J–R; Table S2 ) ., Migration defects observed in Reln mutants mirrored those observed in Dab1 mutants , thus implicating Reelin signalling in the specification of LMC soma position in the ventral spinal cord ., Based on the differential expression and the requirement for its function in LMCm and LMCl neurons , we reasoned that the levels of Dab1 expression , rather than simply its presence or absence , might influence the migration of LMC neurons ., We therefore asked whether increasing Dab1 expression would shift the position of LMC soma laterally ., To do this , we used in ovo electroporation to introduce a Dab1::GFP fusion protein or GFP expression plasmids into the lumbar spinal cord of HH St 17/19 embryos and monitored the position of GFP+ LMC neurons at HH St 29 22 ., Dab1::GFP was expressed with equal efficiency in LMCl and LMCm neurons and did not change their identity nor affect their axon trajectory in the limb ( Figure S6; unpublished data ) ., The mean position of LMCl neurons with elevated Dab1 levels was the same as that of LMCl neurons expressing GFP ( Figure 3A–G , I; Table S3 ) ., However , in four out of five embryos , we observed that LMCm neurons with elevated Dab1 expression were observed in a more ventro-lateral position ( Figure 3E–I; ( ML: 70%; DV: 49% ) ) compared to LMCm neurons expressing GFP ( Figure 3A–D , I; ( ML: 67%; DV: 59% ) , p\u200a=\u200a0 . 0165 , Hotellings T2 test; Table S3 ) , demonstrating that increasing Dab1 expression levels in LMC neurons is sufficient to shift their position laterally ., The myotopic relationship between LMC soma position and axon trajectory within the limb raises the possibility that changes in LMC soma position in Dab1 or Reln mutants could result in the selection of inappropriate limb trajectory by LMC axons ., To examine the LMCl axon limb trajectory in Dab1 mutants , we used the Lhx1tlz marker line 42 and quantified the proportion of LacZ+ LMCl axons projecting into e11 . 5 forelimb dorsal and ventral limb nerves in Dab1−/−; Lhx1tlz/+ , and Lhx1tlz/+ littermate embryos 24 ., In Lhx1tlz/+ embryos we observed ∼99% of LacZ+ axons within the dorsal limb nerves and ∼1% of LacZ+ axons within the ventral limb nerves ( Figure 4A , B , E ) ., The proportions of LacZ+ in dorsal and ventral limb nerves of littermate Dab1−/−; Lhx1tlz/+ embryos were not significantly different ( Figure 4C–E; 98% and 2% , respectively , p>0 . 5 , Students t test ) ., Additionally , in whole mount e12 . 5 Dab1−/−; Lhx1tlz/+ embryos , we did not detect any aberrantly projecting LMCl axons at either limb level ( unpublished data ) ., To trace LMCm axons we used the hcrest/Isl1-PLAP reporter line in which the Isl1 enhancer-promoter drives the expression of placental alkaline phosphatase ( PLAP ) in LMCm neurons at forelimb levels 43 ., PLAP enzymatic reaction was used to detect LMCm axons in Dab1−/−; hcrest/Isl1-PLAP+ and control hcrest/Isl1-PLAP+ e11 . 5 forelimbs , followed by axonal signal quantification ., In hcrest/Isl1-PLAP+ embryos , ∼99% of PLAP+ axons were found in the ventral limb nerve , while ∼1% of PLAP+ axons were found in the dorsal limb nerve ( Figure 4F , G , J ) , proportions not significantly different from Dab1−/−; hcrest/Isl1-PLAP+ embryos ( Figure 4H–J; 99% and 1% , respectively; p\u200a=\u200a0 . 335 , Students t test ) ., LMCm limb trajectory in Reln mutants was also apparently normal ( unpublished data ) , indicating that neither Dab1 nor Reelin are required for the selection of limb trajectory by LMC axons and demonstrating that the LMC soma position can be dissociated from axon trajectory selection ., Since our results indicated that the Dab1 protein level determines the position of LMC neuron somata but not their axon trajectory , we next evaluated whether the deployment of effector pathways governing these processes might be coordinated by a common set of transcriptional inputs ., To determine whether Foxp1 , a transcription factor specifying LMC cell fate , participates in the control of Dab1 expression in LMC neurons , we analyzed the embryonic spinal cords in which Foxp1 is expressed in all motor neurons ( Hb9::Foxp1 transgenic ) as well as in those lacking Foxp1 function 21 , 23 ., We first focused our analysis on upper cervical levels , where Foxp1 and Dab1 expression levels are normally low or undetectable ( Figure 5A–C; Figure S7; unpublished data ) ., In e12 . 5 Hb9::Foxp1+ spinal cords , compared to control embryos , we observed a significant increase in Dab1 mRNA levels ( 30 arbitrary ( arb . ) units versus 16 in controls; p\u200a=\u200a0 . 002 , Students t test; Figure 5A , C , D , F , M ) as well as protein expression levels associated with ectopic Foxp1+ neurons , without any obvious changes in Reelin expression ( Figure 5A , B , D , E , M; Figure S7; 30 arb . units versus 16 in controls; p<0 . 001 , Students t test ) ., To determine whether Foxp1 is required for Dab1 expression , we examined the lower cervical spinal cord of Foxp1 mutant mice at e12 . 5 ., When compared to controls , Foxp1 mutant spinal cords exhibited a significant decrease in Dab1 mRNA levels ( 15 arb . units versus 33 in control littermates; p<0 . 001 , Students t test; Figure 5G , I , J , L , M ) as well as Dab1 protein levels ( Figure 5G , H , J , K , M; Figure S7; 12 arb . units versus 37 in control littermates; p<0 . 001 , Students t test ) , demonstrating that Foxp1 is both sufficient and required for Dab1 expression in migrating LMC neurons ., Although Foxp1 controls Dab1 expression , because of its uniform expression throughout the LMC , it appeared to us an unlikely determinant of the differential level of Dab1 expression in LMCl and LMCm neurons ., LIM homeodomain proteins Isl1 and Lhx1 are determinants of , respectively , LMCm and LMCl neuronal fate , can influence their migration , and can control their axon trajectory by modulating Eph receptor expression ( Figure S8 and Text S1; 22 , 24 , 25 , 42 ) ., We thus hypothesized that while Foxp1 activates Dab1 expression in all LMC neurons , Isl1 and Lhx1 have opposing effects on Dab1: ( 1 ) Isl1 lowers Dab1 expression in LMCm neurons while ( 2 ) Lhx1 elevates Dab1 expression in LMCl neurons ., We tested the first of these hypotheses by electroporating Isl1 and LacZ expression plasmids , or a control LacZ expression plasmid alone into HH St 17/19 chick lumbar spinal cords and measuring changes in Dab1 mRNA levels relative to the unelectroporated control side at HH St 29 22 ., Expression of LacZ did not affect Isl1 or Dab1 mRNA expression while overexpression of Isl1 significantly reduced Dab1 mRNA expression levels in LMC neurons ( Figure S9; e/u values: 1 . 4 for LacZ versus 0 . 7 for Isl1 , p<0 . 001 , Students t test ) indicating that Isl1 can suppress Dab1 mRNA expression ., To test whether Isl1 is required to control Dab1 expression , we examined the effects of siRNAs directed against Isl1 in LMC neurons but observed no significant difference in Dab1 expression when compared to controls ( Figure S9 and Text S1 ) ., Together , these data suggest that Isl1 is sufficient but might be dispensable for the modulation of Dab1 expression in LMC neurons ., We next tested whether Lhx1 is required to specify the position of LMCl neurons by examining embryos with a conditional loss of Lhx1 function in LMC neurons , obtained by crossing Lhx1flox homozygotes with Isl1Cre/+; Lhx1tlz/+ mice , in which Isl1Cre drives Cre recombinase expression in all LMC neurons ., We focused our analysis on e12 . 5 lumbosacral levels in two groups of embryos obtained from these crosses: Lhx1tlz/flox; Isl1Cre/+ , designated as Lhx1COND , and control Lhx1tlz/+ , designated as Lhx1+/− ., Lhx1 loss of function did not affect the total number of LMC or LMCm neurons but resulted in ∼60% of LMCl neurons ( Foxp1+Isl1− ) losing their Lhx1 expression ( Isl1−Lhx1/5+Foxp1+: 37 . 3% versus 95 . 2% in controls; p<0 . 001 , Students t test , Figure 6I , unpublished data ) ., We determined the soma position of three LMC neuronal populations: LMCm , LMCl , and LMCl neurons lacking Lhx1 expression , which were defined as Isl1−Foxp1+Lhx1/5− ( LMCl* ) ., As in control embryos , in which the majority of LMCl neurons settled in the most lateral part of the LMC , in Lhx1COND embryos , a significant proportion of LMCl* neurons settled laterally and the mean position of LMCm , LMCl , or LMCl* neurons was not changed when compared to controls ( Figure 6A–J; Table S4 ) ., However , in Lhx1COND embryos , many LMCl* neurons were found in medial locations , intermingled with LMCm neurons ( Figure 6A–H ) , and these neuronal displacements were more evident when we superimposed the positions of LMCl* , LMCl , and LMCm neurons in images of adjacent control and Lhx1COND spinal cords sections ( Figure S10 ) ., To further characterise the medially displaced population of LMCl* neurons , we counted the number of LMC neurons in four equal quadrants of the LMC ( Figure 6J , K , unpublished data ) ., In both Lhx1 mutant and control embryos the majority of LMCm neurons were in the medial half of the LMC ( unpublished data ) ., In control embryos , 60% of LMCl neurons were in the lateral half of the LMC , compared to 42% of LMCl* neurons in Lhx1 mutants , representing a significant change ( p\u200a=\u200a0 . 003 , Students t test , Figure 6K ) , indicating that Lhx1 is required for LMCl position specification ., To determine whether Lhx1 directs LMCl migration by controlling Dab1 expression , we compared Dab1 protein levels in the lumbar spinal cord of e12 . 5 Lhx1 mutants in which at least 50% of LMCl neurons lost their Lhx1 expression and littermate controls 22 ., Our analysis revealed that in Lhx1 mutants , Dab1 protein expression in LMC neurons was decreased by ∼20% when compared to control embryos ( Figure 7A–H , O; p\u200a=\u200a0 . 038 , Students t test ) ., We also quantified Dab1 mRNA and protein levels in the LMCm , defined as containing >90% of Isl1+Foxp1+ neurons and LMCl defined as Isl1−Foxp1+ ., Within the LMCm , Dab1 mRNA and protein levels were not significantly different from controls , while in LMCl of Lhx1 mutants , relative to controls , Dab1 mRNA was decreased significantly by approximately 40% ( p\u200a=\u200a0 . 01 , Students t test ) and Dab1 protein was decreased significantly by ∼14% ( p\u200a=\u200a0 . 017 , Students t test , Figure 7O ) , indicating that Lhx1 is required for the differential expression of Dab1 in LMC neurons ., Together , our results reveal that Foxp1 and Lhx1 coordinate LMC myotopy through their modulation of expression of neuronal migration and axon guidance effectors ., Following their birth near the ventricular zone , spinal neurons first migrate radially by perikaryal translocation , then tangentially , either in dorsal or ventral direction 14 ., Reelin has been proposed as a radial migration signal; however , our observations argue that the initial , apparently radial trajectory of LMC motor neurons is Reelin signalling independent as is the case of PG and hindbrain motor neurons 36 , 39 ., Thus , in general , the radial migration trajectory of motor neurons might not require Reelin signalling , but once it is terminated , Reelin becomes an important guidance signal , suggesting that unlike cortical neurons that rely on Reelin for their localisation in the radial plane , motor neurons at different rostrocaudal levels of the spinal cord depend on Reelin for the tangential aspect of their migration ., How does Reelin act in motor neuron migration ?, The initial model where Reelin is a migration stop signal has been challenged by observations that Reelin overexpression in the cortical ventricular zone can rescue , at least in part , pre-plate splitting defects associated with Reelin loss of function 34 , 35 ., Likewise , overexpression of Reelin in the ventricular zone of the spinal cord rescues Reln mutant PG neuron migration defects but does not cause an overt phenotype in a wild type background 44 ., In the context of LMC neurons , the Reelin expression domain is intercalated between the emerging postmitotic neurons and their final lateral position , thus precluding a function as a migration stop signal , unless at the time of their early migration LMC motor neurons are insensitive to Reelin ., Our functional Reelin fragment overexpression in the ventral spinal cord resulted in LMCl motor neurons moving beyond their normal lateral position ( E . P . , T . -J . K . , and A . K . , unpublished observations ) ; thus , in the context of motor neurons , Reelin is unlikely to function as a migration stop signal , rather , it likely promotes migration or enables LMC neurons to respond to a cue that provides spatial information ., What is the relationship of the Reelin-mediated LMC position specification to that mediated by cadherins , Eph receptors , and the transcription factor Pea3 27 , 28 , 29 ?, Because of their restricted expression patterns and functional analysis phenotypes , these are thought to operate at the level of motor pools , in contrast to Reelin signalling which appears to specify the position of the entire LMCl division ., Cadherins have been shown to be involved in the clustering of specific motor pools via their combinatorial expression imparting different adhesion properties on specific motor pools ., Similarly , although the early migration of LMC motor neurons in EphA4 mutants appears to be normal , eventually the position of the tibialis motor pool is shifted ., Because of these observations , it is likely that Cadherins , EphA4 , and Pea3 act at a step following Reelin-mediated migration of LMCl neurons ., Unfortunately , since ETS genes , arguably the earliest molecular markers of motor pools , begin to be expressed at the time when LMCl somata attain their lateral position 45 , it is technically difficult to ascertain experimentally whether motor pool clustering precedes or coincides with LMCl lateral migration ., The differences between the LMC position phenotypes in Dab1 and Lhx1COND mutants might shed some light on this hierarchy ., In Dab1 mutants , although shifted medio-ventrally , LMCl neurons remain clustered , in contrast to Lhx1 mutant LMCl motor neurons that can be found intermingled with LMCm neurons ., These observations suggest that while the Dab1 mutation probably only leads to the absence of sensitivity to Reelin , the loss of the transcription factor Lhx1 might have consequences beyond the loss of Dab1 , resulting , for example , in a change in expression of cell surface adhesion molecules allowing LMCl and LMCm neurons to intermingle ., Our findings demonstrate that migration of LMC neurons within the ventral spinal cord requires Reelin signalling through the intracellular adaptor protein Dab1 ., This requirement is principally evident in LMCl neurons and corresponds to the high level of Dab1 protein and mRNA expressed in this population when compared to LMCm neurons ., Other studies have also implicated Dab1 protein levels controlled by Cullin5 and Notch signalling as a determinant of neuronal migration 46 , 47 , raising the question of how might differential Dab1 expression specify LMC soma position in the ventral spinal cord ., Upon activation of the Reelin pathway , Dab1 is phosphorylated and rapidly degraded 30 , 34 ., Therefore , in the presence of Reelin , the low Dab1 protein levels in LMCm neurons might be depleted faster than the higher Dab1 protein levels in LMCl neurons , resulting in the termination of Reelin signalling and thus a migration stop occurring sooner in LMCm neurons than in LMCl neurons ., This mode of Dab1 function assumes that Reelin promotes migration of LMC neurons , or is a factor enabling their reception of a migration cue and is consistent with our observation that both LMCl and LMCm neurons can respond to Reelin ., Thus similar to the Toll-like receptor ( TLR ) 48 and chemokine 49 signalling pathways regulated by the level of expression of a signalling intermediate , Reelin signal is differentially gated in two neuronal populations through opposing levels of Dab1 expression ., In such a model , we would favour the idea that Dab1 concentration , in the presence of Reelin , is an instructive determinant of LMC neuron position , although the formal demonstration of this through , for example , the change of LMCm Dab1 levels to match exactly those in LMCl neurons is technically challenging ., Following its phosphorylation , Dab1 is targeted for polyubiquitination and degradation by Cullin5 47 , raising the possibility that in LMC neurons , Dab1 protein stability might contribute to the differences in Dab1 protein in LMC neurons ., However , since in LMC neurons Cullin5 is apparently expressed at equal levels by LMCl and LMCm neurons ( E . P . and A . K . , unpublished observations ) , and because of the selective enrichment of Dab1 mRNA in LMCl neurons , compared to LMCm neurons , we favour the hypothesis that differential transcriptional regulation of the Dab1 gene or its mRNA stability is an important factor contributing to Dab1 protein levels in LMC neurons ., Our results demonstrate that Dab1 expression levels in LMC neurons are set by Foxp1 and Lhx1 , two transcription factors that are essential for the specification of LMC soma position 21 , 22 , 23 ., Our data suggest the following model of Dab1 expression control in LMC neurons: a basal level of Dab1 expression in LMC neurons is induced or maintained by Foxp1 , while Lhx1 , a transcription factor selectively expressed in LMCl neurons , could act to elevate Dab1 expression in LMCl neurons ., Additionally , based on its ability to suppress Lhx1 22 and Dab1 mRNA expression in LMC neurons , Isl1 might function to diminish Dab1 expression in LMCm neurons ., Thus , although we cannot exclude the influence of other transcription factors or distinguish whether the control of Dab1 expression by Foxp1 and Lhx1 occurs at the level of the Dab1 promoter , through intermediary transcription factors or regulation of Dab1 mRNA stability , we propose that the concerted action of Foxp1 and Lhx1 leads to differential Dab1 expression levels in LMC neurons ., Could transcription factor control of Dab1 expression be a general mechanism gating Reelin signalling in the CNS ?, In the cortex , examples of control of migration effectors by transcription factors include the coupling of neurogenesis to migration by bHLH control of doublecortin and p35 , Tbx20 control of the planar cell-polarity pathway , and Nkx2 . 1 control of Neuropilin2 expression 6 , but to our knowledge , a general link between a specific transcription factor and Dab1 expression has so far only been established for CREB/CREM 50 ., Intriguingly , in the sp | Introduction, Results, Discussion, Materials and Methods | Topographic neuronal maps arise as a consequence of axon trajectory choice correlated with the localisation of neuronal soma , but the identity of the pathways coordinating these processes is unknown ., We addressed this question in the context of the myotopic map formed by limb muscles innervated by spinal lateral motor column ( LMC ) motor axons where the Eph receptor signals specifying growth cone trajectory are restricted by Foxp1 and Lhx1 transcription factors ., We show that the localisation of LMC neuron cell bodies can be dissociated from axon trajectory choice by either the loss or gain of function of the Reelin signalling pathway ., The response of LMC motor neurons to Reelin is gated by Foxp1- and Lhx1-mediated regulation of expression of the critical Reelin signalling intermediate Dab1 ., Together , these observations point to identical transcription factors that control motor axon guidance and soma migration and reveal the molecular hierarchy of myotopic organisation . | Many areas of our nervous system are organized in a topographic manner , such that the location of a neuron relative to its neighbors is often spatially correlated with its axonal trajectory and therefore target identity ., In this study , we focus on the spinal myotopic map , which is characterized by the stereotyped organization of motor neuron cell bodies that is correlated with the trajectory of their axons to limb muscles ., An open question for how this map forms is the identity of the molecules that coordinate the expression of effectors of neuronal migration and axonal guidance ., Here , we first show that Dab1 , a key protein that relays signals directing neuronal migration , is expressed at different concentrations in specific populations of limb-innervating motor neurons and determines the position of their cell bodies in the spinal cord ., We then demonstrate that Foxp1 and Lhx1 , the same transcription factors that regulate the expression of receptors for motor axon guidance signals , also modulate Dab1 expression ., The significance of our findings is that we identify a molecular hierarchy linking effectors of both neuronal migration and axonal projections , and therefore coordinating neuronal soma position with choice of axon trajectory ., In general , our findings provide a framework in which to address the general question of how the nervous system is organized . | neuroscience/motor systems, neuroscience, neuroscience/neurodevelopment, neuroscience/neuronal signaling mechanisms | During embryonic development of the vertebrate motor system, the same transcription factors that regulate axonal trajectories can also regulate cell body migration, thereby controlling topographic map formation. |
journal.ppat.1002912 | 2,012 | Molecular Basis for Nucleotide Conservation at the Ends of the Dengue Virus Genome | Most RNA viruses maintain the specific sequences present at the ends of their genomes ., The 5′ genome end may carry a cap structure to ensure both genome stability and efficient translation 1 ., The 3′-end may carry a poly ( A ) tail or adopt specific 3′-end sequences required for viral replication 2 , 3 ., They are generally copied exactly to avoid loss of genetic information , and have supposedly evolved towards optimum replication efficiency ., Terminal genome damage can be caused by errors introduced by the viral polymerase during initiation and termination , or by cellular ribonucleases 4 ., In addition to special mechanisms to ensure efficient initiation of RNA synthesis , viruses have evolved mechanisms to repair or correct damaged extremities such as the use of abortive transcripts as primers , the generation and use of non-templated primers , and the addition of one or few non-templated nucleotides to the 3′-end by a terminal transferase activity 4 ., However , our knowledge about these mechanisms is still very limited ., Many RNA virus polymerases , which do not use a primer and thus initiate RNA synthesis de novo , generate abortive transcripts during the initiation phase of RNA synthesis 5 , 6 , 7 ., Primer-mediated repair of template extremities was so far only demonstrated for the positive-strand RNA ( +RNA ) turnip crinkle virus ( TCV ) 8 ., Non-templated primer synthesis by the viral polymerase might be involved in the repair mechanism of TCV 9 ., Such mechanism was also proposed as the molecular basis of the reconstitution of 5′-ends of negative-strand RNA ( -RNA ) respiratory syncytial virus ( RSV ) replicons 10 ., In this study we demonstrate how the dengue virus ( DV ) RNA-dependent RNA polymerase ( RdRp ) , which starts RNA synthesis de novo , plays a decisive role in the nucleotide conservation of viral RNA ends ., DV belongs to the Flavivirus genus within the +RNA virus family of Flaviviridae together with viruses of the genera Hepacivirus and Pestivirus 11 ., The Flavivirus genus comprises around 50 virus species 12 including major human pathogens such as DV , yellow fever virus ( YFV ) , West Nile virus ( WNV ) and Japanese encephalitis virus ( JEV ) ., Flaviviruses harbour the RdRp activity in the C-terminal domain ( amino acids 272–900 ) of non-structural protein NS5 13 , 14 , 15 , 16 , 17 ., The N-terminal domain contains methyltransferase activities involved in RNA capping 18 , 19 ., Evidence has been presented that the N-terminal domain of NS5 also harbours the central RNA capping guanylyltransferase activity 20 ., The structure of full-length NS5 is not known but several structures of methyltransferase domains have been determined ( for review see 21 ) ., Likewise , crystal structures of Flavivirus NS5 RdRp domains have been determined for DV 16 and WNV 22 ., All structurally characterized viral RdRps so far adopt the basic fold of the SCOP superfamily of DNA/RNA polymerases ., As the other subgroups of this superfamily , DNA-dependent DNA polymerases ( DdDp , prototype Klenow fragment of the E . coli DdDp I ) , RNA-dependent DNA polymerase ( prototype HIV reverse transcriptase ) and DNA-dependent RNA polymerases ( DdRp , prototype bacteriophage T7 DdRp ) , their apo-structure is usually likened to a right hand comprising fingers , palm and thumb subdomains ., Viral RdRps contain an encircled active site having connecting elements between the fingers and thumb subdomains ., Active sites of viral RdRps performing de novo RNA synthesis are additionally closed in their initiation conformation due to the existence of structural elements allowing the stable positioning of the first NTP into a priming site 23 , 24 ., All Flaviviridae RdRps studied so far initiate RNA synthesis de novo ., Accordingly , Flavivirus RdRp domain structures contain a “priming loop” in the thumb subdomain closing the catalytic site 16 , 22 ., The putative priming loop of DV RdRp was defined as comprising residues 792 to 804 ., Of particular interest are two aromatic residues near the tip of the loop , W795 and H798 , which are conserved in all Flavivirus RdRps ., They might play the role of an initiation platform to which the base of the priming NTP stacks as it was shown for bacteriophage φ6 23 and proposed for HCV and BVDV RdRps 25 , 26 ., Structures of DV RdRp in complex with 3′dGTP as well as two models of de novo initiation complexes of DV and WNV RdRps favor Trp795 in the role of the initiation platform 16 , 22 ., Genomes of Flaviviridae lack a poly ( A ) tail at the 3′-end ., A remarkable trait of Flavivirus genomes is the strict conservation of the 5′- and 3′-end dinucleotides as 5′ AG…CU 3′ ., The molecular basis for this strict conservation of the 5′- and 3′-end dinucleotides and/or the use of the same starting nucleotide for +RNA and -RNA strand synthesis by the viral polymerases is not known ., Its Hepacivirus and Pestivirus counterparts have to display higher nucleotide tolerance ., They are able to initiate with ( A/G ) C and G ( G/U ) , respectively , since the 5′- and 3′-ends of Hepacivirus genomes of different genotypes correspond to 5′ ( A/G ) C…GU 3′ and the genomes of pestiviruses to 5′ GU…CC 3′ ., Interestingly , genomes and antigenomes of non-segmented -RNA ( ns-RNA ) paramyxoviruses , whose RdRps perform de novo RNA synthesis , start with a conserved 5′-AC 10 ., Here we show that the strict sequence conservation of Flavivirus genome ends is entirely polymerase-encoded ., We demonstrate ATP-specific de novo initiation using the RdRp domain of DV protein NS5 ( NS5PolDV ) and specific 10-mer oligonucleotidic RNA templates corresponding to the 3′-end of genomic +RNA and -RNA ., We document the existence of a built-in ATP-specific priming site of NS5PolDV ., This specific site is one of the means by which NS5PolDV ensures that the DV genome and antigenome start with an A , the others being several correction mechanisms including the generation of non-templated pppAG primers as well as the preferential formation and elongation of pppAG even on templates with non-cognate 3′-ends ., Finally , we show that the ATP-specific priming site is part of the putative priming loop coming from the thumb subdomain ., There , residue H798 , and not W795 , is essential for de novo initiation and may act as a priming platform stabilizing the ATP priming nucleotide ., DV RdRp is actively involved in the conservation of the correct ends of the genome proving thus a direct example of how RNA viruses maintain the integrity of their genomes ., The mechanisms described here may more broadly apply to other RNA viruses having viral RdRps able to initiate RNA synthesis de novo ., We set out to study primer synthesis by the RdRp domain of dengue virus protein NS5 ( NS5PolDV ) using small specific templates corresponding to the 3′-ends of the genome ( +RNA ) and the antigenome ( -RNA ) ., Templates are comprised of 10 nucleotides and are predicted to be devoid of stable secondary structure ( see Materials and Methods ) ., Both templates end with the dinucleotide 5′-CU-3′ ., Product formation over time was followed using either ATP and GTP , or all NTPs needed to form a full-length product when synthesis is precisely started at the 3′-end of the template ., Figure 1 shows reaction kinetics of RNA synthesis on DV103′+ corresponding to the 3′-end of the RNA genome 5′-AACAGGUUCU-3′ ( left ) and on DV103′- corresponding to that of the antigenome 5′-ACUAACAACU-3′ ( right ) ., We used either α-32P-GTP ( αGTP , panel A ) or γ-32P-ATP ( γATP , panel B ) as the radioactive nucleotide ., For the catalytic ion , either Mg2+ ( panel A ) or Mg2+ supplemented with Mn2+ ( panel B ) were used at their optimum concentrations 5 mM for Mg2+ and 2 mM for Mn2+ 14 ., Reactions with ATP and GTP render time-dependent accumulation of a short product migrating below the marker G2 ( see panel B ) ., Comparison with authentic unlabeled pppAG ( see Materials and Methods ) visualized using UV-shadowing indicated that it indeed corresponds to pppAG ( not shown ) , the expected product of the first step of de novo RNA synthesis ., When DV103′+ is used as a template , pppAG is formed as well as pppAGA and pppAGAA ., When all NTPs are used , pppAG accumulates with time as does pppAGA in the case of DV103′+ and pppAGU in the case of DV103′- ., After the synthesis of trinucleotides NSPolDV adopts a processive RNA synthesis elongation mode to continue synthesis up to full-length products ( labeled by asterisks in Figure 1 ) ., As we had observed before 14 , when using Mn2+ the reaction is much more efficient and allows for the use of γ-32P-ATP ( γATP ) as radiolabeled nucleotide in order to visualize exclusively de novo RNA synthesis products starting with ATP ., The pattern observed with Mg2+ is reproduced when Mn2+ is present ( Figure 1B ) ., One difference is that the use of Mn2+ results in longer full-length products , which might be caused by an alteration of the terminal nucleotide transferase activity of NS5PolDV 14 , 27 , 28 ., In conclusion , using RNA templates mimicking viral sequences , dinucleotide and trinucleotide products are formed during initiation and before processive RNA elongation , the most abundant being the dinucleotide pppAG ., The first nucleotide of Flavivirus genomes is an adenosine , followed by a guanosine ., This 5′-pppAG sequence is strictly conserved along the Flavivirus genus ., In order to answer the question whether the polymerase ( and/or the correct template ) is at the origin of the conservation of the first nucleotide , we tested a set of DV103′- variants with different 3′-ends ., In addition to the correct DV103′- CU , we used DV103′- CC , DV103′- CA and DV103′- CG in the presence of the corresponding priming NTP and GTP ., The expected primer products are pppAG , pppGG , pppUG and pppCG , respectively ., Figure 2A compares end points of reactions performed in the presence of αGTP and Mg2+ as the catalytic ion ., Remarkably , the CU template only is proficient for product synthesis ( pppAG ) ., RNA primer synthesis on other templates is almost undetectable ., We conclude that in the presence of Mg2+ as a catalytic ion the DV RdRp priming-site accommodates exclusively ATP ., To our surprise , when Mn2+ was used instead of Mg2+ , the pppAG primer was generated even in the absence of the template , albeit to a lower extent ( Figure 2B ) ., This is not the case in the presence of Mg2+ even at ten-fold higher enzyme concentration ( see below Figure 3B ) ., When using Mn2+ and the DV103′- template variants , we therefore included control reactions in the absence of corresponding templates and in the presence of γGTP , which allows exclusive detection of dinucleotides starting with pppG ., Figure 2C shows corresponding reaction kinetics with Mn2+ as the catalytic ion in the absence or the presence of templates using αGTP or γGTP as the radioactive nucleotide ., Again , using DV103′-CU and ATP/GTP , NS5PolDV generates pppAG to a higher extent than without template ., Note that no pppGA product is generated ., When DV103′-CC and GTP is used , NS5PolDV synthesizes pppGG in the presence of the template only ., DV103′-CA , UTP , and GTP lead to the formation of pppUG and pppGU ( see γGTP control reaction ) , the latter by initiation internal to the template ., No product is formed in the absence of the template ., Finally , DV103′-CG allows formation of pppCG which is not formed in the absence of the template ., In conclusion , NS5PolDV keeps the strict preference for an ATP as the priming nucleotide in the presence of Mn2+ when no template is present ., Nevertheless , the use of templates with an altered 3′-nucleotide can force NS5PolDV to start the de novo RNA synthesis with the corresponding base-paired priming nucleotide , and also allows internal initiation ., Collectively , these observations confirm that the priming site of NS5PolDV has a marked specificity for ATP ., This preference is strict in the presence of Mg2+ ., It is equally strict for dinucleotide synthesis in the presence of Mn2+ and in the absence of template ., The specificity for ATP as the starting nucleotide is lost when Mn2+ is used in the presence of templates with incorrect 3′-ends; only then NS5PolDV is able to form pppNG products as efficiently as pppAG ., In the presence of Mg2+ and/or Mn2+ the built-in ATP-specific priming site drives NS5PolDV-mediated RNA synthesis starting with pppA ., The dinucleotide pppAG is accumulated during RNA synthesis on templates with the correct 3′-end ( see Figure 1 ) ., Using Mn2+ this pppAG primer is also formed in the absence of an RNA template ., We asked the question whether NS5PolDV forms and/or elongates pppAG even on templates with incorrect 3′-nucleotides thus enabling to repair incorrect 3′-ends ., First , pppAG formation was tested on the four DV103′- variants in the presence of only ATP and GTP ., Figure 3A shows that NS5PolDV is indeed able to form pppAG in the presence of templates with any 3′-nucleotide and Mn2+ ., In contrast , in the presence of Mg2+ only the natural DV103′- CU template supports pppAG formation even in the presence of an increased concentration of NS5PolDV ( Figure 3B ) ., We then tested pppAG formation exclusively in the presence of Mn2+ on all DV103′- variants in the presence of all nucleotides , a scenario putatively mimicking the situation within the replication complex ., Figure 3C shows that pppAG is always formed in parallel to the dinucleotide , which corresponds to the template ., In the case of the template variant with a -CG 3′-end , pppAG is produced with even higher efficiency than the base-paired dinucleotide ., Note that the dinucleotide pppGU is also produced on all templates by internal initiation ., For the reaction in the presence of all templates and all nucleotides , we quantified all products , which were initiated de novo over the very 3′-end , and found that pppAG is formed as the prominent product ( 32 . 3±1 . 5% , three independent reactions ) ., Note that all templates are present at the same concentration , which should not correspond to the situation in vivo ., We conclude that in the presence of incorrect templates and Mg2+ , NS5PolDV discriminates against these templates and forms pppAG only on the correct template ( see also Figure 2A ) ., In contrast , Mn2+ ions enable NS5PolDV to preferentially generate pppAG even in the presence of incorrect templates , which could represent an indirect way of 3′-end repair ., We then considered the elongation of the correct pppAG primer over templates with incorrect 3′-ends ., We thus tested the elongation of a chemically synthesized pppAG primer ( see Materials and Methods ) either without template or in the presence of the four DV103′- variants ( Figure 4 ) ., The most prominent result is that NS5PolDV is able to productively elongate pppAG on the correct template in the presence of Mn2+ ( Figure 4A ) and Mg2+ ions ( Figure 4B ) ., We also observe that NS5PolDV in the presence of Mn2+ is able to productively elongate pppAG on incorrect templates ( Figure 4A ) , thus demonstrating that the enzyme is able to indirectly correct the error in the template and conserve the 5′-end of the DV genome ., Note that as expected there is no primer elongation detectable in the absence of a template ., NS5PolDV harbors an ATP-specific priming site , which is essential for the formation , accumulation , and elongation of the correct primer pppAG ., Which elements of NS5PolDV form this site ?, The crystal structure of NS5PolDV ( Figure 5A ) allowed the prediction of a priming loop comprising residues 792 to 804 16 , which is expected to provide the priming site during de novo RNA synthesis initiation ., We generated a deletion mutant ( NS5PolDV TGGK ) by replacing residues T794-A799 between T793 and K800 by two glycines ( see close-up in Figure 5A ) ., The overall correct folding of the purified , recombinant mutant protein was verified by a fluorescent thermal shift assay giving identical temperatures of denaturation ( melting temperature Tm ) for both proteins ( wild type ( wt ) NS5PolDV Tm 49 . 0°C ± 0 . 5°C , NS5PolDV TGGK Tm 48 . 4°C ± 0 . 05°C ) ., The TGGK mutant is expected to have an open active site , which impedes correct ATP-specific de novo initiation over the 3′-end of a single-stranded RNA template but may favor the accommodation of double-stranded RNA ., Its RNA synthesis initiation and elongation activity was first tested using a “minigenomic” RNA template consisting of 224 nucleotides of the 5′-end of the DV genome fused to 492 nucleotides of the 3′-end 14 ., It has been shown before using this template and analyzing the products on a denaturing agarose-formaldehyde gel 29 that two types of product are formed ( see wt reaction kinetics in the center panel of Figure 5B ) ., Firstly , the de novo RNA synthesis product is generated corresponding to the size of the template ., Secondly , an elongation product is generated by back-primed RNA synthesis ., There , the 3′-end ( …AACAGGUUCU-3′ ) forms a short hairpin annealing the last di-nucleotide to nucleotides -6 and -7 ( underlined in the sequence ) and is then elongated 29 ., The length of the product is thus ∼twice the size of the template ., Reactions were carried out using either Mg2+ or Mn2+ as catalytic ions ., The left and right panels of Figure 5B show that in both cases the mutant TGGK shows an increased overall activity on this template compared to wt activity ., The center panel shows that this is mainly caused by increased back-priming ., Interestingly , instead of one product species of twice the template size NS5PolDV TGGK produces a range of elongated products of different lengths ., This might be due to the accommodation of long hairpins , which then create longer products than the template but shorter than the elongation product of wt NS5PolDV ., De novo RNA synthesis initiation by wt NS5PolDV and the TGGK mutant were then tested on DV103′- , in the absence of a template and on DV103′+ using Mn2+ as the catalytic ion , ATP and GTP containing αGTP ., Figure 5C ( panel 1 ) shows that in contrast to wt NS5PolDV , NS5PolDV TGGK is not able to catalyze de novo initiation on DV103′- ., Secondly , NS5PolDV TGGK does not catalyze pppAG formation without template ( panel 2 ) ., In contrast , it is able to catalyze de novo initiation on DV103′+ presenting ca ., 32% of wt activity ( panel 3 ) ., In order to understand this apparent contradiction , we used γATP instead of αGTP as radioactive NTP ., It became clear that NS5PolDV TGGK was unable to generate the pppAG primer product ( panel 4 ) ., We conclude that the product observed with αGTP corresponds to pppGA formed by internal de novo initiation being only possible on DV103′+ ., When using Mg2+ as catalytic ion again we did not observe formation of the de novo RNA synthesis initiation product pppAG on either template ( for DV103′- see below Figure 6B ) ., We conclude that NS5PolDV TGGK is unable to pre-form the ATP-specific priming site necessary for de novo RNA synthesis initiation at the very 3′-end ., The predicted priming loop plays indeed an essential role in providing the correct priming site ., We explain the increased activity of NS5PolDV TGGK on minigenomic RNA templates by its increased propensity to catalyze back-priming due its more accessible catalytic site , i . e . to harbor the minigenome in different hairpin conformations allowing 3′ elongation ., Two aromatic residues , W795 and H798 , within the priming loop were proposed to play a particular role in providing an initiation platform to which the base of the priming ATP could establish a stacking interaction 16 ., Residue W795 was given special attention because it was found near the triphosphate moiety of a 3′-dGTP bound to NS5PolDV 16 ., In addition , this tryptophan was better placed than the histidine for stacking a priming ATP in two models of de novo RNA synthesis initiation complexes of NS5PolDV and NS5PolWNV 16 , 22 ., We generated two mutants of NS5PolDV , W795A and H798A ., Overall correct folding of the purified recombinant mutants was equally verified by a fluorescent thermal shift assay giving Tm values corresponding to the wt protein ( wt NS5PolDV Tm 49 . 0°C ± 0 . 5°C , W795A mutant Tm 48 . 6°C ± 0 . 6°C , H798A mutant Tm 48 . 1°C ± 0 . 04°C ) ., The RNA initiation and elongation activities of wt NS5PolDV and the W795A and H798A mutants were tested using the minigenomic RNA template and either Mg2+ or Mn2+ as catalytic ions ( Figure 6A ) ., In both cases the H798A mutant shows an increased activity on this template whereas W795A shows a similar overall activity compared to wt NS5PolDV ., Figure 6B shows the analysis of the reaction products on a denaturing agarose-formaldehyde gel ., The W795A mutant behaves indeed like wt NS5PolDV , the percentage of the de novo RNA synthesis initiation product of template size is unchanged ., In contrast the H798A mutant generates considerably less de novo RNA synthesis product whereas the yield of RNA elongation products is higher ., We then compared the capacities of wt and all mutant NS5PolDV proteins to catalyze de novo RNA synthesis initiation on DV103′- , without template and on DV103′+ using Mn2+ as catalytic ion ( Figure 6C panels 1 , 3 and 4 ) ., Indeed , the H798A mutant is considerably less capable of correct de novo RNA synthesis initiation than wt NS5PolDV whereas W795A behaves as wt NS5PolDV ., Note that the product formed by NS5PolDV TGGK on DV103′+ ( panel 4 ) corresponds to pppGA generated by internal RNA synthesis initiation ( see also Figure 5C ) ; and therefore part of the product formed by the H798A mutant may correspond to pppGA ., When Mg2+ is used on both templates , the same results are obtained ( Figure 6C panel 2 for template DV103′- ) ., We thus conclude that residue H798 is essential for the formation of the correct ATP-specific priming site and may act as a priming platform ., In this study , we present evidence that the dengue virus NS5 polymerase domain ( NS5PolDV ) alone is responsible for maintenance of A and U as first and last nucleotides of the DV genome , respectively ., NS5PolDV was used instead of full-length NS5 in the frame of this study in order to avoid any interference of the RNA-binding , NTP-binding , or enzymatic activities of the N-terminal domain of NS5 ., We report that NS5PolDV is endowed with several structural and mechanistic features converging to the specific de novo synthesis and elongation of the correct ATP-initiated primer even on templates that lack the correct corresponding U at the 3′-end ., The first and last nucleotides of the genome are strictly conserved in the genus Flavivirus thus the results presented here may apply to the entire genus ., We demonstrate the generation of a dinucleotide primer pppAG on both genomic and antigenomic RNA templates ., We have previously observed the production of such dinucleotide primer on homopolymeric templates 14 ., In the following step pppAG ( A/U ) trinucleotides are formed before processive RNA elongation occurs ., During the latter , NS5PolDV continues RNA synthesis to the very end of the template ., We do not know if di- and tri-nucleotide primers as detected in the reaction , originate from a slow but processive RNA synthesis reaction , or are actually released from the complex and re-used by the polymerase acting in a distributive RNA synthesis mode ., We also show that the pppAG primer is effectively elongated in the presence of Mg2+ or Mn2+ and the correct template ., Thus , after initial phosphodiester bond synthesis , the pppAG primer is aligned at the correct position in order to be elongated ., The efficient use of the short primer pppAG reported here is in apparent contrast to the inefficient use of 5′-OH-AG dinucleotide previously reported 13 , 30 ., The 5′-triphosphate moiety of the chemically synthesized pppAG primer is most probably an important binding determinant allowing efficient elongation ( see discussion of the proposed de novo initiation complex Figure 7 ) ., We then demonstrate that in its de novo RNA synthesis initiation state NS5PolDV contains a built-in ATP-specific priming site ., Major structural elements of NS5PolDV contributing to this site reside within residues T794 to A799 ., Their deletion forces NS5PolDV to initiate de novo RNA synthesis internal to the template using GTP as the first nucleotide ( Figure 5C panel 1 ) and to perform primer-dependent RNA synthesis ( Figure 5B ) ., In analogy to the structure of HCV NS5B in complex with a nucleotide in its priming site 31 and because of the amino acid conservation observed within a larger group of de novo RdRps 25 , we expect that NS5PolDV residues R472 ( RdRp catalytic motif F3 , see 14 ) as well as S710 and R729 ( motif E ) are involved in triphosphate binding ., This might explain why de novo RNA synthesis initiation by the loop-deleted mutant is still possible , albeit internal to the template ., We conclude that indeed the T794-A799 loop plays a major role both in correct de novo initiation and in shaping the priming site ., Within the priming loop , residue H798 is essential for primer synthesis ( Figure 6 ) ., We propose that H798 provides the initiation platform against which the priming nucleotide ATP is stacked ., Using the structure of the de novo initiation complex of the RdRp of bacteriophage φ6 23 as a starting point , we generated a model of the initiation complex of DV serotype 2 RdRp in complex with the 3′- end of the genome UUCU and both ATP and GTP as first and second nucleotide , respectively ( Figure 7 ) ., In this model , the triphosphate moiety of ATP indeed interacts with residues S710 , R729 and R737 of the thumb subdomain of NS5PolDV ., The aromatic ring of H798 stacks the adenine nucleobase of ATP in a similar position to a φ6 RdRp tyrosine residue against which the guanine nucleobase of its priming GTP is stacked ., In several protein complex structures histidine has been shown to bind an adenine nucleobase by stacking interactions 32 ., Nevertheless , histidine does not seem to provide any specificity towards adenine versus guanine 33 ., Our model does not propose any obvious specific interaction with the adenine base ., This might be due to the fact that the structure of NS5PolDV has been captured in a pre-initiation state ., In this state , motif F , which provides the upper part of the NTP entry tunnel in the active initiation and elongation conformation of viral RdRps , is not yet correctly positioned 34 ., The fine characterization of the ATP-specific built-in priming site of NS5PolDV awaits the crystal structure of a de novo RNA synthesis initiation complex ., We provide a mechanistic basis for the conservation of nucleotides A and U as the first and last nucleotides of the DV genome , respectively ., Figure 8 summarizes the different levels of control that ensure ATP-specific de novo RNA synthesis initiation ., Firstly , it generates and elongates the bona fide pppAG primer ( red arrows and green arrows on the right ) ., Even in the absence of any template and in the presence of Mn2+ ( Figure 8 left red arrow ) NS5PolDV is able to exclusively synthesize the pppAG primer ( Figure 2B and C , Figure 3A and C ) ., Note that we have also observed pppAG synthesis by full-length NS5 in the absence of a template ( not shown ) ., Since a sufficiently high Mn2+ concentration is present in the cell ( 0 . 1 µM to 40 µM Mn2+ in blood , brain , and other tissues 35 ) , NS5 in the replication complex might already be loaded with pppAG and thus be ready to elongate pppAG on the viral template ., The same pppAG primer is preferentially synthesized in the presence of the correct template irrespective of the metal ion present at the polymerase active site ( Figure 8 right red arrows , Figure 2A and B , Figure 3 ) ., In the presence of Mg2+ , NS5PolDV supports neither formation nor elongation of pppAG on incorrect templates ( Figure 8 blue blocked arrow , Figure 4B ) ., In the presence of Mn2+ , NS5PolDV is able to synthesize cognate dinucleotides on incorrect templates ( Figure 2C ) , but in the presence of all nucleotides and all templates ( a probably biased and more unfavorable set-up compared to the situation in the replication complex in vivo ) , pppAG is still a major product ( Figure 3C ) ., Remarkably , the pppAG/Mn2+-loaded polymerase is able to mismatch and extend pppAG in order to restore the correct 5′-end ( Figure 8 blue arrows , Figure 4 ) ., The selective extension reaction thus refrains synthesis of incorrect RNAs that could occur in the presence of incorrect templates ., All these reactions converge to the formation of pppAG and the conservation of A as the starting nucleotide at the 5′-end of viral genomic and antigenomic RNAs ., Note that the mechanistic basis of the conservation of the second nucleotide G is beyond the scope of this study ., Preliminary results generated in our laboratory indicate that both template and polymerase are important to ensure the specific incorporation of a G as the second nucleotide ( not shown ) ., Several ways of viral RNA genome maintenance and repair concerning terminal damage have been discussed 4 , among others the generation of “non-templated” primers and the use of abortive transcripts as primers ., Here we demonstrate that NS5PolDV uses these two mechanisms ., Non-templated primers are generated only in the presence of Mn2+ ., Abortive transcripts are used as primers in the presence of either Mg2+ or Mn2+ ., A third mechanism observed here is the discrimination against an incorrect template in the presence of Mg2+ ., In addition , in the case that a 3′-end might be shortened , the correction upon de novo initiation should be preceded by the addition of ( a ) nucleotide ( s ) by the terminal transferase activity of NS5 ., This activity has also been listed as another way of repairing terminal damage of viral RNA genomes 4 ., For NS5PolDV we have observed this activity before 14 and now again in the presence of Mn2+ ( Figure 1B ) ., The DV polymerase endows several of the proposed mechanisms to maintain the correct 5′ and 3′-ends of the DV genome and antigenome ., The ability of DV and WNV to restore a U at the very 3′-end of genomes with 3′-end deletions has been demonstrated 2 , 36 ., This observation is in accordance with the existence of an ATP-specific priming site in NS5PolDV ., Tilgner et al . 2 , 36 reported the complete reversion of WNV replicon CA and CG 3′-ends to CU whereas CC was only partially reverted ., Since we have not seen preferential de novo RNA synthesis initiation starting with GG in comparison to UG or CG ( all three are possible in presence of Mn2+ , Figure 2 ) , this might be due to an intrinsic difference between DV and WNV RdRp or caused by different propensities of the erroneous templates to allow pppAG elongation ., Indeed CA and CG 3′-ends allow pppAG elongation more readily than the CC 3′-end ( Figure 4 , two independent reactions were performed ) ., Thus the CC 3′-end might therefore take longer to revert ., Furthermore , Teramoto et al . 2 , 36 observed the correction of the 5′-end from pppGAG to pppAG ., Our work provides a mechanistic explanation for their observation ., The observation of non-templated pppAG formation in the presence of Mn2+ by a viral RdRp has not been reported before using recombinant RdRp assays ., However , previous reports convey the occurrence of non-templated dinucleotide formation ., RSV , a member of the ns-RNA virus family Paramyxoviridae restores the correct 5′-pppA although minireplicons did not encode the correct 3′-U 10 ., The authors propose that RSV RdRp contains a built-in ATP-specific priming site and cite the observation that the RdRp of the related ns-RNA vesicular stomatitis virus ( VSV , Rhabdoviridae ) contains a specific ATP-binding site 37 as an argument in favor of their proposition ., When VSV RdRp assays were carried out using recombinant RdRp in the presence of Mg2+ , non-templated 5′-initiation was not | Introduction, Results, Discussion, Materials and Methods | The dengue virus ( DV ) is an important human pathogen from the Flavivirus genus , whose genome- and antigenome RNAs start with the strictly conserved sequence pppAG ., The RNA-dependent RNA polymerase ( RdRp ) , a product of the NS5 gene , initiates RNA synthesis de novo , i . e . , without the use of a pre-existing primer ., Very little is known about the mechanism of this de novo initiation and how conservation of the starting adenosine is achieved ., The polymerase domain NS5PolDV of NS5 , upon initiation on viral RNA templates , synthesizes mainly dinucleotide primers that are then elongated in a processive manner ., We show here that NS5PolDV contains a specific priming site for adenosine 5′-triphosphate as the first transcribed nucleotide ., Remarkably , in the absence of any RNA template the enzyme is able to selectively synthesize the dinucleotide pppAG when Mn2+ is present as catalytic ion ., The T794 to A799 priming loop is essential for initiation and provides at least part of the ATP-specific priming site ., The H798 loop residue is of central importance for the ATP-specific initiation step ., In addition to ATP selection , NS5PolDV ensures the conservation of the 5′-adenosine by strongly discriminating against viral templates containing an erroneous 3′-end nucleotide in the presence of Mg2+ ., In the presence of Mn2+ , NS5PolDV is remarkably able to generate and elongate the correct pppAG primer on these erroneous templates ., This can be regarded as a genomic/antigenomic RNA end repair mechanism ., These conservational mechanisms , mediated by the polymerase alone , may extend to other RNA virus families having RdRps initiating RNA synthesis de novo . | The 5′- and 3′-ends of RNA virus genomes have evolved towards efficient replication , translation , and escape from defense mechanisms of the host cell ., Little is known about how RNA viruses conserve or restore the correct ends of their genomes ., The Flavivirus genus of positive-strand RNA viruses contains important human pathogens such as yellow fever virus , West Nile virus , Japanese encephalitis virus and dengue virus ( DV ) ., The Flavivirus genome ends are strictly conserved as 5′-AG…CU-3′ ., We demonstrate here the primary role of the DV polymerase in the conservation of the first and last genomic residue ., We show that DV polymerase contains an ATP-specific priming site , which imposes a strong preference for the de novo synthesis of a dinucleotide primer starting with an ATP ., Furthermore , the polymerase is able to indirectly correct erroneous sequences by producing the correct primer in the absence of template and on templates containing incorrect nucleotides at the 3′-end ., The correct primer is productively elongated on either correct or incorrect templates ., Our findings provide a direct demonstration of the implication of a viral RNA polymerase in the conservation and repair of genome ends ., Other polymerases from other RNA virus families are likely to employ similar mechanisms . | virology, viral enzymes, biology, microbiology | null |
journal.ppat.1004348 | 2,014 | Rabies Virus Hijacks and Accelerates the p75NTR Retrograde Axonal Transport Machinery | Rabies virus ( RABV ) is a neurotropic negative-strand RNA virus of the Lyssavirus genus , belonging to the Rhabdoviridae family ., It is transmitted mostly via bites of diseased animals and causes a fatal infection of the nervous system in both animals and humans ., A key step in RABV pathogenesis is rapid transfer to the Central Nervous System ( CNS ) through the Peripheral Nervous System ( PNS ) 1 ., Due to its extraordinary properties in directed axonal transport and trans-synaptic spread , RABV has also been used as a neuro-tracing agent to map neuronal circuitry 2–5 ., Thus , understanding the mechanism of RABV transport is of high significance for both basic and applicative fields ., RABV enters the peripheral nervous system and undergoes long-distance transport arriving at the cell soma and subsequently the CNS 6 ., As peripheral neurons are highly polarized cells with long axons , active intracellular transport is vital to the maintenance of neuronal function and survival 7 , 8 ., Axonal transport is the cellular process of trafficking proteins , organelles , vesicles , RNA and other cellular factors to and from the neuronal cell body ., The molecular motor kinesin drives transport from the cell body anterogradely , supplying proteins , lipids and other essential materials to the cell periphery ., Dynein/dynactin complexes drive retrograde transport , moving damaged proteins for degradation and critical signaling molecules such as neurotrophins to the cell body 9 , 10 ., Although RABV phosphoprotein P , a component of the viral nucleocapsid of infecting virions , was shown to directly interact with a light chain of the dynein motor complex 11 , 12 , axonal RABV transport and CNS infection are independent of that interaction 13 and long distance transport of complete enveloped virions within internalized endosomes is more likely 14 ., However , the cellular and molecular mechanisms involved in RABVs infection and retrograde trafficking are yet to be understood ., Entry of RABV into the cell requires binding of the viral glycoprotein ( G ) and fusion of the virus envelope with the host cell membrane 15 ., Following receptor binding and fusion , RABV may enter the host cell through the endosomal transport pathway ., In neurons , infected cells may mistake RABV particles for cargo and thus recruit trafficking components , allowing viral particles to undergo long-range axonal transport to the neuronal cell body , as was found in the case of adenovirus and the CAR receptor 16 , 17 ., Direct evidence for this notion is still lacking for RABV , as well as the identity and role of the molecular determinants of the axonal transport machinery RABV utilizes ., Both the Neuronal Cell Adhesion Molecule ( NCAM ) and the p75 neurotrophin receptor ( p75NTR ) have been identified as RABV glycoprotein G binding receptors 18 , 19 ., Other membrane-associated components have also been implicated in RABV binding 20 ., By binding one of its receptors , RABV could enter the cell and activate downstream signaling which would allow it to hijack and manipulate axonal transport machineries ., Although p75NTR is known to be involved in the retrograde transport of neurotrophic factors , little is known regarding its direct contribution to viral transport ., It was recently shown , however , that lentiviral vectors pseudotyped with RABV-G are retrogradely transported in motor neurons and co-localize with both p75NTR and NCAM 21 ., The p75NTR contains four cysteine-rich domains ( CRD ) in the N-terminal ectodomain and a type II death domain in its cytoplasmic C-terminal segment ., Rabies virus glycoprotein specifically interacts with high affinity with the first Cysteine-Rich Domains ( CRDI ) of p75NTR 22 ., Neurotrophins , on the other hand , bind to the second and third p75NTR cysteine-rich domains ( CRDII&III ) 23 ., Hence , RABV and neurotrophins do not compete for each others binding site ., However , it was previously reported that treatment of cells with NGF and Neurotrophin-3 , ligands of p75NTR , modulates RABV infection of DRG-originated neurons 24 ., Remarkably , although p75NTR binds RABV with high affinity , it is not essential for its infection 25 , further raising questions regarding the specific role of this interaction ., Here we study the strategies used by RABV to exploit axonal transport mechanisms during CNS invasion ., We tracked RABV entry at nerve terminals and studied its retrograde transport along the axon in comparison to the transport of NGF ., We show that RABV and NGF are internalized in similar time frames at similar domains along nerve tips and that RABV enters the cell along with p75NTR , suggesting common entry machineries ., Then , by tracking the transport of GFP labeled Rabies virions along the axon , we showed it moving in acidic compartments , mostly with neurotrophic factor receptors , yet faster than NGF ., Finally , we determined that p75NTR-dependent transport of RABV is faster and more directed than p75NTR-independent transport ., Our model suggests that RABV may enter the cell by receptor-mediated endocytosis following its binding to p75NTR , after which it enhances the efficiency of the retrograde co-transport of RABV – p75NTR complexes ., The interaction with p75NTR modulates the cellular transport machinery and serves as a mechanism to facilitate movement of RABV to the CNS ., In order to study the mechanism of RABV long distance transport , we used an optimized compartmentalized microfluidic culture chamber ., In this system , murine E12 . 5–13 . 5 DRG explants were plated in one side of the chamber , referred to here as the proximal channel ( Fig . 1A , B ) ., Explants are encouraged to extend axons to the distal axon channel through microgrooves by introduction of a gradient of NGF known to promote DRG axonal growth ( Fig . 1D ) ., By maintaining a difference in media volume we induce directional flow across grooves or channels ., Thus , fluorescent dyes like Sulforhodamine B , introduced into the distal channel that contains less medium , are prevented from reaching the grooves or proximal channel where cell bodies are located ( Fig . 1C ) ., Hence , EGFP-RABV added to the axon terminus , binds exclusively to the distal axon , enabling retrograde tracking of the virus along axons in the groove ., Following serum and trophic factor starvation , EGFP-RABV virions were introduced into the distal channel and 1–2 hours later were observed to move retrogradely towards the cell body , as seen by time lapse imaging ( Fig . 2A , C and Movie S1 ) ., X-Y coordinates of particles moving over time were manually registered and compiled into tracks ., 87 . 29% of RABV particles ( n\u200a=\u200a244 ) , were visible over at least 10 consecutive frames and had an average instantaneous velocity >0 . 1 µm/sec ., These were considered directed particles and characterized further ( Fig . 2E–J ) ., Since RABV-G is known to bind the p75NTR neurotrophin receptor , we asked whether RABV exploits NGFs endogenous transport machinery in order to facilitate its own transport to the cell body , and eventually to the CNS ., To address this question , we applied Quantum-Dot conjugated NGF to axon tips in the distal side of the chamber after starvation , and tracked its retrograde transport along grooves ( Fig . 2B , D and Movie S2 ) ., Detailed transport analysis of RABV and NGF puncta along axons , demonstrates that RABV moves significantly faster ., The average speed of RABV was roughly 40% higher than that of NGF ( 0 . 93±0 . 03 versus 0 . 66±0 . 02 µm/sec , respectively ) ( Fig . 2E ) ., While both RABV and NGF particles presented a “stop and go” motion ( Fig . 2C , D ) RABV particles demonstrated a more processive movement with fewer stops ( Fig . 2F ) ., Though spot duration did not differ significantly ( Fig . 2F ) ., RABV particles spent a larger percentage of their traffic time in direct movement ( Fig . 2H ) ., Moreover , RABVs instantaneous velocity distribution profile was shifted towards the higher velocities ( Fig . 2I ) , averaging at 0 . 785±0 . 008 µm/sec , while the average for NGF was 0 . 546±0 . 007 µm/sec ( n\u200a=\u200a7318 and 5452 , respectively ) ., Hence , RABV moves faster and travels longer distances ( Fig . 2J ) ., These findings suggest that RABV not only exploits the axonal mechanism for neurotrophin transport , but might also increase transport efficiency ., Explants grown in microfluidic chambers tend to vary in the axonal meshwork formed at the distal channel , and consequently in the number of viral particles found in each groove ., We therefore checked whether the number of tracked particles per groove had an effect on measured values ., We found no correlation between number of EGFP-RABV puncta tracked per axon and that of percentage of directed puncta , their speed , displacement or run length ( Fig . S1 ) ., In order to illustrate the differences leading to faster transport of RABV compared to NGF , we proceeded to inquire whether internalization of these ligands occurs over similar time frames ., To this end , we performed a series of live imaging experiments using TIRF microscopy , and tracked fluorescent RABV or NGF particles at the axonal growth cone ., Distinct features of the TIRF evanescent wave allow us to limit our view to the basal surface , an ideal set up for viewing internalization processes occurring at axon tips ., Fluorescently labeled RABV or NGF was applied to DRG explant cultures and their dynamics at the axon tips were tracked ( Fig . 3 ) ., Both RABV and NGF particles demonstrated similar internalization profiles ., They arrived at the axon extremity , anchored to the cell membrane and then travelled for a few seconds before eventually internalizing into the cell ., Internalization was manifested as a gradual decrease in particle intensity until complete disappearance ( see methods for more details ) , ( Fig . 3A , B and Movies S3 , S4 ) ., Quantification of internalization durations revealed similar kinetics for RABV and NGF ( Fig . 3C , D ) ( 9 . 66±1 . 97 seconds and 13 . 25±1 . 58 seconds , respectively , p\u200a=\u200a0 . 148 ) ., RABV-G was shown to bind the p75NTR with high affinity 22 , yet there is no direct evidence demonstrating that p75NTR can act as a receptor to mediate RABV internalization ., We therefore conducted a second series of live TIRF imaging assays where EGFP-RABV was applied to DRG explant cultures along with a fluorescent antibody against the extracellular domain of the p75 receptor ., Dual color TIRF live imaging revealed that RABV and p75NTR were dynamically co-localized at the cell membrane , and were internalized together at the axon tips ( Fig . 4A , C and Movie S5 ) ., Interestingly , tracking dual-color particles indicated that these followed a directed path towards the center of the axonal growth cone prior to their internalization ( Fig . 4B ) ., Thus , RABV binds and is internalized at the axon tip together with p75NTR in a manner similar to that of NGF ., In order to further validate the intimate proximity of p75 and RABV on the cell surface , we used single particle localization algorithms , to determine the relative position of co-localized RABV and p75 spots from live TIRF images at subpixel resolution using Gaussian and radial symmetry fitting ( Fig . 4D–F ) ., Distances between p75 and RABV were measured according to the center positions of the radial symmetry fits , using Parthasarathys Radial Center algorithm 26 , and averaged on 85 . 5±20 . 82 nm ( n\u200a=\u200a7 ) ., This measurement may reflect the actual distance between the rabies virion , roughly 100×200 nm in size , and the p75NTR it binds ., Other factors , such as the size of the labeling antibody and imperfect optical alignment may have also contributed to the measured distance ., Co-localization was further confirmed by stimulated emission depletion ( STED ) microscopy , on DRG explants were treated with either a combination of RABV-EGFP and anti-p75-550 , or RABV-mCherry and anti-p75-488 ( Fig . S2 ) ., Our findings redirect attention to the question of p75NTRs role in RABV infection , in light of previous studies which have shown that p75NTR is not essential for RABV infection but afects clinical manifestation 27 ., To address this question , we measured RABV infection rates in p75NTR-knocked down DRG cultures ., Mixed infection with 4 shRNA constructs against p75NTR decreases p75NTR levels in DRGs ( Fig . S3 ) ., Short inoculation times of 30 and 120 minutes were chosen , to study p75NTRs role in promotion of infection ., Lower infection rates were observed in cultures infected with shRNA as opposed to GFP infected controls ( Fig . 4G ) , at both time points ., Under these conditions , p75NTR expression enhances RABV infection of embryonic sensory neurons ., Similar observations were made upon p75NTR knockdown in the NSC34 motor neuron cell line ( data not shown ) ., As both RABV and NGF traffic retrogradely to the cell body , and use a similar internalization process , we asked whether RABV hijacks the NGF endosomal retrograde transport machinery ., To address this issue we performed dual color live imaging of RABV and NGF retrograde axonal transport in microfluidic chambers ., Tracking RABV and NGF , we analyzed ∼50 events in which RABV and NGF were retrogradely transported together in the same compartment along the axon ( Fig . 5A–F and Movie S6 ) ., Examining the distribution of average track speeds we found that the co-transported RABV/NGF particles could be divided into two populations ( Fig . 5G ) ., Separate characterization of these two populations shows that faster tracks were less prone to pausing mid-way ( 0 . 3±0 . 2 versus 1 . 7±0 . 2 pauses per 100 seconds , respectively ) ., As only 2 pauses were recorded in the fast group , no significant difference was found between either groups stop durations ( Fig . 5H , I ) ., Overall , the fast group , thus spent less time paused ( Fig . 5J ) ., The presence of two distinct populations of RABV/NGF co-transport may suggest to a switch in “drivers” , where NGF leads the slower group while the faster is led by RABV ., We proceeded to further characterize the transport mechanism of RABV particles; seeking to first determine the cellular compartment in which RABV particles are transported , post infection , within DRG axons ., Using microfluidic chambers ( Fig . 1 ) , we infected axons in the distal compartments , while simultaneously treating cells with fluorescent markers of cellular compartments ( Fig . 6 ) ., In order to check that RABV is transported in acidic compartments 28 , possibly late-endosomes , lysosomes or autophagosomes , and to quantify this co-localization , we treated cells with the PH-indicator dye , Lysotracker Red ., In order to observe whether RABV is transported with mitochondria , we treated cells with the Mitotracker Deep Red marker ., Using both markers together with EGFP-RABV , we acquired three channel time- lapse image series of RABV transport ( Fig . 6A–D ) ., Detection and analysis of co-localized fluorescent spots along the axon determined that most of the RABV particles ( 75 . 9±4 . 09% , n\u200a=\u200a3 separate experiments ) were located in acidic compartments ( Fig . 6E ) ., Examination of merged kymographs of RABV and Lysotracker red amplified the outcome of co-localization analysis , as most transient RABV tracks were matched with a Lysotracker red track ( Fig . 6G–I ) ., Thus , RABV particles are located in acidic axonal endosomes that retrogradely move towards the cell body ., In contrast , we could not detect any significant co-localization of RABV and mitochondrial marker ( Fig . 6F ) ., We therefore conclude that while RABV is mostly transported in an acidic compartment , it is not transported along with mitochondria in DRG axons ., Having determined that RABV is internalized with p75NTR and can undergo transport with NGF , we then tested whether it is also transported with more selective neurotrophin receptors ., In order to study this , we infected DRG axons grown in microfluidic chambers with EGFP-RABV , while labeling these cells with fluorescent antibodies against neurotrophin receptors ., Specifically , we used fluorescent-tagged antibodies against the general neurotrophin receptor p75NTR and the specific NGF receptor TrkA and acquired three-channel time-lapse movies of RABV and these receptors ( Fig . 7A–D ) ., Co-localization analysis determined that the over 60% of RABV particles co-localized with p75NTR , yet less than 40% with TrkA ( Fig . 7E , F , H and Movie S7 ) ., Interestingly , when co-localized with neurotrophic factor receptors , mainly p75NTR , RABV tracks showed a greater processivity than that of RABV-only tracks ( Fig . 7G , H ) , suggesting that mutual transport of RABV with NTF receptors induces a more progressive transport ., Although p75NTR may serve as a receptor for RABV ( Fig . 4 and Movie S5 ) and the viral G-protein binds with high affinity to the p75NTR , the receptor is not an absolute requirement for RABV infection , as RABV can also infect p75NTR deficient cells 25 ., Furthermore , RABV could be transported retrogradely along the axon without p75NTR ( Fig . 7 ) ., To assess the precise contribution of p75NTR to RABV transport , we tracked the transport of RABV particles along the axon , with and without p75NTR ( Fig . 8 A–C and Movie S8 ) ., Plotted kymographs of RABV and p75NTR demonstrate that motile RABV tracks tend to co-localize with those of p75NTR ( Fig . 8D–F ) ., Separate characterization of each groups transport ( Fig . 8G–O ) demonstrated that when traveling with p75NTR , RABV particles traveled faster compared to particles negative for the receptor , with respective speeds of 0 . 86±0 . 04 versus 0 . 63±0 . 04 µm/sec ( Fig . 8G ) ., We attribute this alteration in speed to the fact that RABV-p75 puncta are less prone to pausing on their route to the cell body , with an average of 0 . 9±0 . 1 vs . 2 . 3±0 . 2 pauses per 100 seconds for the p75NTR positive and negative groups , respectively ., Moreover , p75 positive particles paused for shorter times ( Fig . 8I ) and overall spent less time paused during their travel ( Fig . 8J ) ., Another factor contributing to their higher respective speed was their instantaneous velocities ., The distribution of instantaneous velocities of RABV particles positive for p75NTR is shifted towards the higher velocities when compared to RABV negative for p75NTR ., RABV ( + ) p75NTR average velocity was higher than that of RABV ( − ) p75NTR , 0 . 71±0 . 008 versus 0 . 49±0 . 007 µm/sec ( n\u200a=\u200a5494 and 3919 , respectively ) ., Another interesting finding was that less than 10% of the recorded velocity events in the p75NTR positive group were anterograde , while over 17% of events in the p75NTR negative groups were anterograde , i . e . moving “backwards” towards their entrance area , the axon tip ., This implies that p75NTR not only has a role in assisting the transport of NTF and viruses , but might also regulate the transport machinery , contributing to vesicle directionality towards the cell body ., Measuring both the area and average intensities of the RABV particles in each group , we found that the p75NTR positive RABV particles were larger in size , ( average area of 1 . 34±0 . 09 µm2 vs . 0 . 81±0 . 07 µm2 , p<0 . 0005 ) , and had stronger intensity of GFP signal , when normalized to the average intensity of RABV particles in each experiment ( 1 . 24±0 . 11 vs . 0 . 61±0 . 1 , p<0 . 001 ) ( Fig . 8K–L ) ., These larger and more prominent particles possibly represent larger endosomes , containing several RABV particles and receptors ., These particles , positive for p75NTR , cover a greater net distance per time unit and are more directed towards the cell body , as shown when we compared their trajectories and mean squared displacements ( Fig . 8N and O , respectively ) ., Taken together , the major differences that were observed between the two RABV groups suggest distinguishable transport mechanisms ., It therefore seems that RABV binding to p75NTR allows the virus to exploit a rapid transport mechanism to facilitate its trafficking to the cell body ., We continued to examine the role of p75 in RABV transport by tracking RABV in axons of a DRG explant after p75 knock-down ., DRG explants were grown in microfluidic chambers and infected with LV-shRNA-p75-EGFP ., mCherry-RABV was added to the distal channel as described before , and after 1 hour of incubation imaged for 1–2 hours ., Although many axons crossed the grooves to the distal channel , only few were found to express GFP , hence were infected with sh-p75 ., Unlike in non-infected axons , where RABV was easily identified when trafficked towards the cell body , RABV transport in sh-p75 axons was less frequent ( Fig . S4A–C ) ., The number of transported RABV particles was reduced in sh-p75 axons when compared to adjacent , non-infected axons or to LV-EGFP infected controls ( Fig . S4D ) or to LV-EGFP infected controls ( not shown ) ., The few RABV particles in sh-p75 axons were less directed than those transported in non-infected cells , as seen by their respective trajectories ( Fig . S4E ) ., A crucial initiating event for the mechanism outlined above is the binding of RABV to p75NTR ., Here we provide direct evidence that p75NTR may serve as a receptor for internalization at axons tips , as well as mediate incorporation into the endosomal neurotrophic transport pathway ., However , RABV does not strictly rely on p75NTR for internalization and may enter the cell in a p75NTR independent pathway , while is also known to bind other receptors 20 , 30 ., Hence there are likely to be additional ways for RABV to merge into the p75NTR-RABV endosome ., Indeed , we observed events where RABV particles merge with p75NTR-positive endosomes en route ( Movie S9 ) ., The p75NTR neurotrophin receptor accelerates RABV transport to the cell body , yet there are instances of fast , processive transport of RABV particles without the receptor ., We assume other identified RABV receptors such as NCAM 30 or other , un-identified ones , may facilitate RABVs retrograde axonal transport within endosomes in a similar fashion ., Our experiment with DRG cultures where p75 was knocked down , show reduced RABV infection and transport ( Fig . 4 and Fig . S4 ) , and further support the role for p75NTR proposed here ., Some viruses such as Herpes Simplex Virus can travel along the axon independently of a membrane compartment , as capsids 31 and control its long distance transport process directly 32–35 ., Interestingly , the RABV phosphoprotein P directly interacts with a dynein light chain 11 , 12 , suggesting a mechanism whereby this interaction is key to RABVs retrograde trafficking ., However , studies on the retrograde transport of RABV enveloped virions 14 and infection of the CNS from the periphery with dynein light chain binding defective virus mutants 13 already showed that such an interaction is not essential for retrograde axonal transport of the virus ., Our data support this finding , as we demonstrated that RABV is transported in acidic compartments ( Fig . 6 ) , and mostly in p75NTR-positive endosomes ( Fig . 7 ) ., A different role for dynein binding should thus be considered ., Dynein is well characterized as a retrograde motor , yet can also act to tether and stabilize dynamic microtubules 36 ., Possibly , RABV binding to dynein tethers projecting microtubules ( MT ) in the cell cortex thereby facilitating its retrograde trafficking from the cell periphery ., Following this tethering , RABV particles can merge into the RABV-p75NTR endosomes and travel to the neuron cell body ., Nonetheless , this RABV-MT interaction could be mediated by binding of RABV to NCAM , which was demonstrated to tether MTs at the synapse 37 ., Suggestions relating to the function of various RABV populations that can either internalize with receptors to endosomes , or act as RNP capsids to manipulate and stabilize the cytoskeleton , require further testing ., We have shown that the RABV and the acidic marker LysoTracker are co-localized and move together along the axon ., This shows that the RABV is transported in membranal compartments similar to neurotrophic factors and is unlikely to be free in the cytoplasm ., Interestingly , low pH induces conformational changes in the viral G protein , suggesting control of membrane fusion events 38 , 39 that may regulate RABV transport in acidic vesicles ., Moreover , it could be that the alteration in RABV G-protein function , as pH changes , provides a signal for RABV-p75NTR complex , leading to its transport acceleration ., The effects of RABV binding to p75NTR on axonal transport processivity and speed , suggest that down-stream signaling is activated and exerts an influence on the retrograde transport process ., Axonal transport can be regulated at four main different levels:, 1 . Microtubule tracks ., 2 . Motor proteins ., 3 . Motor-cargo adaptors ., 4 . ATP supply ., As we have shown that RABV-p75NTR complexes move both instantaneously faster and with fewer pauses , we speculate that more than one regulatory level may be involved ., p75NTR activates different signaling cascades , as a result of binding to several distinct ligands or interactions with various co-receptors ., Recently it was described that structural determinants underlie the signaling specificity of p75NTR to the JNK , RhoA and NF-kB pathways 40 ., Interestingly , c-Jun N-terminal kinases ( JNKs ) can also regulate the axonal transport process in several different ways ., JNK-interacting proteins ( JIPs ) are scaffolding proteins for JNK and serve as linkers between motor proteins and their membrane-associated cargos ., JIP1 serves as a linker between kinesin-1 and dynein to vesicles , and JNK signaling can modulate its transport by regulating the two opposing motors 41 ., JNK , by functioning as a kinesin-cargo dissociation factor , regulates axonal transport 42 ., Additionally , JNK3 phosphorylates kinesin-1 and inhibits its microtubule-binding activity 43 , 44 ., JNK3 and its scaffolding protein Sunday driver ( syd ) are activated after axonal injury and bind to p150 , the regulatory sub-unit of the dynein-dynactin retrograde complex 45 ., Moreover , a recent study has shown that JIP1 phosphorylation serves as a molecular switch to regulate the direction of vesicle transport in neurons , by coordinating kinesin and dynein motors 46 ., These studies show that scaffolding proteins such as the JIPs and JNK play an important role in the regulation of motor proteins and the axonal transport process , and thus might take part in RABV manipulation of the axonal transport machinery ., Another interesting speculation to explain how RABV binding to p75NTR accelerates its transport is the potential involvement of axonal activated NF-kB ., NF-kB can be activated downstream to both p75NTR and RABV in the axon before entering the nucleous , and may affect dynein activity 47–49 ., Interestingly , NF-kB activation after NGF binding to p75NTR enhances neuronal survival 50 , suggesting that p75NTR may regulate NGF retrograde signaling ., In the future , it will be interesting to examine the involvement of axonal p75NTR dependent downstream signaling activated by RABV and its effect on the neuronal cytoskeleton as well as axonal transport ., It is also tempting to consider that local protein synthesis as the result of RABV binding to p75NTR can facilitate this processive transport ., Indeed recently it was demonstrated that efficient retrograde transport of pseudorabies virus , require axonal protein synthesis 51 ., These data reveal an unexpected role for p75NTR that is not necessarily related to its functions as a neurotrophin receptor , but rather to acceleration of RABV-axonal transport ., Whether p75NTR might also provide a fast delivery of other axonal cargos , such as neurotrophic factors , pathogenic prions and tetanus toxin 1 is a question for future research ., This study has addressed the question of how RABV is transported over long distances ., Previous cell biology work in the field was performed mostly on RABV-infected cell lines or neuronal cell bodies , leading to a focus on mechanisms of infection and not on long distance transport 52 ., Although neurotropic viruses need to progress over long distances to reach the CNS 1 , 35 , how RABV performs this task was not clear ., Here , we suggest that RABV hijacks a specific mechanism that enables the neuron to transport cargos over long distances ., Interestingly , p75NTR is internalized by clathrin dependent endocytosis and is sorted into distally transported endosomes after stimulation with NGF 53 , 54 ., Furthermore , RABV internalization was characterized as a dominantly clathrin mediated process 52 , 55 ., Here we show that RABV binds to and is internalized together with p75NTR , forming endocytic compartments which undergo processive long distance transport ., This suggests similar mechanism by which RABV mimics neurotrophins for activation of p75NTR ligand-mediated internalization and transport ., As p75NTR can bind many ligands and various co-receptors , it is possible that binding to some will trigger a signaling effect that will facilitate axonal transport of other cargos and not only RABV ., As we demonstrate here for RABV , p75NTR interaction can modulate the cellular transport machinery and may serve as a novel route to increase transport efficiency and facilitate arrival of cargos to the CNS from the periphery ., ICR mice were bred and maintained at the Tel Aviv University animal care facility until the time of sacrifice ., Spinal cords were dissected from E12 . 5–13 . 5 mice , followed by separation of Dorsal Root Ganglia ( DRG ) from meninges and additional spinal cord ., The Institutional Animal Care Committee at the Tel Aviv University approved all the animal protocols in this work ., Microfluidic chambers were fabricated using methods previously described in detail 56 ., All microfluidic chambers were replica molded using PDMS ( #41201841 Dow Corning ) from masters that were patterned using the photosensitive epoxy SU-8 ( Microchem ) ., All masters consisted of two permanent SU-8 layers on a 3″ silicon wafer and were made in the clean room facility in Tel-Aviv University ., The first layer of SU-8 ( 3 µm depth ) contained the microgrooves , which were patterned by photolithography using a high-resolution chromium mask ( 5 µm minimum feature size; Advance Reproduction Corp . ) ., The second layer of SU-8 ( 100 µm depth ) contained the compartments , which were patterned by photolithography using a 20 , 000 dpi printed transparency mask ( CAD/Art Services , Inc . ) ., Chamber dimensions: channels: length 8 . 25 mm , width 1 . 5 mm; grooves: length 400 µm , width 15 µm , height 5 µm ( Fig . 1B ) ., A single 7 mm well was punctured into the “proximal” or explant channel , into which a “cave” was carved using a scalpel , to prevent explants from floating , 2 additional 1 . 2 mm wells were punctured into the channel on either side of the “explant” well to allow flow ., Two 7 mm wells were punctured into both ends of the “distal” or axons channel to allow control over the distal channel ., Microfluidic devices were cleaned of surface particles using adhesive tape and sterilized in 70% high-grade ethanol for 1 h ., Devices were allowed to completely air dry under sterile conditions , attached to sterile 50 mm glass bottom dishes ( FD5040-100 , WPI ) using gentle pressure and heated to 70°C for 20′ to improve adhesion to glass ., Chambers were coated using 150 µl of 1 . 5 ng/ml Polyornithine ( P-8638 , Sigma ) in PBS for 24 hours , which was replaced with 150 µl Laminin ( L-2020 , Sigma ) 1∶333 in DDW for 24 hours ., Laminin was replaced with culture medium until plating ( 1–3 days ) ., At the day of plating , media were removed from all wells , and a single DRG was inserted to each explant “cave” using a 20 µl tip ., Following 1 hour of incubation at 37°C | Introduction, Results, Discussion, Materials and Methods | Rabies virus ( RABV ) is a neurotropic virus that depends on long distance axonal transport in order to reach the central nervous system ( CNS ) ., The strategy RABV uses to hijack the cellular transport machinery is still not clear ., It is thought that RABV interacts with membrane receptors in order to internalize and exploit the endosomal trafficking pathway , yet this has never been demonstrated directly ., The p75 Nerve Growth Factor ( NGF ) receptor ( p75NTR ) binds RABV Glycoprotein ( RABV-G ) with high affinity ., However , as p75NTR is not essential for RABV infection , the specific role of this interaction remains in question ., Here we used live cell imaging to track RABV entry at nerve terminals and studied its retrograde transport along the axon with and without the p75NTR receptor ., First , we found that NGF , an endogenous p75NTR ligand , and RABV , are localized in corresponding domains along nerve tips ., RABV and NGF were internalized at similar time frames , suggesting comparable entry machineries ., Next , we demonstrated that RABV could internalize together with p75NTR ., Characterizing RABV retrograde movement along the axon , we showed the virus is transported in acidic compartments , mostly with p75NTR ., Interestingly , RABV is transported faster than NGF , suggesting that RABV not only hijacks the transport machinery but can also manipulate it ., Co-transport of RABV and NGF identified two modes of transport , slow and fast , that may represent a differential control of the trafficking machinery by RABV ., Finally , we determined that p75NTR-dependent transport of RABV is faster and more directed than p75NTR-independent RABV transport ., This fast route to the neuronal cell body is characterized by both an increase in instantaneous velocities and fewer , shorter stops en route ., Hence , RABV may employ p75NTR-dependent transport as a fast mechanism to facilitate movement to the CNS . | Rabies virus ( RABV ) is a neurotropic virus that depends on long distance axonal transport in order to reach the central nervous system ( CNS ) ., The strategy RABV uses to hijack the cellular transport machinery is unknown ., Here we use live cell imaging to track RABV entry at nerve terminals and study its retrograde transport along the axon ., First , we demonstrate that RABV interacts with the p75 neurotrophin receptor ( p75NTR ) at peripheral neuron tips to enter the axon ., Then , characterizing RABV retrograde transport along the axon , we showed that the virus moves in acidic compartments , mostly with p75NTR ., Interestingly , RABV is transported faster than NGF , an endogenous p75NTR ligand ., Finally , we determine that p75NTR-dependent transport of RABV is faster and more directed than p75NTR-independent RABV transport ., Hence , RABV not only exploits the neurotrophin transport machinery , but also has a positive influence on transport kinetics , thus facilitating its own arrival at the CNS . | neuroscience, cell biology, biology and life sciences, molecular biology | null |
journal.pgen.1004122 | 2,014 | Coherent Functional Modules Improve Transcription Factor Target Identification, Cooperativity Prediction, and Disease Association | Transcriptional networks are fundamental to many aspects of biology and disease ., Gene expression is a carefully controlled process orchestrated by the activities of transcription factors ( TFs ) which regulate the transcription of each gene ., TFs usually do not work in isolation , but instead multiple factors combine in different ways to regulate groups of genes in a concerted , often cooperative fashion 1–6 ., The ENCODE project has begun to determine the binding locations of many transcription factors using chromatin immunoprecipitation ( ChIP ) followed by high-throughput sequencing ( ChIP-Seq ) 7 , 8 ., Despite the abundance of data about the genomic binding sites for transcription factors , determining transcription factor targets and when factors are active remains challenging ., ChIP-Seq measurements can be noisy and reflect the particular condition in which the experiments are performed ., Collecting more data alone will not solve this problem ., As additional experiments are performed for each factor , critical and frequently used binding regions do become apparent , but it is often difficult to determine a signal threshold to distinguish common sites from condition-specific sites and general non-thematic associations from interesting biology ., For example , NFκB binds to over 15 , 000 regions of the genome covering all possible regulatory targets of the factor ., But in any given biological context , such as a local cooperative interaction with another transcription factor such as Stat1 , only a handful of these genes are actively regulated by NFκB at any one time 4 ., This property of TF function gives the illusion that TFs are operating broadly when in fact they perform specific context-dependent functions–in many cases with specific partners ., These difficulties conspire to make the regulome challenging to study at a global level ., Thus , to understand transcription factor function , there is a need for computationally-efficient methods to, ( i ) improve TF-target identification ,, ( ii ) identify small functional modules that represent context-specific biology , regulated by transcription factors , and, ( iii ) annotate those modules with their functional implications ( e . g . the role of the module in human disease ) ., Recently , a number of methods were developed to derive a network structure to connect sets of genes ( modules ) to the factors that control their expression 9–11 ., These methods use gene expression data to derive the most parsimonious regulatory structure ., However , because of their computational complexity , they can only account for a limited number of factors , require a specific type of input datasets , and , in their current form , cannot integrate other experimental data ( e . g . ChIP-Seq ) ., Thus , these methods may not be sufficient to capture the scale and complexity of the human regulome ., Additionally , their usefulness is hampered by an assumption that the activity of the TF can be estimated by its expression , which , the authors of these methods acknowledge is not true in many cases 9 ., Efficient approaches with the capability to integrate multiple data modalities are needed in order to properly leverage high-throughput experiments in the study of disease ., In this paper , we use factor analysis as a computationally-efficient method to, ( i ) improve the identification of transcription factor targets ,, ( ii ) identify functional modules from gene expression data , and, ( iii ) use these modules to annotate transcription factors and connect them to diseases ., There are several methods for decomposing expression data to find groups of genes that work together ., Network Component Analysis ( NCA ) is a method for inferring transcription factor activity from expression data 12 and has been used to build regulatory networks for model organisms 13 ., However , NCA requires a priori knowledge of the regulatory structure which is often not available , and introduces bias in the associations between TFs and functional components ., On the other hand , independent component analysis ( ICA ) is an unbiased and efficient method for deconvolving the signal from a fixed set of sources measured by a set of sensors ( Figure 1 ) ., In essence , ICA is a computational method for extracting a set of signals from noisy data ., When applied to gene expression data – like those recorded by microarrays – ICA can identify coherent functional modules ( we refer to each ICA component as a module ) ., Importantly , ICA allows genes to participate in multiple modules and thus has some ability to capture different biological contexts ., A set of 423 data-driven modules derived from an ICA of 9 , 395 human expression microarrays covering a wide diversity of human biology was recently reported 14 ., Here , we hypothesize that regulation of each of these ICA-derived modules is controlled by a small set of TFs ., Using our method ( which we call TFICA ) , we associated transcription factors to modules and then analyzed the genes contained within each module ., Intersecting these target modules with ChIP-Seq binding sites improves target identification and elucidates the functional roles of the factors–both individually and in combination ., We compare our approach to traditional methods in three areas: the identification of, ( i ) transcription factor targets ,, ( ii ) TF-TF cooperativity ,, ( iii ) and the functional roles in the context of various diseases ., In each of these cases , we found that our approach significantly outperforms the traditional methods ., Further , we found improved performance when our approach is used in combination with traditional methods , implying that we are capturing an independent modality of transcription factor activity ., Our data-driven approach is unbiased and computationally efficient enabling systematic identification of novel TF-disease relationships ., Finally , we validate one such association between MEF2A and Crohns disease ., We used a set of functional modules derived using ICA 14 ., We then used ENCODE ChIP-Seq experimental data to connect transcription factors to individual modules if the factor bound a significant number of genes in that module ( Figure 2A; see Materials and Methods ) ., For 143 transcription factors and 379 modules , we identified 5 , 002 significant TF-module associations ( with adjusted p<0 . 01 , Fishers exact test; Figure S1 and Table S1 ) for an overall FDR of 16 . 6% ., We hypothesized that the components from ICA represent a single regulatory signal analogous to a single voice recorded by set of microphones ( as in Figure 1 ) ., Thus , modules that associate with only one or a few factors correspond to cleaner signals than those associated with many factors ., We identify 31 associations which we called “high-confidence” as there was only one TF significantly associated with the the module and another 142 “medium-confidence” associations , where the module was associated with three or fewer TFs ., We found that many of the modules nearly or fully overlap with targets of only one or a few transcription factors ( Figure S3A–C ) ., We found that for 171 modules the top associated TF could account for 80% of the targets in the module ( Figure S3D ) ., Additionally , the modules that explain the most variance across the compendium of 9 , 395 gene expression experiments are significantly associated with a larger number of TFs ( Figure S4 ) , and may represent large transcriptional programs ., Used in combination with the ChIP-Seq data , we hypothesized that these modules can improve identification of transcription factor targets ., Specifically , we believe that putative targets ( as determined by ChIP-Seq ) which are also contained within significantly associated modules will be more likely to be “true” targets of the TF ., To test this hypothesis , we used a set of specific GO terms 15 and we considered shared functional annotation as a proxy for a high quality TF-target association ., This strategy has been used successfully in computational evaluation previously 15 ., As expected , we found that TF-target pairs , particularly ones with high ChIP-Seq scores , were enriched for pairs with shared functional annotations ( p<0 . 001; Figure 2B ) ., When considering only the targets in the 5 , 002 modules associated with TFs , we find a significantly higher enrichment for shared annotations ( Figure 2B ) ., This enrichment is maintained across all ChIP-Seq peak scores ( p<0 . 05 ) ., Additionally , if we only consider targets in the 142 medium-confidence or the 31 high-confidence modules , this enrichment is increased further at all peak score thresholds ( p<0 . 001 ) ., In an analogous fashion to the shared functional annotation approach , we used expression analysis to validate our TF-target associations under the assumption that factor expression can be used as a proxy for factor activity ., We hypothesized that for high-confidence modules ( i . e . those associated with just one TF ) the genes within this module should be controlled predominantly by that single factor ., To test this hypothesis we examined the correlation between the expression of the module ( see Methods ) and the expression of the factor across the compendium of 9 , 395 gene expression experiments ., For example , AP-2γ is the sole significant association for module 360 ( OR\u200a=\u200a1 . 66; adjusted p\u200a=\u200a0 . 006 ) and we found a significant correlation between the expression correlation between module 360 and AP-2γ ( Spearman ρ\u200a=\u200a0 . 38 , p<0 . 001 ) ., We systematically evaluated all 5 , 002 TF-module pairs in this manner ( Table S2 ) ., We compared our method to two “naive” approaches for generating TF modules:, ( i ) a “best-module” constructed from only the best ChIP-Seq hits for each TF according to their peak scores and, ( ii ) a “matched-module” constructed from a random sample of the TFs ChIP-Seq targets with the same distribution of peak scores as was found in the ICA module ., For high-confidence and medium-confidence modules TFICA outperforms the best-module method 60% of the time ( binomial p\u200a=\u200a0 . 011 ) , and the matched-module method 77% of the time ( binomial p\u200a=\u200a3 . 2e-11; Figure S5 ) ., In fact , TFICA outperforms both naive approaches even for those modules associated with many transcription factors ( >3 ) ., This holds until modules are associated with 40 or more transcription factors , at which point the individual factor expression signal can no longer be observed ( Figure S5 ) and best-module begins to outperform TFICA ., TFICA outperforms matched-modules regardless of the number of factor associated to the module ( Figure S5 ) ., Additionally , six of the modules that are most enriched for a TFs targets are also the most correlated module for that TF ( OR\u200a=\u200a25 . 4 , Fishers exact P\u200a=\u200a2 . 8e-7 ) , and 15 are in the top 5% of all modules ( OR\u200a=\u200a5 . 6 , Fishers exact P\u200a=\u200a1 . 9e-10 ) ., Finally , we found that the top co-expressed module is significantly enriched ( Materials and Methods ) for ChIP-Seq binding sites for 37 TFs ( OR\u200a=\u200a4 . 9 , Fishers exact P\u200a=\u200a2 . 2e-12 ) ., We exhaustively evaluated the expression correlation between all TFs and all modules to estimate a null distribution , and found that our TFICA TF-module pairs were significantly more correlated than expected by chance ( ρ\u200a=\u200a0 . 05 vs . −0 . 06; t-test p\u200a=\u200a1e-204 ) ., Additionally , in all , 327 TF-module pairs remained significantly correlated when compared to an empirically derived TF-specific null distribution ( p<0 . 05; Figure S5 ) ., Many modules connected to TFs were significantly enriched for functional annotations known to be associated with the factor ( Table 1 and Figure S2 ) ., For instance , sterol regulatory element-binding protein 2 ( SREBP2 ) and module 158 was our most significant TF-module association ( Figure 2A; OR\u200a=\u200a45 . 2; 95% CI\u200a= ( 27 . 8 , 71 . 6 ) ; adjusted p\u200a=\u200a1e-31 ) ., SREBP2 is essential for cholesterol and fatty-acid biosynthesis , and module 158 is significantly enriched for GO terms related to lipid , sterol , cholesterol , and steroid synthesis ( adjusted p<0 . 05; Table 1 ) ., In addition , SREBP2 shares many of the same target modules as SREBP1 ( Figure S2 ) , which are known regulatory partners ., Another example is the association between the transcription factor ZNF274 and module 111 ., Module 111 includes many zinc finger proteins of which ZNF274 is a known regulator 16 ., In addition , ZNF274 clusters near SETDB1 and KAP1 ( Figure S2A ) and has been shown to recruit both of these transcription factors to repress the expression of other zinc finger proteins 17 , 18 ( Table 1 ) ., Module 57 was associated with the greatest number of transcription factors ( 121 different factors; Table 1 , Figure S2 ) ., This module also contains the greatest number of transcription factors as targets ( 14 factors ) ., We found this module to be significantly enriched for DNA binding , regulation of transcription , and transcription regulator activity ( among other regulatory terms; Table 1 ) , and may represent a master regulatory module of other TFs ., As we have demonstrated , we can significantly improve the identification of TF targets using TFICA modules ., Therefore , we hypothesized that TFs that target overlapping modules may function together to regulate gene expression ., We found 3 , 696 transcription factor pairs ( comprising 135 individual TFs ) that share a significant proportion of target modules ( adjusted p<0 . 01 , Fishers exact test; Table S3 ) ., We assessed the putative TF-TF interactions predicted by TFICA using expression correlation , literature co-occurrences , and shared functional annotation ., We compared the predictions of two TFICA similarity metrics ( simple Tanimoto and a weighted approach which places more emphasis on higher confidence TF-module pairs ) to those from a naive method of simply intersecting ChIP-Seq targets ., We evaluated using multivariate linear models and assessed significance with an ANOVA ( Figure 3A ) ., Both TFICA approaches outperform the naive method in all 3 evaluations ( Figure 3A ) with weighted TFICA exhibiting the best performance ., In addition , the combined model of both TFICA-similarity plus shared targets significantly outperforms the naive approach alone in all three of these metrics ( Figure 3A and Figure S6 ) ., TF pairs from TFICA are significantly ( 1 ) more correlated in their overall gene expression across the compendium ( simple: F\u200a=\u200a28 . 6 , p\u200a=\u200a1 . 03e-7; weighted: F\u200a=\u200a41 . 2 , p\u200a=\u200a1 . 75e-10 ) , ( 2 ) more likely to co-occur in Pubmed abstracts ( simple: F\u200a=\u200a22 . 0 , p\u200a=\u200a2 . 92e-6; weighted: F\u200a=\u200a57 . 2 , p\u200a=\u200a6 . 51e-14 ) , and ( 3 ) more likely to share functional annotations ( simple: F\u200a=\u200a67 . 0 p\u200a=\u200a5 . 24e-16; weighted: F\u200a=\u200a119 . 4 , p\u200a=\u200a6 . 50e-27 ) ., Of the top 30 pairs ranked TF-TF pairs according to module similarity , 14 have been previously reported , such as NF-YA and NF-YB , as well as Pol2 with a number of other initiating factors ( Figure 3B ) ., Many of the unreported results may be due to sparse annotation of individual genes ( e . g . CHD2 , CCNT2 , and HEY1 ) , and may indicate new biological links ., For example , CHD2 clusters with CCNT2 and Sin3a , which are known cell cycle regulators ., CHD2 has previously been proposed as involved in the cell cycle 19 consistent with its role as a DNA damage signaling protein ., We used enrichment analysis to associate ICA functional modules to diseases from the Gene Association Database 20 ., Combined with the TFICA analysis , these two datasets allow us to create a transcription factor-disease network ., We created a network of 143 transcription factors connected by their targeted functional modules ( note that this network is naive to any disease associations ) ., TFs clustered together according to the diseases with which they are significantly associated ( Figure 4 ) ., In total , we found 7 , 808 significant associations between 141 transcription factors and 253 diseases ., The average number of diseases we associate to a transcription factor is 36 ( Figure S8A ) with four transcription factors having just one significant disease association ( e . g . BAF170 is associated with macular degeneration ) and p300 associated with the most ( 204 ) diseases ., The number of diseases associated with a transcription factor was significantly related to both the number of targets ( Figure S8B; Spearman ρ\u200a=\u200a0 . 47 , P\u200a=\u200a1 . 2e-7 ) and the number of GO annotations for the factor ( Figure S8C; Spearman ρ\u200a=\u200a0 . 39 , P\u200a=\u200a1 . 4e-5 ) ., The complete list of significant transcription factor-disease associations is available in Table S4 ., Transcription factors with known relationships to disease clustered into distinct groups ( Table 2 ) ., For example , we found a significant association between module 320 and acquired immunodeficiency syndrome ( AIDS ) ( OR\u200a=\u200a12 . 4; 95% CI\u200a=\u200a4 . 7–28 . 7; adjusted p\u200a=\u200a5 . 2E-4 ) ., Three of the TFs associated with this module ( NFKB , IRF4 , and BATF ) are known to be involved in the transcriptional regulation of human immunodeficiency virus 21–23 and cluster together in the interaction network ( Figure 4A ) ., We found a significant association between module 4 and arrhythmogenic right ventricular dysplasia ( OR\u200a=\u200a82 . 2 , 95% CI: 15 . 5–531 . 4 , adjusted p\u200a=\u200a1 . 1e-5 ) ., ER-α , c-Jun , STAT3 , and STAT1 are associated with module 4 , and all have known relationships with arrhythmias 24–28 and clustered together ( Figure 4B ) ., Thus , our network supports the previous suggestion that ER-α may be a promising prognostic marker for the development of atrial fibrillation 24 ., In addition , we found that module 123 was significantly enriched for genes associated with thrombocytopenia ( OR\u200a=\u200a9 . 9; 95% CI\u200a=\u200a2 . 89–27; adjusted p\u200a=\u200a0 . 022 ) ., A number of TFs independently associated with thrombocytopenia , including p300 and GATA1 , cluster together in the interaction network and are associated with module 123 ( Figure 4C ) ., Finally , for breast cancer , we found significant associations with modules 2 , 13 , 46 , and 154 with odds ratios of 8 . 6 ( 95% CI 5 . 6–13 . 4 , adjusted p\u200a=\u200a2 . 3e-19 ) , 3 . 1 ( 1 . 8–5 . 2 , adjusted p\u200a=\u200a0 . 004 ) , 3 . 1 ( 1 . 8–5 . 1 , adjusted p\u200a=\u200a0 . 002 ) , and 4 . 2 ( 2 . 4–7 . 3 , adjusted p\u200a=\u200a2 . 1e-4 ) , respectively ., Based on their connectivity to these modules , the transcription factors E2F6 , CHD2 , NFYA , IRF1 , HEY1 , and E2F1 all cluster together in the TF-TF network ( Figure 4D ) ., We performed an evaluation of our TF-disease associations by comparing our derived associations to an independent standard created by combining ( 1 ) 37 transcription factor-disease associations from the GWAS Catalog , and ( 2 ) 46 associations from OMIM ( see Methods ) ., We assessed overall performance using the Area Under the Receiver Operating Characteristic Curve ( AUROC ) ., TFICA achieved an AUROC of 0 . 712 on this test dataset ( Figure S7 ) ., For comparison , we also evaluated two control strategies: ( 1 ) a simple enrichment analysis on the ChIP-Seq targets associated with each transcription factor , and ( 2 ) the GREAT tool for annotation cis-regulatory elements in the genome 29 ., The simple approach achieved an AUROC of 0 . 612 , whereas GREAT achieved 0 . 687 ( Figure S7 ) ., When combined together in a logistic regression model , TFICA significantly improved the performance of GREAT , increasing the AUROC from 0 . 687 to 0 . 761 ( +10 . 7% , Chi-Squared\u200a=\u200a19 , P\u200a=\u200a1 . 1E-5 ) , suggesting the two approaches are complementary ., Finally , we repeated this analysis using the AUROC50 which is a common measure used to evaluate performance at low false positive rates ( FPR<0 . 5 ) ., We found an AUROC50 value of 0 . 185 for the naive metric , 0 . 248 for GREAT , 0 . 253 for TFICA , and 0 . 292 for the combined metric , indicating again that TFICA is adding an independent source of information for TF binding ., Using the TFICA disease annotations , we visualized the highest confidence transcription factor-disease associations ( see Materials and Methods ) in a regulatory network connecting 62 transcription factors to 253 human diseases ( Figure 5 ) ., As expected , substantial parts of this network reflect known biology ., For example , TFICA associates HNF4G with metabolic disorders , which corresponds to its KEGG annotation ., STAT3s role in fibrotic diseases is well-studied , as it is implicated in the proliferation of fibroblasts and excess ECM proteins 30 ., In total , the network visualization describes 491 relationships , 33 are known associations according GAD , OMIM , and GWAS Catalog and thus the remaining 458 are potentially novel transcription factor-disease relationships ., Transcription factors are connected with an average of 7 . 9 diseases and each disease was associated with an average of 1 . 9 transcription factors ., The high confidence associations visualized in Figure 5 and all of the significant ICA-derived associations are available in Table S4 ., Using our network analysis , we identified MEF2A as the factor with the highest association with Crohns disease ., MEF2A has been previously associated with cardiovascular disease 31 , but is not recognized to have a role in inflammatory bowel disease ., MEF2A was associated with Crohns disease through three modules: 69 , 262 , and 320 ., We validated this association using an independent expression dataset ( not used in the training set ) of 59 patients with Crohns disease and 42 controls 32 ., For each of these modules , the genes comprising the module showed significantly higher levels of differential expression between the two groups compared to genes not in one of the modules ( P\u200a=\u200a0 . 017 , 1 . 4e-06 , and 0 . 0084 , for modules 69 , 262 , and 320 , respectively ) ., We used permutation testing to correct for a potential bias towards higher differential expression for genes contained within functional modules , after which module 262 remained significant ( P\u200a=\u200a0 . 025 ) , which includes genes such as STAT4 , CCR5 , and SMAD3 ., We found that expression of MEF2A itself was significantly higher in Crohns disease patients ( Fig . 4A , P\u200a=\u200a0 . 0013 , Wilcoxon rank-sum test ) ., Additionally , among genes targeted by MEF2A , genes in module 262 also exhibited a higher level of differential expression among patients with Crohns disease ( P\u200a=\u200a0 . 0019 ) ., To investigate the role of module 262 and MEF2A in classification of Crohns disease , we projected the expression values in the Crohns dataset to generate an expression value of the module ( see Materials and Methods ) and found an overall higher expression of the module in patients with Crohns disease ( Fig . 4A , P\u200a=\u200a4 . 6e-09 , Wilcoxon rank-sum test ) ., We evaluated MEF2A and the module expression for their performance in a disease classifier using area under the receiver operating characteristic curve ( AUROC ) ., Both MEF2A and the aggregated expression of module 262 were significantly predictive ( P\u200a=\u200a0 . 0012 and P\u200a=\u200a4 . 5e-06 , logistic regression ) with AUROCs of 0 . 77 and 0 . 86 , respectively ( Fig . 4B ) ., Finally , we combined MEF2A expression and aggregated module expression into a single model and found that this combined statistic outperformed the other two classifiers ( F-test p\u200a=\u200a5 . 8e-11 , AUROC\u200a=\u200a0 . 90 , Fig . 4B ) ., We present a computationally efficient and conceptually simple method that is useful in linking transcription factors to their targets and to disease as well as derive several novel such relationships ., Using current approaches , such an analysis is challenging as TFs in the ENCODE dataset bind near an average of 6 , 050 genes ., Simple enrichment analysis on the full target set often does not reveal coherent functional groups ., Factors may exhibit multifaceted functional roles and target genes in very different cellular contexts , and when all of a TFs targets are grouped together , it becomes difficult to isolate these individual contexts ., Our method overlays data-driven functional module information – from a large compendium of human gene expression data – on top of TF binding data from ChIP-Seq ., We demonstrate that our method ( 1 ) significantly improves TF target identification , ( 2 ) accurately identifies the functional roles of factors both independently and in combination with another factor , and ( 3 ) discovers new disease associations through these functional modules ., We show that TFICA identifies targets that are significantly more functionally coherent than targets identified by naive ( peak-based ) methods ., Importantly , TFICA can identify these targets even in cases that lack strong support from ChIP-Seq binding data ( i . e . sites that are not among the strongest bound peaks ) ., We hypothesized that TFICA would be better able to identify targets that , despite lower binding levels , are biologically important ., Our “matched” analyses tests this hypothesis and we observe that TF-target functional annotation sharing ( Figure 2B ) and expression correlation ( Figure S5 ) is higher for TFICA targets than naively identified targets ., In fact , despite that stronger binding more tightly couples TF and target expression , the expression correlation among modules identified by TFICA are consistent with those of the “best” module ( genes with highest peak scores ) until the modules are associated with more than 40 factors ., Additionally , by linking TFs to established modules of gene expression we identify genes where binding of the factor is not observed , but instead , the TF is exerting indirect genetic control ., In these cases , we hypothesize that the TF may be controlling expression of a module through its direct targets , some of which may be in the module ., Our factor interaction analysis is useful for suggesting the functional roles of TF through guilt by association , particularly for poorly described factors ., For example , CHD2 is a helicase whose function remains to be fully understood ., In our TF-TF network , we found that CHD2 is connected with the cyclin CCNT2 ( Figure 3B ) , supporting the hypothesis that CHD2 plays a role in cell cycle 19 ., It is important to acknowledge that our method for identifying TF-TF interactions , which uses all of the shared modules between two transcription factors may miss those that are unique to particular biological contexts ., Future work will be required to model this type of interaction ., In spite of this limitation , we identify many known relationships and outperform traditional approaches ( Figure 3A ) ., Further , we were able to recapitulate many known TF-disease associations without any prior knowledge linking the factor to the disease ., For example , our factor-disease network ( Figure 5 ) links ER-α to arrhythmogenic right ventricular dysplasia , supporting recent findings that this protein may be used as a prognostic marker 24 ., STAT1 and STAT3 , which we also associate with arrhythmogenic right ventricular dysplasia , were recently found to be elevated in mice with sustained atrial fibrillation 27 ., In fact , we find too many known associations in this network to enumerate here ( refer to Table 2 for an annotated sampling of these associations , and Table S4 for the complete list ) ., Furthermore , we found that the number of diseases associated with a given transcription factor varied widely from just one to hundreds ., We hypothesize that this is related to the roles that a particular factor may play in different cellular contexts , with more general factors ( e . g . p300 , GR , and Pol2 ) associated with more diseases than more specific factors ., We tested this by examining the relationship between the number of diseases associated with a TF and two measures of functional diversity: ( 1 ) the number of targets found for that TF by ChIP-Seq and ( 2 ) the number of unique GO annotations ., In both cases , we found significant positive relationships ( Figure S8B–C ) ., In addition , our factor-disease network suggests a novel role for MEF2A in Crohns disease – an association that would not have been found using the naive method ( naive adj . p\u200a=\u200a1 ) and that we validated using an independent data set and analysis ( Figure 5A–B ) ., Our module analysis suggests MEF2A is promoting inflammatory response via module 262 , which includes STAT4 , CCR5 , and SMAD3 ., It is important to note , however , that this approach is dependent on the quality of the functional networks used ., Other methods for generating cohesive functional networks , including data-driven approaches ( e . g . PPI networks ) and knowledge-based ( e . g . functional annotations from ontologies ) , may complement the approach and improve performance ., Further , we derived our modules using a set of 9 , 395 gene expression experiments without regard to the particular context in which the experiment was performed ., Focusing this analysis using a contextually specific set of experiments ( e . g . only data focused on cardiovascular disease ) could provide further specificity to the disease associations that are derived ., Our work dissects transcription factor function by examining associations with specific gene modules derived from a large compendium of human expression experiments under a wide variety of conditions ., This approach is generally applicable in cases where the biological function can neither be described by a single gene nor the entire genome , but instead operates at an intermediate level – groups of genes or groups of functional pathways ., We demonstrate improved identification of TF targets and construct a regulatory network of human disease ., Finally , we find and validate a novel transcription factor-disease association ., We make three databases publicly available to the community: ( 1 ) a database of 5 , 002 transcription factor-module associations , ( 2 ) a database of 3 , 696 putative transcription factor interacting pairs , and ( 3 ) a database of 7 , 808 transcription factor-disease relationships ., These resources should further enable researchers to explore TF interactions as well as their roles in human disease ., We obtained transcription factor binding data from UCSC , which included 2 , 750 , 490 reproducible binding sites from the ENCODE project 7 , 8 and 41 , 972 gene annotations from RefGene ( build hg19 ) ., 423 gene expression modules ( a . k . a . “components” ) determined previously from independent component analysis ( ICA ) are available at https://simtk . org/home/fcanalysis 14 ., These modules were derived from a compendium of human gene expression data downloaded from the Gene Expression Omnibus ( GEO ) ., All 9 , 395 Affymetrix Human Genome U133 Plus 2 . 0 array deposited in GEO as of May 28 , 2008 were downloaded ., The ICA analysis was performed and published previously 14 ., We obtained gene-disease associations from the Gene Association Database 20 and filtered for positive genome-wide and curated associations , as well as diseases with greater than five genes ., We then used the NCBO Annotator 33 service to map the disease terms to terms in the Disease Ontology , resulting in 34 , 392 distinct associations ., 4 , 267 and 5 , 279 validation associations were downloaded from the NHGRI GWAS catalog 34 ( accessed on 3/31/12 ) and OMIM ( http://omim . org ) , respectively ., For both of these datasets , we mapped local disease terms to standardized terms in the Disease Ontology using the annotator , resulting in 1 , 842 and 9 , 866 annotations ( of which 35 and 46 map to the trans | Introduction, Results, Discussion, Materials and Methods | Transcription factors ( TFs ) are fundamental controllers of cellular regulation that function in a complex and combinatorial manner ., Accurate identification of a transcription factors targets is essential to understanding the role that factors play in disease biology ., However , due to a high false positive rate , identifying coherent functional target sets is difficult ., We have created an improved mapping of targets by integrating ChIP-Seq data with 423 functional modules derived from 9 , 395 human expression experiments ., We identified 5 , 002 TF-module relationships , significantly improved TF target prediction , and found 30 high-confidence TF-TF associations , of which 14 are known ., Importantly , we also connected TFs to diseases through these functional modules and identified 3 , 859 significant TF-disease relationships ., As an example , we found a link between MEF2A and Crohns disease , which we validated in an independent expression dataset ., These results show the power of combining expression data and ChIP-Seq data to remove noise and better extract the associations between TFs , functional modules , and disease . | Transcription factors ( TFs ) are crucial to the precise regulation of many cellular processes and thus , are responsible for many human phenotypes and diseases ., Now that the ENCODE project has mapped hundreds of TFs to their genomic binding locations , extracting functional biological signals is the next step in understanding their role in disease ., In this paper , we present a novel approach to identifying TF targets and use these targets to find regulatory relationships between TFs and diseases ., We present a large open dataset of putative TF-TF interactions and TF-disease associations which includes known connections as well as novel ones ., We validate the association of one of our novel TF-disease associations , MEF2A and Crohns disease , suggesting that our approach generates testable disease association hypotheses ., Integrating these datasets will be crucial for understanding phenotypes and complex diseases . | genome expression analysis, functional genomics, regulatory networks, biology, genomics, computational biology, genomic medicine | null |
journal.pgen.1003351 | 2,013 | Spreading of a Prion Domain from Cell-to-Cell by Vesicular Transport in Caenorhabditis elegans | Transmissible spongiform encephalopathies ( TSEs ) or prion diseases are fatal , age-related , and infectious neurodegenerative disorders that affect humans ( e . g . , Creutzfeldt-Jakob disease ) and animals ( e . g . , scrapie in sheep and bovine spongiform encephalopathy in cattle ) 1 ., At the molecular level , prions propagate by recruitment and conversion of the soluble α-helix-rich cellular PrPC into toxic aggregates of the pathological β-sheet-rich PrPSc isoform , via a mechanism described as seeded or nucleated polymerization 2–5 ., The TSE agent is also infectious at the cellular level , where it transmits from cell-to-cell and infects naïve cells , both from within and outside the central nervous system 6 , 7 ., In yeast , prions can function as heritable epigenetic factors 8–12 upon forming an alternative self-propagating β-sheet-rich state from a soluble α-helix-rich fold ., Yeast and mammalian prion determinants , however , do not share similarities in amino acid sequence , function , or subcellular localization ., Yeast prions are naturally propagated within the cytosol from mother to daughter cells during cell division and require the disaggregase activity of the molecular chaperone Hsp104 to generate seeds and ensure dissemination 13 ., In contrast , cell-surface localized mammalian prions are transmitted from cell-to-cell in terminally differentiated non-dividing cells ., Sup35 , like the majority of yeast prion proteins , contains a glutamine/asparagine ( Q/N ) -rich domain that confers the prion phenotype 14 ., Although the mammalian prion protein PrP lacks this domain , other neurodegenerative disease proteins such as FUS ( Fused in Sarcoma ) and TDP-43 ( TAR DNA-binding protein 43 ) have been shown to contain Q/N-rich prion-like domains 15–17 ., There is increasing evidence that proteins closely associated with the neurodegenerative diseases Alzheimers , Parkinsons , Huntingtons , frontotemporal lobar degeneration ( FTLD ) and amyotrophic lateral sclerosis ( ALS ) , exhibit prion-like properties 18–20 ., Amyloid fibril assembly in general follows a nucleated polymerization reaction in vitro 9 , and the addition of preformed fibrils or pathological brain extract seeds the polymerization of the corresponding monomeric protein in cell culture models or following injection into healthy mouse brains 24–31 ., Furthermore , many proteins that form aggregates and fibrils exhibit cell non-autonomous effects and might spread among tissues within an organism 32–37 ., The cellular processes and mechanisms that underlie cell-to-cell transmission of toxic protein species remain elusive in the current animal models that employ tissue culture cells and mice to investigate prion biology ., The nematode Caenorhabditis elegans has been widely used as a model system to investigate the biology of protein misfolding and toxicity 38–43 , and has the advantage of transparent tissue types including muscle , intestinal , and neuronal tissue ., We aimed to establish a new prion model system in this metazoan to examine the mechanisms of propagation of protein misfolding across tissues in a living organism ., Since there are no known prions in C . elegans , and we wanted to avoid potential complications of infectious mammalian prions , we used the well-characterized cytosolic yeast prion protein Sup35 8 ., Here , we show that a cytosolic prion domain , NM , is highly toxic and can spread among tissues within the animal ., The cell non-autonomous organismal toxicity of Sup35NM was associated with the accumulation of autophagy-derived vesicles , disruption of mitochondrial integrity , and the dynamic movement of the prion domain protein between tissues via autolysosomal vesicles ., Three versions of Sup35NM , corresponding to the full-length wild-type domain , NM , a deletion of the oligorepeat region ( RΔ2-5 ) , and an expansion of the oligorepeats ( R2E2 ) , were fused to YFP ( yellow fluorescent protein ) ( Figure 1A ) and expressed under the control of the body wall muscle ( BWM ) cell specific ( m ) promoter unc-54p ., These NM constructs were selected based on previous observations that deletion of four of the five oligorepeats of the prion domain ( RΔ2-5 ) leads to a strong decrease in prion induction , while expansion of this region ( R2E2 ) significantly increases spontaneous prion formation 44 ., In C . elegans lines expressing approximately similar levels of the transgenes ( Figure 1C ) , NMm::YFP aggregates appeared in early embryonic stages of development and persisted through all larval stages into adulthood ( Figure 1B ) ., The appearance of aggregates was strictly related to the length of the oligorepeats such that R2E2 formed aggregates more rapidly and to higher levels than NM , while deletion of the oligorepeats in RΔ2-5 resulted in expression of a mostly soluble and diffuse protein ( Figure 1B , 1D ) ., The fluorescent foci in NMm::YFP and R2E2m::YFP coincided with higher levels of detergent insoluble protein relative to RΔ2-5m::YFP ( Figure 1D ) ., To further characterize the biochemical and biophysical properties of the NM aggregates , we employed the dynamic imaging method of fluorescence recovery after photobleaching ( FRAP ) ., The diffuse fluorescence observed in RΔ2-5m::YFP expressing animals was shown to correspond to highly mobile protein species by FRAP analysis , in addition to the infrequent appearance of foci that were too small to be assessed by FRAP ( Figure 2A ) ., In contrast , examination of NMm::YFP and R2E2m::YFP foci at high magnification revealed highly diverse shapes and sizes that can be described as long fibril-like species ( ∼10 µm ) , large ( ∼2 µm ) round spherical structures , and small ( ∼0 . 1 µm ) foci ( Figure 2B , 2C ) ., These foci did not exhibit any obvious patterns among the BWM cells and were randomly distributed ., Moreover , each progeny descending from a single hermaphrodite exhibited a unique pattern of R2E2m::YFP foci ( Figure 2C ) suggesting an influence of the individual cellular environment on aggregate phenotypes ., FRAP analysis on animals expressing NM and R2E2 ( Figure 2A ) revealed foci ranging from immobile aggregates that exhibited no FRAP recovery to foci that rapidly recovered fluorescence and thus were comprised of mixed populations of slowly mobile protein species ., These two biophysical states of prion domain aggregates were closely aligned with the distinct visual morphologies , in that every fibril-like aggregate tested was comprised of immobile species , and that spherical aggregates detected in both R2E2 and NM animals corresponded to mobile aggregates that showed recovery following photobleaching ( Figure 2A ) ., R2E2m::YFP animals exhibited a severe reduction in motility relative to wild-type N2 or RΔ2-5m::YFP animals ( Figure 3A; Video S1 , S2 , S3 ) that was associated with a disruption of muscle ultrastructure revealed by rhodamine-phalloidin staining of myofilaments ( Figure 3B ) ., Moreover , nearly all of the R2E2m::YFP and NMm::YFP adults exhibited developmental delay and reduced fecundity , with R2E2m::RFP adults being often sterile ( Ste ) ( Figure 3C , 3D ) ., Whereas adult N2 wild-type and RΔ2-5 animals lay approximately 16 eggs within a 2 . 5 hour period , NM and R2E2 animals laid 8 . 5 and 4 eggs , respectively ( Figure 3C ) ., Furthermore , only 8% of R2E2m::YFP embryos and 1% of NMm::YFP embryos attained adulthood over a three day period at 20°C , relative to >93% achieving adulthood for wild-type N2 or RΔ2-5m::YFP embryos ( Figure 3C ) ., The slightly higher fraction of adult R2E2 animals detected after 72 hours is due to a more severe egg laying defect ( Egl ) of these animals ., R2E2 animals often retained their eggs due to dysfunction of the vulva muscle leading to embryos being laid at later time points of development ., Consequently , eggs laid by R2E2 animals tended to be older than corresponding NM , RΔ2-5 , and wild-type N2 embryos that are deposited at the same time ., NM animals exhibited a more severe embryonic lethal phenotype ( Emb ) than R2E2 , while the latter animals exhibiting increased sterility ( Ste ) and producing fewer total progeny ( Figure 3C ) ., Animals that reached the L4 state of development after 72 hours became adult animals on the next day , whereas younger larvae were more likely to arrest in development ( Figure 3D , Video S1 , S2 , S3 , data not shown ) ., In summary , while the populations of NM and R2E2 animals differed in their distribution among developmental states after 72 hours , expression of the highly aggregation-prone R2E2 resulted in a more severe toxic phenotype than NM ( Figure 3C , 3D; Video S1 , S2 , S3 ) ., Another characteristic of R2E2m::YFP expressing animals was a plethora of morphological defects that included reduced size ( Sma ) , vacuolation ( Vac ) , defective molting ( Mlt ) , clear appearance ( Clr ) , and disrupted gonadal and intestinal morphology ( Figure 3E , and data not shown ) ., Such defects affecting other tissues were observed to a lesser extent in NMm::YFP animals , and not detected in RΔ2-5m::YFP lines or in C . elegans lines expressing other aggregation-prone proteins 38 , 39 , 41–43 ( data not shown ) ., The NM-dependent cellular defects were examined in more detail using transmission electron microscopy ( TEM ) ., Compared to wild-type N2 animals , the muscle cells of R2E2 expressing animals exhibited disrupted sarcomeres , fragmented mitochondria containing a drastically reduced number of cristae , and a complex array of double and single membrane bound organelles ( Figure 4A ) ., These vesicular structures are a hallmark of autophagy ., Surprisingly , the cellular pathology observed in R2E2 expressing animals was not restricted to BWM cells and was also observed in other tissues in which R2E2 was not expressed ., For example , intestinal cells that did not express R2E2 exhibited mitochondrial fragmentation with loss of cristae , and reduced levels of yolk and lipid droplets ( Figure 4B ) ., These studies show that the number of oligorepeats in the prion domain directs the toxicity that results in multiple organismal phenotypes that extend beyond the primary tissue of NM expression ., To examine whether the induction of autophagy is a secondary cellular response due to damage of essential components like mitochondria , or if the prion domain is directly targeted by the autophagy-lysosome pathway ( ALP ) , we employed C . elegans lines expressing markers of specific membraneous organelles ., As the available markers for C . elegans are tagged with green fluorescent protein ( GFP ) , we generated a C . elegans line expressing R2E2 tagged with red fluorescent protein ( RFP ) under the control of the myo-3 promoter for BWM cell specific expression ( R2E2m::RFP ) ., LGG-1::GFP transgenic animals that express the ortholog of the autophagosome marker LC3 ( in mammals ) or ATG8 ( in yeast ) were used to monitor autophagic vesicles 45 ., In R2E2; LGG-1 transgenic lines , we observed co-localization of a subset of R2E2m::RFP foci with autophagosomes ( Figure 5A ) ., We also detected co-localization of R2E2m::RFP foci with RAB-7 positive late endosomes and specifically with LMP-1 positive lysosomes ( Figure 5C , 5D ) ., The majority of these lysosomes exhibited an unusual tubular shape ( Figure 5D ) ., Co-localization was not observed with RAB-5 ( early endosomes ) ( Figure 5B ) , indicating that the R2E2-containing vesicles were derived from the autophagy pathway rather than from endocytosis ., Our studies do not distinguish whether the R2E2m::RFP that co-localizes with vesicular structures corresponds to specific classes of aggregate species described before , as these vesicles have been excluded from FRAP analysis due to their small size ( see materials and methods for more details ) ., These data , together with the TEM analysis , suggest that the prion domain is a target of quality-control autophagy and is transferred from autophagosomes to RAB-7 positive amphisomes and LMP-1 positive autolysosomes , respectively ., Another striking characteristic of the tubular vesicles containing R2E2m::RFP was their dynamic movement within and between muscle cells , monitored by live-cell time-lapse imaging ( Video S4 , S5; Figure 6B ) ., In particular , the over-expression of RAB-5 enhanced ( and facilitated by visualizing single muscle cells ) the detection of cell-to-cell transmission of RFP-positive vesicles between BWM cell quadrants ( Video S5 , Figure 6B ) ., These observations are consistent with previous findings that RAB-5 over-expression increases autophagy 46 ., The intercellular transport of R2E2-containing vesicles was unexpected as C . elegans body wall muscles are composed of individual mononucleated cells that are connected through gap junctions to allow electrical coupling for coordinated body movement 47 ( Figure 6A ) ., No dye coupling has been observed between single muscle cells , implying that there is no unregulated transfer of cytosolic proteins under normal physiological conditions 47 ., This leads us to propose that R2E2 is actively transported by tubular vesicles from cell-to-cell ., As mentioned before , these vesicles are different from the foci described in Figure 2 ., Neither the spherical ( mobile in FRAP analysis ) nor the fibril-like ( immobile in FRAP analysis ) aggregates are moving within or between cells ., Only the small tubular vesicles are getting transmitted and we do not know the conformational state of R2E2 protein within these vesicles ., Nevertheless , misfolding and aggregation is central to the toxicity phenotype , as RΔ2-5m::YFP , which does not form these aggregates , exhibits neither a cell autonomous nor cell non-autonomous toxicity ., The moving , tubular-shaped vesicles were only detected in R2E2m::RFP animals , but not observed with the corresponding proteins tagged with YFP ., In contrast , the diverse aggregate species and other small vesicular structures ( neither tubular nor moving ) were visible with both YFP and RFP ( compare Figure S1A and Figure 1A , Figure 2B and 2C ) ., In transgenic animals expressing only the RFP fluorescent tag in BWM cells ( RFPm ) , no movement of RFP between cells was observed ( Figure 6C , Figure S1B , data not shown ) ., This apparent discrepancy with the fluorescent tags can be explained by RFP being more stable in acidic environments whereas YFP is pH sensitive 48 , indicating that these vesicles might exhibit a low lumenal pH that could explain the lack of similar fluorescent structures in R2E2m::YFP expressing animals ., This speculation is supported by our results that the moving tubular vesicles co-localize with LMP-1::GFP , but not with LGG-1::GFP ( compare Figure 5A and 5D ) ., Indeed , staining of R2E2m::YFP animals with an anti-GFP antibody by indirect immunofluorescence revealed tubular structures in addition to foci visible with YFP fluorescence ( Figure S2 ) ., This further supports our conclusion that acidified lysosomal vesicles containing the prion domain are transported from cell-to-cell ., Muscle cell-expressed R2E2 was also detected in vesicles of coelomocytes and the intestine ( Figure 6D , Figure S3 ) ., Both , the intestine and coelomocytes , have been shown to endocytose molecules from the body cavity ( pseudocoelom ) 49 , 50 , suggesting that the tubular vesicles containing R2E2 could be released from BWM cells into the pseudocoelom and taken up by endocytosis from surrounding coelomocytes or intestinal cells ., While the uptake of proteins from the pseudocoelom into coelomocytes and the intestine is not specific 49 , 50 , the amount of R2E2m::RFP that accumulates in these tissues is much more pronounced than for RFPm ( compare Figure 6D and 6E ) ., These results suggest that R2E2m::RFP is actively released from muscle cells into the pseudocoelom ., To examine the specificity of tissue movement of R2E2 , we expressed RFP-tagged R2E2 in intestinal cells and monitored the dynamics of R2E2i::RFP-containing vesicles ( Figure S4 ) ., Movement of R2E2 was observed by real-time imaging within intestinal cells ( Video S6 ) , and from the intestine into adjacent non-expressing muscle cells ( Figure 6F , Figure S5 , Video S7 ) , thus confirming the spreading of the aggregation-prone prion domain across tissues ., Taken together , these observations reveal that R2E2m::RFP accumulates in tubular vesicles of autolysosomal origin that spread from expressing cells to non-expressing tissues in C . elegans ., Furthermore , R2E2 seems to spread by two different pathways , either by a direct cell-to-cell transport of lysosomes , or through release into and endocytic uptake from the pseudocoelom ( Figure 6B , 6D , 6F , Video S5 and S7 ) ., We next examined whether the prion domain induces aggregation of its soluble isoform in C . elegans ., Such a self-templating or seeding activity forms the basis of amyloid infectivity 9 , 22 ., To address this , we took advantage of the different aggregation properties of the prion domain constructs and used the non-aggregating RΔ2-5m::YFP as a folding sensor ., The seeding model posits a direct interaction of newly forming with preexisting aggregates , which in part is sequence-specific 51 ., To examine this , we introduced in vitro generated recombinant fibrils by microinjection ( Figure S7 ) , into intestinal cells expressing RΔ2-5 , as muscle cells were too small for microinjection ., These studies were based on previous in vitro and cell culture observations that addition of preformed fibrils induces aggregation of the corresponding soluble NM in a sequence-specific manner 9 , 44 , 52 ., We monitored the aggregation state of the RΔ2-5 folding sensor expressed under an intestine-specific ( i ) promoter ( vha-6p ) ., RΔ2-5 and NM constructs exhibited similar patterns of aggregation in the intestine as in muscle cells ( Figure S6 ) ., Analogous to the biophysical properties exhibited in BWM cells , RΔ2-5i::GFP is soluble in intestinal cells ( Figure S6; Figure 7A , 7E ) ., Introduction of recombinant Sup35NM fibrils into intestinal cells resulted in the conversion of RΔ2-5 from a soluble to an aggregated state ( Figure 7A , 7B , 7C , 7D ) that did not co-localize with the injected Sup35NM fibrils ., To address the sequence specificity of these effects , RΔ2-5 animals were also injected with recombinant fibrils of the asparagine-rich yeast prion protein Ure2p with high alpha-helical content 53 , 54 , or β-sheet rich amyloid Aβ1-42 , respectively ( Figure 7E ) ., No aggregation of RΔ2-5 was observed upon injection of either protein ., To test whether cross-seeding occurs when both proteins are co-expressed in C . elegans tissues , we crossed RΔ2-5m::YFP with R2E2m::RFP animals ., Despite being impaired for spontaneous aggregation , RΔ2-5m::YFP readily formed aggregate species that exhibited slow exchange in FRAP when co-expressed with R2E2m::RFP in BWM cells ( Figure 8A , 8B , 8C ) ., RΔ2-5m::YFP aggregates , however , only rarely co-localized with R2E2m::RFP foci ( Figure 8B ) , consistent with observations from the injection experiments ( Figure 7B , 7C ) ., The RΔ2-5 sensor was further employed to test whether protein misfolding spreads from R2E2-expressing muscle cells to the intestine ., Indeed , aggregation of RΔ2-5i::GFP increased when R2E2m::RFP and RΔ2-5i::GFP were co-expressed ( Figure 8E , 8F ) ., The absence of co-localization of RΔ2-5 and R2E2 foci , even when co-expressed ( Figure 8B ) , indicates that aggregation of RΔ2-5 could be due to the global disruption of the folding environment , as seen in tissues co-expressing aggregates of polyglutamine and temperature sensitive mutant proteins 55 , rather than from cross-seeding , which would imply co-aggregation of both proteins into heterologous aggregates 51 ., Indeed , expression of R2E2 in muscle cells accelerated the age-dependent aggregation of an intestinal polyglutamine ( polyQ ) folding sensor ( Q44i::YFP ) 56 ( Figure S8 ) ., Taken together , these results show that R2E2m::RFP aggregates have multiple effects by seeding homologous soluble proteins in a sequence-specific manner and causing an imbalance of organismal proteostasis ., We have developed a metazoan prion model and examined the properties of a Q/N-rich prion domain in non-dividing terminally differentiated cells using C . elegans ., A summary model describing the different aggregate species , vesicular structures and phenotypes observed in the C . elegans prion model , is shown in Figure 9 ., As the mechanism of prion propagation differs between unicellular eukaryotes and metazoans , it was unclear whether the prion propensities of Q/N-rich domains are universal and can adapt to different biological systems of cell-to-cell transmission ., Spreading of the prion domain from an initial site of expression via autolysosomal vesicles occurs through actively regulated cellular processes , involving a direct transport from cell-to-cell and the release and endocytic uptake of these vesicles from the body cavity ., This differs substantially from the propagation of PSI+ in yeast that involves transfer of cytosolic NM propagons from mother to daughter cells during cell divison , that neither requires uptake into membraneous compartments nor active transport ., Rather , the transmission of NM between cells and tissues in C . elegans is reminiscent of mammalian PrPSc propagation between post-mitotic neurons ., Exosomes 57 and tunneling nanotubes 58 have been suggested as possible routes for cell-to-cell transmission of PrPSc ., Either way , cytosolic content also gets transmitted , suggesting that cytosolic and membrane localized prion-like proteins might share some mechanistic aspects of transmission 59 ., Indeed , there is now growing evidence that other disease-associated cytosolic protein aggregates can spread from cell-to-cell 19 , 20 , 26 , 34 , 35 ., The spreading of the prion domain described here in C . elegans will allow us to compare the relative potential of tissue transmission with other aggregation-prone amyloidogenic proteins in our model system ., It remains to be established if all major disease-associated proteins can spread from cell-to-cell themselves in a similar fashion like NM ., Alternatively , prion-like domains might have implications in the spread of pathology throughout the nervous system by allowing a subset of modifiers like FUS and TDP-43 to transmit between cells , which then cause the subsequent aggregation of other disease-linked proteins ., Although motility defects are often associated with the expression of protein aggregates in C . elegans muscle cells 38 , 39 , 41–43 , the expression of NM was unusually toxic compared to the expression of other disease-associated aggregation-prone proteins 38–43 ., Aggregation and toxicity of NM were dependent on the oligopeptide repeats ., Likewise , in yeast and mammals , the oligorepeats affect spontaneous prion induction and disease prevalence , respectively 44 , 60 , 61 ., In yeast and infected tissue culture cells , prions often elicit no toxicity , suggesting that only non-toxic rapidly replicating variants are selected upon infection in these systems 62 ., The unc-54 promoter used to express NM becomes active post-mitotically in 81 of 95 body wall muscle cells 63 , 64 ., The toxicity in C . elegans could therefore reflect the vulnerability of terminally differentiated non-dividing cells ., Autophagy is important for protein quality control and homeostasis in non-dividing neuronal cells 65 , 66 , consequently , autophagic failure has been implicated in prion diseases and other neurodegenerative disorders 67–69 ., While activation of autophagy is beneficial to promote the clearance of disease-associated proteins 70–75 , the chronic induction of autophagy could have deleterious consequences and may be insufficient to suppress toxicity 76–78 , in particular if lysosomal function is already compromised during aging or by the chronic expression of mutant proteins 77–79 ., In line with this , our preliminary results revealed that blocking autophagy by RNAi , to inhibit prion transmission , has only marginal effects to ameliorate NM toxicity in BWM cells ( as measured by motility assays , data not shown ) , indicating that the autophagy-lysosomal pathway has a dual role and also reduces the load of misfolded proteins ., Future studies using genome-wide RNAi screens will identify the cellular pathways that improve fitness in these animals ., One of the most striking consequences attributed to expression of the prion domain in C . elegans was mitochondrial fragmentation and loss of cristae ., An equilibrium of steady fission and fusion events is critical for mitochondrial structure and function , and disruption of this homeostasis has been observed in disease and aging 80 ., Intriguingly , a collapse of mitochondrial function was also observed in lysosomal storage disorders associated with impaired lysosomal degradation 81–83 , and has been proposed to be a common secondary and final mediator of cell death in several diseases associated with autophagic failure and lysosomal dysfunction 81–83 ., It remains to be determined whether related mechanisms are associated with the disruption of mitochondrial ultrastructure observed here for the C . elegans prion model ., There is accumulating evidence that lysosomes have additional roles to their conventional function as digestive organelles ., Lysosomes constitute the exosomes of nonsecretory cells 84 , are exocytosed during plasma membrane repair 85 , and were shown to be transferred via tunneling nanotubes from endothelial progenitor cells to rescue prematurely senescent endothelia 86 ., Our results reveal the involvement of lysosomes in the cell-to-cell transmission of cytosolic aggregation-prone proteins ., Of note , the exocytosis or transfer of lysosomes may represent a specific cellular response to the prion domain as a cargo , because non mitotic aging tissues fail to secrete indigestible lysosomal content , which leads to the characteristic accumulation of lipofuscin 87 ., It is tempting to speculate that proteins with prion domains might trigger a specific cellular response that leads to the release of LMP-1 positive vesicles ., Aggregation of NM in C . elegans occurs spontaneously upon its over-expression , in contrast to observations in bacterial and mammalian models 18 , 88 , 89 ., In yeast , the induction of PSI+ is dependent on the co-existence of PIN+ or other compatible aggregation-prone proteins 90–92 ., Perhaps similar factors such as endogenous Q/N-rich proteins are expressed in C . elegans that can act as PIN+ 93 ., The injection of preformed fibrils or co-expression of aggregation-prone variants seeds the polymerization of the corresponding monomeric protein 12 , 22 , 24 , 28 , 29 by a reaction known as nucleated or seeded polymerization 9 , 22 ., Only Sup35NM fibrils were able to cross-seed RΔ2-5 to form aggregates , whereas injection of fibrillar Abeta1-42 or Ure2p failed to do so , which suggests that seeding of RΔ2-5 is sequence-specific ., However , Sup35NM fibrils or R2E2 aggregates did not co-localize with RΔ2-5 foci ., The absence of co-aggregation might be due to conformational variations resulting from sequence differences within the NM oligorepeats 94 , 95 ., While the different prion domain variants might initially form heterologous seeds below the resolution of our imaging approaches , the preferred coalescence of molecules that have the same conformation might lead to distinct aggregates 51 ., Alternatively , the ability to induce polyQ aggregation in a cell non-autonomous manner , suggests that expression of the aggregation-prone prion domain causes a global disruption of cellular proteostasis , and subsequent misfolding of unrelated metastable proteins , perhaps by titrating chaperones and other anti-aggregation factors 55 , 90 ., Most likely , misfolding of RΔ2-5 upon co-expression of R2E2 in the same or neighboring tissue results from a combinatory effect of sequence-specific cross-seeding together with an overload of the cellular folding capacity ., Under these chronic proteotoxic stress conditions , one misfolded protein can accelerate aggregate formation of another aggregation-prone protein independent of protein sequence homology 41 , 55 ., In summary , this study provides new insights on the intrinsic properties of Q/N-rich prion domains in metazoans ., Although the yeast prion domain NM is not a disease relevant peptide , this novel genetic C . elegans prion model can elucidate cellular pathways underlying the prion-like propagation of conformational changes in proteins between cells and tissues of multicellular organisms in health and disease ., Sup35NM constructs were amplified from the yeast expression vector p316CUP1-3SGFPSG 44 containing either the full-length NM , NM with a deletion of oligorepeats # 2-5 ( aa 56-93 ) ( RΔ2-5 ) , or NM with a two-times extended oligorepeat # 2 ( QGGYQQYNP ) ( R2E2 ) 44 , by PCR standard methods ., Insertion of appropriate restriction sites allowed cloning of the PCR amplicons into pPD30 . 38 39 ., This vector contains the promoter and enhancer elements from the unc-54 gene 96 , as well as EYFP from the vector pEYFP-N1 ( Clontech ) 39 ., For constructing myo-3p::sup35 ( r2e2 ) ::rfp , myo-3p::rfp , vha-6p::sup35 ( rΔ2-5 ) ::gfp , vha-6p::sup35 ( nm ) ::gfp , vha-6p::sup35 ( r2e5 ) ::rfp , unc-54p::cfp::rab-5 , and unc-54p::lmp-1::gfp expression vectors , the MultiSite Gateway Three-Fragment Vector Construction Kit ( Invitrogen ) was used ., NM constructs were amplified from the pPD30 . 38 vectors using appropriate oligonucleotides for gateway cloning and inserted into the pDONR 221 entry vector by recombination ., Likewise , the lmp-1 coding sequence was amplified from a N2 cDNA sample and inserted into the pDONR 221 entry vector ., The plasmid pCZGY#3 ( =\u200apDONR 201_rab-5 ) was a kind gift from Dr . Yishi Jin ., Entry vectors pDONR P4-P1R containing myo-3 ( approx . 2 . 4 kb upstream of the myo-3 gene ) , vha-6 ( approx . 1 . 2 kb upstream of the vha-6 gene ) , or unc-54 ( approx . 1 kb upstream of the unc-54 gene ) promoter region and pDONR P2R-P3 coding for the C-terminal monomeric RFP or GFP tag , were generated accordingly ., The N-terminal CFP was cloned between the unc-54 promoter and rab-5 using appropriate restriction sites ., For co-localization , CFP::Rab-5 was false-colored green ., All pDONR P2R-P3 entry vectors also contained the unc-54 3′UTR ., Promoters , genes of interest and fluorescent tags were finally inserted into the destination vector pDEST R4-R3 in an in vitro recombination reaction ., Wild-type ( N2 , Bristol ) and transgenic worms were handled using standard methods 97 ., If not otherwise indicated , nematodes were grown on NGM plates seeded with the E . coli strain OP50 at 20°C ., The strains NP1129 cdIs131cc1p::gfp::rab-5+unc-119 ( + ) +myo-2::gfp , NP871 cdIs66cc1p::gfp::rab-7+unc-119 ( + ) +myo-2::gfp , NP744 cdIs39cc1p::gfp::rme-1 ( 271alpha1 ) +unc-119 ( + ) +myo-2::gfp , RT258 pwIs50lmp-1p::lmp-1::gfp+Cb-unc-119 ( + ) , and DA2123 adIs2122lgg-1::GFP + rol-6 ( su1006 ) were ordered from the Caenorhabditis Gene Center ( CGC ) ., The strain FY777 lin-15 ( n765ts ) ; grEx170Pmyo-3::gfp::rab-7; lin-15 ( + ) was a kind gift of Dr . Bruce Bamber ., The following strains were generated for this study using germline transformation by microinjection: AM801 rmIs319unc-54p::sup35 ( rΔ2-5 ) ::yfp , AM803 rmIs321unc-54p::sup35 ( nm ) ::yfp , AM806 rmIs324unc-54p::sup35 ( r2e2 ) ::yfp , AM815 rmIs323myo-3p::sup35 ( r2e2 ) ::rfp , AM809 rmEx319vha-6p::sup35 ( rΔ2-5 ) ::gfp+myo-2p::mcherry , AM823 rmEx326vha-6p::sup35 ( rΔ2-5 ) ::gfp , AMf814 rmIs326vha-6p::sup35 ( nm ) ::gfp+myo-2p::mcherry , AM883 rmEx338myo-3p::rfp::rfp , AM887 rmEx339unc-54p::cfp::rab-5 , AM890 rmEx340unc-54p::lmp-1::gfp ., Transgenic lines carrying extrachromosomal arrays were generated by microinjection of the above-mentioned plasmids into N2 wild-type an | Introduction, Results, Discussion, Materials and Methods | Prion proteins can adopt self-propagating alternative conformations that account for the infectious nature of transmissible spongiform encephalopathies ( TSEs ) and the epigenetic inheritance of certain traits in yeast ., Recent evidence suggests a similar propagation of misfolded proteins in the spreading of pathology of neurodegenerative diseases including Alzheimers or Parkinsons disease ., Currently there is only a limited number of animal model systems available to study the mechanisms that underlie the cell-to-cell transmission of aggregation-prone proteins ., Here , we have established a new metazoan model in Caenorhabditis elegans expressing the prion domain NM of the cytosolic yeast prion protein Sup35 , in which aggregation and toxicity are dependent upon the length of oligopeptide repeats in the glutamine/asparagine ( Q/N ) -rich N-terminus ., NM forms multiple classes of highly toxic aggregate species and co-localizes to autophagy-related vesicles that transport the prion domain from the site of expression to adjacent tissues ., This is associated with a profound cell autonomous and cell non-autonomous disruption of mitochondrial integrity , embryonic and larval arrest , developmental delay , widespread tissue defects , and loss of organismal proteostasis ., Our results reveal that the Sup35 prion domain exhibits prion-like properties when expressed in the multicellular organism C . elegans and adapts to different requirements for propagation that involve the autophagy-lysosome pathway to transmit cytosolic aggregation-prone proteins between tissues . | Alzheimers , Parkinsons , Huntingtons , frontotemporal lobar degeneration ( FTLD ) , amyotrophic lateral sclerosis ( ALS ) , and prion diseases are all age-related , fatal neurodegenerative disorders ., Hallmarks of these diseases include the expression of toxic protein species ., The ability to spread and infect naive cells was thought to be limited to prions but has recently been observed for other disease-linked protein aggregates in tissue culture cells and transgenic mice ., The underlying cellular pathways of this cell-to-cell transmission , however , remain elusive ., We have developed a new prion model in the roundworm Caenorhabditis elegans and show that the appearance of aggregate species is associated with cellular toxicity , not only in the expressing cell but as well as in adjacent tissues ., We monitored in real time the spreading of prion domains by autophagy-derived lysosomal vesicles from cell-to-cell ., Given that autophagy and lysosomal degradation have a role in several neurodegenerative diseases , this cellular pathway might be the basis of amyloid infectivity in general . | medicine, biochemistry, infectious diseases, model organisms, neurological disorders, neurology, genetics, biology, molecular cell biology, neuroscience, genetics and genomics | null |
journal.pgen.1006501 | 2,016 | Enrichment of Targetable Mutations in the Relapsed Neuroblastoma Genome | Neuroblastoma is a cancer typically affecting young children arising from the developing sympathetic nervous system , but can occasionally occur in adolescents and adults 1 ., Over half of patients have widely metastatic disease at diagnosis where survival rates are less than 50% despite intensive therapeutic regimens including chemotherapy , radiation therapy and immunotherapy 2 ., There is no standard therapeutic approach for relapsed disease 3 ., Recent next generation sequencing ( NGS ) efforts of matched neuroblastoma samples collected at diagnosis and constitutional DNAs in 373 unique subjects across four studies has clearly documented a relatively low somatic mutation rate in the protein coding portion of the genome 4–7 , challenging the concept of targeting oncogenic drivers with newly developed therapeutics ., The data in neuroblastoma appears to be reflective of pediatric cancers in general 8 ., However , recent studies of diagnostic-relapse-normal DNA “trios” from a limited number of neuroblastoma cases has shown that the mutation rate is much higher after exposure to genotoxic chemoradiotherapy , and that there may be an enrichment of previously subclonal mutations in pathways known to be therapeutically targetable in other diseases 9–11 ., To better understand and characterize the landscape of potentially actionable mutations at both diagnosis and relapse , we analyzed targeted next-generation sequencing data for 151 neuroblastomas from 138 patients that were profiled either at diagnosis , in the midst of therapy , and/or at disease relapse ., Our primary aim was to retrospectively determine the frequency by which a therapeutically relevant lesion was discovered at these time points and to infer if the biopsy procedure followed by DNA sequencing could provide the potential for patient benefit ., We collected sequencing data and clinical information of neuroblastoma patients whose tumor biopsies had been sent to Foundation Medicine for molecular profiling ( see Methods for sample processing details ) ., The only inclusion criterion was the availability of high-quality sequencing data , and we did not filter the cohort further based on disease stage , risk group , age , or presence of molecular lesions ., We analyzed data from 151 samples from 138 individuals at various time points during treatment ( 44 at diagnosis , 42 after the start of treatment , 59 at relapse , and 6 at unknown time points ) and with varying risk status ( Fig 1A , S1 Table ) ., Samples labeled as “diagnosis” were biopsied before the start of treatment , “relapse” samples were taken at the time of disease relapse , and “post-treatment” samples comprised refractory disease , samples collected at definitive surgery , and any other non-relapsed tumors that had been exposed to treatment ( generally 3–4 rounds of chemotherapy ) ., Nine patients had serial biopsies sent for profiling ( Fig 1B , S2 Table ) ., Across the cohort , we identified 1204 unique variants involving 352 unique genes ( Fig 1C ) ., We define “suspected driver” variants as any short variants ( single amino acid substitution or short insertion/deletion ) that appear in the Catalog of Somatic Mutations in Cancer ( COSMIC ) , any copy number alterations ( CNAs ) that have been shown in the literature to be known oncogenic drivers , or any type of variant that disrupts a tumor suppressor gene , falls within a known hotspot , or is a fusion involving oncogenic driver kinases ., Any type of variant uncharacterized in the literature is reported as a variant of unknown significance ( VUS ) ., It is important to note that every sample in this study was taken from a tumor biopsy without a paired germline sample from the same individual ., Therefore , excluding the patients for whom we have data from multiple biopsies–and in whom mutations were only identified in a subset of biopsies–we cannot rule out the possibility that any reported mutation may be from the germline genome ., We detected an average of 9 . 8 total variants per sample , with a range of 2–20 ., The number of suspected driver variants was lower , with an average of 1 . 28 variants per patient , and a range of 0–6 ., High risk tumors had , on average , a higher total number of variants ( 10 . 1 per sample for high risk vs . 6 . 13 per sample for low risk , P = 5 . 49x10-6 ) and a higher number of suspected driver variants ( 1 . 33 per sample for high risk , 0 . 625 per sample for low risk , P = 0 . 0058 ) ( Fig 1D and 1E ) ., There was no statistically significant difference in the number of mutations between high and intermediate risk tumors ., Among the entire cohort of 151 unique neuroblastoma cases , we detected suspected driver single nucleotide variants or short insertions/deletions in 60 genes ( Fig 2A ) ; amplifications in eight genes ( Fig 2B ) , homozygous deletions in nine genes ( Fig 2C ) , and genomic rearrangements or fusions in eight genes ( Fig 2D ) ., ALK was the most commonly mutated gene , with variants occurring in 20% of patients overall ( suspected driver variants in ALK occurred in 16% of patients ) ., Suspected driver ALK variants were present in 3/43 ( 7 . 0% ) of samples at diagnosis , 7/41 ( 17% ) post-treatment samples , and in 11/54 ( 20% ) of samples at relapse ., We observed a similar trend for suspected driver short variants in other genes ( 41/60 genes were mutated in a higher fraction of patients after treatment than at diagnosis; P = 0 . 0062; Fig 2A ) and for suspected driver high-level copy number amplifications ( 8/8 genes were amplified in a higher fraction of patients after treatment than at diagnosis; P = 0 . 0078; Fig 2B ) ., However , homozygous deletions and genomic rearrangements were present at a similar frequency before and after treatment ( Fig 2C and 2D ) ., The observation that a majority of genes were mutated or amplified in tumors that had undergone treatment led us to ask the inverse question; specifically , whether tumors that had undergone treatment were more likely to have a higher number of suspected driver variants ., Although the total number of variants per sample was similar at all treatment time points ( P = 0 . 25 between diagnosis and relapse ) , we found that the average number of suspected driver variants increased from diagnosis to relapse ( 1 . 1 variants per sample at diagnosis vs . 1 . 7 variants at relapse; P = 0 . 038; Fig 2E ) ., The likelihood of a tumor to harbor at least one suspected driver variant also increased from diagnosis to relapse , although to a less significant degree ( 66% of tumors showed at least one variant at diagnosis , while 80% did at relapse; P = 0 . 17; Fig 2F ) ., With a relatively higher frequency of genomic alterations at relapse , it seems likely that there may be more targeted therapeutic options available for patients who have previously undergone conventional treatments ., We therefore identified 40 of the 80 genes containing suspected driver variants as “potentially actionable” by annotating those that have an existing FDA approved or investigational therapy matches ( see Methods and S4 Table ) ., On average , samples taken at relapse had a higher number of potentially actionable variants ( 0 . 57 at diagnosis vs . 0 . 95 at relapse; P = 0 . 048; Fig 2E ) , and each relapse sample was more likely to have at least one actionable variant ( 41% of tumors had at least one actionable variant at diagnosis vs . 58% at relapse; P = 0 . 11; Fig 2F ) ., We next examined trends in the data that might be of clinical relevance ., First , we looked at differences in the mutational profiles between MYCN amplified and non-amplified tumors ., ALK mutations were found in both MYCN amplified and non-amplified cases ( Fig 3A ) , but several other variants were observed only in one context or the other ( i . e . ATRX mutations only in MYCN non-amplified cases , as reported previously 4 , 7 , and STAG2 mutations only in MYCN amplified cases ) ., The majority of these variants only appear in a single patient in the cohort , so these trends towards mutual exclusivity will need to be confirmed in large case series ., We also examined the frequency of alterations in the MAPK pathway ( gene set in S6 Table ) at diagnosis and relapse ., Fifteen unique patients harbored suspected driver alterations in the MAPK pathway at relapse , while only five tumors had a similar aberration at diagnosis ( P = 0 . 076 , Fig 3D ) ., These data support our prior observation that mutations predicted to activate the MAPK pathway are enriched at disease relapse ( Fig 3B ) 9 ., Nine patients had samples submitted for sequencing at various times in therapy ( S2 Table ) ., Of these , two patients did not have tumor samples with driver mutations ., However , paired samples from the remaining seven patients provided insight into neuroblastoma tumor composition and evolution ., Notably , we observed that definitive clonal driver mutations such as MYCN amplification and ALK mutation identified in diagnostic samples frequently persisted as drivers following exposure to chemotherapy ., Interestingly , Patient 82 presented with bilateral adrenal disease , and biopsies were obtained at diagnosis from both the right and left adrenal glands ., Of the 13 total variants detected in both of these samples , 10 variants were shared in common between the two tumors , suggesting that both tumors shared a common cell of origin ., Patients 1 , 3 , 4 , and 36 had multiple biopsies from the same tumor location at different time points during therapy ( Patient 1: cervical lymph node metastasis at diagnosis and relapse; Patient 3 , 4: adrenal gland primary tumor at diagnosis and definitive surgery; Patient 36: tibial metastasis at two relapses ) ., Unsurprisingly , three of the four patients showed similar mutational profiles with only slight differences at each time point; however , Patient 3 had no variants in common between the two biopsies , demonstrating the potentially significant effects of spatial tumor heterogeneity and evolution ., This phenomenon was further illustrated by Patient 2 , who had four serial biopsies at different anatomic sites ( Table 1 ) ., This patient’s diagnostic tumor contained multiple driver lesions , but some appeared clonal ( heterozygous ARID1A mutation with allelic fraction approximating 50% ) , while others appeared to be present in only a subclonal fraction of cells studied ( incomplete ATRX deletion in this male patient and inferred FGFR1 mutation in 30% of cells ) ., Interestingly , two unique ALK mutations emerged at the first and second relapse in two different metastatic locations , and in each case these were subclonal or absent at the time of final relapse following crizotinib treatment , whereas the ATRX and CDK4 copy number lesions appear to have been evolutionarily selected for and were enriched in the final relapse specimen ., This single case demonstrates the absolute requirement of treating physicians to understand that mutations are not binary events , and consideration of whether or not a lesion is clonal or subclonal may have major implications in the outcomes of targeted therapeutic interventions ., Of the 138 unique patients in this cohort , 15 were reported to have received targeted therapy based on FoundationOne sequencing results and 13 had reported outcomes ., Targeted treatments included ALK inhibitors , CDK4/6 inhibitors , MEK inhibitors , Raf inhibitors , mTOR inhibitors , and antibody therapies ., One patient currently has an ongoing complete response , two have ongoing stable disease ( >1 year and >6 months ) , two patients exhibited stable disease ( 4 months and 1 year ) followed by progressive disease , and the remaining nine patients showed progressive disease despite treatment ., Recent papers have demonstrated that there are relatively few somatic genetic alterations in neuroblastomas at diagnosis , but also that the frequency of mutations and other genetic aberrations increases at disease relapse 9 , 10 ., Here we performed a retrospective study of a larger cohort of 138 patients whose tumors had been biopsied and sequenced at diagnosis , second look surgery , relapse , or some combination of these , to gain a better understanding of the mutational landscape of neuroblastoma throughout treatment and disease progression ., The current study provided validation of our original observation that relapsed high-risk neuroblastoma harbor an increased mutational burden 9 ., Gene panel sequencing is not designed to infer mutational burden , but our data are consistent with an enrichment of mutations in known cancer driver genes at relapse , consistent with clonal evolution under the selective pressure of chemoradiotherapy ., We also showed that the prevalence of “potentially actionable” mutations increased at relapse ., Our data provide support for the hypothesis that the neuroblastoma genome evolves under the selective pressure of chemoradiotherapy 9 , 10 , and It is necessary to understand the current genomic landscape of a tumor rather than that at diagnosis in order to make the most informed treatment decisions ., Thus , the clinical practice of acquisition and analysis of relapsed tumor material , with a view to designing prospective clinical trials to determine the impact of matched target therapy on patient outcomes should be encouraged ., In parallel with several recent reports of prospective collection of tumor genomic data using gene panel 12 or whole exome 13–15 sequencing across pediatric cancer , with or without transcriptome sequencing and at diagnosis or relapse , a few general conclusions are apparent ., First , while not evaluable in this retrospective study , the prospective studies noted above all clearly demonstrate that the acquisition of high dimensional genomics data from pediatric cancers is feasible , and to date has been demonstrated to be safe to incorporate into clinical practice ., This is relevant for this retrospective study of neuroblastoma because the majority of relapses in this disease are in the bone or bone marrow compartment , and often require skilled interventional radiologists and pathologists to safely access , process and prepare often small samples for sequencing ., Second , together the data support the conclusion that a significant proportion of patients at relapse will have a potentially actionable oncogenic driver lesion ., In this study , 58% of subjects studied at relapse showed a somatic mutation in one of several genes that are major targets for drug development in other cancers , consistent with other reports 12 , 14 , 15 ., Third , this study did not study the germline genome , but certainly many of the mutations identified in this study have the potential to be germline 16 , and recent studies have shown that up to 10% of pediatric patients harbor likely relevant germline mutations in known predisposition genes 12–14 , 17 ., Paired germline and somatic DNA sequencing should be considered in the genetic evaluation of pediatric cancer subjects ., Finally , our and recent studies together highlight the current dismal reality that , at present , pediatric cancer sequencing efforts are rarely leading to patient benefit ., The reasons for this are multifactorial and applicable to cancer in general , such as lack of adequate mutation-drug matches for many driver lesions ., Further , there is a need to look beyond mutations in the coding genome , and forays into clinical epigenomic profiling and mRNA profiling may yield other targetable genomic alterations ., However , there are also likely childhood cancer specific explanations as well , such as lack of pediatric dosing information and perhaps some degree of risk aversion on the part of physicians and industry partners to expand targeted therapy options in this population of children who most likely will die from their disease ., Taken together , recent efforts in next generation sequencing of pediatric cancer highlight the potential clinical significance of these efforts , but more importantly the urgent need to credential precision medicine approaches to childhood cancer in well designed prospective clinical trials ., Such clinical trials are currently underway , and are investigating mutations not only through next-generation sequencing of tumor samples , but also through collecting and sequencing cell free DNA and circulating tumor DNA 18 , 19 ., This research study was approved by the Childrens Hospital of Philadelphia Institutional Review Board ( IRB 14–011037 ) ., A waiver of consent/assent was granted per 45 CFR 46 . 116 ( d ) because the research involved no more than minimal risk and could not practicably be carried out without a waiver ., The inclusion criteria for this study were a diagnosis of neuroblastoma and availability of targeted sequencing data from tumor DNA through Foundation Medicine , using the FoundationOne or FoundationHeme panels ( see below ) ., We included a total of 151 samples from 138 unique individuals sampled from 0 to 67 years ( all but one patient were diagnosed before age 25 ) ., The samples fell into the following categories: 44 at diagnosis , 36 at definitive surgery after initial induction chemotherapy , six from tumors refractory to induction chemotherapy or at delayed surgery , and 59 at relapse ., Time points were unknown for a further six samples ., For simplicity , unless otherwise specified in the manuscript , we grouped samples into the following categories: diagnosis , relapse , and post-treatment , the latter comprising of definitive surgery , refractory , and post-therapy ., Nine of the 138 patients had multiple samples tested at different time points in their therapy ., The group of 151 tumors included 13 such samples ., A total of 118 samples were collected from “high-risk” patients as defined by the Children’s Oncology Group , 23 from “intermediate risk” , eight from “low risk” , and two patients were not classified ., At the time the samples were analyzed , 130/151 had received some sort of therapy , at least one of: cytotoxic chemotherapy , radiation therapy , MIBG treatment , anti-GD2 immunotherapy , or isotretinoin therapy ., Complete information on the patient cohort can be found in S1 Table ., All details of sequencing and data processing can be found in Frampton , et al 20 ., Foundation Medicine determines percent tumor nuclei through histopathological review ., These results can be found in S5 Table ., FoundationOne and FoundationOne Heme are pan-cancer tests ., FoundationOne interrogates the entire coding sequence of 315 cancer-related genes plus select introns from 28 genes often rearranged or altered in cancer ., The FoundationOne Heme DNA panel interrogates the entire coding sequence of 405 genes and selected introns of 31 genes involved in rearrangements ., These genes are known to be somatically altered in human solid cancers based on recent scientific and clinical literature ., Each gene in the panel is sequenced to a typical median depth of 500X ., Of the 151 tumor samples included in this study , 63 were processed using the FoundationOne panel ( 46 using the T5a baitset , and 17 using the T7 baitset ) , and 88 were processed using FoundationOne Heme ( 24 with the T6b baitset , and 64 with the D2 baitset ) ., All sequenced genes are listed in S3 Table ., Foundation Medicine returns calls for mutations , copy number alterations , and rearrangements in spreadsheet form ., We imposed a constraint on copy number changes , and only considered gains of copy number >10 in our final analysis to rule out false positive calls that may result from aneuploidy and not focal amplification events ( S1 Fig ) ., A search on clinicaltrials . gov was performed ( in August 2016 ) for each of the 80 genes containing suspected driver variants ., A gene was considered “potentially actionable” if a mutation in the gene was an indication for assigning a patient to an arm of a clinical trial ., We identified 40 such genes ., Association between binary variables was assessed using Fisher’s exact test ., Association between continuous variables was assessed using Welch’s T test ., All analyses were performed in the R statistical language ., All code is available at https://bitbucket . org/opadovan/nb_fm_public ., All the data in this manuscript are supplied in S1 Table . | Introduction, Results, Discussion, Materials and Methods | Neuroblastoma is characterized by a relative paucity of recurrent somatic mutations at diagnosis ., However , recent studies have shown that the mutational burden increases at relapse , likely as a result of clonal evolution of mutation-carrying cells during primary treatment ., To inform the development of personalized therapies , we sought to further define the frequency of potentially actionable mutations in neuroblastoma , both at diagnosis and after chemotherapy ., We performed a retrospective study to determine mutation frequency , the only inclusion criterion being availability of cancer gene panel sequencing data from Foundation Medicine ., We analyzed 151 neuroblastoma tumor samples: 44 obtained at diagnosis , 42 at second look surgery or biopsy for stable disease after chemotherapy , and 59 at relapse ( 6 were obtained at unknown time points ) ., Nine patients had multiple tumor biopsies ., ALK was the most commonly mutated gene in this cohort , and we observed a higher frequency of suspected oncogenic ALK mutations in relapsed disease than at diagnosis ., Patients with relapsed disease had , on average , a greater number of mutations reported to be recurrent in cancer , and a greater number of mutations in genes that are potentially targetable with available therapeutics ., We also observed an enrichment of reported recurrent RAS/MAPK pathway mutations in tumors obtained after chemotherapy ., Our data support recent evidence suggesting that neuroblastomas undergo substantial mutational evolution during therapy , and that relapsed disease is more likely to be driven by a targetable oncogenic pathway , highlighting that it is critical to base treatment decisions on the molecular profile of the tumor at the time of treatment ., However , it will be necessary to conduct prospective clinical trials that match sequencing results to targeted therapeutic intervention to determine if cancer genomic profiling improves patient outcomes . | Neuroblastoma is a pediatric cancer that usually affects children within the first five years of life ., The survival rate for the high-risk form of the disease is 40–50% , and patients suffering metastatic recurrences have no known curative therapeutic options ., Drugs targeted to specific genetic alterations in neuroblastoma may be more effective ., Although neuroblastomas generally have few actionable genetic alterations at diagnosis , targetable mutations that confer therapy resistance may be selected for over time ., Here , we analyzed cancer gene panel sequencing data from 151 neuroblastomas acquired at various time points during therapy to further define how the genomic landscape of neuroblastoma evolves ., We found that relapsed tumors tended to have a higher frequency of mutations potentially targetable with currently available therapies , particularly in the RAS/MAPK pathway ., Our data support the concept that therapeutic decisions targeting specific oncogenic mutations should be based on sequencing data obtained as close to the intervention as possible , and not be reliant on archived diagnostic material ., Prospective clinical trials will be required to determine if sequencing data obtained at the time of tumor progression can lend to improved neuroblastoma patient outcomes . | cancer detection and diagnosis, cancer genomics, medicine and health sciences, biopsy, cancer treatment, clinical oncology, blastomas, cancers and neoplasms, basic cancer research, surgical and invasive medical procedures, oncology, mutation, clinical medicine, pharmaceutics, neuroblastoma, drug therapy, chemotherapy, diagnostic medicine, genetics, biology and life sciences, genomics, genomic medicine | null |
journal.pcbi.1006931 | 2,019 | Adaptive multi-view multi-label learning for identifying disease-associated candidate miRNAs | MiRNAs are a group of short non-coding RNAs that mediate post-transcriptional gene silencing1 ., Accumulating evidence has proved that miRNAs play crucial roles in a variety of cancer-related pathways ., Therefore , the identification of miRNA-disease associations can shed new light on understanding possible pathogenesis of diseases ., To compensate for the limitations of experiment-based approaches , a great number of computational models have been proposed to identify potential disease-related miRNAs in recent years2 ., Under the assumption that functionally similar miRNAs tend to be associated with phenotypically similar diseases , Jiang et al . prioritized the entire microRNAome for over a thousand diseases by constructing an integrated phenome-microRNAome network3 ., Chen et al . measured the global network similarity and inferred potential miRNA-disease interactions based on random walk with restart4 ., Shi et al . adopted a similar idea and further integrated the protein-protein interactions into the prediction process5 ., Chen et al . proposed a novel heterogeneous graph inference method by iteratively updating the association probability6 , 7 ., Liu et al . constructed a heterogeneous network in which they integrated the miRNA-target gene and miRNA-lncRNA associations8 ., Specifically , the methods introduced above mainly predicted disease-related miRNAs by applying random walk algorithms to the reconstructed similarity networks9 ., Another family of prediction methods was generally based on network topological characteristics and also achieved remarkable performance ., For instance , Zou et al . computed the similarity score based on walks of different lengths between the miRNA and disease nodes10 ., Sun et al . exploited the potential disease-related miRNAs based on known miRNA-disease network topological similarity11 ., You et al . proposed to measure the association score for a miRNA-disease pair by calculating the accumulative contributions from all paths between them12 ., Li et al . used DeepWalk to enhance the existing associations through a topology-based similarity measure13 ., Chen et al . computed the association possibility between a disease node and a miRNA node in the corresponding graphlet interaction isomers14 ., Although effective , these methods are sensitive to the change of the network topological structures , which might affect the prediction accuracy ., Alternatively , prediction methods that were based on semi-supervised learning as well as supervised learning have been well developed ., Xiao et al . introduced a graph regularized non-negative matrix factorization to effectively discover sparse miRNA-disease associations15 ., Both Chen et al . and Yu et al . adopted matrix completion to recover the potential missing miRNA-disease associations16 , 17 ., Zeng et al . used a derivative algorithm structural perturbation method to estimate the link predictability with structural consistency as the indicator18 ., Chen et al . used an ensemble model where a sequence of weak learners were trained to collectively obtain a predicted association score19 ., Recently , we reconstructed the miRNA and disease similarity matrices based on global linear neighborhoods and then applied label propagation to predict potential associations between diseases and miRNAs20 , 21 ., Chen et al . extracted novel feature vectors for both miRNAs and diseases to train a random forest classifier for the prediction task22 ., Although great efforts have been made to efficiently uncover potential miRNA-disease associations , most existing computational approaches still suffer from several limitations ., Specifically , the inherent noise in the current datasets resulted in incomplete and sparse similarity matrices and thus inevitably affected the prediction accuracies of these methods ., Moreover , the integration of multiple biological data sources in calculating the similarity matrices for both miRNAs and diseases was generally performed by averaging the input similarity information , which might lead to suboptimal results ., Lastly , the predicted association scores from miRNA space and disease space were often updated separately during the learning process ., To solve these problems , in this paper , we propose a novel Adaptive Multi-View Multi-Label ( AMVML ) learning framework to infer disease-related miRNAs ., In particular , our method adaptively learns a new affinity graph for miRNAs and diseases respectively from multiple data sources ( i . e . miRNA sequence similarity , Gaussian interaction profile kernel similarity and so on ) ., In addition , we unify the optimization process for both disease space and miRNA space based on multi-label learning ., The experimental results under several different evaluation metrics clearly demonstrate the superior performance of our method over previous methods ., We further carry out a case study on thyroid cancer to identify potential prognostic biomarkers ., The known human miRNA-disease associations were retrieved from HMDD v2 . 0 database23 ., HMDD is a database for experimentally supported human miRNA and disease associations that were manually collected from all the miRNA-related publications in PubMed ., Each entry in HMDD contains four items , i . e . miRNA name , disease name , experimental evidence for the miRNA-disease association and the publication PubMed ID ., To keep consistent of data from different sources , we also downloaded the annotation information of 4796 human miRNAs released on March 2018 from miRBase24 ., We then downloaded the latest MeSH descriptors from the National Library of Medicine ( https://www . nlm . nih . gov/ ) and only retained items from Category C for diseases , which resulted in 11572 unique items ., After mapping the miRNA names and disease names involved in each association with miRBase records and MeSH descriptors , we finally obtained 6088 associations between 328 diseases and 550 miRNAs for subsequent analysis ( S1 File ) ., Specifically , we classified the 328 diseases based on the Diseases Categories provided in MeSH ., For diseases belonging to multiple categories , we increased the count by one for each category accordingly ., As a result ( Fig 1 , S2 File ) , we can see that most diseases recorded in HMDD were cancers ., For convenience , we used a binary matrix Y ∈ ℝ328×550 to represent the miRNA-disease associations ., For a given disease i and miRNA j , Yij = 1 if i is related to j , and Yij = 0 otherwise ., As described in 25 , the disease semantic similarity can be calculated based on Directed Acyclic Graphs ( DAGs ) ., Specifically , for a given disease d , its DAG is composed of three items , i . e . DAG = ( d , T ( d ) , E ( d ) ) , where T ( d ) represents d itself together with all its ancestor nodes , and E ( d ) contains all direct links connecting the parent nodes to child nodes ., The contribution Dd ( t ) of a disease t in a DAGd to the semantics of disease d was defined as follows:, {Dd ( D ) =1Dd ( t ) =max{0 . 5*Dd ( t′ ) |t′∈childrenoft}ift≠d, ( 1 ), The semantic similarity score between two diseases i and j can then be calculated by:, S ( i , j ) =∑t∈T, ( i ) ∩T, ( j ) ( Di ( t ) +Dj ( t ) ) ∑t∈T, ( i ) Di ( t ) +∑t∈T, ( j ) Dj ( t ), ( 2 ), Moreover , the similarity between a given disease d and a group of diseases Dt = {dt1 , dt2 , … , dtk} was defined by:, S ( d , Dt ) =max1≤i≤k ( S ( d , dti ) ), ( 3 ), Finally , we obtained the semantic similarities for each disease pair according to ( Eq 2 ) ., We denoted the semantic similarity matrix as AD ( 1 ) ∈ ℝ328×328 where ADij ( 1 ) represents the semantic similarity between disease i and disease j ( S3 File ) ., In this subsection , to comprehensively characterize similarities between miRNAs , we adopt three measures using different biological data sources for subsequent predictions26 ., Gaussian interaction profile kernel similarity has been widely used in previous studies and proved effective in measuring both miRNA and disease similarities ., For a given miRNA i or disease j , its interaction profile IP ( mi ) or IP ( dj ) was a binary vector extracted from the i-th row or the j-th column of the association matrix Y . The kernel similarity between two miRNAs or two diseases could then be computed by:, KM ( mi , mj ) =exp ( −βm‖IP ( mi ) −IP ( mj ) ‖2 ), ( 6 ), KD ( di , dj ) =exp ( −βd‖IP ( di ) −IP ( dj ) ‖2 ), ( 7 ), where βm and βd are defined as follows:, βm=βm′/ ( 1550∑i=1550‖IP ( mi ) ‖2 ), ( 8 ), βd=βd′/ ( 1328∑i=1328‖IP ( di ) ‖2 ), ( 9 ), where βm and βd are two parameters controlling the kernel bandwidth ., As a result , we used AM ( 4 ) ∈ ℝ550×550 and AD ( 2 ) ∈ ℝ328×328 to represent the obtained Gaussian interaction profile similarity matrices for miRNAs and diseases , respectively ., We summarize the notations used throughout this paper ., Given a matrix M , Mij and Mi represent its ij-th element and i-th row , respectively ., The transpose of M is denoted by MT . Tr ( M ) denotes the trace of M and the Frobenius norm of M is represented as ||M||F ., For a similarity matrix S , its Laplacian matrix LS is defined as LS=DS−ST+S2 , where DS is a diagonal matrix with its i-th diagonal element equal to ∑j ( Sij + Sji ) /2 ., To systematically evaluate the performance of our method and illustrate its superiority over existing alternatives , we compared AMVML with fourstate-of-the-art methods , i . e . IMCMDA37 , SPMMDA38 , PBMDA12 and EGBMMDA19 under several evaluation metrics ., All these methods have been proved effective in predicting reliable disease-associated miRNAs ., First of all , we adopted the global Leave-One-Out Cross-Validation ( LOOCV ) and five-fold cross-validation to test the general prediction performance ., Specifically , in the framework of global LOOCV , each known miRNA-disease association was selected as a test sample while the remaining associations were considered as training samples ., For five-fold cross-validation , all known miRNA-disease associations were randomly divided into five subsets and each subset was chosen as the test samples ., Besides , the five-fold cross-validation was repeated 10 times to eliminate the potential bias caused by the sample division ., The prediction performance was illustrated by Receiver Operating Characteristic ( ROC ) curve and the accuracy was quantified by the Area Under the ROC Curve ( AUC ) ., As shown in Fig 3 , AMVML achieved the highest accuracy among all methods in both global LOOCV and five-fold cross-validation ., Next , we employed another evaluation metric called Leave-One-Disease-Out Cross-Validation ( LODOCV ) to verify the prediction performance when no prior information is available ., Specifically , for each disease d , we removed all known miRNAs associated with d and carried out predictions based on miRNA association information of the other diseases ., Since there are no known associations for each tested disease in advance , LODOCV is more difficult than global LOOCV and five-fold cross-validation ., We calculated an AUC value for each disease in LODOCV and thus obtained a vector consisting of 328 AUC values for each method ., We then demonstrated the comparison results by density plots ( Fig 4A ) ., As a result , the AUC values obtained by our method mainly concentrated over the interval 0 . 9 , 1 , indicating a better performance than that of the other methods in terms of LODOCV ., Wilcoxon signed-rank test further confirmed the statistical significance of the comparison results ( Table 1 ) ., Lastly , we conducted experiments on real datasets to further demonstrate the prediction ability of our method ., To this end , we first downloaded the older version of HMDD ( v1 . 0 ) which contains 1474 known associations between 129 diseases and 280 miRNAs after filtering ( S7 File ) ., Compared to HMDD v1 . 0 , there were 4614 ( i . e . 6088–1474 ) new miRNA-disease associations , 199 ( i . e . 328–129 ) new diseases and 270 ( i . e . 550–280 ) new miRNAs involved in HMDD v2 . 0 ., In particular , among the 4614 newly recorded associations in HMDD v2 . 0 , 2445 associations were related with miRNAs and diseases already existed in HMDD v1 . 0 , while 2169 associations were related with either new miRNAs or new diseases only contained in HMDD v2 . 0 ., Moreover , the degree distribution of miRNAs as well as that of diseases for the 4614 associations indicating that only a minority of these associations were related with highly connected miRNAs and diseases ( S1 Fig ) ., We then applied each method on HMDD v1 . 0 and validated the prediction results by the 4614 associations newly added in HMDD v2 . 0 ., Therefore , for each method , the greater the number of true positives predicted , the better the performance ., Specifically , we compared the number of true positives in the top-N miRNAs predicted by each method with N ranging from 10 to 50 and an interval of 10 ., As exhibited in Fig 4B , AMVML obtained greater number of validated disease-associated miRNAs than the other methods ., Similar results were also obtained with increased N and larger intervals ( S2 Fig ) ., Taken together , the experimental results under various evaluation metrics proved the effectiveness of our method ., There were two trade-off parameters α and β in our method which balance the learned similarity matrices and the predicted association matrix ., Generally , since our objective function is a minimization problem , setting a large value to α or β indicates a large impact of the label consistency between diseases or miRNAs on the learned disease or miRNA similarity matrix ., To show a reasonable searching range of these two parameters as well as a general trend of the prediction performance affected by varying their values , in this subsection , we analyzed their influences on the prediction accuracy in terms of five-fold cross-validation ( Fig 5A ) ., Similar trends were also observed in global LOOCV ., In particular , when β was fixed , the smaller the α , the better the performance ., In contrast , when α was fixed , the performance varied in a U shape with the change of β ., We can see that the proposed method reached the best performance when both α and β were equal to 1e-4 ., As described in previous section , we have theoretically proved the convergence of our algorithm ., Here we investigated the convergence rate of our method by analyzing the variations of the objective function value in ( Eq 11 ) with respect to the number of iterations ., As demonstrated in Fig 5B , the objective function value reached a steady state within 5 iterations , indicating a fast convergence rate of our method ., In this section , we conducted a case study on thyroid neoplasms to identify potential miRNA biomarkers for this disease ., The overall prediction results and the differential expression analysis for several other diseases were also provided on Github ( https://github . com/alcs417/AMVML ) ., Thyroid cancer is the most common endocrine cancer and its incidence rate has increased rapidly over recent years39 ., We first downloaded the miRNA expression profiles together with the clinical information of thyroid carcinoma from GDC data portal ( https://portal . gdc . cancer . gov/projects/TCGA-THCA ) ., Concretely , the downloaded data contained 506 tumors samples and 59 normal samples and each sample measured the expression level of 1881 miRNAs ., We then applied our method on the given disease to obtain the top-10 predicted miRNAs ( Table 2 ) ., Specifically , we evaluated the classification power of these miRNAs in differentiating tumor samples from normal samples according to their expression profile and the results of five-fold cross-validation illustrated that they could achieve a mean classification accuracy of 0 . 983 ( S3 Fig ) ., Next , we calculated for each miRNA the fold-change as well as the statistical significance of differential expression using the R package edgeR ( Table 2 ) 40 ., Besides , we searched in another two databases dbDEMC and miR2Disease to see if the predicted miRNAs were also recorded in them41 , 42 ., dbDEMC is an integrated database that designed to store and display differentially expressed miRNAs in human cancers detected by high-throughout methods while miR2Disease is a manually curated database providing information about miRNA deregulation in various human diseases ., As a result , the expression level of the top predicted miRNA hsa-mir-181a-2 was significantly altered between tumor samples and normal samples ( log2 fold-change > 1 and adjusted p-value< 0 . 05 ) , which is consistent with the records in both db2DEMC and miR2Disease ., Therefore , we further checked whether this miRNA could serve as a potential biomarker for thyroid cancer ., Specifically , we carried out one-way ANOVA test to validate whether its expression level at different tumor stages also significantly altered ., The tumor stages of all patients were retrieved from the clinical information and there were six pathologic stages after filtering ., As expected , the expression level of hsa-mir-181a-2 varied significantly among different stages ( Fig 6A ) ., Furthermore , the Kaplan-Meier survival analysis confirmed that the survival rates of patients were also significantly related with its expression level ( Fig 6B ) 43 ., Taken together , our results provided new evidence for the functional role of hsa-mir-181a-2 in the development of thyroid cancer ., Identification of disease-associated miRNAs has drawn much attention during the past decade and it still remains a challenging task ., In this study , we proposed a novel computational framework to effectively uncover the potential links between miRNAs and diseases ., Our method integrated datasets from multiple sources and adaptively learned two new similarity graphs ., Specifically , instead of assigning predetermined weight values to each input similarity matrix , the proposed method automatically updated the view weights according to the reliability of each view ., It is also worth mentioning that our method could be easily extended if there are new data sources available ., Besides , our method could simultaneously update the prediction results from both disease space and miRNA space ., The convergence of our method has been proved both theoretically and experimentally ., To demonstrate the utility of our method , we compared AMVML with five state-of-the-art methods and the experimental results confirmed the superiority of our method ., We then applied our method on thyroid cancer and found that hsa-mir-181a-2 could be a potential prognostic biomarker ., Notably , our method is not limited to discover miRNAs for which an association is already known between its paralogous miRNA and the same disease ., In essence , as a semi-supervised learning model , our method could fully take advantage of the limited number of known miRNA-disease associations together with multiple sources of biological datasets to reliably predict novel associations ., Therefore , we anticipate that our method could serve as an effective tool for miRNA-disease association prediction ., The superior performance of our model can be largely attributed to the following two reasons ., First , the consensus similarity matrices obtained from multiple biological datasets based on multi-view learning for both miRNAs and diseases are more robust to outliers and noises compared to existing methods ., Second , the graph-based multi-label learning unified the two prediction spaces into one optimization framework , which enhances the inherent correlations between miRNAs and diseases from the label-consistency point of view ., Nevertheless , our method still has some limitations ., Specifically , there are two parameters α and β in the objective function that need to be tuned in advance , and it is a non-trivial task to find the best combination of the two parameters ., In addition , although our method can adaptively learn a new affinity graph from different data sources , the integration of unreliable similarity matrices might weaken the overall prediction accuracy . | Introduction, Materials and methods, Results, Discussion | Increasing evidence has indicated that microRNAs ( miRNAs ) play vital roles in various pathological processes and thus are closely related with many complex human diseases ., The identification of potential disease-related miRNAs offers new opportunities to understand disease etiology and pathogenesis ., Although there have been numerous computational methods proposed to predict reliable miRNA-disease associations , they suffer from various limitations that affect the prediction accuracy and their applicability ., In this study , we develop a novel method to discover disease-related candidate miRNAs based on Adaptive Multi-View Multi-Label learning ( AMVML ) ., Specifically , considering the inherent noise existed in the current dataset , we propose to learn a new affinity graph adaptively for both diseases and miRNAs from multiple similarity profiles ., We then simultaneously update the miRNA-disease association predicted from both spaces based on multi-label learning ., In particular , we prove the convergence of AMVML theoretically and the corresponding analysis indicates that it has a fast convergence rate ., To comprehensively illustrate the prediction performance of our method , we compared AMVML with four state-of-the-art methods under different validation frameworks ., As a result , our method achieved comparable performance under various evaluation metrics , which suggests that our method is capable of discovering greater number of true miRNA-disease associations ., The case study conducted on thyroid neoplasms further identified a potential diagnostic biomarker ., Together , the experimental results confirms the utility of our method and we anticipate that our method could serve as a reliable and efficient tool for uncovering novel disease-related miRNAs . | MiRNAs are a class of small non-coding RNAs that are associated with a variety of complex biological processes ., Increasing studies have shown that miRNAs have close relationships with many human diseases ., The prediction of the associations between miRNAs and diseases has thus become a hot topic ., Although traditional experimental methods are reliable , they could only identify a limited number of associations as they are in general time-consuming and expensive ., Consequently , great efforts have been made to effectively predict reliable disease-related miRNAs based on computational methods ., In this study , we develop a novel method to discover potential miRNA-disease associations based on Adaptive Multi-View Multi-Label learning ., Considering the inherent noise existed in the current dataset , we propose to learn a new affinity graph adaptively for both diseases and miRNAs from multiple biological data source , including miRNA sequence similarity , miRNA functional similarity and Gaussian interaction profile kernel similarity ., Notably , our method is applicable to diseases without any known associated miRNAs and also obtains satisfactory results ., The case study conducted on thyroid neoplasms further confirms the prediction reliability of the proposed method ., Overall , results show that our method can predict the potential associations between miRNAs and diseases effectively . | linguistics, medicine and health sciences, natural antisense transcripts, gene regulation, endocrine tumors, carcinomas, applied mathematics, cancers and neoplasms, social sciences, biomarkers, simulation and modeling, oncology, algorithms, optimization, micrornas, mathematics, research and analysis methods, lung and intrathoracic tumors, gene expression, thyroid, thyroid carcinomas, biochemistry, rna, anatomy, nucleic acids, semantics, thymic tumors, genetics, endocrine system, biology and life sciences, physical sciences, non-coding rna, neoplasms | null |
journal.pgen.1006414 | 2,016 | Budding Yeast Rif1 Controls Genome Integrity by Inhibiting rDNA Replication | In eukaryotes DNA replication initiates from multiple sites ( origins ) in a characteristic sequential pattern referred to as ‘replication timing’ 1 ., Replication timing is tightly regulated , with some origins being replicated early during S phase and others later 2 , 3 ., Mechanisms that determine replication timing are still unclear , though recent studies point to a model in which limiting factors ( e . g . Sld3 , Sld2 , Dpb11 , Dbf4 ) are sequentially re-distributed to origins with decreased levels of accessibility to these factors , thus generating a temporal program of origin firing 4 , 5 ., In the budding yeast Saccharomyces cerevisiae about one-third of all potential replication origins ( autonomously replicating sequences , or ARSs ) are located within the rDNA repeat array on chromosome XII 6 ., The rDNA array comprises ~150–200 copies of a module containing 35S and 5S rRNA genes separated by two intergenic regions harboring an origin of replication ( rARS ) and a polar replication fork barrier ( RFB ) ( Fig 1A ) ., Interestingly , only ~20% of the rDNA origins fire during a given S phase in wild type cells 6 , 7 , and it has been shown that deregulation of rDNA replication leads to genomic instability 8 , 9 ., The rDNA repeat RFB , which is generated by sequence-specific binding of the Fob1 protein 10 , is believed to prevent head-to-head collisions of the transcription and replication machineries and to mediate rDNA copy number homeostasis ., Replication fork blockage caused by other proteins is observed elsewhere in the genome , for example at tRNA promoters , telomeres , silent mating type loci , dormant origins and centromeres 11–13 ., The replisome protection ( or pausing ) complex , which consists of Tof1/Csm3 in S . cerevisiae , is essential for fork arrest at RFBs both within and outside of the rDNA 13 , 14 and is proposed to act mainly by counteracting the Rrm3 helicase 14 ., Consistent with this idea , genome-wide accentuation of RFBs in rrm3Δ cells leads to fork collapse and breakage , and to loss of viability in combination with mutation of DNA repair genes such as MRE11 , SGS1 , and SRS2 15 , 16 ., Similarly , specific strengthening of the rDNA RFB by Fob1 overexpression decreases viability in mre11Δ mutants 17 ., Apart from its conserved function in DNA double strand break ( DSB ) nucleolytic processing ( resection ) 18 the MRX complex ( Mre11 , Rad50 and Xrs2 ) also participates in replication fork stabilization under conditions of replication stress ( e . g . dNTP depletion; 19 ) and in replisome re-assembly after fork collapse 20 ., Repair of broken replication forks at the yeast rDNA RFB leads to repeat array instability due to recombination-driven gain or loss of copies 10 ., Accordingly , an increase in the efficiency of rDNA origin firing , such as that observed in cells mutated for the histone deacetylase Sir2 6 , 7 , is associated with elevated rDNA instability 8 , 9 , 21 , presumably due to an increase in the number of replication forks arrested at an RFB ., One model suggests that , in addition to down-regulating rDNA origin firing , Sir2 also inhibits unequal sister chromatid exchange , by promoting cohesion binding within the rDNA intergenic spacers 22 , thus defining a second mechanism by which Sir2 promotes rDNA stability ., Rif1 , a budding yeast Rap1-interacting factor , was initially described as an inhibitor of telomerase-dependent lengthening of telomeres in yeast 23 ., Rif1 is highly conserved in eukaryotes 24 , 25 and has more recently been shown to be a regulator of DNA replication initiation in yeast , flies and mammals 26–33 ., We and others found that Rif1 , through its conserved RVxF/SILK motifs , interacts with protein phosphatase 1 ( PP1; Glc7 in budding yeast ) , and that this interaction is crucial for inhibition of replication origin firing by counteracting the activity of the Dbf4-dependent kinase ( DDK ) 26 , 27 , 30 ., Deletion of RIF1 in budding yeast leads to advancement of the replication timing of most late origins 28 ., Importantly , loss of Rif1 in mouse cells leads to defects in S phase progression , hypersensitivity to the DNA polymerase inhibitor aphidicolin , and checkpoint kinase activation 31 , 34 ., Modulation of Rif1 activity may thus provide a valuable tool to study the molecular and cellular consequences of altering the replication-timing program ., Here we show that budding yeast Rif1 inhibits DNA replication initiation at the rDNA locus and thus promotes the stability of rDNA repeat array ., Moreover , the increase of rDNA instability in rif1Δ accounts for the majority of its DNA damage response ( DDR ) -related phenotypes , suggesting that the rDNA is a key target of Rif1 action ., These findings offer a new perspective on the relationship between replication timing , repeated DNA sequences and genome stability ., Driven by the hypothesis that disruption of RIF1 leads to an increase in the number of active replication forks during S phase 26–28 , 30 , we decided to determine where these forks are located ., Chromatin immunoprecipitation ( ChIP ) of epitope-tagged DNA polymerase epsilon ( Pol2 ) from cell cultures synchronously entering S phase provides a read-out of replication fork passage 26 , 27 , 35 ., We thus measured Pol2 association at different loci , comparing wild type to rif1 mutants ., Consistent with our previous observations 27 , the timing of Pol2 recruitment to the early origins was not affected in cells deleted for RIF1 or cells mutated in its Glc7-binding RVxF/SILK motifs ( see ARS305 at the Fig 1B , where in both WT and the rif1 mutants Pol2 is recruited at 45 minutes after the release into S phase ) ., On the other hand , Pol2 was detected over a shorter time interval at the late replicating HMR locus in both rif1 mutants , which might reflect its earlier and/or faster replication ., Given the fact that many dormant replication origins are located within the repetitive rDNA locus 6 , 7 , we also probed for the ARS element there ( referred to as rARS ) ., In wild type cells rARS displayed a peak of Pol2 binding in the middle of S phase ( 60 minutes following release from alpha-factor arrest ) ., Interestingly , deletion of RIF1 or mutation of its Glc7 ( PP1 ) -interacting RVxF/SILK motifs advanced Pol2 binding by ~15 minutes , to a time similar to that of the early ARS305 origin ., Importantly , acute depletion of Rif1 from the nucleus in G1 phase by the anchor-away method ( see Materials and Methods section for details ) also led to advancement of Pol2 binding in the next S phase at rARS , HML and telomere sites , while replication timing of an early origin ( ARS607 ) was not affected ( S1A Fig , left and middle panels ) ., Origins that fire early in S phase , but not late-replicating regions , recruit the pre-replicative complex ( pre-RC ) component Sld3 in G1 , prior to DNA replication initiation 36 ., Significantly , we detected elevated Sld3 recruitment to the late replicating rARS in G1-arrested rif1Δ cells ( Fig 1C ) , whereas the recruitment of Mcm4 , part of the replicative helicase that is loaded synchronously on all origins , was not affected ( S1A Fig , right panel ) ., The decrease in Sld3 recruitment to a non-rDNA early origin ( ARS607 , Fig 1C ) in rif1Δ cells might be due to re-localization of limiting amounts of this protein to the excess of activated late origins , both within the rDNA and elsewhere , due to the absence of Rif1 ., Taken together , these data indicate that Rif1’s interaction with the PP1 phosphatase ( Glc7 ) is responsible for inhibition of rDNA replication , defining the rDNA locus as a novel Rif1-Glc7 target ., To further investigate the role of Rif1 in rDNA replication , we used bromo-deoxyuridine ( BrdU ) incorporation followed by anti-BrdU immunoprecipitation ( IP ) and quantitative PCR ( qPCR ) as a more direct method to measure newly synthesized DNA ., We released G2/M arrested ( nocodazole-treated ) cells into S phase in the presence of 0 . 2 M hydroxyurea ( HU ) and BrdU ( see FACS profiles in S1B Fig ) ., HU slows fork progression and allows one to determine whether late origins fire , or are instead passively replicated by forks coming from nearby early origins ., In accord with a recent genome-wide study 28 , we detected higher levels of DNA synthesis in rif1Δ at late origins ( ARS1212 , ARS522 , HMR locus and telomeres ) , whereas levels of BrdU incorporation at early origins ( ARS305 , ARS607 ) were not affected ( Figs 1D and S1C ) ., We also found higher BrdU incorporation in rif1Δ and rif1-RVxF/SILK mutant cells compared to wild type at and around rARS ( Fig 1D ) ., Importantly , the increased BrdU incorporation in rif1 mutants was specific to the rDNA and not a general feature of repetitive loci , since another repetitive locus ( CUP1 ) incorporated BrdU very similarly in rif1Δ , rif1-RVxF/SILK and wild type cells ., Analysis of the source data from the Peace et al . study 28 also revealed the same trend of higher BrdU incorporation at rARS in rif1Δ compared to wild type ., Next , we used 2D agarose gels 37 to observe directly the replication intermediates at the rDNA locus , again from cells released from a G2/M arrest into S phase in the presence of HU ., Deletion of RIF1 led to a dramatic increase in bubble arc , Y arc , RFB spot , and X-shaped molecules signals at the rDNA in these conditions ( Figs 1E and S1D ) , indicating a higher frequency of rARS firing ., The effect of rif1Δ as seen in asynchronous cultures was less prominent ( S1E and S1F Fig ) , presumably because the increase in fork density in the mutant also increases the rate at which blocked forks are resolved following arrival of a downstream fork moving in the opposite direction , which would convert the replication intermediates into linear molecules ., This effect is nullified when synchronized cells are released into S phase in medium containing HU , which permits early origin firing but severely limits fork elongation ., We thus conclude that the realm of Rif1-dependent inhibition of DNA replication initiation includes the rDNA locus ., As pointed out above , Sir2 plays an important role in several aspects of rDNA biology ., We confirmed the previous observation 7 of an rDNA replication increase ( as detected by BrdU incorporation ) upon deletion of SIR2 , and furthermore found that Sir2’s effect on rARS is quantitatively similar to that of Rif1 ( Fig 2A ) ., However , unlike rif1Δ cells , sir2Δ mutants do not display increased firing at either of the two late-replicating regions we examined , ARS522 or HMR ( ARS317 ) ( Fig 2A ) ., To address the question of whether Rif1 acts independently of Sir2 to inhibit rARS firing , we examined a rif1Δ sir2Δ double mutant , but found no additive effect , suggesting that these two proteins act in a common pathway ., In fact , and quite surprisingly , the rif1Δ sir2Δ double mutant displayed consistently lower BrdU incorporation at both rARS and a site 2 kb distant , compared to both single mutants ., Nevertheless , replication at these sites was still increased at least 2-fold over that observed in wild type cells ., Using 2D gels we also observed more intensive replication and fork pausing in early S phase at the rDNA in sir2Δ cultures , similar to that in rif1Δ ( Fig 2B ) , though with a marked difference in the relative intensity of the arc signals ., Deletion of RIF1 mostly increased bubble arcs , whereas SIR2 deletion and the double deletion of SIR2 and RIF1 led to more Y arcs ., This difference might be due to variations in the spatial pattern of origin activation , fork progression rates , or timing of origin activation in these mutants In conclusion , the above results indicate that Sir2 and Rif1 work in a common pathway to inhibit rARS firing , but suggest in addition that other players may be involved that create a more complex functional relationship between Sir2 and Rif1 ( see Discussion ) ., As indicated above , replication of the rDNA repeats is highly polar in nature due to an orientation-dependent replication fork block ( RFB; see Fig 1A ) ., Replication proceeding rightwards from rARS is efficiently blocked at the RFB , which is thought to prevent potential collisions with an RNA polymerase I ( RNAPI ) complex transcribing the downstream copy of the 35S rRNA gene ., Forks proceeding to the left from rARS , and in the direction of 35S rRNA gene transcription , are free to pass the RFB present at the upstream rDNA copy ., We hypothesized that rDNA locus stability might be sensitive to an increase in origin firing since this leads to a concomitant increase in the number of forks blocked at RFBs ( Fig 1E , Fig 2B ) ., Blocked forks can , with a certain probably , collapse , sometimes generating DNA breaks that will normally be repaired by homologous recombination ( HR ) , non-homologous end joining ( NHEJ ) or alternative break-induced replication ( BIR ) pathways ., Due to the repetitive nature of the rDNA , recombination between different repeats of the same or sister chromatids may lead to a change in the rDNA array size , which is usually referred as ‘rDNA instability’ 38 ., The loss of repeats from the rDNA array can be conveniently measured when a single copy of the ADE2 gene is inserted in the array , in cells where the endogenous ADE2 gene is mutated ., The ADE2 gene confers a white colony-color phenotype , whereas popping-out of this gene from the chromosome ( together with adjacent repeats ) leads to the accumulation of a red pigment when adenine in the medium is limiting , and the appearance of red sectors in colonies 39 ., Indeed , using this colony-color marker-loss assay 39 , we detected higher levels of rDNA instability in rif1Δ cells compared to wild type ( Figs 3A and S2A ) , consistent with a recent report 40 ., To further challenge the idea that the rDNA instability phenotype of rif1Δ is specifically linked to its effect on replication origin firing , we examined the rif1-RVxF/SILK mutant , which we showed previously 27 to result in a loss of the Rif1-Glc7 interaction and increased phosphorylation of two key DDK kinase targets at pre-RCs ., As shown above , the rif1-RVxF/SILK mutant leads to increased and earlier rDNA origin firing ( Fig 1B and 1D ) ., We found that rif1-RVxF/SILK mutant cells also display a higher level of rDNA instability compared to wild type ( Fig 3A ) , though smaller than the increase conferred by complete deletion of RIF1 , perhaps because rif1-RVxF/SILK retains some residual binding to Glc7 27 ., We next hypothesized that strengthening of the RFB by deletion of RRM3 15 , 16 , which encodes a helicase that promotes the passage of replication forks through RFBs 41 , would lead to a further increase in rDNA instability ., As predicted , we observed an additive increase in rDNA instability when combining rif1Δ or rif1-RVxF/SILK with rrm3Δ ( Figs 3B , S2A and S2D ) ., If the effect of Rif1 and Rrm3 on rDNA stability were linked to the RFB , deletion of the FOB1 gene , whose product is required to establish the fork block , would be expected to abolish the instability induced by rif1 mutants , rrm3Δ , or the double mutants rif1Δ rrm3Δ and rif1-RVxF/SILK rrm3Δ ., This is indeed what we found ( Figs 3B , S2A and S2D ) , strongly suggesting that Rif1 , as well as Rrm3 , act through the RFB ., Surprisingly , neither single mutation ( rif1Δ or rrm3Δ ) nor the double mutation rif1Δ rrm3Δ affected cell growth , either under normal conditions or in the presence of DNA damaging agents ( S2C Fig ) , suggesting that DNA repair pathways in these cells are sufficient to cope with the increased load of stalled forks 15 , 16 ., In accordance with its rDNA replication phenotype ( Fig 2 ) , sir2Δ also displays a large increase in rDNA instability that is fully rescued by FOB1 deletion ( Figs 3C and S2B ) ., The increase in rDNA instability caused by sir2Δ is larger than that of rif1Δ and is unaffected by rif1Δ , consistent with a previous report 40 ., Increased instability of the rDNA locus leads to heterogeneity in the size of chromosome XII in a population of the cells 22 ., As expected , then , pulse field gel electrophoresis revealed a heightened smearing ( broader and less sharp band ) of chromosome XII in rif1-RVxF/SILK and rif1Δ cells ( Fig 3D ) , though not to the same extent as in sir2Δ ( S3A Fig ) , consistent with their varying effect on rDNA stability measured by the sectoring assay ., Again as expected , we found that the effect of rif1Δ on chromosome XII heterogeneity was reversed by the fob1Δ mutation ( Fig 3D ) ., Deletion of either RIF1 or RRM3 increases rDNA instability ( Figs 3B and S3C ) , but only rrm3Δ leads to an increase in the ratio of Fob1-dependent blocked forks at the RFB to total forks at rDNA 41 ( compare 2D gels in S1E and S3C Figs ) , since in rif1Δ the increase in RFB signal is paralleled by an increase in the number of forks at the rDNA ( Figs 1E and 2B ) ., These findings further support the argument that Rrm3 acts directly at RFBs , whereas Rif1 primarily acts through controlling DNA replication initiation ., Elevated blockage and collapse of replication forks at the rDNA may also lead to HR-dependent “popping-out” of rDNA repeats in the form of episomal circles 42 , referred to as extrachromosomal rDNA circles ( ERCs ) ., Consistent with elevated rDNA array instability , we observed increased levels of ERCs in rif1Δ , rrm3Δ and sir2Δ cells ( Figs 3E and S3B ) ., It is not known whether the rARS is more or less active on the episomal ERCs , but it is conceivable that the change in ERC number in a cell may affect the apparent rDNA replication phenotype ., Deletion of FOB1 , which has been shown to significantly reduce ERC formation ( 83; Figs 3E and S3B ) abolished ERC accumulation in rif1Δ and sir2Δ mutants ( Fig 3E ) ., However , we found that fob1Δ did not affect the rif1Δ-induced increase in rDNA replication , as detected by BrdU incorporation and 2D gels ( Figs 3F and S3D ) , confirming that the loss of Rif1 influences chromosomal rDNA origin firing ., Taken together , these results show that Rif1 and Sir2 , but not Fob1 , are involved in control of replication initiation at the rDNA locus ., Rif1 was originally identified as a telomere-binding protein involved in TG-tract length regulation ., Deletion of FOB1 did not affect rif1Δ-dependent telomere elongation ( S4A Fig ) , arguing that telomere- and rDNA-related functions of Rif1 are separable ., However , early studies 45 , 47 , 84 , 85 indicated that Rif1 can compete with SIR proteins for binding to the Rap1 C-terminus at telomeres and that this competition can indirectly affect the availability of SIR proteins for binding elsewhere in the genome , in particular at silent mating type loci ( where a Sir2/3/4 complex assembles ) and within the rDNA , where Sir2 binds at two distinct sites ., A more recent report thus suggested that rif1Δ increases rDNA instability indirectly by favoring the re-localization of Sir2 from its binding sites in rDNA to telomeres and silent mating type loci 40 ., To determine whether Rif1 acts directly to affect rDNA stability , or instead works by modulating the distribution of Sir2 at its different target sites ( rDNA , HM loci and telomeres ) , we first assessed rDNA instability in the rif1-RBM mutant , which , like rif1Δ , leads to an increase in telomeric silencing and telomere TG-tract length 43 ., We found that rif1-RBM has no effect on rDNA stability ( Fig 4A ) , suggesting that increased SIR-mediated telomeric silencing and telomeric TG tract length do not lead to rDNA instability ., Furthermore , neither deletion of TEL1 , which reduces telomere length in a rif1Δ background 44 , nor deletion of RIF2 , which further increases telomere length and telomeric silencing 45 , had any effect on rif1Δ-promoted rDNA instability ( Fig 4A ) ., We also examined sir4Δ cells , where the Sir2 protein cannot be recruited to either telomeres or HM loci and is thus liberated for enhanced action within the rDNA 46 ., However , sir4Δ had no significant effect on rif1Δ-induced rDNA instability ( Fig 4B ) ., Taken together , these findings do not support the notion that Rif1 affects rDNA stability by influencing Sir2 distributions in the nucleus , but are instead consistent with Rif1 having a direct effect on rDNA stability ., Next , we measured binding of Sir2 to chromatin by ChIP-qPCR in rif1Δ cells ., We found that Sir2 binding was increased at the HMR silent mating-type locus ( Fig 4C , left panel ) , in line with the idea that Rif1 competes with the SIR complex for Rap1 binding at HMR in wild type cells 47 ., However , in contrast to a recent report 40 that found a small effect of rif1Δ on Sir2 binding at IGS1 ( near the RFB ) using a semi-quantitative ChIP assay , we found no difference in Sir2 binding there by ChIP-qPCR , nor at three other sites along the rDNA locus: at rARS ( which is located in IGS2 ) , at an adjacent region at the 35S rRNA gene promoter , and at a site within the 35S rRNA gene coding sequence ( Fig 4C , left panel ) ., Fob1 ChIP at rDNA was also unaffected by rif1Δ ( Fig 4C , right panel ) ., Taken together , these data suggest that any influence of Rif1 on SIR complex distribution is insufficient to account for its effects on rDNA instability , and instead argue that Rif1 has a direct effect on rDNA stability by maintaining the low level of rARS firing ., Martina et al . 48 recently proposed that Rif1 physically counteracts Rad9 binding to DSBs ., We therefore asked whether rif1Δ-induced rDNA instability stems from unrestrained activity of Rad9 ., However , deletion of RAD9 alone had no effect on rDNA instability in the marker-loss assay and did not alleviate the increased instability caused by rif1Δ ( S4B Fig ) ., We therefore conclude that the effect of rif1Δ on rDNA stability is unrelated to the activity of Rad9 ., Since the rad9Δ mutation abolishes the DNA damage checkpoint ( DDC ) 49 , these results also argue that rif1Δ-dependent elevation in rDNA instability is not a consequence of DDC activation ., Arrested replication forks need to be stabilized and/or restarted to avoid formation of DSBs and/or inappropriate recombination events 13 ., Increased numbers of stalled replication forks might therefore compromise cell viability ., Consistent with this idea , elevating the strength of RFBs , either by removal of the Rrm3 helicase or by overexpression of Fob1 , leads to synthetic sickness in combination with disruption of the MRX complex 15–17 , probably due to a role for MRX in fork repair 50 , fork restart 20 , or fork stabilization at RFBs 17 ., As already reported , deletion of RIF1 also severely compromises growth of mre11Δ cells , both in untreated cells and upon exposure to phleomycin , which generates DSBs ( Fig 5A; 48 , 51 ) ., We reasoned that this effect of rif1Δ might stem , at least in part , from an increased number of replication forks that are pausing at an rDNA RFB in rif1Δ cells , and thus prone to collapse and subsequent DSB formation ., If this were the case , deletion of FOB1 or alleviation of the fork pause through removal of the replisome pausing complex ( Tof1/Csm3 ) would be expected to rescue this synthetic sickness ., Indeed , fob1Δ , tof1Δ , or csm3Δ deletions completely rescued rif1Δ mre11Δ synthetic sickness , both in normal conditions and upon treatment with genotoxic agents ( Figs 5A , S5A and S5B ) ., We next examined the premise that the increased number of RFB-stalled forks in rif1Δ mre11Δ cells stems specifically from the effect of Rif1 on rARS firing ., In support of this notion , we found that mutation of the Rif1 RVxF/SILK motifs alone conferred a synthetic sickness phenotype in combination with mre11Δ that was comparable to that of rif1Δ , whereas rif1-RBM had no such effect ( Figs 5B and S5C ) ., Given our finding that rif1-RVxF/SILK , but not rif1-RBM , increases rARS firing 27 , these data point to a primary effect of Rif1 on rARS firing as the cause for synthetic sickness in combination with mre11Δ ., To test this idea further , we introduced the temperature-sensitive cdc7-4 mutation , which compromises DDK kinase activity and thus decreases replication initiation rates genome-wide 52 , into our rif1Δ mre11Δ strain ., At 30°C , where compromised cdc7-4 activity begins to affect growth in a rif1Δ background , we note significant alleviation of rif1Δ mre11Δ synthetic sickness , both in untreated and phleomycin-treated cells ( Fig 5C ) ., As expected , at 37°C cdc7-4 is unable to support viability in either the rif1Δ or the rif1Δ mre11Δ background ., Finally , we reasoned that if elevated rDNA repeat replication coupled with fork blockage at the RFB were the source of toxicity in rif1Δ mre11Δ double mutants , then removing the rARS/RFB replication system from chromosome XII should improve growth of these cells ., To test this idea we took advantage of a previously described rDNA array deletion strain , which leaves only 2 chromosomal rDNA repeats ( rdn1Δ strain ) 53 ., Survival of this strain is maintained by a multi-copy plasmid harboring both the 35S and 5S rRNA genes but having a 2 μm replication origin instead of rARS ., As predicted , this rDNA repeat- and rARS-deficient strain displayed no evidence of rif1Δ mre11Δ synthetic sickness , nor any effect of FOB1 deletion ( Fig 5D ) ., In line with its strong additive effect in combination with rif1Δ ( Fig 3B ) , rrm3Δ also displays strong synthetic sickness with mre11Δ ., However , consistent with a more general role of Rrm3 at replisome barriers throughout the yeast genome 41 , rrm3Δ synthetic sickness with mre11Δ was more severe than that of rif1Δ mre11Δ and was significantly but not completely suppressed by fob1Δ ( S6A Fig ) ., Moreover , double deletion of RRM3 and RIF1 was inviable in combination with mre11Δ , in line with the additive effects of Rrm3 and Rif1 on rDNA integrity ( S6A Fig ) ., Consistent with the fact that sir2Δ , like rif1Δ , leads to elevated rDNA origin usage and rDNA instability ( Fig 2 , Fig 3C and 3D ) , we found that sir2Δ also shows fob1Δ-suppressible synthetic sickness with mre11Δ ( Fig 5E ) ., The triple mutant rif1Δ sir2Δ mre11Δ was slightly less sick than rif1Δ mre11Δ ( S6B Fig ) , consistent with a partial decrease in replication at the rDNA when combining rif1Δ with sir2Δ ( Fig 2A and 2B ) ., In line with our conclusion that the Rif1 effect on rDNA is independent of Sir2 re-localization , sir4Δ mutation did not alleviate the synthetic sickness of rif1Δ with mre11Δ ( S6C Fig ) as it had no significant effect on rif1Δ-induced rDNA instability ( Fig 4B ) ., We next reasoned that the burden of elevated replication in rif1Δ should lead to a synthetic growth defect in combination with other mutations affecting replisome integrity ., In fact , combining rif1Δ with deletion of CTF4 , which encodes a replisome component that couples CMG helicase and DNA polymerase alpha/primase 54 , led to a strong synthetic sick phenotype ( Fig 5F ) ., The same is true for deletion of MMS22 , whose product has been proposed to be recruited by Ctf4 to the replisome 55 as part of the Rtt101-Mms1-Mms22 ubiquitin-conjugating complex , essential for replisome maintenance at endogenous impediments and upon challenge with genotoxic agents 56 ( Fig 5F; 57 ) ., It worth noting that , together with histone acetyltransferase Rtt109 , Mms22 also influences the downstream repair events at blocked forks 58–60 and participates in the maintenance of rDNA array size 61 ., The synthetic interaction of rif1Δ with both ctf4Δ and mms22Δ was also alleviated by the deletion of FOB1 , again consistent with the idea that the primary defect occurs at the rDNA RFBs ( Fig 5F ) ., All of the above results show that the function of Rif1 in rDNA stability becomes essential for survival when replisome maintenance and/or DNA break repair at rDNA RFBs is compromised ., Treatment of cells with the ribonucleotide reductase inhibitor HU leads to DNA replication checkpoint ( DRC ) activation due to accumulation of single-stranded DNA at stalled replication forks , and is accompanied by Rad53 phosphorylation ( reviewed in ( 62 ) ., Interestingly , the strength of the DRC , as measured by Rad53 phosphorylation ( Rad53-p ) , correlates with the number of arrested replication forks 63 ., We thus reasoned that increased replication in early S phase in rif1Δ cells might lead to higher DRC activation ( Fig 6A ) ., Indeed , we observed reproducibly higher levels of Rad53-p following HU treatment in both rif1Δ cells ( in the W303 background ( Fig 6B–6D ) , and two other backgrounds , S288C and JC482 ( S7A Fig , upper panels ) ) , and in cells where Rif1 was rapidly depleted from the nucleus by anchor-away ( S7A Fig , lower panel ) ., Moreover , only the rif1-RVxF/SILK mutant , but not rif1-RBM , exhibited elevated DRC upon HU treatment ( Fig 6B ) , emphasizing the connection with increased replication ., This effect of Rif1 loss depended on the DRC adaptor Mrc1 , but not the DDR adaptor Rad9 ( S7B Fig ) , indicating that it is a bona fide DRC response 62 ., Interestingly , the kinetics of Rad53 de-phosphorylation upon HU withdrawal ( recovery from the DRC ) were similar in wild type and rif1Δ cells ( S7C Fig ) , suggesting that DRC deactivation is not altered by rif1Δ , consistent with previous observations 28 ., Importantly , despite the elevated DRC , rif1Δ does not confer increased sensitivity to genotoxic agents such as HU , MMS , phleomycin or camptothecin ( S2C and S3 Figs; 48 ) , suggesting that Rif1 has no essential role in the DDR , either in DSB repair , repair of damaged forks , or re-start of stalled replication forks ., Deletion of FOB1 had no effect on either the DRC ( Fig 6C ) or the replication phenotype ( Fig 3F ) of rif1Δ cells , suggesting that their elevated Rad53-p might result from an increase in rDNA replication per se and not from subsequent fork stalling and/or breakage at the rDNA RFBs ., However , we found that removal of the majority of the rDNA repeats , leaving only 20 or 2 copies 53 , 64 , also did not alleviate the rif1Δ-induced increase of Rad53-p upon HU treatment ( Fig 6D ) ., This suggests that even when a large increase in rDNA replication initiation is not possible , the effect of rif1Δ on replication initiation elsewhere is sufficient to lead to elevated Rad53-p , a conclusion consistent with the observation that rif1Δ increases origin activation near telomeres and at other sites in the genome ( this report and 26–28 , 30 ) ., It is important to note that in contrast to rif1Δ , deletion of SIR2 did not increase the level of Rad53-p upon HU ( Fig 6E ) ., This might be due to the fact that in sir2Δ cells the increase in rARS usage is accompanied by a decreased efficiency of firing of genomic origins outside of the rDNA locus 7 , whereas rif1Δ leads to elevation of the total replication load on the genome , due to its effects at both at rDNA and elsewhere ., Thus , rDNA origins are only one example of a general Rif1 inhibitory role in DNA replication initiation ., Nevertheless , due to the intrinsic vulnerability of the repetitive rDNA locus , caused at least in part by the RFB 64 , Rif1’s role in genome integrity is manifested largely at this site ., Genome integrity is maintained by various mechanisms that either prevent damage to DNA or mediate its repair once the damage has occurred 65 ., Inhibition of replication initiation at late-firing or normally dormant origins following DNA damage is one such preventive measure , referred to as the ‘replication checkpoint’ 62 ., The exact benefit of late origin inhibition remains enigmatic ., For example , it is still unclear whether or not a failure to down-regulate origin firing in the wake of exogenous damage leads to a decrease in genome integrity 66 , 67 ., Given the DDR-related genetic interactions and phenotypes of budding yeast Rif1 48 , 51 and the recently discovered role of mammalian Rif1 in the DSB repair pathway choice 68 we sought to determine whether yeast Rif1 participates directly in DSB repair , and if so , through what mechanism ., To our surprise , we found inste | Introduction, Results, Discussion, Materials and Methods | The Rif1 protein is a negative regulator of DNA replication initiation in eukaryotes ., Here we show that budding yeast Rif1 inhibits DNA replication initiation at the rDNA locus ., Absence of Rif1 , or disruption of its interaction with PP1/Glc7 phosphatase , leads to more intensive rDNA replication ., The effect of Rif1-Glc7 on rDNA replication is similar to that of the Sir2 deacetylase , and the two would appear to act in the same pathway , since the rif1Δ sir2Δ double mutant shows no further increase in rDNA replication ., Loss of Rif1-Glc7 activity is also accompanied by an increase in rDNA repeat instability that again is not additive with the effect of sir2Δ ., We find , in addition , that the viability of rif1Δ cells is severely compromised in combination with disruption of the MRX or Ctf4-Mms22 complexes , both of which are implicated in stabilization of stalled replication forks ., Significantly , we show that removal of the rDNA replication fork barrier ( RFB ) protein Fob1 , alleviation of replisome pausing by deletion of the Tof1/Csm3 complex , or a large deletion of the rDNA repeat array all rescue this synthetic growth defect of rif1Δ cells lacking in addition either MRX or Ctf4-Mms22 activity ., These data suggest that the repression of origin activation by Rif1-Glc7 is important to avoid the deleterious accumulation of stalled replication forks at the rDNA RFB , which become lethal when fork stability is compromised ., Finally , we show that Rif1-Glc7 , unlike Sir2 , has an important effect on origin firing outside of the rDNA locus that serves to prevent activation of the DNA replication checkpoint ., Our results thus provide insights into a mechanism of replication control within a large repetitive chromosomal domain and its importance for the maintenance of genome stability ., These findings may have important implications for metazoans , where large blocks of repetitive sequences are much more common . | Rif1 is a conserved eukaryotic protein implicated in regulation of both the temporal pattern of DNA replication initiation and the DNA damage response ( DDR ) ., We found that in budding yeast several of Rif1’s DDR-related phenotypes stem from its ability to interact with the Glc7/PP1 phosphatase and inhibit DNA replication initiation at the highly repetitive and highly transcribed rDNA locus ., Each rDNA copy contains a potential replication origin flanked on one side by a proteinaceous replication fork barrier , a crucial player in rDNA array size maintenance ., Additionally , the rDNA RFB ensures that replication proceeds in the same direction as transcription , thus presumably minimizing collisions between the replication and transcription machineries ., Our results show that inhibition of rDNA origin firing by Rif1-Glc7/PP1 prevents the buildup of an excess of stalled forks within the rDNA locus , which can lead to genome instability and cell death ., These findings highlight the challenges posed by the replication of repetitive loci , and in particular the need to limit DNA replication initiation events at such vulnerable regions ., Our study may have important implications for metazoan genomes , which contain a much higher fraction of repetitive sequences than budding yeast ., Finally , since tumor cells already exhibit elevated levels of replication stress , our results suggest that inhibition of systems that limit DNA replication initiation may jeopardize the viability of these cells and thus prove to be a useful therapeutic strategy . | genetic networks, chromosome structure and function, cell cycle and cell division, cell processes, dna damage, telomeres, fungi, model organisms, dna replication, network analysis, genome analysis, protein structure, dna, synthesis phase, saccharomyces, research and analysis methods, computer and information sciences, chromosome biology, proteins, protein structure networks, molecular biology, genetic loci, yeast, biochemistry, chromosomes, cell biology, nucleic acids, genetics, biology and life sciences, yeast and fungal models, saccharomyces cerevisiae, genomics, computational biology, organisms, macromolecular structure analysis | null |
journal.ppat.1001259 | 2,011 | Selective C-Rel Activation via Malt1 Controls Anti-Fungal TH-17 Immunity by Dectin-1 and Dectin-2 | Fungal infections are a major health threat and incidence of both superficial and invasive infections by Candida species are growing throughout the world due to increasing numbers of at-risk immunocompromised patients , such as transplant recipients and those infected with HIV-1/AIDS , as well as the emergence of strains that are resistant to antimycotic drugs 1 ., Anti-fungal adaptive immunity requires both T helper cell type 1 ( TH1 ) and TH-17 immune responses; IL-17 secreted by TH-17 cells mobilizes neutrophils required for anti-fungal responses 2 , 3 , whereas TH1-produced IFNγ optimally activates neutrophils and subsequent phagocytosis of fungi 4 ., Dendritic cells ( DCs ) are crucial for the induction of T helper cell differentiation 5 , 6 ., Although the requirements for TH-17 differentiation by human DCs are not completely clear , it is evident that secretion of IL-23 , IL-1β and IL-6 are important for TH-17 development 7 , 8 , whereas IL-12p70 skews T helper cell differentiation towards TH1 responses 9 ., Little is known about the molecular mechanisms that underlie the induction of the TH-17-promoting cytokines by DCs after fungal infections ., Pattern recognition receptors ( PRRs ) , such as Toll-like receptors ( TLRs ) and C-type lectins , sense pathogens through conserved pattern-associated molecular patterns ( PAMPs ) , which induce signaling pathways to regulate gene transcription ., C-type lectins are important in fungal recognition by DCs and in induction of anti-fungal TH1 and TH-17 immune responses 5 , 10 ., The cell-wall of many fungi , including Candida species ( spp ) , consists of carbohydrate structures such as chitin , mannan and β-glucan that are recognized by C-type lectins like dectin-1 , dectin-2 , DC-SIGN and mannose receptor 5 , 11 , 12 ., Triggering of β-glucan receptor dectin-1 by C . albicans induces both TH1 and TH-17 immune responses by DCs through Syk-dependent NF-κB activation 10 , 13 , 14 ., Syk induces the assembly of a scaffold consisting of the caspase recruitment domain 9 ( CARD9 ) protein , B cell lymphoma 10 ( Bcl10 ) and mucosa-associated lymphoid-tissue lymphoma-translocation gene 1 ( Malt1 ) 13 , 15 ., This CARD9-Bcl10-Malt1 scaffold couples dectin-1 in human to the canonical NF-κB pathway by activating NF-κB subunit p65 and c-Rel 10 , 13 , whereas dectin-1 triggering also leads to activation of the non-canonical NF-κB RelB pathway 10 ., The balance between p65 and RelB activity is controlled by a distinct Raf-1-dependent pathway that thereby dictates expression of IL-12p70 , IL-1β and IL-23 10 ., It is unclear how the CARD9-Bcl10-Malt1 complex is involved in the activation of the different NF-κB subunits and how this affects TH-17 differentiation ., Although dectin-1-deficient people are more susceptible to mucocutaneous fungal infection , CARD9 deficiency in human causes a more pronounced phenotype with chronic mucoctaneous as well as invasive fungal infections 16 , 17 ., These studies suggest that dectin-1 is not the only receptor that couples CARD9-Bcl10-Malt1 to the defense against fungi ., Indeed , dectin-2 interacts with C . albicans through mannan structures present on both yeast and hyphal forms 18 , 19 and a recent study shows that dectin-2 is involved in the induction of TH-17 responses to C . albicans in mice 19 , 20 ., Dectin-2 indirectly activates Syk through association with the FcRγ chain 12 which results in CARD9-dependent expression of IL-2 , IL-10 and TNF 20 ., Thus , both dectin-1 and dectin-2 are involved in TH-17 development through Syk and CARD9 but the underlying mechanisms and involvement of Bcl10 and Malt1 remain unclear ., It is also unclear whether dectin-1 and dectin-2 are required for a general anti-fungal response to all Candida species ., Here we demonstrate that dectin-1 and dectin-2 convergently contribute to anti-fungal TH-17 immunity by inducing IL-1β and IL-23 production ., Both dectin-1 and dectin-2 triggering leads to Malt1 activation , which specifically activates NF-κB subunit c-Rel that is pivotal to the transcriptional activation of the Il1b and Il23p19 genes ., Syk-dependent recruitment of CARD9 and Bcl10 upon dectin-1 triggering is crucial for activation of all NF-κB subunits , while recruitment of Malt1 and activation of its paracaspase activity is distinctively required for c-Rel but not p65 or RelB activation ., In contrast to dectin-1 , dectin-2 triggering activates only c-Rel , which is also dependent on Malt1 signaling , signifying a specialized function for dectin-2 in TH-17 immunity ., Simultaneous triggering of dectin-1 and dectin-2 by pathogenic fungi promotes the expression of IL-1β and IL-23 to boost TH-17-mediated cellular responses , whereas Malt1 inhibition after Candida infection markedly reduces TH-17 polarization ., Thus , Malt1 activation links dectin-1 and dectin-2 to the c-Rel-dependent expression of IL-1β and IL-23 and directs adaptive anti-fungal immunity ., The recruitment of the CARD9-Bcl10-Malt1 complex by Syk links dectin-1 on DCs to NF-κB activation , thereby controlling anti-fungal TH-17 immunity 10 , 13–15 ., In mice , the pivotal role for Syk and CARD9 in dectin-1 signaling has been established using knock-out models 13 , 14 , while , in contrast , little is known about their role in regulating human adaptive immunity ., Here we investigated the role of the CARD9-Bcl10-Malt1 module in relaying signals from dectin-1 in human primary DCs to induce cytokine responses ., We used the β-glucan curdlan , which is a specific ligand for dectin-1 and induces Syk activation in both mice and human 10 , 14 ., We silenced Syk , CARD9 , Bcl10 and Malt1 by RNA interference ( Figure S1 ) and analyzed expression of cytokines involved in TH1 and TH-17 polarization ., Expression of IL-1β , IL-23p19 , IL-6 , IL-12p35 and IL-12p40 mRNA was completely abrogated by Syk , CARD9 as well as Bcl10 silencing ( Figure 1A and 1B ) ., Notably , Malt1 silencing had distinct effects on the different cytokines; IL-1β and IL-23p19 mRNA expression was strongly decreased , whereas IL-6 and IL-12p35 mRNA was enhanced and IL-12p40 mRNA expression was unaffected by Malt1 silencing ( Figure 1B ) ., TLR4-dependent cytokine expression was unaffected by silencing of either Syk , CARD9 , Bcl10 or Malt1 ( Figure S2 ) ., These data show that Malt1 has a very distinctive function by inducing the TH-17-polarizing cytokines IL-23 and IL-1β , whereas both CARD9 and Bcl10 are more generally required for all dectin-1-induced cytokine responses ., The distinct functions of CARD9 , Bcl10 and Malt1 in cytokine induction after dectin-1 triggering led us to investigate their functions in the activation of NF-κB ., Dectin-1 triggering activates all NF-κB subunits in a Syk-dependent manner , which is crucial to dectin-1-induced cytokine responses 10 ., We first determined nuclear translocation and subsequent DNA binding of the different subunits after dectin-1 triggering ., NF-κB dimers are normally retained inactive in the cytoplasm and translocate into the nucleus upon activation 21 ., In control-silenced DCs , dectin-1 triggering by curdlan resulted in activation of p65 , c-Rel , p52 and RelB , while p50 DNA binding could already be detected in unstimulated cells ( Figure 2A ) ., Silencing of either CARD9 or Bcl10 in DCs completely impaired activation of p65 , c-Rel , RelB and p52 after curdlan stimulation ( Figure 2A ) ., Strikingly , Malt1 silencing specifically abrogated c-Rel activation ( Figure 2A ) , whereas nuclear translocation of the other subunits was unaffected ( Figure 2A ) ., Immunofluorescence stainings showed that Malt1 silencing interfered with the nuclear translocation of c-Rel but neither with p65 nor RelB ( Figure 2B ) ., These data strongly suggest that Malt1 is required for selective activation of c-Rel-containing NF-κB dimers , whereas recruitment of CARD9 and Bcl10 are an absolute requirement for activation of all NF-κB subunits ., We next used chromatin immunoprecipitation ( ChIP ) assays to investigate the effect of Malt1-induced c-Rel activation on the DNA binding of the NF-κB subunits to different cytokine promoters ., Our data show that the NF-κB site of the Il1b promoter was solely occupied by c-Rel , while both c-Rel and p65 were bound to the Il23 , Il6 and Il12a promoters after curdlan stimulation of control-silenced DCs , albeit in different ratios ( Figure 2C ) ., Malt1 silencing completed abrogated binding of c-Rel to the Il1b , Il23 , Il6 and Il12a promoters after dectin-1 triggering ( Figure 2C ) , consistent with a pivotal role for Malt1-mediated signaling in c-Rel activation ., The absence of c-Rel activation allowed binding of p65 to the promoters as is evident from the higher p65 association with the different promoters ( Figure 2C ) ., Notably , c-Rel binding to the Il1b promoter was completely replaced by p65 binding after Malt1 silencing ( Figure 2C ) ., This suggests that the Il1b promoter is preferentially bound by c-Rel and that c-Rel is a stronger activator of Il1b transcription than p65 , since c-Rel replacement by p65 after Malt1 silencing resulted in significantly reduced IL-1β expression ( Figure 1B ) ., While Malt1 silencing abolished c-Rel binding to both the Il23 and Il12a promoter after dectin-1 triggering , loss of c-Rel activation had opposite effects on Il23 and Il12a transcription as IL-23p19 mRNA levels were severely decreased , while IL-12p35 mRNA was enhanced ( Figure 1B ) ., These results are consistent with our previous findings showing that c-Rel is a stronger transactivator of Il23 but a weaker transactivator of Il12a transcripton than p65 10 ., IL-6 expression was enhanced after Malt1 silencing ( Figure 1B ) , suggesting that c-Rel functions as an inhibitory factor when bound to the Il6 promoter ., The Il12b promoter was not bound by c-Rel in either control- or Malt1-silenced cells after curdlan stimulation ( Figure 2C ) , consistent with the similar IL-12p40 mRNA levels in both control- and Malt1-silenced cells after dectin-1 triggering ( Figure 1B ) ., In order to further demonstrate the importance of c-Rel in the transcriptional regulation of the Il1b , Il23 , Il6 and Il12a genes , we measured cytokine expression in c-Rel-silenced DCs ( Figure S1 ) after curdlan stimulation ( Figure 2D ) ., Similar to Malt1 silencing , c-Rel silencing strongly decreased IL-1β and IL-23p19 mRNA , while enhancing IL-6 and IL-12p35 mRNA levels compared to control-silenced cells after dectin-1 triggering ( Figure 2D ) ., IL-12p40 mRNA expression was independent of c-Rel activation ( Figure 2D ) , as was LPS-induced cytokine expression ( Figure S2 ) ., These results strongly suggest that Malt1-mediated c-Rel activation leads to the expression of IL-1β and IL-23 , key cytokines in TH-17 differentiation ., Since Malt1 has paracaspase activity 22 , 23 , we next investigated whether the adaptor or protease function of Malt1 is involved in the selective activation of c-Rel after dectin-1 triggering ., We used z-VRPR-FMK , a compound which blocks the proteolytic activity of Malt1 22 ., Inhibition of Malt1 proteolytic activity completely abolished activation of c-Rel without affecting the other NF-κB subunits ( Figure 3A ) , which is similar to Malt1 silencing ( Figure 2A ) , Immunofluoresence stainings confirmed that Malt1 inhibition specifically interferes with nuclear translocation of c-Rel after curdlan stimulation ( Figure S3 ) ., Malt1 paracaspase inhibition also markedly reduced both IL-1β and IL-23p19 mRNA levels and slightly enhanced IL-6 and IL-12p35 mRNA production after curdlan stimulation ( Figure 3B ) , similarly to Malt1 silencing ( Figure 1B ) ., IL-12p40 mRNA production was neither dependent on Malt1 expression nor activation ( Figure 3B ) ., We next measured cytokine production and found that Malt1 inhibition severely reduced IL-1β and IL-23 protein expression , without affecting IL-12p70 expression and only slightly enhancing IL-6 expression ( Figure 3C ) , indicating that dectin-1-induced cytokine expression is primarily regulated at the transcriptional level ., These results show that Malt1 protease activity is required for specific c-Rel activation and plays a central role in the induction of TH-17-polarizing cytokines by dectin-1 ., Dectin-1 plays an important role in anti-fungal immunity through the induction of TH1 and TH-17 differentiation 10 , 14 ., Since Candida albicans infections are amongst the most common causes of invasive fungal infections in immunocompromised patients 24 , 25 , we used two different C . albicans strains to investigate the importance of Malt1 signaling in anti-fungal immune responses ., Consistent with its function in the induction of TH-17-polarizing cytokines , Malt1 activation is required for expression of IL-1β and IL-23 by DCs in response to both C . albicans strain CBS8781 and CBS2712 ( Figure 4A ) ., As observed with curdlan stimulation , the expression of IL-6 was slightly upregulated as a result of Malt1 protease inhibition , whereas IL-12p70 production was unaffected by Malt1 signaling after C . albicans stimulation ( Figure 4A ) ., To elucidate the contribution of dectin-1 signaling to the Malt1-dependent cytokines responses , we treated DCs with C . albicans in the presence of blocking dectin-1 antibodies ., Notably , we observed that C . albicans CBS8781-induced cytokine expression was completely abrogated after blocking dectin-1 , whereas cytokine production after C . albicans CBS2712 stimulation was only partially inhibited by dectin-1 antibodies ( Figure 4A ) ., Malt1 inhibition decreased C . albicans CBS2712-induced IL-1β and IL-23 expression more strongly than dectin-1 inhibition ( Figure 4A ) ., These results suggest that fungal infections trigger not only dectin-1 but also other receptors to induce anti-fungal TH-17 responses via Malt1 ., To further investigate this , we used two different Candida species , C . lusitaniae and C . nivarienis , both emerging pathogenic fungi causing opportunistic infections in transplant and immunocompromised patients 24 , 26 ., C . lusitaniae CBS4413 induced cytokine production in a dectin-1-dependent manner , while C . nivariensis CBS9983 was only partially dependent on dectin-1 signaling for the production of IL-1β , IL-23 , IL-6 and IL-12p70 ( Figure 4B ) ., Both IL-1β and IL-23 production by C . lusitaniae and C . nivariensis was largely dependent on Malt1 protease activity ( Figure 4B ) ., Noteworthy , Candida spp ., that trigger Malt1 activation via dectin-1 in combination with other unidentified receptor ( s ) induce higher levels of IL-1β and IL-23 in DCs than those that trigger only dectin-1 ( Figure 4A and 4B ) ., Similar to the C . albicans strains , C . lusitaniae and C . nivariensis stimulation resulted in slightly enhanced IL-6 expression after Malt1 inhibition , while IL-12p70 was unaffected ( Figure 4B ) ., These data suggest that c-Rel activation by Malt1 signaling controls anti-fungal TH-17 immunity to Candida spp ., We next set out to identify the fungal PRR ( s ) on DCs that are triggered by C . albicans CBS2712 and C . nivariensis to induce Malt1 activation independently of dectin-1 ., The C-type lectin dectin-2 has been shown to participate in fungal TH-17 immunity in mice 19 , 20 ., We explored a role for dectin-2 in cytokine responses to Candida spp ., by using blocking antibodies against dectin-2 ., Notably , induction of IL-1β and IL-23p19 mRNA by both C . albicans CBS2712 and C . nivariensis was partially abolished by dectin-2 antibodies , while blocking both dectin-1 and dectin-2 completely abrogated expression of IL-1β and IL-23p19 ( Figure 5 ) ., These data strongly suggest that dectin-2 signaling contributes together with dectin-1 to the induction of these TH-17-polarizing cytokines by C . albicans CBS2712 and C . nivariensis ., Blocking dectin-2 triggering by C . albicans CBS2712 or C . nivariensis slightly increased IL-6 but greatly enhanced IL-12p35 mRNA expression ( Figure 5 ) , most likely reflecting the negative influence of c-Rel binding to the respective promoters on Il6 and Il12a transcription ., C . nivariensis-induced IL-6 and IL-12p35 mRNA expression was dependent on both dectin-1 and dectin-2 , while C . albicans CBS2712 induced IL-6 and IL-12p35 expression via dectin-1 , as blocking both dectin-1 and dectin-2 had no additional effects compared to blocking dectin-1 alone ( Figure 5 ) ., IL-12p40 mRNA expression by C . albicans CBS2712 and C . nivariensis was independent of dectin-2 triggering ( Figure 5 ) ., These data suggest that dectin-2 signaling through Malt1 controls only c-Rel-dependent gene expression without affecting c-Rel-independent transcription ., The residual expression of IL-6 , IL-12p35 and IL-12p40 induced by C . albicans CBS2712 after either blocking dectin-1 or dectin-1 plus dectin-2 suggests additional involvement of other receptors , such as TLRs ( Figure 5A ) ., As expected , dectin-2 antibodies did not interfere with cytokine expression induced by C . albicans CBS8781 and C . lusitaniae , since cytokine induction was completely inhibited by blocking dectin-1 antibodies ( Figure 4 and 5 ) ., Cytokine protein levels confirmed the mRNA expression data ( Figure S4 ) ., Our data demonstrate that both dectin-1 and dectin-2 contribute to anti-fungal TH-17-polarizing cytokine responses to various Candida spp ., We next triggered dectin-2-FcRγ signaling by crosslinking dectin-2 with antibodies and investigated cytokine expression induced by human DCs ., Dectin-2 crosslinking induced high levels of IL-1β and IL-23p19 mRNA expression ( Figure 6A ) ., Remarkably , dectin-2 crosslinking did neither induce IL-6 , IL-12p35 nor IL-12p40 mRNA expression ( Figure 6A ) , consistent with our observations when blocking dectin-2 binding by Candida spp ., ( Figure 5 ) and strongly suggesting that dectin-2 triggering specifically induces IL-1β and IL-23p19 ., These results confirm that dectin-1 and dectin-2 signaling converge to boost the expression of TH-17-polarizing cytokines , as we observed after Candida spp ., stimulation ., We next investigated whether dectin-2 crosslinking induces NF-κB activation ., Notably , dectin-2 triggering resulted in the specific activation of c-Rel , whereas the other NF-κB subunits p65 , RelB and p52 were not activated ( Figure 6B ) ., Consistently , c-Rel-silenced DCs exhibited a defect in the induction of IL-1β and IL-23p19 mRNA expression after dectin-2 crosslinking ( Figure 6C ) ., We next silenced Malt1 expression to investigate whether dectin-2-FcRγ signaling , like dectin-1 , employs Malt1 to specifically activate c-Rel and induce c-Rel-dependent cytokine expression ., Similar to c-Rel silencing , Malt1 silencing completely abolished IL-1β and IL-23p19 mRNA production in response to dectin-2 crosslinking ( Figure 6C ) ., These data demonstrate that dectin-2 has a specialized function in adaptive immunity and specifically contributes to the induction of IL-1β and IL-23p19 , emphasizing the importance of the Malt1-c-Rel activation axis in TH-17 immunity ., Since Malt1 links dectin-1 and dectin-2 to the expression of the TH-17-polarizing cytokines IL-1β and IL-23 via the activation of c-Rel , we investigated whether Malt1 activation affects adaptive immunity to Candida spp ., We first co-cultured curdlan-primed DCs with CD4+ T cells and measured IL-17 secretion after 5–12 days of co-culture 7 ., Malt1 inhibition markedly reduced the capacity of curdlan-primed DCs to induce IL-17 expression in CD4+ T cells ( Figure 7A , 7B and 7C ) ., Thus , Malt1 activity is essential for the induction of TH-17-polarizing cytokines in DCs via dectin-1 triggering and subsequent TH-17 skewing ., The ability of DCs primed by the different Candida spp ., to promote IL-17 expression in CD4+ T cells was completely blocked when Malt1 activation was inhibited in the DCs ( Figure 7D and 7E ) ., This effect of Malt1 inhibition on the ability of Candida-primed DCs to induce TH-17 polarization was irrespective of the involvement of either dectin-1 alone ( C . albicans CBS8781 and C . lusitaniae ) or combined dectin-1 and dectin-2 triggering ( C . albicans CBS2712 and C . nivariensis ) , consistent with a general role for Malt1 in inducing the TH-17-polarizing cytokines IL-1β and IL-23 ., Thus , the marked impact of Malt1 inhibition on IL-1β and IL-23p19 expression in response to fungal infections translates to a block in TH-17 immunity ., Our data demonstrate that Malt1 signaling to c-Rel activation drives anti-fungal TH-17 responses ., C-type lectins are amongst the most important innate receptors on DCs to induce anti-fungal TH-17 immunity 5 , 11 , 14 ., Expression of cytokines upon dectin-1 triggering by fungi requires NF-κB activation through Syk-dependent CARD9-Bcl10-Malt1 signaling 13 , 15 ., Here we demonstrate that Malt1 activation by dectin-1 and dectin-2 on human DCs induces the expression of TH-17-polarizing cytokines IL-1β and IL-23 through selective activation of the NF-κB subunit c-Rel ., c-Rel is crucial for optimal transcription of the Il1b and Il23p19 genes ., Dectin-1-induced activation of p65 , RelB and c-Rel is completely dependent on the recruitment of CARD9 and Bcl10 ., Notably , Malt1 through its proteolytic paracaspase activity is specifically involved in activation of c-Rel , but dispensable for p65 and RelB activation ., Malt1 activation of c-Rel is similarly essential in the induction of TH-17-polarizing cytokines by dectin-2 ., Strikingly , dectin-2 signaling , unlike dectin-1 , only induces strong c-Rel , but not p65 and RelB activation , strongly suggestive of a specific TH-17 polarizing function of dectin-2 ., Furthermore , the involvement of dectin-1 and detcin-2 in anti-fungal immunity by human DCs depends on the Candida species ., Our data strongly suggest that dectin-2 is crucial in recognition of some pathogenic Candida species to boost dectin-1-induced TH-17 responses via Malt1 ., Thus , Malt1-dependent activation of c-Rel dictates adaptive TH-17 immunity to fungi by dectin-1 and dectin-2 ., Protective immunity against fungal infections via TH-17 cellular responses requires the expression of IL-1β , IL-23 and IL-6 by DCs ., Here we demonstrated that selective activation of NF-κB family member c-Rel by dectin-1 and dectin-2 signaling in response to fungi was essential to expression of IL-1β and IL-23 and consequently TH-17 immunity ., Our data showed that loss of c-Rel binding to the Il1b and Il23p19 promoters strongly decreased IL-1β and IL-23p19 expression even though p65 bound to the NF-κB binding sites in the absence of c-Rel activation ., These data further showed that c-Rel was the stronger activator of Il1b and Il23 transcription ., Furthermore , c-Rel had an inhibitory effect on Il6 transcription , although p65-driven IL-6 expression allows for sufficient IL-6 to direct TH-17 polarization ., Both dectin-1 and dectin-2 triggering resulted in c-Rel activation and c-Rel-dependent IL-1β and IL-23 expression ., Engagement of dectin-1 by fungal ligands leads to phosphorylation of the immunoreceptor tyrosine-based activation motif ( ITAM ) -like sequence within its cytoplasmic domain 15 , 27 and subsequent association of the spleen tyrosine kinase Syk ., Syk activation by dectin-1 is required for NF-κB activation via the assembly of the CARD9-Bcl10-Malt1 module 13 , 14 ., Unlike dectin-1 , dectin-2 requires pairing with the adaptor molecule FcRγ to induce signaling 12 , 20 ., Dectin-2 triggering results in phosphorylation of the ITAM of FcRγ and activation of Syk signaling , which induces cytokine expression 20 ., Dectin-2 signaling is CARD9-dependent , however a role for Bcl10 and Malt1 remains to be established 20 ., In antigen receptor signaling , oligomerization of CARD11 ( CARMA1 ) triggers the formation of a scaffold that physically bridges the CARD11-Bcl10-Malt1 complex with downstream signaling effectors , such as TRAFs and TAK1 , to activate the NF-κB-regulating IKK complex 28 ., Here we demonstrated that c-Rel activation by dectin-1 and dectin-2 is completely dependent on Malt1 activation ., Malt1 is an unique protein as it is the only human paracaspase known 22 , 23 and our data showed that its paracaspase activity was essential to the activation of c-Rel by dectin-1 and dectin-2 ., Malt1 has a distinctive function within the CARD9-Bcl10-Malt1 complex induced upon dectin-1 and dectin-2 triggering since silencing of CARD9 and Bcl10 by RNA interference completely abrogated the activation of all NF-κB subunits , while Malt1 silencing selectively abrogated c-Rel activation ., It is unclear how Malt1 specifically activates c-Rel ., A similar observation has been reported for B cell receptor signaling , which uses the CARD11-Bcl10-Malt1 complex for NF-κB activation 29 , while Malt1 is involved in RelB activation after BAFF stimulation in specific B cell subsets 30 ., In T cell receptor signaling , the paracaspase activity of Malt1 partially accounts for the amount of NF-κB activation 22 , which might reflect the c-Rel-dependency in T cell receptor responses ., Only two substrates for Malt1 are known , its binding partner Bcl10 and A20 that functions as an inhibitor of NF-κB activation 22 , 23 , but it remains to be determined if they have any role in the selective activation of c-Rel via Malt1 ., We showed that dectin-2 signaling only induced strong c-Rel activation , while dectin-1 triggering activated all NF-κB subunits; possibly the differential use of downstream molecules like TRAFs by dectin-1 and dectin-2 might underlie these differences in NF-κB activation ., Crosstalk between signaling pathways triggered by recognition of different PAMPs by various PRRs is essential to the induction of immune responses 5 , 6 , 11 ., Here we demonstrated that dectin-1 and dectin-2 play distinct roles in immunity to fungi ., While dectin-1 triggering induced cytokines involved in promoting both TH1 and TH-17 polarization , dectin-2 triggering resulted specifically in IL-1β and IL-23p19 expression , which enhanced IL-1β and IL-23 expression in response to different pathogenic Candida spp ., This suggests that dectin-1 functions more broadly as an anti-fungal receptor inducing protective immunity , while dectin-2 is more specialized in boosting TH-17 cellular responses ., Our data also demonstrated that even related pathogenic fungi triggered different sets of PRRs , likely contributing to tailoring of pathogen-specific immunity ., C . albicans strain CBS8781 and C . lusitaniae induced cytokine expression in a dectin-1-dependent manner ., In contrast , C . albicans strain CBS2712 and C . nivariensis triggered both dectin-1 and dectin-2 and showed higher IL-1β and IL-23 responses , strongly suggesting that dectin-1 and dectin-2 signaling pathways converge to enhance TH-17 immunity ., Other Candida species might preferentially trigger dectin-2 but not dectin-1 for IL-1β and IL-23p19 protein expression as shown in the study of Saijo et al . 19 ., Notably , C . albicans CBS2712 also induced dectin-1- and dectin-2-independent expression of IL-6 , IL-12p35 and IL-12p40 ., The contribution of TLR signaling , especially via TLR2 , and collaboration with dectin-1 signaling has previously been recognized in cytokine responses in Candida infections 11 , 31 ., However , C-type lectin triggering seems to be more specialized in IL-23p19 and IL-1β induction ., The situation in mice might be more complex as murine TLRs seem to induce c-Rel activation 32 , while human TLRs do not 33 ., Our data emphasize that immune responses are tailored not only to pathogens from different species but even within species ., Thus , interpretation of data obtained with a single pathogen should be done with caution ., Research into the role of dectin-1 in fungal infections using knock-out mice has resulted in conflicting data 34 , 35 and the use of different yet related fungi might underlie these differences ., Genetic variation within the Candida clade might not only account for differences in pathogenicity 36 but also for the differential recognition by innate receptors ., We have demonstrated here that even closely related C . albicans strains trigger different sets of PRRs to activate adaptive immune responses ., Malt1-mediated c-Rel activation might be a general mechanism for induction of protective TH-17 immunity against fungi and other microbes , since the Card9-Bcl10-Malt1 complex might couple other C-type lectins besides dectin-1 and dectin-2 to NF-κB activation ., Furthermore , the carbohydrate specificities of dectin-1 and dectin-2 for β-glucans and high mannoses , respectively , signify their importance in more general anti-fungal immunity against species from the phylum Ascomycota that contain mannan , chitin and glucan structures in their cell-wall 37 , 38 ., Many pathogenic ascomycetes such as Candida spp ., , Aspergillus spp ., , Coccidiodides spp ., , Pneumocystis jirovecii ( previously known as Pneumocystis carinii ) , Histoplasma capsulatum , Trichophyton rubrum and Microsporum audouinni have been identified as dectin-1 and/or dectin-2 ligands 12 , 20 , 34 , 35 , 39–41 ., In contrast , the cell-wall of fungi from the phylum Basidiomycota , such as Cryptococcus and Malassezia spp , differs from that of ascomycetes , as it is enfolded in a glucuronic acid-rich carbohydrate capsule or consists of lipophilic structures , respectively 42 , 43 ., TH-17 responses to Cr ., neoformans have been reported 44 but are not mediated by dectin-1 45 ., Thus , dectin-1 and dectin-2 control c-Rel activation distinctively via Malt1 activation to induce IL-1β and IL-23 expression and as such tailor TH-17 immune responses against fungal pathogens ., Given the pivotal role of TH-17 responses not only in protective immunity against fungi but also in the pathology of human autoimmune diseases like Crohns disease , ulcerative colitis , psoriasis and in vaccine development against tuberculosis 3 , 46 , our results might benefit therapeutic developments as Malt1 presents a rational target for immunomodulatory drugs ., This study was performed in accordance with the ethical guidelines of the Academic Medical Center ., Immature DCs ( iDC , day 6 and 7 ) were generated as described previously 10 ., DCs were stimulated with 10 µg/ml curdlan ( Sigma ) , heat-killed Candida spp ., 47 ( multiplicity of infection ( MOI ) 10 ) and 10 ng/ml LPS from Salmonella typhosa ( Sigma ) ., Dectin-2 triggering was induced by pre-incubating DCs for 2 h at room temperature with 5 µg/ml anti-dectin-2 ( MAB3114; R&D Systems ) , followed by crosslinking on goat-anti-mouse IgG ( 115-006-0710; Jackson ) -coated culture plates ., Cells were preincubated with blocking antibodies or inhibitor for 2 h with 20 µg/ml anti-dectin-1 ( MAB1859; R&D Systems ) , 20 µg/ml anti-dectin-2 ( MAB3114; R&D Systems ) or 75 µM z-VRPR-FMK ( Malt1 inhibitor 22; Alexis ) ., DCs were transfected with 25 nM siRNA using transfection reagent DF4 ( Dharmacon ) , and used for experiments 72 h after transfection ., ‘SMARTpool’ siRNAs used were: Syk ( M-003176-03 ) , CARD9 ( M-004400-01 ) , Bcl10 ( M-004381-02 ) , Malt1 ( M-005936-02 ) , c-Rel ( M-004768-01 ) and non-targeting siRNA ( D-001206-13 ) as a control ( Dharmacon ) ., This protocol resulted in nearly 100% transfection efficiency as determined by flow cytometry of cells transfected with siGLO-RISC free-siRNA ( D-001600-01 ) and did not induce IFN responses as determined by quantitative real-time PCR analysis 10 ., Silencing of expression was verified by real-time PCR and flow cytometry ( Figure S1 ) ., mRNA isolation , cDNA synthesis and PCR amplification with the SYBR green method in an ABI 7500 Fast PCR detection system ( Applied Biosystems ) were performed as described 10 ., Specific primers were designed using Primer Express 2 . 0 ( Applied Biosystems; Table S1 ) ., The Ct value is defined as the number of PCR cycles where the fluorescence signal exceeds the detection threshold value ., For each sample , the normalized amount of target mRNA was calculated from the obtained Ct values for both target a | Introduction, Results, Discussion, Materials and Methods | C-type lectins dectin-1 and dectin-2 on dendritic cells elicit protective immunity against fungal infections through induction of TH1 and TH-17 cellular responses ., Fungal recognition by dectin-1 on human dendritic cells engages the CARD9-Bcl10-Malt1 module to activate NF-κB ., Here we demonstrate that Malt1 recruitment is pivotal to TH-17 immunity by selective activation of NF-κB subunit c-Rel , which induces expression of TH-17-polarizing cytokines IL-1β and IL-23p19 ., Malt1 inhibition abrogates c-Rel activation and TH-17 immunity to Candida species ., We found that Malt1-mediated activation of c-Rel is similarly essential to induction of TH-17-polarizing cytokines by dectin-2 ., Whereas dectin-1 activates all NF-κB subunits , dectin-2 selectively activates c-Rel , signifying a specialized TH-17-enhancing function for dectin-2 in anti-fungal immunity by human dendritic cells ., Thus , dectin-1 and dectin-2 control adaptive TH-17 immunity to fungi via Malt1-dependent activation of c-Rel . | Fungal infections are a major health threat and the incidence is growing worldwide ., There is a need for efficient antifungal vaccines ., Adaptive immune responses and in particular T helper cell type 17 ( TH-17 ) responses are crucial in the defence against fungal infections ., Human dendritic cells ( DCs ) induce TH-17 responses after interaction with fungi ., DCs express C-type lectins dectin-1 and dectin-2 that interact with the carbohydrate structures present in the cell-wall of fungi ., It is unclear how signaling by these C-type lectins leads to specific TH-17 responses ., Here we demonstrate that the signaling molecule Malt1 present in the CARD9-Bcl10-Malt1 complex is responsible for TH-17 induction by selectively activating the NF-κB transcription factor c-Rel , which drives transcription of the TH-17-polarizing cytokines ., Inhibition of either Malt1 or c-Rel prevents TH-17 induction in response to fungi ., Furthermore , we show that the C-type lectin dectin-2 selectively activates c-Rel , signifying a specialized TH-17-enhancing function for this C-type lectin ., Thus , novel vaccination strategies that target dectin-2 or activate Malt1 can induce predominant TH-17 responses ., Since aberrant TH-17 responses underlie the pathology of atopic dermatitis and various autoimmune diseases , Malt1 is a rational therapeutic target to attenuate anomalous adaptive immune responses . | immunology/immunomodulation, infectious diseases/fungal infections, immunology/innate immunity, immunology/leukocyte signaling and gene expression, immunology/immunity to infections | null |
journal.pntd.0004183 | 2,015 | Vibrio cholerae Serogroup O139: Isolation from Cholera Patients and Asymptomatic Household Family Members in Bangladesh between 2013 and 2014 | During 1992–93 , V . cholerae O139 was first recognized in Bangladesh , India and other countries in Southeast Asia as a causative agent of epidemic cholera 1–3 ., Prior to this , the O1 serogroup was considered the sole cause of cholera epidemics 4 ., The isolation of O139 from clinical cases declined quickly after the initial outbreak , with the exception of one epidemic in August 2002 in Dhaka city 5 ., Following on from this , isolates of the O139 serogroup were also isolated sporadically from clinical and environmental samples from various regions of Bangladesh during 2005 , although no large-scale outbreaks of cholera attributed to O139 serogroup V . cholerae were reported during this time 5 , 6 ., Since then , clinical cholera in Bangladesh has been caused entirely by the V . cholerae O1 serogroup , with an unexplained disappearance of V . cholerae O139 5 ., With the recent publication of whole genome-based phylogenies of V . cholerae , we are able to see how the isolates responsible for global cholera relate to each other 7 ., It is clear from these data that the isolates causing the current ( seventh ) pandemic of cholera form a highly related monophyletic lineage dominated by isolates of the El Tor biotype and of the O1 serogroup ., This phylogeny , based on whole genome sequence of clinical isolates , shows three overlapping global expansions of V . cholerae since the pandemic began , denoted wave I , II , and III ., From sequencing of a limited number of O139 isolates , it has been shown that they form a single distinct phylogenetic branch that falls within wave II of the seventh pandemic El Tor lineage 7 ., At one time , O139 isolates were thought to represent a new lineage of V . cholerae that would spread globally and perhaps even replace the O1 serogroup ., Since the O139 serogroup of V . cholerae was first recognized , it has been included in cholera surveillance initiatives and in vaccine design efforts 1 , 8 ., However , pandemic wave II O139 and O1 serogroup isolates are increasingly rare , being replaced almost exclusively by O1 serogroup wave III strains , causing disease globally ., Here we report the isolation , characterization and sequence analysis of recent isolates of V . cholerae O139 recovered from stools of patients hospitalized at the icddr , b diarrheal hospital , as well as from asymptomatic members of the patients’ households ., The aim of this study was to compare these new V . cholerae O139 isolates to existing O139 and O1 isolate sequence data to determine if these new cases were caused by new O139 variants , or the persistence of strains that belong to the known O139 lineage ., This will inform the management of future cholera epidemics ., This study was carried out in patients presenting to the icddr , b diarrheal hospitals in the Mohakhali and Mirpur neighborhoods of Dhaka , as well as from asymptomatic household members of the cholera patients between December 2013 to March 2014 ., These patients were enrolled either from the systematic surveillance system for enteric pathogens at the icddr , b Mohakhali hospital , through other ongoing cholera studies 9–11 , or through passive surveillance for cholera being conducted at these health facilities as part of a vaccination campaign with oral killed cholera vaccine Shanchol in Dhaka , Bangladesh 11 ., Demographic , socioeconomic , and clinical data were obtained from all study participants ., Trained study staff or hospital physicians performed a clinical examination of all study participants ., Study participants were assessed for degree of dehydration according to WHO guidelines 12 and treatment was provided according to the icddr , b protocols 13 ., All participants gave informed consent for collection of stool/rectal swab specimens ., Rectal swabs were collected from case 1 and case 3 because fresh stool specimens were not available ., Fresh stool specimens as well as stool swabs were collected from case 2 and case 4 ., All rectal swabs were placed in a Cary Blair medium and transported to the icddr , b laboratory at room temperature ., Two swabs were obtained from each patient ., In the laboratory , the first rectal swab taken from each patient was cultured directly on to taurocholate tellurite gelatin agar ( TTGA ) and the second swab was enriched in alkaline peptone water and incubated at 37°C overnight 14 ., After incubation and further culture on TTGA plates , suspected colonies resembling V . cholerae were tested by slide agglutination with monoclonal antibodies specific for V . cholerae O1 and O139 15 , as well as by biochemical assays ., Specimens that were positive for V . cholerae O139 were stored at −70°C and later examined by a multiplex PCR assay for concurrent detection of rfb sequences specific for O139/O1 genes of V . cholerae and for ctxA-specific sequences 16 ., Toxigenic V . cholerae O139 ( CIRS 134B ) and V . cholerae O1 El Tor Inaba ( strain N16961 ) and classical Inaba ( strain 569B ) serotypes were used as positive controls for the multiplex PCR assay ., Strains were tested for antimicrobial resistance by disk diffusion method using azithromycin , ciprofloxacin , ceftriaxone , erythromycin , mecillinam , norfloxacine , nalidixic acid , trimethoprim sulfamethoxazole , and tetracycline ., The four V . cholerae O139 isolates were tested in the rabbit ileal loop assay 17 to detect fluid accumulation and enterotoxicity ., V . cholerae O1 strains 569B and N16961 were used as positive controls ., Detection of cholera toxin ( CT ) was performed by ganglioside GM1-specific enzyme linked immunosorbent assays ( ELISA ) 18 and differentiation of classical and El Tor biotype CT was made using MAMA PCR described previously 19 ., Multiplex PCR assays were performed on a Thermo cycler C-1000 instrument ( Bio-Rad ) ., Two sets of oligonucleotide primer pairs were used ., The first was O139 rfb-F ( 5´- AGCCTCTTTATTACGGGTGG-3´ ) , O139 rfb-R ( 5´-GTCAAACCCGATCGTAAAGG-3´ ) , and the second one was ctxA-F ( 5´-CTCAGACGGGATTTGTTAGGC-3´ ) , ctxA-R ( 5´TCTATCTCTGTAGCCCCTATTA-3´ ) ; these pairs were used to amplify O139 rfb ( amplicon size 449 bp ) and ctxA ( amplicon size 302 bp ) genes respectively using previously described procedures 16 ., The product was analysed on 1% agarose gel using Gel Red ( BioTium , USA ) stain for visualization ., Genomic DNA was extracted from eight V . cholerae strains collected in Bangladesh; Strain 5 , Strain 6 , Strain 7 and Strain 8 in 1993 and Strain 9 to 12 in 2002 as well as the four V . cholerae O139 isolates collected in 2013 and 2014 ., Genomic DNA was prepared by incubating a fresh V . cholerae colony from a gelatin agar plate into 5 mL of LB broth with overnight shaking at 37°C at 150 rpm ., Genomic DNA was extracted with a DNA pure extraction kit ( QIAGEN , Germany ) according to the manufacturer’s recommendations ., Specimen DNA was stored at -70°C and shipped in dry ice to the Wellcome Trust Sanger Institute for sequencing and whole genome analysis ., Isolates were sequenced as multiplexed libraries on an Illumina MiSeq machine , producing 150 nucleotide paired-end reads ., Whole genome sequence analysis on the four V . cholerae O139 isolates collected in 2013 and 2014 was carried out and compared with data from the eight strains collected in Bangladesh in 1993 and 2002 ( S1 Table ) ., The data generated were combined with previously published V . cholerae genome sequence data 7 from representative O1 seventh pandemic El Tor global wave I , II and III strains , as well as three wave II O139 seventh pandemic El Tor isolates , and used to construct a whole genome single nucleotide polymorphisms ( SNP ) -based phylogeny ., To achieve this , reads for all isolates were mapped to the V . cholerae O1 El Tor strain N16961 ( accession AE003852/AE003853 ) reference sequence using SMALT v0 . 7 . 4 20 , and with GATK for indel realignment 21 ., SNPs were called using a combination of SAMtools 22 mpileup and BCFtools as described previously 23 ., SNPs falling in regions identified as being recombinant and so not likely to reflect the underlying phylogeny of the bacterium were excluded from this analysis as described in Croucher et al . 24 and a phylogenetic tree was drawn using the non-recombinant SNPS with RAxML 25 ., Draft de novo genome assemblies were created using Velvet 26 and scaffolded using SSPACE 27 and Gap Filler 28 ., The hospital surveillance activities of icddr , b were approved by the Research Review Committee ( RRC ) and Ethical Review Committee ( ERC ) of icddr , b ., According to the icddr , b hospital surveillance system , we only require verbal consent from patients undergoing routine investigation for collecting stool specimens ., Consent was documented in the surveillance questionnaire in the hospital surveillance system ., Consent was also obtained in accordance with other ongoing studies approved by the RRC/ERC of icddr , b ( # PR-10061 , # PR-11041 ) Based on the above , verbal consent was obtained from one study participant ( case 3 ) while written informed consent was obtained from two cholera patients ( cases 1 , 2 ) and one asymptomatic household contact ( case 4 ) ., The socioeconomic status of the four cases was lower middle class and all were individuals who lived in high risk settings in urban slums in and around Dhaka city ., The annual income of the adults ( and the parent of the child; case 2 ) ranged from 15 , 000–30 , 000 Taka ( ~USD 200–300 ) ., Two of the participants worked in garment factories while the other two families were self-employed in small businesses ., All of the cases reported that they consumed stored tap water in their homes or workplaces and shared kitchens and toilets in the community ., The four V . cholerae O139 isolates ( strains 1–4 ) collected in this study , having tested strongly positive for the O139 lipopolysaccharide ( LPS ) O-antigen by the rapid dipstick assay , according to the manufacturer’s recommendations ( Span Diagnostics Ltd . , India ) , were further characterized ., The serogroup was further confirmed: all four O139 isolates were found to be positive for the O139 rfb gene ., PCR was used to assay for the presence of the cholera toxin gene , ctxA ., Strains 2 , 3 and 4 were found to be positive for ctxA , all of which when sequenced were characteristic of the classical biotype ( see methods ) ., The production and type of toxin was further confirmed by ELISA ., The serogroup O139 isolates taken from case 1 ( strain, 1 ) was negative for both the classical and El Tor biotype of cholera toxin by both PCR and ELISA ., The rabbit ileal loop assay to detect fluid accumulation and enterotoxicity showed that three O139 strains were strongly positive for toxin production ( strains 2–4 ) , while the isolate from case 1 ( strain, 1 ) failed to induce fluid accumulation and enterotoxicity in the rabbit ileal loop ., All four V . cholerae O139 isolates were found to be resistant to nalidixic acid but were sensitive to all of the other antibiotics tested ., To understand the detailed genetic relationships between the O139 isolates taken from these four patients , we extracted the DNA and sequenced their genomes ., For comparison , we also sequenced an additional eight O139 isolates taken in Bangladesh in previous outbreaks in 1993 and 2002 ., For the isolates obtained from cases 2 , 3 , and 4 , 95 . 6% of their sequence read data mapped to the genome of the V . cholerae pandemic 7 strain N16961 serogroup O1 reference sequence ( S1 Table ) ., These three genomes differed by between 112–114 ( 313–316 before removing putative recombination ) single nucleotide polymorphisms ( SNPs ) from the reference sequence ., When compared to the O139 strain MO10 ( accession AAKF03000000 ) mapped to the N16961 reference sequence , there were between 46–47 SNPs differentiating MO10 from these three isolates ., To determine the phylogenetic relationship of the V . cholerae O139 isolates we constructed a whole genome core phylogeny from these three V . cholerae O139 isolates taken from cases 2–4 along with the O139 isolates collected in previous years , and including those previously described 7 ., The sequence data are deposited in the European Nucleotide Archive with accessions ERS452533 –ERS452544 ., The phylogenetic relationships of the isolates sequenced in this study , with the exception of that from case 1 , are consistent with previous data 7 ., Fig 1 shows that the majority of the isolates fell in the O139 branch of the seventh pandemic El Tor phylogenetic tree , along with the previously published O139 sequences from India and Bangladesh ( accessions ERS013124 , ERS013125 , AAKF03000000 ) and the majority of isolates sequenced in this study ., The new O139 sequences , with the exception of the 2013 isolate from case 1 , cluster with a strong temporal signature with isolates in the previous study 7 including the MO10 isolate from India in 1992 ., Whole genome analysis of the non-toxigenic 2013 strain ( isolated from case, 1 ) was distinct , with only 83 . 6% of the sequence reads mapping to the N16961 reference genome , and showing 98 , 743 and 97 , 313 ( approximately 124 , 200 and 122 , 400 before removing putative recombination ) SNP differences when compared to N16961 or MO10 using the mapping to the N16961 genome , respectively ., The genome sequence data showed that the case 1 isolate ( strain, 1 ) lacked the ctxAB genes but possessed the O139 specific rfb gene , confirming the PCR and phenotypic results described above ., Based on the mapped genome data , this isolate was highly divergent from all other sequenced O1 and O139 isolates in this study and described previously ., Although this isolate was confirmed as belonging to the O139 serogroup using traditional techniques , sequence analysis showed that the genes for the O-antigen biosynthesis genes of this isolate were different from both the O1 and other O139 isolates ., To investigate this in greater detail , the 84 contigs of the previously sequenced O139 cholera isolate MO10 were ordered against the O1 N16961 reference using ABACAS 8 ., The Artemis Comparison Tool 29 was used to compare the ordered MO10 genome against the case 1 genome; the contiguated sequence of one area of the LPS operon in the case 1 isolate was used to correct the ordering of one of the MO10 contigs ., A search in the NCBI database using blastn 30 of the LPS operon sequence of the MO10 isolate found a hit against a 46 . 7kb sequence of Vibrio cholerae genes for O-antigen synthesis , O139 strain MO45 ( accession AB012956 . 1 ) ., Comparative sequence analysis of the O139 O-antigen biosynthesis genes in the case 1 isolate showed that it was distinct from those within the O1 N16961 and O139 MO10 isolates , MO45 O-antigen biosynthesis genes , and the O-antigen biosynthesis genes from the genome obtained from case 3 in this study ( Fig 2 ) ., We report the isolation and characterization of four isolates belonging to the V . cholerae serogroup O139 in an eleven month period between December 2013 and March 2014 in Dhaka , Bangladesh ., This represents the first report in Bangladesh since 2005 of clinical cases of cholera caused by V . cholerae O139 infection ., This included two patients who presented with acute diarrhea who were ultimately hospitalized ., To put these four isolates in context , icddr , b conducts epidemiological and ecological surveillance for cholera in different parts of Bangladesh ., Between 2010–2012 , 500 clinical and environmental V . cholerae strains were isolated , 496 were confirmed as O1 and four as V . cholerae O139; all of those four previous O139 isolates were obtained from environmental samples 5 ., Given the association between V . cholerae O139 and previous epidemics , the persistence and newly identified sporadic cases of both toxigenic and non-toxigenic V . cholerae O139 in the environment and in symptomatic and asymptomatic infections is notable , and may have future implications for the diagnosis and prevention of cholera in this region ., Recently , a strain of V . cholerae O139 was isolated from a cholera patient in Beijing in China in May 2014 31 highlighting the continued low level presence of this lineage in different locales ., Interestingly , in two of our patients , V . cholerae O139 strains were isolated from asymptomatic household members of V . cholerae O1 infected cholera patients ., In our previous studies , we have shown that household contacts of an index case of cholera are approximately three times more at risk of infection with V . cholerae 11 ., However , we have previously identified non-O1/non-O139 isolates in household members of patients with O1 cholera 32 or even a different O1 Inaba or Ogawa serotype from that isolated from an index case ., The V . cholerae O139 strains presented here were only resistant to nalidixic acid and therefore differ from the V . cholerae O1 that currently predominate global infections , which are also resistant to trimethoprim-sulfamethoxazole and tetracycline 31 ., Whole genome sequence analysis showed that isolates from cases 2–4 fell on the O139 branch of the seventh pandemic El Tor phylogenetic tree , along with the previously published O139 sequences from India and Bangladesh ( accessions ERS013124 , ERS013125 , AAKF03000000 ) 7 ., These data also highlighted the existence of one isolate , from case 1 , that was typed genotypically and phenotypically as serogroup O139 , but phylogenetically represented a distant non-El Tor pandemic 7 V . cholerae lineage ., By comparing the LPS O-antigen operons it was apparent that this isolate , although highly divergent from the previously sequenced O139 isolates , possessed part of the O139 O-antigen gene cluster both targeted by the diagnostic O139 PCR test and which phenotypically appears sufficient to produce a O139 positive result by ELISA and the rapid dipstick typing methods ., Further work will be required to determine fully the significance of this subtype to human health ., At the time of writing this report , we had isolated two more strains of V . cholerae O139 from patients hospitalized with cholera between October and November 2014 ., These strains are being further characterized at present , and preliminary data suggest that they are phenotypically and genotypically similar to the isolate from case 1 . We are at present carrying out detailed analysis of these strains using genomic techniques ., In summary , our data suggest that V . cholerae O139 strains persist not only within the environment , but also are associated with occasional causes of acute watery diarrhea ., Since previous infection with V . cholerae O1 does not provide protection against O139 , and vice versa , our data suggest that O139 could re-emerge as a significant cause of cholera in areas where the pathogen persists ., Of note , oral killed cholera vaccine Shanchol is bivalent , and contains components of both O1 and O139 organisms , while other currently commercially available cholera vaccines are monovalent , providing protection against O1 alone ., We are continuing with the surveillance of patients with acute watery diarrhea for detection of V . cholerae O139 to monitor emergence of new variants and also to detect any new and reemerging outbreaks or epidemics using microbiological and genomic analysis ., This is extremely important for planning future strategies for immunoprophylactic preventive measures . | Introduction, Materials and Methods, Results, Discussion | Cholera is endemic in Bangladesh , with outbreaks reported annually ., Currently , the majority of epidemic cholera reported globally is El Tor biotype Vibrio cholerae isolates of the serogroup O1 ., However , in Bangladesh , outbreaks attributed to V . cholerae serogroup O139 isolates , which fall within the same phylogenetic lineage as the O1 serogroup isolates , were seen between 1992 and 1993 and in 2002 to 2005 ., Since then , V . cholerae serogroup O139 has only been sporadically isolated in Bangladesh and is now rarely isolated elsewhere ., Here , we present case histories of four cholera patients infected with V . cholerae serogroup O139 in 2013 and 2014 in Bangladesh ., We comprehensively typed these isolates using conventional approaches , as well as by whole genome sequencing ., Phenotypic typing and PCR confirmed all four isolates belonging to the O139 serogroup ., Whole genome sequencing revealed that three of the isolates were phylogenetically closely related to previously sequenced El Tor biotype , pandemic 7 , toxigenic V . cholerae O139 isolates originating from Bangladesh and elsewhere ., The fourth isolate was a non-toxigenic V . cholerae that , by conventional approaches , typed as O139 serogroup but was genetically divergent from previously sequenced pandemic 7 V . cholerae lineages belonging to the O139 or O1 serogroups ., These results suggest that previously observed lineages of V . cholerae O139 persist in Bangladesh and can cause clinical disease and that a novel disease-causing non-toxigenic O139 isolate also occurs . | Vibrio cholerae serogroup O1 is thought to be the sole causative agent for cholera in Bangladesh and most of the high risk developing countries ., Whilst historically Vibrio cholerae serogroup O139 has been seen to cause sporadic disease , the overall numbers of reported O139 clinical cases are low , with none reported in Bangladesh since 2005 ., Here we report four patients suffering from cholera attributed to serogroup O139 V . cholerae ., Cases 1 and 2 were symptomatic ( isolated strains 1 , 2 ) , and cases 3 and 4 were asymptomatic ( isolated strains 3 , 4 ) ., All cases were from urban Dhaka and represented a range of age groups ., Cases 2–4 presented with no sign of dehydration whereas case 1 showed some signs of dehydration ., Phenotypic and whole genome sequence data indicates that one of the four O139 V . cholerae isolates represents a novel O139 subtype ., Since natural infection with V . cholerae O1 or vaccination with currently available licensed cholera vaccines ( e . g . , Dukoral ) provides little protection against O139 , we conclude that V . cholerae O139 remains in circulation and is still causing a low incidence of cholera ., Therefore , further studies looking at the significance of these isolates towards the total burden of cholera in Bangladesh is warranted , including clinical evaluation , genome sequencing and immunobiochemistry . | null | null |
journal.pcbi.1000869 | 2,010 | Continuous Attractors with Morphed/Correlated Maps | Multiple state variables can be encoded in the same network ., An example is offered by the place representations of several environments 20 , 21 ., To each environment corresponds a neural map which is encoded in the synaptic efficacies ., Sensory inputs would then select the correct representation , i . e . both the environment and the position in the environment ., The selected map wins the competition with the other maps stored in the network , and a localized pattern appears ., In this case the network only maintains information about one of the several encoded state variables ., A more peculiar property of multiple continuous attractors , is their ability to represent simultaneously the values of several state variables ., This property was explored in 28 , where two partially overlapping neural populations ( representing discrete features ) , are assigned two uncorrelated maps ., Another example is provided in the study of 29 , where a single network stores and represents simultaneously a continuous and discrete attractors ., In principle , given the existence of multiple representations in different brain regions ( either one per region , or many in one region ) , a brain area downstream would necessarily encode several state variables ., In light of a Hebbian interpretation on how this encoding takes place , it seems natural to distinguish between two cases ., When multiple representations provide a simultaneous input to a region , the result is probably encoded multiplicatively 29 , or , in general , non-linearly ., For inputs happening non concurrently , as for instance when walking through several rooms sequentially , an additive encoding of each room is expected 21 ., In the following we will analyze additive encoding ., The present contribution addresses the issue of encoding correlated maps ., The motivations come from recent experimental results on place cells recording in morphed environments 30–32 , where place fields remapping along a sequence of morphed arenas is experimentally tested , and from theoretical and experimental studies concerning the morphing of discrete attractors 33–35 ., In general , we would consider the encoding of manifolds , each of dimension , where ., We will refer to a single manifold as a map , once a coordinate system is chosen ., The use of uppercase ( e . g . ) or lowercase ( e . g . ) will distinguish between the whole map and a single point on it respectively ., Given a pre-synaptic neuron indexed by , and a post-synaptic , the encoding of a single map is obtained using a synaptic matrix , and is such that a continuous attractor representation would arise if it were the only map ., We assume , as mentioned above , that the complete encoding arises from a linear superposition of the matrices , ., The statistical properties of the maps , and in particular the correlation between them , can be fully specified by providing the probability density ., The general problem is too difficult to be studied analytically ., Some results can be obtained for the case of uncorrelated maps on the same manifold 27 , though the system can be explored by simulating the full microscopic networks ( see e . g . 21 for the uncorrelated case and 36 for simulation results of the correlated case ) ., In order to simplify the analysis , while retaining the basic structure of the problem , we focus on the case of representations , on a 1-dimensional circular manifold ( i . e . the ring model 16 , 37 ) ., The correlation between the maps is constructed by limiting the distance between the single neuron locations on the two maps ., We devise a simple method to generate a morph sequence between two uncorrelated maps , by linearly modifying the neurons locations between the original maps ., This method also suggests a way to test the network response to the exposure of intermediate maps between the two stored correlated maps ., For concreteness , one could think about maps of two similar circular arenas , and reason in term of spatial coding ., In this context , we are interested in clarifying how the information about the position in the current environment is represented by the network , when varying the constitutive parameters of the model; And how the representation changes when the network is exposed to environments along a morph sequence ., In the following we will describe with mean-field ( MF ) theory the attractor landscape of a network , i . e . the stable solutions in absence of any place specific input ., We then consider the behavior of the solutions when a spatially tuned input is present ., We will establish the approximate relationship between two strongly correlated maps and the encoding of a morph sequence between two reference rings , and study the behavior of the solutions in presence of a tuned input varying along the sequence ., Finally we will verify the results with microscopic simulations of finite networks ., The network properties can be tested experimentally to confirm ( or falsify ) the attractor hypothesis ., In this Section we analyze the fixed point solutions of the system , and heuristically describe the region of stability of these solutions ., A more rigorous description of the stability can be found in Methods - Stability ., In Methods - Reduced dynamics we derive the dynamics of the order parameters from Eqs ., 2 . We report here the result ( 5 ) where the function is defined as ( 6 ) i . e . the rescaled steady state activity profile Eq ., 3 . Note that can be eliminated from the right hand sides of the Eqs ., 5 , rotating the integration variable ., This is possible because there is no spatial dependence in the external input to the network ., The first four equations in Eqs ., 5 can then be solved independently of the fifth one , since the right hand sides do not depend on ., We show in Methods - Solutions properties that , once we have the solution for the variables ( ) , the last equation reduces to ., We can thus restrict the analysis to four out of five equations in Eqs ., 5 ., The elimination of one angular degree of freedom is a consequence of the rotation invariant structure of the encoding , and is the hallmark of continuous attractors arising from spontaneous symmetry breaking ., On the other hand , the integrals over in Eqs ., 5 are not over the whole circle and we cannot rotate away ., Before analyzing the fixed point solutions of the system described by Eqs ., 5 , we briefly mention an uninteresting region in the parameters space which can be found also in the classical ring model ., This region corresponds to the homogeneous solution , i . e . all the neurons in the network are active at a constant level , and can be obtained from Eq ., 2 . The expression corresponding to the line of separation in the plane between the homogeneous solution and the spatially localized bump ( see Fig . 3A , curve surrounding the Homogeneous region ) , is ( 7 ) where ., This result is obtained in Methods - Stability , see also below ., Let us start by imposing , a restriction that will be addressed later on ., The first tree equations at steady state from Eqs ., 5 become then equations for the three order parameters : ( 8 ) The first two equations determine the shape of the bump ., Given the map specific modulation in the coupling and the distance between the maps , we can derive from the first two equations the size of the bump and the order parameter , representing how close the network representations are to the stored environments and ., The last equation gives us the amplitude of the network activity , which also depends on the parameter ., As mentioned in Results - Phase diagram of the model , the order parameter can be chosen arbitrarily , due to the rotation invariance of the problem; for simplicity we choose ., We deal first with the equation concerning the amplitude of the solution ., Given that the activity can be rescaled by changing the value of the applied external current , we are not interested in actually solving the equation ., The only requirement is that in order for the solution to be meaningful , i . e . no negative amplitudes are allowed ., This requirement translates to a constraint on the inhibition : ( 9 ) We show with stability analysis ( Methods - Stability ) that the critical value , obtained by choosing the equality in the previous expression , corresponds to the onset of amplitude instability; given a choice for the parameters , which specifies the bump shape , for values of the inhibition weaker than the solution grows to infinity ., This qualitative behavior was present also in the classical ring model ., Fig . 3B shows the values of as a function of for various choices of ., In order to stabilize the solutions , the inhibition must grow with increasing and decreasing ., Note that it is reasonable to consider the previously mentioned homogeneous solution as a bump with maximal size ., In this case the critical can be explicitly computed , and turns out to be ., Now we focus on the possible solution ., It is easy to see that when , the second of Eqs ., 8 is automatically satisfied due to the symmetry of the integrand in ( and ) ; This means that the solution exists everywhere in the parameter space ., The steady state activity Eq ., 3 with ( and , our initial assumption ) reads ( 10 ) which corresponds to a packet of activity localized in the coordinate , and modulated in , see Fig . 2B for a plot of the activity profile ., The remaining fixed point equation can be used to obtain ., We refer to the case as a single ring solution; the ring is spanned by the freedom of choice in the angle ., In this regime of activity the network is not able to represent separately the environments and , but only the middle environment described by ., Even though the solution exists everywhere , it is destabilized in some regions of the parameter space , as shown in the phase diagram ( Fig . 3A , Single ring region ) ., By looking at the maximal bump size , we can expect to reproduce the curve separating the homogeneous solution from the single ring ., Inserting in the first of Eqs ., 8 , it is possible in this case to compute explicitly the integral , which in fact yields Eq ., 7 ., In order to find the region of existence of the solutions with , we can solve numerically Eqs ., 8 in the parameters plane ., The result is shown in Fig . 4 , where the color code represents for a given choice of the parameters ., It can be seen that there is only a narrow region of high ( low correlation ) and low where such a solution exists ., It is important to note that the equations used to find are invariant under the symmetry ., This means that both solutions ( ) representing map or are possible ., The steady state activity profile in this case looks like: ( 11 ) Given the freedom of choice for the phase , each of this solutions lives on a ring; we call the solution , double ring ., An instance of the network activity in this regime is shown in Fig . 2C ., The curve separating representations preferring one of the two maps ( ) , and , can be obtained by expanding the second of Eqs ., 8 to first order in : ( 12 ) where is the Heaviside step function , and ., Dividing by , we get rid of the solution ., By finding the zeros of the integral , we select the curve in the parameter space corresponding to the onset of existence of the double ring solution ., This curve is shown in Fig . 4 ., We have found that the stability of the double ring solution coincides , empirically , with the region of existence of such solution ( compare the phase diagram in Fig . 3A , Double ring region with Fig . 4 ) ., Finally , we examine the meaning of the equation for , the order parameter linked to the location of the maximum of the bump in ., We have assumed for simplicity , given that a rotation in the integrands in Eqs ., 5 is in general not viable due to the restricted range of integration in ., Note though , that when the size of the bump is small enough , it is possible to perform the rotation without affecting the value of the integrals; the only requirement is that the rotation keeps the bump from touching the boundaries ., In Methods - Solutions properties we verify that there are no solutions with both and different from ., We can therefore set in the steady state activity Eq ., 3 , and impose the activity itself to be zero on the boundary to findThis equation corresponds to the curve of separation in the plane ( using the relationship , Eq . 8 ) between the single ring solution and a cylinder solution ( Fig . 3A , curve surrounding the C region ) ., In this regime , in addition to the freedom of choice for the location of the bump in , the solution is also partially marginal in ., The bump can be freely moved on a segment and a circle , defining a cylinder; the activity profile in this case is described by Eq ., 4 , see an instance in Fig . 2D ., This region extends in the high limit and covers the whole range of correlations ., Despite the fact that each of the maps and defines a ring , it shouldnt come as a surprise that the topology of the attractor is a cylinder instead of a torus ., The correlation between maps gives rise by definition to a cylinder structure , as can be seen for instance by inspecting Fig . 2B , II ., It can be shown that when the cylinder solution degenerates in a torus; the bump of activity can be in any location of the coordinates ( hence , also in ) ) ., This regime is linked to the observation of an activity bump simultaneously localized in two environments in network simulations 39 , and the study in 28 ., Fig . 3 summarizes the results obtained so far ., When is low , the only solutions is a constant level of activity which spreads over the whole network ( Homogeneous region ) ., As is increased , the interplay between the short range excitation and long range inhibition creates a pattern of localized activity in the middle map ( Single ring , see also Fig . 2B ) or , if the correlation between maps is small enough , a localized pattern in either or ( Double ring , Fig . 2C ) ., Intuitively , the network “remembers” the two maps separately ( , two solutions ) if they are weakly correlated ( ) ., When the maps are more similar , the network represents just an average between them ( ) ., The bump size decreases with increasing ., When is further increased , instead of having a reduced size of the localized activity in just one of the maps , the presence of two stored maps in the synaptic structure and the inhibition produce a packet of activity which looks localized in both maps ( Cylinder solution , Fig . 2D ) ., Three particular values of the distance deserve a special mention ., The case , corresponding to the encoding of two identical maps , can be shown to be identical to the ring model 37 , as expected ., In particular , besides the homogeneous solution and the amplitude instability region , the system can only exhibit the single ring solution ., The case , corresponding to the encoding of two uncorrelated maps , does not have the single ring regime as a possible solution ., The double ring solution in this case is depicted in Fig . 2A , where it can be seen that the bump is perfectly localized in either maps or , lacking any spatial tuning in the other map ., This is the desired outcome in the “multi-chart” approach of 21 ., The third case is ., We will see in Results - Morphing maps that this case is closely related to the behavior of a network storing a morph sequence between two uncorrelated maps ., As can be seen in the phase diagram , the double ring solution is not possible in this regime ., How the environment , and the position in the environment , are represented by the network activity ?, For the single ring ( Eq . 10 ) and the double ring ( Eq . 11 ) solutions , both characterized by , it is evident that the position is coded by the order parameter ., The identity of the environment can only be represented with the ambiguity in the choice of the sign of when the network operates in the double ring regime ., In the cylinder regime , it is not clear how the information about the environment is represented in the network , since now the solution is described by and ., The following Section is mainly devoted to explore the link between the state variable ( eventually time-dependent ) in the active environment , and the behavior of the solution in this novel regime , by introducing a spatially tuned external input ., Until now we considered the condition in which the only external input to the network , , was steady and uniform ., Let us introduce a tuned input , for instance in map at position :For simplicity we assume the shape of the external input to be ., The parameter measures the strength of the tuned component of the external input as a fraction of the constant baseline we adopted so far ., In general what we are interested in , and what is experimentally observable , are the tuning curves of the neurons i . e . their profile of activity as a function of the input angle in the active environment ., It is easy to see the effect on the dynamics of the order parameters ( Eqs . 5 ) when the location specific external current is inserted in the original dynamics for the network activity , Eq ., 2 . The dynamics keeps the same form as in Eq ., 5 , with the exception of the threshold-linear term in , which now reads ( 13 ) where correspond to the choice of map in the input , and for map ., With the input at a constant location , one can see that a solution of Eqs ., 5 for the single and double ring regime ( ) , is , i . e . the input pinpoints the location of the bump ., This implies that , assuming a weak tuned input , the tuning curve of a neuron can be written in the single and double ring regime ( from Eqs . 10 , 11 ) as ( 14 ) and ( 15 ) respectively ., The tuning curve in the single ring regime has a maximum for ( hence is the preferred angle for a neuron ) , independently of which map is being used in the external input , as can be seen from Eq ., 14 ., This implies that each neuron has identical tuning curves in both environments , and that the preferred angle of a neuron does not coincide with either the assigned or but with their average ., For the double ring regime , the preferred angle assumes the form ( maximizing Eq . 15 in ) In this case each neuron has two different tuning curves according to the map used in the external input ., The preferred angles coincide with the assigned ones ( ) only when the stored maps are uncorrelated ( , hence ) ., In the cylinder regime ( , not necessarily ) , a solution for Eqs ., 5 in presence of a tuned input is ., For an input in map , , the tuning curve would then be proportional to ( from Eq . 4 ) Note that the dependence on means that the external stimulus does not determine completely the network activity , in contrast to what happens in the previously examined regimes ., Neurons that respond maximally to the tuned input are then , and , hence ., This means that the tuned external input pinpoints the location of the bump maximum in map but the bump is free to stabilize anywhere along the other map given the freedom of choice in ( see activity example in Fig . 2D ) ., If several randomly selected external locations in one of the maps are presented to the network , once at time and starting from random initial conditions , the tuning curves would be an average over :where the allowed range for is , see Methods - Solutions properties ., The cylinder regime extends the region of existence of two tuning curves per neurons to an higher correlation between the stored maps; the difference is that the coding becomes unreliable: during a single exposure to a given value of the input angle , a neuron could remain silent even if its average tuning curve would predict a response ., When the representation refers to the location in an environment , it is natural to think about a smoothly varying location ., With a moving input like , the tuning curve depends as before on which map is stimulated , but in a novel way ., Assume for simplicity to start from a initial condition , corresponding to ( ) ., A moving input in the map would tend to move the bump along that map ( i . e . increase the of the solution ) , while keeping constant ( hence the bump will move to ) ., This movement is possible only until the bump reaches the part of configuration space not occupied by neurons due to the distance between maps , see Fig . 2D ., At that point , the bump will start to move equally along and , maintaining , which is proportional to , and increasing ( proportional to ) ., A similar scenario , but with , is obtained when stimulating the map ., If the size of the bump is sufficiently small , this effect has dramatic consequences ., The small bump will move along neurons with when a moving stimulus is presented in environment , and viceversa neurons with will be active only when the moving stimulus is presented in environment ., As a consequence , neurons will essentially just have a tuning curve ( or field ) , only in one map , and will be silent in the other one ., We refer to this phenomenon as dynamical pattern separation ( see Fig . 5 for an example ) ., The separation of the activity patterns is essentially a dynamical phenomenon , dependent on the history of the inputs ., The figure shows also the robustness of the dynamical pattern separation behavior to the addition of Gaussian -correlated noise in the external current ( see Methods - Numerical Methods ) ., Note that neurons characterized by ( i . e . ) , will have tuning curves in the same location ., The number of neurons with tuning curves in both environments grows with the size of the bump ., Note though that by changing the sign of the velocity in the moving input , the behavior would reverse; neurons with positive ( negative ) would be active during a stimulation in map ( ) ., In order to maintain the dynamical pattern separation and the analogy with place coding , one could think about two circular environments , as we did so far , with the additional constraint that the environments can only be traveled , for instance , in the counter-clockwise direction ( CCW ) ., As an alternative , the two environments may be thought as the same circular arena , but traveled clockwise ( CW , environment ) and CCW ( ) ; this interpretation would give rise to place fields with directional selectivity ( see Discussion ) ., The dynamical pattern separation is basically dependent on the history of the input ( positive or negative velocity ) , in addition to the identity of the map used in the stimulation ., This history dependence is present also for non smooth time-dependent stimuli , as for instance the sequential presentation of stimuli with an intervening delay period ., In this case the history dependence gives rise to a memory effect: the current location of the bump following a stimulation depends on the location attained after the previous stimulus presentation ., Let us consider a basic example of this phenomenon , where the tuned external input is always presented in map ., Consider for simplicity the state of the network being characterized by , as a result of the presentation of stimulus sometime in the past ., If we now present a stimulus , the bump will move , through the shortest arc on the map , to the new location ., Depending on the stimuli , this movement can happen in two ways ., If the shortest arc from to is directed CCW , the bump will move with a positive velocity and will end up being located in the region ( as we previously saw in the case of moving tuned input ) ., If the shortest arc is directed CW , then the movement will happen with a negative velocity , and the final location of the bump will be in the region ., Hence , by looking at the activity resulting from the presentation of , we know whether the shortest way on the ring to it from is CW or CCW ., A similar result can be obtained if the stimulus presentation alternates between map and ., If we vary the manifolds on which the maps live , for example to segments instead of circles , the history dependence changes accordingly ., For instance , on segments the activity would give us information about the second stimulus being greater/smaller than the first one ( see Discussion ) ., In the next section we present a simple ( albeit artificial ) delayed discrimination task which the network can perform by exploiting the memory effect ., Let us suppose to have a screen with a circle on it ., A first stimulus ( a dot ) appears on the circle at some random location ( described by an angle , ) , for the duration of ., This first stimulus is then removed for a delay period of ., Then a second stimulus appears at another random angle ; the subjects task is to determine whether the shortest path on the circle from angle to is CW or CCW ., The basic idea is that it is enough to look at the network activity ( location of the bump in the axis ) , to determine the relationship between the first and the second stimulus ( see Results - Tuned external input for a description of the idea ) ., To test the ability of the network to solve this task , we numerically solve the dynamics for the order parameter ( Results - Phase diagram of the model ) with an external input ( Results - Tuned external input ) mimicking the presentation of the stimuli , for a sequence of trials ., We used no inter-trial interval , i . e . the presentation of the second stimulus in the -th trial is immediately followed by the presentation of the first stimulus in trial ., The time courses of the bump location on the axis ( ) in two example trials for which , are shown in Fig . 6A ., When looking at the location of the bump in the axis at the end of a trial , there is a clear difference between the two cases of shortest CW , corresponding to positive ( in the specific example ) , or CCW arcs ( , where ) ., Fig . 6B shows that the bump location at the end of trial , can be used to easily discriminate between the two possible answers ( except for the cases in which the first and second stimuli are relatively close to each other ) ., Note that this result has been obtained without any activity reset to new initial conditions during the inter-trial intervals ., How do the results described so far change when , instead of storing just two correlated maps , the network encodes a sequence of maps gradually morphed between two uncorrelated ones ?, Let us start by constructing two random uncorrelated maps , and ., We would like to define the intermediate maps as gradual rotations between the two extreme ones; since we are dealing with circles , the rotation should be performed along the shortest arc between and ( see Eq . 21 , Methods - Inverse transformation ) ., We assume here to have already transformed the variables in such a way that we can write directly ( 16 ) where indexes the maps along the morph sequence ., Hence a neuron with label in the first map , will rotate along the sequence to its location on the last map , following the shortest path on the circle ., With this choice of the morphing procedure , each neuron is still characterized by just two quantities , its labels in the extreme maps ., We store the whole morph sequence by a superposition of the synaptic structures generated in each map separately , as for the case of two correlated maps previously described ., For the sake of analytical tractability , we study the resulting coupling in the limit ( 17 ) Introducing the definition of two uncorrelated maps ( Eq ., ( 1 ) with ) into Eq ., ( 16 ) , we can rewrite the angles in the intermediate maps as , We can now integrate Eq ., ( 17 ) ( 18 ) Making use of the Euler formula for the functionit is possible to deriveThe first term of the infinite product in the Euler formula , or the first term in the limit sum , gives us ., Comparing the coupling in Eq ., ( 18 ) , and the one derived for two maps , Eq ., ( 2 ) , we see that to first order , the synaptic coupling induced by the storage of the whole morph sequence , is equivalent to the storage of two correlated maps with ., In Fig . 7 , we compare the network activity generated by the approximated coupling and the full result of Eq ., 18 , when the external input is constant ., The results are qualitatively similar but the full morph case reaches the cylinder regime for lower compared to the case ., Note that the network storing the morph sequence shows the same dynamical pattern separation observed in the two maps case ( Fig . 8 ) , see next Section for a simulation example in a finite network with a finite number of encoded maps ., The important difference , is that while the very correlations between maps forced the absence of neurons with certain labels , hence constraining the permissible region for a marginal solution in , here the neurons cover the entire ( ) space ., The result is purely due to the process of storing multiple maps along the morph sequence ., This morphing algorithm also yields a way of stimulating the network with positions in environments intermediate between and ( with or without the intermediate maps encoded in the network ) ., It is sufficient to use as a place specific input what we had in Eq ., 13This time , the suitable range for the variable indexing the morph sequence is the whole range , if using as an approximation for the morphed case , or the restricted if the network is storing just two correlated maps ., In the reference frame defined by the original coordinates ( ) , a change in the stimulated environment corresponds to a rotation of the axis representing the maximal external input; between a vertical axis ( stimulus localized in environment , to an horizontal axis , stimulus localized in environment . ) In the experiment of 32 , the rat is trained until it develops two separate place coding for a single arena with different light configurations ( representing two distinct environments ) ., The advantage of this setup is that it allows , for instance , to slowly morph the light configuration between the two environments familiar to the rat ., The experimental results shows a sharp transition around the middle of the light morphing ( lasting ) between the place representation in light configuration and ., A link to these experimental results is provided by the use of time-varying external environment , where represents the duration of the morphing and denote the upper and lower bounds of the range ., An example usage of this protocol is shown in Fig . 8 for the approximated whole morph sequence storage , for two slightly correlated maps in the cylinder region of the parameter range and for the double ring regime ., For each run we show the dynamics of the relevant order parameter for the regime under consideration , for the double ring case and for the cylinder solution ., In addition , we numerically solve the dynamics for a moving stimulus in either environment or ., We use this as a reference for computing , at each time step , the correlation coefficient between the network activity during the morphing protocol and the activity in the fixed environment ., The transition is sharpest for the storage of two slightly correlated maps ., Note that similar results would be obtained by testing the network separately in each environment of the sequence ( see e . g . 31 ) ., The sharp transition is maintained when increasing the amplitude of the external tuned input , because a small tilt in the tuned input towards either map or is sufficient to generate the dynamical pattern separation described in the previous Section ., The transition in the cylinder regime occurs few seconds later than the one occurring in the double ring regime , which in turn happens in the middle of the morphing ( ) ., This delay is due to the time required for the bump to move from the region of to , or viceversa ( see also Fig . 5B ) ., This result could be compared with the experimental results of 32 ., The delay does not occur when testing the network in separate environments along the morph sequence ., There are two additional observations to be made ( data not | Introduction, Results, Discussion, Methods | Continuous attractor networks are used to model the storage and representation of analog quantities , such as position of a visual stimulus ., The storage of multiple continuous attractors in the same network has previously been studied in the context of self-position coding ., Several uncorrelated maps of environments are stored in the synaptic connections , and a position in a given environment is represented by a localized pattern of neural activity in the corresponding map , driven by a spatially tuned input ., Here we analyze networks storing a pair of correlated maps , or a morph sequence between two uncorrelated maps ., We find a novel state in which the network activity is simultaneously localized in both maps ., In this state , a fixed cue presented to the network does not determine uniquely the location of the bump , i . e . the response is unreliable , with neurons not always responding when their preferred input is present ., When the tuned input varies smoothly in time , the neuronal responses become reliable and selective for the environment: the subset of neurons responsive to a moving input in one map changes almost completely in the other map ., This form of remapping is a non-trivial transformation between the tuned input to the network and the resulting tuning curves of the neurons ., The new state of the network could be related to the formation of direction selectivity in one-dimensional environments and hippocampal remapping ., The applicability of the model is not confined to self-position representations; we show an instance of the network solving a simple delayed discrimination task . | How is your position in an environment represented in the brain , and how does the representation distinguish between multiple environments ?, One of the proposed answers relies on continuous attractor neural networks ., Consider the web page of your campus map as a network of pixels ., Every pixel is a neuron , and nearby pixels excite each other , while distant pairs are inhibited ., As a result of their interactions , a bunch of close-by pixels will light up , indicating your current position as suggested by your web-cam ( the sensory input ) ., When you travel to another campus , the common assumption holds that pixels are completely scrambled and the excitatory/inhibitory pattern of connections is summed to the existing one ., Now these connections and the sensory input will activate the pixels corresponding to your location in the new campus ., The active pixels will look like noise in the old map ., But what if the campuses are similar , i . e . the pixels are not completely scrambled ?, We show that the network has a novel way of distinguishing between the environments , by lighting up distinct subsets of pixels for each campus ., This emergent selectivity for the environment could be a mechanism underlying hippocampal remapping and directional selectivity of place cells in 1D environments . | neuroscience/theoretical neuroscience | null |
journal.pcbi.1005249 | 2,016 | Improved Prediction of Non-methylated Islands in Vertebrates Highlights Different Characteristic Sequence Patterns | DNA methylation is known to play an important role in vertebrate gene regulation 1 , 2 ., Most of the human genome is usually methylated , however over 30 years ago a relatively small number of non-methylated regions were identified using methylation-sensitive restriction enzymes 3 ., These non-methylated regions were found to have a higher than expected number of CpG dinucleotides when compared to the rest of the genome , and it was suggested that this is because methylated CpGs are more likely to be mutated to CpAs or TpGs than non-methylated CpGs , leading to the reduction of CpG dinucleotides in most of the genome 4 , 5 ., Due to the increasing availability of genomic sequence data this sequence-based definition of non-methylated regions became popular , and they have come to be referred to as CpG islands ., One of the first sequence-based definitions of CpG islands was proposed by Gardiner-Garden at al ., 6 , which defines CpG islands as regions of the genome that have a length of >200bp , GC content >50% , and a ratio of observed CpGs to expected CpGs ( or CpG Ratio ) >0 . 6 ., A variant of this method is still used to provide an annotation of CpG islands in the popular UCSC Genome Browser 7 , and for many years it has been used as a proxy for non-methylated regions of the genome ., CpG ratios are also used in vertebrates to classify genes into those with high CpG promoters and low CpG promoters , due to the observation that there is a clear bimodal pattern in the CpG ratios of human promoters 8 , 9 ., This bimodality is also observed in several other vertebrates including humans , chicken , frog and zebrafish 10 ., However , the overall percentage of promoters that overlap a CpG island was later shown to be much lower in cold-blooded vertebrates ( <20% ) when compared to warm-blooded vertebrates ( >40% ) 11 ., This suggested that very few promoters are unmethylated in cold-blooded vertebrates , or that the role of CpG-rich regions may differ between these two groups of organisms ., Later , tissue-specific methylated regions were identified and found to be correlated with tissue-specific gene expression ., One earlier study 12 showed that out of 150 differentially methylated regions that were studied , 100 of the regions overlapped with predicted CpG islands , many of which were shown to be methylated in most tissues ., More recent papers have reported similar results on a genome-wide scale 13 , 14 ., This dynamic methylation , even at CpG islands , made it clear that some disagreement between true unmethylated regions and CpG island predictions should be expected since CpG island predictions were not tissue-specific ., Additionally , fish CpG islands were found to be GC-poor though thought to be still have a high CpG observed/expected ratio 15 , and a study of CpG islands in mouse and human showed that the GC content of islands in the two organisms differs 16 ., These findings made it clear that a single model of CpG islands for all of these species would not be appropriate , and at the very least the cutoffs for calling a region a CpG island would need to be adapted when applying the prediction to other organisms ., Fortunately , experimental methods have now been developed that are able to identify methylated or non-methylated regions ( also called non-methylated islands , or NMIs ) in a given cell line or tissue genome-wide ., These methods include bisulfite sequencing 17 , 18 as well as the Bio-CAP method 19 , the latter of which is specifically for identifying regions that are unmethylated ., Recently , Bio-CAP was used to determine the location of NMIs in several vertebrates , including both warm-blooded ( human , mouse , chicken and platypus ) and cold-blooded ( lizard , frog and zebrafish ) vertebrates , in multiple tissues including testes and liver 20 ., It was noted that the number of non-methylated regions that overlap with predicted CpG islands was in most cases quite low ( e . g . ≈ 20% overlap in Zebrafish ) , leading the authors of that study to conclude that the computational methods that are commonly used to identify CpG islands are not able to accurately identify NMIs in vertebrate genomes ., The same study 20 also showed that in contrast to the low percentage suggested by CpG island predictions , over 50% of promoters in all vertebrates actually do have non-methylated regions at their TSSs , but that existing CpG island prediction methods are simply not able to identify them in cold-blooded vertebrates ., We investigated whether it is possible to use computational methods to predict the location of NMIs in the genomes of six vertebrates using only the DNA sequence itself ., The method we used differs from methods that are used for the computational prediction of CpG islands in two ways: first , we use newly available experimentally determined non-methylated island regions as training data and apply a supervised learning approach in order to learn from these regions ., By training our model separately on each organism and tissue , it is possible to predict tissue-specific NMIs from genomic sequence , despite the fact that genomic sequence is identical between tissues , as well as adapt to different organism specific mechanisms and genome-wide properties such as differing background GC content ., This was not possible when classical CpG island prediction methods were developed because experimental methods that could identify non-methylated regions in a large scale did not yet exist , so these methods had to rely on unsupervised learning approaches which do not leverage experimentally determined NMI regions ., Second , as a classifier we use a string kernel support vector machine ( SVM ) which allows us to easily use longer subsequences of DNA as features , rather than looking only at dinucleotide frequencies ., Support vector machines have been used successfully to identify other types of regulatory sequences 21 , 22 as well as in the prediction of DNA methylation itself 23–25 ., However , with the exception of a cursory analysis of NMIs 26 , this is the first genome-wide analysis of the predictive power of DNA sequence across such a wide range of vertebrate species using uniformly generated and analyzed data ., Our study reveals that there are strongly predictive sequence features that differentiate NMIs from the rest of the genome in all six organisms that we studied ( area under the ROC curve of a genome-wide classifier 0 . 91–0 . 99 ) ., In contrast to predicting NMIs using the regions’ CpG ratios only , the SVM is able to maintain a much lower FDR for all organisms , especially cold-blooded vertebrates ( lizard , frog and zebrafish ) , where CpG ratio-based predictions contain an overwhelming number of false positives ., Additionally , we show that longer sequence features ( approximately 6 bp subsequences or longer ) are required to predict NMIs in cold-blooded vertebrates , while shorter subsequences ( 2–4 bp ) are sufficient in warm-blooded vertebrates ., The most highly weighted sequence features are shown to contain a high number of C/G nucleotides in warm-blooded vertebrates , and a low number of C/G nucleotides in cold-blooded vertebrates ., Clearly , when NMIs are known for a given biological sample it is not necessary to predict them ., Therefore , besides the lessons we have learned from constructing the classifier in the first place , we also assessed the ability of our classifier to be trained on one organism and to predict the NMIs in another ., We show that cross-species prediction performs well across warm-blooded vertebrates , where the classifier is essentially an improved CpG island detector ., However , in cold-blooded organisms where we have shown that the sequence features required to identify NMIs are more complex , this cross-species prediction is more difficult ., It performs better in lizard than in frog and zebrafish , but none of the cold-blooded vertebrates have their NMIs predicted with the same level of accuracy as the NMIs in warm-blooded vertebrates ., To investigate how informative DNA sequence is in determining whether or not a given genomic region is a non-methylated island , we used a subset of each organism’s chromosomes to train an SVM to differentiate between known NMI sequences from testes and the rest of the genome ( see Methods section for details ) ., This classifier was then used to predict NMIs in the rest of the genome and its performance was evaluated ., This process was carried out separately for each organism , which allowed us to identify differences in predictive performance and important sequence features between the six species ., As a baseline , the performance of the classifier was compared to the observed versus expected CpG ratio of the sequence , as well as to two existing CpG island predictions: UCSC’s predictions based on the Gardiner-Garden and Frommer method 6 and a more recent hidden Markov model-based method 27 ., We compare the performance of the methods using ROC curves and Precision-Recall curves in order to understand the performance of the classifiers when classifying a given random genomic window as an NMI or not , as well as whether the classifiers can annotate the entire genome while controlling the number of false positives ., Our ROC curves show that our classifier was able to identify NMIs based on sequence alone with high performance ( see Fig 1 and Table 1 ) ., A simple CpG ratio is also quite predictive in humans , mice , and chickens ( AUROC 0 . 96–0 . 98 ) , though our SVM performs better in all cases ( AUROC 0 . 98–0 . 99 ) ., In cold-blooded vertebrates our SVM is able to improve NMI prediction even more markedly over the CpG ratio ( SVM AUROC 0 . 91–0 . 98 versus CpG ratio AUROC 0 . 76–0 . 88 ) ., The UCSC and Wu HMM methods are both very conservative with their predictions ., While they both maintain low false positive rates , they also have very low average true positive rates of between 7% and 49% ., The two methods also achieved higher true positive rates in warm-blooded vertebrates than cold-blooded vertebrates , with an average of 3 times more true positives being identified in warm-blooded vertebrates than in cold-blooded vertebrates ., Our Precision-Recall curves ( PRCs ) more clearly show that the SVM outperforms the other methods , and despite the fact that the ROC curves for the SVM and CpG Ratio looked quite similar the PRC shows that the SVM is much better at controlling false positives genome-wide ( see Fig 2 ) ., In the case of all cold-blooded vertebrates ( lizard , frog and zebrafish ) the CpG ratio-based predictions consist almost exclusively of false positives , regardless of the scoring cutoff that is used ., In contrast , the SVM-based method is still able to control the amount of false positives , though at an admittedly low recall ., The UCSC and Wu HMM methods both have high precision in warm-blooded vertebrates ( >0 . 71 in all cases ) , but their precision drops in cold-blooded vertebrates ., The UCSC method still manages an average of 0 . 52 precision in lizard and zebrafish , but this decreases to 0 . 15 in frog ., The Wu HMM method has lower performance , with an average precision of 0 . 3 in lizard and 0 . 15 in zebrafish ., The Wu HMM software was not able to produce a prediction for frog ( the software crashed repeatedly ) ., While some of the false positives which lead to low precision for many of the predictions could in fact be false negatives from the Bio-CAP method , we have no way of knowing if this is the case from the Bio-CAP data alone ., We next sought to investigate why CpG island prediction methods based on CpG ratios perform so poorly at predicting NMIs in cold-blooded vertebrates in comparison to their higher predictive accuracy in warm-blooded vertebrates ., Additionally , we wanted to understand why CpG ratios perform so poorly in cold-blooded vertebrates while our classifier is better able to control false positives ., To address these questions we looked into the performance of our classifier on the parameter tuning subset of the data , which contained the same number of windows for all organisms ( 30 , 000 ) and a fixed ratio of non-NMI windows to NMI windows ( 5:1 ) ., For all datasets we calculated the AUROC and AUPRC when using k-mers of increasing length ( see Methods section ) as features ( see Fig 3 ) ., The results show that short k-mers are already very predictive of NMI status in warm-blooded vertebrates ( AUROC >0 . 97 ) , but that longer k-mers are required for reasonable performance in cold-blooded vertebrates , especially frogs and zebrafish ., Using longer k-mers that led to a high AUROC in all organisms we observed that the 20 highest scoring k-mers were almost completely devoid of A/T nucleotides in warm-blooded vertebrates , while nearly every k-mer in cold-blooded vertebrates contained an A or T nucleotide ( see Table 2 ) ., Two example regions in frog and lizard are shown in Fig 4 , where in some windows overlapping an NMI the CpG ratio is low , the UCSC and Wu HMM methods both perform poorly , but the SVM classifier is still able to correctly identify the majority of the NMI region ., We also investigated how well a classifier trained in one organism can predict NMIs in another organism ., This is particularly useful because it would potentially allow us to improve NMI annotation in species whose genomic sequence is known but genome-wide methylation experiments such as Bio-CAP or whole-genome bisulfite sequencing have not yet been performed , which is the majority of species ., In order to make it easier to compare the results across organisms , we fixed the SVM parameters and used a k-mer length of 6 for all organisms ., This way we can attribute any differences that we observe to the differences between organisms rather than differences in parameters ., Additionally , after observing the clear grouping of samples into cold and warm-blooded vertebrates in the previous results , we decided to add two more training sets to the analysis: a set of sequences sampled from all warm-blooded vertebrate species ( human , mouse and chicken ) , and a set of sequences from all cold-blooded species ( lizard , frog and zebrafish ) ., As shown in Table 3 , the best predictions ( AUPRC ) are always achieved when training and testing on the same species ., The pooled samples which include the test organism consistently achieve the second best performance ., The one exception to this was in chicken , where the prediction based on pooled warm-blooded sequences was slightly better than the predictions using chicken sequences alone , though this difference is very small ( 0 . 003 AUPRC ) ., In general the warm-blooded vertebrates are all fairly well predicted by other warm-blooded vertebrates as well as by lizard ., A classifier trained on frog and zebrafish sequences performed relatively poorly when trying to predict NMIs in any of the warm-blooded vertebrates , though pooled cold-blooded sequences managed to achieve a high AUPRC in chicken ( 0 . 792 ) ., Cross-species prediction of NMIs in cold-blooded species was overall very difficult ., The only partial exception to this is the lizard , whose NMIs were predicted modestly well ( > 0 . 3 AUPRC ) by classifiers trained in any organism or set of organisms , with the exception of frog and zebrafish ., These results also show that best cross-species predictor for warm-blooded organisms is always the organism that has the most recent common ancestor: mouse and human are more closely related than chicken , and also achieve the highest cross-species predictive performance ., Additionally , the predictor based on pooled warm-blooded vertebrate species would be useful for predicting NMIs in other warm-blooded vertebrates as well as lizard , a species whose last common ancestor with the three warm-blooded organisms was roughly 300 million years ago ., This is not the case in the remaining cold-blooded vertebrates , where there is no trend of better cross-species prediction when training on more evolutionarily close organisms ., Our results demonstrate that highly informative DNA sequence features are contained within experimentally determined non-methylated regions of DNA in vertebrates , including cold-blooded vertebrates ., Using known non-methylated regions of the genome to both train and evaluate our methods , we were able to show that DNA sequence can be used as a predictor for non-methylated regions in all six vertebrates that were examined ., This was not entirely expected because existing sequence-based CpG island predictions have a low overlap with true non-methylated regions , suggesting that DNA sequence might not be highly predictive of non-methylated regions 20 ., The high performance of our classifier in all organisms proves that despite the low overlap with existing predictions , the use of a more complex model with longer subsequences as features and a supervised learning approach demonstrates that there are DNA features in all vertebrates that are predictive of NMI regions ., Secondly , these informative DNA features can be used to predict the location of non-methylated islands genome-wide better than existing CpG island prediction methods ., Given the fact that NMIs only make up 2–4% of the genome , only a classifier that can control false positives will be useful for predicting non-methylated regions genome-wide ., The precision-recall curves show that for all organisms we are able to predict NMIs better than all existing methods , and that for cold-blooded vertebrates we are uniquely able to identify a modest number of non-methylated regions while maintaining a low false discovery rate ., Additionally , we have shown that NMI predictions made by training a classifier in one species and predicting in another can be highly accurate when performed between warm-blooded species and to a lesser extent lizard , while cross-species prediction in zebrafish and frog proved to be very difficult ., One explanation for the differences we are observing in cross-species prediction accuracy is that NMI regions in all six vertebrates are identified based on two factors: first , simple CpG-richness , and second , more complex sequence features ( potentially including transcription factor binding sites ) that are species and/or tissue-specific ., These two factors have a different level of importance in each organism ., In warm-blooded vertebrates , the first factor is far more important than the second , while in cold-blooded vertebrates the second is more important than the first ., This is why a simple classifier using dinucleotide frequencies works so well in warm-blooded vertebrates , and why this simple classifier is easily transferable to other warm-blooded vertebrates ., In contrast , in zebrafish and frog CpG content still contributes to the prediction of NMIs , but the contribution of more complex sequence features is more important ., These complex sequence features are species-specific , and therefore do not transfer well between species , resulting in poor cross-species prediction accuracy in zebrafish and frog ., Lastly , the importance of the two factors in lizard is more balanced ., CpG content is more important than in other cold-blooded vertebrates , but complex sequence features are more important than in warm-blooded vertebrates ., This explains why we observe fairly good prediction of NMIs in lizard after training in warm-blooded vertebrates , which are identifying the simple CpG richness that is somewhat important in lizard NMIs ., On the other hand , the classifiers trained in other cold-blooded vertebrates have learned to identify complex species-specific sequence features and place less importance on CpG richness , and therefore perform poorly at predicting lizard NMIs ., The importance of longer , more complex sequence features , especially in frog and zebrafish , suggests that different mechanisms for the establishment and maintenance of non-methylated regions may be dominant in these organisms ., To investigate this , the k-mers that contribute the most to the classification of NMIs in each organism were compared to all known transcription factor binding motifs in the JASPAR 2016 vertebrates database 28 ., In warm-blooded vertebrates as well as lizard , the resulting enriched transcription factor motifs included a number of zinc finger proteins and other DNA binding proteins with GC-rich binding motifs ( e . g . SP1 , SP2 , and E2F family proteins ) , as had been previously observed 2 ., In the remaining cold-blooded vertebrates there were essentially no enriched motifs within NMI regions ( see data in S2 Table ) ., The important k-mers identified by the SVM did show clear differences between warm-blooded and cold-blooded organisms though , with the highest scoring k-mers containing almost no A/T nucleotides in warm-blooded organisms , and nearly every high scoring k-mer containing an A/T nucleotide in cold-blooded organisms ( see Table 3 ) ., This disagreement between the importance of the k-mers , which suggests a role for sequence-specific transcription factors , and the lack of known JASPAR motifs could suggest that the factors involved in establishing and maintaining NMIs in cold-blooded organisms may have changed along with composition of cold-blooded NMI sequences themselves ., And because of this our databases do not adequately reflect those transcription factor binding sites in cold-blooded vertebrates ., Overall , these findings are complemented by those presented in a new paper which was published during the review process of this article 29 ., In that paper the authors experimentally evaluated whether stretches of genomic sequence containing NMIs from one organism can also evade methylation when placed into another organism ., Their study demonstrated that transplanting a segment of genomic DNA from human into mouse , or from mouse into zebrafish , resulted in the majority of NMIs within these regions remaining unmethylated ., In particular , these findings show that the fish is capable of protecting mouse CpG islands from being methylated ., Nevertheless , the typical fish NMI sequence has evolved in the direction of the sequence patterns that we identified in our analysis ., This again raises the point that there should be a particular functional pressure that has led it to do so , for example transcription factors whose motifs are not currently known in cold-blooded vertebrates ., The locations of experimentally determined NMIs in the testes of six vertebrates were taken from a recent paper 20 , which we used as a gold standard set of true NMIs ., These NMIs were identified using the Bio-CAP method 19 , which provides a genome-wide estimate of the proportion of unmethylated CpG dinucleotides ., In comparison to another popular method for measuring methylation levels , whole genome bisulfite sequencing , Bio-CAP relies on affinity purification of non-methylated CpGs with a zinc finger CxxC domain , has lower resolution and in at least one set of experiments identifies a substantially lower number of non-methylated regions ( approximately 41 , 000 human NMIs with Bio-CAP 20 compared to 51 , 572 human NMIs with bisulfite sequencing 14 ) , suggesting that Bio-CAP results may contain substantial number of false negative regions ., Despite these potential shortcomings , this data set was selected because it is the only one available that covers a sufficiently wide range of vertebrate species in matching tissues ., The six vertebrates that we analyzed , along with their UCSC genome version , were Homo sapiens ( hg19 ) , Mus musculus ( mm9 ) , Gallus gallus ( galGal3 ) , Anolis carolinensis ( anoCar2 ) , Xenopus tropicalis ( xenTro3 ) , and Danio rerio ( danRer7 ) ., Reference genomes were downloaded from UCSC to match the versions used in 20 ., We omitted the platypus because its genome assembly is highly fragmented , consisting of over 200 thousand separate sequence fragments , and only approximately 17% of the organism’s genome could be mapped to a chromosome and ordered 30 ., Experimentally determined NMI locations from liver tissue in all six organisms were also analyzed and the resulting trends were the same as those for testes , so we omitted those results from the paper for the sake of conciseness ., CpG island predictions from UCSC’s Genome Browser were downloaded from the UCSC FTP site ., These predictions are based on a variant of the original Garden-Gardiner and Frommer method 6 , with a number of modifications: the length of the region must be ≥ 200bp , the percent of G or C nucleotides must be ≥ 50% , observed/expected CpG ratio must be ≥ 0 . 6 , and an additional running score is calculated that must remain above 0 for the entire length of the island ., The running score is computed by adding 17 for each CpG , and subtracting 1 for every other base ., Also , islands are cut in half at their maximum running score and each half is evaluated separately ( for details see http://genomewiki . ucsc . edu/index . php/CpG_Islands ) ., Additionally , CpG island predictions from Wu et al . 27 were downloaded from the paper’s website if available ( Gallus gallus , Mus musculus , Homo sapiens ) , or calculated from scratch using default parameters with the exception of Xenpous tropicalis , which repeatedly crashed and was therefore excluded from our analysis ., The method uses a hidden Markov model in 16 bp steps , with the surrounding 256 bp windows of the genome as observations , and models three states: Alu repeat elements , CpG Islands , and “baseline” ( the rest of the genome ) , with the regions belonging to the Alu state provided as known ., It should be noted that both methods use an unsupervised learning approach , meaning that they do not use a training set of genomic regions that are known to be either methylated or non-methylated ., This is because they were both developed prior to the creation of experimental methods that are able to identify large sets of regions which would be suitable for this training set ., As a proven supervised machine learning method we used a support vector machine with a spectrum kernel to predict non-methylated islands genome-wide ., A support vector machine ( SVM , 31 ) finds a boundary between two sets of data: positive and negative training data ., This boundary can then be used to classify new data points by checking to see which side of the boundary the data points fall on ., The spectrum kernel 32 is a string kernel that is known to perform well on many biological applications 21 , 33 , especially those where strings of genomic sequence are the primary input ., A spectrum kernel is one of the simplest string kernels , and as features it counts the frequency of k-mers ( k-length subsequences ) in each of the input sequences ., The similarity between two sequences is then the dot product of the vectors of these frequencies ., Feature weights ( w ) for each k-mer are calculated as outlined in the original spectrum kernel paper of Leslie et al . 32:, w = ∑ support\xa0vectors x i α i y i Φ ( x i ) ( 1 ), where α is the optimal solution to the SVM dual problem , yi is the NMI’s class ( -1 or 1 ) , and Φ is the spectrum of k-mers for the sequence xi ., The spectrum kernel implementation we used is contained in the Shogun machine learning toolbox 34 , and SVM solver that was used is SVMlight 35 ., Each genome was pre-filtered to remove regions that are not uniquely mappable using the GEM mappability program 36 ( version 20130406-045632 ) ., This was done because the gold standard Bio-CAP data is sequencing-based , and therefore is blind to NMIs that are in regions that are not uniquely mappable ., GEM was run with default parameters with the following exceptions , which were chosen in order to match the sequencing data and read mapping settings that were used when defining NMI peak regions 20: generate an index that includes the reverse complement DNA sequence , use a read length of 51bp , and allow a maximum of 2 mismatches ., We additionally removed a region of 500bp centered at the borders of each NMI from the analysis ., This was done because Bio-CAP is not a high resolution method , and therefore there is some uncertainty about the exact location of the NMI borders ., The border regions were removed so that we could avoid training the classifier on mislabeled regions , and to hopefully reduce the number of false positive and false negative regions in our data sets overall ., The performance of all NMI classifiers was evaluated using two metrics: the area under the receiver operating characteristic curve ( AUROC ) , and the area under the precision-recall curve ( AUPRC ) ., Both methods require a set of regions which have been scored according to their likelihood of being an NMI or not , as well as the known NMI class of the regions ., Receiver operating characteristic curves plot the true positive rate versus the false positive rate of the classifier as the score threshold that is used to determine whether a region is an NMI or not is varied across all possible values ., The area under this curve is then calculated and can be interpreted as the probability that a randomly selected NMI region will score higher than a randomly selected non-NMI region ., Values near 0 . 5 indicate that the classifier performs similarly to randomly selecting a class , and an AUROC of 1 . 0 is a perfect classifier ., Precision-recall curves were also used to evaluate our classifiers ., This is because ROC curves are known to be insensitive to class imbalance 37 , or the ratio of NMI regions versus the rest of the genome ., Because NMIs are quite rare in the genome , only making up roughly 2-4% of each organism’s genome , when trying to predict their location genome-wide we are in a situation where there is a severe class imbalance ., This is a difficult situation because a classifier that appears to perform well in the balanced setting might be nearly useless in the unbalanced setting , because it is very hard to control the number of false positives in such a situation ., To give a better idea of how practical these methods are for the genome-wide prediction of NMIs it is better to look at precision-recall curves , which plot the fraction of predicted NMIs that are true NMIs ( precision ) versus the fraction of all true NMIs that are predicted as NMIs ( recall ) as the score cutoff for calling a region an NMI is varied across all possible values ., This gives us an idea of the false discovery rate ( 1—precision ) of the method at different score thresholds and whether the genome-wide predictions the classifier makes are likely to contain many false positives ., Parameter tuning , model training and performance evaluation was carried out separately on each organism as explained below ., Two exceptions to this is in the calculation of the top 20 k-mers for each organism , and the cross-species predictions , in which case the k-mer length parameter was fixed at 6 and SVM soft margin penalty parameter C fixed to 1 . 0 to make the results more easily comparable across organisms ., In order to select the appropriate parameters for our SVM and to evaluate its performance when identifying NMIs , we split each organism’s filtered genome into three sets: a parameter tuning set which was used to select optimal model parameters , a separate training set and a test set to evaluate the performance of the model on held out data ., A random subset of each organism’s chromosomes was set aside as the test set , such that approximately half of the organism’s genomic sequence was contained in the test set ., The remaining chromosomes were then divided into NMI and non-NMI regions , and those regions were split into 750 bp windows of which a random 50% were used for parameter tuning and the other 50% were used for training ., This window size was selected after considering the length distribution of the NMIs to ensure that the majority of NMIs would be longer than or equal to the window length ., In order to control for the differing length of each organism’s genome as well as the different ratio of background-to-NMI sequences , the tuning and training windows were subset to have a to | Introduction, Results, Discussion, Methods | Non-methylated islands ( NMIs ) of DNA are genomic regions that are important for gene regulation and development ., A recent study of genome-wide non-methylation data in vertebrates by Long et al . ( eLife 2013;2:e00348 ) has shown that many experimentally identified non-methylated regions do not overlap with classically defined CpG islands which are computationally predicted using simple DNA sequence features ., This is especially true in cold-blooded vertebrates such as Danio rerio ( zebrafish ) ., In order to investigate how predictive DNA sequence is of a region’s methylation status , we applied a supervised learning approach using a spectrum kernel support vector machine , to see if a more complex model and supervised learning can be used to improve non-methylated island prediction and to understand the sequence properties of these regions ., We demonstrate that DNA sequence is highly predictive of methylation status , and that in contrast to existing CpG island prediction methods our method is able to provide more useful predictions of NMIs genome-wide in all vertebrate organisms that were studied ., Our results also show that in cold-blooded vertebrates ( Anolis carolinensis , Xenopus tropicalis and Danio rerio ) where genome-wide classical CpG island predictions consist primarily of false positives , longer primarily AT-rich DNA sequence features are able to identify these regions much more accurately . | DNA methylation is one of the most heavily studied epigenetic modifications , with active research into the topic taking place for over 30 years ., Over that time a relationship between non-methylated stretches of genomic DNA and the regulation of nearby genes has been identified , and it was noted that approximately half of all human genes have an unexpectedly high number of unmethylated CpG dinucleotides in their promoters ., Nevertheless , until relatively recently we haven’t been able to directly measure methylation gnome-wide , leading to the widespread use of CpG-rich regions ( or CpG Islands ) as a proxy for unmethylated genomic regions ., Fortunately , recently developed experiments finally allow us to look at DNA methylation genome-wide , and have shown that CpG islands do not overlap true non-methylated islands to the extent we would have hoped , especially in cold-blooded vertebrates ., We show that by using these experimentally determined non-methylated regions as training data , we are able to improve the prediction of non-methylated regions genome-wide in comparison to classical CpG island prediction methods , in both warm-blooded and cold-blooded vertebrates ., Additionally , we show that while CpG content is an important predictor of methylation status in warm-blooded vertebrates , in cold-blooded vertebrates higher predictive accuracy can be achieved using longer stretches of relatively AT-rich DNA sequence . | amphibian genomics, vertebrates, neuroscience, animals, animal models, osteichthyes, reptile genomics, model organisms, amphibians, artificial intelligence, epigenetics, dna, mammalian genomics, dna methylation, chromatin, research and analysis methods, computer and information sciences, fishes, chromosome biology, gene expression, support vector machines, chromatin modification, dna modification, animal genomics, biochemistry, zebrafish, machine learning, cell biology, nucleic acids, genetics, biology and life sciences, genomics, cognitive science, frogs, organisms | null |
journal.pbio.1002265 | 2,015 | Neuromodulation to the Rescue: Compensation of Temperature-Induced Breakdown of Rhythmic Motor Patterns via Extrinsic Neuromodulatory Input | Maintaining neural function at different temperatures is a particularly difficult challenge for the nervous system since all biophysical processes are temperature dependent ., General knowledge about how this task is achieved remains rather limited , in particular since all biological processes , including those that govern signal transduction and neuronal excitability , vary substantially in their response to temperature 1 ., Central pattern generators ( CPGs ) are a class of neural networks that generate rhythmic activity patterns ., CPG activity has to be particularly resilient against perturbations because many CPGs drive vital behaviors such as respiration , swallowing , and locomotion ., CPG activity depends on the coordinated interplay between synaptic and cell-intrinsic ionic conductances 2 ., Many conductance combinations can give rise to rhythmicity , allowing networks to individually vary in conductance levels while remaining within the permissive conductance space for rhythmic activity ., It has been suggested that compensation of temperature perturbations may be achieved by keeping conductance levels within this permissive range via a balanced coregulation of cellular and synaptic properties that result in opposing effects on network output ., For example , phase constancy in the pyloric rhythm of crabs over a wide temperature range is accompanied by a balanced change of two opposing conductances ( Ih and IA; 3 , 4 ) ., In Aplysia , release of a neuromodulator that modulates muscle contraction drops 20-fold at higher temperatures , but this drop is partially counterbalanced by an increase in modulator efficacy 5 ., More recent studies indicate that neuromodulators may contribute to temperature compensation: Thuma et al . 6 show that dopamine modulation can restore muscle force after temperature-induced loss of muscle contractions ., This study tests the hypothesis that temperature compensation is conditional and under control of extrinsic neuromodulatory input fibers eliciting compensatory changes that oppose temperature-induced changes in intrinsic conductance levels ., For this , we use the well-characterized pyloric ( filtering of food ) and gastric mill ( chewing ) CPGs in the crustacean stomatogastric ganglion ( STG; Fig 1A; 7 , 8 ) , which , like most CPGs , are modulated by well-regulated extrinsic neuromodulatory pathways 9 ., The triphasic pyloric motor pattern is driven by a three-neuron pacemaker ensemble ( the single anterior burster AB and two pyloric dilator PD neurons ) that allows it to be continuously active with and without modulatory input 10 ., The phase relationship of the pyloric rhythm is maintained constant over a broad temperature range ( 7°C to 31°C; 3 , 4 ) ., In contrast , the gastric mill rhythm is two-phasic , episodic , and driven by half-center oscillations of Interneuron 1 ( Int1 ) and the lateral gastric ( LG ) neuron ( Fig 1A; 11 , 12 ) ., Rhythmic gastric mill activity requires modulatory input from descending projection neurons in the commissural ganglia ( CoG; 8 , 13 ) ., Modulatory commissural neuron 1 ( MCN1 ) , for example , mediates various sensory responses and elicits a robust gastric mill rhythm 7 , 12 ., We provide direct evidence that temperature compensation in the gastric mill network depends on a balanced change of circuit-intrinsic properties with opposing function that are regulated by descending modulatory input from MCN1 ., Modest temperature increase by 3°C led to a cessation of CPG activity that was caused by a concomitant increase in leak currents in LG ., Dynamic clamp-mediated subtraction of the leak was sufficient to rescue the rhythm and achieve temperature compensation ., Compensation was also achieved by stronger extrinsic neuromodulatory input from MCN1 or by augmenting MCN1’s influence with bath application of MCN1’s peptide cotransmitter C . borealis tachykinin-related peptide Ia ( CabTRP Ia ) 15 , 16 ., CabTRP Ia release from MCN1 activates a modulator-induced current ( IMI 17 ) in LG , and we show that this current effectively acts as a negative leak current to counterbalance the detrimental effects of the leak increase ., Thus , temperature compensation is under extrinsic neuromodulatory control , allowing conditional compensation of rapid temperature influences ., To test the role of circuit extrinsic neuromodulatory inputs for counterbalancing temperature-induced changes on CPG activity , we altered the temperature of the STG motor circuits but kept the CoGs at a constant temperature ., The CoGs contain descending projection neurons that provide extrinsic modulatory input to the STG circuits ., This approach is fundamentally different from previous studies 3 , 4 in which extrinsic neuromodulatory input from CoG projection neurons as well as STG motor circuits were affected by temperature changes ., Here , we thermally isolated the STG circuits from the rest of the nervous system by building a petroleum jelly well around the STG ., Extrinsic input fibers such as the descending CoG projection neurons remained mostly unaffected by these temperature changes ., However , the axon terminals of some projection neurons have local synaptic interactions within the STG 11 and may show ectopic spike initiation ., Temperature effects on these interactions were not investigated in this study ., CPG activity in the STG was recorded extracellularly on three different motor nerves containing the axons of pyloric and gastric neurons ( pyloric: PD , LP , and PY on the lvn; gastric: LG on the lgn and DG on the dgn ) ., We found that a moderate temperature increase from 10°C to 13°C had distinct effects on the two rhythms ( Fig 1B ) : the pyloric rhythm was resilient to temperature changes and continued its regular activity , while the spontaneous gastric mill rhythm terminated , as can be seen by the sporadic and nonrhythmic activity of the gastric mill neurons LG and DG on the lgn and dgn ( Fig 1B ) ., The pyloric rhythm had previously been shown to be “resistant” against temperature perturbations in vivo and in vitro ( although for much larger temperature ranges , up to 26°C or more; 3 , 18 ) in that the phase relationship of the pyloric neurons remains constant while cycle period decreases ., In all previous studies , however , temperature affected CPG as well as its input fibers , making it unclear whether extrinsic inputs from other parts of the nervous system are necessary to maintain the rhythm or not ., Despite the fact that in our experiments the temperature perturbation exclusively affected the STG , we found similar results for the pyloric rhythm as described previously ., In none of our experiments did the pyloric rhythm cease or show any obvious change from its canonical pattern ( see S1 Fig ) ., In fact , this was true even when we increased the temperature up to 19°C ( N = 6 ) ., To conclude , our results predict that the broad temperature range of the pyloric rhythm is likely to be intrinsic to the STG circuit and that the permissive temperature range of this rhythm is independent of temperature effects on other areas of the nervous system , such as the CoGs ., The gastric mill rhythm depends on the activity of upstream modulatory projection neurons in the CoGs 7 , 8 ., About 20 CoG projection neurons innervate the STG via the unilateral stomatogastric nerve ( stn , see Fig 1A , 19 ) ., Sensory input like olfactory or mechanosensory stimuli 20–22 as well as sensory feedback from proprioceptors 20 , 23 , 24 activates CoG projection neurons and elicits gastric mill activity ., Even individual projection neurons can start the gastric mill rhythm in vitro 12 and in vivo 25 ., One particularly well-characterized projection neuron is MCN1 11 , 12 , 25 , a bilaterally symmetric neuron in each CoG with axonal projections to the STG ( Fig 1A ) ., To study the mechanism of the temperature-induced breakdown of the gastric mill rhythm , we first decentralized the nervous system to remove the influence of all CoG projection neurons ., To initiate a gastric mill rhythm , we then stimulated MCN1 tonically with the lowest frequency eliciting a gastric mill rhythm at 10°C ( = threshold frequency , see Materials and Methods ) ., MCN1 was activated by extracellular stimulation of the inferior oesophageal nerve ( ion ) , which contains the axons of only two projection neurons , MCN1 and MCN5 ., MCN1 has the lower stimulation threshold of the two and can thus be activated selectively 11 , 12 ., For the analysis , we focused on LG since this neuron is part of the core pattern generator of the gastric mill rhythm and has a strong influence on all other gastric mill neurons: if spiking is prevented in LG , the gastric mill rhythm stops 11 ., Similarly to the spontaneous gastric mill rhythm , LG activity was rhythmic at 10°C during MCN1 stimulation but became substantially reduced and irregular at 13°C ( Fig 2A ) ., In four of ten experiments , LG spiking ceased completely ( Fig 2B ) ., This effect was reversible , i . e . , the rhythm returned to its original regularity and strength when temperature was decreased back to 10°C ., To quantify temperature effects on LG firing rate , we counted the number of LG bursts in 100 s and the number of LG spikes per burst ( see Materials and Methods ) ., We found that at 10°C , LG showed rather regularly spaced action potentials during the bursts followed by relatively long interburst intervals ( Fig 2B ) ., In contrast , at 13°C LG was either not active or its firing was erratic or tonic ., LG activity was never rhythmic at this temperature ( Fig 2B ) ., Concurrently , the number of LG bursts , as well as the number of LG spikes per burst , dropped significantly ( Fig 2C ) ., When temperature was returned to 10°C , in all preparations rhythmicity was recovered , and the number of LG bursts and spikes per burst returned to control values ( Fig 2C ) ., In these experiments , temperature was changed by ~1°C/min , and measurements were taken at 10°C and 13°C ., Given that physiological temperature changes might occur over a longer time period than in our experiments , we performed an additional set of experiments in which we slowly changed the temperature ( 1°C/h ) ., The goal of this set of experiments was to examine if homeostatic processes exist that compensate ( slow ) temperature perturbations ., Specifically , we kept the temperature at 10°C for 1 h , then slowly increased temperature by ~1°C/h , and recorded LG activity at 10°C and 13°C ., However , we found no obvious difference as compared to the faster temperature ramps used earlier: again , there was a significant decrease in the number of LG bursts and spikes per burst ( Fig 2D ) , demonstrating that the cessation of LG rhythmicity is not counterbalanced by homeostatic processes in vitro ., Next , we tested if the termination of the rhythm occurs abruptly or in a graded fashion ., Like in previous experiments , MCN1 stimulation frequency was determined at 10°C ., Stimulation was then stopped , and the temperature was lowered to 8°C ., Stimulation was then restarted to elicit rhythmicity , and temperature was continuously increased by 1°C/min until LG firing ceased ., We found that the number of LG spikes per burst continuously decreased linearly as temperature was increased ( Fig 2E , green trace ) ., In the example shown , all burst activity stopped as temperature reached 12 . 5°C ., The linear decrease in LG spike activity was consistent across preparations , and on average LG bursting stopped at 12 . 7 ± 0 . 2°C ( N = 5 ) ., Additionally , we tested the permissive temperature range for normal operation of this system by performing experiments with a broader temperature range ( 8°C to 16°C , N = 5 ) ., In these experiments , we increased the MCN1 stimulation frequency ( 175% threshold frequency ) to facilitate LG rhythmicity at temperatures above 13°C ., Again , we found that the number of LG spikes per burst continuously decreased in a highly linear fashion as temperature was increased ( Fig 2E , purple trace ) ., In the example shown , all burst activity stopped as temperature reached 15 . 8°C ., On average , this happened at 15 . 9 ± 0 . 4°C ( N = 5 ) ., The linear response to temperature changes indicates that the gastric mill CPG is unable to compensate moderate temperature influences ., The permissive temperature range , however , increased with higher MCN1 stimulation frequencies ., In summary , even a moderate temperature increase led to a consistent disruption of the gastric mill rhythm , which was in stark contrast to the robust behavior of the pyloric rhythm ., The pyloric and gastric mill rhythms share the same main function: digestion of food ., Proper digestion is vital for the animal’s survival , and it is intuitive to assume that both rhythms are equally important and that mechanisms exist to prevent cessation of both rhythms ., For the pyloric rhythm , it has been suggested that physiological temperature compensation is achieved by opposing temperature dependencies of membrane currents ( Ih and IA; 3 ) ., The gastric mill rhythm , in contrast , apparently lacks adequate compensation despite the fact that pyloric and gastric mill neurons are located in the same ganglion and comprise comparable ion channels and membrane currents ( e . g . , Ih and IA can be found in pyloric and gastric mill neurons; 26 , 27 ) ., To determine what provoked the termination of gastric mill activity , we asked whether intrinsic factors contributed to the observed temperature-induced changes in LG activity ., We first compared the intracellular response of LG to temperature changes: we found that LG’s resting potential hyperpolarized at 13°C ( Fig 3A ) , with an average drop of 2 . 67 ± 1 . 34 mV ( 10°C: −67 . 01 ± 3 . 00 mV , 13°C: −69 . 68 ± 3 . 50 mV , N = 13 ) ., Hyperpolarization was continuous and linear with temperature increase ( Fig 3A , right ) ., Also , LG spike amplitude decreased significantly by 4 . 73 ± 1 . 22 mV ( Fig 3B , 10°C: 17 . 02 ± 5 . 58 mV , 13°C: 12 . 28 ± 4 . 36 mV , N = 13 ) ., Next , we looked at the electrical postsynaptic potential ( ePSP ) , which LG receives from MCN1 12 ., For this , MCN1 was stimulated with frequencies that did not elicit gastric mill rhythms , but rather only individual ePSPs ( typically 1 Hz or below ) ., Fig 3C shows an example of the change in ePSP amplitude when temperature was increased ., On average , ePSP amplitude was reduced by 2 . 47 ± 0 . 54 mV at 13°C ( 10°C: 8 . 73 ± 3 . 06 mV , 13°C: 6 . 26 ± 2 . 52 mV , N = 13 ) ., All effects were reversible when temperature was decreased back to 10°C ., The MCN1 to LG gap junction has been shown to be voltage sensitive such that more hyperpolarized LG membrane potentials lead to smaller ePSPs ., This could have possibly contributed to the diminishment of the ePSP amplitudes ( since the resting potential hyperpolarized ) ., However , this effect only leads to an average ePSP amplitude change of 0 . 14 mV/1 mV membrane potential change 12 ., The temperature-induced change in ePSP amplitude in our experiments was almost six times larger ( 0 . 82 mV/1 mV ) ., Thus , the observed change in LG membrane potential was not sufficient to explain the diminished ePSP amplitudes ., The changes in resting potential , spike , and ePSP amplitude indicated that the input resistance of LG might have changed , causing a shunt of all of LG’s responses ., We found that input resistance decreased significantly by 4 . 12 ± 1 . 4 MΩ ( 34 . 34 ± 10% , 10°C: 12 . 03 ± 2 . 38 MΩ , 13°C: 7 . 92 ± 2 . 12 MΩ , N = 13 ) when temperature was increased ( Fig 3D ) ., Changes in input resistance can be due to changes in leak currents , voltage-gated currents , or synaptic input ., The latter appears unlikely to have contributed , since the STG was decentralized ., Decentralization removes most spontaneous activity of descending projection neurons that may cause synaptic input to LG , and it stops the gastric mill rhythm and silences or strongly diminishes the pyloric rhythm ., Consequently , because of the lack of STG activity in this condition , synaptic input from other STG neurons was also unlikely to have contributed to the observed change in input resistance ., To further reduce STG and projection neuron input to LG , we blocked action potentials with tetrodotoxin ( TTX ) ( 0 . 1 μM; N = 4 ) ., The result was the same: we obtained a decrease in input resistance when temperature was increased ., Hence , the decrease in input resistance was independent of synaptic input ., We also tested a broader range of current amplitudes by injecting 10 s long current pulses into LG , ranging from 1 to 3 nA in both the depolarizing and the hyperpolarizing direction ., Fig 3E shows that the voltage deflections of all current steps were smaller at 13°C than at 10°C ., This was true for all preparations tested ( N = 5 ) ., Fig 3F shows the change in LG membrane potential as a function of the injected current ., We noted a difference between LG’s voltage response to current injections in the positive and negative direction ., While this has not been reported directly before , it is most likely a result of the aforementioned voltage dependence of the gap junction between LG and MCN1 12 ., Importantly , the resulting skew of LG’s voltage response in the tested current range was small in comparison to the shunting effect of temperature increase ., Since many , if not all , processes in the nervous system are temperature dependent , a causal connection between temperature effects on a specific process and the output of a motor circuit is difficult to show ., Our data so far show that when temperature increases ( 1 ) leak conductance of LG increases , associated with ( 2 ) a hyperpolarization of LG’s resting potential ., To test whether either of these two effects or both could contribute to the termination of the gastric mill rhythm at higher temperature , we first tested the effects of a change in membrane potential ., For this , we recorded LG intracellularly and measured the membrane potential at 10°C and 13°C ., We then elicited a gastric mill rhythm via MCN1 stimulation at 10°C and hyperpolarized LG to resting potential values obtained at 13°C ( ΔVm = 2 . 88 ± 1 . 55 mV , N = 4 ) ., The gastric mill rhythm was not affected by this manipulation ., Thus , the observed change in membrane potential at 13°C was not a significant contributor to the termination of the gastric mill rhythm ., We next tested whether an increase in leak conductance is sufficient to explain the termination of the gastric mill rhythm by using the dynamic clamp technique 28 ., We either added an artificial leak conductance at 10°C or subtracted leak conductance at 13°C ., First , we measured LG input resistance and resting potential at 10°C and 13°C and used the difference to calculate the leak conductance increase ( = Δleak , see Materials and Methods ) ., We then elicited a gastric mill rhythm at 10°C , and after several gastric mill cycles , we turned the dynamic clamp on and injected the appropriate amount of additional leak ( +Δleak ) ., Immediately after the onset of the artificial leak conductance , LG bursting ceased ( Fig 4A ) ., Thus , an increase in leak conductance as caused by a temperature increase of 3°C was sufficient to terminate the gastric mill rhythm ., Consequently , a reduction of leak conductance at high temperature should also be sufficient to restore the rhythm ., We tested this prediction by carrying out the reverse experiment ( Fig 4B ) ., We stimulated MCN1 at 13°C with the threshold frequency that was sufficient to elicit a rhythm at 10°C ., As seen in our previous experiments , no gastric mill rhythm was elicited at 13°C despite the continuous MCN1 stimulation ., We then turned on the dynamic clamp and subtracted the appropriate leak ( −Δleak ) ., Immediately , LG regained its spiking ability , and rhythmicity was restored ., In two out of the four experiments , LG firing stopped completely when artificial leak was added at 10°C ( Fig 4C ) ., In the other half , LG either generated sporadic action potentials or infrequent bursts of a few action potentials with varying interburst intervals ., Thus , in all experiments MCN1 stimulation elicited a gastric mill rhythm at 10°C but failed to do so at 13°C ( similar to our previous findings; see Fig 2 ) ., When leak was subtracted ( 13°C − Δleak ) , all preparations recovered the rhythm ( Fig 4C , right ) ., Consistent with the previous experiments ( Fig 3 ) , adding leak diminished action potential ( AP ) and ePSP amplitudes , while subtracting leak increased them ., In summary , our results demonstrate that a temperature-induced increase in leak conductance was sufficient to terminate the rhythm ., Accordingly , bursting in LG could be restored at elevated temperatures by adding a negative leak ., C . borealis , the animal used for this study , experiences substantial temperature fluctuations in its habitat 29 , 30 ., One would thus assume that the nervous system should be able to cope with the small temperature fluctuations we applied in our experiments ., To test if there may be mechanisms to compensate for the temperature-induced termination of the gastric mill rhythm in vivo , we implanted extracellular electrodes in intact animals and recorded the main motor nerve ( lvn ) ., Recordings typically lasted for several days ., We investigated temperature effects on the gastric mill rhythm using two approaches: first , the temperature of the water was changed at a rate comparable to the in vitro experiments ( 1°C/min ) —a velocity that has previously been shown to be sufficiently slow to cause similar changes at the STG somata 18 ., Since we were interested in spontaneous gastric mill rhythms , i . e . , rhythms that were independent of artificial stimulation , temperature was only increased after a gastric mill rhythm was present at 10°C ., The rhythm was then continuously monitored during the temperature change ., Fig 5A shows the rhythm obtained at 10°C and 13°C ., In contrast to the in vitro condition , the rhythm persisted in vivo and showed no signs of irregularity ., The number of LG spikes per burst declined slightly with increasing temperature ( Fig 5B ) , but the number of LG bursts increased at the same time , which was in stark contrast to the isolated nervous system ., Yet , physiological temperature changes might occur over longer periods ., Thus , in a second set of experiments , the temperature was kept at 10°C for at least 1 h and then slowly increased at a rate of 1°C/h to 13°C ( similar to Fig 2D ) ., We found that in these conditions spontaneous gastric mill rhythms occurred at 13°C ( N = 7 , Fig 5C ) , implying the existence of mechanisms that compensate the temperature-induced changes in the gastric mill circuit in vivo ., To mechanistically understand the adaptations rescuing the gastric mill rhythm , we went back to the in vitro preparation ., Since there was no apparent compensation within the STG circuit , we focused on one of LG’s modulatory input , namely MCN1 ., MCN1 had been the only projection neuron providing input to LG in our experimental setup ( see also 11 ) , and our initial experiments had already indicated that increasing MCN1 stimulation frequency increased the dynamic range of the gastric mill rhythm ( Fig 2E ) ., To scrutinize this idea and to test if a temperature-dependent up-regulation of MCN1 projection neuron activity could counterbalance the termination of the gastric mill rhythm , we first determined MCN1 activity at different temperatures ., We recorded spontaneous MCN1 spike activity in preparations in which feedback from the pyloric and gastric mill CPGs in the STG was severed to exclude ascending influences on the activity of MCN1 31 ., In contrast to the previous experiments , we now altered the temperature of the CoG ., MCN1 activity was recorded extracellularly from the ion stump connected to the CoG ., We found that MCN1 activity increased at 13°C ., In the example in Fig 6A , MCN1 firing frequency increased by 57 . 59% ., Note that the activity of both MCN1 neurons in a given nervous system preparation was analyzed to determine if temperature affects both MCN1 copies similarly ., Although MCN1 firing frequency at 10°C was quite variable between the two MCN1 neurons within a given preparation and across animals , in seven of eight preparations , firing frequency of both MCN1 neurons increased at 13°C ( by 51 . 29 ± 26 . 59% , N = 8 , n = 16 , Fig 6B ) ., Next , we asked whether the temperature-induced up-regulation in MCN1 firing frequency is sufficient to counterbalance the increase in LG leak conductance and to prevent the termination of the gastric mill rhythm at elevated temperatures ., To test this , we went back to the original experimental setup in which we decentralized the STG from all CoG inputs and stimulated the ion on the STG side of the nerve transection to elicit gastric mill rhythms ( Fig 1A ) ., Specifically , we stimulated MCN1 at 10°C with threshold frequency , observed the rhythm , and monitored its cessation after increasing the STG temperature to 13°C ., We then raised the MCN1 stimulation frequency in 1 Hz steps to mimic the increase in MCN1 firing frequency observed at 13°C ., Fig 6C shows that an increase in MCN1 firing frequency from 7 to 10 Hz ( 42 . 85% ) was sufficient to restore the rhythm in this particular example ., On average , a 56 . 07 ± 11 . 99% ( N = 10 ) increase in MCN1 stimulation frequency rescued the gastric mill rhythm ., This was true although threshold MCN1 stimulation frequencies varied considerably between preparations ( 5–9 Hz at 10°C , 8–15 Hz rescue frequency at 13°C ) ., We found a significant decrease in the number of LG bursts and LG spikes per burst at 13°C , but both returned to control values when MCN1 stimulation frequency was increased ( Fig 6D ) ., In fact , when we measured the minimum MCN stimulation frequency at different temperatures ( Fig 6E ) , we found that a linear increase in MCN1 frequency of 0 . 96 Hz/1°C was sufficient to rescue the rhythm at increasing temperatures ., What is the mechanism that allows MCN1 to rescue the rhythm at 13°C ?, Our previous results indicate that subtracting a leak conductance is sufficient to achieve this goal ( Fig 4B ) ., Bursting in LG is mainly driven by the release of MCN1’s peptide cotransmitter CabTRP Ia 11 ., In the STG , CabTRP Ia is exclusively found in the MCN1 terminals and thus is specific to MCN1 ., Like many other modulators in the STG , CabTRP Ia activates a well-characterized voltage-gated cation conductance ( IMI , modulator-induced current; 17 ) ., IMI supports membrane potential oscillations because of its inverted bell-shaped voltage-current relationship 32 ., Importantly , IMI has recently been suggested to act as a negative leak conductance because of the linear falling edge of its voltage-current relationship 33 ., Could the CabTRP Ia-activated IMI be sufficient to rescue the rhythm by counterbalancing the temperature-induced leak increase in LG ?, To test this , we bath applied CabTRP Ia ( 1 μM; 11 ) as a means to increase IMI and measured the response of LG during MCN1 stimulation at 13°C ., The release concentration and dynamics of CabTRP Ia are unknown , and the effective concentrations of peptide transmitters on STG neurons differ greatly between neuron types 34 ., As current responses to peptide modulators also vary substantially from animal to animal 35 , we made no attempt to determine the CabTRP Ia threshold concentration ., Rather , and most importantly for our purposes , we used a concentration shown to be effective in activating IMI 15 ., We first elicited a rhythm at 10°C , then increased the temperature to 13°C to elicit the termination , and finally applied CabTRP Ia ., Fig 7A shows that CabTRP Ia application indeed can restore the rhythm ., In the example shown , 7 Hz MCN1 stimulation elicited a gastric mill rhythm at 10°C ( Fig 7A ,, i ) , but not at 13°C, ( ii ) ., We then stopped the MCN1 stimulation and applied CabTRP Ia ., CabTRP Ia alone never elicited a gastric mill rhythm nor did it cause LG action potentials, ( iii ) ., CabTRP acts specifically on IMI , a G-protein coupled voltage-dependent inward current 14 , 34 ., Hence , to cause sustained LG activity , an additional depolarization of the membrane potential would be required ., We noted a consistent small depolarization of the membrane potential ( 2 . 01 ± 0 . 97 mV , N = 8 ) , which is consistent with earlier findings 15 and neuronal release of CabTRP Ia by MCN1 11 , indicating that the concentration used was within the physiological range used by MCN1 ., We also observed subthreshold oscillations in the LG membrane potential as a result of rhythmic disinhibitions from Int1 that were triggered by increased pyloric activity in the presence of CabTRP Ia 17 ., When MCN1 stimulation was turned on, ( iv ) , however , LG responded immediately to the threshold stimulation frequency and generated rhythmic bursts of action potentials at 13°C ., The effects of CabTRP Ia on LG were reversible, ( v ) , i . e . , MCN1 stimulation at 13°C with the threshold frequency after CabTRP Ia washout was neither sufficient to elicit LG spikes nor to start a gastric mill rhythm ., Across animals ( N = 4 ) , we found that CabTRP Ia application always restored rhythmic LG activity at 13°C when MCN1 was activated with the threshold frequency ., Correspondingly , the number of LG bursts and the number of LG spikes per burst first decreased significantly at 13°C ( Fig 7B ) and then increased in the presence of CabTRP Ia ., We noted similar but weaker effects in experiments with lower CabTRP Ia concentrations ( N = 4 ) ., In two preparations , rhythmic LG activity recovered at 100 nM , and LG firing frequency was not significantly different from the 10°C control ., Lower concentrations did not elicit spiking in LG at elevated temperatures in those two experiments ., In the other two experiments , the rhythm recovered at 10 nM , but the firing frequency of LG was lower when compared to the 10°C control condition , indicating that 10 nM did not fully recover LG rhythmicity at elevated temperatures ., Only with a simultaneous increase in MCN1 stimulation frequency were control values reached ., MCN1 activation in all experiments was necessary to elicit the rhythm , independently of whether CabTRP Ia was present or not ., Thus , MCN1s additional transmitter release and network effects ( such as activating the LG half-center antagonist Int1 ) were necessary to start the rhythm ., In summary , the CabTRP Ia-induced IMI broadened the permissible temperature range of the gastric mill rhythm and allowed LG to generate rhythmic bursts of activity ., Our results show that CabTRP Ia is sufficient to counterbalance the temperature-induced leak current in LG and to rescue the rhythm at 13°C , presumably by its known effect on IMI ., We tested the dynamics and range of this compensation by using computational models of the gastric mill network with the known connectivity of the circuit ( Fig 1A; 36 ) ., Pyloric influences on the gastric mill network were modeled by driving the pyloric pacemaker neuron AB with constant sinusoidal currents ., Initially , using values within the physiological range , we set leak and IMI conductances such that the model produced oscillations that were similar to the biological network ( Fig 8A ,, i ) ., To mimic the effects of a temperature increase , we then added an additional leak conductance to the model LG ., With twice the amount of leak , rhythmicity was absent , and LG was completely silent, ( ii ) ., All other parameters were intentionally kept constant ., No temperature-dependent changes other than the increase in the leak current ( as indicated by Fig 3F and Fig 4 ) were added since the goal was to test the interplay between leak and IMI conductances rather than the effect of temperature on synaptic and membrane properties other than leak ., When we increased the IMI maximum conductance by 50% , rhyt | Introduction, Results, Discussion, Materials and Methods | Stable rhythmic neural activity depends on the well-coordinated interplay of synaptic and cell-intrinsic conductances ., Since all biophysical processes are temperature dependent , this interplay is challenged during temperature fluctuations ., How the nervous system remains functional during temperature perturbations remains mostly unknown ., We present a hitherto unknown mechanism of how temperature-induced changes in neural networks are compensated by changing their neuromodulatory state: activation of neuromodulatory pathways establishes a dynamic coregulation of synaptic and intrinsic conductances with opposing effects on neuronal activity when temperature changes , hence rescuing neuronal activity ., Using the well-studied gastric mill pattern generator of the crab , we show that modest temperature increase can abolish rhythmic activity in isolated neural circuits due to increased leak currents in rhythm-generating neurons ., Dynamic clamp-mediated addition of leak currents was sufficient to stop neuronal oscillations at low temperatures , and subtraction of additional leak currents at elevated temperatures was sufficient to rescue the rhythm ., Despite the apparent sensitivity of the isolated nervous system to temperature fluctuations , the rhythm could be stabilized by activating extrinsic neuromodulatory inputs from descending projection neurons , a strategy that we indeed found to be implemented in intact animals ., In the isolated nervous system , temperature compensation was achieved by stronger extrinsic neuromodulatory input from projection neurons or by augmenting projection neuron influence via bath application of the peptide cotransmitter Cancer borealis tachykinin-related peptide Ia ( CabTRP Ia ) ., CabTRP Ia activates the modulator-induced current IMI ( a nonlinear voltage-gated inward current ) that effectively acted as a negative leak current and counterbalanced the temperature-induced leak to rescue neuronal oscillations ., Computational modelling revealed the ability of IMI to reduce detrimental leak-current influences on neuronal networks over a broad conductance range and indicated that leak and IMI are closely coregulated in the biological system to enable stable motor patterns ., In conclusion , these results show that temperature compensation does not need to be implemented within the network itself but can be conditionally provided by extrinsic neuromodulatory input that counterbalances temperature-induced modifications of circuit-intrinsic properties . | All physiological processes are influenced by temperature ., This is a particular problem for the nervous system , as temperature changes can disrupt the well-balanced flow of ions across the cell membrane necessary for maintaining nerve cell function ., Possessing compensatory mechanisms that counterbalance detrimental temperature effects and maintain vital behaviors is especially important for poikilothermic animals , because they do not actively maintain their body temperature and can experience substantial temperature fluctuations ., In this study , we analyze the mechanisms that allow the nervous system to maintain rhythmic activity over a range of different temperatures ., To do so , we use the well-characterized central pattern generator of the stomatogastric nervous system of the crab that controls the motion of the gut ., In this system , when experimentally isolated from the rest of the nervous system , even a small temperature increase can lead to termination of rhythmic activity due to a change in the balance of ionic conductances at elevated temperatures ., However , the intact animal can compensate for these detrimental temperature effects ., We demonstrate that such compensation can be achieved by restoring the balance of ionic conductance via an increase in neuromodulator release from projection neurons that control the motor circuits ., We conclude that temperature compensation via neuromodulation may be a widespread phenomenon since it allows quick and flexible compensation of temperature influences on the nervous system . | null | An electrophysiology and modelling study reveals how temperature can affect the balance of ionic conductances in neural circuits and how neuromodulators can compensate for detrimental temperature effects. |
journal.pcbi.1005586 | 2,017 | GINOM: A statistical framework for assessing interval overlap of multiple genomic features | A fundamental question in genome biology is whether two or more genomic features , for example gene promoters/bodies and histone modifications , are associated ., These associations can shed light on fundamental regulatory mechanisms , such as the epigenetic regulation of gene expression ., Genomic features can be represented as intervals , and thus the question becomes whether one set of genomic intervals , the query set , overlaps another set or sets of intervals , the reference set ( s ) , significantly more or less than what would be expected by chance ., The results of this test can guide the exploration of the underlying nature of the association; for example , if a histone modification is found to be associated with promoters of transcribed genes , one might hypothesize that the modification is required for active transcription ., Thus , there is a widespread need for an accurate and computationally-efficient statistical test of genomic interval overlap ., Several statistical strategies to examine genomic interval overlap have been developed 1–5 ., One study tested for associations of transposable element insertions in the genome with various epigenomic features 1 ., Using transposon mutagenesis , 1 created a database of transposable element insertions across the mouse genome and subsequently tested for insertion site bias with respect to a randomized control set ., Another method , MULTOVL , performs a Monte Carlo shuffling of the intervals uniformly throughout the genome to obtain an empirical null distribution of overlap lengths in order to test for significance 2 ., The Binary Interval Search ( BITS ) algorithm also uses a Monte Carlo simulation by uniformly shuffling the query intervals many times and obtaining an empirical null distribution of the intersection count 3 ., The Genome Association Test ( GAT ) presented in 4 is another statistical test of overlap length based on Monte Carlo simulations as in 2 , but the randomization procedure used to form this empirical null distribution can be designed to exclude regions such as gaps and repetitive sequences ., Another method called GenometriCorr is an R package that includes four different statistical tests for spatial relationships of genomic intervals 5 ., Two of the tests ( the absolute distance test and the Jaccard test ) rely on Monte Carlo randomization to formulate and test against an empirical null , and the other two tests ( the relative distance test and the projection test ) use analytical null distributions ., The main limitation of these statistical strategies is their pairwise treatment of genomic intervals , where a query interval set is compared to one or multiple reference sets individually ., These methods cannot reveal any higher order associations , i . e . any association between a query interval set and multiple reference sets simultaneously ., We present a robust statistical framework called GINOM ( Genomic INterval Overlap Model ) that adds more flexibility to the study of associations between genomic intervals ., We impose a parameterized probability model , i . e . a density function , on query interval location with respect to any number of reference sets ., Given query interval data , the model parameters are estimated through likelihood-based methods ., Each parameter value is interpreted as the amount of departure from the null distribution on the genomic loci and is indicative of the query interval overlap with a certain reference set or group of reference sets ., To specifically address the inclusion of higher order associations , we design the model in GINOM to consider any possible combination of reference sets rather than restricting to only pairwise comparisons ., Since it is possible that some combinations will have no effect on query interval location , we provide an automatic selection procedure to keep only the model terms that best describe the data ., Our goal is to define the probability density function of a query interval starting point location conditional on a given query interval length with respect to known sets of reference intervals ., This density will be defined over all possible genomic loci and formulated as a mixture of deviations from a given null distribution ., These deviations from the null occur only on the genomic index sets that would indicate an overlap of a query interval of the given length with one or more of the reference sets ., Define the discrete set of nucleotides that comprise an organism’s genome as G = { 1 , … , L } , where L is the length of the genome ., A query interval q ( X , Y ) ⊆ G is defined from two discrete random variables—the starting point location X and the length ( or cardinality ) Y—and is given as q ( X , Y ) = {X , X + 1 , … , X + Y − 1} ., From here onward , we use the notation of capital X and Y when referring to the random variables and lower case x and y when referring to observations of the respective random variables ., Since the distribution of query interval starting points depends on the query interval length , we are concerned with modeling the conditional distribution of X|Y ., We denote f ( x|y ) as the probability density function of X|Y ., Let R1 , R2 , … , RN be N known reference sets , where each Ri is defined as the union of a set of intervals on G . In practice , each Ri is set to represent a particular feature of the genome that may influence the distribution of X|Y; for example , Ri could consist of all the loci that lie within genic regions ., We say that a query interval q ( x , y ) overlaps Ri if the intersection of q ( x , y ) and Ri is non-empty ., Let ri ( x|y ) be the overlap indicator function for q ( x , y ) and Ri , which is given by, r i ( x | y ) = 1 , ifq ( x , y ) ∩ R i ≠ ∅ 0 , otherwise ., ( 1 ), In other words , ri ( x|y ) equals 1 over all x such that a query interval q ( x , y ) would overlap Ri , and it equals 0 over all x where q ( x , y ) would not overlap Ri ., Note that it is possible for a query interval to overlap more than one reference set at a time; therefore , we must define the overlap indicator function for multiple reference sets ., Consider a non-empty subset of reference set indices given by π ⊆ {1 , 2 , … , N} ., For example , if N = 3 , all possible values of π are {1} , {2} , {3} , {1 , 2} , {1 , 3} , {2 , 3} , and {1 , 2 , 3} ., The overlap indicator function indexed by the set π is defined as rπ ( x|y ) = ∏i∈π ri ( x|y ) ., It is the indicator function of the simultaneous overlap of q ( x , y ) with all the reference sets given in the index set π ., E . g . if π = {1 , 2 , 3} , then r{1 , 2 , 3} ( x|y ) equals 1 over all values of x where q ( x , y ) would overlap reference sets R1 , R2 , and R3 simultaneously and equals 0 otherwise ., Suppose further that , conditional on the length Y = y , we have defined a null distribution f0 ( x|y ) of query interval starting point locations ., For example , the null distribution could be defined as the discrete uniform distribution over G 0 ( y ) ⊆ G , the set of all x where q ( x , y ) would lie completely within the mappable regions of G . That is ,, f 0 ( x | y ) = 1 / | G 0 ( y ) | ifx ∈ G 0 ( y ) 0 otherwise , ( 2 ), where | ⋅ | denotes set cardinality ., This null distribution suggests that a query interval has an equal chance of lying anywhere inside the mappable regions of the genome and no chance of lying outside ., The restriction to the mappable regions is due to the impossibility of observing a query interval in any non-mappable region ., Although we develop the theory behind GINOM to incorporate any arbitrary null distribution , in practice we use the null provided above in Eq ( 2 ) ., In our model we take the density of a query interval to be, f ( x | y ) = c ( θ , y ) f 0 ( x | y ) exp ∑ π θ π r π ( x | y ) , ( 3 ), where c ( θ , y ) is the normalizing constant and the summation is over the non-empty subsets of {1 , 2 , … , N} , i . e . all possible combinations of reference sets ., Here , θ is the vector of model parameters ., We choose a model density with log-linear form for various reasons ., First , since the equation is formulated as an additive mixture of effects ( in the logarithmic sense ) , model interpretation is intuitive and achieved through examining the value of θ ., The components of θ describe different types of departures from the null distribution ., As θπ increases ( decreases ) , fixing all other components of θ , the probability that a random query interval intersects all of the reference sets in π increases ( decreases ) ., As θπ approaches infinity ( negative infinity ) , fixing all other components of θ , the probability approaches 1 ( 0 ) ., A further advantage to the log-linear form is that the model equation belongs to an exponential family of distributions ., This distribution family satisfies many regularity conditions that allow for a straightforward application of various techniques for statistical inference and hypothesis testing ., Finally , in this formulation each component of θ can take any value in R , and thus an optimization over θ is unconstrained and computationally more efficient than a constrained optimization ., One can use θ to define a query interval enrichment profile across the genome in the following manner ., The ratio f ( x | y ) / f ( x ˜ | y ) gives the relative likelihood of a random query interval of length y having its left endpoint at x versus another location x ˜ ., Comparing the model in Eq ( 3 ) with the null distribution f0 , we say that the endpoint x has been enriched ( depleted ) relative to x ˜ under our model if the ratio f ( x | y ) / f ( x ˜ | y ) is greater than ( less than ) f 0 ( x | y ) / f 0 ( x ˜ | y ) ., The function, h ( x | y ) = ∑ π θ π r π ( x | y ) , ( 4 ), which is a summation of certain components of θ , may be regarded as giving an enrichment profile ., If h ( x|y ) is greater than ( less than ) h ( x ˜ | y ) , then it follows that f ( x | y ) / f ( x ˜ | y ) is greater than ( less than ) f 0 ( x | y ) / f 0 ( x ˜ | y ) , and thus x is enriched ( depleted ) relative to x ˜ ., Moreover , the ratio between the two ratios above , ( f ( x | y ) / f ( x ˜ | y ) ) / ( f 0 ( x | y ) / f 0 ( x ˜ | y ) ) , is equal to exp { h ( x | y ) - h ( x ˜ | y ) } ., This quantity provides a measure of the degree of enrichment ( depletion ) at x relative to x ˜ under our model ., Using this expression , it may be seen that increasing ( decreasing ) the value of θπ increases ( decreases ) the degree of enrichment for all locations x where q ( x , y ) intersects all of the reference sets in π relative to all other locations x ˜ ., Note that the expressions presented in the above paragraph simplify when using the standard null distribution provided in Eq ( 2 ) ., Assuming that locations x and x ˜ are both mappable , then f 0 ( x | y ) / f 0 ( x ˜ | y ) = 1 . We say that x is enriched ( depleted ) with respect to x ˜ if f ( x | y ) / f ( x ˜ | y ) = exp { h ( x | y ) − h ( x ˜ | y ) } is greater than ( less than ) 1 , i . e . if h ( x|y ) is greater than ( less than ) h ( x ˜ | y ) ., Typically , we are interested in studying the enrichment of x with respect to a mappable x ˜ such that q ( x ˜ , y ) would not overlap any of the reference sets ., In this case of x ˜ in the so-called “background” portion of the genome , h ( x ˜ | y ) = 0 , and , therefore , x is enriched ( depleted ) simply if h ( x|y ) is positive ( negative ) ., From this point onward , we exclusively use this simplified condition when determining the enrichment of x ., To help with the above model interpretation , we provide a simple example with N = 2 and f0 as in Eq ( 2 ) ., Since N = 2 , the model terms are given by the set indices π = {1} , {2} , {1 , 2} ., Suppose that θ = ( θ{1} , θ{2} , θ{1 , 2} ) ′ = ( 0 . 5 , 0 . 7 , − 0 . 2 ) ′ , where the prime denotes vector transpose ., Now , for an x such that q ( x , y ) overlaps R1 but not R2 , the enrichment is given by h ( x|y ) = θ{1} = 0 . 5 , which is positive ., Thus , the set of all such x’s are enriched , and the probability of q ( x , y ) is e0 . 5 = 1 . 65 times greater than q ( x ˜ , y ) , where x ˜ is in the background ., Similarly , the set of all x’s such that q ( x , y ) overlaps R2 but not R1 are enriched , and the probability of q ( x , y ) is e0 . 7 = 2 . 01 times that of q ( x ˜ , y ) ., For an x such that q ( x , y ) overlaps both R1 and R2 , the enrichment is given by h ( x|y ) = θ{1} + θ{2} + θ{1 , 2} = 1 . 0 ., In this case , the probability of q ( x , y ) is e1 . 0 = 2 . 72 times that of q ( x ˜ , y ) ., Notice that even though the individual model parameter θ{1 , 2} is less than zero , there is still enrichment in these x’s rather than depletion ., The model parameter θ{1 , 2} represents the interaction effect of R1 and R2—the effect of R1 and R2 beyond that of simply adding the two individual effects together ., In this case , the overall effect of R1 and R2 is slightly less than their additive effect due to the negative interaction term ., For more details on model interpretation , see the Model Interpretation section of S1 Text ., In general , since there are 2N − 1 possible non-empty subsets of {1 , 2 , … , N} , there are that many possible model parameters in Eq ( 3 ) ., In practical implementations , 2N − 1 can be quite a large number , and the inclusion of all possible parameters can yield an unnecessarily complex model ., Therefore , to avoid overfitting , we typically restrict to a smaller submodel by setting the less important parameters equal to zero and thus effectively dropping those terms from the model equation ., For example , for ease of interpretation , we could decide that any term with third-order interactions or higher—that is , if π contains three or more reference set indices—is too complex to include in the model ., Thus , we would automatically exclude those terms from consideration in the model equation , and the summation would be over all π such that |π|, ≤ 2 . From here on , we denote d as the number of model parameters included in the model ., If all possible parameters are included in the model , i . e . if d = 2N − 1 , then we say that f is the full model ., Suppose we are given data in the form of n query intervals {q ( xi , yi ) , i = 1 , … , n} , and let us assume that the xi’s are conditionally independent given the yi’s ., In order to fit the model to the data , we seek the maximum likelihood estimate ( MLE ) of θ given the data ., To compute the MLE θ ^ , we maximize over the likelihood function , which is equal to the joint density function, ∏ i = 1 n f ( x i | y i ) = ∏ i = 1 n c ( θ , y i ) f 0 ( x i | y i ) exp { θ ′ T } , ( 5 ), where the prime denotes vector transpose and T is a length d vector with the π-th component given by t π = ∑ i = 1 n r π ( x i | y i ) ., The value tπ is the number of query intervals in the dataset that overlap all of the reference sets given in the index set π simultaneously ., After taking the logarithm of the above function and dropping terms that are constant in θ , we are left with the following objective function:, ℓ ( θ ) = θ ′ T + ∑ i = 1 n log ( c ( θ , y i ) ) ., ( 6 ), The MLE is thus computed as, θ ^ = argmax θ ∈ R d ℓ ( θ ) ., ( 7 ), See S1 Text for an efficient method to compute the optimization in Eq ( 7 ) as well as approximate confidence intervals for each component of θ ., Now that we have a means for computing the MLE , and since the density f satisfies all the necessary regularity conditions , we can use the generalized likelihood ratio test ( GLRT ) for hypothesis testing ( see Ch . 8 of 6 ) ., In particular , we use the GLRT to test whether a specific component of θ differs from zero or not , i . e . whether a certain reference set or combination of reference sets affects query interval location ., In order to perform this statistical test for the π-th component , one must solve Eq ( 7 ) twice—once allowing θπ to be unconstrained in the optimization and once with the constraint that θπ = 0 ., If the value of ℓ ( θ ^ ) changes enough with addition of this constraint , i . e . with the removal of the π-th model term , then we claim that θπ is not equal to zero ., For more detailed information on the GLRT , including the computation of a p-value for the test , see S1 Text ., In order to select the model parameter configuration that best describes the data , we implement the widely-used bidirectional stepwise model selection algorithm 7 ., This algorithm systematically adds and removes model terms one at a time in an effort to find a configuration that minimizes the Bayesian Information Criterion ( BIC ) 8 , which is given by B I C = d log n − 2 ℓ ( θ ^ ) ., The BIC is a log-likelihood function that is penalized by the number of model parameters; therefore , the model configuration with optimal BIC is one that offers a high likelihood value with a small number of model parameters ., The result of the stepwise algorithm is a model that contains only the most influential parameters , allowing for an easy yet biologically meaningful model interpretation ., We analyzed the performance of GINOM on query interval sets simulated directly from the model equation Eq ( 3 ) using the nine reference sets listed in Table 1 ., This is the same set of features used in the LINE-1 insertion bias study we discuss later and contains a combination of genomic and epigenomic features associated with genome accessibility and transcriptional regulation ., For simplicity in our simulations , we fix the query interval length to be Y = 1 unless stated otherwise and assume all loci in G are mappable when forming the null distribution in Eq ( 2 ) ., A simulation is done by first selecting N reference sets to consider , then optionally restricting the domain to a subset of the entire genome , followed by setting the true values for each component of θ , and then finally generating a random sample of size n via acceptance/rejection algorithm ., The result of the simulation is a set of n query intervals that are independently and identically distributed according to the specified model ., We provide an application for GINOM using real data ., For this example , we examine somatic insertions of the active Long INterspersed Element-1 ( LINE-1 or L1 ) retrotransposon into the genome and test for insertion biases ., LINEs are mobile elements present in many eukaryotes; their number can increase through a copy-and-paste mechanism and they have been implicated in several diseases 9 ., Recently , somatic retrotransposition was identified as a frequent event in many human cancers and in the adult human brain 10–18 ., A major biological question arising from this work is whether LINE-1 , the active human retrotransposon , displays a bias in the locations of somatic retrotransposition ., Various and sometimes contradictory biases for L1 retrotransposon insertions have been identified ., Prior experimental work on the L1 protein machinery identified a retrotransposition bias for accessible DNA 19 and an L1 endonuclease recognition motif ( TTAAAA ) 20 ., High-throughput studies of somatic retrotransposition uncovered a disproportionate bias towards affecting protein-coding genes in the brain and towards genes commonly mutated in cancer 11 , 17 ., Contradicting this view , high-throughput studies of germ-line retrotranspositions were identified to display depletion in protein-coding regions 21 , 22 ., One other interesting bias is that somatic L1 retrotranspositions in cancer are enriched towards regions that also display DNA hypomethylation 17 ., We aim to rigorously test the biases associated with somatic L1 retrotranspositions ., To do so , we compiled a dataset consisting of 2540 somatic L1 retrotranspositions available from a single tissue that were identified in lung cancer from two independent studies 13 , 18 ., The Helman et al study identified 363 somatic retrotransposition events in lung adenocarcinoma ( 39 events ) and lung squamous cell carcinoma ( 324 events ) from The Cancer Genome Atlas ( TCGA ) 13 ., Tubio et al identified 2177 L1 retrotransposition events from various lung cancer sources including primary tumor samples , cell-lines , and TCGA samples 18 ., Here , we examine the overlaps of the lung cancer somatic L1 retrotransposition query set: 2540 events ( hg19 coordinates ) with respect to nine curated reference sets that represent features of protein-coding genes , euchromatin , and heterochromatin ., We reasoned that the ability of novel LINE-1 copies to insert themselves into a given region might be affected by this region’s accessibility; hence , we selected a set of genomic and epigenomic features that are known to be associated with genome accessibility and active transcription ., The gene feature reference set includes RefSeq gene bodies ( UCSC version update 10/06/15 ) ., Euchromatic tracks include broad peaks identified by the ENCODE project for histone marks H3K36me3 , H3K79me2 , and H3K4me3 in lung cancer cell-line A549 23 ., Our selected heterochromatic marks include additional ENCODE broad peaks for histone marks H3K9me3 and H3K27me3 in A549 cells 23 , which have a role in repressing gene expression ., We also included a track for two additional heterochromatic features , lamina associated domains ( LADs ) and late DNA replicating regions that , although defined in fibroblasts , are partially conserved between cell-types 24 , 25 ., Finally , we included a track for regions hypomethylated in cancer cells 26 ., We compared our method with four other methods: BITS , MULTOVL , GAT , and GenometriCorr ., However , unlike our method , which can examine multiple features simultaneously , these other methods test for significant overlap in a pairwise manner ., Somatic L1 retrotranspositions in lung cancer displayed a bias for heterochromatic marks including H3K9me3 , H3K27me3 , lamina-associated domains ( LADs ) , and late-replicating regions according to BITS , MULTOVL , GAT , and GenometricCorr by the Jaccard measure ( S2 Table ) ., In addition , consistent with Lee et al . 17 , somatic L1 retrotransposition also displayed a preferential bias towards regions hypomethylated in cancer according to BITS , MULTOVL , GAT , and GenometricCorr by the Jaccard measure ( S2 Table ) ., Conversely , and consistent with studies of germ-line retrotransposition , somatic L1 insertions displayed a bias against gene regions and euchromatin marks including H3K4me3 , H3K36me3 , and H3K79me2 according to MULTOVL , GAT , and GenometricCorr by the Jaccard measure ( S2 Table ) ., Therefore , using the pairwise approach , we did not observe a bias of lung somatic L1 retrotranspositions towards gene regions ., We hypothesized that our new method GINOM , which can examine combinations of overlaps , might be able to recover gene regions that account for previously observed biases of somatic retrotranspositions towards cancer mutations 17 ., We examined the lung L1 somatic retrotransposition query set with respect to all nine reference sets using GINOM with the BIC penalty in model selection ( see Model Selection subsection within Materials and methods section ) ., Our null model is that of Eq ( 2 ) with G 0 ( y ) being the set of loci x such that q ( x , y ) would lie entirely within the mappable regions of G . GINOM reports significant model parameters and their associated MLE and p-value of the GLRT , and we show the results in Table 4 ., For ease of interpretation , we label the significant model terms as either a primary effect or a secondary effect ., A primary effect indexed by model term π is such that there exists no other significant model parameter indexed by π ˜ where π ˜ ⊂ π ., For example , model term {2 , 7} is a primary effect because no model term with its index set contained in {2 , 7} ( e . g . either term {2} or term {7} ) is significant ( Table 4 ) ., A secondary effect is a model term that has at least one primary effect contained within its index set ., For example , model term {6 , 7 , 8} is a secondary effect because primary effects {6} and {8} are contained within {6 , 7 , 8} ( Table 4 ) ., The primary effects in the model selected by GINOM under the BIC penalty were hypomethylation , late-replicating , LADs , H3K9me3 , and the combination of gene regions and H3K27me3 ( Fig 3 ) ., Here , all primary effects showed enrichment with respect to the background , and , interestingly , the strongest of these enrichments occured in hypomethylation , model term {5} , which was consistent with Lee et al . 17 ( Fig 3 ) ., With respect to x ˜ in the background , the enrichment of a location x such that q ( x , y ) overlaps R5 only ( and not any other reference set ) is h ( x | y ) = θ ^ { 5 } = 0 ., 7654 ., Therefore , the probability of a query interval q ( x , y ) is e0 . 7654 = 2 . 15 times that of q ( x ˜ , y ) ., The model was also able to recover gene regions when paired with the facultative heterochromatin mark H3K27me3 , given by model term {2 , 7} ., The euchromatin marks were also frequently incorporated into secondary effects ., For example , model term {2 , 4 , 6 , 9} , which represents the combination of gene regions , H3K79me2 , H3K9me3 , and late-replicating regions , was assigned a parameter value of θ ^ { 2 , 4 , 6 , 9 } = - 1 ., 668 ., It therefore acts antagonistically to and with greater magnitude than the sum of the effects contained within it ( terms {6} and {9} ) so that the overall combined effect of R2 , R4 , R6 , and R9 is negative ., Specifically , for a location x such that q ( x , y ) simultaneously overlaps R2 , R4 , R6 , and R9 only , the enrichment is given by h ( x | y ) = θ ^ { 6 } + θ ^ { 9 } + θ ^ { 2 , 4 , 6 , 9 } = - 0 ., 7932 ., Therefore , such an x is depleted , with the probability of q ( x , y ) being e−0 . 7932 = 0 . 4524 times that of q ( x ˜ , y ) in the background , which is consistent with somatic transpositions being biased against euchromatin ., Finally , we sought to more closely examine the main effect resulting from the intersection of gene regions and H3K27me3 ., It was recently reported that L1 retrotranspositions display a bias towards cancer genes 17 ., To test this in our dataset , we utilized the cancer gene lists from Lee et al , which include the COSMIC cancer gene census and the Memorial Sloan-Kettering Cancer Center cancer gene list ( S3 Table ) ., In the L1 lung cancer insertion dataset , we observed that 759 genes bodies overlap a somatic L1 retrotransposition ., Of these , a significant proportion ( 37/759 ) are cancer genes ( p-value = 0 . 027 , hypergeometric test , S1 Text ) ., We then looked more specifically at the L1 insertions in gene bodies marked by H3K27me3 , which represented a significant main effect in the GINOM model ., Of the 759 gene bodies that overlap a somatic L1 retrotransposition , 309 genes also intersect a H3K27me3 broad peak , and are significantly enriched for cancer genes ( 22/309 , p-value 0 . 001 , hypergeometric test , S1 Text ) ., Hence , we speculate that L1 insertion bias towards cancer genes could be a consequence of L1 bias towards genes located within the facultative heterochromatin marked by H3K27me3 ., Together our results reconcile some of the discrepancies in observations between somatic L1 retrotranspositions and germ-line retrotransposition events ., Overall , somatic L1 retrotranspositions display a bias towards features typically associated with constitutive heterochromatin , similar to germline transposition events ., However , some genes , e . g . cancer genes and genes marked by H3K27me3 , also display a bias for somatic L1 transposition events in lung cancer ., We developed a Genomic INterval Overlap Model that allows for the interrogation of significant associations between many genomic features simultaneously ., Unlike prior methods , which test for associations in a pairwise manner , GINOM treats query interval location as a random variable of log-linear distribution with model terms formed from any possible combination , or interaction , of multiple reference sets ., In this fashion , GINOM can uncover any higher-order interaction among reference sets that has a significant effect on query interval location ., Through an implementation of a stepwise model selection routine , GINOM can handle an arbitrarily large number of reference sets simultaneously as input and subsequently output a reduced model that satisfies an optimal trade-off between number of model terms and likelihood value ., The end result is a set of significant model terms with associated parameter values that defines a profile of query interval enrichment at a level of detail beyond that of pairwise comparisons ., To highlight this unique capability of GINOM , we fit the model using a query interval set of lung cancer somatic L1 retrotranspositions and a collection of reference sets representing protein-coding genes , euchromatin , and heterochromatin features ., The output of GINOM indicates enrichment towards the individual heterochromatic reference sets , as do other current methods ., However , GINOM uncovers more nuanced associations than the other methods by identifying a significant enrichment within genes only when paired with the H3K27me3 heterochromatic mark ., Conversely , it also indicates depletion within genes when paired with certain euchromatic marks ., The association of L1 somatic retrotranspositions and gene bodies , when marked by H3K27me3 , is not recovered by other methods because they only consider pairwise associations ., Our results demonstrate GINOM’s ability to test for significance of interval overlap between multiple genomic features ., As more data of this type becomes available , it will provide an effective method to screen for yet-uncharacterized higher-order associations between genomic features . | Introduction, Methods, Results/Discussion | A common problem in genomics is to test for associations between two or more genomic features , typically represented as intervals interspersed across the genome ., Existing methodologies can test for significant pairwise associations between two genomic intervals; however , they cannot test for associations involving multiple sets of intervals ., This limits our ability to uncover more complex , yet biologically important associations between multiple sets of genomic features ., We introduce GINOM ( Genomic INterval Overlap Model ) , a new method that enables testing of significant associations between multiple genomic features ., We demonstrate GINOM’s ability to identify higher-order associations with both simulated and real data ., In particular , we used GINOM to explore L1 retrotransposable element insertion bias in lung cancer and found a significant pairwise association between L1 insertions and heterochromatic marks ., Unlike other methods , GINOM also detected an association between L1 insertions and gene bodies marked by a facultative heterochromatic mark , which could explain the observed bias for L1 insertions towards cancer-associated genes . | The age of genomics has made a large number of datasets available for the wider scientific community ., Many of these datasets come in the form of genomics tracks , represented as features associated with a collections of genomic intervals along chromosomes ., A common talk in genomics is to identify putative associations between these features that can lead to new insights about genome organization and function ., For example , activity of certain classes of genes might be influenced by the presence of specific combinations of chromatin modifications and binding of transcription factors at their promoters or enhancers ., Here , we present a novel methodology , named GINOM ( Genomic INterval Overlap Model ) , to test for the significance of these associations ., We apply it to the problem of detecting biases of the locations along chromosomes where mobile genetic elements tend to insert themselves , and identify a potential preference for L1 elements towards gene bodies marked by a facultative heterochromatic mark . | null | null |
journal.pntd.0007380 | 2,019 | Modelling population dynamics and seasonal movement to assess and predict the burden of melioidosis | Melioidosis is an infection caused by the Gram-negative bacillus Burkholderia pseudomallei , which exhibits marked seasonality in most settings where it is endemic , including Southeast Asia and Northern Australia 1 ., Melioidosis is a communicable disease that is usually transmitted via contaminated soil or water , and is highly prevalent in Northeast Thailand 2 ., Most of the population at risk of melioidosis lives in rural areas , especially those people who frequently come into contact with soil or water , such as rice farmers 3 , 4 ., In Thailand , the highest number of melioidosis reported cases are often in January and October 5 ., Infection with B . pseudomallei shows great clinical diversity , spanning asymptomatic infections , localized skin ulcers or abscesses , chronic pneumonia mimicking tuberculosis , and fulminant septic shock with abscesses in multiple internal organs 6 ., Both humans and animals are susceptible to B . pseudomallei , and may be infected by percutaneous inoculation , inhalation , or ingestion ., Person-to-person spread and zoonotic infections of humans are very rare 7 ., The incubation period is between 1–21 days ( average 9 days ) 8 , and is believed to be influenced by the inoculation dose , mode of infection , host risk factors , and probably differential virulence of the infecting organisms ., Most cases result from recent infections , although latency with reactivation has been described up to 62 years following exposure 8 , while the median times to relapse and reinfection are 21 weeks and 111 weeks , respectively ., The risk of relapse is related to a patient’s adherence to treatment and the initial extent of disease , but not to any underlying conditions 9–11 ., Melioidosis seems to be more severe in older people with lower immunity or chronic underlying conditions , such as diabetes 12 ., The risk of contracting melioidosis in diabetic individuals is 12 times higher than for non-diabetic individuals 13 , 14 ., Currently , the global burden of melioidosis is estimated to be 165 , 000 cases per year ( 95% credible interval 68 , 000–412 , 000 ) , with 89 , 000 deaths ( 36 , 000–227 , 000 ) 15 ., Thailand’s Bureau of Epidemiology ( BoE ) launched a melioidosis surveillance system in 2001 ( Report 506 ) 5 ., Approximately 80% of reported melioidosis cases were from Northeast Thailand 5 ., In the past , the number of cases shown in the surveillance system was heavily relied on provincial and regional hospitals voluntarily report , very few were reported from private hospitals 16 ., In general , melioidosis is diagnosed by testing for antibodies to B . pseudomallei using an indirect hemagglutination ( IHA ) technique , which has been found to have low sensitivity and specificity 17 ., This surveillance system was revised in 2010 in order to capture more health data items ., There has been an increase in usage of bacterial culture 16 which could give rise to an increase in total number of culture-confirmed cases ., In addition , there has been an improvement to access to healthcare ., Nevertheless , the true number of cases is still under-reported because of diverse clinical manifestations and inadequate bacterial identification methods ., A previous estimation suggested cases in Thailand were in excess of 7 , 000 cases per year 15 , while the BoE reported just 3 , 242 cases in 2015 5 ., B . pseudomallei is resistant to a wide range of antimicrobials , and ineffective treatment may result in death in 70% of cases 18 ., The treatment for melioidosis consists of an intensive phase of at least 10–14 days of ceftazidime , meropenem , or imipenem , administered intravenously , followed by oral eradication therapy , usually with trimethoprim–sulfamethoxazole ( TMP-SMX ) for 3–6 months 19 ., There is currently no vaccine against melioidosis 20 , 21 ., The demographics of Thailand are currently in a transition phase , becoming more like those of developed countries , with rapid changes in population structure , reductions in birth and mortality rates , and a low rate of population growth ., Urbanization is accelerating , and there are large annual population movements ., These types of changes have been shown to have important impacts on public health and the disease burden of both non-communicable 22 and communicable diseases 23 ., The population at highest risk of contracting melioidosis is the working age group ., There is appreciable seasonal movement among this group as they go about their working lives ., The internal migration of Thai people involves a number of distinct forms of movement within each year ., Three forms have been identified in previous research 24: a single movement , seasonal movement , and repeated movement ., Seasonal migration involves people moving from the North and Northeast regions of Thailand towards the Bangkok metropolis and the Central region during the dry season ( from March through to June ) , and in the reverse direction during the wet season ( June to September ) 24 ., 40% of people from the Northeast are classified as seasonal migrants ( a transient population ) 25 ., It is obvious that for person-to-person transmissible infections , there are significantly more infections when such transient individuals are considered 26–29 ., However , very few studies were trying to look at the effect of transient populations on an infectious disease from a primarily environmental source which will help better describe the temporal and spatial changes of the incidence of such a disease 30 ., Developed countries are also observing an emergence of melioidosis related to travelling and importation of cases 1 ., To date , only a few approaches have been applied to determine the melioidosis burden , including simple maps of melioidosis 1 , maps of the global distribution of B . pseudomallei , and estimates of the total incidence and mortality due to melioidosis worldwide using a statistical model 15 ., Only one study has used mathematical modelling , exploring the use of childhood seroprevalence data as a marker of intensity of exposure 31 ., In this study , we used mathematical modelling to predict the incidence of melioidosis in the Thai population , taking account of population changes , seasonal movement , and incidence of diabetes ., The model provides multi-dimensional forecasting of melioidosis , which could be useful for targeting intervention strategies in this setting ., We generated a deterministic demographic sub-model to predict the size of the total population ( see S1 Figure A ) ., We stratified the population by age and sex into 100 annual interval classes , from 0 to 100 years old ., The population in each class followed the actual population structure of Thailand between 1980 and 2015 , based on birth , death , and migration rate data from the Population and Housing Census 32 , 33 , and using the 1980 census data as the initial condition ., All females in the age classes between 15 to 50 years old were considered to be capable of reproduction , with the fertility rate ( fr ) 34 , while the death rate was age-related 35 ., Members of the population were assumed to die upon reaching 100 years of age ., Crude net migration rates ( immigrant minus emigrant per 1 , 000 population ) for each year had an impact on all age and sex compartments 36 ., Most of the at-risk population for melioidosis lives in rural areas , especially in Northeast Thailand , so we modelled internal migration by classifying the population of Thailand into three independent groups ., These were: those from the Northeast region who live at home for more than 6 months in a year ( NE ) , the transient group or the people from the Northeast region who move seasonally between home and other parts of the country and spend less than 6 months in a year at home ( T ) and lastly the non-Northeast group , who live somewhere other than the Northeast ( Non_NE ) ., We created the seasonal movement sub-model to overlay with the demographic sub-model to estimate the rates of movement among them ( see S1 Figure B ) ., We solved a large set of ordinary differential equations ( ODE ) for the deterministic demographic sub-model and the seasonal movement sub-model , defined in S1 Information on Demographic sub-model and Seasonal movement sub-model , respectively ., The demographic and seasonal movement sub-model was overlaid with the melioidosis infection model , defined in S1 Information on Melioidosis infection sub-model ., In the melioidosis infection model ( a susceptible , exposed , infected , recovered , susceptible , or SEIRS , model ) , the population was further divided into eight health compartments: susceptible ( S ) , diabetic susceptible ( SDM ) , exposed ( E ) , symptomatic ( Sym ) , asymptomatic ( Asym ) , severe ( Sev ) , treatment ( Treat ) , and recovered ( R ) ( see Fig 1 ) ., Melioidosis case data stratified by age , sex , and geographical area were obtained from the Thai annual epidemiological surveillance reports from 2008 to 2015 5 ., Key assumptions for our model were as follows ., First , the transient population data used within this model referred only to the movement of the Thai population ., The movement of migrant workers from other countries could be significant but was omitted in this study for simplicity 24 ., Second , diabetes progression was assumed to be irreversible , i . e . people could not move from diabetic to non-diabetic ., Third , we did not consider pre-diabetes or impaired glucose tolerance ., Fourth , we assumed that incidence rates of diabetes were constant over time but varied by age ., Fifth , we did not focus on chronic symptoms ( those of duration greater than two months ) , including such presentations as chronic skin infections , chronic lung nodules , or pneumonia , which only accounted for around 10% of melioidosis patients 12 ., Finally , we did not focus on any behavioral factors such as excessive alcohol use ., We used R software version 3 . 3 . 3 to run and analyze the model outputs , and the deSolve package to solve the differential equations 37 ., The initial parameter values were calculated from population data and disease burden ., Model fitting was carried out using the Markov Chain Monte Carlo ( MCMC ) method , implemented with the Bayesian Tools R package as defined in S2 Information on the Bayesian framework 38 ., The demographic and seasonal movement sub-models were run from 1980 ( see S2 Figure A ) to calibrate the model by fitting to the average migration data , including the population in the Northeast moving to non-Northeast , and the reverse direction from 2005 to 2015 25 ., We estimated seasonal movement parameters from the transient population model ( see S1 Table A ) and used them to run the melioidosis infection model from 2005 ., The model was run and fitted to the annual incidence of melioidosis by age , sex , and area by year , and seasonally by month , from 2008 to 2015 5 ., For model fitting , the DEzs method in the Bayesian Tools package allowed automatic parallelization on three cores to be used for sampling ., This method allowed fewer chains to be used for estimated a large number of parameters and thus optimized the computational time 39 ., Number of iterations and burn in were decided upon the model convergence by analyzing the differences between multiple Markov chains ., The convergence was assessed by several measures including the standard procedure of Gelmal-Rubin 40 , 41 and the target acceptance rates 42 ., Thirty-three parameters were estimated and the median values and credibility intervals were reported ., These parameters were those representing the infection rates among both sexes in the Northeast , transient , and non-Northeast populations , ( βaNE , βaT , βaN ) respectively , proportion of symptomatic cases ( pE ) , recovery rate from asymptomatic ( σ ) , recovery rate from symptomatic ( γ ) , Relative susceptibility to melioidosis among diabetic individuals when compare with non-diabetic ( q ) , mortality/death rate for melioidosis ( μM ) , amplitude ( Ainc ) , phase angle ( φinc ) and proportion of reporting ( Report ) ( see S1 Table A ) ., Note that the proportion ( 1- Report ) was defined as “Under-reporting” i . e . those symptomatic melioidosis patients that have been seen by a physician , but the physician did not report them to the public health authority for some reasons e . g . improperly diagnosis or missing report ., The model was further used to predict the 20-year age-specific incidence of melioidosis among males and females in Thailand , sampling all 33 parameters from the posterior chains ., The model predictions were reported as age , gender , and area-specific incidence rates over time ., The demographic sub-model was able to reproduce the past population structure of Thailand from 1980 to the present ( see S2 Figure A ) ., The parameters that characterized seasonal movement were estimated by fitting the model to the population movement data ( see S2 Figure B ) ., The model showed that majority of movements were made by Northeast individuals who moved to non-Northeast areas , approximately 13 , 600 persons per 100 , 000 population per month , or 34% of all movements within a month ( see S1 Table A ) ., Moreover , the majority of movements were among those aged between 15 and 60 years old , about 19 , 000 persons per 100 , 000 population per month , or 51% of all movements within a month ( see S2 Figure C ) ., The fitting performance is shown in Fig 2 ., Melioidosis cases occurred seasonally , with a peak in the wet season that lasted from May to October ., The infection parameters that minimized the fit statistic , using the Bayesian method , are shown in Table 1 ., The highest infection rate was estimated to be 6 cases per 100 , 000 population per month among males aged 45–59 years old in the Northeast ., The lowest rate was 0 . 4 cases per 100 , 000 population per month among females aged 15–44 years old in the non-Northeast region ., Surprisingly , we found that the infection rate among the transient male population aged 15–44 years was higher than the non-Northeast population ( 0 . 8 compared with 0 . 08 per 100 , 000 persons per month ) ., Overall 46% of melioidosis cases were symptomatic ., Recovery rates for untreated symptomatic cases and asymptomatic patients were estimated by the model , with the average period of infection estimated at around 9 and 5 months , respectively ., The susceptibility to melioidosis among DM population is 10 . 84 95% CI 8 . 42–12 . 23 times greater than the non-DM population ., If patients’ treatment failed and they developed severe melioidosis , they could die within two weeks ., We estimated 80% and 50% under-reporting of cases in 2008–2009 and 2010–2015 , respectively ., Projections of the numbers of melioidosis cases between 2015 and 2035 are given in Fig 3 ., Total melioidosis incidence per year was projected to increase by 70% , from 6 , 569 ( 4 , 834–8 , 701 ) in 2015 to 11 , 173 ( 8 , 207–14 , 773 ) in 2035 ., The largest increase of melioidosis was projected to occur among the population aged 45–59 years old ., The predicted incidence among males was two-fold greater than that of females ., The majority of melioidosis cases were seen to occur in the population from the Northeast region of Thailand ., The predicted incidence among non-diabetic was two-fold greater than that of diabetic population ., In Fig 4 , total melioidosis incidence rates were projected to increase by approximately 10% by 2035 , from 11 . 42 ( 8 . 5–13 . 4 ) in 2015 to 12 . 78 ( 9 . 6–14 . 9 ) per 100 , 000 population in 2035 ( see Table 2 ) ., The highest incidence rates were predicted to be among those aged between 45–59 years old , followed by those age 60 years old and above ., The incidence was almost double among males compared with females in both Northeast and other regions ., The incidence rate among the Northeast population was more than double compared with the transient population , and almost ten times higher when compared with the other regions ., This study also highlighted the importance of melioidosis among the transient population who temporally live in the risk area but had almost six times higher incidence compared with other regional populations ., From diabetes prospective , the incidence of melioidosis among diabetes was predicted to be as high as 60 per 100 , 000 population ., To summary , the risk of melioidosis among the aging population with some chronic diseases such as diabetes is presenting an increasing trend ., The risk of infection among transient population , who temporary get some disease exposure during the agricultural seasons , cannot be ignored ., Population dynamics , seasonal movement , melioidosis infection rates , and under-reporting are important components of melioidosis incidence patterns ., The increases seen in melioidosis cases are partly attributable to demographic changes as working , transient , and aging population groups are more prone to develop melioidosis ., The key findings from our study are firstly , the increasing trend of melioidosis incidence , especially among males aged 45–59 years old , is predicted to continue; and secondly , the male , Northeast , and transient populations aged 45–59 years old were at the highest risk of melioidosis infection ., We anticipate that the modelling methods described here could be used in similar settings , especially those with reliable census data , to estimate the future melioidosis burden , as well as the potential effects of under-reporting ., In addition , this modelling approach could be adapted to study other infectious diseases , behavioral changes , and seasonal movements , where demographic factors are important drivers of a population’s disease burden . | Introduction, Methods, Results, Discussion | Melioidosis is an infectious disease that is transmitted mainly through contact with contaminated soil or water , and exhibits marked seasonality in most settings , including Southeast Asia ., In this study , we used mathematical modelling to examine the impacts of such demographic changes on melioidosis incidence , and to predict the disease burden in a developing country such as Thailand ., A melioidosis infection model was constructed which included demographic data , diabetes mellitus ( DM ) prevalence , and melioidosis disease processes ., The model was fitted to reported melioidosis incidence in Thailand by age , sex , and geographical area , between 2008 and 2015 , using a Bayesian Markov Chain Monte Carlo ( MCMC ) approach ., The model was then used to predict the disease burden and future trends of melioidosis incidence in Thailand ., Our model predicted two-fold higher incidence rates of melioidosis compared with national surveillance data from 2015 ., The estimated incidence rates among males were two-fold greater than those in females ., Furthermore , the melioidosis incidence rates in the Northeast region population , and among the transient population , were more than double compared to the non-Northeast region population ., The highest incidence rates occurred in males aged 45–59 years old for all regions ., The average incidence rate of melioidosis between 2005 and 2035 was predicted to be 11 . 42 to 12 . 78 per 100 , 000 population per year , with a slightly increasing trend ., Overall , it was estimated that about half of all cases of melioidosis were symptomatic ., In addition , the model suggested a greater susceptibility to melioidosis in diabetic compared with non-diabetic individuals ., The increasing trend of melioidosis incidence rates was significantly higher among working-age Northeast and transient populations , males aged ≥45 years old , and diabetic individuals ., Targeted intervention strategies , such as health education and awareness raising initiatives , should be implemented on high-risk groups , such as those living in the Northeast region , and the seasonally transient population . | Melioidosis is an infectious disease caused by the Gram-negative bacillus Burkholderia pseudomallei , which exhibits marked seasonality in most settings where it occurs , such as Southeast Asia and Northern Australia ., Most of the population at risk of contracting melioidosis lives in rural areas; particularly at risk are those who are exposed to soil and water , such as rice farmers ., Thailand’s demography is in a transient phase , with older age groups set to double within a decade ., Social impacts of lifestyle changes are reflected in seasonal movement and increasing urbanization ., In this study , we used mathematical modelling to examine the impacts of such demographical changes on an important infectious disease and to dynamically predict the disease burden in a developing country setting , namely Thailand ., We found that melioidosis incidence was significantly higher among working-age Northeast and transient populations , specifically among males aged ≥45 years old and individuals with diabetes ., Improved health education and awareness raising should be implemented on a national scale , with a focus on high-risk groups living in endemic areas , as well as those who move seasonally between these and other areas . | death rates, medicine and health sciences, melioidosis, population dynamics, geographical locations, diabetes mellitus, bacterial diseases, age groups, endocrine disorders, population biology, public and occupational health, infectious diseases, endocrinology, metabolic disorders, people and places, population metrics, asia, biology and life sciences, population groupings, thailand | null |
journal.pgen.1006122 | 2,016 | A Powerful Procedure for Pathway-Based Meta-analysis Using Summary Statistics Identifies 43 Pathways Associated with Type II Diabetes in European Populations | Genome-wide association study ( GWAS ) has become a very effective way to identify common genetic variants underlying various complex traits 1 ., The most commonly used approach to analyze GWAS data is the single-locus test , which evaluates one single nucleotide polymorphism ( SNP ) at a time ., Despite the enormous success of the single-locus analysis in GWAS , proportions of genetic heritability explained by already identified variants for most complex traits still remain small 2 ., It is increasingly recognized that the multi-locus test , such as gene-based analysis and pathway ( or gene-set ) analysis , can be potentially more powerful than the single-locus analysis , and shed new light on the genetic architecture of complex traits 3 , 4 ., The pathway analysis jointly tests the association between an outcome and SNPs within a set of genes compiled in a pathway according to existing biological knowledge 4 ., Although the marginal effect of a single SNP might be too weak to be detectable by the single-locus test , accumulated association evidence from all signal-bearing SNPs within a pathway could be strong enough to be picked up by the pathway analysis if this pathway is enriched with outcome-associated SNPs ., Various pathway analysis procedures have been proposed in the literature , with the assumption that researchers could have full access to individual-level genotype data 5–9 ., In practice , pathway analysis usually utilizes data from a single resource with limited sample size , as it can be challenging to obtain and manage individual-level GWAS data from multiple resources ., As a result , pathway analysis often fails to identify new findings beyond what have already been discovered by the single-locus tests ., To maximize the chance of discovering novel outcome-associated variants by increasing sample size , a number of consortia have been formed to conduct single-locus meta-analysis on data across multiple GWAS 10–14 ., The single-locus meta-analysis aggregates easily accessible SNP-level summary statistics from multiple studies ., Similarly , the pathway-based meta-analysis 15–21 that integrates the same type of summary data across participating studies could provide us a greater opportunity for detecting novel pathway associations ., Future association studies focusing on identified pathways would have a much-reduced multiple-comparison burden in searching for novel variants with main or complicated nonlinear joint effects on the outcome of interest ., In this paper , we developed a pathway-based meta-analysis procedure by extending the adaptive rank truncated product ( ARTP ) pathway analysis procedure 9 , which was originally developed for analyzing individual-level genotype data ., The new procedure , called Summary based ARTP ( sARTP ) , accepts input from SNP-level summary statistics , with their correlations estimated from a panel of reference samples with individual-level genotype data , such as the ones from the 1000 Genomes Project 22 , 23 ., This idea was initially used in conducting gene-based meta-analysis 24 , 25 or conditional test 26 ., As will be shown in the Results Section , sARTP usually has a power advantage over its competitors ., In addition , sARTP is specifically designed for conducting pathway-based meta-analysis using SNP-level summary statistics from multiple studies ., In real applications ( e . g . , the type 2 diabetes example described below ) , it is very common that different studies could have genotypes measured or imputed on different sets of SNPs ., As a result , the sample size used in the pathway-based meta-analysis on each SNP can be quite different ., Ignoring the difference in sample sizes across SNPs in a pathway-based meta-analysis would generate biased testing results ., Pathway analysis generally targets two types of null hypotheses 4 , including the competitive null hypothesis 15 , 16 , 18–20 , i . e . , the genes in a pathway of interest are no more associated with the outcome than any other genes outside this pathway , and the self-contained null hypothesis 17 , 21 , i . e . , none of the genes in a pathway of interest is associated with the outcome ., The sARTP procedure focuses on the self-contained null hypothesis , as our main goal is to identify outcome-associated genes or loci ., Also , as pointed out by 27 , tests for the competitive null hypothesis often assume that genotype measured at different genes are independent when evaluating the association significance level ., This assumption , which is generally invalid in practice , is unnecessary for sARTP when testing the self-contained null hypothesis ., One may refer to 27 and 4 for more discussion and comparison of these two types of hypotheses ., The pathways defined in many public databases can consist of thousands of genes and tens of thousands of SNPs ., To make the procedure applicable to large pathways , or pathways with high statistical significance , we implement sARTP with efficient and parallelizable algorithms , and adopt the direct simulation approach ( DSA ) 28 to evaluate the significance of the pathway association ., We demonstrated the validity and power advantage of sARTP through simulated and empirical data ., We applied sARTP to conduct a pathway-based meta-analysis on the association between type 2 diabetes ( T2D ) and 4 , 713 candidate pathways defined in the Molecular Signatures Database ( MSigDB ) v5 . 0 ., The analysis used SNP-level summary statistics from two sources with European ancestry ., One is generated from the Diabetes Genetics Replication and Meta-analysis ( DIAGRAM ) consortium 13 , which consists of 12 , 171 T2D cases and 56 , 862 controls across 12 GWAS ., The other one is based on a T2D GWAS with 7 , 638 T2D cases and 54 , 319 controls that were extracted from the Genetic Epidemiology Research on Aging ( GERA ) study 29 , 30 ., The novel T2D-associated pathways detected in the European population were further examined in Asians using summary data generated by the Asian Genetic Epidemiology Network ( AGEN ) consortium meta-analysis , which combined 8 GWAS of T2D with a total of 6 , 952 and 11 , 865 controls from eastern Asian populations 10 ., Here we describe the proposed method sARTP for assessing the association between a dichotomous outcome and a pre-defined pathway consisting of J genes ., The same procedure can be applied to study a quantitative outcome with minor modifications ., Firstly , we conducted a simulation study to evaluate the empirical size of sARTP and MsARTP ., Secondly , we compared empirical powers of different strategies for carrying out pathway-based meta-analysis that integrated summary statistics from multiple studies ., We also evaluated whether results from sARTP were consistent with the ones from MsARTP ., Thirdly , we compared our method to the recently developed method aSPUsPath 8 that can be used for pathway-based meta-analysis ., We used the R package , aSPU ( version 1 . 39 ) , with the default settings given in 8 , 17 to conduct the aSPUsPath test ., To demonstrate the consistency between results obtained by sARTP using SNP-level summary statistics and the ones by ARTP using individual-level genotype data , we compared pathway analysis results from three different procedures on the 4 , 713 candidate pathways using the GERA GWAS data ., Details on how those 4 , 713 pathways were pre-processed are given in the Results of T2D Pathway Analysis Section ., We applied sARTP to the SNP-level summary statistics generated from the GERA study , using either an internal or an external reference panel ., We also obtained the pathway p-values by directly applying the ARTP method to the individual-level GERA GWAS data ., Fig 1 shows the comparison among p-values from these three analyses , and demonstrates that all three approaches can generate very consistent results ., The URLs for data and software presented herein are as follows: DIAbetes Genetics Replication And Meta-analysis ( DIAGRAMv3 ) , http://diagram-consortium . org/ Genetic Epidemiology Research on Aging ( GERA , dbGaP Study Accession: phs000674 . v1 . p1 ) , http://www . ncbi . nlm . nih . gov/projects/gap/cgi-bin/study . cgi ?, study_id=phs000674 . v1 . p1 Molecular Signatures Database ( C2: curated gene sets ) , http://software . broadinstitute . org/gsea/msigdb/collections . jsp#C2 BioMart ( Homo sapiens genes NCBI36 and GRCh37 . p13 ) , http://feb2014 . archive . ensembl . org/ IMPUTE2 , https://mathgen . stats . ox . ac . uk/impute/impute_v2 . html GWAS Catalog , http://www . ebi . ac . uk/gwas/ 1000 Genomes Project ( Phase 3 , v5 , 2013/05/02 ) , ftp://ftp . 1000genomes . ebi . ac . uk/vol1/ftp/release/20130502/ aSPU , https://cran . r-project . org/web/packages/aSPU/index . html GTEx Portal v6 , http://gtexportal . org/home/ GeneCards Human Gene Database , http://www . genecards . org/ Ingenuity Pathway Analysis , http://www . ingenuity . com/ LocusZoom , http://locuszoom . sph . umich . edu/locuszoom/ ARTP2 package , https://cran . r-project . org/web/packages/ARTP2/ Web-based tool of ARTP2 , http://analysistools . nci . nih . gov/pathway/ | Introduction, Materials and Methods, Results, Discussion | Meta-analysis of multiple genome-wide association studies ( GWAS ) has become an effective approach for detecting single nucleotide polymorphism ( SNP ) associations with complex traits ., However , it is difficult to integrate the readily accessible SNP-level summary statistics from a meta-analysis into more powerful multi-marker testing procedures , which generally require individual-level genetic data ., We developed a general procedure called Summary based Adaptive Rank Truncated Product ( sARTP ) for conducting gene and pathway meta-analysis that uses only SNP-level summary statistics in combination with genotype correlation estimated from a panel of individual-level genetic data ., We demonstrated the validity and power advantage of sARTP through empirical and simulated data ., We conducted a comprehensive pathway-based meta-analysis with sARTP on type 2 diabetes ( T2D ) by integrating SNP-level summary statistics from two large studies consisting of 19 , 809 T2D cases and 111 , 181 controls with European ancestry ., Among 4 , 713 candidate pathways from which genes in neighborhoods of 170 GWAS established T2D loci were excluded , we detected 43 T2D globally significant pathways ( with Bonferroni corrected p-values < 0 . 05 ) , which included the insulin signaling pathway and T2D pathway defined by KEGG , as well as the pathways defined according to specific gene expression patterns on pancreatic adenocarcinoma , hepatocellular carcinoma , and bladder carcinoma ., Using summary data from 8 eastern Asian T2D GWAS with 6 , 952 cases and 11 , 865 controls , we showed 7 out of the 43 pathways identified in European populations remained to be significant in eastern Asians at the false discovery rate of 0 . 1 ., We created an R package and a web-based tool for sARTP with the capability to analyze pathways with thousands of genes and tens of thousands of SNPs . | As GWAS continue to grow in sample size , it is evident that these studies need to be utilized more effectively for detecting individual susceptibility variants , and more importantly , to provide insight into global genetic architecture of complex traits ., Towards this goal , identifying association with respect to a collection of variants in biological pathways can be particularly insightful for understanding how networks of genes might be affecting pathophysiology of diseases ., Here we present a new pathway analysis procedure that can be conducted using summary-level association statistics , which have become the main vehicle for performing meta-analysis of individual genetic variants across studies in large consortia ., Through simulation studies we showed the proposed method was more powerful than the existing state-of-art method ., We carried out a comprehensive pathway analysis of 4 , 713 candidate pathways on their association with T2D using two large studies with European ancestry and identified 43 T2D-associated pathways ., Further examinations of those 43 pathways in 8 Asian studies showed that some pathways were trans-ethnically associated with T2D ., This analysis clearly highlights novel T2D-associated pathways beyond what has been known from single-variant association analysis reported from largest GWAS to date ., We also identify a novel locus for T2D in the European populations at chromosome 17q21 ( rs1058018 , p = 3 . 06 × 10−8 ) . | genome-wide association studies, medicine and health sciences, variant genotypes, carcinomas, cancers and neoplasms, gastrointestinal tumors, liver diseases, genetic mapping, epidemiological methods and statistics, oncology, research design, mathematics, statistics (mathematics), test statistics, genome analysis, gastroenterology and hepatology, research and analysis methods, case-control studies, mathematical and statistical techniques, epidemiology, statistical methods, hepatocellular carcinoma, epidemiological statistics, heredity, meta-analysis, genetics, biology and life sciences, physical sciences, genomics, genomics statistics, computational biology, human genetics | null |
journal.pgen.1001300 | 2,011 | Genome-Wide Association Study of Coronary Heart Disease and Its Risk Factors in 8,090 African Americans: The NHLBI CARe Project | Coronary heart disease ( CHD ) is the leading cause of mortality in African-American men and women 1 ., The risk factors for CHD in African Americans are similar to those reported in Caucasians , but their relative impact varies between the two ethnic groups ., Multiple studies have reported that smoking , type-2 diabetes ( T2D ) , hypertension , and LDL- and HDL-cholesterol ( LDL-C and HDL-C ) are significant independent risk factors for CHD in African Americans 2–5 ., In general , hypertension and LDL-C have a larger and smaller impact on CHD risk , respectively , in African Americans compared with Caucasians 3 ., There is also extensive evidence of the role of genetic factors in the familial aggregation of CHD and its predictors in African Americans 6 ., However , the underlying genes remain largely unknown ., Recent advances in genome-wide association studies ( GWAS ) have made spectacular advances in identifying genes contributing to numerous common chronic diseases in Europeans and European Americans 7 ., There are multiple loci convincingly associated with CHD risk in Caucasians , including many genes involved in lipid metabolism , as well as novel chromosomal regions that do not appear to contribute to risk through traditional risk factors 7–14 ., However , there have been no large-scale GWAS for CHD and its risk factors in African Americans ., GWAS in African Americans is important because new genes may be identified as a result of genetic variation private to populations of African-descent , differences in allele frequencies and in patterns of linkage disequilibrium ( LD ) , differences in the relative impact of risk factors to disease , or differences in gene-environment interactions ., Here we report a large ( and for most phenotypes first ) GWAS for CHD , type-2 diabetes ( T2D ) , hypertension , LDL-C and HDL-C , and smoking in 8 , 090 African Americans as part of the National Heart , Lung , and Blood Institute ( NHLBI ) -sponsored Candidate gene Association Resource ( CARe ) Project 15 ., We genotyped 909 , 622 single nucleotide polymorphisms in 9 , 119 African Americans from the ARIC ( N\u200a=\u200a3 , 269 ) , CARDIA ( N\u200a=\u200a1 , 209 ) , CFS ( N\u200a=\u200a704 ) , JHS ( N\u200a=\u200a2 , 200 ) , and MESA ( N\u200a=\u200a1 , 737 ) population-based cohorts , on the Affymetrix Genome-Wide Human SNP Array 6 . 0 platform ., Genotypes were called using Birdseed v1 . 33 16 , and stringent quality-control filters were applied ( Tables S1 and S2 ) ., For samples that passed quality control ( N\u200a=\u200a8 , 100 ) , principal component analysis ( PCA ) using EIGENSTRAT 17 revealed only ten population outliers across all cohorts; these samples were also excluded from the analysis ( Text S1 and Figure S1 ) ., Overall , a total of 8 , 090 African Americans with very high genotype quality ( average genotype success rate of 99 . 65% ) were available for analysis ., The demographics of these participants by cohort are shown in Table, 1 . To increase our coverage of common genetic variation and statistical power , and to facilitate comparisons across genotyping platforms , we imputed genotypes in the CARe African-American populations using MACH taking into account the admixed nature of the population ( Text S1 ) 18 , 19 ., For all cohorts except CFS , single marker genetic association tests were performed by study using PLINK v1 . 06 20 under an additive genetic model ., We used linear regression for quantitative traits ( HDL-C , LDL-C , and smoking ) and logistic regression for dichotomous phenotypes ( CHD , hypertension , and T2D ) ., For CFS , family structure was modeled using linear mixed effects ( LME ) models and generalized estimating equations ( GEE ) for quantitative and dichotomous phenotypes , respectively 21 ., For all analyses , the first ten principal components were used as covariates to account for global admixture and population stratification ., A detailed description of the analysis methods and the phenotypic definitions used can be found in Text S1 ., Power calculations for the different phenotypes analyzed are summarized in Table S3; we have excellent power to find strong signals , but low to modest power for variants with weak phenotypic effects ., The inflation factors ( λs ) observed were all near unity ( Table S4 ) , suggesting that most confounders , including population stratification , were well-controlled ., We applied genomic control to the individual cohorts results and combined them using the inverse variance meta-analysis method 22 ., Inflation factors of the meta-analysis results were modest and were again scaled using genomic control ( Table S4 ) ., Quantile-quantile ( QQ ) plots of the six different meta-analyses after double genomic control corrections show that the test statistics follow the null expectations , except for the HDL-C and LDL-C meta-analyses , which show an upward departure from the null distributions at the lowest P-values ( Figure 1 ) ., This departure is caused by known genetic variants with large effects on lipid levels ( Figure S2 ) ., The main goal of this study was to identify new genetic risk factors for CHD and its predictors in African Americans ., For five traits analyzed ( we could not identify African-American replication cohorts for smoking ) , we identified SNPs with the strongest evidence of association in the CARe meta-analysis – SNPs were selected after accounting for LD to limit association signals redundancy – and sought replication using in silico data or direct genotyping in independent African-American cohorts ( Table 1 ) ., Combined results from a meta-analysis of the CARe and replication data are presented in Tables S5 , S6 , S7 , S8 , S9 and summarized in Table, 2 . We identified one novel locus that reached the generally accepted level for genome-wide significance ( P≤5×10−8 ) : SNP rs7801190 in the potassium/chloride transporter gene SLC12A9 and hypertension ( OR\u200a=\u200a1 . 31 , combined P\u200a=\u200a3 . 4×10−8 ) ., Despite reaching genome-wide significance , we are cautious in highlighting this association because it was identified using imputed genotypes ( imputation quality r2_hat\u200a=\u200a0 . 70 ) and the replication result , also obtained by imputation , was not statistically significant ( P\u200a=\u200a0 . 29 ) ., Indeed , when we directly assessed the quality of the imputation by directly genotyping rs7801190 in ARIC African-American samples ( N\u200a=\u200a2 , 572 ) , we failed to validate the observed association with hypertension ., This result suggests that the association between rs7801190 and hypertension status observed in the CARe African-American datasets is likely due to chance ., To validate our phenotype modeling and analytical strategy , we sought to replicate in the CARe meta-analyses genetic associations previously reported in populations of European ancestry ., We retrieved all index SNPs associated at genome-wide significance level with CHD , T2D , hypertension , HDL-C , LDL-C , and smoking in Caucasians as well as their proxy SNPs ( defined as markers with an r2≥0 . 5 with the index SNPs in HapMap samples of European ancestry ( CEU ) ) ( Table S10 ) 23 ., We then determined whether there was also evidence of association for the same signals in this large sample of African Americans ., We detected modest to strong evidence of replication for one locus associated with CHD , one locus with T2D , nine with HDL-C , and six with LDL-C ( Table 3 and Table S11 ) ., We did not replicate signals associated with smoking or hypertension ., Furthermore , the top ten associated SNPs in a recent hypertension GWAS performed in African Americans 24 were not associated with hypertension in the CARe meta-analysis ( different direction of effect and/or P>0 . 05 ) ., Since these hypertension association signals did not replicate in the original publication , non-replication here may result from their being falsely positive in the original report ., Although replication of some of the above loci in African-derived populations had been reported previously 25 , for most of them , the CARe results represent the first replication in populations of African ancestry ., Taking advantage of the LD patterns in African Americans ( LD breakdown over shorter distances compared with Caucasians ) , we assessed whether we could fine-map some of the associations previously reported in Caucasians ., For this , we evaluated SNPs that were correlated with the index SNP in HapMap CEU ( r2≥0 . 5 ) , but largely uncorrelated with it in HapMap samples of African descent ( YRI ) ( r2≤0 . 1 ) ., In most cases , the same signals were responsible for the associations in Caucasians and African Americans ( Table 3 and Table S11 ) ., However , we found five examples where the predominant association signals were at SNPs strongly correlated with the index SNPs in HapMap CEU but weakly or not correlated with the index SNPs in HapMap YRI: the CDKN2A/CDKN2B locus for CHD and the FADS1-3 , PLTP , LPL , and ABCA1 loci for HDL-C ( Table S12 ) ., Using available genetic association results for myocardial infarction 10 and HDL-C 26 in Caucasians , we illustrate in Figure 2 and Figure S3 how our results in African Americans can help refine association signals ., For instance , for the FADS locus , the index SNP in Caucasians ( rs174547 ) is in strong LD with the top SNP in the CARe African-American meta-analysis ( rs1535 ) in HapMap CEU ( r2\u200a=\u200a1 ) but not in HapMap YRI ( r2\u200a=\u200a0 . 09 ) ., The region of strong LD around rs174547 in HapMap CEU is 113 kb wide and includes the three FADS genes , whereas rs1535 , located in an intron of FADS2 , is in strong LD with no other markers in HapMap YRI ( Figure 2 ) ., Comparison of association signals regionally in African Americans and European-derived individuals can thus be useful in two ways: ( 1 ) they may suggest smaller chromosomal regions for re-sequencing experiments to attempt to identify causal variant ( s ) that underlie shared signals between African- and European-derived chromosomes or ( 2 ) they may indicate that the index SNPs for African and European populations are linked to distinct causal variants ., A third potentially interesting result from trans-ethnic comparison of association results is the identification of ethnic-specific association signals ., For instance , at the ABCA1 locus , three SNPs in LD ( rs4743763 , rs4149310 , and rs2515629 ) are associated with HDL-C in CARe African Americans ( P<1×10−5 ) , but not in Caucasians ( Figure S3D ) ., The optimal analytical strategy for GWAS in recently admixed populations has not been established ., In African Americans , an ideal test statistic would incorporate both genotype information as traditionally used in GWAS , but also , at each locus , the probability that a given individual has zero , one , or two copies of a European ( or African ) chromosomal segment ., This method would be particularly informative in a case where , for example , the causal allele is not in LD with any markers on the genotyping array , but is at higher frequency on one ancestral background ., To explore the benefits of such a statistical framework , we designed and applied a novel method that combines evidence of association from genotypes and local ancestry estimates; the method is described in details in Text S1 ., Briefly , we use a panel of ancestry informative markers across the genome and a new implementation of the software ANCESTRYMAP 27 to estimate , for each of the CARe African Americans genotyped , the probabilistic proportion of European ancestry ( 0–100% ) at the locus for each of the ∼900 , 000 SNPs genotyped on the Affy6 . 0 platform ., For each SNP , we can then compute association between the phenotype and both the SNP genotype and the SNP estimate of local ancestry to generate a combined score that summarizes allelic variation and admixture ., This method was used to produce the association data presented in Table 4 ., Our method to assess combined SNP- and ancestry-association was tested explicitly on CHD and its risk factors in the CARe African-American samples ( Figures S4 , S5 ) ., For each SNP , we compared the test statistic obtained using the SNP-alone or the SNP+admixture information ( in both methods , global ancestry is included as a covariate ) , focusing on markers that would not have been prioritized for follow-up replication when considering only SNP genotype association results ( Figure S6 ) ., Across the six phenotypes , we identified 12 SNPs outside the previously known loci with a P≤1×10−6 in this SNP+admixture test statistic ( Table 4 ) ., Most of these SNPs have a large allele frequency difference between the HapMap CEU and YRI individuals , suggesting that local ancestry might confound simple SNP association testing ., For instance , the frequency of the C-allele at rs8078633 near the APPBP2 gene is 100% and 18% in CEU and YRI , respectively ., The association between this SNP and HDL-C levels is weak when considering only allelic variation ( P\u200a=\u200a0 . 98 ) but becomes highly significant when evidence from the genotype and the estimate of local ancestry is combined ( P\u200a=\u200a3 . 6×10−7 ) ( Table 4 ) ., This composite approach also identified a SNP near the phospholipase B1 gene ( PLB1 ) that is strongly associated with LDL-C levels ( P\u200a=\u200a4 . 1×10−8 ) , but that would not have been noticed using traditional genotype-only association testing ( P\u200a=\u200a0 . 23 ) ( Table 4 ) ., As more large-scale GWAS in individuals of African ancestry are completed , it will be important to replicate these results ., Most large-scale genetic efforts to identify risk factors for CHD have focused so far on populations of European ancestry ., Given the prevalence of the disease in African Americans , and the development of better genotyping platforms that more completely survey common genetic variation in African-derived genomes 16 , it is now both pertinent and timely to investigate the genetics of CHD in populations of African ancestry ., The CARe Project was launched four years ago with the specific goal to create a resource for association studies of various heart- , lung- , and blood-related phenotypes across different ethnic groups 15 ., In this article , we present results from the largest GWAS to date for CHD and its risk factors in African Americans ., Despite being the largest , the size of our GWAS is modest compared to that of some European-derived consortia ., As a consequence , we had limited discovery power and did not identify novel loci specifically associated with CHD or its risk factors that reach genome-wide significance in our African-American dataset ., We also attempted to replicate in the CARe African-American participants genetic associations to CHD and its risk factors previously identified in Caucasians ., We could replicate 17 of those associations; for many of them , this was the first replication in a non-European-derived population ( Table 3 ) ., For five of these 17 associations , we showed how cross-ethnic comparisons of genetic association results may help refine genomic intervals carrying causal alleles ( Figure 2 and Figure S3 ) ., There were , however , a large number of loci originally found in Caucasians that were not replicated in the CARe meta-analyses presented in this manuscript ( Table S11 ) ., Because our sample size was relatively modest , that we used stringent statistical thresholds to declare replication in order to control our false positive rate , and that effect sizes could be weaker for given loci across different ethnic groups , our limited power probably explains why many loci did not replicate in the CARe African Americans ., Alternatively , some of these non-replications could be explained by the absence of variants within these loci associated with these traits in African Americans ., Our data does not allow us to distinguish these two possibilities , and larger replication studies in African-American cohorts will be needed to draw informative conclusions ., Taken together , our results suggest that CHD risk in African Americans is not influenced by loci with major phenotypic effect on disease risk , but rather by multiple variants of weak effect , as we have observed for CHD and other traits in Caucasians ., Because opportunities for replication and meta-analysis with other African-American cohorts are evolving rapidly , the CARe dataset is an outstanding public resource that provides a strong base for discovery of genetic contributors to CHD in non-European-derived populations ., All participants gave informed written consent ., The CARe project is approved by the ethic committees of the participating studies and of the Massachusetts Institute of Technology ., African-American participants for the GWAS were drawn from five population-based studies: Atherosclerosis Risk in Communities ( ARIC; N\u200a=\u200a3 , 269 ) , Coronary Artery Risk Development in young Adults ( CARDIA; N\u200a=\u200a1 , 209 ) , Cleveland Family Study ( CFS; N\u200a=\u200a704 ) , Jackson Heart Study ( JHS; N\u200a=\u200a2 , 200 ) , and Multi-Ethnic Study of Atherosclerosis ( MESA; N\u200a=\u200a1 , 737 ) ., Although longitudinal data is available for most participants , only information collected at recruitment was considered in this GWAS ., Replication results for top SNP associations were obtained using in silico or de novo genotyping from four African-American and African-Caribbean population-based cohorts ( Health , Aging , and Body Composition Study ( Health ABC; N\u200a=\u200a1 , 119 ) , National Health and Nutrition Examination Survey III ( NHANES III , N\u200a=\u200a1 , 720 ) , Jamaica Spanish Town ( SPT , N\u200a=\u200a1 , 746 ) and Jamaica GXE ( N\u200a=\u200a969 ) , one nested case-control panel from the population-based Multiethnic Cohort ( MEC , N\u200a=\u200a2 , 184 ) , and two case-control panels ( Cleveland Clinic , N\u200a=\u200a620 , and PennCATH , N\u200a=\u200a491 ) ., A detailed description of all cohorts and phenotype definitions used in this study is provided in Text S1 ., All discovery samples ( GWAS ) were genotyped on the Affymetrix Genome-Wide Human SNP array 6 . 0 according to the manufacturers protocol ., For replication , the MEC samples were genotyped by Taqman , and the NHANES III , Jamaica SPT , Jamaica GXE , Cleveland , and UPENN samples were genotyped using Illuminas Oligos Pool All ( OPA ) technology ., The Health ABC samples were genotyped on the Illumina Human1M-Duo BeadChip array as part of an independent GWAS; SNP results for the replication of the CARe findings were extracted and analyzed ., Several quality control ( QC ) filters were applied to the genome-wide genotype data: DNA concordance checks; sample and SNP genotyping success rate ( >95% , minor allele frequency ≥1% ) ; sample heterozygosity rate , identity-by-descent analysis to identify population outliers ( Figure S1 ) , problematic samples , and cryptic relatedness; Mendel errors rate in CFS and JHS , and SNP association with chemistry plates ., For replication , SNPs and samples with genotyping success rate <90% were excluded ., Because of the admixed nature of the participants , SNPs were not removed solely because they departed from Hardy-Weinberg equilibrium ., A detailed description of the quality control checks applied to the discovery ( GWAS ) and replication genotyping data can be found in Text S1 ., To increase coverage and facilitate comparison with other datasets , we imputed genotype data using MACH v1 . 0 . 16 19 ., We built a panel of reference haplotypes using HapMap phase II CEU and YRI data , and imputed SNP genotypes using all Affymetrix 6 . 0 SNPs that passed the QC steps described above ., Using and independent dataset of ∼12 , 000 SNPs genotyped on the same DNA but with a different platform , we estimated an allelic concordance rate of 95 . 6% ( Text S1 ) ., SNP-only based genetic association analysis of quantitative ( HDL-C , LDL-C , smoking ) and dichotomous ( coronary heart disease , type-2 diabetes , hypertension ) traits were carried out using linear and logistic statistical framework , respectively , in PLINK ( unrelated cohorts: ARIC , CARDIA , JHS , MESA , UPENN , Cleveland , MEC , NHANES III , and Health ABC ) or using R scripts that model family structure ( related cohort: CFS ) 28 ., For the cohorts with genome-wide genotyping data available , the first ten principal components were included in each analysis to account for population stratification and admixture ., The method to estimate local ancestry was implemented in ANCESTRYMAP and is described in details in Text S1 ., To combine allelic and local ancestry information ( Table 4 ) , we calculated a chi-square statistic with two degrees-of-freedom ., Association results were combined across cohorts using an inverse variance meta-analysis approach as implemented in metal . CARe: http://www . broadinstitute . org/gen_analysis/care/index . php/Main_Page; MACH: http://www . sph . umich . edu/csg/abecasis/MACH; METAL: http://www . sph . umich . edu/csg/abecasis/Metal/index . html; PLINK: http://pngu . mgh . harvard . edu/~purcell/plink . | Introduction, Results, Discussion, Materials and Methods | Coronary heart disease ( CHD ) is the leading cause of mortality in African Americans ., To identify common genetic polymorphisms associated with CHD and its risk factors ( LDL- and HDL-cholesterol ( LDL-C and HDL-C ) , hypertension , smoking , and type-2 diabetes ) in individuals of African ancestry , we performed a genome-wide association study ( GWAS ) in 8 , 090 African Americans from five population-based cohorts ., We replicated 17 loci previously associated with CHD or its risk factors in Caucasians ., For five of these regions ( CHD: CDKN2A/CDKN2B; HDL-C: FADS1-3 , PLTP , LPL , and ABCA1 ) , we could leverage the distinct linkage disequilibrium ( LD ) patterns in African Americans to identify DNA polymorphisms more strongly associated with the phenotypes than the previously reported index SNPs found in Caucasian populations ., We also developed a new approach for association testing in admixed populations that uses allelic and local ancestry variation ., Using this method , we discovered several loci that would have been missed using the basic allelic and global ancestry information only ., Our conclusions suggest that no major loci uniquely explain the high prevalence of CHD in African Americans ., Our project has developed resources and methods that address both admixture- and SNP-association to maximize power for genetic discovery in even larger African-American consortia . | To date , most large-scale genome-wide association studies ( GWAS ) carried out to identify risk factors for complex human diseases and traits have focused on population of European ancestry ., It is currently unknown whether the same loci associated with complex diseases and traits in Caucasians will replicate in population of African ancestry ., Here , we conducted a large GWAS to identify common DNA polymorphisms associated with coronary heart disease ( CHD ) and its risk factors ( type-2 diabetes , hypertension , smoking status , and LDL- and HDL-cholesterol ) in 8 , 090 African Americans as part of the NHLBI Candidate gene Association Resource ( CARe ) Project ., We replicated 17 associations previously reported in Caucasians , suggesting that the same loci carry common DNA sequence variants associated with CHD and its risk factors in Caucasians and African Americans ., At five of these 17 loci , we used the different patterns of linkage disequilibrium between populations of European and African ancestry to identify DNA sequence variants more strongly associated with phenotypes than the index SNPs found in Caucasians , suggesting smaller genomic intervals to search for causal alleles ., We also used the CARe data to develop new statistical methods to perform association studies in admixed populations ., The CARe Project data represent an extraordinary resource to expand our understanding of the genetics of complex diseases and traits in non-European-derived populations . | genetics and genomics/complex traits, cardiovascular disorders/hypertension, genetics and genomics/population genetics, cardiovascular disorders/coronary artery disease, cardiovascular disorders/myocardial infarction, diabetes and endocrinology/type 2 diabetes | null |
journal.pgen.1005111 | 2,015 | Host Genetic Variation Influences Gene Expression Response to Rhinovirus Infection | Rhinovirus ( RV ) is the most prevalent human respiratory pathogen 1 ., It was discovered as the predominant cause of the common cold over 50 years ago 2 ., Longitudinal studies indicate that nearly all individuals experience at least one RV infection by two years of age 3 ., Each year following , pre-school age children experience six 4 and adults experience two to three 5 RV infections on average ., Recent studies have shown that infection with RV results in a broad spectrum of illness severity , ranging from asymptomatic infections to severe lower respiratory illnesses such as bronchiolitis and pneumonia 6 ., In addition , RV infections contribute to the morbidity of chronic respiratory illnesses such as asthma , chronic obstructive pulmonary disease ( COPD ) , and cystic fibrosis ( CF ) 6 ., The diversity in response to RV infection is likely attributable to , at least in part , inter-individual variation in the host genome ., Indeed , genetic variation in the promoter region of the IL-10 gene was shown to influence severity of RV illnesses in a small sample of 18 subjects 7 and polymorphisms at the 17q12-q21 asthma locus were associated with both the occurrence and number of RV wheezing illnesses in early life 8 ., Beyond these few associations however , the genetic and/or mechanistic basis for the vast inter-individual variation in the response to RV infection is not well understood ., Many cell types are involved in the immune response to RV 9 ., Genome-wide gene expression response to RV infection was previously studied in bronchial epithelial cells 10 , 11 and in nasal epithelial scrapings 12 ., While the nasal mucosa is considered the primary site of RV replication 13 , RV genomes were found in pericardial fluid , stool , urine , plasma , and serum samples of children with respiratory illnesses , suggesting that systemic infection of RV occurs 14–16 ., In addition , RV infection induces cytokine production from monocytes and macrophages without productive viral replication 17 , raising the possibility that RV-associated respiratory illnesses may result from virus-induced inflammatory cytokines rather than to cytopathic effects of RV per se 18 ., To date , however , there have been no genome-wide studies of gene expression response to RV infection in peripheral blood mononuclear cells ( PBMCs ) , or studies characterizing the genetic architecture of inter-individual regulatory variation in gene expression response to RV ., To begin addressing this gap , we have collected and analyzed gene expression data in PBMCs of 98 individuals , before and after RV infection in vitro ., Our study design allowed us to provide a comprehensive view of the regulatory variation involved in gene expression levels ( eQTLs ) in uninfected and RV-infected PBMCs ., We also identified genetic variations that are specifically associated with differences in gene expression response ( reQTLs ) to RV infection; these are loci that interact , directly or indirectly , with the infection process and likely include a subset of loci that contribute to the inter-individual variation in the clinical response to RV infection ., We obtained genome-wide array-based gene expression data from uninfected and RV-infected PBMCs from 98 unrelated adults ( GEO accession number: GSE53543 ) ., We detected as expressed 13 , 881 autosomal probes ( targeting 10 , 893 genes; see Materials and Methods ) ., Across the 196 samples ( uninfected and RV-infected paired samples from 98 individuals ) , gene expression profiles clearly clustered into two major groups by treatment status ( by PCA; see Methods; S1 Fig and S2 Fig ) ., We next identified the differentially expressed genes between uninfected and RV-infected PBMCs; of the 10 , 893 expressed genes , 2 , 242 were up-regulated and 3 , 779 were down-regulated in RV-infected compared to uninfected PBMCs ( at the Bonferroni corrected significance threshold of P<4 . 60x10-6; Fig . 1A , S1 Dataset ) ., We note that with a samples size of 98 individuals we have considerable power to detect differences in gene expression levels ., Indeed , the vast majority of effect sizes we detected as significant were minor ( Fig . 1A ) ., We focused , therefore , on the 271 genes that were both significantly differentially expressed and showed at least two-fold difference in expression between uninfected and RV-infected PBMCs ., Although a two-fold threshold is arbitrary , it is consistent with the thresholds used in other studies of gene expression response to RV 11 , 12 and thus provides some measure of consistency with previous reports ., In fact , considering our data in the context of other studies we observed a large overlap between the genes that respond to RV in different cell types ( 86 . 5% overlap between PBMCs and bronchial epithelial cells S1 Table and 73 . 3% overlap between PBMCs and nasal epithelial scrapings S2 Table ) ., Additionally , pathway analysis of the 271 genes revealed enrichment for immune and viral response related pathways , as expected ( Fig . 1B , Fig . 1C , S3 Table , S4 Table ) ., We next identified local ( putative cis ) eQTLs in uninfected and RV-infected PBMCs by testing for associations between expression levels and genetic variation at loci within 1 Mb windows from the nearest annotated end of each expressed gene ., We performed this analysis separately in uninfected and RV-infected PBMCs , and in both cases we used the SNP with the minimum P value observed for each gene to assess the evidence of an eQTL at that locus ., At a false discovery rate ( FDR ) of 5% based on permutations ( which also consider the minimum P value for each gene ) , we identified 521 genes with local eQTLs in uninfected PBMCs ( Fig . 2A , S2 Dataset ) and 523 genes with local eQTLs in RV-infected PBMCs ( Fig . 2A , S2 Dataset ) ., For each significant local eQTL-gene expression pair in uninfected and in RV-infected PBMCs , we compared the evidence of eQTL association across conditions ( S3 Fig ) ., Pearson correlation of eQTL association P values was 0 . 84 for the 521 eQTLs in uninfected cells ( P<10–15 ) and 0 . 81 for the 523 eQTLs in RV-infected cells ( P<10–15 ) , suggesting that a significant proportion of genetic regulation of gene expression is maintained between uninfected and RV-infected PBMCs ., We reasoned that SNPs that are specifically associated with gene expression response to RV infection ( reQTLs ) are likely to have a role in RV-specific response ., To identify reQTLs , we tested associations between the expression response ( estimated as Log2 fold change in gene expression response to RV infection ) and genetic variation in loci that are within 1 Mb of the nearest end of each expressed gene ., This analysis revealed local reQTLs for 38 genes at the FDR 5% threshold ( Fig . 2A , S2 Dataset ) ., For 25 of the genes with reQTLs , genotypic effect on gene expression at FDR 5% threshold was present only in uninfected or only in RV-infected cells ( S2 Dataset ) ., For the remaining 13 genes , genotypic effect on gene expression was present in both uninfected and RV-infected cells at FDR 5% threshold , but the effect size of association was significantly different between conditions ( S2 Dataset ) ., Overall , expression levels of genes with reQTLs were slightly higher in the condition where absolute genotypic effect size of the reQTL was larger ( S4 Fig ) ., This analysis raises the possibility that slight differences in power to map eQTLs may explain a subset of our observations , but we note that the average difference in expression level is small and is not consistent across all genes with reQTLs ., The 38 genes with reQTLs included those with known functions in viral response , such as UBA7 , OAS1 , IRF5 ( Fig . 2B ) ., The protein product of UBA7 ( ubiquitin-like modifier activating enzyme 7 ) is involved in the activation of a critical antiviral protein ISG15 ( interferon-stimulated gene 15 ) 19 , 20 ., OAS1 ( 2’-5’-oligoadenylate synthase 1 ) activates latent RNase L following viral infections and results in degradation of viral RNA 21 , 22 ., Similarly , IRF5 ( interferon regulatory factor 5 ) protein is a direct transducer of virus-mediated signaling 23 ., In addition , a subset of the 38 genes with local reQTLs have previously been associated with immune or RV-associated diseases ( such as asthma , COPD , and CF; as catalogued by the Genetic Association Database 24 ) ., Specifically , eight genes ( IRF5 , UBA7 , TMTC1 , GSTM3 , FBN2 , OAS1 , PODXL , ITGA2 ) were previously associated with immune system diseases in at least one study 25–29 ., Notably , three genes ( IRF5 , GSTM3 , FBN2 ) have been associated with asthma 29–31 , two ( GSTM3 , MSR1 ) with COPD 32 , 33 and one ( GSTM3 ) with CF 34 ., It should be noted , however , that while these observations represent a slight enrichment compared to genome-wide expectations ( S5 Fig ) , enrichment P values were not significant ., In an attempt to fine map the causal reQTLs , we next repeated reQTL mapping for the 38 genes with significant reQTLs using imputed genotype data ( see Material and Methods for details of imputation ) ., We compared the most significant reQTL for each gene based on imputed vs . genotyped data ( S5 Table ) ., For 16 of the genes , the most significant reQTL based on genotyped data remained the most significant reQTL based on imputed genotype data ., However , for 12 of the genes , multiple SNPs had the smallest reQTL P value in the imputed genotypes , suggesting that fine mapping efforts using imputed genotype data in a sample size of 98 individuals is not sufficient to identify the causal reQTL ., Previous studies have suggested that reQTLs can often be found within the binding sites of transcription factors that have different levels of activity between the conditions being tested 35 , 36 ., We examined this in our data by considering the overlap between known protein binding sites ( based on ENCODE ChIP-Seq data 37 ) and the 38 reQTLs identified in our study ., This analysis is challenging because of the uncertainty in identifying the causal reQTL association ., We therefore considered for this analysis the reQTLs and all other common SNPs in high LD with the reQTLs ( r2>0 . 8 and MAF>0 . 1 based on 1000 Genomes Phase I European Population ) as input SNPs ., We then extracted 1 , 000 control SNPs ( matched based on MAF , gene density , distance to nearest gene , number of SNPs in LD ) per each input SNP and examined the overlap between protein binding sites and each unique SNP in the input and control SNP lists ., This analysis revealed that STAT2 was the most substantially enriched transcription factor ( 16 . 30 fold; P<10–6; S6 Table and Fig . 3A ) with binding sites overlapping reQTL loci SNPs relative to the control SNPs ( see Materials and Methods for details of enrichment analysis ) ., STAT2 is a critical regulator of anti-viral response 38 ., In our study , STAT2 gene expression increased by 3 . 68 fold ( S1 Dataset; P<10–77 ) in response to RV infection ., Based on this observation , we hypothesized that reQTL loci SNPs that lie within STAT2 binding sites should have stronger regulatory effects on gene expression in RV-infected cells relative to uninfected cells ., As expected , for all five targets of STAT2 among our reQTL regions , the eQTL effects on gene expression were stronger in RV-infected cells compared to uninfected cells ( Fig . 3B and S6 Fig ) ., Consistent with this observation is that all the STAT2 ChIP-Seq signals ( based on ENCODE ChIP-Seq data ) in our reQTL loci were identified in the IFNα treated human K562 cell line ( S7 Table and Fig . 3A ) ., Additionally , our results imply that the SNPs residing in STAT2 binding sites are more likely to be the causal SNP in the reQTL loci of EXOSC9 , SLFN5 , PRR24 , OAS1 , and ARL5B ., In fact , causality of the reQTL of SLFN5 gene , rs11080327 , was recently demonstrated by luciferase reporter assays 35 ., Lastly , we searched for distant ( putatively trans ) eQTLs and reQTLs by testing associations between SNP-gene combinations for which the SNP distance from the nearest end of the gene was more than 5 Mb ., We used the minimum P value observed per gene to assess the significance of the distant eQTL or reQTL ., We identified 11 and 15 genes with distant eQTLs in uninfected and in RV-infected PBMCs , respectively ( at an FDR 5% based on permutations; S3 Dataset ) ., We found no direct evidence for distant reQTLs when we considered genetic associations with the expression response to infection ( S3 Dataset ) , yet 6 of the trans eQTLs were specific to either the infected or uninfected cells ., Because trans eQTL findings based on microarray expression data tend to suffer more from a high degree of false-positives , partly due to cross-hybridizations , we considered carefully each significant finding ., For each gene that was putatively associated with a trans eQTL , we re-mapped each of its probes within 2 Mb of the trans eQTL using SHRiMP 39 , as previously described 40 ., We discarded all trans eQTLs whose associated trans-probe mapped in its vicinity ., After exclusions , only 4 and 6 genes with distant eQTLs in uninfected and RV-infected cells , respectively , remained; 3 of them were common to both conditions ( S8 Table ) ., Thus , 4 trans eQTLs may be reQTLs , yet considerations of incomplete power for detecting trans eQTLs in our study call for caution in this interpretation ., Mapping of condition-specific or response eQTLs in a genome-wide level is an emerging area of research with the promise of understanding biology of infectious diseases 41 , 42 , response to pharmaceutical treatment 36 , and inter-individual variation in immune response more broadly 35 , 43 ., Here , we report 38 local reQTLs that are associated with inter-individual variation in gene expression response to RV infection ., These loci are likely to interact , directly or indirectly , with the infection process ., Because infection with rhinovirus causes significant changes in regulatory activity of the identified reQTLs , these variants represent promising candidates for susceptibility and response to RV infections in vivo ., In support of this reasoning , we pointed to eight genes with RV-reQTLs that have been previously associated with immune diseases , and four that were previously associated with respiratory diseases ( either asthma , COPD , or CF ) ., We note that while the numbers of disease-associated genes among genes with reQTLs were greater than the genome-wide expectations , none of the enrichment P values were significant ., That said , it is important to emphasize that RV infections , in general , are thought to affect the morbidity of the chronic respiratory illnesses rather than the risk of developing the disease 6 ., Therefore , it is possible that the reQTLs identified here are more likely to influence the severity of the respiratory illnesses rather than the occurrence , which has been the focus of the majority of the association studies performed thus far ., Similarly , risk of developing the common cold might be more directly influenced by the reQTLs identified here because RV is causally associated with development of the common cold ., However , to our knowledge , there have been no association studies on the frequency or severity of common colds ., It is also possible that at least some of the reQTLs identified here are functional in response to a broader range of stimuli , including other pathogens ., In fact , six ( SLFN5 , ARL5B , SPTLC2 , IRF5 , ADCY3 , CCDC146 ) of the 38 genes implicated in our reQTL mapping were also identified in a recent reQTL mapping study for Escherichia coli lipopolysaccharide , influenza , and interferon-β in dendritic cells 35 ., Further comparative studies of reQTLs will be necessary to disentangle stimulus-specific and shared reQTLs ., We also note that some of the reQTLs identified in our study may be confounded due to cell type heterogeneity in PBMCs ., In our study , we were unable to count cell subsets of PBMCs before and after RV infection ., If for instance , cell subset proportions of PBMCs change in response to RV infection , statistical power to identify cell-type specific eQTLs may differ between uninfected and RV-infected PBMCs and this may potentially lead to identification of cell-type specific eQTLs as reQTLs ., Similarly , it is possible that cell type heterogeneity in PBMCs may mask identification of cell-type specific reQTLs , especially those that are specific to rare cell subsets of PBMCs ., However , we also appreciate the fact that gene expression response in a complex system of interacting cells as in PBMCs might be more relevant to true physiological responses than those observed in purified cell subsets ., The enrichment of STAT2 binding sites among reQTL regions highlights the role of condition-specific transcription factors in gene-by-environment interactions ., Our results suggest that transcription factor activation upon RV-infection reveals SNPs with regulatory activity that could not be identified in uninfected PBMCs ., This hypothesis is also supported by the fact that all STAT2 ChIP-Seq signals in our reQTL regions were identified in the IFNα treated human K562 cell line ., IFNα is involved in the innate immune response against viral infections and our results therefore suggest that RV infection may activate STAT2 through the IFNα signaling pathway ., In conclusion , we have provided a comprehensive genome-wide view of host genetic variation that is associated with gene expression response to rhinovirus infection ., The reQTLs identified here are promising candidates to influence both the frequency and the severity of RV related respiratory illnesses ., Additionally , our results contribute to the field of genotype-by-environment interactions and might further help to disentangle stimulus-specific and shared reQTLs ., One hundred unrelated adult volunteers ( 49 males and 51 females; age range 19–60 ) were recruited between July and November 2011 to study the genotype-specific effects of RV infection on gene expression patterns in PBMCs ., Informed written consent was obtained from each study participant ., This study was approved by the Institutional Review Board at the University of Chicago ., Twenty ml of blood was drawn from each participant ., PBMCs were isolated from whole blood samples by Ficoll-Paque separation 44 ., From each subject , 4x106 PBMCs were treated with media alone for 24 hours and 4x106 PBMCs were treated with media containing RV16 for 24 hours ., The multiplicity of infection was 10 plaque-forming units per cell ., DNA was extracted on the day of sample collection using QIAamp DNA Blood Mini Kit; concentrations were measured on a Nanodrop ND-100 Spectrophotometer ., Genotyping of 100 individuals was performed using the Axiom Genome Wide Human Array Plate–CEU , which interrogates 669 , 059 SNPs ., Genotype calls were extracted from the raw data using Affymetrix Power Tools software ., One individual with genotype call rate less than 97% and one individual that failed Affymetrix gender call were excluded from all further analyses ., After exclusions , the sample size was 98 ( 49 males and 49 females; age range 19–60 ) ., Following quality-control checks ( Hardy-Weinberg equilibrium P>10–6 , MAF>0 . 1 , SNP call rate>95% ) , 382 , 855 SNPs were retained and 373 , 312 autosomal SNPs with unique SNP identifiers were used in subsequent analyses ., Global proportions of European , Asian , and African ancestry in our samples were estimated by using the program ADMIXTURE 45 and assuming 3 ancestral populations ( K = 3 ) ( S7 Fig ) ., Subjects from the Phase 3 HapMap CEU ( Utah residents with Northern and Western European ancestry from the CEPH collection ) , CHB ( Han Chinese in Beijing , China ) , JPT ( Japanese in Tokyo , Japan ) , and YRI ( Yoruba in Ibadan , Nigeria ) were included as reference populations ., 85 of the individuals had over 80% European ancestry , six of them had over 80% Asian ancestry and the remaining seven individuals had relatively more mixed ancestry fractions ., Ancestry estimates were taken into account in further analyses to correct for the potential effects of ancestry on gene expression profiles ., 373 , 312 SNPs that passed the genotyping quality-control checks were used to perform the pre-phasing of the chromosomes using SHAPEIT ( v2 . r790 ) 46 ., Imputation was performed using IMPUTE2 ( 2 . 3 . 1 ) 47 over genomic regions of 5 Mb , as recommended ., For both pre-phasing and imputation , 1000 Genomes Phase 3 data were used as the reference panel ., 78 , 091 , 231 autosomal variant sites were imputed across 98 individuals ., After quality-control checks ( Hardy-Weinberg equilibrium P>10–6 , MAF>0 . 1 , SNP call rate>95% ) , 3 , 722 , 989 of the imputed variants were retained ., Total RNA was extracted after 24-hour incubation , using the RNeasy Plus Mini Kit; concentrations were measured on a Nanodrop ND-100 Spectrophotometer and quality was assessed using an Agilent 2100 Bioanalyzer ., Genome wide gene expression profiling of uninfected and RV-infected PBMCs was obtained using Illumina HumanHT-12 v4 Expression BeadChip arrays , which targets 47 , 305 probes ., The cDNA synthesis , labeling , and hybridization of RNA to the microarrays were performed at the University of Chicago Functional Genomics Core ., The process of quality-control checks ( probes that mapped to unique Ensembl gene IDs , probes that did not contain any HapMap SNPs with MAF>0 . 01 in the CEU population , probes that targeted autosomal chromosomes ) resulted in retention of 26 , 440 probes ., Among those , 13 , 881 probes that were detected as expressed in PBMCs ( detection P<0 . 05 in at least 25% of the samples ) were used in subsequent analyses ., Low-level microarray analyses were performed in R ( http://www . R-project . org ) , using the Bioconductor software package lumi 48 ., Probe intensity estimates were log2-transformed and rank-invariate normalized ., Linear models were used to test the relationship between each known covariate ( gender , virus batch , processing day , chip number , PBMC count , age , and ADMIXTURE’s 45 Q estimates ( the ancestry fractions ) ) and the principal components that explain at least 5% of the total variance in the gene expression data ., ( S1 Fig ) ., Processing day was significantly associated with the Principal Component 2 and hence it was included as an “adjustment variable” when performing SVA 49 ., SVA analysis yielded no significant surrogate variables when “adjustment variable” of processing day was used with “variable of interest” of treatment ., The effects of processing day were regressed out of the gene expression data prior to further analyses ( S2 Fig ) ., Median probe intensity estimates per gene were calculated as the expression levels for 10 , 893 genes and used in all further analyses ., Differentially expressed genes between uninfected PBMCs and RV-infected PBMCs were identified using a paired t-test in R statistical environment ., Significance was calculated using the Bonferroni correction at α = 0 . 05 ( P<4 . 60x10-6 ) ., The DAVID bioinformatics database 50 was used to test for enrichment of Gene Ontology ( GO ) categories ( BP_ALL ) and Kyoto Encyclopedia of Genes and Genomes ( KEGG ) pathways among the genes that were both statistically differentially expressed and had ≥2-fold change in response to RV infection ., 10 , 893 genes that were detected as expressed in our data were used as the background set in all enrichment analyses ., Enrichment P values were calculated using a modified Fisher Exact test ( EASE Score ) ., Associations between local ( defined as SNP-gene pairs within 1 Mb window from the nearest end of the gene ) and distant ( defined as SNP-gene pairs for which the SNP distance from the nearest end of the gene was more than 5 Mb ) SNPs and gene expression in uninfected and in RV-infected cells were tested using linear regressions with additive genotype effects and taking ancestry estimates into account ., Similarly , associations between local and distant SNPs and gene expression response ( Log2 fold change in gene expression response to RV infection ) were tested using linear regressions with additive genotype effects and taking ancestry estimates into account ., All the analyses were performed as implemented in “linear” model of the Matrix eQTL package 51 ., The minimum P value observed for each gene was recorded and used as the evidence of eQTL or reQTL association ., To estimate the FDR , each of the phenotype data ( gene expression or gene expression response ) were shuffled 100 times and linear regressions were repeated using each set of permuted data ., The minimum P value for each gene was recorded and used as the empirical null distribution ., Permutation-based FDR was calculated using fdrci package 52 in R . Genes previously reported to be associated with immune diseases ( Disease Class ) and RV-related respiratory illnesses ( asthma , COPD , CF ) ( Disease Terms ) were downloaded from the Genetic Association Database ( GADCDC data as of 08/18/2014 ) ., Enrichment P values were calculated using Fishers exact test ., SNP ‒ human ChIP-Seq ( based on ENCODE data 37 ) annotation data were downloaded from HaploReg v2 53 in December 2014 ., For each ChIP annotation across all available cell types and conditions , the frequency of overlap between unique reQTL loci SNPs was compared with the frequency of overlap with a set of unique control SNPs ., reQTL loci SNPs included the list of reQTLs and all other common SNPs in high LD with the reQTLs; r2>0 . 8 and MAF>0 . 1 based on 1000 Genomes Phase I European Population ., 1 , 000 control SNPs ( matched based on MAF ±5% point , gene density ±50% , distance to nearest gene ±50% , number of SNPs in LD ±50% using r2 0 . 5 ) per each input SNP were pulled using the SNPsnap webserver 54 ., Binomial P values and fold-enrichments were calculated for proteins with at least two ChIP-Seq signal overlapping with the reQTL loci SNPs ., For STAT2 ChIP annotation , the frequency of overlap between reQTL loci SNPs and the matching control SNPs was additionally compared within IFNα-30 minute and IFNα-6 hour treated human K562 cell line ., To minimize the false-positive trans eQTL findings , each probe targeting the genes with significant trans eQTLs was re-mapped using SHRIMP 39 with relaxed mapping setting that was previously described; match score of 10 , mismatch score of 0 , gap open penalty of −250 , gap extension penalty of −100 , and minimal Smith-Waterman score of 30% 40 ., Trans eQTLs whose associated trans-probe mapped in its vicinity were excluded . | Introduction, Results, Discussion, Materials and Methods | Rhinovirus ( RV ) is the most prevalent human respiratory virus and is responsible for at least half of all common colds ., RV infections may result in a broad spectrum of effects that range from asymptomatic infections to severe lower respiratory illnesses ., The basis for inter-individual variation in the response to RV infection is not well understood ., In this study , we explored whether host genetic variation is associated with variation in gene expression response to RV infections between individuals ., To do so , we obtained genome-wide genotype and gene expression data in uninfected and RV-infected peripheral blood mononuclear cells ( PBMCs ) from 98 individuals ., We mapped local and distant genetic variation that is associated with inter-individual differences in gene expression levels ( eQTLs ) in both uninfected and RV-infected cells ., We focused specifically on response eQTLs ( reQTLs ) , namely , genetic associations with inter-individual variation in gene expression response to RV infection ., We identified local reQTLs for 38 genes , including genes with known functions in viral response ( UBA7 , OAS1 , IRF5 ) and genes that have been associated with immune and RV-related diseases ( e . g . , ITGA2 , MSR1 , GSTM3 ) ., The putative regulatory regions of genes with reQTLs were enriched for binding sites of virus-activated STAT2 , highlighting the role of condition-specific transcription factors in genotype-by-environment interactions ., Overall , we suggest that the 38 loci associated with inter-individual variation in gene expression response to RV-infection represent promising candidates for affecting immune and RV-related respiratory diseases . | Rhinovirus ( RV ) is the predominant cause of the common cold ., However , infections with RV result in a broad spectrum of effects ranging from asymptomatic infections to severe lower respiratory illnesses ., We hypothesized that diversity in response to RV-infections is , at least in part , due to variation in the host genome ., To address this , we mapped the genetic variations that are associated with gene expression response ( reQTLs ) to RV-infection in PBMCs ., Here , we report local reQTLs for 38 genes including those with known functions in viral response such as UBA7 , OAS1 , IRF5 and those that have been previously associated with immune and RV-related diseases ( e . g . , ITGA2 , MSR1 , GSTM3 ) ., We also show that reQTL regions are enriched for binding sites of the virus-activated STAT2 transcription factor , suggesting a potential mechanism of action for five of the reQTLs identified ., Overall , the reQTLs we identified represent promising candidates to affect individual’s immune response to RV infections and further targeted studies of the reQTL regions might lead to improved control and treatment of RV-associated immune and respiratory diseases . | null | null |
journal.pgen.1003242 | 2,013 | Gene Copy-Number Polymorphism Caused by Retrotransposition in Humans | In recent years it has become apparent that changes in gene copy-number introduced by genomic duplication and deletion events are an important force driving adaptive evolution 1 ., Examples of adaptive gene gains and losses have been found in a variety of organisms , including humans 2–4 and Drosophila melanogaster 5 , 6 ., Much attention has focused on gene duplications in particular , as they may facilitate the evolution of new gene functions 7 , 8 ., Given that all new gene duplicates must arise as polymorphisms , and the fact that genomic duplications and deletions can have negative phenotypic consequences 9–11 , massive efforts have been made to identify regions of the genome differing in copy-number , referred to as copy-number variants ( CNVs ) , among humans 2 , 12–15 and other species ( e . g . , refs . 16–18 ) ., These studies have revealed extensive copy-number variation especially within humans , with any two African individuals differing in copy-number at over 100 genes 2 , 19 ., It has been suggested that in humans the vast majority of gene duplications contributing to this variation result in a new copy located adjacent to the original gene 14 ., However , a substantial number of new duplicates are inserted far from the original locus in humans and other mammals 20 , 21 , including genes duplicated by retrotransposition 22 , 23 ., These retrocopies , which are created when a messenger RNA transcript is reverse-transcribed and reinserted into a different location in the genome , are an especially interesting class of gene duplicate for several reasons ., First , a new retrocopy will contain an entire coding sequence except when derived from an incomplete transcript ., In addition , retrocopies occasionally carry promoter elements located downstream of the retrotranscribed transcripts transcription start site but located upstream of an alternative transcription start site 24 ., Evidence that a substantial proportion of gene retrotransposition events result in functional gene copies , called retrogenes , come from both mammals 25 , 26 and Drosophila 27 ., In addition , patterns of gene movement onto and off of the X chromosome in mammals and off of the X in D . melanogaster suggest that many retrogenes are subject to positive selection ( e . g . , refs . 28–30 ) ., Finally , processed pseudogenes , inactivated gene copies created by retrotransposition , have also been shown to influence expression levels of the parental gene copy , potentially disrupting its function 31 , 32 ., Despite the potentially important evolutionary and phenotypic consequences of retrogenes , current CNV-detection approaches are largely unable to find them ., In fact , only one study of copy-number variation in humans was able to detect any polymorphic retrogenes 2 ., Previously , we developed a method capable of leveraging next-generation sequence data to detect gene copy-number variants caused by retrotransposition , or retroCNVs , and used it to reveal that 13% of gene copy-number polymorphisms in D . melanogaster are caused by retrotransposition 30 ., Although a similar method has been applied to detect retroCNVs in humans 33 , there has been no detailed analysis of retroCNVs in humans to date ., Here we apply an improved method to a number of sequenced human genomes , including data from the 1000 Genomes Project 34 ., We find a surprising amount of variation due to retroCNVs within the human population—accounting for ∼12 genes differing in copy-number between any two individuals ., By comparing retroCNV patterns to retrogene divergence , we reveal that retrotransposition is an important source of both adaptive and deleterious mutations in humans ., We also find evidence that some of these retroCNVs may currently be under positive selection in humans ., These findings underscore the functional and evolutionary importance of gene duplication via retrotransposition , and suggest that further study of retrogenes will illuminate the extent to which these retroCNVs affect human phenotypes and drive adaptive evolution ., In order to detect polymorphic retrocopies of protein coding genes segregating in human populations , we searched for evidence of retrocopy insertion sites using sequence reads from two human genomes that we sequenced ourselves with the SOLiD technology ( denoted AAC and SJS ) , and additional genomes from the 1000 Genomes Project 34 ., Briefly , this approach works by searching for paired-end reads spanning insertion sites of retrocopies present in the reference genome but absent from a resequenced genome ( Figure 1a ) , or vice-versa ( Figure 1b ) ., We also searched low-coverage genomes resequenced for the 1000 Genomes Project 34 for exon-exon junction-spanning reads indicative of retroCNVs ( Figure 1c ) , similar to our previous approach 30 ., Because the whole genome must be searched in order to discover retroCNV insertions absent from the reference genome , such retroCNVs were initially discovered using a smaller set of 17 individuals ( Table S1; Materials and Methods ) ., These retroCNVs were then genotyped using paired-end sequence data from three subpopulations from the 1000 Genomes Project: 52 Yoruban individuals in Nigeria ( referred to as the YRI subpopulation ) , 41 individuals of European ancestry in Utah ( referred to as CEU ) , and 56 Han Chinese individuals and Japanese individuals from Tokyo ( referred to as ASI ) ., Because of this ascertainment scheme , these retroCNVs are expected to be biased towards higher frequencies than if they were discovered using the entire set of sequenced genomes ., RetroCNVs present in the reference genome were identified using paired-end reads from all individuals sequenced for the 1000 Genomes Project , and are therefore unaffected by any ascertainment bias ., We correct for this difference in ascertainment schemes where necessary in the analyses presented here ., We find that our computational approach for retroCNV identification has high specificity and sensitivity , allowing us to estimate the contribution of retrotransposition to gene copy-number polymorphism in humans ., We identified 91 retroCNVs in total , finding that these polymorphisms account for 11 . 9 genes differing in copy-number between any two African individuals on average ., Given that a recent comparison of pairs of individual human genomes has revealed gene copy-number differences at 105 genes on average ( based on data from ref . 2 ) , our results suggest that retroCNVs could account for a sizable minority of human gene copy-number polymorphisms ( although retroCNVs may often be non-functional ) ., We were able to determine the insertion sites of 39 retroCNVs ( 18 present in the reference genome; 21 absent from the reference ) , and verify that retrocopy presence was the derived state for each of these ( Materials and Methods ) ; the remaining 52 retroCNVs were identified from reads spanning exon-exon junctions only and therefore have unknown insertion loci ., While many of these retrocopies may contain only fragments of coding sequence , perhaps due to the low processivity of reverse-transcriptase or partial degradation of the mRNA used as template , we found that at least 41 . 8% ( accounting for ∼6 complete gene copy-number differences between any two African genomes ) of the retrocopies across all genomes are complete or near-complete retrogenes which may have the potential to be functional ( see Materials and Methods ) ., To estimate the fraction of false positive retrogenes in our analysis , we attempted to validate all retroCNVs with known insertion sites by PCR amplification followed by sequencing ., We confirmed 10 of 11 retroCNVs present in the reference genome ( 90 . 9% ) that we were able to assay , and 17 of 21 ( 80 . 5% ) retroCNVs absent from the reference genome ., In the case of retroCNVs absent from the reference genome our experimental design does not allow us to differentiate between false positives and retroCNVs we could not amplify due to experimental difficulties such as low primer specificity ( Materials and Methods ) , and most retroCNVs we could not amplify ( whether present or absent in the reference ) were flanked by repetitive elements ., It therefore seems plausible that some or all of the four retroCNVs absent from the reference genome that we could not confirm are actually true positives ., However , even if we conservatively assume that these four cases are false positives , our false positive rate across the set of 39 retroCNVs with known insertion loci is acceptably low ( 15 . 6%; validation results are listed in Table S2 and genomes used for validation are listed in Table S3 ) ., The remaining 52 retroCNVs may contain a higher fraction of false positives , and their relatively high fraction of singletons ( 67 . 3% ) is consistent with this ., However , we have previously shown that the exon-exon junction approach used to detect these retroCNVs is quite accurate 30; thus , many of these 52 retroCNVs are likely true events , and the large number of singletons could in part be explained by somatic mutations in the cell lines used to obtain DNA for the individuals in the 1000 Genomes Project , in addition to false positives ., In any case , the omission of these retroCNVs does not qualitatively affect any of the analyses described below ., We estimate that the approach using paired-end reads to discover retroCNVs ( whether present in or absent from the reference genome ) was able to detect at least 77 . 4% of singleton retroCNVs inserted in non-repetitive sequence in the 17 discovery genomes ., The false negative rate decreases dramatically for retroCNVs present more than once in the discovery set—we estimate that retroCNVs present in just two samples would be discovered ∼95% of the time ( Materials and Methods ) ., In addition , the exon-exon junction approach has previously been shown to be highly sensitive 30; this implies that our dataset contains the vast majority of retroCNVs present in the genomes we examined during the discovery phase of our study ., All retroCNVs included in our dataset , and their insertion coordinates when known , are listed in Table S2 ., The sets of genome sequences and retroCNVs included in each of our analyses are summarized in Table S4 ., In contrast to tandem duplications caused by replication slippage , or sometimes by non-allelic homologous recombination ( NAHR ) , retrotransposition results in a new gene duplicate located far from the parental copy ., Unlike our previous examination of gene retrotransposition in D . melanogaster 30 , in this study we were able to locate the insertion site of new retrocopies and therefore to examine precise patterns of gene movement caused by this type of duplication ., Although there is an excess of fixed retrogene movements onto and off of the human and mouse X chromosomes relative to expectations 29 , we do not see such a pattern in our set of retroCNVs ( Table 1 ) , suggesting differences in the contribution of adaptive evolution to polymorphic and fixed retrogenes ., As we have previously done in D . melanogaster , here we conducted a statistical test for differences in patterns of movement between retroCNVs and fixed functional retrogenes ., If gene movements onto and off of the X are neutral , then we expect the same proportion of such events among polymorphic retrocopies and fixed functional retrogenes; however , if movements involving the X chromosome are often adaptive , then we will observe a higher fraction of this class of movements among fixed retrogenes ., We do in fact find a significantly higher fraction of fixed functional retrogenes than retroCNVs moving to and from the X chromosome ( P\u200a=\u200a0 . 0067; Fishers exact test using fixed retrogene data from ref . 29 ) , lending further support to the hypothesis that natural selection is driving gene movement to and from mammalian X chromosomes 29 ., This result remains significant when we only examine retroCNVs discovered in females ( P\u200a=\u200a0 . 0079 ) , and is therefore not an artifact of reduced power to detect X-linked retroCNVs in males ., Because retroCNVs absent from the reference genome were discovered using a different ascertainment scheme than retroCNVs present in the reference genome , combining them in this analysis could impact our results ., However , this would only result in a deficit of retroCNVs moving to or from the X chromosome if such retroCNVs were more likely to be confined to lower allele frequencies by purifying selection than other retroCNVs , and there is no reason to expect such a difference in selective pressures ., Moreover , after imposing the same ascertainment scheme on both retroCNVs present in and absent from the reference genome ( Materials and Methods ) , we observe a similar but non-significant deficit of retroCNVs moving to or from the X ( none of the 9 retroCNVs in this set involve movements to or from the X; P\u200a=\u200a0 . 11 ) ., When we test separately for an excess of fixed functional retrogenes moving off of the X or moving onto the X , we do not see significance in either case ( P\u200a=\u200a0 . 150 for movements off of the X; Table S5; P\u200a=\u200a0 . 0650 for movements onto the X; Table S6 ) ., However , although we have lower statistical power in these comparisons , we do observe trends suggestive of natural selection ., Moreover , the excess of fixed functional retrogenes moving off of the X is significant when we compare retroCNVs to data from ref ., 35 ( P\u200a=\u200a0 . 0077; Table S5 ) ; when we examine all retroCNVs , including those with an unknown insertion site , we also see a significant excess of fixed retrogenes originating on the X chromosome when comparing our data to both ref ., 29 and ref ., 35 ( P\u200a=\u200a0 . 032 and P\u200a=\u200a3 . 6×10−4 respectively; Table S7 ) ., Combined with the observation that processed pseudogenes do not exhibit a bias of movement from the X 29 , our data strongly suggest that natural selection is responsible for the excess of functional retrogenes moving off of the X chromosome in mammals , and perhaps onto the X chromosome as well ., These observations could be the result of positive selection driving the fixation of new functional retrogenes moving to or from the X , selection to maintain such genes once they are established , or both of these mechanisms ., While it is widely believed that gene duplicates created by retrotransposition are almost always dead-on-arrival pseudogenes because they do not carry all regulatory elements from the parental copy with them , it has been shown that a retrocopy inserted into another gene will often exploit that genes regulatory machinery in order to be expressed 26 ., We therefore examined the insertion point of our retroCNVs to determine how many were inserted into existing genes ., We found that over one-half ( 20 of 39 ) of retroCNVs were inserted into genes , with all but one of these retroCNVs being inserted into an intron ( Table S2 ) ., This does not represent a significant deviation from what one would expect if retrocopy insertions were distributed uniformly across the genome , as introns make up roughly 40% of the human genome ( P\u200a=\u200a0 . 60; χ2 test ) ., Although there does not appear to be a strong bias in polymorphism data , we compared retroCNVs to the 7 , 831 retrocopies ( functional or otherwise ) identified in the reference genome ( Materials and Methods ) , nearly all of which are fixed , and found a deficit of fixed human retrocopies in introns compared to retroCNVs: 50 . 0% of retroCNVs versus 31 . 8% of fixed retrocopies are found in introns ( Table 2; P\u200a=\u200a0 . 022; Fishers exact test; P\u200a=\u200a0 . 012 using fixed retrocopies from ref . 26 with dS<0 . 1 when compared to their parent gene ) ., Again , similar to the reasoning laid out above , this implies that retrocopies inserted into introns are often deleterious , as was suggested by Vinckenbosch et al . 26 ., Indeed , the results in Table 2 suggest that roughly one-half of intronic retrocopy insertions are eliminated by purifying selection ., A similar deficit of fixed intronic retrocopies is observed when we impose the same ascertainment scheme on all retroCNVs , as described in Materials and Methods ( 62 . 5% of retroCNVs found in introns versus 31 . 8% of fixed retrocopies ) , although this comparison is no longer significant ( P\u200a=\u200a0 . 12 ) , perhaps in part due to diminished statistical power ., Because this is a comparison of patterns of retroCNVs that may not be functional to fixed retrocopies that are mostly pseudogenes , the simplest interpretation of this result is that the insertion of retrocopies into genes may often be deleterious even when the inserted retrocopy is non-functional ., Thus , intronic insertions may often be deleterious regardless of the content of the inserted sequence ., This interpretation is supported by the observation that tandem duplications occurring within introns are often subject to purifying selection in Drosophila 17 ., If the above interpretation is correct , then it could imply that roughly half of the genic retroCNVs we detect here are deleterious and would not be allowed by selection to reach fixation ., This interpretation is substantiated by the lower allele frequencies of intronic versus intergenic retroCNVs when examining only retroCNVs present in the reference genome ( avg . frequency in YRI is 0 . 46 for intronic and 0 . 72 for intergenic retroCNVs; P\u200a=\u200a0 . 75 ) or absent from the reference genome ( 0 . 11 for intronic versus 0 . 16 for intergenic; P\u200a=\u200a0 . 95 ) ., We performed this comparison separately for retroCNVs present and absent from the reference genome in order to control for ascertainment bias , as these retroCNVs had different ascertainment schemes ., While these differences are not significant , they are consistent with selection acting against intronic insertions , especially given evidence that non-retroCNV insertions within introns are often deleterious as discussed above ., Consistent with this interpretation , it has been noted that fixed retrocopy insertions are less likely to be intronic than expected if retrocopies are inserted with uniform probability across the genome 26 , although there is evidence of an insertion bias associated with chromatin accessibility in Drosophila 36 ., Overall , there is substantial evidence that insertions of retrocopies or other sequence into introns are often deleterious ., Since one would presume that retrocopies inserted into introns are also more likely to be expressed , our results suggest that retrotransposition could be an important source of new functional gene copies as well as potentially deleterious mutations ., An additional possible functional consequence of the insertion of retroCNVs into introns is the formation of sense-antisense pairs , as we previously suggested 37 ., Consistent with this possibility , we find that 10 of 20 retrocopies inserted into another gene are on that genes minus strand ( Table S2 ) ., We also find that one retroCNV , a copy of RPL3 , switches strands mid-sequence , most likely due to 5′ inversion during retrotransposition 38 ., Another interesting consequence of the insertion of a retrocopy into an intron of a host gene is the possibility of chimeric transcription of the host and the retrocopy ., Chimeric genes are likely an important source of new gene functions 39 , and the large fraction of retroCNVs inserted into introns suggests that retrotransposition could be an important source of these genes ., Indeed , there are several known cases of retrotransposition resulting in functional chimeric genes in humans 40 , 41 and Drosophila 6 , 42 , 43 , with some of these genes showing evidence for adaptive evolution 6 , 44 ., In order to search for evidence of chimeric transcripts among the 20 retroCNVs inserted within existing genes , we examined RNA-seq data from lymphoblast tissues from 60 HapMap individuals of European descent 45 ., We found that 20% ( 4 of 20 ) of these retroCNVs show evidence of chimeric expression ., The chimeric transcript CBX3-C15orf57 , where the CBX3 retroCNV is inserted in-between the second and third exons of C15orf57 , shows evidence of expression as a chimera in 20 individuals ., The chimeric combination SDHC-RPA1 forms a sense-antisense pair , with SDHC inserted in-between the fifth and sixth exon of RPA1; the chimeric transcript is expressed in 6 individuals ., UQCR10-C1orf194 , in which UQCR10 is inserted into the second exon of C1orf194 is expressed in a single individual ., An examination of the sequencing read confirming the validity of this retroCNV reveals that the UQCR10 portion of this transcript is not in proper reading frame ., The RPL18A-TXNRD1 combination , in which RPL18A is inserted in-between the third and fourth exons of TXNRD1 , was also found to be expressed in one individual ., We also found evidence of chimeric transcripts derived from SKA3-DDX10 in a breast cancer cell line and in a lymphoid cell line ( HCC1954 and HCC1954-BL from ref . 46 ) , both derived from an individual genotyped for SKA3 ., The SKA3 retroCNV is inserted in-between the tenth and eleventh exons of DDX10 , forming a sense-antisense pair ., Because three of these chimeric transcripts involve either a sense-antisense pair or the retroCNV apparently being inserted out of reading frame , they may be nonfunctional and perhaps deleterious ., Alternatively , it has been suggested that chimeric transcripts could result in novel protein coding regions even if they are not in sense-sense orientation or proper reading frame 25 ., In addition , we have only examined expression data for chimeric transcripts from lymphoblast cell lines for the majority of our retroCNVs , and two additional cell lines for a single retroCNV ( SKA3; Materials and Methods ) , and may therefore be underestimating the number of segregating chimeric genes caused by the incorporation of retroCNVs into existing genes ., While further work is required to determine the number of these new genes and their functional consequences , our results suggest that retrotransposition could be a source of evolutionary novelty creating not only new gene duplicates but new genes with potentially novel functions ., In order to examine the population dynamics of retroCNVs , we used both insertion presence/absence information at retroCNV insertions and evidence of retrotransposition from exon-exon junction-spanning reads to genotype 39 retroCNVs whose insertions we were able to locate ., After estimating allele frequencies for these retroCNVs in three human populations ( Materials and Methods ) , we noticed that several had very high derived-allele frequencies ( Figure 2; frequencies listed in Table S2 ) ., While this observation is consistent with positive selection driving retroCNVs to fixation , the fact that many of our retroCNVs were ascertained in a sample of 17 genomes ( AAC , SJS , and 15 individuals from the 1000 Genomes Project ) biases our frequency spectra towards higher frequency variants ., We therefore searched for more direct evidence of adaptive natural selection acting on individual retroCNVs ., Although previous genome-wide studies of copy-number variation have searched for evidence of natural selection sweeping duplications towards fixation 2 , 14 , these searches were conducted at regions containing the parental copy and not necessarily the daughter copy ., This was because location of the daughter locus was not known , and was simply assumed to be proximate to the parental locus ., These approaches would therefore fail to detect evidence of positive selection on dispersed duplications , a limitation that does not affect our analysis because we have identified the exact location of the new duplicates ., Conversely , if the insertion sites of duplicates are not known , many previous studies of ongoing selective sweeps in humans 47 , 48 may have detected the signature of positive selection on an inserted sequence that was not known to lie in the selected region ., In addition to examining the correct locus , testing for adaptive evolution requires accurate genotyping ., We therefore genotyped all 39 retroCNVs with known insertion sites as homozygous for retroCNV presence , heterozygous , or homozygous absent using our short-read sequences ., In order to assess our genotyping accuracy , we initially compared our genotyping results for the retroCNV of DHFR to those of Conrad et al . 2 , who were able to genotype this retroCNV as well ., We found that our genotypes agreed for 100% of individuals genotyped as homozygous for retroCNV presence by Conrad et al . , for 85% of individuals genotyped as heterozygous , and for 98% of individuals genotyped as homozygous absent ., Because Conrad et al . 2 may have committed genotyping errors as well , these percentages can be thought of as a lower bound on our genotyping accuracy , suggesting that our genotyping is highly accurate ., In order to gain additional confidence in our genotyping accuracy , we analyzed the genotypes of two available trios from the 1000 Genomes Project , finding that no analyzed retroCNVs violated Mendelian inheritance ( Table S8 ) , although these genomes had higher coverage than the rest of our data set ., In addition , we experimentally validated the genotypes of DHFR and GNG10 ( discussed below ) in 36 individuals ( Table S3 ) and found that our genotyping is also accurate in genomes with lower coverage , with 94 . 4% and 91 . 7% of genotyping calls confirmed for these two retroCNVs , respectively ., At these two retroCNVs we correctly genotyped 85 . 3% of heterozygous individuals and 100% of homozygotes , similar to our results in comparison to those of Conrad et al . 2 ., The action of positive selection on an allele results in a rapid increase in the frequency of the haplotype containing the selected allele in the population ., The swift nature of this rise in frequency results in a decrease in genetic diversity among chromosomes containing the selected allele compared to neutral expectations ., We therefore examined nucleotide diversity ( π ) in regions flanking retroCNV insertions , finding several retroCNVs with a marked reduction in diversity among haplotypes containing the retroCNV relative to the other haplotypes in the population ( Materials and Methods ) ., However , a deficit of diversity is expected among haplotypes sharing a derived allele regardless of its selective importance 49 ., With this in mind , we used coalescent simulations 50 to ask whether the ratio of π among haplotypes containing a retroCNV to π among haplotypes lacking it , which we refer to as πder/πanc , was lower than expected under neutrality ( Materials and Methods ) ., This is similar to the haplotype-based test first suggested by Hudson et al . 51 , the sole difference being that we contrast π between the derived and ancestral allelic classes , rather than the number of segregating sites ., For a polymorphism segregating in the absence of selection , we expect the observed ratio of πder/πanc to be typical when compared to those generated from the neutral coalescent for derived alleles of the same sample frequency ., For a polymorphism sweeping to fixation , on the other hand , relatively little diversity is expected among chromosomes containing the selected allele that is rapidly rising in frequency , and this allelic class would therefore exhibit a lower πder/πanc ratio than polymorphisms of the same frequency simulated under neutrality ., We were able to perform this test on 17 retroCNVs in the CEU subpopulation , 16 in YRI , and 13 in ASI ( Materials and Methods ) ., Two retrocopies are candidates for positive selection according to this test: the retrocopy of DHFR appears to be experiencing positive selection in individuals of European descent ( P\u200a=\u200a0 . 0083; Figure 3 ) , as does a retrocopy of GNG10 in both Europeans ( P\u200a=\u200a0 . 0094; Figure S1 ) and Africans ( P<1 . 1×10−4; Figure S2 ) ., If we correct for multiple testing by conservatively assuming that all 46 tests for selection that we conducted were independent—even though many tests were of the same retroCNVs but in different subpopulations—the false discovery rate ( FDR ) for the DHFR and GNG10 retroCNVs in Europeans is 0 . 14 , while the FDR for the GNG10 retroCNV is 0 . 0051 in Africans ., As stated above , a deficit of diversity is expected within haplotypes containing a new mutation under the neutral coalescent ., However , this deficit is less pronounced for polymorphisms with relatively high derived-allele frequencies such as the DHFR and GNG10 retroCNVs because the amount of diversity associated with any allele is proportional to its frequency ., The reductions in heterozygosity shown in Figure 3 , Figure S1 , and Figure S2 may therefore be suggestive of positive selection; this interpretation is supported by the results of our coalescent-based test that takes allele frequency into account ., The DHFR retroCNV , previously discovered by Anagnou et al . 52 , is inserted into the fourth intron of PSM8 , forming a sense-antisense pair ., The ORF of this retrocopy perfectly matches that of the parental copy of DHFR in the reference genome 53 ., DHFR codes for dihydrofolate reductase , deficiency of which causes megaloblastic anemia and neurological disease 54 , and is required for nucleotide synthesis 55 ., DHFR has an important role in cell growth , and its inhibition has been used in antibacterial 56 and antitumor drugs 57 ., This retrocopy also exhibited a similar reduction in nucleotide diversity in the Asian subpopulation , although this pattern was not significant by our test ( P\u200a=\u200a0 . 099; Figure S3 ) ., GNG10 , which has been associated with melanoma 58 , has a retrocopy that forms a sense-sense pair with SBF2 , which has been implicated in Charcot-Marie-Tooth disease 59 ., To gain further confidence in these results , we compared the πder/πanc ratios observed for these candidates to those calculated from random regions flanking SNPs with similar derived allele frequencies , finding that relatively few SNPs in the human genome exhibited lower πder/πanc ratios than these retroCNVs , even though some of these loci are likely themselves under positive selection ., For example , just 2 . 5% and 5 . 5% of loci in the genome exhibited lower ratios of πder/πanc than the DHFR retroCNV in Europeans and the GNG10 retroCNV in Africans , respectively ( Materials and Methods ) ., Although we experimentally determined that our genotype calls at these two retroCNVs were quite accurate , genotyping error could still affect the analyses described above ., We therefore conducted a further test based on integrated haplotype scores ( iHS ) , a statistic designed to detect extended haplotypes characteristic of ongoing sweeps , around these two retroCNV insertions 48 ., Importantly , this test is not dependent on our genotype assignments ., We find that only 1 . 2% of random genomic regions exhibit stronger biases toward extreme iHS values than the region containing the GNG10 retroCNV in Africans , the strongest candidate identified by our coalescent-based test ( Materials and Methods ) ., Additionally , only 5 . 7% of random genomic regions exhibit more extreme iHS values than the DHFR retrocopy in Asians , where we observed a suggestive but non-significant signal of selection in our coalescent-based test ., We cannot know with certainty that natural selection is responsible for the patterns of diversity around these two retroCNVs , or that the retroCNVs themselves rather than polymorphisms in linkage disequilibrium with them are the targets of any such selection ., Nonetheless , our findings that the haplotypes containing these retroCNVs exhibit reduced diversity and reside within regions identified by an extended haplotype test suggest that these retroCNVs should be considered candidates for adaptive natural selection ., This evidence that multiple retroCNVs currently segregating in human subpopulations could potentially confer an increase in fit | Introduction, Results/Discussion, Materials and Methods | The era of whole-genome sequencing has revealed that gene copy-number changes caused by duplication and deletion events have important evolutionary , functional , and phenotypic consequences ., Recent studies have therefore focused on revealing the extent of variation in copy-number within natural populations of humans and other species ., These studies have found a large number of copy-number variants ( CNVs ) in humans , many of which have been shown to have clinical or evolutionary importance ., For the most part , these studies have failed to detect an important class of gene copy-number polymorphism: gene duplications caused by retrotransposition , which result in a new intron-less copy of the parental gene being inserted into a random location in the genome ., Here we describe a computational approach leveraging next-generation sequence data to detect gene copy-number variants caused by retrotransposition ( retroCNVs ) , and we report the first genome-wide analysis of these variants in humans ., We find that retroCNVs account for a substantial fraction of gene copy-number differences between any two individuals ., Moreover , we show that these variants may often result in expressed chimeric transcripts , underscoring their potential for the evolution of novel gene functions ., By locating the insertion sites of these duplicates , we are able to show that retroCNVs have had an important role in recent human adaptation , and we also uncover evidence that positive selection may currently be driving multiple retroCNVs toward fixation ., Together these findings imply that retroCNVs are an especially important class of polymorphism , and that future studies of copy-number variation should search for these variants in order to illuminate their potential evolutionary and functional relevance . | Recent studies of human genetic variation have revealed that , in addition to differing at single nucleotide polymorphisms , individuals differ in copy-number at many regions of the genome ., These copy-number variants ( CNVs ) are caused by duplication or deletion events and often affect functional sequences such as genes ., Efforts to reveal the functional impact of CNVs have identified many variants increasing the risk of various disorders , and some that are adaptive ., However , these studies mostly fail to detect gene duplications caused by retrotransposition , in which an mRNA transcript is reverse-transcribed and reinserted into the genome , yielding a new intron-less gene copy ., Here we describe a method leveraging next-generation sequence data to accurately detect gene copy-number variants caused by retrotransposition , or retroCNVs , and apply this method to hundreds of whole-genome sequences from three different human subpopulations ., We find that these variants account for a substantial number of gene copy-number differences between individuals , and that gene retrotransposition may often result in both deleterious and beneficial mutations ., Indeed , we present evidence that two of these new gene duplications may be adaptive ., These results imply that retroCNVs are an especially important class of CNV and should be included in future studies of human copy-number variation . | genetic polymorphism, natural selection, population genetics, biology, evolutionary biology, genomic evolution | null |
journal.ppat.1007852 | 2,019 | The landscape of transcription initiation across latent and lytic KSHV genomes | Regulated gene expression is an essential process for all eukaryotic cells as well as the pathogens that infect them ., A fundamental step in the regulation of gene expression is the initiation of RNA synthesis ., RNA Polymerase II ( RNAP II ) transcription begins with the binding of gene-specific regulatory factors within the core promoter located near the transcription start site ( TSS ) ., Sequence elements found within core promoters include the TATA element , BRE ( Transcription factor II B recognition element ) , Initiator ( Inr ) , and downstream promoter element ( DPE ) 1 ., These elements serve as binding sites for subunits of the transcription machinery and their nucleotide composition can impact the efficiency of transcription initiation ., Models of transcription initiation often depict RNA synthesis initiating from a single nucleotide ., However , the application of high-throughput 5’-end sequencing technologies , such as Cap Analysis of Gene Expression ( CAGE ) , have demonstrated that most RNAP II promoters have an array of closely spaced TSSs instead of the expected single TSS 2–7 ., Thus , promoters can be more accurately described as a distribution of initiation events ( transcription start site clusters , TSCs ) on a stretch of given nucleotides 5 ., Moreover , the nucleotide sequences that comprise the core promoter can influence the distribution of TSSs as well as the expression profile for a given gene ., This has led to the classification of animal promoters into two major groups , narrow and broad 8 , 9 ., For example , more broadly distributed TSSs are correlated with CpG islands and ubiquitously expressed genes , whereas promoters harboring TATA and TATA-like sequences exhibit a narrow distribution of initiation sites and frequently drive expression of tissue-specific genes 4 , 10–15 ., Viruses employ a parasitic lifestyle thus they exploit many , and in some cases , all , of the host gene expression machineries ., The co-option of cellular processes allows viruses to dramatically downsize their genomes , however , it necessitates the incorporation of host cell-specific gene regulatory mechanisms and features that enable the use of such mechanisms ., Additionally , the compact nature of viral genomes requires the encoding of viral proteins and RNAs within a relatively small amount of genomic space ., Kaposi sarcoma-associated herpesvirus ( KSHV ) is a human oncogenic virus and a member of the gammaherpesvirus subfamily 16 , 17 ., While KSHV infection is generally unproblematic in the healthy adult population , in the context of immunosuppression such as iatrogenic immunosuppression or advanced HIV-infection resulting in AIDS , the virus is associated with several malignancies 18 , 19 ., In fact , it is estimated that approximately 1% of all human tumors are associated with KSHV infection and the World Health Organization ( WHO ) has classified the virus as a class I carcinogen 18 , 20 ., KSHV is the etiological agent of all forms of Kaposi sarcoma 21–23 , a complex and highly vascularized solid tumor of endothelial origin 24 , as well as the B cell lymphoproliferative disorders multicentric Castleman’s disease ( MCD ) and primary effusion lymphoma ( PEL ) 25 , 26 ., As with other herpesviruses , KSHV displays two distinct phases of its viral lifecycle , latency and lytic infection ., During latent infection KSHV exists in a dormant state in which viral gene expression is dramatically restricted such that few viral antigens and no viral particles are produced ., In contrast , during the lytic cycle , which is driven by the viral-encoded transcription factor replication and transcription activator ( RTA ) 27 , 28 , the full repertoire of viral genes is expressed and new infectious virions are produced ., KSHV has a large DNA genome and similar to other DNA viruses has a complex gene organization including overlapping genes and polycistronic mRNAs ., The current annotation of the KSHV genome has over 90 genes that encode for viral proteins and both long and small noncoding RNAs 29 ., Despite identification of KSHV over two decades ago we still lack fundamental knowledge regarding how viral gene expression is controlled at the level of transcription initiation ., Additionally , although several studies have employed RNA-seq to define the transcriptional output of the KSHV genome as well as the temporal kinetics of viral gene expression , it is often difficult to accurately identify TSSs from traditional RNA-seq data alone ., Identification of TSSs is particularly important in complex transcriptomes , such as that of KSHV , where there is extensive overlapping transcription of viral genes and the appearance of polycistronic RNAs ., To fill this gap in fundamental knowledge we have mapped all transcription initiation events on the viral genome using RNA annotation and mapping of promoters for analysis of gene expression ( RAMPAGE ) 30 ., Leveraging RAMPAGE , we now provide transcriptome-wide nucleotide resolution TSC annotations for viral genes in two widely used models of KSHV infection , namely the iSLK . 219 31 , 32 and TREx-BCBL1-RTA 33 systems ., We confirm 48 of 50 previously known TSSs as well as identify over 100 novel TSCs in each cell line ., Moreover , these analyses identify cell-type specific differences in TSC positions as well as promoter strength , and define sequence elements comprising viral core promoters ., Collectively , we greatly expand the transcriptional landscape of the KSHV genome and identify transcriptional control mechanisms at play during KSHV lytic reactivation ., Precise promoter annotation is required for understanding the mechanistic basis of condition- and tissue-specific gene regulation ., While the transcriptional landscape of KSHV has been characterized by RNA-seq there is still a gap in fundamental knowledge regarding the location of viral TSSs , and thus it has not been feasible to comprehensively define viral promoters and the sequence motifs embedded within them ., Moreover , whether there are condition ( i . e . latent vs . lytic ) or cell type specific differences in TSS usage is not known ., To fill this gap in knowledge we took advantage of two well established models of KSHV infection , namely , iSLK . 219 and TREx-BCBL1-RTA cells ., While iSLK . 219 cells are of clear cell renal cell carcinoma origin and thus of lesser biological relevance 34 , the iSLK system has proven invaluable to KSHV research as it is routinely used in studies addressing viral gene function through bacterial artificial chromosome ( BAC ) mutagenesis ., iSLK . 219 cells are infected with a recombinant virus , KSHV . 219 , in which GFP , expressed from the elongation factor 1-α promoter , and RFP , expressed from the viral lytic gene PAN promoter , were inserted into the viral genome 31 , 32 ., In contrast , TREx-BCBL1-RTA cells are a genetically engineered derivative of KSHV-infected B cells isolated from a patient with PEL 33 ., A key feature in both systems is the integration of a doxycycline ( Dox ) -inducible version of the major viral transcription activator RTA ., Thus , upon introduction of Dox into the cell culture media KSHV enters the lytic cycle ., Leveraging iSLK . 219 and TREx-BCBL1-RTA cells we mapped TSSs transcriptome-wide at nucleotide resolution across a 96 h Dox-induced reactivation time course using RAMPAGE ., RAMPAGE , which combines the two orthogonal 5’-selection approaches of template-switching and cap-trapping , enables TSS identification with high specificity through the high-throughput sequencing of 5′‐complete complementary DNAs ( Fig 1A ) ., Additionally , an important distinction from CAGE is that RAMPAGE allows for pair-ended sequencing and thus yields extensive transcript connectivity information allowing the annotation of genes with their TSS 30 ., Total RNA was isolated from cells in a latent state , or at 12 h , 24 h , 48 h , 72 h , and 96 h post-lytic reactivation and RAMPAGE libraries were prepared ., These time points were chosen as we observed >70% of cells in the lytic stage by 72 h post-dox induction as well as a significant increase in virion associated DNA in the culture supernatant ( Fig 1B and 1C ) ., The resultant RAMPAGE sequencing reads were aligned to the human ( GRCh38 ) and KSHV ( GQ994935 . 1 ) genomes and TSCs were identified by Paraclu 5 and quantified in tags per million reads mapped ( TPM ) ., To assess the specificity of our RAMPAGE data for 5′ ends we examined the distribution of the raw TSS 5’ signal over cellular annotated transcripts ., Importantly , the metaprofile of signal density clearly demonstrates enrichment near the 5’ end and confirms the specificity of our data ( Fig 1D and 1E ) ., Within 12 h of Dox addition to the culture media there was an increase in TSSs that continued through the time course ( Figs 2 and 3 ) ., In total , we identified 164 and 292 TSCs in iSLK . 219 and TREx-BCBL1-RTA cells , respectively ( Figs 2 and 3 , S1 Table ) ., TSSs for 50 viral transcripts have been previous mapped or annotated 27 , 29 , 35–53 ., Within our RAMPAGE data we observed prominent TSSs at the exact position or within a few nucleotides of 48 of the 50 previously identified TSSs , with the exceptions being the TSSs driving ORF17 and ORF42 expression ( S1 Fig ) ., We surmise that the slight variations in TSS architecture observed between previous studies and our data is related to the use of histone deacetylase ( HDAC ) inhibitors to promote KSHV reactivation in the previous reports ., An advantage of RAMPAGE over other 5’-end sequencing technologies is that sequencing is performed in a pair-ended format and thus more information regarding transcript content is present in the data ., Leveraging the pair-end information of RAMPAGE we were able to assign many TSSs to ORFs that were previously lacking TSS information ( S1 Table ) ., For example , ORF7 , ORF10 and ORFK7 were assigned TSSs ., This high-resolution map of viral transcription initiation refines the current annotation of KSHV transcriptome , and reveals a much more complex transcriptional landscape than previously appreciated ., Transcriptional regulation is highly complex and chromatin structure and modifications directly contribute to transcription initiation ., The open chromatin landscape of the KSHV genome in PEL has been investigated by formaldehyde-assisted isolation of regulatory elements sequencing ( FAIRE-seq ) 54 ., Previous FAIRE-seq analyses found that regions of open chromatin were not restricted to transcriptionally active loci , but that transcriptionally inactive loci were also nucleosome depleted ., Moreover , the transcriptional repressor CTCF occupied the majority of nucleosome depleted regions regardless of transcriptional activity 54 ., Having mapped the transcription initiation landscape of PEL cells , we intersected FAIRE-seq and CTCF chromatin immunoprecipitation sequencing ( ChIP-seq ) data sets with the RAMPAGE defined TSCs from TREx-BCBL1-RTA cells ( Fig 4A ) ., Consistent with previous work , we observed that TSCs of both latent and lytic reactivated TREx-BCBL1-RTA cells were highly correlated ( p < 0 . 00001 ) with regions of open chromatin ., Furthermore , we similarly observed enrichment of CTCF at both latent and lytic TSCs ( p < 0 . 00001 ) ., In fact , more than 50% of KSHV TSCs are located within 250 bp of a CTCF or FAIRE-seq peak in TREx-BCBL1-RTA cells ( Fig 4B and 4C ) ., TREx-BCBL1 cells have been the subject of extensive investigation regarding the presence and location of various chromatin modifications , including the activating mark H3K4me3 and the repressive mark H3K27me3 , as well as the KSHV latent protein LANA and cellular RNAP II 55 ., By intersection analysis we visualized the previous ChIP-seq data with our comprehensive TSC map ( S2 Fig ) ., Interestingly , although we analyzed H3K4me3 ChIP-seq data from latent TREx-BCBL1-RTA cells , TSCs belonging to both latent and lytic genes are located in close proximity to H3K4me3 signal ., However , given that a small percentage of PEL cells are continually undergoing spontaneous reactivation it is possible that the H3K4me3 signal over the lytic genes is derived from this population of cells ., We next sought to determine whether KSHV promoter usage is regulated in a cell-type specific manner ., Intersection analysis of TSCs identified in iSLK . 219 and TREx-BCBL1-RTA cells revealed that viral gene expression is highly cell-type specific ., We identified 129 TSCs were present in both cell lines , we identified 35 and 163 TSCs present exclusively in either iSLK . 219 or TREx-BCBL1-RTA cells , respectively ( Fig 5A ) ., Importantly , the difference in numbers of TSCs identified is likely not due to sequencing depth as in both iSLK . 219 and TREx-BCBL1-RTA a similar number of reads were mapped to the viral genome ( S3 Fig ) ., Interestingly , among the 129 TSCs present in both cell-types we identified three novel TSCs that have a corresponding RTA ChIP-seq peak immediately preceding the TSC , suggesting the transcripts are RTA-dependent ( Fig 5B ) ., Importantly , the RTA ChIP-seq peaks identified were present in two separate data sets analyzed ., In support of a role for RTA in the expression of these TSCs , RTA was capable of activating the expression of a luciferase reporter harboring these genomic fragments as promoters ( S4 Fig ) ., While this data does not directly demonstrate the TSCs are RTA-dependent , these data do demonstrate that these genomic fragments harbor the potential to serve as promoters that drive productive transcription initiation ., Inspection of our data uncovered that many immediate early genes , such as K6 , tend to be associated with multiple TSCs , all of which are shared between both iSLK . 219 and TREx-BCBL1-RTA cells ., Previous studies have demonstrated that immediate early genes can be expressed at a low basal level during latency , and are subsequently robustly induced upon lytic reactivation ., Given that both multiple shared TSCs are present in both cell-types we hypothesized that basal and induced gene expression may be controlled by different promoters ., Indeed , inspection of TSCs that drive expression of K6 uncovered multiple promoters , with a prominent switch in TSC usage upon lytic reactivation ( Fig 5C ) ., As shown in Fig 5C , K6 is expressed from three major promoters ( P1 , P2 and P3 ) ., While P3 was preferentially used during latency , usage of P1 and P2 increased substantially upon lytic induction ., The conservation of regulated TSC selection for K6 between cell-types suggests this may be an underlying mechanism controlling the proper regulation of K6 expression ., As noted above we also identified cell-type specific TSCs ., For example , in TREx-BCBL1-RTA cells we identified 11 ORFs expressed by TSCs only present within TREx-BCBL1-RTA cells ., In contrast , we identified a distinct set of 11 ORFs that were expressed from TSCs only identified in iSLK . 219 cells ., For instance , we identified a TSC for LANA2 ( K10 . 5 ) in TREx-BCBL1 cells while it was absent from iSLK . 219 cells ( S5 Fig ) ., This is consistent with a previous report demonstrating LANA2 expression is restricted to PEL and MCD cells 56 ., Moreover , in TREx-BCBL1-RTA cells we identified a prominent TSC initiating synthesis of the ORF58 mRNA ( Fig 5D ) , however , in iSLK . 219 cells a homologous TSC is not present and ORF58 would be need to be expressed from a bicistronic mRNA ORF58/59 ( Fig 5D ) ., However , it is equally possible that ORF58 is not expressed in iSLK . 219 cells , and we are unaware of any studies detecting ORF58 protein expression in iSLK . 219 cells ., In highly complex genomes with extensive genic overlap , 5’-end RNA sequencing is more accurate in measuring transcript expression than traditional RNA-seq as it avoids ambiguous read assignment and transcript length normalization ., We thus quantified viral gene expression in both cell lines based on the intensity of the 5’ RAMPAGE signal ( Fig 6A ) ., Moreover , since we did not identify prominent TSCs for all ORFs we also quantified viral gene expression by mapping all of the reads directly to the ORFs ( S6 Fig ) ., Latency in iSLK . 219 cells is extremely tight , and only RNAs belonging to known latent genes were expressed ., In contrast , in latent TREx-BCBL1-RTA cells we observed a much broader transcriptional profile , with the promoters of several lytic genes , including ORF11 , ORF50 , ORFK8 , and ORF57 , producing mRNAs ., The broader expression profile in TREx-BCBL1-RTA cells is consistent with low level spontaneous reactivation in PEL cells ., Unsupervised hierarchical clustering analysis further demonstrated that viral reactivation in iSLK . 219 and TREx-BCBL1 cells follows different kinetic programs ( Fig 6B ) ., While the activity of most promoters in TREx-BCBL1-RTA cells gradually increases over the 96 h time course , expression from the majority of promoters in iSLK . 219 cells peaked between 48 h-72 h ( Fig 6A and 6B ) ., Consistent with this analysis , in iSLK . 219 cells the number of RAMPAGE reads mapping to the viral genome peak at 48 h , while in TREx-BCBL1-RTA cells this occurs at 96 h post-reactivation ( S3 Fig ) ., Given that unsupervised hierarchical clustering demonstrated distinct kinetic programs we next sought to investigate whether the individual programs were associated with unique core promoter compositions ., It has previously been suggested that TSCs within a 50 bp window are under control of the same transcription initiation complex 57 , thus we merged TSCs within a 50 bp window into a single cluster and performed K-means analysis on the resulting TSCs ., K-means analysis separated the iSLK . 219 and TREx-BCBL1-RTA expression profile in to three and four distinct classes , respectively ( Fig 7A and 7B ) ., The clusters defined revealed sets of genes with matched peak expression across the time points ( Fig 7C and Fig 7D ) ., The identification of variable cluster numbers between iSLK . 219 and TREx-BCBL1-RTA transcription programs supports the notion of cell-type specific kinetic programs ., To identify sequence elements associated with the distinct clusters we extracted nucleotide sequences 50 bp upstream and downstream of the maximally expressed nucleotide within each cluster ( MaxTSN ) and searched for position-specific ultra-short motifs by kplogo analysis ., All clusters , with the exception of TREx-BCBL1-RTA cluster 2 , display a preference for a Py-Pu dinucleotide at the −1/+1 position and a similar preference has been observed at both mouse and human TSCs 4 ., TREx-BCBL1-RTA cluster 2 is also unique in that a GG motif is present at the +7/8 position ., Interestingly , when all cellular TSCs are ranked by expression the presence of a downstream GG motif is associated with higher expression ( S7 Fig ) ., Consistent with this analysis , cluster 2 displays the highest mean expression value across all time points ., While this motif has been previously noted in CAGE data its association with more robust expression was not reported ., The molecular basis for this observation is not known although its distance from the MaxTSN is approximately one helical turn of DNA and thus may be phased accordingly with the MaxTSN ., iSLK . 219 clusters 1 and 3 , which exhibit maximum expression 48 h and 96 h post-induction , respectively , and TREx-BCBL1-RTA cluster 1 , which displays maximum expression at 96 h post-induction , all harbor an AT rich region 25–31 upstream of the MaxTSN ( Fig 7E and 7F ) ., Moreover , motif analysis using HOMER identified an enrichment of TATA binding protein ( TBP ) motif within these clusters ., However , we are cautious to interpret this as indicating that TBP is involved in the regulation of genes displaying maximum induction at later time points as viral late genes contain a variant TATA motif , TATT , that can also be defined by HOMER as a TBP motif ., Thus , it is more likely that the enrichment of TBP motifs within these clusters reflects a combination of TSCs that are driven by TBP and the viral late gene initiation complex ., While neither an AT rich region or TBP motif was found in the other clusters , an initiator ( Inr ) motif ( CA+1 ) is found enriched at TSCs within TREx-BCBL1-RTA cluster four ., An Inr motif is similar to a TATA motif in that it can directly recruit the general transcription factor TFIID ., However , the Inr recruits TFIID through interactions with TAF1 and TAF2 rather than TBP ., The enrichment of Inr motifs in cellular genes that lack TATA motifs has been previously observed and it has been speculated that promoters with an Inr are less dependent on a TATA box and vice versa 58 ., Thus , our observation mirrors what has been reported in humans and suggests preferential use of TATA or Inr motifs on viral promoters ., Thus , we hypothesize that KSHV uses a combination of sequence elements to recruit TFIID and promote viral gene expression ., We also determined whether the presence of Inr motif was associated with a higher maximum of gene expression ., When ranking TSC expression an expanded consensus Inr motif ( BBCA+1BW ) can be found dispersed among the moderately and highly expressed cellular TSCs within iSLK . 219 cells ( S7 Fig ) ., In contrast , the expanded Inr motif is more commonly identified within higher expressed cellular TSCs in TREx-BCBL1-RTA cells ., This suggests that distinct transcriptional control mechanisms and/or preferences may be at play within these two commonly used models of KSHV infection ., With regard to viral TSCs , no clear enrichment of the Inr motif is found among genes with a distinct expression profile ( S7 Fig ) ., We also investigated the relationship between the presence of a TATA motif with expression strength ., Analysis of the TATA motif density plot clearly shows enrichment of this motif approximately 30 bp upstream of the MaxTSN in both iSLK . 219 and TREx-BCBL1-RTA cells ( Fig 8A ) ., Interestingly , similar to what was observed for the Inr motif , a TATA motif can be found dispersed among the moderately and highly expressed cellular TSCs within iSLK . 219 cells , while it is more prevalent among the highest expressed genes in TREx-BCBL1-RTA cells ., With regard to viral TSCs , a TATA motif is not associated with expression ., The TATT motif is a noncanonical TBP-like motif that enables late gene expression via a specialized viral transcription complex 59 , 60 ., Interestingly , while we observed clear enrichment of this motif approximately 30 bp upstream of the viral MaxTSN in motif density plots , there also appears to be a preference for it upstream of the cellular MaxTSN ( Fig 8B ) ., To further investigate this , we first quantified the presence of TATA and TATT motifs within cellular and viral promoters ( Fig 8C ) ., Consistent with previous reports we observed TATA motifs within only a small percentage ( 2~3% ) of cellular promoters ., Moreover , a TATT motif was also observed in a similar portion of cellular promoters ., In contrast , we observed a higher percentage of viral promoters harboring TBP-like motifs ., Specifically , we observed a TATA motif within 16% of viral promoters and a TATT motif within 7% ., We next calculated the relative frequency of TATA and TATT motifs 50 bp upstream and downstream of cellular and viral MaxTSNs to search for positional enrichment ( Fig 8D ) ., Within viral promoters we identified a clear enrichment for these motifs 40–30 bp upstream of the MaxTSN ., Interestingly , a second minor enrichment of the TATT motif can be found 25–20 bp upstream of the MaxTSN ., The function significance of this second cluster is unclear ., With regard to cellular TSSs , the TATA motif is clearly enriched 35–30 bp upstream of the MaxTSN ., Unexpectedly , we also observe a peak of TATT motifs at a similar position within cellular promoters ., Moreover , there is a clear depletion of this motif immediately upstream of the cellular MaxTSN ., We anticipate this suggests that TATT motifs are engaged by cellular TBP for productive transcription and hypothesize that the viral transcription initiation complex may be similarly capable of regulating a subset of cellular genes due to the present of the its preferred binding motif , TATT ., The presence of TATA- and Inr-like motifs is primarily associated with tissue-specific or terminally differentiated cell-specific gene expression ., Moreover , the presence of these motifs tends to focus transcription initiation events into more well defined clusters ., Thus , we compared the breadth of cellular and viral TSCs ., The width distribution of viral TSCs is similar between iSLK . 219 and TREx-BCBL1-RTA cells ( Fig 8E ) ., When compared with the host , we find that the spatial distribution of viral TSCs in both iSLK . 219 and TREx-BCBL1-RTA cells is more confined than host TSCs , indicating viral gene expression is under a higher degree of spatial control ( Fig 8F ) ., Comprehensive genome annotations are critical for understanding the biology of all organisms , as well as viruses ., Moreover , precise annotations of key gene regulatory features such as TSSs and core promoters , provides the opportunity to define mechanistic features of gene expression regulation ., Our transcriptome-wide nucleotide resolution mapping of KSHV TSSs has greatly expanded the transcriptional landscape of KSHV and has identified core promoter sequences associated with TSC architecture and regulation ., Moreover , the identification of over 100 novel TSCs in both iSLK . 219 and TREx-BCBL1-RTA cells highlights the breadth of the KSHV transcriptome that remains uncharacterized ., Future studies must be directed towards understanding the functional significance of the newly defined viral RNAs as they may function as noncoding RNAs or encode proteins that impact the KSHV lifecycle and thus KSHV-associated disease progression ., The high-resolution map of TSSs described here more than triples the number of known transcripts that are expressed from the KSHV genome ., However , without full-length RNA-sequencing data it is difficult to fully assess whether the novel TSSs contribute to the expression of yet to be identified RNAs , or whether these extend the 5’-UTRs of known mRNAs ., Though analyses here indicate that both are true as we observe novel splice junctions associated with newly discovered TSSs , as well as novel TSSs in which all mate pair reads are associated with a known ORF ., Along this line , over 300 previously unknown transcripts were recently identified within the EBV transcriptome through the integration of multiple RNA-sequencing approaches , including CAGE-seq and full-length RNA sequencing 61 ., While determining the function of novel transcripts will likely require a detailed description of transcript structure coupled with biochemical assays , we surmise that extension of 5’-UTR sequences impacts gene expression regulation ., Indeed , alternative TSS selection has been shown to lead to large differences in translation efficiency of RNA in both yeast and mouse embryonic fibroblasts 62 , 63 ., Moreover , RNA structures within 5’-UTRs can influence RNA half-live 64 , 65 ., Thus , in addition to extending the number of RNAs transcribed off the KSHV genome this study also uncovers additional points of entry for post-transcriptional control processes ., Furthermore , the use of multiple TSSs to expand the landscape of viral 5’-UTRs appears to be common among viruses as recent studies leveraging CAGE-seq and precision nuclear run-on sequencing ( PRO-Seq ) have observed multiple TSSs for many EBV and cytomegalovirus ( CMV ) transcripts 61 , 66 , 67 ., While transcription initiation is often viewed as originating from a single nucleotide , this is an antiquated model ., Transcriptome-wide 5’-sequencing methodologies , such as CAGE-seq , PRO-seq , and RAMPAGE , have demonstrated that transcription initiation occurs over a range of nucleotides within promoters 2 , 3 , 30 ., Additionally , low level transcription initiation is also frequently observed within the coding regions of human genes ., Given that this has been observed with multiple technologies it is unlikely to be an experimental artifact ., Indeed , RAMPAGE analysis on KSHV infected cells reveals that KSHV transcription initiation is similarly structured , with multiple TSSs driving expression of individual genes as well as transcription initiation within protein coding genes ., The use of multiple TSSs for an individual protein coding gene enables additional mechanisms of post-transcriptional control of gene expression mediated through the various 5’-UTR sequences ., The extent to which KSHV gene expression is regulated via 5’-UTR based mechanisms is unclear , however , our study highlights the need for consideration of these mechanisms when investigating the regulation of viral gene expression ., Using our experimentally defined TSSs we identified unknown similarities and differences in TSC usage between iSLK . 219 and TREx-BCBL1-RTA cells ., For example , leveraging previously published RTA ChIP-seq data sets we identified three novel TSCs present in both cell types that harbor prominent RTA binding peaks immediately upstream suggesting their expression is regulated by RTA ., Moreover , when cloned upstream of a luciferase reporter these genomic fragments confer RTA-inducible luciferase expression ., As we noted earlier these data do not conclude that the three novel TSCs are RTA-dependent ., However , they do demonstrate the genomic fragments have the potential to drive productive RTA-dependent transcription initiation events ., We also identified clear examples where a TSC was specifically used in only one cell-type ., For example , in TREx-BCBL1-RTA cells a TSC for ORF58 is identified while in iSLK . 219 cells this TSC is absent ., While it is possible that that ORF58 is translated from a bicistronic ORF58/59 transcript in iSLK . 219 cells , it is equally plausible to hypothesize that ORF58 protein is not expressed there ., Core promoter nucleotide composition is known to influence various aspects of transcription , including the mechanisms of transcription factor recruitment , expression level , and breadth of transcription initiation sites ., Leveraging our RAMPAGE mapped TSSs we have defined the nucleotide composition of cellular and viral promoters and defined motifs associated with transcription breadth and strength ( Fig 8A , 8B and 8F and S7 Fig ) ., We find that viral promoters more heavily rely on TATA-like motifs than human promoters , and K-means clustering identified a distinct set of viral genes that harbor an Inr motif ., TATA-like and Inr motifs are associated with more focused transcription initiation profiles and indeed we discovered that viral promoters drive more focused transcription initiation than cellular promoters ., We hypothesize that the high gene density of the KSHV genome , and likely other viruses , necessitates a tighter control on the spatial distribution of transcription initiation events ., During KSHV infection the TATT motif is recognized by the viral TBP mimic ORF24 to facilitate late gene expression 59 ., Interestingly , quantification of the TATT motif frequency within cellular promoters identifies a minor enrichment of this sequence approximately 35–30 bp upstream of cellular promoters ., TATT can serve to recruit TBP ., For example , the mouse a4 IFN promoter contains a well-defined TATT motif 68 ., Whether cellular genes that contain this motif are bound by the viral initiation complex is not known although it would be very interesting to investigate this ., While it would likely be unfavorable for the virus to activate the expression of an interferon via the viral initiation complex , we hypothesize that cellular promoters that harbor TATT motifs contain sequences that recruit additional factors that are required for their expression ., Thus , in this scenario , the binding of the viral late gene complex would not necessarily be activating , but rather could sterically inhibit the recruit of the additional factors and inhibit gene expression ., Our study prov | Introduction, Results, Discussion, Materials and methods | Precise promoter annotation is required for understanding the mechanistic basis of transcription initiation ., In the context of complex genomes , such as herpesviruses where there is extensive genic overlap , identification of transcription start sites ( TSSs ) is particularly problematic and cannot be comprehensively accessed by standard RNA sequencing approaches ., Kaposis sarcoma-associated herpesvirus ( KSHV ) is an oncogenic gammaherpesvirus and the etiological agent of Kaposi’s sarcoma and the B cell lymphoma primary effusion lymphoma ( PEL ) ., Here , we leverage RNA annotation and mapping of promoters for analysis of gene expression ( RAMPAGE ) and define KSHV TSSs transcriptome-wide and at nucleotide resolution in two widely used models of KSHV infection , namely iSLK . 219 cells and the PEL cell line TREx-BCBL1-RTA ., By mapping TSSs over a 96 h time course of reactivation we confirm 48 of 50 previously identified TSSs ., Moreover , we identify over 100 novel transcription start site clusters ( TSCs ) in each cell line ., Our analyses identified cell-type specific differences in TSC positions as well as promoter strength , and defined motifs within viral core promoters ., Collectively , by defining TSSs at high resolution we have greatly expanded the transcriptional landscape of the KSHV genome and identified transcriptional control mechanisms at play during KSHV lytic reactivation . | Kaposis sarcoma-associated herpesvirus ( KSHV ) is an oncogenic gammaherpesvirus and the etiological agent of Kaposi’s sarcoma and the B cell lymphoma primary effusion lymphoma ( PEL ) ., Despite identification of the virus over 20 years ago there is still an incomplete understanding of how many RNAs are transcribed from the viral genome and the location these RNAs are derived from ., To fill this gap in knowledge we determined the landscape of transcription initiation on the KSHV genome ., Our analyses more than tripled the number of known TSCs and thus viral-expressed RNAs ., Furthermore , we identified key sequence features associated with the regulation of viral transcription start sites ., This study provides the first transcriptome-wide characterization of KSHV transcription initiation sites as well as a framework for future studies to define functions of novel viral transcripts and viral gene regulatory elements . | medicine and health sciences, pathology and laboratory medicine, gene regulation, pathogens, microbiology, dna transcription, viruses, dna viruses, genome analysis, sequence motif analysis, molecular biology techniques, nucleotide mapping, microbial genetics, herpesviruses, research and analysis methods, sequence analysis, genomic libraries, transcriptional control, genomics, gene mapping, bioinformatics, medical microbiology, gene expression, microbial pathogens, kaposis sarcoma-associated herpesvirus, molecular biology, viral genetics, virology, database and informatics methods, viral pathogens, genetics, viral gene expression, biology and life sciences, computational biology, organisms | null |
journal.pcbi.1005674 | 2,017 | Noise correlations in the human brain and their impact on pattern classification | The development of fMRI has made it possible to observe the human brain noninvasively as it responds to stimuli or engages in cognitive tasks ., For example , participants might be presented with a series of stimuli drawn from two or more categories ( e . g . , faces and scenes ) , while the blood oxygenation level-dependent ( BOLD ) contrast is measured over time from tens of thousands of volumetric pixels ( voxels ) ., Different events in the experiment can then be linked to changes in BOLD activity , permitting inferences about the neural basis of cognition ( in the example above , about category-selective object perception ) ., However , this is a challenging endeavor because both the physiological processes underlying BOLD activity and the measurement of BOLD activity with fMRI are noisy , and because the resulting datasets can be large and statistically complex 1 , 2 ., Traditionally , fMRI analyses have focused on the information contained in the timecourse of individual voxels or regions ., Such methods are “univariate” because they seek to relate experimental events to single dimensions of BOLD variability , such as the activity averaged across voxels in a region of interest ( ROI ) ., Univariate methods have long been the dominant approach when using brain-imaging data to draw inferences about the neural basis of different aspects of cognition 3 , including: object perception 4 , episodic memory 5 , and cognitive control 6 , 7 ., However , given that cognitive processes are often realized in highly distributed 8 and dynamic 2 ways in the brain , and given that fMRI data have considerable spatial resolution and thus natively live in a high-dimensional space 9 , performance achievable with univariate methods may be inherently limited ., A different class of analyses , multivariate pattern analysis ( MVPA ) , was developed to examine such complex neural representations , treating patterns of BOLD activity across voxels and their link to experimental events as a classification problem 1 , 10 ., MVPA involves training a simple statistical model , in a supervised fashion , to extract regularities in patterns of BOLD activity obtained from different experimental conditions ., The trained model is then used to classify or decode the condition under which previously unanalyzed test data were obtained ., MVPA has led to a wide range of discoveries about the human brain that often go beyond those achievable by applying univariate methods to the same data , including about: perception 11 , 12 , attention 13–15 , memory 16–19 , language processing 20 , 21 and decision-making 22 , 23 ., Although MVPA has been successful across a range of applications , why it is successful has been harder to pin down 10 , 24 , 25 ., One early and still prominent proposal is that MVPA is sensitive to local biases in the manner in which sub-voxel information is represented across populations of voxels 8 , 11 , 12 ., For example , orientation information in the primary visual cortex is represented in sub-millimeter columns 26 , 27 and thus would be obscured at the level of voxels , which typically span a couple of millimeters ., However , because the distribution of orientation columns across voxels is irregular , any given voxel may have a random over-representation of , and thus a weak bias toward , a particular orientation ., Prior studies have argued that by aggregating such weak biases across a population of voxels , the orientation of a stimulus can be reliably decoded using MVPA 11 ., Another possibility is that MVPA allows for the identification of information represented at a larger scale that spans multiple , spatially disparate voxels ., For instance , it is possible to decode stimulus orientation based on the systematic way in which areas of retinotopic visual cortex over-represent the orientation perpendicular to the radius from the fovea 28 ., Regardless of the scale of neural representations , the assumption underlying this prior work is that considering patterns of activity across voxels rather than averaging over them ( as in univariate ROI analyses , for example ) provides additional or different sensitivity ., These theories view neural representations as points in a high-dimensional activity space , with each voxel in the pattern representing a potentially informative dimension ., Although two stimulus categories may be hard to distinguish along any one dimension in this space , jointly considering many voxels allows for better inference by exploiting more dimensions of information ., This interpretation of MVPA downplays an important factor known to influence the representation of information in populations of neurons—that neural variability is correlated in vivo 29–33 ., Both experimental 29 , 30 , 32 and computational 34–36 studies have shown that correlations in neural variability have a significant impact on the information content of neural populations; see 37 , 38 for reviews ., More relevant for present purposes , accurate decoding depends on taking such noise correlations into account 39 , 40 ., Given that noise correlations are important for neural decoding , they may also influence decoding of fMRI data ., Indeed , noise correlations amongst voxels are widespread in fMRI , both during rest 41 , 42 and in the background of tasks 43 , 44 , driven in part by anatomical connections 45 ., Yet , prevailing interpretations of why multivariate decoding is effective have not sufficiently acknowledged the relevance of noise correlations to the decoding of information from populations of voxels ., This is not to say that the classification algorithms themselves disregard correlations among voxels ., Indeed , in most cases these algorithms are sensitive to the presence of correlations 46 , and decoding performance is influenced by them ., Our argument is instead that prevailing interpretations of why MVPA is effective generally center on the benefits of aggregating the information conveyed by patterns of mean activity across voxels , and overlook the influence of correlations ., Even when theories have explicitly considered the influence of correlations , they have generally considered signal correlations: moment-to-moment correlations in the representation of task-dependent stimulus information across multiple voxels in the population ( i . e . , overlap in the representation of the underlying signal across multiple voxels ) ., For instance , if two voxels contain the same signal across training patterns , classification algorithms such as support vector machines ( SVM ) and regularized logistic regression can assign one voxel a higher weight than the other 47 ., Here we propose that noise correlations—which exist persistently before , during , and after experimental events—help explain the effectiveness of MVPA ., In contrast to signal correlations , noise correlations reflect the extent to which noise in the activity of a voxel is correlated with noise in the activity of other voxels in the population ., The theory that motivates this hypothesis is from a recent computational study 36 ., This study showed that the impact of noise correlations on multivariate decoding depends on whether the correlations are between neurons from homogeneous vs . heterogeneous populations , with the latter being beneficial and the former being detrimental ., When considering homogeneous populations—neurons that code for the same stimulus variable—decoding performance worsens as noise correlations increase ., That is , when neurons in a population are selective for the same stimulus , lower noise correlations between them allow the decoder to exploit more dimensions of information ., Indeed , experimental 29 , 30 , 32 and computational studies 48 have found a relation between lower noise correlations in homogeneous populations and increased information ., Importantly , in contrast to homogeneous populations , decoding performance for heterogeneous populations of neurons that code for different stimulus variables can improve as noise correlations increase 36 ., The intuition is that , given a constant amplitude of noise , the presence of noise correlations between neurons coding for different stimulus variables allows a multivariate decoder to recognize that the correlated ( or shared ) variance can be attributed to dimensions that are irrelevant for discriminating between the variables , and can thus be ignored ., This reduces the dimensionality of the classification problem and , more importantly , the amount of overlap between the categorical distributions , thereby improving performance 49 ., Indeed , a recent theoretical study 46 similarly argued that weight vectors in decoding models , such as MVPA , are influenced by both the signal and noise in brain imaging data , thereby suggesting a similar influence of heterogeneous noise correlations on classification performance ., Here we extend this theory—developed 36 and supported 50–52 at the level of neurons—to populations of voxels in fMRI ( Fig 1 ) ., Two challenges arise from this extension: First , it is impossible to know whether a given voxel contains a homogenous neuronal population and even whether multiple voxels with similar selectivity can be considered truly homogenous ., Thus , we focus on the theoretical predictions associated with decoding from heterogeneous populations ( i . e . , that noise correlations among voxels selective for different stimuli will improve decoding of these stimuli ) ., Second , the theory was developed to account for the influence of positive noise correlations ., However , negative correlations can arise in fMRI ( e . g . , depending on preprocessing steps ) , so our analyses consider the influence of both positive and negative noise correlations ., We find that MVPA decoding performance is influenced not only by the selectivity of individual voxels but also by noise correlations between heterogeneous populations of voxels ., Across several analyses of an fMRI dataset , we demonstrate a positive relationship between the magnitude of noise correlations and decoding performance , and we show that as expected with such classifier algorithms 46 , 49 , MVPA exploits noise correlations by assigning higher weights to voxels with higher noise correlations ., We also show that selectivity and noise correlations influence decoding in a complementary fashion—as long as there is signal in the data , performance is modulated by the magnitude of noise correlations ., Indeed , voxels that were highly selective for one class also exhibited higher noise correlations with voxels selective for the other class ., Finally , using a simple model of BOLD activity , we simulate different levels of selectivity and noise correlations in artificial data and show that the benefit of noise correlations for decoding is a ubiquitous property of fMRI data beyond the example dataset ., We used a subset of the data from an fMRI study on attentional control 53 ., Seventeen participants were presented with blocks of face or scene stimuli interleaved with blank periods during two “localizer” runs ., In addition , data were collected during two “rest” runs in which participants only fixated a central point ., Using one of the localizer runs , we fit a general linear model ( GLM ) to the activity observed in ventral temporal cortex , and labeled each voxel as either face-selective or scene-selective based on whether that voxel had greater activation in response to the presentation of face vs . scene stimuli ., Then , we used the rest runs to compute noise correlations , since there were no stimuli or tasks in these runs ., We were specifically interested in heterogeneous noise correlations ( i . e . , noise correlations between voxels with different selectivity ) and thus calculated , for every voxel , the average correlation between its timecourse and the timecourse of all voxels selective for the opposite category ., Finally , to examine how these noise correlations influenced decoding performance , we selected voxels from both face- and scene-selective populations with either high or low noise correlations , and used the other , separate localizer run to train and cross-validate a multi-way ( face/scene/blank ) classifier based on the patterns of activity from these voxels ., If MVPA is sensitive to noise correlations , then classification accuracy should be better for patterns of activity from voxels that are strongly vs . weakly correlated with voxels selective for the opposite category ., As a first pass , we focused on voxels with the highest vs . lowest 1% of noise correlations ( Fig 2A ) and found that classification was better for voxels with the highest noise correlations ( t16 = 7 . 24 , p < 0 . 0001 ) ., This sorting was based on raw values ( high more positive , low more negative ) , but the same result was obtained when we analyzed positive correlations ( high more positive , low closer to zero; t16 = 4 . 12 , p < 0 . 001 ) and negative correlations ( high closer to zero , low more negative 54; t16 = 3 . 19 , p < 0 . 01 ) ., For a more continuous sense of this relationship , we divided voxels into percentiles of raw noise correlations ( Fig 2B ) ., Classification accuracy improved monotonically as MVPA was applied to voxel sets with greater noise correlations ( slope vs . 0: t16 = 6 . 66 , p < 0 . 0001 ) ., Taken together , these results demonstrate a clear influence of the magnitude of heterogeneous noise correlations on decoding performance ., We chose an arbitrary , small bin size of voxels ( 1% ) in the analyses above ., To examine how this parameter affected our findings , we repeated the analysis of raw values with larger bin sizes of high and low noise correlations: 6% , 12 . 5% , 25% , 37 . 5% and 50% ( Fig 3 ) ., While overall decoding performance improved with increasing bin size , decoding was consistently better for patterns of activity from voxels with high vs . low noise correlations ( ps < 0 . 02 ) ., A 2 ( noise correlation magnitude: high vs . low ) x 6 ( bin sizes ) repeated-measures ANOVA revealed that the difference was greater for smaller bin sizes: In addition to main effects of noise correlation magnitude ( F1 , 16 = 28 . 57 , p < 0 . 0001 ) and bin size ( F5 , 80 = 164 . 30 , p < 0 . 0001 ) , there was a reliable interaction between these variables ( F5 , 80 = 14 . 12 , p < 0 . 0001 ) ., This interaction is also consistent with the monotonic relationship across percentiles reported above ( Fig 2B ) : As bin size increased , both the high and low sets included more voxels with intermediate magnitudes of noise correlation , thereby bringing performance closer to the mean across magnitudes ., The analyses above use an L2-norm regularized logistic regression classifier for MVPA ., Such regularization helps avoid over-fitting—which was a risk given that the number of samples in the training set was much smaller than the number of voxels whose weights were learned—by constraining the learning process ., In the case of L2-norm regularization , the sum of squares of the voxel weights is penalized ( here , penalty parameter = 1 ) ., Because all voxels contribute to this sum , this regularization induces interactions between voxels when determining weights ., It could be possible that the influence of noise correlations on decoding performance reflects their effects on such interactions per se rather than the placement of the classifier boundary ., To evaluate this possibility , we repeated the bin size analysis with regularization turned off ., Classification accuracy decreased across the board ( presumably because of over-fitting ) , but we still found greater accuracy for high vs . low noise correlations ( S1 Fig ) ., This suggests that the benefit of noise correlations was not an artifact of regularization ., So far , we have calculated noise correlations from the rest runs and performed classification on the localizer runs ., In using a different run to compute noise correlations , we tacitly assumed that they were stationary across rest and localizer runs ., However , noise correlations may depend on the task condition or may be most closely tied to decoding when actually obtained from the data being decoded ., To examine this possibility , we computed noise correlations between voxels during the localizer run used for crossvalidation ., This is challenging because stimulus-evoked responses can induce signal correlations ., Thus , we first regressed out these responses ( and global noise sources ) and examined BOLD correlations in the residuals ., This “background connectivity” approach has been used successfully across a range of tasks to study noise correlations 44 , 55–57 ., We again identified face- and scene-selective voxels from one localizer run , but then calculated heterogeneous noise correlations ( i . e . , in the residuals ) and classified the other localizer run ., The pattern of results was nearly identical to that obtained when noise correlations were calculated from the separate rest runs , as seen by repeating the bin size analysis ( Fig 4A ) ., Classification accuracy was again consistently better for high vs . low noise correlations ( ps < 0 . 01 ) , and there were main effects of noise correlation magnitude ( F1 , 16 = 18 . 28 , p < 0 . 001 ) and bin size ( F5 , 80 = 152 . 77 , p < 0 . 0001 ) , and an interaction ( F5 , 80 = 6 . 78 , p < 0 . 0001 ) ., In fact , the heterogeneous noise correlation for a given voxel was fairly stable across rest and localizer runs ( Fig 4B ) ., This was quantified with Spearman’s rank order correlation across voxels within participant ( mean rho = 0 . 21; t16 = 5 . 30 , p < 0 . 0001 ) ., Given these results , and because the rest dataset was fully separate , we returned to using the rest runs for calculating noise correlations in the remaining analyses ., We next compared the classification accuracy obtained by selecting voxels with high or low noise correlations in the rest runs across the six bin sizes to classification accuracy obtained for sets of voxels of equal size chosen randomly ( irrespective of noise correlation ) ., If MVPA automatically exploits noise correlations in a given population of voxels , as long there are enough voxels in the population with high noise correlations , MVPA should assign high weights to these voxels and achieve similar performance to a classifier trained only on voxels with high correlations ., For the smallest bin size of 1% , the high noise correlation set produced better decoding performance than the random set ( t16 = 2 . 38 , p = 0 . 03 ) , consistent with the notion that there were not enough voxels with high noise correlations in the random set ( Fig 5 ) ., However , starting at the 6% bin size , decoding performance was indistinguishable between high noise correlation and random sets ( ps > 0 . 09 ) ., Critically , highlighting the efficiency of MVPA at exploiting noise correlations , the random sets exceeded the low noise correlation sets at all bin sizes ( ps < 0 . 001 ) ., Taken together , these results suggest that a small number of voxels with high correlations dominate MVPA decoding performance even when considering large sets of voxels ., We assumed in the previous analysis that MVPA as typically applied ( i . e . , without explicitly considering noise correlations during feature selection ) performed as well as MVPA over voxels with high noise correlation because it automatically assigned these voxels higher weights ., Here we test this directly by carrying out MVPA over all ventral temporal voxels without feature selection and examining the relationship between assigned classifier weights and average heterogeneous noise correlations ., That is , if a voxel was determined to be face-selective in one localizer run , how correlated was, ( a ) its average noise correlation with scene voxels in the rest runs , with, ( b ) its weight assigned for the face category in a classifier trained on the second localizer run ?, We first summarize this relationship using a median-split analysis on the noise correlations ( Fig 6 ) , which revealed that voxels with higher noise correlations were assigned higher weights ( t16 = 3 . 96 , p = 0 . 001 ) ., Another way to look at this relationship is to calculate the Spearman rank order correlation between noise correlation and classifier weight across voxels ., This correlation was reliable across participants ( mean rho = 0 . 045; t16 = 3 . 58 , p = 0 . 002 ) ., The analyses above demonstrate that MVPA decoding performance is enhanced when voxels with high vs . low noise correlations ( measured during rest or in the background of the task ) are selected during classifier training , and that voxels which receive high classification weights in MVPA tend to exhibit high noise correlations with voxels selective for the other class being discriminated against ., However , in addition to the magnitude of noise correlations , decoding performance is also influenced by the selectivity of individual voxels ( i . e . , how differently a voxel responds to the conditions being classified ) ., In this section , we examine the relative influence of selectivity on MVPA decoding performance ., We first consider the extent to which selectivity and noise correlations interact ., For instance , when we divided voxels in our dataset into percentiles of raw noise correlations , we observed a monotonic improvement in MVPA decoding performance with an increase in the magnitude of noise correlations ( Fig 2B ) ., How does selectivity vary across these sets of voxels ?, To answer this question , we took the absolute value of the selectivity scores that had been used to identify face and scene voxels in one localizer run ( i . e . , for determining which voxels should count as having opposite selectivity when calculating heterogeneous noise correlations ) ., As a reminder , these scores reflect the face vs . scene contrast from the GLM , specifically the z-scored difference of the parameter estimates modeling the average evoked response from face and scene blocks , respectively ., Average selectivity increased monotonically ( Fig 7 ) as we moved from voxels with low noise correlations to voxels with high noise correlations ( slope vs . 0: t16 = 5 . 03 , p < 0 . 001 ) , and the Spearman rank order correlation between noise correlation and selectivity across voxels was reliable ( mean rho = 0 . 076; t16 = 4 . 17 , p < 0 . 001 ) ., In other words , voxels with higher selectivity for one of the two categories also had higher noise correlations with voxels selective for the other category ., Given the link between selectivity and noise correlations across voxels in our empirical dataset , we next sought to examine their cumulative influence on decoding ., We selected voxels with the top vs . bottom 12% of noise correlations , and within each set selected voxels with high vs . low selectivity based on a median split of voxel selectivity from the GLM ., We then examined MVPA classification accuracy for the patterns of activity from voxels in each of the resulting four bins with 6% of voxels ( Fig 8 ) ., Of particular note in this analysis is the comparison between low noise correlation/low selectivity and high noise correlation/low selectivity , which had comparable levels of selectivity ( 1st and 3rd columns of Fig 8A ) but dramatically different classification accuracy ( same columns of Fig 8C ) ., This suggests that as long as there is a minimum amount of signal conveyed by selectivity , which allows for above-chance classification , noise correlations can be sufficient to increase decoding performance ( same columns of Fig 8B ) ., This claim is further reinforced by the comparison of low noise correlation/high selectivity to high noise correlation/low selectivity ., Although there was a dramatic difference in signal conveyed via selectivity ( 2nd and 3rd columns of Fig 8A ) , classification accuracy did not differ and was in fact numerically in the opposite direction ( same columns of Fig 8C ) , suggesting that the selectivity difference was offset by the reverse difference in noise correlations ( same columns of Fig 8B ) ., Taken together , these results support the notion that when selectivity differences are present , noise correlations can influence classification accuracy ., So far , we have used an existing fMRI dataset to demonstrate that MVPA is highly attuned to noise correlations between voxels , and that decoding performance may be sensitive to the information carried both by the selectivity of individual voxels and the noise correlations between them ., We next sought to expand upon these findings in two ways: First , as described above , selectivity and noise correlations were inherently confounded in the empirical dataset ., How might we better examine the cumulative contributions of noise correlations and selectivity to decoding performance ?, Second , all of the findings reported above were based on one fMRI dataset with particular characteristics ., To what extent do our conclusions apply to other datasets and reflect a general principle about the computational underpinnings of MVPA ?, To address these issues , we developed a simple model of selective coding in the presence of noise correlations , wherein we could independently vary voxel selectivity and heterogeneous noise correlations ., By performing MVPA over artificial BOLD activity generated from this model , we could then simulate the influence of different parameters ., The model included a set of voxels roughly matched in number to the 1% bin size in our earlier analyses ., By construction , half of the voxels responded preferentially to face stimuli and the other half to scene stimuli ., The mean responses , variances , and correlations of all voxels in the model were drawn from the range observed in our empirical dataset , ensuring that the simulated voxels produced physiologically realistic activity ., Following Azeredo da Silveira & Berry ( 2014 ) , we used a Gaussian approximation in each of 100 model “participants” to sample data for time points from face and scene blocks ( matched to the number of face and scene TRs in the empirical dataset ) ., To ensure that the resulting timecourses were temporally autocorrelated like real BOLD activity 58 , we convolved them with a canonical hemodynamic response function ( HRF ) ., Finally , we performed cross-validated MVPA over the artificial patterns of activity obtained from the simulated voxels ., We first sought to examine the influence of heterogeneous noise correlations on decoding performance ., Noise correlations across pairs of voxels varied according to whether the voxels were drawn from the pool of face-selective voxels , the pool of scene-selective voxels , or one from each of the pools ., We performed 20 simulations manipulating the magnitude of across-pool noise correlations linearly between 0 and 0 . 22 ( i . e . , the range of positive noise correlations in the empirical dataset ) , while holding all other parameters constant ., There was a monotonic increase in classification accuracy as the magnitude of heterogeneous noise correlations increased ( blue curve in Fig 9A ) ., This is precisely the pattern predicted by the computational theory on which our study was based 36 , and is similar to the pattern of results observed in our empirical dataset ., Notably , by allowing noise correlations to vary while the selectivity of the voxels in the two pools was held constant , these results show that noise correlations are sufficient to influence above-chance decoding performance ., To examine the influence of voxel selectivity , we repeated the analysis above but further manipulated the strength of face and scene selectivity in the mean responses of voxels from the two pools , over a fixed range of noise correlations ., As expected , when voxel selectivity decreased across three levels , overall decoding performance also decreased ( Fig 9A ) ., However , at all levels , we observed the same monotonically increasing relationship between classification accuracy and the magnitude of noise correlations ., Notably , selectivity affected classification accuracy even with near-zero noise correlations , but the effect of selectivity was stronger in the regime of stronger noise correlations ., We next sought to examine the extent to which these results depend on the specific parameters used in our simulations ., For instance , in the simulations described thus far , the variance of all voxels was matched to the median variance observed in our empirical dataset ., Given that overall noise in the system , correlated or otherwise , is ultimately a function of the variability in the activity of individual voxels , we examined the extent to which our results depended on the magnitude of voxelwise variance ., Repeating our analysis across three levels of variance , spanning the range observed in our empirical dataset , we found a similar influence of noise correlations on classification accuracy ( Fig 9B ) ., Specifically , as voxel variance increased , thereby increasing noise in the system , overall decoding performance went down; however , at every level of variance , we observed the same relationship between classification accuracy and the magnitude of noise correlations ., Another modeling choice we made was to sample activity within the face- and scene-selective voxel pools based on homogeneous mean , variance , and correlation values matched to population averages from our empirical dataset ., We next examined the influence of introducing heterogeneity in the response properties of the simulated population of voxels ., We generalized our model to include greater voxelwise diversity by randomly varying the population covariance matrix according to a Gaussian distribution with SD equal to 10% of the original value ., We similarly varied the mean responses of individual voxels ( while maintaining selectivity ) in each population according to a Gaussian distribution with SD matched to the mean within-population SD from our empirical dataset ., We obtained the same pattern of results from MVPA , with classification accuracy increasing monotonically as the magnitude of noise correlations increased ( Fig 9C ) ., Indeed , greater population diversity led to a steeper increase in classification accuracy , consistent with the notion that heterogeneity can be beneficial , especially at higher levels of noise correlation ., MVPA has proven useful for decoding information from brain imaging data 1 , 10 , with insights often extending what has been learned from univariate methods ., Although the effectiveness of MVPA has been widely acknowledged , which aspects of neural representation MVPA taps into are still debated 2 , 10 , 24 , 25 , 28 ., Prior theories argued that MVPA benefits from aggregating signals across voxels—either local biases in the mapping of micro-scale representations onto voxels 11 , 25 or more global , macro-scale representations that span multiple voxels 28 ., In both cases , the argument was that MVPA exploits the distribution of weak or uncertain feature-selective signals to identify regularities that discriminate experimental conditions ., Our findings show that this interpretation is incomplete: Instead of thinking of each voxel as making a distinct contribution to the information represented collectively by the population of voxels , MVPA is also highly attuned to noise correlations between voxels ., This reflects the mechanics of classification algorithms 49 and builds on neurophysiological studies showing both that noise correlations impact the information content of neural populations 36–38 and that accurate decoding of this information requires taking these noise correlations into account 39 , 40 ., Specifically , our study was inspired by a recent computational theory 36 , which proposed that multivariate decoding is enhanced for heterogeneous neural populations with high noise correlations ., Extending this proposal to the problem of multivariate decoding with fMRI data , we show that noise correlations between heterogeneous populations of voxels influence MVPA ., The same result was | Introduction, Results, Discussion, Materials and methods | Multivariate decoding methods , such as multivoxel pattern analysis ( MVPA ) , are highly effective at extracting information from brain imaging data ., Yet , the precise nature of the information that MVPA draws upon remains controversial ., Most current theories emphasize the enhanced sensitivity imparted by aggregating across voxels that have mixed and weak selectivity ., However , beyond the selectivity of individual voxels , neural variability is correlated across voxels , and such noise correlations may contribute importantly to accurate decoding ., Indeed , a recent computational theory proposed that noise correlations enhance multivariate decoding from heterogeneous neural populations ., Here we extend this theory from the scale of neurons to functional magnetic resonance imaging ( fMRI ) and show that noise correlations between heterogeneous populations of voxels ( i . e . , voxels selective for different stimulus variables ) contribute to the success of MVPA ., Specifically , decoding performance is enhanced when voxels with high vs . low noise correlations ( measured during rest or in the background of the task ) are selected during classifier training ., Conversely , voxels that are strongly selective for one class in a GLM or that receive high classification weights in MVPA tend to exhibit high noise correlations with voxels selective for the other class being discriminated against ., Furthermore , we use simulations to show that this is a general property of fMRI data and that selectivity and noise correlations can have distinguishable influences on decoding ., Taken together , our findings demonstrate that if there is signal in the data , the resulting above-chance classification accuracy is modulated by the magnitude of noise correlations . | A central challenge in cognitive neuroscience is decoding mental representations from patterns of brain activity ., With functional magnetic resonance imaging ( fMRI ) , multivariate decoding methods like multivoxel pattern analysis ( MVPA ) have produced numerous discoveries about the brain ., However , what information these methods draw upon remains the subject of debate ., Typically , each voxel is thought to contribute information through its selectivity ( i . e . , how differently it responds to the classes being decoded ) , with improved sensitivity reflecting the aggregation of selectivity across voxels ., We show that this interpretation downplays an important factor: MVPA is also highly attuned to noise correlations between voxels with opposite selectivity ., Across several analyses of an fMRI dataset , we demonstrate a positive relationship between the magnitude of noise correlations and multivariate decoding performance ., Indeed , voxels more selective for one class , or heavily weighted in MVPA , tend to be more strongly correlated with voxels selective for the opposite class ., Furthermore , using a model to simulate different levels of selectivity and noise correlations , we find that the benefit of noise correlations for decoding is a general property of fMRI data ., These findings help elucidate the computational underpinnings of multivariate decoding in cognitive neuroscience and provide insight into the nature of neural representations . | medicine and health sciences, statistical noise, diagnostic radiology, functional magnetic resonance imaging, engineering and technology, signal processing, neuroscience, magnetic resonance imaging, simulation and modeling, hemodynamics, mathematics, statistics (mathematics), brain mapping, neuroimaging, research and analysis methods, computer and information sciences, imaging techniques, animal cells, hematology, cellular neuroscience, radiology and imaging, diagnostic medicine, cell biology, neurons, software engineering, information theory, biology and life sciences, cellular types, physical sciences, background signal noise, preprocessing | null |
journal.pcbi.1000500 | 2,009 | Parallel Computational Subunits in Dentate Granule Cells Generate Multiple Place Fields | Neurons possess highly branched , complex dendritic trees , but the relationship between the structure of the dendritic arbor and underlying neural function is poorly understood 1 ., Recent studies suggest that dendritic branches form independent computational subunits: Individual branches function as single integrative compartments 2 , 3 , generate isolated dendritic spikes 4 , 5 linking together neighbouring groups of synapses by local plasticity rules 6–8 ., Coupling between dendritic branches and the soma is regulated in a branch-specific manner through local mechanisms 9 , and the homeostatic scaling of the neurotransmitter release probability is also regulated by the local dendritic activation 10 ., The computational power of active dendrites had already been demonstrated by several computational studies 11–16 , but how local events influence the output of the neuron remained an open question ., Using the cable equation 17 or compartmental modelling tools one can calculate the current or voltage attenuation between arbitrary points in a dendritic tree 14 , which is in good agreement with in vitro recordings ., However , cortical networks in vivo are believed to operate in a balanced state 18 , 19 , where the inhibitory drive is continuously adjusted such that the mean activity of the population is nearly constant 20 , 21 ., In this case , the firing of an individual neuron is determined , beyond its own input , by the activity distribution of the population ., A simple cascade model 22 incorporating numerous dendritic compartments allowed us the statistical estimation of the activity distribution of neurons within the population ., We used this model to study how localized dendritic computations influence the output of the neuron ., The present study focuses on hippocampal granule cells ., Compared to pyramidal neurons granule cells have relatively simpler dendritic arborization: They lack the apical trunk and the basal dendrites , but are characterized by several , equivalent dendritic branches , extended into the molecular layer 23 ( Figure 1A ) ., Recordings from freely moving rats revealed that like pyramidal neurons , granule cells exhibit clear spatially selective discharge 24 , 25 ., However , granule cells had smaller place fields than pyramidal cells , and had multiple distinct subfields 24 , 26 ., It has also been recently shown that these subfields are independent , i . e . , their distribution was irregular and the transformation of the environment resulted in incoherent rate change in the subfields 26 ., The dendritic morphology of granule cells suggest that parallel dendritic computations could contribute to the generation of multiple , distinct subfields of these neurons ., In the present study we analyzed how synaptic input arriving to dendritic subunits influence the neuronal output ., First , we introduce the model used in this study and we define statistical criteria to measure if a dendritic branch alone is able to trigger somatic spiking ., We show , that generally neurons perform input strength encoding i . e . , input to the whole dendritic tree but not activation of a single branch is encoded in the somatic firing ., Next we demonstrate that if the local response is enhanced by active mechanisms ( dendritic spiking and synaptic plasticity ) then neurons switch to feature detection mode during which the firing of the neuron is usually triggered by the activation of a single dendritic branch ., Furthermore we show that moderately branched dendritic tree of granule cells is optimal for this computation as large number of branches favor local plasticity by isolating dendritic compartments , while reliable detection of individual dendritic spikes in the soma requires low branch number ., Dendritic branches of dentate granule cells could therefore learn different inputs; and the cell , activated through different dendritic branches , could selectively respond to distinct features ( locations ) , participating in different memories ., Finally using spatially organized input we illustrate that our model explains the multiple independent place fields of granule cells and these dendritic computations increase the pattern separation capacity of the dentate gyrus ., Supposing that firing rates of presynaptic neurons ( uj ) are independent and identically distributed we assume that the total input of the dendritic branches Ui\u200a=\u200aΣjwijuj is drawn randomly from a Gaussian distribution with mean μ and variance σ2: ( 5 ) where pU indicates a probability distribution over U ( Figure 1C; see Eq . 17 in Methods for parameters specific to hippocampal granule cells ) ., More specifically , indicates the distribution of the magnitude of possible total inputs to a single dendrite over many different instances ., Based on the distribution of the total input , we can compute the distribution of the somatic activation and determine the firing threshold ( β ) according to the proportion of simultaneously active cells ( the sparseness of the representation , spDG ) in the DG 24 ., First , we rearrange Eq ., 3 using the input distribution to express the distribution of : ( 6 ) where indicates that the inputs of the dendritic branches are randomly sampled from a Gaussian distribution ., We substitute Eq ., 6 into Eq ., 4 , and we get ( 7 ) We can assume again , that the inputs ( Ui ) of the dendritic branches are independent and identically distributed variables ., ( Note , that while the activations are not independent because of the back-propagation of currents from the soma , the inputs are . ), If N is high enough , we can approximate the sum in Eq ., 7 with a Gaussian distribution , and rewrite the equation: ( 8 ) where indicates a probability distribution over , while μF and are the expected value and the variance of the dendritic integration function F ( U ) given the input distribution : ( 9 ) ( 10 ) We calculated the integrals 9–10 with two different forms of dendritic integration of synaptic inputs: a linear and a quadratic function ( Figure 1C ) ., The details of these calculations are in the Supporting Information ( Text S4 ) ., In this paper we do not model inhibitory neurons in the dentate gyrus , however , we assume , that they play a substantial role in continuously adjusting the firing threshold of principal neurons and regulating the activity of the network 20 , 21 ., As a result of this regulation always the most depolarized neurons are able to fire , and the proportion of simultaneously active neurons is characteristic for different hippocampal areas 24 , 29 ., Given that all neurons share a common input statistics and have similar internal dynamics , equation 8 also describes the distribution of across the granule cell population at a given time ., If only the most depolarized 1–5% of the population are able to fire 29 , this also means that only those neurons exceed their firing threshold whose activation is within the uppermost 1–5% of the distribution described by Eq ., 8 ., Therefore , the proportion of simultaneously active neurons within the dentate gyrus spDG 24 , 29 also determine the firing threshold β for granule cells ., We approach the dendritic independence by focusing on the statistical distributions of the input to dendritic branches , as these branches form the basic computational subunits in our model ., We ask whether the input of a single branch could be sufficiently large to significantly depolarize not only the given branch but also the soma of the neuron ., We defined two conditions to study whether the spiking of the neuron is caused by the activation of a single dendritic branch or by the simultaneous depolarization of multiple branches ., First , the conditional probability is the probability of firing given that any branch k has total input Uk\u200a=\u200aΣjwkjuj , while inputs to all other branches are random and independent samples from the distribution of ( Figure 2A ) ., At those Uk values where this probability is close to 1 the cell tends to fire when any of the dendritic branches gets that input ., Second , the conditional distribution is the distribution of the synaptic input of the most active branch at the time the depolarization of the soma exceeds the firing threshold ( β ) , where U* is the total synaptic input arriving to the most active branch ( Figure 2A ) ., K ( U* ) can be regarded as the marginal distribution of above the firing threshold ( Figure 2B ) ., The probability mass of this function shows the typical maximal input ( U* ) values when the neuron fires ., These two conditions together determine whether a single branch can be sufficiently depolarized to trigger somatic spike or not ., If the probability of firing is high ( H ( U ) ≈1 ) at typical input values ( K ( U* ) ) then the firing of the cell is caused by a single branch ., With the definition of Gasparini and Magee 30 we call this form of information processing as independent feature detection ., On the other hand , if the firing probability is low ( H ( U ) ≪1 ) even if one of the branches receive extremely large input ( U* is high ) then the cell mostly fires when the overall dendritic activation is high , and even the most depolarized branch usually fails to make the neuron fire ., We use the expression input strength encoding 30 to denote this second type of computation ., The calculation of the two functions H ( U ) and K ( U* ) is described in the Methods section ., First we chose unstructured synaptic input , i . e . , the firing of entorhinal neurons were independent and the strength of all synapses were equal ., In this case we approximated the total synaptic input U to a branch with a Gaussian distribution ( Eq . 5 , Figure 1C ) ., Given the input distribution we asked whether the excitation of single branches can be sufficiently large to cause significant depolarization in the soma ., The typical largest input values , indicated by the probability mass of K ( U* ) ( Figure 2C–D ) are unable to sufficiently depolarize the soma and determine the neuronal output ( indicated by the low H ( U ) values ) in the case of both the linear ( Figure 2C ) and the quadratic ( Figure 2D ) integration functions ., Wherever K ( U* ) has high values , H ( U ) is low in both cases , which indicate , that these branches are not able to independently influence the output of the neuron ., Only coactivation of several branches could make the neuron fire in this case , and the output of the neuron encodes the strength of all dendritic inputs ., As H ( U ) converges to 1 for high input values extremely high inputs to a single dendrite could reliably trigger somatic firing ., In the next sections , however , we study how synaptic plasticity selectively modifies individual synapses and contributes to the sparse occurrence of extraordinarily high input values ., During Hebbian learning synapses contributing to postsynaptic activation are potentiated while other synapses may experience compensatory depression 31 , 32 ., We simulated the learning process by showing a finite number of uncorrelated samples from the input distribution ( see Methods ) to the model neuron initiated with uniform synaptic weights ., The synaptic weights of those dendritic branches where the activation exceeded a threshold , βd were modified according to the following Hebbian plasticity rule 33 that incorporates heterosynaptic depression 31: ( 11 ) where is the local dendritic activation , uj is the presynaptic firing rate and wij is the synaptic strength ., is the Heaviside function and γ<1 is a constant learning parameter ., Note , that the learning rule is local to the dendritic branches: the synaptic change depends on the local activation but not on the somatic firing ., Next , we calculated the total input to the branches Ui\u200a=\u200aΣjwijuj after modification of synapses ( Figure 3A ) , and recalculated the two functions H ( U ) and K ( U* ) defined previously with the new input distribution ( Eq . 18 ) ., As shown on Figure 3A the total synaptic input in response to a learned pattern increases significantly after learning ( compare blue and grey curves on Figure 3A ) , while untrained patterns generate smaller synaptic inputs ( compare grey and black curves on Figure 3A ) ., The main consequence of synaptic plasticity is that the trained patterns generate much larger local response than untrained patterns , which raise the possibility of their detection in the soma ., Note , that an unspecific increase of synaptic weights would result in an upward shift of both the input distribution ( Eq . 5 ) and the firing threshold , but would not affect the somatic detection of individual dendritic events ., The neuron is able to selectively respond to the dendritically learned patterns if a single branch , when facing with its preferred input , is able to induce significantly more depolarization at the site of the action potential initiation compared with the case when all of the branches get random , not learned input ., Figure 3B–E shows the dendritic input and the activation of the soma after learning ., If the maximal input U* is small ( left bumps on Figure 3B , D ) and none of the branches got its preferred input then the somatic activation is usually small ., If U* is high ( Figure 3B , D; right bumps ) , which means that one of the branches receives its preferred input pattern , then the somatic activation is increased ., The increase of the somatic activation with learned input is only moderate in the linear case ( Figure 3B , C ) resulting in an incomplete separation of learned and not learned inputs by the somatic firing threshold ., However , if synaptic inputs are supra-linearly ( quadratically ) integrated within the dendritic branches , efficient separation is possible: the probability that the presentation of a learned pattern elicits subthreshold somatic response , called dendritic spike detection probability was over 95% ( Figure 3D , E ) ., In this case the output of the neuron encodes whether or not one of the stored features was present in the neurons input and not simply the strength of the total input arriving to the whole dendritic tree ., In other words , if dendritic nonlinearity enhance the response of a given branch to its preferred input , then this branch alone is able to trigger somatic spiking ., In the following sections we use the term dendritic spiking to refer to these supra-linear dendritic events ., Although there is no data available on the synaptic induction of local dendritic spiking in hippocampal granule cells , voltage dependent Ca2+ currents are present in the membrane of granule cells 34 , 35 and whole-cell recordings from these neurons suggest that T-type Ca2+ channels can generate dendritic action potentials at least in young neurons 36 or under hyper-excitable conditions 34 , 37 ., Next , we explored how the independent feature detection ability of the model depends on the resistance between the somatic and dendritic compartments with nonlinear dendritic integration ., In the passive cable model of dendritic trees the space constant of the membrane λm≈ ( Rm/Ri ) 1/2 plays a substantial role in determining the voltage attenuation among two sites ., Consequently , an increase in the intracellular resistivity Ri or a similar decrease in the membrane resistance Rm will contribute to the separation of dendritic subunits by decreasing the membranes space constant λm ., In the present study we used the inverse of the space constant R≈Ri/Rm to characterize the degree of electrical resistivity between the somatic and dendritic compartments ., Indeed , an increased resistivity ( R ) between the compartments ( smaller space constant ) induced larger degree of electrical isolation as the somatic response to the same amount of dendritically applied current decreased ( compare Figure 4A left and right panels ) ., However , this isolation did not modify the dendritic spike detection probability in the soma: Large dendritic spikes localized to a single compartment could be reliably separated from subthreshold events with a somatic firing threshold at a large range of resistances R ( Figure 4A–B ) ., This was also true for the selective alternation of the somatic or the dendritic membrane resistance ( Figure 4B ) ., On the other hand , the resistance parameter had a substantial impact on the isolation of different dendritic compartments which might be necessary for the independence of synaptic plasticity ., To measure the isolation of the dendritic subunits we calculated the influence of other compartments on the activation of a given branch ( external influence ) quantified by the standard deviation of ., Figure 4C shows the activation of a dendritic branch in the function of its input at different R values ., If the resistance is small ( , Figure 4C , left ) , then the local activation depends only slightly on the local input and the external influence is high ( Figure 4D ) ., In this case the local input spread out to the entire dendritic tree and activates similarly all branches ., On the other hand , if the resistance is high ( R\u200a=\u200a1 , Figure 4C , right ) then the external influence is small , and the depolarization of a dendritic branch depends mostly on the local input ., Interestingly , decreasing the resistance of the perisomatic membrane ( ) alone was more efficient in separating the dendritic subunits than decreasing the resistance of the dendritic membrane or both ( Figure 4D ) ., The extensive GABAergic 38 , 39 and glutamatergic 40 innervation of the proximal dendritic and perisomatic region of granule cells may therefore contribute significantly to the isolation of the dendritic compartments ., The impact of a single branch on the somatic activation , and also the coupling between dendritic branches may depend highly on the structure of the dendritic tree ., Therefore we varied the number of dendritic subunits , N , and calculated the probability of detecting dendritic spikes in the soma and the external influence on the dendritic subunits ( Figure 5 ) ., The probability of detecting a dendritic spike in the soma decreased gradually after a few ( N≈30 ) number of branches from 1 to 0 . 3 ( N≈1000 , Figure 5A–B ) ., If the number of branches was low , then the effect of a single branch on the soma was relatively high , and the somatic detection of single dendritic events was reliable ., Conversely , one out of hundreds of branches had relatively low impact on the neurons output even if the local depolarization was significant ., The electrical coupling between the dendritic subunits characterized by the external influence on the local activation also decreased with the number of branches , ( Figure 5C–D ) ., In the model the branches are connected through the somatic compartment , and because the variance of the somatic activation decreases if N increases ( Eq . 8 ) , the external influence will also decrease ., However , in a complex dendritic tree containing higher number of subunits the branches are electronically more isolated which is required for local plasticity ., To keep the probability of dendritic spike detection high and the dendritic coupling low at the same time , the number of branches should therefore be as high as possible , but not higher than N≈60 ., As we showed on Figure 4 , the dendritic coupling depends on the resistance R , as high resistance separates better the subunits ., Therefore we conclude , that a medium number of branches with relatively high resistance is ideal for parallel dendritic computations ., The optimal number of dendritic subunits , however , depends on the size of the dendritic event determined by the local integration of the synaptic inputs ( Figure 5B ) ., Appropriate detection of dendritic responses to learned patterns with linear integration is possible only in very small dendritic trees , whereas supra-linear integration allows the detection of individual dendritic events also in a larger dendritic arbor ., Nonlinear integration by dendritic spiking therefore permits the neuron to selectively respond to a larger number of distinct input pattern ., During the calculation above we assumed , that the activity of the presynaptic neurons are independent and that the samples from the distribution are uncorrelated ., It is known , however , that the firing of entorhinal neurons are not independent: At least half of layer II cells in the medial entorhinal cortex ( EC ) are grid cells , whose firing depend mostly on the position of the animal 27 ., Moreover , in reality animals do not face with discrete uncorrelated samples , but they experience the continuous change of their environment which is mirrored by the activity of the entorhinal neurons ., In order to test our model under more realistic conditions , we simulated the activity of the rodents EC during exploratory behavior as input to our modeled granule cell ., The EC consisted of two neuron population: A population of grid cells ( 1000 neurons , 5 spacing , 5 orientations ) representing a path integrator system 41 and a population of visual cells ( 1200 units ) , representing highly processed sensory information available in the EC 42 ., In these simulations we used the Webots mobile robot simulator 43 ., The firing statistics of the entorhinal neurons was the same as used in the analytical calculation except that the activity of the neurons was location dependent ., Moreover , as we simulated the trajectory of the rat during continuous foraging for randomly tossed food pellets 26 the subsequent input patterns were highly correlated ., We simulated a single granule cell with N\u200a=\u200a20 dendritic branches each of them receiving a total number of M\u200a=\u200a100 synaptic contacts from entorhinal neurons ., The resistance was R\u200a=\u200a1 , we used the quadratic integration function and the neuron was tested in 5 different environments ., During the 5 min . learning period ( while 2000 spatial locations was sampled with an average running speed of 0 . 22 m/s ) 0–8 branches learned usually at different spatial locations in each of the 5 environments ., In most of the time synaptic plasticity in different branches occurred at different places , therefore the subunits were able to learn independently ., Moreover , learning occurred only in naive branches , i . e . , each branch learned only in one environment at a specific location and synapses of trained branches did not engage in learning at a different location ., After the training period the synaptic weights of those branches that were subthreshold for synaptic plasticity ( βd\u200a=\u200a1 . 11 ) in all environments were scaled down manually ., Next we studied the spatial activity pattern of the somatic and dendritic compartments while the robot was moving on a different track in the same environments ., The dendritic branches responded with high activation ( “dendritic spikes” ) to subsequent visit of places close to their preferred locations leading to the formation of dendritic place fields ( Figure 6 ) ., Moreover , since the activation of the soma was substantially increased in each of these dendritic place fields , the neuron had a multi-peaked activity map in several environments ( Figure 6 ) ., Finally we explored the effect of the size of the dendritic tree on the spatial firing pattern of the neuron ( Figure 7 ) ., If there were only a few functional dendritic subunit than the neuron obviously had a small number of dendritic place fields ( Figure 7A ) , but the individual branches had strong influence on the somatic activity ., Therefore the correlation between the somatic activation as and the maximal dendritic input U* was high ( Figure 7B , C ) , as predicted by the analytical calculations ., On the other hand , in neurons with large number of dendritic subunits there were more dendritic place fields ( Figure 7A ) , but a single branch had only a little impact on the activity of the neuron ( Figure 7D ) ., Accordingly , the correlation between the maximal dendritic input and somatic activation was reduced ( Figure 7B ) ., In these cases the cell fired when the overall excitation was high or when more than one branch were simultaneously excited ., Therefore , the moderately branching dendritic tree of granule cells seems optimal for parallel dendritic computations since extensive branching inhibits the detection of individual dendritic events ., We conclude , that clustered plasticity together with dendritic spiking may be an adequate cellular mechanism to explain the generation of multiple place fields in the DG 24 , 26 ., Dendritically generated spikes mediated by voltage-gated Na+ 3 and/or Ca2+ channels 44 as well as glutamate-activated N-methyl-D-aspartate ( NMDA ) channels 45 have been described in a variety of neurons ( for a review see 46 or 47 ) including hippocampal granule cells 34–37 ., We used a quadratic integration function in order to analytically model supra-linear dendritic integration 15 which differs from the sigmoid form of nonlinearity realized by dendritic spiking ( Text S1 , 3 , 4 , 45 ) ., We believe , however , that at this level of abstraction the exact form of nonlinearity does not affect our results: As that is the difference between the dendritic responses to learned and not learned patterns that influence the somatic detection of dendritic events , a sigmoid integration function give qualitatively similar results ( Text S2 ) ., Moreover , we studied only passive interactions between individual dendritic events as the effect of voltage and calcium dependent currents ( including A-type and Ca2+-dependent potassium 48 and the H-current 49 ) regulating the propagation of dendritic spikes were not included in the model ., Future studies using a compartmental model equipped with dendritic spiking could support our results and clarify further details ., Our analysis has revealed that a moderately branched dendritic tree is optimal for the independent branches model , and we have shown that this mechanism could contribute to the spatial firing properties of granule cells in the DG ., The dendritic tree of cerebellar Purkinje cells as well as the apical dendrites of hippocampal and neocortical pyramidal cells is typically larger , and more ramifying 50 ., Their morphology is suitable for local plasticity within single branches 6 , 8 , and although it seems that individual branches may function as single integrative compartments 3 , 4 , 51 , 52 , dendritic spikes localized to these compartments fail to propagate to the soma and directly influence the neurons output 53 ., Larger dendritic events , active spread of dendritic spikes towards the soma or interactions among dendritic subunits could contribute to the generation of somatic action potentials in this case ., The dendritic tree of pyramidal neurons is , however , far more complex than that of granule cells: it has several morphological and functional subregions with different afferent inputs and membrane excitability 50 ., Understanding how their spatial firing characteristics arise from their cellular properties would require at least a different model structure and is beyond the scope of this paper ., Whether individual dendritic events influence the output of the neuron depends - beyond the structure of the dendritic tree - on the size and the frequency of the large dendritic events and the output sparsity ., The size of the events depends on the exact form of the dendritic integration function and the plasticity rule while the input statistics determine the frequency of such events ., We have shown that given the sparseness of the output , sufficiently large , localized dendritic events arriving with appropriate frequency are able to separately determine the output of the neuron ., Whether a local event is sufficiently large depends on the geometry of the dendritic tree: A smaller event may be sufficient if there are only a few subunits , or if the events actively propagate to a large part of the entire dendritic tree ( e . g , the apical tuft in pyramidal neurons , 54 ) ., Conversely , in neurons such as cerebellar Purkinje cells with large , ramifying dendritic tree , where individual events are localized to small branches , very large dendritic spikes would be required to influence the output ., Indeed , detailed compartmental modelling of dendritic morphology revealed that the forward propagation of the action potential initiated in the apical trunk of pyramidal neurons was very effective , while in Purkinje cells dendritic action potentials were rapidly attenuated 53 ., Clustered plasticity allows the neuron to simultaneously learn several different patterns but requires the electrical and/or biochemical isolation of the dendritic compartments 47 , 55 ., However , the intracellular resistance ( ) in dentate granule cells is relatively low and granule cells are usually regarded as electrically compact neurons 28 ., Indeed , signal propagation from somata into dendrites in vitro is more efficient in granule cells compared with CA1 pyramidal cells and distal synaptic inputs from entorhinal fibers can efficiently depolarize the somatic membrane of granule cells 28 ., However , in vitro studies do not take into account that neurons are embedded in a network of spontaneously active cells ., As thousands of synapses bombard the dendritic tree in vivo , the dendritic membrane becomes “leakier” and , consequently , the membranes space constant decreases significantly 56 ., Moreover perisomatic inhibition 57 and feed-back excitation ( via hilar mossy cells 40 ) further decrease the resistance of the proximal membrane contributing to the separation of the somatic and dendritic compartments 54 , 58 ., More specifically , we predict , that the membrane resistance of granule cells is considerably smaller at the perisomatic region than in the distal dendrites ., Indeed , computational studies predict a 7–30 fold increase in the somatic leak conductance due to the synaptic background activity 59 ., On the other hand , large space constant at long terminal branches facilitate interactions among synapses distributed on the same branch ., Therefore the long dendritic branches of dentate granule cells may act as single integrative computational subunits , separated from each other by the perisomatic region of the cell ., Furthermore , in the present paper we used steady-state approximations and we neglected temporal characteristics of the input and the integration ., For rapidly varying inputs the coupling between dendritic sites and the soma is much smaller than for slowly varying currents since the distributed capacitance throughout the tree will absorb the charge before it reaches the soma 14 ., Therefore dendritic compartments in a passive tree are more isolated for transient events such as dendritic spikes than for steady-state current ., Finally , biochemical compartmentalization is likely to play a substantial role in the cooperative induction of LTP in both hippocampal 60 and neocortical neurons 7 ., If , on the other hand , dendritic branches are not isolated during the learning process and synapses across the whole dendritic tree are modified simultaneously then different dendritic branches will be sensitive for different component ( modalities ) of the same episode ., A new episode with partial overlap with the previously learned one may trigger dendritic spiking in the corresponding dendritic branch ., As the somatic detection probability of dendritic spikes does not depend on the degree of electrical isolation ( Figure 4 ) , individual branches trigger somatic spiking , and , in this way the dentate gyrus contributes to the associative recall of the previously encoded episode in the hippocampus ., Since the first description of LTP at perforant path - granule cell synapses 61 synaptic plasticity has become widely accepted as the physiological basis of memory 62 ., As Hebbian plasticity is intrinsically unstable , simply because it is a positive feed-back mechanism multiple stability-promoting mechanisms h | Introduction, Model, Results, Discussion, Methods | A fundamental question in understanding neuronal computations is how dendritic events influence the output of the neuron ., Different forms of integration of neighbouring and distributed synaptic inputs , isolated dendritic spikes and local regulation of synaptic efficacy suggest that individual dendritic branches may function as independent computational subunits ., In the present paper , we study how these local computations influence the output of the neuron ., Using a simple cascade model , we demonstrate that triggering somatic firing by a relatively small dendritic branch requires the amplification of local events by dendritic spiking and synaptic plasticity ., The moderately branching dendritic tree of granule cells seems optimal for this computation since larger dendritic trees favor local plasticity by isolating dendritic compartments , while reliable detection of individual dendritic spikes in the soma requires a low branch number ., Finally , we demonstrate that these parallel dendritic computations could contribute to the generation of multiple independent place fields of hippocampal granule cells . | Neurons were originally divided into three morphologically distinct compartments: the dendrites receive the synaptic input , the soma integrates it and communicates the output of the cell to other neurons via the axon ., Although several lines of evidence challenged this oversimplified view , neurons are still considered to be the basic information processing units of the nervous system as their output reflects the computations performed by the entire dendritic tree ., In the present study , the authors build a simplified computational model and calculate that , in certain neurons , relatively small dendritic branches are able to independently trigger somatic firing ., Therefore , in these cells , an action potential mirrors the activity of a small dendritic subunit rather than the input arriving to the whole dendritic tree ., These neurons can be regarded as a network of a few independent integrator units connected to a common output unit ., The authors demonstrate that a moderately branched dendritic tree of hippocampal granule cells may be optimized for these parallel computations ., Finally the authors show that these parallel dendritic computations could explain some aspects of the location dependent activity of hippocampal granule cells . | neuroscience/theoretical neuroscience, computational biology/computational neuroscience | null |
journal.pcbi.1001059 | 2,011 | A Computational and Experimental Study of the Regulatory Mechanisms of the Complement System | The complement system is pivotal to defending against invading microorganisms ., The complement proteins recognize conserved pathogen-associated molecular patterns ( PAMPs ) on the surface of the invading pathogens 1 to initiate the innate immunity response ., The complement activity also enhances adaptive immunity 2 , 3 and participates in the clearance of apoptotic cells 4 as well as damaged and altered self tissue ., The complement proteins in the blood normally circulate as inactive zymogens ., Upon stimulation , proteases in the system cleave the zymogens to release active fragments and initiate an amplifying cascade of further cleavages ., There are three major complement activation routes: the classical , the lectin and the alternative pathways 5 ., Regardless of how these pathways are initiated , the complement activity leads to proteolytic activation and deposition of the major complement proteins C4 and C3 , which induces phagocytosis , and the subsequent assembly of the membrane attack complex which lyses the invading microbes ., However , complement is a double-edged sword; adequate complement activation is necessary for killing the bacteria and removing the apoptotic cells , while excessive complement activation can harm the host by generating inflammation and exacerbating tissue injury ., Dysregulation of the balance between complement activation and inhibition can lead to rheumatoid arthritis 6 , systemic lupus erythematosus 7 , Alzheimers disease 8 and age-related macular degeneration 9 ., Since the final outcome of complement related diseases may be attributable to the imbalance between activation and inhibition 10 , manipulation of this balance using drugs represents an interesting therapeutic opportunity awaiting further investigation ., In light of this potential , complement inhibitors such as factor H and C4b-binding protein ( C4BP ) are critical since they play important roles in tightly controlling the proteolytic cascade of complement and avoiding excessive activation ., Therefore , a systems-level understanding of activation and inhibition , as well as the roles of inhibitors , will contribute towards the development of complement-based immunomodulation therapies ., Complement is usually initiated by the interaction of several pattern-recognition receptors with the surface of pathogens ., C-reactive protein ( CRP ) 11 and ficolins are two initiators of the classical and lectin pathways , which boost immune responses by recognizing phosphorylcholine ( PC ) or N-acetylglucosamine ( GlcNAc ) , respectively , displayed on the surface of invading bacteria 12 , 13 , 14 ., Recently , it was discovered that under local infection-inflammation conditions as reflected by pH and calcium levels , the conformations of CRP and L-ficolin change which leads to a strong interaction between them 15 ., This interaction triggers crosstalk between classical and lectin pathways and induces new amplification mechanisms , which in turn reinforces the overall antibacterial activity and bacterial clearance ., On the other hand , C4BP , a major complement inhibitor is synthesized and secreted by the liver ., The estimated plasma concentration of C4BP is 260 nM under normal physiological condition 16 but its plasma level can be elevated up to four-fold during inflammation 17 , 18 ., Through its α-chain 19 , 20 , C4BP modulates complement pathways by controlling C4b-mediated reactions in multiple ways 21 , 22 , 23 ., Further , C4BP has been proposed as a therapeutic agent for complement-related autoimmune diseases on the premise that mice models supplemented with human C4BP showed attenuation in the progression of arthritis 24 ., Therefore , it is important to understand the systemic effect and the underlying inhibitory mechanism of C4BP ., With this background , we constructed a detailed computational model of the complement network consisting of a system of ordinary Differential equations ( ODEs ) ., The large model size and the many unknown kinetic rate parameters lead to significant computational challenges ., Using the technique developed in 25 , we approximated the ODE dynamics as a dynamic Bayesian network 26 and used it to estimate the model parameters ., After constructing the model , we investigated the enhancement mechanism induced by local inflammation and its interplay with the inhibition mechanism induced by C4BP ., Our studies confirmed and further elucidated the previous experimental findings 15 ., Specifically , using our model we established a detailed relationship between the antimicrobial response and the strength of the crosstalk between CRP and L-ficolin as determined by various combinations of the pH and calcium levels ., We also found that C4BP prevents complement over-activation and restores homeostasis , but it achieves this in two distinct ways depending on whether the complement activity was initiated by PC or GlcNAc ., Finally , the computational model suggested that the major inhibitory effect of C4BP is to potentiate the natural decay of C3 convertase ( C4bC2a ) ., These findings regarding the role of C4BP were experimentally validated ., An earlier mathematical study 27 of the complement system focused on the classical pathway ., This study assumed the dynamics to be linear , which is a severe restriction ., A later study by Korotaevskiy et al 28 more realistically assumed the dynamics to be non-linear ., It also included the alternative pathway ., The main focus was to derive quantitative conclusions regarding the lag time of the immune response as the initial concentrations of the constituent proteins were varied ., Relative to 28 , our model additionally includes the lectin pathway and the recently identified amplification pathways induced by the crosstalk between CRP and L-ficolin 15 ., On the other hand , given our focus on the up- and down- regulation mechanisms of the complement , we do not model the alternative pathway in detail since its role is to maintain a basal level of complement activation ., Instead , this basal activity and the effects of other mechanisms such as C2 bypass 29 are implicitly captured by the kinetic parameters in our model ., Given our focus on the amplification and down-regulation mechanisms of complement , we included in our model only the key proteins in the classical and lectin pathways ., The basal activity maintained by the alternative pathway and other mechanisms are implicitly captured by the kinetic parameters in our model ., A schematic representation of the model structure is shown in Figure 1A ., The cascade of events captured by the model can be described as follows ., The classical pathway is initiated by the binding of antibodies or CRP to antigens or PAMPs ., In our model , in order to decouple the involvement of adaptive immune response , the classical pathway is triggered by the binding of CRP to PC , which is a ligand often displayed on the surface of the invading bacteria 30 , 31 ., Deposited CRP then binds to C1-complex ( formed by C1q , two molecules of C1r , and two molecules of C1s ) that is further activated ., The activated C1-complex recruits C4 leading to the cleavage of C4 to its fragments , C4b and C4a ., After binding of C2 to C4b , the same protease complexes are responsible for generating fragments , C2a and C2b , by cleaving C2 ., The C2a and C4b then form the C4bC2a complex , which is an active C3 convertase , cleaving C3 to C3a and C3b ., The formation of C3b exposes a previously hidden thioester group that covalently binds to patches of hydroxyl and amino groups on the bacterial surface 32 ., The surface-deposited C3b plays a central role in all subsequent steps of the complement cascade: ( 1 ) it acts as an opsonin that enhances the binding and leads to the elimination of bacteria by the phagocytes , ( 2 ) it induces the formation of membrane attack complex leading to the lysis of bacteria ., Since the concentration of the deposited C3 reflects the antibacterial activity of complement , we terminated our model at this step to simplify the network ., On the other hand , the lectin pathway is initiated by the binding of mannose-binding lectin ( MBL ) or ficolins to PAMPs on the pathogen surface ., In our model , we focused on the lectin pathway initiated by L-ficolin as it can interact with CRP and induce crosstalk between classical and lectin pathways ., L-ficolin recognizes various PAMPs on the bacterial surface via the acetyl group on the GlcNAc moiety 33 , 34 ., Therefore , in our model the lectin pathway was triggered by binding of L-ficolin and GlcNAc onto the bacterial surface ., Subsequently , a protease zymogen called MASP-2 is recruited and activated ., Activated MASP-2 cleaves C4 and C2 to form C4bC2a which is C3 convertase ., At this point , the classical pathway and lectin pathway merge at the cleavage step of the central complement protein , C3 , and hence constitutes the endpoint of our model ., As discovered in 15 , infection-induced local inflammation conditions ( slight acidosis and hypocalcaemia ) provoke a strong crosstalk between CRP and L-ficolin 15 ., This elicits two new complement-amplification pathways , which reinforce the classical and lectin pathways ., Since we aimed to study the complement activation and modulation under pathophysiological conditions , we included these two amplification pathways ( Figure 1A , purple ) in our model ., Infection by bacteria containing PC will induce the CRP:L-ficolin mediated amplification pathway: PC→CRP:L-ficolin→MASP2→C4→C2→C3 ., On the other hand , infection by bacteria containing GlcNAc will induce the CRP:L-ficolin mediated amplification pathway: GlcNAc→CRP:L-ficolin→C1→C4→C2→C3 ., The complement allows a rapid attack to intruding bacteria while at the same time protecting the host cells from over-activation ., C4BP , a major inhibitor of complement activation , was reported to either accelerate the decay of the convertases or aid proteolytic inactivation of key players in the pathway into inactive forms such as factor H 32 but the systemic effect of C4BP has remained unclear ., Hence , in our model , we included this major multifunctional inhibitor ., Upstream of the complement cascade , C4BP competes with C1 for the immobilized CRP 23 ., Downstream to this , C4BP binds to C4b and serves as a cofactor to the plasma serine protease factor I in the cleavage of C4b both in the fluid phase and when C4b is deposited on bacterial surfaces 21 ., In addition , C4BP is able to prevent the assembly of the C3 convertase and accelerate the natural decay of the complex 35 ., All of the above effects of C4BP are considered in our model and the relevant components are depicted as red bars in Figure 1A ., The reaction network diagram of the model is shown in Figure 1B ., Processes such as protein association , degradation and translocation are modeled with mass action kinetics and processes such as cleavage , activation and inhibition with Michaelis-Menten kinetics ., The resulting ODE model consists of 42 species , 45 reactions and 85 kinetic parameters with 71 unknown ., The details can be found in the supporting information ( Text S1 ) ., Due to the large model size and many unknown kinetic parameters , tasks such as parameter estimation and sensitivity analyses became very challenging ., Hence , we applied the probabilistic approximation technique developed by Liu et al 25 to derive a simpler model based on the standard probabilistic graphical formalism called Dynamic Bayesian Networks ( DBNs ) 26 ., Briefly , this approximation scheme consists of the following steps:, ( i ) Discretize the value space of each variable and parameter into a finite set of intervals ., ( ii ) Discretize the time domain into a finite number of discrete time points ., ( iii ) Sample the initial states of the system according to an assumed uniform distribution over certain intervals of values of the variables and parameters ., ( iv ) Generate a trajectory for each sampled initial state and view the resulting set of trajectories as an approximation of the dynamics defined by the ODEs system ., ( v ) Store the generated set of trajectories compactly as a dynamic Bayesian network and use Bayesian inference techniques to perform analysis ., A more detailed description of this construction can be found in the Methods section while we explain in the Discussion section how we fixed the number of trajectories to be generated and the maximum time point upto which the trajectories are to be constructed ., In the ODE model the PC-initiated and GlcNAc-initiated complement cascades are merged for convenience ., By suppressing these two cascades to one at a time ( by setting the corresponding expressions in the reaction equations to zero ) , we constructed two dynamic Bayesian networks; one for the PC-initiated complement cascade and the other for GlcNAc-initiated complement cascade ., The range of each variable and parameter was discretized into 6 non-equal size intervals and 5 equal size intervals , respectively ., The time points of interest were set to {0 , 100 , 200 , … , 12600} ( seconds ) ., Each of the resulting DBN approximations encoded trajectories generated by sampling the initial values of the variables and the parameters from the prior , which was assumed to be uniform distributions over certain intervals ., The quality of the approximations relative to the original ODEs dynamics was sufficiently high and the details can be found in the supporting information ( Figure S1 ) ., The values of initial concentrations and 14 kinetic parameters were obtained from literature data ( Table S1 and Table S2 ) ., To estimate the remaining 71 kinetic parameters , we generated test data by incubating human blood under normal and infection-inflammation conditions with beads coated with PC or GlcNAc followed by immunodetection of the deposited CRP , C4 , C3 and C4BP in time series ., For PC-beads , the concentration levels of deposited CRP , C4 , C3 and C4BP were measured at 8 time points from 0 to 3 . 5 h ( Figure 2A , B , red dots ) ., For GlcNAc-beads , the concentration levels of deposited MASP-2 , C4 , C3 and C4BP were also measured at 8 time points from 0 to 3 . 5 h ( Figure 2C , D , red dots ) ., To estimate unknown kinetic parameters , a two-stage DBN based method 25 was deployed ., In the first stage , probabilistic inference applied to the discretized DBN approximation was used to find the combination of intervals of the unknown parameters that have the maximal likelihood , given the evidence consisting of the test data ., As mentioned above , each unknown parameters value space was divided into 5 equal intervals and the inference method called factored frontier algorithm 35 was used to infer the marginal distributions of the species at different time points in the DBN ., We then computed the mean of each marginal distribution and compared it with the time course experimental data ., To train the model by iteratively improving fitness to data , we modified the tool libSRES 36 and used its stochastic ranking evolutionary strategy ( SRES ) , to search in the discretized parameter space consisting of 571 combinations of interval values of the unknown parameters ., The result of this first stage was a maximum likelihood estimate of a combination of intervals of parameter values ., In the second stage we then searched within this combination of intervals having maximal likelihood ., Consequently , the size of the search space for the second stage was just 1/571 of the original search space ., We used the SRES search method and the parameter values thus estimated are shown in Table S2 ., In principle , given the noisy and limited experimental data and the high dimensionality of the system , one could stop with the first stage 37 and try to work an interval of values for each parameter rather than a point value ., However , in our setting we wanted to use the ODE model too for conducting in silico experiments such as varying initial concentrations including the down and over expression of C4BP ., This would have been difficult to achieve by working solely with our current DBN approximation ., We address this point again in the Discussion section ., Figure 2A–2D shows the comparison of the experimental time course training data ( red dots ) with the model simulation profiles generated using the estimated parameters ( blue lines ) ., The model predictions fit the training data well for most of the cases ., In some cases , the simulations were only able to reproduce the trends of the data ., This may be due to the simplifications assumed by our model and further refinement is probably necessary ., We next validated the model using previously published experimental observations 15 ., In particular , normalized concentration level of deposited C3 was used to predict the antibacterial activity since C3 deposition initiated the opsonization process and the lysis of bacteria ., We first simulated the concentration level of deposited C3 at 1 h under different conditions ., We next normalized the results so that the maximum value among them equals to 95% which is the maximum bacterial killing rate reported in the experimental observations 15 ., The normalized values were then treated as predicted bacterial killing rates ., The simulation results are shown in Figure 2E and 2F as black bars ., Consistent with the experimental data ( Figure 2E , grey bars ) , our simulation showed that under the infection-inflammation conditions , the P . aeruginosa , a clinically challenging pathogen , can be efficiently killed ( 95% bacterial killing rate ) by complement whereas under the normal condition , only 28% of the bacteria succumbed ( Figure 2E , black bars ) ., Consistent with experimental data , our simulation results show that in the patient serum , depletion of CRP or ficolin induced a significant drop in the killing rate from 95% to 33% or 25% respectively , indicating that the synergistic action of CRP and L-ficolin accounted for around 40% of the enhanced killing effect ., However , in the normal serum , depletion of CRP or ficolin only resulted in a slight drop in the killing rate from 28% to 18% or 10% respectively ., Furthermore , simulating a high CRP level ( such as in the case of cardiovascular disease ) under the normal healthy condition did not further increase the bacterial killing rate ., As shown in Figure 2F , the simulation results matched the experimental data ., Thus , our model was able to reproduce the published experimental observations shown in both Figure 2E and 2F with less than 10% error ., This not only validated our model thus promoting its use for generating predictions , but also yielded positive evidence in support of the hypothesized amplification pathways induced by infection-inflammation condition ., It also suggested that the antibacterial activity can be simulated efficiently by the level of deposited C3 and this was used to generate model predictions described in later sections ., We performed local and global sensitivity analysis of the model to identify species and reactions that control complement activation during infection , and to evaluate the relative importance of initial concentrations and kinetic parameters for the model output ., To identify critical species , we first calculated the scaled absolute local sensitivity coefficients 38 for initial concentrations of major species using the COPASI tool 39 ., The model outputs were defined as the peak amplitude ( maximum activation ) and integrated response ( area under the activation curve that reflects the overall antibacterial activity ) of C3 deposition ., The results are shown in Figure 3A ., Both the peak amplitude and integrative response were strongly influenced by initial concentrations of C2 and C3 , and were mildly influenced by initial concentrations of C4BP , C1 and C4 ., In contrast , the low sensitivities of CRP , MASP-2 and L-ficolin indicate that over-expression of these proteins is unlikely to increase the antibacterial activity ., Interestingly , it was observed that the integrative response was more sensitive than the peak amplitude to the changes in the initial concentration of PC ., Since the concentration of PC is correlated to the amount of invading bacteria , this result implies that the maximum complement response level may not increase as the amount of bacteria increases but the overall response ( i . e . the area under the curve obtained by integrating the response level over time ) will be enhanced to combat the increased number of bacteria ., In order to identify critical reactions , we next computed global sensitivities for kinetic parameters ., To reduce complexity , we used the DBN approximations ., Multi-parametric sensitivity analysis ( MPSA ) 40 was performed on the DBN for PC-initiated complement cascade ( the details are presented in the Materials and Methods section ) ., The results are shown in Figure 3B ., Strong controls over the whole system are distributed among the parameters associated with the immobilisation of C3b with the surface , interaction between CRP and L-ficolin , cleavage of C2 and C4 , and the decay of C3 convertase ( see Figure 1B , reactions labeled in red ) ., The sensitivity of reactions associated with C3 , C2 and C4 is consistent with the local sensitivity analysis , which highlighted the significant role of major complement components ., The high sensitivity of interaction of CRP and L-ficolin confirms that the overall antibacterial response depends on the strength of the crosstalk between the classical and lectin pathways ., In addition , since the decay of C3 convertase is one of the regulatory targets of C4BP , the sensitivity of the system to a change in the rate of decay of C3 convertase suggested that the regulatory mechanism by C4BP plays an important role in complement ., Since the critical reactions identified are common in PC- and GlcNAc-initiated complement cascades , MPSA results using the other DBN will produce similar results and hence this analysis was not performed ., We next focused our investigation on the enhancement mechanism by the crosstalk and the regulatory mechanism by C4BP ., Under infection-inflammation conditions where PC-CRP:L-ficolin or GlcNAc-L-ficolin:CRP complex is formed , the amplification pathways are triggered ., Model simulation showed that if C1 and L-ficolin or CRP and MASP-2 competed against each other , the antibacterial activity of the classical pathway or lectin pathway might be deprived of the amplification pathways ( see Figure S2 ) ., Therefore , in order to achieve a stable enhancement , C1 and L-ficolin ( or CRP and MASP-2 ) must simultaneously bind to CRP ( or L-ficolin ) ., Further , the abilities of CRP and L-ficolin to trigger subsequent complement cascade were not affected by the formation of this complex ., This is consistent with the previous experimental observation that two amplification pathways co-exist with the classical and lectin pathways 15 ., According to 15 , slight acidosis and mild hypocalcaemia ( pH 6 . 5 , 2 mM calcium ) prevailing at the vicinity of the infection-inflammation triggers a 100-fold stronger interaction between CRP and L-ficolin compared to the normal condition ( pH 7 . 4 , 2 . 5 mM calcium ) ., This can be explained by the fact that the pH value and calcium level influence the conformations of CRP and L-ficolin which in turn govern their binding affinities ., Therefore , the overall antibacterial response which is influenced by the binding affinity of CRP and L-ficolin will be sensitive to the pH value and calcium level ., To confirm this and further investigate the effects of pH and calcium on the antibacterial response , we simulated the complement system under different pH and calcium conditions ., Based on the previous biochemical analysis 15 , we first estimated functions using polynomial regression to predict the binding affinity of CRP and L-ficolin for different pH values and calcium levels ( Figure 4A , B , right panel ) ., In the right panels of Figure 4A and 4B , the reported binding affinities 15 were normalized and are shown as dots ., By curve fitting the dots , we estimated polynomial functions that can be used to predict the binding affinity ., The curves of these functions are shown in red ., We then simulated the C3 deposition dynamics using the predicted binding affinities at pH ranging from 5 . 5 to 7 . 4 in the presence of 2 mM and 2 . 5 mM calcium ., The simulation time was chosen to be 3 . 5 h which is the time frame of the response peaks ., The results are shown in Figure 4A and 4B ., Under both 2 mM and 2 . 5 mM calcium conditions , decreasing pH induces not only the increase of the peak amplitude ( maximum activation ) but also hastens the peak time ( time of maximum activation ) ., To further compare the effects of the two calcium levels , the dose-response curves were generated as shown in Figure 4C ., The antibacterial response was predicted by simulating the system for 1 . 5 h ., At 2 mM calcium ( blue curve ) , the antibacterial response was clearly greater than at 2 . 5 mM calcium ( pink curve ) indicating that slight hypocalcaemia enhanced the antibacterial activity in a stable manner ., In addition , the pH-responses were reaching saturation levels when pH was near 5 . 5 ( Figure 4C ) , implying that the undesirable complement-enhancement by extreme low pH condition can be avoided ., This also suggests that the saturation of the pH-response was influenced by the calcium level in the milieu ., We next investigated the complement regulation by the major inhibitor , C4BP , under infection-inflammation conditions ., We varied the initial concentration of C4BP and simulated the PC- and GlcNAc- initiated complement under infection-inflammation conditions ., The simulation time was chosen to be 5 h which is slightly beyond the largest time point of our training experimental data ., The predicted effects of the initial concentration of C4BP on the antibacterial response in terms of C3 deposition are shown in Figure 5A and 5B ., For PC-initiated complement activation , when the starting amount of C4BP was perturbed around the normal level of 260 nM 16 , increasing C4BP level only delayed the peak time but did not decrease the peak amplitude significantly ., In contrast , reducing the initial C4BP level clearly hastened the complement activation and maximized the activity ., Interestingly , the GlcNAc-initiated complement activation ( Figure 5B ) behaved differently from the PC-mediated complement activation ( Figure 5A ) ., Around the normal level of 260 nM , perturbing the initial C4BP changed the maximum activity but did not affect the peak time , suggesting that C4BP plays distinct roles in regulating the classical and lectin pathways ., To experimentally verify the model predictions , we perturbed the initial amount of C4BP in the patient sera by, ( i ) spiking with purified C4BP ( high C4BP ) and, ( ii ) reducing it by immunoprecipitation ( low C4BP ) ., The resulting C4BP levels in the normal and patient sera are shown in Figure S3 ., The sera were then incubated with PC- or GlcNAc-beads to initiate complement ., Unaltered serum served as the normal control ., The time profiles of the deposited C4BP level was measured over 4 h using Western blot ( Figure 5C ) ., Comparing the kinetic profiles in the C4BP deposition initiated by both PC and GlcNAc , we observed the following order of peak time: high C4BP>normal C4BP>low C4BP , indicating that the pre-existing initial level of C4BP was indeed the driving force controlling the deposition of complement components onto the simulated bacterial surface ., We then measured the time profiles of deposited C3 ., Figure 5D shows that with PC-beads , high C4BP sera induced an early peak and low C4BP delayed the peak of C3 deposition ., The peak amplitude for all three conditions was at a similar level ., These observations are consistent with the simulation results shown in Figure 5A ., With GlcNAc-beads , reducing C4BP led to a slight increase in the peak height although the peak coincided with the normal condition ., In contrast , spiking the sera ( high C4BP ) delayed and lowered the peak amplitude of C3 deposition ., Thus the experimental results broadly agree with our model predictions presented in Figure 5B ., We next investigated how C4BP mediates its inhibitory function ., As shown in Figure 1A , the inhibitory effects of C4BP target different sites in complement:, ( a ) binding to CRP and blocking C1 ,, ( b ) preventing the formation of C4bC2a by binding to C4b ,, ( c ) acting as a cofactor for factor I in the proteolytic inactivation of C4b , and, ( d ) accelerating the natural decay of the C4bC2a complex , which prevents the formation of C4bC2a and disrupts already formed convertase ., To identify the dominant mechanism , we employed in silico knockout of the reactions involved for each mechanism and performed simulations ., Figure 6A–6D shows the model predictions ., Among the four inhibitory mechanisms , only the knockout of reaction, ( d ) significantly enhanced the complement activation suggesting that facilitating the natural decay of C4bC2a ( C3 convertase ) is the most important inhibitory function of C4BP ., This is consistent with our previous observations derived from sensitivity analysis , which identified the decay of C3 convertase as a critical reaction ., In addition , as the inhibitory effect of reaction, ( d ) is stronger than others , knocking out reaction, ( a ) and, ( b ) can even reduce the complement activity , which is counter-intuitive and emphasizes the significance of the systems-level understanding ., To confirm our hypothesis that the major inhibitory role of C4BP relies on accelerating the decay of C3 convertase , we measured the C4 cleavage at different time points ., Figure 6E ( black triangles ) indicates the inactive C4b fragments presented from the time points of 20 , 30 and 90 min under high , normal , and low C4BP conditions , suggesting that C4BP aided cleavage and inactivation of C4b , and thereby caused the natural decay of the C4bC2a ., Here , we developed an ODE-based dynamic model for the complement system accompanied by DBN-based approximations of the ODEs dynamics to understand how the complement activity is boosted under local inflammation conditions while a tight surveillance is established to attain homeostasis ., Previously published models of complement system have focused on the classical and alternative pathways 27 , 28 ., Our model includes the lectin pathway and more interestingly , the recently identified amplification pathways induced by local inflammation conditions 15 ., It also encompasses the regulatory effects of C4BP in the presence of enhanced complement activity ., The ODE model incorporated both the PC-initiated and GlcNAc-initiated complement together for convenience ., By setting the corresponding expressions to zero one at a time , two DBN approximations were then derived; one for the PC-initiated complement cascade and the other for GlcNAc-initiated complement cascade ., For constructing the DBN approximation from an ODE model , one needs to fix , the maximal time point upto which each trajectory is to be explored and , the number of trajectories to be generated ., is set to be suitably beyond the largest time point for which experimental data is available ., In the present study 3 . 5 h , is the largest time point of our training experimental data ., Based on this we set to be 5 h ., After constructing the model , we simulated the system upto 10 h and found no relevant dynamics after 3 . 5 h ., As for the choice of , the number of trajectories , ideally one would like to specify the acceptable amount of error between the actual and the approximated dynamics and use to determine ., This is however difficult to achieve due to the following: The dynamic Bayesian network we construct is a factored Markov chain ., It approximates the idealized Markov chain induced by the ODEs dynamics ., This i | Introduction, Results, Discussion | The complement system is key to innate immunity and its activation is necessary for the clearance of bacteria and apoptotic cells ., However , insufficient or excessive complement activation will lead to immune-related diseases ., It is so far unknown how the complement activity is up- or down- regulated and what the associated pathophysiological mechanisms are ., To quantitatively understand the modulatory mechanisms of the complement system , we built a computational model involving the enhancement and suppression mechanisms that regulate complement activity ., Our model consists of a large system of Ordinary Differential Equations ( ODEs ) accompanied by a dynamic Bayesian network as a probabilistic approximation of the ODE dynamics ., Applying Bayesian inference techniques , this approximation was used to perform parameter estimation and sensitivity analysis ., Our combined computational and experimental study showed that the antimicrobial response is sensitive to changes in pH and calcium levels , which determines the strength of the crosstalk between CRP and L-ficolin ., Our study also revealed differential regulatory effects of C4BP ., While C4BP delays but does not decrease the classical complement activation , it attenuates but does not significantly delay the lectin pathway activation ., We also found that the major inhibitory role of C4BP is to facilitate the decay of C3 convertase ., In summary , the present work elucidates the regulatory mechanisms of the complement system and demonstrates how the bio-pathway machinery maintains the balance between activation and inhibition ., The insights we have gained could contribute to the development of therapies targeting the complement system . | The complement system , which is the frontline immune defense , constitutes proteins that flow freely in the blood ., It quickly detects invading microbes and alerts the host by sending signals into immune responsive cells to eliminate the hostile substances ., Inadequate or excessive complement activities harm the host and may lead to immune-related diseases ., Thus , it is crucial to understand how the host boosts the complement activity to protect itself and simultaneously establishes tight surveillance to attain homeostasis ., Towards this goal , we developed a detailed computational model of the human complement system ., To overcome the challenges resulting from the large model size , we applied probabilistic approximation and inference techniques to train the model on experimental data and explored the key network features of the model ., Our model-based study highlights the importance of infection-mediated microenvironmental perturbations , which alter the pH and calcium levels ., It also reveals that the inhibitor , C4BP induces differential inhibition on the classical and lectin complement pathways and acts mainly by facilitating the decay of the C3 convertase ., These predictions were validated empirically ., Thus , our results help to elucidate the regulatory mechanisms of the complement system and potentially contribute to the development of complement-based immunomodulation therapies . | computational biology/systems biology, immunology/innate immunity | null |
journal.pntd.0005649 | 2,017 | Sterol 14α-demethylase mutation leads to amphotericin B resistance in Leishmania mexicana | The leishmaniases are a complex of diseases caused by parasitic protozoa of the genus Leishmania which are transmitted between people via the bite of an infected sandfly 1 ., The specific disease caused by the parasites depends upon which Leishmania species is responsible and ranges from a self-limiting cutaneous form , through a mucocutaneous disease and a frequently fatal visceral form 2 ., Control is largely dependent upon the use of chemotherapy ., In recent years the polyene amphotericin B ( AmB ) has emerged as the treatment of choice where available , particularly in the liposomal formulation which abrogates some of the toxic effects associated with the parent compound itself 3 ., The specificity of AmB relates to its mode of action being mediated through a binding to the membrane sterol ergosterol , which is the primary sterol of fungal and Leishmania membranes , while binding with less avidity to cholesterol 3 , the principal sterol of mammalian host membranes ., It was suggested that AmB molecules polymerise at membranes where they bind , forming pores that cause membrane leakage to various ions and this has been proposed as a key cause of death 4 , although binding to ergosterol alone is sufficient to cause death in fungi 5 , 6 ., AmB in a liposomal formulation , AmBisome , has emerged as a treatment of choice because of the enhanced efficacy of the drug against macrophage-resident Leishmania parasites and the accompanying reduction in host toxicity 7 , 8 ., Several trials using the drug as a combination with other leishmanicides 9 indicate that AmB containing combinations offer promise for future therapies 10 ., Other trials 11 have indicated that a single injection of AmBisome is efficacious , particularly in India where other drugs , including pentavalent antimony 12 and miltefosine 13 are suffering from treatment failure and increasing resistance ., The fact that the incidence of resistance to AmB in fungi has been slow to emerge , in spite of over 50 years of use 14 , has underpinned a belief that the fitness costs associated with any resistance might protect against the problem 15 ., However , there are increasing reports of AmB resistance in fungi 16–18 ., Moreover , several reports of AmB treatment failure have been reported in leishmaniasis patients in India 19 , 20 and in immunocompromised patients in France 21 and resistance to the drug has been reported to occur in at least one field isolate already 19 ., Resistance in that case was proposed to relate to several phenotypic changes to the parasite , notably a change in sterol metabolism 19 and increase in defence against oxidative stress 22 ., In common with several reports in L . mexicana 23 and L . donovani 19 , 24 , selection of resistance was associated with the replacement of ergostane-type sterols with cholestane-type sterols , the latter being less avid binders of AmB 3 ., Other studies into changes occurring in selected AmB resistance in Leishmania point to alterations in enzymes of cellular thiol 24 , 25 and ascorbate 22 metabolism leading to an enhanced resistance to oxidative stress being associated with selection ., Although changes to C24-Δ-sterol methyl transferase gene expression suggested a possible genetic marker for resistance 19 , 26 direct corroboration is lacking , and no specific gene mutations have yet been described that correlate unequivocally with resistance ., Understanding molecular mechanisms of drug resistance provides potential biomarkers to assess the spread of resistance and can also offer routes to slow the emergence of resistance or even bypass the problem ., Due to the lack of economic incentivisation for new drug development , it is essential to retain existing drugs for neglected tropical diseases , such as the leishmaniases , if we are to achieve aims of bringing the disease under control ., Here we use the complementary high throughput data approaches of metabolomics and whole genome sequencing to reveal a gene whose mutation causes resistance to AmB in Leishmania mexicana ., Promastigotes of L . mexicana strain MNYC/BZ/62/M379 were cultured in Homem ( GIBCO ) medium 27 supplemented with 10% foetal bovine serum—Gold ( FBS ) ( PAA Laboratories GmbH ) starting at a density of 1 x 105 cells/ml and maintained at 27°C , passaging once every 72 hours ., The cells were selected for AmB resistance by increasing concentrations of the drug , initially exposing cells to 0 . 0135 μM of AmB ( Sigma-Aldrich ) with stepwise doubling of the drug concentration to a final concentration of 0 . 27 μM ., Cells able to grow in the presence of the drug were cloned under drug pressure by limiting dilution to 1 cell/ml in 20 ml of growth medium and plated out into 96-well plates ., Susceptibility of the cells to various drugs was determined using an adaptation of the Alamar Blue assay 28 ., A starting density of 1 × 106 cells/ml were incubated at 27°C in the presence of various drug concentrations for 72 hours in a 96-well microtiter plate ., Resazurin ( Sigma-Aldrich ) in 1× phosphate-buffered saline ( PBS ) ( Sigma-Aldrich ) pH 7 . 4 solution was added to a concentration of 44 . 6 μM and cells incubated for a further 48 hours ., The fluorescence of the reacted dye was measured on a FLUOstar OPTIMA ( BMG LabTech , Germany ) spectrometer set at excitation and emission wavelengths of 530 nm and 590 nm , respectively ., The drugs used in the susceptibility assays , unless stated otherwise , were bought from Sigma-Aldrich ., To assess the sensitivity to H2O2 , wild-type and derived AmB resistant cells at a starting density of 2 × 106 cells/ml were exposed to 20 μM , 200 μM , 500 μM and 1 mM of H2O2 29 in growth medium in 6-well plates ., The response by the two cell lines to H2O2 were compared by observation under a light microscope at different time points over 72 hours ., An alternative test for H2O2 sensitivity involved glucose oxidase as described previously 30 ., Briefly , 180 μl of cells at 5 x 105cells/ml were plated into a 96-well plate and 20 μl of glucose oxidase solution ( Sigma-Aldrich ) were added in varying concentrations then tested using the AlamarBlue assay described above ., The cell body length of L . mexicana promastigotes was determined using the SoftWoRx 5 . 5 software on a DeltaVision Applied Precision Olympus IX71 microscope ., Smears of cells from late log phase culture were spread onto a microscope slide ., The cells were fixed in absolute methanol overnight at -20°C then rehydrated with 1 ml of 1 × PBS for 10 minutes ., 50 μl of PBS containing 1 μg/ml 4 , 6-diamidino-2-phenylindole ( DAPI ) ( Sigma-Aldrich ) and 1% 1 , 4-Diazobicyclo- ( 2 , 2 , 2 ) octane ( DABCO ) ( Sigma-Aldrich ) were added to stain the cells ., The cell length was measured from the anterior end to the posterior end of the cell body ., Cells were grown to mid-log phase and 1 × 108 cells for each sample collected and metabolism was quenched rapidly by cooling them to 4°C in a dry ice/ethanol bath while mixing vigorously to avoid freezing and possible cell lysis 31 ., Cells were separated from medium by centrifugation at 1 , 250g for 10 minutes at 4°C and 5 μl of supernatant was used for spent medium analysis ., Metabolites were extracted from the cell pellet by addition of 200 μl of chloroform-methanol-water ( 1:3:1 ) solution and mixed vigorously at 4°C for 1 hour ., The metabolites were separated from the cell debris by centrifugation at 13 , 000g for 5 minutes at 4°C and the samples were stored under argon gas at -80°C until analysis ., Separation and mass detection of the metabolites was performed according to 32 , using the DionexUltiMate3000 Liquid chromatography system using a SeQuant ZIC-HILIC column coupled to the Orbitrap Exactive mass spectrometer at Glasgow Polyomics , University of Glasgow ., Raw data was processed and analyzed using the mzMatch 33 and IDEOM 34 software platforms ., Metabolite identifications were given at Level 2 according to the Metabolomics Standards initiative ( MSI ) where accurate masses and predicted retention times were used to yield putative annotations but when retention times of authentic standards were available , the identification should be considered as Level 1 35 ., Metadata to support the identification of each metabolite is available in the IDEOM file for each study ( S1 Table ) ., It is important to note that many of the metabolite names given in the IDEOM file are generated automatically as the software provides a best match to database entries of the given mass and formula ., In the absence of additional information these must be considered as putatively annotated hits; the confidence score in the column adjacent to that hit serves as a guide to this ., Clearly it is beyond the scope of any study to provide authenticated annotations to many hundreds of detected compounds , but the full datasets are included in the spirit of open access data ., Mid log phase cells were taken and washed in PBS before 1 . 5 ml 25% KOH in 60% ethanol was added to each 100 mg cells in glass tubes ., Samples were incubated at 85°C for one hour then an equal volume of n-heptane was added ., Samples were vortexed then incubated at room temperature for 10 min ., The top layer containing the sterols was transferred to a new glass vial for analysis ., One microlitre of heptane extract sample was injected into a Split/Splitless ( SSL ) injector at 270°C using splitless injection ( 1 minute ) into Trace 1310 gas chromatograph ( Thermo Scientific ) ., Helium carrier gas at a flow rate of 1 . 2 ml/min was used for separation on a TraceGOLD TG-5SILMS 30 m length with 5 m safeguard × 0 . 25 mm inner diameter × 0 . 25 μm film thickness column ( Thermo Scientific ) ., The initial oven temperature was held at 50°C for 2 min ., Separation of sterols was performed using a gradient of 20°C/min from 50 to 325°C with an 8 . 5 minutes final temperature hold at 325°C ., Eluting peaks were transferred through an auxiliary transfer temperature of 275°C into a Q-Exactive GC mass spectrometer ( Thermo Scientific ) ., Electron ionisation ( EI ) was at 70 eV energy , with an emission current of 50 μA and an ion source of 230°C ., A filament delay of 3 . 5 minutes was used to prevent excess reagents from being ionised ., Full scan accurate mass EI spectrum at 60 , 000 resolution were acquired for the mass range 50 to 750 m/z ., Peak detection used the Xcalibur software ( Thermo Scientific ) ., Masses were compared to those in the NIST/EPA/NIH Mass Spectral Library ( EI ) ., Cells for genomic DNA extraction were grown to mid-log phase in a 10 ml culture and harvested by centrifugation at 1 , 250g for 10 minutes and washed once in 1 × PBS ., The cells were re-suspended in 500 μl NTE buffer ( 10 mM Tris-HCl pH 8 . 0; 100 mM NaCl; 5 mM EDTA ) to which 25 μl of 10% SDS and 50 μl of 10 mg/ml RNase A ( Sigma-Aldrich ) were added and warmed to 37°C ., The solution was mixed by inverting and incubated at 37°C for 30 minutes ., After addition of 25 μl of 20 mg/ml pronase ( Sigma-Aldrich ) the lysates were incubated at 37°C overnight ., The samples were then extracted twice with phenol:chloroform:isoamyl alcohol ( 25:24:1 ) ( Sigma-Aldrich ) and chloroform , with mixing for 5 minutes between extraction steps ., The aqueous phase was obtained after centrifugation at 16 , 000g for 10 minutes and the DNA was precipitated with absolute ethanol and washed once with 70% ethanol ., The DNA was dried in the fume hood and after being dissolved in water the concentration was determined using a NANODROP 1000 spectrophotometer ( Thermo Scientific ) ., Paired-end samples of the genomic DNA for the progenitor wild-type and derived AmB resistant cells were sequenced using Illumina GAIIx next generation DNA sequencing platform and analysed at Glasgow Polyomics , University of Glasgow ., All DNA sequence information is deposited at the European Nucleotide Archive ( ENA ) under project number PRJEB10872 ., The expression vector pNUS-HnN for Crithidia fasciculata and Leishmania 36 was used to express the WT sterol 14α-demethylase gene ( LmxM . 11 . 1100 ) fused to the His-tag at the N-terminus in AmB resistant L . mexicana ., The vector pNUS-GFPcN was used for expression of both WT and N176I CYP51 with the Green Fluorescent Protein ( GFP ) tag at the C-terminus in both resistant and WT L . mexicana ., The genes were amplified by PCR using Phusion High-Fidelity DNA polymerase ( New England Biolabs ) ., Primers for pNUS-HnN incorporating NdeI and XhoI restriction sites ( underlined ) for WT CYP51 were forward 5 GCATATGATGATCGGCGAGCTTCTCC3 and reverse 5CTCGAGCTAAGCCGCCGCCTTCT3 ., For expression of the WT and N176I CYP51 in pNUS-GFPcN , the forward and reverse primers were 5’CATATGATGATCGGCGAGCTTCTCCT3’ and 5’AGATCTAGCCGCCGCCTTCTTC3’ , respectively , with NdeI and BglII restriction sites ., Sterol C14-reductase ( LmxM . 31 . 2320 ) was expressed in pNUS-GFPcN using forward 5’CATATGATGGCAAAACGCAGAGGTACTG3’ and reverse 5’AGATCTGTATATGTACGGGAACAGCC3’ primers , respectively ., The genes were initially sub-cloned into pGEM-T Easy vector ( Promega ) and multiplied in XL1 Blue E . coli competent cells ( Promega ) prior to cloning into the pNUS vectors ., The presence of the gene fragments was confirmed by their PCR amplification using vector-specific primers designed from the vector sequences on http://www . ibgc . u-bordeaux2 . fr/pNUS/index . html ., Thus , presence of WT CYP51 in the pNUS-HnN was verified with forward 5CATCATCATCATCACAGCAGC3 and reverse 5GTCGAAGGAGCTCTTAAAACG3 primers , while the presence of both the WT and N176I CYP51 in pNUS-GFPcN was verified with the forward 5’TATCTTCCACTTGTCAAGCGAAT3’ and reverse 5’CCCATTCACATCGCCATCCAGTTC3’ primers ., Similarly , the presence of these genes was confirmed by PCR in DNA extracted from the transfectants ., The presence of mutated chromosomal CYP51 in AmB resistant L . mexicana expressing the WT CYP51 gene was confirmed by PCR amplification using a forward primer ( 5CGCGAAATAGATATAAAGCACACG3 ) starting from 43 bp upstream of the start codon of CYP51 or 569 bp from the point mutation and a reverse primer ( 5TCGCGAGCGATGATAATCTCG3 ) starting 213 bp downstream the mutated base resulting in a 788 bp PCR fragment ., PCR amplification fragments were sequenced at Eurofins MWG Operon , Germany and aligned using CLC workbench Genomics software ., All primers were purchased from Eurofins MWG Operon , Germany ., L . mexicana promastigotes were grown to log phase and 1 × 107 cells were harvested and washed then re-suspended in 100 μl transfection buffer ( 90 mM NaPO4 pH7 . 3; 5 mM KCl; 50 mM HEPES pH7 . 3; 0 . 15 mM CaCl2 ) and added to 10 μg of plasmid DNA in a cuvette before transfection with an Amaxa biosystems NucleofectorII ( Lonza ) using program U-033 ., The cells were incubated on ice for 10 minutes and then transferred into pre-warmed 10 ml of Homem supplemented with 10% FBS-Gold and left to recover for 18 hours at 27°C ., G418 disulfate salt ( Sigma-Aldrich ) at 50 μg/ml was added to select cells carrying the plasmids ., Clones of the selected cells were obtained in the presence of 50 μg/ml G418 in growth medium by limiting dilution ., Immunofluorescence microscopy was performed with WT and AmB resistant cell lines expressing GFP-CYP51 or tagged sterol reductase ( GFP-SR ) from episomal vectors ., 200 μl of mid-log phase cells were collected and washed two times with PBS , fixed in 1% formaldehyde ( methanol-free , Thermo Scientific ) for 30 minutes ., Triton X-100 ( Sigma-Aldrich ) was added up to final concentration 0 . 1% and incubated for 10 minutes , afterwards , glycine was added to the final concentration of 0 . 1 M and incubated for an additional 10 minutes ., Cells were centrifuged , resuspended in PBS , spread on microscopy slides and left to dry ., Slides were washed with PBS and blocked with PBS , 0 . 1% Triton X-100 , 0 . 1% BSA ( Sigma-Aldrich ) for 10 minutes ., Primary antibody against the ER specific chaperone BiP 37 , a gift from Professor J . Bangs ( University of Buffalo , New York ) , was applied in dilution of 1:5000 , overnight , at 4°C ., Subsequently , slides were washed three times with PBS and incubated with secondary anti-rabbit Alexa Fluor antibody ( Molecular Probes ) ., Following 1 hour incubation , slides were washed three times with PBS , dried and mounted with 5 μM 4’ , 6-diamidino-2-phenylindole ( DAPI ) ., Microscopy was performed using Axioscope , Volocity software and processed with ImageJ software ., For mitochondrial staining , cells were incubated with 100 nM MitoTracker red ( Molecular Probes ) at 25°C for 30 minutes ., Subsequently , cells were fixed as described above and mounted with DAPI ., AmB resistant L . mexicana promastigotes were selected by stepwise increase in drug concentration in culture medium over 18 passages stretching over six months ., During this period , a 23-fold increase ( P = 0 . 0007 ) in EC50 value above that of the wild-type ( WT ) was observed ( Fig 1 and S1 Fig ) ., Sustained growth in a drug concentration above 0 . 27 μM could not be achieved ., The acquired AmB resistance was stable over at least 15 passages in drug free medium ., There was no appreciable difference in the growth phenotype between the resistant and the WT cells , although during the process of resistance induction , the derived AmB resistant cells required at least five passages of adaptation to a given drug concentration before they would grow at similar rates to WT ., The cells showing the highest resistance level had a significantly reduced cell body length compared to the WT cells ( P < 0 . 0001 ) ., The late log-phase WT cells and the resistant clone had average cell body lengths of 11 . 16 ± 0 . 19 μm ( n = 126 ) and 9 . 86 ± 0 . 16 μm ( n = 126 ) , respectively ., The AmB resistant cells exhibited mild cross-resistance to potassium antimony tartrate ( PAT ) and miltefosine with fold change in EC50 values of 2 . 9 and 3 . 9 representing significant differences ( P = 0 . 0005 and P < 0 . 0001 , respectively ) to the WT ( Table 1 and S2 Fig ) ., A marginal 1 . 7-fold increase in the EC50 value ( P = 0 . 0007 ) to ketoconazole ( an inhibitor of sterol synthesis at the sterol 14α-demethylase step ) was also observed in the AmB resistant line ., Interestingly , the AmB resistant cells were found to be more susceptible to pentamidine ( P = 0 . 0001 ) with decreases in EC50 values of 13 . 3-fold ., We also tested the effect of various reagents causing oxidative stress ., Exposure of both cell lines to methylene blue , a stress inducing agent 39 , showed that AmB resistant cells were far more susceptible to this agent with EC50 values of 0 . 117 ± 0 . 001 μM against 4 . 20 ± 0 . 22 μM for WT ( P < 0 . 0001 ) ., Addition of 500 μM H2O2 directly to cells induced swelling resulting in their assuming rounded shapes , and sluggish to no movement ., Resistant cells recovered from exposure more slowly than WT cells ( as judged by inspection of flagellar motility ) and by 72 hours following exposure reached average densities of 4 × 106 cell/ml whilst WT cells were at 9 × 106 cells/ml ., Because H2O2 is labile , we also tested the effect of glucose oxidase in medium ., Glucose oxidase produces H2O2 continuously 30 and the concentration of enzyme added to medium can , therefore , act as a surrogate for quantitation of susceptibility to the peroxide ., 4 . 5 mU/ml of glucose oxidase were required to inhibit growth of WT cells by 50% whilst the EC50 for the resistant cell line was just 1 . 8 mU/ml , confirming the increased sensitivity to stress of the resistant cell line ., Using an untargeted liquid chromatography-mass spectrometry ( LC-MS ) metabolomics approach we compared the two cell lines ., Principal components analysis ( PCA ) revealed the WT and resistant lines to have clear differences ( Fig 2A ) ., Among the most significant changes were alterations around the sterol metabolic pathway ., Previous studies in both L . donovani 19 , 24 and L . mexicana 23 have also identified changes to sterols ( specifically an increase in cholesta-5 , 7 , 24-trien-3β-ol in L . donovani and 4 , 14-dimethyl-cholesta-8 , 24-dienol and other methyl sterols in L . mexicana ) ., A metabolite of m/z 394 . 32 putatively identified as ergosta-5 , 7 , 22 , 24 ( 28 ) -tetraen-3β-ol was diminished in resistant cells compared to WT cells ( Fig 2B and S1 Table ) while a metabolite of m/z 410 . 35 putatively identified as 4 , 4-dimethylcholesta-8 , 14 , 24-trien-3β-ol and another with m/z 426 . 35 consistent with 4α-formyl-4β-methylcholesta-8-24-dien-3β-ol were more abundant in resistant cells compared to WT cells ( Fig 2B and S1 Table ) ., While the LC-MS approach taken allows comprehensive coverage of the metabolome and is thus ideally suited to initial identification of those areas of metabolism changing in response to biological perturbation , the hydrophilic interaction liquid chromatography ( HILIC ) -based liquid chromatography platform is not suitable for separation and robust identification of individual lipids ., Having identified that changes in sterol metabolism were key , we adopted a gas chromatography ( GC ) -MS approach since this methodology had been previously applied to the identification of sterol metabolism in L . mexicana 40 ., Fig 3 shows chromatograms obtained from the GC-MS , and identities of the detected peaks are indicated in Table 2 based on matches with the NIST library ( https://www . nist . gov/srd/nist-standard-reference-database-1a-v14 ) ., The major difference between WT and AmB resistant cells is depletion of peak 6 representing ergosterol ( the most abundant sterol in WT ) and concomitant increase in peak 5 , corresponding with 14-methylergosta-8 , 24 ( 28 ) -dien-3β-ol in the resistant cell line ., Peak 4 , not detected in WT , is abundant in resistant cells , and corresponds to 4 , 4-dimethylcholesta-8 , 14 , 24-trien-3β-ol , a product of the sterol 14α-demethylase reaction ( note , however , that isomers of this compound exist that we cannot distinguish ) ., In addition , peaks 3 and 8 are increased 50-fold and 80-fold , respectively , in the AmB resistant cell line , putatively identified as ergosta-5 , 24 ( 28 ) -dien-3β-ol and 4 , 14-dimethylergosta-8 , 24 ( 28 ) -dien-3β-ol ., The level of cholesterol was unchanged because it is acquired from the medium rather than synthesised 41 ., The detected sterols were mapped onto a pathway based on work by Roberts et al . , 42 and metacyc . org database ( Fig 4 ) ., Overall , components from the upper part of the pathway are accumulated whereas intermediates of the downstream steps are decreased ., Replacement of ergostane-type sterols in WT cells with cholestane-type sterols in resistant derivatives , are similar to those noted in L . donovani promastigotes selected for resistance 24 and also a field isolate of L . donovani from a refractory patient 19 ., In a separate study , L . mexicana selected for resistance had also lost ergosterol , but in this case the sterol species that accumulated was 4 , 14-dimethylcholesta-8 , 24-dienol 23 , indicating that there may be several distinct ways whereby loss of ergosterol synthesis can be achieved , with the accompanying accumulation of other sterols ., In principle , the loss of any enzyme in the ergosterol synthetic pathway could lead to loss of production of that sterol and acquisition of resistance to AmB ., For example , Pourshafie et al . 26 point to possible mutations in C-24-Δ-sterol-methyltransferase causing the increase in cholesta-5 , 7 , 24-trien-3β-ol they identified ., The related parasite Trypanosoma brucei accumulates exogenous cholesterol for membrane biogenesis 43 and L . donovani deficient in a cytochrome P450 enzyme related to sterol 14α-demethylase , termed CYP5122A1 , produce less ergosterol than WT cells and grow less well but recover WT growth rates if medium is supplemented with exogenous ergosterol 44 ., We therefore tested whether addition of exogenous ergosterol would accumulate in membranes of our resistant cells and re-sensitise them to AmB ., However , addition of exogenous ergosterol ( 7 . 6 μM , as used in reference 44 ) for 5 passages prior to testing drug efficacy , failed to re-sensitise , rather it further reduced sensitivity ( 2 . 84-fold increase in EC50 value for AmB ( P = 0 . 0004 ) ) which could be attributed to the drug binding ergosterol in medium ., Whole genome sequencing of the resistant line and its WT progenitor ( passaged in parallel during the course of resistance selection ) resulted in more than 50% of the reads being aligned in both cases to the reference L . mexicana MHOM/GT/2001/U1103 genome ( 17 , 138 , 430 and 15 , 392 , 124 reads out of totals of 32 , 238 , 036 and 27 , 514 , 220 reads which were obtained for the WT and the AmB resistant clone , respectively , were aligned ) ., Comparison of read coverage depth between AmB resistant and WT cells showed variations in chromosome copy numbers ., An extra copy was observed to have been gained for chromosomes 05 , 19 , 22 , and 27 and there was loss of a copy for chromosomes 12 and 17 of the AmB resistant cells as compared to the parental WT cells ., In addition , a total of 5 , 047 single nucleotide polymorphisms ( SNP ) distinguish the WT and resistant lines ., Since the metabolomic data indicated changes in sterol metabolism we looked specifically for any genetic alterations in genes encoding enzymes of this pathway in AmB resistant cells ., A single homozygous SNP was found among genes encoding the enzymes of the ergosterol pathway , namely in sterol 14α-demethylase ( EC 1 . 14 . 13 . 70 ) , where a non-synonymous mutation from A to T results in an amino acid alteration from asparagine to isoleucine ( N176I ) in this enzyme ., Lines of several Candida species resistant to AmB also had mutations to sterol 14α-demethylase ( ERG11 ) 45–47 which lead to a cessation of ergosterol production ., These cells , however , were also resistant to azoles that target the demethylase whilst our Leishmania cell line was not ., Since the Candida lines also accumulate lanosterol ( the substrate of sterol 14α-demethylase ) whilst our Leishmania cell line accumulated the enzyme’s product , we conclude that the Candida mutants have lost enzyme activity whilst our mutant retains demethylase activity but fails to provide the product into later steps of the ergosterol synthetic pathway ., Interestingly , 4 , 4-dimethylcholesta-8 , 14 , 24-trien-3β-ol is the product of the sterol 14α-demethylase reaction and it was increased in the resistant cells , indicating that the enzyme itself was functional , but that its metabolic product accumulates and no longer feeds the remainder of the ergosterol pathway ., Modelling of the site of the mutation revealed it to reside on an external loop of the enzyme , some way from the active site ( Fig 5A ) ., This is compatible with the enzyme’s retaining activity , but somehow becoming divorced from other features required to progress the product further through the ergosterol pathway ., In L . donovani , gene knockout experiments with sterol 14α-demethylase concluded that the enzyme was essential and double knockout only possible in the presence of an expressed episomal version of the gene 48 ., By contrast , null mutants were made in L . major and the cells were viable , albeit hyper sensitive to temperature stress 49 ., An alignment of the primary sequences of sterol 14α-demethylase from different trypanosomatids , yeast and humans ( Fig 5B ) revealed that the mutated asparagine residue is conserved among trypanosomatid species analysed and not in yeast ( Saccharomyces cerevisiae ) or humans ( Homo sapiens ) ., The conservation could indicate important function across the Kinetoplastidae , beyond enzyme activity , for instance it could be important for protein-protein interactions ., Connection of ergosterol synthesis with AmB resistance was reported in previous studies , and here we observed substantial changes in the ergosterol biosynthetic pathway including loss of ergosterol , and mutation in sterol 14α-demethylase ( CYP51 ) ., We therefore re-expressed the WT allele in resistant cells to see if ergosterol synthesis could be restored and whether reversion to AmB sensitivity occurred ., Indeed , LC-MS revealed the restoration of the key marker of ergosterol synthesis ( Fig 6 ) , and concomitant reversion to AmB sensitivity was also associated with expression of WT CYP51 in resistant cells ., The line now expressing WT CYP51 was found to be less sensitive to ketoconazole with a fold change in EC50 value of 2 . 6 ( P = 0 . 0072 ) as compared to WT cells indicating a possible over-expression of the demethylase , a known target of ketoconazole , in the re-expressor line ( Table 3 ) ., The implication of overexpression would be the presence of more protein , requiring more drug to achieve the same level of inhibition of demethylase activity ., The AmB resistant cells expressing the WT gene for sterol 14α-demethylase were found to have lost hypersensitivity to pentamidine by reverting back to the WT EC50 value for this drug and sensitivity to miltefosine was also restored ., Since the mutation that affects sterol 14α-demethylase falls outside the active site and the enzyme retains activity as assessed by the ability of cells to convert lanosterol to its product , the mutation in the enzyme must prevent entry of the product into the remainder of the sterol pathway ., Altering the subcellular localisation of the enzyme such that the product is divorced from the next enzyme in the pathway would offer a means to allow this ., We therefore tested the subcellular localisation of both WT and mutant enzyme by tagging both with green fluorescent protein ( GFP ) at the C-terminus , the N-terminus having been proposed as important to localisation 44 , 50 ., A staining of the GFP-tagged CYP51 expressing cells with antibodies to the endoplasmic reticulum ( ER ) specific protein BiP , revealed localisation to the ER ., There was no indication that the mutated enzyme localises differently from the WT enzyme at this resolution ( Fig 7 ) hence the mutation did not seem to affect the broad compartmental localisation of the enzyme ., We also tested localisation of the next enzyme in the sterol pathway , sterol C14-reductase , by tagging with GFP and it too was found in the endoplasmic reticulum in both WT and AmB resistant cells ( Fig 7 ) ., The point mutation in CYP51 therefore has no effect on organellar targeting of that protein , nor the next enzyme in the pathway , although we have not been able to ascertain whether these enzymes are linked in either cell line ., Higher resolution microscopy or the use of different tagging system might yield more information on the localisation of the enzymes ., The leishmaniases represent a significant health burden in many parts of the tropical and sub-tropical world ., Elimination is a public health priority ., Treatment of diagnosed patients is central to elimination plans ., AmB has , in recent years , gained favour as a first line treatment for the leishmaniases , particularly in its liposomal formulation , AmBisome , which can be given in lower doses and is substantially less toxic than non-liposomal formulations of the drug ., Efficacy is such that a single injection of AmBisome ( 10 mg/kg ) is currently proposed for primary intervention 11 ., A single dose regimen carries the public health benefit of assured compliance with no need for prolonged hospitalisation ., However , the policy also brings with it a risk of resistance selection ., Future plans for sustained therapeutic intervention with combination therapies 9 , 10 , 51 , 52 once the best combination regimens have been chosen , might mitigate against this risk ., However , where AmB is part of the combination , selection of resistance during the single shot monotherapy phase of the control programme would be calamitous ., The fact that resistance to AmB has not emerged to any great extent in the treatment of fungal infections , in spite of over 50 years of use 14 , 15 , coupled to various laboratory based experiments corroborating | Introduction, Materials and methods, Results, Discussion | Amphotericin B has emerged as the therapy of choice for use against the leishmaniases ., Administration of the drug in its liposomal formulation as a single injection is being promoted in a campaign to bring the leishmaniases under control ., Understanding the risks and mechanisms of resistance is therefore of great importance ., Here we select amphotericin B-resistant Leishmania mexicana parasites with relative ease ., Metabolomic analysis demonstrated that ergosterol , the sterol known to bind the drug , is prevalent in wild-type cells , but diminished in the resistant line , where alternative sterols become prevalent ., This indicates that the resistance phenotype is related to loss of drug binding ., Comparing sequences of the parasites’ genomes revealed a plethora of single nucleotide polymorphisms that distinguish wild-type and resistant cells , but only one of these was found to be homozygous and associated with a gene encoding an enzyme in the sterol biosynthetic pathway , sterol 14α-demethylase ( CYP51 ) ., The mutation , N176I , is found outside of the enzyme’s active site , consistent with the fact that the resistant line continues to produce the enzyme’s product ., Expression of wild-type sterol 14α-demethylase in the resistant cells caused reversion to drug sensitivity and a restoration of ergosterol synthesis , showing that the mutation is indeed responsible for resistance ., The amphotericin B resistant parasites become hypersensitive to pentamidine and also agents that induce oxidative stress ., This work reveals the power of combining polyomics approaches , to discover the mechanism underlying drug resistance as well as offering novel insights into the selection of resistance to amphotericin B itself . | Antimicrobial resistance threatens to reverse many of the great strides made against pathogens responsible for disease ., Understanding the molecular processes underlying resistance is crucial to quantifying and tackling the problem ., Here we select resistance in Leishmania parasites to amphotericin B , an antileishmanial drug of increasing importance ., We then combine genome sequencing with untargeted and targeted metabolomics analyses to identify a gene , sterol 14α-demethylase , mutation of which drives a change in sterol metabolism and loss of ergosterol , the molecular target of amphotericin B . Accumulation of a downstream intermediate of ergosterol biosynthesis indicated the enzyme itself retains activity , but the pathway to ergosterol is truncated ., Expression of wild-type sterol 14α-demethylase in the resistant cells restored amphotericin B sensitivity and normal ergosterol production . | antimicrobials, cell physiology, medicine and health sciences, oxidative stress, drugs, microbiology, antifungals, cell metabolism, parasitic protozoans, metabolomics, protozoans, leishmania, pharmacology, drug metabolism, lipids, antimicrobial resistance, mycology, amphotericin, sterols, pharmacokinetics, biochemistry, cell biology, microbial control, biology and life sciences, metabolism, organisms | null |
journal.pcbi.1002926 | 2,013 | Dynamic Modeling of Cell Migration and Spreading Behaviors on Fibronectin Coated Planar Substrates and Micropatterned Geometries | Understanding cell migration mechanisms is a critical issue in many biophysical phenomena , including angiogenesis , tumor growth , metastasis , and wound healing 1–3 ., Cell migration is a complex multifaceted process , triggered by chemotaxis and haptotatic responses from the extracellular environment 4 ., Initially , a thin lamellipodium protrudes due to actin polymerization at the leading edge , followed by actin depolymerization at the lamellipodium base 5–8 ., Focal adhesions ( FAs ) are assembled between the lamellipodium base and the extracellular matrix ( ECM ) ., FAs are composed of FA molecules ( such as FAK , paxillin , vinculin , Zyxin , VASP , and talin ) , and transmembrane proteins , especially integrins αvβ3 and αvβ5 that link the ECM to the cytoskeleton via FA molecules 9 , 10 ., Afterwards , contractile bundles of actin filaments , called stress fibers ( SFs ) , extend from nascent FAs and some of which connect to the nucleus 11 ., The corresponding motor activity exerts force on the FAs fore and aft 12 , enabling the generation of a traction force and the release of FAs in the rear of the cell , creating the cell bodys forward movement ., The following individual processes of these steps of cell migration have been studied extensively in the literature: actin polymerization and depolymerization 6–8 , focal adhesion dynamics 13 , 14 , and motor activity of contractile myosin 15 , 16 ., Furthermore , both experiments and computational models from those prior works mostly involve 2-dimensional migration on a flat substrate ., However , it still remains a challenge to elucidate how these mechanisms work together to mimic 2-D cell migratory behaviors , which have been observed in existing experimental works ., The current work is motivated by two experimental works; one on Chinese Hamster Ovary ( CHO ) cell migration on 2-D ( Figure S1-A ) fibronectin coated substrate 17 , and the other on cells spreading on 2-D ( Figure S1-B ) fibronectin coated micropatterns on chips 18 ., Cell migration experiments have indicated that three separate variables , such as substratum ligand density , cell integrin expression level and integrin–ligand binding affinity , significantly affect changes in cell migration speed ., For example , when cells migrate on various fibronectin coating concentrations , the cell migration speed takes a maximum at a particular ligand density ( ∼1140 molecules/µm2 ) with a biphasic curve 17 ., On the other hand , cell spreading experiments have revealed that interactions between a cells cytoskeleton and micropatterned geometries impinge on cell morphology and mechanics 18 ., For example , when cell spreading occurs on a crossbow pattern , the cell exhibits locally high traction forces at three corners of the pattern , which may be due to concentrated ventral SFs ., Explaining complex interactions with 3-D ECM structure ( Figure S1-C&D ) entails a proper model mechanism of cell spreading because the cell morphology in 3-D ECM is strikingly different from that on 2-D ECM surfaces as the cell is elongated with the highest directionality and highest velocity of migration in 3-D ECM , but the cell forms peripheral lamellae with an increased random migration on 2-D plastic or fibronectin-coated substrates 19 ., To this end , we have built a computational 3-D cell migration model on 2-D curved ECM surfaces and discovered that the cell migration speed differs depending on the diameter of a sprout , and explained the mechanism 20 ., It is interesting to note that there is an optimal sprout diameter that creates the highest speed of cell migration ., In a similar way as on 2-D curved surfaces , we first aim to look at 3-D cell migration model on 2-D planar surfaces with various fibronectin coating concentrations to understand relationship between the migratory speed and ligand surface density ., After verifying our 3-D model with 2-D cell migratory mechanism , we then aim to look at 3-D cell spreading model on various 2-D fibronectin-coated patterns ., This entails, a ) deformation mechanics of both cell membrane and nucleus ,, b ) 3-D interactions between transmembrane integrins and ECM ligands , leading to focal adhesion formation ,, c ) SF formation and traction generation , and, d ) lamellipodium protrusion at the leading edge of the cell ., Integration of these key mechanisms is pivotal for elucidating the aforementioned migratory and spreading behaviors ., Several prior works have incorporated multiple force-generating systems in their cell migratory models 21–23 ., These works , however , have considered only frictional forces with the substrate rather than focal adhesion ( FA ) dynamics 24 , 25 , which generate a mechanical traction force due to a gradient in degraded ligand matrix density during the formation and rupture of ligand-receptor bonds 13 , 24 , interplay between Rac-mediated membrane protrusion and adhesions at the leading edge 25 ., To explain these mechanisms , a model having ligand-receptor bonds distributed across the cell membrane is necessary ., Thereby , we have applied FA dynamics to our cell migratory model ., Furthermore , our 3-D computational cell spreading model differs from other existing 2-D models 26–29 in that we incorporate aforementioned FA dynamics , cell membrane and nuclear remodeling , actin motor activity , and lamellipodia protrusion ., Additionally , our model can predict 3-D spatiotemporal behavior of cell spreading on 2-D micropatterns as well as spatiotemporal distribution of two kinds of actin stress fibers ( SFs ) , one is a SF connected to the nucleus and the other is a ventral SF , in 3-D intracellular domain ., To our knowledge , neither a cell migration or a spreading model integrating focal adhesion dynamics , cell membrane and nuclear remodeling , actin motor activity , and lamellipodia protrusion has been published that reflects 3-D spatiotemporal dynamics of both cell spreading and migration , all interfaced with a 2-D planar surface and fibronectin coated patterns ., In the following , numerical simulations demonstrate the diverse migration and spreading behaviors in relation to the various ligand densities of migrating 2-D surfaces and micropatterns , respectively ., We model the geometric structure of a cell as a double mesh structure: the outer mesh representing the cell membrane and the inner mesh for the nucleus membrane ., See Figure 1-A ., Each mesh consists of N nodes connected elastically to adjacent nodes , forming a double elastic membrane ., The inner and outer mesh nodes may be connected when SFs are formed between the nucleus and the cell membrane 30 , 31 ., Multiple transmembrane integrins are bundled together and placed at each node on the outer mesh ., They can bind to ligands on the matrix substrate , forming a focal adhesion , to which a SF is connected ( Figure 2-A ) ., Furthermore , the model also includes ventral SFs which extend between two focal adhesions ., Figure 1-B shows the free body diagram of the i-th node of the cytoskeleton , called the i-th integrin node , where a bundle of integrins is formed ., Double membranes in the integrated cell migration model move in Lagrangian approach ., Acting on this node are force vectors due to frictional dissipative force , focal adhesion force , elastic energy force , SF force , and lamellipodium force ., The equation of motion for each integrin node is given by ( 1 ) where is the velocity vector of the i-th integrin node ., Similarly , the equation of motion for each node of the nucleus is given by ( 2 ) where , and are frictional dissipative force , elastic energy force and SF force at the i-th nuclear node , respectively , and is the velocity of the i-th nuclear node ., The velocities and are expressed as ( 3 ) where and represent coordinates of the i-th integrin node and the i-th nuclear node , respectively ., Most of the frictional dissipative term arises from the rupture of stretched ligand-receptor bonds; when they rupture , the stored strain energy is released and dissipated ., Similarly , also arises from the energy stored in SFs that , when F-actin is depolymerized , the stored strain energy is released and dissipated ., These dissipative forces can be written as ( 4 ) where and are friction coefficients associated with the energy dissipation at the integrin node and the nuclear node , respectively ., In the literature these coefficients are estimated as 0 . 001 Ns/m 21 , 32 , 33 ., comes from the binding and rupture of ligand-receptor bonds and cannot easily be measured 34 ., It should be noted that the sum of forces is zero because the motion is quasi-static in time ( Text S1 , Figure S2 ) , thus Equations ( 1 ) – ( 4 ) can be simplified to the following two force balance equations: ( 5 ) ( 6 ) Formation of a focal adhesion is described by a stochastic process due to binding kinetics between receptors and ligands on the surface of ECM ., Monte Carlo simulation methods have been established for various ligand-receptor binding kinetics in the literature 35–37 ., We apply a similar technique to cell migration and spreading on planar surfaces ., First we represent the 2-D planar surface and a micropatterned geometry using a mesh of triangles , over which ligands are distributed ( Figure S3 ) ., Each focal adhesion consists of a bundle of ligand-receptor bonds ( Figure 2-B ) , each of which ruptures and binds stochastically ., Let be the probability with which a single receptor binds to a ligand on the substrate during a time interval ., ( 7 ) ( 8 ) where is the forward reaction rate ( 1 molecule−1 s−1 ) , represents the density of bound ligands , the original density of the ligands ( molecules area−1 ) , and the area associated with the integrin node under consideration ., Note that represents the number of unbound ligands available for bonding in the vicinity of the integrin node ., In simulations , a triangular mesh of approximate side lengths of 0 . 75 µm were used for area ., ( See Figure S3 ) ., Similarly , existing ligand-receptor bonds may rupture with probability during a time interval , ( 9 ) where is the kinetic dissociation rate at a distance from the force equilibrium location ., Here , is the equilibrium distance of an integrin when it is unstressed ( 20–30 nm ) 38 , represents the stretched distance from the equilibrium ( See Figure 2-B ) ., We utilized the Bells model to run stochastic simulation of bond rupturing and bonding , Bells equation for the kinetic dissociation rate is defined by 39 ( 10 ) where is the kinetic dissociation rate ( 1 s−1 ) under unstressed conditions with an equilibrium distance , is a force applied to the bond , is the transition distance ( 0 . 02 nm ) , is the Boltzmann constant , and T is absolute temperature 39 ., The number of ligand-receptor bonds , i . e . the size of each focal adhesion , can be simulated with these binding and rupture probabilities ., Let be the number of ligand-receptor bonds at the i-th integrin node , and be the number of ligands on the j-th local surface near the i-th integrin node ., The initial value of is calculated by multiplying and ., The number of bonds and available ligands vary stochastically ., By drawing a random number , , between 0 and 1: If Pran1< , then one bonding occurs , update and ., Similarly , the rupture of ligand-receptor bonds can be simulated by drawing a random number , : If Pran2< , then one rupture occurs , update and ., Above bonding-rupture tests continue in subsequent time until the bond breaks completely ( ) ., Once is known , the focal adhesion force of the i-th integrin node is computed as ( 11 ) where is an effective spring constant for a single ligand-receptor bond ( ∼1 pN/nm ) 40 , and is a unit normal vector representing the i-th integrin nodes direction on the cell membrane ( See Figure 2-B ) ., This focal adhesion force acts between the i-th integrin node and the point on the ECM surface where the extension of the unit normal vector intersects with the ECM surface ., From Figure 2-B this intersection position , that is , the root location of receptor and ligand bonds ( ) , is given by ( 12 ) where is the bond length , is the unit normal vector of the ECM surface , and is the gap between the i-th integrin node and the ECM surface , as shown in Figure 2-B ., These expressions are valid only when and the gap is less than a critical height ( ) of 300 nm ( <10 ) : ., The latter condition is to restrict the formation of receptor-ligand bonds within the upper limit ., The first set of cell migration simulations was aimed at comparing the integrated model against the experimental data published previously ., Palecek et al . 17 performed CHO cell migration experiments in 2-D planar plates under various fibronectin coating concentrations ., They found that the observed cell migration speed significantly depends on substratum ligand level , cell integrin expression level and integrin–ligand binding affinity ., Interestingly , CHO cell migration speed exhibits a biphasic dependence on extracellular-matrix ligand concentration regardless of integrin expression level ( the α5β1 receptor on fibronectin ) 17 ., The simulation results , too , showed similar behaviours of the biphasic dependence on fibronectin coating concentrations ., Figure 3-A show samples of trajectories and morphologies of simulated cell migrations along the planar surface of five different fibronectin surface densities of 19 . 4 , 192 , 568 , 1140 and 1522 molecules µm−2 for three hours ( see Videos S1 , S2 , S3 , S4 and S5 ) ., The ligand densities used for the simulations matched those of the available experiment data; ligand surface densities of fibronectin were converted from fibronectin plate concentrations ( µg ml−2 ) using the relationship between plating concentration and ligand surface density of fibronectin 41 ., First the total path length of each trajectory was obtained and was divided by the travelling time , 3 hours , to obtain the time-averaged cell migration speed ., In the experiments , the speed of CHO cell migration was monitored in every 15 minutes , and was time averaged over the entire migration period ( 12 h ) for each of fibronectin concentrations ., Figure 3-B compares the average migration speed between the experiment and simulations ., Here an error bar indicates a SE ( standard error ) of means ., The experimental data show that the cell migration speed is the lowest when migrating in the lowest ligand density , increases with increasing the ligand density , reaches a maximum value at the ligand density of 1140 molecules µm−2 , and then decreases as the ligand density becomes too dense ( Figure 3-B ) 17 , 41 ., The simulated cell migration speed , too , shows a trend similar to the experiments: slow for a very low ligand density , the fastest at the particular ligand density of 1140 molecules µm−2 , then slower again for the highest simulated ligand density ., Both experiments and simulations attain the fastest speed at the particular ligand density of 1140 molecules µm−2 ., Overall both the simulation and experiment show an excellent agreement over the ligand density range of 10∼1500 molecules µm−2 ., Statistical analysis of linear regression was performed by comparing the experiment and the simulation in terms of the mean values of time-averaged cell migration speed for the same ligand density ., As shown in Figure 3-C , good correlations were found between the two with R2\u200a=\u200a0 . 767 ., Therefore , the model validates and , in turn , is validated by showing that cell migration speeds are strongly dependent on ligand density ., The second set of cell spreading simulation was intended to compare the integrated model against the recent experimental data published by Tseng et al . 18 ., They developed a method to micropattern ECM proteins on poly-acrylamide gels in order to impinge on cell morphology and mechanics simultaneously , and have reported that measured traction forces differ considerably depending on the shape of micropatterns ., In particular , in the case of the crossbow shaped micropatterns , concentrated cell traction forces are repeatedly located in the bottom part of the vertical bar ., The simulation of the integrated model also showed similar spreading cell morphologies on micropatterned models and traction force distributions on the cell surface ( Figure 4-A , B and C ) ., Figure 4 shows spreading cell morphologies with traction force contours and oriented SFs on three micropatterned geometries ( a disk , a “pacman” shape , and a crossbow shape ) , after 60 minutes of spreading time for all shapes ( see Videos S6 , S7 and S8 ) ., Initially , all cell models start spreading from a spherical shape ., The dimensions of micropatterns used for the simulations matches those of experiments for quantitative comparisons regarding contour plots of traction forces ( or ) and spatial distributions of SFs inside of the cell; we obtained traction stress per a cell ( unit: Pa ) by dividing summations of tangential component of at i-th integrin node by a total area of ventral cell surface where focal adhesions are formed ( Figure 4-D and E ) ., Outside of the micropatterns , it was assumed that the ligand density was zero such that focal adhesion and lamellipodia protrusive forces only existed within the micropatterns ., Both experiments and simulations reveal similar trends in terms of concentrated traction forces on local areas of the ventral cell surface ( Figure 4-A , B and C ) as well as the order of higher traction stress per a cell among the three micropatterns ( Figure 4-D and E ) ., For the disk shaped micropattern , a few concentrated traction stress areas were observed at the ridge of the disk ( Figure 4-A , two yellow circles ) ., However , locations of concentrated traction forces on the disk shaped micropattern stochastically varied with time ( see Video S6 ) ., This time-varying inconsistent distribution of stress on the pattern may be due to the smooth ridge of the shape , which gives a short length of receptor-ligand bonds such that the traction energy dissipates quickly ., In the case of the “pacman” shaped micropattern , two sites of concentrated traction stress ( Figure 4-B , two yellow circles ) with SFs connected to the nucleus ( Figure 4-B , black arrows a ,, b ) and an oriented ventral SF was observed in between the sharp edges of the “pacman” mouth , as seen in experimental observations ( Figure 4-B , black arrow, c ) although additional concentrated traction forces were located in the smooth ridge of the shape like the disk shaped micropattern ., Interestingly , this behaviour was visualized to be persistent over time ( see Video S7 ) ., In the case of the crossbow shaped micropattern , ventral SFs were aligned along the top roof and the bottom bar , as seen in experimental observations ( Figure 4-C , black arrows e , f , g , h ) , and three sites of concentrated traction stress were observed at right and left end tips of the top roof and a bottom part of the vertical bar ( Figure 4-C , three yellow circles ) ., In addition , the strongest traction stress resulted from the contractile activity of SFs ( Figure 4-C , black arrow d ) at the bottom part of the vertical bar ., As the activity of actin SFs are stronger , the length of receptor-ligand bonds is stretched more at the leading edge , which results in stronger traction stress ., The animation of cell spreading simulation on the crossbow shaped micropattern , too , shows concentrated traction force at theses three sites ( see Video S8 ) ., Since a cell tends to migrate toward the stiffer gel region from the more compliant one 42 , the cell may sense locally increased tension at the sharp edge of the micropatterns as the fibronectin bundles are anchored to the plate 43 ., Thereby , larger areas of FAs are formed at the corners of the micropatterns while smaller areas of FAs are observed at the round boundary ., From the agreement between simulation and experimental results on these micropatterned shapes , the model validates and , in turn , is validated by showing persistent high stress concentrations at sharp geometrically patterned edges ., It has been reported that nascent adhesions ( smaller than ∼0 . 25 µm ) initiate the adhesion of protrusions of the leading edge of the cell , followed by the disassembly of a subpopulation of nascent adhesions within a minute and growth of the remainder into focal complexes ( ∼0 . 5 µm in size ) and then focal adhesions ( 1–5 µm in size ) within 5 minutes 44 ., Afterwards , focal adhesions either disassemble or mature within the ventral surface of the cell membrane within 10–20 minutes 45 , 46 ., Furthermore , it is known that the maturation and turnover of focal adhesions involves protein recruitment and elongation , followed by protein disengagement and shrinkage 46 ., In the current integrative cell migration model , the disengagement of actin stress fibers from integrins bound to the ECM is assumed to occur when a force-transmitting structural linkage ruptures ( =\u200a0 ) ( see Figure 2-B ) ., With the onset of motor activity after actin polymerization , the generated force is transmitted to the focal adhesions , and receptor-ligand bonds at the focal adhesions are subsequently stretched , resulting in an increases in both traction force and rupture probability for a receptor-ligand bond according to Bells law 39 ., As shown in Figure 5-A , the situation differs at the leading and trailing edges , in large part due to the location of the nucleus closer to the rear of the cell ., Note that the angle between the inclined stress fiber and the horizontal plane of the substrate at the trailing edge is higher than that at the leading edge of the cell ., If we assume that the stress fibers all exert comparable levels of force then the normal force component will be larger at the trailing edge and therefore have a higher probability of rupture , thereby allowing forward motion of the cell ., To test this hypothesis , 266 stress fibers connected to the nucleus at the leading edge and 245 stress fibers connected to the nucleus at the trailing edge were monitored and statistically analysed during three hours of simulated cell migration on the plate with fibronectin density of 200 molecules/µm2 ( Figure 5-A , Video S9 ) ., Consistent with this hypothesis , we found the lifetime of stress fibers at the trailing edge to be less than that at the leading edge of the cell; 32 . 00±2 . 78 s at the leading edge and 24 . 92±2 . 17 s at the trailing edge ( Figure 5-B ) ., Therefore , we propose that increased magnitude of normal force on the adhesion site at the trailing edge plays a key role in accelerating the rupture of receptor-ligand bonds , leading to an increase in cell migration speed ., Our modelled stress fiber lifetime physically represents a contractile SF period which is related with the turnover time of the three main dynamic components consisting of SF-actin , alpha-actinin , and myosin ., However , it should be noted that there is total lifetime of stress fiber which includes multiple periods of the lifetime of its constituent until it fully disappears ., Recently , Hotulainen and Lappalainen 47 have observed highly dynamic associations and dissociations of these components in the SF by FRAP analysis ., They found recovery times for actin , alpha-actinin , and myosin light chain ( MLC ) in bleached regions of the SF were 323 s , 123 s , and 223 s , respectively ( see fig 7A in 47 ) ., Interestingly , all components of the SF ( see fig . 7A in 47 , white boxes ) disappeared at the time of +4 s ( depolymerization occurs ) after SFs contractile motion got started at the time of −20 s ., Thus , it seems to us that this time period of 24 s may be related with contractile period of the SF among full periods of the SF ( actin polymerization , SF contractile motion , and actin depolymerization ) ., Additionally , time periods for actin polymerization and actin depolymerization in our model were set to be 180 s and 1–5 s , respectively , and time period for SF contractile motion in the model was determined to be ∼30 s ., Summation over the full period yields ∼215 s , which is within a similar range of the recovery times for the three main components of a SF ., It should be noted that most nonmotile cell types contain thick , non-dynamic stress fibers , whereas most motile cell types contain very few and thin stress fibers 47 or few and large stress fibers on the soft substrata 48 ., In case of nonmotile cells , most SFs are known to form at the ventral surface of the cell , and its movements are very slow ., However , in case of motile cells , it is possible to assemble ventral SFs by the interaction with preassembled dorsal SFs and transverse arcs within the period of 27 min ( see fig . 5 in 47 ) ., During the course of the assembly of ventral SF in motile cells , three major processes ( actin polymerization , SF contractile motion , and actin depolymerization ) are periodically repeated due to the turnover of actin in either dorsal SF or transverse arcs and SFs alignments were dynamically varied due to actin motor activity ., Thus , it should be emphasized that there exist three main highly dynamic processes of the SF ., In addition , it has been known that rapid SF depolymerization occur because of cell shortening 49 or SF detachment via localized application of trypsin at focal adhesions 50 , 51 ., Note that for the sake of video visualization of the processes of actin polymerization and bundling , the frame-to-frame time scale is 360 s while the simulation time step used is 0 . 001–0 . 01 s ., Because the frame rate is greater than the SFs dynamic period ( ∼215 s ) , the simulated SF dynamics may appear discontinuous , when they are , in fact , not ., Although there are differences in cell migration speeds between the model and experiment , we are interested in similar trends across a range of the ligand density , and linear regression between the cell migration speed of both the model and experiment with identical ligand density confirms good agreement between the model and experimental data ., Additionally , we also simulated cell migration models in which SFs are disconnected from the nuclear membrane on the substrates under five ligand surface densities ( Figure S4 ) , which resulted in lower cell migration speed than cell migration model with SFs connected to the nuclear membrane ( Figure 3-B ) ., Thus , our simulated results reveal that these SFs connected to the nucleus play an important role in cell migration ., In the literature 30 , the authors also demonstrated that nesprin-1 depleted endothelial cells showed decreased migration speed with no SFs connected to the nuclear membrane ., Furthermore , Khatau , et al . 43 highlighted the interplay between cell shape , nuclear shape , and cell adhesion mediated by the perinuclear actin cap ., We also found that the cell migration speed is limited by ligand density and integrin density ( Figure S5 ) ., They work together to promote adhesion of the cell , and in turn , cell speed ., This example shows how either value alone is enough to act as a bottle neck and limit the migration speed ., If the ligand density is high ( 950 molecules/µm2 ) , but the integrin density is insufficient ( ≤137 molecules/µm2 ) , the cell speed will be limited ., Similarly , if the integrin density is high ( 205 molecules/µm2 ) but the ligand density is insufficient ( 200 molecules/µm2 ) , then the migration speed is again limited ( Figure S5 ) ., We believe that the integration of focal adhesion dynamics ( receptor-ligand bonds ) and actin motor activity is important to observe and predict maximum cell migration speeds ., In addition , as cells contacting area on the substrate becomes larger , the numbers of focal adhesion sites such that ventral SFs anchored at FAs is increased ., That is to say , two resultant forces from focal adhesions and actin SFs are increased and they are important to capture the maximum cell migration speed dependent on substrate geometry as well as ligand surface density ., Figure 6-A shows samples of trajectories and morphologies of simulated cell migrations along the planar surface of fibronectin surface density of 1140 molecules/µm2 for three hours under nine different cases of polymerization times with 60 , 180 , and 300 s ( rows ) and depolymerization times with 1 , 10 , and 30 s ( columns ) ., First , simulated data were compared with different depolymerization times for the three values ( rows ) of polymerization times of 60 , 180 , and 300 s ., Cell migration speed at each value ( row ) of polymerization time increases as the depolymerization time becomes larger ( Figure 6-B ) ., In the case of the polymerization time of 60 s , especially , the morphologies of cells were observed to be round ., This phenomenon results from faster actin motor activity with the inclusion of a shorter polymerization process ., Thereby , the occurrence of more frequent actin motor activity prevents the cell from stretching more than the other cases of polymerization times of 180 and 300 s ., On the other hand , as the polymerization time becomes larger , the cell tends to stretch more and its morphology is changed to wider crescent-shape from the rounded shape ., Next , simulated data were compared with different polymerization times for three values ( columns ) of depolymerization times of 1 , 10 , and 30 s ( Figure 6-B ) ., As for cases of depolymerization times of 1 and 10 s , cell migration speed increases as polymerization time decreases ., In our model , a shorter polymerization process represents faster FA component ( integrin and vinculin ) renewal within FAs due to increased level of myosin II activation per FA ., Contraction could pull these components out of FAs ., It has been reported that faster turnover rates of vinculin and integrin due to further increase in actomyosin contractility are correlated with faster cell migration speed at the intermediated ligand surface density 45 ., However , in case of depolymerization time of 30 s , cell migration speed takes a maximum at an intermediated value of polymerization time of 180 s , which suggest that a balance between adhesion strength and myosin II activity is required for optimal cell migration 45 ., The elastic forces , and , are obtained by using the virtual work theory in structural mechanics ., To this end , the elastic energy stored in the cell and nucleus membranes are obtained ., Two types of elastic energy are considered ., One is the elastic energy associated with distance changes between surface nodes 52 , 53: ( 13a ) ( 13b ) where is the length of the i-th line of the cell membrane mesh , and is that of the nucleus ., Both are updated at every time-step ., and are their relaxed ( zero force ) lengths ., and are effective stiffness constants of the line elements of the cell membrane ( 5 . 0×10−5 N/m ) 32 , 54 and nucleus ( 5 . 0×10−3 N/m ) 55 , respectively ., Similarly , the elastic energy associated with area changes is given by ( 14a ) ( 14b ) where is the i-th mesh area of the cell membrane and is that of the nucleus ., and are their relaxed values ., Parameters and are effective stiffness constants of area elements of the cell membrane ( 1 . 0×10−4 N/m2 ) and nucleus ( 1 . 0×10−4 N/m2 ) , respectively 53 ., Elastic forces and can be obtained by differentiating the total energy , ( 15a ) ( 15b ) where and indicate total stored energies of the cell membrane and nucleus , respectively , and , , and are obtained analytically ., An actin SF is a bundle of actin microfilaments assembled by actin-myosin II interactions ., It is known that at least one end of each SF is connected to focal adhesion molecules , such as vinculin , talin , paxillin , zyxin , and FAK 38 , and the other end of a SF can be connected to the nuclear membrane 30 , transmitting a force to the nucleus ., In the model , the i-th integrin node is connected to the j-th nuclear node by a SF ., Its connection to the j-th nuclear node is determined by the n | Introduction, Results, Discussion, Model | An integrative cell migration model incorporating focal adhesion ( FA ) dynamics , cytoskeleton and nucleus remodeling , actin motor activity , and lamellipodia protrusion is developed for predicting cell spreading and migration behaviors ., This work is motivated by two experimental works: ( 1 ) cell migration on 2-D substrates under various fibronectin concentrations and ( 2 ) cell spreading on 2-D micropatterned geometries ., These works suggest ( 1 ) cell migration speed takes a maximum at a particular ligand density ( ∼1140 molecules/µm2 ) and ( 2 ) that strong traction forces at the corners of the patterns may exist due to combined effects exerted by actin stress fibers ( SFs ) ., The integrative model of this paper successfully reproduced these experimental results and indicates the mechanism of cell migration and spreading ., In this paper , the mechanical structure of the cell is modeled as having two elastic membranes: an outer cell membrane and an inner nuclear membrane ., The two elastic membranes are connected by SFs , which are extended from focal adhesions on the cortical surface to the nuclear membrane ., In addition , the model also includes ventral SFs bridging two focal adhesions on the cell surface ., The cell deforms and gains traction as transmembrane integrins distributed over the outer cell membrane bond to ligands on the ECM surface , activate SFs , and form focal adhesions ., The relationship between the cell migration speed and fibronectin concentration agrees with existing experimental data for Chinese hamster ovary ( CHO ) cell migrations on fibronectin coated surfaces ., In addition , the integrated model is validated by showing persistent high stress concentrations at sharp geometrically patterned edges ., This model will be used as a predictive model to assist in design and data processing of upcoming microfluidic cell migration assays . | Cell migration is a complex , multifaceted process , triggered by chemotaxis and haptotatic responses from the extracellular matrix ( ECM ) ., It is triggered by a thin lamellipodium protrusion at the leading edge , followed by the assembly of a number of focal adhesions between the lamellipodium base and the ECM ., Afterwards , actin stress fibers extend from nascent focal adhesions , some of which connect to the nucleus ., In this work , we have developed a dynamic model of cell migration incorporating these four mechanisms of cell biology , such as remodeling of cell and nuclear membranes , focal adhesion dynamics , actin motor activity , and lamellipodia protrusion at the leading edge ., We successfully compared our model with existing experimental works of cell migration on ( 1 ) substrates with various fibronectin coating concentrations , and ( 2 ) cell spreading on three patterned surfaces ., Finally , our model demonstrates how actin stress fibers anchored at the trailing edge play a key role , leading to an increase in cell migration speed ., Thereby , the model will not only provide new insights on better building such an experiment , but also further experiments will allow us to better validate the model . | systems biology, biophysic al simulations, biology, computational biology, biophysics simulations, biophysics, biomechanics | null |
journal.pcbi.1003742 | 2,014 | Membrane Interaction of Bound Ligands Contributes to the Negative Binding Cooperativity of the EGF Receptor | The epidermal growth factor receptor ( EGFR ) , a member of the Her ( ErbB ) family of cell-surface receptors , is critical to a variety of cellular processes and is implicated in the development of several forms of cancer and other diseases 1–3 ., In normal cells , EGFR activation is initiated by the binding of extracellular ligands from the epidermal growth factor ( EGF ) family 4–7 , giving rise to the formation of active EGFR dimers , which transmit intracellular signals ., It was first shown 30 years ago 8–12 that the Scatchard plots of EGF binding to EGFR are nonlinear ( concave up ) , which is indicative of heterogeneous binding affinity ., It has been further suggested that the heterogeneity in EGFR ligand binding may play an important role in determining the signaling response to different ligand concentrations 12–14 ., More recently , in a study conducted by Pike and colleagues 15 , a characterization of EGFR ligand binding based on a simultaneous fitting of binding isotherms from cells with different levels of EGFR expression showed that EGFR ligand binding may be described by a simple model ( shown in Fig . 1A ) ., In this model , which is consistent with earlier results 16 , negative cooperativity underlies the heterogeneity of EGFR ligand binding 15 , 16: The binding affinity of a ligand at one EGFR binding site is smaller when the other site is occupied ) ., The structural origin of this negative cooperativity has been unclear ., The existence of two structurally distinct binding sites in the doubly liganded dimer of Drosophila EGFR ( dEGFR ) 17 is consistent with its binding cooperativity ., In crystal structures of the doubly liganded human EGFR dimer , however , the two binding sites are structurally virtually identical 18 , 19 ., A recent investigation , based on structural and biochemical analyses 20 , suggested that the ligand-binding cooperativity may be explained by a conformational change in the ectodomain dimer ., Although it is very plausible that such a scenario explains some of the negative cooperativity in EGFR , it is not clear that it represents the only , or even the main , contribution ., Here we use molecular dynamics ( MD ) simulations to investigate the structural basis of the negative cooperativity in the ligand binding of human EGFR ., We simulated the human EGFR ectodomain dimer anchored to a lipid membrane by EGFR transmembrane ( TM ) helices ., In our simulations of both singly and doubly liganded ectodomains , the dimer began in an upright orientation , with the dimers long axis perpendicular to the membrane , then spontaneously rotated and lay down on the membrane in such a way that one of the binding sites faced the membrane , while the other faced the bulk solvent ., The ligand in the membrane-facing site developed extensive favorable interactions with the membrane , and our approximate free energy calculations suggest that these interactions contribute a significant fraction of the ligand binding free energy ., These findings are consistent with Förster resonance energy transfer ( FRET ) experiments 21 , 22 , which showed that some EGFR-bound ligands are positioned within 40 Å of the membrane , while others are positioned beyond 70 Å ., In further simulations of glycosylated EGFR ectodomains , we found that the ectodomain orientation and the membrane interaction of the bound ligand appear compatible with the in vivo glycosylation of EGFR ., Based on our simulation findings , we suggest that the negative cooperativity in human EGFR ligand binding may arise in part from broken symmetry between the two bound ligands in an orientation in which the ectodomain rests on the membrane ., Specifically , the simulations showed that the membrane favorably interacts with ligands bound to EGFR ectodomains resting on the membrane , and the results suggest that the high-affinity binding to an unliganded EGFR dimer may be attributed to this previously largely overlooked ligand-membrane interaction ( Fig . 1A ) ., Such a structural explanation is supported by the experimental finding that when high-affinity ligand binding is abolished , the distance between bound ligands and the membrane increases 22 ., The mechanism we propose here also offers a straightforward explanation of the observation that negative ligand-binding cooperativity in human EGFR is observed only when the receptor is embedded in the membrane ., We first simulated a dimer of the ectodomains ( domains I , II , III , and IV ) and the single-helix TM segment of EGFR ., Based on available crystal structures 18 , 19 , 23 , the ectodomains were prepared in the form of an EGF-bound , back-to-back symmetric dimer , and the TM helices were prepared in the form of a TM dimer , with the N-terminal GxxxG-like motifs ( where G represents glycine , or another amino acid with a small side chain ) as the dimer interface 24–26 ., The ectodomains were initially positioned upright , approximately perpendicular to the membrane ( Fig . 1B ) ., In the simulation , the ectodomains lay down toward the membrane surface , and approximately 1 . 7 µs into the simulation , one of the bound ligands developed extensive interactions with the membrane ( Fig . 1C ) in a partial-resting orientation of the ectodomain dimer ., Later in the course of the simulation , the ectodomains lay flat on the membrane , producing a full-resting orientation ( Fig . 1D ) ., The simulation observations are supported by FRET measurements 21 , 22 , which have indicated that the EGFR ectodomain dimer may rest on the membrane ., The simulation indicates that such orientations of the ectodomains are made possible because the linker segments between the ectodomains and the TM helices are not fully rigid , as has been previously suggested 23 ., We previously simulated EGFR ectodomain monomers tethered to membrane-embedded TM helices ., Although the ectodomain is ligand-free in the simulations , it came to rest on the membrane from an upright orientation in such a way that , if it were ligand-bound , the ligand would be in contact with the membrane ( Fig . 6A in ref . 26 ) ., Notably , the orientation of the ectodomain dimer with respect to the membrane broke the symmetry between its two bound ligands: One of the ligands ( the membrane-facing ligand ) but not the other ( the solvent-facing ligand ) was in contact with the membrane ( Fig . 1C ) ., Once the ectodomain dimer rested on the membrane , the membrane-facing ligand developed extensive favorable interactions with the lipids ., In particular , the hydrophobic residues of the EGF ligand , such as Pro7 and Leu8 , were found to enter the hydrophobic interior of the membranes extracellular leaflet ( Fig . 1C , inset ) ., As discussed in detail in later sections , similar membrane interactions were also observed for one of the bound ligands in two other EGFR dimer simulations we performed ., Although it is challenging to accurately calculate biomolecular binding free energies in simulation , generalized Born models can often provide a rough estimate ., We estimated the binding free energy of each bound ligand using the molecular mechanics/generalized Born volume integration ( MM/GBVI ) model ( see Methods ) 27 ., We initially calculated what we refer to as the ligand-protein interaction energy , which is an estimate for the free energy if there is little change in the protein structure on binding ., As shown in Fig . 1E , the simplest application of the MM/GBVI method yields an interaction energy of an EGF ligand with EGFR of ∼80 kcal mol−1 ., The ligand-membrane interaction energy contributes an additional ∼25 kcal mol−1 to the membrane-facing ligand but almost nothing to the solvent-facing one ., The EGF-EGFR interaction energy of over 80 kcal mol−1 is much higher than the experimental value of the EGF binding free energy ( 10–15 kcal mol−1 15 ) ., As noted above , however , the computational quantity does not include the conformational free energy cost incurred when EGFR adopts its ligand-bound conformation during EGF binding ., Using the structures of the active 18 , 19 and inactive 26 dimers , the MM/GBVI model estimates the cost of EGFRs transition to the ligand-bound conformation to be 89 kcal mol−1 per monomer in the EGFR dimer ., This is qualitatively consistent with our previously reported simulations 26 , in which the ligand-bound conformation was not stable without the EGF ligands , and it suggests that the favorable EGF-EGFR interaction is approximately canceled out by EGFR adopting the unfavorable ligand-bound conformation ., Despite this apparently satisfactory cancellation , we suspect that the individual canceling terms may still be overestimates , and we regard the use of the generalized Born model in the present context as qualitative ., To assess whether interactions with the membrane could contribute significantly to the experimental ligand-binding free energy , we do not only look at the large ( ∼25 kcal mol−1 ) estimated binding energy , but we also look at how this energy compares to the estimated ligand-protein interaction energy ., Since the protein-membrane energy is a substantial fraction even of the large ligand-protein interaction energy , the generalized Born model supports the conclusion that the membrane-facing ligand binds to the EGFR dimer with higher affinity than does the solvent-facing one ., Intriguingly , we found that the hydrophobic patch formed by Pro7 and Leu8 is largely conserved in vertebrate EGF molecules ( Fig . 1F ) ., Pro7 is especially well conserved in vertebrate EGF ., Leu8 is less conserved , but this position is hydrophobic in the majority of vertebrate EGF members despite being solvent accessible ., In vivo , human EGFR is glycosylated on the ectodomains 28 , and 10 of the receptors 12 potential glycosylation sites are found to be fully or partially occupied by a variety of large , branching glycans 29 , 30 ., It is conceivable that the relatively bulky glycans may preclude an EGFR ectodomain dimer from resting on the membrane ., To test whether this is the case , we modeled and simulated an EGFR ectodomain–TM dimer system with full glycosylation ( Fig . 2A ) ., We decorated the EGFR ectodomains at the 10 identified glycosylation sites 29 , 30 with three types of glycans ( BiS1F1 , Man6 , and Man8 ) that are common in EGFR glycosylation ( see Methods , Fig . S1 , and Table S1 ) ., The simulation shows that the glycosylation does not disrupt the ligand-membrane contacts ( Fig . 2B ) we observed in nonglycosylated EGFR ., In the course of the simulation starting from the partial-resting orientation ( Fig . 1C ) , the orientation of the ectodomains with respect to the membrane remained unchanged , with the membrane-facing ligand embedded in the membrane and the other ligand facing the solvent ., The simulation showed that , in addition to interglycan interactions , the polar glycans interact extensively with the protein and the lipid head groups ., The flexibility of glycans allowed the ectodomains to rest on the membrane ., The glycans were found to be distributed adjacent to the protein surfaces , rather than protruding into the solvent ( Fig . S2 ) ., The membrane-facing ligand of the glycosylated EGFR exhibited the same degree of membrane interactions in the simulation ( Fig . 2C ) as that of the nonglycosylated EGFR ( as indicated by the values of t\u200a=\u200a0 in Fig . 2C ) ., In fact , the membrane embedding of the glycosylated-EGFR membrane-facing ligand was slightly deeper than that of the membrane-facing ligand in nonglycosylated EGFR ., We calculated that the membrane interaction contributes an additional estimated ∼30% to the membrane-facing ligands MM/GBVI binding energy but virtually nothing to that of the solvent-facing ligand ( Fig . 2C ) ., We thus conclude that robust interaction between EGFR-bound ligands and the membrane , which may contribute to the heterogeneous ligand binding in an EGFR dimer , is accessible to glycosylated as well as nonglycosylated EGFR ., Having demonstrated that the ectodomains in a two-ligand EGFR dimer may rest on the membrane and that the two ligands may differ in their interactions with the membrane due to the ectodomains orientation , we further simulated the one-ligand EGFR dimer ., These simulations suggest that the ectodomains may rest on the membrane and that the ligand in a one-ligand EGFR dimer may also develop favorable interactions with the membrane , thus providing a structural model for high-affinity binding in the one-ligand EGFR dimer ( Fig . 1A ) ., Because a crystal structure of a singly liganded ectodomain of an EGFR dimer is not yet available , we made a model based on the crystal structure of the two-ligand ectodomain dimer by removing one bound ligand from the crystal structure 26 ., We here simulated the one-ligand ectodomain dimer in this conformation , connected with the TM segments , three times ., In all three simulations , the ectodomain dimer , which was initially in an upright orientation , spontaneously lay down on the membrane ( Fig . 3B ) , allowing the ligand to come into contact and develop extensive interactions with the membrane ., We again calculated the MM/GBVI energy of the ligands interaction with the membrane ( Fig . 3C ) ., The results suggest that the membrane interaction is energetically favorable , and that the free energy increase associated with the ligands membrane interaction is a significant fraction of the free energy arising from its interaction with EGFR ., The favorable nature of the ligand-membrane interaction strongly suggests that the membrane-facing binding sites are associated with high-affinity binding ., This notion is notably supported by the observation from FRET experiments that abolishing the high-affinity binding leads to a significant increase in the average distance between the ligands and the membrane 22 ., Assuming a thermodynamically equilibrated system , in one-ligand EGFR dimers the ligands predominantly occupy the high-affinity sites facing the membrane ( Fig . 1A ) ., We also performed a similar simulation of a fully glycosylated , one-ligand ectodomain dimer attached to the TM segments ., Starting form an upright conformation , the ectodomain dimer again spontaneously lay down , with its bound ligand coming into contact with the membrane ( Fig . 3B and 3C ) ., This simulation suggests that a glycosylated one-ligand dimer may also prefer to rest on the membrane in such a way that its bound ligand faces the membrane , and that the ligand-membrane interactions are energetically favorable A simulation study of EGFR 22 has previously suggested that ectodomain interactions with the membrane may be at the root of the observed negative cooperativity of ligand binding ., It was further suggested that the negative cooperativity may arise from the ectodomains transition to a dEGFR-like asymmetric conformation , induced by interactions with the membrane ., Although our simulations also suggest the important role of EGFR ectodomain interactions with the membrane , our simulations did not show a robust transition from a symmetric to an asymmetric conformation in the ectodomain dimer ., Fig . 4A shows that , other than minor deviations due to the inherent flexibility of the loop regions , the dimers two ectodomain subunits were nearly conformationally identical in our simulations ., In particular , the conformations of domain II in the two subunits are highly similar , whereas in the dEGFR dimer the domain II is straight in one subunit and bent in the other , which ultimately leads to the different conformations of the two binding sites ., This is illustrated in Fig . 4B , where the angle characterizing the bending of domain II is plotted ., The angles of the two EGFR subunits were approximately the same in our simulations of the two-ligand dimer , much as they are in crystal structures 18 , 19 ., Our MM/GBVI calculation supports the notion that the two receptors of the two-ligand EGFR dimer maintain similar binding-site conformations while resting on the membrane: The two ligands have comparable MM/GBVI interaction energies with the receptors ( Fig . 5B ) , including cases in which the receptors are glycosylated ( Fig . 2C ) ., On the other hand , for the one-ligand dimer , which assumes asymmetric conformations , the angles differ significantly between the two subunits ., Similarly , in our simulations the average root-mean-square deviation ( RMSD ) of the Cα atoms of domains I , II , and III between the two subunits was significantly lower for the two-ligand ectodomain dimer ( 2 . 7±0 . 3 Å when glycosylated and 2 . 6±0 . 5 Å when nonglycosylated ) than for the one-ligand dimer ( 4 . 4±0 . 2 Å when glycosylated and 4 . 7±0 . 3 Å when nonglycosylated ) ., We performed three independent simulations of the two-ligand , nonglycosylated EGFR dimer ., As discussed above , in one of these simulations , the ectodomain dimer first assumed a partial-resting orientation and eventually lay flat on the membrane ( Fig . 1; also shown as Simulation 1 in Fig . 5 ) ., In another simulation ( Fig . 5A , Simulation 2 ) , the system arrived at a similar partial-resting orientation and remained there to the end of the simulation ., In the third simulation ( Fig . 5A , Simulation 3 ) , the ectodomain dimer was found to rest sideways on the membrane surface ., What is common to all three simulations , however , is that only one bound ligand made extensive contact with the membrane ( Fig . 5B , upper panels ) despite the variation in the orientation of the ectodomain dimer ., The buried surface area of the membrane-facing ligand was consistently greater than that of the solvent-facing ligand ( Fig . 5B ) : 2 , 300±100 Å2 versus 1 , 600±100 Å2 ., A substantial portion ( up to 200 Å2 ) of the approximately 700-Å2 difference is due to the embedding of Pro7 and Leu8 in the membrane ., Our MM/GBVI calculations also consistently suggest that ligand-membrane interactions contribute a significant fraction to the free energy of ligand-receptor binding ( Fig . 5B ) ., Earlier FRET measurements indicated that EGF bound to EGFR dimers falls into two groups: one in which the bound ligand is close to the membrane , and another in which it is farther away ., Specifically , the FRET results showed that the N termini of the EGF molecules in the “close” group are no more than 35–40 Å from the membrane , and the N termini of the EGF molecules in the “far” group are no closer than 69–71 Å from the membrane 21 , 22 ., Our simulation results ( Figs . 2C and 5B ) agree with these data ( see the description of the distance measurements in the Methods section ) and the simulation conformations are similar to those in the structural model proposed by Kästner et al . based on their FRET results 31 ., In all of our simulations , the membrane-facing ligands were close to the membrane surface ( ∼10 Å ) , and thus belong to the former population ., This population may also include the solvent-facing ligands in cases in which the ectodomain dimer rests flat on the membrane , such as at the end of Simulation 1 , where the membrane distance is ∼40 Å for the solvent-facing ligand ., The latter population , on the other hand , may consist of the solvent-facing ligands in dimers such as those observed in Simulations 2 and 3 , as well as those in the upright dimers ( with distances of ∼80–120 Å ) ., Combining the observations from the simulations with those from the FRET measurements , we suggest that it is unlikely that the negative cooperativity of ligand binding can be attributed to a single specific orientation of the ectodomain dimer ., We instead suggest that the cooperativity is associated with an ensemble of ectodomain-dimer orientations , with the shared feature that the high-affinity ligand binding occurs at the membrane-facing binding site ., This provides a straightforward explanation for the experimental observation that abolishing high-affinity ligand binding increases the average ligand-membrane distance 22 ., Additionally , our simulations showed that free EGF molecules may interact favorably with and be attached to the membrane ( Fig . S3 ) ., This simulation finding , combined with the observation that Spitz ligands ( which bind to dEGFR ) need to be palmitoylated ( and thus attached to membrane ) to activate dEGFR in vivo 32 , raises the possibility that the ligand-binding process of EGFR may occur at the membrane surface ., Our simulations suggest that an EGFR ectodomain dimer may rest on the membrane , and that the interaction between a bound ligand and the membrane may lead to a breaking of the symmetry between the two ligands , thus contributing to the negative cooperativity of EGFR ligand binding ( Fig . 1A ) ., Our investigation is in part inspired by the FRET measurements of ligand distance from the membrane; based on these results , the orientation of EGFR ectodomains relative to the membrane was suggested to affect ectodomain conformations and give rise to the negative cooperativity 21 , 22 , 31 ., The mechanism we propose here is particularly supported by the FRET finding that abolishing high-affinity ligand binding leads to a significant increase in the average distance between EGFR-bound ligands and the membrane 22 ., Our simulations of glycosylated EGFR ( to our knowledge the first simulations of a fully glycosylated receptor ) showed that the mechanism we propose is compatible with EGFR glycosylation: A glycosylated ectodomain dimer may also rest on the membrane , and the attached glycans do not preclude interactions between the EGFR-bound ligand and the membrane ., In this investigation , we have largely focused on EGFR dimers because they are central to the negative cooperativity of EGFR ligand binding 14 ., EGFR monomers may also bind ligands at a high affinity comparable to that of EGFR dimers 15 , but the ectodomain structure of the ligand-bound EGFR monomer has not yet been resolved ., In previous MD simulations , we showed that an EGFR monomer is similar to an EGFR dimer in that its ectodomains also rest on the membrane in a way that would allow membrane contact with the bound ligand 26 ., From this observation , which is independent of any specific conformation of the ectodomains , it may be inferred that the high affinity of ligand binding in EGFR monomers could also be explained by favorable interactions between the membrane and the bound ligands ., Our simulations suggest that the ectodomains of an EGFR dimer may rest on the membrane and that a bound EGF ligand may be in direct and energetically favorable contact with the membrane ., Our earlier simulations also suggest that EGFR monomer ectodomains may also rest on the membrane 26 ., This does not imply , however , that the ectodomains are fixed on the membrane in well-defined orientations ., It is likely that , on a timescale much longer than our simulations , the ectodomains convert from one orientation to another in a dynamic equilibrium ., While the orientations in which the ectodomains rest on the membrane may predominate , the ectodomains likely access the other orientations that could be crucial to the process of ligand binding or EGFR dimerization ., A recent study 20 proposed that a conformational change from the so-called “flush” to the “staggered” arrangement between the two extracellular subunits in an EGFR dimer ( Fig . 6A ) may be at the root of the binding cooperativity of EGFR ., While such a binding-cooperativity mechanism differs from the mechanism we propose here , these two mechanisms are not mutually exclusive ., In agreement with the finding of Liu et al . 20 based on crystal structures , our simulations show that the two-ligand EGFR dimer prefers the staggered conformation and that the one-ligand and ligand-free EGFR dimers prefer the flush conformation 26 ., Intriguingly , the ectodomain interaction with the membrane and the glycosylation of EGFR appear to strengthen this trend ( Fig . 6B ) ., From this observation , we suggest that the membrane may be of critical importance to the negative cooperativity of EGFR ligand binding , not only for its asymmetric interactions with the bound ligands , but also for its effect on the accessible conformational space of the ectodomain dimers ., Further investigation is certainly needed to quantify the contribution of the conformational dynamics of the ectodomains and the contributions of ligand-membrane interactions to the ligand-binding cooperativity of EGFR ., Further investigation would also be needed to clarify whether the membrane interactions of the ectodomains have any role in autoinhibition ., We have not addressed this question , but we have previously shown that the membrane interactions of the EGFR kinase domain do play an autoinhibitory role 25 , 26 ., Experiments have shown that the ligand-binding cooperativity of EGFR is apparently missing for isolated EGFR dimer ectodomains in solution 12 ., It was shown that the negative cooperativity may be partially recovered when the membrane is included in experiments of EGFR ectodomains attached to the TM helices 33 ., Our suggested mechanism for the negative binding cooperativity , in which the membrane plays a central role , offers a straightforward explanation for these findings ., If the asymmetry between the bound ligands in an EGFR dimer , and thus the binding cooperativity , is indeed associated with the difference in the interactions of bound ligands with the cell membrane , the absence of the membrane would naturally eliminate the binding cooperativity ., Likewise , the lack of cooperativity for detergent-solubilized EGFR 34 may be explained by the absence of an extended membrane capable of interacting with EGFR-bound ligands ., It has been shown that mutations at the intracellular domains of EGFR yield nearly linear Scatchard plots 33 ., Although these Scatchard plots could reflect a weakened negative cooperativity due to these mutations , and thus suggest that the root of the negative cooperativity may lie beyond the ectodomains and the membrane , there is an alternative explanation: that the dimerization prior to ligand binding , which is a prerequisite of the binding cooperativity 15 , was weakened , leading to both a near-linear Scatchard plot and a difficulty in using the plot to reliably quantify binding cooperativity 20 ., Our investigation of the relationship between the EGFR ectodomains and the cell membrane using atomistic , long-timescale MD simulations suggests a structural mechanism for the negative cooperativity of ligand binding of EGFR dimers; in this mechanism , the ectodomains may rest on the membrane , and the presence of the membrane may break the symmetry between the two binding sites ., These results add further support to the emerging view that interactions between EGFR and the membrane play a central role in many aspects of the regulation of EGFR signaling 25 , 26 , 34–36 ., The simulations were performed on a special-purpose supercomputer , Anton 37 , using the Amber ff99SB-ILDN 38–40 force field , combined with the ff99SB* backbone correction 41 for proteins , the CHARMM C36 force field 42 for lipids , and TIP3P 43 as the water model ., The simulated systems were solvated in water with 0 . 15 M NaCl , with residue protonation states corresponding to pH 7 ., Additional Na+ ions were included to neutralize the net charges of the proteins ( −3 for the extracellular domains of each EGFR , −4 for each EGF ligand ) and the POPS lipids ., As an equilibration stage , the protein backbone atoms were first restrained to their initial positions using a harmonic potential with a force constant of 1 kcal mol−1 Å−2 ., The force constant was linearly scaled down to zero over at least 50 ns ., Simulations were performed in the NPT ensemble with T\u200a=\u200a310 K and P\u200a=\u200a1 bar using the MTK algorithm 44 with 20-ps relaxation time ., Water molecules and all bond lengths to hydrogen atoms were constrained using M-SHAKE 45 ., The simulation time step was 1 fs for the equilibration stage and 2 fs for production simulations; the r-RESPA integration method was used , with long-range electrostatics evaluated every 6 fs 46 ., The glycosylation of EGFR was modeled based on the mass-spectrometry analysis of the CL1-0 cell line 30 , which is broadly consistent with similar analysis on CL1-5 and A431 cell lines 29 , 30 ., Since EGFR glycan attachments in the cell are very diverse—for every glycosylation site there is a large number of different glycan types that can be attached to it—we chose glycans among the most commonly found at the specific sites ., These three common types are BiS1F1 , Man6 , and Man8 ( Fig . S1 and Table S1 ) ., The glycan structures for the initial models were obtained using the Glycam web service 47 and then adjusted in VMD 48 to avoid clashes with protein and membrane ., The simulations were performed with the GLYCAM06 force field 49 applied to the glycans ., The simulated systems included the ectodomain–TM dimers with two EGF molecules bound ( three simulations of 2 . 6 , 1 . 2 , and 2 . 1 µs; ∼315 , 000 atoms ) and with one EGF molecule bound ( three simulations of 2 . 5 , 2 . 3 , and 0 . 9 µs; ∼300 , 000 atoms ) , a two-ligand glycosylated ectodomain–TM dimer ( 3 . 0 µs; ∼310 , 000 atoms ) , a one-ligand glycosylated ectodomain–TM dimer ( 8 . 3 µs; ∼300 , 000 atoms ) , and a single EGF molecule ( see SI; two simulations of 8 . 9 and 8 . 3 µs; ∼62 , 000 atoms ) ; a membrane was included in every case ., Each system is set up such that each dimer is at least 25 Å from its periodic image ., The model membrane consisted of POPC lipids , with 30% ( molar ) POPC randomly replaced by POPS in the intracellular leaflet of the bilayer ( only for the ectodomain–TM simulations ) to approximately mimic the charge distribution in the cellular membrane 26 , 50 ., Modeling , analysis , and visualization were performed using VMD 48 ., The distance between the EGF N terminus and the membrane , namely the distance from the N terminus to the plane through the phosphates of the extracellular lipid layer , was computed in a manner consistent with the FRET measurements 22 ., The EGF-EGFR interaction energy estimation was based on the molecular mechanics/generalized Born volume integration ( MM/GBVI ) model 27 and performed using MOE software ( Chemical Computing Group ) 51 ., The EGF-receptor binding energy was calculated for each snapshot from the difference of the energy of the EGF-receptor complex and the sum of isolated EGF and receptor energies ., The EGF-membrane energy was calculated analogously ., The conformational free energy of EGFR extracellular dimers was estimated based on the published coordinates of the full-length ligand-bound and ligand-free EGFR dimers 26 after energy minimization ., Our calculations included domains I , II , III , and IV ., The MM/GBVI energy is −34287 . 4 kcal mol−1 for the ligand-free dimer and −34110 . 2 kcal mol−1 for the ligand-bound dimer ( the EGF ligands were not included in the calculation ) , and thus the conformational free energy cost for each monomer is 88 . 6 kcal mol−1 . | Introduction, Results, Discussion, Methods | The epidermal growth factor receptor ( EGFR ) plays a key role in regulating cell proliferation , migration , and differentiation , and aberrant EGFR signaling is implicated in a variety of cancers ., EGFR signaling is triggered by extracellular ligand binding , which promotes EGFR dimerization and activation ., Ligand-binding measurements are consistent with a negatively cooperative model in which the ligand-binding affinity at either binding site in an EGFR dimer is weaker when the other site is occupied by a ligand ., This cooperativity is widely believed to be central to the effects of ligand concentration on EGFR-mediated intracellular signaling ., Although the extracellular portion of the human EGFR dimer has been resolved crystallographically , the crystal structures do not reveal the structural origin of this negative cooperativity , which has remained unclear ., Here we report the results of molecular dynamics simulations suggesting that asymmetrical interactions of the two binding sites with the membrane may be responsible ( perhaps along with other factors ) for this negative cooperativity ., In particular , in our simulations the extracellular domains of an EGFR dimer spontaneously lay down on the membrane in an orientation in which favorable membrane contacts were made with one of the bound ligands , but could not be made with the other ., Similar interactions were observed when EGFR was glycosylated , as it is in vivo . | Epidermal growth factor receptor ( EGFR ) molecules are of central importance in cellular communication ., Embedded in the cell membrane , these receptors bind epidermal growth factor ( EGF ) molecules outside the cell and translate this binding into specific biochemical signals inside the cell , which in turn trigger cell proliferation , migration , or differentiation ., EGFR dysfunction has been implicated in a variety of cancers , and EGFR-targeting drugs are commonly used in cancer treatments ., It has been widely assumed that the extracellular portion of an EGFR molecule protrudes perpendicularly from the cell membrane ., In detailed , atomic-level computer simulations , however , we find that it lies down on the membrane , placing its EGF-binding site adjacent to the membrane surface ., We further show that EGF may interact with EGFR in two distinct ways ( with or without the involvement of the membrane ) ., This may explain the experimental finding that an EGF molecule binds to EGFR more weakly at higher EGF concentration ., This phenomenon , which is a manifestation of an underlying negative cooperativity , is an important but poorly understood characteristic of EGFR activity ., In this study , we also model and analyze the glycan chains attached to EGFR , which are integral to its behavior in living cells . | biophysics, biology and life sciences, computational biology | null |
journal.pgen.1004822 | 2,014 | The Evolution of Sex Ratio Distorter Suppression Affects a 25 cM Genomic Region in the Butterfly Hypolimnas bolina | In 1930 , Fisher noted that the strength of selection on the sex ratio was frequency dependent , echoing earlier findings of Düsing 1 , 2 ., As a well-mixed outbreeding population progressively deviates from a 1∶1 sex ratio , selection on individuals to restore the sex ratio to parity becomes stronger ., In natural animal populations , a common cause of population sex ratio skew is the presence of sex ratio distorting elements , in the form of either sex chromosome meiotic drive 3 , or cytoplasmic symbionts 4 ., In some cases , these elements can reach very high prevalence , distorting population sex ratios to as much as 100 females per male 5 , and producing intense selection for restoration of the individual sex ratio to 1 female per male ., The most common consequence of this selection pressure is the evolution of systems of suppression – host genetic variants that prevent the sex ratio distorting activity from occurring ., Suppressor factors are known for a wide range of cytoplasmic symbionts and meiotic drive elements 3 , 6 , 7 ., The evolution of suppression of Wolbachia induced male-killing activity in the butterfly Hypolimnas bolina represents a compelling observation of intense natural selection in the wild ., Female H . bolina can carry a maternally inherited Wolbachia symbiont , wBol1 , which kills male hosts as embryos 8 ., The species also carries an uncharacterised dominant , zygotically acting suppression system that allows males to survive infection 6 ., Written records and analysis of museum specimens indicate this symbiont was historically present , and active as a male-killer , across much of the species range , from Hong Kong and Borneo through to Fiji , Samoa and parts of French Polynesia 9 ., Evidence from museum specimens also indicates that host suppression of male-killing had a very restricted incidence in the late 19th century , with infected male hosts ( the hallmark of suppression ) being found in the Philippines but not in other localities tested ., By the late 20th century , suppression of male-killing was found throughout SE Asia , but not in Polynesian populations where the male-killing phenotype remained active 10 ., The most extreme population was that of Samoa , where 99% of female H . bolina were infected with male-killing Wolbachia , resulting in a sex ratio of around 100 females per male within the population 5 ., However , following over 100 years of stasis on Samoa , the rapid spread of suppression of male-killing activity of the bacterium was finally observed between 2001 and 2006 , restoring both individual and population sex ratio to parity 11 ., When strong selection occurs at a locus , it is expected to leave a genomic imprint beyond the target of selection , as a result of genetic hitch-hiking ., A neutral ( or even deleterious ) variant that is initially present in the haplotype in which the favoured allele arose ( i . e . is linked to the site of selection ) , will also increase in frequency 12 ., When selection is very strong , the frequency of linked variants may increase across a broad genomic region 13 ., Importantly , the extent of the chromosome over which this effect will occur depends on the selection pressure in the first few generations; before recombination has broken down associations between the target of selection and linked variants ., Where sex ratio distorters are common , the selection pressure in these first generations may be very strong indeed ( before the sex ratio becomes less biased through spread of the suppressor ) ., It is thus likely that selection on the sex ratio will influence linked material over a broader genomic region compared to many other selective regimes ., That is , the episode of selection is likely to have a very wide genomic impact ., In this paper , we first mapped a genomic region in SE Asian butterflies that was required for male survival in the presence of Wolbachia ., We then investigated the impact of the recent spread of the suppressor in Samoa on the pattern of variation around this region ., To this end , we initially developed theory to predict the impact of suppressor spread on linked genetic variation ., We then directly observed changes in the frequency of genetic variants surrounding the suppressor locus by comparing the pattern of genetic variation in H . bolina specimens collected in Samoa before ( 2001 ) and after the selective sweep ( 2006 and 2010 ) ., By examining post-sweep samples at two time points we were additionally able to track allele frequency changes following the initial sweep ., The data revealed changes in the pattern of genetic variation over a 25 cM region surrounding the suppressor locus ., We further suggest that the suppressor was probably derived by immigration , and that the sweep may have introduced deleterious material that was subsequently subject to purifying selection ., Hypolimnas bolina has 31 chromosomes and a total genome size of 435 MB 14 , 15 ., Previous work established that the rescue of male zygotes from Wolbachia induced killing was dominant , and potentially a single locus trait 6 ., Genetic markers spanning the genome were developed using a targeted gene approach informed by conservation of synteny in Lepidoptera , with the sequence of H . bolina orthologs obtained through Roche 454 transcriptome sequencing ( see Methods and Materials , NCBI SRA accession: SRP045735 ) ., These markers were then tested for co-segregation with suppression in order to identify the linkage groups associated with male host survival ., Female butterflies from South East ( SE ) Asia that carried both Wolbachia and the suppressor allele , were crossed with males from the French Polynesian island Moorea ( where suppression is absent ) ., The resulting F1 daughters ( who inherited Wolbachia from their SE Asian mother ) were then backcrossed to Moorea males to create a female-informative family for identification of loci linked to the suppressor ., The absence of recombination in female Lepidoptera means that a SE Asia allele on any chromosome that is necessary for male survival will be present in all of the surviving sons of this female ( as if they lack it , they die ) , but this allele will show normal 1∶1 segregation in her daughters ( S1 Figure ) ., Initially 10 loci from across the genome were screened ., Of these , one locus orthologous to sequence on chromosome 25 in the moth Bombyx mori showed this unusual pattern of inheritance ., For this locus , all 16 sons carried the same maternal allele of SE Asia origin while 8 daughters showed Mendelian segregation ( probability of observing this pattern of segregation in sons on the null hypothesis of no association\u200a= ( 1/2 ) 16: p<0 . 0001 ) ., We then obtained an additional 11 markers in this linkage group ., Candidates were identified initially via synteny to B . mori , and then confirmed as showing co-segregation with the original marker and as being associated with male survival , in the female-informative family ., In this way , a suite of 12 suppressor-linked markers ( A-L ) were developed , all of which followed the presumed pattern of inheritance of the suppressor - that of presence in all 16 sons and half of the daughters ., The remaining 9 non-suppressor-linked markers ( M-U ) , representing 8 separate linkage groups , were developed to investigate potential genome-wide effects ., Marker information and accession numbers are given in S1 Table and S2 Table ., A linkage map for this chromosome , the suppressor linkage group ( SLG ) , was then constructed ., The region required for male survival was identified by the exclusion of recombinants ., This was achieved by examining the segregation of alleles from sons of the SE Asia x Moorea cross above that were mated to Wolbachia-infected Moorea ( non-suppressor ) females ( creating a male-informative family ) ., 307 recombinant daughters were obtained , which were used to create a linkage map of the 12 suppressor-linked markers ( data used to create linkage map in S6 Table ) ., The markers were estimated to cover a 41 cM recombination distance and were syntenic with B . mori ( Fig . 1 ) ., The suppressor locus was localized to a region within this chromosome by excluding linked loci where the SE Asia derived paternal allele was absent in one or more sons ( indicating that the genomic region containing the SE Asia allele was not necessary for male survival ) ., Three suppressor-linked alleles ( D , E and F ) , all in the +11 to +12 region , were retained in all 60 sons , whereas the 9 markers proximal and distal to these were excluded by the presence of one or more recombinants ( Fig . 1 ) ., The probability of observing retention of a marker in a sample of 60 on the null hypothesis of no association between the +11/+12 genomic region and male survival is 0 . 560\u200a=\u200a9×10−19 ., Thus we posit that the suppressor lies between marker C at +8 ( excluded by one recombinant ) and marker G at +17 ( excluded by two recombinants ) - a region of approximately 10 cM ., Our data also indicate that while this genomic region is necessary for male survival , presence of this locus was not always associated with male survival , with the number of surviving sons obtained being one quarter , rather than one half , of the number of daughters obtained in our cross ( 60 sons vs 307 daughters ) ., Our data identified a 10 cM genomic region on chromosome 25 of SE Asian H . bolina that was necessary for a male butterfly to survive Wolbachia induced male-killing ., This region was also a focus of selection during the spread of suppression of male-killing between 2001 and 2006 in Samoa ., During this episode , patterns of allelic variation were observed to be altered over a 25 cM region of chromosome 25 , with increases in frequency of one allele at each locus creating the vast majority of heterogeneity between time points ., The largest magnitude of change occurred in markers that co-segregated with suppression in SE Asia , and in this region the overall genetic diversity ( as measured by AE and π ) declined - the classical signature of a selective sweep ., Three further features implicate the role of selection in altering allele frequency across this 25 cM region ., First , the changes in allele frequency are too large to be accounted for by drift , even under conservative assumptions for population size and generation time ., Second , LD is generated across this region , as predicted under a model of strong selection ., Third , 9 markers unlinked to the suppressor linkage group showed no evidence of changes in the frequency of allele variants between 2001 and 2006 , implying that demographic factors were not the major force driving changes in allele frequency ., While we observed changes consistent with the operation of selection over a very broad genomic area , the degree of change was less than that predicted from our model ., This is true both of the magnitude of allele frequency change at loci located near the suppressor locus , and the breadth of the region of chromosome over which changes in allele frequency occurred ., Our model , which presumes a panmictic model and no cost to carrying the suppressor , predicts the suppressor should fix ( and take alleles within 5 cM distance to frequency in excess of 87% ) , and that allele frequency changes should be observed chromosome-wide ., In contrast , the swept allele at locus D ( which lies within 5 cM of the target of selection ) attains a frequency of just 0 . 67 ( n\u200a=\u200a172 , CI 0 . 59–0 . 74 ) in 2006 and 2010 samples ., Further , we observed only very small changes in allele frequency at the most distant locus from the region containing the suppressor , locus L . We suggest there are three non-mutually exclusive explanations for this lack of fit with the model ., First , the suppressor mutation in natural populations diffuses spatially following its initial arrival , and each generation of spatial diffusion is associated with a narrower local sweep ., The principle impact of spatial diffusion will be to narrow the genomic region that is affected by selection compared to that predicted in a panmictic model , and to reduce the magnitude of change at loci far from the target of selection ., For a locus 25 cM distant from the suppressor , association with the suppressor allele may last just one or two generations , such that changes in allele frequency occur only near the point of origin , and are diluted by absence of any selection on these loci in the majority of the species range ., However , spatial diffusion represents a poor explanation for the lower than expected frequency of variants at tightly linked loci post-sweep , which are expected to maintain strong association during the spread of the suppressor across the island , as this occurs in about 10 generations ., A second possibility is the involvement of other loci in the genome , as enhancers of suppressor action ., Our data indicate that the genomic region in question is necessary for male survival , but do not rule out involvement of other loci in enhancing suppression ., If other loci are involved , either as required elements or enhancers of male survival , this would slow suppressor spread , and might account for the narrowness of the sweep observed compared to model predictions ., However , the requirement of the genomic region for male survival in the presence of Wolbachia again makes this a poor explanation for the failure of tightly linked loci to reach high frequency ., Because it is necessary , it should become fixed in the population , and closely associated allele variants should in consequence attain very high frequency ., A third possibility is that there is a cost to being homozygous for the suppressor mutation , either in both sexes , or in female hosts only ., A cost such as this could prevent fixation of the suppressor allele , and thus also help account for the decreased magnitude of effect at loci tightly linked to the suppressor ., If the suppressor allele reaches >0 . 75 frequency , then males lacking the suppressor would be sufficiently rare that the population sex ratio would be near parity ., Biologically , costs to suppressor carriage may be directly associated with the suppression system itself ., Modification of a sex determination gene , for instance , might rescue males but be deleterious in females , or when homozygous ., Alternatively , costs may be associated with linked mutations ., The presence of deleterious loci in linkage with the suppressor is supported by our observation that some material that had been initially swept into the population was lost between 2006 and 2010 ., Finer-scale investigation of this linkage group , especially within the region identified as required for male survival , is necessary to illuminate the precise dynamics that occurred during this episode of selection ., In our data , we observed concordance between the position of the suppressor ascertained in SE Asian butterflies , and the genomic region subject to selection during spread of suppression through the Samoan population of the butterfly ., This observation has two possible interpretations ., First , the suppressor mutation may have been introduced into Samoa by migration ., Given that the suppressor is absent in the nearest island groups , American Samoa and Fiji , suppressor introduction would be associated with a long distant migrant ., Second , the genomic region identified here may represent a hotspot for suppressor mutation , derived independently in Samoa by de novo mutation ., This may be an identical mutation to that found in SE Asia , or an alternative mutation in the same gene , which still confers suppression ., Alternatively , there may be a suppression-conferring mutation in a different gene within the region identified as containing the suppressor ., The presence of novel swept alleles at loci linked to the suppressor indicates that migration is the most parsimonious explanation for suppressor origin ., Variants not present in the 2001 sample were observed to be the main ‘swept’ allele at 4 of the 11 loci at which significant change was detected ( indicated with green arrows in Fig . 3 ) ., At two of these loci ( A & I ) , the invading allele was defined by a single nucleotide polymorphism ( SNP ) being absent from the 2001 sample , whereas the other two alleles represented different combinations of existing SNPs ., The four loci were in three genomic locations spaced over 17 cM and showed no evidence of linkage disequilbrium in the 2001 pre-sweep sample , and thus they can be treated as independent from each other ( Fig . 5 ) ., They therefore support ( but do not definitely prove ) a migratory origin ., None of the loci tested in this study are likely to be the suppressor locus itself ( markers were selected that spanned chromosome 25 and had conserved exon sequence – with several being housekeeping genes ) ., Future research should aim to establish the actual nature of the suppressor mutation in both Samoa and SE Asia through fine-scale genetic mapping ., Such a project will allow the source of suppression on Samoa ( migration or in situ mutation ) to be clarified ., Beyond this , it will reveal the actual target of selection in this system ., It has been widely conjectured that the evolution of sex determination systems might occur in response to the presence of sex ratio distorting microbes 20 ., It is notable that a strong candidate gene – doublesex – resides within the equivalent genomic area in Bombyx mori , and with conservation of synteny being profound in Lepidoptera , is likely to reside in this area in Hypolimnas ., Doublesex represents a tempting candidate as it is known that splicing of this gene is altered in the presence of male-killing Wolbachia in another lepidopteran , the moth Ostrinia scapulalis 21 ., We utilized high-throughput sequencing of the transcriptome of H . bolina to obtain coding sequence from multiple loci across the genome ., Following total RNA extraction from 1 male and 1 female adult H . bolina , mRNA library construction and sequencing using the Roche 454 sequencing platform ( http://www . 454 . com ) , 450 bp reads were de novo assembled into contigs using the Newbler assembler to create the first set of Expressed Sequence Tags ( EST ) for H . bolina ., The trimmed reads have been deposited as one male , and one female , partial transcriptome datasets in the NCBI SRA database , accession SRP045735 ., In the absence of any annotated genome or transcriptome for H . bolina , the moth Bombyx mori was used as a proxy reference genome , this being the only available resource for Lepidoptera at the time of the study ., There is a high level of synteny of gene location in the Lepidoptera 22 allowing a targeted gene approach , in which several genes could be selected from each chromosome across the genome ., Coding sequence of highly conserved genes such as ribosomal proteins and housekeeping genes from B . mori were initially targeted and then retrieved from NCBI ( http://www . ncbi . nlm . nih . gov ) ., To determine putative H . bolina orthologs a local tBLASTx was then performed against the H . bolina EST set ., Only genes that returned a single tBLASTx hit were included , reducing the likelihood of the inclusion of paralogs in our marker set ., The orthologous H . bolina contigs were then translated into amino acid sequences using the ExPASY online tool ( http://web . expasy . org/translate ) , with the sequence lacking mid-sequence stop codons chosen as the most likely translation ., In a final test for paralogs , a reciprocal BLAST was performed of coding sequence from the orthologous H . bolina contigs as queries against the B . mori genome using the INPARANOID8 search tool ( 23; http://inparanoid . sbc . su . se/ ) ., Where present , intronic regions were targeted for marker development , as they are likely to have a higher degree of nucleotide diversity ., Again , conservation of synteny in Lepidoptera genome organisation allowed the intron/exon boundaries in H . bolina genes to be inferred using the B . mori genome ., Through tBLASTx analysis of the B . mori coding sequence of the targeted gene against the B . mori WGS ( Whole Genome Shotgun contigs ) database in NCBI , exonic regions were identified ( as only these regions will align ) ., The translated orthologous H . bolina contig and the corresponding B . mori amino acid sequence were aligned using ClustalW 24 and the position of the intron/exon boundaries subsequently located ., Once intron/exon boundaries were identified in B . mori genes , and extrapolated to the H . bolina orthologous sequences , primers were designed for H . bolina that spanned introns of size 500–1000 bp ( Bombyx size approximation ) ., This size range was chosen to enable successful amplification of the intronic region during PCR ., Marker optimisation was performed using three test H . bolina samples and successful PCR products were sequenced using Sanger technology ., In order to investigate the genetic architecture of male-killing suppression in H . bolina and determine markers in linkage with the suppressor locus , we crossed females of a butterfly population ( the Philippines ) that were Wolbachia-infected and homozygous for the male-killing suppressor allele ( SS ) to males from a Wolbachia-infected population ( Moorea , French Polynesia ) that lacked the suppressor ( ss ) , to create suppressor-heterozygous Wolbachia-infected offspring ( Ss ) ( S1 Figure ) ., Recombination does not occur during female meiosis in the Lepidoptera 25 , permitting the progeny of Ss females to be used to identify the linkage group ( SLG , Suppressor Linkage group ) in which the dominant suppressor allele was carried ., To this end , Ss females were crossed with ss males to produce the female-informative families ., For inclusion in the SLG , markers linked to the suppressor locus are characterized by being present in all surviving sons of the Ss heterozygous mother , rather than the 50% expectation from Mendelian segregation with random survival ., Initially each marker was sequenced in the F1 parents ( Ss female×ss male ) ., In each case , SNPs were chosen that were heterozygous in the female and homozygous in the male – following the presumed pattern of the suppressor ., These same SNPs were then scored in 16 male and 8 female F2 progeny ., Once a marker had been found that was present in half of the daughters ( following Mendelian inheritance ) but all of the sons ( for a son to survive it must have at least one copy of the suppressor , and hence linked marker allele ) , further markers were developed for that same chromosome based on synteny with B . mori ., A final suite of 12 markers that produced clean sequence and that spanned the suppressor-associated chromosome were developed to form the SLG ., Recombination does occur in male H . bolina , and thus crosses of Ss males to ss females ( the male-informative families ) allow, a ) mapping of genetic markers within a chromosome relative to each other and, b ) mapping of the suppressor within the linkage group , in terms of a region of the chromosome that is always present in surviving sons ., To this end , the 12 linked markers were sequenced in the female F2 ( n\u200a=\u200a307 ) from one male informative cross ( Ss male×ss female ) and a linkage map created using JoinMap ( version 3 . 0; Haldane mapping function ) 26 ., To place the suppressor locus within the map F2 males ( n\u200a=\u200a60 ) from this cross were analysed using the same 12 markers ., Absence of recombinants in a core subset of markers , flanked by markers with an increasing numbers of recombinants , indicated the position of the suppressor locus ( Fig . 1 ) ., A population sample of butterflies from three time points ( 2001: n\u200a=\u200a48 , 2006: n\u200a=\u200a48 , 2010: n\u200a=\u200a46 ) were collected from the Samoan island of Upolu ., For each individual , DNA was extracted using the Qiagen DNeasy kit ( www . qiagen . com ) , and the suite of 12 suppressor-linked markers amplified using PCR ., Following Sanger sequencing of the amplicons through both strands , the resultant marker sequences were alignment in Codoncode ( www . codoncode . com/ ) ., SNPs present within and between the population samples were then identified and scored for each individual butterfly ., Using the SNP data ( given in S7 Table ) , the alleles present at each marker in each population sample were estimated using the haplotype reconstruction software PHASE ( version 2 . 1 27 , 28 ) with 1000 iterations , a thinning interval of 100 and 1000 burn-in iterations ., Allele frequencies at each marker for each time group could then be calculated and compared ., Output was also examined by eye , with alleles identified first where there was no ambiguity ( either homozygous , or a SNP separating into two defined alleles ) ., Thereafter , alleles were assumed identical to those already identified where possible ., The low allele diversity meant this visual analysis produced very similar result to PHASE output , which can thus be considered robust ., Patterns of genetic differentiation were estimated using GENEPOP 29 ., First , heterogeneity of allele frequency distributions between pairs of time points was estimated using a G test based on allele frequency distribution ., Where allele distributions were heterogeneous , we ascertained the allele whose frequency change made the greatest contribution to heterogeneity as that with the largest standardized residual within the heterogeneity test 18 ., This allele was then removed ( it was an allele increasing in frequency in each case ) , and the data retested to ascertain if the population samples were then homogeneous , or whether there was evidence for a second allele that changed in frequency ( a second allele was identified in three cases ) ., We additionally used FST standardized population genetic differentiation to quantify the magnitude of change between allele frequency distributions between the two samples ., In each case , the rare individuals where sequence could not be obtained for particular alleles , or not inferred accurately , were coded as missing information ., DNA polymorphism statistics and estimates of nucleotide diversity ( number of segregating sites , number of haplotypes , pi , theta , the average number of nucleotide sequences ( k ) , Tajimas D , haplotype diversity ( Hd ) ) for each marker for each time point were conducted in DnaSP ( version 5 ) 30 ., These statistics were estimated using sequence data excluding gaps i . e . indel mutations were not used ( present in 8 of 9 unlinked markers ) ., Nine unlinked markers , from 8 different chromosomes , were also sequenced for the 2001 and 2006 population samples to investigate the degree to which changes were observed in the wider genome and as a control for demographic effects ., These were tested for the presence of heterogeneity between time points using a G test based on allele frequency distributions , for differentiation using the FST statistic , and several polymorphism statistics as described above for the SLG markers ., We additionally analysed evidence for alteration in the pattern of linkage disequilibrium , again using GENEPOP ., The significance of LD between all possible combinations of loci was tested in the 2001 and 2006 samples separately ., We do not report the magnitude of LD , as this is not a standardized measure , being dependent on the allele frequency distribution at each locus . | Introduction, Results, Discussion, Materials and Methods | Symbionts that distort their hosts sex ratio by favouring the production and survival of females are common in arthropods ., Their presence produces intense Fisherian selection to return the sex ratio to parity , typified by the rapid spread of host ‘suppressor’ loci that restore male survival/development ., In this study , we investigated the genomic impact of a selective event of this kind in the butterfly Hypolimnas bolina ., Through linkage mapping , we first identified a genomic region that was necessary for males to survive Wolbachia-induced male-killing ., We then investigated the genomic impact of the rapid spread of suppression , which converted the Samoan population of this butterfly from a 100∶1 female-biased sex ratio in 2001 to a 1∶1 sex ratio by 2006 ., Models of this process revealed the potential for a chromosome-wide effect ., To measure the impact of this episode of selection directly , the pattern of genetic variation before and after the spread of suppression was compared ., Changes in allele frequencies were observed over a 25 cM region surrounding the suppressor locus , with a reduction in overall diversity observed at loci that co-segregate with the suppressor ., These changes exceeded those expected from drift and occurred alongside the generation of linkage disequilibrium ., The presence of novel allelic variants in 2006 suggests that the suppressor was likely to have been introduced via immigration rather than through de novo mutation ., In addition , further sampling in 2010 indicated that many of the introduced variants were lost or had declined in frequency since 2006 ., We hypothesize that this loss may have resulted from a period of purifying selection , removing deleterious material that introgressed during the initial sweep ., Our observations of the impact of suppression of sex ratio distorting activity reveal a very wide genomic imprint , reflecting its status as one of the strongest selective forces in nature . | The sex ratio of the offspring produced by an individual can be an evolutionary battleground ., In many arthropod species , maternally inherited microbes selectively kill male hosts , and the host may in turn evolve strategies to restore the production or survival of males ., When males are rare , the intensity of selection on the host may be extreme ., We recently observed one such episode , in which the population sex ratio of the butterfly Hypolimnas bolina shifted from 100 females per male to near parity , through the evolution of a suppressor gene ., In our current study , we investigate the hypothesis that the strength of selection in this case was so strong that the genomic impact would go well beyond the suppressor gene itself ., After mapping the location of the suppressor within the genome of H . bolina , we examined changes in genetic variation at sites on the same chromosome as the suppressor ., We show that a broad region of the genome was affected by the spread of the suppressor ., Our data also suggest that the selection may have been sufficiently strong to introduce deleterious material into the population , which was later purged by selection . | signatures of natural selection, genetic polymorphism, natural selection, symbiosis, genetics, biology and life sciences, population genetics, species interactions, evolutionary biology, evolutionary processes, evolutionary genetics | null |
journal.pgen.1000623 | 2,009 | Genome-Wide Association Study Implicates Chromosome 9q21.31 as a Susceptibility Locus for Asthma in Mexican Children | Asthma ( OMIM 600807 ) is a leading chronic childhood disease with prevalence rates reaching a historically high level ( 8 . 9% ) in the United States and continuing to increase in many countries worldwide 1 , 2 ., Asthma is characterized by airway inflammation and bronchoconstriction leading to airflow obstruction , but the mechanisms leading to asthma development remain unknown ., Genetic risk factors likely play a central role in asthma development ., Twin studies support a strong genetic component to asthma ( especially childhood asthma ) with heritability estimates suggesting that 48–79% of asthma risk is attributable to genetic risk factors 3 ., In an effort to localize disease susceptibility genes , genome-wide linkage studies have identified at least 20 linkage regions potentially harboring disease genes 4 , and over 100 positional and biological candidate genes have been tested for association with asthma 3 ., However , no genes have been definitely shown to influence this complex disease ., Genome-wide association studies ( GWASs ) have emerged as a powerful approach for identifying novel candidate genes for common , complex diseases ., In the first asthma GWAS , using 307 , 328 single nucleotide polymorphisms ( SNPs ) , Moffatt et al . found highly statistically significant associations of SNPs in adjacent genes ORM1-like ( S . cerevisiae ) ( ORMDL3; OMIM 610075 ) and gasdermin B ( GSDMB or GSDML; OMIM 611221 ) with risk of childhood asthma in German and British populations 5 ., Meta-analysis of the Moffatt et al . study and five subsequent replication studies , including our own study , supports the association of ORMDL3 and GSDML SNPs with risk for childhood asthma across various populations 6 ., More recently , using 518 , 230 SNPs , Himes et al . implicated SNPs in phosphodiesterase 4D , cAMP-specific ( phosphodiesterase E3 dunce homolog , Drosophila ) ( PDE4D; OMIM 600129 ) with risk of asthma in whites from the United States and replicated this finding in two other white populations 7 ., Using only 97 , 112 SNPs , Choudhry et al . implicated chromosome 5q23 SNPs for association with asthma in Puerto Ricans 8 , but no other Puerto Rican cohorts are available for replication ., Few genetic studies of asthma have included Hispanic populations , and replication of positive genetic findings is scarce across Hispanic groups ., Hispanics have differing proportions of Native American , European , and African ancestries ., This admixture introduces special considerations ( such as controlling for population stratification in association studies ) , but admixture in Hispanic populations also provides a unique opportunity to use ancestry analysis to evaluate our genetic association findings 9 , 10 ., Mexico City is one of the most polluted cities in the world , and its inhabitants experience chronic ozone exposure , which has been linked to asthma development in children and adults and asthma exacerbations 11–13 ., We conducted a GWAS to identify novel candidate susceptibility genes associated with childhood asthma in case-parent trios from Mexico City and tested the most significantly associated SNPs in an independent study of trios of Mexican ethnicity ., GWAS findings were then examined in the context of ancestry analysis and genome-wide expression data to provide supportive evidence for associations with childhood asthma ., The 520 , 767 autosomal SNPs passing quality control were tested for association with childhood asthma using additive modeling with the log-linear method in 492 children with asthma and their biological parents from Mexico City ., Not surprisingly given the study size , no SNP met genome-wide significance using a conservative Bonferroni adjustment ., Nonetheless , the comparison of observed and expected p-values in the quantile-quantile plot ( Figure 1 ) shows several top SNPs with some deviation from expectation ., These deviations may occur by chance or may represent a true excess of small p-values ., Figure 2 shows the observed p-values plotted against chromosomal location ., An intergenic SNP on chromosome 16 had the most significant association with childhood asthma rs1867612 ( p\u200a=\u200a1 . 55×10−6 ) , followed by an intronic SNP in potassium voltage-gated channel , Shab-related subfamily , member 2 ( KCNB2; OMIM 607738 ) on chromosome 8 rs2247572 ( p\u200a=\u200a1 . 94×10−6 ) and two intergenic SNPs on chromosome 20 rs6063725 ( p\u200a=\u200a3 . 52×10−6 ) and rs720810 ( p\u200a=\u200a5 . 13×10−6 ) with only moderate linkage disequilibrium ( LD ) ( r2\u200a=\u200a0 . 59 ) ., The next most significant SNP ( rs2378383 ) highlights a cluster of SNPs on chromosome 9q21 . 31 ranking among the top GWAS SNPs ., This cluster of SNPs spans transducin-like enhancer of split 4 ( E ( sp1 ) homolog , Drosophila ) ( TLE4; OMIM 605132 ) and its upstream region ., LD analysis of SNPs with p≤0 . 001 in this cluster shows two large LD blocks in this region with one block encompassing TLE4 and the other block encompassing the upstream region ( Figure S1 ) ., Eleven of the 18 most significantly associated SNPs met our criteria to be selected for replication in 177 case-parent trios of Mexican ethnicity from the Genetics of Asthma in Latino Americans ( GALA ) study 14 ., The GWAS p-values for the 11 SNPs selected for replication testing ranged from 3 . 30×10−5 to 1 . 55×10−6 ( Figure 2 ) ., There were no significant deviations in Hardy-Weinberg equilibrium ( HWE ) for the replication SNPs in either the GWAS or replication study ( p>0 . 12 ) , and minor allele frequencies ( MAFs ) were similar between the two studies ( Table 2 ) ., The replication study had at least 70% power to detect an association with four SNPs ( rs2378377 , rs4674039 , rs1830206 , and rs3814593 ) and at least 80% power to detect an association with the remaining seven SNPs ( rs1867612 , rs2247572 , rs6063725 , rs720810 , rs2378383 , rs6951506 , and rs3734083 ) for similar MAFs and relative risk ( RR ) estimates observed in the GWAS ., Association results in the GWAS and replication studies are compared in Table 2 ., No SNPs were significant with conservative Bonferroni correction for multiple testing , but two SNPs were associated with childhood asthma in the replication study with a p-value close to 0 . 05 ., The chromosome 9q21 . 31 SNP rs2378383 , which is located 147 kb upstream of TLE4 in an intergenic region between coiled-coil-helix-coiled-coil-helix domain containing 9 ( CHCHD9 ) and TLE4 , was associated with childhood asthma in the replication study with p\u200a=\u200a0 . 03 ., Meta-analysis of rs2378383 in the two studies gave a combined p-value of 6 . 79×10−7 , and the RR estimate for carrying one copy of the rs2378383 minor allele ( G ) compared to carrying no copies in the GWAS RR , 0 . 61; 95% confidence interval ( CI ) , 0 . 49–0 . 76 was quite similar to the RR estimate in the replication study ( RR , 0 . 63; 95% CI , 0 . 41–0 . 96 ) ., The SNP rs2378377 , a neighboring intergenic SNP in moderate LD with rs2378383 ( r2\u200a=\u200a0 . 73 ) , had a marginal association with childhood asthma in the replication study with p\u200a=\u200a0 . 06 ( combined p\u200a=\u200a2 . 68×10−6 ) ., RR estimates for the rs2378377 minor allele ( G ) were also similar between the GWAS ( RR , 0 . 64; 95% CI , 0 . 53–0 . 79 ) and the replication study ( RR , 0 . 71; 95%CI , 0 . 50–1 . 02 ) ., None of the other nine SNPs were associated in the replication study ( Table 2 ) ., Association results from additive modeling for SNPs in the region spanning TLE4 and its upstream region ( chromosome 9 nucleotide positions from 81 , 114 , 500 to 81 , 531 , 500 , NCBI build 36 . 3 ) were obtained from the previous GWASs of asthma 5 , 7 , 8 ., Our top two SNPs from this region were genotyped only in the GWAS in whites from the United States 7 , where they were not associated with asthma ( p\u200a=\u200a0 . 59 for rs2378383 and p\u200a=\u200a0 . 65 for rs2378377 ) ., Eighty-nine other SNPs were available in this region , and the smallest p-value was observed for rs1328406 ( p\u200a=\u200a0 . 056 ) ., There were 54 SNPs available in this region from the GWAS in German and British populations 5 , with the smallest p-values observed for rs2807312 ( p\u200a=\u200a0 . 0041 ) , rs7849719 ( p\u200a=\u200a0 . 018 ) , rs7862187 ( p\u200a=\u200a0 . 033 ) , rs10491790 ( p\u200a=\u200a0 . 043 ) , and rs946808 ( p\u200a=\u200a0 . 049 ) ., From the 26 SNPs available from the GWAS in Puerto Ricans 8 , the smallest p-value was 0 . 19 ., In our GWAS in Mexicans , there were several SNPs in TLE4 and its upstream region with small p-values , and the SNPs listed above are located in close proximity to many of our associated SNPs ., Similar to our GWAS and replication study , only cases with childhood-onset asthma were included in both GWASs in white populations 5 , 7 , and the cases from Himes et al . were predominantly atopic ( 91 . 2% ) as defined by at least one positive skin prick test 7 ., In contrast , the GWAS in Puerto Ricans included both childhood-onset and adulthood-onset asthma cases , and 83% of the cases were considered atopic as defined by total IgE>100 IU/mL 8 ., Associations of rs2378383 and rs2378377 were examined in data from the GWAS population stratified by residential ambient ozone exposure ( 199 trios with ≤67 ppb and 214 trios with>67 ppb annual average of the maximum 8 hour averages ) and by current parental smoking ( 253 trios with and 233 trios without current parental smoking ) ., The minor alleles of both SNPs were inversely associated with asthma at p<0 . 05 in all strata ( results not shown ) , thus giving no evidence for effect modification in the presence of these environmental exposures ., Among the 445 cases with skin test data available , 408 can be classified as atopic by virtue of having at least one positive skin test ., We repeated the GWA scan in this subset of 408 trios ., Chromosome 9q21 . 31 SNPs predominated among the top ranked SNPs , with rs2378383 ( p\u200a=\u200a7 . 18×10−7 ) and rs2378377 ( p\u200a=\u200a1 . 08×10−6 ) being the two top ranking SNPs ., In addition to smaller p-values , the magnitudes of association with asthma were slightly stronger for rs2378383 ( RR , 0 . 54; 95% CI , 0 . 42–0 . 69 ) and rs2378377 ( RR , 0 . 54; 95% CI , 0 . 42–0 . 72 ) in the subset of trios where the asthmatic child is also known to be atopic ., The chromosome 9q21 . 31 SNPs rs2378383 and rs2378377 were tested for association with the number of positive skin tests as a quantitative measure of the degree of atopy in the trios with skin test data ., Both SNPs were associated with degree of atopy ( p\u200a=\u200a0 . 0018 for rs2378383 and p\u200a=\u200a0 . 0010 for rs2378377 ) ., Their RR estimates indicate an inverse association in which carrying one copy of the minor allele is associated with a decreasing number of positive skin tests ( RR , 0 . 92; 95% CI , 0 . 87–0 . 97 for rs2378383 and RR , 0 . 92; 95% CI , 0 . 88–0 . 97 for rs2378377 ) , consistent with the direction of association for asthma ., The Mexican subjects in the GWAS had mean ancestral proportions of 69 . 5±15 . 6% for Native American , 27 . 3±14 . 3% for European , and 3 . 2±3 . 0% for African ancestries ., Given the predominance of Native American ancestry , we evaluated Native American transmission in the GWAS along the chromosomal arm ( 9q ) containing the replicated SNP ( rs2378383 ) by ancestry analysis ., As shown in Figure 3 , there is a significant under-transmission of Native American ancestry at rs2378383 ( z-score\u200a=\u200a−2 . 21 and two-sided p\u200a=\u200a0 . 028 ) and surrounding SNPs ., The deficiency in Native American ancestry at this locus suggests that a protective allele occurred at a higher frequency in the Native American ancestral population than in the European and African ancestral populations ., An examination of this SNP in the HapMap and Human Genome Diversity Panel ( HGDP ) data reveals that the frequency of the G allele is generally low in European , African , and East Asian populations ( 0 . 125 in HapMap European , 0 . 033 in HapMap African , 0 . 122 in HapMap Chinese , and 0 . 273 in HapMap Japanese ) , while its frequency is much higher in Native American populations ( 0 . 57 in HGDP Pima and 0 . 36 in HGDP Mayan ) ., This pattern suggests that the G allele may be tagging a protective allele in the Native American ancestral population ., This conclusion is consistent with the finding that the G allele is associated with a decreased risk for childhood asthma in the GWAS and replication analyses ( Table 2 ) ., We examined gene expression patterns in 51 diverse human tissues in the context of GWAS findings to determine whether genes expressed in asthma relevant tissues ranked higher than genes not expressed in such tissues and thus to assess the biological plausibility of our overall GWAS findings ., These results are presented in Figure 4 ., In the 14 , 330 genes with GWAS SNPs in the gene or nearby , median false discovery rate q-values ( derived from the GWAS p-values ) were compared between genes expressed versus genes not expressed in each of 51 human tissues ., Among the 51 tissues , the most significant deviation between the median q-values was found between 3 , 618 genes expressed in the lung versus 10 , 712 genes not expressed in the lung ( Figure 4 , p\u200a=\u200a0 . 00025 ) ., This finding remains significant even after a conservative Bonferroni correction for multiple testing ., The q-values for the 3 , 618 lung-expressed genes are presented in Table S1 ., TLE4 did not contribute to the observation of significantly lower GWAS q-values in lung-expressed genes , as TLE4 was classified as not expressed ., TLE4 displays a nearly ubiquitous expression pattern with similar low intensity levels across many tissues , so even though it is present in the lung , its expression level in the lung did not exceed our 75th percentile threshold to be classified as expressed ., Other tissues in the respiratory and immune system also showed deviations in the median GWAS q-values for expressed versus not expressed genes , including thymus and lymph node at the p<0 . 01 level ( uncorrected ) and fetal lung , trachea , tonsil , smooth muscle , and bronchial epithelium at the p<0 . 05 level ( uncorrected ) ., Thus , SNPs in genes more highly expressed in tissues related to the pathogenesis of asthma and allergies tend to give more significant GWAS results than genes more highly expressed in other tissues ., These results give biological credibility to our overall GWAS findings and are consistent with the multigenic etiology of asthma ., A two-dimensional cluster analysis was conducted to identify the implicated tissues with correlated gene expression patterns ., As shown in Figure S2 , lung tissue is grouped in a cluster with fetal lung and placenta tissues , thus suggesting that gene expression patterns in lung are most similar to gene expression patterns in fetal lung and placenta and that their signals are correlated ., There are 2 , 385 genes classified as expressed in all three tissues – lung , fetal lung , and placenta ., The gene expression patterns of other implicated tissues are also highly correlated , including the immune tissues tonsil , lymph node , and thymus ( Figure S2 ) ., Genetic studies of asthma are few in Hispanic populations , and to our knowledge , this work presents the first asthma GWAS in Mexicans and the most extensive coverage of genetic variation for an asthma GWAS in any Hispanic population ., The GWAS included 492 Mexican case-parent trios ., Given the moderate GWAS sample size , no SNP met genome-wide significance ., However , the ranking of GWAS SNPs highlighted a potentially important candidate region for childhood asthma susceptibility , chromosome 9q21 . 31 ., Several chromosome 9q21 . 31 SNPs with small GWAS p-values were located in TLE4 and its upstream region , and two of these SNPs ( rs2378383 and rs2378377 ) were tested for replication in an independent study of 177 case-parent trios of Mexican ethnicity ., Despite the small sample size for replication , both SNPs gave p-values close to 0 . 05 and the same direction and magnitude of association as the GWAS ., Neither rs2378383 nor rs2378377 have a known impact on TLE4 expression , but given their location upstream of TLE4 , it is possible that these SNPs reside in a TLE4 regulatory region ., Ancestry analysis in this chromosomal region provided supportive evidence that rs2378383 ( G ) is associated with a decreased risk of childhood asthma in Mexicans ., Ancestry and transmission-based association analyses provide complementary but not completely independent lines of evidence ., At each SNP , the log-linear method only used parents who were heterozygous in genotype , while ancestry analysis used all parents who are heterozygous in ancestry , including parents who are homozygous in genotype ., We did not a priori expect that ancestry analysis results would corroborate log-linear association results ., Our ancestry analysis uses the same principles that underlie admixture mapping and relies on the key assumption of different risk allele frequencies between the ancestral populations , primarily Native American and European in this study ., Under this assumption , individuals with disease in the admixed population would be expected to share an excess of ancestry from the population with the highest risk allele frequency at the disease locus 9 ., In contrast , individuals with disease in the admixed population would be expected to share a shortage of ancestry from the population with the highest protective allele frequency at the disease locus ., At chromosome 9q21 . 31 , there was less Native American ancestry than expected , suggesting that the Native American ancestral population had a higher frequency of the protective rs2378383 allele ( G ) ., Ancestry analysis implicated chromosome 9q21 . 31 as a chromosomal region that may underlie ethnic differences in childhood asthma ., Complex diseases with differing disease prevalence rates in the ancestral populations are most suitable for ancestry analysis 15 ., Prevalence rates of childhood asthma in the true ancestral Native American , European , and African populations are unknowable , but it is interesting to note that Mexicans have the highest Native American ancestry and the lowest asthma prevalence rate among Hispanic populations 16 ., Differing frequencies of genetic risk factors in the ancestral populations presumably contribute to the differing prevalence rates of childhood asthma in modern populations ., Our study found an association between Native American ancestry and a lower disease risk ., Similarly , Native American ancestry was associated with milder asthma in a previous study of subjects of Mexican ethnicity from the GALA study 17 ., These findings collectively suggest that the Native American ancestral population had higher frequencies of alleles that decrease prevalence and severity of asthma in the modern Mexican population ., A comparison of asthma prevalence and severity among modern Native Americans , Europeans , and Africans would further support this interpretation , but such data are scarce 17 ., The evidence for locus-specific ancestry around rs2378383 has implications for replication ., Because rs2378383 ( G ) occurs at relatively low frequency in European , African and East Asian populations , genetic association studies in these populations are likely to suffer from lack of power at this locus ., In contrast , the G allele occurs at moderate frequency in the Native American populations surveyed in HGDP 18 ., Such disparate allele frequencies facilitate ancestry analysis in the region and improve the statistical power of transmission-based tests , as there are many more heterozygous parents in the Mexican population than a European , African or East Asian population ., In fact , we obtained association results for SNPs in the chromosome 9q21 . 31 region from previous GWASs and found that SNPs in this region had only nominal evidence for association with asthma in the GWASs in white populations 5 , 7 ., It is not surprising that substantial evidence for replication was not found given the ethnicity differences ( whites for Moffatt et al . and Himes et al . 5 , 7 and Puerto Ricans for Choudhry et al . 8 ) ., Future replication and fine-mapping of the region would be most effective if performed in Native American populations , or admixed populations with high Native American ancestral contributions ., The chromosome 9q21 . 31 SNPs associated with childhood asthma in the GWAS map to TLE4 and its upstream region ., The TLE family of proteins in humans is homologous to the Drosophila Groucho protein , which participates in cell fate determination for neurogenesis and segmentation ., The highly conserved structure among the Drosophila Groucho and human TLE gene products suggest similar functions as transcriptional regulators in cell fate determination and differentiation 19 ., Six genes encode the TLE family of proteins in humans ( TLE1 , TLE2 , TLE3 , TLE4 , TLE5 , TLE6 ) , as deposited in the NCBI database ., The distinct expression patterns among the TLE genes suggest a complex mechanism in humans involving non-redundant roles for the TLE genes 20 ., The TLE4 gene , in particular , shows ubiquitous expression across many tissues 19 , 21 , and TLE4 functions as a transcriptional co-repressor in several key developmental pathways 22 ., More specifically , TLE4 has been implicated in early B-cell differentiation ., TLE4 interacts with the transcription factor Paired box 5 ( PAX5; OMIM 167414 ) , which activates B-cell specific genes and represses alternative lineage fates 23 ., A spliced version of TLE4 acts as a negative regulator for the PAX5/TLE4 function 23 ., An alteration of B-cell differentiation involving TLE4 could be relevant to immune development and thus asthma ., TLE interacts with Runt-related transcription factor 3 ( RUNX3; OMIM 600210 ) in a manner that may be directly relevant to asthma ., In mice , loss of RUNX3 function results in an allergic asthma phenotype due to accelerated dendritic cell maturation and resulting increased efficacy to stimulate T cells 24 ., Interaction with TLE is required for RUNX3 to inhibit dendritic cell maturation 25 ., A recent paper provides support for the interaction of RUNX3 specifically with TLE4 26 ., Interestingly , the chromosome 9q21 . 31 SNPs rs2378383 and rs2378377 near TLE4 are associated with asthma as well as degree of atopy in our data , and their associations with asthma became more pronounced when considering only the asthmatic children with atopy and their parents ., These findings suggest that the influence of TLE4 on asthma may be related to its influence on immune system development ., Childhood asthma is a complex disease , and there are likely many susceptibility genes influencing immune system development and asthma in the Mexican population ., The examination of GWAS in the context of genome-wide expression illustrated the biological plausibility of our GWAS findings and showed consistency with the involvement of multiple genes ., Genes expressed in the lung show association signals that differ most significantly from the association signals from genes not expressed in the lung when compared to 50 other human tissues ., The lung represents a major pathogenic site for asthma , and this finding implies that multiple genes expressed in the lung are collectively associated with an increased risk of childhood asthma ., Some of the other implicated tissues may represent false positives , but several of the highlighted tissues are biologically plausible for childhood asthma , including trachea , bronchial epithelium , smooth muscle , and immune tissues such as thymus , tonsil , and lymph node ., Other GWASs have implicated different susceptibility loci ., Several SNPs implicated in the first asthma GWAS by Moffatt et al . in the ORMDL3 region 5 were associated with childhood asthma in our GWAS including rs9303277 ( p\u200a=\u200a0 . 036 ) , rs11557467 ( p\u200a=\u200a0 . 014 ) , rs8067378 ( p\u200a=\u200a0 . 020 ) , rs2290400 ( p\u200a=\u200a0 . 037 ) , and rs7216389 ( p\u200a=\u200a0 . 042 ) but were not ranked among our top 5 , 000 SNPs ., More recent GWASs have implicated loci other than ORMDL3 ., The PDE4D SNPs implicated by Himes et al . 7 were not associated with childhood asthma in our GWAS at p<0 . 05 ., Two nearby SNPs , not in LD with the implicated SNPs , were associated rs13158277 ( p\u200a=\u200a0 . 030 ) and rs7717864 ( p\u200a=\u200a0 . 015 ) but were also not ranked among our top 5 , 000 SNPs ., Chromosome 5q23 SNPs implicated by Choudhry et al . 8 were not associated with childhood asthma in our GWAS at p<0 . 05 ., Initial GWAS findings regarded as replicated may not be ranked among the front runners in a genome-wide scan in the replication populations for statistical 27 as well as other biological reasons ( such as ethnic differences , phenotypic heterogeneity , genetic heterogeneity , differing patterns of interacting environmental exposures , or multigenic etiology ) ., This trend in discordant GWAS findings is quite common for various complex diseases 28 , and follow-up studies are crucial in separating true genetic associations from false positives ., The major limitation of this study is the sample size for the GWAS and replication study ., The Mexican population is largely under-studied given its size , and only moderate sample sizes are currently available for the study of asthma genetics ., In our study , no SNPs met genome-wide significance , and no replication SNPs met the significance threshold when using a conservative Bonferroni correction for multiple testing ., Despite this limitation , top GWAS findings , replication in an independent population , and ancestry analysis taken together implicate a novel region for association with asthma in Mexican children ., This study has several strengths ., The case-parent trio design and the log-linear analysis protects against bias due to population stratification 29 , so our GWAS results are not confounded by population stratification in this admixed population ., Also , disease misclassification is minimal ., Although bronchial hyper-reactivity was not tested , children with asthma were given reliable diagnoses based on clinical grounds by pediatric allergists at a pediatric allergy specialty clinic ., The allergy clinic is a tertiary referral clinic , so the children with asthma were previously seen by a generalist and a pediatrician over time for recurrent asthma symptoms ., Physician diagnosis of asthma has been shown to have a high level of validity in children after the first few years of life 30 ., Further , the asthmatic children were predominantly atopic to aeroallergens based on skin prick testing limiting heterogeneity of the disease phenotype ., The GWAS and replication association results and the supporting ancestry analysis implicate chromosome 9q21 . 31 as a novel susceptibility locus for childhood asthma in the Mexican population ., This region contains a biologically plausible novel susceptibility gene for childhood asthma , TLE4 , but further work is needed to decipher whether TLE4 or a nearby gene explain the signals from the chromosome 9q21 . 31 region ., Further , childhood asthma is a complex disease with a proposed multigenic etiology , but most single studies will not have sufficient power to examine such complex relationships ., Identification of important interacting risk factors in childhood asthma and other complex diseases will require very large sample sizes ., This work identifies chromosome 9q21 . 31 ( including TLE4 ) as a novel candidate susceptibility locus for childhood asthma , suggests that this region may underlie ethnic differences in childhood asthma , and emphasizes the presence of multiple genetic risk factors in the complex mechanism leading to childhood asthma ., The study protocol was approved by the Institutional Review Boards of the Mexican National Institute of Public Health , Hospital Infantil de Mexico Frederico Gomez , and the US National Institute of Environmental Health Sciences ( NIEHS ) ., Parents gave written informed consent for the childrens participation , and children gave their assent ., Children with asthma ( aged 5–17 ) and their biological parents were recruited between June 1998 and November 2003 from a pediatric allergy specialty clinic at a large public hospital in central Mexico City , Hospital Infantil de Mexico Frederico Gomez ., The case-parent trio design protects against bias due to population stratification in this admixed population 29 , 31 ., Blood samples were collected from enrolled children and their parents for DNA extraction ., Children were diagnosed with asthma by a pediatric allergist at the referral clinic based on clinical symptoms and response to treatment 32 ., Asthma severity was rated as mild ( intermittent or persistent ) , moderate , or severe by the pediatric allergist according to symptoms in the Global Initiative on Asthma schema 33 ., Questionnaires on the childrens asthma symptoms and risk factors , including environmental tobacco smoking , were completed by parents , nearly always the mother ., The clinical evaluation also included skin prick testing to measure atopy and pulmonary function testing , as previously described 6 ., A battery of 24 aeroallergens common in Mexico City was used for skin prick testing ., Histamine was used as a positive control , and the test was considered valid if the histamine reaction was 6 mm or greater 34 ., Glycerin was used as a negative control ., Children were considered atopic if the diameter of skin reaction to at least one allergen exceeded 4 mm ., Pulmonary function testing was performed at a later date using the EasyOne spirometer ( ndd Medical Technologies , Andover , Massachusetts ) according to American Thoracic Society guidelines 35 ., Children were asked to withhold asthma medications on the morning of the test ., The best test of three technically acceptable tests was selected ., Percent predicted forced expiratory volume in 1 second ( FEV1 ) was calculated using spirometric prediction equations from a childhood population in Mexico City 36 ., Measurements of ambient ozone were obtained from the Mexican governments air monitoring station closest to each childs residence ( within 5 km ) ., The annual average of the daily maximum 8 hour averages of the ozone level was collected for the year prior to study entry and dichotomized at the median for stratified analyses ., Further details on the ozone measurement protocol have been described elsewhere 6 ., Peripheral blood lymphocytes were isolated from whole blood , and DNA was extracted using Gentra Puregene kits ( Gentra Systems , Minneapolis , Minnesota ) ., A total of 498 complete case-parent trios with previously confirmed parentage and sufficient amounts of DNA were genotyped for 561 , 466 SNPs using the Illumina HumanHap 550 K BeadChip , version 3 ( Illumina , San Diego , California ) at the University of Washington , Department of Genome Sciences ., Genotypes were determined using the Illumina BeadStudio Genotyping Module , following the recommended conditions ., Results for three unrelated study subjects fell below the genotyping call rate threshold of 95% resulting in exclusion of three trios ., The remaining study subjects were genotyped with an average call rate of 99 . 7% ., Quality control analyses were conducted using PLINK ( http://pngu . mgh . harvard . edu/~purcell/plink/ ) 37 , unless otherwise stated ., In preliminary SNP-level quality control , SNPs were excluded due to poor chromosomal mapping ( N\u200a=\u200a173 ) , missingness>10% ( N\u200a=\u200a988 ) , MAF<0 . 001% ( N\u200a=\u200a253 ) , HWE p-value ( in parents only ) <1×10−10 ( N\u200a=\u200a557 ) , Mendelian errors in more than two families ( N\u200a=\u200a4 , 945 ) , and heterozygous genotype calls for chromosome X SNPs in more than one male ( N\u200a=\u200a380 ) ., All SNP exclusions were ma | Introduction, Results, Discussion, Materials and Methods | Many candidate genes have been studied for asthma , but replication has varied ., Novel candidate genes have been identified for various complex diseases using genome-wide association studies ( GWASs ) ., We conducted a GWAS in 492 Mexican children with asthma , predominantly atopic by skin prick test , and their parents using the Illumina HumanHap 550 K BeadChip to identify novel genetic variation for childhood asthma ., The 520 , 767 autosomal single nucleotide polymorphisms ( SNPs ) passing quality control were tested for association with childhood asthma using log-linear regression with a log-additive risk model ., Eleven of the most significantly associated GWAS SNPs were tested for replication in an independent study of 177 Mexican case–parent trios with childhood-onset asthma and atopy using log-linear analysis ., The chromosome 9q21 . 31 SNP rs2378383 ( p\u200a=\u200a7 . 10×10−6 in the GWAS ) , located upstream of transducin-like enhancer of split 4 ( TLE4 ) , gave a p-value of 0 . 03 and the same direction and magnitude of association in the replication study ( combined p\u200a=\u200a6 . 79×10−7 ) ., Ancestry analysis on chromosome 9q supported an inverse association between the rs2378383 minor allele ( G ) and childhood asthma ., This work identifies chromosome 9q21 . 31 as a novel susceptibility locus for childhood asthma in Mexicans ., Further , analysis of genome-wide expression data in 51 human tissues from the Novartis Research Foundation showed that median GWAS significance levels for SNPs in genes expressed in the lung differed most significantly from genes not expressed in the lung when compared to 50 other tissues , supporting the biological plausibility of our overall GWAS findings and the multigenic etiology of childhood asthma . | Asthma is a leading chronic childhood disease with a presumed strong genetic component , but no genes have been definitely shown to influence asthma development ., Few genetic studies of asthma have included Hispanic populations ., Here , we conducted a genome-wide association study of asthma in 492 Mexican children with asthma , predominantly atopic by skin prick test , and their parents to identify novel genetic variation for childhood asthma ., We implicated several polymorphisms in or near TLE4 on chromosome 9q21 . 31 ( a novel candidate region for childhood asthma ) and replicated one polymorphism in an independent study of childhood-onset asthmatics with atopy and their parents of Mexican ethnicity ., Hispanics have differing proportions of Native American , European , and African ancestries , and we found less Native American ancestry than expected at chromosome 9q21 . 31 ., This suggests that chromosome 9q21 . 31 may underlie ethnic differences in childhood asthma and that future replication would be most effective in populations with Native American ancestry ., Analysis of publicly available genome-wide expression data revealed that association signals in genes expressed in the lung differed most significantly from genes not expressed in the lung when compared to 50 other tissues , supporting the biological plausibility of the overall GWAS findings and the multigenic etiology of asthma . | genetics and genomics/gene expression, genetics and genomics/complex traits, genetics and genomics/genetics of disease, respiratory medicine/asthma, respiratory medicine/respiratory pediatrics | null |
journal.pntd.0002924 | 2,014 | Protein Kinase C and Extracellular Signal-Regulated Kinase Regulate Movement, Attachment, Pairing and Egg Release in Schistosoma mansoni | Protein kinases C ( PKCs ) and extracellular signal-regulated kinases/mitogen-activated protein kinases ( ERKs/MAPKs ) are signalling enzymes that play a critical role in regulating cellular processes , such as gene expression , the cell cycle , growth , development and differentiation , cellular motility , survival and apoptosis 1 , 2 ., PKC/ERK signalling occurs in response to various stimuli , including ligands that bind receptor tyrosine kinases ( RTKs ) and G-protein coupled receptors ( GPCRs ) 1 , 2 ., Putative PKCs and ERKs exist in kinomes of the blood flukes Schistosoma mansoni 3 , 4 , S . japonicum 5 and S . haematobium 6 ., These parasites cause human schistosomiasis , a neglected tropical disease ( NTD ) characterised by inflammatory granulomatous reactions in the host organs that occur in response to entrapped eggs from adult female worms 7 ., The global significance of human schistosomiasis is huge; more than 200 million people have the disease , and 0 . 8 billion are at risk of infection 8 , 9 ., Chemotherapy relies upon praziquantel ( PZQ ) treatment , but this compound kills adult worms and not juvenile stages , and does not prevent re-infection 10 , 11; emergence of drug resistant strains is also possible 12 ., Although anti-schistosome vaccine targets exist 13 , some worms exhibit antigenic polymorphism , making targeting difficult 14 ., Thus , there is considerable interest in identifying new anti-schistosome drug targets , and the protein kinases are potential candidate molecules 15 , 16 ., In humans 10 PKCs exist , including PKCβI and PKCβII , which arise from alternate gene splicing 17 ., PKCs comprise a C-terminal serine/threonine kinase domain linked through a flexible hinge segment to an N-terminal regulatory region that contains diacylglycerol ( DAG ) -sensitive and Ca2+-sensitive domains 18 ., Conventional PKCs ( cPKCs: PKCα , PKCβI/βII , and PKCγ ) are DAG sensitive and Ca2+ responsive; novel PKCs ( nPKCs: PKCδ , PKCε , PKCη , and PKCθ ) are DAG sensitive but Ca2+ insensitive; and atypical PKCs ( aPKCs: PKCζ and PKCι , or λ-murine ) are insensitive to both DAG and Ca2+ 18 ., Phosphorylation on specific amino acid residues is also crucial to PKC activation 19 ., Previously , a DAG/Ca2+-dependent PKC-like activity and a phospholipid/phorbol ester sensitive kinase activity were detected in adult S . mansoni homogenates 20 , 21 , and a PKC ( SmPKC1 ) homologous to human PKCβ was characterised molecularly 22 ., Previously , we identified four putative PKCs in the S . mansoni genome with homology to human PKCs , particularly within functional domains 23; two proteins were similar to human cPKCβI , one to nPKCε and one to aPKCζ 23 , with PKCε also being designated PKCη 4 ., Using phospho-specific antibodies , we showed that activated PKCβ associated with the neural mass , tegument , ciliated plates and germinal cells of miracidia , and that PKC activation restricted development to mother sporocysts that parasitize the snail intermediate host 23 ., MAPK pathways exist in all eukaryotes , with components being conserved among yeast , invertebrates and mammals 24–29 ., The ERK pathway features Ras as a monomeric G-protein , Raf as a MAPKKK , MAPK/ERK Kinase ( MEK ) as a MAPKK , and ERK as a MAPK , the last three forming a hierarchical kinase cascade 30 ., Humans and many other organisms express ERK1 and ERK2 ( p44 and p42 MAPK ) to varying extents in tissues and more than 150 ERK1/2 substrates exist 2 , including cytosolic , membrane , nuclear and cytoskeletal proteins 30 ., Phosphorylation of ERK1/2 on threonine and tyrosine resides within the Thr-Glu-Tyr ( TEY ) motif in the activation loop is essential for activation ., In S . mansoni , Ras GTPase activator protein- ( Ras-GAP ) and ERK1/2-like proteins have been detected 31 , and a Ras homologue has been characterised 32 ., Activation of the S . mansoni epidermal growth factor receptor ( EGFR; SER ) by human EGF leads to ERK2 phosphorylation in Xenopus oocytes 33 , and hypothetical ERK pathways for S . mansoni and S . japonicum have been reconstructed in silico 3 , 4 , 34 , supporting that ERK signalling is intact in schistosomes ., Identifying signalling molecules that play a fundamental role in cellular communication and function of schistosomes represents one of the great challenges of the schistosome post-genomic era 16 ., The aims of the current study were to characterise global PKC and ERK signalling in S . mansoni and to determine whether PKC and ERK are critical to schistosome function ., We demonstrated several PKC and ERK isotypes , profiling activities in different life-stages ., We determined PKC and ERK responses to kinase activators and inhibitors in adult worms , and determined the localization of active kinases in intact worms in situ through functional kinase mapping ., Finally , we showed physiological roles for PKC and ERK in schistosomes through the modulation of kinase activities using pharmacological agents ., The findings , together with those also presented for the effects of PZQ on PKC/ERK signalling , establish a role for PKC and ERK in worm homeostasis and behaviour , and identify these kinase groups as potential anti-schistosome drug targets ., Laboratory animal use was within a designated facility , regulated under the terms of the UK Animals ( Scientific Procedures ) Act , 1986 , complying with all requirements therein; regular independent Home Office inspections occurred ., The Natural History Museum Ethical Review Board approved experiments involving mice and work was carried out under Home Office project license 70/6834 ., Anti-phospho antibodies ( Ab ) used to detect phosphorylated PKC and ERK in S . mansoni were anti-phospho-PKC ( pan ) ( βII Ser660 ) rabbit Ab ( #9371 ) , anti-phospho-PKC ( pan ) ( ζ Thr410 ) ( 190D10 ) rabbit mAb ( #2060 ) , immobilized anti-phospho p44/p42 MAPK ( ERK1/2 ) ( Thr202/Tyr204 ) ( D13 . 14 . 4E ) XP rabbit mAb ( #3510 ) , and anti-phospho-p44/42 MAPK ( ERK1/2 ) ( Thr202/Tyr204 ) rabbit mAb ( #9101 ) ( Cell Signalling Technology , New England Biolabs ) ., The p44/p42 MAPK ( ERK1/2 ) ( non-radioactive ) immunoprecipitation/kinase assay kit , phorbol 12-myristate 13-acetate ( PMA ) , cell lysis buffer , RIPA buffer , lambda phosphatase , and anti-rabbit horseradish peroxidase ( HRP ) -linked secondary antibodies were also purchased from Cell Signalling Technology ., The Omnia S/T peptide 8 kinase assay kit ( KNZ1081 ) , RPMI-1640 medium , foetal bovine serum , antibiotics/antimycotics , and anti-rabbit Alexa Fluor 488 secondary antibodies were purchased from Invitrogen ., Pre-cast Precise 10% polyacrylamide gels , West Pico chemiluminescence substrate , Restore Western blot stripping buffer and Halt protease/phosphatase inhibitor cocktail were from Pierce , whereas nitrocellulose membrane was from GE Health ., U0126 , GF109203X , and PKC catalytic subunit derived from rat brain were purchased from Merck ., Vectashield mounting medium was from Vector Laboratories ., PKC and ERK gene candidates were identified from the S . mansoni genome assembly , relying on existing annotations ( http://www . genedb . org/genedb/smansoni ) , Andrade et al . 4 , and our existing analysis 23 ., Protein sequences were assessed for similarity with other organisms using pBLAST ( http://www . uniprot . org ) ., The detection site of the anti-phospho antibodies was identified within the putative S . mansoni PKC and ERK sequences and was aligned to human PKC and ERKs using MUSCLE ( www . ebi . ac . uk/Tools/msa/muscle ) ., The Belo Horizonte strain of S . mansoni was maintained in Biomphalaria glabrata and albino female mice ( BKW strain ) ., Miracidia and in vitro transformed sporocysts were collected , concentrated , and processed for Western blotting as previously described 35 , 36 ., Briefly , cercariae were collected after they emerged from patent snails and were transferred to 15 ml conical tubes and cooled on ice for 15 min prior to centrifugation ( 200× g ) ., Concentrated cercariae were transferred to 1 . 5 ml microfuge tubes , pelleted by pulse centrifugation , and homogenized on ice in 25 µl RIPA buffer containing Halt protease/phosphatase inhibitors ., An aliquot was removed for protein quantification ( Bradford assay with Bovine serum albumin ( BSA ) as protein standard ) , 5× SDS-PAGE sample buffer added and samples heated at 95°C for 5 min ., Extracts were then either electrophoresed immediately or stored at −80°C ., Adult worms were collected by hepatic portal perfusion of mice 40–42 days post infection and were gently washed with pre-warmed RPMI 1640 and either frozen in liquid nitrogen and stored at −80°C for immunoprecipitation/kinase assay , fixed on ice in acetone and stored at 4°C for immunohistochemistry , or placed in RPMI 1640 at 37°C ., Following equilibration at 37°C for 30 min in RPMI 1640 , live adult worms were treated for various durations ( 0 , 15 , 30 , 60 , or 120 min ) with either 20 µM GF109203X , 1 µM PMA , 1 µM U0126 , dimethyl sulphoxide ( DMSO; vehicle for PMA or U0126 , 0 . 2% or 0 . 1% , respectively ) , or were left untreated ( RPMI 1640 only ) ., In addition , live adult worms were also incubated in 1 µM PMA or DMSO for 24 h at 37°C/5% CO2 in RPMI containing glutamine , glucose , antibiotic/antimycotic mixture ( 100 U penicillin , 100 µg streptomycin and 0 . 25 µg amphotericin B/ml ) and 10% FBS ., GF109203X is a PKC inhibitor that competes for the ATP binding site and therefore only catalytically inactive proteins are inhibited ., The effect of GF109203X on PKC was thus established by pre-incubating worms in 20 µM GF109203X or RPMI 1640 for 120 min prior to exposure to 1 µM PMA for 30 min at 37°C ., The MEK inhibitor U0126 inhibits active and inactive MEK1/2 blocking ERK phosphorylation ., Immediately after treatment , medium was removed and worms were either homogenized in 30 µl RIPA buffer and processed for Western blotting or were acetone-fixed for immunohistochemistry in similar ways to that for cercariae ( above ) ., Perfused adult worms ( 100 or 20 pairs for PKC or ERK assay , respectively ) were transferred to microfuge tubes and washed twice with RPMI 1640 at 37°C ., Worms were then exposed to 1 µM PMA or DMSO ( vehicle ) in RPMI 1640 for 30 min at 37°C prior to being snap frozen in liquid nitrogen and stored at −80°C ., When required , worm pairs were defrosted and homogenized on ice in 100 µl cell lysis buffer with Halt protease/phosphatase inhibitor cocktail ., The homogenate was then centrifuged for 15 min at 4°C , the supernatant recovered , and the remaining pellet re-homogenized in 50 µl cell lysis buffer with inhibitors and centrifuged for a further 10 min ., Supernatants were pooled and equal quantities of protein from each sample incubated overnight at 4°C in either anti-phospho PKC ( pan ) ( ζ Thr410 ) , anti-phospho PKC ( pan ) ( βII Ser660 ) antibodies , or immobilized anti-phospho p44/p42 MAPK ( ERK1/2 ) ( Thr202/Tyr204 ) ( D13 . 14 . 4E ) XP primary antibodies ( each at 1/25 dilution ) ., The next day , 50 µl protein A agarose beads were added to homogenates and agitated for 5 min at 4°C except when the pre-immobilized anti-phospho p44/p42 MAPK ( ERK1/2 ) antibody was used; such rapid immunocapture reduces non-specific binding while permitting efficient immunocomplex adsorption 35 , 37 ., Following brief centrifugation , samples were washed twice each in 500 µl ice-cold lysis buffer and 500 µl ice-cold kinase buffer ( 25 mM Tris ( pH 7 . 5 ) , 5 mM β-glycerolphosphate , 2 mM DTT , 0 . 1 mM Na3VO4 and 10 mM MgCl2 ) ., Negative control immunoprecipitations lacking primary antibodies were also performed ., The Omnia ( Ser/Thr 8 peptide ) PKC assay was performed as per the manufacturers instructions to detect immunoprecipitated PKC activity ., Master mix , containing ATP , Ser/Thr 8 substrate , DTT , kinase buffer and water was added to individual wells of black 96-well microtitre plates ( Nunc ) and reactions started by adding immunocomplex ., Accumulation of phosphorylated substrate was measured ( excitation 355 nm , emission 460 nm ) every 30 s for up to 200 min at 30°C using a FLUOstar OPTIMA ( BMG Labtech ) microplate reader ., Positive control reactions contained 2 ng recombinant human PKC; negative controls either lacked the substrate or comprised samples prepared without the immunoprecipitation antibody ., Immunoprecipitated ERK activity was detected using the p44/p42 MAPK ( ERK1/2 ) assay kit ., Immunocomplexes were re-suspended in 20 µl kinase buffer supplemented with 200 µm ATP and 2 µg Elk-1 fusion protein and incubated for 30 min at 30°C ., Reactions were terminated with 10 µl 3× SDS-PAGE sample buffer and samples heated at 95°C for 5 min in preparation for SDS-PAGE and Western blotting with anti-phospho Elk-1 antibodies to reveal ERK1/2 activity ., Negative controls comprised samples without immunoprecipitation antibody ., Adult worms were placed in individual wells of a 24-well culture plate ( Nunc; 3–6 worm pairs per well ) in 1 . 5 ml RPMI 1640 containing glutamine , glucose , antibiotic/antimycotic mixture and 10% FBS at 37°C/5% CO2 22 for 1 h to equilibrate ., To determine effects of PKC inhibition/activation or ERK inhibition on adult worms , 1 µM , 5 µM , 20 µM or 50 µM GF109203X or U0126 , or 1 µM PMA were added to cultures ., Control groups containing RPMI 1640 only or 0 . 5% DMSO ( vehicle for PMA/U0126 ) were also included and all media and components were replenished daily ., Worms were observed at various times over 96 h using an Olympus SZ54045 binocular dissecting microscope and avi-format movies captured using a JVC TK-1481 composite colour video camera linked to Studio Launcher Plus for Windows software ., Worm behaviour including pairing status and ventral sucker attachment to the base of the culture plate was determined ., Egg release by worms was also enumerated ., Detailed analysis of worm movement was done using image J 38 ( http://rsbweb . nih . gov/ij/ ) ; the diameter of the well in pixels was calibrated to mm and the distance travelled by the posterior tip of each worm in 10 s was manually tracked and measured enabling translation into speed of movement ( velocity ) of pixels/s to mm/s ., Worm coiling ( Figure 10A ) was determined by counting the number of coils that persisted during 10 s visualization ., Results are representative of four independent experiments with a minimum of three replicates each; 30 or more parasites per treatment were scored , except for DMSO controls ( n\u200a=\u200a24 ) ., Racemic PZQ powder ( Shin Poong Pharmaceutical ) was dissolved in DMSO ., Freshly perfused adult worm pairs were incubated in RPMI 1640 containing 0 . 2 µg/ml PZQ , 0 . 1% DMSO ( vehicle ) , or RPMI 1640 alone at 37°C for 15 , 30 or 120 min ., Next , worms were either processed for Western blotting or immunohistochemistry as detailed above ., The final concentration of PZQ used was similar to a study 39 that reported 0 . 2 µg/ml PZQ to be the lowest concentration needed to induce maximal muscular contraction in worms , a phenotype observed with PZQ treatment ., Equal amounts of sample protein ( 12 µg ) were separated on 10% Precise SDS–PAGE gels , semi-dry transferred to nitrocellulose membranes and stained with Ponceau S to confirm homogeneous transfer ., After blocking for 1 h in 5% non-fat dried milk , membranes were washed three times in 0 . 1% Tween-Tris-buffered saline ( TTBS ) and incubated in anti-phospho PKC ( pan ) ( ζ Thr410 ) , anti-phospho PKC ( pan ) ( βII Ser660 ) or anti-phospho p44/p42 MAPK ( ERK1/2 ) ( Thr202/Tyr204 ) antibodies ( 1∶1000 in TTBS ) overnight at 4°C ., After TTBS wash , blots were incubated for 2 h at room temperature with HRP-conjugated secondary antibodies ( 1∶3000 in TTBS ) and exposed to West Pico chemiluminescence substrate ., Immunoreactive bands were visualized with a GeneGnome imaging system ( Syngene ) and relative band intensities quantified using GeneTools software ., Protein loading was determined by incubating blots with anti-actin antibodies ( 1∶1000 ) ., When necessary blots were stripped with Restore Western blot stripping buffer and incubated in an alternative antibody ., In addition , to confirm that the anti-phospho PKC and anti-phospho ERK antibodies detected only the phosphorylated kinases , blots were treated with lambda phosphatase ( 400 U/ml in TTBS containing 1% BSA and 2 mM MnCl2 ) for 4 h before incubation in primary antibodies; secondary antibody labeling and detection were then performed ., Acetone-fixed worms were washed twice with PBS , treated with 100 mM glycine for 15 min and blocked in 10% goat serum for 1 h ., After a further wash , worms were incubated in anti-phospho PKC ( pan ) ( ζ Thr410 ) , anti-phospho PKC ( pan ) ( βII Ser660 ) or anti-phospho p44/p42 MAPK ( ERK1/2 ) ( Thr202/Tyr204 ) antibodies ( 1∶50 dilution in PBS with 5% BSA ) for 3 days on a rotator ., Worms were then washed twice in PBS for 1 h each and incubated in Alexa Fluor 488 secondary antibodies ( 1∶500 in 5% BSA ) and 2 µg/ml rhodamine phalloidin for 24 h followed by two 1 h washes in PBS ., Next , specimens were mounted on slides in Vectashield ., All incubations were carried out at room temperature and washes were done in microfuge tubes ., Adult worms were visualized on a Leica TCS SP2 AOBS confocal laser scanning microscope using 20× dry , 40× or 62× oil immersion objectives and images captured ., The signal received from negative controls ( i . e . worms incubated with secondary antibody only ) was negated from that of the positive samples ., This was achieved by reducing the power level of the photomultiplier tube , which was then kept constant for all observations ., Statistical comparisons were performed with one-way analysis of variance ( ANOVA ) using Minitab ( version 16 ) ., All data were expressed as mean ±SEM , and statistical significance was determined by a Fishers multiple pair-wise comparison ., The phospho-specific antibodies used in this study were anti-phospho PKC ( pan ) ( ζ Thr410 ) , anti-phospho PKC ( pan ) ( βII Ser660 ) and anti-phospho p44/p42 MAPK ( ERK1/2 ) ( Thr202/Tyr204 ) ., Using the S . mansoni genome , four putative PKCs 23 and five putative ERKs displaying homology to human isotypes were identified from full-length sequences ., Comparative analysis revealed that , except for Smp_131700 ( ∼95 kDa PKC ) , the antibody recognition sequence was conserved in the putative PKCs and ERK1/2 , with the key phosphorylation motif retained; similar levels of sequence homology exist between the human isotypes recognized by these antibodies ( Figure 1 ) ., Such motif conservation is common , because of the critical nature of the phosphorylation site in activation of the respective kinases 1 , 2 , 18 , 40 ., As with human PKCs , in S . mansoni , the sequence surrounding the phosphorylated Thr residue within the PDK1 consensus motif was more conserved than that surrounding the Ser phosphorylation site in the bulky ring motif ( Figure 1 ) ., These antibodies were then used to detect phosphorylated ( activated ) PKCs and ERK1/2 in homogenates ( 12 µg protein ) of four different life stages ( miracidium , sporocyst , cercaria and adult worm ) of S . mansoni by Western blotting ., The anti-phospho PKC ( pan ) ( ζ Thr410 ) antibodies have been used previously to detect activated PKCs in invertebrates other than S . mansoni ( e . g . 41 ) ., They detect PKCα , βI , βII , γ , δ , ε , η , θ , and ζ/ι isotypes only when phosphorylated at a site homologous to Thr410 of human PKCζ ( Figure 1 ) that is critical to catalytic competency and PKC activation 42 ., In adult worms and cercariae , three immunoreactive bands were consistently detected , namely an ∼78/∼81 kDa doublet and a ∼132 kDa band with greater immunoreactivity in adult males than cercariae ( Figure 2A ) ., A weak band was also seen just above the ∼132 kDa band; however , this was not quantified in later experiments with worm pairs , as it was not always evident ., Occasionally , a weak immunoreactive band was also detected at ∼116 kDa in male worms and worm pairs ( Figure 2A ) ., In sporocysts and miracidia , a single ∼81 kDa band was detected which appeared highly phosphorylated in sporocysts compared to the other life stages studied ( Figure 2A ) ., Previously , we validated anti-phospho PKC ( pan ) ( βII Ser660 ) antibodies to detect phosphorylated ( activated ) PKC in S . mansoni miracidia and sporocysts 23; they have also been used on other invertebrates , such as the snail Lymnaea stagnalis 43 , 44 ., These antibodies recognize PKCα , βI , βII , δ , ε , η , and θ isotypes only when phosphorylated at a residue homologous to Ser660 of human PKCβII ( Figure 1 ) that is crucial to activation 45 ., Anti-phospho PKC ( pan ) ( βII Ser660 ) antibodies detected up to three immunoreactive bands in homogenates of adult S . mansoni ( Figure 2B ) ., Two faint bands were observed at ∼78 kDa and ∼132 kDa , similar to with anti-phospho PKC ( pan ) ( ζ Thr410 ) antibodies; in addition , a more immunoreactive band of ∼116 kDa was seen , occasionally also detected with anti-phospho PKC ( pan ) ( ζ Thr410 ) antibodies ., This ∼116 kDa protein was only detected in cercariae and adult worms , with greater immunoreactivity in worm pairs ( Figure 2B ) ., No immunoreactive proteins were detected in sporocyst homogenates with anti-phospho PKC ( pan ) ( Ser660 ) antibodies , and in miracidia only one protein was detected ( ∼78 kDa ) ( Figure 2B ) ., This finding is in accord with our previously published work that demonstrated that this PKC became inactive during the miracidium-to-mother-sporocyst transition 23 ., Finally , anti-phospho p44/42 MAPK ( ERK1/2 ) ( Thr202/Tyr204 ) antibodies have been used extensively to detect ERK phosphorylation ( activation ) in invertebrates including flies , snails and nematodes 44 , 46–49 , but not in S . mansoni ., These antibodies detect ERK1/2 when phosphorylated at residues homologous to Thr202/Tyr204 of human ERK1 ( Figure 1 ) with phosphorylation critical to activation ., Three proteins were detected using these antibodies across the four life stages of S . mansoni investigated ( Figure 2C ) ., Proteins of ∼48 kDa and ∼43 kDa were detected in adult worm homogenates , with greater immunoreactivity consistently observed in males ., In cercariae and sporocysts , a ∼43 kDa band and a weaker ∼38 kDa band were evident but the ∼48 kDa protein was not seen ., The ∼43 kDa band was more intense in sporocysts than in cercariae , female worms or adult worm pairs ( Figure 2C ) ., Strikingly , there were no visible immunoreactive ERKs in miracidia homogenates ( Figure 2C ) , when probing equal protein amounts from each life stage ., Treatment of blots containing protein extracts ( 20 µg ) of adult worm pairs with lambda phosphatase for 4 h prior to exposure to each of the anti-phospho antibodies resulted in either a total loss ( Figures 2B , 2C ) or substantial reduction ( Figure 2A ) in immunoreactivity , demonstrating that the antibodies react specifically with the phosphorylated forms of these proteins ., Next , we determined whether the phosphorylation ( activation ) state of the bands detected with the anti-phospho-PKC antibodies could be modulated using the PKC activator PMA or inhibitor GF109203X , reagents that we used previously to characterise PKC signalling in S . mansoni miracidia 23 and L . stagnalis haemocytes 49 ., Phosphorylation of all bands increased significantly ( p≤0 . 05 ) when live adult worms were exposed to 1 µM PMA for 30 min ( Figure 3A ) , demonstrating that the endogenous PKC-like proteins are not fully activated in adults under physiological conditions ., The ∼116 kDa and ∼132 kDa proteins showed the greatest increases in phosphorylation and are therefore likely cPKCs , which are characteristically highly responsive to DAG , Ca2+ and phorbol esters such as PMA 19 ., A biphasic effect of PMA on PKC phosphorylation was also evident with increased phosphorylation observed between 15 and 30 min , followed by a decrease to basal levels at 60 min increasing again at 120 min ( data not shown ) ., GF109203X inhibits PKC activity via competition with the ATP binding site and does not directly inhibit active forms of the enzyme ., Therefore , the inhibitory effect of GF109203X on S . mansoni PKCs was assessed by pre-incubating worms in GF109203X for 120 min , followed by exposure to 1 µM PMA for 30 min ., Under these conditions , the PMA-induced phosphorylation of PKCs was blocked ( Figure 3A ) ., Then , to determine whether the immunoreactive phosphoproteins possessed PKC activity an immunoprecipitation/kinase assay was conducted ., Adult worm proteins immunoprecipitated with either anti-phospho PKC antibody were able to phosphorylate the PKC substrate ( Ser/Thr 8 peptide ) , with higher activity achieved with anti-phospho PKC ( pan ) ( ζ Thr410 ) antibodies than with anti-phospho PKC ( pan ) ( βII Ser660 ) antibodies; activity also increased when worms were exposed to 1 µM PMA for 30 min ( Figure 3B ) ., The differences in kinase activity , seen as a result of the different antibodies having been used , is most likely due to the antibody affinity towards , and access to , the various epitopes in the native protein , together with the differential recognition of PKC isoforms displayed in Figure 3A ., Activated MEK1/2 is known to phosphorylate ERK1/2 in the TEY motif 30 ., Therefore , to evaluate the effect of MEK inhibition on ERK phosphorylation , live adult worms were exposed to 1 µM of the MEK inhibitor U0126 for 30 , 60 or 120 min ., Immunoblotting revealed that U0126 significantly attenuated ERK phosphorylation over time , compared to controls ( p≤0 . 05; Figure 4A ) ., After 30 min , phosphorylation of the ∼48 kDa and ∼43 kDa proteins was attenuated by ∼27% and 72% , respectively ( p≤0 . 01 and p≤0 . 001 ) , and after 120 min ERK phosphorylation was almost completely blocked ( Figure 4A ) ., Furthermore , adult worm proteins immunoprecipitated with anti-phospho p44/42 MAPK ( ERK1/2 ) ( Thr202/Tyr204 ) antibodies phosphorylated the ERK substrate Elk-1 ( Figure 4B ) , demonstrating that the detected ERK proteins display ERK activity ., Collectively these findings , when combined with knowledge of molecular weights ( below ) and antibody recognition sites , are entirely consistent with the detected proteins being S . mansoni PKCs and ERK1/2 , with MEK acting upstream of ERK as in other organisms ., Having determined that the immunoreactive proteins behave like PKCs and ERKs , it was possible to consider these proteins in relation to the S . mansoni genome data sets ., Although the precise assignment of Smp identifiers to immunoreactive bands was beyond the scope of this study , antibody recognition sites support a tentative assignment to the PKC class ( Figure S3 ) ., As we have previously reported 23 , the S . mansoni genome contains four annotated PKCs , two similar to human cPKCβI ( Smp_128480 , 75 . 6 kDa; Smp_176360 , 114 . 9 kDa ) , one to nPKCε ( Smp_131700 , 94 . 9 kDa ) and one to aPKCζ/ι ( Smp_096310 , 76 . 7 kDa ) , with PKCε also being similar to PKCη 4 ., Of these molecules , the ∼78 kDa and ∼116 kDa phosphorylated PKCs detected by anti-phospho PKC ( Ser 660 ) antibodies are most likely the β-type cPKCs ( Smp_128480 and Smp_176360 ) , because the crucial Ser residue recognized by these antibodies is conserved in these Smps and because the remaining S . mansoni nPKCε/aPKCι proteins do not possess this Ser autophosphorylation site ( Figure 1 ) ., Moreover , the ∼81 kDa band detected exclusively with anti-phospho PKC ( pan ) ( ζ Thr410 ) antibodies most likely corresponds to the atypical ι type PKC ( Smp_096310 ) ( Figure S3 ) as the Thr phosphorylation site ( Thr489 ) is conserved but , as with human PKCι , the Ser phosphorylation site is not ( Figure 1 ) ., The lack of strong activation of this PKC by PMA ( Figure 3A ) further supports this finding , given that aPKCs do not respond directly to DAG/PMA ., Available transcriptomic data for cercariae , schistosomules and adult worms 50 ( www . genedb . org ) also confirm that these PKCs are expressed in these life stages , with expression being developmentally regulated ( cf . Figure S3 ) ., The large ∼132 kDa PKC-like protein consistently detected in adult worms and cercariae using anti-phospho PKC ( pan ) ( ζ Thr410 ) antibodies was activated by PMA and inhibited by GF109203X , confirming its PKC-like nature ., Although absent from mammals , such high molecular weight PKCs are common in “lower” animals and , using similar antibodies , have been detected in the sea urchin Lytechnus pictus ( 135–140 kDa ) 51 and Caenorhabditis elegans ( 122 kDa ) 52 ., PKC activation profiles were notably different among the four life stages studied , with four , two , and one PKC detected consistently in cercariae and adult worms , miracidia and mother sporocysts , respectively ., This information suggests more complex roles for activated PKCs in the human host-infective/-dependent life stages ., Wiest et al . 21 found nine-fold greater total PKC activity in adult S . mansoni than in larval stages , and it is established that PKC expression is developmentally regulated in other invertebrates 53 ., We have shown that the ∼78 kDa β-type PKC of miracidia becomes inactive during development to the asexually reproducing , parasitic mother sporocyst stage and that PKC activity restricts this transformation 23 ., Here , using anti-phospho ( pan ) ( ζ Thr410 ) antibodies , we see that the ∼81 kDa PKC is substantially activated in 48 h mother sporocysts compared to miracidia , highlighting a possible role for this PKC in asexual reproduction and signifying the dynamic nature of PKC activation during schistosome development ., The ∼48 kDa and ∼43 kDa proteins detected with anti-phospho p44/42 MAPK antibodies likely correlate to genome sequences Smp_142050 ( 45 . 3 kDa , 70% identity to human ERK1 ) and Smp_047900 ( SmERK2 , 40 . 83 kDa , 68% identity to human ERK1 ) , due to molecular weight and antibody detection site similarity ( Figures 1 , 2C and S3 ) ; these proteins are also expressed in the requisite life-stages 50 ( www . genedb . org ) ( Figure S3 ) ., The recognition site of the p44/42 MAPK antibodies was conserved among three other putative S . mansoni ERK-like proteins ( Smp_133490 , 58 . 7 kDa; Smp_133500 , 82 . 7 kDa; Smp_134260 , 70 kDa ) , however , phosphorylated ERK-like proteins of this size were not detected on blots ., The ∼38 kDa protein detected in mother sporocysts and weakly in cercariae is possibly a non-specific band , because ∼38 kDa ERK-like proteins could not be detected in the S . mansoni genome ., The ERKs also appear differentially activated in the different S . mansoni life stages studied ( Figure 2C ) , with ERK1/2 ( ∼48/∼43 kDa ) detected only in adult worms ., No ERK activation was detected in miracidia , and only the ∼43 kDa ERK was phosphorylated ( using 12 µg protein ) in mother sporocysts and cercariae ., Thus , the ∼48 kDa ERK in male worms may play a specific role in growth and development and/or host-interactions , particularly given that human EGF activates EGFR in S . mansoni 33 and influences ERK signalling and proliferation in other parasites , including Echinococcus multilocularis 54 and Trypanosoma cruzi 55 ., The ERK pathway also interacts with the transforming growth factor β ( TGFβ ) pathway in schistosomes , possibly restricting interaction of SmSmad4 with receptor-activated Smad2 56 ., Therefore , because TGFβ signalling plays a part in mitotic activity , parasite development and egg embryogenesis 57 , 58 , ERK activity in particular cell types might suppress TGFβ signalling , with concomitant effects on development and reproduction ., Cross talk between PKC and ERK signalling , with PKC upstream of ERK is common in many organisms , including invertebrates such as C . elegans 59 ., To determine whether ERK and PKC signalling are connected in S . mansoni , ERK phosphorylation was determined after live adult worms were exposed to PKC modulators , and vice-versa ., PMA ( 1 µM ) increased phosphorylatio | Introduction, Materials and Methods, Results/Discussion | Protein kinases C ( PKCs ) and extracellular signal-regulated kinases ( ERKs ) are evolutionary conserved cell signalling enzymes that coordinate cell function ., Here we have employed biochemical approaches using ‘smart’ antibodies and functional screening to unravel the importance of these enzymes to Schistosoma mansoni physiology ., Various PKC and ERK isotypes were detected , and were differentially phosphorylated ( activated ) throughout the various S . mansoni life stages , suggesting isotype-specific roles and differences in signalling complexity during parasite development ., Functional kinase mapping in adult worms revealed that activated PKC and ERK were particularly associated with the adult male tegument , musculature and oesophagus and occasionally with the oesophageal gland; other structures possessing detectable activated PKC and/or ERK included the Mehlis gland , ootype , lumen of the vitellaria , seminal receptacle and excretory ducts ., Pharmacological modulation of PKC and ERK activity in adult worms using GF109203X , U0126 , or PMA , resulted in significant physiological disturbance commensurate with these proteins occupying a central position in signalling pathways associated with schistosome muscular activity , neuromuscular coordination , reproductive function , attachment and pairing ., Increased activation of ERK and PKC was also detected in worms following praziquantel treatment , with increased signalling associated with the tegument and excretory system and activated ERK localizing to previously unseen structures , including the cephalic ganglia ., These findings support roles for PKC and ERK in S . mansoni homeostasis , and identify these kinase groups as potential targets for chemotherapeutic treatments against human schistosomiasis , a neglected tropical disease of enormous public health significance . | Parasitic blood flukes , also called schistosomes , cause human schistosomiasis , a neglected tropical disease and major public health problem in developing countries , especially sub-Saharan Africa ., Sustainable control of schistosomiasis is difficult , mainly because the complex life cycle of the parasite involves a freshwater snail host , and the ability of the parasite to evade the immune response of the human host and to survive for many years ., Little is yet known about the cellular mechanisms in schistosomes and how they regulate parasite homeostasis , development and behaviour ., In this paper , the nature of intracellular signalling by protein kinases C ( PKCs ) and extracellular signal-regulated kinases ( ERKs ) in schistosomes is studied and these proteins are found to be vital for the coordination of processes fundamental to parasite survival , such as muscular activity and reproductive function ., Our results contribute to an understanding of molecular events regulating schistosome function and identify PKCs and ERKs as possible targets for the development of new chemotherapeutic treatments against schistosomiasis . | erk signaling cascade, protein kinase signaling cascade, protein kinase c signaling, signal transduction, infectious diseases, medicine and health sciences, helminth infections, cell biology, biology and life sciences, ras signaling, parasitic diseases, molecular cell biology, zoology, cell signaling, helminthology, signaling cascades | null |
journal.pcbi.1004137 | 2,015 | The Neuroglial Potassium Cycle during Neurotransmission: Role of Kir4.1 Channels | Astrocytic processes enwrap more than half of CA1 hippocampal synapses to form tripartite synapses 1 , 2 ., Perisynaptic astroglial processes are enriched in ionic channels , neurotransmitter receptors and transporters , enabling astrocytes to detect neuronal activity via calcium signaling 3 and ionic currents with various components , such as glutamate and GABA transporter 4–7 or potassium ( K+ ) 8–10 ., Thus astrocytes regulate neuronal activity through multiple mechanisms , involving signaling or homeostasis of extracellular space volume , glutamate , GABA or K+ levels 11 ., Interestingly , membrane depolarization was the first activity-dependent signal identified in glial cells and was attributed to K+ entry across their membrane 10 ., Such K+ entry was suggested to contribute to K+ spatial buffering , consisting in glial uptake of excess extracellular K+ ( K+o ) , redistribution via gap-junction astroglial networks and subsequent release at sites of low K+o 12 ., Modeling studies have mostly investigated astroglial regulation of K+o during pathological conditions to clarify its impact on aberrant neuronal activity ., In particular astrocytes , by regulating K+o , have been shown to contribute to initiation and maintenance of epileptic seizures 13–15 , as well as to the severity of ischemia following stroke , with a neuroprotective or neurotoxic role , depending on K+o 16 , 17 ., In addition , experimental data suggest that several K+ channels or transporters contribute to astroglial K+ clearance , such as inward rectifier 4 . 1 and two pore K+ channels ( Kir4 . 1 and K2P , respectively ) or Na/K ATPases 18 , 19 ., Remarkably , recent work suggest that Kir4 . 1 channels play a prominent role in astroglial regulation of K+o 20–23 ., However , the mouse model used to draw these conclusions , i . e . conditional Kir4 . 1 knockout mice directed to glial cells ( GFAP-Cre-Kir4 . 1fl/fl mice , Kir4 . 1-/- ) , exhibits several limitations:, 1 ) Kir4 . 1 channels are not specifically deleted in astrocytes , but also in other glial cells such as oligodendrocytes or retinal Müller cells 22;, 2 ) astrocytes are severely depolarized 21 , 22;, 3 ) Kir4 . 1-/- mice die prematurely ( ~3 weeks ) and display ataxia , seizures , hindleg paralysis , visual placing deficiency , white matter vacuolization and growth retardation 22 , highlighting that chronic deletion of Kir4 . 1 channels induces multiple brain alterations and possibly compensations ., Thus , the specific and acute contribution of astroglial Kir4 . 1 channels to K+o and to the moment to moment neurotransmission is still unclear ., To decipher the acute role of astrocytes in controlling K+ homeostasis and neuronal activity , we built a tri-compartment model accounting for K+ dynamics between neurons , astrocytes and the extracellular space ., We quantified K+ neuroglial interactions during basal and high activity , and found that Kir4 . 1 channels play a crucial role in K+ clearance and astroglial and neuronal membrane potential dynamics , especially during repetitive stimulations , and prominently regulate neuronal excitability for 3 to 10 Hz rhythmic activity ., To model K+ ions dynamics during neuronal activity , we built a biophysical model that includes three compartments: the neuron , the astrocyte and the extracellular space ( Fig . 1A ) ., As performed in several studies 13 , 16 , 24 , the neuron is approximated by a single compartment conductance-based neuron containing Na+ and K+ voltage-gated channels , enabling action potential discharge ., The associated neuronal membrane potential is coupled with the dynamics of intracellular and extracellular Na+ and K+ levels via the dependence of the neuronal currents to the Nernst equation ., The ion concentrations depend also on the activity of neuronal and astroglial Na/K ATPases , which maintain resting K+i by balancing K+ and Na+ fluxes ., Similarly , the astrocyte is approximated by a single compartment conductance-based astrocyte containing Kir4 . 1 channels , which are inward rectifier K+ channels strongly expressed in astrocytes that generate dynamic K+ currents 25 ., In the model , neurons and astrocytes are separated by a homogenous extracellular space compartment ., The model is based on balancing ionic fluxes between the three compartments ( Fig . 1B ) ., The model starts with the induction of a synaptic current ( Iapp , see Materials and Methods ) ., This current is the initial input of a classical Hodgkin-Huxley model , which describes the neuronal membrane potential dynamics ( entry of Na+ and exit of K+ ) ., Released extracellular K+ is taken up by astrocytes through Kir4 . 1 channels and Na/K ATPases ( Fig . 1B and Materials and Methods ) ., Because Kir4 . 1 channels are strongly involved in K+ uptake 22 , we fitted the I-V curve of K+ ions through Kir4 . 1 channels using equation 22 ( see materials and methods ) and predicted the I-V curve at various values of K+o ( Fig . 1C ) ., We obtain that K+ fluxes through Kir4 . 1 channels vanish around astrocytic resting membrane potential ( ~-80 mV ) and are outward during astrocytic depolarization for a fixed K+o ( 2 . 5 mM , Fig . 1C ) ., However , they become inward when K+o increases ( 5–10 mM , Fig . 1C ) ., Using this model , we shall investigate quantitatively the contribution of Kir4 . 1 channels to K+ uptake in relation to neuronal activity associated with different K+o ., To validate our tri-compartment model , we compared simulation results with electrophysiological recordings ., To account for the synaptic properties of CA1 pyramidal neurons , we generated a synaptic current ( Iapp ) using the depression-facilitation model ( equation 1 ) ( see Materials and Methods with input f, ( t ) = δ, ( t ) ) ( Fig . 2A , E , I ) ., We first investigated responses to single stimulation ., Using the Hodgkin-Huxley model , this synaptic current induces a firing activity ( S1A Fig . ) , resulting in a ~ 0 . 9 mM increase of K+o within 300 milliseconds , which slowly decayed back to baseline levels during 10 seconds ( S1B Fig . ) ., The extracellular K+ dynamics was associated in our model with a small astrocytic depolarization of ∆V = −1 . 35 mV ( equations 22 , 23 , 25 ) ( Fig . 2C ) ., Using electrophysiological recordings of evoked field excitatory postsynaptic potential ( fEPSP ) by a single stimulation of Schaffer collaterals in acute hippocampal slices ( Fig . 2B ) , we measured astroglial membrane potential depolarization and found that it reached ~ 1 . 3 mV ( 1 . 3 ± 0 . 2 mV , n = 6 ) ( Fig . 2C ) , confirming the result of our simulation ., After validating the responses of the tri-compartment model to basal stimulation , we investigated the impact of trains of stimulations on the dynamics of astroglial membrane potential ., During tetanic stimulation ( 100 Hz for 1 second ) , variations in neuronal membrane potential described by the Hodgkin-Huxley equation show a bursting activity during ~ 1 second ( S1C Fig . ) ., This is associated with a depolarization of astrocytic membrane potential of ~ 5 mV , which lasts ~ 6 seconds ( Fig . 2G , H ) and an increase in K+o that reaches a peak value of 4 . 4 mM ( S1D Fig . ) ., For repetitive stimulations ( 10 Hz for 30 seconds ) , the neuron exhibited firing activity during the whole stimulation ( S1E Fig . ) ., This was associated with an astroglial depolarization of ~ 12 mV ( Fig . 2K ) and an increase in K+o peaking at 6 . 9 mM after 17 . 5 seconds of stimulation ( S1F Fig . ) ., Although the stimulation lasted 30 seconds , the astrocytic depolarization started to decay after 17 seconds ( Fig . 2K ) ., The kinetics of astroglial membrane potential dynamics obtained with the numerical simulations are comparable to the results obtained with electrophysiological recordings performed in individual astrocytes during single stimulation ( rise time: 48 . 4 ms for numerical stimulation , 42 ± 19 ms n = 6 for experiments; time of peak: 740 ms for numerical simulation , 730 ± 60 ms n = 6 for experiments; decay time: 3 . 67 s for numerical simulation , 4 . 50 s ± 0 . 2 n = 6 for experiments , Fig . 2D ) , tetanic stimulation ( rise time: 610 ms for numerical simulation , 491 ± 122 ms n = 5 for experiments; time of peak: 1 . 07 s for numerical simulation , 1 . 05 s ± 0 . 25 n = 5 for experiments; decay time: 4 . 18 s for numerical simulation , 4 . 55 s ± 0 . 45 n = 5 for experiments , Fig . 2H ) and repetitive stimulation ( rise time: 1 . 5 s for numerical simulation , 1 . 27 s ± 0 . 18 n = 5 for experiments; time of peak: 6 . 8 s for numerical simulation , 5 . 2 s ± 0 . 9 n = 5 for experiments; decay time: 7 . 95 s for numerical simulation , 8 . 3 s ± 0 . 4 n = 5 for experiments , Fig . 2L ) ., These data show that the dynamics of astroglial membrane potential changes obtained from numerical simulations and from electrophysiological recordings are similar ., Thus our model captures the key players sufficient to mimic the evoked astroglial membrane potential dynamics observed experimentally in different regimes of activity ., We investigated the dynamics of the K+ cycle between neurons , extracellular space and astrocytes induced by neuronal activity to decipher the time needed to restore basal extracellular and intra-neuronal K+ levels ., We studied K+ redistribution induced by single , tetanic ( 100 Hz , 1, s ) and repetitive ( 10 Hz , 30, s ) stimulations , and found that the general behavior of K+ dynamics was divided into three phases ( phases 0 , 1 and 2; Fig . 3 ) ., During phase 0 ( t = 0 to t1 ) , neuronal K+ is released in the extracellular space ( peak K+o during phase 0: 0 . 9 mM at 300 ms for single stimulation; 1 . 9 mM at 1 . 3 s for tetanic stimulation; 4 . 4 mM at 30 s for repetitive stimulation , Fig . 3A , D , G ) ., Compared to basal K+ levels in each compartment ( at t = 0 ) , the relative transient evoked increase in K+ concentration is prominent only in the extracellular space ( ~+37% for single stimulation , +76% for tetanic stimulation and +168% for repetitive stimulation , Fig . 3B , E , H ) ., During phases 0 and 1 , released K+ is then mostly buffered by astrocytes ( ~80 to 99% at the end of phase 1 ) during the different regimes of activity ( time t2 ( at the end of phase 1 ) for single stimulation: 8 . 2 s; tetanic stimulation: 8 . 7 s; repetitive stimulation: 34 . 2 s , Fig . 3C , F , I ) ., The astroglial net K+ uptake increases with the activity-dependent K+o transient rises ( S2A–C Fig . ) evoked by the different regimes ( S2D Fig . ) ., Neurons slowly re-uptake only ~5–10% of their released K+ at the end of phase 1 ( Fig . 3C , F , I ) ., Remarkably , although K+o increases with the strength of stimulation ( from 0 . 9 to 4 . 4 mM , Fig . 3A , D , G and S2A–C Fig . ) , the time needed for astrocytes to buffer the released K+ is not proportional to K+o rises ( Fig . 3C , F , I ) , as shown by the phase diagram illustrating astroglial K+ uptake as a dynamic function of activity-dependent changes in K+o evoked by the different stimulations ( S2D Fig . ) , but is to the square root of K+o ( equation 22 ) ., In addition , at the end of phase 1 , K+o is almost back to baseline levels , whereas intra-astroglial K+ levels reach their peak value ( Fig . 3C , F , I ) ., Finally during phase 2 ( t2 to end ) , astroglial buffered K+ is slowly redistributed back to neurons , which ends the K+ cycle ., The long-lasting phase 2 is marked by an inversion of K+ fluxes in astrocytes , suggesting moderate K+ release by astrocytes over time ., Indeed , K+ redistribution to neurons depends on K+ release through Kir4 . 1 channels , which is limited by the low outward rectification of these channels ( Fig . 1C ) ., Altogether , these data suggest a slow , but dynamic and efficient astroglial clearance capacity for the different regimes of activity ., To study quantitatively the acute and selective role of astroglial Kir4 . 1 channels in neuroglial K+ dynamics , we inhibited the Kir4 . 1 current in our tri-compartment model ., Because Kir4 . 1-/- mice display altered synaptic plasticity compared to wild type mice 22 , 26 , we recalibrated the synaptic current ( Iapp ) parameters τrec and τinact in equations 1 , 2 ( see Table 1 ) for the facilitation-depression model to get an optimal fit to the recorded postsynaptic responses 26 ., Another change in the model consisted in setting at zero both the Kir4 . 1 current and the leak term ., In addition , to compensate for the loss of K+ fluxes through astroglial Kir4 . 1 channels , we added in equation 27 a constant K+ flux to maintain K+o at an equilibrium value of 2 . 5 mM ., Consequently , the astrocytic membrane potential displayed no change during stimulation , in agreement with electrophysiological recordings 21 , 22 ., The numerical simulations show that inhibition of astroglial Kir4 . 1 channels leads to higher transient peak increase in K+o during repetitive and tetanic stimulation compared to control conditions ( Fig . 4E , F , I , J ) , while no difference is observed for single stimulation ( Fig . 4A , B ) ., In addition , for all regimes of activity , the rise and decay times of the K+o were increased when Kir4 . 1 channels were inhibited ( single stimulation , control: rise time 136 ms , decay time 3 . 4 s; Kir4 . 1 inhibition: rise time 232 ms , decay time 4 . 2 s; tetanic stimulation , control: rise time 638 ms , decay time 4 s; Kir4 . 1 inhibition: rise time 753 ms , decay time 6 s; repetitive stimulation , control: rise time 6 . 8 s; Kir4 . 1 inhibition: rise time 20 . 2 s , Fig . 4B , F , J ) ., Finally , Kir4 . 1 channel inhibition only slightly increased neuronal firing induced by single stimulation ( Fig . 4C , D ) and tetanic stimulation ( Fig . 4G , H ) , while it had major effect on neuronal excitability during repetitive stimulation ( Fig . 4K , L ) ., Indeed , although firing frequency was only slightly increased during the first 8 seconds of repetitive stimulation when K+o reached 10 mM ( Fig . 4I ) , action potential amplitude and firing rate then progressively decreased due to neuronal depolarization ( from-33 mV to-19 mV after 14 and 30 seconds of stimulation , respectively ) , suppressing neuronal firing after 14 seconds of stimulation ( Fig . 4K ) ., Altogether , these data show that astroglial Kir4 . 1 channels are prominently involved in K+ buffering during high level of activity , and thereby have a major impact on neuronal resting membrane potential controlling firing during trains of stimulations ., To investigate the effect of astroglial Kir4 . 1 channels on endogenous physiological rhythmic activity , we generated probabilistic firing induced by sub-firing stimulation coupled to neuronal Brownian noise ( Fig . 5A , B ) ., To simulate the firing activity , we generated a sub-firing periodic stimulation ( 5 ms squared stimulus ) , which defines the applied synaptic intensity in our tripartite compartment model ( Fig . 5A ) , and added a Brownian noise in the neuronal membrane potential ( equation 21 , Fig . 5B ) ., Such stimulation induces an increase in K+o ( Fig . 5C ) , and thus firing over time ( Fig . 5D-E ) ., We found that astroglial Kir4 . 1 channels had no effect on the firing probability ( computed over 100 simulations ) for basal ( 0 . 1 Hz , Fig . 5F ) , low ( 1 Hz , Fig . 5G ) and high ( 50 Hz , Fig . 5K ) frequency stimulations ., However , Kir4 . 1 channels directly regulate the firing probability for 3 and 5 Hz stimulations after 7 and 12 s of sub-firing stimulation , respectively ( Fig . 5H , I ) ., In contrast , Kir4 . 1 channels regulate only transiently the firing probability induced by 10 Hz stimulation ( Fig . 5J ) ., These data suggest a prominent and specific involvement of astroglial Kir4 . 1 channels in regulation of firing during theta rhythmic activity ., Several models have investigated extracellular K+ regulation of neuronal activity , including glial uptake mechanisms 13–17 , 24 , 33 , 34 ., To study seizure discharges and spreading depression , a first tri-compartment model including the neurons , astrocytes and extracellular space was proposed 24 , although the astrocytic membrane potential was not taken into account , and K+ accumulation in the interstitial volume was controlled by a first-order buffering scheme that simulated an effective glial K+ uptake system ., With such model , after evoked firing , it took ~17 s for the neuronal membrane potential to return to resting values , via activation of Na/K ATPases ., The model also predicted that elevated K+o have a key role in the initiation and maintenance of epileptiform activity ., In our study , we accounted for the astroglial modulation of K+ buffering capacity regulated by its membrane potential , and found that the biophysical properties of astrocytic membranes including Kir4 . 1 channels were sufficient to account for the long-lasting clearance of extracellular K+ ., Interestingly , we confirm that alteration in K+ clearance leading to an extracellular K+ accumulation induces epileptiform activity , and show specifically that Kir4 . 1 channel acute inhibition leads to such pathological bursting activity during repetitive stimulation ., A similar tri-compartment model has been simplified as a one-dimensional two-layer network model to study how neuronal networks can switch to a persistent state of activity , as well as the stability of the persistent state to perturbations 13 ., In this model , Na+ and K+ affect neuronal excitability , seizure frequency , and stability of activity persistent states ., In particular , the quantitative contribution of intrinsic neuronal currents , Na/K ATPases , glia , and extracellular Na+ and K+ diffusion to slow and large-amplitude oscillations in extracellular and neuronal Na+ and K+ levels was revealed ., In the model , the estimated K+o during epileptiform activity are comparable to the ones observed experimentally 35 , 36 ., Although this model does not account for astroglial Kir4 . 1 channels , it shows that a local persistent network activity not only needs balanced excitation and inhibition , but also glial regulation of K+o 15 ., Finally , a model accounting for the extracellular space and astroglial compartments has quantified the involvement of several astroglial ionic channels and transporters ( Na/K ATPase , NKCC1 , NBC , Na+ , K+ , and aquaporin channels ) in the regulation of firing activity 34 ., To account for K+ dynamics between neurons , astrocytes and the extracellular space , we built for the first time a tri-compartment model , where we included neuronal voltage-gated channels , Na/K pumps and astrocytic Kir4 . 1 channels according to their biophysical properties , as well as membrane potential of astrocytes ., Because functional expression of voltage-gated calcium channels on hippocampal mature astrocytes in situ in physiological conditions and its impact on astrocytic functions is still a matter of debate 37 , such channels were not included in our model ., However , many other astroglial K+ channels ( such as two pore domain K+ channels ( K2P ) ( TWIK-1 , TREK-1 , TREK-2 and TASK-1 ) , inward rectifier K+ channels ( Kir2 . 1 , 2 . 2 , 2 . 3 , 3 . 1 , 6 . 1 , 6 . 2 ) , delayed rectifier K+ channels ( Kv1 . 1 , 1 . 2 , 1 . 5 , 1 . 6 ) , rapidly inactivating A-type K+ channels ( Kv1 . 4 ) , calcium-dependent K+ channels ( KCa3 . 1 ) ) , but also other channels , transporters or exchangers ( such as Cx hemichannels , Na+/K+/Cl- co-transporter ( NKCC1 ) K+/Cl- exchanger , glutamate transporters ) 16 , 38 , 39 could also play a role in the regulation of activity-dependent changes in K+i or K+o ., Functional evidence of the contribution of these channels , transporters or exchangers in astroglial K+ clearance is actually scarce , although K2P channels have been suggested to participate in astroglial K+ buffering 40 , while NKCC1 were recently shown in hippocampal slices not to be involved in activity-dependent K+ clearance 41 ., Similarly , adding slower timescale K+ dependent conductances in the neuron model could modulate the slow redistribution of K+ to neurons , and thus the duration of the neuroglial potassium cycle , and is of interest to implement in future development of the model ., In our study , the aim was to simplify the system to capture in the model the minimal set of astroglial channels and pumps accounting for our experimental data related to activity-dependent changes in astroglial membrane potential ., In addition our tri-compartment model , as most existing models , did not account for the complex multiscale geometry of astrocytes and neurons ., Incorporating in our current model additional astroglial and neuronal channels , as well as complex cell geometry is of particular interest to identify modulatory effects of other specific channels and of microdomain geometry on the neuroglial potassium cycle ., In accordance with previous studies , where Kir4 . 1 channels were chronically deleted genetically in glial cells 20 , 21 , 23 , we found that acute inhibition of Kir4 . 1 channels leads to altered regulation of extracellular K+ excess and affects the kinetics of K+o ( Fig . 4I , J ) ., However , in contrast to these studies , we found that Kir4 . 1 channel inhibition also alters significantly K+o peak amplitudes during repetitive stimulation , suggesting that Kir4 . 1-/- mice may display some compensatory mechanisms attempting to maintain extracellular K+ homeostasis ., In addition , our model reveals that specific and acute inhibition of Kir4 . 1 channels slows down , but does not abolish , astroglial uptake of excess K+ during single , tetanic and repetitive stimulations , confirming that astroglial Na/K ATPases , included in our model , also contribute to K+ clearance 41 ., Contrary to action potentials , characterized by a very fast dynamics in the order of a few milliseconds , astroglial K+ buffering lasts tens of seconds ., As shown in the present study , most of extracellular K+ released by neurons is first cleared by astrocytes through Kir4 . 1 channels ., To determine the factors controlling the slow timescale of astroglial K+ clearance , we focused on Kir4 . 1 channels ., Because the astroglial leak conductance ( equation 23 ) is six times smaller than the Kir4 . 1 channel conductance , we neglected it ., The dynamics of astrocytic membrane potential VA is described by equation 23 , where the membrane capacitance is CA ≈ 15 pF and the maximal Kir4 . 1 channel conductivity isGKir≈60pS ., In that case , using equation 23 , the time constant of Kir4 . 1 channel-mediated return to equilibrium of astroglial membrane potential τA is defined as We obtain the following approximation τA ≈ 0 . 6s using equation 23 and the parameters of table 1 ., This time constant is consistent with the fitted exponential decay time obtained in our simulations and experiments for a single stimulation where we obtained τ ≈ 0 . 7s ., However , simulations for stronger stimulations indicate an increase of τ to approximatively 4 seconds ( tetanic stimulation ) and 9 seconds ( repetitive stimulation ) ., This increase in clearance duration is due to the dependence of the Kir4 . 1 current to K+o , as illustrated by the IV relation ( Fig . 1C ) and described in equation 22 ., The Nernst potential VKA increases for strong stimulations ( tetanic and repetitive ) , which slow down the kinetics of astrocytic membrane potential VA through the term 1+exp ( VA−VKA−V2AV3A ) in equation 22 ., We conclude that the slow time scale of K+ clearance is in part due to the availability of Kir4 . 1 channels at low and high K+o ., This clearance timescale is much longer than the glutamate clearance rate of τglu ≈ 15 ms that we previously reported 42 ., Moreover , the redistribution of K+ released by neurons during the different regimes of activity shows that the higher the activity , the lower the proportion of released K+ remains transiently in the extracellular space ., This suggests that Kir4 . 1 channels have a strong uptake capacity , especially for high regimes of activity ( K+o up to 5–6 mM ) ., Remarkably , our model reveals that astroglial Kir4 . 1 channels strongly regulate neuronal firing induced by high stimulation regime such as repetitive stimulation ., Kir4 . 1channels are crucially involved in regulation of K+o during this regime of activity , most likely because such stimulation triggered long-lasting neuronal release of K+ ( 20 mM over 30 seconds , Fig . 3G ) resulting in a sustained , but moderate increase in K+o ( >6 mM for ~22 s , Fig . 4I , J ) , compared to the neuronal release ., These data suggest that during repetitive stimulation , astrocytes can buffer up to ~14 mM of K+o and thereby preserve neuronal firing ., However , astroglial Kir4 . 1 channels slightly impact neuronal firing induced by single and tetanic stimulations , probably because these stimulations only triggered transient neuronal K+ release ( 0 . 9 mM over 300 ms ( Fig . 3A ) and 1 . 9 mM over 1 . 3 s ( Fig . 3D ) , respectively ) , resulting in a short and small increase in K+o ( >2 . 7 mM for ~450 ms for single stimulation ( Fig . 4A , B ) , and >3 . 5 mM for 1 . 5 s for tetanic stimulation ( Fig . 4E , F ) ) ., Nevertheless , we show a prominent and specific involvement of astroglial Kir4 . 1 channels in probabilistic firing activity induced by 3 to 10 Hz sub-firing stimulations ( Fig . 5 ) , suggesting a key role of these channels in sustained theta rhythmic activity ., Interestingly , these data imply that Kir4 . 1 channels can contribute to fine tuning of neuronal spiking involving low , but long-lasting , increase in K+o ., Thus besides gliotransmission , regulation of K+o by Kir4 . 1 channel provides astrocytes with an alternative active and efficient mechanism to regulate neuronal activity ., Several studies have reported decreased Kir4 . 1 protein levels and Kir functional currents in sclerotic hippocampus from epileptic patients 43–46 ., Whether these changes are the cause or the consequence of epilepsy is still an open question ., However , Kir4 . 1-/- mice display an epileptic phenotype 22 , 47 and missense mutations in KCNJ10 , the gene encoding Kir4 . 1 , have been associated with epilepsy in humans 48 , 49 ., These data thus suggest that impairment in Kir4 . 1 function leading to alterations in K+o dynamics , as shown in our study , may cause epilepsy ., Remarkably , dysfunction of K+o regulation by Kir4 . 1 channels is likely involved in other pathologies , since it contributed to neuronal dysfunction in a mouse model of Huntington’s disease 50 and the presence of antibodies against Kir4 . 1 channels in glial cells was recently found in almost 50% of multiple sclerosis patients 51 ., Thus astroglial Kir4 . 1 channels may well represent an alternative therapeutic target for several diseases ., Experiments were carried out according to the guidelines of European Community Council Directives of January 1st 2013 ( 2010/63/EU ) and our local animal committee ( Center for Interdisciplinary Research in Biology in Collège de France ) ., All efforts were made to minimize the number of used animals and their suffering ., Experiments were performed on the hippocampus of wild type mice ( C57BL6 ) ., For all analyses , mice of both genders and littermates were used ( PN19–PN25 ) ., Acute transverse hippocampal slices ( 400 μm ) were prepared as previously described 42 , 52–54 from 19–25 days-old wild type mice ., Slices were kept at room temperature ( 21–23°C ) in a chamber filled with an artificial cerebrospinal fluid ( ACSF ) composed of ( in mM ) : 119 NaCl , 2 . 5 KCl , 2 . 5 CaCl2 , 1 . 3 MgSO4 , 1 NaH2PO4 , 26 . 2 NaHCO3 and 11 glucose , saturated with 95% O2 and 5% CO2 , prior to recording ., Acute slices were placed in a recording chamber mounted on a microscope including infra-red differential interference ( IR-DIC ) equipment , and were bathed in ACSF perfused at 1 . 5 ml/min ., ACSF contained picrotoxin ( 100 μM ) , and connections between CA1 and CA3 regions were cut to avoid epileptic-like activity propagation ., Extracellular field and whole-cell patch-clamp recordings were obtained using glass pipettes made of borosilicate ., Astroglial and postsynaptic responses were evoked by Schaffer collateral stimulation ( 0 . 05Hz ) in the CA1 stratum radiatum region with glass pipettes filled with ACSF ( 300–700 kΩ ) ., Astrocytes from stratum radiatum were recognized by their small soma size ( 5–10 μm ) , very low membrane resistance and hyperpolarized resting membrane potentials ( ≈- 80 mV ) , passive properties of their membrane ( linear I-V ) , absence of action potentials , and large coupling through gap junctions ., Field excitatory postsynaptic potentials ( fEPSPs ) were obtained in 400 μm slices using pipettes ( 4–6 MΩ ) located in the stratum radiatum region ., Stimulus artifacts were suppressed in representative traces ., Whole-cell recordings were obtained from CA1 astrocytes , using 4–6 MΩ glass pipettes containing ( in mM ) : 105 K-Gluconate , 30 KCl , 10 HEPES , 10 Phosphocreatine , 4 ATP-Mg , 0 . 3 GTP-Tris , 0 . 3 EGTA ( pH 7 . 4 , 280 mOsm ) ., Prolonged repetitive stimulation was performed for 30 s at 10 Hz ., Post-tetanic potentiation was evoked by stimulation at 100 Hz for 1 s in the presence of 10 μM CPP ( ( Rs ) -3- ( 2-Carboxypiperazin-4-yl- ) propyl-1-phosphonic acid ) ., Recordings were performed with Axopatch-1D amplifiers ( Molecular Devices , USA ) , at 10 kHz , filtered at 2 kHz , and analyzed using Clampex ( Molecular Devices , USA ) , and Matlab ( MathWorks , USA ) softwares ., The data represent mean ± SEM ., Picrotoxin was from Sigma and CPP from Tocris ., We present here the biophysical model we have built to describe K+ dynamics during neuronal activity and specifically the role of astroglial Kir4 . 1 channels ., After Schaffer collateral stimulation , excitatory synapses release glutamate molecules that activate postsynaptic neurons ., We modeled this step by classical facilitation/depression model 55 ., The resulting postsynaptic activity triggers ionic release in the extracellular space and a change in the astrocytic membrane potential through ion uptake ., We used the average neuronal potential and mass conservation equations for ionic concentrations to model changes in astrocytes ., We have built a tri-compartment model , which accounts for: 1 ) the neuron , 2 ) the astrocyte and 3 ) the extracellular space ., We included voltage gated channels , Na/K pumps and astrocytic Kir4 . 1 channels ., To account for the stimulation of Schaffer collaterals that induce a postsynaptic response in the CA1 stratum radiatum region , we used a facilitation-depression model 55–57 ., where f is the input function ., For a single stimulation generated at time tstim , f ( t ) = δ ( t-tstim ) ., A stimulation instantaneously activates a fraction Use of synaptic resources r , which then inactivates with a time constant τinac and recovers with a time constant τrec In the simulations , at time t = tstim , r and e respectively decreases and increases by the value User ., The synaptic current Iapp is proportional to the fraction of synaptic resources in the effective state e and is given by Iapp = Asee ( the parameter Ase is defined in table 1 ) ., We used the following definitions for the input function f: The dynamics of the neuronal membrane potential , VN , follows the classic Hodgkin Huxley ( HH ) equations 58 ., with rate equations, αn ( VN ) =0 . 01 ( VN+10 ) exp ( 0 . 1 ( VN+10 ) ) −1, ( 12 ), βn ( VN ) =0 . 125exp ( VN/80 ), ( 13 ), αm ( VN ) =0 . 1 ( VN+25 ) exp ( 0 . 1 ( VN+25 ) ) −1, ( 14 ), βm ( VN ) =4exp ( VN/18 ), ( 15 ), αh ( VN ) =0 . 07exp ( VN/20 ), ( 16 ), βh ( VN ) =1exp ( 0 . 1 ( VN+ 30 ) ) +1, ( 17 ), Vrest is the resting membrane potential and VKN and VNaN are respectively the K+ and Na+ equilibrium potentials and are given by the Nernst equations, VNaN=RTFln ( Na0NaN ), ( 18 ), VKN=RTFln ( K0KN ), ( 19 ), where Na0 and NaN are respectively the extracellular and neuronal sodium concentrations , and K0 and KN are respectively the extracellular and neuro | Introduction, Results, Discussion, Materials and Methods | Neuronal excitability relies on inward sodium and outward potassium fluxes during action potentials ., To prevent neuronal hyperexcitability , potassium ions have to be taken up quickly ., However , the dynamics of the activity-dependent potassium fluxes and the molecular pathways underlying extracellular potassium homeostasis remain elusive ., To decipher the specific and acute contribution of astroglial Kir4 . 1 channels in controlling potassium homeostasis and the moment to moment neurotransmission , we built a tri-compartment model accounting for potassium dynamics between neurons , astrocytes and the extracellular space ., We here demonstrate that astroglial Kir4 . 1 channels are sufficient to account for the slow membrane depolarization of hippocampal astrocytes and crucially contribute to extracellular potassium clearance during basal and high activity ., By quantifying the dynamics of potassium levels in neuron-glia-extracellular space compartments , we show that astrocytes buffer within 6 to 9 seconds more than 80% of the potassium released by neurons in response to basal , repetitive and tetanic stimulations ., Astroglial Kir4 . 1 channels directly lead to recovery of basal extracellular potassium levels and neuronal excitability , especially during repetitive stimulation , thereby preventing the generation of epileptiform activity ., Remarkably , we also show that Kir4 . 1 channels strongly regulate neuronal excitability for slow 3 to 10 Hz rhythmic activity resulting from probabilistic firing activity induced by sub-firing stimulation coupled to Brownian noise ., Altogether , these data suggest that astroglial Kir4 . 1 channels are crucially involved in extracellular potassium homeostasis regulating theta rhythmic activity . | Neural excitability relies on precise inward and outward ionic fluxes ., In particular , potassium ions , released by neurons during activity , have to be taken up efficiently to prevent hyperexcitability ., Astrocytes , the third element of the synapse , play a prominent role in extracellular potassium homeostasis ., Thus unraveling the dynamics of the neuroglial potassium cycle during neurotransmission and the underlying molecular pathways is a key issue ., Here , we have developed a tri-compartment model accounting for potassium dynamics between neurons , astrocytes and the extracellular space to quantify the specific and acute contribution of astroglial Kir4 . 1 channels to extracellular potassium levels and to the moment-to-moment neurotransmission ., We demonstrate that astroglial Kir4 . 1 channels are sufficient to account for the slow membrane depolarization of astrocytes and crucially contribute to extracellular potassium clearance during basal and high activity ., We also show that astrocytes buffer in less than 10 seconds more than 80% of the potassium released by neurons , leading to recovery of basal extracellular potassium levels and neuronal excitability ., Remarkably , we found that Kir4 . 1 channels also prominently regulate slow 3 to 10 Hz rhythmic firing activity ., Altogether , these data show that Kir4 . 1 channels acutely regulate extracellular potassium and neuronal excitability during specific patterns of activity . | null | null |
journal.pgen.0030049 | 2,007 | A Dinucleotide Deletion in CD24 Confers Protection against Autoimmune Diseases | Multiple sclerosis ( MS ) is a chronic , inflammatory neurodegenerative disease of the central nervous system of unknown etiology ., There is evidence to support the hypothesis that MS is an autoimmune process modulated by both genetic and environmental factors 1–6 ., An increased risk of MS among MS relatives has been found in numerous prospective epidemiological studies 2 , 4 , 7 ., Twin studies from different populations consistently indicate that a monozygotic twin has a 5- to 6-fold higher risk of MS than a dizygotic twin 1 , 2 , 8 ., Collectively , these findings would implicate that , at least in part , the risk for developing this disorder and possibly its progression are mediated by multiple genetic factors ., Several whole-genome screens were performed in MS affected families ., These studies confirmed the association of MS with the HLA class II DR2 haplotype ( HLA-DRB1*1501-DQA1*0102-DQB1*0602 ) , but failed to confirm other major putative loci in MS 9–11 ., Systemic lupus erythematosus ( SLE ) is a classic systemic autoimmune disease with diverse clinical symptoms , including fatigue , joint pain and swelling , skin rashes , and chest pain ., Severe SLE complications include nephritis , central nervous system vasculitis , pulmonary hypertension , interstitial lung disease , and stroke ., Whole-genome scans have revealed multiple chromosomal regions 12–17 ., However , the identity of most susceptibility genes are unknown 18 ., CD24 is a glycosylphosphatidylinositol-anchored cell surface protein with expression in a variety of cell types that can participate in the pathogenesis of MS and SLE , including activated T cells 19 , 20 , B cells 21 , macrophages 22 , and dendritic cells 23 ., CD24 , as a candidate locus 10 , was shown to be essential for the induction of experimental autoimmune encephalomyelitis ( EAE ) in mice 24 ., Interestingly , CD24 controls a checkpoint of EAE pathogenesis after the autoreactive T cells are produced 24 ., Recently , we showed that CD24 is essential for local clonal expansion and persistence of T cells after their migration into the central nervous system and that expression of CD24 on either hematopoietic cells or nonhematopoietic antigen-presenting cells in the recipient is sufficient to confer susceptibility to EAE 25 ., These findings suggest that CD24 is essential for susceptibility to EAE ., Human CD24 ( CD24 ) mRNA has a 0 . 24-kb ORF and a 1 . 8-kb 3′ UTR ., A CT single nucleotide polymorphism ( SNP ) at position 170 from the CD24 translation start site ( P170 ) in the CD24 putative cleavage site for the glycosylphosphatidylinositol anchor ( −1 position ) 26 results in the nonconservative replacement of alanine with valine ., The P170TT genotype expressed higher cell-surface CD24 than the P170CT or P170CC genotypes , which had an increased risk and more rapid progression of MS 27 ., Thus , the CD24 SNP may influence MS pathogenesis by affecting the expression of CD24 ., The potential contribution of CD24 to SLE has not been studied ., However , since CD24 has emerged as a major checkpoint of homeostatic proliferation in lymphopenic hosts 28 , 29 , and since leucopenia is a defining feature of SLE 30 , it would be of great interest to evaluate whether CD24 polymorphism may affect susceptibility to SLE ., Interestingly , the long 3′ UTR of mouse Cd24 mRNA plays an important role in controlling CD24 expression 31 ., Two cis elements of mouse Cd24 mRNA , a negative and a positive cis element , regulate the stability of mouse Cd24 mRNA expression and determine cell-surface CD24 expression 31 ., Our sequencing analysis of the 3′ UTR of CD24 revealed four polymorphisms in the Ohio population ., Considering the importance of CD24 in the development and progression of MS , we investigated the association of the CD24 polymorphisms at the 3′ UTR with the susceptibility to both organ-specific and systemic autoimmune diseases ., Our study revealed a dinucleotide deletion in the 3′ UTR of human CD24 that confers significant protection against the risk and progression of MS and the risk of SLE by destabilizing CD24 mRNA ., CD24 has been identified as an autosomal gene located in Chromosome 6q21 32 , with intronless pseudogenes in Chromosomes 1 , 15 , and Y . In addition to a lack of introns , it has been reported that Chromosome Y DNA sequence differs from CD24 cDNA at 23 positions , with two changes in the coding regions and the remaining ones scattered through the 1 . 6-kb cDNA region 32 ., The CD24 gene sequence , as assembled by Celera , is presented in Figure S1 ., However , a recent update in the National Center for Biotechnology Information ( NCBI ) database placed CD24 on Chromosome Y with partial intron 1 sequence and exon 2 identical to the cDNA except for eight changes in the region corresponding to the 3′ end of the cDNA ., We used PCR primers antisense to a portion of the intronic sequence and the 3′ end of the CD24 mRNA to amplify from genomic DNA , using as templates genomic DNA from eight unrelated normal individuals ( four males and four females ) ., Since the primers would amplify both the Chromosome 6 sequence and the putative Chromosome Y sequence ( Figure S2 ) , we sequenced five clones from each of the eight individuals in order to determine which annotation is correct ., We found that none of the 40 sequences matched the putative Chromosome Y sequence , regardless of the sex of the donor , and all sequences matched the intron 1 and exon 2 sequence of the CD24 gene as located on Chromosome 6 32 ( Figure S2 ) ., These results indicated that the putative Chromosome Y sequence is likely incorrect and that the PCR primer pair amplifies the autosomal CD24 gene ., Analysis of five clones from each of the eight individuals also revealed four SNPs , three of which were reported in the NCBI database ( P1056 A/G , P1527 TG/del , and P1626 A/G ) ., As shown in Figure 1 , following an anchored PCR designed to eliminate the contribution of intronless CD24 pseudogenes 32 , these three polymorphic sites could be identified by restriction enzyme digestion of individual PCR products ., The accuracy of the PCR-restriction length fragment polymorphism ( RFLP ) was confirmed by sequencing the PCR products from 32 individuals ., The PCR-RFLP analysis was therefore adopted for genotyping ., The genotype distributions of these polymorphisms did not deviate from the Hardy-Weinberg equilibrium ( Table S1 ) ., Moreover , the genotype distributions are essentially the same among males and females among the large set of samples tested ( Table S1 ) ., We also used Merlin software ( http://www . sph . umich . edu/csg/abecasis/Merlin ) to detect potential genotyping errors 33 ., No Mendelian inconsistency and obligatory double recombination were found ., Taken together , these data ruled out the possibility that the Chromosome Y locus contributes to the data presented in this study and confirmed the accuracy of the genotypes presented ., We examined the association of the CD24 polymorphisms in the 3′ UTR with MS using DNA from independent Caucasian participants with MS and race- , age- and gender-matched controls from Central Ohio ( Table 1 ) ., A summary of the CD24 allele and genotype analyses of the MS patients compared with controls is shown in Table 2 ., A significant difference in the allelic frequencies between the MS patients and the controls was found for P1527 ( p = 0 . 006 ) , but not for P1056 or P1626 ., Remarkably , an approximately 2-fold decrease in the risk for MS was found in participants with the P1527TG/del or P1527del/del genotypes compared with the P1527TG/TG genotype ., These data suggest that the dinucleotide deletion may confer protection against MS risk ., Since the above case-control results could potentially be due to population admixture , even though we have restricted the analysis to only the Caucasian samples , we also considered transmission disequilibrium tests using family data , as such a test is still valid under population admixture ., We used a total of 150 pedigrees , including 49 pedigrees from the Multiple Sclerosis Genetics Group ( MSGG ) ( with 63 informative nuclear families ) and 101 from central Ohio ( with 93 informative nuclear families ) , to determine whether the CD24 polymorphisms are associated with MS risk ., The family compositions of both cohorts are shown in Table S2 ., A strong association in P1527 was found using TRANSMIT ( http://www-gene . cimr . cam . ac . uk/clayton/software ) 34 , 35 ( p = 0 . 002 ) ., No significant association was observed with other SNPs ., Linkage disequilibrium ( LD ) analysis of the four SNPs using the 150 MS and 187 SLE family samples revealed a surprisingly low LD between P170 and P1527 ( Figure 2A and 2B ) ., Considering the short distance between the SNPs , it is possible that a recombination hotspot may exist in the CD24 gene ., These results , together with the fact that P1056 and P1626 are not significantly associated with MS susceptibility , suggest that P170 and P1527 are independently associated with MS risk ., Such an interpretation is plausible since the allele frequencies of the SNPs are not very similar , which diminishes the power of detecting association even for a nearby SNP in high LD with the causal SNP ., The significance of the origin of the participant in the association of P170 has been highlighted in recent studies by us and others 27 , 36 , 37 ., MS disease severity is usually measured according to the expanded disability status scale ( EDSS ) ., MS patients who have lost the ability to walk without aid have reached EDSS 6 . 0 ., For the majority of the patients , their EDSS 6 . 0 status was based on a follow-up visit to our center ., A few of the cases were based on case history ., Because this is one of the most traumatic events in a patients life , most can recall accurately the time when their disease reached EDSS 6 . 0 ., We then tested whether the CD24 genotype affected the time span it took the patients to reach EDSS 6 . 0 from the day of the first symptom of MS . Clinical data from 275 independent Caucasian MS patients in the Ohio cohorts , but not those from the MSGG , were available for the survival analysis ., The Kaplan-Meier curves provide estimates of the distribution of the time it took to reach EDSS 6 . 0 for patients with different genotypes ., As shown in Figure 3 , patients with the P1527TG/del or P1527del/del genotype had a more delayed disease progression pattern than those with the P1527TG/TG genotype ( p = 0 . 0188 ) ., In addition , the patients with the P1626AA genotype also showed faster progression ( p = 0 . 0105 ) ., No significant result was found in the patients with the P1056 genotype ., We used a Caucasian cohort of age and sex-controlled samples ( Table 3 ) to test the potential association between CD24 polymorphism and risk of SLE ., A summary of the CD24 allele and genotype analyses in the SLE patients against controls is shown in Table 4 ., A significant difference in the allelic frequencies between the SLE patients and the controls was found for P1527 ( p = 0 . 00003 ) , but not for P1056 or P1626 ., Remarkably , a 2 . 6-fold decrease in the risk for SLE was found in participants with the P1527TG/del or P1527del/del genotype compared to the P1527TG/TG genotype ., These data suggest that the dinucleotide deletion may confer protection against SLE risk ., We used a total of 187 pedigrees to determine whether the CD24 polymorphisms are associated with SLE risk ( with 187 informative nuclear families ) ., A strong association in P1527 was found using TRANSMIT ( p = 0 . 002 ) , but it did not show evidence for transmission disequilibrium for P170 , P1056 , or P1626 ( p > 0 . 05 ) ., Linkage disequilibrium analysis of the four SNPs using the 187 family samples also revealed a low LD between P170 and P1527 ( Figure 2B ) , suggesting that P1527 is independently associated with SLE risk , the same as in MS . Since P1527 resides in the 3′ UTR , its polymorphism may affect the accumulation of its mRNA ., To address whether mRNA transcribed from the P1527del allele presents a decrease in its expression levels in vivo , we established an allele-specific real-time PCR ( RT-PCR ) to measure the allele-specific transcripts ., As shown in Figure 4A , the primers designed for the P1527TG allele detected CD24 mRNA in the P1527TG/TG , but not in the P1527del/del individuals , and vice versa ., These results demonstrate complete specificity of the primers used ., In addition , the conditions used led to the amplification of CD24 cDNA in a strictly dose-dependent fashion over six logs of magnitude ( Figure 4B ) ., We therefore used this method to measure the allele-specific expression of two CD24 alleles in eight P1527TG/del individuals ., As shown in Figure 4C , the P1527del transcripts were 2 . 5-fold less than the P1527TG transcripts ., Since the two alleles were present in the same cells and therefore were transcribed at the same rate , our data demonstrate that the P1527 variant has a strong impact on mRNA expression of CD24 in vivo , most likely by post-transcriptional mechanisms ., P1527 is located in the 3′ UTR that modulates mRNA stability 31 ., To test if this polymorphism modulates CD24 mRNA stability , we constructed two plasmids ( pTracer CMV2-CD24TG and pTracer CMV2-CD24del; Figure 5 , top panel ) and transfected Chinese hamster ovary ( CHO ) cells with the two constructs ., Starting at 48 hours after transfection , the synthesis of RNA was blocked by actinomycin D , and the half life of mRNA was measured by RT- PCR ., The levels of GFP mRNA were used as internal controls for transfection efficiency ., Prior to actinomycin D treatment , there was significantly higher mRNA expression for the CD24TG cDNA in comparison to the CD24del cDNA ., Using the pre-treatment mRNA levels as 100% , we measured the decay kinetics of two mRNA from two different cDNA ., As shown in Figure 5 ( lower panel ) , the decay patterns of CD24TG were significantly more gradual than those of CD24del ( p < 0 . 001 ) , and the differences in the rates of decay were significantly different at all time points starting from 0 . 5 h ( all p < 0 . 001 ) ., In particular , mRNA from the CD24del cDNA had a half life of less than 1 h , while that derived from the CD24TG had a half life of more than 4 h ., Thus , the dinucleotide deletion at the P1527 position destabilized CD24 mRNA ., It is well established that polymorphisms of immune-related genes modulate host susceptibility to autoimmune diseases , including MS and SLE 27 , 38–42 ., Historically , most studies have focused on polymorphisms that result in the replacement of amino acids 27 , 38 , 40 ., More recently , substantial information has been accumulated that demonstrates that polymorphisms at the promoter and intron regions can also have a significant impact on susceptibility ., These alterations modulate either RNA synthesis ( transcription ) or splicing 41 , 42 ., Although it is well established that the 3′ UTR plays a major role in RNA stability , we are not aware of any study reporting that polymorphism at the 3′ UTR modulates susceptibility to autoimmune diseases by changing mRNA stability ., Our data presented in this study revealed that a destabilizing dinucleotide deletion in the 3′ UTR of the CD24 gene may confer a significant protection against the risk and progression of MS and against the risk of SLE ., Our conclusion is based on five lines of evidence ., First , a population study with 275 independent Caucasian MS patients and a comparable size of normal controls revealed that individuals with the deletion in at least one allele had about a 2-fold less relative risk in comparison to those without the deletion ., Thus , the CD24 P1527del allele may be a protective genetic susceptibility factor for the onset of MS . This is more remarkable in light of the fact that polymorphisms at sites that were only 100–500 bp apart did not have a significant impact on the risk of MS . The strong association at P1527 , but not at the nearby SNPs , suggests that the deletion was causatively related to the reduced MS susceptibility ., This interpretation is consistent with the fact that the frequencies of the associated alleles at the two nearby ( flanking ) loci are very different from that of the protective allele ., A recent study showed that the power to detect the association in such loci is diminished even when there is high linkage disequilibrium 43 ., This also leads to a reasonable explanation as to why two loci in high LD are not both associated with the disease ., Second , using data from two independent cohorts of families , we also established a strong association of the CD24 P1527 polymorphism with MS . The P1527TG allele was preferentially transmitted to affected individuals ., This result strongly supports the conclusion from the case-control analysis that the P1527del allele may be a protective genetic susceptibility factor for the onset of MS . Both of these results remain significant after multiple-testing adjustments ., Within the Ohio State University ( OSU ) cohort , our previous data revealed that the P170T allele was preferentially transmitted into affected individuals among multiplex families with two or more MS patients 27 ., This result continues to hold with our expanded OSU family set , although not with the MSGG cohort ( data not shown ) ., In summary , results from both of population and the family studies confirm our earlier conclusion that the CD24 locus is a major modulator for MS risk ., Third , survival analysis revealed a significant association ( even after correcting for multiple tests ) of CD24 P1527 with MS disease progression; MS patients with the P1527del allele had a significantly delayed progression ., This finding further confirms that the P1527del allele is a protective genetic factor for MS . An interesting issue is whether P1527 is associated with the progression of MS because of its linkage to P170 ., We consider it very unlikely as our analysis of LD revealed that there is little LD between the two sites despite their close proximity to one another , perhaps due to a recombination hotspot within the CD24 gene ., Moreover , P1056 , which is closer to P170 , is not associated with the progression of MS . We therefore consider it likely that P170 and P1527 are independently associated with the progression of MS . Since P1626 is less than 100 bp away and shows a strong LD with P1527 , it remains possible that its association with MS progression may be due to its proximity to P1527 ., This interpretation is favored as P1626 shows no association with MS risk ., Since our analysis has now covered all known CD24 polymorphisms in the exons , it is likely that P1527 , rather than other SNPs , is related to protection against autoimmune diseases ., Fourth , in addition to MS , which is an organ-specific autoimmune disease , we also observed that the CD24 P1527del allele is preferentially transmitted to unaffected individuals in the SLE family data ., It is worth noting that the SLE data should not be regarded as a replication of MS data per se ., Rather , our data suggest that the protective effect of the dinucleotide deletion extends to systemic autoimmune diseases ., Thus , in addition to its critical role for T-cell proliferation in the central nervous system 25 , CD24 may play a role in the development of multiple autoimmune diseases ., Based on the observed data pattern and the structure of the family cohorts , we have chosen TRANSMIT soft ware to detect association between CD24 polymorphism and risk of autoimmune diseases to maximize the statistical power ., However , we caution that TRANSMIT may have inflated type-I error due to its inferences of missing parental genotypes 44 ., Nevertheless , we do not believe the core finding is due to type-I errors , as statistically significant association can also be find with FBAT that deletes data from families without parental information ( MS dataset , p = 0 . 04; SLE data set , p = 0 . 01 ) ., Fifth , the dinucleotide deletion reduced steady levels of CD24 mRNA by more than 2-fold ., Thus , in heterozygous patients , the mRNA from the alleles with the deletion was only 50% of that of the alleles without the deletion ., This is recapitulated in transfection studies ., Analysis of RNA decay kinetics revealed that the half life for the CD24 transcript with the dinucleotide deletion was at least 4-fold shorter than that of the wild-type allele ., Since CD24 was expressed at high levels among some lineages of hematopoietic cells and in the transfected CHO cells , the reduction in the steady levels may underestimate the reduction in other cell types , such as T cells , in which CD24 is expressed at lower levels and is therefore less likely to saturate the degradation system ., The low expression of CD24 in T cells is essential for homeostatic proliferation of T cells , which has been implicated in the development of autoimmune diseases ., In summary , we demonstrated that a dinucleotide deletion at the 3′ UTR of the human CD24 gene confers significant protection against the risk and progression of MS and the risk of SLE ., These results not only provide insight into the genetic basis of MS and SLE susceptibility , but , perhaps more importantly , to our knowledge , this is the first report that shows how polymorphisms at the 3′ UTR modulate susceptibility to autoimmune diseases by regulating RNA stability ., Since CD24 is a checkpoint for homeostatic proliferation of T cells 29 , which is implicated in other autoimmune diseases 45 , it will be of great interest to test the contribution of CD24 to the risk and progression of other autoimmune diseases ., All sample collection and experimentation was approved by the Institutional Review Board , and informed consents from all participants were obtained before sample collection ., Some of the participants had been enrolled in the previous study 27 ., Patients with definite MS , as diagnosed by K . W . R . and D . J . L . at OSUMultiple Sclerosis Center according to the McDonald criteria 46 , were offered the opportunity to participate ., The clinical diagnosis of MS type and the EDSS score 47 were determined by three of the authors ( K . W . R . , D . J . L . , and N . G . ) ., The time of disease onset and the time when a walking aid was first prescribed ( EDSS 6 . 0 ) were determined retrospectively by the analysis of case records without knowing CD24 genotype ., In the case-control study , we selected a consecutive series of 829 participants including 361 MS patients and 468 normal controls ., For the case-control analysis based on participants with the same genetic background , we only used 275 independent Caucasian cases and age and gender-matched 443 Caucasian controls ( Table 1 ) ., All MS patients were recruited at the OSU Medical Center between January 2000 and March 2006 and agreed to participate in this study ., All donors gave written informed consent ., The control participants were obtained from the American Red Cross ( Columbus , Ohio ) between September 1999 and January 2006 using leftover peripheral blood samples ., In the Ohio family study , 101 pedigrees of MS families were used for association analysis ., Of the 346 participants from the families , 135 were MS patients and 211 were non-MS relatives ., All MS patients and their unaffected family members were recruited at the OSU Medical Center , and all agreed to participate in this study ., We interviewed all MS patients for family history of MS . Consenting family members with or without MS provided blood samples as well ., In rare cases , when family members were at other locations , samples were obtained by local physicians or nurses and transported or mailed to our center ., Ascertainment of the presence or absence of MS among the relatives was by history alone ., Relatives who provided blood samples were not subjected to neurological evaluation or an MRI at our center ., These participants were selected between October 2001 and May 2005 ., In the MSGG family study , 321 participants from 49 pedigrees of multiplex MS families were obtained from the MSGG through the University of California San Francisco ., Of the participants , 119 were MS patients and 202 were from non-MS relatives ., These participants were selected between October 1997 and January 2003 ., Demographics and disease characteristics of the MS patients and controls are summarized in Table 1 ., The sex ratio and average age of the OSU MS patients were not significantly different from those of normal controls ( p = 0 . 506 in sex; p = 0 . 970 in age ) ., In all of the OSU MS patients , as well as in each of the familial and sporadic groups , there were no significant correlations among age , age at onset , EDSS , duration of the disease , and clinical course ( all p > 0 . 05 ) ., In the OSU MS patients , no information was obtained for the EDSS score in four patients and the clinical course in three patients ., The group of patients with some missing phenotypic information was included in our genetic analysis to be detailed below ., The comparison of clinical and demographic features between OSU and MSGG family MS patients did not show any significant differences ( p > 0 . 05 ) ., Although there was no significant difference in the ethnicity between the MS patients of the OSU and the MSGG families , the MS patients of the MSGG families were from a number of other countries besides the United States ., Demographics and disease characteristics of the SLE patients and controls are summarized in Table 3 ., A total of 264 unrelated SLE patients were consecutively recruited at the Columbus Childrens Hospital and Research Institute , OSU , and followed in the Ohio SLE Study ., SLE cases for case-control analysis were all independent Caucasians ., Healthy race- , sex- , and age-matched participants ( 270 ) with no history of autoimmune disease were enrolled from the American Red Cross ( Columbus , Ohio ) ., The sex ratio and average age of SLE patients were not significantly different from those of normal controls ( p = 0 . 435 in sex; p = 0 . 990 in age ) ., The healthy participants were completely independent from the control participants in the MS group ., Both case and control samples were collected between 1999 and 2006 ., A large collection of 187 pedigrees of SLE families was obtained from the Columbus Childrens Hospital and Research Institute , OSU , with predominantly one affected offspring per family ., Of the 555 participants from the families , 187 were SLE patients and 368 were non-SLE relatives ., Samples from both parents were available for 36% of the families , and samples from siblings were also collected where available ( Table S2 ) ., In the case of single-parent families , samples were always taken from siblings ., An extensive questionnaire and interview with a trained physician were completed by unaffected family members to determine the absence of SLE ., The SLE patients were diagnosed according to the classification criteria of the American College of Rheumatology 30 , 48 ., Only those that were diagnosed as definitive SLE were included in the study ., The demography and clinical data for the samples were listed in Table 3 , using kidney involvement and WHO classifications for disease severity 48 ., All participants were Caucasians who gave written informed consent ., Approval for human study protocols was obtained from the human subjects review board at OSU and the Institutional Review Board ., The genomic DNA was isolated from peripheral blood leukocytes ( PBL ) by using the QIAamp DNA Blood Minikit ( Qiagen , http://www . qiagen . com ) ., We searched for polymorphisms in the 3′ UTR of exon 2 using PCR and DNA sequencing , and these polymorphisms were further determined by DNA cloning and sequencing ., Since several intronless CD24 pseudogenes have been identified in the human genome 32 , the functional CD24 locus was selectively amplified by nested PCR ( Figure 1 ) ., The first PCR amplification ( Invitrogen http://www . invitrogen . com ) was from intron 1 to the end of exon 2 by using a forward primer ( 5′-CTA AAG AGA ATG ACC TTG GTG GGT TGA G-3′ ) and a reverse primer ( 5′-CAC AGT AGC TTC AAA ACT GTT CGA-3′ ) ., The PCR conditions were as follows: 94 °C for 30 s , 55 °C for 30 s , and 68 °C for 2 min for 20 cycles ., The predicted CD24 PCR fragment was 2 , 017 bp long ., The identity of the PCR products to the CD24 gene sequence on Chromosome 6 , but not the putative Chromosome Y locus sequence as well as the SNP in the region was confirmed by cloning and sequencing of the PCR products ( Figure S2 ) ., The second PCR amplification ( Promega , http://www . promega . com ) was based on each polymorphic site using the primers as follows: a forward primer ( 5′-CTA AAG AGA ATG ACC TTG GTG GGT TGA G-3′ ) and a reverse primer ( 5′-GGA TTG GGT TTA GAA GAT GGG GAA A-3′ ) for 170 C/T polymorphism ( P170 ) from the CD24 translation start site , a forward primer ( 5′-GGC ATT TCC TAT CAC CTG TTT-3′ ) and a reverse primer ( 5′-AAT CTA CCC CCA GAT CCA AGC A-3′ ) for 1056 A/G polymorphism ( P1056 ) , a forward primer ( 5′-GCA ATT TTG CCT TCA AAA CAG-3′ ) and a reverse primer ( 5′-TTT AGG CTT AGG ACC AGG TTC-3′ ) for 1527∼1528 TG/del polymorphism ( P1527 ) , and a forward primer ( 5′-CAA CTA TGG ATC AGA ATA GCA ACA AT-3′ ) and a reverse primer ( 5′-GGAACATCTAAGCATCAGTGTGTG-3′ ) for 1626 A/G polymorphism ( P1626 ) ., The PCR conditions were as follows: 94 °C for 30 s , 55 °C for 30 s , and 72 °C for 30 s , for 35 cycles ., The PCR products were digested overnight with BstXI ( 50 °C ) for P170 , BstUI ( 60 °C ) for P1056 , BsrI ( 65 °C ) for P1527 , and MfeI ( 37 °C ) for P1626 ( New England Biolabs , http://www . neb . com ) and then electrophoresed on 3 . 0% agarose gels ( Figure 1 ) ., The genotypes were designated as “C , ” “A , ” “del , ” or “A” when the restriction sites of BstXI , BstUI , BsrI , and MfeI were respectively absent , and as “T , ” “G , ” “TG , ” or “G” when each restriction site was respectively present ( Figure 1 ) ., The validity of the PCR-RFLP analysis was confirmed by direct sequencing of several PCR samples with each genotype ., CD24 cDNA was amplified from the peripheral blood leukocyte of individuals with the P1527TG/del genotype by RT-PCR ( Invitrogen ) ., The following primers were used: a forward primer ( 5′-ATG GGC AGA GCA ATG GTG-3′ ) and a reverse primer ( 5′-CAC AGT AGC TTC AAA ACT GTT CGA-3′ ) ., The PCR products ( 1 , 842 bp ) were cloned into the TOPO-pCDNA2 . 0 vector ( Invitrogen ) , which was digested with KpnI/NotI , and then the PCR products with the additional KpnI/NotI site were cloned into the pTracer CMV2 vector ( Invitrogen ) , thus generating two plasmids , pTracer CMV2-CD24TG and pTracer CMV2-CD24del ., The sequence of two CD24 cDNA inserts was confirmed by DNA sequencing ., To exclude potential confounding factors , we selected the same sequence at the P170 , P1056 , and P1626 sites between the two plasmids ., To test the expression efficiency of the CD24 alleles , we transfected varying concentrations of plasmids into CHO cells using FuGENE 6 ( Roche , http://www . roche . com ) ., For the RNA stability experiment , 48 h after transfection , CHO cells were treated with actinomycin D ( 5 μg/ml ) ( Sigma , St . Louis , Mo . ) for 0 . 5 , 1 , 2 , 3 , and 4 h ., Untreated cells were used as control at the 0 h time point ., We isolated total RNA from 1 × 106 transfected CHO cells using a commercially available kit ( Qiagen ) ., We exposed RNA samples to DNase digestion before cDNA synthesis ., For gene-specific PCR , 1 μl of first-strand cDNA product was amplified with platinum Taq polymerase ( Invitrogen ) according to the manufacturers instructions ., We designed primers specific for CD24 ( forward: 5′-CCC ACG CAG ATT TAT TCC AGT-3′ , reverse: 5′-TGG TGG TGG CAT TAG TTG GAT-3′ ) and for GFP ( forward: 5′-GGT GAT GTT AAT GGG CAC AA-3′ , reverse: 5′-TAG TGA CAA GTG TTG GCC ATG-3′ ) and performed a 30-cycle , three-step PCR ( denaturation at 95 | Introduction, Results, Discussion, Materials and Methods, Supporting Information | It is generally believed that susceptibility to both organ-specific and systemic autoimmune diseases is under polygenic control ., Although multiple genes have been implicated in each type of autoimmune disease , few are known to have a significant impact on both ., Here , we investigated the significance of polymorphisms in the human gene CD24 and the susceptibility to multiple sclerosis ( MS ) and systemic lupus erythematosus ( SLE ) ., We used cases/control studies to determine the association between CD24 polymorphism and the risk of MS and SLE ., In addition , we also considered transmission disequilibrium tests using family data from two cohorts consisting of a total of 150 pedigrees of MS families and 187 pedigrees of SLE families ., Our analyses revealed that a dinucleotide deletion at position 1527∼1528 ( P1527del ) from the CD24 mRNA translation start site is associated with a significantly reduced risk ( odds ratio = 0 . 54 with 95% confidence interval = 0 . 34–0 . 82 ) and delayed progression ( p = 0 . 0188 ) of MS . Among the SLE cohort , we found a similar reduction of risk with the same polymorphism ( odds ratio = 0 . 38 , confidence interval = 0 . 22–0 . 62 ) ., More importantly , using 150 pedigrees of MS families from two independent cohorts and the TRANSMIT software , we found that the P1527del allele was preferentially transmitted to unaffected individuals ( p = 0 . 002 ) ., Likewise , an analysis of 187 SLE families revealed the dinucleotide-deleted allele was preferentially transmitted to unaffected individuals ( p = 0 . 002 ) ., The mRNA levels for the dinucleotide-deletion allele were 2 . 5-fold less than that of the wild-type allele ., The dinucleotide deletion significantly reduced the stability of CD24 mRNA ., Our results demonstrate that a destabilizing dinucleotide deletion in the 3′ UTR of CD24 mRNA conveys significant protection against both MS and SLE . | When an individuals immune system attacks self tissues or organs , he/she develops autoimmune diseases ., Although it is well established that multiple genes control susceptibility to autoimmune diseases , most of the genes remain unidentified ., In addition , although different autoimmune diseases have a common immunological basis , a very small number of genes have been identified that affect multiple autoimmune diseases ., Here we show that a variation in CD24 is a likely genetic factor for the risk and progression of two types of autoimmune diseases , including multiple sclerosis ( MS ) , an organ-specific autoimmune disease affecting the central nervous system , and systemic lupus erythematosus , a systemic autoimmune disease ., Our data indicated that if an individuals CD24 gene has a specific two-nucleotide deletion in the noncoding region of CD24 mRNA , his/her risk of developing MS or SLE is reduced by 2- to 3-fold ., As a group , MS patients with the two-nucleotide deletion will likely have a slower disease progression ., Biochemical analysis indicated that the deletion leads to rapid decay of CD24 mRNA , which should result in reduced synthesis of the CD24 protein ., Our data may be useful for the treatment and diagnosis of autoimmune diseases . | genetics and genomics, rheumatology, neurological disorders, vertebrates | null |
journal.pcbi.1005517 | 2,017 | Automatically tracking neurons in a moving and deforming brain | Optical neural imaging has ushered in a new frontier in neuroscience that seeks to understand how neural activity generates animal behavior by recording from large populations of neurons at cellular resolution in awake and behaving animals ., Population recordings have now been used to elucidate mechanisms behind zebra finch song production 1 , spatial encoding in mice 2 , and limb movement in primates 3 ., When applied to small transparent organisms , like Caenorhabditis elegans 4 , Drosophila 5 , and zebrafish 6 , nearly every neuron in the brain can be recorded , permitting the study of whole brain neural dynamics at cellular resolution ., Methods for segmenting and tracking neurons have struggled to keep up as new imaging technologies now record from more neurons over longer times in environments with greater motion ., Accounting for brain motion in particular has become a major challenge , especially in recordings of unrestrained animals ., Brains in motion undergo translations and deformations in 3D that make robust tracking of individual neurons very difficult ., The problem is compounded in invertebrates like C . elegans where the head of the animal is flexible and deforms greatly ., If left unaccounted for , brain motion not only prevents tracking of neurons , but it can also introduce artifacts that mask the true neural signal ., In this work we propose an automated approach to segment and track neurons in the presence of dramatic brain motion and deformation ., Our approach is optimized for calcium imaging in unrestrained C . elegans ., Neural activity can be imaged optically with the use of genetically encoded calcium sensitive fluorescent indicators , such as GCaMP6s used in this work 7 ., Historically calcium imaging was often conducted in head-fixed or anesthetized animals to avoid challenges involved with imaging moving samples 4 , 8 , 9 ., Recently , however , whole-brain imaging was demonstrated in freely behaving C . elegans 10 , 11 ., C . elegans are a small transparent nematode , approximately 1mm in length , with a compact nervous system of only 302 neurons ., About half of the neurons are located in the animal’s head , which we refer to as its brain ., Analyzing fluorescent images of moving and deforming brains requires algorithms to detect neurons across time and extract fluorescent signals in 3D ., Automated methods exist for segmenting and tracking fluorescently labeled cells during C . elegans embryogenesis 12 , and semi-automated methods are even able to track specific cells during embryo motion 13 , but to our knowledge these methods are not suitable for tracking neurons in adults ., Generally , several strategies exist for tracking neurons in volumetric recordings ., One approach is to find correspondences between neuron positions in consecutive time points , for example , by applying a distance minimization , and then stitching these correspondences together through time 14 ., This type of time-dependent tracking requires that neuron displacements for each time step are less than the distance between neighboring neurons , and that the neurons remain identifiable at all times ., If these requirements break down , even for only a few time points , errors can quickly accumulate ., Other common methods , like independent component analysis ( ICA ) 15 are also exquisitely sensitive to motion and as a result they have not been successfully applied to recordings with large brain deformations ., Large inter-volume motion arises when the recorded image volume acquisition rate is too low compared to animal motion ., Unfortunately , large inter-volume brain motion is likely to be a prominent feature of whole-brain recordings of moving brains for the foreseeable future ., In all modern imaging approaches there is a fundamental tradeoff between the following attributes: acquisition rate ( temporal resolution ) , spatial resolution , signal to noise , and the spatial extent of the recording ., As recordings seek to capture larger brain regions at single cell resolution , they necessarily compromise on temporal resolution ., For example , whole brain imaging in freely moving C . elegans has only been demonstrated at slow acquisition rates because of the requirements to scan the entire brain volume and expose each slice for sufficiently long time ., At these rates , a significant amount of motion is present between image planes within a single brain volume ., Similarly , large brain motions also remain between sequential volumes ., Neurons can move the entire width of the worm’s head between sequential volumes when recording at 6 brain-volumes per second , as in 10 ., In addition to motion , the brain also bends and deforms as it moves ., Such changes to the brain’s conformation greatly alter the pattern of neuron positions making constellations of neurons difficult to compare across time ., To track neurons in the presence of this motion , previous work that measured neural activity in freely moving C . elegans utilized semi-automated methods that required human proof reading or manual annotation to validate each and every neuron-time point 10 , 11 ., This level of manual annotation becomes impractical as the length of recordings and the number of neurons increases ., For example , 10 minutes of recorded neural activity from 10 , had over 360 , 000 neuron time points and required over 200 person-hours of proofreading and manual annotation ., Here , we introduce a new time-independent algorithm that uses machine learning to automatically segment and track all neurons in the head of a freely moving animal without the need for manual annotation or proofreading ., We call this technique Neuron Registration Vector Encoding , and we use it to extract neural signals in unrestrained C . elegans expressing the calcium indicator GCaMP6s and the fluorescent label RFP ., We introduce a method to track over 100 neurons in the brain of a freely moving C . elegans ., The analysis pipeline is made of five modules and an overview is shown in Fig 1 ., The first three modules , “Centerline Detection , ” “Straightening” and “Segmentation , ” collectively assemble the individually recorded planes into a sequence of 3D volumes and identify each neuron’s location in each volume ., The next two modules , “Registration Vector Construction” and “Clustering , ” form the core of the method and represent a significant advance over previous approaches ., Collectively , these two modules are called “Neuron Registration Vector Encoding . ”, The “Registration Vector Construction” module leverages information from across the entire recording in a time-independent way to generate feature vectors that characterize every neuron at every time point in relation to a repertoire of brain confirmations ., The “Clustering” module then clusters these feature vectors to assign a consistent identity to each neuron across the entire recording ., A final module corrects for errors that can arise from segmentation or assignment ., The implementation and results of this approach are described below ., Worms expressing the calcium indicator GCaMP6s and a calcium-insensitive fluorescent protein RFP in the nuclei of all neurons were imaged during unrestrained behavior in a custom 3D tracking microscope , as described in 10 ., Only signals close to the cell nuclei are measured ., Two recordings are presented in this work: a new 8 minute recording of an animal of strain AML32 and a previously reported 4 minute recording of strain AML14 first described in 10 ., The signal of interest in both recordings is the green fluorescence intensity from GCaMP6s in each neuron ., Red fluorescence from the RFP protein serves as a reference for locating and tracking the neurons ., The microscope provides four raw image streams that serve as inputs for our neural tracking pipeline , seen in Fig 2A ., They are: ( 1 ) low-magnification dark-field images of the animal’s body posture ( 2 ) low-magnification fluorescent images of the animal’s brain ( 3 ) high-magnification green fluorescent images of single optical slices of the brain showing GCaMP6s activity and ( 4 ) high-magnification red fluorescent images of single optical slices of the brain showing the location of RFP ., The animal’s brain is kept centered in the field of view by realtime feedback loops that adjust a motorized stage to compensate for the animal’s crawling ., To acquire volumetric information , the high magnification imaging plane scans back and forth along the axial dimension , z , at 3 Hz as shown in Fig 2B , acquiring roughly 33 optical slices per volume , sequentially , for 6 brain-volumes per second ., The animal’s continuous motion causes each volume to be arbitrarily sheared ., Although the image streams operate at different volume acquisition rates and on different clocks , they are later synchronized by flashes of light that are simultaneously visible to all cameras ., Each image in each stream is given a timestamp on a common timeline for analysis ., Each of the four imaging streams are spatially aligned to each other in software using affine transformations found by imaging fluorescent beads ., An example of the high magnification RFP recording is shown in S1 Movie ., The animal’s posture contains information about the brain’s orientation and about any deformations arising from the animal’s side-to-side head swings ., The first step of the pipeline is to extract the centerline that describes the animal’s posture ., Centerline detection in C . elegans is an active field of research ., Most algorithms use intensity thresholds to detect the worm’s body and then use binary image operations to extract a centerline 16–18 ., Here we use an open active contour approach 19 , 20 to extract the centerline from dark field images with modifications to account for cases when the worm’s body crosses over itself as occurs during so-called “Omega Turns . ”, In principle any method , automated or otherwise , that detects the centerlines should be sufficient ., At rare times where the worm is coiled and the head position and orientation cannot be determined automatically , the head and the tail of the worm are manually identified ., The animal’s centerline allows us to correct for gross changes in the worm’s position , orientation , and conformation ( Fig 3a ) ., We use the centerlines determined by the low magnification behavior images to straighten the high magnification images of the worm’s brain ., An affine transform must be applied to the centerline coordinates to transform them from the dark field coordinate system into the coordinate system of the high magnification images ., Each image slice of the worm brain is straightened independently to account for motion within a single volume ., The behavior images are taken at a lower acquisition rate than the high magnification brain images , so a linear interpolation is used to obtain a centerline for each slice of the brain volume ., In each slice , we find the tangent and normal vectors at every point of the centerline ( Fig 3b ) ., The points are interpolated with a single pixel spacing along the centerline to preserve the resolution of the image ., The image intensities along each of the normal directions are interpolated and the slices are stacked to produce a straightened image in each slice ( Fig 3c ) ., In the new coordinate system , the orientation of the animal is fixed and the gross deformations from the worm’s bending are suppressed ., More subtle motion and deformation , however , remains ., We further reduce shearing between slices using standard video stabilization techniques 21 ., Specifically , bright-intensity peaks in the images are tracked between neighboring image slices ., The coordinates of these peaks are used to calculate the affine transformations between neighboring slices of the volume using least squares ., All slices are registered to the middle slice by applying these transformations sequentially throughout the volume ., Each slice would undergo transformations for every slice in between it and the middle slice to correct shear throughout the volume ., A final rigid translation is required to align each volume to the first volume of the recording ., The translations are found by finding an offset that maximizes the cross-correlation between each volume and the initial volume ., A video of straightening is shown in S1 Movie ., Straightened images are used for the remaining steps of the analysis pipeline ., Only the final measurement of fluorescence intensity is performed in the original unstraightened coordinated system ., Before neuron identities can be matched across time , we must first segment the individual neurons within a volume to recovers each neuron’s size , location , and brightness ( Fig 3d and 3e ) ., Many algorithms have been developed to segment neurons in a dense region 22 , 23 ., We segment the neurons by finding volumes of curvature in fluorescence intensity in the straigthened volumes ., After an initial smoothing , we compute the 3D Hessian matrix at each point in space and threshold for points where all of the three eigenvalues of the Hessian matrix are negative ., This process selects for regions around intensity peaks in three dimensions ., In order to further divide regions into objects that are more likely to represent neurons , we use a watershed separation on the distance transform of the thresholded image ., The distance transform is found by replacing each thresholded pixel with the Euclidean distance between it and the closest zero pixel in the thresholded image ., This approach is sufficient to segment most neurons ., Occasionally neurons are missed or two neurons are incorrectly merged together ., These occasional errors are corrected automatically later in the pipeline ., Extracting neural signals requires the ability to match neurons found at different time points ., Even after gross alignment and straightening , neurons in our images are still subject to local nonlinear deformations and there is significant movement of neurons between volumes ., This remaining motion and deformation is clearly visible , for example , in S1 Movie ., Rather than tracking neurons sequentially in time , the neurons in each volume are characterized based on how they match to neurons in a set of reference volumes ., Our algorithm compares constellations of neurons in one volume to unannotated reference volumes and assigns correspondences or “matches” between the neurons in the sample and each reference volume ., We modified a point-set registration algorithm developed by Jian and Vemuri 24 to do this ( Fig 4a ) ., The registration algorithm represents two point-sets , a sample point-set denoted by X = {xi} and a reference point-set indicated by R = {ri} , as Gaussian mixtures and then attempts to register them by deforming space to minimize the distance between the two mixtures ., In their implementation , each point is modeled by a 3D Gaussian with uniform covariance ., Since we are matching images of neurons rather than just points , we can use the additional information from the size and brightness of each neuron ., We add this information to the representation of each neuron by adjusting the amplitude and standard deviation of the Gaussians ., The Gaussian mixture representation of an image is given by ,, f ( ξ , X ) =∑iAiexp ( −‖ξ−xi‖22 ( λσi ) 2 ) , ( 1 ), where Ai , xi , and σi are the amplitude , mean , and standard deviation of the i-th Gaussian ., These parameters are derived from the brightness , centroid , and size of the segmented neuron , while ξ is the 3D spatial coordinate ., A scale factor λ is added to the standard deviation to scale the size of each Gaussian ., This will be used later during gradient descent ., The sample constellation of neurons is then represented by the Gaussian mixture f ( ξ , X ) ., Similarly , the reference constellation’s own neurons are represented as a f ( ξ , R ) ., To match a sample constellation of neurons X with a reference constellation of neurons R , we use the non rigid transformation u : I R 3 ↦ I R 3 ., The transformation maps X to uX such that the L2 distance between f ( ξ , uX ) and f ( ξ , R ) is minimized with some constraint on the amount of deformation ., This can be written as an energy minimization problem , with the energy of the transformation , E ( u ) , written as, E ( u ) = ∫ f ( ξ , u X ) - f ( ξ , R ) 2 d ξ + E Deformation ( u ) ., ( 2 ) Note that the point-sets X and R are allowed to have different numbers of points ., We model the deformations as a thin-plate spline ( TPS ) ., The TPS transformation equations and resulting form of EDeformation ( u ) are shown in the methods ., The minimization of E is found by gradient descent ., Working with Gaussian mixtures as opposed to the original images allows us to model the deformations and analytically compute the gradients of Eq 2 making gradient descent more efficient ., The gradient descent approach used here is similar to that outlined by Jian and Vemuri 25 ., Since the energy landscape has many local minima , we initially chose a large scale factor , λ , to increase the size of each Gaussian and smooth over smaller features ., Gradient descent is iterated multiple times with λ decreasing multiple times ., After the transformation , sample points are matched to reference points by minimizing distances between assigned pairs using an algorithm from 14 ., The matching is not greedy , and neurons in the sample that are far from any neurons in the reference are not matched ., A neuron at xi is assigned a match vi to indicate which neuron in the set R it was matched to ., For example if xi matched with rj when X is registered to R , then vi = j ., If xi has no match in R , then vi = ∅ ., The modified non-rigid point-set registration algorithm described above allows us to compare one constellation of neurons to another ., In principle , neuron tracking could be achieved by registering the constellation of neurons at each time-volume to a single common reference ., That approach is susceptible to failures in non-rigid point-set registration ., Non-rigid point-set registration works well when the conformation of the animal in the sample and the reference are similar , but it is unreliable when there are large deformations between the sample and the reference , as happens with some regularity in our recordings ., In addition , this approach is especially sensitive to any errors in segmentation , especially in the reference ., An alternative approach would be to sequentially register neurons in each time volume to the next time-volume ., This approach , however , accumulates even small errors and quickly becomes unreliable ., Instead of either of those approaches , we use registration to compare the constellation of neurons at each time volume to a set of reference time-volumes that span a representative space of brain conformations ( Fig 4b ) , as described below ., The constellation of neurons at a particular time in our recording is given by Xt , and the position of the i-th neuron at time t is denoted by xi , t ., We select a set of K reference constellations , each from a different time volume Xt in our recording , so as to achieve a representative sampling of the many different possible brain conformations the animal can attain ., These K reference volumes are denoted by {R1 , R2 , R3 , … , RK} ., We use 300 volumes spaced evenly through time as our reference constellations ., Each Xt is separately matched with each of the references , and each neuron in the sample , xi , t , gets a set of matches v i , t = { v i , t 1 , v i , t 2 , v i , t 3 , ., ., v i , t K } , one match for each of the K references ., This set of matches is a feature vector which we call a Neuron Registration Vector ., It describes the neuron’s location in relation to its neighbors when compared with the set of references ., This vector can be used to identify neurons across different times ., We find that 300 reference volumes creates feature vectors that are sufficiently robust to identify neurons in our recordings ., What determines the optimal number of reference volumes ?, As long as the reference volumes contain a representative sample of the space of brain conformation occupied during our recordings , the number of reference volumes needed to create a robust feature vector depends only on the size of this conformation space ., Because the conformation space of a real brain in physiological conditions is finite , there exists some number of reference volumes beyond which adding more reference volumes provides no additional information ., Crucially , the worm brain seems to explore this finite conformation space quickly relative to the time scales of our recordings ., As a result , the number of required reference volumes should not depend on recording length , at least for the minutes-long timescales that we consider here ., The neuron registration vector provides information about that neuron’s position relative to its neighbors , and how that relative position compares with many other reference volumes ., A neuron with a particular identity will match similarly to the set of reference volumes and thus that neuron will have similar neuron registration vectors over time ., Clustering similar registration vectors allows for the identification of that particular neuron across time ( Fig 4c and 4d ) ., To illustrate the motivation for clustering , consider a neuron with identity s that is found at different times in two sample constellations X1 and X2 ., When X1 and X2 have similar deformations , the neuron s from both constellations will be assigned the same set of matches when registered to the set of reference constellations , and as a result the corresponding neuron registration vectors v1 and v2 will be identical ., This is true even if the registration algorithm itself fails to correctly match neuron s in the sample to its true neuron s in the reference ., As the deformations separating X1 and X2 become larger , the distance between the feature vectors v1 and v2 also becomes larger ., This is because the two samples will be matched to different neurons in some of the reference volumes as each sample is more likely to register poorly with references that are far from it in the space of deformations ., Crucially , the reference volumes consist of instances of the animal in many different deformation states ., So while errors in registering some samples will exist for certain references , they do not persist across all references , and thus do not effect the entire feature vector ., For the biologically relevant deformations that we observe , the distance between v1 and v2 will be smaller if both are derived from neuron s than compared to the distance between v1 and v2 if they were derived from s and another neuron ., We can therefore cluster the feature vectors to produce groups that consist of the same neuron found at many different time points ., The goal of clustering is to assign each neuron at each volume to a cluster representing that neuron’s identity ., Clustering is performed on the list of neuron registration vectors from all neurons at all times , {vi , t} ., Each match in the vector , v i , t k , is represented as a binary vector of 0s with a 1 at the v i k - th position ., The size of the vector is equal to the number of neurons in Rk ., The feature vector {vi , t} is the concatenation of all of the binary vectors from all matches to the K reference constellations ., For computational efficiency , a two-step process was used to perform the clustering: First agglomerative hierarchical clustering was used on the neurons from an initial subset of volumes to define the clusters ., Next , neurons from all volumes at all times were assigned to the nearest cluster as defined by correlation distance to the clusters’ center of mass ., Assignments were made in such a way so as to ensure that a given cluster is assigned to at most one neuron per volume ., Details of this clustering approach are described in the methods ., Each cluster is given a label {S1 , S2 , S3 , …} which uniquely identifies a single neuron over time , and each neuron at each time xi , t is given an identifier si , t corresponding to the cluster to which that neuron-time belongs ., Neurons that are not classified into one of these clusters are removed because they are likely artifactual or represent a neuron that is segmented too poorly for inclusion ., Neuron Registration Vector Encoding successfully identifies segmented neurons consistently across time ., A transient segmentation error , however , would necessarily lead to missing or misidentified neurons ., To identify and correct for missing and misidentified neurons , we check each neuron’s locations and fill in missing neurons using a consensus comparison and interpolation in a TPS deformed space ., For each neuron identifier s and time t⋆ , we use all other point-sets , {Xt} to guess what that neuron’s location might be ., This is done by finding the TPS transformation , ut→t⋆: Xt ↦ Xt⋆ , that maps the identified points from Xt to the corresponding points in Xt⋆ excluding the point s ., Since the correspondences between neurons has already been determined , ut→t⋆ can be found by solving for the parameters from the TPS equation ( see methods ) ., The position estimate is then given by ut→t⋆ xi , t with i selected such that si , t = s ., This results in a set of points representing the set of predicted locations of the neuron at time t⋆ as inferred from the other volumes ., When a neuron identifier is missing for a given time , the position of that neuron s is inferred by consensus ., Namely , correct location is deemed to be the centroid of the set of inferred locations weighted by the underlying image intensity ., This weighted centroid is also used if the current identified location of the neuron s has a distance greater than 3 standard deviations away from the centroid of the set of locations inferred from the other volumes , implying that an error may have occurred in that neuron’s classification ., This is shown in Fig 5 , where neuron 111 is correctly identified in volume 735 , but the the label for neuron 111 is incorrectly located in volume 736 ., In that case the weighted centroid from consensus voting was used ., To assess the accuracy of the Neuron Registration Vector Encoding pipeline , we applied our automated tracking system to a 4 minute recording of whole brain activity in a moving C . elegans that had previously been hand annotated and published 10 ., A custom Matlab GUI was used for manually identifying and tracking neurons ., Nine researchers collectively annotated 70 neurons from each of the 1519 volumes in the 4 minute video ., This is much less than the 181 neurons predicted to be found in the head 26 ., The discrepancy is likely caused by a combination of imaging conditions and human nature ., The short exposure time of our recordings makes it hard to resolve dim neurons , and the relatively long recordings tend to cause photobleaching which make the neurons even dimmer ., Additionally , human researchers naturally tend to select only those neurons that are brightest and are most unambiguous for annotation , and tend to skip dim neurons or those neurons that are most densely clustered ., We compared human annotations to our automated analysis in this same dataset ., We performed the entire pipeline including detecting centerlines , worm straightening , segmentation , and neuron registration vector encoding and clustering , and correction ., Automated tracking detected 119 neurons from the video compared to 70 from the human ., In each volume , we paired the automatically tracked neurons with those found by manual detection by finding the closest matches in the unstraightened coordinate system ., A neuron was perfectly tracked if it matched with the same manual neuron at all times ., Tracking errors were flagged when a neuron matched with a manual neuron that was different than the one it matched with most often ., The locations of the detected neurons are shown in Fig 6A ., Only one neuron was incorrectly identified for more than 5% of the time volumes ( Fig 6B ) ., The locations of neurons and the corresponding error rates are shown in Fig 6B ., Neurons that were detected by the algorithm but not annotated manually are shown in gray ., Upon further inspection , it was noted that some of the mismatches between our method and the manual annotation were due to human errors in the manual annotation , meaning the algorithm is able to correct humans on some occasions ., GCaMP6s fluorescent intensity is ultimately the measurement of interest and this can be easily extracted from the tracks of the neuron locations across time ., The pixels within an approximate 2 μm radius sphere around each point are used to calculate the average fluorescent intensity of a neuron in both the red RFP and green GCaMP6s channels at each time ., This encompasses regions of the cell body , but excludes the neuron’s processes ., The pixels within this sphere of interest are identified in the straightened RFP volume , but the intensity values are found by looking-up corresponding pixels in the unstraightened coordinate system in the original red- and green-channel images , respectively ., We use the calcium-insensitive RFP signal to account for noise sources common to both the GCaMP6s and the RFP channel 10 ., These include , for example , apparent changes in intensity due to focus , motion blur , changes in local fluorophore density arising from brain deformation and apparent changes in intensity due to inhomogeneities in substrate material ., We measure neural activity as a fold change over baseline of the ratio of GCaMP6s to RFP intensity ,, Activity = Δ R R 0 = R - R 0 R 0 , R = I GCaMP 6 s I RFP ., ( 3 ) The baseline for each neuron , R0 , is defined as the 20th percentile value of the ratio R for that neuron ., Fig 7 shows calcium imaging traces extracted from new whole-brain recordings using the registration vector pipeline ., 156 neurons were tracked for approximately 8 minutes as the worm moves ., Many neurons show clear correlation with reversal behaviors in the worm ., The Neuron Registration Vector Encoding method presented here is able to process longer recordings and locate more neurons with less human input compared to previous examples of whole-brain imaging in freely moving C . elegans 10 ., Fully automated image processing means that we are no longer limited by the human labor required for manual annotation ., In new recordings presented here , we are able to observe 156 of the expected 181 neurons , much larger than the approximately 80 observed in previous work from our lab and others 10 , 11 ., By automating tracking and segmentation , this relieves one of the major bottlenecks to analyzing longer recordings ., The neuron registration vector encoding algorithm primarily relies on the local coherence of the motion of the neurons ., It permits large deformations of the worm’s centerline so long as deformations around the centerline remain modest ., Crucially , the algorithm’s time-independent approach allows it to tolerate large motion between consecutive time-volumes ., These properties make it well suited for our neural recordings of C . elegans and we suspect that our approach would be applicable to tracking neurons in moving and deforming brains from other organisms as well ., Certain classes of recordings , however , would not be well suited for Neuron Registration Vector Encoding and Clustering ., The approach will fail when the local coherence of neuron motion breaks down ., For example , if one neuron were to completely swap locations with another neuron relative to its surroundings , registration would not detect the switch and our method would fail ., In this case a time-dependent tracking approach may perform better ., In addition , proper clustering of the feature vectors requires the animal’s brain to explore a contiguous region of deformation space ., | Introduction, Results, Discussion, Methods | Advances in optical neuroimaging techniques now allow neural activity to be recorded with cellular resolution in awake and behaving animals ., Brain motion in these recordings pose a unique challenge ., The location of individual neurons must be tracked in 3D over time to accurately extract single neuron activity traces ., Recordings from small invertebrates like C . elegans are especially challenging because they undergo very large brain motion and deformation during animal movement ., Here we present an automated computer vision pipeline to reliably track populations of neurons with single neuron resolution in the brain of a freely moving C . elegans undergoing large motion and deformation ., 3D volumetric fluorescent images of the animal’s brain are straightened , aligned and registered , and the locations of neurons in the images are found via segmentation ., Each neuron is then assigned an identity using a new time-independent machine-learning approach we call Neuron Registration Vector Encoding ., In this approach , non-rigid point-set registration is used to match each segmented neuron in each volume with a set of reference volumes taken from throughout the recording ., The way each neuron matches with the references defines a feature vector which is clustered to assign an identity to each neuron in each volume ., Finally , thin-plate spline interpolation is used to correct errors in segmentation and check consistency of assigned identities ., The Neuron Registration Vector Encoding approach proposed here is uniquely well suited for tracking neurons in brains undergoing large deformations ., When applied to whole-brain calcium imaging recordings in freely moving C . elegans , this analysis pipeline located 156 neurons for the duration of an 8 minute recording and consistently found more neurons more quickly than manual or semi-automated approaches . | Computer algorithms for identifying and tracking neurons in images of a brain have struggled to keep pace with rapid advances in neuroimaging ., In small transparent organism like the nematode C . elegans , it is now possible to record neural activity from all of the neurons in the animal’s head with single-cell resolution as it crawls ., A critical challenge is to identify and track each individual neuron as the brain moves and bends ., Previous methods required large amounts of manual human annotation ., In this work , we present a fully automated algorithm for neuron segmentation and tracking in freely behaving C . elegans ., Our approach uses non-rigid point-set registration to construct feature vectors describing the location of each neuron relative to other neurons and other volumes in the recording ., Then we cluster feature vectors in a time-independent fashion to track neurons through time ., This new approach works very well when compared to a human . | fluorescence imaging, invertebrates, classical mechanics, caenorhabditis, neuroscience, animals, animal models, vector construction, caenorhabditis elegans, model organisms, microscopy, experimental organism systems, damage mechanics, dna construction, molecular biology techniques, neuroimaging, research and analysis methods, imaging techniques, animal cells, deformation, molecular biology, molecular biology assays and analysis techniques, physics, gene expression and vector techniques, calcium imaging, cellular neuroscience, transmission electron microscopy, dark field imaging, cell biology, neurons, electron microscopy, nematoda, biology and life sciences, cellular types, physical sciences, organisms | null |
journal.ppat.0030050 | 2,007 | Transcriptional Regulation of Chemical Diversity in Aspergillus fumigatus by LaeA | Aspergillus fumigatus is a saprophytic filamentous fungus with no known sexual stage ., Prolific production of asexual spores ( conidia ) and nearly ubiquitous distribution in the environment ensures constant host exposure to its spores , at a density of 1 to 100 conidia/m−3 1 ., The innate immune system enables spores to be eliminated from lung epithelial tissue with ease in immunocompetent vertebrates ., However , immunocompromised individuals are at risk for pulmonary disease as a consequence of A . fumigatus infection ., Of particular concern is invasive aspergillosis , which occurs when hyphal growth proliferates throughout pulmonary or other tissues ., Invasive aspergillosis has an associated mortality rate ranging from 50% to 90% depending on the patient population 2 ., As the number of immunocompromised patients has increased in recent decades due to immunosuppressive chemotherapy treatments , HIV/AIDS , and solid organ and bone marrow transplantation , the incidence of invasive aspergillosis has increased more than 4-fold in developed nations 2 ., Several A . fumigatus secondary metabolites or natural products ( e . g . , conidial melanins and mycotoxins ) have been implicated as affecting virulence 3–7 ., However , the exact mechanisms by which many of these compounds might affect disease outcome are unknown , nor is it clear in most cases whether these factors play direct or indirect roles in pathogenicity ., In contrast to most genes involved in primary metabolism , genes encoding secondary metabolite biosynthetic enzymes exist in contiguous clusters within the genome 8 , 9 ., LaeA was originally identified as a transcriptional regulator of secondary metabolite gene clusters in Aspergillus nidulans and A . fumigatus 10 , 11 , including gliotoxin in the latter ., Gliotoxin has long been suggested to be a major virulence attribute in invasive aspergillosis 12–14 ., However , whereas a ΔlaeA mutant shows reduced virulence in a mouse model of invasive aspergillosis 11 , inactivation of gliotoxin biosynthesis alone does not 15–17 ., Therefore , we reasoned that because LaeA is a transcriptional regulator , perhaps acting at a chromatin remodeling level 9 , 18 , a microarray experiment comparing the transcriptomes of ΔlaeA , wild-type , and complemented ΔlaeA control strains would yield further insight into LaeA-mediated A . fumigatus virulence attributes ., We uncovered an unprecedented view of LaeA global regulation of mycotoxin islands , nearly all found in nonsyntenic regions of the Aspergillus genome ., Because secondary metabolite gene cluster regions are evolutionarily diverse and may affect virulence attributes , LaeA is a novel target for comprehensive modification of chemical diversity ., Transcriptional profiles of the wild-type , ΔlaeA , and ΔlaeA complemented strain were determined by comparisons of relative transcript levels between ( 1 ) ΔlaeA versus wild-type and ( 2 ) wild-type versus complemented control strain ., All strains were grown under identical conditions ( 25 °C , liquid shaking culture , glucose minimal media , 60 h ) for three biological replicates ., The condition and time point were chosen on the basis of optimal production of secondary metabolites 10 , 11 ., The comparison of ΔlaeA versus wild-type was used to determine gene expression patterns specific to the ΔlaeA mutant , while the wild-type versus complemented strain comparison was conducted as a control , because the difference between these two strains is the presence of an ectopic copy of a selectable marker for hygromycin resistance ., The processed signal intensity ratios for the three ΔlaeA versus wild-type replicates were analyzed using the significance analysis of microarrays ( SAM ) method 19 , as described in Materials and Methods ., In total , 943 genes were significantly differentially expressed ., Figure S1 shows a heat map of a subset of these loci , depicting normalized expression ratios for the three ΔlaeA versus wild-type experiments and the three wild-type versus complemented control experiments ., The high quality of the data is indicated by the consistency of color between the replications and the relative lack of color in the control lanes ., Of the 943 genes showing significant differences in expression between ΔlaeA and wild-type by SAM analysis , 415 showed increased expression in ΔlaeA and 528 showed decreased expression ., Table 1 and Figure S2 indicate functional categories for these genes ( defined as described by the Gene Ontology Consortium , http://www . geneontology . org ) ., The most remarkable discovery was the near-global suppression of secondary metabolite gene expression in the ΔlaeA mutant ., Nearly all ( 97% ) of the secondary metabolite gene cluster loci showed decreased expression in ΔlaeA , with a mere three genes in this category showing increased expression in ΔlaeA ., This was in contrast to all other functional categories , which showed substantial proportions of both increased and decreased expression in the mutant , possibly reflecting indirect effects due to loss of production of multiple metabolites ., In addition to genes with unknown function ( 39% ) and genes involved in secondary metabolism ( 11% ) , other major categories included genes encoding proteins involved in transmembrane transport ( 8% ) and those involved in information processing ( 4% ) , and cell wall biogenesis ( 4% ) ., Statistical analysis of the overrepresentation of different Gene Ontology categories and Pfam protein domains within the set of 943 differentially regulated genes is shown in Tables S1 and S2 , respectively ., Interestingly , LaeA appeared to influence expression of a subset of species- and lineage-specific genes not strongly conserved with other fungal species ., Only 18% and 44% of all genes significantly differentially expressed in the mutant have putative orthologs in Saccharomyces cerevisiae and Neurospora crassa , respectively , compared to an average of 33% and 58% of all A . fumigatus genes ., Many , but not all , of these genes were classified as secondary metabolism genes ., Moreover , there are about 120 differentially expressed genes; again , most , but not all , are present in secondary metabolism clusters ( Table 2 ) , which have no detectable orthologs in Aspergillus oryzae and A . nidulans ., Considering this overwhelming tight and directed transcriptional control of secondary metabolite loci by LaeA , below we focus on such genes as possible members of the LaeA-regulated A . fumigatus pathogenicity arsenal ., Although initial genome analysis suggested the presence of 26 secondary metabolite gene clusters 20 , subsequent analysis ( G . Turner , N . D . Fedorova , V . Joardar , J . R . Wortman , and W . C . Nierman , unpublished data ) has provided support for only 22 clusters ., Of the 13 secondary metabolite gene clusters whose expression was influenced by LaeA in the condition used for microarray analysis , ten are particularly strongly affected , with a majority of genes within these clusters being significantly down-regulated in ΔlaeA as indicated by SAM ., Three additional clusters have at least one gene encoding a critical enzyme such as a nonribosomal peptide synthetase ( NRPS ) or a polyketide synthase showing decreased expression in ΔlaeA ., Additionally , 38% ( 23 of 71 ) of all P450 monooxygenases show differential expression in ΔlaeA , also associated with secondary metabolite biosynthesis and/or detoxification ., Fifteen of these genes encoding P450 monooxygenases are found in secondary metabolite gene clusters ., Table S3 gives normalized expression ratio values for all 22 gene clusters in A . fumigatus ., Table 2 summarizes the current state of knowledge regarding function of LaeA-regulated secondary metabolite gene clusters ., These include clusters dedicated to production of conidial melanins , fumitremorigens , gliotoxin , and ergot alkaloids such as festuclavine , elymoclavine , and fumigaclavines A , B , and C ( Table, 2 ) 4 , 17 , 21–27 ., Figure 1 depicts the chromosomal landscape of those regions most strongly regulated by LaeA ., To confirm these microarray results , quantitative real-time reverse-transcription ( RT ) -PCR ( QRT-PCR ) was performed on one major class of secondary metabolite genes , those encoding NRPSs 28 ., As indicated in Table 3 , relative expression levels for NRPSs that showed differential expression between mutant and wild-type in the microarray study also were dramatically reduced upon QRT-PCR analysis ., In all cases , complementation of the laeA defect restored NRPS gene expression to wild-type levels ( Table 3 ) ., Notably , because the microarray analysis determines only relative expression and not absolute levels of transcript , we could not conclude whether secondary metabolite clusters not showing differential expression in mutant versus wild-type were not affected by LaeA or were simply not induced under the growth condition used ., To further examine these possibilities , we assessed the expression of a subset of NRPSs thought to encode siderophore-biosynthesizing enzymes ., Although siderophores do not fit neatly into a definition of secondary metabolites , which are dispensable in laboratory growth conditions 29 , these molecules are produced from clustered genes and are critical for pathogen growth in blood serum 30 ., Because iron was included in the media used for the microarray study , we investigated whether the Δlae mutant was deficient in expression of siderophore gene cluster NRPSs under iron-limiting conditions ., As previously reported 29 , low iron conditions induced transcriptional upregulation of several NRPS genes known or predicted to be involved in siderophore biosynthesis ( Table 4 ) ., Normalized expression levels of the siderophore NRPSs in the low iron condition relative to high iron conditions were NRPS2/sidC , 2 . 245 ± 0 . 449; NRPS3/sidE , 68 . 595 ± 13 . 725; and NRPS4/nps6/sidD , 28 . 509 ± 4 . 704 ., NRPS7 transcripts were not detectable in these experiments ., In contrast to Reiber et al . 29 , NRPS3/sidE showed the highest induction to the low iron conditions in our experiments ., This discrepancy might be explained by differential sensitivity of the semiquantitative RT-PCR method used by Reiber et al . compared to our QRT-PCR methodology or subtle differences in culture conditions ., We also noted that the complemented control strain with an ectopic copy of laeA showed increased expression of NRPS3/sidE in both low and high iron conditions ., Comparison of the ΔlaeA mutant and controls by QRT-PCR analysis indicated differential expression of NRPS3/sidE in low ( inducing ) iron conditions ., In high iron conditions , NRPS4/nps6/sidD , NRPS3/sidE , and possibly NRPS2/sidC showed decreased expression in the ΔlaeA mutant ( Table 4 ) ., Interestingly , NRPS7 was not detectable in these experiments ., In the low iron condition , expression of actin did decrease in the ΔlaeA mutant ., However , the dramatic decrease in expression of NRPS3/sidE ( 1 , 000-fold less ) seen in the ΔlaeA background strongly suggests that LaeA regulates the expression of at least this NRPS ., Little is known about the function of SidE , although it has been speculated to be involved in siderophore biosynthesis on the basis of homology to SidC 29 ., It remains to be determined whether NRPS3/sidE is involved in siderophore production , a process known to be critical to virulence 31 , 32 , or whether it synthesizes an iron-responsive compound with a distinct function ., Regardless of the function of SidE , these experiments show that LaeA is also involved in controlling expression of other secondary metabolite clusters not induced by the environmental conditions used in the microarray experiments ., Cluster 18 ( Figure, 1 ) on Chromosome 6 , strongly differentially expressed in ΔlaeA , encodes the genes required for gliotoxin biosynthesis ., Gliotoxin is arguably the most well-studied mycotoxin produced by A . fumigatus ., First identified in 1936 , this compound has immunosuppressive properties in vitro 12 and in vivo 13 , 14 , although its direct contribution to pathogenicity is only beginning to be understood 15–17 ., Like all other compounds in the epipolythiodioxopiperazine class , gliotoxin is a cyclic dipeptide with an internal disulfide bridge that can undergo redox cycling ( for a recent review , see 33 ) ., Immunosuppressive activity of gliotoxin is due at least in part to negative regulation of the transcription factor nuclear factor–κB , which occurs by inhibition of proteasome-mediated degradation of the nuclear factor–κB inhibitor IκBα 34 , 35 ., Gliotoxin is also known to be cytotoxic and can evoke both apoptotic 36–39 and necrotic 40 , 41 cell death ., Recently , gliotoxin was shown to trigger the release of apoptogenic factors by the host mitochondrial protein Bak 42 ., The secondary metabolism gene cluster responsible for gliotoxin production was recently identified by bioinformatic analysis 43 and has been experimentally confirmed 15 , 17 ., Despite the known immunosuppressive activities of the molecule and its detection in blood serum of patients with invasive aspergillosis 44 , three recent studies using genetic mutants of the gliotoxin gene cluster demonstrated that gliotoxin is not a virulence factor in murine models of invasive aspergillosis 15–17 ., However , these same studies presented evidence that gliotoxin could adversely affect T cells , neutrophils , and mast cells and , we offer , likely acts synergistically with other LaeA-regulated toxins ., The ΔlaeA mutant is impaired in gliotoxin production during growth in culture as well as growth in vivo in murine models of invasive aspergillosis 10 , 15 , and the microarray results presented here confirm that LaeA strongly influences expression of genes in this cluster under the condition investigated ., Secondary metabolite cluster 1 on Chromosome 1 , which is differentially expressed in ΔlaeA , contains an atypical NRPS called Afpes1 that is required for virulence in an insect model of invasive aspergillosis 4 ., Afpes1 shows greatest homology to NRPSs that produce siderophores or destruxins , including one paralog required for virulence of the plant pathogen Alternaria brassicae 45 ., However , the Afpes1 cluster is thought to be unlikely to produce either of these compounds , because destruxin toxin has not been detected in A . fumigatus 4 and expression of Afpes1 was not responsive to iron levels 4 , 21 ., Deletion of Afpes1 alters conidial morphology and hydrophobicity as well as melanin synthesis and results in increased susceptibility to reactive oxygen species , implying altered conidial melanin and/or rodlet composition 4 ., Most of these characteristics are common to the ΔlaeA phenotype 11 , possibly implicating a role of the Afpes1 metabolite in the attenuated virulence of ΔlaeA ., A . fumigatus synthesizes several clavine ergot alkaloids , compounds that can be partial agonists or antagonists of serotonin , dopamine , and α-adrenalin receptors , thus affecting nervous , circulatory , reproductive , and immune system function 46 ., The role of these compounds in invasive aspergillosis has not been determined ., In addition to having the receptor-modulating activities mentioned , the festuclavine ergot alkaloid produced by A . fumigatus is cytostatic and is directly mutagenic in the Ames assay 47 , 48 ., Recently , Coyle and Panaccione 25 showed that deletion of an A . fumigatus dimethylalleletryptophan synthase ( DMAT synthase ) homologous to dmaW of the ergot-producing species Claviceps purpurea eliminated all known ergot alkaloids , confirming its predicted function in the first committed step of ergot alkaloid production ( i . e . , addition of dimethylallyl diphosphate to l-tryptophan to result in 4-methylallyl-tryptophan ) ., The biochemical activity of the A . fumgiatus DmaW enzyme was also confirmed by Unsöld and Li 22 , who subsequently characterized a reverse prenyltransferase in the same gene cluster that converts fumigaclavine A to fumigaclavine C 23 ., These genes are located in secondary metabolite gene cluster 4 on Chromosome 2 , which is strongly differentially expressed in ΔlaeA ., Melanins found in conidia are one of the few described virulence factors in A . fumigatus 5 , 6 , 24 ., Lack of melanins leads to increased susceptibility to reactive oxygen species produced by the host innate immune response during infection as well as altered ( smooth ) conidial morphology 5 , 7 ., However , the scarcity of nonpigmented A . fumigatus spores in nature has drawn into question the clinical relevance of melanins as virulence factors 1 ., Conidia of ΔlaeA are pigmented , but altered expression of alb1 in the mutant has been reported previously and at least one unidentified spore metabolite is missing in ΔlaeA 11 ., There is significant differential expression of the 1 , 8-dihydroxynapthalene–melanin gene cluster in ΔlaeA under the condition investigated in this study ., Expression of this gene cluster is also regulated by cAMP/protein kinase A signaling 49 as is LaeA itself 10 , perhaps a suggestion that in this case LaeA control of this cluster may be both directly and indirectly mediated by protein kinase A . Additionally , a LaeA-regulated supercluster on Chromosome 8 is likely to produce multiple compounds ., Recently , two genes in this cluster have been reported to encode biosynthetic enzymes for the tremorgenic mycotoxin fumitremorgin B and related compounds 26 , 27 ., The cyclo-l-Trp-l-Pro derivative fumitremorgin B is cytotoxic , inhibiting cell cycle progression at G2/M , and thus has been of interest as a potential anticancer agent ., The pathway involves generation of the cyclic dipeptide brevianamide F by the NRPS brevianamide synthetase 27 , prenylation of brevianamide F by the prenyltransferase FtmPT1 to tryptrostatin B 26 , and subsequent conversion in several steps to fumitremorgen B . Thus , LaeA-mediated influence on expression of ftmPT1 and ftmPT2 would govern the production of this entire class of diketopiperozine compounds ., Once again , however , the specific effects of these compounds on pathogenicity during invasive aspergillosis are unknown ., The fact that LaeA promotes expression not only of these secondary metabolite gene clusters but an additional eight others confirms its role as a master controller of secondary metabolism ., The importance of several of these compounds in toxicity studies also underscores relevance of LaeA during infection 11 ., We suggest the possibility that virulence attributes are not influenced as much by individual metabolites as by the blend of LaeA-regulated toxins , which , in combination , may confer an advantage to the pathogen ., Comparative genomic analysis between A . fumigatus and related species indicates overlap between A . fumigatus–specific genes and genes differentially expressed in ΔlaeA ( N . D . Fedorova and W . C . Nierman , unpublished data ) ., In total , 68% of A . fumigatus secondary metabolite genes do not have orthologs in the closely related species A . clavatus ( N . D . Fedorova and W . C . Nierman , unpublished data ) ., Additional secondary metabolite genes do not have orthologs in more distantly related Aspergilli such as A . oryzae and A . nidulans 20 ., The variability of secondary metabolite clusters may be explained by the fact that many of them are located in highly divergent telomere-proximal regions characterized by frequent chromosomal rearrangements 20 , 50 ., For example , 54% of the clusters showing differential expression in ΔlaeA in the conditions described here were found within 300 kb of telomeres ., It should be noted that , in addition to the secondary metabolite clusters , other genes with significantly lower expression in ΔlaeA also show some positional specificity within the genome but to a much lesser extent ( unpublished data ) ., Further analysis also showed that A . fumigatus telomere-proximal clusters tend to have larger numbers of genes than clusters located closer to the centromeres , suggesting that the former may accumulate additional genes more easily ( N . D . Fedorova , J . R . Wortman , and W . C . Nierman , unpublished data ) ., Initial comparative genome analyses indicate that the telomere-proximal regions ( and to a lesser extent , synteny breakpoints and intrasyntenic regions ) appear to be a hotbed of diversity , not only between Aspergillus species but even between different strains of the same species 51 , 52 ., The genomes of two A . fumigatus strains have been sequenced: the clinical isolate Af293 ( by The Institute for Genomic Research , Rockville , Maryland , United States ) and isolate CEA10 ( under contract from Elitra Pharmaceutical by Celera Genomics and made available by Merck; B . Jiang and W . C . Nierman , personal communication ) ., These strains show an overall divergence of 2% , and the majority of this variation is in telomere-proximal and synteny breakpoint regions ., Similarly , microarray experiments also supported high divergence in these regions when Af293 was compared to the unsequenced A . fumigatus strains Af294 and Af71 20 ., Many secondary metabolite clusters appear to be associated with transposons ( known to be active in A . fumigatus 51 ) and transposase-like sequences ( Table S3 ) ., The finding that these transposable elements often flank or are embedded in many of the clusters may represent one mechanism for generating the diversity of secondary metabolites in aspergilli ., Whether or not there is a connection between LaeA function and transposon activity has yet to be established ., In total , these analyses suggest that secondary metabolite clusters are located in the regions that undergo extensive rearrangements , which may result in subsequent alterations in secondary metabolite production and , therefore , have major impacts on niche adaptation between different species of fungi or between strains of the same species ., Other examples include a non–aflatoxin-producing clade of A . flavus , better known as the food-fermenting A . oryzae used in the production of traditional Asian products such as miso and soy sauce , which may have arisen as a result of telomere-proximal rearrangements 53 ., Similarly , genotypic variability between strains of Fusarium compactum also proved to be a major determinant of metabolite production and geographic distribution 54 ., In Fusarium graminearum , the major cause of wheat and barley head blight , intraspecific polymorphic variations in a trichothecene mycotoxin gene cluster were correlated with chemotype differences , host range , and fitness 55 ., In light of such examples , it is interesting to speculate about the role of LaeA in chemotype evolution and niche adaptation ., It is possible that variation at any particular secondary metabolite gene cluster could result in less efficient control by LaeA ., This potential has been demonstrated in A . nidulans 18 ., Conversely , LaeA itself is a major target for comprehensive changes in the entire complement of secondary metabolites ., The clustering of secondary metabolite biosynthetic genes has been suggested to reflect their evolutionary history 8 , 9 , 20 , 51 , 56 , 57 ., Several models have been proposed to explain the establishment and maintenance of secondary metabolic gene clusters in filamentous fungi ., The “selfish cluster” hypothesis proposes that selection occurs at the level of the cluster and promotes maintenance of the cluster as a unit , possibly through horizontal transfer events 56 ., However , there is only limited evidence for widespread horizontal transfer of secondary metabolism gene clusters , with penicillin being a notable exception 58 ., Alternative models suggest that clusters are maintained due to coregulation mechanisms , likely at the level of chromatin regulation 8 , 9 ., LaeA may provide a mechanistic means of secondary metabolism gene cluster coregulation and maintenance ., Certainly LaeA demonstrates a positional bias for local gene regulation , as transfer of genes into or out of a secondary metabolite cluster leads to respective gain or loss of transcriptional regulation by LaeA 18 ., This has been speculated to occur through regulation of nucleosome positioning and heterochromatin formation 9 ., Our results confirm that LaeA plays a central role in regulation of chemical diversity in A . fumigatus ., Furthermore , genomic regions that are transcriptionally controlled by LaeA are species and even strain specific , suggesting that they may serve as niche adaptation factors ., The loss of laeA results in a great decrease in repertoire of secondary metabolites , which appears to impact the infection process ., Therefore , LaeA constitutes a novel target for the production of an array of factors critical to success during pathogenesis ., Furthermore , LaeA is a tool to identify metabolite gene clusters that may impact virulence , allowing the correlation of specific secondary metabolite clusters with virulence even in absence of knowledge about the mycotoxin itself ., Three prototrophic A . fumigatus fungal strains were used in this study ., Af293 ( the wild-type clinical isolate used in the A . fumigatus genome sequencing project 20 ) , TJW54 . 2 ( ΔlaeA ) 11 , and a complemented control strain TJW68 . 6 ( ΔlaeA + laeA ) 11 were grown in triplicate at 25 °C in liquid minimal media 59 with shaking ( 280 rpm ) for 60 h ., Profiles of secondary metabolites extracted from the media with chloroform were compared by thin-layer chromatography , and the results confirmed that the ΔlaeA strain showed reduced levels of multiple secondary metabolites under this condition ( 10 and unpublished data ) ., Total RNA was isolated from fungal mats , labeled , and hybridized with a DNA whole-genome amplicon microarray 20 , 60 in three independent biological replicates ., To analyze siderophore NRPS gene expression under low- or high-iron conditions , 50-ml liquid cultures were grown as described 29 , with low-iron media containing 25 g/L glucose , 3 . 5 g/L ( NH4 ) 2SO4 , 2 . 0 g/L KH2PO4 , 0 . 5 g/L MgSO4 ( heptahydrate ) , and 8 mg/L ZnSO4 ( heptahydrate ) ( pH 6 . 3 ) ., High-iron media was identical except for the addition of Fe ( III ) Cl3 to a final concentration of 300 μM ., Cultures were grown at 37 °C , 280 rpm , and samples were collected at 24 h postinoculation ., All glassware was subjected to sequential treatment with 1 mM and 5% HCl as described 29 ., Total RNA was extracted from Aspergillus strains by use of TriZOL reagent ( Invitrogen , http://www . invitrogen . com ) according to the manufacturers instructions ., RNA was further purified by two extractions with phenol:chloroform:isoamyl alcohol ( 25:24:1 ) and then labeled with Cy-3 or Cy-5 dye and hybridized as previously described 20 ., The generation of the whole genome array has been described 20 ., QRT-PCR was used to ( 1 ) confirm the expression level trends observed in the microarray experiment and ( 2 ) investigate NRPS gene expression under iron-limiting conditions ., Expression of select NRPSs putatively regulated by LaeA was examined ., Total RNA from two or three biological replicates was pooled in equal amounts ( 2 μg per sample ) for each Aspergillus strain , wild-type AF293 , TW54 . 2 , and TW68 . 6 , and treated with Ambion Turbo DNA-free DNase I ( Ambion , http://www . ambion . com ) to remove contaminating genomic DNA ., A total of 500 ng of DNase I–treated total RNA from each sample was reverse transcribed with Superscript III reverse transcriptase ( Invitrogen ) ., Real-time RT-PCR was conducted with 20-μl reaction volumes with the iQ SYBR green supermix ( Bio-Rad , http://www . bio-rad . com ) , 2 μl of a 1:6 dilution of first-strand cDNA , and 0 . 4 μl of each 10 μM primer stock ., Primer sequences were previously reported 28 ., No reverse transcriptase controls ( NRT ) were used to confirm elimination of contaminating genomic DNA ., Real-time RT-PCR was performed using an iQ Cycler Real-Time PCR detection system ( Bio-Rad ) ., PCRs for each NRPS were done in triplicate and melt curve analysis was performed immediately following the PCR to confirm the absence of nonspecific amplification products and primer dimers ., The relative expression levels of NRPS genes between A . fumigatus wild-type strain AF293 , the ΔlaeA mutant , and the complemented control strain were calculated using 2−ΔΔCt method with iQ cycler system software ., All values were normalized to expression of the A . fumigatus actin gene and relative to the wild-type strain for each condition analyzed ., Gene expression ratios were determined for triplicate comparisons of ( 1 ) wild-type and ΔlaeA and ( 2 ) ΔlaeA and the complemented control strain ., Prior to statistical analysis , the LOWESS normalization method was used to remove any systematic bias from the raw expression ratios 61 ., Loci showing significantly different expression were identified using the SAM method for one-class designs that has been previously described in detail 19 , implemented in the TM4 suites MultiExperiment Viewer ( http://www . tm4 . org ) 62 , 63 ., This allowed identification of genes whose mean expression across experiments is significantly different from a user-specified mean ( log2 = 0 , corresponding to identical mRNA levels in the mutant and wild-type strains ) ., Genes with scores above the significance threshold and exceeding the cutoff value of zero for the false discovery rate ( the most conservative setting ) were designated as significantly differentially expressed between mutant and wild-type ., The delta value cutoff in SAM was chosen to capture the maximum number of significant genes while maintaining the reported estimated false discovery rate at zero ., Genes down-regulated in ΔlaeA were further analyzed by the Expression Analysis Systematic Explorer ( EASE ) 64 within TM4 to identify overrepresented Gene Ontology terms and Pfam domains ., Fishers exact test probabilities and step-down Bonferroni corrected probabilities are reported from the EASE analysis to indicate which terms are overrepresented in the down-regulated gene set . | Introduction, Results/Discussion, Materials and Methods | Secondary metabolites , including toxins and melanins , have been implicated as virulence attributes in invasive aspergillosis ., Although not definitively proved , this supposition is supported by the decreased virulence of an Aspergillus fumigatus strain , ΔlaeA , that is crippled in the production of numerous secondary metabolites ., However , loss of a single LaeA-regulated toxin , gliotoxin , did not recapitulate the hypovirulent ΔlaeA pathotype , thus implicating other toxins whose production is governed by LaeA ., Toward this end , a whole-genome comparison of the transcriptional profile of wild-type , ΔlaeA , and complemented control strains showed that genes in 13 of 22 secondary metabolite gene clusters , including several A . fumigatus–specific mycotoxin clusters , were expressed at significantly lower levels in the ΔlaeA mutant ., LaeA influences the expression of at least 9 . 5% of the genome ( 943 of 9 , 626 genes in A . fumigatus ) but positively controls expression of 20% to 40% of major classes of secondary metabolite biosynthesis genes such as nonribosomal peptide synthetases ( NRPSs ) , polyketide synthases , and P450 monooxygenases ., Tight regulation of NRPS-encoding genes was highlighted by quantitative real-time reverse-transcription PCR analysis ., In addition , expression of a putative siderophore biosynthesis NRPS ( NRPS2/sidE ) was greatly reduced in the ΔlaeA mutant in comparison to controls under inducing iron-deficient conditions ., Comparative genomic analysis showed that A . fumigatus secondary metabolite gene clusters constitute evolutionarily diverse regions that may be important for niche adaptation and virulence attributes ., Our findings suggest that LaeA is a novel target for comprehensive modification of chemical diversity and pathogenicity . | Patients with suppressed immune systems due to cancer treatments , HIV/AIDS , or organ transplantation are at high risk of infection from microbes ., Some of the most deadly infections for such patients arise from a fungal pathogen , Aspergillus fumigatus ., This species , like several of its close relatives , can produce an array of small chemical compounds that influences both the infection process and its environmental niche outside of the host ., The genes dedicated to production of each compound are clustered adjacent to each other in the genome ., One protein named LaeA is a master regulator of such clustered small molecule genes , and removal of the gene encoding LaeA cripples the organisms ability to infect ., We conducted a genome-wide microarray experiment to identify small molecule gene clusters controlled by the presence of LaeA in A . fumigatus ., In doing so , we identified actively expressed gene clusters critical for small molecule production and potentially involved in disease progression ., These results also provide insight into evolutionary events shaping the organisms collection of chemical compounds . | secondary metabolism, aspergillus fumigatus, mycotoxin, fungal, immunology, microbiology, computational biology, pathogenicity, gene expression regulation, genetics and genomics | null |
journal.pgen.1003997 | 2,013 | Comprehensive Analysis of Transcriptome Variation Uncovers Known and Novel Driver Events in T-Cell Acute Lymphoblastic Leukemia | T-cell acute lymphoblastic leukemia ( T-ALL ) is an aggressive malignancy that accounts for approximately 15% of pediatric and 25% of adult ALL cases ., Despite improved outcome over the years , about 25% of children and 50% of adults still fail to respond to intensive chemotherapy protocols or relapse 1 ., Improved understanding of T-ALL biology through the identification and characterization of oncogenic lesions is expected to lead to a better prognostic classification and the development of new targeted therapeutic strategies ., T-ALL is caused by the accumulation of multiple oncogenic mutations that have been identified through characterization of chromosomal aberrations and candidate gene sequencing 2 ., Chromosomal translocations in T-ALL frequently involve the T-cell receptor ( TCR ) loci , whereby TCR regulatory elements become juxtaposed to genes that are normally not expressed in T-cells 3 , 4 ., In this way , a specific set of recurrently over-expressed transcription factors ( TFs ) have been documented , including TLX1 , TLX3 , TAL1 , LMO1 , HOXA , and NKX family members 5 ., T-ALL samples expressing each of these transcription factors show a distinctive gene expression signature and as such these transcription factors define distinct molecular subtypes in T-ALL 6 ., Chromosomal rearrangements can also lead to large chromosomal deletions and amplifications; to focal gene deletions or amplifications , such as CDKN2A deletion and MYB duplication 7 , 8; and to in-frame fusion genes encoding chimeric proteins with oncogenic properties such as the constitutively active NUP214-ABL1 fusion kinase 9 ., In addition , point mutations and small insertions/deletions ( INDELs ) have also been described leading to oncogenic events , such as mutations activating NOTCH1 that occur in more than 60% of T-ALL cases 10 , or mutations in cytokine receptors and tyrosine kinases such as IL7R and JAK3 11–17 ., The latter may lead to new opportunities for molecularly tailored therapies with kinase inhibitors 12 , 16 , 18 , 19 ., With the advent of next generation sequencing ( NGS ) technologies , our sequencing capacity has significantly improved in the past five years ., It is now possible to apply targeted re-sequencing , exome sequencing ( Exome-seq ) , whole genome sequencing ( WGS ) , whole transcriptome sequencing ( RNA-seq ) or a combination of these , to investigate individual genomes , especially those related to disease 20 ., Also for T-ALL , these NGS approaches have recently proven their value in the discovery of novel driver genes 13 , 14 , 17 , 21 ., We previously identified a spectrum of new oncogenic driver genes using Exome-seq on 67 T-ALLs , and described clear differences between pediatric and adult patients 17 ., In particular , we identified CNOT3 as a tumor suppressor mutated in 8% of adult T-ALL cases and mutations affecting the ribosomal proteins RPL5 and RPL10 in 10% of pediatric T-ALLs 17 ., Similarly , whole genome sequencing of early T-cell precursor ALL cases led to the identification of mutations in several new oncogenes and tumor suppressor genes affecting cytokine signaling , T-cell development and histone-modifying genes 2 , 13 ., However , the potential of RNA-seq for the discovery of driver genes in T-ALL remains unexplored ., In the present study , we applied paired-end RNA-seq on 49 T-ALL samples ( 31 patients , 18 cell lines ) to gain insights in the transcriptome landscape of T-ALL ., First , we show that identification of somatic single nucleotide variants ( SNV ) and recurrently mutated driver genes is feasible on RNA-seq data , even without matched normal samples ( e . g . , germlines or remission DNA ) ., We identify STAT5B , H3F3A , and PTK2B as candidate cancer genes in T-ALL ., This becomes possible when ( 1 ) optimal read mapping and SNV calling procedures are applied; and ( 2 ) functional annotation , gene expression , or additional sequencing data from other cohorts is used to prioritize the true driver genes ., Next , we optimized gene expression measurements using multiple normalization strategies , and showed that classical gene expression studies ( e . g . , clustering ) are feasible on normalized RNA-seq data ., We also detected new fusion genes ( SSBP2-FER and TPM3-JAK2 ) and used gene expression data to determine the consequence of observed chromosomal rearrangements on the over-expression of key driver genes ., Finally , we searched for significant alternative transcript events ( ATE ) but besides one coherent exon-skipping event in SUZ12 , we found relatively few candidate ATEs in T-ALL ., In conclusion , through a combination of the analysis of gene expression levels , fusion transcripts , SNVs , and INDELs , we could identify known and new driver alterations in T-ALLs and novel potential targets for therapy ., We performed paired-end RNA-seq on 31 T-ALL patients , 18 T-ALL cell lines , and 1 normal thymus sample ., We obtained on average ∼110 million reads per sample , leading to an average coverage of ∼88× ( Table S1 . A ) ., To assess the quality of detecting SNVs from the RNA-seq data , we compared the RNA-seq to Exome-seq data ., For 16/18 of the cell lines and for 20/31 patient samples we had exome data available ( previously generated 17 or obtained for this study , Table S2 ) ., For the exome data analysis , we followed the pipeline of mapping , SNV and somatic mutation detection that we validated previously 17 ( using BWA , GATK , SomaticSniper , and Variant Effect Predictor ( VEP ) ) 22–25 ., For the RNA-seq data we used TopHat2 26 for mapping , SAMTools 27 for SNV detection , and VEP 25 for variant annotation ( Figure 1 . A ) ., By comparing positions that had a coverage of at least 20× in both RNA-seq and Exome-seq , combined with Sanger re-sequencing of a subset of positions , we found that the accuracy of SNV calling in RNA-seq strongly depends on the read mapping , corroborating earlier observations 28 , 29 ( Figure S1 ) ., We found that mapping RNA-seq reads to the genome ( as used by TopHat version 1 . 3 . 3 ) is prone to errors when dealing with paralogous genes , as observed by the prediction of false positive SNVs in KIF4A and GLUD1 due to erroneous mapping to KIF4B and GLUD2 ( both pseudogenes with no introns ) ( Figure S1 ) ., However , these errors were resolved by mapping to the transcriptome ., In the case of the RPMI8402 cell line , 877 SNVs were found by mapping to the genome , while this number was reduced to 283 SNVs when mapping to the transcriptome ., Mapping to the transcriptome did not only reduce the number of RNA-seq exclusive calls but also increased the overlap with the Exome-seq calls ( Figure 2 , Figure S2 ) ., However , transcriptome mapping also has limitations as it relies on current gene and isoform annotation ., We observed that a combination of transcriptome and genome mapping provides the best solution ., It is important that all reads are mapped twice to the genome , independently of each other; once as entire read and once as split read ., This has become possible in TopHat2 by setting the option “read-realign-edit-dist” to zero ., Our analysis reveals that this mapping approach results in the best overlap of SNVs compared to exomes ( Figure 2 , Figure S3 ) ., This mapping strategy not only improves the alignment accuracy by preventing misalignment to pseudogenes , but also leads to identification of the most likely isoform structure of a gene by mapping the reads independently both to the transcriptome and to the genome and then selecting the best possible alignment ., Using the optimized mapping and filtering strategy we identified 436 , 974 SNVs across 49 samples ., By using samples for which both the exome and the transcriptome were sequenced several aspects of SNV detection in RNA-seq data can be evaluated , such as sensitivity , specificity , and allelic imbalance ., Regarding sensitivity , we found that on average , 32% of the SNVs that are called in Exome-seq were also called by the RNA-seq ( Table S3 ) ., Similar ratios were observed when comparing validated somatic SNVs from Exome-seq/WGS to RNA-seq SNVs: 36% in a triple negative breast cancer study 30 , and 41% in a lymphoma study 31 ., We observed that the sensitivity varies considerably between samples , and strongly correlates with the average depth of coverage of the sample ( Figure S4 ) ., Regarding specificity , we found that the remaining RNA-seq-only and Exome-seq-only SNVs ( for positions where both have at least 20× coverage ) are found mainly with a low variant allele frequency ( VAF ) and are therefore likely due to arbitrary VAF and coverage thresholds ., For example , on the RPMI8402 and TLE79 samples , many RNA-seq-only SNVs ( 9/18 and 61/88 respectively ) have a VAF below 40% ., Regarding allelic imbalance , we found that of all heterozygous Exome SNVs with more than 20× coverage , the majority ( 2 , 914/4 , 043 or 72% ) were also heterozygous SNVs in RNA-seq ., Of the remaining SNVs , many ( 988/4 , 043 ) are homozygous reference in the RNA-seq ( i . e . , not detected ) ., A small fraction we can almost certainly attribute to allelic imbalance , namely the 141/4 , 043 SNVs ( 3 . 5% ) that are homozygous variant in the RNA-seq , indicating that for those only the variant allele is expressed ( or the gene is only expressed in cancer cells that harbor the variant ) ., Next we asked whether small insertions and deletions ( INDELs ) can be detected from RNA-seq data ., As with the SNVs , we used the Exome-seq data for assessing the quality of our INDEL detection strategy ., On average , 47 . 5% of the INDELs that were detected by RNA-seq were also found in the Exome-seq ( unfiltered ) INDEL calls ., However , only 4% of the Exome-seq INDELs ( for which the region containing INDEL is covered by at least 3 reads in RNAseq data ) were found back in the RNA-seq calls ( Table S3 ) ., To investigate this sensitivity issue , we evaluated ten validated INDELs that we previously identified with Exome-Seq 17 ( Table S4 ) ., Three of the ten INDELs were also identified in the RNA-seq data using the default SAMTools parameters ( see Materials and Methods ) ., Of the seven missed INDELS , two are found in a gene that is not expressed; another two are clearly present in the RNA-seq data when inspected manually with IGV , but did not reach the default threshold ( see Materials and Methods ) ; and the last three are effectively discordant between RNA-seq and Exome-seq , as they show only reads with reference sequence ( Figure S5 ) ., Re-mapping of the reads with BWA 22 on the transcriptome followed by BLAT 32 on the genome improved the INDEL identification , now revealing the KDM6A INDEL in TLE87 and PTEN INDEL in TLE92 , which were previously missed ( Figure S6 . A–B ) ., It is notable that the combination of TopHat2 ( to transcriptome only ) and BLAT does not correctly detect these two INDELS ( Figure S6 . C–D ) ., We conclude that INDEL detection on RNA-seq data is feasible , yet technically challenging and that the fraction of INDELs compared to SNVs is moderate ( see also the next Section and Figure 3 ) ., Our next aim was to select candidate driver genes using the collected SNVs and INDELS ., To remove germline variants we initially removed all SNPs present in dbSNP 33 , 1000genomes 34 , the Complete Genomics genomes 35 , and those detected in our own exome data from normal samples ( 39 from our earlier work 17 and 6 from this study ) ., We , however , retained those variants also present in the COSMIC 36 database , since SNP databases are known to contain also some disease-specific SNVs ., Some examples of SNVs that are likely driver mutations , but that are also present in polymorphism databases are: JAK3 A572V in R7 , and FBXW7 R425C in TUG1 ., With this filtering , we obtained a final list of 10 , 403 protein-altering SNVs and 430 protein-altering INDELs , with a median of 63 SNVs and 4 INDELS per sample ( Table S1 . B ) ., Cell lines harbored significantly more mutations than patient samples ( Mann-Whitney test p-value\u200a=\u200a1 . 095E-05 ) , as previously also observed by Exome-seq 17 ., As a first approach to identify candidate T-ALL driver genes , we selected all genes that contained a protein-altering mutation in at least two of the 31 patient samples ( for recurrence we did not take cell lines into account ) ., This process resulted in the selection of 213 genes ( Table S5 ) ., We found that this list is strongly enriched for genes related to T-ALL and to cancer in general , with “precursor T-cell lymphoblastic leukemia-lymphoma” as the most highly enriched function ( p-value\u200a=\u200a1 . 35E-11 by Ingenuity Pathway Analysis ) ( Table S6 ) ., The list of 213 candidates contained many known T-ALL driver genes ( Figure 3 ) , such as NOTCH1 , BCL11B , FBXW7 , IL7R , JAK1 and JAK3; and it also contained the drivers CNOT3 and RPL10 , recently identified in our exome re-sequencing study 17; and CTCF , which was recently reported to be recurrently mutated in ETP-ALL 13 ., In addition , the candidate list contained two established cancer driver genes involved in other cancer types , but not yet reported to be mutated in T-ALL , namely H3F3A and CIC ., These genes were reported recently by Vogelstein 37 to be true cancer drivers ., We identified two patient samples ( TLE76 and TUG6 ) with H3F3A mutations both on the K28 residue that is a mutational hotspot in glioblastoma 38 ., This mutation was confirmed somatic in the TUG6 sample ., Sequencing of this hotspot in additional T-ALL samples indicated a low frequency of H3F3A K28 mutation in T-ALL ( detected in 3 of 102 cases ) ., Next we asked if we could identify additional genes in the candidate list that could be linked to T-ALL ., We wanted to utilize the genes that are known to be involved in T-ALL as a guide for identifying additional candidates ., To this end we used our gene prioritization approach ENDEAVOUR 39 , which scores candidate genes based on a set of training genes ., It builds a profile based on the training genes ( integrating information on protein-protein interactions , genetic interactions , gene expression , text-mining , sequence homology , Gene Ontology , and protein domains ) and then prioritizes the candidate genes for their similarity to the derived profile ., As training set we used all known drivers , and as test set we used all the 213 candidates with at least two patient mutations ( excluding the genes that are in the training set ) ., We reasoned that this would reveal the genes with strong similarity to the known drivers and such genes would be good candidate drivers ., We found 45 significantly ranked genes with two interesting genes at the top of the ranking , namely PTK2B and STAT5B that are involved in JAK/STAT signaling ( Table S7 ) ., Furthermore , the list contained genes for which we had identified single T-ALL cases with a somatic mutation in our previous exome study: ANKRD11 , CTCF , DOCK2 , H3F3A , and HADHA ., We did not select these genes before in our Exome-seq cohort 17 because they were only mutated in one of the 39 samples we analyzed ., Now , with the RNA-seq cohort , we thus found additional samples with mutations in these genes ., T-ALL is characterized by the overexpression of transcription factors ( TFs ) , such as TLX1 , TLX3 , TAL1 , and the HOXA family members 6 ., Therefore , identifying and analyzing expression perturbations in a T-ALL cohort is highly relevant ., To obtain accurate gene expression levels from the mapped RNA-seq reads , we followed the procedure outlined in Figure 1 . B , including read aggregation , GC-normalization , length normalization , and between-sample normalization ( see Materials and Methods ) ., In addition , we removed a batch effect that was clearly present in the data set using a Generalized Linear Model ( GLM , see Materials and Methods ) ( Figure S7 ) ., It is notable that transcript-based expression analysis conducted with cufflinks revealed the same batch effect linked to the origin of the sample , thereby confirming a technical bias in the data set ( Figure S7 . B , see Materials and Methods ) ., We next looked at the expression values of TLX1 , TLX3 , TAL1 , and other important TFs in T-ALL ., Clustering of TLX1 , TLX3 , and TAL1 expressing samples confirmed that the correct samples ( based on karyotyping and molecular analysis ) showed over-expression of the respective TF ( Figure 4 . A ) ., Indeed , 8 samples that harbored a STIL-TAL1 rearrangement showed high TAL1 expression ( Figure 4 . D ) ., Note that also other samples with high TAL1 expression were detected ., This fits with a previously reported observation of TAL1 over-expression in the absence of a translocation in T-ALL 6 , 40 ., To assess the accuracy of our expression values obtained after normalization , batch effect removal and clustering , we tested whether previously published gene signatures associated with TAL1 , TLX ( TLX1 and TLX3 ) and LYL1 can be detected also in our data set 41 ., We used 13 gene signatures obtained by Soulier et al using a microarray study on 92 primary T-ALL samples 41 ., Gene set enrichment analysis shows that our TAL1 expressing cases are significantly associated with TAL1 signatures , whereas our TLX over-expressing cases are associated with the TLX signature 7 , 8 and the LYL1 cases with the LYL1 signature 10 , 11 ., This analysis confirms that the obtained expression data represent meaningful values and sample clustering produces gene lists that are biologically meaningful ( Figure 4 . B ) ., We next used the gene expression information as a guide to assist in the detection of relevant mutations ., We found that the expression profile of PTK2B , a candidate driver identified above by ENDEAVOUR , significantly correlated with the JAK3 expression profile ( PTM , with p-value threshold at 1E-05 , see Materials and Methods ) ( Figure 4 . C ) ., Indeed , PTK2B was previously implicated in IL-2 mediated signaling and JAK/STAT signaling , and was shown to physically interact with JAK3 42 ., These data warrant further investigation of PTK2B as an important tyrosine kinase in T-ALL case with activated JAK/STAT signaling ., To our knowledge , only very few cancer specific alternative transcript events ( ATE ) have been described for any cancer type 43–45 , and no ATE is reported for T-ALL ., In contrast to SNVs , INDELS , copy number variations , and fusions , which are all curated and present in large numbers in public cancer mutation databases ( e . g . , COSMIC 36 , CENSUS 46 ) , we could not find driver ATEs in those databases ( although splice sites represent an important class of cancer mutations ) ., If ATEs represent an important , yet underestimated , type of somatic variation in cancer , we would expect at least some of the known cancer driver genes to present a significant ATE ., We thus asked whether novel variations could be found in these genes in the form of ATEs ., To this end , we applied cufflinks and cuffdiff ( see Materials and Methods ) and found significant ATEs in 12 of the 47 known driver genes ( BCL11B , FLT3 , IL7R , LCK , MYB , NKX2-1 , SFTA3 , RPL10 , RUNX1 , SETD2 , SUZ12 , and TAL1 ) ( Table S8 ) ., However , when we manually inspected these events in IGV , we found only two interesting cases ., One case represents an unambiguous skipping of exon 7 in SUZ12 , occurring in several patient samples , but most significant ( cuffdiff p-value\u200a=\u200a5 . 10E-05 ) in the R5 patient sample , and absent in the Thymus ( Figure 4 . E ) , and a potential , but less clear , skipping of exon 8 in LCK in three samples ( Figure S8 ) ., Exon 7 of SUZ12 is a canonical exon ( present in all known isoforms ) according to RefSeq , Ensembl , and UCSC annotation ., The ATE we observe is a heterozygous event with the wild-type junction supported by 90 reads and the novel junction supported by 71 reads ., RT-PCR clearly confirmed the exon-skipping event in R5 and to a minor extent in other samples , while being absent in the thymus ( Figure 4 . F ) ., The functional consequences of these splice variants remain to be determined , but the fact that these variants are both in-frame suggests that these proteins could be functional protein isoforms ( Figure S8 and S9 ) ., Overall , relatively few significant ATEs are detected , and no obvious ATEs are found with consequences on the protein structure , therefore T-ALL presents robust isoform usage at the current resolution of sequencing and analysis ., Most of the T-ALL cases harbor chromosomal rearrangements that lead to the generation of fusion genes or ectopic expression of genes due to juxtaposition to strong promoters or regulatory sequences ., Chromosomal translocations involving the TCR genes are largely underestimated by karyotyping and the TCR partner genes remained unidentified in several cases 4 , 47 ., On the other hand , a multitude of mechanisms other than translocations could cause ectopic expression of oncogenes 48 ., To detect fusion transcripts , we used the defuse algorithm on our entire dataset 49 ., Briefly , this method identifies candidate gene fusions by discordant alignments produced by spanning reads ( each read in the read pair aligns to a different gene ) and by split reads ( reads that harbor a fusion boundary ) ., The total number of predicted fusions initially was 1 , 160 and 1 , 265 in patient and cell line samples , respectively ., Also in normal thymus RNA , 60 fusion transcripts were detected ., Next , we implemented additional filters , considering only predictions supported by 8 or more spanning reads and 5 or more split reads ., Furthermore , we removed fusions involving ribosomal genes , mitochondrial genes and fusions between adjacent genes , as these could be caused by read-through or trans-splicing 50 , 51 ( Figure 1 . C ) ., After applying these filters , we obtained an average of 5 . 5 fusion events per patient sample and 11 . 1 per cell line ( Table S1 . C ) ., In total , 397 candidate genes are involved as potential partner in a gene fusion ( Table S9 ) ., Details on the fusion breakpoints and validation of the novel candidate fusion transcripts are reported in Tables S9 and S12 ( see also Materials and Methods: RT-PCR and Sanger Sequencing ) ., First , to determine the relevance of these predicted fusion transcripts we looked at functional enrichment of these genes ., 278 of 397 genes correspond to functionally annotated protein-coding genes according to DAVID functional enrichment 52 , 53 ., Furthermore , this set is strongly enriched for cancer-related genes , and more specifically for genes involved in Acute Myeloid Leukemia ( p-value\u200a=\u200a4 . 48E-10 ) and T-ALL ( p-value\u200a=\u200a4 . 47E-05 ) , including TP53 , STAT5B , NOTCH1 , IL7R , IKZF1 , CDKN2A , MLLT10 , ETV6 , and ABL1 ., Second , we specifically analyzed the 27 in-frame fusions , predicted to encode chimeric proteins ( Table S10 ) ., This list contained known oncogenic fusion genes , including NUP214-ABL1 ( n\u200a=\u200a2 ) , MLL-FOXO4 ( n\u200a=\u200a1 ) , PICALM-MLLT10 ( n\u200a=\u200a1 ) , ETV6-NCOA2 ( n\u200a=\u200a1 ) and SET-NUP214 ( n\u200a=\u200a1 ) ., In addition , we identified 3 novel chimeric transcripts in T-ALL , namely NUP98-PSIP1 ( n\u200a=\u200a1 ) , TPM3-JAK2 ( n\u200a=\u200a1 ) and SSBP2-FER ( n\u200a=\u200a1 ) and a novel DDX3X-MLLT10 fusion transcript ( n\u200a=\u200a1 ) recently described in a pediatric T-ALL patient 54 ., Conventional cytogenetic analysis confirmed the presence of a t ( X;10 ) in the case with the DDX3X-MLLT10 fusion , whereas it failed to detect the chromosomal rearrangements for the TPM3-JAK2 , NUP98-PSIP1 and SSBP2-FER fusions , demonstrating the power of RNA-seq to identify cryptic fusion genes and to provide genetic information even in patients with uninformative cytogenetics ., Reassuringly , RT-PCR and Sanger sequencing confirmed the presence of these fusion transcripts ( Table S12 ) ., The TPM3-JAK2 and SSBP2-FER fusions encode typical tyrosine-kinase fusions that join the tyrosine-kinase domain of JAK2 or FER to the dimerization units of TPM3 or SSBP2 , respectively ( Figure 5 . A ) ., To assess whether the TPM3-JAK2 and SSBP2-FER fusions encode oncogenic proteins , we tested their transforming properties in the IL-3–dependent Ba/F3 cell line 55 ., Both TPM3-JAK2 and SSBP2-FER transformed Ba/F3 cells to IL-3–independent growth , with even faster kinetics than the JAK1 A634D mutant , which is a known transforming kinase 18 ( Figure 5 . B ) ., Western blot analysis confirmed the constitutive auto-phosphorylation of the JAK2 and FER fusion proteins , as well as the downstream STAT proteins ( Figure 5 . C ) ., Ba/F3 cells transformed by the TPM3-JAK2 fusion were sensitive to a JAK kinase inhibitor , documenting the potential application of JAK2 kinase inhibitors for the treatment of T-ALL cases with JAK2 fusion genes ., No specific FER inhibitors were available to test their activity ., Both TPM3-JAK2 and SSBP2-FER fusion were screened in 50 additional T-ALL samples , but no additional case with these fusions was found ., Third , we also analyzed the identified fusions that did not seem to encode chimeric proteins ( out-of-frame fusions ) , and which were the majority of fusions detected in T-ALL ., These fusion events can be used as surrogate markers for the identification of chromosomal rearrangements , providing accurate information on the precise chromosomal breakpoints ., In combination with the gene expression data obtained by RNA-seq , these data can identify genes that are located close to such potential breakpoints and for which the expression is significantly up- or down-regulated ., As expected , we identified the STIL-TAL1 fusion in several T-ALL cases ( n\u200a=\u200a8 ) ., We also identified and validated 6 fusion events involving TCR genes ., In 4 of these cases , the TCR gene was found to be fused to the potential oncogene ( NOTCH1 , IL7R , PLAG1 , and TLX1 ) ., In the two other cases ( R4 , TLE90 ) , the TCR gene was fused to RIC3 or SFTA3 , resulting in the ectopic expression of LMO1 and NKX2-1 , respectively , as indicated by RNA-seq gene expression data ( Figure 5 . D and E ) ., Similarly , we could better characterize the t ( 10;14 ) in ALL-SIL cell line that expresses TLX1 at high level ., In addition to the TCR gene rearrangements , also other fusions were associated with overexpression ., We detected out-of-frame fusion transcripts that joined exon 4 of CDK6 to exon 2 of HOXA11-AS and exon 5 of CDK6 to sequences downstream of EVX1 ., In the same patient we also detected a fusion joining DPY19L1 on chromosome 7p14 to HOXA11 on chromosome 7p15 ., The gene expression analysis documented high expression of genes of the HOXA cluster ( i . e . HOXA9 , -A5 , -A13 , -A10 , -A11 ) ., Moreover , other fusions identified in this study , such as CLINT1-MEF2C , HNRP-ZNF219 ( n\u200a=\u200a2 ) , ZEB1-BMI1 and AHI1-MYB ( n\u200a=\u200a2 ) were also associated with transcriptional activation of MEF2C , ZNF219 , BMI1 and MYB as confirmed by the expression data ( Table S9 and S12 , and Figure S10 ) ., Increased MYB expression in T-ALL was previously observed as a consequence of MYB duplication ( including in the BE-13 cell line ) , which may also explain the detected AHI1-MYB fusion 8 , 56 ., Finally , we also found out-of-frame fusion transcripts leading to the potential inactivation of tumor suppressor genes , such as TP53-TBC1D3F ( ALLSIL cell line ) , PTEN-RNLS ( LOUCY cell line ) , IKZF1-ABCA13 and CDKN2A-miR31HG ( R6 case ) , indicating a third class of fusion events ( Figure S10 ) ., FISH analysis performed in the R6 case confirmed the p15/p16 deletion ., As the genes are in close proximity , the IKZF1-ABCA13 was presumably generated by deletion although no material was available to confirm this hypothesis ., The landscape of genomic variation underlying T-ALL has recently been investigated by sequencing candidate genes 14 , 21 , whole exomes 17 and whole genomes 13 ., The results of these studies , combined with a large body of gene-by-gene evidence collected over the last decade , provide a growing comprehension of the T-ALL genome ., The T-ALL genome is mainly characterized by the over-expression of TF , such as TLX1/3 and TAL1 , in combination with gain-of-function NOTCH1 mutations , and with additional mutations in chromatin modifiers , cellular signaling factors such as those involved in the JAK-STAT signaling pathway 57 , tumor suppressor genes ( TP53 , PTEN , WT1 ) , or in other genes such as ribosomal genes 17 ., Since the majority of observed mutations are point mutations and gene fusions ( much more than copy number variations 13 ) we reasoned that RNA-seq would be effective to identify many of these mutations , certainly those associated with ( over- ) expressed oncogenes ., Indeed , exome sequencing allows identifying point mutations but not gene fusions; and low coverage whole-genome sequencing allows identifying structural variation ( gene fusions ) but not point mutations ., In this study we present RNA-seq analyses on a heterogeneous group of 31 T-ALL samples and 18 T-ALL cell-lines and demonstrate that RNA-Seq is indeed a very powerful approach to detect gene mutations and fusions as well as expression perturbations ., Our first challenge with regards to the accurate identification of point mutations was finding the optimal analysis pipeline – from read mapping to SNV calling and filtering – to avoid too many false positive SNVs ., By exploiting whole-exome sequencing data for a subset of our samples we obtained a recovery ratio of 32% when compared to the exome derived SNVs; a ratio that is comparable with previous RNA-seq studies 30 , 31 ., However , this concordance could only be achieved by using the optimal read mapping methods and parameters: ( 1 ) use of a recent version of TopHat2 ( v . 2 . 0 . 5 . or higher ) and ( 2 ) forcing this aligner to map all reads twice to the genome ( once directly and once using split reads ) and once to the transcriptome ., Indeed , the computational task of sequence read mapping is more challenging for RNA-seq data because a large fraction of the obtained reads need to be split to allow reads that overlap exon-exon boundaries in the cDNA to be mapped to the genome ., In this way , RNA-seq is more prone to the identification of false SNVs due to the erroneous mapping of reads , for example to highly similar non-spliced pseudogenes ., For example , in the RPMI8402 cell line , 603 RNA-seq exclusive SNVs were found with the genome mapping strategy , while only 35 when using combined mapping strategy ., Among the previously published large scale RNA-seq cancer studies , only a handful performed variant calling on the RNA-seq data 30 , 31 , 58 , 59 ., A combined mapping strategy was followed in all cases either by mapping the reads to a customized genome reference file ( by the addition of exon junction segments ) or mapping the reads twice ( once to the genome and once to the transcriptome ) ., Variant calling pipelines also showed diversity: Morin et al and Shah et al used SNVMix 60 for variant calling , while Seo et al and Berger et al implemented filters based on alignment on the non-reference bases ., To our knowledge there is no extensive benchmarking study evaluating aligners and variant callers for RNA-seq data , but a review paper by Quinn et al compared the performance of two variant callers ( GATK 23 and SAMTools 27 ) with the optional duplicate removal step ( pre and post alignment ) , and concluded that post-alignment duplicate removal and variant calling with SAMTools achieved the best performance in terms of sensitivity and specificity 61 ., We have also followed the same strategy in our study and we could achieve a comparable recovery ratio of 32% when compared to Exome-seq calls ., A second challenge in identifying point mutations was the prioritization of candidate driver mutations versus passenger mutations ., Due to the lack of matched germline RNA for each patient as control , we used a la | Introduction, Results, Discussion, Materials and Methods | RNA-seq is a promising technology to re-sequence protein coding genes for the identification of single nucleotide variants ( SNV ) , while simultaneously obtaining information on structural variations and gene expression perturbations ., We asked whether RNA-seq is suitable for the detection of driver mutations in T-cell acute lymphoblastic leukemia ( T-ALL ) ., These leukemias are caused by a combination of gene fusions , over-expression of transcription factors and cooperative point mutations in oncogenes and tumor suppressor genes ., We analyzed 31 T-ALL patient samples and 18 T-ALL cell lines by high-coverage paired-end RNA-seq ., First , we optimized the detection of SNVs in RNA-seq data by comparing the results with exome re-sequencing data ., We identified known driver genes with recurrent protein altering variations , as well as several new candidates including H3F3A , PTK2B , and STAT5B ., Next , we determined accurate gene expression levels from the RNA-seq data through normalizations and batch effect removal , and used these to classify patients into T-ALL subtypes ., Finally , we detected gene fusions , of which several can explain the over-expression of key driver genes such as TLX1 , PLAG1 , LMO1 , or NKX2-1; and others result in novel fusion transcripts encoding activated kinases ( SSBP2-FER and TPM3-JAK2 ) or involving MLLT10 ., In conclusion , we present novel analysis pipelines for variant calling , variant filtering , and expression normalization on RNA-seq data , and successfully applied these for the detection of translocations , point mutations , INDELs , exon-skipping events , and expression perturbations in T-ALL . | The quest for somatic mutations underlying oncogenic processes is a central theme in todays cancer research ., High-throughput genomics approaches including amplicon re-sequencing , exome re-sequencing , full genome re-sequencing , and SNP arrays have contributed to cataloguing driver genes across cancer types ., Thus far transcriptome sequencing by RNA-seq has been mainly used for the detection of fusion genes , while few studies have assessed its value for the combined detection of SNPs , INDELs , fusions , gene expression changes , and alternative transcript events ., Here we apply RNA-seq to 49 T-ALL samples and perform a critical assessment of the bioinformatics pipelines and filters to identify each type of aberration ., By comparing to exome re-sequencing , and by exploiting the catalogues of known cancer drivers , we identified many known and several novel driver genes in T-ALL ., We also determined an optimal normalization strategy to obtain accurate gene expression levels and used these to identify over-expressed transcription factors that characterize different T-ALL subtypes ., Finally , by PCR , cloning , and in vitro cellular assays we uncover new fusion genes that have consequences at the level of gene expression , oncogenic chimaeras , and tumor suppressor inactivation ., In conclusion , we present the first RNA-seq data set across T-ALL patients and identify new driver events . | null | null |
journal.pcbi.1004360 | 2,015 | Optimal Prediction of Moving Sound Source Direction in the Owl | Predicting the future position of an object in the environment is a common and critical component of many tasks that involve reaching or orienting toward moving targets 1–4 ., To execute these prediction tasks successfully , motor plans must extrapolate beyond accumulated sensory input to account for delays in sensory and motor processing , as well as for the future movements of the object ., The ability to make accurate predictions of the location of a moving target is especially critical in prey capture ., Prey capture for moving targets has been studied at the behavioral and neural levels for animals that rely on visual 5–9 and auditory 10–12 information ., For example , salamanders use visual input to predict the position of moving prey , make a head orienting movement toward the target , and then generate a ballistic movement of the tongue to capture the prey 7 ., Barn owls also visually track their prey when possible 13 , but are additionally able to use auditory information to capture moving prey 10 ., After estimating a sound source’s trajectory , the owl makes a head orienting movement to localize a moving target before preparing to bring its feet forward to strike the prey 10 ., Interestingly , salamanders and barn owls have neurons with similar specialized receptive fields that shift in time to mediate predictive prey capture 12 , 6 , 8 ., These specializations occur in the fast-OFF retinal ganglion cells of the salamander 6 , 8 and the auditory spatially selective neurons in the optic tectum ( OT ) of the barn owl 12 ., The receptive fields of these neurons shift toward a moving source , where the amount of shift is sufficient to account for delays in sensory and motor processing ., Furthermore , it has been shown in the salamander that it is possible to read out the predicted location of a moving target from the fast-OFF retinal ganglion cells using a population vector average ( PV ) 8 ., Here , we use the PV to model the computations performed by the barn owl as it tracks a moving sound source and address how such a neural circuit may approach optimal performance ., These studies open several questions about the neural basis of predictive behaviors ., What information is represented in these populations of neurons ?, Is the observed neural representation an optimal solution to the prey capture problem faced by each species ?, An optimal solution to the prediction problem would take into account the source dynamics , sensory statistics , and prior information to guide the solution ., This approach to an optimal solution can be formulated as Bayesian prediction 14 ., There is support for Bayesian models of perception and behavior in diverse tasks across multiple species 15–18 ., Additionally , there have been multiple proposals for how neural systems can implement Bayesian inference 19–23 , 16 , 24–26 ., In particular , several studies have addressed the problem of inference in time in the context of hidden Markov models 20 , 27 , 28 and tracking using the Kalman filter 29 , 19 , 22 , 30 , 31 ., However , it remains unknown how a neural system can perform Bayesian prediction ., Here we specify how a population of neurons should respond to a moving stimulus to allow for a Bayesian prediction to be decoded from the neural responses ., We approach this question in the context of auditory-based prey capture by the barn owl ., The Bayesian prediction problem we consider is that of predicting a sound source’s future direction , given a sequence of sensory observations and a prior distribution for direction and angular velocity ., It has been shown that the owl’s sound localization for brief sounds is consistent with a Bayesian model 24 ., Here , evolutionary pressure for optimality may be expected , given the dependence of owls on successful sound localization during hunting ., The success of the PV in decoding predictive movements of visual targets in the salamander 8 and dragonfly 32 makes this a viable candidate mechanism for implementing Bayesian prediction ., It has been shown that the owl’s map of auditory space decoded by a PV is consistent with the owl’s localization behavior for brief stationary stimuli 24 ., More generally , it has been shown that a population code can encode the statistical properties of the environment to allow a PV to match a Bayesian estimator 23 , 26 , 24 ., This model for the neural implementation of Bayesian inference is attractive because it matches the common observation that population codes are adapted to natural statistics 33 ., However , the applicability of the PV model to Bayesian inference in time is unknown ., Here , we determine the conditions under which a population of neurons with spatial-temporal receptive fields can perform Bayesian prediction for moving sound sources ., We consider the problem of predicting the future direction of a moving source from a temporal sequence of auditory observations ., Specifically , the prey capture problem is that of predicting the direction of a moving sound source a short time in the future based on the sequence of interaural time difference ( ITD ) measurements from the sounds reaching the left and right ears ( Fig 1 ) ., ITD is the difference in the arrival time of sounds at the two ears and is a primary cue for localization in the horizontal dimension 34 , 35 ., The Bayesian filtering approach to predicting at time k the direction at a point n time steps in the future θk+n , given the sequence of observations up to time k , ITD1:k = ITD1 , ITD2 , … ITDk , is to compute an estimate from the posterior distribution pk+n ( θ , ω|ITD1:k ) ., The form of the posterior distribution is determined by a model for the dynamics of the moving target and the statistical relationship between the state of the target and the ITD observations ., The temporal dynamics of the horizontal direction of the moving target are modeled as, θk=θk−1+Δtωk−1+ηk, ωk=ωk−1+νk, where θk is the target direction , ωk is the angular velocity , ηk is a zero-mean circular Gaussian noise process , νk is a zero-mean Gaussian noise process , k is the current time step , and Δt is the time step duration ., The sensory information ITD is modeled as a sinusoidal function of direction that is corrupted by noise:, ITDk=Asin ( 2πfθk ) +ξk, where ξk is a zero-mean Gaussian noise process with standard deviation 12 . 5 μs and the amplitude A and frequency f are determined by the shape of the owl’s head and facial ruff 24 ., The sinusoidal mapping between direction and ITD is based on direct measurements of ITD for the barn owl 36 ., All noise processes are assumed to be mutually independent and uncorrelated across time ., The noise process νk influencing the prey velocity depends on the type of behavior displayed by the prey ., A large standard deviation of the noise corresponds to irregular fleeing behavior displayed by prey under close attack when there is no place to hide 37 ., A small standard deviation produces a smoother trajectory for the prey , which corresponds to escape toward cover 37 ., Here we use a velocity-noise standard deviation of 0 . 125 deg/s corresponding to mouse escape behavior under close-distance owl attack where prey trajectories are smooth 37 ., This parameter value has the effect of keeping the velocity roughly constant over a short period of time ., The prior depends on both the natural prey behavior and the owl’s bias as determined by the behavioral cost function 24 ., Here we assume that the prior emphasizes directions at the center of gaze 24 and slow source velocities ., We also assume that there is a weak negative correlation between direction and velocity such that there is a bias for sources moving into the center of gaze 38 , 39 ., The form of the prior is a Gaussian with zero mean for both direction and velocity ., The standard deviation for direction is 23 . 3 deg 24 , the standard deviation for velocity is 50 deg/s , and the correlation between direction and velocity is -0 . 05 ., The parameter values for the velocity standard deviation in the prior ( σv0=50 deg/s ) and during movement ( σvk=0 . 125 deg/s , k ≥ 1 ) describe a situation where the initial velocity can take on a wide range of values , but the velocity will be roughly constant over a short period of time ., The Bayesian prediction at time k of the direction at a point n steps in the future , θk+n given the sequence of observations ITD1:k is computed as the mean of the posterior n steps in the future pk+n ( θ , ω|ITD1:k ) Because we are estimating a circular variable , the Bayesian prediction is the direction of the Bayesian prediction vector , defined as the vector that points in the direction of the mean value of the direction n steps in the future:, BVk=∫u ( θ ) pk+n ( θ|ITD1:k ) dθ ,, where u ( θ ) is a unit vector pointing in direction θ ( Methods ) ., Solving the Bayesian prediction equations may be computationally difficult for nonlinear or non-Gaussian models 40 ., If the system is linear with Gaussian noise , then the Kalman filter can be used for Bayesian prediction 41 ., Our model includes Gaussian noise but the mapping from direction to ITD is nonlinear ., We found that relationship between direction and ITD is nearly linear for sound sources in the frontal hemisphere ( Fig 2 ) ., The root-mean-square ( RMS ) error between the measured ITD and the linear approximation ITD = 2 . 67 μs / deg × θ was 15 . 1 μs for directions between -100 deg and 100 deg ., We therefore used the Kalman filter to perform Bayesian prediction for computational simplicity ( Methods ) ., The Bayesian model successfully predicts future directions of the prey for smoothly moving sources ( Fig 3 ) ., We chose the prediction time step n in order to predict the source direction 100 ms in the future 12 ., Initially the Bayesian prediction is dominated by the prior distribution , which emphasizes central directions ( Fig 3A ) ., Because of the influence of the prior , the posterior does not initially lead the source direction ., However , after a short delay the posterior pk+n ( θ , ω|ITD1:k ) predicts the future direction of the source ( Fig 3A and 3B ) ., Note that the performance of the Bayesian prediction differs from the Bayesian tracking estimate ., Whereas the tracking algorithm seeks to place the center of posterior at the current source position ( Fig 3C ) , the prediction algorithm seeks to place the center of posterior at the future position of the source ., Also , the predictive posterior ( Fig 3A ) is wider than the posterior for tracking ( Fig 3C ) because uncertainty increases as the time window for prediction increases beyond the current time where observations are available ., It has been shown that the owl’s map of auditory space decoded by a PV is consistent with the owl’s localization behavior for brief stationary sounds 24 ., Here we investigate conditions on a population of neurons with spatial-temporal receptive fields under which the PV will match the Bayesian prediction in time ., The PV at time k is given by an average of weighted preferred direction vectors:, PVk=1N∑j=1Na ( ITD1:k|θ ( j ) , ω ( j ) ) u ( θ ( j ) ), where the preferred directions θ ( j ) are defined by the motor output ., The PV at time k depends on the sequence of past ITD measurements and predicts the future direction of the target ., By associating each neuron with a fixed preferred direction θ ( j ) , we are making the assumption that the motor neurons that the OT neurons ultimately influence are fixed ., This assumption means that the effect of a given level of response for an OT neuron on the motor output stays constant ., The rate function a ( ITD1:k|θ ( j ) , ω ( j ) ) is the firing rate of the jth neuron in response to the sequence of ITD values ITD1:k ., We now state our main result , which specifies sufficient conditions so that the PV will approximate the Bayesian prediction estimate ., The first prediction derived from our result is that neurons implementing Bayesian prediction using this type of population code will have receptive fields that shift in time towards the moving source ( Fig 4A–4D ) ., This is the type of shift that is necessary to compensate for delays and allow for the owl to capture the moving source 6 , 12 , 8 ., These delays include signal processing in the brain as well as motor delays , and total approximately 100 ms 12 ., While the receptive fields shift in time , there is a delay to the onset of the shift of the receptive field ., This delay in the shift occurs in the Bayesian model because the response is initially dominated by the prior before sufficient sensory information has been accumulated ., Therefore , the predictive posterior initially lags behind the source direction ( Fig 3A ) ., It is only after a delay that the predictive posterior leads the current source direction ., The model also predicts that receptive fields get sharper with time ., The sharpening of the receptive fields follows the sharpening of the posterior as more sensory information is collected ( Fig 3A ) ., Additionally , the model predicts that the shift of the receptive field depends on the speed of the moving source ., Faster source velocities lead to larger shifts , while slower source velocities correspond to smaller shifts of the receptive field ( Fig 5A ) ., This prediction follows from the fact that the posterior shifts faster for faster sources ., The receptive field shifts predicted by the model are consistent with experimental results in the barn owl 12 ( Figs 4 and 5 ) ., Neurons in the owl’s OT that are involved in generating head orienting movements show shifting receptive fields for moving sources 12 ( Fig 4E and 4F ) ., The receptive field shifts in the owl are consistent with the Bayesian prediction model in that the shift toward the source is not instantaneous , but occurs after a delay ( Fig 4E and 4F ) ., Receptive fields of midbrain neurons also get sharper in time , as predicted by the model 12 , 43 ., Additionally , the size of the shift varies with the speed of the moving source ( Fig 5B ) ., The time course and magnitude of the observed shifts correspond well to the predicted shifts in the model ., The model predicts an asymmetry in the shifts of the receptive fields for sounds moving into and out of the center of gaze that increases with the eccentricity of the receptive field ( Fig 6 ) ., For neurons with receptive fields at the center of gaze , the shifts for clockwise and counterclockwise sources are mirror images ( Fig 6A–6C ) ., For neurons with more peripheral receptive fields , the shifts for clockwise and counterclockwise moving sources are asymmetric ( Fig 6D–6I ) ., For neurons with peripheral receptive fields , the initial shift of the receptive field for sources moving into the center of gaze is in the opposite direction than one would expect based on the idea that receptive fields should move towards the source ., This occurs because of the effect of the prior on the performance of the posterior ( Fig 3A ) ., Initially , the posterior is dominated by the prior and thus at stimulus onset is not leading the source by the desired 100 ms . The asymmetry of the receptive field shifts for peripheral OT neurons has not been investigated in the owl ., However , neurons in the owl’s external nucleus of the inferior colliculus ( ICx ) do have an asymmetry in their direction selectivity for sounds moving into and out of the center of gaze , which may be related to asymmetric shifts 38 , 39 ., Testing this prediction will require further study ., The prediction of asymmetry in the receptive field shift for clockwise moving and counterclockwise moving sources depends on the presence of a prior that emphasizes central directions ., We found that predicted receptive field shifts were symmetric for clockwise moving and counterclockwise moving sources in both central and peripheral neurons when the prior in the model was uniform ( Fig 7 ) ., As noted above , the asymmetry is caused by the initial dominance of the prior on the location of the peak in the posterior ., When the prior is uniform , this effect is removed and the posterior can quickly lead the source direction for motions both into and out of the center of gaze ., The receptive field shifts predicted by the model were robust to parameter variation ( Fig 8 ) ., We examined the receptive field shifts for different standard deviations of the noise terms and different prior standard deviations for direction and velocity ., The model predicted similar magnitudes of shifts for the chosen values ( center column ) and when each parameter was halved ( left column ) or doubled ( right column ) ., Changing the standard deviation of the noise corrupting ITD had the greatest effect on the receptive fields ( Fig 8A–8C ) ., This parameter influences the width of the posterior and therefore influences the width of the receptive field ., The net effect of the receptive field shifts is that the activity moves across the population so that it predicts the future direction of the moving source ( Fig 9A ) ., It is this activity that must be decoded by the PV to approximate the Bayesian prediction ., To test the PV implementation of Bayesian prediction , we constructed a model of 5000 Poisson neurons with receptive fields that shift according to the posterior ( Methods ) ., The PV matched the Bayesian prediction closely for different stimulus conditions ( Fig 9 ) ., The PV approximated the Bayesian prediction to within 3 degrees ( root-mean-square ( RMS ) error ) for velocities up to 125 deg/s ( Fig 9B ) ., The RMS error in the approximation of the Bayesian prediction by the PV depended strongly on the fraction of time the predicted source direction was in the frontal hemisphere ( spearman rank correlation = 0 . 92; Fig 9C ) ., Since all of the preferred directions of the model neurons are in the frontal hemisphere , the model will necessarily fail when the posterior is localized at source directions behind the head ., We also computed the RMS error using a population of deterministic neurons to determine the contribution of the Poisson variability of the neurons to the error ( Fig 9D ) ., The Poisson variability increased the RMS error for many trajectories ( mean ± s . d . ratio of RMS error for deterministic neurons to RMS error for Poisson neurons 0 . 43 ± 0 . 23 ) ., However , the largest errors in the approximation are primarily due to the limited range of preferred directions of neurons in the population ., The pattern of error as the initial direction and velocity of a moving source varied is explained by larger errors occurring when the predicted source trajectory spends more time behind the head ., We showed that the PV can read out the Bayesian prediction in time from a population of neurons ., The PV will approximate the Bayesian prediction when the population has specialized responses with shifting receptive fields ., The types of shifting receptive fields predicted by our analysis are observed in the OT of the owl 12 and the retina of the rabbit 6 and salamander 6 , 8 ., This result shows that with the appropriate encoding of the stimulus , a simple decoding algorithm can perform complex computations 44 , 19 , 8 ., Our work provides a theoretical framework in which to interpret observations about circuits underlying prediction ., Previous work identified neurons in the OT 12 and retina 6 , 8 with shifting receptive fields that account for delays in neural processing ., Leonardo and Meister ( 2013 ) further showed that decoding a population of such responses with a PV can predict a moving target position ., Our work shows that this type of network computation can be optimal and capture the statistics of a dynamic target ., This work shows that a non-uniform population code model with a PV decoder can implement Bayesian inference for stationary and moving sources ., The non-uniform population code model proposes that a prior distribution is encoded in the distribution of preferred stimuli and that the statistics of the sensory input are encoded by the pattern of neural responses across the population 23 , 24 ., Here we extend this model to show that the dynamics of a population code can represent the statistics of a dynamical system ., This is an important extension of the non-uniform population code model due to the dynamic nature of ethologically relevant stimuli ., We make several predictions about the receptive field shifts necessary for optimal prediction ., First , we predict that neurons have receptive fields that shift towards a moving source where the shift increases with the source velocity ., This prediction is consistent with observations in the OT 12 and retina 6 , 8 ., We also predict that the shift is sluggish when a non-uniform prior is present ., This is consistent with responses of OT neurons 12 ., Our analysis also leads to several predictions that have not been tested in the auditory or visual systems ., In particular , we predict an asymmetry in the shifts of receptive fields for sources moving into and out of the center of gaze when a prior emphasizes the center of gaze ( Fig 6 ) ., We also predict that for noisier stimuli , the magnitude of the shift will decrease and the receptive fields will become wider ( Fig 8A–8C ) ., Finally , we predict that receptive fields should become narrower over time to reflect the accumulation of sensory information ., Studies of neurons thought to support predictive behaviors have not yet investigated all of these response features predicted by our model ., Bayesian theories of perception propose that neural systems represent statistical models of the environment , where the models may contain many parameters ., The parameters of these models may be learned by an animal over multiple time scales ., For the owl , information about the prior and the basic relationship between sound localization cues and source directions is primarily due to a combination of genetic changes over an evolutionary time scale and learning over the life of the animal 45 ., There is evidence , however , that the owl adjusts to the noise level of the stimulus on a trial-to-trial basis 46 ., We therefore predict that the noise-level parameter of the model is learned rapidly , leading to wider and more slowly shifting receptive fields in high noise environments ., Future work is required to determine how the parameters of the model are learned in the owl’s auditory system ., Previous studies have shown that a cascade model with a gain control component can produce the experimentally observed shifting receptive fields 6 , 12 , 8 ., This model involves a negative feedback loop , causing the neural response at each time step to be influenced by its predecessors ., This model is phenomenological , but it suggests that a recurrent network within the OT is sufficient to generate the receptive field shifts necessary for Bayesian prediction ., However , neurons upstream from OT in ICx show direction selectivity 39 , 47 and it is therefore possible that shifting receptive fields originate in ICx ., Furthermore , the asymmetric direction selectivity observed in ICx may possibly be explained by single-cell adaptation 39 rather than by a network effect ., Therefore , the mechanism underlying receptive field shifts in OT remains an open question ., Previous work has addressed inference in time using the Kalman filter 19 , 22 , 30 , 48 ., While we determine how a population of neurons should respond to a moving stimulus but did not specify a mechanism for implementing the responses , these studies constructed networks to represent the Kalman filter estimate and variance as a function of time ., One type of model produces a population code where the estimate of the target location is at the peak of a symmetric population response 22 , 30 ., This is accomplished through a nonlinear encoding model involving divisive normalization ., It is possible to read out the estimate using a center-of-mass decoder , but the model is limited to Gaussian distributions ., Another model encodes the target estimate and variance using a linear probabilistic population code 48 ., This model also relies on divisive normalization to implement the Kalman filter , but requires a nonlinear decoder to determine the estimated location from the activities ., The model of Eliasmith and Anderson ( 2003 ) utilizes nonlinear responses and linear decoding ., However , unlike the preferred direction vectors in the PV , the linear decoders are not in general equal to the preferred directions and are obtained using supervised learning ., These models may be extended to consider the case of prediction , but the responses of neurons performing prediction in these schemes has not been investigated ., Our model differs from the previous models in that the preferred direction at the peak of the population activity profile will not in general equal the PV estimate ( Fig 9 ) ., This occurs because our model includes a non-uniform population , whereas previous models use a uniform population ., An additional distinction between our model and previous models is that our predictions apply to nonlinear and non-Gaussian models ., It has previously been shown that the PV performs poorly when decoding arm movements from motor cortical responses 49 ., The work presented here does not conflict with this previous finding ., We show that the PV will perform well in tracking and prediction when the receptive fields of the neurons encode the state dynamics with shifting receptive fields ., This is not a general-purpose decoder , but rather must be used to read out the activity of a specialized population with shifting receptive fields such as those in the OT ., Experimental evidence suggests that populations of neurons with response properties that are adapted to the natural statistics are important for perception and behavior ., The work presented here shows how network properties tailored to the dynamics of moving prey allow for optimal Bayesian prediction by a population of neurons ., The Bayesian prediction at time k of the direction at a point n steps in the future θk+n , given the sequence of observations ITD1:k is computed from the posterior n steps in the future pk+n ( θ , ω|ITD1:k ) ., To construct the posterior at time k+n we first compute the posterior at the current time step pk ( θ , ω|ITD1:k ) , then predict n steps in the future using the transition probability density pk+n|k ( θk+n , ωk+n|θk , ωk ) ., Using the dependence relationships between direction , velocity , and ITD indicated in Fig 1 , the posterior at time k+n is given by, pk+n ( θ , ω|ITD1:k ) =∬pk+n|k ( θ , ω|θk , ωk ) pk ( θk , ωk|ITD1:k ) dθkdωk ., The Bayesian prediction of the direction of the sound source at time k+n conditioned on the observations ITD1:k is the mean of the predictive posterior over direction pk+n ( θ|ITD1:k ) ., This posterior is found by marginalizing pk+n ( θ , ω|ITD1:k ) over the angular velocity ω ., Because we are estimating a circular variable , the Bayesian prediction is the direction of the Bayesian prediction vector , defined as the vector that points in the direction of the mean value of direction n steps in the future:, BVk=∫u ( θ ) pk+n ( θ|ITD1:k ) dθ ,, where u ( θ ) is a unit vector pointing in direction θ ., We used the Kalman filter to compute the Bayesian prediction for simulations where the linear approximation to the relationship between direction and ITD was valid ., The Kalman filter computes the mean and covariance of the posterior when the system is linear with Gaussian noise 41 ., Given that the relationship between azimuth and ITD is nearly linear for the frontal hemisphere , a linear model is a reasonable approximation to our system ., The dynamical system for the moving source can be described as:, xk=Axk−1+ςk, where the state vector consists of the direction and angular velocity xk=θkωk , the matrix A=1Δt01 describes the state dynamics , and the noise vector contains the noise for direction and velocity ςk=ηkνk ., The noise at time k ≥ 1 is Gaussian with zero mean and covariance matrix Q and is uncorrelated across time ., The output of the system is a linear approximation to the mapping from direction to ITD plus noise:, ITDk=Cxk+ξk, where C = 2 . 67 0 and ξk is a Gaussian noise process with zero mean and variance R that is referred to as the observation error ., The Kalman filter is used to compute the mean and covariance of the posterior at each time ., Define x^i|j and Σi|j to be the mean and covariance , respectively , of the posterior at time i given observations up to time j ., The mean of the posterior distribution is computed recursively through a process of prediction and updating ., The prediction one step ahead in time is computed as, x^k|k−1=Ax^k−1|k−1, Σk|k−1=AΣk−1|k−1AT+Q ., Updating the estimate with a new observation is computed as, x^k|k=x^k|k−1+LkITDk−Cx^k|k−1and, Σk|k= ( I−LkC ) Σk|k−1, where the Kalman gain is, Lk=Σk|k−1CTCΣk|k−1CT+R−1 ., When an estimate has been made for the state x^k|k , it is possible to use that estimate as a basis for predicting future states at time k+n ., This requires the estimate at time k to be multiplied by the state transition matrix n times:, x^k+n|k=Anx^k|k ., The covariance of the posterior at time k+n is computed as, Σk+n|k=∑m=1nAm−1Q ( Am−1 ) T+AnΣk|k ( An ) T ., We used a particle filter to compute the Bayesian prediction for simulations where the linear approximation to the relationship between direction and ITD was not valid ., Particle filtering algorithms are sampling-based approaches to approximating the posterior distribution that are valid for nonlinear and non-Gaussian models 40 ., The particle filter algorithm we used was adapted from 49 ., The algorithm is given by the following steps: The neural population model consists of 5000 Poisson neurons with receptive fields that shift according to the prediction given in the proposition proved in the results ., The preferred directions of the neurons were drawn from the prior Gaussian distribution with mean zero and standard deviation 23 . 3 deg ., These preferred directions match the model of Fischer and Peña ( 2011 ) ., To generate the neural responses to a sequence of ITD inputs we first computed the predictive posterior pk+n ( θ , ω|ITD1:k ) as described above ., We then used our main result specifying that the activities are proportional to the ratio of the posterior and prior to generate the spiking probabilities for the population of neurons ., We scaled the ratio of the posterior to the prior so that firing rates would be approximately 10 spikes/s for neurons with peak responses ., Spike counts were generated for the population at each time step using independent Poisson neurons with the specified rate ., The direction of the PV was used to estimate the predicted source direction at each time ., The PV was tested for counterclockwise source trajectories with initial directions covering -180 deg to 180 deg in 10 deg steps and angular velocities ranging from 0 deg/s to 150 deg/s in 25 deg/s steps ., We calculated the RMS error between the PV estimate θPVA ( t ) and the Bayesian prediction θBayes ( t ) to quantify the approximation error where, RMS=1T∫0T ( θBayes ( t ) −θPVA ( t ) ) 2dt . | Introduction, Results, Discussion, Methods | Capturing nature’s statistical structure in behavioral responses is at the core of the ability to function adaptively in the environment ., Bayesian statistical inference describes how sensory and prior information can be combined optimally to guide behavior ., An outstanding open question of how neural coding supports Bayesian inference includes how sensory cues are optimally integrated over time ., Here we address what neural response properties allow a neural system to perform Bayesian prediction , i . e . , predicting where a source will be in the near future given sensory information and prior assumptions ., The work here shows that the population vector decoder will perform Bayesian prediction when the receptive fields of the neurons encode the target dynamics with shifting receptive fields ., We test the model using the system that underlies sound localization in barn owls ., Neurons in the owl’s midbrain show shifting receptive fields for moving sources that are consistent with the predictions of the model ., We predict that neural populations can be specialized to represent the statistics of dynamic stimuli to allow for a vector read-out of Bayes-optimal predictions . | Many behaviors require predictive movements ., Predictive movements are especially important in prey capture where a predator must predict the future location of moving prey ., How sensory information is transformed to motor commands for predictive behaviors is an important open question ., Bayesian statistical inference provides a framework to define optimal prediction and Bayesian models of the brain have received experimental support ., However , it remains unclear how neural systems can perform optimal prediction in time ., Here we use a theoretical approach to specify how a population of neurons should respond to a moving stimulus to allow for a Bayesian prediction to be decoded from the neural responses ., This provides a novel theoretical framework that predicts properties of neural responses that are observed in auditory and visual systems of multiple species . | null | null |
journal.pcbi.1001007 | 2,010 | Using Sequence-Specific Chemical and Structural Properties of DNA to Predict Transcription Factor Binding Sites | An important step in characterizing the genetic regulatory network of a cell is to identify the DNA binding sites recognized by each transcription factor ( TF ) protein encoded in the genome ., A TF typically activates and/or represses genes by associating with specific DNA sequences ., Although other factors , such as metabolite binding partners and protein-protein interactions ( for example , between a TF and RNA polymerase or a second TF ) , can affect gene expression 1 , it is important to identify the sequences directly recognized by TFs to the best of our ability to understand which genes are controlled by which TFs ., A better understanding of gene regulation , which plays a central role in cellular responses to environmental changes , is a key to manipulating cellular behavior for a variety of useful purposes , as in metabolic engineering applications 2 ., A number of computational methods have been developed for predicting TF binding sites given a set of known binding sites 3–10 ., Commonly used methods involve the definition of a consensus sequence or the construction of a position-specific weight matrix ( PWM ) , where DNA binding sites are represented as letter sequences from the alphabet {A , T , C , G} ., More sophisticated approaches further constrain the set of potential binding sites for a given TF by considering , in addition to PWMs , the contribution of each nucleotide to the free energy of protein binding 3 and additional biologically relevant information , such as nucleotide correlation between different positions of a sequence 8 or sequence-specific binding energies 6 ., Perhaps not as widely used as sequence analysis , the idea of employing structural data for predicting TF binding sites has been considered 11–15 ., Most of these methods use protein-DNA structures rather than DNA by itself ., Acquiring training sets large enough to be useful is problematic for even well-studied TFs , for which only small sets of known binding sites ( on the order of 10 sites ) are typically available 8 ., New high-throughput technologies have been used to identify large numbers of binding sites for particular TFs 16–18 , but there remains a need for methods that predict TF binding sites given a small number of positive examples ., Such methods can be used , for example , to complement analysis of high-throughput data ., Binding sites detected by high-throughput in vitro methods , such as protein-binding microarrays 16 , can be compared with predicted binding sites to prioritize studies aimed at confirming the importance of sites in regulating gene expression in vivo ., The fine three-dimensional ( 3D ) structure of DNA is sequence dependent and TF-DNA interactions depend on various physicochemical parameters , such as contacts between nucleotides and amino acid residues and base pair geometry 19 ., These parameters are not accounted for by conventional methods for predicting TF binding sites , which rely on sequence information alone ., Letter representations of DNA sequences do not capture the biophysics underlying TF-DNA interactions ., Given that a TF does not read off letters from a DNA sequence , but interacts with a particular sequence because of its chemical and structural features , we hypothesized that better predictions of TF binding sites might be generated by explicitly accounting for these features in an algorithm for predicting TF binding sites ., The mechanisms by which TFs recognize DNA sequences can be divided into two classes: indirect readout and direct readout 19 ., For indirect readout , a TF recognizes a DNA sequence via the conformation of the sequence , which is determined by the local geometry of base pair steps , the distortion flexibility of the DNA sequence , and ( water-mediated ) protein-DNA interactions 20 , 21 ., For direct readout , a TF recognizes a DNA sequence through direct contacts between specific bases of the sequence and amino acid residues of the TF 22 , 23 ., These two classes of recognition mechanisms are not mutually exclusive ., In this study , we introduce a method , SiteSleuth , for predicting TF binding sites on the basis of sequence-dependent structural and chemical features of short DNA sequences ., By using molecular dynamics ( MD ) methods to calculate these features , we can map a set of known or potential binding sites for a given TF to vectors of structural and chemical features ., We use features of positive and negative examples of TF binding sites to train a support vector machine ( SVM ) to discriminate between true and false binding sites ., Negative examples are derived from randomly selected non-coding DNA sequences ., Positive examples are taken from RegulonDB 24 , which collects information about TFs in Escherichia coli ., Classifiers for E . coli TFs developed through the SiteSleuth approach are evaluated by cross validation , and the classifier for Fis is tested against chromatin immunoprecipitation ( ChIP ) -chip assays of Fis binding sites 17 ., Combining ChIP with microarray technology , ChIP-chip assays provide information about DNA-protein binding in vivo on a genome-wide scale 25 ., We also evaluate the performance of SiteSleuth against four other computational methods: the method of Berg and von Hippel ( BvH ) 3 , MATRIX SEARCH 5 , Match 7 , and QPMEME 6 ., The BvH , MATRIX SEARCH , and Match methods rely on the PWM approach to capture TF preferences for binding sites ., The QPMEME method is similar to SiteSleuth in that it employs a learning algorithm ., In the case of Fis , we show that SiteSleuth generates significantly fewer estimated false positives and provides higher prediction accuracy than the other computational approaches ., Let us use X\u200a=\u200a{x1 , … , xN} to represent the set of training data , where xk ( k\u200a=\u200a1 , … , N ) is a real-valued n-dimensional feature vector that characterizes the kth training example and n is the number of features considered ., The features considered are described below ., Given input xk and scalar output yk\u200a=\u200a{−1 , 1} , which identifies a training example as a positive or negative example of a binding site , classifier training produces an ( n−1 ) -dimensional hyperplane in the space of features that satisfies the equation wTx+d\u200a=\u200a0 and a set of linear inequality constraints , each involving a slack variable ., The parameters w and d and the slack variables ξk ( k\u200a=\u200a1 , … , N ) are found by solving the minimization problem ( 1a ) subject to the following constraints ( 1b ) where C+ and C− are penalty parameters 27 ., These parameters are introduced to balance the contributions of negative and positive training examples to the objective function ( Eq . 1a ) , as we typically have available many more negative examples than positive examples ., The penalty parameters are determined for each TF via a grid search over ranges of C− and C+ values as part of a 3-fold cross-validation procedure for each classifier ., In 3-fold cross validation , we randomly divide the training set into three subsets of roughly equal size ., One subset is then used to test the accuracy of the classifier trained on the remaining two subsets until each subset has been used in testing ., We used the F-measure to assess accuracy ., The F-measure is the harmonic mean of precision ( p ) and recall ( r ) :Precision is the fraction of predicted binding sites that are actually binding sites and recall is the fraction of actual binding sites predicted to be binding sites:where TP , FP , and FN represent true positives , false positives and false negatives from 3-fold cross validation ., To find values of C− and C+ that maximize the F-measure , we first performed a coarse grid search over the following grid points: C−\u200a=\u200a2−5 , 2−3 , … , 215 and C+\u200a=\u200a2−5 , 2−3 , … , 215 ., We then performed fine grid searches using progressively smaller grid spacing ( 2 , 20 . 5 , 20 . 125 , … ) around the best C− and C+ values found in the coarse grid search ., Once trained , a classifier for a TF , taken to recognize binding sites of length L , is used for prediction as follows ., The classifier is used to scan an organisms genome for binding sites of length L . Given a feature vector xm for a potential binding site m , we calculate the quantity wTxm+d ., The decision function of the classifier is the sign of wTxm+d ., Thus , if the sign of this quantity is positive , then site m is predicted to be a TF binding site ., Conversely , a negative quantity indicates that m is not a binding site ., This step is repeated for all non-coding sequences in the E . coli genome of length L . The length L was chosen for each TF based on information in RegulonDB 24 ., Structural and chemical features of short DNA sequences were defined based on the predicted 3D structures of these DNA sequences , which were determined via MD simulations ., MD simulations of solvated nucleic acids have been performed for almost three decades 28 , 29 ., Simulations of DNA oligomers have been studied systematically and results have been discussed in multiple publications 30–32 ., Our approach is similar to that used in Refs ., 30–32 and is described below ., Because the available experimental data are incomplete ( i . e . , structures are unavailable for all 4-mers , at least in the Nucleic Acid Database 33 ) and available structures have been determined under various experimental conditions ( e . g . , free or bound to protein ) , we used simulated structures rather than experimentally determined structures for determining structural and chemical features ., Predicted structures were obtained for a common condition in a uniform manner ., For comparison , we implemented four other computational TF binding site prediction methods: the method of Berg and von Hippel ( BvH ) 3 , Match 7 , MATRIX SEARCH 5 , and QPMEME 6 ., These methods were implemented as described in the cited papers and , for the 54 TFs studied , a list of binding sites predicted by each method can be found online at http://cellsignaling . lanl . gov/EcoliTFs/SiteSleuth/ ., For completeness , each method is briefly presented below ., To discuss these methods we will need to first introduce a few quantities ., For a set of N DNA binding sites of a particular TF , the length of each binding site is denoted by L . The value of L is set equal to the length of binding sites reported in RegulonDB for a given TF ., In the case of Fis , we set L\u200a=\u200a21 ., We define to be the number of times base b appears in the jth position in the sequences of the binding sites , and to be the corresponding frequency ., We denote as the overall background frequency of base b ., We use S to denote a potential TF binding site of length L and we use Sj ( j\u200a=\u200a1 , … , L ) to denote the jth base of sequence S . For the BvH method , we denoted the number of occurrences of the most common base in position j of the set of binding sites by ., Using a training set of N binding sites , the BvH method calculates the score of each binding site as the summation over every position of the log-odds score of observing a base of S versus the most frequent base in the corresponding position of the sequence ., Thus , the score is given byA pseudocount of 0 . 5 is used in the formula 3 ., A cutoff threshold is defined as the mean score of the N positive training examples ., To evaluate whether a new sequence S is a binding site , the score of S is calculated based on the above formula and compared with the cutoff threshold ., If the score of sequence S is greater than the cutoff threshold , it is predicted to be a binding site ., For the Match method , a set of N training examples is used to define an information vector , which describes the conservation of the position j in a binding site from the training set:The information vector is used to evaluate whether a new sequence S is a binding site or not by calculating a score defined asand min and max are calculated using the lowest and highest nucleotide frequency in each position , respectively ., A cutoff threshold is defined as the mean score of the N positive training examples ., If the score for a new sequence S is larger than the cutoff threshold , S is predicted to be a binding site ., Using a set of N binding sites as training examples , the MATRIX SEARCH method calculates the score of each binding site S as the summation over every position of the log-odds score of observing a base in S versus the overall background frequency of that base in the corresponding position of the sequences ., Thus , the score is given byA pseudocount of 0 . 01 is used in the formula 5 ., A cutoff threshold is determined as the mean of the N scores calculated from the training data ., A new sequence S is predicted to be a binding site if its score is greater than the cutoff threshold ., The QPMEME ( Quadratic Programming Method of Energy Matrix Estimation ) method defines a weight for each base b at position j in S . The score for a sequence S is defined asThe weight is estimated via a learning algorithm that only uses positive examples ., The learning algorithm minimizes the variance subject to the constraint that the score for each known binding site is less than a predefined cutoff value ., Consistent with the Methods section of Djordjevic et al . 6 , we used −1 for the cutoff value in our implementation of QPMEME , which constrains all known binding sites to one side of a hyperplane ., Mathematically , the learning algorithm is described byfor every S in the training data set ., To make a preliminary assessment of our hypothesis that we can produce better predictions if we consider the chemical and structural features of sequence-specific DNA , we examined the features of various sequences and found that the same base in the same position in a sequence can have different chemical and structural features depending on its environment ., We illustrate this finding in Figure 3 , which shows sequence-specific DNA structures ., From the structures , one can see the context-dependent variation in the twist angle between the center two base planes ., The center base pair is the same in each structure , but the twist angle for the left structure of Figure 3A is −20 . 4° , whereas the twist angle for the right structure of Figure 3A is −4 . 3° ., Figure 3A demonstrates that different local structural features may characterize the same nucleotide at the same position in a sequence ., The feature vectors for TGG and AGA are given in Table S2 ., Similarly , Figure 3B demonstrates that different nucleotides in the same position may be characterized by the same local structural features ., The twist angles of the middle base pairs of the two structures in Figure 3B are the same , even though the base pairs are different ., These observations suggested to us that chemical and structural features may capture sequence correlations relevant for TF-DNA interactions that are not apparent from sequence data alone and encouraged us to build classifiers that separate negative and positive examples of TF binding sites based on their positions in chemical and structural feature space ., This approach , which we call the SiteSleuth method , combines DNA structure prediction , computational chemistry and machine learning ., To demonstrate the reliability of MD simulations for prediction of structural features of DNA oligomers , we calculated the propeller feature using, 1 ) available experimental structural data ( obtained from the Nucleic Acid Database 33 ) and, 2 ) predicted structures obtained via MD simulations , and we found significant correlation ( about 0 . 8 ) ., The results are shown in Figure S2 ., As described in the Methods section , binary SiteSleuth classifiers were developed to identify and predict the binding sites of 54 TFs based on TF binding sites documented in RegulonDB ., The input to a classifier is a vector of structural and chemical features generated from DNA sequences , each labeled as either a positive or negative example ., Negative examples were taken from randomly chosen non-coding sequences of the E . coli genome ., The classifiers were then used to scan both strands of non-coding sequences in the E . coli genome from 5′ to 3′ to identify potential TF binding sites ., For comparison , we also considered four other computational TF binding site prediction methods: BvH 3 , MATRIX SEARCH 5 , Match 7 , and QPMEME 6 These methods are each briefly described in the Methods section ., The accuracy of predictions of each method was evaluated through a 3-fold cross-validation procedure , described in the Methods section ., For each method , the mean cross-validation score , V , for the 54 TFs considered are listed in Table S4 and classifier accuracy is summarized in Figure 4 ., Recall that V is the fraction of positive examples predicted to be true binding sites in the cross-validation procedure ., Figure 4 is a heat map showing the cross-validation score , , produced by each of the five computational methods ., Brighter red indicates a higher cross-validation score and black represents ., A cross-validation score of indicates perfect prediction , whereas a cross-validation score of zero indicates that the method fails to predict any TF binding sites correctly ., Of the 54 TFs studied , SiteSleuth outperforms all the other methods in 28 cases , equals the next best method in 11 cases , and performs more poorly in 15 cases ., Based on the number of times a method outperformed all the other methods , SiteSleuth ( 28 ) performed better than QPMEME ( 8 ) , which performed better than MATRIX SEARCH ( 2 ) , which equaled the performance of BvH ( 2 ) , which performed better than Match ( 0 ) ., In one case , IcsR , SiteSleuth is the only method for which ., The data used to construct Figure 4 are given in Table S4 ., Interestingly , Figure 4 reveals that all methods give cross-validation scores of zero for several TFs: CysB , GcvA , OxyR , RcsAB , and Rob ., This observation suggests that methods that rely on DNA sequence information , including SiteSleuth , are insufficiently equipped to predict the binding sites for these TFs ., Some of these TFs , such as GcvA 42 , may perhaps recognize DNA indirectly via interaction with a second protein that recognizes DNA directly ., Another explanation could be that some of these TFs , such as Rob 43 , may be recognizing very short sequences ., The total number of TF binding sites predicted by each computational method is given in Table S3 ., For most TFs , QPMEME and Match both predict a large number of TF binding sites in the E . coli genome ., The BvH and MATRIX SEARCH methods predict fewer binding sites , but still more than the number of predictions generated by SiteSleuth ., In Figure 5 , we show the performance of SiteSleuth relative to that of BvH for the TFs with five or more known binding sites ., The relative performance ( RP ) score shown in Figure 5 is defined as the number of TF binding sites predicted by BvH divided by the number of TF binding sites predicted by SiteSleuth ., This score indicates how many times more TF binding sites are predicted by BvH than by SiteSleuth ., For example , BvH predicts 23 times more TF binding sites for MetJ than does SiteSleuth ., For reference , the log transformed number of TF binding sites predicted by SiteSleuth is also indicated in Figure 5 and a solid line is drawn at RP\u200a=\u200a1 ., As can be seen in Figure 5 , 41 TFs have RP>1 and 13 TFs have RP<1 ., Thus , there is a large class of TFs for which SiteSleuth predicts fewer binding sites than BvH ( RP>1 ) and , by extension , the other computational methods ., From these results alone , it is not clear whether fewer predictions are a result of fewer false positives or more false negatives ., To examine this question , we considered ChIP-chip data for Fis binding to DNA 17 , which , as shown in Figure 5 , has RP>1 ., Our findings are discussed in the next section ., As described in the Methods section , we also generated ROC curves and calculated AUC to compare classifiers ., For each of the five computational methods and for TFs in RegulonDB with 20 or more known binding sites , the AUC values are tabulated in Table S6 ., We find that SiteSleuth had the largest AUC for 60% of the TFs tested , BvH had the largest AUC for 25% of the TFs , MATRIX SEARCH had the largest AUC for 10% of the TFs tested , QPMEME had the largest AUC for 5% of the TFs tested , and Match had the largest AUC for 0% of the TFs tested ., ChIP-chip assays have identified 894 DNA sequences that bind Fis in E . coli 17 , which we used to validate the Fis binding sites predicted by each method ., Looking at SiteSleuth results for Fis , SiteSleuth predicted 129 , 150 binding sites for Fis from a positive training set of 133 binding sites published in RegulonDB ( Table S3 ) , the second largest training set available for the 54 TFs we studied ., The relative performance of SiteSleuth for Fis binding site prediction is close to one for three of the other methods under consideration ( RPBvH\u200a=\u200a1 . 56 , RPMatch\u200a=\u200a2 . 03 , RPMATRIX SEARCH\u200a=\u200a1 . 55 , and RPQPMEME\u200a=\u200a11 . 67 ) ., SiteSleuths cross-validation score for Fis ( V\u200a=\u200a0 . 33 ) is low ( Table S4 ) ., The availability of empirical data on Fis binding , including a larger number of known binding sites in RegulonDB for training , and the indirect recognition mechanisms of Fis binding to DNA 33 suggested that Fis may provide a good example to test whether SiteSleuth , which accounts for DNA structure , performs better than the other methods , despite its low cross-validation score ., Predictions of Fis binding sites from each computational method are compared to experimentally identified DNA sequences that bind Fis in E . coli in ChIP-chip assays 17 ., We assume that the sequences found in this study contain , to a first approximation , the complete set of Fis binding sites ., For each method , the approximate number of false positives was determined by subtracting the number of predictions that matched experimentally defined Fis binding sequences from the total number of predictions made by the method ., Figure 6 shows the number of false positives generated by each computational method ( black bars ) ., As can be seen , the QPMEME method produced more than 1 . 5 million estimated false positives ., Match generated approximately 261 , 000 false positives and BvH and MATRIX SEARCH both generated roughly 200 , 000 false positives ., SiteSleuth produced the fewest false positives , over 70 , 000 fewer than the next best method , a reduction of 35% in the estimated false positive rate ., In absolute terms , QPMEME predicted a binding site within 889 of the 894 experimentally defined Fis binding sequences ( 99 . 44% ) ., However , the predictions are not practically useful , since they are hidden within over 1 . 5 million estimated false positive results ., The gray bars in Figure 6 report the percentage of TF binding sites correctly predicted by the five computational methods normalized by the total number of predictions ., After normalization , QPMEME was the lowest performer for Fis ., The BvH , Match , and MATRIX SEARCH methods gave approximately equivalent results ., SiteSleuth outperformed these methods , showing a 41% improvement over MATRIX SEARCH , the next best method ., We postulated that a better TF binding site prediction method could be developed on the basis of chemical and structural features , instead of letter sequences ., To test this hypothesis , we developed the SiteSleuth method , in which potential TF binding sites are associated with DNA sequence-specific structural and chemical features ., These features are then used to build classification models for and to predict TF binding sites ., Compared to the other computational methods we tested , including the three methods that use a PWM representation of TF binding sites ( BvH , Match , and MATRIX SEARCH ) , our method provides a higher cross-validation accuracy ., For 72% of the TFs studied , SiteSleuth cross-validation accuracy is as high as or higher than any other method ( Table S4 ) ., SiteSleuth also generates 35% fewer estimated false positive results ( Figure 6 ) , and gives more accurate predictions ( 41% improvement over the next best method ) for TF binding sites ( Figure 6 ) ., In addition , the four other methods considered here each rely on the additivity assumption , which states that each nucleotide in a DNA binding site contributes to binding affinity in an independent fashion ., In the study of Benos et al . 44 , the additivity assumption was tested ., In general , the additivity assumption holds rather well as shown by ddG measurements of mutated DNA sites in several protein-DNA complexes 44 ., However , it was shown that additivity is a poor assumption for some cases 44 ., SiteSleuth does not rely on the additivity assumption , which may partially explain its better performance ., It must be noted that none of the methods for predicting TF binding sites considered here can be deemed reliable when used alone ., In Figure 6 , although SiteSleuth indeed produces the highest fraction of correct predictions , the fraction of correct predictions is still small at 0 . 4% ., Nonetheless , SiteSleuth constitutes an advance over existing methods and the approach warrants further investigation ., The chemical and structural features we have considered are crude and additional determinants of specificity and other biologically relevant features , such as amino acid side chain interaction energy with DNA , could be incorporated into the SiteSleuth approach in the future ., It may also be possible to incorporate experimental measurements of short DNA sequence properties into the SiteSleuth framework ., A mechanistic understanding of TF binding to DNA could guide the design of novel model features ., For example , a recent study of Fis showed that the shape of the DNA minor groove affects Fis-DNA binding 45 ., This property is hard to capture using only DNA letter sequences , but could be captured by defining a new feature in SiteSleuth based on the available structural data ., Presently , the features defined in SiteSleuth are unable to capture the effects of the minor groove on Fis binding , which may account for SiteSleuths poor performance in absolute terms ., The QPMEME method is similar to the SVM-based approach of SiteSleuth ., Both methods involve a quadratic programming minimization procedure with linear inequality constraints ., QPMEME maps sequences of L bases into 4L multidimensional spaces with energy terms for each dimension and constructs a hyperplane such that all positive examples are located on one side of the plane ., This quadratic optimization procedure defines a separating hyperplane by minimizing the variance of energies in an energy matrix so as to minimize the number of random sequences lying on the side of the plane that contains the positive examples ., In contrast , the separating hyperplane of an SVM divides true binding sites from nonbinding sites with maximum margin ., The distinction between random sequences , considered in QPMEME , and negative examples , considered in SiteSleuth , is important because sequences do not appear with equal probability in the E . coli genome , as is shown in Figure S1 ., SiteSleuth used negative examples directly sampled from non-coding regions of the E . coli genome ., In the report of Djordjevic et al . 6 , the QPMEME method is applied to non-ORF regions of the E . coli genome to predict binding sites for 34 TFs , including Fis ., For Fis , Table 1 of Ref ., 6 indicates that QPMEME predicts 255 Fis binding sites , compared to the 1 . 5 million found with QPMEME in our hands ( Table S3 ) ., To ensure that our implementation was correct , we applied QPMEME using the same training data set used by Djordjevic et al . 6 from DPInteract and were able to reproduce their weight matrix 6 ., For Fis , RegulonDB reports 133 binding sites , compared to only 19 reported Fis binding sites in DPInteract ., This difference in the size of the training data set ( 19 versus 133 positive examples of Fis binding sites ) may be responsible for the difference in number of predicted binding sites ( 255 vs . 1 . 5 million ) ., As can be seen by comparing the common entries in Table 1 of Ref ., 6 and in Table S3 , Fis is not an isolated example of QPMEME predicting a larger number of TF binding sites when the number of positive training examples is larger ., It is also the case for the TFs ArcA , ArgR , CRP , CytR , DnaA , FadR , FarR , Fnr , FruR , GalR , GlpR , H-NS , IHF , LexA , LRP , MetJ , NagC , NarL , OmpR , SoxS , and TyrR ., The QPMEME method may perform poorly for TFs with relatively large numbers of known binding sites because QPMEME requires that all positive examples be located on one side of a hyperplane in the space spanned by an energy matrix 6 ( see Methods section ) ., Thus , known binding sites that are outliers in this space may potentially expand the range of sequences considered to be binding sites , such that recall is maximized at the expense of precision ., We have not systematically investigated the reasons underlying our observation that QPMEME performs poorly for the TFs identified above when using positive training data from RegulonDB , as such an investigation was beyond the intended scope of our study ., In summary , how TFs selectively bind to DNA is one of the least understood aspects of TF-mediated regulation of gene expression ., An ability to better predict TF binding sites from small training data sets may advance our understanding of TF-DNA binding , and may reveal important insights into TF binding specificity , regulation and coordination of gene expression , and ultimately into gene function ., A long-standing problem has been how to identify new TF binding sites given known binding sites ., The accuracy and usefulness of computational methods for genome-wide TF binding site prediction has been limited by the inability to validate , verify , and inform these methods ., Only recently has technology matured to the point that we can assay for TF binding sites on a genome-wide scale ., This capability should allow us to critically evaluate predictions from computational methods and to develop methods that are more predictive than those currently available ., Toward this end , the work presented here provides a starting point for future investigations of how TF binding site prediction can be improved by considering the physical and chemical aspects of TF-DNA binding . | Introduction, Methods, Results, Discussion | An important step in understanding gene regulation is to identify the DNA binding sites recognized by each transcription factor ( TF ) ., Conventional approaches to prediction of TF binding sites involve the definition of consensus sequences or position-specific weight matrices and rely on statistical analysis of DNA sequences of known binding sites ., Here , we present a method called SiteSleuth in which DNA structure prediction , computational chemistry , and machine learning are applied to develop models for TF binding sites ., In this approach , binary classifiers are trained to discriminate between true and false binding sites based on the sequence-specific chemical and structural features of DNA ., These features are determined via molecular dynamics calculations in which we consider each base in different local neighborhoods ., For each of 54 TFs in Escherichia coli , for which at least five DNA binding sites are documented in RegulonDB , the TF binding sites and portions of the non-coding genome sequence are mapped to feature vectors and used in training ., According to cross-validation analysis and a comparison of computational predictions against ChIP-chip data available for the TF Fis , SiteSleuth outperforms three conventional approaches: Match , MATRIX SEARCH , and the method of Berg and von Hippel ., SiteSleuth also outperforms QPMEME , a method similar to SiteSleuth in that it involves a learning algorithm ., The main advantage of SiteSleuth is a lower false positive rate . | An important step in characterizing the genetic regulatory network of a cell is to identify the DNA binding sites recognized by each transcription factor ( TF ) protein encoded in the genome ., Current computational approaches to TF binding site prediction rely exclusively on DNA sequence analysis ., In this manuscript , we present a novel method called SiteSleuth , in which classifiers are trained to discriminate between true and false binding sites based on the sequence-specific chemical and structural features of DNA ., According to cross-validation analysis and a comparison of computational predictions against ChIP-chip data available for the TF Fis , SiteSleuth predicts fewer estimated false positives than any of four other methods considered ., A better understanding of gene regulation , which plays a central role in cellular responses to environmental changes , is a key to manipulating cellular behavior for a variety of useful purposes , as in metabolic engineering applications . | molecular biology/bioinformatics, mathematics/statistics, computational biology/sequence motif analysis, computational biology/transcriptional regulation | null |
journal.pgen.1002133 | 2,011 | Identification of a Mutation Associated with Fatal Foal Immunodeficiency Syndrome in the Fell and Dales Pony | The Fell and Dales are related sturdy pony breeds traditionally used as pack animals to carry goods over the difficult upland terrain of northern England ., Both breeds experienced near extinction during WWII , and the current populations are descended from very few animals ., It is likely that this genetic bottleneck , together with the use of prominent sires , was responsible for the emergence of a fatal Mendelian recessive disease , FIS , which currently affects up to 10% of Fell and 1% of Dales foals ( data from UK breed societies ) ., Both of these breeds are registered with the Rare Breeds Survival Trust due to their falling numbers , and important position in the UKs agricultural heritage ., FIS was first described in 1998 as a unique syndrome in which affected Fell foals develop diarrhoea , cough and fail to suckle 1 ., Despite an initial response to treatment , the infections persist and were shown to be due to a primary B-cell deficiency 2 associated with reduced antibody production , with tested immunoglobulin isotypes including IgM , IgGa , IgGb and IgG ( T ) being significantly reduced 3 ., Paradoxically , circulating T-lymphocyte numbers are normal 4 ., The reduced antibody levels in affected foals are consistent with an inability to generate an adaptive immune response , resulting in immunodeficiency once colostrum-derived immunoglobulin levels decrease at 3–6 weeks of age ., This loss of maternally derived antibodies correlates with typical onset of FIS signs at 4–6 weeks ., Concurrently , affected foals develop a non-hemolytic , non-regenerative progressive profound anemia 5 , in itself severe enough to cause death and the main marker for euthanasia decisions by vets ., As a result of FIS , foals die or are humanely destroyed between 1–3 months of age , the disease being 100% fatal ., In 2009 , this condition was reported in the Dales breed 6; it is likely that the mutation has passed between the breeds given the similarity between them and the practice of occasional interbreeding ., The clinical and pathological findings for FIS are compatible with a primary defect of genetic origin 1 , and this is supported by extensive genealogical studies 6 , 7 ., FIS has a pattern of inheritance typical of an autosomal recessive disease , and the likely founder animal , which features in both the Fell and Dales studbooks , has been traced by pedigree analysis ., Primary immunodeficiences , which include depleted levels of lymphocytes and/or immunoglobulins , have previously been reported in the horse ., The recessive defect ‘severe combined immunodeficiency’ ( SCID ) , which is found in the Arabian breed , comprises a fatal deficiency in T- and B-lymphocyte numbers and function ., The underlying lesion was found to be a 5 base-pair deletion in the gene coding DNA-dependent kinase , catalytic subunit DNA-PKCS 8 , a protein involved in V ( D ) J recombination required for adaptive immunity 9 ., Like FIS foals , SCID foals have a markedly reduced thymus and have reduced numbers of germinal centers in secondary lymphoid organs 10; unlike SCID foals however , FIS foals have apparently normal numbers of circulating T-cells 4 ., Primary aggamaglobinemia is rare in horses and comprises of a complete absence of immunoglobulin and reduced peripheral B-lymphocyte levels , with normal T-lymphocyte activity 11 ., In this respect there is a similarity to FIS , however primary aggamaglobinemia is only observed in males and is X-linked ., Furthermore , profound anemia in combination with B-lymphopenia has not previously been reported in the horse or any other species , and as such FIS appears to be a unique disease process ., Here we report the mapping and identification of the genetic lesion that causes FIS ., An initial scan using microsatellite markers identified the chromosome region responsible ., The opportune production of a SNP chip , which utilized the SNP variants generated during the sequencing of the equine genome ( http://www . broadinstitute . org/mammals/horse ) then allowed a confirmatory association scan ., This was followed by re-sequencing of the implicated region in order to identify the causal mutation ., A genome-wide microsatellite scan was performed on 41 individuals taken from five pedigrees of Fell ponies in which FIS was segregating ( Figure S1 ) , using a panel of 228 markers ( Table S1 ) ., The data were examined both for loss of heterozygosity and for linkage ( Table S2 ) ., Only one microsatellite , at 30 . 25 Mb on ECA26 , showed a significant loss of heterozygosity ( χ2\u200a=\u200a7 . 15 , P\u200a=\u200a0 . 028 ) and significant linkage ( LOD score =\u200a3 . 29 at θ\u200a=\u200a0 ) to the disease ., The location of the lesion was confirmed and refined using a genome-wide association analysis with an equine SNP array ( Illumina EquineSNP50 Infinium BeadChip ) , which contains 54 , 602 validated SNPs ., After applying quality control ( see Materials and Methods ) , data were available for 42 , 536 SNPs in 49 individuals ( 18 FIS-affected and 31 controls ) ., To consider whether there was any population stratification among the samples , a multi-dimensional scaling plot of the genome-wide identity-by-state distances was performed ( Figure S2 ) ; there was no significant difference between the affected and control samples for the first two components ( P\u200a=\u200a0 . 553 ) ., In addition , a quantile-quantile plot ( Figure S3 ) to compare the expected and observed distributions of –log10 ( P ) , obtained by a basic association test , showed that there was little evidence of inflation of the test statistics ( genomic inflation factor λ\u200a=\u200a1 . 04 ) , indeed the test statistics appear to be marginally depressed rather than inflated ., No correction was considered necessary ., Two SNPs on ECA26 showed genome-wide significance after Bonferroni correction for multiple testing ( Figure 1A ) ., These are BIEC2-692674 at 29 . 804 Mb ( Praw\u200a=\u200a2 . 88×10−7 ) and BIEC2-693138 at 32 . 19 Mb ( Praw\u200a=\u200a1 . 08×10−6 ) ., The associated region spanned 2 . 6 Mb from position 29 . 6 Mb to 32 . 2 Mb on ECA26 ( Figure 1B ) ., In a subsequent fine-mapping phase , 62 additional SNPs within the region were genotyped on 13 FIS-affected samples ., Several novel SNPs were identified ( dbSNP ss295469621-295469629 ) ., In addition , two further microsatellites were also genotyped ( Table S1 ) ., The homozygous affected haplotype was shared by these animals over a 992 kb segment ( Figure 2A ) ., According to ENSEMBL gene prediction , fourteen genes lie in this interval ( Figure 2B ) ., Five selected individual animals were re-sequenced over this critical region using sequence capture by NimbleGen arrays followed by GS FLX Titanium sequencing ( GenBank submission under study accession no . ERP000492 ) ., The five individuals comprised one affected foal ( A13 on Figure 2 ) , the two obligate carrier parents , one apparently clear animal and one obligate carrier chosen for maximal homozygosity across the region ., The FIS carrier status and familial relationships were confirmed for each individual by parentage verification ., The last animal proved particularly useful in eliminating many potential causal variants ., In total , eight verified SNPs were identified in the affected foal , narrowing the critical region to 375 , 063 bp ( ECA26: 30 , 372 , 557 – 30 , 747 , 620 bp ) ., Coverage of this critical interval was increased from 92 . 9% to 98 . 4% using Sanger sequencing; none of the remaining gaps fell within 200 bp of protein-coding sequence ., Only one variant , a SNP at 30 , 660 , 224 bp , segregated as expected for a causal recessive mutation within the five sequenced samples ., In addition there was no evidence of DNA rearrangement , duplication or insertion/deletion seen in the affected foal ( Figure S4 ) ., The segregating SNP was assessed for validity as an FIS marker in equine populations ., Subsequently all 38 available affected foals ( 37 Fell , 1 Dales ) were shown to be homozygous for the affected allele and all 21 available obligate carriers were heterozygous ., A selection of Fell and Dales samples which were submitted to the Animal Health Trust for parentage verification between 2000 and 2010 were anonymously screened for the affected allele: 82 / 214 ( 38% ) of the Fells and 16 / 87 ( 18% ) of the Dales were heterozygous for the lesion and no homozygous affecteds were discovered ., These carrier rates are consistent with the approximate observed disease prevalence of 10% in the Fell and 1% in the Dales populations ., In addition , a selection of horse breeds ( 184 individuals from 11 breeds consisting of Thoroughbred ( n\u200a=\u200a29 ) , Appaloosa ( n\u200a=\u200a8 ) , Arab ( n\u200a=\u200a21 ) , Warmblood Sport Horse ( n\u200a=\u200a17 ) , Lipizzaner ( n\u200a=\u200a2 ) , Cleveland Bay ( n\u200a=\u200a20 ) , Dartmoor Pony ( n\u200a=\u200a19 ) , Icelandic Horse ( n\u200a=\u200a8 ) , New Forest Pony ( n\u200a=\u200a20 ) , Sheltand Pony ( n\u200a=\u200a20 ) and Shire ( n\u200a=\u200a20 ) ) which were considered unlikely to have interbred with either the Fell or Dales was genotyped and all proved homozygous wild-type ., The identification of a mutation that segregates 100% with the disease has enabled a diagnostic test to be developed and offered to breeders and owners , allowing them to avoid carrier-carrier matings , and consequently drastically reduce the numbers of FIS-affected foals born each year ., A gradual reduction in the use of carrier animals will , over time , lead to a reduction in the affected allele frequency in the population , while conserving the gene pool as much as possible ., In addition , other equine breeds that may have interbred with the Fell or Dales will now be screened for FIS carriers ., The FIS-associated SNP falls within the single exon of the sodium/myo-inositol co-transporter gene ( SLC5A3 , also known as SMIT ) , which is a cell membrane transporter protein responsible for the co-transport of sodium ions and myo-inositol ., This SNP is non-synonymous , causing a P446L substitution in SLC5A3; this amino acid residue ( equivalent residue 451 in the human protein ) is conserved in all 11 placental mammals for which high-coverage sequence is now available ( selection shown in Figure 3 ) ., Similarly , this residue is conserved in other solute carrier family 5 ( SLC5 ) paralogs in the horse which share similar structural homology ( Figure 3 ) ., The crystal structure of a bacterial homolog of SLC5A1 ( sodium/glucose co-transporter 1 ) has recently been elucidated 12 and shows this member of the SLC5 family to have 14 transmembrane helices; the structural conformations adopted during transport , and the precise positions of substrate binding during transfer are now being identified 13 ., Alignment of the protein sequences of the SLC5 family suggests that P446 in equine SLC5A3 is located in a transmembrane helix which is involved in forming the substrate cavity 12 and which tilts during substrate transfer ., The two prolines at positions 445 and 446 may be required for effective substrate binding by closing the substrate binding site after the substrate is bound ., Proline residues introduce structural destabilisation into alpha helices and have long been obvious candidates for points of conformational change required for substrate binding and release 14 ., Indeed , replacement of prolines in the transmembrane helices of transport proteins has shown that some residues are profoundly important in transport , affecting either substrate affinity or substrate movement 15 ., SLC5A3 is an osmotic stress response gene , which acts to prevent dehydration caused by increased osmotic pressure in the extracellular environment ., Dehydration causes the disruption of numerous cellular functions by denaturation of intracellular molecules and damage to sub-cellular architecture 16 ., Extreme osmotic conditions are found in the kidney , although osmotic response mechanisms have also been found in numerous tissues , and in particular are critical for lymphocyte development and function 17 , 18 ., The mechanism by which the osmotic stress response is mediated in mammals is not completely understood , but involves a signaling cascade comprising Rho-type small G-proteins , p38 Mitogen-activated protein kinase ( p38MAPK ) and the transcription factor , Nuclear Factor of Activated T-cells 5 ( NFAT5 ) 19 ., NFAT5 directly stimulates the transcription of hyperosmolarity-responsive genes , of which SLC5A3 is one ., These act to counterbalance the effects of extracellular osmotic pressure by transporting small organic osmolytes , such as myo-inositol , into the cell , thereby maintaining isotonicity with respect to extra-cellular conditions 20 ., NFAT5 is the only known transcriptional activator of hyperosmolarity response genes , and was shown to be essential for normal lymphocyte proliferation and adaptive immunity 18 ., Targeted knockout of NFAT5 in mice results in late gestational lethality whereas partial loss of function leads to defects in adaptive immunity and a substantially reduced spleen and thymus 18 ., Furthermore , transgenic studies identify loss of T-cell mediated immunity as the prime deficiency ensuing from aberrant NFAT5 activity 21 , 22 ., Similarly FIS-affected foals have markedly reduced thymus and spleen with a lack of germinal centers 1 , 23 ., However , FIS disease immunopathology indicates that FIS foals have apparently normal circulating T-cell numbers with only peripheral blood B-lymphocyte numbers significantly depleted; currently , there are no data available indicating NFAT5 activity in specific B-lymphocyte functions ., Studies are now required to demonstrate the functional differences between the Pro446 and Leu446 forms of the protein ., This will be achieved by introducing this mutation into transgenic mice and assessing transport function ., Further investigation into the physiological consequences of this mutation will then also be possible ., In particular , it will be important to identify how the mutation leads to profound anemia and B-lymphopenia whilst neutrophils and T-cell numbers ( including CD4/CD8 ratio ) and function ( responses to mitogens PHA and Con A ) appear normal 4 ., Importantly , it must be investigated whether there is a defect in T-cell function that is currently undetected or whether the antigen presenting function of B-cells is so suppressed that the T-cells cannot respond ., In addition , it must be noted that the lymphoid organs in FIS foals have depleted thymus tissue and poor germinal centre development in spleen and lymph nodes , which suggests that there may be some unidentified T-cell dysfunction ., Alternatively , any T-cell defects could be due to severe inflammatory responses in these very sick foals ., SLC5A3 is not associated with any described mammalian disease , although a role in the pathogenicity of Down Syndrome is suggested 24 ., The effect of loss of SLC5A3 activity has not been comprehensively studied , however SLC5A3 knockout mice die shortly after birth due to hypoventilation 25 , probably due to failure of the peripheral nervous system 26; similarly FIS-affected foals have peripheral ganglionopathy 1 ., There is relatively little literature on SLC5A3 function in hemopoietic or immunological tissues , in any species , although a role for osmotic control in developing cells is likely ., Due to uncertainty regarding tissue distribution and function of SLC5A3 , it cannot be assumed that the profound anemia and severe loss of circulating B-lymphocytes in FIS is directly due to a functional change in SLC5A3 expression or function; formal proof that this is the case will entail functional studies ., Whilst there is no doubt that the mutation in this gene is predictive of carrier or disease status , the mechanism by which this amino acid change could lead to the two described pathologies is speculative ., It is , of course possible that the mutation site is close to another , as yet unidentified , mutation that is ultimately responsible for the severe hematological and immunological changes in homozygotes ., However , all coding sequence within the critical region has been fully investigated and this is the only variant that segregates with the disease ., We hypothesize that the phenotype seen in FIS-affected foals is either a result of partial loss or subtle alteration in SLC5A3 activity that has deleterious effects on B-lymphocyte and erythroid development but cannot discount the involvement of other genetic variants in the critical region ., Whichever is the case , further analysis of FIS is justified as this genetically determined combination of immune phenotypes has not previously been reported in any other species ., Procedures were limited to the collection of blood by jugular venipuncture or hairs pulled from the mane or tail ., Blood samples were taken as veterinary diagnostic procedures as all study animals were equine patients presenting with clinical signs suggestive of FIS or were healthy related or unrelated animals that were blood tested for anemia and/or B-lymphocyte deficiency ., Many of the Fell and Dales ponies used in this study have been described previously 6 , 7 ., Study animals were all equine patients presenting with clinical signs suggestive of FIS or were healthy related or unrelated Fell or Dales ponies that were blood tested for anemia and/or B-lymphocyte deficiency ., Several FIS foals presented subsequent to euthanasia ., Pedigree information was available for many of the Fell ponies ( Figure S1 ) , and these samples ( n\u200a=\u200a41 ) were used in the linkage and homozygosity mapping analysis ., An additional ten samples were added to these for the association study; these were isolated samples for which no pedigree information and/or samples from immediate family were available ., Any adult Fell pony was eligible as a control for the association study ., FIS diagnosis was based on breed , age of animal ( 4–8 weeks at presentation ) , and profound anemia with no other predisposing cause , and was confirmed on pathology ., Specifically , this indicated severely reduced numbers ( or absence ) of germinal centers in spleen and regional lymph nodes ., B-lymphocyte deficiency was also used for FIS diagnosis ., Many accompanying clinical signs were also reported , primarily related to opportunistic infections , but these were not considered diagnostic alone ., Blood samples were collected in EDTA collection tubes from all of the Fell pony individuals indicated in Figure S1 , and from a Dales foal and its parents 6 ., Genomic DNA was isolated from the samples using a Nucleon™ BACC Genomic DNA Extraction Kit ., A panel of 228 markers , distributed as evenly as possible over the equine genome and described in Table S1 , was used ., Two further markers , TKY1155 and TKY2012 , which were located in the implicated region , were subsequently genotyped ., The genome scan was performed in multiplexes of three markers ., Four PCR reactions , each utilising a different fluorescent dye , were pooled together post-PCR to form a panel of 12 markers for analysis ., An 18 bp tail ( 5′-TGACCGGCAGCAAAATTG-3′ ) was added to the 5′ end of the forward primer and a complementary fluorescent labelling primer was included in the PCR reaction as a means of making the reactions more efficient and to reduce costs 27 ., Amplification was performed in 6 µl volumes , using 2 . 5 pmol of reverse , 1 pmol of tailed-forward , 5 pmol of the labelled universal primer ( either 6-FAM , VIC , NED , or PET ) , 20 ng genomic DNA , 0 . 75 unit AmpliTaq Gold ( Applied Biosystems ) , 1× GeneAmp PCR buffer II ( Applied Biosystems ) , 1 . 5 mM MgCl2 , and 200 µM each dNTP ., After denaturation at 94°C for 10 min , a 30-cycle PCR of 94°C for 1 min , 55°C for 1 min , and 72°C for 1 min , followed by 8 cycles of 94°C for 1 min , 50°C for 1 min , and 72°C for 1 min was performed , followed by a final extension at 72°C for 30 min ., Genotyping analysis was performed on an ABI3100 ( Applied Biosystems ) according to the manufacturers instructions ., Genotyping data was analysed with GeneMapper version 4 . 0 ( Applied Biosystems ) ; alleles were assigned to pre-defined bins and automatically given an appropriate integer value ., Mendelian inheritance was checked ., We used 41 ponies ( 14 FIS-affected , 17 obligate carriers , 10 adults of unknown carrier status ) for which pedigree information and DNA was available , in the linkage analysis ( Figure S1 ) ., A parametric linkage analysis was carried out using SUPERLINK v . 1 . 5 28 assuming an autosomal recessive mode of inheritance ., The disease allele frequency was estimated at 0 . 1 , with 100% penetrance ., Pearsons chi2 test of independence was used to identify markers where homozygosity varied significantly between the cases and controls ., An A×2 ( where A\u200a= number of alleles at a given locus ) contingency table with A-1 degrees of freedom was used ., Expected and observed heterozygosity values were computed for cases and controls for all markers exhibiting a positive LOD score , using ARLEQUIN 29 ., Statistical significance was assessed by calculating the one-tailed probability of the chi squared distribution ., SNP genotyping on 51 genomic DNA samples was performed using standard manufacturers protocols by Cambridge Genomic Services ( University of Cambridge , UK ) ., The Illumina EquineSNP50 Infinium BeadChip , which contains 54 , 602 validated SNPs , was used; information on this array is available at http://www . illumina . com/documents/products/datasheets/datasheet_equine_snp50 . pdf ., Quality control and genotype calling were performed using GenomeStudio v . 2009 . 2 ( Illumina Inc . ) ., Samples with a call rate <95% were discarded ( n\u200a=\u200a2 ) ., We performed a basic case-control association analysis on the remaining 49 samples ( 18 affected and 31 controls ) ., Analysis was performed with the software package PLINK 30 ., SNPs with low minor allele frequency ( <0 . 02 ) or genotyping rate ( <90% ) were excluded; this left 42 , 536 SNPs for analysis ., The presence of population stratification was assessed using multi-dimensional scaling ( Figure S2 ) and quantile-quantile plots were drawn to confirm that there was no over-inflation of the test statistics ( Figure S3 ) ., A total of 62 polymorphic SNPs were studied in 13 affected individuals ., A subset of these helped to delineate recombination breakpoints and these are identified in Figure 2 ., Information regarding these SNPs can be found at http://www . broadinstitute . org/ftp/distribution/horse_snp_release/v2/equcab2 . 0_chr26_snps . xls ., PCR amplification of the target sequence containing each informative SNP was performed in 12 µl volumes containing 20 ng genomic DNA , 0 . 75 unit AmpliTaq Gold , 1× GeneAmp PCR buffer II , 1 . 5 mM MgCl2 , 200 µM each dNTP , 10 pmol of reverse and of tailed-forward primer ., A PCR program of 94°C for 10 min , followed by 30 cycles of 94°C for 1 min , 58°C for 1 min , and 72°C for 2 min , and then an extension of 72°C for 10 min was used ., The PCR products were purified ( MultiScreen PCR96 filter plates; Millipore ) before sequencing in a 6 µl volume using 0 . 5 µl of 5× BigDye Terminator v3 . 1 ( Applied Biosystems ) , 5–20 ng PCR template , 1 µl of 1× BigDye sequencing buffer and 3 . 2 pmol universal sequencing primer ( Sigma-Aldrich ) ., Templates >500 bp were also sequenced in the reverse direction ., Sequencing was performed using cycle sequencing: 96°C for 0 . 5 min , 44 cycles of 92°C for 4 s , 58°C for 4 s and 72°C for 1 . 5 min ., Purification was performed by isopropanol precipitation followed by sequencing on an ABI3100 according to the manufacturers instructions ., Sequences were viewed using STADEN 31 ., This was performed at the Centre for Genomic Research ( University of Liverpool , UK ) ., A region of 3 Mb ( ECA26: 28 , 942 , 655 – 31 , 942 , 655 Mb ) was selected for re-sequencing which encompassed the critical region ., Custom tiling 385 k NimbleGen Sequence Capture arrays ( http://www . 454 . com/products-solutions/experimental-design-options/nimblegen-sequence-capture . asp ) which covered 92 . 9% of the target were designed from the horse reference sequence using standard repeat-masking algorithms ., Five individuals were selected for re-sequencing consisting of one affected pony ( A13 in Figure 2 ) , its parents , one obligate carrier selected for maximal homozygosity over the region and one individual apparently homozygous wild-type ., Sequencing was performed using GS FLX Titanium Series chemistry and assembled using Roche Newbler software v2 . 0 . 00 ., An average 34-fold read depth was obtained ., Sequence from each of the five sequenced animals was aligned to the EquCab2 reference sequence using the Artemis Comparison Tool ( ACT ) 32 to identify possible rearrangements or insertion/deletions ( Figure S4 ) ., MySQL ( Oracle Corporation ) was used to interrogate the data ., The critical region was narrowed using heterozygous variants in the affected foal and Sanger sequencing subsequently verified these ., The narrowed critical region was then interrogated for variants that segregated as expected for a recessive mutation; putative causal variants were confirmed or disproved using Sanger sequencing . | Introduction, Results, Discussion, Materials and Methods | The Fell and Dales are rare native UK pony breeds at risk due to falling numbers , in-breeding , and inherited disease ., Specifically , the lethal Mendelian recessive disease Foal Immunodeficiency Syndrome ( FIS ) , which manifests as B-lymphocyte immunodeficiency and progressive anemia , is a substantial threat ., A significant percentage ( ∼10% ) of the Fell ponies born each year dies from FIS , compromising the long-term survival of this breed ., Moreover , the likely spread of FIS into other breeds is of major concern ., Indeed , FIS was identified in the Dales pony , a related breed , during the course of this work ., Using a stepwise approach comprising linkage and homozygosity mapping followed by haplotype analysis , we mapped the mutation using 14 FIS–affected , 17 obligate carriers , and 10 adults of unknown carrier status to a ∼1 Mb region ( 29 . 8 – 30 . 8 Mb ) on chromosome ( ECA ) 26 ., A subsequent genome-wide association study identified two SNPs on ECA26 that showed genome-wide significance after Bonferroni correction for multiple testing: BIEC2-692674 at 29 . 804 Mb and BIEC2-693138 at 32 . 19 Mb ., The associated region spanned 2 . 6 Mb from ∼29 . 6 Mb to 32 . 2 Mb on ECA26 ., Re-sequencing of this region identified a mutation in the sodium/myo-inositol cotransporter gene ( SLC5A3 ) ; this causes a P446L substitution in the protein ., This gene plays a crucial role in the regulatory response to osmotic stress that is essential in many tissues including lymphoid tissues and during early embryonic development ., We propose that the amino acid substitution we identify here alters the function of SLC5A3 , leading to erythropoiesis failure and compromise of the immune system ., FIS is of significant biological interest as it is unique and is caused by a gene not previously associated with a mammalian disease ., Having identified the associated gene , we are now able to eradicate FIS from equine populations by informed selective breeding . | Foal Immunodeficiency Syndrome ( FIS ) is a genetic disease that affects two related British pony breeds , namely the Fell and the Dales ., Foals with FIS appear to be normal at birth but within a few weeks develop evidence of infection such as diarrhoea , pneumonia , etc ., The infections are resistant to treatment , and the foals die or are euthanized before three months of age ., The foals also suffer from a severe progressive anemia ., Being a recessive condition , the disease is difficult to control without a diagnostic DNA test to identify symptom-free carrier parents ., Within the last few years the horse genome has been sequenced , and this has allowed the development of tools to identify genetic mutations in the horse at high resolution ., In this article we demonstrate the use of these new tools to identify the location of the FIS mutation ., The presumptive causal lesion was then identified by sequencing this region ., This has enabled us to develop a test that can be used to identify carrier ponies , allowing breeders to avoid FIS in their foal crop . | animal types, animal genetics, immunology, lymphoid organs, veterinary diagnostics, veterinary science, veterinary medicine, biology, immune system, veterinary immunology, genetics of the immune system, large animals, bone marrow, genetics, genetics of disease, genetics and genomics, immunoglobulins | null |
journal.ppat.1000831 | 2,010 | Rapid Evolution of Pandemic Noroviruses of the GII.4 Lineage | Norovirus ( NoV ) , a member of the Caliciviridae family , is now considered the most common cause of viral gastroenteritis outbreaks in adults worldwide 1 ., In the US , NoV has been identified as the cause of over 73% of outbreaks of gastroenteritis 1 ., Furthermore , outbreak NoV strains spread rapidly causing great economic burden on society due to medical and social expenses ., Consequently , a vaccine or treatment for NoV would be useful in reducing its transmission and alleviating disease symptoms ., Our current knowledge of NoV replication and evolution has made it difficult to predict the efficacy of a treatment or longevity of a vaccine , as evidence is emerging that NoV , like many other RNA viruses , exists as a dynamic , rapidly evolving and genetically diverse population 2 , 3 , 4 ., The high level of genetic diversity in RNA viruses is recognised as the basis for their ubiquity and adaptability 5 ., Therefore , in order to develop a successful treatment or control program it is first necessary to understand the mechanisms behind NoV replication and evolution ., NoV is a small round virion of 27–38 nm in diameter and possesses a single-stranded , positive-sense , polyadenylated , RNA genome of 7400–7700 nucleotides 6 ., The human NoV genome is divided into three open reading frames ( ORFs ) ., ORF1 encodes for the non-structural proteins , including an NTPase , 3C-like protease and RNA-dependent RNA polymerase ( RdRp ) 7 ., The two structural proteins VP1 , the major capsid protein , and VP2 , the minor capsid protein are encoded by ORF2 and ORF3 , respectively 8 , 9 ., NoV is a highly diverse genus with up to 61% VP1 amino acid diversity between its five genogroups ( GI to GV ) 10 ., Up to 44% amino acid diversity over VP1 is also observed within the genogroups and has resulted in the further subgrouping of GI , GII and GIII into 8 , 17 and 2 genotypes , respectively 10 ., VP1 exhibits the highest degree of sequence variability in the genome 11 , 12 ., It consists of three domains , namely the shell ( S ) domain connected by a flexible hinge ( P1 domain ) to a protruding domain ( P2 ) 13 ., The highly conserved S domain forms the backbone of the capsid structure 13 , while the moderately conserved P1 domain encodes the flexible hinge that connects the S and P2 domains ., The protruding P2 domain possesses motifs that are involved in binding to the host cell , and hence , the P2 domain is responsible for the antigenicity of the virus 14 , 15 ., The most clinically significant of the five genogroups is GII , as it is the most prevalent human NoV genogroup detected and more frequently associated with epidemics compared with other genogroups ., Of particular interest is GII genotype 4 , ( GII . 4 ) , because this lineage accounts for 62% of all NoV outbreaks globally 14 , 15 and has also caused all five major NoV pandemics in the last decade ( 1995/1996 , U5-95_US strain; 2002 , Farmington Hills; 2004 , Hunter; 2006 , 2006a virus; and 2007 , 2006b virus ) 16 , 17 , 18 , 19 ., The basis for the increased epidemiological fitness 20 of the GII . 4 strains , as determined by its high incidence and ability to cause pandemics , is currently unknown ., Investigations with influenza indicate a link between increased viral evolution and increased viral incidence 21 , 22 ., However , because of the non-culturable nature of human NoV , variations in rates of evolution have not been calculated for different NoVs and consequently this has not been investigated as a factor in determining viral incidence and epidemiological fitness ., Replication efficiency and genetic diversity are both important parameters in viral fitness 23 ., The aim of this study was to determine if these two parameters are contributing to the increased epidemiological fitness of the GII . 4 strains ., Replication efficiency and genetic diversity are primarily determined by the viral RdRp , as it controls the rate new sequence is introduced into the genome ., Therefore using in vitro RdRp assays together with bioinformatics , the replication efficiency , mutation rate and rate of evolution of GII . 4 viruses was compared with other NoV GII genotypes ., The results of this study suggest that , like influenza A , the increased incidence of the pandemic GII . 4 lineage may be a result of the combined influence of a high mutation , replication and evolution rate which , together culminate in an increased epidemiological fitness for the GII . 4 strains ., Stool samples containing NoV were obtained from the Department of Microbiology , Prince of Wales Hospital , Sydney , Australia , with the exception of the stool specimen that contained NoV/Mc17/01/Th ( GenBank accession numbers AY237413 ) ., This stool specimen was obtained from McCormic Hospital , Chiang Mai , Thailand 16 ., The six genetically diverse NoV strains used in this study included: three GII . 4 pandemic strains; NoV/Sydney 348/97/AU ( of the NoV/US95_96 GII . 4 pandemic lineage ) 16 , NoV/NZ327/06/NZ ( NoV/2006a GII . 4 lineage ) 17 and NoV/NSW696T/06/AU ( NoV/2006b GII . 4 lineage ) 17 ., Two recombinant strains; NoV/Sydney C14/02/AU ( GII . b ORF1 and GII . 3, ORF2/3 commonly referred to as GII . b/GII . 3 ) 16 and NoV/Sydney4264/01/AU ( GII . 4, ORF1 and GII . 10, ORF2/3 , GII . 4/GII . 10 ) 16 , and a GII . 7, NoV , NoV/Mc17/01/Th associated with rare sporadic cases of gastroenteritis 24 ., In this study , the RdRp enzymes are referred to by their genotype , except in the case of the GII . 4 strains , which are referred to by their pandemic name , eg ., GII . 4 2006b-RdRp ( see Table 1 ) ., RdRps from recombinant strains are indicated by an ‘r’ in front of the nomenclature ., Viral RNA was extracted from 140 µl of 20% faecal suspension using the QIAmp Viral RNA kit according to manufacturers instructions ( Qiagen , Victoria , Australia ) ., RNA was resuspended in 50 µl of Baxter Steri-pour H2O and stored at −80°C ., cDNA synthesis was performed as described previously 16 ., The full length capsid gene , P2 domain and RdRp regions were amplified with specific primers ( Table 2 ) using reverse transcriptase - polymerase chain reaction ( RT-PCR ) methods described in 17 ., The amplified RdRp genes were cloned into pGEM-T Easy vector ( Promega , Wisconsin , United States ) ., Plasmids and PCR products were purified by PEG precipitation and washed with 70% ethanol ., Products were sequenced directly on an ABI 3730 DNA Analyzer ( Applied Biosystems , Foster City , CA , US ) using dye-terminator chemistry ., pGEM-T Easy vectors containing 1736 bp from the 3′ end of ORF1 were purified using the Quantum prep® plasmid miniprep kit ( BioRad , California , United States ) and used as template DNA for the construction of expression vectors ., Strain specific primers incorporating restriction enzyme sites , were designed to amplify the precise RdRp region of each strain ( Table 2 ) ., PCR was performed as described previously 17 ., PCR products were digested with their corresponding restriction enzymes and cloned into the expression vector pTrcHis2A ( Invitrogen , Mount Waverley , Australia ) ., Constructs containing the hepatitis C virus ( HCV ) genotype 3a RdRp ( pVRL69 ) and HCV genotype 1b RdRp ( pVRL75 ) , were used as controls and have been described previously 25 ., Site directed mutagenesis of residue 291 in the GII . 4, US95_96-RdRp and the GII . 4 2006a-RdRp was carried out with the Stratagene Quickchange II mutagenesis kit , according to manufacturers instructions ( Stratagene , La Jolla , United States ) ., The primers used to introduce the mutation into the plasmid are listed in Table 2 ., The NoV RdRps and control HCV RdRps were expressed in Escherichia coli , as described previously 25 , except expression of the NoV RdRps was performed for 4 hr at 30°C ., Purity was checked by SDS-PAGE and the identity of the RdRp was confirmed by western blot with an anti-six histidine antibody and peptide sequencing performed by the Bioanalytical Mass Spectrometry Facility ( University of New South Wales , Australia ) ., Recombinant RdRp was quantified with a Nanodrop ND-1000 Spectrophotometer ( Nanodrop , Wilmington , United States ) ., Kinetic RdRp assays were performed in a final volume of 15 µl and contained 20 mM Tris-HCl ( pH 7 . 4 ) , 2 . 5 mM MnCl2 , 5 mM DTT , 1 mM EDTA , 500 ng of homopolymeric C RNA template , 2 U RNasin ( Promega ) , 4 mM sodium glutamate and increasing concentrations of 3H-GTP ( Amersham Biosciences , Little Chalfont , UK ) ranging from 2 µM to 60 µM ., Reactions were initiated with the addition of 50 nM of RdRp and incubated for 9 mins at 25°C ., The reactions were terminated by adding EDTA to a final concentration of 60 mM , 10 µg herring sperm DNA and 170 µl of 20% ( w/v ) trichloroacetic acid ., The incorporated radionucleotides were precipitated on ice for 30 min and then filtered through a 96 well GF/C unifilter microplate ( Falcon , Franklin Lakes , United States ) by a Filtermate harvester ( Packard BioSciences , Melbourne , Australia ) ., Using the harvester , the filters were washed thoroughly with water and left to dry ., The filter wells were each filled with 25 µl of Microscint scintillation fluid ( Packard Biosciences ) and radioactivity measured using a Packard liquid scintillation counter ( TopCount NXT; Packard Biosciences ) ., Background measurements for each assay consisted of reactions without RdRp and were subtracted from the count per minute ( CPM ) values obtained for the individual enzyme assays ., Results were plotted and statistical analysis performed with the Mann Whitney Test ( one-tailed , 95% confidence interval ) in GraphPad Prism version 4 . 02 ( GraphPad Software , San Diego , CA ) ., An in vitro fidelity assay was developed to measure mutation rates and was adapted from Ward et al . 26 ., The RdRp assay was performed using conditions described above with a homopolymeric C RNA template , except 82 . 1 pmoles of 3HUTP ( 2 µCi ) or 3HATP ( 4 µCi ) ( Amersham Biosciences ) were added ( as the non-complementary nucleotides ) with an equimolar amount of GTP ( 82 . 1 pmoles ) ( Promega ) added as the complementary nucleotide ., The total amount of ribonucleotide incorporated was calculated in a parallel experiment with the addition of 1 µCi ( 164 . 2 pmoles ) 3HGTP ( Amersham Biosciences ) as the correct nucleotide ., The assay was incubated for 50 min at 25°C ., Error frequency of the RdRp was determined by calculating the total number ( pmoles ) of non-complementary ribonucleotides incorporated and dividing by the total number ( pmoles ) of 3HGTP ribonucleotides incorporated ., In order to determine the rate of evolution of the rGII . 3 , GII . 3 , GII . 4 and GII . 7 capsids , the nucleotide sequences of ORF2 were analysed ., RNA capsid sequences used for the analysis included eight from this study and 76 sequences from GenBank , with the oldest strains available dating back to 1987 ., The strains used and their GenBank accession numbers are listed in Text S1 ., The rate of evolution ( substitutions/nucleotide site/year ) for GII . 3 , GII . b/GII . 3, GII . 4 and GII . 7, NoVs was determined by calculating the number of nucleotide substitutions in ORF2 compared to an ancestral strain and this was plotted against time 27 ., The rate of evolution was determined by linear regression with the program GraphPad PRISM® version 4 and was equivalent to the gradient of the line ., Pairwise alignments of RNA sequences and evolutionary distances between sequences were carried out using the Maximum Composite Likelihood model in Mega 4 . 0 28 ., Bootstrapped trees ( 1000 data sets ) were constructed using the Neighbour-joining method , also with the program Mega 4 . 0 ., In order to determine the amount of selection each genotype is under , the average Ka/Ks ratio was calculated for each genotypes capsid gene ( GII . 4 , GII . b/GII . 3 and GII . 7 ) ., The Ka/Ks ratio is a measure of nonsynonymous amino acid changes compared to synonymous ( silent ) changes ., Ka/Ks>1 indicates that positive selection is occurring ., Ka/Ks\u200a=\u200a1 is interpreted as neutral evolution and Ka/Ks<1 is indicative of negative or purifying selection ., The program Sliding Windows Alignment Analysis Program ( SWAAP ) version 1 . 0 . 2 29 was utilised ., The Nei-Gojobori model was used to calculate Ka and Ks values 30 ., The window size was set at 15 bp ( 5 aa ) and the step size was 3 bp ( 1 aa ) ., Predicted secondary structure analysis of the RdRps and capsid protein VP1 were performed by generating a Protein Data Bank ( PDB ) file from the amino acid sequence in FastA format using software on the CPHmodels 2 . 0 Server 31 ., Three dimensional structures were then generated from the PDB files with PyMol 32 ., The GenBank accession numbers for the RdRp and capsid genes described in this paper are listed in Text S1 ., Over the last decade five NoV pandemics have occurred approximately every two years and all pandemics have been associated with a single NoV genotype , GII . 4 16 , 17 , 19 , 34 ., The reason for the predominance of the GII . 4 strains has been the subject of much speculation but is currently unknown primarily due to a limited understanding of NoV population dynamics and evolution 4 , 15 , 35 ., Studies with other RNA viruses indicate that viral fitness is dependent on many factors , such as , viral mutation , replication efficiency , population size and host factors ( reviewed in 2 ) ., To date progress has been made in understanding the role host factors have on NoV prevalence with several studies indicating that variations in viral docking to the blood group antigens may affect infectivity of individuals within a population ( reviewed in 36 ) ., In particular , GII . 4 viruses bind to all blood group antigens , whereas , GII . 1 and GII . 3 viruses bind fewer blood group antigens and this could account for higher prevalence of GII . 4 viruses 36 ., This paradigm however remains controversial , especially for GII NoV , as not all studies show an association between blood group antigens and clinical infection 37 , 38 , 39 ., Apart from the host/viral interaction , no other factors have been affiliated with NoV fitness ., Recent studies performed with poliovirus have shown that an increase in fidelity leads to less genetic diversity and subsequently a reduction in viral fitness and pathogenesis because of a reduced adaptive capacity of the virus 40 , 41 ., It has been hypothesised that viruses are fitter if they are able to produce a more robust ( diverse ) population ( reviewed in 42 , 43 , 44 ) ., In the current study we examined whether there was a link between epidemiological fitness , as defined by their incidence , and the rate and accuracy of viral replication ., In the present study error rates were assessed directly by examining the mutation rate of the viral RdRp and by analysing the rate of evolution for selected GII lineages ., Our results are consistent with mutation rates for the poliovirus RdRp 26 and retrovirus reverse transcriptases 45 , which range between 10−3 to 10−5 ( Table 1 ) ., The more prevalent GII . 4 strains had a 5 to 36-fold higher mutation rate compared to the less frequently detected GII . b/GII . 3 and GII . 7 strains , as determined by in vitro enzyme assays ., Consistent with this , the rate of evolution of the capsid was on average 1 . 7-fold higher in GII . 4 viruses compared to GII . 3 , GII . b/GII . 3 and GII . 7 viruses ., The GII . 4 capsids also had a larger Ka/Ks ratio than the GII . b/GII . 3 and GII . 7 strains suggesting that the increased incidence/epidemiological fitness of the GII . 4 strains maybe through greater antigenic drift , a consequence of the higher mutation rate of the GII . 4, RdRp ., The mutation rates for the control HCV RdRps ( average of 1 . 6×10−3 substitutions per nucleotide site , Table, 1 ) were 2-fold higher compared to the GII . 4, RdRps ., Evaluation of previously published rates of evolution for the HCV hypervariable region 1 ( HVR1 ) within the envelope 2 glycoprotein ( E2 ) were also higher ( 6–fold ) than the NoV GII . 4 rates of evolution calculated in this study 46 ( Table 1 ) ., HVR1 was chosen for comparison because , like the NoV capsid gene , it is the most variable region in the genome and under the greatest immune selection ., Mutation rate and rate of evolution cannot be directly compared as they are indirectly related due to the increased complexity of evolution in vivo 20 ., However , in this study we did find a common trend between the two different measurements of diversity with HCV displaying the highest diversity rate for both measurements compared to NoV ., Interestingly , the majority of non-synonymous mutations in the P2 domain for all three NoV genotypes were localised to six common structural sites ., These six hypervariable regions within the P2 domain were consistent with hypervariable sites for GII . 4 capsids already identified in other studies 4 , 19 ., We demonstrated that GII . 7 and GII . 3 viruses shared two and four common hypervariable sites , respectively , with GII . 4 viruses ( Fig . 5 ) ., Substitutions at one of these sites ( residue 395 ) have been shown to alter GII . 4 strains antigenic profiles 4 ., Localization of the hypervariable sites to common regions on the surface of the P2 domain suggests that these regions are likely to be under immune pressure possibly from a neutralizing antibody response 39 ., The lower number of amino acid changes at these sites for viruses with a GII . 3 capsid may explain why GII . b/GII . 3 is predominantly associated with gastroenteritis cases in children 47 ., This suggests that GII . b/GII . 3 viruses are not as efficient at escaping herd immunity compared to GII . 4 strains and therefore only hosts immunologically naïve to GII . 3 infection are susceptible ., Similarly , we propose that the low prevalence of the GII . 7 strain is also a consequence of a low mutation rate in the RdRp resulting in limited antigenic drift and an inability to escape herd immunity ., Apart from mutation rate , replication rate is considered to be another major determinant in viral fitness 48 ., Replication rates are important because an increased replication rate would produce a larger heterogenous population than a slower replicating virus in the same unit of time , given the same mutation rate ., Interestingly , the RdRps from the recent 2006 GII . 4 pandemic strains had a higher nucleotide incorporation rate than the recombinant GII . 4, RdRp and the US95/96-like pandemic GII . 4, RdRp , which could be associated with a point mutation in the RdRp ( Thr291Lys ) ., Residue 291 is located in the finger domain , which is comprised of five β sheets that run parallel and strongly interact with each other ., The innermost of these five β sheets contains motif F which interacts directly with incoming nucleotides 49 ., Therefore , it is plausible that substitutions at residue 291 affects the orientation of motif F due to the strong interaction between the five β sheets and subsequently alters the binding affinity to the incoming nucleotide ., Fixation of the Thr291Lys point mutation in the GII . 4 lineage after 2001 has been paralleled with a reduction in the period of stasis between the emergence of new antigenic variants 4 ., Alterations in residue 291 after 2001 could have led to an increase in the rate of evolution of GII . 4 strains by increasing the replication rate , however this did not seem to have an effect on mutation rate ( Table 1 ) ., High replication rates did not always correlate with epidemiological fitness as the NoV strain , GII . 7 , had the highest incorporation rate but is considered to be the least fit due to it having the lowest incidence ., Therefore , this study suggests mutation rate in combination with a high replication rate are key determinates in epidemiological fitness ., Influenza research also indicates a relationship between rate of evolution and epidemiological fitness ( reviewed in 21 ) ., New antigenic influenza A variants arise every one to two years and cause more annual epidemics than influenza B , as well as the more devastating pandemics 21 ., Once a population has accumulated mass herd immunity to a virus the virus is forced to alter its antigenic determinants , a possibility for viruses with poor fidelity and fast replication rates , or face extinction 50 , whereas , viruses such as influenza B , which have higher fidelity and slower antigenic change , are more often associated with sporadic cases 21 ., In this study a parallel can be seen in the epidemiology between NoV and influenza , in particular between GII . b/GII . 3 viruses and influenza B and GII . 4 viruses and influenza A . In summary , this study supports the hypothesis that epidemiological fitness is a consequence of the ability of the virus to generate genetic diversity , as the NoV pandemic GII . 4 strains were associated with an increased replication and mutation rate ., Therefore , it would seem that GII . 4 viruses , as opposed to GII . b/GII . 3 and GII . 7 viruses , have reached a balance in their replication rate and mutation rate that is better suited to viral adaptation ., In contrast , it would seem that the GII . 7 lineage , despite having a high replication rate , has a low mutation rate that limits its adaptation and therefore its incidence ., It is important to improve our understanding of the mechanisms underlying NoV epidemiological fitness as future pandemics are expected . | Introduction, Materials and Methods, Discussion | Over the last fifteen years there have been five pandemics of norovirus ( NoV ) associated gastroenteritis , and the period of stasis between each pandemic has been progressively shortening ., NoV is classified into five genogroups , which can be further classified into 25 or more different human NoV genotypes; however , only one , genogroup II genotype 4 ( GII . 4 ) , is associated with pandemics ., Hence , GII . 4 viruses have both a higher frequency in the host population and greater epidemiological fitness ., The aim of this study was to investigate if the accuracy and rate of replication are contributing to the increased epidemiological fitness of the GII . 4 strains ., The replication and mutation rates were determined using in vitro RNA dependent RNA polymerase ( RdRp ) assays , and rates of evolution were determined by bioinformatics ., GII . 4 strains were compared to the second most reported genotype , recombinant GII . b/GII . 3 , the rarely detected GII . 3 and GII . 7 and as a control , hepatitis C virus ( HCV ) ., The predominant GII . 4 strains had a higher mutation rate and rate of evolution compared to the less frequently detected GII . b , GII . 3 and GII . 7 strains ., Furthermore , the GII . 4 lineage had on average a 1 . 7-fold higher rate of evolution within the capsid sequence and a greater number of non-synonymous changes compared to other NoVs , supporting the theory that it is undergoing antigenic drift at a faster rate ., Interestingly , the non-synonymous mutations for all three NoV genotypes were localised to common structural residues in the capsid , indicating that these sites are likely to be under immune selection ., This study supports the hypothesis that the ability of the virus to generate genetic diversity is vital for viral fitness . | Since 1995 , norovirus has caused five pandemics of acute gastroenteritis ., These pandemics spread across the globe within a few months , causing great economic burden on society due to medical and social expenses ., Norovirus , like influenza virus , has over 40 genotypes circulating within the population at the same time ., However , it is only a single genotype , known as genogroup II genotype 4 ( GII . 4 ) , that causes mass outbreaks and pandemics ., Very little research has been conducted to determine why GII . 4 viruses can cause pandemics ., Consequently , we compared the evolution properties of several pandemic GII . 4 strains to non-pandemic strains and found that the GII . 4 viruses were undergoing evolution at a much higher rate than the non-pandemic norovirus strains ., This phenomenon is similar to influenza virus , where an increase in antigenic drift has been associated with increased outbreaks ., This discovery has important implications in understanding norovirus incidence and also the development of a vaccine and treatment for norovirus . | virology/mechanisms of resistance and susceptibility, including host genetics, virology/virus evolution and symbiosis, virology/emerging viral diseases | null |
journal.pgen.1007581 | 2,018 | Cooperation, cis-interactions, versatility and evolutionary plasticity of multiple cis-acting elements underlie krox20 hindbrain regulation | Enhancers are short , cis-acting regulatory elements that modulate transcription of target genes , relatively independently of their orientation or distance with respect to the promoter ., They act as platforms to recruit multiple transcription factors 1 that interact with the transcription machinery at the promoter via cofactors 2 ., A single gene can be controlled by multiple enhancers that show different activity profiles , providing both diversity and specificity of expression 3 , or redundant profiles that may be required to ensure transcriptional robustness 4 ., Interactions between enhancers can occur through different modes of cooperation: additive , synergistic , repressive , hierarchical or competitive 5 ., Multiplicity of enhancers is a common feature among developmental genes 6 and is likely to play a major role in the evolution of gene expression , as it provides the necessary flexibility for pattern evolution 7 ., For many years , the functions of enhancers have been mainly investigated through analysis of transgenic constructs carrying a reporter gene driven by a minimal promoter and linked to the enhancer 8 ., Although fruitful , this approach is based upon the assumption that enhancer function can be recapitulated by the activity profile deduced from such an assay ., However , it has not been established that this is always the case ., In recent years , the advent of easy and efficient genome editing techniques , in particular those based on the CRISPR/Cas9 system , have facilitated mutation of putative enhancers in their natural genomic context 9 , 10 , enabling the direct dissection of enhancer function in various species , including vertebrates 11 ., Hindbrain segmentation is a highly conserved morphogenetic process in vertebrate development 12 ., Among the regulatory genes involved in segmentation , Krox20 ( also known as Egr2 ) plays a particularly important role ., It encodes a zinc-finger transcription factor and is specifically and precisely expressed in two developing hindbrain segments , rhombomeres ( r ) 3 and 5 13–15 ., Krox20 is responsible for the formation and specification of these rhombomeres 16–19 ., The regulation of Krox20 expression in the developing hindbrain provides an attractive model to study the functions and evolution of cis-acting elements involved in control of patterning in vertebrates ., Three evolutionarily conserved enhancer elements active in the hindbrain have previously been identified near the Krox20 gene , termed A , B and C 20 ., Analysis of chicken element A revealed that it is active in r3 and r5 and requires Krox20 binding for this activity 20 , suggesting that it acts as an autoregulatory element ., Indeed , deletion of element A in the mouse leads to a complete loss of Krox20 expression at late stages without affecting early stages , a phenotype very similar to Krox20 loss-of-function 21 ., In contrast , chicken element B enhancer activity is Krox20-independent , and is restricted to r5 20 , 22 , making it a prime candidate for the initiation of Krox20 expression in r5 ., Finally , chicken enhancer C is active in the r3-r5 region , also in a Krox20-independent manner , suggesting that it might contribute to the initiation of Krox20 expression in r3 and/or r5 20 , 23 ., Surprisingly , deletion of element C in the mouse does not affect Krox20 expression at early stages , but leads to a loss of maintenance of Krox20 at late stages in r3 24 ., This loss of maintenance is due to cooperation in cis between element A and C , leading to increased accessibility of element A and potentiation of its autoregulatory activity in r3 24 ., This unexpected function of element C , unlike a classical enhancer , clearly illustrates the necessity of mutating putative cis-regulatory elements in their chromosomal context to decipher their true function ., In spite of these observations , previous analyses do not provide a complete global picture of Krox20 regulation in the hindbrain ., In particular , they do not explain the basis for early Krox20 expression in r3 ., We therefore decided to engage in a systematic search and analysis of Krox20 cis-regulatory elements ., For this purpose , we turned to the zebrafish , which allows easier identification and functional characterisation of regulatory elements and evolutionary comparisons with existing data from other vertebrates ., This approach has revealed a complex cis-regulatory landscape , with 6 elements controlling zebrafish krox20 expression in the hindbrain ., Three of these are the homologues of the previously identified mouse and chicken elements A , B and C . Combining transgenic reporter analyses and CRISPR/Cas9-mediated mutagenesis in the chromosomal context , we assign precise functions to each of these 6 elements and provide a comprehensive view of krox20 cis-regulation in the hindbrain ., Three important features of gene regulation emerge ., First , cooperation and redundancy between multiple cis-elements play a major role in regulation ( for instance , 4 elements cooperate to maintain autoregulation ) ., Second , unexpected versatility of several elements allows them to be involved in different aspects of expression control ., Third , this versatility is underlain by major plasticity across vertebrate evolution , despite the highly conserved pattern of Krox20 expression ., These characteristics of Krox20 cis-regulation are likely shared by other developmental genes and are therefore of broad significance ., To study krox20 cis-regulation in detail in the zebrafish hindbrain , we first analysed its expression pattern by in situ hybridization , to provide a reference for comparison with the activities of putative enhancers ., As krox20 regulation in the hindbrain has been shown to involve a positive feedback loop 20 , we examined both wild type embryos and those carrying a homozygous point mutation in the krox20 coding sequence that abolishes Krox20 function and thereby prevents autoregulation ( krox20fh227 allele 21 , 25 ) ., In agreement with previous studies 21 , krox20 expression is dynamic between the 95% epiboly and 20-somite stages ( 20s ) ., A positive feedback loop contributes to the amplification and maintenance of expression , as in absence of active protein , krox20 mRNA disappears from r3 between 5s and 10s , and from r5 between 10s and 15s ( Fig 1A ) ., In contrast , in the wild type , the mRNA is maintained in both rhombomeres beyond 20s ., krox20 is also expressed in neural crest cells leaving the neural tube from the r5/r6 region ( Fig 1A , arrowhead ) ., To identify the transcriptional enhancers responsible for krox20 expression in the hindbrain , we undertook a systematic approach based on the observation that active cis-regulatory sequences typically show greater DNA accessibility than other sequences ., We assessed chromatin accessibility within the krox20 locus and its vicinity by ATAC-seq 26 ., ATAC-seq was performed on either 95% epiboly whole embryos or on micro-dissected regions ( whole hindbrain , including r3 and r5 , or a posterior region devoid of krox20-expressing cells; Fig 2 ) from 5s and 15s embryos ., These conditions correspond to key moments in krox20’s expression dynamics: at the very beginning of gene activation ( 95% epiboly ) , after activation with limited ( 5s hindbrain ) or full ( 15s hindbrain ) contributions of the autoregulatory loop , and in regions where the gene remains silent ( posterior regions ) ., The analysis revealed 7 major peaks that are present outside of the promoter and coding sequence when krox20 is active ( Fig 2 ) ., As all 7 peaks were located in non-repetitive regions and additional enhancers may have been missed by ATAC-Seq , we extended our survey of functional enhancers to all non-repetitive intergenic regions ., This led to the selection of 22 sequences ( ranging from 720 to 1726 bp ) , 7 containing one of the identified accessibility peaks ( Fig 2 , blue boxes ) ., To evaluate transcriptional enhancer activities associated with the 22 selected sequences , each was cloned into the Zebrafish Enhancer Detection ( ZED ) plasmid 27 , upstream of a GFP reporter gene driven by the gata2 minimal promoter ., These constructs were co-injected with transposase mRNA into one-cell stage zebrafish embryos and GFP fluorescence was monitored ., Among the 22 cloned sequences , 6 led to hindbrain-specific GFP expression ( Fig 1B ) , suggesting that each harboured a transcriptional enhancer ., These 6 sequences were named A to F according to their positions along the locus ., Each sequence included one of the accessibility peaks , demonstrating that assessment of chromatin accessibility by ATAC-seq is a powerful approach to identify cis-regulatory elements ( Fig 2 ) ., All of these peaks ( with the exception of the one corresponding to element F ) were reduced at 15s in the krox20-negative posterior region of the embryo and two of them ( corresponding to elements D and E ) were very small at 95% epiboly ( Fig 2 ) , suggesting that in most of these regions , DNA accessibility is correlated with gene activity ., In silico analysis of the 7th accessible region , located close to the promoter , revealed a putative binding site for the architectural protein CTCF 28 that may participate in increasing chromatin accessibility ., Elements A to E are located upstream of krox20 , whereas element F is located downstream ., Elements A , B , C and F show sequence similarity with the previously identified mouse and chicken hindbrain enhancers A , B and C 20 and the mouse NE element 24 , respectively , and occupy the same relative positions along the locus ( Fig 2 ) ., Sequence conservation between species is relatively high for elements B and F , reduced for element C , and low for element A ( Figs 2 and S1 ) ., Sequences weakly homologous to element E were also identified in the vicinity of the mouse and chicken krox20 gene , again at the same relative positions ( Figs 2 and S1 ) ., No sequences homologous to element D were detected in the mouse or chick ( Fig 2 ) ., To further investigate the activity of the 6 zebrafish elements , embryos injected with each construct were used to generate stable transgenic lines , whose profiles of GFP expression during hindbrain development were established by in situ hybridization ( Fig 1B ) ., At least two independent lines were analysed for each element , with the exception of element F , for which only one line was obtained ., The patterns of GFP expression were identical for the different lines corresponding to the same element ., We found that element A is weakly active in r3 between 3s and 10s , and much stronger in r5 from 3s to beyond 20s ., Element B is active only in r5 between 3s and approximately 10s ., At 10s , element B also drives GFP expression in neural crest cells migrating posteriorly to r5 ( Fig 1B ) ., Element C activity , first observed in r3 at 3s , later extends into r4 at 5s and then into r5 at 10s , and vanishes thereafter ., Elements D and E are both active in r3 and r5 between 3s and 20s; D is more efficient in r3 at early stages , whereas E shows more activity in r5 at late stages ., Finally , element F activity is restricted to r3 , with very early onset ( 95% epiboly ) but rapid extinction ( at around 10s ) ., This enhancer assay suggests that among the 22 non-repetitive intergenic sequences located within and around the krox20 locus , 6 are likely to have hindbrain enhancer activities that reflect aspects of the normal hindbrain expression of the gene ., This conclusion is further supported in that 3 of these elements , A , B and C , appear to show both structural and functional homology to previously characterised mouse and chicken enhancers 20 ., Indeed , the patterns of activity of the homologous elements in the three species are very similar: B is restricted to r5 , A is active in both r3 and r5 , and C is active in a domain extending from r3 to r5 ., Sequences homologous to elements E and F also occur in the chick and mouse genomes , at the same relative positions as in the zebrafish ., Together , the 6 zebrafish cis-acting elements appear to recapitulate all aspects of krox20 expression , in particular early activity in r3 and r5 for F and B , respectively , and intermediate or late activities in both r3 and r5 for all others ., Finally , the fact that almost all major accessibility peaks identified by ATAC-seq correspond to hindbrain enhancers constitutes a strong validation of the use of this procedure to identify novel transcriptional cis-acting elements ., To determine the roles played in krox20 hindbrain regulation by the various cis-acting elements identified in the vicinity of the gene , we generated stable zebrafish lines with deletions of each element using CRISPR/Cas9 technology ., Mutations were obtained by injecting into one-cell stage embryos the Cas9 protein together with two guide RNAs that targeted sequences flanking each element , resulting in its deletion ., Stable lines were then selected; the deletions were characterised by PCR cloning and sequencing ( S2 Fig ) and the lines were used to obtain homozygous mutant embryos ., The generation of stable lines carrying deletions of several elements was sometimes problematic ., In such cases , we used an alternative approach that allowed us to obtain mutations in both alleles , directly in the injected embryo ., Embryos were injected with the Cas9 protein , together with a mix of 3–4 guide RNAs that targeted evolutionarily conserved short sequences and/or putative binding sites for transcription factors , located within a 150–450 bp region presumably corresponding to the core enhancer ( S2 Fig ) ., This procedure was very efficient , allowing the introduction of deletions within both alleles at the same time , as demonstrated by the absence of fragments corresponding to the wild type allele following PCR amplification and further analysis of the DNA sequences ( S3 Fig ) ., Although the deletions introduced in both alleles might be different ( S3B Fig ) , in all the cases analysed they led to complete or almost complete inactivation of the element , as judged by the homogeneity of the phenotypes associated with the mutations and their similarity to those corresponding to germ-line deletion of the same element ( S4 Fig ) ., Genotypes of mutated embryos generated through this approach ( somatic deletion ) are noted with the * symbol following the inactivated element ., To grossly map cis-acting elements governing krox20 in the hindbrain , we first generated a line carrying a deletion , ∆ ( A-E ) , that completely eliminated a 75 kb intergenic region between krox20 and nrbf2 , including the 5 identified upstream elements , but excluding the krox20 promoter region ( S2 Fig ) ., The expression of krox20 in embryos carrying a homozygous ∆ ( A-E ) deletion was dramatically affected: krox20 expression was initiated in r3 , but krox20 mRNA levels rapidly decreased in this rhombomere and no expression was ever observed in r5 ( Fig 3 , ∆ ( A-E ) ) ., This result indicates that cis-acting sequences sufficient for initiation of krox20 expression in r3 are located outside of the deleted region ., In contrast , cis-acting elements necessary for initiation in r5 and maintenance in r3 are located within this region ., There is an obvious candidate for governing krox20 initiation in r3: the downstream element F , which shows enhancer activity at early stages specifically in this rhombomere ( Fig 1 ) ., Indeed , in homozygous mutants with a deletion of element F , krox20 expression was completely abolished in r3 at all stages , whereas r5 expression was unaffected ( Fig 3 , ∆F ) ., Therefore , element F is absolutely required for initiation of krox20 expression in r3 , consistent with its enhancer activity there at early stages ( Fig 1B ) ., In absence of any initiation , the feedback loop cannot be engaged and so no expression is observed at later stages either ., We next sought to identify the cis-acting sequences involved in the initiation of krox20 expression in r5 that are located within the ∆ ( A-E ) deleted region ., For this purpose , we generated zebrafish lines carrying deletions of each of the elements A , B , C , D or E . No phenotype was observed with any deletion in the heterozygous state ., When affecting both alleles , two deletions , ∆A and ∆B , appeared to delay initiation of krox20 expression in r5 ( Fig 3 ) ., In the case of ∆B , r5 expression was completely abolished at 3s and dramatically reduced at 5s , but at later stages , normal levels of expression were gradually reached ( Fig 3 ) ., There was no effect on r3 ., Note that B is the only element whose deletion also obliterates krox20 expression in neural crest cells derived from the r5/r6 region ( Fig 3 , arrowhead ) ., For ∆A , r5 expression was also affected at 3s and 5s , although less severely than in the ∆B mutant ., However , ∆A also led to a slight reduction of expression in both r3 and r5 at later stages ., The other deletions ( ∆C , ∆D and ∆E ) did not affect krox20 expression in r5 at early stages ( Fig 3 ) ., To determine whether elements A and B are the only contributors to the initiation of krox20 expression in r5 , we examined the effect of deleting both , by introducing a deletion of B ( ∆B’ ) in a ∆A background ( S2 Fig ) ., Embryos carrying homozygous deletions of both elements ( ∆A ∆B’ ) show a stronger phenotype than embryos with a single mutation: expression in r5 is only detected after 5s and late expression is also severely affected , presumably because the feedback loop cannot be appropriately established due to late and very poor initiation ( Fig 4 ) ., As there was still limited expression maintained in r5 in the double mutant , we wondered whether a third element might be involved in the initiation step ., Three elements show enhancer activity in r5: C , D and E ( Fig 1 ) ., However , for elements D and E , this activity appears to be totally dependent on the presence of functional Krox20 protein ( Fig 5A ) ., This is not the case for element C , raising the possibility that it could cooperate with elements A and B to initiate krox20 in r5 ., We therefore combined deletions in C with deletions of A and/or B and examined whether any expression remained ., The combination of homozygous B and C deletions did not increase the severity of the phenotype associated with B deletions ( Fig 4 ) ., However , the combination of homozygous A , B and C deletions led to an almost complete loss of expression in r5: only a very low level of mRNA was reproducibly observed at 12s ( Fig 4 ) ., In conclusion , this analysis identified the cis-acting elements involved in the initiation phase of krox20 expression 21: their homozygous mutation affects the hindbrain expression of krox20 at very early stages ( at around 3s ) , before any significant involvement of the autoregulatory loop ( Fig 1A ) ., In r3 , a single element , F , is absolutely required ., In r5 , however , the situation is more complex and elements show partial redundancy ., Although element B appears as the major contributor , elements A and C are also involved and the mutation of all three elements is required to essentially abolish krox20 r5 expression ., Residual expression could be due to very weak activity of a non-characterised fourth element or to the fact that the internal mutations in enhancers A and C do not totally inactivate them ( Figs 4 and S2 ) ., Finally , among the identified elements , in the neural crest derived from r5/r6 , enhancer B is the only one required for krox20 expression ., Three of the krox20 cis-acting elements , A , D and E , appear to share similar characteristics: they act as enhancers in both r3 and r5 , and are active at late stages ( up to 20s ) ., Furthermore , deletions ∆A and ∆E lead to a slight decrease in krox20 mRNA levels in both r3 and r5 after 5s ( Fig 3 ) ., These features suggest that they are involved in the maintenance of krox20 expression and possibly in the underlying positive feedback loop 21 ., In addition , the chick and mouse orthologues of element A contain Krox20 binding sites that are required for enhancer activity 20 , 21 , and mouse element A is absolutely necessary for krox20 autoregulation 21 ., To investigate whether zebrafish elements A , D and E could be involved in direct krox20 autoregulation , we first examined the activity of these elements in the absence of the Krox20 protein ., As indicated above , without Krox20 , the enhancer activities of elements D and E were completely abrogated in both r3 and r5 ( Fig 5A ) , demonstrating that these elements are Krox20-dependent and are likely to be involved in the feedback loop ., In the case of element A , in the absence of Krox20 protein , r3 enhancer activity was completely eliminated , but some r5 activity was maintained , although severely reduced ( Fig 5A ) ., These data indicate that element A possesses a dual function: Krox20-dependent enhancer activities in both r3 and r5 and a Krox20-independent enhancer activity specifically in r5 ., This latter activity is likely to contribute , together with elements B and C , to the initiation of krox20 expression in r5 ( Fig 4 ) ., To determine whether the Krox20-dependent activities of elements A , D and E might involve direct binding of the Krox20 protein , we looked for potential binding sites for Krox20 within the enhancer sequences ., For each , we identified several putative binding sites ( S1 and S2 Figs ) ., Oligonucleotides corresponding to sequences from each enhancer and carrying two of these binding sites were synthesized and used to perform gel retardation experiments in the presence of the Krox20 protein , together with specific or non-specific competitors ., In each case , there was at least one strong retarded band , corresponding to a specific complex with Krox20 ( S5 Fig ) , indicating that these elements contain high affinity Krox20 binding sites and supporting the idea that their enhancer activity is dependent on direct binding of Krox20 ., As the phenotypes associated with the single homozygous mutation of elements A , D or E are limited , it is likely that these elements cooperate to establish full autoregulation ., We tested this hypothesis by combining the different mutations ., Indeed , the combination of two homozygous deletions , affecting A and D , A and E , or D and E severely reduced krox20 expression at 12s and 22s ( Fig 5B ) ., Furthermore , elimination of the three enhancers , either by introduction of deletions affecting each one ( Fig 5B ) or by combination of a deletion of element A with a deletion of the D-E region ( S2 and S4 Figs ) led to complete loss of krox20 expression at 12s and 22s ., Note that neural crest expression at 12s , which relies on element B , is maintained in all cases ( Fig 5B ) ., In conclusion , our data establish that elements A , D and E all carry Krox20-dependent enhancer activities ., Furthermore , these elements cooperate to generate the positive feedback loop that maintains late expression of krox20 ., Finally , these activities are likely to involve direct binding of the Krox20 protein to each enhancer ., On the basis of the above analysis , element C appears somehow peculiar ., Like elements A , D and E , it shows enhancer activity in r3 and r5 , but this activity is Krox20-independent ( Figs 1B and 5A ) ., Furthermore , its activity is not restricted to r3 and r5 , but also covers r4 , with a dynamic anterior-posterior pattern ( Fig 1B ) ., Deletion experiments have shown that element C is a minor contributor to initiation of krox20 expression in r5 ( Figs 3 and 4 ) ., It is also involved in late krox20 expression , as its deletion leads to a slight decrease in krox20 mRNA levels in both r3 and r5 after 5s ( Fig 3 ) , although this is not likely to occur via direct autoregulation ( Fig 5A ) ., To determine whether element C interacts with other elements at late stages , we combined its deletion with mutations in A , D and E . Inactivation of element C did not exacerbate the late phenotype associated with the elimination of element A ( Fig 6 , compare ∆A and ( ∆A C* ) ) ., In contrast , the phenotype was more severe when mutation of element C was combined with mutations of elements D and E ( Fig 6 , compare ( D* E* ) and ( ∆C D* E* ) ) ., In fact , this latter genotype leads to a phenotype similar to that of ( ∆A D* E* ) , although slightly less severe in r5 ( Fig 6 ) , probably due to the more significant involvement of element A in the initiation of krox20 in r5 , as compared to element C . Together , these data are consistent with element C contributing to autoregulation by modulating the activity of element A . Similar cooperation was previously observed in the mouse , where the orthologue of element C , although not directly participating in the positive feedback loop , cooperates in cis with element A to potentiate its autoregulatory activity 24 ., To investigate whether such a cis-cooperation exists between A and C in zebrafish , we generated embryos homozygous for D and E mutations and heterozygous for A and/or C deletions ( Fig 6 ) ., The latter were introduced by crossing ∆A and ∆C homozygous lines and were therefore present on different chromosomes ., When both heterozygous deletions for A and C were present ( ∆A/+ +/∆C D* E* ) , krox20 expression at late stages was affected in a manner similar to the combination ( ∆A D* E* ) , where the deletion of element A is homozygous ., In contrast , when only the heterozygous deletion of C was introduced in the ( D* E* ) background ( ∆C/+ D* E* ) , it did not significantly increase the severity of the ( D* E* ) phenotype ( Fig 6 ) ., These results support the existence of a cis interaction between C and A , required to allow A to participate in the autoregulatory loop ., Together , these data indicate that element A does not take part in autoregulation when a functional element C is not present on the same chromosome ., Therefore , element C cooperates with element A to potentiate its autoregulatory activity , just as in the mouse ., However , in the zebrafish , two additional cis-regulatory elements , D and E , directly participate in the feedback loop ., In contrast to element A , element D and E are not likely to depend on element C to exert their enhancer activities ., Zebrafish element A acts both as a Krox20-independent initiator element in r5 and as an autoregulatory element in r3 and r5 ., The existence of these dual activities is surprising in view of what we know of its chicken and mouse orthologues ., Chicken element A is totally dependent on Krox20 binding for its enhancer activity , as demonstrated by comparison of a reporter transgene in mouse Krox20 null and wild type backgrounds , and by mutation of element A Krox20 binding sites with enhancer activity assessed in chick embryos 20 ., In addition , while deletion of mouse element A completely abolishes the positive feedback loop , it has no effect on early expression in r5 in this species 21 ., Therefore , element A does not appear to act as an initiator element in chick nor mouse , suggesting its enhancer activity has been modified during vertebrate evolution ., To investigate whether a coherent pattern of evolution of the element might be identified , we analysed the activities of orthologues of element A from several key species in the vertebrate phylogenetic tree ( Fig 7 ) ., We cloned the orthologues of zebrafish element A ( zA ) identified by sequence alignments from koi carp Cyprinus rubrofuscus ( kA ) , spotted gar ( sA ) , Xenopus tropicalis ( xA ) , chicken ( cA ) and mouse ( mA ) into the ZED GFP expression vector , generated stable zebrafish transgenic lines ( at least two independent ones for each species ) and determined the patterns of GFP expression by in situ hybridization ., In a wild type zebrafish background , despite the heterospecific character of the assay , all elements behaved similarly and could direct GFP expression in r3 and r5 , although there were some relative variations in the expression level between the two rhombomeres ( Fig 7 ) ., To determine whether any of these enhancer activities were dependent on the Krox20 protein , we injected transgenic embryos from each line with the Cas9 protein and guide RNAs targeting the sequences encoding the three zinc fingers of the Krox20 protein , which constitute the DNA binding domain ( S2 Fig ) ., This treatment effectively abolishes krox20 expression at 12s ( S6 Fig ) , and allows the assessment of Krox20-independent enhancer activity ., A large proportion of the activities of the enhancers was Krox20-dependent ( Fig 7 ) ., However , limited Krox20-independent activities were maintained in some cases ., Surprisingly , their patterns appeared different from one species to another and incoherent with the phylogenetic tree: the zebrafish and spotted gar elements remained active in r5 only , whereas the koi carp element was only active in r3; the mouse element was weakly active in both r3 and r5 , whereas no activity was detected with the Xenopus and chick elements ( Fig 7 ) ., In conclusion , this analysis shows that the features required for Krox20-dependent expression of element A are likely to have been largely conserved during the course of vertebrate evolution ., In contrast , the capacity of this element to combine its autoregulatory activity with Krox20-independent initiator functions appears highly contingent , with no clear correlation with the course of evolution ., Furthermore , this Krox20-independent activity can occur in r3 , in r5 or in r3 and r5 , revealing a surprising plasticity of element A for acquiring and losing additional functions during evolution ., Several studies , mostly performed in Drosophila , have recently shown that cooperation between cis-regulatory elements is a common feature in the regulation of developmental genes and can occur according to different modes , including in particular additive , synergistic or hierarchical interactions 3 , 5 , 24 , 29 , 30 ., The present study provides examples of such co-operations in vertebrates , in the initiation of krox20 expression in r5 as well as in the positive feedback loop ( Fig 8 ) ., Although we have not performed quantitative analyses of the contributions of each cis-acting element to the different aspects of krox20 hindbrain expression , in the case of initiation in r5 , this cooperation appears to occur through an additive mode: deletion of each element leads to reduced expression ( with B>A ) , and a drastic decrease requires combination of both deletions ., The third element , C , appears only as a minor contributor to this activity ., More generally , the transcriptional activity of each of these r5 initiating elements shows specificities ( Fig 1B ) that may reflect differences in which transcription factors act on them 20 , 22 , 23 , 31 ., Considering autoregulation , three elements ( A , D and E ) can directly bind Krox20 protein ( S5 Fig ) ., Elimination of each one alone leads only to a mild phenotype ( with E>A>D , Fig 3 ) ., However , combined knockdown of any two elements results in a major decrease in late expression ( Fig 5B ) , suggesting the existence of a strong synergistic component in this co-operation ., Therefore , in this case , synergy and redundancy are not exclusive , as an almost full activity is already reached with two elements ., Redundancy in the cis-acting elements controlling zebrafish krox20 autoregulation differs remarkably from the situation in the mouse , in which element A is absolutely required for late Krox20 expression 21 ., While we were not able to detect sequences homologous to element D in the mouse Krox20 locus , there is a poorly conserved mouse orthologue of element E , although , it cannot rescue the deletion of element A in this species ., Overlapping activities between regulatory elements add robustness to the expression of developmental genes 3 , 32 ., We speculate that the difference in redundancy in the control of the krox20 feedback loop between zebrafish and mouse might reflect differences in both external and internal conditions that require addit | Introduction, Results, Discussion, Materials and methods | Cis-regulation plays an essential role in the control of gene expression , and is particularly complex and poorly understood for developmental genes , which are subject to multiple levels of modulation ., In this study , we performed a global analysis of the cis-acting elements involved in the control of the zebrafish developmental gene krox20 ., krox20 encodes a transcription factor required for hindbrain segmentation and patterning , a morphogenetic process highly conserved during vertebrate evolution ., Chromatin accessibility analysis reveals a cis-regulatory landscape that includes 6 elements participating in the control of initiation and autoregulatory aspects of krox20 hindbrain expression ., Combining transgenic reporter analyses and CRISPR/Cas9-mediated mutagenesis , we assign precise functions to each of these 6 elements and provide a comprehensive view of krox20 cis-regulation ., Three important features emerged ., First , cooperation between multiple cis-elements plays a major role in the regulation ., Cooperation can surprisingly combine synergy and redundancy , and is not restricted to transcriptional enhancer activity ( for example , 4 distinct elements cooperate through different modes to maintain autoregulation ) ., Second , several elements are unexpectedly versatile , which allows them to be involved in different aspects of control of gene expression ., Third , comparative analysis of the elements and their activities in several vertebrate species reveals that this versatility is underlain by major plasticity across evolution , despite the high conservation of the gene expression pattern ., These characteristics are likely to be of broad significance for developmental genes . | Animal development relies on the early delimitation of specific embryonic territories that will later participate in the formation of tissues and organs ., This process is governed by sets of so-called developmental genes ., The activities of the genes are themselves controlled by associated DNA sequences called enhancers ., In vertebrates a part of the embryonic brain is delimited by the activity of the gene krox20 ., In this study , we have performed a comprehensive analysis of the krox20 regulatory landscape in the zebrafish vertebrate model ., We show that 6 enhancers cooperate according to different modes to establish the complete pattern of krox20 activity ., Furthermore , these enhancers appear unexpectedly versatile , combining different types of activities ., This versatility is underlain by major plasticity across vertebrate evolution , despite the high conservation of the delimitation process ., These observations are likely to be of broad significance for developmental genes . | fish, medicine and health sciences, in situ hybridization, molecular probe techniques, gene regulation, brain, vertebrates, animals, animal models, osteichthyes, developmental biology, model organisms, hindbrain, experimental organism systems, molecular biology techniques, embryos, research and analysis methods, embryology, probe hybridization, gene expression, molecular biology, zebrafish, eukaryota, enhancer elements, anatomy, genetics, biology and life sciences, evolutionary biology, organisms, evolutionary developmental biology | null |
journal.pgen.1004610 | 2,014 | Tissue-Specific RNA Expression Marks Distant-Acting Developmental Enhancers | Development and function of mammalian tissues rely on the dynamic control of tissue-specific gene expression , a process largely regulated by distant-acting transcriptional enhancers 1–3 ., Disruption of enhancer sequences can lead to severe phenotypes in mouse models 4–9 ., Furthermore , population-scale genetic studies indicate that a large proportion of sequence variants associated with human diseases affect non-coding functions in the genome , of which enhancers are a major category 10 ., Despite their functional relevance , the genome-scale identification of enhancers that are active in vivo in developmental and disease processes remains challenging ., In principle , genome-wide profiling of enhancer-associated epigenomic marks ( e . g . H3K27ac and CBP/p300 ) enables the genome-scale identification of enhancers predicted to be active in a given cell type or tissue 2 , 3 , 11–15 ., However , none of these marks is unique to enhancer regions or found at all enhancers and ChIP-based technology has well-documented limitations with sensitivity and specificity 3 , 14–16 ., Recently , expression of short non-coding transcripts has been described as a feature of many enhancers with a possible tight correlation between cell type-specific enhancer activity and eRNA expression levels 17–20 ., Using cap analysis of gene expression ( CAGE ) in a collection of human tissues and cell type , Andersson et al . 21 identified over 40 , 000 candidate enhancers marked by bidirectional capped RNA expression suggesting that RNA transcription can provide a complementary approach for de novo enhancer discovery ., Anecdotal evidence suggests a functional requirement for such eRNAs in enhancer-mediated gene regulation 22 , 23 ., Regardless of the molecular mechanisms underlying eRNA-mediated regulatory functions , the prevalence of eRNA transcription at the whole transcriptome level in vivo and whether eRNA expression signatures can potentially be used as an independent mark for in vivo enhancer discovery remain poorly explored ., In this study , we compare eRNA expression profiles determined via total RNA sequencing across developmental mouse tissues and demonstrate highly tissue-specific genome-wide expression signatures of eRNAs in vivo ., We find that eRNA expression globally correlates with tissue-specific enhancer activity and that RNAs transcribed from in vivo enhancers constitute a major proportion of tissue-specifically expressed non-coding RNAs ., Finally , we demonstrate through application of reporter assays in transgenic mice that differential expression of eRNAs can correctly predict tissue-specific in vivo enhancer activities independent of other chromatin-associated marks ., To test the hypothesis that eRNA transcription marks active in vivo enhancers in a tissue-specific manner , we first measured eRNA expression from 15 intergenic enhancers active in mouse embryonic forebrain or limb buds that were randomly selected from a larger collection of previously identified in vivo enhancers 24 ., We assessed eRNA expression from each enhancer by quantitative RT-PCR across three different embryonic mouse tissues including forebrain , limb , and heart as a negative control ( Figure 1 ) ., While baseline expression of each eRNA was detected in all three tissues , in 80% of cases eRNAs from tissue-specific enhancers showed highest expression in the predicted tissue compared with the other two tissues ( 12/15; p\u200a=\u200a0 . 0006 , Fishers exact test ) , suggesting that eRNAs are commonly expressed from tissue-specific developmental enhancers with a quantitative relationship between eRNA transcription and tissue-specific enhancer activity ., To study eRNA expression from in vivo enhancers beyond this small-scale qPCR screen , we examined genome-wide total RNA transcription in embryonic heart and limb , two tissues with different developmental origins and trajectories , and with divergent in vivo enhancer landscapes as assessed by epigenomic marks 25–27 ., We extracted total RNA from limb and heart tissues microdissected at mouse embryonic day E 11 . 5 ., Following ribosomal RNA depletion , we used a strand-specific total RNA sequencing protocol to generate more than 200 million sequencing reads from each tissue ( see Methods , Table S1 ) ., While the majority of sequencing reads ( 53% in heart , 60% in limb ) mapped to annotated mouse cDNA sequences , a considerable proportion ( 38% in heart , 30% in limb ) mapped to introns as well as intergenic regions , consistent with a possible association with in vivo enhancers ., Examination of individual genomic loci containing known enhancers revealed examples of bidirectional tissue-specific eRNA expression from validated intergenic and intragenic enhancers consistent with their in vivo activity ( Figure 2A and Figure S1 ) ., These results indicate widespread transcription from non-coding sequences in vivo and anecdotally support correlation of in vivo enhancer activity with tissue-specific eRNA transcription ., In order to assess tissue-specific eRNA expression more systematically , we examined eRNA expression associated with a large collection of in vivo-validated tissue-specific enhancers 24 , 26 , 28 ( http://enhancer . lbl . gov ) ., To avoid confounding factors arising from the presence of pre-mRNAs , we restricted this analysis to intergenic in vivo enhancers ( see Methods ) ., We examined a total of 145 such enhancers that are active in heart or limb ., In general , enhancers were substantially enriched in uniquely mapped reads , and they were nine times as likely as random non-coding regions to contain ten or more independent reads within 1 kb of the enhancer midpoint ( p\u200a=\u200a5 . 5E-108 based on background distribution; see Table S2 and Methods ) ., While 41% of enhancers met this stringent threshold , overall 92% of enhancers showed evidence of at least weak transcription ( ≥1 uniquely mapped reads; p\u200a=\u200a2 . 3E-15 based on background distribution , see Table S2 and Methods ) ., Consistent with our small-scale sampling of enhancers by quantitative PCR ( Figure 1 ) , 79% of heart enhancers and 83% of limb enhancers showed higher eRNA expression in the tissue where enhancer activity was observed in vivo ( Figure 2B; p<10−8 , Fishers exact test ) ., We next examined tissue-derived RNA signatures at intergenic regions enriched for enhancer-associated p300 and H3K27ac epigenomic marks 27 , 29 from the same tissues ( see Methods ) ., Similar to known in vivo enhancers , eRNA transcription was highly enriched around the center regions defined by ChIP-Seq , and tissue-specific eRNA expression patterns correlated with the predicted enhancer activity based on tissue-specific p300 or H3K27ac signature in the same tissues ( Figure 2C–F; See Methods ) ., This global correlation between tissue-specific eRNA expression and enhancer activity corroborates previous observations derived from CAGE analysis of human cell types and tissues 21 and supports the possibility that eRNA expression profiling from tissues may provide an effective approach for identifying tissue-specific in vivo enhancers ., To explore the potential of eRNA profiling for de novo enhancer discovery , we first used a sliding window approach to identify candidate intergenic regions enriched for RNA expression ., Known coding and intronic regions and unannotated transcripts were removed , which led to the identification of 3 , 422 and 3 , 775 intergenic regions in heart and limb , respectively , that showed marked RNA expression at a conservatively chosen threshold of ≥10 uniquely mapped reads ( see Methods; Figure S2A–B and Table S2 ) ., These regions included 834 heart-specific and 1 , 078 limb-specific loci ( tissue-specifically transcribed regions , TSTRs ) that were differentially expressed in these two tissues ( Figure 3A–B and Table S3 ) ., Most of these ∼2 , 000 TSTRs were located distal to the nearest transcription start site ( Figure S2C ) ., There is substantial overlap between TSTRs identified from developing mouse tissues in this study and candidate transcription start sites ( TSSs ) captured by CAGE from mouse cells and tissues 30 ., Overall , 45% of heart TSTRs and 55% of limb TSTRs overlap with at least one CAGE-derived TSS candidate ., This represents a strong enrichment compared to random control sequences ( 8% and 8 . 3% , respectively; p<4 . 3E-68 , Fishers exact test , see Methods ) , but also indicates that large numbers of additional enhancer candidates were identified by analysis of ex vivo tissue at relevant developmental stages ., Tissue-specific expression of a panel of 22 candidate TSTRs was tested and in all cases confirmed by quantitative RT-PCR ( Figure 3C–D , see Methods ) , demonstrating that these RNA-seq data sets accurately identified non-coding TSTRs across tissues ., To assess whether these TSTRs may represent in vivo enhancers , we first examined their evolutionary sequence constraint , a feature associated with many distant-acting enhancers 25 , 31 , 32 ., We found that 69% and 73% of TSTRs in heart and limb , respectively , overlap with elements under evolutionary constraint as compared to 28% and 27% of random control sequences ( p<2 . 0E-62 , Fishers exact test; Figure 4A and Figure S2D ) ., Additionally , heart TSTRs are enriched near genes critical for cardiovascular and heart development , whereas limb TSTRs are enriched near genes involved in muscle tissue development and limb development/morphogenesis ( Table 1 ) ., Heart and limb TSTRs are also enriched for different sets of transcription factor binding motifs related to development of the respective tissues compared with random genomic sequences ( Table S5 and Table S6 ) ., Finally , we compared tissue-specific TSTR expression with mRNA levels of nearby genes in two tissues ( see Methods ) ., The strongest correlation was observed between TSTRs and their nearest genes ( Pearson correlation: R\u200a=\u200a0 . 68 for heart , R\u200a=\u200a0 . 55 for limb ) , and decreased substantially for more distant genes ( Figure 4B ) ., These results support that TSTRs may represent regulatory elements coordinating the transcription of nearby genes ., To evaluate the overlap of TSTRs with enhancer-associated epigenomic marks , we examined p300 and H3K27ac enrichment ( Figure 4C–D ) ., We find that 36% and 46% of heart and limb TSTRs are marked by p300 and/or H3K27ac ., TSTRs with and without epigenomic enhancer marks show similar expression level and substantial evolutionary constraint ( Figure 4A and Figure S3A–B ) ., However , the transcription of TSTRs with enhancer marks tends to be more balanced in both directions , whereas TSTRs marked by tissue-specific RNAs only are more biased toward one direction ( Figure S3C–D ) ., In addition , TSTRs negative for p300 and/or H3K27ac are more distal to the nearest transcription start sites ( Figure S3E ) ., These results indicate a substantial overlap of extragenic TSTRs with enhancer-like regions ., However , this does not exclude the possibility that subsets of the observed TSTRs represent other classes of regulatory elements or unannotated non-coding loci ., To directly assess the potential of TSTRs identified by transcriptome profiling for the de novo discovery of tissue-specific in vivo enhancers , we used a transgenic mouse enhancer assay previously shown to reliably capture in vivo enhancer activity 25 , 27 , 33 ., In an initial retrospective comparison , we found that heart- or limb-specific TSTRs overlap with 12 tested elements that had previously been examined due to increased conservation or enhancer associated epigenomic marks 24 ., Of these elements , 9/12 ( 75% ) were annotated as tissue-specific positive enhancers in vivo ( Table S7 , http://enhancer . lbl . gov ) ., Next , we performed transgenic mouse assays for another set of 19 TSTRs that had not previously been tested ( Table S8 ) and exhibited tissue-specific RNA expression ., This panel included elements both with and without detectable p300 and/or H3K27ac signal in ChIP-Seq experiments ( Table S8 ) that were chosen blind to the identity of nearby genes ., Mouse genomic DNA for individual TSTRs with up to 2 kb of flanking sequence was cloned upstream of a minimal heat shock promoter fused to a lacZ reporter gene and transgenic mice were assayed by whole-mount staining for the expression of lacZ reporter at E11 . 5 25 ( see Methods ) ., Only elements that drove reproducible reporter gene expression pattern in at least three embryos were considered positive enhancers ., In total , 8/19 ( 42% ) candidate enhancers predicted by tissue-specific RNA expression functioned as positive enhancers in vivo ( Figure 5 , Table S8 and Figure S4 ) ., In all cases , the observed tissue-specific in vivo enhancer activity was consistent with the tissue specificity of the corresponding TSTR ., As representative examples , transgenic whole-mount embryos and transverse sections for elements mm1052 , mm1018 , mm1054 and mm1064 are shown in Figure 5 ., In these examples , reproducible LacZ reporter activities were detected in both atrial and ventricular regions of the heart ( Figure 5A–C ) and anterior regions of the fore- and hindlimb ( Figure 5D ) ., Combining the results from newly performed enhancer assays and retrospective comparisons with pre-existing in vivo data sets , 17 of 31 TSTRs ( 55% ) represented in vivo enhancers , and for 15 of these 17 enhancers ( 88% ) the tissue specificity of eRNA expression correctly predicted the in vivo enhancer activity patterns ., These results support the general utility of eRNA profiling as an informative mark for in vivo enhancer prediction ., Recent large-scale transcriptome studies suggest that up to 80% of mammalian genomes may be actively transcribed 34–37 ., While many of these transcripts show differential expression signatures across cell types and tissues , the majority of non-coding transcripts have not been associated with in vivo functions ., In the present study , we explored the in vivo expression dynamics of tissue-specific non-coding RNAs using a total RNA-Seq strategy that captures both coding and non-coding transcripts 18 ., Our results suggest that the majority of enhancers show evidence of tissue-specific eRNA transcription ., In addition , de novo identified tissue-specifically transcribed non-coding regions ( TSTRs ) showed major characteristics of canonical enhancers ., These results indicate that enhancers are a predominant function associated with differentially expressed non-coding loci across developing tissues ., CAGE analysis from human cell lines and tissues showed that incorporating enhancer expression data can increase the validation rate of ENCODE enhancer predictions and that bidirectional capped RNA signatures can in principle be used to identify de novo cell-specific enhancers 21 ., However , in the absence of sizable in vivo validation data sets , the quantitative correlation between tissue-specific eRNA expression and in vivo enhancer activity in mammalian developmental processes has remained unclear 21 ., We have tested a set of 19 candidate enhancers predicted by tissue-specific RNA expression in transgenic mouse assays and 42% showed reproducible enhancer activity in vivo , demonstrating the general utility of eRNA-based enhancer prediction in a developmental mammalian system ., Of note , two of the tissue-specific enhancers reported in this study ( mm1052 and mm1061 ) did not overlap with any CAGE peaks collected from 399 mouse samples 30 despite the scope of the tissue and cell type panels examined in these previous studies ., Considering the dynamics of the enhancer landscape in developing tissues and organs 29 , it appears likely that many additional enhancers active during development will be identifiable by whole transcriptome analysis of tissues across different developmental stages ., While a substantial proportion of extragenic transcription appears linked to enhancer activity , our observation of several TSTRs that were not active in the transgenic enhancer reporter assays supports the hypothesis that eRNA-like transcripts can also originate from other non-coding elements , such as inactive enhancers ., These observations are consistent with recent mechanistic studies on eRNAs showing that eRNA transcription precedes the establishment of H3K4me1/2 38 , suggesting that eRNA transcription may occur before enhancer activation ., TSTRs without supportive p300/H3K27ac marks show significant , though slightly decreased conservation , less bi-directional transcription , and are more distal to the nearest coding genes ( Figure S3 ) , suggesting that they may have different biological functions ., Consistent with this observation , a larger proportion of TSTRs with supportive p300/H3K27ac marks were active in vivo compared to TSTRs without such marks , although this difference was not significant at the sample size examined ( p\u200a=\u200a0 . 15 , Fishers exact test; Table S8 ) ., While the results of our study do not permit strong conclusions about the functionality of intergenic loci that exhibit transcription but no accompanying enhancer epigenomic signatures , it is possible that these regions are less likely to be active enhancers ., Transcription may be occurring due to other processes or at a different class of regulatory element than active enhancers ., Together , our data suggest that additional criteria such as bi-directional transcription , conservation and independent enhancer marks may further increase the performance of eRNA-based enhancer predictions ., Nonetheless , considering the overall substantial correlation between TSTRs and tissue-specific in vivo enhancer activity , our results corroborate that short non-coding transcripts are commonly associated with the regulation of cell type- and tissue-specific gene expression ., Enhancer RNAs may be very unstable and sensitive to exosome degradation 21 , 39 , resulting in low steady-state level in cells ., This may explain why eRNAs represent a small proportion of the transcriptome profile ( Figure S2A ) , despite the large number of sites from which they originate ., At current sequencing depth , many enhancers may still be missed ( Figure S2B ) , which is consistent with the notion that a great proportion of mammalian genomes may be actively transcribed and cis-regulatory genomic elements may represent major sites of extragenic non-coding transcription 34–37 , 39 ., Recently , Andersson et al . showed that depletion of a co-factor of the exosome complex resulted in an over 3-fold average increase of eRNA abundance 21 ., Thus , a combination of in-depth transcriptome profiling and exosome depletion may provide a more sensitive method for eRNA-based enhancer discovery ., Emerging evidence indicates that eRNA transcripts can be required for enhancer-mediated gene activation ., Targeted knock-down of specific eRNAs has been shown to affect the expression of enhancer target genes in cell-based assays , providing a potential strategy for altering gene expression in experimental and therapeutic applications 22 , 23 , 40 ., Through in-depth transcriptome profiling , we have shown extensive eRNA expression in developing tissues , as well as a global correlation of eRNA expression with tissue-specific in vivo enhancer activity ., Our results highlight the widespread and potentially important role of eRNAs in orchestrating gene expression , providing support for the general feasibility of eRNA-based targeting of in vivo gene expression ., Embryonic heart or limb tissue was isolated from CD-1 strain mouse embryos at E11 . 5 by microdissection in cold PBS 27 ., A single sample consisting of tissue pooled from multiple embryos was analyzed for either tissue ., After washing , about 1 ml TRIzol reagent ( Life Technologies , 15596-026 ) was added to every 100 mg of tissue sample , followed by homogenization using a glass dounce homogenizer ., Total RNA from individual tissues were extracted following the manufacturers instructions ., Genomic DNA contamination was removed by using the TURBO DNA-free kit ( Applied Biosystems , AM1907 ) following manufactures protocol , and the RNA samples were stored at −80°C before further processing ., In order to perform the transcriptome analysis by Illumina sequencing , ribosomal RNAs was removed from total RNA ( 5∼10 µg per reaction ) by using two rounds of the RiboMinus Eukaryote Kit for RNA-Seq ( Life Technologies , A10837-08 ) following the manufacturers instructions ., The quality of total RNA after rRNA removal was analyzed on RNA 6000 Pico chip ( Agilent , 5067-1513 ) to assure that rRNA contamination was less than 30% ., 100 ng total RNA after rRNA removal were used to construct the individual sequencing libraries for Illumina sequencing ., Strand-specific RNA-Seq libraries were created following in-house protocols ., Briefly , RNA samples were fragmented with 10×Fragment buffer ( Ambion , AM9938 ) to achieve an average fragment size of 200–300 nt ., First strand cDNA synthesis was performed with random hexamer and Superscript II reverse transcriptase ( Life Technologies , 18064-014 ) ., During the second strand synthesis , dUTP was used instead of dTTP to introduce strand-specificity ., After adaptor ligation and size selection , the second strand containing dUTP was cleaved by AmpErase UNG ( Life Technologies , N8080096 ) ., The resulting strand-specific cDNA was subjected to 12 cycles of PCR amplification and sequenced with HiSeq 2000 instrument ., 50 sequencing cycles were carried out ., Raw Illumina reads ( 50 bp ) were first filtered using the Illumina CASAVA-1 . 8 FASTQ Filter module ( http://cancan . cshl . edu/labmembers/gordon/fastq_illumina_filter/ ) ., The remaining sequence tags were mapped back to the mouse genome ( NCBI build 37 , mm9 ) using bowtie2 41 , and the alignments were extended to 200 bp in the 3′ direction to account for the average length of DNA fragments ., Repetitively mapped reads were excluded from the following analysis ., For de novo peak calling , a sliding window method EnrichedRegionMaker module from USEQ 42 was employed ., For eRNA-based enhancer predictions , a conservative threshold of 10 or more reads ( without considering strand specificity ) was chosen based on the observation that in retrospective comparison with in vivo validated enhancers , 40 . 7% of enhancers met or exceeded this expression threshold , compared to 4 . 5% of random control regions ( p\u200a=\u200a5 . 5E-108 , Table S2 ) ., Enriched regions overlapping with refGene , mouse mRNA , or ESTs ( mm9 ) were also removed before the downstream analysis ., This process was performed individually for heart and limb RNA-Seq data ., To generate Figure S2B , 10% to 100% of sequencing reads were randomly selected from the raw sequencing data , and de novo peak calling was individually performed to identify the enriched intergenic regions ., Among raw enriched regions , tissue-specifically transcribed regions ( TSTRs ) were defined as non-coding regions with significantly higher expression in this tissue compared with the other tissue ( p<0 . 01 , two-proportion z-test; Figure 3A–B ) 43 with the equation shown below:where ( n represents mappable reads within each TSTR in heart or limb , and N represent the total number of mappable reads excluding ribosomal regions in the corresponding tissue ) and ., RPKM<2−9 were arbitrarily set to 2−9 for visualization purposes in Figure 3A–B ., Candidate transcription start sites ( TSSs ) marked by CAGE peaks were downloaded from http://fantom . gsc . riken . jp/5/ 30 and extended to 1 kb each side from the peak midpoint ., For each TSTR ( 1 kb around the peak center ) , the overlapping candidate TSSs were identified by BEDTools 44 ., Random control peaks were also generated using BEDTools with the same number and size of sequences and excluding known genes , mouse mRNAs and ESTs ., We compared tissue-derived RNA signatures at intergenic regions to enhancer-associated p300 27 and H3K27ac marks from the same tissues and time-point ., H3K27ac ChIP-Seq datasets are described in more detail in Nord et al . 29 and Attanasio et al . 45 ., Candidate tissue-specific intergenic enhancers were predicted by ChIP-Seq of p300 ( 171 in heart , 656 in limb ) or H3K27ac ( 6965 in heart , 2174 in limb ) as described previously 27 ., Briefly , uniquely aligned sequencing reads were extended to 300 bp in the 3′ direction ., Enriched regions ( peaks ) were identified with MACS 46 ( p≤1E-5 ) using matched input as controls ., Peaks overlapping with repetitive regions , known genes , mouse mRNAs and ESTs were removed for further analysis ., Summary eRNA coverage plots were generated for p300- and/or H3K27ac-derived intergenic enhancers within a 10 kb window , centering on the maximum ChIP-seq coverage ., Using the mapped reads , normalized mean eRNA coverage values were calculated for 25 bp windows across the 10 kb regions scaled by total mapped reads ., For mean calculations , only the 5th–95th percentiles were used to reduce the effect of outliers ., Coverage was calculated separately for antisense and sense reads , and as a combined value ., For the summary plots , a loess best fit line was plotted for each of the eRNA datasets ( limb and heart ) , separating into sense and antisense reads ( Figure 2C–F ) ., Pre-computed conservation scores ( phastCons scores ) generated from 30 vertebrate genome alignments were download from the UCSC Genome Browser 47 ., For each TSTR ( 1 kb around the peak center ) , the conservation score was defined as the most highly constrained overlapping phastCons element in the mouse mm9 genome ., Random control peaks were generated using BEDTools with the same number and size of sequences and excluding known genes , mouse mRNAs and ESTs 44 ., The percentages of TSTRs and random control regions overlapping phastCons elements were plotted in Figure 4A ., Tissue-specific TSTRs were classified as enriched in p300 and/or H3K27ac if the relative ChIP-seq coverage was equal to or greater than the 95th percentile of experiment background coverage estimated across 1 Mb of unique sequence ., After classification , coverage heatmaps were generated for ChIP-seq data using normalized coverage values , with input corrections ., Coverage was plotted for 25 bp windows centered on the peak RNA coverage and extending 25 kb on either side ., For plotting purposes , coverage was centered and scaled using mean and SD in order to compare signal across datasets ., TSTRs were organized as no H3K27ac and p300 signal , enriched in H3K27ac signal only , enriched in p300 signal only and enriched in both marks from the top to the bottom in Figure 4C–D ., Known heart or limb enhancers were downloaded from Vista Enhancer Browser ( http://enhancer . lbl . gov ) ., For known enhancer regions , the expression level of individual eRNAs was defined as the mapped sequencing reads within a 2 kb window around the center of in vivo tested enhancers ., For eRNAs only expressed in one tissue , the mapped number of reads was arbitrarily set to 1 in the other tissue in order to compute the absolute fold change for plotting purposes in Figure 2B ., Fold change was defined as higher expression level divided by lower expression of each eRNA in two tissues ., For the volcano plot , y axis represents p-value for the expression differences of each known enhancer , which was computed by two-proportion z-test 43 ., Coverage of randomly selected control regions ( excluding known genes , mRNA and ESTs ) was also computed and iterated 100 times to estimate the genome-wide background based on normal distribution ., The percentages of enhancers or the average percentage of control regions with indicated numbers of uniquely mapped reads in either tissue are listed in Table S2 , as well as associated p-values ., After peak calling , for each individual TSTR , normalized RPKM ( Reads Per Kilobase per Million mapped reads ) was calculated in two tissues ( heart and limb ) with the raw mapped RNA-Seq data within a 2 kb window around the center of each TSTR ., Then , a tissue-specificity index was computed as ( s−u ) / ( s+u ) , in which s is the expression of TSTR in the matching tissue and u is its expression in the other tissue ., The expression of mouse refGene ( mm9 ) was also analyzed in the same way by computing the RPKM across annotated cDNA regions in two tissues ., The tissue-specific expression correlation between TSTRs and their nearby genes was computed as described 18 with minor modifications ., Briefly , we paired each TSTR with the nearby genes ., For each set of genes with the same ranked distance to TSTRs ( the first to the fifth closest genes ) , genes were ranked based on tissue-specificity indices and grouped into 20 genes per bin ., Average tissue-specificity indices from each bin were used to compute the correlation ., The Pearson correlation between nearby genes and the corresponding TSTRs was conducted with the statistics module in the R package ( http://cran . r-project . org/ ) ., Gene ontology analysis for the genes near TSTR regions was performed by GREAT version 2 . 02 48 ., Enriched GO biological processes with a binomial p-value and fold enrichment were listed in Table 1 ., For TSTRs in heart and limb , enriched motifs were computed within a 2 kb window around the center of individual TSTRs by the motif finding module of HOMER ( Hypergeometric Optimization of Motif EnRichment ) 49 ., Known motifs for transcription factors with a p-value less than 10−2 compared with random genomic sequences were reporter in Table S5 and Table S6 ., For directionality analysis , the expression of individual TSTRs in sense and antisense strands was defined as the strand-specific mapped sequencing reads within a 2 kb window around the center of TSTRs in either heart or limb ., Then the directionality index was defined as |f−r|/ ( f+r ) , in which f is the expression of TSTR in one strand and r is its expression in the other strand in the same tissue ., Total RNA was extracted from independently collected pools of heart or limb tissues with the same method as described before and synthesized into cDNA by reverse transcription using the SuperScript First-Strand Synthesis System ( Invitrogen ) ., Candidate TSTRs for RT-PCR validations were randomly selected from the top 30% differentially expressed regions ranked by Z scores ., Expression analysis of candidate TSTRs was carried out by real-time PCR using gene-specific primers ( Table S4 ) and KAPA SYBR FAST qPCR Master Mix ( KAPA Biosystems ) on a Roche LightCycler 480 ., All primers were designed in silico using Primer3 ( http://primer3 . wi . mit . edu/ ) and tested for amplification efficiency ., Target gene expression was calculated with the 2−ΔΔCT method 50 and normalized to the Gapdh housekeeping gene ., Candidate enhancers for in vivo testing were selected randomly from TSTRs with a p-value less than 0 . 01 ., The tested regions included up to 2 kb genomic DNA flanking the TSTRs on either sides ., This general transgenic procedure has been described before 25 , 27 ., Briefly , the selected regions were PCR amplified from mouse genomic DNA and cloned into the Hsp68-promoter-LacZ reporter 51 , 52 ., Genomic coordinates and the PCR primers for the cloned regions are listed in Table 8 ., The transgenic embryos were assayed at E11 . 5 for expression patterns ., A positive enhancer is defined as an element with reproducible expression pattern in at least three embryos resulting from independent transgenic integration events 27 ., For histological analysis , selected embryos were embedded in paraffin and sectioned using standard methods ., RNA-seq data is available through GEO under accession number GSE58157 ., In vivo transgenic data is available through the Vista Enhancer Browser under the identifiers used throughout this study ( http://enhancer . lbl . gov ) . | Introduction, Results, Discussion, Methods | Short non-coding transcripts can be transcribed from distant-acting transcriptional enhancer loci , but the prevalence of such enhancer RNAs ( eRNAs ) within the transcriptome , and the association of eRNA expression with tissue-specific enhancer activity in vivo remain poorly understood ., Here , we investigated the expression dynamics of tissue-specific non-coding RNAs in embryonic mouse tissues via deep RNA sequencing ., Overall , approximately 80% of validated in vivo enhancers show tissue-specific RNA expression that correlates with tissue-specific enhancer activity ., Globally , we identified thousands of tissue-specifically transcribed non-coding regions ( TSTRs ) displaying various genomic hallmarks of bona fide enhancers ., In transgenic mouse reporter assays , over half of tested TSTRs functioned as enhancers with reproducible activity in the predicted tissue ., Together , our results demonstrate that tissue-specific eRNA expression is a common feature of in vivo enhancers , as well as a major source of extragenic transcription , and that eRNA expression signatures can be used to predict tissue-specific enhancers independent of known epigenomic enhancer marks . | Up to 80% of mammalian genomes are actively transcribed , producing large numbers of non-coding RNAs without known functions ., One particularly exciting category of such non-coding transcripts are the recently discovered enhancer RNAs ( eRNAs ) transcribed from distant-acting enhancer elements ., Studies in cell-based paradigms suggest a functional requirement for such eRNA in enhancer-mediated gene regulation ., In this study , we explored the in vivo expression dynamics of tissue-specific non-coding RNAs in embryonic mouse tissues via in-depth transcriptome profiling ., Our results suggest that enhancers may be a predominant function associated with differentially expressed non-coding loci across developing tissues , and that differential eRNA expression signatures from total RNA-Seq can be used to identify uncharacterized tissue-specific in vivo enhancers independent of known epigenomic marks ., Our results highlight the widespread and potentially important role of eRNAs in orchestrating gene expression and the necessity for functional studies in interpreting genome-wide enhancer predictions . | genetics, biology and life sciences, developmental biology | null |
journal.pgen.1007842 | 2,018 | RNA-on-X 1 and 2 in Drosophila melanogaster fulfill separate functions in dosage compensation | In eukaryotic genomes several long non-coding RNAs ( lncRNAs ) are associated with chromatin and involved in gene expression regulation , but the mechanisms involved are largely unknown ., In both mammals and fruit flies , they are required to specifically identify and mark X-chromosomes for dosage compensation , a mechanism that helps maintain balanced expression of the genome ., The evolution of sex-chromosomes , for example the X and Y chromosome pairs found in mammals and flies , leads to between-gender differences in gene dosage ., Although some genes located on the X-chromosome are expressed in a sex-specific mode , equal expression of most of the genes in males and females is required 1 , 2 ., Thus , gradual degeneration of the proto-Y chromosome causes an increasing requirement to equalize gene expression between a single X in males and two X-chromosomes in females ., X-chromosome expression must also be balanced with expression of the two sets of autosomal chromosomes ., Several fundamentally different mechanisms that solve the gene dosage problem and provide such balance have evolved 1–4 ., In mammals , one of the pair of X-chromosomes in females is largely silenced through random X-chromosome inactivation , a mechanism that involves at least three lncRNAs 5 , 6 ., One , the long noncoding Xist RNA , plays a key role in marking one of the X-chromosomes and recruiting Polycomb repressive complex 2 , thereby mediating its inactivation by histone H3 lysine 27 methylation 7 ., In fruit flies , the gene dosage problem has been solved in an apparently opposite way , as X-chromosomal gene expression is increased by approximately a factor of two in males 2 , 3 ., This increase is mediated by a combination of general buffering effects that act on all monosomic regions 8–10 and the specific targeting and stimulation of the male X-chromosome by the male-specific lethal ( MSL ) complex ., The MSL complex consists of at least five protein components ( MSL1 , MSL2 , MSL3 , MLE , and MOF ) and two lncRNAs , roX1 and roX2 3 , 11 , 12 ., Although the mammalian and fly compensatory systems respectively inactivate and activate chromosomes in members of different sexes , both rely on lncRNA for correct targeting ., Results of UV-mediated crosslinking analyses suggest that only one species of roX is present per MSL complex in Drosophila 13 ., Furthermore , inclusion of a roX species is essential for maintaining correct targeting of the MSL complex to the X-chromosome 14 ., Upregulation of the male X-chromosome is considered to be partly due to enrichment of histone 4 lysine 16 acetylation ( H4K16ac ) , mediated by the acetyltransferase MOF ., The increased expression of X-linked genes in male flies is generally accepted , but the mechanisms involved have not been elucidated ., Proposed mechanisms , which are hotly debated 15–17 , include increased transcriptional initiation 18 , 19 , increased elongation 20 , 21 or an inverse dosage effect 22 ., The roX1 and roX2 RNAs differ in sequence and size ( 3 . 7 kb versus 0 . 6 kb ) but can still individually support assembly of a functional MSL complex ., In an early study of roX1 and roX2 , a short homologous stretch was detected 23 , which subsequently led to the definition of conserved regions shared by the two RNAs named roX-boxes , located in their 3’ ends 24–26 ., Confirmatory genetic studies have shown that expression of six tandem repeats of a 72-bp stem loop region from roX2 is sufficient for mediation of the MSL complex’s X-chromosome binding and initiation of H4-Lys16 acetylation in the absence of endogenous roX RNA 24 ., The roX RNAs are not maternally deposited and transcription of roX1 is initiated in both male and female embryos at the beginning of the blastoderm stage 27 ., Females subsequently lose roX1 expression and a few hours after roX1 is first detected roX2 appears , but only in males 28 ., Despite differences in size , sequence and initial expression , the two roX RNAs are functionally redundant in the sense that mutations of either roX1 or roX2 alone do not affect male viability and they both co-localize with the MSL complex along the male X-chromosome 23 , 27 ., In contrast , double ( roX1 roX2 ) mutations , which cause a systematic redistribution of the MSL complex , are lethal for most males 29–32 ., It should be noted that in roX1 roX2 mutant the reduction in MSL complex abundance on the male X-chromosome is dramatic; more pronounced than the reductions observed in mle or mof mutants 14 ., Nevertheless , some roX1 roX2 mutant males may survive , while mle , msl1 , msl2 , msl3 or mof loss-of-function mutations are completely male-lethal 29–31 ., Whether other RNA species can fulfill the role of roX RNAs in these instances or the MSL complex can function without RNA species remains to be clarified ., Furthermore , the degree of lethality in roX1 roX2 mutant is highly sensitive to several modifying factors , such as expression levels of MSL1 and MSL2 33 , expression of hairpin RNAs 34 , 35 , presence and parental source of the Y-chromosome 31 , and a functional siRNA pathway 36 ., The observations that roX1 roX2 mutations are not completely lethal and there are several modifying factors suggest an additional layer of redundancy in the role of lncRNAs in chromosome-specific targeting ., To further our understanding of the role of lncRNAs ( particularly specific roles and redundancies of roX1 and roX2 ) in chromosome-specific regulation we here provide a comprehensive expression analysis of roX1 , roX2 and roX1 roX2 mutants to explore the redundancy as well as the differences between the two lncRNA species ., We show that roX1 and roX2 have partly separable functions in dosage compensation ., In larvae , roX1 is the most abundant variant and the only variant present in the MSL complex when the complex is transmitted ( physically associated with the X-chromosome ) in mitosis ., Loss of roX1 results in reduced expression of the genes on the X-chromosome , while loss of roX2 leads to MSL-independent upregulation of genes with male-biased testis-specific transcription ., In roX1 roX2 mutant , gene expression is strongly reduced in a manner that is not related to proximity to high-affinity sites ., Initial evidence on localization of roX RNAs originates from immunostaining experiments on polytene chromosomes ., Indeed , both roX1 and roX2 are expressed in salivary gland cells and co-localize on polytene chromosomes close to perfectly ( Fig 1A ) ., Overall , the intensities of roX1 and roX2 RNA in situ hybridization signals correlate closely , and the localization patterns along the X-chromosome are nearly identical , except at cytological band 10C , where the roX2 signal is notably stronger than the roX1 signal ., As cytological band 10C is the location of the roX2 gene , this implies that roX2 is favored in MSL complexes targeting the roX2 region rather than roX1 ., At the onset of dosage compensation in the early male embryo , expression of roX is differentially regulated 27 , 28 ., A burst of roX1 transcription in the blastoderm stage is the initial step preceding assembly of the MSL complex ., This occurs independently of roX2 expression , which does not begin until 2 h after the MSL complex is first detectable on the X-chromosome ., In Schneider 2 cells , roX2 is expressed more strongly than roX1 and is detectable by FISH in 95% of them , while roX1 signals , although bright , are visible only in a small fraction of the cells 37 ., We therefore asked whether roX1 and roX2 are expressed in different Schneider 2 cells ., Simultaneous detection of both roX RNAs showed that the rare cells that express roX1 also express roX2 ( Fig 1B ) ., Therefore , in contrast to salivary glands and embryos , only a small fraction of S2 cells express both roX RNAs and all those expressing roX2 also express roX1 ., To investigate roX localization and targeting in cells undergoing mitosis we subjected neuroblasts of male larvae and 5–6 h embryos to RNA in situ hybridization analysis ., While both roX1 and roX2 were clearly visualized in the “X-territory” in most interphase cells , only roX1 signals were detected on the distal part of the metaphase X-chromosome ( Fig 1C and 1D and S1A Fig ) ., We also observed targeting of MLE to the distal part of the mitotic chromosome ( S1A Fig ) , and such targeting by MSL2 and MSL3 has been previously shown 38 , 39 ., We conclude that expression and/or targeting of roX RNAs is differentially regulated depending on the cell type and cell cycle stage , and roX1 RNA is the dominant roX RNA bound to the X-chromosome as part of MSL complexes during mitosis ., The roX2 mutant allele Df ( 1 ) 52 , the most commonly used roX2 loss-of-function allele , carries a deletion spanning a gene-dense region , including roX2 30 ., Removal of this region is lethal , so it is compensated with a rescuing cosmid , frequently P{w+ 4Δ4 . 3} ., Nevertheless , roX2 is not the only gene affected by the widely used combination Df ( 1 ) 52 P{w+ 4Δ4 . 3} , and genes carrying it differ considerably in genetic background from roX1 and wild type flies ., In a previous microarray analysis , potential background problems were solved by comparing roX1 roX2 mutant flies with roX2 flies as controls 40 ., Here , to analyze differences in expression profiles of single ( roX1 and roX2 ) mutants and double ( roX1 roX2 ) roX mutants we decided to create a deletion mutant of roX2 without affecting adjacent genes ., Such a mutant would permit analysis of single and double mutants using a roX1+ roX2+ strain as a control and facilitate various other genetic analyses ., To create the desired mutant allele , we used the CRISPR-Cas9 technique to induce two double-strand breaks simultaneously in the roX2 locus and recovered four roX2 deletion mutant strains ( Fig 2A and S2 Fig ) ., All deletions in these mutants span the longest exon of roX2 , including two conserved roX-boxes ., As expected , all four mutant strains were viable and fertile ., Further analysis was performed with the roX29-4 allele , hereafter designated as the roX2 mutant ., This deletion does not uncover the intergenic regions flanking roX2 and therefore it is less likely to affect the flanking genes nod and CG11650 ., The breakpoints are located almost precisely at the sites of double-strand breaks , deleting the region from 7 bp upstream of the annotated transcription start site to 60 bp upstream of the annotated gene end ., RNA in situ hybridization confirmed the absence of roX2 RNA in salivary glands ( Fig 2B ) , while the roX1 signal intensity and binding pattern were apparently unchanged in the roX2 mutant ., In larval brain of roX1 mutants the roX2 RNA was still observed in the X-territory of interphase cells , however it was not detected on the metaphase X-chromosome ( S1B Fig ) ., We recombined the newly made roX29-4 allele with the roX1ex6 mutant allele 30 to obtain the roX1ex6 roX29-4 double mutant flies , hereafter roX1 roX2 mutant ., As observed with other mutant alleles , removal of both roX RNAs resulted in high male-specific lethality beginning at the third instar larvae stage and continuing through pupal development , although a small number of adult males hatched ., The next experiments were designed to investigate the specific roles ( if any ) of the roX RNA species in dosage compensation and assess potential additional functions in regulation of gene expression ., For this , we sequenced ( using an Illumina platform ) polyadenylated RNA from wildtype , roX1 mutant , roX2 mutant and roX1 roX2 mutant 1st instar male larvae ., This developmental stage was chosen to minimize indirect effects of dosage compensation failure in the roX1 roX2 mutant , as roX1ex6 roX29-4 1st instar larvae are healthier than those of later stages ., The four genotypes compared are not isogenic , however , the outcrosses as described in Material and methods ensure that the entire autosomal complement is heterozygous in all genotypes and half of it will have identical origin ., Still , we cannot fully exclude that remaining differences in genetic background could be a contributing factor to the observed changes in expression for some genes ., In wildtype larvae , roX1 RNA was approximately ten times more abundant than in roX2 mutant larvae ( Fig 2C ) ., Notably , we observed increases in abundance of both roX RNAs in response to absence of the other , but not establishment of wildtype roX levels , in the single mutants ., More specifically , we recorded 89% reductions in roX RNA levels in the roX1 mutant , while removal of roX2 RNA ( which normally constitutes only 7% of the total roX RNA complement ) resulted in a 45% increase in roX1 RNA abundance on average ., Therefore , the single mutants differ considerably in levels of roX RNA ., Moreover , although viability and fitness are not affected in either of the single mutants , the efficiency of dosage compensation is significantly compromised in the roX1 mutant ., The average log2 expression ratio of the X-chromosome in this mutant was -0 . 13 , corresponding to an 8 . 6% reduction in average expression of X-chromosome genes relative to genes on the four major autosomes ., In the roX2 mutant , the average expression ratio for X-chromosome genes was lower than that of autosomal genes , but density distributions for X and autosomal expression ratios were very similar ( Fig 3A and 3B and S3 Fig ) ., A Mann-Whitney U-test confirmed that the two populations cannot be differentiated in terms of these expression parameters , so global X-chromosome transcription is not significantly affected in the roX2 mutant ., In conclusion , the roX2 mutant shows no lack of compensation and has roX levels comparable or even higher than wildtype ., Thus , it is not clear whether the total amount of roX or the type of roX is responsible for the observed reduction in average expression of X-chromosome genes in the roX1 mutant ., The results also implies that the observed increase in levels of roX1 RNA in the roX2 mutant ( Fig 2C ) does not lead to hyper-activation of the X-chromosome but is enough to maintain proper X-chromosome expression ., We and others have previously shown that in absence of roX RNAs , the MSL-complex become less abundant on the X-chromosome and relocated to heterochromatic regions including the 4th chromosome 14 , 30 , 37 , 40 ., In fact , the fourth chromosome is related to the X-chromosome and evolutionary studies have shown that the 4th chromosome was ancestrally an X-chromosome that reverted to an autosome 41 , 42 ., Importantly , upon analysis of the 4th chromosome we detected weak but significant downregulation of genes on the fourth chromosome as a specific consequence of roX2 deletion ( Fig 3A ) , but not the previously reported downregulation of the fourth chromosome in the roX1 roX2 mutant flies 43 ., As expected , strong downregulation of X-linked genes occurred in the roX1 roX2 mutant ( Fig 3C ) ., However , it was more severe ( a 33% reduction relative to wildtype levels ) than previously reported in microarray studies 40 , and following RNAi depletion of MSL proteins 9 , 43–45 ., The distribution plot shows that the vast majority of genes were downregulated in the roX1 roX2 mutant and the entire distribution of X-chromosomal gene expression was shifted approximately -0 . 56 on log2 scale relative to the expression of genes on the four major autosomal arms ., The expression ratios of X-linked genes varied widely , especially in the roX1 roX2 mutant ( Fig 3C ) ., It has been proposed that MSL complexes are assembled at the sites of roX RNA transcription , then spread to the neighboring chromatin in cis direction , as well as diffusely , gradually binding to more distant loci ., In addition , our in situ hybridization results indicate enrichment of roX2 RNA at cytological region 10C ., We therefore tested if dosage compensation has a distinct spatial pattern along the X-chromosome ., We observed some clustering of genes related to sensitivity to roX1 or roX2 RNAs , but it appeared to be randomly distributed spatially , except for a gradual decrease in expression of genes in the proximal X-chromosome region in the roX1 mutant , and the 10C region in the roX2 mutant ( Fig 4A ) ., A number of studies have estimated that the MSL complex binds specifically to roughly 250 chromatin entry sites , high-affinity sites ( HAS ) or Pion-X sites ., Since roX RNAs are important for the spreading of the MSL complex from these high-affinity sites we asked whether the extent of genes’ differential expression in roX mutants correlates with their distances from these sites ., Dot plots of genes’ expression ratios against their distances from HAS or Pion-X sites showed weak trends , but were difficult to interpret due to high variation ( S4 Fig ) ., Thus , for more informative visualization we grouped the genes into bins with increasing distance from HAS ( Fig 4B ) ., In roX1 mutant , the average expression ratio was not significantly affected by the distance from HAS ., This was also true for genes located within approximately 30 kb from HAS in roX2 and roX1 roX2 mutant ., However , more remote genes had higher average expression ratios in roX2 and roX1 roX2 mutant , and thus are less suppressed in the double mutant and even upregulated in the roX2 mutant ., On polytene chromosomes in the roX1 roX2 mutant we still observed MSL targeting on the X-chromosome , but only at HAS 14 ., This might suggest that genes close to HAS would retain dosage compensation function also in the absence of roX RNAs ., On the contrary , our results show that genes within approximately 30 kb from HAS are strongly and equally affected while genes more distal to HAS are less sensitive to the absence of roX and absence of bound MSL complex ., We next asked if the roX-dependent dosage compensation depends on the binding strength of the MSL complex , using publicly available chromatin immunoprecipitation data on MSL1 , MOF and MSL3 46 to correlate with our differential expression data ( Fig 5A–5C and S5 Fig ) ., All X-chromosome genes were ranked in order of increasing MSL complex enrichment and divided into five bins with equal numbers of genes ., Thus , bin 1 included unbound and weakly bound genes , while bin 5 included genes highly enriched in MSL proteins ., We found that genes in bins 1 and 2 responded more variably to removal of either or both roX RNAs , a pattern that is probably related to their low expression levels ( Fig 5H ) ., In the single roX mutants , expression ratios did not correlate with enrichment of MSL proteins ( Fig 5A and 5B and S5A , S5B , S5D and S5E Fig ) , indicating that MSL complex-regulated genes uniformly respond to the absence of one roX RNA , regardless of the enrichment levels in wildtype flies ., Strikingly , strong and significant upregulation of genes classified as non- or weakly MSL complex-binding was detected in the roX2 mutant , similarly to genes located far from HAS ( Fig 5B and S5B and S5E Fig ) ., In roX1 roX2 mutant , these weakly MSL complex-binding genes are still suppressed , but much less than strongly binding genes ., Since the MSL complex is still enriched at HAS in the absence of roX it is surprising that dosage compensation by roX RNA-free MSL complexes has low efficiency even for genes with the highest MSL enrichment ., The genes highly enriched in MSL1 and MSL3 ( bin 5 ) were slightly less down-regulated , but this trend was not seen with MOF enrichment bins ( S5F Fig ) ., Since genes with low MSL complex-binding levels are less suppressed than others in the roX1 roX2 mutant , and upregulated in the roX2 mutant , we asked whether dosage compensation in the absence of roX depends on genes’ expression level ., For this , we divided the X-chromosome genes into 12 equally sized bins according to their expression levels ., In accordance with observations regarding genes that weakly bind the MSL complex , we observed upregulation of weakly expressed genes in the roX2 mutant and less pronounced reduction in their expression in the roX1 roX2 mutant ( Fig 5D–5F ) ., High-affinity sites are defined as those that retain incomplete MSL complexes in msl3 , mle or mof mutants 45 , 47–51 , and it has been suggested that MSL complex-binding is directed by hierarchical affinities of target sites 49 , 50 ., In the roX1 roX2 mutant we observed more pronounced reductions in MSL complex abundance on the male X-chromosome than those reported in msl3 , mle or mof mutants , but the remaining MSL targets in the roX1 roX2 mutant were highly reminiscent of those described in msl3 , mle and mof mutants 14 , 30 ., We observed reduced expression of strongly MSL-binding genes in the roX1 roX2 mutant , which is intriguing as these genes are assumed to retain the MSL complex 14 ., Thus , to test the suggestion , we explored correlations between the MSL binding bins and 263 high affinity sites defined by targeting in mle , mof or msl3 mutants , or following their depletion 45 , 51 , 52 ., In parallel we analyzed the 208 peaks we previously identified in the absence of roX1 roX2 14 ., The previously defined 208 peaks in the roX1 roX2 mutant overlap 405 genes on the X-chromosome , 309 of which are among the 328 genes in bin 5 ( Fig 5G ) ., We conclude that the 208 MSL peaks defined in the roX1 roX2 mutant correspond more strongly with genes in the highest MSL binding class than the previously defined HAS do ( Fig 5G ) ., Intriguingly , expression of X chromosomal genes also correlates with MSL1 binding enrichment ( Fig 5H ) , and thus overlap with HAS ., This suggests that the distribution of MRE motifs and consequently MSL complex-binding is governed by gene expression in a manner that promotes adequate dosage compensation in males ., In higher eukaryotes replication timing is connected to the chromatin landscape and transcriptional control 53 ., Generally , early replicating regions are associated with active transcription 54–56 whereas late replicating regions are associated with inactive regions and heterochromatin 57 ., Genome-wide studies on cultured Drosophila cells have revealed dependency of male-specific early replication of the X-chromosome on the MSL complex 56 , 58 ., We therefore asked whether X-chromosomal or genome-wide sensitivity to a specific roX mutant condition correlate with replication timing ., Using available data on replication timing from analyses of S2 and DmBG3 ( male ) and Kc167 ( female ) cells 58 we classified the genes as early or late replicating ., Based on our RNA-seq data we then calculated expression ratios for genes grouped by their chromosome location ( autosomal or X-chromosomal ) and their replication timing as determined in the three cell types ., Conceivably , early and late X chromosomal replication domains ( determined from analyses of S2 and DmBG3 male cell cultures ) are respectively associated with genes bound and unbound by the MSL complex , and thus are affected in similar manners by roX mutations ( Fig 6 and S6 Fig ) ., In female Kc167 cells the relation between sensitivity to roX and replication timing is generally similar to that observed in male cell cultures ., However , in Kc167 cells the X-chromosome has a slightly different pattern of replication domains , which shifts the average expression ratio ( Fig 6 and S6 Fig ) ., In particular , the distribution of distinctively upregulated X chromosomal genes in the roX2 mutant only corresponds with the distribution of late-replication regions in male cells ., Notably , in larval neuroblasts and embryonic cells ( Fig 1C and 1D ) , we only detected roX1 RNA ( no roX2 RNA ) on mitotic X-chromosomes , suggesting that roX1-containing MSL complexes mediate dosage compensation in the G1 phase , when replication timing is established 59 ., It is tempting to speculate that selective transmission of roX1-containing MSL complexes through mitosis enables the cells to quickly and efficiently establish the correct chromatin state and hence maintain correct replication timing ., Transcription upregulation of the X-chromosome in the roX2 mutant is associated with genes classified as having low expression levels , late replication and weak MSL complex-binding ., We asked if this observed upregulation is caused by mis-targeting of MSL complexes associated with excess of roX1 , i . e . , if the upregulated genes are enriched in MSL complexes due to increases in roX1 levels and/or loss of roX2 ., To test this possibility , we assessed relative enrichments of MSL1 and H4K16ac on the upregulated genes by ChIP-qPCR analyses ., In the roX2 mutant , none of the eight genes we tested became targeted by MSL1 or enriched in H4K16ac at a comparable level to known MSL target genes ( S7 Fig ) ., In contrast , enrichment levels were similar to those detected on the autosomal control genes RpS3 and RpL32 ., We therefore conclude that stimulation of weakly expressed X chromosomal genes in the roX2 mutant is not mediated by induced targeting of the MSL complex ., Further analysis of upregulated genes in the roX2 mutant showed that they included not only X chromosomal genes but also late-replicating autosomal genes ., This , together with the absence of MSL complex-enrichment on these genes , indicates that the upregulation is a roX2-specific effect and at least partly separable from MSL complex-mediated gene regulation ., Intriguingly , we discovered that these upregulated genes in the roX2 mutant strain include high proportions of genes ( both X-chromosomal and autosomal ) with male-biased testis-specific transcription ( Fig 7 ) ., Whether roX2 has a specific role in transcriptional regulation of genes involved in spermatogenesis or the observed phenomenon is an indirect consequence of roX2 mutation is an intriguing question that warrants further investigation ., The dosage compensation machinery involving roX1 and roX2 RNAs provides a valuable model system for studying the evolution of lncRNA-genome interactions , chromosome-specific targeting and gene redundancy ., LncRNAs differ from protein coding genes and are often less conserved at the level of primary sequence , as expected due to their lack of protein-coding restrictions ., Like those encoding other lncRNAs , rapid evolution , i . e . , low conservation of the primary sequences of roX genes has complicated comparative studies 24 , 60 ., Despite their differences in length and primary sequences , roX1 and roX2 have also been considered functionally redundant in Drosophila melanogaster ., However , remarkably considering their rapid evolution and apparent redundancy , orthologs for both roX1 and roX2 have been found in all of 26 species within the Drosophila genus with available whole genome assemblies 60 ., Models that explain evolutionarily stable redundancy have been proposed 61 suggesting that the presence of both roX1 and roX2 in these diverged species may be attributable to differences in targets , affinities and/or efficiency or additional functions ., On polytene chromosomes , binding patterns of roX1 and roX2 are more or less indistinguishable , except in region 10C where roX2 is almost exclusively present ., In the roX2 mutant , genes located in the 10C bin are on average downregulated , but similar downregulation of genes in many other bins is observed , so the effect cannot be directly attributed to loss of roX2 ., In wildtype 1st instar larvae , levels of roX1 RNA are much higher than levels of roX2 RNA ., Interestingly , in roX1 mutant larvae the absolute amount of roX2 RNA increases , but only to ~10% of wildtype levels of total roX RNA ., This appears sufficient to avoid lethality , but still causes a significant decrease in X-chromosome expression ., However , despite the huge difference in amounts , not only in number but even more considering the size of the two roX RNAs , the staining intensities of roX RNA on roX1 mutant and wildtype polytene chromosomes seem to be roughly equal ., On mitotic chromosomes we only observed roX1 RNA in the MSL complexes bound to the distal X-chromosome and this binding is not redundant ., This indicates that just after cell division roX1 RNA will be the dominating variant in assembled MSL complexes ., Taken together , our results suggest that roX2 RNA has higher affinity than roX1 RNA for inclusion in MSL complexes ., Moreover , varying amounts of the two species with different affinities at given cell cycle stages may support proper transmission , spreading of assembled MSL complexes and maintenance of appropriate levels of the complexes ., It should be noted that some male roX1 roX2 mutant escaped , so loss of roX is not completely male-lethal , unlike loss of mle , msl1 , msl2 , msl3 or mof 29–31 , 62 ., The complete male lethality in these mutants is attributed to reductions in dosage compensation that have been measured in several studies and observed not only in msl mutants but also following RNAi-mediated depletion of MSL proteins 9 , 43–45 ., Notably , the average reduction of X-chromosome expression , relative to wildtype levels , calculated in these cases has varied from ca ., 20 to 30%; substantially less than the 35% reduction we observed in the roX1 roX2 mutant ., Some of the reported differences may be due to use of different techniques and bioinformatics procedures ( including use of different cut-offs for expression and developmental stages ) ., However , the reasons why some males can survive the very dramatic imbalance observed in expression of a large portion of the genome are unclear ., Furthermore , the reduction in expression of X-chromosome genes observed in the roX1 mutant is not accompanied by any reported phenotypic changes , indicating that D . melanogaster has high intrinsic ability to cope with significant imbalances in X-chromosome expression ., We speculate that in parallel with a compensation mechanism that addresses dosage imbalances the fly has evolved a high degree of tolerance to mis-expression of the X-chromosome ., The 4th chromosome in D . melanogaster ( the Muller F-element ) is related to the X-chromosome ., Evolutionary studies have shown that sex chromosomes do not always represent terminal stages in evolution—in fact , the 4th chromosome was ancestrally an X-chromosome that reverted to an autosome 41 , 42 ., Moreover , the fly shows high and unusual tolerance to dosage differences 63 and mis-expression 8 , 64–66 of the 4th chromosome ( although much smaller than the tolerance to those of the X-chromosome ) ., These observations suggest that tolerance of mis-expression is a common outcome in the evolution of sex-chromosomes and this property has been retained with respect to the 4th chromosome , even after its reversion to an autosome ., We propose that high tolerance of mis-expression in the absence of full functional dosage compensation may be selected for during evolution of sex-chromosomes ., This is because gradual degeneration of the proto-Y chromosome will be accompanied by an increasing requirement to equalize gene expression between a single X- ( in males ) and two X-chromosomes ( in females ) , but changes in genomic location of highly sensitive genes will be favored during periods of incomplete ( or shifting ) dosage compensation ., On transcript level , responses to reductions in dosages of X-chromosome genes have been found to be similar to those of autosomal genes 67 ., Thus , potential mechanisms for the higher tolerance are post-transcriptional compensatory mechanisms or selective alterations in gene composition ( changes in genomic locations ) , similar to those proposed for the observed demasculinization of the Drosophila X-chromosome 68 ., Prompted by the strong relationship between orchestration of the X- and 4th chromosomes by the MSL complex and POF system 2 , 14 , 69–71 , respectively , we also measured effects of roX suppression on chromosome 4 expression in roX mutants ., We observed weak but significant reduction of expression in the roX2 mutant , but the cause of this reduction remains elusive ., In roX2 mutant we also observed transcriptional upregulation of X-chromosome genes classified | Introduction, Results, Discussion, Material and methods | In Drosophila melanogaster , the male-specific lethal ( MSL ) complex plays a key role in dosage compensation by stimulating expression of male X-chromosome genes ., It consists of MSL proteins and two long noncoding RNAs , roX1 and roX2 , that are required for spreading of the complex on the chromosome and are redundant in the sense that loss of either does not affect male viability ., However , despite rapid evolution , both roX species are present in diverse Drosophilidae species , raising doubts about their full functional redundancy ., Thus , we have investigated consequences of deleting roX1 and/or roX2 to probe their specific roles and redundancies in D . melanogaster ., We have created a new mutant allele of roX2 and show that roX1 and roX2 have partly separable functions in dosage compensation ., In larvae , roX1 is the most abundant variant and the only variant present in the MSL complex when the complex is transmitted ( physically associated with the X-chromosome ) in mitosis ., Loss of roX1 results in reduced expression of the genes on the X-chromosome , while loss of roX2 leads to MSL-independent upregulation of genes with male-biased testis-specific transcription ., In roX1 roX2 mutant , gene expression is strongly reduced in a manner that is not related to proximity to high-affinity sites ., Our results suggest that high tolerance of mis-expression of the X-chromosome has evolved ., We propose that this may be a common property of sex-chromosomes , that dosage compensation is a stochastic process and its precision for each individual gene is regulated by the density of high-affinity sites in the locus . | In humans and fruit flies , females and males have different sets of sex chromosomes ., This causes gene dosage differences that must be compensated for by adjusting the expression of most genes located on the X-chromosome ., Long non-coding RNAs are central in this compensation and in fruit flies this is mediated by two non-coding RNAs , roX1 and roX2 which together with five proteins form the male-specific lethal complex ., The complex recognizes and upregulates gene transcription on the male X-chromosome ., While non-coding RNAs are are engaged in numerous biological processes and critical for compensation their precise functions remain elusive ., To understand the function of long non-coding RNAs we analysed the expression of all genes in roX1 , roX2 and roX1 roX2 mutants to explore the roles of long non-coding RNAs ., These mutants have different impacts on the genome-wide expression ., Our results also suggest that the X-chromosome is highly tolerant to mis-expression and we speculate that this tolerance evolved in parallel with compensation mechanisms and may be a common property of sex-chromosomes ., We propose that dosage compensation is a stochastic process that depends on the distribution of specific binding sites which will be selected for and optimized depending on the genes’ individual expression levels . | dosage compensation, invertebrates, molecular probe techniques, gene regulation, animals, long non-coding rnas, invertebrate genomics, animal models, developmental biology, drosophila melanogaster, model organisms, experimental organism systems, molecular biology techniques, drosophila, research and analysis methods, sex chromosomes, probe hybridization, chromosome biology, animal studies, gene expression, life cycles, x chromosomes, molecular biology, insects, rna hybridization, animal genomics, arthropoda, biochemistry, rna, eukaryota, cell biology, nucleic acids, genetics, biology and life sciences, genomics, non-coding rna, larvae, organisms, chromosomes | null |
journal.pgen.1004342 | 2,014 | Genome-Wide Inference of Ancestral Recombination Graphs | At each genomic position , orthologous DNA sequences drawn from one or more populations are related by a branching structure known as a genealogy 1 , 2 ., Historical recombination events lead to changes in these genealogies from one genomic position to the next , resulting in a correlation structure that is complex , analytically intractable , and poorly approximated by standard representations of high-dimensional data ., Over a period of many decades , these unique features of genetic data have inspired numerous innovative techniques for probabilistic modeling and statistical inference 3–9 , and , more recently , they have led to a variety of creative approaches that achieve computational tractability by operating on various summaries of the data 10–17 ., Nevertheless , none of these approaches fully captures the correlation structure of collections of DNA sequences , which inevitably leads to limitations in power , accuracy , and generality in genetic analysis ., In principle , the correlation structure of a collection of colinear orthologous sequences can be fully described by a network known as an ancestral recombination graph ( ARG ) 18–20 ., An ARG provides a record of all coalescence and recombination events since the divergence of the sequences under study and specifies a complete genealogy at each genomic position ( Figure 1A ) ., In many senses , the ARG is the ideal data structure for population genomic analysis ., Indeed , if an accurate ARG could be obtained , many problems of interest today—such as the estimation of recombination rates or ancestral effective population sizes—would become trivial , while many other problems—such as the estimation of population divergence times , rates of gene flow between populations , or the detection of selective sweeps—would be greatly simplified ., Various data representations in wide use today , including the site frequency spectrum , principle components , haplotype maps , and identity by descent spectra , can be thought of as low-dimensional summaries of the ARG and are strictly less informative ., An extension of the widely used coalescent framework 1 , 2 , 9 that includes recombination 21 is regarded as an adequately rich generative process for ARGs in most settings of interest ., While simulating an ARG under this model is fairly straightforward , however , using it to reconstruct an ARG from sequence data is notoriously difficult ., Furthermore , the data are generally only weakly informative about the ARG , so it is often desirable to regard it as a “nuisance” variable to be integrated out during statistical inference ( e . g . , 22 ) ., During the past two decades , various attempts have been made to perform explicit inference of ARGs using techniques such as importance sampling 19 , 22 ( see also 23 ) and Markov chain Monte Carlo sampling 24–27 ., There is also a considerable literature on heuristic or approximate methods for ARG reconstruction in a parsimony framework 28–35 ., Several of these approaches have shown promise , but they are generally highly computationally intensive and/or limited in accuracy , and they are not suitable for application to large-scale data sets ., As a result , explicit ARG inference is rarely used in applied population genomics ., The coalescent-with-recombination is conventionally described as a stochastic process in time 21 , but Wiuf and Hein 36 showed that it could be reformulated as a mathematically equivalent process along the genome sequence ., Unlike the process in time , this “sequential” process is not Markovian because long-range dependencies are induced by so-called “trapped” sequences ( genetic material nonancestral to the sample flanked by ancestral segments ) ., As a result , the full sequential process is complex and computationally expensive to manipulate ., Interestingly , however , simulation processes that simply disregard the non-Markovian features of the sequential process produce collections of sequences that are remarkably consistent in most respects with those generated by the full coalescent-with-recombination 37 , 38 ., In other words , the coalescent-with-recombination is almost Markovian , in the sense that the long-range correlations induced by trapped material are fairly weak and have a minimal impact on the data ., The original Markovian approximation to the full process 37 is known as the sequentially Markov coalescent ( SMC ) , and an extension that allows for an additional class of recombinations 38 is known as the SMC ., In recent years , the SMC has become favorite starting point for approximate methods for ARG inference 39–42 ., The key insight behind these methods is that , if the continuous state space for the Markov chain ( consisting of all possible genealogies ) is approximated by a moderately sized finite set—typically by enumerating tree topologies and/or discretizing time—then inference can be performed efficiently using well-known algorithms for hidden Markov models ( HMMs ) ., Perhaps the simplest and most elegant example of this approach is the pairwise sequentially Markov coalescent ( PSMC ) 42 , which applies to pairs of homologous chromosomes ( typically the two chromosomes in a diploid individual ) and is used to reconstruct a profile of effective population sizes over time ., In this case , there is only one possible tree topology and one coalescence event to consider at each genomic position , so it is sufficient to discretize time and allow for coalescence within any of possible time slices ., Using the resulting -state HMM , it is possible to perform inference integrating over all possible ARGs ., A similar HMM-based approach has been used to estimate ancestral effective population sizes and divergence times from individual representatives of a few closely related species 39–41 ., Because of their dependency on a complete characterization of the SMC state space , however , these methods can only be applied to small numbers of samples ., This limits their utility with newly emerging population genomic datasets and leads to reduced power for certain features of interest , such as recent effective population sizes , recombination rates , or local signatures of natural selection ., An alternative modeling approach , with better scaling properties , is the product of approximate conditionals ( PAC ) or “copying” model of Li and Stephens 43 ., The PAC model is motivated primarily by computational tractability and is not based on an explicit evolutionary model ., The model generates the th sequence in a collection by concatenating ( noisy ) copies of fragments of the previous sequences ., The source of each copied fragment represents the “closest” ( most recently diverged ) genome for that segment , and the noise process allows for mutations since the source and destination copies diverged ., The PAC framework has been widely used in many applications in statistical genetics , including recombination rate estimation , local ancestry inference , haplotype phasing , and genotype imputation ( e . g . , 44–48 ) , and it generally offers good performance at minimal computational cost ., Recently , Song and colleagues have generalized this framework to make use of conditional sampling distributions ( CSDs ) based on models closely related to , and in some cases equivalent to , the SMC 49–52 ., They have demonstrated improved accuracy in conditional likelihood calculations 49 , 50 and have shown that their methods can be effective in demographic inference 51 , 52 ., However , their approach avoids explicit ARG inference and therefore can only be used to characterize properties of the ARG that are directly determined by model parameters ( see Discussion ) ., In this paper , we introduce a new algorithm for ARG inference that combines many of the benefits of the small-sample SMC-based approaches and the large-sample CSD-based methods ., Like the PSMC , our algorithm requires no approximations beyond those of the SMC and a discretization of time , but it improves on the PSMC by allowing multiple genome sequences to be considered simultaneously ., The key idea of our approach is to sample an ARG of sequences conditional on an ARG of sequences , an operation we call “threading . ”, Using HMM-based methods , we can efficiently sample new threadings from the exact conditional distribution of interest ., By repeatedly removing and re-threading individual sequences , we obtain an efficient Gibbs sampler for ARGs ., This basic Gibbs sampler can be improved by including operations that rethread entire subtrees rather than individual sequences ., Our implementation of these methods , called ARGweaver , is efficient enough to sample full ARGs on a genome-wide scale for dozens of diploid individuals ., Simulation experiments indicate that ARGweaver converges rapidly and is able to recover many properties of the true ARG with good accuracy ., In addition , our explicit characterization of the ARG enables us to examine many features not directly described by model parameters , such as local times to most recent common ancestry , allele ages , and gene tree topologies ., These quantities , in turn , shed light on both demographic processes and the influence of natural selection across the genome ., For example , we demonstrate , by applying ARGweaver to 54 individual human sequences from Complete Genomics , that it provides insight into the sources of reduced nucleotide diversity near functional elements , the contribution of balancing selection to regions containing very old polymorphisms , and the relative influences of direct and indirect selection on allele age ., Our ARGweaver software ( https://github . com/mdrasmus/argweaver ) , our sampled ARGs ( http://compgen . bscb . cornell . edu/ARGweaver/CG_results ) , and genome-browser tracks summarizing these ARGs ( http://genome-mirror . bscb . cornell . edu; assembly hg19 ) are all freely available ., The starting point for our model is the Sequentially Markov Coalescent ( SMC ) introduced by McVean and Cardin 37 ., We begin by briefly reviewing the SMC and introducing notation that will be useful below in describing a general discretized version of this model ., The SMC is a stochastic process for generating a sequence of local trees , and corresponding genomic breakpoints , such that each describes the ancestry of a collection of sequences in a nonrecombining genomic interval , and each breakpoint between intervals and corresponds to a recombination event ( Figure 1B ) ., The model is continuous in both space and time , with each node in each having a real-valued age in generations ago , and each breakpoint falling in the continuous interval , where is the total length of the genomic segment of interest in nucleotide sites ., The intervals are exhaustive and nonoverlapping , with , , and for all ., Each is a binary tree with for all leaf nodes ., We will use the convention of indexing branches in the trees by their descendant nodes; that is , branch is the branch between node and its parent ., As shown by Wiuf and Hein 36 , the correlation structure of the local trees and recombinations under the full coalescent-with-recombination is complex ., The SMC approximates this distribution by assuming that is conditionally independent of given , and , similarly , that depends only on and , so that , ( 1 ) where is the effective population size , is the recombination rate , and it is understood that ., Thus , the SMC can be viewed as generating a sequence of local trees and corresponding breakpoints by a first-order Markov process ., The key to the model is to define the conditional distributions and such that this Markov process closely approximates the coalescent-with-recombination ., Briefly , this is accomplished by first sampling the initial tree from the standard coalescent and setting , and then iteratively, ( i ) determining the next breakpoint , , by incrementing by an exponential random variate with rate , where denotes the total branch length of ;, ( ii ) sampling a recombination point uniformly along the branches beneath the root of , where is a branch and is a time along that branch;, ( iii ) dissolving the branch above point ; and, ( iv ) allowing to rejoin the remainder of tree above time by the standard coalescent process , creating a new tree ( Figure 1B ) ., As a generative process for an arbitrary number of genomic segments , the SMC can be implemented by simply repeating the iterative process until then setting equal to and equal to ., Notice that , if the sampled recombination points are retained , this process generates not only a sequence of local trees but a complete ARG ., In addition , a sampled sequence of local trees , , is sufficient for generation of aligned DNA sequences corresponding to the leaves of the trees ( Figure 1C ) ., Augmented in this way , the SMC can be considered a full generative model for ARGs and sequence data ., We now define an approximation of the SMC that is discrete in both space and time , which we call the Discretized Sequentially Markov Coalescent ( DSMC ) ., The DSMC can be viewed as a generalization to multiple genomes of the discretized pairwise sequentially Markov coalescent ( PSMC ) used by Li and Durbin 42 ., It is also closely related to several other recently described discretized Markovian coalescent models 39 , 40 , 50 ., The DSMC assumes that time is partitioned into intervals , whose boundaries are given by a sequence of time points , with , for all ( ) , and equal to a user-specified maximum value ., ( See Table 1 for a key to the notation used in this paper . ), Every coalescence or recombination event is assumed to occur precisely at one of these time points ., Various strategies can be used to determine these time points ( see , e . g . , 50 ) ., In this paper , we simply distribute them uniformly on a logarithmic scale , so that the resolution of the discretization scheme is finest near the leaves of the ARG , where the density of events is expected to be greatest ( see Methods ) ., Each local block is assumed to have an integral length measured in base pairs , with all recombinations occurring between adjacent nucleotides ., The DSMC approaches the SMC as the number of intervals and the sequence length grow large , for fixed and ., Like the SMC , the DSMC generates an ARG for ( haploid ) sequences , each containing nucleotides ( Figure 1B ) ., In the discrete setting , it is convenient to define local trees and recombination events at the level of individual nucleotide positions ., Assuming that denotes a recombination between and , we write , with for positions and ., Notice that it is possible in this setting that and ., Where a recombination occurs ( ) , we write where is the branch in and is the time point of the recombination ., For simplicity and computational efficiency , we assume that at most one recombination occurs between each pair of adjacent sites ., Given the sparsity of variant sites in most data sets , this simplification is likely to have , at most , a minor effect during inference ( see Discussion ) ., Like the SMC , the DSMC can additionally be used to generate an alignment of DNA sequences ( Figure 1C ) ., We denote such an alignment by , where each represents an alignment column of height ., Each can be generated , in the ordinary way , by sampling an ancestral allele from an appropriate background distribution , and then allowing this allele to mutate stochastically along the branches of the corresponding local tree , in a branch-length-dependent manner ., We denote the induced conditional probability distribution over alignment columns by , where is the mutation rate ., In this work , we assume a Jukes-Cantor model 53 for nucleotide mutations along the branches of the tree , but another mutation model can easily be used instead ., Notice that , while the recombinations are required to define the ARG completely , the probability of the sequence data given the ARG depends only on the local trees ., In the case of an observed alignment , , and an unobserved ARG , , the DSMC can be viewed as a hidden Markov model ( HMM ) with a state space given by all possible local trees , transition probabilities given by expressions of the form , and emission probabilities given by the conditional distributions for alignment columns , ., The complete data likelihood function of this model—that is , the joint probability of an ARG and a sequence alignment given model parameters —can be expressed as a product of these terms over alignment positions ( see Methods for further details ) : ( 2 ) This HMM formulation is impractical as a framework for direct inference , however , because the set of possible local trees—and hence the state space—grows super-exponentially with ., Even with additional assumptions , similar approaches have only been able to accommodate small numbers of sequences 32 , 35 , 54 ., Instead , we use an alternative strategy with better scaling properties ., The key idea of our approach is to sample the ancestry of only one sequence at a time , while conditioning on the ancestry of the other sequences ., Repeated applications of this “threading” operation form the basis of a Markov chain Monte Carlo sampler that explores the posterior distribution of ARGs ., In essence , the threading operation adds one branch to each local tree in a manner that is consistent with the assumed recombination process and the observed data ( Figure 2 ) ., While conditioning on a given set of local trees introduces a number of technical challenges , the Markovian properties of the DSMC are retained in the threading problem , and it can be solved using standard dynamic programming algorithms for HMMs ., The threading problem can be precisely described as follows ., Assume we are given an ARG for sequences , , a corresponding data set , and a set of model parameters Assume further that is consistent with the assumptions of the DSMC ( for example , all of its recombination and coalescent events occur at time points in and it contains at most one recombination per position ) ., Finally , assume that we are given an th sequence , of the same length of the others , and let The threading problem is to sample a new ARG from the conditional distribution under the DSMC ., The problem is simplified by recognizing that can be defined by augmenting with the additional recombination and coalescence events required for the th sequence ., First , let be represented in terms of its local trees and recombination points: ., Now , observe that specifying the new coalescence events in is equivalent to adding one branch to each local tree , for , to obtain a new tree ( Figure 2 ) ., Let us denote the point at which each of these new branches attaches to the smaller subtree at each genomic position by , where indicates a branch in and indicates the coalescence time along that branch ., Thus , the coalescence threading of the th sequence is given by the sequence ., To complete the definition of , we must also specify the precise locations of the additional recombinations associated with the threading—that is , the specific time point at which each branch in a local tree was broken before the branch was allowed to re-coalesce in a new location in tree ., Here it is useful to partition the recombinations into those that are given by , denoted , and those new to , which we denote ( Figure 3A&B ) ., Each is either null ( ) , meaning that there is no new recombination between and , or defined by , where is a branch in and is the time along that branch at which the recombination occurred ., We call the recombination threading of the th sequence ., For reasons of efficiency , we take a two-step approach to threading: first , we sample the coalescence threading , and second , we sample the recombination threading conditional on ., This separation into two steps allows for a substantially reduced state space during the coalescence threading operation , leading to significant savings in computation ., When sampling the coalescence threading ( step one ) , we integrate over the locations of the new recombinations , as in previous work 42 , 50 ., Sampling the recombination threading ( step two ) can be accomplished in a straightforward manner independently for each recombination event , by taking advantage of the conditional independence structure of the DSMC model ( see Methods for details ) ., The core problem , then , is to accomplish step one by sampling the coalescence threading from the distribution , ( 3 ) where the notation indicates that random variable is held fixed ( “clamped” ) at a particular value throughout the procedure ., This equation defines a hidden Markov model with a state space given by the possible values of each , transition probabilities given by and emission probabilities given by ( Figure 3C ) ., Notice that the location of each new recombination , , is implicitly integrated out in the definition of ., Despite some unusual features of this model—for example , it has a heterogeneous state space and normalization structure along the sequence—its Markovian dependency structure is retained , and the problem of drawing a coalescent threading from the desired conditional distribution can be solved exactly by dynamic programming using the stochastic traceback algorithm for HMMs ., Additional optimizations allow this step to be completed in time linear in both the number of sequences and the alignment length and quadratic only in the number of time intervals ( see Methods for details ) ., The main value of the threading operation is in its usefulness as a building block for Markov chain Monte Carlo methods for sampling from an approximate posterior distribution over ARGs given the data ., We employ three main types of sampling algorithms based on threading , as described below ., We implemented these sampling strategies in a computer program called ARGweaver , that “weaves” together an ARG by repeated applications of the threading operation ., The program has subroutines for threading of both individual sequences and subtrees ., Options allow it to be run as a Gibbs sampler with single-sequence threading or a general Metropolis-Hastings sampler with subtree threading ., In either case , sequential sampling is used to obtain an initial ARG ., Options to the program specify the number of sampling iterations and the frequency with which samples are recorded ., The program is written in a combination of C++ and Python and is reasonably well optimized ., For example , it requires about 1 second to sample a threading of a single 1 Mb sequence in an ARG of 20 sequences with 20 time steps ., Our source code is freely available via GitHub ( https://github . com/mdrasmus/argweaver ) ., To summarize and visualize samples from the posterior distribution over ARGs , we use two main strategies ., First , we summarize the sampled ARGs in terms of the time to most recent common ancestor ( TMRCA ) and total branch length at each position along the genome ., We also consider the estimated age of the derived alleles at polymorphic sites , which we obtain by mapping the mutation to a branch in the local tree and calculating the average time for that branch ( see Methods ) ., We compute posterior mean and 95% credible intervals for each of these statistics per genomic position , and create genome browser tracks that allow these values to be visualized together with other genomic annotations ., Second , we developed a novel visualization device for ARGs called a “leaf trace . ”, A leaf trace contains a line for each haploid sequence in an analyzed data set ., These lines are ordered according to the local genealogy at each position in the genome , and the spacing between adjacent lines is proportional to their TMRCAs ( Figure S2 ) ., The lines are parallel in nonrecombining segments of the genome , and change in order or spacing where recombinations occur ., As a result , several features of interest are immediately evident from a leaf trace ., For example , recombination hot spots show up as regions with dense clusters of vertical lines , whereas recombination cold spots are indicated by long blocks of parallel lines ., Having demonstrated that ARGweaver was able to recover many features of simulated ARGs with reasonable accuracy , we turned to an analysis of real human genome sequences ., For this analysis we chose to focus on sequences for 54 unrelated individuals from the “69 genomes” data set from Complete Genomics ( http://www . completegenomics . com/public-data/69-Genomes ) 58 ., The 54 genome sequences were computationally phased using SHAPEIT v2 59 and were filtered in various ways to minimize the influence from alignment and genotype-calling errors ., They were partitioned into ∼2-Mb blocks and ARGweaver was applied to these blocks in parallel using the Extreme Science and Engineering Discovery Environment ( XSEDE ) ., For this analysis , we assumed generations , , and , implying ., We allowed for variation across loci in mutation and recombination rates ., For each ∼2-Mb block , we collected samples for 2 , 000 iterations of the sampler and retained every tenth sample , after an appropriate burn-in ( see Methods for complete details ) ., The entire procedure took ∼36 hours for each of the 1 , 376 2-Mb blocks , or 5 . 7 CPU-years of total compute time ., The sampled ARGs were summarized by UCSC Genome Browser tracks describing site-specific times to most recent common ancestry ( TMRCA ) , total branch length , allele ages , leaf traces , and other features across the human genome ., These tracks are publicly available from our local mirror of the UCSC Genome Browser ( http://genome-mirror . bscb . cornell . edu , assembly hg19 ) ., Several decades have passed since investigators first worked out the general statistical characteristics of population samples of genetic markers in the presence of recombination 21 , 80–83 ., Nevertheless , solutions to the problem of explicitly characterizing this structure in the general case of multiple markers and multiple sequences—that is , of making direct inferences about the ancestral recombination graph ( ARG ) 19 , 20—have been elusive ., Recent investigations have led to important progress on this problem based on the Sequentially Markov Coalescent ( SMC ) 17 , 37–42 , but existing methods are still either restricted to small numbers of sequences or require severe approximations ., In this paper , we introduce a method that is faithful to the SMC yet has much better scaling properties than previous methods ., These properties depend on a novel “threading” operation that can be performed in a highly efficient manner using hidden Markov modeling techniques ., Inference does require the use of Markov chain Monte Carlo ( MCMC ) sampling , which has certain costs , but we have shown that the sampler mixes fairly well and converges rapidly , particularly if the threading operation is generalized from single sequences to subtrees ., Our methods allow explicit statistical inference of ARGs on the scale of complete mammalian genomes for the first time ., Furthermore , the sampling of ARGs from their posterior distribution has the important advantage of allowing estimation of any ARG-derived quantity , such as times to most recent common ancestry , allele ages , or regions of identity by descent ., Despite our different starting point , our methods are similar in several respects to the conditional sampling distribution ( CSD ) -based methods of Song and colleagues 49–52 ., Both approaches consider a conditional distribution for the th sequence given the previous sequences , and in both cases a discretized SMC is exploited for efficiency of inference ., However , the CSD-based methods consider the marginal distribution of the th sequence only given the other sequences and never explicitly reconstruct an ARG , while ours considers the joint distribution of an ARG of size and the th sequence , given an ARG of size and the previous sequences ., In a sense , we have employed a “data augmentation” strategy by explicitly representing full ARGs in our inference procedure ., The main cost of this strategy is that it requires Markov chain Monte Carlo methods for inference , rather than allowing direct likelihood calculations and maximum-likelihood parameter estimation ., The main benefit is that it provides an approximate posterior distribution over complete ARGs and many derived quantities , including times to most recent common ancestry , allele ages , and distributions of coalescence times ., By contrast , the CSD-based methods provide information about only those properties of the ARG that are directly described by the model parameters ., We view these two approaches as complementary and expect that they will have somewhat different strengths and weaknesses , depending on the application in question ., Our explicit characterization of genealogies can be exploited to characterize the influence of natural selection across the genome , as shown in our analysis of the Complete Genomics data set ., In particular , we see clear evidence of an enrichment for ancient TMRCAs in regions of known and predicted balancing selection , reduced TMRCAs near protein-coding genes and selective sweeps , and reduced allele ages in sites experiencing both direct selection and selection at closely linked sites ., Interestingly , the genealogical view appears to have the potential to shed light on the difficult problem of distinguishing between background selection and hitchhiking ., Our initial attempt at addressing this problem relies on a genealogy-based summary statistics , the relative TMRCA halflife ( RTH ) , that does appear to distinguish effectively between protein-coding genes and partial selective sweeps identified by iHS ., However , more work will be needed to determine how well this approach generalizes to other types of hitchhiking ( e . g . , complete sweeps , soft sweeps , recurrent sweeps ) and whether additional genealogical information can be used to characterize the mode of selection more precisely ., Additional work is also needed to determine whether our ARG-based allele-age estimator—which is highly informative in bulk statistical comparisons but has high variance at individual sites—can be used to improve functional and evolutionary characterizations of particular genomic loci ., A related challenge is to see whether our genome-wide ARG samples can be used to improve methods for association/LD mapping ( see 34 , 84–88 ) ., In addition to natural selection , our methods for ARG inference have the potential to shed light on historical demographic processes , an area of particular interest in the recent literature 16 , 17 , 51 , 52 , 89 ., To explore the usefulness of ARGweaver in demography inference , we attempted to infer a population phylogeny with admixture edges for the 11 human populations represented in the Complete Genomics data set , based on the genealogies sampled under our naive ( panmictic ) prior distribution ., We extracted 2 , 304 widely spaced loci from our inferred ARGs , obtained a consensus tree at each locus , and reduced this tree to a subtree with one randomly selected chromosome for each of the 11 populations ( see Text S1 for details ) ., We then analyzed these 11-leaf trees with the PhyloNet program 90 , 91 , which finds a population tree that minimizes the number of “deep coalescences” required for reconciliation with a given set of local trees , allowing for both incomplete lineage sorting and hybridization ( admixture ) events between groups ., PhyloNet recovered the expected phylogeny for these populations in the absence of hybridization and generally detected complex patterns of gene flow where they are believed to have occurred , but it had difficulty reconstructing the precise relatinonships among source and admixed populations ( Supplementary Figure S20 ) ., These experiments suggested that the posterior distribution of ARGs does appear to contain useful information about population structure even when a noninformative prior distribution is used , but tha | Introduction, Results, Discussion, Methods | The complex correlation structure of a collection of orthologous DNA sequences is uniquely captured by the “ancestral recombination graph” ( ARG ) , a complete record of coalescence and recombination events in the history of the sample ., However , existing methods for ARG inference are computationally intensive , highly approximate , or limited to small numbers of sequences , and , as a consequence , explicit ARG inference is rarely used in applied population genomics ., Here , we introduce a new algorithm for ARG inference that is efficient enough to apply to dozens of complete mammalian genomes ., The key idea of our approach is to sample an ARG of chromosomes conditional on an ARG of chromosomes , an operation we call “threading . ”, Using techniques based on hidden Markov models , we can perform this threading operation exactly , up to the assumptions of the sequentially Markov coalescent and a discretization of time ., An extension allows for threading of subtrees instead of individual sequences ., Repeated application of these threading operations results in highly efficient Markov chain Monte Carlo samplers for ARGs ., We have implemented these methods in a computer program called ARGweaver ., Experiments with simulated data indicate that ARGweaver converges rapidly to the posterior distribution over ARGs and is effective in recovering various features of the ARG for dozens of sequences generated under realistic parameters for human populations ., In applications of ARGweaver to 54 human genome sequences from Complete Genomics , we find clear signatures of natural selection , including regions of unusually ancient ancestry associated with balancing selection and reductions in allele age in sites under directional selection ., The patterns we observe near protein-coding genes are consistent with a primary influence from background selection rather than hitchhiking , although we cannot rule out a contribution from recurrent selective sweeps . | The unusual and complex correlation structure of population samples of genetic sequences presents a fundamental statistical challenge that pervades nearly all areas of population genetics ., Historical recombination events produce an intricate network of intertwined genealogies , which impedes demography inference , the detection of natural selection , association mapping , and other applications ., It is possible to capture these complex relationships using a representation called the ancestral recombination graph ( ARG ) , which provides a complete description of coalescence and recombination events in the history of the sample ., However , previous methods for ARG inference have not been adequately fast and accurate for practical use with large-scale genomic sequence data ., In this article , we introduce a new algorithm for ARG inference that has vastly improved scaling properties ., Our algorithm is implemented in a computer program called ARGweaver , which is fast enough to be applied to sequences megabases in length ., With the aid of a large computer cluster , ARGweaver can be used to sample full ARGs for entire mammalian genome sequences ., We show that ARGweaver performs well in simulation experiments and demonstrate that it can be used to provide new insights about both demographic processes and natural selection when applied to real human genome sequence data . | genome scans, neutral theory, population genetics, mutation, mathematics, statistics (mathematics), effective population size, genome analysis, biostatistics, ecological metrics, genomics, genetic polymorphism, evolutionary modeling, genetic drift, haplotypes, population size, ecology, natural selection, genetics, biology and life sciences, gene flow, computational biology, evolutionary biology, physical sciences, evolutionary processes, human genetics | null |
journal.ppat.1004571 | 2,015 | Secreted Herpes Simplex Virus-2 Glycoprotein G Modifies NGF-TrkA Signaling to Attract Free Nerve Endings to the Site of Infection | Herpes simplex virus type 1 and 2 ( HSV-1 and HSV-2 , respectively ) are highly prevalent , neurotropic human pathogens 1 ., Initial infection occurs in epithelial cells , generally within the skin and the mucosa of the oral tract and genitalia 1 ., Then , HSV reaches and infects free nerve endings ( FNE ) of sensory neurons and colonizes ganglia of the Peripheral Nervous System ( PNS ) ., The mechanism ( s ) facilitating HSV neurotropism , which is crucial for latency and pathogenesis , are not well understood ., Since herpesviruses are highly adapted pathogens that modify several aspects of both the immune and nervous systems , it is conceivable that they may modulate factors influencing neuronal functions to gain access to the nervous system ., Several axonal guidance cues and neurotrophic factors involved in neural targeting have been identified 2 ., Among them , neurotrophins are a family of secreted proteins that play relevant roles in neuronal survival , axonal growth and guidance in the PNS ., Members of this family include nerve growth factor ( NGF ) , brain-derived neurotrophic factor ( BDNF ) , neurotrophin 3 ( NT3 ) and NT4/5 3 ., Each neurotrophin binds with high affinity and activates tyrosine kinase receptors known as Trks ., NGF binds TrkA , BDNF and NT4/5 bind TrkB , and NT3 binds TrkC ., Moreover , NT3 can also bind TrkA and TrkB , although with lower affinity 3 ., Both mature neurotrophins and immature precursors ( proneurotrophins ) also bind p75 neurotrophin receptor ( p75NTR ) , a member of the tumor necrosis factor ( TNF ) receptor superfamily ., p75NTR has multiple and diverse functions 4 ., Another important family of neurotrophic factors is the glial cell line-derived neurotrophic factors ( GDNF ) family ligands ( GFLs ) formed by GDNF and artemin among others ., GFLs interact with co-receptors of the GDNF Family Receptor α ( GFRα ) protein family , allowing the activation of the tyrosine kinase receptor RET ( rearranged during transfection ) 5 ., Peripheral neurons innervating skin and mucosa show a strong dependency on neurotrophic factors both ex vivo and in vivo 6 , 7 ., In order to colonize the PNS , HSV must reach FNE , dynamic structures capable of degeneration and regeneration 8 in response to neurotrophic factors 6 , 7 ., The possible relevance of neurotrophic factors in the initial steps of HSV infection in neurons is not completely understood ., We hypothesized that HSV could modify nerve ending navigational cues during the early steps of PNS colonization ., HSV glycoprotein G ( gG ) is the least conserved of the glycoproteins shared by HSV-1 and HSV-2 9 ., HSV-1 gG ( gG1 ) and HSV-2 gG ( gG2 ) have a similar C-terminal domain present at the virion and at the surface of the infected cells , composed by an extracellular proline-rich domain , a transmembrane region and a short cytoplasmic tail ., The N-terminal domain of HSV-2 gG , but not that of gG1 , is proteolytically cleaved and secreted ( termed here SgG2 ) during infection 10 , 11 ., Whether these differences between gG1 and gG2 have any functional consequence is unknown ., Recombinant soluble gG1 ( SgG1 ) and SgG2 bind chemokines with high affinity enhancing chemokine function 12 , in sharp contrast to all previously described viral chemokine-binding proteins ( vCKBPs ) that inhibit chemotaxis ., Since neurotrophic factors are also secreted proteins regulating many aspects of peripheral neurons , and have been previously involved in immune related-functions 13 , we investigated whether HSV gG could bind neurotrophic factors modifying their activity ., Here we show that SgG2 interacts with several neurotrophic factors ., SgG2 , but not SgG1 or M3 , a viral chemokine binding protein from murine gamma herpesvirus 68 ( MHV-68 ) 14 , transiently enhances NGF-dependent axonal growth of superior cervical ganglion ( SCG ) neurons ex vivo ., The molecular mechanism beneath this enhancement involves modulation of the NGF receptor TrkA ., SgG2 alters TrkA localization in lipid rafts , and NGF-dependent TrkA internalization , signaling and retrograde transport ., In vivo , both infection with HSV-2 and expression of SgG2 in the external layers of the epidermis modifies the termination zone of the TrkA dependent FNE ., Our data indicate that SgG2 may reconfigure neurotrophin signaling during HSV-2 primary infection to attract specific terminal axons to the infection site and may facilitate neural invasion ., To test whether gG from HSV-1 or HSV-2 interacts with neurotrophic factors we performed surface plasmon resonance ( SPR ) assays ., As a control we used another vCKBP from MHV-68 , M3 ., Recombinant SgG2 interacted with members of the neurotrophin family such as NGF and those of the GFLs family , like artemin or GDNF ( Fig . 1A and S1 Table ) ., SgG1 and M3 bound neurotrophic factors but saturable binding was not demonstrated in these SPR experiments and therefore non-specific binding cannot be excluded ( S1A Fig . and S1 Table ) ., However , binding specificity was suggested by the lack of interaction of the vCKBPs tested with interferon ( IFN ) -α , TNF-α , and interleukin ( IL ) -1 , and the negative binding of M3 to artemin ( Fig . 1B , S1B Fig . and S1 Table ) ., Saturation experiments using SPR further demonstrated the specific , high affinity interaction of SgG2 and NGF ( Fig . 1 D and S1 Table ) ., We could also show binding of SgG1 and SgG2 to NGF coupled in a chip ( S1C Fig . and Fig . 1C , respectively ) , with a similar affinity ( KD 1 . 4 x 10−8 M ) to that calculated for NGF interacting with SgG2 coupled in a chip based on kinetics analysis ( not shown ) , supporting the specificity of the interaction ., HSV-2 glycoprotein D ( gD ) was used as a negative control for NGF binding ( Fig . 1C ) ., We confirmed binding of these vCKBPs to NGF by crosslinking assays using iodinated rat NGF ( 125I-rNGF ) and soluble , recombinant M3 , SgG1 and SgG2 ( S1D Fig . and S1 Text ) ., The formation of the vCKBP-NGF complex was competitively inhibited with increasing concentrations of unlabeled NGF ( S1E , F Fig . ) ., Addition of increasing amounts of cold NGF resulted in the formation of higher molecular weight bands that probably correspond to NGF oligomers since they appear also in the absence of viral proteins ., The formation of NGF oligomers at high concentration may affect the pattern of binding to the viral proteins in our SPR assays ., Since HSV-1 and HSV-2 gG also bind chemokines , we tested whether NGF and chemokine binding takes place through different regions or whether the binding sites overlap and the interaction of NGF may be competitively inhibited with chemokines ., We injected CXCL12β or NGF alone or in combination with NGF in a chip containing SgG2 ., The binding detected after simultaneous injection of chemokine and NGF nearly corresponded to the sum of the binding obtained when the chemokine and NGF were injected independently ( Fig . 1D ) ., Furthermore , injection of one of the analytes ( NGF or CXCL12β ) in an SgG2-coupled chip preincubated with the other analyte ( CXCL12β or NGF ) caused no displacement of the prebound analyte ( not shown ) ., It is important to note that the affinity of the interaction of SgG2 with CXCL12β ( 2 . 2x10−9 M ) 12 is higher than that determined for NGF ( 2 . 2x10−8 M ) ., We addressed whether the vCKBP could affect the function of neurotrophic factors ., FNE of sensory neurons are the main targets of HSV-1 and HSV-2 in the skin and mucosa ., We focused on NGF and artemin due to their relevance in epidermal homeostasis and innervation 6 , 7 ., In order to have a feasible and robust model we used mouse sympathetic neurons from SCG that express high levels of NGF and artemin receptors , and depend on NGF and artemin both in vivo and in culture 15 ., We cultured SCG as 3D explants using collagen matrix containing recombinant viral proteins , and we measured the area comprised by axons normalizing to the ganglion perimeter ., In the presence of NGF , SgG2 enhanced axonal growth of SCG neurons compared to HEPES control ( Fig . 2A , middle panels ) ., On the contrary , M3 ( Fig . 2A , middle panels ) did not induce such increase ., No changes in axonal growth induced by SgG2 or M3 were detected in the absence of trophic factors ( Fig . 2A , upper panels ) , or when artemin was used ( Fig . 2A , lower panels ) ., Next , we addressed whether SgG1 had the same effect on NGF as SgG2 ., SgG1 did not significantly increase axonal growth of SCG neurons 24 h post-incubation ( Fig . 2B ) ., SgG2 enhancement of NGF-mediated axonal growth was also observed when using dissociated SCG neurons ( Fig . 2C ) ., To further characterize this effect and to discriminate between axonal growth and directionality , transfected HEK-293T cells were co-cultured with SCG in 3D explants with NGF , and the proximal/distal ( P/D ) ratio was analyzed ., HEK-293T cells express endogenous axonal repulsive cues like semaphorin 3A 16 and several class-A ephrins 17 ., Accordingly , using low concentrations of NGF ( 0 . 25 nM ) , control transfected HEK-293T cells induced repulsion of SCG axons ( Fig . 2D ) ., Such repulsion was also observed when HEK-293T cells were transfected with a V5-tagged M3-expressing plasmid ( V5-M3 ) ., On the contrary , the expression of V5-SgG2 significantly reduced the repulsion of SCG axons ( Fig . 2D middle panel ) ., Viral protein expression was detected by Western blot ( Fig . 2D ) ., Altogether , these data showed that SgG2 specifically increases axonal growth of SCG neurons in an NGF-dependent manner ., We addressed whether the increase in NGF-dependent axonal growth mediated by SgG2 was related to changes in TrkA ., Since SgG2 bound NGF and modulated its function we decided to address whether it could interact with TrkA ., We incubated dissociated SCG neurons with NGF in the presence or absence of SgG2 ., Following immunoprecipitation of TrkA , SgG2 was detected only when NGF was present , indicating that SgG2 and TrkA belong to the same complex in the presence of NGF and the interaction of SgG2 with NGF is required ( Fig . 3A ) ., Then we focused on TrkA signaling ., Dissociated cultures of SCG neurons were incubated with NGF , M3 or SgG2 alone or NGF together with the viral proteins , and the downstream signaling was analyzed by Western blotting ., Incubation with SgG2 or M3 in the absence of NGF did not induce TrkA phosphorylation or activation of downstream pathways ( Fig . 3B–F ) ., As expected , NGF promoted TrkA phosphorylation 3 ., SgG2 significantly increased NGF-dependent TrkA phosphorylation at tyrosine 490 ( Tyr490 ) at 15 min , but not at 120 min post-incubation , when compared to control and M3 conditions ( Fig . 3B , C ) ., Extracellular signal-regulated kinases ( ERK ) 1/2 activation in response to NGF , but not that of AKT ( also known as protein kinase B ) , was significantly increased in the presence of SgG2 both at 15 and 120 min post-incubation ( Fig . 3B , D , E ) ., We also tested the activation status of cofilin , an actin-severing protein responsible of actin turnover that is inactivated by phosphorylation 18 ., We detected higher levels of cofilin phosphorylation induced by SgG2-NGF 120 min post-NGF stimulation when compared to NGF alone ( Fig . 3B , F ) ., At the cell surface there are different and segregated types of rafts depending on their lipid and glycosphingolipid composition , leading to diverse biological functions 19 ., A fraction of TrkA accumulates in GM1 lipid rafts , and TrkA appears to signal through GM1 rafts 20 , 21 ., We addressed whether SgG2 could affect TrkA recruitment to GM1-enriched lipid rafts in SCG dissociated neurons ., In the absence of NGF , a small fraction of TrkA co-localized with GM1 staining ( Fig . 4A , top row ) ., As predicted , NGF stimulation maintained and increased TrkA co-localization with GM1 staining 2 min post NGF stimulation ( Fig . 4A , third row ) ., SgG2 disrupted NGF-dependent and-independent TrkA localization in GM1 rafts at 2 ( Fig . 4A , fourth row ) and 10 min post-stimulation ( Fig . 4B , fourth row ) ., We hypothesized that SgG2 could divert TrkA to other types of rafts such as GM3 rafts 19 ., Stimulation with NGF reduced the localization of TrkA in GM3-rich rafts ( Fig . 4C , D , third row ) concomitantly with NGF-dependent TrkA increase in GM1 rafts , when compared to HEPES treatment ., However , addition of SgG2 retained TrkA within GM3 rafts at 2 min post-incubation ( Fig . 4C , fourth row ) and even more at 10 min post-incubation ( Fig . 4D , fourth row ) ., Overall , our results indicated that SgG2 recruits TrkA to GM3-enriched lipid rafts , thereby altering the normal NGF-TrkA signaling that usually occurs within GM1 rafts ., The NGF receptor p75NTR is present in GM1 rafts 20 and interacts with TrkA in an NGF-dependent manner 3 , 22 ., We addressed the effect of SgG2 on the interaction between TrkA and p75NTR ., As shown in S2 Fig ., , SgG2 disrupts the NGF-induced TrkA-p75NTR interaction in agreement with TrkA raft relocation mediated by SgG2 , whereas SgG1 did not ., As TrkA recruitment to GM3 appeared to be independent of the presence of NGF in the culture we wanted to determine whether SgG2 alone was sufficient to induce changes in TrkA signaling ., To explore this , we preincubated dissociated SCG neurons with HEPES or with SgG2 for 10 min followed by NGF stimulation in the presence or absence of SgG2 ., Preincubation with SgG2 did not affect NGF dependent TrkA or ERK phosphorylation ( S3A–C Fig . ) and , in agreement with Fig . 3B , changes on signaling were detected only when SgG2 was added together with NGF ., These results indicated that SgG2 induces TrkA raft relocalization that may be necessary , but not sufficient , to enhance NGF-TrkA signaling ., Careful examination of the results obtained following 10 min of NGF incubation ( Fig . 4B , D ) suggested a possible effect of SgG2 at the level of plasma membrane TrkA ., To confirm this observation we analyzed the internalization rate of TrkA in response to NGF and SgG2 ., Similar amounts of TrkA were detected in almost every experimental condition in the absence of NGF ( Fig . 5A , time 0 min ) ., Addition of 1 nM NGF promoted maximal TrkA internalization at 15 min post NGF incubation in both control ( HEPES ) and M3 conditions ( Fig . 5A ) ., However , when SgG2 and NGF were added simultaneously we observed higher levels of TrkA at the plasma membrane ( Fig . 5A ) ., Following 120 min of NGF exposure , this difference was maintained although TrkA at the neuronal surface increased both in control and M3 conditions , probably due to receptor recycling ( Fig . 5A ) ., In vivo , neurotrophins are secreted by specific tissues and only the axon terminals are directly exposed to them 23 ., For a suitable signaling on neurons to occur , neurotrophins must be transported in a retrograde manner from distal axons to their cell bodies ., TrkA internalization seems to be required for TrkA retrograde transport 23 ., Cofilin is a critical mediator of TrkA retrograde transport 18 ., Stimulation of SCG neurons with NGF and SgG2 induced significant higher levels of cofilin phosphorylation , resulting in its inactivation 24 , at late time points ( Fig . 3 ) ., We addressed the effect of SgG2 on TrkA retrograde transport by monitoring the localization of phosphorylated TrkA ( p-TrkA ) on SCG neurons , grown in microfluidic devices , whose axon terminals were incubated with NGF alone or in combination with vCKBPs during 120 min ., We found very low levels of p-TrkA in the distal axons or cell bodies of neurons not exposed to NGF ( Fig . 5B , upper row ) ., As expected , exposure of distal axons to NGF during 120 min promoted a moderate increase on p-TrkA staining in distal axons and a high increase of p-TrkA staining in the cell bodies ( Fig . 5B , second row ) ., Similar results were obtained when distal axons were exposed to NGF and M3 ( Fig . 5B , fourth row ) ., Exposure of distal axons to NGF and SgG2 induced a significant accumulation of p-TrkA staining in distal axons whereas low level staining was present in the cell bodies ( Fig . 5B , third row and graph ) ., Moreover , the growth cone had a normal morphology following 120 min exposure to NGF in both control and M3 conditions , with large actin-rich filopodial protrusions ., The presence of SgG2 severely disturbed growth cone morphology , presenting only few filopodia ( NGF-HEPES: 88 , 8% spread growth cones , NGF-M3 94 , 7% spread growth cones and NGF-SgG2: 6 , 2% spread growth cones; P = 0 , 00303681 , P = 0 , 00052529 , respectively ) ., Altogether , these data indicated that SgG2 partially blocks NGF-dependent TrkA endocytosis and retrograde transport with p-TrkA accumulating at the distal axon in the presence of NGF-SgG2 ., Also , the actin cytoskeleton is perturbed ., We hypothesized that SgG2-modification of NGF function could attract axons to infected sites in vivo ., NGF is secreted by keratinocytes in the skin 6 ., In adult mice glabrous skin there are two morphologically different intra-epidermal FNE , with distinct termination zones ( 25 , Fig . 6A ) :, ( i ) FNE that reach and meander through the stratum granulosum , an external layer of epidermis ( Fig . 6A , solid arrowhead ) , most of them being non-peptidergic and expressing RET; and, ( ii ) FNE with straight trajectory that remain in the inner stratum basale or stratum spinosum ( Fig . 6A , open arrowhead ) , most of them being from peptidergic TrkA+ neurons 25–27 ., Since SgG2 specifically modifies NGF-dependent TrkA+ signaling , we analyzed peptidergic FNE in vivo during HSV-1 or HSV-2 infection ., We topically applied PBS , HSV-1 or HSV-2 onto mice hindpaw ( glabrous skin ) following mild skin exfoliation ., We did not detect any macroscopical difference in the skin between the different conditions 48 h post-infection ., We detected HSV infection with an anti-gBgD antibody and the peptidergic FNE using an anti-calcitonin gene related peptide ( CGRP ) antibody ., The number of peptidergic FNE was significantly reduced in skin infected with both HSV-1 and HSV-2 ( Fig . 6B ) ., Most peptidergic FNE remained in the stratum basale or in the stratum spinosum in PBS-treated skin , as described 25 ( Fig . 6C left panel , 6D left panel ) ., HSV-1 infection promoted a subtle , although not significant , change in the trend of the termination zone of the remaining peptidergic FNE ( Fig . 6C , right panel ) ., However , following HSV-2 infection nearly half of the remaining peptidergic FNE increased in length , reaching the stratum granulosum ( Fig . 6D , right panel ) ., We hypothesized that changes in straight peptidergic FNE termination zone induced by HSV-2 , but not by HSV-1 , could be mediated by SgG2 through its interaction with NGF ., We transfected mice hindpaw skin with empty vector or viral cDNA coding for V5-SgG2 or V5-M3 ., Transfection did not affect epidermal layers , general nerve arrangement or the number of peptidergic FNE per field in all the conditions tested ., In vector-transfected skin , nearly all straight trajectory-peptidergic FNE remained in the stratum basale or in the stratum spinosum , ( Fig . 6E , left panel , 6F left panel ) ., However , in the fields where V5-SgG2 expression was detected , around 20% of straight trajectory-peptidergic FNE showed increased length , reaching the stratum granulosum ( Fig . 6E , right panel ) ., This phenomenon was not detected when V5-M3 was expressed ( Fig . 6F , right panel ) ., Altogether these data suggested that HSV-2 , but not HSV-1 , modifies the termination zone of the straight peptidergic FNE and that SgG2 may account , at least partially , for this phenomenon ., Neurotropism is a major evolutionary advantage for HSV ., The infection of neuronal FNE permits HSV to establish latency in sensory ganglia and persist for the lifetime of the individual ., We have detected that infection of glabrous skin with HSV-1 and HSV-2 causes a reduction in the number of peptidergic FNE ., This could be due to the secretion of neurotoxic molecules following infection , such as IL-1β and TNF-α 28 , 29 ., HSV may have developed molecular mechanisms to counteract this toxicity ., The presence , plasticity and topology of FNE during development and adult remodeling depend on axonal navigational cues and trophic factors like neurotrophins ., Our results show that HSV-2 SgG specifically binds to several neurotrophic factors from the neurotrophin family ( NGF , BDNF and NT3 ) and the GFL family ( GDNF and artemin ) ., Interactions of HSV-1 gG and MHV-68 M3 with neurotrophic factors were also observed but we cannot rule out the possibility that these interactions are non-specific ., Further research may identify biological relevance for these interactions in the biology of HSV-1 and MHV-68 ., Only the interaction between SgG2 and NGF is biologically relevant in our experimental model resulting in an increase in FNE growth of TrkA+ neurons ., This constitutes the first description , to our knowledge , of a protein expressed by a human pathogen with the ability to modulate neurotrophic factors and we hypothesize that this interaction may contribute to HSV neurotropism ., To test this hypothesis we focused our functional studies on neurotrophic factors regulating innervation of the epidermis and that are important in skin homeostasis and inflammation , like NGF and artemin 6 , 7 , 13 ., Using SCG explants and dissociated cultures , only SgG2 , but not SgG1 or M3 , modified NGF-dependent axonal growth ex vivo ., In vivo only SgG2 modified the growth and termination site of the TrkA+ FNE , corresponding to peptidergic neurons ., Importantly , similar results were observed when infecting the skin with HSV-2 but not with HSV-1 ., One of the most intriguing results of our work is the difference between HSV-1 and HSV-2 ., HSV-1 seroprevalence is higher than that of HSV-2 30–32 ., HSV-1 is normally transmitted during childhood and is linked to facial herpes whereas HSV-2 is normally sexually transmitted and associated with genital herpes ., However , both viruses can infect these two anatomical areas , and genital herpes due to HSV-1 is increasing 31 , 33 ., Of note , HSV-1 reactivates in the genital tract less frequently than HSV-2 34 ., Therefore , differences in prevalence , transmission route and recurrent disease between HSV-1 and HSV-2 in the oro-labial or genital area exist ., The lack of effect with HSV-1 in our in vivo experiments does not rule out the possibility that HSV-1 may modulate neurotrophic factors in another setting or utilizes a different mechanism to modulate axonal growth ., In this regard , HSV-1 latency associated transcript induces axonal regeneration and growth in a post-entry phase following serum deprivation 35 or NGF starvation 36 , and this could facilitate release of HSV-1 into the peripheral tissue following anterograde transport from the neuronal cell body 36 ., From a molecular perspective , the different activity of gG1 and gG2 could be due to their structure and location ., HSV-1 gG is a transmembrane protein that remains anchored at the surface of infected cells and the virion envelope whereas gG2 is proteolytically processed secreting an N-terminal domain , SgG2 , with the potential of reaching neighboring cells ., Mature peptidergic neurons express TrkA ., These type of FNE represent 40% of the total FNE in glabrous skin 25 , but they are the predominant population in mucosa and in the internal organs 37 , 38 ., The fact that HSV-2 displays a specific molecular mechanism involved in attracting peptidergic FNE points to possible implications on a more efficient colonization of different anatomical niches or subsets of neurons ., In this regard , recent evidence indicates that sensory A5+ neurons ( corresponding to CGRP+/TrkA+ neurons ) support HSV-2 productive infection while they are non-permissive for HSV-1 productive infection in vitro 39 ., This difference in susceptibility is dependent on the latency associated transcript 40 ., Whether the lack of HSV-1 effect on TrkA+ FNE shown here may influence A5+ neuron susceptibility requires further investigation ., The genitalia are enriched in TrkA-expressing peptidergic innervation 41 , 42 ., For instance , TrkA+ projecting neurons rise up to 60% from first sacral ganglia to rat penis 43 ., We propose that SgG2 may facilitate the infection of FNE innervating the genitalia and/or subsequent spread ., In agreement with this hypothesis , pharmacological modulation of capsaicin-sensitive peptidergic neurons reduces the severity of cutaneous HSV-2 genital infections both in female and male pigs 44 ., Alternatively , infection of peptidergic neurons may be relevant for nociception ., In summary , we cannot conclude from our study that HSV-2 SgG-mediated modulation of NGF is essential for the infection of FNE and subsequent colonization of the nervous system , but our data lead us to propose that SgG2 may facilitate the infection of specific subsets of neurons and this may have consequences for transmission and disease ., Deletion or disruption of us4 , the gene encoding gG , in HSV-1 results in partial attenuation in mouse models of infection 45–47 ., There are no reports analyzing the role of gG2 in vivo using animal models ., Transient enhancement of NGF-mediated axonal growth by SgG2 involves several related events affecting TrkA localization and trafficking ., SgG2 alone modifies TrkA association with gangliosides , increasing its relative presence in GM3- versus GM1-rich rafts ., This effect could be mediated by the SgG2 ability of binding glycosaminoglycans ( N . M . -M . , A . V . -B . and A . A . submitted manuscript ) ) ., However SgG2 mediated TrkA recruitment to GM3 is not sufficient to induce changes in TrkA signaling by itself ., SgG2 must be bound to NGF to promote the described changes as previously observed for SgG2 enhancement of chemokine function 12 ., GM1 is the preferential ganglioside for canonical NGF-TrkA signaling 21 ., GM1 is a marker of caveolae and caveolae-like membranes 20 ., The SgG2-mediated TrkA exclusion from GM1 rafts may influence the function of TrkA and could explain the reduced TrkA-p75NTR interaction upon NGF stimulation since p75NTR is located mainly within caveolae and caveolae like membranes 20 ., TrkA-p75NTR interaction appears to be required for NGF-TrkA endocytosis in some models 22 and could account at least partially for the reduced TrkA endocytosis detected in the presence of SgG2 during NGF stimulation ., The fact that TrkA is a raft resident protein 48 , whereas RET remains outside lipid rafts and is transiently recruited to rafts upon binding to GFLs in cis 49 , 50 , could explain the lack of SgG2 effect on Artemin function ., Since TrkA endocytosis is a pre-requisite for its retrograde transport 18 , 51 , this could explain the blockage of NGF-mediated TrkA retrograde transport caused by SgG2 ., Finally , accumulation of p-TrkA in distal axons mediated by NGF-SgG2 could induce a local increase in axon length as proposed for NT-3 activation of TrkA 51 , 52 ., The SgG2-NGF-TrkA complex promotes an aberrant downstream signaling ., The enhanced NGF-mediated TrkA phosphorylation in the presence of SgG2 could be due to the presence of higher levels of TrkA at the plasma membrane , to differences in NGF-TrkA sensitivity , or to the modification of TrkA-ganglioside association since the raft environment determines the interaction of receptors with specific signaling components 19 ., One of the most intriguing questions that arise from SgG2-NGF-TrkA signaling resides in the long-term inactivation of cofilin through phosphorylation ., Cofilin is an actin-severing protein that regulates actin turnover 24 ., Besides , cofilin is a key component of the TrkA retrograde transport complex 18 ., Our data showed that SgG2 blocks NGF-induced TrkA retrograde transport and disturbs growth cone morphology ., Both results are in agreement with the observed phosphorylation of cofilin ., An increased amount of p-TrkA in distal axons may lead to a rapid and transient axonal growth ., Growth cones appear to function as a probe 53 , and exposure to SgG2-NGF reduces filopodia resulting in a blunt growth cone ., These blunt growth cones could be less responsive to repulsive navigational guidance cues and , together with an increased TrkA and ERK signaling , may reach distant , non-permissive sites ., All these molecular events promoted by SgG2 , raft relocation of TrkA , reduced internalization of TrkA in response to NGF and activation of modified NGF signaling pathway , appear to be interconnected finally resulting in the invasion of external epidermal layers by peptidergic FNE when SgG2 is ectopically expressed in mouse footpad skin , probably facilitating HSV-2 access to this type of FNE ., We have shown that HSV gG is a vCKBP with the unique property of enhancing the activity of chemokines 12 , immune proteins involved in the migration and activation of leukocytes 54 ., Both gG1 and gG2 potentiate chemokine-mediated migration in vitro and in vivo , and this may facilitate the infection of cells recruited to areas of infection and thereby virus dissemination to other tissues ., The migration of uninfected epithelial cells towards HSV infected sites has been shown 55 supporting the ability of HSV to modulate cell migration ., Here we demonstrate that HSV SgG2 interacts with neurotrophic factors and enhances NGF-dependent growth of FNE to sites of infection probably facilitating transmission to the PNS ., Interestingly , SgG2 uses a common mechanism to enhance chemokine and NGF activity , causing the relocalization of their receptors to specific ganglioside-rich sites in the plasma membrane and delaying the internalization rate of the receptors , increasing their levels at the cell surface ( this report and N . M . -M . , A . V . -B . and A . A . , manuscript submitted ) ., SgG2 seems to bind both chemokines and NGF simultaneously and this property could play a relevant role during in vivo infection due to the contribution of chemokines and neurotrophins in the crosstalk between the immune and nervous system ., Chemokines are essential in the antiviral response but they can also modulate the responsiveness of axons to guidance cues , acting as antagonists of axonal repulsion and inducing axonal sprouting 56–58 ., Neurotrophic factors are essential elements of the nervous system and also play relevant roles in inflammation and pain modulation 59 ., NGF is the major contributor of axonal growth of peptidergic neurons and a key component of the neurogenic inflammatory response in the epidermis in response to stress or under pathological conditions , permitting a coordinated response between immune cells and peptidergic neurons 60–62 ., Our results in vitro , ex vivo and in vivo , lead us to propose that SgG2 modifies NGF-TrkA signaling to induce peptidergic neurons to grow to external layers of the epidermis , facilitating access of HSV-2 to peptidergic neurons ., This is , to our knowledge , the first example of a viral protein , encoded by a highly prevalent neurotropic human pathogen that interacts with components of both immune and nervous systems , and both activities may help HSV-2 to reach FNE to establish latency in neurons ., A similar strategy may be used by other relevant human pathogens ., The interaction of SgG2 with neurotrophic factors sheds light on the complex network of virus-host interactions and uncovers a new molecular framework to investigate the colonization of the nervous system by HSV-2 , an important human pathogen ., All animal experiments were performed in compliance with national and international regulations and were approved by the Ethical Review Board of the Centro de Biología Molecular Severo Ochoa under the project number SAF2009–07857 ., The procedures employed complied with the National ( “Real Decreto” 1201/2005 and 53/2013 ) and European ( Directives 86/609/CEE and 53/2013 ) regulations ., The interactions betwee | Introduction, Results, Discussion, Materials and Methods | Herpes simplex virus type 1 ( HSV-1 ) and HSV-2 are highly prevalent viruses that cause a variety of diseases , from cold sores to encephalitis ., Both viruses establish latency in peripheral neurons but the molecular mechanisms facilitating the infection of neurons are not fully understood ., Using surface plasmon resonance and crosslinking assays , we show that glycoprotein G ( gG ) from HSV-2 , known to modulate immune mediators ( chemokines ) , also interacts with neurotrophic factors , with high affinity ., In our experimental model , HSV-2 secreted gG ( SgG2 ) increases nerve growth factor ( NGF ) -dependent axonal growth of sympathetic neurons ex vivo , and modifies tropomyosin related kinase ( Trk ) A-mediated signaling ., SgG2 alters TrkA recruitment to lipid rafts and decreases TrkA internalization ., We could show , with microfluidic devices , that SgG2 reduced NGF-induced TrkA retrograde transport ., In vivo , both HSV-2 infection and SgG2 expression in mouse hindpaw epidermis enhance axonal growth modifying the termination zone of the NGF-dependent peptidergic free nerve endings ., This constitutes , to our knowledge , the discovery of the first viral protein that modulates neurotrophins , an activity that may facilitate HSV-2 infection of neurons ., This dual function of the chemokine-binding protein SgG2 uncovers a novel strategy developed by HSV-2 to modulate factors from both the immune and nervous systems . | Herpes simplex virus type 1 and 2 ( HSV-1 and HSV-2 , respectively ) establish latency in peripheral sensory ganglia , where they remain for the lifetime of the infected individual ., Understanding the mechanisms that allow these viruses to colonize the nervous system will permit devising antiviral strategies ., We show that HSV-2 glycoprotein G ( SgG2 ) binds to and increases the function of nerve growth factor ( NGF ) , a neurotrophin expressed in the skin and mucosa essential for axonal growth and neuronal survival ., This constitutes the first description , to our knowledge , of a human pathogen with the ability to augment neurotrophic factor function ., The enhancement in NGF activity results in an increase in axonal growth of neurons expressing the receptor for NGF ., These results were obtained in vitro , ex vivo and in the infected mouse , suggesting that this effect may permit a more efficient infection of NGF dependent free nerve endings by HSV-2 ., Absence of a similar function for HSV-1 gG may indicate a preference for the infection of particular subsets of neurons by these viruses ., These results shed light on the modulation of neurotrophic factors by relevant human pathogens and on the mechanisms of colonization of the nervous system by HSV . | null | null |
journal.pbio.2002940 | 2,017 | The Ink4a/Arf locus operates as a regulator of the circadian clock modulating RAS activity | Earth’s rotation within repetitive cycles of 24 hours ( h ) led to the evolution of an endogenous timing system across all phyla: the circadian clock , which allows organisms to anticipate and adapt to environmental changes such as light and darkness ., At the molecular level , the generation of circadian rhythms in each cell is based on a complex interplay of positive and negative transcriptional/translational feedback loops ., In mammals , two main feedback loops have been dissected in greater detail 1 ., In the period ( PER ) /cryptochrome ( CRY ) loop , a heterodimer of the proteins circadian locomotor output cycles kaput ( CLOCK ) and brain and muscle aryl hydrocarbon receptor nuclear translocator-like protein ( BMAL ) drives the expression of genes from the Per and Cry families ., The resulting proteins translocate back to the nucleus , form complexes , and bind to the CLOCK/BMAL heterodimer , thereby inhibiting their own synthesis ., CLOCK/BMAL also induces the transcription of the nuclear receptor gene families reverse strand of ERBA ( Rev-Erb ) and RAR-related orphan receptors ( Ror ) within the ROR/Bmal/REV-ERB loop ., REV-ERB and ROR compete for the binding to ROR response elements ( ROREs ) in the promoter region of Bmal1 , regulating its transcription in antagonistic ways ., By controlling rhythmic RNA and protein abundance , the cellular endogenous clock regulates circadian rhythms of a variety of biological processes such as rest/activity cycles , metabolism , hormone level , immune functions , and the metabolism of drugs 2 , 3 ., Disturbances of the circadian system are associated with many pathological phenotypes including cancer 4 , 5 , though the effect of the circadian clock in tumourigenesis is still an issue of ongoing debate ., Several epidemiological studies reported increased occurrence of cancer in long-term shift workers 6 , 7 , indicating that disrupted circadian rhythms may constitute a cancer risk factor ., However , a causative connection is still disputed 8 ., Particularly at the molecular level , there is a need for further investigation regarding the putative role of the circadian clock in tumourigenesis ., An oncogenic-driven mouse model identified CLOCK and BMAL1 as playing a key role in regulating proliferation and differentiation pointing to a complex role of the circadian clock in cancer progression 9 ., Perturbations of the circadian clock ( via experimental chronic jet lag ) in mice led to accelerated tumour growth , which could be counterbalanced by regular timing of food access 10 ., While the disruption of core-clock genes such as Per1 and Per2 has also been associated with cancer promoting mechanisms , 11 , 12 , their role in tumourigenesis is still debatable because a recent study showed that Per1 or Per2 deficiency does not lead to more tumour-prone phenotypes in mice 13 ., In turn , cancer strongly influences circadian rhythms of biological processes , such as fatty acid and cholesterol biosynthesis , and metabolic oscillations 14 , 15 ., Current attempts of chronobiological cancer treatment provide promising results , leading to a reduced toxic effect of drugs and increasing survival times of cancer patients 16 , 17 ., One of the processes via which the circadian clock could potentially influence tumourigenesis , namely by triggering malignant proliferation , might be the cell-division cycle in whose regulation the clock is involved 18 ., Several links between the circadian clock and the cell cycle have been reported for mammalian cells 19–22 ., The clock was found to unidirectionally gate the cell cycle in mouse liver cells via a circadian expression of the cell cycle regulator Wee1 19 , whereas in mouse fibroblasts , the circadian clock was reported to be phase-shifted by mitosis , possibly via concentration changes of the PER-CRY complexes 20 ., More recently , the non-POU domain containing octamer binding ( NONO ) -PER protein complex has been reported to couple the clock and the cell cycle by activating the rhythmic transcription of the cell cycle checkpoint gene Ink4a 22 , which encodes , as part of the Cdkn2a locus , the cyclin-dependent kinase ( CDK ) inhibitor p16Ink4a 23 ., p16Ink4a can be activated by the oncogene RAS , leading to cell cycle arrest in the Gap 1 ( G1 ) phase , a tumour-suppressive mechanism counteracting abnormal cell proliferation 24 ., A correlation between the expression level of Per2 and the expression of Ink4a mRNA has also been reported 25 , indicating that PER is a positive regulator of Ink4a expression and might be responsible for its circadian expression ., In addition , Cdkn2a encodes for another tumour suppressor and cell cycle regulator protein , alternate open reading frame ( ARF ) 26 , which is mainly activated by mitogenic stimulation through the cell cycle checkpoint protein avian myelocytomatosis viral oncogene homolog ( MYC ) 27 , but can also be induced via oncogenic RAS 28 ., ARF interacts in a circadian time ( CT ) -dependent manner with a mouse double minute 2 homolog ( MDM2 ) , a negative regulator of the tumour suppressor gene p53 29 ., p53 , in turn , directly modulates the expression of Per2 by binding to a response element in its promoter region , which is overlapping with the E-box cis-element essential for the CLOCK/BMAL binding and the transcriptional activation of Per2 30 ., In addition , the transcription factor E2F was described as a putative bridging element between the circadian clock and the cell cycle 31 ., Recently , we demonstrated that RAS induces a dysregulation of the mammalian circadian clock in HaCaT keratinocyte cells 32 ., We showed that the induction of RAS leads to a lengthening of the circadian period while inhibition of the RAS/mitogen-activated protein kinases ( MAPK ) pathway causes a period shortening ., These findings were supported by using a mathematical model to analyse the rhythmic properties of the core-clock network in silico 1 , 32 ., Other mathematical models have been developed that investigate the cell cycle-related functions of the tumour suppressor genes p53 and Arf , as well as Mdm2 33–35 ., Despite the accumulating experimental and in silico evidence pointing to a role of the circadian clock as a suppressor of aberrant cell proliferation , the underlying molecular mechanisms are not yet fully understood ., Although mathematical models exist that couple elements of the cell cycle to the components of the circadian clock 36–42 , none of them include the cell cycle regulatory genes Ink4a and Arf , nor do they include specific parameters for the investigation of the role of oncogenic-mediated signalling via RAS ., Using mouse embryonic fibroblasts ( MEFs ) as a model system , we investigated the Ink4a/Arf- and Bmal1-dependent influence of RAS on the circadian phenotype ., To attain a deeper understanding concerning the molecular interactions of the circadian clock with components of the cell cycle , we constructed a single-cell semi-quantitative mathematical model ., The model allows for the coupling of components of the cell division cycle with the core-clock network , enabling the interpretation of the RAS-mediated effect on the circadian clock phenotype observed in Ink4a/Arf knockout MEFs as compared to their wild-type ( WT ) counterparts ., In particular , the model was used to attain a better understanding of the mechanism by which INK4a and ARF might influence properties of the circadian clock ., Furthermore , we performed a comprehensive bioinformatics analysis to investigate the systemic effects of different experimental perturbations at the cellular level ., Expression data of a set of core-clock genes , as well as in silico simulations , show that RAS overexpression influences the transcriptional expression and period of core-clock genes such as Bmal and Per differentially in the Ink4a/Arf+/+ and Ink4a/Arf-/- scenarios , which points to a regulatory role of Ink4a/Arf in the RAS-mediated effect on the circadian clock ., By analysing the relevance of selected cell cycle components in mediating the RAS-induced change on the circadian period via Ink4a/Arf in silico , we show that the presence of the transcription factors E2F1 and p53 is necessary for simulating this phenotype ., Genome-wide transcriptional profiling data from Ink4a/Arf-/- MEFs and their corresponding WT MEFs ( Ink4a/Arf+/+ ) are in line with the mathematical predictions showing that Ink4a/Arf play an essential role in the RAS-mediated effect on the gene expression levels of core-clock and cell cycle-related genes ., Hence , our findings resulting from a combined computational and experimental approach provide a deeper insight into the dynamics of the cross-talk between oncogene-induced perturbations of the circadian system and the cell cycle and reveal a novel role for Ink4a/Arf as an important regulator of the RAS-mediated effect on the circadian clock phenotype ., Recently , we showed that oncogenic RAS influences properties of the circadian clock by lengthening the circadian period in different cell types 32 ., To attain a deeper mechanistic insight on how RAS perturbs the circadian clock and on its subsequent effect on the cell cycle and cell proliferation , we used a well-established cellular model system of MEFs ., We compared the circadian phenotypes of MEFs from WT mice ( Ink4a/Arf+/+ ) to their littermate knock-out MEFs ( Ink4a/Arf-/- ) , which carry a targeted deletion of exons 2 and 3 of the tumour suppressor gene Cdkn2a , disrupting both Ink4a and Arf 23 ., We analysed the effect of RAS on the circadian clock phenotype by bioluminescence recordings and the induction of senescent cell cycle arrest by SA-ß-Gal staining ., MEFs were lentivirally transduced with a Bmal1-promoter driven luciferase construct ( Bmal1:Luc ) and bioluminescence was recorded for 5 days as schematically represented in Fig 1A ., The bioluminescence data shows similar clock phenotypes for Ink4a/Arf-/- MEFs and their corresponding Ink4a/Arf+/+ littermates with average period values of around 24 h ( T = 24 . 7 ± 0 . 3 h for Ink4a/Arf+/+ MEFs and T = 24 . 2 ± 0 . 2 h for Ink4a/Arf-/- MEFs , n = 5; mean and SEM; representative results in Fig 1B and 1C , summary of the data in Table 1 ) , indicating that the period of the circadian clock is not influenced by the knock-out of Ink4a/Arf ., The viability of the cells was not affected by the knockout ( S1A Fig ) ., To specifically study the interaction of the RAS/MAPK pathway with components of the circadian clock , oncogenic RAS ( H-RAS G12V ) was overexpressed in both cell types ., Interestingly , upon RAS overexpression , we observed a change of the circadian clock phenotype: the RAS overexpressing Ink4a/Arf+/+ cells ( T = 26 . 7 ± 0 . 7 h , n = 5; mean and SEM ) show a 2 h longer period as compared to the WT condition ( T = 24 . 7 ± 0 . 3 h , n = 5; mean and SEM; p = 0 . 004 , Student t test ) , whereas in the RAS-infected Ink4a/Arf-/- MEFs , a decrease of 2 . 5 h in the period was observed ( T = 21 . 7 ± 0 . 6 h , n = 5; mean and SEM; p = 0 . 0003 , Student t test ) as compared to cells with unmodified RAS levels ( T = 24 . 2 ± 0 . 2 h , n = 5; mean and SEM; representative results in Fig 1B and 1C , summary of the data in Table 1 ) ., To investigate a putative effect of RAS on core-clock components , Bmal1 was downregulated by short hairpin RNA ( shRNA ) prior to the stable expression of oncogenic RAS in both Ink4a/Arf-/- MEFs and their WT littermates ., As expected , the downregulation of Bmal1 disrupted circadian rhythmicity ( representative results in Fig 1D and 1E , summary of the data in Table 1 ) ., Bioluminescence data from several independent experiments ( five WT mice and Ink4a/Arf-/- littermates , with two independent transductions per condition ) are summarised in Fig 1F and Table 1 and reproduced our representative results shown in Fig 1B and 1C ., Ink4a/Arf-/- MEFs proliferate faster than the Ink4a/Arf+/+ MEFs , independent of the downregulation of Bmal1 ( Fig 1G ) ., To investigate the effect of RAS overexpression and Bmal1 downregulation on cellular senescence , we performed SA-ß-Gal staining using Ink4a/Arf-/- MEFS and WT littermates ( n = 3 WT mice and Ink4a/Arf-/- littermates ) and tested for senescence-associated galactosidase activity ., While the Ink4a/Arf+/+ MEF population has a low amount of senescence cells ( 7 ± 1 . 44% , n = 3; mean and SEM ) , RAS overexpression leads to a strong increase ( 94 . 5 ± 1 . 15% , n = 3; mean and SEM; Fig 1H and 1I ) ., The results are in agreement with published data showing that RAS overexpression increases the percentage of senescent cells in WT MEFs 24 ., In comparison , the population of Ink4a/Arf-/- MEFs shows the lowest number of senescent cells ( 0 . 83 ± 0 . 44% , n = 3; mean and SEM ) independent of RAS overexpression ( 1 . 5 ± 0 . 5% , n = 3; mean and SEM; Fig 1H and 1I ) ., The downregulation of Bmal1 shows no significant effect on the senescence phenotype ( Fig 1H and 1I ) ., To explore the observed Ink4a/Arf-dependent effect of RAS overexpression on the clock—an increased circadian period in WT MEFs but a decreased period in Ink4a/Arf-/- cells ( Fig 1B–1E ) —we developed a novel semi-quantitative mathematical model coupling the mammalian cell cycle and the circadian clock using ordinary differential equations ( ODEs ) ., The model contains all elements of our previously published single-cell model of the mammalian circadian core-clock 1 , the cell cycle elements INK4a and ARF , and a minimal selection of their respective interaction partners , which allow for the connection to the core-clock ., These include the cell cycle checkpoint regulators Myc ( G1/S ) and Wee1 ( G2/M ) and components of two INK4a- and ARF-dependent signalling pathways: the INK4a/ retinoblastoma-associated protein 1 ( RB1 ) /E2F1 pathway ( module, 1 ) and the ARF/MDM2/p53 pathway ( module 2 ) ., The resulting regulatory network includes nine additional elements ( MYC , WEE1 , ARF , MDM2 , INK4 , CDK/Cyc , p53 , RB1 , E2F1 ) and their corresponding transcriptional and translational components ( Fig 2A , S2 Fig ) ., In total , the model comprises 46 variables and 170 parameters ., In the model , the CLOCK/BMAL complex activates the transcription of Wee1 and represses Myc transcription as reported in the literature 19 , 43 , 44 ., The transcriptional activation of Ink4a by the PER-NONO complex in a circadian manner 45 , 46 is modelled as a positive interaction between the PER/CRY complex and Ink4a transcription ., By forming a complex with D-type CDKs CDK4 and CDK6 , INK4a prevents the interaction with the cell cycle checkpoint regulator Cyclin D ( CycD ) , thereby inhibiting the subsequent phosphorylation of the RB1 , a key regulator of the E2F family of transcription factors ( E2F1 , E2F2 , and E2F3a ) 47–51 ., The dephosphorylated active form of RB1 inhibits the dissociation of the RB1/E2F complex leading to the inactivation of E2F-mediated transcription of cell cycle genes ., The cell cycle/core-clock loop in module 1 is completed by a predicted transcriptional activation of Bmal by E2F because E2F potentially binds to the promoter region of Bmal1 ( as reported by MotifMap , the genome-wide map of candidate regulatory motif sites for humans 52 ) ., The prediction of E2F as an important binding element between the clock and the cell cycle is further supported by data from the unicellular red alga Cyanidioschyzon merolae in which time-dependent phosphorylation of E2F promotes the G1/S transition and a mutation of the E2F phosphorylation sites results in an uncoupling of cell cycle progression from the circadian clock 31 ., Altogether , these results point to a putative connection between E2F and the clock which—given the existing data—can be assumed to happen via Bmal1 ., Module 2 contains the ARF/MDM2/p53 pathway and represents an indirect feedback from ARF to the core-clock ., ARF is encoded by the same gene locus as INK4a and can be activated by oncogenic MYC or oncogenic RAS 47 ., Accumulated ARF associates with the p53 inhibitor MDM2 and leads to its rapid proteasomal degradation ., This decreases MDM2-mediated ubiquitination of the tumour suppressor p53 and promotes its stabilisation , which in turn activates the transcription of Mdm2 53 , 54 ., A recent study describes a p53 response element located in the promoter region of Per2 , which overlaps with the E-box cis-elements crucial for CLOCK/BMAL-mediated Per2 transcription 30 ., Thus , the binding of p53 strongly represses the transcription of Per2 by competing with the CLOCK/BMAL binding to its promoter 30 ., Apart from the strong negative influence of p53 on Per2 , PER2 is also known to transcriptionally modulate p53 55 , and this regulation is thought to be positive 39 ., For simplicity , we have merged the mutual influence into one negative interaction from p53 to Per2 ., In addition , p53 inhibits the phosphorylation of RB1 via the p21/CDK/CycE/RB1 pathway 56 , 57 ., Although we cannot exclude the possibility that other elements may also be involved in connecting the clock and the cell cycle elements INK4a and ARF , the chosen modules represent a minimal functional set well-supported by published data 44 , 58–60 that enables us to investigate the properties of the system in silico ., The model development and the complete network scheme is described in greater detail in the supplementary information ( S1 Text , S2 Fig ) ., The resulting model is robust to perturbations in the range of ± 10% as shown by the control coefficient analysis over all parameters ( S1 Text ) ., The period of the model system was adjusted to 23 . 65 h and the phase of Bmal mRNA expression was set to CT 21 h ., To examine whether simulations of the model are in agreement with biological phenotypes of the clock , we compared the peak phases of the in silico mRNA expression patterns of core-clock genes and the PER/CRY protein complexes with experimental data retrieved from the literature 1 ., The peak phases for all core-clock genes are within the range of published experimental peak phases of core-clock mRNAs ( Fig 2B , Table 2 ) ., Furthermore , the model successfully reproduces the correct phase relations among the core-clock components ( Fig 2C ) ., Data from our previous work points to a role of RAS as a regulator of the circadian clock period 32 as was also reported by other studies 61 ., With our novel mathematical model , we investigated whether the observed RAS-induced change in the circadian clock period ( Figs 1B and 2C ) can be simulated both in the Ink4a/Arf WT and in the knock-out condition ( Fig 2D ) ., For the simulation , we adapted a method from our previous work on RAS-mediated dysregulation of the circadian clock in cancer , in which a factor ktt was introduced to the activation/inhibition rates describing CLOCK/BMAL-mediated transcription: ktt = 1 describes a normal RAS expression level whereas ktt < 1 indicates a reduction in the transcriptional activity of CLOCK/BMAL caused by RAS overexpression 32 ., The double knock-out of Ink4a/Arf was achieved by setting the initial conditions of INK4a and ARF mRNAs , cytoplasmic and nuclear proteins , as well as their rate of change to 0 ( S1 Text , equations 1 , 2 , 8 , 10 , 14 , 19 = 0 ) ., We measured the period in our model for a transient region , defined as the mean of the time between the first four peaks ( three periods ) after introducing the perturbation of RAS ( represented by ktt < 1 ) to the system ., In this transient region , there are still fluctuations of the modelled system that can represent the observed biological noise of retrovirus-mediated RAS overexpression ., More information concerning the model analysis can be found in S1 Text ., The model predicts a slightly longer circadian period for the Ink4a/Arf-/- system ( 23 . 68 h compared to the adjusted period of 23 . 65 h in the WT ) when RAS is expressed at WT levels ( ktt = 1; Fig 2D ) , which results in a phase shift over time ( S3A Fig ) ., As expected from the experimental data shown in Fig 1 and our previously published results 32 , there is a lengthening of the period upon RAS overexpression in Ink4a/Arf+/+ MEFs ( Fig 2D ) ., Interestingly , the opposite effect on the period is predicted by simulations in the Ink4a/Arf-/- system ., Thus , the in silico period changes are in agreement with the experimentally measured phenotypes in Ink4a/Arf-/- MEFs and their WT littermates ( Fig 1B and 1C ) and show the same tendency of an increase/a decrease of the period length in response to RAS overexpression ., The simulations were not fitted to exactly reproduce the values of the experimentally measured periods , which were obtained from MEFs originating from different mice and therefore show a certain biological variation ., The simulated Ink4a/Arf-/- system shows a nonmonotonic dependency of the period length on the strength of the RAS overexpression ., With increasing RAS , the period length first decreases , reaching its minimum for ktt = 0 . 7 , only to increase afterwards in response to even higher simulated levels of RAS ( 0 . 4 < ktt < 0 . 7; Fig 2D ) ., After introducing the perturbation of RAS to the system by varying the parameter ktt , there are nonmonotonic phase shifts of Bmal expression that depend on the strength of RAS overexpression ( S1 Text ) ., The change of the ktt value causes Bmal oscillations to peak at different times: The system with the strongest RAS overexpression and lowest simulated ktt value ( ktt = 0 . 4 ) peaks the earliest for the first iteration but is then superseded by the system with ktt = 0 . 7 for the following iterations ., The WT system always peaks last ., These phase shifts are especially prominent for the Ink4a/Arf-/- system , which represents one possible explanation for the observed nonmonotonic period changes ( Fig 2D ) ., Yet the effect of RAS overexpression on the length of the period is also dependent on the time when the perturbation is introduced to the model system ( S3B Fig ) ., Interestingly , the model predicts a longer period for an inhibition of RAS ( ktt > 1 ) in the Ink4a/Arf-/- system ( S1 Text ) ., We were able to confirm this prediction experimentally by inhibiting RAS in the Ink4a/Arf-/- MEFs by using a MEK inhibitor that blocks the downstream chain in the RAS signalling pathway ., Contrary to the shortened period observed when overexpressing RAS , we now observed an increase of the period ( T = 23 . 76 ± 0 . 1 h for Ink4a/Arf-/- MEFs and T = 25 . 24 ± 0 . 1 h for Ink4a/Arf-/- -RAS MEFs; n = 3; mean and SEM; S1C–S1E Fig ) ., Furthermore , we simulated the overexpression of RAS in the Ink4a/Arf+/+ condition ( ktt = 0 . 6 ) ., The model predicts that RAS overexpression leads to an increase of the expression level of Ink4a as compared to the normal RAS scenario ( Fig 2E ) ., This is in line with published data showing that RAS activates Ink4a , leading to cell cycle arrest 24 , 27 and correlates with the RAS-induced increased number of senescent cells shown in Fig 1I ., Moreover , we downregulated RAS in the Ink4a/Arf-/- MEFs and observed a long period phenotype , as opposed to the period change observed for RAS overexpression ( S1C–S1E Fig ) ., Furthermore , we used an RAS-inducible construct to investigate the effects of different levels of RAS induction ., Our data indicates that both longer and shorter periods can be observed , in a RAS-dependent manner , in agreement with the simulations from our mathematical model ( S1F–S1J Fig ) ., Taken together , the results demonstrate that the model reproduces important circadian properties , which are in agreement with experimental data , and that it can be used to further elucidate the mechanism of RAS-induced and Ink4a/Arf-dependent changes of the circadian period ., In order to investigate the relative influence of the INK4a/RB1/E2F1 ( module, 1 ) and ARF/MDM2/p53 ( module, 2 ) pathways in mediating the RAS-induced effect on the core-clock in silico , we tested whether the presence of key elements in the network and the oscillations of both modules are necessary to reproduce the experimentally determined clock period phenotypes of Ink4a/Arf+/+ and Ink4a/Arf-/- MEFs ., As shown in Fig 3A , module 1 represents the connection of INK4a to the clock via the INK4a-dependent inhibition of E2F , a transcription factor that we predict to regulate the transcription of Bmal in the model ., Still , we cannot exclude the possibility that such a regulation may happen via additional elements that were not investigated within the scope of this study ( S3 Text ) ., However , additional elements would not change the validity of the conclusions derived from the model as long as the delays in the expression values between the defined core-clock elements remain ., These delays or phase differences , which were retrieved from published experimental data , were used as constraints in our model ( S1 Text ) ., Module 2 ( Fig 3B ) represents the ARF-mediated activation of the transcription factor p53 , which is known to repress the CLOCK/BMAL-mediated transcription of Per2 ., In both modules , the nuclear protein elements show oscillations in their simulated expression patterns according to the circadian rhythm of the core-clock system ( Figs 3C and 1D ) ., The connection of module 1 to the core-clock was disrupted by setting the concentration of E2FN to 0 ( S1 Text , equations 7 , 12 , 22 = 0; initial concentration of E2FN = 0 ) , thereby removing its predicted regulation of Bmal in the model ., Ink4a/Arf knockout was modelled as described above ., Upon the disconnection of module 1 , both the Ink4a/Arf+/+ and the Ink4a/Arf-/- system acquire shorter periods when RAS is overexpressed ( ktt = 0 . 7 ) as compared to normal RAS conditions ( ktt = 1; Fig 3E ) ., These results are not in line with our previous simulations for the full network ( Fig 2D ) and the experimental observations ( Fig 1A and 1B ) , which show an Ink4a/Arf-dependent effect on the period upon RAS overexpression ., Thus , it seems that in our modelling scenario the predicted connection between E2F and Bmal is indeed necessary to reproduce the observed period changes ., When comparing the expression and the period length of Bmal oscillations before and after the perturbation by RAS , it becomes evident that the knockout of module 1 results in a lower period value of 22 . 86 h for both the Ink4a/Arf+/+ and the Ink4a/Arf-/- system ., This results in a phase shift of Bmal oscillations when compared to the oscillations in the WT , causing the differing effect on the Bmal period length upon the perturbation by RAS ( S1 Text ) ., We further investigated the relevance of the oscillations in module 1 for the RAS-mediated effect on the circadian period ., The connection between E2FN and Bmal was maintained , but the expression of E2FN was clamped to the constitutive concentration of its mean value ( S1 Text , E2FN = 5 . 7 , equations 7 , 12 , 22 = 0 ) ., The constant expression of E2F leads to similar period lengths of 23 . 63 h for the Ink4a/Arf+/+ and the Ink4a/Arf-/- systems ( Fig 3F ) ., In this case , RAS overexpression causes a nonmonotonic change of the period length depending on the value of ktt , but independent of the Ink4a/Arf status resulting in slightly shorter periods for 0 . 7 ≤ ktt < 1 and longer periods for 0 . 4 ≤ ktt < 0 . 7 ., Again , this does neither reproduce the results of the previous simulation nor the experimental observations ., Bmal oscillation profiles of both systems shows that there is only a very small phase shift between the Ink4a/Arf+/+ and the Ink4a/Arf-/- systems , potentially explaining why both systems react similar upon perturbation by RAS ( S1 Text ) ., These results indicate that in the model , both the connection of the INK4a/RB1/E2F1 module to the core-clock via E2F regulation of Bmal and the low-amplitude oscillations of E2F are crucial to simulate the contrary effect of RAS overexpression on the circadian period ., Next , we tested whether the connection of module 2 to the core-clock is also necessary to reproduce the experimentally observed phenotype by setting the concentration of the component p53N to 0 ( S1 Text , equation 16 = 0 ) ., The expression profile of Bmal shows that the perturbation is introduced at a similar oscillation phase for both systems , which differs from that of the WT Bmal oscillations ( S1 Text ) ., As before , both the Ink4a/Arf+/+ and the Ink4a/Arf-/- system show a nonmonotonic change of the period dependent on the strength of the simulated RAS overexpression: for 0 . 7 ≤ ktt < 1 , there is a decrease in period length , followed by a slight increase for 0 . 4 ≤ ktt < 0 . 7 ., This does not reproduce the experimentally observed RAS-induced period changes indicating that the connection of the ARF/MDM2/p53 pathway to the core-clock is crucial as well ., To simulate a constitutive connection of module 2 to the core-clock , we clamped the oscillatory expression of p53N in module 2 to its average expression value ( S1 Text , equation 16 = 0 . 6 ) ., Interestingly , by using a constitutive concentration of p53N , we were able to simulate circadian phenotypes similar to the simulations of the whole network and the experimental observations ( Fig 3H ) ., Although p53 is no longer oscillating , there is a similar phase shift between the Ink4a/Arf+/+ and the Ink4a/Arf-/- system as in the original ( S1 Text ) ., This suggests that the oscillation of the ARF/MDM2/p53 pathway plays only a minor role in regulating the period in response to RAS overexpression and can instead be substituted by a constitutive expression ., Taken together , these results indicate that upon RAS overexpression , the connections of both the ARF/MDM2/p53 and the INK4a/RB1/E2F1 pathway to the core-clock are necessary to produce the observed period phenotypes ., The simulations from the mathematical model predict that a circadian expression of the INK4a/RB1/E2F1 pathway is crucial for reproducing the Ink4a/Arf-dependent change in rhythmicity in response to RAS overexpression ., This again points to dynamical effects such as phase shifting in gene expression , which are hard to identify experimentally because even small phase shifts can cause large effects in a feedback loop ., Furthermore , we simulated the expression of representative components of the core-clock ( Per ) , module 1 ( E2f ) , and module 2 ( p53N ) under four different conditions ( Ink4a/Arf+/+ , Ink4a/Arf+/++RAS , Ink4a/Arf-/- , and Ink4a/Arf-/-+RAS ) and compared them to experimental 24-h time-course measurements of Per2 , E2f , and p53 under the same conditions ., The knockout of INK4a and ARF and the overexpression of RAS ( ktt = 0 . 6 ) were modelled as described above ., Both the in silico simulations and the experimental measurements show that Per is oscillating with a circadian rhythm in all conditions ( Fig 3I ) ., As expected , there is a phase shift in the simulated Per expression upon the knockout of Ink4a and Arf , with the knockout peaking slightly later than the WT , which is in line with the phase shift observed in the simulations of Bmal expression ( S3A Fig ) ., The same tendency can be observed in the experimental time-course measurements of Per2 ., RAS overexpression reduces the amplitude of Per oscillations both in the WT and the Ink4a/Arf-/- system ( Fig 3I ) ., E2f and p53 exhibit only low expression changes ., The simulated expression patterns of E2f show oscillations with similar amplitudes in all four conditions while in the experimental measurements , the WT has a higher fold change than the perturbed conditions ( Fig 3J ) ., The in silico expression of p53N shows low amplitude oscillations in the WT conditions ( both with and without RAS overexpression ) , which are out-of-phase with Per oscillations ( S1 Text ) in agreement with a recent work that modelled the spatiotemporal regulation of p53 by Per2 39 ., In the knockout condition , its expression is constitutive and does not change upon overexpression of RAS ( Fig 3K ) , which is reflected in the low fold change observed for p53 ., This supports our prediction that the constitutive , but not the oscillatory expression of the ARF/MDM2/p53 pathway i | Introduction, Results, Discussion, Materials and methods | The mammalian circadian clock and the cell cycle are two major biological oscillators whose coupling influences cell fate decisions ., In the present study , we use a model-driven experimental approach to investigate the interplay between clock and cell cycle components and the dysregulatory effects of RAS on this coupled system ., In particular , we focus on the Ink4a/Arf locus as one of the bridging clock-cell cycle elements ., Upon perturbations by the rat sarcoma viral oncogene ( RAS ) , differential effects on the circadian phenotype were observed in wild-type and Ink4a/Arf knock-out mouse embryonic fibroblasts ( MEFs ) , which could be reproduced by our modelling simulations and correlated with opposing cell cycle fate decisions ., Interestingly , the observed changes can be attributed to in silico phase shifts in the expression of core-clock elements ., A genome-wide analysis revealed a set of differentially expressed genes that form an intricate network with the circadian system with enriched pathways involved in opposing cell cycle phenotypes ., In addition , a machine learning approach complemented by cell cycle analysis classified the observed cell cycle fate decisions as dependent on Ink4a/Arf and the oncogene RAS and highlighted a putative fine-tuning role of Bmal1 as an elicitor of such processes , ultimately resulting in increased cell proliferation in the Ink4a/Arf knock-out scenario ., This indicates that the dysregulation of the core-clock might work as an enhancer of RAS-mediated regulation of the cell cycle ., Our combined in silico and in vitro approach highlights the important role of the circadian clock as an Ink4a/Arf-dependent modulator of oncogene-induced cell fate decisions , reinforcing its function as a tumour-suppressor and the close interplay between the clock and the cell cycle network . | In mammals , the circadian clock controls the punctual regulation of biological processes , which , in turn , affect physiology and behaviour , allowing for the synchronisation of internal time to environmental light-dark cycles ., Malfunctions of the circadian clock are associated with pathological phenotypes including cancer ., Given the range of molecular time-dependent processes , including metabolism , DNA repair , and the cell cycle , the clock is hypothesised to act as a tumour suppressor ., With the help of mathematical modelling and whole-genome analysis combined with machine learning , we investigated the RAS-dependent dysregulation of the circadian clock ., We find that the tumour-suppressor Ink4a/Arf acts as a key mediator of RAS oncogene-induced changes in the circadian system , thereby mediating the interplay between the clock and the cell cycle . | cell cycle and cell division, cell processes, biological cultures, mathematical models, circadian oscillators, chronobiology, molecular biology techniques, research and analysis methods, sw480 cells, mathematical and statistical techniques, gene expression, hyperexpression techniques, cell lines, molecular biology, molecular biology assays and analysis techniques, gene expression and vector techniques, circadian rhythms, biochemistry, cell biology, genetics, biology and life sciences | null |
journal.pcbi.1005084 | 2,016 | Vulnerability-Based Critical Neurons, Synapses, and Pathways in the Caenorhabditis elegans Connectome | Understanding the fundamental architectural design of nervous systems will provide insights into nervous system function and how mental illness arises from its dysfunction ., The design derives from a trade-off between physical wiring costs and functional complexity 1–4 ., Nervous systems mediate behavior to achieve the organism’s goals , and thus the network must be composed of behavioral circuits ( e . g . , to forage , to avoid danger , to mate ) ., The punitive nature of wiring costs , however , demands highly efficient solutions , producing short and shared pathways whenever possible ., How then do nervous systems solve the problem of having distinct behavioral circuits , on the one hand , and near maximized sharing of pathways and information , on the other ?, To determine this , it is useful to examine a nervous system that is both well specified and highly tractable , making the nematode roundworm Caenorhabditis elegans an attractive model system , especially since its complete connectome is available for analysis ., Considerable progress has been made in delineating behavioral circuits 5–9 of the C . elegans including , for example , mechanosensation 10–12 , chemosensation 13 , feeding 14 , 15 , exploration 14 , 15 , egg laying 16–18 , mating 7 , 19–22 , and tap withdrawal 23 , 24 ( i . e . , avoidance of a vibrating “tap” ) ., These circuits appear to follow a general pattern of sensory neurons ( S ) to 1–3 layers of interneurons ( I ) , with the final interneuron layer being command interneurons ( i . e . , direct control of motor neurons ) , to one or more motor neurons ( M ) that directly control muscle activity ( i . e . , S → I → M pattern ) 7 , 14 , 25 ., Some subcircuits have also been characterized , for example , for both forward and backward locomotion , and more generally for foraging behavior 6 , 14 ., Nonetheless , many details remain unclear , not only with respect to the specific functions of every neuron and synapse , but also the extent to which the behavioral circuits are largely separate and parallel versus having a more integrative and serial design ( i . e . , input → central processing → output ) 7 , 14 ., These issues are especially exacerbated in the C . elegans connectome because it is highly interconnected , with evidence that behavioral functions such as navigation entail activity across a large fraction of the nervous system 26 ., Graph theoretic analysis of neuronal networks provides a normative approach to help quantitatively specify the network structure and dynamics—in which neurons are network nodes , and synapses ( whether chemical or gap junctions ) are connections or edges between the nodes ., Important network properties of the connectome have been established , including a layout that is nearly optimized to minimize wiring costs 1 , 3 , 4 , 8 , 27 ., The near optimal wiring suggests strong selection pressure on efficiency , and thus , near optimally efficient information processing: e . g . , path sharing whenever possible ., And characteristics have been identified that reflect evolved complexity in light of this constraint , especially small-worldness ( i . e . , high clustering for functional specialization , together with significant interconnections across the network for efficient information integration and transfer ) , and nonrandom distributions of highly connected neurons as well as those that support high volume traffic 4 , 7 , 8 , 27–30 ., Larger information processing structures in the network have also been identified , such as a set of highly connected interneurons that form hubs and “rich clubs” that integrate information to drive locomotion 27 , 31 ., Modularity analyses have also attempted to delineate the main functional modules based on high interconnectivity within modules versus lower across them 8 , 27 , 28 ., Although specific results across studies differ based on the modularity algorithm used , these studies demonstrate that identified modules align with known behavioral circuits , providing some evidence for separable behavioral circuits ., At the same time , the results also show elements of an integrative and serial design ( i . e . , to the extent modules align with input , central , and output processing layers ) ., Taken together , current findings may suggest a fundamentally mixed architectural design ., An important next step , then , is to build on these findings by unequivocally identifying specific critical substructures of the connectome , which will in turn help clarify the architectural design principles ., To this end , we take an evolutionary and developmental ( evo-devo ) perspective and seek to identify the most critical elements of this information processing system , i . e . , those whose addition provided the most ‘bang for the buck’ for information propagation , integration and processing 32–34 ., To determine this , in the current study we conducted a computational network-vulnerability analysis that systematically removed network components and calculated the effects of each loss on key network properties , with the view that functional loss upon removal reveals what their addition enables ., We also take the perspective that critical components in real-world nervous systems may not always depend on local individual properties independent of context , such that significance may emerge from this context ., The vulnerability analysis also enables an assessment of these contextual effects 35 , 36 ., Using this approach , we tested all neurons and synapses to assure no misassignments: i . e . , some considered critical when not ( e . g . , highly connected but redundant ) , or those that may appear not , but in fact are , from a more global network perspective ., We analyzed the most recent published connectome of the C . elegans hermaphrodite as a combined ( i . e . , chemical synapses and gap junctions ) , directed , and weighted network ( see Methods and S1 Text ) 5 , 15 , 29 ., The S1 Text contains an examination of the general network properties of the intact network , including circular wiring diagrams , comparisons of the three neuron classes ( sensory , interneurons , and motor ) on different network properties , and robustness and information propagation analyses ., We then sought to identify the most critical neurons in the C . elegans connectome ., To do this , we asked which ones had the greatest impact on network function when they are lost ., Thus , we conducted a vulnerability analysis , attacking each of the 279 neurons in the connectome and calculating the change in global network properties when the neuron is removed ( see Methods for analysis details and explanation of notation used; S1 File ) 35 , 36 ., Three fundamental properties that determine efficient information processing in a network are ( 1 ) specialized processing in the form of local clusters , ( 2 ) global information integration and coordination via short path lengths between nodes , and ( 3 ) the control of information flow by providing a route that is shared by multiple pathways ., These three properties were assessed by calculating the network vulnerability , V , ( i . e . , the relative change in value after elimination of the network component ) with respect to the clustering coefficient ( VC ) , global efficiency ( VE ) , which assesses the network’s average path length , and the average betweenness centrality ( VB ) , which represents the average of number of shortest paths that travel through each network component ( i . e . , neurons here , synapses below ) ( detailed in the Methods and S1 Text ) ., To determine the critical neurons , we reasoned that the most influential ones should yield network vulnerability scores after deletion that are clear outliers from the rest of the population , reflecting extreme functional loss in the network with respect to the specific fundamental network property ., The most conservative criterion used in the Towlson et al . 27 study to identify the rich club was selecting those that were 3SD or more above the average , which we adopted here ( Fig 1; S2 File ) ., This stricter criterion enabled the identification of the most critical neurons , it provided a standard value often used in identifying outliers ( 3SD ) , and it continued the convention established by Towlson et al . 27 , enabling a direct comparison of findings across studies ., Thus , we defined a neuron as critical for a given network property when a deletion of the neuron produced a vulnerability over three SDs above the mean vulnerability value of all 279 neuronal attacks ., Our analysis identified 2 , 5 , and 8 critical neurons for VC , VE , and VB , respectively , yielding 12 ( i . e . , 4 . 3% ) critical neurons in all ( Table 1; S1 and S2 Files ) ., Of the 12 critical neurons , 9 are interneurons , 2 are motor neurons , and 1 is sensory , reflecting the general importance of interneurons in information processing in the C . elegans connectome , as has been reported previously 27 ., Moreover , of the 7 interneurons with known function , 5 are considered command interneurons that control locomotion 27 ., To obtain further insight into the critical neurons , we examined their development times ( Fig 2A; Table 1 ) 27 ., All critical neurons for both the VC ( 2 interneurons ) and VE ( 5 interneurons ) analyses developed early ., For VB , 5 interneurons and one motor neuron ( DA01 ) developed early , whereas 2 neurons developed late ( after hatching ) : one sensory ( AQR ) and one motor ( VD01 ) ., Taken together , all interneurons and 1 motor neuron ( DA01 ) developed early , whereas , one sensory and 1 motor neuron developed late ., Thus , as found by Towlson et al . 27 , we also found that all of the critical interneurons formed in the early development phase before twitching ( 470 min after fertilization; first visible motor activity ) , presumably attesting to their importance in establishing the connectome topology ., In addition , we note that the early developing motor neuron ( DA01 ) is less peripheral than the late developing one with respect to functional connectivity ( VD01 ) in the fully formed adult connectome 6 ., Thus , our results also suggest a general ‘inside-out’ developmental pattern , with the most peripheral neurons tending to develop last , presumably requiring environmental experience after hatching 27 , 37 ., Examining the specific results for VC , E , B , the AVA neurons stand out , being critical for multiple network properties ( Table 1; S1 File ) ., AVAL and AVAR were critical for both VC and VE , meaning that they most affect network clustering and efficiency ., AVAL was additionally critical for VB , although this measure ( i . e . , VB ( i ) ) actually decreased when AVAL was removed ( without taking the absolute value ) ( S1 File ) ., This decrease indicates that a main pathway segment was removed with AVAL , leaving longer alternative routes ( resulting in increased B ( i ) overall , XB ( i ) ) ., Thus , the results show that the AVA neurons ( and AVAL in particular ) are dominant ones in the connectome , critically influencing network clustering , interconnectivity , and information flow ., Given AVA’s known role in backward movement control , these results suggest that backward movement likely plays a key role in locomotor maneuverability in potentially multiple biological functions; and to the extent backward movement in response to environmental stimuli might be construed as avoidance behavior , the results might also highlight the significance of avoidance behavior in general ., We discuss possible biological function further below ., For VE , 5 neurons were found to have the greatest effect on global efficiency: besides the two AVA neurons , DVA ( I ) and PVCL/R ( I ) ( i . e . , both left and right ) were also identified ( Table 1; S2 File ) ., All five critical neurons for VE are interneurons , which would indeed be those expected to have significant influence on global efficiency ( yet only these 5 were found to be critical ) ., Since all have been implicated in either forward locomotion for avoiding aversive stimuli or backward locomotion , it suggests that the most influential nodes on global efficiency in the connectome perform locomotion control , and again , especially with respect to backward locomotion and avoidance behavior ., For VB , 8 neurons were found to have the greatest effect on betweenness centrality: 1 sensory ( AQR ) , 5 interneurons ( AVAL , AVER , AVHL , DVC , PVPR ) , and 2 motor ( DA01 , VD01 ) ( Table 1; S1 File ) ., Thus , with respect to network information propagation , all neuronal types ( S , I , and M ) , related to potentially multiple biological functions are implicated , which we examine further in the synapse analysis ., During this study , AVHL and DVC were conspicuous critical neurons whose biological function had been unknown ( there has since been new evidence for DVC involvement in locomotion ) 38 ., We also note that due to the integrative nature of the C . elegans connectome , it is possible that many neurons ( especially interneurons ) participate in multiple functions ., We therefore conducted an analysis based on known biological function of neighbors to help determine the possible functions of AVHL and DVC ( S1 Text; S8 Fig ) ., Indeed , our analysis suggests multiple possible functions for both neurons , most notably ventral cord pioneering , chemotaxis , and locomotion for AVHL , and locomotion and ventral cord pioneering for DVC ., We discuss possible biological functions for these neurons further below ., We also note that although ventral cord pioneering is important for axon guidance during neuronal development of the C . elegans connectome , if these neurons remain integrated in the connectome in the fully developed adult , they likely participate in functions other than ventral cord pioneering ., This consideration would then point to chemotaxis and locomotion for AVHL , and locomotion for DVC in the adult neuronal network ., We next attempted to better understand the general characteristics of the critical neurons ., For example , were they the most highly connected ?, Degree does appear important , with 6 critical neurons being in the top 10 for degree ( these 6 being members of the rich club 27 ) ., However , we also found 6 other critical neurons , and none were among the next 18 for degree , with all 12 critical neurons being only in the top 113 ( of 279 total ) ( Tables D and E in S1 Text ) ., To further examine the properties of the critical neurons , we first compared all critical neurons to the noncritical ones on multiple network factors using the Mann-Whitney U test ( Fig 2B ) ., The critical neurons have higher degree ( D ) , strength ( Str ) , average weight ( AW ) , nodal efficiency ( E ( i ) ) , and nodal betweenness centrality ( B ( i ) ) values than the noncritical neurons ., Only the nodal clustering coefficient ( C ( i ) ) showed no differences between the critical and noncritical groups ., Thus , as a group , the critical neurons were more connected , more strongly connected , more closely connected to others in the network ( i . e . , the shortest average path lengths ) , and generally positioned as important control centers regarding the number of shortest pathways passing through them and thus the number of pathways influenced by them ., These results support others that have shown the significance of nodal network properties such as degree and betweenness centrality in the connectome 7 , 8 , 27–29 ., Overall , multiple factors distinguish the critical neurons , but it is important to clarify their impact on each vulnerability measure VC , E , B individually ., One might assume that each vulnerability measure would be most affected by its corresponding nodal property: e . g . , individual neuron values for clustering coefficient ( C ( i ) ) determining global VC ., However , this is not what we found—in fact in no case was the same local nodal property the leading factor influencing global network vulnerability ., To evaluate the relationship of global VC , E , B to nodal network properties of the intact network , we examined correlations of D , Str , AW , C ( i ) , E ( i ) , B ( i ) to VC , E , B without taking the absolute values ( Table F in S1 Text ) ., The S1 Text contains a detailed discussion of the results ., In sum , neurons most critical for network clustering ( VC ) send dual ( or more ) projections to other neurons that project to each other , producing the largest clusters ., Those most critical for global efficiency ( VE ) generally have the most influence—or control—in the network , regarding the number of pathways sharing them and thus the number influenced by them ., Thus , the critical VE neurons do not necessarily have high individual efficiency values , but they do normally have high individual betweenness values ., And those critical for network betweenness ( VB ) also appeared to be those with most control in the network , but the effect is more contextual—it depends on the other possible available routes when the node is lost ( detours ) , the loss in nodal betweenness centrality of the attacked node , and the changes generated in others due to the loss of the attacked connections ., Most critical neurons for VB also tended to be connected to other nodes with high nodal betweenness centrality , creating a linked control structure ., The previous analysis of critical neurons determined the significance of each neuron’s constellation of connections to the network ., An individual synapse analysis provides a finer grain analysis of each constituent element 39 , 40 ., We therefore next conducted a vulnerability edge analysis by attacking each one of the 2 , 990 synaptic connections ( S3 File ) ., Because the criterion used to identify critical neurons ( 3SD or above ) proved too liberal for the synapses ( potentially labeling too many as most critical and obscuring differences among them ) , we used the following procedure ., First , for each list of VC , VE , and VB values ( Fig 3A–3C ) , we transformed the values to the distance from the mean in SD integer units ( i . e . , 1SD , 2SD , etc . ) , and ordered this list from highest to lowest ., Second , we created a histogram distribution for each ordered list and identified the first major change in the slope ( see Fig 3D ) ., Third , we compared the SD values of the first change points for VC , VE , and VB and used the most conservative value ( the highest SD distance ) as the criterion for all three vulnerability measures ., From this procedure , 6SD was identified from the VE values and used as the criterion to identify the critical synapses ( Fig 3A–3D ) ., The analysis identified 7 , 13 , and 17 synapses as critical for VC , VE , and VB , respectively , implicating 29 ( i . e . , ~1% ) critical synapses ( Table 2; Fig 3 ) ., 28 of the 29 synapses are excitatory , suggesting that they promote information transmission ., The one exception , VD01 ( M ) → DVC ( I ) , is inhibitory , which we note further below ., In addition , 26 of the 29 pairs ( 89 . 7% ) contain at least one interneuron , 14 ( 48 . 3% ) contain at least one motor neuron , and 5 ( 17 . 2% ) contained one sensory neuron ., Thus , the critical synapses also appear to relate predominately to interneurons , but also included motor and sensory neurons ., Since inherently costly long-range connections should be minimized , one might expect an overrepresentation of these among the most critical synapses , as they would provide critical waypoints for the dispersed nervous system ., Indeed , we did find a bias for long-range connections as calculated by the direct Euclidean distance of soma positions between two neurons: 16 of the 29 ( 55 . 2% ) , compared to roughly 10% in the entire connectome ( Fig 3E ) 1 , 3 , 4 ., These connections were similar for all vulnerability results: for VC , 4/7 ( 57% ) ; VE , 6/13 ( 46% ) ; VB , 9/17 ( 53% ) ., Finally , regarding connectome laterality , we did find asymmetries between the left and right neuronal pairs , in which certain neuron types were critical only for the left or right ( e . g . , neurons AVER , AVHL , PVPR ) ., Overall , however , we found that there is no clear overall laterality bias in the critical neurons ( 4 left , 5 right ) or synapses ( source node: 10 left , 12 right; sink node: 6 left , 8 right ) ., The left/right differences that we did find , however , warrant future examination of laterality in the C . elegans connectome 41 ., We next examined the vulnerability edge results for each network property VC , E , B more closely ( Table 2 ) ., For VC , of 7 total , 2 were AVA connections to each other ( i . e . , AVAL ( I ) ←→ AVAR ( I ) ) , 4 were AVAL/R ( i . e . , both left and right ) to motor neurons DA07 ( M ) and AS08 ( M ) , and 1 was motor to motor ( VA08 ( M ) → DD04 ( M ) ) ., With respect to biological function , AVA , DAn , and VAn have well-established involvement in backward locomotor control , and DDn and ASn in the coordination of locomotion ( between forward and backward ) ., Thus , the results show that the most dominant synapses that drive coordinated local processing occurs with locomotor control , and in particular , backward locomotion ., The analysis of VB identifies 17 critical synapses with the largest influence on betweenness centrality: 15/17 ( 88% ) contain at least one interneuron in the pair , 6/17 ( 35% ) at least one motor neuron , and 3/17 ( 18% ) have one sensory neuron , with 7 of the 17 ( 41% ) links being interneuron to interneuron ( I → I ) , and the remaining 10 being some combination of sensory , interneurons , and motor neurons ., To better understand the significance of these critical segments in the network for overall betweenness centrality , we examined their distribution ., If these segments were independent , they should be fairly evenly distributed throughout the connectome ., If , however , there were an interdependency among them , some pattern may emerge ., In fact , we found that the synapses formed three separate pathways ( see blue lines in Fig 3E ) ., To see these more clearly we collapsed left and right neuron pairs ( e . g . , AVAL or AVAR as AVA ) ., Justification for the left/right collapsing is taken from the biological work on the C . elegans , which normally shows symmetric function for neuron pairs 3 , 6 ., As shown in Fig 4A , the first AVA-based critical pathway begins with RIB ( I ) → AVE ( I ) → AVA ( I ) and has two routes after this ., Route, ( a ) ( from RIB ( I ) to VD01 ( M ) ) is in fact a significant segment of the well-established circuit for backward locomotion ., Although the role of route, ( b ) ( RIB to PVC ) has been suggested to be secondary coordination of backward with forward locomotion , the VB analysis nonetheless suggests that this is an important pathway for locomotor control ., As shown in Fig 4B , the second PVP-based critical pathway begins with AQR ( S ) ←→ PVP ( I ) and has two routes after this ., For route, ( a ) , the results suggest a critical sensory ( AQR ) to motor ( VD01 ) circuit ., To date , AQR has been implicated in aerotaxis ( O2 and CO2 ) , regulating social feeding ( i . e . , behavior when individuals aggregate at bacterial patches ) and bordering behavior ( i . e . , aggregation in densest part of bacterial patch ) , and suppressing innate immunity 5 , 42–44 ., Since evidence shows that aerotaxis also regulates social feeding 43 , we summarize these functions as the regulation of ( 1 ) social feeding 42 , 43 and ( 2 ) internal immunity responses 44 ., The motor neuron VD01 is implicated in locomotion 5 , 6 , 15 ., Given that route, ( a ) ends in locomotor behavior , the results suggest that the identified circuit is involved in social feeding ., Moreover , the results thus predict involvement of both PVP and DVC in this behavioral circuit , whose functional roles in the fully developed adult network to date remain unclear ( although there is evidence for PVP involvement in ventral cord pioneering during development; and there is new evidence for DVC involvement in locomotion ) 38 , 45–47 ., More specifically , if we use the general behavioral circuit structure as a guide , S --> I1 --> I2 --> M , it provides further suggestive evidence for the functions of PVP ( I1: sensorimotor integration ) and DVC ( I2: information integration , motor control ) 7 , 14 ., The behavioral function of route, ( b ) ( AQR ( S ) ←→ PVP ( I ) → AVH ( I ) ) is less clear ., However , since route, ( a ) implicates PVP in social feeding behavior , this may also implicate route, ( b ) in the same , suggesting a possible functional role for AVH , whose function currently remains unknown ., The more local analysis of the potential biological functions of AVH and DVC based on the functions of their neighbors ( reported in the S1 Text ) also appears to further corroborate these findings by implicating both AVHL and DVC in locomotor control ( among other possible functions , which is also to be expected ) ., Finally , the VB analysis also identified the RMD ( M ) → OLL ( S ) link as a critical pathway segment ( Fig 4C ) ., OLL has been implicated in aversive stimulus sensation and RMD in avoidance responses of the nose/head ., Interestingly , the motor to sensory link as an important pathway highlights the loop structure in the C . elegans circuitry , with the motor to sensory link closing the loop 7 ., This loop structure component suggests that multiple motor signals likely converge on sensory processing , suggesting significant action-influenced perception ., Indeed , given the bidirectional links in the two main pathways between motor and interneurons , interneurons among themselves , and interneurons and sensory neurons , the VB results suggest that these are important recurrent connections in the connectome circuitry , and further suggest that recurrent feedback loops for central processing ( middle layers ) and action influences on perception are important general principles of nervous system function 7 ., The VE synapse analysis identified 13 critical links with the greatest effect on global efficiency , with all nodes being interneurons or motor neurons , and all connections containing at least one interneuron and over ½ at least 1 motor neuron: 6 I → I , 4 I → M , 3 M → I ( see red lines in Fig 3E ) ., The interneuron-to-interneuron links are expected to be critical information processing structures in the network , and the others show the significance of motor control ., More specifically , two critical pathways were identified , which were again AVA- and PVP-based ( Fig 5A and 5B ) ., Thus , the VE results show that components of both the AVA- and PVP-based pathways are important control segments that have the greatest influence on global network efficiency ., Combining the results for all critical neurons , synapses and vulnerabilities , they appear to converge on three main pathways ., These pathways are most clearly defined by the VB results ( Fig 4A ) , which identify the synapses with the largest effect on network betweenness centrality , the VE results then highlight the components with the largest effect on global efficiency , and the VC results show where the most critical clustering occurs ., First , there is an AVA-based pathway , beginning with RIB ( I ) → AVE ( I ) , and ending with either motor neurons or PVC ., This pathway strongly matches the backward locomotion control circuit , attesting to the significance of backward locomotion in the C . elegans nervous system ., Second , there is a PVP-based pathway ., This pathway is composed of sensory ( AQR ) , interneurons ( PVP , DVC , AVH ) , and a motor neuron ( VD01 ) ., The sensory to motor structure reveals a complete behavioral circuit , and implicates social feeding behavior ., The third critical pathway segment is RMD ( M ) → OLL ( S ) that again identifies avoidance control as important—in this case , aversive stimulus avoidance with respect to nose/head ., The motor to sensory directional link also highlights the importance of the loop structure in the C . elegans circuitry 7 ., To better understand the nature of the critical synapses , we examined the two main local properties of edges: strength ( Str ) ( i . e . , the weight of the given synapse ) and edge betweenness centrality ( EBC , a betweenness measure for edges; see S1 Text ) ., We first compared the critical and noncritical synapses on Str and EBC using the Mann-Whitney U test ( Fig 4C ) , and the critical synapses indeed showed significantly higher Str and EBC than the noncritical synapses ., Thus , like the critical neurons , the critical synapses have stronger connections and were those with the most shortest paths running through them—i . e . , with the highest volume traffic ., As stated previously , constituents with such high volume traffic can be considered as control structures by virtue of their influence over the multiple pathways passing through them ., To clarify how Str and EBC related to each vulnerability measure VC , E , B , we next examined the correlations ( without absolute value for VC , E , B ) , and all were statistically significant ( Table F in S1 Text ) ., The S1 Text contains a detailed discussion of the results ., In sum , like neurons , the synapses most critical for betweenness centrality generally have the highest control themselves , and also tend to link control nodes together ., For global efficiency , also like neurons , they generally are those with the highest control in the network , regarding the number of pathways sharing them ., There were , however , five exceptions for global efficiency and betweenness centrality ( detailed in the S1 Text ) , and they suggest that , like betweenness centrality , global efficiency can also be critically affected by context ., Finally , for clustering ( VC ) , critical synapses appeared to participate in shared projections to neighbors , producing clusters ., Function derives from form , and with unrelenting natural selection , nervous systems must achieve complexity as efficiently as possible ., Network analysis provides a normative approach to analyze the solution , and vulnerability analysis helps provide a more experimental and less assumption-laden analysis of individual component contributions to information processing ., In this study , we tested the effects of attacks on every neuron and synaptic connection in the C . elegans connectome to identify and characterize the most critical constituents of the network ., To our knowledge , this is the first study to analyze network robustness for individual node and edge attacks on an entire nervous system ., We identified 12 neurons and 29 synapses critical for clustering , information integration and propagation ., Although one might have expected the clustering , efficiency , and betweenness values for individual neurons and synapses to be the most important factors determining these same properties at the network level , control structures—i . e . , those that influence multiple others—prove the most important ., From an evolutionary-development ( evo-devo ) perspective , the additions of higher-order control structures can produce the largest effects on information processing , providing the most ‘bang for the buck’ ., Thus , to most affect clustering , new neurons should project paired synapses to neighbors ., To most affect global efficiency , new neurons or synapses should be placed in central positions that shorten the largest number of prior existing pathways ., These more centralized neurons not only contribute to overall efficiency and information integration , they also become traffic centers , and thereby have greater influence ( and thus control ) in the connectome ., Finally , to most affect network traffic flow ( i . e . , global betweenness ) , new neurons or synapses should provide new pathways that decrease a larger number of path lengths and should also normally link to other control structures , leading to the formation of chains of control units ( i . e . , node1 → node2 → node3 ) ., Not only were there critical control structures at th | Introduction, Results, Discussion, Methods | Determining the fundamental architectural design of complex nervous systems will lead to significant medical and technological advances ., Yet it remains unclear how nervous systems evolved highly efficient networks with near optimal sharing of pathways that yet produce multiple distinct behaviors to reach the organism’s goals ., To determine this , the nematode roundworm Caenorhabditis elegans is an attractive model system ., Progress has been made in delineating the behavioral circuits of the C . elegans , however , many details are unclear , including the specific functions of every neuron and synapse , as well as the extent the behavioral circuits are separate and parallel versus integrative and serial ., Network analysis provides a normative approach to help specify the network design ., We investigated the vulnerability of the Caenorhabditis elegans connectome by performing computational experiments that, ( a ) “attacked” 279 individual neurons and 2 , 990 weighted synaptic connections ( composed of 6 , 393 chemical synapses and 890 electrical junctions ) and, ( b ) quantified the effects of each removal on global network properties that influence information processing ., The analysis identified 12 critical neurons and 29 critical synapses for establishing fundamental network properties ., These critical constituents were found to be control elements—i . e . , those with the most influence over multiple underlying pathways ., Additionally , the critical synapses formed into circuit-level pathways ., These emergent pathways provide evidence for, ( a ) the importance of backward locomotion , avoidance behavior , and social feeding behavior to the organism;, ( b ) the potential roles of specific neurons whose functions have been unclear; and, ( c ) both parallel and serial design elements in the connectome—i . e . , specific evidence for a mixed architectural design . | One of the most important scientific aims is to uncover the functional design principles of nervous systems ., To reach this aim , it is useful to examine a complex nervous system that is both well specified and highly tractable , making the nematode roundworm Caenorhabditis elegans an attractive model system , especially since it is the only complete connectome currently available for analysis ., In this computational study , we tested the effects of individual attacks on every neuron and synaptic connection in the C . elegans connectome to identify and characterize the most critical constituents of the network by quantifying the changes in key network properties of the connectome that influence information processing ., Our analysis identified 12 neurons and 29 synapses critical to clustering , information integration and propagation ., These critical constituents formed circuit-level structures that control network processing in the C . elegans connectome ., We believe our study provides a significant advance in the understanding of the network topology of the C . elegans connectome , and provides insights into the fundamental architectural design of complex nervous systems . | invertebrates, medicine and health sciences, neural networks, nervous system, caenorhabditis, electrophysiology, neuroscience, animals, motor neurons, animal models, caenorhabditis elegans, model organisms, brain mapping, interneurons, research and analysis methods, computer and information sciences, animal cells, connectomics, sensory neurons, cellular neuroscience, neuroanatomy, cell biology, anatomy, synapses, physiology, neurons, nematoda, biology and life sciences, cellular types, neurophysiology, organisms | null |
journal.pcbi.1003101 | 2,013 | PUMA: A Unified Framework for Penalized Multiple Regression Analysis of GWAS Data | Genome-wide association studies ( GWAS ) have identified many susceptibility loci underlying the molecular etiology of complex diseases 1 ., These studies have been responsible for the discovery of many individual genes that contribute to disease risk 2–10 , for discoveries on the front line of personalized medicine 11 , 12 , and for discovering novel pathways important for the progression of complex heritable diseases 13 ., The expense of each GWAS that is capable of finding well-supported disease loci is considerable and , as a consequence , each robust and interpretable association discovered in a GWAS is valuable , not only from the point of view of scientific discovery but also in terms of return on investment 14 , 15 ., A clear picture that has an important bearing on the investment-discovery tradeoff in GWAS experiments is that the associations identified to date generally explain only a small to moderate fraction of total heritability 16 , 17 ., Recent analyses have suggested that a considerable amount of this ‘missing’ heritability can be accounted for by rare variants or variants with weak effects 18–20 ., This suggests that there is an opportunity to identify more risk loci through studies that require even greater investment , by including larger sample sizes and/or by incorporating higher genetic marker coverage of the genome by using next-generation sequencing ( NGS ) ., The novel associations discovered by large consortia GWAS studies support this supposition 7–10 ., Another complementary strategy that leverages both the current and future investment in GWAS experiments is the application of new statistical analyses that can reliably identify weaker associations 21–25 ., Although there has been an explosion of methods in this area 26 , 27 , few have produced robustly supported associations that are not detectable by single marker tests of association 1 , 26–29 ., Here , we report a general framework for applying a family of GWAS analysis methods that is extremely promising for detection of weak associations yet has not been widely applied to learn novel biology from GWAS datasets: penalized multiple regression ( PMR ) methods ., PMR methods work by simultaneously incorporating tens to hundreds of thousands of genetic markers in a single statistical model where a penalty is incorporated to force most marker regression coefficients to be exactly zero , so that only a small fraction are estimated to make a contribution to disease risk 22 , 30–39 ., By jointly analyzing markers , PMR methods are able to consider the correlation of each marker with the phenotype , conditional on all other relevant markers ., This can increase the power to detect weak associations compared to single marker methods due to the smaller residual variance and the fact that the conditional correlation of a marker with the phenotype can often be substantially higher than the marginal correlation 40 ., The latter effect is a consequence of non-zero correlation structure between associated markers when the underlying genetic architecture is polygenic 41 ., These methods therefore model the underlying biology more accurately than single marker tests , by explicitly modeling the polygenic architecture of complex phenotypes to account for the effects of multiple susceptibility loci ., They also leverage the same type of statistical model used in single marker testing methods that have demonstrated reliability in the identification of strong associations 1 , 28 , 29 ., Yet , despite theoretical power of PMR methods , the large body of statistical literature exploring their theoretical properties ( see reviews 42 , 43 ) , and the recent interest in the methods development community 22 , 30–39 , these methods have not been successful in GWAS analysis ., This is due to a combination of limitations:, 1 ) inability to scale for very large GWAS datasets 22 , 32 , 34 , 35 ,, 2 ) poor performance on simulated data 22 , 31 ,, 3 ) they often find too many ‘hits’ to be biologically plausible for a given GWAS sample size 22 , and, 4 ) they do not identify novel , well-supported associations that are not detectable by standard methods 22 , 31 ., In order to address these issues , we present a combined algorithmic and heuristic framework for PUMA ( Penalized Unified Multiple-locus Association ) analysis that optimizes these methods for reliable detection of weak associations when applied to large GWAS datasets ., The complete PUMA framework includes an extremely efficient implementation of a new minorize-maximization ( MM ) algorithm 44 for generalized linear models ( GLM ) 45 , a theoretically motivated data-adaptive heuristic approach to determine penalty strength and for model selection , and post hoc methods for assessing the rank of identified associations ., Within PUMA , we implement all sparse feature selection , penalized regression approaches proposed for GWAS analysis to date , including four penalties implemented in a maximum likelihood framework ( i . e . Lasso , Adaptive Lasso , NEG , MCP ) , as well as theoretically justified penalties that have not been previously applied to GWAS ( i . e . LOG ) ( Figure 1 ) ., We demonstrate the power of our framework for detecting weaker associations that are invisible to individual marker testing through analysis of simulated GWAS data that mirror observations from analyses of real GWAS data ., We also demonstrate that our approaches correct issues with all current PMR methods where software is available for GWAS analysis , where we find that all of these currently available PMR GWAS methods can perform worse than single marker testing for our simulation conditions ., As an illustration of the value of PUMA for mining existing GWAS data for novel associations , we apply these methods to the original Wellcome Trust Case Control Consortium ( WTCCC ) 2 GWAS datasets for type 1 diabetes , Crohns disease and rheumatoid arthritis ., Our re-analysis identifies weak associations that implicate additional susceptibility loci for these autoimmune diseases , which did not appear significant by standard single marker tests of association in these datasets , yet were, 1 ) identified in an independent GWAS of the same phenotype that did not include WTCCC data ,, 2 ) previously known to play a role in disease etiology , or, 3 ) known to function in a relevant biological pathway ., Our results demonstrate that appropriately tuned PMR methods can provide a complementary approach to large meta-analyses 4–10 to identify susceptibility loci with weak associations ., We also provide a discussion concerning how the framework can be extended to perform penalized analysis of epistasis , to incorporate mixed model analysis , and to address challenges of genome-wide genotypes provided by whole-genome next-generation sequencing ., The methods implemented in our PUMA framework are orders of magnitude faster than existing software when assigned identical computational tasks and no pre-screening of markers is performed ( Table 1 ) ., This substantial boost in computational speed allows PUMA to perform a dense two-dimensional search of tuning parameter values for non-convex penalties ( i . e . MCP , NEG , LOG ) and examine upwards of 1 million total modes of the likelihood surface for simulated case/control dataset of 5 , 000 individuals and 650K genetic markers in less than 24 hours on a 6 core Intel® Xeon® W3690 @ 3 . 47 GHz with 12 Gb memory when a pre-screening p-value cutoff of 0 . 01 from single marker analysis is applied ( Table 2 ) ., This is a huge improvement compared to existing software for non-convex PMR methods 22 , 32 which only examine a single mode ., While pre-screening markers based on a p-value cutoff may initially seem to detract from the purpose of a multiple-locus analysis , it is supported by statistical theory , is necessary for large scale analysis and has almost no impact on the set of markers identified as associated ., In a seminal paper , Fan and Lv 46 demonstrate that pre-screening by ranking the marginal correlation of each variable with the response will retain the relevant variable asymptotically with probability tending to 1 ., Fan and Song 43 extend this result to generalized linear models ., Moreover , Tibshirani , et al . 47 and El Ghaoui , et al . 48 establish exact pre-screening methods for linear and logistic Lasso models where relevant variables are guaranteed to be retained for finite sample sizes and demonstrate that the number of variables can be reduced by up to 3 orders of magnitude ., Intuitively , both the asymptotic 43 , 46 and exact pre-screening methods 47 , 48 rely on the fact that a variable is unlikely to have a very small marginal correlation with the response but a large and very significant conditional correlation for a particular sample size when the relevant variables explain only a small fraction of the variation in the response ., Moreover , pre-screening is often computationally necessary because storing 650 K markers for 5000 samples requires 26 Gb of memory ., Finally , we note that pre-screening is used by previous applications of PMR methods to GWAS data 22 , 31 in order to handle genome-scale data ., We use a pre-screening p-value cutoff based on single marker analysis , because, 1 ) it retains all relevant variables asymptotically 43 , 46 ,, 2 ) it approximates the exact methods proposed for Lasso 47 , 48 , which cannot be easily adapted to other penalties ,, 3 ) it reduces memory requirements so that very large datasets can be analyzed on a high-end desktop computer ,, 4 ) it substantially reduces the computational burden ,, 5 ) by using a p-value it is naturally calibrated to the sample size and the fraction of variation in the response being explained , and, 6 ) it has very little empirical effect on the results ., We demonstrate this final and most important point in two complementary simulation studies ., First we consider a simple two-step forward regression method , which is known to approximate penalized multiple regression 49 , 50 and , under a range of biologically motivated simulation conditions , demonstrate that variables that do not cross an initial p-value threshold have a very low probability of being significant in the second step ( Figure S1 ) ., Second we demonstrate that the pre-screening has no noticeable effect on the performance of Lasso and MCP methods but substantially reduces the computational time ( Figure S2 ) ., We analyzed 960 simulated GWAS datasets to assess the performance of our PUMA framework compared to other published methods for PMR GWAS analysis ., We note that these simulations , while far more extensive than other published works on PMR GWAS analysis 22 , 30–38 are not meant to be exhaustive or to capture all the possible complexities in a GWAS but rather to:, 1 ) serve as a baseline for comparing GWAS analysis methods and, 2 ) provide an estimate of the expected performance for these methods when applied to GWAS data under relatively ideal experimental conditions ., Our goal therefore was not to attempt to model a broad spectrum of possible GWAS data complexities ( e . g . stratified experimental sampling schemes , known or cryptic population structure effects on phenotype , relatedness among individuals , measured or latent covariates , etc . ) but rather to simulate data that captured the most basic components of a GWAS experiment ( see Methods for details ) ., In simulated data a causal variant is defined as a variant whose coefficient value is nonzero , so that number of minor alleles at this marker contributes to the phenotype ., In order to mimic the fact that true causal variants are not available from array-based genotyping , the simulated causal variants were removed from the dataset so that they are not considered by the tests of association ., Therefore , just like in all array-based genotyping datasets , our simulations identify associations based on markers in linkage-disequilibrium with the ( omitted ) causal variant ., We assessed the performance of PMR methods for which there is available software ., We compared the performance of the Lasso penalty from Wu , et al . 31 , the NEG penalty as implemented in the HyperLasso program 22 , and a permutation-based approach to selecting tuning parameter values for the MCP penalty 32 , 51 that we term perm-MCP ., We note that we only considered PMR approaches that are designed to handle the specific challenges of GWAS data and that also perform feature selection , such that we do not consider ridge , elastic net , or group-penalties since they set many correlated markers to have nonzero coefficients and thus complicate the generation of interpretable p-values 30 , 52 ., We also did not consider Markov Chain Monte Carlo ( MCMC ) approaches 34 , 35 since they could not efficiently scale to genome-wide data while exploring a range of tuning parameter values ., We ran the HyperLasso program 22 with standard settings ( see Text S1 ) ., We applied the method of Wu , et al . 31 , setting the number of selected markers to the true number of causal markers in each simulation since Wu , et al . 31 do not specify a criterion for selecting the model size ., As a benchmark , we also ran a single marker analysis implemented by applying a logistic regression model to each marker individually ., We used a pre-screening p-value cutoff of 0 . 01 from single marker analysis for the PMR methods to make them computationally tractable ., Simulations indicate that HyperLasso 22 and the Lasso of Wu , et al . 31 are generally less powerful than a standard single marker test ( Figure 2 , S3 , S4 , S5 , S6 , S7 , S8 , S9 , S10 ) ., While Lasso is sometimes comparable or slightly more powerful than a single marker test for low FDR , the performance of the method benefits from the fact that the number of selected markers is set using information not available in real data ., Setting the marker number to 10 ( the default in the implementation of Wu , et al . 31 ) or another arbitrary value results in poor performance and is not competitive with a single marker test ( results not shown ) ., The performance of HyperLasso is especially poor as is it suffers from the fact that the choice of tuning parameters has a huge effect on performance , but the method does not implement a search over tuning parameter values ., Moreover , HyperLasso does not include a way to evaluate the significance of a selected marker , so we used their default approach of using coefficient values from selected markers to assess performance ., Alternatively , perm-MCP was the most powerful in our simulations ., We note that for perm-MCP , by setting the expected false positive rate ( eFPR ) and using permutations to obtain the value of the tuning parameter based on this rate , perm-MCP generates a single model with relatively few nonzero coefficients while explicitly addressing the multiple testing problem ., Yet in practice this result indicates that perm-MCP may assign p-values to only a handful of markers so that the method may not identify any novel associations for a particular dataset ., Since the number of nonzero coefficients is directly related to the specified eFPR and the pre-screening cutoff , we examined multiple eFPR values ( , , , , ) and cutoff values ( 0 . 1 , 0 . 01 , 0 . 001 ) , and selected the values the yielded the highest power ( eFPR\u200a= , cutoff\u200a=\u200a0 . 001 ) to present in Figure 2 , where other cutoff combinations produce poorer performance ( see Figure S11 for a representative plot showing the results for all cutoffs ) ., We note that the eFPR value is based on the number of markers that pass the pre-screening cutoff , not the total number of markers ., Therefore the performance of perm-MCP is sensitive to the eFPR and cutoff values , yet there is no clear method to optimally specify this value a priori ., Furthermore , determining the appropriate cutoff for a desired eFPR for correlated high-dimensional data is the subject of current research 53 , and its application to permutation methods for selecting a tuning parameter remains an open question ., We also note that the performance achieved with PUMA methods does not require the optimal determination of eFPR and pre-screening cutoffs ., In addition , we note that while Ayers and Cordell 32 have previously shown that penalized regression methods can perform well on simulated data , the datasets we address here are orders of magnitude larger ., Ayers and Cordell 32 conducted two simulation studies , one with 4000 markers and the other with no more than 228 ., By considering such a small set of markers , which is not the product of a pre-screening step , they were able to use standard R packages and apply a permutation method to select tuning parameters on the full dataset ., Moreover , the multiple testing problem is less severe in their analysis ., For the HyperLasso program , Ayers and Cordell 32 selected the tuning parameter as described by Hoggart , et al . 22 ., However , using these settings for the genome-scale datasets examined here caused the HyperLasso program to crash ( Text S1 ) and so we use the default program settings ., We note that the program worked as expected for smaller datasets ., It is unclear whether this problem is an issue with the underlying algorithm or the specifics of the implementation ., Thus the difference between the performance of methods in Ayers and Cordell 32 and the current study is the scale of the data , the large multiple-testing burden for genome-scale data and the necessity of a pre-screening step for genome-scale data ., PUMAs statistical power is due to its data-adaptive properties ., PUMA, 1 ) performs a two dimensional search of the tuning parameter space, 2 ) selects the number of nonzero coefficients based on both the fit to the data and the sample size , and, 3 ) uses a heuristic methods to assess the significance of correlated markers ., Conversely , perm-MCP fixes one of the tuning parameters , does not incorporate the sample size , and does not address the issues of testing the significance of correlated markers ., Moreover , perm-MCP relies on setting the eFPR despite problem of determining an appropriate value a priori for high dimensional data ., For the 960 simulated GWAS datasets we analyzed , almost all PMR GWAS approaches implemented in PUMA except NEG and adaptive Lasso outperformed single marker analysis under simulation conditions with sufficient sample size ( Figure 3 , Figures S3 , S4 , S5 , S6 , S7 , S8 , S9 , S10 ) ., Quite critically , the performance is far greater even when using a conservative control of FDR that is commonly employed in GWAS studies ., Moreover , the improvement of the PMR methods in PUMA is most noticeable for causal variants with intermediate marginal heritability ., Overall , these simulations demonstrate that the advantage of PMR methods over a single marker test increases with sample size , but decreases with the number of susceptibility loci ( Figure S3 , S4 ) ., While the penalized methods implemented in our PUMA framework consistently had higher power than single marker analysis as a function of FDR under most simulation conditions , none of the penalties consistently stood out as the most powerful ., However , our PUMA framework , which includes a fast novel algorithm for penalized maximum likelihood estimation in generalized linear models , data-adaptive tuning of tuning parameters , heuristics for model selection and novel method of assigning p-values ( see Methods ) increased the power of PMR methods compared to current approaches using the same penalties 22 , 31 ., We note that our implementation of the NEG penalty showed a substantial increase in power over the HyperLasso program 22 and indicates that our search over tuning parameter values and heuristic approach for model selection was successful ., Moreover , our search of one or both tuning parameter values for MCP ( termed 1D-MCP and 2D-MCP , respectively ) showed that our approach to applying MCP ( i . e . 2D-MCP ) can be more powerful than that of Ayers and Cordell 32 ., The fact that our implementation of Lasso had higher power than the version of Wu , et al . 31 confirms the usefulness of our data-adaptive approach for selecting penalty strength and our novel method for assigning p-values ., We also note that for comparison we applied a conditional regression test and our previously published algorithm VBAY , a variational Bayes approach for fitting a mixture prior penalty 33 ., We found that perm-MCP and VBAY had similar performance to our PMR methods and while the conditional test of association was sometimes more powerful than single marker analyses it was generally not as powerful as the PUMA PMR methods ., In our re-analysis of type 1 diabetes , Crohns disease and rheumatoid arthritis datasets , we applied a single-marker analysis and all PMR analysis approaches ( Lasso , Adaptive Lasso , NEG , LOG , 1D-MCP , 2D-MCP , perm-MCP ) using all the recommended components of our framework ., We included sex and the first two principal components as unpenalized covariates , applied a pre-screening cutoff of 0 . 01 on the p-values from the single marker test , and ran 100 reorderings for the non-convex penalties ., Quantile-Quantile ( QQ ) plots of p-values from a standard single marker analysis indicate that the effects of any remaining population structure is minimal ., Moreover , including the subset of significantly associated markers identified by the PMR methods as covariates in a single marker analysis of remaining markers does not yield an inflation of the QQ plots and thus indicates that the PMR methods are not overfitting the data ( Figure S12 ) ., We also note that due to the complex LD around the MHC on chromosome 6 , while we included this region in our analysis , we omit this region from any post hoc analysis and discussion ., Our single-marker re-analysis of type 1 diabetes , Crohns disease and rheumatoid arthritis datasets reproduced the same associations as reported in the original analysis ( Figure S13 ) ., Our PMR methods recapitulated almost all of the associations identified by single marker analysis , although there were differences among the methods ., The PUMA Lasso and Adaptive Lasso identified almost no additional associations compared to single marker tests , and while LOG , NEG and 1D-MCP identified more , almost all of the associations found by these five methods ( Lasso , Adaptive Lasso , LOG , NEG , 1D-MCP ) were identified by 2D-MCP ( Figure 4 , S14 ) ., We note that perm-MCP identified very few associations ( 12 overall , across the three diseases ) , all but one of which was identified by a single marker test , and all were identified by 2D-MCP ., We therefore discuss the associations found by 2D-MCP , where we consider three categories of interest ( Table 3 ) : those concordant with single marker tests , those that recapitulate associations identified in external GWAS studies but not by single marker analysis of the WTCCC , and novel associations , of which many were deemed to be biologically interpretable in terms of the current knowledge of disease etiology ., In the absence of functional validation , the presence of a feasible biological interpretation lends more credibility to these novel findings ., A critical point to note about the performance of our PUMA framework for PMR analysis of GWAS data is that these methods not only result in the correct identification of more loci than a single marker testing analysis ( when controlling the false discovery rate at the same level ) , but also lead to re-orderings of the rank of markers that are considered the most significant when compared to a single marker analysis ( Figure 5 ) ., As a consequence , we are able to identify etiologically relevant and replicated disease loci that are too weak to be detected by single marker analysis , yet show strong signals of association by PMR analysis ., This means that our PMR GWAS analysis is not simply taking advantage of the lower residual variance to improve performance , but is also taking advantage of the fact that conditional correlation of a relevant marker with the phenotype is often more significant than the marginal correlation ., When the coefficients for multiple markers , each tagging different susceptibility loci throughout the genome , have nonzero values in the PMR framework , their association with the phenotype becomes more significant ., Our framework can therefore identify disease susceptibility loci in a GWAS with weak associations with phenotype , when they are invisible to a single marker testing approach ( i . e . they have p-values in a single marker test that would never be considered significant ) ., The associations identified by PUMA generally recapitulate associations identified by single marker analysis , and the PUMA hits have perfect concordance for strong associations ., Overall 2D-MCP recapitulates the largest number of associations , while the union of the other PMR methods ( considered here for illustrative purposes due to the high degree of concordance with each other , and the fact that 2D-MCP identifies almost all of the associations they find ) had a lower degree of concordance with the single marker analysis ( Figure 4 , 6 , S14 , Table S1 ) ., Of the 6 associations identified by a single marker analysis that were missed by our methods , 5 were from type 1 diabetes and 1 was from Crohns disease ( Table S2 ) ., One of these associations was borderline significant by 2D-MCP with a p-value of 1 . 42×10−7 ., We compared associations identified by our PUMA methods that were not detected by single marker tests in the WTCCC dataset to markers identified by independent studies in the HuGE database of published GWAS 53 in order to find associations identified in both our analysis and an independent study that did not include WTCCC data ., Such replications are considered the gold standard for validating a putative association 54 ., In the ideal case the same marker would show an association in both the WTCCC dataset and those summarized in the HuGE database ., However , given, 1 ) the lack of overlap of marker-sets between genotyping platforms ,, 2 ) that the HuGE database reports only the most significant marker in an associated LD block , and, 3 ) that PMR methods tend to select only a single marker within a LD block , we considered a marker to recapitulate a known association if the two are within 0 . 1 cM 6 ., A representative example from Crohns disease is shown in Figure 7 where only 2D-MCP is able to identify STAT3 as a susceptibility locus in the WTCCC dataset ( Figure 7a ) ., While this association has also been replicated in non-independent datasets 6 , which included WTCCC data , the role of STAT3 in Crohns and other autoimmune disease is well established 55 , 56 ., While all PUMA methods and a single marker test are able to replicate associations from independent studies , LOG , NEG and 1D-MCP , stood out in terms of identifying associations replicated by non-independent studies , but not detected in the WTCCC dataset by a single marker analysis ., These counts reflect the results when the number of markers considered as ‘hits’ was set to be equal across methods so that they reflect the ordering of markers by PMR methods rather than the number of associations ., When comparing the total number of significant hits from each method to associations identified in either independent studies or non-independent external studies that incorporated WTCCC data , 2D-MCP is the only PMR method to identify as many total replicated associations as a single marker test ( Tables 4 , S3 , S4 , Figures S15 , S16 , S17 , S18 , S19 ) ., However , 2D-MCP is able to replicate known associations that cannot be replicated by a standard single marker test in this dataset , thus demonstrating that PMR methods can extract biologically relevant information that is overlooked by standard analyses ( Table S5 ) ., These results demonstrate that PMR methods overall are able to identify replicated associations in this dataset that are invisible to a standard single marker test ., Moreover , our methods provide an opportunity to replicate previously unreplicated associations by re-analyzing existing GWAS datasets ., Re-analysis of type 1 diabetes , Crohns disease and rheumatoid arthritis datasets from the original WTCCC 2 with our PUMA methods revealed novel associations that have not been identified in previous GWAS of these diseases ( Table 5 , Figures S14 , S20 , S21 , S22 ) ., These methods , most notably 2D-MCP , identify novel associations in or near genes which have been previously associated with etiologically related diseases or which are known to function in biologically relevant pathways based on public databases and disease literature ( Tables 5 , 6 ) ., In addition , PUMA also identified associations without a clear biological link to the disease phenotype ( Tables S6 , S7 , S8 ) ., PUMA methods identified novel susceptibility loci for type 1 diabetes involved in pancreatic function , insulin pathways and immune cell function and for Crohns disease that are involved in pro- and anti-inflammatory pathways ( Table 6 ) ., 2D-MCP identified a gene functioning in apoptosis as a susceptibility locus for rheumatoid arthritis ( Table 6 ) ., These genes are known to function in relevant pathways or have been previously implicated in the etiology of the disease but have not been found by previous GWAS of each disease ., A representative example is shown in Figure 7b where only 2D-MCP identifies an association that implicates SLC30A1 ., This gene is a zinc transporter related to SLC30A8 , which has been implicated in type 2 diabetes , and zinc transport plays a role in insulin secretion by pancreatic -cells 57 , 58 ., Each GWAS discovery that has a well supported association produces valuable information for understanding the etiology of the disease phenotype and such discoveries are regularly used as the foundation for studies that use the locus as a starting point 59 , 60 ., Given that GWAS involving a thousand to several thousands of individuals seldom return more than a few to a dozen well-supported associations ( depending on the disease ) the monetary , time , and resource investment in these studies often translates to a considerable expenditure per discovery ., This is true even when considering additional discoveries that may occur as individual GWAS are combined together into large meta-analysis studies 4–10 ., We have demonstrated that our PUMA framework has the potential to produce added investment return for GWAS studies by discovering additional well-supported disease loci associations that are invisible to the standard single marker analysis methods responsible for almost all reported GWAS 1 , 53 ., For example , our re-analysis of type 1 diabetes , Crohns disease and rheumatoid arthritis from the original Wellcome Trust Case Control Consortium ( WTCCC ) 2 demonstrates that PUMA methods can identify associations that are not detectable by single marker analysis approaches but which replicate associations known from independent studies , which did not include WTCCC data , as well as novel loci with strong links to known disease etiology ., These included 10 novel associations identifying genes that are linked to primary pathways of these autoimmune diseases , specifically 6 genes involved in pancreatic function , insulin pathways and immune-cell function in type 1 diabetes; 4 genes ( in 3 association regions ) functioning in pro- and anti-inflammatory pathways in Crohns disease; and 1 gene involved in apoptosis pathways in rheumatoid arthritis ., Applying our PUMA framework therefore has the potential to add a significant number of discoveries for a given GWAS ., A critical property of our PUMA framework is it does not return the same ordering of significant markers produced by a standard single marker analysis ., By simultaneously accounting for the associations of multiple loci and better refle | Introduction, Results, Discussion, Methods | Penalized Multiple Regression ( PMR ) can be used to discover novel disease associations in GWAS datasets ., In practice , proposed PMR methods have not been able to identify well-supported associations in GWAS that are undetectable by standard association tests and thus these methods are not widely applied ., Here , we present a combined algorithmic and heuristic framework for PUMA ( Penalized Unified Multiple-locus Association ) analysis that solves the problems of previously proposed methods including computational speed , poor performance on genome-scale simulated data , and identification of too many associations for real data to be biologically plausible ., The framework includes a new minorize-maximization ( MM ) algorithm for generalized linear models ( GLM ) combined with heuristic model selection and testing methods for identification of robust associations ., The PUMA framework implements the penalized maximum likelihood penalties previously proposed for GWAS analysis ( i . e . Lasso , Adaptive Lasso , NEG , MCP ) , as well as a penalty that has not been previously applied to GWAS ( i . e . LOG ) ., Using simulations that closely mirror real GWAS data , we show that our framework has high performance and reliably increases power to detect weak associations , while existing PMR methods can perform worse than single marker testing in overall performance ., To demonstrate the empirical value of PUMA , we analyzed GWAS data for type 1 diabetes , Crohnss disease , and rheumatoid arthritis , three autoimmune diseases from the original Wellcome Trust Case Control Consortium ., Our analysis replicates known associations for these diseases and we discover novel etiologically relevant susceptibility loci that are invisible to standard single marker tests , including six novel associations implicating genes involved in pancreatic function , insulin pathways and immune-cell function in type 1 diabetes; three novel associations implicating genes in pro- and anti-inflammatory pathways in Crohns disease; and one novel association implicating a gene involved in apoptosis pathways in rheumatoid arthritis ., We provide software for applying our PUMA analysis framework . | Genome-wide association studies ( GWAS ) have identified hundreds of regions of the human genome that are associated with susceptibility to common diseases ., Yet many lines of evidence indicate that many susceptibility loci , which cannot be detected by standard statistical methods , remain to be discovered ., We have developed PUMA , a framework for applying a family of penalized regression methods that simultaneously consider multiple susceptibility loci in the same statistical model ., We demonstrate through simulations that our framework has increased power to detect weak associations compared to both standard GWAS analysis methods and previous applications of penalized methods ., We applied PUMA to identify novel susceptibility loci for type 1 diabetes , Crohns disease and rheumatoid arthritis , where the novel disease loci we identified have been previously associated with similar diseases or are known to function in relevant biological pathways . | genome-wide association studies, mathematics, statistics, genetics, biology, statistical methods | null |
journal.pcbi.1004010 | 2,014 | Exploring O2 Diffusion in A-Type Cytochrome c Oxidases: Molecular Dynamics Simulations Uncover Two Alternative Channels towards the Binuclear Site | Cytochrome c oxidases ( Ccoxs ) are the terminal enzymes of the respiratory chain in eukaryotes and in aerobic prokaryotes ( reviewed in 1 ) ., These integral membrane proteins belong to the heme-copper oxidases superfamily and couple dioxygen ( O2 ) reduction to the translocation of protons across the membrane ., Ccox takes up four electrons from cytochrome c ( cyt c ) in the positively charged side of the membrane ( the inter-membrane space in mitochondria or the periplasm in bacteria ) and eight protons from the negatively charged side ( eq ., 1 ) 2 , 3: ( 1 ) where the subscripts P and N refer to the positive and negative sides of the membrane , respectively ., Four of the eight protons reported in equation 1 are used to reduce one O2 molecule and form two water molecules 2 , 3 , whereas the remaining protons are pumped from the negative to the positive side of the membrane ., This overall process contributes to the generation and maintenance of a transmembrane electrochemical proton gradient , which can be further utilized for several energy-requiring processes , such as ATP synthesis 4 ., Based on structural and phylogenetic analysis , the heme-copper oxidases superfamily is currently divided into three major subfamilies 5: A , B and C . The main differences between the three families are the pathways and mechanisms of proton transfer/pumping ., The A-type Ccoxs , which are the subject of this work , are widespread through all kingdoms of life 5 and among them are the most thoroughly explored Ccoxs 3 , 6 , such as the bovine heart mitochondria , the Paracoccus ( P . ) denitrificans and the Rhodobacter ( R . ) sphaeroides enzymes ., These Ccoxs contain , in the catalytic subunit ( subunit I ) , a low spin heme a and a heterodinuclear center named binuclear center , BNC ( Fig . 1A ) ., The BNC is deeply buried in the core of the protein and it is formed by a high-spin heme a3 and a copper ion ( CuB ) ., In subunit II , these Ccoxs contain only one redox center , a binuclear copper center named CuA , which accepts electrons from the soluble cyt c and then transfers them to the BNC via heme a ., It is believed that protons ( both chemical and pumped ) are transported from the N-side of the membrane to the BNC via two special proton conducting pathways 3: the D- and K-pathways ( Fig . 1A ) ., A third putative proton-conducting pathway , the H-pathway , was proposed for the mammalian Ccox only 6 , 7 , and it was suggested to be exclusively used for the transfer of the pumped protons 8 ., Several high-resolution crystallographic structures of the A-type family are nowadays available in the literature ( e . g . mammalian 7 , 9–11 and bacterial Ccoxs 12–15 ) and , based on these structures , it is known that all A-type members share a remarkable structural similarity of the core functional unit formed by subunits I and II ( Fig . 1A ) ., Subunit I consists of twelve transmembrane α-helices and contains the BNC and the heme a center ., Subunit II is formed by a solvent exposed globular β-sheet domain ( which functions as a docking surface for cyt c ) and two transmembrane α-helices ., It contains only one redox center , the binuclear copper center ( CuA ) ., Moreover , at the interface between subunits I and II , Ccox has one Mg+2 ion whose function is still not well understood , but it was suggested to be part of the exit pathway for the pumped protons and for water formed in the BNC 16 , 17 ., Subunit III , although not considered to be part of the core functional unit , is also highly conserved among the A-type subfamily ., Nevertheless , its absence significantly increases the probability of suicide inactivation 18 , 19 and thereby reduces the catalytic lifespan of Ccox ( in 600-fold or more ) 19 ., Based in the X-ray data available ( eg . 7 , 12 , 13 ) , a putative O2 channel for the A-type family was proposed ( Fig . 1B ) ., Iwata and co-workers , after pressurizing R . sphaeroides Ccox crystals with xenon , were able to identify a continuous hydrophobic channel that starts in the membrane region of subunit I 13 ., This putative O2 channel has two possible entrances that merge together in a region close to the proton-gating residue , E286I ( the residues are numbered according to the R . sphaeroides Ccox sequence and the subscript indicates the subunit number ) ., This pathway presents a constriction point which does not allow the access of O2 to the BNC , at least without the occurrence of some conformational change in the protein ., Unfortunately , until now , none of the mutagenesis and biochemical studies performed in this channel 20–22 was able to clearly demonstrate that it serves as an O2 route into the BNC ., All the tested mutations were located too close to the BNC 20 , 21 , which made the interpretation of the results difficult and did not allow to unambiguously distinguish between the structural obstruction of the O2 channel and the perturbation of the BNC binding kinetics ., However , and contrary to the A-type family , in the B-type family the channel used by O2 to reach the BNC is nowadays considered to be well established ., The crystallographic studies ( with xenon pressurization ) performed in the Thermus ( T . ) thermophilus ba3 enzyme 23–25 , lead to the identification of a “Y-shaped” hydrophobic channel that runs from the membrane region towards the BNC ., This channel , although located roughly at the same position of the putative O2 channel in the A-type Ccox , does not possess a constriction point close to the BNC ., In the A-type Ccoxs , the narrowing of the O2 channel is mainly caused by two conserved bulky residues ( W172I and F282I in R . sphaeroides 13 ) , whereas in the B-type Ccox , smaller residues occupy these positions ( Y133I and T231I in T . thermophilus 23 , 24 ) ., The differences between the A- and B-type regarding the O2 channel are thought to reflect the different functional environments of each type of Ccox ., Although the static crystal structures have been a valuable tool for providing insights into the O2 diffusion and for identifying potential O2 channels in Ccox , the elucidation of the molecular basis of O2 diffusion requires the knowledge of the Ccox conformational dynamics ., Transiently formed cavities and openings inside the protein ( frequently regulated by side chain rotation or by water movements ) are not visible in the static X-ray structures , but have already been shown to be very relevant for ligand diffusion ( see for example 26 ) ., In this context , molecular dynamics ( MD ) simulation techniques ( with sufficient simulation time and conformational sampling ) appear as an alternative for studying the dynamic behavior of proteins and to determine their ligand occupation probabilities inside the protein ., In the last decade , computational methods have been widely used to study gas migration in a number of proteins and MD simulations have successfully allowed the identification of several alternative routes for ligand diffusion ( e . g . hydrogenase 26–28 , myoglobins 29 , 30 , oxidases 31 , 32 and laccases 33 ) ., Moreover , the combination of MD simulations with Implicit Ligand Sampling ( ILS ) 29 calculations allows the calculation of the energy cost of transferring any small , apolar molecule ( like O2 or H2 ) from the solvent to the protein and consequently to compute a 3D free-energy landscape for that specific ligand molecule ( e . g 29 , 33 , 34 ) ., Over the last decades , most of the Ccox research using computational methods focused on the mechanisms and energetics of reduction and/or proton pumping ( e . g 17 , 35–56 ) ., In the A-type Ccox , little emphasis has been given to the identification of the routes used by O2 to move from the solvent towards the BNC , a question only addressed , to our knowledge , by Hofacker and Schulten 31 and by Farantos and co-workers 31 ., In the first work , Hofacker and Schulten 31 used MD simulations to study O2 diffusion in the vicinity of the BNC in a bacterial Ccox from P . denitrificans and in the bovine CcOx enzyme ., Their simulations revealed a unique , well-defined O2 diffusion channel starting in the membrane-spanning surface of subunit I , close to the interface with subunit III ., More recently , Farantos and co-workers 32 have applied the ILS method in order to study the binding of several small gas molecules around the BNC region in the A-type Ccox from P . denitrificans and in the B-type Ccox enzyme from T thermophilus ., From these calculations , the authors were able to identify several cavities around the heme a3 region that are conserved in both the A-type and B-type enzymes ., This study is however limited to the BNC region , not including other parts of the protein and , consequently , not allowing the analysis of the whole O2 permeation process ., The main objective of this work is to identify the O2 channels in the fully reduced Ccox from R . sphaeroides 15 using a combination of MD simulations ( with and without explicit O2 ) and ILS calculations ., Our results revealed the existence of three putative O2 diffusion channels ., One of channels correlates very well with the channel inferred from the X-ray data available , whereas the other two are alternative routes for O2 to reach the BNC , and were not observed in the X-ray structures pressurized with xenon ., Both alternative channels start in the membrane phase and terminate close to Y288I ., Although A-type Ccoxs have been widely studied during the last four decades , the details of the O2 diffusion mechanism are still very incomplete ., In particular , the existence and the characteristics of the channel ( s ) used by O2 to travel from the solvent/membrane to the BNC are still unclear ., In this study , we have used an integrated strategy of all-atom MD simulations ( with and without explicit O2 molecules ) and ILS calculations , designed to examine and characterize the O2 delivery channels in fully reduced Ccox from R . sphaeroides ., Altogether , our results suggest that O2 does not diffuse unspecifically inside this protein and instead , uses three well-defined channels running from the interior of the membrane ( where O2 solubility is higher than in the aqueous phase ) towards the Ccox core ., The first pathway has two entrance points , located between helices 5 and 8 and helices 11 and 13 of subunit I , which converges into the constriction point just before the BNC ., This channel correlates very well with the channel inferred from the available X-ray structures ., The second pathway has only one entry located between the transmembrane helices 13 and 16 of subunit I and it terminates close to Y288I ., The third identified pathway approaches the BNC from the subunit II side ., This channel runs parallel to the heme a3 hydroxylethylfarnesyl tail and also terminates just below Y288I ., According to our observations , the hydrophobic channel detected in the X-ray structures does not constitute the most likely ( energetically preferred ) entrance point for the O2 molecules in this Ccox ., From the O2 affinity map , O2 accesses the BNC via the alternative dynamic channels formed by transient hydrophobic cavities , whose opening and closure is regulated by the thermal fluctuations of the protein ., This may be the reason why these channels were not visible in the static X-ray structures ., In summary , our results suggest that the original hypothesis ( based on static X-ray structures and mutational studies on A-type Ccox ) that proposed , that O2 permeation occurs via a unique , continuous and permanently open channel , is indeed a simplification ., Our current work does not rule out the role of the X-ray inferred channel , but suggests other alternative routes to the BNC ., Furthermore , it emphasizes the need to take into account the dynamic behavior of the protein in order to obtain a more complete description of the O2 putative channels and a more detailed picture of the mechanisms underlying O2 diffusion in these Ccoxs ., The 2 . 15 Å resolution crystal structure of the fully reduced Ccox from R . sphaeroides ( pdb code: 3FYE ) 15 was used as the starting point for this work ., This X-ray structure only contains the minimum functional unit ( subunits I and II ) for Ccox ., Only the water molecules with a relative accessibility to the solvent lower than 50% were kept ., The relative accessibility of water was computed using the program ASC 66 , 67 , resulting in the selection of 240 water molecules ., Since the GROMOS 54A7 force-field 68 lacks the proper parameterization for the Ccox redox centers , the atomic partial charges for reduced CuA , heme a and BNC centers were calculated using quantum mechanical calculations with the software Gaussian09 69 and RESP fitting 70 , as described in detail in S1 Text in section 1 ., The van der Waals parameters for the iron atom ( located in the two heme groups ) were taken from the universal force field 71 whereas the remaining bonded and van der Waals parameters for the metal centers were adapted from the GROMOS 54A7 force field 68 ., The protonation state of each individual protonable group at pH 7 . 0 was determined using a combination of Poisson-Boltzmann calculations , performed with the package MEAD ( version 2 . 2 . 5 ) 72–74 , and Metropolis Monte Carlo simulations , using the program PETIT ( version 1 . 3 ) 75 ., These calculations were performed using the methodologies described in 75 , 76 ., For details related with the determination of the protonation state of the protonatable residues , see section 2 in S1 Text ., Subunits I and II of Ccox were inserted in a pre-equilibrated dimysristoylphosphatidylcholine ( DMPC ) lipid membrane ( for details related with the membrane construction , equilibration and characterization see 77 ) ., The optimal position of the protein relative to the membrane was determined using the location of the charged residues in the transmembrane helices as a reference ., After Ccox insertion into the membrane , all the DMPC molecules located within a cut-off distance of 1 . 2 Å from the protein atoms were removed , as described in detail elsewhere 77 , 78 ., Subsequently , the system ( protein , membrane and crystallographic waters ) was hydrated in a orthorhombic box using a pre-equilibrated box of SPC water molecules 79 ., The water molecules misplaced in the center of the membrane ( formed by the highly hydrophobic lipid tails ) , were removed upon visual inspection ., The final system contained the reduced Ccox embedded in a 175 DMPC lipid membrane surrounded by 19 , 645 water molecules , in a total of 75 , 178 atoms ., All MD simulations were performed using the software package GROMACS 4 . 0 . 4 80 together with the united atom GROMOS 54A7 force-field 68 for the protein and lipids and the previously described atomic partial charges and parameters for the redox centers ., The simple point charge ( SPC ) water model was used 79 ., Periodic boundary conditions were applied to all simulations ., Non-bonded interactions were calculated using a twin range method 81 with short and long-range cut-offs of 8 and 14 Å , respectively ., A reaction field correction 82 , 83 was applied for the truncated electrostatic interactions , considering a dielectric constant of 62 84 ., The SETTLE algorithm 85 was used to constraint the bond lengths and angle in water molecules , while the LINCS algorithm 86 was used to keep all remaining bonds constrained ., The time step for integrating the equations of motion was 0 . 002 ps and the neighbor list was updated every 5 steps ., The simulations were performed at the constant temperature of 310 K , which is above the phase transition temperature for the DMPC lipids ( Tm\u200a=\u200a296–297 K ) in order to ensure that the membrane is in the liquid crystalline state 87 ., A Berendsen heat bath 88 was used , with separate couplings for the protein , membrane and solvent , using a relaxation time constant of 0 . 1 ps ., The pressure was coupled semi-isotropically ( coupling constant of 5 . 0 ps and isothermal compressibility of 4 . 6×10−5 bar−1 84 ) , resulting in an independent coupling of the lateral ( Px+y ) and perpendicular ( Pz ) pressures ., For all simulations , the x+y and z pressure components were kept at 1 atm and no surface tension was applied 84 ., These simulation conditions were shown by Poger et al . 84 , 89 to correctly reproduce several experimental measurements for this type of membranes ., The system was energy minimized with the steepest-descent method in order to remove excessive strain by performing 5000 steps of steepest-descent minimization with harmonic restraints applied to all non-hydrogen atoms ( protein and lipids ) , followed by further 5000 steps restraining the non-hydrogen atoms of the protein , ending with 5000 steps with restraints applied to the Cα atoms only ., After the minimization procedure , and in order to allow proper repacking of the lipids around the protein , a 20 ns MD relaxation was executed in three steps ., First , a 0 . 5 ns simulation was performed with position restraints to all non-hydrogen atoms of the protein and solvent , at constant temperature and pressure ., Afterwards , an additional 0 . 5 ns simulation was performed , with position restraints applied to the non-hydrogen atoms of the protein only ., Finally , only the Cα atoms were restrained for a period of 19 ns ., A force constant of 1000 kJ mol−1nm−2 was used for all the steps that included harmonic position restraints ., The unrestrained simulations started after these 20 ns of restrained simulation ., In order to reduce the well known sampling problems in membrane-protein simulations , five MD simulations , 100 ns each , were performed , resulting in 0 . 5 µs of total simulation time ., All replicates were initiated with different sets of random velocities ., These simulations will be hereafter designated as O2-free simulations ., After 20 ns of restrained simulations , we randomly added 84 molecules of dioxygen ( O2 ) in the solvent zone of each system ., No O2 was placed inside the protein nor inside the hydrophobic core of the membrane ( see S4 Figure in S1 Text ) ., This new set of simulations will be , hereafter , designated as O2 simulations ., The water molecules within a 2 Å distance from the O2 molecules were deleted , similarly to the procedure described in 33 ., In order to allow the solvent to adapt to the newly added O2 molecules , a 0 . 5 ns MD simulation with position restraints on all non-hydrogen atoms ( force constant of 1000 kJ mol−1nm−2 ) was performed ., After this initialization procedure , unrestrained MD simulations were carried out and the simulation conditions and parameters were similar to the ones described previously for the MD simulations without O2 , except for the temperature coupling groups used ., In this set of simulations , the O2 molecules were included in the same group as the protein ., 5 MD simulations , 100 ns each , were performed ., The parameters for the O2 molecules were taken from the previously published work of Victor et al 90 ., The 84 O2 molecules added to the system corresponds to an O2 concentration of ∼0 . 235 M , which is higher than the experimental solubility of this gas in water ., However , this high O2 concentration does not affect the structural properties of the protein as shown in S5 Figure in S1 Text ., Moreover , the use of this high number of O2 molecules is necessary to obtain reliable statistics within a reasonable simulation time ., Sites with high O2 affinity were determined using the ILS method , as previously described in 29 ., In this method , the potential of mean force for placing an O2 molecule in any position inside the protein is calculated according to: ( 2 ) In equation 2 , the implicit ligand potential of mean force , , is an average over a finite number of protein and solvent configurations ( ) and over a number of different equally probable orientations of the ligand ( ) ., Moreover , is the Boltzmann constant , is the absolute temperature , and is the interaction energy between the protein and solvent configuration ( ) with the ligand located at position with the orientation ., In our case , the O2 free energy map was constructed using the last 50 ns ( for each replicate ) of the O2-free simulations ., For the calculations , all 50005 conformations ( =\u200a10001 conformations x 5 replicates in equation 2 ) were fitted to the X-ray structure using the Cα atoms ., A grid of 51×55×87 dimensions was used with a grid spacing of 1 Å and 400 O2 insertions were performed per grid point ( =\u200a400 in equation 2 ) ., All calculations were carried out using a version of GROMACS 4 . 0 . 4 Widom TPI algorithm , modified almost in the same way as described in 33 ., The only difference is that the ligand insertions here were made within the whole space of the grid cube ( the grid cube is centered at the insertion point and with edge length equal to the grid spacing ) , while in the previous work ( described in 33 ) the insertions were only possible within the inscribed sphere on the grid cube ., The 3D free energy map obtained describes the Gibbs free energy of moving an O2 molecule from vacuum to a given position in the system , ΔGvac→prot ( O2 ) ., This map was then converted into the ΔGwat→prot ( O2 ) map of interest using a ΔGwat→prot ( O2 ) calculated as described in 33 ., The secondary structure assignment was performed with the program DSSP 91 ., To determine the percentage of secondary structure loss relative to the X-ray structure , the secondary structure classes considered were: α-helix , 310-helix , 5-helix , β-sheet and β-bridge ., For the energy landscape analysis , we used the method described in 33 ., In short , this method classifies the energy landscape into energy basins through a steepest-descent tessellation and , afterwards , identifies the lowest-energy point within the boundaries between each pair of neighboring basins , i . e . the saddle point between those basins ., After this procedure , a network of paths between all energy minima of the landscape can be constructed using the steepest-descent paths from the saddle points to the minima ., A cutoff of 20 kJ·mol−1 was used for the network construction ., The errors of the free energy profiles were calculated using two blocks: the first block corresponds to the frames ranging from 50 ns to 75 ns ( for all five replicates ) whereas the second block contains all the frames ranging from 75 ns to 100 ns ., The errors were determined as half the difference of the energies observed between the two blocks for each minina and for each transition ., The method used for error calculation assumes that similar minima and pathways can be identified in the two blocks ., However , for the O2 channel 1 , the pathways connecting M6 to M8 and M10 to M11 were not visible in one of the blocks and by this reason their associated errors could not be calculated . | Introduction, Results/Discussion, Materials and Methods | Cytochrome c oxidases ( Ccoxs ) are the terminal enzymes of the respiratory chain in mitochondria and most bacteria ., These enzymes couple dioxygen ( O2 ) reduction to the generation of a transmembrane electrochemical proton gradient ., Despite decades of research and the availability of a large amount of structural and biochemical data available for the A-type Ccox family , little is known about the channel ( s ) used by O2 to travel from the solvent/membrane to the heme a3-CuB binuclear center ( BNC ) ., Moreover , the identification of all possible O2 channels as well as the atomic details of O2 diffusion is essential for the understanding of the working mechanisms of the A-type Ccox ., In this work , we determined the O2 distribution within Ccox from Rhodobacter sphaeroides , in the fully reduced state , in order to identify and characterize all the putative O2 channels leading towards the BNC ., For that , we use an integrated strategy combining atomistic molecular dynamics ( MD ) simulations ( with and without explicit O2 molecules ) and implicit ligand sampling ( ILS ) calculations ., Based on the 3D free energy map for O2 inside Ccox , three channels were identified , all starting in the membrane hydrophobic region and connecting the surface of the protein to the BNC ., One of these channels corresponds to the pathway inferred from the X-ray data available , whereas the other two are alternative routes for O2 to reach the BNC ., Both alternative O2 channels start in the membrane spanning region and terminate close to Y288I ., These channels are a combination of multiple transiently interconnected hydrophobic cavities , whose opening and closure is regulated by the thermal fluctuations of the lining residues ., Furthermore , our results show that , in this Ccox , the most likely ( energetically preferred ) routes for O2 to reach the BNC are the alternative channels , rather than the X-ray inferred pathway . | Cytochrome c oxidases ( Ccoxs ) , the terminal enzymes of the respiratory electron transport chain in eukaryotes and many prokaryotes , are key enzymes in aerobic respiration ., These proteins couple the reduction of molecular dioxygen to water with the creation of a transmembrane electrochemical proton gradient ., Over the last decades , most of the Ccoxs research focused on the mechanisms and energetics of reduction and/or proton pumping , and little emphasis has been given to the pathways used by dioxygen to reach the binuclear center , where dioxygen reduction takes place ., In particular , the existence and the characteristics of the channel ( s ) used by O2 to travel from the solvent/membrane to the binuclear site are still unclear ., In this work , we combine all-atom molecular dynamics simulations and implicit ligand sampling calculations in order to identify and characterize the O2 delivery channels in the Ccox from Rhodobacter sphaeroides ., Altogether , our results suggest that , in this Ccox , O2 can diffuse via three well-defined channels that start in membrane region ( where O2 solubility is higher than in the water ) ., One of these channels corresponds to the pathway inferred from the X-ray data available , whereas the other two are alternative routes for O2 to reach the binuclear center . | biomacromolecule-ligand interactions, biochemistry, proteins, electron transport chain, biology and life sciences, bioenergetics, biophysics, biophysical simulations | null |
journal.pgen.1001344 | 2,011 | Viral Genome Segmentation Can Result from a Trade-Off between Genetic Content and Particle Stability | A biological clone of foot-and-mouth disease virus ( FMDV ) , termed C-S8c1 , evolved in BHK-21 cell culture infections at high multiplicity of infection ( MOI ) , towards a population dominated by defective genomic forms that were infectious by complementation in the absence of standard size ( ST ) genomes 1 ( Figure 1 and Figure S1 ) ., By passage 260 , the population ( C-S8p260 ) was composed mainly of two classes of genomes that included internal in-frame deletions , Δ417 plus Δ999 and the minority genome Δ1017 ( with deletions of 417 , 999 and 1017 nucleotides , respectively , at the capsid-coding region ) ., ST genomes were not detected in C-S8p260 , and it was estimated that their frequency in C-S8p260 was lower than 10−4-fold the frequency of genomes with deletions ., The segmented genome version was stable at least up to passage 460 at high MOI ., However , when population C-S8p260 was subjected to low-MOI infections , that impeded coinfection of cells by the complementing genome classes , a ST genome termed C-S8p260p3d was selected as a result of recombination between Δ417 and Δ999 RNAs ( Figure S1 ) 2 ., The dominance of a population of complementing defective genomes that did not require ST genomes for replication was regarded as the first step of an evolutionary transition towards viral genome segmentation , an event likely to have occurred at some point of the evolutionary history of RNA viruses 1 , 3–6 ., A critical question in the displacement of a ST genome by defective , complementing genomes , is the molecular basis that underlies the superiority of the segmented forms over the ST genome ., As C-S8p260 and C-S8p260p3d share a common genetic background ( similar set of point mutations relative to the parental C-S8c1 ) , this dual viral system constitutes a suitable model to address this fundamental question 2 ., Here we provide evidence that the segmented C-S8p260 is endowed with a non replicative advantage over its unsegmented counterpart C-S8p260p3d that does not reside in the rate of either RNA genome replication or of virus-specific protein synthesis ., Unexpectedly , an increased virion stability conferred a higher specific infectivity and longer lifespan on the segmented virus ., The relative fitness advantage that led to dominance of the segmented population C-S8p260 over its ST ancestor was determined using virus-competition assays between C-S8p260 and C-S8p260p3d , or each in competition with C922L150 , another C-S8c1-derived clone of lower fitness ( see Materials and Methods ) ., Additionally , C-S8p260 was competed against the ST population derived from passage 460 ( C-S8p460p5d ) 2 ., The results ( Table 1 and Figure 2 ) indicate that population C-S8p260 displayed a two-fold higher fitness ( or relative selection coefficient , see Materials and Methods ) than C-S8p260p3d and C-S8p460p5d ., Both , C-S8p260 and C-S8p260p3d won their respective competitions against C922L150 , but C-S8p260 displayed a 1 . 7-fold higher fitness than C-S8p260p3d ., The outcome of these competitions strongly suggests that the segmented genetic system confers approximately a two-fold additional fitness advantage ( relative to the corresponding unsegmented genome version ) , in agreement with its reaching dominance in the C-S8c1 lineage ., To identify the step of the virus life cycle that was associated with the fitness advantage of C-S8p260 over C-S8p260p3d , the intracellular and extracellular concentrations of viral RNA , in the course of infections with both viruses , were determined ., BHK-21 cells were infected at a MOI of 20 PFU/cell , in order to maintain the coinfection of cells by the two components of C-S8p260 , and to restrict the measurements to a single round of cell infections ., In independent infections carried out in parallel ( Figure 3A ) , C-S8p260 and C-S8p260p3d did not differ significantly in their exponential increase of intracellular viral RNA ., Their respective growth rate constants ( see Materials and Methods ) were rC-S8p260\u200a=\u200a0 . 065±0 . 007 RNA molecules/cell·min and rC-S8p260p3d\u200a=\u200a0 . 054±0 . 006 molecules/cell·min ( ANOVA , F1 , 12\u200a=\u200a1 . 29 , p\u200a=\u200a0 . 278 ) ., To test whether the selective advantage of C-S8p260 was manifested only upon coinfection with the unsegmented form , the specific RNA concentrations of C-S8p260 and C-S8p260p3d were measured in cells coinfected at high-MOI at the stages of cell entry ( Figure 3B ) , intracellular replication ( Figure 3C ) , and virus release to the extracellular medium ( Figure 3D ) ., Both types of RNA were rapidly uptaken by the BHK-21 cells , following application of the viruses to the cells ( Figure 3B ) , and then the intracellular viral RNA levels increased rapidly and reached a maximum at 15 minutes post-infection ( pi ) ., The viral RNA levels remained approximately constant up to minute 60 pi , and then they increased exponentially ., The uptake process was parallel for the two viruses ., A similar result was observed upon measurement of the intracellular level of the two types of RNA during the exponential growth phase ( Figure 3C ) ., The slope of the exponential increase of intracellular RNA was parallel: the genomic intracellular RNA of C-S8p260 and C-S8p260p3d increased at a rate rC-S8p260\u200a=\u200a0 . 050±0 . 004 RNA molecules/cell·min and rC-S8p260p3d\u200a=\u200a0 . 064±0 . 008 molecules/cell·min , respectively ( F1 . 26\u200a=\u200a2 . 38 , p\u200a=\u200a0 . 13 , Figure 3C ) ., The same culture samples were used to measure the release of viral RNA into the extracellular culture medium ( Figure 3D ) ., The results show that beginning at minute 135 pi , the concentration of extracellular viral RNA increased very rapidly at a similar rate of rC-S8p260\u200a=\u200a0 . 076±0 . 008 RNA molecules/cell·min and rC-S8p260p3d\u200a=\u200a0 . 096±0 . 010 molecules/cell·min ( F1 . 26\u200a=\u200a2 . 59 , p\u200a=\u200a0 . 12 ) ., RNA samples were treated with RNase A ( under the assumption that encapsidated RNA is RNase-resistant and non-encapsidated RNA is RNase-sensitive 7 ) , prior to the specific quantification of the two types of RNA ., The treatment did not alter the measurements significantly ( see Materials and Methods ) ., Thus , the segmented and unsegmented forms of FMDV followed parallel kinetics of RNA synthesis , not only at the early steps of infection , but also during genome replication and release of RNA from the cell ., The synthesis of viral proteins was analyzed at different times post-electroporation of BHK-21 cells with either RNA transcribed from plasmid pMT260p3d ( that gives rise to C-S8p260p3d ) or with an equimolar mixture of RNA obtained from plasmids pMT260Δ417ns and pMT260Δ999ns ( that give rise to C-S8p260 ) , constructed as described in Text S1 ( Figure 3E , 3F and Figure S2 ) ., Electroporated and mock-electroporated cells were metabolically labeled with 35SMet-Cys , protein expression was monitored every 30 minutes , between 1 . 5 and 4 . 5 hours post-electroporation , and the proteins were resolved by SDS-PAGE and fluorography ( Figure 3E ) ., The analysis revealed a parallel expression kinetics of viral proteins , with few minor differences ., In both cases , the maximum level of viral proteins was detected between 2 and 2 . 5 hours post-electroporation ( Figure 3E , 3F ) ., The expression of structural proteins VP1 and VP3 was lower in cells electroporated with RNA transcripts pMT260Δ417ns and pMT260Δ999ns than in cells electroporated with pMT260p3d ., This may be a consequence of the deletion in pMT260Δ999ns that affects both the VP1 and VP3-coding regions ., The kinetics of expression of non-structural proteins followed parallel curves during the time of the measurements , as determined by the label present in 3D and 3CD ( Figure 3F ) ., The results exclude that the selective advantage of population C-S8p260 can be due to faster kinetics in viral protein expression ., To determine whether C-S8p260 and C-S8p260p3d displayed different specific infectivity ( see Materials and Methods ) , viral genomic RNA molecules and infectivity ( PFU/ml ) were measured in both populations ., Viral RNA production was 2 . 8±2 . 1 higher in C-S8p260p3d population relative to C-S8p260 population ( repeated measures ANOVA: F1 , 8\u200a=\u200a17 . 53 , p<0 . 01 , Table 2 ) ., In agreement with this result , the production of viral particles was two-fold higher for C-S8p260p3d than for C-S8p260 , as previously measured by quantitative electron microscopy 1 ., Since both viruses , however , showed no differences in viral titer production ( ANOVA: F1 , 12\u200a=\u200a0 . 0018 , p\u200a=\u200a0 . 97 , Table 2 ) , the specific infectivity of C-S8p260 is 2 . 6-fold higher than that of C-S8p260p3d ., Of note , this difference coincides with the fitness differences between C-S8p260 and C-S8p260p3d ( compare Table 1 ) ., The ratio between the specific infectivities of C-S8p260 and C-S8p260p3d was additionally determined by using an alternative approach based on estimating the proportion of genomes that enter the cell for a given initial inoculum 8 ., The ratio of C-S8p260 to C-S8p260p3d genomic RNA was measured in a mixture prepared with equal PFUs of C-S8p260 and C-S8p260p3d ., Then , BHK-21 cells were infected with the mixture , and the ratio of segmented to ST RNA was determined 60 minutes after infection ( before the onset of exponential replication ) ., When complete cytopathic effect was reached , the ratio of the two types of RNA was measured again ., The resulting cell lysate was used to infect new BHK-21 cells , and the process repeated to attain a total of three sequential cell entry events ( Figure 4 ) ., The ratio of C-S8p260 to C-S8p260p3d RNA varied in a step-wise fashion , with 2-fold increases occurring only between each infection and the corresponding virus entry inside the cell ., The magnitude of the step-wise increases confirmed the difference in specific infectivity between the two viruses ( compare Figure 4 and Table 2 ) ., The ratio of the amount of the two types of RNA remained constant from each cell entry event up to the corresponding cell lysis , in agreement with the results of viral RNA kinetics ( compare Figure 4 and Figure 3 ) ., The results strongly suggest that the viral population with the segmented genome is more infectious than the population with the ST genome ., Upon elimination of the “Entry” points from the data in Figure 4 , a graph coincident with that of a standard fitness determination is obtained ( inset in Figure 4 ) , which again indicates a two-fold higher fitness of C-S8p260 relative to C-S8p260p3d ., To investigate whether the increase of specific infectivity in the segmented-genome FMDV population could be attributed to an increase in the stability of the viral particles , the loss of infectivity of C-S8p260 and C-S8p260p3d at 37°C was quantitated ., The results ( Figure 5 ) show that the inactivation rate constant ( see Materials and Methods ) of C-S8p260 was k\u200a=\u200a0 . 0156±0 . 0005 min−1 ( corresponding to a half-life of 44 min ) , while the inactivation rate constant of C-S8p260p3d was k\u200a=\u200a0 . 0190±0 . 0007 min−1 ( corresponding to a half-life of 33 min ) , a statistically significant difference ( ANOVA , F1 . 25\u200a=\u200a14 . 47 , p<0 . 001 ) ., It must be noticed that both rates represent an extremely fast kinetics of infectivity decay , a well known feature of FMDV 9 ., To establish a link between the stability of viral particles and the observed relationship of the fitness and infectivity differences between C-S8p260 and C-S8p260p3d , the infectivity of both populations was monitored again using the step-wise RNA level technique ( described in Figure 4 ) ., Between the first and the second round of infection ( that is , after the lysis event ) , the viral population was incubated for one additional hour at 37°C ., The results indicate that the incubation at 37°C accentuated the increased infectivity of C-S8p260 relative to C-S8p260p3d ( Figure 5C ) ., Thus , measurements of specific infectivity , virion stability , and the analysis of different steps involved in the virus life cycle suggest that increased stability of the particles harboring RNA with internal deletions is the phenotypic trait that conferred a selective advantage of the segmented virus over its ST counterpart ., The empirical observations described can be synthesized in a simple computational model summarized in Figure 6 ., The standard type is termed population S , and the two complementing defective viruses are generically termed populations A ( C-S8p260Δ417 ) and B ( C-S8p260Δ999 ) ., Experiments on the kinetics of RNA and protein synthesis indicate that , in the replicative period inside the cell , particles of either type ( S , A , or B ) replicate at the same rate , conditional on at least one of their complementing counterparts being present for classes A and B . This condition is implemented as follows ., Suppose that , initially , a cell holds nS , nA and nB particles of types S , A , and B , respectively ., After replication , the viral population is formed by r×MA particles of type A , r×MB particles of type B , and r×nS particles of type S , where MA is the smallest quantity of nS+nB and nA , and MB is the smallest quantity of nS+nA and nB ., On the other hand , experiments on specific infectivity and virion stability show that the segmented population is more infectious ( due to its higher stability ) than the standard counterpart ., This is implemented as a decay factor dS<1 that reduces the total number of S infective particles between replicative periods ., In the particular case of the current experiments , a value of 0 . 47 can be estimated for the decay factor dS ( see Text S1 ) ., Those two steps ( replication at the same rate and differential infection ) quantify the process described in Figure 4 ., Finally , the average number of particles that infect cells is a constant m that stands for the MOI in the experimental system ., In the experiments where the defective complementary form displaces S , m\u200a=\u200a20 has been used ., The two key parameters are m and dS , which represent antagonistic selection pressures ., At low m , S populations are at an advantage because complementation is rare , while high m benefits segmented forms ., However , the latter would be unable to displace the S population if both types were equally infectious ., The increase in viral infectivity is truly beneficial for a value of m high enough that replication is not strongly limited by complementation ., This is the behavior summarized in Figure 6 , where the two different outcomes of the competition as a function of m and dS are shown ., Above a critical line of m values , there is co-dominance of standard and defective populations , while below that line the S population disappears ., The model correctly predicts that the standard form will be displaced by the complementary , defective population in the experimental situation ( see Figure 6B ) , where the pair ( dS , m ) = ( 0 . 47 , 20 ) ., It is important to emphasize that this result is independent of the fractions of S , A , and B present at the outset of the experiment or in the computational initial condition ., Moreover , the inverse of the decay corresponds to the stepwise increase in the frequency of each virus , as displayed in Figure 4: ., This value is in good agreement with empirical findings ( see model prediction in Figure 4 ) ., Genome segmentation is a major evolutionary transition from independent towards complementing transmission of genetic information ., Two main proposals for the evolutionary advantage of genome segmentation have been made on the basis of theoretical studies ., One is that genome segmentation is a form of sex that counteracts the effect of deleterious mutations 5 , 10 ., Another , not mutually exclusive but mechanistic proposal , is that genome segmentation may ensue from the selection of shorter RNA molecules whose replication is completed in a shorter time than replication of the corresponding full length genome 3 , 6 ., Evidence supporting selection for deleted RNA was obtained in experiments involving in vitro replication of Qβ RNA , without the requirement to express viral proteins or to produce infectious particles 11 , 12 ., In the case of FMDV there is no evidence of an advantage of the segmented over ST genome at the stage of RNA genome replication , protein expression or production of infectious virus , in agreement with previous descriptions for positive strand defective viruses 13 , 14 ., The lack of replicative advantage of C-S8p260 is also reflected in the fact that the segmented virus had a constant two-fold additional fitness advantage , over the ST virus , independently of the genetic background of the competitor virus ., Thus , all evidences point towards a non-replicative trait , virion stability , behind the selective advantage of the genome version with internal deletions ., The slower inactivation rate of the segmented virus correlates with the difference of specific infectivity between C-S8p260 and C-S8p260p3d , and such a difference is , in turn , at the origin of the fitness difference ., Due to the exponential nature of the infectivity decay 15 , even a modest increase in virion stability can account for the enrichment of the population in the more stable forms 13 ., The implementation of the model with the actual values of the MOI and decay rates ( see Text S1 ) , estimated for the segmented and ST viruses , predicted a decay value ( ) in the parameter space where the ST virus is driven to extinction ., The inverse of this decay value ( ) represents the relative increase per infection of the ST virus ., This value confirms that the increased stability of the segmented virus can account for the differences in fitness , specific infectivity , and the step-wise dynamics observed ( 2 . 5 , 2 . 6 , and 1 . 9 , respectively ) ., The model also explores the limits imposed by the MOI on the complementing system , and predicts the minimal MOI required for the segmented forms to displace the ST virus ., The molecular basis for the higher thermal stability and fitness of the Infectious C-S8p260 population relative to the ST virus is unclear ., However , some evidence indicates that thermal inactivation of FMDV may be due to a conformational change in the virion 9 ., We suggest that the amount of RNA inside the virion may Influence the kinetic barrier of the inactivation process because of packaging considerations ., The volume occupied per unit mass ( Vm ) 16 of full-length RNA inside the FMDV virion is about Vm\u200a=\u200a1 . 95 Å3/Da ( see calculation of the RNA packaging density in the Text S1 ) ., This corresponds to a very high packing density , slightly higher than that of RNA in a molecular crystal ( about 2 . 1 Å3/Da ) , and substantially higher than the density of other icosahedral RNA viruses 17 ., This measurement implies that genomes inside the capsid may be partially dehydrated ., Packaging 5%–12% shorter RNAs ( in the C-S8p260 virions ) would lead to Vm values of about 2 . 05 Å3/Da–2 . 20 Å3/Da , and thus may involve no dehydration ., Based on these estimates , and ignoring other energetic effects on RNA packaging which are more difficult to predict , one could surmise that C-S8p260 virions would be at an energetically lower state than the ST virions harboring longer RNAs ., The extra energy needed to trigger the putative conformational rearrangement that may lead to FMDV inactivation could be higher for the C-S8p260 virions than for the ST virions , rendering C-S8p260 virions more resistant to thermal inactivation , as experimentally observed ., Accordingly , an increase in the length of an internal oligoadenylate tract in the viral RNA was shown to have a negative effect on FMDV fitness 18 , and thermal stability 19 ., Thus , variations of the RNA length could destabilize or stabilize the infectious virion conformation by reducing or increasing thermostability and fitness because of excessive RNA packing density , or by relaxing the RNA packaging constraints , respectively ., Packaging constraints of genome length in icosahedral viruses have been previously described , including adenovirus vectors 20 or the strongly pressurized capsids of some double-stranded DNA viruses 21 ., A study including multiple DNA and RNA bacteriophages concluded that genome packaging density was negatively correlated with virus stability 22 ., Gene overlapping in viruses is thought to have evolved as a consequence of physical constraints of genome length in the capsid 23 ., Accordingly , our results contribute a new model for the fitness advantage of RNA genome segmentation , a key evolutionary transition in RNA genetics ., We propose that segmentation is a molecular solution that counteracts the trade-off between capsid stability and genome length in geometrically constrained viral particles ., The relaxation of the genome packaging of segmented genomes maximizes the genetic content in the virion without the associated loss of particle stability ., The origin of baby hamster kidney 21 ( BHK-21 ) cells and procedures for cell growth , infection of cell monolayers with FMDV in liquid medium , and for plaque assays in semisolid agar medium have been previously described 18 , 24 ., FMDV C-S8c1 is a plaque-purified virus of natural isolate C1 Santa-Pau Spain 70 24 ., FMDV C-S8p260 and C-S8p460 are the viral populations obtained after 260 and 460 serial cytolytic passages , respectively , of C-S8c1 at high MOI in BHK-21 cells ( 2×106 BHK-21 cells infected with the virus contained in 200 µl of the supernatant from the previous infection , that include about 2·106 to 4·107 PFU ) ( Figure S1 ) 1 , 2 ., FMDV C-S8p260p3d and C-S8p460p5d are the viral populations obtained after three serial cytolytic passages of C-S8p260 and five serial cytolytic passages of C-S8p460 , respectively , at low MOI in BHK-21 cells ( 2×106 BHK-21 cells infected with 200 µl of a 10−3 dilution of the supernatant from the previous infection; MOI of about 10−3 PFU/cell ) 1 , 2 ., FMDV C922L150 was obtained after 150 population passages of clone C922 ( a subclone of C-S8c1 p2 ) in BHK-21 cells ( MOI of 0 . 1–10 PFU/cell ) 25 ., The production of lytic plaques ( used to determine the viral titer ) in a population of complementing viruses follows a two-hit kinetics as described in 26 and in the Text S1 ., Viral RNA was extracted from the viral samples using Trizol ( Invitrogen ) ., Intracellular RNA was extracted by direct addition of Trizol to the cell monolayer , after removing the cell culture medium ., RNA was quantified by real-time RT-PCR with the Light Cycler instrument ( Roche ) using the Light Cycler RNA Master kit ( Roche ) , as previously described 27 ., Purified RNA from FMDV C-S8p260p3d or pMT260Δ999ns was used as standard ., Reverse transcription was performed with AMV reverse transcriptase ( Promega ) , and PCR amplification was carried out using Ampli-Taq polymerase ( Perkin-Elmer ) , as specified by the manufacturers ., The pairs of sense and antisense oligonucleotides , respectively , that amplify specific viruses are the following: C-S8p260p3d , 5′-CTACCCATGGACGCCAGACCCG-3′ ( sense ) /5′-GTGTTGGTTGTGTGTGCAG-3′ ( antisense ) ; C-S8p260 , 5′-CACGAATTCACGGGCAAAGGCTACTGG-3′ ( sense ) /5′-GAGAAGAAGAAGGGCCCAGGGTTG-3′ ( antisense ) ., When necessary , the supernatants of infected cells were treated for 1 hour with 1 µg/ml of pancreatic RNase A as previously described 7 , to eliminate non-encapsidated RNAs ., Six independent samples of the supernatant of BHK-21 cells coinfected by C-S8p260 and C-S8p260p3d were either untreated or treated with RNase , prior to the specific quantification of the two types of RNA by real-time RT-PCR , as described in 7 ., The ratio of segmented to unsegmented genomic RNA was 1 , 23±1 , 11 in RNase A-treated and 1 , 64±0 , 49 in untreated samples ., Thus , no significant differences ( ANOVA , F1 . 10\u200a=\u200a0 . 68 , p>0 . 43 ) could be detected between the amounts of encapsidated and non-encapsidated genomic RNA released into the culture medium by the segmented and ST viruses ., Specific infectivity is defined as the ratio between the number of infectious viruses , measured in PFUs , and the total amount of viral RNA , determined by quantitative RT-PCR ., BHK-21 monolayers of 5·105 cells were infected in parallel with 1·107 PFU of C-S8p260 , C-S8p260p3d or the mixture of both ( a total of 2·107 PFU ) ., At each specific time point , at least three plates were withdrawn for the determination of intracellular and extracellular FMDV RNA ., The RNA values obtained were normalized to the number of cells ., The increase of the amount of viral RNA over time was fitted to the equation: x ( t ) =\u200ax0·e ( r·t ) , where r is the growth rate constant measured in RNA molecules/passage ., During coinfections , the concentration of C-S8p260 was measured by RT-PCR amplification using primers that specifically amplify the genome harboring deletion Δ999 ., Since this genome has been estimated in a proportion of 40% in the C-S8p260 population 1 , viral RNA of C-S8p260 population has been calculated as 2 . 5 times the concentration of Δ999 RNA ., Viral protein synthesis was analysed by metabolic pulse-labelling with 35S Met-Cys , followed by SDS-PAGE electrophoresis and fluorography ., Proteins were labelled by the addition of 60 µCi of 35S Met-Cys ( Amersham ) per ml in methionine-free DMEM ., Fluorography , autoradiography and western blot procedures of the gels were carried out as previously described 28 ., The amount of actin in the sample was determined using anti-β-actin MAb AC-15 ( Sigma ) , and corresponded to a concentration of protein in the linear region of the relationship between the western blot signal and the protein concentration ., FMDV-specific proteins were identified using MAbs 29 , as previously described 30 ., Growth-competitions between two viruses in BHK-21 cells were carried out as previously described 25 , with minor modifications ., A cell monolayer is infected with a mixture of a problem and reference virus in a proportion of 1∶1000 ( unless otherwise stated ) and at a MOI\u200a=\u200a10–20 ., When the complete cytopathic effect is reached the supernatant ( containing the virus ) is collected and used for a new infection ., The fitness of several clones used in the present study is considerably high ., Moreover , fitness determination of a multipartite virus does not have a standardized protocol ., Using the ratio 1∶1000 allows performing 6 to 8 serial infections ( the typical number being 3 to 4 ) before the exponential variation of the genotypes frequency reaches a saturation point ., The exponential increase of the proportion of the problem genomes is fitted to an exponential curve ., The slope of the curve gives the selection coefficient for one strain relative to the other 31 ., This value is often used in virology as a measure of the relative fitness 32 ., The proportion of the two genomes at different passages was determined by specific real-time RT-PCR ., The equations for each competition and fitness values are given in the legend of Figure 2 and in Table 1 , respectively ., Equal volumes of viral samples of C-S8p260 and C-S8p260p3d were incubated at 37°C and aliquots were collected at different time points and rapidly chilled to 0°C ., Viral titer decay as a function of time was determined and fitted to an exponential curve following the equation: viral titer ( at time t ) =\u200aviral titer ( at time t\u200a=\u200a0 ) · e−k·t , where k ( h−1 ) is the inactivation rate constant of the infectious FMDV virion 33 ., The equations that define the inactivation rate and the average life of the segmented and ST viruses are given in the legend of Figure 5 ., One Way ANOVA were calculated using the Statistica 6 . 0 software package ( StatSoft 2001 ) . | Introduction, Results, Discussion, Materials and Methods | The evolutionary benefit of viral genome segmentation is a classical , yet unsolved question in evolutionary biology and RNA genetics ., Theoretical studies anticipated that replication of shorter RNA segments could provide a replicative advantage over standard size genomes ., However , this question has remained elusive to experimentalists because of the lack of a proper viral model system ., Here we present a study with a stable segmented bipartite RNA virus and its ancestor non-segmented counterpart , in an identical genomic nucleotide sequence context ., Results of RNA replication , protein expression , competition experiments , and inactivation of infectious particles point to a non-replicative trait , the particle stability , as the main driver of fitness gain of segmented genomes ., Accordingly , measurements of the volume occupation of the genome inside viral capsids indicate that packaging shorter genomes involves a relaxation of the packaging density that is energetically favourable ., The empirical observations are used to design a computational model that predicts the existence of a critical multiplicity of infection for domination of segmented over standard types ., Our experiments suggest that viral segmented genomes may have arisen as a molecular solution for the trade-off between genome length and particle stability ., Genome segmentation allows maximizing the genetic content without the detrimental effect in stability derived from incresing genome length . | Genome segmentation , the splitting of a linear genome into two or more segments , is a major evolutionary transition from independent towards complementing transmission of genetic information ., Many viruses with RNA as genetic material have segmented genomes , but the molecular forces behind genome segmentation are unknown ., We have used foot-and-mouth disease virus to address this question , because this non-segmented RNA virus became segmented into two RNAs when it was extensively propagated in cell culture ., This made possible a comparison of the segmented form ( with two shorter RNAs enclosed into separate viral particles ) with its exactly matching non-segmented counterpart ., The results show that the advantage of the segmented form lies in the higher stability of the particles that enclose the shorter RNA , and not in any difference in the rate of RNA synthesis or expression of the genetic material ., Genome segmentation may have arisen as a molecular mechanism to overcome the trade-off between genomic content and particle stability ., It allows optimizing the amount of genetic information while relaxing packaging density . | computational biology/evolutionary modeling, genetics and genomics/microbial evolution and genomics, virology/virus evolution and symbiosis | null |
journal.pntd.0005758 | 2,017 | The tradition algorithm approach underestimates the prevalence of serodiagnosis of syphilis in HIV-infected individuals | Syphilis is an ancient human disease caused by Treponema pallidum , which is mostly transmitted by sex activity ., It remains a worldwide public health concern as there has been a global increase in the incidence of syphilis , especially among men who have sex with men ( MSM ) ., MSMs are a unique population that experience disproportionately high rates of HIV infection ., Although clinical profiling of symptoms is important , serologic tests are still considered the mainstay of syphilis diagnosis ., Serological tests for syphilis can be categorized into two types: the non-treponemal tests ( NTT ) such as rapid plasma reagin ( RPR ) , toluidine red unheated serum test ( TRUST ) , and Venereal Disease Research Laboratory ( VDRL ) tests ., Other treponemal tests ( TT ) include the T . pallidum particle agglutination assay ( TPPA ) , T . pallidum hemagglutination assay ( TPHA ) , treponemal ELISA , and chemiluminescence methodologies1 ., Currently , there are three algorithms for screening of syphilis ., First , the traditional screening algorithm commences with a non-treponemal assay followed by a confirmation with a treponemal test ., Second , the reverse algorithm starts with a treponemal assay , and a reactive treponemal screening assay is followed by a quantitative non-treponemal assay ., Third , the European Centre for Disease Prevention and Control ( ECDC ) algorithm-a modified reverse algorithm: a reactive treponemal screening test is followed by a second ( and different ) treponemal test but is not accompanied by a non-treponemal test2 ., All testing algorithms possess certain advantages and limitations ., Consequently , there is no generally recognized diagnostic algorithm3 ., For those infected with both syphilis and HIV , the situation is even more complex4 ., For example , unusual serologic responses such as the prozone and sreofast phenomenon have been observed in HIV-infected individuals5 ., To the best of our knowledge , no studies have analyzed the different algorithms for detecting syphilis in HIV-positive people ., Therefore , this study aimed to compare the results of three syphilis screening algorithms in an attempt to evaluate their screening performance in this unique population ., Moreover , we examined whether the different screening algorithms significantly influenced the seroprevalence of syphilis in HIV-positive patients ., We conducted a cross-sectional study to assess the impact of different syphilis screening algorithms in a HIV-positive population ., Sample size was estimated to be 677 using N=Zα2P ( 1−P ) ⁄δ2 , assuming 19 . 8% syphilis prevalence in HIV infected patients6 , with 3% precision and 95% level of confidence ., We collected a convenience sample of discarded serum specimens from HIV patients undergoing serologic evaluation for HIV virus load in The First Affiliated Hospital , Medical College of Zhejiang University ., The patients’ HIV infection status was confirmed by the detection of HIV antibodies in blood using enzyme-linked immunosorbent assay ( ELISA ) and western blot analysis ., The following data abstracted from the hospital electronic medical record: age , sex , racial and ethnic identity , the route of HIV transmission , and the stage of AIDS ( the name were anonymized in the supporting information ) ., This study was approved by the Institutional Ethics Committee of The First Affiliated Hospital , Medical College of Zhejiang University and complied with the Declaration of Helsinki guidelines ., TRUST ( Rongsheng Biotech Co . , Ltd , Shanghai , China ) was used as the non-treponemal test , and TPPA ( Fujirebio INC , Tokyo , Japan ) and TP-EIA ( Wantai Biological Pharmacy Enterprise Co . , Ltd , Beijing , China ) were used as the treponemal tests ., Every sample ( one per patient ) was tested by TRUST , TPPA , and TP-EIA simultaneously ., All testing was performed according to the manufacturer’s instruction ., The performing assay technician was unaware of the results of other testing , and all the results were reported independently ., The results of syphilis serological testing were interpreted following different algorithms respectively ., The definition of serological diagnosis of syphilis under different algorithms is illustrated in Fig 1 ., In the traditional algorithm , samples were screened by TRUST test , and the positive samples would be checked by TPPA test ., If the TPPA test also gave a positive result , the sample will be considered as positive for syphilis by serodiagnosis ., In the reverse algorithm , samples were screened by TP-EIA test , and the positive samples would be referred to the results of TRUST test ., If the TRUST test is positive , the sample is thought to be infected by syphilis ., When an inconsistent result was got , the sample would be judged by TPPA test in addition ., In the ECDC algorithm , samples were screened by TP-EIA test , and the reactive samples were confirmed by TPPA test ., According to the results of TPPA assay , the positive percent agreement , negative percent agreement and total percent agreement , each with 95% confidence interval ( CI ) , of the TP-EIA and TRUST assays were calculated by standard 2 x 2 contingency tables ., In addition to percent agreement , kappa coefficients were calculated as a secondary measure of agreement ., The seroprevalence of syphilis using traditional and reverse algorithms were compared using McNemars test for paired proportions ., Statistical analysis was performed using SPSS , version 20 ( version 20; IBM Corp . , Armonk , NY , USA ) ., As shown in Table 1 , the 865 HIV infected individuals had a mean age of 40 . 7 ( range 17–81 ) years , and the male accounted for 82% ., The majority of them ( 87 . 1% ) were of Han ethnicity ., More than half ( 58 . 7% ) of the HIV infected individuals were transmitted by heterosexual ., Among the 865 patients , 382 ( 37 . 9% ) patients were in AIDS stage and 1 patient’s stage of HIV infection was unavailable ., The serological test results of syphilis are illustrated in Fig 2 ., Overall , 123 subjects had TP-EIA +/TPPA+/TRUST+ results , and 602 subjects had TP-EIA −/TPPA−/TRUST− results ., 90 patients were TP-EIA+/TPPA+/TRUST− ., In order to exclude the prozone phenomenon , the TRUST tests for the 90 TP-EIA+/TPPA+/TRUST− subjects were repeated with serum samples diluted from 1:1 to 1:32 , and no subjects were found activate with TRUST after dilution ., The total percentages of agreement and corresponding kappa values of each assay’s results compared with those of TPPA were as follows: for TP-EIA , 98 . 4% , 0 . 960; for TRUST , 85 . 2% , 0 . 566 ., These data indicated that there was a very good strength of agreement between the TPPA test and the TP-EIA ., Using the TPPA test as the standard test , the TP- EIA had 100% positive percent agreement and 97 . 9% negative percent agreement ( Table 2 ) ., When the data was analyzed in the AIDS group and non-AIDS group , the results were similar to the total HIV infected individuals ( S1 Table ) ., In the traditional algorithm ( Fig 1 ) , 161 ( 18 . 6% ) samples were reactive with TRUST ., Of 161 TRUST positive samples , 123 ( 76 . 4% ) were confirmed as positive by TPPA and were suggestive of syphilis ., 38 ( 23 . 6% ) were considered to be false-positive by the TRUST ., The rate of serodiagnosis of syphilis was 14 . 2% ( 95% confidence interval CI , 11 . 9%– 16 . 6% ) using the traditional algorithm ., In the reverse algorithm ( Fig 1 ) , 227 ( 26 . 2% ) samples tested positive with TP-EIA ., 125 ( 55 . 1% ) of the 227 TP-EIA positive samples were TRUST-positive ., Discordant samples ( n = 102 ) were tested with TPPA and 90 ( 88 . 2% ) tested positive ., 12 ( 11 . 8% ) samples had negative TPPA results ., The rate of serodiagnosis of syphilis was 24 . 9% ( 95% CI , 22 . 0%– 27 . 7% ) using the reverse algorithm ., In the ECDC algorithm ( Fig 1 ) , 213 ( 93 . 8% ) of the 227 TP-EIA positive samples were confirmed by TPPA ., The rate of serodiagnosis of syphilis was 24 . 6% ( 95% CI , 21 . 7%– 27 . 5% ) using the ECDC algorithm ., Of the 213 samples diagnosed by ECDC algorithm , 123 ( 57 . 7% ) samples were active with TRUST ., The reverse algorithm demonstrated significantly higher seroprevalence of syphilis than the traditional algorithm ( 24 . 9% vs . 14 . 2% , p < 0 . 001 ) in the 865 HIV infected patients ., The 123 patients diagnosed by the traditional algorithm were also confirmed by the reverse screening algorithm , while the reverse screening algorithm detected an additional 92 patients that could not be detected using the traditional algorithm ., Compared to the reverse algorithm , the traditional algorithm also had a missed serodiagnosis rate of 42 . 8% ., The situation is similarly when compared the traditional algorithm with the ECDC algorithm ( with a missed serodiagnosis rate of 42 . 3% ) ., Among the 92 patients , 90 patients were TP-EIA+/TPPA+/TRUST− and 2 patients were TP-EIA+/TPPA−/TRUST+ ., The seroprevalence of syphilis screened by traditional algorithm , reverse algorithm and ECDC algorithm in AIDS stage and non-AIDS stage group was 13 . 1% vs . 14 . 9% ( p = 0 . 46 ) , 26 . 2% vs . 24 . 1% ( p = 0 . 48 ) , and 25 . 9% vs . 23 . 9% ( p = 0 . 50 ) , respectively ., Both in AIDS stage and non-AIDS stage group , the traditional algorithm showed significantly lower seroprevalence of syphilis than the reverse algorithm and ECDC algorithm ( Fig 3 ) ., The total percentages of agreement and corresponding kappa values of each algorithm’s results compared with those of reverse algorithm were as follows: for tradition algorithm , 89 . 4% , 0 . 668; for ECDC algirithm , 99 . 8% , 0 . 994 ., Using the reverse algorithm as the standard test , the tradition algorithm had 57 . 2% positive percent agreement and 100% negative percent agreement ( Table 3 ) ., Compared the traditional algorithm with the reverse algorithm , the positive percent agreement between the non-AIDS group and AIDS group had no statistically significant difference ( 62% vs . 50% , p = 0 . 08 , S2 Table ) ., Serological testing of syphilis remains an important component in the diagnosis of syphilis ., Latent syphilis , which is without clinical symptoms , is mainly detected by the non-treponemal and treponemal serologic tests ., Treponemal tests become positive in the 2–4 weeks after infection , and it can be detected after successful treatment , even persist lifetime ., Non- treponemal tests become positive about 2 weeks later than Treponemal tests ., Titers of non-treponemal tests are generally related to disease activity , and it can be declined to negative after successful therapy ( except for serofast phenomenon ) ., Non-treponemal tests are mainly used to monitor disease activity and assess the response to treatment ., Non-treponemal tests are not sensitive for latent , primary , tertiary syphilis and neurosyphilis , as well as successful treated syphilis ., Our study showed a very good strength of agreement between TP-EIA and TPPA , while the TRUST only have a 57 . 7% positive percent agreement with TPPA ( κ = 0 . 566 ) in HIV positive patients ., These results indicated the insensitive situations of TRUST in HIV infected individuals are common , especially in the AIDS group ., Treponemal tests first or non-treponemal tests first ?, Does the Order Matter ?, That is the key difference between the tradition algorithm and the reverse algorithm ., Nowadays , there is no uniform screening method for syphilis ., Public health decisions on which algorithm should be employed depending on many factors , including disease prevalence , cost , ease of use , and suitability for automation ., It is important to consider the screening abilities of different algorithms in the same population ., Matthew7directly compared the traditional and reverse syphilis screening algorithms in a population with a low prevalence of syphilis ., Their results showed that among 1000 patients tested , 6 patients were falsely reactive by reverse screening , compared to none by traditional testing ., However , reverse screening identified 2 patients with possible latent syphilis that were missed by traditional testing ., In HIV-positive individuals , the situation is more complex ., This is the first direct comparison of the reverse and traditional syphilis screening algorithms in a HIV-infected population ., Our present study found that among 865 patients tested , the reverse screening algorithm diagnosed an additional 92 patients that could not be observed using the traditional algorithm ., The missed diagnosis rate of the traditional screening algorithm was 42 . 8% compared with the reverse screening algorithm , which is higher than the study by Tong 8 ., That was a large survey conducted in an area with a high prevalence of syphilis ( 11 . 4% ) ., Previous studies9 have suggested that reverse screening can yield a high false-positive rate , while many early studies lacked parallel traditional screening on the same samples ., Our study found the false-positive rate of reverse screening was lower than traditional screening ( 1 . 4% vs . 4 . 4% ) and our finding were consistent with Tong’s findings ., The prevalence of syphilis of the participants may contribute to the difference ., We and Tong’s study were carried out in a population with a high prevalence of syphilis ., Our study showed there was a very good strength of agreement between the reverse and ECDC algorithm , and demonstrated that the seroprevalence of syphilis using the reverse algorithm ( or the ECDC algorithm ) was significantly higher than the traditional algorithm for HIV-positive individuals ., The 92 patients missed by traditional algorithm contribute to this difference , of which , 90 TRUST−/TP-EIA+/TPPA+ patients were the majority ., Patients with discordant TRUST and TP-EIA serological results are confirmed by TPPA ., If TPPA is non-reactive , it is considered to be false-positive ., When TPPA is reactive , there are 3 interpretations, ( i ) successfully treated syphilis infection;, ( ii ) early/late or latent syphilis , when the sensitivity of TRUST is low;, ( iii ) the prozone phenomenon , especially in secondary syphilis ., The prozone phenomenon in syphilis testing refers to a false-negative response resulting from an excess of antibody , which prevents visible agglutination in agglutination or precipitation tests ., Beyond our expectation , no prozone phenomenon were found among the 90 TRUST−/TP-EIA+/TPPA+ patients in the present study , which is lower than Jeffrey’s 10study ( 0 . 90% , 2/223 ) ., May be it is due to the small sample size , and it needs to evaluate the rate of prozone phenomenon in HIV infected individuals in a larger sample size ., There are several limitations to our study and the results should be interpreted with caution ., First , all specimens were obtained from hospital patients and there is consequent sample selection bias ., Second , the study was conducted from the perspective of serological diagnosis , and both the clinical diagnosis and prior history of syphilis were not analyzed ., In conclusion , screening of HIV-populations using different algorithms may result in a statistically different seroprevalence of syphilis ., When comparing the prevalence of syphilis in HIV-infected individuals from different surveys , it is important to assess which screening method is employed ., Finally , we advocate the reverse algorithm ( or the ECDC algorithm ) approach for the screening of syphilis in HIV-infected populations , given its sensitivity for early/late and latent syphilis ., The quantitative non-treponemal tests were recommended to determine serological activity of syphilis in ECDC algorithm ., The tradition algorithm approach underestimates the prevalence of syphilis in HIV-infected individuals . | Introduction, Materials and methods, Results, Discussion | Currently , there are three algorithms for screening of syphilis: traditional algorithm , reverse algorithm and European Centre for Disease Prevention and Control ( ECDC ) algorithm ., To date , there is not a generally recognized diagnostic algorithm ., When syphilis meets HIV , the situation is even more complex ., To evaluate their screening performance and impact on the seroprevalence of syphilis in HIV-infected individuals , we conducted a cross-sectional study included 865 serum samples from HIV-infected patients in a tertiary hospital ., Every sample ( one per patient ) was tested with toluidine red unheated serum test ( TRUST ) , T . pallidum particle agglutination assay ( TPPA ) , and Treponema pallidum enzyme immunoassay ( TP-EIA ) according to the manufacturer’s instructions ., The results of syphilis serological testing were interpreted following different algorithms respectively ., We directly compared the traditional syphilis screening algorithm with the reverse syphilis screening algorithm in this unique population ., The reverse algorithm achieved remarkable higher seroprevalence of syphilis than the traditional algorithm ( 24 . 9% vs . 14 . 2% , p < 0 . 0001 ) ., Compared to the reverse algorithm , the traditional algorithm also had a missed serodiagnosis rate of 42 . 8% ., The total percentages of agreement and corresponding kappa values of tradition and ECDC algorithm compared with those of reverse algorithm were as follows: 89 . 4% , 0 . 668; 99 . 8% , 0 . 994 ., There was a very good strength of agreement between the reverse and the ECDC algorithm ., Our results supported the reverse ( or ECDC ) algorithm in screening of syphilis in HIV-infected populations ., In addition , our study demonstrated that screening of HIV-populations using different algorithms may result in a statistically different seroprevalence of syphilis . | Syphilis remains a worldwide public health concern as there has been a global increase in the incidence of syphilis ., Serologic tests are still considered the mainstay of syphilis diagnosis ., Currently , there are three algorithms for screening of syphilis- traditional algorithm , reverse algorithm and European Centre for Disease Prevention and Control ( ECDC ) algorithm ., But there is no uniform screening method for syphilis ., Different surveys use different screening algorithm ., Will the different screening algorithm influence the seroprevalence of serodiagnosis of syphilis in this unique population ?, For those infected with both syphilis and HIV , the situation is even more complex ., To the best of our knowledge , no studies have analyzed the different algorithms for detecting syphilis in HIV-positive people ., Therefore , we compared the results of the three syphilis screening algorithms in an attempt to evaluate their screening performance in this unique population ., Our results supported the reverse ( or ECDC ) algorithm in screening of syphilis in HIV-infected populations ., In addition , our study demonstrated that the tradition algorithm approach underestimates the prevalence of serodiagnosis of syphilis in HIV-infected individuals . | urology, hiv infections, medicine and health sciences, pathology and laboratory medicine, pathogens, applied mathematics, tropical diseases, microbiology, simulation and modeling, algorithms, treponematoses, bacterial diseases, immunodeficiency viruses, viruses, retroviruses, mathematics, rna viruses, sexually transmitted diseases, neglected tropical diseases, bacterial pathogens, research and analysis methods, infectious diseases, serology, medical microbiology, aids, hiv, serodiagnosis, microbial pathogens, treponema pallidum, diagnostic medicine, genitourinary infections, viral pathogens, biology and life sciences, viral diseases, physical sciences, lentivirus, organisms, syphilis | null |
journal.ppat.1002093 | 2,011 | Mycobacterium tuberculosis Uses Host Triacylglycerol to Accumulate Lipid Droplets and Acquires a Dormancy-Like Phenotype in Lipid-Loaded Macrophages | One-third of the world population is latently infected with Mycobacterium tuberculosis ( Mtb ) and this vast reservoir is expected to contribute towards an increasing incidence of tuberculosis ( TB ) disease ., The World Health Organization estimated recently that there were 11 million prevalent cases of the disease and 1 . 8 million deaths annually due to TB , including 0 . 5 million deaths in HIV-positive patients 1 ., Mtb , the causative agent , is inhaled as an aerosol and enters the lung where it infects the alveolar macrophages and eludes host defenses ., The primary immune response of the host controls bacillary multiplication and causes the pathogen to enter a state of dormancy and become phenotypically antibiotic tolerant leading to latent TB 2 , 3 , 4 ., As a result of the host immune response , the pathogen is contained within the granuloma which is made up of infected macrophages surrounded by foamy lipid-loaded macrophages , mononuclear phagocytes and lymphocytes enclosed within a fibrous layer of endothelial cells 5 , 6 , 7 ., Mtb can persist inside the host for decades until the host immune system is weakened and then reactivates to cause active disease 3 ., It was established several decades ago that Mtb inside the host uses fatty acids as the major source of energy 8 ., Isocitrate lyase ( icl ) , which has been known to be a key enzyme of the glyoxylate cycle used by organisms that live on fatty acids 9 , was shown to be vital for the pathogens persistence inside the host demonstrating the critical role of fatty acids as an energy source for Mtb 10 ., Based on the observation that fatty acids are normally stored as triacylglycerol ( TAG ) in the adipose tissues of mammals , seed oils of plants and as lipid inclusion bodies in prokaryotes for use as energy source during and after dormancy/ hibernation , TAG was postulated to be the storage form of energy for latent Mtb 11 ., Intracellular lipid inclusion bodies were initially observed in Mtb more than six decades ago and were more recently detected in mycobacteria isolated from the sputum of TB patients 12 , 13 ., We showed that TAG accumulation is a critical event of Mtb dormancy and reported the discovery of triacylglycerol synthase 1 ( tgs1 ) as the primary contributor to TAG synthesis within the pathogen and that the deletion of tgs1 led to a nearly complete loss in TAG accumulation by Mtb under in vitro dormancy-inducing conditions 11 , 14 , 15 ., Recent observations from other groups have shown that the tgs1 gene is upregulated and TAG accumulates in dormant Mtb found in the sputum of TB patients and in the widespread , multi-drug resistant W/Beijing strain of Mtb 16 , 17 ., The source of fatty acids for synthesis of the TAG that accumulates as lipid droplets in the pathogen remains unknown ., In humans with untreated pulmonary TB , caseous granulomas in the lungs were shown to contain lipid-loaded foamy macrophages which harbored acid-fast bacilli 7 ., Such lipid-loaded macrophages which are found inside the hypoxic environment of the tuberculous granuloma contain abundant stores of TAG and are thought to provide a lipid-rich microenvironment for Mtb 5 , 6 ., Human macrophages cultured under hypoxia ( 1% O2 ) accumulate TAG in lipid droplets 18 ., Mtb-infected human alveolar macrophages are most likely enclosed in a hypoxic environment within the granuloma where the pathogen becomes dormant ., It was shown recently that tuberculous granulomas in guinea pigs , rabbits and non-human primates were hypoxic 19 ., It is well recognized that nonpulmonary tissue oxygen concentrations within the human body are far below the oxygen concentration in ambient room air and the typical oxygen level in standard in vitro cell cultures is much higher than that encountered by macrophages inside the human body 20 , 21 ., Furthermore , the oxygen concentration in the phagosome of activated macrophages was shown to be lower than the extracellular oxygen concentration 22 ., Dissemination of Mtb to distal sites such as the adipose tissue may also provide a TAG-enriched host environment for Mtb to go into dormancy 23 ., We postulate that Mtb inside lipid-loaded macrophages might import fatty acids derived from host TAG to accumulate TAG inside the bacterial cell and provide evidence to support this hypothesis ., We infected human peripheral blood mononuclear cell ( PBMC ) -derived macrophages and THP-1 derived macrophages ( THPM ) with Mtb and incubated them under hypoxia ( 1% O2 ) in order to mimic the microenvironment within the human lung granuloma ., We demonstrate that the macrophages accumulate lipid droplets under hypoxia ., Using single and double isotope labeling methods to metabolically label the host TAG , we determined that Mtb imports fatty acids released from host TAG to accumulate TAG within the bacterial cell ., Host fatty acids were incorporated intact into Mtb TAG ., We also show that host TAG that was metabolically labeled with a fluorescent fatty acid was imported by Mtb and accumulated as fluorescent lipid droplets within the bacterial cell ., Deletion of tgs1 resulted in a drastic decrease in radiolabeled and fluorescent TAG accumulation within Mtb inside THPM thereby revealing that synthesis of TAG within the pathogen from fatty acids released from host TAG constitutes the major pathway of TAG accumulation by Mtb inside the host ., We demonstrate that Mtb cells within lipid-loaded macrophages accumulate lipid droplets containing TAG , lose acid-fast staining and become phenotypically resistant to the two frontline antimycobacterial drugs , rifampicin ( Rif ) and isoniazid ( INH ) , all of which are thought to be indicative of the dormant state of the pathogen 2 , 11 , 15 , 24 , 25 ., Taqman real-time PCR analysis of gene transcripts of Mtb recovered from lipid-loaded macrophages revealed that genes thought to be involved in dormancy and lipid metabolism were upregulated within the pathogen ., Human alveolar macrophages , in which Mtb multiplies , probably reach a hypoxic environment within the granuloma , in which the pathogen goes into a latent state ., Such macrophages are likely to be lipid-loaded as a consequence of hypoxia and Mtb infection , both of which have been reported to induce lipid accumulation in macrophages in vitro 6 , 18 , 26 ., It is well known that nonpulmonary tissue oxygen concentrations within the human body are much lower than the oxygen level in ambient air and that caseous granulomas in rabbits are hypoxic 19 , 21 ., In order to mimic the hypoxic microenvironment within the granuloma , we infected human PBMC-derived macrophages and THPM with low numbers of Mtb ( MOI 0 . 1 to 5 ) and incubated them under 1% O2 , 5% CO2 ., About 3% of the host cells were infected at MOI 0 . 1 as determined by the CFUs recovered from the infected host cells after 4 h infection ., Oil Red O-staining lipid bodies increased upto 5 days in Mtb-infected macrophages as well as uninfected macrophages incubated under 1% O2 ( Figure 1A , D ) ., In contrast , lipid bodies increased moderately in macrophages incubated under 21% O2 ( Figure 1A , D ) ., TAG was the major lipid that accumulated in THPM lipid droplets under hypoxia and maximal levels were reached by day 5 ( Figure 1B and C ) ., Longer incubations resulted in greater loss of THPM from the adhered monolayer ( data not shown ) ., TAG accumulation in lipid bodies was also strongly induced under hypoxia in human PBMC-derived macrophages ( Figure 1 D , E ) ., Lipid droplets containing TAG increased greatly in size and number with time of culturing under hypoxia but only moderately under normoxia , and when normalized to viable macrophage cell counts , it was observed that TAG levels in hypoxic macrophages were much higher than that in normoxic macrophages ., There were considerable differences in lipid body formation between macrophages in the same population ., Since the photomicrographs showing selected fields of Oil Red O-stained macrophages do not adequately represent the TAG levels in the whole macrophage population , we relied on the analysis by thin-layer chromatography ( TLC ) of the TAG levels in the lipid extracts from the total population ., Since it was reported earlier that oxygenated mycolic acids , which are found only in virulent mycobacteria but absent in the non-virulent Mycobacterium smegmatis , were necessary for lipid body formation in macrophages under normoxic conditions 6 , we tested whether such a mechanism may be involved in lipid body formation in human macrophages under hypoxia ., Our observations indicate that macrophages accumulate TAG upon hypoxic stress alone since uninfected macrophages accumulated lipid droplets containing TAG to significantly higher levels under hypoxia than under normoxia ( Figure 1D , E ) ., We observed that the levels of TAG were slightly lower in M . smegmatis-infected macrophages than in Mtb-infected macrophages under hypoxia ., However , this difference was not significant under normoxic conditions ., Macrophages obtained from human PBMCs after differentiation for 7 days contained varying levels of small lipid droplets between different individual donors suggesting donor-to-donor variations in macrophage characteristics ., Moreover , the increase in lipid body size and number under hypoxia varied by different degrees between donors ., It has been well established by our group and others that intracellular TAG is accumulated inside Mtb under in vitro dormancy-inducing conditions and in Mtb from sputum of human TB patients and that the tgs1 gene product of Mtb is a major contributor to this process 11 , 13 , 14 , 15 , 16 , 17 ., In order to determine whether Mtb cells inside lipid-loaded macrophages utilized host lipids for accumulating TAG inside the bacterial cell , we radiolabeled macrophages with 1–14Coleate ( 10 µCi/ 7×106 THPM or 10 µCi/ 4×106 human PBMC-derived macrophages ) under 1% O2 , 5% CO2 for 24 h prior to infection with Mtb ( 5 bacilli per macrophage ) ., The infected macrophages were then incubated under 1% O2 , 5% CO2 for 3 days ., Mtb were recovered by lysing the host cells and centrifuging the lysate at 3500 x g ., As described in Methods , the 3500 x g pellets containing Mtb were washed thoroughly with mild detergent to remove host TAG adhering to the outside of the Mtb cells and any remaining host TAG was removed by enzymatic hydrolysis by TAG lipase which was followed by further detergent washes ., The 3500 x g pellet of uninfected host cell lysate was used as a control for background TAG levels ., We observed that radioactivity in TAG in the 3500 x g pellets of the infected human PBMC-derived macrophages and THPM was significantly higher than background controls suggesting that Mtb inside the host cells was utilizing the radiolabeled host lipids to accumulate TAG within the bacterial cell ( Figure 2A ) ., Moreover , TAG levels increased with time in live Mtb cells during infection of THPM under hypoxia but not in heat-killed Mtb cells indicating that intracellular TAG accumulation required active processes in live Mtb ( data not shown ) ., In order to determine whether the TAG that accumulates in Mtb cells inside radiolabeled macrophages is indeed intra-bacterial TAG , we labeled 7×106 THPM with about 18×106 dpm 9 , 10-3Holeic acid ( per data point ) under 1% O2 for 24 h prior to infection with Mtb ., About 17×106 dpm ( 98% ) of the radiolabeled oleic acid was taken up by the host cells under these conditions and incorporated into TAG ., The radioactivity in THPM TAG ( 1 . 5×106 dpm ) accounted for nearly 32% of total macrophage lipids ( 4 . 7×106 dpm ) ., The radiolabeled THPM were infected with Mtb at an MOI of 5 . 0 and incubated 3 days under 1% O2 ., As shown in Figure 2B , the detergent washes and lipase treatment were effective in removing radioactive material from the exterior of the Mtb cells ., Extraction and TLC analysis of lipids from the washed Mtb cells revealed that radiolabeled macrophage-derived fatty acids were indeed imported and stored as TAG inside Mtb ( Figure 2B , Mtb TAG ) ., In order to determine whether Mtb is capable of importing fatty acids derived from TAG outside the bacterial cell and confirm the above findings which suggested that Mtb utilized radiolabeled fatty acids from host TAG to accumulate radiolabeled TAG inside the bacterial cell , we incubated Mtb with radiolabeled TAG in culture medium ., A mid-log phase culture of Mtb was labeled with 14C-triolein for 2 h under aerobic conditions ., The Mtb cells were then washed with detergent and treated with lipase to remove any radiolabeled TAG that may be adhered to the extracellular surface of the Mtb cells ., Intracellular Mtb lipids and lipids in the washes prior to and after lipase treatment were resolved on silica-TLC ., The autoradiogram shown in Figure 2C reveals that the washes combined with lipase treatment were effective in removing TAG adhered to the exterior of the Mtb cell ., Post-lipase washes had almost no TAG ., Bacterial lipids were not removed by the lipase treatment and washes ., Most importantly , the lipid extract from the washed Mtb cells showed that TAG is stored inside the bacterial cell ( Figure 2C , Intra-Mtb lipids ) ., It is also evident that the radiolabeled fatty acids imported from extracellular TAG into Mtb are utilized by the bacteria for synthesis of other Mtb lipids ., In order to determine whether the major triacylglycerol synthase gene of Mtb ( tgs1 ) is involved in TAG accumulation inside the bacilli within hypoxic lipid-loaded THPM , we infected THPM radiolabeled with oleate with Mtb wild-type or Mtb Δtgs1 mutant and incubated the infected host cells under hypoxia for 3 days ., Deletion of tgs1 resulted in a severe reduction , but not a complete loss , of radiolabeled TAG accumulation by Mtb inside THPM ( Figure 2D ) ., This finding suggested that the TAG accumulated inside Mtb within lipid-loaded macrophages was synthesized mainly by re-esterifying host lipid-derived fatty acids into TAG by Mtb tgs1 gene product ., The TAG that accumulates in the tgs1 mutant is probably generated by the other Mtb tgs gene products ., In order to obtain an independent confirmation of our findings above that indicated the import of host TAG-derived fatty acids by Mtb and their subsequent accumulation as TAG within the bacterial cell , we metabolically labeled host TAG with the fluorescently-tagged fatty acid , BODIPY 558/568 C12 ., The fluorescent fatty acid was incorporated into the lipid bodies accumulating in THPM under 1% O2 ( Figure 3A , D ) ., When these THPM containing fluorescent lipid bodies were infected with Mtb and incubated under 1% O2 , the pathogen became loaded with well defined , highly fluorescent lipid bodies inside the cells ( Figure 3B , C , E ) ., Infection of THPM at MOIs of 0 . 1 and 0 . 25 yielded similar results and we selected an MOI of 0 . 25 in order to have higher numbers of Mtb cells for statistical purposes ., Optical cross-sectioning of the 3D image of Mtb containing fluorescent fatty acid-labeled lipid droplets revealed that the lipid droplets were intracellular suggesting that the pathogen generates intracellular TAG using fatty acids derived from host TAG ( Figure 3E ) ., TLC analysis of lipids extracted from fluorescent fatty acid-labeled THPM and Mtb recovered from such THPM revealed that TAG was the predominant lipid in both ( Figure 3F ) ., Thus , Mtb imports the fluorescently labeled fatty acids derived from the host TAG and accumulates fluorescent , intracellular lipid droplets comprised mostly of TAG ., Microscopic measurement of fluorescence in a population of 250 individual Mtb cells recovered from THPM showed that after 2 days of infection , most of the Mtb cells contained lipid droplets with intermediate levels of fluorescence and a smaller number contained intensely fluorescent lipid droplets ( Figure 3G ) ., At day 3 , the cellular distribution of fluorescence intensities was clearly bimodal , with about 8% of the cells in a high-fluorescence-level subpopulation ., Most of these cells in this subpopulation contained well defined lipid bodies ., This fraction of the population was reduced to about 1% in mutant Mtb cells lacking tgs1 ., Overall , the fluorescent TAG accumulation was severely reduced in the Δtgs1 mutant at all time points compared to the wild type ., These results suggest that tgs1 plays a significant role in the synthesis of TAG within Mtb from host TAG-derived fatty acids ., To examine whether Mtb inside THPM imported intact host TAG or fatty acids from host TAG , we metabolically labeled THPM with dual isotope labeled triolein glycerol-1 , 2 , 3–3H , carboxyl-1-14C under 1% O2 for 24 h prior to infection with Mtb at an MOI of 5 . 0 ., If Mtb hydrolyzed host TAG and the fatty acids were used for TAG synthesis within Mtb , the 3H:14C ratio of the TAG from Mtb recovered from THPM should be different and probably less than that of the host TAG ., If host TAG was taken up intact by the pathogen , the isotopic ratio of TAG in the pathogen should be the same as that of host TAG ., Mtb were recovered from THPM after 3 days in 1% O2 , washed with mild detergent and treated with TAG lipase to remove contaminating extracellular host TAG prior to lipid extraction ., Uninfected background controls did not contain TAG thereby demonstrating the efficacy of detergent washes and lipase treatment in removing host TAG contamination in the 3500 x g pellets containing Mtb ( Figure 4 , lanes UI ) ., TLC analysis showed that the radioactivity in the total lipid extracts of THPM was primarily in TAG and that Mtb accumulated radiolabeled TAG inside THPM ( Figure 4A and Table 1 ) ., Dual isotope-labeled TAG was purified from total lipid extracts of host ( 3500 x g supernatant ) and Mtb ( 3500 x g pellet ) by silica-TLC and the 3H:14C ratios were determined ., As indicated in Table 1 , the ratio of Mtb TAG was significantly lower than THPM TAG suggesting that TAG found in Mtb inside THPM was synthesized mainly from fatty acids released from host TAG ., The Mtb cells inside THPM also accumulated wax esters , albeit to a lower extent ( Figure 4A and B ) ., As can be clearly seen in the TLC analyses of the lipids in Figure 4 , it is evident that the lipid profile of Mtb recovered from radiolabeled macrophages ( lanes “Mtb” in Figure 4A and B ) , is markedly different from that of the host cells ( lanes “THPM” in Figure 4A and B ) and uninfected background controls ( lanes “UI” in Figure 4A and B ) ., Within the Mtb cell , the radiolabel was found distributed among TAG , fatty acids ( breakdown products of TAG ) , polar lipids ( synthetic products incorporating fatty acids ) and wax esters ( storage lipids containing fatty acids ) ., Thus , at the time of recovery of Mtb from the host cells , the Mtb cells were in the process of metabolizing the radiolabeled fatty acids derived from the host ., In order to determine whether fatty acids released from host TAG were incorporated intact into TAG , we metabolically labeled THPM with 9 , 10–3H , 1–14Coleic acid and infected them with Mtb at an MOI of 5 . 0 ., TLC analysis revealed that the radioactivity in THPM lipids was primarily in TAG and that the total lipid profiles of host and Mtb were markedly different ( Figure 4B , Table 2 ) ., If the host TAG-derived fatty acids were catabolized to acetate which was then used for fatty acid synthesis within Mtb , isotopic ratio of Mtb TAG should indicate a loss of 3H ., We found that the 3H:14C ratio of Mtb TAG was nearly identical to THPM TAG indicating that host TAG-derived fatty acids were incorporated intact into Mtb TAG ( Table 2 ) ., We compared the fatty acid compositions of 1-14Coleic acid-derived THPM TAG and TAG of Mtb recovered from THPM in order to obtain additional confirmation of the import of host fatty acids into Mtb TAG ., By resolving the fatty acid methyl esters of the THPM and Mtb TAG on argentation-TLC ( Figure 5A ) and reversed-phase silica-TLC ( Figure 5B ) , we found that all of the 14C in TAG from THPM and from Mtb isolated from THPM was found in oleic acid suggesting that host TAG-derived fatty acids were being incorporated into the TAG that accumulated within Mtb ( Figure 5A , B ) ., The fatty acid composition of unlabeled host and Mtb TAG was determined by purifying and analyzing the fatty acid methyl esters derived from TLC-purified TAG by capillary gas chromatography ., The fatty acid composition of the TAG from the pathogen was nearly identical to that of the host TAG ., C16:0 , C18:0 and C18:1 fatty acids were the dominant components in both the pathogen and the host ( Figure 5C ) ., Longer chain saturated fatty acids ( C24 , C26 and C28 ) that were present in very low amounts in the pathogen TAG were absent in the host TAG ., We conclude that the TAG that accumulated in the pathogen consists predominantly of fatty acids derived from the host TAG ., We assessed host cell numbers and viability in uninfected and infected THPM under hypoxia and normoxia and conclude that THPM cells incubated under 1% O2 are viable hosts for Mtb for upto 5 days at an MOI of 0 . 1 and upto 3 days at an MOI of 5 ., As determined by trypan blue dye exclusion method , THPM cell viability was about 90% in both cases ( Figure 6A ) ., At day 3 under 1% O2 , about 85% of the original THPM population infected with Mtb at an MOI of 0 . 1 remained adhered as a monolayer and 80% of the THPM infected at an MOI of 5 . 0 remained adhered ( Figure 6B ) ., By day 5 under 1% O2 , about 45% of the original Mtb-infected THPM population remained adhered as a monolayer loaded with lipid droplets after infection at an MOI of 0 . 1 ( Figure 6B ) ., Nearly half the host cells had perished under hypoxia and Mtb infection ., Of the THPM incubated under 21% O2 after infection with Mtb at an MOI of 0 . 1 , 80% remained adhered on day 3 but only 40% by day 5 ( Figure 6B ) ., Viability of these infected cells was about 70% at days 3 and 5 under 21% O2 ( Figure 6A ) ., THPM infected at an MOI of 5 . 0 and incubated under 21% O2 were completely overcome by Mtb multiplication by day 3 ( data not shown ) ., Viability of PBMC-derived macrophages ( infected and uninfected ) in the adhered monolayer was about 97% at 3 days and 92 % at 5 days under 1% O2 at both MOIs ., At 3 and 5 days under normoxia , about 95% of the adhered human macrophages ( infected and uninfected ) were viable ., The total viable cell counts ( by Trypan Blue dye exclusion method ) of adhered hypoxic human macrophages at 3 and 5 days were about 30% of the 0-day count of 5×105 macrophages per well of a 12-well plate and the respective counts for normoxic samples were about 45% of the 0-day count ., We determined the rate of Mtb multiplication within macrophages under 21% O2 and 1% O2 ., After normalization to the respective THPM cell counts , Mtb CFUs inside THPM under 1% O2 at day 5 , increased to about 5-fold of 0-day values ., In contrast , Mtb CFUs inside normoxic THPM increased to about 30-fold of 0-day values by day 5 ( Figure 6C ) ., Mtb CFUs in the extra-cellular medium were much lower than those inside adhered THPM monolayer ( data not shown ) ., Mtb replication within PBMC-derived macrophages under hypoxia was even more restricted than that inside hypoxic THPM ., At day 5 under hypoxia , Mtb CFUs in PBMC-derived macrophages normalized to macrophage cell count was about 3-fold of 0-day values ., In contrast , Mtb CFUs increased to about 34-fold of 0-day values at day 5 inside human PBMC-derived macrophages incubated under normoxia ., If the microenvironment inside hypoxic lipid-loaded macrophages mimics what happens in the hypoxic environment of the granuloma , we might expect Mtb within such macrophages to develop phenotypic drug resistance which is a key indicator of dormancy 2 , 4 , 15 ., To test for this possibility , we examined whether such phenotypic tolerance may be developed by Mtb within THPM and inside human PBMC-derived macrophages under 1% O2 ., At 0 , 3 and 5 days after infection , Mtb cells inside macrophages were exposed to antibiotic for 2 additional days , under the same conditions , prior to lysis of the host cells and recovery of the bacilli ., The antibiotic resistance , as a percentage of untreated control incubated for the same time-period under the same oxygen concentration , was determined by CFU determination after agar plating ., As shown in Table 3 , we found that phenotypic tolerance of Rif and INH of Mtb recovered from hypoxic THPM increased with time and reached maximal levels by 5 days under 1% O2 when about 8% of the total Mtb population was resistant to 5 µg/ml Rif and about 49% was resistant to 0 . 8 µg/ml INH ., Further incubation ( upto 16 days ) in 1% O2 decreased the percentage of antibiotic-resistant Mtb ( data not shown ) ., In contrast , Mtb inside normoxic THPM did not develop phenotypic tolerance of the antibiotics ( data not shown ) ., Mtb inside human PBMC-derived macrophages incubated under hypoxia also developed phenotypic tolerance to Rif and INH , as observed in THPM ( Table 3 ) ., Phenotypic resistance to Rif and INH increased to 18% and 43% respectively at day 7 inside hypoxic PBMC-derived macrophages ., In contrast , Mtb inside normoxic human macrophages showed much lower phenotypic tolerance to Rif ( 4% ) and negligible phenotypic tolerance ( 0 . 5% ) to INH at day 7 under normoxia ., Log-phase Mtb cultures used for infection and Mtb recovered from macrophages after 4 h infection and treated in vitro with antibiotics under normoxia for 2 days showed no resistance to Rif and INH ., Thus , Mtb developed phenotypic drug tolerance in hypoxic THPM as well as in hypoxic human PBMC-derived macrophages ., It has been established previously that dormant Mtb loses acid-fast staining and accumulates Nile Red-staining lipid droplets 13 , 15 , 16 , 25 ., In order to determine whether such a phenotype is developed by Mtb inside hypoxic lipid-loaded macrophages , Mtb cells recovered from human PBMC-derived macrophages after 0 , 3 and 5 days in 1% O2 were stained with Auramine-O and Nile Red ., We observed that , in addition to the bacilli that stained with either stain , there was a subset of bacilli in the total population that retained both stains ., The fraction of the Mtb population that stained with the green acid-fast stain ( Auramine-O ) decreased from about 86% at 0-day to about 40% at day 5 ., In contrast , Mtb cells that stained red with the lipid stain ( Nile Red ) increased with time from about 35% at 0-day to about 81% at 5-day inside hypoxic human macrophages ( Figure 7A–D ) ., Thus , by day 5 inside hypoxic macrophages , the fraction of acid-fast staining bacilli in the Mtb population decreased to half the level of the 0-day control , while the fraction that stained with Nile Red increased more than two-fold ., Moreover , at day 5 inside hypoxic macrophages , Mtb cells were markedly elongated in shape when compared to the 0-day controls ., In order to stain Mtb inside intact host cells , infected THPM after 5 days in 1% O2 were fixed with 4 % paraformaldehyde and stained with Auramine-O followed by Nile Red ., Mtb cells inside such intact THPM showed loss of acid-fastness and accumulation of Nile Red staining lipid droplets similar to the Mtb cells that were recovered from the macrophages before staining ( Figure 7E–G ) ., We examined the changes in transcript levels of selected Mtb genes that have been shown to be upregulated in a variety of in vitro and in vivo experimental models that mimicked dormancy 27 ., The gene for isocitrate lyase ( icl ) was induced ( Figure 8 ) , consistent with the idea that the pathogen in THPM utilizes fatty acids as the energy source ., Induction of dormancy- and stress-responsive genes , dosR ( Rv3133c ) and hspX ( Rv2031c ) , implicates the attainment of the dormant state by Mtb inside hypoxic , lipid-loaded THPM ., In our hypoxic THPM model , tgs1 ( Rv3130c ) , Rv3088 ( tgs4 ) , Rv1760 , Rv3371 and Rv3087 ( data not shown for this gene ) were found to be highly up-regulated at 72 h after infection ., It is noteworthy that lipY , that was previously reported to be involved in TAG mobilization 28 , was highly induced ., Induction of other lipase and cutinase-like genes suggests their possible involvement in the hydrolysis of host lipids ., The fatty acyl-coenzyme A reductase ( fcr ) genes Rv3391 and Rv1543 , that are involved in wax ester biosynthesis ( 15 , unpublished results ) were also upregulated ., Mtb can persist for decades inside the human body in the dormant state and reactivate when the hosts immune system weakens 4 ., HIV infection increases the risk of reactivation leading to the deadly synergy between AIDS and TB 3 , 29 , 30 ., Currently , there is no drug that can kill latent TB and the development of such antibiotics is critical to the cure and eradication of the disease 2 , 31 ., Novel drugs that target dormancy-specific metabolic pathways may enable the treatment of patients with multi- and extremely-drug resistant Mtb and drastically shorten the currently used , very long-term treatment period to cure TB ., Understanding of dormancy-specific processes and a model system to test for inhibition of such processes are required to discover such drugs ., The pathogen is likely to go into a dormant state within macrophages that are in the hypoxic environment of the granuloma 15 , 32 , 33 ., Such macrophages might be loaded with TAG-containing lipid bodies 6 , 7 , 18 ., Since one of our objectives was to develop an in vitro model that mimics the in vivo situation and is suitable for high-throughput screening , we used THPM as host cells in order to avoid the well known donor-to-donor variations in primary human macrophages and the technical difficulties involved in obtaining large , homogenous populations of alveolar macrophages for experimental purposes ., We validated our results obtained with hypoxic THPM by demonstrating similar observations in human macrophages which were derived from mononuclear cells isolated from the peripheral blood of healthy volunteers and subjected to hypoxia ., THPM , which are capable of lipid accumulation , were reported to faithfully model the apoptotic response of human alveolar macrophages in response to Mtb infection 34 , 35 ., Furthermore , the antimycobacterial activity of INH in THPM was similar to that in human monocyte-derived macrophages 36 ., The assumption that Mtb-infected human alveolar macrophages most likely reach a hypoxic environment within the granuloma serves as the basis for the well-studied in vitro hypoxic model of Mtb dormancy 33 ., Moreover , oxygen concentrations in healthy tissue within the human body are thought to range between 5 to 71 Torr and are well below the oxygen concentration of 157 Torr in ambient room air 20 , 21 ., The oxygen tension in caseous granulomas of rabbits was measured to be approximately 2 Torr ( ∼0 . 3 % O2 ) 19 ., Hypoxic , lipid-loaded macrophages may provide a lipid-rich sanctuary for Mtb during its dormancy ., The killing of Mtb by macrophages inside the hypoxic regions of the granuloma is likely to be severely inhibited since superoxide and NO production by macrophages are greatly diminished by hypoxia 21 , 37 ., Furthermore , electron paramagnetic resonance-based measurements have shown that oxygen concentration in the intraphagosomal compartment was significantly lower than the extracellular environment 22 ., However , macrophages infected in vitro with Mtb are currently incubated in normoxic environments where the oxygen level is far higher than that encountered by Mtb-infected macrophages inside the human lung granuloma ., Consequently , Mtb inside those macrophages are not subjected to the hypoxic stress encountered inside the granuloma and do not develop phenotypic tolerance of antibiotics such as Rif and INH 36 , 38 which is a key indicator of dormancy 2 , 4 , 15 ., In order to mimic the hypoxic micro-environment within the granuloma , we infected macrophages with Mtb and incubated them in a 1% O2 , 5% CO2 environment ., Under such co | Introduction, Results, Discussion, Materials and Methods | Two billion people are latently infected with Mycobacterium tuberculosis ( Mtb ) ., Mtb-infected macrophages are likely to be sequestered inside the hypoxic environments of the granuloma and differentiate into lipid-loaded macrophages that contain triacylglycerol ( TAG ) -filled lipid droplets which may provide a fatty acid-rich host environment for Mtb ., We report here that human peripheral blood monocyte-derived macrophages and THP-1 derived macrophages incubated under hypoxia accumulate Oil Red O-staining lipid droplets containing TAG ., Inside such hypoxic , lipid-loaded macrophages , nearly half the Mtb population developed phenotypic tolerance to isoniazid , lost acid-fast staining and accumulated intracellular lipid droplets ., Dual-isotope labeling of macrophage TAG revealed that Mtb inside the lipid-loaded macrophages imports fatty acids derived from host TAG and incorporates them intact into Mtb TAG ., The fatty acid composition of host and Mtb TAG were nearly identical suggesting that Mtb utilizes host TAG to accumulate intracellular TAG ., Utilization of host TAG by Mtb for lipid droplet synthesis was confirmed when fluorescent fatty acid-labeled host TAG was utilized to accumulate fluorescent lipid droplets inside the pathogen ., Deletion of the Mtb triacylglycerol synthase 1 ( tgs1 ) gene resulted in a drastic decrease but not a complete loss in both radiolabeled and fluorescent TAG accumulation by Mtb suggesting that the TAG that accumulates within Mtb is generated mainly by the incorporation of fatty acids released from host TAG ., We show direct evidence for the utilization of the fatty acids from host TAG for lipid metabolism inside Mtb ., Taqman real-time PCR measurements revealed that the mycobacterial genes dosR , hspX , icl1 , tgs1 and lipY were up-regulated in Mtb within hypoxic lipid loaded macrophages along with other Mtb genes known to be associated with dormancy and lipid metabolism . | Two billion people are latently infected with Mycobacterium tuberculosis ( Mtb ) ., Cure and possible eradication of tuberculosis are limited by the lack of availability of any drug that can kill dormant Mtb ., Understanding of the processes critical for dormancy and a reliable dormancy model suitable for high throughput screening of chemicals will help to discover drugs that can kill dormant Mtb ., Storage of lipids for utilization as energy source is critically needed for dormancy ., In the human lung , Mtb-infected macrophages are sequestered inside the hypoxic environments of the physical enclosure called granuloma in which Mtb becomes dormant ., None of the currently used cell culture models of Mtb infection mimic this situation ., We developed a model that mimics the environment inside the human granuloma by incubating Mtb-infected macrophages under hypoxia ., We found that , under these conditions , macrophages accumulate lipid droplets and Mtb within these macrophages acquire a dormancy phenotype ., We report how the pathogen inside the macrophages utilizes the host lipids to store lipids within the pathogen and acquire the hallmark traits of dormant Mtb ., Thus , our novel model of Mtb dormancy may enable better understanding of the metabolic processes vital for the dormant pathogen and help to discover drugs that can kill latent pathogens . | biochemistry, biology, microbiology, molecular cell biology | null |
journal.pcbi.1004930 | 2,016 | A Detailed Data-Driven Network Model of Prefrontal Cortex Reproduces Key Features of In Vivo Activity | The prefrontal cortex ( PFC ) is a key structure in higher-level cognitive functions , including working memory , rule and concept representation and behavioral flexibility 1–6 , and has been linked to impairments of these functions in psychiatric disorders like schizophrenia 7–10 or attention-deficit/hyperactivity disorder 11 ., Our understanding of the computational and dynamic mechanisms underlying these cognitive functions , their neuromodulation , and their aberrations in psychiatric disorders , is still very limited , however ., Computational network models are a highly valuable tool for driving forward such an understanding , as data from many different levels of experimental analysis can be integrated into a coherent picture ., With respect to psychiatric conditions , it is of particular importance that models incorporate sufficient biological detail and exhibit physiological validity in order to serve as explanatory tools ., Psychiatric conditions like schizophrenia are characterized by a multitude of abnormalities in diverse cellular and synaptic properties , transmitter systems , and neuromodulatory input 7–10 ., Moreover , pharmacological treatment options target the neurochemical and physiological level , yet they are supposed to change functionality at the behavioral and cognitive level ., It is thus crucial to gain insight into the explanatory links between behavioral functions and the underlying neurobiological “hardware” , a task that requires sufficient physiological detail in the model specification , in particular realistic assumptions about anatomical structure and cell type diversity ., Ultimately , the physiological validity of a computational model ought to be reflected in the degree to which it can reproduce and predict detailed aspects of the neural activity observed in vivo ., That is , from a statistical perspective , one may define a good , physiologically valid model as one that accurately ( i . e . , quantitatively ) captures distributions compiled from the electrophysiological activity ( spiking , field potentials , membrane voltages ) produced by networks in vivo , but not necessarily as one that captures every detail of membrane biophysics or receptor kinetics ., In our perception , such requirements are currently not met even by sophisticated cortical network models which do include a lot of biophysical detail 12–14 , as these are often only loosely compared to in vivo data or test only specific aspects of those ., In this work , we present a computational network model of the PFC which has high physiological validity and predictivity both at the single-neuron- ( in vitro ) and at the network- ( in vivo ) level , yet is still simple enough to be computationally tractable ., Its anatomical structure , neural , and synaptic properties are completely derived from the experimental literature and our own experimental data ., The activity of the network is compared with a range of statistics derived from in vivo data , including spike trains , local field potentials , and membrane potential fluctuations ., The model turns out to reproduce these data quantitatively , and also exhibits robustness with respect to moderate changes in parameters ., The network model introduced in Materials and Methods aims to combine computational tractability with physiological validity ., This balance is achieved by embedding a simple , reduced two-dimensional single neuron model into a realistic network architecture that is derived from the experimental literature ., All model parameters were directly estimated from our own in vitro data and the experimental literature ( see Materials and Methods for details ) , and no specific parameter tuning was necessary to bring the network model closer to in vivo-like behavior ., At the single-cell level , the network is based on an approximation ( simpAdEx 15 ) to the adaptive exponential integrate-and-fire model ( AdEx 16 ) which yields closed-form expressions for instantaneous and steady-state firing rates , thus allowing for fast and fully automatized fitting to f-I and V-I curves from physiologically recorded cells ( Fig 1A ) ., We had shown previously that this cell model is able to accurately predict spike times of recorded neurons driven by in vivo-like fluctuating currents not used for model fitting 15 ( Fig 1B ) and , like the full AdEx 17 , can generate a wide range of spike patterns ., In vitro recordings from ∼200 L2/3 and L5 pyramidal cells , fast-spiking and bitufted interneurons from the medial PFC of adult rodents were used to generate a distribution of model cells that reflects the diversity of neurons in the real PFC ( see Materials and Methods for details ) ., The resulting model parameters ( Table 1 ) follow broad distributions ( Fig 1C ) , mostly of Gaussian shape , with the exception of ΔT and τw which are best described by a Gamma distribution , and b which follows an exponential distribution ( red curves in Fig 1C indicate distributions from which model parameters were drawn ) ., Anatomically , the network is divided into two laminar components , representing the superficial layers L2/3 and deep layer L5 ( Fig 2A ) ., Neurons are distributed over the five cell types in each layer based on estimates from the literature ( Table 2 ) ., The neurons are randomly connected with different connection probabilities pcon for each pair of cell types according to the literature 18–27 , including local clusters of higher connectivity 28 , 29 ., The neurons are assumed to be organized in a single column and horizontal spatial distance is not taken into account ., However , all neurons receive a constant background current ( i . e . , without fluctuations ) that represents synaptic connections from outside the network , both within and outside the same column ( see section “Admissable and realistic range of input currents” below ) ., Since these currents were constant , all irregularity was produced intrinsically within the simulated network ., Neurons are connected by conductance-based synapses ( AMPA , GABAA and NMDA ) with kinetics estimated from electrophysiological data , short-term synaptic plasticity 30 that is matched to the types of the connected neurons 31 , 32 , synaptic delays and random failure of synaptic transmission 33–36 ., Distributions of synaptic weights ( log-normal 37 ) and delays ( Gaussian ) were extracted from the literature ( Table 3 ) ., The average connection strength ( connectivity pcon times synaptic peak conductance gmax ) between pyramidal cells and interneurons in the different columns and layers is indicated by the width of the arrows in Fig 2A ., Wherever possible , we used data from the rodent prefrontal cortex , or at least agranular cortices such as the motor cortex , which in rodents shows a similar layered anatomy as the PFC ., Apart from the missing granular layer 4 , specific features of the rodent PFC that are modeled here include an increased fraction of reciprocal compared to unidirectional connections 32 , longer NMDA time constants than in other areas 38 , 39 , and a uniquely prefrontal distribution of short-term synaptic plasticity properties for connections among pyramidal cells 32 ., To assess whether the network model can reproduce the dynamics of real prefrontal neurons in vivo , we compared measures computed from the model with those from electrophysiological data , as well as with a number of findings from the literature ., Unless otherwise stated , we simulate a single column with 1000 neurons and apply a constant DC current of 250 pA to all pyramidal cells and 200 pA to all interneurons ., These currents are the only parameters that are not directly obtained from experimental data ., As discussed below , appropriate values for these currents were derived by inferring from lumped-population input simulations the amount of current produced by a network of realistic size , set up with the very same structure as the explicitly modeled network ., Spike-train statistics ., All experimentally recorded spike trains ( kindly provided by Dr . Christopher Lapish , Indiana University Purdue University , Indianapolis , see 40 for details ) were first segregated into statistically stationary segments to yield estimates of spike train statistics that reflect in vivo baseline activity , free from task-related responses ( not modeled here ) or other potential confounds 41 ., For consistency , the same procedure was applied to the simulated spike trains , although , strictly , these were stationary by simulation setup ., From all jointly stationary segments , the mean 〈ISI〉 , coefficient of variation CV , and autocorrelation function of the inter-spike intervals ( ISIs ) were computed for each individual spike train , as well as the zero-lag cross-correlation CC ( 0 ) between pairs of neurons ( Fig 3 ) ., The in vivo data show very low zero-lag cross-correlations between neuron pairs ( 2 . 4 ⋅ 10−4 ± 2 . 5 , mean ± SD ) and CV s near one ( 1 . 04 ± 0 . 33 ) , consistent with the proposal of an “asynchronous-irregular”, ( AI ), state of cortical dynamics, ( although the correlations theoretically proposed for the AI state are usually even at least one order of magnitude larger than obtained here 42 ) ., The average single-cell ISIs follow a monotonically decreasing distribution with a mean comparable in size to the standard deviation, ( 570 ± 610 ms ) , but with a heavy tail that is better described by a log-normal or beta-2 distribution 43 rather than an exponential distribution ., The autocorrelation function shows a rapid decay with small negative flanks, ( half-width at half maximum: 10 . 1 ± 1 . 1 ms , minimum: 64 . 6 ± 69 . 9 ms , mean ± SD ) ., Without further tuning of network parameters beyond their derivation from slice-physiological and anatomical data , all these in vivo statistics are well reproduced by the model, ( Fig 3 ) ., Two-sample Kolmogorov-Smirnov tests did not find notable differences between experimental and simulated distributions in any of the statistics, ( CV : p = 0 . 26 , KS ( 29 ), = 0 . 28; mean ISI: p = 0 . 4 , KS ( 29 ), = 0 . 23; CC: p = 0 . 4 , KS ( 29 ), = 0 . 24 ) , indicating that simulated distributions were not statistically distinguishable from the experimental ones ., The asynchronous- irregular firing with low rates is also seen in the raster plot of spike times, ( Fig 3B ) ., Low fraction of spiking neurons and layer-dependent firing rates ., Fig 3B reveals a relatively low fraction of spiking pyramidal cells in both layers—only 22% of the cells emitted more than 10 spikes during the 30s of simulated time , which will be used as the definition of “spiking neurons” throughout the paper , in line with 44–46 ., Comparing the neural and synaptic parameters of those neurons which fire at a sufficiently high rate, ( > 0 . 33 Hz ), and those which do not, ( ≤ 0 . 33 Hz ) , we find that only the rheobase, ( and the cell parameters that contribute to it ), differs between the two populations: Spiking neurons have rheobases at the lower end of the distribution, ( 42 . 9 ± 2 . 1 pA ) , compared to 69 . 0 ± 1 . 6 pA for non-spiking neurons, ( mean ± SEM; p = 3 . 5 ⋅ 10−20 , t ( 997 ), = 9 . 4 , two-sided t-test ) , some of them even firing spontaneously, ( called “generator neurons” 47 ) ., While neurons firing at very low rates may go undetected using extracellular single-unit recordings , recording techniques that are less biased toward spiking neurons , such as calcium imaging or in vivo patch-clamp , often reveal a large fraction of neurons that are mostly silent, ( “dark matter theory” of neuroscience , 44–46 ) ., Consistent with these results , the fraction of neurons with more than 10 spikes rarely exceeded 40% in simulations with in vivo-like firing patterns, ( see section “Admissible and realistic range of input currents” below ) ., This can be explained by the way the neurons are activated: While most neurons receive a background current above their rheobase , the high firing rates of the interneurons, ( Fig 3B ), lead to an average membrane potential in the pyramidal cells below the firing threshold, ( mean difference: -17 . 3mV , range: -37 . 2 to -2 . 2mV for the example shown in Fig 3 ), that is occasionally kicked above threshold by random fluctuations ., This means that the firing rate is mostly determined by the amplitude of the fluctuations of the membrane potential, ( see below for statistics ) ., These results are qualitatively conserved across the range of input currents for which the overlap between experimental and simulated distributions is reasonably high ., Membrane potential and local field potential statistics ., In addition to the spike data , we also compared the membrane potential statistics and LFP signals between simulation and experiments ., For the simulated network , we observed a broad range of membrane potential fluctuations, ( after removing spike events; Fig 4A; 3 . 28 ± 0 . 72 mV , mean ± SD; range between 0 . 72 mV and 11 . 23 mV ) ., We compared this distribution of standard deviations with those from in vivo patch-clamp recordings from 10 putative pyramidal cells during up-states in anesthetized adult rodent PFC, ( kindly provided by Dr . Thomas Hahn , Central Institute of Mental Health and BCCN Heidelberg-Mannheim ) ., The simulated distribution is less than one SEM away from the average of the experimental distribution, ( pooled over all data sets ), for most bins , and a Kolmogorov-Smirnov test, ( see Materials and Methods ), does not show a significant difference, ( p = 0 . 45 , KS ( 29 ), = 0 . 23 ) ., The range of membrane potential fluctuations in the model and in the recordings used here is also consistent with values found in the literature 48 , 49 ., The local field potential, ( LFP ), in the model was estimated as the sum of all synaptic currents, ( allowing excitatory and inhibitory currents to partially cancel ) ., This is a reasonable approximation to the standard model of the LFP 50 under the assumption that all neurons are confined in a small volume of cortical space ., We computed the power spectral density of this model-derived signal and of the LFP signals obtained from the in vivo recordings, ( Fig 4B ) ., Up to a constant offset, ( that has been removed in the figure ) , the spectrum of the simulated LFP is less than one SEM away from the average estimated from the experimental recordings, ( from awake , behaving animals , also provided by Dr . Christopher Lapish 40 ), at most of the frequencies ., Both spectra follow a 1/f power law for frequencies below 60 Hz and change their scaling behavior for higher frequencies , consistent with LFP spectra described in the literature 51–53, ( the fluctuations in the simulated curve are stochastic in nature , i . e . there is no systematic deviation from the 1/f behavior across different simulations ) ., For frequencies beyond 60 Hz , the experimental spectrum is well described by a 1/f2 power law , while the simulated one rather follows a 1/f3 relation ., Both scaling behaviors have been reported in the literature, ( 1/f2 52 , 54 , 1/f3 51 ) , and the difference may result from the simplifications made in the computation of the simulated LFP , e . g . neglecting the spatial integration of currents in extracellular space or the contribution of active currents 14 ., Transient information transfer and the role of neuronal heterogeneity ., We next examined how neurons in L2/3 and L5 would respond to a simple stimulus simulated by a brief series of spikes at high rate, ( 250 spikes within 5 ms ), from a virtual, ( not explicitly simulated ), “input population” connected to 10% of the pyramidal cells in L2/3 ( cf . Table 3 ) ., The stimulus induces a number of spikes in L2/3 , and with a short delay also in L5 ( Fig 5A ) ., The delays ( L2/3: 8 . 9 ± 1 . 1 ms; L5: 17 . 7 ± 1 . 2 ms , mean ± SD ) are similar to values that have been reported in the literature ( e . g . 3 . 4 ± 0 . 5 ms in L2/3 and 16 . 6 ± 1 . 2 ms in L5 55 ) ., Note that these delays are significantly longer than the fixed synaptic delays ( below 2 ms , see Materials and Methods ) and arise from the dynamics of the neurons and the kinetics of the synapses ( c . f . 56 ) ., For a sufficiently strong stimulus ( e . g . 500 spikes within 5 ms ) , the neurons in L2/3 show a brief period ( 100–150 ms ) of persistent activity ( Fig 5B ) ., The transmission of transient stimuli between layers crucially depends on the heterogeneity of the neuronal parameters ., With a 80% reduction in the variance of all parameter distributions ( but no change in the means ) , the stimulus only elicits a response in L2/3 , but is not transmitted to the output layer L5 anymore ( Fig 5C ) ., Indeed , L2/3 activity is almost independent of neuronal variability , whereas the number of spikes in L5 systematically decreases as the standard deviation of neuronal parameters is reduced ( Fig 5D ) ., To further examine the transmission dynamics , we reproduced an in vitro experiment with suppressed inhibition 57 which showed that input in L2/3 resulted in an epileptiform spread of activation across the whole network under this condition , whereas the same input in L5 did not ., We mimicked this setup by reducing the inhibitory synaptic weights in the network to 30% of their original values and inducing a strong stimulus ( see above ) in each of the two layers , while varying the peak conductance gmax of the synaptic connection between the mimicked Poisson input population and the network ., For moderate connection strengths ( gmax = 2 ) , only a fraction of the network responds , and the number of spikes elicited by the network is much larger if the stimulus is injected in L2/3 ( 404 ± 116 , mean ± SD ) compared to a stimulus in L5 ( 118 ± 33 ) ., Higher connection strengths ( gmax = 20 ) reliably drive the network into an “epileptic state” ( transient high-rate response from all neurons in the network ) for a stimulus in L2/3 ., In contrast , this state was never reached for an input in L5 , consistent with the experimental results in 57 ., In the previous section we showed that the model can reproduce a wide range of characteristics of neural activity in vivo ., Here , we assess how the reproduction quality of in vivo-like behavior depends on those parameters of the model which were only loosely constrained by experimental data ., We restrict this analysis to the spike series statistics 〈ISI〉 , CV and CC ( 0 ) ., Admissible and realistic range of input currents ., The background currents I = I ex L23 , I ex L5 , I inh L23 , I inh L5 have so far been treated as free parameters , as such estimates are difficult to obtain or at least have not been reported experimentally ., We address this in two ways: First , we systematically vary these four currents and assess the similarity between experimental and simulated spike time distributions using Kolmogorov-Smirnov statistics as before ., Second , we estimate the required background currents from the simulation itself , using the assumption that the simulated network is embedded in a larger , but structurally identical network from which these currents originate ., Fig 6A shows the Kolmorgorv-Smirnov test statistic DKS as a function of Iex and Iinh , where I ex L23 = I ex L5 = I ex and I inh L23 = I inh L5 = I inh ( see below for a discussion of laminar differences in the input currents ) ., The figure reveals that the overlap between experimental and simulated distributions is acceptable ( p > 0 . 05 for the two-sample Kolmogorov-Smirnov test , i . e . failure to reject the null hypothesis H0 of equal distributions for the two samples , see Materials and Methods ) for a wide region of Iex and Iinh values ( delimited by the black isocline in Fig 6A , associated with DKS values below 0 . 4 ) ., More specifically , simulated CV and mean ISI distributions become indistinguishable from their experimental counterparts as Iex increases , while the overlap with the ISI distribution decreases again for very high Iex values ( Fig 6A , left inset ) ., Both CV and mean ISI deviate from the experimental distributions as Iinh increases ., CC , on the other hand , matches well with the experiments for high Iinh values ( Fig 6A , lower inset ) ., As mentioned above , the fraction of firing neurons is quite low in most networks showing in vivo-like firing patterns , typically between 20 and 30% , as shown in Fig 6B ( blackly delimited region gives the empirically acceptable parameter regime copied from Fig 6A ) ., The ratio of inputs into the two layers , I ex L23 / I ex L5 and I inh L23 / I inh L5 , does not have a strong influence on these results within the tested range ( mean DKS ± SEM: 0 . 31 ± 0 . 05 , 0 . 32 ± 0 . 05 and 0 . 35 ± 0 . 06 for ratios of 1 , 2 and 4 , respectively ) , but does of course affect the relative firing rate between the two layers ., In vivo experiments found that firing rates are considerably higher in L5 compared to L2/3 pyramidal cells ( 3–20 times 27 ) ., This condition is fulfilled in our model as long as L2/3 receives less or the same input as L5 ( Fig 6C ) ., To estimate which range of I values could be realistically assumed , we tested whether a substantially larger network than the 1000-neuron-network simulated here would produce mean synaptic currents that are large enough to self-sustain in vivo-like activity ( i . e . within the blackly circumscribed regions in Fig 6A ) ., In this case , the activity in the large network and the small network ( the latter driven by the currents from the larger one ) would be indistinguishable , and the in vivo-like activity would be supported by the larger network ., We increased the size of the network either by changing the density of neurons or by adding input from nearby columns ( see “Estimation of background currents” in Materials and Methods ) ., Fig 7 shows the mean synaptic current into pyramidal cells and interneurons in L2/3 and L5 that would result from the reduced equivalent-population input models described in Materials and Methods if the network size was varied through the number of columns ( Fig 7A ) or the density of neurons within columns ( Fig 7B , both figures showing currents averaged over the values of the other independent variable , i . e . neural density or number of columns , respectively ) ., The shaded areas show the ranges for Iex ( blue ) and Iinh ( red ) within which these currents would produce in vivo-like activity ( DKS < 0 . 4 ) ., Note that it is sufficient that one of the two layers receives a current above the lower bound , as it will push the other layer into the right regime by cross-layer synaptic connections ., The upper bound , on the other hand , may not be exceeded by either of the two layers , as this would push the other layer beyond its upper bound as well ., It is apparent that these conditions are fulfilled already for ( spatially ) relatively small networks ( ∼ 5 columns ) , and currents saturate as network size grows further ( Fig 7A ) ., By increasing the neuron density , on the other hand , the input currents increase monotonically over a wide range ( Fig 7B , averaged over all column numbers ≥ 5 ) ., Mean synaptic currents sufficient to drive the network into the experimentally observed regime arise for densities between 19 , 000 and 44 , 000 neurons per mm3 ., This range overlaps with densities found in anatomical studies ( 30 , 000 to 90 , 000 neurons per mm3 58–60; horizontal dotted lines ) ., Variation of synaptic parameters ., We attempted to estimate all synaptic parameters from data reported in the literature ., Given that these come with some uncertainty and variation , however , we explored how sensitive the network behavior is with respect to changes in mean synaptic peak conductances and their distribution , synaptic time constants , and the GABAA reversal potential ., All these parameter variations were performed for a range of different background currents and averaged results are reported ., The GABAA reversal potential E rev GABA was initially set to -70mV , which is well within the range of the values reported in the literature 19 , 24 , 55 , 61 ., Within the physiologically reasonable range from -90 to -60mV 62 , the divergence between simulated and experimental distributions ( as assessed by the KS test statistic ) increases with E rev GABA ( Fig 8A ) ., At the same time , the standard deviation of the membrane potential decreases ., The time constants of the synaptic kinetics also turned out to be important for the agreement with in vivo data: While small changes are acceptable , both very fast and very slow GABAA kinetics strongly diminish the agreement with the experimental data ( DKS = 0 . 99 for τon = 0 . 6 ms and τoff = 8 ms and DKS = 0 . 85 for τon = 12 ms and τoff = 160 ms ) ., The NMDA time constants have less effect , unless they are very strongly increased ( DKS = 1 . 0 for τon = 17 . 2 ms and τoff = 300 ms , compared to values of τon ≤7 ms and τoff ≤ 100 ms reported in the literature 38 , 39 ) ., The effects of the mean synaptic peak conductances are shown in Fig 8B ., While small to moderate changes ( ± 50% ) have no significant effect , a strong decrease in the inhibitory synaptic efficiencies leads to a significant mismatch with the in vivo statistics ., Apart from the mean , we also analyzed how the distribution of the synaptic peak conductances affected in vivo-like behavior by either reducing the variability or drawing them from a normal rather than a log-normal distribution ( conserving mean and standard deviation ) ., Reducing the variability of the synaptic weights increased the mismatch between empirical and simulated distributions ( Fig 8C , gray line ) ., Surprisingly , just changing the form of the underlying distribution from log-normal to normal , without changing its mean or standard deviation , had a similarly strong effect as a pronounced reduction in standard deviation ( Fig 8C , black line ) , so both the variability as well as the functional form of the synaptic conductance distribution are crucial for reproducing spiking dynamics as observed in vivo ., Without the long tail of the log-normal distribution , the network activity becomes much more synchronized ( CC ( 0 ) : 0 . 017 ± 0 . 006 , mean ± SD ) and exhibits strong bursts ( CV : 1 . 81 ± 0 . 09 , mean ± SD ) , while mean firing rates are not much affected ( 〈ISI〉: 420 ± 558 ms , mean ± SD ) ., The current model has a strong focus on its tight connection to data ., Many existing network models of the neocortex are based on neurobiological findings as well 13 , 14 , 63–65 , but the present model differs from them in two respects: The strict way in which the in vitro data is used to fix or systematically infer every detail of the model , and , more importantly , the quantitative test of the model’s validity on a wide range of in vivo findings ., Recently , a few studies have also moved in this direction ., Fisher et al . 66 proposed a model for the short-term memory circuit in the oculo-motor system of the adult goldfish ., They fitted the model simultaneously to a range of anatomical , physiological and behavioral data ., This approach gives a coherent picture of this particularly well-defined non-cortical system ., Furthermore , Potjans and Diesmann 27 proposed a model of a sensory cortex network where the connection probabilities are thoroughly derived from in vitro studies ., While the neuron parameters are generic and homogeneous , their focus is on the precise laminar and horizontal organization of the synaptic connections ., They compare the results of their simulations with the baseline firing rates and flow of transient information through the different layers in vivo ., These comparisons to experimental data are qualitative in character , as it is the case for most existing large-scale models of cortical networks 12–14 , 67 ., However , a few recent studies also made statistical comparisons on partial aspects of physiological data 68 , 69 ., It would be interesting to assess these models on a wider range of in vivo data as we proposed here , to see which degree of biological detail is sufficient to predict their key properties ., An important simplification made in the present model is the reduction to two laminar components , leaving out layer 4 and 6 as well as the long-range fiber bundles and interneurons in layer 1 ., While layer 4 is missing in rodent PFC , layer 6 is only weakly connected to the other layers in our reference connectivity maps , which are based on the motor cortex 57 , 70 , 71 ., Thus , its inclusion in the network should not have a major impact on the results shown here ., This is probably different in sensory networks , where layer 6 strongly interacts with both pyramidal cells and interneurons in layer 4 72 ., The model exhibits a low fraction of spiking neurons , consistent with results from recording methods such as calcium imaging , which are not biased towards high firing rates ( “dark matter theory” of neuroscience 44–46 ) ., As described above , this may partly result from the variance-driven firing of the neurons: The membrane potential is on average well below the spiking threshold , but large fluctuations still lead to occasional spiking ., The size of the fluctuations and the low-rate , Poisson-like firing ( CV ≈ 1 ) of the neurons is consistent with the high-conductance state theory 48 and balanced-state theory 42 , 73 ., We note that the irregular and highly asynchronous firing of the neurons 74 observed here is a generic property of the network that simply emerged from its parametrization through in vitro and anatomical findings ., There are two main determinants of the high-fluctuation regime of the model: First , variability in the membrane potential requires variability in the synaptic parameters and in particular , the fat tail of the log-normal distribution of the synaptic weights ., Second , the range between the firing threshold Vup and the GABAA reversal potential E rev GABA must be sufficiently large , because below E rev GABA , all synaptic currents depolarize the cell , so the dynamical range for a balanced , variance-driven state is constrained between these two values ., Using the multivariate distributions of neuron parameters obtained from our in vitro recordings , we also observed that decreased cellular heterogeneity has a profound effect on the processing of transient stimuli ., It prevents the transmission of stimulus-induced activity from L2/3 to L5 ., This phenomenon can be understood if one considers the rheobase distribution: Reduced heterogeneity removes those neurons that originally had a very low or even negative rheobase ., These are the ones which are highly susceptible to even small inputs and form a small but significant fraction of L5 neurons that were activated by the transient synaptic input from the L2/3 cells ., Given that L5 provides the majority of output to other brain areas , impaired transfer of stimuli to this layer may lead to major impairments in information processing ., Thus , apparently quite subtle changes in the distributional properties of synaptic and cellular parameters ( not affecting their means ) may lead to major changes in network dynamics and functional connectivity among columns or areas , effects that have been proposed to underlie major psychiatric conditions like schizophrenia 8 , 9 ., By varying the total input from a virtual population designed according to the same principles as the actually simulated network , we provided evidence that a larger network than the one actually simulated with anatomically realistic neuron densities should be capable of self-sustaining in vivo-like spiking modes ., Although we did not demonstrate self-consistency in a strict sense , we have shown that the background currents into the smaller , simulated network needed to yield in vivo-like behavior are consistent with the range produced by a much larger network of anatomically reasonable size ., For currents within the blackly delimited region of Fig 6A , the spike train distributions are statistically indistingu | Introduction, Results, Discussion, Materials and Methods | The prefrontal cortex is centrally involved in a wide range of cognitive functions and their impairment in psychiatric disorders ., Yet , the computational principles that govern the dynamics of prefrontal neural networks , and link their physiological , biochemical and anatomical properties to cognitive functions , are not well understood ., Computational models can help to bridge the gap between these different levels of description , provided they are sufficiently constrained by experimental data and capable of predicting key properties of the intact cortex ., Here , we present a detailed network model of the prefrontal cortex , based on a simple computationally efficient single neuron model ( simpAdEx ) , with all parameters derived from in vitro electrophysiological and anatomical data ., Without additional tuning , this model could be shown to quantitatively reproduce a wide range of measures from in vivo electrophysiological recordings , to a degree where simulated and experimentally observed activities were statistically indistinguishable ., These measures include spike train statistics , membrane potential fluctuations , local field potentials , and the transmission of transient stimulus information across layers ., We further demonstrate that model predictions are robust against moderate changes in key parameters , and that synaptic heterogeneity is a crucial ingredient to the quantitative reproduction of in vivo-like electrophysiological behavior ., Thus , we have produced a physiologically highly valid , in a quantitative sense , yet computationally efficient PFC network model , which helped to identify key properties underlying spike time dynamics as observed in vivo , and can be harvested for in-depth investigation of the links between physiology and cognition . | Computational network models are an important tool for linking physiological and neuro-dynamical processes to cognition ., However , harvesting network models for this purpose may less depend on how much biophysical detail is included , but more on how well the model can capture the functional network physiology ., Here , we present the first network model of the prefrontal cortex which has not only its single neuron properties and anatomical layout tightly constrained by experimental data , but is also able to quantitatively reproduce a large range of spiking , field potential , and membrane voltage statistics obtained from in vivo data , without need of specific parameter tuning ., It thus represents a novel computational tool for addressing questions about the neuro-dynamics of cognition in health and disease . | action potentials, medicine and health sciences, neural networks, prefrontal cortex, membrane potential, brain, electrophysiology, neuroscience, ganglion cells, network analysis, interneurons, computer and information sciences, animal cells, cellular neuroscience, cell biology, pyramidal cells, anatomy, physiology, neurons, biology and life sciences, cellular types, neurophysiology | null |
journal.ppat.1005238 | 2,015 | A New Glycan-Dependent CD4-Binding Site Neutralizing Antibody Exerts Pressure on HIV-1 In Vivo | Although the envelope glycoproteins ( Env ) of primate immunodeficiency viruses have extremely variable sequences 1 , most of them engage CD4 as the primary cellular receptor to initiate the viral life cycle 2 ., The consequence is that the CD4 binding site ( CD4bs ) is a comparatively well-conserved region of Env that serves as a critical neutralization epitope and an appealing vaccine target ., The introduction of single cell antibody cloning techniques 3 , 4 yielded dozens of broad and potent CD4bs antibodies from infected individuals , some of which neutralize ~90% of HIV-1 strains in vitro 5–7 ., Some of these antibodies are also effective at reducing viral load when used to treat infected humanized mice ( hu-mice ) 8 , macaques 9–11 and humans 12 ., The most potent group of CD4bs antibodies characterized to date is derived from two VH genes , IGVH1-2 5 , 7 , 13 and IGVH1-46 6 , 7 , 14–16 ., These antibodies engage many of the same Env residues as CD4 ., For example , residue Arg71HC in VRC01-like bNAbs interacts with residue Asp368gp120 on Env , and thereby mimics how Arg59CD4 interacts with the same residue when CD4 binds to gp120 6 , 7 , 13 , 16 ., Although the light chains are less restricted in their origin , specific alterations are required for activity , including mutations and deletions 6 , 13 , 16 ., Overall , the restricted origins and complex development of these bNAbs from their inactive germline precursors may explain why it has been so difficult to elicit them by vaccination ., A second , far more diverse group of CD4bs-directed antibodies is often referred to as ‘CDRH3-dominated class of CD4bs antibodies’ ., These antibodies use their CDRH3-loop regions to engage Env 15 ., These include b12 17 , HJ16 18 , CH103 19 and the recently described VRC13 and VRC16 15 ., Structural analyses indicate that all CDRH3-dominated antibodies use loop-based recognition mechanisms , with the CDRH3 contributing 50%-70% of the paratope interface 15 , 19 , 20 ., They are not VH-restricted since their CDRH3s are randomly assembled from IgH variable , diversity and joining segments during V ( D ) J recombination 21 ., In keeping with their diverse origins , CDRH3-dominated antibodies seem to employ different mechanisms of recognition and they also vary in the angles with which they approach the CD4bs 15 ., To isolate new CD4bs bNAbs , we sought HIV-1 infected donors whose sera contained potent neutralizing antibodies that appeared to target the CD4bs ., One such donor was EB179 ., By sorting peripheral blood mononuclear cells ( PBMCs ) from this individual we isolated a new antibody , 179NC75 , that is encoded by IGVH3-21 and IGVL3-1 gene segments ., In TZM . bl neutralization assays 179NC75 showed an overall IC80 of 0 . 42 μg/ml against 120 Tier-2 HIV-1 ., Binding assays using various Env-based proteins indicated that 179NC75 is glycan-dependent and belongs to the same sub-class of CDRH3-dominated CD4bs antibodies as HJ16 ., These glycan-dependent CD4bs antibodies have not yet been tested for activity in vivo ., To do so we treated humanized mice infected with HIV-1YU2 with 179NC75 and found that it selects for escape variants with mutations in the potential N-linked glycosylation site at gp120 position 276 ., Similar mutations were also found in the autologous isolate from the 179NC75 donor , suggesting that selection pressure had been exerted in the human host ., For the human studies , The Rockefeller University Institutional Review Board approved all studies involving patient enrollment , sample collection , and clinical follow-up ., Donor EB179 was selected from a group of long-term non- progressors that was followed at the Ragon Institute in Boston , and is also referred to as subject 330183 ., The subject described in this study provided written informed consent prior to participating in this study ., For the mouse studies , this study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health ., The protocol was approved by the Institutional Animal Care and Use Committee ( IACUC ) of The Rockefeller University , and in accordance with established guidelines and policies at The Rockefeller University ( protocol number 13618-H ) ., Purified IgG samples from 394 HIV-1-infected long-term non-progressors were screened for neutralizing activity against a panel of 14 viruses representing 8 different clades or inter-clade recombinants ., IgG purified from donor EB179 was exceptional in its neutralization potency and breadth , ranking within the top 2% of the cohort , and neutralized 11 out of the 14 viruses in the panel ( S1A Table ) ., A single leukapheresis sample was obtained 4 . 5 years after initial diagnosis with clade B HIV-1 infection , at age 44 ., At the time the sample was collected , donor EB179 had 1038 CD4 T cells/mm3 and a viral load of 3180 copies/ml and was not receiving antiretroviral therapy ., Molecular HLA typing revealed HLA A*02:01 , 68:02; B*07:02 , 53:01; Cw04:01 , 07:02; DRB11:01 , 15:01 ., Single-cell sorting of 2CC core+CD19+IgG+ B cells from donor EB179’s PBMCs was conducted as previously described 3 ., Briefly , we sorted memory B cells using the gp120 2CC core protein as bait 22 ., Rescue primers were used to amplify both heavy chains 7 and Igλ genes 23 ., All PCR products were sequenced and analyzed for Ig gene usage , CDR3 , and the number of VH/VL somatic hypermutations ( IgBLAST , http://www . ncbi . nlm . nih . gov/igblast/ and IMGT , http://www . imgt . org/ ) ., Multiple sequence alignments were performed using the MacVector program ( v . 13 . 5 . 5 ) with the ClustalW analysis function ( default parameters ) , and then used to generate dendrograms by the neighbor-joining method ( with best tree mode and outgroup rooting ) ., To specifically isolate members of the 179NC75 clone we used the following forward primers for the heavy chains: 5’-CTGCAACCGGTGTACATTCTGAAATGAGATTGGAAGAAT-3’ and 5’-CTGCAACCGGTGTACATTCTGAGGTCCAGTGTGAAGAA-3’ ( in a 1:1 mix ) ; and for the light chains: 5’-ATGGCCTGGATCCCTCTACTTCTC-3’ and 5’- ATGGCATGGATCCCTCTCTTCCTC-3’ ( in a 1:1 mix ) ., The reverse primers were the same as described previously for Ig gene amplification 7 ., Purified , digested PCR products were cloned into human Igγ1- , IgK or Igλ-expression vectors as previously described 24 ., Antibodies were produced by transient transfection of IgH , IgK and IgL expression plasmids into exponentially growing HEK 293T cells ( ATCC; CRL-11268 ) using polyethyleneimine ( PEI ) -precipitation 24 ., IgG antibodies were affinity purified using Protein G Sepharose beads according to the manufacturer’s instructions ( GE Healthcare ) ., High-binding 96-well ELISA plates ( Costar ) were coated overnight with 5 μg/ml of purified 2CC core , gp120 . YU2 ( wild type or mutants ) or gp140 . YU2 foldon trimer in PBS ., After washing 6 times with PBS + 0 . 05% Tween 20 , the plates were blocked for 2 h with 2% BSA , 1 μM EDTA and 0 . 05% Tween-PBS ( “blocking buffer” ) , and then incubated for 1 h with IgGs that were added as seven consecutive 1:4 dilutions in PBS from an initial concentration of 4 μg/ml ., After additional washing , the plates were developed by incubation with goat HRP-conjugated anti-human IgG antibodies ( Jackson ImmunoResearch ) ( at 0 . 8 μg/ml in blocking buffer ) for 1 h followed by HRP chromogenic substrate ( ABTS solution; Invitrogen ) ., For competition ELISAs , the plates were coated with 5 μg/ml 2CC core , gp120 or gp140 foldon , washed , blocked for 2 h with blocking buffer and then incubated for 1 h with IgGs added as seven consecutive 1:4 dilutions in PBS from an initial concentration of 32 μg/ml , and in the presence of biotinylated 179NC75 antibody at a constant concentration of 4 μg/ml ., The plates were then developed using HRP-congugated streptavidin ( Jackson ImmunoReseach ) ( at 1 μg/ml in blocking buffer ) ., For ELISAs using BG505 SOSIP . 664-D7324 trimers , the plates were coated overnight with 5 μg/ml of D7324 antibody as previously described 25 , washed and then incubated with 500 ng/ml of the trimer 25 , 26 ., After a further wash , IgGs were added for 1 h as seven consecutive 1:4 dilutions in PBS from initial concentrations of 4 μg/ml ., The endpoint was generated by incubation with goat HRP-conjugated anti-human IgG antibodies , as described above ., All experiments were performed at least 3 times ., For EndoH ELISA , the plates were coated overnight at 4°C with 5 μg/ml of EndoH-treated or untreated gp120 in 100 mM sodium bicarbonate/carbonate buffer , pH 9 . 6 ., They were then washed with TBS + 0 . 05% Tween 20 and blocked for 1 h in the same buffer supplemented with 3% ( w/v ) BSA , and washed again before test antibodies were added for 2 h ., After a final wash , the endpoint was generated using goat HRP-conjugated anti-human IgG antibodies , again as described above ., HIV-1 neutralization was evaluated using the luciferase-based TZM . bl cell assay as described previously 27 ., Briefly , envelope pseudoviruses were incubated with fivefold serial dilutions of single antibodies and applied to TZM . bl cells that carry a luciferase-reporter gene ., After 48 h cells were lysed and luminescence was measured ., IC50 and IC80 reflect single antibody concentrations that caused a reduction in relative luminescence units ( RLU ) by 50% and 80% , respectively ., NOD Rag1−/−Il2rgnull ( NOD . Cg-Rag1tm1Mom Il2rgtm1Wjl/SzJ ) mice were purchased from The Jackson Laboratory and bred and maintained at the Comparative Bioscience Center of The Rockefeller University according to guidelines established by the University’s Institutional Animal Care and Use Committee ., All experiments were performed under protocols approved by the same committee ., Hu-mice were treated with 1 mg of 179NC75 sub-cutaneously ( s . c . ) on day 0 , followed by 0 . 5 mg s . c . injections twice-weekly for a period of 5 weeks 8 ., The gp120 sequences from escape variant viruses were obtained as previously described 8 ., The autologous virus from donor EB179 was isolated as previously described 28 ., Briefly , CD19 and CD8-depleted mononuclear cells were cultured at a concentration of 5 × 106 cells/ml in Iscoves modified Dulbecco’s medium ( IMDM; Gibco ) supplemented with 10% fetal bovine serum ( FBS; HyClone , Thermo Scientific ) , 1% GlutaMAX ( Gibco ) , 1% penicillin/streptomycin ( Gibco ) , and 1 μg/ml phytohaemagglutinin ( Life Technologies ) at 37°C and in an atmosphere containing 5% CO2 ., After 2–3 days , 5 × 106 cells were transferred into IMDM supplemented with 10% FBS , 1% penicillin/streptomycin , 5 μg/ml polybrene ( Sigma ) , and 100 IU/ml of IL-2 ., The medium was replaced weekly and the HIV-1 content of culture supernatants was quantified using the Lenti-X p24 Rapid Titer Kit ( Clontech ) according to the manufacturer’s instructions ., Env genes from the autologous virus were cloned by reverse transcriptase PCR as described elsewhere 29 ., Single , double and triple mutations were introduced into wild-type HIV-1YU2 envelope using the QuikChange ( multi- ) site-directed mutagenesis kit , according to the manufacturer’s specifications ( Agilent Technologies ) ., Polyclonal IgG purified from donor EB179 had exceptional neutralization capacity , with respect of potency and activity against 11 of 14 Tier-2 viruses in a small cross-clade panel ( S1A Table ) ., To map the predominant NAb specificities , we tested EB179 IgG against HIV-1YU2 mutants that are resistant to NAbs targeting the trimer apex ( N160K ) , the CD4bs ( N280Y ) or the base of the V3 loop ( N332K ) 8 , 30–32 ., Among these mutants , only HIVYU2N280Y was resistant to EB179 IgG ( S1B Table ) ., We conclude that at least a proportion of the neutralization activity present in this serum is directed to the CD4bs ., To isolate and characterize the NAbs present in EB179 , we used flow cytometry to sort memory B cells that bound to 2CC core , a gp120 antigen that presents the CD4bs in an exposed and stable conformation 22 ., Among CD19+IgG+ B cells , ~0 . 2% bound strongly to 2CC core ., Of the 372 cells sorted , 87 produced paired heavy and light chains , 36 of which represented ten clonally related families ( Fig 1A ) ., Antibody sequences obtained from the expanded B cell clones contained higher numbers of somatic mutations compared to antibodies obtained from B cells that appeared only once ( S1 Fig ) ., The average number of nucleotide mutations in the heavy chain of clonal sequences was 44 . 76 ( ± 3 . 66 , N = 36 ) compared to 20 . 82 ( ± 1 . 39 N = 51 ) for unique sequences ( S1A Fig ) ., A similar trend was observed when the light chain sequences were analyzed ( S1B Fig ) ., Representative variants from each of the clonal families were selected for further analysis ( S2 Table ) ., These variants were expressed as IgG1 antibodies that were tested for binding to a HIV-1YU2 gp140 foldon protein 33 or 2CC core 22 , and for neutralizing activity ., Except for 179NC9055 , all the antibodies bound strongly to the HIV-1YU2 gp140 and/or 2CC core proteins ( Fig 1B ) , and members of clones 1 , 2 , 3 , 4 , 6 , and 7 neutralized the Tier-1 ( i . e . , neutralization-sensitive ) HIV-1BAL virus ( Fig 1C ) ., While antibodies from clones 3 and 7 were only weakly active against the other viruses in the panel , one representative of the most expanded clone 1 ( 179NC75 ) strongly neutralized four of the five viruses tested ( IC50 ≤0 . 05 μg/ml , Fig 1C ) ., To isolate additional 179NC75 variants , we amplified cDNA from the 2CC core+CD19+IgG+ single-sorted B cells using specific VH and VL forward primers ( see Methods ) ., We obtained a total of 23 heavy chain and 25 light chain variants from the 179NC75 clonal family ., The heavy and light chain sequences carried 34% and 29% amino acid mutations on average , respectively , compared to their germline gene segments IGVH3-21 and IGVL3-1 ., The various sequences of the 179NC75 clone were similar by up to 73% from clonal members ( Fig 2A and 2B ) ., The CDRH3 and CDRL3 regions were 24 and 10 residues long , respectively ( Fig 2A and 2B , S2 Table ) ., There were no insertions or deletions ., Variants 179NC 54 , 60 , 65 , 75 , 21 and 1055 ( indicated in Fig 2A and 2B ) were tested for activity against a panel of nine Tier-2 viruses , including three from clade B , one from clade C , two from clade A , two clade A/G recombinants and one clade A/E recombinant ., 179NC75 and two closely related variants , 179NC54 and 179NC60 , potently neutralized 6 of these 9 viruses , whereas the other antibodies had lesser or no neutralization activity ( Fig 2C ) ., Accordingly , we selected 179NC75 for additional analyses ., When tested against an extended cross-clade panel of 120 Tier-2 viruses , 179NC75 neutralized viruses from clades B particularly strongly ( S3 Table ) ; its geometric mean IC50 and IC80 values were 0 . 113 μg/ml and 0 . 291 μg/ml , respectively ( S4 Table ) ., When compared to other CD4bs bNAbs against a panel of 22 Tier-2 clade , B viruses , 179NC75 was more potent than b12 against 13 viruses , than HJ16 against 15 viruses , than VRC01 against 8 viruses , and than CH103 against 6 viruses ( Fig 2D ) ., Its overall breadth of activity across the clade B virus panel was 70% ( S4 Table ) ., To map the epitope targeted by 179NC75 and its clonal variants , we performed a series of ELISAs ., All members of the 179NC75 clonal family bound to HIV-1YU2 gp120 , gp140 foldon 34 and 2CC core 22 proteins ( S2 Fig ) ., In a competition ELISA , soluble CD4 ( sCD4 ) and most CD4bs antibodies competed with 179NC75 for binding to gp120YU2 , whereas PGT121 , PGT128 and 10–1074 did not ( Fig 3A upper and lower panels ) ., The 8ANC195 bNAb , which binds an epitope adjacent to the CD4bs 7 , 35 , inhibited 179NC75 binding by ~ 50% ( Fig 3A , lower panel ) ., We conclude that the 179NC75 epitope is proximal to the CD4bs ., We next tested how different mutations in the CD4bs affected 179NC75 binding ., The D368R single mutation was not sufficient to affect the gp120-binding of 179NC75 family members , but the D368R and N280Y double mutation substantially impaired their binding ., In contrast , VRC01 is sensitive to the single D368R substitution ( Fig 3B , right upper and lower panels ) ., The Asn276 glycan site is important for the binding of two different bNAbs: the CD4bs antibody HJ16 36 and the gp120-gp41 specific antibody 8ANC195 7 , 35 ., The 8ANC195 epitope lies outside the CD4bs and this antibody binds Env in the presence of CD4 7 , 35 ., Since HJ16 strongly inhibited 179NC75 binding ( Fig 3A , upper panel ) and 8ANC195 did so weakly ( Fig 3A , lower panel ) , we assessed whether the binding of 179NC75 family members was affected by the N276D substitution and found that it had a profound impact ( Fig 3B , lower left panel ) ., In contrast , the N276D change had no effect on VRC01 binding , as previously reported 37 ( Fig 3B ) ., When monomeric gp120s from both YU2 and the clade A/E virus 93TH057 38 were treated with EndoH , a glycosidase that removes N-linked oligomannose glycans , the binding of 179NC75 and its clonal variants was completely abolished ( Fig 3C ) ., To further probe the nature of the glycan-dependency of 179NC75 , we tested binding of the Fab to BG505 SOSIP . 664 trimers , ( fully glycosylated , cleaved , native-like , soluble trimers 25 ) produced in HEK293-6E cells in the presence and in the absence of the mannosidase I inhibitor kifunensine ., HEK293-6E cells fully process glycans resulting in a mixture of complex-type and high-mannose N-glycans , while HEK293-6E cells treated with kifunensine , produce protein containing only high-mannose N-glycans ., We observed that 179NC75 binds to BG505 SOSIP . 664 trimer with processed glycans with a KD of ~90 nM , ( S3 Fig ) but cannot bind to trimers containing only high mannose glycans ( S3 Fig , S6 Table ) ., Hence , we conclude that 179NC75 is a glycan-dependent antibody that binds to the CD4bs in a way that involves the Asn276 residue and depends on the presence of complex glycans ., In these respects , its epitope is similar to that of the HJ16 CD4bs bNAb ., We compared the neutralization potencies of 179NC75 to the ones of HJ16 18 ., For the 53 Tier-2 viruses that were tested against both HJ16 and 179NC75 , 179NC75 neutralized more viruses than HJ16 ( 26 compared to 19 ) , and was 20-fold more potent ( IC50 of 0 . 118 μg/ml compared to 2 . 326 μg/ml ) ( Fig 3D ) ., Previous reports show that neutralizing antibodies bind BG505 SOSIP . 664 trimers with higher affinity as opposed to non-neutralizing antibodies 25 ., Therefore , as expected , the more potent variants of the 179NC75 clone , 179NC75 and 179NC1055 , bound strongly to BG505 SOSIP . 664-D7324 trimers in capture ELISA , while 179NC65 and 179NC21 bound weakly or not at all , respectively ( S4 Fig ) ., Most predicted germline versions of CD4bs antibodies are unable to bind Env antigens 7 ., To test whether the germline version of 179NC75 could bind the BG505 SOSIP . 664-D7324 trimers , and assess the role of CDRH3 in trimer binding , we generated a germline version of 179NC75 ( 179NC75gl ) ., The predicted germline version of the antibody was made as previously described by reverting the V and J segments of the heavy and light chains to their predicted germline sequences , while retaining the CDRH3 sequence as found in the mutated antibody 7 , 39 , 40 ., For comparison , we used the previously published predicted germline versions of VRC01 39 , 40 , 3BNC60 7 , 1NC9 7 , CH103 19 and HJ16 ( constructed in the course of this study ) ., Although all of the above mature CD4bs bNAbs bound the BG505 SOSIP . 664-D7324 trimers , the only predicted germline antibody able to do so was 179NC75gl ( Fig 4 ) ., An implication is that 179NC75 binding principally involves contacts made by the CDR3s , particularly the exceptionally long ( 24-residue ) heavy chain CDR3 ., The loop binding , glycan-dependent CD4bs bNAbs have not been tested for their activity in vivo ., To address this issue , we treated six HIV-1YU2–infected hu-mice with 179NC75 for 5 weeks 8 , 29 ., Monotherapy with 179NC75 resembled monotherapy with other bNAbs , in that there was a transient decrease in viral load in most of the treated animals followed by a rapid rebound 8 , 29 , 41 ( Fig 5A and 5B ) ., Viral env genes were cloned and sequenced from the day-28 plasma of 179NC75-treated mice , a time point where viremia had universally rebounded to levels similar to the day-0 value ., Two types of mutations were consistently observed , both proximal to the CD4bs: the first eliminated the glycan-site at position N276; the second involved residues G459 or K460 ( Fig 5C ) ., In total , 13 sequences had only a mutation affecting the N276 glycan site , whereas 8 contained mutations in the region near position 460 and 7 sequences contained mutations in both regions ( Fig 5C and 5D ) ., In all mice the rebounding viruses carried at least one of these mutations ., In mouse 1107 mutations in both areas were observed , resulting in the loss of the N276 glycan but the introduction of a potential N-linked glycosylation site ( PNGS ) at position 460 ( Fig 5C ) ., To confirm that the most commonly observed mutations did confer resistance , HIV-1YU2 Env-pseudoviruses containing one or both of the N276D and K460N substitutions were tested for their sensitivity to 179NC75 ., All three of the virus mutants were found to be 179NC75-resistant ( Fig 5E ) ., We also tested the HIVYU2 N280Y , N332K , N160K and G459D virus mutants ., As expected , and consistent with the ELISA data , the N280Y substitution conferred complete resistance to 179NC75 , while the N332K and N160K changes had no effect ., The G459D mutant was also 179NC75-sensitive ( Fig 5E ) ., We conclude , that 179NC75 is a potent neutralizing antibody that exerts selection pressure on HIV-1YU2 in vivo and drives the emergence of resistant viruses with sequence changes proximal to the CD4bs ., To test whether 179NC75 exerted selective pressure on the autologous virus found in subject EB179 , we cloned env genes from the donor’s T cells obtained at the time of the leukapheresis ., All nine gp120 sequences obtained contained Asn at position 460 , introducing a PNGS at that position in eight of the nine sequences ( Fig 6A–6C ) ., Five sequences contained an Asn-Gly-Thr insertion immediately N-terminal to position N460 , resulting in the sequence NGTNET , and therefore adding another PNGS to the one that was already at position 460 ., Five other sequences contained the N276S mutation , eliminating the PNGS at position 276 ., One of the nine sequences included both the Asn-Gly-Thr insertion at position 460 and the N276S change ( Fig 6C ) ., Of note is that this pattern of sequence changes is highly similar to the escape mutations seen in the env genes of the 179NC75-treated , HIV-1YU2-infected hu-mice ( Fig 5 ) ., To test whether the autologous virus from patient EB179 is resistant to 179NC75 , we cultured the donor’s CD4+ T cells from the same leukapheresis sample that was used for the antibody isolation ., Outgrown virus was then tested for neutralization in the TZM . bl assay for neutralization by the EB179 polyclonal IgG ( from the same time point ) , as well as by 179NC75 and other known bNAbs including the CD4bs antibody 3BNC117 7 , the V3-stem binding antibody 10–1074 42 and the V1/V2 apex-binding antibody PG16 43 ( Fig 6D ) ., As expected , the EB179 polyclonal IgG failed to neutralize the autologous virus ., Amongst the two CD4bs antibodies , the autologous virus was fourfold more resistant to 179NC75 than to 3BNC117 , suggesting that the EB179 antibody repertoire has CD4bs antibodies that differ from 3BNC117 and VRC01-class bNAbs ., Interestingly , the autologous virus was also resistant to PG16 and 10–1074 , indicating that the patient may have additional neutralizing antibodies bearing similar specificities in his antibody repertoire ., Taken together , the data imply that loop-based , glycan-dependent CD4bs bNAbs of the 179NC75 family exert selective pressure on HIV-1 in vivo ., The CD4bs is a highly conserved epitope on the HIV-1 Env and an important potential target for neutralizing antibodies ., Although this site evolved to avoid antibody accessibility , two major groups of CD4bs bNAbs have been discovered 15 ., The first group , exemplified by VRC01 , is VH-restricted , IGVH1-2 or IGVH1-46 , with the heavy chains positioned in a CD4-like orientation and CDRH2 making significant contacts with gp120 6 , 7 , 15 ., The CDRL3 7 , 21 , 44 , and in some cases also CDRL1 6 , of the corresponding light chains have to be short and compact to minimize potential interference and clashes with the glycans that surround the CD4bs ., The emergence of these antibodies involves many somatic hypermutations , some of which are in the framework regions 45 ., The second group of CD4bs bNAbs , which includes b12 and HJ16 , is far more heterogeneous ., These antibodies bind to gp120 via a CDRH3-dominated , loop based mechanism 15 ., As might be expected , members of this group of CD4bs bNAbs arise from different VH segments and carry fewer somatic mutations 17–19 ., The new antibody described in this study , 179NC75 , is a loop binder that is closely related to HJ16 ., Similarly to HJ16 , its Env-binding and virus-neutralizing activities are dependent on the N276 glycan 36 ., Consistent with the CDRH3 loop-based mechanism of recognition that was described for antibodies that are not VH-restricted 15 , when we generated the predicted germline version of 179NC75 , where all mutations were reverted but the CDRH3 was retained , the antibody bound to BG505 SOSIP . 664 trimers ., This could indicate that any residual mutations present in the CDR3s of the reverted antibody might allow binding ., Interestingly , the germline version of HJ16 also had some binding to BG505 SOSIP . 664 trimers ( Fig 4 ) , however this binding was lower that the one of 179NC75 , which could be attributed to a shorter CDRH3 ( 19 versus 24 residues ) ., Serum antibodies that are CD4bs-specific and N276-dependent have been described in HIV-1-infected individuals in two separate studies 32 , 46 ., In the first study , an HJ16-type of CD4bs antibody response was found to be part of the second wave of serum neutralization in the CAP257 patient 46 ., Viruses cloned from CAP257 after the emergence of these CD4bs antibodies carried an N276D or T278A mutation that were considered to be responses to antibody selection pressure 46 ., In a second study , serum from individual VC1004 contained CD4bs-targeted NAbs that were sensitive to the N276D substitution but not D368R 47 ., However , as the antibodies responsible for the serum activity were not cloned in either study much of what we know about the in vivo activity of these N276-dependent class of CD4bs antibodies is inferential ., Our 179NC75 therapy experiments in HIV-1–infected hu-mice demonstrate that escape variants contain very similar , and sometimes identical , mutations to ones present in the autologous virus isolated from the infected human from whom the 179NC75 antibody was also derived ., We conclude that the CDRH3-dominted N276-dependent CD4bs antibodies are effective at suppressing viremia in vivo and thence driving the emergence of escape variants . | Introduction, Materials and Methods, Results, Discussion | The CD4 binding site ( CD4bs ) on the envelope glycoprotein is a major site of vulnerability that is conserved among different HIV-1 isolates ., Many broadly neutralizing antibodies ( bNAbs ) to the CD4bs belong to the VRC01 class , sharing highly restricted origins , recognition mechanisms and viral escape pathways ., We sought to isolate new anti-CD4bs bNAbs with different origins and mechanisms of action ., Using a gp120 2CC core as bait , we isolated antibodies encoded by IGVH3-21 and IGVL3-1 genes with long CDRH3s that depend on the presence of the N-linked glycan at position-276 for activity ., This binding mode is similar to the previously identified antibody HJ16 , however the new antibodies identified herein are more potent and broad ., The most potent variant , 179NC75 , had a geometric mean IC80 value of 0 . 42 μg/ml against 120 Tier-2 HIV-1 pseudoviruses in the TZM . bl assay ., Although this group of CD4bs glycan-dependent antibodies can be broadly and potently neutralizing in vitro , their in vivo activity has not been tested to date ., Here , we report that 179NC75 is highly active when administered to HIV-1-infected humanized mice , where it selects for escape variants that lack a glycan site at position-276 ., The same glycan was absent from the virus isolated from the 179NC75 donor , implying that the antibody also exerts selection pressure in humans . | CD4bs is a central viral vulnerability site and isolation of new anti-HIV-1 CD4bs broadly neutralizing antibodies ( bNAbs ) provides information about viral escape mechanisms ., Here we describe a new anti-HIV-1 bNAb that was isolated from an HIV-1 infected donor ., The antibody , 179NC75 , targets the CD4 binding site in a glycan-dependent manner ., Although many CD4bs antibodies have been already described , a glycan-dependent mode of recognition is unusual for anti-HIV-1 CD4bs bNAbs ., The glycan-dependent CD4bs antibodies have never been tested for their ability to neutralize HIV-1 in vivo ., We infected humanized mice with HIV-1YU2 and treated them with 179NC75 three weeks after infection ., We observed a drop in viral load immediately after treatment followed by a viral rebound ., The viral rebound was associated with specific escape mutations in the plasma virus envelope , resulting in a deletion of N276 glycan , and in some cases a glycan shift from position 276 to position 460 ., Similar signature mutations were found in the envelope of the autologous virus cloned from patient’s plasma ., This defines the escape pathways from 179NC75 , and shows that they are the same in humans and in HIV-1YU2 infected humanized mice . | null | null |
journal.pntd.0003681 | 2,015 | Whole Genome Comparisons Suggest Random Distribution of Mycobacterium ulcerans Genotypes in a Buruli Ulcer Endemic Region of Ghana | Buruli ulcer ( BU ) is a neglected tropical disease caused by infection with Mycobacterium ulcerans ., Each year 5000–6000 cases are reported from 15 of the 33 countries where BU cases have been reported , predominantly from rural regions across West and Central Africa 1 ., The disease involves subcutaneous tissue and has several manifestations but necrotic skin ulcers are a common presentation , caused by the proliferation of bacteria beneath the dermis by virtue of a secreted bioactive lipid called mycolactone 2 ., The role of mycolactone in the natural ecology of M . ulcerans is not understood , but it has been shown to possess several specific activities against mammalian cells from activating actin polymerization , blocking secreted protein translocation , to interacting with neuronal angiotensin type II receptors causing hypoesthesia 3–5 ., These collective biological activities of mycolactone , while diverse , might collectively help explain the tissue destruction , lack of inflammation , and painlessness associated with BU ., BU is rarely fatal and early diagnosis followed by combined antibiotic therapy ( rifampicin and streptomycin ) is key to preventing complications that can arise from severe skin ulceration 6 ., Epidemiological studies frequently link BU occurrence with low-lying and wetland areas and human-to-human transmission seems rare , suggesting an environmental source of the mycobacterium 7–23 ., Frustratingly however , the environmental reservoir ( s ) and mode ( s ) of transmission of M . ulcerans remain unknown ., M . ulcerans has the genomic signature of a niche-adapted mycobacterium , indicating that it is unlikely to be found free-living in diverse aquatic ( or other ) environments , but more likely in close association with a host organism ., In south eastern Australia , native marsupials have been identified as both susceptible hosts and reservoirs of M . ulcerans , with high numbers of the bacteria shed in the feces of infected animals ., Mosquitoes have also been found to harbor the bacteria in this region and a zoonotic model of disease transmission has been proposed involving possums , biting insects and humans 24–26 ., No such animal reservoir has yet been identified in African BU endemic areas and studies of BU lesion distribution are thought not consistent with mosquito biting patterns 22 , 27 ., On the other hand , case-control studies in Cameroon have shown that bed nets are protective , supporting a role for insects in transmission 28 ., A feature of M . ulcerans is the close correlation between genotype and the geographic origin of a strain , but its restricted genetic diversity has limited the application of traditional molecular epidemiological methods such as VNTR-typing to discriminate between isolates at the village or even regional scales ., The advent of low cost genomics has opened up new possibilities to explore and track the movement and spread of this pathogen within communities 29 , 30 ., Agogo is the principal town of 30 , 000 inhabitants in the Asante Akim North ( AAN ) district within the Ashanti region of Ghana and BU has been reported in about half of the sixty-four communities in this district since mid-1975 10 ., The AAN district covers an area of 650 km2 in the forest belt of Ghana and it is the third most endemic district in Ghana 31 ., Five of the communities ( Ananekrom , Serebouso , Nshyieso , Serebuoso and Dukusen ) in this district are among the communities reported with the highest burden of the disease in Ghana 31 ., About 120 laboratory-confirmed new cases are reported annually in this district 31 ., Subsistence farming and petty trading are the principal occupations of inhabitants of these endemic communities ., People generally live in simple dwellings constructed from local materials ., Houses are often close together with 3–5 households in a compound ., Many inhabitants raise animals such as goats , sheep , and pigs in the immediate vicinity of their houses ., Farming is the main occupation with some people engaged in fishing and petty trading ., Farms may be distant , ranging 5–20 km from a given domicile ., Fishing is usually undertaken close to home ., Water sources are of two types ., Water for drinking and cooking is usually fetched from bore holes fitted with mechanical pumps , within or near a village ., Water for bathing and domestic chores such as washing of clothes is drawn from local natural water sources ( rivers , streams , ponds ) ., These natural sources are usually no more than 500 metres from a given village ., In this study we sequenced and compared the genomes of 18 M . ulcerans isolates obtained from 10 BU endemic villages in the AAN district and uncovered genetic evidence supporting the introduction of a foreign clone of M . ulcerans to this region ., This observation indicates that M . ulcerans can be mobilized and spread throughout a region , indicating that reservoirs of the bacterium are themselves potentially highly mobile ., M . ulcerans isolates were obtained from BU diagnostic samples , collected as part of routine laboratory diagnosis ., Ethical approval to interview patients and use bacterial isolates resulting from diagnostic specimens for research was obtained from the ethical review board of the Noguchi Memorial Institute for Medical Research , University of Ghana , Legon , Accra , Ghana ( FWA 00001824 ) , with written informed consent obtained from all adult patients or the parents/guardians of the participating children ., The study was carried out in ten endemic villages including Ananekrom , Nshyieso , Serebouso , Dukusen , Afreserie , Afreserie OK , Baama , Nysonyameye , Kwame Addo and Bebuso , in the Asante Akim North ( AAN ) district of Ghana ( Table 1 ) ., These are small villages and hamlets , 5 to 10 km from each other with populations between 120–1500 inhabitants ., Ananekrom is the largest of these communities and is the closest ( 15 km ) to the district capital , Agogo ., An asphalt road connects Agogo to Ananekrom , Dukusen and Afriserie , while the other communities are located off this main road and are connected to each other by unmade roads and foot-tracks ., A community health centre Ananefromh ( near Ananekrom ) is usually the first point of call for patients seeking medical treatment ., Patients suspected of having BU are referred to the Agogo Presbyterian Hospital for diagnosis and treatment ., Patient information including name and place of residence were obtained from hospital records and patients were visited in their homes for more detailed interviews that included questions about possible travel to other BU endemic areas outside the AAN district ., GPS coordinates in the vicinity of each patient’s residence were recorded in order to map the spatial distribution of cases in the villages , based on the assumption that the patient acquired their infection near their domicile ., The isolates examined in this study are listed in Table 1 and were recovered from fine needle aspirates ( FNA ) or swabs , obtained from pre-ulcerative lesions and ulcers respectively ., Specimens were stored in transport medium and PBS and transported in cool boxes to the Noguchi Memorial Institute for Medical Research ( NMIMR ) for diagnosis 32 , 33 ., Tubes containing swabs were vortexed in 3 ml of transport medium for 30 sec and the swabs removed ., A volume of 250μl of the transport medium from either specimen type was transferred into 1 . 5 ml microfuge tubes and decontaminated using the oxalic acid method as previously described 34 ., The pellets were resuspended in 100 μl phosphate buffered-saline ( PBS ) and 100 μl volume of the decontaminated sample was inoculated onto Löwenstein Jensen ( LJ ) slopes and incubated at 33°C ., The cultures were observed weekly for growth ., Suspected M . ulcerans colonies were harvested and DNA extracted as described above 35 ., The DNA extract was tested with the IS2404 PCR for the identification of M . ulcerans 36 ., Colonies positive for IS2404 were suspended in 1 ml of Middlebrook 7H9 broth and stored at -80°C ., All 18 bacterial samples analyzed were selected from this stored collection and were subcultured on LJ medium and DNA for whole genome sequencing was extracted from resulting growth as described 35 ., The isolation date refers to the date when colonies became visible on LJ medium following primary cultivation ., DNA sequencing was performed using two methods ., The Ion Torrent Personal Genome Machine was employed , with a 316 chip and 200bp single-end sequencing chemistry ( Life Technologies ) ., Genomic libraries for Ion Torrent sequencing were prepared using Ion Express , with size selection using the Pippin Prep ( Sage Sciences ) and emulsion PCR run using a One-Touch instrument ( Life Technologies ) ., The Illumina MiSeq was also used , with Nextera XP library preparation and 2x250 bp sequencing chemistry ., Read data for the study isolates have been deposited in the European Nucleotide Archive ( ENA ) under accession ERA401876 ., Prior to further analysis , reads were filtered to remove those containing ambiguous base calls , any reads less than 50 nucleotides in length , and containing only homopolymers ., All reads were furthermore trimmed removing residual ligated Nextera adaptors and low quality bases ( less than Q10 ) at the 3 end ., Resulting sequence Fastq sequence read files from either platform were subjected to read-mapping to the M . ulcerans Agy99 reference genome ( Genbank accession number CP000325 ) using Bowtie2 v2 . 1 . 0 37 with default parameters and consensus calling to identify SNPs ( indels excluded ) using Nesoni v0 . 109 , a Python utility that uses the reads from each genome aligned to the core genome to construct a tally of putative differences at each nucleotide position ( including substitutions , insertions , and deletions ) ( www . bioinformatics . net . au ) ., Those positions in the Agy99 reference genome that were covered by at least 3 reads from every isolate defined a core genome ., Note that the pMUM001 plasmid ( required for mycolactone synthesis ) was not included in the reference genome 38 ., Testing of the plasmid sequences revealed less then 10 polymorphic sites among the genomes under investigation and the highly repetitive sequence structure of the mycolactone genes impaired unambiguous read-mapping ., An unpaired t test with Welch’s correction was used to assess the differences between mean nucleotide pairwise identities for different groups of genomes ., The null hypothesis ( no difference between means ) was rejected for p<0 . 01 ., The inputs for subsequent phylogenomic analyses were the nucleotide sequence alignments of the concatenated variable nucleotide positions for the core genome among all isolates ., A maximum-likelihood ( ML ) phylogeny was inferred using RAxML v 7 . 2 . 8 , with the GTR model of nucleotide substitution ( plugin within Geneious v 8 . 0 . 4 ) ., We performed 1000 rapid pseudo-replicate bootstrap analyses to assess support for the ML phylogeny ., We used Consensus-Tree-Builder ( Geneious v8 . 0 . 4 ) to collapse nodes in the tree with bootstrap values below a set threshold of 70% ., The resulting phylogenomic tree was exported in Newick format and visualized using FigTree v1 . 4 . 0 ( tree . bio . ed . ac . uk/software/figtree ) ., A haplotype network was derived using the median-joining algorithm as implemented in SplitsTree v4 . 13 . 1 39 , 40 ., A correction to the source attribution of the M . ulcerans Agy99 reference genome was also made in the course of this study , where it was realized that this isolate was actually obtained from a patient attending St Martin’s Hospital in Agroyesum ( Amansie West ) and not the Ga District Hospital as originally published 41 ( K . Asiedu and J . , Hayman pers comms ) , thereby explaining the inconsistent geographic clustering reported in previous molecular epidemiological studies 30 , 42 , 43 ., Eighteen M . ulcerans isolates were randomly selected for whole genome sequencing ., The isolates represented 20% ( total of 92 isolates from 2010–2012 ) of all culture-confirmed BU cases referred to the Agogo Presbyterian Hospital between 2010 and 2012 ( Table 1 ) ., There were no differences in colony phenotype or growth characteristics among the isolates ., The DNA sequence reads for each genome were mapped to the M . ulcerans Agy99 reference sequence ., Sequencing and read-mapping summary statistics are given in Table, 1 . In addition to the 18 Agogo isolates sequenced here , 15 other genomes ( including some previously described ) were included in comparisons making a total of 33 isolates ( Table 1 ) ., These additional genomes were from M . ulcerans isolates in other regions of Ghana and from surrounding countries to provide appropriate genetic context for interpreting the diversity and evolution of M . ulcerans isolates from around Agogo ., Read-mapping and SNP identification revealed 320 variable nucleotide positions across a 5 . 2Mb core genome for the 33 isolates ., A phylogeny was inferred from this alignment , showing the clustering typical of M . ulcerans genotypes with geographic origin ( Fig . 1 ) ., A separate SNP alignment was performed taking the genome sequences for only the 18 isolates from the Agogo region , and 10 of them ( called Agogo-1 ) clustered with isolates from the neighboring district of Amansie West and also the Ivory Coast , the country which borders this region to the west ( Fig . 1 ) ., This close relationship is indicative of a local clone that has spread and persisted within the greater region for some time ., Unexpectedly however , this analysis also revealed the presence of a second distinct M . ulcerans genotype co-circulating with Agogo-1 ., This second genotype ( called Agogo-2 ) was substantially more diverse from all other Ghanaian M . ulcerans genotypes ( 138 SNPs ) , suggesting the re-introduction of M . ulcerans to the Agogo region , potentially from a source outside Ghana ( Fig . 1 , S1 Table ) ., The intra-genotype variation within either cluster was low ., The mean nucleotide pairwise identity was 94 . 7% ( SEM ± 0 . 4 ) for Agogo-1 versus 97 . 2% ( SEM ± 0 . 4 ) for Agogo, 2 . The mean pairwise nucleotide identity was significantly lower for Agogo-2 genomes compared with Agogo-1 ( p<0 . 001 ) ., To investigate the possible origin of the Agogo-2 isolates we compared SNP profiles among our panel of M . ulcerans genomes from across West and Central Africa ., The closest match obtained was to isolate ITM102686 , obtained from a patient originating from Ibadan , Nigeria , with 29 SNPs different when only this genome was compared to the Agogo-2 cluster ., This close association may indicate that Nigeria was the source of the Agogo-2 cluster ., Some circumspection is needed when interpreting these data , as only two M . ulcerans genomes were sampled from countries east of Benin ., There is however a compelling patient history behind this isolate to support Nigeria as the correct origin ., The Caucasian patient , a long-term resident in Ibadan and an employee of a non-government organisation , believes he was infected on an Ibadan golf course , when he was bitten by black biting flies ( his description suggests they may have been moth flies Psychodidae ) that began plaguing the course when ground works started adjacent to a lake on the course ., The patient developed a painful ulcer on the site of the insect bites ., A couple of months later he developed a second ulcer on an adjacent site on the same limb that was microbiologically diagnosed as a Buruli ulcer ., This patient history , combined with the documented cases of BU in Ibadan , with cases occurring around the Ibadan University campus and other nearby institutions 44 , support Ibadan as the likely origin of M . ulcerans isolate ITM102686 ., We next explored the distribution of M . ulcerans genotypes in the Agogo region at the village scale and observed no obvious pattern or relationship between genotype , patient , strain and village ( Fig . 2 ) ., There is complete intermixing of Agogo-1 and Agogo-2 clusters amongst the population ., Median-joining-network analysis suggested the independent radiation of the two clusters throughout the region ( Fig . 2 ) ., Furthermore , within either cluster there was a broad distribution of cluster subtypes across the region ., For example , isolates F70 and S38 ( Agogo-1 ) have identical SNP profiles but the patients came from Baama and Serebouso , villages separated by 10–15 km ., Similarly isolates F74 and 1510 ( Agogo-2 ) , came from patients who live in two different villages ( Fig . 2 ) ., Patient interviews did not identify any travel histories or other epidemiological links that might explain these distribution patterns ., An 11-year old girl from Serebouso was the third child within her family to have BU ( isolate 212 , November 2012 ) , eight years after two of her siblings had the disease ., The family of this child lived very close to that of another BU patient , a 3 year-old infant ( isolate S77 , February 2010 ) ., Both isolates belonged to the Agogo-2 cluster but their genome sequence differed in nine nucleotide positions , a significant amount of genetic variation given that S77 shares a near identical genotype with F74 and 1510 ., Again , we could not identify any specific activity or travel history such as attending a common community event that was shared by Agogo-2 genotype patients ., These data suggest that, ( i ) the disease is acquired locally ,, ( ii ) multiple M . ulcerans genotypes are circulating simultaneously within the local region and, ( iii ) a single clone can have the propensity to spread through a region ., Further support for local acquisition of infection comes from observations of infants with no travel history with BU such as a locally-born 2-year old infant from Ananekrom identified over the time of this study ., The clonal population structure of M . ulcerans has made identifying and comparing genetic variation in isolates at anything less than a continental scale very difficult ., Here we have used the high resolution afforded by comparative genomics to explore the molecular epidemiology of BU at the regional and village scale ., Like recent studies using a single polymorphic genetic locus or whole genome sequence comparisons to assess M . ulcerans genetic diversity across a range of African countries , we found a highly significant relationship between the genotype of an isolate and its geographic origin at a national and regional scale 42 , 45 ., These repeated observations indicate that M . ulcerans , when introduced to an area , remains localized and isolated for a sufficiently long period to allow mutations to become fixed in the bacterial population and a local genotype to evolve ., It is reasonable to infer therefore , that the environmental reservoirs of the bacterium in these areas are also likely to be somewhat localized and isolated ., However , the current study has shown for the first time how this focal distribution pattern breaks down at a local scale with the presence of identical genotypes appearing concurrently in separate areas of the same district ., There was no discernible distribution pattern for either the Agogo-1 or Agogo-2 genetic clusters , with both M . ulcerans genotypes appearing at the same times and within the same villages across the region ., Interestingly , there were several examples of isolates with identical genome sequences ( e . g . isolates F74 , 1510 or F85 , F65 ) that were obtained from patients living in four different villages , each separated by distances in excess of 10km ( Fig . 2 ) ., There are several potential explanations for these patterns ., The bacteria ( or a vector spreading the bacteria ) may be widely distributed across the region and infections are being acquired locally , or it may be that people are traveling and becoming infected from a common point source ., Patient interviews and travel histories did not reveal any common activity that might explain a point-source transmission scenario , although the long incubation time for this disease ( 4-months ) is likely to make recall of any such events unreliable 46 ., However , on balance the former scenario seems most likely , and we suggest that each genotype of M . ulcerans has now spread equally widely across the region ., If this assumption is correct , then the lack of genetic variation among isolates suggests that the spread of M . ulcerans throughout the region has occurred relatively rapidly , with insufficient time elapsed for mutations to accumulate ., Reliable mutation rates for M . ulcerans have not been established and some solid data here would allow inferences regarding the time particular clones have been extant within a population ., To our knowledge , this is the first report to employ whole genome sequencing to explore the molecular epidemiology of BU at a local scale ., A previous study utilizing high-resolution SNP assays to explore M . ulcerans genetic variation did uncover some suggestion of local genotype clustering and a recent report used VNTR to examine the link between human and environmental sources of M . ulcerans 30 , 47 ., However such approaches rely on variable nucleotides that have been defined from a limited reference genome set ., If this reference genome set does not represent the genetic variation of the isolates under investigation then data analysis can be flawed , with phenomena such as long-branch attraction and phylogenetic discovery bias confounding analyses 48 ., Whole genome sequencing and comparisons of all isolates under investigation as in our study here overcomes the potential weaknesses of targeted SNP-based typing ., SNP-typing could however be employed to classify patient samples as Agogo-1 and Agogo-2 genotypes without relying on sequencing of cultured isolates , as culture sensitivity is only around 30% , depending on transport duration ., Future studies could thus search for clinical phenotypes between these two distinct bacterial genotypes , although no differences were observed in pathology or treatment outcomes among the patients associated with this study ., There are interesting parallels between M . ulcerans and Mycobacterium leprae , the causative agent of leprosy , where genomics has shown that the leprosy bacillus is another example of a niche-adapted , highly clonal , zoonotic mycobacterial pathogen , with the potential to spread from environment-to-human 49–52 ., Mycobacterium tuberculosis might also be considered in a similar context , with genomic population analysis also suggesting interactions among genetically distinct M . tuberculosis lineages 53 , 54 ., One potential issue arising from this study is the risk of incorrectly attributing Nigeria as the origin M . ulcerans genome sequence ITM102686 , as it represents only one isolate ., While the patient history makes a persuasive argument for Ibadan as the source of the infection , additional M . ulcerans isolates are clearly required from patients in different BU endemic regions of Nigeria and surrounding countries , to further explore the relationship and disease transmission patterns we propose here ., Regardless of the precise origin of Agogo-2 isolates , the data presented here suggest that M . ulcerans can be introduced into a region and then be spread extensively ., How might M . ulcerans be imported into a region ?, We speculate that movements of people or perhaps animals between countries could be one likely means , where infected individuals with BU lesions that can contain very high bacterial burdens might inadvertently contaminate aquatic environments during bathing or other water contact activities ., Now is the time to undertake more intensive and extensive whole-genome M . ulcerans sequencing surveys across West Africa , to assess the extent of genotype admixture such as we’ve revealed here ., Enriching our genome data will also inform other research programs that are identifying reservoirs of M . ulcerans , leading to the new knowledge required to design interventions and stop the spread of BU . | Introduction, Methods, Results, Discussion | Efforts to control the spread of Buruli ulcer – an emerging ulcerative skin infection caused by Mycobacterium ulcerans - have been hampered by our poor understanding of reservoirs and transmission ., To help address this issue , we compared whole genomes from 18 clinical M . ulcerans isolates from a 30km2 region within the Asante Akim North District , Ashanti region , Ghana , with 15 other M . ulcerans isolates from elsewhere in Ghana and the surrounding countries of Ivory Coast , Togo , Benin and Nigeria ., Contrary to our expectations of finding minor DNA sequence variations among isolates representing a single M . ulcerans circulating genotype , we found instead two distinct genotypes ., One genotype was closely related to isolates from neighbouring regions of Amansie West and Densu , consistent with the predicted local endemic clone , but the second genotype ( separated by 138 single nucleotide polymorphisms SNPs from other Ghanaian strains ) most closely matched M . ulcerans from Nigeria , suggesting another introduction of M . ulcerans to Ghana , perhaps from that country ., Both the exotic genotype and the local Ghanaian genotype displayed highly restricted intra-strain genetic variation , with less than 50 SNP differences across a 5 . 2Mbp core genome within each genotype ., Interestingly , there was no discernible spatial clustering of genotypes at the local village scale ., Interviews revealed no obvious epidemiological links among BU patients who had been infected with identical M . ulcerans genotypes but lived in geographically separate villages ., We conclude that M . ulcerans is spread widely across the region , with multiple genotypes present in any one area ., These data give us new perspectives on the behaviour of possible reservoirs and subsequent transmission mechanisms of M . ulcerans ., These observations also show for the first time that M . ulcerans can be mobilized , introduced to a new area and then spread within a population ., Potential reservoirs of M . ulcerans thus might include humans , or perhaps M . ulcerans-infected animals such as livestock that move regularly between countries . | In this study we use the power of whole genome sequence comparisons to track the spread of Mycobacterium ulcerans , the causative agent of Buruli ulcer , through several villages in the Ashanti region of Ghana , providing new insights on the behaviour of this enigmatic and emerging pathogen . | null | null |
journal.pbio.2005345 | 2,018 | Inter-subunit interactions drive divergent dynamics in mammalian and Plasmodium actin filaments | Actin is a highly conserved cytoskeletal protein with essential roles in cell division , contraction , and motility ., Cell motility is an important process in biological development , cancer metastasis , and tissue penetration by both immune cells and pathogens ., Many eukaryotic cell types have the ability to move continuously in a substrate-dependent ameboid manner in both 2D and 3D environments , typically by deforming their cellular shape and protruding the cell’s leading edge 1–3 ., Gliding motility is an alternative mode of locomotion displayed by some bacteria and protists that is independent of cell shape changes 4–6 ., The malaria-causing parasite Plasmodium employs gliding motility in several phases of its life cycle ., Gliding is essential for successful penetration of host organs such as the mosquito midgut and salivary glands as well as the mammalian liver 7 ., Prior to liver infection , high motility speeds of 1–3 μm/s allow the parasite to escape from the dermis , where it is deposited by the biting mosquito , and thereby evade infiltrating neutrophils , the fastest migrating human cells , which move much slower ( 1–5 μm/min ) 8 , 9 ., Despite these striking differences in locomotion modes and speeds , the fundamental requirement for these cells is the dynamic turnover of actin filaments 10–12 ., The actin monomer has a highly conserved structure that consists of four subdomains and a central nucleotide ( adenosine triphosphate ATP , adenosine diphosphate with inorganic phosphate ADP + Pi , or adenosine diphosphate alone ADP ) binding cleft ( Fig 1A ) 13 ., Actin possesses the ability to self-assemble from monomers ( G-actin ) to form filaments ( F-actin ) , which in turn can form higher order filamentous structures ., Particular regions in the actin subdomains , such as the hydrophobic plug ( H-plug ) of subdomain 3 and the highly flexible DNAse I-binding loop ( D-loop ) of subdomain 2 , as well as the nucleotide state , have been implicated as major contributors to the formation and stability of filaments 14–16 ., Actin isoforms of most eukaryotes inherently form long and stable filament structures ( greater than 1 μm in length ) , display high sequence conservation across species ( >90% similarity from yeast to humans ) , and are regulated by a large set of actin binding proteins ( ABPs ) 17 ., These ABPs can affect the monomer-filament ratio and fulfil a variety of regulatory roles , including but not limited to nucleation and elongation ( e . g . , formin ) , monomer binding ( e . g . , profilin ) , filament binding ( e . g . , coronin ) , and filament severing ( e . g . , cofilin/actin depolymerising factor ADF family ) ., In contrast , protozoan parasites express divergent actins ( 60%–80% identity with vertebrate actins ) that typically differ in their ability to form actin filaments ., Giardia , trypanosomid , and apicomplexan actins are refractory to actin polymerisation modulating compounds , such as latruculin , and their structures are difficult to visualise with phalloidin 18–21 ., Similarly , actin 1 of Plasmodium has fundamental shifts in its functional properties: despite a similar monomer tertiary structure 22 to mammalian actin , Plasmodium actin only forms short filaments of approximately 100 nm in length , has a noncanonical filament structure that is dynamically unstable 16 , 22–28 , displays slow polymerisation yet rapid depolymerisation rates 29 , and is regulated by a highly reduced set of predicted ABPs 30 ., Such altered properties are crucial for intracellular parasite growth 31–33 and efficient parasite motility 12 , 31 , 33–35 ., These fundamental differences make Plasmodium actin a useful model for the comparative assessment of key amino acid residues that result in altered filament properties ., Here , we combined multiscale molecular dynamics ( MD ) simulations with three newly developed molecular genetic screens to identify the contribution of particular amino acid residues to the altered properties of canonical rabbit actin and the evolutionarily distant Plasmodium actin 1 ., The results reveal a distinct region in subdomain 3 as the primary contributor to divergent actin behaviour and filament incorporation ., Plasmodium actin 1 is one of the most divergent eukaryotic actins known ( 80% sequence identity compared to rabbit alpha actin , S1 Fig ) ., To assess the structural and dynamic differences between actin monomers from parasites , as represented by Plasmodium actin 1 , and canonical mammalian actin monomers , as represented by rabbit actin , we first employed all-atom MD simulations ., Apart from a more flexible region in subdomain 4 , comparison of the breathing dynamics of the actin monomers revealed no marked differences in behaviour between the actins from the two species ( S2 Fig ) ., To assess actin filament dynamics , we constructed 15-mer Plasmodium and rabbit actin filaments based on electron microscopy ( EM ) data for the rabbit alpha actin filament 36 and performed five coarse-grained ( CG ) MD simulations of 10-μs duration for each filament ., These simulations recapitulated the modified filament architecture observed in previous EM studies , in terms of alpha angle and subunits per crossover 16 , 22 , 27 ( S3 Fig ) ., Moreover , our models allowed for the comparative assessment of the dynamics of the molecular interactions both within and between subunits in the filaments ., Comparison of the filament models for the two species revealed that both filaments displayed essential common interaction hot spots in regions involved in intermolecular contacts ( Fig 1B–1E ) ., Two interaction regions in subdomains 1 and 3 ( designated S1b and S3a ) are highly conserved across species ( over 90% sequence identity of Plasmodium compared to rabbit alpha actin ) ., Other interaction regions corresponded to sequences in subdomains 2 , 3b , and 4 ( Fig 1E , overall sequence identities of 78% , 72% , and 77% , respectively ) ., Interestingly , within these regions there are residue differences between Plasmodium and rabbit actin filaments and thus altered contacts between filament subunits ( Fig 1E ) ., Thus , whereas the locations of the interfaces are conserved , the residues are , on average , more divergent at the interfaces than elsewhere ( S1 Table ) ., We sought to assess the contribution of these divergent regions to the altered behaviour of actin species in their cellular context ., As a first unbiased screen , chimeras containing exchanges between Plasmodium and mammalian actin were generated and tested for their ability to replace the endogenous actin 1 gene in the parasite ( Fig 2A and 2B ) ., To do so , we established a two-step genetic methodology that allowed for replacement of the endogenous copy while leaving the desired locus selection marker free and with minimal flanking changes in the parasite genome ( S4 Fig ) ., For rabbit actin-based chimeras , our screening approach revealed that any changes to this actin were insufficient to obtain viable parasite lines , strongly indicating that no one region is capable of full restoration of parasite-like actin function from a canonical backbone ., Even for exchanges introduced into the parasite actin , the majority of the parasite actin chimeras were unable to functionally replace the endogenous gene ., Importantly , changes to subdomains 2 and 3 of the parasite actin resulted in parasite death , indicating that these regions provide essential features required for parasite blood stage growth ( the stage at which the genetic manipulations are performed , Fig 2A ) ., Further , mutants with single point mutations in these subdomains ( P42Q and K270M ) were also lethal , strongly implicating the essentiality of these regions for normal parasite function in the blood stages of infection ., However , two of the most divergent regions could be exchanged with the canonical mammalian actin equivalent regions: an N-terminal domain swap ( PbS1aOc , residues 1–33 , 67% sequence identity with rabbit alpha actin with one additional acidic residue on its N-terminus ) and a subdomain 4 replacement ( PbS4Oc , residues 182–263 , 77% sequence identity with rabbit alpha actin ) ., Interestingly , these transgenic parasites displayed asexual growth rates in the blood that were comparable to the wild-type replacement control ( Fig 2C ) , suggesting that modification of these regions does not significantly affect intracellular growth and erythrocyte invasion ., The malaria parasite needs to propagate and disseminate in a variety of tissue environments and different temperatures , suggesting that the required actin dynamics of these other stages could be different 37 , 38 ., We thus tested if the PbS1aOc and PbS4Oc chimeric lines were effective in infection of Anopheles mosquitoes ., After transmission to the mosquito midgut , the parasite develops into midgut-penetrating ookinetes , which can traverse the epithelia to establish oocysts on the basal lamina ( Fig 2A ) ., These oocysts produce thousands of sporozoites that are subsequently released into circulation and actively invade the insect’s salivary glands 7 ., PbS1aOc produced slower moving ookinetes and reduced oocyst numbers ( Fig 3A and 3B ) ., Furthermore , while the numbers of sporozoites in the mosquito circulation were within the usual range , suggesting normal oocyst development , a marked reduction in salivary gland occupancy and sporozoite motility was observed ., PbS1aOc displayed a 50% reduction in parasite numbers in the salivary glands ( Fig 3C ) ., While this line was capable of infecting naive mice at similar rates to wild-type controls ( Fig 3D and 3E ) and showed a similar motile population compared to wild-type controls ( Fig 3F ) , these parasites moved at a lower average speed ( Fig 3G ) ., Thus , a modification of the N-terminal region of actin , with a concomitant increase in the number of acidic residues , resulted in a decreased motility of two different parasite stages , which affected their ability to penetrate the organs of the mosquito ., Alteration of the amino acid residue composition of subdomain 4 ( PbS4Oc ) resulted in a much more pronounced defect in salivary gland invasion ( Fig 3C ) ., Again , this line was still able to cause infection in mice by natural mosquito transmission , although the decreased numbers of parasites in these glands resulted in a concomitant delay in infection ( Fig 3D and 3E ) ., The motile population of salivary gland–resident parasites was reduced and these moved in a discontinuous fashion ( Fig 3F–3H ) ., Interestingly , these parasites often paused during motility , a phenomenon not yet described for any Plasmodium mutant ( Fig 3H ) ., Intriguingly , such events were often accompanied by short reversals in direction and detachment at the rear of the parasite ( Fig 3I and 3J , S1 Movie and S2 Movie ) ., While pausing did not change the average speed compared to wild-type , there was an increased range of speeds obtained by individual parasites ( Fig 3G ) ., Changing the subdomain 4 region involved the alteration of 20 amino acid residues to canonical equivalents ., Strikingly , the pausing phenotype was also observed by combined modification of only three residues in this region ( three combined parasite to mammalian residue mutants H195T , G200S , and E232A: mutant HGE/TSA ) ( Fig 3 and S5 Fig ) ., CG MD simulations with this triple mutant indicate that a change of these three residues results in a shift of filament parameters toward rabbit actin ( S3 Fig ) and an atypical interaction profile that is not observed in either species: residues F54 and Y189 display enhanced contacts with Y170 and H174 , respectively , which could have consequences for actin filament turnover ( S6 Fig ) 16 , 36 ., Taken together , altering actin dynamics by mutation of nonessential actin regions results in parasites that , while behaving normally in the mammalian host , have prominent inabilities to effectively colonise the mosquito vector ., Plasmodium actin cannot be visualised using typical imaging approaches 39 , yet expressing the F-actin binding Plasmodium coronin fused with mCherry enabled visualisation of F-actin in motile sporozoites 37 ., Coronin localises at the periphery of nonmotile sporozoites and relocalises to the rear in motile sporozoites in an actin filament–dependent fashion 37 ., To investigate a possible change in localisation , we transfected PbS4Oc-expressing parasites with mCherry-tagged coronin under the control of a sporozoite-specific promoter ( Fig 4A , S7 Fig ) ., Curiously , coronin localised to the periphery in an actin-independent manner in these sporozoites resembling the localisation in nonmotile wild-type sporozoites or after treatment with the actin filament modulator cytochalasin D ( Fig 4B ) ., This indicates that the majority of the tagged coronin was unable to bind the modified actin , similar to observations with a coronin actin binding mutant and when parasites are treated with filament stabiliser , jasplakinolide 37 ., Intriguingly , overexpression of coronin , but not overexpression of profilin and ADF2 , rescued efficient motility and salivary gland penetration of PbS4Oc sporozoites ( Fig 4C–4E ) ., Overexpression of coronin in WT sporozoites also increased their capacity to move 37 , suggesting that in PbS4Oc sporozoites , coronin overexpression is compensating for its reduced binding ability to a mutated actin ., To further probe the domain contributions to this rescue , we tested whether mutations to coronin would improve motility and invasion of the PbS4Oc parasite line ., Mutations in the N-terminal actin binding domain ( coronin mutant R349E , K350E ) , as expected , could not rescue the PbS4Oc phenotype ., We also tested if the actin binding N-terminal domain of coronin ( residues 1–388 , lacking the unique region and coiled-coil domain ) , which was shown to bind and bundle actin 40 , affected motility and invasion ., This also could not improve the invasion and motility of the PbS4Oc parasite , suggesting a cooperative involvement of both N- and C-terminal regions of coronin in mediating rescue ( Fig 4C–4E ) ., Above , we identified subdomains 2 and 3 as key contributors to essential processes in blood stage parasites ( Fig 2B ) ., In order to assess the contribution of these regions to altered actin dynamics , we performed a second screen by expressing these variants ( PbS2Oc , OcS2Pb residues 34–71 , PbS3bOc , and OcS3bPb residues 264–338 ) as mCherry-tagged additional copies under the control of a sporozoite stage-specific promoter ( Fig 5A , S8 Fig and S9 Fig ) ., This allowed for careful phenotypic characterisation in the parasite without the lethality observed in the initial screen ., Tagged actins revealed interesting differences: an additional copy of parasite actin was present throughout the parasite cell , as expected 41 ., Yet , mammalian actin was only cytoplasmic with a notable absence in the nuclear region ( Fig 5B ) ., Furthermore , the two ‘control’ parasite lines ( expressing either parasite or canonical actin ) differed in their response to jasplakinolide: parasite mCherry-actin accumulated at both the front and rear of the cell ( Fig 5B ) , sites implicated for F-actin formation ( in the case of formin 1 , which is located at the front; S8 Fig ) and turnover ( rear ) 37 , 40 , 41 , 42 ., In contrast , mammalian mCherry-actin displayed no change in localisation ( Fig 5B ) ., This suggests that tagged mammalian actin is not readily incorporated into jasplakinolide-stabilised parasite F-actin in the sporozoite ., We next assessed localisation and jasplakinolide responsiveness with the actin chimeras ., Unexpectedly 22 , the chimeras containing exchanges in subdomain 2 ( a highly flexible and divergent region between species ) were similar in behaviour to their corresponding controls ( Fig 5C ) ., In contrast , exchanging subdomain 3 resulted in reciprocal effects: a mammalian actin containing this corresponding parasite region ( residues 264–338 with 21 changes; 72% sequence identity ) resulted in a changed localisation under jasplakinolide , thus suggesting more efficient incorporation to endogenous filaments ( Fig 5D ) ., Changing the same region in the parasite actin resulted in a cellular distribution and jasplakinolide response similar to mammalian actin , suggesting that subdomain 3 contains the contacts necessary and sufficient for incorporation into divergent filaments ., Importantly , the effect appears to be independent of the H-plug , as changing the two divergent amino acid residues on the loop and the two most divergent residues outside this loop back to residues in the canonical actin ( K270M , A272S , E308P , and T315Q ) still rendered an additional copy of actin that behaved similarly to Plasmodium actin ( Fig 5D ) 16 ., Parasites expressing rabbit actin displayed a 50% reduction in parasite average speed ( Fig 5E ) ., Interestingly , all chimeras of Plasmodium actin consistently moved with a slightly higher average speed compared to their corresponding mammalian chimeras ., OcS3bPb , the only mammalian actin chimera that responded to jasplakinolide treatment , moved at a similar speed to the wild-type Plasmodium actin control ( Fig 5E ) ., To test whether the specific regions that we identified above are common contributors to actin dynamics in other cell types , we performed a third screen by transfecting the panel of chimeras into different mammalian cells as additional copies ., Higher eukaryotic cells did not readily incorporate parasite actin into stable filamentous structures ( Fig 6A and 6B , S10 Fig ) ., Furthermore , cells containing parasite actin moved slightly faster in a random migration assay ( Fig 6C ) ., These observations indicate that malaria parasite actin can be used as a template to identify the minimum determinants required for filament incorporation in higher eukaryotic cells ., Strikingly , changing subdomain 2 or 3 into the corresponding mammalian equivalent resulted in the parasite GFP-actin more readily incorporating into filament networks ( Fig 6A and 6B ) ., In contrast , exchanging individual regions of mammalian actin did not result in decreased incorporation , suggesting that more than one region can provide the required minimal contacts necessary for incorporation ., To identify key amino acid residues contributing to filament incorporation , multiple mutations ( whereby parasite actin residues were converted to conserved mammalian counterparts ) in both subdomain 2 and subdomain 3 were generated and analysed ( Fig 6A and 6B ) ., Consistent with the results of the parasite lines , single and multiple mutations of both the H-plug and regions typically on the surface of the actin filament in subdomain 3 did not result in improved filament incorporation , suggesting an extensive interface in the subdomain 3 region required for incorporation ., Remarkably , a single residue change ( N41H ) in subdomain 2 of Plasmodium actin was sufficient to rescue filament incorporation ( Fig 6A and 6B ) ., This indicates that the presence of an imidazole group alone provides sufficient interaction for more efficient filament incorporation ( Fig 6A and 6B ) ., Thus , crucial differences between actin species , including the ability for a monomer to incorporate into divergent filaments , are due to the changes in particular divergent regions in specific subdomains only ., Furthermore , subdomain 3 is the key common contributor to filament incorporation in two very diverse cellular systems ., While , structurally , the Plasmodium actin 1 monomer is similar to that of canonical mammalian actins 22 , Plasmodium actin filaments are shorter and more dynamic 22 , 26 , 27 ., Here , we have shown that Plasmodium actin is unable to efficiently incorporate into actin filaments of mammalian cells and that tagged mammalian and Plasmodium actin localise differently in parasites ., Using MD simulations as well as chimeric and mutagenesis genetic screens , we identified the fundamental contributors that underlie these different characteristics ., Our data independently confirm that these two actins occupy a similar structural space and identify divergent amino acid residues responsible for the differences in filament dynamics ., Plasmodium was unable to tolerate allelic changes in subdomains 2 and 3 , suggesting that these regions have crucial features required for parasite biology and are , particularly P42 and the H-plug , key contributors to contact sites ., The data suggest that modifications of these two regions result in filaments that are too stable and thus lethal for the parasite , similar to the effects observed with jasplakinolide treatment or genetic ablation of the actin regulator , ADF1 , and the alpha capping protein 31 , 32 , 43 , 44 ., Unlike exchanges in subdomains 2 and 3 , the parasite readily tolerated the exchange of highly divergent regions of subdomain 4 and the N-terminal residues ( Fig 2 ) ., These parasite lines grew like wild-type parasites in the blood stage and were only affected in their progression in the mosquito host ( Fig 3 ) ., Importantly , this indicates that red blood cell invasion , while dependent on actin 33 , is more tolerant of minor changes to its sequence ., This observation is consistent with ablating ABP expression having little effect on the parasite in the mammalian host but having a strong effect once the parasite infects the insect vector 37 , 44–48 ., While other studies looked indirectly with actin regulators , our study investigated the core of the machinery itself across the life cycle , revealing that modest changes to dynamics have important consequences in mosquito colonisation ., This indicates that the greatest selection pressure on the parasite actin sequence could be during active organ penetration of the mosquito ., Changes to the N-terminal sequence resulted in a consistent decrease in speed between the two parasite stages that employ gliding motility in the mosquito ( Fig 3B and 3G ) ., Given the well-known interaction between the actin N-terminus and myosin 49–56 , it is reasonable to suggest that the decrease in cell speed is due to a reduced interaction between these two proteins ., We thus propose that parasites containing this change in acidic residue content have reduced force transmission of actomyosin , which results in reduced average speeds ., This hypothesis could be probed in vitro using classical myosin sliding filament assays with the respective Plasmodium machinery components 57 ., Previous studies in yeast have indicated a relatively low threshold in opisthokonts to accommodate changes to its actin subdomain 1 sequence , in which the addition of a single acidic residue to the N-terminus rendered viable yet sick yeast cells , while two additional acidic residues were lethal for the cell 58 ., In comparison , the addition of another acidic residue to parasite actin was surprisingly well tolerated ( Figs 2 and 3 ) : these parasites could complete the life cycle and move sufficiently well to cause infection ( albeit at a lower overall efficiency in the mosquito ) ., Interestingly , yeast cells with changes to subdomain 1 and the N-terminus were sensitive to changes in growth temperature 58–61 ., Given the marked differences in body temperatures between mosquito and mammal , it is possible that changes in parasite actin sequence affect the well-tuned actin dynamics required at different temperatures ., In vitro polymerisation assays with these mutants at both 37°C and 21°C would provide important insights into differences in the polymerisation kinetics at different temperatures ., These detailed kinetic studies are especially interesting given that hybrid actins between rabbit muscle and yeast equivalents can produce unexpected changes to certain biochemical properties 58 ., Interestingly , modifications of three amino acid residues in subdomain 4 rendered a parasite that frequently paused and reversed direction during migration ( Fig 3H–3J ) ., We propose that this change in motility is due to more stable actin filaments , which might be misoriented and hence allow partial rearward motility ., Similar movements can be observed if Toxoplasma parasites are treated with high concentrations of jasplakinolide , a filament-stabilising drug 62 ., We showed previously that coronin can bind to actin filaments in motile sporozoites but not to those treated with jasplakinolide 37 ., Intriguingly , in the PbS4Oc sporozoites , coronin did not bind actin filaments and hence this parasite to some extent phenocopies jasplakinolide treatment ., These parasites could also not enter salivary glands efficiently , for which well- tuned actin dynamics are important 37 , 45 ., Yet , overexpression of coronin rescued both motility and invasion phenotypes ( Fig 4 ) ., This indicates a central role for coronin in rapid filament recycling in the highly motile parasite ., We propose that the discontinuous movements in subdomain 4 mutants ( PbS4Oc and HGE/TSA ) result from reduced actin filament recycling due to altered architecture and interactions at the filament interface , particularly by enhanced interactions of F54 and Y189 ., By alteration of subdomain 4 residues , the resulting contacts by conserved amino acid residues render the parasite actin filament less prone to disassembly as choreographed by coronin ( S11 Fig ) ., Reduced invasion and aberrant motility were only rescued by full-length coronin , while mutated or truncated coronins were insufficient to rescue these phenotypes ., This may suggest a cooperative requirement of the coronin domains for actin turnover ( Fig 4C–4E ) ., This result was somewhat surprising , because a recent publication on the biochemical properties of P . falciparum coronin suggested that much of the classical coronin function could be carried out by the N-terminal actin binding domain alone 40 ., Our work indicates that , in the context of the P . berghei actin mutated sporozoite , both regions of coronin are required for full functionality ., The mechanism by which coronin could be mediating Plasmodium filament recycling could be similar to that in higher eukaryotes ., In these , coronin mediates a spatial-temporal recruitment of other filament regulators 63–65 ., Coronin binds first and , depending on the nucleotide state of the actin 63 , 66 , can serve to shield or allow access of ADFs to the filament ., Binding of these additional factors together results in destabilisation of the filament ., Given the close proximity of coronin and ADF on canonical filaments , it was also suggested that coronin could interact directly with ADF in the filament 64–67 ., It is tempting to speculate that the rescue mediated by coronin could be due to a stepwise recruitment by coronin , with depolymerisation factors on the highly dynamic Plasmodium actin filament , which thus enhances severing ., Our rescue experiments thus hint at a possibly conserved spatial-temporal coronin mechanism for filament regulation beyond what has currently been demonstrated for opisthokonts ., The C-terminal region of coronin appears important for this rescue , which suggests that this domain could allow for an interaction with other ABPs 68 ., Indeed , it has been suggested for coronin of the related parasite T . gondii that the C-terminal coiled-coil domain could function as a molecular recruitment hub 69 ., Alternatively , the C-terminal region simply provides an improved interaction with the filament , resulting in efficient turnover in the parasite ., Together , our observations suggest interplay between both halves of Plasmodium coronin to mediate efficient rescue in motile sporozoites ., A recent high-resolution cryo-EM structure of the jasplakinolide-stabilised Plasmodium actin filament 16 is in good agreement with our models and generally fits with our biological observations ., However , some of our in vivo observations could not be directly inferred from the static structure ., For example , the combined mutation of H195 and G200 ( to Thr and Ser , respectively ) at a lateral interface in Plasmodium actin had no significant effect on the parasite across the entire complex life cycle ., A third residue , E232 , which is not at an interface , with the exception of one published structure 70 , needed to be mutated in order for an effect on parasite biology to be observed , suggesting a dynamic interplay across several residues in this region ( Fig 3C–3H ) ., It is possible that E232 could provide a compensatory interaction when H195 and G200 are mutated ., Our systematic approach , which spans from molecular models to cells , has thus identified novel cooperative interactions within actin filaments ( Fig 7 ) ., In this study , we made use of additional copy expression systems in highly divergent cells ., It is important to note that these assays can only provide insights into monomer incorporation into filaments ., Thus , we can make no direct inferences regarding whether each actin mutant , if present as the primary untagged actin in the cell , can fully complement wild-type function ., For example , PbS2Oc was lethal as a replacement yet behaved similarly to wild-type Plasmodium actin as an additional copy in the parasite ( Fig 5 ) ., Newly available genetic tools , such as inducible knockout systems 33 , might be adapted to probe these otherwise lethal mutants for their functional consequences in cells ., Our observation that parasite actin does not efficiently incorporate into the mammalian cell actin network provided a useful tool in understanding the minimal requirements for network incorporation ., Interestingly , subdomain 4 conversion of parasite actin did not improve network incorporation efficiency , while changing subdomain 2 or 3 resulted in a considerable incorporation into filamentous structures ., Subdomain 2 , with its highly dynamic extendibility , has been implicated to act as the primary interaction arm , after initial contact by subdomain 4 , to allow the incoming monomer to bind to the barbed end 36 ., The N41H mutant in Plasmodium actin subdomain 2 alone shows improved incorporation into mammalian actin filaments ., Histidine has the capacity to have more contacts than asparagine and thus this position might be the key prominent interaction to allow optimal contact with the filament 16 ( Fig 1E ) ., Indeed , oxidation of H40 results in an actin monomer that is unable to polymerise 71 , 72 ., Here , we extend this understanding to include this contact as important for an incoming monomer to dock onto a filament ., We , with others , have shown that subdomain 3 provides important contact sites across the strand once the monomer has been incorporated and also for the incoming monomer to be included in the filament 15 , 36 , 73 ( Fig 1 and Fig 7 ) ., Furthermore , we cannot exclude that subdomain 3 could be providing the required contact interface to allow for increased ABP-mediated incorporation 74 ., We propose that having at least one of these important subdomain sites ( 2 or 3b ) is sufficient for effective incorporation in higher eukaryotic cells ., Identification of altered ABP binding between actin species could shed light on the primary factors responsible for monomer incorporation ., Notably , there is a consistent feature between divergent systems: alteration of subdomain 3 allowed for a mammalian equivalent to be more efficiently incorporated into the highly divergent parasite cytoskeleton , indicating that this region could be a common ‘site of recognition’ across differently behaving actin species ., This could be tested with other highly divergent actins to assess whether this region acts as a common recognition point across species ., Through our comparative approaches , we have identified essential contributors to the differential behaviour of two highly divergent actin species ( Fig 7 ) ., We have identified a region of subdomain 3 outside the H-plug as providing important contacts to allow an otherwise divergent actin to be more efficiently incorporated into canonical actin filaments , thus enhancin | Introduction, Results, Discussion, Materials and methods | Cell motility is essential for protozoan and metazoan organisms and typically relies on the dynamic turnover of actin filaments ., In metazoans , monomeric actin polymerises into usually long and stable filaments , while some protozoans form only short and highly dynamic actin filaments ., These different dynamics are partly due to the different sets of actin regulatory proteins and partly due to the sequence of actin itself ., Here we probe the interactions of actin subunits within divergent actin filaments using a comparative dynamic molecular model and explore their functions using Plasmodium , the protozoan causing malaria , and mouse melanoma derived B16-F1 cells as model systems ., Parasite actin tagged to a fluorescent protein ( FP ) did not incorporate into mammalian actin filaments , and rabbit actin-FP did not incorporate into parasite actin filaments ., However , exchanging the most divergent region of actin subdomain 3 allowed such reciprocal incorporation ., The exchange of a single amino acid residue in subdomain 2 ( N41H ) of Plasmodium actin markedly improved incorporation into mammalian filaments ., In the parasite , modification of most subunit–subunit interaction sites was lethal , whereas changes in actin subdomains 1 and 4 reduced efficient parasite motility and hence mosquito organ penetration ., The strong penetration defects could be rescued by overexpression of the actin filament regulator coronin ., Through these comparative approaches we identified an essential and common contributor , subdomain 3 , which drives the differential dynamic behaviour of two highly divergent eukaryotic actins in motile cells . | Actin is one of the most abundant and conserved proteins across eukaryotes ., Its ability to assemble from individual monomers into dynamic polymers is essential for many cellular functions , including division and motility ., In most cells , actin is able to form long and stable filaments ., However , an actin of the malaria-causing parasite Plasmodium , while having a very similar monomer structure to actins from other eukaryotes , forms only short and unstable filaments ., These short and dynamic filaments are crucial in allowing the parasite to move very rapidly in tissue ., Here we investigated the basis of these differences ., We used molecular dynamics simulations of actin filaments to investigate the actin–actin interfaces in filaments from Plasmodium and rabbit ., We next engineered parasites to express chimeric actins that contained different parts of rabbit and parasite actin and thereby identified actin residues important for parasite viability and progression across the life cycle ., We could rescue the most prominent defect specifically with overexpression of the actin binding protein coronin ., This suggests that the more stable actin harms the parasite and that coronin helps in recycling filaments ., By screening the effects of actin chimeras in mammalian cells , we also identified regions that allow these different actins to efficiently interact with each other ., Taken together , our results improve our understanding of the interactions required for actin to incorporate into filaments across divergent eukaryotes . | cell motility, parasite groups, medicine and health sciences, actin filaments, plasmodium, vertebrates, rabbits, parasitic diseases, animals, mammals, parasitology, animal models, developmental biology, apicomplexa, experimental organism systems, amniotes, dynamic actin filaments, research and analysis methods, contractile proteins, actins, proteins, life cycles, biochemistry, cytoskeletal proteins, leporids, eukaryota, cell biology, biology and life sciences, sporozoites, organisms, parasitic life cycles | null |
journal.pgen.1000231 | 2,008 | Predicting Unobserved Phenotypes for Complex Traits from Whole-Genome SNP Data | Results from linkage analyses and , more recently , genome-wide association studies ( GWAS ) imply that a large number of loci underlie the genetic architecture of complex traits 1–15 ., GWAS are usually multi-staged , have mostly focused on gene discovery and typically set very stringent type-I error rates in the first stage to avoid false positives ., Analysis is most frequently performed one SNP at a time ., Consequently , these studies may not properly capture all of the genetic variation that is present in the samples , The initial wave of GWAS has found many genetic variants that are robustly associated with disease or quantitative traits , but these variants typically explain only a small fraction of the genetic variance , and so the utility of predictions made using this information can be limited ., An alternative to gene discovery is to focus on the prediction of phenotypes using all genotypic ( SNP ) information across the whole genome simultaneously ., The prediction of phenotypes is useful in a range of fields , from artificial selection programs 16 to risk prediction in human medicine 17 and forensics ., To predict phenotypes , identification or genotyping of causal variants is not necessary , as long as there are variants genotyped that are in linkage disequilibrium ( LD ) with the causal variants 16 , 17 ., To predict phenotypes from genomic data , the relationship between genome-wide marker data and phenotypes needs to be modeled ., The single SNP regression approach that is often applied in conjunction with stringent thresholds would be expected to inaccurately estimate the proportion of variance that can be explained from genotypic data ., Instead , model selection approaches are required to find the set of SNPs that best explains and predicts variation in phenotype ., Such approaches have already been proposed for mapping multiple quantitative trait loci ( QTL ) 18–23 and recently a method was suggested for the simultaneous analysis of all SNPs in a GWAS 24 ., In this study , we use statistical modeling to fit multiple SNP effects from a GWAS and derive the best model with a Bayesian model selection approach termed Reversible Jump Markov Chain Monte Carlo ( RJMCMC ) 25 ., We predict unobserved phenotypes for individuals based on genome-wide SNP data only , family information ( without genetic data ) only , or on a combination of the two ., Publicly available data including pedigree , genotypic and phenotypic information on heterogeneous stock mice were used ( 26; http://gscan . well . ox . ac . uk/ ) ., The total number of animals was 2 , 296 from 85 unrelated families ., The available pedigree spanned four generations , generating complex relationships ., In the last generation , there were 172 full sib families with an average size of ∼11 ( SD ∼8 ) ., Genotypes were available for 12 , 112 SNPs on most animals in the pedigree , and we used the 11 , 730 SNPs on the autosomal chromosomes ., Phenotypes were already adjusted for the environmental fixed effects , e . g . sex , age , year and season 26 , 27 ., We chose three phenotypes , coat colour as a complex trait with a number of known causal loci ( estimated h2≈0 . 72 ) , and percentage of CD8+ cells ( %CD8 ) as a quantitative trait having high heritability ( estimated h2≈0 . 99 ) , and mean cellular haemoglobin ( MCH ) as a quantitative trait having moderate heritability ( estimated h2≈0 . 55 ) ., Coat colour , as used here , is a measure of the darkness of the coat from white to black ., For more detail about the data , see 26 , 27 ., We fitted a range of linear mixed models , with multiple SNPs as fixed effects and , in some models , a polygenic effect to account for additive genetic effects not detected by the SNPs ., These polygenic effects are estimated from the pedigree ., The effect of a SNP genotype on the phenotype was modeled either by fitting the additive term of one of the alleles or by fitting both additive and dominance terms ., In the additive genetic model , phenotypic observations are a linear function of fixed effects , a polygenic term representing the sum of unidentified additive genetic effects , the additive effects due to SNPs associated with QTL and residuals ., The linear model can be written as , ( 1 ) where y is a vector of length Nr , with single trait phenotypes for all animals corrected for fixed environmental effects ( Nr\u200a=\u200ano . observations in Table 1 ) , nq is the number of SNPs associated with the QTL involved in phenotypic expression , μ is the overall mean , is a vector of Nr ones , u is a vector of N random polygenic effects for N animals ( N\u200a=\u200a2296 ) , αi is the fixed effect of the ith SNP and e is a vector of residuals ., Z is an incidence matrix for the random polygenic effects relating observations to individual animals , with dimensions Nr×N ., Note that N>Nr as some animals have a polygenic effect estimated based upon phenotypic information from relatives without having a phenotypic observation themselves ., Λi is a column vector of length Nr having coefficients 0 , 1 or 2 representing indicator variables of the genotype for each animal at the ith SNP ., The variance structure of phenotypic observations is written as , where A is the numerator relationship matrix , I is a identity matrix , is polygenic additive genetic variance and is error variance ., In the model containing additive and dominance effects , all terms are the same as the additive genetic model except that dominance effects due to SNPs are added ., The model is written as , ( 2 ) where δi is the dominance effect of the ith SNP and Δi is a column vector having coefficients that are 1 for a heterozygous genotype and 0 for a homozygous genotype at the ith SNP ., We predicted phenotypes of individuals by using information on relatives and/or the estimated effects of their SNP genotypes ., Prediction of phenotypes was based on BLUP of polygenic values 29 using pedigree and phenotype information only ( i . e . ( 1 ) with the third term omitted ) , or on additional genomic information where the prediction model was based on additive effects only ( model A ) or on both additive and dominance effects ( model AD ) ., Both A and AD models were fitted with and without the effects of additional polygenic factors from pedigree relationships ., For the single SNP analyses , the prediction was performed from a multiple regression analysis using those SNPs that were selected previously from the single SNP analyses ., As for the AD model , the single SNP analyses also fitted polygenic effects , plus the additive and dominance effect of the SNP ., To assess how well we predicted unobserved phenotypes , we used one part of the data for estimation and the remaining part for prediction and validation ., Approximately half of the phenotypic data for each trait were randomly selected ., Using only half of the phenotypes and all genotypes , the other half of the phenotypes ( i . e . future , unobserved , phenotypes ) were predicted with the proposed genetic model using whole genome SNP data ., We tested how well we could predict phenotypes from genetic data in two ways ., The first prediction was within families , using phenotypic data from approximately half of the animals in each full sib family to predict the phenotypes of the other half of the animals ( intra-family comparison ) ., The second prediction was across families , using phenotypic data from approximately half of the 85 unrelated families to predict the phenotypes of the animals in the other half of the families ( inter-family comparison ) ., The latter prediction could also be used for data sets that lack pedigree information ., When fitting the pedigree only , i . e . not using any marker data , there is no ability to predict the phenotypes of animals in other , unrelated families , so the accuracy of the inter-family prediction is zero ., For each comparison , we correlated the predicted genotype of an animal in the prediction set with its phenotype ( which was not used in the estimation phase ) ., We term the correlation between predicted phenotypes and actual phenotypes as the accuracy of prediction ., To gauge the precision with which this correlation is estimated we performed 10 replicates ., For each replicate , the estimation and prediction sets were sampled and analyzed ., In addition to performing the model selection procedure and prediction from the entire autosomal SNP genotype set , we also investigated how well genotypic data from a single chromosome could predict phenotypes ., For individual chromosome analyses , the AD model was used for the inter-family prediction with a single replicate per trait ., The total number of original phenotypes , the number of phenotypes used in the estimation analysis and the number of phenotypes to be predicted but not used in the estimation step are shown in Table 1 ., For the prediction set , on average approximately 700 ( %CD8 ) to 950 ( coat colour ) observations were used ., Table 2 shows the correlation between true and estimated phenotypes of the three different traits when using the intra- or inter-family prediction ., It shows that the use of genomic information substantially increases the accuracy of predicting unobserved phenotypes , compared to BLUP ( fitting only the pedigree ) , and a substantial accuracy was achieved even with inter-family prediction , where genomic and phenotypes data in some families was used to predict phenotypes in other families ., The accuracy of prediction is highest with intra-family prediction when using genomic information and phenotypic information from relatives to predict an individual phenotype ., For example , for %CD8 and an additive model of gene action and fitting the pedigree , the correlation between predicted and observed phenotype is 0 . 71 whereas it is 0 . 64 when using only pedigree information ., The accuracies of prediction with the model AD are generally greater than those with model A for intra- and inter-family prediction ., The difference between the accuracies with and without considering dominance varies across the traits ., For coat colour , the accuracy of prediction substantially increases in both intra-family ( 0 . 72 to 0 . 89 ) and inter-family ( 0 . 58 to 0 . 87 ) prediction ., For %CD8 , the accuracy increases slightly for the intra-family prediction ( 0 . 71 to 0 . 73 ) ., The increase due to inclusion of dominance is larger for the inter-family prediction ( 0 . 50 to 0 . 58 ) ., For MCH , the accuracy slightly increases for both intra- and inter-family prediction ., These results are consistent with a substantial amount of dominance variance for coat colour , some dominance variance for %CD8 and little dominance variance for MCH ., When omitting polygenic terms in the genetic model and using whole genome marker information only , the correlations between predicted and actual phenotypes are generally decreased for the intra-family prediction , and practically unchanged in inter-family prediction except for coat colour ( Table 2 ) ., The bottom two rows of Table 2 for the inter-family prediction show that phenotypes can be predicted from marker data and phenotypes observed in ‘unrelated’ families ., For coat colour and the AD model , the prediction is very good ( correlation of 0 . 81 ) ., The precision with which phenotypes can be predicted from genetic data is , of course , limited by how much of the variation between individuals is due to genetic factors ., Prediction of unobserved phenotypes from genetic data will never be accurate for traits with a low heritability , even if the prediction of the genetic effect is 100% accurate ., To quantify how much of the variation between individuals due to genetic effects we detected , we scaled the accuracy of predicting phenotypes by h , the square root of the heritability ., This parameter represents the correlation between additive genetic value and phenotype , and is a key parameter in artificial selection programs 30 ., The scaled accuracy is an estimate of the precision with which additive genetic values are predicted ., When using an additive genetic model and whole genome information ( Model A ) , this estimated correlation between predicted and inferred genetic values for the intra-family prediction was 0 . 84 , 0 . 71 and 0 . 71 for coat colour , %CD8 and MCH , respectively , and 0 . 68 , 0 . 50 and 0 . 47 for the inter-family prediction ( Table 3 ) ., When using an additive and dominance genetic model and whole genome information ( Model AD ) , the estimated correlation between predicted and inferred additive genetic values for the intra-family prediction was 1 . 05 , 0 . 73 and 0 . 75 for coat colour , %CD8 and MCH , respectively , and 1 . 02 , 0 . 59 and 0 . 48 for the inter-family prediction ( Table 3 ) ., Therefore a large proportion of existing genetic variation was detected and exploited by our application ., It should be noted that the values for model AD should be scaled by the square root of the broad-sense heritability which was , however , unknown ., Instead , we scaled the values for the AD model by narrow-sense heritability , which may result in an overestimation of accuracy depending on the amount of dominance variance ., Figures 1 , 2 and 3 show that the accuracy of prediction is higher when considering whole genome information compared with using information from one chromosome at a time ., Even with coat colour , a single gene or a single chromosome does not determine all variation in phenotypic expression ( Figure 1 ) ., Although the accuracy of prediction when considering chromosome 7 alone is high ( 0 . 79 ) , the accuracy can be improved when using whole genome information ( 0 . 88 ) ., With %CD8 ( Figure 2 ) , the accuracy of prediction obtained by considering each chromosome at a time ranges from 0 . 05 to 0 . 50 , implying that most chromosomes contribute to variation in this complex phenotype ., When considering the entire genome simultaneously , the accuracy of prediction increases to 0 . 63 ., With MCH , the accuracy obtained from individual chromosomes varies up to 0 . 23 ( Figure 3 ) ., However , again the accuracy of prediction is highest ( 0 . 40 ) when using whole genome information ., The estimated negative correlations between actual phenotypes and predictions based upon a single chromosome ( e . g . , Figure 1 ) is most likely due to sampling error ., Chromosomal analyses were done for a single replicate ., The whole genome approach based on fitting multiple SNPs and using RJMCMC for model selection provides a posterior density of each SNP being associated with the phenotype ., Therefore , the positions of trait loci can be estimated ( e . g . Figure 4A , C and E ) ., For comparison , the method using regression on single SNPs that considers one position at a time was used ., This method yields a likelihood ratio ( LR ) for each SNP which was plotted against genomic position ( Figure 4B , D and F ) ., Averages of the posterior QTL density or LR from the 10 replicates are shown for the inter-family prediction ., For coat colour , high posterior densities are shown for the regions around ∼159 Mb on chromosome 2 , ∼80 Mb on chromosome 4 and ∼80 Mb on chromosome 7 ( Figure 4 A , C and E ) ., These regions agree very well with the positions of a number of known genes for variation in coat colour 31 ( diamonds in Figure 4 ) ., Specifically , the non-agouti gene is at 154 Mb on chromosome 2 , tyrosinase-related protein is at 79 Mb on chromosome 4 , and the tyrosinase and Rab38 genes are at 81 Mb and 82 Mb , respectively , on chromosome 7 ., The LR profiles from the single SNP method are similar to that from the multiple SNP method ( Figure 4B , D and F ) ., However , correlated estimates due to linkage disequilibrium between the causal genes and multiple SNPs cause a broad confidence interval when using the single SNP method ., For %CD8 , high posterior densities are shown for the regions around ∼170 Mb on chromosome 1 , ∼125 Mb on chromosome 2 and ∼30 Mb on chromosome 17 ( Figure 4A , C and E ) ., Some of these estimated positions agree with putative QTL region previously reported by 26 ( also see http://gscan . well . ox . ac . uk/ ) ( diamonds in Figure 5 ) ., The LR pattern from the single SNP method is similar to that from the multiple SNP method ( Figure 5B , D and F ) , but again the mapping resolution is lower ., For MCH , high posterior densities are observed for the region near ∼155 Mb on chromosome 1 , ∼82 Mb on chromosome 8 and ∼65 Mb on chromosome 14 ( Figure 6A , C and E ) ., Estimated positions agree well with putative QTL region previously reported 26 ( diamonds in Figure 6 ) ., As with the other traits , the single SNP method has lower map resolution ., Convergence of the parameter estimates was diagnosed from the pattern of the accuracy values after 100 , 1000 , 10000 and 100000 iterations when using intra-family prediction for a single replicate ., The burn-in period was 10% of the total number of iterations ., Figure 7 shows that the accuracy rapidly increases in early iteration rounds , and generally becomes a stable value after 10 , 000 iterations ., A similar pattern was observed in the inter-family prediction , i . e . the accuracy reached a stable value after ∼10 , 000 iterations ( result not shown ) , indicating that only a moderate number of iterations are required to achieve the accuracies of predicted phenotype shown in the results ., The pattern of convergence of the estimated parameters ( e . g . variances ) was similar to that of the accuracy ( result not shown ) , which was expected because accuracy was closely related to the estimated parameters ., In this study , we used only a single starting value in order to save computing time due to many different situations to be tested with many analyses ., However , for a single intensive analysis , it is always desirable to use multiple starting values to make sure that estimates reach apparent convergence ., We have proposed a method to simultaneously analyse whole genome SNP data for association with phenotypes , applied this method to three traits measured in a heterogeneous mouse stock and successfully predicted unobserved phenotypes ., The precision of the prediction of unobserved phenotypes depends on the actual genetic architecture of the traits ( heritability , number of genes , distribution of effect sizes and mode of gene action ) , the marker density and experimental sample size ., For the qualitative trait ( coat colour ) and the highly heritable quantitative trait ( %CD8 ) , the accuracies of predicting phenotypes were high , even when using genomic information from unrelated families in the same population ., This is a valuable result with important applications in medicine , agriculture and forensics ., Reversible jump theory is well established for solving model selection problem 20 , 25 , 32–34 ., We found that RJMCMC in genomic selection was computationally efficient and gave reliable estimates ., For the data set on mice ( ∼2200 individuals and ∼10 , 000 SNP ) , it took ∼15 minutes with a single CPU ( ∼2 GHz ) , which compares favourably to a number of other computing strategies on the same data set 35 ., Assuming that computing time increases linearly with the number of individuals and markers , the method would run within one week even if the data set was large ( e . g . 10 , 000 individuals with 1 , 000 , 000 SNPs ) ., More time may be required to adequately monitor convergence , however parallel computing strategies would be useful here , e . g . 36 ., Therefore , the methods described in this study can scale up to much larger data sets ., There are several approaches for whole genome association studies such as Bayesian random effect approaches 37 , ridge regression or shrinkage estimators 38 , 39 ., However , most of these approaches are computationally intensive ( as reported by 37–39 , and some statistical properties are ill-defined ( as discussed in 38 ) ., Data sets used in those studies 37–39 were much smaller than what we have used here ., Nevertheless , we recognize that improvements to our model are possible , for example using random QTL effects , and that these may lead to even better results ., Very recently , a fast analysis of all SNPs in a genome-wide association study was described using a method akin to a penalised likelihood approach 24 ., This method was implemented to find a subset of SNPs that best explains case-control status in a disease study subject to a specified type-I error rate , but can also be used to select a subset for the prediction of phenotypes ., When comparing results between different prediction strategies , the accuracies of the intra-family prediction were generally higher than those for inter-family prediction ( Table 2 ) ., There are three possible explanations for this observation ., Firstly , the prediction of phenotype within families can use both linkage ( family ) and linkage disequilibrium ( population ) information for detected gene effects , whereas the prediction across families can exploit only LD in the population ., Secondly , there may be polygenic effects which were not captured by the SNPs but these can be captured when using the phenotypes of close relatives ., Thirdly , in the data set that we used , effects due to the common environment shared by littermates are confounded with genetic effects ., Therefore , if there are such non-genetic effects that cause resemblance between relatives ( in particular fullsibs ) , then these could be partially captured by the polygenic terms and even by SNP genotype effects ., Importantly , such non-genetic common family effects do not affect the inter-family prediction ., It was shown that the difference of the accuracy of prediction with and without polygenic terms based on pedigree information was large for the intra-family prediction whereas it was much smaller for the inter-family prediction for %CD8 and MCH ( Table 2 ) ., This observation makes sense in that polygenic or common environmental effects can be informative for the prediction within families , but are not relevant for prediction across families ., For coat colour , this pattern was not evident , presumably because the phenotypes are not affected by non-genetic family effects ., Given phenotype and pedigree data , narrow- or broad-sense heritability ( h2 ) for the quantitative traits can be estimated in the classical genetic model 40 ., However , since the data set used in this study consisted of full sib families with no replicates for maternal performance of dam , maternal environmental effects or family non-genetic effects may not be well-separated from genetic effects estimated in the classical model using pedigree information ., Therefore , our estimate of heritability from the polygenic additive model may be biased upwards ., We also tried to fit epistatic effects for pairs of SNP in addition to additive and dominance effects ( see 23 for more detail on the method used ) ., However , the model including epistasis did not improve the accuracy of prediction for any trait ( results not shown ) ., This was probably because the sample size was not sufficient to capture epistatic effects or , alternatively , because epistatic interactions do not contribute much to genetic variance in our data set 41 ., We showed the strength of the multiple SNP method used in this study , compared to a set of SNPs obtained from the single SNP regression method , which is currently widely used in standard genome scans ( Figures 4 , 5 and 6 ) ., Compared to the multiple SNP method , the single SNP analyses generate more apparently significant SNPs but our results suggest that it would be much more difficult to determine the number and location of causal variants ., Both methods can provide SNP sets to predict unobserved phenotypes ., The accuracies of prediction using SNPs obtained from single SNP regression were generally lower than those with the multiple SNP method ( Table 4 ) ., This was probably due to the fact that the choice of SNPs was not optimum ., For example , selecting only one significant SNP in a region might ignore the possibility of having two QTL in a region , or alternately that multiple SNPs are required to explain the variance due to a single QTL ., In contrast , the multiple SNP method used Bayesian model selection which tested all possible models with a proper acceptance ratio according to the appropriate posterior distribution ., We used a prior of Poisson distribution with mean μn\u200a=\u200a1 for the number of QTL ( nq ) in the RJMCMC ., This might be a conservative way of detecting QTL , avoiding false positives and reducing random noise if there was no apparent prior information about the number of QTL ., We also tested the performance of the RJMCMC with a different prior which was Poisson distribution with mean μn\u200a=\u200a14 ., Note that the estimated number of QTL was ∼15 , ∼13 and ∼14 for coat colour , %CD8 and MCH , respectively ( Table 4 ) ., Table 5 shows that the average number of SNP fitted simultaneously in each RJMCMC round was much larger with a prior mean of 14 than that with a prior mean of 1 ., However , the accuracy ( correlation ) was not much different whether using a prior mean of 1 or 14 ., Although the number of SNP simultaneously fitted in each RJMCMC round was smaller when using a prior mean of 1 , all or most of the significant SNPs were found and fitted in the model over many iterations ., This is why the accuracy with a prior mean of 1 is very close to that when using a prior mean of 14 ., This agrees with conclusions from previous studies 32 , 34 , 42 that estimation of QTL positions and effects are robust with respect to prior values ., We also used a flat uniform prior for ρ ( assuming that there was no prior information for the QTL positions ) , and ML estimates for α and δ were obtained given nq and ρ ( also see online Supporting Information text S2 ) ., If there is apparent and useful information about priors , the RJMCMC can implement the information , which may give better results ., For our main RJMCMC analyses , we fixed the value of the polygenic heritability for computational reasons ., We tested the sensitivity of this procedure on the accuracy of predicted phenotypes ., For the polygenic heritability , three fixed values were compared , the previously used fixed value , half that value and a heritability of 0 ., In addition , we estimated heritability in every MCMC round ., Table 6 shows that the accuracies are not dramatically different between estimates although zero heritability , equivalent to no polygenic effect fitted , results in slightly lower accuracies ., We tested intra-family prediction only as this may be affected by value of the polygenic heritability ., Our results are based on ∼50% cross-validation ., If more than 50% of the data are used for estimation then the accuracy of prediction may improve because estimates of marker effects will be more precise ., We tested this by using 90% of the data for estimation stage and 10% for assessing the accuracy of prediction ., Because families vary in size , it is not possible to select exactly 10% of each family ., Therefore , we randomly divided the animals into 10 sets regardless of the family information ., This generates a structure intermediate between inter- and intra-family prediction ., We used 90% for estimation and 10% for validation and used 10 replicates without overlap in the validation sets ., The correlation between true and predicted phenotypes and their SD were 0 . 91 ( 0 . 02 ) , 0 . 73 ( 0 . 04 ) and 0 . 61 ( 0 . 06 ) , for coat colour , %CD8 and MCH , respectively ., The corresponding values for 50% cross-validation were 0 . 89 ( 0 . 03 ) , 0 . 73 ( 0 . 02 ) and 0 . 55 ( 0 . 02 ) ., Hence , the accuracy for coat color and MCH are higher when using 10% cross-validation than when using 50% cross-validation , but the accuracy for CD8% is not much different ., Standard deviation over 10 replicates tend to be larger with 10% cross-validation than that with 50% cross-validation ., This is probably due to the fact that 90% discovery gives better estimation of marker effects and therefore we pick up a larger correlation , but that 10% validation gives larger sampling variance for the correlation ., Our MCMC method used estimated rather than sampled values for some parameters , which is known as an empirical Bayesian approach 43 ., For a given QTL model , based on sampled values for the number of QTL , their effects and their positions , we obtained ML estimates for the remaining model parameters ., This differs from the full Bayesian approach where in the MCMC algorithm all model parameters are sampled conditional on data and other parameters ., Hence , the posterior distribution for the model parameters could differ somewhat from those of a full Bayesian approach ., The empirical Bayesian approach has a large computational advantage as for sampled values for QTL number , effects and positions , no time is wasted with evaluating all possible values of Θ but rather evaluation is at the most likely value ., Estimates converge more quickly compared to the full Bayesian approach ., It is unlikely that much information is lost in this empirical Bayesian approach because parameters in Θ have smooth distributions and it is not likely that critical information exists at values with lower probability density ., Casella 43 discussed the empirical Bayesian procedure for a hierarchical model where in an iterative procedure ML estimates were obtained for hyper parameters and other parameters were sampled conditional on these ML estimates ., He justified this procedure statistically by showing that it implies an Expectation Maximization algorithm ., In our approach , ML estimates for Θ and the likelihood of the data given the model parameters are used in RJMCMC to get the posterior density of QTL parameters across model dimensions ., The justification for our procedure is shown in 20 ., The method used here for prediction of phenotypes would be useful in many situations but the accuracy achieved is expected to vary ., The mouse population was formed from crossbreeding inbred lines and so LD is expected to exist over considerable distance ., In species with much less LD , for example humans , more markers and more phenotypic records are needed to achieve the same level of accuracy ., In conclusion , the prediction of unobserved phenotypes for complex traits from genome-wide marker data is feasible and can be accurate ., Applications of our method are plentiful: in artificial selection programs it may lead to faster response to selection , by increasing the precision with which polygenic values are predicted 16 , in human medicine it can be used to identify individuals that are most at risk for disease 17 , and in forensics it can help to build a phenotypic profile from DNA evidence . | Introduction, Methods, Results, Discussion | Genome-wide association studies ( GWAS ) for quantitative traits and disease in humans and other species have shown that there are many loci that contribute to the observed resemblance between relatives ., GWAS to date have mostly focussed on discovery of genes or regulatory regions habouring causative polymorphisms , using single SNP analyses and setting stringent type-I error rates ., Genome-wide marker data can also be used to predict genetic values and therefore predict phenotypes ., Here , we propose a Bayesian method that utilises all marker data simultaneously to predict phenotypes ., We apply the method to three traits: coat colour , %CD8 cells , and mean cell haemoglobin , measured in a heterogeneous stock mouse population ., We find that a model that contains both additive and dominance effects , estimated from genome-wide marker data , is successful in predicting unobserved phenotypes and is significantly better than a prediction based upon the phenotypes of close relatives ., Correlations between predicted and actual phenotypes were in the range of 0 . 4 to 0 . 9 when half of the number of families was used to estimate effects and the other half for prediction ., Posterior probabilities of SNPs being associated with coat colour were high for regions that are known to contain loci for this trait ., The prediction of phenotypes using large samples , high-density SNP data , and appropriate statistical methodology is feasible and can be applied in human medicine , forensics , or artificial selection programs . | Results from recent genome-wide association studies indicate that for most complex traits , there are many loci that contribute to variation in observed phenotype and that the effect of a single variant ( single nucleotide polymorphism , SNP ) on a phenotype is small ., Here , we propose a method that combines the effects of multiple SNPs to make a prediction of a phenotype that has not been observed ., We apply the method to data on mice , using phenotypic and genomic data from some individuals to predict phenotypes in other , either related or unrelated , individuals ., We find that correlations between predicted and actual phenotypes are in the range of 0 . 4 to 0 . 9 ., The method also shows that the SNPs used in the prediction appear in regions that are known to contain genes associated with the traits studied ., The prediction of unobserved phenotypes from high-density SNP data and appropriate statistical methodology is feasible and can be applied in human medicine , forensics , or artificial breeding programs . | genetics and genomics/complex traits, genetics and genomics/animal genetics | null |
journal.pgen.0030117 | 2,007 | The Unconventional Xer Recombination Machinery of Streptococci/Lactococci | Chromosome replication is a key function in living cells , and any factor that impedes progression of replication forks can result in mutagenesis and genome instability ., Several strategies have evolved to rescue replication forks stalled by DNA damage ., Most of these depend on homologous recombination pathways but are not necessarily accompanied by strand exchange 1 ., However , in cases where replication fork repair does lead to sister chromatid exchange , bacteria with circular chromosome ( s ) are faced with a potential topological problem because an odd number of crossovers between sister chromatids generates chromosome dimer , which must be converted back to monomers for a correct segregation to daughter cells ., In E . coli , chromosome dimer formation occurs in 15% of the cell population 2 , 3 , and conversion to monomers is carried out by the Xer site-specific recombination system ( for recent reviews see 4 , 5 ) ., This is composed of two paralogous tyrosine recombinases ( integrases ) , XerC and XerD , which cooperatively catalyze strand exchanges at a 28-bp DNA sequence , the dif site , which must be located at the junction of the two replichores to be functional 3 , 6–8 ., Xer recombination is intimately coupled to cell division 9 through the septal protein FtsK 10–12 , a DNA translocase 8 with an essential N-terminal transmembrane domain involved in its localization at the septum 13 , and a C-terminal DNA motor domain involved in positioning and synapsing the two dif sites of the chromosome dimer at the division septum 12 , 14–19 as well as in activating the strand exchange 8 by direct interaction with XerD 20 , 21 ., Homologs of XerCD recombinases and FtsK are found in most eubacterial phyla and some archeal lineages 22 as well as the canonical dif site 23 ., Moreover , interactions between the E . coli dif site and the XerCD recombinases of Haemophilus influenzae 24 , Pseudomonas aeruginosa 25 , Bacillus subtilis 26 , Proteus mirabilis 27 , and Caulobacter crescentus 28 have been experimentally demonstrated in vitro ., These observations led to the general view that Xer recombination is a function conserved among bacteria harboring circular chromosome ( s ) ., However , regulation of strand exchange may differ , depending on the bacterial species: FtsK-mediated activation of Xer recombination in H . influenzae obeys the E . coli paradigm 21 , whereas in B . subtilis , the model bacteria for firmicutes ( formerly known as low GC-content Gram-positive bacteria ) , neither of the two FtsK analogs ( SpoIIIE and YtpT ) appears able to drive Xer recombination 26 ., Several attempts have been made to identify the Xer recombination machinery in Streptococci , a taxonomic group belonging to firmicutes and comprising major pathogens 29 as well as innocuous food-grade species of major industrial importance 30 , 31 ., These studies revealed putative XerCD recombinases but failed to identify a dif site 32 , 33 ., We show here , by comparative genomics and functional analyses in L . lactis , S . pneumoniae , and E . coli , that Streptococci possess alternative Xer recombination machinery phylogenetically unrelated to the classical XerCD/dif system ., This machinery involves a single tyrosine recombinase ( XerS ) and an atypical dif site ( difSL ) , both located on a single genetic module ., We also show that , in contrast to B . subtilis , the streptococcal FtsK protein localizes at the division septum and controls the XerS/difSL recombination ., Assuming that Xer recombination is highly conserved in eubacteria with a significant homology of the dif sites even between distantly related species 26 , we performed an in silico search for putative dif sites in several completely sequenced firmicutes genomes ., Candidate dif sites should:, ( i ) have a significant similarity with that of B . subtilis ( difBS ) ,, ( ii ) occur only once per genome , and, ( iii ) be localized in the replication terminus ( terC ) , defined as the chromosomal region located opposite the replication origin ( oriC ) where compositional strand biases switch sign 34 , 35 ., Using these rules , a canonical dif site could be identified in all species analyzed except for Streptococci and Lactococci ( Table 1 ) ., We therefore used an alternative three-step approach based on comparative genomics to identify the streptococcal/lactococcal dif site ( Figure 1 ) ., The terC region for three streptococcal genomes was localized ( Figure 1A ) using a cumulative GC skew diagram 34 , and a comparison of the 10-kb region encompassing the GC skew shift was performed ( Figure 1B ) ., This analysis revealed a 2-kb segment that showed significant similarity within the three species ( >70% identity at the DNA level ) and included a 356-amino-acid tyrosine-recombinase–encoding gene ( annotated ymfD on the L . lactis IL1403 genome 36 but hereafter renamed xerS ) preceded by a ∼50-bp highly ( >90% ) conserved sequence ., When used to scan 49 genomes of other streptococcal species ( Figure 1C ) , this ∼50-bp fragment revealed a 31-bp consensus sequence ( hereafter named difSL ) with weak homology to the B . subtilis or E . coli dif sites , but with an overall structure resembling the DNA targets for tyrosine recombinases ( i . e . , two imperfect inverted repeats separated by a 6–8-bp central sequence ) ., Comparative analysis of the genetic context in the 10-kb terC region of different streptococcal species revealed notable features strongly suggesting that streptococcal Xer recombination machinery is defined by one genetic module , corresponding to the difSL site followed by one of its dedicated recombinases ( Figure 2 ) ., The physical link between difSL and xerS open reading frame ( ORF ) was found to be preserved among all Streptococci for which there is sequence data , and no genetic element other than difSL-xerS was conserved in the 10-kb terC region ., In addition , the genes surrounding difSL-xerS did not show a preferred orientation that might indicate possible cotranscription with the xerS gene ., Moreover , the xerS ORF often exhibits a putative ρ-independent transcription terminator at its end ., These observations indicate that xerS is unlikely to be part of an operon and suggest that the difSL-xerS pair behaves as an individual genetic module ., The candidate difSL site was tested for its ability to support site-specific recombination in L . lactis and S . pneumoniae by intermolecular recombination assays ., A 37-bp synthetic sequence encompassing the putative 31-bp lactococcal difSL site was cloned in a plasmid that does not replicate in firmicutes ( pCL52 , Table S1 ) ., The resulting construction ( pCL235 , Table S1 ) was used to transform a wild-type ( WT ) strain of L . lactis ( MG1363 , Table S1 ) ., In contrast to pCL52 , which did not yield transformant , pCL235 produced transformants at an efficiency representing 1% of the efficiency attained with a replicative plasmid ( unpublished data ) ., This demonstrates that the putative difSL site was capable of rescuing pCL235 , presumably by promoting integration of exogenous DNA into the lactococcal chromosome ., When transformed into a recA strain ( VEL1122 , Table S1 ) , plasmid pCL235 was also rescued with the same efficiency as in the WT strain ( unpublished data ) , indicating that plasmid integration occurred in a RecA-independent manner ., Moreover , as judged by Pulsed-Field gel Electrophoresis analysis ( Figure S1 , lanes 2 and 4 ) , pCL235 integrated into the chromosome of both strains at the location predicted for the native difSL site ., Thus , the 37-bp sequence appeared to contain the DNA target of a site-specific recombination system ., The difSL-mediated site-specific integration was also demonstrated to be a general process in Streptococci , since plasmid pGh9 , a temperature sensitive replication ( repAts ) mutant containing either the lactococcal 37-bp sequence described above ( pCL231 , Table S1 ) or its pneumococcal counterpart ( pCL403 , Table S1 ) , integrated into the chromosomes of L . lactis and S . pneumoniae under nonpermissive conditions with comparable efficiencies ( respectively 4 . 88 × 10−2 ± 2 . 33 × 10−2 cell−1 and 2 . 67 × 10−2 ± 1 . 55 × 10−2 cell−1 ) ., However , it should be mentioned that location of the insertion site of pCL403 has not been verified in S . pneumoniae ., The minimal size of the difSL site was determined in L . lactis by scoring the integration efficiency of pGh9 containing variants of the difSL site ( Table 2 ) ., Reducing the length of the difSL region from 48 to 31 bp did not alter the plasmid integration efficiency , indicating that the strongly conserved T located 13 bp away from the 31-bp consensus sequence ( Figure 1C ) was not part of the difSL site ., However , removing the two external bp from both sides of the 31-bp consensus sequence ( Table 2 , dif-8 ) led to a 100-fold decrease in integration efficiency , though this sequence was still proficient in site-specific recombination at the native difSL site ( Figure S1 , lanes 3 and 5 ) ., Finally , deleting two nucleotides at either side of the 31-bp consensus sequence led to a 2-fold ( Table 2 , compare dif-7 to dif-5; Wilcoxon test , p < 0 . 003 ) or 4-fold ( Table 2 , compare dif-6 to dif-5; Wilcoxon test , p < 0 . 01 ) reduction in integration frequency ., Together , these results led us to propose that the 31-bp consensus sequence defines the authentic difSL site ., Given that predictive analyses revealed XerS as the prime candidate for the actual Xer recombinase , a recombination assay was performed in S . pneumoniae to test if XerS was needed for recombination at difSL ., This streptococcal species was selected mainly for its facility to construct mutants compared to L . lactis ., The recombination at the difSL site was totally abolished in a xerS mutant ( strain S501 , Table S1; Materials and Methods ) , with undetectable integration of pCL403 , demonstrating that XerS was one catalytic recombinase of the XerS/difSL system ., To test whether XerS was the only recombinase involved in the Xer catalytic machinery , XerS/difSL recombination was assayed in E . coli using an excision assay previously developed for the genetic analysis of the E . coli dif site activity 37 ., Briefly , the native E . coli dif site was replaced by a cassette containing two directly repeated lactococcal difSL sites flanking a kanamycin resistance ( KmR ) gene ( strain E359 , Table S1 ) , and the excision frequency ( cell−1 generation−1; Materials and Methods ) was determined by counting the number of KmS recombinants at different generations during serial cultures ( Figure 3A ) ., In absence of the lactococcal XerS recombinase , almost no recombination was observed ( <0 . 006% cell−1 generation−1 ) indicating that XerCD of E . coli do not recombine difSL ., In contrast , introduction of a plasmid expressing the lactococcal xerS gene ( pCL297 , Table S1 ) increased the excision frequency to 10% cell−1 generation−1 ( Figure 3A ) , a value close to the excision frequency observed in E . coli when using the native XerCD/dif system 3 ., In addition , the excision frequency was not significantly altered in E . coli xerC or xerD mutant ( Figure 3A ) ., This indicates that fortuitous interaction between XerS and E . coli XerC or XerD recombinases is unlikely to account for recombination at difSL sites ., However , as recombination assay has not been performed on a xerC xerD double mutant , this cannot be totally ruled out ., XerS also promoted intermolecular recombination between one lactococcal difSL site replacing the native E . coli dif site on the chromosome ( strain E368 , Table S1 ) and a second difSL site located on a nonreplicative plasmid ( unpublished data ) ., Together , these data demonstrated that XerS is the only streptococcal tyrosine recombinase required to catalyze intra- and intermolecular recombination between difSL sites ., A phylogenetic analysis of all tyrosine recombinases present in the sequenced genome of five streptococcal species revealed another integrase conserved among Streptococci ., This atypical recombinase , more related to phages integrases ( Figure S2 ) and previously identified as XerD in S . pneumoniae 33 , lacks the extreme C-terminal region and two amino acids of the catalytic tetrad R-H-R-Y 38 ., When tested alone in E . coli , YnbA ( the lactococcal ortholog of S . pneumoniae XerD ) showed no intra- or intermolecular recombination activity on difSL and did not influence the recombination process when coexpressed with XerS ., Moreover , it did not bind specifically to the lactococcal difSL site in vitro ( unpublished data ) ., Therefore , YnbA is unlikely to belong to the streptococcal Xer system ., Although the XerS/difSL system involves only one recombinase , as do the Cre/loxP and Flp/FRT systems , its location at the terC region of streptococcal chromosomes strongly suggests that it functions to resolve chromosome dimers ., To examine whether XerS/difSL can substitute the XerCD/dif system in E . coli , we used the growth competition assays ( Figure 3B ) previously developed to show that XerCD/dif resolved chromosome dimers in E . coli 3 , 39 ., For that purpose , we constructed E . coli strains containing or not one lactococcal difSL site replacing the native dif site ., The strain harboring a complete streptococcal Xer system ( E368 , Table S1 ) showed a growth advantage of 10% generation−1 when competed with its isogenic strain missing the difSL site ( E367 , Table S1 ) ., As found for XerCD in E . coli 37 , this selective benefit matches the excision frequency of the KmR cassette in the excision assay described above ., In addition , strain E368 showed no growth defect compared to a strain harboring a functional XerCD/dif system ( E375 , Table S1 ) ., These results were correlated with cell morphology changes: strain E367 retained the filamentation phenotype of an E . coli strain defective in Xer recombination , while the strain harboring the complete XerS/difSL system displayed a WT cell morphology ( unpublished data ) ., Thus , the XerS/difSL system can substitute XerCD/dif in E . coli to resolve chromosome dimers ., Chromosome dimers in E . coli are mostly formed by homologous recombination 2 ., As a recA mutation also drastically reduces the XerCD-mediated recombination at dif 2 , 37 , this argued toward the fact that chromosome dimer is mandatory for creating the conditions necessary for a recombination between two directly repeated dif sites ., Such dependence was investigated for the XerS system using the same E . coli excision assay 37 ., The frequency of the KmR cassette excision by the XerS system fell from 10% to less than 0 . 6% cell−1 generation−1 ( Figure 3A ) in a recA derivative of the E359 strain ( strain E379 , Table S1 ) ., As this 20-fold reduction of recombination efficiency was similar to that observed in E . coli 37 , this strongly suggests that , as for the XerCD/dif system , XerS can perform efficient recombination only when difSL sites are located on chromosome dimers ., All streptococcal genomes sequenced so far contain one ORF encoding a protein homologous to the 787-amino-acid B . subtilis protein SpoIIIE ., These SpoIIIE-like proteins ( hereafter named FtsKSL ) range from 758 ( S . mutans ) to 816 ( S . agalactiae ) amino acids in length and retain the structural signatures of proteins from the FtsK-HerA superfamily 40: they contain a ∼180-amino-acid N-terminal region of weak similarity that includes four predicted transmembrane domains and a strongly conserved ∼500-amino-acid C-terminal region corresponding to the DNA translocase domain ( unpublished data ) ., The cellular localization of FtsKSL was determined in L . lactis using GFP fusions , corresponding to the GFP protein fused to the C-terminal of the full-length ( FtsK1 − 763 − GFP ) or N-terminal region ( FtsK1 − 181 − GFP ) of lactococcal FtsKSL ., Both GFP fusions clearly localized at the septum of L . lactis ( Figure 4 ) , indicating that as expected , FtsKSL localizes at the lactococcal division septum and its 181 amino acids containing the four transmembrane domains were sufficient to drive this localization ., The control of XerS-mediated recombination by FtsK was examined in S . pneumoniae and E . coli ., For that purpose , strains expressing FtsK proteins deleted of their C-terminal part were constructed ( ftsKC mutants , only the first 405 amino acids and the first 316 amino acids of FtsK are synthesized in S . pneumoniae and E . coli respectively; Materials and Methods ) ., Surprisingly , during the construction of the ftsKC mutants in S . pneumoniae , insertions of the Mariner minitransposon were obtained into the C-terminal or the N-terminal domain of FtsKSL ., This suggests that neither the C- nor the N terminus is essential for the growth of this bacterium , though all pneumococcal ftsK mutants were severely impaired in growth rate and cell viability as xerS mutants ( unpublished data ) ., In S . pneumoniae , XerS/difSL recombination depended on the C-terminal part of FtsKSL , because integration of the repAts plasmid containing the pneumococcal difSL site ( pCL403 ) , though not totally abolished as for the xerS mutant , became severely impaired in the ftsKC mutant ( S502 , Table S1 ) , with an efficiency of 1 . 52 × 10−4 ( ± 3 . 5 × 10−4 ) cell−1 corresponding to less than 1% of the integration efficiency of the WT strain ., Similar observations were made in the ftsKC mutant of E . coli ( E372 , Table S1 ) , using the excision assay ., The excision frequency of the difSL-KmR- difSL cassette dropped from 10% in WT strain to 0 . 1% cell−1 generation−1 in the ftsKC strain ( Figure 3A ) ., This decrease was unambiguously associated to the lack of the C-terminal part of FtsK , since expression of the full-length E . coli ftsK gene carried on a pBAD-derivative plasmid ( pCL263 , Table S1 ) restored KmR cassette excision to nearly the WT frequency ( Figure 3A ) ., In addition , results from growth competition assays ( Figure 3B ) or cell morphology observations ( unpublished data ) also showed that the XerS system was unable to resolve chromosome dimers in E . coli in absence of the C-terminal domain of FtsK ., Together our data demonstrated that XerS/difSL recombination in S . pneumoniae and E . coli , as well as dimer resolution in E . coli , depends on FtsK ., In this work , we provide experimental evidence that Streptococci possess an unconventional Xer recombination machinery that requires only one tyrosine recombinase , XerS , to catalyze the site-specific recombination at a 31-bp sequence difSL ., This raises an important question as to whether this system is orthologous to the “classical” E . coli XerCD system found in most bacterial species , including many other firmicutes ., Not only does XerS differ significantly in length and primary sequence from members of the XerCD recombinases family ( unpublished data ) , but difSL also differs in length and shows a weak similarity with the E . coli or B . subtilis dif sites ( Figure 1C ) ., Moreover , difSL is located immediately upstream the xerS coding sequence in all streptococcal and lactococcal species analyzed ., Such genetic organization contrasts with that of classical XerCD systems , with the two recombinases encoded by genes located far from each other and distant from dif , and is more comparable to the integration modules of mobile elements such as integrons 41 , bacteriophages such as P1 42 or mycobacteriophage L5 43 , and some ICEs such as the clc element from Pseudomonas 44 or CTnDOT from Bacteroides 45 ., As Streptococci and Lactococci ( together defining the taxonomic family of Streptococcaceae ) represent a rather homogeneous phylogenetic group among firmicutes when compared to other genera such as Clostridium or Lactobacillus 46 , 47 , we speculate that acquisition of the XerS system might have replaced the “classical” XerCD system at the time of or soon after their split from other firmicutes , with this event representing one landmark of this separation ., As demonstrated in this study , the cis-organization of the difSL-xerS module is not mandatory for efficient recombination , but this probably reveals a selective pressure to maintain that arrangement ., Although at present the xerS transcription start point location is unknown , we speculate that difSL either lies between the xerS ORF and its promoter or is part of the xerS promoter ., If this is true , this unusual arrangement might reflect a regulatory mechanism in which binding of XerS to difSL might autoregulate xerS expression ., Alternatively , as it has been recently observed that some filamentous phages 48 or genetic islands 49 can divert the XerCD recombination system to integrate themselves at the chromosomal dif site of several bacteria , another hypothesis could be that the difSL-xerS arrangement might serve to prevent insertion of additional genetic material at difSL , because such event would separate the xerS ORF from its promoter and lead to inactivation of the chromosome dimer resolution system ., With only one catalytic recombinase involved in the recombination reaction , the XerS system is more similar to Cre/loxP and Flp/FRT than to XerCD/dif ., However , XerS retains particular features that could indicate alternative mechanism in the recombination process ., For instance , in vivo characterization of the difSL site in L . lactis revealed an asymmetry in its organization , with left and right arms differing in length ( the left arm being 2-bp longer than the right one ) as well as in nucleotide sequence ( Table 2 ) ., This differs from loxP and FRT sites , which contain two perfectly symmetrical 13-bp arms surrounding the core region 50 , 51 ., How a single recombinase can accommodate dissimilar binding sites to perform the DNA strand exchange reaction without accessory factor has to be analyzed , but we speculate that asymmetry of the difSL site might have a role in the control of this strand exchange order ., Though we did not provide direct experimental data demonstrating that XerS/difSL is involved in chromosome dimer resolution in Streptococci , several lines of evidence strongly suggest that dimer resolution is the primary task of this site-specific recombination system ., First , classical XerCD recombinases and canonical dif site are not present in streptococcal/lactococcal genomes but substituted by the XerS/difSL recombination module at the chromosomal location predicted for a site-specific recombination system acting on chromosome dimer resolution ., Second , to catalyze the strand exchange reaction XerS seems to require at least one of the two difSL sites located on the chromosome , because recombination between two difSL sites contained within a multicopy plasmid with theta replication ( pSC101 derivative in E . coli , and pAMB1 in L . lactis , unpublished data ) could be detected neither in E . coli nor in L . lactis ., At last , not only is XerS/difSL able to resolve chromosome dimers in E . coli as efficiently as the native XerCD/dif system ( Figure 3B ) , but XerS/difSL recombination also hinged on formation of chromosome dimers , as revealed by the RecA-dependency of the KmR cassette excision ( Figure 3A ) , with the excision efficiency exactly matching the frequency of chromosome dimers resolution ., We also demonstrated that , in contrast to SpoIIIE from B . subtilis that only infrequently ( ∼6% ) concentrates at the vegetative septum 52 and is not involved in the Xer recombination 26 , the streptococcal FtsKSL protein localizes at the division septum and still directs the XerS/difSL recombination , as dimer resolution and intra- or intermolecular recombination were drastically inhibited in E . coli and S . pneumoniae cells lacking the C-terminal part of FtsK ., Although our preliminary analyses of the pneumococcal ftsK mutants need to be confirmed , the ability to obtain viable cells depleted of FtsK suggests that neither the N- nor the C terminus of the protein is essential in Streptococci ., As essentiality of FtsK seems to be species dependent , with only the N-terminal part in E . coli and the C terminus in C . crescentus being essential 53 , 54 , we hypothesize that activity of FtsK , though still involved in cell division and DNA translocation , could slightly differ among the different bacteria ., However , it appears that no correlation can be done between essentiality and localization of FtsK to the division septum ., In E . coli and some other γ-proteobacteria , the C-terminal part of FtsK drives the XerCD recombination reaction in two ways: by participating to the formation of the recombination synapse through its DNA translocase activity 14–16 and activating the recombination reaction by direct interaction with XerD 8 ., Some of our data strongly indicated that such interaction between FtsKSL and XerS is unnecessary to activate the XerS/difSL recombination in Streptococci , though this cannot be totally ruled out ., First , the XerS-mediated intramolecular recombination frequency at difSL in E . coli ( Figure 3A ) was close to that measured with XerCD/dif 3 , suggesting no species specificity for FtsK requirement ., This observation contrasts to that made in E . coli where the H . influenzae FtsK was inefficient in activation of the E . coli XerD and vice versa , implying that the FtsK-XerD interaction is highly species specific in these bacteria 21 ., In addition , both pneumococcal and lactococcal XerS protein sequences do not contain the amino acids motif ( RQ–QQ ) found in E . coli XerD and involved in its specific interaction between with FtsK 20 ., At last , the cassette excision by recombination at difSL in E . coli , as well as plasmid integration in S . pneumoniae , was not totally abolished in ftsKC mutants but dropped to 1% of the recombination activity of WT strains , suggesting that some productive recombination synapses might form independently of FtsK , most probably by the random collision of two dif sites ., This observation also contrasts the results obtained with the cassette excision assay performed with the E . coli XerCD/dif system , wherein no recombination was detected in an ftsKC mutant 55 , suggestive of the FtsK-mediated activation of the recombination ., Our data are more easily accommodated to a model where XerS is unable to form a productive synapse and requires the DNA translocase activity of FtsKSL to bring the two difSL sites of a chromosome dimer close to each other and in an active geometry before performing the strand exchange ., However , the recombination would not need direct activation by protein interaction between FtsKSL and XerS ., However , as for the XerCD model 11 , our model cannot provide satisfactory explanation to how FtsK is involved in the intermolecular recombination between one difSL site located on a suicide ( or repAts ) plasmid and the chromosomal difSL site , and the mechanism of the FtsKSL-mediated control has to be analyzed further ., In conclusion , the discovery of a Xer recombination system phylogenetically unrelated to the classical XerCD system reinforces the idea that chromosome dimer resolution can be viewed as a housekeeping function conserved among bacteria with circular chromosome ( s ) , but that some species can use functional analogs to perform this task ., We expect that other bacterial species among those whose genome ( s ) are missing a canonical dif site also contain alternative chromosome dimer resolution systems ., Finally , we note that the particularity of the XerS system makes it a valuable candidate for the development of new antibacterial drugs specifically directed against the pathogenic Streptococci ., The plasmids and bacterial strains used in this study are listed in Table S1 ., E . coli strains and plasmids containing synthetic lactococcal or pneumococcal variants of difSL sites were constructed using the procedure provided in Text S1 ., E . coli strains were grown at 30 °C in LB medium ., Antibiotics were used at the following concentrations: erythromycin ( Em ) 150 μg ml−1 , chloramphenicol ( Cm ) 20 μg ml−1 , spectinomycin ( Spc ) 100 μg ml−1 , kanamycin ( Km ) 50 μg ml−1 , and ampicillin ( Ap ) 25 μg ml−1 ., L . lactis strains were grown semi-anaerobically at 30 °C in M17 broth ( Merck KGaA , http://www . merck . de ) supplemented with 0 . 5% ( w/v ) glucose ( GM17 ) and transformed as previously described 56 ., Antibiotics used for selection of lactococcal transformants were: Em 1 μg ml−1 , Cm 5 μg ml−1 , and Spc 200 μg ml−1 ., S . pneumoniae strains were grown in Todd-Hewitt broth ( Difco/BD Biosciences , http://www . bdbiosciences . com ) supplemented with 0 . 5% yeast extract ( THY ) and transformed using synthetic competence-stimulating peptide ( CSP ) as described 57 ., Antibiotic concentrations used for selection of pneumococcal transformants were: Em 0 . 2 μg ml−1 and Km 500 μg ml−1 ., Restriction and modification enzymes were purchased from New England Biolabs ( http://www . neb . com ) and used as recommended by the supplier ., Plasmid DNA from E . coli was isolated with the Qiaprep spin kit according to the manufacturers instructions ( Qiagen , http://www . qiagen . com ) ., Chromosomal DNA from E . coli , L . lactis , and S . pneumoniae was isolated with the DNeasy tissue kit according to the manufacturers instructions ( Qiagen ) ., Preparation of lactococcal genomic DNA embedded in agarose matrix , Pulsed-Field gel Electrophoresis , and Southern blot with dried agarose gels were performed as previously described 58 ., Hybridization signals were detected with a Bioimaging BAS1000 analyzer system ( FUJI Photo Film , http://www . fujifilm . com ) and analyzed with TINA version 2 . 07c software ( Raytest Isotopenβgeräte GmBH , http://www . raytest . de ) ., Nucleotide sequences were obtained from NCBI ( http://www . ncbi . nlm . nih . gov/genomes/static/eub_g . html ) , JGI ( http://genome . jgi-psf . org/tre_home . html ) , and the Sanger Institute ( http://www . sanger . ac . uk/Projects/Microbes ) ., Cumulative GC skews were obtained from an in-house build program ( Laurent Lestrade , Laboratoire de Biologie Moléculaire des Eucaryotes , Toulouse , France ) ., Multiple DNA comparison was performed using MultiPipMaker program 59 ., Chromosomal integration assays of repAts plasmid pGh9 or difSL-containing derivatives in L . lactis were performed according to 60 ., For S . pneumoniae , frozen strains containing the pGh9 or derivatives were grown at 39 °C ( water bath ) to an OD600 = 0 . 3 in THY without antibiotics ., Appropriate dilutions were plated on 10 ml of D medium 57 containing 2% of defibrinated horse blood ( bioMérieux , http://www . biomerieux . com ) and supplemented when appropriate with Em , and plates were incubated at 39 °C ., Integration of the pGh9 plasmid was undetectable in S . pneumoniae ( no colonies were observed when plating 0 . 1 ml of the undiluted bacterial culture ) ., The integration frequency per cell ( ipc ) was calculated as the geometric mean of the ratio of colonies on selective versus nonselective plates obtained from five to 19 independent experiments ., Mutagenesis was carried out as described 61 ., The target DNA for in vitro transposition of the KmR mariner minitransposon pR410 61 were obtained by PCR reactions using R800 chromosomal DNA as template ., The sizes of PCR fragments were: 2 . 012 bp for xerS gene ( primers: forward , 5′-TAgAAAACCgATTCTCAgAAATgAgATC-3′; reverse , 5′-gAAgAAgAATTggCCgA AATCAA-3′ ) and 4 . 053 bp for ftsKSL gene ( primers: forward , 5′-AAAACAAAgCCTTggTggTgC CT-3′; reverse , 5′-CTTgCgACAAgAAAgggAAA-TTT-3′ ) ., The mutagenized PCR fragments were then transformed into strain R800 ., For each mutagenesis , ten KmR transformants were checked by PCR and shown to carry a mariner insertion ., The accurate insertion position of the transposon , as well as its orientation , was determined by PCR and DNA sequencing as described 61 ., The resulting chromosome structures of the selected mutants were: R800 xerS , insertion of mariner 167 bp downstream the ATG ( insertion allows the synthesis of only the fir | Introduction, Results, Discussion, Materials and Methods, Supporting Information | Homologous recombination between circular sister chromosomes during DNA replication in bacteria can generate chromosome dimers that must be resolved into monomers prior to cell division ., In Escherichia coli , dimer resolution is achieved by site-specific recombination , Xer recombination , involving two paralogous tyrosine recombinases , XerC and XerD , and a 28-bp recombination site ( dif ) located at the junction of the two replication arms ., Xer recombination is tightly controlled by the septal protein FtsK ., XerCD recombinases and FtsK are found on most sequenced eubacterial genomes , suggesting that the Xer recombination system as described in E . coli is highly conserved among prokaryotes ., We show here that Streptococci and Lactococci carry an alternative Xer recombination machinery , organized in a single recombination module ., This corresponds to an atypical 31-bp recombination site ( difSL ) associated with a dedicated tyrosine recombinase ( XerS ) ., In contrast to the E . coli Xer system , only a single recombinase is required to recombine difSL , suggesting a different mechanism in the recombination process ., Despite this important difference , XerS can only perform efficient recombination when difSL sites are located on chromosome dimers ., Moreover , the XerS/difSL recombination requires the streptococcal protein FtsKSL , probably without the need for direct protein-protein interaction , which we demonstrated to be located at the division septum of Lactococcus lactis ., Acquisition of the XerS recombination module can be considered as a landmark of the separation of Streptococci/Lactococci from other firmicutes and support the view that Xer recombination is a conserved cellular function in bacteria , but that can be achieved by functional analogs . | In bacteria , genetic information is mainly carried by a single circular chromosome ., The replication of this circular molecule sometimes leads to the formation of a chromosome dimer unable to segregate in the daughter cells during the division process ., In the bacterial model E . coli , chromosome dimers are resolved in monomers by site-specific recombination: two recombinases ( XerC and XerD ) cooperatively catalyze the recombination at a chromosomal site ( dif ) , located at the junction of the two replication arms ., This recombination is intimately coupled to cell division by the control of the septal protein FtsK ., Xer recombination machinery as described in E . coli appears highly conserved among bacterial species ., We show by comparative genomics and genetic studies that Streptococci use an alternative Xer recombination system , renamed XerS/difSL , which is composed of a single recombinase phylogenetically unrelated to XerCD proteins and a noncanonical dif site ., We also demonstrate that the streptococcal FtsK protein localizes at the division septum and operates the XerS/difSL recombination ., This is the first identification of an alternative Xer recombination system in prokaryotes to out knowledge , which might indicate that other unusual chromosome dimer resolution systems could exist in bacterial phyla where a canonical dif site is not detected . | molecular biology, genetics and genomics, microbiology, eubacteria | null |
journal.pcbi.1003956 | 2,014 | Epidemic Spreading Model to Characterize Misfolded Proteins Propagation in Aging and Associated Neurodegenerative Disorders | Misfolded proteins ( MP ) are associated with aging processes and several human neurodegenerative diseases 1–3 ., The prion-like hypothesis explains the neurodegenerative progression by the intercellular transfer of pathogenic proteins 4–6 , under the perspective that MP behave like infectious-like agents that propagate from a few initial host regions to other brain regions ., For instance , in Alzheimers disease ( AD ) , soluble Amyloid-ß ( sAß ) oligomers are thought to be the principal seeds that carry the misfolding process from region to region , accelerating the production/deposition of new misfolded proteins 7–9 and thus contributing to drive the pathology to new areas of the brain 10 , 11 ., The associated Aß toxicity has a relevant impact on AD development and progression 12–18 ., The cell-cell transference is possible because sAß oligomers are very small assemblies of MP , which can be absorbed by axonal processes and transported to cell bodies , causing cytotoxicity in the receiving cells 10 , 11 , 19 ., Also , sAß oligomers that are immersed in the extracellular fluid are subjected to the principles of molecular diffusion processes in the brain , i . e . a highly anisotropic movement along the axis of nervous fibers 20 ., Consequently , sAß propagation , and the subsequent Aß deposition and cytotoxicity effects , occurs mainly between anatomically interconnected areas or between neighboring neuronal cells 10 , 11 , 21 , 22 ., Neuropathologic evidence supports the idea that each neurodegenerative disorder is linked to the misfolding of a specific protein or group of proteins 5 , 23–25 ., Thus , the network degeneration hypothesis proposes that misfolded proteins mechanisms should present disease-specific anatomical patterns 26–29 ., Two recent studies showed that specific functional and structural covariance subnetworks of the healthy brain are in correspondence with the spatially dissociable cortical atrophy patterns of five distinct dementia syndromes 27 , 30 ., The reported link between structural/functional brain connectivity patterns and neurodegenerative damage supports the network degeneration hypothesis ., This also emphasizes the strategic importance of developing molecular pathological approaches capable of reproducing MP propagation , which might not only be conducive to a better understanding of MP spreading factors , but could also help to evaluate their contribution to disease progression in relation with other postulated pathological mechanisms ( e . g . the neuronal activity dependent degeneration 31–33 ) ., In this context , a Network Diffusion Model of disease progression in dementia was proposed 34 , where the pathogenic proteins propagation follows the regional concentration gradients under the spatial constraints defined by the brains connectional anatomy ., Consistent with their theoretical predictions , the authors found that specific anatomical sub-modules are in correspondence with characteristic cortical atrophy patterns in AD and behavioral frontal temporal dementia ., However , the ability of this model to replicate real MP propagation/deposition patterns remained unexplored ., A potential limitation of this model is that it does not consider possible defense mechanisms of the brain ., Rather , the disease factors can accumulate gradually , without system resistance , while inducing cellular death and cortical atrophy ., Conversely , immunologic brain responses have been demonstrated to combat MP accumulation 35–38 ., For instance , Aß clearance by macrophages and microglia cells are responsible in part for the remarkable fluctuations in neurological functions that AD patients present even during the same day 26 , 35 , 39 ., Furthermore , recent evidence indicates that initial Aß related processes could have a protective role on the nervous system 40 , 41 , which suggests non Aß related neurodegenerative effects ( e . g . cellular death and cortical atrophy ) at all the Aß propagation states but only after an abnormal accumulation process ., Considering the relevance that both intercellular MP transfer and associated clearance defenses have toward the development of neurodegenerative disorders , here we proposed a stochastic epidemic spreading model ( ESM ) to describe the dynamic interactions between MP infectious-like agents and the brains clearance response ., The validity/applicability of the proposed hypothesis and model was explored using 733 individual PET Aß datasets from the Alzheimers Disease Neuroimaging Initiative ( ADNI ) ., We found that the ESM is able to reconstruct individual/group Aß deposition patterns ., Most importantly , ESM predicts that it is not an increased Aß production but mainly a deficit in Aß clearance processes and an early Aß onset age that result in the formation of an excessive Aß deposition pattern , and in the conjectured acceleration of the preceding tauopathy ., Additionally , our results highlight the strategic role of the MP outbreak regions and their connectional architecture on the diseases temporal progression , as well as the impact of individual genetic and demographic properties on intra-brain Aß propagation ., We developed a stochastic epidemic spreading model ( ESM ) to describe intra-brain Aß propagation and deposition processes ( Methods section ) ., Then , we proceeded to explore the ability of the model to reproduce Aß deposition patterns in healthy and pathological brains ., Figure 1 illustrates the key processing steps of our approach ., First , we used Florbetapir ( 18F-AV-45 ) PET data to quantify Aß deposition patterns in a cohort of 733 subjects with non-Hispanic Caucasian ancestry ( Table S1 ) from the ADNI database ( Methods , Study participants , Dataset 1 ) ., Each participant was previously diagnosed as healthy control ( HC , n\u200a=\u200a193 ) , early mild cognitive impairment ( EMCI , n\u200a=\u200a233 ) , late mild cognitive impairment ( LMCI , n\u200a=\u200a196 ) or probable AD ( n\u200a=\u200a111 ) ., For each subject , the baseline 18F-AV-45 PET scan was employed to calculate the Aß deposition probabilities for 78 regions covering all the gray matter 42 , and these were used to define the individual Aß deposition pattern ( Methods , Regional Aß deposition patterns ) ., Next , we used the developed ESM , and region-region anatomical connectivity information from 60 healthy young subjects ( Methods , Study participants , Dataset 2 ) , to generate multiple hypothetical regional courses of Aß propagation/deposition ., Each hypothetical generation corresponded to a specific set of sAß spreading seed regions , up to a maximum of 6 regions consisting of all possible combinations , and a set of model parameters from which we simulated 50 years of propagation starting with Aß presence only in the seeds ., A selective iterative algorithm ( Methods , Model exploration/validation ) was used to identify the seed regions that better explained the PET-based Aß deposition patterns across the study cohort , as well as the individualized model parameters that maximized the similarity between the generated and the individual reference Aß deposition patterns ., In sum , a set of the most likely Aß outbreak regions were identified , assuming the same set of regions for the whole sample , whereas for each subject four different model parameters were estimated: Aß production rate ( ) , Aß clearance rate ( ) , onset age of Aß outbreak ( Ageonset ) , and model noise level ( σ ) ., For further details see Methods ( Model exploration/validation subsection ) and Figure 1 ., Consistent with the hypothesis of an intra-brain Aß epidemic spreading behavior , our propagation/deposition model reproduced , from the remote non-binding states , the characteristic Aß deposition patterns in the adult cohort ( Figures 2A , B ) ., It explained between 46 . 4∼56 . 8% ( all P<10−10 ) of the variance in mean regional Aß deposition probabilities ( adjusted by age , gender , and educational level ) in HC , EMCI , LMCI and AD groups ., See Table S2 for a comparison with previous approaches ., In addition , it identified the posterior and anterior cingulate cortices as the most probable starting seed regions for the Aß propagation process ( see Table S3 for examples of other tested combinations of regions , based on previous reports ) ., The cingulate cortex , particularly its posterior area , is considered a core node of the default mode network ( DMN ) , and is thought to be involved in self-relevant/affective decisions , mental simulation , and integration tasks 43 , 44 ., This result is in agreement with the large amount of evidence suggesting the critical role of the DMN on the genesis and propagation of AD 31 , 45 , 46 ., For a complementary seeds identification analysis , see Discussion section ( Identification of the MP propagation epicenter subsection ) and Figure S1 ., Next , we re-evaluated the competence of the ESM framework to reproduce prion-like spreading mechanisms , but now based on the idea that , if the ESM is describing real intra-brain propagation of MP , then alterations in the structural connectional information should affect the models results negatively ., We tested this by comparing the capability of the ESM to explain advanced Aß deposition patterns , using the available connectivity information and alternatively using “non-informative” connectional information ., For this , 100 randomized versions of the original anatomical connectivity matrix were created ( preserving its weight , degree and strength distributions 47 ) , and the propagation model was evaluated for each of these versions ., We observed a significantly higher model competence ( all P<10−5 ) to explain the Aß deposition patterns when the original anatomical connectional information was used ( Table S4 ) ., This result supports the ability of the SEM to describe real MP spreading processes , based on the central interrelation between biological factors directly related to these pathogenic proteins ( e . g . Aß production and clearance ) and the complex connectional architecture of the human brain ., Finally , the statistical robustness and predictive power of the introduced ESM was tested via a repeated random sub-sampling cross-validation ., Each clinical group ( HC , EMCI , LMCI and AD ) was randomly split into training and test data of the same size ., For each such split , the model values derived at the group level for the training data were used to test the predictive validity of the model on the validation group ., We observed significant predictive power across the different clinical states ( Figure 2C ) , with prediction accuracy values ranging from 40 . 7% ( 95% CI: 36 . 3 , 45 . 0 ) for the HC group to 31 . 4 ( 95% CI: 28 . 4 , 34 . 2 ) for the AD group ( Table S5 ) ., Slightly lower prediction accuracy was observed for the AD group ., We attribute this to the smaller sample size , in comparison with the other groups , and as will be analyzed in the next subsections , to a larger period of Aß propagation/deposition processes ( with a significantly earlier propagation onset ) ., This larger period of the phenomenon to be modeled can be consequently associated to a larger accumulation of model errors ., Historically , the identification of outbreak nodes has been considered a primary step towards the spatiotemporal understanding of epidemic phenomena 48 ., In the context of brain neurodegenerative disorders , functional proximity to epicenter regions implies greater disease-related regional vulnerability 30 ., This suggests an organized pattern for propagation of disease agents , in accordance with the trans-neuronal network-based MP spread hypothesis , and supports the key role of specific epicenter regions in the disease progression processes ., Those results were obtained using an indirect measure of MP presence , i . e . gray matter atrophy quantified using voxel-based morphometry ., However , the relation between gray matter atrophy and MP effects is still unclear , and , in addition , the former can also be caused by multiple different factors ( e . g . , vascular dysregulation ) ., To obtain direct evidence of MP dispersion as a function of proximity to an epicenter , we first explored the relation between the PET-based regional Aß deposition patterns and the effective anatomical distances to the identified Aß outbreak regions ( anterior and posterior cingulate cortices; Methods , Model exploration/validation ) ., We observed a significant negative linear relationship between these two variables , across the four clinical states ( Figure 3A ) ., Interestingly , best-fit lines for the different clinical groups displayed a consistent co-linearity ( Figure 3A and Figure S2 ) ., The relationships were characterized by similar slope but different Aß deposition intercepts that increase according to disease progression ., We verified that these associations are not explainable by the spatial proximity between regions ( Table S6 ) ., These results support the role of the outbreak regions as centers of radial disease factor propagation , which is modulated by the brains connectional architecture ., Next , we used the spatiotemporal information provided by the ESM to analyze the link between regional Aß arriving times ( ) and the effective anatomical distances ., For each brain region i , was calculated as the time at which the Aß probability deposition of this region reached a given threshold ( e . g . <0 . 9 implying no deposition , ≥0 . 9 implying deposition ) ., In line with the previous results , we found a significant linear predictive relationship between the effective anatomical distances and the values ( Figure 3B ) ., The shape of this relationship was invariant to the selection of different Aß deposition thresholds ., Notably , these results correspond with the linear predictive relationship reported for effective distances in human social networks and disease arrival times for real epidemics propagation data 49 ( e . g . 2009 H1N1 pandemic ) ., This parallelism between intra-brain MP propagation mechanisms and epidemic propagation in human disease networks 49 , supports our hypothesis of an intra-brain epidemic spreading behavior of MP propagation ., Furthermore , these model-based findings clarify the distance-vulnerability effects observed for gray matter atrophy 30 and Aß deposition ( Figure 3A ) ., In terms of regional vulnerability to disease pathological effects , recent studies have also suggested a direct link between structural/functionally connectivity levels and regional vulnerabilities 31 , 50 , 51 ., Highly connected brain regions are usually known as “hubs nodes” of the brain network ( for review see 50 ) ., Buckner et al . , 2009 , showed a high correspondence between Aß deposition levels and functional connectivity in the brain hubs ., Further evidence , based on meta-analyses of published magnetic resonance imaging data about 26 different brain disorders , suggest that pathological brain lesions ( i . e . gray matter atrophy lesions ) are mainly concentrated in structural hub regions , independently of the studied disorder 51 ., This fact is considered a consequence of the high topological centrality and biological cost of the hubs , which make them more vulnerable to a diverse range of pathogenic processes 31 , 51 ., In order to explore if the introduced ESM can clarify this connectional-pathogenic association , we analysed the relation between regional anatomical connectivity degrees and Aß arrival times , as measures of hubness ( Methods , Anatomical connection probability ) and temporal vulnerability to receive aberrant disease factors , respectively ., We observed significant negative correlations ( all P<10−9 ) between these two variables , independently of the selection of different Aß deposition thresholds ( see Figure S3 ) ., This suggests that regions with higher anatomical connectivity degrees experience early Aß arrival and , consequently , larger periods of exposition to the negative effects of this aberrant protein ., For decades , Aß propagation and accumulation has been thought to have a causal role on the cascade of cognitive/clinical events leading to AD 52 , 53 ., For instance , Aß toxicity has been causally associated with brain oxidative stress 14 , 18 , mitochondrial dysfunction 18 , synapse and spine loss 13 , widespread neuronal dysfunction and cell death 12 , synaptic plasticity and memory impairment 16 , 17 ., To test the potential clinical impact that progressive Aß presence can have on the pathologys progression , we studied whether model variables controlling intra-brain Aß propagation/deposition are related to AD and intermediate cognitive/clinical states ., For this , we considered the clinical diagnosis ( HC , EMCI , LMCI or AD ) as a dependent variable in a Multinomial Logistic Regression model with Aß production/clearance rates , noise and onset age as independent variables ( controlling by gender , age and educational level ) ., Then , the contribution of each regressor was evaluated using a robust metric of relative importance in prediction analysis ( Methods , Statistical Analysis ) ., We observed a statistically significant relationship between clinical diagnosis and Aß clearance rate and onset age ( Figure 4 and Table S7 ) ., The clearance rate was found more related to the clinical diagnosis than the other model parameters ( Figure 4A ) , explaining 8 . 45% ( 95% CI: 4 . 88 , 11 . 89 ) of its inter-subject variance ., A closer look at the differences between the four clinical groups ( Figures 4B-E and Table S8 ) , revealed that the clearance rate is also the model variable with the most consistent variance across the different clinical diagnoses ., It is followed by the onset age of Aß accumulation , which reflects a decreasing transition from HC to EMCI-AD states ( Figure 4D ) and explains 6 . 77% ( 95% CI: 3 . 42 , 9 . 88 ) of the variance in clinical diagnoses ., We observed a significant decreasing trend on the Aß production rate from HC to EMCI-AD states ( Figure 4B ) , however the impact of this effect on the clinical diagnosis was not significant ( 95% CI: −0 . 76 , 3 . 77; for further analysis , see Text S1 ) ., While the individualized global Aß production rate can be seen as a measure of the lifetime individual regional potential to produce Aß infectious-like factors , the corresponding clearance rate reflects the lifetime intrinsic capacity to combat the Aß accumulation ., Therefore , these results suggest that although a significantly earlier onset age of Aß accumulation and a non-significantly lower production rate of Aß agents are associated to AD , a deficiency associated to Aß clearance might be the most determining Aß-mediated factor for the development of the disease ., Apolipoprotein E ( APOE ) e4 genotype is considered a relevant genetic risk factor for AD and intermediate MCI states 54 ., It has been associated to Aß aggregation into fibrils 55 , the hindered clearance of sAß 56 , and neurodegeneration by directing toxic Aß oligomers to synapses 54 , 57 ., Using our ESM , we explored how different APOE e4 genotypes impact Aß propagation and deposition ., For each model parameter , we performed a three-way ANOVA test considering as grouping parameters the number of APOE e4 allele copies , as well as the gender and educational level of the participants ., We observed a significant effect of APOE e4 genotype on the Aß production/clearance rates and on the onset age ( Figure 5A and Table S9 ) ., In particular , we found that APOE e4 genotype had highest impact on the onset age , decreasing it proportionally to the number of APOE e4 allele copies ( Figures 5A , E ) , and explaining 13 . 21% of its inter-subject variance ( P\u200a=\u200a1 . 12×10−24 , F\u200a=\u200a59 . 57 ) ., This result is in line with previous reports associating APOE e4 genotype with an earlier age at disease onset and a faster AD pathological progression 58 , 59 ., In addition , we observed a significant decrease in Aß clearance rate with regard to the number of APOE e4 allele copies ( Figures 5A , C ) , explaining 10 . 48% of its inter-subject variance ( P\u200a=\u200a2 . 24×10−19 , F\u200a=\u200a45 . 60 ) ., This supports our previous result associating AD onset with an Aß clearance deficiency and , more importantly , evidences that this clearance deficiency partly has a genetic cause 56 ., We also found significant effects of APOE e4 on Aß production rate ( Figures 5A , B; P\u200a=\u200a5 . 38×10−19 , F\u200a=\u200a21 . 98 ) , which reflects the multi-factorial influence of this genotype on the evolution of AD and intermediate MCI states 54–57 ., Further statistical analyses were performed to assess how the specific number of APOE e4 allele copies impact on Aß propagation and deposition ( Table S10 ) ., We found that the effects due to the presence of two e4 allele copies are more relevant ( in terms of the model parameters ) than the effects due to the presence of only one copy ( Figures 5B–E and Table S10 ) ., This is in agreement with the reported semi dominant inheritance effect of APOE genotype on developing AD 60 ., When investigating the relationship of the model parameters with the demographic variables , we also found a significant impact of gender on Aß production rate ( P\u200a=\u200a1 . 90×10−3 , F\u200a=\u200a9 . 68 ) , Aß clearance rate ( P\u200a=\u200a1 . 35×10−3 , F\u200a=\u200a19 . 19 ) and Aß onset age ( P\u200a=\u200a2 . 06×10−3 , F\u200a=\u200a36 . 84 ) ( Tables S9 and S11 ) ., For all these cases , female gender was associated with significantly lower model parameter values ( Table S11 ) ., This result is in high correspondence with the fact that women are more likely to develop AD than men 61 , 62 ., Furthermore , we found a significant interaction between APOE e4 genotype and gender , which together are modulating the Aß onset age ( P\u200a=\u200a1×10−5 , F\u200a=\u200a9 . 30 ) ., This is consistent with the higher propensity for women to develop AD across most ages and APOE genotypes 62 , with the most pronounced detrimental effect of APOE e4 on DMN connectivity and CSF tau levels 61 , and with the reported greatest amyloid plaque and neurofibrillary tangle pathology for women 63 ., Finally , when investigating the relationship of the noise parameter σ with APOE e4 and the demographic variables ( Table S9 ) we found that female subjects with a higher educational level have a higher noise level ( P\u200a=\u200a0 . 019 , F\u200a=\u200a5 . 45 ) ., In conjunction with a significant impact of gender and educational level on the Aß onset age ( P\u200a=\u200a0 . 01 , F\u200a=\u200a5 . 45 ) and a non-significant trend effect of educational level on Aß clearance rate ( P\u200a=\u200a0 . 093 , F\u200a=\u200a2 . 82 ) , this may be reflecting the complex relationship that exists between Aß binding , gender , cognitive reserve and clinical state 64 ., The larger variability in Aß deposition patterns associated with higher noise , gender and educational level , could explain the disputed results of the cognitive reserve hypothesis 65–67 ., CSF measures of Aß , total tau ( t-tau ) and phosphorylated tau ( p-tau ) are considered the most relevant early biomarkers of AD 68 , 69 ., Although Aß and tau proteins were historically considered to arise and act independently , now it is thought that both proteins are strongly interrelated 13 ., Based on different converging evidences , it has been suggested that Aß pathophysiology might drive and accelerate pre-existing tauopathy 70 ., Here , we aimed to re-evaluate this interrelation hypothesis under the assumption that , if the intra-brain ESM of Aß propagation/deposition can reflect Aß pathophysiology accurately , then abnormalities in CSF Aß , t-tau and p-tau concentrations should be correctly reflected by the individualized model parameters ., For this , we used CSF Aß1-42 , t-tau and p-tau181 measures from a subsample of 307 healthy and non-healthy subjects belonging to the 18F-AV-45 PET scanned group ( Methods , CSF measures ) ., For each CSF measure , we performed a seven-way ANOVA test , considering the model parameters , age , sex and educational level as modulatory factors ., The results ( Figure 6A and Table S12 ) show a significant impact of Aß production/clearance rates on CSF Aß1-42 , explaining 10 . 40% ( P\u200a=\u200a1 . 24×10−12 , F\u200a=\u200a55 . 02 ) and 11 . 85% ( P\u200a=\u200a4 . 83×10−14 , F\u200a=\u200a62 . 66 ) respectively , of its across-subject variance ( see also Text S1 ) ., We also found that the Aß onset age and the chronological age are significant modulators of CSF Aß1-42 , explaining 2 . 31% ( P\u200a=\u200a5 . 36×10−4 , F\u200a=\u200a12 . 24 ) and 2 . 97% ( P\u200a=\u200a9 . 29×10−5 , F\u200a=\u200a15 . 69 ) respectively , of its variance ., Together , all considered modulators accounted for 28 . 82% of the CSF Aß1-42 variance ., Aß production/clearance rates were also found to have significant impact on CSF t-tau , explaining 4 . 45% ( P\u200a=\u200a1 . 68×10−5 , F\u200a=\u200a19 . 33 ) and 2 . 77% ( P\u200a=\u200a6 . 32×10−4 , F\u200a=\u200a11 . 93 ) respectively of its variance ., However , in this case the higher impacts correspond to the Aß onset age and chronological age ( Figure 6B ) , with 5 . 08% ( P\u200a=\u200a4 . 41×10−6 , F\u200a=\u200a21 . 87 ) and 5 . 45% ( P\u200a=\u200a2 . 07×10−6 , F\u200a=\u200a23 . 44 ) respectively , of explained variances ., Similar effects were observed for CSF p-tau ( Figure 6C ) , for which Aß onset age was the strongest modulator and accounted for 4 . 44% ( P\u200a=\u200a2 . 07×10−5 , F\u200a=\u200a18 . 23 ) of its variance ., According to these results , while Aß production/clearance rates might be influencing the deposition and recirculation of Aß and subsequently its inter-relationship with tau proteins , the observed Aß onset age and chronological age effects on t-tau and p-tau may be reflecting the time duration of such inter-relationship ., These results are consistent with the idea of an interrelated pathway between amyloid pathophysiology and tauopathy 70 , 71 and , in combination with results from the previous subsections , they are also consistent with the notion of an associated failure to clear mislfolded proteins 70 , 72 ., The prion-like hypothesis explains the neurodegenerative progression by the intercellular transfer of pathogenic factors 4 , 73 ., This perspective presents a striking similarity with the spread of real infectious diseases in biological populations ., Social networks constitute a common structural substrate over which infectious factors propagate , reaching in some cases an epidemic/uncontrollable behavior 74 ., Independently of the pathogenic agents characteristics , its propagation dynamics are always constrained by the connectivity structure of the attacked system ., It is in this context that we hypothesized the Aß proteins propagation and deposition as a natural epidemic spreading event , whose dynamics are determined by infectious-like agents and immunologic response actions that compete under a restrictive anatomical network ( the structural human connectome ) ., Note , however , that the term infectious does not necessarily imply the presence of fully negative propagating factors , since the genesis and role of MP in the brain are not completely understood 40 ., Previous studies have used the brains structural and functional connectivity to explain neurodegenerative atrophy patterns ( for recent reviews see 75 , 76 ) ., We extended previous connectivity-based approaches 27 , 30 , 34 by combining pathogenic factors actions ( production and spreading ) with possible defense responses , including also the influence of stochastic or undetermined processes ., The inclusion of basic biological variables ( e . g . MP production/clearance rates , time of propagation ) provides a more realistic characterization and understanding of the studied phenomenon , allowing not only to reproduce the MP dynamics but also to identify the genetic , structural , and demographic factors associated to it ., For purposes of comparing different methods , we applied the Network Diffusion Model ( NDM ) 34 to the same Aß datasets and connectivity information ( for further details see Table S2 ) ., We found that NDM also identified the posterior and anterior cingulate cortices as the most probable starting seed regions for the Aß propagation process ., However , even when the obtained mean regional explained variance for the NDM was around 27–33% , with a significant statistical association ( p<0 . 05 ) , the corresponding root mean square errors ( RMSEs ) were considerably high , reflecting large absolute differences between estimated and reference Aß concentration patterns ., In addition , Akaike Information Criterion ( AIC ) values evaluated for both models ( ESM and NDM ) revealed a significantly lower accuracy performance for the NDM ( P\u200a=\u200a7 . 13×10−8 , Z\u200a=\u200a−5 . 26 ) , independently of the number of models parameters ., We noted that although the NDM is capable of dispersing the initial infectious-like factors from the seed regions to the rest of the brain network , such dispersion is at the expense of the local concentrations , which after the initial exchange decreases continuously ., As a consequence , the total Aß concentration is never higher than the “injected” amount and after a given time the propagation of the factors stops ., This behavior is not physiologically realistic as shown in the literature 77 ., Note that this issue is a consequence of the absence of a source term in the NDM , which is included in the ESM ., In addition , consistent with reported associations between functional proximities to a pathogenic epicenter and gray matter atrophy levels 27 , 30 , we found that effective anatomical distances to the Aß outbreak regions can predict regional Aß depositions and arrival times values ., In terms of prediction accuracy , anatomical connectional proximities to the epicenter seem to be more interrelated to Aß levels than functional proximities to gray matter atrophy levels ( Table S2 ) ., This might be responding to several possible causes , such as:, a ) a tentative higher impact of the anatomical connectivity ( implying only direct links ) than the functional connectivity ( implying both direct and indirect links ) on pathogenic agents propagation ,, b ) the use on 27 , 30 of indirect measures of MP presence to evaluate prion-like mechanisms , i . e . gray matter atrophy quantified using voxel-based morphometry , and, c ) the fact that gray matter atrophy can be caused by multiple pathogenic factors ( e . g . , vascular and metabolic dysregulations ) ., In addition , in these studies the nodes of the analyzed networks were obtained based on a priori statistical selection of the significantly affected brain regions in the diseased group , ignoring other brain regions , which may have introduced a bias in the posterior atrophy level vs functional proximity analysis ., The intercellular transference of pathogenic proteins ( e . g . across axonal projections 10 , 19 , or the extracellular space that is constrained by the connectional architecture 20 ) , is a major statement of the prion-like hypothesis ., The ability of ESM to reconstruct Aß deposition patterns from early to advanced disease states , suggest the methodological importance of considering the structural connectional information on the modeling of MP propagation/deposition mechanisms ., However , these alone do not offer an evaluation of the real contributions/advantages of using or not the connectional information on MP propagation modeling ., We tested this contribution by comparing the ability of the introduced ESM framework to explain advanced Aß deposition patterns , using the available connectivity information and alternatively using equivalent non-informative connectional information ( Results , first subsection , and Table S4 ) ., The results supported a significantly higher model competence to explain Aß deposition patterns | Introduction, Results, Discussion, Methods | Misfolded proteins ( MP ) are a key component in aging and associated neurodegenerative disorders ., For example , misfolded Amyloid-ß ( Aß ) and tau proteins are two neuropathogenic hallmarks of Alzheimers disease ., Mechanisms underlying intra-brain MP propagation/deposition remain essentially uncharacterized ., Here , is introduced an epidemic spreading model ( ESM ) for MP dynamics that considers propagation-like interactions between MP agents and the brains clearance response across the structural connectome ., The ESM reproduces advanced Aß deposition patterns in the human brain ( explaining 46∼56% of the variance in regional Aß loads , in 733 subjects from the ADNI database ) ., Furthermore , this model strongly supports, a ) the leading role of Aß clearance deficiency and early Aß onset age during Alzheimers disease progression ,, b ) that effective anatomical distance from Aß outbreak region explains regional Aß arrival time and Aß deposition likelihood ,, c ) the multi-factorial impact of APOE e4 genotype , gender and educational level on lifetime intra-brain Aß propagation , and, d ) the modulatory impact of Aß propagation history on tau proteins concentrations , supporting the hypothesis of an interrelated pathway between Aß pathophysiology and tauopathy ., To our knowledge , the ESM is the first computational model highlighting the direct link between structural brain networks , production/clearance of pathogenic proteins and associated intercellular transfer mechanisms , individual genetic/demographic properties and clinical states in health and disease ., In sum , the proposed ESM constitutes a promising framework to clarify intra-brain region to region transference mechanisms associated with aging and neurodegenerative disorders . | Misfolded proteins ( MP ) mechanisms are a characteristic pathogenic feature of most prevalent human neurodegenerative diseases , such as Alzheimers disease ( AD ) ., Characterizing the mechanisms underlying intra-brain MP propagation and deposition still constitutes a major challenge ., Here , we hypothesize that these complex mechanisms can be accurately described by epidemic spreading-like interactions between infectious-like agents ( MP ) and the brains MP clearance response , which are constrained by the brains connectional architecture ., Consequently , we have developed a stochastic epidemic spreading model ( ESM ) of MP propagation/deposition that allows for reconstructing individual lifetime histories of intra-brain MP propagation , and the subsequent analysis of factors that promote propagation/deposition ( e . g . , MP production and clearance ) ., Using 733 individual PET Amyloid-ß ( Aß ) datasets , we show that ESM explains advanced Aß deposition patterns in healthy and diseased ( AD ) brains ., More importantly , it offers new avenues for our understanding of the mechanisms underlying MP mediated disorders ., For instance , the results strongly support the growing body of evidence suggesting the leading role of a reduced Aβ clearance on AD progression and the modulatory impact of Aß mechanisms on tau proteins concentrations , which could imply a turning point for associated therapeutic mitigation strategies . | connectomics, medicine and health sciences, dementia, neuroanatomy, anatomy, mental health and psychiatry, computational neuroscience, nervous system, neurology, neurodegenerative diseases, biology and life sciences, alzheimer disease, computational biology, autosomal dominant diseases, neuroscience, huntington disease, neuroimaging, clinical genetics | null |
journal.pgen.1003939 | 2,013 | Pathways-Driven Sparse Regression Identifies Pathways and Genes Associated with High-Density Lipoprotein Cholesterol in Two Asian Cohorts | Much attention continues to be focused on the problem of identifying SNPs and genes influencing a quantitative or dichotomous trait in genome wide scans 1 ., Despite this , in many instances gene variants identified in GWAS have so far uncovered only a relatively small part of the known heritability of most common diseases 2 ., Possible explanations include the presence of multiple SNPs with small effects , or of rare variants , which may be hard to detect using conventional approaches 2–4 ., One potentially powerful approach to uncovering the genetic etiology of disease is motivated by the observation that in many cases disease states are likely to be driven by multiple genetic variants of small to moderate effect , mediated through their interaction in molecular networks or pathways , rather than by the effects of a few , highly penetrant mutations 5 ., Where this assumption holds , the hope is that by considering the joint effects of variants acting in concert , pathways GWAS methods will reveal aspects of a diseases genetic architecture that would otherwise be missed when considering variants individually 6 , 7 ., In this paper we describe a sparse regression method utilising prior information on gene pathways to identify putative causal pathways , along with the constituent variants that may be driving pathways association ., Sparse modelling approaches are becoming increasingly popular for the analysis of genome wide datasets 8–11 ., Sparse regression models enable the joint modelling of large numbers of SNP predictors , and perform ‘model selection’ by highlighting small numbers of variants influencing the trait of interest ., These models work by penalising or constraining the size of estimated regression coefficients ., An interesting feature of these methods is that different sparsity patterns , that is different sets of genetic predictors having specified properties , can be obtained by varying the nature of this constraint ., For example , the lasso 12 selects a subset of variants whose main effects best predict the response ., Where predictors are highly correlated , the lasso tends to select one of a group of correlated predictors at random ., In contrast , the elastic net 13 selects groups of correlated variables ., Model selection may also be driven by external information , unrelated to any statistical properties of the data being analysed ., For example , the fused lasso 14 , 15 uses ordering information , such as the position of genomic features along a chromosome to select ‘adjacent’ features together ., Prior information on functional relationships between genetic predictors can also be used to drive the selection of groups of variables ., In the present context , information mapping genes and SNPs to functional gene pathways has recently been used in sparse regression models for pathway selection ., Chen et al . 16 describe a method that uses a combination of lasso and ridge regression to assess the significance of association between a candidate pathway and a dichotomous ( case-control ) phenotype , and apply this method in a study of colon cancer etiology ., In contrast , Silver et al . 17 use group lasso penalised regression to select pathways associated with a multivariate , quantitative phenotype characteristic of structural change in the brains of patients with Alzheimers disease ., In identifying pathways associated with a trait of interest , a natural follow-up question is to ask which SNPs and/or genes are driving pathway selection ?, We might further ask a related question: can the use of prior information on putative gene interactions within pathways increase power to identify causal SNPs or genes , compared to alternative methods that disregard such information ?, One way to answer these questions is by conducting a two-stage analysis , in which we first identify important pathways , and then in a second step search for SNPs or genes within selected pathways 18 , 19 ., There are however a number of problems with this approach ., Firstly , highlighted variants are then not necessarily those that were driving pathway selection in the first step of the analysis ., Secondly , the implicit ( and reasonable ) assumption is that only a small number of SNPs in a pathway are driving pathway selection , so that ideally we would prefer a model that has this assumption built in ., The above considerations point to the use of a ‘dual-level’ sparse regression model that imposes sparsity at both the pathway and SNP level ., Such a model would perform simultaneous pathway and SNP selection , with the additional benefit of being simpler to implement ., A suitable sparse regression model enforcing the required dual-level sparsity is the sparse group lasso ( SGL ) 20 ., SGL is a comparatively recent development in sparse modelling , and in simulations has been shown to accurately recover dual-level sparsity , in comparison to both the group lasso and lasso 20 , 21 ., SGL has been used for the identification of rare variants in a case-control study by grouping SNPs into genes 22; for the identification of genomic regions whose copy number variations have an impact on RNA expression levels 23; and to model geographical factors driving climate change 24 ., SGL can be seen as fitting into a wider class of structured-sparsity inducing models that use prior information on relationships between predictors to enforce different sparsity patterns 25–27 ., Hierarchical and mixed effect modelling approaches have also been suggested as a means of leveraging pathways information for the simultaneous identification of SNPs or genes within associated pathways ., Brenner et al . 28 propose such a method for identifying SNPs in a priori selected candidate pathways by comparing results from multiple studies in a meta-analysis ., This approach is similar in motivation to the two-stage methods described above ., The method proposed by Wang et al . 29 is closer in spirit to our own , in that it provides measures of pathway significance , and also ranks genes within pathways ., Both of these methods however use results from univariate tests of association at each gene variant as input to the models , in contrast to our joint-modelling approach ., Here we describe a method for sparse , pathways-driven SNP selection that extends earlier work using group lasso penalised regression for pathway selection ., This latter method was previously shown to offer improved power and specificity for identifying associated pathways , compared with a widely-used alternative 30 ., In following sections we describe our method in detail , and demonstrate through simulation that the incorporation of prior information mapping SNPs to gene pathways can boost the power to detect SNPs and genes associated with a quantitative trait ., We further describe an application study in which we investigate pathways and genes associated with serum high-density lipoprotein cholesterol ( HDLC ) levels in two separate cohorts of Asian adults ., HDLC refers to the cholesterol carried by small lipoprotein molecules , so called high density lipoproteins ( HDLs ) ., HDLs help remove the cholesterol aggregating in arteries , and are therefore protective against cardiovascular diseases 31 ., Serum HDLC levels are genetically heritable 32 ., GWAS studies have now uncovered more than 100 HDLC associated loci ( see www . genome . gov/gwastudies , Hindorff et al . 33 ) ., However , considering serum lipids as a whole , variants so far identified account for only 25–30% of the genetic variance , highlighting the limited power of current methodologies to detect hidden genetic factors 34 ., We arrange the observed values for a univariate quantitative trait or phenotype , measured for N unrelated individuals , in an response vector ., We assume minor allele counts for P SNPs are recorded for all individuals , and denote by the minor allele count for SNP j on individual i ., These are arranged in an genotype design matrix ., Phenotype and genotype vectors are mean centred , and SNP genotypes are standardised to unit variance , so that , for ., We assume that all P SNPs may be mapped to L groups or pathways , , , and begin by considering the case where pathways are disjoint or non-overlapping , so that for any ., We denote the vector of SNP regression coefficients by , and additionally denote the matrix containing all SNPs mapped to pathway by , where , is the column vector of observed SNP minor allele counts for SNP j , and is the number of SNPs in ., We denote the corresponding vector of SNP coefficients by ., In general , where P is large , we expect only a small proportion of SNPs to be ‘causal’ , in the sense that they exhibit phenotypic effects ., A key assumption in pathways analysis is that these causal SNPs will tend to be enriched within a small set , , of causal pathways , with , where denotes the size ( cardinality ) of ., We denote the set of causal SNPs mapping to pathway by , and make the further assumption that most SNPs in a causal pathway are non-causal , so that , where denotes the size ( cardinality ) of ., A suitable sparse regression model imposing the required , dual-level sparsity pattern is the sparse group lasso ( SGL ) ., We illustrate the resulting causal SNP sparsity pattern in Figure 1 , and compare it to that generated by the group lasso ( GL ) , a group-sparse model that we used previously in a sparse regression method to identify gene pathways 17 , 30 ., With the SGL 20 , sparse estimates for the SNP coefficient vector , are given by ( 1 ) where and are parameters controlling sparsity , and is a pathway weighting parameter that may vary across pathways ., ( 1 ) corresponds to an ordinary least squares ( OLS ) optimisation , but with two additional constraints on the coefficient vector , , that tend to shrink the size of , relative to OLS estimates ., One constraint imposes a group lasso-type penalty on the size of ., Depending on the values of and , this penalty has the effect of setting multiple pathway SNP coefficient vectors , , thereby enforcing sparsity at the pathway level ., Pathways with non-zero coefficient vectors form the set of ‘selected’ pathways , so thatA second constraint imposes a lasso-type penalty on the size of ., Depending on the values of and , for a selected pathway , this penalty has the effect of setting multiple SNP coefficient vectors , , thereby enforcing sparsity at the SNP level within selected pathways ., SNPs with non-zero coefficient vectors then form the set of selected SNPs in pathway l , so thatThe set of all selected SNPs is given byThe sparsity parameter controls the degree of sparsity in , such that the number of pathways and SNPs selected by the model increases as is reduced from a maximal value , above which ., The parameter controls how the sparsity constraint is distributed between the two penalties ., When , ( 1 ) reduces to the group lasso , so that sparsity is imposed only at the pathway level , and all SNPs within a selected pathway have non-zero coefficients ., When , solutions exhibit dual-level sparsity , such that as approaches 0 from above , greater sparsity at the group level is encouraged over sparsity at the SNP level ., When , ( 1 ) reverts to the lasso , so that pathway information is ignored ., For the estimation of we proceed by noting that the optimisation ( 1 ) is convex , and ( in the case of non-overlapping groups ) that the penalty is block-separable , so that we can obtain a solution using block , or group-wise coordinate gradient descent ( BCGD ) 35 ., A detailed derivation of the estimation algorithm is given in the accompanying Supplementary Information S1 , Section, 3 . From ( S . 9 ) and ( S . 10 ) , the criterion for selecting a pathway l is given by ( 2 ) and the criterion for selecting SNP j in selected pathway l by ( 3 ) where and are respectively the pathway and SNP partial residuals , obtained by regressing out the current estimated effects of all other pathways and SNPs respectively ., The complete algorithm for SGL estimation using BCGD is presented in Box, 1 . We test the hypothesis that where causal SNPs are enriched in a given pathway , pathway-driven SNP selection using SGL will outperform simple lasso selection that disregards pathway information in a simple simulation study ., We simulate genetic markers for individuals ., Marker frequencies for each SNP are sampled independently from a multinomial distribution following a Hardy Weinberg equilibrium frequency distribution ., SNP minor allele frequencies are sampled from a uniform distribution ., SNPs are distributed equally between 50 non-overlapping pathways , each containing 50 SNPs ., We then test each competing method over 500 Monte Carlo ( MC ) simulations ., At each simulation , a baseline univariate phenotype is sampled from ., To generate genetic effects , we randomly select 5 SNPs from a single , randomly selected pathway , to form the set of causal SNPs ., Genetic effects are then generated as described in Supplementary Information S1 , Section S3 ., To enable a fair comparison between the two methods ( SGL and lasso ) , we ensure that both methods select the same number of SNPs at each simulation ., We do this by first obtaining the SGL solution , , with and , which ensures sparsity at both the pathway and SNP level ., We use a uniform pathway weighting vector ., We then compute the lasso solution using coordinate descent over a range of values for the lasso regularisation penalty , , and choose the setwhere is the number of SNPs previously selected by SGL , and is the number of SNPs selected by the lasso with ., We measure performance as the mean power to detect all 5 causal SNPs over 500 MC simulations , and test a range of genetic effect sizes ( see Supplementary Information S1 , Section S3 ) ., In a follow up study , we compare the performance of the two methods in a scenario in which pathways information is uninformative ., For this we repeat the previous simulations , but with 5 causal SNPs drawn at random from all 2500 SNPs , irrespective of pathway membership ., Results are presented in Figure, 2 . Referring to Figure 2 , we see that where causal SNPs are concentrated in a single causal pathway ( Figure 2 - left ) , SGL demonstrates greater power ( and equivalently specificity , since the total number of selected SNPs is constant ) , compared with the lasso , above a particular effect size threshold ( here ) ., Where pathway information is not important , that is causal SNPs are not enriched in any particular pathway ( Figure 2 - right ) , SGL performs poorly ., To gain a deeper understanding of what is happening here , we also consider the power distributions across all 500 MC simulations corresponding to each point in the plots of Figure, 2 . These are illustrated in Figure, 3 . The top row of plots illustrates the case where causal SNPs are drawn from a single causal pathway ., Here we see that there is a marked difference between the two distributions ( SGL vs lasso ) ., The lasso shows a smooth distribution in power , with mean power increasing with effect size ., In contrast , with SGL the distribution is almost bimodal , with power typically either 0 or 1 , depending on whether or not the correct causal pathway is selected ., This serves as an illustration of the advantage of pathway-driven SNP selection for the detection of causal SNPs in the case that pathways are important ., As previously found by Zhou et al . 6 in the context of rare variants and gene selection , the joint modelling of SNPs within groups gives rise to a relaxation of the penalty on individual SNPs within selected groups , relative to the lasso ., This can enable the detection of SNPs with small effect size or low MAF that are missed by the lasso , which disregards pathways information and treats all SNPs equally ., Where causal SNPs are not enriched in a causal pathway ( bottom row of Figure 3 ) , as expected SGL performs poorly ., In this case SGL will only select a SNP where the combined effects of constituent SNPs in a pathway are large enough to drive pathway selection ., Finally , with many pathways methods an adjustment to pathway test statistics is made to account for biases due to variations in pathway size , that is the number of SNPs in a pathway 6 ., We explore potential biases using SGL for pathway selection using the simulation framework described above , but this time allowing for varying pathway sizes , ranging from 10 to 200 SNPs ., We find no evidence of a pathway size bias ( see Supplementary Information S1 , Section 5 for further details ) ., We discuss the issue of accounting for pathway size and other potential biases in pathway and SNP selection when using real data in a later section ., The assumption that pathways are disjoint does not hold in practice , since genes and SNPs may map to multiple pathways ( see ‘Pathway mapping’ section below ) ., This means that typically for some ., In the context of pathways-driven SNP selection using SGL , this has two important implications ., Firstly , the optimisation ( 1 ) is no longer separable into groups ( pathways ) , so that convergence using coordinate descent is no longer guaranteed 35 ., Secondly , we wish to be able to select pathways independently , and the SGL model as previously described does not allow this ., For example consider the case of an overlapping gene , that is a gene that maps to more than one pathway ., If a SNP mapping to this gene is selected in one pathway , then it must be selected in each and every pathway containing the mapped gene , so that all pathways mapping to the gene are selected ., We instead want to admit the possibility that the joint SNP effects in one pathway may be sufficient to allow pathway selection , while the joint effects in another pathway containing some of the same SNPs do not pass the threshold for pathway selection ., A solution to both these problems is obtained by duplicating SNP predictors in , so that SNPs belonging to more than one pathway can enter the model separately 30 , 36 ., The process works as follows ., An expanded design matrix is formed from the column-wise concatenation of the sub-matrices , , to form the expanded design matrix of size , where ., The corresponding parameter vector , , is formed by joining the pathway parameter vectors , , so that ., Pathway mappings with SNP indices in the expanded variable space are reflected in updated groups ., The SGL estimator ( 1 ) , adapted to account for overlapping groups , is then given by ( 4 ) With this overlap expansion , the model is then able to perform pathway and SNP selection in the way that we require , and the corresponding optimisation problem is amenable to solution using the BCGD estimation algorithm described in Box, 1 . However , for the purpose of pathways-driven SNP selection , the application of this algorithm presents a problem ., This arises from the replication of overlapping SNP predictors in each group , , that they occur ., Consider for example the simple situation where there are two pathways , , containing sets of causal SNPs and respectively ., Here the indicates that SNP indices refer to the expanded variable space ., We begin by assuming that and contain the same SNPs , so that in the unexpanded variable space , ., We then proceed with BCGD by first estimating ., We assume that the correct SNPs are selected , so that , and otherwise ., For the estimation of , the estimated effect , of these overlapping causal SNPs is removed from the regression , through its incorporation in the block residual ., Since no other causal SNPs exist in pathway , so that the criterion for pathway selection , ( 2 ) is not met ., That is is not selected ., Now consider the case where additional , non-overlapping causal SNPs , possibly with smaller effects , occur in , so that in the unexpanded variable space , ., In other words , causal SNPs are partially overlapping ( see Figure 4 ) ., This is the situation for example where multiple causal genes overlap both pathways , but one or more additional causal genes occur in ., During BCGD pathway is then less likely to be selected by the model , than would be the case if there were no overlapping SNPs , since once again the effects of overlapping causal SNPs , , are removed ., For pathways-driven SNP selection , we will argue that we instead require that SNPs are selected in each and every pathway whose joint SNP effects pass a revised pathway selection threshold , irrespective of overlaps between pathways ., This is equivalent to the previous pathway selection criterion ( 2 ) , but with the additional assumption that pathways are independent , in the sense that they do not compete in the model estimation process ., We describe a revised estimation algorithm under the assumption of pathway independence below ., We justify the strong assumption of pathway independence with the following argument ., In reality , we expect that multiple pathways may simultaneously influence the phenotype , and we also expect that many such pathways will overlap , for example through their containing one or more ‘hub’ genes , that overlap multiple pathways 37 , 38 ., By considering each pathway independently , we aim to maximise the sensitivity of our method to detect these variants and pathways ., In contrast , without the independence assumption , a competitive estimation algorithm will tend to pick out one from each set of similar , overlapping pathways , and miss potentially causal pathways and variants as a consequence ., We illustrate this idea in the simulation study in the following section ., One potential concern is that by not allowing pathways to compete against each other , specificity may be reduced , since too many pathways and SNPs may be selected ., We discuss the issue of specificity further in the context of results from the simulation study ., A detailed derivation of the SGL model estimation algorithm under the independence assumption is given in Supplementary Information S1 , Section, 2 . The main results are that the pathway ( 2 ) and SNP ( 3 ) selection criteria become ( 5 ) respectively ., The key difference is that partial derivatives and are replaced by , that is each pathway is regressed against the phenotype vector ., This means that there is no block coordinate descent stage in the estimation , so that the revised algorithm utilises only coordinate gradient descent within each selected pathway ., For this reason we use the acronym SGL-CGD for the revised algorithm , and SGL-BCGD for the previous algorithm using block coordinate gradient descent ., The new algorithm is described in Box, 2 . Finally , we note that for SNP selection we are interested only in the set of selected SNPs in the unexpanded variable space , and not the set ., Since , under the independence assumption , the estimation of each does not depend on the other estimates , , we do not need to record separate coefficient estimates for each pathway in which a SNP is selected ., Instead we need only record the set of SNPs selected in each selected pathway ., This has a useful practical implication , since we can avoid the need for an expansion of or , and simply form the complete set of selected SNPs as We now explore some of the issues raised in the preceding section , specifically the potential impact on pathway and SNP selection power and specificity of treating the pathways as independent in the SGL estimation algorithm ., We do this in a simulation study in which we simulate overlapping pathways ., The simulation scheme is specifically designed to highlight differences in pathway and SNP selection with the independence assumption ( using the SGL-CGD estimation algorithm in Box 2 ) and without it ( using the standard SGL estimation algorithm in Box 1 ) ., SNPs with variable MAF are simulated using the same procedure described in the previous simulation study , but this time SNPs are mapped to 50 overlapping pathways , each containing 30 SNPs ., Each pathway overlaps any adjacent ( by pathway index ) pathway by 10 SNPs ., This overlap scheme is illustrated in Figure 5 ( top ) ., As before we consider a range of overall genetic effect sizes , ., A total of 2000 MC simulations are conducted for each effect size ., At MC simulation , we randomly select two adjacent pathways , where ., From these two pathways we randomly select 10 SNPs according to the scheme illustrated in Figure 5 ( bottom ) ., This ensures that causal SNPs overlap a minimum of 1 , and a maximum of 2 pathways , with ., The true set of causal pathways , , is then given by , or ( although simulations where will be extremely rare ) ., Genetic effects on the phenotype are generated as described previously ( Supplementary Information S1 , Section S3 ) ., SNP coefficients are estimated for each algorithm , SGL-BCGD and SGL-CGD , using the same regularisation with and for both ., The average number of pathways and SNPs selected by SGL-BCGD and SGL-CGD across all 2000 MC simulations is reported in Table, 1 . As expected , for both models , the number of selected variables ( pathways or SNPs ) increases with decreasing effect size , as the number of pathways close to the selection threshold set by increases ., For each model , at MC simulation we record the pathway and SNP selection power , and respectively ., Since the number of selected variables can vary slightly between the two models , we also record false positive rates ( FPR ) for pathway and SNP selection as and respectively ., The large possible variation in causal SNP distributions , causal SNP MAFs etc . makes a comparison of mean power and FPR between the two methods somewhat unsatisfactory ., For example , depending on effect size , a large number of simulations can have either very high , or very low pathway and SNP selection power , masking subtle differences in performance between the two methods ., Since we are specifically interested in establishing the relative performance of the two methods , we instead illustrate the number of simulations at which one method outperforms the other across all 2000 MC simulations , and show this in Figure 6 ., In this figure , the number of simulations in which SGL-CGD outperforms SGL , i . e . where SGL-CGD power>SGL-BCGD power , or SGL-CGD FPR<SGL-BCGD FPR , are shown in green ., Conversely , the number of simulations where SGL-BCGD outperforms SGL-CGD are shown in red ., We first consider pathway selection performance ( top row of Figure 6 ) ., For both methods , the same number of pathways are selected on average , across all effect sizes ( Table 1 ) ., At low effect sizes , there is no difference in performance between the two methods for the large majority of MC simulations , and where there is a difference , the two methods are evenly balanced ., As with SGL Simulation Study 1 , this is the region ( with ) where pathway selection fairs no better than chance ., With , SGL-CGD consistently outperforms SGL , both in terms of pathway selection sensitivity and control of false positives ( measured by FPR ) ., To understand why , we turn to SNP selection performance ( bottom row of Figure 6 ) ., At small effect sizes , in the small minority of simulations where the correct pathways are identified , SGL-BCGD tends to demonstrate greater power than SGL-CGD ( Figure 6 bottom left ) ., However , this is at the expense of lower specificity ( Figure 6 bottom right ) ., These difference are due to the slightly larger number of SNPs selected by SGL-BCGD ( see Table 1 ) , which in turn is due to the ‘screening out’ of previously selected SNPs from the adjacent causal pathway during BCGD , as described previously ., This results in the selection of a larger number of SNPs when any two overlapping pathways are selected by the model ., In the case where two causal pathways are selected , SNP selection power is then likely to be higher , although at the expense of a greater number of false positives ., When pathway effects are just on the margin of detectability , SGL-CGD is more often able to select both causal pathways , although this doesnt translate into increased SNP selection power ., This is most likely because at this effect size neither model can detect SNPs with low MAF , so that SGL-CGD is detecting the same ( overlapping ) SNPs in both causal pathways ., Note that once again SGL-BCGD typically has a higher FPR than SGL-CGD , since more SNPs are selected from non-causal pathways ., As the effect size increases , the number of simulations in which SGL-CGD outperforms SGL-BCGD for SNP selection power grows , paralleling the former methods enhanced pathway selection power ., This is again a demonstration of the screening effect with SGL-BCGD described previously ., This means that SGL-CGD is more often able to select both causal pathways , and to select additional causal SNPs that are missed by SGL ., These additional SNPs are likely to be those with lower MAF , for example , that are harder to detect with SGL , once the effect of overlapping SNPs are screened out during estimation using BCGD ., Interestingly , as before SGL-CGD continues to exhibit lower false positive rates than SGL ., This suggests that , with the simulated data considered here , the independence assumption offers better control of false positives by enabling the selection of causal SNPs in each and every pathway to which they are mapped ., In contrast , where causal SNPs are successively screened out during the estimation using BCGD , too many SNPs with spurious effects are selected ., The relative advantage of SGL-CGD over SGL-BCGD on all performance measures starts to decrease around , as SGL-BCGD becomes better able to detect all causal pathways and SNPs , irrespective of the screening effect ., One issue that must be addressed is the problem of selection bias , by which we mean the tendency of SGL to favour the selection of particular pathways or SNPs under the null , where no SNPs influence the phenotype ., Possible biasing factors include variations in pathway size or varying patterns of SNP-SNP correlations and gene sizes ., Common strategies for bias reduction include the use of dimensionality reduction techniques and permutation methods 39–42 ., In earlier work we described an adaptive weight-tuning strategy , designed to reduce selection bias in a group lasso-based pathway selection method 30 ., This works by tuning the pathway weight vector , , so as to ensure that pathways are selected with equal probability under the null ., This strategy can be readily extended to the case of dual-level sparsity with the SGL ., Our procedure rests on the observation that for pathway selection to be unbiased , each pathway must have an equal chance of being selected ., For a given , and with tuned to ensure that a single pathway is selected , pathway selection probabilities are then described by a uniform distribution , , for ., We proceed by calculating an empirical pathway selection frequency distribution , , by determining which pathway will first be selected by the model as is reduced from its maximal value , , over multiple permutations of the response , ., This process is described in detail in Supplementary Information S1 , Section 4 ., We note that alternative methods for the construction of ‘null’ distributions , for example by permuting genotype labels , have been used in existing pathways analysis methods 6 ., In the present context we choose to permute phenotype labels in order to preserve LD structure , since we expect this to be a significant source of bias with our data ., Our iterative weight tuning procedure then works by applying successive adjustments to the pathway weight vector , , so as to reduce the difference , , between the unbiased and empirical ( biased ) distributions for each pathway ., At iteration , we compute the empirical pathway selection probability distribution , determine for each pathway , and then apply the following weight adjustmentThe parameter controls the maximum amount by which each can be reduced in a single iteration , in the case that pathway l | Introduction, Materials and Methods, Results | Standard approaches to data analysis in genome-wide association studies ( GWAS ) ignore any potential functional relationships between gene variants ., In contrast gene pathways analysis uses prior information on functional structure within the genome to identify pathways associated with a trait of interest ., In a second step , important single nucleotide polymorphisms ( SNPs ) or genes may be identified within associated pathways ., The pathways approach is motivated by the fact that genes do not act alone , but instead have effects that are likely to be mediated through their interaction in gene pathways ., Where this is the case , pathways approaches may reveal aspects of a traits genetic architecture that would otherwise be missed when considering SNPs in isolation ., Most pathways methods begin by testing SNPs one at a time , and so fail to capitalise on the potential advantages inherent in a multi-SNP , joint modelling approach ., Here , we describe a dual-level , sparse regression model for the simultaneous identification of pathways and genes associated with a quantitative trait ., Our method takes account of various factors specific to the joint modelling of pathways with genome-wide data , including widespread correlation between genetic predictors , and the fact that variants may overlap multiple pathways ., We use a resampling strategy that exploits finite sample variability to provide robust rankings for pathways and genes ., We test our method through simulation , and use it to perform pathways-driven gene selection in a search for pathways and genes associated with variation in serum high-density lipoprotein cholesterol levels in two separate GWAS cohorts of Asian adults ., By comparing results from both cohorts we identify a number of candidate pathways including those associated with cardiomyopathy , and T cell receptor and PPAR signalling ., Highlighted genes include those associated with the L-type calcium channel , adenylate cyclase , integrin , laminin , MAPK signalling and immune function . | Genes do not act in isolation , but interact in complex networks or pathways ., By accounting for such interactions , pathways analysis methods hope to identify aspects of a disease or traits genetic architecture that might be missed using more conventional approaches ., Most existing pathways methods take a univariate approach , in which each variant within a pathway is separately tested for association with the phenotype of interest ., These statistics are then combined to assess pathway significance ., As a second step , further analysis can reveal important genetic variants within significant pathways ., We have previously shown that a joint-modelling approach using a sparse regression model can increase the power to detect pathways influencing a quantitative trait ., Here we extend this approach , and describe a method that is able to simultaneously identify pathways and genes that may be driving pathway selection ., We test our method using simulations , and apply it to a study searching for pathways and genes associated with high-density lipoprotein cholesterol in two separate East Asian cohorts . | null | null |
journal.pgen.1003824 | 2,013 | In Vivo Analysis of Lrig Genes Reveals Redundant and Independent Functions in the Inner Ear | Protein-protein interactions are critical for diverse and complex biological functions throughout the animal kingdom , including nervous system development , cell adhesion and signaling , tissue morphogenesis , the immune response and human disease 1–4 ., This functional diversity is accomplished by superfamilies of proteins harboring combinations of common protein recognition motifs ., For instance , the human genome encodes hundreds of proteins with extracellular leucine rich repeats ( LRR ) , a 20–30 amino acid motif that forms a characteristic horseshoe structure for protein-protein interactions 5 , 6 ., Similarly , the large immunoglobulin ( Ig ) superfamily of cell adhesion molecules is defined by the presence of Ig domains , which can mediate highly specific homophilic and heterophilic binding 7 , 8 ., Despite their abundance , LRR and Ig motifs are rarely found in the same protein , with only several dozen mammalian genes encoding LRR-Ig proteins that fall into twelve gene families 3 , 9 , 10 ., Most of these proteins are vertebrate-specific and show discrete expression in the developing nervous system , suggesting that expansion of the LRR-Ig family may have contributed to the increased complexity of the vertebrate nervous system ., Consistent with this idea , several LRR-Ig proteins have been shown to control highly specific cell-cell interactions underlying synapse formation and other aspects of nervous system development 2 ., The invertebrate-specific Kekkon proteins , on the other hand , modulate signaling by binding to and downregulating EGF receptors 11 , 12 ., Within the LRR-Ig family , only the Lrig subfamily contains both invertebrate and vertebrate members 3 , indicating that analysis of this family may provide general insights into the evolution of LRR-Ig proteins ., The leucine-rich repeat and immunoglobulin-like domain proteins ( Lrigs ) are single pass transmembrane proteins with extracellular domains containing fifteen LRRs , three Ig-like domains and intracellular domains of varying length 13 ., The fly and worm genomes each contain a single Lrig gene ., This family is expanded in the vertebrate genome , which encodes for three family members 14: Lrig1 ( formerly Lig1 ) , Lrig2 , and Lrig3 ., The extracellular domains are highly conserved within the family , but the cytoplasmic domains diverge significantly , with no motifs common to flies , worms , or vertebrates ., This suggests that Lrig family members may interact with similar binding partners yet ultimately exert distinct downstream effects ., Most of what is known about Lrig function has come from analysis of Lrig1 , which is downregulated in several human cancers 15 ., Consistent with its proposed role as a tumor suppressor gene , Lrig1 can control the activity of several receptor tyrosine kinases ( rTKs ) with important effects on cell proliferation and survival ., For instance , Lrig1 negatively regulates members of the ErbB family of receptors by promoting receptor degradation 16–18 ., In support of this , Lrig1 regulates EGFR levels in primary human keratinocytes 19 , and loss of Lrig1 results in increased EGF signaling and excess intestinal stem cell proliferation , tumor formation and psoriasis-like hyperplasia in mice 20–22 ., However , Lrig1 can also inhibit Met and Ret rTK activation 23 , 24 , suggesting that Lrig1 activity extends beyond regulation of EGF signaling ., How any Lrig protein functions at the molecular level remains a mystery ., Whether Lrig3 shares some properties with Lrig1 remains an open question ., As predicted by homology in their extracellular domains , both Lrig1 and Lrig3 can bind to ErbB receptors 25 ., However , although downregulation of Lrig3 in human glioma cells caused enhanced EGFR levels 26 , more recent studies indicate that Lrig3 actually opposes Lrig1s effects on EGF signaling 18 ., In addition , similar to Lrig1s ability to interact with a variety of receptors , Lrig3 also binds to FGF receptors and regulates FGF and Wnt signaling in Xenopus 27 ., Whereas several phenotypes reported in Lrig1 mutant mice have been associated with changes in EGF signaling , loss of Lrig3 leads to a disruption in the three-dimensional structure of the inner ear that is not easily explained by altered ErbB signaling 25 , 28 ., Thus , it is not yet clear how the functions identified for Lrig1 and Lrig3 in vitro translate to their actions in vivo ., Comparison of Lrig1 and Lrig2 , on the other hand , has suggested key differences ., First , reduction of Lrig2 either lowers or has no effect on EGFR levels in vitro 18 , 29 ., Consistent with this observation , Lrig2 does not behave like a typical tumor suppressor in humans ., For instance , Lrig2 expression can be increased in some human tumors , and a combination of high levels of Lrig2 and low levels of Lrig1 correlates with a poor prognosis for a type of early-stage squamous cell carcinoma 30 ., Similarly , overexpression of Lrig2 correlates with invasiveness of pituitary adenoma 31 ., In addition , studies of Lrig protein expression in human tumors have revealed fundamental differences in the subcellular distribution of these family members 32 , 33 ., Although Lrig2 phenotypes have not yet been described in mice , loss of LRIG2 causes Urofacial Syndrome in humans , which is characterized by abnormal bladder function and altered facial expression , possibly due to abnormal innervation 34 ., In order to clarify whether Lrig genes mediate common biological functions in vivo , we have taken a genetic approach in mice ., We have focused our analysis on the development and function of the inner ear , an exquisitely complex structure whose perfect form and function is crucial for the senses of hearing and balance 35 , 36 ., The spiral-shaped cochlea mediates the sense of hearing ., Head position and motion is sensed by movement of fluid within the vestibular system , which consists of three semicircular canals oriented in the three dimensions of space , and a saccule and utricle that detect linear acceleration and gravity ., The inner ear contains six sensory epithelia , which contain the sensory hair cells ., Vestibular hair cells in the two maculae and three cristae detect motion of the head , whereas auditory hair cells in the organ of Corti respond to specific frequencies of sound ., Vestibular and auditory information is transmitted from the inner ear to the brain by primary sensory neurons in the vestibular or spiral ganglia respectively ., The inner ear provides an unusually sensitive system for analysis of gene function since small changes in the formation or structure of the inner ear can cause profound functional deficits in hearing and balance ., For instance , Lrig3 mutant mice exhibit hyperactivity and run in circles due to truncation of a single semicircular canal 28 ., Further , Lrigs have been shown to modulate BMP , FGF , and Wnt signaling pathways , which all play important roles in the morphogenesis and patterning of the inner ear 35 ., Thus , analysis of the inner ear provides an ideal opportunity to uncover the in vivo actions of the Lrigs ., Here , we analyzed several features of inner ear development and function in single and double Lrig mouse mutants ., Our results suggest that Lrig1 and Lrig3 cooperate during morphogenesis ., Lrig1 and Lrig2 , on the other hand , control largely distinct aspects of inner ear development and function , yet act redundantly to ensure proper innervation of the cochlea ., To be able to compare and contrast Lrig gene function in the inner ear , it is critical to know when and where each family member is expressed ., Any sites of overlap offer an opportunity to examine redundancy , whereas unique sites of expression can be used to reveal the biological significance of individual family members ., For instance , Lrig3 is the only family member expressed in the developing lateral semicircular canal and Lrig3 mutant mice circle due to defects in this canal ., However , although Lrig3 is also expressed in other regions of the inner ear , Lrig3 mutant mice exhibit normal auditory responses , with no other obvious changes in the structure or function of the inner ear 28 ., This raises the possibility that other Lrig genes compensate for the loss of Lrig3 ., Therefore , to begin to determine whether these three family members play overlapping functions , we compared their expression patterns in the inner ear , either by in situ hybridization ( Lrig1 ) or by examining the expression of βgeo reporter genes inserted into the Lrig2 ( Figure S1 ) and Lrig3 28 loci ., Given the known role for Lrig3 in canal morphogenesis , we first compared expression patterns at embryonic day 12 . 5 ( E12 . 5 ) , just before the canals begin to acquire their mature morphology ., The inner ear develops from the otic vesicle , a simple sphere of epithelium that invaginates from the epidermis overlying the hindbrain beginning around E9 in mouse 35 ., Over the next several days , the vestibular apparatus and endolymphatic duct develop from the dorsal half of the otic vesicle , while the cochlea extends ventrally ( Figure 1A ) ., Beginning around E12 , the semicircular canals are sculpted from the vertical and lateral pouches ., The utricle and saccule develop from an intermediate region called the atrium 37 ., In parallel , signaling events establish restricted sensory regions , which ultimately produce hair cells and support cells in the mature sensory epithelia in the canals ( the cristae ) , the utricle and saccule ( the maculae ) , and the cochlea ( the organ of Corti ) ., Non-sensory regions in the cochlea go on to form the lateral wall , inner sulcus , and Reissners membrane ., Consistent with previous studies 28 , Lrig1 and Lrig3 showed remarkably restricted yet related patterns of expression at E12 . 5 , overlapping both in the atrium and in the non-sensory domain of the cochlea ( Figure 1B , C ) ., In contrast , Lrig2-βgeo activity was evident throughout the early otic epithelium ( Figures 1C and S2 ) ., Indeed , Lrig2-βgeo expression appeared nearly ubiquitous at all stages examined , although the levels varied in different tissues ( Figure S2 ) ., To determine whether Lrig1 and Lrig2 , like Lrig3 , help determine the three-dimensional structure of the inner ear , we generated and analyzed Lrig1 and Lrig2 mutant mice ., Lrig1 mutant mice harbor a gene trap insertion in the third intron of the Lrig1 locus , and Lrig2 mutants contain a gene trap insertion after exon 11 ( Figure S1 ) ., These gene trap insertions are predicted to interfere with normal splicing of endogenous transcripts , instead producing transmembrane fusion proteins that are targeted to the lysosome and therefore unlikely to exert any effect 38 ., Western blot and immunostaining studies confirmed that Lrig1 and Lrig2 protein levels are severely reduced in each mutant background ( Figures S1 and S3 ) ., In contrast to Lrig3 mutants , however , both Lrig1 and Lrig2 single mutant animals exhibited normal inner ear morphologies at E14 . 5 ( Fig . 2B , C , E ) ., Given the striking co-expression of Lrig1 and Lrig3 , we wondered whether combined loss of these two family members would provide any evidence for similar functions ., Indeed , inner ear development is more severely disrupted in Lrig1−/−;Lrig3−/− double mutant mice than in either single mutant ( Table 1 ) ., For example , the utricle and saccule fail to separate ( Figure 2G , arrowhead ) , consistent with the co-expression of Lrig1 and Lrig3 in the embryonic atrium ( Figure 1 ) ., In addition , the posterior canal is abnormally small and misshapen ( Figure 2H , I ) ., To see whether Lrig1 and Lrig3 also cooperate in the lateral canal , we took advantage of the fact that the lateral canal phenotype is only partially penetrant in Lrig3 mutants maintained on this background , with truncation or thinning observed in only 33% of the animals ( Figure 2D , Table 1 ) ., However , loss of either one or two copies of Lrig1 did not strongly enhance this phenotype ( Figure 2D , G , I; Table 1 ) , consistent with the fact that Lrig1 and Lrig3 are not obviously co-expressed in the lateral canal epithelium 28 ., The fact that new phenotypes emerge only in sites of Lrig1/Lrig3 co-expression strongly suggests that these two family members act redundantly during inner ear morphogenesis ., In contrast , Lrig1 and Lrig2 do not appear to cooperate here , as Lrig1−/−;Lrig2−/− double mutant ears developed normally ( Figure 2F ) despite the extensive co-expression of Lrig1 and Lrig2 at E12 . 5 ( Figure 1B , D ) ., To gain a broader view of genetic interactions among Lrig family members , we asked whether either Lrig2 or Lrig3 exert overlapping functions with Lrig1 in other regions of the inner ear ., In support of this idea , unlike either single mutant , Lrig1−/−; Lrig3−/− double mutant animals die at or before birth ( Table, 2 ) and suffer from an array of morphogenetic phenotypes , including microphthalmia and skeletal malformations ( data not shown ) ., Although the presence of new defects suggests that Lrig1 and Lrig3 likely work together in many other tissues , this lethality prevented analysis of any other aspects of inner ear function ., In contrast , Lrig1−/− , Lrig2−/− , and Lrig1−/−;Lrig2−/− mutant mice survive past the onset of hearing ., We therefore focused the rest of the analysis on Lrig1 and Lrig2 ., As previously reported 20 , Lrig1−/− mice frequently die within the first postnatal week when maintained on an inbred background ( Table 2 ) ., Lrig2−/− mice were born in normal Mendelian ratios and showed no obvious defects ( Table 2 ) ., However , very few Lrig1−/−;Lrig2−/− double mutant animals survived to six weeks of age ., A small percentage of Lrig1−/−; Lrig2−/− double mutants survived to adulthood ( Table, 2 ) and were noticeably runty during adolescence ., More strikingly , half of the double mutant survivors exhibited a mild vestibular defect with circling behavior ( 3 of 6 animals , see video S1 ) ., Since neither surviving Lrig1−/− nor Lrig2−/− animals showed any signs of circling , this observation suggests that Lrig1 and Lrig2 may work together in the vestibular system ., To investigate this possibility , we performed a more detailed analysis of expression in the vestibular system by double labeling with an anti-Lrig1 antibody and an anti-β-galactosidase antibody to detect Lrig2-βgeo ., Lrig2 gene trap heterozygotes were used due to the lack of antibodies that reliably detect Lrig2 protein in tissue ., Lrig2-βgeo should serve as an accurate read-out of the pattern of Lrig2 expression , but it should be noted that there may be subtle differences in the stability of the Lrig2-βgeo protein compared to endogenous Lrig2 ., However , our observations of Lrig2-βgeo expression match previous reports of Lrig2 transcription 10 , so any discrepancies are likely to be minor ., No Lrig1 labeling was detected in Lrig1 mutant tissue , confirming that this antibody detects only this family member ( Figure S3 ) ., Consistent with results from in situ hybridization and X-gal staining ( Figures 1 and S2 ) , double labeling at E12 . 5 revealed highly restricted expression of Lrig1 protein in the atrium and cochlea , with broad Lrig2-βgeo expression throughout the otic epithelium and in the surrounding mesenchyme ( Figure 3A , B ) ., Within the atrium , Lrig1 was restricted to non-sensory regions , which flank the Sox2-positive sensory patches that eventually give rise to the maculae ( Figure 3C ) ., This pattern was maintained after formation of the utricle and saccule , with expression in the transitional epithelium adjacent to the utricular macula and in the extramacular epithelium of the saccule at E16 . 5 ( Figure 3D , F ) , E18 . 5 ( Figure S3 ) and P15 ( Figure 3H–J ) ., Lrig2-βgeo , in contrast , was expressed throughout sensory and non-sensory regions of the vestibular organs at E16 . 5 and continuing through the first postnatal week ( Figure 3E , G and S2 ) ., However , by P15 , Lrig2-βgeo levels were noticeably enhanced in the utricular and saccular maculae as well as the cristae ( Figure 3E , G and data not shown ) ., In contrast , Lrig1 protein was not detected in the vestibular sensory epithelia at any stage ., Thus , Lrig1 and Lrig2 are co-expressed in non-sensory regions of the utricle and saccule , but only Lrig2 seems to be present in the sensory epithelia ., One prominent site of overlapping expression was the vestibular ganglion , which communicates head position information to the brain ., Lrig1 was present at low levels in the neuronal cell bodies , with intense expression in projections to the utricle , saccule , and lateral crista at E16 . 5 ( Figure 3D ) , E18 . 5 ( Figure S3 ) , and P15 ( Figure 3H–J ) ., Lrig2-βgeo was also present in the vestibular ganglion at all stages , with enriched expression at P15 ( Figures 3H and S2 , and data not shown ) ., The co-expression of Lrig1 and Lrig2 in the vestibular ganglion , particularly at postnatal stages , may explain why Lrig1−/−;Lrig2−/− double mutant animals display occasional circling behavior , since the gross structure of the inner ear is unaffected ( Figure 2F ) and Lrig1 and Lrig2 are not co-expressed in the sensory epithelia at any stage ( Figure 3 ) ., Thus , it is possible that Lrig1 and Lrig2 act redundantly in the vestibular ganglion neurons or non-sensory epithelium , though they do not cooperate during the initial formation of the vestibular apparatus ., As in the vestibular system , Lrig1 and Lrig2 showed largely distinct patterns of expression in the cochlea , overlapping only in non-sensory regions ., Lrig1 protein was restricted to non-sensory regions of the cochlea at all stages , with maintained expression only in Reissners membrane , a structure that regulates the endolymph environment that is critical for cochlear function ( Figure 4A–C ) 39 ., Lrig2-βgeo , on the other hand , appeared ubiquitous in the cochlear epithelium and surrounding mesenchyme at E12 . 5 and E16 . 5 ( Figure 4A′–B′ ) ., However , similar to the vestibular system , expression was elevated in sensory and neural tissues postnatally ( Figures 4C′ and S2 ) ., Although Lrig1 was not detected in the spiral ganglion neurons or their projections at any stage , expression was apparent in the mesenchyme in the region that the spiral ganglion neuron neurites grow through to reach the cochlear duct ( Figure 4A , and data not shown ) ., In summary , although Lrig1 and Lrig2 are at times co-expressed in the vestibular system and cochlea , these two family members show fundamentally different expression patterns , which contrasts with the obvious similarities in the expression of Lrig1 and Lrig3 at all stages examined ( Figure 1 and 28 ) ., To assess the relative contributions of Lrig1 and Lrig2 to cochlear function , we tested auditory responsiveness in single and double mutant mice using two complementary assays ., First , we recorded Distortion Product Otoacoustic Emissions ( DPOAEs ) , which are generated by the cochlea in response to simultaneous presentation of two slightly dissimilar pure tone frequency stimuli ., Production of DPOAEs depends on outer hair cell ( OHC ) function , and DPOAE thresholds will increase if hair cells are missing , damaged , or cannot be properly stimulated due to changes in cochlear mechanics ., Second , we recorded Auditory Brainstem Responses ( ABRs ) ., ABRs reflect the sum of neuronal activity in response to sound stimulation , starting with the initial activation of spiral ganglion neurons ( wave, 1 ) and following with activation in the auditory brainstem ( waves 2–5 ) ., Sensitivity is assessed by determining the lowest intensity sound stimulus ( i . e . the threshold ) that is able to generate an ABR response ., In addition , the strength of the neuronal response can be evaluated by measuring the latency and amplitude of the first wave ., By altering the frequency of the pure tone stimuli , function can be tested along the length of the cochlea , from high frequencies in the base to low frequencies in the apex ., Together , these tests offer a sensitive way to identify impairments in the ability of the cochlea to detect and respond to acoustic stimuli ., DPOAE and ABR measurements revealed that Lrig1 , but not Lrig2 , is necessary for normal auditory sensitivity ., Lrig1 mutants showed significantly elevated DPOAE and ABR thresholds in response to 11 . 3 and 16 kHz stimuli , which typically elicit the lowest threshold responses in control animals ( Figure 5 and Table S2 ) ., Whereas control animals reliably detected 16 kHz DPOAE stimuli as quiet as 15 dB , mutants did not respond until the sounds were 45 dB , which is ∼30 times more intense ., Thresholds were also elevated in response to lower ( 5 . 6 and 8 kHz ) and higher ( 22 . 6 and 32 kHz ) frequencies , but these differences were not statistically significant since sensitivity is already reduced in these regions of control cochleae ( for example , 57 . 43±2 . 37 dB for wild-type vs . 71 . 86±4 . 5 dB for Lrig1−/− animals presented with a 32 kHz stimulus ) ., Lrig2 mutants , on the other hand , responded with the same sensitivity as control littermates ., Similarly , Lrig1+/−;Lrig2−/− mutants also demonstrated normal thresholds ., However , loss of either one or two copies of Lrig2 from Lrig1 mutants strongly enhanced the effect , such that the outer hair cell response of Lrig1−/−;Lrig2+/− and Lrig1−/−;Lrig2−/− animals only occurred in response to sounds greater than 55 dB across all frequencies ( Table S1 ) ., To understand how loss of Lrig2 might exacerbate the Lrig1 phenotype , we looked more closely at the nature of the ABR waveforms in all single and double mutant combinations ( Figure 6 ) ., As expected , in Lrig1 mutants the amplitude of the first wave was significantly diminished in response to a range of frequencies and sound intensities ( Figure 6C , D , and Tables S3 and S4 ) ., Combined with the increased thresholds , this suggests that the neural response is decreased because the cochlea is not able to detect sounds with sufficient sensitivity ., Remarkably , despite the lack of any effect on thresholds , Lrig2 mutants showed a similar response: the amplitude of the first wave was significantly decreased relative to controls at multiple frequencies and across sound intensities ( Figure 6C , D , and Tables S3 and S4 ) ., Latencies were also increased ( Figure 6D ) ., Thus , whereas Lrig1 is critical for the initial detection of sound , Lrig2 is required for the subsequent neuronal response ., Since Lrig2 is uniquely enriched in the spiral ganglion neurons throughout life , these findings suggest that Lrig1 and Lrig2 control distinct aspects of cochlear function ., Amplitudes and latencies were even more affected in double mutants , as expected based on the increased thresholds ., Although Lrig1−/−;Lrig2−/− double mutants exhibit a fully penetrant auditory response deficit , the cochlea showed no gross malformations either at E19 ( Figure S4A , B ) or in adults ( data not shown ) ., The cochlear duct had a normal histological appearance , consistent with the absence of any morphological defect at E14 . 5 ( Figure 2 ) ., In addition , immunostaining confirmed the presence of hair cells and neurons in each turn of the cochlea , with spiral ganglion neurites extending to contact hair cells in the organ of Corti ( Figure S4C , D ) ., Similarly , in the few double mutant animals that survived past early postnatal stages , there was no obvious change in the number or organization of hair cells and spiral ganglion neurons ( data not shown ) ., However , the overall pattern of cochlear innervation was clearly disrupted in double mutants , as revealed by immunolabeling for neurofilament , which labels both afferent and efferent neurites ( Figure 7A–C ) ., Whereas control neurites aligned in regularly spaced radial bundles that were clearly separated from each other ( Figure 7A ) , the mutant neurites were noticeably defasciculated and the gaps between the bundles were smaller and present only intermittently ( Figure 7C ) ., More strikingly , the inner spiral bundle ( ISB , bracket ) was reduced , indicating a possible change in the innervation of the cochlea by efferent neurons from the hindbrain ., In contrast , no obvious changes were apparent in Lrig1 or Lrig2 single mutants ( Figure S4E–G ) ., To determine whether Lrig1 and Lrig2 might act redundantly in certain contexts , we looked more closely at the efferent innervation of the cochlea by staining for choline acetyltransferase ( ChAT ) 40 and synaptophysin in single and double mutant animals ., Consistent with results from neurofilament-staining , efferent innervation of the cochlea was noticeably sparser in double mutant animals ( n\u200a=\u200a4 ) compared to controls ( n\u200a=\u200a8 ) ( Figure 7D , F ) ., In contrast , cochleae from Lrig1−/− ( n\u200a=\u200a2 ) and Lrig1+/−;Lrig2−/− ( n\u200a=\u200a4 ) animals were unaffected ( Figure S4G , H ) ., Due to the nature of the crosses used to generate sufficient numbers of double mutant animals , Lrig2−/− single mutant animals were not available for analysis of efferent innervation ., However , the normal pattern of neurofilament staining ( Figure S4G ) together with the lack of defects in the Lrig1+/−;Lrig2−/− cochlea ( Figure S4H′ ) indicates that Lrig2 is not required on its own and that Lrig1 can fully compensate for reduced Lrig2 activity ., On the other hand , cochleae from Lrig1−/−;Lrig2+/− animals ( n\u200a=\u200a4 ) exhibited an intermediate phenotype ( Figure 7E ) , which fits with their diminished auditory responsiveness ., Taken together , these findings indicate that Lrig1 and Lrig2 exert overlapping functions during cochlear innervation , perhaps uncovering a novel role for Lrig proteins in the nervous system ., Moreover , the absence of any obvious morphogenetic or gross cochlear patterning defects argues against the idea that Lrig1 and Lrig2 act redundantly to control any of the major signaling pathways , consistent with their distinct effects in vitro and in cancer ., Here , we used genetic analysis in mice to compare and contrast the effects of Lrig2 and Lrig3 to the founding member of the family , Lrig1 ., By analyzing multiple aspects of inner ear development and function , we found that Lrig1 and Lrig3 cooperate to control inner ear morphogenesis , whereas Lrig1 and Lrig2 appear to affect largely distinct aspects of inner ear function ., Our results highlight the biological significance of all three Lrig genes in vivo and provide insights into the functional diversity of the LRR-Ig superfamily of proteins ., Our findings add to a growing body of work underscoring the similarities between Lrig1 and Lrig3 ., At the molecular level , both Lrig1 and Lrig3 can bind multiple members of the EGF receptor family and show a similar subcellular distribution , with expression on the cell surface and in intracellular vesicles 25 , 41 ., Moreover , both family members also interact with other rTKs 23 , 24 , 27 , indicating that the Lrig ectodomain does not mediate selective binding ., In addition , in vitro studies suggest that both Lrig1 and Lrig3 can act as negative regulators of signaling pathways 16 , 17 , 23 , 24 , 26 , 27 ., Our findings suggest that Lrig1 and Lrig3 also exhibit common activities in vivo ., For instance , Lrig1 and Lrig3 show strikingly similar patterns of expression within multiple tissues throughout development 10 , 28 ., Moreover , Lrig1−/−;Lrig3−/− double mutants exhibit much more dramatic phenotypes than either single mutant ., Importantly , new phenotypes emerge at sites of co-expression , such as the developing utricle and saccule ., Conversely , the strongest phenotype in the Lrig3 mutant ear is in the lateral canal , which is one of the few sites where Lrig1 and Lrig3 do not overlap ., Curiously , although Netrin1 is a key effector of Lrig3 activity in the lateral canal , the atrium develops normally in Netrin1−/− mice ( A . M . N . and L . V . G . , unpublished observation ) , suggesting that Lrig1 and Lrig3 mediate their effects through additional molecules in this region of the inner ear ., Consistent with this idea , neither the anterior nor posterior canal was truncated in Lrig1−/−;Lrig3−/− double mutant inner ears , despite the known role of Netrin1 there 42 ., One likely explanation is that Lrig1 and Lrig3 modulate a broadly active signaling pathway that controls expression of Netrin1 in the lateral canal , but that other target genes are responsible for effects elsewhere in the inner ear ., Indeed , our results suggest that both of these Lrig proteins mediate their effects through key signaling pathways underlying morphogenesis , as Lrig1−/−;Lrig3−/− double mutants die at or before birth with obvious morphogenetic malformations in multiple tissues ., A much more detailed analysis of each affected tissue will be needed to pinpoint the pathways involved ., Although Lrig1 and Lrig3 appear to cooperate during inner ear morphogenesis , each protein also has its own distinct biological functions ., Indeed , the phenotypes already reported in Lrig1 mutant mice indicate that this family member may play a particularly prominent role in EGF signaling and cell proliferation 20–22 ., Similarly , despite the extensive overlap of Lrig1 and Lrig3 in the ear , loss of Lrig1 is sufficient to cause a significant auditory phenotype , as evidenced by increased DPOAE and ABR thresholds ., No auditory phenotypes were detected in Lrig3 mutant mice , in contrast 28 ., Moreover , a role for ErbB receptors in inner ear morphogenesis has not been described , and in fact , broad inhibition of ErbB activity has no effect on canal formation in chicks 25 ., On the other hand , BMP and FGF signaling is critical for inner ear morphogenesis 35 ., Thus , one possibility is that Lrig1 and Lrig3 work together to modulate signaling through BMP or FGF pathways , but that Lrig1 is the dominant regulator of the EGF pathway in vivo ., In support of this idea , Lrig1 and Lrig3 can actually exert opposing effects on ErbB receptor levels in vitro , with Lrig1 reinforcing its effects by decreasing Lrig3 levels 18 ., Hence , the added loss of Lrig3 might not be expected to exacerbate the effects of Lrig1 on EGF signaling in vivo ., Whether Lrig1 and Lrig3 also exert reciprocal effects on the FGF receptor or other putative targets has not yet been examined ., An important step towards resolving these apparent differences will be to determine the nature of the pathways affected by both Lrig1 and Lrig3 in vivo ., Analysis of Lrig2 indicates that this family member has acquired particularly independent functions ., Unlike Lrig1 and Lrig3 , Lrig2 seems to be expressed nearly ubiquitously , although final confirmation awaits the production of reliable anti-Lrig2 antibodies ., Such broad expression is not typical for proteins that function in developmental signaling pathways , which tend to show more restricted patterns of expression ., Notably , despite the fact that Lrig2 is apparently present in every site of Lrig1 expression , no new morphogenetic phenotypes are uncovered in Lrig1−/−;Lrig2−/− double mutant mice ., Thus , Lrig2 is not sufficient to compensate for the combined loss of Lrig1 and Lrig3 , whereas Lrig3 can direct proper inner ear morphogenesis even in the absence of both Lrig1 and Lrig2 ., Although final proof will require analysis of triple mutant animals , the contrasting phenotypes seen in each set of double mutants strongly suggest that Lrig2 does not affect the same pathways as Lrig1 or Lrig3 ., These genetic results fit with previous reports that Lrig2 behaves differently in vitro and in human tumors 18 , 29–31 , 33 ., It is also possible that some residual function persists in Lrig2 gene trap mice , thereby obscuring redundant effects ., However , this seems unlikely given the sensitivity of the inner ear to even subtle changes in signaling levels , as well as the fact that we were able to detect effects on auditory responsiveness in Lrig2−/− mutants | Introduction, Results, Discussion, Materials and Methods | Lrig proteins are conserved transmembrane proteins that modulate a variety of signaling pathways from worm to humans ., In mammals , there are three family members – Lrig1 , Lrig2 , and Lrig3 – that are defined by closely related extracellular domains with a similar arrangement of leucine rich repeats and immunoglobulin domains ., However , the intracellular domains show little homology ., Lrig1 inhibits EGF signaling through internalization and degradation of ErbB receptors ., Although Lrig3 can also bind ErbB receptors in vitro , it is unclear whether Lrig2 and Lrig3 exhibit similar functions to Lrig1 ., To gain insights into Lrig gene functions in vivo , we compared the expression and function of the Lrigs in the inner ear , which offers a sensitive system for detecting effects on morphogenesis and function ., We find that all three family members are expressed in the inner ear throughout development , with Lrig1 and Lrig3 restricted to subsets of cells and Lrig2 expressed more broadly ., Lrig1 and Lrig3 overlap prominently in the developing vestibular apparatus and simultaneous removal of both genes disrupts inner ear morphogenesis ., This suggests that these two family members act redundantly in the otic epithelium ., In contrast , although Lrig1 and Lrig2 are frequently co-expressed , Lrig1−/−;Lrig2−/− double mutant ears show no enhanced structural abnormalities ., At later stages , Lrig1 expression is sustained in non-sensory tissues , whereas Lrig2 levels are enhanced in neurons and sensory epithelia ., Consistent with these distinct expression patterns , Lrig1 and Lrig2 mutant mice exhibit different forms of impaired auditory responsiveness ., Notably , Lrig1−/−;Lrig2−/− double mutant mice display vestibular deficits and suffer from a more severe auditory defect that is accompanied by a cochlear innervation phenotype not present in single mutants ., Thus , Lrig genes appear to act both redundantly and independently , with Lrig2 emerging as the most functionally distinct family member . | The mammalian genome encodes three Lrig family members - Lrig1 , Lrig2 , and Lrig3 ., Lrig proteins share a characteristic extracellular domain that can bind to a variety of signaling receptors , but the three family members show little homology in the cytoplasmic domain ., Lrig1 is a tumor suppressor gene required for normal EGF signaling ., Whether Lrig2 and Lrig3 play similar roles is not known ., To address this gap in knowledge , we compared the expression and function of Lrigs in the mouse inner ear , which is responsible for hearing and balance ., Even subtle changes in the inner ear cause easily detected deficits in hearing and balance , making it an ideal system for analysis of gene function ., We find that Lrigs can act both redundantly and independently in the inner ear , with Lrig1 and Lrig3 cooperating to control morphogenesis and Lrig1 and Lrig2 acting independently to ensure proper cochlear function ., However , loss of both Lrig1 and Lrig2 causes a more severe auditory response deficit and additionally causes a vestibular defect , suggesting some overlapping activities ., Our findings provide new insights into the in vivo functions for the Lrig genes , which play important roles in vertebrate development and disease . | null | null |
journal.pcbi.1006179 | 2,018 | Community interactions and spatial structure shape selection on antibiotic resistant lineages | The human body is home to extraordinarily diverse microbial communities , or microbiomes 1 ., Metabolic interactions among microbial members are now known to play a critical role in host health , including beneficial effects such as protection against pathogens 2 , but also detrimental effects such as obesity , diabetes , and enhanced virulence in polymicrobial infection sites 3–10 ., When a pathogen arises within such diverse and dynamic ecosystems , recent evidence suggests that the efficacy of drug treatment not only depends on the target species and the drug treatment regimen used , but also on the other species present and on the nature of their interaction 11–16 ., Key to controlling antibiotic resistance and managing microbiome health is therefore to understand which treatment strategy is most effective and under what conditions 17–19 ., A major concern when using antibiotics is the potential emergence of de novo antibiotic resistant mutants and/or the competitive release of new or existing antibiotic resistant strains 19 ., Competitive release–when one species or strain increases in density due to the decline in density of its competitors–is for instance one of the main causes of C . difficile infection , especially following treatment with broad-spectrum antibiotics ., Such antibiotic therapy disrupts the normal gut microbiome composition , killing protective resident species , thus leading to the overgrowth of C . difficile 2 ., Competitive release after drug treatment has also been demonstrated in rodent mixed strain malaria infections consisting of genetically distinct drug-resistant and drug-sensitive Plasmodium chabaudi clones 20–22 , with the resistant strain rising in frequency due to the inhibition of its drug-sensitive competitor 20 ., A key factor mediating the strength of competition between susceptible and resistant P . chabaudi strains is resource availability 23 , 24 ., For instance , recent work has shown that resource abundance can lead to the competitive release of the resistant strain and increased virulence 23 ., While there has been a strong focus on competition as a driver of antibiotic resistance and virulence , mutualistic and exploitative interactions among co-infecting bacteria have also been associated with enhanced virulence 4 , 25 , 26 and , in some cases , antibiotic resistance 11–14 , 27 ., For example , Vega et al . ( 2013 ) showed that the pathogenic S . typhimurium was able to enhance its antibiotic tolerance by sensing indole produced by the commensal E . coli 12 ., Another example of cross-species protection against antibiotics involves the beta-lactam susceptible S . aureus , which was protected from beta-lactam antibiotics when enclosed within a layer of resistant beta-lactamase producing P . aeruginosa 13 ., In addition to cross-protection that arises when detoxifying enzymes are released into their local extracellular environment 13 , 28–30 , recent studies have shown that cross-protection can also occur via intracellular detoxification 14 , 31 , 32 ., For instance , Sorg et al . ( 2017 ) recently showed that chloramphenicol-resistant S . pneumoniae can protect chloramphenicol- susceptible S . pneumoniae by degrading chloramphenicol intracellularly , which then lowers the extracellular concentration of antibiotic in their neighbouring environment ., Together , these examples highlight the importance of the interplay between species and strain metabolic interactions , their spatial arrangement , and the mode of resistance in shaping the outcome of antibiotic resistance ., Although we use the terms strain and species interchangeably , we anticipate that competitive interactions will dominate among strains of the same species , whereas more diverse ecological interactions will be more common among species with more distinct metabolic profiles ., Top-down sequencing approaches have revealed important correlations between microbiome composition and host health ( e . g . , 1 ) , yet bottom-up approaches are indispensable for identifying the causal mechanisms underlying microbiome-mediated effects on their host ., A major challenge of using a bottom-up approach , however , is that microbiomes are highly diverse resulting in large networks of microbial interactions that become substantially more complex as diversity increases ., In order to make sense of such complexity , many studies—as the ones described above- have focused on more tractable , well-defined microbiomes with a reduced diversity and therefore smaller interaction networks ., These studies have provided valuable insights into the causal links between microbiome structure and function and host health ., For instance , previous work using two-species co-infection models have revealed the role of co-infection for increased virulence 4 , 9 , 33 and antibiotic tolerance 8 , 12 , 34 ., Here we use a similar qualitative , two species model approach to develop an understanding of the basic principles of antibiotic perturbations on population structure ., Since microbes typically grow in multispecies , surface attached micro-colonies and biofilms 35 , 36 , it is therefore key to understand the impact of species interactions and spatial arrangement on the dynamics of resistance 37 , 38 ., Here we examine this idea by extending an established individual-based computer simulation model of bacterial biofilm growth on an inert surface 39 ., Specifically , we investigate how the nature of the ecological relationship between antibiotic-resistant and sensitive strains or species across the conflict-mutualism continuum affects the community response to antibiotic treatment , and what is the role of spatial structure for such outcome ., To investigate the mechanistic and demographic underpinnings of community-mediated resistance in a spatial context , we implemented a mechanistically-explicit individual-based model consisting of an antibiotic-resistant ( R ) and a susceptible ( S ) species ( or strain ) ., Our model extends an established individual-based framework that simulates the growth and division of cells on an inert surface with explicit diffusion of nutrients and metabolites ( 40 , see Methods for further details on the model ) ., Specifically , here we devise four distinct , metabolically-explicit , media that correspond to four distinct ecological relationships along the conflict-mutualism continuum , namely: “interference competition” where strains or species compete for shared limiting nutrients and release toxins that inhibit the growth of other species ( e . g . , bacteriocin production by , and toxic to , E . coli strains 41 ) ; “exploitation competition” where there is competition for shared nutrients but no release of toxins; “non cross-feeding” where strains or species do not compete for shared nutrients and also do not release by-products; and “cross-feeding” where species release metabolic by-products that can be used by other species for growth ( see Fig 1A for a schematic ) ., As a baseline , we assume that all interactions are symmetric , that is , the nutrients consumed by resistant and by susceptible cells have the same nutritional value , and the toxins have the same inhibitory effect ., We first confirm that our metabolic interactions defined in Fig 1A lead to the expected ecological relationships , by contrasting the growth of each lineage in co-culture with growth in monoculture under equivalent drug and interaction environment ( Figs 1B and S1 ) ., As expected , the interference and exploitation competition media generate strong competition ( more growth in mono-culture than in co-culture ) ., In the non-cross feeding medium , both species are weakly harmed by coculture growth despite the lack of competition for nutrients or the production of antimicrobials that harm the other partner , suggesting competition for space ( Fig 1B ) ., Such space competition arises from competition to gain access to the growing front of the biofilm where nutrient concentrations are highest ., When the density of cells is low , competition for space is non-existent or very weak but it intensifies as the density of cells increases and the nutrient-rich front becomes more crowded ., In contrast , cross-feeding leads to mutualism as seen by an increase in population densities when grown in coculture compared to when grown alone ( Fig 1B ) ., We next ask: how does the type of ecological interaction between resistant and susceptible strains influence the community response to antibiotic perturbation ?, Generally , we find that the density of susceptibles decreases with increasing levels of antibiotics , regardless of their interaction with the resistant strain ( Fig 1C ) ., Whether the density of the resistant strain increases or decreases following antibiotic treatment , however , depends on whether susceptibles harm or help the resistant type ., Specifically , we find that antibiotic perturbation leads to an increase in resistant cell densities when resistant and susceptible species are competitors ( competitive release ) but to a reduction in resistant density when they are mutualists ( mutualist suppression , Fig 1B and 1C ) ., Moreover , mutualistic cross-feeding weakens the negative impact of antibiotic exposure on the susceptible species—that is , susceptibles grow better in the presence of the resistant species than when alone , but this cross-species protection decreases with increasing antibiotic level ( Figs 1B and 1C and S1 ) ., Note that cross-protection is purely measured by growth rates , and is therefore agnostic to the mechanism ., As such , it does not imply a reduction of antibiotic inhibition , and can be due to an increase in intrinsic growth rate , due for instance to the supply of food by cross-feeding ., In the non cross-feeding media , we see a weak competitive release of the resistant type , with the susceptible species doing worse when co-cultured with the resistant species than when alone , again due to competition for space ( Figs 1B , 1C and S1 ) ., This negative effect of coculture growth on susceptibles is likely explained by the fact that resistant cells are able to invade the growing ( nutrient-rich and high antibiotic ) front first due to their growth advantage over susceptible cells whose growth is inhibited by antibiotics ( Fig 1C and 1D ) ., And as a result , the distance between the nutrient-rich front and susceptible cells gradually increases , ultimately leading susceptible cells to starvation ., In sum , our results suggest that antibiotic perturbation leads to the competitive release of the resistant species when the susceptible and resistant species are competitors and to the suppression of the resistant species when the two species are mutualists ., In the model presented above , we assume that there are no fitness costs of resistance ., The no-cost scenario reflects cases where the resistant strain has either acquired compensatory mutations that alleviate the cost of resistance 42 or when resistance is intrinsic , in which case resistant and susceptible strains likely belong to different species 43 ., But in many cases , resistance comes at a fitness cost , for instance , due to the acquisition of resistance via a plasmid or due to chromosomal resistance mutations with epistatic effects 44 , 45 ., Adding cost of resistance to our model , we find that in the absence of antibiotics , the greater the cost of resistance and the strength of competition , the more quickly the susceptible strain outcompetes the resistant strain , as expected ( Figs 2 and S2 ) ., In contrast , we find that the cost of resistance has little impact on resistant:susceptible ratios under strong antibiotic selection , and/or when the two types are mutualists ( Figs 2 and S2 ) ., The results in Fig 2 suggest that competitive release and mutualistic suppression of antibiotics are generally robust to costs of resistance ( a result also robust to changes in inoculum size , S3 Fig ) ., Next , we examine if they are also robust to variation in the initial configuration of the two species community ., In an antibiotic-free environment , competitive communities are known to be more sensitive to initial conditions , while mutualistic communities are more robust 39 ., Specifically , in a community with two equal competitors the most common species is usually favoured ( 39; see also S4A Fig ) ., The reason for this positive frequency-dependent invasion effect is intuitive: all else equal , the species that starts at a higher frequency is more likely to dominate the nutrient-rich front , and cut off its competitor access to nutrients ., In contrast , mutualistic community composition is robust to initial conditions , as partner co-dependency generates negative frequency dependence 39 ., How does applying antibiotics influence competition-mediated sensitivity and mutualism-mediated robustness to initial conditions ?, Unlike in our simple equal competitor scenario without antibiotics , applying antibiotics to the competitive community switches the system to a state where the resistant species is generally favoured independently of initial proportions and spatial distribution ( S4A Fig ) ., This result suggests that the competitive release of a resistant species following an antibiotic perturbation is stable to varying initial demographic conditions ( S4 Fig ) ., Note that this effect occurs because the resistant and sensitive strains have equal competitive abilities ( resistance is cost-free ) ., In the more biologically expected scenario of resistance being costly , the sensitive strain would be favoured at no or low antibiotics ., But as the antibiotic concentration increases and passes some threshold , the benefits of being resistant outweigh its costs , and as such , the resistant strain would now be favoured over the sensitive strain ., Cross-feeding ( and mutualistic ) communities , however , continue to show a robust signature of negative frequency dependence ( i . e . the rare species is favoured ) , reaching a stable equilibrium proportion of susceptibles that is independent of initial proportions and intermixing but depends on the antibiotic concentration ( lower proportion of sensitives for higher antibiotic concentrations ) ( S4A Fig ) ., Moreover , despite a surface blanket of resistant cells ( Fig 1D ) , susceptible cells remain generally intermixed with resistant cells ( S4B Fig ) and their densities remain positively correlated irrespective of initial proportions and intermixing , thereby suggesting that the suppression of mutualists is also robust to varying initial demographic conditions ., So far we have assumed that antibiotic resistance in a focal cell has no impact on the abundance of antibiotic encountered by other cells ., However , resistance often occurs through antibiotic degradation , leading to a reduction in the levels of antibiotics present in the local environment of a focal resistant cell ( i . e . detoxification ) 14 , 31 , 32 , 46 ., In this context , the presence of antibiotics can change the nature of an ecological relationship between species , for example turning a resistant lineage from competitor to protector 31 , 38 , 47 ., Next , we test the impact of protective detoxification by extending our model to allow resistant cells to remove the antibiotic through a process that mimics intracellular enzymatic degradation ( consistent with typical periplasmic beta-lactamase activity or with cytoplasmic antibiotic-modifying or -degrading enzymes , there is no release of antibiotic degrading enzyme; see Methods ) 14 , 32 ., Such process leads to the detoxification of the environment surrounding the antibiotic detoxifying resistant cells , thereby benefiting any nearby susceptible cells ., Previous work has shown that competitors generally tend to segregate as they grow and divide while mutualists tend to mix ( 39 , 48 , 49; and also S4B Fig ) ., Therefore , we predict that cross-protection will be more effective in mutualistic communities because mutualism drives mixing of susceptible and resistant cells , allowing the susceptible cells to benefit from their partner’s local detoxification ., In contrast , because competition leads to species segregation , such spatial separation will limit the ability of susceptible cells to profit from their competitor’s local detoxification ., Consistent with our prediction , we find that detoxification by the resistant lineage provides the greatest rescue effect for susceptible cells when the lineages are engaged in mutualistic exchange ( Fig 3 ) , leading to greater intermixing ( Fig 4 ) ., Detoxification coupled with cross-feeding leads to a reduction in the concentration of antibiotics within the biofilm to levels much lower than those reached in other media ( S5 Fig ) ., This occurs because antibiotic degradation is growth-dependent , and as such the growth-promoting effect of cross-feeding leads to further antibiotic detoxification ., Put another way , susceptibles feed their detoxifier resistant partner , and in turn , benefit from not only increased food provision but also increased detoxification ., When resistant and susceptible species are competitors , however , antibiotic detoxification is not sufficient for susceptibles to persist and they are quickly outcompeted ( Fig 3 ) ., Even in the non cross-feeding medium , where susceptibles are able to persist for longer , their proportion decreases slowly with time ( Fig 3 ) ., And because of their unavoidable growth disadvantage compared to resistant cells , we expect that without a mechanism to drive species mixing , susceptible cells will quickly be buried and starved at the bottom of the biofilm ., To further test the impact of spatial patterning , we ran a new set of simulations where the resistant and susceptible types are now initially segregated ., We predict that spatial segregation should accelerate the decline of non cross-feeding susceptibles because there are no mutual benefits driving the mixing of susceptible cells with detoxifying resistant cells , but only the exploitation of detoxification by susceptibles ., Consistent with this prediction , we find that in the non cross-feeding media , the proportion of susceptibles initially decreases more quickly when resistant and susceptible cells start segregated ( Fig 4A ) ., But in the cross-feeding media , despite a decrease in the proportion of susceptibles at the start of the experiment , susceptible cells mix with their detoxifying partner and reach a stable equilibrium proportion ( Fig 4A–4C ) ., This supports the idea that , in structured communities , the mutual benefits from cross-feeding are crucial to drive species mixing and allow susceptible cells to exploit the antibiotic detoxification of their neighbouring resistant cells ., Furthermore , we can see that when susceptible and resistant species are competitors , the proportion of susceptible cells declines more slowly when they are initially segregated from resistant cells than when mixed , and likely occurs because spatial segregation weakens competition ( Fig 4A ) ., Microbes live in metabolically-connected and spatially extended multispecies communities 35 ., Understanding how microbial communities respond to antibiotic assault is therefore central to improving human health ., Our model suggests that antibiotic perturbation can lead to the competitive release of resistant competitors , or to the mutualistic suppression of resistant partner species–with the outcome tuned by costs of resistance , spatial patterning , and potential for detoxification ., How can these findings inform the development of strategies that aim at promoting microbiome health ?, Crucially , whether one wants to enhance or attenuate competitive release and/or mutualistic suppression depends on the species that are present and the impact they have on their host , that is , whether they help or harm their host 37 ., Target infections are commonly polymicrobial 8 , 33 , 50 , particularly for chronic infections such as in cystic fibrosis lung infections 51 or chronic wounds 52 ., Even for acute and clonal infections , the antibiotic administration still has a strong community context due to impacts on commensal microbiomes , often accompanied by unintentional collateral harm to the host 53 , 54 ., Competitive release occurs when two species compete for shared limiting resources and the removal of one species liberates resources that can be used by the other species , which then increases in density ., Because of higher density , the probability of getting mutations that further improve the fitness of the resistant strain increases , potentially leading to higher rates of resistance 17 ., The role of spatial competition for competitive release and the spread of drug resistance has recently been studied experimentally in microbial colonies 55 ., In the absence of antibiotics , de novo resistant clones may remain trapped and starved within the inner region of the colony , layered with growing sensitive cells ., When high antibiotic concentrations are applied , however , the sensitive cells are killed and eventually removed from the growing front , thus freeing space and resources that can then be taken up by the resistant cells ., A similar finding has been observed in the context of chemotherapy using a computer simulation of tumor growth 56 ., We can therefore see how competitive release is a major concern for health if the resistant strain is a pathogen , as such release may facilitate the rise of antimicrobial-resistant parasites and virulence 17 , 20 , 57 ., Under such conditions , the medical priority is therefore to reduce the rise of resistant strains via competitive release ., Two candidate mechanisms for preventing the emergence of antibiotic resistance under such competitive scenario include maintaining competitive suppression , that is ensuring that cells competing with resistant cells are not inhibited or killed by the antibiotic , and/or targeting resistant cells only , for instance by combining antibiotics with phage therapy 58 , 59 ., Although ubiquitous , microbial competition is not universal 36 ., The oral and gut microbiome , for instance , are replete with species that benefit from the presence of other microbial species 3 , 4 , 60–63 ., Within a polymicrobial infection context , examples are growing where co-infecting bacteria enhance each others’ growth 8 , 26 , 33 ., Such dense , mutualistic communities are of particular concern for controlling infections as their higher densities may hinder clearance of targeted infections , and also , increase the likelihood of emergence of antibiotic-resistant mutants ., So what are the consequences of disrupting communities of antibiotic-resistant and susceptible mutualists ?, As antibiotic concentration increases , we find that the density of the susceptible species decreases , causing the decline of the resistant species ( mutualist suppression ) , but despite such decline in density , the susceptible species grows better in coculture than in monoculture , illustrating a continued impact of the mutualistic exchange ( cross-species phenotypic resistance ) ., From a medical intervention perspective , this means that one can knock down resistant bacteria by hitting mutualistic susceptible species , but also , that one can protect susceptible bacteria by not hitting resistant species that support their growth ., If the susceptible species is a pathogen , such cross-species protection is therefore likely to reduce the efficacy of the antibiotic treatment ., The potential for susceptible cells to be protected against antibiotics is in principle enhanced when resistant bacteria can detoxify their environment ., Empirical evidence of cross-protection of susceptible cells by antibiotic-resistant detoxifying cells is accumulating in the literature , and has been documented both within the same species 14 , 28 , 64–67 and between different species 30 , 68 , and also in different contexts , including cases in which the antibiotics were produced by other community members 69 or exogenously added to the growth medium 14 , 28 , 31 , 64–66 , 68 ., Our work shows that , in spatially extended environments , the emergent spatial arrangement of resistant and susceptible cells influences greatly whether susceptible cells can benefit from detoxification ., Importantly , our results suggest that this protective effect is much more effective in mutualistic communities ., The reason for this effect is that mutualistic partners tend to spatially mix , thus allowing susceptible cells to fully benefit from the detoxification by their spatially proximate mutualistic partner ., In contrast , competition leads to segregation , which ultimately prevents susceptible cells to profit from detoxification ., This finding is in line with previous work on the evolution of cooperation in microbial biofilms showing that competition leads to the formation of clonal groups ( high segregation ) that insulates enzyme-secreting strains from non-secreting strains , thus precluding non-secretors from receiving the benefits of the secreted products 70 , 71 ., Our model assumes that the antibiotic slows down the growth of susceptible cells ( bacteriostatic ) ., Generally , we expect the effect of protective detoxification to be stronger in the presence of bacteriostatic than bactericidal ( killing ) antibiotics ., This is because , with bacteriostatic antibiotics , most cells will eventually be able to resume growth once the concentration of antibiotic drops to levels low enough to permit growth ., In contrast , with bactericidal antibiotics , only the cells that survive the antibiotic assault will be able to grow ., But this killing effect will be reduced in communities where susceptible and resistant lineages intermix , and so less relevant in mutualistic communities ., Although our simulations do not look at the effect of bacteriocidal antibiotics , it would be interesting to test these ideas both theoretically and experimentally ., Our results are based on the assumption that the antibiotic is applied at time 0 , thus before any interaction between resistant and susceptible cells has taken place ., In a clinical context , however , antibiotics will likely be applied to an already established microbial community with a given ecological and spatial structure that reflects the no antibiotic case ., How does the timing of antibiotic administration impact our results ?, We ran a new set of simulations where the antibiotic is now added after 4h , 12h , or 24h of biofilm growth , and found that adding antibiotics at later stages of biofilm development generally favours the susceptible strain , and can even prevent the competitive release of the resistant strain ( S8 Fig ) ., This positive effect of delayed antibiotic administration on the susceptible lineage was stronger at low levels of antibiotics and with detoxification by resistant cells , as expected ., Our model assumes that all the interactions are symmetric ., Any deviations from the baseline dynamics are therefore due to the effect of the antibiotic and/or the cost or resistance ., Interactions between resistant and susceptible strains may , however , be asymmetric , potentially changing the benefits and costs of interactions ., How can asymmetries impact the outcome of antibiotic treatment ?, When resistant and susceptible species are mutualists , their interests are largely aligned ., As such , the cross-feeding and cross-protection benefits received by susceptible cells depend on the growth of its mutualist resistant partner ., If susceptibles start outgrowing the resistant type , the reciprocal benefits of mutualism are diminished , and this will ultimately harm susceptibles due to the lower provision of food and lower detoxification by resistant cells ., This negative feedback can help stabilize the mutualistic interaction , and consequently , the response to antibiotic treatment ., Clearly , hosts play a crucial role in shaping the composition and structure of their microbiomes , not only by providing shelter and food to their resident microbes , but also by producing antimicrobial cells and molecules that inhibit or kill potential enemies 72 ., In turn , microbes affect their host’s fitness and behaviour in various ways , including aid with digestion and supplementation of essential nutrients 73 , as well as protection from pathogens , either directly through competition , or indirectly by eliciting the host’s immune response 74 , 75 ., While a considerable amount of work has been done to understand how within-host community dynamics shape host health , including virulence evolution and drug resistance , the majority of these studies have focused on interactions between parasites and in competition ., Our work suggests that broadening our view of microbe-microbe and host-microbe interactions to include the full conflict-mutualism spectrum is important to elucidate the causes and consequences of intra- and interspecific interactions in host health ., Our work focuses on a two-strain or two-species microbial community living in a simple environment ( one/two resources and a single antibiotic ) , which is undeniably an oversimplified view of natural microbial ecosystems ., Although our results are likely not generalizable to the large diversity of microbiomes , such minimal microbiome approach allows us to identify testable principles of community-mediated antibiotic resistance which can lay the foundations for further research on more complex communities ., For instance , it would be interesting to extend our model to investigate the outcome of antibiotic treatment in a community consisting of isogenic resistant and susceptible strains plus a third resistant or susceptible strain that acts as a mutualist or competitor ., Testing these ideas in more diverse communities and complex environments will help elucidate both general and system-specific principles that determine the outcome of antibiotic therapy ., In sum , our results suggest that the interplay between the metabolic and spatial relationships of resistant and susceptible strains within a community plays an important role in shaping the outcome of antibiotic treatment ., Understanding this relationship can therefore be key to develop effective control strategies ., We expect that the spatial segregation and lower density of competitive communities should facilitate the clearing of an infection because the target sensitive species is isolated from the resistant species , and as such , more vulnerable to antibiotic clearance ., In such competitive scenario , a priority is to maintain competitive suppression , and therefore using narrow-spectrum antibiotics may be more effective than broad-spectrum antibiotics ., In contrast , the spatial mixing and higher densities of mutualistic communities will make it harder to clear the target species ., Under such mutualistic conditions , narrow-spectrum and bacteriostatic antibiotics may therefore be less effective as cross-protection and cross-feeding increase the likelihood that sensitive cells will be able to resume growth once the concentration of the bacteriostatic antibiotic falls below levels permissive for growth ., One potential treatment strategy for the control of mutualistic communities would be to first disrupt the mutualism , e . g . through a diet change that induces a shift in metabolic interaction 76 , to lower mixing of resistant and susceptible cells , thus facilitating clearance ., Testing these ideas experimentally would be an important step towards effectively leveraging the power of antibiotics to promote microbiome health ., Our model is an extension of an established , empirically tested individual-based framework that simulates the growth and division of cells on an inert surface ., This modelling framework has been used extensively over the past decade to understand the behaviour , ecology , and evolution of microbial communities , including studies looking at the drivers of genotypic spatial segregation in biofilms and its consequences for within-species 70 , 77 and between-species 71 cooperation , th | Introduction, Results, Discussion, Methods | Polymicrobial interactions play an important role in shaping the outcome of antibiotic treatment , yet how multispecies communities respond to antibiotic assault is still little understood ., Here we use an individual-based simulation model of microbial biofilms to investigate how competitive and mutualistic interactions between an antibiotic-resistant and a susceptible strain ( or species ) influence the two-lineage community response to antibiotic exposure ., Our model predicts that while increasing competition and antibiotics leads to increasing competitive release of the antibiotic-resistant strain , hitting a mutualistic community of cross-feeding species with antibiotics leads to a mutualistic suppression effect where both susceptible and resistant species are harmed ., We next show that the impact of antibiotics is further governed by emergent spatial feedbacks within communities ., Mutualistic cross-feeding communities can rescue susceptible members by subsidizing their growth inside the biofilm despite lack of access to the nutrient-rich and high-antibiotic growing front ., Moreover , we show that antibiotic detoxification by resistant cells can protect nearby susceptible cells , but such cross-protection is more effective in mutualistic communities because mutualism drives mixing of resistant and susceptible cells ., In contrast , competition leads to segregation , which ultimately prevents susceptible cells to profit from detoxification ., Understanding how the interplay between microbial metabolic interactions and community spatial structuring shapes the outcome of antibiotic treatment can be key to effectively leverage the power of antibiotics and promote microbiome health . | Pathogens -microorganisms that make us sick- often live within dynamic and complex multispecies communities , where they may not only compete for limiting resources but also exchange beneficial resources or services with other resident species ., While antibiotics are commonly used to get rid of such harmful microbes , the community-wide effects of antibiotic treatment and its consequences for antibiotic resistance are still not well understood ., How do competitive or mutually beneficial interactions between antibiotic resistant and susceptible species influence community resistance to antibiotics ?, Here we investigate this question using a computational model ., We find that antibiotic exposure favours the resistant lineage when resistant and susceptible strains are competitors but harms both types when they are mutualists ., With antibiotic-detoxifying resistant cells , cross-protection of susceptible cells is more effective in mutualistic communities because mutualism drives mixing of susceptible and resistant cells ., In contrast , competition leads to their segregation , precluding susceptible cells to profit from their competitor’s local detoxification ., Our findings highlight that knowing not only what species are present but also how they interact with each other and arrange themselves in space is central to understanding antibiotic resistance and to informing the development of strategies that promote microbiome health . | biofilms, antimicrobials, medicine and health sciences, toxins, microbiome, pathology and laboratory medicine, drugs, microbiology, toxicology, toxic agents, antibiotic resistance, nutrition, antibiotics, pharmacology, microbial genomics, antimicrobial resistance, medical microbiology, detoxification, genetics, microbial control, biology and life sciences, species interactions, genomics, nutrients | null |
journal.pgen.1006059 | 2,016 | The Great Migration and African-American Genomic Diversity | The history of African-American populations is marked by dramatic migrations within Africa , through the transatlantic slave trade , and within the United States ( US ) ., By 1808 , when the transatlantic slave trade was made illegal in the US , approximately 360 , 000 Africans had been brought forcibly into the US in documented voyages 1 ., International and domestic slave trade continued to impose long-distance migration on enslaved African-Americans until the end of the Civil War , in 1865 ., By 1870 , the US census reported 4 . 88 million “colored” individuals of which 90% lived in the South 2 ., Despite the ban on slavery , economic and social perspectives for most African-Americans remained bleak ., Better opportunities in the North ( Northeast and Midwest ) and West led millions of African-Americans to leave the South between 1910 and 1970 3 ., This demographic event known as the Great Migration profoundly reshaped African-American communities across the US 4 ., Today , 45 million Americans identify as Black or African-American ., A history of slavery and of systemic discrimination led to increased social , economic , and health burdens in many African-American communities ., Health disparities continue to be compounded by poverty , unequal access to care , and unequal representation in medical research ., To reduce health disparity in research , many cohorts are currently being assembled to encompass more of the diversity within the US 5 , 6 ., These cohorts create opportunities in both medical and population genetics; they also require an understanding of genetic diversity within diverse cohorts ., However , the large-scale migrations and incomplete genealogical records for African-Americans present a challenge for such an understanding ., Previous studies have described the proportions of African , European , and Native American ancestries across individuals 7–13 , the amount of diversity in sequence data 9 , 14 , 15 and inferred admixture models 12 , 16 , 17 ., However , because previous cohorts were not representative of the general African-American populations , they provided limited information about population structure among African-Americans ., Here , we use cohorts including 3 , 726 African-Americans and a total of 13 , 199 individuals geographically distributed across the contiguous US to investigate nation-wide population structure among African-Americans ., We first confirm and refine previous estimates of admixture proportions and timing in the population , and find significant differences in ancestry proportions between US regions ., We then investigate relatedness among African-Americans and European-Americans through identity-by-descent analysis , and identify long- and short-range patterns of isolation-by-distance ., We introduce quantitative models , incorporating both census data and fine-scale migration , to describe these isolation-by-distance patterns and infer migratory patterns in the population ., Integrating quantitative models for admixture , relatedness information , and historical data , we identify ancestry-biased migrations during the Great Migration as a driving force for ancestry and relatedness variation among African-Americans ., The analysis of geographically distributed cohorts through detailed mathematical modeling therefore helps us understand the distribution of genetic diversity in large cohorts and provides new insights into recent human demography ., We analyzed data from three cohorts:, ( a ) Health and Retirement Study 18 ( HRS ) , with 1 , 501 African-Americans and 9 , 308 European-Americans sampled representatively across all US states , and including urban and rural regions;, ( b ) Southern Community Cohort Study 19 ( SCCS ) , including 2 , 128 African-Americans sampled within the southern US in rural locations;, ( c ) 1000 Genomes Project cohort of 97 individuals of African ancestry from the southwest USA 20 ( ASW ) ., Genotypes were obtained on Illumina Human Omni 2 . 5M and Human 1M-Duo platforms , and joint analyses were performed on a common set of 553 , 795 high-quality SNPs ( for detailed information , see Materials and Methods and S1 and S2 Tables ) ., Individual genomes carry genetic material from multiple ancestral lineages , and each diploid locus derives ancestry from two distinct lineages ., We used RFMix 11 together with 1000 Genomes Project panels from Africa , Europe , and Asia to identify the most likely continental ancestry at each locus for individuals in the cohorts ( Fig 1D , S2 Fig and Materials and Methods ) ., Here , continental ancestry is defined as the inferred location of the ancestral lineage prior to the advent of transatlantic travel ., The overall proportion of African ancestry is substantially higher in the SCCS and HRS than in the ASW and the recently published 23andMe cohort 12 ( Table 1 ) ., The HRS cohort can be thought of as representative of the entire African-American population , while the SCCS focuses primarily on individuals attending community health centers in rural , underserved locations in the South ., By contrast , the sampling for the ASW and 23andMe did not aim for specific representativeness , and the ascertainment in the 23andMe cohort might have enriched for individuals with elevated European ancestry ( see Materials and Methods and discussion in 12 ) ., In the HRS , average African ancestry proportion is 83% in the South and lower in the North ( 80% , bootstrap p = 6 × 10−6 ) and West ( 79% , p = 10−4 ) ( Fig 1 ) ., Within the SCCS , African ancestry proportion is highest in Florida ( 89% ) and South Carolina ( 88% ) and lowest in Louisiana ( 75% ) with all three significantly different from the mean ( Florida p = 0 . 006 , South Carolina p = 4 × 10−4 , and Louisiana p < 10−5; bootstrap ) ., The elevated African ancestry proportion in Florida and South Carolina is also observed in the HRS and in the 23andMe study 12 , but Louisiana is more variable across cohorts ( Fig 1E ) ., As expected , European ancestry proportions largely complement those of African ancestry across the US ., Because recombination breaks down ancestral haplotypes over time ( Fig 1D ) , the length of continuous ancestry tracts is informative of the time of admixture , with shorter tracts reflecting older admixture ., We inferred the timing of admixture using Tracts 16 , which fits a demographic history to the observed distribution of tract lengths ( see Materials and Methods for details and S4 Table for confidence intervals ) ., Because of the small number of Native American tracts , even a small amount of spurious Native American ancestry assignments can bias the inference ., Thus , we first considered a model with two source populations: African and non-African ., Assuming a single admixture event , we estimated the time of admixture onset g , where g = 1 means that the parents of the individual are the founders of the admixed population and that the current individual represents the first admixed generation ., For HRS , we inferred a timing of g = 5 . 8 generations ago ( S8 Fig ) ., The estimated year of birth of the first admixed children is T = Ts − ( g − 1 ) τ , where Ts = 1939 . 8 is the average year of birth of HRS individuals and τ is the generation time ., Individuals born τ years earlier should be 1 generation closer to the onset of admixture ., Correlating birth year and inferred admixture time within our cohort ( Fig 2D ) , we inferred τ = 27 . 4 ( r2 = 0 . 88 , p = 10−7 ) , which leads to an admixture year of 1808 ( bootstrap 95% CI: 1805 . 5 , 1810 . 4 ) ., Note that 1808 represents the admixture time that best explains the data under the assumption of a single admixture event ., The narrow confidence interval is , therefore , no guarantee that something exceptional occurred between 1805 and 1810 ., To investigate the role of modeling assumptions in admixture time estimate , we considered more general models ., A model allowing for two phases of European admixture outperforms the single-pulse model for HRS and SCCS ( see Materials and Methods ) ., In HRS , it suggests a first admixture event in 1740 ( 8 . 3 generations ago; bootstrap 95% CI: 1711 . 6 , 1744 . 2 ) and a second pulse , of approximately equal size , in 1863 ( 3 . 8 generations ago; bootstrap 95% CI: 1852 . 9 , 1865 . 9 ) ( S8 Fig and Materials and Methods ) ., Mean birth year in SCCS is Ts = 1946 . 9 , supporting a single admixture event in 1802 ( 6 . 3 generations ago; bootstrap 95% CI: 1799 . 2 , 1803 . 6 ) , or two events in 1714 and 1854 ( 9 . 5 and 4 . 4 generations ago; bootstrap 95% CIs: 1704 . 6 , 1739 . 7 and 1849 . 8 , 1868 . 7 ) ( Fig 2A , S8 Fig and S4 Table for confidence intervals ) ., The two-pulse model remains a coarse simplification of the historical admixture process , but the data strongly supports ongoing admixture , predominantly before or around the end of the Civil War ., This is consistent with historical accounts of “a marked decline in both interracial sexual coercion and interracial intimacy” 21 at the end of the Civil War ( see also Ref . 22 and references therein ) ., The limited role of early 20th century admixture is further supported by the similarity in the inferred single-pulse time to admixture in all HRS census regions ( between 5 . 4 and 6 . 2 generations ago , S11 Fig ) and all cohorts , which is easily explained if most admixture occurred in the South prior to the Great Migration ., The similar levels of African ancestry for all age groups within the HRS also support limited European admixture between 1930 and 1960 ( Fig 2D ) ., Importantly , more recent admixture is not represented in the SCCS and HRS cohorts; only two participants were born after 1970 ., Time estimates point to admixture occurring when most ancestors to present-day African-Americans lived in the South ., The regional differences in ancestry seen in Fig 1 are therefore unlikely to be caused by differences in recent admixture rates , and the large influx of migrants from the South would have strongly attenuated any earlier differences ., An alternate explanation for regional differences in ancestry proportions is that individuals with higher European ancestry were more likely to migrate to the North and West during the Great Migration , a scenario we refer to as ancestry-biased migration ., To validate the ancestry-biased migration model , we compared ancestry proportions of HRS individuals according to their region of birth , residence , and migration status ., European ancestry proportions in African-Americans who left the South ( 16 . 5% ) is elevated compared to individuals who remained in the South ( 15 . 3% , bootstrap p = 0 . 04 ) , confirming that ancestry-biased migrations continued at least to the mid-20th century ., These migrants had substantially less European ancestry than African-Americans already established in the North ( 20 . 9% ) and West ( 25 . 0% ) ( Fig 2E ) ., Since the latter two groups received large contributions from the first wave of the Great Migration , this suggests that the proportion of European ancestry in first-wave migrants was higher than in the second wave—i . e . , that there was stronger ancestry bias during the first wave of migration ., This change over time in ancestry-biased migration is consistent with historical accounts that southern African-American migrants to northern cities during the later stages of the Great Migration had darker complexion than North-born African-Americans ( see 23 , p . 179 ) ., The change could be explained by better social opportunities available to individuals with higher levels of European ancestry: Individuals with wealth and education were much more likely to migrate in the first wave of the migration ( see 23 , p . 167 ) ., Fig 2E shows that despite the ongoing ancestry bias , the migrations of HRS participants led to more uniform ancestry proportions across regions ., Interestingly , the proportion of African ancestry among African-Americans increased in all four US regions between the time of birth and the time of survey of participants: The ancestry bias caused migrants to have levels of admixture between those of the South-born and North-born individuals ., Their departures and arrivals both increased the regional African ancestry proportions ., Out of 1 , 491 non-Hispanic African-Americans in HRS , 11 individuals have more than 5% Native American ancestry ., Within SCCS , this proportion is only 8 out of 2 , 128 individuals ., The ASW cohort , with 8 out of 97 individuals above this threshold , is a clear outlier ., The other 89 individuals , however , have similar amounts of Native American ancestry to the other studies ., If we filter out individuals contributing more than 5% Native American ancestry from each cohort , the proportion of Native American ancestry in the remaining individuals is close to 1 . 1% in the SCCS , in all HRS census regions , and in the ASW ., The filtered SCCS Louisianans have significantly more Native American ancestry ( 1 . 6% , bootstrap p = 2 × 10−5 ) , and South Carolinians have less ( 0 . 09% , p = 2 × 10−5 ) , than the mean Native American ancestry ., We did not find a global correlation between European and Native American ancestry , except within Louisiana ( S4 Fig ) ., A three-population admixture model accounting for Native American admixture confirmed the predominantly early , multiple-phase European admixture and suggested that Native American admixture occurred even earlier , consistent with previous findings 12 ., Inferred dates of admixture onset are 1494 ( bootstrap 95% CI: 1478 . 8 , 1516 . 0 ) for the HRS ( S9 and S10 Figs ) and 1486 ( bootstrap 95% CI: 1475 . 4 , 1499 . 4 ) for the SCCS ( Fig 2B and 2C ) , as described in Materials and Methods ., The presence of a small amount of spurious , short segments of inferred Native American ancestry could bias the inference toward these unrealistically early dates ., The lack of longer Native American segments nevertheless suggests that most Native American ancestry in African-Americans results from contact in the early days of slavery ( see , e . g . , 24 ) ., The three-population model suggests more recent European admixture dates than the two-population model , but with a higher proportion of migrants in the earlier migration ., Finally , a three-population model with continuous European admixture provided qualitatively similar estimates to the two-pulse model , with an early onset of Native American admixture ( 1482 ) and European migration spanning the period between 1758 to 1887 ., Direct admixture between African-Americans and Native Americans is further supported by the observation that the proportion of Native American ancestry in HRS African-Americans ( 1 . 2% ) is comparable to that in HRS European-Americans ( 1 . 5% ) ., This proportion is therefore much higher than would be expected if the Native American contribution occurred through European admixture ., Despite substantial disagreement as to the specific dates , all models agree on European admixture occurring predominantly prior to the Civil War ., Along the X chromosome in the HRS , we estimate 84 . 82% African ancestry , 12 . 89% European ancestry , and 2 . 29% Native American ancestry ( bootstrap 95% CI 2 . 14% , 2 . 45% ) ., The higher proportion of African ancestry along the X compared to autosomes is consistent with previous studies 12 , 17 and the historical record of early admixture occurring predominantly through coerced sexual interaction between European-American males and African-American females 21 ., A model with a single pulse of admixture ( as considered in 12 ) applied to the present data suggests 28 . 6% Europeans among male contributors , but only 5 . 2% among female contributors ., By contrast , it suggests almost no contribution from Native American males , and 3% from Native American females ., The US Census includes a separate category for Hispanic/non-Hispanic ethnicity ., In HRS , 32 African-Americans have self-identified as Hispanics ( of which only 10 are within the contiguous US ) ., Hispanics often trace ancestry to regions colonized by Spain and Portugal , and where Native American populations contributed a higher proportion of the present-day gene pool compared than in the US ., Genetic ancestry within this group is indeed distinct from the bulk of the non-Hispanic African-American population in at least two ways: elevated Native American ancestry and a higher genetic similarity to southern European populations ( S5 and S6 Figs ) ., The correlation between southern European and Native American ancestries also holds in individuals who do not self-identify as Hispanic , particularly in Louisiana ( see Materials and Methods ) ., Individuals with elevated Native American and southern European ancestry would not be identified by self-reported ethnicity or by genetic estimates of African/non-African ancestry , yet they may have distinct response patterns to medical tests 25 , 26 ., The classical isolation-by-distance model predicts that genetic relatedness between individuals decreases as their geographic distance increases 27 ., However , large-scale migrations can dramatically alter this picture 28 ., To investigate the effect of recent migrations on patterns of genetic relatedness within African-Americans , we consider genetic segments that are identical-by-descent ( IBD ) between pairs of individuals ., We focus on long IBD segments ( l ≥ 18cM ) , which correspond to an expected common ancestor living within the last 8 generations ( see Materials and Methods ) and are therefore informative of recent demography ., Fig 3A , 3B , S12 and S15 Figs show the mean pairwise relatedness among seven geographic regions in the US for African-Americans and European-Americans ., Here , the relatedness of two individuals is defined as the total length of the genome shared through long IBD segments ., These recent relatedness patterns differ markedly between African-Americans and European-Americans ( compare Fig 3A and 3B ) : African-Americans exhibit a distinct enrichment in South-to-North relatedness along the main historical migration routes ., To compare these relatedness patterns with recent migration data , we used the 20th century US census data and a simple coalescent model to estimate the expected relatedness between geographic regions ( see Materials and Methods ) ., Census-based predictions ( Fig 3D ) are correlated with IBD-based observations ( Fig 3A ) if we consider non-identical pairs of regions ( Mantel test p = 0 . 019 ) ., Limiting the comparison to the South-to-North and South-to-West relatedness , to capture migration routes specific to the Great Migration , yields p = 0 . 063 ( using the 2010 region of residence ) and p = 0 . 015 ( using place of birth ) ( see Materials and Methods ) ., Fig 3C and S16 Fig show the relatedness between African-Americans and European-Americans ., African-Americans across the US are more related to European-Americans from the South than to those from the North or West ( bootstrap p < 0 . 0002 ) ., In addition , European-Americans from the South tend to be more related to African-Americans in the North than to those in the South ( bootstrap p = 0 . 11 ) ., This increased relatedness with increased distance is unusual in population genetics , but is easily explained: The ancestry-biased migration is also a relatedness-biased migration ., The reduced relatedness between northern European-Americans and African-Americans may also be reinforced by recent European migration , because the new migrants were more likely to settle in the North but were less likely to be related to African-Americans ., Despite the unusual long-range relatedness patterns , identity-by-descent decays with distance within African-American communities in the South , reflecting isolation-by-distance ( S19 Fig ) ., To understand how migrations affect isolation-by-distance and identity-by-descent , we introduce a quantitative model taking into account a diploid population density n and spatial diffusion constant D . In short , the displacement between parental birthplace and offspring birthplace of individuals is modeled as an isotropic random walk; the distribution of the times t to the most recent common ancestor of two individuals separated by distance R is calculated under a coalescent model; and the amount of genetic material shared IBD given a common ancestor at time t is computed as in Ref ., 29 ., Under this model , we can calculate the expected fraction of genome shared IBD between two randomly chosen individuals separated by a distance R . If we consider only IBD segments of length in ℓ = lmin , lmax ( in Morgans ) , we find, E ℓ f | R = 1 16 π n D 2 K 0 R r min - K 0 R r max + R r min K 1 R r min - R r max K 1 R r max ( 1 ), where r min , max = D / l min , max , and Kα ( x ) is the modified Bessel function of the second kind 30 ( see Materials and Methods ) ., Fig 3E shows the presence of a background level of IBD relatedness in both African-Americans and European-Americans even at long distances ., This could be attributed to false positives in IBD calling , to relatedness originating prior to ancestral migrations from Europe and Africa into the Americas , or to a small amount of distance-independent migration ., We account for these effects in our model by introducing an additional distance-independent ( constant ) term ., Using IBD segments longer than 18cM , we estimate the background IBD for African-Americans and European-Americans in HRS to be bAFR = 0 . 048cM and bEUR = 0 . 011cM respectively ( see Materials and Methods for details ) ., We estimate population density nAFR = 1 . 9km−2 and diffusion constant DAFR = 63 . 5km2/generation for African-Americans across the US , and nEUR = 7 . 6km−2 and DEUR = 59 . 6km2/generation for European-Americans ( Fig 3E ) ., The ratio of European- to African-American inferred population density is therefore 3 . 9 ., According to the 2010 US Census , 13% of the total population have self-identified as “Black or African American alone” and 72% self-identified as “White alone” ., The ratio of European- to African-American population size from the census is 5 . 5 , in good agreement to our estimate above ., Interestingly , the root mean squared displacement per generation , 2 D × 1 generation ∼ 15 − 16 km , shows comparable local migration rates in European-Americans and African-Americans despite the different histories and population densities ., This root mean square ( RMS ) displacement is much less than the contemporary RMS parent-offspring dispersal in the US , estimated at 989km , but within the range of other modern human populations ( 2 . 6–300km ) 31 ., RMS displacement is heavily influenced by the largest displacement , and the latter study found approximately 27% of parent-offspring displacements in the US to be over the 1000km range ., Such long-range migrations did not appear to leave a strong signature of isolation-by-distance in our IBD data and were captured by the uniform background term in our model ., The RMS displacement in our model therefore does not account for such long-range migrations ., The history of African-American populations combines strong ties to place with large-scale migrations 4 ., This comprehensive study shows the combined effects of fine-scale population structure , large-scale migrations , and admixture in shaping genetic diversity among African-Americans ., Detailed models of genomic diversity recapitulate known historical events , such as the travel routes used during the Great Migration 4 , 23 and the timing , amount , and geography of admixture between African , European , and Native ancestors 22 , 24 , 32–34 ., They also quantify demographic effects that were less well characterized , such as ancestry-biased migration and the geographic patterns of relatedness among African-Americans ., The observed ancestry-biased migrations of African-Americans suggest that the differences in social opportunity afforded to individuals with different levels of European ancestry at the time of the Great Migration 23 contributed to shaping the genetic population structure of contemporary African-Americans ., The observed patterns of relatedness have consequences for genetics research ., Long IBD segments are often inherited from a recent common ancestor and are likely to carry shared but recent mutations ., Such variants are more likely to be deleterious than older variants and are therefore prime targets for disease-mapping studies of rare traits 35 ., Considering our analysis of long-range IBD sharing across the US , we expect rare monogenetic traits to be more often shared over long distances among African-Americans than among European-Americans , particularly along the routes of the Great Migration ., Yet , their spatial distributions over short ranges should be as structured as in European-Americans ., Despite the overall correlation in regional admixture proportions among the SCCS , HRS , and 23andMe cohorts , significant differences remain in nation-wide and regional ancestry proportions ., Such differences likely result from sampling biases that correlate with existing population structure through geography , urban/rural status , wealth , education level , and identity ., Detailed sampling and sociodemographic modeling should therefore inform the design and analysis of large genetic cohorts that include African-Americans , as well as further efforts to understand the genetic makeup of African-American communities ., The use of these samples for the present study was approved by the IRB at McGill University and Stanford University , where the analyses were performed ., We used the genotype data of 12 , 454 individuals from the Health and Retirement Study 18 ( HRS ) , genotyped on the Illumina Human Omni 2 . 5M platform , and of 2 , 169 African-American individuals from the Southern Community Cohort Study 19 ( SCCS ) , genotyped on either Illumina Human Omni 2 . 5M or Human 1M-Duo platforms ., The HRS cohort includes 1 , 649 individuals who self-identified as African-Americans ( non-ambiguously in both HRS Tracker and dbGaP databases ) and 10 , 432 individuals who self-identified as European-Americans ., There are also 366 individuals labeled as “Others” whom we have not used in our main analyses ( except in a PCA analysis , discussed below ) ., The remaining 7 individuals have ambiguous , non-matching race identifiers in HRS Tracker and dbGaP , and we have , thus , excluded them from our analyses ., We performed comparisons with data from 23andMe 12 and from 97 individuals of African ancestry from the southwest USA ( ASW ) from the 1000 Genomes Project ( at ftp://ftp . 1000genomes . ebi . ac . uk/vol1/ftp/release/20130502/supporting/hd_genotype_chip/ ) 20 ., The 23andMe cohort includes many African-American individuals and has been the subject of a detailed population genetic analysis 12 , and the ASW cohort has been a reference African-American population in recent studies ., However , these two cohorts were not meant to be representative of the US population ., The 23andMe database has a complex ascertainment scheme , which may cause biases in ancestry and socioeconomic status ., In particular , biases in regional representation and a small amount of survey response errors might lead to a lower European ancestry proportion ., These possible biases are described in detail in 12 ., Similarly , the ASW cohort was assembled from duos and trios with at least one Oklahoma resident , but with no attempt to reach geographic or demographic representativeness ( Morris Foster , personal communication ) ., For comparisons with the 23andMe study , we used the global ancestry proportions reported in 12 , because the genotype data is not publicly available ., The global ancestry proportions reported in the 23andMe study are calculated by first using their in-house local ancestry assignment pipeline and then aggregating the results across the genome , as described in detail in 12; we employ a similar scheme , described below in detail ., The HRS genotype data that we received had been already quality controlled , filtered , and phased ., The SCCS cohort comprises data from 648 individuals in a breast cancer study ( genotyped on Illumina Omni 2 . 5M platform ) and 760 individuals in a prostate cancer study , 484 individuals in a lung cancer study , and 277 individuals in a colorectal cancer study ( genotyped on Illumina Human 1M-Duo ) ., All genotyped individuals were either cases or controls in their respective nested case-control studies ., We converted the lung cancer dataset from human genome assembly hg18 to hg19 using the LiftOver utility from the UCSC Genome Bioinformatics Group and merged the four separate SCCS datasets into one using PLINK 1 . 9 36 ., During the merge process , we removed markers to which more than one name was assigned at the same position along a chromosome; removed markers with missing genotype calls; corrected unambiguous strand misassignments and removed ambiguous strand ( mis ) assignments; removed multi-allelic markers; and , finally , filtered the data for missing calls 37 first based on genotypes ( PLINK argument --geno 0 . 0125 ) and then based on call rates per individual and minor allele frequency ( PLINK arguments --mind 0 . 0125 --maf 0 . 01 ) ., The final SCCS dataset contains 2 , 128 individuals and 585 , 527 variants after these steps ., We then used the same process to merge the HRS data with those of SCCS and ASW , resulting in a single dataset in PLINK format with 14 , 679 individuals and 553 , 795 variants ., Performing a PCA on the data ( pruning for LD leaves 77 , 902 markers ) , we found no batch effects ( see S1 Fig ) ., We then phased the merged data with SHAPEIT2 38 ( default arguments ) , and converted the output to PLINK format ( while preserving the phasing information ) using genetic map information from the 1000 Genomes Project data ( at http://mathgen . stats . ox . ac . uk/impute/data_download_1000G_phase1_integrated_SHAPEIT2_9-12-13 . html ) ., Geographic information in HRS is usually provided in the form of US census regions and divisions ., We have used these locales in the ancestry analyses ., ZIP code information for HRS study participants is available , but use of this data is restricted ., We used zip code data only for the fine-scale spatial analysis of identity-by-descent relatedness ., For SCCS , latitude and longitude coordinates of clinics were available ., In the IBD analysis , we assigned the ASW individuals to the West South Central census division ( see , e . g . , https://catalog . coriell . org/1/NHGRI/Collections/1000-Genomes-Collections/African-Ancestry-in-SW-USA-ASW ) ., In terms of geographic locations , we restrict our analyses to the census divisions in the contiguous United States ( i . e . , Pacific , Mountain , West North Central , East North Central , Middle Atlantic , New England , West South Central , East South Central , South Atlantic ) ., For the individuals in HRS , we only consider the ones born in the contiguous US who , at the time of sampling in 2010 , also lived in the contiguous US; this reduces our sample size in HRS to 10 , 974 individuals of which 1 , 501 are self-identified African-Americans and 9 , 308 are self-identified European-Americans ( with the remaining individuals being classified as “Others” ) ., There are 4 additional individuals satisfying the geographic constraints above but who have discordant race identifiers in two different data files provided with the cohort data; these were removed from any downstream analysis ., Among the unambiguous self-identified African-Americans and European-Americans mentioned above , there are respectively 10 and 427 individuals also self-identifying as Hispanics ., The former 10 individuals are only included in our analysis of Hispanics status ., In S1 Table , we summarize a few characteristics of the HRS African-American and SCCS cohorts , namely , the number of sampled individuals , the number of males and females , the number of Hispanics ( if specified ) , and the locale ., African-American sample sizes in the Ne | Introduction, Results, Discussion, Materials and Methods | We present a comprehensive assessment of genomic diversity in the African-American population by studying three genotyped cohorts comprising 3 , 726 African-Americans from across the United States that provide a representative description of the population across all US states and socioeconomic status ., An estimated 82 . 1% of ancestors to African-Americans lived in Africa prior to the advent of transatlantic travel , 16 . 7% in Europe , and 1 . 2% in the Americas , with increased African ancestry in the southern United States compared to the North and West ., Combining demographic models of ancestry and those of relatedness suggests that admixture occurred predominantly in the South prior to the Civil War and that ancestry-biased migration is responsible for regional differences in ancestry ., We find that recent migrations also caused a strong increase in genetic relatedness among geographically distant African-Americans ., Long-range relatedness among African-Americans and between African-Americans and European-Americans thus track north- and west-bound migration routes followed during the Great Migration of the twentieth century ., By contrast , short-range relatedness patterns suggest comparable mobility of ∼15–16km per generation for African-Americans and European-Americans , as estimated using a novel analytical model of isolation-by-distance . | Genetic studies of African-Americans identify functional variants , elucidate historical and genealogical mysteries , and reveal basic biology ., However , African-Americans have been under-represented in genetic studies , and relatively little is known about nation-wide patterns of genomic diversity in the population ., Here , we study African-American genomic diversity using genotype data from nationally and regionally representative cohorts ., Access to these unique cohorts allows us to clarify the role of population structure , admixture , and recent massive migrations in shaping African-American genomic diversity and sheds new light on the genetic history of this population . | hispanic people, african americans, population genetics, census, mathematical models, ethnicities, research design, paleontology, population biology, research and analysis methods, random walk, geography, native americans, mathematical and statistical techniques, paleogeography, people and places, population metrics, survey research, earth sciences, genetics, population groupings, biology and life sciences, population density, evolutionary biology | null |
journal.pgen.1002842 | 2,012 | Rescuing Alu: Recovery of New Inserts Shows LINE-1 Preserves Alu Activity through A-Tail Expansion | Long INterspersed Element-1 ( LINE-1 or L1 ) and the Short INterspersed Element ( SINE ) Alu are non-long-terminal-repeat ( non-LTR ) retroelements that are responsible for approximately one third of the human genome 1 ., Due to their ability to randomly insert throughout the genome 2 , both L1 and Alu are capable of disrupting critical genes and causing a large diversity of genetic diseases 3–6 ., The creation of an engineered L1 assay system specifically designed to rescue de novo L1 inserts in a culture system demonstrated that L1 insertion contributes significantly to genetic instability through retrotransposition-mediated deletions and rearrangements 7–10 ., This assay has the added advantage of providing a valuable tool for analyzing aspects of the L1 insertional mechanism under controlled experimental conditions 11–13 ., Computational analyses further corroborated that both Alu and L1 insertions are associated with genomic loss , rearrangements and structural variation in humans 14–16 ., Prior to our development of a similar assay system for SINES , there are very few published details of recovered de novo SINE insertions in culture ., Two previous reports account for a total of 12 fully characterized de novo Alu insertion events in culture 17 , 18 ., One of these approaches utilized an untagged AluSx to transfect cells and the Alu inserts were then detected by “panhandle” PCR amplification using an anchor that is attached to the restriction digested cellular DNA ., The researchers evaluated a total of 101 PCR products and found that seven were bona fide Alu insertion events 18 ., The other five Alu insertion events were recovered using a tagged Alu and inverse PCR approach 17 , 18 ., An additional published report describes eight inserts from two tagged rodent SINEs 19 ., Thus , only 20 de novo SINE inserts from cell culture have been characterized prior to the work reported here ., Because these data arose from different approaches , using different SINEs , and different cell lines , generalizations from the data become difficult ., New high-throughput approaches have yielded large amounts of data on mobile element insertion , including somatic events observed in cancer samples 20 and brain 21 ., However , these approaches are limited by short sequence reads , the inability to sequence through homopolymeric A-tails , and high difficulty of recovery and validation of “singleton” events ( very rare events , i . e . , somatic insertions ) due to the inability to refer back to a reference clone ., Although in silico and high throughput sequencing analyses provide valuable insights into retroelement activity , a tissue culture assay system provides a controlled genetic environment during retrotransposition that confers the ability to distinguish between retrotransposition-mediated events and those that occur post-insertionally with the added advantage of being able to manipulate SINE sequences for experimental evaluation ., Here , we present the adaptation and development of an engineered recovery-construct that allows for the rescue of inserted tagged SINE elements in a tissue culture assay system and provide detailed data from over 200 rescued de novo Alu inserts ., Because SINEs are transcribed by RNA polymerase III ( pol III ) , several obstacles introduced by the RNA pol III transcriptional requirements must be overcome to develop experimental methods to investigate the mechanistic aspects of Alu retrotransposition ., Due to these constraints , prior methods for the recovery of SINE inserts in culture have been mostly limited to inverse PCR 17 , 18 ., As an alternate approach , we have developed an Alu recovery system by redesigning the existing Alu-neoTET vector 17 , following the strategy used to create the L1 recovery vector 7 , 8 ., The principle of the method is shown in Figure 1A ., We performed extensive modifications and adaptations of the Alu construct 17 ( Figure 1B ) ., First , a bacterial promoter ( EM7 ) was inserted upstream of the neoTET cassette to obtain kanamycin resistance in bacterial cells ., We then introduced a minimal γ origin of replication ( 305 bp ) of plasmid R6K 22 , 23 , which was sequence modified to allow RNA polymerase III ( pol III ) transcription ., The R6KγORI was selected due to its smaller size ., Specific sections of the R6KγORI were changed by site directed mutagenesis to eliminate runs of four or more thymidine residues that could function as internal RNA pol III terminators ( details in Materials and Methods ) ., Finally , in order to analyze A-tail expansion , we substituted the original homopolymeric A-tail with a dA-rich sequence containing non-A disruptions ( Figure 1B ) ., As expected , the added sequence length ( 439 bp ) and the variation in A-tail length and composition 24 reduced the retrotransposition efficiency of the Alu rescue construct to close to 50% when directly compared to the parental construct ( Figure 1C ) ., The retrotransposition efficiency of the Alu rescue construct increases when using a highly efficient driver vector expressing only L1 ORF2p ( Figure 1C ) ., However , the added length to the tagged Alu RNA did not appear to contribute to 5′ truncation of the Alu inserts , as fewer than five percent were truncated ( see details below ) ., We recovered a total of 226 Alu inserts from transfected HeLa cells ( complete sequence details of the insertions are available in Text S1 and Table S1a ) ., Because transfection of the L1 ORF2 protein alone supports Alu retrotransposition in HeLa 17 , 25 , we wanted to determine if ORF2-driven Alu inserts differed from those driven by a full length L1 ., We analyzed de novo Alu inserts driven by ORF2 alone ( N\u200a=\u200a178 ) or driven by full-length L1 ( N\u200a=\u200a48 ) for comparison between the sets ., Overall , we found no significant differences between Alu inserts driven by full-length L1 vs . ORF2 alone ( Tables 1 and 2 and Figure 2 ) ., Therefore , we primarily report the combined observations of all Alu inserts ., We obtained sequences from both 5′ and 3′ genomic flanking sequence of the inserts ( Text S1 and Table S1a ) ., Of the fully characterized de novo Alu inserts , the vast majority ( ∼96% ) exhibited the hallmark characteristics of retrotransposition: direct repeats flanking the insert , a 3′ oligo dA rich tail and a target site resembling the L1 endonuclease consensus sequence 7 , 9 , 26–28 ., Atypical insertions ( lacking the retrotransposition hallmarks ) were associated with genomic deletions or rearrangements ( details below ) ., The observed target consensus site for the inserts ( 5′-TTTT/AA-3′ ) is identical to the known preferred L1 endonuclease cleavage site 8 ( Figure 2A ) , suggesting that most Alu inserts in our culture system initiated by the conventional endonuclease-dependent target primed reverse transcription ( TPRT ) mechanism ., The direct repeats ranged from 5–27 bp , with a 14 . 0±3 . 0 bp average ( Table 1 ) ., Eight of the recovered events ( 3 . 5% ) resulted in an Alu insert with a 5′ truncation ., This is less than half of what is observed in the genome ( ∼10% of Alu elements are 5′ truncated ) 1 , 29 ., As proof of the versatility of the method , we modified our construct to the study of other SINE elements ., We recovered seven inserts from two rodent SINEs by substituting the BC1 or B2 sequences for the Alu sequence in the rescue vector 30– ., Sequence analysis revealed that the fully characterized de novo inserts ( five B2 and one BC1 ) also contained the endonuclease target site and insertion characteristics of typical L1-mediated retrotransposition ( Text S2 and Table S1c ) Our analyses of the recovered Alu inserts determined that less than four percent of the inserts ( 8 of 226; 3 . 5% ) lack the typical characteristics of TPRT-mediated Alu insertions ., Six of these insertions ( 2 . 7% ) contain two features indicating that the insertion likely completed through recombination with an existing Alu present at the genomic site ( Text S1 and Table S1a ) ., First , the recovered sequences of these clones contain a chimeric sequence between the genomic and the tagged Alu ., Secondly , they lack the characteristic flanking direct repeat ., In several cases , the recombination caused a loss or a rearrangement of the genomic sequence ( Figure S1 ) ., This type of retrotransposition mediated deletion has been previously described for L1 7–9 and Alu 14 , 34 ., For one particular example , clone 57 , the immediate 3′ and 5′ genomic sequences flanking the insert are 99 kb apart in the reference genome assembly ., PCR analysis of the transfected and untransfected HeLa DNA confirmed that this genomic rearrangement was not pre-existing in the HeLa cell line , but instead is likely associated with the Alu insertion ( Figure S2 ) ., Interestingly , clone 57 is the only insert in our data set with no identifiable A-tail ., An additional two inserts of the fully characterized Alus ( 0 . 8% ) also lacked the canonical endonuclease cleavage sites and direct repeats of TPRT insertion ( clones 108 and 203 ) , suggesting an endonuclease independent mechanism of insertion 11 , 35 ., These clones were also associated with potential genomic rearrangements ( details in Text S1 ) ., We used the 5′ flanking genomic sequence from the 226 rescued inserts to determine their genomic location ., Alu insertions were recovered from all chromosomes ( Figure 2B ) ., The distribution of Alu inserts across chromosomes appears largely random ( Figure 2C ) , in agreement with previous reports of L1 insertions from tissue culture 9 ., A previous study showed an enrichment of L1 inserts associated with the c-myc gene on chromosome 8 36 ., However , we did not observe Alu insertions associated with the c-myc gene ., We analyzed the G+C and repeat element sequence content of the pre-insertion loci in 20 kb intervals of all 226 Alu inserts ( Table 2 ) ., Relative to the genomic average and modified HeLa karyotype , we find that the overall pattern of Alu pre-insertion sites is consistent with a previous analysis of de novo tagged L1 inserts 36 ., Pre-insertion sites were Alu rich and L1 poor , although the tagged L1s inserted into comparatively more L1 poor regions ( 13 . 3% for L1 inserts from Gasior et al . 2006 compared to 15 . 5% for Alu inserts in the present study ) ., Alu insertions that were driven by ORF2 alone landed in genomic regions that were more L1 poor than insertions driven by full-length L1 ( 13 . 9% compared to 17 . 0% ) ., However , this observed difference is not statistically significant ( two sample , two-tailed t-test , p\u200a=\u200a0 . 172 ) ., We next assessed the distribution of recovered inserts relative to annotated genes in the human reference genome ., We find that 57 . 7% of all combined inserts landed in genic sequence compared to 42 . 3% that were intergenic ( Table 3 ) ., As previously indicated , there is no significant difference between the genic/intergenic distribution of L1 and ORF2 driven Alu inserts ( Pearson X2\u200a=\u200a3 . 41; p\u200a=\u200a0 . 065 ) ., Six of the Alu inserts landed within exons , but only two caused disruption of coding sequences , while the other four landed in the 5′ or 3′ untranslated regions ( UTRs ) of coding exons ( Table 3 ) ., Just over a third ( 36 . 2% ) of genic de novo Alus inserted in the sense strand , compared to ( 63 . 8% ) on the opposite strand ., This observation is slightly more skewed than the 55% antisense to 45% sense strand intronic distribution of the sequenced human genome 37 , 38 ., Overall , these data are consistent with an antisense bias ( binomial probability , p\u200a=\u200a0 . 002 ) ., To further analyze Alu insertion preferences , we assessed the de novo Alu inserts relative to features that were found to associate with the genome-wide distribution of Alus in a previous evolutionary analysis 39 ., In this approach , the 226 de novo Alu inserts observed here are localized within a system of 2765 non-overlapping human genome 1 Mb windows as employed in 39 and statistically evaluated for association with previously described genomic features ( details in Materials and Methods ) ., Nine genomic features were selected to evaluate genome landscape , recombination and natural selection ( details in Table 4 and Table S2 ) ., For each feature , we contrasted 203 insert-containing windows and the 2562 insert-free windows , using the Mann-Whitney-Wilcoxon test ( see Materials and Methods ) ., After Bonferroni correction for multiple testing , our results ( Table 4 ) indicate that the de novo Alus integrated in genomic regions that:, ( a ) are rich in genes and highly conserved elements ( suggesting function ) ,, ( b ) have high GC-content ,, ( c ) contain a 13-mer associated with recombination hotspots and genome instability ( Myers et al . 2008 ) and, ( d ) are enriched with SINEs , confirming that our observations of the 2-kb flanking regions ( Table 2 ) may extend up to 1 Mb ., We repeated the analysis using random subsets of the de novo inserts and the results remained consistent ( data not shown ) ., Some transcripts containing Alu sequences have been reported to be subjected to RNA editing 40–43 ., However , these examples refer to Alu sequences within RNA pol II generated transcripts ., Thus , we evaluated our data for evidence of editing of RNA pol III transcripts ., A total of 52 , 039 bp of de novo Alu inserts were analyzed , which excluded the middle A-rich region of the Alu sequence from the analysis ., We only found six point mutations ( ∼0 . 01% ) , three clustering within a 20 bp sequence of a single Alu insert ., None of the changes reflected the expected sequence changes due to RNA editing and may reflect errors introduced during reverse transcription by the L1 ORF2 or random mutations ., Our observations are consistent with previously published data showing no evidence of editing by three APOBECs ( 3A , 3B or 3G ) on the neomycin cassette sequence from inserts of a tagged Alu 44 , 45 ., An intriguing observation associated with SINE insertion events is the reported increase in A-tail length of new inserts relative to their source element 17–19 ., We used constructs with non-A disruptions within the A-tail to further investigate the underlying mechanisms of A-tail expansion in recovered de novo Alus ., We used two constructs containing different A-tails ( Figure 1B ) to determine if differences in A-tail disruptions or length might differentially affect new insert A-tail sequence ., The shorter A-tail construct ( A30D ) contains three polyA segments of 10 adenosines , separated by two different disruptions ( CT and TAC , Figure 1B ) ., The longer A-tail construct ( A70D ) is more than twice as long as the A30CT tail ( 82 bp compared to 35 bp ) and contains four polyA segments of 17 or 18 adenosines separated by three different disruptions ( CATTAC , G , and CACAC , Figure 1B ) ., We fully analyzed A-tail sequence data from 14 Alu inserts generated from the construct with the short A30D A-tail and 91 inserts from the longer A70D construct ( Figure 3 ) ., Overall , the de novo Alu inserts showed extensive A-tail expansion relative to the parental Alu ., A-tail and insert characteristics for the individual inserts are detailed in Table S1b ., Because the length of the A-tail at the 3′ end of the recovered inserts can vary depending on where priming occurs within the RNA molecule during TPRT ( see Figure 3A ) , we grouped inserts based on the priming location ., Internal priming has previously been observed for L1 inserts 2 ., Priming location was inferred by the absence/presence of the non-adenosine disruptions ., We define polyA segments of new inserts as “terminal” when the segment is used as the priming location for TPRT ., Note that the “terminal” polyA segment of a new insert can be any one of the polyA segments from the reference parental element ( shaded orange in Figure 3C ) and that internal priming events can generate inserts with shorter individual polyA segments as well as shorter A-tails in general ., Figure 3C shows examples of the four types of A-tails generated by construct A70D ., Although the A30D data set is much smaller , many of the observed characteristics were shared between both data sets ., Figure 3B shows the A-tail length results for the A30D data set ( data for the larger A70D set is provided in the Table S1b ) ., Surprisingly , when the construct with this shorter A-tail was used , all but two Alu inserts ( #123 and #125 ) primed at the most 3′ end polyA segment ( Figure 3B ) ., These two Alu30D inserts with A-tails lacking one or both of the non-adenosine nucleotide disruptions were likely the result of internal priming during TPRT ( as illustrated in Figure 3A ) ., In contrast , the majority of the priming occurred internally in the A70D dataset , but very few primed in the first or most “internal” polyA or segment #1 ( 8 out of 91 inserts , Figure 3D ) ., The A30D and A70D data sets are significantly different with respect to having “complete” A-tails ( all disruptions and polyA segments present ) ( Pearson X2 , p<0 . 001 ) ., It is possible that the added length of the A70D A-tail led to an increased frequency of internal priming by expanding the available area for priming to occur ., In both sets , priming seldom occurred at a distance of less than 20–25 bp downstream from where the polyA segment initiates ., The A70D data set also has significantly fewer than expected priming events within the most 3′ polyA segment , under the null hypothesis that priming locations are randomly distributed across the A-tail ( Chi-square goodness-of-fit , p<0 . 0001 ) ., We observed significant extension of the polyA segment length in both data sets ., Closer inspection of the individual segment sizes revealed that the terminal segments are considerably longer than internal segments ., The median terminal segment length ( 41 . 5 bp ) for the A30D set is about 4 times longer than the median for internal segments ( 11 bp ) ( Mann-Whitney U test , p<0 . 0001 ) ., Similar to the A30D data , the 91 A-tails from the A70D data set showed a bias to 3′ end elongation of the inserts when the length of the internal polyA segments is compared to terminal segments ( Figures 3D and E ) ., The histogram ( Figure 3E ) shows the overall size distribution of all four polyA segments , separated into internal ( white bars ) vs . terminal ( black bars ) segments ., Although both internal and terminal polyA segments increased in length , terminal segments are significantly longer ( medians of 23 . 0 and 42 . 0 , respectively; Mann-Whitney U test , p<0 . 0001 ) ., Almost all of the A70D terminal polyA segments ( 95 . 9% ) show expansion of four adenosines or more , while just over half of the internal segments exceed this level of expansion ( 55 . 6% ) ., Although there is a bias toward larger expansions occurring at terminal segments ( gray bars , Figure 3D , Table S1b ) , all of the internal polyA sections showed at least a minor increase in length relative to parental segments ( indicated by the dashed horizontal line , Figure 3D ) with medians of 22 or more adenosines per polyA segment ., In contrast , shortening only occurs in terminal segments , as we observed 17 inserts with shorter terminal polyA segments than the parental construct ( Figure 3E , black bars left of the vertical dashed line and Table S1b ) ., This suggests that the shorter terminal A-stretches may be a result of internal priming within the terminal polyA segments during the initial step of reverse transcription by ORF2p ( Figure 3A ) ., To determine if the observed A-tail expansion may have occurred at the RNA level , we generated cDNA clones using 3′ RACE ( RT-PCR ) from isolated RNA of transiently transfected cells using either construct specific primers or a generic anchored oligo-polydT primer ( details in Materials and Methods ) ., Sequence analysis of these clones clearly showed that the insert A-tail elongation could not be explained by RNA transcript variation ., We observed only slight transcript sequence differences of 1–3 adenosine losses or gains per polyA segment ( Figure S3 ) ., Moreover , we observed more than twice as many adenosine losses than gains in the cloned cDNA sequence derived from the transcripts , suggesting that the A-tail variation introduced by transcription or by our recovery and cloning methodology is more likely to lead to shorter A-tails ., Analysis of clones recovered from PCR amplification of a DNA template also revealed a similar change in adenosine numbers ( Figure S4 ) , possibly indicating that these sequence differences in the cDNA are introduced during the bacterial growth or amplification steps during the RT-PCR steps of the 3′ RACE and are not reflective of the actual RNA sequence ., It is noteworthy that we did not observe the large adenosine amplifications in our analysis of RNA transcripts , making it unlikely that changes in the Alu RNA template are a significant mechanism for the A-tail expansion observed in our recovered clones ., During the Alu rescue process , many of the loci containing the Alu inserts were independently recovered multiple times ., If expansion of polyA segments is an artifact of the cloning process , we would expect to see segment length variation between independently recovered samples ., Instead , we observed minimal sequence variation between the recovered samples derived from the same Alu insert ., In eight randomly chosen A-tail examples with a combined 2444 bp , only one sample with a single adenosine insertion was observed ( Figure S5 ) ., This observation is in stark contrast to the consistent and large A-tail length expansion of the Alu inserts ., Our data strongly indicate that the recovery assay is unlikely to contribute to the large A-tail expansions observed ., Our SINE recovery method provides a complementary approach for comprehensive analysis of the impact of Alu on the human genome that can give novel insights into the biological mechanisms governing SINE amplification ., In summary , the recovery of de novo tagged Alu inserts in HeLa cells revealed that ( 1 ) L1 and ORF2 driven Alu inserts are indistinguishable in human cells; ( 2 ) Alu insertion mediated deletions and rearrangements lack the hallmarks of retrotransposition , likely due to an alternate mechanism to resolve insertion intermediates; ( 3 ) inserts show an apparently random distribution across chromosomes , although a bias exists favoring insertion near other SINEs , highly conserved elements and genes; ( 4 ) de novo Alu inserts show no evidence of RNA editing; ( 5 ) TPRT priming was not observed within the first 20 bp ( most 5′ ) of the A-tail , suggesting the possible interference of bound proteins to the transcript or an unknown spacing requirement needed to engage the RT , RNA and priming sequence; ( 6 ) L1 ORF2 protein may show slippage during reverse transcription , leading to the expansion of de novo Alu element A-tails; and ( 7 ) expansion occurs across the entire length of the A-tail , but with a bias toward the 3′ end ., A major advantage of our approach is the ability to study inserts that have experienced little or no selection and the ability to compare de novo inserts to the known reference source element ., By comparing inserts from our tissue culture system to genomic Alu inserts , we can better understand how selection has shaped the current distribution of human Alu sequences ., Our results indicate that different genomic features might be important for initial Alu integration , as studied here , vs . long-term evolutionary survival of Alu insertions in the genome 39 ., In particular , here we show that Alus integrate in genomic regions rich in genes and in sequences categorized as “most conserved” 46 , suggesting an integration preference into or near functional elements ., The association of Alu integrations with gene-dense regions is intriguing and is consistent with the previously reported enrichment of Alus near housekeeping genes 47 , 48 ., Although speculative , this integration preference suggests Alu is a highly efficient mutagen of human genes ., In addition , targeting gene rich regions may provide fertile ground for added damage due to genomic rearrangements generated during insertion 49 ., Interestingly , among these significant features , only enrichment of the genome instability 13-mer 50 was also a significant positive predictor of the distribution of human-specific AluY elements , as identified in an evolutionary analysis 39 ., This suggests that , except for this one common predictive feature , there are substantial differences between Alu integration and fixation preferences; while the present analysis largely captures integration , the published Alu distribution properties 39 reflect both integration and fixation ., Increased Alu insertion near other SINEs provides a mechanism explaining the clustering of Alus in the human genome 51 and the common occurrence of tandem Alu inserts 52 ., Having a higher density of Alu elements may facilitate non-allelic homologous recombination ( NAHR ) , leading to the uneven genetic exchange between alleles that cause both deletions and duplications 49 ., The importance of the genome instability 13-mer motif correlating with both Alu distribution and integration is that it highlights a convergence of recombination hotspots and high Alu density regions potentially contributing to Alu-mediated NAHR 49 , 53–55 ., Consistent with the observations obtained from genomic data mining 56 , we have found that Alu retrotransposition is associated with genomic deletions and rearrangements ., However , the lack of the structural retrotransposition hallmarks suggests that alternate means of resolving retrotransposition intermediates , such as recombination 8 , 9 , 14 , 57 or non-homologous end joining 7 , 8 , 58 is likely contributing to the Alu-mediated genomic rearrangements/deletions ., Overall , our findings validate the tissue culture system as a robust method to study SINE biology ., An important feature of our Alu rescue system is that we were able to directly compare de novo Alu insert A-tails to the parental source A-tails with engineered disruptions ., This approach allowed us to determine that A-tail elongation occurs during reverse transcription by ORF2p , leading to expansion across the length of the A-tail , but with disproportionate expansion closer to the 3′ end ., The portion of the A-tail used for base pairing during TPRT priming was likewise not random , with the majority of priming locations at least 25 or more bases away from the 5′ end of the A-tail ., This priming location preference may reflect a physical constraint such as bound proteins that limit where annealing for reverse transcription can occur ., Although speculative , a potential protein candidate could be polyA binding protein ( PABP ) , which is known to associate with SINE RNPs 59 , 60 ., We present a model of slippage by ORF2p during TPRT ( Figure 4A ) favoring A-tail expansion ., We propose that the beginning of TPRT only provides a weak interaction between the Alu transcript and the cleaved DNA strand through limited hydrogen bonding between base pairs ., At this early stage , the complex may become dissociated , pausing reverse transcription until the interaction is re-established in a manner somewhat reminiscent of the reiterative synthesis of telomerase during reverse transcription ., This is similar to the model proposed for the I factor , a non-LTR element in Drosophila 61 ., In addition , telomerase slippage has been reported in Saccharomyces 62–64 , T . thermophila 65 and Candida albicans 66 ., Previous in vitro data also highlighted similarities between the L1 protein and telomerase by demonstrating that L1 ORF2 can initiate reverse transcription on oligonucleotide adapters simulating telomere ends 67 ., Our model depicts two non-mutually exclusive mechanisms by which slippage can occur: either ( 1 ) complete dissociation occurs followed by re-annealing , or ( 2 ) partial dissociation occurs , causing the cDNA strand to “loop out” before base pairing can once again secure the complex ., Interestingly , previous observations on the reverse transcription activity of the Bombyx mori R2 element demonstrate the incorporation of additional nucleotides that appear to arise from multiple rounds of the reverse transcriptase engaging the 3′ end of full length RNA templates 68 ., However , potentially untemplated residues can be incorporated depending on the length and composition of the extreme 3′ end of the RNA ., As cDNA length increases , the additional hydrogen bonding between the molecules stabilizes the process and reduces or eliminates slippage ., This increased stability with cDNA extension provides a simple explanation for our observation of preferential 3′ A-tail expansion , as the probability of dissociation and expansion diminishes as the nascent cDNA strand grows in length ., In order for our model to favor A-tail expansion over shortening , re-annealing and/or “looping out” must preferentially occur as depicted in Figure 4A to duplicate A-tail nucleotides rather than delete them ., Specifically , re-annealing of the cDNA strand must be further 3′ on the Alu RNA strand , with the cDNA strand looping out ., We propose that the presence of proteins bound to the Alu RNA could affect re-annealing dynamics ., For example , a potential candidate is poly ( A ) binding protein ( PABP , as shown in Figure 4A ) , which may play an important role in favoring A-tail sequence duplication over deletion , serving as a physical barrier that promotes 3′ re-annealing and/or prevents the Alu RNA from looping out ., Because our construct contains non-adenosine residues , sequence duplications can be easily identified ., We recovered five Alu inserts with duplicated non-A disruptions in the A-tail sequence ( Figure 4A ) ., Duplication of 3′ sequences was previously observed for a recovered L1 sequence 9 , indicating that this type of event also occurs during L1 insertion ., Several data support our proposed model ., First , no expansion of the polyA segments is observed at the RNA level ., Second , A-tail expansion occurs across all polyA segments ., These observations are not consistent with RNA polyadenylation or template switching , as these processes would lead exclusively to expansion of the terminal polyA segment ., Finally , duplications of the non-A disruptions are a strong indicator of slippage ., Although polyadenylation of Alu transcripts and template switching may occur , our data indicate that these types of events are not the main processes contributing to the A-tail expansion of de novo Alu inserts in this assay system ., In contrast to L1 , A-tail expansion of new Alu inserts has a significant biological impact on the perpetuation of active Alu elements in the human genome ( Figure 4B ) ., Although there are over one million Alu elements in the genome , the vast majority are inactive and unable to generate new copies ., Several factors , including intrinsic nucleotide composition and adjacent genomic sequences , determine Alu retrotransposition capability 24 , 69 ., One such requirement for efficient Alu retrotransposition is the presence of an A-tail 70 ., Because RNA polymerase III transcribed Alu RNA does not undergo enzymatic polyadenylation like mRNAs , Alu depends on the 3′ encoded polyA sequence to generate A-tail containing Alu transcripts ., Previous work has shown that A-tails of individual Alu elements mutate rapidly leading to smaller and more heterogeneous tails 71 , 72 and limiting retrotransposition capability 24 ., As time progresses , the A-tails of active Alu source elements shrink and degrade , decreasing their ability to support retrotransposition ., Therefore , without the reintroduction of new Alu copies with expanded A-tail sequence to counteract the rapid evolutionary loss of homogeneity and length , active Alu copies would be lost , leading to the eventual ext | Introduction, Results, Discussion, Materials and Methods | Alu elements are trans-mobilized by the autonomous non-LTR retroelement , LINE-1 ( L1 ) ., Alu-induced insertion mutagenesis contributes to about 0 . 1% human genetic disease and is responsible for the majority of the documented instances of human retroelement insertion-induced disease ., Here we introduce a SINE recovery method that provides a complementary approach for comprehensive analysis of the impact and biological mechanisms of Alu retrotransposition ., Using this approach , we recovered 226 de novo tagged Alu inserts in HeLa cells ., Our analysis reveals that in human cells marked Alu inserts driven by either exogenously supplied full length L1 or ORF2 protein are indistinguishable ., Four percent of de novo Alu inserts were associated with genomic deletions and rearrangements and lacked the hallmarks of retrotransposition ., In contrast to L1 inserts , 5′ truncations of Alu inserts are rare , as most of the recovered inserts ( 96 . 5% ) are full length ., De novo Alus show a random pattern of insertion across chromosomes , but further characterization revealed an Alu insertion bias exists favoring insertion near other SINEs , highly conserved elements , with almost 60% landing within genes ., De novo Alu inserts show no evidence of RNA editing ., Priming for reverse transcription rarely occurred within the first 20 bp ( most 5′ ) of the A-tail ., The A-tails of recovered inserts show significant expansion , with many at least doubling in length ., Sequence manipulation of the construct led to the demonstration that the A-tail expansion likely occurs during insertion due to slippage by the L1 ORF2 protein ., We postulate that the A-tail expansion directly impacts Alu evolution by reintroducing new active source elements to counteract the natural loss of active Alus and minimizing Alu extinction . | SINEs are mobile elements that are found ubiquitously throughout a large diversity of genomes from plants to mammals ., The human SINE , Alu , is among the most successful mobile elements , with more than one million copies in the genome ., Due to its high activity and ability to insert throughout the genome , Alu retrotransposition is responsible for the majority of diseases reported to be caused by mobile element activity ., To further evaluate the genomic impact of SINEs , we recovered and characterized over 200 de novo Alu inserts under controlled conditions ., Our data reinforce observations on the mutagenic potential of Alu , with newly retrotransposed Alu elements favoring insertion into genic and highly conserved elements ., Alu-mediated deletions and rearrangements are infrequent and lack the typical hallmarks of TPRT retrotransposition , suggesting the use of an alternate method for resolving retrotransposition intermediates or an atypical insertion mechanism ., Our data also provide novel insights into SINE retrotransposition biology ., We found that slippage of L1 ORF2 protein during reverse transcription expands the A-tails of de novo insertions ., We propose that the L1 ORF2 protein plays a major role in minimizing Alu extinction by reintroducing active Alu elements to counter the natural loss of Alu source elements . | transposons, retrotransposons, biology, molecular cell biology, molecular biology | null |
journal.pcbi.1002886 | 2,013 | Interpretation of Genomic Variants Using a Unified Biological Network Approach | Advances in next-generation sequencing technologies have considerably reduced the cost of genome sequencing ., As a result , there has been an avalanche of personal genomic data with numerous individual genomes sequenced in the last few years 1–4 ., Variants in protein-coding genes are of special interest due to their stronger likelihood of functional effects ., A comprehensive understanding of the functional impact of variants in coding genes requires their integration with various levels of annotations , such as primary sequence of the gene , three-dimensional structures of its protein products and biological networks where genes interact with each other ., Functional annotation of single nucleotide variants ( SNVs ) at genomic sequence level results in their classification as nonsynonymous ( which includes missense and nonsense ) , splice site disrupting or synonymous ., Similarly , small insertions and deletions ( indels ) in coding genes can be classified as frame-shift or in-frame ., Nonsense and splice site disrupting SNVs as well as frame-shift indels are mostly assumed to lead to loss-of-function ( LoF ) of genes 5 ., On the other hand , missense SNVs and in-frame indels may or may not be damaging 6 ., It is well understood that genes and their protein products rarely act in isolation but rather work closely with other genes and/or their products to form various networks and pathways which accomplish specific goals , for example , signal transduction , metabolism etc ., Thus , a comprehensive understanding of the functional impact of variants necessitates the inclusion of these interactions between genes ., Network-based approaches are thus often used to study human disease 7 ., One feature that has emerged from past studies of disease genes and networks is that protein products of genes associated with similar disorders have a higher likelihood of physical interaction with each other 8 ., It has also been noted in many studies that functionally essential genes are more likely to encode for hub ( i . e . highly connected ) proteins in the physical protein-protein interaction ( PPI ) network in both yeast 9 and humans 8 ., Moreover , hub proteins are likely to be under stronger negative selection constraints in humans and positive selection tends to occur on network periphery 10 ., Similar studies on signaling pathways have revealed that as one goes from extracellular space to the nucleus in the cell , negative selection constraints on genes encoding corresponding proteins tend to increase 11 ., Selection studies have also been performed on metabolic pathways where enzyme connectivity signifies the number of other metabolic enzymes that produce the enzymes reactants or consume its products ., For example , in a yeast network of 584 metabolites and comprising about 16% of all yeast genes , Vitkup et al found that highly connected enzymes evolve slower than less connected enzymes 12 ., Montanucci et al also reported that genes encoding highly connected enzymes in N-glycosylation metabolic pathway exhibit stronger purifying selection constraints and tend to evolve slowly in primates 13 ., In order to obtain a higher resolution understanding of the relationship between selection constraints and networks , some studies have also integrated three-dimensional protein structures with PPI network to obtain structural interaction network ( SIN ) ., Kim et al showed that in yeast , hubs in PPI with more than two interaction interfaces are more likely to be essential than those with two or less interfaces 14 ., Using structurally resolved human PPI network , Wang et al showed that disease-causing missense SNVs and in-frame insertions and deletions tend to be enriched at the interaction interfaces of proteins associated with corresponding disorders 15 ., They also showed that the disease specificity of different mutations of the same gene can be explained by their location on the interaction interfaces ., Another important feature that has emerged from studies of genomic variants on protein structure ( without consideration of network interactions ) is that benign missense polymorphisms tend to occur at solvent exposed sites on protein structure , while disease-causing missense SNVs tend to be more buried 16 ., Previous studies examining the relationship of functional significance and selection properties of genes with network topology have mostly focused on networks with a singular mode of interactions between genes or their protein products , for example physical protein-protein interactions ., However , a gene and its protein products can be involved in various biological networks and its role and consequently its centrality can vary across these networks ., For example , SIX5 is a transcription factor gene that targets 360 genes in the human regulatory network but interacts with only one protein in the physical PPI network 17 , 18 ., This gene is of high functional significance since its disruption causes Branchio-oto-renal syndrome , a developmental disorder characterized by the association of branchial arch defects , hearing loss and renal anomalies 19 ., In this study we examine the relationship of functional essentiality and selection with various biological networks – protein-protein interaction ( PPI ) , phosphorylation , signaling , metabolic , genetic and regulatory ., This enables us to understand the functional importance and selection constraints on genes in a global systemic approach ., Moreover , although it has been shown that low evolutionary conservation of LoF- tolerant genes and their large distance from recessive disease genes in PPI network can be used to predict disease causation of variants 5 , their unique properties in diverse biological networks have not been exploited before ., Here , we use the distinguishing network and evolutionary properties of functionally essential and LoF-tolerant genes to build a predictive model for global damage caused by novel variants ., Using this model , we are able to compute functional indispensability scores for all protein-coding genes ., The biological networks studied in this work include – PPI , phosphorylation , metabolic , signaling , genetic and regulatory ( Materials and Methods ) ., Some of these networks represent direct physical interactions between proteins , for example , PPI ., On the other hand , genetic and regulatory networks contain indirect interactions between gene pairs ., Additionally , some networks such as phosphorylation , metabolic , signaling and regulatory are directional with an upstream and downstream gene , whereas PPI and genetic interactions are undirected ., While a gene can have a vital role in one pathway or network , it might not be as crucial in another network ., Therefore , we pool together data from all the above-mentioned biological networks to construct a unified global network , which we term Multinet ( Materials and Methods ) ., The Multinet enables the analyses of the genes via their roles in the individual networks and the combined network ., We note that some interactions between two different networks can be shared ., For example , an interaction in which gene A phosphorylates gene B can occur in both phosphorylation and PPI networks ., However , we find that out of ∼110 , 000 interactions in our data set , only 881 interactions occur in more than one network ., Thus the vast majority of interactions in our data are unique or non-redundant ., This observation reiterates the fact that interactions of genes vary across different networks and it is crucial to include all the networks while analyzing the relationship between functional importance and selection constraints with global network centrality ., The distribution of 881 interactions which occur in more than one network is shown in Supporting Figure S1 ., The numbers of genes and unique interactions in each network are shown in Supporting Table S1 ., In this section we investigate the relationship between functional significance of genes and their properties in various biological networks ., All human protein-coding genes are divided into four categories based on their known disease susceptibilities and functional impact ., A ‘gene significance score’ ranging from 3 to 0 is assigned to each gene: 3 for essential genes , 2 for all genes with disease-causing mutations in HGMD , 0 for LoF-tolerant genes and 1 for all the remaining genes that do not fit into any of the above categories ( Materials and Methods ) ., We then correlate these significance scores with the degree centralities of the genes in all networks ., Degree centrality of a gene in any network is defined as the number of its interacting partners in that network ., In order to estimate the total number of interacting partners of a gene , we use its connectivity ( number of interactions ) in the Multinet ( Materials and Methods ) ., We find that gene significance scores show positive correlation with degree centralities in most networks , though it is statistically significant only in PPI and signaling network and Multinet ( Figures 1 and 2A; Supporting Table S2 ) ., Thus , in general , essential genes tend to be more connected in biological systems consistent with previous findings 8 ., Surprisingly , we find a small but significant negative correlation between gene significance score and metabolic degree ( Spearman correlation coefficient or SCC\u200a=\u200a−0 . 07 , pvalue\u200a=\u200a0 . 028 ) ., We also find that , unlike most other degree centralities , the metabolic degree centrality of genes shows a significant positive correlation with the number of paralogs ( duplicated copies ) ( SCC\u200a=\u200a0 . 15; pvalue\u200a=\u200a8 . 26e-07 ) ( Supporting Table S3; Materials and Methods ) ., Thus , it is possible that in case of a LoF mutation in a participating enzyme , the metabolic pathway can be re-routed to an alternate path , possibly involving a duplicated gene of the disabled enzyme ., Our observation in the human metabolic network is in agreement with a previous study by Vitkup et al , in which they found that highly connected enzymes are no more likely to be essential than less connected enzymes in yeast metabolic network 12 ., In this study we find that not only are essential genes unlikely to be highly connected in human metabolic network , LoF-tolerant genes ( whenever present in metabolic network ) are indeed more connected than essential genes ( Supporting Table S7 ) ., This result demonstrates a major contrast between the structure of the metabolic network and other networks ., In most biological networks , highly connected genes tend to have fewer duplicated copies; hence LoF mutations in them can have serious phenotypic consequences ., Since this distinct trend of high degeneracy at hub proteins is observed only in the metabolic network , we further posit that this might be an evolutionary mechanism to increase tolerance towards damaging mutations ., The uniqueness of such a ‘protective’ effect somewhat suggests an implicit level of greater functional importance of metabolic pathways as compared to other networks of gene interactions ., Interestingly , we find that gene significance scores are positively correlated with the number of networks the gene is involved in ( Figures 1 and 2B ) ., This indicates that genes involved in many networks can act as information bottlenecks between different systems and thus tend to be more essential ., We next examine the relationship between selection constraints on genes and their network properties ., We estimate evolutionary constraints over long time-scale by dN/dS ( ratio of missense to synonymous substitution rates ) computed from human-chimp ortholog alignments ( Materials and Methods ) ., dN/dS<1 indicates purifying selection while values close to 1 indicate neutral selection and dN/dS>1 indicates positive selection ., We find that dN/dS values of genes are negatively correlated with their degree centralities in all networks , though they reach significance in PPI , phosphorylation , regulatory and Multinet networks ( Supporting Table S4 ) ., This shows that highly connected genes tend to be under stronger purifying selection constraints over long evolutionary time-scale , in agreement with previous studies 10 ., Furthermore , we analyze patterns of genetic variation in modern-day humans in relation to biological networks ., We compute average heterozygosity of each gene to estimate its genetic variability using missense SNPs ( single nucleotide polymorphisms ) and their corresponding allele frequencies in three sets of populations from 1000 Genomes Pilot data ( Materials and Methods ) 4 ., We find that there is a significant negative correlation between Multinet degree and heterozygosity of missense SNPs for all three populations , indicating more genetic variation at the periphery of networks ( the correlation is also significant for some populations in PPI , phosphorylation and regulatory networks ) ( Supporting Table S5 ) ., Interestingly , we do not find a significant correlation of heterozygosity of synonymous SNPs with Multinet degree ( Supporting Table S6; Materials and Methods ) ., Putting together , these results suggest that reduced genetic variability of highly connected genes with respect to missense SNPs is indeed due to selection constraints ., When network edges between two genes correspond to physical interactions between their protein products , molecular level details of the interaction can be obtained by integrating three-dimensional protein structures with the underlying network data ., Therefore , in order to understand the reasons for selection constraints in PPI network at higher resolution , we integrated three-dimensional protein structures with network interaction data to create structural interaction network ( SIN ) ( Figure 3A; Materials and Methods ) 14 , 15 , 20 ., SIN is a subset of the larger PPI network and consists of 2 , 102 genes and 11 , 433 interactions ., SIN construction allows us to estimate the number of interfaces used by a protein to interact with other proteins ( Figure 3A; Materials and Methods ) ., We find that there is a significant positive correlation between gene significance scores and the number of interfaces used by their protein products in SIN ( Figure 1 ) ., Thus , protein products of essential genes tend to use more interaction interfaces than those of LoF-tolerant genes ., We also find that the number of interfaces used by the protein to interact with other proteins in SIN is positively correlated with their degree centrality in PPI network ( SCC\u200a=\u200a0 . 18 , pvalue\u200a=\u200a1 . 06e-09 ) ., This shows that hub proteins tend to have more interaction interfaces ., Thus , it is likely that higher number of interfaces possessed by protein products of essential genes could partly be a result of their higher degree centrality in PPI network ., We next examine the impact of missense SNVs on protein structure in relation to SIN ., We find that , in general , residues with disease-causing missense SNVs tend to be more buried inside protein structure than polymorphic residues ( Figure 3B ) ., Our observation is consistent with previous findings which have reported that missense mutations buried inside protein structure tend to be more deleterious than those on surface 16 ., However , these previous studies treated all proteins equally and did not differentiate between hub and non-hub proteins in PPI network ., When we treat hub ( degree centrality>\u200a=\u200a50 ) and non-hub proteins separately , we find that accessible surface area for residues with missense disease mutations is higher for hub proteins ( Wilcoxon rank sum pvalue\u200a=\u200a0 . 014; Supporting Figure S2 ) ., We also observe a significant positive correlation between the degree centrality of protein and the accessible surface area of their residues undergoing disease mutations ( SCC\u200a=\u200a0 . 028 , pvalue\u200a=\u200a3 . 12e-03 ) ., These results show that hub proteins tend to have a higher fraction of missense disease mutations on their exposed surface ., This result is very reasonable in light of our observation that hub proteins tend to have more interaction interfaces ( see preceding paragraph ) , thereby having a higher fraction of their exposed surface under selection constraints ., In order to further examine the close correlation of network and evolutionary properties with gene essentiality we use a logistic regression model to differentiate essential genes from LoF-tolerant genes ( Materials and Methods ) ., Network features used to train the logistic regression model include degree centralities in Multinet and all networks separately ( PPI , phosphorylation , signaling , metabolic , genetic and regulatory ) , number of networks the gene is involved in and number of interfaces used in SIN ., Selection properties used in the model include human-chimp dN/dS ratios and average heterozygosities of both synonymous and missense SNPs in modern human populations ., The average values of these features for LoF-tolerant and essential genes along with corresponding Wilcoxon rank sum pvalues are provided in Supporting Table S7 ( see also Figure 1 ) ., Using these features we obtain an excellent classification accuracy for 140 LoF-tolerant and 115 essential genes with AUC\u200a=\u200a0 . 914 ( Figure 4A; Materials and Methods ) ., Network properties that contribute significantly to the model include degree centralities in regulatory , genetic and metabolic networks as well as number of networks the gene is involved in ( Materials and Methods ) ., On further examination of network participation of LoF-tolerant and essential genes , we find that most LoF-tolerant genes are not involved in any network and some of them are involved in a very small number of networks ( Figure 4B ) ., On the other hand , most essential genes are involved in many networks ( Figure 4C ) ., Genes involved in a variety of networks serve as information bottlenecks between different systems and hence are more likely to be essential ., We note that absence in some networks could partially be due to missing network data in our study and/or a bias in existing databases ., Essential genes are more likely to have been the focus of previous research studies , for example PPI studies , and hence more likely to be present in our PPI network ., They also tend to have more regulatory interactions and thus are more likely to be present in our regulatory network ( which consists of 118 transcription factors and their target genes: the most comprehensive human regulatory network available to our knowledge ) 17 ., However , the strength of our model lies in its use of many different network properties to minimize the biases resulting from the use of a single network property or data resource ., Furthermore , to test the robustness of our model , we computed the AUC for separation of LoF-tolerant and essential genes multiple times – each time randomly removing 10% of the edges from a network and rebuilding the Multinet ., After repeating this for all the networks , we find minimal change in the AUC ( ranging from 0 . 914 to 0 . 912 ) , which shows that our model is quite robust to changing some edges in individual networks ., We next perform an independent validation of our model by applying it on all genes that are neither LoF-tolerant nor essential ., Interestingly , we find that predicted functional indispensability scores are in the following order: genes with known disease-causing mutations have significantly higher scores than genes identified in genome-wide association ( GWA ) studies ( Wilcoxon rank sum pvalue\u200a=\u200a7 . 62e-05 ) , which are in turn significantly higher than all the remaining neutral genes ( Wilcoxon rank sum pvalue<2 . 2e-16 ) ( Figure 4D ) ., Genes identified in GWA studies are associated with phenotypic consequences , while they are not necessarily the causal genes ., Hence it is reassuring that genes with known disease-causing mutations in HGMD receive significantly higher scores than those identified in GWA studies ., This validation exercise demonstrates that our model can help researchers pick candidate disease genes in clinical sequencing studies ., We have provided the predicted scores for all the genes in Supporting Table S8 ., We note that the predicted functional indispensability scores are continuous scores unlike the discrete gene significance scores used to compute correlations in an earlier section ., Genes and their protein products work in collaboration with other genes to form biological systems that perform specific tasks ., For a systemic understanding of the role a gene plays , there is a need to integrate different modes of gene interactions ., In this work we pool together interaction data from various biological systems ( PPI , phosphorylation , signaling , metabolic , genetic and regulatory ) to create a unified Multinet , enabling the computation of degree centrality of the genes in their individual networks and in the context of the entire Multinet ( Supporting Table S8 ) ., Subsequent analysis of functional significance and evolutionary properties of genes allows us to relate genomic sequence variants in individual genes to their functional effects in individual and global networks ., We find that highly connected genes in the Multinet and genes that participate in many biological systems tend to be more functionally significant , have fewer paralogs and resist mutations in healthy humans ., While we also observe similar trends in most of the constituent networks of the Multinet , the metabolic network seems to be an exception ., Highly connected genes in the metabolic network tend to have more paralogs and are more tolerant to LoF mutations ., Next , we combine three-dimensional protein structural information with PPI network to create structural interaction network ( SIN ) and understand selection on protein structure at molecular level detail ., We find that functionally essential genes ( which are more likely to encode for hub proteins ) tend to use more interfaces to interact with other proteins ., We also observe that hub proteins in PPI network contain a higher fraction of disease-causing mutations on their solvent exposed surface , as compared to non-hub proteins ., Thus , although generally missense SNVs on exposed protein surface are more likely to be benign , our results show that those on the surface of hub proteins are more likely to be deleterious 21 ., Finally , we integrate network and selection properties of genes to build a logistic regression model which can separate LoF-tolerant and essential genes with high accuracy ( AUC\u200a=\u200a0 . 91 ) ., Application of the model on all genes shows that it predicts higher functional indispensability scores for genes with known disease-causing mutations than genes identified in GWA studies , which themselves have higher scores than remaining neutral genes ., The predicted functional indispensability scores for all genes are made publicly available and can be used to predict candidate disease genes in future clinical studies ., These scores are indicators of global damage caused by deleterious mutations in coding genes – including nonsense and missense SNVs and in-frame and frame-shift indels ., As mentioned above , nonsense SNVs and frame-shift indels are mostly assumed to disable gene function ., However , missense SNVs and in-frame indels are more complex since they may or may not have a deleterious impact ., Various methods exist to predict the functional effects of missense SNVs , for example , SIFT and PolyPhen 21 , 22 ., While these methods examine the tolerance of individual sites in genes to missense mutations , they do not take into account the functional significance of the entire gene ., For example , a moderately deleterious missense SNV in a highly significant gene can be equally or more damaging than a strongly deleterious missense SNV in a less significant gene ., Our method to compute functional indispensability scores for entire genes can be combined with scores predicted by SIFT and PolyPhen to obtain a more comprehensive view of the functional effects of genomic variation ., We note that even though our model is very robust to the removal of some edges in individual networks , the incomplete and biased nature of existing biological networks data may constitute a caveat in our study ., However , to our knowledge , this is the first comprehensive genome-wide study linking genetic variants at population scale as well as disease variants with a vast body of available network resources ., Models developed and applied in this study can be further expanded as more interaction data is obtained and further population genetics projects are undertaken , particularly with the future phases of the 1000 Genomes project ., Human protein-protein interaction and genetic interaction networks were extracted from BIOGRID ( release 3 . 1 . 83 ) 18 containing 43 , 722 and 263 interactions , respectively ., Regulatory network ( relationship between transcription factors and target genes ) was from ENCODE data 17 ., Metabolic enzyme network contained directed linkages from upstream enzymes to downstream enzymes , based on compound reactions in KEGG 23 ., Phosphorylation network in human contains 28 , 637 directed kinase-substrate interactions between 2 , 392 genes 24 ., The signaling network in this study is constructed based on 1 , 011 interactions and 527 proteins ( downloaded July 2011 ) from human signaling pathways obtained from the SignaLink database ( http://www . signalink . org/ ) 25 ., SignaLink offers an easily-downloadable and well-curated set of interactions from eight major signaling pathways found in humans that are not tissue-specific , namely EGF/MAPK , Ins/IGF , TGF-β , Wnt , Hedgehog , JAK/STAT , Notch and NHR ( Nuclear Hormone Receptors ) ., Manual data curation was performed in SignaLink by extensive literature survey of primary experimental evidence of these interactions , resulting in expansion of verified interaction data for the corresponding signaling pathways in protein interaction databases such as the KEGG 26 , Reactome 27 and NetPath 28 , while maintaining substantial overlaps with these databases ., A detailed description of the curation process and comparisons between these databases and SignaLink can be found in 25 ., Throughout the article , connectivity of the gene in PPI , phosphorylation , signaling and metabolic networks refers to connectivity of the protein product of the gene ., Interactions from all the above networks were combined to create Multinet ., If a gene pair interacts in multiple networks or shows both upstream and downstream connection in a directional network , the interaction is counted once in Multinet ., The list of 140 LoF-tolerant genes was obtained from MacArthur et al 5 ., This list contains genes that show homozygous LoF mutations in at least one individual in 1000 Genomes pilot data 4 ., The list of 115 essential genes was obtained from Liao et al 29 ., These genes exhibit clinical features of death before puberty or infertility when LoF mutations occur ., The list of 2 , 451 disease genes was obtained from HGMD ( Human Gene Mutation Database ) 30 ., All the genes with any disease-causing mutation ( DM tag in HGMD ) were used ., If any gene occurred in more than one category , its category was decided in a hierarchical fashion as follows: essential , followed by disease followed by LoF-tolerant ., The remaining 19 , 267 genes were assigned the category of neutral ., The list of genes identified in GWA studies was obtained from the NHGRI GWAS catalogue ( https://www . genome . gov/26525384#download ) ., Number of paralogs for each gene and dN/dS values for human-chimp orthologs were obtained from Ensembl using BioMart 31 ., SNPs in modern-day humans and their allele frequencies were obtained from the low-coverage pilot phase of the 1000 Genomes Project 4 ., This phase consisted of 60 individuals of CEU ( Utah residents with Northern and Western European Ancestry ) , 59 individuals of YRI ( Yoruba in Ibadan , Nigeria ) and 60 individuals of CHB+JPT ( Han Chinese in Beijing , China and Japanese in Tokyo , Japan ) populations ., Heterozygosity value is calculated as 2pq , where p and q correspond to the frequencies of the two alleles ., Average heterozygosity for a gene is defined as the average heterozygosity of the SNPs in that gene , where heterozygosities of missense and synonymous SNPs are computed separately . | Introduction, Results, Discussion, Materials and Methods | The decreasing cost of sequencing is leading to a growing repertoire of personal genomes ., However , we are lagging behind in understanding the functional consequences of the millions of variants obtained from sequencing ., Global system-wide effects of variants in coding genes are particularly poorly understood ., It is known that while variants in some genes can lead to diseases , complete disruption of other genes , called ‘loss-of-function tolerant’ , is possible with no obvious effect ., Here , we build a systems-based classifier to quantitatively estimate the global perturbation caused by deleterious mutations in each gene ., We first survey the degree to which gene centrality in various individual networks and a unified ‘Multinet’ correlates with the tolerance to loss-of-function mutations and evolutionary conservation ., We find that functionally significant and highly conserved genes tend to be more central in physical protein-protein and regulatory networks ., However , this is not the case for metabolic pathways , where the highly central genes have more duplicated copies and are more tolerant to loss-of-function mutations ., Integration of three-dimensional protein structures reveals that the correlation with centrality in the protein-protein interaction network is also seen in terms of the number of interaction interfaces used ., Finally , combining all the network and evolutionary properties allows us to build a classifier distinguishing functionally essential and loss-of-function tolerant genes with higher accuracy ( AUC\u200a=\u200a0 . 91 ) than any individual property ., Application of the classifier to the whole genome shows its strong potential for interpretation of variants involved in Mendelian diseases and in complex disorders probed by genome-wide association studies . | The number of personal genomes sequenced has grown rapidly over the last few years and is likely to grow further ., In order to use the DNA sequence variants amongst individuals for personalized medicine , we need to understand the functional impact of these variants ., Deleterious variants in genes can have a wide spectrum of global effects , ranging from fatal for essential genes to no obvious damaging effect for loss-of-function tolerant genes ., The global effect of a gene mutation is largely governed by the diverse biological networks in which the gene participates ., Since genes participate in many networks , no singular network captures the global picture of gene interactions ., Here we integrate the diverse modes of gene interactions ( regulatory , genetic , phosphorylation , signaling , metabolic and physical protein-protein interactions ) to create a unified biological network ., We then exploit the unique properties of loss-of-function tolerant and essential genes in this unified network to build a computational model that can predict global perturbation caused by deleterious mutations in all genes ., Our model can distinguish between these two gene sets with high accuracy and we further show that it can be used for interpretation of variants involved in Mendelian diseases and in complex disorders probed by genome-wide association studies . | systems biology, genomics, genetics, biology, computational biology, genetics and genomics | null |
journal.pbio.1001422 | 2,012 | Recombination Modulates How Selection Affects Linked Sites in Drosophila | Homologous meiotic recombination has an important role in molecular evolution ., Sufficient recombination uncouples the evolution of different sites on the same chromosome allowing positive or negative selection at one site to act independently from selection at another site ., If there is less than effectively free recombination between two selected sites , then linkage results in selection at one site interfering with selection at another site ., This has been termed “Hill–Roberson interference” 1–6 ., Hill–Robertson interference increases the probability of fixation of deleterious mutations , decreases the probability of fixation of advantageous mutations , and reduces overall DNA sequence diversity ., Thus , the breakdown of linkage disequilibrium between loci experiencing Hill–Robertson interference allows selection to act more efficiently , purging deleterious mutations and accelerating adaptation 1–6 ., Such indirect effects of recombination on the genome 7 result in a positive association between the rate of recombination and adaptive evolution 8–10 ., For example , recombination rate is positively associated with codon usage bias , whereby those codons coded by the most abundant tRNAs are “preferred” and used more often 11 , 12 ., Recombination has direct effects on a genome sequence as well , because recombination influences base composition through biased gene conversion and the distribution of repetitive elements , hotspot sequences , and indels 7 , 13–17 ., Understanding the magnitude of indirect effects in light of these direct effects has proved challenging 7 ., One striking association is a positive relationship of local recombination rate and nucleotide diversity 13 , 18 , 19 ., Originally described in Drosophila melanogaster 13 , the positive relationship between recombination rate and nucleotide diversity has been demonstrated in a wide range of taxa , including humans , mice , yeast , maize , and tomatoes ( reviewed in 20 ) ., It is not fully understood how much of this relationship results from recombinations indirect versus direct effects on the genome ., For instance , mutations created during crossing over or double-strand break repair may generate new polymorphisms and hence increase diversity 21–27 ., Alternatively , recombination may indirectly influence genetic diversity by mitigating the genomic footprint of selective sweeps and background selection 28–30 ., One way to distinguish between these general explanations is to evaluate the relationship of between-species nucleotide divergence at neutral sites and local recombination rate , because truly neutral mutations are substituted at the same average rate between species as they appear between generations , even if linked to sites under selection 31 , 32 ., This allows us to predict that both within-species nucleotide diversity and between-species nucleotide divergence would have a positive relationship with local recombination rate 13 , if the recombination–diversity association was purely caused by mutation ., In contrast , selective sweeps and background selection will cause an association between recombination and within-species nucleotide diversity , but not a relationship between recombination and between-species nucleotide divergence 30 , 32 ., The absence of an association of between-species nucleotide divergence and local recombination rate suggests that variation in recombination rate translates to variation in the efficiency of selection 13 ., Past work relating nucleotide divergence to recombination rate found no relationship between these two variables in several species of Drosophila , mouse , beet , yeast , and other species 13 , 20 , 33–37 ., Furthermore , in several species , evidence indicates that segregating ancestral polymorphisms may be responsible for correlations between divergence and recombination rate ( 38–40 , also suggested by 25 , 41 ) ., The test above , however , implicitly assumes that local recombination rates are conserved between the two species used to generate the nucleotide divergence measure ., If recombination rate has diverged between the two species , no relationship between local recombination rate and nucleotide divergence may be detected even when recombination is mutagenic ( see Figure S1 ) ., Recombination rates , especially at fine scales , are often not conserved among closely related species , as is the case between humans and chimpanzees 42–44; thus , the assumption of conservation of recombination rates may be violated in previous studies , and a more definitive understanding of the diversity–recombination association awaits estimates that are free from this assumption ., Though there are theoretical expectations concerning how recombination rate should affect selection efficiency 45 , 46 , it is unclear empirically whether variation in local recombination rates translates into significant variation in the efficiency of selection 7 ., Several empirical studies have tackled this problem 12 , 38 , 47–52 , and many findings suggest that recombination rate influences the efficiency of positive or negative selection in regions of moderate or high recombination ., Still , various confounding factors ( e . g . , biased gene conversion , gene density ) may produce spurious correlations between both recombination and substitution rate , and some authors suggest that there is no strong empirical evidence for recombination affecting the efficiency of selection ( apart from reduced selection in regions with essentially no recombination 7 ) ., The Drosophila pseudoobscura system is ideal for pursuing questions about recombination rate variation and its molecular evolutionary consequences ., The average crossover rate of D . pseudoobscura ( about 7 cM/Mb in females ) is over twice that of D . melanogaster 53 ., There is also considerable fine-scale ( <200 kb windows ) variation in the local recombination rate within the genome of D . pseudoobscura and within the genome of its sister species , D . persimilis 25 , 33 , 54 ., While some recombination data are available for D . pseudoobscura and D . persimilis , these sister taxa interbreed in the wild 55–57 and are , therefore , not ideal for examining the divergence–recombination association ., For example , shared polymorphism due to hybridization and recent speciation may be responsible for the positive divergence–recombination association found in a previous study 25 ( see also 38 , 39 ) ., Fortunately , a third species exists ( D . miranda ) that is phylogenetically close to D . pseudoobscura but does not interbreed with D . pseudoobscura ., Since there is still some residual shared ancestral polymorphism 58 , we also obtained the genome sequence for a slightly more distantly related outgroup species , D . lowei ( Figure S2 ) ., Sequence from D . lowei is useful for generating a proxy for neutral mutation rate across the genome ., In this work , we generate and compare two fine-scale recombination maps for D . pseudoobscura , which each cover approximately 43% of the D . pseudoobscura physical genome and one fine-scale recombination map that covers approximately 31% of the D . miranda physical genome ., In order to circumvent the assumption of classic studies , we analyze the relationship of local recombination rate to nucleotide diversity and divergence in regions with very similar recombination rates between the two species ., By employing a linear model framework to account for multiple covariates , we conclude that the contribution of recombination to diversity is significant and positive , but recombination contributes little to divergence ., This indicates that recombination is likely to modulate the footprint of selection in the genome ., Next , we tested the impact of recombination rate on the efficiency of selection ., We examined whether recombination rate ( 1 ) affects the distribution of nonsynonymous substitutions across the genome and ( 2 ) affects the pattern of diversity around nonsynonymous and synonymous substitutions ., In particular , we use a generalized linear model to test how recombination modulates the magnitude and physical extent of the loss of diversity surrounding substitutions ., Our analysis of these putative selective sweeps should be less sensitive to common confounding factors such as gene expression and GC content than previous measures ., In total , this work allowed us to determine that recombination rate has an important impact on how selection shapes diversity across the genome of Drosophila pseudoobscura and its close relatives ., We generated linkage maps for chromosome 2 and parts of the X chromosome for D . pseudoobscura and D . miranda ., Using a backcross design and inbred lines , we developed two replicate recombination maps ( referred to here as “Flagstaff” and “Pikes Peak” ) for D . pseudoobscura and one recombination map for D . miranda using the Illumina BeadArray platform to distinguish heterozygotes from homozygotes of the inbred lines used in the backcross design ., These maps ( Table S1 ) measure recombination rate across <200 kb windows , and we refer to these as “fine-scale” maps ., Recombination was surveyed across approximately 43% of the D . pseudoobscura physical genome and about 31% of the D . miranda physical genome ( Tables S1 and S2 ) ., For each of the three maps , nearly the entire assembled region of chromosome 2 ( 97 . 8%–99 . 4% ) , the majority of the XR chromosome arm ( 70 . 8%–89 . 4% ) , and part of the XL chromosome arm ( ∼22%–23% ) were surveyed ( Table S2 ) ., After removal of likely erroneous putative double recombinants , ambiguous genotypes , and markers that did not work or gave inconsistent genotypes , recombination was measured for three different crosses for 1 , 158–1 , 404 individuals per map ( Table S1 ) ., Excluding larger intervals at the telomeres and centromeres , intervals between markers had a median size across the three maps of 141–148 kb for chromosome 2 and 146–160 kb for the XR chromosome arm ( Table S1 ) ., For chromosome 2 , recombination rates ranged from 0–30 . 8 cM/Mb in D . pseudoobscura and 0–24 . 0 cM/Mb in D . miranda ( Table S2 ) ., The number of individuals surveyed is often slightly different per interval; therefore , for all intervals where no recombination was detected , we report 0 cM/Mb ., The recombination rate for those intervals with “0 cM” should be interpreted as <1 recombination event per total number of individuals surveyed for each interval ( Dataset S1 ) ., Recombination near the telomere and centromere was measured at a broader scale than the remainder of chromosome 2 because we expected these regions to have lower crossover rates than the center of the chromosome ( chromosome 2 is telocentric ) ., Because of this limitation , comparisons of recombination rates between the ends of the chromosome and the center are more tentative ., Nonetheless , examining recombination across roughly 3 Mb of sequence at the telomeric end and 3 Mb at the centromeric end , we found up to an 8 . 9-fold difference between the recombination rates at the middle of chromosome 2 relative to the centromeric end ., The Pikes Peak D . pseudoobscura map exhibited the largest reduction of recombination at the telomeric or centromeric ends relative to the center of the chromosome for all three maps , though in the Flagstaff D . pseudoobscura map and the D . miranda map , recombination rates were reduced by at least 2 . 6-fold in the centromere and telomere relative to the center of the chromosome ( Table S3 ) ., For the XR chromosome arm , recombination rates ranged from 0–25 . 2 cM/Mb in D . pseudoobscura and 0–32 . 3 cM/Mb in D . miranda ( Figure S3 presented with 95% confidence intervals; see also Dataset S1 , Table S2 ) ., The number of crossovers per individual for both chromosome 2 and the XR arm was close to 1 ( 1 . 01–1 . 06 ) for D . pseudoobscura and was 1 . 40–1 . 54 for D . miranda , illustrating that a greater overall recombination rate in D . miranda relative to D . pseudoobscura is observed in both an autosome and a sex chromosome ., The XL chromosome arm was not surveyed as intensively ( ∼22%–23% of the XL arm in Pikes Peak and D . miranda and ∼60% of the XL arm in Flagstaff; Figure S4 presented with 95% confidence intervals; Dataset S1 ) ., The number of crossovers per individual appears consistent with ∼1 crossover per chromosome arm , as in D . pseudoobscura XR and chromosome 2 , but the average number of crossovers per individual on the XL reflects how much of the arm was surveyed ., For example , when ∼22%–23% of the arm was surveyed , crossovers per individual ranged from 0 . 23–0 . 26 ( Table S2 ) ., A binomial Generalized Linear Model ( GLM ) with size of the interval as a covariate and interval identity as a factor in the model indicated significant heterogeneity in recombination rate among intervals for chromosome 2 , XR , and XL ( each tested separately ) for each of the three maps ( each tested separately , interval identity p<0 . 00001 , χ2≥64 . 67 , df≥3 , in all cases ) ., Furthermore , 95% confidence intervals ( generated via the same method in 54 ) do not overlap in many cases between different intervals ( shown in Figures 1 , S3 , S4; Dataset S1 ) ., Overall , we observe heterogeneity in fine-scale recombination rates within each of the three maps ( see Figures 1 , S3 , and S4 with 95% confidence intervals plotted; Dataset S1; statistical quantification between maps given in section below ) , and we note a reduction in recombination rate around the telomeric and centromeric ends consistent with other studies in Drosophila 33 ., Our three fine-scale crossover maps utilized markers on average 141–160 kb apart ( median interval size for each of the three maps , with the exception of XL where the median distance between markers was 200–1 , 775 kb for the three crosses ) ., We additionally examined three regions on chromosome 2 in more detail ., Each of these regions spanned a total of 99–125 kb , and we placed markers every ∼20 kb within the region ( 16 total intervals; Tables S4 and S5 ) ., These regions were originally picked because previous data 25 , 33 indicated that recombination rates for each of these regions differed ( regions are referred to as 6 Mb , 17 Mb , and 21 Mb , which indicate approximate location on chromosome 2 ) ., We refer to these as “ultrafine-scale” maps ., For these ultrafine maps , we followed the same backcross scheme as above , and we scored approximately 10 , 000 individuals for each marker ( Table S5 ) ., For the 16 ultrafine intervals ( Tables S4 and S5 ) , each interval was on average 20 . 61 kb long ( range 12 . 6–27 . 4 kb ) ., Recombination rates range from 1 . 6–21 . 2 cM/Mb for these ∼20 kb intervals ( Figure 2; see Table S5 for 95% CI ) ., The ultrafine-scale map uncovered variation in recombination rates that was not apparent with the fine-scale maps ., For example , for the 17 Mb ultrafine-scale region on chromosome 2 , the recombination rates for the two fine-scale intervals spanning this region ( 17 . 5–17 . 7 Mb ) are 5 . 6 and 4 . 4 cM/Mb ., The ultrafine-scale recombination rates , in contrast , ranged from 3 . 5–21 . 2 cM/Mb ( markers spanning 17 . 5–17 . 7 Mb ) ., This heterogeneity in recombination rates within the ultrafine regions was statistically significant ( binomial GLM similar to that described in fine-scale section above: p\u200a=\u200a0 . 0011 , df\u200a=\u200a14 , χ2\u200a=\u200a35 . 91; 95% confidence intervals given in Table S5 ) and highlights the fact that “broader” scale measures of recombination rates ( such as the fine-scale measures here ) are averages of true variation in recombination rate ., For comparisons of recombination rates between fine-scale maps , we restricted our analysis to intervals that were condensed to have nearly identical physical marker placement between the three fine-scale maps ( Figures S5 and S6; Table S6 ) ., Recombination was estimated as detailed above , using the number of crossovers spanning the newly defined physical intervals ., After condensing across all three maps , 97 intervals remained for chromosome 2 and 44 intervals for XR ( see Tables 1 and S6 fornumber of individuals , size , range of these condensed intervals , and base pairs between markers on each map ) ., The XL chromosome arm was not included in the analysis that used condensed intervals across maps because too few intervals overlapped between all three maps ., When comparing two maps , intervals were condensed between those two maps only ( see Datasets S2 and S3 for rare events logistic regressions for all two-map and three-map comparisons ) ., Recombination rates did not differ significantly between the two D . pseudoobscura maps for either the XR or chromosome 2 for the two-map comparisons ( each chromosome analyzed separately , rare events logistic regression , absolute value of z>0 . 3901 , p>0 . 236 , in both cases; Dataset S2 ) ., For chromosome 2 , one interval was significantly different in recombination rate after correcting for multiple tests 59 ., For the XR , no intervals between the two D . pseudoobscura maps were significantly different in recombination rate after correcting for multiple tests ., The 95% confidence intervals for the odds ratio of the difference between maps were narrow and located around zero , indicating that the maps are likely very similar ( chromosome 2 , 0 . 87–1 . 10; XR , 0 . 94 , 1 . 28; within-species two map comparison ) ., It is unlikely that the single significant difference observed within the same species is because of slight differences in marker placement between the two maps ., The marker placement for this interval was nearly identical between the two maps ( left marker , 102 nucleotides different between maps; right marker , 17 nucleotides ) ., For both chromosome 2 and the XR chromosome arm , Drosophila miranda had significantly higher recombination rates than both D . pseudoobscura maps ( Figure S5 , Table 1 , Datasets S2 and S3 ) ., A rare events logistic regression of two-map comparisons indicated that the recombination rate of the D . pseudoobscura crosses we surveyed is about 76%–78% of the D . miranda recombination rate we observed on chromosome 2 ( absolute z value>4 . 5374 , p<0 . 001 for D . miranda relative to either D . pseudoobscura map , Table 1 ) ., The recombination rate of D . pseudoobscura is about 68%–71% of the D . miranda recombination rate on the XR chromosome arm ( rare events logistic regression absolute z value>5 . 101 , p<0 . 001 for D . miranda relative to either D . pseudoobscura map , Table 1 ) ., After the global difference between D . miranda and D . pseudoobscura is accounted for by the rare events logistic regression , recombination rates within and between species appear very similar for chromosome 2 ( Figure S5; Datasets S2 and S3 ) ., None of the intervals for the two-map comparison between D . miranda and D . pseudoobscura–Flagstaff were significantly different after correction for multiple tests , though power to detect significant differences on a per interval basis was likely weak ( see confidence intervals in Datasets S2 and S3 ) ., For example , 15 of the 115 intervals on chromosome 2 showed at least a 3-fold difference in recombination rate between maps ( Datasets S2 and S3 ) , though this magnitude of difference was not significant in our rare events logistic regression after correcting for multiple tests ., Likewise , only one of the intervals for the two-map comparison between D . miranda and D . pseudoobscura–Pikes Peak was significantly different after correction for multiple tests , but 19 of the 123 intervals exhibited at least a 3-fold difference in recombination rate between maps for chromosome 2 ., The XR chromosome exhibited a qualitatively larger difference in recombination rate between species than chromosome 2 ., After the global difference between D . miranda and D . pseudoobscura is accounted for by a rare events logistic regression , two of the intervals between D . miranda and D . pseudoobscura–Flagstaff for the two-map comparison and seven of the intervals between the D . miranda and D . pseudoobscura–Pikes Peak two-map comparison were significantly different after correction for multiple tests ., Six of the 72 intervals between D . miranda and D . pseudoobscura–Flagstaff two-map comparison exhibited at least a 3-fold difference , and 12 of 102 intervals between D . miranda and D . pseudoobscura–Pikes Peak exhibited at least a 3-fold difference ( Dataset S2 ) ., Twenty-seven of 97 condensed intervals ( three-map comparison , condensed between all three maps ) for chromosome 2 were considered to be “conserved” within and between species ., This means that they displayed a nonsignificant difference across all three maps when analyzed with a rare events logistic regression and had an odds ratio between 0 . 62 and 1 . 615 after the effect of map identity was taken into account ., These “conserved” intervals were used for further downstream analyses ( see “Diversity , Divergence , and Recombination”; Table S7 ) ., For the XR , seven of 44 intervals condensed between all three maps were conserved within and between species according to the criteria outlined above ., In sum , we observe strong conservation in recombination rates within a single species , while between species , we see globally elevated recombination rates in D . miranda ., Once the global difference is accounted for , there are few intervals with significant differences in recombination rate within and between species ., Thus , it is possible and parsimonious that recombination rate is generally conserved at the scale examined here ( ∼180 kb ) over moderate evolutionary timescales ( 2–2 . 5 my ) ., We used various Illumina platforms to resequence genomic DNA from 10 D . pseudoobscura lines using virgin females from lines that were inbred for five or more generations with full-sibling single-pair mating ( Table S8 ) ., Drosophila pseudoobscura populations across North America display very little differentiation , as indicated by low FST values ( always<0 . 10 , often<0 . 05 for loci located outside of the inversion polymorphisms of the third chromosome ) 60 , 61 ., Therefore , the choice of strains sequenced for estimating diversity covered much of the species range but was fairly random ., We also sequenced two lines of D . persimilis ( one of these was provided by S . Nuzhdin ) , two lines of D . pseudoobscura bogotana ( one of these was provided by S . Nuzhdin ) , one line of D . lowei , and three lines of D . miranda ( two provided by D . Bachtrog , Table S8; Short Read Archive accession numbers SRA044960 . 1 , SRA044955 . 2 , and SRA044956 . 1; see also http://pseudobase . biology . duke . edu/ ) ., The divergence between D . persimilis and D . lowei was used to generate measures of a proxy for neutral mutation rate across the genome ., In all diversity and divergence calculations , the reference sequences for the D . pseudoobscura and D . persimilis genomes were both included 62 , 63 ., Details of diversity and divergence calculations are discussed in Text S1 ( see section titled “Fine-Scale Recombination Maps: Computational Methods for Diversity and Divergence Measures” ) ., Briefly , average pairwise diversity and divergence was calculated for 4-fold degenerate sites , focusing exclusively on unpreferred codons 64 , though we obtained very similar results when using all 4-fold degenerate sites ., Overall , recombination is significantly and positively associated with average pairwise diversity but not average pairwise divergence at 4-fold degenerate sites of unpreferred codons ., We examined this relationship in several ways ., We analyzed each chromosome for each uncondensed recombination map independently using a generalized linear model for diversity and a separate model for divergence ( Tables S9 , S10 , and S11 ) ., After accounting for multiple covariates , diversity at 4-fold degenerate sites of unpreferred codons shows a significant , positive relationship with recombination , while divergence at 4-fold degenerate sites of unpreferred codons does not ( Tables S9 and S10 ) ., This result is consistent for each of the three recombination maps ( D . pseudoobscura–Flagstaff , D . pseudoobscura–Pikes Peak , and D . miranda ) for both chromosome 2 and the XR chromosome arm ( Tables S9 and S10 ) ., The XL chromosome arm contained too few intervals for analysis for D . pseudoobscura–Flagstaff ., For D . pseudoobscura–Pikes Peak and D . miranda , diversity showed a significant , or nearly significant , positive relationship with recombination , while divergence did not ( Table S11 ) ., The analysis above suggests that the recombination–diversity relationship is probably the result of the effect of recombination on selection at linked sites ( sensu 13 , 18 ) ; however , inadvertently including regions with discordant recombination rates between species in the analysis above could result in a pattern that supports this hypothesis—even when recombination is predominantly mutagenic ( Figure S1 ) ., To resolve this potential bias , we restricted analysis to only regions that exhibited conserved recombination rates between all three chromosome 2 maps ( N\u200a=\u200a27 intervals; described above ) and examined recombination in association with average pairwise D . pseudoobscura diversity at 4-fold degenerate sites of unpreferred codons ( Table 2; Figures S7 and S8 ) and average pairwise D . pseudoobscura–D ., miranda divergence at 4-fold degenerate sites of unpreferred codons ( Table 3; Figures S7 and S8 ) ., The effect of recombination on diversity was significant when the analysis was restricted to only those regions with the most conserved recombination rates ( quasibinomial GLM , F\u200a=\u200a6 . 123 , p value\u200a=\u200a0 . 024 ) , and the effect of recombination on divergence remained nonsignificant ( quasibinomial GLM , F\u200a=\u200a0 . 138 , p value\u200a=\u200a0 . 714 ) ., These regions contained only one interval within 4 Mb of the telomeric end and no intervals within 4 Mb of the centromeric end of the chromosome; thus , these results are not a function of broad-scale regional recombination rate differences across the chromosome ., These results support the hypothesis that recombination affects diversity through the effect of selection on linked sites ., We did not perform an analysis on conserved windows for the X chromosome , as only seven intervals were conserved within and between species ., To determine the impact of recombination rate on selection at linked sites in the genome , we used two generalized linear models to analyze the relationship of recombination rate and several measures that may be indicative of the efficiency of selection: ( 1 ) abundance of nonsynonymous substitutions and ( 2 ) average pairwise nucleotide diversity at 4-fold degenerate sites around nonsynonymous substitutions ., We analyzed the association of recombination rate with these two measures in a generalized linear model framework to account for covariates such as gene density , GC content , and a proxy for neutral mutation rate ., Biased gene conversion may influence substitution rates; thus , we controlled for GC content in all of the analyses below 7 , 16 , 65 , 66 ., We did not consider gene expression as a covariate , though some studies point to a negative relationship with recombination rate 67 ., The relationship of recombination rate to nonsynonymous substitution abundance was examined with the D . pseudoobscura Flagstaff fine-scale recombination maps ., Nonsynonymous substitution abundance was measured as the nonsynonymous substitutions on the branch leading to D . pseudoobscura+D ., persimilis as identified with PAML ., The response variable was the number of nonsynonymous substitutions in each gene , and the covariates of the linear model included ( 1 ) the number of synonymous substitutions in the gene in question allowing for inclusion of genes where Ks\u200a=\u200a0 ( sensu 50 ) , ( 2 ) , GC content of the gene , ( 3 ) gene density of 50 kb on either side of the midpoint of the gene , and ( 4 ) average pairwise divergence at 4-fold degenerate sites of unpreferred codons between D . persimilis and D . lowei as a proxy for neutral mutation rate within the gene ., We found no relationship ( Table, 4 ) between recombination and nonsynonymous substitution abundance with the fine-scale data ( generalized linear model with Poisson distribution , z\u200a=\u200a−0 . 614 , p\u200a=\u200a0 . 539 ) ., In response to selective sweeps , a trough in diversity should be visible around selected variants 30 , 68–72 ., We analyzed diversity surrounding the nonsynonymous substitutions along the lineage leading to D . pseudoobscura+D ., persimilis identified by PAML ., We compared the average pairwise diversity patterns at 4-fold degenerate sites surrounding these substitutions in relation to the Flagstaff recombination rate and distance in basepairs from the substitution ( Text S1 ) ., In regions with high recombination rates , the footprints of selection are thought to be narrower than in regions with low recombination rates , where strong linkage between sites will create a stronger signature of sweeps 39 , 69 , 71 , 73 ., As a control , similar analyses were performed using synonymous substitutions along the D . pseudoobscura+D ., persimilis lineage following 68 ., Synonymous substitutions , in many cases , evolve in a more neutral fashion than nonsynonymous substitutions ( 68 , but see 74 , 75 ) ., In a recent genome-scale analysis conducted with data similar to what are presented here , little reduction in diversity was seen around synonymous substitutions 68; this study instead saw an increase in diversity , which disappeared after correction for local mutation rates ., We considered 60 kb on either side of the substitution along the D . pseudoobscura lineage divided into 1 , 000 bp nonoverlapping windows ( sensu 68 ) ., For each 1 , 000 bp window , the response variable was the number of polymorphic 4-fold degenerate sites ., The generalized linear model included the following covariates: ( 1 ) total 4-fold degenerate sites , ( 2 ) GC content , ( 3 ) proportion of coding bases , ( 4 ) divergence of D . lowei–D ., persimilis at 4-fold degenerate sites as a proxy for neutral mutation rate , and ( 5 ) proportion of bases that were nonsynonymous substitutions ., The identities of each nonsynonymous substitution were included as random effects ., This generalized linear mixed model with Poisson distribution included the following factors: absolute physical distance from the substitution , fine-scale-derived estimates of recombination rate , and the interaction between these two factors ., A negative interaction term means that short distances from a substitution and high recombination rates have similar effects on diversity as large distances and low recombination rates ., We expect the interaction term for distance and recombination rate to be much reduced in magnitude for synonymous substitutions in comparison to the nonsynonymous analysis ., We found a small but significant negative interaction term of physical distance from the nonsynonymous site and recombination rate on nucleotide diversity around nonsynonymous substitutions ( Poisson GLMM , z\u200a=\u200a−7 . 52 , p<0 . 001; Table 5 , Figures 3 and S9 ) ., In other words , higher rates of recombination allow for recovery of diversity at shorter physical distances from the nonsynonymous site than lower recombination rates ( Figure S9 ) ., In contrast , a weaker interaction was detected for the interaction of distance and recombination rate on diversity around synonymous substitutions along the D . pseudoobscura lineage ( Poisson GLMM , z\u200a=\u200a−2 . 43 , p\u200a=\u200a0 . 015; Table 6 , Figures 3 and S9 ) ., GLM plots for the very low recombination rates of <0 . 5 cM/Mb show wider dips in diversity ( and more associated noise; Figure S9 ) than plots for recombination rates of >0 . 5 cM/Mb ( Figure S9 ) ., Distance from a substitution had a positive , significant ef | Introduction, Results, Discussion, Materials and Methods | One of the most influential observations in molecular evolution has been a strong association between local recombination rate and nucleotide polymorphisms across the genome ., This is interpreted as evidence for ubiquitous natural selection ., The alternative explanation , that recombination is mutagenic , has been rejected by the absence of a similar association between local recombination rate and nucleotide divergence between species ., However , many recent studies show that recombination rates are often very different even in closely related species , questioning whether an association between recombination rate and divergence between species has been tested satisfactorily ., To circumvent this problem , we directly surveyed recombination across approximately 43% of the D . pseudoobscura physical genome in two separate recombination maps and 31% of the D . miranda physical genome , and we identified both global and local differences in recombination rate between these two closely related species ., Using only regions with conserved recombination rates between and within species and accounting for multiple covariates , our data support the conclusion that recombination is positively related to diversity because recombination modulates Hill–Robertson effects in the genome and not because recombination is predominately mutagenic ., Finally , we find evidence for dips in diversity around nonsynonymous substitutions ., We infer that at least some of this reduction in diversity resulted from selective sweeps and examine these dips in the context of recombination rate . | Individuals within a species differ in the DNA sequences of their genes ., This sequence variation affects how well individuals survive or reproduce and is transmitted to their offspring ., Genes near each other on individual chromosomes tend to be passed to offspring together—neighboring genes are unlikely to be separated by exchanges of genetic material derived from different parents during meiotic recombination ., When genes are inherited together , however , the evolutionary forces acting on one gene can interfere with variation at its neighbors ., Thus , variation at multiple genes can be lost if natural selection acts on one gene in close proximity ., Recombination can prevent or reduce this loss of variation , but previous tests of this phenomenon failed to account for recombination rate differences between species ., In this study , we show that some parts of the genome differ in recombination rate between two species of fruit fly , Drosophila pseudoobscura and D . miranda ., Avoiding an assumption made in previous studies , we then examine sequence variation within and between fly species in those parts of the genome that have conserved recombination rates ., Based on the results , we conclude that recombination indeed preserves variation within species that would otherwise have been eliminated by natural selection . | genome evolution, neutral theory, population genetics, genome sequencing, mutation, genome databases, genome complexity, genetic polymorphism, biology, genetic drift, sequence databases, natural selection, genetics, genomics, evolutionary biology, genomic evolution, evolutionary processes, genetics and genomics | Recombination rate in Drosophila species shapes the impact of selection in the genome and is positively correlated with nucleotide diversity. |
journal.pntd.0000656 | 2,010 | Estimating Contact Process Saturation in Sylvatic Transmission of Trypanosoma cruzi in the United States | Since the Brazilian physician Carlos Chagas discovered the parasite Trypanosoma cruzi in 1909 , much research has been devoted throughout the Americas to the study of its transmission and control , primarily in the domestic and peridomestic settings in which it is passed to humans , via triatomine insect vectors of the subfamily Triatominae ( Hemiptera: Reduviidae ) ., Although control measures have succeeded in preventing new infections among humans in some areas of Brazil , Uruguay , Chile , and Argentina , the parasite , which is native to the Americas , remains endemic in sylvatic settings as far north as the United States , being limited only by the habitats of the several vector species ., In each region , the epidemiology of sylvatic T . cruzi transmission differs in important particulars , as each host and vector species has certain peculiarities—behaviors or immunities—which have led to adaptations in the ways by which the infection is maintained ., In the United States , sylvatic hosts ( which rapid urbanization often brings into peridomestic settings ) include primarily raccoons ( Procyon lotor ) and opossums ( Didelphis virginiana ) in the southeast and woodrats ( Neotoma micropus ) in Texas , although dogs and armadillos have also been cited as significant , and the parasite is also found in skunks , foxes , squirrels , mice , and other Neotoma spp ., ( Vectors do feed upon birds , reptiles and amphibians as well , but these are refractory to T . cruzi infection 1 , and hence incompetent hosts . ), There are over 130 species of triatomine vectors , of which 11 are known to inhabit the southern United States , 8 of them in Texas 2 ., Two of the most important in the southeastern U . S . 2 , 3 are Triatoma sanguisuga , found from central Texas all the way east to islands off the Atlantic coast , and Triatoma gerstaeckeri , associated primarily with woodrat nests and domestic settings from central Texas south into Mexico as far as the state of Queretaro 4 ., In addition , there are different strains of T . cruzi circulating in these populations ., Strains are classified within six major groups known as Type I and Type IIa through IIe ., Of these , only Types I and IIa are known to circulate in the United States 5 , and it is widely believed ( primarily from experiments in mice , e . g . , 6–8 ) that the strains circulating in the U . S . are less virulent than those in Latin America , where the incidence of Chagas disease in humans is much higher: an estimated 16–18 million people ( only a handful of autochthonous cases have been diagnosed in the United States 9 , though it has also been estimated that as many as half a million people in the U . S . may harbor the parasite , due to migration from Latin America ) ., Among sylvatic hosts in the United States , raccoons and other placental mammals are associated with Type IIa infections , while opossums are associated with Type I infections 5 ., T . cruzi may be transmitted in a number of ways ., Historically , the primary infection route , especially in South America , has involved the vectors feeding process , in which a bloodmeal from an infected host can transmit the parasite to the vector , where it lives in the insects gut , and defecation by an infected vector on the host following the bloodmeal can result in stercorarian transmission to the host ., In sylvatic hosts this may occur when the animal scratches the bite and inadvertently rubs the parasite-contaminated matter into the lesion ., However , among humans there have recently been other transmission avenues of greater concern: the parasite can be passed from one human to another through blood transfusion and organ transplants , congenitally from mother to child through the placenta , and oral transmission by consumption of food contaminated by vectors has been blamed for outbreaks in South America ., In fact , these avenues of transmission may also be important for sylvatic hosts as well: vertical ( congenital ) transmission has been verified experimentally among rats 10 and supported by circumstantial evidence among lemurs 11 and other animals , and oral transmission to hosts through their predation upon vectors ( raccoons , opossums , and even woodrats are opportunistic feeders that commonly include insects in their diets ) has even been suggested by some 12 , 13 to be the primary means of T . cruzi transmission to hosts in some cycles in the U . S . Indeed , T . sanguisuga and T . gerstaeckeri are known to be so cautious in their feeding behavior as to avoid climbing up entirely onto hosts during feeding 3 , and often defecate 30 minutes or more after feeding ends , making them likely to be rather inefficient at stercorarian transmission to hosts ., Both oral and stercorarian transmission to hosts , however , as well as bloodborne transmission to vectors , may be amplified by changes in vector behavior caused by infection with T . cruzi ., Many disease vectors are known to increase their feeding rate when infected , due to parasites building up inside their digestive tracts and impeding feeding ., This behavior has been verified for one species of triatomine vector and trypanosome 14 , but not documented for Chagas vectors and T . cruzi ., Many of the still-unanswered questions regarding sylvatic T . cruzi transmission cycles may be exceptionally difficult to address through direct observation in the laboratory and field: for instance , which of the several transmission pathways is really dominant in each cycle ?, ( We may think of a cycle as a specified host , vector , parasite strain , and geographic region , although in practice such cycles communicate with each other , primarily via vector dispersal . ), Mathematical models have proven a useful tool in many fields , including ecology and epidemiology , as they can describe , predict , and provide evaluation measures for phenomena which may be difficult to observe directly ., Population biology models consisting of dynamical systems ( usually systems of differential equations , see , e . g . , 15 ) , which describe the spread and growth of populations over time , have made notable contributions to disease control beginning notably with Ronald Rosss study of malaria transmission in the early 1900s 16 , for which he later won the Nobel Prize ., Such mathematical modeling of T . cruzi transmission has to date involved primarily household-based modeling of vector infestations and human infection ( but see below for a notable exception ) , although in the past decade geospatial models have been developed to describe vector distribution , disease risk , and relevant ecological niches 2 , 17 ., The ability of mathematical models to explain and predict depends not only on the underlying assumptions about the biological processes ( demographic , infection-related and other ) used to construct them , but also on knowing the values of certain fundamental parameters , most of which can be observed directly: information such as average lifespan , population density , or the probability of a host becoming infected from consuming an infected vector ., For instance , the ability of a given population to invade or persist in a habitat often depends on threshold quantities such as a reproductive number ( which can be calculated in terms of these fundamental parameters ) being above or below a critical value ., The best-known of these is the basic reproduction number for an infection or population 18 , 19 , denoted , which typically signals persistence of the population precisely when ., In practice , however , the parameters values for a given transmission cycle change seasonally , from one region to another , and even from study to study ( especially if sample sizes are small ) ., As a result , the critical link between theoretical models and empirical data provided by parameter estimation requires a broad perspective and familiarity with a range of empirical literature ., As noted above , numerous mathematical modeling studies have been published of T . cruzi transmission to humans ( e . g . , 20–22 ) , but almost none have been published on the sylvatic transmission cycles that maintain the parasite ., Decades of studies have established details of the life cycles of T . cruzi hosts and vectors in the United States , but studies focused on measuring infection parameters are only just beginning to appear ( e . g . , 13 ) ., Mathematical models can bridge this gap by facilitating calculation of these parameters using enzootic prevalence observations together with known information on the life histories of host and vector species ., The aims of the present study are to estimate values for those measures of host and vector life histories and T . cruzi infection which have been observed directly in the literature via an extensive review , and then to illustrate a method by which other key infection-related parameters can be calculated using mathematical models ., One of the important aspects of the sylvatic T . cruzi transmission cycle which models can help investigate is density dependence in the infection rates ., ( In this paper the term “rate” refers to a frequency per unit of time at which an event occurs . The term “proportion” will be used to refer to ratios which do not involve time , such as disease prevalence . ), Infectious disease transmission is driven by contact processes between susceptible and infective individuals , and sylvatic transmission of T . cruzi in particular depends on both the vector-initiated process of taking bloodmeals and the host-initiated process of predation on vectors ., The rates at which these two contacts occur depend in part on the host and vector population densities , and in part on the ratio of those densities , due to the saturation that occurs when this ratio is too high or too low ., That is , the per capita contact rate is a function of the vector-host density ratio , so that the total contact rate is the product of this function and the respective ( host or vector ) density ., Ratio-dependent contact rates , which were used in epidemiological models as early as Rosss classic malaria model 16 , are also a well-established notion in the study of predator-prey systems 23 , 24 , and the present study will illustrate how these correspond to the density-dependent effects observed in the transmission of T . cruzi ( e . g . , 25 ) ., Saturation in contact processes—the notion that given rates can increase only up to a certain point—has also been studied extensively in the contexts of both predator-prey systems ( e . g . , 26 ) and mathematical epidemiology ( leading to the distinction between mass-action incidence for low densities and standard incidence for high densities ) ., Predation and infection are superimposed in the transmission of vector-borne infections , and empirical studies 25 , 27 have observed a corresponding density dependence in which per-vector biting rates decrease at high vector-host ratios ., Per capita contact rates thus increase with the density ratio only up to a certain limit , so that the total contact rates ( per capita rates multiplied by host or vector density ) then become functions of one density or the other alone ., When the ratio of vectors to hosts is low , hosts are plentiful relative to vectors , so on the one hand each vector can feed as often as it wants ( that is , at its preferred feeding frequency ) , but on the other hand an average host has a hard time finding vectors to consume , making both contact processes limited by the number of vectors ., When the ratio of vectors to hosts is high , however , there are not enough hosts upon which for the vectors to feed at their desired frequency ( requiring them to find other blood sources ) , but the hosts are able to eat until reaching satiation , so that both contact processes are limited by hosts ., One recent theoretical study 28 developed a mathematical model for sylvatic transmission of T . cruzi and determined that the way in which the two contact processes saturate can affect not only vector population densities but also whether the infection cycle persists ., Another study 29 found that such a model coupled to one involving human infection explained observed domestic prevalence data better than a model of exclusively domestic transmission ., In order for a mathematical model to predict the rate at which new infections occur , it is necessary to derive quantities such as threshold density ratios from empirical data , so as to understand in what phase of saturation the causative contact processes are operating ., This paper presents a way to do so ., This paper derives estimates for the key biological parameters needed to model sylvatic Trypanosoma cruzi transmission cycles in Texas and the southeastern United States involving raccoons , Virginia opossums , woodrats , and the two vector species Triatoma sanguisuga and Triatoma gerstaeckeri ., Many of these parameters can be estimated directly via an extensive literature review , but infection and contact rates will be estimated indirectly using estimated prevalence levels and a few properties of some relatively simple dynamical population models ., The results will also be used to address the issue of saturation in the two infectious contact processes ., The intention is to provide well-informed direct estimates of as many quantities as possible and a method for computing other estimates which can be applied to models designed to address a broad spectrum of questions ., An exhaustive literature review was used to derive estimates for basic demographic information on host and vector species , as well as those epidemiological parameters for which direct estimation is possible ., The review initiated with a Medline search on “Triatoma sanguisuga” , “Triatoma gerstaeckeri” , or “Trypanosoma cruzi” , together with “United States”—or , for general demographic information on hosts , keywords used were “raccoon” , “opossum” and “woodrat” ., From the over 1000 resulting articles , only those ( approximately 80 ) which reported data on one of the quantities estimated in the Results section of this paper were kept ., The vast majority of the papers discarded focused exclusively on genetics or microbiology , rather than population biology , and were discarded from the title and abstract; the full text of all other articles was examined for relevant data ., Results were found ( and kept ) in English , Spanish , and Portuguese ., References in the sources were then checked manually as well ., Gray literature was not specifically sought except for non-Chagas-related demographic information on host species not identified in scientific literature , but was checked when it appeared as a reference in another source ., Additional references were added at reviewers suggestions ., Well-established properties of nonlinear dynamical systems models were then used to estimate infection rates based on prevalence and known parameters , and to frame the estimation of the threshold population-density ratios that determine whether host or vector population densities drive each type of infectious contact ., ( Specific simple models are used as illustrations in the Results section , but the approach outlined can be applied to a wide variety of dynamical systems , and results are not meant to be limited to the models given . ), Models were used ( and will be discussed ) only where necessary to help estimate relevant quantities ., In every case , epidemiological quantities were estimated as time-averaged values over an entire year , in order not to allow seasonal fluctuations ( which impact both host and vector populations significantly ) to prevent study of endemic steady states and prevalence ., Basic demographic information on host and vector species is necessary for all modeling of T . cruzi transmission cycles ., Numerous studies have published data supporting the estimation of average lifespans for raccoons 30–34 , opossums 12 , 34 , 35 , and woodrats 36 , and references therein; reproductive rates for raccoons 30–32 , opossums 34 , 37 , and woodrats 37; population densities for raccoons 32 , 38–47 , opossums 40 , 41 , 48 , and woodrats 36 , 49 , 50; average lifespans for T . sanguisuga 3 , 51 and T . gerstaeckeri 3 , 52 , 53; reproductive rates for T . sanguisuga 3 , 12 , 51 and T . gerstaeckeri 3 , 53; and , in a single case , vector population density 54 ., Discussion and development of estimates for these quantities are provided in Text S1 ., Table 1 summarizes these estimates ( including SI equivalents ) for the demographic parameters of each species ., Vertical transmission of T . cruzi has been widely documented in humans , and estimated to occur with frequency between 1 and 10 percent in Latin America 55–58 ., Because the parasite is transmitted through the placenta and blood supply to the fetus , vertical transmission is possible among placental mammals , but it is generally not believed to occur among marsupials ., A study in Venezuela found a vertical transmission rate among Wistar rats ( Rattus norvegicus ) of 9 . 1% for a strain of T . cruzi isolated from dogs , but none at all for a strain isolated from humans 10 ., Another study in Georgia ( USA ) found that a Type IIa strain of T . cruzi isolate from Georgia was twice as likely to be vertically transferred in mice as a Type I isolate from South America 11 ., In the absence of any data on vertical transmission among raccoons , we might reasonably estimate that Type IIa strains are transmitted congenitally roughly 10% of the time ( as a proportion , ) , with Type I strains transmitted as much as an order of magnitude less frequently ( say ) ., There is almost no published data on rates of oral infection with T . cruzi ( which could be estimated directly by multiplying the predation rate of hosts upon vectors by the probability of infection following consumption of an infected vector ) , although the possibility of oral transmission has long been documented ., Olsen et al . , writing in the early 1960s , referenced a “postulate” that oral transmission was the primary route of infections for opossums in Alabama , with insects consisting of 43% of opossums diet by mass , and 60% by volume 12; Roellig et al . recently extended this notion to include raccoons as well 13 ., One recent source wrote , “Animals can easily become infected with T . cruzi when an infected triatomine bug is ingested . ” 59 However , despite a significant body of research on what raccoons , opossums and woodrats eat , a literature review revealed no data on how much ( or how often ) they eat ( in order to estimate predation frequency ) ., Rabinovich et al . 60 observed 33 instances of predation when each of 13 female white-eared opossums ( Didelphis albiventris ) was placed with 10 infected Triatoma infestans for a day , but the rather high predation rate estimate that would result from this data is skewed by the experimental conditions , e . g . , the fact that both opossums and bugs were starved for a period of time prior to the experiment , and the opossums had no other available food ., Since predation is opportunistic and there are other insects available to the hosts as well , we will therefore estimate predation to occur for all hosts no more often than one triatomine every 3 or 4 days , which equates to an upper bound of about vectors/yr/host ., However , it may also be orders of magnitude lower ., ( Woodrats are of course much smaller than raccoons and opossums , and hence eat less , but vectors are found much more easily in woodrat nests , at least by humans , so we will assume opportunity balances out total volume . ), The probability ( or proportion ) of infection of a host following consumption of an infected vector can be estimated from three experiments in which uninfected hosts were fed vectors infected with T . cruzi ., Yaeger conducted 11 trials of an experiment in which an uninfected Virginia opossum ( D . virginiana ) was fed two Rhodnius prolixus vectors 61 infected with a Type IIe strain; 3 of these trials resulted in infection , yielding an estimate for of ., Roellig et al . 13 conducted 2 trials of an experiment in which an uninfected raccoon was fed 3 R . prolixus vectors infected with strain IIa; both trials resulted in infection ( yielding an estimate for of 1 ) ., Finally , the aforementioned study by Rabinovich et al . 60 produced its own estimate of 0 . 075 for the infection probability of white-eared opossums by eating T . infestans infected with an unspecified strain of T . cruzi ( presumably not IIa ) ; since their experiment combined oral and stercorarian transmission ( all 6 of the 13 opossums who ate a bug were also verified to have been bitten by at least one other bug , except for the opossum who ate all 10 of the bugs placed with her ) , it is impossible to disentangle the raw oral transmission data in a way that can be pooled with the other two experiments ., Yaegers estimate for opossums is precisely twice that of Rabinovich et al . , although the difference is not inordinate ., Roellig et al . s data is based on so few trials that no great significance can be ascribed to the resulting high estimate for raccoons , but it is nevertheless suggestive that the probability of oral transmission may vary significantly by host species and by parasite strain ( opossums appear not to become infected when exposed to Type IIa T . cruzi 62 , and hence may be more difficult to infect with any Type II strain ) —not to mention vector species—which is entirely consistent with the speculation of some biologists that North American strains may have adapted in response to local conditions ., Obtaining a single estimate for opossums requires an assumption that differences due to species ( D . virginiana vs . D . albiventris ) , vector species , and possibly parasite strain are negligible , in which case we can take a weighted average of ., To estimate oral infection probability for raccoons we are left with either the above 100% estimate or else an average across all host species ( including opossums ) of ., There is likewise no published research on the extent to which infection with T . cruzi increases vector behaviors in T . sanguisuga or T . gerstaeckeri that promote infection ., Añez and East 14 found that triatomine bugs of the genus Rhodnius , a common T . cruzi vector in South America , probed or bit an average of 6 . 5 times as often when infected with the parasite Trypanosoma rangeli as when uninfected , prior to engorging ., This differential behavior may amplify by a factor ( say ) not only the biting rate of infected vectors but also their availability for predation due to increased mobility driven by hunger , so that the effective vector density for infection behaviors is rather than ., However , DAlessandro and Mandel 63 found no difference in the feeding behaviors of R . prolixus infected by T . cruzi ., Although such frequencies can be expected to vary widely by species ( of parasite as well as vector ) , it would be consistent with research on South American species to expect no differential behavior in infected T . sanguisuga or T . gerstaeckeri ., In the case where we wish to investigate the possible effects of such an amplification factor , however , it is worth noting Añez and Easts value ., Research suggests that in general sylvatic hosts do not suffer mortality from T . cruzi infections , even though high mortality rates have been reported for dogs , and the long-term risks have been verified for humans ., Also , the mice which die from T . cruzi infections in laboratory experiments are often injected with considerably higher concentrations than a single horizontal transmission is likely to produce initially ., We may therefore assume ( following , e . g . , 64 ) that in general the sylvatic hosts under study have no significant additional mortality caused by infection with T . cruzi ., Table 2 summarizes these parameter estimates ., ( Table 3 defines additional variables and parameters used in later sections . ), Estimation of the per capita infection rates for vector transmission must be made indirectly , as at present there are few published data on both the vector biting rate and the proportion of feedings which result in an infection in each direction ( host to vector and vice versa ) ., ( Two notable exceptions are 65 , which estimated the probability of vector infection per feeding for a specific South American cycle , and 60 , which estimated the probability of stercorarian infection of opossums D . albiventris at 0 . 06 95% CI: 0 . 023 , 0 . 162 per infected T . infestans bite ) ., Instead , given the long history of established T . cruzi infections in the regions of interest , we shall assume that the parasite has reached endemic equilibrium in the host and vector populations , and use published data to estimate endemic prevalence in both host and vector ., This will allow us to use the formulas derived from our population dynamics model which express endemic equilibrium prevalence as a function of model parameters , to calculate the infection rates necessary to produce those endemic levels ., With prevalence levels and all other parameter values known , it will be possible to solve for the infection rates ., But first we must estimate prevalence ., Reported prevalences are given in Tables 4–8 for raccoons , opossums , woodrats , T . sanguisuga and T . gerstaeckeri in the southeastern United States and northern Mexico ., Asterisks ( * ) denote studies which published paired estimates of host and vector prevalence ., For host prevalence , the method of diagnosis is given as hemoculture , serology ( IFAT\u200a=\u200aIndirect Fluorescent Antibody Test , IHA\u200a=\u200aindirect hemagglutination assay ) , either ( both culture and serological tests were performed , and a single positive is reported as positive ) , blood smear ( BS ) , or xeno diagnosis ., The dagger after the citations to Lathrop and Ominsky 66 marks joint prevalence reported for a mixed population of 6 T . sanguisuga and 9 T . gerstaeckeri ., As evidenced by Table 4 , dozens of studies have reported prevalence figures for the infection of raccoons with T . cruzi in the past fifty years , in states throughout the southeastern quarter of the United States ., As observed by several researchers , notably Yabsley et al . 67 , the method used to determine infection can have a significant effect on the results: in particular , the parasite is often only found in the blood ( by hemoculture or blood smears ) during the initial ( acute ) period of infection , while the immune system takes some time to develop antibodies to T . cruzi , so that serological tests like IFAT and ELISA are more likely to detect chronic infections ., It is therefore best to use both methods in order to capture both acute and chronic infections ., Most studies reported prevalence based only on blood cultures until about ten years ago , and as can be seen in Table 4 there is a marked difference in the prevalences reported based on hemoculture studies as compared to serological or both ., Ten of the sixteen blood-based studies reported prevalences of 15% or less ( seven of these reported prevalences of 1 . 5% or less , and the mean of all 16 values is under 20% ) , whereas apart from a single , small-sample ( n\u200a=\u200a12 ) zero value , the studies which included serological results reported a mean of over 50% prevalence ., There is also some notable geographic variation ., Infection rates near the central part of the country appear to be relatively high , with studies from Kentucky , Missouri , Oklahoma and central Tennessee all reporting prevalences of well over 50% , with a total prevalence of 106/163 or 65% ., On the other hand , the region directly east of that , from the mountains to the Atlantic , has little or no infection: studies from Maryland , Virginia , West Virginia and even eastern Tennessee adjacent to Virginia all report effectively zero prevalence , the exception being a study of raccoons in the suburban area of Fairfax County , Virginia , near Washington , D . C . , where increased opportunity for foraging results in a higher raccoon population density ., Prevalence among raccoons in Georgia and neighboring South Carolina ranges from 33% to 60% except for one hemoculture-based study which reported 22% ., Pooling these 7 studies yields an overall prevalence of 351/908 or 38 . 7% , heavily weighted by the large study of Brown et al . 68 ., Moving west along the Gulf Coast , there is no data apart from Olsen et al . s study from eastern-central Alabama in the early 1960s until we reach Texas , where there are only two small studies from 1977–1978 ., We shall take the figure of 24% from central Texas , rather than that of 0% from south Texas , as being representative of prevalence among raccoons in the central and eastern part of the state ., Examining the reported prevalences for opossums , there is a clear tendency for the studies which used both blood culture and serology to report higher prevalences ( see Table 5 ) , with the exception of the early datum from Texas , which was of such a small sample size ( n\u200a=\u200a8 ) that it cannot be claimed to be representative ., There is nearly an order of magnitude difference in sample size between the three largest studies 68–70 and the next largest , and these three show , on the one hand , nearly identical hemoculture-based prevalences between Texas ( 16% ) and Florida and Georgia ( 17% , consistent with the more recent Georgia figure of 15 . 4% 71 ) , and , on the other hand , a prevalence that nearly doubles when both hemoculture and serology are taken into account ( 28% in Georgia 68 ) ., Although some of the smaller studies suggest that in places the prevalence of T . cruzi in opossums may be much higher than this , we shall use Brown et al . s 28% figure as representative of prevalence in both the southeast and Texas ., The four earliest reported prevalences of T . cruzi infection in Texas woodrats are relatively close to each other ( ranging from 21 . 4% to 34 . 9% , see Table 6 ) but used hemocultures or blood smears rather than serology , which may imply an underestimate; the two reports from west Texas , both serological , are higher but come from much smaller samples ., We shall nevertheless pool the data to obtain an overall prevalence of 225/678 or 33 . 2% ., Very few studies have reported infection prevalence for the vector T . sanguisuga east of Texas ( see Table 7 ) ., The studies published by Hays , Olsen and their collaborators in the 1960s give prevalences of around 6% in eastern central Alabama , but the two more recent studies in Georgia and Louisiana agree on values an order of magnitude higher ., It is likely that infection prevalence does vary by location , but for an overall average we shall pool the two more recent reports , for a total prevalence of 56 . 5% in the southeast ., In Texas , reported prevalences appear to fluctuate within a range of 17% to 44% ., Pooling all but the first two studies ( since the second gave no absolute numbers ) yields an overall prevalence of 135/543 or 24 . 9% ., Early studies had T . cruzi prevalence in the vector T . gerstaeckeri varying widely from 5% to 92% ( see Table 8 ) , and despite some slight convergence , results continue to fluctuate from 26 . 5% to 77 . 4% , even among relatively large ( ) samples ( we exclude from further discussion the small sample from Queretaro in central Mexico ) ., Since these studies typically collected vectors from woodrat nests , it is likely that there may be considerable variation in infection proportions from one nest to another ., The three reports from the state of Nuevo León , Mexico , just south of Texas , also fit within this range ., We will therefore pool all studies for which raw data is given ( noting that the rate given in Galavíz et al . is close to that in the study by Martínez-Ibarra et al . , on which Galavíz was second author , and that the data in deShazo is likely incorporated into the study by Sullivan et al . given the dates , and the fact that deShazo | Introduction, Methods, Results, Discussion | Although it has been known for nearly a century that strains of Trypanosoma cruzi , the etiological agent for Chagas disease , are enzootic in the southern U . S . , much remains unknown about the dynamics of its transmission in the sylvatic cycles that maintain it , including the relative importance of different transmission routes ., Mathematical models can fill in gaps where field and lab data are difficult to collect , but they need as inputs the values of certain key demographic and epidemiological quantities which parametrize the models ., In particular , they determine whether saturation occurs in the contact processes that communicate the infection between the two populations ., Concentrating on raccoons , opossums , and woodrats as hosts in Texas and the southeastern U . S . , and the vectors Triatoma sanguisuga and Triatoma gerstaeckeri , we use an exhaustive literature review to derive estimates for fundamental parameters , and use simple mathematical models to illustrate a method for estimating infection rates indirectly based on prevalence data ., Results are used to draw conclusions about saturation and which population density drives each of the two contact-based infection processes ( stercorarian/bloodborne and oral ) ., Analysis suggests that the vector feeding process associated with stercorarian transmission to hosts and bloodborne transmission to vectors is limited by the population density of vectors when dealing with woodrats , but by that of hosts when dealing with raccoons and opossums , while the predation of hosts on vectors which drives oral transmission to hosts is limited by the population density of hosts ., Confidence in these conclusions is limited by a severe paucity of data underlying associated parameter estimates , but the approaches developed here can also be applied to the study of other vector-borne infections . | The parasite Trypanosoma cruzi , transmitted by insect vectors , causes Chagas disease , which affects millions of people throughout the Americas and over 100 other mammalian species ., In the United States , infection in humans is believed rare , but prevalence is high in hosts like raccoons and opossums in the southeast and woodrats in Texas and northern Mexico ., The principal U . S . vector species appear inefficient , however , so hosts may be primarily infected by congenital transmission and oral transmission caused by eating infected vectors ., Mathematical models can evaluate the importance of each transmission route but require as inputs estimates for basic contact rates and demographic information ., We estimate basic quantities via an exhaustive review of T . cruzi transmission in the southern and southeastern U . S . , and use properties of mathematical models to estimate infection rates and the threshold ( saturation ) population-density ratios that govern whether each infection process depends on host or vector density ., Results ( based on extremely limited data ) suggest that oral transmission is always driven by host density , while transmission to vectors depends upon host density in cycles involving raccoons and opossums , but upon vector density in cycles involving woodrats , which live in higher concentrations . | infectious diseases/neglected tropical diseases, ecology/population ecology, mathematics/nonlinear dynamics | null |
journal.pntd.0006188 | 2,018 | An agent-based model of tsetse fly response to seasonal climatic drivers: Assessing the impact on sleeping sickness transmission rates | The tsetse fly ( genus: Glossina ) is the vector for human African trypanosomiasis ( HAT ) or sleeping sickness , a neglected tropical disease caused by two sub-species of the protozoan parasite Trypanosoma brucei s . l . : T . b ., rhodesiense , in eastern and southern Africa and T . b ., gambiense in West Africa 1 ., T . b ., rhodesiense HAT ( rHAT ) is a zoonosis , affecting a wide range of wildlife 2 , 3 and domestic animals , particularly cattle 4 , presenting in humans as an acute disease 5 ., The history of HAT in sub-Saharan Africa is characterised by long periods of endemicity where the disease self-sustains at low background levels , with periodic epidemics in regional foci 6 ., As sleeping sickness is a neglected tropical disease , treatments are often out-of-date , difficult to administer , physically invasive and partially validated , with the prospect for future developments of more effective treatments being limited ( e . g . 7–11 ) ., Furthermore , where tools are available , HAT is rarely prioritised due to competing public health interests 12 ., In terms of disease prevention , there is currently no immunological prophylaxis to stop infection in humans 13 , made difficult to produce due to the parasite being able to evade the hosts immune response by altering the antigenic character of its glycoprotein surface coat 14 ., Given these difficulties with preventing and treating HAT infection in humans , it is not surprising that mitigation strategies focused on vector control have seen success ( e . g . 15–18 ) , given that the tsetse fly is not only required for transmission , but also for several stages of parasite development 19 , 20 ., Despite such efficacy , the control of the disease in tsetse ( and , therefore , wildlife ) in game reserves and other protected areas is complicated by ecological , conservationist and environmental considerations 21–23 ., Gaining a greater understanding of the population dynamics in a tsetse population appears to be an attractive goal , considering that such an understanding could lead to the development of more targeted vector control strategies which have a less adverse ecological impact , while also allowing a more plausible understanding of the rHAT transmission system ., For the latter , demographic growth ( through the availability of food and habitat ) and climate changes ( affecting tsetse development and mortality rates ) are two factors which could affect tsetse population dynamics , and ultimately affect the transmission system 24 , 25 ., As a result of the significant role that a tsetse population has in determining the rate and distribution of rHAT transmission , this paper considers the tsetse sub-component of the larger rHAT transmission system in detail , with the ultimate goal being the creation of a more accurate representation of the transmission system as a whole ., Collecting comprehensive data on populations of tsetse in the field is expensive , complex and time consuming and , consequently , numerous attempts have been made to model tsetse populations as part of vector control or HAT transmission studies ( e . g . 26–29 ) ., Some models incorporate climatic drivers which create fluctuations in the tsetse population through the seasons ( e . g . 30–33 ) ., One recent example used agent-based modelling ( ABM ) techniques to simulate a simple fluctuation in tsetse population size through different seasons by altering the length of a predetermined lifespan for tsetse , depending on whether the tsetse emerges in the dry ( 2 months ) or wet season ( 3 months ) 31 ., Incorporating more detail , 33 used known relationships between temperature and different life events and processes , such as mortality and the length of the pupation period , as parameters when constructing a population model for vector control ., ABMs are “a computerized simulation of a number of decision-makers or agents , and institutions , which interact through prescribed rules” 34 ., ABMs have been described as a “third way” of conducting scientific research , incorporating both deductive since ABMs start with basic assumptions , and inductive approaches , as they produce simulation data to analyse 35 ., However , Epstein 36 suggests that rather than inductive or deductive , ABMs should be considered as “generative” tools in that , through the initialisation of a population of autonomous agents in a relevant spatial environment , one can allow the agents to interact given a simple set of local rules , and generate , from the bottom up , the macroscopic behaviour and regularity of the population as a whole ., Such an approach lends itself well to both the investigation of the HAT transmission system as a whole and the tsetse populations and their dynamics as a component ., Starting with tsetse population dynamics , much is written about how varying climatic conditions have different impacts on various tsetse life events and processes e . g . : pupal period duration ( e . g . 37 ) , probability of pupal death ( e . g . 38 , 39 ) , and time between oviposition ( e . g . 40 , 41 ) ., Representing observations made from samples acquired both in the field and laboratory studies , these patterns provide us with a solid framework to model the larger population , for which comprehensive data are much more difficult , if not impossible , to acquire ., By initialising a tsetse population as individuals , each abiding by rules set by the above behavioural patterns ( and others relating to feeding , mating and age-dependent mortality ) , plausible population level outcomes such as fluctuations in population size should be observable as the simulation progresses ., When the HAT transmission system is incorporated into an ABM for acquiring preliminary knowledge of the disease transmission system , the constructed model becomes a representation of a complex system ( e . g . 42–44 ) , given that the prevalence of the disease is a complicated emergent phenomenon produced by relatively simple , individual specific rules ( both vector and host ) concerning movement and resource acquisition ., In a complex system , the causes of emergent phenomena cannot easily be decoupled and explained by specific parts of the system 45 with , in this case , the model landscape and agent behaviour creating variation in the timing , location and probability of infection as a result of their influence on variability in contact patterns between vector and host 46 , 47 ., In this way , ABMs could be considered the most appropriate way to investigate both the HAT transmission system , and tsetse fly dynamics as a sub-component , allowing the representation of interdependent processes such as how individuals interact with each other and their environment through space and time more easily than is possible through more traditional epidemiological techniques 48 ., In previous work , an ABM of rHAT transmission was produced using a spatialized approach , incorporating factors often overlooked ( e . g . human behaviour and activity-based movement; density and mobility of vectors; and the contribution of additional hosts ) 27 ., This paper presents the first ABM which considers the effect of climatic factors on individual tsetse and their life processes in detail , while also considering the effect this has on rHAT transmission in a large study area in Eastern Province , Zambia ., Through the incorporation of seasonality parameters into an existing fine spatial and temporal scale ABM of rHAT transmission in the region 27 , the aim was to develop a greater understanding of tsetse population dynamics through simulation , and subsequently produce a more plausible model of rHAT transmission ., The incorporation of such data is vital where transmission rates , and indeed the transmission system as a whole , are to be explored over multiple years ., The existing model provided a suitable starting point for the simulation of these seasonal parameters by modelling tsetse flies at the individual level , along with different life events for which durations and probability of occurrence can be climatically constrained ., Ultimately , the modified model was implemented with the aim of answering the following research questions: throughout the year , how does the tsetse fly population fluctuate both as a whole , and within different life stages ( e . g . pupal , teneral , mature ) ?, Under the caveat that a plausible model has been produced , what rates of disease transmission are observed , and how do these vary seasonally ?, Such a model will allow for future exploration of long-term mitigations strategies , alterations to the demographic make-up of the study area , and climate change scenarios ., Eastern Province , Zambia is situated in southern Africa , sharing borders with Malawi ( to the East ) and Mozambique ( to the South ) ., The Luangwa Valley is an extension of the Great Rift Valley of East Africa , traversing the Zambian Eastern , Northern and Muchinga Provinces ., The valley is characterised as a flat bottomed valley bounded by steep , dissected escarpments which rise to a plateau at approximately 900–1000 m 49 ., Different types of vegetation are observed at different altitudes , with valley areas consisting mainly of mopane woodland and patches of grassland , while the natural vegetation on the escarpment and plateau is miombo woodland , interspersed with munga woodland 50 ., The study area spans a sparsely populated region of the Luangwa Valley ., Villages are small ( between 5 and 20 households ) and inhabitants are predominantly subsistence farmers ., The data collection area and region to be modelled consists of a 75 km transect which starts close to Mfuwe airport in the north , and runs southwards along the Lupande River and its distributaries ( Fig 1 ) ., Average monthly temperature and rainfall measurements collected at the Mfuwe airport ( 1982–2012 ) weather station are reproduced in Fig 2 51 ., There are three main seasons in Zambia’s tropical climate: the rainy season spans November to April ( wet and warm ) with mean monthly rainfall peaking at 210 mm in January ., After the rains , a cold and dry period occurs prior to August , in which May is the hottest and wettest month , with mean temperatures below 23°C and mean rainfall below 3 mm ., The hot and dry season usually spans August , September and October , with mean temperatures reaching 28°C in October accompanied by 17 mm of rainfall on average , the first after four dry months in succession 49 , 51 ., The Luangwa River and its main tributaries are perennial , and although flash flooding occurs in all rivers during the wet season , the smaller rivers which drain the valley floor dry out during the dry season and flow during the rains 52 ., rHAT is endemic in the Luangwa Valley , first being reported in 1908 53 ., G . m ., morsitans was not originally considered a vector of rHAT in the valley , despite 50% of domestic and game animals in the Valley having been observed to harbour trypanosomes 54 ., In the early 1970s , a large rHAT outbreak occurred in Isoka ( 241 case in 3 years ) attributed to fly encroachment from Luangwa 55 ., Wildlife had been observed to reside in Isoka for several months during the rainy season , migrating away during the dry season ., In 1973 , early diagnosis and improved treatment methods were introduced , and case numbers fell 56 ., Today , cases of rHAT continue to be reported in the Luangwa Valley ., Mid-Luangwa Valley has recently experienced increased immigration of people seeking fertile land ., Land pressure has resulted in human settlement in increasingly marginal , tsetse-infested areas , previously avoided for fear of disease risk to introduced livestock ., Households grow cotton as a cash crop and maize and groundnuts for home consumption 49 ., These anthropogenic changes have the potential to destabilise current trypanosomiasis transmission cycles , resulting in increasing prevalence of trypanosomiasis in both human and animal hosts , and the spread of rHAT into previously unaffected areas ., Risk factors include human proximity to the large wildlife reservoir in the South Luangwa National Park to the north-west 2 , and ever-increasing livestock and human density on the plateau ., Little is known concerning tsetse-trypanosome-human interaction in the region ., Therefore , the ABM has the potential to enable exploration of contact risk within communities ., Furthermore , with climate changes expected to occur in the near future , such as reduced annual rainfall , increased storm events and increased temperature 5758 , it is becoming increasingly important to understand how climate factors can affect tsetse populations , particularly in areas such as this , where increases in temperature could see the tsetse habitat spreading further up the valley to more populous areas ., This paper describes a new , seasonally sensitive ABM for rHAT/animal African trypanosomiasis ( AAT ) , based on an earlier , non-seasonal model that was constructed using data derived from a detailed rHAT , AAT , and G . m ., morsitans ecological survey , undertaken in 2013 , in Eastern Province , Zambia 27 ., Due to the fine spatial and temporal scales used to model the system , and the number of mechanisms incorporated ( e . g . , tsetse reproduction , tsetse feeding , human agent movements using real-world routines and pathfinding techniques 59 ) , the model was complex and its data inputs were numerous ., As a result , only new data and modifications to the original model are described here ., A detailed description of the original , non-seasonal model framework , and the data used to construct it , can be found in 27 ., The previous iteration of the model included a longer pupal duration in males than in females , as suggested in the literature ( e . g . 37 , 60 ) , and so for each larva deposited during the simulation , a 35 and 30 day pupal period was included for males and females , respectively , represented as a period of inactivity ., However , pupation is known to be temperature sensitive with pupal periods decreasing with increasing temperature , a relationship observed by Phelps and Burrow’s laboratory experiments at constant temperatures 37 ., Hargrove 41 utilised the data to present a near perfect fit for pupal duration at temperatures between 16°C and 32°C ( r2 = 0 . 998 ) ( see Fig 5 ) , represented by Eq 2:, r=k31+e ( a+bt ) ,, pupalduration=1r ,, Eq 2, Where: t = temperature , for males: a = 5 . 3 , b = -0 . 24 and k3 = 0 . 053 and for females: a = 5 . 5 , b = -0 . 25 and k3 = 0 . 057 ., Given the excellent fit to the data and the large variation in pupal periods expected within the temperature range found in the study area ( 19°C = ~60 days , 28°C = ~20 days ) , variation in pupal duration with temperature is clearly an important factor to incorporate in the model ., The previous , non-seasonal ABM provided the majority of the methods and data used in the current version of the simulation , and so readers are referred to 27 for greater detail and only a summary is provided here ., Census data were used to locate and initialise the human and animal populations living in the households shown in Fig 1 ., A sample of resource-seeking routines sorted by gender and age was taken in the field ( see supplementary information of 27 ) , and a set of plausible paths from each village to each resource was created using a pre-processing A* pathfinding technique 59 ., For tsetse , an estimate of the total apparent population size , density and distribution was provided ., Four agent types were included in the ABM , together with an areal representation of wildlife ., Humans , cattle , other domestic animals and tsetse used in the ABM were constructed as four separate classes , with populations modelled 1:1 with the data collected in the census ( e . g . 16 , 024 human agents ) and the estimated tsetse population discussed previously ., Each class had its own initial information and storage structures for events that occurred through the simulation ., The ABM was written in Python 2 . 7 using an object-oriented framework , and run on the Lancaster University High End Computing ( HEC ) Cluster , with all spatial data being processed using Quantum GIS 1 . 8 . 0 ., The subsequent sections draw attention to any modifications between the original , non-seasonal modelling framework and the new ABM model , while also describing how the climatic drivers affecting the tsetse population were incorporated into the model ., The initial iteration of the model was split into 2 , 400 time-step ( or tick ) days , as the more frequent the tick , the smaller the jumps made by agents as the simulation updates , and the less chance of missing potential interactions ., However , this method was restrictive in terms of memory usage and CPU time required to run just six months of the simulation ., Further tests were carried out to establish how coarse the temporal resolution could be made before the number of simulated domestic host-vector contacts was reduced , and a greater daily probability of wildlife feed was required to maintain the tsetse population levels ., It was established that 600 ticks per day ( 2 . 4 minutes per tick ) allowed the simulation to progress with no obvious effect on human , cattle and other domestic animal bite numbers , while requiring a very similar daily wildlife bite probability to produce a stable tsetse population ( 37% chance per day of a hungry tsetse taking a wildlife bite , compared with 35% in the previous version ) ., As a result , 600 ticks per day were used to produce the results of this investigation , which required approximately 4 . 5 GB of RAM per simulation run on the high performance machine , and 24 hours of CPU time per simulated year ., To capture the effect of seasonality on the tsetse fly population , daily temperature was calculated every 24 hours using the interpolation method discussed previously , and set as a global variable for the simulation ., For each female , once mated , the number of days since mating was compared with the birth interval calculated using Eq 1 and the daily temperature ., If and when the number of days since mating exceeded the interval calculated on a given day , a pupa was deposited ., A count of the number of days since mating was replaced with a count of the number of days since last offspring , and Eq 1 was used again on a daily basis ( using the alternative constants for further births ) , until another birth occurred ., This process was repeated for the duration of a female tsetse fly’s lifespan ., There was an equal chance of each tsetse offspring being male or female , and each pupa was deposited in a bush area in which the female tsetse rested during the previous night ., A rolling average of the temperature that each pupa has experienced since birth was calculated and attributed to each individual ., This temperature was used to determine each individual’s pupal duration , given that if a pupa’s age exceeded the pupal duration calculated using Eq 2 , the pupa would emerge as a teneral fly ., It was considered important to use a rolling average of temperature here as the length of a pupal period can span months with quite different temperatures ., As described previously , rather than a single probability used to decide whether a pupa would die during its entire pupal period , a variable daily probability of pupal death was included , increased in some months to account for losses observed in the rainy season ., Should the probability be exceeded for a pupa , that tsetse was removed from the simulation ., Death could result from pupal mortality , starvation , or if a tsetse fly exceeded the daily mortality rate calculated by sex , age and temperature ( Eq 6 , Figs 7 and 8 ) ., The mortality rate was calculated individually for each teneral and mature fly , and if the probability was exceeded , the tsetse was removed from the simulation ., Starvation occurred if a tsetse tried and failed to feed before a given period of time had elapsed ., The starvation element was more strict for teneral flies ( 3 days instead of 5 days ) highlighting their increased vulnerability and reduced flight strength ., In the previous version of the simulation , 75 teneral files were added to the simulation for the first 35 days to account for pupae deposited prior to the start of the simulation ., As this version of the simulation started in August , and the simulated climate quickly became hostile for teneral flies as temperature increased , 500 teneral tsetse were required per day for the first 45 days , which is representative of average simulated pupal maturation rates during September as the simulation progressed ( see Results ) ., In the original model , in the absence of climatic factors , a scaling factor for adult fly mortality was required to offset fly starvation within the simulation ., This value was set at 55% ., Although this scaling factor is still required in this iteration of the model due to the same starvation element , the incorporation of temperature dependent mortality , and more detailed mechanisms for modelling pupae , has reduced the required level of scaling to 80% To allow the model to initialise and stabilise , the simulation was run for a year before the results for this paper were produced , allowing a ‘burn-in’ period ., For example , the results presented below are representative of years 2–4 of the simulation ., 100 repeat simulations were used to produce the results presented here ., At the end of the three year simulation , a relatively stable population record was observed in both the male and female tsetse populations , with both exhibiting a double peak in response to the climatic driver ( Fig 9 ) ., Each year , until peak temperature was reached in October and November , the population slowly increased , with each gender’s population size increasing by approximately 2000 flies ., Such population increases during this hot and dry season could be attributable to the absence of a boosted pupal mortality which is observable during the rainy season 61 , with increasing temperatures having a greater impact in reducing pupal duration and the period between births , than increasing tsetse teneral and mature tsetse mortality ., During the rainy season ( November-April ) , this population gradually fell to an annual low , a result of peaks in pupal mortality at the start of the rainy season , and high temperatures causing increased mortality in the annual peak population of teneral flies ( see Fig, 10 ) ( now emerged after a high period of births discussed previously—birth numbers can be seen in Fig 11 ) ., During this period , with a reduced number of pupae to develop , and teneral tsetse to mature and start reproducing , the higher temperatures no longer aided a growth in population as there were fewer pupal maturations and birth rates to ‘accelerate’ ( Fig, 11 ) At the end of the rainy season , the tsetse population gained a small boost due to a plateau in temperature , and the drop in population slowed through the cool and dry season ( May to July ) , although recovery did not start during this period as temperatures were too low to aid rapid repopulation of the tsetse , and the pupal population was still recovering ( Fig 10 ) ., Fig 12 presents the different possible modes of tsetse death included in the model , and how the rates varied as the simulation progressed ., Non-starvation death represented the deaths attributable to the age-temperature dependent mortality model defined by Eq 6 , and was consistently responsible for the largest number of daily deaths , peaking in the period of highest temperature with approximately 350 deaths per day ., Unsurprisingly , given its temperature dependency , the mortality shape closely aligned to mean monthly temperature , except for a period in February and March after the pupal population was reduced by a period of high pupal mortality during the rainy season , resulting in a reduced teneral population and , therefore , fewer adult deaths ., Deaths due to starvation followed closely the general pattern of population size , with teneral starvations being particularly low–likely a result of the low daily teneral population size ( ranges between 100 and 400 –Fig, 10 ) and the teneral tsetse population having the highest age-temperature dependent mortality rate ., Using the Ackley and Hargrove model 61 for pupal mortality produced peaks prior to the rainy season and , to a lesser extent , after the wettest months ( Fig 12 ) ., The ratio of pupae to mature tsetse was approximately 2:1 at any given time , with the mature to teneral population ranging between 15:1 at the peak of population size and 25:1 when population sizes were generally lower ., Across the three year simulation , the approximate incidence rate for human and cattle rHAT infections was 0 . 355 per 1000 person-years ( SE = 0 . 013 ) , and 0 . 281 per 1000 cattle-years ( SE = 0 . 025 ) ., There were 11 human infections each year on average ( i . e . per year , per run ) , and 2 cattle infections ., Fig 13 illustrates how these infections clustered spatially and by season ., The aggregate number of infections across all years and each of the 100 repeats was used to produce this heat map due to the low infection numbers ., There was not much spatial variability through the seasons despite the variation in tsetse population size ., However , the number of infections reduced during the second half of the rainy season with the lowest density of infections observed during the cool and dry months ., Two hotspots are visible in each of the seasons , each with elongated elements suggesting that frequently used paths were sources of interaction between vector and human host ., This is possibly most visible in the north as east-to-west movement here could represent movement between villages and the river , a hypothesis which is given support by observations of infection by activity ( Table 1 ) which suggest that in each season , water collection accounted for approximately 25% of human infections , second only to school trips which accounted for 49% to 51% of infections ., No human infections were acquired whilst watering or grazing cattle , while the third highest number of infections occurred when farming ., There was little variation in infections by activity between the seasons ., With the observed high proportion of infections coming from school trips , it is unsurprising that 5–10 year olds and 10–18 year olds had the highest infections rates ( Table 2 ) ., Infection rates were generally lower in the cool and dry season , peaking in the hot and dry season ., Table 3 shows that the highest incidence rates were observed amongst immigrant tribes , with the only indigenous tribe ( the Kunda ) exhibiting one of the lowest infection rate across each time period , despite making up over 70% of the population ., Infection rates observed by gender and cattle ownership were comparable across time periods , with males and cattle owning households exhibiting marginally higher infections rates in comparison to females and households without cattle ( Table 4 ) ., Infections acquired and matured within the tsetse population fluctuated as the three year simulation progressed , with a small year-on-year increase in average infections both in the midgut and salivary gland ( Fig 14 ) ., On average , the peak time of salivary gland infection development was at the beginning of the rainy season , which reflects the period of highest tsetse densities plus a time-lag for development of mature infections in the fly ., The first plausible individual-based model representation of a real world tsetse population was created allowing a simulation of the system over multiple years ., The model was specified using temperature-dependent parameters derived from the literature , detailed human and animal information from acquired datasets , and expert opinion , and an estimate of the initial tsetse population size and distribution ., For example , the pupal population which was completely emergent from the model ( as no initial pupae data were inputted ) corresponded with literature findings that pupae are comparatively difficult to find in the rainy season , and that the pupal population will be greater than that of the developed flies 38 , unsurprising considering that the parameters suggest that pupae are ‘safer’ than teneral flies , pupal duration is at least 3 weeks , and a constant flow of developing pupae is required to replace teneral files which are dying or maturing ., In addition , the ratio of female-to-male tsetse fluctuated around 2:1 , a change from the simpler , non-seasonal model 27 , but more in line with estimates in the literature 67 , possibly as a result of running the simulation for longer , and with the addition of climate-driven parameters ., The shape of the mature population was comparable to samples of tsetse collected in the region of the South Luangwa National Park ( Regional Tsetse and Trypanosomiasis Control Programme ( RTTCP ) data reported in 22 ) , Eastern Province , Zambia 68 , G . pallidipes in neighbouring Zimbabwe 61 , and similar , yet less detailed , ABM studies 31 ., The peak adult population of around 6500 flies suggests that the relatively crude technique used to extrapolate sample data from tsetse surveys for initial model construction ( see 27 for more detail ) produced a reasonable estimate with 5250 flies ., Furthermore , the small teneral population observed is perhaps not a surprise , given that the teneral stage is a brief transition with a gradual input of developing pupae , and high mortality rates coupled with maturation to adult fly on first feed as outputs ., The decrease in pupal population during the rainy season , combined with a consistently small teneral population highlights how one or two years with a very hot and wet rainy season could have serious consequences for a tsetse population , with a reduction in pupal development during periods of high mortality , and high temperatures killing more teneral tsetse reducing the birth rate over subsequent months ., Similarly , such a relationship could occur over the coming years in response to climate change , with IPCC reports suggesting that more extreme rainfall events could occur , along with a rise in temperature over the next 50 years ( e . g . 57 , 58 ) ., As a result , it is not surprising that some studies have suggested that certain tsetse fly populations could face extinction within the next 50 years 69 ., Future studies will consider using the present model as the basis to test future climate change scenarios and examine the response in the tsetse population to such perturbations ., The model suggested similar incidence rates for rHAT infection in humans and cattle , which is likely to be a response to both the fact that the majority of the cattle were in households at the south of the transect , away from the tsetse zone ( only approximately 550 of 2925 cattle were within close proximity of the tsetse zone ) 27 , and that humans were modelled to be much more active than cattle in the simulation , travelling more frequently away from the home ., The latter point is corroborated by similar observations of human incidence rate in both cattle owning and non-cattle owning households , particularly as no human infections occurred while tending to cattle in the field or by the river ., As with observations in the previous study , collecting water and school attendance provided the highest proportion of infections by some margin , and is likely to be in response to the high frequency of both trips within the simulation and , for schools , the longer distances travelled to a sparse resource , and the time of day of the trips coinciding with tsetse activity ., In support of these simulated observations , a recent study of rHAT infections in Zambia found that almost half of the observed female infections were found in school-age children 70 ., The data for males suggested fewer infections in children ., This perhaps reflects that school attendance in the model is ove | Introduction, Methods, Results, Discussion | This paper presents the development of an agent-based model ( ABM ) to incorporate climatic drivers which affect tsetse fly ( G . m . morsitans ) population dynamics , and ultimately disease transmission ., The model was used to gain a greater understanding of how tsetse populations fluctuate seasonally , and investigate any response observed in Trypanosoma brucei rhodesiense human African trypanosomiasis ( rHAT ) disease transmission , with a view to gaining a greater understanding of disease dynamics ., Such an understanding is essential for the development of appropriate , well-targeted mitigation strategies in the future ., The ABM was developed to model rHAT incidence at a fine spatial scale along a 75 km transect in the Luangwa Valley , Zambia ., The model incorporates climatic factors that affect pupal mortality , pupal development , birth rate , and death rate ., In combination with fine scale demographic data such as ethnicity , age and gender for the human population in the region , as well as an animal census and a sample of daily routines , we create a detailed , plausible simulation model to explore tsetse population and disease transmission dynamics ., The seasonally-driven model suggests that the number of infections reported annually in the simulation is likely to be a reasonable representation of reality , taking into account the high levels of under-detection observed ., Similar infection rates were observed in human ( 0 . 355 per 1000 person-years ( SE = 0 . 013 ) ) , and cattle ( 0 . 281 per 1000 cattle-years ( SE = 0 . 025 ) ) populations , likely due to the sparsity of cattle close to the tsetse interface ., The model suggests that immigrant tribes and school children are at greatest risk of infection , a result that derives from the bottom-up nature of the ABM and conditioning on multiple constraints ., This result could not be inferred using alternative population-level modelling approaches ., In producing a model which models the tsetse population at a very fine resolution , we were able to analyse and evaluate specific elements of the output , such as pupal development and the progression of the teneral population , allowing the development of our understanding of the tsetse population as a whole ., This is an important step in the production of a more accurate transmission model for rHAT which can , in turn , help us to gain a greater understanding of the transmission system as a whole . | African trypanosomiasis is a parasitic disease which affects humans and other animals in 36 sub-Saharan African countries ., The disease is transmitted by the tsetse fly , and the human form of the diseases is known as sleeping sickness ., In an attempt to improve our understanding of the mechanisms which contribute to sleeping sickness transmission , a detailed , seasonally driven model of the tsetse fly has been produced , with the theory that a greater understanding of the disease vector’s life cycle will allow developments in our knowledge of disease transmission ., The model incorporates previously developed spatial data for the Luangwa Valley case study , along with demographic data for its inhabitants ., Tsetse and potential human and animal hosts are modelled at the individual level , allowing each contact and infection to be recorded through time ., Through modelling at a fine-scale , we can incorporate detailed mechanisms for tsetse birth , feeding , reproduction and death , while considering what demographics , and which locations , have a heightened risk of disease . | death rates, invertebrates, medicine and health sciences, ruminants, african trypanosomiasis, tropical diseases, vertebrates, parasitic diseases, animals, mammals, parasitic protozoans, glossina, simulation and modeling, seasons, developmental biology, protozoans, pupae, neglected tropical diseases, tsetse fly, population biology, insect vectors, research and analysis methods, infectious diseases, zoonoses, life cycles, protozoan infections, trypanosomiasis, insects, disease vectors, arthropoda, population metrics, trypanosoma, eukaryota, earth sciences, biology and life sciences, species interactions, cattle, amniotes, bovines, organisms | null |
journal.pcbi.1006202 | 2,018 | Real-time decision-making during emergency disease outbreaks | The responsibilities of policymakers during infectious disease outbreaks include the difficult task of choosing between multiple control interventions based on an uncertain future ., Mathematical and simulation models have proved a useful tool to aid decision-making during disease outbreaks by both generating forecasts of outbreak severity and comparing different control strategies 1–6 ., Using mathematical and simulation models , however , requires estimation of model parameters , and in the early stages of an outbreak , when decisions are most critical and data are most scarce , significant parametric uncertainty may limit the ability of models to provide informed advice ., Recent research has outlined frameworks that combine Bayesian parameter estimation and a mathematical model for generating real-time forecasts of outbreak severity throughout the course of an epidemic , aiming to assimilate available surveillance data into model estimates as rapidly as possible 7–10 ., The efficacy of these frameworks has typically been evaluated using the accuracy of forecasts ., For decision-makers , however , the accuracy of model forecasts per se , are not the best measure of success , it is the selection of the optimal control interventions in the face of uncertainty that counts , and ultimately what they are judged on ., However , projections of the impact of alternative control interventions are not always performed alongside projections of the burden of infection ., For several infectious diseases it is not possible , or relevant , to include projections of control interventions , for instance when interventions are directly related to estimates of the burden of disease or when control interventions are related to the timing of the peak of an epidemic ( such as in influenza ) ., Including projections of interventions does allow disease control problems to be phrased as optimization problems and therefore allows the determination of whether optimal control choice is dependent upon the underlying state of the outbreak ., In contrast to real-time forecasting approaches , we demonstrate real-time decision-making , which requires integration of all information until now plus the potential future impact of candidate control interventions ., In this work , we simulated the process of real-time decision-making by fitting a dynamic epidemic model to the observed ( confirmed ) , herd-level , infection data at weekly intervals throughout two historical outbreaks of foot-and-mouth disease ( FMD ) , a viral disease of economically important livestock , and compared forward simulations of the impact of alternative culling and vaccination interventions under a management objective of minimizing total culls so as to gain disease-freedom ., We repeated these forward simulations at each time point using parameter estimates from a model fitted to the complete outbreaks , thereby comparing the best we could have done at the time with the best we could have done in retrospect ., Forward simulations predicted the final total culls ( number of animals culled ) from having taken a single control intervention from a particular date into the future ., Control interventions included culling of infected premises only ( IP ) , culling of infected premises and dangerous contacts ( IPDC ) , culling of infected premises , dangerous contacts and contiguous premises ( IPDCCP ) , ring culling in areas surrounding infected premises at 3 and 10 km radii ( RC3 and RC10 respectively ) , and vaccination in areas surrounding infected premises at 3 and 10km radii ( V3 and V10 respectively ) ( see Materials and Methods for further details of the control interventions ) ., This set of intervention strategies includes those that governments have implemented , or considered , in the past and that are consistent with other studies on foot-and-mouth disease ., In contrast to previous work in real-time forecasting 7–10 , we used an individual-based model and included uncertainty regarding the location of infected and undetected farms 11 , 12 ., As surveillance data become available , we improve our understanding , in a cumulative manner , of both how the outbreak is unfolding , via estimates of transmission parameters , and of where likely new infections may be located , via estimates of the spatial distribution of infected but undetected farms ., Forward simulations are therefore conditional upon both the actual state of the outbreak ( i . e . the pattern of confirmed infected cases and the pattern of inferred , undetected infections ) and having enacted the actual outbreak controls until the time point in question ., The outbreaks of FMD in the UK in 2001 and Miyazaki , Japan in 2010 contrast in a number of ways and thus provide a valuable comparison for investigating how policy recommendations are affected by additional information throughout an outbreak ., Firstly , the UK outbreak had over 2000 infected premises over an area that included England , Wales , and Scotland ( roughly 230000km2 ) 13 whereas the outbreak in Miyazaki affected only 290 premises contained within the Miyazaki prefecture ( less than 8000km2 ) 14 , 15 ., Secondly , at the time of confirmation of FMD , there were multiple foci of infection dispersed throughout the UK but only one focus of infection in the Miyazaki outbreak ( although additional foci occurred later ) ., Finally , control interventions deployed during each outbreak were different; UK control interventions only included culling strategies whereas control of the outbreak in Miyazaki began with culling and shifted to vaccination after 5 weeks , necessitated by constraints on the disposal of accumulating carcasses 16 ., In the early stages of both the UK and Miyazaki outbreaks , estimates of the instantaneous risk of onward transmission are highly uncertain ( Figs 1 , S3 and S4 ) ., Our analysis highlights that uncertainty reduces through time in several regards ., Firstly , estimates of transmission parameters change through time ( Figs 1 , S3 and S4 ) ., Secondly , our understanding of how transmission parameters relate to one another change through time ( S5 and S6 Figs ) ., Finally , our confidence in the locations of premises we believe to be infected also changes through time ( Figs 1 , S7 and S8 ) ., Given that the relationships between parameters change through time , marginal distributions of parameters ( S3 Fig ) do not tell the whole story , and we therefore summarize how the shape of the multidimensional posterior distribution evolves through time via a measure of instantaneous risk of onwards transmission ., Note that the instantaneous risk of transmission indicates the overall relative risk of transmission , which does not have a direct epidemiological interpretation but provides a direct comparison across weeks ., Projections of the final total number of animals culled that were made in the first three weeks of each outbreak were highly uncertain , grossly overestimating the size of the outbreak in the UK and , on average , tending to underestimate the size of the outbreak in Miyazaki ( Figs 2 and 3 ) ., In the UK , the mean outbreak size for vaccination strategies was estimated to be ten times larger when using one week’s worth of data compared to when using all of the data ( Fig 2; accrued vs complete , week 1 ) ., In Miyazaki , bimodal distributions of outbreak size were generated by forward simulations started in the initial weeks as the spatial extent of the outbreak was highly uncertain ( Fig 3 ) ., This was particularly marked under strategies of non-ring control culling ( IP and IPDC ) ., However , in both outbreaks , despite projections being highly variable , the relative performance of control interventions was resolved early on ( Figs 2B and 3B ) ., The projected rankings of control interventions between those based on available data and those based on all outbreak data were identical at five weeks into the UK epidemic ( S9 and S10 Figs ) , and differences in the rankings of control interventions after week 5 were minor ., Using only available data , the top three ranked control strategies ( vaccination at 3km , 10km , and IPDCCP culling , respectively ) do not change from weeks 5–8 , at which point vaccinating at 10km becomes optimal over vaccinating at 3km in week 9 ., Vaccination interventions ( either at 3km or 10km ) are always ranked as optimal throughout the whole UK outbreak regardless of the week at which projections were made and regardless of the amount of data available ., If vaccination was deemed politically unpalatable then culling of infected premises , dangerous contacts , and contiguous premises ( IPDCCP ) was consistently ranked as the next best intervention to minimize total culls from weeks 4–10 of the outbreak in the UK ( regardless of whether available or all data were used to estimate transmission parameters ) ., Although the rankings of control interventions did change in the Miyazaki outbreak , the relative distribution of expected outbreak sizes across different control interventions remained consistent , with 3km and 10km vaccination strategies consistently ranked as optimal ( Figs 3 and S11 ) and consistently performing better than the other strategies when compared in repeated bootstrap simulations ( Figs 3C and S11C ) ., Ring culling at 3km was ranked as optimal for weeks 1–5 if vaccination interventions were not considered , regardless of the amount of data used to estimate transmission parameters ., IP culling was optimal in the final stages of both outbreaks when there were few or no more infected premises ( S1 , S2 and S10 Figs ) ., The consistency in the relative rankings of control interventions lends support for trusting model comparisons of interventions at outbreak onset ., For the UK outbreak on week 2 ( 5 March 2001 ) , when only using available data , projections of the magnitude of the four best performing control interventions occupied comparable overlapping ranges: V3 , V10 , IPDC and IPDCCP ( accrued information; Fig 2 ) ., Such a set of projections may provide a strong case in support of only enacting IPDC , the minimum under EU and Japanese law , as this avoids the latent underlying costs and consequences of culling contiguous premises or emergency vaccination , for only a small increase in the expected number of total animals culled ., However , had all the data for the outbreak been available at this date ( complete information; Fig 2 ) , then the expected performance of IPDC was revealed to be very poor compared to these other three control interventions , with the results supporting either vaccination ( V3 or V10 ) or culling in some proximity around infected farms ( RC3 or IPDCCP ) to quickly halt the outbreak ., Here , despite variability in the absolute value of projections , our initial rankings and estimates of the relative performance of controls were again found to be robust ., In the outbreak in Miyazaki the three best control interventions were correctly identified by week 2 using only the available data despite highly uncertain projections of total animals culled ( Figs 3 and S11 ) ., Relative performance of control interventions was largely unchanged after this point , with the exception of the emergence of IP culling as optimal in the final weeks due to there only having been 1 infected case on 4th July , before which there had not been a case since 18 June 2010 14 ., Consistency in the performance of vaccination strategies is highlighted by noting that greater than 50% of the time vaccination actions are optimal when bootstrap samples are compared across the distributions of simulated total culls for weeks 2–5 ( Figs 3C and S11C ) ., Bimodal distributions in outbreak size did reappear later in the outbreak between weeks 6–7 ( 1–8 June ) coincident with the occurrence of additional foci of infection ( S11 Fig ) , although these additional outbreaks were effectively controlled and policy recommendations were unchanged ., This is in contrast to the UK 2001 outbreak where , at the time of the first case being reported , there were already infected farms in multiple foci spread across the country with no significant new emerging foci as the epidemic progressed that would give rise to the bimodal predictions that we observe for Miyazaki ., Overall , outbreak size and optimal control choice for the Miyazaki outbreak were generally more straightforward to predict than for the UK outbreak , and policy recommendations for minimizing outbreak size are in line with what was actually implemented in 2010 ( vaccination at a 10km radius , V10 ) ., Changes in policy recommendations and large uncertainty in model projections may be caused by uncertainty associated with the parameters governing the dynamics of the outbreak and/or uncertainty associated with the locations of infected farms ., By looking at projections made using parameters generated with complete outbreak data , isolating the effect of changing the arrangement of infected farms ( since transmission parameters are fixed in these simulations ) , we see that policy recommendations are still changing ., Between 5 and 12 March 2001 in the UK outbreak ( weeks 2–3 ) , there was a marked resolution in the performance of control interventions when parameters were estimated using only available data ( accrued information ) ., The relative performance of control interventions at this time point comes to resemble analogous projections made using all outbreak data ( Figs 2 , S9 and S10 ) despite the absolute value of such projections still being highly uncertain ., Here , this switch may have been caused by a constrained estimate of where the outbreak is–an improved estimate of the location of undetected infected premises ( S7 Fig ) ., Our best estimate of the location of undetected infections on 5 March 2001 ( week, 2 ) includes undetected infections in the north of Scotland ., Additionally , the number of counties with a non-zero expected number of undetected infections is overestimated compared to analogous estimates if all the data were available ( S7 Fig ) ., Our best estimate of undetected infections on 12 March 2001 ( week, 3 ) had foci of infection in Cumbria , Midlands , Devon , North Wales , and Essex; these foci were in the same areas as estimates using all outbreak data ., Furthermore , those counties that have a non-zero expected number of undetected infections are contiguous to counties that are estimated as a foci of infection ( S7 Fig ) ., At this point , by excluding an allowance for additional culling in Scotland , ring culling at 3km became second only to vaccination actions ., Ultimately , the change in our understanding of the locations of infected farms ( and having the resource capability to respond ) , drove this improved prediction of the relative performance of control interventions ., Such a dramatic change in the relative performance of control interventions was not seen in the Miyazaki outbreak , despite starting with high variability in projections of the efficacy of control interventions ., The Miyazaki outbreak started and largely stayed in one location , the Tsuno township ., Though bimodal distributions in outbreak size did arise in simulations ( S11 Fig ) as additional foci of infection occurred later at the Ebino township , Miyakonojo , and Miyazaki city ( all over 40km away from the original point of infection 17 ) , the projections were largely seeded by an outbreak with a state characterized as a single point of infection , with compact and radial spread of infection ( Figs 3 , S8 and S11 ) ., Accounting for the possibility of a larger outbreak size in Miyazaki did not stretch the simulated culling or vaccination resources , leaving policy recommendations for how we should respond unchanged ., As with results from the UK outbreak , uncertainty associated with the locations of undetected infected premises , in the form of additional foci of infection , seemed to have a large effect on forward projections in the Miyazaki outbreak ., Our results have shown that policy recommendations using real-time parameter estimates and forward simulations can be correct from an early stage in an outbreak despite highly uncertain projections of epidemic severity , in line with previous research on one-time decisions for the control of Ebola 18 ., Projections from mathematical and simulation models are useful for informing policy-making insofar as they help to identify what is the best course of action and we have included two summaries of how simulation results may be communicated , dependent upon whether policymakers have an objective of, 1 ) optimizing an expectation in outbreak severity ( Figs 2B and 3B ) or, 2 ) optimizing the number of times that an intervention was predicted to be optimal ( Figs 2C and 3C ) ., In many situations , estimating the burden of infection and selection of the optimal control intervention are directly related ., However , our results are a reminder that this may not always be the case; in many cases it is how projections influence change in the recommended control intervention , rather than the value of projections themselves , that matter and so , where appropriate , it is important to include projections of intervention efficacy when making real-time projections of disease burden ., Our results showed that changes in the estimated relative performance of control interventions was strongly influenced by the spatial distribution of both observed and undetected infectious premises and culled premises ( the state of the outbreak ) ., Although our analysis only looked at one change in control intervention ( from what was being applied historically ) this dependence highlights that , should one allow multiple changes of control through time , then the optimal policy is likely to be dynamic ., In our results , even if accurate parameter estimates had been available from the first confirmed case ( such as those derived from the whole outbreak ) , it would still have been necessary to perform state-dependent control to find the best choice of control intervention as the outbreaks evolve ., In practice , estimating the state of the outbreak ( i . e . prediction of locations of undetected infections ) depends on estimates of transmission parameters and , therefore , continual re-evaluation of control interventions alone makes little sense without also re-evaluating , and re-fitting , model parameters ., Thus , control recommendations should adapt to the changing state of an outbreak ., Our results included a marked stabilization in recommended control interventions in simulations for the outbreak in the UK ., It remains an avenue for future research to investigate what are the drivers of such stabilization in recommendations in control intervention , and whether it is logistically feasible to measure ( and therefore act upon ) such drivers , beyond an in silico experiment ., Such analyses would require a more general approach to seeding outbreaks , and generating landscapes , to more thoroughly explore the state-space of potential future outbreaks ., Here , we only investigated two outbreaks , simulations for each of which were seeded from historical starting points ., Often , the need for rapid response during an outbreak is discussed in terms of the initial response 13 , but not as often in terms of adapting as circumstances change throughout an outbreak ., In both outbreaks we investigated , there was significant variability in epidemiological predictions during the early stages of disease outbreaks highlighting that real-time updating of parameters is vital in order to obtain accurate predictions of epidemic size and extent ., However , additional foci of infection may be seen as independent outbreaks and , therefore , the rapid response that is called for at the start of an outbreak needs to be reapplied when such foci are discovered ., For the outbreaks that we investigated , identifying the locations of undetected infected premises was of greater importance to determining the best course of action than identifying the underlying disease dynamics; this calls for increased vigilance in surveillance during an outbreak and highlights the importance of methodology for predicting undetected infections ( e . g . 11 , 12 ) , assimilating surveillance data into parameter estimates for individual-based models , and of confirming negative cases , especially when such premises are in locations that would be considered as additional foci of infection ., Simply because an outbreak has progressed beyond its initial stages does not mean that the need to act swiftly according to changes in the outbreak are in any way diminished ., At a coarse level , for one-time decisions , the idea of state-dependent control has already been adopted or discussed by several agencies , such as the Department for Environment , Food , and Rural Affairs , UK ( DEFRA ) and the United States Department of Agriculture ( USDA ) in flowcharts for determining when to perform emergency vaccination 19 , 20 , the dependence of different phases in smallpox eradication upon smallpox prevalence 21 , management of wildlife diseases 22 , and the use of adaptive surveillance of herds in the eradication of rinderpest 23 ., Decision-making frameworks , such as adaptive management , have also highlighted the utility of modeling and optimization as tools for generating state-dependent policies 24 ., We note that optimization methods such as dynamic programming or reinforcement learning would be required to generate state-dependent policies and that such methodologies , although a more complex optimization than what we have presented , would allow the possibility of changing control intervention repeatedly through time , or generating optimal interventions that are a combination of those defined here , such as county-specific interventions ( e . g . 25 ) ., Our analysis shows there is potential to make large gains in the effectiveness of the response to an outbreak by adopting state-dependent control that is on a much more nuanced scale than one-time binary decisions , and that mathematical and simulation models can play a significant role in policy preparedness by investigating such strategies before an outbreak occurs ., Both outbreaks used a similar model structure ., The infectious pressure , λj ( t ) , on a susceptible farm j at time t is, λj ( t ) =ϵ ( t ) +∑i∈I ( t ) βijh ( Ij−Ii ) +∑i∈N ( t ) βij*h ( Ij−Ii ) ,, where I ( t ) and N ( t ) are the sets of infected and notified farms at time t respectively ( t = 0 , 1 , 2 , … T ) ., Models of similar structure have been previously used for modeling FMD ( e . g . 1 , 4 , 11 ) ., A summary of notation used to describe the model is given in Table 2 ., Notification time is assumed to mean the time of laboratory confirmation of FMD , and removal time is assumed to mean the date when the farm is both culled and disposed of ., We count t in days; day 0 is the first confirmed infected case in the data and T is the day of the final parameter estimate ( 24 December 2001 for the UK outbreak and 6 July 2010 for the outbreak in Miyazaki ) ., By including both S ( t ) and R ( t ) , as the sets of susceptible and removed farms at time t respectively , we then have four sets essentially giving the state of the epidemic on the population at a given point in time , St=⟨S ( t ) , I ( t ) , N ( t ) , R ( t ) ⟩ ., The infection time of the kth farm is denoted Ik , while Nk and Rk denote notification and removal times of farm k respectively ., Note that in the next section we assume that for It , Nt , and Rt , the t subscript denotes the set of infection , notification , or removal times of all premises respectively; here the time subscript is dropped for succinctness ., All times are listed in days unless otherwise specified ., Infectious pressure can be decomposed into contributions from infectious ( but not yet notified ) farms ,, βij=γ1q ( i;ξ ) w ( j;ζ ) δ ( δ2+ρij2 ) ω , i∈I ( t ) , j∈S ( t ) ,, ( 1 ), and contributions from notified farms βij*=γ2βij ( i∈N ( t ) , j∈S ( t ) ) ., We define, q ( i;ξ ) = ( cic¯ ) ψ1+ξ2 ( pip¯ ) ψ2+ξ3 ( sis¯ ) ψ3, and, w ( j;ζ ) = ( cjc¯ ) ϕ1+ζ2 ( pjp¯ ) ϕ2+ζ3 ( sjs¯ ) ϕ3, to be the ‘infectivity’ of farm i and ‘susceptibility’ of farm j respectively , where ck , pk , and sk are the numbers of cattle , pigs , and sheep on farm k , with mean numbers of cattle , pigs , and sheep , denoted by c¯ , p¯ , s¯ ., The effect of distance is captured using a Cauchy-type kernel where ρij is the Euclidean distance between farms i and j , and δ the decay of transmission rate with distance ., Baseline infectious pressure distinguishes between periods before and after the movement ban was implemented ,, ϵ ( t ) ={ϵ1t<movementbanϵ1ϵ2otherwise ,, and the latency of the disease was modeled using the function, h ( t ) ={1t<40otherwise ., An estimate at time t of the complete parameter set , of 16 parameters ( 12 parameters in Miyazaki ) , is denoted θt ., Parameter ω was assumed fixed at 1 . 3 in both outbreaks , as were the initial priors ., The priors for each parameter were elicited via expert opinion and assigned the following distributions: π0 ( δ ) ∼Gamma ( 4 , 8 ) , π0 ( ϵ1 ) , π0 ( ϵ2 ) ∼Gamma ( 1E−7 , 1 ) , π0 ( ψ1 ) , π0 ( ψ2 ) , π0 ( ψ3 ) , π0 ( ϕ1 ) , π0 ( ϕ2 ) , π0 ( ϕ3 ) , π0 ( ξ2 ) , π0 ( ξ3 ) , π0 ( ζ2 ) , π0 ( ζ3 ) ∼Gamma ( 0 . 5 , 1 ) , π0 ( γ1 ) , π0 ( γ2 ) ∼Gamma ( 1E−4 , 1 ) ., Gamma ( a , b ) is the gamma distribution with shape parameter a and rate parameter b ., Infection to notification time was assumed to be distributed according to a Gamma distribution such that Ni → Ii = di∼Gamma ( 4 , b ) with parameter b = 0 . 5 governing the scale of the probability distribution ., Notification to removal time is an observed quantity since both events are recorded ., Let the data ( demographics and event history ) observed up to time t be Xt− , with πt ( θt , It|Xt− ) denoting the joint posterior distribution of the model parameters and infection times ( including infection times of undetected infections ) immediately at time t ., We assume that infections occurred according to a continuous-time non-homogeneous Poisson process , where infection rate is assumed to be λj ( t ) as above ., Let I , N , and R be vectors of infection , notification , and removal times for individuals 1 , … , nI who were infected by observation time Tobs ., Conditioning on this event time , the joint posterior distribution over parameters θ = {ϵ1 , ϵ2 , γ1 , γ2 , ξ , ζ , ψ , ϕ , δ , b} is, π ( θ|I , N , R ) ∝∏j=1nIλj ( Ij− ) exp∫IκTobs∑j∈Pλj ( t ) dt, ×∏j=1nI ( Nj−Ij ) a−1e−b ( Nj−Ij ), ×∏p=1|θ|fθp ( θp ) ,, where the first line represents the infection process , the second line represents the detection ( infected to notified ) process , and the third line represents independent prior distributions for all components of θ ., Ij− represents the time immediately before the infection time of the jth premises ., Parameters in bold , represent the corresponding set of species-specific parameters , e . g . ζ = {ζ2 , ζ3} ., κ denotes the initial infective , and P denotes the set representing all individuals in the population ., The joint posterior distributions used in the forward simulations and figures of the instantaneous risk of onward spread represent the MCMC output sub-sampled to 2000 parameter coordinates ., Multisite adaptive Metropolis-Hastings was used to draw from the conditional posterior distributions of {ϵ1 , ϵ2 , γ1 , , γ2 , δ} , ψ and ϕ ., Furthermore , since the infection times are unobserved , we updated them component-wise using Metropolis-Hastings , with a reversible-jump update to explore the posterior over the presence of undetected infections ., See 11 , 12 for further details ., The introduction of the 3-species model ( in comparison to 11 and 1 ) results in a non-linearly correlated posterior distribution , particularly between γ1 and both of ξ and ζ ., This presents particular difficulties for Metropolis-based MCMC algorithms , since the optimal proposal distribution scale changes with location in the posterior parameters space ., To approximately orthogonalize the posterior distribution , and hence improve the convergence properties of our adaptive Metropolis algorithm , we employed the following non-centered update for ξ and ζ ., To construct an efficient proposal distribution for ξ , we seek to exploit the shape of the posterior distribution with respect to γ1 ., This can be thought of as a joint update of γ1 and ξ , respecting the contour of the joint posterior ., First , we write the equation for q ( i;ξ ) ( see above ) as, γ1q ( i;ξ ) =γ1Ci+γ1ξ2Pi+γ1ξ3Si, =A{i , 1}+A{i , 2}+A{i , 3}, where Ci= ( ci/c¯ ) ψ1 and similarly for Pi and Si ., We then let, R=A1+A2+A3=∑i=1nI ( A{i , 1}+A{i , 2}+A{i , 3} ), We then propose, ( A2*+A3* ) T∼MVN2 ( ( A2+A3 ) T , 2 . 3822B ), and, A1*=R−A2*−A3*, where , with probability 0 . 05 , B = I2 the 2x2 identity matrix , and with probability 0 . 95 , B = Σk the empirical covariance matrix of the MCMC samples for ( A1 , A2 ) T up to iteration k ( see 30 ) ., We then solve for γ1 and ξ ., Similarly , we update γ1 and ζ ., Uncertainty in the joint posterior distribution of transmission parameters was summarized using a measure of the instantaneous risk of onward spread , defined as the instantaneous force of infection from an average-sized infectious farm to an average-sized susceptible farm at time t ( gt ) ., This risk measure was calculated at each time point as the integral over the joint posterior distribution of parameters at that point in time and over a distance of 20km ( from the infectious to the susceptible farm ) :, gt=∫θ∫020γ1ξ2 , t+ξ3 , tζ2 , t+ζ3 , tδt ( δt2+ρ2 ) ωdρdθ ., The equation for the instantaneous risk of onward spread is taken from the equation for the infectious pressure ( Eq 1 ) substituting the average number of cattle , pigs , and sheep for ci , pi , and si respectively ., Each control intervention , a , is evaluated according to the expected total number of animals culled , U ( Yt|a ) , where the Bayesian predictive distribution of the ongoing epidemic , Yt , is given by the integral, fYt ( Yt|Xt− , a ) =∫Θ∫IfYt ( Yt|Xt− , θt , It , a ) πt ( θt , It|Xt− ) dIdθ ., This was estimated by simulating forward from fYt ( Yt|Xt− , θt , It , a ) using draws from πt ( θt , It|Xt− ) ., The above formulation is using parameter estimates from the time point in question , so-called ‘accrued information’ ., Forward simulations were also generated with parameters estimated using all data , fYt ( Yt|XT− , a ) , so-called ‘complete information’ , which were generated in an analogous fashion instead using draws from πT ( θt , It|XT− ) in the above integral ., The optimal control strategy at⋆ is chosen as that which minimizes the mean expected total animals culled, at⋆=argminameanU ( Yt|a ) , Vaccine efficacy was assumed to be 90% , whereby animal numbers on a vaccinated farm were reduced by 90% after a delay until conferment of immunity of 4 days ., Delay from infection to notification time was simulated as 9 days , with a 4-day latency period ., Notification to removal time was simulated as 1 day for infected premises and 2 days for culling of dangerous contacts or contiguous premises ., The MCMC algorithm above was implemented in C++ ( GCC version 4 . 8 . 3 ) embedded within an R package ., General-purpose graphics processing unit ( GPU ) using NVIDIA CUDA 7 . 5 was used to implement the calculation of the likelihood and speed up inference ., The software is available under the GPLv3 license at http://fhm-chicas-code . lancs . ac . uk/InFER/InFER/tags/InFERfmd-v1 . 0 . | Introduction, Results, Discussion, Materials and methods | In the event of a new infectious disease outbreak , mathematical and simulation models are commonly used to inform policy by evaluating which control strategies will minimize the impact of the epidemic ., In the early stages of such outbreaks , substantial parameter uncertainty may limit the ability of models to provide accurate predictions , and policymakers do not have the luxury of waiting for data to alleviate this state of uncertainty ., For policymakers , however , it is the selection of the optimal control intervention in the face of uncertainty , rather than accuracy of model predictions , that is the measure of success that counts ., We simulate the process of real-time decision-making by fitting an epidemic model to observed , spatially-explicit , infection data at weekly intervals throughout two historical outbreaks of foot-and-mouth disease , UK in 2001 and Miyazaki , Japan in 2010 , and compare forward simulations of the impact of switching to an alternative control intervention at the time point in question ., These are compared to policy recommendations generated in hindsight using data from the entire outbreak , thereby comparing the best we could have done at the time with the best we could have done in retrospect ., Our results show that the control policy that would have been chosen using all the data is also identified from an early stage in an outbreak using only the available data , despite high variability in projections of epidemic size ., Critically , we find that it is an improved understanding of the locations of infected farms , rather than improved estimates of transmission parameters , that drives improved prediction of the relative performance of control interventions ., However , the ability to estimate undetected infectious premises is a function of uncertainty in the transmission parameters ., Here , we demonstrate the need for both real-time model fitting and generating projections to evaluate alternative control interventions throughout an outbreak ., Our results highlight the use of using models at outbreak onset to inform policy and the importance of state-dependent interventions that adapt in response to additional information throughout an outbreak . | Mathematical and simulation models may be used to inform policy in the early stages of an infectious disease outbreak by evaluating which control strategies will minimize the impact of the epidemic ., In these early stages , significant uncertainty can limit the ability of models to provide accurate predictions , and policymakers do not have the luxury of waiting for data to alleviate this state of uncertainty ., For policymakers , however , what is most important is the selection of the optimal control intervention , rather than accuracy of model predictions ., We fit an epidemic model to observed , spatially-explicit , infection data at weekly intervals throughout two historical outbreaks of foot-and-mouth disease , UK in 2001 and Miyazaki , Japan in 2010 , and compare forward simulations of the impact of alternative control interventions ., These are compared to policy recommendations generated in hindsight using data from the entire outbreak ., Our results show that the optimal control policy is identified accurately from an early stage in an outbreak , despite high levels of uncertainty in projections of epidemic size , and that the relative performance of control strategies is strongly mediated by our understanding of the locations of infected farms , rather than improved estimates of transmission parameters . | livestock, medicine and health sciences, animal diseases, foot and mouth disease, ruminants, immunology, vertebrates, animals, mammals, simulation and modeling, preventive medicine, infectious disease control, vaccination and immunization, zoology, research and analysis methods, public and occupational health, infectious diseases, swine, sheep, epidemiology, agriculture, eukaryota, biology and life sciences, amniotes, organisms | null |
journal.pcbi.1007073 | 2,019 | A new framework for assessing subject-specific whole brain circulation and perfusion using MRI-based measurements and a multi-scale continuous flow model | Applying traditional compartment models to in vivo hemodynamic measurements provides clinically valuable parameters in a wide range of medical conditions , e . g . Alzheimer disease 1 , stroke 2 , 3 or cancer 4 , 5 ., In these approaches , a pharmacokinetic compartment model is fitted to tracer evolution time curves from perfusion acquisitions to extract estimates of physiological parameters ., The methodology is applied to entire organs , or regional- or voxel-wise , depending on the application ., In the case of tracer-based measurements of brain hemodynamics , cerebral blood flow ( CBF , or perfusion ) , cerebral blood volume ( CBV ) , and mean transit time ( MTT ) are commonly extracted parameters from one-compartment ( 1C ) models ., However , a fundamental drawback of these methods has been previously pointed out: determining CBF from traditional 1C models is scale dependent , hence the results depend on discretization level 6–8 ., In 9 it was demonstrated that 1C models are prone to substantial errors when applied to smaller computational units connected in space instead of entire organs ., This implies that measurements of perfusion on different discretization scales can provide considerably varying results depending on the choice of imaging device and post-processing software ., The major reason for scale dependency of traditional 1C models is the lack of spatial connectivity in the model , hence allowing for repeated counts of the same fluid volume when applying the model to spatially connected units ., Recognition of the deficiencies of traditional compartment models has led to the development of multiscale , continuous blood flow models ., Such models are highly relevant for improved understanding of the conditions affecting both global blood flow and microrheology in disease states , such as , e . g . cerebral aneurysms and sickle cell anemia 10 ., Treatment of patients , such as by means of neurosurgery , may also benefit from individualized models that describe complex geometrical phenotypes 11 ., Multiscale blood flow models may also contribute to a better understanding of angiogenesis and interstitial flow in simulated tumor microvascular networks , thus providing a more comprehensive and descriptive model for drug delivery 12–15 ., A challenging topic within multiscale flow models is the precise mathematical formulation of perfusion within a continuous flow model ., A model of perfusion should be in accordance with the physiological interpretation of perfusion being considered as a feeding arterial flow of oxygenated blood into the tissue or an organ ., As a solution , we adopt a continuous flow model in which perfusion is regarded as the volume flux of oxygenated blood , which transits from arterial to the venous side in a two-compartment ( 2C ) model 16–19 ., This understanding of perfusion is both mathematically strict and physiologically sound ., The vascular system is a geometrically highly complex tubular network connecting vessels at different spatial scales ., One particular challenge of whole brain simulation of perfusion is how to connect flow on the various scales , ranging from the carotid artery lumen with diameter close to 6 mm 20 down to capillaries with diameters of approximately 6 μm 21 ., A suitable continuum model for flow simulations is expected to care for both tissue inhomogeneity and anisotropy , with the inconvenience of requiring a large number of unknown modelling parameters ., One common approach is therefore to represent the vessels as an inexpensive 1D flow model coupled with a 3D continuum model for the brain tissue 19 , 22–24 ., The simple geometry of a 1D model reduces the number of required modelling parameters , while the vascular geometry can be observed in dedicated MR acquisitions ., However , well-posedness and stability of the solution at the interface between the 1D and 3D model is challenging 22 , 23 ., In the current work , we address this problem by introducing a local flow distribution region where interface conditions are governed by a mass conserving , smooth support function , hence ensuring stability of the system ., The purpose of the numerical simulations is twofold ., First , by examples to illustrate working principles of our algorithm , and secondly , to demonstrate scale invariance ., With this in mind , we have chosen the geometry of a frog tongue as example data 27 ., This data set exhibits a realistic vascular geometry ., Practically , the data set was scanned from a written source 27 ., Preprocessing steps included semi-automatic segmentation of each of the vascular networks ., Length of the tongue was measured to be 35 mm , and field of view ( FOV ) was set accordingly ., As an approximation we consider the data to be almost two-dimensional ., The tongue is also stretched between the nails pinning it to the surface , leading to a deformed geometry compared to a unprepared frog tongue ., A visualization of the arterial and venous network , as well as the tongue tissue is shown in Fig 2 ., Input data can be found in Supporting Information S1–S3 Data ., Furthermore , we used a full human brain for simulation of flow ., Acquisition parameters and postprocessing steps of these data are described in the following ., In the coupled model combining 1D flow in larger vessels with 3D Darcy flow in the brain , the majority of tissue is modelled as a porous medium where pressure driven flow is restricted by fluid mass balance and generic assumptions about the vascular microstructure of the arterioles , venules and the capillary system ., In order to describe perfusion mathematically , we work under the assumption of two parallel 3D systems ( or compartments ) , one accounting for arterial and one accounting for venous flow ., The perfusion is interpreted as the delivery of oxygenated blood from the first to the latter compartment ., Further details regarding the 3D model are given below ., For simplicity , we currently omit the index β = {a , v} and consider either the arterial or the venous vascular network consisting of nodes Ni and edges Ejk ., Each node Ni has an associated position x ˜ i and pressure p ˜ i Pa ., An edge Ejk is the connection between the pair of nodes ( Nj , Nk ) , and is associated with a tubular length Ljk m and a constant tubular radius Rjk m ., The length Ljk is a geodesic distance measured along the tubular medial axis , and Rjk is computed as the average tubular radius along the structure Ejk ., Each edge Ejk mediates an absolute flow q ˜ j k m3 s−1 from node Nj to node Nk ., Algorithmically , the network is represented as a connectivity matrix of an undirected graph ., A schematic illustration of a vascular network with proper notation is shown in Fig 4 ., Each node Ni is connected to a set of neighboring nodes N ( N i ) ., A terminal node is defined as any node with only one neighbour , i . e . N T ≔ { i : | N ( N i ) | = 1 } for the cardinality |⋅| , while interior nodes are nodes with more than one neighbour , N I ≔ { i : | N ( N i ) | > 1 } ., We further split the terminal nodes into root terminals and interior terminals ., Root terminals NR are pressurized terminal nodes with imposed Dirichlet boundary conditions ., Algorithmically , the root terminals are the intersection points of the large vessels with the brain-background interphase ∂ Ω B , N R = { i : x ˜ i ∈ ∂ Ω B } ., Interior terminals are terminal nodes placed within the domain , mediating flow from/to the vascular tree into the 3D domain ., The flow in the interior terminals is a Neumann boundary condition for the microvascular flow in the 3D domain ., For the remaining , we refer to interior terminals as terminals ., Finally , the set of all nodes is the union of interior nodes , terminals and root terminals , NI ∪ NT ∪ NR ., For modelling of flow through larger vessels we approximate the vessels as straight tubes of constant , circular cross-sections ., We also assume laminar flow of an incompressible , Newtonian fluid ., The assumption of laminar flow is supported by Reynold numbers < 200 for any of the middle cerebral arteries and penetrating arterioles 40 ., Under these assumptions , the vascular flow in larger vessels can be modelled using the Hagen-Poiseuille equation , relating a pressure drop Δ p ˜ j k ≔ p ˜ j - p ˜ k of an incompressible fluid of viscosity μ Pa ⋅ s through Ejk to the flow q ˜ j k between two nodes Nj and Nk 41 ,, Δ p ˜ j k = 8 μ L j k q ˜ j k π R j k 4 ., ( 9 ), Fluid mass balance must be ensured for each interior node, ∑ j ∈ N ( N k ) q ˜ j k = 0 , k ∈ N I ., ( 10 ), Denoting κ j k ≔ π R j k 4 / 8 μ L j k , and combining ( 9 ) with ( 10 ) yields the relation, ∑ j ∈ N ( N k ) κ j k ( p ˜ j - p ˜ k ) = 0 , k ∈ N I ( 11 ), for the fluid mass balance of each interior node ., Within the arterial and networks we assume full porosity , hence collapsing into one-compartment flow ., The root terminals represent the outer boundaries of the complete flow system , and each of the root terminals are therefore assigned a user-provided Dirichlet boundary condition p ˜ 0 p ˜ k = p ˜ 0 ∀ k ∈ N R ., ( 12 ), From the definition of a terminal node , terminal nodes have only one neighbour , i . e . only one edge connection , and the flow in each of the terminal nodes can be expressed as, q k ˜ = κ j k ( p ˜ j - p ˜ k ) , j ∈ N ( N k ) ∀ k ∈ N I ( 13 ), providing Neumann boundary conditions of flow continuity between the arterial/venous network and the brain ., This topic is further addressed in the next section ., The intersection point connecting the flow in the 1D tubular network with the flow in the generic 3D tissue yields a singularity both in terms of physics and numerics ., Physically , there is a modelling gap between the explicit , macro scale representation of arterial/veinal flow and the micro circulation in the 3D domain ., In our model , we fill this gap using a volumetric source term Qϵ m3 s−1 m−3 ., For ϵ = 0 this is essentially a Dirac measure, Q 0 ( x ) = ∫ Ω Q ( y ) δ ( x - y ) d y ( 14 ), where all fluid is distributed within an infinitely small point ., However , assume that the blood from the terminals is distributed along a fine scale network which is not visible at the imaging resolution ., The idea is to replace the Dirac measure , which is not physically sound , with a more realistic model of the source region with a characteristic radius ϵ ., To this end , define the support function, η ϵ ( x ) ≔ 1 ϵ n η ( x ϵ ) ( 15 ), where the shape function η is any positive and continuously differentiable function satisfying ∫ Ω B η d x = 1 ., In principle , the shape function η should reflect the structure of the sub-resolution arterial and venous trees , but due to the lack of such data we adopt the generic choice, η ( x ) ≔ { C exp ( 1 | x | 2 - 1 ) if | x < 1 , 0 if | x | ≥ 1 ., ( 16 ), An appropriate expression for the source terms then becomes, Q ϵ ( x ) = ∫ Ω Q ( y ) η ϵ ( x - y ) d y ., ( 17 ), Due to the properties that both the Dirac delta distribution and the source distribution integrate to unity , we note that the total integral over ( 14 ) and ( 17 ) remains the same and global mass balance is ensured from the terminals ., Moreover , ( 17 ) converges to ( 14 ) in the case where ϵ → 0 , justifying the notation Q0 in ( 14 ) ., For the remaining , adopt the notation of β indicating characteristics of the arterial ( β = a ) or venous ( β = v ) tree , e . g . single nodes Nk → Nβ , k or the set of nodes N j → N β j , j ∈ { I , T , R } ., Considering the volumetric source/sink terms Q β , k ϵ in ( 8 ) , the total flow contribution ( 17 ) from terminal k can be approximated as, Q β , k ϵ ( x ) = Q β , k η β , k ϵ ( x - x ˜ k ) | N β , k | ., ( 18 ), where |Nβ , k| is the node volume ., The volumetric flow Qβ , k through terminal k relates to the absolute flow q ˜ β , k by, Q β , k = q ˜ β , k / | N β , k | , ( 19 ), thus providing the relation, Q β , k ϵ ( x ) = q ˜ β , k η β , k ϵ ( x - x k ) ., ( 20 ), In the flow-distribution region around each interior terminal |x − xk| < ϵ we require that the pressure drop between a terminal node Nβ , k and the surrounding brain tissue satisfying |x − xk| < ϵ scales with the terminal flow up to a user-provided constant γβ q ˜ β , k = γ β μ ( p ˜ β , k - ∫ Ω η β , k ϵ ( x - x k ) p β d x ) ∀ k ∈ N β I ., ( 21 ), The coefficient γβ has the interpretation of an effective resistance in the unresolved network extending from the terminal node ., A higher value of γβ will enforce a lower pressure drop between the vascular tree and the microvascular model , and vice versa ., This closes the coupled modelling system , where the complete flow model is formed from ( 7 ) , ( 8 ) , ( 11 ) , ( 12 ) , ( 13 ) , ( 20 ) , ( 21 ) ., For the real brain application , voxelwise maps of the perfusion proportionality factor α ( 7 ) were generated with higher values in grey matter than in white matter 42 , ensuring a physiologically reasonable map ., We illustrate our approach by applying the method also to simulate flow in a frog tongue ., In this specific example there is no grey or white matter , and we use α constant everywhere ., The equations in the preceding sections describe blood propagation from the arterial to the venous side of the brain vascularity ., In order to simulate a real contrast enhanced MR acquisition we also introduce a model for transport of a tracer in the bloodstream ., All quantities are assumed to be in SI units , and later converted to more appropriate units for presentation whenever needed ., Observable or volumetric tracer concentration C ( x , t ) mol/m3 is a linear function of the fractional volumetric tracer concentrations Cβ ( x , t ) for each of the compartments, C ( x , t ) = C a ( x , t ) + C v ( x , t ) ., ( 22 ), Furthermore , tracer distribution volume is different from a geometric volume whenever ϕβ < 1 , leading to the relation, C β ( x , t ) = ϕ β ( x ) c β , b ( x , t ) ( 23 ), connecting blood tracer concentration cβ , b ( x , t ) mol/m3 to volumetric tracer concentration Cβ ( x , t ) ., The following criteria are assumed to hold: The tracer is homogeneously distributed in the fluid within a small distribution volume Ωi ( i . e . a voxel or a node ) , all physiological and structural parameters are stationary within the time of acquisition , and tracer transport by diffusion is not considered ., Under these assumptions , the influx of tracer into Ωi is determined by the product of fluid tracer concentration cβ ( x , t ) and flux uβ, ( x ), - ∫ ∂ Ω i c β ( u β · n ) d A ( 24 ), where n is the outward pointing surface normal of Ωi ., The rate of change of tracer within the control volume yields, d d t ∫ Ω i C β ( x , t ) d x ., ( 25 ), In addition , one must account for volumetric source terms ., Combining ( 24 ) with ( 25 ) and allowing for perfusion and inflow and outflow of tracer from/to all interior terminals , an upstream finite volume model for tracer mass balance can be phrased as, ∫ Ω i ϕ a ∂ c a ∂ t d x = - ∫ ∂ Ω i c a ( u a · n ) d A - ∫ Ω i c ″ P d x + ∑ k ∈ N a I ∫ Ω i c a ′ q ˜ a , k η a , k ϵ ( x - x k ) d x x ∈ Ω B for c a ′ = { c ˜ a , k ( t ) if q ˜ a , k ≥ 0 ( Node is upstream ) c a ( x , t ) if q ˜ a , k < 0 ( Brain tissue is upstream ) ∫ Ω i ϕ v ∂ c v ∂ t d x = - ∫ ∂ Ω i c v ( u v · n ) d A + ∫ Ω i c ″ P d x + ∑ k ∈ N v I ∫ Ω i c v ′ q ˜ v , k η v , k ϵ ( x - x k ) d x x ∈ Ω B for c v ′ = { c v ( x , t ) if q ˜ v , k < 0 ( Brain tissue is upstream ) c ˜ v , k ( t ) if q ˜ v , k ≥ 0 ( Node is upstream ) and where c ″ = { c a ( x , t ) if p a ≥ p v c v ( x , t ) if p a < p v ., ( 26 ), For the edges Eβ , jk we construct a finer discretization in order to facilitate graded tracer concentration along the edges ., Hence , split each edge Eβ , jk into nβ , jk subsegments associated with medial axis voxels , and assign every remaining voxel in the edge to the closest medial axis point , leading to disc-like discretization volumes Eβ , jk , i referring to subsegment i within edge Eβ , jk ., Also , assume the order of subsegments is downstream with increasing index i ., In particular , c β , j k , n β , j k refers to the tracer concentration in the last subsegment of edge Eβ , jk , which is identical to the first subsegment upstream of node k ., Similar equations as ( 26 ) apply to the nodes under the assumption of full porosity within the distributing node volume, ∫ N a , k ϕ ˜ a , k ∂ c ˜ a , k ∂ t d x = Δ f a , k Δ f a , k = { - ( c ˜ A I F - c ˜ a , k ) q ˜ a , k if k ∈ N a R ( root terminal , q ˜ a , k < 0 ) ∑ j ∈ N ( N a , k ) c ′ q ˜ a , j k if k ∈ N a I ( interior node ) Δ c q ˜ a , k if k ∈ N a T ( terminal ) c ′ = { c ˜ a , j k , n a , j k if q ˜ a , j k > 0 ( incoming fluid ) c ˜ a , k if q ˜ a , j k < 0 ( outgoing fluid ) Δ c = { c ˜ a , j k , n a , j k - c ˜ a , k if q a , k > 0 ( terminal is a source ) ∫ Ω c ˜ a , k η a , k ϵ ( x - x k ) d x - c ˜ a , k if q a , k < 0 ( terminal is a sink ) ∫ N v , k ϕ ˜ v , k ∂ c ˜ v , k ∂ t d x = Δ f v , k Δ f v , k = { ( c ˜ v , j k , n v , j k - c ˜ v , k ) q ˜ v , k if k ∈ ( N v I ∪ N v R ) ( root terminal , q ˜ a , k > 0 ) ∑ j ∈ N ( N v , k ) c ′ q ˜ v , j k if k ∈ N v I ( interior node ) Δ c q ˜ v , k if k ∈ N v T ( terminal ) c ′ = { c ˜ v , j k , n v , j k if q ˜ v , j k > 0 ( incoming fluid ) c ˜ v , k if q ˜ v , j k < 0 ( outgoing fluid ) Δ c = { c ˜ v , j k , n v , j k - c ˜ v , k if q v , k > 0 ( terminal is a source ) ∫ Ω c ˜ v , k η v , k ϵ ( x - x k ) d x - c ˜ v , k if q v , k < 0 ( terminal is a sink ) ( 27 ), where fβ , k is the incoming/outgoing tracer flux of node k ., Within the edges , tracer concentrations in each subsegment follows accordingly, ∫ E β , j k , i ϕ ˜ β , j k , i ∂ c ˜ β , j k , i ∂ t d x = ( c ˜ β , k - c ˜ β , j k , i ) q ˜ β , j k i = 1 ∫ E β , j k , i ϕ ˜ β , j k , i ∂ c ˜ β , j k , i ∂ t d x = ( c ˜ β , j k , i - 1 - c ˜ β , j k , i ) q ˜ β , j k i = 2 , … , n β , j k ( 28 ), for β = {a , v} ., Note that the first subsegment relates to the upstream node ., The hematocrit factor Hct connects blood tracer concentration cβ , b with plasma tracer concentration cβ according to, c β , b = c β ( 1 - H c t ) ., ( 29 ), Tracer can only distribute within the arterial and the venous compartment , and the observable tracer concentration becomes, C ( x , t ) = ( c a ( x , t ) ϕ a + c v ( x , t ) ϕ v ) ( 1 - H c t ) ., ( 30 ), when applying ( 22 ) , ( 23 ) , and ( 29 ) ., In the current model , the hematocrit is independent of vessel scale , and therefore only has the role as a global scaling factor of the tracer concentration curves ., Eqs ( 26 ) , ( 27 ) , ( 28 ) , and ( 30 ) form the model for indicator dilution ., Further details on the numerical implementation are shown in Supporting Information ., Integrating ( 8 ) over a control volume ( voxel ) Ωi ⊂ Ω and applying the divergence theorem yields, - ∫ ∂ Ω i ( λ a ∇ p a ) · n d A = ∑ k ∈ T a I ∫ Ω i Q a , k ϵ d x - ∫ Ω i P d x - ∫ ∂ Ω i ( λ v ∇ p v ) · n d A = ∑ k ∈ T v I ∫ Ω i Q v , k ϵ d x + ∫ Ω i P d x ( 31 ), for the conductivities λβ ≔ kβ/μ , β = {a , v} ., The elliptic term of Eq ( 8 ) was discretized using finite volume TPFA ( two-point flux approximation ) , leading to a linear relation in the transmissibilities tij and pressure difference pβ , i − pβ , j between a center voxel xi and an adjacent neighbor xj ., TPFA is widely applied in reservoir mechanics , and the reader is referred to 43 for more details ., Following TPFA , Eq ( 31 ) can be approximated as a linear system, ∑ j ∈ N i t i j ( p a , i - p a , j ) + α i ( p a , i - p v , i ) | Ω i | - ∑ k ∈ T a I q ˜ a , k η a , k ϵ ( x i - x k ) | Ω i | = 0 q ˜ a , k = κ a , j k ( p ˜ a , j - p ˜ a , k ) , j ∈ N ( N a , k ) ∀ k ∈ N a I Continuity in flow ∑ j ∈ N i t i j ( p v , i - p v , j ) - α i ( p a , i - p v , i ) | Ω i | - ∑ k ∈ T v I q ˜ v , k η v , k ϵ ( x i - x k ) | Ω i | = 0 q ˜ v , k = κ v , j k ( p ˜ v , j - p ˜ v , k ) , j ∈ N ( N v , k ) ∀ k ∈ N v I Continuity in flow ( 32 ), when also applying Eq ( 20 ) ., Further define the voxel neighborship around an interior node N v ( N β , k ) ≔ { j : η β , k ϵ ( x j - x k ) > 0 , k ∈ N β I } including all voxels close to terminal Nβ , k receiving a nonzero fluid distribution ., The network Eqs ( 11 ) and ( 12 ) are readily discretized , while the condition on pressure continuity Eq ( 21 ) becomes, q ˜ β , k = γ β μ ( p ˜ β , k - ∑ j ∈ N ( N β , k ) η β , k ϵ ( x j - x k ) p β , j | Ω j | ) ∀ k ∈ N β I ( 33 ), A linear system A x = d was created ,, A = ( A ( D a → D a ) A ( D a → D v ) A ( D v → D a ) A ( D v → D v ) A ( D a → N a ) 0 0 A ( D v → N v ) A ( N a → D a ) 0 0 A ( N v → D v ) A ( N a → N a ) 0 0 A ( N v → N v ) ) ( 34 ), where x is the concatenation of the voxelwise pressure values pβ , i and nodal pressure values p ˜ β , i ., The argument D refers to the Darcy equation in the continuum , and N refers to the nodes ., The subscript indicates arterial or venous compartment/tree ., The arrow indicates interactions , e . g . the submatrix A ( D a → N a ) contains the interaction between the arterial compartment and arterial-tree nodes ., Right hand side d depends on Dirichlet boundary conditions on the pressure ., The linear system of equations was solved using GMRES 44 with a tolerance of 10−6 , and a LUP decomposition for preconditioning ., A first approximation of the forward time step was initially computed from the largest possible time step satisfying the CFL conditions of the Darcy domain , the nodes , and the segments ., However , due to the implementation of a backward Euler solver , we were able to use significantly longer time steps , leading to a sequence of time points ti = iδt , i = {0 , 1 , … , n} where δt was ten times the maximum time step according to the CFL condition ., Total number of iterations became n = floor ( 120/δt ) , where 120 is maximum simulation time ., Forward simulation of tracer evolution was performed by creating a discrete linear system of equations from ( 26 ) , ( 27 ) , ( 28 ), c i + 1 = c i + δ t ( A c i + b i ) , i = { 2 , 3 , … , n } , c 0 = 0 ( 35 ), in the variable, c i = c D , a , c D , v , c N , a , c N , v , c E , a , c E , v i T ( 36 ), containing the concatenation of discrete variables of tracer concentration at time point ti in the Darcy domain cD , β , the nodes cN , β , and the edges cE , β , and where B is a block-diagonal matrix, ℬ = ( ℬ ( D a → D a ) ℬ ( N a → D a ) ℬ ( D v → D a ) ℬ ( D v → D v ) ℬ ( D a → N a ) 0 0 ℬ ( D v → N v ) 0 0 0 0 ℬ ( N a → D a ) 0 0 ℬ ( N v → D v ) ℬ ( N a → N a ) 0 0 ℬ ( N v → N v ) ℬ ( N a → E a ) 0 0 ℬ ( N v → E v ) 0 0 0 ℬ ( E a → N a ) 0 0 ℬ ( E v → N v ) ℬ ( E a → E a ) 0 0 ℬ ( E v → E v ) ) ( 37 ), with similar notation as ( 34 ) , in addition to E ( ⋅ ) referring to the edges ., The constant vector bi depends on AIF values cAIF , i ., A backward Euler updates the concentration at time point ti according to, ( I - δ t A ) c i + 1 = c i + δ t b i ( 38 ), where I is the identity matrix ., The matrix ( I - δ t B ) is fixed over iterations , and a GMRES solver was used as a solver with a LUP preconditioner with the previous iterate as initial guess in the consecutive time iteration ., As arterial input function we used a gamma-variate function, c ˜ A I F ( t ) = C 0 ( t - t 0 ) A e - ( t - t 0 ) / B ( 39 ), with constants A = 3 , B = 1 45 ., Tracer simulation time was 120 s , with a delay t0 = 7 . 5 min ., All program code was written in MATLAB ., We performed a numerical sensitivity analysis to examine how uncertanties in the input parameters are propagating through the model and affecting the output parameters ., For a function y = f, ( xi ) depending on a set of variables xi , i = {1 , 2 , …} , the relative sensitivity coefficient, c i * ≔ x i y ∂ y ∂ x i ( 40 ), is a measure of how the input parameter xi affects the outcome y ., The derivative was computed around an expected xi with a 1% variation on xi ., We report c i * for the perfusion parameter α , the fluid viscosity μ , as well as the arterial and venous components of the porosity ϕβ , the permeability kβ , and the pressure drop parameter γβ ., As investigated output parameters we used the arterial and venous pressures pa and pv , respectively , the perfusion ( P ) , time to peak ( TTP ) , and the mean transit time ( MTT ) ., Time to peak and mean transit time were computed according to standard definitions from tracer kinetic modelling ., Time to peak is the average time in seconds to maximum height of the contrast enhancement curves ., Mean transit time becomes CBV/CBF , which is equivalent to MTT = ( ϕa + ϕv ) /P in terms of our notation ., The current section accounts for simulation of circulation and perfusion in the frog tongue previously described ., The vascular networks were segmented in terms of a binary mask , nodes , edges , and medial axes ., In Fig 5 we have aligned these structures with the support function ηϵ ( x − xk ) ( 15 ) ., Simulation parameters used in the numerical simulations are shown in Table 1 ., Several of the parameters are not accurately known , and literature references were used to find appropriate estimates ., The parameters kβ and ϕβ are field parameters , but were held constant in space within each compartment ., The perfusion proportionality factor α was held constant everywhere within the frog tongue ., Obtained pressure maps of the arterial and venous compartments are depicted in Fig, 6 . Pressure conditions in the vessel network are completely determined by the node pressure , but for visualization purposes the pressure was approximated along the edges using linear interpolation between connecting nodes ., The obtained map of perfusion is shown in Fig, 7 . For the applied set of parameters an average perfusion of 65 ml/min/100ml was obtained , in the same order as human brain perfusion 50 ., Average fluid tracer concentration of the arterial input function , and the arterial and venous compartments are shown in Fig, 8 . Voxelwise , volumetric tracer concentration C mmol L−1 ( 30 ) as a function of time is shown in Fig, 9 . We used a time step Δt = 5 s between each time frame for plotting ., In order to demonstrate scale invariance of the algorithm , perfusion was recomputed within a smaller FOV with different resolutions represented by multiplicative resolution scales Si , i = {1 , 2 , 4 , 6 , 8 , 10 , 12 , 14 , 16} ., See the red rectangle in Fig 5 for the applied FOV with matrix size S1 = ( 100 × 100 ) ., The matrix size at scale i becomes Si = ( 100 × 100 ) i ., With except from the FOV , same parameter settings as reported in Table 1 were used for these simulations ., Average perfusion for each scale was computed within the frog tongue , and obtained values are shown in Fig, 10 . For all practical means , perfusion remains constant over the resolution scales ., The current section describes numerical simulation of circulation and perfusion in a complete human brain where the geometry , including grey and white matter , as well as the vascular networks were extracted from MRI data ., Simulation parameters are shown in Table 2 ., All figures show the same image slice ( no . 180 ) of the 3D image stack , with except from the 3D rendering in Fig 11 ., A volume rendering of the T1-weighted input data with superimposed arterial and venous masks is shown in Fig 11 52 ., Vascular permeability kβ and porosity ϕβ were assigned constant values within grey and white matter for each compartment , according to Table 2 ., The perfusion proportionality factor α ( x ) ( 7 ) is plotted in Fig 12 for one axial slice ., The grey matter value of α was set 1 . 6 times higher than the white matter value in order to resemble regional distribution of human brain perfusion 42 ., The piecewise constant parameter maps of kβ ( x ) , ϕβ ( x ) , and α ( x ) were smoothed using a Gaussian convolution with radius 2 . 5 mm and standard deviation 1 . 5 mm to impose smoothness in the white matter/grey matter boundary ., The Gaussian smoothing is an attempt to simulate partial volume effects in MR , where a voxel situated on the boundary between different tissue will possess properties reflecting both tissue types ., Calculated pressure maps of the arterial and venous compartments are shown in Fig, 13 . The voxelwise map of perfusion for a single axial slice is shown in Fig, 14 . Obtained values of average perfusion , arterial and venous pressure , time to peak ( TTP ) , and mean transit time ( MTT ) are reported in Table 3 for the entire brain , as well as for white , and grey matter ., The ratio of white matter perfusion to grey matter perfusion is 1 . 45 , not far away from the expected ratio of approximately 1 . 6 ., The total number of arterial and venous nodes found in the data set was 335 and 1222 , respectively ., Spatially averaged tracer concentration-time-curves are shown in Fig 15 for the arterial input function , as well as the arterial and venous compartments ., Run-time for the whole brain simulation was 2 . 5 d on a 32 multicore 2 . 29 GHz linux server with 355 Mb RAM without use of parallel computing environments ., The relative sensitivity coefficient c i * according to ( 40 ) was computed for each output variable for the frog tongue and the 3D human brain example ., Spatially averaged relative sensitivity coefficients of the human brain example are shown in Fig 16 ( left panel ) , where it is found that the perfusion proportionality parameter α has a strong positive relation to the perfusion P ( grid position ( 1 , 3 ) ) , and a negative relation to MTT ( grid position ( 1 , 5 ) ) ., Venous porosity ϕv is strongly positively correlated with MTT ( grid position ( 3 , 5 ) ) ., An example of a voxelwise map of c i * for the frog tongue is shown in Fig 16 ( right panel ) , demonstrating a local variability in the coefficients ., A proper mathematical model of circulation and perfusion is essential for simulating the pathway of blood , nutrients , oxygen , and drugs within the brain and other organs ., In this respect , a comprehensive simulation tool for entire organs should address certain requirements:, ( i ) It needs to be scale invariant ,, ( ii ) it should reflect a clinical understanding of perfusion , and, ( iii ) it should apply to an entire organ ., We claim that existing simulation tools are lacking at least one of these requirements ., Traditional pharmacokinetic compartment models are useful to describe perfusion within entire organs , but they are inaccurate for voxelwise descriptions 6–9 , hence , they are short of, ( i ) ., On the other hand , numerous simulation studies have described the intertwined processes of angiogenesis , drug delivery and interstitial flow in artificial tumor microvascular networks 12–15 ., However , these models are not explicitly expressing perfusion as a clinical parameter , and they have not been developed for the whole brain , hence lacking in, ( ii ) and, ( iii ) ., In this context , there is a need for precise mathematical models describing perfusion both locally and globally , also requested in 53 ., The current study is an attempt to bridge this gap ., Our main contribution is a comprehensive , data-driven and scale invariant model for whole brain circulation and perfusion using a mathematically strict definition of perfusion in line with its clinical understanding ., Thus , our method is simultaneously addressing, ( i ) -, ( iii ) , and it represents a new generation of simulation tools for predicting transport of nutrients , oxygen and drugs in live , human tissue ., Simulation of flow in live , human tissue is a challenging task where compartment models with suitable modifications can be transferred across organs and physiology 18 ., In this context , our model is generic and can be modified for description of other organs than the brain , as well as pathological conditions ., At present , our model assumes no leakage and an intact blood-brain barrier , and the focus for discussion is restricted to brain perfusion in the absence of pathological conditions ., In the current study we demonstrate scale in | Introduction, Methods, Results, Discussion | A large variety of severe medical conditions involve alterations in microvascular circulation ., Hence , measurements or simulation of circulation and perfusion has considerable clinical value and can be used for diagnostics , evaluation of treatment efficacy , and for surgical planning ., However , the accuracy of traditional tracer kinetic one-compartment models is limited due to scale dependency ., As a remedy , we propose a scale invariant mathematical framework for simulating whole brain perfusion ., The suggested framework is based on a segmentation of anatomical geometry down to imaging voxel resolution ., Large vessels in the arterial and venous network are identified from time-of-flight ( ToF ) and quantitative susceptibility mapping ( QSM ) ., Macro-scale flow in the large-vessel-network is accurately modelled using the Hagen-Poiseuille equation , whereas capillary flow is treated as two-compartment porous media flow ., Macro-scale flow is coupled with micro-scale flow by a spatially distributing support function in the terminal endings ., Perfusion is defined as the transition of fluid from the arterial to the venous compartment ., We demonstrate a whole brain simulation of tracer propagation on a realistic geometric model of the human brain , where the model comprises distinct areas of grey and white matter , as well as large vessels in the arterial and venous vascular network ., Our proposed framework is an accurate and viable alternative to traditional compartment models , with high relevance for simulation of brain perfusion and also for restoration of field parameters in clinical brain perfusion applications . | An accurate simulation of blood-flow in the human brain can be used for improved diagnostics and assignment of personalized treatment regimes ., However , current algorithms are limited to simulation of blood flow within tumours only , and in terms of parameter estimation , traditional compartment models have limited accuracy due to lack of spatial connectivity within the models ., As a remedy , we propose a data-driven computational fluid dynamics model where the geometric domains for simulation are defined from state-of-the art MR acquisitions enabling a segmentation of large arteries and veins ., In the capillary tissue we apply a two-compartment porous media model , where the perfusion is pressure-driven and is defined as the transition of blood from arterial to venous side ., In addition , we propose a model for dealing with the intermediate scale problem where the vessels are undetectable and the flow does not adhere to requirements of porous media flow ., For this scale , we propose a support function distributing the fluid in a nearby region around the vessel terminals ., Combining these elements , we have developed a novel full human brain blood-flow simulator . | medicine and health sciences, classical mechanics, fluid mechanics, nervous system, vertebrates, animals, simulation and modeling, amphibians, tongue, materials science, network analysis, digestive system, research and analysis methods, computer and information sciences, fluid dynamics, blood pressure, continuum mechanics, fluid flow, physics, porosity, eukaryota, mouth, anatomy, central nervous system, biology and life sciences, physical sciences, material properties, vascular medicine, frogs, organisms | null |
journal.pcbi.1000549 | 2,009 | Tipping the Balance: Robustness of Tip Cell Selection, Migration and Fusion in Angiogenesis | Sprouting angiogenesis 1 is integral to a plethora of normal and pathological biological processes , such as embryonic development , wound healing and cancer ., Sprouts are headed by migrating tip cells , which exert pulling forces on their neighbouring stalk cells ., Tip cells are guided by , amongst other signals , external VEGF gradients released by nearby oxygen deficient tissue 2 ., Tip cells meet and fuse , forming blood vessel loops ( anastomosis ) which can support flow , though the full mechanism remains unclear 3 ., A dense vascular network builds up , which is later pruned by remodelling processes related to flow 4 ., Under pathological conditions , such as cancer and diabetic retinopathy , overly high VEGF concentrations are observed , resulting in aberrant sprouting and a tortuous , leaky network with poor perfusion 5 , 6 ., Normalisation of blood vessel development in disease has important therapeutic potential ., Improving vascularisation and blood flow in cancer can reduce the occurrence of metastatic cells by lowering tumour hypoxia and mutation rate 6 ., Normalisation also improves drug delivery by creating better supply lines to the tumour cells 6 ., Our overall aim is to first use a systems approach to understand how changes in VEGF result in a switch between normal and abnormal tip cell selection , sprouting and fusion and then predict methods to tip the balance back to normal angiogenesis in disease ., Our main approach is computational modelling , which has been highlighted as essential when the dynamics of a process are so intricate and multi-faceted that predicting the outcome of perturbations is no longer intuitive 7 ., However , whilst the model we present here is an adaptable , modular , generic computer algorithm , easily transferable to the investigation of cell motility/signalling dynamics in a diverse range of biological applications , it is highly important that models maintain focus if we are to answer specific questions in a chosen biological domain ., Therefore , we keep our parsimonious in silico model closely tied to quantitative in vivo angiogenesis imaging data throughout the paper; an approach recently highlighted as imperative for understanding complex developmental systems 8 ., Angiogenesis has previously been modelled using a variety of approaches including cellular automata and continuum models 9–14 ., Individual-based approaches , such as agent-based modelling used here , have been highlighted as more appropriate over continuum methods for sprouting simulations , which involve low numbers of cells 15 ., Other models considering the early stages of angiogenic sprouting have focussed on endothelial cells themselves releasing diffusible attractants and inhibitors 15–18 ., However , recent evidence shows Dll4-Notch juxtacrine signalling drives sprout initialisation and external gradients of attractants released by oxygen deficient tissue guides the sprouting process 2 , 19 ., Recently a simple binary model of tip cell selection was used in a sprouting model 20 , but did not involve the essential ingredient of dynamic Dll4-Notch lateral inhibition and instead allocated tip cell status based on exceeding a single threshold of environmental VEGF ., In our work , by modelling the full interconnected pathways and protein interactions underlying tip cell selection and migration , we elucidate the combined , system level effects of ongoing lateral inhibition with migration , fusion and increasing VEGF concentrations ., The central , novel feature of this model is that tip/stalk cell fates are continually in flux , due to the realistic modelling of ongoing Dll4/Notch lateral inhibition , known to control tip cell selection 19 ., Fates are not permanently fixed , as in other models 16 , 20 ., This yields significant new insights into the subtle dynamics , sensitivities and recursive nature of the process ., Moreover , to the Authors knowledge , this is the first computer model to consider physical mechanisms of filopodia ( long , thin , dynamic actin-based membrane protrusions ) led migration of adhered endothelial cells in vessels , the complex morphological changes that occur in cell shape and junction size during sprouting/fusion and the spatial feedback this gives to the ongoing signalling networks ., An agent-based model is utilised here as it provides a simple , effective methodology for capturing the distributed hierachical and morphological aspects of the multiscale sprouting process ., Physical tension in the membrane is modelled by the inclusion of springs following Hookes law , connecting the memAgents , see Fig 1 ( A ) , facilitating realistic modelling of filopodia-led cell migration ., Springs have been employed to great effect in a number of other biological models , such as in nematode sperm migration 21 , where springs simulate internal actin filament cross-connections driving the lamellapodia of a single cell ., Another example uses springs to mimic forces between cells , of migrating enterocytes in the intestine 22 and epithelial cells dividing and sorting in the intestinal crypt 23 ., However , here we make the significant extension that springs model both the internal actin-based cortical tension and filopodia driven migration in single cells , but also the higher , cell-cell level interactions of multiple , adhered cells ., Fig . 1 ( B ) shows the pathways we include in the model in a simplified manner: the actin driven migration pathway and the Notch-driven selection pathway; both require activation of the receptor VEGFR-2 ., Migration is known to occur through the kinases p38 and PI3K , leading to actin polymerisation and filopodia-led migratory behaviour 24–26 ., It can be seen as a positive feedback loop as further actin-driven movement forward will increase the input signal of VEGF ., The Notch pathway is a negative feedback loop: VEGF binds to VEGFR-2 receptors on the endothelial cells , which up-regulates expression of the ligand Dll4 27 ., Dll4 binds to Notch receptors on neighbouring cells via juxtacrine signalling and results in the down-regulation of VEGFR-2 receptors in the neighbour cell 28–30 ., Under normal conditions this lateral inhibition has been shown , in silico and in vivo , to generate an evenly scattered arrangement of tip cells separated by not more than two stalk cells ( termed a ‘salt and pepper pattern’ ) ., Moreover , under pathological conditions we previously predicted that cell fates can oscillate 31 ., This is consistent with observations made in other systems , where negative feedback combined with high input signals and delays ( caused by rates of gene expression and protein transport ) produces oscillations 32 ., Three separate investigations were performed with the model leading to the following main conclusions/predictions ., First , analysis of emergent behaviour observed in the model led to the prediction that , under normal VEGF conditions , tip cell fusion locally disrupts stabilised tip/stalk selection patterns , by altering the network of interconnected cells , and causes a local flip in tip/stalk fates; the new junction formed when tip cells fuse creates a new influence on the ongoing lateral inhibition ., In contrast , oscillations in tip/stalk fate , occurring in high VEGF , are shown to be robust to anastomosis events ., Secondly we predict tip cell migration alone has a destabilising effect on Dll4/Notch selection as it causes lengthening and distortion of cell-cell junction sizes ., Junction size changes affect the strength of Dll4/Notch signalling between different neighbours ., This is shown to contribute to the switch between normal and pathological angiogenesis ., We go on to make a novel prediction for therapeutic intervention from this study: reducing migration can normalise sprouting in pathologically high VEGF environments ., Finally , validation of a novel filopodia-adhesion driven mechanism for migration , based on the filopodia-adhesion driven migration of neuronal growth cones described in 33 , is presented by comparing our in silico model with in vivo data from zebrafish vascular development ., MemAgents have two states , which define their overall type and determine the behaviours they can choose to implement:, 1 ) their physical state and, 2 ) their filopodia-related location ., The physical state is set as either: node , spring or surface ., Node agents have springs connecting them to other node agents , forming the main mesh , and are part of the physical structure of the membrane ., Spring and surface agents do not apply forces ., They are created and deleted on-the-fly , sitting upon the mesh , to maintain a continuous membrane surface as the mesh changes shape , they are explained in the next section ., A memAgents physical type remains set for its lifetime ., The filopodia state is set as either: none , base , shaft or top , see Fig . 1 ( C ) ., All memAgents are initialised in state ‘none’ ( not in a filopodia ) ., Each memAgent is allocated a number of VEGFR-2 , Notch and Dll4 by its cell agent at each time step , as described in 31 and text S1 ., It then changes its filopodia state based on the accumulation of actin , accrued by local VEGF activating its VEGFR-2 receptors ., See Tables 2 , 3 and 4 for the attributes and initial settings of each agent type in the system ., As the square mesh representing the cell expands and stretches , the space between nodes increases ., In order to represent the continuous nature of the cell surface , when snapped to the 3D gridded lattice , a voxelisation algorithm was used ., Without it holes would occur as nodes get stretched further apart by migration forces , leaving apparent gaps in the mesh ., To cover the surface , the space between each four-node square of the spring-agent mesh is divided into two triangles ., The triangles are then voxelised , creating a new memAgent at each grid site that the triangle plane passes through , see Supporting Information for details of the voxelisation algorithm ( text S3 , fig S2 , S3 , S4 , S5 ) ., The new memAgents created to complete the surface are allocated the physical state ‘surface’ as opposed to ‘node’ ., Only node agents are allowed to extend filopodia - as these are the only agents which can convey forces through the mesh ., The ‘surface’ agents instead sit on top of the mesh , associated with a particular surface triangle ., They pass actin they accrue from their receptor activation to the nearest node agent of their triangle ., For continuous coverage of filopodia springs , a new memAgent is created in every grid site that the filopodia passes through ., These ‘spring’ agents pass any actin they accrue upwards towards the top node , which can then use them to extend ., When springs in a filopodium are deleted and new ones created , e . g . in shaft node addition/deletion ( described below ) , the spring agents associated with the deleted springs are reassigned to the new springs ., This maintains continuity in actin token possession along the filopodia ., All node memAgents that do not have an adhesion to the substrate , update their position each time step , according to the forces being applied to them by their springs ., MemAgents with an adhesion remain stationary , anchored to their current position ., The force on a node of its springs is given by: ( 1 ) where is the vector position of the neighbour node at the other end of the spring and is the specific spring constant for that type of spring ., There are three types of spring in the model , normal mesh , filopodia and junction springs ., See Table 1 for the specific spring parameters of each type ., Spring constants were estimated by matching morphological curvature of tip and stalk cells with confocal images ., If is too high , then the curve of the cell is too sharp and straight , thus callibrating these constants against images yielded realistic curvature and migration speeds in the model ., A full sensitivity analysis of these spring constants was also performed , details are given in the results section ., SN is given by: ( 2 ) where is the natural rest spring length for that type of spring ., The new position of node agent is given by ., However , since biological membranes can contract without springing back out , springs in the model do not generate a force when the spring length goes below the set ideal length ( ) ., Junctions between cells are represented by a specific type of ‘junction spring’ ., To keep junctions between cells tight , junction springs have a stronger spring constant , see Table 1 ., Biological cells are able to extend a number of different types of protrusion 37 ., Endothelial tip cells almost exclusively favour filopodia , which contain long parallel bundles of actin filaments 2 , these help the cell to navigate 38 ., We base the filopodia dynamics in the model principally on insights from sensory growth cone studies reported in 33 , where filopodia are shown to have three morphologically distinct regions: basal , shaft and tip ( hereafter tip is referred to as top to avoid confusion with tip cells ) ., Each region has its own type of adhesion with distinct functionality ., Basal adhesions are critical to filopodia initiation , shaft adhesions inhibit veil advance ( migratory advance of the cell body ) and the top adhesions can signal the release of all shaft adhesions when a certain environmental stimulus is found 33 ., This release allows the cell body to quickly advance in the optimal direction ., As tip cells migrate , once the veil advance mechanism has been triggered , they meet other tip cells and ‘fuse’ forming a new junction , termed anastomosis ., To implement new junction formation: if any two non-filopodia memAgents , from different tip cells , are located on adjacent grid sites , they change their status to ‘junction’ and create a new ‘junction’ spring between them ., If either memAgent is a ‘surface’ agent , then the new junction spring is created from the nearest node of the surface triangle to which it belongs ., If a new junction spring is created between two agents , then new junction springs are also created between all direct neighbour nodes in the mesh , with the corresponding node neighbours on the other tip cell , which helps zipper up the newly formed junction ., To limit excessive junction springs , a new spring is only created if both nodes contain no other junction springs to the that cell ., Figure 3 ( A ) shows blood vessels in the retina sat upon the astrocytic network , a network of inter-connected star-shaped astrocyte cells ., The isoform is known to adhere to astrocytes rather than freely diffusing through interstitial space , as opposed to the isoform 42 ., In this work , only the isoform is modelled ., Two simple astrocyte patterns were used in the simulations ., The first was based on a section of real retina , which biases the second and fourth cells to become tips , in a row of four cells , see Fig 3 ( B , C ) ., This topology was used when observing system behaviour , and where selection dynamics and bias were not a concern ., To implement this , the state ‘astrocyte’ was assigned to grid sites within a predefined ‘A shape’ pattern as seen in confocal image Fig 3 ( B ) ., The second astrocyte pattern was a simple unbiased square latticework; all cells experience the same VEGF levels and the same layout of astrocytes ahead ., To implement this the state ‘astrocyte’ was assigned to grid sites within a predefined pattern of interlacing strutts with equal dimensions , width 9 , depth 4 grid sites , extending to the grid boundaries ., MemAgents could not move into ‘astrocyte’ grid sites; however , filopodia , given their thin representative size , could co-exist in an astrocyte grid site ., VEGF adherence to the astrocytes was modelled by allocating a number of VEGF molecules to grid sites with astrocytes in their local neighbourhood , starting ahead of the vessel , ( only for grid sites where their coordinate is greater than , see Table 1 ) ., VEGF increased linearly in the axis ., However , this alone could only guide cells straight upwards , and in parallel ., To force cells to meet , in simulations involving anastomosis , a short-range , local gradient from a point source of VEGF was overlayed ., The use of local gradient ‘signposts’ fits with biological observations 2 , 43 ., Thus the amount of VEGF in each grid site was given by ( 3 ) where is the position of the grid site on the y axis , is the distance of the grid site from the point source , is a small random fluctuation , the value of which is randomly chosen from zero to 0 . 3 ( the final VEGF level was not allowed to go below zero ) ., and are the VEGF levels for the linear gradient and any point sources respectively ., If the distance exceeds or is greater than the point sources maximum range , is reset to equal and thus no point source VEGF is allocated in that grid site ., The ‘normal’ level of was set to 0 . 2 and to 0 . 15 for all runs with point source VEGF ., Point sources were placed at a distance of from the vessel , similar to the distance between fusing tip cells in the zebrafish and mouse retina data presented ., Referring back to Fig . 2 the whole model is simulated as follows ., Within each timestep all memAgents are updated by:, 1 ) assessing their nearest neighbours in the look up lattice-grid;, 2 ) calculating their new receptor activation levels ( VEGFR-2 is activated by local lattice site VEGF levels , Notch is activated by local Dll4 in memAgents from a neighbouring cell ) then, 3 ) deciding whether to implement any filopodia behaviours ., Subsequently , once all memAgents have been updated ,, 4 ) the off-lattice positions of all node agents are recalculated; implementing Hookes law and applying to all springs ., Once all memAgents have updated , each cell agent:, 1 ) calculates new total levels of active receptors and ligands ( VEGFR-2 receptors are down-regulated by Notch activity , and Dll4 is up-regulated by VEGFR-2 activity , see text S1 for exact equations ) ;, 2 ) removes spring and surface agents and calculates new coverage of the mesh based on new node positions then;, 3 ) Updated levels of receptors and ligands are equally distributed to all cell memAgents after a delay representing transcription/translation rates ( Notch and Dll4 are localised to memAgents at junctions only ) ., The process then repeats ., Dll4-Notch lateral inhibition between endothelial cells is responsible for tip cell selection ., However , lateral inhibition is an ongoing process , continuing throughout angiogenesis ., The interaction between ongoing lateral inhibition and the further stages of angiogenesis , migration and anastomosis , have not yet been investigated in vivo , in vitro and not least in silico ., An interesting , unexpected , emergent property of the system was observed in simulations when migration , anastomosis and lateral inhibition occur concurrently ., The formation of a new junction between fusing tip cells during anastomosis provides a new border where the two tip cells now battle , via Dll4/Notch signalling , to inhibit each other ., This causes a destabilisation of the current tip/stalk pattern and results in one of the fusing tip cells being inhibited , see video S2 in Supporting Information ., We wanted to investigate whether this emergent property could cause a destabilisation along the entire vessel , and whether different starting selection patterns and VEGF concentrations would have an effect ., A vessel containing seven cells was initialised with two point sources of VEGF placed such that they may induce two loops , see Fig . 4 ., However , the number of cells in the vessel preclude more than one loop forming initially; only two or three tips can be stably selected from seven cells ., Thus if two loops form a flip must have occurred in a stalk cell , caused by destabilisation from anastomosis of the first loop ., To allow time for two loops to form , the simulation was run for a maximum of 8000 timesteps ., See video S3 for a typical simulation run ., It was found that only two of all the possible initial tip/stalk patterns could generate a flip in a fate and produce two loops instead of one: S T S T S S T and S T S S T S T . This is because they contain an adjacent set of stalk cells ., If the tip cell inhibited during anastomosis is next to the adjacent stalk cells a row of four adjacent stalk cells is produced ., This allows one stalk cell to flip and become a new tip cell ., In the first pattern , the tip cell required to be inhibited for the flip to occur is cell 4 , and in the second arrangement , cell 5 ( furthermore called the ‘correct cell’ inhibited ) ., The stalk cell that flips is necessarily cell 5 and cell 4 respectively , as the others all still have a strong tip cell neighbour ., Fig . 5 ( A ) shows that the correct cell inhibited causes a flip in fate of the neighbouring stalk cell to a new tip , forming a second loop in the majority of cases; if the incorrect tip cell is inhibited , which could happen due to stochastic events , i . e . cell 2 or 7 respectively , then no second loop is observed ., An arrangement such as S T S T S T S was found to never result in two loops; a flip of any tip cell to a stalk cell results in an equally viable arrangement , thus no stalk cell will flip ., Using VEGFR-2 levels as a marker for cell fate , it can be clearly seen from Fig . 5 ( B ) that a cells fate changes during the simulation ., Once the first loop forms , the central cell indeed flips fate , triggering the second loop formation ., Fig . 5 ( C ) shows that when 10 times the normal VEGF level is used , the behaviour of the system is markedly different; a synchronous oscillation is observed ., Interestingly under these pathological conditions , although new junctions form at a continuous rate , flipping of cell fates does not occur: oscillations are impervious to the changes in the network caused by anastomosis ., The higher incidence of new junction springs in high VEGF , leads to sheet formation of multiple tip cells , rather than simple fusion of two non-adjacent tip cells ., This is due to all cells attempting to become tip cells during each period of the oscillation and fusing together ., See video S4 for a typical simulation in high VEGF ., This emergent sheet-like sprouting morphology is interestingly consistent with observations of developing mouse retinas in conditions of excessive VEGF concentrations by intraocular injection 2 ., Mathematical models of lateral inhibition tend to naturally focus on regular , symmetric arrangements of cells to avoid any intrinsic or unintended bias 44 , 45 ., However , the junctions between real endothelial cells during sprouting are known to be highly contorted , irregular and indeed dynamic as cells stretch and move , as seen in Fig . 6 ( A , B ) , 3 ., This irregularity could , we hypothesised , lead the unequal Dll4-Notch interactions to bias certain cells to become tips and thus stabilise a salt and pepper pattern faster ., To investigate this , cells were initiated with irregular junctions and the stability of the system evaluated ., Contrary to our initial simple hypothesis however , the emergent behaviour of the system exposed a far more important , subtle role for dynamic junction size in robustness of tip/stalk patterning ., The simulations were performed on a vessel comprising ten segment regions ( wide ) , but instead of segments representing one cell as before , they were divided into two cells ., This facilitated unequal boundaries between neighbouring cells ., Four distinct arrangements of cells were used: two ‘control’ arrangements , with regular , equal sized junctions and two irregular settings ., These different arrangements were achieved by simply assigning a boundary offset parameter ( in radians ) for the position of the junction between each segments two cells , see Fig . 6 ( C ) ., The ‘Unequal Neighbours’ setting was expected to give the fastest stabilisation , due to the extra bias of different number of neighbour cells ., As the cells were now half the size of those in previous simulations , the receptor and ligand levels per cell were also halved ., Here we show that the ‘veil advance’ mechanism for migration and navigation , based on filopodia adhesion signalling described in 33 , yields realistic tip cell morphology and behaviour during migration and anastomosis ., Figure 3 shows characteristic retinal tip cell morphology; filopodia can be very long and curved ., Often tip cells display large numbers of filopodia without any evidence of migration of the main cell body ( Fig ., 3 ( F ) green arrow ) ., Fig . 3 ( E ) white arrow , and ( F ) yellow arrow , show that where filopodia from different tip cells appear to have met , significantly more cell body has been pulled up along them ., Figure 3 ( D ) shows that switching off the filopodia-led veil advance mechanism yields unrealistic tip cell morphology ., This is for two reasons , firstly the filopodia are entirely straight , having no mechanism for curvature without adhesions ., Secondly , with adhesions veil advance can be delayed until the optimal direction for migration has been found; instead , without adhesions , cells instantly begin migrating in all directions , along every filopodia , with multiple environmental cues this would be highly inefficient in vivo ., Without the veil advance mechanism , or a similar process , the morphology indicated by the green arrow in Fig . 3 ( F ) , with filopodia but no cell body advance , would not be possible ., Subsequently , time lapse in vivo movies were taken of developing intersomitic vasculature in the trunk region of zebrafish embryos ( the retina cannot currently be live imaged ) to validate the simulated veil advance mechanism ( see video S1 ) ., Quantifications were made in the zebrafish and simulations concerning:, a ) the number of contacts between tip cell filopodia;, b ) how long they remained in contact and, c ) how long it took between, 1 ) the first contact made between filopodia and fusion of the two tip cells , and, 2 ) the final filopodia contact and fusion of the two tip cells ., See Supporting Information ( text S4 , fig S6 and S7 ) for the live imaging method , specific parameter settings and the methods used in the simulations ., The graphs in Fig . 10 show that a similar frequency distribution of contact lifetimes is obtained in simulation to in vivo , with a smoother curve in simulation due to averaging over fifty runs compared to eight anastomosis events in vivo ., The in silico and in vivo contact time frequency were found to have no significant difference ( ) ., Fig . 10 ( C ) and ( D ) show the time between the first filopodia contact and tip-tip fusion , and the time from the last filopodia contact and fusion are also similar in simulation to in vivo ., It takes slightly longer to fuse from the first contact in simulation , however , the average time from final contact until fusion is almost exact ., Here we have presented the significant extension of physical tension to our existing agent-based model , and considered the minimal , interlaced , pathways of ongoing lateral inhibition-driven tip cell selection , actin-driven migration and sprout fusion in normal and pathological angiogenesis ., Through three separate investigations we arrived at a number of significant predictions ., First , we reported emergent phenomena observed in simulations , showing anastomosis destabilises the established tip/stalk pattern and causes cell fates to flip ., We also showed , in high VEGF , oscillations are robust to fusion events ., Based on these simulation results the phrase ‘cell fate’ no longer seems appropriate for tip/stalk cells ., Instead , given their reversible nature we suggest they be viewed as ‘phenotypes in flux’ ., The second investigation focused on the effects of realistic , unequal cell-cell junction sizes on tip cell selection ., This study gave surprising and intriguing results: if average junction size is inversely proportional to the VEGF level , then normal selection is possible , regardless of the VEGF level ., Our initial hypothesis was that unequal junctions would simply increase selection rate due to the inherent bias ., Instead , actual junction size , rather than relative size difference , was found to be a key , novel factor for fast and stable selection ., Migration was shown to pull and stretch junctions , widening the range of sizes , which in turn feeds back to reduce stability of tip/stalk patterning in high VEGF ., We showed that reducing migration , via the use of simulated actin polymerisation inhibitors , can reduce junction stretching and allow normal selection to occur in pathologically high VEGF ., Lastly we quantitatively validated the novel filopodia-adhesion mechanism , seen only before in neuronal growth cones , against new in vivo live imaging data in zebrafish , establishing it as a plausible model deserving further investigation ., These results suggest migration is a key positive feedback loop involved in switching between normal and pathological angiogenesis ., Figure 1 ( B ) shows how the migration and selection protein pathways inter-relate ., It is clear , just in terms of protein-pathways , that the positive feedback of migration will affect the functioning of the negative-feedback selection pathway; increased VEGFR-2 activity levels caused by migration will increase Dll4 expression , which can lead to abnormal oscillations ., Positive feedback is known to make oscillations generated by negative feedback more robust in other biological systems , e . g . in the cell cycle , circadian rhythms and in tunable synthetic biological systems 52–54 ., Interestingly it appears that the angiogenic sprouting system may exhibit more robustness in its pathological behaviour than in its normal functioning and may explain why it exhibits such sensitivity to VEGF levels , falling into abnormal development at just two times the normal amount 55 ., Modularity , the localisation of functionally distinct processes to independent pathways , is known to improve the robustness of a system 56; mutations or perturbations in one process are unable to disrupt those in a separate pathway ., However , the sprouting pathway architecture , Fig . 1 ( B ) clearly exhibits a lack of modularity; VEGFR-2 is central to migration and selection ., Disruptions in VEGFR-2 activation will cause both processes to be affected , which in turn affects fusion events and the network integrity ., It has been argued that biological systems , in particular morphogenesis , are intrinsically robust 56–58 ., Thus , it is interesting to uncover a case where the system shows sensitivity under normal conditions and may instead fall into robust pathological behaviour ., However , it has also been argued that robustness may be a by-product of morphogenesis evolution , rather than directly selected for 59; as such its omission in a particular developmental process is not unlikely , if a non-robust mechanism had other selective advantages ., Indeed we have shown that the sensitivity of selection to fusion events , disrupting cell fates , has the advantage that new tip cells are rapidly selected , allowing the recursive generation of sprout loops , without requiring a long wait for cell division to provide new tips ., Insights into such trade-offs between robustness and performance , and why pathological systems fail , has been highlighted for its therapeutic value 56 ., Here it has enabled us to arrive at a new insight: that therapeutic intervention of the migration pathways in endothelial cells could aid normalisation of angiogenesis in diseases characterised by high VEGF levels ., Our computer model , with its unique blend of spatial cell migratory morphological plasticity and local signal interpretation capacity , is central to making these observations , which could not have been achieved with any other model currently available ., Nevertheless , as with all computer models , | Introduction, Model, Results, Discussion | Vascular abnormalities contribute to many diseases such as cancer and diabetic retinopathy ., In angiogenesis new blood vessels , headed by a migrating tip cell , sprout from pre-existing vessels in response to signals , e . g . , vascular endothelial growth factor ( VEGF ) ., Tip cells meet and fuse ( anastomosis ) to form blood-flow supporting loops ., Tip cell selection is achieved by Dll4-Notch mediated lateral inhibition resulting , under normal conditions , in an interleaved arrangement of tip and non-migrating stalk cells ., Previously , we showed that the increased VEGF levels found in many diseases can cause the delayed negative feedback of lateral inhibition to produce abnormal oscillations of tip/stalk cell fates ., Here we describe the development and implementation of a novel physics-based hierarchical agent model , tightly coupled to in vivo data , to explore the system dynamics as perpetual lateral inhibition combines with tip cell migration and fusion ., We explore the tipping point between normal and abnormal sprouting as VEGF increases ., A novel filopodia-adhesion driven migration mechanism is presented and validated against in vivo data ., Due to the unique feature of ongoing lateral inhibition , ‘stabilised’ tip/stalk cell patterns show sensitivity to the formation of new cell-cell junctions during fusion: we predict cell fates can reverse ., The fusing tip cells become inhibited and neighbouring stalk cells flip fate , recursively providing new tip cells ., Junction size emerges as a key factor in establishing a stable tip/stalk pattern ., Cell-cell junctions elongate as tip cells migrate , which is shown to provide positive feedback to lateral inhibition , causing it to be more susceptible to pathological oscillations ., Importantly , down-regulation of the migratory pathway alone is shown to be sufficient to rescue the sprouting system from oscillation and restore stability ., Thus we suggest the use of migration inhibitors as therapeutic agents for vascular normalisation in cancer . | Abnormal vasculature exacerbates many diseases such as cancer and diabetic retinopathy ., In angiogenesis new blood vessels , headed by a migrating tip cell , sprout from pre-existing vessels in response to chemical signals ., The signals are released from newly oxygen deficient tissue ., The signals are known to be different in disease and are thought to cause the process of angiogenesis to progress abnormally , though the reasons for this remain unclear ., Normalisation of angiogenesis has great potential as a therapeutic strategy; it has been shown to reduce metastasis and improve drug delivery in tumours ., Here we focus on the behaviours of three inter-related initial angiogenic pathways associated with changes in tissue signal conditions , utilising both in silico and in vivo approaches ., By the construction and implementation of a novel computational model for cell motility and signal processing we present a new theory on why angiogenesis exhibits such sensitivity to signal changes and show that the behaviour in disease is surprisingly more robust than normal functioning ., This we attribute to the positive feedback of cell migration reinforcing abnormal oscillations in cell fate selection ., We make the unique prediction that normalisation could be achieved by reducing cell migration alone . | computer science/applications, developmental biology/morphogenesis and cell biology, cell biology/cell signaling, cell biology/morphogenesis and cell biology, cell biology/developmental molecular mechanisms, developmental biology/pattern formation, ophthalmology/diabetic retinopathy, developmental biology/cell differentiation, cell biology/cell adhesion, computational biology/systems biology, cardiovascular disorders, cell biology/cytoskeleton | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.