Datasets:

griffin

/

ChemSum

like 13

Characterization of the T4 gp32–ssDNA complex by native, cross-linking, and ultraviolet photodissociation mass spectrometry

Protein-DNA interactions play crucial roles in DNA replication across all living organisms. Here, we apply a suite of mass spectrometry (MS) tools to characterize a protein-ssDNA complex, T4 gp32$ssDNA, with results that both support previous studies and simultaneously uncover novel insight into this noncovalent biological complex. Native mass spectrometry of the protein reveals the co-occurrence of Znbound monomers and homodimers, while addition of differing lengths of ssDNA generates a variety of protein:ssDNA complex stoichiometries (1 : 1, 2 : 1, 3 : 1), indicating sequential association of gp32 monomers with ssDNA. Ultraviolet photodissociation (UVPD) mass spectrometry allows characterization of the binding site of the ssDNA within the protein monomer via analysis of holo ions, i.e. ssDNAcontaining protein fragments, enabling interrogation of disordered regions of the protein which are inaccessible via traditional crystallographic techniques. Finally, two complementary cross-linking (XL) approaches, bottom-up analysis of the crosslinked complexes as well as MS1 analysis of the intact complexes, are used to showcase the absence of ssDNA binding with the intact cross-linked homodimer and to generate two homodimer gp32 model structures which highlight that the homodimer interface overlaps with the monomer ssDNA-binding site. These models suggest that the homodimer may function in a regulatory capacity by controlling the extent of ssDNA binding of the protein monomer. In sum, this work underscores the utility of a multi-faceted mass spectrometry approach for detailed investigation of non-covalent protein-DNA complexes.

characterization_of_the_t4_gp32–ssdna_complex_by_native,_cross-linking,_and_ultraviolet_photodissoci

Introduction<!>Materials and reagents<!>Sample preparation<!>Direct infusion experiments<!>Bottom-up LC-MS/MS of crosslinked samples<!>Data analysis<!>Results and discussion<!>Conclusions

<p>Protein-DNA interactions are fundamental for transcriptional regulation, DNA repair, and replication. 1,2 The genome is organized in double-stranded DNA (dsDNA); however, genetic information is accessed for transcription or replication as a single-stranded DNA (ssDNA) intermediate. ssDNA is inherently less stable and must be protected to preserve genomic integrity. All organisms, from viruses to humans, have evolved single-stranded DNA binding proteins (SSBs) to protect and stabilize ssDNA intermediates. SSBs bind ssDNA independently of sequence and with high affinity, and are critical for the sequestration and stabilization of ssDNA, which occur in preparation for DNA replication and DNA repair processes. [3][4][5] The bacteriophage T4 gene protein 32 (also known as T4 gp32) is a prototypical member of this family and is required for viral replication. 3,6 Gp32 is divided into two subdomains, subdomain I and subdomain II, which are linked through a connecting region. 7 An oligonucleotide/oligosaccharide-binding (OB) fold forms the structural core. 7 An electropositive cle consisting of aromatic and basic residues within the connecting region facilitates non-specic ssDNA binding. [7][8][9] Gp32 discriminates between ssDNA and dsDNA based on the hydrophobic interactions that are created between a pocket of aromatic side chains of gp32 within the electropositive cle and the bases of a ssDNA substrate. 9 An X-ray crystal structure of gp32 revealed a zinc nger consisting of His64, Cys77, Cys87 and Cys90 within the Cterminal tail. 7 This motif maintains the subdomain I structure. 7 To bind to ssDNA, the acidic C-terminal domain of the gp32 protein must undergo a conformational change that exposes the positively-charged region of its core domain, which in turn, interacts with the negatively charged ssDNA backbone. 10,11 Gp32, and all other SSBs, bind ssDNA transiently. 9 In the apo (unbound) state, gp32 is a mixture of monomers and dimers, owing largely to the inhibitory effect of the C-terminal domain on the binding cle. 9 Gp32 binds ssDNA in a cooperative manner, in which the binding of one gp32 molecule increases the binding affinity such that additional gp32 monomers bind to the ssDNA substrate in a contiguous process. 9,12 These ndings indicate that gp32 binding to ssDNA likely occurs via selfdissociation of gp32 dimers and subsequent monomer-bymonomer binding. 12,13 Subdomain II (N-terminal region) facilitates cooperative binding on ssDNA via electrostatic interactions between the core domains of adjacent gp32 molecules. 9,11 Here, we re-examine gp32 self-assembly and ssDNA interactions via a suite of advanced mass spectrometry (MS) methods. [14][15][16][17] Electrospray ionization (ESI) facilitates the transfer of intact proteins and protein-DNA complexes from the solution into the gas phase for subsequent mass spectrometric interrogation, in many cases preserving both the structure and stoichiometry of the non-covalently bound complexes. 18,19 Many studies have exploited ESI-MS as a means to preserve and facilitate analysis of native interactions, including DNA-small molecule ligand complexes, [20][21][22][23] DNA-templated silver clusters, [24][25][26][27] DNA duplex and quadruplex complexes, [28][29][30] and protein-DNA complexes. [31][32][33][34][35][36][37][38][39] Native ESI-MS studies of DNA complexes have enabled the determination of DNA-ligand stoichiometry, 32,38 ligand selectivity, 21 and DNA-binding pathways. 35 Analyses of native complexes such as these are contingent upon the exceptional performance of high-resolution and high-mass accuracy mass spectrometers. 40 UV-induced crosslinking MS has been developed for interrogation of protein-DNA complexes via bottom-up LC-MS/MS analysis of the DNApeptide crosslinks aer proteolytic digestion. 41 In general, compared to ESI-MS of multiprotein complexes, fewer studies have reported the analysis of protein-DNA complexes. [31][32][33][34][35][36][37][38][39] Analysis of protein-nucleic acid complexes containing large DNA or RNA strands (>20 nucleotides) in positive-ion mode typically generates spectra that are complicated by the presence of highly heterogeneous ion populations arising from cation adduction. 42 However, the addition of volatile salts (e.g., ammonium acetate) minimizes the prevalence of cation adduction for protein-DNA complexes. 43,44 While single-stage mass spectrometry (MS1) experiments of native DNA-protein complexes in volatile salt solutions provide basic mass information and binding stoichiometry, MS/MS, oen via collisionally activated dissociation (CAD), is typically used to glean more detailed structural information. The use of MS/MS for structural interrogation of oligonucleotidecontaining complexes has emerged as useful for a few studies focused on ligand localization and conformational changes. 22,26,28,31,34,45 However, in the case of non-covalent complexes, collisional activation methods, although providing some sequence information, do not afford extensive information related to ligand localization and overall complex structure. To combat these decits, alternative activation methods that are more suitable for the analysis of non-covalent macromolecular complexes have been developed, including surfaceinduced dissociation (SID), 46 electron activation methods, 47 and ultraviolet photodissociation (UVPD). 48 The latter uses high-energy UV photons to promote cleavages along multiple positions of the protein backbone, generating a diverse series of fragment ions including a, a + 1, b, c, x, x + 1, y, y À 1, and z ions. 49,50 Through access to these higher-energy fragmentation pathways, UVPD yields both ligand-containing (holo) ions and ligand-free (apo) product ions, thus enabling the determination of ligand-binding sites and revealing conformational reorganization. 51,52 Complementary to the information generated via UVPD of native protein-ligand complexes is the insight that can be obtained from cross-linking mass spectrometry (XL-MS). XL-MS has been previously utilized to unveil protein-protein interaction networks and to monitor conformational changes of multimeric complexes. 53,54 The covalent linkages created by crosslinking allows mapping of protein-protein binding interfaces, 55 especially when used in conjunction with molecular docking soware. 56 Crosslinking of non-covalent protein complexes can be analyzed using traditional bottom-up approaches to gain distance information between neighboring residues, or via analysis of the intact crosslinked complexes to decipher the overall complex stoichiometry of crosslinked subunits.</p><p>Herein, we showcase the use of native UVPD-MS data to characterize viral protein-ssDNA complexes comprised of bacteriophage T4 gene product 32 (gp32) and ssDNA substrates (dT12, dT20). Complementary information gathered from native UVPD-MS and XL-MS unravel the unique characteristics of gp32's ssDNA binding mechanism. Using these approaches, we establish with individual amino acid precision how gp32 monomers bind a ssDNA substrate. In addition, we show the non-sequence specic gp32 binding to ssDNA. Integrating insight obtained from bottom-up methods and MS1 analysis of intact crosslinked complexes demonstrates that gp32 does not bind ssDNA as a multimer (e.g. dimer or trimer), and rather dissociates into monomers prior to ssDNA binding. Bottom-up XL-MS enables generation of two model dimer structures in which the ssDNA-binding region of each monomer overlaps with the homodimer interface. This work substantiates native MS, UVPD-MS and XL-MS as complementary techniques for probing protein-nucleic acid interactions with unprecedented structural resolution.</p><!><p>Gp32 (pIF89) and gp32-DCTD (amino acids 1-254, pIF898) were cloned with a C-terminal intein-chitin binding domain. pIF89 was transformed into BL21 ArcticExpress E. coli cells (Agilent) and supplemented with 50 mg mL À1 carbenicillin and 20 mg mL À1 gentamycin and grown at 30 C. Cells were induced with 0.5 mM IPTG and cultured overnight at 12 C. Resulting pellets were resuspended in resuspension buffer (50 mM Tris-HCl pH 7.5, 500 mM NaCl, 1 mM EDTA, 10% sucrose (w/v), 1 mM PMSF) and sonicated to lyse. Lysate was loaded on 5 mL chitin resin (NEB) equilibrated with buffer A (50 mM Tris pH 7.5, 100 mM NaCl, 1 mM EDTA). Resin was extensively washed with buffer B (50 mM Tris pH 7.5, 1 M NaCl, 1 mM EDTA). Intein cleavage was performed on resin by incubating with buffer B supplemented with 50 mM DTT overnight at 4 C. Resulting elution was concentrated with a 10 kDa MWCO concentrator (Amicon) and dialyzed in storage buffer (20 mM Tris pH 7.5, 150 mM NaCl, 10% glycerol (v/v)). Protein was ash frozen in liquid nitrogen and stored at À80 C. Final protein molecular weight is approximately 33.5 or 28.5 kDa based on SDS-PAGE results (Fig. S1 †). The molecular weight results obtained from SDS-PAGE depend not only on the protein theoretical molecular weight but also on amino acid composition, 57 and are in agreement with gp32 SDS-PAGE results provided by various T4 gp32 protein vendors.</p><p>Oligonucleotides were purchased from Integrated DNA Technologies (Coralville, IA, USA). LC solvents including LC-MSgrade water, formic acid, and acetonitrile were acquired from Sigma-Aldrich (St. Louis, MO). Bis(sulfosuccinimidyl)suberate (BS3), DMTMM, MS-grade trypsin, formic acid (99.5+%) and Pierce™ C18 Spin Columns utilized for bottom-up XL-MS sample clean-up were purchased from Thermo-Fisher Scientic (Waltham, MA). LC analytical columns (15 cm, 75 mm inner diameter) containing C18 stationary phase (3 mm diameter) were packed in-house. Micro Bio-Spin™ P-6 Gel Columns (Bio-Rad Laboratories Inc., Hercules, CA) were used for desalting, buffer exchange, and size exclusion chromatography (SEC).</p><!><p>For native MS, gp32 and gp32-DCTD solutions were diluted to 10 mM in 50 mM ammonium acetate (AmAc), whereas denatured samples were diluted to 10 mM in 50/50 acetonitrile/water with 0.1% formic acid. Similarly, dT12 and dT20 strands were diluted to 10 mM in water and added to the gp32-and gp32-DCTD-AmAc solutions for an incubation period of approximately 10 minutes at 25 C. The resulting gp32 and gp32 + dT solutions were desalted, buffer exchanged, and subjected to SEC using 6 kDa SEC lters. For XL-MS experiments, gp32 was reconstituted to 0.1 mM into a 1 phosphate buffered saline solution at pH 7.2. 10 mM BS3 or 10 mM DMTMM stock was diluted to 5 mM in water and then allowed to react with gp32 at protein/crosslinker molar ratio of 1 : 10 (1 hour incubation at 25 C) for the samples subjected to proteolytic digestion and bottom-up LCMS/MS analysis, or 1 : 100 (10 minutes incubation at 25 C) for the MS1 analysis of the intact complexes. 58,59 For the MS1 analysis of the intact complexes, the crosslinking reactions were quenched with 0.5% formic acid aer the 10 minute crosslinking reactions to avoid excessive crosslinking. Excess (unreacted) crosslinker and monomer were removed by passing the reaction solution through 30 kDa SEC lters. The crosslinked samples prepared for bottom-up analysis were quenched with AmAc in 50 excess of the cross-linker. In preparation for trypsin digestion, these samples were further diluted with 150 mM ammonium bicarbonate. Trypsin was added at a protein/protease molar ratio of 1 : 40 and incubated for 16 h at 37 C. 58 Prior to LC separation, samples were cleaned up with C18 spin columns, dried with a SpeedVac, and reconstituted in 2% acetonitrile.</p><!><p>Equimolar ($10 mM each) gp32 or gp32-DCTD + ssDNA solutions were loaded into Au/Pd-coated nanospray borosilicate static tips (prepared in-house) for nESI. A heated capillary set to 200 C was used to desolvate the protein-DNA complexes, aiding their transmission into the gas phase. All direct infusion experiments were conducted on a Thermo Scientic Q Exactive UHMR mass spectrometer customized for the implementation of UVPD as described earlier, 60 except for the MS/MS experiments undertaken on denatured proteins which were performed on a Thermo Orbitrap Elite mass spectrometer, also equipped with a 193 nm excimer laser for UVPD as previously described. 61 The robust sensitivity of the UHMR mass spectrometer to ions of high m/z and its optimized optics for the retention of electrostatic complexes were ideal for analysis of intact complexes and afforded information about the heterogeneity of protein complexes with high molecular weights. [62][63][64] Modications made to the higher-energy collisional dissociation (HCD) cell allowed the implementation of UVPD using a 193 nm ArF excimer laser (Excistar, Coherent, Santa Cruz, CA). Modulation of the injection atapole and interatapole voltages allowed in-source trapping (IST), enabling front-end collisional activation that improved analysis of native proteins and protein-DNA complexes. 60,[65][66][67][68] MS1 analysis of intact crosslinked proteins were analyzed on the UHMR instrument. MS1 spectra were collected at R ¼ 1563 and averaged over 25 scans. All HCD and UVPD mass spectra were collected using an isolation width of 5 m/z, a resolution of 200 000 (@m/z 200), a trapping gas value of 1 and IST of À50 V. UVPD was performed with a single pulse at 1-3 mJ per pulse. UVPD conditions were optimized to maximize coverage, and 1 pulse at 3 mJ per pulse was found to provide the highest number of fragment ions (Fig. S2 and S3 †). All UVPD mass spectra were collected in triplicate for each laser energy condition.</p><!><p>Tryptic digests of cross-linked proteins were separated on a Thermo Scientic Dionex UltiMate 3000 nano-LC system equipped with a house-packed C18 trap (3 cm  100 mm i. d.) and analytical columns (15 cm, 75 mm i. d.) and analyzed using a Thermo Scientic Orbitrap Fusion Lumos Tribrid mass spectrometer. The resolution was set to 60 000 and 30 000 (@m/ z 200) for MS1 and MS2 spectra, respectively. MS/MS data collection was performed at top-speed mode with a 3 s cycle time, 1  10 5 intensity threshold, 50 ms maximum injection time, xed mode HCD, 25% collision energy, 5  10 4 AGC target, and 2 mscans per scan. Dynamic exclusion was utilized for ions of the same m/z observed 3 times in the MS1 spectra within a rolling 20 s elution window, with an exclusion duration of 20 s and a mass tolerance of 25 ppm. Cross-linked peptides were separated using a gradient of 2% B to 35% B to 90% B over the course of 60 minutes. Mobile phases A and B consisted of 0.1% formic acid in water and 0.1% formic acid in acetonitrile, respectively. A ow rate of 300 nL min À1 was used throughout the 60 minute separation.</p><!><p>Deconvolution of all high-resolution MS/MS spectra was performed using the Xtract algorithm (Thermo Fisher) at a S/N of 3, while deconvolution of low-resolution MS1 spectra was performed using UniDec. 69 ProSight Lite was used for all sequence coverage analysis using an error tolerance of 10 ppm. Mass shis of +79.94, +61.91 and +3589.62 Da, +6024.00 Da (with 10 ppm tolerance) and combinations of these mass shis were used for the C-terminal covalent S 2 O modication, Zn 2+ cofactor (e.g., addition of one Zn atom and loss of two hydrogen atoms), dT12 (Fig. S4a †), and dT20 (Fig. S4b †), respectively. To determine the backbone cleavage yields generated by UVPD, the abundances of the holo ions and their corresponding apo ion series were collectively summed from triplicate runs acquired using three different UVPD conditions (one laser pulse applied at 1 mJ, 2 mJ or 3 mJ). To allow direct comparison across all spectra, the identied holo/apo ions were normalized to the total ion current of the spectrum, as previously described. 60,70,71 All structural representations of gp32 are based on the X-ray crystal structure produced by Shamoo et al., which contains residues 22-239 of the entire 301 amino acid sequence (PDB 1GPC). 7 Backbone cleavages derived from the identied holo fragment ions were plotted onto the gp32 crystal structure (PDB 1GPC) using UV-POSIT 72 and a series of Python and MATLAB (MathWorks) scripts. For this analysis, backbone cleavages derived from holo ions (containing both Zn 2+ and DNA) were included only if seen at least twice within the three sets of UVPD data (obtained from triplicate runs using 1 pulse, 3 mJ per pulse).</p><p>Bottom-up XL-MS data were searched against appropriate protein databases using Byonic soware (Protein Metrics, San Carlos, CA) with a tolerance of 10 ppm for precursor and fragment ions, a maximum of two missed cleavages, and a 1% FDR cutoff. Strict qualications were utilized during the K-K and D/ E-K crosslink ltering process. In brief, peptides were grouped by unique peptides, ltered with a score cutoff of 300 or higher, and were further substantiated with a Pep2D of 9.9  10 À5 or lower. The cross-linked peptide hits of the gp32 homodimer were veried using ClusPro 2.0 protein-protein docking soware (Boston, MA) in conjunction with PyMOL Molecular Graphics System, Version 2.0 (Schrodinger, LLC). ClusProgenerated gp32 homodimer structures that displayed interprotein K-K crosslinks of 30 Å or lower and D/E-K crosslinks of 15 Å or lower were selected.</p><!><p>To investigate the stoichiometry, structure and mechanism of this viral protein-ssDNA system, native MS was used to analyze solutions containing gp32 or gp32$ssDNA complexes. These species were characterized using MS/MS, including both HCD and UVPD. Crosslinking experiments were undertaken to explore the gp32 homodimer interface, including: (1) direct infusion of the intact complexes formed via crosslinking of the gp32 homodimer and subsequent incubation with ssDNA, and (2) crosslinking of gp32 followed by a traditional bottom-up workow with tryptic digestion and LC-MS/MS analysis for identication of cross-linked peptides.</p><p>We observed gp32 monomers and homodimer ions via native ESI-MS (Fig. 1a), in agreement with a prior FRET study. 10 Each gp32 monomer retained one zinc atom, resulting in 1 : 1 or 2 : 2 gp32$Zn complexes (see Fig. S5 † for deconvoluted spectrum). Zn(II) chelation is directly correlated to gp32$ssDNA binding, indicating that gp32 remains folded during MS analysis. 73 MS/MS characterization of the gp32$Zn complex (11+) resulted in 55% sequence coverage by UVPD (1 pulse, 3 mJ) (Fig. 1b and S6 †) and 5% coverage by HCD (Fig. S7 †) based on production and consideration of both apo sequence ions (no Zn) and holo sequence ions (Zn retained). In agreement with previous studies, [74][75][76] we observed little change in UVPD sequence coverage based on precursor charge state (Fig. S8a and b †). C-terminal modication of gp32 was also identied based on the UVPD mass spectrum and veried via subsequent bottom-up LC-MS/MS analysis (Fig. S9 †) and additional topdown MS/MS analysis of denatured gp32 (Fig. S10 †). This Cterminal modication is consistent with incorporation of a disulfur monoxide moiety at Cys302 and is attributed to the intein reaction which occurs during protein purication.</p><p>We detected a disulde bond between Cys87 and Cys90 based on the sequence maps derived from the UVPD and HCD mass spectra of denatured gp32 (Fig. S10a-c †). Upon addition of the reducing agent TCEP (Fig. S10d †), this disulde bond is disrupted, resulting in the observation of two new backbone cleavages occurring between the two previously linked cysteines (Fig. S10e and f †). Additionally, markedly increased sequence coverage was obtained for the disulde-reduced proteins upon HCD and UVPD (32% and 41% sequence coverage, respectively) (Fig. S10e and f †) compared to the non-reduced proteins (HCD 21% sequence coverage, UVPD 33% sequence coverage) (Fig. S10b and c †), consistent with previous reports indicating that the presence of disulde bonds impedes fragmentation of intact proteins. 77 The ion signal originally dispersed among oxidized and reduced proteoforms may be concentrated into the reduced form aer addition of the reducing agent, thus increasing the abundance of the precursor ion available for MS/ MS analysis. This "concentration" of the precursor into a more homogeneous form likely contributes to the increased sequence coverage observed for the reduced protein compared to the nonreduced protein. These results, along with the notable absence of a disulde bond in the native protein (Fig. S6b †), suggest that the Cys87-90 disulde bond forms upon denaturation of gp32 and loss of the Zn 2+ atom, which is consistent with previous studies which highlight the role of Zn in the coordination of residues Cys87 and Cys90 in the native gp32 protein and stabilization of subdomain I. 73 We next turned to examination of gp32 in complex with ssDNA oligonucleotides. Addition of ssDNA oligonucleotide, dT12, to the gp32 solution resulted in production of 1 : 1 gp32$dT12 complexes, as observed in several charge states (all containing one Zn as seen in Fig. 1c see Fig. S11a † for deconvoluted mass spectrum). Nearly all of the 1 : 1 and 2 : 2 gp32$Zn complexes originally seen in Fig. 1a shied to the ssDNA-bound species. Increasing the oligonucleotide length to dT20 resulted in detection of 2 : 1 and 3 : 1 gp32$dT20 complexes (with two and three Zn, respectively) (Fig. 1d and deconvoluted spectrum in Fig. S11b †). UVPD mass spectra of the 2 : 1 and 3 : 1 gp32$dT20 complexes are displayed in Fig. S12, † both of which show the production of numerous large-size fragment ions, some of which may correspond to portions of two gp32 proteins held together by non-covalent interactions in addition to ones with and without dT20 and Zn. Given the high probability of false positives when searching for ions that include sub-portions of two or more molecules of the same protein, these types of ions were excluded from the searches. Only fragment ions corresponding to sequence ions from apo gp32, gp32 + Zn, gp32 + dT20, or gp32 + Zn + dT20 were considered, as shown in Fig. S12c and f. † These results indicate the ability of native MS to monitor the oligomerization of gp32 via binding to ssDNA as the length of the oligonucleotide increases, and are consistent with an approximately 6,7nucleotide footprint, as reported in previous gp32 studies. 10,78,79 These results likewise establish that native MS can distinguish the stoichiometries of native protein-DNA complexes and reveal the oligomerization of gp32 and its binding to ssDNA as a function of the DNA strand length.</p><p>Next, we used UVPD to interrogate the structures of the gp32$ssDNA complexes. UVPD generates both apo (without DNA) and holo (with DNA) fragment ions that enable localization of the gp32$dT12 interactions with individual amino acid precision. 51,52 To minimize mis-assignment of ions, we focused on the holo fragment ions that retained the entire mass of the DNA sequence. Restricting the holo fragment ion searches to only those that retain the entire ssDNA streamlines the searches and xes the mass shi of the holo fragment ions to a dened value (in this case +3589.62 Da for dT12), much in the same way that searches and localization of specic post-translational modications are successfully executed in other MS/MS analyses of proteins. 80,81 Prior UVPD studies of other protein-ligand complexes have either not observed fragmentation of the ligand, as the amide backbone of the protein is a signicant UV chromophore, or have not undertaken searches for fragment ions containing sub-portions of the ligands owing to the enormous search space and potential for false positives. 52,60,70,71,82,83 In essence, considering only holo fragment ions which contain the entire mass of the ssDNA ligand may exclude the identication of some meaningful ions but importantly avoids increasing the false positive rate for identication of misassigned spurious holo fragment ions.</p><p>Each gp32$ssDNA complex can generate three types of holo fragment ions: ones containing only Zn (e.g., loss of DNA and retention of Zn), ones containing only DNA (e.g., loss of Zn and retention of DNA), and ones containing both Zn and DNA. Znfree product ions may arise from backbone cleavages in stretches of the protein remote from the Zn binding site, or may arise from ejection of Zn from Zn-coordinated fragment ions during UVPD. The summed distributions of apo ions and these three types of holo ions were categorized based on whether they contain the N-terminus (all a/a + 1, b, c ions) or C-terminus (all x/x + 1, y/y À 1, z ions) and directly compared to the distribution of apo and holo (Zn) ions produced from the DNA-free gp32 precursor (Fig. 2a). While the fragment ion distributions from gp32 (11+) display a nearly equal distribution of apo N-and Cterminal fragment ions, gp32$ssDNA (11+) shows a shi towards preferential production of apo N-terminal ions. With respect to the production of holo fragment ions (i.e., Zncontaining product ions from the gp32 precursor or fragment ions containing Zn, DNA, or DNA + Zn from the gp32$ssDNA complex), the number of holo ions containing the C-terminus is signicantly greater than the number of holo ions containing the N-terminus upon UVPD of the gp32$ssDNA complex, whereas gp32 produces more holo fragment ions containing the N-terminus compared to those containing the C-terminus. These results suggest that Zn in gp32 is coordinated closer to the N-terminus, whereas ssDNA is coordinated closer to the Cterminus in the gp32$dT12 complex.</p><p>All of the holo ions produced from the 1 : 1 gp32$dT12 complexes were further grouped by holo ion type (i.e., containing DNA, DNA + Zn, or Zn) (Fig. 2b) as well as by specic ion type (a/a + 1, b, c, x/x + 1, y/y À 1, z) based on the UVPD spectra shown in Fig. S13. † In general, the various ions types (a/a + 1, b, c, x/x + 1, y/y À 1, z) are observed both with and without retention of Zn, as well as with and without retention of dT12. It is this large array of assignable fragment ions that facilitates localization of dT12 as the gp32$dT12 complexes disassemble and release fragments that retain dT12. These patterns of different holo ions again highlight that dT12 is coordinated closer to the C-terminus, as there are a signicantly greater number of C-terminal holo ions than N-terminal holo ions. This contrasts the apo ion data shown in Fig. 2b for UVPD of the 1 : 1 gp32$dT12 complexes in which N-terminal apo ions are nearly twice as abundant as C-terminal apo ions. Fig. 2c shows an even more detailed breakdown of the various types of fragment ions produced from the gp32$ssDNA complexesa-type ions are most abundant for the apo fragment ions, whereas x-and y-type ions are most abundant for the various holo ion types. The observed shi towards a higher number of C-terminal ions for the holo fragment ions vs. apo fragment ions produced from gp32$ssDNA (Fig. 2b) as well as the shis in distribution of holo and apo fragment ions upon UVPD of gp32 versus gp32$ssDNA (Fig. 2a) underscores how the presence of dT12 changes the fragmentation pattern of this protein.</p><p>In order to localize the ssDNA ligand binding site, the DNAcontaining holo ions (ones containing a portion of the gp32 sequence plus the entire dT12 sequence, with or without retention of Zn) were collectively summed based on the UVPD mass spectra acquired for the 1 : 1 gp32$dT12 complex (11+). This holo ion mapping method with UVPD has been previously used to determine binding sites for protein-small molecule ligand complexes but is here used for the rst time to interrogate noncovalent protein-DNA interactions. 52,84 The sequence map that illustrates the backbone cleavage sites that lead to the resulting dT12-containing fragment ions is shown in Fig. 3a, demarcated based on whether the fragment ions contain the Cterminus (red) or N-terminus (blue), or in some cases complementary N-terminus and C-terminus ions (green). The corresponding sequence coverage map showing all the backbone cleavage sites that generate all the identied DNA-containing holo ions is displayed in Fig. S14. † As observed from the map, most of the C-terminal holo ions contain over 200 residues. Most N-terminal holo ions, although few in total, are equally large, containing over 200 amino acids. It is reasonable that these long stretches of the protein would retain the DNA. However, there are also a number of fragment ions whose compositions are consistent with just a few residues of gp32 bound to the entire dT12, such as (y 6 $dT12) 2+ (a short Cterminal holo ion) and (a 5 $dT12) 3+ (a short N-terminal holo ion) (see conrmations of these ions via isotopic ts in Fig. S15. †).</p><p>The backbone cleavage sites from which the C-terminal and N-terminal holo ions originate were plotted onto the gp32 crystal structure (PDB ID: 1GPC) 7 (Fig. 3b, and sequence motif map in Fig. S16 †) and color-coded to match the cleavage sites in the companion sequence map. Red, blue and green correspond to the backbone positions marking the C-terminal, N-terminal, and bidirectional DNA-containing holo ions, respectively. An almost continuous series of C-terminal holo ions is observed near the protein N-terminus in addition to a number of Nterminal holo ions at the protein C-terminus, indicating a large breadth of ssDNA contact throughout the entire central structural region of the protein (Fig. 3). These results indicate that gp32 binds ssDNA with the majority of residues of subdomain I, the connecting region, and subdomain II. 7 Furthermore, the array of small bidirectional holo ions observed at both the N-and C-terminus of the protein suggests that both the Nterminus and C-terminus of gp32 interact with dT12. Overall, this data sheds new light on the extent and reach of ssDNA binding with gp32, expanding the binding site from the electropositive cle, and implicates interactions of both the N-and C-terminal tails with dT12, adding new insight to the partiallysolved X-ray structure of gp32 (1GPC). 7 These results highlight the ability of MS-based structural methods to interrogate disordered regions of proteins which may be inaccessible via traditional techniques.</p><p>Next, we investigated the hypothesis that gp32 homodimers dissociate into monomers prior to binding to ssDNA. 9,12 First gp32 was incubated with Lys-Lys crosslinker BS3, and the resulting MS1 spectrum shown in Fig. 4a substantiates that the intact crosslinked products are predominantly dimers with an estimated average of 8 cross-linked and/or dead-end modied lysines per molecule of gp32 (MW exp of intact cross-linked gp32 dimer ¼ 69 755 Da, see deconvoluted spectrum in Fig. S17 †). The similar mass shis of each dead-end modication, which add a mass shi of +158 Da, and each crosslink (+138 Da), prevents specic differentiation of the two modications given the large mass of the dimer and the resolution of the mass Fig. 3 (a) Sequence of gp32 with the backbone cleavage sites leading to N-terminal ((a, a + 1, b, c) blue), C-terminal ((x, x + 1, y, y À 1, z) red) or bidirectional (green) dT12-containing holo fragment ions generated upon UVPD (combined data from 1 pulse at 3 mJ) of the 1 : 1 gp32$dT12 complex (m/z 3386, 11+ charge state, containing 1 Zn). Greyed out sequence areas correspond to regions of the sequence that are unresolved in the crystal structure but contain many confirmatory backbone cleavages that lead to assignable sequence ions. The C-terminal cysteine is shaded in gold to denote the disulfur monoxide modification. (b) Schematic model of the gp32 crystal structure (PDB ID: 1GPC, residues 22-239) with the residues corresponding to backbone cleavage sites that lead to holo fragment ions shaded in the respective colors (red for backbone cleavages that lead to C-terminal sequence ions, blue for backbone cleavages that lead to N-terminal sequence ions, and green for backbone cleavages that result in both C-terminal and N-terminal ions). The approximate ssDNA binding cleft is shaded in purple.</p><p>spectrometer. The deconvoluted MS1 spectrum (Fig. S17 †) indicates that the crosslinked dimer contains a maximum of 17 crosslinks (with no dead-ends) or a maximum of 15 dead-ends (no crosslinks), or various combinations of crosslinks and dead-ends between these two extremes. Because of the different possible mass additions related to the crosslinking reactions (i.e., dead-end modication, protein cross-link), it is not possible to determine whether the cross-linked gp32 dimer retains Zn. Given the four crosslinks identied later in Fig. 5, the cross-linked sample likely contains on average 4 crosslinks and 12 dead-end modications per dimer. The MS1 spectrum obtained aer addition of ssDNA (dT12) to the BS3-crosslinked gp32 solution is displayed in Fig. 4b. No gp32$dT12 complexes are observed, instead several dT12 dimers and crosslinked gp32 dimers are evident (Fig. 4b, S18a and b †). The inability of the cross-linked gp32 homodimer to form complexes with dT12 suggests that the binding cle of gp32 is blocked or disrupted upon crosslinking, thus disabling interactions with dT12 in a way that prevents the formation of the gp32$dT12 complexes. Additionally, the charge state distribution of the cross-linked gp32 dimer shis towards the lower m/z range (higher charge states) upon the addition of dT12 to the solution (Fig. 4b). We hypothesize that this shi in the charge state distribution may be due to the relatively high concentration of free dT12 in solution compared to the heterogeneous population of crosslinked gp23 dimers. Cross-linked gp32 dimer was also analyzed in the presence of dT20 (Fig. S19 †) with similar results, in that no binding was observed between dT20 and the crosslinked gp32 homodimer. In this case, the addition of dT20 did not result in a notable shi in the charge state distribution of the cross-linked gp32 dimer, and free dT20 is not observable in the mass spectrum. The corresponding MS1 spectrum acquired for a denaturing solution containing crosslinked gp32 and dT12 is shown in Fig. S20, † showing both denatured gp32 monomer and dimer. The charge state distributions of denatured, cross-linked gp32 dimers in Fig. S20a † ($14+ to 20+) is somewhat broadened relative to the charge state distribution of the native-like non-cross-linked gp32 dimers (Fig. 1a, $14+ to 18+), and the charge state distribution of denatured crosslinked gp32 monomers ($11+ to 14+ in Fig. S20a †) are extremely shied relative to the charge states of denatured gp32 monomers (Fig. S21, † $18+ to 31+). The intramolecularly crosslinked gp32 monomers may better maintain a more Fig. 4 Native MS1 spectra of (a) the solution containing gp32 after BS3 cross-linking and (b) the solution containing gp32 after BS3 cross-linking followed by addition of dT12, showing the absence of gp32$dT12 complexes. Clean-up of the solutions using 30 kDa SEC filters after crosslinking removed most of the monomeric gp32 and unreacted crosslinker. compact structure, preventing unfolding and charging, in addition to the conversion of the basic lysine side-chains to various less basic hydrolyzed deadend groups.</p><p>In order to localize the crosslinks in the gp32 dimer, a bottom-up LC-MS/MS strategy was used. Gp32 was incubated with either Lys-Lys crosslinker BS3 or Asp/Glu-Lys crosslinker DMTMM. The resulting solutions containing crosslinked gp32 were then subjected to tryptic digestion and MS/MS analysis to map the locations of the crosslinks. The base peak LC-MS chromatograms and the MS/MS spectra of the crosslinked peptides are shown in Fig. S22 and S23. † Four BS3-crosslinked peptides were identied, with crosslinks occurring between Lys residues 51 and 104, 104 and 104, 178 and 178, 190 and 67, and 190 and 71, while one DMTMM-crosslinked peptide was iden-tied with the linkage between D102 and K141 (Fig. S23a †). The crosslinked peptides were t to potential gp32 dimer structures generated by protein-protein docking. Fig. 5 shows the two optimized gp32 homodimer structures based on validation of the ve cross-linked peptide assignments. Model 1 incorporates 4 out of the 5 identied BS3 crosslinks, whereas Model 2 incorporates 4 out of the 5 identied BS3 crosslinks plus the DMTMM crosslink (D102-K141). All potential intramolecular crosslinked peptide hits were discarded based on comparing Lys-Lys and Asp/Glu-Lys distances in the model monomer vs. dimer structures with BS3 or DMTMM crosslinking distance constraints. Comparison of the two homodimer models in Fig. 5 to the gp32 crystal structure in Fig. 3b reveals overlap of the homodimer interfaces in each of the two models with the gp32 binding cle, suggesting that a large portion of the gp32 ssDNA-binding cle is involved in the gp32 homodimer interface. This result implies that the ssDNA binding cle is not accessible for ssDNA binding when gp32 is homodimeric. Our MS data supports a dimerization model that gp32 dimers are unable to bind ssDNA and that dimer dissociation is required prior to binding. We additionally puried a gp32 construct with a C-terminal tail deletion (gp32-DCTD, amino acids 1-254) to test whether the C-terminal acidic tail impacts dimerization, which is unresolved in the gp32 crystal structure. We still observed dimerization in the gp32-DCTD constructs via native MS1 (Fig. S24a †), but did not observe evidence of ssDNA binding (Fig. S24b †). Our data suggests that the acidic tail is dispensable for dimer formation, similar to the T7 phage gp2.5, 85 but is essential for ssDNA binding. We favor a model where the gp32 acidic tail mediates multiple protein-protein interactions, including with T4 gp59. 86</p><!><p>In summary, combining native MS, UVPD-MS and XL-MS expanded our molecular understanding of how T4 gp32 binds ssDNA and how this binding is inhibited for the crosslinked homodimeric gp32 complex. Native UVPD-MS combined with XL-MS provides direct evidence that the gp32 dimer dissociates into monomers prior to binding to ssDNA. The presence of 3 : 1 gp32$dT20 complexes observed in Fig. 1d also indicates a sequential association of gp32 monomers with ssDNA owing to the odd number of gp32 molecules that are bound to the ssDNA. Additionally, we characterized the ssDNA binding cle size and developed a model structure for the ssDNA-inactive gp32 dimer. We conjecture that gp32 dimers act as an inactive storage compartment and may also regulate the extent of ssDNA binding. Cellular processes that sequester gp32 into dimers also prevent ssDNA binding by this protein. Based on the successful characterization of the well-studied T4 gp32-ssDNA complexes, we anticipate that this multi-pronged MS approach will provide further in-depth characterization of other novel protein-DNA systems which may have unsolved structures and unknown binding modes.</p>

Royal Society of Chemistry (RSC)

A combined 3D-QSAR and docking studies for the In-silico prediction of HIV-protease inhibitors

BackgroundTremendous research from last twenty years has been pursued to cure human life against HIV virus. A large number of HIV protease inhibitors are in clinical trials but still it is an interesting target for researchers due to the viral ability to get mutated. Mutated viral strains led the drug ineffective but still used to increase the life span of HIV patients.ResultsIn the present work, 3D-QSAR and docking studies were performed on a series of Danuravir derivatives, the most potent HIV- protease inhibitor known so far. Combined study of 3D-QSAR was applied for Danuravir derivatives using ligand-based and receptor-based protocols and generated models were compared. The results were in good agreement with the experimental results. Additionally, docking analysis of most active 32 and least active 46 compounds into wild type and mutated protein structures further verified our results. The 3D-QSAR and docking results revealed that compound 32 bind efficiently to the wild and mutated protein whereas, sufficient interactions were lost in compound 46.ConclusionThe combination of two computational techniques would helped to make a clear decision that compound 32 with well inhibitory activity bind more efficiently within the binding pocket even in case of mutant virus whereas compound 46 lost its interactions on mutation and marked as least active compound of the series. This is all due to the presence or absence of substituents on core structure, evaluated by 3D-QSAR studies. This set of information could be used to design highly potent drug candidates for both wild and mutated form of viruses.

a_combined_3d-qsar_and_docking_studies_for_the_in-silico_prediction_of_hiv-protease_inhibitors

Background<!>Results and discussion<!>Statistics of the ligand-based models<!><!>Statistics of the receptor based models<!>Contour maps of CoMFA<!><!>Contour maps of CoMFA<!>Contour maps of CoMSIA<!><!>Docking results<!><!>Docking results<!><!>Docking results<!><!>Conclusion<!>Dataset preparation<!>CoMFA & CoMSIA 3D-QSAR models<!>Molecular docking by GOLD<!>Abbreviations<!>Competing interests<!>Authors’ contributions<!>Additional file 1<!>

<p>Human immunodeficiency virus (HIV) is a retrovirus that is peril to human health, responsible to cause AIDS, an immunodeficiency syndrome. The disease presents a serious health care challenge because each year it affects an increasing number of people across the globe [1]. To combat disease, several new drugs were approved by FDA which reduces the morbidity and mortality of HIV infection. These drugs are categorized as HIV-Reverse transcriptase (HIV-RT), HIV-Integrase (HIN-IN) & HIV-Protease inhibitors (HIV-PIs), the major targeted enzymes of HIV life cycle. HAART (highly active anti-retroviral therapy) is the most promising anti-AIDS therapy including these inhibitors in combination. The major obstacle in the use of HAART therapy is resistance that virus develops [2]. The hyper-mutability of HIV, drug resistance and their side effects are the biggest challenge to develop an effective anti-AIDS therapy.</p><p>HIV-1 Protease is emerging as one of the major druggable target for the development of new chemotherapeutics. HIV protease inhibitors, restrain the viral maturation by preventing the formation of structural and functional proteins and form immature, non-infectious virus. However, it is highly prone to develop mutations, since it is a homodimer and a single mutation of gene causes double mutation of enzyme [3]. Structurally, HIV protease is a homodimer protein, containing 99 amino acids in each chain, with an active site located at the dimer interface [4]. The protein is composed of three regions; catalytic core (Asp25, Gly27, Ala28, Asp29 and Asp30), flap (Ile47, Gly48, Gly49, and Ile50) and the C-terminal region (Pro81, and Ile84). From literature, Asp25, Gly27, Ala28, Asp29 and Gly49 are known to be highly conserved residues to which a potent inhibitor may bind strongly. Mutations of HIV protease at Val32, Ile50 and Ile84 (hydrophobic residues, close to binding pocket) are responsible for the resistance to most FDA approved drugs due to loss of Vander Waal interactions [5]. Almost all FDA approved anti-AIDs drugs are resistant to I84V mutant virus and became ineffective against disease.</p><p>The failure of drug therapies against mutated virus protein encouraged the scientists to develop more potent, effective and stable second generation HIV-PIs, but still the HIV-PI therapies are associated with the serious problems that limit their significance and effectiveness [6]. In order to take a forward step for prediction and guidance of more effective drug, 3D-QSAR studies were conducted as primitive step in finding new inhibitors using a dataset of 102 (R)-hydroxyethylamino sulfonamides derivatives from literature [7].</p><p>3D-QSAR technique is subdivided into ligand-based and structure-based methods. Ligand-based approach is frequently applicable in the absence of experimentally resolved protein crystal structure whereas, structure-based method extract the protein bound ligand information for the generation of align model [8-10]. In the present work, both strategies were applied to generate the CoMFA and CoMSIA models and their comparison with reference to the most active moiety Darunavir (hydroxyethylamino sulfonamides derivatives). Extensive research is ongoing that used different scaffolds, methodology and algorithms for predicting better results. Darunavir (DRV) is one of the most attracting targets as it is the most active molecule among eleven FDA approved drugs of present time [11]. The obtained models revealed the significance of stereoelectronic properties, hydrogen bonding characteristics and structure variations leading to changes in the interaction profile. The influences of grid distances, alignment methods and combination of charges were explored out of which the best model was selected. Additionally, molecular docking of compounds explored the binding affinity of highly active and least active compounds with its receptor by using GOLD docking suit [12]. The purpose of the study was to validate the experimental results obtained with Darunavir derivatives and to predict the compound that may developed into a more potent HIV inhibitor based on outcomes extracted from the current study.</p><!><p>Protease active site is composed of catalytic triad having two C2 symmetrical monomeric units, Asp25 (25')-Thr26 (26')-Gly27 (27'). This triad is surrounded by amino-acids, classified into S1 (1') and S2 (2') sub-sites, which mostly include the hydrophobic amino-acids [13]. However, on ligand binding, Protease behaves as asymmetrical monomer [14]. Darunavir, an FDA approved drug has shown extensive hydrogen bonding with protease backbone, especially with S2 sub-site of protease, moreover it also retained interaction with mutated protein [15].</p><p>In the present work, the additive model of Jorissen R.N. et.al., [7] was further subjected to 3D-QSAR using CoMFA & CoMSIA techniques and the generated contour maps were further validated by molecular docking.</p><!><p>The reliability of CoMFA and CoMSIA models were highly dependent on the better alignment of molecules in a three dimensional space. The database alignment implemented in Sybyl7.3 [16] was used to align 102 compounds using most active compound 32 as a template. The core structure of compound 32 was chosen as a structural element for superimposition of all other compounds (Figure 1a). The alignment is shown in Figure 1b and c. The statistical model of training and test tests (Tables 1 and 2) generated for the initial data set was depicted in Table 3. From the results, it can be deduced that lowering the grid space showed negative impact on the model. The default value of the grid space was selected as best and was used for further studies. To validate the model by external test set, activities of 24 compounds were predicted and the residual values for external and internal data sets were evaluated (Table 2). The best model with convincing statistical results is shown in Table 3 and the residual value for the best model was found to be less than 1 in both training and test sets as mentioned in Tables 1 and 2. Furthermore CoMSIA was applied on the same dataset and the results are tabulated in Table 4.</p><!><p>Core structure and dataset alignment. a) Core structure of danuravir derivatives with marked points used for alignment, b) Ligand-based alignment by using most active compound 32 as template, c) Structure-based alignment using cognate ligand of 3QOZ.pdb as reference.</p><p>Ligand-based and structure-based, actual and predicted pIC50 values of training set generated by CoMFA model along with their residuals</p><p>Ligand-based and structure-based, actual and predicted pIC50 values of test set generated by CoMFA model along with their residuals</p><p>The statistics of all generated CoMFA models in order to obtained the best model</p><p>Where: GS grid spacing, q2: cross validated correlation coefficient, SEP Standard Error of Prediction, C optimal number of Components, r2: non-cross validated correlation coefficient, SEE Standard Error of Estimation, F Fischer test values, rpred2: prediction of external test set for validation.</p><p>Ligand-based and structure-based CoMSIA models along with percentage contribution of their descriptors</p><p>Where: q2: cross validated correlation coefficient, r2: non-cross validated correlation coefficient, rpred2: prediction of external test set for validation, F Fischer test values, C optimal number of Components, SEE Standard Error of Estimation, SEP Standard Error of Prediction, %1-5: percentage contribution of descriptors in the field, respectively, S Steric field, ES Electrostatic field, H Hydrophobic descriptor, D hydrogen bond Donor field, and A hydrogen bond Acceptor field.</p><!><p>In ligand-based approach, several combinations of charges and grid spacing were used. Among them, the model generated by using MMFF94 charges was retrieved as the best model with q2 value of 0.74, standard error of prediction was 0.99 and the r2 value of 0.96. The results are summarized in Table 4. For structure-based method, the bound conformation of Darunavir in the crystal structure of HIV protease (PDB: 3QOZ) [17,18] was used as a template to align the series of 102 compounds (Figure 1c). As shown in Table 1, the structure-based QSAR method returned with the q2 value of 0.682, r2 of 0.938, F value of 178.46 and lower standard error of estimate and standard error of prediction with an average residual values of 0.077. While the r2 value of the test set was 0.947. This statistical evaluation showed that the performance of the structure-based method was comparable to the ligand-based approach for CoMFA studies (Table 3).</p><p>In CoMSIA, cross validated value of 0.664 and 0.751 was obtained for the structure-based and ligand-based methods, respectively. The CoMSIA analysis is tabulated in Table 4. Similarly, predictive r2 value was 0.927 and 0.929 for structure-based and ligand-based methods, respectively.</p><!><p>CoMFA contours of different colors represented different fields i.e. steric (bulky favored- green whereas yellow is indicative of bulky disfavored area). Similarly, blue and red regions described electron donating and accepting groups would be favored or disfavored, respectively.</p><p>Figure 2a and 2c displayed CoMFA generated steric and electrostatic contour maps for ligand-based and structure-based models, respectively. The most active compound 32 was superimposed on the steric and electrostatic contours maps for clear illustration.</p><!><p>CoMFA contour maps. The contour maps of CoMFA modeling, sterically favored areas are represented by green isopleths while yellow regions are served for sterically unfavorable regions. However, electropositivity and electronegativity are represented by blue and red contours, respectively. a-b are representative of ligand-based CoMFA descriptors of most active (comp-32) and least active (comp-46) whereas c-d demonstrate structure-based CoMFA contour maps with active and in-active compounds, 32 and 46, respectively.</p><!><p>The analysis of contour maps generated by ligand and structure-based methods showed that the electronegativity (red polyhedral) is favored at R1 position in compound 32 where 3-phenyloxaolidin-2-one ring is present. While, the presence of prop-1-ene group at this position in compound 46 has a negative effect on the biological activity depicted in Figure 2b (ligand-based) and 2d (structure-based). Similarly, electropositivity (blue contours) is favored between benzene ring and nitrogen of 3-phenyloxaolidin-2-one in compound 32. The increase or decrease in electronegativity, represented by red contours at R1, indicated its effect on observed biological activities. If we compared compounds 28–31 with 7–14, it was found that they have huge difference in their inhibitory activity due to the difference in number of electronegative fluorine at R1 position which buried near red isopleth. Even the compounds having propanone moiety at same position, more declined activity was observed. Second red polyhedral was observed near R2 position, surrounded the isobutane moiety of compound 32, which demonstrated that the substitution of electronegative element at this position could further enhance the biological activity of the compound 32.</p><p>At R2 position, less bulky group would be favorable for biological activity, indicated by yellow polyhedral. Compounds 4, 6, 21 and 55 contained bulky group at this position and considered as less effective with inhibitory activity as compared to active. Similarly, comparison of compound 43 with template 32, it was revealed that replacement of 2-methyl thiophene with less bulky substituent at R2 position would help to enhance its inhibitory activity. A large green polyhedral found near R3 position indicating if replaced anisole moiety of compound 32 with more bulkier group would be beneficial for better activity.</p><p>Presence of methoxy phenyl at para position of compound 32, strongly favored the inhibitory activity as electronegative and bulky group is required at R3 position. Compounds which pose methoxy phenyl group at this position, showed activity not less than 8.38. While compounds 43 and 46 contained isoxazole group at this position, could be the reason of their reduced activity.</p><!><p>The CoMSIA steric and electrostatic descriptors were found to be identical with the CoMFA generated models, which proved the consistency of the results. Moreover, the results of other three descriptors of CoMSIA also improved the drug prediction. The hydrogen bond donor and acceptor descriptors revealed the reason of higher activity of compound 32. At R1 position of 32, the purple polyhedral is surrounded which showed that this is donor disfavored region. In compound 32, this donor disfavored region is supplemented by the presence of highly electronegative elements in 3-phenyloxazolidin-2-one ring. At R2 position hydrogen bond acceptor is disfavored (red polyhedral); at this position an alkyl chain is present in compound 32. At R3 position a hydrogen bond acceptor is favored (magenta polyhedral), which is supplemented by the presence of methoxy group. In contrast, these properties are absent in least active compound 46 which possibly the reason of its lower activity.</p><p>The hydrophobic descriptor of CoMSIA is important to evaluate the hydrophobicity required to sustain the biological activity of any compound. At R1 position, hydrophobicity is highly disfavored (white isopleth) whereas R3 is hydrophobic favored (yellow contours) region. As shown in Figure 3c, compound 32 contained nitrogen containing hydrophilic moiety at R1 position while this hydrophilic moiety is absent in compound 46 (Figure 3f). In compound 32, the R3 position is substituted with the phenyl-methoxy group while compound 46 contained hetero-atomic methyl-isoxazole moiety at R3 position, showed that hydrophilic substitution at R3 position would decrease the biological activity of compound 46. The CoMSIA contour maps of compound 32 and 46 with ligand-based and structure-based approaches are presented in Figures 3 and 4, respectively.</p><!><p>CoMSIA ligand-based descriptors. Representation of ligand-based CoMSIA descriptors with most active and least active compounds. a-c depicted steric & electrostatic, acceptor & donor and hydrophobic descriptor maps of most active compound, respectively (32), whereas d-f showed all five descriptor contours with least active compound (46).</p><p>CoMSIA structure-based maps. Illustration of structure-based CoMSIA descriptors. Upper portion marked as a-c displayed steric & electrostatic, acceptor & donor as well as hydrophobic contour maps of compound 32 claimed as most active. However, d-f are representative of compound 46's descriptor maps marked as least active compound within the series.</p><!><p>To validate the 3D-QSAR results, docking simulation was performed and the most active compound 32 and least active compound 46 was evaluated for their binding interactions in the active site of protease and results were compared. Initially, the performance of docking software was tested by re-docking experiment. For this purpose, crystal structures of two proteins with their cognate ligands were retrieved from PDB and the cognate ligands were re-docked. The results are summarized in Table 5. The superimposed view of docked conformation and the reference ligand is presented in Figure 5a-b. Based on the re-docking results, GOLD was used for docking. The comparison of the scores attributed by two scoring functions as Gold-Score and Chem-Score also showed the compound 32 to be more active than 46 in both wild and mutated proteins. However, Gold-score showed drastic difference between the scores of two compounds which can be assumed on this basis to more accurate than Chem-Score.</p><!><p>Re-docking and docking results of wild type and mutated with most active and least active compounds</p><p>Re-docking poses and RMSD values. Re-docking results of a) wild type (3EKV) and b) mutated (3NU9) proteins with RMSD of 1.255Å and 1.32Å, respectively.</p><!><p>On the basis of docking analyses, it was revealed that compound with highest activity (32) ranked at top position as compared to least active compound 46. The docking scores were in correlation with 3D-QSAR and experimental results. The docked conformation of compound 32 in wild type (Figure 6a-b) and mutated proteins (Figure 6c-d) revealed that compound interacted with the binding pocket residues of targeted proteins through several favorable interactions including polar, hydrophobic, hydrogen bonding and the weak Van der Waal contacts.</p><!><p>2D and 3D docking representations. A representation of docking interactions and poses of most active and least active molecules with wild type and mutated HIV-protease protein via 2D and 3D representations. a-d most active compound (32) interacted with important active site residues of wild type and mutated proteins, respectively. Similarly, e-h represents interactions of least active compound within binding pocket of wild type and mutated proteins to show how the compound 46 lost its interactions and activity due to conformational change occurred in response to mutation.</p><!><p>The carbonyl oxygen of the core structure near R1 position also mediated strong hydrogen bonding with the backbone amino group of Asp29' and Asp30'. Moreover, hydrophobic interactions were observed between Ile50 and core group of compound 32 and acetophenone with Arg8. Pro81' also mediated hydrophobic interaction with the methyl group of methoxybenzene present at R3 position. Furthermore side-chains of S1' residue Val82' mediated CH--π contact with the hydrophobic portion of the ligand at R3 position. Val32' mediated CH3--π interactions with the core benzene of compound 32.</p><p>The observed docked conformation of compound 32 in the mutated protein (I84V) was flipped at ~90°, showed in Figure 6c-d. Even with this orientation, the ligand was found to be interacting with several important residues including Gly27, Gly27', Asp25, Asp25', Asp29, Asp29', Ile50, GLy49' and Ile50'. In this case the Gly27 interacted with R2 substitution and Gly27' with core structure of compound 32. A hydrogen bond was observed between the side chain oxygen of Asp25' and the hydroxyl of compound 32 (2.04Å). Furthermore Asp29' mediated a strong hydrogen bond with oxygen atom of R1 3-phenyloxazolidin-2-one ring with the distance of 1.95Å. Moreover, the compound is stabilized by the hydrophobic interactions offered by Ile50, Gly49' and Ile50'. These interaction patterns of compound 32 with the wild type and mutated forms of protein suggested that the modification at R2 position could increase the activity of compound. This hypothesis further confirms the results obtained by CoMFA.</p><p>The docked conformation of compound 46 in the wild type protein (Figure 6e-f) revealed that it formed CH3--π interaction with side chain of Ile50 and Val82', however, core benzene of compound 46 also mediated aromatic interaction with Pro81'. On the other hand, Asp25 interacted with hydroxyl oxygen of core structure whereas Ile50' attracted towards oxygen of sulfonamide near R3 substituent.</p><p>The terminal methoxy oxygen at R1 mediated interactions with the wild type protein's amino group of Asp29 and Asp30 with the distance of 2.29Å and 1.7Å, respectively. The interactions of compound 46 with these residues were lost upon mutation (Figure 6g-h). The binding orientations of compound 32 and 46 (Figure 6) revealed that compound 32 maintained its interactions with the active site residues in wild type as well as in mutated protein while compound 46 lost most of its binding interactions in mutated protein as shown in Table 6.</p><!><p>Protein-ligand binding interactions with specific conserved residues</p><p>Note: Distances of important interactions are shown in Å, however, all interactions mentioned here having distances of less than 3Å.</p><!><p>In the present work, comparison of ligand and structure-based 3D-QSAR using CoMFA and CoMSIA were derived for HIV-1 protease inhibitors. The statistics of both models were convincing and comparable. The model was significantly favored by internal and external predictions as well as visualization of contour maps. The effect of important structural characteristic of the potent inhibitor was predicted by the generated model. From the predictions, it was evident that at R1 position electronegativity is favored due to presence of Asp29 in its vicinity and hydrophobicity is disfavored which is relevant with the presence of methyloxazolidione ring in compound 32. Docking results also showed that terminal methoxy oxygen at R1 mediated bidentate interactions with the amino group of Asp29 and Asp30 which was lost in compound 46. At R2 position, bulkiness is disfavored whereas at R3; hydrophobicity is favored which is evident by presence of methoxy phenyl in compound 32. The docking studies of most potent and least active inhibitors further verified the generated 3D-QSAR models and can be used as guidance for better drug development.</p><!><p>The dataset of 102 compounds was retrieved from literature reported by Jorissen R.N. et al., [7] and available in Additional file 1. 2D structures were drawn by Chem-Draw [19] and converted into 3D by MOE (Molecular Operating Environment) program [20]. The biological activities of all compounds were shown in Table 1 along with its negative logarithmic units, pIC50 values. Stereochemistry and atom typing were confirmed for each compound. Three different charges i.e., GH, AM1BCC and MMFF94 were applied to the dataset and all three sets were subjected to the database alignment by using sybyl7.3 [16]. The database alignment is depicted in Figure 1. The core structure of most active compound 32 (pIC50 = 12.10) was used as a template for alignment [21] in ligand-based QSAR. On the other hand, for structure-based QSAR, bound conformation of original compound was used as template for alignment.</p><!><p>The dataset of 102 compounds were segregated into training and test sets containing 78 and 24 compounds, respectively (Tables 1 and 2). Each set was constructed on basis of regular distribution of biological activities (Table 1). Comparative Molecular Field Analysis (CoMFA) and Comparative Molecular Similarity Indices Analysis (CoMSIA) with 2Å grid spacing, sp3 carbon probe atom with a charge of +1 and VdW radius of 1.52Å was used to calculate steric and electrostatic field descriptors. In order to reduce noise and improve efficiency, column filtering of 2.0 kcal mol-1 was used [16]. A default cutoff of 30 kcal mol-1 was used for field energy calculations. Subsequently partial least square (PLS) analysis was performed to obtain 3D-QSAR model.</p><p>The optimal number of components was determined by leave-one-out procedure (Cross validation) to build the statistical significant regression model. The quality of the model was judged by cross-validated coefficient q2 which should not be less than 0.5. The external predictivity was calculated by conventional correlation coefficient r2[22,23].</p><!><p>The dataset of 102 compounds was subjected to docking in order to validate the QSAR results via GOLD docking suit [12]. The emphasis was totally on most active and the least active compounds to evaluate their quality of interaction as HIV-1 protease inhibitors. For docking, wild type (PDB: 3EKV) [24], and mutated I84V (PDB: 3NU9) [25] proteins were retrieved from Protein Data Bank (PDB) [26] in order to check the consistency of ligand's interactions even if mutated viral attack is present.</p><p>The cognate ligand and water molecules were removed, and polar hydrogens were added. Software was validated by re-docking and root mean square deviation (RMSD) calculations shown in Table 5 and Figure 5. Default GOLD docking parameters were used with Gold-score and Chem-score as scoring and rescoring functions. For each ligand, ten docked poses were saved and analyzed.</p><!><p>3D-QSAR: 3-dimentional quantitative structure-activity relationship; HIV: Human immunodeficiency virus; AIDs: Acquired immunodeficiency syndrome; FDA: Food and drug administration; HAART: Highly active antiretroviral therapy; RT: Reverse transcriptase; PIs: Protease inhibitors; IN: Integrase; DRV: Darunavir; CoMFA: Comparative molecular field analysis; CoMSIA: Comparative molecular similarity index analysis; GOLD: Genetic optimization for ligand docking; MOE: Molecular operating environment; PLS: Partial least square; RMSD: Root mean square deviation.</p><!><p>The authors declared no competing interests.</p><!><p>ZQ supervised, conceived and guided the whole project and the manuscript. SU and HS carried out the work and drafted the manuscript with UM and SAH. All authors have read and approved the final manuscript.</p><!><p>Darunavir derivatives with all sibstitutions. Core structure of darunavir with positions marked for substitutions and structures of substituents at R1, R2 and R3 positions along with their experimental inhibitory activities.</p><!><p>Click here for file</p>

PubMed Open Access

Oxygen tolerant RAFT polymerisation initiated by living bacteria

Living organisms can synthesize a wide range of macromolecules from a small set of natural building blocks, yet there is potential for even greater materials diversity by exploiting biochemical processes to convert unnatural feedstocks into new abiotic polymers. Ultimately the synthesis of these polymers in situ might aid the coupling of organisms with synthetic matrices, and the generation of biohybrids or engineered living materials. The key step in biohybrid materials preparation is to harness the relevant biological pathways to produce synthetic polymers with predictable molar masses and defined architectures under ambient conditions. Accordingly, we report an aqueous, oxygen-tolerant RAFT polymerization platform based on a modified Fenton reaction which is initiated by Cupriavidus metallidurans CH34, a bacterial species with iron reducing capabilities. We show the synthesis of a range of water-soluble polymers under normoxic conditions, with control over the molar mass distribution, and also the production of block copolymer nanoparticles via polymerization-induced selfassembly. Finally, we highlight the benefits of using a bacterial initiation system by recycling the cells for multiple polymerisations. Overall, our method represents a highly versatile approach to producing well-defined polymeric materials within a hybrid natural-synthetic polymerization platform and in engineered living materials with properties beyond those of biotic macromolecules.

oxygen_tolerant_raft_polymerisation_initiated_by_living_bacteria

Introduction<!>Results and Discussion<!>C.met<!>Conclusions

<p>Nature exploits a vast array of biological pathways to produce biotic macromolecules (polysaccharides, proteins, DNA, RNA etc.) derived from a small subset of monomers (e.g. sugars, amino acids, nucleobases etc.). In contrast, the chemical industry has made available an enormous stock of monomers, particularly those with reactive double bonds, to provide routes to an almost limitless set of abiotic macromolecules. Polymers derived from vinylic or acrylic functionality have found use in biomedicine, 1, 2 and as energy, 3 and information storage materials. 4,5 Combining biosynthetic pathways with abiotic monomers could therefore generate an even greater diversity of materials and, if conducted in the presence of an organism with appropriate biochemical functionality, allow hybrid synthetic/natural interfaces and engineered living materials (ELMs) to be formed. Cellular metabolism is underpinned by electron transport via redox pathways. We and others have shown that these pathways can be used in cell-activated polymerization. [6][7][8][9][10][11] Prior reports have focused on the metal reducing activity of bacteria (e.g. E. coli, C. metallidurans, S. oneidensis) to mediate the active and dormant states of copper, iron and other metallic catalysts for atom transfer radical polymerizations (ATRP). [6][7][8][9]11 However, ATRP suffers a disadvantage of requiring careful tuning of the concentrations of bacteria and metal complexes to control the balance of growing and dormant chains for desirable kinetics and molar mass distribution. 12 In contrast, RAFT polymerization, which is a chain-transfer agent mediated polymerization, requires instead a constant flux of external radicals. In many biological environments, a source of radicals is readily available, thus RAFT might be inherently easier to control than cell instructed ATRP, which is adversely affected by alternate indirect initiation pathways from bacterial cultures. 13 Whilst it has been shown that the generic reducing environment of bacteria can be used to produce organic radicals from the reduction of an aryl diazonium salt, which initiates the RAFT process, 10 this has been achieved so far only under anoxic conditions, hindering translation to biological applications. Conversely, many oxygen tolerant RAFT polymerisations have been reported, 14 either by polymerizing directly through oxygen [15][16][17] or utilising a scavenger such as an enzyme [18][19][20] or oxygen trap [21][22][23][24][25] , which has enabled ultralow reaction volumes, 17,19,22 3D/4D printing 21,26 and high throughput platforms, 22 but to the best of our knowledge have not been applied in a bacterially initiated RAFT polymerisation.</p><p>Accordingly, in this study, we present a new oxygen tolerant bacteria-initiated RAFT polymerization, by utilizing an adapted Fenton polymerisation. 27,28 Ourapproach harnesses the substantially faster reaction rate (4-5 orders of magnitude) between hydrogen peroxide and Fe 2+ than with Fe 3+ to produce hydroxyl radicals to mediate the RAFT process. While a typical Fenton polymerization procedure directly implements Fe 2+ to avoid this, we postulated that we could use the Fe 3+ reducing capabilities of C. metallidurans CH34 metabolism, which instructs the in situ formation of Fe 2+ , and accelerate the formation of hydroxyl radicals to initiate the RAFT process. To achieve oxygen tolerance, we were inspired by previous studies which utilized glucose oxidase (GOx) to deoxygenate transiently the reaction media from a glucose feedstock. 18,19 This approach provided a dual benefit, as a key byproduct from GOx deoxygenation is hydrogen peroxide which could be fed into our bacterially instructed Fenton reaction (Scheme 1). 30 Using this approach, we report the optimization and mechanistic evaluation of our bacterially mediated Fenton polymerisation. We highlight this through the synthesis of a range of well-defined RAFT polymers and polymer nanoparticles in open-to-air vessels under aqueous conditions. Scheme 1. Fenton GOx RAFT process initiated by reducing agents: ascorbic acid (AscA) or bacteria. D-Glucose (DG) is converted to D-Glucanolactate (DGA) by glucose oxidase (GOx) which consumes O2 in the process to form H2O2. Without the presence of reducing agents, polymerisation should not take place. GOx protein image from PDB ID: 3QVP.A. 29</p><!><p>Before conducting our bacteria mediated Fenton RAFT polymerisations, we initially evaluated the viability of C. metallidurans CH34 cells in the presence of a range of water-soluble monomers to ensure any observable polymerization was not caused by cell lysis (Figure S1 and Table S1). Both N,N-dimethylacrylamide (DMA) and N-hydroxyethylacrylamide (HEA) exhibited an MIC50 above 100 mM. However, N-acryloyl morpholine (NAM) displayed some toxicity towards the bacterial cultures (MIC50 = 42.5 mM). As a result of this, a concentration of 25 mM NAM was employed as this ensured c. 70% bacterial viability, a similar viability was observed at a monomer concentration of 100 mM for DMA and HEA.</p><!><p>To test our bacteria instructed Fenton-RAFT hypothesis, we incubated a mixture of DMA monomer, carboxyethyl propanoic acid trithiocarbonate (CEPTC) water soluble RAFT agent, FeCl3 as the Fe 3+ source, glucose oxidase and glucose with a C. metallidurans culture (1.7 x 10 10 colony forming units (CFU) mL -1 ) in phosphate buffered saline (PBS) ([DMA]:[CTA]:[FeCl3]:[GOx]:[Glucose] = 200:1:5.3:0.002:0.8) and heated the suspension to 30°C in an open to air vessel under normoxic conditions for 24 h. Aside from its iron reducing properties, C. metallidurans lacks the glucose transporter, thus we deemed it unlikely that the bacterial cells were reducing the glucose concentration through metabolization. 31 Conducting the polymerisations in PBS instead of growth medium also mitigated the risk of incorporating additional reducing agents which may contribute to redox based radical initiation pathways. After removal of the bacteria and iron oxide precipitate, 1 H NMR spectroscopy confirmed the presence of polymer, with monomer conversion reaching 53% (Figure 1a). SEC analysis indicated a monomodal molecular weight distribution with low dispersity (Ð = 1.12) and low molar mass (Mn,SEC = 19,900 g mol -1 ) as is expected for RAFT polymerisation. Crucially, control experiments omitting FeCl3 or with C. metallidurans cultures which were heat killed (3.6 x 10 2 CFU mL -1 ) displayed no monomer conversion indicating the importance of metabolically active cells for successful polymerization (Table S2). Noticeably, reaction mixtures containing FeCl3 but in the absence of bacteria yielded a small level of polymerization (10% monomer conversion) which we suspect is due to the slower Fe 3+ mediated Fenton reaction, producing a low concentration of hydroxyl radicals which still contribute to conversion (Figure 1b, Figure S2. Polymerisations in the absence of CTA yielded substantially higher molar masses (Mn,SEC = 451,000 g mol -1 ) and high dispersity (Ð = 2.11) following a conventional free radical mechanism (Figure 1c). When hydroxyl radicals are generated from the bacterially produced Fe 2+ , Fe 3+ is regenerated during the Fenton reaction. We, therefore, postulated that the bacteria could recycle the available Fe 3+ for further Fenton polymerisations at a reduced FeCl3 concentration. Accordingly, the polyDMA produced in polymerisations conducted at 7 μM maintained narrow dispersities (Đ ~ 1.28, Figure 1d) and still achieved moderate monomer conversions (44%). Therewas an increasing trend correlating FeCl3 concentration with monomer conversion between 7 and 700 µM, reaching a maximum of 66.2%, also resulting in an increase in Ð from 1.28 to 1.49. All polymers had unimodal molar mass distributions with similar Mn,SEC to their Mn,th values (Figure 1e). Strikingly at 7 mM we observed a substantial reduction in monomer conversion to 9%, much broader molar mass distributions (Ð = 2.11) and Mn,SEC 50-fold higher than the Mn,th which is more consistent with free radical polymerization likely caused by excess oxidation of free RAFT agent and possible toxicity towards C. metallidurans. For this reason, we adopted Fe concentrations of 7 µM for the remaining experiments.</p><p>Bacteria assisted Fenton RAFT polymerizations with HEA and NAM (conducted at 100 mM and 25 mM monomer solutions respectively) displayed similar monomer conversions to DMA (37 and 40% respectively), albeit with higher dispersities (Ð ~ 1.6 for both polymerisations, compared to 1.28 for DMA) (Figure 2a,Table S3, Figure S3). Although HEA polymers displayed moderately similar experimental and theoretical molar masses, the NAM analogues were 10-fold higher in molar mass than expected, attributed either due to the difference in monomer concentration or the poorer cell tolerability described above. To probe this, we performed a copolymerization produced a copolymer with similar experimental and theoretical molar masses and low dispersity (Ð = 1.21), suggesting this was due to the overall monomer concentration not NAM toxicity. We then investigated the polymerization kinetics of our bacteria-initiated RAFT polymerisation by sampling a DMA polymerization at 1 h, 2 h and 24 h, monitoring monomer conversion and Mn,SEC. Notably, we observed the polymerization did not proceed above 41% monomer conversion under these conditions (Figure 2b). This conversion is in line with other bacterial radical polymerization systems, 7,8,10 and we anticipate is due to the low initial monomer conversion, which quickly depletes retarding the ensuing polymerization reaction, compounded by the consumption of the glucose feedstock by GOx. Although a uniform molar mass distribution (Ð < 1.40) and retention of the trithiocarbonate was observed across all time points indicating contribution by the chain transfer agent (Figure S4b), only partial linear evolution between Mn,SEC and monomer conversion for RAFT polymerizations was observed, suggesting some RAFT characteristics.(Table S3 and (Figure 2c). This is supported by the first order kinetic plot (Figure S4), which indicates a fast linear reaction between 0 and 2 h, which then reached a plateau in terms of rats after 35% monomer conversion (Figure S4a). Although the relatively low monomer conversion of this polymerization is a potential limitation, the necessity for active metabolism and living cells to initiate polymerization, a notable difference compared to previous strategies, 10 means conversion is correlated to the tolerability of the chosen monomers.</p><p>One of the major advantages of RAFT polymerisations is the ability to prepare block copolymer nanoparticles with relative ease, 32 which have enormous potential in drug delivery 33 and other applications. 34 An extremely versatile route that has been explored for the last decade is the polymerization-induced self-assembly (PISA), enabling the preparation of well-defined nanoparticles in situ during the polymerization which can be conducted under completely aqueous conditions (Figure 2d). 35,36 Given the success of thisapproach and our encouraging results with bacteria-initiated solution polymerisations, we explored if we could utilize the methodology presented here to produce block copolymer nanoparticles via PISA. The pDMA75 mCTA was extended with a target 200 units of diacetone acrylamide, a monomer known to undergo PISA, [37][38][39] reaching quantitative monomer conversion as is expected in PISA due to the high local monomer concentration within the growing particles. Particle size analysis via both DLS and TEM indicates successful nanoparticle preparation with corroborative sizes between the two techniques (Figure 2e). However, due to the low concentrations used in our PISA reaction no molar mass information could be obtained from dried particles. The ability to produce nanoparticles using this system could in future offer the potential of biomimetic extracellular vesicles, which are achievable through PISA 40 which could for instance transport innate quorum sensing molecules. 41 One of the major advantages of utilising living systems to initiate chemical reactions or indeed polymerisations is their ability to be reused or expanded through culture to remove feedstock requirements, important for the sustainability of these processes. Hence, we subsequently investigated if the initial C. metallidurans culture could be recycled for several polymerization reactions by pelleting the cells through centrifugation and resuspension with a new polymerization mixture (Figure 3a). It was found that the initial bacterial culture could be reused at least three times using without supplementing with growth media or nutrients. Interestingly the monomer conversion and Mn,SEC was variable between each cycle at 40, 80 and 50 % for the three consecutive polymerisations and 18,800, 32,500 and 26,500 g mol -1 respectively, each with low (Ð ~ 1.3) in all cases. While further investigation is required to understand fully these differences, we anticipate that some bacterial proliferation or changes in bacterial metabolism may affect final conversion. (Figure 3b, Figure 3c). A similar phenomenon was reported by Keitz and co-workers for the bacteria mediated Cu(I)-catalysed azide-alkyne cycloaddition, where subsequent cycles yielded different reactionconversions to the first cycle, which they suggested was due to bacterial growth between cycles 1 and 2.</p><!><p>In conclusion, we have developed an oxygen tolerant bacterially initiated polymerization method which can be used to produce macromolecules with defined length via RAFT polymerisation. To achieve this, we utilized the reducing capabilities of C. metallidurans to produce Fe2+ in situ and a simultaneous glucose oxide catalysis pathway to generate hydrogen peroxide from a glucose feedstock, which then reacts to produce hydroxyl radicals and initiate polymerization. We found that high monomer conversion could only be achieved with actively metabolising bacteria and in the presence of Fe3+, supporting our proposed mechanism. Synthesized polymers exhibited the characteristics of conventional RAFT polymerisations such as narrow molecular weight distributions, retention of end-group fidelity and similar average molar masses, albeit with some limits in terms of blocking efficiencies. We exemplified this polymerization technique by utilising monomers known to undergo polymerization-induced self-assembly to produce bacterially synthesized polymer nanoparticles. Finally, we showcased the ability for the bacteria to be a reusable component for radical generation and thus polymerization. This microbial redox pathway to produce well defined polymers could open the potential for hybrid natural and non-natural material platforms and thus new engineered living materials.</p>

ChemRxiv

“Beating speckles” via electrically-induced vibrations of Au nanorods embedded in sol-gel

Generation of macroscopic phenomena through manipulating nano-scale properties of materials is among the most fundamental goals of nanotechnology research. We demonstrate cooperative ''speckle beats'' induced through electric-field modulation of gold (Au) nanorods embedded in a transparent sol-gel host. Specifically, we show that placing the Au nanorod/sol-gel matrix in an alternating current (AC) field gives rise to dramatic modulation of incident light scattered from the material. The speckle light patterns take form of ''beats'', for which the amplitude and frequency are directly correlated with the voltage and frequency, respectively, of the applied AC field. The data indicate that the speckle beats arise from localized vibrations of the gel-embedded Au nanorods, induced through the interactions between the AC field and the electrostatically-charged nanorods. This phenomenon opens the way for new means of investigating nanoparticles in constrained environments. Applications in electro-optical devices, such as optical modulators, movable lenses, and others are also envisaged.

“beating_speckles”_via_electrically-induced_vibrations_of_au_nanorods_embedded_in_sol-gel

<!>Results<!>Discussion<!>Methods

<p>A mong the basic goals in nanotechnology research has been the development of systems in which macroscopic phenomena are induced and modulated by the intrinsic properties of nano-scale assemblies. The ''nanoscale-macroscale'' relationship is apparent in many processes and fundamental applications, including piezoelectricity, magnetism, heat conductivity in solids, and also in more specialized phenomena such as quantum coherence. This generic relationship has been particularly encountered in coupled photonic/electronic devices [1][2][3][4][5] . For example, an interesting new concept for realizing nanometric photonic devices on silicon chips focuses on implanting a single gold nanoparticle (Au NP) along the silicon waveguide while externally controlling its photonic tunability 6,7 . In another recent demonstration, realization of optically reconfigurable properties was obtained by illumination of metal nanorod assemblies 8 .</p><p>In this study, we exploit electro-optic modulation on the nanoscale to form a novel macroscopic photonic modulation concept, based upon encapsulation of Au nanorods in a sol-gel host matrix. Sol-gel materials constitute a unique class of porous and transparent host materials which can readily embed a broad range of chemical and biological guest molecules. In particular, numerous experiments have utilized the transparency of sol-gels, making possible the application of diverse spectroscopic and optical techniques for analysis of gelembedded guest species 9,10 . The porous gel framework also enables co-encapsulation of solvent molecules, thus allowing mobility of embedded solute molecules and particles 11 .</p><p>Nanoparticles (NPs) were incorporated within sol-gel host matrixes and varied reports underscore both the scientific and practical potential of such composite systems 12,13 . Au NP/sol-gel architectures, in particular, have been employed in sensing 14,15 , controlled release 16,17 , and other applications. Utilization of Au NP/sol-gel systems as photonic or electronic conduits, however, has not been reported. Here we report a unique electro-optic phenomenon, directly linked to encapsulation of Au nanorods within sol-gel pores. Specifically, we observed light ''speckle beats'' induced through placing the Au nanorod/sol-gel assembly in an alternating-current (AC) electric field while simultaneously illuminating the sample with a laser beam. The concept we present, besides being a demonstration of interlinked nano-macro phenomena, opens the way to applications in photonic devices, biological imaging, nanostructure characterization, and others.</p><!><p>The experimental setup is depicted in Figure 1. Au nanorods, synthesized through conventional protocols 18 were interspersed with the silica precursors of the sol-gel matrix, producing after gelation a transparent sol-gel matrix in which the Au nanorods were embedded within the porous framework. Repeated washing cycles did not remove the nanorods from within the sol-gel matrix, confirming their encapsulation within the gel. Following co-assembly, the Au nanorod/sol-gel construct was placed between two electrodes producing an alternating current (AC) electric field for which both the voltage and frequency could be externally modulated. Simultaneously with application of the AC field, the nanorod/sol-gel sample was also illuminated by a focused monochromatic laser beam (l5532 nm) and the scattered light was recorded by a conventional chargecoupled device (CCD) camera.</p><p>Figure 1B presents a picture of the Au nanorod/sol-gel electrode setup, showing the sol-gel droplet between the two electrodes. The distance between the electrodes could be engineered to facilitate optimal power input and usage of minute sample volumes (microliter range). Importantly, the experiments were carried out before complete dehydration of the sol-gel, thus motion of the Au nanorods was still enabled within the aqueous solution trapped in the porous sol-gel framework 19 .</p><p>Application of the AC field gave rise to dramatic modulation of the scattered light. Figure 2 depicts still images of a small region within the speckle pattern, extracted from a video recorded using the setup depicted in Figure 1 (the representative video is provided in the Supplementary Information file). Specifically, Figure 2 shows visible ''beats'' of the monochromatic light scattered from the gel/nanorod assembly (e.g. ''speckles'' 20 ). Note in particular the intensity modulation of the speckle pattern shown in Figure 2 indicated by the white rectangle in Figure 2; the beat frequency was precisely correlated with the applied AC.</p><p>We carried out a temporal analysis of the recorded images of the back-reflected speckle patterns, such as shown in Figure 2. Specifically, we determined through simple image analysis both the frequency of speckle motion, as well as the amplitude span of the speckle beats (Figure 3). Significantly, Figure 3 demonstrates that the amplitude of the speckle vibration (i.e. difference between minimum and maximum speckle intensities) was dependent upon the AC voltage amplitude (Figure 3A), while the frequency of the speckle beats was directly correlated to the AC periods (Figure 3B). These relationships establish a direct link between the light modulation and the AC field.</p><p>A particularly striking result is the observation of higher-order harmonics upon Fourier transformation of the speckle beats (Figure 3C). Importantly, the distinct colors in Figure 3C correspond to different pixels in the pulse area (e.g. analysis of different speckles), demonstrating the universal nature of this phenomenon. The observation of higher harmonics for the beats of the speckle patterns is ascribed to the sinusoidal dependence of nanorod motion within the constrained sol-gel environment, according to the model introduced by Coussot et al. [21][22][23] . Specifically, according to the model:</p><p>Where g is the viscosity, t and _ c are the shear stress and shear rate magnitudes, respectively, depending on the current flocculation state (l) of the material (e.g. degree of ''jamming''). The value of l is determined by solving the differential equation 24,25 :</p><p>In which h is the flocculation characteristic time and a is the material parameter. Thus, since the shear stress t is proportional to the applied force (in our case it is a sinusoidal force induced by the AC field with radial frequency of v 0 ), the shear rate _ c as well as the viscosity are depicted as a nonlinear function of the applied sinusoidal force, e.g. includes higher harmonics than the radial frequency of the applied sinusoidal force. Mathematically, using the above relations one obtains:</p><p>In which t 0 is the amplitude of the applied sinusoidal shear stress (proportional to the applied force) and</p><p>The significance of the analysis introduced by Coussot et al. underscoring the generation of a higher order harmonic peak -is the establishment of a clear quantitative link between the modulated scattered light, the physico-chemical characteristics of the Au nanorod/sol-gel system (e.g. motion of the Au nanorods in the constrained gel environment), and the electric field parameters. Indeed, the theory we present here for the generation of the higher harmonic in the temporal Fourier domain conforms the experimental results in Figure 2C depicting harmonics at frequencies which are multiples of the frequency of the stimulating electrical AC field.</p><p>It should be noted that other models have been discussed in the literature, characterizing non linear effects in viscous solutions, particularly focused on rheology 26 . The study described in ref. [26], however, analyzes Laponite suspensions which exhibit much lower viscosity than the sol-gel system presented here. This distinction is important, since the model outlined by Coussot et al. is more pertinent to high viscosity systems such as sol-gels. Nevertheless, the intriguing phenomena we report echo the observations that non linear rheology effects are apparent also in lower viscosity liquids, forming the basis to the optical modulation shown in Figures 2 and 3.</p><p>Additional experiments were carried out to confirm that the ''speckle beats'' were directly linked to the motion of Au nanorods within the sol-gel construct induced be the AC field. Specifically, speckle modulation was not observed when an ''empty'' sol-gel (not containing Au nanorods) was placed in an AC field and illuminated by the laser beam. Similarly, no speckle beats were observed when a direct current (DC) field was applied, nor when Au nanorods in an aqueous solution (not inside a sol-gel) or in a dried film were placed in the AC field (data not shown). These observations and the mathematical analysis of the frequency modulation outlined above underscore the correlation between the speckle beating phenomenon and applied AC field. This correlation is ascribed to the electrostatic-charge of the Au nanorod surface -as they are coated by a bilayer of cetyltrimethylammonium bromide (CTAB), which exhibits partial positivelycharged headgroups 27 . Consequently, the Au nanorods interact with the externally-applied alternating electrical field. It is important to note that encapsulation of the nanorods within the porous gel framework results in restricted mobility, leading to localized vibrations with amplitude and periodicity that are determined by the AC field. These vibrations are manifested through the spatial-temporal analysis of the back-reflected optical speckle patterns depicted in Figure 3. Indeed, the ensemble of embedded Au nanorods moving in tandem with the applied AC field is the fundamental driving force generating the observed speckle pattern beats.</p><p>To further support the above description we carried out in situ environmental scanning electron microscopy (ESEM) experiments designed to visualize the mobility of the gel-embedded Au nanorods while an externally-modulated AC field is applied (Figure 4). Figure 4A depicts still ESEM images of encapsulated Au nanorods, recorded in progressive time-points upon application of the sinusoidal AC field (the complete video file is provided in the Supporting Information). As apparent in Figure 4A (and the video file in the Supporting Information), the frequency of the vibration motion of the nanorods echoes the applied AC frequency.</p><p>To quantitatively verify the correlation between the nanorod motion recorded in the ESEM experiment and the applied AC field we carried out a Fourier transformation over the temporal changes visible in the spatial region of interest (Figure 4B), similar to the procedure described above for the optical speckle images (Figure 3C). Importantly, the Fourier transformation yielded a second harmonic frequency peak appearing in the temporal domain, confirming that the AC-induced gold nanorod vibrations recorded in the ESEM experiment indeed correspond to nanorod motion, rather than vibrations of the electron beam, which might be induced by the electric field.</p><p>It should be emphasized that the proposed model does not make a direct link between the movements of the gold NPs and the eventual speckle formation due to the ensemble of moving particles. While we have previously established a direct relationship between backreflected speckle patterns and positions of the reflecting particles 28 , the model outlined above is still suggestive and not yet concrete.</p><!><p>This work demonstrates a unique phenomenon in which ''beats'' of optical speckle patterns are introduced through illuminating a Au nanorod/sol-gel assembly placed within an AC electric field. This electrical/optical effect is due to interactions between the applied field and the electrostatically-charged Au nanorods, embedded in the constrained environment of the sol-gel pores. The presented results underscore a unique macroscopic physical property (visible vibrations of the speckle pattern) that is directly traced to a nanoscale material organization (Au nanorods embedded in sol-gel pores).</p><p>The data presented point to direct utilization of the Au NP/sol-gel system as an optical modulator. These applications have attracted significant scientific and technological interest and several reports depicted varied liquid crystal (LC)-based modulator designs 29 . Although LC modulators exhibit significant electro-optic effects, they are significantly smaller than the modulation effect that can be obtained with the Au NPs, since in liquid crystals the electrical field modifies the refraction index while in the system depicted here modulation is based upon NP mobility. Indeed, we previously showed that transversal movement of only 10 nm is sufficient to obtain modulation with extinction ratio of above 10 dB for axial interaction of about 20 microns 6 . This is not obtainable with LCs.</p><p>An important parameter concerns the concentration of the Au NPs for which the optical phenomenon arises. As one of the important applications we envision are nano photonic modulators, the minimal concentration would constitute a single nano particle within the cross section of the optical mode multiplied by axial distance of about 20 microns. In case of silicon waveguides this corresponds to a volume of around 200 nm3400 nm320,000 nm, which is 1.6 (mm 3 ) 6 .</p><p>The sol-gel modulator exhibits notable advantages as a possible photonic platform. The sol-gel matrix constitutes a stable, rigid scaffold which allows high repeatability of the modulation effect. The Au nanorod/sol-gel assembly is furthermore simple to manufacture and, as demonstrated above, requires relatively low modulation voltages even if the electrodes are positioned quite far from the waveguide. In addition, although a sol-gel based waveguide would exhibit a lower refractive index 30 , it displays much lower evanescent tails thereby resulting in smaller attenuation from the external electrodes. Furthermore, the Au NP/sol-gel concept might be also used for molecular imaging since via the temporal profile of the mechanical movement of the NPs one could identify the specific type of tissue even without actually imaging it, as e.g. done in conventional imaging approaches.</p><!><p>Materials. Au(III) chloride trihydrate ($99.9%), trisodium citrate, sodium borohydride (98%), L-ascorbic acid (99%), tetramethyl orthosilicate, TMOS, (98%), and trizma base were purchased from Sigma-Aldrich. Cetyltrimethylamonium bromide, CTAB, (98%) was received from Alfa Aesar (USA). Silver nitrate was purchased from Metalor Technologies Ltd (UK). Hydrochloric acid (HCl) was purchased from Gadot Biochemical Industries Ltd (Israel). All chemicals were used as received. Water used in all experiments was de-ionized up to 18.2 MV cm resistivity (Barnstead/Thermolyne Corporation, Dubuque, IA).</p><p>Au nanorod synthesis. Positively-charged Au nanorods were synthesized according to the two-step seed-mediated growth method described by Nikoobakht and El-Sayed. First, seed solution was prepared by mixing 5 mL of 0.2 M CTAB solution with 5.0 mL of 0.0005 M chlorauric acid (HAuCl 4 ). To the stirred solution, 0.60 mL of icecold 0.010 M sodium borohydride (NaBH 4 ) was added, resulted in the formation of a brownish yellow solution. Vigorous stirring of the seed solution was continued for several minutes. Following stirring, the solution was kept at room temperature. Growth solution was prepared by mixing 5 mL of 0.2 M CTAB solution with 0.2 mL of 0.0040 M silver nitrate (AgNO 3 ) solution at room temperature. 5.0 mL of 0.0010 M HAuCl 4 was subsequently added, and after gentle mixing 70 mL of 0.0788 M ascorbic acid was added. As a result the growth solution changed its color from dark yellow to colorless. The final step was the addition of 12 mL of the seed solution to the growth solution at 30uC. The color of the solution gradually changed within 50 min. In order to separate nanorods from excess CTAB, the solution was centrifuged at 13000 rpm for 10 min. The supernatant was removed and the nanorods were re-suspended in appropriate volume of double de-ionized water (DDW). The washing process was repeated 3 times.</p><p>Au nanorod/sol-gel preparation. The sol-gel precursor solution was prepared by mixing 4.41 ml TMOS, 2.16 ml DDW, and 0.06 ml HCl (0.62 M) at 4uC for 1 h. The resulted prehydrolyzed TMOS sol was kept at 220uC before use. For preparing the composite Au nanorod/sol-gel sample, equal amounts of the prehydrolyzed sol and Au nanorod solution were mixed together. Subsequently, Tris buffer (50 mM, pH58) was added to the nanorods/sol solution in a 151 volume ratio at room temperature. Gelation occurred within minutes after buffer addition, accordingly all measurements were carried out immediately after sample preparation.</p><p>Transmission electron microscopy (TEM). 10 mL of the nanorod solution was applied to 400-mesh copper formvar/carbon grid (Electron Microscope Sciences, Hatfield, PA, USA), and then allowed to dry in air. Afterwards, Au nanorod samples were characterized using 200 kV JEOL JEM-2100F Transmission electron microscope (USA).</p><p>Electro-optical experimental setup. Arrays of gold electrodes on silicon or glass surfaces (electrode spacing 500 mm) were fabricated using the appropriate shadow masks and gold evaporation, followed by welding low-noise electrical cables to the electrodes. The light modulation experiment was videoed under microscope while a green laser at wavelength of 532 nm and power of 50 mW illuminated the sample.</p><p>In-situ environmental scanning electron microscopy (ESEM). Experiments were carried out on an FEI Quanta 200 field emission gun (FEG) environmental SEM (ESEM). The samples in the ESEM were mounted on a Peltier cooling stage, which was externally water-cooled. This allowed imaging the samples at wet mode (initial conditions typically were 2uC and 5.3 torr), while applying external AC electric fields on the sample electrodes via chamber feed-through leads connected to an external voltage supplier. Video recording was done in a frequency of 10 Hz, which was greater than the electric field frequency (0.2 Hz) in order to avoid coupling between the ESEM scanning and the external electric field. Data analysis. Temporal light beat analysis was carried out through recording the speckle pattern reflected from the Au nanorod/sol gel, and then choosing the spatial region of interest. The temporal information (flickering of the speckle) was Fourier transformed. In the spectrum we analyzed both the temporal frequency at which the flickering occurred as well as the amplitude at which the speckle patterns were flickering.</p>

Scientific Reports - Nature

Transcriptome-based reconstructions from the murine knockout suggest involvement of the urate transporter, URAT1 (slc22a12), in novel metabolic pathways

URAT1 (slc22a12) was identified as the transporter responsible for renal reabsorption of the medically important compound, uric acid. However, subsequent studies have indicated that other transporters make contributions to this process, and that URAT1 transports other organic anions besides urate (including several in common with the closely related multi-specific renal organic anion transporters, OAT1 (slc22a6) and OAT3 (slc22a8)). These findings raise the possibility that urate transport is not the sole physiological function of URAT1. We previously characterized mice null for the murine ortholog of URAT1 (mURAT1; previously cloned as RST), finding a relatively modest decrement in urate reabsorptive capacity. Nevertheless, there were shifts in the plasma and urinary concentrations of multiple small molecules, suggesting significant metabolic changes in the knockouts. Although these molecules remain unidentified, here we have computationally delineated the biochemical networks consistent with transcriptomic data from the null mice. These analyses suggest alterations in the handling of not only urate but also other putative URAT1 substrates comprising intermediates in nucleotide, carbohydrate, and steroid metabolism. Moreover, the analyses indicate changes in multiple other pathways, including those relating to the metabolism of glycosaminoglycans, methionine, and coenzyme A, possibly reflecting downstream effects of URAT1 loss. Taken together with the available substrate and metabolomic data for the other OATs, our findings suggest that the transport and biochemical functions of URAT1 overlap those of OAT1 and OAT3, and could contribute to our understanding of the relationship between uric acid and the various metabolic disorders to which it has been linked.

transcriptome-based_reconstructions_from_the_murine_knockout_suggest_involvement_of_the_urate_transp

INTRODUCTION<!>Gene expression determinations<!>Determination of enriched functions in differentially expressed or variable genes<!>Gene expression-based metabolic network reconstruction<!>Functional annotations enriched among genes differentially expressed or variable in the mURAT1 knockout<!>Comparison of metabolic reconstructions derived from wild-type and mURAT1 knockout renal transcriptomes<!>DISCUSSION

<p>The renal transporter, URAT1 [1], encoded by the human ortholog of the previously identified mouse gene, RST [2], was identified as the principal agent of proximal tubular reabsorption of uric acid, a medically important metabolite linked to renal and cardiovascular dysfunction as well as gout [3,4,5]. However, recent biochemical as well as genetic evidence indicates that several other transporters, including, prominently, GLUT9 [6,7,8,9], as well as ABCG2, NPT1, NPT4, and MRP4, make major contributions to the renal handling of urate (reviewed in [10,11,12]). Moreover, URAT1 transports not only urate, but also multiple other compounds (including endogenous metabolites such as acetoacetate, lactate, and orotate), a number of which are also substrates of the closely related multispecific renal organic anion transporters, OAT1 and OAT3 [13,14]; additionally, URAT1, like OAT1 and OAT3, is an exchanger that can mediate the bidirectional transport of its substrates [1,15]. Collectively, these findings raise the possibility that URAT1 might have significant functions distinct from urate reabsorption.</p><p>RST was established as the murine homolog of URAT1, mURAT1, on the basis of sequence homology [14,16] and functional similarity [15]. We characterized mice null for mURAT1, finding, consisted with the studies noted above, a relatively modest loss of urate reabsorptive capacity [17]. We subsequently performed metabolomic analyses of plasma and urine in these knockouts. Targeted measurements of the most abundant small organic anions did not reveal any differences between wild-type and mURAT1 knockout mice. However, global (untargeted) metabolomic measurements revealed significant changes in the plasma and urine concentrations of multiple unknown small molecules, suggesting an altered metabolite profile in the knockouts [17].</p><p>Although the latter molecules have not yet been identified, here we have computationally defined the biochemical networks consistent with transcriptomic data from the null mice.. Computational reconstruction of metabolic networks based on the constraints imposed by extant patterns of gene expression have been demonstrated to have predictive value in a variety of model systems (reviewed in [18]). We previously implemented such approaches to identify metabolic networks associated with OAT1 and its near paralog, OAT3, via analyses of the corresponding knockout mice. These studies implicated OAT function in multiple metabolic pathways, including the pentose phosphate shunt, the Krebs cycle, and the polyamine pathway (OAT1; [18]), xenobiotic hydroxylation and glucuronidation, the metabolism of flavonoids and other products of the gut microbiome, and prostaglandin, cyclic nucleotide, and glycosaminoglycan metabolism (OAT3; [19]), findings generally consistent with the transport and metabolomic data for OAT1 and OAT3. In order to better understand the role of URAT1 beyond urate transport, we have now performed analogous global metabolic reconstructions based on transcriptomic data from mURAT1 knockout mice. Our findings suggest the involvement of mURAT1 in metabolic processes overlapping those associated with OAT1 and OAT3, including pathways relating to carbohydrate, nucleotide, glycosaminoglycan, and coenzyme A metabolism, among others.</p><!><p>The generation and breeding of mURAT1 knockout mice and the collection of microarray data on renal gene expression in mUrat1 and the corresponding wild-type mice was previously reported (Eraly* et al., 2008). In brief, RST-null mice were generated by homologous recombination and then back-crossed to C57BL/6J mice for seven generations to yield the progenitors from which experimental animals (wild-type as well as knockout) were descended. Quantitative PCR analysis of the expression of mOat1, mOat3, and mUrat1 was performed as follows: RNA from wild-type and mUrat1 knockout kidneys (n = 3 per group) was purified on RNeasy columns (Qiagen, Valencia, CA), reverse transcribed using SuperScript III (Invitrogen, Carlsbad, CA), and the resulting cDNA samples was subjected to duplicate real-time PCR reactions at the University of California, San Diego/Veterans Affairs Medical Center's (UCSD/VAMC) Center for AIDS Research Genomics Core laboratory. Gene expression values were normalized to that of GAPDH in the corresponding cDNA samples. Gene-specific primer sequences (5′ to 3′) were as follows [please note that the first 18 bases (ACT GAA CCT GAC CGT ACA) on each forward primer correspond to the "Z sequence" that is complementary to the "uniprimer" used in the Amplifluor system]: mOat1 (mSlc22a6), ACT GAA CCT GAC CGT ACA GCA TGA CTG CCG AGT TCT ACC (forward) and CAG CGC CGA AGA TGA AGA G (reverse); mOat3 (mSlc22a8), ACT GAA CCT GAC CGT ACA GCA GCC CTT CAT CCC TAA TG (forward) and CCT CCC AGT AGA GTC ATG GTC AC (reverse); and mUrat1 (mSlc22a12), ACT GAA CCT GAC CGT ACA CCA TGC TAG GGC CTT TGG TA (forward) and GCA TCC AGG AGC CAT AGA CAC (reverse).</p><!><p>Renal gene expression patterns in wild-type and mURAT1 knockout mice as determined by microarray analysis were compared using VAMPIRE (http://genome.ucsd.edu/microarray; accessed February 11th, 2015), which takes variability differences between genotypes (which can be substantial (Eraly, 2014)) into account in determining the statistical significance of any gene expression differences (Hsiao et al., 2005, Hsiao et al., 2004). Differentially expressed genes were identified as those meeting the statistical significance threshold of p<0.05 following Bonferroni correction for multiple comparisons. Functional annotations enriched among differentially expressed genes were determined using the AmiGO tool (http://amigo1.geneontology.org/; accessed February 11th, 2015), with the p value threshold set at 10-5 and the background set drawn from the Mouse Genome Informatics database (http://informatics.jax.org; accessed February 11th, 2015).</p><p>There appear to be, generally, highly significant differences between knockout and wild-type mice in mean gene expression variability (Eraly, 2014). Specifically, in each of multiple microarray comparisons of gene expression in wild-type and diverse knockout mice, the mean log coefficient of variation (CV) ratio (the mean, for the various measured genes, of the logs of the ratios of CV of expression in knockout to CV in wild-type – equivalent to calculating the geometric mean of the CV ratio; please see (Eraly, 2014) for the rationale for this procedure) was found to typically deviate highly significantly from zero, the expected value if there were no differences between knockout and wild-type mice in gene expression variability. Moreover, the distribution of the log CV ratios was found to be approximately Gaussian. As such, in the current study, we considered genes to be differentially variable if they fell within the top or bottom 1% of the expected distribution of the log CV ratios; i.e., had variability differences significantly greater or lesser than the mean variability difference between mURAT1 knockout and wild-type mice. Functional annotations enriched among these differentially variable genes were also determined using AmiGO, as described above for differentially expressed genes.</p><!><p>Microarray gene expression detection p-values were mapped to their corresponding reactions in the genome scale computational metabolic reconstruction, iMM1415 ((Sigurdsson et al., 2010), the murine version of the previously developed human metabolic reconstruction, Recon1 (Duarte et al., 2007)), based on gene-protein-reaction associations (Thiele and Palsson, 2010). The Gene Inactivity Moderated by Metabolism and Expression algorithm (GIMME; (Becker and Palsson, 2008)) provides summations of the most probable network flux states consistent with actively expressed genes and capable of achieving defined objective functions, thereby permitting quantification of the consistency of gene expression data with various metabolic objectives. In this study biomass was defined as the objective function, and the GIMME algorithm was used to generate the metabolic reconstructions most consistent with the wild-type and mURAT1 knockout mouse gene expression data, as previously described (Wu et al., 2013). Renal-specific uptake and secretion exchange constraints (as previously used to analyze blood pressure regulation (Chang et al., 2010)) were used across all conditions, so that calculated differences were only a function of changes in gene expression profiles. Each network reaction was then set as a required metabolic objective and the range of achievable flux states, hereafter referred to as flux-span, was calculated, providing a measure of the likelihood that the corresponding reaction was functional. Reactions having flux-span increased or decreased by two fold or greater in the mURAT1 knockout relative to wild-type mice were determined, and the proportions of these increased or decreased reactions within each of the various metabolic subsystems of iMM1415 were calculated.</p><!><p>We previously analyzed renal gene expression in mURAT1 knockout mice [17] focusing on genes known at that time to be involved in urate metabolism and transport (including XDH, xanthine dehydrogenase; HGPRT, hypoxanthine guanine phosphoribosyl transferase; ADA, adenosine deaminase; AMPD2, adenosine monophosphate deaminase 2; PNP, purine nucleoside phosphorylase; UAT, urate transporter (galectin 9); MRP2 & 4, multidrug resistance proteins 2 & 4 (ATP-Binding Cassette Sub-Family C Members 2 & 4); NPT1, sodium-phosphate transport protein 1 (slc17a1); OAT1 and OAT3), and found no substantial changes in expression level. Examination of genes discovered since that prior report to contribute to the renal handling of urate (including glucose transporter type 9 (GLUT9; slc2a9), ATP-binding cassette, sub-family G, member 2 (ABCG2), organic anion transporter 10 (OAT10; organic cation transporter like 3; slc22a13), and sodium-phosphate transport protein 4 (NPT4; slc17a3); reviewed in [10,11,12]) also did not reveal any significant changes (Fig. 1A). (Note that while organic anion transporter 4 (OAT4; slc22a11) has also been implicated in renal urate transport, the murine genome appears to lack an ortholog for this transporter ([20].) Moreover, there was no significant difference between wild-type and mURAT1 knockout mice in the renal expression of the mURAT1-related transporters, OAT1 and OAT3, as determined by quantitative PCR (Fig. 1B).</p><p>However, there was differential expression in mURAT1 knockout mouse kidneys of genes involved in various aspects of metabolism, biosynthesis, cell cycle progression, apoptosis, and development, among other processes (Supplementary Table 1, [17]). Now that several studies have indicated that other proteins may be equally if not more important uric acid transporters than URAT1, and because a number of other URAT1 substrates have been identified, we used, as a first step, the gene ontology analysis tool AmiGO (http://amigo1.geneontology.org/; accessed February 11th, 2015) to quantify the enrichment of general functional annotations among the differentially expressed genes. We found the greatest enrichment to be for the terms "cellular metabolic process" (p=3.26×10−11) and "organic substance metabolic process" (p=1.03×10−9) (Fig. 2A).</p><p>We also employed an additional method based on recent data suggesting that differences in gene expression variability may contribute to phenotype independent of any changes in average gene expression (reviewed in [21]): we identified genes having significantly greater or lesser variability in the knockout compared to the wild-type – essentially, those at the extremes of the distribution of the gene by gene ratios of the coefficient of variation (CV) in knockout to CV in wild-type (please see the Methods). As with the differentially expressed genes, the differentially variable genes encompassed diverse functions, including those related to metabolism, cell cycle progression, growth factor signaling, and transcriptional regulation (Supplementary Table 2). Again as with the differentially expressed genes, though, the most significant enrichment was noted for the term "cellular metabolic process" (p=5.48×10−12) (Fig. 2B), providing additional support for the existence of metabolic alterations in the mURAT1 knockout, and thus for a consequential role for mURAT1 in cellular metabolism.</p><!><p>In order to obtain a finer-grained picture of the metabolic changes induced by deletion of mURAT1, we computationally defined the biochemical networks consistent with the transcriptomic data from the knockout mice. Specifically, we assembled and analyzed murine-specific biochemical network reconstructions based on global (microarray-derived transcriptomic) renal gene expression profiles in mURAT1 knockout and wild-type mice. Gene expression data were used as weighting constraints on network reactions, and the metabolite flux-spans, which may be considered to be measures of reaction functionality, were calculated for the various reactions as previously described ([18,19]; please also see Methods). Of the ~3400 reactions in the genome scale metabolic network, iMM1415 (please see Methods), 448 (13.18%) manifested some degree of alteration in the knockout compared to wild-type reconstructions (Supplementary Table 3). Among these, 102 had flux-spans increased by greater than two fold in the knockout relative to the wild-type (Table 1), and 12 had flux-spans decreased by greater than two fold (Table 2).</p><p>The proportions of these increased and decreased reactions in the various metabolic subsystems of the reconstructions were then determined (Fig. 3). (As an example, the chondroitin sulfate metabolism subsystem comprises 44 reactions, the flux-spans of 38 of which were increased by greater than two fold in the knockout compared to the wild-type, corresponding to a proportion of 0.86; Table 1 and Fig. 3A.) This analysis suggested significant increases in the mURAT1 knockout in metabolic functions related to glycosaminoglycan metabolism (aminosugar metabolism and chondroitin and keratan sulfate degradation), methionine metabolism, and lysosomal and extracellular transport (as indicated by a significant proportion of the reactions in these reconstructed sub-systems manifesting two fold or greater increase in flux-span in the knockout relative to wild-type; Table 1 and Fig. 3A). Conversely, the metabolic reconstructions featured significant decreases in functions related to coenzyme A biosynthesis (Table 2 and Fig. 3B).</p><p>Since mURAT1 is an extracellular membrane transporter, we also examined all extracellular transport reactions altered to any degree in the knockout reconstructions (and not just those having greater than two fold flux-span increases or decreases). There were 46 extracellular transport reactions, corresponding to 40 transported metabolites, having flux-span increases in the mURAT1 knockout relative to wild-type reconstructions (Table 3), and 3 reactions, corresponding to 3 metabolites, having flux-span decreases (Table 4). About a third of the compounds overall comprise small organic anions of the kind that typically (although not exclusively [18,19,22,23]) comprise substrates and inhibitors of the multispecific OATs (Tables 3 and 4). Alterations in the reconstructions of transport of the remaining compounds might reflect downstream consequences of the loss of URAT1.</p><p>In vitro data on the interactions of endogenous substrates with URAT1 provide experimental support for these analyses. Four of the compounds manifesting altered transport in the reconstructions have been demonstrated to interact with URAT1 in vitro; these comprise urate as well as 2-oxoglutarate (α-ketoglutarate), acetoacetate, and lactate [13]. Another two – pyruvate and progesterone – are known to interact with other members of the organic anion transporter (OAT) family of which URAT1 is a member [13,14,23]. Notably, nearly half (four of ten) of the URAT1-interacting compounds of endogenous origin that were listed in a comprehensive review [13] manifested altered transport in the knockout reconstructions (Table 5), supporting the validity of our analyses.</p><!><p>Multiple lines of evidence indicate an important role for URAT1 in the renal reabsorption of urate; in particular loss of function URAT1 mutations have been repeatedly associated with decreased urate reabsorption, resulting in hypouricemia, exercise-induced renal failure, nephrolithiasis, and as recently demonstrated, endothelial dysfunction [24,25,26,27,28]. However, as noted in the Introduction, recent data indicates that other transporters also make significant contributions to urate reabsorption [11,17,29,30] and that URAT1 has other substrates besides urate (Table 5). These findings suggest additional functions for URAT1 beyond urate transport. Consistent with this notion, the metabolic reconstructions presented here, based on computational definition of the biochemical pathways consistent with the mURAT1 knockout transcriptomic data, suggest multiple metabolic alterations in the mURAT1 knockout mouse distinct from urate handling. As discussed further below, the set of metabolic reactions associated with mURAT1 manifests some commonalities with those previously reported for the related transporters, OAT1 and OAT3 [18,19]. It is also notable that many more reactions had increased functionality in the mURAT1 knockout mouse reconstructions than had decreased functionality, possibly indicating that loss of mURAT1 induces a heightened state of metabolic compensation. This is in contrast to the OAT1 and OAT3 knockout reconstructions which had more reactions with decreased than increased functionality [18,19].</p><p>Metabolites having altered extracellular membrane transport in the mURAT1 knockout mouse reconstructions included several that are plausible URAT1 substrates on the basis of their previously demonstrated in vitro interactions with this transporter and/or other OAT family members (Tables 3 and 4) – indeed, four of the ten known URAT1-interacting substrates of endogenous origin were represented among these molecules (Table 5). Overall, these findings suggest the potential involvement of URAT1 in bioenergetic pathways (via transport of acetoacetate, α-ketoglutarate, lactate, and pyruvate), nucleotide metabolism (via transport of urate), and steroid signaling or metabolism (via transport of or interaction with progesterone). With regard to the latter process, progestgerone and two other intermediates in steroid metabolism ((20S)-20-hydroxypregn-4-en-3-one and 4-methylpentanal) comprised the three compounds manifesting decreased transport in the knockout reconstructions (each by two fold or greater).</p><p>Alterations in the reconstructions of transport of non-OAT substrates and of the non-transport metabolic subsystems presumably reflect possible secondary or downstream effects of mURAT1 loss, including those due to genomic regulatory feedback. These changes involved multiple cellular processes and biochemical pathways taking place in the cytosol, lysosomes, and mitochondria. Prominent among them were increases in functionalities relating to lysosomal turnover of glycosaminoglycans (polymers of aminosugars that constitute the glycan component of glycosylated proteins [31,32]): A significant proportion of the reactions belonging to the subsystems of aminosugar metabolism and degradation of the glycosaminoglycans, chondroitin and keratan sulfate, had flux-span increases greater than two fold in the mURAT1 knockout compared with wild-type reconstructions (Fig. 3A). Moreover, in the lysosomal transport sub-system, which also had a significant proportion of increased reactions in the knockout reconstructions, all of the affected reactions involved transport of glycosaminoglycans or aminosugars (Fig. 3A and Table 1). Notably, altered glycosaminoglycan metabolism was also a feature of our gene expression-constrained reconstructions of metabolism in the OAT1 and OAT3 knockouts [18,19].</p><p>Glycosaminoglycans are not only important elements of the extracellular matrix but are also critical to cell-cell communication mediated by integral membrane proteins and soluble factors [32]. For example, emerging evidence indicates that extracellular matrix glycosaminoglycans may modulate growth factor signaling during branching morphogenesis in the kidney (reviewed in [33]). Thus, our findings suggesting altered metabolism of these molecules in mURAT1 knockout mice raise the possibility of a role for mURAT1 in renal development, though this hypothesis is somewhat mitigated by the lack of obvious developmental anomalies in the knockouts [17]. Moreover, altered glycosylation is a virtually universal feature of malignancy [32] so that URAT1 involvement in glycosaminoglycan metabolism may also have implications for pathophysiological mechanisms in cancer. Notably, glycosaminoglycans bind lipoproteins and may thereby modulate cholesterol levels [32]. Thus involvement in glycosaminoglycan metabolism may help explain the relationship between uric acid and cardiovascular disease.</p><p>Methionine and associated one-carbon metabolism also manifested increased functionality in the mURAT1 knockout reconstructions (Fig. 3A and Table 1). Methionine is a precursor to homocysteine, elevated plasma levels of which represent a cardiovascular risk factor [34]. Notably in this regard, reactions involving the metabolism of folate, which is necessary for the recycling of homocysteine to methionine and is thus cardio-protective, were also increased in the knockout reconstructions, though not to a statistically significant degree (Fig. 3A and Table 1). As with glycosaminoglycan metabolism, the possible involvement of URAT1 in methionine and folate metabolism may contribute to the links between uric acid and cardiovascular disease.</p><p>Conversely, functions related to the biosynthesis of coenzyme A, required in the Krebs cycle and for fatty acid metabolism, were significantly decreased in the mURAT1 knockout reconstructions (Fig. 3B and Table 2), consistent with the role for mURAT1 in renal cellular bioenergetics that was suggested by the alterations in the reconstructions of transport of metabolic intermediates. Of note, the URAT1-related transporters, OAT1 and OAT3, were also implicated in cellular bioenergetics in our prior network analyses [18,19]). The possible involvement of URAT1 in coenzyme A metabolism could have implications for the various neurodegenerative and metabolic disorders that involve coenzyme A dysregulation [35,36,37].</p><p>Accumulating data indicate that OAT1 and OAT3, previously primarily studied in terms of their role in mediating the renal secretion of numerous important organic anionic pharmaceuticals (e.g., non-steroidal anti-inflammatory drugs, β-lactam antibiotics, loop and thiazide diuretics), in fact handle a diverse array of metabolically and clinically significant endogenous substrates (e.g., Krebs cycle intermediates, uremic toxins, enterobiome products, cyclic nucleotides, prostaglandins, and steroid conjugates) [13,14,38,39], suggesting their function in various physiological processes. Indeed, mice null for OAT3 are relatively hypotensive, indicating a role for this transporter in the regulation of blood pressure [23]. Moreover, the expression of these and related transporters in widely dispersed tissues (including liver, olfactory mucosa, and choroid plexus of the brain, in addition to kidney) and their transport of signaling molecules (as listed above) raise the possibility that they participate in organism-wide communication networks (which has been termed the "remote sensing and signaling hypothesis") [38,39,40]. The metabolic reconstructions presented here are in line with likewise physiological roles for mURAT1.</p><p>In summary, in the context of a growing number of URAT1 substrates other than urate, the identification of other clinically important uric acid transporters (e.g., slc2a9, ABCG2), and the presence of metabolomic changes in the mURAT1 knockout (although the involved metabolites could not be identified), we have used gene expression data from mURAT1 knockout mice to computationally constrain pathway reconstructions, thereby allowing global characterization of metabolic networks potentially associated with mURAT1. While there is need for follow up physiological studies analyzing the knockout mice under conditions of perturbed as well as basal homoeostasis, our findings suggest that mURAT1 has additional functions beyond urate reabsorption, including in bioenergtic and biosynthetic metabolism; as detailed above many of these functions overlap those of the closely related transporters, OAT1 and OAT3. Moreover, since mURAT1 and hURAT1 are at least as phylogenetically similar as most mouse-human orthologous gene pairs (they manifest 81% sequence similarity and 74% identity at the amino acid level [1,2,15], slightly superior to the mean 70.1% identity across all mouse-human orthologs [41]), these additional functions may also occur in humans. While the much higher levels of urate in humans than mice could result in greater competitive inhibition by urate of these other URAT1 functions, such inhibition would not be expected to be complete. Urate concentrations in human plasma generally fall in the range 200-500 uM, the midpoint of which, 350 μM, is approximately equivalent to the apparent affinity (Km) of urate for URAT1, which was estimated to be 371 μM [1]. Thus, on average, circulating urate might be expected to occupy about half of the available URAT1 transporters, so that transport of any other substrates would be reduced but not eliminated. Accordingly, our findings in URAT1 knockout mice could potentially have implications for human physiology, including for our understanding of the relationship between uric acid and the various metabolic disorders to which it has been linked.</p>

PubMed Author Manuscript

High Accuracy Semi-Empirical Quantum Models Based on a Reduced Training Set

There exists a great need for computationally efficient quantum simulation approaches that can achieve an accuracy similar to high-level theories while exhibiting a wide degree of transferability. In this regard, we have leveraged a machine-learned force field based on Chebyshev polynomials to determine Density Functional Tight Binding (DFTB) models for organic materials. The benefit of our approach is two-fold:(1) many-body interactions can be corrected for in a systematic and rapidly tunable process, and (2) high-level quantum accuracy for a broad range of compounds can be achieved with ∼0.3% of data required for one advanced deep learning potential (ANI-1x). In addition, the total number of data points in our training set is less than one half of that used for a recent DFTB-neural network model (trained on a separate dataset).Validation tests of our DFTB model against energy and vibrational data for gas-phase molecules for additional quantum datasets shows strong agreement with reference data from either hybrid density-functional theory, coupled-cluster calculations, or experiments. Preliminary testing on graphite and diamond successfully reproduce condensed phase structures. The models developed in this work, in principle, can retain most of the accuracy of quantum-based methods at any level of theory with relatively small training sets. Our efforts can thus allow for high throughput physical and chemical predictions with up to coupled-cluster accuracy for materials that are computationally intractable with standard approaches.

high_accuracy_semi-empirical_quantum_models_based_on_a_reduced_training_set

<p>materials science. These calculations can provide atomic level detail of physical phenomena and chemical reactions that can help to understand experimental observations. With modern computing advances, computationally intensive simulations using quantum mechanical methods such as hybrid density-functional theory (DFT), Møller-Plesset second-order perturbation theory (MP2), or coupled-cluster approaches can be performed to provide accurate descriptions of electronic and vibrational states, as well as thermodynamic quantities for a diverse set of systems. The high computational expense of these approaches, however, generally limits their application to static gas-phase clusters or small, sub-nanometer systems sizes. This can be far below the spatial scales of most experimental studies, which frequently probe nanometers or beyond and involve dynamic measurements. For example, the "gold standard" CCSD(T) (coupled-cluster considering single, double, and perturbative triple excitations) method scales as O(N 7 ), where N is the number of basis functions involved in the calculation. Consequently, it is generally only applicable to systems with tens of atoms or less, precluding its use for larger biomolecules or condensed phases. Neural network (NN) approaches can be used to completely parameterize the quantum mechanical interactions, 1-3 bypassing the need for direct electronic state calculation and thus yielding improved scaling and efficiency. However, NNs tend to require extremely large training sets and can perform poorly outside of their training regime. 1 This need for large data sets, in particular, makes such approaches challenging to implement and use effectively. Therefore, there is a widespread need for the development of computational methods that can maintain the accuracy of high level approaches while yielding substantial gains in training set size, computational costs, and scaling.</p><p>Semi-empirical methods, for example Density Functional Tight Binding (DFTB), [4][5][6][7] are a potential strong alternative quantum approach. The DFTB Hamiltonian is derived directly from an expansion of the Kohn-Sham DFT total energy, yielding a good balance between approximate quantum mechanics and empiricism. This can result in calculations requiring only a small fraction of the computational cost compared to DFT or other high-level quantum approaches. Here, the DFTB total energy is written as:</p><p>where E BS corresponds to the band structure energy, E Coul is the charge fluctuation term, and E rep is the repulsive energy. E BS is calculated as a sum over occupied electronic states from the DFTB Hamiltonian. In practice, DFTB Hamiltonian matrix elements are computed from pre-tabulated Slater-Koster tables derived from reference calculations with a minimal basis set. The repulsive energy, E rep , corresponds to ion-ion repulsions, as well as Hartree and exchange-correlation double counting terms. This term can be expressed as an empirical function where parameters are fit to reproduce high-level quantum or experimental reference data. A pairwise potential energy function is often used for the repulsive energy term, 8,9 though many-body interaction terms are required in some cases. 10,11 DFTB is approximately three orders of magnitude more efficient than DFT calculations and exhibits O(N 3 ) scaling.</p><p>Its combination of approximate quantum mechanics with empirical functions can allow for a high degree of flexibility in terms of optimization approaches, desired accuracy, and transferability across element types and diverse conditions. 12-14 DFTB models have been created for a broad range of materials, though the repulsive energy largely has been tuned to relatively low-level DFT data for condensed phases. [15][16][17][18][19] Recent efforts have been made to enhance the accuracy of DFTB through creation of more sophisticated and systematic approaches for determining the repulsive energy term. and the Curvature Constrained Splines methodology 14 have been used to create strictly pair-wise additive repulsive energies for several organic and inorganic systems. However, these methods can struggle for systems where greater than two-body interactions in E rep are needed. 14 NNs have been proposed as a promising method to include many-body interactions into the DFTB repulsive energy. 22,23 The resulting DFTB-NN models have the capability of predicting molecular properties for a wide range of compounds and element types. However, the drawback of utilizing NNs for E rep is similar to their use for classical force fields in that they require large amounts of training data and can have slow parameterization due to the presence of quasi-degenerate local minima, 24 making their development exceedingly challenging.</p><p>Here, we explore the possibility of creating DFTB models that can leverage the relative simplicity of linear regression machine learning in the recently developed Chebyshev Interaction Model for Efficient Simulation (ChIMES) method. ChIMES is a many-body force field based on linear combinations of Chebyshev polynomials. 25 It has been shown that ChIMES models yield good agreement with DFT reference method for a wide range of properties and materials under both ambient and extreme conditions. [26][27][28] The main advantage of ChIMES is that it is completely linear in fitted coefficients, allowing for rapid parameterization to a global minimum. The reliance on Chebyshev polynomials, which are orthogonal, allows the complexity of a ChIMES model to be systematically tuned to an arbitrary degree of accuracy and transferability, while also providing straightforward methods for regularization to minimize overfitting. 16 In this study, we determine an optimal DFTB/ChIMES model for C, H, N, O-containing systems using high level quantum chemical reference data. We use an iterative scheme to systematically expand our training set where at each iteration, a small fraction of the force configurations with largest deviation in our validation set are included in the next training set iteration. The accuracy and transferability of the resulting model are investigated for a wide variety of gas-phase clusters as well as some carbon solids. We find that use of a small fraction of our chosen data set (∼0.3% of similar NN efforts) yields DFTB/ChIMES models that maintain close to hybrid functional, coupled-cluster, and/or experimental accuracy for the gas-phase clusters studies here, and compares favorably to previous DFTB-NN efforts for similar systems.</p><p>For our DFTB/ChIMES models, the total energy is determined as the sum of the standard DFTB energy with an additional ChIMES contribution:</p><p>For this work, DFTB calculations are performed using the 3ob-3-1 parameter set, which contains a third-order expansion about the charges and is considered an optimal DFTB starting point for most organic system. 22,29 The ChIMES energy is written as a many-body expansion:</p><p>where n a is the number of atoms in the system. The atomic energies E i are constants used to match energies from reference data, and two-body (pairwise) energies are expressed as linear combinations of Chebyshev polynomials of the first kind. 30,31 Higher-bodied interactions are determined through products of a cluster's constituent pair-wise polynomials. 32 ChIMES parameters are determined by fitting to the difference between the reference energies and atomic forces and those computed from DFTB alone using the following objective function:</p><p>where N d is the total number of data entries, given by forces) are smaller than the variations between wB97X-DFT itself and higher levels such as CCSD(T) and MP2 (4.9/5.9 kcal/mol for energies and 4.6/5.9 kcal/mol-Å for forces). 3</p><p>The performance of DFTB/ChIMES in comparison to coupled-cluster reference data is also provided in Table 1. Here, we have selected the ISO34 data set 35 and DFTB-NN rep (DFTB-NN with deep tensor neural networks). 22 One can see that the accuracy of DFTB/ChIMES is much better than that for standard DFTB, is slightly improved over that from DFTB-NN rep , and approaches the PBE0 data given in Reference 22.</p><p>To test the performance of our model on high accuracy force data specifically, we compare DFTB/ChIMES with the CCSD(T)/cc-pVTZ data for 2000 configurations of ethanol in the GDML data set 38 (54,000 data points total). Again our DFTB/ChIMES gives an improvement over standard DFTB as MAE and RMSE are both reduced by ∼40%. A direct force comparison to DFTB-NN rep or the ISO34 reference was unavailable.</p><p>To probe the smoothness in the potential energy surface from DFTB/ChIMES, we have also computed the potential energy profile for rotation around the dihedral angles in alka-nes. The torsional profile for n−butane is shown in Figure 2 of the metastable minimum (at ±70 • ) in good agreement with wB97X reference data with deviations of less than 0.5 kcal/mol. DFTB, however, underestimates the torsional barriers by 0.9 and 1.7 kcal/mol for the lower-energy (at ±120 • ) and main barrier (at 0 • ), respectively. DFTB/ChIMES is more accurate overall, with deviations of less than 0.5 kcal/mol for predicting all energy barriers discussed here.</p><p>Next, we compare the the vibrational frequency predictions of DFTB and DFTB/ChIMES on 342 gas phase molecules from the Computational Chemistry Comparison and Benchmark Database or CCCBDB (https://cccbdb.nist.gov/). The reference data is at the MP2/cc-pVTZ level of theory. We also make comparisons with several DFT methods. The functionals chosen here are wB97XD, 39 which is the same as wB97X with an additional dispersion correction, and the Perdew-Burke-Ernzerhof (PBE) functional. 40 The predicted vibrational frequencies for those DFT functionals are also taken from CCCBDB. Figure 3 shows the distribution of frequencies for each computational method. wB97XD gives good agreement with MP2 reference data with MAE/RMSE = 20/36 cm −1 . DFTB and PBE underestimate the vibrational stretching frequencies by about 100 cm −1 on average, where the MAE/RMSE are 77/114 and 61/79 cm −1 , respectively. DFTB/ChIMES yields smaller errors in the frequency prediction with MAE/RMSE = 36/61 cm −1 , showing notable improvement over PBE and comparable accuracy to wB97XD.</p><p>Though the DFTB/ChIMES model developed here is trained on molecular (gas phase) data, we have also tested its performance in reproducing the structural properties of graphite and diamond. These systems were chosen due to the fact that they contain a single element only while still probing different types of chemical bonds. In conclusion, we have shown that ChIMES can be used to extend DFTB to hybrid functional accuracy or greater. ChIMES parameters are determined rapidly through linear optimization, creating a beyond-pairwise interaction potential for DFTB. DFTB/ChIMES has the capability of reproducing vast quantities of high-level reference data while requiring only a small fraction of it for training. The accuracy of DFTB/ChIMES is discussed for total energies, atomic forces, isomerization energies, and vibrational frequencies across the vast conformational diversity of organic molecules in several popular datasets, as well as for the dihedral rotation energy profile of n−butane. Preliminary testing on solid carbon allotropes at ambient conditions show that DFTB/ChIMES is able to reproduce the experimental structure of graphite (a well-known challenge for standard DFT) as well as bulk diamond properties, while having been determined from gas-phase cluster data, only. On the basis of the results presented here, DFTB/ChIMES represents a promising direction for developing general purpose quantum models that are applicable to a wide range of materials and conditions. The small training set required by our approach, as shown in this study, could yield significant advantages for future development of computational models with a coupled cluster accuracy, significantly improved scaling, and high efficiency. The utility and ease of parameterization of DFTB/ChIMES allows for high-level quantum theory accuracy in the systems where traditional methods are far too computationally intensive for use.</p>

ChemRxiv

Rotationally inelastic scattering of CD3 and CH3 with He: comparison of velocity map-imaging data with quantum scattering calculations

Rotationally inelastic scattering of methyl radicals (CD 3 and CH 3 ) in collisions with helium is examined by a combination of velocity map imaging experiments and quantum scattering calculations. In the experiments a beam of methyl radicals seeded in Ar intersects a beam of He atoms at 90 at a collision energy of 440 AE 35 cm À1 (CD 3 + He) or 425 AE 35 cm À1 (CH 3 + He). The methyl radicals are prepared photolytically in a gas expansion that cools them to 15 K, giving a distribution over a small number of initial (low) rotational angular momentum states. By resonance-enhanced multi-photon ionization detection, we obtain velocity map images which are specific to a single rotational angular momentum quantum number n 0 of the methyl radicals, but averaged over a small subset of the projection quantum number k 0 . We extract resolved angular scattering distributions for n 0 ¼ 2-9 (for CD 3 ). We compare these to predictions of scattering calculations performed based on a recent potential energy surface [P. J. Dagdigian and M. H. Alexander, J. Chem. Phys. 2011, 135, 064306] in which the methyl radical was fixed at its equilibrium geometry. The fully (n, k) / (n 0 , k 0 ) resolved differential cross sections obtained from the calculations, when combined in weighted sums over initial (n, k) levels corresponding to the 15 K experimental radical temperature, and final k 0 levels that are not resolved in the spectroscopic detection scheme, show excellent agreement with the experimental measurements for all final states probed. This agreement gives confidence in the calculated dependence of the scattering on changes in both the n and k quantum numbers.

rotationally_inelastic_scattering_of_cd3_and_ch3_with_he:_comparison_of_velocity_map-imaging_data_wi

Introduction<!>A. Experimental apparatus<!>B. Density-to-ux transformation<!>D. REMPI detection and state distribution in incident beams<!>E. Quantum scattering calculations<!>A. DCSs for CD 3 + He collisions<!>B. DCSs for CH 3 + He collisions<!>Discussion<!>Conclusions

<p>The methyl radical plays an important role in the combustion of hydrocarbons, 1,2 chemical vapour deposition (CVD) of diamond lms, 3 and in the chemistry of the atmospheres of the outer planets in the solar system. 4 In addition, CH 3 has been detected in the interstellar medium via its infra-red emission bands. 5 Observation of methyl radicals in the upper atmospheres of Saturn 6 and Neptune 7 indicates it is a reactive intermediate in the hydrocarbon photochemistry of these planets: methyl radicals are created by vacuum ultraviolet photodissociation of methane, and the self-recombination reaction is believed to be the only photochemical source of ethane. The atmospheres of giant planets such as Saturn and Neptune are composed mainly of molecular hydrogen and helium, with trace amounts of other substances. Therefore, photochemically generated methyl radicals will undergo elastic and inelastic collisions with He and H 2 before reactive loss, with reaction cross sections that can depend on the internal energy of the radicals.</p><p>The inelastic scattering of labile free radicals using molecular beams and laser spectroscopic techniques was reviewed in the mid 1990s. [8][9][10] However, considerable advances have been made since then using methods such as ion imaging 11 and velocity map imaging (VMI) 12 with laser spectroscopic detection of the nal levels. Most of these experimental studies have concentrated on scattering dynamics of diatomic radicals, with spectroscopic probes used to measure state-resolved integral cross sections (ICSs), and more recently differential cross sections (DCSs). The most extensively studied free radicals have been NO [13][14][15][16][17][18][19][20][21][22][23][24] and OH, [25][26][27][28][29][30][31][32][33] the latter because of its important role in atmospheric chemistry, astrochemistry and combustion. Sarma et al. 34 used VMI to obtain fully quantum-state-specied product angular distributions for OH scattered by He and Ar, and VMI methods were also used to study collision induced alignment 35 and orientation 36 in NO -Ar scattering. Radicalradical scattering studies are rarer, but Kirste et al. 37 recently reported state-to-state ICSs for collisions of Stark decelerated OH with NO.</p><p>Nevertheless, measurements of DCSs for inelastic scattering of free radicals other than NO are rare, and to the best of our knowledge have not been reported for reactive radicals larger than diatomics. Macdonald and Liu, 38,39 and Lai et al. 40 examined the inelastic scattering of the linear NCO radical with He and Ar, respectively, but concentrated on ICSs for spin-orbit conserving and spin-orbit changing collisions. ICSs have also been reported for rotationally inelastic collisions of NH 2 with He. 41 Greater attention has been paid to the inelastic scattering of closed-shell triatomic and polyatomic molecules, as illustrated by determinations of DCSs for scattering of ammonia [42][43][44] and deuterated ammonia 45 with rare gases and molecular hydrogen, and for water with helium 46 and hydrogen. 47 Along with advances in experimental techniques, there have been many quantum scattering calculations of ICSs, and also DCSs, employing high-quality potential energy surfaces (PESs). These have mostly concerned collisions of diatomic and stable polyatomic molecules. 14,18,21,[25][26][27][28][30][31][32][33]42,[46][47][48] Dagdigian 49 recently reviewed work on quantum scattering calculations of collisional rotational and vibrational energy transfer in small hydrocarbon intermediates and highlighted studies involving methylene (CH 2 ) 50,51 and methyl. 52,53 The pathways for energy transfer in collisions of a polyatomic are more complicated than for collisions of a diatomic molecule. There is only one type of anisotropy in an atom-diatom interaction, namely the difference in interaction energy for end-on vs. side-on approach. By contrast, for collisions of a nonlinear polyatomic molecule there are two types of anisotropies, corresponding to approach of the collision partner in or perpendicular to the molecular plane.</p><p>Dagdigian and Alexander 54 recently investigated rotational energy transfer of methyl in collisions with a helium atom through quantum scattering calculations on a computed PES. This PES was calculated with a coupled-cluster method that includes all single and double excitations, as well as perturbative contributions of connected triple excitations [RCCSD(T)]. Because of the anisotropy of the PES due to the repulsion of the He atom by the three H atoms on methyl, a strong propensity was found for Dk ¼ AE3 transitions, where k is the body-frame projection of the rotational angular momentum n.</p><p>Ma et al. 53 extended this work to the study of the vibrational relaxation of the n 2 mode of methyl by a He atom. Vibrational relaxation was found to be nearly two orders of magnitude less efficient than pure rotational relaxation. The vibrational relaxation rate also depends strongly upon the rotational quantum numbers n and k. Although methyl has an unpaired electron, these calculations 53,54 ignored the electron spin. The dependence of the cross sections on the spin can be determined in a straightforward manner by assuming that it is a spectator during the collision. 49 In the present work DCSs for collisions of CH 3 and CD 3 with helium are determined experimentally through the use of crossed molecular beam (CMB) and VMI methods. The measured DCSs are compared with quantum scattering calculations that use the computed PESs mentioned above. 53,54 Comparison between experiment and theory at the level of stateresolved DCSs provides a critical test of the inuence of both short-range repulsive and long-range attractive intermolecular interactions. Anticipating this comparison, we nd excellent agreement between the measured and computed state-to-state DCSs. This conrms the accuracy of the computed PES.</p><p>This paper is organized as follows: Method Sections A and B describe the details of the experimental determination of the angular distributions. Method Sections C and D present, respectively, a brief description of the rotational levels of CH 3 and CD 3 , specically the different nuclear spin modications, and the spectroscopic intricacies of the detection scheme. Section E describes the quantum scattering calculations. Results Sections A and B present and compare the measured and theoretical cross sections for collision of CD 3 and CH 3 with He, respectively. Discussion and Conclusions sections then follow.</p><!><p>Experimental measurements were carried out using a newly constructed, compact crossed molecular beam machine with velocity map imaging detection, based on the design of Strecker and Chandler. 55 Pulsed molecular beams of jet-cooled methyl radicals (either CD 3 or CH 3 ) and helium crossed at 90 in a high vacuum chamber, and the velocities of the scattered methyl radicals were imaged following resonance enhanced multiphoton ionization (REMPI) with rotational level resolution. The resultant images were corrected by density to ux conversion prior to analysis of angular dependences to derive quantumstate-resolved differential cross sections. A schematic diagram of the top and side view of the instrument is depicted in Fig. 1, and the component parts are described in greater detail below.</p><p>Molecular beams were formed by supersonic expansion of gas samples through a pair of pulsed valves (General Valve Series 9), and the expansions were collimated by skimmers. The pulsed valves operated at 10 Hz repetition rate, the nozzle diameters were 0.8 mm, and the skimmer orices were 1 mm diameter. The nozzle-skimmer distances could be adjusted in the range 0.2 cm to 4.3 cm, and for the current experiments were selected to be 3.2 cm. The distance from each skimmer to the scattering centre was 3.7 cm. The source chambers were evacuated by a turbomolecular pump (Edwards nEXT300D). The two skimmed beams propagated horizontally and crossed at a 90 intersection angle. The gas ows in both beams passed through the scattering region and were directed straight into two additional turbomolecular pumps (Pfeiffer HiPace 80). A typical base pressure for the scattering chamber was <10 À8 Torr with the pulsed valves turned off, and this rose to $10 À7 Torr when the valves were operating. Additional turbomolecular pumps were mounted just above the ion optics, and in the detector chamber.</p><p>The primary molecular beam was formed by expansion of a mixture of 3% CH 3 I or CD 3 I in Ar at a stagnation pressure of 4 bar. The secondary beam was formed by expansion of 3 bar of pure helium. Methyl radicals were generated by 266 nm photolysis of the CH 3 I or CD 3 I precursor, using 40 mJ per pulse of the fourth harmonic of a Nd:YAG laser (Surelite SLII-10). Both photolytic and electric discharge sources of methyl radicals were tested, but the discharge source was found to produce radicals with a rotational temperature of 120 K, that was too high for inelastic scattering experiments.</p><p>The front face of the valve used to generate the primary molecular beam was modied to accommodate prisms above and below the nozzle orice, such that a UV laser beam could be passed across the gas expansion. Hence, methyl iodide molecules seeded in the argon expansion were photolysed in close proximity to the nozzle, where the high number density ensured the resultant methyl radicals underwent numerous collisions with atoms of the carrier gas. As is shown in ESI, † the methyl radicals cooled to a rotational temperature of $15 K with this method of preparation. Residual methyl iodide in the molecular beam was observed, but did not affect the velocity map images for methyl radical scattering because any photolysis of the methyl iodide by the probe laser generated methyl radicals with velocities substantially higher than those from the inelastic scattering of interest here.</p><p>The intersection region of the two molecular beams was located within a vertically mounted stack of electrodes forming an ion optics assembly (Fig. 1c) for VMI. Scattered methyl radicals were ionized by a probe laser focused at the intersection of the beams, and the electric eld created by the ion optics accelerated the ions upwards towards a position-sensitive detector. A 51 cm long eld-free dri region was located between the end of the stack of ion optics and the detector. The ion detector (Photek) consisted of a pair of microchannel plates (MCPs) (10 mm pore on 12 mm pitch), a phosphor screen (P46 phosphor) and CCD camera (BASLER A310f, 782 Â 582 pixels). The voltages on the front and rear MCPs were maintained at 830 and 1300 V, respectively. An additional 500 V was applied to the rear MCP (making a total of 1800 V) for 20 ns to time-gate the detector (Photek GM-MCP-2) so as to detect only the narrowest possible, central portion of the scattered methyl radical Newton sphere. The voltage applied to the phosphor screen was 6300 V. Precise velocity calibration of the ion optics and imaging detector assembly was carried out using O 2 multi-photon excitation and dissociation at a wavelength of 224.999 nm, 56,57 which is a standard method to establish pixel radius to speed conversion factors in VMI. 12,58,59 UV radiation in the wavelength range 285-288 nm required for REMPI detection of the methyl radicals was generated by frequency doubling the output of a tuneable pulsed dye laser (Sirah), operating with Pyromethane 597 dye, using a KDP doubling crystal. The dye laser was pumped by 532 nm light from an Nd:YAG laser (Surelite SLI-20). The energy, repetition rate and duration time of this probe laser beam were 4.5 mJ per pulse, 10 Hz and 4-6 ns, respectively, the laser linewidth was 0.0027 nm (0.32 cm À1 ) and the laser polarization was parallel to the scattering plane. The (2 + 1) REMPI scheme was used both to optimize the production of cold methyl radicals in the primary molecular beam, and to observe the inelastic scattering of these radicals in collisions with the He in the secondary beam. The maximum Doppler shi of inelastically scattered CD 3 radicals is 0.1 cm À1 , which is smaller than the laser linewidth. The laser was therefore xed at a wavelength corresponding to the centre of the chosen REMPI line, as determined from the PGOPHER simulation described in Section D. Nozzle and skimmer alignments were set using two diode lasers, and the probe laser was then aligned and focused to the crossing point of the two molecular beams, with minor adjustment to optimize signal levels. The probe laser position remained the same for all image determinations.</p><p>The ion optics assembly for VMI, shown in Fig. 1c, consisted of 20 electrodes and was designed for direct current (DC) slice imaging. 60 The ion optics were adapted from the design proposed by Lin et al. 59 Careful consideration was given to the number of electrodes, their spacing and applied voltages, while limiting their diameters so that the assembly tted within the connes of the miniature CMB machine. All these parameters were optimised using SIMION simulations, while ensuring precise velocity map imaging across the intersection volumes of the molecular beams and probe laser. 12 Voltages on the electrodes in the bottom section of the ion optics stack formed the extraction and focusing elds, and these voltages could be adjusted separately, with 1 kV applied to the repeller electrode. The electrodes forming the accelerating eld were connected by resistors that ensured a gradually decreasing eld to the nal ground electrode.</p><p>The electrodes forming the homogeneous acceleration eld stretched the methyl ion packet along the ight axis according to the initial velocities of the neutral methyl radicals. The aforementioned short voltage pulse at the MCPs then allowed only a thin centre slice of the ion packet to be recorded, corresponding to methyl radicals scattered within, or close to the plane of the crossed molecular beams. Direct analysis of this slice image gave the three dimensional velocity distribution of the inelastically scattered methyl radicals without need for image reconstruction techniques, although some correction of the raw images was necessary to derive product uxes, as is discussed further in Section B. The nozzle producing the secondary beam of pure He was operated in a repeating mode of 50 shots on and 50 shots off, thereby separately recording total ion signal and any background contributions from the interaction of the probe laser with the primary beam. The desired scattering signal was then obtained by subtraction of the background image from the total signal image.</p><!><p>A Newton diagram for inelastic scattering of CD 3 with He is shown in Fig. 2 and illustrates the laboratory frame velocities of CD 3 [v CD 3 ¼ 550 AE 30 m s À1 ], and helium [v He ¼ 1710 AE 80 m s À1 ], and the pre-and post-collision centre-of-mass (CM) frame velocities of the methyl radical u CD 3 and u 0 CD3 , respectively. The CM-frame scattering angle q is dened as the angle between the CM-frame velocities of CD 3 before and aer a collision. Analysis of experimental images requires a density-to-ux transformation because the detection efficiency of the scattered products depends on their laboratory frame velocity. This detection bias leads to an asymmetry in the measured images with respect to the relative velocity vector (dashed line in Fig. 2).</p><p>To correct the images for this detection bias, we employ the method of Monte Carlo simulation of the experiment, using a modication of the computer program of Eyles and Brouard. 61 This code simulates an instrument function that determines a relative detection efficiency of scattered molecules that depends on their nal laboratory frame velocity. The changes made to the Monte Carlo program include simulation of slicing of the central part of the Newton sphere. For reliable use of the simulation program, values for various parameters characterizing the experimental apparatus were carefully determined. These parameters included the speed and angular divergence distributions, temporal proles, and spatial widths of the two molecular beams, as well as the temporal prole, Rayleigh range and beam waist of the focused probe laser. By sampling $2 Â 10 8 sets of initial conditions from the distributions of molecular beam and laser beam properties, the Monte Carlo program was used to simulate the instrument function. As will be discussed in Section D, the k projection number is not resolved in the REMPI spectra of methyl radicals. A set of instrument functions was therefore simulated for individual k projection numbers and each was weighted according to the 2-photon line strength factors for the given detection line. The effect of varying k projection number is found to be negligible for low n, where the difference in energy between individual k-states is small, while it is more pronounced for higher n. A nal, corrected image was then obtained by dividing the raw experimental image by this sum of instrument functions over k projection quantum numbers. As part of this analysis, the angular resolution of the experiment was calculated and is reported in Fig. S1 In this subsection, we briey describe the rotational levels of the CH 3 and CD 3 radicals, and their nuclear spin symmetries. The rotational energies for the lower levels of CH 3 and CD 3 are plotted in Fig. 3. The methyl radical is an oblate symmetric top. We use n and k to designate, respectively, the rotational quantum number and its body-frame projection. Because the three H (D) atoms are equivalent, the ground vibronic state of CH 3 has two nuclear spin modications, labelled ortho and para. 54,62 The ortho levels have nuclear spin symmetry A 1 and include rotational levels for which k is a multiple of 3 (k ¼ 0, 3, 6, .). In particular, the rotational levels with k ¼ 0 and odd n do not exist for the ground vibronic state. The para levels have nuclear spin symmetry E and include all rotational levels for which k is not a multiple of 3 (k ¼ 1, 2, 4, 5, .).</p><p>The CD 3 radical has three nuclear spin modications. The A 1 nuclear spin functions are those with rotational levels with k ¼ 0 and odd n and with levels for which k is a multiple of 3. The A 2 nuclear spin functions correspond to rotational levels with even n and k ¼ 0, and also with levels for which k is a multiple of 3. Thus, rotational levels with k ¼ 3, 6, . have two components (A 1 and A 2 ). Finally, the E nuclear spin functions include all levels for which k is not a multiple of 3.</p><!><p>The distribution of rotational levels in the supersonic incident CH 3 and CD 3 beams and the inelastically scattered nal levels were detected using (2 + 1) REMPI spectroscopy through the 0 0 0 band of the 4p 2 A 00 2 ) X2 A 00 2 transition. 63,64 Lines in the Q branch are by far the strongest and were employed to optimize the experimental conditions but were of no value in determining state-to-state DCSs because of spectral overlap.</p><p>The level distributions in the incident radical beams were determined by comparison of experimental spectra with spectra simulated using the PGOPHER program. 65 The simulation incorporated the effects of nuclear spin statistics of the three equivalent H or D atoms. The procedures used are described in the ESI, † which also contains an example spectrum (Fig. S3 †). The derived relative populations are listed in Table 1 and correspond to a rotational temperature of 15 K. As expected from the smaller rotational constants of CD 3 , more rotational levels have signicant populations than in the CH 3 isotopologue.</p><p>It should be noted that the lines in the REMPI spectrum of methyl are resolved in the n rotational quantum number, but not in the k projection quantum number. Depending upon the spectroscopic branch, and hence Dn (s0) of the line used to detect the nal level, the k projection levels of a given n contribute differently. The levels associated with the DCSs reported below are denoted by n k 1 k 2 . to indicate that the unresolved n k 1 , n k 2 , . levels have been detected on the given transition. The relative contributions of the different k projection levels to the measured REMPI intensity were determined by calculating 2-photon line strength factors using the PGOPHER program.</p><p>The levels of the excited 4p 2 A 00 2 electronic state are predissociated, and the linewidths for the CH 3 isotopologue are larger than for CD 3 . Hence, the efficiency of detection of CH 3 rotational levels is lower, and it was possible to determine DCSs for fewer nal levels than for CD 3 . In addition, the computed integral cross sections for formation of higher CH 3 rotational levels were found to be smaller than for CD 3 levels of comparable rotational angular momentum.</p><!><p>We used the HIBRIDON suite of programs 66 to carry out fully quantum, close-coupled, state-resolved differential cross section calculations for collisions of CH 3 and CD 3 with He. The bulk of the calculations employed our previously computed 54 PES for the interaction of CH 3 , xed at its equilibrium geometry, with helium. Since the centre of mass of methyl is located at the carbon atom, this PES could be used without modication for the CD 3 isotopologue. Some additional calculations on CD 3 -He were carried out with a PES for which the CD 3 geometry was averaged over the probability distribution for the n 2 umbrella coordinate, using our previously determined 53 4-dimensional PES involving this degree of freedom.</p><p>Rotational energies were computed with a rigid rotor symmetric top Hamiltonian using spectroscopic studies by Yamada et al. 67 for CH 3 and Sears et al. 68 for CD 3 . The methyl radical is an open-shell species, with doublet spin multiplicity, so that each rotational level, with rotational angular momentum n, is split into spin doublets, with total angular momentum j ¼ n AE 1/2. We have ignored spin in our scattering calculations since the spin-rotation splitting and hyperne splittings are small 69 and not resolved in the REMPI spectra. Separate calculations were carried out for each of the three nuclear spin modications since they are not interconverted in collisions with closed-shell species without nuclear spin.</p><p>We checked convergence of the differential cross sections with respect to the size of the rotational basis and the number of partial waves in the calculation. Rotational levels whose energies were less than 1100 cm À1 were included in the channel basis, and the calculations included total angular momenta J # 130ħ.</p><p>Since the CH 3 and CD 3 incident beams each contained several rotational levels, DCSs for formation of a specic nal rotational level n k were determined by weighting the computed state-to-state DCSs at the experimental collision energy by the experimentally determined rotational level populations in the incident beam, which are listed in Table 1. Since the k projection number of the scattered radicals is not resolved in the REMPI spectra, theoretical DCSs for comparison with the experimental measurements were weighted according to the 2-photon line strength factors for the given detection line. More details can be found in ESI. †</p><p>We observed previously that CH 3 + He integral cross sections computed with the rigid-molecule PES were virtually identical (within 1%) to those computed using a PES in which the umbrella motion was averaged over the n 2 ¼ 0 probability distribution. 53 We nd a similar excellent agreement of CD 3 + He DCSs computed with rigid-molecule and umbrella-averaged PESs. The only exception was a slight increase ($5%) of the forward (q < 10 ) scattering for low-Dn, Dk ¼ +3 transitions, e.g. 1 0 / 3 3 as shown in Fig. S6 of the ESI. † Consequently, for comparison with experiment we used the DCSs calculated with the rigid-molecule PES.</p><!><p>From the measured beam speed distributions (see ESI †), the mean CD 3 + He collision energy was calculated to be 440 cm À1 with an expected spread of AE35 cm À1 . Fig. 4 presents the raw images recorded for detection of CD 3 nal rotational levels for n 0 ¼ 2-9. These images were each typically accumulated for 8 to 10 hours. The scattered products cannot be observed in the portions of the images corresponding to the forward direction because of imperfect subtraction of background signals arising from unscattered CD 3 radicals present in the parent molecular beam. The incomplete cooling of the methyl radicals to the lowest rotational level of a given nuclear spin modication during the supersonic expansion is the origin of this masking of the scattering signals and is more signicant when lower rotational levels are probed. Nevertheless, from examination of the images, we see that the scattering is conned relatively close to the incident beam direction for detection of low n 0 nal levels. This suggests that the scattering is largely in the forward direction for these levels. By contrast, signicant intensity extends to a much larger range of scattering angles for high n 0 nal levels.</p><p>The recorded images were analysed following the procedures described in Method Section B to correct for the density-to-ux transformation in order to derive the DCSs. Fig. 5 displays the determined DCSs for nal levels n 0 ¼ 2-4, while Fig. 6 presents the DCSs for n 0 ¼ 5-9. In both gures, the unresolved k 0 projections are specied for each n 0resolved DCS. Also shown in Fig. 5 and 6 are theoretical DCSs. The experimental DCSs are not shown for q < 30 for nal levels with n 0 ¼ 2 and 3 and for q < 20 for nal levels with higher n 0 because of contributions to these angles from unscattered radicals in the parent beam, as discussed above. The calculated DCSs show pronounced Þ for final rotational levels from n 0 ¼ 2-9, with unresolved final k 0 projection levels as discussed in the text. Y denotes the spectroscopic branch. The rotational angular momentum of the incident CD 3 beam was predominantly n ¼ 1 (see Table 1 for the complete incident level distribution). The orientation of the relative velocity vector v rel is indicated in one panel. diffraction oscillations in this strongly forward scattered region. Unfortunately, even with greater initial state purity the angular resolution of the experiments, imposed by the velocity and angular spreads of the two molecular beams, would be insufficient to resolve these structures clearly. For the current experiments on the CD 3 + He system, this angular resolution ranges varies from 3 to 16 depending on the scattering angle (see ESI † for further details).</p><p>For quantitative comparison with the theoretical calculations, the experimental angular distributions are normalized by scaling the experimental value to match the theoretical value at either 90 for levels with n 0 # 4, or at 180 for higher n 0 states. These choices were made to ensure that the comparison was done at angles where the experimental signal levels were strongest. The error bars associated with the experimental DCSs were determined by combining the standard deviation determined from comparison of several (typically 3) measured images for a single nal state with the uncertainty introduced by application of the density-to-ux transformation. The latter factor was quantied by comparing DCSs extracted from the two halves of the image separated by the relative velocity vector (which should be symmetric aer perfect transformation). The theoretical DCSs reproduce satisfactorily all the features of the measured DCSs to within the experimental uncertainty. The n 0 -dependent DCSs are a sensitive probe of both attractive and repulsive parts of the potential. The near quantitative agreement for all nal n 0 levels conrms the high quality of the calculated ab initio PES 54 and the accuracy of the close-coupling treatment of the scattering dynamics.</p><p>In most cases, different spectroscopic branches can be used to detect a given methyl rotational level n 0 . The unresolved k 0 projections of the given n 0 , weighted according to the 2-photon line strength factors (see Method Section D) will give rise to slightly different predicted DCSs. As an illustration of this subtle effect, we see in Fig. 5 and 6 that the experimental and theoretical DCSs for a given n 0 are slightly different for detection of this level on REMPI lines of different Dn 0 . This is most dramatically illustrated in Fig. 5 in the comparison of detection of n 0 ¼ 4 on the O(4), P(4), and S(4) lines. The differing measured DCSs reect the fact that the k 0 projection levels of nal rotational level n 0 ¼ 4 have different DCSs. This point will be taken up again in the Discussion.</p><!><p>The collision energy of 425 AE 35 cm À1 for inelastic scattering of CH 3 radicals (seeded in excess Ar) with He is slightly smaller than for the scattering of CD 3 by He, because of the smaller reduced mass. The spectroscopic lines for CH 3 are also not resolved in the k projection quantum number, although the spacings between the lines detecting different k levels of the same n are larger than for CD 3 . In addition, the CH 3 REMPI transitions are more broadened by predissociation of the intermediate Rydberg state, which lowers the detection efficiency. Consequently, velocity map images for CH 3 scattering were recorded only for the three strongest spectroscopic lines. Fig. 7 presents representative examples of these images. The DCSs derived from the density-to-ux transformation (Method Section B) and normalized to the theoretical DCSs are presented in Fig. 8. In two of the panels in Fig. 8, experimental DCSs derived from images accumulated on different days are compared. We see that these DCSs lie almost entirely on top of each other, which demonstrates excellent reproducibility. Fig. 8 also compares these experimental DCSs with theoretical calculations. There is agreement for the n 0 k 0 ¼ 3 123 and 2 1 nal levels, except for small angle scattering. However, the experimental DCS for scattering into n 0 k 0 ¼ 2 012 does not agree with the comparable theoretical DCS for q # 90 , computed under the assumption that the probe laser excites all three k 0 projection levels. Re-measurement of the images conrms the reproducibility of the experimentally determined DCSs (red and blue lines in Fig. 8b). However, careful examination of the 2-photon transition wavenumbers reveals that lines originating from these k 0 levels are separated by more than their widths. If we assume that the k 0 ¼ 2 level was preferentially excited, then we obtain good agreement of the computed DCS [green curve in Fig. 8b] with the experimental DCS.</p><!><p>The agreement between the measured and calculated DCSs for scattering of both CH 3 and CD 3 radicals with He at respective collision energies of 425 and 440 cm À1 lends considerable condence to the quality of the theoretical treatment outlined in Method Section E. We can therefore derive insights into the scattering dynamics not only from comparison with experiment (as we have done in the case of Fig. 5, 6 and 8), but also by analysis of the calculated, fully state resolved DCSs. The raw experimental images immediately reveal a trend that is borne out by the derived DCSs and by the scattering calculations: for both CD 3 or CH 3 the angular distributions for transitions into n 0 ¼ 2-4 (averaged over k 0 ), where the degree of translational to rotational energy transfer is small, peak in the forward hemisphere, whereas those with n 0 $ 5 (i.e. intermediate to large energy transfer) are predominantly sideways and backwards scattered. The degree of backward scattering increases steadily with increasing n 0 , and hence with Dn, because the initial levels of CD 3 populated have mostly n ¼ 1 (Table 1).</p><p>Similar behaviour is observed in the inelastic scattering of atoms such as He with diatomic molecules, 19,22,23 and reveals that low impact parameter collisions (with, presumably, larger angles of deection) are necessary for large changes in the rotational angular momentum. Classically, the rotational angular momentum of the molecule is induced by a torque that acts on the molecule for the duration of the collision. 70 The magnitude of the torque is proportional to the gradient of the intermolecular potential, which is largest at short range.</p><p>Beyond the Dn dependence of the DCSs, we can also explore the efficiency of changes in the projection quantum number k. The experimental DCSs determined using CD 3 REMPI lines originating from the same n 0 level but corresponding to different spectroscopic branches probe different groups of k 0 projection quantum numbers. Thus, as highlighted earlier, Fig. 5 illustrates the DCSs for transition into n 0 ¼ 4 obtained by O(4), P(4) and S(4) transitions, which probe, respectively, k 0 projections 0-2, 1-3, and 0-4. The varying DCSs obtained indicate clearly that there is a signicant variation with k 0 . The calculations reveal the fully k 0 -resolved dependence: compare, for example, the 1 1 / 4 1 , 4 2 and 4 4 DCSs displayed in Fig. 9 for scattering of both CH 3 and CD 3 .</p><p>In the previous theoretical study, 54 the 3-fold corrugation of the PES resulting from repulsion of the helium atom by the three H atoms on the methyl radical gave rise to a strong</p><p>Þ for final rotational levels with n 0 ¼ 2 and 3 and unresolved final k 0 projection levels as discussed in the text and shown with each image. Y denotes the spectroscopic branch. The rotational angular momentum of the incident CH 3 beam was predominantly n ¼ 0 and 1 (see Table 1 for the complete incident level distribution). The orientation of the relative velocity vector v rel is indicated in one panel. propensity for Dk ¼ AE3 in the integral CH 3 + He inelastic cross sections. To illustrate this corrugation, Fig. 10 presents the atom-molecule separation R at which the interaction energy is equal to 440 cm À1 (the collision energy) as a function of the angles (q He , f He ) that dene the orientation of the He atom (see Fig. 1 of ref. 54). The heavy line in Fig. 10 shows the 3-fold corrugation of the PES for approach of the He atom in the molecular plane.</p><p>It is also interesting to ask how the periodic corrugation of the PES might be manifested in the angular dependence of the state-to-state scattering. Fig. 11 presents computed state-tostate DCSs for scattering of the lowest rotational levels of each nuclear spin modication of CH 3 and CD 3 into the n 0 ¼ 2 and 3 nal rotational levels. The sums of such DCSs weighted by the distribution of populations over initial CD 3 or CH levels given in Table 1, but resolved by nal n 0 and k 0 level, are plotted in the ESI † section. The DCSs plotted in Fig. 11 for the same initial rotational level of the two isotopologues are similar. Consequently, we henceforth concentrate on the CD 3 isotopologue.</p><p>The DCSs for all Dk ¼ 0 transitions have a similar shape, namely a reasonably sharp forward peak and a broad, lower intensity peak in the backward hemisphere. These transitions in all the nuclear spin modications are induced primarily by the v 20 term in the expansion of the PES. 54 By contrast, the |Dk| s 0 transitions all display a broad DCS, starting from zero intensity at q ¼ 0 and extending over the entire angular range, with oscillations at small angles (q # 45 ). In the case of the levels of A 1 and A 2 nuclear spin symmetry, these transitions involve Dk ¼ 3 and are enabled by the strong v 33 term in the angular expansion of the PES, as discussed previously in some detail. 54 For the levels of E symmetry, many Dk ¼ 1 transitions, e.g. 1 1 / 2 2 , can be described as a Dk ¼ AE3 transition from the k ¼ AE1 component of the initial level to the k 0 ¼ H2 component of the nal level. 54 Again, this transition is enabled by the v 33 term. Hence, for the low Dn, Dk s 0 transitions, the shapes of the DCSs are closely connected with the change in the k projection quantum number in the collision.</p><p>To gain further insight into the origin of the differences in the DCSs for Dk ¼ 0 vs. AE3 transitions, we can consider the dependence of the scattering upon the impact parameter. Since the initial and nal rotational quantum numbers n and n 0 are small compared to the orbital angular momentum L, the orbital angular momentum is approximately equal to the total angular momentum J. The partial cross sections (the contribution to the ICS from each value of J) give information about which range of impact parameters contributes to a particular transition. Fig. 12 presents a plot of the partial cross sections from the 1 0 level to the 3 0 and 3 3 levels, and to the higher 5 0 and 6 6 levels. The scattering into both n 0 ¼ 3 nal levels is dominated by the same range of J (i.e., classically, by the same range of impact parameters), but the partial cross sections for the 3 0 level exhibit a bimodal distribution peaking at J ¼ 16 and 33. The high-J and low-J peaks are related to the strong forward and backward peaks seen in the DCS for this transition. In contrast, the partial cross sections for the 3 3 level display a single peak at J ¼ 22. Similarly, for this transition, only a single broad peak appears in the DCS. Also, we observe that the scattering into the higher rotational levels, exemplied here by the 5 0 and 6 6 levels, occurs at smaller values of J (i.e. impact parameters). The torque required for a large Dn transition requires a hard collision at small impact parameters.</p><p>Since the contribution to the integral cross sections into the 3 0 and 3 3 levels occurs over the same range of impact parameters, it is not possible to use the partial cross sections to explain fully the difference in the DCSs for these two levels. Possibly, the orientation of the CD 3 molecule plays a signicant role. The rotational angular momentum of the CD 3 molecule in the 3 3 level is oriented along the C 3 symmetry axis, so that the molecule is rotating perpendicular to this axis. Excitation of this rotational motion will be most readily induced by a collision in which the He atom approaches in the plane of the three D (or three H) atoms (i.e. q He z 90 in Fig. 10). In contrast, in the 3 0 level the CD 3 molecule rotates about an axis perpendicular to the C 3 symmetry axis. Excitation of this motion will require a collision with a He atom approaching out of the hydrogenic plane (i.e. q He away from 90 in Fig. 10).</p><p>For all nal levels measured, the experimental DCSs for inelastic scattering of CD 3 are somewhat more forward peaking than those for CH 3 (see Fig. S7 of ESI † for direct comparisons). As mentioned above and shown in Fig. 9, the state-to-state DCSs of the two isotopologues are, however, very similar. The differences in the measured DCSs may be explained by considering the initial level population in the primary molecular beam. For the CH 3 radical the 1 0 level is missing, and the 0 0 level is the most populated, whereas for CD 3 the most populated levels have n ¼ 1. Accessing a particular nal n 0 level therefore involves a smaller Dn 0 for CD 3 than in the case of CH 3 .</p><p>Fig. 13 shows theoretical DCSs for inelastic scattering of CD 3 out of the various k ¼ 1 and 2 rotational levels with n ¼ 1-3 into levels with n 0 k 0 ¼ 4 4 and 5 5 . Different behaviour is seen, as a function of the initial rotational angular momentum k of the CD 3 molecule. For the 4 4 nal level, the DCSs for k ¼ 1 initial levels peak in the forward hemisphere, while those for k ¼ 2 peak in the backward hemisphere. The former and latter involve Dk ¼ 3 and 6 transitions (the latter probably involving two virtual Dk ¼ 3 transitions), respectively. The situation for the 5 5 nal level is reversed; in this case, transitions out of the k ¼ 1 and 2 initial levels involve Dk ¼ 6 and 3 transitions, respectively. Fig. 14 displays theoretical state-to-state DCSs for inelastic scattering of CD 3 out of the n k ¼ 1 1 rotational level into k 0 ¼ 1 rotational levels with n 0 ¼ 2-7. These Dk ¼ 0 collisions induce additional rotation about an axis perpendicular to the C 3 symmetry axis but do not change the projection k of the rotational angular momentum along the symmetry axis. The corresponding DCSs provide insight into which collisions lead to a change in the magnitude of n while preserving its component along the C 3 axis. We observe that as n 0 (and hence Dn) increases, the bimodal distribution present for the 2 1 and 3 1 nal levels develops into a single peaked backward distribution.</p><p>We can also explore the dependence of scattering on the initial angular momentum by comparison of the dependence on n of state-to-state DCSs for a xed increase in n. Fig. 15 shows theoretical state-to-state DCSs for inelastic scattering of CD 3 by He for Dk ¼ 0 transitions involving different k ¼ 1 initial levels and Dn ¼ 1. When normalized to the maximum value, the DCSs appear remarkably similar. There is very little sensitivity to the initial level, even though the energy transfer associated with these transitions is larger for higher initial n, as can be seen in Table 2.</p><p>The comparisons made in Fig. 9 and 11-15 are far from exhaustive, but provide examples of insights that can be drawn from the state-to-state DCSs. They illustrate the consequence of the angular periodicity in the PES, and the additional complexity that arises when the collision partner is nonlinear.</p><!><p>We have presented an in-depth comparison of differential cross sections for rotationally inelastic scattering of methyl radicals (both CH 3 and CD 3 ) with He. The measurements made use of a newly constructed crossed molecular beam and velocity map imaging instrument. The results presented here represent (to our knowledge) the rst reported determinations of DCSs for inelastic scattering of a polyatomic free radical. The agreement is excellent with the predictions of close-coupling scattering calculations on a recently reported ab initio potential energy surface. CD 3 radical scattering was examined for nal n 0 levels up to n 0 ¼ 9 in the vibrational ground state, whereas experiments for CH 3 radicals were limited to n 0 ¼ 2 and 3. Because of limitations in the REMPI detection scheme, the CM-frame experimental angular distributions represent sums over an unresolved group of k 0 projection levels.</p><p>The accuracy of the calculated, fully state-resolved state-tostate DCSs out of initial states with n k ¼ 0 0 and 1 1 is conrmed by comparison with the less highly resolved experimental data. The theoretical calculations also allow an instructive study of how the features of the underlying PES inuence the relative magnitude of the differential scattering into various n 0 and k 0 nal states.</p><p>The excellent agreement between theory and experiment is very satisfying, but limited (so far) to the single collision energy used here, and to the use of He as a single collision partner. We are initiating comparable experimental and theoretical studies of the scattering of methyl radicals with Ar, H 2 and D 2 to ascertain whether we can achieve a comparable level of agreement and understanding for collisions of CD 3 (CH 3 ) with heavier or more structurally complicated collision partners.</p>

Royal Society of Chemistry (RSC)

Neoadjuvant chemotherapy using platinum- and taxane-based regimens for bulky stage Ib2 to IIb non-squamous cell carcinoma of the uterine cervix

PurposeThere are no reports on the use of neoadjuvant chemotherapy (NAC) in non-squamous cell cervical carcinoma. We examined the effectiveness and safety of paclitaxel/carboplatin (TC) and docetaxel/carboplatin (DC).MethodsStage Ib2 to IIb disease was present in 23 patients scheduled for radical hysterectomy. We administered 1–3 courses of either the TC or the DC regimen. Anti-tumor effects were found superior by Response Evaluation Criteria in Solid Tumors. Safety was assessed with National Cancer Institute Common Terminology Criteria for Adverse Events.ResultsMedian age was 50 years (range 32–63 years), with stage Ib2 in 6 cases (26.1 %) and IIb in 17 cases (73.9 %). Complete response was achieved in 5 cases (21.7 %), partial response in 13 (56.5 %), stable disease in 5 (21.7 %); the response rate was 78.3 %, and surgery completion rate was 78.3 %. Leukopenia or neutropenia ≥grade 3 was seen in 12 (52.2 %) and 21 (91.3 %) cases, respectively, with grade 3 febrile neutropenia in 2 cases (8.7 %) and no anemia or thrombocytopenia ≥grade 3. Median progression-free survival was 26 months (95 % Cl, 13.5–38.5 months); median overall survival was 35 months (95 % Cl, 20.9–49.1 months).ConclusionNAC for non-squamous cell cervical carcinoma showed potent anti-tumor effects and manageable adverse events.

neoadjuvant_chemotherapy_using_platinum-_and_taxane-based_regimens_for_bulky_stage_ib2_to_iib_non-sq

Introduction<!>Subjects<!>Inclusion criteria<!>Exclusion criteria<!>Regimen<!>Criteria for initiating the second course of treatment<!>Carboplatin dose-reduction criteria<!>Paclitaxel dose-reduction criteria<!>Docetaxel dose-reduction criteria<!>Supportive therapy<!>Primary treatment<!>Postoperative therapy<!>Outcome evaluation<!>Evaluation of anti-tumor response<!>Evaluation of adverse events<!>Statistical analysis<!><!>Adverse events<!>Surgery completion and adjuvant therapy<!><!>Discussion<!>

<p>The methods used to treat bulky stage Ib2 to IIb cervical cancers differ between Japan and Western countries. In Western countries, concurrent chemoradiation therapy (CCRT) has been recommended as a standard treatment for such tumors, based on the results of multiple large-scale randomized trials and meta-analyses [1–7]. In Japan, Korea, and Italy, among other countries, the neoadjuvant chemotherapy (NAC) approach has been introduced to clinical practice and is extensively utilized [8, 9]. An Italian phase III, controlled study involving patients with locally advanced stage Ib2 to IIb squamous cell carcinoma of the cervix showed that NAC prior to radical hysterectomy improves patient outcomes compared with conventional radiation therapy alone [10].</p><p>There are no previous reports on the use of NAC for bulky non-squamous cell carcinoma of the cervix. We present the results of an ongoing pilot study on its efficacy and safety.</p><!><p>We studied 23 patients with locally advanced non-squamous cell carcinoma of the uterine cervix (clinical stage Ib2 to IIb) between January 2002 and September 2011. All patients were scheduled to undergo radical hysterectomy and gave informed consent for this study.</p><!><p>The following inclusion criteria were employed: (1) histologically verified non-squamous cell carcinoma of the uterine cervix; (2) locally advanced disease, stage Ib2–IIb; (3) between 20 and 74 years of age; (4) Eastern Cooperative Oncology Group performance status 0–2; (5) no prior treatment; (6) presence of a measurable bulky mass in the uterine cervix on magnetic resonance imaging (MRI); (7) hematologic and biochemical findings within the following parameters, WBC count ≥4,000/mm3, neutrophil count ≥2,000/mm3, platelet count ≥100,000/mm3, hemoglobin ≥10.0 g/dL, AST and ALT levels ≤2 times the upper limit of normal reference range, serum total bilirubin level ≤1.5 mg/dL, serum creatinine ≤1.5 mg/dL, and creatinine clearance ≥60 mL/min; (8) life expectancy ≥6 months; and (9) written informed consent personally given by the subject.</p><!><p>Exclusion criteria were as follows: (1) overt infection; (2) serious complication(s), for example, cardiac disease, poorly controlled diabetes mellitus, malignant hypertension, bleeding tendency; (3) multiple active cancers; (4) interstitial pneumonia or pulmonary fibrosis; (5) pulmonary effusions; (6) history of unstable angina or myocardial infarction within 6 months after registration, or a concurrent serious cardiac arrhythmia requiring treatment; (7) contraindications to treatment with paclitaxel, docetaxel, or carboplatin; (8) intestinal paralysis or ileus; (9) pregnancy, breast-feeding, or desire for future pregnancy; (10) history of serious drug hypersensitivity or drug allergy; and (11) judged unsafe for participation by the attending physician.</p><!><p>The choice of regimen was left to the attending physician. Paclitaxel/carboplatin (TC) therapy was administered to 4 patients and DC therapy to 19 patients. Courses of treatment were administered 21 days apart, with a intravenous paclitaxel dose of 175 mg/m2 or a docetaxel dose of 70 mg/m2 administered on Day 1, and intravenous carboplatin with area under the curve (AUC) 6 mg/mL per min also administered on Day 1. As a rule, maximum 3 courses of treatment were administered to each patient.</p><!><p>The second course was postponed by a maximum of 2 weeks when blood analysis performed within 2 days prior to the planned start did not satisfy the following criteria: (1) neutrophil count ≥1,000/mm3; (2) platelet count ≥75,000/mm3.</p><!><p>The carboplatin dose for the second course was reduced from AUC 6 mg/mL per min to AUC 5 mg/mL per min if the patient experienced grade 4 thrombocytopenia or grade 3 thrombocytopenia accompanied by bleeding. If signs of toxicity remained after this dose reduction, the third course of treatment was reduced to AUC 4.</p><!><p>The paclitaxel dose for the second course was reduced from 175 to 135 mg/m2 in patients exhibiting grade 2 or higher severe peripheral nerve toxicity during the first course. If this grade of nerve toxicity persisted after the dose reduction, the paclitaxel dose for the third course was reduced to 110 mg/m2.</p><!><p>The docetaxel dose for the second course was reduced from 70 to 60 mg/m2 if the patient experienced grade 4 neutropenia lasting 7 days or longer or febrile neutropenia lasting 4 days or longer. If signs of toxicity remained after this dose reduction, the docetaxel dose for the third course was reduced to 50 mg/m2.</p><!><p>A granulocyte colony-stimulating factor (G-CSF) preparation was administered to patients who developed grade 4 neutropenia during the first course of NAC. These patients were permitted prophylactic G-CSF during the second and subsequent courses of NAC. Anti-emetics were used for the preventive purpose.</p><!><p>Patients with stage Ib2–IIb cervical carcinoma underwent radical hysterectomy unless the tumor responded to preoperative treatment with progressive disease (PD), at which time the tumor was up-staged. In cases in which surgery was not possible, CCRT was adopted.</p><!><p>Postoperative radiotherapy, postoperative chemotherapy, or CCRT was additionally administered in patients with a positive surgical margin at the vaginal stump, lymphadenopathy, invasion of the cardinal ligament, or evident invasion of the vasculature.</p><!><p>The primary endpoint was anti-tumor response. Secondary endpoints comprised adverse events, the surgery completion rate, the progression-free survival (PFS) period, and the overall survival (OS) period. Hematologic tests and urinalysis were performed before the start of treatment and, as a rule, once weekly after starting treatment. Electrocardiograms and chest radiographs were obtained before the start and at the end of treatment.</p><!><p>Anti-tumor response was evaluated using Response Evaluation Criteria in Solid Tumors guidelines. The baseline MRI findings were compared with the findings at the conclusion of treatment. For our efficacy evaluation, we adopted the best rating, without incorporating the response period.</p><!><p>Adverse events were evaluated employing the National Cancer Institute Common Terminology Criteria for Adverse Events, version 3.0.</p><!><p>Progression-free survival (PFS), defined as the time from the start of the study treatment to documented tumor progression or death, and overall survival (OS), defined as the time from the start of treatment to the date of death, were calculated by the Kaplan–Meier method. The statistical data were obtained using StatMate III.</p><!><p>Patient characteristics (n = 23)</p><p>DC docetaxel + carboplatin, TC paclitaxel + carboplatin</p><p>Response</p><p>CR complete response, PR partial response, SD stable disease, PD progressive disease, TC paclitaxel + carboplatin, DC docetaxel + carboplatin</p><p>Adverse events of TC/DC therapy</p><p>TC paclitaxel + carboplatin, DC docetaxel + carboplatin</p><!><p>In 3 cases (13.0 %), the second course of treatment was postponed due to a low neutrophil count; in all 3 patients, the second course was initiated within 7 days of its scheduled time. Both patients (10.0 %) with grade 3 febrile neutropenia for 4 days or longer had received DC therapy prior to the development of this complication. In these 2 cases, doses were reduced for the second course of treatment: docetaxel from 70 to 60 mg/m2 and carboplatin from AUC 6 to AUC 5.</p><!><p>Radical hysterectomy after NAC was completed in 18 of the 23 patients, giving a surgery completion rate of 78.3 %. Adjuvant therapy after radical hysterectomy consisted of no treatment in 3 cases (13.0 %), radiotherapy in 2 cases (8.7 %), chemotherapy in 15 cases (65.2 %), and CCRT in 3 cases (13.0 %).</p><!><p>Kaplan–Meier curves for progression-free survival (a) and overall survival (b). The median PFS for all patients was 26 months (95 % CI, 13.5–38.5 months), and the median OS was 35 months (95 % CI, 20.9–49.1 months)</p><!><p>The incidence of non-squamous cell carcinoma of the uterine cervix has been steadily rising in Japan, currently accounting for approximately 10–15 % of all cervical cancer cases. Lymph node metastasis is more frequent with this disease, compared with invasive squamous cell carcinoma [11], and its sensitivity to radiotherapy and chemotherapy is considered to be lower [12]. Thus, squamous and non-squamous cell carcinomas must be analyzed separately. It is advisable and desirable to try new therapeutic strategies in non-squamous cell carcinoma, but the number of published studies involving this type of cervical cancer is small, with the number of cases analyzed in these reports also small. Thus, no high-level evidence regarding treatment has been obtained for this type of cervical carcinoma.</p><p>The response rates of adenocarcinoma are reportedly 20 % to cisplatin [13], 15 % to ifosfamide [14], 14 % to 5-fluorouracil [15], and 12 % to oral etoposide [16]; these response rates are lower than those of squamous cell carcinoma. According to Curtin et al. [17], however, the response rate of adenocarcinoma to paclitaxel is as high as 31 %, even when the agent is used independently. Docetaxel has also been attracting interest as an agent of NAC. Nagao et al. evaluated the efficacy of combined chemotherapy using a DC regimen (docetaxel 60 mg/m2 and carboplatin at AUC 6 on day 1, repeating the combination every 21 days) in 17 patients with advanced or recurrent cervical cancer, including 6 with adenocarcinoma and 1 with adenosquamous carcinoma. A partial response was obtained in 6 of the 7 cases with adenocarcinoma (including the case of adenosquamous carcinoma); the response rate was 86 % [18]. Considering these findings, we conducted a pilot study involving standard regimens of TC and DC, conventionally used for the treatment of ovarian cancer.</p><p>In the analysis of adverse events, severe neutropenia developed in 91.3 % of patients, but subsided in response to short-term treatment with a G-CSF preparation. During the first course of DC therapy, grade 3 febrile neutropenia developed in 2 cases; the dose of both agents was reduced for the next course of treatment. All signs, specific to taxanes, of peripheral neuropathy were grade 1 or less, allowing for continuation of treatment while preserving the quality of life of the individual patients. No serious adverse events occurred, and the response rate was 78.3 %. This study demonstrated a high response rate of bulky non-squamous cell carcinoma of the cervix to NAC using taxanes (paclitaxel or docetaxel) and carboplatin. It also demonstrated the safety of the medications in this regimen. The completion rate of radical hysterectomy, however, was only 78.3 %; thus, the treatment outcomes in this study were not satisfactory. Possible reasons for the low surgery completion rate include the rapid progression of non-squamous cell carcinoma, frequent invasion of tissues and organs surrounding the uterus, and frequent lymph node metastasis.</p><!><p>Treatment results and outcomes of all patients</p><p>ASC adenosquamous cell carcinoma, MAC mucinous adenocarcinoma, EDC endometrioid adenocarcinoma, CCC clear cell adenocarcinoma, DC docetaxel + carboplatin, TC paclitaxel + carboplatin, CR complete response, PR partial response, SD stable disease, NT no treatment, CT chemotherapy, RT radiotherapy, CCRT concurrent chemoradiation therapy, PFS progression-free survival, OS overall survival, NED no evidence of disease, AWD alive with disease, DOD died of disease</p>

PubMed Open Access

Identification of a Maleimide-Based Glycogen Synthase Kinase-3 (GSK-3) Inhibitor, BIP-135, that Prolongs the Median Survival Time of \xce\x947 SMA KO Mouse Model of Spinal Muscular Atrophy

The discovery of upregulated glycogen synthase kinase-3 (GSK-3) in various pathological conditions has led to the development of a host of chemically diverse small molecule GSK-3 inhibitors, such as BIP-135. GSK-3 inhibition emerged as an alternative therapeutic target for treating spinal muscular atrophy (SMA) when a number of GSK-3 inhibitors were shown to elevate survival motor neuron (SMN) levels in vitro and to rescue motor neurons when their intrinsic SMN level was diminished by SMN-specific short hairpin RNA (shRNA). Despite their cellular potency, the in vivo efficacy of GSK-3 inhibitors has yet to be evaluated in an animal model of SMA. Herein, we disclose that a potent and reasonably selective GSK-3 inhibitor, namely BIP-135, was tested in a transgenic \xce\x947 SMA KO mouse model of SMA, and found to prolong the median survival of these animals. In addition, this compound was shown to elevate the SMN protein level in SMA patient-derived fibroblast cells as determined by western blot, and was neuroprotective in a cell-based, SMA-related model of oxidative stress-induced neurodegeneration.

identification_of_a_maleimide-based_glycogen_synthase_kinase-3_(gsk-3)_inhibitor,_bip-135,_that_prol

<!>Effects of BIP-135 in a Model of Oxidative Stress<!>SMA Animal Studies<!>Drug<!>Kinase Selectivity<!>Animals<!>Motor Function Tests<!>Statistical analysis<!>Primary Neurons, Cell Cultures, and Neuronal Viability Assays (Oxidative Stress-Induced Neurodegeneration)<!>Intracellular Total Glutathione Measurements<!>SMA Patient Fibroblast, Cell Cultures, Immunostaining and Immunoblot Analysis<!>

<p>Over the years, glycogen synthase kinase-3 (GSK-3), a member of the serine/threonine kinase family, has been extensively studied as a drug target, because upregulated GSK-3 has been linked to a number of human pathological conditions.1 This enzyme is known to regulate a diverse array of intracellular processes through the phosphorylation of its protein substrates.2 In mammals, GSK-3 is expressed in two highly homologous isoforms, namely GSK-3α and GSK-3β. This multifunctional kinase is constitutively active in resting cells and can be physiologically inhibited by various signaling pathways, including the PI3K/Akt mediated apoptosis cascade in stimulated cells through phosphorylation of residues Ser-21/Ser-9 (GSK-3α/β).3 Inhibition of GSK-3 results in the activation of its downstream constituents, such as β-catenin, c-Jun, and the cyclic AMP response element binding protein (CREB), which consequently upregulate Tcf/Lef gene transcription and CREB-induced gene transcription of neurotrophic factors, such as brain-derived neurotrophic factor (BDNF).4,5 BDNF helps to support the survival of existing neurons and promote neurogenesis.6 Thus, GSK-3 inhibition has shown neuroprotective effects in various models of neurodegenerative disease.7,8 Motor neuron diseases (MNDs), such as amyotrophic lateral sclerosis (ALS), also experienced neuroprotection in vitro.9 Significant delay in symptom onset and an increase in survival time was observed in an animal model of ALS when a GSK-3 inhibitor was administered to the animals.10 As such, neuroprotection, offered by GSK-3 inhibition, is a validated therapeutic strategy in treating many other MNDs, such as spinal muscular atrophy (SMA).</p><p>SMA is a major cause of lethality among infants in their early stage of life, and it remains as an untreatable disease. Several therapeutic strategies are currently under investigation for correcting some of the symptoms associated with SMA.11,12 A characteristic defect that occurs in SMA is a deficiency in the survival motor neuron (SMN) protein, constituted by the mutation or deletion of the survival motor neuron 1 (SMN1) gene responsible for SMN protein expression. Its absence progresses into the degeneration of the lower motor neurons found in the anterior horn of the spinal cord that is commonly found in SMA patients.11,12 The severity of SMA in patients is inversely dependent on the expression level of SMN2, a paralog of SMN1.12 All SMA patients contain at least one copy of SMN2 and thus, pharmacologic agents capable of elevating endogenous level of SMN2-derived SMN proteins would be highly desirable for the development of SMA therapeutics. 11,12</p><p>Recently, inhibition of GSK-3 was shown to elevate cellular SMN levels in SMA type I (a severe form of SMA often diagnosed before 6 months of age) patient-derived fibroblasts and also in motor neurons in which the intrinsic SMN levels were diminished by SMN-specific short hairpin RNA (sHRNA).13 The connection between GSK-3 inhibition and SMA was first demonstrated in an image-based screen, conducted by Rubin and coworkers, to identify compounds capable of elevating SMN level in vitro. Independently of these findings, and prior to becoming aware of the then unpublished research of Rubin, we had submitted several of our GSK-3 inhibitors for screening to the SMA Foundation, and received reports showing their efficacy in animal models. We now disclose the results of these SMA animal experiments, along with in vitro biology conducted by the Rubin group.</p><p>To date, a plethora of GSK-3 inhibitors have been reported, a few of which are shown in Figure 1A.14–17 Additionally, we have disclosed a variety of selective, maleimide-bearing GSK-3 inhibitors that are active in both cell and animal models of neurological and psychiatric disorders, including Parkinson's disease and bipolar disorder (Figure 1B).16–18 Based upon the usual cycle of design, synthesis, and kinase testing, we identified a potent ATP-competitive GSK-3 inhibitor, namely 3-(5-bromo-1-methyl-1H-indol-3-yl)-4-(benzofuran-3-yl)pyrrole-2,5-dione, or BIP-135. Ultimately, this compound was tested against a total of 62 kinases (see supporting information).16BIP-135 was found to be relatively selective for GSK-3β (21 nM). However, due to the high sequence homology between GSK-3α and GSK-3β, it was not surprising to find similar inhibitory activity against both isoforms. Among all the other kinases tested, this compound was found to show some activity toward PKCβ (β1: 980 nM; β2: 219 nM), DYRK1B (590 nM), and PI3Kα (870 nM). Given the structural resemblance of BIP-135 to known PKCβ inhibitors of the staurosporine class and the high homology between PKCβ1 and β2, this off-target activity was not unexpected.19 In the case of the modest inhibition shown by BIP-135 of DYRK1B, this may not be a concern. A number of articles on DYRK1B suggest that it plays a key role in cancer biology, as well as in muscle differentiation.20 In particular, overexpression of the DYRK1B gene appears to be associated with pancreatic cancers as a consequence of its downstream effect on oncogenic K-ras.20 As for PI3Kα, it is one of a number of isozymes in the PI3K family that can activate Akt. The activation of Akt has been shown to downregulate FOXO, a protein that is over-expressed in type I SMA that contributes to muscle atrophy.21 As a weak inhibitor of PI3Kα, it is unlikely that BIP-135 can influence the expression level of the FOXO protein.</p><p>As reported previously by Rubin, a set of commercially available GSK-3 inhibitors including alsterpaullone, CHIR98014, and AR-A014418, were shown to elevate SMN levels in vitro.13 Under the same conditions, western blot analysis employing BIP-135 led to an elevation in SMN protein levels. Exposure of human SMA patient fibroblasts to a dose of 25 µM of the test compound for 72 h led to a 7-fold increase in SMN levels compared to vehicle treated cells (Figure 2). However, the typical bell-shaped dose-response curve was observed due to the observation of some toxicity at higher concentrations.</p><!><p>As oxidative stress also appears to play a role in the dysfunction of motor neurons in SMA, we accordingly sought to explore the possibility that BIP-135 might also work in preventing cell loss under such conditions.22–24 Specifically, it has been suggested that free radicals generated under conditions of oxidative stress lead to the production of reactive lipid aldehydes, resulting in increased levels of cell damage and cellular death in cells that have lower SMN levels.25 Thus, to examine the therapeutic efficacy of BIP-135 in a neuronal model of oxidative stress, primary immature cortical neurons were exposed to the glutamate analog, homocysteic acid (HCA; 5 mM), which was used to deplete the cortical neurons of the antioxidant glutathione. Since glutathione is a major cellular antioxidant, its depletion allows the gradual accumulation of endogenouslyproduced oxidants, thereby inducing neuronal degeneration over a period of approximately 24 h.23 Using this model of oxidative stress, the survival of the HCA-treated cortical neurons was reduced to approximately 25 percent (Figure 3). On the other hand, the BIP-135 treated cells were protected to the extent of about 80% at a concentration of 20 µM. No significant toxicity was observed when the neurons were exposed to BIP-135 alone, in the absence of HCA (Figure 3A). Other commercially available GSK-3 inhibitors, such as AR-A011418, also provided some protection in this model; however, BIP-135 was able to protect at a lower concentration, demonstrating that BIP-135 is a superior neuroprotective agent in this model of oxidative stress (Figure 3B and C). Micrographs taken of the cortical neurons after 24 h exposure to the GSK-3 inhibitors, BIP-135, SB-216763, or AR-A011418, either alone or in combination with HCA are shown in Figure 3D. These images visually reflect the important neuroprotective effect of BIP-135.</p><p>The dependence of this oxidative stress model on impaired cysteine/cystine transport and glutathione depletion is well-documented.26 However, one could argue that the observed neuroprotective effects of BIP-135 may be due to the inhibition of one or both of these two processes. To rule out these possibilities, the total intracellular glutathione levels (reduced glutathione and oxidized glutathione) were measured after the neurons were treated with HCA, with or without the GSK-3 inhibitors. The data obtained revealed that the GSK-3 inhibitors, did not alter the overall glutathione levels (Figure 4), demonstrating that the observed neuroprotection afforded by BIP-135 is independent of antioxidant production. A possible explanation to the observed neuroprotection is the activation of BDNF by GSK-3 inhibition, which has offered neuroprotection in many other models of neurodegenerative diseases.6 This compound may also increased neuron's resistance to oxidative damage by disrupting the regulatory role of GSK-3 in destabilizing the anti-apoptotic protein Bcl-2 in the intrinsic apoptotic pathway6 and thus, enhanced the overall Bcl-2 level that leads to neuronal survival.27,28 These findings further support the possible therapeutic use of BIP-135 in SMA, at least in an animal model, where Bcl-2 deficiency was found on the spinal cord of SMA mice.28 In fact, increasing evidence has suggested that neuronal oxidative stress plays a significant role in cell death and dysfunction associated with neurological diseases, including SMA.23,24 Thus, neuroprotection in this oxidative model, demonstrated by BIP-135, may also be a useful predictor in qualifying other GSK-3 inhibitors for use in treating SMA.</p><!><p>While the in vitro results are promising, the possibility to move a compound like BIP-135 forward to the clinic requires a demonstration that it works in animal models of the disease. To date, animal models of SMA have been limited in large part because SMA is a disease exclusive to humans. In general, murine models of SMA often require homozygous mutation or deletion of the Smn gene, followed by insertion of the SMN2 gene with or without SMN2 cDNA lacking exon 7 (SMN2Δ7+/+).29 This will ensure that the correct phenotypes of the SMA disease are present in the mice. In the present study, the latter murine species (with SMNΔ7+/+) was used to analyze the in vivo effect of BIP-135. In addition, this particular strain of SMA transgenic mice mimics the phenotypes that resembles the human SMA Type I disease.30</p><p>In the Δ7 SMA neonatal mouse model of SMA30,31, the impact of BIP-135 on several phenotypes observed within the human SMA disease, such as (1) loss of motor function; (2) body weight deficiency; and (3) survival, were investigated using three different doses (25, 75, and 125 mg/kg). Among these doses, the in vivo potency of BIP-135 is most promising at 75 mg/kg (intraperitoneal injection, 100% DMSO as vehicle). At this dose, this GSK-3 inhibitor did not appear to be toxic and was well-tolerated by the animals (no decrease in body weight) (Figure 5B). More importantly, BIP-135 caused a modest extension in the median survival of SMA KO animals by two days, suggesting a valid in vivo protective effect for this compound when the median lifespan of these animals was approximately 14 days (Figure 5A). The motor function of the transgenic animals was evaluated based on their performances in the geotaxis and tube test (Figure 5C and D).31 In brief, geotaxis test examined the ability of the animals to orient itself from a downward- to an upward-facing position when placed on an inclined platform. Tube test analyzed the hind-limb strength of the animals when rising from a laying to a standing position. Based on the results gathered from the geotaxis and tube test, BIP-135 did not improve the overall motor function, but it did increase the number of SMA KO animals that completed the geotaxis test (Figure 5C, b). Due to the short lifespan of these animals (≤ 14 days), it may be too difficult to detect motor function improvement from the BIP-135 treatment, especially when the endpoint of the model is the death of the animals. A mouse model with milder SMA phenotypes (type III or IV) and longer lifespan may be more amenable to evaluate the therapeutic effect of BIP-135 in SMA, as well as its effect on SMN protein level in vivo.32</p><p>In summary, we have demonstrated that BIP-135 is a potent and reasonably selective GSK-3 inhibitor that is neuroprotective in a cortical neuron model of oxidative stress. More importantly, BIP-135 was able to elevate SMN protein levels in vitro and was found to extend the median survival period of transgenic mice bearing a severe SMA phenotype. This is the first report of a GSK-3 inhibitor to exert protective effects in an animal model of SMA, and this finding will hopefully support other efforts to test validated GSK-3 inhibitors in SMA. In addition, the evaluation of BIP-135 in a less severe mouse model of SMA is being pursued.</p><!><p>To examine the in vitro and in vivo effects of BIP-135 in SMA, this compound was re-synthesized using our previously reported methods (see supporting information).15 The staurosporine-related maleimide BIP-135 is a potent GSK-3 inhibitor (IC50 = 7–21 nM, tested in the presence of 10 µM ATP). The overall purity of BIP-135 used in both the in vitro and in vivo experiments exceeds 98% according to high-performance liquid chromatography.</p><!><p>All kinase inhibition assays were conducted at Reaction Biology Corporation, Inc. (http://www.reactionbiology.com).</p><!><p>Male and female SMN2+/+;SMN2Δ7+/+;Smn+/− (heterozygote knockout for Smn gene, HET) mice were purchased from Jackson laboratories, Bay Harbor, ME, USA (stock number 5025) and were bred to generate a self-sustaining colony of SMN2+/+;SMN2Δ7+/+;Smn+/− breeder mice. The breeder mice then generated the SMN2+/+; SMN2Δ7+/+; Smn−/− (SMA model mice, homozygote knockout for Smn, KO), as well as the SMN2+/+; SMN2Δ7+/+; Smn+/+ (wild type for Smn gene, WT) and SMN2+/+; SMN2Δ7+/+; Smn+/− (HET) control mice for behavioral phenotyping. All mice included in the present study were homozygous for human SMN2+/+ and SMN2Δ7+/+. One male was housed with 1–3 female mice until vaginal plugs were observed. The male was then removed. Pregnant females were housed individually in Plexiglas cages and were provided with nesting materials and enriched environments containing a plastic igloo, a flexible gnaw bone and 'Envirodri' bedding. Food and water were available ad libitum. All mice were maintained at a temperature of 21 °C on a 12 h light/dark cycle. All studies were approved by an Institutional Animal Care and Use Committee (IACUC) established at PsychoGenics Inc. according to the rules set out by the Public Health Service Office of Laboratory Animal Welfare.</p><p>At birth (defined as postnatal day 0, or P0), litters were randomly culled to 10 with equal numbers of males and females removed. Pups were tattooed using non-toxic ink applied under the skin and a tail snip sample was taken for genotyping. Genotyping was performed by Transnetyx Inc. Genotype data were normally available within 48 h after birth. Once the genotype results are known, the litters were further culled to a maximum of 8 pups per litter by removing the HET animals at P3. Litters with less than 6 pups at P3 were voided (thus litter size used ranged from 6–8 pups). Both body weight and survival were monitored daily for these mice. Mice were dosed once a day between 08:30 to 09:30 AM via intraperitoneal injection (IP) starting at P3 and continued until the KO pups died. Body weights for unused (not dosed) littermate WT and HET were taken at P10, P12 and P14 only to monitor the litter and dam overall health. Motor function assessments (negative geotaxis followed by the tube test) were performed in the afternoon at least 4 hours post morning drug injection. The study end point for each KO mouse was death of the animal.</p><p>The Δ7 SMA mouse model closely mimics the human SMA genotype and results in mouse phenotype that resembles the human SMA Type I disease. A battery of tests was used to assess the effect of GSK-3 inhibition on the body weight, life span, and motor function (negative geotaxis and hind limb suspension tests, a.k.a. tube test) in the KO animals. Ten KO females and 13 males were treated with the GSK-3 inhibitor BIP-135 at 75 mg/kg and 12 females and 13 males with vehicle (100% DMSO). As much as possible, littermate KO animals received different treatments. BIP-135 was dissolved in DMSO and injected once per day with a dosing volume of 2.5 ml/kg, IP. One female KO treated with BIP-135 was found dead at P4 (i.e. a day after receiving the first dose). This animal showed significant delay in body weight growth prior to the commencement of the treatment and thus was excluded from the analysis.</p><!><p>Motor function tests (negative geotaxis and tube test) were performed, as previously described.31</p><!><p>Survival evaluation in the SMA study was performed using Kaplan-Meier analysis with Logrank (Mantel-Cox) and Breslow-Gehan-Wilcoxon as the post-hoc tests. Body weight, negative geotaxis and tube test parameters were analyzed using the Mixed Effects Model (also known as Mixed ANOVA Model) which is more robust to missing values caused by fatalities over time, and is based on likelihood estimation rather than moment estimation as in the typical repeated-measures ANOVA analysis. The treatment and gender were analyzed as independent and trials and age as dependent factors. Mixed model ANOVA was followed by simple effect and Tukey's post hoc tests when indicated. ANOVAs were performed using the PROC MIXED procedure in SAS 9.1.3 (SAS Institute, Cary, NC). Values are presented as mean ± SEM. A p value of < 0.05 was considered statistically significant.</p><!><p>Primary neurons, cell cultures, and neuronal viability assays were prepared and performed, as previously described.26,33</p><!><p>Intracellular total glutathione [glutathione (GSH) + oxidized glutathione (GSSG)] measurements of primary neuron cultures were determined using the GSH-Glo Glutathione Assay kit, (Promega) according to the manufacturer's protocol.26</p><!><p>SMA patient fibroblast, cell cultures, immununostaining and immunoblot analysis were prepared and performed, as previously described.13</p><!><p>spinal muscular atrophy</p><p>glycogen synthase kinase-3</p><p>motor neuron disease</p><p>amyotrophic lateral sclerosis</p><p>survival motor neuron 1</p><p>full length</p><p>survival motor neuron 2 cDNA lacking exon 7</p><p>embryonic stem</p><p>homocysteic acid</p><p>fast axonal transport</p><p>fetal bovine serum</p><p>glutathione</p><p>oxidized glutathione</p><p>intraperitoneal</p><p>cyclic AMP response element binding protein</p><p>brain-derived neurotrophic factor</p><p> Author Contributions </p><p>Synthesis, purification, and kinase selectivity of BIP-135 were completed by P.C.C. and I.N.G. Preparation and phenotypic analysis of the SMA transgenic mice were performed by B.F.E. and S.R. Immunostaining and immunoblotting of SMA patient-derived fibroblast were accomplished by N.R.M and L.L.R. Experiment design, data analysis, writing, and editing were completed by P.C.C., I.N.G., B.F.E, S.R., N.R.M., L.L.R., and A.P.K.</p><p> ASSOCIATED CONTENT </p><p> Supporting Information </p><p>Synthetic scheme and kinase selectivity table of BIP-135. This information is available free of charge via the Internet at http://pubs.acs.org.</p>

PubMed Author Manuscript

Morphology, Activation, and Metal Substitution Effects of AlPO4-5 for CO2 Pressure Swing Adsorption

Aluminophosphate, AlPO4-5, an AFI zeotype framework consisting of one-dimensional parallel micropores, and metal-substituted AlPO4-5 were prepared and studied for CO2 adsorption. Preparation of AlPO4-5 by using different activation methods (calcination and pyrolysis), incorporation of different metals/ions (Fe, Mg, Co, and Si) into the framework using various concentrations, and manipulation of the reaction mixture dilution rate and resulting crystal morphology were examined in relation to the CO2 adsorption performance. Among the various metal-substituted analogs, FeAPO-5 was found to exhibit the highest CO2 capacity at all pressures tested (up to 4 bar). Among the Fe-substituted samples, xFeAPO-5, with x being the Fe/Al2O3 molar ratio in the synthesis mixture (range of 2.5:100–10:100), 5FeAPO-5 exhibited the highest capacity (1.8 mmol/g at 4 bar, 25°C) with an isosteric heat of adsorption of 23 kJ/mol for 0.08–0.36 mmol/g of CO2 loading. This sample also contained the minimum portion of extra-framework or clustered iron and the highest mesoporosity. Low water content in the synthesis gel led to the formation of spherical agglomerates of small 2D-like crystallites that exhibited higher adsorption capacity compared to columnar-like crystals produced by employing more dilute mixtures. CO2 adsorption kinetics was found to follow a pseudo–first-order model. The robust nature of AlPO4-5–based adsorbents, their unique one-dimensional pore configuration, fast kinetics, and low heat of adsorption make them promising for pressure swing adsorption of CO2 at industrial scale.

morphology,_activation,_and_metal_substitution_effects_of_alpo4-5_for_co2_pressure_swing_adsorption

Introduction<!>Materials<!>Growth of AlPO4-5 and MeAPO-5 Adsorbents<!>Characterization<!>CO2 Adsorption<!>Activation of Adsorbents by Calcination and Pyrolysis<!><!>Activation of Adsorbents by Calcination and Pyrolysis<!><!>Activation of Adsorbents by Calcination and Pyrolysis<!><!>Effect of Metal Substitution<!><!>Effect of Metal Substitution<!><!>Effect of Metal Substitution<!><!>Effect of Metal Substitution<!>Effect of Metal Content<!><!>Effect of Metal Content<!><!>Effect of Metal Content<!><!>Effect of Reaction Mixture Dilution and Associated AlPO4-5 Morphology<!><!>Effect of Reaction Mixture Dilution and Associated AlPO4-5 Morphology<!><!>Effect of Reaction Mixture Dilution and Associated AlPO4-5 Morphology<!><!>Isosteric Heat of Adsorption<!><!>Isosteric Heat of Adsorption<!>CO2 Adsorption Kinetics<!><!>CO2 Adsorption Kinetics<!>Conclusions<!>Data Availability Statement<!>Author Contributions<!>Conflict of Interest

<p>The severe environmental concern related to the greenhouse effects is mainly attributed to gases that absorb and emit radiation within the thermal infrared range. The primary greenhouse gases are carbon dioxide, methane, water vapor, nitrous oxide, and ozone. CO2 is the most important contributor coming mainly from combustion of fossil fuels and being released into the atmosphere from the power plants, steel plants, cement industry, and other large-scale industrial operations, as well as from transportation (Kontos et al., 2014). Notably, the increase of total primary energy supply, reaching 150% in 2014 compared to 1971, is caused by the associated worldwide economic growth with fossil fuels holding the most significant share for that energy produced (Metz et al., 2005; IEA, 2016; Tabish et al., 2020). Under these circumstances, it is evident that carbon dioxide emissions will keep increasing as energy demands will continue being covered primarily by fossil fuels at least for the next 50 years.</p><p>CO2 capture, utilization, and storage (CCUS) technologies have been proposed to stabilize the concentrations of greenhouses gases in the atmosphere and mitigate their impact. Among other proposed solutions, such as retrofits of existing units, usage of fuels with less carbon dioxide footprint, and nuclear and renewable energy, CCUS is the most promising one in terms of compatibility with the energy production, continued dependence on fossil fuels, and delivery infrastructure. Focusing on postcombustion, different technologies have been used. Employing amine-based solvents is the most mature technology. Monoethanolamine, diethanolamine, and N-methyldiethanolamine are commonly used alkanolamines. These lean amines, commonly diluted along with water of content in the order of about 70%, have a high reactivity toward CO2. Still, the technology has certain drawbacks mainly associated with high energy consumption for the regeneration step, requirement of large voluminous equipment, high corrosion rates, and solvents slippage to the atmosphere (Resnik et al., 2004; Haszeldine, 2009). As promising alternatives to solvent-based systems, which have constituted the main industrial practice for several decades, technologies based on adsorption and membrane separation are gaining considerable attention (Pilatos et al., 2010; Labropoulos et al., 2015; Kueh et al., 2018). Concerning adsorption, emphasis is put on porous materials. These can be divided into two categories: (i) physical adsorbents, such as porous carbons, zeolites, and metal–organic frameworks, and (ii) chemical adsorbents, such as functionalized materials with surface agents such as amine moieties (Choi et al., 2009; Sayari et al., 2011; Pokhrel et al., 2018; Varghese and Karanikolos, 2020). In this front, various materials are being discovered and explored, yet a platform of suitable materials/systems to treat a wide range of industrial emissions at a large scale still remains a challenge. The reason is that multiple factors need to be met at the same time, i.e., adsorbents need to exhibit (i) high capacity; selectivity compared to other components such as NOx, SO2, and H2O vapor; fast adsorption/desorption kinetics; and low energy consumption; and (ii) chemical and thermal stability, sustainable performance for many cycles, low manufacturing cost, and mechanical robustness at large scale.</p><p>Zeolites have been studied for CO2 capture, particularly involving dry CO2, based on their relatively high adsorption capacity (Chue et al., 1995; Sircar and Golden, 1995; Siriwardane et al., 2001; Chou and Chen, 2004), low-cost of production, and excellent thermal stability (Musyoka et al., 2015). Yet, due to polar zeolite surfaces, these materials are highly hydrophilic resulting in lower CO2 capture capacity and selectivity and early saturation in the presence of moisture (Corma, 2003). To battle this problem, less hydrophilic zeolite-type materials need to be explored. Aluminophosphates (AlPOs) are among the most notable adsorbents for this duty. AlPOs were discovered in 1982 by Wilson and partners of Union Carbide (Wilson et al., 1982). The framework of AlPOs, such as AlPO4-5, consists of alternating Al3+ and P5+ connected by oxygen atoms (Wilson et al., 1982; Stoeger et al., 2012). The main feature is the charge neutrality of the framework, which occurs from the equal ratio of alumina to phosphorus (Al/P = 1). This constant ratio producing materials of a net neutral electric charge prevents ion exchange, and as a result, AlPOs tend to be only slightly hydrophilic due to the absence of acidic sites (Cundy and Cox, 2003; Carmine, 2012). Indeed, in our recent work, we showed that AlPO4-5 is rather hydrophobic particularly at relatively low water partial pressures, where water molecules occupy niches close to pore walls, followed later by the filling of the central pore area (Schlegel et al., 2018). However, AlPOs retain their CO2 sorption capacity potential due to the high concentration of physisorption sites. Therefore, the limited hydrophilicity, concentration of physisorption sites, and more linear shape of the CO2 adsorption isotherms as compared to other zeolites suggest that these materials would have a longer process lifetime extending over many adsorption/regeneration cycles by pressure-mediated tuning between adsorption/desorption cycles [pressure swing adsorption (PSA)]. Liu et al. (2011) examined 8-member ring AlPOs (AlPO4-17, AlPO4-18, AlPO4-53, and AlPO4-25) at different temperatures and determined CO2 uptake capacities in the range of 1.52–2.32 mmol/g at 273–293 K and 100 kPa. Among the above structures, AlPO4-53 possessed a higher CO2 affinity and a CO2/N2 selectivity of 98.4 with low interaction with water molecules as compared to the benchmark zeolite 13X. Zhao et al. (2009) investigated AlPO4-14 and reported CO2 adsorption capacities ranging from 2.0 to 2.7 mmol/g within the temperature range of 273–300 K at 100 kPa, exhibiting also relatively high CO2 over CH4 selectivity. Delgado et al. (2013) studied the adsorption behavior of AlPO4-11 for CO2, CH4 and N2 and reported a CO2 adsorption capacity of 0.31-0.7 mmol/g at 298–338 K and 100 kPa.</p><p>In addition to high adsorption capacity and selectivity, fast adsorption/desorption kinetics and a low heat of adsorption are key factors for an industrially prominent adsorbent candidate in PSA CO2 capture applications. The work reported herein studies the synthesis and modification/functionalization of AlPO4-5 and its metal substituted analogs as potential candidates for CO2 capture. The crystal lattice of AlPO4-5 possesses a hexagonal symmetry and a monodirectional channel morphology extending along the c-axis. The main channels are created by 12-member rings of alternating tetrahedra of [AlO4]− and [PO4]+ having a diameter of 7.2 Å (Rajic and Kaucic, 2002; Guo et al., 2005; Karanikolos et al., 2008). Here, we assess the impact of various factors on CO2 sorption, namely, (a) the effect of various heteroatoms used in isomorphic substitution in the AlPO4-5 framework at varying concentrations, (b) the pore activation method and in particular the affinity of remnant carbon species present into the inner pore surface after structure-directing agent (SDA) removal by two different thermal treatment methods, i.e., partial oxidation and pyrolysis, and (c) the impact of AlPO4-5 crystal morphology and hydrothermal synthesis mixture composition and in particular the water content in the mixture.</p><!><p>Aluminum isopropoxide (Merck), orthophosphoric acid (85% in H2O, Sigma Aldrich), and triethylamine (TEA, Merck) were used as precursors for AlPO4-5 growth. Tetraethyl orthosilicate (Merck), magnesium chloride (Merck), cobalt (II) acetate tetrahydrate (98%, Sigma-Aldrich), and iron (III) nitrate non-ahydrate (Merck) were used as metal precursors for the synthesis of the metal-substituted AlPOs (MeAPO-5).</p><!><p>The AlPO materials were grown hydrothermally from reaction mixtures starting from a gel composition of 1Al2O3:1.3P2O5:1.2TEA:xMe:yH2O, where x and y refer to the metal/Al2O3 and water/Al2O3 molar ratios, respectively. Desired amount of aluminum isopropoxide was dissolved in deionized water under stirring for 3 h. To this solution, orthophosphoric acid was added dropwise, and the mixture was stirred for another 1 h. TEA, as the SDA, was then added dropwise, and the mixture was stirred for 24 h. For the preparation of the metal or ion substituted AlPOs, the metal precursor was added right after the addition of the SDA. The reaction gel having a pH ranging from 5 to 6 was transferred to a Teflon-lined stainless steel autoclave and placed inside a preheated oven at a temperature of 160°C for 24 h. After growth, the autoclave was quenched, and the solid product was collected after repeated centrifugation/washing cycles. The obtained crystals were dried at 80°C for 6 h and were subsequently calcined in air using a ramping rate of 2.5°C/min until temperature reached 80°C, keeping temperature stable for 30 min there, and further increasing it up to 600°C, where it was kept constant for 5.5 h. Activation at various temperatures under air flow (calcination) or nitrogen (pyrolysis) in a tubular furnace was also performed in order to parametrically explore decomposition/removal of the SDA occluded into the pores using the above temperature ramping program. A water/Al2O3 molar ratio of 100:1 was used in these experiments. The MeAPOs were synthesized using different metals/ions (Fe, Mg, Co, and Si) into the AlPO4-5 framework with water/Al2O3 molar ratio of 100:1 and metal/Al2O3 ratio of 5:100. For the FeAPO-5 adsorbents, additional metal contents were studied as well.</p><p>The following naming code of samples was applied throughout the various sets of experiments in this work:</p><p>For the activation set of experiments: TAlPO4-5.P and TAlPO4-5.C, where T is the thermal treatment temperature in °C, and P and C stand for pyrolysis under inert atmosphere and calcination under airflow, respectively.</p><p>For the metal substitution set of experiments: The molar ratios were placed ahead of the sample names. For example, 5FeAPO-5 indicates a metal/Al2O3 ratio of 5:100.</p><p>For the experiments on varying water content in the synthesis mixture: The H2O/Al2O3 molar ratio was placed ahead of the sample names. For example, 400AlPO4-5 indicates a H2O/Al2O3 ratio of 400:1.</p><!><p>The crystallinity of the synthesized samples was investigated by X-ray diffraction (XRD) using a Panalytical X'Pert Pro Powder Diffractometer. A Cu-Kα monochromatized radiation source with wavelength λ = 1.5406 Å, power 40 kV, and current of 40 mA was utilized, and the scan speed was set to 0.02 degrees/s. Morphology evaluation was performed by scanning electron microscopy (SEM) using an FEI Quanta 200 microscope. Samples were placed on a carbon tape and were coated by gold to enhance the conductivity and to allow observation at 30 kV. Ultraviolet-visible diffuse reflectance spectroscopy (UV-Vis DRS) of the adsorbents was recorded using a Varian Analytical Cary 5000 UV-Vis-NIR spectrometer, equipped with diffuse reflectance accessory used to record electronic spectra from 200 to 800 nm. Thermogravimetric analysis (TGA) was performed using TA instruments Trios V3.1 analyzer under 50 mL/min of air flow with a ramping rate of 10°C/min from 25 to 900°C. Fourier-transform infrared spectroscopy (FTIR) was carried out on a BRUKER TENSOR II series FTIR spectrometer with the aid of diamond attenuated total reflectance crystal, where thin sections of the samples were scanned in the wavenumber region of 4,000 to 400 cm−1 with a resolution of 4 cm−1 by undergoing 32 scans. Nitrogen adsorption–desorption analysis was performed at 77 K using a Micromeritics 3-Flex analyzer. Before the analysis, the samples were degassed at 220°C for 3 h. The surface area was determined by the Brunauer–Emmett–Teller (BET) method. Pore size distribution and pore volume were obtained by processing adsorption–desorption data using the Barrett, Joyner, Halenda (BJH) and Horvath–Kawazoe (HK) models via the MicroActive software.</p><!><p>CO2 adsorption experiments were carried out using a Rubotherm gravimetric sorption analyzer (IsoSORP STATIC 3xV-MP). The magnetic suspension balance measures sample weight and gas dosing to determine the adsorption equilibrium. Approximately 50 mg of each activated adsorbent was first heated under vacuum at 150°C for 3 h to remove moisture and other volatile substances until a constant sample mass was obtained. Buoyancy correction was performed using helium gas and incorporated into the sample weight. The sample in a stainless-steel holder was pressurized with CO2 gas up to 4 bar (in steps of 1 bar), and continuous measurement of sample mass gain during CO2 adsorption was determined at 25°C under equilibrium/saturation conditions. The gravimetric analyzer software uses the recorded mass changes to estimate the sorption capacity at each pressure step and incorporates the buoyancy correction as well in order to generate the final CO2 sorption isotherm. For the adsorption kinetics study, the adsorbents were measured again but with a single step directly up to 4 bar taking capacity values at regular time intervals. The procedure was repeated for three different temperatures, namely, 25, 45, and 60°C.</p><!><p>Upon growth of porous crystalline materials such as zeolites, SDAs (or organic templates) are typically used to direct the formation of the pores. Following growth, removal of these molecules from the pores needs to take place as to activate the open porosity of the materials. Whether the occluded molecules are completely decomposed and removed or remnants of carbon still exist into the pores may affect the affinity and interaction of the internal surface with CO2. This was first investigated in our study by activating the resulting AlPO4-5 crystals via calcination and partial calcination (air flow) or pyrolysis (nitrogen atmosphere) at various temperatures. The latter was employed based on our previous work revealing that graphitic carbon in the form of carbon nanotubes (CNTs) into the pores of oriented AlPO4-5 films affected CO2 sorption, as well as permeance behavior through the resulting CNT membranes (Labropoulos et al., 2015). In addition, the occluded amine SDA is partially decomposed and removed upon partial calcination, which might create extra spacing/porosity within the larger AlPO4-5 channels. The objective was to examine whether remaining carbon species would act favorably toward CO2 adsorption via interaction of CO2 with both AlPO4-5 and carbon surfaces, or complete removal of SDA would be preferred due to maximizing the pore volume and carbon-free AlPO4-5 surface.</p><p>Consequently, for calcination, five different temperatures were used, namely, 300, 400, 500, 600, and 700°C under airflow. The distinct color of the materials produced at the different calcination temperatures is shown in Supplementary Figure 1. The as-synthesized AlPO4-5 crystals had a white color. The material calcined between 300 and 500°C had a distinct brown coloration that turned softer as the temperature increased, which indicates the existence of remaining carbon species at the external crystal surface at these relatively low calcination temperatures with carbon content being decreased with increase in temperature. At temperatures above 600°C the materials turned white again, indicating complete carbon removal.</p><p>Pyrolysis of the synthesized AlPO4-5 samples was carried out under nitrogen flow at temperatures of 240, 400, and 700°C. The aim was to investigate possible affinity enhancement effects between CO2 and carbon species resulting from the TEA pyrolysis in the pores or formation of additional carbon porosity/surface within the AFI channels. The creation of porous carbon-based structure and possibly extra porosity inside the channels of AlPO4-5 might enhance affinity toward CO2 due to the existence of pyrolytic carbon. Analogous treatment yielded formation of single-wall CNTs inside the AFI pores of powder AlPO4-5 crystals (Tang et al., 1998), as well as in oriented membrane configuration for membrane-based gas separation, as demonstrated in our previous work (Labropoulos et al., 2015). The pyrolysis temperature of 240°C was selected as it represents a point within the TEA decomposition region based on TGA data (Wan et al., 2000).</p><p>XRD analysis of selected calcined and pyrolyzed samples are shown in Figure 1A. The water/Al2O3 molar ratio was fixed at 100 for this activation effect study. All samples treated at temperatures up to 700°C exhibit the characteristic diffraction peaks of AFI (Karanikolos et al., 2008; Basina et al., 2018), confirming the stability and structural integrity of the materials up to this temperature. Notably, the sample pyrolyzed at the lowest temperature (240°C), in addition to the AFI peaks, exhibits an additional peak shoulder at 22.6° and a minor peak at 12.1°, which are attributed to remnant carbon clusters (Sawant et al., 2017). At tested temperatures of 400°C and above, the above peaks disappear indicating that the remnant carbon content is significantly reduced or eliminated. Furthermore, such peaks do not exist in the untreated AlPO4-5 material either (Supplementary Figure 2). Thermal stability of a potential adsorbent is one of the key quality features for industrial application ensuring robustness upon thermal stresses that may occur during preparation as well as application. e.g., upon regeneration or other processing steps. The high temperature treatment applied here and the obtained stability of the AlPO4-5 adsorbents confirm their robustness and suitability for high temperature application. In order to explore the upper temperature stability limit, we also calcined the material at 850°C. There, we observed a structure collapse and transformation into a dense AlPO4-tridymite phase (Stoeger et al., 2012).</p><!><p>(A) XRD patterns and (B) TGA profiles of calcined and pyrolyzed AlPO4-5 adsorbents treated at different temperatures.</p><!><p>The carbon content of the thermally treated samples and their corresponding thermal behavior in comparison to untreated AlPO4-5 was further studied by TGA, which was performed under oxidative conditions (air). According to the obtained results (Figure 1B), the pyrolyzed and calcined adsorbents exhibit almost the same TGA profiles as revealed for the corresponding samples treated at 400 and 700°C Notably, the only significant weight loss experienced by both of the above sets of samples is the one at the low temperature range of up to 100°C, which is attributed to removal of adsorbed moisture. The absence of any noticeable transition due to amine decomposition indicates that the occluded amine molecules have been almost completely removed from the pores at these temperatures. In addition, the samples treated at 700°C exhibit a more hydrophobic behavior containing 1.3 wt% of moisture compared to the ones treated at 400°C that contain approximately 4.2 wt% of moisture (Table 1). This is attributed to the fact that polar functional groups on the AlPO4-5 surface, such as Al-OH and P-OH (Peri, 1971), are being compromised by the higher temperature treatment, thus suppressing hydrophilicity. The untreated AlPO4-5 adsorbent and the one pyrolyzed at 240°C exhibit two distinct weight loss transitions after the removal of the adsorbed water. The first transition is attributed to decomposition of the TEA molecules from the pores and occurs in the temperature range of 120–300°C for the untreated sample, while it starts at a higher temperature (165°C) for the pyrolyzed one. This temperature difference is due to the fact that portion of the amine has already been decomposed in the latter sample. The second transition occurs at a considerably higher temperature (500–650°C) and is attributed to the removal of trapped and possibly graphitized carbon species from the pores. As per Table 1, the total organic content in the untreated sample is 7.8 wt%, whereas that of the sample pyrolyzed at 240°C is 5.66 wt%. The existence of remnant carbon in the latter sample is in agreement to XRD evidence discussed above.</p><!><p>Characteristic TGA transitions and corresponding weight loss percentages for the various calcined and pyrolyzed AlPO4-5 adsorbents.</p><!><p>The effect of the calcination and pyrolysis treatment on the CO2 adsorption capacity is depicted in Figure 2 (data in Supplementary Table 1). The sample calcined at 700°C exhibits higher CO2 capacity (1.53 mmol/g at 4 bar) than the one calcined at 400°C. This observation is consistent throughout the pressure range examined (up to 4 bar) and indicates that maximizing porosity in the channels of the adsorbent is critical. The sample pyrolyzed at 700°C exhibits higher CO2 capacity compared to the other pyrolyzed samples throughout the pressure range of 0 to 4 bar, with a maximum capacity of 1.48 mmol/g at 4 bar. This is attributed to the total decomposition and removal of TEA from the framework, and the creation of clean pores in the zeolite to host the CO2 molecules. The remaining two isotherms corresponding to the lower pyrolysis temperature samples (240AlPO4-5.P and 400AlPO4-5.P) are of interest since an inversion in the CO2 adsorption capacity can be observed. Specifically, the sample which was treated at 240°C exhibits higher capacity at low pressures (up to 2 bar), whereas at higher pressures (2–4 bar) the sample treated at 400°C adsorbs more CO2. In addition, a noticeable observation with respect to low pressure CO2 capture application is that, up to 1 bar, the sample pyrolyzed at 240°C exhibits almost same capacity as the one pyrolyzed at the highest temperature tested, i.e., 700°C, which is also very close to the capacity observed for the calcined sample at 700°C. This behavior indicates that the carbon species remnants into the pores created by low temperature SDA pyrolysis interact efficiently with CO2 at low pressures, whereas capture capacity at higher pressures is more favored by the increased pore volume and carbon-free surface created upon higher temperature thermal treatment. Conclusively, if the AlPO4-5 adsorbents are to be used for low-pressure CO2 capture, high thermal activation is not required, as similar capture capacity can be achieved by pyrolysis treatment at significantly lower temperatures, e.g., 240°C, thus saving energy and safeguarding the thermal stability of the adsorbents. The relatively high capacity resulting from such pyrolysis-based activation treatment is attributed to enhanced interaction of CO2 at low pressures with carbon species that are remnant in the pores upon partial SDA decomposition.</p><!><p>CO2 adsorption isotherms of calcined and pyrolyzed AlPO4-5 adsorbents at 25°C.</p><!><p>Ion-substituted AlPO4-5 were prepared by incorporating Fe, Mg, Co, and Si into the framework of AlPO4-5. Ion substitution in AlPO4-5 takes place through various mechanisms depending on the substituting element, with divalent and trivalent ions substituting Al, whereas tetravalent ions such as silicon substitutes predominantly P at low Si content, whereas at higher Si ratios a pair of Si ions substitutes adjacent Al and P ions (Gaber et al., 2020). In the present set of experiments, a water/Al2O3 molar ratio of 100:1 and a metal/Al2O3 ratio of 5:100 were used. The XRD patterns of the resulting MeAPOs (Figure 3A) are in accordance with the AFI structure possessing all the AFI peaks, including the characteristic ones at 2θ of approximately 20, 22, and 23° that correspond to the reflections from the (210) (002) and (211) crystallographic planes, respectively (Karanikolos et al., 2008; Stoeger et al., 2012; Basina et al., 2018). Furthermore, the materials exhibit high crystallinity as no broad peaks or shoulders and no evidence of secondary crystalline phases/impurities were noticed. A closer look at the XRD patterns reveals that some peaks of the ion substituted AlPOs are not perfectly aligned with those of the AlPO4-5. Indeed, Figure 3B shows the three characteristic AFI peaks extending between 19 and 23°, which are attributed to the crystallographic (210), (002) and (211) planes where minor shifts are evident. These shifts are attributed to the ion substitution and incorporation into the framework since heteroatoms with different ionic radii are inserted into the lattice substituting Al and/or P. The ionic radius of alumina is 0.53 Å with all substituting metals having larger ionic radii, i.e., Fe: 0.645 Å, Mg: 0.72 Å, and Co: 0.745 Å. Silica possesses an ionic radius is 0.4 Å and has the ability to substitute phosphorus, which has an ionic radius 0.38 Å (SM2 mechanism). Notably, Fe exhibits the biggest expansion along the c-dimension. These shifts confirm the successful substitution of the metals into the AlPO framework. Analogous results were reported in our previous work, where AlPO4-5 metal substitution resulted in change of lattice parameters (Gaber et al., 2020).</p><!><p>(A) XRD patterns of MeAPOs with Me/Al2O3 molar ratio in the synthesis mixture of 5:100, and (B) closer view of the AFI characteristic peaks between 19 and 23°.</p><!><p>SEM images of calcined AlPO4-5 and MeAPO-5 are depicted in Figure 4. The materials are comprised of crystalline particles, thus confirming the XRD findings. All images correspond to adsorbents generated from dense reaction mixtures (H2O/Al2O3 molar ratio of 100), except Figure 4c that corresponds to dilute reaction mixture (H2O/Al2O3 molar ratio of 400) and the crystals exhibit columnar morphology. According to the low magnification images, the particles of the materials with H2O/Al2O3 molar ratio of 100 have spherical shape with uniform sizes that range between 20 and 30 μm. In high magnification, it is evident that the spherical particles are agglomerates of flat rectangular-like small crystals. In FeAPO-5 (Figures 4d,f), an analogous morphology is observed, whereas it is evident that the crystal aggregates tend to acquire a more hexagonal shape (Figure 4f) following the AFI crystal structure. This reveals that the particles in this adsorbent have been formed not by simple physical aggregation of the small individual crystallites, but rather a considerable degree of interaction and coalescence have taken place upon crystal growth.</p><!><p>SEM images of (a,b) 100AlPO4-5, (c) 400AlPO4-5, (d–f) FeAPO-5, (g,h) MgAPO-5, (i,j) CoAPO-5, and (k,l) SAPO-5 adsorbents.</p><!><p>The MgAPO-5 crystal aggregates (Figures 4g,h) exhibit spherical shape with an average particle diameter of 20 μm. The particle size of MgAPO-5 is smaller than that of FeAPO-5, whereas the aggregates seem to have a more perfect spherical shape compared to the hexagonal configuration observed in FeAPO-5. This reveals that interaction and coalescence between the individual small crystallites in MgAPO-5 is lower compared to that in FeAPO-5. CoAPO-5 (Figures 4i,j) also mainly consists of spherical particles, but they are quite heterogeneous since some of them have holes at their axis. The majority of SAPO-5 crystals are spherical, but some larger, irregularly shaped aggregates have also formed by further merging of the spherical particles, as shown in Figures 4k,l.</p><p>The morphology of AlPO4-5 prepared at high dilution rate (H2O/Al2O3 molar ratio in the synthesis mixture of 400, Figure 4c) reveals mainly columnar, hexagonal monocrystals (Cheung et al., 2012; Gaber et al., 2020). This is attributed to the fact that dilution of the reaction gel decreases the nucleation rate (Du et al., 1997), and favors preferential growth along the c-axis of the crystals. Indeed, the spheres formed in the samples using dense synthesis mixtures (H2O/Al2O3 molar ratio of 100) are comprised of smaller rectangular-like crystals that are tightly interconnected. The progress of crystallization controls the size of the spherical aggregates and prove to be rather homogeneous. FeAPO-5 particles tend to acquire a hexagonal shape in accordance to the AFI morphology compared to the rest of materials grown using dense reaction mixtures, thus indicating a closer interaction and coalescence among the individual flat-like crystallites that comprise each aggregate. This can be correlated to the XRD observations (Figure 3B), where FeAPO-5 was shown to exhibit the largest shift in the (002) peak, which may be also partially associated to the interface Fe ions shared among coalesced crystallites. Increasing the dilution slows down the crystallization rate, thus forming hexagonal, rod-like shape crystals following preferential growth along the c-axis (Iwasaki et al., 2003; Karanikolos et al., 2008). However, due to slow crystallization, it is possible that these samples may also possess some amorphous areas (Utchariyajit and Wongkasemjit, 2008) and low crystallinity regions.</p><p>The CO2 adsorption results of the metal substituted AlPOs and the parent AlPO4-5 are shown in Figure 5 (data tabulated in Supplementary Table 2). Overall, MeAPO-5 samples exhibit higher capacity than the parent AlPO4-5 indicating that metal substitution promotes CO2 adsorption. Among all MeAPO-5 samples, FeAPO-5 exhibits the highest capacity throughout the pressure range tested, reaching a maximum value of 1.8 mmol/g at 4 bar, which is 15% higher than that of the parent AlPO4-5 at the same conditions, owing to the substitution of Al3+ with Fe3+ in the lattice. At the lower pressure range of up to 3 bar, the Fe- and Co-substituted analogs exhibit higher capacity than the Mg-substituted one due to the fact that the presence of transition metal elements (Fe and Co) in the lattice tend to enhance the affinity between CO2 with the inorganic framework compared to alkali earth metals (Mg) and weak metals (Al) (Yu et al., 2018).</p><!><p>CO2 adsorption of various metal substituted AlPOs at 25°C.</p><!><p>The CO2 capacity of MgAPO-5 is the lowest at low CO2 partial pressures even when compared to the parent AlPO4-5 adsorbent, whereas at pressures higher than 2.5 bar it displays an increase reaching at 4 bar at an almost same value as that of FeAPO-5. Strong ionic character and bond length in this case (Mg-O 1.969 Å) could influence CO2 sorption (Caskey et al., 2008; Yazaydin et al., 2009), with a possible clustering effect of CO2 molecules in the vicinity of Mg within the pores taking place at higher pressures. In addition, diffusion limitations within the pores as well as in the interstitial spaces among the crystallites in each particle agglomerate could be overcome as pressure increases. CoAPO-5 and SAPO-5 contain acidic sites within their framework due to charge imbalance upon ion substitution. Their sorption capacity is slightly higher compared to the parent AlPO4-5 adsorbent (by 6% and 4%, respectively). Analogous results were reported using in situ IR spectroscopy studies on SAPO-56 adsorbents possessing a high concentration of acid sites (Cheung et al., 2012). Effect of ion substitution on the lattice parameters may play also a role in CO2 adsorption. The substituting metals have two choices in AlPO4-5, i.e., to substitute Al and P positions. In general, Si will preferentially substitute for P, and divalent and trivalent ions will replace Al (Wilson et al., 1982). Substitution by transition metals with higher ionic radius (Fe, Mg, and Co) will generally increase the value of lattice parameters than low ionic radius elements such as Si (Kaneko and Rodríguez-Reinoso, 2019), thus enhancing the CO2 adsorption capacity.</p><!><p>The enhanced CO2 adsorption capacity of FeAPO-5 compared to all other AlPO samples tested led us to further investigate this material by varying the molar composition of iron in the precursor mixture. An initial molar ratio of Fe/Al2O3 of 5:100 was used as a basis, and a set of materials with three more Fe/Al2O3 molar ratios, namely, 2.5:100, 7.5:100, and 10:100, were additionally synthesized and studied.</p><p>The liquid N2 adsorption–desorption isotherms of the FeAPO-5 adsorbents are shown in Figure 6A, while the resulting physical and pore properties estimated from BET analysis, BJH, and HK methods are presented in Table 2. All adsorbents exhibit a steep nitrogen uptake at a low relative pressure (P/Po) confirming their microporous nature. The adsorbents display a type-IV isotherm with hysteresis loops extending across relative pressures of 0.45 to 0.9, which confirm also the existence of mesopores. Both micropore and mesopore size distributions are shown in insets i and ii, respectively. Among the tested samples, the steepest increase at high relative pressures (P/Po > 0.8) and the most extensive hysteresis loop was observed for 5FeAPO-5, which are indicative of an extended mesoporous network (Figure 6B). According to the micropore size distribution, all-metal substituted samples possess a narrow peak centered at ~0.7 nm, which confirms their AFI structure (Karanikolos et al., 2008). From Table 2, the BET surface area of FeAPO-5 adsorbents varies from 120 to 264 m2/g, the pore volume from 0.06 to 0.18 cc/g, and the average pore diameter from 1.8 to 3.5 nm. The relatively wide ranges observed reveal the considerable affect that the metal incorporation induces to the physical properties of the adsorbents.</p><!><p>(A) Liquid N2 adsorption–desorption isotherms of FeAPOs. Inset (i): Micropore size distribution determined by the HK method. Inset (ii) BJH-derived mesopore size distribution. (B) BET surface area and mesopore volume as a function of Fe/Al2O3 molar content (with respect to %Al2O3) in the synthesis mixture.</p><p>Physical properties of FeAPO-5.</p><!><p>To shed more light on the structural effects of the metal content, we also performed XRD, FTIR, and UV-Vis DRS analysis of the FeAPO-5 adsorbents. According to the XRD patterns (Figure 7A), the structural integrity of the AFI framework is retained for the whole range of Fe/Al2O3 % molar ratios tested (2.5–10), with no evidence of metal incorporation effects in the AFI crystallinity or appearance of secondary phases/impurities. FTIR spectra are shown in Figure 7B. The band at 3,535 cm−1 is ascribed to –OH functional groups, whereas the bands at 1,219, 716, and 615 cm−1 are ascribed to the asymmetric and symmetric stretching vibrations of the Al-O-P units (Zhang et al., 2020). Comparing the FeAPO-5 samples with the non-substituted AlPO4-5, the band at 562 cm−1 due to Al-O or P-O bending modes of AlPO4-5 framework is significantly suppressed upon Fe incorporation, whereas the band at 1,112 cm−1, which corresponds to stretching vibration of Al–O in combination with P–O (Chen and Jehng, 2003), is shifted to lower wavenumber (cm−1) values. Comparison among the various FeAPO-5 samples does not reveal any noticeable differences in FTIR bands as to differentiate among the Fe loadings and/or possible Fe segregation. UV-Vis DRS spectra of the Fe-substituted adsorbents are shown in Figure 7C. A dominant peak centered at 263 nm is evident for all FeAPO-5 samples with a shoulder at 235 nm attributed to the ligand to metal charge transfer of Fe3+ in [FeO4]− tetrahedral geometry (Mohapatra et al., 2002). The intensity of the above band for the unsubstituted AlPO4-5 is negligible compared to the iron containing samples (Feng et al., 2016). Notably, for the sample with the highest metal content (10FeAPO-5) the above peak is shifted to higher wavelengths, which is indicative of increased amount of octahedral complexes in extra-framework positions, and the distribution state for the Fe species corresponding to 280 nm in particular can be isolated or clustered (Feng et al., 2016). The broad band between 400 and 500 nm is due to Fe d–d transitions and is also indicative of clustering of iron species (Wei et al., 2008). Notably, the intensity of the above band for the 5FeAPO-5 sample is minimal.</p><!><p>(A) XRD, (B) FTIR, and (C) UV-Vis DRS analysis of FeAPO-5 adsorbents.</p><!><p>The FeAPO-5 adsorbents exhibit both micropores with an average diameter of ~0.7 nm, which is bigger than the kinetic diameter of CO2 (0.33 nm), thus posing no kinetic restriction for CO2 adsorption, as well as a relatively extended mesoporous network (Figure 6A). In such configuration, at low partial pressure of CO2, monolayer adsorption and adsorption on the micropores are anticipated to occur first, whereas mesopores are filled as pressure increases and adsorbate–adsorbate interactions are being enhanced. The CO2 adsorption capacity of the FeAPO-5 adsorbents with different metal concentrations are shown in Figure 8 (data also in Supplementary Table 3). According to Figure 6B, the BET surface area decreases with Fe content until a Fe/Al molar ratio of 7.5:100 and then it increases again. The mesopore volume follows an opposite trend, i.e., it increases with Fe content, exhibits a maximum at Fe/Al ratio of 5:100, and then it decreases. As shown by UV-Vis DRS analysis above, for high metal concentration in the synthesis mixture, a portion of the available Fe forms clusters, and it is not incorporated into the AFI framework as tetrahedrally coordinated ions by substituting Al3+. The existence of amorphous Fe3+ phase for high metal loading has been confirmed also before via various spectroscopic and chemical probe methods (Das et al., 1992). Consequently, the CO2 adsorption capacity does not exhibit a regular trend as Fe content increases, rather it exhibits a maximum for the sample corresponding to Fe/Al2O3 ratio of 5:100, which exhibits the higher volume of mesopores and the minimum portion of extra-framework or clustered iron as revealed by the UV-Vis analysis. Indeed, the adsorption capacity of 5FeAPO-5 at low pressure (1 bar) is 0.64 mmol/g and at high pressure (4 bar) it becomes 1.8 mmol/g, corresponding to an incremental increase of 1.16 mmol/g from the above low to high pressure values, which is the highest increment among the FeAPO-5 adsorbents tested. This is attributed to the contribution of mesopores in the adsorption capacity which becomes dominant at higher pressures. Conclusively, Fe incorporation enhances the mesoporosity despite the fact that it decreases the surface area. The latter is more important for relatively low metal content causing a reduction in CO2 capacity compared to pure AlPO4-5, yet as metal content increases, the mesoporosity formation becomes dominant and brings the capacity to higher values than that of the pure AlPO4-5 at same conditions. However, for high metal loadings, some Fe prefers to cluster upon growth forming amorphous Fe3+ phases and/or extra-framework species thus associated to reduced mesoporosity and lower CO2 capacity.</p><!><p>CO2 adsorption of FeAPO-5 adsorbents with different metal content at 25°C.</p><!><p>In our previous work we have shown that a critical parameter that strongly affects crystal morphology upon AlPO4-5 growth is the water content in the synthesis mixture (Karanikolos et al., 2008). Specifically, it was demonstrated that dense reaction mixtures favor growth perpendicular to the AFI channels, i.e., along the a-b directions, while diluted reaction mixtures induce preferential growth along the c-direction, i.e., parallel to the channels. Accordingly, we employed here two different dilutions in the reaction mixture, namely, H2O/Al2O3 molar ratio of 100 (sample 100AlPO4-5) and 400 (sample 400AlPO4-5), which resulted in spherical agglomerates of small flake-like crystallites, and columnar monocrystals, respectively (Figures 4A–C). From XRD analysis (Figure 9A), it is evident that both samples possess the AFI structure, yet the lower-dilution sample displays a higher crystallinity when compared to the high-dilution one.</p><!><p>(A) XRD patterns for AlPO4-5 crystals corresponding to two different dilution rates in the reaction mixture, i.e., H2O/Al2O3 molar ratio of 100 and 400 (100AlPO4-5 and 400AlPO4-5). (B) N2 adsorption isotherms of 100AlPO4-5 and 400AlPO4-5 at 77K, inset (i): micropore size distribution determined by the HK method, inset (ii): BJH mesopore size distribution.</p><!><p>The liquid N2 adsorption isotherms and pore size distributions of the 100AlPO4-5 and 400AlPO4-5 adsorbents are shown in Figure 9B, and the associated physical and porosity properties are presented in Table 3. It is evident that diluted reaction mixtures yield AlPO4-5 with lower surface area and micropore volume, yet enhanced mesoporosity. The shape of the isotherms also confirms this observation, as the isotherm of 100AlPO4-5 approaches that of type-I revealing a strongly microporous nature with negligible contribution from larger pores, whereas that corresponding to 400AlPO4-5 is a type-IV isotherm with a steep rise in adsorption at P/P0 > 0.85 and an enhanced hysteresis loop revealing capillary condensation in mesopores. Indeed, the surface area drops from 294 to 125 m2/g while the mesopore volume increases from 0.0133 to 0.4153 cc/g for the 100AlPO4-5 and 400AlPO4-5 adsorbents, respectively. The micropore size distribution (Figure 9B, inset ii) is narrow averaging at ~ 0.7 nm for both materials, in accordance to AFI structure, whereas the portion of larger pores is relatively low for 100AlPO4-5, and becomes significant for 400AlPO4-5, for which an average mesopore width of 25 nm is evidenced from Figure 9B, inset.</p><!><p>Physical properties of 100AlPO4-5 and 400AlPO4-5.</p><!><p>The CO2 adsorption isotherms and associated capacities for the AlPO4-5 adsorbents corresponding to low and high dilutions of the reaction mixture are shown in Figure 10. It is evident that dense reaction mixtures (100AlPO4-5) yield adsorbents that exhibit higher capacity compared to diluted ones (400AlPO4-5). Indeed, 100AlPO4-5 exhibits an adsorption capacity of 1.57 mmol CO2/g at 4 bar and 25°C, which is almost double than that of 400AlPO4-5. As discussed earlier for the Fe-substituted adsorbents, a high mesopore volume is favorable for enhancing CO2 capacity. Nevertheless, this does not seem to be the case for the non-substituted materials where substituting heteroatoms do not exist. Here, the effect of BET surface area reduction for the adsorbent corresponding to high dilution rate of the reaction mixture is more dominant than the mesoporosity formation, thus causing a decrease in CO2 capacity compared to the adsorbent originated from dense mixture. In addition, a high average mesopore diameter (25 nm in the case of 400AlPO4-5) might decrease the interaction potential, affinity between CO2-CO2, and retention upon multilayer adsorption.</p><!><p>CO2 adsorption of AlPO4-5 adsorbents corresponding to two different reaction mixture dilution rates and associated crystal morphologies at 25°C.</p><!><p>The isosteric heat of adsorption concerns the amount of energy that is released when CO2 adsorbs onto the adsorbent surface at a fixed coverage. Consequently, at least the same amount of energy must be added in order for the CO2 to be desorbed from the adsorbent upon regeneration. In addition, the isosteric heat of adsorption values provide a quantifiable indicator of the affinity of an adsorbate molecule toward the adsorbent surface. CO2 adsorption isotherms were collected at three different temperatures in order to calculate the heat of adsorption of the AlPO adsorbents at different CO2 loadings. The isosteric heat of adsorption values were determined based on the Clausius–Clapeyron equation using the obtained isotherm data, and the results are depicted in Figure 11.</p><!><p>Isosteric heat of adsorption of different AlPO4-5 and FeAPO-5 adsorbents.</p><!><p>The heat of adsorption values for all the studied adsorbents are lower than 25 kJ/mol throughout the tested coverages, which indicates that the mechanism of CO2 adsorption in these materials is physisorption, in line to the reported literature (Simmons et al., 2011). As such, they hold a high potential for PSA-based capture application as desorption can easily be enabled by pressure swing without the need of thermal regeneration. According to the obtained profiles, the heat of adsorption increases as the CO2 loading increases for all the studied adsorbents, implying that, as coverage increases, adsorbate–adsorbate (CO2-CO2) interactions are stronger than adsorbate–adsorbent interactions on CO2-philic surface sites. Concerning the Fe-substituted analogs in particular, the ones corresponding to relatively low Fe/Al2O3 molar ratio in the synthesis mixture (2.5:100 and 5:100) exhibit the highest heat of adsorption values compared to the rest of tested adsorbents. This indicates that the incorporated metal enhances the physical binding, a fact that is more pronounced for 5FeAPO-5 at low coverage indicating a stronger CO2 interaction on the adsorbent surface upon monolayer formation. Notably, the above mentioned adsorbent exhibited the highest uptake compared to the rest of the ion-substituted materials tested as well as to the pure AlPO4-5 (Figures 5, 8). Furthermore, the heat of adsorption curve of 5FeAPO-5 is almost flat, which is attributed to high crystallinity and homogenous surface with a uniform distribution of the Fe heteroatoms into the framework lattice that enhance binding. Notably, this adsorbent contains the lowest portion of extra-framework or clustered iron among all FeAPO-5 materials tested (Figure 7C). As Fe content in the synthesis mixture increases the heat of adsorption, and thus the CO2 binding strength with the adsorbent surface decreases, a result that is in agreement to the observed reduced CO2 adsorption capacity for 7.5FeAPO-5 and 10FeAPO-5, which is attributed to the tendency of the readily available Fe ions at these high metal concentrations to cluster in the synthesis mixture and yield formation of amorphous Fe3+ phases upon crystal growth (Das et al., 1992), as also discussed above (Figure 7C), and/or remain as extra-framework species partially blocking active surface sites.</p><p>The low value of heat of adsorption for the 400AlPO4-5 adsorbent, which corresponds to diluted reaction mixture, indicates that the morphology and textural and surface properties of this material, including the large pore width, large mesopore volume, and rod-like crystal morphology, promote a very weak binding with the surface, which is in agreement to the low CO2 adsorption capacity obtained for this material. In addition, the heat of adsorption curve as a function of coverage for this adsorbent exhibits the largest slope among all tested adsorbents for the low coverage range (up to 0.22 mm/g). This is indicative of the initially weak interaction of CO2 with the surface and during monolayer adsorption, which becomes stronger as coverage increases due to more dominant adsorbate–adsorbate interactions. At higher coverages (>0.22 mmol/g), the slope becomes lower and almost equalizes with that of the other tested adsorbents.</p><!><p>Understanding of the adsorption kinetics of an adsorbent is vital in designing PSA systems (Loganathan et al., 2014). Various adsorption kinetic models have been employed, among which, the pseudo–first-order kinetics or Lagergren model (Liu et al., 2014) is widely used for CO2 adsorption. The pseudo–first-order equation models a reversible interaction between adsorbent and adsorbate. This is suitable for describing the physical adsorption of CO2 on solid adsorbents (Serna-Guerrero and Sayari, 2010; Loganathan et al., 2014). The pseudo–first-order model is governed by Equation (1):</p><p>where qe and qt (mg/g) represent the amount of CO2 adsorbed at equilibrium and at a given time "t", respectively, and k (min−1) is the first order rate constant. The relevant boundary conditions are BC1: t = 0, qt = 0 and BC2: t = ∞, qt = qe. Applying the above boundary conditions to Equation (1) while solving the aforementioned differential equation leads to Equation (2).</p><p>The higher CO2 adsorption capacity and rate of the 100AlPO4-5 adsorbent compared to the 400AlPO4-5 analog, as confirmed by both equilibrium data at various pressures (Figure 10), and kinetic data up to 4 bar (Supplementary Figure 3) led us to further investigate the kinetic behavior of this adsorbent by conducting adsorption experiments at three different temperatures, i.e., 25, 45, and 60°C up to a pressure of 4 bar (Supplementary Table 5 and Supplementary Figure 4). The CO2 adsorption capacity decreases as temperature increases confirming the physisorption mechanism. In order to calculate the adsorption rate constant k for the three different temperatures, the adsorption data at the different temperatures were used to plot log (qe – qt) vs. t/2.303 (Supplementary Figure 5). The resulting linear fit slope is the adsorption rate constant (k), and the y-intercept is log qe (Liu and Shen, 2008; Liu et al., 2014). The adsorption rate constant increases as temperature increases (Table 4) yet with relatively slight differences indicating that the adsorption of CO2 on AlPO4-5 is rather thermodynamically limited. The equilibrium adsorption capacity (qe) was also calculated from the pseudo–first-order model and compared with the experimentally measured one (Supplementary Table 5). The model predicted the equilibrium adsorption capacity with an average relative error of 1.8–6.5%, and the value of the correlation coefficient (R2) of the model is around 0.98. Thus, the pseudo–first-order model proves to be qualitatively and quantitatively suitable for modeling of CO2 adsorption kinetics of the studied AlPO4-5 adsorbents as shown Figure 12A.</p><!><p>Adsorption rate constant (k) at different temperatures and up to 4 bar for 100AlPO4-5.</p><p>(A) Comparison of experimental data vs. model fitting data. (B) Arrhenius plot for the kinetic rate constant obtained by the linear Lagergren model for the 100AlPO4-5 adsorbent.</p><!><p>The temperature dependence of rate constant (k) can be described by the Arrhenius equation (Equation 3), where A is the pre-exponential factor, E is the activation energy, R is the universal gas constant, and T is the temperature in absolute units.</p><p>The plot of ln(k) vs. 1/T (Figure 12B) exhibits a linear profile, as anticipated from the linearized form of the Arrhenius equation. From the obtained slope, the activation energy for the CO2 adsorption on 100AlPO4-5 at 4 bar was calculated to be 4.7 kJ/mol. The obtained activation energy value is in close agreement to the reported values of CO2 adsorption on activated carbon and Zeolite 13X, where activation energies of 3.9 and 4.8 kJ/mol, respectively, were estimated up to CO2 pressure of 3 bar (Zhang et al., 2010).</p><!><p>Effects of metal substitution, synthesis mixture composition and associated morphology manipulation, and activation procedure on AlPO4-5 were studied for CO2 adsorption. Activation by calcination in air at high temperature was found to completely open up the pores resulting in higher CO2 capacity compared to calcination at lower temperatures. Yet, upon activation by pyrolysis in inert atmosphere, low temperature (240°C) treatment enhanced CO2 interaction with the surface at low pressures (up to 1 bar) due to the existence of remnant carbon species in the pores from the partial decomposition of the SDA. Ion substitution by Fe, Mg, Co, and Si induced changes in lattice parameters and morphology, whereas the Fe-substituted adsorbents exhibited the highest capacity compared to the rest of metal-substituted analogs. Parametric variation of the Fe content in the synthesis mixture revealed that, at relatively low metal concentrations, mesopore volume increases and microporosity and BET surface area decrease with metal content. At these conditions, the strong interaction of CO2 with the framework Fe and the adsorbate–adsorbate interactions in the formed mesopores increased the CO2 capacity. For high metal concentrations, some Fe prefers to cluster upon growth forming amorphous Fe3+ phases and/or extra-framework species, thus resulting in reduced mesoporosity and lower CO2 capacity. Dense reaction mixtures yielded AlPO4-5 consisting of spherical agglomerates of 2D-like crystallites that were almost exclusively microporous in nature, whereas diluted reaction mixtures resulted in crystals of columnar morphology exhibiting significant mesoporosity. In contrast to the metal-substituted analogs and thus in the absence of metal-CO2 interactions, the microporosity and high surface area of the AlPO4-5 adsorbents corresponding to dense reaction mixtures were dominant factors compared to the enhanced mesoporosity of the ones grown from diluted mixtures, thus yielding higher CO2 capacity for the former materials.</p><p>The isosteric heat of adsorption values for all the studied adsorbents were lower than 25 kJ/mol indicating that the mechanism of CO2 adsorption is physisorption, which is suitable for PSA application. The Fe-substituted analog corresponding to a Fe/Al2O3 molar ratio in the reaction mixture of 5:100 exhibited the highest heat of adsorption at low coverage, indicating affinity and stronger physical binding of CO2 with the framework-incorporated Fe. Notably, this adsorbent exhibited the lowest portion of extra-framework or clustered iron. Kinetic analysis revealed that a pseudo–first-order model could describe well the CO2 adsorption kinetics of the AlPO4-5 adsorbents, with an activation energy of 4.7 kJ/mol at 4 bar and adsorption rate constants increasing with temperature, yet with rather slight differences. AlPO4-5–based adsorbents, though they exhibit rather moderate capacities, they are robust, thermally and chemically stable materials, they exhibit low heat of adsorption, thus potential for PSA application, and limited hydrophilicity compared to classical zeolites that makes them important for CO2 capture from wet streams. The parametric investigation performed in this work sheds light on main optimization factors and paves the way for further studies toward implementation at industrial scale.</p><!><p>All datasets generated for this study are included in the article/Supplementary Material.</p><!><p>AP performed the synthesis/growth, characterization of the adsorbents, and drafted part of the manuscript. KR performed the CO2 evaluation and drafted part of the manuscript. DK edited the manuscript and contributed in the supervision of AP. DR edited the manuscript and contributed in the supervision of the activities. YA contributed in the supervision of the activities, provided feedback and ideas, contributed in the design of the experiments, and edited the manuscript. GK set up and designed the project, attracted the funding, supervised the activities, and drafted/edited the manuscript. All authors contributed to the article and approved the submitted version.</p><!><p>The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.</p>

PubMed Open Access

PHOSPHOLIPIDOMIC STUDIES IN HUMAN CORNEA FROM CLIMATIC DROPLET KERATOPATHY

Climatic droplet keratopathy (CDK) is an acquired degenerative disease predominantly affecting males over 40 years old. It results in progressive corneal opacities usually affecting both eyes. CDK is multifactorial and its etiology remains unknown. Our recent findings are consistent with CDK pathology being driven by environmental factors with oxidative stress playing an important role (for example, contributing to lipid peroxidation) rather than climate factors. The changes in corneal lipid composition affected by environmental factors remain understudied. The purpose of this study was to systematically investigate phospholipids profile [phosphatidylcholine (PC) and phosphatidylserine (PS)] in corneas from CDK patients using tandem mass spectrometry. Samples from CDK areas and from non-affected areas were obtained from patients diagnosed with CDK who underwent cataract surgery, were subjected to lipid extraction using a modified Bligh and Dyer method; protein concentrations were determined using the Bradford\xe2\x80\x99s method. Lipids were identified and subjected to ratiometric quantification using TSQ Quantum Access Max triple quadrupole mass spectrometer, using appropriate class specific lipid standards. All phospholipid classes showed lower total amounts in affected areas compared to control areas from CDK\xe2\x80\x99s corneas. Comparative profiles of two phospholipid classes (PC, PS) between CDK areas and control areas showed several common species between them. We also found a few unique lipids that were absent in CDK areas compared to controls and vice versa. Lower amount of phospholipids in CDK areas compared to control areas could be attributed to the lipid peroxidation in the affected corneal regions as a consequence of increased oxidative stress.

phospholipidomic_studies_in_human_cornea_from_climatic_droplet_keratopathy

<!>Tissue samples procurement<!>Lipid extraction<!>Mass spectrometry<!>Statistical analysis<!>RESULTS<!>DISCUSION

<p>Climatic droplet keratopathy (CDK) is a bilateral degenerative corneal disease related to advanced age, lifestyle, and multiple environmental factors such as corneal micro traumas, low humidity and lack of adequate protection for exposure to ultraviolet radiation (UVR) (Suarez et al., 2015). CDK most commonly affects males over 40 years old, in the form of progressive opacities of the anterior layers of the cornea. This disease is rare in temperate latitudes, and it is commonly linked to overexposure to UVR, being endemic in certain rural communities around the world. Since the first description of the disease in 1898 (Baquis), CDK has been reported in different parts of the world following different criterions (presumed aetiology, geographic area, by eponym, clinical presentation, nature of corneal deposits and patient's activities) and more recently in the Patagonia region of Argentina, as has been reviewed by Serra et al. (2015).</p><p>CDK slowly progresses to corneal opacity, through three grades of increasing severity. In the initial stages (grade 1), multiple tiny and tightly confluent translucent extracellular sub epithelial deposits, located in the Bowman layer close to the temporal and/or nasal limbus, giving the affected cornea a tarnished, hazy aspect. In grade 2, haziness spreads over the inferior 2/3rds of the cornea, affecting the central cornea. At this stage, visual acuity may be moderate to severely affected due to the visual axis involvement. Grade 3 is characterized by the presence of large golden sub-epithelial droplets of different sizes (some of them are 1 mm in diameter) grouped in clusters that grow and cover the cornea as the disease progresses. In advanced cases, areas of vascularized anterior stromal opacification or fibrosis may be observed. In general, corneal sensitivity and visual acuity are severely affected at this stage (Freedman, 1965; Urrets-Zavalia et al., 2007).</p><p>Another clinical finding observed in some of our patients in the Argentinian Patagonia is a mild to severe atrophy of the 1/2 inferior iris stroma, more frequently observed in grades 2 and 3 (Urrets-Zavalía et al., 2007). Despite the harsh environmental conditions in which individuals from this region live, dry eye was not common among CDK patients, or controls (Urrets-Zavalía et al., 2007).</p><p>We have investigated the biological features of matrix metalloproteinases (MMPs) and their inhibitors TIMPs in patients with CDK, as these molecules control the degradation of the corneal epithelium and stroma. Our studies showed enhanced MMP-2 and MMP-9 levels and a decreased expression of TIMP-1 in CDK patients' tears (Holopainen et al., 2011). Immunohistochemistry showed that MMP-2 was expressed at the basement membrane zone in both control and affected corneas, but also marked the edges of the granular CDK deposits; MMP-9 expression was restrained to basal layers of the epithelium and was markedly induced in CDK corneas (Holopainen et al., 2012). We also investigated the effect of UVR in the production of MMPs and cytokines using an in vitro cellular model of immortalized human corneal epithelial cells (HCE). Exposure of HCE cells to UVR significantly increased MMP and pro-inflammatory cytokine secretion, suggesting an active participation of the corneal epithelium on CDK's pathogenesis (Holopainen et al., 2012).</p><p>Many advances that could help to understand ethiopathogenic mechanisms involved in CDK have been made in the last years, recently reviewed by Serra et al (2015).</p><p>Lipids constitute a unique group of biomolecules that mediate a large number of functional and structural activities in the cell, tending to maintain homeostasis (Checa et al., 2015), and constitute about 5% of the weight of mammalian cells (Fahy et al., 2009). Cellular lipids are highly complex and dynamic (Yang and Han, 2016). There is evidence that phospholipids may play an important role in the regulation of several ocular homeostatic mechanisms due to their presence in aqueous humor and subsequent changes under injury conditions (Liliom et al., 1998). Lipids are also important components of the tear film. The lipid layer is highly organized, consisting of a monolayer of phospholipids at the water-air interface, which provides a hydrophobic interface on which non-polar lipids expand (Rantamäki et al., 2011). Defects in the lipid layer of the tear film results in increased evaporation, which perpetuates ocular surface inflammation and damage. Also, lipids have been involved in ocular surface pathologies such as dry eye, allergic keratoconjunctivitis, infections and glaucoma (Ham et al., 2004; Robciuc et al., 2014; Fujishima et al., 2013; Aribindi et al., 2013a; Edwards et al., 2014).</p><p>There are no previous reports on identification or determination of superficial corneal lipid composition of patients with CDK. Lipid peroxidation may occur at the epithelium of the cornea exposed to prolonged exposure to UVR without adequate protection, in certain individuals under certain circumstances, which could result in an oxidative stress. Our aim was to systematically study lipids in corneal epithelial cells from CDK corneas using tandem mass spectrometry, and identify a differential composition, if any, between healthy corneal epithelial areas and CDK affected areas.</p><!><p>The samples were procured according to the protocol approved by institutional review process and the tenets of the Declaration of Helsinki. Corneal epithelial cell specimens from the eyes of two CDK patients (a 62-year old and 63-year old males) (Figure 1) at the moment of a planned cataract surgery were obtained and kept at −70° C until they were further processed. One CDK affected area, and one control non-affected area, were collected by scraping these regions of the cornea using a crescent knife. No abnormalities of the iris were observed biomicroscopically, and the ocular tensions, as well as the ocular fundus were normal in both eyes patients.</p><!><p>Corneal tissue was subjected to lipid extraction using a modified Bligh and Dyer method (Iverson et al., 2001). The organic phase with extracted lipids was dried in a Speed-Vac (Model 7810014, Labconco, Kansas City, MO, USA). Samples were flushed with argon gas to prevent oxidation. All extractions and subsequent handling were made using glass vials; polyvinyl plastic was avoided completely to prevent contaminating impurities. Corresponding aqueous phase extracted proteins were subjected to determination of concentration using Bradford's method (Bradford, 1976), and these concentrations were used to normalize lipids per amount of proteins.</p><!><p>Dried lipid samples were re-suspended in LC-MS grade Acetonitrile: Isopropanol (1:1). Samples were infused with a flow rate of 5μl/min, using Tri Versa Nanomate (Advion Inc., Ithaca, NY, USA), a chip-based electrospray ionization machine controlled with Chipsoft8.3.3 version software.</p><p>A triple quadrupole electrospray mass spectrometer (TSQ Quantum Access Max; Thermo Fisher Scientific, Pittsburgh, PA, USA) was used for analysis of lipids in infusion mode using TSQ Tune software that is part of the Xcaliber 2.3 software package. Samples were analyzed for 2.00 minutes with a 0.500-second scan. Scans typically ranged from 200 to 1000 m/z. A peak width was set at 0.7 and collision gas pressure was set at 1 mTorr. Sheath gas (nitrogen) was set to 20 arbitrary units. Auxiliary gas (Argon) was set to 5 arbitrary units. Settings for analyses of different phospholipid classes were established based on previous studies (Enriquez-Algeciras and Bhattacharya, 2013). The 0.1% formic acid (FA) was used as an additive method for analyses of lipids in the positive ion mode only (that is, for analyses of PC but not for PS). Control and CDK corneal epithelial cells were utilized for each of the two different class of phospholipid analyzed. Class specific lipids were quantified using class specific quantitative lipid standards:1,2-ditridecanoyl-sn-glycero-3-phosphocholine and 1,2-dioleoyl-sn-glycero-3-phospho-L-serine (sodium salt) (Avanti Polar Lipids, Inc; Alabaster, AL, USA). Approximately 5 scans each with and without internal standard (usually in the range of 0.1–10pmol) were performed for each sample. Ratiometric quantification was achieved using the MZmine 2.9 program. Lipid concentration was normalized to protein amount determined from the corresponding aqueous phase as described above.</p><p>Representative spectra for each sample were carefully and manually inspected by two independent observers from 5 spectra collected for each sample with and without the internal standard (total 10 spectra) and then used for further analyses. Spectra were converted to netCDF files from Thermo RAW files using the Xcalibur 2.3 software suite, subsequently imported into MZmine 2.9 (Edwards et al., 2014). Identification was carried out against a custom database created from the Lipid Maps Database (LMDB).</p><p>Identified lipids were subjected to analysis for determination of common and unique species using an Excel macro (Aribindi et al., 2013b). We defined unique when a given lipid species was found in only one group (control areas or CDK affected areas). Common species are those that have been found in samples from both control and CDK affected areas of both patients. All unique lipid experimental readings (the amount of lipid species pM/μg protein) were found to be significantly different from 0.0 by one sample t-test (p≤0.05).</p><!><p>Student's t-test was used for comparison between lipid concentrations in control areas vs CDK affected areas. A p-value ≤0.05 was considered significant.</p><!><p>We obtained lipid profiles for two classes of phospholipids: phosphatidylcholine (PC) and phosphatidylserine (PS), using established parameters (Enriquez-Algeciras and Bhattacharya, 2013). A representative PC and PS spectrum for control corneal samples without and with ratiometric standards (Fig. 2 A, 3A and Fig. 2 B, 3B, respectively) are shown. Using Excel Macros all data from two phospholipids were analyzed to determine the presence of common and unique lipid species in control areas and CDK affected areas.</p><p>All phospholipids classes showed higher total amounts in control areas compared to CDK affected areas. In grade 1, total phospholipids amount was 7.6 times higher in control than in CDK areas. In grade 2, phospholipids concentration was 35 times higher in control areas than in CDK affected areas. PC and PS concentrations were higher in the control than in CDK area in grade 1 (4.7 times and 10 times, respectively) as well as in grade 2 (5.7 times and 36 times, respectively). The total amount of all two classes of phospholipids normalized to total amount of proteins in the corresponding aqueous phase extractions are presented in table 1.</p><p>We found 35 unique PC species in the control areas and 38 species in CDK affected areas from grade 1 patient's sample. No unique PC species were found in samples from grade 2 (table 2, figure 4). For common lipids, we found 47 common PC species in control and CDK affected areas in grade 1 and 85 common PC species for grade 2 (table 3, figure 4).</p><p>In case of PS, 13 unique species were found in control areas and 8 in CDK affected areas from grade 1. For grade 2, 19 PS and 9 PS species were found in control and CDK areas, respectively (table 2, figure 4). Forty PS and 36 species were common between control and affected areas from grade 1 and grade 2, respectively (table 3, figure 4).</p><!><p>CDK is a rare corneal degenerative disease characterized by progressive opacity because of accumulation of translucent material in the Bowman layer of the cornea within the interpalpebral fringe (Gray et al., 1992; Urrets-Zavalía et al., 2006, 2007). We have previously shown that CDK occurred in individuals who have lived their entire life under certain unfavorable environmental conditions such as high exposure to UVR, chronic micro erosions of the cornea, low levels of ascorbic acid (AA) in their diets and serum, and lack of protection with sunglasses or hats (Suarez et al., 2015). For those reasons we have previously proposed that this corneal disease should be call environmental proteinaceous corneal degenerative disease instead of climatic droplet keratopathy (Suarez et al, 2015).</p><p>The precise chemical nature of globules accumulated under the corneal epithelium remains uncertain despite many efforts have been made to unravel its composition (Tabbara, 1986; Kaji et al., 2007; Menegay et al., 2008; Kaji et al., 2010). It has been suggested that corneal deposits are derived from plasma proteins that, after reaching an inflamed cornea from the limbal vessels, are degraded by excessive exposure to UVR (Gray et al., 1992). Although no mechanistic link between UVR exposure and CDK has been established, Holopainen et al. have addressed this issue analyzing tears and corneal specimens obtained from CDK affected eyes. They found that CDK samples have higher levels of MMP-2, MMP-9, and pro inflammatory cytokines than unaffected individuals (Holopainen et al., 2012). They hypothesized that these proteins are produced by the corneal epithelium, and showed that UVR-B is capable of inducing pro-inflammatory response in these cells with concomitant increase in MMP-2 and MMP-9 levels. This is then likely to increase the apoptotic/ necrotic response of the corneal epithelial cells, harm the integrity of the basement membrane, and, eventually, lead to visible changes in the corneal architecture.</p><p>Recently, we performed a study in which two groups of guinea pigs were exposed during 30 months to daily doses of UVR-B similar to the one received by CDK patients living in Patagonia Argentina, fed with AA sufficient or AA partially deficient diets, respectively. Superficial corneal debridement wounds were made in one eye using a rotating burr (Algerbrush II, Ambler Surgical, Exton, PA, USA) at the nasal or temporal corneal limbus (once a week, alternating each region) previous topical anesthetic applied on the ocular surface (proparacaine ophthalmic solution, Alcon laboratories Argentina, Buenos Aires, Argentina). We found that guinea pig corneal epithelial cell levels of MMP-9 were increased, and that ALDH3A1 activity (most abundant soluble protein in the cornea, Estey et al., 2007) was increased during the first 18 months and then it decayed. The lipid peroxidation marker malondialdehyde (MDA) (Grotto, 2009) concentration followed the same pattern than ALDH3A1 activity. Moreover, we found that animals fed with deficient AA diet showed more pronounced abnormalities in the cornea and in the crystalline lens (unpublished data). All these findings are compatible with a state of chronic oxidative stress.</p><p>It is well known that combination of near continual exposure of the cornea and ocular surface to UVR and molecular oxygen can provoke oxidative stress and tissue damage. Oxidative stress can be defined as an imbalance between reactive oxygen species (ROS) and antioxidants (Cejka and Cejkova, 2015). Among the antioxidants we can mention enzymatic and non-enzymatic molecules. Some of the enzymatic antioxidants are constituted by aldehyde dehydrogenases (ALDHs), catalase, and superoxide dismutase (Chance et al., 1979; Fridovich, 1995; Harrison and Arosio, 1996; Chen et al., 2013), whereas AA, reduced glutathione (GSH), α-tocopherol and NAD(P)H are among the non-enzymatic antioxidants (Dickinson and Forman, 2002; Machlin and Bendich, 1987).</p><p>Oxidative stress can cause peroxidation and further damage of lipids, nucleic acids, bases, and proteins. The eye is one of the major targets of the ROS and reactive nitrogen species (RNS) attack due to exposition to several environmental factors like high pressure of oxygen, light, UVR, ionizing radiation, chemical pollutants and pathogenic microbes, which are able to shift the redox status of a cell towards oxidizing conditions. There is increasing evidence indicating that persistent oxidative stress contributes to the development of many ocular diseases such as certain types of keratitis, pterygium, and pinguecula, among others (Kruk et al., 2015).</p><p>Our hypothesis for CDK genesis is that individuals with prolonged corneal exposure to multiple unfavorable environmental conditions (e.g., excessive UVR-B exposure, lack of vegetation/shade, dry/windy climate, particle bombardment, AA partial nutritional deficiency, lack of eye protection, genetic factors, etc.) would develop inflammatory processes and oxidative stress leading to progressive degradation and accumulation of proteinaceous material in Bowman's layer and, in advances cases of the disease, in the superficial stroma and deep epithelium (Urrets-Zavalia et al., 2012; Holopainen et al., 2012; Serra et al., 2015).</p><p>In the present study, we have determined phospholipids (PC and PS) present in control and CDK affected areas from patients' corneas using triple quadrupole mass spectrometry, in parent-ion and neutral loss scan modes with parameters previously established in the lipid field and widely used for ocular tissue in the study of some ophthalmological diseases (Han et al., 2012; Bhattacharya, 2013; Wang et al., 2016). Phospholipids are the basic building blocks of cell membranes, arranged as bilayer membranes. Membrane phospholipids create a hydrophobic environment for transmembrane protein function and communication. Some membrane lipids are substrates for production of lipid second messengers, which are metabolized by enzymatic activity from phospholipid precursors (Fahy et al., 2011; Aribindi et al., 2013b; Edwards et al., 2014; Li et al., 2015).</p><p>Corneal lipid content was studied in humans and animals many years ago (Feldman, 1967; Broekhuyse, 1968; Bazan and Bazan, 1984). Bazan et al. (1984) established that corneal rabbit epithelium has larger phospholipid content and more saturated than in the rest of the cornea. Conversely, corneal endothelium has a higher unsaturation level, according with water permeability of this layer (three times higher than in epithelium). Herein, we report data that clearly shows decreased levels of total phospholipids, PC and PS in the affected CDK epithelium area (in grade 1 as well as grade 2) compared to the non-affected areas. We also describe phospholipids uniquely enriched in control areas, and others only present in CDK affected areas (table 2). At the same time, corneal epithelial cells from CDK grade 1 and grade 2 patients' eyes present a common PC and PS composition (table 3).</p><p>It is known that UVR can induce lipid peroxidation in many cell types including human corneal epithelial cells. UVR leads to the production of ROS that initiate lipid peroxidation by attacking the polyunsaturated fatty acids chains in cell membrane phospholipids, and may cause the accumulation of reactive aldehydes. Unlike ROS, aldehydes are generally long-lived compounds that can diffuse through the cell and react with biological targets distant from the site of origin (Estey et al., 2007). Lipid peroxidation has been involved in several disorders such as atherogenesis, neurodegenerative diseases, cataractogenesis, and retinopathy (Awasthi et al., 1996; Witting et al., 1999; Butterfield et al., 2001; Kumar et al., 2001; Totan et al., 2001). Lipid peroxidation as a consequence of oxidative stress may have been taking place in corneas from CDK patients. This could account for the reduction in phospholipid concentrations that was observed.</p><p>In summary, we have studied for the first time phospholipid composition in corneal epithelial cells from CDK patients using shotgun lipidomics (direct infusion in triple quadrupole mass spectrometry), and we have found new evidences in favor of our hypothesis about the etiopathogenesis of CDK. The lower total amount of phospholipids observed in affected areas compared to control areas, and a differential composition between healthy corneal epithelial areas and CDK affected areas, could be signs of an active oxidative stress process occurring in CDK corneas.</p>

PubMed Author Manuscript

Predicting spike protein NTD mutations of SARS-CoV-2 causing immune escape by molecular dynamics simulations

The emergence of coronavirus disease 2019 (COVID-19) pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has been bringing the world to a standstill. Beyond all doubt, the most striking therapeutic target for antibody development is the spike (S) protein on the surface of virus. In contrast with an immunodominant receptor-binding domain (RBD) of the spike protein, little is known about neutralizing antibodies binding mechanisms of N-terminal domain (NTD), let alone the effect of NTD mutation on antibody binding and risk of immune evasion.Employing various computational approaches in this study, we investigated critical residues for NTD-antibody bindings and their detailed mechanism. The results showed that some residues on NTD including Y144, K147, R246 and Y248 are critically involved in the direct interaction of NTD with many monoclonal antibodies (mAbs), indicating that the viruses harboring these residue mutations may have high risk of immune evasion. Binding free energy calculations and the interaction mechanism study revealed that R246I, which is present in Beta (B.1.351) variant, may decrease or even abrogate the efficacies of many antibodies. Therefore, special attention should be paid to the mutations of the 4 residues for future antibody design and development.

predicting_spike_protein_ntd_mutations_of_sars-cov-2_causing_immune_escape_by_molecular_dynamics_sim

INTRODUCTION<!>Mapping the binding interface of S protein to 11 NTD antibodies<!>Key residues in NTD for neutralizing mAbs binding<!>Impact of R246I mutation on S protein for NTD-specific antibodies binding<!>Molecular mechanism for 5-24 binding to R246I mutant NTD<!>CONCLUSION<!>Preparation of mAb-S protein complexes<!>System preparation<!>Molecular dynamics (MD) simulations<!>Binding free energy calculation

<p>The coronavirus disease 2019 (COVID-19) pandemic induced by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) poses a serious threat to public health with severe socio-economic damage. As of August 2021, there are more than 213 million confirmed cases and about 4.44 million deaths worldwide 1 .</p><p>The pathogenic agent, namely SARS-CoV-2, is a kind of positive-sense RNA virus, consisting of spike (S) glycoprotein, membrane protein, envelope glycoprotein, lipid bilayer and inner structures 2 . It belongs to the betacoronavirus genus whose family members include Middle East respiratory syndrome coronavirus (MERS-CoV) and SARS-associated coronavirus (SARS-CoV). Similar to the two family members, SARS-CoV-2 infects host cells through S protein which can interact with the angiotensin-converting enzyme 2 (ACE2) entry receptor on host membranes with high affinity [3][4] . As a consequence, the S protein becomes the main target of neutralizing antibodies. S1 and S2 subunits are two functional components of the S protein, of which the former (S1) is further divided into receptor-binding domain (RBD) necessary for ACE2 binding, N-terminal domain (NTD) and other domains [5][6] . So far, plenty of mAbs recognizing RBD have been discovered such as REGN10933 7 , S2M11 8 . Compared with RBD-specific mAbs, a small number of antibodies targeted to NTD have been developed. Nevertheless, it is reported that some NTD-targeting mAbs are capable of neutralizing SARS-CoV-2 infection in vitro with high potency [9][10] . For instance, McCallum et al. described 41 NTD-specific human mAbs, among which S2X333 neutralized SARS-CoV-2 with an IC50 of 2 ng/mL, on par with the first-rate RBD mAbs S2E12 and S2M11 11 . Therefore, NTD may be another promising therapeutic target site.</p><p>Although prophylactic and/or therapeutic drugs are being developed at an unprecedented pace, it is still unknown when the epidemic will be under effective control all over the world. One of the most crucial reasons is that a larger number of prevalent mutations and deletions in S glycoprotein have emerged since the start of the outbreak and this rapid viral evolution could either facilitate transmissibility, or lead to reduction in protective efficacy of vaccines and mAbs [12][13][14] . Typically, SARS-CoV-2 Beta (B.1.351/501Y.V2) variant, emerging in late 2020 in Eastern Cape, South Africa (SA), contains 9 S protein mutations including three mutations (K417N, E484K, N501Y) in RBD, a cluster of NTD mutations (e.g., 242-244del & R246I) and so on 15 .</p><p>Wang et al. found that this variant is refractory to neutralization by convalescent plasma and vaccine sera 16 . McCarthy et al. found that 90% of deletion mutations occupied four discrete sites within the NTD 12 . What is worth mentioning is that SARS-CoV-2 Lambda (B.1.621) variant , a new variant of interest with higher infectivity and immune resistance, has T76I, L452Q, F490S and a unique 7-amino-acid insertion mutation (RSYLTPGD246-253N) in the N-terminal domain 17 . All of above have made the current drug development situation even grimmer. Hence, it matters laying emphasis on immunogenicity of different S protein domains and the specific mechanism of mAbs targeting them, including NTD.</p><p>Up to now, there have already been numerous studies to learn the effects and mechanism of mutations in the RBD domain [18][19][20] . In particular, RBD mutations containing E484K, N501Y and K417N have attracted considerable attention from researchers [21][22][23] . Whereas, the details of the NTD mutations have been still elusive until now, owing to the absence of attaching great importance to this antigenic site. When it comes to researches about NTD, recent reports elucidated that several currently circulating variants, comprising Beta (B.1.351) and Alpha (B.1.1.7), harbors some NTD mutations like Y144del, R246I, etc., and these lineages will partially or completely escape neutralization mediated by a variety of mAbs 16 revealed that the RSYLTPGD246-253N insertion mutation in the N-terminal domain of the Lambda S protein, is responsible for evasion from neutralizing antibodies 17 . By comparison, even though A222V possesses a high mutation rate according to statistic by GISAID sequence database 24 , it makes no difference to immunogenicity 11 .</p><p>Considering this, it is of great significance to understand binding modes of NTD antibody-antigen and distinguish detrimental mutations effectively from so much data.</p><p>Complementary to time-consuming wet lab study, computational methods are capable of providing a lot of details about protein-protein interaction binding pattern more than binding affinity. To our knowledge, no one has systematically studied the details of antibody-antigen binding and mutant effects for NTD yet using computing methods. Herein, we pay attention to in silico approaches to explore the crucial residues for NTD-antibody binding and predict mutational implications for neutralizing effectiveness. Methodically analyzing binding patterns by molecular dynamics (MD) simulation and end-point molecular mechanics generalized Born surface area (MM/GBSA) 25 binding free energy calculation helps to shed light on the binding mechanisms of NTD-antibody. Our results revealed that mutations of some residues on NTD including Y144, K147, R246, Y248 have the high hazard of immune evasion for many antibodies and R246I mutation may reduce the efficacies of most current NTD antibodies through abolishing the hydrogen bond and electrostatic interaction with antibodies. Our research findings could be beneficial for drug design and mAbs-based therapeutics in clinical.</p><!><p>Taking the integrity of structures and the availability of biological experiment data into account, we have studied 11 systems consisting of diverse antibodies in this paper (Table S1 for details). Through mapping the interface residues on S protein NTD, we found that the interface includes about 40 residues, which are Y144, R246, Y145, H146, R158, T250, K147, Y248, L249, P251, D253, Q14, V16, N148, S247, G252, S254, S255, C15, N17, K150, W152, S256, V143, E156, L18, T19, G142, H245, G257, N74, T76, K77, F140, E154, L244, W258, G75, T20 and T73. These residues constitute antibodies' epitopes, which concentrate in N terminus (residue 14-19), a β -sheet spanning residue 144-158 and a loop formed by residue 246-256, collectively forming an antigenic site on the pinnacle of the NTD (Figure 1, red surface).</p><p>Binding interfaces of all 11 antibodies are highly overlapping and flanked by four oligosaccharides at position N17, N74, N122 and N149, which may exert an impact on antibodies' binding.</p><!><p>We performed all-atom molecular dynamics simulation for 11 prototype glycosylated NTD-antibody systems, and the frames extracted from 30-60ns trajectories were applied to subsequent analysis. The key binding residues were identified from analysis of the MM/GBSA binding free energy decomposition results, and the residues with energy contribution < -1.00 kcal/mol were chosen as pivotal residues.</p><p>In total, there are 38 key residues in S protein NTD for all 11 systems (Figure 2A).</p><p>The greater the energy contribution of the residues means that its mutations are more likely to affect the effectiveness of antibodies. As indicated in Figure 2A, R246 takes the most vital part in five systems viz. 5-24 (PDB ID: 7L2F), 4A8 (PDB ID: 7C2L), FC05 (PDB ID: 7CWS), 2-51 (PDB ID: 7L2C), 1-87 (PDB ID: 7L2D) (Figure 2A, Table S2). In the five systems, the energy contribution of R246 is over 9.90 kcal/mol (Table S2). Therefore, the mutations of R246 have the highest risk of immune escape.</p><p>In addition, mutants of K147 may also affect the potencies of some antibodies. Research results by McCallum et.al already showed that K147T mutant weakens the potencies of 4A8 (PDB ID: 7C2L), S2X333 (PDB ID: 7LXY) and S2M28 (PDB ID: 7LY2), but has little implication in S2L28 (PDB ID: 7LXZ) 11 , which is in accordance with our prediction about K147 (Figure 2A).</p><p>When it comes to occupancy frequencies of key residues among 11 systems, there are 14 residues that have direct interaction with 3 or more antibodies, and 9 residues, concentrating in residues 144-147 and 246-252, that play an important role in 5 or more systems (Figure 2B). In particular, Y144 is involved in almost all mAbs binding except DH1052 (PDB ID: 7LAB), and the mutants of Y144 like Y144del and Y144F are already present in the real world 26 . Y144, K147, R246 and Y248 make a relatively large contribution in 8 or more antibodies binding, hinting mutations of these four residues have the high possibility to affect binding affinities of most NTD mAbs. Thus, we should pay special attention to these sites when optimizing antibodies to NTD.</p><p>In consideration of both binding free energy contribution and occupancy frequency, R246 on S protein has the strongest binding affinity to the antibodies among the first four residues with the high frequency, arousing our interest to do further investigation.</p><!><p>In light of the research of Wibmer et al., R246I has already appeared in SARS-CoV-2 Beta (B.1.351) variant 13 . To evaluate the binding affinity between various antibodies and R246I mutant NTD, MM/GBSA calculations were carried out with wild type NTD as control based on 2 independent MD runs lasting 60ns.</p><p>The results of binding free energy (ΔG) with glycans' contribution are hard to reproduce in each experiment (Table S3). In order to exploring more details, 500ns MD simulation was executed for 7L2F complex. As indicated in Figure S1, the three flexible sugars (N17, N74, N122 glycan), especially N17 glycan, move intensely and stay closed to the antibody within 3 angstroms most of the time for wild type complex (Figure S1A), while for R246I complex, the sugars are more than 5 angstroms away from the antibody in most of the time (Figure S1B). What's more, the energy contribution of glycans varies between -2.75 and -26.82 kcal/mol in the wild type complex, but only between -9.02 and 2.76 kcal/mol in the R246I mutant (Figure S1C).</p><p>In conclusion, the flexibility of glycans allows their binding free energy contribution to vary greatly over time and have a big gap between the two systems. Whereas, provided that single point mutations don't cause large conformational changes, the contribution of sugars in the wild type and mutant should be similar. It's hard to sample statistically significant conformations of complexes with sugars at different positions over a limited period of simulation time. In this article we focus on the effect of residue mutations on antibody binding, rather than the role of sugars. Thus, when calculating binding affinities between NTD and antibodies, we excluded the contribution of sugars.</p><p>As shown in Figure 3 (Table S4 for details), ΔG of wild type NTD is higher than that of R246I mutant in 8 systems except for 4-18 (PDB ID: 7L2E), DH1052 (PDB ID: 7LAB), S2X333 (PDB ID: 7LXY), suggesting that R246I mutant has the potential to impact the effectiveness of these 8 antibodies. As was expected, the energy contribution changes of R246 and I246 to the overall binding energy show the same tendency as ΔG (Figure 4). Moreover, we performed 500ns MD simulation for 7L2F complex, whose trends of ΔΔG in different length of simulation time are consistent with the result of 30-60ns (Table S5). This demonstrated that 60ns simulation is enough to obtain the variation tendency. Recent studies show that R246I can weaken the affinity of some antibodies, to varying degrees, including 5-24 (PDB ID:7L2F), 4A8 (PDB ID: 7C2L), 2-17 (PDB ID: 7LQW), FC05 (PDB ID: 7CWS) 16,27 , and our predictions of percentage change in the relative binding free energy for above four systems are -53.76%, -40.58%, -13.92%, -19.13% respectively (Figure 3, Table Therefore, our predicted trends of the relative binding free energy percentage are in good agreement with existing experimental findings, proving that our method is reliable to a certain extent.</p><p>According to our prediction, the percentage of relative binding free energy between wild type NTD and R246I mutant to antibodies are -28.85%, -16.95%, -13.87% and -13.72% for 1-87 (PDB ID:7L2D), 2-51 (PDB ID: 7L2C), S2M28 (PDB ID: 7LY2) and S2L28 (PDB ID:7LXZ) respectively, implying R246I mutation is possible to impair the efficacies of these four antibodies (Figure 3, Table S4). One of the most noteworthy things is that the affinity decreases the most for antibody 1-87 (PDB ID: 7L2D). These results suggest that it is best to avoid using these four types of antibodies alone, especially 1-87, against variants with R246I mutations. Moreover, the ΔΔG for 4-18 (PDB ID:7L2E), DH1052 (PDB ID:7LAB) and S2X333 (PDB ID:7LXY) are around 0 kcal/mol (Table S4), which is indicative of negligible impact on binding affinity (P value > 0.01). These three antibodies may not be susceptible to R246I mutation, so when targeting a mutant strain of SARS-COV-2 that has R246I mutation, they may work even better. Our findings can provide guidance on the clinical use of NTD antibodies. The binding free energy in WT is filled with blue, and the binding free energy in R246I mutant is filled with dark red. The relative binding free energy percentage (ΔΔG/ΔGWT =(ΔGR246I-ΔGWT)/ ΔGWT) between WT and R246I is filled with cyan. 251 snapshots from 30-60ns trajectories are used for per residue energy decomposition.</p><!><p>In order to understand the underlying mechanism of R246I mutation, we carried out a more detailed analysis for the trajectory of glycosylated NTD-(5-24) complex (PDB ID: 7L2F). Through energy decomposition of per residue, we found that R246 has a large negative value (-9.95±1.28 kcal/mol), while I246 has an energy contribution close to zero (-0.15±0.07 kcal/mol), which illustrates I246 has much less contact with 5-24 compared with R246 (Figure 5A). What's more, the contribution of residue D105 in 5-24 tends to be reversed before and after the mutation (Figure 5A), suggesting that R246 are in close contact with D105.</p><p>Based on above results, we analyzed hydrogen bonds and electrostatic interaction between D105 in 5-24 and R246 or I246 in NTD in the 500ns MD production phase.</p><p>Statistically, R246 has various amounts of stable hydrogen bonds with D105, which stabilizes at 2 hydrogen bonds and goes up to 3 sometimes (Figure 5B). However, I246</p><p>has no such interaction with antibody during all the simulation time (data not shown).</p><p>As expected, the positively charged R246 forms a strong electrostatic interaction with D105, while R246I mutation with a neutral charge abolishes that interaction (Figure 5C). To further demonstrate the stability of the electrostatic interaction, we analyzed the distance between R246 and D105 during the whole simulation process. The distance between R246 and D105 is less than 3.75 Å during 92% of the simulation time, while I246 is more than 6.00 angstroms away from D105 all the time (Figure 5D), corroborating the idea that this interaction between R246 and D105 is very stable and intense.</p><p>It should be noticed that the phenomenon of hydrogen bonds and electrostatic interaction formed between R246 and antibody observed in the 7L2F is not unique. We noted that R246 in NTD is in close contact with glutamic acid in several antibodies, viz 4A8 (PDB ID: 7C2L), FC05 (PDB ID: 7CWS), 2-51 (PDB ID: 7L2C), 1-87 (PDB ID: 7L2D), with a mode similar to R246-D105 as highlighted in Figure S2. Because of the above reasons, R246I mutation will decrease the binding affinity to many antibodies.</p><!><p>In the face of the severe epidemic of COVID-19, it is urgent to have a deep understanding of molecular mechanisms of harmful mutations and the vital antigenic epitopes when we are developing treatment and prevention methods. This paper applied a series of in silico methods to explore the binding patterns of different antibodies to S protein NTD. By per residue energy decomposition, some residues, including R246, Y144, K147 and Y248, are found to play a crucial role in multiple NTD-specific antibodies binding. This result reminds us that mutations of these residues has the potential to cause immune evasion. By the means of MD simulation and MM/GBSA calculation, we predicted that R246I mutation may decrease or even invalidate the effectiveness of some mAbs. Further analysis of the molecular mechanism revealed that the immune escape of R246I mutation from 5-24 could be, to a great degree, attributed to the abolishment of the strong hydrogen bonds and electrostatic attractions between R246 of NTD and D105 of antibody. The findings in this study makes it possible to optimize existing therapeutic antibodies, thus making them more efficacious against COVID-19.</p><!><p>As of May 1, 2021, we have retrieved 14 NTD-specific antibodies bound to S protein from the Protein Data Bank. Considering the integrity of structures and the availability of biological experiment data, we have studied 11 systems in this paper.</p><p>The NTD domain (residue 14 or 27-291) were truncated from the full-length S protein and both terminals are capped with ACE and NME, respectively. In order to get an intact structure, missing residues in flexible loops were modeled using SWISS-MODEL [28][29] . The interfaces analysis was carried out using scripts of pymol2.5 30 with a cutoff of 0.75. Glycosylated NTD wild type and R246I mutation models were generated using Glycan Reader module [31][32][33] available within CHARMM-GUI 34 according to Table S6.</p><!><p>Protonation states were assessed using H++ 3.2 [35][36] at pH 7.4. A cubic explicit water box described using the TIP3P model was used to solvated the complex system, which was extended by 12 Å from the solute. An atmosphere of 150 mM NaCl was also included in all simulations. The generated models were parametrized using CHARMM36 all-atom additive force fields for protein and glycans [37][38] . 8,000 steps of minimization including 6,000 steps of steepest descent minimization and 2,000 steps of conjugate gradient minimization were performed to remove bad contacts during the energy minimization phase. Subsequently, the temperature was incrementally changed from 0 to 300 K for 125ps at 1 fs/step. Equilibration in NPT ensemble was run at 1.0 bar and 300 K for 500,000 steps at 2fs/step. Pmemd program implement in Amber18 software package 39 was used to run the minimization, heating simulations with position constraints (1 kcal/mol/Å 2 ) on protein and dihedral Angle constraints on carbohydrates.</p><!><p>Pmemd.cuda in Amber18 was used to perform MD simulations at 300 K, 1 bar for all NTD-antibody complexes. Temperature and pressure were controlled by Langevin thermostat 40 and a Nosé-Hoover Langevin barostat [41][42] . Bonds involving hydrogen atoms were fixed by the SHAKE algorithm 43 . The cutoff distance applied for van der Waals interactions was 12 Å. All simulations were performed using particle-mesh Ewald (PME) for long-range electrostatic interactions 44 . Cpptraj module in Amber18 was used for trajectory processing.</p><!><p>Binding free energy (ΔG) of NTD-antibody complexes was calculated by MM/GBSA method 25 using GB OBC model(igb=2)with a salt concentration of 150 mM. In this study, the internal and external dielectric constants were set to 1.0 and 78.5</p><p>separately. When calculating the binding affinity excluding glycans' contribution, we made use of Amber18 built-in program Cpptraj to remove glycans in every trajectory, and pymol2.5 30 to add hydrogens for glycosylation sites. The free energy decomposition analysis was carried out using an internal program with idecomp=1.</p>

ChemRxiv

4-Aryl pyrrolidines as a novel class of orally efficacious antimalarial agents. Part 1: Evaluation of 4-aryl-N-benzylpyrrolidine-3-carboxamides

Identification of novel chemotypes with antimalarial efficacy is imperative to combat the rise of Plasmodium species resistant to current antimalarial drugs. We have used a hybrid target-phenotype approach to identify and evaluate novel chemotypes for malaria. In our search for drug-like aspartic protease inhibitors in publicly available phenotypic antimalarial databases, we identified GNF-Pf-4691, a 4-aryl-N-benzylpyrrolidine-3-carboxamide, as having a structure reminiscent of known inhibitors of aspartic proteases. Extensive profiling of the two terminal aryl rings revealed a structure-activity relationship in which relatively few substituents are tolerated at the benzylic position, but the 3-aryl position tolerates a range of hydrophobic groups and some heterocycles. Out of this effort, we identified (+)\xe2\x88\x9254b (CWHM-1008) as a lead compound. 54b has EC50 values of 46 nM and 21 nM against drug sensitive Plasmodium falciparum 3D7 and drug resistant Dd2 strains, respectively. Furthermore, 54b has a long half-life in mice (4.4 h) and is orally efficacious in a mouse model of malaria (q.d.; ED99 ~ 30 mg/kg/day). Thus, the 4-aryl-N-benzylpyrrolidine-3-carboxamide chemotype is a promising novel chemotype for malaria drug discovery.

4-aryl_pyrrolidines_as_a_novel_class_of_orally_efficacious_antimalarial_agents._part_1:_evaluation_o

Introduction<!>Synthesis.<!>Amide benzyl ring SAR.<!>Pyrrolidine aryl ring SAR.<!>Pyrrolidine Stereochemical SAR.<!>Revisiting the amide aryl ring SAR.<!>Pyrrolidines are potent on the drug-resistant Dd2 strain of P. falciparum.<!>Inhibition of Aspartic Proteases.<!>Affinity for the hERG channel.<!>In vitro and in vivo pharmacokinetics.<!>54b is orally efficacious in a mouse model of malaria.<!>Conclusions<!>General.<!>(\xc2\xb1)-(3R,4S)-methyl1-benzyl-4-(6-(trifluoromethyl)pyridin-3-yl)pyrrolidine-3-carboxylate (46a).<!>(\xc2\xb1)-(3R,4S)-1-tert-butyl3-methyl4-(6-(trifluoromethyl)pyridin-3-yl)pyrrolidine-1,3-dicarboxylate (47a).<!>(+)-(3R,4S)-1-(tert-butoxycarbonyl)-4-(6-(trifluoromethyl)pyridin-3-yl)pyrrolidine-3-carboxylic acid (51a) and (\xe2\x88\x92)-(3S,4R)-1-(tert-butoxycarbonyl)-4-(6-(trifluoromethyl)pyridin-3-yl)pyrrolidine-3-carboxylic acid (51a).<!>(+)-(3R,4S)-N-(4-(dimethylamino)benzyl)-4-(6-(trifluoromethyl)pyridin-3-yl)pyrrolidine-3-carboxamide (53).<!>(\xe2\x88\x92)-(3S,4R)-N-[(1S)-1-[4-(dimethylamino)phenyl]ethyl]-4-[6-(trifluoromethyl)pyridin-3-yl]pyrrolidine-3-carboxamide (52a).<!>(\xe2\x88\x92)-(3S,4R)-N-[(1S)-1-[4-(dimethylamino)phenyl]ethyl]-4-[4-(trifluoromethyl)phenyl]pyrrolidine-3-carboxamide (52b).<!>(+)-(3R,4S)-N-[(1S)-1-[4-(dimethylamino)phenyl]ethyl]-4-[6-(trifluoromethyl)pyridin-3-yl]pyrrolidine-3-carboxamide (54a).<!>(+)-(3R,4S)-N-[(1S)-1-[4-(dimethylamino)phenyl]ethyl]-4-[4-(trifluoromethyl)phenyl]pyrrolidine-3-carboxamide (54b).<!>(+)-(3R,4S)-N-[(1R)-1-[4-(dimethylamino)phenyl]ethyl]-4-[6-(trifluoromethyl)pyridin-3-yl]pyrrolidine-3-carboxamide (55).<!>Biological Assays.<!>In vitro Antimalarial Assays (3D7 and Dd2).<!>In vivo Antimalarial Efficacy Suppressive Assay.<!>Fluorescence polarization hERG assay.

<p>Efforts at reducing incidents of malaria have stalled in recent years after a decade of progressive reduction in disease burden. The majority of malaria-related deaths occur in Africa and are caused by Plasmodium falciparum, the most lethal to man of the Plasmodium species. The introduction of artemisinin and artemisinin combination therapies (ACT) in 2005 led to a remarkable 60% reduction in mortality rates between 2000–2015.1 However, according to the World Health Organization, there were ~219 million new cases of malaria and 435,000 deaths in 2017, highlighting a leveling out of progress in reducing global malarial disease burden since 2015.2 Furthermore, there are reports of emerging resistance to artemisinin in Southeast Asia.1, 3–4 As the artemisinins are the only fully effective class of antimalarial drugs available today, it is crucial that additional antimalarial drugs be developed with new mechanisms of action as the next line of defense to combat developing resistance to known drugs. These efforts, along with efforts to control malaria transmission will be needed if successful global eradication of malaria is to be achieved.</p><p>The Plasmodium parasite has a complex lifecycle including sexual replication in the mosquito stage and asexual replication in the human liver and blood stages. These lifecycle stages involve numerous potential opportunities for intervention. The challenge is to identify unexploited biological targets that will result in effective parasite killing. A broad class of proteases, the aspartic proteases, have been successfully exploited for the treatment of AIDS—more than 10 FDA approved drugs have been developed that inhibit the HIV aspartic protease. Plasmodium has multiple aspartic proteases that play key roles in the survival of the parasite in its human host.5–6 Identification of inhibitors of the Plasmodium aspartic proteases has the potential to provide a novel mechanism for anti-malarial therapies. However, a major challenge to developing protease inhibitors as drugs is identifying compounds with cellular activity commensurate with their enzyme inhibition potency. This is often due to poor physicochemical properties, efflux mechanisms, protein binding or poor target-related biochemical efficiencies.</p><p>Rather than focus on a substrate-based inhibitor design approach that may provide potent but non-drug-like peptidomimetics, we elected to identify known, drug-like aspartic protease inhibitors with inherent antimalarial cell activity for optimization. We termed this approach as a "hybrid target-phenotype drug discovery". Scientists at GlaxoSmithKine (GSK),7–8 Novartis,9 and St. Jude's Children's Hospital10 recently screened large chemical libraries for antimalarial activity in a standard Plasmodium falciparum 3D7-infected red blood cell assay. This effort identified ~20,000 compounds with antimalarial activity, and the data were made public. Since a number of aspartic proteases have been shown to play essential roles in the Plasmodium life cycle, we hypothesized that aspartic protease inhibitors with antimalarial activity may be acting through one or more Plasmodium aspartic proteases. Such inhibitors would provide a powerful starting point for drug discovery: phenotypic hits (cell active) with a limited but efficacious set of potential drug targets. By mining these databases for aspartic protease-inhibiting chemotypes, we and others have found multiple inhibitor classes with drug-like properties that may be acting on the parasite through an aspartic protease mechanism due to their structural similarity to known aspartic protease inhibitors.11–13 Indeed, we have shown drug-like compounds from the aminohydantoin series inhibits plasmespins II, IV, V, and X.6, 11 The pyrrolidine class described herein also appears to be particularly drug-like and is ideal for lead optimization to identify novel clinical candidates for treatment of malaria.</p><p>In our previous work, we evaluated a set of spiropiperidine hydantoins such as CWHM-505 (1) (Fig. 1).12 While we were able to improve the potency of this series to sub-100 nanomolar, they suffered from particularly poor metabolic stability, presumably due to the cleavage of the benzylic piperidine. A number of pyrrolidine and piperidine-based inhibitors of aspartic proteases BACE, renin and Pf plasmepsin 2 (PM-II) have been reported in the literature (2–5).14–18 These inhibitors have been shown to bind to the aspartic acid residues in the active site via the protonated nitrogen of the pyrrolidine (e.g., PDB ID: 3UFL).14</p><p>We were attracted to pyrrolidine as an antimalarial pharmacophore as it would avoid the primary metabolic liability in our previous work on benzyl piperidines (e.g., 1) as the basic nitrogen is not substituted in these compounds. Indeed, renin-inhibiting pyrrolidine (4) had been reported to have a two-hour half-life in rats.15 Using the simple unsubstituted pyrrolidine core (Fig. 2), we searched the GSK TCAMS and Novartis malaria phenotypic screening databases and found 45 hits containing this pharmacophore. Many of these hits were aminoquinoline analogs of chloroquine and amodiaquine and were thus eliminated from our consideration for lack of novel mechanism of action. Six compounds, however, were members of the unique antimalarial pharmacophore represented by racemic Novartis compound GNF-Pf-4691 (6; Fig. 2). We resynthesized GNF-Pf-4691 (6) and assayed it in a SYBR Green Pf 3D7 assay19 and found it to have an IC50 of 385 nM, within 3-fold of the reported database value. To our delight, we found this compound to have a 100-fold cytotoxicity selectivity index and an improved metabolic stability profile in mouse liver microsomes. On the basis of this data, we elected to initiate structure-activity relationship (SAR) studies to determine whether this pyrrolidine series should be pursued for full lead optimization. We independently varied the amide benzyl ring and the pyrrolidine aryl ring to develop some preliminary SAR and identified the preferred stereoisomers of the pyrrolidine core. Using this SAR, we combined preferred groups at each position to identify lead compounds suitable for in vivo efficacy studies.</p><!><p>It is well established that trans-3,4-disubstituted pyrrolidines can be readily and stereo selectively synthesized through a 3+2 cycloaddition reaction using N-(methoxymethyl)-N-(trimethylsilylmethyl)benzylamine.20–21 Accordingly, racemic trans-α,β-unsaturated esters (8) were prepared by the Wittig reaction and then converted to the corresponding pyrrolidines ((a) through the unstabilized azomethine ylide generated by TFA (Scheme 1). To simplify preparation of the final products, the benzyl group was replaced with Boc (9b). Ester hydrolysis followed by amide coupling and Boc-deprotection yielded the final trans-pyrrolidines (11) as racemates.</p><!><p>Holding the right-hand side constant as 4-trifluoromethylphenyl, we systematically modified the terminal amide aryl ring. Initially new analogs were synthesized as racemates to more efficiently profile the SAR and develop our understanding of the pharmacophore. Representative analogs are shown in Fig. 3. Substitution on the phenyl ring is essential for potency. 4-Methoxy (6) and 4-methyl (12; GNF-Pf-3587) groups give 10-fold increases in potency relative to H (13). The most potent substituent in this position identified to date is dimethylamine (14), approximately two-fold more potent than the lead methoxy analog. Conversion of the dimethyl aniline into an indoline ring (15) was tolerated but extending it as a diethyl aniline (16) reduced potency by 2–3-fold.</p><!><p>Holding the amide aryl ring constant as the dimethyl aniline, we systematically explored SAR of the pyrrolidine aryl ring(Fig. 4). This position is also quite sensitive to substitution patterns. 4-CF3 (14) is the preferred substituent with t-butyl (19) being an appropriate replacement, but replacement with methyl (17) or chloro (18) led to 3- to 4-fold reductions in potency. We also explored potential replacements for the trifluoromethyl group such as difluoromethyl (20), CF2CH3 (21), and pentafluorosulfide (22–23). These groups were suitable replacements with IC50 values ranging from 130 to 470 nM.</p><p>3,4-Disubstitution is tolerated and may even be favorable for enhancement of potency (24–29). For example, with an IC50 of 83 nM, the 4-CF3-3-Cl analog (26) is one of the most potent racemates we have identified. 2,4-Disubstitution was less well tolerated leading to 6–8-fold reductions in potency (25, 27).</p><p>Extension of the aryl ring with phenylether was tolerated in the 4-position (30) but not the 3-position (31) or as a pyridylether (32). Extending the aryl ring itself led to loss of potency (33).</p><p>The phenyl ring itself can be replaced with pyridine without loss in potency (34). Incorporation of a second ring nitrogen (e.g., pyrimidine 35) leads to modest erosion of potency. A number of other heterocycles were explored (36–44). Benzothiophenes (36–37, 39), benzofuran (38), and 2-quinoline (40) were all essentially equipotent with 4-trifluoromethylphenyl (14). Thiophene analogs designed to mimic the CF3 and t-butyl phenyl group were tolerated as t-butyl (41) but not so as CF3 (42). Other heterocycles such as thiazole (43) and pyrazole (44) led to >8-fold losses in potency.</p><!><p>The stereochemistry of the pyrrolidine was also investigated(Fig. 5). The cycloaddition chemistry leads to stereospecific formation of the pyrrolidine ring as a racemic mixture of trans isomers (46a-b). After replacement of the benzyl groups for Boc groups (47a-b) and hydrolysis of the methyl esters to the acids, the (S)-benzyloxazolidinone chiral auxiliary coupled to enable resolution of the resulting diastereomers 48a-b (first eluent) and 49a-b (second eluent) by silica gel chromatography. The isolated diastereomers were then hydrolyzed to furnish the carboxylic acids 50a-b and 51a-b as single enantiomers. These acids were then coupled to benzylic amines and deprotected to give the final products 52–55. The absolute stereochemistry was determined by x-ray crystallography for 54b to be the (3R,4S)-configuration (Fig. 5B).</p><p>There is a modest three-fold difference in potency between (3R,4S)-54a and (3S,4R)-52a enantiomers with the (3R,4S)-configuration being preferred (Fig. 5A). In contrast, stereochemistry at the α-benzyl carbon has a dramatic effect on potency with the (S)-Me group being tolerated (53) and the (R)-Me group being detrimental to potency by >30-fold (55). Comparing pyridine analog 54a versus phenyl analog 54b shows a nearly 3-fold preference for phenyl. A similar relationship holds for pyridine analog 52a versus phenyl analog 52b.</p><!><p>Having identified the (3R,4S) stereochemistry as preferred, we revisited our studies on the amide aryl ring SAR using the 4-trifluoropyridine moiety as the pyrrolidine aryl ring substituent as this would give lower lipophilicity. As described in the synthesis section, we utilized enantiomer 50a as a synthetic intermediate to derive a series of analogs. SAR with the 4-trifluoropyridine moiety is shown in Fig. 6. As with the 4-trifluoromethylphenyl series, the dimethylaniline is preferred (34). Modest changes to pyridine (56), pyrrolidine (57), and N-pyrazole (58) led to modest to dramatic losses in potency. Other substitutions (59–61) and extension of the dimethylamine (62) were not tolerated.</p><!><p>With an IC50 value of 51 nM, compound 54b was selected as our lead compound for further profiling. 54b was profiled in extensive side-by-side studies with chloroquine (CQ) in the multi-drug resistant Dd2 strain of P. falciparum. 54b has equivalent potency to CQ in the Pf 3D7 strain (IC50 = 46 ± 6 Nm vs. CQ IC50 = 38 ± 2 nM) and 10-fold greater potency than CQ against the multi-drug resistant Dd2 (IC50 = 21 ± 1 nM vs. CQ IC50 = 196 ± 14 nM).</p><!><p>Since our original hypothesis was that these pyrrolidines might be aspartic protease inhibitors, we profiled a select set of seven compounds for inhibition of human β-secretase (BACE1), Pf plasmepsin II (PM-II) and Pf plasmepsin IV (PM-IV) enzymes (Table 1; entries 1–7). However, none of the seven compounds inhibited these aspartic proteases. 6 was also tested for inhibition of Pf PM-V, PM-IX and PM-X in a knockdown assay (PM-V) and in western blots looking for PM-IX and X substrate processing in compound-treated parasites but was found to be inactive against these proteases (data not shown).6,11,22–23 Thus it appears that these pyrrolidines are not inhibitors of plasmepsins II, IV, V, IX or X. To date, we have not identified a biomolecular target for these pyrrolidines, and we cannot rule out inhibition of other Plasmodium aspartic proteases.</p><!><p>A potential concern with this series is its potential for binding the hERG channel given all of these compounds contain a basic amine pyrrolidine core.24–25 To address this, we evaluated ten compounds in this series for hERG binding in a competitive binding assay (Table 1). Compounds tested have binding affinities for hERG ranging from 2 μM to >50 μM. While some of the compounds have modest 8- to 20-fold hERG/3D7 ratios, our best compounds had hERG binding affinities of >20 μM and selectivity ratios of 90- to 600-fold. Compounds with a pyrrolidine phenyl ring (entries 1–6) tended to have stronger affinity for the hERG channel versus pyridine (34 and 54a) and pyrimidine (35) (entries 7–9). The exception to this trend is 54b (entry 10) which had a >600-fold selectivity vs. the hERG channel. It is possible that this is due to this compound being tested as a single enantiomer or the chiral α-methyl group may also reduce hERG potency.</p><!><p>Six compounds were profiled for metabolic stability in mouse liver microsomes (MLM; Table 1). All six compounds were moderately to very stable with MLM half-lives ranging from 31 to 161 min. 12 and 34 and 54b were selected for mouse PK studies (Table 2). 12 has relatively high clearance and a modest half-life in mice. In contrast, and 34 and 54b have good half-lives and low clearance in mice. 54b was found to have suitable oral bioavailability for in vivo efficacy studies.</p><!><p>Given the oral bioavailability of 54b, we evaluated it as a tool compound to provide in vivo proof-of-concept in the murine Peters 4-day suppressive test using P. chabaudi ASCQ (a CQ-resistant strain).19, 26–27 NIH mice were inoculated with P. chabaudi ASCQ parasitized red blood cells. After 4 h, 54b was dosed orally at 3, 10 and 30 mg/kg/day once daily for four days. Parasitemia levels were determined 24 h after the last treatment (Table 3). At 30 mg/kg/day, 54b inhibits parasitemia at 98.7%. The dose-response data allow us to approximate an ED90 of ~20 mg/kg/day and an ED99 of ~30 mg/kg/day. Compound concentrations in the plasma were determined at 1 h, 6 h and 24 h post dose on the last day of treatment. Compound concentrations in the plasma (110 nM at 24 h) remained above the Pf 3D7 IC50 (46 nM) for the full 24 h period only for the most fully efficacious dose of 30 mg/kg/day.</p><!><p>Herein, we describe the profiling of aryl pyrrolidine carboxamides as a novel class of antimalarial agents. We have generated an extensive SAR for the terminal aryl rings. The SAR is rather narrow in that relatively few substituents are permissible without significant losses in potency. On the carboxamide aryl ring, dimethylaniline is vastly superior to nearly every functional group evaluated. On the pyrrolidine aryl ring, para-trifluoromethyl can be replaced by a limited set of small lipophilic moieties, including t-butyl, CHF2, and CF2CH3. Addition of a halogen in the meta-position is beneficial. Replacement of the phenyl ring with heterocycles such as pyridine, pyrimidine, benzofuran and thiophene, is also tolerated. (3R,4S) is the preferred stereochemistry on the pyrrolidine ring and (S)-Me is preferred in the benzyl position. Out of this effort, we identified 54b as a lead compound for further profiling. 54b is more potent on the drug resistant Dd2 strain, has suitable selectivity over the hERG channel, has a long half-life in mice, and is orally efficacious in a mouse model of malaria (q.d.; ED99 ~ 30 mg/kg/day). Thus, compound 54b (CWHM-1008) presents a promising lead for optimization as an antimalarial drug with a low molecular weight, modest lipophilicity, excellent antimalarial potency, and long half-life oral bioavailability in mice.</p><!><p>Commercially obtained reagents were used without further purification. All reactions were monitored by TLC with silica gel-coated plates. Chemical structures and IUPAC names were generated using CambridgeSoft ChemDraw Ultra 10.0 or CDD Vault (www.collaborativedrug.com). Specific rotation value were recorded on AUTOPOL®μOJ-H, 4.6mm × 250mm 5 IV-T (γ= 589 nm, 50 mm cell, 20°C). MS analyses were performed on the API 2000 electrospray mass spectrometer in positive/negative ion mode. The scan range was 100–1000d. 1H NMR spectra were recorded on a Bruker AV-400 or 500 MHz spectrometer. Chemical shifts (δ) are given in relative to tetramethylsilane (δ 0.00 ppm) in CDCl3. Coupling constants, J, were reported in hertz unit (Hz). All compounds were ≥95% pure by HPLC conducted on an Agilent 1260 system using a reverse phase C18 column with diode array detector and a methanol/water (0.1% NH4OH) gradient unless stated otherwise. HRMS spectra were recorded on an ABSciex 5600+ instrument.</p><p>Compounds were filtered for PAINS substructures using the PAINS (Pan-assay interference compounds) filters available at http://fafdrugs4.mti.univ-paris-diderot.fr. 38 of the 47 final compounds described here have the dialkyl aniline substructure PAIN alert. However, this substructure is only considered an interference substructure for AlphaScreen technology as they can be quenchers of singlet oxygen and are not considered general PAINS.28 The assays used herein are cell-based assays and include correlation of compound potency from cell based assays to in vivo efficacy models.</p><!><p>To a stirred solution of methyl (triphenylphosphoranylidene)acetate (20.1 g,60.0 mmol) in dichloromethane (100 mL) was added 6-(trifluoromethyl)nicotinaldehyde (10.0g,57.1 mmol) at 0 °C, and then the resulting mixture was stirred at room temperature for 4 h . The reaction was monitored by TLC (10% ethyl acetate in petroleum ether). The reaction was concentrated in vacuo. Flash chromatography (petroleum ether/ethyl acetate: 95/5) afforded (E)-methyl 3-(6-(trifluoromethyl)pyridin-3-yl)acrylate (13.2 g, 99 % yield) as a white solid.</p><p>To a stirred solution of (E)-methyl 3-(6-(trifluoromethyl)pyridin-3-yl)acrylate (13.2 g, 57.1 mmol) in dichloromethane (150 mL) was added N-(methoxymethyl)-N-(trimethylsilymethyl)-benzylamine (17.8 g, 74.3 mmol). The resulting mixture was cooled to 0 °C and a solution of TFA (0.8 mL, 0.1 eq) in dichloromethane (2.0 mL) was added dropwise. The reaction mixture was allowed to warm to room temp and stirred for 16 h. The solvent was removed by evaporation in vacuo and the resulting oil was purified by flash column chromatography (petroleum ether/ethyl acetate: 80/20) to give the title compound (16.8 g, 81% yield). MS: m+1=365.3.</p><!><p>To a flask was added 46a (16.8 g, 46.2 mmol), ammonium formate (9.2 g,146 mmol), Pd/C (1.7 g, 10%) and MeOH (200 mL). The reaction flask was flushed with argon three times and then heated at 70°C for 30min. The reaction was cooled to room temp and filtered through Celite®, rinsing with MeOH (50 mL). To this mixture was added triethylamine (23.3 g, 231 mmol), then cooled to 0 °C and di-tert-butyl decarbonate (30.2 g ,139 mmol) was added dropwise. The reaction mixture was allowed to warm to room temp and stirred for 16 h. The solvent was removed by evaporation in vacuo and the resulting oil was purified by flash column chromatography (petroleum ether/ethyl acetate: 80/20) to give the title compound (15.6 g, 90% yield). MS: m+1=375.2.</p><!><p>To a flask was added 47a (15.6 g, 44.4 mmol) and methanol (150 mL). A solution of LiOH (4.7 g, 111 mmol) in water (100 mL) was added dropwise. The reaction mixture was stirred at room temperature for 4 h. The most of methanol was removed by evaporation, diluted with water and basified with hydrochloric acid (2M) to pH 4. The mixture was extracted with EtOAc (3 × 150 mL). The combined organic extracts were washed with water then brine, dried over sodium sulfate, filtered and concentrated to afford (±)-(3R,4S)-1-(tertbutoxycarbonyl)-4-(6-(trifluoromethyl)pyridin-3-yl)pyrrolidine-3-carboxylic acid (14.5 g, 91% yield).</p><p>To a suspension of (±)-(3R,4S)-1-(tert-butoxycarbonyl)-4-(6-(trifluoromethyl)pyridin-3-yl)pyrrolidine-3-carboxylic acid (14.5 g, 40.3 mmol) in anhydrous THF (300 mL) was added triethylamine (10.2 g, 101 mmol). The resulting mixture was cooled to −20 °C and a solution of pivaloyl chloride (5.8g, 48.4 mmol) in THF (20 mL) was added dropwise. The reaction mixture was stirred at −20 °C for 2 h and then added LiCl (1.8 g, 44.3 mmol) was added followed by benzyl-2-oxazolidinone (7.1 g, 40.3 mmol). The reaction mixture was allowed to warm to room temp and stirred for 4 h. The solvent was removed by evaporation in vacuo and the resulting oil was purified by flash column chromatography (petroleum ether/ethyl acetate: 90/10) to separate the mixture of 48a and 49a.</p><p>The first eluting compound was (3R,4S)-tert-butyl3-((S)-4-benzyl-2-oxooxazolidine-3-carbonyl)-4-(6-(trifluoromethyl)pyridin-3-yl)pyrrolidine-1-carboxylate (48a) obtained as a white solid (6.7 g, 32% yield). TLC Rf = 0.52 (petroleum ether/ethyl acetate: 3/1). To a flask was added 48a (6.7 g, 12.9 mmol) and THF (100 mL). A solution of LiOH (1.1 g, 25.8 mmol) in water (25 mL) was added dropwise. The reaction mixture was stirred at room temperature for 4 h. The most of methanol was removed by evaporation, diluted with water and basified with hydrochloric acid (2M) to pH 4. The mixture was extracted with EtOAc (3×150 mL). The combined organic extracts were washed with water then brine, dried over sodium sulfate, filtered and concentrated to afforded (+)-(3R,4S)-1-(tert-butoxycarbonyl)-4-(6-(trifluoromethyl)pyridin-3-yl)pyrrolidine-3-carboxylic acid (50a) (3.8 g, 82% yield). MS: m-1=359.2. 1H NMR (400 MHz, DMSO-d6) δ ppm 8.74 (s, 1H), 8.09 (d, J = 8.0 Hz, 1H), 7.86 (d, J = 8.0 Hz, 1H), 3.81 (t, J = 8.8 Hz, 1H), 3.73 (td, J = 8.8,1.6 Hz, 1H), 3.67 (m, 1H), 3.49–3.36 (m, 2H), 3.29 (t, J = 10.0 Hz, 1H), 1.41 (d, J = 8.4 Hz, 9H). [α]D20 +30.1 (c 0.09, MeOH).</p><p>The second eluting compound was diastereomer (3S,4R)-tert-butyl3-((S)-4-benzyl-2-oxooxazolidine-3-carbonyl)-4-(6-(trifluoromethyl)pyridin-3-yl)pyrrolidine-1-carboxylate (49a) obtained as a white solid (7.1 g, 34% yield). TLC Rf = 0.41 (petroleum ether/ethyl acetate: 3/1). 49a was hydrolyzed with LiOH as described above to give (−)-(3S,4R)-1-(tert-butoxycarbonyl)-4-(6-(trifluoromethyl)pyridin-3-yl)pyrrolidine-3-carboxylic acid (51a) (3.7 g, 75% yield). MS: m-1=359.2. 1H NMR (400 MHz, DMSO-d6) δ ppm 8.75 (s, 1H), 8.10 (d, J = 8.0 Hz, 1H), 7.87 (d, J = 8.0 Hz, 1H), 3.81 (t, J = 8.8 Hz, 1H), 3.77–3.66 (m, 2H), 3.44 (br, 2H), 3.30 (t, J = 9.6 Hz, 1H), 1.41 (d, J = 9.6 Hz, 9H). [α]D20 −30.4 (c 0.13, MeOH).</p><!><p>To a suspension of 50a (400 mg, 1.11 mmol), 4-(aminomethyl)-N,N-dimethylaniline (200 mg, 1.33 mmol), HATU (630 mg, 1.66 mmol) in dichloromethane (10 mL) was added triethylamine (0.31 ml, 2.22 mmol). The reaction mixture was stirred at room temp for 16 h. The mixture was diluted with dichloromethane (30 mL) and washed with saturated sodium carbonate solution then brine, dried over sodium sulfate, filtered and concentrated. The residue was purified by flash column chromatography (petroleum ether/ethyl acetate: 1/1) to give (330 mg, 61% yield) of (3R,4S)-tert-butyl3-(4-(dimethylamino)benzylcarbamoyl)-4-(6-(trifluoromethyl)pyridin-3-yl)pyrrolidine-1-carboxylate as a white solid.</p><p>To a suspension of (3R,4S)-tert-butyl3-(4-(dimethylamino)benzylcarbamoyl)-4-(6-(trifluoromethyl)pyridin-3-yl)pyrrolidine-1-carboxylate (330 mg, 0.67 mmol), in dichloromethane (6 mL) was added TFA (3.0 ml). The reaction mixture was stirred at room temp for 2 h. The solvent was removed by evaporation in vacuo and the resulting oil was diluted with dichloromethane (30 mL) and washed with saturated sodium carbonate solution then brine, dried over sodium sulfate, filtered and concentrated. The residue was purified by flash column chromatography (DCM/NH3 in MeOH: 20/1) to give the title compound 160 mg (61% yield). 1H NMR (500 MHz, DMSO-d6) δ ppm 8.67(s, 1H), 8.25(t, J=5.5, 1H),7.96 (d, J=8.0, 1H), 7.83 (d, J=8.0, 1H), 6.86 (d, J=8.5, 2H), 6.58(d, J=8.5, 2H), 4.18(dd, J=14.5, 6.0, 1H), 4.01(dd, J=14.5, 6.0, 1H),3.52(q, J=10.5, 1H), 3.26 (t, J=10.0, 1H), 2.99 (q, J=8.0, 1H), 2.97(q, J=8.0, 1H),2.85 (t, J=10.0, 1H), 2.83 (s, 6H). [α]D20 +73.2 (c 0.11, MeOH). HRMS (ESI) m/z: [M + H]+ Calcd for C20H24F3N4O 393.1902; found 393.1889.</p><!><p>The title compound was synthesized from 51a and (S)-4-(1-aminoethyl)-N,N-dimethylbenzenamine as according to the method described for 53. Yellow solid (110 mg, 54% yield). 1H NMR (400 MHz, DMSOd6) δ ppm 8.66 (s, 1H), 8.15 (d, J = 8.0 Hz, 1H), 7.97 (d, J = 8.0 Hz, 1H), 7.84 (d, J = 8.0 Hz, 1H), 7.06 (d, J = 8.4 Hz, 2H), 6.65 (d, J = 8.4 Hz, 2H), 4.79 (m, 1H), 3.36 (m, 1H), 3.30 (1H, overlapped with the peak of H2O), 3.20 (m, 1H), 2.90–2.84 (m, 9H), 1.23 (s, 1H), 1.19 (t, J = 6.8 Hz, 3H). [α]D20 −122.4 (c 0.12, MeOH). HRMS (ESI) m/z: [M + H]+ Calcd for C21H26F3N4O 407.2058; found 407.2043.</p><!><p>The title compound was synthesized from 51b and (S)-4-(1-aminoethyl)-N,N-dimethylbenzenamine as according to the method described for 53. White solid, 95 mg (56% yield). 1H NMR (400 MHz, DMSO-d6) δ ppm 8.10 (d, J = 8.4Hz, 1H), 7.65 (d, J = 8.0 Hz, 2H), 7.47 (d, J = 8.0 Hz, 2H), 7.06 (d, J = 8.4 Hz, 2H), 6.66 (d, J = 8.4 Hz, 2H), 4.79 (m, 1H), 3.49 (q, J = 16.0 Hz, 1H), 3.30 (1H, overlapped with the peak of H2O), 3.18 (t, J = 8.0 Hz, 1H), 2.90–2.80 (m, 7H), 2.75 (dd, J = 14.4,8.8 Hz, 1H). [α]D20 −156.7 (c 0.10, MeOH). HRMS (ESI) m/z: [M + H]+ Calcd for C22H27F3N3O 406.2106; found 406.2102.</p><!><p>The title compound was synthesized from 50a and (S)-4-(1-aminoethyl)-N,N-dimethylbenzenamine as according to the method described for 53. White solid, 78 mg (48% yield). 1H NMR (400 MHz, DMSOd6) δ ppm 8.62 (s, 1H), 8.15 (d,J = 8.0 Hz , 1H), 7.90 (dd, J = 8.0 Hz, 1H), 7.81 (d, J = 8.0 Hz, 1H), 6.78 (d, J = 8.4 Hz, 2H), 6.51 (d, J = 8.8 Hz, 2H), 4.79 (m, 1H), 3.41 (m, 1H), 3.30 (2H, overlapped with the peak of H2O), 3.20 (t, J = 7.2 Hz, 1H), 2.98 (m, 2H), 2.85 (t, J = 6.4 Hz, 1H), 2.83 (s, 6H), 1.24 (d, J = 7.2 Hz, 3H). [α]D20 +38.3 (c 0.12, MeOH). HRMS (ESI) m/z: [M + H]+ Calcd for C21H26F3N4O 407.2058; found 407.2048.</p><!><p>The title compound was synthesized from 50b and (S)-4-(1-aminoethyl)-N,N-dimethylbenzenamine as according to the method described for 53. White solid, 125 mg (63% yield). 1H NMR (500 MHz, DMSO-d6) δ ppm 8.09 (d, J = 8.0 Hz, 1H), 7.61 (d, J = 8.0 Hz, 2H), 7.41 (d, J = 8.0 Hz, 2H), 6.83 (d, J = 8.5 Hz, 2H), 6.53 (d, J = 8.5 Hz, 2H), 4.81 (m, 1H), 3.39 (q, J = 17.0 Hz, 1H), 3.28 (t, J = 9.5 Hz, 1H), 3.19 (t, J = 8.0 Hz, 1H), 2.97 (t, J = 8.0 Hz, 1H), 3.39 (q, J = 16.0 Hz, 1H), 2.82 (s, 6H), 2.77 (t, J = 10.0 Hz, 1H), 1.25 (d, J = 8.0 Hz, 3H). [α]D20 +39.1 (c 0.14, MeOH). HRMS (ESI) m/z: [M + H]+ Calcd for C22H27F3N3O 406.2106; found 406.2098.</p><!><p>The title compound was synthesized from 50a and (R)-4-(1-aminoethyl)-N,N-dimethylbenzenamine as according to the method described for 53. White solid, 102 mg (58% yield). 1H NMR (400 MHz, DMSO-d6) δ ppm 8.66 (s, 1H), 8.16 (d, J = 8.0 Hz, 1H), 7.95 (d, J = 8.0 Hz, 1H), 7.84 (d, J = 8.0 Hz, 1H), 7.05 (d, J = 8.0 Hz, 2H), 6.65 (d, J = 8.4 Hz, 2H), 4.78 (m, 1H), 3.50 (q, J = 15.2 Hz, 1H), 3.30 (2H, overlapped with the peak of H2O), 3.20 (t, J = 8.0 Hz, 1H), 2.87–2.67 (m, 8H), 1.18 (d, J = 6.8 Hz, 3H). [α]D20 +149.6 (c 0.12, MeOH). HRMS (ESI) m/z: [M + H]+ Calcd for C21H26F3N4O 407.2058; found 407.2057.</p><!><p>Aspartic protease enzyme assays, mouse liver microsome assays, in vivo PK studies, in vivo efficacy studies were performed as we have previously described.11</p><!><p>In vitro antimalarial activity was determined by a malaria SYBR Green I-based fluorescence (MSF) method described previously by Smilkstein et al.29 with slight modification.30 Stock solutions of each test drug were prepared in DMSO at a concentration of 20 mM. The drug solutions were serially diluted with culture medium and distributed to asynchronous parasite cultures on 96-well plates in quadruplicate in a total volume of 100 μl to achieve 0.5% parasitemia with a 2% hematocrit in a total volume of 100 μl. The plates were then incubated for 72 h at 37°C. After incubation, 100 μl of lysis buffer with 0.2 μl/ml SYBR Green I was added to each well. The plates were incubated at 37°C for an hour in the dark and then placed in a 96-well fluorescence plate reader (Spectramax Gemini- EM; Molecular Diagnostics) with excitation and emission wavelengths at 497 nm and 520 nm, respectively, for measurement of fluorescence. The 50% inhibitory concentration (IC50) was determined by nonlinear regression analysis of logistic dose-response curves (GraphPad Prism software). Antimalarial potency of compounds was determined by this technique for both Plasmodium falciparum 3D7 (CQ-sensitive) and Dd2 (multi-drug resistant) strains. 3D7 assay results are given as an average of at least three replicates.</p><!><p>In vivo antimalarial activity was determined for 54b against the rodent CQ-resistant Plasmodium chabaudi ASCQ strain according to the 4-day suppressive test.19, 31 Briefly, 5-week old female NIH mice (n = 6 per group) were inoculated i.p. with 2×107 parasitized (Plasmodium chabaudi ASCQ) red blood cells. Thereafter, the compounds were administered as a suspension in 0.5% carboxymethyl cellulose orally to the animals once daily at 4 h, 24 h, 48 h and 72 h post inoculation. Groups including a vehicle control and chloroquine (CQ) as a reference drug were included. Parasitemia levels were determined on the day following the last treatment (Day 4). Plasma drug concentration was determined by LCMS using 100 |μL orbital blood plasma sample from Day 3, which was collected at 1 h, 6 h, 24 h after dosing.</p><p>The Institutional Animal Care and Use Committee at the Guangzhou Institutes of Biomedicine and Health, Chinese Academy of Sciences, reviewed and approved the animal use in these studies. The animal care and use program is run entirely according to Association for Assessment and Accreditation of Laboratory Animal Care, International (AAALACi) standards and is assured by the Office of Laboratory Animal Welfare (OLAW identification number A5748-01).</p><!><p>The inhibition activity of compounds on hERG potassium channel was determined using Predictor® hERG Fluorescence Polarization Assay Kit (Life Technologies, Carlsbad, CA, USA). Experiments were performed according to the manufacturer's instructions. In brief, the reactions were carried out in 384-well plates including 10 μL of Predictor® hERG Membrane and 10 μL Predictor® hERG Tracer Red with appropriate amount of compound or positive control. Reactions were incubated for 2 h at room temp and then read on an EnVision Multilabel Reader (Perkin Elmer, Inc.) using polarized excitation and emission filters. Data were analyzed using Graphpad Prism 5 (GraphPad Software Inc., San Diego, CA).</p>

PubMed Author Manuscript

3D MALDI mass spectrometry imaging reveals specific localization of long-chain acylcarnitines within a 10-day time window of spinal cord injury

We report, for the first time, the detection and specific localization of long-chain acylcarnitines (LC ACs) along the lesion margins in an experimental model of spinal cord injury (SCI) using 3D mass spectrometry imaging (MSI). Acylcarnitines palmitoylcarnitine (AC(16:0)), palmitoleoylcarnitine (AC(16:1)), elaidic carnitine (AC(18:1)) and tetradecanoylcarnitine (AC(14:1)) were detected as early as 3 days post injury, and were present along the lesion margins 7 and 10 days after SCI induced by balloon compression technique in the rat. 3D MSI revealed the heterogeneous distribution of these lipids across the injured spinal cord, appearing well-defined at the lesion margins rostral to the lesion center, and becoming widespread and less confined to the margins at the region located caudally. The assigned acylcarnitines co-localize with resident microglia/macrophages detected along the lesion margins by immunofluorescence. Given the reported pro-inflammatory role of these acylcarnitines, their specific spatial localization along the lesion margin could hint at their potential pathophysiological roles in the progression of SCI.

3d_maldi_mass_spectrometry_imaging_reveals_specific_localization_of_long-chain_acylcarnitines_within

<!>Results<!>3D MSI demonstrates the heterogeneous shape of lesion.<!>Discussion<!>Maldi MSI.

<p>SCI is a devastating medical condition leading to irreversible damage of the central nervous system (CNS). Traumatic SCI can lead to paralysis with complete or partial loss of neurological functions below the injury site, and this can result from several different causes such as road traffic crashes, falls, and violence 1 . Nowadays, the increased incidence of trauma may be related to popular sports such as ice hockey, American football, rugby, horse riding and diving 2,3 . Currently, there are no effective therapies available for SCI patients. The long-standing challenge facing researchers is to develop effective strategies to prevent further tissue loss, maintain the health of living cells, and replace cells that have died to enable axonal growth and reestablish synapses that restore neural circuits essential for proper functional recovery 4 .</p><p>A key factor for effective therapy is elucidation of the distinct phases involved in SCI and the cellular and molecular events underlying them 3 . Diverse groups of cells and molecules from the nervous, immune, and vascular systems are implicated. Most participating cells reside in the spinal cord; however, others are translocated to the site of injury from the circulatory system. Thus, after primary trauma, cellular and molecular injury and inflammatory cascades are initiated, causing activation of resident microglia and astrocytes coupled with infiltration of innate immune cells including lymphocytes and monocytes. Furthermore, the local release of cytokines and chemokines by microglia, macrophages and neural cells induces a particular environment that can be either neurotoxic or neurotrophic [4][5][6] . During acute phase, macrophages phagocyte cell debris and glial scar formation is hypothesized to protect healthy tissue 7 . Chronic inflammatory processes (weeks post trauma) lead to aberrant tissue remodeling and nerve tissue dysfunction. Various cellular and molecular events designed to heal the injury can paradoxically lead to further neuronal injury or even cell death. The site of injury may spread to adjacent areas of the spinal cord, sometimes extending four spinal segments above and below the initial lesion site. The affected area markedly expands, becomes filled with immune cells, and a "scar" is formed 7 .</p><p>One of the approved clinical treatments for SCI is administration of methylprednisolone that can modulate the inflammatory process. However, a high-dose of methylprednisolone is often associated with severe immunosuppression and side effects, such as pulmonary or urinary tract infections 8,9 . In addition to mono-therapies, more complex cellular therapies are being suggested carrying several advantages and targeting several SCI-associated conditions such as: to bridge cavities or cysts, to replace dead cells, to create a favorable environment, and to allow axonal regeneration [8][9][10] . However, none of these provides a total understanding of the injury-inflammatory mechanisms involved in the lesioned spinal cord and proximities that can be used for a temporal and segment-specific target in SCI treatment. Thus, the molecular cross-talk occurring among cellular inhabitants at the lesion site and the adjacent segments needs to be investigated for this purpose.</p><p>Thus, in order to get an accurate view of the injury-driven mechanisms where the inflammatory process and neural injury are implicated, we have extended our previous analysis 5 to involve a spatiotemporal lipidomic evaluation by carrying out 3D Matrix-Assisted Laser Desorption/Ionization (MALDI) MS imaging across the SCI tissue. Combined with the most advanced tools for processing and statistical analysis of MSI datasets, we demonstrate the advantage of this molecular imaging method in probing SCI to provide novel insights into its pathophysiological mechanism.</p><!><p>2D MSI reveals lesion-specific lipids after SCI. 2D MALDI MS imaging of uninjured rat spinal cord typically shows distinct distribution of different lipid species. They are contained within the gray and white matter, resulting in spectra clustering according to these two regions, e.g., distribution of PC [16:0/16:0], m/z 830.5 and m/z 768.6 in sections taken from the cervical lower C5-C6 (R2) and lumbar L6-S1 segments (C3) that were unaffected by the lesion, where the distinct localization of these signals in the dorsal horn, the rest of the gray matter (intermediate zone and ventral horn), and the white matter, respectively, can be observed (Supplementary Fig. S1). When the pixels are colored based on these clusters, the well-recognizable "butterfly" shape of the gray matter is evident, with the posterior and anterior gray horns well-defined (Fig. 1a). Upon induction of SCI at thoracic segments Th8-Th9, disruption of this distribution is observed at the lesion center and penumbra, leading to the disappearance of high MW signals (ie., m/z 750-1000) and generation of low-MW signals (ie., m/z 300-600) in all areas affected by the lesion, i.e., at the lesion site (L) itself as well as in the caudal parts of R1 (adjacent segment rostral to injury) and rostral parts of C1 (adjacent segment caudal to injury). Consequently, the area Figure 1. Mapping of spatial segmentation results using all imaged tissue sections. (a) Segmentation was obtained by k-means clustering using Euclidean distance as metric. The segmentation maps represent the two main branches comprised of spectra from the gray matter (flesh) and white matter (light blue) regions. Sections in L show significant disruption of molecular content indicated by the loss of the "butterfly" shape typical of the gray matter region (red arrows mark the beginning and end of loss of butterfly shape). The order by which the sections were taken from the spinal cord is from left to right on each row (indicated by the violet arrow). (b) Stacking of the dissected ion images of m/z 844.5 and m/z 768.5 representing the gray and white matter regions, respectively. Distortion of ion distribution due to the lesion is more prominent in the gray matter. attributed to the gray matter disappears or appears to be distorted in sections taken close to the lesion center at this and at 7 and 10 days post-SCI (Fig. 1b).</p><p>Automatic elucidation of co-localized m/z by Pearson's correlation analysis confirms the lower intensities of high-MW signals at the lesion site (Supplementary Table S1). At the lesion site instead, Pearson's correlation highlights that predominantly co-localized signals are low-MW. Among these, LC ACs AC(16:1) (m/z 398.3), AC(16:0) (m/z 400.3), and AC(18:1) (m/z 426.4) and an oxidized long-chain fatty acid FA 26:1(9OH,10OH) (m/z 427.4) are specific to this area, in addition to being co-localized at the lesion site. This finding is cross-validated by ROC analysis (Supplementary Table S2). Indeed, pairwise comparison of spectra taken from the lesion (L) segments at each time point post-SCI reveals that these signals are present at all time points. Comparison of the intensities of these m/z, normalized against the total ion count (TIC), and grouped according to the different time points and segments, shows that most of them (AC(16:1), AC(18:1), lysophosphatidylcholines lysoPC(16:0) (m/z 518.3) and lysoPC(18:1) (m/z 544.3), m/z 534.3 and m/z 562.3) have more elevated intensities in the lesion segment and at 3 days post lesion (Fig. 2). On the other hand, few m/z (m/z 721.6, m/z 722.5, and m/z 725.5) also exhibit increased intensities at the lesion segment but are more intense at 10 days after lesion.</p><p>Examination of the time point-specific, lesion-discriminative m/z shows that most of the signals associated with the SCI lesion at 3 days are low-MW, while those associated with the lesion at 10 days are high-MW (examples shown in Fig. S2). The latter includes PC [16:0/20:4] (m/z 820.5), previously identified as one of the arachidonic acid-containing PCs and temporarily elevated at the SCI impact site 11 . Other m/z include those observed in normal spinal cord (m/z 770.5, m/z 772.6), as the injured site 10 days post-lesion is less extended most probably due to glial scar formation 4 .</p><p>Detailed examination of the lesion site, by imaging of the lesion-co-localized ions, highlights further sub-structuring within the lesion. As shown in Fig. 3 at 3 days post-SCI, AC(16:0) (as well as AC(16:1) and AC(18:1)) presents a distinct distribution surrounding the lesion site. On the other hand, lysoPC(18:1) (as well as lysoPC(16:0) and lysoPC(18:0) (m/z 546.3), and m/z 534.3) are distinctly distributed within the lesion core not overlapping the surrounding ions. In addition, the heme fragment of hemoglobin (m/z 616.2) is also observed distinct from that of the other ions indicating that the specific distribution of lipids is not related to the hemorrhage. Similar patterns of lipid distribution were found at 7 and 10 days post-injury confirming that this particular distribution is not specific to a single time point (Supplementary Figs S3, S4).</p><!><p>The heterogeneity of the distribution of AC(16:0) can be further assessed by generating the 3D projection of the ion images. In the rostral end of the lesion, AC(16:0) ion distribution initially appears as a localized spot observable at the dorsal column, particularly at the site of descending corticospinal tract/pyramidal tract (CST, Fig. 3). In succeeding sections approaching the lesion center, the distribution is expanded and begins to outline affected tissue, eventually covering the remnants of the disrupted gray matter, until it borders the entire girth of the spinal cord at the lesion center. From then on and going towards the caudal end of the lesion, the outline thickens, with the inner diameter increasing as the site of affected tissue recedes. At a certain point along the caudal end of the lesion, the thickening collapses and the ring contracts and reverts to outlining the site of injury, eventually terminating as a spot observable at the dorsal column as was observed in the rostral part. Thus, following AC(16:0) shows a hollow, turnip-shaped distribution across the segment with position-dependent variation in thickness, outlining the site of lesion (Fig. 3). In SCI lesion at 3 days, the distribution of AC(16:0) covers the entire length of the lesion extending further into R1 and C1. On the other hand, at 7 (Supplementary Fig. S3) and 10 days (Supplementary Fig. S4) the lesion-specific signals are primarily restricted to the lesion segment except at 10 days for a weak signal in C1. It can also be noted that in sections along the caudal part of L, the distribution of AC(16:0) is interrupted in sections where the outline of the gray matter is discernible. In the 7-day SCI sample, the lipid disruption distribution is difficult to deduce because of the discontinuity of the signal observed in some sections taken from L, although it can be observed that the distribution of AC(16:0) appears to be interrupted also in regions outlined by the gray matter (Supplementary Fig. S3). Virtual dissection along the rostro-caudal axis of the composite 3D reconstructed volume of AC(16:0) and lysoPC(16:0) summarizes these findings (Fig. 4). AC(16:0) outlines the lesion, revealing its turnip shape. The outline is well defined at the rostral part of the lesion and becomes thicker along the caudal part after surpassing the lesion center, which approximately covers the entire girth of the spinal cord. Inside this hollow structure can be found lysoPC(16:0). At 3 days after SCI, the heme-containing fragment of the hemoglobin subunit m/z 616.2 can also be observed within the lesion region. This signal can also be observed in SCI at 7 days but in very weak intensity and in fewer sections only, while in SCI at 10 days it is barely detectable (Supplementary Figs S3, S4). The 3D images were then projected in the T7-T9 of the spine of the rat to visualize the lesion and adjacent segments of the spinal cord in the context of their anatomical location (Fig. 5a). Side views of the composite 3D plots of AC(16:0) and lysoPC(16:0) and AC(16:0) and lysoPC(18:1) in the lesion segment show how the turnip-shaped volume outlined by AC(16:0) (red) appears along T8 and encasing both lysoPC(16:0) (blue) and lysoPC(18:1) (green), as revealed by the overlap of m/z 400.3 with the two masses (pink in Fig. 5b,c and yellow in Fig. 5d,e). Front views of the same projections are shown in Fig. 5c and e, where it can be observed that at the lesion center, the extent of overlap almost covers the entire girth of the spinal cord. A video summarizing the main outcomes of this 3D temporal MS imaging study is shown in Supplementary File 1, using the ion distribution of m/z 844.5 to represent the gray matter and that of m/z 822.6 for the white matter. 1 and Supplementary Fig. S5). Other MS n as well as MS/MS assignments using MALDI-TOF/TOF and Q-TOF obtained in this work can be found in Supplementary File 2.</p><p>Long-chain acylcarnitines co-localize with resident microglia/macrophages found at the lesion margin. The presence of LC ACs along the lesion margins only could indicate attempts to sequester the injury. Hines et al. have previously demonstrated that resident microglia are able to perform this sequestration in laser-induced lesions by extending filopodia and covering the lesion entirely 12 . Ren and co-workers have recently demonstrated that these are macrophages derived from resident microglia, and distinguished them from bone marrow-derived ones (which penetrate the lesion interior instead) by their distinct responses to CX3C chemokine receptor 1 (CX3CR1) and beta-galactoside-binding S-type lectin galectin-3 (Mac-2) immunostaining 13,14 . We thus performed immunostaining of CX3CR1 and Mac-2 and co-registered the images with those from MSI. In order to reconstruct the entire image of the spinal cord section, a mosaic of bright field images was acquired using a confocal microscope, and the entire mosaic was reconstructed and co-registered with the ion image of m/z 400.3. Results revealed that CX3CR1 +++ /Mac-2 +/− type staining corresponding to resident microglia is present in the border of the lesion (Fig. 6a and insets). CX3CR1 +/− /Mac-2 ++ type staining corresponding to BMDMs is detected in the core of the lesion. Along the lesion margin, cells exhibiting CX3CR1 +++ /Mac-2 +/− staining co-localize with m/z 400.3 ion distribution (Fig. 6a insets). Reactive microglia exhibit distinct morphology and are hypertrophic with retracted processes along the lesion margins (Fig. 6b and inset). In contrast, microglia detected in regions of the tissue unaffected by the lesion exhibit a ramified morphology with extended processes, indicating that they are in a resting state (Fig. 6c and inset).</p><!><p>In this work we describe using 3D MS imaging techniques the distinct localization to the lesion site of lipids whose abundances are enhanced 3, 7 and 10 days after SCI. Among these are the LC ACs which are intermediates in the acylcarnitine/carnitine shuttle of free fatty acids to the inner mitochondrial membrane during β-oxidation. Accumulation of LC ACs can be a result of mitochondrial damage leading to decreased viability. The LC ACs themselves are pro-inflammatory, and together with medium chain ACs has been demonstrated in insulin resistance and type 2 diabetes mellitus to induce cytokine production and activate cyclooxygenase-2 (COX-2), as well as induce phosphorylation of JNK and ERK via the downstream effector MyD88, and even contribute to ROS production via undeduced mechanisms in RAW 264.7 cells 15 . In addition, the LC ACs have also been shown to exhibit neurotoxicity, and have been implicated in myelinated axon degradation when released by mitochondria-aberrant Schwann cells in the peripheral nervous system (PNS) 16 . Recently, it has also been shown in the context of obesity that LC ACs are capable of activating a reporter gene critical to NF-κB signaling, and indicating that LC ACs could probably perform other functions in relation to chronic inflammation 17,18 . This is further supported by the fact that LC ACs have restricted cellular localization than short chain ACs such as acetyl carnitine, which diffuse more freely in the cell membrane and surprisingly exhibits anti-inflammatory, neurite outgrowth-promoting and neuroprotective effects 19 .</p><p>The specific distribution of LC ACs in different areas of the lesion could hint at a well-partitioned immune response. Like LC ACs, lysoPCs are also pro-inflammatory. However, lysoPCs are derived directly from PCs via phospholipase A2 (PLA2)-catalyzed hydrolysis; the latter can be induced during SCI 20 . Mitochondrial dysfunction can also be induced during SCI; however, the distinct distribution of accumulated LC ACs only along the lesion border suggests that only a specific population of cells, albeit localized in this region, is affected. Further experiments are needed to confirm whether the LC ACs are produced by the co-localized resident microglia/macrophages themselves in response to the injury after their recruitment 12 . Nonetheless, the presence of pro-inflammatory LC ACs in this localized microenvironment is one barrier that can possibly lead to aberrant polarization of resident microglia from the M2 to the M1 phenotype 13 , and consequent abrogation of their anti-inflammatory function.</p><p>The turnip-shaped distribution of LC ACs along the lesion periphery confirms that the immune response along the rostro-caudal axis is heterogeneous 6 . Rostral to the lesion site where the LC AC and lysoPC distribution is distinct, the microenvironment is neurotrophic and the spinal cord continues to receive brain-derived factors promoting recruitment of T regulator lymphocytes to promote neurite outgrowth and axonal repair 6 . Caudal to the lesion site, where the distinct LC AC and lysoPC distribution is disrupted, the microenvironment is predominantly inflammatory and marked by the presence of neutrophils and absence of neurotrophic signals.</p><p>Nevertheless, it also has to be considered that the rostral and caudal segments farthest from the lesion site present the same molecular pattern, and with the ability to promote neurogenesis 5 . Inflammation originates from the lesion and spreads to the two adjacent segments (R1 and C1) with a clear distinct pattern 6 . Within a 12-day SCI time course, this pattern of regionalization can be observed as early as 24 h 3 , and is in line with microglia, macrophage and neutrophil recruitment followed by T regulator cells. In fact, as we recently demonstrated, transcription factors from the Smad family involved in neurite outgrowth promotion only arrive after 24h in DRG cells 3 . Inflammation started to propagate from the lesion site to the caudal segment 1-3 days after SCI, which is in line with ROS and β-oxidation activation.</p><p>Others have previously reported on the use of MSI to examine SCI as well as traumatic brain injuries (TBI), using MALDI, DESI and other ion sources. Setou and co-workers have reported the rapid decrease and eventual disappearance of docosahexanoic acid (DHA)-containing PCs, PC(16:0/22:6) (m/z 844), PC(18:0/22:6) (m/z 872), PC (16:0/20:4) (m/z 820) and PC (18:0/20:4) (m/z 848), and temporary elevation of arachidonic acid-containing PCs, PC (16:0/20:4) (m/z 820) and PC (18:0/20:4) (m/z 848) as well as lysoPC(18:0) by MALDI MSI beginning 1 day post SCI 11,21 . Even though spectral acquisition was performed between m/z 400 and 1,000, analysis focused on the highly abundant signals at m/z 700-900, thus missing the mass range where LC ACs are normally observed 22 . Likewise, Woods and co-workers applied similar MSI analyses on TBI, revealing brain-specific cerebrosides enhanced by injury, but likewise missing the LC AC mass range 23 . On the other hand, Cooks and co-workers applied DESI imaging on contusion SCI models and succeeded in detecting and mapping deprotonated of forms of important lipid peroxidation products such as AA and DHA in negative mode 24 . However, this is the first report on the detection of LC ACs in response to SCI and its distinct distribution along the rostro-caudal axis.</p><p>Although LC ACs have been implicated in various chronic inflammatory diseases and including SCI, interpretation of its accumulation as a result of faulty mitochondrial function remains controversial. Here, we demonstrate that in fact, this accumulation can be considered in terms of its spatial context, highlighting the advantage of the use of MSI for its mapping. Inhibition studies and mapping of LC ACs in other models of CNS injuries are underway, to determine the potential of these lipids as candidate targets for SCI therapy. Experimental SCI procedure. SCI was induced using the modified balloon compression technique in adult male Wistar rats, described previously 25 . Manual bladder expression was performed 2 times a day for 3-10 days after injury. Experimental SCI rats at 3, 7 and 10 days were sacrificed by isoflurane anesthesia followed by decapitation. Pressure expression of each spinal cord was done by injecting sterile saline (10 ml) throughout the vertebral canal along the caudo-rostral axis. Afterwards, each spinal cord was inspected to verify the centering of the lesion at the Th8-Th9 level on the longitudinal axis. The entire spinal cord was divided into 1-cm segments and those that contained the lesion site (L segment Th7-Th11) and the segments rostral (R1 segment, C1-Th6) and caudal (C1 segment, Th12-L6) to the L segment were subjected to MSI (Fig. S1C). The segments were embedded in optimal cutting temperature (OCT) compound (CML, France) and frozen in isopentane cooled with liquid nitrogen then stored at −80 °C until use.</p><!><p>Imaging data acquisition. The entire R1, L and C1 segments were cut into 12-μm sections using a cryostat (Leica Microsystems, Nanterre, France). Sections obtained after every 200 μm (approx.) were subjected to MSI. These were mounted on indium tin oxide (ITO)-coated slides and placed under vacuum in a dessicator for 15 min. DHB was used as matrix, and was prepared at a concentration of 20 mg/mL in 70:30 methanol/0.1% TFA in H 2 O. 12 layers of matrix were deposited using SunCollect plus (SunChrom, Friedrichsdorf, Germany) programmed to spray at gradually increasing flow rates from 10 to 50 µL/min.</p><p>Lipid imaging was performed on an AutoFlex Speed instrument (Bruker Daltonics, Bremen, Germany) equipped with a FlashDetector TM . The instrument was equipped with a Smartbeam ™ -II laser capable of operating up to 2 kHz and was controlled using FlexControl 3.3 (Build 108) software (Bruker Daltonics). The datasets were recorded in positive reflector mode and 500 laser shots were accumulated for each raster point. The laser focus was set to medium, and deflection of masses was deactivated. Spectra were acquired at a lateral resolution of 40 μm. External calibration was performed using the PepMix standard (Bruker Daltonics). Spectra were acquired between m/z 300-1100.</p><p>2D Image processing and data analysis. The MSI datasets were analyzed with SCiLS Lab 3D, version 2016b (SCiLS, Bremen, Germany) 26 . The baseline was removed by iterative convolution and the data were normalized based on the total ion count (TIC) method 27 . Subsequently, orthogonal matching pursuit (OMP) was used to detect peaks; these peaks were aligned to the mean spectrum by centroid matching and automatic spatial segmentation was performed by using the bisecting k-means algorithm 28 .</p><p>A first overview of the spatial regions of white and gray matter was achieved by applying hierarchical clustering using MSI data from all three segments simultaneously. The depth of clustering was selected interactively. Co-localized m/z values with the lesion, white matter and gray matter regions were determined by using Pearson's correlation analysis 29 . Receiver Operating Characteristic (ROC) analysis was performed to detect m/z values which discriminate between the R1/C1 segments versus the L segment, and L segments at different time points (3 days vs 7 days, 7 days vs 10 days and 3 days vs 10 days) 30 . For the first analysis, a random subset of spectra taken from both R1 and C1 was used to allow the comparison of similar groups of spectra.</p><p>3D Image construction, volume visualization and rendering on rat anatomy. For 3D image construction, the MS images were acquired from 12-µm sections taken after cutting about 16 sections from the segment. This resulted in about 36 different sections per segment. Thus, the 3D model was generated out of 36 different sections per segment (R1, L and C1) for all three time points resulting in a model with 324 sections altogether. Registration of different tissue sections was done interactively with rigid 2D alignment. The registered m/z images were passed into the 3D software suite blender for volume visualization. The used rendering technique is ray tracing, i.e., the light is simulated as it passes different media, and thereby various interactions are considered. Materials are modeled by volumetric particles and an intermediate surface which is usually derived from a binary segmentation function. The analyses also include an anatomical model of a rat. The skeleton and the single vertebrate are taken from a processed X-ray CT dataset originally used for 3D printing 31 . For the volumetric m/z image a scalar density function D (x, y, z) of spatial m/z intensities was used. High m/z values give high emission and low m/z values give low or no emission. MS/MS. MS/MS spectra were acquired using off-tissue and direct approaches. The off-tissue approach involves liquid-liquid extraction using the Folch method. A 12-µm-thick spinal cord section was suspended in 60 µL MeOH and to this 30 µL CHCl 3 and 30 µL H 2 O were added. The mixture was vortexed for 1 min, then centrifuged at 10,000 g for 10 min, with the centrifuge temperature maintained at 4 °C. CHCl 3 extracts were collected and 1 µL was spotted on a MALDI target and combined with 1 µL of matrix alone or containing 5 mg/mL LiCl. The matrix used was 20 mg/mL DHB dissolved in 70:30 MeOH/0.1% TFA. MALDI LTQ orbitrap XL (Thermo Fisher Scientifique, Bremen, Germany) and UltraFlex II MALDI TOF/ TOF were used for both full scan and MS/MS spectral acquisition. The MALDI LTQ Orbitrap XL is equipped with a commercial N 2 laser (LTB Lasertechnik, Berlin, Germany) operating at λ = 337 nm with a maximum repetition rate of 60 Hz. The hybrid configuration replaces the heated capillary of the electrospray source with a q00 that sends packets of ions into a linear trap for collision-induced fragmentation, with the fragment ions then being transferred to the orbitrap for high-resolution mass analysis. The maximum energy per pulse was set to 10 µJ. Full scans were acquired at 60,000 resolution (at m/z 400) in the positive mode at the same mass range used during imaging acquisition (m/z 300-1100), by averaging 10 or 20 scans acquired with 2 microscans per step. Acquired gain control (AGC) was turned on to determine the optimal number of laser shots per step. Precursor ion isolation was performed using an isolation window between ±1 and ±4 Da and the fragments scanned with a maximum accumulation time between 120 and 200 ms. External calibration was performed using the ProteoMass MALDI Calibration Kit (Sigma-Aldrich, St. Quentin-Fallavier, France).</p><p>On the other hand, the MALDI TOF/TOF instrument is equipped with a smartbeam laser capable of operating with a maximum repetition rate of 200 Hz. Full scans were obtained by post source decay, in positive reflector mode. MS/MS spectra were obtained using the LIFT cell with the precursor ions generated with 1000 laser shots and accelerated to 2 kV, and the fragments with 4000 shots and accelerated to 18.80 kV. The reflector potential was set at 29.50 kV. External calibration was performed using the PepMix 6 calibrant (LaserBiolabs, Sophia-Antipolis, France).</p><p>The Synapt G2s (Waters, Manchester, UK) coupled to the SPIDERMASS instrument was used for direct analysis of spinal cord tissue sections 32 . 1 μL 5 mg/mL LiCl was spotted on the section and dried prior to analysis. SPIDERMASS uses a tunable wavelength laser system (Radiant version 1.0.1, OPOTEK Inc., Carlsbad, USA) set at 2940 nm and connected to a 10-ns pulse duration Nd:YAG laser pump (Quantel). The laser system is connected to a biocompatible laser fiber (450 µm inner diameter (ID), length of 1 m, HP Photonics IR Fiber) with a 2-cm focusing lens attached at the end. 5 s irradiation at 2 and 4 kJ/m² induces analyte ablation at the tissue surface. The ablated material is aspirated and transported by a Tygon tubing (2.4 mm ID, 4 mm outer diameter (OD), Akron, USA) directly to the inlet of the mass spectrometer by an atmospheric pressure interface 33 . Spectra were acquired in positive mode with a scan time of 500 ms. MS/MS spectra were acquired by CID, with the precursor ion selection window controlled automatically by the instrument. Each spectrum was acquired by averaging 10 scans with each scan acquired for 1 s.</p><p>Immunofluorescence. Immunohistochemical staining was performed on tissue sections which had served for MSI after removal of the MALDI matrix using EtOH washing. After matrix removal, the sections were immersed in blocking buffer (PBS 1x containing 1% bovine serum albumin, 1% ovalbumin, 2% Triton, 1% NDS, and 0.1 M Glycine) for 1 h. The primary antibodies polyclonal rabbit Anti-CX3CR1 (1:30, Santa Cruz Biotechnology, CliniSciences, Nanterre, France) and mouse Anti Mac-2 (1:30, Euromedex, Souffelweyersheim, France) were diluted with the blocking buffer and applied to the sections except for the negative control where only the blocking buffer was applied. The sections were then incubated overnight at 4 °C. The following day, the sections were washed three times with PBS 1×, and incubated for 1 h at 37 °C with the secondary antibodies Alexa fluor donkey anti rabbit, and Alexa fluor donkey anti mouse (Life Technologies, ThermoFisher Scientific, Courtaboeuf, France) both diluted in blocking buffer without 0.1 M glycine. Afterwards, the sections were further washed with several changes of PBS 1×, stained with Sudan black 0.3% for 10 min in order to decrease the background signal generated by lipids, and were eventually counterstained with Hoechst solution (1:10,000). The slides were then washed with PBS 1×, and Dako fluorescent mounting medium was applied on the sections before putting cover slips. Mosaic widefield images composed of 100-150 image scans of 512 × 512 dpi resolution were obtained using a confocal microscope (Leica Biosystems, Nussloch, Germany). Confocal images were also obtained at a confocal depth of 3 μm.</p>

Scientific Reports - Nature

One-step ionic liquid-based ultrasound-assisted dispersive liquid–liquid microextraction coupled with high-performance liquid chromatography for the determination of pyrethroids in traditional Chinese medicine oral liquid preparations

In this study, a simple one-step ionic liquid-based ultrasound-assisted dispersive liquid–liquid microextraction technique was coupled with high-performance liquid chromatography for the analysis of four pyrethroids in three kinds of traditional Chinese medicine oral liquid preparations: simotang oral liquid, kangbingdu oral liquid, and huaji oral liquid. The extraction parameters were examined to improve extraction efficiency. The optimum extraction conditions were 50 μL of 1-octyl-3-methylimidazolium hexafluorophosphate utilized as the extraction solvent and 800 μL of acetonitrile applied as the dispersive solvent. The extraction was assisted by ultrasonication for 8 min. The limits of detection for the four pyrethroids were within 0.007–0.024 mg L−1, and the limits of quantitation ranged between 0.023 and 0.080 mg L−1. The accuracy of the pyrethroid determination ranged from 80.1 to 106.4%. It was indicated that the proposed ionic liquid-based ultrasound-assisted dispersive liquid–liquid microextraction method had an easy operation and was accurate and environmentally friendly. This approach has potential for the analysis of pyrethroids in traditional Chinese medicine oral liquid preparations.

one-step_ionic_liquid-based_ultrasound-assisted_dispersive_liquid–liquid_microextraction_coupled_wit

Introduction<!><!>Apparatus<!>Chromatographic conditions<!>Optimization of IL-DLLME extraction method<!>IL-DLLME procedure<!>Calculations<!>Preparation of spiked samples<!><!>IL volume<!>Type of dispersive solvent<!>Dispersive solvent amount<!>Ultrasonic extraction time<!><!>Conclusions<!>

<p>Traditional Chinese medicine (TCM) is widely employed in the treatment of a variety of diseases, including cough, hyperlipidemia, hypertension and infectious diseases [1, 2]. During the cultivation of Chinese herbal medicine, pesticides are commonly used to control pests and diseases. Currently, synthetic pyrethroid insecticides are more frequently used than traditional organophosphate [3], organonitrogen [4], organochlorine [5], and carbamate pesticides [6] because of their strong insecticidal activity and good stability upon exposure to light and air [7]. However, numerous studies have indicated that these pyrethroid pesticides are toxic to the nervous, reproductive, immune and cardiovascular systems [8]. Oral liquid is one of the most commonly used TCM preparations, and residues of pyrethroid pesticides in TCM oral liquid preparations greatly affect the patients' health and course of therapy. Because TCM preparation contains a great many of herbal components, the interference of complex matrix to the pyrethroid residues and the limitation of current analytical methods will make the residue analysis of pyrethroid pesticides very difficult. Therefore, an accurate measurement method for the pyrethroid pesticides in TCM oral liquid preparations is urgently required.</p><p>Varieties of methods have been exploited for the measurement of pyrethroid residues, and the potential analytical methods include gas chromatography with electron capture detection [9] or mass spectrometry (MS) [10] and high-performance liquid chromatography (HPLC) with ultraviolet (UV) detection [11], diode array detection [12], or MS [13]. MS significantly improved the analysis of pyrethroid residues owing to its high sensitivity; however, it has stricter instrumentation requirements and is not suitable for some typical analytical laboratories. Among these techniques, HPLC–UV has been frequently employed in the analysis of pyrethroid residues [14–16]. However, the analysis of pyrethroid residues in TCM oral liquid preparations is difficult because of the extremely low pyrethroid concentrations and the complexity of the TCM sample. Therefore, pretreatment of the sample before HPLC analysis is crucial for the whole analysis process. Several approaches have been employed for the extraction of pyrethroids from samples, and these methods include the Soxhlet extraction [17], ultrasonic extraction [18], liquid–liquid extraction [19], and solid-phase extraction [20]. However, the extraction approaches have certain limitations, including large organic solvent consumption and a time-consuming extraction procedure.</p><p>Recently, ionic liquids (ILs)—semi-organic molten salts with an organic or inorganic anion and an organic cation—have emerged as alternative extraction solvents for sample treatment because of their advantages of strong thermal stability, good miscibility with organic and aqueous solvents, low vapor pressure, and good solubility for both organic and inorganic compounds. ILs have been utilized for the analysis of several kinds of organic compounds, such as benzoylurea insecticides, neonicotinoid insecticides, and endocrine-disrupting compounds [21–23]. Compared with conventional extraction methods, less organic solvent was consumed during IL dispersive liquid–liquid microextraction, and a higher extraction efficiency was achieved within a shorter extraction time.</p><p>The current study was performed to exploit a one-step ionic liquid dispersive liquid–liquid microextraction (IL-DLLME) for the sensitive measurement of the pyrethroid insecticide in TCM oral liquid preparations. In the current research, ultrasound technology was utilized to cause the ILs to disperse into the aqueous phase as well as to enrich the efficiency. The extraction conditions were examined to improve extraction efficiency. The current approach was then employed in the trace measurement of four pyrethroid insecticides in TCM oral liquid preparations.</p><!><p>Chemical structures of four pyrethroids</p><!><p>A KQ2200DE ultrasonic generator provided by Kunshan Ultrasonic Instruments Co., Ltd. (Jiangsu, China) was operated with an output power and frequency of 100 W and 40 kHz, respectively. An AXTGL16M desktop high-speed refrigerated centrifuge was purchased from Anxin Technologies Inc. (Jiangsu, China).</p><!><p>The determination of pyrethroids was implemented on an HPLC system (Waters Corporation, USA) with a 1525 HPLC pump and a 2489 UV/visible detector. A Diamonsil C18 column (5 μm, 4.6 mm id × 150 mm) from Dikma Technologies Inc. (Beijing, China) was used. Eluent A was water/acetonitrile (95/5, v/v), and eluent B was water/acetonitrile (5/95, v/v). The mobile phases were eluted according to the following program: 12% (A) from 0 to 9.0 min, followed by 12–0% (A) from 9.0 to 35.0 min. The flow rate was 0.6 mL min−1, and the column temperature was 30 °C. The detection was monitored at 210 nm.</p><!><p>The effects of extraction conditions, including the type of IL, IL volume, type of dispersive solvent, dispersive solvent amount, and ultrasonic extraction time, on the recoveries were optimized by single-factor experiments. The experiments were all carried out in triplicate.</p><!><p>The TCM oral liquid preparations were centrifuged at 8000 rpm for 30 min, and the supernatant was filtered by a membrane filter (0.22 μm) before the IL-DLLME procedure. Then, 50 μL of [C8MIM][PF6] and 800 μL of acetonitrile were measured by microsampler and pipette respectively, and added to 5 mL of the filtered sample solutions in a conical tube. The ultrasound-assisted extraction (output power of 100 W and frequency of 40 kHz) was carried out for 8 min. Subsequently, 8 mL of the sample was centrifuged at 6153×g for 5 min. The pyrethroids were extracted into a droplet of IL settled at the bottom of the tube. A syringe was used to remove the upper aqueous phase. The IL phase containing the analytes was diluted with 70 μL of acetonitrile. Ultimately, 10 μL of the resultant solution was delivered into the chromatographic system for analysis.</p><!><p>The enrichment factor (EF), defined as the ratio of the final concentration in the sediment phase (Cfin) to the initial target component concentration in the TCM oral liquid preparation (Cini), was calculated as:\documentclass[12pt]{minimal} sepackage{amsmath} sepackage{wasysym} sepackage{amsfonts} sepackage{amssymb} sepackage{amsbsy} sepackage{mathrsfs} sepackage{upgreek} \setlength{\oddsidemargin}{-69pt} egin{document}$$ { ext{EF}} = rac{{{ ext{C}}_{ ext{fin}} }}{{{ ext{C}}_{ ext{ini}} }} $$\end{document}EF=CfinCini</p><p>The extraction recovery (ER), which was utilized to estimate the pretreatment procedure under various experimental conditions, was calculated as follows:\documentclass[12pt]{minimal} sepackage{amsmath} sepackage{wasysym} sepackage{amsfonts} sepackage{amssymb} sepackage{amsbsy} sepackage{mathrsfs} sepackage{upgreek} \setlength{\oddsidemargin}{-69pt} egin{document}$$ { ext{ER}} = rac{{{ ext{C}}_{ ext{fin}} imes { ext{V}}_{ ext{fin}} }}{{{ ext{C}}_{ ext{ini}} imes { ext{V}}_{ ext{ini}} }} imes 100\% $$\end{document}ER=Cfin×VfinCini×Vini×100%where Vfin is the final target component concentration in the sediment phase and Vini is the initial target component concentration in the TCM oral liquid preparation [23].</p><!><p>Spiked samples were prepared by spiking appropriate amount of the standard solutions in the TCM oral liquid preparations to yield final concentrations of 20, 50 and 100 μg L−1 for four pyrethroids, respectively. Then the samples were subsequently prepared according to the upper IL-DLLME procedure.</p><!><p>Effect of IL type (a), IL volume (b), dispersive solvent type (c), dispersive solvent volume (d) and ultrasonic extraction time (e) on the extraction recovery</p><!><p>When a smaller amount of IL was used, a small amount of precipitation formed, which indicated that the target compound could not be extracted efficiently and that repeatability was poor. In contrast, excess IL may decrease the enrichment factor and sensitivity of the analytical method. The optimum extraction volume of [C8MIM][PF6] was examined by the comparison of three different volumes (40 μL, 50 μL, and 60 μL). It could be observed that higher recoveries were obtained when 50 μL and 60 μL of [C8MIM][PF6] were used (Fig. 2b). Considering the enrichment factor and sensitivity of the analytical method, a lower volume of IL was preferred, and 50 μL of [C8MIM][PF6] was employed during the extraction.</p><!><p>To reduce interfacial tension and increase surface area between the two phases, a proper dispersive solvent with excellent miscibility during the one-step ionic liquid-based ultrasound-assisted dispersive liquid–liquid microextraction (IL-UA-DLLME) process is necessary. Methanol, acetonitrile and acetone were investigated (Fig. 2c). A higher extraction efficiency was achieved when the dispersive solvent was acetonitrile. Thus, acetonitrile was applied for the dispersive solvent.</p><!><p>The effect of the acetonitrile amount was examined by varying the amount from 700 to 900 μL (Fig. 2d). The highest extraction recovery was achieved with 800 μL of acetonitrile. Therefore, 800 μL of acetonitrile was added during the extraction procedure.</p><!><p>The ultrasonic extraction time was investigated from 6 to 9 min (Fig. 2e). When the ultrasonic extraction time was changed from 6 to 8 min, the recovery increased. However, with further extension of the extraction time from 8 to 9 min, the extraction recovery did not greatly vary. As a consequence, 8 min was selected as the ultrasonic extraction time.</p><!><p>Analytical characteristics of the IL-DLLME method combined with HPLC–UV analysis</p><p>Analysis of the TCM oral liquid preparations and spiked recoveries (n = 3)</p><p>Typical chromatograms of four pyrethroids in oral liquids—a simotang oral liquid, b kangbingdu oral liquid, c huaji oral liquid—using optimum conditions: (1) beta-cyfluthrin, (2) fenvalerate, (3) tau-fluvalinate, and (4) bifenthrin. In chromatograms (a, c, d), the spiked levels were 0, 20, 100 μg L−1, and b shows the standard solution</p><p>Comparison of IL-UA-DLLME with other methods for the determination of pyrethroids in liquid samples</p><p>aDispersive liquid–liquid microextraction</p><p>bIonic liquid dispersive liquid–liquid microextraction</p><p>cUltrasound-assisted dispersive liquid–liquid microextraction</p><p>dLiquid–liquid extraction-dispersive solid-phase extraction</p><p>eGas chromatography-flame ionization detector</p><!><p>In the current work, a sensitive analytical approach was investigated for the measurement of four pyrethroids in TCM oral liquid preparations by the utilization of IL-UA-DLLME coupled with HPLC. The extraction parameters were investigated to improve the extraction efficiency, and excellent enrichment performance was achieved. The chromatographic conditions were also tested, and the chromatographic determination was achieved within 35 min. Compared with previous studies, although the sample matrix is more complex on account of the various of herbal component in TCM oral liquid, the proposed method achieved similar LODs with the utilization of less types and lower volume of toxic organic solvents during the microextraction procedure. The results reveal that the method is an accurate, simple, and environmentally friendly method for analyzing the pyrethroids in TCM oral liquid preparations.</p><!><p>1-butyl-3-methylimidazolium hexafluorophosphate</p><p>1-hexyl-3-methylimidazolium hexafluorophosphate</p><p>1-octyl-3-methylimidazolium hexafluorophosphate</p><p>enrichment factor</p><p>extraction recovery</p><p>ionic liquid dispersive liquid–liquid microextraction</p><p>ionic liquids</p><p>ionic liquid-based ultrasound-assisted dispersive liquid–liquid microextraction</p><p>traditional Chinese medicine</p><p>Publisher's Note</p><p>Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.</p>

PubMed Open Access

Computational strategies and challenges for using native ion mobility mass spectrometry in biophysics and structural biology

Native mass spectrometry (MS) allows the interrogation of structural aspects of macromolecules in the gas phase, under the premise of having initially maintained their solution-phase non-covalent interactions intact.In the more than 25 years since the first reports, the utility of native MS has become well established in the structural biology community. The experimental and technological advances during this time have been rapid, resulting in dramatic increases in sensitivity, mass range, resolution, and complexity of possible experiments.As experimental methods are improved, there have been accompanying developments in computational approaches for analysing and exploiting the profusion of MS data in a structural and biophysical context. Here, based on discussions within the EU COST Action BM1403 on Native MS and Related Methods for Structural Biology with broad participation from Europe and North America, we consider the computational strategies currently being employed by the community, aspects of best practice, and the challenges that remain to be addressed.

computational_strategies_and_challenges_for_using_native_ion_mobility_mass_spectrometry_in_biophysic

Introduction<!>Computational considerations in converting native IM-MS data into information<!>Calculating CCSs from structures and models<!>Modelling protein structures using IM-MS data<!>Combining molecular dynamics with native IM-MS<!>Summary and outlook

<p>Native mass spectrometry (MS) involves the transfer of proteins and other macromolecules intact into the gas phase with minimal disruption to the non-covalent interactions that are present in their solvated form. This then allows a range of experiments to probe the macromolecules' higher-order structure, including their fold, assembly and non-covalent interactions [1][2][3][4] . Native MS has helped elucidate various aspects of biomolecular structure, including the subunit composition, stoichiometry and stability of complexes, as well as the dynamic behaviour they display. When combined with ion mobility (IM), where ions are separated based on their mobility through an inert buffer gas kept at constant pressure and temperature under a weak electric field, the size, in the form of a rotationally averaged collision cross section (CCS), of a macromolecule can be probed 5 .</p><p>By virtue of being inherently dispersive, native IM-MS has a unique capability to characterize individual states in heterogeneous and dynamic systems, such as co-populated conformations or assembly states of complexes.</p><p>Thus, native IM-MS has enabled a large number of insights into a diverse array of macromolecular systems, encompassing proteins, nucleic acids, carbohydrates and lipids, and combinations thereof [6][7][8] .</p><p>Proteins and other macromolecules are typically dynamic, in that they populate a range of interconverting structures at equilibrium. Frequently, this heterogeneity is such that macromolecules are better described as structural ensembles (of conformations and/or assemblies), defined by the free-energy landscape accessible at given conditions. IM-MS is sensitive to some of this complexity, providing sparse data that can be a powerful descriptor of molecular states. These data on their own are not sufficient for characterizing molecular structure at atomic detail, but they can, in combination with other information, provide insight into the native state and surrounding free-energy landscape [9][10] .</p><p>Native IM-MS is conducted in the absence of bulk solvent, a factor which may induce some structural changes in the molecules under analysis. Because the gas-phase structures of large biomolecules are dictated by numerous non-covalent interactions -many of which are far from the molecular surface -they hence typically retain the vast majority of their solution-phase character [11][12] . However, the removal of solvent and acquisition of charges alters the physico-chemical environment of the protein, and leads to some degree of restructuring into different conformations, particularly for states that are intrinsically disordered or only marginally stable reflecting -and dependent on -the gas-phase interaction strengths of residues involved 14 .</p><p>The large body of work developing and employing native IM-MS has indicated that a wealth of information is obtainable from such experiments. Yet structural interpretation and translation of the data into structural biology information is often not straightforward. Here we give a perspective on the computational frameworks that must be put in place to address this challenge, and we describe the current thinking and state-of-the-art of the approaches that are being developed. We chart where we believe the field stands in terms of progress in five key computational themes (and their interconnections) namely: 1) IM-MS data extraction and analysis, 2) CCS calculation, 3) determining charge locations, 4) computational modelling, and 5) gas-phase molecular dynamics (MD) (Figure 1). Our thoughts are heavily influenced by the discussions and contributions from the wider native IM-MS community, nucleated through the EU COST Action BM1403, and we refer to our companion article that details and directs the reader to specific software that will aid users in extracting the most, and most reliable, information from their data.</p><!><p>The first step in using IM-MS data is to extract the raw data into a format from which it is possible to determine the key physical properties of the ions under investigation. At the most basic level, this comprises the mass, charge and mobility. All of these properties do not have single values but rather populate distributions, reflecting at least in part the heterogeneity of the system at hand (Table 1).</p><p>While instrument manufacturers' software typically allows the transformation of the measured mass-to-charge (m/z) spectrum onto a mass axis via the assignment of charge states, the frequent complexity of native MS data can make this process difficult. Charge-state assignment can be ambiguous for high charge states, and residual adducts are typical for large macromolecules 15 . Moreover, the samples under analysis themselves frequently contain multiple components, and can sometimes be extremely heterogeneous. Another challenge is that spectral peaks can be poorly resolved, due to the gentle nature of the ionization process employed. To overcome these challenges, both researchers and instrument vendors have developed software and algorithms tailored specifically to native MS data in order to aid users in their analysis (see our companion article for a comprehensive catalogue of available tools).</p><p>While calibration of the m/z axis is straightforward, in order to transform the mobility information (typically acquired in the form of an arrival time distribution, ATD) into a CCS axis, a calibration procedure is typically required 16 . In the overwhelming majority of cases, this is achieved using reference standards appropriate to the target analyte 17 . This process is sensitive to the conditions under which the experiments are performed, and care must be taken to minimise biases associated with the choice of solution and sampling conditions, instrument settings, selection of standards, and the calibration procedures 16 . The information encoded in the CCS (and in CCS distributions, CCSDs) is often used to infer structural properties of a given analyte and can inform computational modelling and (in principle) molecular dynamics (MD) simulations. It also enables direct comparisons of molecular states without additional calibration and computational modelling -as systematic biases cancel when making relative measurements. Nonetheless in all these uses, an important (but underexplored) consideration is the appropriate incorporation of uncertainties associated with the native IM-MS measurement and its transformation into CCS.</p><p>The ATDs and corresponding CCSDs can differ considerably in profile and width, reflecting (after accounting for instrument-dependent resolving power and other effects 18 ) the conformational heterogeneity of the analyte [19][20] . The width of these distributions can be exploited directly, or deconvolved into multiple Gaussian contributions in the case of feature-rich peak shapes 21 . IM-MS experiments can be data-rich, but objective deconvolution of complex ATDs into information of value remains challenging. The difficulty arises in having to decide on the number of conformational families present in the data, and the selection of appropriate width for each Gaussian. Higher resolution IM instrumentation and/or use of tandem IM-MS approaches might enable the separation and resolution of overlapping populations, at least for certain types of samples [22][23][24][25] .</p><!><p>The translation of CCS data obtained during native IM-MS experiments into structural information involves several challenges, including determining how best to obtain the CCS values of the relevant reference structure of the computational model (generation of structures and models is outside the scope of this review). For instance, the user may wish to compare their experimental CCS to available atomic coordinates or to use the CCS to distinguish between various structural hypotheses. A number of approaches exist and selection of the most appropriate method depends on a multitude of factors including the chemical nature of the system under investigation, its shape and intrinsic dynamics, and experimental conditions such as the IM buffer gas 5 . A practical consideration is a trade-off between computational expediency and accuracy in CCS estimations: building a large number of models lets one screen a wider structural space, while performing higher accuracy calculations necessitates screening a smaller range of structures.</p><p>In its most simplistic form, the CCS can be viewed as the rotationally averaged projected area ("shadow") of an object 26 , plus a layer having a thickness related to the gas radius and its polarizability 5 . For any convex object, the projected area is equal to a quarter of its surface area 27 . This simple analytical relationship is useful when considering protein structure at an extremely coarse-grained level 28 . However, when considering protein structure at higher resolution, it is however clear that they are not convex, but feature cavities and protrusions that can lead to multiple collisions or occlude portions of the protein surface from collisions with the buffer gas 29 . On a finer scale, the surface roughness due to the amino acids that decorate the exterior influence the drag a protein experiences during the IM-MS experiment and severs the relation between surface area and projected area. Furthermore, the charge on the protein is inherently non-zero in ion mobility and is expected to impact on CCSs, modulated by the dipole moment and polarizability volume of the gas. The exact distribution of charge can in principle affect the mobility 30 , but appears to have a minor effect on the CCSs of proteins [31][32] . For moderate charge states (i.e. the low amount of charge per unit mass typical in native mass spectra), the CCS appears to be relatively constant in He, but less so in N2 [31][32] . How this phenomenon manifests itself for proteins of all sizes and shapes, and for other types of macromolecules, is currently not known, but neglecting these effects is unlikely to be the major source of bias; more important perhaps are the perturbations the charges make to the structure (see below). Nevertheless, given the increasingly sophisticated questions that IM-MS is being used to answer, and the higher performance IM-MS instruments that have become available [33][34] , considerable scope remains to ensure that local charges and interaction potentials are effectively accommodated in CCS calculations. Different computational approaches (and implementations thereof) for estimating CCSs from structures exist, at differing levels of complexity and computational cost (see our companion paper, and others 5,16 ). The simplest and fastest approach is to consider a protein in terms of its area when projected from different viewpoints. Here the gas atoms are represented by hard spheres that are 'fired' through the sampling volume, and the projected area is calculated from the fraction of trajectories that collide with the protein. A bit more advanced, the exact hard spheres scattering model computes the angle of deflection of the gas to calculate the corresponding deflection (momentum transfer) for the ion. Both approaches ignore electrostatic interactions, and they ignore London dispersion forces acting at long range.</p><p>In the methods at the other end of the complexity spectrum however (several methods are found between these extremes), the short-and long-range interactions of the protein with the gas molecules are modelled explicitly, accounting for both the physico-chemical properties (polarizability, charge, Van der Waals interactions, and potentially internal degrees of freedom) of the gas and of the atoms in the protein, requiring numerical integration of gas-particle trajectories with numerous iterations for each such trajectory. While this more rigorous and explicit consideration of the physical processes underpinning the IM separation might provide more accurate CCSs for atomistic structure models, it does not readily lend itself to coarse-grained structural representations, whereas it is readily achievable to calculate the projected area of e.g. SAXS-derived bead models or iso-surfaces from electron microscopy 20,35 . Consequently, the nature of the structure model can effectively narrow the repertoire of applicable methods for CCS calculation.</p><p>The difference in computational cost between these two extremes currently spans several orders of magnitude, with the most complex approaches taking hours to converge when applied to macromolecules. This renders them intractable for assessing the hundreds of thousands of models needed to explore adequately the rototranslational space associated with structure modelling, or the thousands of frames from MD simulations. As a result, it is often only feasible to use simpler approaches, potentially compromising on the accuracy of the CCS estimation. However, in order to deduce ion shapes from IM-MS, what matters is not so much the accuracy of the absolute calculated values but rather how accurately they can be matched to experiment. For example, for large and globular proteins the simplest projection approximation method can be generally parameterised (i.e. scaled, or calibrated) to reproduce the results from the most computationally costly trajectory method with a relative error within 1% 20 , and experimental drift-tube helium CCS values to within 3% RMSD 36 . In general, appropriate parameterization of the CCS calculation is as important as the underlying physical model that is being used 16 , and one must pay attention to the type and size of system for which a given parameterization was developed, as well as to the type of experiment it was designed to match. For example, no simple parameterization has been thoroughly validated for proteins that are grossly convex, intrinsically disordered, or in extreme charge states. For smaller systems, the relative effect of surface interactions will be proportionally greater than for very large ones. For highly concave structures, a simple projection approach will not take into account "parachute" effects on ion friction. In all these cases, or whenever in doubt, more expensive methods are necessary for good accuracy [37][38] .</p><!><p>Computational methods are needed to exploit native IM-MS data for validating or modelling three-dimensional protein structures. A typical workflow involves distinct steps: converting the experimental data acquired into modelling restraints, building models that sample the conformational space of individual proteins or protein assemblies, and evaluating the models in light of the data. Currently, there are two strategies for building models using MS and other related structural datasets. The first strategy filters models generated by computational methods based on their "goodness-of-fit" to the experimental datasets [39][40][41][42] . The second strategy samples models by directly integrating the experimentally derived restraints with an appropriate scoring function a into the computational workflow -i.e. using the restraint to optimise dynamically the model building [43][44] .</p><p>For modelling analysis, it is important to use appropriate "building blocks". In general, the individual subunits and or complexes can be represented as atomic coordinates (e.g. crystal structures, homology models), as coarse-grained models (e.g. spheroids), or as density maps. Furthermore, it can be important to consider multiple alternative starting structures to ensure that the space is suitably explored 45 . This is pertinent for proteins or complexes that are particularly flexible or are characterised by intrinsically dynamic regions, and where maybe only one particularly stable or abundant structure has been characterized previously e.g. by Xray crystallography. In such cases, developing robust methods for building alternative starting structures for downstream model building becomes a critical aspect of the computational workflow.</p><p>An important aspect of any modelling pipeline is the consideration of the uncertainty introduced at each step of the analysis. First, one must consider ambiguity in the data caused by the limited resolving power of the instruments, the conformational heterogeneity of the protein (which manifests itself as a CCSD broader than the instrumentation limit), and the possibility of low-quality data which can compromise the discriminatory ability of the CCS measurements 46-47 . a A modelling restraint is defined as an assembly/protein feature (e.g. volume, shape, flexibility) quantified with respect to the data used to generate it. It represents the 'force' that glues the individual subunits and forms configurations consistent with the input data. The scoring function sums up all restraints and may be thought as the force field that enables to make up the assembly.</p><p>values if proteins undergo a significant degree of structural change upon transfer to the gas phase, and these discrepancies bring challenges for modelling. Side chains that are solvent-exposed in solution take advantage of the low permittivity of vacuum to collapse onto the surface by forming new interactions [48][49][50] . In the case of protein ions that are intrinsically malleable, e.g. hollow structures, those with hinges, or low charge states of intrinsically disordered proteins, these additional (non-native) non-covalent interactions can lead to unstimulated compaction of the overall protein structure {Rolland, 2019 #179;Hall, 2012 #115;Hansen, 2018 #90;Pacholarz, 2014 #102;Pagel, 2013 #104;Landreh, 2017 #9;van der Spoel, 2011 #56}. Gas-phase induced unfolding happens when the native intramolecular interactions are too weak compared to the repulsion between like charges, and is more likely to occur for high charge states (and at higher activation energies). Gas-phase structural changes require some energy barriers to be overcome, which in turn depends on the native interactions, on the charge state adopted during electrospray, on the internal energy uptake and on the time spent in the mass spectrometer. Despite notable advances made {Konermann, 2017 #178;Marchese, 2012 #67}, gas-phase structural changes remain hard to fully predict, and thus contribute to the uncertainty of the CCS calculation.</p><p>Uncertainty from computations that aim to match experimental data to structural models comprises contributions from the choice of representations [55][56] , the completeness of the information available, the use of the appropriate scoring function, and the biases of individual sampling algorithms (e.g. if they don't accurately capture the data). Finally, measurable errors may be introduced by the post-processing step which typically scores models based on how well they match the input datasets, which may include clustering approaches for generating an ensemble of computational models. A final challenge comes in weighting the merits, and biases, of individual methods based on their ability to contribute to accurate models. As such, the final output of a combined experimental and modelling effort is best represented by an ensemble of structures that encapsulates the convolution of both the inherent conformational heterogeneity of the protein and the various sources of uncertainty in the IM-MS pipeline 42,55 . Benchmarking studies have provided some ways of efficiently integrating the different methods by taking into account the relative uncertainty of the different methods [57][58] , such that it is becoming increasingly possible to bring together the individual techniques in a single workflow</p><!><p>The integration of native IM-MS experiments with molecular dynamics (MD) simulations is highly desirable, as the two methods are complementary with respect to the resolution of structural information they provide, and the timescales that they operate on 9 . In the first instance, solvent-free MD plays an important role in understanding the fundamentals of MS and for interpreting MS data 12,50 . For example, the effects of solvent, temperature and charge on protein structure have been studied in this way, and there are numerous examples of system-specific investigations where MD has been used together with MS 9 . The most widespread MD methods have been developed mainly for condensed-phase calculations, which presents specific challenges when applying them to simulations in vacuum. For example, electrostatic interactions are significant over much longer distances in the absence of solvent which, if taken into account slows down the calculations considerably, thus limiting the sampling and simulation timescales. Moreover, the commonly used force fields are designed to match the solution phase, and hence the effective polarization at the solution interface might not reflect gas-phase conditions. The magnitude of this inaccuracy is currently unquantified, however employing polarizable force fields could be a means to mitigate such errors at an additional computational cost 50 .</p><p>Another challenge stems from considering how charge is distributed on a macromolecule. While the locations of charges do not appear critical for CCS calculations on large molecules, they remain an integral part of the physical model and help determine the system dynamics at the atomic level, thereby greatly influencing the accuracy of the simulations. This, of course, reflects the fact that the location of charges to a large extent 'drives' the structural dynamics, and vice versa. For macromolecules, charging in electrospray takes place via the protonation of basic sites, and deprotonation of acidic sites b -with the note that additional sites become available during electrospray due to their high gas-phase basicity or acidity 60 , that Zwitterionic states are frequently stable in the gas phase [61][62] , and that, depending on solution conditions, charged buffer components can act as charge carriers. Experimentally pinpointing the location of charges is extremely difficult however, b Note that 'basic/acidic sites' is here used according to the Brønsted-Lowry definition, that is, their ability to accept or donate a proton. As such, aspartate and glutamate residues are basic sites, as they are corresponding bases to aspartic acid and glutamic acid, whereas they are typically considered to be acidic residues in biochemistry, regardless of protonation state. and one cannot assume that protonation states simply carry over from solution to the gas phase. Depending on the conditions under which the electrospray process generates charged particles, particularly the presence/absence of protic solvent and the time frame of ionization, the removal of solvent greatly affects the energetics of both the protonated and deprotonated form. However, because of a certain amount of kinetic trapping, the site might still carry some "memory" of its protonation state in solution over the experimental time scales 63 .</p><p>The number of possible charge isomers grows rapidly with the number of (de)protonatable sites, meaning that a complete consideration of isomers is usually not feasible. In lieu of complete enumeration, Monte-Carlo approaches, where protons are moved randomly between basic sites to generate new charge isomers, have been developed to address this issue 51,64 . While the details in how the energies are evaluated and in how the charge isomers are sampled differ between the different approaches, they all compute energy as the sum of the proton affinities for all protonated sites and the electrostatic interactions between charged sites and their surroundings (including other charged sites). The interplay between charge and conformation means that even if the lowestenergy charge isomer can be identified for a crystal structure, relaxation of sidechain conformations, as well as on higher structural levels, might shift the energy considerably 64 . Therefore, care must be taken to not let the rich structural detail in a crystal structure, obtained under considerably different conditions, bias the calculations towards "incorrect" charge isomers.</p><p>Hybrid MD and Monte-Carlo approaches have been developed for the combined search of conformer and charge-isomer space in the gas phase. These have shown that side chains have a propensity to fold onto the protein surface with consequent structure contraction and formation of new charged and neutral hydrogen bonds and salt bridges 62 . These structural rearrangements promote self-solvation and are compatible with maintenance of a native-like fold. An interesting feature in the emerging picture of folded protein ions in the gas phase is the capability to compensate for the energetic penalty of charge separation in vacuo with favourable, conformation-specific intramolecular interactions, in line with growing experimental and theoretical evidence [65][66] . Persistence of zwitterionic states in protein structures provides a rationale for conformational stability in the gas phase and conformational effects on charge-state distributions and is a feature that simulation methods should accommodate.</p><p>In addition to the combinatorial challenges in choosing a "correct" charge isomer, there may be several coexisting charge isomers, and protons could in principle transfer between sites in the gas phase (the "mobile proton model" 67 ), following or promoting structural transitions 68 . As classical MD typically disallows the breakage or cleavage of chemical bonds, protonation dynamics cannot readily be incorporated into such simulations. Recently there has been progress in accommodating proton mobility, with simulations being stopped at regular intervals, and charges being transferred at random towards charge isomers of lower energy 61, 69-71 . Current implementations of this approach are however not truly thermodynamic, in the sense that they do not adhere to Boltzmann statistics, and consequently, they might be error-prone in quantifying how probable the different charge isomers are. Nevertheless, this represents an important step towards accommodating the important role of charges in gas-phase MD, and future integration with popular MD software will be instrumental for the community. Combined quantum mechanics/molecular dynamics (QM/MM) would be a more accurate way to account for proton transfer 72 ; although computationally much more costly than force field MD, it may prove valuable to IM-MS modelling in the future.</p><p>The transition from solution to the gas phase can also incur changes in the structure of the protein. Though these are often small in amplitude 73 , they can significantly alter the contacts made between amino acids 50 .</p><p>This, together with the need to consider electrostatic interactions over long distances, means that MD might struggle to explore experimentally relevant parts of the conformational landscape 50,54 . Experimental data from solution-phase methods are frequently used to restrain the MD simulations, facilitating the transition from the starting structure to the conformations that pertain to the question at hand. In principle, experimentally derived CCSs can be used in a similar fashion, but the considerable overhead required for continuously calculating the CCS during the simulation, and comparing with a given reference value has so far limited the use of CCSbased restraints 9 . Instead, other, more computationally expedient quantities, such as the radius of gyration or solvent accessible surface area (SASA), have been used as proxies for the CCS 38,[74][75] . Recent speed increases in CCS calculations might enable explicit CCS restraints, strengthening the link between simulation and experiments, especially for systems where non-globular structures or conformational transitions might complicate the relationship between proxies and CCSs.</p><!><p>Native IM-MS has the potential to significantly impact structural biology, analogous to the revolution that MS has enabled in proteomics. It is also clear that native MS-derived information benefits from being combined with results obtained from other, orthogonal techniques. These can be other MS-based approaches, such as chemical cross-linking, hydrogen-deuterium exchange, and covalent-labelling (footprinting) approaches, or other structural biology techniques altogether. The resulting "hybrid" strategies enable more accurate and confident structure modelling, particularly in the absence of high-resolution atomistic structures, and extend the validity of these models by sampling heterogeneous conformational and assembly space. However, in order to maximise the potential of native IM-MS, computational strategies that facilitate the translation of the raw data it produces into structural models with associated dynamics, as well as providing a deep understanding of the processes that occur between the protein in solution and its detection in the mass spectrometer, will be instrumental.</p><p>We imagine an era of structural proteomics where macromolecular structures can be computed in a highthroughput manner by exploiting native IM-MS data. Here, we have reviewed the key challenges to achieve this aim (Box 1). The high pace of activity in the field augurs well for these issues being resolved in the nottoo-distant future. These efforts will benefit from the complementary perspectives of the structural MS community, who bring insight into gas-phase effects derived from decades of study on small molecules, and computational structural biologists, who are aware of the priorities and sensitivities in modelling and MD.</p><p>Success in this endeavour will ultimately enable deeper and more quantitative insights from harnessing MS data into understanding the structure, dynamics and interactions of biomolecules, impacting on our understanding of biological (mal)function as well.</p>

ChemRxiv

The Distribution of Aluminum Species in HZSM-5 Catalysts

The structure of aluminum-containing moieties in and within H-ZSM-5 catalysts is a complex function of the elemental composition of the catalyst, synthesis conditions, exposure to moisture, and thermal history. 27Al NMR data collected at field strengths ranging from 7.05 \xe2\x80\x93 35.2 Tesla, i.e., 1H Larmor frequencies from 300 to 1500 MHz, reveals that Al primarily exists as framework or partially-coordinated framework species in commercially-available dehydrated H-ZSM-5 catalysts with Si/Al ranging from 11.5 to 40. Quantitative direct-excitation and sensitivity-enhanced 27Al NMR techniques applied over the wide range of magnetic field strengths used in this study show that prior to significant hydrothermal exposure, detectable amounts of non-framework Al species do not exist. Two-dimensional 27Al multiple-quantum magic-angle spinning (MQMAS) along with 1H-27Al and 29Si-27Al dipolar correlation (D-HMQC) NMR experiments confirm this conclusion, and show that generation of non-framework species following varying severities of hydrothermal exposure are clearly resolved from partially-coordinated framework sites. The impact of hydration on the appearance and interpretation of conventional direct-excitation 27Al spectra, commonly used to assess framework and non-framework Al, is discussed. Aluminum sites in dehydrated catalysts, which are representative of typical operating conditions, are characterized by large quadrupole interactions and are best assigned by obtaining data at multiple field strengths. Based on the results here, an accurate initial assessment of Al sites in high-Al content MFI catalysts prior to any hydrothermal treatment can be used to guide reaction conditions, anticipate potential water impacts, and identify contributions from hydroxyl groups other than those associated with the framework bridging acid site.

the_distribution_of_aluminum_species_in_hzsm-5_catalysts

Introduction<!>Catalyst samples.<!>NMR hardware and sample packing.<!>1D 27Al NMR.<!>2D 27Al-1H correlation.<!>2D MQMAS.<!>2D 27Al-29Si.<!>Results and Discussion<!>Al assignments via 2D NMR on different catalyst preparations.<!>Al and proximate H\xe2\x80\x99s in different catalysts.<!>Presence of Al(III) species.<!>Water exposure, steaming, and Al(IV)-2 vs. EFAls.<!>Al(IV)-2 and catalysis.<!>Potential dynamics of Al coordination.<!>Conclusions

<p>Zeolite catalysts have been successfully implemented in industrial processes for decades, but economic and environmental drivers for lifetime, selectivity, and reactivity improvements continue to motivate structure-property research.1–5 A key aspect to understanding zeolite structure-activity relationships is the location of aluminum atoms in the catalyst framework, the details of aluminum bonding in the framework, proximity of framework Al sites, and the structure and distribution of Al species in non-framework structures.6–11 In addition, proximity between framework and non-framework Al species may also be important to overall reactivity, selectivity, and potential impacts from water.12–16 In typical zeolite analyses, direct excitation 27Al MAS (magic-angle spinning) NMR of catalysts under ambient hydration are used to probe the distribution of Al atoms between framework and non-framework sites, but such studies are limited in that they may only emphasize those Al species whose bonding environment approximates a spherical, e.g. tetrahedral or octahedral, electronic distribution.17 Accurate detection of all Al species is even more complex when the catalyst is dry, due to the large quadrupolar interaction of Al in distorted tetrahedral, pentavalent, and trivalent coordinations, or when Al occurs at framework sites in which it lacks full coordination to other framework Si atoms through oxygen bridges. The latter case, termed partially-coordinated Al,18,19 has recently been detected in H-ZSM-5 based on ultrahigh-field and multidimensional NMR experiments.20</p><p>In order to develop meaningful structure-activity relationships for H-ZSM-5 catalysts, as well as address the current literature debate about water's impact on both reactivity and stability of the various Al moieties and their associated hydroxyl groups,21,22 there exists a need to systematically assess and discuss the accurate detection of all Al species in industrially-relevant H-ZSM-5 catalysts in both hydrated and dehydrated states. Excellent descriptions of the unique requirements for doing quantitative NMR on quadrupolar nuclei already exist in the literature.23–25 Here, the primary goal is to identify how signals from the different Al species in H-ZSM-5 catalysts manifest themselves as a function of catalyst history, exposure to moisture, and the specific NMR experiments used to detect them. Particular focus will be given to the types of Al species that exist when the catalyst is dehydrated, as such conditions should more accurately reflect the high-temperature reactions relevant to hydrocarbon catalysis. A combination of single-pulse, spin-echo, quadrupolar Carr-Purcell-Meiboom-Gill (Q-CPMG), multi-quantum MAS (MQMAS), and dipolar heteronuclear multiple-quantum correlation (D-HMQC) experiments reveal that prior to any hydrothermal treatments, there is no evidence of non-framework Al species. Rather, Al exists primarily as fully coordinated framework Al (denoted Al(IV)-1) and partially coordinated framework Al (denoted Al(IV)-2) in the catalysts with Si/Al ranging from 11.5 to 40 used in this study. In addition, depending on the level of catalyst hydration, the data show that signals that arise from Al(IV)-2 can be mistakenly assigned to Al(V) species. Finally, no evidence for Al(III) species was found in dry H-ZSM-5 catalysts using a variety of NMR experiments, including high-sensitivity Q-CPMG experiments at 19.6 T.</p><!><p>Zeolite ZSM-5 samples with different aluminum contents (Si/Al=40 8014, Si/Al=15 CBV 3024E, Si/Al=11.5 CBV 2314) were obtained from Zeolyst in the ammonium-exchanged form. Dehydrated HZSM-5 zeolite samples were prepared from the ammonium form in a glass reactor body via a stepwise vacuum procedure to a final temperature of 723 K, under 2×10−5 torr using an Edwards EO4K diffusion pump. AHFS-washed catalysts were obtained by washing as-received NH4-ZSM-5 with ammonium hexafluorosilicate (AHFS), with detailed procedures described previously.26 The "mild-steamed" catalyst was prepared by heating at 500 °C for 72 h in a home-built flow reactor, under a 12 mL/min dry N2 flow with 17 torr water vapor (saturated vapor pressure at ambient lab temperature). Catalysts exposed to "severe steaming" conditions followed the same procedure, but at 600 °C. The steamed samples were dehydrated following the same procedure for dehydrating as-received samples as described above.</p><!><p>NMR experiments were recorded on a variety of instruments at 7.05, 9.4, 14.1, 19.6 and 35.2 T (1H Larmor frequencies at 300, 400, 600, 830 and 1500 MHz), with Bruker Avance II, III or NEO consoles. On the 7.05 and 9.4 T instruments, samples were packed in 4 mm rotors with grooved Teflon spacers for further sealing, as described previously,27 with typical spinning speeds of 10–15 kHz. On the 14.1, 19.6 and 35.2 T instruments at NHMFL, samples were packed in Al-free 3.2 mm pencil rotors purchased from Revolution® NMR, with spinning speeds of 16 or 18 kHz when magic-angle spinning (MAS) was needed. Verifying that 27Al signals from rotors or stators do not influence interpretation of chemical species is critical, since many commercial systems present background signals that can lead to erroneous interpretations of broad Al quadrupolar lineshapes. As an additional measure against moisture ingress for those samples that needed long-term sample storage, sulfur powders were used to provide an additional seal to the sample, as illustrated in the scheme in Figure S1. The finite vapor pressure of sulfur was exploited to prevent moisture adsorption into the rotor, as confirmed by the T1 data in Figure S1.</p><!><p>For 1D single-pulse 27Al spectra, 8 and 5μs dead-time was used at 14.1 and 19.6 T, respectively, to avoid ring-down baseline distortions. A 1.6 μs pulse was used for central-transition (CT) 90° and 16° excitation flip angles, respectively, corresponding to 27Al radio-frequency (RF) field strengths of 52 kHz and 10.4 kHz, and consistent with known nutation behavior illustrated in Figure S2. Spin-echo spectra were acquired with 90- and 180-pulse of 1.6 and 3.2 μs, respectively. In all cases, a recycle delay of 0.2 s was used when quantitative measurement was desired.</p><!><p>Robust 17Al- or 1H-excited D-HMQC sequences were used,28 as it requires lower power and shorter mixing time for recoupling and importantly, distance measurements can be obtained from the Al-H buildup curves.25,29,30 For the D-HMQC experiment, recoupling was applied with the SR412 sequence at various mixing time as stated in the text accordingly. For 27Al detected D-HMQC experiments, an adiabatic WURST (wideband, uniform rate, smooth truncation) type of pulse was applied before each transient for a 2-fold signal sensitivity enhancement.31 80 and 50 rotor-synchronized t1 increments were required in 27Al- and 1H- detected experiments, respectively, to avoid truncations in the indirect dimension.</p><!><p>The shifted-echo32 MQMAS sequence was used instead of a z-filter sequence33 as it provides increased sensitivity and better preserves the lineshape for the large CQ (~17MHz) site in the catalysts studied here. MAS spinning at 16 and 18 kHz was used on the 14.1 and 19.6 instruments, respectively, with 12 rotor-synchronized t1 increments and 0.1 s of recycle delay. The MQMAS data were processed with a Q-shear transformation to expand the F1 spectral width.34</p><!><p>An 27Al-excited D-HMQC sequence was used for 27Al-29Si correlation. Following iterative optimization, a recoupling time of 4.4 ms was implemented. A WURST pulse was applied before each transient for signal enhancement, with a total recycle delay of 0.3 s.</p><!><p>Typically, aluminum speciation in H-ZSM-5 catalysts is accomplished via single-pulse 27Al MAS NMR experiments on hydrated samples, using a small flip-angle pulse to quantitatively excite the central transition. Spectra similar to that shown in the top slice of the Figure 1 stack plot are obtained, and the relative amounts of tetrahedrally-coordinated framework Al at bridging-acid sites (BASs) vs. non-framework, octahedrally-coordinated Al (Al(VI)) are assumed from the integrated areas of the ∼55 and 0-ppm peaks, respectively. However, by now it is clear that more than one type of framework Al site can exist, which is not reflected in the simple two-site assignment of either framework or non-framework Al afforded by such spectra.20,35,36 Examining all of the spectra in Figure 1, beginning with the completely dehydrated spectrum in the bottom trace, shows that the signals are complex and may span up to 80 ppm. Based on past work, two types of framework Al are known to exist, i.e., Al(IV)-1 and Al(IV)-2, but such assignments are not possible by direct inspection of any of the spectra in Figure 1.20,35 Moreover, Figure 1 shows that depending upon the state of catalyst hydration, the apparent amount of 0-ppm non-framework signal changes, suggesting that the formation of octahedral Al species are reversible at room temperature or that their coordination changes with water adsorption. It is also important to notice that Al(IV)-2 converges to the 55-ppm region upon hydration due to reduced CQ, but not to 0 ppm. In addition to the questions that Figure 1 generates concerning the potentially transient nature of Al distribution in the catalyst versus hydration, the concomitant changes in the number and type of hydroxyl groups, i.e. acidic vs. non-acidic, are also important to understanding catalyst function. While the bottom spectrum in Figure 1 of the completely dry sample appears to be the least informative due to the dominant quadrupolar broadening, it should represent the fewest number of possible Al species due to the absence of water as discussed in more detail below. The process illustrated by the series of spectra in Figure 1 is reversible, in that subjecting the hydrated sample in the top slice to simple vacuum-dehydration under relatively mild temperatures (e.g., 350–450°C) will generate again the bottom spectrum.</p><p>Figure 2 summarizes 27Al and 1H one-dimensional NMR data acquired on dry H-ZSM-5 catalysts following different sample histories, and using static or MAS data acquisition. The series of catalysts used for Figure 2 include as-received H-ZSM-5 with Si/Al of 15 or 40, as-received followed by AHFS washing under mild conditions that should not significantly impact the number of framework Al species, and as-received followed by the mild-steaming protocol described in the Experimental section. AHFS has been reported as an effective method for removing non-framework Al species in zeolite catalysts,37,38 and under appropriately mild conditions, e.g., low temperature and short exposure times, can be used without damaging the framework.26 To introduce which Al species may exist in a catalyst, and whether Al species that are removed by AHFS under typical conditions are only non-framework species, Figure 2a and b show direct-excitation MAS and static spin-echo 27Al NMR spectra for a dehydrated Si/Al = 15 catalyst, respectively, before (red or upper trace) and after (blue or lower trace) washing with AHFS. The 19.6 T MAS spectra in 2a yield quantitative information on Al species, due to the high field strength and small-flip angle excitation. From 2a, there is a ca. 30% decrease in the total integrated signal area for the AHFS-washed catalyst relative to the initial catalyst. Static whole-echo spin-echo spectra acquired at 14.1 T are shown in Figure 2b, revealing a similar ca. 35–40% intensity loss before and after washing. Figure S2 shows that quantitative conditions and constant lineshapes may be achieved by small-flip angle, 90° excitation, or spin-echo methods. A proposed interpretation of the data in 2a and 2b is that two types of Al species exist, which are further verified in subsequent experiments (vide infra), one of which is characterized by a large quadrupole interaction giving rise to the double-horned powder pattern with singularities near 200 and −200 ppm in 2b, and a second species giving rise to the relatively narrow signal centered near 50 ppm. This second component represents a species with a smaller quadrupolar interaction, and more importantly, is the species that is predominantly removed by AHFS washing. Figure 2a supports these conclusions, with the smaller quadrupolar-interaction species contributing to the narrower peak component centered near 40 ppm. Again, it is this species that is predominantly removed by AHFS washing.</p><p>Analysis of the dehydrated MAS spectra in 2a can support this assignment, in which elements arising from appropriate quadrupolar interaction and asymmetry parameters may be fit to the data. Figure S4 shows the results for appropriate quadrupolar parameters, in which the amounts of Al(IV)-1 and Al(IV)-2 were found to be 69% and 31%, respectively, in agreement with the amounts determined via the AHFS-washing treatment described in the preceding paragraph. Given the expected distributions in chemical shift as well as T2 anisotropy inherent to fitting large quadrupolar linewidths, multiple experimental methods and samples, in addition to simulations, provide optimum results. Therefore, additional experiments on post-synthetically modified catalysts were pursued. Figure 2c shows static 27Al spectra for a second as-received catalyst with Si/Al=40, i.e. a lower Al content than the samples in 2a and b. Again, the initial catalyst (blue or lower trace) appears to contain the same two spectral components as observed in 2a, albeit in a slightly different ratio. Two different field strengths were used for the spectra in 2a and 2c, as noted in the figure caption, which explains the different chemical shift scales. Hydrothermal treatment, or steaming, of a zeolite catalyst is known to lead to framework hydrolysis and creation of non-framework Al species. Of course, varying steaming conditions can lead to different outcomes. Similar to the mild AHFS treatment, in which minimal impacts on framework integrity were targeted, the initial Si/Al=40 catalysts was subjected to the mild hydrothermal treatment described above, leading to the second spectrum in Figure 2c (red or upper trace). As was noted with the data in Figure 2a and b, the first Al component characterized by the large quadrupolar interaction and maximum width singularities appears unchanged after steaming, as shown in 2c. However, the narrow-lineshape component associated with the second Al species increases in magnitude, with a relative ratio of the two Al species now more closely mirroring the ratio for the as-received Si/Al = 15 catalyst in 2a. To clarify, the spectra in Figure 2a–c are plotted to show changes in relative amounts of Al species; the total amount of Al cannot be calculated from these spectra alone due to disparate changes in the species with large and small CQ. As an additional control, static 27Al spectra were acquired at five different field strengths, ranging from 7.05 T to 35.2 T, or from 300 to 1500 MHz 1H Larmor frequency. The results are shown in Figure S5, and are consistent with the field-dependent response expected for two Al species as proposed above.</p><p>Finally, changes in the number and type of Al species should impact the number and type of OH groups present in the catalyst, and Figure 2d shows 1H MAS NMR spectra on the same dry catalysts whose Al NMR data were shown in 2a-c. The two well-known peaks at ca. 1.8–2 ppm and 4.2 ppm are the non-acidic silanol (SiOH) and acidic bridging-acid sites (BAS), respectively. However, additional peaks are present, some of which are observed in the 10–15 ppm range as recently reported.39 These other peaks in Figure 2d appear near 2.8 ppm, from 5–10 ppm, and from 10–15 ppm. Previously, the species giving rise to the 1H signal at ca. 2.8 ppm, and the broad signals in the 5–10 ppm region, have been assigned to a variety of hydrogen-bonded species,40–42 strongly adsorbed water in the framework,43 or hydroxyls on the crystallite surfaces.9 Critically, for both the Si/Al = 15 and 40 catalysts, the appearance and disappearance of these other peaks with either AHFS or mild-steaming treatments, excluding the 2 and 4.2 ppm peaks, responds in a manner identical to that of the more narrow Al signal in 2a-2c.</p><!><p>Figure 3 summarizes results from 27Al 2D MQMAS experiments, which support the assignment of two types of Al as proposed from the data in Figure 2. Figure 3a is the same dehydrated Si/Al=15 sample as previously discussed in 2a, for which two signals are observed off the diagonal in the MQMAS spectra denoted Al(IV)-1 and Al(IV)-2. Along the diagonal in 3a-c, a signal corresponding to a small amount of framework Al(IV) species with a small (near zero) quadrupolar coupling constant CQ is observed, which arises from some residual NH4+ cations associated with tetrahedral framework Al sites and is denoted Al(IV)NH4. The magnitude of this signal is likely artificially enhanced in the MQMAS spectra relative to the other larger CQ (ca. 10–17 MHz) species, and 1H NMR data in Figure 2d shows that the total amount of residual ammonium cation is low. Moreover, as shown in Figure 3d, the signal is easily removed with complete calcination or subsequent thermal treatments, and will not be discussed further. Also, following complete hydration of the tetrahedral framework Al sites, which would give rise to a 1D 27Al NMR spectra similar to that shown in the top trace of Figure 1, the MQMAS intensity for those Al sites would also appear along the diagonal in the same position as the Al(IV)NH4 signal. Indeed, it is possible that trace moisture ingress into the dehydrated catalyst could contribute to this signal in Figures 3a–c.</p><p>Figure 3a and 3b were obtained on the same sample before and after AHFS-washing, directly complimenting the results shown above in 2a and 2b. The signal labeled Al(IV)-1 is the tetrahedral framework Al site known to give rise to the bridging-acid site hydroxyl group, for which previous MQMAS data have been reported.44 Figure 3b shows that this signal is essentially independent of the AHFS treatment, in agreement with Figure 2a and 2b. The Al(IV)-2 signal, which was previously assigned to framework Al sites that were partially coordinated,20 is essentially removed by the washing step as shown by comparing 3a and 3b. Similarly, Figures 3c and 3d illustrate that Al(IV)-2 sites can be created in an H-ZSM-5 catalyst where they did not initially exist. In the as-received Si/Al=11.5 sample, Figure 3c shows that Al(IV)-2 sites are absent, but can be created following the "severe-steaming" treatment described in the Experimental section. Importantly, what is termed severe-steaming here is mild compared to industrial hydrothermal treatments. Figure 3d shows that the Al(IV)-2 species are created by the steaming step, as well as Al(V) "pentacoordinated" species. As a large fraction of Al(IV)-1 is converted to Al(IV)-2 and its MQMAS efficiency is low due to large CQ, the signal for the remaining Al(IV)-1 is not prominent. The signal assigned as Al(IV)-2 exhibits chemical shift and quadrupolar parameters consistent with known values for tetrahedral Al sites in H-ZSM-5 catalysts that are dehydrated. Importantly, the data in Figure 3 shows that prior to formation of non-framework Al(V) or Al(VI) species in a catalyst that only has Al(IV)-1 sites, significant amounts of Al(IV)-2 species can be present. Coupled with the data in Figure 2c and 2d, these data indicate that partially-coordinated framework Al(IV)-2 sites, and their associated hydroxyl groups, can exist in amounts comparable to framework Al(IV)-1 and BAS's prior to the formation of any EFAl species. As a guide to the reader, some representative structures for the Al(IV)-1 and Al(IV)-2 sites are depicted below in Scheme 1, with the blue termini denoting lattice attachment.</p><!><p>Data in Figures 2, 3, and S4 indicated that Al(IV)-2 species can be present in relatively large amounts in as-received H-ZSM-5 catalysts, on the order of 30–40% of the total Al spins, and that their presence correlates with the many of the hydroxyl signals in the high-resolution 1H MAS data shown in Figure 2d. Bonding relationships that define interatomic H and Al atom distances can be investigated by 2D dipolar coupling experiments like those shown in Figure 4. Specifically, dipolar-mediated heteronuclear multiple-quantum coherence (D-HMQC) data for 27Al and 1H spins in the dehydrated Si/Al=15 H-ZSM-5 sample are shown, using either 27Al (4a) or 1H (4b) as the detected spin population. As expected, the latter exhibits better signal-to-noise due to the larger 1H gyromagnetic ratio. Comparing the summed projections of the contours on the 1H and 27Al axes in 4a and 4b indicates that the data for the directly and indirectly-detected case for each spin are comparable. Importantly, the same H-Al correlations are detected in each experiment, although the aluminum lineshapes vary based on differences in the 27Al (shorter) and 1H (longer) T2 values operative in each directly-detected dimension. In 4a and 4b, there exists a 27Al/1H chemical shift correlation at approximately 38/2.8 ppm, indicated by the dashed red lines. Recall, in Figure 2b, the 27Al signal whose maximum intensity appeared at 38 ppm was disproportionally affected by AHFS washing, and Figure 4 shows that this Al signal is correlated to protons with a chemical shift of ca. 2.8 ppm. Similarly, Figure 2d showed that the 2.8 ppm signal, as well as other signals in the 5–10 and 11–15 ppm regions, were also removed or significantly attenuated by AHFS washing. Figure 4c shows results from the same HMQC experiment on Si/Al=11.5 catalyst that was used to obtain the Si/Al=15 data in 4a. Note the complete absence of intensity in the regions corresponding to the 1H and 27Al signals proposed to arise from the Al(IV)-2 species in Figure 4c, as highlighted by the boxed areas, which were filled in 4a. Further, the previous data in Figures 2 and 3 showed that in catalysts where neither the Al nor H signals giving rise to the dipolar correlations shown in Figure 4 were present, they could be created by mild hydrothermal treatment of the catalyst. Thus, Figure 4 further supports the assignment that the Al(IV)-2 signal and these correlated 1H signals originate from the same species. The cross-peak intensities in the HMQC data are a function of the dipolar evolution, or recoupling, time in the pulse sequence as demonstrated by the 27Al slices at the 4.2 ppm 1H chemical shift in 4d. In order to quantify Al-H through-space distances for these and all species detected in the Figure 4 HMQC experiments, variable dipolar recoupling time experiments were acquired and analyzed in detail, as shown in Figure 5.</p><p>Analyzing the time-dependence of two-spin coherence arising from heteronuclear dipolar coupling is a well-known approach for estimating through-space distances of isolated spin pairs,30 in which it is assumed that homonuclear dipolar coupling between spins with large gyromagnetic ratios and concomitant complications from spin-diffusion may be ignored. Previously, several groups have used one-dimensional echo methods, e.g., spin-echo double-resonance (SEDOR),45–47 or rotational echo adiabatic passage double resonance (REAPDOR),48,49 to probe Al-H distances in zeolites. Two-dimensional HMQC data provides additional opportunities to probe site specificities by resolving chemically distinct Al and H species in the second dimension that are not visible in standard 1D data. For example, Figure 5a shows the analysis of four key regions of the HMQC results, as indicated by the rectangular boxes at different 1H chemical shift positions, and their corresponding signal intensities in 5b obtained using 13 different recoupling times. Integrating intensities at each of these positions, and analyzing either the full 27Al intensity or the intensity at the right or left horn of the quadrupolar powder pattern, according to the dipolar coupling equation shown in Figure S6 yields quantitative estimates of the internuclear distance for the Al-H spin pair in that species. As in all spectra of anisotropic powders, T2 anisotropy exists, and since that is a term in the dipolar coupling equation shown in Figure S6, slightly different fitting approaches are required whether one chooses to use the entire powder pattern or to fit the intensity at one of the singularities. There are theoretical drawbacks in each method. Fitting the entire powder pattern requires using an average or global T2 value, which introduces uncertainty due to the fact that T2 is anisotropic. Fitting at each singularity allows the use of a known, fixed T2 value, but the result is impacted by the incomplete integration over the polar angle that defines the relative 1H-27Al dipolar and 27Al quadrupolar interactions. However, the results presented here indicate that both approaches yield similar results, suggesting that these effects have a negligible impact on the measured dipolar coupling. Here, a single averaged-T2 value for the series of spectra like those previously shown in Figure 4d was used, i.e., 0.7 to 0.8 ms depending on the species, based on experimental data and the full spectral fits. This method was compared to fitting the right or left horn, with typical T2's of 0.6 and 0.9 ms, respectively. As shown by comparing Figures S6 and S7, the internuclear distances for all species were comparable using either approach.</p><p>Fits to the Figure 5 data are shown in Figures S6 and S7 along with the calculated internuclear Al-H distances. The Al(IV)-1 results for the well-known BAS, denoted by the box labeled ii in Figure 5, provide a meaningful reference since that internuclear distance has been previously reported, ranging from 0.238–0.255 nm.45,50–52 In comparing results for the full-spectrum/averaged-T2 vs. fitting the horn/single-T2 methods in Figure S7, the HMQC experiments yield an Al-H distance for the BAS at the framework Al(IV)-1 position of 0.242 and 0.255 nm, respectively, which are within the expected range. As an additional control, fitting the full slice intensities for the BAS Al(IV)-1 species in both the dehydrated Si/Al = 15 and Si/Al = 11.5 catalysts yielded Al-H distances of 0.242 and 0.250 nm, respectively. A further 2D SEDOR experiment at 19.6 T, in which the static 27Al echo was measured as a function of evolution time in the presence of 1H dipolar coupling, yielded a dipolar Pake pattern from which the dipolar coupling constant D=1906 Hz could be directly extracted. The results of this experiment are shown in Figure S8, and yields an internuclear 1H-27Al distance of 0.254 nm in agreement with the HMQC dipolar evolution results. More important than the absolute distances, which are subject to potential errors from the assumptions described above, is the comparison between the internuclear distance of the Al(IV)-1 and its BAS proton to the other Al-H species giving rise to signals in the boxed regions labeled i, iii, and iv in 5a. These are the signal regions that Figures 2–4 indicated arise from Al(IV)-2 and its associated hydroxyl groups. From the initial growth of the dipolar evolution curves in Figure 5 and the fits in Figure S6–7, the Al-H distances for the intensity in the i and iii regions of Figure 5 are essentially the same as for the Al(IV)-2/BAS proton at 0.255 nm, while that for the signal in region iv is 0.278 nm. Obviously, the signal-to-noise in region iv of Figure 5a is very low, making the latter value most subject to error as can be seen by examining the individual data points for all buildup curves in Figure S6. The important conclusion from this analysis, however, is that the Al and H spin pairs exist for the proposed Al(IV)-2 species, with interatomic distances similar to that found for the framework BAS's at the Al(IV)-1 sites.</p><!><p>In total, the data in Figures 1–5 and their accompanying SI figures support the recent assignment that partially-coordinated framework Al(IV)-2 species can exist in significant amounts in H-ZSM-5, prior to the formation of any detectable non-framework Al moeities.20 To further confirm that the Al and associated hydroxyl species discussed above do not arise from Al(III) species, sensitivity-enhanced static high-field experiments with extensive signal averaging were acquired to yield Q-CPMG data with excellent signal-to-noise ratios on dehydrated Si/Al=15 catalysts, with several examples illustrating the increased sensitivity shown in Figure S9. For reference, the expected quadrupolar lineshape simulated using CQ = 25 MHz and asymmetry parameters ηQ = 0.1 and 0.6 are superimposed on the experimental data in Figure S9b and S9c; it is known that tri-coordinate Al species are characterized by large quadrupolar coupling constants.53–56 The expected Al(III) quadrupolar linewidth significantly exceeds the experimental one for the Si/Al=15 catalyst. Figure S8c shows the baseline expansion of the experimental spectrum obtained following 128K scans, for which there is no detectable signal in the regions associated with Al(III) species, i.e., 200–300 and −200 to −400 ppm. The static Q-CPMG approach with extensive signal averaging is an attractive approach for addressing this question, since the quadrupolar powder pattern expands toward both low and high frequencies which favors the differentiation of species with varying CQ values, and additionally, the spinning sidebands typically interfere with those signal regions.</p><!><p>Figure 1 previously illustrated the effect of hydration on the 27Al spectra of as-received H-ZSM-5, while Figures 2c–d and 3c–d showed how the 1D and 2D spectra of dehydrated catalysts revealed Al(IV)-2 and some EFAl's could be introduced by steaming. Spectra obtained for the dehydrated Si/Al=11.5 catalysts following the severe-steaming procedure described in the Experimental section, and then exposed to ambient moisture, are shown in Figure 6 as a function of field strength. This is the same sample used to obtain the dehydrated MQMAS data shown previously in Figures 3d, but following ambient moisture exposure. Although the CQ for Al(IV)-2 is smaller in the dehydrated catalysts than the CQ for Al(IV)-1, ca. 11 MHz vs. 17 MHz, respectively, it is larger than that for Al(IV)-1 once the catalyst is hydrated, ca. 5–6 MHz vs. 1–2 MHz.20,36 Recall, the dashed line shown previously in Figure 1 reflects this change in CQ for Al(IV)-2, i.e., from ca. 11 MHz to 5–6 MHz. This creates potential problems for interpretation of standard single-pulse 27Al spectra as illustrated by comparing the three spectra in Figure 6. At 9.4 T, the quadrupolar-broadened Al(IV)-2 lineshape extends into the 20–40 ppm region, which is often assigned to non-framework Al(V) species. As the Al(IV)-2 quadrupolar broadening decreases at higher fields, the presence of some Al(V) species in the steamed sample becomes evident, but the presence of Al(IV)-2 is no longer obvious since the Al(IV)-1 and Al(IV)-2 lineshapes overlap due to the relatively smaller quadrupolar interaction at 19.6 T. Figure 6 demonstrates how standard 1D 27Al MAS NMR data on hydrated zeolites can be misleading, resulting in incorrect or incomplete assignment of important Al species. MQMAS experiments can resolve both species in the hydrated catalysts, even at relatively high field strengths, as shown in Figure S10.</p><p>As additional evidence that simple 1D experiments on hydrated catalysts can be misleading, Figure 7 compares single-pulse Al NMR data with Al species detected via 2D 27Al{29Si} D-HMQC experiments, both acquired at 14.1 T. As a control, Figure 7a compares single-pulse quantitative excitation with a 1D dipolar-coupled spectrum, in which no Al(VI) species at 0 ppm are present in the latter since they are not in the Si-rich framework. This Si/Al=11.5 catalyst has only been subjected to mild-steaming, so no Al(V) species are present.</p><p>Comparing the 2D results for the mild-steamed catalyst in 7b to that of the initial one in 7c shows that the Al(IV)-2 species are created by mild-steaming, as indicated by the signal in the 40–50 ppm regions of the Al dimension in 7b. Further, the relative intensity in the F2 projection in Figure 7b suggests that the Al(IV)-2 species are dipolar-coupled to multiple Si atoms similar to that for Al(IV)-1, while no correlation is observed for any EFAl signals. Figure 7b is particularly useful in making assignments, since the ca. 55-ppm signal in the 1D spectrum in 7a appears symmetrical, and it is difficult to discern contributions to the lineshape from Al(IV)-2 that are clearly evident in the contour plot.</p><!><p>Previous work has shown that H-ZSM-5 can exhibit increased reaction rate constants and conversions for H/D exchange, dehydration, and alkane cracking reactions when all Al species are present in the catalyst, i.e., prior to catalyst treatments designed to extract EFAl species.20,26,37 Moreover, in some cases the addition of trace water amounts leads to increased reaction rates, but again, only in catalysts that have not been exposed to solvent washing traditionally associated with EFAl removal.57,58 Previous reports have suggested that EFAl species are present in some of the same commercial catalyst used in this study, and that Bronsted-Lewis and Bronsted-Bronsted synergies are responsible for enhancements to reaction rates or conversions.59–62 In reviewing this literature, it is probable that partially-coordinated framework Al(IV)-2 and associated OH groups were present in some HZSM-5 catalysts. Moreover, EFAl species should not be present in typical commercial H-ZSM-5 catalysts prior to any hydrothermal treatments based on the results above. Of course, partially-coordinated framework Al(IV)-2 can serve as precursors to hydrothermal evolution of EFAl species under high-temperature reaction conditions, as discussed recently by Bokhoven,63 but examination of past literature in view of our detailed results suggests that Al(IV)-2 and its OH groups can contribute to catalysis through increased rates for probe reactions.20 Further, it appears likely that Al(IV)-2 species can play a role as sites for phosphate tethering to the catalyst in phosphorus-based modifications of H-ZSM-5 catalysts, which are commercially practiced and designed to increase selectivity and hydrothermal stability.64,65 The detailed experiments described here show that AHFS washing can remove Al(IV)-2 species while leaving Al(IV)-1 species intact, that mild steaming can reintroduce Al(IV)-2 without generating significant amounts of EFAls, and that severe steaming leads to both Al(IV)-2 and EFAl species like Al(V) and Al(VI).</p><!><p>Labile tetrahedral framework Al species in the presence of water has been discussed extensively.66–70 Specifically, interconversion between tetrahedral and framework-associated octahedral Al species in the presence of moisture has been reported for zeolites Mordenite and Beta.66–69 Here, all assignments related to Al(IV)-2 in our work were based first on the analysis of dry samples, in which 1H T1 measurements confirmed that water was present at most only in trace amounts. The results presented here are not inconsistent with the findings in MOR and Beta, in that Al(IV)-2 could be involved in such a process, but the 27Al/29Si dipolar correlation results in Figure 7a suggest that the 0-ppm Al(VI) signal is not associated with the framework as might be necessary for a reservoir of framework-associated Al(VI) that easily interconverts with Al(IV)-2 species. It is also possible that the 0-ppm Al(VI) species interconverts with EFAls by hydration/dehydration. Some preliminary MQMAS experiments on a limited sample set suggest that more than one type of Al(VI) species might be present, but additional work is required to confirm such an assignment. However, Al(IV)-2 species could be involved in such processes. Most importantly, the data presented here show that Al(IV)-2 species survive dehydration, and are distinct relative to Al(IV)-1 sites.</p><!><p>The structure of aluminum-containing moieties in and within H-ZSM-5 catalysts as a function of the elemental composition of the catalyst, synthesis conditions, exposure to moisture, and thermal history has been investigated using a variety of 1D and 2D 1H, 27Al, and 29Si NMR experiments at field strengths ranging from 7.05 – 35.2 Tesla, i.e., 1H Larmor frequencies from 300 to 1500 MHz. In addition to traditional framework BAS Al species, i.e. Al(IV)-1, significant amounts of partially-coordinated framework Al, Al(IV)-2, were detected in some catalysts based on their Si/Al and hydrothermal history, and in both dry and hydrated catalysts. Particular focus on understanding 27Al signals in dehydrated catalysts revealed that the Al(IV)-2 species are more clearly identified when moisture is absent, but still exist following hydration. Importantly, Al(IV)-2 was abundant in in multiple catalysts where no extra-frameowork Al species could be detected. Quantitative direct-excitation and sensitivity-enhanced 27Al NMR techniques applied over a wide range of magnetic field strengths showed that prior to significant hydrothermal exposure, detectable amounts of non-framework Al species do not exist. In addition, it was shown that Al(IV)-2 species could be mistakenly assigned as Al(V) in spectra acquired under ambient hydration. Examination of these results in the context of prior literature supports that framework Al(IV)-2 and its OH groups can contribute to catalysis based on increased catalyst/hydrocarbon H/D exchange rates at low temperature, and increased hexane cracking conversion at high temperature. An accurate initial assessment of Al(IV)-1 and Al(IV)-2 sites in MFI catalysts prior to any hydrothermal treatment can be used to guide reaction conditions, anticipate potential water impacts, and identify contributions from hydroxyl groups other than those associated with the framework bridging acid site.</p>

PubMed Author Manuscript

Lanthanide Luminescence Revealing the Phase Composition in Hydrating Cementitious Systems

The hydration process of Portland cement in a cementitious system is crucial for development of the high-quality cementbased construction material. Complementary experiments of Xray diffraction analysis (XRD), scanning electron microscopy (SEM) and time-resolved laser fluorescence spectroscopy (TRLFS) using europium (Eu(III)) as an optical probe are used to analyse the hydration process of two cement systems in the absence and presence of different organic admixtures. We show that different analysed admixtures and the used sulphate carriers in each cement system have a significant influence on the hydration process, namely on the time-dependence in the formation of different hydrate phases of cement. Moreover, the effect of a particular admixture is related to the type of sulphate carrier used. The quantitative information on the amounts of the crystalline cement paste components is accessible via XRD analysis. Distinctly different morphologies of ettringite and calciumÀ silicateÀ hydrates (CÀ SÀ H) determined by SEM allow visual conclusions about formation of these phases at particular ageing times. The TRLFS data provides information about the admixture influence on the course of the silicate reaction. The dip in the dependence of the luminescence decay times on the hydration time indicates the change in the structure of CÀ SÀ H in the early hydration period. Complementary information from XRD, SEM and TRLFS provides detailed information on distinct periods of the cement hydration process.

lanthanide_luminescence_revealing_the_phase_composition_in_hydrating_cementitious_systems

Introduction<!>Sample Preparation<!>Powder X-Ray Diffraction Analysis (XRD)<!>Scanning Electron Microscopy (SEM)<!>XRD<!>SEM<!>TRLFS -Influence of Hydration Time on Luminescence Decay Times τ<!>TRLFS -Influence of Hydration Time on Luminescence Spectra and Relative Fractional Contributions<!>Connecting Spectroscopic Parameters with Location of Eu(III) in CSI and CSII<!>Discussion<!>Conclusions

<p>Chemical admixtures based on organic acids and polycarboxylate ether (PCE) retard the hydration process of Portland cement, thus influencing the properties of the hardened material. The determination of the involved reactions is challenging due to the complex phase composition of the hydrated cementitious matrix. A deeper mechanistic understanding of the admixture influence during the hydration process is needed to avoid undesirable consequences of their application. Further, tailoring the admixture chemistry enables controlling both the properties of the fresh cement paste and the development of the hardened cement matrix.</p><p>The non-hydrated Portland cement consists of four main phases: tricalcium silicate (Ca 3 SiO 5 , C 3 S), dicalcium silicate (Ca 2 SiO 4 , C 2 S), tricalcium aluminate (Ca 3 Al 2 O 6 , C 3 A) and dicalcium aluminateÀ ferrite (Ca 2 (Al x ,Fe (1-x) ) 2 O 5 , C 2 (A,F)). [1] Thus, two main types of reactions occur during the cement hydration: the hydration of calcium silicate phases (silicate reaction) and the hydration of calcium aluminate phases (aluminate reaction). During the latter, aluminate phases C 3 A and C 2 (A,F) react with sulphate ions released form the dissolution of the sulphate carrier (CaSO 4 , its hemi-CaSO 4 • 0.5H 2 O or dihydrate CaSO 4 • 2H 2 O) (Eq. 1). The product of these reactions within first 24 h is the aluminateÀ ferrite tri (AFt) 1 hydrate phase ettringite [Ca 3 Al(OH) 6 ] 2 • (SO 4 ) 3 • 26H 2 O (Eq. 2). During further ageing, ettringite can transform into the aluminateÀ ferrite mono (AFm) 2 hydrate phases. This process can be understood as redistribution of available sulfate ions to dissolved aluminates in the cementitious system.</p><p>The silicate reactions produce calcium silicate hydrate (CÀ SÀ H) phases, which can have the variable stoichiometry and cause the main binding capacity of hardened cement paste (Eq. 2).</p><p>Carboxylic acids are used as retarding admixtures and can be found in the compositions of flow modifiers. [2] Citric and tartaric acids are known to retard the hydration of silicates by slowing down the dissolution of C 3 S. [3,4] Citric acid reduces further the early strength of the hardened cement paste. [5] There are evidences for the influence of citric acid on the hydration of aluminate and formation of ettringite. [6] The origin of these influences is discussed contradictorily. [3,6] G. Möschner et al. explained the retarded dissolution of the clinker grains by the sorption of citrate to the clinker surface and the formation of a protective layer around the clinker grains. [3] Tartaric acid was found to delay the ettringite formation by poisoning the nuclei growth or by formation of a calcium tartrate layer on C 3 A. [4,7] The influence of PCEs, which are used as effective superplasticizers, is described as retardation of the stiffening. The reason for this seems to be the preferential adsorption of these comb-like polymers on the unhydrated aluminate phases, which are the crystallization places for the ettringite formation. [2,8,9] Evidences of the changes in the aluminate reactions are available analysing their products, which are crystalline hydrate phases (AFt, AFm, portlandite (Ca(OH) 2 )). For this, X-ray diffraction analysis (XRD) was used to obtain information on crystalline substances. On the other hand, CÀ SÀ H are poorly crystalline making XRD of limited use. Here, complementary analytical techniques are needed to characterize the influence of admixtures. Recently, europium-doped CÀ SÀ H phases were investigated using Eu(III) luminescence to probe structural properties. [10] Using time-resolved laser fluorescence spectroscopy (TRLFS), structural information on the CÀ SÀ H phases can be obtained regardless of their crystallinity. The identification of the Eu(III) location within the cement matrix is improved by comparison with respective luminescence data of the pure (reference) hydrate phases.</p><p>Here, we present results obtained in the investigation of the admixture influence in the early stage (first 24 h) of the Portland cement hydration as well as after 28 d. For the first time, we applied XRD and TRLFS as complementary techniques to investigate the formation of the different phases (and their time evolution) during the early cement hydration and the effects of different organic admixtures on it. The achieved results were supported by the scanning electron microscopy (SEM). SEM images allowed morphological phase analysis based on the distinct morphological differences between hydrate phases. Each of the analytical techniques has its marked advantages to evaluate certain parts of the hydration process. By combining the complementary techniques, we could show that an improved, more detailed, description can be achieved, in which differences in the formation of CÀ SÀ H and ettringite were resolved depending on the presence and the type of admixture used.</p><!><p>A mixture of 90 % wt. Portland cement (CEM I Milka, HeidelbergCement AG, Germany) and 10 % wt. calcium aluminate cement (Fondu ciment, Kerneos, France) was prepared. Afterwards, 4 % wt. of i) calcium sulfate hemihydrate CaSO 4 • 0.5H 2 O (Schwarze Pumpe, Germany) (CSI) or ii) of calcium sulfate CaSO 4 (BASF, Germany) (CSII) was added to this basic cementitious system (CS). Finally, 3 % wt. of rutile TiO 2 was added to each cement system as internal standard. The resulting absolute component amounts for both cementitious systems are given in Table 1. Series of cement paste samples were prepared from each CS with water-to-cement ratio (w/c) of 0.5. In each sample series, organic admixtures based on carboxylic acids (citric acid, tartaric acid) as well as the complex hydration control agent (HCA) [11] were added with mixing water. The HCA was dry mixed from 2-hydroxy-2-sulfonato acetic acid, ethylene carbonate and polycarboxylate in the ratio of 1 : 1 : 0.25, respectively. The polycarboxylate was a copolymer from acrylic acid and 2acrylamido-2-methylpropansulfonic acid. The concentrations of the admixture solutions were chosen specifically for each CS and are given in Table 1. Additionally, reference samples were prepared without any admixture. For luminescence studies, the cement paste samples were doped with aqueous EuCl 3 solutions. The concentration of EuCl 3 was chosen to achieve an Eu(III) amount of 10 μmol/g CS. The water containing in the EuCl 3 solutions was considered by calculation of w/c. After different hydration times (1, 2, 4, 6, 10, 16, 24 h, and 28 d), the samples were gently dried in a vacuum dryer at room temperature to remove the excess water and thus to stop the hydration process. We avoided applications of further drying methods typically used to stop the cement hydration (e. g. solvent drying, freeze drying), since all of them change the phase composition of hydrated cement paste, although to a different extent. [12] Dried cement samples were ground by hand in the agate mortar and used for analyses. In the meantime, the sample material was stored in tightly closed glass vials.</p><!><p>XRD was carried out to achieve information on the composition of crystalline phase in the cement paste samples. The powder samples were filled into the plate sample holders. The measurements were carried out on a Bragg-Brentano diffractometer (D8 Advance, Bruker AXS, Germany) with a copper X-ray source (Cu Kα ) and a LYNXEYE XE-T detector. The tube was operated at 40 keV and 40 mA. The diffractograms were acquired in the 2θ range of 5-60°with the step size of 0.016°and 1 s per step exposure time.</p><p>The qualitative phase analysis was carried out using DIFFRAC.EVA software (Bruker AXS, Germany) with implemented PDF-2 2016 database. Based on the results of the phase identification, Rietveld refinements were performed using the Topas software (Bruker AXS, Germany) to determine the quantitative phase composition of the crystalline components of the hydrated cement samples. The determined percentages of crystalline phases are related to the total crystalline part of the cement paste, thus excluding the nanocrystalline/amorphous part. The structural data for the Rietveld refinement analyses was taken from the Inorganic Crystal Structure Database (ICSD, FIZ Karlsruhe Leibnitz Institute for Information Infrastructure, Germany). First, the Rietveld analyses of unhydrated CSI and CSII were carried out by including the lattice parameters, Lorentzian crystallite size, and the atomic positions of unhydrated cement phases in the refinement. The results of the phases analyses on the CSI and CSII are summarized in the Table SI1. For the quantification of the hydrated samples, the refined structural parameters of the unhydrated phases were adopted and fixed during the refinement. Following parameters were refined: background, zero error, scale factor, and Lorentzian crystal size. Additionally, lattice parameters and atomic positions of the hydrate phases were refined.</p><!><p>SEM imaging was carried out at a scanning electron microscope (FEI XL30, Philips, USA) using secondary electron detector. The images were captured at the acceleration voltage of 20 kV and the working distance of 6.5 mm. The powders prepared from the dried cement paste were sputtered with gold to improve the surface conductivity. Ettringite and CÀ SÀ H phases could be identified on their characteristic morphology, which is well documented being hexagonal prismatic needles for ettringite and fibrous for CÀ SÀ H phases. [13][14][15][16] Luminescence Spectroscopy Time-resolved laser fluorescence spectroscopy (TRLFS) was used to analyse the luminescence of Eu(III) in 1 g of each cement sample. A pulsed Nd:YAG laser (QuantaRay, Spectra Physics, USA) with a repetition rate of 20 Hz was coupled to an optical parametric oscillator (OPO, GWU Lasertechnik, Germany), serving as light source for the Eu(III) excitation at λ ex = 394 nm ( 5 L 6 ! 7 F 0 transition). Water and the organic compounds have no absorption band at the used wavelength. An influence of photochemical reactions of the organic compounds due to the temperature or laser power were ruled out by observing a stable luminescence signal over time. An iCCD camera (Andor Technology, iStar iCCD-3202, DH-720-18H-13, United Kingdom) was connected to a spectrograph (Oriel Instrument, MS257, USA) equipped with a 1200 lines/nm and 500 nm blazed grating. The Eu(III) luminescence emission was detected in the spectral range of 575 nm < λ em < 635 nm. The time-resolved luminescence kinetics were measured by using the boxcar technique with an initial delay time of t l = 10 μs. A total of 100 accumulations were recorded for each emission spectra. For a full luminescence kinetics, 150 emission spectra collected with increasing the delay times in linear variable steps were acquired (exposure time t exposure = 1.7 ms). The gate pulse widths were chosen specifically for each sample and are shown in Table SI2.</p><p>The luminescence decay times of Eu(III) were calculated from the decrease of the time-dependent integrals of the 5 D 0 ! 7 F 1 and 5 D 0 ! 7 F 2 emission band by using a multi-exponential function (Eq. 5). In the data fitting procedure, the decay constants of both emission bands were shared.</p><p>Here, y(t l ) is the (integral) luminescence intensity of Eu(III) 5 D 0 ! 7 F 1 or 5 D 0 ! 7 F 2 emission band and t l the corresponding delay time after the laser flash. The background signal is given by y 0 and the amplitude of the luminescence decay of the i th component by A i .</p><p>Eq. 6 can be used to determine the fraction f j of the luminescence decay with the amplitude A j for component j.</p><p>The asymmetric ratio R was determined according to Eq. 7 using the integrated luminescence intensity of the 5 D 0 -7 F 2 and 5 D 0 -7 F 1 band.</p><p>Furthermore, time-resolved area-normalized emission spectra (TRANES) were calculated from emission spectra measured at 10 μs, 100 μs, 200 μs, 500 μs, 1000 μs, 2000 μs, 3000 μs and 4000 μs delay time (gate widths are shown in Table SI2). Here, up to 2000 accumulations were aquired for one emission spectrum. This high accumulation number was chosen to significantly improve the signal-to-noise ratio. Differences in the spectral intensity distribution of area-normalized emission spectra after different delay times point towards the existence of different Eu(III) species in the sample. Each sorbed Eu(III) species is correlated with a specific spectral signature. This fact was further used for a factor analysis of the 5 D 0 -7 F 2 emission bands to analyse the relative amount of the existing Eu(III) species.</p><p>In the analysis of the spectra, complementary information on the composition of the cement system was used. In the early state of cement hydration, Eu(III) could be sorbed by portlandite, ettringite (AFt) and CÀ SÀ H. Additionally, Eu(III)-carbonates and Eu(III)-hydroxides can be formed due to a reaction with CO 2 from the atmosphere and the high pH value. Moreover, in CSII Eu(III) could be sorbed by CaSO 4 . Based on this information, a Gaussian fitting for each separate reference system was done first. Afterwards, all obtained Gaussian curves were summed up. The resulting function represented the Eu(III) luminescence spectrum of all species (Eq. 8, for more details see Eq. SI1) and was used for the factor analysis of the measured emission spectra. We further used the specific advantage of a time-gated detection scheme. In the analysis, the emission spectra recorded with a delay time of t l = 1000 μs were used. In that case, Eu(III)-hydroxide species were not observed due to their short luminescence decay times (below 100 μs) and consequently, these Eu(III) species could be excluded (A 1 = 0) to slightly simplify the highly complex system. In Eq. SI1 the second term A 2 (…) corresponded to Eu(III) in portlandite, the third A 3 (…) to Eu(III)-carbonates, the fourth A 4 (…) to Eu(III) in CÀ SÀ H, the fifth A 5 (…) to Eu(III) in ettringite and the sixth A 6 (…) to Eu(III) in CaSO 4 .</p><p>In the case of CSI, no sorption of Eu(III) on CaSO 4 was observed and therefore, A 6 was zero.</p><p>y</p><p>For each reference, the area ratio A k(ratio) of all Gaussian curves was determined. These ratios were fixed in Eq. 8 in order to identify the relative amount of each reference compound in the measured 5 D 0 -7 F 2 emission band of Eu(III) in the analysed cement sample. In Eq. 8, A k(ratio) is a free running area factor and contains information about the amount of the k th reference in the cement sample, y 0 is the offset of the 5 D 0 -7 F 2 emission band, x m -the wavelength of the peak maximum, and A m -the area of the m th Gaussian function. Furthermore, in Eq. 8 the obtained peak maxima of the Gaussian curves in the m th reference were fixed to the wavelength listed in Table SI3. The peak width w m of the m th Gaussian function and the peak width at half maximum FWHM m are connected as described by Eq. 9.</p><p>Eq. 10 was used to calculate the relative amount of each Eu(III) reference f k in the 5 D 0 -7 F 2 emission band of the cement samples at a delay time of t l = 1000 μs.</p><p>Here, A k is the area of the Gaussian peak in the k th reference in the sample, while A k(ref) is representing the area sum of all Gaussian curves of the k th reference Φ k is the quantum yield of Eu(III) in the k th reference.</p><p>A parallel factor analysis (PARAFAC) algorithm, written in the optimization toolbox of Matlab 2018 (MathWorks, USA), was used to extract the emission spectra, decay times and speciation of each Eu(III) species in the CSI and CSII samples. [17] In the hydration time dimension, the luminescence decay time of each component has been constrained to a monoexponential decay. [18] For the calculations, the spectral and speciation dimension was set to nonnegative.</p><!><p>The percentages of the crystalline unhydrated cement phases and their hydrates determined using the Rietveld analysis are plotted over time and shown in Figure 1. The consumption of the unhydrated phases and formation of the hydrates was analysed related to the corresponding hydration reactions. The latter can be divided into: i) the hydration of the calcium silicate phases leading to the formation of CÀ SÀ H and portlandite; the amount of calcium silicates includes both tri-and dicalcium silicate phases (Figure 1, right); ii) the hydration of the calcium aluminate phases; the amount of calcium aluminate phases includes C 3 A and C 2 (A,F); the product of these reactions within first 24 h is found to be ettringite (Eq. 2) (Figure 1, right). During further ageing, ettringite transformed partially into monocarbohydroxyaluminate iii and monocarbohydroxyaluminateÀ ferrite 3 , which were identified in the samples aged for 28 days.</p><p>The amounts of the related phases for the silicate reaction are shown in Figure 1, left, and for the aluminate reaction -in Figure 1, right for each analysed cementitious system. The consumption of sulfate carriers (gypsum in CSI or anhydrous calcium sulfate in CSII) is shown in Figure SI1 in comparison with the amount of ettringite formed.</p><p>In the CSI samples aged for 1 h, ettringite is found to have an amount of 12 % to 14 %. Then, the amounts of ettringite achieved values between 19 % and 22 % after 24 h of hydration and increased slightly to 22 %-25 % after 28 days of ageing. It shows, that most of the ettringite was formed in CSI within the first hour after the hydration begins. The lowest ettringite amount was observed regardless of the ageing time in the reference sample without any admixture. CSII samples showed a similar pattern of the ettringite formation: 11 %-13 % of ettringite were formed within first hour of hydration. In the age of 28 days, the amount of ettringite increased to 20 %-24 %.</p><p>Portlandite is a crystalline product of the silicate reactions. In the CSI-based samples, it was first identified in the reference sample after 4 h ageing. In the samples with admixtures, portlandite was identified: after 2 h ageing -in the sample with HCA, after 6 h of ageing -in the sample with tartaric acid, and as longest after 16 h -in the sample with citric acid. In the age of 28 days, similar amounts of 18 %-20 % of portlandite were identified in the analysed samples. The consumption of calcium silicates is estimated at the total content of C 3 S and C 2 S and differs significantly depending on the admixture added. It is pronounced in the samples with citric (33 %) and tartaric (32 %) acids showing the lowest amounts of remaining calcium silicates in the age of 28 days. For comparison, the reference sample showed 38 % of calcium silicates remaining after 28 days. For CSII, portlandite formation was observed solely in the samples with and without admixtures aged for 28 days, with similar amounts being between 9 %-11 %. The highest amount of portlandite (11 %) is formed in the reference sample and corresponds to 43 % of calcium silicates remaining in this sample. Interestingly, the noticeable consumption of calcium silicates is observed within first 24 h of hydration, where no portlandite crystallisation is identified. The amounts of portlandite formed after 28 days ageing in the samples with admixture added are lower than in the reference sample without admixture. Furthermore, the presence of admixture led to higher consumption of calcium silicates in these samples than in the reference sample. This observation leads to the question, what calcium silicates are consumed for within first 24 h of hydration and in the samples with admixtures -after 28 days of hydration. Since no other crystalline products of the silicate reactions are identified, another question is, in which form the products of the calcium silicate dissolution persist in the cementitious matrices.</p><!><p>The SEM images were taken for the samples aged for 1 h, 4 h, 6 h and 24 h (Figure SI2-3) to analyse the formation of the hydrate phases based on their morphology. The hydrate phases ettringite and CÀ SÀ H are identified in the analysed samples and are exemplarily shown in Figure 2. Their presence depends on the ageing time of the samples and on the addition of admixtures. In general, ettringite is identified after shorter ageing times than CÀ SÀ H giving an estimation of the comparable speed of the aluminate versus silicate reaction. The results of the SEM evaluation are discussed below.</p><p>Ettringite is formed in both analysed cement systems already in the samples aged for 1 h, thus supporting the results of XRD. The number of crystals, their size and morphology differ depending on the kind of cement system and the presence of admixture. In CSI samples, ettringite crystals are well-shaped hexagonal prisms (Figure SI2). In absence of admixtures, numerous short prismatic crystals are formed already after 1 h covering the surface of the unhydrated cement grains (Figure SI2, a). In the samples based on CSII, the formed ettringite crystals are significantly thinner and longer, appearing as a straw-like cover on the cement grain surfaces (Figure SI3, a). If an admixture was added, the formation of larger ettringite crystals than in the reference samples is observed in both cement systems. In CSI, the lower number of ettringite crystals forms bushes of hexagonal prisms in contrast to crystal cover in the reference samples (Figure SI2, b-d). In the CSII, visually less ettringite crystals are formed either, if compared to the related CSII reference samples (Figure SI3, b-d). Additionally, the crystals in the CSII samples with admixture addition are thicker and clearly hexagonal prismatic in contrast to the reference samples. The decrease of the number of ettringite crystals formed in presence of admixtures can indicate the suppression of the nucleation by admixtures. This effect was observed for tartaric acid by Zhang et al.. [4] The formation of the CÀ SÀ H phase is identified on its characteristic tiny fibrous crystals covering the cement grains. In the CSI samples, the CÀ SÀ H forms a honeycomb-like cover on the surface of the unhydrated cement grains. In all samples except those with citric acid, the CÀ SÀ H phase is found after 4 h ageing. The addition of the citric acid delays its formation: The CÀ SÀ H is found in this sample first after 6 h of ageing. In the CSII, the formation of the CÀ SÀ H is week and its morphology is not fibrous. This makes the identification of the CÀ SÀ H in the CSII samples based on SEM difficult.</p><!><p>The luminescence decay kinetics of Eu(III) in CSI and CSII with water, citric acid, tartaric acid as well as HCA in dependence on the hydration time were analysed according to Eq. ( 5) with N = 3, assuming Eu(III) to be located in at least three distinctly different coordination environments. The obtained luminescence decay times τ 2 and τ 3 showed a relatively sharp decrease of the respective value between 4 h and 6 h hydration time in each case of CSI and between 4 h and 10 h in case of CSII (Figure 3, grey area). After this dip, a further upward trend of the luminescence decay times with increasing hydration time was observed. Especially, this dip is pointing to a strong and sudden (on the time scale of tenths of minutes to an hour) change of the Eu(III) coordination environment.</p><p>Apart from the shift of the dip, in principle a similar hydration time evolution of the luminescence behaviour of Eu (III) in both cement systems with water or tartaric acid was observed (Figure 3). In the case of citric acid, the addition of CaSO 4 (CSII) compared to CaSO 4 • 0.5H 2 O (CSI) led to a longer τ 3 for the samples with hydration times < 4 h, but to shorter ones at longer hydration times. In the case of HCA addition, the trend was however the other way around. Here, the addition of CaSO 4 (CSII) led to shorter τ 3 in the first 4 h and longer ones at higher hydration times.</p><p>The short luminescence decay time τ 1 was nearly constant in all cases (CSI and CSII, Figure 3). In the case of CSI with water (Figure 3a, top row) and citric acid (Figure 3b, top row), similar luminescence decay times τ 3 were observed, whereas lower values in the first 6 h due to addition of tartaric acid were found (Figure 3c, top row). After 10 h hydration time, τ 3 of Eu(III) in CSI with tartaric acid reached similar values as with water or citric acid. However, the addition of HCA led to a decrease of τ 3 in the first 24 h of hydration (Figure 3d, top row). After 672 h, τ 3 was similar for CSI with water and tartaric acid as well as for citric acid and HCA. Furthermore, similar values were observed for the luminescence decay times τ 2 of Eu(III) in CSI after addition of water and tartaric acid. However, τ 2 of Eu(III) was shorter in the first 24 h after addition of HCA and longer in the first 6 h if citric acid was added. After 10 h hydration time, τ 2 of Eu(III) in the case of citric acid addition reached similar values as for water or tartaric acid addition. In the case of CSII, the addition of water and HCA led to similar luminescence decay times (τ 3 and τ 2 ) of Eu(III) in dependence to the hydration time (Figure 3a and 3d bottom row). In comparison, τ 3 was longer in the first 6 h and shorter between 10 h and 24 h in the case of citric acid addition (Figure 3b bottom row). After 672 h (28 d) an influence of citric acid was not observed. In the first 6 h τ 3 reached similar values by the addition of tartaric acid as of water, whereas they are shorter between 10 h and 16 h (Figure 3c bottom row). Afterwards, no difference of τ 3 was found. Compared to water, the addition of tartaric acid led to a slightly increased luminescence decay times τ 2 in the first 6 h, whereas no influence was observed afterwards. A similar behaviour was observed for citric acid addition. Here, τ 2 increased significantly in the first 10 h and is similar to values obtained for water afterwards.</p><!><p>In all cases of CSI and CSII, the relative fraction of τ 1 decreased, while the one of τ 2 as well as τ 3 raised with increasing hydration time (Figure SI4). Furthermore, the highest fraction was observed for τ 1 and the lowest for τ 3 .</p><p>Moreover, a change in the spectral signature of the Eu(III) emission spectrum (evaluated from TRANES) as well as a rise of the asymmetric ratio R with increasing hydration time was observed. Similar results were obtained for both analysed cement systems with all admixtures. At least two Eu(III) species can be distinguished due to the differences in the spectral intensity distribution only based on discrimination by increasing delay time used to record the emission spectrum. In Figure 4a-4c the TRANES of Eu(III) in CSI with HCA after different hydration times and the related change of the asymmetric ratio (Fig- ure 4d) are shown as an example. Upon application of a more sophisticated data analysis method (PARAFAC, Figure SI5) the contribution of three different Eu(III) species are resolved complementary to the analysis of the luminescence decay kinetics (vide supra).</p><!><p>The observed spectroscopic parameters (decay kinetics, spectral intensity distribution, fractional intensities) are the result of Eu (III) to be located in different molecular environments. Different sorption sites of Eu(III) on or in the early hydration phases, e. g. portlandite, ettringite and CÀ SÀ H, are possible. Due to the similar coordination environments of Eu(III) on or in these phases, only small variations in the luminescence decay times are expected (Table SI4). However, to be able to specify the different Eu(III) species, the spectral information was analysed in more detail by dissembling the emission spectra with Gaussian curves (Eq. 8). Figure 5a shows the Gaussian fitting of the 5 D 0 - 7 The relative fractions of the single Eu(III) components (Eu(III) in portlandite, ettringite and CÀ SÀ H) in CSI and CSII were calculated using Eq. 10 and are shown in Figure 6. These fractions show on which hydration phase the Eu(III) preferentially interacts. In the time-resolved luminescence experiments the initial gate step was applied as a selection criterion, e. g. sorbed species were detected in favour over hydroxide species (the latter do have a relatively short luminescence time and hardly contribute to the luminescence signal detected at large initial delays). Consequently, the fractions shown are only relative and directly correlated to the gating settings.</p><p>High fractions of Eu(III) in portlandite were observed for CSI for the whole analysed hydration time period (Figure 6a, top). For Eu(III) in CSII with citric acid or tartaric acid similar results were found (Figure 6b, top). In contrast, for CSII with water or HCA an Eu(III) sorption on portlandite was observed earliest after 24 h.</p><p>In the case of CSI with citric acid as well as HCA Eu(III) sorbs to CÀ SÀ H (Figure 6a, middle) starting after 1 h hydration time and was observed over the whole analysed hydration time period as well. However, in the case of CSI with tartaric acid an Eu(III) sorption to CÀ SÀ H was found earliest after 4 h and with water addition even later after 10 h. In the case of CSII an Eu(III) sorption to CÀ SÀ H was observed with tartaric acid addition after 24 h, with citric acid after 10 h. In contrast with water as well as with HCA Eu(III) on CÀ SÀ H was detected after only 1 h hydration time (Figure 6b, middle).</p><p>Eu(III) sorbed to ettringite was found in both cement systems with water in the whole hydration time period, whereas this sorption was retarded for all admixtures (Figure 6a and 6b, bottom). For CSI with tartaric acid or HCA Eu(III) sorbed to ettringite was observed earliest after 4 h and with citric acid after 16 h (Figure 6a below). The addition of HCA to CSII led to a sorption of Eu(III) to ettringite after 6 h (Figure 6b below). A fraction of Eu(III) sorbed to ettringite was observed after 10 h for CSII with citric acid and after 672 h (28 d) with tartaric acid.</p><p>In general, the fractions of Eu(III) sorbed to CÀ SÀ H or ettringite are in both cement systems (CSI and CSII) lower compared to Eu(III) in portlandite (Figure 6).</p><p>The formation of Eu(III)-carbonate and the sorption of Eu(III) to CaSO 4 (only in CSII) are side reactions and do not play a role in the assessment of the additive influence on the hydration process. For reasons of completeness, the fractions of both in the cement samples are illustrated in Figure SI8 and the 5 D 0 -7 F 2 emission band of Eu(III) in CaSO 4 in Figure SI9.</p><!><p>The aim of the work was to analyse the influences of different admixtures on the hydration of cement paste. For this, cement paste samples after different ageing duration were analysed by XRD, SEM, and TRLFS to obtain detailed and complementary information on the phase composition. The latter included the unhydrated cement phases, as well as their hydrates formed during the hydration process over hours up to 28 days. The phases of the unhydrated cement are crystalline and accessible for conventional XRD. The hydrate phases include different crystalline phases also accessible by the XRD, and poorly crystalline CÀ SÀ H mostly hidden for this analytical method. The XRD analysis allows to determine the percentages of the crystalline phases after fixed ageing times and based on these data to compare the role of admixtures on the corresponding hydration reactions. Thus, strong influences of the admixtures on the aluminate reactions were identified by analysing the increase of ettringite percentages and corresponding decrease of the aluminate percentages in different cement paste samples. Controversial dependence of the ettringite percentage from the kind of admixture was found for both analysed cement systems.</p><p>In CSI (containing CaSO 4 • 0.5H 2 O), the presence of admixtures led to an intensive aluminate hydration as the higher amount of ettringite was found to be formed in contrast to the reference sample. In CSII, the formation of ettringite was decreased in presence of admixtures. From all three admixtures used, the citric acid was the one most distinctly changing the amount of ettringite in contrast to the reference sample for both analysed cement systems. In CSI, it led to the formation of the highest amount of ettringite, and in CSII -of the lowest. For the assessment of the silicate reaction, the changes in the amounts of its products portlandite and CÀ SÀ H are needed to be considered. Similar to the aluminate reaction, the silicate reaction was intensified in CSI and decelerated in CSII as estimated on the amount of portlandite. Here, the tartaric acid had the strongest influence on the amount of portlandite formed, leading to the formation of highest amount of portlandite in CSI and lowest in CSII in comparison to all other samples. The strongest increase of the portlandite percentage in CSI was found in the cement paste with tartaric acid and lowest in the reference sample. Controversially for CSII, the highest percentage of portlandite was found in the reference cement paste and the lowest in the cement paste with citric acid. In summary, according to the XRD data, the presence of admixtures has similar influence on the crystalline components of the analysed cement systems: the aluminate and silicate reactions are accelerated in CSI and decelerated in CSII, whereas citric or tartaric acids were the admixtures with the most significant effects in CSI or CSII, respectively. Moreover, the presence of admixtures seems to change the kinetics of the hydration reaction and consequently the amounts of related hydrates formed.</p><p>SEM imaging revealed the morphology of the crystals formed on the surfaces of the non-hydrated cement particles. It was seen that the surface coverage of the initial cement particles with hydration products was growing with increasing hydration time. Morphological changes of ettringite crystals were observed depending on the cement system (CSI vs. CSII) and on the presence of admixtures. In CSI, the ettringite crystals were well-shaped hexagonal prisms, whereas in CSII they appeared thin and forming a straw-like cover on the cement particles. In presence of admixtures, the visibly lower number of larger ettringite crystals was formed. Obviously, a lower number of seeds can be formed, if an admixture is added to the cement system. This makes it attractive to assume, that admixtures seem to distort the ettringite nucleation. In our previous work, we could show the occupation of the ettringite nucleation places by the admixtures carrying the negative charges (like negatively charged carboxylic groups available in each admixture analysed in this study). [9] The nucleation of ettringite is possibly shifted from the cement particle surface into the surrounding solution, thus explaining the rise of the ettringite amount in presence of such admixtures as discussed above. Additionally, a lower number of larger crystals could be visually identified on the SEM images. Obviously, the aluminate reaction is balanced between nucleation on the cement surface and in the surrounding solution. Consequently, the amount of ettringite formed changes depending on how favourable the conditions are for its crystallization. According to SEM imaging, the formation of CÀ SÀ H was more pronounced in the cement samples based on CSI. The presence of any admixture in each cement system led to the formation of the CÀ SÀ H cover, that was visually less dense in comparison to the corresponding admixture-free reference sample.</p><p>Eu(III) serve as a luminescence sensor due to its specific luminescence response on the coordination environment. Especially, poorly crystalline hydrate phases (such as CÀ SÀ H) with a high Eu(III) sorption affinity become accessible. In a control experiment ICP-MS analysis of Eu(III) sorbed on CÀ SÀ H (C/S = 1.4, pH = 12.6, 10 μmol Eu(III)/g CÀ SÀ H) showed an exchange of every 1670 Ca(II) ion against one Eu(III) ion. This underlines that only an influence of the Eu(III)-sorption to the hydration process of the CSs is not expected. Therefore, additional information on the formation of hydrate phases can be obtained complementary to XRD and SEM data.</p><p>By repeating the TRLFS measurements at different sites of the grounded cement sample and after a certain time similar data were observed and leads to representative results. Three different Eu(III) species were identified based on the temporal and spectral analysis of the luminescence data. The formation of Eu(III)-hydroxide is indicated by the calculated short decay times τ 1 (Figure 3), the spectral signature of the 5 D 0 -7 F 2 emission band (at a delay time of t l = 10 μs) (Figure 4) and was observed in absence of influences of admixture, type of the sulphate carrier (CaSO 4 or CaSO 4 • 0.5H 2 O) or hydration time. This is in good agreement with the observations by Burek et al.. [10] The luminescence of the other two Eu(III) species is characterized by their long luminescence decay times (τ 2 and τ 3 ) indicating a significantly decreased quenching rate due to a reduced number of OH-oscillators in the coordination sphere. Therefore, this is pointing to surface sorption on or incorporation of Eu(III) into the hydrate phases that are formed over time. A similar sorption process of Eu(III) on CÀ SÀ H was observed before. [19][20][21][22] A shorter luminescence decay time τ 3 of Eu(III) was observed in CSI with tartaric acid (in the first six hours) and with HCA (in the first 24 hours) compared to CSI reference or with citric acid. This effect points to more OH-vibrations in the first coordination sphere of Eu(III) and, consequently, to less incorporation of Eu (III) into the hydrate phases, thus indicating a variation of the hydrate phase composition (Figure 3).</p><p>Studies of the cement hydration process describe a beginning CÀ SÀ H formation between 1 h and 6 h. [23][24][25] This corresponds to the time range of the observed dip of hydration-related time dependence of the luminescence decay times that was found for the early hydration times (s. Figure 3). As discussed above, this dip indicates a strong short-term change of the Eu(III) coordination environment and could result from a changed hydrate phase composition. Except for CSI with water and CSII with citric acid, the factor analysis (Figure 6a and 6b, middle) suggests the existence of CÀ SÀ H in the hydration time range between 1 h and 6 h for all the other analysed systems. In good agreement with the dip is only the beginning Eu(III) sorption on CÀ SÀ H in the case of CSI with tartaric acid. The SEM analysis, however, shows a visible CÀ SÀ H formation in CSI not before 4 h hydration time. It is described in the literature, that CÀ SÀ H phases differ in their structure and density at early hydration stages. At the hydration begin, an intermediate CÀ SÀ H phase is observed, which differs structurally from the more mature CÀ SÀ H. Thus, the observed dip in the hydration-related time dependence of the luminescence decay times and its controversy to SEM observations can be related to the transformation of the intermediate CÀ SÀ H to the mature product, thus explaining the abrupt change in the nearest coordination environment of Eu(III). [26][27][28] The change of the luminescence decay times with prolonged ageing indicates a progressive sorption of Eu(III) on the surface of as well as an incorporation into the hydrate phases and a decrease in the amount of Eu(III)-hydroxide (Figure SI4). Furthermore, the amount of Eu(III) surface complexation is higher than that of Eu(III) incorporation. The change of the spectral signature of the Eu(III) emission in the TRANES with increasing hydration times indicates a change of the Eu(III) species. Moreover, it is a consequence of a progressive sorption and of a change of the hydrate phase composition (Figure 4).</p><p>Similar sorption sites can be further distinguished by the detailed analysis of the information hidden in luminescence spectra by fitting of Gaussian curves. This analysis shows for both cement systems only a small amount of Eu(III) localized in CÀ SÀ H or ettringite, but a higher one in portlandite (Figure 6). This indicates a strong sorption affinity of Eu(III) to portlandite considering the fact of a higher amount of ettringite in the system according to the XRD data (Figure 1). Moreover, the addition of admixtures to both cement systems retards the ettringite formation (Figure 6). In the case of CSI, tartaric acid and HCA have similar effects, and citric acid -the strongest one, whereas in the case of CSII tartaric acid has the strongest influence. In contrast, for CSI a similar retarded effect of both acids and the strongest by the addition of HCA was observed by XRD. In good agreement are the SEM results, which indicate a faster formation of ettringite for CSI and CSII with water (reference system) compared to the samples with admixture presence due to the formation of numerous smaller crystals. A retarded ettringite formation by the addition of citric and tartaric acid was already observed before. [3,7] Furthermore, the TRLFS data show an increasing amount of ettringite with increasing hydration time for CSI and are in good agreement with the observation by XRD. However, SEM and XRD are pointing to an early formation of ettringite after 1 h hydration time for both cement systems, whereas this was not observed in all cases by TRLFS. This could be caused by a preferred sorption of Eu(III) to portlandite (vide supra) and to CaSO 4 (in CSII), as well as by the formation of Eu(III)-carbonate, which all compete with the sorption to ettringite. Therefore, a strong incorporation of Eu(III) into ettringite and the specific influence on the spectral signature is weak after 1 h hydration time.</p><p>The TRLFS data showed an accelerated CÀ SÀ H formation in CSI due to the addition of admixtures. Here, citric acid and HCA had the strongest influence. For CSII, the addition of citric or tartaric acid, however, led to a retarded CÀ SÀ H formation, whereas citric acid had the most pronounced effect. Furthermore, a retarded CÀ SÀ H formation in CSII explains the later dip of the luminescence decay times τ 2 and τ 3 as compared to the one observed for CSI (Figure 3).</p><p>TRLFS indicates portlandite formation in both cement systems already after shorter hydration times if compared to XRD. The reason could be the high sorption affinity of Eu(III) to this hydration phase. For CSI, the sorption of Eu(III) to portlandite was observed for the whole hydration time range investigated and no influence of the admixtures was found. In CSII, water (reference system) and HCA retarded the portlandite formation significantly.</p><p>It is tempting to attribute the contradictory influence of admixtures on the CÀ SÀ H as well as portlandite formation in CSI and CSII to the different types of sulphate carrier used in the preparation. The anhydrous CaSO 4 in CSII has a lower solubility than CaSO 4 • 0.5H 2 O in CSI. Consequently, less Ca(II) (important for CÀ SÀ H and portlandite formation) and SO 4 -ions (for ettringite formation) are freely available in CSI. The discussed above distinct differences in the morphology of the ettringite crystals in both cement systems confirm this effect.</p><p>The complementary use of XRD, SEM and TRLFS allowed deep insights into the admixture action in cementitious systems differing by the kind of the sulphate carrier (CSI with CaSO 4 • 0.5H 2 O; CSII with CaSO 4 ). The changes in the amounts of the specific crystalline reaction products are available from the XRD data. According to them, the most significant intensification of the aluminate reaction was observed in CSI, if citric acid was added. The silicate reaction was the most intensive after the addition of tartaric acid. Reverse effects were observed with these admixtures in CSII. The addition of citric acid slowed down the aluminate reaction in CSII; the addition of tartaric acid most significantly delayed the silicate reaction. However, it should be noted, that the extent of the silicate reaction could be estimated on the amount of portlandite. Another important product of this reaction, which is the main binding phase in hardened cement systems, the CÀ SÀ H phase, was inaccessible via XRD due to its poor crystallinity. The use of SEM allowed visual assessments of the admixture effects based on the appearance of the hydration products. Significant differences were revisited between CSI and CSII reference samples, as well as those with particular admixture if correspondingly compared. The use of SEM allowed first insights into the formation of CÀ SÀ H, since this phase formed a characteristic cover on the unhydrated cement particles. Further information on CÀ SÀ H was however still inaccessible. The application of TRLFS allowed detailed insights into the phase composition of the sample based on the specific sensitivity of Eu(III) used as a sensor to its chemical surrounding. The identification of the phases contributed to the luminescence reply of cement systems was achieved by comparison with the spectra of corresponding pure compounds and following deep mathematical analysis of the spectral data of the analysed samples by considering the conclusions from the XRD data. The extent of the Eu(III) sorption affinity to particular phases was needed to be considered during the data analysis. TRLFS results confirmed the con-clusions from the XRD data about the reverse effects of admixtures on silicate and aluminate reactions in both analysed cement systems. Additionally, the formation of portlandite, which indicates the ongoing silicate reaction, was identified earlier than by XRD due to high sensitivity of the Eu(III) used as luminescence sensor. Furthermore, the luminescence data allowed to identify the formation of CÀ SÀ H, as well as the change in its structure between 4 h and 6 h of ageing. Thus, the detailed comparison of the information obtained from XRD, SEM and TRLFS enabled comprehensive insights into the phase formation during the hydration taking place in the complex cementitious systems influenced by chemical admixtures.</p><!><p>We have presented complementary data of different analytical methods (XRD, SEM, TRLFS) to describe the effect of organic admixtures on the hydration process of two cement systems. Different observation scales of the methods applied allow to reach an extensive complementary information about the changes in the cement paste composition due to the influences of the chemical admixtures. XRD allows to monitor the crystalline amounts of the cement paste components and SEM -to analyse the formation of hydrate phases based on their morphology. The results show, if admixtures were used, a changed hydration reaction kinetics, whereas the hydration products remain the same and differ exclusively in their amounts considering the specific ageing time. With the use of Eu(III) as an optical probe in TRLFS analysis, it was possible to add additional information on the hydration phase dynamics due to the complementary selectivity of the Eu(III) sorption affinity to the hydration phases. Furthermore, the TRLFS also allowed to monitor minor phase (portlandite), which has a high sorption affinity for Eu(III). It was for the first time possible to get insights into the influences of admixtures on the CÀ SÀ H formation using luminescence spectroscopy. The dip in the luminescence decay times indicates possibly the transition of the low-density CÀ SÀ H formed at the hydration begin into high-density CÀ SÀ H characteristic for mature cement paste. Further work is needed to directly evidence this dip to specific structural changes of CÀ SÀ H during the hydration process.</p><p>In the future, the combination of SEM, XRD and TRLFS might become a key tool for the cement industry. Especially, online and inline luminescence measurements in combination with fiber optics may pave the road to novel cement materials based on an improved understanding of the cement hydration and how this process can be directed. Here, the outstanding sensitivity and the non-invasive character of optical methods will help to further decipher molecular processes during the hydration of cement. Luminescence spectra can be measured in a few seconds and it is foreseen to also test additional luminescence probes (such as other lanthanide, but also organic dyes maybe of interest) in order to establish a mulitparameter sensing for a further improved understanding of cement hydration kinetics. Here, it can be expected that the achievable time-resolution is determined by mixing times of the cement components, which probably is on the time-scale of minutes.</p>

Chemistry Open

Enhancing nicotine vaccine immunogenicity with liposomes

A major liability of existing nicotine vaccine candidates is the wide variation in anti-nicotine immune responses among clinical trial participants. In order to address this liability, significant emphasis has been directed at evaluating adjuvants and delivery systems that confer more robust potentiation of the anti-nicotine immune response. Toward that end, we have initiated work that seeks to exploit the adjuvant effect of liposomes, with or without Toll-like receptor agonist(s). The results of the murine immunization study described herein support the hypothesis that a liposomal nicotine vaccine formulation may provide a means for addressing the immunogenicity challenge.

enhancing_nicotine_vaccine_immunogenicity_with_liposomes

<p>Chronic tobacco use, reinforced by nicotine dependence, is a widespread problem with costly impact. Current therapies to promote smoking cessation and prevent relapse include psychosocial interventions, antidepressants, and partial agonists. Despite these treatment options, relapse rates are still unacceptably high.</p><p>An alternative therapeutic strategy that has recently received intense attention is known as immunopharmacotherapy.1,2 This involves eliciting antibodies that are highly specific for a particular drug of abuse. Antibodies bind the drug in the periphery, and the resulting antibody-drug complexes are unable to cross the blood-brain barrier (BBB). As a result, brain exposure to the drug is reduced. The extent of this exposure reduction is related to both the quality and quantity of anti-drug antibodies elicited by the vaccine. We have carried out a wide variety of immunopharmacotherapeutic campaigns against several drugs of abuse, including cocaine,3 heroin,4 methamphetamine,5 and nicotine.6</p><p>In the case of nicotine, various candidate vaccine formulations have been or are currently being evaluated in human clinical trials.7 The hitherto lackluster performance of nicotine vaccines in the clinic underscores the notion that a sufficiently high concentration of nicotine-specific antibodies must be generated in order to provide broad efficacy in studies aimed at promoting cessation and preventing relapse.8 We continue to investigate tactics for enhancing nicotine vaccine immunogenicity.</p><p>We have previously scrutinized other parameters, such as linker position and length,9 hapten conformational constraint,10 and variation of protein carrier.6,11,12 Recently, a distinctly different tactic was employed for the generation of nicotine-specific antibodies: immunization of mice with an adeno-associated virus (AAV) vector carrying the genetic information for expressing the Fab fragment of NIC9D9, an anti-nicotine monoclonal antibody.13,14 While highly promising, the existing regulatory and developmental hurdles of such a gene therapy approach mean that further work within the more traditional active vaccination manifold merits pursuit. In this latter context, the vital role of adjuvant(s) for promoting robust nicotine-specific antibody production remains to be fully explored. The purpose of the present study was to evaluate the impact of liposomal delivery of a typical hapten-protein conjugate (i.e. AM1-KLH) on the resulting anti-nicotine immune response.</p><p>A means by which any weakly immunogenic vaccine can be rendered more immunogenic is through the use of adjuvants.15 Since the 1970s, it has been known that liposomal presentation of antigens can confer greater immunogenicity compared to antigen alone.16,17 Liposomal vaccines carry many advantages, including recruitment of various components of the immune system,18 opportunity for dose-sparing of antigen, and "plasticity" (tremendous flexibility) with regard to lipid composition.19 In addition to serving as an adjuvant by virtue of its direct stimulation of the immune response (e.g. macrophages and dendritic cells), liposomes also provide a physical means for either delivering encapsulated antigens or presenting surface-associated antigens with the ability to modulate epitope density and homogeneity. Also, orthogonal adjuvants, such as Toll-like receptor (TLR) agonists, may be encapsulated within the vesicles, or incorporated into the lipid bilayers. In other words, the liposome platform is one that boasts a great diversity of manipulations.</p><p>Generally speaking, a wide variety of surface functionalized liposomes have been studied for coupling with immunoglobulins, proteins, and peptides.20 Molecular Express, Inc. have developed VesiVax® Conjugatable Adjuvant Lipid Vesicles (CALVs) that offer a practical means for liposomal delivery of antigens of interest, by virtue of direct membrane anchorage through reactive functional groups. For example, the VesiVax® TLR4 conjugatable liposome contains a TLR4-specific agonist, monophosphoryl lipid A, and maleimide groups on the outer surface that allow covalent conjugation with sulfhydryl-containing molecules (e.g., cysteine groups on peptides or proteins). The VesiVax® TLR4 platform allows the user to covalently conjugate the desired target antigen for an easy-to-use antigen delivery method.</p><p>As a matter of course, we first evaluate new formulations for their ability to generate antibodies in mice, using results obtained from murine immunization studies to inform subsequent (e.g. pharmacokinetic and behavioral) studies in rats. There are several reports of active immunization against nicotine in mice.9,10,21,22 This is justifiable based on the relatively low cost output (compared to rats) and rapid immune response (1–2 months) exhibited by mice used in these studies.</p><p>This study began with the preparation of immunogen via hapten-protein conjugation. AM1 nicotine hapten6,23 was activated using EDC and sulfo-NHS in DMF/H2O (10:1 v/v) at room temperature for 4–6 h. After DMF removal under reduced pressure, activated AM1 was mixed with either bovine serum albumin (BSA) or keyhole limpet hemocyanin (KLH) in 0.1 M MOPS saline pH 7.2 at 4 °C for 18–20 h. The resulting hapten-protein conjugates, AM1-BSA and AM1-KLH, were purified by dialysis using PBS pH 7.2 at 4 °C. Protein concentrations were determined by BCA assay,24 and copy number (23–27) for AM1-BSA was determined by MALDI-TOF mass spectrometry. AM1-KLH was employed for immunization, while AM1-BSA was used for enzyme-linked immunosorbent assay (ELISA).</p><p>Next, liposomes were prepared. 1,2-Dimyristoyl-sn-glycero-3-phosphocholine (DMPC) was purchased from Lipoid GmbH (Ludwigshafen, Germany), 1,2-dimyristoyl-sn-glycero-3-phospho-(1′-rac-glycerol) (DMPG) was purchased from Nippon Fine Chemical (Osaka, Japan), cholesterol was purchased from NOF (Japan), lipid-maleimide linker was synthesized by Molecular Express (Rancho Dominguez, CA), monophosphoryl lipid A (MPLA) was purchased from Avanti Polar Lipids (Alabaster, AL), and S-[2,3-bis(palmitoyloxy)-(2RS)-propyl]-N-palmitoyl-(R)-cysteinyl-alanyl-glycine (Pam3CAG) was purchased from Bachem (Torrance, CA). Briefly, VesiVax® CALVs25 were prepared by mixing DMPC, DMPG, cholesterol, lipid-maleimide, with or without MPLA and/or Pam3CAG in chloroform/methanol (1:1 v/v). The organic solvents were removed under a stream of nitrogen gas at 65 °C, and residual solvents were removed in vacuo for >24 h. Unilamellar liposomes were formed by hydrating the lipid films with 100 mM sodium phosphate pH 7.026–28 and probe sonicating until translucent. The particle sizes of the liposomes were 40–60 nm, as measured by dynamic light scattering (UPA 150, Microtrac, Montgomeryville, PA). The ratio of lipids was optimized to afford a maleimide functional group density of approximately 700 per liposome. For each liposome formulation evaluated, the abbreviation LP# indicates which type of TLR agonist was incorporated during lipid hydration and vesicle formation. LP2 refers to liposomes with TLR2 agonist Pam3CAG.29,30 LP4 refers to liposomes with TLR4 agonist MPLA.31,32 LP24, designed for agonism of both TLR2 and TLR4, refers to liposomes with both Pam3CAG and MPLA. LP0 refers to liposomes without any TLR agonist.</p><p>With the AM1-KLH immunogen and the set of CALVs in hand, union of the former with the latter commenced. The hapten-protein conjugate AM1-KLH was pre-treated with 40 mM dithiothreitol (DTT, final concentration) in PBS pH 7.2 at room temperature for 2 h in order to generate reactive sulfhydryl groups. Excess DTT was removed with an Econo-Pac 10DG desalting column (Bio-Rad, Hercules, CA). Immediately following the desalting process, the reduced hapten-protein conjugate was mixed with each of the four CALVs described above. After incubation at room temperature for 1 h, conjugation efficiencies were confirmed by HPLC; however, aggregation was observed. Bicinchoninic acid (BCA) assay of the conjugated samples (post-filtration, 0.22 μm syringe filter) suggested that ~70% of the AM1-KLH was lost, while the concentration of lipid components remained unchanged (HPLC analysis). A second conjugation was performed to reach the desired AM1-KLH dosage, using reduced and sterile-filtered AM1-KLH. Aggregation was again observed. However, no secondary filtration was performed since the reduced AM1-KLH and liposome samples were sterilized pre-conjugation. It is known that KLH is unstable under certain conditions. This protein is sourced from an oceanic creature that lives in a high-salt environment, and has a propensity to precipitate under conditions of lower salinity.33 Also, each intact KLH didecamer (8 MDa) is a hollow cylinder approximately 35 nm in diameter by 40 nm in length.34 For comparison, liposomes fabricated for this study were 40–60 nm in diameter.</p><p>To evaluate these liposomal nicotine vaccine formulations, groups of n=5 BALB/c mice (9 weeks, 25–30 g) were immunized subcutaneously on days 0, 14, and 35 with 200 μL of one of the following: AM1-KLH + Sigma Adjuvant System® (SAS, Sigma-Aldrich, St. Louis, MO), AM1-KLH-LP2, AM1-KLH-LP24, AM1-KLH-LP4, or AM1-KLH-LP0. Sera were collected via retro-orbital bleed on days 21, 42, and 56 (see Figure 1).</p><p>Serum anti-nicotine antibody titers were determined by ELISA.6 Mouse sera were serially diluted in 1% BSA across Costar 3690 plates pre-coated with AM1-BSA. After incubation at 37 °C in a moist chamber for 90 min, plates were washed with deionized H2O, treated with goat anti-mouse antibody horseradish peroxidase conjugate (Southern Biotech, Birmingham, AL) for 30 min, and washed again with deionized H2O. Plates were developed using 3,3′,5,5′-tetramethylbenzidine and H2O2 (TMB Substrate Kit, Thermo Pierce, Rockford, IL). Color development was halted by the addition of 2 M H2SO4, and plates were read on a plate reader (SpectraMax M2e, Molecular Devices, Sunnyvale, CA). By plotting absorbance versus log dilution, mid-point titers – dilutions affording an absorbance reading 50% of the maximum value – were obtained.</p><p>A summary of titer measurements is provided in Figure 2. Importantly, the data suggest superior performance of liposome formulations that were fortified with TLR4 agonist MPLA (with or without the help of TLR2 agonist Pam3CAG), compared to the conventional hapten-protein immunoconjugate approach. This is consistent with numerous studies underscoring the advantage of using liposomal lipid A.35–40 Notably, the non-liposomal group received SAS adjuvant, a component of which is, in fact, MPLA. For this group of mice that received AM1-KLH + SAS, each injection furnished 50 μg MPLA per animal, whereas, for the groups that received either AM1-KLH-LP24 or AM1-KLH-LP4, each injection furnished ~40 μg MPLA per animal. Despite containing a lesser amount of potent adjuvant MPLA, these two liposomal vaccine formulations outperformed the non-liposomal formulation. Finally, the vaccine formulations possessing LP2 and LP0 were comparable, suggesting little advantage in using the TLR2 agonist Pam3CAG alone.</p><p>Nicotine-specific serum antibody binding affinities and antibody concentrations were determined by competitive radioimmunoassay (RIA) using 3H-labeled nicotine.6,21,22 The protocol described herein is an adaptation of the procedure described by Müller.41 First, the serum dilution that binds ~50% of 3H-labeled nicotine is determined. Then, the affinity constant is calculated by competition with unlabeled nicotine. Since the sera were pooled for each vaccine group, the measured affinity constants are average affinities for each group.</p><p>Competitive RIA was carried out in a 5kDa MWCO Equilibrium Dialyzer-96 (Harvard Apparatus, Holliston, MA) to allow easy separation of bound and free L-[N-methyl-3H]-nicotine tracer; SA = 81.7 Ci/mmol (PerkinElmer, Boston, MA). Mouse sera were pooled and diluted in 2% BSA to a concentration that bound ~50% of ~30,000 dpm of 3H-nicotine tracer. Each sample chamber was loaded with 75 μL of diluted sera and 75 μL of radiolabelled tracer (~30,000 dpm), and each buffer chamber was loaded with 150 μL of unlabeled (−)-nicotine at varying concentrations in 1% BSA. The chamber contents were equilibrated on a plate rotator (Harvard Apparatus, Holliston, MA) at room temperature for at least 22 h. A 75 μL aliquot was removed from each sample/buffer chamber and added to 5 mL scintillation fluid (Ecolite(+)™, MP Biomedicals, Santa Ana, CA). Radioactivity (dpm) of each aliquot was measured in a Beckman LS 6500 Scintillation Counter.</p><p>Figures 3 and 4 show results from competitive RIA of pooled sera from each of the five vaccine groups. Due to the limited volume of sera available, each measurement was performed twice. Additional replicates would further galvanize the Kd values and permit a more nuanced interpretation of the results. Nevertheless, one broad conclusion can be made. The CALV platform can indeed be exploited in the nicotine vaccine arena for production of murine polyclonal sera exhibiting nicotine-specific Kd values that are on par with others that have been reported. For instance, immunization of BALB/c mice with 3′-EstNic-rCTB + Alum resulted in anti-nicotine antisera with Kd ~56 nM,21 and immunization of BALB/c mice with 3′-EstNic-rQbVLP + Alum resulted in anti-nicotine antisera with Kd ~46 nM.22</p><p>Serum anti-nicotine antibody concentrations are also on par with reports of a comparable nature. For instance, immunization of rats with AM1-KLH + AS-03 elicited antisera with anti-nicotine [Ab] ~40 μg/mL,6 and immunization of rats with 3′-AmNic-rEPA + complete/incomplete Freund's adjuvant elicited antisera with anti-nicotine [Ab] ~184 μg/mL.42 It has been stated that an efficacious nicotine vaccine should be able to elicit a serum anti-nicotine antibody concentration of at least 200 μg/mL in animals. However, Freund's adjuvant, despite being highly potent, is not approved for human use. The results herein demonstrate that, without the aid of TLR2 and/or TLR4 agonism, the anti-nicotine antibody response is minimal.</p><p>Based on these encouraging initial results, we will continue to evaluate VesiVax® CALVs for the rapid generation of additional nicotine vaccine formulations. In this vein, efforts to integrate a carrier protein (or peptide) other than KLH are underway, and results will be reported in due course. Additionally, adjustment of antigen and TLR agonist dosages and screening of other TLR agonists may lead to further increases in immunogenicity.</p>

PubMed Author Manuscript

Luminescent Di and Polynuclear Organometallic Gold(I)-M (Au2, {Au2Ag}n and {Au2Cu}n) Compounds Containing Bidentate Phosphanes as Active Antimicrobial Agents

The reaction of new dinuclear gold(I) organometallic complexes containing mesityl ligands and bridging bidentate phosphanes [Au2(mes)2(\xce\xbc-LL)] (LL = dppe: 1,2-Bis(di-phenylphosphano)ethane 1a, and water-soluble dppy: 1,2-Bis(di-3-pyridylphosphano)ethane 1b) with Ag+ and Cu+ lead to the formation of a family of heterometallic clusters with mesityl bridging ligands of the general formula [Au2M(\xce\xbc-mes)2(\xce\xbc-LL)]A (M = Ag, A = ClO4\xe2\x88\x92, L-L = dppe 2a, dppy 2b; M = Ag, A = SO3CF3\xe2\x88\x92, L-L = dppe 3a, dppy 3b; M = Cu, A = PF6\xe2\x88\x92, L-L = dppe 4a, dppy 4b). The new compounds were characterized by different spectroscopic techniques and mass spectrometry The crystal structures of [Au2(mes)2(\xce\xbc-dppy)] 1b and [Au2Ag(\xce\xbc-mes)2(\xce\xbc-dppe)]SO3CF3 3a were determined by a single-crystal X-ray diffraction study. 3a in solid state is not a cyclic trinuclear Au2Ag derivative but it gives an open polymeric structure instead, with the {Au2(\xce\xbc-dppe)} fragments \xe2\x80\x9clinked\xe2\x80\x9d by Ag(\xce\xbc-mes)2 units. The very short distances of 2.7559(6) \xc3\x85 (Au-Ag) and 2.9229(8) \xc3\x85 (Au-Au) are indicative of gold-silver (metallophillic) and aurophilic interactions. A systematic study of their luminescence properties revealed that all compounds are brightly luminescent in solid state, at room temperature (RT) and at 77 K, or in frozen DMSO solutions with lifetimes in the microsecond range and probably due to the self-aggregation of [Au2M(\xce\xbc-mes)2(\xce\xbc-LL)]+ units (M= Ag or Cu; LL= dppe or dppy) into an extended chain structure, through Au-Au and/or Au-M metallophylic interactions, as that observed for 3a. In solid state the heterometallic Au2M complexes with dppe (2a\xe2\x80\x934a) show a shift of emission maxima (from ca. 430 to the range of 520\xe2\x80\x93540 nm) as compared to the parent dinuclear organometallic product 1a while the complexes with dppy (2b\xe2\x80\x934b) display a more moderate shift (505 for 1b to a max of 563 nm for 4b). More importantly, compound [Au2Ag(\xce\xbc-mes)2(\xce\xbc-dppy)]ClO4 2b resulted luminescent in diluted DMSO solution at room temperature. Previously reported compound [Au2Cl2(\xce\xbc-LL)] (L-L dppy 5b) was also studied for comparative purposes. The antimicrobial activity of 1\xe2\x80\x935 and AgA (A= ClO4\xe2\x88\x92, OSO2CF3\xe2\x88\x92) against Gram-positive and Gram-negative bacteria and yeast was evaluated. Most tested compounds displayed moderate to high antibacterial activity while heteronuclear Au2M derivatives with dppe (2a\xe2\x80\x934a) were the more active (MIC 10 to 1 \xce\xbcg/mL). Compounds containing silver were ten times more active to Gram-negative bacteria than the parent dinuclear compound 1a or silver salts. Au2Ag compounds with dppy (2b, 3b) were also potent against fungi.

luminescent_di_and_polynuclear_organometallic_gold(i)-m_(au2,_{au2ag}n_and_{au2cu}n)_compounds_conta

Introduction<!>1. Chemistry and characterization<!>2. Luminescence Studies<!>3. Antimicrobial Activity<!>Conclusion<!>1. Synthesis and Characterization of the Polynuclear Metal Complexes<!>[Au2(mes)2(\xce\xbc-LL)] (LL = dppe 1a; dppy 1b)<!>[Au2Ag(\xce\xbc-mes)2(\xce\xbc-LL)]ClO4 (LL = dppe 2a; dppy 2b)<!>[Au2Ag(\xce\xbc-mes)2(\xce\xbc-LL)]SO3CF3 (LL = dppe 3a; dppy 3b)<!>[Au2Cu(\xce\xbc-mes)2(\xce\xbc-LL)]PF6 (LL = dppe 4a; dppy 4b)<!>2. Single-Crystal X-ray Diffraction Studies<!>3. Luminescence studies<!>4. Microbial Toxicity Assays

<p>The photophysical and luminescent properties of closed-shell d10 gold(I) compounds have been widely investigated over the last decades.[1] The studies have focused on finding a correlation between the emission properties of polynuclear gold(I) derivatives and aurophilic interactions. Aurophilicity[2] or the weak Au(I)-Au(I) interaction displayed in most polynuclear gold(I) derivatives has been attributed to relativistic effects (increase in the effective nuclear charge when high-speed electrons are moving close to a heavy atomic nucleus).[2,3] This relativistic effect involves a contraction of the less-diffuse orbitals(s- and p- orbitals) and an expansion of the more difusse orbitals (d- and f- orbitals) and it reaches the maximum for gold.[2–4] The tendency in gold(I) to form metal-metal interactions with other Au(I) centers or other metals of similar charge (e.g. Ag(I), Cu(I), Tl(I)) is attributed to the sub-bonding interaction introduced through the stabilization of the filled 5d-orbital-based molecular orbitals with the empty molecular orbitals of appropriate symmetry derived from the 6s and 6p orbitals by configuration mixing.[1–3] Polynuclear homometallic gold(I) and heterometallic gold(I)-M (e.g. M = Ag(I), Cu(I), Tl(I)) derivatives constitute an important family of luminescent metal compounds.[1] The presence of gold in these derivatives enhances the spin-orbit coupling of the system, which in turn facilitates the access to triplet excited states by intersystem crossing. Relaxation of the triplet excited state by radioactive decay would usually result in phosphorescence with large Stokes shifts.[1a] Since the first reports[4,5] on the photoluminescence of [Au2(μ-dppm)2]2+, di and polynuclear gold(I) phosphane derivatives have been widely studied,[1,6] including organometallic compounds,[7] compounds with sulfur-[6g,n,8] and nitrogen-containing[9] ligands and chalcogenide centred gold derivatives.[10] In some cases the short gold(I)…gold(I) distances in these derivatives may not play a decisive role in determining the emission energy, as the auxiliary counter anion or solvent can dramatically affect their photophysical properties.[6h,g,m] Other luminescent polynuclear gold(I) complexes without phosphanes incorporating carbeniate,[11] sulfur ligands,[12] ylide,[13] stibinne,[14] and most recently N-heterocyclic carbene[15] ligands have been described. Examples of luminescent gold-containing polymers and gold nanoparticles are also known.[16]</p><p>Heterometallic di and polynuclear gold(I)-M (M = Ag, Cu) with[1c,17,18] and without[1c,19,20] phosphane ligands have also displayed interesting luminescent properties. These properties may be associated with different factors such as the nature of the ligand and the heterometal, or the presence or absence of metallophilic interactions. In general for Au-M (M = Ag, Cu) compounds with weak Au-M interactions (not clusters) the metallophilic interactions are mainly responsible for the photophysical properties observed. The number of polynuclear homometallic[4,6a,c,f,m–p,r,7b,c,e,f,8b,f,9a,c,11,14,15] or heterometallic[1c,17b–e,h,18b,c,19a,b,d,f,h,j,l,20c] gold(I) compounds luminescent in solution at room temperature is more limited. Most of these complexes are only brightly emissive in the solid state at room temperature.[e.g: 1a,6h] The measurements in solution are mostly performed in degassed solvents like CH2Cl2, CHCl3, CH3CN, Me2CO or THF. Examples of polynuclear gold(I) compounds luminescent in DMSO or aqueous solution at RT are limited to [Au2(μ-G).(μ-dmpe)](KBr)0.75.2H2O (G = guaninato dianion, dmpe = 1,2-bis(dimethylphosphano)ethane)[9b] and some N-heterocyclic dinuclear gold(I) carbene compounds.[15b] Interestingly, some of these carbene gold(I) derivatives which also displayed potential antitumor properties (targeting mitochondria) were used in luminescence studies of intracellular distribution.[15b] More recently, Au-Ag alkynyl phosphane aggregates (encapsulated by silica nanoparticles) which exhibit intense phosphorescence free from oxygen quenching, have been applied in two-photon imaging in human mesenchymal stem cells.[17b]</p><p>Gold(I)[21] and silver(I)[22] derivatives have been studied in the last few decades for their potential applications in medicine. While silver(I) derivatives are used mainly as antibacterial agents,[22,23] gold(I) compounds (especially those containing thiolates and phosphanes) have been used in the treatment of rheumatoid arthritis.[21a] More recently, gold(I)-phosphane and carbene complexes have been studied as potential antitumor,[15,24] antiparasitic[25] and antimicrobial agents.[26,27] Silver carbene derivatives have also displayed high activity against selected tumor cancer cells in vitro[27] and Gram-positive and Gram-negative bacteria.[27,28] It is however surprising that the biological activity of heterometallic gold-silver derivatives has been scarcely investigated[29] when a synergistic or cooperative effect of the two metals (as described for other heterometallic systems like Ti-Ru or Ti-Au in anticancer therapy)[30] could be anticipated.</p><p>We report here on the preparation of organo-heterometallic derivatives with mesityl and bis-phosphane bridging ligands of the general formula [Au2M(μ-mes)2(μ-LL)]A (M = Ag, A = ClO4−, L-L = dppe 2a, dppy 2b; M = Ag, A = SO3CF3−, L-L = dppe 3a, dppy 3b; M = Cu, A = PF6−, L-L = dppe 4a, dppy 4b) (Scheme 1) from new organometallic dinuclear complexes [Au2(μ-mes)2(μ-LL)] (L-L = dppe 1a, dppy 1b). These compounds display very short Au-Au and Au-M distances that imply metallophylic interactions and, as a result, are luminescent in solid state at room temperature and at 77 K, or in frozen solutions. The compounds seem to aggregate into polymers (Scheme 1) in the solid state or concentrated solutions. The compounds are soluble in DMSO and mixtures of DMSO /H2O and their antimicrobial activity against Gram-positive and Gram-negative bacteria and yeast has been evaluated. Most tested compounds display moderate to high antibacterial activity while heteronuclear Au2M derivatives with dppe (2a-4a) are the more active (MIC 10 to 1 μg/mL) and more active than the dinuclear Au2 parent compound (1a) against Gram-negative bacteria. Compounds containing silver, Au2Ag (2a, 3a) are also more active than silver salts (AgX; X = ClO4−, OSO2CF3−) against both Gram-negative and Gram-positive bacteria. Au2Ag compounds with dppy (2b, 3b) are also potent against fungi. Compound 2b is luminescent in diluted DMSO solutions at room temperature and could potentially be used in the future in studies of fluorescence microscopy to track these types of derivatives inside yeast or mammalian cells.</p><!><p>Some years back we and others reported on the use of mesityl gold complexes of the type [Au(mes)L] (L = AsPh3, PPh3) as precursors to polynuclear homo and heterometallic gold(I) derivatives.[31–34] The mesityl group (2,4,6-Me3C6H2) can act as a terminal ligand or as a bridge between two or more metal centres, affording different bonding modes (the most common one a three-centre-two-electron bond). Thus dinuclear Au2,[35] Au-Ag[31] and trinuclear Au2M (M = Ag, Cu)[32] gold(I) derivatives with mesityl bridging ligands were prepared by addition of weakly coordinated gold, silver or copper compounds (such as Ag(OSO2CF3), Ag(OClO3), [{Au(PPh3)}2(μ-Cl)2]ClO4 or [Cu(CNMe)4]PF6 to derivatives of the type [Au(mes)L] or [Ag(mes)]4.[35] Some of the compounds were structurally characterized and displayed supported metal-metal interactions (some of them can be considered formally as metal-metal bonds). The cluster [Au(mes)]5 was prepared by different ways[33,35–37] and addition of Ag+ and Cu+ ions to this and the related cluster [Au(trip)]6 (trip = 2,4,6-{(CH3)CH]3C6H2))[34] resulted in heterometallic gold(I) compounds with very interesting structural features and unsupported Au-Ag bonds like that of [Au6Ag(trip)6](CF3SO3).[34] More recently, the addition of silver perfluorocarboxylates to [Au(mes)]5 afforded heterometallic complexes of the type [AuAg4(RCO2)4(tht)x]n with supported Au…Ag interactions and the mesityl ligands bridging 3 centers in an unprecedented situation.[38] We also prepared Au(I) mesityl derivatives with the bidentate phosphane dppm (1,2-Bis(diphenylphosphano)methane)[31] and dppe (1,2-Bis(diphenylphosphano)ethane).[39] In the case of dppm we obtained the mononuclear derivative [Au(mes)(dppm)] which was used as a precursor to Au-Ag compounds with terminal mesityl and bridging dppm which did not display formal Au-Ag bonds.[40] In the case of dppe dinuclear [Au2(mes) 2(μ-dppe)] 1a was obtained instead.[39] The addition of AgA (A = OSO2CF3; OClO3) or [Cu(CNMe)4]PF6 to 1a lead to the formation of heterometallic compounds (Scheme 1) with mesityl bridging ligands of the general formula [Au2M(μ-mes)}2(μ-dppe)]A (M = Ag, A = ClO4−, 2a; A = SO3CF3−, 3a; M = Cu, A = PF6−, 4a). These cationic air-stable heterometallic compounds are soluble in DMSO and could potentially be used in biological studies. Besides, we noticed that they are brightly luminescent in the solid state under a common vis-UV lamp. Although the synthesis and partial characterization of 1a-4a had been described before by one of us,[39] the results were never published and crystallographic, luminescence and biological studies of these complexes had not been undertaken.</p><p>Complexes 1a–4a are air- and moisture-stable white (1a) or yellow solids (2a–4a). Acetone solutions of cationic compounds 2a–4a display conductivities typical of 1:1 electrolytes. The IR spectra show absorptions arising from the anions ClO4− (2a) at 1088 (br, vs), 623 (s) cm−1, CF3SO3− (3a) at 1262 (br), 1221 (s), 1154 (s) cm−1 and PF6− (4a) at 839 (br, vs) cm−1. The 31P{1H} NMR (CDCl3) of 1a shows a singlet at 42.5 ppm. In 2a (44.9 ppm), 3a (45.2 ppm) and 4a (44.0 ppm) the signal is down-field displaced from that of 1a as in other polynuclear gold phosphane complexes with mesityl bridges. The single signal indicates that all phosphorous in the molecule are chemically equivalent. The 1H NMR spectra of 2a–4a show three singlets for the mesityl ligands, slightly displaced from the resonances of the starting material 1a, and in a consistent ratio with the phenyl and methylene resonances of the ancillary ligands. There is only one type of mesityl ligand for every compound. The mass spectra (FAB+) for 1a does not show the parent peak but signals that can be assigned to polynuclear fragments of similar and higher molecular weight could be observed instead: [Au2(mes)(dppe)]+ [M – mes]+ at m/z = 911 (100%) and [Au3(mes)2(dppe)]+ [M + Au]+ at m/z = 1227 (42%). This is an indication that, the preparation of trinuclear derivatives from 1a, is feasible. The mass spectra (FAB+) show, for all the heterometallic compounds, the ion peak [M – A]+ with 100% intensity at m/z 1136 (2a, 3a) and at m/z 1092 (4a).</p><p>The structure of compound 3a has been determined by an X-ray diffraction study (Figure 1); selected bond lengths and angles are given in table 1.</p><p>The crystal structure of 3a consists of an "open" polymer of the type [(μ-Ag){Au2(μ-mes)2(μ-dppe)}]n(CF3SO3)n·1.6n(H2O). Crystals of 3a were highly unstable when removed from their mother liquor. The structure was solved by direct methods (more details in the SI section), which revealed the positions of the heavy atoms and of a subset of the C, O and F sites.</p><p>The remainder of the structure was located and refined in an alternating series of least-squares refinements and difference Fourier maps. The crystallographic asymmetric unit was found to include 2.5 units of the building block of the cationic polymer, [(μ-Ag){Au2(mes)2(μ-dppe)}]+, 2.5 triflate anions, CF3SO3−, and six water sites, of which four were assigned occupancies of 0.5. The section of the complex polymer in the chosen asymmetric unit is bounded at one end by a silver atom, Ag1, located on a crystallographic two-fold axis at (0.25, y, 0.5). Beginning at this point, the chain extends with two gold atoms, Au1 and Au2, each of which is ligated by one mesityl ligand and one P atom of a dppe ligand that bridges the two gold centers. The basic three-metal unit, Ag…Au…Au, is repeated (Ag2, Au3, Au4) with all three metal centers on general positions. Finally, the part of the chain in the asymmetric unit ends with Ag3 and Au5, which belong to a third link whose Au5…Au5ii component straddles a crystallographic two-fold axis at (0.75, y, 1.0). The dppe ligand that bridges the Au5…Au5ii unit is disordered about the two-fold axis. For the asymmetric unit we chose a chemically connected dppe, of which one P atom, P5, is bonded to Au5 with no symmetry applied and P6 is bonded to Au5ii (ii: 1.5-x, y, 2-z). The atoms of this dppe ligand, P5, P6 and C53 through C78, were assigned occupancies of 0.5 in line with the two-fold disorder. Two of the triflate sites were found to be fully occupied and the third, which includes atoms S3 and C153, was found near a crystallographic two-fold axis at (0.5, y, 0.25) and was assigned occupancy of 0.5. Six single-atom sites were located in intermolecular space and modeled as water sites. Based on considerations of the refined displacement parameters and chemically unreasonable contacts, four of these sites were refined with occupancies fixed at 0.5. These may not represent the exact stoichiometry of the crystal, and indeed we cannot guarantee that the stoichiometry remained constant in the brief time between when the highly unstable sample was removed from its mother liquor and when it was placed in the cold stream of the diffractometer.</p><p>Every silver center is bonded to the ipso carbon atoms of the mesityl groups and also bridges two {Au2(μ-dppe)} fragments with an Ag-Au distance which ranges from 2.7560(6) to 2.8506(13) Å (Table 1). The shorter distances (ca. 2.75 to 2.78 Å) are of the same order as those found in complexes with formal supported silver-gold bonds,[41] especially in the most closely related example with mesityl ligands [{Au(μ-mes)AsPh3}2Ag](ClO4)[32] (2.7758(8) Å). The longer distances Ag-Au found in 3a of 2.80 to 2.85 Å are of the same order of distances found in complexes where a formally nonbonding Ag….Au interaction has been proposed like in related mesityl complexes such as [{(Ph3P)Au(μ-mes)Ag(tht)}2](SO3CF3)2 [2.8245(6) Å][31] or [AuAg4(mes)(RCO2)4(tht)x]n (x = 1, R = CF3, CF2CF3, x = 3, CF2CF3)[38] which range from 2.8140(8) to 3.0782(6) Å (depending on the carboxylate). In some of these latter complexes one mesityl ligand is bridging one Au and two silver centers[38] and this is one of the reasons the Ag-Au distances are considerably longer. Thus, we can postulate appreciable silver-gold bonding interactions in 3a. In general the distances Ag-Au in compounds with supported silver-gold interactions are longer than those with unsupported ones and usually the derivatives with those supported gold-silver interactions do not display luminescence attributable to the metallophilic interactions. The distances Au-Au in 3a of 2.9226(8) and 2.9228(8) Å are quite short indicating a strong aurophilic interaction.[42] Similar and mostly longer distances have been found in luminescent polynuclear gold(I) derivatives with bis-phosphanes like [Au2(dppm)2]2+ (2.931(1)–2.962(1) Å depending on the counter ion),[5] [Au2(dmpe)2]2+ (dmpe = bis(dimethylphosphano)ethane; 2.9265(5)-2.974(3) Å depending on the counter ion),[6r] [Au3(dmmp)2]3+ (dmmp = bis(dimethylphosphanomethyl)methylphosphane; 2.962(1) and 2.981(1) Å),[6p] [Au2(dpephos)]2+ (dpephos: bis-(2-diphenylphosphano)phenylether); 2.9764(13)-3.0038 (6) Å depending on the counter ion),[6f] [Au2(xantphos)Cl2] (xantphos = 9,9-dimethyl-4,5-bis(diphenylphosphano)xanthene; 2.9947(4) Å), [6a] or [m-C6H4(OCH2CCAu)2(μ-dppm)] (3.049(1) Å).[7d] The Au2Ag derivatives described here (2a,b; 3a,b) which display quite short Ag-Au and Au-Au distances (as demonstrated for 3a) are pale yellow and brightly yellow emissive in solid state as described next. Gold atoms are in almost linear environments. The M-C bond lengths (Au-C distances range from 2.069(14) to 2.098(15) Å and Ag-C from 2.252(14) to 2.368(14) Å) are similar to those found in the mesityl heterometallic complexes mentioned above.[31,32,38]</p><p>We prepared the analogue di- (1b) and trinuclear (2b–4b) mesityl organometallic gold compounds with water soluble diphosphane dppy: 1,2-Bis(di-3-pyridylphosphano)ethane (Scheme 1). All complexes are air- and moisture-stable white (1b), pale yellow (2b–3b) or green solids (4b) which crystallize with molecules of water (see experimental). The heterometallic complexes 2b–4b are not soluble in CHCl3 or CH2Cl2 but they are soluble in CH3CN and DMSO. CH3CN solutions of cationic compounds 2b–4b display conductivities typical of 1:1 electrolytes. The IR spectra show absorptions arising from the anions ClO4− (2b) at 1082 (br, vs), 616(s) cm−1, CF3SO3-− (3b) at 1257 (br,vs), 1158 (m) cm−1 and PF6− (4b) at 839 (br, vs) cm−1. The 31P{1H} NMR (CDCN3) of 1b shows a singlet at 34.2 ppm. In 2b (32.9 ppm), 3b (32.9 ppm) and 4b (33.9 ppm) the broad signals are high-field displaced from that of 1b. However, in the 1H NMR spectra in CH3CN of 2b–4b the three singlets for the mesityl ligands are slightly displaced from the resonances of the starting material 1b (like in other heterometallic complexes with bridging mesityl ligands) and in a consistent ratio with the pyridyl and methylene resonances of the ancillary ligands. Like in the previous case only one type of phosphorous atom is observed in the 31P{1H} NMR spectra and only one type of mesityl ligand in the 1H NMR spectra for compounds 1b–4b in CH3CN. More detailed NMR experiments of these complexes are explained later on. The mass spectra (ESI+) for 1b shows the parent peak [M] at m/z: 1057 [100%] as well as signals that can be assigned to polynuclear fragments of similar molecular weight like: [Au2(mes)(dppy)]+ [M – mes]+ at m/z = 915 (14%). The mass spectra (ESI+) show, for all the heterometallic compounds the ion peak [M – A]+ with 100% intensity at m/z 1141 (2b, 3b) and at m/z 1097 (4b).</p><p>The structure of compound 1b has been determined by an X-ray diffraction study (Figure 2); selected bond lengths and angles are given in the legend of figure 2. The molecule is a dimer with two symmetric units (see Figure 2). The gold atoms are in a nearly linear environment (C(1)-Au(1)-P(1) 178.0(1)) and the distance Au-C is similar to the distances in other compounds where the mesityl group acts as a terminal ligand like in [Ag(μ-dppm)2{Au(mes)2}]ClO4 (2.083(10) and 2.080(9) Å).40 The distance Au-P is quite common for Au(I)-phosphane derivatives (including dinuclear complexes with bisphosphanes) and does not deserve further comment. No intra- or intermolecular gold-gold interactions are observed. Since the molecular conformation would be expected to be flexible, with unhindered rotation possible about either the Au1---P1 or Au1---C1 bond, it is pertinent to ask what factors contribute to the observed centric linear molecular topology, in which the torsion angle Au1…P1…P1′…Au1′ is 180° but in which the planes of the mesityl ligand and one of the pyridyl groups are parallel to each other to give a torsion angle C6---C1…P1---C17 of 2.9°. The extended topology mitigates for relief of steric effects, while the parallel orientations of the two rings brings the methyl group at C9 as close as possible to the C18---H18 bond of the pyridyl ring. We can identify both intra- and intermolecular interactions that serve to explain these observations.</p><p>As the positions of hydrogen atoms are important to this discussion, we point out here that, while all H atoms were located at calculated positions and refined as riding atoms (with free rotation for the methyl groups), the final locations of the relevant H atoms were confirmed to be correct through the use of omit maps – difference Fourier maps in which the contribution of the atom or atoms in question is omitted from F(calc). Contoured maps were made for the two unique pyridyl rings to confirm their correct orientations, and for the methyl groups at C7 and C9. Most relevant is that the methyl group at C7 was thus confirmed to be correctly oriented. With that orientation, H7A makes its closest possible approach to Au1, 2.71 Å (no standard uncertainty is provided, since the local C---H distance was constrained during refinement). On the opposite side of Au1, H18 of the pyridyl ring that is parallel to the mesityl ligand, and the position of which is not in doubt, makes its closest possible approach to Au1, 2.89 Å. The observed molecular conformation brings the methyl group at C9 as close as possible to the C18---H18 unit of the nearby pyridyl group, although the H…H contacts thus formed are both greater than 2.6 Å and thus not indicative of steric problems. We note that an omit map for the methyl group at C9 showed that the H atoms are located at positions between two closely-spaced low maxima, with one H parallel to the plane of the mesityl phenyl ring but distal to Au1, and with the other two H atoms thus straddling the extended pyridyl plane containing the neighboring C18---H18 bond.</p><p>Whether the two Au1…H contacts are important in establishing the molecular conformation observed in the crystal is open to discussion, but these contacts must serve as stabilizing influences. The recent unambiguous observation by neutron diffraction of an O---H…Pt hydrogen bond with an H…Pt distance of 2.885(3) Å[43] demonstrated that such interactions are bona fide hydrogen bonds and are potentially important in molecular solids.[44] As for intermolecular contacts in this Van der Waals' solid, the shortest ring…ring contact has a Cg…Cg distance of 4.695(3) Å; and we conclude that these are of no structure-directing importance. There are four C---H…π contacts, of which one has clearly a stablizing influence; one is within the limits normally considered for establishing the potential importance of such contacts, and two are borderline and of dubious importance (see Table in SI).</p><p>The nitrogen atoms in the pyridyl groups of the dppy phosphane could potentially coordinate metallic centers when 1b interacts with silver or copper salts. In the IR spectra of the Au2Ag complexes (2b and 3b) in solid state we observe broad signals at ca. 1640 cm−1 that seem to correspond to the coordination of metal fragments to the nitrogen of the 3-pyridyl group. This band appears in the IR spectra of the free phosphane dppy at 1568 cm−1 and in compound 1b (both as solids) at 1571 cm−1. In the IR spectra (of the solid 2b and 3b) we observe a broader band at 1576 (2b) or 1570 cm−1 (3b) and a new broad band at 1642 cm−1 for both complexes. Both similar bands are observed for the Au2Cu complex (4b), with that at 1575 cm−1 being weaker. The presence of these two bands could indicate the existence of two types of pyridyl groups in these compounds, with or without coordination to the metallic centers. Similar displacement of the pyridine band has been described in cyclo- and polyphosphazene with pyridine side groups after the coordination of the nitrogen atom to Au and Ag centers.[45] In our case we believe that these N-Ag or N-Cu interactions do not occur in solution since we do not observe a change in the 1H NMR signals of the protons from the pyridyl ring and we just observe one signal in the 31P{1H} NMR spectra for the compounds (in CD3CN or d6-DMSO). This signal is, however quite broad (as opposed to the sharp signal for 1b) which may indicate a fluxional behavior of the compounds at RT. These complexes (2b–4b) are not soluble in deuterated CHCl3, CH2Cl2, acetone, MeOH or THF. The low temperature 31P{1H} NMR for these complexes in CD3CN only displays one broad signal. Unfortunately with acetonitrile we only could get reliable data up to −35 °C due to the relatively high freezing point of this solvent. It is very reasonable to assume than in solution the polar solvent CH3CN or DMSO (media where the heterometallic complexes are soluble) may facilitate the metal decoordination to the nitrogen atom. A similar effect has been described for heterocyclic carbene silver compounds containing pendant pyridyl groups. In this case the interactions of the Ag atoms with the N of the pendant pyridyl groups are lost in CD3CN solutions.[19i]</p><p>In order to obtain more information we performed an X-ray analysis of crystals obtained while crystallizing compound 2b in a dichloromethane-acetone-DMSO solution by slow diffusion of n-hexane at −5°C. We have to mention that the crystals obtained (2c) are colorless and non luminescent, as opposed to compound 2b itself which is pale yellow and brightly yellow luminescent at RT in solid state and in solution. In the crystallization tube a yellow, luminescent non crystalline powder was also observed which may correspond to compound 2b, whose IR spectrum and powder-diffractogram are different to those of compound 2c, as it will be mentioned later.</p><p>The structure of 2c was analyzed by single-crystal diffraction; and while the quality of the data permitted an unambiguous determination of the connectivity and general geometrical parameters, low resolution and weak diffraction prevented a full anisotropic refinement. A detailed description of the analysis and of the results is given in the supplementary material. The basic unit of the cationic polymer is a trinuclear (AgAu2) fragment of formula [AgAu2(μ-mes)2(μ-dppy)], which crystallizes with one equivalent of perchlorate, ClO4− and two equivalents of DMSO. But the crystallographic asymmetric unit is comprised of three links of the polymer chain along with the corresponding three units of ClO4− and 6 DMSO sites. One of this individual units is depicted in Figure 3. The individual units of the polymer, labeled "A," "B," and "C" in the atoms list, are shown in a figure in the SI. The A unit has a significantly shorter Au1…Au2 distance [Au1A…Au2A 3.3079(9) Å] than do the B and C units [Au1{B,C}…Au2{B,C} 3.4533(10), 3.4720(9) Å]. Within the asymmetric unit, the trinuclear core of the B fragment is rotated roughly 180° about the crystallographic c-axis with respect to the orientation of the A and C fragments. The latter are roughly parallel, although the difference in the Au1…Au2 distances between the A and C fragments ameliorates the severity of the pseudo-symmetry to at least some extent. The distances Au-Ag range from 2.822(1) to 2.998(1) Å with one Ag coordinated to two Au atoms in every individual unit (in unit A: Au1-Ag= 2.998(1) Å, Au2-Ag = 2.917(1) Å; in unit B: Au1-Ag= 2.933(1) Å, Au2-Ag = 2.917(1) Å; in unit C: Au1-Ag= 2.931(1) Å, Au2-Ag = 2.822(1) Å). These distances are similar or longer than the longest distances Ag-Au found in 3a of 2.80 to 2.85 Å. As commented before these values are of the same order of the distances found in complexes where a formally nonbonding Ag….Au interaction has been proposed. The Au-Au distances found in this structure are long enough for not being considered aurophilic interactions. The lack of metallophilic interactions may be the reason why these colorless crystals are not luminescent. There are two mesityl ligands acting as bridges between the Ag and the two Au atoms in every unit and the distances M-C are similar to those found in mesityl heterometallic complexes described before.[31,32,38] Importantly, in this case the silver atom of every unit is coordinated to the nitrogen atom of one pyridyl group of the dppy ligand of another unit (see Figure 3 and figures in the SI).</p><p>It seems plausible from the structure of 2c that the core {Au2Ag(μ-mes)2(μ–dppy)}+ is the same as in compound 2b in solid state and in solution. However, and as mentioned before, for 2b in solution an interaction between nitrogen atoms of the pyridyl groups and silver atoms is not observed.</p><p>For luminescent 2b in the solid state, it seems more reasonable that the compound may be a polymer which displays shorter Au-Au and/or Au-Ag interactions (as in 3a) as well as the coordination of Ag atoms to one or more N atoms from the pyridyl groups (as indicated by the IR spectrum of 2b and 3b). In fact, the powder-diffractogram of the luminescent and yellow compound 2b indicates that its structure is completely different to that of 2c (see supporting information). The IR spectrum of 2c is similar to that of 2b, (although it is not exactly the same) showing also a band at ca. 1654 cm−1 due to the coordination of the silver to the nitrogen of the 3-pyridyl group. Unfortunately we have not been able to obtain crystals of enough quality from 2b to assess their structure in the solid state. In the crystallization attempts to obtain monocrystals of 2b, white crystals of 2c are always obtained instead. What it seems clear is that in all these luminescent heterometallic complexes the mesityl is acting as a bridging ligand between the gold and other metal (Ag or Cu) centers in a 3c-2e− fashion and that the units {Au2M(μ-mes)2(μ–LL)} may assemble in the space through metallophilic interactions (Scheme 1). The compounds with dppy have the possibility of coordination of one or more nitrogen atoms to the silver or copper atom in the solid state. Different counterions and solvents may have an effect also in the structure in the solid state as they can also coordinate to the second metallic center. This effects may be the reason for differences found in the luminescence studies between compounds with dppe and dppy (a versus b) and even, in compounds with ddpy with different counterions (2b and 3b).</p><!><p>We have studied the luminescence of all new compounds 1–4. Previously reported compound [Au2Cl2(μ-L-L)] (L-L = dppy 5b[46]) was also studied for comparative purposes. The luminescence of 5a (L-L = dppe[47]) in solid state at RT[8g,9a] had been reported before (data in Table 3).</p><p>All new complexes are luminescent both at room temperature and at 77 K in solid state and in frozen solutions. Although complexes 2b and 3b are brightly luminescent under a common UV lamp (ex 365 nm) in concentrated solutions at RT, only 2b luminesces in a diluted DMSO solution (5 × 10−4 M) at RT. The excitation and emission data as well as the lifetimes for excited states are summarized in Table 3. The lifetimes are all relatively long (7–57 μs), which indicate the emission transitions are all forbidden and phosphorescent. The free phosphanes display a broad band with emission maxima at 433 (dppe) and 493 nm (dppy) in solid state at RT, upon excitation between 270–310 nm and 300–410m, respectively. Complexes 2a–4a, with dppe, are strongly luminescent in solid state or in frozen solutions (dichloromethane or DMSO). All of these complexes display a similar optical behavior both in solid state and in frozen solutions. The spectra show a simple excitation profile with its maximum located between 401 (2a) and 414 nm (4a), leading to a maximum emission band appearing between 522 (2a) and 533 nm (4a) in solid state at RT (see figure 4a and 4b), which is lightly red shifted when the temperature is lowered to 77 K. This last shift observed with decreasing temperature is a standard phenomenon in luminescent gold complexes with metallophillic interactions and has been related to a thermal contraction of the gold-gold interactions that leads to a reduction in the band gap energy.[48] This result is indicative of transitions influenced by the gold-gold or gold-heterometal interactions, as has also been observed in other gold(I)-heterometal compounds displaying Au-M interactions.[19e,19f,19l]</p><p>In contrast, the luminescence profile and features (excitation and emission maxima) of gold precursor 1a are very different, suggesting a different origin. For this dinuclear gold compound, an emission pattern of four peaks is observed (see figures 4c), which become a better resolved vibrational fine structure when the temperature is lowered to 77 K (figure 4d). This structured band is due to vibrations of phenyl groups (the energy difference between neighboring bands being ca. 1330 cm−1) and is similar to that showed in the emission spectrum of [AuCl(PPh3)][49] and other dinuclear gold complexes containing phosphane ligands,[4,9a,9c] whose emission was attributed to the ligand based π-π* transitions or to metal to ligand charge transfer (MLCT, Au 5d →PR3 π*). The energy of the emission for this complex 1a resembles that obtained for the free phosphane, dppe (433 nm). Thus, an intraligand transition modified by the coordination to gold is probably responsible for the luminescence in 1a. The same conclusion can be reasonably applied to 1b. Its spectrum shows very broad bands with emission maximum at 505 nm (excitation maximum at 415 nm), which is also similar to that of free phosphane, dppy (493 nm), and for which its crystalline structure show no intra- or intermolecular gold-gold interactions, as mentioned before.</p><p>On the other hand, unlike what we observed for 1a, the absorption spectra of the 2a-4a solid also differ from those of its solutions, which is also indicative that the absorption and emission properties of these heterometallic compounds are supramolecular in nature and that the extended metallic interactions in the solids are crucial for the observation of luminescence. The diffuse-reflectance UV (DRUV) spectrum of 1a in solid state displays bands with maxima at λ < 300 nm and upon its coordination to heterometallic atoms in 2a–4a, broad bands appear at ca. 400 nm (see also supporting information). These new bands lead to the observed emissions and could correspond to charge transfer processes between the metallic atom and the ligands or to metal centered transitions.[50]</p><p>In fact, none of the degassed solutions of compounds 2a–4a are luminescent at room temperature, neither in dichloromethane nor DMSO, in which the molecular self-aggregation of [Au2M(μ-mes)2(μ-LL)]+ units (M= Ag or Cu) through metallic interactions could be lost. Taking into account the spectroscopic data shown in the Chemistry Section, we assume that the simplest repeated unit in diluted solutions is [{Au2M(μ-mes)}2(μ-LL)]+, which wouldn't therefore give an emission as a result of the interactions between metals. We have also carried out a study of the absorption and emission spectra at different concentrations (from 5 × 10−4 until 10−2 M) for 2a in order to detect a possible deviation of the Lambert-Beer's Law, which would be consistent with the presence of molecular aggregation through aurophilic interactions in fluid solution, as was first detected by Laguna and coworkers,[19l] and which has also been observed in other Au-Ag complexes due to argento-aurophilic bonding.[19k, 19f] Unfortunately, we have not been able to obtain conclusive evidence from these studies at different concentrations. However, the observation of a similar luminescence spectrum for 2a–4a in frozen solution (dichloromethane or DMSO) and in solid state seems to be indicative that molecular aggregation through metallophilic atractions is implicated, as observed in solid state for 3a. In fact, the blue shift of the emission and excitation when measurements are carried out in frozen solution relative to that in solid state can be explained by a higher aggregation in solid state, as has also been observed in other gold(I)-heterometal compounds displaying extended Au-M interactions [19e, 51]</p><p>Complexes 2b–4b, with dppy, are also strongly luminescent in solid state or in frozen DMSO solutions. However, their behavior is not similar. The main facts we observed for these compounds are as follows:</p><p>1) Complexes 3b and 4b show a similar behavior to those of analogue complexes 3a and 4a, which have dppe instead of dppy. The only observed difference is that for complex 4b the band seen at RT (range of emission 475–675 nm, with the maximum at 563 nm) unfolds into two others which are not so well defined when the measurement is taken at 77 K (at 522 and 595 nm). For these complexes, the absorption spectra of the solid also differ from those of the solutions. (In the supporting information section the DRUV spectra of compounds 1b–4b and the excitation and emission spectra of 3b and 4b in the solid state at RT are shown). These facts also indicate that the metallophilic interactions are crucial for the observation of luminescence. 2) Complex 2b behaves differently depending on the aggregation state and temperature. In the solid state at RT this compound exhibits two emissions, whose intensities depend on the excitation wavelength and temperature (see figure 5). Thus, at 315 nm, it exhibits two emission bands with the maxima at 360 and 520 nm, respectively. Excitation at a longer wavelength, 398 nm, gives only the emission band at 520 nm, which is similar to that observed in the rest of heterometallic complexes 2a–4a and 3b–4b, suggesting a similar origin. Thus, the metallophilic interactions between [Au2M(μ-mes)2(μ-LL)]+ units are probably responsible for the luminescence in these heterometallic complexes. In fact, as we have already mentioned, the colorless crystals 2c obtained from a solution of 2b show no gold-gold or gold-silver interactions and are not luminescent.</p><p>A similar behavior at room temperature was observed by Yoshida et al[52a], which was attributed to the presence of two kinds of [Au(CN)2]n oligomers, the lower energy band being associated to the larger oligomers. Thus, the emission band at 360 nm might be assigned to oligomeric species including a small number of [Au2Ag(μ-mes)2(μ-dppy)]+ units such as dimer and trimer. The low energy band might originate from the larger oligomers. Upon freezing, 2b display intense luminescence with a maximum at 522 nm, which is comparable to the wavelength of the low energy band observed at RT. This result gives further confirmation that the low energy band originate from the extended chain structures.</p><p>Moreover, this complex 2b is the only one that luminesces in a diluted DMSO solution (5 × 10−4 M) at RT showing an emission band with a maximum at 524 nm (excitation maximum at 339 nm), which seems to indicate that for this complex the metallophilic interactions are not lost in a diluted solution. We have also carried out a study of the absorption and excitation spectra at different concentrations (from 5 × 10−5 to 10−3 M) for 2b in order to detect a possible deviation of the Lambert-Beer's law. Thus, an increase of concentration from 5 × 10−4 to 10−3 M at RT produces the displacement of the excitation band to 366 nm. The absorption spectra of 2b in DMSO show that the weak shoulder appearing near 325 nm does not obey Lambert-Beer's law, either, and appears at slightly lower energy when the concentration increases (see supporting information).[52b] This deviation from Lambert-Beer's law is consistent with molecular aggregation in fluid solution because, as the number of metallophilic interactions increase the HOMO-LUMO gap is reduced. These results are consistent with an extended chain structure for the solid 2b, similar to that observed to 3a, and in fact, similar spectral changes have also been previously reported in other gold-silver complexes with extended chain structures with gold–gold or gold–silver interactions [19f, 19k, 19l and refs. therein] The observed emission energy in diluted solution (524 nm) is similar to that of lower energy observed for the solid (520 nm), which seems to indicate that the extended chain structure of 2b is kept in diluted solution.</p><p>In summary, all these results are consistent with an extended chain structure as that observed for 3a. However, from these data it is not possible to assign the nature of the transition associated because there are several possibilities, including metal to ligand charge transfer (MLCT) or metal centered (MC).[52c] Both of them have been used to explain the luminescent properties observed in other complexes with extended gold(I)-M interactions, which have emission and excitation maxima similar to those of our compounds. [19e,19f,19l,19m,20c]</p><!><p>The antimicrobial activity of the new compounds (1–4) as well as that of previously described compounds 5a, 5b and the silver salts AgOClO3 and AgOSO2CF3 were evaluated against Gram-negative (Salmonella typhimurium and Escherichia coli), Gram-positive (Bacillus cereus and Staphyloccocus aureus) bacteria and yeast (Saccharomyces cerevisiae) (Table 4). A number of the compounds analyzed in this work exhibited Minimum Inhibitory Concentration (MIC) values in the 1 ug/ml – 100 ug/ml range (Table 4). None of the compounds tested showed significant activity against the eukaryotic S. cerevisiae at the concentrations tested. The compounds with the dppe ligand (a) had greater activity against bacteria than the compounds with the dppy ligand (b). The parent dinuclear compounds [Au2(mes)2(μ-LL)] 1a, 1b were insoluble or inactive against Gram-negative bacteria at concentrations of 100 μg/mL or lower. While compound 1a (with dppe) was toxic (10 μg/mL) for Gram-positive bacteria, 1b (with dppy) was inactive at concentrations of 100 μg/mL or lower. In the case of previously reported compounds [Au2Cl2(μ-LL)] 5a, 5b the derivative with the water soluble phosphane (dppy) was active (100 μg/mL) against Gram-negative and Gram-positive bacteria while 5a (with dppe) was inactive for all bacteria and yeast at the same concentration. The heretometallic compounds containing copper [Au2Cu(μ-mes)2(μ-LL)]PF6 4a, 4b were toxic for Gram-positive bacteria. Derivatives containing dppe (4a) were 10 times more active than those with dppy (4b) although their toxicity was similar to that of parent dinuclear compound 1a. 4b was more toxic to Gram-positive bacteria than parent compound 1b. The most toxic compounds were the heterometallic gold-silver compounds [Au2Ag(μ-mes)2(μ-LL)]A (2, 3). All these complexes were toxic to Gram-negative and Gram-positive bacteria. There were not significant differences depending on the counterion (the toxicity may come from the heterometallic Au2Ag cation). The derivatives with dppe (2a, 3a) were 10 (or even 100) times more active than those with dppy. However the derivatives with dppy (2b, 3b) were active against yeast at concentrations of 100 μg/mL.</p><p>Since we had observed that the heterometallic compounds decoordinate the "naked" metallic center over time in DMSO solutions, we tested the activity of the silver salts AgOClO3 and AgOSO2CF3 for comparison. We did not evaluate the activity of the air, moisture and light sensitive Cu(I) derivative [Cu(CH3CN)4]PF6 since it would have decomposed in DMSO solution immediately upon addition. The silver salts were toxic to Gram-positive and Gram-negative bacteria and yeast (100 μg/mL) and especially to Gram- positive bacteria B. cereus (10 μg/mL). The toxicity was the same for the Au2Ag derivatives with dppy (2a, 3a). However compounds Au2Ag with dppe resulted more toxic (10 times) than the silver salts for all Gram-positive and Gram-negative bacteria pointing out that, in this case, the toxicity does not come from the dissociation of the silver perclorate or triflate from the heterometallic derivative. The heterometallic compounds were also more toxic to Gram-negative bacteria than parent dinuclear compounds 1a and 1b. In this case it is obvious that there is a synergistic or cooperative effect between the gold and silver metals. The toxicity for parent dinuclear compound (1a) is similar for Gram-positive bacteria but the compound is not active for Gram-negative bacteria at concentrations of 100 μg/mL or lower. The cationic heterometallic Au2Ag compounds 2a, 3a allow for a better solubilization in the cell culture media and display high toxicity against Gram-negative bacteria (10 μg/mL) and Gram-positive bacteria (10 μg/mL). Remarkably, the complex with ClO4− (2a) displays a much higher toxicity for Gram-positive bacteria B. cereus (1 μg/mL) that any other of the compounds (parent dinuclear gold, silver salts or other heterometallic derivatives) tested.</p><p>The organometallic dinuclear gold(I) compound [Au2(mes)2(μ-dppe)] 1a resulted more active against Gram-positive bacteria and also more active than previously described coordination derivative [Au2Cl2(μ-dppe)] 5a. The fact that gold compounds are more toxic for Gram-positive bacteria than Gram-negative or fungi has been noted previously with auranofin[53a] and some other gold(I) phosphane derivatives including some described by us.[24e,26a,53b–f] However the nature of other ancillary ligands coordinated to the gold centers (besides the phosphanes) plays a decisive role. For instance a more complicated pattern is found for gold-thiol-phosphane derivatives and the antimicrobial potency can be higher for Gram-negative or fungi depending on the combination thiol-phosphane.[53a] We thought that the incorporation of water soluble phosphane dppy in the polynuclear gold(I) compounds would increase the solubility in DMSO and mixtures DMSO/water of the organometallic analogue compound and thus the toxicity against microorganisms. We found however that water-soluble derivative [Au2Cl2(μ-dppy)] 5b was more toxic than novel organometallic compound [Au2(mes)2(μ-dppy)] 1b as opossed to what was seen for dppe (5a vs 1a). Related complexes of gold(I) with dppe and pyridyl-substituted diphosphanes like [Au(LL)2]Cl (LL= dppe, dppy) had been described in the literature as cytotoxic agents against a panel of cisplatin-resistant human ovarian carcinoma cell lines.[54] Their citotoxicity was strongly dependent upon their lipophilicity being more cytotoxic and hepatotoxic those compounds more lipophilic (with dppe).[54a]</p><p>As mentioned before, the heterometallic Au2Ag compounds with dppe (2a, 3a) were the most active. They showed strong activity, MIC of 10 ug/ml or less, against both Gram-positive and Gram-negative bacteria. Dinuclear silver(I)-oxygen bonding complexes derived from camphanic acic ligands have displayed a wide spectrum of effective antimicrobial activity.[23d] Some N-heterocyclic carbene Ag(I) complexes inhibit more efficiently the growth of Gram-positive bateria including B. subtilis,[28b–d] Gram-negative E. coli,[28c–g] and P. aeruginosa[28c,e,f] and even antibiotic resistant strains of S. aureus.[28a,d] Binuclear and polymeric silver(I) complexes with a tridentate heterocyclic N- and S- ligand 8-(9pyridin-3-yl)methylthio) quinoline were good inhibitors of Gram-positive bacteria and some Gram-negative bacteria (P. aeruginosa) but they were very poor against E. coli.[55] The behaviour of the gold-silver compounds 2a, 3a and AgA (A = OClO3 −, OSO2CF3 −) is similar to that of silver-carbene derivatives,[28] silver fluorinated tris(pyrazolyl)borate complexes[56] and AgNO3 and silver(I) sulfadiazine[56] inhibiting grow of both Gram-negative and Gram-positive bacteria. However in the case of the tris(pyrazolyl)borate complexes it was demonstrated that the effect on Gram-positive species was due to the ligand whereas the activity for Gram-negative species was truly due to the silver ion.[56] In our case it seems that the activity displayed is due to the presence of the metals. The MIC of 2a and 3a for Gram-negative E. coli and Gram-positive S. aureus of 10 μg/mL is very low, in the order of those obtained recently for highly active antimicrobial silver nanoparticles (12.5 μg/mL)[22b] and some silver heterocyclic carbenes (1–32 μg/mL)[28d–g] for the same microorganisms. The value of 1 μg/mL of 2a (nanomolar range) for Gram-positive B. cereus is quite remarkable.</p><p>We tested the ability of these two compounds in Minimun Bactericidal Concentration (MBC) assays against E. coli and S. aureus (Figure 6). Both 2a and 3a showed strong bactericidal activity, reducing cell counts by several orders of magnitude. We did not observe complete killing within 24 hours with 2a and 3a which suggests that either longer exposure times may be required to get complete killing or that the availability of active compound has dropped sharply in the 24 hours of the assay. We cannot rule out the development of some resistance within the bacterial population although this is very unlikely given the short exposure time to the compounds. Overall the heterometallic Ag2Au compounds behave as highly active antibacterial agents for Gram-positive and Gram-negative bacteria with toxicity against Gram-negative bacteria much higher than that of the parent dinuclear Au2 compound or the AgA salts alone.</p><!><p>In conclusion we have demonstrated that gold(I) organometallic complexes containing mesityl ligands and bridging bidentate phosphanes can serve as excellent precursors to highly luminescent heterometallic Au2M (M = Ag, Cu) derivatives and that the luminescence displayed, whether in the solid state or in frozen solutions, results from the formation of aggregated species that self-associate through metallophilic attractions. In solution, where the ions are dispersed, the complexes are non luminescent except compounds 2b and 3b in concentrated solutions which can be seen photoluminescent when excited with a hand-held UV lamp. The compounds display moderate to high antibacterial activity while heteronuclear Au2M derivatives with dppe (2a-4a) are very active (MIC 10 to 1 μg/mL) and more active than the dinuclear Au2 parent compound (1a) against Gram-negative bacteria. They are also more active than silver salts (AgX; X = ClO4−, OSO2CF3−) against both Gram-negative and Gram-positive bacteria. Au2Ag compounds with dppy (2b, 3b) are also potent against fungi. The new trinuclear Au2Ag compounds possess peculiar chemicophysical properties with respect to their precursors responsible for the observed biological effects. Compound 2b is also luminescent in diluted DMSO solutions at room temperature and could potentially be used in the future in studies of fluorescence microscopy to track these types of derivatives inside yeast or mammalian cells. These preliminary results warrant further studies on the synthesis of different luminescent biologically active Au-Ag heterometallic complexes.</p><!><p>Solvents were purified by use of a PureSolv purification unit from Innovative Technology, Inc.; all other chemicals were used as received. Elemental analyses were carried out on a Perkin-Elmer 2400-B microanalyser. Infrared spectra (4000-400 cm−1) were recorded on a Nicolet 380 FT-IR infrared spectrophotometer on KBr pellets. The 1H, 13C{1H} and 31P{1H} NMR spectra were recorded in CDCl3, CD3CN or d6-DMSO solutions at 25 °C on a Bruker 400 and spectrometer (δ, ppm; J, Hz); 1H and 13C{1H} were referenced using the solvent signal as internal standard while 31P{1H} was externally referenced to H3PO4 (85%). The mass spectra (FAB+, matrix: 3-nitrobenzylalcohol) were recorded from CH2Cl2 solutions on a VG Autopsec spectrometer and the mass spectra (ESI+, matrix: DCTB) on a MALDI-TOF MICROFLEX (Bruker) spectrometer. Conductivity was measured in an OAKTON pH/conductivity meter in (CH3)2CO and CH3CN solutions. Compounds [Au(mes)(AsPh3)], [31] 5a,[33] 5b,[34] and water-soluble phosphane dppy[57] were prepared as previously reported. All other chemicals and solvents were purchased from Sigma-Aldrich and Strem Chemicals.</p><!><p>To dppe (0.50 mmol, 0.199 g) or water-soluble phosphine dppy (0.24 mmol, 0.098 g) in 15ml of CH2Cl2, [Au(mes)(AsPh3)][31] (0.5 mmol, 0.622 g, to obtain 1a or 0.489 mmol, 0.304 g to obtain 1b) was added and the mixture stirred for 20 min at rt. The solution was then evaporated to 3 mL and 20 mL of n-hexane was added. A compound precipitated which was filtered and dried in vacuo to afford pure 1a (white solid) or 1b (pale yellow solid). 1a: Yield: 0.491 g, 91%. Anal. Calc. for C44H46P2Au2 (1030): H, 4.55; C, 51.75; found H, 4.50; C, 51.3; MS(FAB+): [M/Z, (100%)]: 911 [M – mes]+. 31P{1H} NMR (CDCl3): δ = 42.0 ppm, (d6-dmso): δ= 42.7 ppm. 1H NMR (CDCl3): 2.30 (p-Me), δ = 2.64 (o-Me), 2.85 ((CH2)2), 6.97 (m-H) ppm. 13C{1H} (CDCl3): δ = 21.1 (p-Me), 24.2 ((CH2)2), 26.8 (o-Me), 126.6 (Mes, C3, C5), 127.8 (Mes, C1), 129.4 (Ph, C4), 131.5 (Ph, C3, C5), 133.3 (Ph, C2, C6), 135.5 (Ph, C1), 145.8 (Mes, C2, C4, C6) ppm. Conductivity (acetone): 4.8 S cm2 mol−1(neutral). 1b: Yield: 0.162 g, 93%. Anal. Calc. for C40H42N4P2Au2·H2O (1052): H, 4.21; C, 45.64; N, 5.33; found H, 3.89; C, 45.27; N, 5.75. MS(ESI+) [M/Z, (100%)]: 1057 [M + Na]. 31P{1H} NMR (CD3CN) δ = 34.4 ppm; (d6-DMSO): δ = 34.7 ppm. 1H NMR (CD3CN): δ = 2.23 (p-Me), 2.48 (o-Me), 3.02 ((CH2)2), 6.83 (m-H), 7.48 (H3), 8.09 (H4), 8.74 (H2), 9.01 (H6) ppm; (d6-dmso) : δ = 2.16 (p-Me), 2.34 (o-Me), 2.51 ((CH2)2), 6.71 (m-H), 7.57 (H3), 8.23 (H4), 8.76 (H2), 9.08 (H6) ppm. 13C{1H} (CDCl3): δ = 21.3 (p-Me), 23.7 ((CH2)2), 27.2 (o-Me), 124.4 (Py, C3), 127.0 (Mes, C3), 136.4 (Py, C5), 140.9 (Py, C4), 145.4 (Mes, C1), 153.0 (Py, C2), 153.7 (Py, C6) ppm. Conductivity (CH3CN): 13.2 S cm2 mol−1(neutral).</p><!><p>To a solution of [Au(mes)(μ-dppe)] 1a (0.20 mmol, 0.206 g) for the preparation of 2a or [{Au(mes)}2(μ-dppy)] 1b (0.15 mmol, 0.155 g) for the preparation of 2b in 20 mL of CH2Cl2 was added to a solution of AgClO4 (0.20 mmol, 0.415 g 2a, 0.15 mmol, 0.031g 2b) in 10 mL of Et2O. The solution was stirred at 0°C protected from the light for 20 min. The solvent was then reduced to 2–3 mL and 20 mL of n-hexane was added. A yellow solid precipitated and was filtered and dried in vacuo to afford pure 2a (yellow solid) or 2b (pale yellow solid). 2a: Yield 0.204 g, 82%. Anal. Calc. for C44H46P2Au2AgClO4 (1136): H, 3.75; C, 42.69; found H, 3.45; C, 42.25. MS(FAB+) [M/Z, (%)]: 1136 [M]+. IR: ν(ClO4−) = 1088 (br, vs), 623(s) cm−1. 31P{1H} NMR (CDCl3) δ = 44.9 ppm. (d6-DMSO): δ = 46.9 ppm. 1H NMR(CDCl3): δ = 2.30 (p-Me), 2.34 (o-Me), 3.07 ((CH2)2), 7.04 (m-H) ppm. 13C{1H} NMR(CDCl3): δ = 21.0 (p-Me), 23.2 ((CH2)2), 27.2 (o-Me), 127.9 (Mes, C3, C5), 129.6 (Ph, C2, C4, C6), 130.1 (Mes, C1), 133.6 (Ph, C3, C5), 136.6 (Ph, C1), 145.3 (Mes, C4), 154.3 (Mes, C2, C6) ppm. Conductivity (acetone): 160 S cm2 mol−1 (1:1 electrolyte). 2b: Yield: 0.135 g, 72%. Anal. Calc. for C40H42N4P2Au2AgClO43H2O (1296): H, 3.73; C, 37.07; N, 4.32; Found H,3.34; C, 36.86; N,4.23. MS(FAB+) [M/Z, (%)]: 1141 [M]+. ν(ClO4−) = 1082 (br, vs), 616(s) cm−1. 31P{1H} NMR (CD3CN): δ = 32.7 ppm; 1H NMR (CD3CN): δ = 2.41 (o-Me), 2.24 (p-Me), 6.86 (m-H), 3.14 ((CH2)2), 7.53 (H3), 8.12 (H4), 8.74 (H2), 9.04 (H6) ppm. 13C{1H} NMR (CDCN3): δ = 20.8 (o-Me), 21.3 (p-Me), 26.6 ((CH2)2), 127.0 (Mes, C3, C5), 124.3 (Py, C3), 133.3 (Mes, C2), 141.1 (Py, C4), 153.0 (Py, C2), 153.4 (Py, C6) ppm. Conductivity (CHCN3): 121.6 S cm2 mol−1 (1:1 electrolyte).</p><!><p>To a solution of [Au(mes)(μ-dppe)] 1a (0.20 mmol, 0.206 g) for the preparation of 3a or a solution of [{Au(mes)}2(μ-dppy)] 1b (0.10 mmol, 0.103 g) for the preparation of 3b in 20 mL of CH2Cl2 was added a solution of AgSO3CF3 (0.20 mmol, 0.514 g, 3a or 0.10 mmol, 0.026 g, 3b) in 10 mL of Et2O. The reaction mixture was stirred at 0°C protected from the light for 20 min (3a) or 90 min (3b). The solvent was then reduced to 2–3 ml and 20 mL of n-hexane was added. A yellow solid precipitated and was filtered and dried in vacuo to afford pure 3a (yellow solid) or 3b (pale yellow solid). 3a: Yield: 0.198 g, 76%. Anal. Calc. for C45H46P2Au2Ag SO3F3 (1286): H, 3.60; C, 41.97; S, 2.49; found H, 3.6; C, 41.85: S, 2.9. MS(FAB+) [M/Z, (%)]: 1136 [M]+. IR: ν(SO3CF3−) = 1262 (br), 1221 (s), 1154 (s) cm−1. 31P{1H} NMR (CDCl3): δ = 45.2 ppm. (d6-DMSO): δ = 47.0 ppm. 1H NMR (CDCl3): δ = 2.35 (p-Me), 2.65 (o-Me), 3.13 (P-(CH2)2-P), δ=7.06 (m-H) ppm. 13C{1H} NMR(CDCl3): δ = 21.4 (p-Me), 27.4 (o-Me), 30.9 ((CH2)2), 127.3 (Mes, C3, C5), 129.7 (Ph, C3, C5), 132.1 (Ph, C4), 133.6 (Ph, C2, C6) (Mes, C1 and Ph, C1 not showing) ppm. Conductivity: 107 S cm2mol−1 (1:1 electrolyte). 3b: Yield: 0.105 g, 82%. Anal. Calc. for C41H42N4P2Au2AgSO3F34H2O (1362): H, 3.55; C, 36.16; N, 4.11, found H, 3.11; C, 35.74: N, 3.96. MS(FAB+) [M/Z, (%)]:1141 [M]+. IR: ν(SO3CF3−) = 1257 (br,vs), 1158 (m) cm−1. 31P{1H} NMR (CD3CN): δ = 32.5 ppm. 1H NMR (CD3CN): δ = 2.25 (p-Me), 2.40 (o-Me), 3.16 ((CH2)2), 6.89 (m-H), 7.53 (H3), 8.13 (H4), 8.75 (H2), 9.05 (H6) ppm (3-pyridyl). 13C{1H} NMR (CDCN3): δ = 20.4 (p-Me), 20.7 ((CH2)2), 25.7 (o-Me), 127.0 (Mes, C3, C5), 124.6 (Py, C3), 141.4 (Py, C4), 153.1 (Py, C2), 153.7 (Py, C6) ppm. Conductivity (CHCN3): 181.0 S cm2 mol−1 (1:1 electrolyte).</p><!><p>To a 20 mL of CH2Cl2 solution of [Au(mes)(μ-dppe)] 1a (0.20 mmol, 0.206 g) for the preparation of 4a or [Au(mes)(μ-dppy)] 1b (0.10 mmol, 0.103 g) for the preparation of 4b, [Cu(CH3CN)4]PF6 (0.21 mmol, 0.078 g, 4a; 0.10 mmol, 0.037 g, 4b) was added and stirred for 20 min at r.t. The solvent was then reduced to 2–3 mL and 20 mL of n-hexane was added to obtain a precipitate which was filtered and dried in vacuo to afford pure 4a (greenish yellow solid) or 4b (green solid). 4a: Yield 0.162 g, 62%. Anal. Calc. for C45H46P3Au2F6Cu (1092): H, 3.74; C, 42.65; found H, 3.75; C, 42.25. MS(FAB+) [M/Z, (%)]:1092 [M]+. IR: ν(PF6−) = 834 (br, vs) cm−1. 31P{1H} NMR (CDCl3): δ = 44.0, −144.1 ppm (PF6, sept); (d6-DMSO): 42.6, −144.1 ppm (PF6, sept);. 1H NMR (CDCl3): δ = 2.31 (p-Me), 2.69 (o-Me), 3.03 ((CH2)2), 7.01 (m-H) ppm. 13C{1H} NMR(CDCl3): δ = 21.6 (p-Me), 22.2 ((CH2)2), 27.3 (o-Me), 127.5 (Mes, C3, C5), 128.3 (Ph, C3, C5), 129.2 (Mes, C1), 129.9 (Ph, C4), 132.4 (Ph, C2, C6), 133.4 (Ph, C1), 144.5 (Mes, C4), 154.0 (Mes, C2, C6) ppm. Conductivity (acetone): 130 S cm2 mol−1 (1:1 electrolyte). 4b: Yield: 0.09 g, 90%. Anal. Calc. for C40H42N4P3Au2F6Cu.5H2O (1333): H, 3.93; C, 36.03; N, 4.20. found H, 3.48 C, 35.72, N, 4.13. MS(FAB+) [M/Z, (%)]:1097 [M]+. IR: ν(PF6−) = 839 (br, vs) cm−1. 31P{1H} NMR (CD3CN): δ = 34.1 ppm; 1H NMR(CD3CN): δ = 2.22 (o-Me), 2.43 (p-Me), 6.81 (m-H), 3.05 ((CH2)2), 7.49 (H3), 8.11 (H4), 8.73 (H2), 9.01 (H6) ppm. 13C{1H} (CDCN3): δ = 20.0 (o-Me), 20.0 (p-Me), 26.1 ((CH2)2), 126.2 (Mes, C3, C5), 124.6 (Py, C3), 141.1 (Py, C4), 152.6 (Py, C2), 153.7 (Py, C6) ppm. Conductivity (CHCN3): 170.4 S cm2 mol−1 (1:1 electrolyte).</p><!><p>Data for 1b (CCDC 841439), 2c (CCD 841440) and 3a (CCDC 841438) were collected by using a Xcalibur S3 diffractometer (graphite monochromated Mo-Kα radiation, λ = 0.71073 Å, φ-ω scans). These data can be obtained free of charge from The Cambridge Crystallographic Data Center via www.ccdc.cam.ac.uk/data_request/cif.</p><p>1b: Crystals of 1b (yellow prisms with approximate dimensions 0.42 × 0.26 × 0.17 mm) where obtained when a solution of 1b in CH2Cl2 was crystallized by slow diffusion of n-hexane at −5 °C. As the crystals were found to be unstable when removed from their mother liquor, a single crystal of 1b was harvested using a Mitegen support with a drop of perfluorinated oil and then placed directly on the diffractometer at a temperature of 173 K. 2c: Crystals of 2c (colorless prisms with approximate dimensions 0.16 × 0.03 × 0.02 mm) where obtained when a solution of 2b in dichloromethane-acetone-DMSO solution by slow diffusion of n-hexane at −5°C. As the crystals were found to be highly unstable when removed from their mother liquor, a single crystal of 2c was harvested using a Mitegen support with a drop of perfluorinated oil and then placed directly on the diffractometer at a temperature of 100 K. 3a: Crystals of 3a (yellow sphenoids with approximate dimensions 0.39 × 0.18 × 0.12 mm) where obtained when a solution of 3a in CH2Cl2 was crystallized by slow diffusion of n-hexane at −5 °C. As the crystals were found to be highly unstable when removed from their mother liquor, a single crystal of 3a was harvested using a Mitegen support with a drop of perfluorinated oil and then placed directly on the diffractometer at a temperature of 150 K. The diffraction pattern was considered to be acceptable, although weak. The data for 1b, 2c and 3a were gathered and processed using the usual procedures for the Xcalibur S3 diffractometer (graphite monochromated Mo-Kα radiation, λ = 0.71073 Å, φ-ω scans) including multi-scan corrections for systematic errors. The structure was solved by direct methods, which revealed the positions of the heavy atoms and of a subset of the C, O and F sites. The remainder of the structure was located and refined in an alternating series of least-squares refinements and difference Fourier maps. Details of the solution and refinements for 1b and 3a compound are presented in Table 5 and in the SI while the data and full analysis of the results for 2c can be found in the SI.</p><p>The room-temperature powder diffraction pattern for Compound 2b was measured by the X-ray Diffraction and Fluorescence Service of the University of Zaragoza, using a Rigaku DMax 2500 using graphite-monochromated Cu radiation. Data were recorded from 5 – 60° 2-theta with a step size of 0.03°. The powder diffraction pattern for compound 2c was simulated using the program Platon, with the simulation based on the single-crystal diffraction analysis of 2c at T = 100 K. The difractograms are collected in the SI.</p><!><p>Absorption spectra in solution were recorded with a Unicam UV-4 spectrophotometer. Diffuse-reflectance UV (DRUV) spectra were recorded in the same equipment using a Spectralon RSA-UC-40 Labsphere integrating sphere. The solid samples were mixed with silica gel and placed in a homemade cell equipped with a quartz window. The intensities were recorded in Kubelka-Munk units: log[R/(1−R)2], where R= reflectance.</p><p>Steady-state photoluminescence spectra were recorded with a Jobin-Yvon Horiba Fluorolog FL-3-11 spectrofluorimeter using band pathways of 3 nm for both excitation and emission. Phosphorescence lifetimes were recorded with a Fluoromax phosphorimeter accessory containing a UV xenon flash tube at a flash rate between 0.05 and 25 Hz. The lifetime data were fit using the Jobin-Yvon software package and the Origin 7.0 program. The solid samples were prepared by mixing the compounds with silica gel.</p><!><p>Bacteria and yeast were stored as glycerol stocks at −80°C and streaked onto Meuller-Hinton plates prior to each experiment. Colonies from these newly prepared plates were inoculated into 5 ml of media (tryptic soy broth for E. coli, S. salmonella, S. aureus and B. cereus and YPD broth for S. cerevisiae) and grown overnight at 37°C (30°C for S. cerevisiae and 25°C for B. cereus). The overnight cultures were diluted to an OD600 <0.01 (ThermoSpectronic, Genesys 8 spectrophotometer) in 2 ml of fresh media in sterile culture tubes. The compounds (1–5, AgClO4 and AgSO3CF3) were brought up in DMSO to a concentration of 1 mg/ml and immediately added into the cell culture tubes at 1 μg/ml, 10 μg/ml or 100 μg/ml. Control tubes for each cell type were inoculated with DMSO alone. The culture tubes were then placed at the appropriate temperature in a shaking incubator (Lab-Line Orbital Shaker), at ~200 RPM. Cell growth was monitored by taking OD600 readings at several time points over the course of 24 hours. The minimal inhibitory concentration (MIC) was determined to be the concentration at which there was negligible increase in the OD600 value from the initial reading after 24 hours. All samples were tested in triplicate. The bactericidal activities of compounds 2a and 3a against S. aureus and E. coli were determined by plating dilutions of bacterial cultures on to Mueller-Hinton plates at each time point and counting the resulting number of colonies.</p>

PubMed Author Manuscript

Synthesis and optoelectronic properties of 2,6-bis(2-anilinoethynyl)pyridine scaffolds\xe2\x80\xa0

A series of sixteen bisphenylureas based on a 2,6-bis(2-anilinoethynyl)pyridine scaffold have been synthesized for use as potential anion sensors. Prior work with one of these receptors revealed a distinct fluorescence response in the presence of a suitable anion source with specificity towards chloride anion. This study demonstrates that the fluorescent properties of these receptors can be tuned through the systematic variation of the pendant functional groups.

synthesis_and_optoelectronic_properties_of_2,6-bis(2-anilinoethynyl)pyridine_scaffolds\xe2\x80\xa0

Introduction<!>Results and discussion<!>Conclusion<!>

<p>Supramolecular anion sensors have received considerable attention in recent years, specifically those displaying a colorimetric and/or fluorescence response in the presence of specific anions.1–7 A typical design motif for these sensors takes advantage of non-covalent interactions such as electrostatics,8–10 hydrogen bonding,11–13 anion-π,14–16 and hydrophilicity/hydrophobicity. 17–20 To obtain a colorimetric and/or fluorescent outcome, a fluorophore unit is often appended to the sensor scaffold, which changes its fluorescent response upon a binding event; however, the pendant fluorophores are often bulky and exhibit a limited range of responses.21 An alternative approach would be to functionalize a fluorophore backbone with typical anion binding motifs.22–24 This would eliminate the need for a receptor to have a pendant fluorophore and could allow for greater tunability of both fluorescence response and binding strength.</p><p>The 2,6-bis(arylethynyl)pyridine scaffold has found utility in numerous areas of chemistry. The inherent properties of such ethynylpyridines (conjugation, absorption/emission, pH-dependent speciation, metal binding capability, rigidity, etc.),25–27 have been exploited for applications such as liquid crystals,28 lightemitting materials,29 rotaxane-type structures,30 molecular magnets,31 antiangiogenic activity,32 polymer composites33 and coordination complexes.24,34,35 In contrast, the supramolecular chemistry of 2,6-bis(arylethynyl)pyridines has received little attention, perhaps in part due to the scarcity of uniting the fields of molecule/ion recognition with the synthesis of highly-conjugated, carbon-rich materials. Most of the host/guest studies reported have focused on exploiting the pyridine lone pair to bind metal ions22 or organoiodides.3 Alternatively, we hypothesized that the unique absorption/emission properties in tandem with the structural rigidity of (arylethynyl)pyridines aptly positions them to function as small molecule or ion receptors. Moreover, we envisioned that the 2,6-bis(arylethynyl)pyridine scaffold and its derivatives would serve as a versatile building block for the development of receptor molecules that target a variety of guests depending on the protonation state of the pyridine and the style of functional group appended to the arylethynyl unit.36</p><p>Our initial studies focused on sulfonamide derivatives of 2,6-bis(2-aniloethynyl)pyridine to form an anion binding cavity (e.g., 1, Fig. 1). To our delight, this class of receptors showed the remarkable propensity of forming 2 + 2 dimers with either water, halides or both depending on the protonation state of the receptor.37 Subsequent studies examined urea derivatives such as 2 (Fig. 1), which possessed much simpler solution speciation with halides, as the two urea binding groups obviate the need for higher order aggregates to form to satisfy the anion's hydrogen bonding preference. Crystallographic data indeed showed that the urea receptors formed a 1 : 1 complex with chloride starting from TFA-protonated 2 (R = H).35 More importantly, the urea derivatives exhibited switchable fluorescent and colorimetric responses upon protonation. For example, the p-methoxy derivative of 2 displayed an "on-off" fluorescence behavior in the presence of chloride, while the p-nitro derivative exhibited an "off-on" fluorescence behavior in organic solvents.34 The magnitude of the fluorescence event was dictated by the anion, resulting in a rare, fully organic "turn-on" fluorescent sensor for chloride, which typically quenches fluorescence. While the cause of this behavior is still under investigation, it would be desirable to tune these systems so that the "turn on" behavior could be observed even in competitive solvents such as DMSO or water. To achieve this goal we describe herein the synthesis of a small library of sixteen different 2,6-bis(2-aniloethynyl)pyridine bisureas (2–5/a–d). We also report an exhaustive study of their absorption and emission profiles in CH3CN, and ultimately show that a similar "turn-on" behaviour can be observed even in relatively polar solvents.</p><!><p>To obtain a full range of electron-poor and electron-rich scaffolds, we examined 2,6-bis(2-aniloethynyl)pyridine derivatives functionalized with tert-butyl (6), carboethoxy (7), trifluoromethyl (8), and methoxy (9) groups located at the 4-position on the aniline rings, to be consistent with our previous studies. Each derivative was synthesized via a twofold Sonogashira cross-coupling reaction between the respective 4-functionalized-2-ethynylanilines with 2,6-dibromopyridine as shown in Scheme 1; thus, protodesilylation of known anilines 10–1338,39 in basic MeOH (EtOH for 11) followed by cross-coupling afforded the 2,6-bis(2-aniloethynyl)pyridine cores 6–9 in good to excellent yield. Each of these derivatives was then reacted with 4-methoxyphenyl isocyanate (a), 4-nitrophenyl isocyanate (b), phenyl isocyanate (c) and pentafluorophenyl isocyanate (d) to furnish the sixteen bisureas 2–5/a–d.</p><p>Interestingly, the purification and characterization of some of these compounds proved problematic. First, the yields for reactions with 7 and 8 were in general considerably lower than those of their electron-rich counterparts, 6 and 9 (Table 1). This was anticipated since the aniline nitrogens on 7 and 8 should be less nucleophilic due to the presence of the para-substituted electron-withdrawing groups. However, these bisureas were accompanied by myriad side products, as revealed by TLC analysis. In the case of 4d, a small amount of a 2-quinazolinone derivative was isolated along with the desired bisurea; the X-ray crystal structure of this unanticipated material and discussion of its formation are given in the Supporting Information†.</p><p>Second, the 1H NMR spectra of certain bisureas were unexpectedly complex. For example, the proton spectrum of 3d (which appears as one spot by thin layer chromatography) is shown on the bottom of Fig. 2a; however, when this same compound is prepared as the TFA salt, the spectrum simplifies to afford the anticipated set of resonances (Fig. 2a top). Interestingly, when TFA is added to 3d while in solution, the 1H NMR spectrum does not resolve, suggesting the presence of a kinetically stable aggregate. This behaviour was observed more often in scaffolds containing electron-withdrawing units at R and/or R′, which is an indication that the urea pKas are lowered, leading to increased hydrogen bond donor ability. Clearly, protonation and addition of an anionic guest to 3d breaks up the self-association/ aggregation, leading to the simplified 1H NMR spectrum; therefore, compounds 2–5/a–d were isolated as their corresponding trifluoroacetic acid (TFA) salts.</p><p>Another unusual problem was encountered while attempting to determine binding constants and stoichiometries of some of the bisureas with various tetrabutylammonium salts. For example, the 1H NMR titration of 3d with tetrabutylammonium chloride is shown in Fig. 2b. During the course of the titration the 1H NMR spectra go from sharp and resolved prior to the addition of chloride to completely obscured at approximately one equiv. of chloride. At the half equivalent point visible aggregate is present in the NMR tube, but remarkably, upon addition of excess chloride the host goes back into solution and the 1H NMR spectra resolve.</p><p>The most likely explanation for the observed phenomena is an equilibrium involving aggregation in solution. This behavior was observed more often in scaffolds containing electron-withdrawing units at R and/or R′, which is an indication that when the urea pKas are lowered, hydrogen bond donor ability increases and aggregation increases. This phenomenon prevented the accurate determination of binding constants for this class of sensors, but it can clearly be seen that there is a strong propensity for these molecules to form complexes/aggregates in the presence of a suitable anionic guest.</p><p>This hypothesized self-aggregation in solution can also be observed in the X-ray crystal structure of 2b.40 In the solid-state a solvent water molecule resides inside the molecular cavity of 2b, forming two solvent Ow–H…N, N–H…Ow hydrogen bonds and one bridging intermolecular Ow–H…O H-bond (Fig. 3 left). The molecules are also joined together by three N–H…O H-bonds between the −NO2 group in one molecule and NH groups in another one. In the crystal structure the molecules assemble into 1-D molecular strips that form layers perpendicular to the b axis (Fig. 3 right). Even in the case of such a complex molecular shape, all H atoms that can form H-bonds are involved in intra-and intermolecular H-bonds in the crystal structure, providing another example of the Hamilton rule.41</p><p>With the library of compounds now in hand, we examined the absorption and emission properties of the bisureas as well as those of the scaffold cores. Given that an ideal anion sensor is one that is capable of functioning in polar media, such as DMSO, H2O or CH3CN, all UV-vis and fluorescence studies were performed in CH3CN solutions. Fig. 4 shows the absorption spectra for the bis(arylethynyl)pyridines 6–9. Depending on the electronic nature of the substituents attached to these cores a red or blue shift in absorption is observed. For example, the trifluoromethyl substituted core 8 has a blue-shifted absorbance compared to the tert-butyl substituted core 6, while the methoxy substituted core 9 is red-shifted. This same red and blue shifting behavior is seen in the corresponding emission spectra (Fig. 4). The electron-withdrawn cores are blue shifted and the electron rich cores are red shifted. One interesting phenomenon occurs in the case of the methoxy-substituted core 9. It contains two distinct fluorescence bands, which are likely the result of dual emission from one species in solution.</p><p>The photoluminescence quantum yield (PLQY) for each of the four ethynylpyridines followed the trend in which the most electron rich cores had the lowest PLQY and the most electron poor cores had the highest PLQY (Table 2).</p><p>The neutral bisureas of each core were investigated in a similar manner. Fig. 5 shows the UV-vis spectra for the tert-butyl derivatives 2a–d. Interestingly, the substituted phenyl ureas had virtually no effect on the resulting UV-vis λmax indicating that the ethynylpyridine cores are primarily responsible for the observed adsorption behavior of the bisurea scaffolds.</p><p>The neutral bisurea scaffolds were also investigated with respect to their fluorescence, with careful attention paid to those scaffolds containing electron withdrawing R and R′ groups (e.g. 3b, 3d, 4b, and 4d). Given that an ideal sensor is one that would go from an "off" state to an "on" state, we were primarily interested in electron withdrawing scaffolds that were nonemissive in their unbound state. Based on these criteria, compound 3d was further scrutinized with respect to its ability to sense chloride, whereas 4d was not studied further due to its tendency to decompose into the 2-quinazolinone byproduct described in the Supporting Information†.</p><p>Gratifyingly, compound 3d exhibits a fluorescence "off to on" response in the presence of chloride despite being in a polar MeCN solvent (Fig. 6). Although compound 3d is not soluble in water, these results provide evidence that this particular ethynylpyridine scaffold can be tuned to exhibit a fluorescence response even in the presence of a highly competitive/polar solvent. Further studies will focus on determining the exact cause of this fluorescence and will aim to exhibit this same response in an aqueous solvent system.</p><!><p>The inherently fluorescent ethynylpyridine scaffolds presented herein can be tuned to exhibit either a red or blue shifted fluorescence response depending on the electron donating or withdrawing ability of the pendant functional groups. Similarly, the hydrogen bonding behavior of the bisurea scaffolds is highly dependent on the pendant functional groups. NMR spectroscopic and solid state structural studies indicate that without a suitable guest these receptors hydrogen bond strongly with themselves, suggesting that these sensors could allow for larger binding constants for anionic guests and potentially greater selectivity. The simple derivatization of this class of receptors allows for a "turn-on" fluorescence response to analytes in increasingly polar solvents, a feature often lacking in small molecule organic anion probes.24</p><!><p>Electronic supplementary information (ESI) available: synthetic procedures and spectral data for 2–9; discussion of and X-ray data for the 2-quinazolinone derivative. CCDC reference number 844848. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c2sc00975g</p>

PubMed Author Manuscript

The regulation of hepatic Pon1 by a maternal high-fat diet is gender specific and may occur through promoter histone modifications in neonatal rats\xe2\x98\x86

The antioxidant (AOX) defense system is critical for combating whole-body oxidative stress, and the present study aimed to determine the consequences of a maternal high-fat (HF) diet on neonatal hepatic lipid accumulation, oxidative stress, the expression of AOX genes, as well as epigenetic histone modifications within Pon1, an AOX enzyme. Hepatic thiobarbituric acid reactive substances were significantly increased and nonesterified fatty acids decreased in offspring of HF-fed dams, while triglycerides increased in male but not female HF offspring when compared to controls (C). Pon1, Pon2, Pon3 and Sod2 were significantly increased in offspring of HF-fed dams when compared to C. However, the increase in Pon1 and Pon3 was only significant in male but not female offspring. When compared to C, the hepatic Pon1 promoter of male and female HF offspring had significantly more acetylated histone H4 as well as dimethylated histone H3 at lysine residue 4, which are both involved in transcriptional activation. Trimethylation of histone H3 at lysine residue 9, which is involved in transcriptional repression, was only associated with genes in females. Results from the present study reveal that a maternal HF diet affects hepatic metabolism in the neonate in a gender-specific manner, and these differences, in association with epigenetic modification of histones, may contribute to the known gender differences in oxidative balance.

the_regulation_of_hepatic_pon1_by_a_maternal_high-fat_diet_is_gender_specific_and_may_occur_through_

Introduction<!>Animal and experimental design<!>Neonatal hepatic triglyceride (TG), nonesterified fatty acid (NEFA) and thiobarbituric acid reactive substances (TBARS) analysis<!>RNA isolation and real-time polymerase chain reaction (RT-PCR) analysis<!>Chromatin immunoprecipitation (ChIP) analysis<!>Statistical analysis<!>Statement of ethics<!>Maternal gestational characteristics<!>Offspring observations<!>Gene expression<!>Histone modifications related to transcription activation<!>Histone modifications related to transcription repression<!>Discussion

<p>The intake of a high-fat (HF) diet during pregnancy impacts fetal development[1–3], and specifically, a recent study has demonstrated that markers of antioxidant (AOX) defense capacity were decreased in adult offspring of HF-fed dams [4]. Because the liver is infiltrated with byproducts of fatty acid oxidation and because altered handling of hepatic oxidation products is associated with liver damage and disease [5], the inappropriate programming of this crucial protective mechanism by a maternal HF diet may be deleterious for hepatic development and health. Additionally, because hepatic oxidative balance is often related to systemic antioxidant status [6–8], the altered production of these in the liver will likely have numerous consequences for all major organ systems.</p><p>The paraoxonase (PON) family of enzymes is critical for whole-system oxidative balance [9]. The liver is responsible for the synthesis of these enzymes that protect both itself as well as the rest of the body from prooxidant damage. Specifically, the PON1 enzyme, which has lactonase and esterase activities and is synthesized in the liver, has been shown to be a critical component of the AOX defense system, and its synthesis was altered in patients with hepatosteatosis [10]. Therefore, it is not surprising that modulations in PON1 enzyme activity in liver as well as serum have been correlated to liver disease and altered lipoprotein metabolism [11–14]. Furthermore, studies have suggested that gender differences in the response to oxidative stress may contribute to the differential susceptibility of males and females to developing chronic diseases [15–20], and the AOX defense system is likely responsible for these differences.</p><p>While the adult epigenome can be modified by various elements and toxins in the environment [21], studies have also shown that in utero exposures can result in significant epigenetic modifications, and these epigenetic marks can persist into adulthood to permanently control gene expression [22–24]. The covalent modification of histone tails – including acetylation (Ac) or methylation (Me) – represents a class of epigenetic changes that regulate transcriptional activity of genes by altering the state of the chromatin [25]. In primates, the chronic consumption of a maternal HF diet led to an increase in H3K14Ac and a trend of increase in H3K9Ac, H3K18Ac, H3K9Me2, trimethylated histone H3 at lysine residue 9 (H3K9Me3) and H3K27Me3, and these fetal hepatic tissues had an increase in triglycerides and nonalcoholic fatty liver disease [26]. Furthermore, a maternal HF diet in rats resulted in decreased association of acetylated histone H3 (H3Ac), dimethylated histone H3 at lysine residue 4 (H3K4Me2), H3K9Me3 and H3K27Me3 within the promoter of the hepatic gluconeogenic Pck1 gene, along with a decrease in H3K9Me3 and an increase in acetylated histone H4 (H4Ac) and H3K4Me2, and all of these modifications were associated with increased transcription and transcriptional rate of the Pck1 gene [27]. While studies have authenticated the importance of epigenetics in controlling the expression of AOX genes [28–30], few data are available showing the effect of gestational diet on this system.</p><p>A maternal HF diet in nonhuman primates was shown to increase hepatic oxidative stress in the offspring [31], and because the PON1 enzyme is synthesized in the liver and provides protection against system-wide oxidative damage [32], the current study aimed to determine its transcription, as well as the transcription of several AOX genes, in livers of both male and female offspring in response to a maternal HF diet. Furthermore, we wanted to investigate the fetal hepatic epigenome by focusing on Pon1 as a model for gender-specific histone modifications that occur in response to a maternal HF diet. In order to accomplish this, the present study utilized an obesity-resistant (OR) rat model without the concurrent development of gestational obesity. Our results demonstrate that the in utero exposure to an HF diet affects fetal hepatic oxidative balance and has a gender-specific effect on the expression of AOX genes in association with epigenetic modifications within the tested gene.</p><!><p>Timed-pregnant female rats were obtained from Charles River (Wilmington, MA, USA) on embryonic day 2. The OR strain [Crl:OR(CD)] in the current experiment was developed from an outbred line of Crl:CD(SD) rats. This model does not become obese when fed HF diets. Dams were separated into two dietary treatments: five control (C; 64% carbohydrates, 20% protein, 16% fat), and five high fat (HF; 35% carbohydrates, 20% protein, 45% fat), where the HF diet had approximately 3.3, 3.17 and 1.8 times more saturated, monounsaturated and polyunsaturated fat, respectively, per kilocalorie consumed when compared to the C diet [27]. Pregnant animals were fed these diets ad libitum until embryonic day 20, when they were fasted overnight and underwent cesarean delivery to collect livers from offspring. All fetal liver tissue samples were immediately stored in liquid nitrogen and kept for further analysis. Fetal weights and gender were also recorded at this time.</p><!><p>Frozen liver samples (100 mg) were ground using a mortar and pestle with liquid nitrogen and mixed with 0.3 ml saline (0.9% w/v NaCl) as previously reported [33]. Homogenized samples were quickly frozen in liquid nitrogen and kept in −70 °C. The samples were quickly thawed at 37 °C and diluted five times in saline to 1.5 ml. Twenty microliters of the diluted samples was incubated with 20 μl 1% deoxycholate at 37 °C for 5 min, and 10 μl of the samples was used to analyze either liver TG or NEFA. Hepatic TG was analyzed via the Thermo Infinity Triglycerides Liquid Stable Reagent (Thermo Fisher Scientific, Rockford, IL, USA) following company protocol and using a commercially available standard reference kit (Verichem Laboratories, Providence, RI). Hepatic NEFA concentration was determined using a commercially available kit (HR-2 Series, Wako Diagnostics, Richmond, VA, USA). TBARS assay (catalog no. 10009055, Cayman Chemical, Ann Arbor, MI, USA) was performed per company protocol to determine the concentration of malondialdehyde (MDA), a product of lipid peroxidation, within neonatal hepatic tissues. TG, NEFA and TBARS data are presented as the amount of TG, NEFA or MDA (in milligrams) per gram of liver.</p><!><p>Livers from 10 offspring (5 male and 5 female) from each dietary group were randomly chosen for sampling, with all litters being represented. Total RNA was isolated using the GenElute Mammalian Total RNA Miniprep Kit (Sigma-Aldrich, St Louis, MO, USA), treated with DNase I to eliminate any DNA contamination and quantified using Nano Drop ND-1000 Spectophotometer while ensuring that the ratio of 260/280 was >1.9. cDNA was synthesized from 2 μg of RNA in a 20-μl reaction volume using the High Capacity cDNA Reverse Transcription Kit with random primers (Applied Biosystems, Foster City, CA, USA) and a thermal cycler (Applied Biosystems 2700, Foster City, CA, USA), with the following program: 25 °C for 10 min, 37 °C for 120 min, 85 °C for 5 s and a 4 °C hold. RT-PCR was performed using 25 ng cDNA as the template, SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA, USA), and 5 μmol/L of each forward and reverse primer (Table 1) in the 7300 Real-Time PCR System (Applied Biosystem, Foster City, CA, USA), with the following program: 95 °C for 10 min, 95 °C for 15 s, 60 °C for 1 min, 95 °C for 15 s, 55 °C for 1 min and 95 °C for 15 s, with 40 cycles of steps 2 and 3. A serial dilution was used to create a standard curve for quantification, a dissociation curve was analyzed following each reaction, and a no-template control was included with all reactions to ensure that no additional products were synthesized during the PCR reaction. mRNA level of ribosomal protein L7a was utilized as the internal control.</p><!><p>The ChIP assay was previously reported in great detail and is therefore not described here [27]. As previously stated, livers from 10 offspring (5 male and 5 female) from each dietary group were randomly chosen for sampling and were the same samples as those used for mRNA analysis. Antibody information is listed in Table 2.</p><!><p>Results are reported as means±S.E.M. Neonatal liver TG, TBARS and gene expression data were analyzed using two-way analysis of variance (ANOVA) with interactions (SAS software, Cary, NC, USA) with diet and gender as main effects. When an interaction was found between factors (P<.05), the post hoc Tukey test was used to identify specific differences within diet or within gender. For hepatic ChIP analysis in offspring (n=10, 5 male and 5 female), each gender was analyzed individually to determine the effect of diet on each antibody using Student's t test, while the effect of antibody between IgG and H3K9Me3 was determined using two-way ANOVA with antibody and gender as main effects.</p><!><p>We certify that all applicable institutional and governmental regulations regarding the ethical use of animals were followed during this research (University of Illinois Institutional Animal Care and Use Committee approval #09112).</p><!><p>The following has been previously reported [27]; therefore, the data are not shown here. Briefly, gestational food intake and body weight did not differ between C and HF dams. Dams fed the HF diet gained significantly more weight than C, possibly because the offspring of these dams were heavier. However, when fetal and placental weights were subtracted from this weight to determine actual mass gained during gestation, the two groups did not differ.</p><!><p>As previously reported [27], there were no differences in litter size between the two dietary groups, and at the time of cesarean delivery on gestational day 21, the offspring of dams fed the HF diet were significantly heavier when compared to the offspring of dams that were fed the C diet during gestation. At birth, the offspring of HF-fed dams had significantly higher (P<.05) hepatic TBARS content when compared to the offspring of C dams, and there was no effect of gender or an interaction between diet and gender (Table 3). There was no effect of diet or gender on hepatic TG content in offspring at birth; however, there was a significant interaction between the factors (P<.05), which was due to the fact that while male offspring had a significant (P<.05) increase in hepatic TG in response to a maternal HF diet, female offspring did not respond to diet (Table 3). Hepatic NEFA content was significantly lower (P<.05) in offspring of HF-fed dams when compared to C, and there was no effect of gender or interaction between diet and gender (Table 3). Acox1 mRNA content was measured as a marker of fatty acid oxidation and was significantly higher in female than in male offspring (P<.05), and there was a significant interaction between diet and gender, which, similarly to hepatic TG content, was due to the fact that while male offspring had a significant (P<.05) increase in hepatic Acox1 in response to a maternal HF diet, female offspring did not respond to diet (Table 3).</p><!><p>Hepatic Pon1 mRNA content was significantly higher (P<.005) in offspring of HF-fed dams when compared to C offspring, and there was a significant (P<.05) interaction between the factors, which was due to the fact that male HF offspring had significantly (P<.005) higher Pon1 expression than C offspring, while females did not respond to diet. Additionally, HF female offspring had significantly (P<.05) lower Pon1 expression than HF males (Table 4). Offspring of HF-fed dams had significantly higher (P<.05) Pon2 mRNA content, with no effect of gender and no interaction between diet and gender (Table 4). Similar to Pon1 expression, hepatic Pon3 mRNA content was significantly higher (P<.005) in offspring of HF-fed dams when compared to C offspring, and there was a significant (P<.05) interaction between the factors, which was due to the fact that male HF offspring had significantly (P<.005) higher Pon3 expression than C offspring, while females did not respond to diet. Additionally, C female offspring had significantly (P<.05) higher Pon3 expression than C males (Table 4). There was no effect of diet or gender, or an interaction between the two factors on hepatic Sod1 expression (Table 4). However, Sod2 mRNA was significantly (P<.05) higher in female than male offspring and significantly (P<.005) higher in offspring of HF-fed dams when compared to C offspring (Table 4).</p><!><p>The ChIP assay was performed to investigate the association of modified histones to Pon1 gene expression. Normal rabbit IgG antibody was used as the negative control, indicating nonspecific binding of protein–DNA. Proteins were considered negative for binding if the resulting value was equal to or less than the IgG value. At the Pon1 promoter, H3Ac was not significantly different between C and HF offspring in males (Fig. 1A) but slightly decreased in females (P=.11, Fig. 1B). At the coding region, H3Ac was also not significantly different between C and HF offspring in males (Fig. 1C) but slightly decreased in females (P=.11, Fig. 1D). At the promoter, H4Ac was significantly higher in both male (P<.05, Fig. 1A) and female (P<.01, Fig. 1B) HF offspring when compared to C offspring; whereas there was no significant change in H4Ac within the Pon1 coding region in either males (Fig. 1C) or females (Fig. 1D). At the promoter, H3K4Me2 was also significantly higher in HF male (P<.01, Fig. 1A) and female (P<.01, Fig. 1B) offspring when compared to C, and there was no significant change in H3K4Me2 within the Pon1 coding region in either male (Fig. 1C) or female (Fig. 1D) offspring. The L7a ribosomal protein gene was used as a control to normalize all mRNA data because its expression was not affected by either diet or gender, and histone modifications within the L7a promoter are also presented to demonstrate any histone modifications occurring within the promoter of a gene whose expression was not affected by diet or gender. No histone modifications corresponding to transcription activation were observed in response to diet within the L7a promoter in male (Fig. 1E) or female (Fig. 1F) offspring.</p><!><p>Methylation of lysine 9 of histone H3 (H3K9Me3) is a marker of condensed, inactive chromatin. There was no significant difference between H3K9Me3 and IgG in male offspring at the promoter (Fig. 2A) and coding region (Fig. 2C) of the Pon1 gene and at the promoter (Fig. 2E) of the L7a gene. This indicates that lysine 9 on histone H3 was not methylated in male offspring. However, in female offspring, the differences between H3K9Me3 and IgG at the promoter (Fig. 2B) and coding region (Fig. 2D) of the Pon1 gene and at the promoter (Fig. 2F) of the L7a gene were all significant (P<.05). This indicates that histone H3 was modified by methylation of lysine 9 in female offspring to induce a more inactive chromatin state. Although H3K9Me3 was significantly reduced in livers of HF female offspring when compared to C in a non-gene-specific manner (Fig. 2B, D, F), it was still significantly higher than nonspecific IgG binding, indicating a generally condensed chromatin state in females.</p><!><p>Maternal nutrition is a key regulator of fetal growth and development, and our study is the first, to our knowledge, to suggest that a maternal HF diet, independent of gestational obesity, increases fetal hepatic oxidative stress, and the AOX system that responds to this stimulus appears to be regulated epigenetically. In rodents, these phenotypic gender differences have been shown to also exist in adulthood [34,35], and our findings provide the first potential mechanistic explanation for these differences. It is likely that the early hit of oxidative stress and the premature activation of the AOX system may influence the efficacy of this system in the future, thereby regulating the adult offspring's response to stress events.</p><p>Previous reports in rats have suggested that males may be more susceptible to the deleterious effects of a maternal HF diet, which include their tendency to have increased oxidative stress when compared with females [36]. While both males and females in the present study had increased hepatic TBARS, a marker of lipid peroxidation, the differential genetic response may be an indicator of innate gender differences in the mechanisms that govern both the induction and response to oxidative stress. Both male and female offspring had decreased hepatic NEFA content, which may result from an increase in hepatic fatty acid oxidation. We previously reported that gluconeogenesis was increased in livers of these neonates [27], which agrees with the current data, since an increase in gluconeogenesis and fatty acid oxidation typically occur concurrently. However, while fatty acids were decreased in both genders in response to a maternal HF diet, Acox1, which encodes the enzyme responsible for the regulatory step of palmitoyl fatty acid oxidation, was approximately twofold increased in male but not female offspring in response to maternal HF feeding. In males, hepatic TGs were also significantly increased, but this was not observed in females. An increase in TG synthesis occurs in response to excess NEFA availability, suggesting that males accumulate and potentially uptake more lipids in response to a maternal HF diet when compared to their basal state than do females, which may also explain the increase of Acox1 expression as an attempt to deal with this excess. Therefore, it appears that the differential regulation of the AOX system occurs in direct response to the inherent differences in lipid metabolism between males and females exposed to a maternal HF diet.</p><p>The PON1 enzyme is synthesized in the liver and acts as a potent antioxidant in circulation, especially in atherosclerotic plaques, where lipid peroxidation is increased. Polymorphisms in the human PON1 gene are associated with an increase in various diseases, including diabetes, heart disease and hyperlipidemia [37–40]. Gender differences in the activity of the PON enzymes in response to HF feeding have been previously demonstrated. A study in adult animals fed an HF (cafeteria diet, 64% fat) diet showed that males and females had gender-specific responses in PON1 serum activity, with no response in the male and a significant decrease by HF feeding in females, bringing the level of PON1 activity to that of males [41]. However, there was no effect of the HF diet on the expression of Pon1 in the liver in either males or females. In adults, PON1's AOX function is most often associated with HDL in circulation where it is thought to contribute to the prevention of atherosclerosis development [42]. However, in the neonate, circulating HDL has been shown to be quite low [43], suggesting that the increase in hepatic Pon1 gene expression in the current study was likely a local response to oxidative stress and an appropriate marker of enzyme activity. In addition to Pon1, we observed that the expression of the other two PONs (Pon2 and Pon3) as well as Sod2 (MnSOD) was also elevated in response to a maternal HF diet, but only Pon3 followed Pon1's expression pattern – being increased in males but not females in response to a maternal HF diet.</p><p>Nonalcoholic fatty liver disease (NAFLD) and cirrhosis are typically accompanied by increased hepatic oxidative stress [44], and NAFLD has been associated with decreased AOX capacity and a decrease in the expression of AOX genes [45]. Our previous study showed that adult offspring of HF-fed mothers had decreased hepatic AOX gene expression, which is generally though to signal decreased antioxidant defense capacity [4]. These differences were observed in offspring consuming a control postweaning diet, suggesting that the in utero environment has the potential to program adult phenotypes, regardless of the postnatal diet. Because the current study utilized OR rats to isolate the effects of fat intake from those of maternal adiposity typically observed in animal models of diet-induced obesity, it is possible that this current study and our previous study using Sprague–Dawley rats actually model different pathologies. However, it is also possible that the overexpression of the AOX system early on increases the basal expression of these genes and disrupts the feedback mechanisms that regulate their transcription in adulthood. Further studies will be critical for determining the exact role that maternal diet has in the oxidative balance in offspring and whether altered PON expression in livers of neonates corresponds to its decreased activity in the plasma of adult animals.</p><p>Pon1 was used in the present study as a model for studying histone modifications in a gene that had a gender-specific pattern of transcription in response to a maternal HF diet. Additionally, unlike the other genes studied, the PON1 enzyme is a known extrahepatic AOX, so its expression in liver impacts systemic AOX balance. As previously discussed, the essential role of PON1 during AOX defense has been confirmed; the regulation of the PON1 gene remains poorly understood. Our histone modification analysis of the Pon1 promoter and coding region presents a novel mechanism for the in utero activation of this gene by a maternal HF diet. Both the male and female promoters showed a histone code that correlates to the active transcription of genes. The acetylation of histone H4 and the dimethylation of histone H3 at lysine residue 4 are associated with active transcription, and these were both significantly increased by a maternal HF diet in the Pon1 promoter of male and female offspring. However, the Pon1 gene was only significantly increased by a maternal HF diet in male livers but not female, which may be due to the presence of H3K9Me3, a known potent inhibitor of transcription, within the female but not male promoter as well as coding regions. We observed a similar pattern of H3K9Me3 modification within the L7a promoter, indicating a female-specific modification. However, unlike the promoter and coding regions of Pon1, there were no transcription activation-associated modifications present within L7a in response to diet. These complex histone interactions and patterns are likely responsible for differences in the expression of these genes [46]. Future in-depth analyses of gene-wide histone modifications, as well as studies of other epigenetic mechanisms, will be essential for determining the precise regulation of Pon1, as well as of other AOX genes in response to maternal HF diet.</p><p>In the current study, we focused on the epigenetic regulation of the Pon1 gene to demonstrate the consequence of a maternal HF diet for the histone code of a critical AOX gene. However, we observed that other AOX genes were also increased in the livers of offspring in response to a maternal HF diet, so it is reasonable to presume that these genes are also regulated at the epigenetic level in response to the in utero environment, and in some of these genes, the changes may be gender specific. More studies are needed to fully unravel the role of epigenetics in the regulation of the AOX system, but our preliminary results suggest that the in utero environment is a potent regulator of AOX genes, which can potentially result in the dysregulation of the AOX system in adult offspring. These outcomes could result in altered handling of prooxidants in adult offspring whose mothers consumed an HF diet and could potentially explain the systemic differences often observed between the male and female response to oxidative stress.</p>

PubMed Author Manuscript

Mixed-Carbene Cyclometalated Iridium Complexes with Saturated Blue Luminescence

Cyclometalated iridium complexes have emerged as top-performing emitters in organic lightemitting diodes (OLEDs) and other optoelectronic devices. A persistent challenge has been the development of cyclometalated iridium complexes with deep blue luminescence that have the requisite color purity, efficiency, and stability to function in devices. In this work we report a new class of cyclometalated iridium complexes with saturated blue luminescence. These complexes have the general structure Ir(C^C: NHC )2(C^C: ADC ), where C^C: NHC is an Nheterocyclic carbene (NHC) derived cyclometalating ligand and C^C: ADC is a different type of cyclometalating ligand featuring an acyclic diaminocarbene (ADC). The complexes are prepared by a cascade reaction that involves nucleophilic addition of propylamine to an isocyanide precursor followed by base-assisted cyclometalation of the ADC intermediate. All three emit deep blue light with good quantum efficiencies and color profiles very close to the ideal primary blue standards for color displays.

mixed-carbene_cyclometalated_iridium_complexes_with_saturated_blue_luminescence

<!>Results and Discussion<!>Conclusions<!>Methods

<p>Work in the late 1990's showed that showed that phosphorescent emitters can lead to higher efficiencies for organic light-emitting diodes (OLEDs) when compared to their fluorescent counterparts, on account of their ability to utilize both singlet and triplet excitons. 1,2 Since that time, cyclometalated iridium complexes have emerged as the champion phosphors for OLEDs, with many examples of top-performing devices doped with such compounds. [3][4][5][6][7] Advantages of cyclometalated iridium complexes include their high photoluminescence quantum yields, which directly factor into overall device efficiency, their short lifetimes which are necessary for sharp color displays, and their color tunability which allows facile preparation of devices covering the entire visible spectrum. The largest remaining technological challenge in OLED research is the lack of deep blue phosphorescent compounds with adequate performance metrics -i.e. color purity, quantum efficiency, and stability -to function in commercial devices. The physical origins of this challenge are summarized in Figure 1. Cyclometalated iridium complexes luminesce from excited states that typically involve contributions from triplet ligand-centered ( 3 LC) states localized on the cyclometalating ligand and triplet metal-to-ligand charge transfer ( 3 MLCT) states, which mix through configuration interaction in the low-energy T1 state. 4 The lowest-energy ligand-field or metal-centered state, which involves transitions between Ircentered d orbitals, is also a triplet state ( 3 MC). 8 One of the reasons bis-cyclometalated iridium complexes are so successful as triplet emitters is that the 3 MC state is typically much higher in energy than the T1 state and thus cannot be populated. However, in deep-blue-emitting compounds with high-energy T1 states, the 3 MC state is thermally accessible, and population of this nonradiative state decreases quantum yield and also promotes electrons into metal-ligand σ* orbitals, leading to ligand-loss degradation pathways that limit device longevity. 9 As is also shown in Figure 1, incorporation of strong σ-donor ligand sets can destabilize the deleterious 3 MC state, allowing efficient luminescence from T1 and greater photostability. As a result, N-heterocyclic carbenes (NHCs) have emerged as popular design elements in blueemitting cyclometalated iridium complexes, and several recent breakthroughs in the design of blue phosphorescent devices involve NHC-containing C^C: cyclometalating ligands, chelated to iridium through the NHC and a phenyl group. [10][11][12][13][14][15] We recognized the possibility that acyclic diaminocarbenes (ADCs), which are known to be even stronger σ-donors than NHCs on account of the greater 2p character in their σ orbital, [16][17][18][19] could potentially destabilize 3 MC states to an even greater extent than is possible with NHCs. We have shown that ADCs can be installed onto cyclometalated iridium complexes via on-complex, nucleophilic addition to coordinated isocyanides. In this way, we introduced two new classes of ADC-containing cyclometalated iridium complexes, one with chelating bis(ADC) "Chugaev-type" ancillary ligands accessed via chelative insertion of hydrazine into cis-oriented isocyanides, 20 and a second class with a cyclometalated ADC ancillary ligand, prepared via a cascade reaction involving nucleophilic addition of an amine to an isocyanide followed by C-H activation. 21 Versions of these complexes exhibit intense blue luminescence when immobilized in polymer films, [20][21][22] with quantum efficiencies as high as 79% for a cyclometalated ADC complex. 21 However, the biggest limitation of these first-generation compounds, which will preclude any practical device applications, is their rather poor color purity, with significant sky blue to blue-green coloration in each case. In addition, the blue-emitting ADC-containing complexes we prepared all use the C^N cyclometalating ligand 2-(2,4-difluorophenyl)pyridine (F2ppy), which is known to degrade in OLED devices via ligand defluorination. 9 In the present work, we address these two limitations and improve upon our previous designs, introducing three new compounds which are devoid of sp 2 C-F bonds and give rise to high-purity blue photoluminescence. These mixedcarbene compounds have the general formula Ir(C^C: NHC )2(C^C: ADC ), where C^C: NHC is an Nheterocyclic carbene (NHC) derived cyclometalating ligand and C^C: ADC an ADC-containing cyclometalating ligand. CIE coordinates 23 for these blue-emitting compounds are very close to National Television System Committee and International Electrotechnical Commission standards for color displays, making these compounds excellent candidates for incorporation into OLEDs.</p><!><p>Synthesis and structural characterization. Figure 2 outlines the synthesis of the compounds described in this work. Preparation of the mixed-carbene complexes begins with precursors of the type [Ir(C^C: NHC )2(μ-Cl)2] (1a-c); the versions 1a with C^C: NHC = 1-phenyl-3-methyl imidazole-2-ylidine (pmi) and 1b with with C^C: NHC = 1-phenyl-3-methyl benzoimidazole-2ylidine (pmb) have been previously described, 14 whereas 1c is new to this work. Treatment of dimers 1a-c with 4-trifluoromethylphenyl isocyanide forms mononuclear complexes of the type Ir(C^C: NHC )2(CN-p-C6H4CF3)(Cl) (2a-c), which then are treated with excess propylamine to form the final ADC-containing complexes 3a-c. Complexes 3a-c form via nucleophilic addition of the amine to the coordinated isocyanide, followed by base-assisted cyclometalation to form the final neutral, tris-chelated heteroleptic products. The identity and purity of the final products was ascertained by 1 H, 19 F, and 13 Further confirmation of the structures of 3a-c comes from single-crystal X-ray diffraction studies. Whereas the structure of 3c could be solved but refined poorly, structures of 3a and 3b refined reasonably well and are included here in Figure 2. X-ray crystallography confirms the pseudo-octahedral coordination geometry and trans arrangement of the NHC ligands and also verifies the chelated nature of the ADC ancillary ligand. Bond metrics of the ADC ligand, in particular the N-C bond distances and the carbene N-C-N bond angles, are very similar to the bond metrics observed in previously described complexes from our group with the same ancillary ligand partnered with pyridine-or thiazole-based C^N cyclometalating ligands. 21 There also not any substantial differences between Ir-C(NHC) and Ir-C(ADC) distances in these structures, with the latter distances falling in the range spanned by the Ir-C(NHC) distances. Photoluminescence. Figure 4 overlays the room-temperature photoluminescence spectra of complexes 3a-c, whereas Table 1 summarizes the data. The complex Ir(pmi)2(C^C: ADC ) (3a) is negligibly fluorescent in fluid solution, but when immobilized in poly(methyl methacrylate) (PMMA) at 2 wt% deep blue luminescence is observed, with a wavelength of maximum emission of 418 nm and a photoluminescence quantum yield (ΦPL) of 0.13. Replacing pmi with pmb in complex 3 has no impact on the observed wavelength of emission, but in this case the quantum efficiency is much higher. In CH2Cl2 solution complex 3b weakly emits, with a quantum yield of 0.013, and in PMMA the spectrum narrows and the quantum yield increases dramatically to 0.31. Finally, in complex 3c, where the pyridyl-substituted cyclometalating ligand pmp is used, the photoluminescence bathochromically shifts and is environmentally dependent. In solution complex 3c emits in the green region of the spectrum, with λ = 511 nm and a quantum yield of 0.39. The photoluminescence blue shifts significantly in PMMA film, with λ = 459 nm, and the quantum yield increases to 0.48. The photoluminescence lifetimes also vary across the series. When measured in PMMA film, the lifetimes decrease in the order 3a (τ = 6.1 μs), 3b (τ = 1.8 μs), and 3c (τ = 0.85 μs). Emission spectra were also recorded at 77 K in rigid solvent glass, and under these conditions the luminescence maximum displays a significant rigidochromic blue-shift in each case, signifying substantial charge-transfer character in each case. The spectra of 3a and 3b are much sharper with pronounced vibronic structure at 77 K, consistent with significant ligand-centered 3 ππ* excited-state character as well, whereas the spectrum of 3c is devoid of vibronic structure even at 77 K, indicating a T1 state that is mainly a metal-to-ligand charge transfer ( 3 MLCT) state. Using the thin-film photoluminescence data, CIE coordinates 23 were determined and are summarized in Figure 5. All three compounds fall in the blue region of the spectrum. The redshifted emission in complex 3c engendered by the pyridyl-substituted NHC results in significant sky-blue coloration with CIE (x, y) coordinates of (0.16, 0.18). In contrast, the color profiles for 3a and 3b represent pure blue luminescence, with CIE coordinates of (0.16, 0.10) for 3b and (0.16, 0.07) for 3a. These coordinates are nearly perfect matches for the blue color specified by the most widely used industry standards, including the IEC sRGB standard (0.15, 0.06), and the primary blue standards proscribed by NTSC (0.14, 0.08) and SMPTE C (0.155, 0.07). The color profiles of 3a-c represent substantial improvements over our first two classes of cyclometalated iridium ADC complexes, all of which exhibited luminescence with significant sky blue to bluegreen coloration rendering them unsuitable for device applications. 20,21 To further contextualize the photoluminescence properties and the CIE coordinates, we can compare complexes 3a-c to homoleptic mer-Ir(C^C: NHC )3 complexes which use the same three NHC ligands. Immobilized in a thin film, compound 3c exhibits nearly identical photoluminescence to that of mer-Ir(pmp)3, 12 with the same featureless profile and nearly identical maximum wavelengths and CIE coordinates (CIE (x, y) = (0.16, 0.16) for mer-Ir(pmp)3). The photoluminescence attributes of 3a and 3b are also quite similar to their respective mer-Ir(C^C: NHC )3 (C^C: NHC = pmi, pmb) analogues. These compounds also emit with maxima near 400 nm, and although precise quantum yield values in polymer film were not reported, they were estimated to be greater than 0.1-0.2. 11 The photoluminescence quantum yields of 3a (Φ = 0.13) and 3b (Φ = 0.31) are in the same range or greater, suggesting the photoluminescence efficiencies of 3a and 3b are very similar to or better than the homoleptic analogues. In addition, CIE coordinates for devices made from mer-Ir(pmb)3 were found to be (0.17, 0.06) and (0.17, 0.08), 10 very similar to the color profile observed for the photoluminescence of 3a and 3b. One other point of comparison are homoleptic mer-Ir(C^C: NHC )3 complexes with CF3-substituted pmi ligands. 13 The photluminescence quantum yields of these compounds were measured in PMMA films, with one isomer having ΦPL = 0.14 and another with ΦPL = 0.47, similar values to those of 3a and 3b. CIE coordinates for devices made from these CF3-substituted Ir(pmi)3 analogues are also nearly identical to those observed for 3a and 3b. Taken together, these comparisons indicate that installing cyclometalated ADC ancillary ligands preserves the favorable photluminescence characteristics of the carbene-based iridium cyclometalates that have been used in some of the top-performing blue devices. 10,12,13 In addition, installation of the ADC moeity offers some potential advantages over homoleptic NHC-based phosphors. The stronger σ-donating nature of ADCs is expected to destabilize the 3 MC states in 3a-c, which may lead to greater stability of these compounds in OLEDs. In addition, the alkyl group derived from the primary amine that undergoes nucleophilic addition with the isocyanide (see Figure 2), which is a propyl group in 3a-c, can in principle be substituted with many other nonpolar or polar sterically unhindered groups. This ability to easily alter that substituent may provide an avenue for controlling the alignment of the phosphor when deposited into an OLED device, another important criterion for optimizing device performance. 24 Both of these potential advantages will be scrutinized in future work when 3a-c and related analogues are tested as dopants for blue OLEDs.</p><!><p>In this work we address the long-standing challenge of designing effective blue phosphors for OLED applications. The compounds we describe are heteroleptic tris-cyclometalated iridium complexes with mixed-carbene ligation, prepared from isocyanide precursors by a nucleophilic addition/cyclometalation cascade sequence. The photoluminescence attributes of these compounds are admirable. In PMMA thin films, deep blue luminescence is observed for two members of the series, with CIE coordinates that match well with industry standards for blue phosphors. The photoluminescence quantum yields are good and the lifetimes are in the microsecond range, also important criteria for device applications. The acyclic diamino carbene (ADC) ancillary ligand is expected to destabilize deactivating metal-centered excited states, and the ligand sets are devoid of sp 2 C-F bonds, both of which are anticipated to improve the long-term stability of these compounds in OLEDs.</p><!><p>General synthetic procedure for 3a-c. In a typical reaction, the precursor Ir(C^C: NHC )2(CN-p-C6H4CF3)(Cl) (2a-c) was dissolved in CH2Cl2, and excess propylamine was added to the solution.</p><p>After stirring for 4 d at room temperature, the volatiles were removed under vacuum and the crude product was purified by silica gel column chromatography, eluting with CH2Cl2, followed by recrystallization from CH2Cl2/hexane.</p><p>Physical Characterization. Steady-state emission and excitation spectra were recorded using a Horiba FluoroMax-4 spectrofluorometer. Samples for solution emission were prepared in a glovebox and housed in 1 cm quartz cuvettes with septum-sealed screw caps. PMMA thin-film samples were prepared in the glovebox by dissolving PMMA (98 mg, 35 kDa) in dichloromethane (1.0 mL), and then adding the respective iridium complex 3a-c (2 mg, 2 wt. %). The solution was drop-cast onto a quartz slide and kept under nitrogen until immediately before measurement.</p><p>Emission quantum yields in solution were determined relative to a standard of quinine sulfate in 0.05 M sulfuric acid, which has a reported fluorescence quantum yield (ΦF) of 0.52. 51 The absolute quantum yields of complexes doped into PMMA thin films were recorded using a Spectraloncoated integrating sphere (3.2 inch diameter, Horiba) attached to the Fluoromax-4 fluorometer.</p><p>Phosphorescence lifetimes were measured on a Horiba DeltaFlex Lifetime System, using 330-nm pulsed diode excitation sources.</p>

ChemRxiv

Hematopoietic Stem Cell Mobilization and Homing after Transplantation: The Role of MMP-2, MMP-9, and MT1-MMP

Hematopoietic stem/progenitor cells (HSPCs) are used in clinical transplantation to restore hematopoietic function. Here we review the role of the soluble matrix metalloproteinases MMP-2 and MMP-9, and membrane type (MT)1-MMP in modulating processes critical to successful transplantation of HSPC, such as mobilization and homing. Growth factors and cytokines which are employed as mobilizing agents upregulate MMP-2 and MMP-9. Recently we demonstrated that MT1-MMP enhances HSPC migration across reconstituted basement membrane, activates proMMP-2, and contributes to a highly proteolytic bone marrow microenvironment that facilitates egress of HSPC. On the other hand, we reported that molecules secreted during HSPC mobilization and collection, such as hyaluronic acid and thrombin, increase MT1-MMP expression in cord blood HSPC and enhance (prime) their homing-related responses. We suggest that modulation of MMP-2, MMP-9, and MT1-MMP expression has potential for development of new therapies for more efficient mobilization, homing, and engraftment of HSPC, which could lead to improved transplantation outcomes.

hematopoietic_stem_cell_mobilization_and_homing_after_transplantation:_the_role_of_mmp-2,_mmp-9,_and

1. Hematopoietic Stem/Progenitor Cells (HSPCs) Transplantation<!>2. Mobilization of HSPC<!>3. Homing of HSPC<!>4. Matrix Metalloproteinases<!>5. MMPs in HSPC Mobilization<!>6. MMPs in HSPC Homing<!>7. Therapeutic Strategies to Improve Transplantation Outcomes

<p>HSPC transplantation is a clinical procedure in which HSPCs capable of reconstituting normal bone marrow (BM) function are administered intravenously to a patient who has undergone preparative regimens including chemotherapy and/or irradiation. Approximately 60,000 autologous and allogeneic HSPC transplants are performed annually worldwide to treat various cancers and diseases of the blood and immune system [1]. During steady-state homeostasis, approximately 0.06% of BM HSPCs circulate continuously in the peripheral blood (PB) [2], but this number can be increased with the use of chemotherapeutic drugs (e.g., cyclophosphamide) and/or growth factors and cytokines (e.g., granulocyte colony-stimulating factor (G-CSF)) that "mobilize" HSPC from BM into the PB [3]. Currently mobilized (m)PB HSPC collection has almost replaced BM harvest for autologous and most allogeneic transplantations because it is relatively easy to collect by apheresis in an outpatient setting and because engraftment after transplantation is faster. G-CSF is the most commonly used mobilizing agent in the clinic with regimens using 10 μg/kg/day for five days when used alone, or 10 to 14 days when used in combination with chemotherapeutic agents [3]. Randomized trials of mPB transplantation have shown that neutrophil and platelet engraftment generally occurs at a median of 9–14 days compared to 21 days with BM [4]. This has been attributed to the higher number of HSPC collected and transplanted. A limitation of mPB HSPC transplantation is that patients' responses to G-CSF vary: 5–10% of allogeneic normal donors mobilize poorly and up to 40% of autologous patients fail to mobilize depending on their disease and the intensity/number of prior chemotherapy regimens [5]. Hence, elucidation of the molecular mechanisms of HSPC mobilization could lead to more efficient mobilizing agents and development of better protocols.</p><p>An alternative source of HSPC is cord blood (CB) obtained at the time of childbirth, after the umbilical cord has been detached from the newborn. Since the first CB transplant in 1988, more than 30,000 CB transplants have been performed worldwide in pediatric and adult patients [1, 6]. CB has several advantages over BM and mPB as source of HSPC for transplantation. CB contains lower numbers of more immature, immunocompetent T cells and thus requires less stringent HLA matching; this means that a mismatch at one or two loci can be tolerated without significant increase in graft versus host disease (GvHD) or decrease in overall survival [6, 7]. However, the main limitation of CB transplantation is the low CD34+ cell dose available in one CB unit which is generally insufficient to support engraftment in adult patients. Retrospective analyses of outcomes of CB and BM transplantation in adults have reported delayed neutrophil engraftment: 27 days with CB versus 18 days with BM and platelet engraftment: 60 days with CB versus 29 days with BM [7]. Currently efforts are being made to increase CB cell dose in order to speed up engraftment and hematopoietic recovery. Strategies to use more than one CB unit [7] or to expand CB CD34+ cells ex vivo [8], however have met with limited success. A more comprehensive knowledge of CB HSPC biology and the mechanisms of their homing is expected to improve CB transplantation outcomes.</p><!><p>During homeostasis, continuous traffic of HSPC between the BM and PB contributes to normal hematopoiesis. The ability to enhance these physiological processes and, for example, enforce egress (mobilization) of HSPC from the BM to circulation has been invaluable in clinical transplantation [9]. Several cytokines (G-CSF, granulocyte macrophage-CSF, Flt-3 ligand, interleukin (IL)-8, and stem cell factor (SCF)/kit-ligand) and chemokines (stromal cell-derived factor (SDF)-1 and GROβ) can trigger mobilization in varying degrees [3, 10]. We recently demonstrated that G-CSF also increases plasma hepatocyte growth factor (HGF) levels in mobilized patients and expression of its receptor c-Met in HSPC and myeloid cells, suggesting that GCSF-mediated HSPC mobilization occurs in part through the HGF/c-Met axis [11].</p><p>Another important axis that influences mobilization is the SDF-1/CXCR4 axis. This axis is essential for retention of HSPC in the BM, and perturbation of the SDF-1 gradient in the BM as well as a decrease in the responsiveness of HSPC to SDF-1 leads to mobilization [12].</p><p>The proteases carboxypeptidase M (CPM) [13] and CD26 [14] cleave the C-terminus of SDF-1 resulting in attenuated chemotactic responses of HSPC; moreover, G-CSF-induced mobilization is impaired in CD26-deficient mice [14]. Desensitization of CXCR4 by the urokinase-mediated plasminogen activation system during G-CSF-induced mobilization has been demonstrated [15]. In murine studies, SDF-1 concentration in the BM decreased following G-CSF administration and correlated with mobilization, suggesting that a physiological drop in SDF-1 level in the BM is a critical step in HSPC mobilization [12, 16]. A decrease in BM SDF-1 levels has been reported to coincide with a peak of proteolytic activity of neutrophil elastase (NE), cathepsin G (CG), and matrix metalloproteinase (MMP)-9 [17]. Both SDF-1 and CXCR4 are targets of degradation by NE, CG, MMP-9, MMP-2, and membrane type (MT)1-MMP [18–20] and can be inactivated by proteolytic cleavage. The roles of these MMPs will be discussed in detail in subsequent sections.</p><p>Recent studies have demonstrated that thrombolytic agents such as microplasmin, tenecteplase, and recombinant tissue plasminogen activator enhance G-CSF-induced mobilization in murine models [21]. Adhesion molecules also play an important role in the retention of HSPC in the BM. Very late antigen (VLA)-4 is expressed by HSPC and its ligand, vascular cell adhesion molecule (VCAM)-1, is constitutively expressed by endothelial and stromal cells [22]. Disruption of the VCAM-1/VLA-4 axis with a small molecule inhibitor of VLA-4 resulted in the mobilization of more HSC over basal levels [23]. CD44, a polymorphic integral membrane glycoprotein, binds to several extracellular matrix (ECM) components such as hyaluronic acid (HA), fibronectin, and collagen and mice treated with anti-CD44 antibody or lacking CD44 exhibit impaired mobilization in response to G-CSF [24].</p><p>Components of innate immunity also participate in G-CSF-induced HSPC mobilization [25]. During G-CSF-induced mobilization, the complement cascade (CC) is activated by the classical pathway. However, in the early steps of the CC, C1q, C3, and its cleavage fragments C3a and desArgC3a increase retention of HSPC in BM while in the later steps of the CC, C5, and its cleavage fragments C5a and desArgC5a enhance HSPC mobilization and their egress into PB [26–28].</p><!><p>The success of clinical HSPC transplantation relies on the inherent ability of transplanted HSPC to home efficiently to the BM niche and engraft. Interactions between HSPC and their niches that are disrupted during mobilization need to be reestablished during their homing to the BM and its repopulation. Previously it was believed that mobilization and homing were mirror images of each other; however, emerging evidence now suggests that this is not the case, although both processes involve many of the same adhesive and chemotactic interactions. The homing of HSPC to BM is a rapid process occurring within hours after their transplantation and is a prerequisite for their repopulation and engraftment [29]. Homing involves three consecutive steps: extravasation from PB through the BM endothelium, migration through stroma, and lodgement into niches.</p><p>The extravasation of circulating HSPC within the BM requires a set of molecular interactions that mediates the recognition of circulating HSPC by the endothelium of BM sinusoids [30]. The mechanisms of HSPC extravasation are similar to those of leukocytes into inflamed tissues in that they are mediated by adhesion molecules. The BM endothelium constitutively expressing P-selectin, E-selectin, and VCAM-1 mediates rolling and tethering of HSPC to the blood vessel wall prior to their extravasation [31]. HSPC expressing CD44 and VLA-4 receptors interact with their cognate ligands on the endothelial surface. The coordinated action of these adhesion molecules is triggered by SDF-1 presented at the surface of endothelial cells. SDF-1 mediates activation of lymphocyte function-associated antigen-1, VLA-4 and VLA-5, converting the rolling of HSPC into stable arrest on the endothelium [32].</p><p>SDF-1—CXCR4 signalling is critical in the regulation of HSPC homing, engraftment, and retention in the BM. Blockade of CXCR4 was shown to inhibit HSPC homing and engraftment, whereas overexpression of CXCR4 by CB and mPB CD34+ cells leads to increased SDF-1 induced in vitro migration and homing in NOD/SCID mouse models [33, 34]. We recently showed that mPB CD34+ cells that had higher responsiveness to SDF-1 had high CXCR4 expression and could compensate for a lower CD34+ cell dose in achieving faster hematopoietic engraftment after transplantation [35]. SDF-1 binding to CXCR4 activates phosphatidylinositol 3-kinase (PI3K), the phospholipase C-γ (PLC-γ)/protein kinase C (PKC) cascade, and p44/42 mitogen-activated protein kinase (MAPK) [36]. SDF-1—CXCR4 signalling also activates the atypical PKC subtype PKCζ which mediates cell polarization, adhesion, MMP-9 secretion, and chemotaxis of CD34+ cells [37]. Interactions mediated by the Rho family GTPases (Rac, Rho, and Cdc42) have also been implicated in HSPC homing and engraftment. Following engraftment, deletion of Rac-1 and Rac-2 GTPases led to massive mobilization of HSPC from BM. Knocking out Rac-1 significantly reduced migration towards SDF-1 and attenuated the homing of murine HSC to the endosteum, which is essential for long-term HSC repopulation [38]. We previously demonstrated that Rac-1 colocalization with CXCR4 in membrane lipid rafts of HSPC promotes their in vivo homing in a murine model [39]. Moreover, in vitro treatment of HSPC with supernatants of leukapheresis products (SLPs) or their components, such as C3a or platelet-derived microvesicles (PMVs), modulates the SDF-1—CXCR4 axis and speed up in vivo homing in murine models [39, 40].</p><p>Proteases regulate HSPC migration and tissue localization and have been shown to also play important functions in HSPC homing. The roles of proteolytic enzymes particularly MMP-2, MMP-9, and MT1-MMP in this process will be discussed in detail in subsequent sections.</p><!><p>MMPs belong to a family of Zn2+-binding, Ca2+-dependent endopeptidases whose essential function is proteolysis of the ECM, a process that is required in several cellular processes [41]. Currently, 24 human MMPs have been identified that have structural similarities but vary in their expression profiles and substrate specificities. MMPs are classified based on substrate recognition into stromelysins (MMP-3, MMP-10, and MMP-11), matrilysins (MMP-7, MMP-26), gelatinases (MMP-2 and MMP-9), and collagenases (MMP-1, MMP-8, MMP-13, and MMP-14) [42]. Apart from ECM molecules, MMPs act on a whole array of substrates including other proteinases and MMPs, proteinase inhibitors, growth factors, cytokines, chemokines, cell surface receptors, and cell adhesion molecules and regulate many processes such as cell migration, proliferation, apoptosis, angiogenesis, tumor expansion, and metastasis [42–44]. The expression and function of MMPs are regulated at different levels. Generally expressed at low levels, MMPs are upregulated during tissue remodeling, inflammation, wound healing, and cancer progression [45, 46]. They are synthesized as latent enzymes that are either secreted or membrane-anchored. Six MT-MMPs have been identified so far, of which four, MT1-/MMP-14, MT2-/MMP-15, MT3-/MMP-16, and MT5-/MMP-24, have a transmembrane domain while the other two, MT4-/MMP-17 and MT6-/MMP-25, have a glycosylphosphatidylinositol domain [42]. Their membrane anchoring allows them to carry out pericellular proteolysis. MMPs are activated by the proteolytic release of the N-terminal propeptide domain. Once active, they can be inhibited by endogenous tissue inhibitors of metalloproteinases (TIMPs), the reversion-inducing cysteine-rich protein with Kazal motifs (RECK), and tissue factor pathway inhibitor-2 as well as by plasma inhibitor (α2-macroglobulin) [47, 48]. Therefore, a balance between MMPs and their inhibitors is important to ECM remodeling in the tissue and in HSPC migration.</p><p>The gelatinases MMP-2 and MMP-9 have been extensively studied in cancer and other diseases. MMP-2 is secreted by fibroblasts, endothelial cells, epithelial cells, and transformed cells whereas MMP-9 is produced predominantly by leukocytes [49]. MMP-2 and MMP-9 are required for physiological processes such as ECM remodeling during growth and development, inflammation, wound healing, angiogenesis, and leukocyte mobilization [41, 46]. They are also involved in pathological processes such as cancer, inflammation, and neural and vascular degenerative diseases [45, 46]. Although MMP-2 and MMP-9 are secreted by cells in the developing embryo, mice deficient in these gelatinases are viable. However, mice deficient in MMP-2 exhibit defects in developmental angiogenesis whereas mice deficient in MMP-9 show delayed vascularization and ossification resulting in moderate skeletal abnormalities [49]. MMP-2 and MMP-9 are similar in many respects, but differ in their glycosylation pattern, activation, and substrate specificity. The 92 kDa MMP-9 has two glycosylation sites in the prodomain and the catalytic domain whereas the 72 kDa MMP-2 is a nonglycosylated protein [42]. MMP-2 activation is a cell surface event mediated by the formation of a ternary complex containing MT1-MMP, TIMP-2, and proMMP-2 [42]. The N-terminal domain of TIMP-2 binds to MT1-MMP whereas the C-terminal domain binds the hemopexin domain of proMMP-2. An adjacent MT1-MMP free of TIMP-2 subsequently activates proMMP-2 by cleaving its propeptide domain. On the other hand, MMP-9 activation is mediated by a proteolytic cascade involving MMP-3, MMP-2 and MMP-13 [42]. MMP-3 is activated by plasmin generated from plasminogen by urokinase-type plasminogen activator (uPA) bound to its receptor on the cell surface [42]. Similarly, MMP-2 activates proMMP-13 which then activates proMMP-9 [50]. MMP-2 and MMP-9 share similar proteolytic activities and degrade a number of ECM molecules such as gelatin, collagen types IV, V, and XI, and laminin. In addition MMP-2 also degrades collagen types I, II and III. Both also cleave several non-ECM molecules such as SDF-1, tumor necrosis factor (TNF)-α, transforming growth factor (TGF)-β, plasminogen, and uPA, among many others [42–44].</p><p>MT1-MMP also plays important roles in both physiological and pathological processes. Mice deficient in MT1-MMP exhibit craniofacial dysmorphism, dwarfism, osteopenia, fibrosis of soft tissues, arthritis, and premature death, emphasizing the function of MT1-MMP in ECM remodelling during development [51]. On the other hand, elevated MT1-MMP expression has been observed in a wide variety of cancers including lung, breast, cervical, brain, liver, head, and neck, indicating MT1-MMP's role in tumor progression and metastasis [45]. In addition, it has been implicated in angiogenesis, bone development, atherosclerosis, inflammation, and wound healing [44] by virtue of its ability to degrade several ECM macromolecules including collagens, laminins, fibronectin, aggregan, and fibronectin and to activate proMMP-2 and proMMP-13 [42, 44, 52]. MT1-MMP cleaves CD44, processes α v integrin to its mature form, and degrades tissue transglutaminase, thus modifying the immediate cell environment and affecting cellular functions in a variety of ways [44]. MT1-MMP has a domain structure composed of a propeptide region, catalytic domain, hinge region, hemopexin-like domain, and a type I transmembrane domain with a cytoplasmic tail at the C-terminus which anchors it to the cell surface [42]. MT1-MMP is also expressed in its latent form and cleavage of the propeptide region by furin or related protein convertases renders it active. Activation of MT1-MMP takes place in the trans-Golgi network complex during secretion, and the enzyme is expressed in its active form at the cell surface [53]. Active MT1-MMP is inhibited by RECK [54], TIMP-2, TIMP-3 and TIMP-4 but not by TIMP-1 [48]; however, TIMP-2 has a dual role. On the one hand it inhibits MT1-MMP and on the other promotes activation of proMMP-2 [47]. At the cell surface, the 60 kDa active MT1-MMP molecule undergoes self-proteolysis by removal of its catalytic domain which results in an inactive 44 kDa species [55]. A high level of truncated MT1-MMP coincides with high proMMP-2 activation. In migrating tumor cells, MT1-MMP localizes predominantly in the lamellipodia through its interaction with CD44 [56]. It has been reported that this interaction and the consequent shedding of CD44 stimulate cell motility [57]. The hemopexin domain of MT1-MMP interacts with the stem region of CD44 which interacts with F-actin through its cytoplasmic domain. Binding of MT1-MMP with CD44 indirectly links the proteinase to the cytoskeleton and thereby enables its localization to the lamellipodia [58]. MT1-MMP can be internalized by both clathrin-dependent and caveolae-dependent pathways, and recycled back to the surface, and the internalization process is essential for the enzyme to promote cell migration [58]. The redistribution of MT1-MMP to sites of degradation, such as the lamellipodia of endothelial cells and the invadopodia of tumor cells, is highly complex and involves a dynamic interplay between endocytic and exocytic processes [59]. In most cell types surface expression of MT1-MMP is low due to rapid endocytosis resulting in intracellular accumulation of the proteinase in the early and late endosomes. This dynamic regulation suggests that the spatiotemporal recruitment of MT1-MMP to specialized domains plays a critical role in its invasive properties [60]. Various growth factors, chemokines, and inflammatory mediators have been reported to modulate MT1-MMP expression in malignant and normal cells. In the fibrosarcoma cell line HT1080, Rac-1 modulation of MT1-MMP and its processing to its 44 kDa form correlated with proMMP-2 activation [61]. Furthermore, Rac-1 has been demonstrated to promote hemophilic complex formation of MT1-MMP, recruitment to the lamellipodia-rich cell surface, and subsequent proMMP-2 activation [62]. In Lewis lung carcinoma cell line, type 1 insulin growth factor-1 increases invasiveness of these cells through increased MT1-MMP expression via the PI3K/AKT/mTOR pathway [63]. The SDF-1−CXCR4 axis promotes melanoma cell invasion and metastasis by upregulating MT1-MMP through Rac-1 and RhoA-GTPases [64]. Moreover, through activation of Rac-1 and its downstream effector ERK1/2, SDF-1 intracellularly upregulates MT1-MMP; however, when these cells were in contact with Matrigel, a PI3K-dependent transient redistribution of MT1-MMP to the cell surface was observed [65]. In endothelial cells, VEGF-mediated upregulation of MT1-MMP at the mRNA level occurs through the MAPK/JNK pathway, whereas protein expression is regulated by PI3K [66, 67]. MT1-MMP clustering on the cell surface is dependent on cortical actin polymerization which is regulated by PI3K, and this clustering of MT1-MMP has been suggested to be more important than its internalization [59]. In mesenchymal stromal cells (MSCs), we reported that MT1-MMP mediates their chemotactic migration towards SDF-1 and HGF [68]. Increased chemoinvasion of MSC through upregulation of MT1-MMP by cytokines TGF-β, TNF-α, and IL-1 was also reported [69]. In hematopoietic cells, MT1-MMP has been demonstrated to mediate trans-endothelial migration of monocytes, and the interaction of monocytes with fibronectin and endothelial ligands, such as VCAM-1 and intracellular cell adhesion molecule (ICAM)-1, increased MT1-MMP clustering and localization into membrane protrusions or lamellipodia [70].</p><!><p>MMPs have been traditionally considered to facilitate cell migration by breaching basement membrane barriers comprised of ECM proteins. The mobilizing agent G-CSF induces neutrophil proliferation, activation, and degranulation with the subsequent release of the serine proteases NE, CG, and MMP-9 into the BM, making it a highly proteolytic microenvironment [17, 71]. MMP-9 has been reported to be elevated in plasma after mobilization with G-CSF or IL-8 [71, 72]. Mobilization is thought to occur predominantly through the action of neutrophils and release of granulocytes from the BM always precedes mobilization of HSPC [73]. However, HSPC themselves contribute to this process by secreting MMP-2 and MMP-9 [74]. We have reported that while BM HSPCs in steady-state do not secrete MMPs, upon stimulation with G-CSF, among other growth factors and cytokines, HSPCs secrete both MMP-2 and MMP-9 leading to their enhanced migration through reconstituted basement membrane (Matrigel) [74]. Both mature leukocytes as well as immature CD34+ cells from in vivo G-CSF-mobilized PB highly express MT1-MMP compared to their steady-state (ss)BM and ssPB counterparts. Furthermore, cell surface expression of MT1-MMP in in vivo mobilized polymorphonuclear cells (PMNs) and monocytes was 8 and 12 times higher than in their respective ssPB counterparts. G-CSF exerts its effects on other signalling axes that also result in upregulation of MMP-9 and MT1-MMP. For example, we observed that HGF increases in the plasma of mobilized patients and that G-CSF, through induction of HGF, upregulates MMP-9 and MT1-MMP expression in BM PMN [11]. In addition, activation of the CC by G-CSF triggers a series of reactions generating various bioactive peptides including the C5 cleavage fragments (C5a and desArgC5a) which have been shown to play an important role in mobilization [25, 75]. In particular, we recently demonstrated that C5a increases MMP-9 and MT1-MMP expression in PMN and mononuclear cells (MNCs), contributing to a microenvironment that is conducive to the egress of HSPC from the BM [28]. MT1-MMP causes pericellular degradation by processing several ECM components (such as gelatin, fibronectin, laminin, vitronectin, and fibrillar collagens) [41, 43, 52]. This is further substantiated by observation that G-CSF-induced migration of MNC and HSPC through Matrigel (containing collagen type IV, laminin, entactin) is MT1-MMP dependent. MT1-MMP expressed by HSPC activates proMMP-2 secreted by BM stromal cells [76, 77]. Active MMP-2 could subsequently initiate a cascade of activation of other MMPs including MMP-9 and MMP-13, which not only degrade the ECM but also disrupt adhesive interactions between HSPC and their niches. One of the most potent of these interactions is between SDF-1, produced by stromal cells, including endothelial cells, and osteoblasts, and its receptor CXCR4, expressed by HSPC. Recently, we demonstrated that C5a, apart from upregulating MT1-MMP, decreases CXCR4 expression in PMN, which disrupts their chemotaxis towards SDF-1. Moreover, this impaired chemoattraction is partially restored by the potent MT1-MMP inhibitor EGCG (which also inhibits proMMP-2 activation), suggesting that MT1-MMP contributes towards reduced retention of HSPC in the BM microenvironment [28].</p><p>In addition to matrix remodeling, MT1-MMP cleaves adhesion molecules, such as CD44 and integrins, and chemokines such as SDF-1. Reduced retention of HSPC in the BM during G-CSF-induced mobilization is also due to a decrease in active SDF-1 in the BM which coincides with peak proteolytic activity [12, 16]. SDF-1 can be cleaved and inactivated by several proteases which are activated during mobilization such as NE, CG, MMP-2 and MMP-9, CD26, CPM, and MT1-MMP [13, 14, 17–20]. MT1-MMP cleavage of SDF-1 results in loss of binding to CXCR4 and reduced chemoattraction for CD34+ cells [18]. Therefore, during G-CSF-induced mobilization, upregulation of MT1-MMP expression in HSPC has consequences that could lead to their reduced retention in the BM and resultant egress into PB.</p><p>A low SDF-1 level in the BM during cytokine-induced mobilization has been suggested to result from suppression of SDF-1-producing osteoblasts, and although the mechanism of this phenomenon is not completely understood [78], it has been suggested to occur in light of an observed increase in osteoclast activity [79]. Active bone remodeling takes place during G-CSF mobilization, and increased cathepsin K and MMP-9 secretion by osteoclasts has been implicated in high bone turnover [79, 80]. Similarly, MT1-MMP also plays an important role in bone remodeling as shown by the fact that MT1-MMP−/− mice exhibit severe skeletal defects [51]. MT1-MMP is also known to be expressed by osteoclasts and recently it was demonstrated that MT1-MMP was necessary for macrophage fusion during multinucleated osteoclast formation and differentiation [81]. Accordingly, 8-day-old MT1-MMP−/− mice exhibited impaired osteoclast function, which did not, however, result in increased bone mass since osteoblast function is also compromised in these mice [51, 81]. These results suggest that MT1-MMP could have a dual role in bone development. Nevertheless, the role of MT1-MMP in bone remodeling during G-CSF-induced mobilization requires further investigation.</p><p>It was postulated that MT1-MMP expression in HSPC is regulated by the endogenous inhibitor RECK and that high MT1-MMP and low RECK levels in HSPC resulted in the egress of BM progenitors into circulation [54, 77]. In a recent publication, we provided evidence that MT1-MMP expression on the surface of HSPC is regulated by its incorporation into membrane lipid rafts, and that both MT1-MMP expression and proMMP-2 activation are PI3K-dependent [76]. Moreover, in co-cultures of fibroblasts with BM CD34+ cells, proMMP-2 is not activated; however, pre-incubation of CD34+ cells with G-CSF highly upregulated MT1-MMP, which was then able to activate proMMP-2 in similar co-cultures. Consistent with this, in co-cultures of stromal cells with mPB CD34+ cells that had been transfected with MT1-MMP siRNA, active MMP-2 was not detectable [76]. The MT1-MMP activation of proMMP-2 in the BM microenvironment is important because active MMP-2 not only initiates the activation of other MMPs that play a role in matrix remodeling, but also inactivates SDF-1 and CXCR4 and adhesion molecules, processes which facilitate egress of HSPC from BM niches across the ECM and subendothelial membranes.</p><p>MT1-MMP promotes cell motility by pericellular ECM degradation [41, 52, 53]. In this respect we demonstrated the role of MT1-MMP in the migration of CD34+ cells and MNC across reconstituted basement membrane as specific inhibition of MT1-MMP by siRNA significantly abrogated their migration [76]. On the other hand, upregulation of MT1-MMP expression in HSPC by G-CSF could lead to their reduced retention in the BM and enhanced egress into PB [77]. The definitive proof of this would involve studies using MT1-MMP−/− mice, but such mice have severe developmental abnormalities resulting in their early death [51]; hence in vivo mobilization experiments in MT1-MMP knockout mice are difficult. However, these experiments have been carried out in chimeric mouse models [77].</p><p>The initial cell response to cytokine or chemokine stimulation is the reorganization of the actin cytoskeleton during which several signalling pathways are activated. MT1-MMP has been shown to be modulated by PI3K, MAPK, and Rho family GTPases depending on the cell type and stimulant [64, 82]. We demonstrated that although both PI3K and MAPK are activated in hematopoietic cells upon G-CSF stimulation, MT1-MMP expression and proMMP-2 activation are only PI3K dependent. Furthermore, inhibition of the PI3K-AKT axis also inhibits cell polarization, co-localization of MT1-MMP with F-actin, and trans-Matrigel migration of mPB CD34+ cells [76]. Murine HSPC studies have shown that cytokine-mediated lipid raft clustering activated the AKT-FOXO signalling pathway, which is essential for entry into cell cycle, and inhibition of lipid raft formation by MβCD led to repression of this pathway and hibernation-like state of HSPC [83]. The recruitment of the regulatory PI3K subunit p85 to lipid rafts after cytokine stimulation is required for activation of the downstream effectors of the PI3K-AKT axis and is dependent on lipid raft integrity [84]. Consistent with these findings, we showed that G-CSF-induced PI3K activation in lipid rafts leads to reorganization of the actin cytoskeleton, recruitment of MT1-MMP into lipid rafts, and proMMP-2 activation [76], which increases the degradation of basement membrane and interstitial matrix in the BM microenvironment and eventually the egress of HSPC into circulation as shown in Figure 1(a). Mobilizing signals such as G-CSF expand the number of myeloid cells in the BM, which secrete proteolytic enzymes (including MMP-2 and MMP-9) that disrupt interactions that retain the HSPC in the BM (e.g., SDF-1/CXCR4, VCAM-1/VLA-4, ECM). Permeabilization of the endothelial barrier occurs with granulocytes "paving the way" for HSPC. Chemoattractants in the blood (e.g., bioactive lipids such as sphingosine 1-phosphate, HGF) further facilitate egress into PB of HSPC [73] (Figure 1(a)).</p><!><p>MMPs degrade various ECM molecules and facilitate HSPC transmigration across basement membrane barriers. SDF-1 and other growth factors induce the secretion of MMP-2 and MMP-9, thus facilitating in vitro migration of HSPC across reconstituted basement membrane towards SDF-1 [73]. Incubation of CB-derived HSPC with SCF induces MMP-2 and MMP-9 secretion and homing in NOD/SCID mice [85]. We demonstrated that MT1-MMP mediates migration of CB CD34+ cells and megakaryocytic progenitors towards an SDF-1 gradient [86]. Recently another group reported, using a chimeric mouse model, that engraftment levels of MT1-MMP−/− c-Kit+ cells were significantly lower than those of wild-type cells, and inhibition of MT1-MMP by monoclonal antibody attenuated homing of human HSPC in a NOD/SCID mouse model [77].</p><p>We have demonstrated that HSPC primed by SLP or their components (fibrinogen, fibronectin, complement C1q, complement C3a, PMV, HA, thrombin) respond better to an SDF-1 gradient [26, 39, 40, 87, 88]. Furthermore, murine HSPC that have been primed with C3a, PMV, or SLP before transplantation engrafted faster in lethally irradiated mice [39, 40, 87]. This priming effect was due to increased incorporation of CXCR4 and Rac-1 GTPase into membrane lipid rafts and to increased MMP-2 and MMP-9 secretion, indicating that the SDF-1—CXCR4 axis cooperates with MMPs during HSPC homing [39].</p><p>We recently investigated the effects of two priming molecules, HA and thrombin, on the modulation of MT1-MMP expression and its activity [88]. First, we found that HA and thrombin upregulated MT1-MMP expression in CB HSPC. Secondly, HA- and thrombin-primed chemoinvasion of HSPC towards a low SDF-1 gradient was MT1-MMP dependent. It has been established that the SDF-1—CXCR4 axis promotes the chemotaxis not only of normal but also of tumor cells [70, 89]. For example, the coordinated interaction of CXCR4 and MT1-MMP is required for melanoma cell metastasis to lungs [65]. CXCR4 was required for the initial phases of melanoma cell chemotaxis and their arrival in the lungs, whereas MT1-MMP was necessary for subsequent invasion and dissemination of the tumor. We can therefore speculate that a coordinated interaction between CXCR4 and MT1-MMP is also required by HSPC for their homing to the BM. MT1-MMP expressed in CB HSPC (upregulated by HA and thrombin), activates proMMP-2 secreted by endothelial cells. Active MMP-2 could thereby participate in the extravasation process and facilitate homing by activating other MMPs and degrading ECM barriers [73]. In this respect, MT1-MMP has been shown to facilitate trans-endothelial migration of monocytes through clustering of MT1-MMP at the lamellipodia upon contact with activated endothelial cells or the immobilized endothelial ligands VCAM-1 and ICAM-1 [70]. Priming of homing-related responses of CB HSPC showed that HA and thrombin activated the PI3K-AKT signalling axis and Rac-1 GTPase [88]. MT1-MMP expression and proMMP-2 activation were dependent on both these signalling pathways. We demonstrated that intracellular crosstalk between these pathways leads to signal amplification of a low SDF-1 gradient, leading to enhanced MT1-MMP expression on the cell surface of CB HSPC and increase chemoinvasion towards SDF-1 [88]. Thus, agents such as HA and thrombin that positively regulate the SDF-1—CXCR4 axis may also prime the homing-related responses of HSPC by upregulating MT1-MMP. HSPCs then attach/tether to and extravasate the sinusoid endothelium. MT1-MMP promotes activation of proMMP-2 that degrades ECM barriers. HSPCs are chemoattracted to their BM niches due to amplified chemotactic response towards SDF-1 produced by osteoblasts and stromal cells (Figure 1(b)).</p><!><p>Accumulating evidence indicates an important role for MMP-2, MMP-9, and MT1-MMP in HSPC mobilization and homing. In particular, modulation of MT1-MMP could become a potential target for development of therapeutic strategies that could improve transplantation outcomes. First, critical to optimizing clinical mobilizing regimens is an understanding of the molecular mechanisms that regulate HSPC mobilization [73, 90, 91]. Such studies have already led to the development of new mobilizing agents that resulted in a rapid collection of more HSPC for transplantation. For example, recent use of Plerixafor (AMD3100) which reversibly binds CXCR4 and disrupts SDF-1—CXCR4 interactions has demonstrated that a combination of AMD3100 and G-CSF results in greater mobilization efficacy compared to G-CSF alone [92, 93].</p><p>The main problem with G-CSF-induced mobilization is the variable kinetics of mobilization as shown by the significant number of patients/donors who either mobilize poorly or fail to mobilize. Recent findings indicate that positive regulation of MT1-MMP increases migration/mobilization of HSPC, and hence the development of mobilizing agents that increase MT1-MMP expression could enhance mobilization efficiency. For example, cytokines such as HGF that upregulate MT1-MMP expression in HSPC [11] could also synergize with G-CSF and increase mobilization efficiencies in patients who mobilize poorly with G-CSF alone. This could be tested in a clinical setting following confirmatory results from murine models. On the other hand, potential inhibitors such as green tea polyphenol EGCG, which inhibits MT1-MMP expression and proMMP-2 activation, and statins, which like methyl-β-cyclodextrin disrupt lipid raft formation, could inhibit MT1-MMP incorporation into lipid rafts and thereby negatively affect HSPC mobilization. This could result in lowering the number of HSPC collected, and therefore we suggest that overcoming these inhibitory effects could improve transplantation outcomes.</p><p>Secondly, a successful transplantation outcome also depends on the ability of a large number of intravenously injected HSPC to rapidly find their way to the BM. This has led to the design of strategies to increase the number of HSPC available by ex vivo expansion of HSPC before transplantation. Another strategy is to enhance the homing potential of a limited number of HSPC by their ex vivo exposure to agents such as C3a and PMV that prime their chemotactic responses by positively regulating the SDF-1—CXCR4 axis, as we previously showed in murine models [40, 87]. Interestingly, this strategy of ex vivo exposure of CB HSPC to C3a is already being evaluated in clinical trial [94]. Therefore we suggest that other priming agents that enhance the responsiveness of CB HSPC towards SDF-1 and upregulate MMPs expression could be used for ex vivo short-term treatment/priming of CB HSPC and be evaluated in clinical trials.</p><p>Lastly, a better understanding of the molecular mechanisms of HSPC migration is not only beneficial for designing novel therapeutic strategies discussed above but could also be applied to enhancement of the homing properties of other types of stem cells such as MSC which share common migration mechanisms.</p>

PubMed Open Access

An RNase P-based assay for accurate determination of the 5\xe2\x80\x99-deoxy-5\xe2\x80\x99-azidoguanosine-modified fraction of in vitro transcribed RNAs

Chemo-enzymatic approaches are important for generating site-specific, chemically-modified RNAs, a cornerstone for RNA structure-function correlation studies. T7 RNA polymerase (T7RNAP)-mediated in vitro transcription (IVT) of a DNA template containing the G-initiating class III \xce\xa66.5 promoter is typically used to generate 5\xe2\x80\x99-chemically-modified RNAs by including a guanosine analog (G-analog) initiator in the IVT. However, the yield of 5\xe2\x80\x99-G-analog-initiated RNA is often poor and variable due to the high ratios of G-analog:GTP used in IVTs. We recently reported that a T7RNAP P266L mutant afforded a ~3-fold increase in fluorescent 5\xe2\x80\x99-thienoguanosine-initiated pre-tRNA compared to the wild type. Here, we further explored the utility of T7RNAP P266L to generate 5\xe2\x80\x99-deoxy-5\xe2\x80\x99-azido-guanosine (azG)-initiated RNA and found that the mutant generated ~4-fold more azG-initiated pre-tRNACys than the wild type in an IVT containing a 10:1 ratio of azG:GTP. For accurate quantitation of the 5\xe2\x80\x99-azG-initiated RNA fraction, we employed RNase P, an endonuclease that catalyzes the removal of the 5\xe2\x80\x99-leader in pre-tRNAs. Importantly, we show how RNase P can be leveraged for assessing 5\xe2\x80\x99-G-analog incorporation in any RNA by rendering the target, upon its binding to a customized external guide sequence RNA, into an unnatural substrate of RNase P. Such an approach in conjunction with T7RNAP P266L-based IVTs should aid chemo-enzymatic methods that are designed to generate 5\xe2\x80\x99-chemically-modified RNAs.

an_rnase_p-based_assay_for_accurate_determination_of_the_5\xe2\x80\x99-deoxy-5\xe2\x80\x99-azidoguan

<!>Overexpression and purification of recombinant proteins used in this study<!>Synthesis of azG (5\xe2\x80\x99-deoxy-5\xe2\x80\x99-azidoguanosine)<!>In vitro transcription reactions<!>Post-transcriptional removal of 5\xe2\x80\x99-ppp-G-initiated RNAs<!>PRORP-based assays to determine percent incorporation of azG in pre-tRNACys<!>PRORP-based assays to determine percent incorporation of azG in MR1 and MR2

<p>RNAs with site-specific chemical modifications are important tools in RNA structure-function studies.[1] The chemo-enzymatic approach[1–2] to site-specifically introduce a modified nucleoside/nucleotide monophosphate analog at the 5'-end of RNAs during in vitro transcription (IVT) is particularly attractive due to its low cost and its applicability to synthesize any full-length RNA (> 25 nt) at reasonable yield. T7 RNA polymerase (T7RNAP) has long been preferred in IVTs that have been employed to generate such 5'-chemically modified RNAs because of its tolerance to initiate RNAs with guanosine analogs[3] (G-analogs; see Table S1 for a representative list of analogs used in previous IVTs with T7RNAP and the G-initiating class III Φ6.5 promoter).</p><p>The ease of T7RNAP-based IVTs to incorporate 5'-G-analogs in RNAs is counterweighed by the low yields that result from (i) use of high G-analog:GTP ratios in IVTs, and (ii) the propensity of T7RNAP to abort transcription when RNAs are initiated with G-analogs.[4] We recently demonstrated[4] that the decreased yield could be partly overcome by use of the Pro266Leu (P266L) mutant derivative of T7RNAP, which has a decreased propensity for abortive transcription.[5] Use of the P266L mutant compared to the wild type (WT)[4] afforded a 3-fold increase in the IVT yield of pre-tRNACys initiated with thienoguanosine (thG), a fluorescent guanosine surrogate.[6] When an IVT performed with this mutant was coupled with a one-pot multi-enzyme (OPME) method (Figure S1),[4] which entails a post-transcriptional clean-up step using sequentially a pyrophosphohydrolase and an exoribonuclease to specifically degrade 5'-GTP-initiated pre-tRNACys, we obtained near-homogeneous 5'-thG-initiated pre-tRNACys at good yield (40 μg/100 μL IVT).</p><p>In our recent study,[4] we chose to study 5'-thG-initiated RNAs because the intrinsic fluorescence of thG together with Abs260 measurements permitted an easy quantitation of the fraction of in vitro transcribed RNAs initiated with either thG or GTP. While exploring the potential of T7RNAP P266L to synthesize RNAs initiated with non-fluorescent G-analogs, we recognized the need for an accurate method to assess the fractional content of GTP- versus G-analog-initiated RNAs present at the end of an IVT. To this end, we describe the development and validation of a new assay to measure the fraction of RNAs initiated with a G-analog in IVTs by exploiting RNase P,[7] an endonuclease that cleaves the 5'-leader in pre-tRNAs. Here, we demonstrate the utility of T7RNAP P266L in synthesizing 5'-deoxy-5'-azidoguanosine (azG)-initiated RNAs. Additionally, we showcase the value of the OPME approach in generating near-homogeneous 5'-azG-initiated RNA through targeted depletion of GTP-initiated RNAs.</p><p>This work was motivated by the fact that despite the increasing popularity of using click chemistry to conjugate spectroscopic or chemical (e.g., biotin) handles through the reaction between an alkyne and an azide to generate a 1,4-disubstituted 1,2,3-triazole,[1–2, 8] there is an incomplete understanding of the variability in the fractional content and yield of 5'-modified RNAs generated during IVTs. In one study, 5'-azG was reported to be the exclusive initiator of a 66 nt RNA (IVT containing 4:1 azG:GTP), an inference based on the complete conversion of this RNA to a "clicked" end product;[8a] in this instance, yields were not reported. In another study, increasing the azG:GTP ratio from 5:1 to 100:1 led to a ~2-fold increase (from 36% to 81%) in azG incorporation and a ~2-fold decrease (from 7.6 to 3.6 μg/200 μL IVT) in yield of a 76 nt-long 5'-azG-initiated RNA;[8b] the latter was quantitated using the colorimetric reaction with 6-heptynoyl p-nitroaniline. The high azG incorporation resulting from use of a 100:1 azG:GTP in the IVT was offset by a very low yield, although this is partly attributable to the use of 0.2 mM rNTPs in these specific IVTs.[8b] In a comprehensive study that examined incorporation of O-(5'-guanosine)-O-propargyl monophosphate (GMPPrg) as the initiator in five different RNAs (~60 nt),[9] a 9:1 GMPPrg:GTP ratio led to 21 to 38% G-analog incorporation and yields ≤ ~20 μg/300 μL IVT. These yields, which were calculated post-HPLC purification of the "fluor-clicked" end product,[9] were likely enhanced by use of a fed-batch method[10] that entailed periodic supplementation of GTP during the IVT.</p><p>Here, we focused on azG incorporation by T7RNAP P266L and developed a method to determine the fraction of RNAs initiated with either azG or GTP. We first considered using click chemistry with an alkyne-bearing fluorophore to detect the fraction of 5'-azG-initiated RNAs; however, we abandoned this idea because any non-covalently bound fluorophore that remains even after purification of the modified RNAs would render azide measurements inaccurate. We then explored a spectrophotometric assay that detects azides on the basis of a light-absorbing, ferric-azide complex (λmax = 450 nm);[11] although we obtained a good standard curve for azG (data not shown), we found that reliable detection required high (> 0.1 mM) concentrations of azG-initiated RNAs. These roadblocks inspired us to design an alternative method, preferably with general applicability to studies of many G-analog initiators.</p><p>We chose to exploit the expected difference in electrophoretic mobility of small RNAs containing either 5'-triphosphate or a 5'-hydroxyl, since the latter lacks the negative charges present in GTP. Because IVTs are typically used to generate RNAs > 25 nt, any method that leverages electrophoretic mobility differences in standard denaturing PAGE requires an additional step to focus on a short (< 10 nt) segment corresponding to only the 5' region of the target RNA.[12] We address this challenge using a two-phase approach: after confirming the expectation that small RNAs containing either a GTP or an uncharged G-analog at the 5'-end are resolvable by PAGE and easily quantitated, we customized an endonuclease to cleave longer RNAs and release short 5'-fragments for subsequent analysis by PAGE.</p><p>RNase P is an endonuclease that catalyzes the removal of 5'-leaders from precursor tRNAs; it functions as either a ribonucleoprotein (powered by a catalytic RNA) or an RNA-free, protein-only form.[13] For the studies described here, we have employed exclusively the biochemically well-characterized Arabidopsis thaliana protein-only RNase P 3 (AtPRORP3, 60 kDa).[14] We chose pre-tRNAs for our initial tests because cleavage of the 5'-leader of a pre-tRNA, initiated with either azG or GTP, was expected to produce as cleavage products small RNAs bearing either 5'-azG or 5'-GTP, which in turn could be resolved by denaturing PAGE and quantitated to obtain a reliable measurement of the percent of 5'-azG-initiated RNA. To test this idea, we performed IVTs (which included or excluded azG) to generate A. thaliana pre-tRNACys with a 5 nt leader and cleaved it with recombinant AtPRORP3 (Figure 1A). Indeed, the two expected products (5'-azGGUUU-3' and 5'-ppp-GGUUU-3') are easily separated (Figure 1). We included α−32P-UTP in our IVTs to enable visualization and quantitation of the two 5 nt leaders. To establish size markers for the expected products, we used AtPRORP3 to cleave pre-tRNACys, which was generated in a standard IVT, to yield 5'-ppp-GGUUU-3' (Figure 1B). We then further processed this sample with a calf-intestinal phosphatase (CIP) treatment to dephosphorylate 5'-ppp-GGUUU-3' and generate 5'-OH-GGUUU-3', which, as expected, migrated similarly to 5'-azGGUUU-3'; this migration pattern is expected since 5'-OH- and 5'-azG-GGUUU-3' lack the 5'-triphosphate-associated negative charges (Figure 1B). We independently ascertained the presence of 5'-azG in this RNA using the colorimetric ferric chloride-based assay[11] (data not shown).</p><p>With the PRORP-based assay in hand, we investigated potential advantages afforded by T7RNAP P266L compared to the WT in generating 5'-azG-initiated pre-tRNACys. For pairwise comparisons, we used ~2 μg of T7RNAP WT and P266L in our IVTs as described previously.[4] First, we examined if aborted transcription with T7RNAP WT was indeed a roadblock in IVTs containing azG. For a sensitive visualization in this regard, we conducted "hot", small-scale transcriptions with α−32P-UTP. With T7RNAP WT, we observed a substantial decrease in the full-length (likely due to an increase in aborted transcripts), when the azG:GTP ratio was increased progressively from 0:1 to 4:1 to 10:1 (Figure S2). In contrast, T7RNAP P266L produced more full-length pre-tRNACys than T7RNAP WT regardless of the azG:GTP ratio used (Figure S2); this finding mirrors our previous observation with thG as the G-analog initiator in IVTs.[4]</p><p>Second, for the purpose of assessing changes in the yield of RNAs from IVTs performed with the T7 RNAP P266L mutant compared to the WT, we set up "cold" IVTs using different ratios of azG:GTP. At 4:1 azG:GTP, we found that T7RNAP WT incorporated azG into 43 ± 3% of pre-tRNACys transcripts and yielded 22 ± 2 μg (total RNA)/100 μL IVT; at 10:1 azG:GTP, there was only a modest increase in percent incorporation of azG (52 ± 1%) while the yield decreased by 2.2-fold to 10 ± 1.5 μg/100 μL IVT (Figures 1B, 1C, and S3). In contrast, when T7RNAP P266L was used in an IVT with a 4:1 ratio of azG:GTP, we observed 60 ± 1% incorporation of azG into pre-tRNACys and a total RNA yield of 42 ± 7 μg/100 μL IVT; at 10:1 azG:GTP, percent incorporation of azG was 76 ± 1% and total RNA yield as 29 ± 1 μg/100 μL IVT (Figures 1B, 1C, and S4). Thus, the use of T7RNAP P266L compared to the WT in IVTs performed with either 4:1 or 10:1 azG:GTP led to ~2.5- or 4-fold, respectively, increased yield of 5'-azG-initiated pre-tRNACys. At least three independent replicates confirmed this trend.</p><p>Lastly, because GTP cannot be excluded from any IVT, some fraction of GTP-initiated transcripts is inescapable regardless of the ratio at which GTP is used with a G-analog initiator. This caveat motivated the design of the OPME approach (Figure S1),[4] which we tested here using the pre-tRNACys generated in an IVT using T7RNAP P266L and a 10:1 ratio of azG:GTP. The OPME method led to 20 ± 1 μg/100 μL IVT of nearly homogeneous (93 ± 3%) 5'-azG-initiated pre-tRNACys (Figures 1B, 1C, and S5).</p><p>Building on the value of using AtPRORP3 to accurately determine the fraction of 5'-azG-initiated pre-tRNACys (natural substrate of APRORP), we sought to extend the utility of this approach to any RNA generated in an IVT. To this end, we exploited the finding that any RNA could be tailored for targeted, site-specific cleavage by RNase P if binding of the target RNA to a customized external guide sequence (EGS) leads to a sequence- and structure-specific complex resembling the pre-tRNA (Figure 2A).[15] While various aspects of substrate recognition by AtPRORP3 have been studied,[13a,13c] the ability of this enzyme to cleave bipartite substrates was unknown, a gap that we first addressed.</p><p>To render any target RNA·EGS complex an efficient non-natural substrate of AtPRORP3, the EGS design must take into account the substrate-recognition determinants[13a] of AtPRORP3. Deletion of the anticodon (AC) stem-loop in a pre-tRNAGly had no effect on either the binding [KM(STO)] or the rate of cleavage (kreact) by AtPRORP3 under single-turnover (STO) conditions;[13a] in contrast, deletion of the D stem-loop increased KM(STO) ~57-fold and decreased kreact 5-fold. However, substitution of the D stem-loop with a 9-nt loop in a model substrate (not a pre-tRNA) increased KM(STO) ~27-fold and had no effect on kreact. These findings inspired us to design an EGS that would result in a final substrate where the AC loop is disrupted and the D stem-loop is replaced by a 9-nt loop (Figure 2A).</p><p>To test our bipartite-substrate strategy, we designed a Model RNA I (MR1, 30 nt) and an External Guide Sequence 1 (EGS1) customized for MR1 (Figure 2A). Recognition and cleavage of the MR1·EGS1 complex by AtPRORP3 was expected to produce 5 nt (5'-GGUUU-3') and 25 nt products. When 5'-[32P]-MR1 was hybridized to EGS1, we observed ~90% cleavage by AtPRORP3 under the appropriate assay conditions (Figure 2B). We then performed IVTs with T7RNAP P266L, α−32P-UTP, and a 4:1 ratio of azG:GTP to generate internally radiolabeled MR1. Cleavage of the MR1·EGS1 complex by AtPRORP3 and quantitation of the resulting products revealed that 58 ± 4% of the RNAs was initiated with 5'-azG (Figure 2C; mean and standard deviationdetermined from three independent measurements).</p><p>AtPRORP3 has been shown to use only N-1 and N-2 in guiding its cleavage between N-1 and N+1 (N+1 is the first position in the mature tRNA).[14c] Thus, it is unlikely that the fraction of 5'-azG-initiated RNA determined above could be influenced by a potential bias in the recognition/processing of 5'-azG- versus 5'-GTP-initiated MR1 by AtPRORP3 given that the modification is located at N-5. Regardless, we tested this potential bias by performing a time-course assay using MR1 RNA that was transcribed by T7RNAP P266L using either a 0:1 or 4:1 ratio of azG:GTP (Figures S6). Under single-turnover conditions, the rate of cleavage of MR1 transcribed under these two conditions was roughly the same (Figure S6). As expected, the percent of MR1 cleaved varied depending on the time of incubation (from 17 to 87% in this experiment), every time point yielded approximately the same quantitation with respect to the fraction of 5'-azG-initiated MR1; averaging all time points resulted in a mean ± standard deviation of 62 ± 6%. Since even partial cleavage of a target RNA by AtPRORP3 suffices for a reliable measure of 5'-G-analog incorporation, there is no need to optimize the EGS and/or assay conditions to obtain near-complete cleavage of different target RNAs. Tweaking the annealing conditions, however, may be necessary for long target RNAs where the EGS binding site is not accessible.</p><p>We then used the AtPRORP3- and EGS-based assay to determine the fraction of 5'-azG-initiated RNAs in a second model RNA (MR2, 45 nt), which upon assembly with EGS1 and cleavage by AtPRORP3 was expected to yield 8- nt (5'-GGUUUAUU-3') and 37-nt products (Figure S7). After obtaining MR2 RNA from IVTs containing a 4:1 ratio of azG:GTP, the PRORP-based assay revealed that 84% of MR2 RNA was 5'-azG-initiated (Figure S7). While the electrophoretic mobility differences for the 8 nt 5'-azG versus 5'-GTP were expectedly less pronounced than the 5 nt counterparts obtained from cleavage MR1, the two products from MR1 were easy to resolve and quantitate (Figure S7).</p><p>The PRORP (RNase P)-based assay described here provides a sensitive and easy method to calculate the fraction of 5'-G-analog-initiated RNAs generated in an IVT. Although we have provided proof of principle with 5'-azG, we expect the method to be broadly applicable to any G-analog whose mass and charge difference with GTP will result in an electrophoretic mobility difference. In practice, the determination of the exact fraction of 5'-G-analog-initiated RNAs will require an EGS RNA and recombinant AtPRORP3. The EGS can be either purchased or easily generated from an IVT using a pair of synthetic oligonucleotides, and the protein can be isolated using a single affinity purification step. This investment, however, will lead to various payoffs.</p><p>Foremost, an accurate determination of the extent of 5'-modification will decrease chemo-enzymatic synthesis costs by dialing down the stoichiometric excess of fluors or spin labels (or another probe) based on the exact amount of 5'-modified RNA rather than the total RNA as is customary. Second, our work reveals that empirical fine-tuning, based on accurate quantitation, is important for success of the chemo-enzymatic strategy. For example, our tests of 4:1 and 10:1 ratio of azG:GTP in IVTs revealed that T7RNAP P266L yields more 5'-azG-initiated RNA than WT, with the 4:1 ratio engendering the best trade-off between RNA yield and percent of 5'-azG-initiated RNA. Third, if a probe-tethered RNA is needed in large amounts for high-throughput enzyme assays or for structural studies, maximizing yield is critical. If the experimental work plan of the labeled RNA allows some latitude in terms of the 5' sequence in the target RNA, assessment of different 5'-leader sequences with the same EGS will help identify an initiating sequence that allows the highest fraction of G-analog-initiated RNA generated in an IVT. Such a recommendation is based on our finding that T7RNAP P266L generated 84% 5'-azG-initiated MR2 compared to 58% with MR1 and pre-tRNACys. While pre-tRNACys and MR1 have Gs at +1, +2, +6, and +7, MR2 has Gs at only +1 and +2. Since G-analog-initiated RNAs are already more prone to result in aborted transcripts, the additional stress of multiple G residues in the first 8 nt of the nascent transcript probably accentuates aborted transcription due to the low concentrations of GTP used in these IVTs to favor initiation with the G-analog. Last, use of the OPME approach to enrich 5'-G-analog-initiated RNA is required only if there is a large fraction of 5'-GTP-initiated RNA, again lending importance to an accurate quantitation of that fraction.</p><p>In addition to validation of a customizable assay that affords a simple approach to calculate the fraction of 5'-G-analog-initiated RNAs at the end of an IVT, our results provide a second instance where (i) use of T7RNAP P266L in lieu of WT results in increased total RNA yield and percent of 5'-G-analog-initiated RNA, and (ii) the OPME approach can enrich the percentage of 5'-G-analog-initiated RNA to near homogeneity. These findings with azG parallel those obtained when thG was the G-analog, and underscore the tolerance of T7RNAP P266L to accept two different G-analogs. Our work should motivate studies to test other G-analogs (especially non-azide/alkyne initiators such as 5'-hydrazinyl-5'-deoxyguanosine[16]), and determine if a common 10-nt leader/initiating sequence rich in A, C, U would afford high yields of RNAs initiated with different G analogs. Use of programmable RNases (e.g., Argonautes loaded with defined DNA guides[17]) for site-specific cleavage of large RNAs and generation of smaller cleavage products bearing the 5'-G-analog initiator also merits consideration.</p><!><p>Both WT and P266L T7RNAPs were overexpressed and purified as described earlier,[4] while Δ9 AtPRORP3 followed a previously published protocol but with only the first affinity chromatography step which sufficed to yield nearly homogeneous protein.[14c]</p><!><p>azG was generated as described previously,[18] by reacting 5'-deoxy-5'-iodoguanosine with sodium azide under aqueous conditions. The IR and 1H-NMR data (not shown) confirmed the purity and identity of the final product. The azG stock used for IVTs was prepared in water; solubilization was facilitated by a 5 min incubation at 55°C.</p><!><p>A PCR DNA template containing a class III Φ6.5 promoter (5'-TAA TAC GAC TCA CTA TA-3') and the gene encoding an A. thaliana mitochondrial pre-tRNACys with a 5 nt 5'-leader and no trailer (Figure 1) was prepared as previously described.[4, 19] DNA templates containing a class III Φ6.5 promoter and the genes for either MR1 or MR2 model RNAs were generated using a pair of DNA oligonucleotides for each RNA (Table S2; Sigma). These oligonucleotides were designed to be complementary at their respective 3'-termini; the 17-bp complement allowed fill-in with Phusion DNA Polymerase (NEB). The PCR amplicons were purified using a QIAquick PCR Purification Kit (QIAGEN) and the amplicon concentration was determined using A260 measurements on a NanoDrop 2000c spectrophotometer.</p><p>The IVT reaction conditions were identical to our previous study[4] with respect to buffer, enzyme amounts, etc. IVTs containing azG were performed using 4.8 mM azG and either 1.2 mM GTP (4:1 ratio) or 0.48 mM GTP (10:1 ratio). To determine the total RNA yields of pre-tRNACys, 100-μL IVTs were performed at 37°C for 5 h. Post-transcription, the DNA template was digested by addition of 10 U of DNase I (Roche) and the reaction was then subjected to phenol-chloroform extraction. Next, the samples were extensively dialyzed against water at 4°C to remove unincorporated ribonucleotides; we used a 3,500-Da cut-off dialysis membrane. The RNA samples were recovered and then precipitated by ethanol precipitation. The resulting pellets were re-suspended in water and total RNA yield was determined using Abs260 measurements on a NanoDrop 2000c spectrophotometer and the extinction coefficient of pre-tRNACys.</p><p>To internally label pre-tRNACys, MR1, and MR2 with 32P, 20-μL IVTs were performed as described above, except with the inclusion of 20 μCi of α-[32P]-UTP [3000 mCi/mmol; PerkinElmer, Waltham, MA]. We chose α-[32P]-UTP because each of the expected cleavage products contained multiple U residues, thus enabling visualization of the 5'-leader post-PAGE regardless of whether the RNA had been initiated with azG or GTP. IVTs containing α-[32P]-UTP were performed at 37°C for 3 h and subsequently quenched with an equal volume of loading dye containing 7 M urea, 10 mM EDTA, and 20% (v/v) phenol. The samples were then electrophoresed on an 8 or 12% (w/v) polyacrylamide/7M urea gel at 30 mA for 1 h. After identification by autoradiography, the full-length RNAs were excised and eluted in 20 mM Tris-HCl (pH 7.5), 100 mM NaCl, 1 mM EDTA, 0.01% SDS (w/v). The elution was performed in two steps: 37°C for 3 h, and overnight at 4°C. The eluents were filtered through 0.22-μm cellulose acetate Spin-X centrifuge tube filters (Corning Costar, Corning, NY), and then subjected to ethanol precipitation. The radioactivity of the resulting RNA pellets was counted using a Bioscan QC-4000 XER radioisotope counter (Bioscan, Inc., Washington, D.C.) and the pellets were then typically resuspended in 10 to 20 μL of water.</p><!><p>The OPME approach was performed essentially as described earlier[4] but with two differences: the Xrn-1 incubation was performed exclusively at 37°C for 1.5 h, and the reactions were performed in a heat block.</p><!><p>To determine the fraction of pre-tRNACys initiated with either azG or GTP, we first performed "hot" IVTs and purified the full-length RNA as described above. Then, the RNAs were cleaved with recombinant AtPRORP3. The cleavage assays (10 μL) were performed by incubating internally-labeled pre-tRNACys RNA (~2000 dpm) with 2 μM AtPRORP3 in 20 mM HEPES (pH 7.5), 150 mM NaCl, 1 mM MgCl2, 4 mM DTT, and 5% (v/v) glycerol for 45 min at 25°C. Reactions were quenched by adding an equal volume of quenching [7 M urea, 20% (v/v) phenol, 0.2% (w/v) SDS, 10 mM EDTA, 0.05% (w/v) bromophenol blue, 0.05% (w/v) xylene cyanol] and then electrophoresed at 30 mA on a 20% (w/v) polyacrylamide/7 M urea gel (40 cm length x 20 cm width) for 1 h. The gel was then exposed overnight to a phosphorimager screen (at −80°C), and subsequently visualized using an Amersham Typhoon 5 Biomolecular Imager (GE Healthcare). Percent incorporation of azG was determined using ImageQuant 5.1 software-based quantitation of the slower (RNAs initiated with 5'-azG) and faster (RNAs initiated with 5'-triphosphate) migrating RNAs; these two RNA products were generated by AtPRORP3-mediated cleavage of pre-tRNACys.</p><!><p>To determine if AtPRORP3 would cleave an RNA·EGS hybrid (Figure 2B), MR1 and EGS1 mimicking part of mitochondrial At pre-tRNACys were first assembled to form a non-covalent bipartite substrate. These RNAs were synthesized by Sigma Aldrich. A mixture of 400 nM MR1 (spiked with 2,000 dpm 5'−32P-MR1 that was gel purified post-labeling with T4 polynucleotide kinase and γ-[32P] ATP) and 4 μM EGS1 were denatured at 95°C for 3 min, followed by annealing at 37°C for 10 min. Next, an equal volume of reaction buffer [for a final concentration of 20 mM HEPES-KOH (pH 7.5), 150 mM NaCl, 10 mM MgCl2 and 4 mM DTT] was added and the substrates were incubated for an additional 30 min at 25°C. The cleavage reaction was initiated by adding an equal volume of 8 μM AtPRORP3 in reaction buffer supplemented with a final concentration of 5% (v/v) glycerol. The cleavage assays (10 μL) were performed using final concentrations of 100 nM MR1, 1 μM EGS1, and 4 μM AtPRORP3. After defined periods of incubation at 25°C, the reaction was quenched using 1.5 volumes of quenching solution [7 M urea, 20% (v/v) phenol, 0.2% (w/v) SDS, 10 mM EDTA, 0.05% (w/v) bromophenol blue, 0.05% (w/v) xylene cyanol] and products were electrophoresed on a 20% polyacrylamide/7 M urea gel (20 cm length x 20 cm width) at 30 mA for 1 h. The product sizes were confirmed using for reference the product generated by 100 nM recombinant E. coli RNase P;[20] this reaction was performed at 37°C for 20 min in 10 mM HEPES-KOH (pH 7.5), 400 mM NH4OAc, 10 mM Mg(OAc)2, 5% (v/v) glycerol and 0.01% (v/v) Igepal-40.</p><p>To test if the PRORP-based assay could be used to determine the fraction of 5'-azG or 5'-GTP-initiated MR1/MR2 (Figure 2C and S7), we performed "hot" IVTs and purified the full-length RNAs as described above. The PRORP-based assay was performed as described above one change: cold MR1 or MR2 supplements were not included.</p>

PubMed Author Manuscript

Dimeric and Trimeric Catenation of Giant Chiral [8+12] Imine Cubes Driven by Weak Supramolecular Interactions

Mechanically interlocked structures, such as catenanes or rotaxanes are fascinating synthetic targets and are the basis of molecular switches and machines. Today, the vast majority of catenated structures are built upon macrocycles and only a very few examples of threedimensional shape-persistent organic cages forming such structures are reported. However, the catenation in all these cases was based on a thermodynamically favoured π-π stacking under certain reaction conditions. Here, we present our findings that catenane formation can be driven by even less directional dispersion (Keesom) interactions of methoxy-groups during the synthesis of chiral [8+12] imine cubes, giving dimeric and also for the first time trimeric catenated organic cages. To further elucidate the underlying driving forces, twelve differently 1,4-substituted benzene dialdehydes have been reacted with a chiral triamino tribenzotriquinacene under various conditions to study whether monomeric cages or catenated cage dimers are the preferred products.

dimeric_and_trimeric_catenation_of_giant_chiral_[8+12]_imine_cubes_driven_by_weak_supramolecular_int

Main<!>Results and Discussion<!>Conformational analysis by DFT calculations (Chapter 13 Supplementary Information) of<!>Conclusions<!>Synthesis of (OMe-cube)2

<p>Since the first report of Wasserman in the early 1960s of a catenane as a statistical occurring by-product during a macrocyclization via benzoin condensation, 1 the interest in interlocked molecular structures developed rapidly in the last decades, 2,3 especially because such compounds build the fundamental knowledge for molecular switches and machines. 4,5,6 Although Lüttringhaus and Schill introduced already rational synthetic approaches towards a number of interlocked structures in the 1960s, 7 the real ignition of this research field began with the work by Jean-Pierre Sauvage and coworkers with a high-yielding catenane synthesis exploiting the templated coordination of two molecular strands by a metal ion before closing these to two interlocked macrocycles via Williamson ether synthesis. 8,9 This concept of using a template was and still is the most frequently applied strategy for the synthesis of more complex interlocked structures such as borromean rings, 10 various knots, 11,12,13,14 a Star of David catenane, 15 poly[n]catenanes, 16 or interlocked coordination cages. 17,18,19,20,21 Besides ligand to metal ion coordination also weaker and less directing supramolecular interactions, such as hydrogen bonding or π-π stacking have been used to arrange molecular precursors in the right fashion to synthesize interlocked structures. 22 In contrast to the relative large number and diversity of interlocked coordination cages, 18 there are still only a few examples of purely organic cage catenanes reported to date. The first example was reported by Beer et al. 23 They exploited a template effect of sulfate anions, interacting with carbamate units to prearrange two tripodal precursor molecules in such a way that by the end-capping of these via a copper-mediated 1,3-dipolar cycloaddition a triply interlocked cage dimer was formed in 21% yield. One year later, in 2010, Cooper and coworkers described that by changing conditions for the synthesis of a [4+6] imine cage by adding catalytic amounts of trifluoroacetic acid to the reaction solution in acetonitrile or dichloromethane, these [4+6] imine cages form triply interlocked dimers, 24 which was proven by single-crystal X-ray diffraction. It was suggested that π-π stacking most probably is the driving force for the catenane formation and if a competing aromatic solvent was present in certain amounts, this indeed suppressed the catenane formation. In 2014, the formation of a quadruply interlocked dimer of giant [12+8] boronic ester cage was described, 25 which was clearly characterized by singlecrystal X-ray diffraction. The only difference between the interlocked cage dimer and a corresponding monomeric [12+8] boronic ester cage 26 published before is the position and length of solubilizing alkyl-chains in the molecular precursors, which led to the hypothesis that additionally weak dispersion interactions may be responsible for the catenane formation to overcome any entropic penalty. Similar but more distinct, this entropic penalty was balanced by dispersion interactions in the formation of a hydrocarbon cage and its catenated dimer made by alkyne metathesis. 27 Depending on concentration of reacting monomers, the equilibrium between monomeric and interlocked cage could be shifted towards the one or the other metathesis product. The authors assumed that a triply interlocked structure is energetically more favored than a singly one due to a maximization of filled space. 2015, Li et al. exploited the hydrophobic effect to achieve an interlocked cage dimer via a hydrazone bond formation in water. 28 Very recently, the group of Shaodong Zhang presented the formation of a triply interlocked catenane of a [2+3] imine cage. 29,30 Again, it was concluded that the driving force is the energetic benefits of additional π-π stacking. In contrast to the before mentioned examples, here the dimer formation has been studied more detailed by kinetic NMR measurements and time-dependent mass spectroscopy; however, no thermodynamic assumptions were corroborated by experiments.</p><p>During our ongoing work on using chiral triamino tribenzotriquinacenes (TBTQs) in the condensation with aromatic aldehydes to study self-sorting of cages, 31,32 we serendipity found an unprecedented substituent driven formation of dimeric and trimeric cage catenanes by very weak supramolecular interactions, which is described herein.</p><!><p>Inspired by Warmuth's chiral cube, 33 based on the condensation of eight molecules of cyclotriveratrylene (CTV) trisaldehyde and para-phenylene diamine we intended to use a chiral TBTQ precursor instead, which is in contrast to the CTV structurally fixed and cannot racemise during cage formation. Indeed, the condensation of enantiopure triamino TBTQ (P)-1 34 with 2,4-dihydroxy terephthalaldehyde 2 under typical conditions we used before for similar systems (cat. TFA, CDCl3 room temperature) 33,35 gave clean chiral [8+12] cage OH-cube in 88% isolated yield (Fig. 1) and was identified by NMR spectroscopy and mass spectrometry. Originally we were interested in post-stabilizing the OH-cube by Pinnick-oxidation to turn imine bonds into amide bonds. 36 As reported before, this does not work with the phenolic hydroxy groups present. To avoid a 24-fold post-synthetic Williamson etherification on OHcube, 37 we instead condensed TBTQ (P)-1 with dimethoxy terephthalaldehyde 3 under the same conditions (Fig. 2a). In contrast to the reaction with aldehyde 2, here the 1 H NMR spectrum of the crude product was very complex with a large number of peaks in the aromatic as well as in the aliphatic region (Fig. 2b). The corresponding MALDI-TOF MS revealed that beside the [8+12] OMe-cube (m/z = 5623.24), a [16+24] condensation product (m/z = 11245.57) was found and even a small peak with m/z = 16868.52 was detected (Fig. 2c), suggesting that a larger [24+36] species may have formed. Taking into consideration the complex 1 H NMR spectra reported for triply interlocked cages before, 24 it was assumed that these species are most likely catenated dimer (OMe-cube)2 and trimer (OMe-cube)3 rather than larger more symmetric and non-interlocked species. By applying recycling gel permeation chromatography (r-GPC) with dichloromethane as solvent, it was possible to separate the three compounds after multiple cycles (Fig. 2d, for details, see Supplementary Information). As described in literature before, the equilibrium between monomeric and catenated cage shifts towards the latter by increasing the concentration of reactants and vice versa to the monomeric cage by decreasing it. Therefore, a screening of the reaction at different concentrations (between 0.42 mM and 42.8 mM) was performed and analyzed mainly by MALDI-TOF MS (Supplementary Table 1). As expected, with higher concentration more catenated compounds (OMe-cube)2 and (OMe-cube)3 are found and the concentration needs to be 0.42 mM or below that to avoid the formation of those and to form monomeric cage OMe-cube exclusively. For comparison; reactions with dihydroxy terephthaldehyde 2 gave under no concentration (up to 42.8 mM) any catenated species at all and in each experiment only monomeric cage OH-cube was detected by 1 H NMR spectroscopy (Supplementary Fig. 248). It is worth mentioning that as soon as a solution of monomeric cage OMe-cube was concentrated by rotary evaporation (50 ºC, reduced pressure), the equilibrium immediately shifted towards the catenated products (OMe-cube)2 and (OMe-cube)3 as found by NMR and r-GPC analysis. On one hand, this clearly demonstrated the dynamic covalent chemistry character and thus thermodynamically driven formation of the catenane. 38 On the other hand, it made the separation and characterization of monomeric cage OMe-cube more challenging. Despite of these findings, we were able to develop a synthetic protocol to isolate OMe-cube in 85% yield, avoiding long reaction times, certain concentration and temperature thresholds and exploiting the low solubility of the cage in acetonitrile (Fig. 3a and Supplementary Information). On the other hand, running the reaction of 3 and 1 in dichloromethane instead of CHCl3 at 10 mM concentration and 80 °C for 3 days allowed us to push the equilibrium towards the tricatenane (OMe-cube)3, which was isolated in 80% yield (Fig. 3a). The best results for the dimeric cage (OMe-cube)2, was achieved, when dialdehyde 3 and triamine 1 were reacted at 10 mM scale.</p><p>However, it still needed to be separated by r-GPC from OMe-cube and (OMe-cube)3 at 35 o C to be obtained in 47% isolated yield (Fig. 3a).</p><p>After reinjection the once separated fractions each again, three distinct peaks each of nearly 187). In contrast to the relatively simple 1 H NMR spectrum of monomeric OMe-cube (Fig. 3h), the one of the [16+24] species was much more complex (Fig. 3i). Nevertheless, despite the large number of signals, most of them are sharp and did not superimpose, allowing a more detailed analysis of the structure (Fig. 4a and for detail structural analysis see Supplementary Information). By 2D NMR experiments, eight different types of imine-protons and eight different methoxy-protons were identified (Figs. 4b and 4c). This is exactly the number expected for a triply interlocked cage dimer (see model in Figs. 4h-i.) A singly interlocked dimer can clearly be ruled out. Here, two sets of twelve instead of eight peaks would be expected (Supplementary Fig. 329). By DOSY NMR in dichloromethane at 295 K only one trace of signals for (OMe-cube)2 confirmed that this is a single species. The diffusion coefficient D = 4.27 x 10 -10 m 2 s -1 corresponds to a solvodynamic radius of 12.3 Å (Supplementary Fig. 188). This is slightly larger than for OMecube (10.1 Å) which is consistence with its slightly larger size. The trimeric interlocked cage (OMe-cube)3 shows unfortunately much less resolved multiple broad peaks in the 1 H NMR spectrum in contrast to dimer (OMe-cube)2. Independent if this compound is a chain-like catenane (OMe-cube)@(OMe-cube)@(OMe-cube) or a multiple catenane [(OMe-cube)@(OMe-cube)]@(OMe-cube) (for models, see Supplementary Figs. 336 and 337) a total number of 72 imine proton peaks are expected due to the lack of any applicable symmetry operations (C1-symmetry). The same is true for all other chemically 'equivalent' protons making it very difficult or even impossible to distinguish between the two possibilities. Furthermore, the generation of positional isomers cannot strictly be ruled out.</p><p>However, since the trimeric catenane (OMe-cube)3 is still very good soluble under reaction conditions and no larger oligomers such as tetrameric and pentameric cages (OMe-cube)4 or (OMe-cube)5 are found by mass spectrometry, it seems to be more likely a multiple catenane [(OMe-cube)@(OMe-cube)]@(OMe-cube) and not a chainlike triple catenane (OMecube)@(OMe-cube)@(OMe-cube). If it would be the latter motif, we would expect at least some formation of longer oligomers. On the other hand, a multiple tetrameric catenane is out of steric reasons simply not possible, which once more favors the multiple catenane [(OMecube)@(OMe-cube)]@(OMe-cube) model. DOSY NMR of (OMe-cube)3 again shows a single trace with a diffusion coefficient D = 4.27 x 10 -10 m 2 s -1 . The calculated solvodynamic radius of 11.2 Å was found to be almost similar with the dicatenane (OMe-cube)2 (12.3 Å) once more suggesting a tightly packed interlocked structure (Figure S189). It is worth mentioning that for OH-cube, OMe-cube, (OMe-cube)2 as well as (OMe-cube)3 innumerous large single-crystals from various solvents have been obtained. Unfortunately, even with synchrotron radiation no resolution has been obtained to elucidate the solid state structures.</p><p>We were interested to get further insight into the driving force of the unique catenation of methoxy cage OMe-cube to dimer (OMe-cube)2 and even to trimer (OMe-cube)3 and why we do not see any such catenation for the hydroxyl substituted OH-cube under any concentration.</p><p>Due to the triply interlocked catenation of dimer (OMe-cube)OMe-cube) in favour of a possible singly interlocked dimer (OMe-cube)-(OMe-cube), the aforementioned π-π stacking as driving force, found for almost all other yet in literature described interlocked organic cages, was excluded (see above), otherwise singly interlocked catenation should have been formed preferably. If π-π stacking would have been the driving force, for OH-cube a higher tendency of dimerization would have been expected than for OMe-cube, because intramolecular Hbonding of the hydroxyl imine is stiffening the π-backbone and strongly enhances intermolecular π-π stacking. 39 This assumption is strengthened by the fact that under various reaction conditions (different acid concentration, different concentration of reactants, different solvents, different and elevated temperature, different reaction times (up to several months!) no catenane formation was found for OH-cube (Supplementary Information). A kinetic formation driven by precipitation was also ruled out, because the reaction mixture of 1, 2 and OH-cube were at all times clear solutions. 38 Since π-stacking was ruled out as driving force, it was hypothesized that dipole-dipole induced dispersion interactions (so-called Keesom interactions) 40 of the methoxy groups are responsible for the catenation as e.g. found in single crystals of methoxy-substituted π-systems 41 In this respect, it is worth mentioning that Cooper et al. described the unexpected formation of a knot, when originally achieving cages based on dimethoxy terephthaldehyde 3, 42 which may rely on the same weak interactions. Indeed, a closer look at the X-ray structure show the same methoxy methoxy interacting motif, albeit with a larger distance between the functional groups of d(MeO•••CH3O) = 3.5 Å (Supplementary Fig. 330).</p><!><p>OMe-cube as well as NOESY cross peaks of imine CH and the aromatic TBTQ protons revealed a low barrier of rotation of the linker units at room temperature, which is also present in the triply interlocked dimer (OMe-cube)2 allowing the mechanically interlocked molecule to adopt conformations that have three such methoxy-methoxy interactions (Fig. 5c).</p><p>If methoxy groups are absent, no catenane formation should occur. Thus triamine 1 was reacted with non-substituted terephthalaldehyde 4 (Fig. 5a) under different conditions (various solvents, Supplementary Fig. 252) and no catenane formation was observed. Pure H-cube was isolated in 90% from THF. By adding two methyl substituents instead of two methoxy groups to the aldehyde (5) still almost no catenane formation is observed by 1 H NMR (Supplementary Fig. 253) and monomeric Me-cube is formed in 84% yield. As soon as the alkyl substituents at the dialdehyde precursor (6) get longer (here ethyl), the possibility of intermolecular dispersion interactions 43 (Fig. 5b) is increased and now some catenane (Et-cube)2 was found by 1 H NMR spectroscopy as well as MS (Supplementary Fig. 254) besides monomeric Et-cube (which still is the main product). Comparing the different results of Me-cube versus Et-cube, based on the simple elongation of the alkyl chains by one methylene unit each, electronic effects to foster ππ-stacking can be ruled out, because the methyl-as well as the ethyl-substituents have almost the same Hammett parameters (σm(Me) = -0.07; σm(Et) = -0.07; σp(Me) = -0.17; σp(Et) = -0.15). 44 As for OMe-cube and (OMe-cube)2, the ratio of catenane (Et-cube)2 versus monomeric cage Et-cube was also strongly solvent dependent for the reaction of triamine 1 and aldehyde 6 and in THF the amount of catenane was higher than e.g. in CHCl3 and both compounds (monomer and catenane) were selectivity achieved by adjusting the conditions. The reaction in CHCl3 at room temperature gave monomeric cage Et-cube in 75% isolated yields, whereas running the reaction in THF gave after separation 35% of the catenated dimer (Etcube)2 in pure form.</p><p>To further exclude pure electronic effects, we reacted triamine 1 with diethoxy-and diisopropoxy dialdehydes 7 and 8, (Fig. 5a) where the substituents have comparable Hammett parameter as in dimethoxy dialdehyde 3 (σm(OMe) = 0.12; σm(OEt) = 0.10; σm(O i Pr) = 0.10), but are of different steric demand. Whereas for the diethoxy dialdehyde 7 some catenane formation of (OEt-cube)2 was observed, for diisopropoxy dialdehyde 8 no catenane (OiPrcube)2 occurred (Supplementary Fig. 255 and 256), supporting once more the hypothesis that the catenane formation is mainly driven by additional weak dipole-dipole or dispersion interactions and in case of the latter steric repulsion is stronger than the weak attraction (Figure 5b, Charton steric parameter for Me, Et, and i Pr are νMe = 0.52; νEt = 0.56; νiPr = 0.76). 45 By increasing these weak interactions, the equilibrium may be shifted towards the interlocked structures. Sulfur containing organic compounds are known to interact via sulfur-sulfur interactions. 46 And indeed, by using dimethylthioether 9 in the condensation with triamine 1 in CDCl3 almost exclusively the catenated dimer (SMe-cube)2 was formed (Fig. 5a and Supplementary Fig. 257). Again, to rule out electronic effects based on the thioalkyl substituent donating to the aromatic dialdehyde, di-tert-butylthioether substituted dialdehyde 10 with two sterically demanding tert-butyl groups was investigated in the reaction (νMe = 0.52 vs νtBu = 1.24). 45 As expected, only clean monomeric SC(CH3)3-cube was formed and isolated in 75% yield (Supplementary Fig. 258). Finally, we investigated the reaction of dibromo dialdehyde 11 with triamine 1, to see whether halogen bond formation 47 can also induce catenation. Although the mass spectrum of the reaction mixture in CD2Cl2 showed a pronounced peak at m/z = 13591.6, which is the double amount of the monomeric Br-cube (m/z = 6796.4), in the correlated 1 H NMR spectrum only small detectable peaks of any interlocked species are present besides mainly those signals of pure monomeric Br-cube (Supplementary Fig. 259). However, in contrast to all other reactions, here a precipitate was formed of very low solubility, which may contain insoluble (Br-cube)2.</p><p>To correlate the weak interactions responsible for catenation, the systems where catenation occurred have been studied by concentration dependent NMR spectroscopy (see Supplementary information), to estimate the Gibb's enthalpy of cage to catenane transformation. With ΔG298 = -26.7 kJ/mol the reaction of 2 SMe-cube → (SMe-cube)2 is about 6 kJ/mol higher as for the methoxy cages 2OMe-cube → (OMe-cube)2, ΔG298 = -20.8 kJ/ mol) and almost 10 kJ/mol higher than found for the ethoxy cages 2 OEt-cube → (OEt-cube)2, ΔG298 = -15.7 kJ/mol) (Fig. 5d), which is the same trend as expected for these weak interactions. 46 Unfortunately, the amount of (Et-cube)2 besides Et-cube was too small to determine reliable numbers by this method.</p><!><p>An unprecedented dimeric and trimeric cage catenane formation based on week (and a priori non-directing) local dipole-dipole-interactions (so called Keesom interactions) of methoxysubstituents has been observed. By varying the substituents of the used 1,4-benzene dialdehydes electronically as well as by steric demand, the hypothesis was strengthened, because as soon as different alkoxy-substituents with similar Hammett parameters but of various bulkiness were present, those dialdehydes with bulky substituents (such as i OPr) did not form any catenanes, whereas those with methoxy-and ethoxy-substituents did. Changing the methoxy groups to less polar ethyl groups decreased catenane formation significantly, because now the dispersion interaction is decreased. In case there is only a methyl-or no substituent at the dialdehyde the intermolecular forces are too week to foster catenane formation. Finally, we concluded that dialdehydes with substituents that can undergo other (stronger) interactions such as chalcogenchalcogen or halogen-halogen bond formations, should be beneficial to catenane formation and indeed with thiomethyl substituents a clear reaction to (SMe-cube)2 was observed, having a ~6 kJ/mol higher Gibb's enthalphy for catenane formation than the (OMe-cube)2.</p><p>This motif of week dispersion interactions as driving force for catenation of shape-persistent organic cages allow to further study the influence of subtle structural changes to understand events of dynamic covalent chemistry of larger and more complex structures as well as to construct e.g. poly[n]catenated cages with n > 3.</p><!><p>To a solution of TBTQ 1 (20 mg, 0.0429 mmol, 1 equiv) and 2,5dimethoxy-terephthaldehyde 3 (12.6 mg, 0.0644 mmol, 1.5 equiv) in chloroform (deuterated, 4 mL) in a screw-capped 8 mL glass vial, a catalytic amount of TFA (0.4 µL, 0.0052 mmol, 0.01 equiv, 12 mol%) was added and the reaction mixture was stirred at RT for 3 days. Afterwards, the crude reaction mixture was washed with aq. K2CO3 solution (0.25 M, 3 × 2 mL), dried over</p>

ChemRxiv

Structure-Activity Relationship Studies to Identify Affinity Probes in Bis-aryl Sulfonamides that Prolong Immune Stimuli

Agents that safely induce, enhance, or sustain multiple innate immune signaling pathways could be developed as potent vaccine adjuvants or co-adjuvants. Using high-throughput screens with cell-based nuclear factor kappa B (NF-\xce\xbaB) and interferon stimulating response element (ISRE) reporter assays, we identified a bis-aryl sulfonamide bearing compound 1 that demonstrated sustained NF-\xce\xbaB and ISRE activation after a primary stimulus with lipopolysaccharide or interferon-\xce\xb1, respectively. Here, we present systematic structure-activity relationship (SAR) studies on the two phenyl rings and amide nitrogen of the sulfonamide group of compound 1 focused towards identification of affinity probes. The murine vaccination studies showed that compounds 1 and 33 when used as co-adjuvants with monophosphoryl lipid A (MPLA) showed significant enhancement in antigen ovalbumin-specific immunoglobulin responses compared to MPLA alone. SAR studies pointed to the sites on the scaffold that can tolerate the introduction of aryl azide, biotin and fluorescent rhodamine substituents to obtain several affinity and photoaffinity probes which will be utilized in concert for future target identification and mechanism of action studies.

structure-activity_relationship_studies_to_identify_affinity_probes_in_bis-aryl_sulfonamides_that_pr

Introduction:<!>Results and Discussion:<!>Conclusions<!>Materials.<!>Instrumentation.<!>General procedure A for the syntheses of select site A and site B modified compounds.<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxybenzenesulfonamide (1).<!>N-(4-chloro-3-methoxyphenyl)-4-ethoxybenzenesulfonamide (4).<!>N-(4-chloro-2-methoxyphenyl)-4-ethoxybenzenesulfonamide (5).<!>N-(2,5-dimethoxyphenyl)-4-ethoxybenzenesulfonamide (6).<!>N-(4-bromo-2,5-dimethoxyphenyl)-4-ethoxybenzenesulfonamide (7).<!>N-(4-chloro-3,5-dimethoxyphenyl)-4-ethoxybenzenesulfonamide (8).<!>N-(2,5-dimethoxy-4-nitrophenyl)-4-ethoxybenzenesulfonamide (9).<!>N-(4-amino-2,5-dimethoxyphenyl)-4-ethoxybenzenesulfonamide (10).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-hydroxybenzenesulfonamide (11).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-methoxybenzenesulfonamide (12).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-propoxybenzenesulfonamide (13).<!>4-Butoxy-N-(4-chloro-2,5-dimethoxyphenyl)benzenesulfonamide (14).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-(prop-2-yn-1-yloxy)benzenesulfonamide (15).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-propylbenzenesulfonamide (16).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-nitrobenzenesulfonamide (17).<!>4-Amino-N-(4-chloro-2,5-dimethoxyphenyl)benzenesulfonamide (18).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-cyanobenzenesulfonamide (19).<!>tert-butyl (4-(N-(4-chloro-2,5-dimethoxyphenyl)sulfamoyl)benzyl)carbamate (20).<!>4-(aminomethyl)-N-(4-chloro-2,5-dimethoxyphenyl)benzenesulfonamide (21).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-phenoxybenzenesulfonamide (22).<!>N-(4-chloro-2,5-dimethoxyphenyl)-3-methoxybenzenesulfonamide (23).<!>N-(4-chloro-2,5-dimethoxyphenyl)-2-methoxybenzenesulfonamide (24).<!>3-Bromo-N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxybenzenesulfonamide (25).<!>Methyl 4-(N-(4-chloro-2,5-dimethoxyphenyl)sulfamoyl)benzoate (26).<!>4-(N-(4-chloro-2,5-dimethoxyphenyl)sulfamoyl)benzamide (27).<!>4-(N-(4-chloro-2,5-dimethoxyphenyl)sulfamoyl)benzoic acid (28).<!>Ethyl 4-(N-(4-chloro-2,5-dimethoxyphenyl)sulfamoyl)benzoate (29).<!>4-(N-(4-chloro-2,5-dimethoxyphenyl)sulfamoyl)-N-methylbenzamide (30).<!>4-Chloro-N-(4-ethoxyphenyl)-2,5-dimethoxybenzenesulfonamide (31).<!>General procedure B for the syntheses of site C modified compounds 33\xe2\x80\x9347 and 49\xe2\x80\x9352.<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxy-N-methylbenzenesulfonamide (33).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxy-N-propylbenzenesulfonamide (34).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxy-N-butylbenzenesulfonamide (35).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxy-N-pentylbenzenesulfonamide (36).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxy-N-hexylbenzenesulfonamide (37).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxy-N-heptylbenzenesulfonamide (38).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxy-N-dodecylbenzenesulfonamide (39).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxy-N-isopropylbenzenesulfonamide (40).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxy-N-isobutylbenzenesulfonamide (41).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxy-N-(prop-2-yn-1-yl)benzenesulfonamide (42).<!>N-(but-3-yn-1-yl)-N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxybenzenesulfonamide (43).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxy-N-(pent-4-yn-1-yl)benzenesulfonamide (44).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxy-N-(2-(2-(2-(prop-2-yn-1yloxy)ethoxy)ethoxy)ethyl)benzenesulfonamide (45).<!>N-benzyl-N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxybenzenesulfonamide (46).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxy-N-phenethylbenzenesulfonamide (47).<!>N-(4-chloro-2,5-dimethoxyphenyl)-N-((4-ethoxyphenyl)sulfonyl)acetamide (48).<!>Ethyl N-(4-chloro-2,5-dimethoxyphenyl)-N-((4-ethoxyphenyl)sulfonyl)glycinate (49).<!>Ethyl 4-((N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxyphenyl)sulfonamido)butanoate (50).<!>N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxy-N-(3-hydroxypropyl)benzenesulfonamide (51).<!>tert-Butyl (3-((N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxyphenyl)sulfonamido)propyl)carbamate (52).<!>4-((N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxyphenyl)sulfonamido)butanoic acid (53).<!>4-((N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxyphenyl)sulfonamido)-N-ethylbutanamide (54).<!>N-(3-aminopropyl)-N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxybenzenesulfonamide hydrochloride (55).<!>N-(6-((3-((N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxyphenyl)sulfonamido)propyl)amino)-6-oxohexyl)-3\xe2\x80\x99,6\xe2\x80\x99-dihydroxy-3-oxo-3H-spiro[isobenzofuran-1,9\xe2\x80\x99-xanthene]-5-carboxamide (56).<!>N-(9-(2-carboxy-4-((3-((N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxyphenyl)sulfonamido)propyl)carbamoyl)phenyl)-6-(diethylamino)-3H-xanthen-3-ylidene)-N-ethylethanaminium (57).<!>N-(3-((N-(4-chloro-2,5-dimethoxyphenyl)-4-ethoxyphenyl)sulfonamido)propyl)-5-((3aR,4R,6aS)-2-oxohexahydro-1H-thieno[3,4-d]imidazol-4-yl)pentanamide (58).<!>N-(4-azido-2,5-dimethoxyphenyl)-4-ethoxybenzenesulfonamide (59).<!>3-Amino-N-(4-chloro-2,5-dimethoxyphenyl)-4-methoxybenzenesulfonamide (60).<!>3-Azido-N-(4-chloro-2,5-dimethoxyphenyl)-4-methoxybenzenesulfonamide (61).<!>N-(4-azido-2,5-dimethoxyphenyl)-4-ethoxy-N-(prop-2-yn-1-yl)benzenesulfonamide (62).<!>N-(3-aminopropyl)-N-(4-azido-2,5-dimethoxyphenyl)-4-ethoxybenzenesulfonamide (63).<!>N-(3-((N-(4-azido-2,5-dimethoxyphenyl)-4-ethoxyphenyl)sulfonamido)propyl)-5((3aR,4R,6aS)-2-oxohexahydro-1H-thieno[3,4-d]imidazol-4-yl)pentanamide (64).<!>Biology: Cell lines and reagents<!>Measurement of NF-\xce\xbaB activation using THP1-Blue\xe2\x84\xa2 NF-\xce\xbaB cells<!>Measurement of ISRE activity in ISRE-bla THP-1 cells<!>Cell viability assay<!>Animals<!>In vivo adjuvant activity study<!>Statistical analysis

<p>Vaccines consisting of antigen and adjuvant rely primarily on adjuvants for enhancement of immune stimuli.1 These adjuvants include ligands for pattern recognition receptors (PRRs) such as Toll-like receptors (TLR) −2, −4, −7, −8, and −9, nucleotide-binding oligomerization domain-like receptors (NLRs), RIG-I-like receptors (RLRs) and cytokines such as interferon-α (IFN-α), IFN-γ and IL-12.2–13 Some of these adjuvants have been approved for human use by the U.S. Food and Drug Administration (FDA) including the TLR-4 agonist monophosphoryl Lipid A (MPLA),14 the TLR-9 agonist CpG 1018,15 and other adjuvants with different mechanisms of action such as alum and squalene based adjuvants. Despite the availability of approved adjuvants, the need for co-adjuvants is evident since single adjuvant vaccines often do not generate long lasting protective immunity.16 Alum has been used as an effective single adjuvant for decades primarily due to its safety record and induction of increased humoral immunity;17, 18 however it induces only weak cellular immunity and predominantly a T helper (Th) type 2 associate response; whereas in some cases a T helper type 1 response would be more effective for protection. In addition, it is not always sufficient for vaccinating immunocompromised and elderly populations.17, 19 A co-adjuvant is a substance that may or may not be an adjuvant by itself but can work with a known adjuvant to offer synergistic effects such as enhanced antibody response. For example, IL-2 has been shown to be a co-adjuvant with alum-adsorbed hepatitis B vaccine.20 Similarly, combination adjuvants can be obtained using a PRR or NLR ligand, an immunogenic protein, a delivery system or another adjuvant with a complementary mechanism of action.16 One such combination AS04 (Adjuvant System 04), consisting of MPLA and alum, has been FDA approved in a hepatitis B vaccine Fendrix® and human papillomavirus vaccine Cervarix®.21–24 Alternative combinations involving approved adjuvants, TLR agonists, NOD agonists, and delivery systems are being explored.25–28</p><p>Our approach towards identifying novel co-adjuvants focused on small molecules that may not lead to immune activation by themselves but may enhance the primary immune activation such as nuclear factor kappa B (NF-κB) or IFN stimulating response element (ISRE) activation induced by a TLR-4 agonist (LPS or MPLA). The rationale behind the approach is as follows: Upon vaccine administration, local antigen presenting cells (APCs) at the site of injection, such as dendritic cells and Langerhans cells, are activated by the TLR-4 agonist. These APCs engulf antigen and travel to local draining lymph nodes where the antigen is presented to T cells.29 The activation levels of APCs induced by a TLR-4 agonist peaks at 2–6 hours and then decays due to negative feedback mechanisms.30–37 Because it takes approximately 12–24 hours for an APC to travel to the lymph node after vaccination,38 APCs are arriving during the decay phase of the activation. This rationale is well supported from a report that showed that the absence of interleukin-1 receptor associated kinase M (IRAK-M, a negative regulator of TLR signaling)39 increases NF-κB activation, improves migration of dendritic cells (DCs) to lymph nodes thereby increasing the lifespan of the activated DCs and secretion of Th1-skewed cytokines and chemokines.31 Thus, we hypothesize that prolonging or sustaining the activation of APCs induced by the TLR-4 agonist for 12–24 hours will lead to optimal presentation of antigen to the T cells which would enhance the initial immune response and potentially allow for a longer lasting response. Our hypothesis is supported by reports that enhanced responses to vaccinations were observed in mice with genetic disruption of either IRAK-M, an inhibitor of the NF-κB pathway,31 or of UBP43, a negative regulator of type 1 IFN signaling.40 Thus, to address this issue, we sought HTS methods directed towards identification of co-adjuvants that prolonged activation of an immune response induced by a primary stimulus.41, 42</p><p>These cell based HTS campaigns tested protraction of a TLR-4 agonist lipopolysaccharide (LPS) stimulus through the NF-κB pathway41 or of IFN-α signaling via the interferon stimulating response element (ISRE)42 pathway. Compounds that prolonged LPS induced NF-κB signaling included a distinct set of pyrimido[5,4-b]indoles that were also found to be effective co-adjuvants with MPLA, an FDA approved adjuvant, in murine vaccination studies.41 In parallel, compounds that prolonged IFN-α induced ISRE signaling in vitro were also evaluated as co-adjuvants in vivo42 which led to identification of a potent bis-aryl sulfonamide compound 1 (Fig. 1) bearing 4-chloro-2,5-dimethoxy and 4-ethoxy substituted phenyl groups connected by a sulfonamide functional group. Compound 1 did not possess any NF-κB or ISRE activity when tested alone but it enhanced their activation when tested in presence of LPS or IFN-α respectively, compared to the stimulus alone. The further drug development of such hits identified through cell-based phenotypic assays and involved in cell signaling pathways is hampered without the knowledge of the target receptor or the compound's mechanism of action.43</p><p>This necessitated SAR studies focused towards identification of affinity probes which involved evaluation of structural variations for compound 1 that were unexplored in the HTS with an aim to identify positions on the scaffold that can tolerate the introduction of small functional groups such as aryl azide or diazirine to make photoreactive probes or large substituents such as biotin and fluorescence moieties to generate affinity and fluorescent probes, respectively.44–50</p><p>Exploration of several different functional groups and substituents will allow us to systematically identify the position and size of the affinity probe as well as the reactive handle to be used for introducing these probes. These chemical probes would then be useful tools for future mechanistic and functional receptor studies. In addition, the chemical handle would allow one to covalently conjugate the small molecule to peptides or protein antigens to make self-adjuvanting vaccine constructs which are widely explored in vaccine development.51–56</p><!><p>Approximately 3400 differently substituted bis-aryl sulfonamide compounds were screened in the original HTS libraries and a scatter plot showing activation data for these compounds in both cell-based NF-κB and ISRE assays is shown in Supporting Information Fig. S1. These results provided preliminary SAR analysis indicating the substituents on the two aryl rings necessary for activity and pointed to compound 1 as an advanced lead. Hence, we approached further SAR studies on compound 1 by first identifying three areas (sites A, B and C) of potential modification as shown in Fig. 1. To standardize the reaction, we began with synthesis of compound 1 by reaction of 4-ethoxysulfonyl chloride (3a) and 4-chloro-2,5-dimethoxy aniline (2a) in the presence of an organic base (Scheme 1). However, the reaction not only provided the desired compound 1, but also formed the bis-sulfonamide side-product in high yields. This undesired side-product was formed in situ by further reaction of compound 1 with another equivalent of 4-ethoxysulfonyl chloride (3a). We were able to isolate this bis-sulfonamide side-product but observed that it was somewhat unstable. Limited hydrolysis by lithium hydroxide facilitated the complete conversion of this bis-sulfonamide side-product to compound 1 without further hydrolysis of the mono-sulfonamide bond thereby improving reaction yields for compound 1 (Scheme 1). This reaction strategy was utilized for synthesis of several site A and site B modified compounds for SAR analysis.</p><p>SAR studies were initiated by modifying the substituents at site A (Fig. 1). These compounds were synthesized according to Scheme 1 using different anilines (2a-g). We probed the removal of one aryl substituent at a time to obtain compounds 4, 5, and 6 lacking the 2-methoxy, 3-methoxy and 4-chloro substituent, respectively. Replacement of 4-chloro by a 4-bromo substituent gave compound 7 and migration of the 2-methoxy substituent to the 3-position gave compound 8. These compounds were evaluated for sustained activation of both NF-κB and ISRE pathways using LPS and IFN-α as primary stimuli, respectively. The SAR studies pointed to the importance of the methoxy substituents at the 2 and 5 positions of the aryl ring, because either removal of any one of the substituents as in compound 4 and 5 or its displacement to another position on the ring as in 8 led to complete loss of activity. Removal of the 4-chloro as in compound 6 or its replacement with a spatially larger bromo substituent as in compound 7 retained activity (Table 1). Thus, to further explore position 4 on the phenyl ring, we synthesized analogs with 4-nitro (9) substitution and its 4-amino (10) derivative. However, both these analogs were inactive suggesting that only hydrophobic substituents at this site are tolerated (Table 1).</p><p>Next, we focused on site B as shown in Fig. 1. The compounds were synthesized as discussed earlier (Scheme 1) using different aryl sulfonyl chlorides (3a-p) and 4-chloro-2,5-dimethoxy aniline (2a). Some of the aryl sulfonyl chlorides were commercially available, while the others were synthesized as shown in Supporting Information Scheme S1. We first probed the homologous series of 4-O-alkylated compounds starting with 4-hydroxy analog 11, 4-methoxy analog 12, 4-propoxy analog 13 and 4-butoxy analog 14 compared to 4-ethoxy analog compound 1. Bioactivity evaluation of these compounds showed that only the smaller homolog as in 4-methoxy compound 12 was tolerated while the hydrophilic interaction with hydroxy group of 11 without any hydrophobic alkyl group was not tolerated. The higher 4-alkoxy chains showed gradual loss of activity (Table 2). While the 4-propoxy substituted compound was weakly active, the 4-propargyloxy compound 15, designed to use the alkyne as a handle for click chemistry reactions, was found to be inactive. Removal of the ether oxygen to obtain 4-propyl substituted compound 16 also led to loss of activity suggesting a crucial role of hydrogen bond interaction by the ether oxygen. Other functional groups that could be involved in such hydrogen bond interactions led to the syntheses of 4-nitro analog 17 and its amine bearing derivative 18 (Scheme 2) obtained by reduction of the nitro group. Also, the 4-nitrile analog 19, N-Boc methylamine derivative 20 obtained by in situ N-Boc protection during the reduction of the nitrile group and its free methylamine derivative 21 (Scheme 2) were synthesized. All these compounds were also evaluated but found to be either weakly active or completely inactive. A prior report indicated that analogs bearing a 4-O-phenyl substitution exhibited ubiquitin ligase inhibition activity,57 so we synthesized the 4-O-phenyl analog 22, but this compound was inactive. Encouraged by the activity of 4-methoxy substituted analog 12, we synthesized 3-methoxy and 2-methoxy substituted compounds 23 and 24, respectively. However, none of these molecules was active. In order to find an additional handle for modification, bromine was introduced to obtain a 3-bromo-4-methoxy substituted compound 25, which was also found to be inactive. Learning from the requirement of a hydrogen bonding functional group at site B for activity, we probed the addition of another oxo-containing group to obtain the 4-methylester analog 26 and an amide analog 27. Ester hydrolysis of compound 26 yielded the 4-carboxyl derivative 28 (Scheme 2). While the methyl ester bearing analog 26 was active, the hydrolyzed carboxylic acid analog 28 and the amide linked compound 27 lost activity (Table 2).</p><p>Hypothesizing that the lack of hydrophobic alkyl group interaction could be a cause for the loss of activity, compound 28 was further derivatized to obtain the ethyl ester analog 29, and the N-methylamide analog 30 (Scheme 2). While analog 29 retained partial activity, compound 30 was completely inactive suggesting that only hydrogen bond accepting substituents were tolerated (Table 2). An additional analog (compound 31, Scheme 1) was synthesized by inversing the sulfonamide bond obtained by reaction of 2-ethoxyaniline and 4-chloro-2,5-dimethoxybenzenesulfonyl chloride, but the inactivity of this analog suggested that the positioning of the sulfonamide functional group was also critical for activity.</p><p>Moving forward, we wanted to probe the expansion at site C on the nitrogen of the sulfonamide function of compound 1. These compounds were synthesized by derivatization of compound 1 as shown in Scheme 3. The first extensive series of compounds were the N-alkylated derivatives including N-methyl (33), N-propyl (34), N-butyl (35), N-pentyl (36), N-hexyl (37), N-heptyl (38), and N-dodecyl (39). A clear correlation of bioactivity with the alkyl chain length was observed with potency gradually decreasing with increased alkyl chain length and compounds bearing alkyl chain lengths greater than N-pentyl were completely inactive (Table 3). We also probed the effect of steric bulk around the core structure by synthesizing N-isopropyl (40) and N-isobutyl (41) derivatives. Steric bulk closer to the core structure, as in compound 40, eliminated the NF-κB activity while retaining ISRE activity. In contrast, spacing the isopropyl group away by one methylene unit as in compound 41 regained the activity in both the NF-κB and ISRE assays. Encouraged by these results, we desired to synthesize alkyne bearing compounds with an additional aim to utilize the functional group as a biorthogonal reactive site. A homologous series of alkyne bearing molecules including N-propargyl (42), N-butynyl (43), and N-pentynyl (44) were synthesized (Scheme 3). Activity data showed that while N-alkyl derivatization with increasing alkyl chain length led to dramatic loss of activity, the corresponding N-alkynyl derivatives retained activity almost equivalent to that of compound 1 (Fig. 2, Table 3) for the corresponding alkyl chain length. As shown in Fig. 2, the retention of activity for the N-alkynyl compounds compared to loss in activity for the analogous N-alkyl derivatives for the same carbon unit chain length suggested the possible involvement of π-π interactions in near proximity with the target receptor(s). We were therefore interested in evaluation of a triethyleneglycol linked alkyne derivative (45) which would allow us to conveniently place the reactive functional group distant from the core. However, the 12-atom chain length equivalent to N-dodecyl compound 39 was too long to retain activity.</p><p>These results for the alkyne bearing compounds led us to propose making compounds where substituents can form enhanced π-π interactions. Thus N-benzyl (46) and N-phenethyl (47) derivatives were synthesized and were also found to be potent analogs (Table 3). Since the N-isopropyl analog 40 was inactive, we desired to determine if steric bulk was the only reason for its inactivity and if that could be mitigated by some hydrogen bonding functional group such as acetyl. Thus, the N-acetyl derivative (48) was synthesized and the bioactivity assays showed that the compound was active. However, before proceeding with syntheses of additional acylated analogs, we wanted to evaluate its stability in stock solutions since during the assay this compound could behave as a prodrug by undergoing deacetylation to release active compound 1. While the stock of compound 48 in DMSO was stable, incubation of compound with assay media showed release of compound 1 (data not shown), suggesting that the bioactivity could be due to a prodrug effect and not true interaction with the receptor. Thus, syntheses of additional acylated analogs were not pursued.</p><p>Since the hydrophobic alkyl and alkynyl groups were well tolerated at site C, we examined if incorporating a hydrophilic group that could serve as a handle for further chemical modification would be acceptable for activity. A pair of compounds bearing a precursor to a reactive handle such as carboxylic esters were synthesized by alkylation of compound 1 to obtain the N-ethyl glycinate (49) and N-ethyl butanoate (50) analogs (Scheme 3). Our attempts to make a stable propionate analog failed after several attempts likely due to retro Michael type reaction, and despite isolating a few milligrams of the tert-butyl propionate ester derivative, activity studies were not pursued due to stability concerns. Both the ethyl ester substituted compounds 49 and 50 retained dual NF-κB and ISRE activities (Table 3). To avoid additional substitution closer to the core sulfonamide pharmacophore, we chose a propylene spacer for further analogs. A terminal hydroxy bearing analog as in N-propan-3-ol (51) and the N-Boc protected aminopropane analog (52) were then synthesized. The ethyl ester of compound 50 was de-esterified using lithium hydroxide to obtain its carboxylic acid analog 53, which was converted to the ethyl amide analog 54 (Scheme 4). Similarly, a free amine bearing molecule was obtained by N-Boc deprotection of 52 to obtain compound 55. Biological evaluation showed that the terminal hydroxy analog 51 retained activity in both the assays while the N-Boc protected compound 52 showed reduction in activity, which was recovered when the N-Boc group was removed as in compound 55. Both the free carboxylic acid and ethyl amide derivatives retained activity, which was more skewed towards the NF-κB pathway (Table 3).</p><p>All these compounds were evaluated for toxicity using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assays. All the active compounds showed viability between 69% and 81%. Some of the inactive compounds were completely non-toxic. Compound 47 with the N-phenethyl substitution was an exception showing somewhat higher toxicity (% viability = 44%) suggesting that an aryl group connected by an ethylene unit near the core sulfonamide structure may lead to toxicity (Tables 1–3).</p><p>The bioactivity data from both the assays for all the compounds were plotted to verify the correlation between the chemical structure and bioactivity. Most of the compounds were active in both NF-κB and ISRE bioassays and showed a good correlation (Pearson two-tailed, R2=0.6812, P<0.0001, Fig. 3). The SAR trends however varied depending on the site of modification. Site A modifications involving removal of the methoxy substituent (compounds 4 and 5) led to significant loss of activity (Fig. 3). On the other hand, nonpolar modifications at position 4 of site A (compounds 7 and 8) showed slightly skewed ISRE activity compared to compound 1, while hydrogen bond forming substituents at this position led to loss of activity (compounds 9 and 10). Most of the site B modified compounds were inactive suggesting restricted SAR tolerance due to limited spatial availability in the target receptor. Only short alkyl groups connected via ether-linkage as in compounds 1, 12 and 13 or carboxyl (ester)-linkage as in compounds 26 and 29 retained activity. A good correlation was seen, however, between the two assays for these compounds. In contrast, most of the site C modified compounds were active in both the bioassays suggesting that only a part of the substituent may be involved in receptor interaction and the rest of the group subtends out of the target receptor(s). A notable variation was observed in sterically hindered bulky groups close to the core structure as in compound 40 which led to a loss of NF-κB activity, while still retaining ISRE activity. On the other hand, another subset of compounds bearing a reactive handle such as carboxylic acid analog 53 and its amidated derivative 54, showed reduction in ISRE activity while retaining the NF-κB activity (Fig. 3). This suggested that a negative charge on the compound may be a deterrent for ISRE activity.</p><p>Continuing with the focus on compounds that retain dual NF-κB and ISRE activity similar to original hit compound 1, we selected site B modified compound 12, site C modified compound 33 and an aliphatic amine bearing compound 55 for dose response experiments and EC50 determination as these compounds are nearly equipotent in both the assays when evaluated at 5 μM concentration. As shown in Fig. 4, both compounds 12 and 33 showed relatively higher NF-κB activity at 5 μM concentration, but the activity of compound 12 decreased faster at lower concentrations which led to EC50 value of 1.85 μM. Compound 1 and 33 were almost equipotent with EC50 values of 0.60 μM and 0.69 μM, respectively. Compound 55 was relatively weaker with EC50 of 3.32 μM. The potency trends for these compounds remained the same in ISRE activity with compounds 1, 12 and 33 exhibiting EC50 of 0.66 μM, 1.4 μM, and 0.84 μM, respectively, and compound 55 with EC50 = 3.04 μM. Even though the activity of compound 55 was slightly attenuated, the amine handle can be utilized for derivatization to obtain affinity probes.</p><p>It was important to examine the adjuvanticity of the most potent compounds to verify if prolongation of immune stimulus by this chemotype leads to enhancement of in vivo antibody responses and if prolonged activation of the innate immune system could lead to systemic inflammation that may be harmful to the host.58, 59 In addition, it was also important to verify that the modifications on the scaffold that yielded potent compounds in vitro would retain the adjuvanticity in vivo as well. All the compounds administered to mice had low toxicity in the MTT assays. Since these vaccine co-adjuvants are designed to be administered locally (mostly intramuscularly) and show negligible toxicity (based on MTT data), we did not anticipate an excessive systemic inflammatory response. In the past, we utilized LPS as a widely recognized activator of the innate immune system and well characterized TLR-4 ligand to screen over 160,000 compounds for their ability to enhance APC activation.41, 42 However, to test these compounds for potency as co-adjuvants, we switched to the FDA approved TLR-4 adjuvant, MPLA for in vivo evaluation. Immunization experiments in mice (8 mice/group) were performed to evaluate the co-adjuvanticity of the lead compounds 1, 12 or 33 using ovalbumin (OVA) as a model antigen and MPLA as an adjuvant. Amine handle bearing compound 55 was not selected for immunization since it was designed for further derivatization as an intermediate to make probes as discussed below. Examination of OVA-specific IgG antibodies showed that co-immunization of MPLA with compounds 1 and 33 induced statistically significant increases in antigen-specific antibody titers when compared to mice immunized with MPLA alone (Fig. 5), without demonstrable systemic toxicity, as indicated by behavior change or weight loss. These results verified our approach that selected bis-aryl sulfonamide compounds that prolong immune stimulation could enhance the adjuvanticity of MPLA and that modified compounds that retained potency in vitro were equally potent in vivo as well.</p><p>Now that we had confirmed the in vitro and in vivo potency of selected active compounds, we were further interested to utilize the SAR studies for designing affinity probes. The activity data guided us to utilize site C for the introduction of an identifiable tag by derivatizing compound 55. Although compound 55 was less potent than compound 1, the changes in the hydrophobic interaction after amine derivatization may improve the potency. Compound 55 was derivatized to obtain fluorescein labeled compound 56, rhodamine labeled compound 57 and biotin labeled compound 58 (Scheme 5). In primary screens, the biotin labeled compound 58 was equipotent to compound 1 and thus could serve as the affinity probe (Fig. 6, Table 4). The rhodamine analog 57 showed reduced activity compared to compound 1 in both the NF-κB and ISRE assays likely due to the presence of a fixed charge on the molecule similar to the amine bearing compound 55. In contrast, the fluorescein analog 56 was completely inactive in both the assays (Fig. 6, Table 4).</p><p>Having validated specific site C modifications that tolerated the introduction of a trackable tag, we were interested to find a position where a photoreactive group such as aryl azide could be introduced to make photoaffinity probes. This prompted us to derivatize compounds 10 and 25, even though these were inactive but surmising that a change in the hydrogen bonding properties may have an opposite effect. The aromatic amine on position 4 at site A of compound 10 was converted to aryl azide using diazotization reaction to obtain compound 59 (Scheme 6). In parallel, the 3-bromo substitution at site B of compound 25 was reacted with sodium azide using copper catalyzed reaction. However, the major product of this reaction was aromatic amine analog 60, which was further converted to azide using the earlier described diazotization chemistry to obtain compound 61 (Scheme 6). The photoreactive aryl azide bearing compounds 59 and 61 and the aromatic amine analog 60 were then evaluated in the primary screens. While compound 61 was inactive just like its precursor bromo analog 25, the reversal of hydrogen bonding capacity in compound 60 led to resurgence of activity in both the assays possibly due to hydrophilic interaction with the aromatic amine (Fig. 6, Table 4). In contrast, the reversal of hydrogen bonding capacity of compound 10 led us to a potent aryl azide bearing analog 59 which was then utilized for making photoaffinity probes (Fig. 6, Table 4).</p><p>Using the methods utilized earlier, compound 59 was derivatized to obtain an alkyne analog 62, and a biotin analog 64 was obtained via an aliphatic amine derivative 63 (Scheme 6). Evaluation of these compounds in our primary screens showed that the alkyne probe 62 was very potent while the biotin probe 64 showed relatively weak activity in both the NF-κB and ISRE assays (Fig. 5, Table 4). Also, all the affinity probes had viability in the same range as the potent compounds in this series making them ideal candidates for future studies.</p><p>Our systematic SAR studies on bis-aryl sulfonamides that sustain NF-κB and ISRE activation have led to the identification of not only rhodamine labeled affinity fluorescent probe 57 and biotin-tagged affinity probe 58, but also alkyne and biotin labeled photoaffinity probes 62 and 64, respectively. These affinity probes will be utilized in concert for target identification and cell trafficking experiments.</p><!><p>Compound 1 was identified from HTS campaigns, that screened for agents capable of prolonging immune signaling, and was shown to be a potent co-adjuvant with MPLA in vivo. Here, we presented systematic SAR studies consisting of design, syntheses and evaluation of analogs of compound 1 to identify sites on the scaffold that can tolerate modification while still retaining dual NF-κB and ISRE enhancing activities in order to obtain affinity and photoaffinity probes. SAR studies pointed to key substitutions at site B and site C that retain potency in vitro and in vivo, while site A allowed the introduction of photoreactive aryl azide functionality. In addition, observed SAR trends at site C allowed the introduction of trackable tags such as rhodamine or biotin. This led to syntheses of several affinity probes which will be utilized to determine the mechanism of action and receptor target for this bis-aryl sulfonamide series of compounds that sustain NF-κB and ISRE activation.</p><!><p>Reagents were purchased as at least reagent grade from commercial vendors unless otherwise specified and used without further purification. Solvents were purchased from Fischer Scientific (Pittsburgh, PA) and were either used as purchased or redistilled with an appropriate drying agent. All the reagents 2a-g and 3g-o were purchased from commercially available vendors while reagents 3a-f were synthesized from commercially available reagents as shown in Supporting Information. Compounds used for structure-activity studies were synthesized according to methods described below and all the compounds were identified to be least 95% pure using HPLC.</p><!><p>Analytical TLC was performed using precoated TLC silica gel 60 F254 aluminum sheets purchased from EMD (Gibbstown, NJ) and visualized using UV light. Flash chromatography was carried out using with a Biotage Isolera One (Charlotte, NC) system using the specified solvent. Microwave reaction was performed using Biotage Initiator+ (Charlotte, NC). Reaction monitoring and purity analysis were done using an Agilent 1260 LC/6420 Triple Quad mass spectrometer (Santa Clara, CA) with Onyx Monolithic C18 (Phenomenex, Torrance, CA) column. Purity of all final compounds was above 95% (also see LC-MS spectra in Supporting Information for all final compounds). All final compounds were analyzed by high resolution MS (HRMS) using an Agilent 6230 ESI-TOFMS (Santa Clara, CA). 1H and 13C NMR spectra were obtained on a Varian 500 with XSens probe (Varian, Inc., Palo Alto, CA). The chemical shifts are expressed in parts per million (ppm) using suitable deuterated NMR solvents.</p><!><p>To a solution of a substituted phenyl sulfonyl chloride (reagent 3, 1 eq.) in anhydrous CH2Cl2 were added, triethylamine (2 eq.) and a solution of substituted aniline (reagent 2, 2 eq.) in CH2Cl2. The reaction mixture was stirred at room temperature overnight and then poured into water and acidified with 3N HCl followed by extraction with EtOAc. The EtOAc fraction was then dried over MgSO4, and solvent was removed under vacuum. The resultant residue was dissolved in MeOH and THF, followed by the addition of lithium hydroxide monohydrate (15 eq.) in water and stirred at room temperature until bis-sulfonamide side product is converted to the desired product. The solvent was then removed, dissolved in EtOAc, washed with water and brine, dried under vacuum to obtain the residue which was purified by column chromatography to obtain the final product.</p><!><p>Compound 1 was synthesized using 4-chloro-2,5-dimethoxyaniline (2a, 1.7 g, 5.3 mmol) and 4-ethoxybenzenesulfonyl chloride (3a, 1g, 4.5 mmol) after recrystallization in EtOH as pink crystals (1.2 g, yield = 71%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.66 (d, J = 8.80 Hz, 2H), 7.24 (s, 1H), 6.91 (s, 1H), 6.86 (d, J = 8.80 Hz, 2H), 6.77 (s, 1H), 4.04 (q, J = 6.93 Hz, 2H), 3.87 (s, 3H), 3.60 (s, 3H), 1.42 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.6, 149.2, 143.6, 130.0, 129.4, 125.2, 117.8, 114.4, 113.1, 106.3, 64.0, 56.8, 56.4, 14.6. HRMS for C16H17ClNO5S [M – H−] calculated 370.0521, found 370.0523.</p><!><p>Compound 4 was synthesized using 4-chloro-3-methoxyaniline (2b, 142.84 mg, 0.92 mmol) and 4-ethoxybenzenesulfonyl chloride (3a, 100 mg, 0.46 mmol) as off-white solid (83 mg, yield = 54%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.70 (d, J = 8.80 Hz, 2H), 7.17 (d, J = 8.56 Hz, 1H), 7.01 (br. s., 1H), 6.89 (d, J = 9.05 Hz, 2H), 6.80 (d, J = 2.20 Hz, 1H), 6.51 (dd, J = 2.20, 8.56 Hz, 1H), 4.05 (q, J = 7.09 Hz, 2H), 3.83 (s, 3H), 1.42 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.7, 155.3, 136.3, 130.4, 129.6, 129.4, 119.0, 114.6, 113.9, 105.8, 64.0, 56.2, 14.6. HRMS for C15H15ClNO4S [M − H]− calculated 340.0416, found 340.0416.</p><!><p>Compound 5 was synthesized using 4-chloro-2-methoxyaniline (2c, 142.84 mg, 0.92 mmol) and 4-ethoxybenzenesulfonyl chloride (3a, 100 mg, 0.46 mmol) as tan solid (100 mg, yield = 64%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.66 (d, J = 8.80 Hz, 2H), 7.45 (d, J = 8.56 Hz, 1H), 6.83 – 6.91 (m, 2H), 6.85 (d, J = 8.80 Hz, 2H), 6.72 (d, J = 1.96 Hz, 1H), 4.04 (q, J = 7.09 Hz, 2H), 3.65 (s, 3H), 1.41 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.5, 150.0, 130.4, 130.2, 129.4, 124.8, 121.9, 121.0, 114.3, 111.3, 63.9, 55.9, 14.6. HRMS for C15H16ClNO4SNa [M + Na+] calculated 364.0381, found 364.0382.</p><!><p>Compound 6 was synthesized using 2,5-dimethoxyaniline (2d, 138.8 mg, 0.92 mmol) and 4-ethoxybenzenesulfonyl chloride (3a, 100 mg, 0.46 mmol) as off-white solid (117 mg, yield = 76%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.70 (d, J = 8.80 Hz, 2H), 7.14 (d, J = 2.93 Hz, 1H), 7.01 (s, 1H), 6.85 (d, J = 8.80 Hz, 2H), 6.65 (d, J = 8.80 Hz, 1H), 6.53 (dd, J = 2.93, 9.05 Hz, 1H), 4.03 (q, J = 6.85 Hz, 2H), 3.75 (s, 3H), 3.62 (s, 3H), 1.40 (t, J = 7.09 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.4, 153.8, 143.4, 130.4, 129.4, 126.8, 114.3, 111.4, 109.5, 106.8, 63.9, 56.2, 55.8, 14.6. HRMS for C16H19NO5SNa [M + Na+] calculated 360.0876, found 360.0877.</p><!><p>Compound 7 was synthesized using 4-bromo-2,5-dimethoxyaniline (2e, 105.2 mg, 0.45 mmol) and 4-ethoxybenzenesulfonyl chloride (3a, 50 mg, 0.23 mmol) as purple solid (59 mg, yield = 62%). 1H NMR (500 MHz, HLOROFORM-d) δ 7.67 (d, J = 8.80 Hz, 2H), 7.21 (s, 1H), 6.93 (s, 1H), 6.92 (s, 1H), 6.85 (d, J = 9.05 Hz, 2H), 4.04 (q, J = 7.09 Hz, 2H), 3.86 (s, 3H), 3.61 (s, 3H), 1.41 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.6, 150.2, 143.7, 130.0, 129.4, 126.0, 115.8, 114.4, 106.2, 105.8, 64.0, 56.9, 56.4, 14.6. HRMS for C16H18BrNO5SNa [M + Na+] calculated 437.9981, found 437.9979.</p><!><p>Compound 8 was synthesized using 4-chloro-3,5-dimethoxyaniline (2f, 50 mg, 0.27 mmol) and 4-ethoxybenzenesulfonyl chloride (3a, 29 mg, 0.13 mmol) as white solid (30 mg, yield = 61%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.71 (d, J = 8.80 Hz, 2H), 6.89 (d, J = 9.05 Hz, 2H), 6.76 (br. s., 1H), 6.35 (s, 2H), 4.06 (q, J = 7.01 Hz, 2H), 3.80 (s, 6H), 1.43 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.8, 156.3, 136.1, 129.7, 129.5, 114.6, 107.2, 98.2, 64.0, 56.4, 14.6. HRMS for C16H17ClNO5S [M − H]− calculated 370.0521, found 370.0519.</p><!><p>Compound 9 was synthesized using 3,5-dimethoxy-4-nitroaniline (2g, 50 mg, 0.27 mmol) and 4-ethoxybenzenesulfonyl chloride (3a, 29 mg, 0.13 mmol) as yellow solid (30 mg, yield = 61%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.69 – 7.86 (m, J = 8.80 Hz, 2H), 7.44 (s, 1H), 7.40 (s, 1H), 7.31 (s, 1H), 6.90 – 6.95 (m, 2H), 4.07 (q, J = 6.93 Hz, 2H), 3.93 (s, 3H), 3.82 (s, 3H), 1.43 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 163.1, 149.4, 140.9, 133.1, 132.8, 129.5, 129.4, 114.8, 108.3, 103.2, 64.1, 57.0, 56.5, 14.5. HRMS for C16H19N2O7S [M + H+] calculated 383.0907, found 383.091.</p><!><p>To a solution of compound 9 (128 mg, 0.33 mmol) in EtOAc were added, a catalytic amount of palladium on carbon and some sodium sulfate. The reaction was subjected to Parr hydrogenation apparatus using hydrogen gas at 50 psi pressure for 6 hours. The solvent was then removed, and the residue was purified using silica gel column chromatography (5% MeOH/CH2Cl2) to obtain 84 mg of compound 10 as tan solid (yield = 72%). 1H NMR (500 MHz, METHANOL-d4) δ 7.55 (d, J = 8.80 Hz, 2H), 6.91 (d, J = 8.80 Hz, 2H), 6.87 (s, 1H), 6.26 (s, 1H), 4.06 (q, J = 7.01 Hz, 2H), 3.78 (s, 3H), 3.33 (s, 3H), 1.38 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, METHANOL-d4) δ 162.2, 147.8, 140.9, 135.9, 131.2, 129.2, 114.4, 113.5, 110.3, 99.0, 63.6, 55.2, 54.7, 13.5. HRMS for C16H20N2O5SNa [M + Na+] calculated 375.0985, found 375.0988.</p><!><p>Compound 11 was synthesized using 4-chloro-2,5-dimethoxyaniline (2a, 383 mg, 2.03 mmol) and 4-hydroxybenzenesulfonyl chloride (3b, 195 mg, 1.02 mmol) as dark brown solid (237 mg, yield = 68%). 1H NMR (500 MHz, DMSO-d6) δ 10.44 (s, 1H), 9.39 (s, 1H), 7.54 (d, J = 8.80 Hz, 2H), 7.02 (s, 1H), 6.97 (s, 1H), 6.83 (d, J = 8.80 Hz, 2H), 3.66 – 3.78 (m, 3H), 3.48 (s, 3H). 13C NMR (126 MHz, DMSO-d6) δ 161.2, 148.1, 146.0, 129.9, 129.2, 125.5, 117.3, 115.3, 113.9, 109.2, 56.5, 56.4. HRMS for C14H13ClNO5S [M − H]− calculated 342.0208, found 342.0205.</p><!><p>Compound 11 was synthesized using 4-chloro-2,5-dimethoxyaniline (2a, 450 mg, 2.40 mmol) and 4-methoxybenzenesulfonyl chloride (3c, 248 mg, 1.20 mmol) as light brown solid (189 mg, yield = 44%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.68 (d, J = 8.80 Hz, 2H), 7.24 (s, 1H), 6.92 (s, 1H), 6.88 (d, J = 9.05 Hz, 2H), 6.77 (s, 1H), 3.87 (s, 3H), 3.83 (s, 3H), 3.61 (s, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 163.1, 149.2, 143.6, 130.3, 129.4, 125.2, 117.9, 114.0, 113.1, 106.3, 56.8, 56.4, 55.6. HRMS for C15H16ClNO5SNa [M + Na+] calculated 380.033, found 380.0326.</p><!><p>Compound 13 was synthesized using 4-chloro-2,5-dimethoxyaniline (2a, 106 mg, 0.57 mmol) and 4-propoxybenzenesulfonyl chloride (3d, 66 mg, 0.28 mmol) as off-white solid (65 mg, yield = 59%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.66 (d, J = 8.80 Hz, 2H), 7.23 (s, 1H), 6.92 (s, 1H), 6.86 (d, J = 9.05 Hz, 2H), 6.76 (s, 1H), 3.92 (t, J = 6.60 Hz, 2H), 3.87 (s, 3H), 3.60 (s, 3H), 1.76 – 1.85 (m, 2H), 1.02 (t, J = 7.46 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.8, 149.2, 143.5, 130.0, 129.3, 125.2, 117.8, 114.4, 113.1, 106.2, 69.9, 56.8, 56.4, 22.3, 10.4. HRMS for C17H20ClNO5SNa [M + Na+] calculated 408.0643, found 408.0641.</p><!><p>Compound 14 was synthesized using 4-chloro-2,5-dimethoxyaniline (2a, 191 mg, 1.02 mmol) and 4-butoxybenzenesulfonyl chloride (3e, 127 mg, 0.51 mmol) as off-white solid (95 mg, yield = 47%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.66 (d, J = 8.80 Hz, 2H), 7.23 (s, 1H), 6.91 (s, 1H), 6.85 (d, J = 8.80 Hz, 2H), 6.76 (s, 1H), 3.96 (t, J = 6.48 Hz, 2H), 3.87 (s, 3H), 3.60 (s, 3H), 1.76 (quin, J = 7.20 Hz, 2H), 1.47 (sxt, J = 7.40 Hz, 2H), 0.97 (t, J = 7.46 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.8, 149.2, 143.5, 130.0, 129.3, 125.3, 117.8, 114.4, 113.1, 106.2, 68.1, 56.8, 56.4, 31.0, 19.1, 13.8. HRMS for C18H22ClNO5SNa [M + Na+] calculated 422.0799, found 422.0802.</p><!><p>Compound 15 was synthesized using 4-chloro-2,5-dimethoxyaniline (2a, 66 mg, 0.35 mmol) and 4-(prop-2-yn-1-yloxy)benzenesulfonyl chloride (3f, 40.7 mg, 0.18 mmol) as off-white solid (24 mg, yield = 32%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.69 (d, J = 9.05 Hz, 2H), 7.23 (s, 1H), 6.96 (d, J = 8.80 Hz, 2H), 6.90 (s, 1H), 6.76 (s, 1H), 4.72 (d, J = 2.20 Hz, 2H), 3.87 (s, 3H), 3.59 (s, 3H), 2.55 (t, J = 2.45 Hz, 1H). 13C NMR (126 MHz, DMSO-d6) δ 160.2, 148.1, 146.4, 132.5, 128.9, 125.1, 117.7, 114.9, 113.9, 109.9, 78.8, 78.5, 56.4, 55.8. HRMS for C17H15ClNO5S [M − H]− calculated 380.0365, found 380.0365.</p><!><p>Compound 11 was synthesized using 4-chloro-2,5-dimethoxyaniline (2a, 210 mg, 1.12 mmol) and 4-propylbenzenesulfonyl chloride (3g, 100 μL, 0.56 mmol) as white solid (121 mg, yield = 59%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.64 (d, J = 8.31 Hz, 2H), 7.23 (s, 1H), 7.21 (d, J = 8.31 Hz, 2H), 6.92 (s, 1H), 6.76 (s, 1H), 3.87 (s, 3H), 3.53 – 3.58 (m, 3H), 2.60 (t, J = 7.70 Hz, 2H), 1.61 (sxt, J = 7.60 Hz, 2H), 0.91 (t, J = 7.34 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 149.2, 148.6, 143.6, 136.0, 128.9, 127.2, 125.1, 117.9, 113.1, 106.4, 56.8, 56.3, 37.8, 24.1, 13.7. HRMS for C17H20ClNO4SNa [M + Na+] calculated 392.0694, found 392.0695.</p><!><p>Compound 17 was synthesized using 4-chloro-2,5-dimethoxyaniline (2a, 847 mg, 4.51 mmol) and 4-nitrobenzenesulfonyl chloride (3h, 500 mg, 2.25 mmol) as yellow solid (182 mg, yield = 22%). 1H NMR (500 MHz, DMSO-d6) δ 10.15 (s, 1H), 8.37 (d, J = 8.80 Hz, 2H), 7.93 (d, J = 8.80 Hz, 2H), 7.04 (s, 1H), 7.01 (s, 1H), 3.76 (s, 3H), 3.35 (s, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 150.2, 149.4, 144.5, 144.0, 128.4, 124.1, 123.5, 119.6, 113.2, 107.4, 56.9, 56.3. HRMS for C14H12ClN2O6S [M − H]− calculated 371.011, found 371.0104.</p><!><p>To a solution of compound 17 (150 mg, 0.4 mmol) in EtOAc were added, a catalytic amount of palladium on carbon and some sodium sulfate. The reaction was subjected to hydrogenation on Parr hydrogenation apparatus using hydrogen gas at 50 psi pressure for 6 hours. The solvent was then removed, and the residue was purified using silica gel column chromatography (9% MeOH/CH2Cl2) to obtain 79 mg of compound 18 as tan solid (yield = 58%). 1H NMR (500 MHz, DMSO-d6) δ 9.07 (s, 1H), 7.31 – 7.41 (m, J = 8.56 Hz, 2H), 6.97 (s, 1H), 7.01 (s, 1H), 6.47 – 6.56 (m, J = 8.80 Hz, 2H), 5.98 (s, 2H), 3.70 (s, 3H), 3.54 (s, 3H). 13C NMR (126 MHz, DMSO-d6) δ 152.9, 148.1, 145.5, 128.9, 126.1, 124.6, 116.4, 113.9, 112.4, 108.0, 56.6, 56.4. HRMS for C14H15ClN2O4SNa [M + Na+] calculated 365.0333, found 365.0335.</p><!><p>Compound 19 was synthesized using 4-chloro-2,5-dimethoxyaniline (2a, 372 mg, 1.98 mmol) and 4-cyanobenzenesulfonyl chloride (3i, 100 mg, 0.50 mmol) as white solid (40 mg, yield = 23%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.83 (d, J = 8.56 Hz, 2H), 7.73 (d, J = 8.31 Hz, 2H), 7.25 (s, 1H), 6.92 (s, 1H), 6.79 (s, 1H), 3.90 (s, 3H), 3.57 (s, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 149.4, 144.0, 142.9, 132.6, 127.8, 123.5, 119.5, 117.1, 116.7, 113.1, 107.4, 56.8, 56.2. HRMS for C15H12ClN2O4S [M − H]− calculated 351.0212, found 351.021.</p><!><p>To a crude solution of compound 19 (280 mg, 0.75 mmol) in methanol was added, a catalytic amount of palladium on carbon and di-tert-butyl dicarbonate (326 mg, 1.5 mmol). The reaction was subjected to hydrogenation on Parr hydrogenation apparatus using hydrogen gas at 50psi pressure overnight. The solvent was then removed, and the residue was purified using silica gel column chromatography (9% MeOH/CH2Cl2) to obtain 65 mg of compound 20 as white solid (yield = 20%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.70 (d, J = 8.07 Hz, 2H), 7.33 (d, J = 8.07 Hz, 2H), 7.25 (s, 1H), 6.93 (br. s., 1H), 6.76 (s, 1H), 4.94 (br. s., 1H), 4.34 (d, J = 5.87 Hz, 2H), 3.87 (s, 3H), 3.57 (s, 3H), 1.46 (s, 9H). 13C NMR (126 MHz, CHLOROFORM-d) δ 155.8, 149.3, 144.9, 143.7, 137.5, 127.5, 127.5, 124.8, 118.2, 113.1, 106.5, 80.0, 56.8, 56.3, 44.0, 28.3 HRMS for C20H25ClN2O6SNa [M + Na+] calculated 479.1014, found 479.1018.</p><!><p>Compound 20 (11 mg, mmol) was stirred in a solution of 4N HCl in dioxane for 1h. The solvent was then removed to obtain compound 21 in quantitative yield as hydrochloride salt (grey solid). 1H NMR (500 MHz, METHANOL-d4) δ 7.71 – 7.88 (m, J = 8.31 Hz, 2H), 7.51 – 7.67 (m, J = 8.31 Hz, 2H), 7.23 (s, 1H), 6.87 (s, 1H), 4.17 (s, 2H), 3.82 (s, 3H), 3.51 (s, 3H). 13C NMR (126 MHz, METHANOL-d4) δ 150.5, 147.2, 142.1, 139.5, 130.5, 129.3, 126.3, 120.3, 114.7, 110.4, 57.3, 57.0, 43.7. HRMS for C15H18ClN2O4S [M + H+] calculated 357.067, found 357.0674.</p><!><p>Compound 22 was synthesized using 4-chloro-2,5-dimethoxyaniline (2a, 140 mg, 0.74 mmol) and 4-phenoxybenzenesulfonyl chloride (3j, 100 mg, 0.37 mmol) as light yellow solid (51 mg, yield = 33%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.69 (d, J = 8.80 Hz, 1H), 7.41 (t, J = 7.95 Hz, 1H), 7.25 (s, 2H), 7.23 (t, J = 7.60 Hz, 2H), 7.03 (d, J = 7.58 Hz, 2H), 6.95 (s, 1H), 6.94 (d, J = 8.80 Hz, 2H), 6.80 (s, 1H), 3.87 (s, 3H), 3.64 (s, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 161.9, 154.8, 149.3, 143.6, 132.2, 130.2, 129.4, 125.1, 125.0, 120.3, 118.0, 117.2, 113.1, 106.3, 56.8, 56.4. HRMS for C20H18ClNO5SNa [M + Na+] calculated 442.0486, found 442.0489.</p><!><p>Compound 23 was synthesized using 4-chloro-2,5-dimethoxyaniline (2a, 182 mg, 0.97 mmol) and 3-methoxybenzenesulfonyl chloride (3k, 100 mg, 0.48 mmol) as brown solid (189 mg, yield = 54%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.29 – 7.35 (m, 2H), 7.21 – 7.26 (m, 2H), 7.00 – 7.09 (m, 1H), 6.94 (s, 1H), 6.78 (s, 1H), 3.87 (s, 3H), 3.77 (s, 3H), 3.58 (s, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 159.6, 149.3, 143.7, 139.8, 129.9, 124.9, 119.5, 119.3, 118.2, 113.1, 111.7, 106.5, 56.8, 56.4, 55.6. HRMS for C15H16ClNO5SNa [M + Na+] calculated 380.033, found 380.0329.</p><!><p>Compound 24 was synthesized using 4-chloro-2,5-dimethoxyaniline (2a, 182 mg, 0.97 mmol) and 2-methoxybenzenesulfonyl chloride (3l, 100 mg, 0.48 mmol) as tan solid (121 mg, yield = 70%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.87 (td, J = 1.70, 7.70 Hz, 1H), 7.58 (s, 1H), 7.49 (tt, J = 1.50, 7.70 Hz, 1H), 7.22 (s, 1H), 6.99 (t, J = 7.70 Hz, 1H), 6.95 (d, J = 8.31 Hz, 1H), 6.78 (s, 1H), 3.95 (s, 3H), 3.80 (s, 3H), 3.73 (s, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 156.4, 149.2, 142.9, 135.1, 130.9, 126.1, 125.7, 120.3, 116.8, 113.0, 111.8, 104.9, 56.7, 56.6, 56.1. HRMS for C15H16ClNO5SNa [M + Na+] calculated 380.033, found 380.0329.</p><!><p>Compound 25 was synthesized using 4-chloro-2,5-dimethoxyaniline (2a, 180 mg, 0.70 mmol) and 3-bromo-4-methoxybenzenesulfonyl chloride (3m, 100 mg, 0.35 mmol) as brown solid (68 mg, yield = 58%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.99 (d, J = 2.20 Hz, 1H), 7.64 (dd, J = 1.96, 8.56 Hz, 1H), 7.22 (s, 1H), 6.93 (s, 1H), 6.85 (d, J = 8.56 Hz, 1H), 6.79 (s, 1H), 3.93 (s, 3H), 3.89 (s, 3H), 3.65 (s, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 159.4, 149.3, 143.7, 132.4, 131.4, 128.5, 124.6, 118.4, 113.1, 112.0, 111.0, 106.6, 56.8, 56.6, 56.4. HRMS for C15H14BrClNO5S [M − H]− calculated 433.947, found 433.9469.</p><!><p>Compound 26 was synthesized using 4-chloro-2,5-dimethoxyaniline (2a, 80 mg, 0.42 mmol) and of methyl 4-(chlorosulfonyl)benzoate (3n, 50 mg, 0.21 mmol) as tan solid (31 mg, yield = 38%). 1H NMR (500 MHz, CHLOROFORM-d) δ 8.08 (d, J = 8.31 Hz, 2H), 7.80 (d, J = 8.31 Hz, 2H), 7.25 (s, 1H), 6.94 (s, 1H), 6.76 (s, 1H), 3.94 (s, 3H), 3.89 (s, 3H), 3.55 (s, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 165.5, 149.3, 143.8, 142.5, 134.1, 130.0, 127.2, 124.1, 118.9, 113.1, 107.0, 56.8, 56.2, 52.7. HRMS for C16H15ClNO6S [M − H]− calculated 384.0314, found 384.0308.</p><!><p>Compound 27 was synthesized using 4-chloro-2,5-dimethoxyaniline (2a, 171 mg, 0.77 mmol) and 4-carbamoylbenzenesulfonyl chloride (3o, 100 mg, 0.46 mmol) as white solid (14 mg, yield = 8%). 1H NMR (500 MHz, DMSO-d6) δ 9.83 (br. s., 1H), 8.13 (br. s., 1H), 7.96 (d, J = 8.31 Hz, 2H), 7.76 (d, J = 8.31 Hz, 2H), 7.60 (br. s., 1H), 7.02 (s, 1H), 6.99 (s, 1H), 3.74 (s, 3H), 3.38 (br. s., 3H). 13C NMR (126 MHz, DMSO-d6) δ 166.7, 148.2, 146.7, 142.4, 138.0, 128.0, 126.7, 124.5, 118.3, 114.0, 110.7, 56.5, 56.3. HRMS for C15H14ClN2O5S [M − H]− calculated 369.0317, found 369.0315.</p><!><p>To a solution of compound 26 (25.0 mg, 0.06 mmol) in MeOH and THF was added a solution of LiOH (40.1 mg, 0.97 mmol) in water. The reaction was stirred overnight and the solvent was then removed under vacuum. The residue was mixed with acidified water (3N aq. HCl) extracted with EtOAc, dried over MgSO4 and concentrated to dryness under vacuum to obtain compounds 28 as white solid (21 mg, yield = 87%). 1H NMR (500 MHz, DMSO-d6) δ 10.44 (s, 1H), 9.39 (s, 1H), 7.54 (d, J = 8.80 Hz, 2H), 7.02 (s, 1H), 6.97 (s, 1H), 6.83 (d, J = 8.80 Hz, 2H), 3.66 – 3.78 (m, 3H), 3.48 (s, 3H). 13C NMR (126 MHz, DMSO-d6) δ 166.4, 148.3, 147.0, 143.9, 134.5, 129.9, 127.1, 124.4, 118.7, 114.0, 111.1, 56.6, 56.3. HRMS for C15H13ClNO6S [M − H]− calculated 370.0158, found 370.0152.</p><!><p>To a solution of compound 28 (20 mg, 0.05 mmol) in anhydrous EtOH was added trimethylsilyl chloride (68.3 μL, 0.54 mmol) and the reaction was stirred at room temperature until completion. The reaction mixture was poured into water and extracted with EtOAc. The organic layer was dried over MgSO4 and concentrated under vacuum to obtain the residue which was purified by silica gel column chromatography (25% EtOAc/hexanes) to obtain compound 29 as white solid. (17.6 mg, yield = 82%). 1H NMR (500 MHz, CHLOROFORM-d) δ 8.09 (d, J = 8.31 Hz, 2H), 7.80 (d, J = 8.31 Hz, 2H), 7.26 (s, 1H), 6.95 (s, 1H), 6.76 (s, 1H), 4.39 (q, J = 7.17 Hz, 2H), 3.89 (s, 3H), 3.56 (s, 3H), 1.40 (t, J = 7.09 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 165.0, 149.3, 143.8, 142.4, 134.5, 130.0, 127.2, 124.2, 118.8, 113.1, 106.9, 61.8, 56.8, 56.3, 14.2. HRMS for C17H17ClNO6S [M − H]− calculated 398.0471, found 398.0467.</p><!><p>To a solution of compound 23 (25 mg, 0.07 mmol) in anhydrous DMF were added, 2M methyl amine solution in THF (67 μL, 0.13 mmol), triethylamine (38 μL, 0.27 mmol) and HATU (31 mg, 0.08 mmol). The reaction was stirred at room temperature until completion. The reaction mixture was poured into water and extracted with EtOAc. The organic layer was dried over MgSO4 and concentrated under vacuum to obtain the residue which was purified by reverse-phase C18 column chromatography (56% MeOH/H2O with 0.1% CF3CO2H) to yield compound 30 as white solid (3 mg, yield = 12%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.79 (s, 4H), 7.26 (s, 1H), 6.94 (s, 1H), 6.76 (s, 1H), 6.14 (s, 1H), 3.89 (s, 3H), 3.56 (s, 3H), 3.03 (d, J = 4.89 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 166.4, 149.3, 143.8, 141.2, 138.8, 127.5, 127.4, 124.2, 118.8, 113.1, 106.9, 56.8, 56.3, 27.0. HRMS for C16H16ClN2O5S [M − H]− calculated 383.0474, found 383.0468.</p><!><p>Compound 31 was synthesized using the general procedure A using p-phenetidine (2h, 166.04 μL, 1.21 mmol) and 4-Cl-2,5-dimethoxybenzenesulfonyl chloride (3p, 100 mg, 0.61 mmol) as brown solid (98 mg, yield = 43.6%). 1H NMR (500 MHz, DMSO-d6) δ 9.74 (br. s., 1H), 7.35 (s, 1H), 7.28 (s, 1H), 6.97 (d, J = 8.80 Hz, 2H), 6.75 (d, J = 9.05 Hz, 2H), 3.88 (q, J = 7.10 Hz, 2H), 3.86 (s, 3H), 3.77 (s, 3H), 1.24 (t, J = 6.85 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 157.5, 149.7, 149.0, 128.4, 128.4, 125.4, 125.1, 114.9, 114.9, 113.7, 63.6, 57.3, 56.8, 14.8. HRMS for C16H18ClNO5SNa [M + Na+] calculated 394.0486, found 394.0489.</p><!><p>To a solution of compound 1 (1 eq.) in anhydrous DMF were added, potassium carbonate (2 eq.), and reagent 32 (1.1 eq.). The reaction was then heated at 45°C with stirring until completion. The suspension was extracted with EtOAc and brine. Then the organic layer was isolated, dried over MgSO4 and concentrated in vacuo. The crude material was purified by chromatography to obtain the final compounds.</p><!><p>Compound 33 was synthesized using compound 1 (10 mg, 0.03 mmol) and iodomethane (32a, 1.85 μL, 0.03 mmol) as white solid (10 mg, yield = 95%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.61 (d, J = 8.80 Hz, 2H), 6.96 (s, 1H), 6.92 (d, J = 8.80 Hz, 2H), 6.83 (s, 1H), 4.09 (d, J = 6.85 Hz, 2H), 3.85 (s, 3H), 3.39 (s, 3H), 3.19 (s, 3H), 1.45 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, dichloroethane) δ 162.1, 150.4, 148.6, 130.6, 129.7, 128.0, 122.5, 116.1, 114.0, 113.8, 63.9, 56.7, 55.6, 37.8, 14.6. HRMS for C17H20ClNO5SNa [M + Na+] calculated 408.0643, found 408.0641.</p><!><p>Compound 34 was synthesized using compound 1 (10 mg, 0.03 mmol) and 1-iodopropane (32b, 2.9 μL, 0.03 mmol) as tan solid (11 mg, yield = 96%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.59 (d, J = 8.80 Hz, 2H), 6.86 – 6.92 (m, 3H), 6.81 (s, 1H), 4.07 (q, J = 7.09 Hz, 2H), 3.84 (s, 3H), 3.51 (br. s., 2H), 3.36 (s, 3H), 1.44 (sxt, J = 7.30 Hz, 5H), 0.89 (t, J = 7.34 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.0, 150.7, 148.6, 131.7, 129.6, 125.7, 122.6, 117.4, 113.9, 113.7, 63.9, 56.8, 55.6, 51.4, 22.2, 14.6, 11.2. HRMS for C19H24ClNO5SNa [M + Na+] calculated 436.0956, found 436.0954.</p><!><p>Compound 35 was synthesized using compound 1 (10 mg, 0.03 mmol) and 1-iodobutane (32c, 2.9 μL, 0.03 mmol) as white solid (11 mg, yield = 96%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.60 (d, J = 8.80 Hz, 2H), 6.87 – 6.93 (m, 3H), 6.82 (s, 1H), 4.08 (q, J = 6.93 Hz, 2H), 3.85 (s, 3H), 3.48 – 3.61 (m, 2H), 3.37 (s, 3H), 1.45 (t, J = 6.85 Hz, 3H), 1.39 (dd, J = 7.46, 14.79 Hz, 2H), 1.29 – 1.35 (m, 2H), 0.87 (t, J = 7.09 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.0, 150.8, 148.6, 131.7, 129.6, 125.6, 122.6, 117.4, 113.9, 113.7, 63.9, 56.8, 55.6, 49.4, 31.0, 19.8, 14.6, 13.7. HRMS for C20H26ClNO5SNa [M + Na+] calculated 450.1112, found 450.1106.</p><!><p>Compound 36 was synthesized using compound 1 (10 mg, 0.03 mmol) and 1-iodopentane (32d, 3.9 μL, 0.03 mmol) as off-white solid (12 mg, yield = 98%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.59 (d, J = 8.80 Hz, 2H), 6.87 – 6.92 (m, 3H), 6.82 (s, 1H), 4.08 (q, J = 7.09 Hz, 2H), 3.85 (s, 3H), 3.54 (br. s., 2H), 3.36 (s, 3H), 1.45 (t, J = 6.97 Hz, 3H), 1.37 – 1.42 (m, 2H), 1.26 – 1.30 (m, J = 3.70 Hz, 4H), 0.85 (t, J = 6.85 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.0, 150.8, 148.6, 131.7, 129.6, 125.6, 122.6, 117.4, 113.9, 113.7, 63.9, 56.8, 55.5, 49.7, 28.7, 28.5, 22.3, 14.6, 14.0. HRMS for C21H28ClNO5SNa [M + Na+] calculated 464.1269, found 464.1265.</p><!><p>Compound 37 was synthesized using compound 1 (10 mg, 0.03 mmol) and 1-iodohexane (32e, 4.4 μL, 0.03 mmol) as off-white solid (8 mg, yield = 65%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.59 (d, J = 8.80 Hz, 2H), 6.86 – 6.91 (m, 3H), 6.81 (s, 1H), 4.07 (q, J = 7.09 Hz, 2H), 3.84 (s, 3H), 3.47 – 3.62 (m, 2H), 3.36 (s, 3H), 1.44 (t, J = 7.10 Hz, 3H), 1.39 (quin, J = 7.60 Hz, 2H), 1.16 – 1.34 (m, 6H), 0.85 (t, J = 6.85 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.0, 150.8, 148.6, 131.7, 129.6, 125.6, 122.6, 117.4, 113.9, 113.7, 63.9, 56.8, 55.5, 49.7, 31.4, 28.8, 26.2, 22.6, 14.6, 14.0. HRMS for C22H30ClNO5SNa [M + Na+] calculated 478.1425, found 478.1422.</p><!><p>Compound 38 was synthesized using compound 1 (10 mg, 0.03 mmol) and 1-iodoheptane (32f, 4.9 μL, 0.03 mmol) as white solid (12 mg, yield = 95%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.59 (d, J = 8.80 Hz, 2H), 6.87 – 6.91 (m, 3H), 6.81 (s, 1H), 4.07 (q, J = 6.85 Hz, 2H), 3.84 (s, 3H), 3.45 – 3.63 (m, 2H), 3.36 (s, 3H), 1.44 (t, J = 7.10 Hz, 6H), 1.39 (quin, J = 7.40 Hz, 1H), 1.14 – 1.33 (m, 10H), 0.85 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 161.9, 150.7, 148.5, 131.6, 129.5, 125.6, 122.6, 117.3, 113.9, 113.6, 63.9, 56.7, 55.5, 49.6, 31.7, 28.9, 28.9, 26.5, 22.6, 14.6, 14.1. HRMS for C23H32ClNO5SNa [M + Na+] calculated 492.1582, found 492.1578.</p><!><p>Compound 39 was synthesized using compound 1 (10 mg, 0.03 mmol) and 1-bromododecane (32g, 7.1 μL, 0.03 mmol) as white solid (14 mg, yield = 96%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.59 (d, J = 8.80 Hz, 2H), 6.86 – 6.92 (m, 3H), 6.81 (s, 1H), 4.07 (q, J = 7.09 Hz, 2H), 3.84 (s, 3H), 3.45 – 3.63 (m, 2H), 3.36 (s, 3H), 1.44 (t, J = 6.97 Hz, 3H), 1.38 (quin, J = 7.30 Hz, 2H), 1.23 – 1.30 (m, 10H), 0.88 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 161.9, 150.7, 148.5, 131.6, 129.6, 125.6, 122.6, 117.3, 113.9, 113.6, 63.9, 56.7, 55.5, 49.7, 31.9, 29.7, 29.6, 29.6, 29.6, 29.4, 29.2, 28.9, 26.6, 22.7, 14.6, 14.1. HRMS for C28H43ClNO5S [M + H+] calculated 540.2545, found 540.2549.</p><!><p>Compound 40 was synthesized using compound 1 (10 mg, 0.03 mmol) and 2-iodopropane (32h, 2.96 μL, 0.03 mmol) as white solid (5 mg, yield = 48%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.76 (d, J = 8.80 Hz, 2H), 6.94 (s, 1H), 6.92 (d, J = 8.80 Hz, 2H), 6.70 (s, 1H), 4.39 (spt, J = 6.70 Hz, 1H), 4.09 (q, J = 7.09 Hz, 2H), 3.81 (s, 3H), 3.61 (s, 3H), 1.46 (t, J = 6.97 Hz, 3H), 1.13 (d, J = 6.60 Hz, 3H), 0.99 (d, J = 6.60 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 161.9, 152.9, 148.3, 132.9, 129.8, 123.3, 122.6, 118.5, 114.0, 113.9, 63.9, 56.8, 55.8, 51.9, 22.2, 20.9, 14.6. HRMS for C19H24ClNO5SNa [M + Na+] calculated 436.0956, found 436.0957.</p><!><p>Compound 41 was synthesized using compound 1 (10 mg, 0.03 mmol) and 1-iodo-2-methylpropane (32i, 3.4 μL, 0.03 mmol) as white solid (9 mg, yield = 77%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.56 (d, J = 8.80 Hz, 2H), 6.92 (s, 1H), 6.89 (d, J = 8.80 Hz, 2H), 6.80 (s, 1H), 4.07 (q, J = 7.09 Hz, 2H), 3.85 (s, 3H), 3.29 – 3.51 (m, 2H), 3.33 (s, 3H), 1.59 (spt, J = 7.00 Hz, 2H), 1.44 (t, J = 6.97 Hz, 3H), 0.91 (br. s., 6H). 13C NMR (126 MHz, CHLOROFORM-d) δ 161.9, 150.6, 148.5, 131.5, 129.6, 126.0, 122.5, 117.2, 113.9, 113.7, 63.9, 57.1, 56.8, 55.5, 27.6, 20.1, 14.6. HRMS for C20H26ClNO5SNa [M + Na+] calculated 450.1112, found 450.1109.</p><!><p>Compound 42 was synthesized using compound 1 (10 mg, 0.03 mmol) and propargyl bromide solution in toluene (32j, 2.8 μL, 0.03 mmol) as white solid (9 mg, yield = 81%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.62 (d, J = 8.80 Hz, 2H), 6.96 (s, 1H), 6.90 (d, J = 8.80 Hz, 2H), 6.85 (s, 1H), 4.44 (br. s., 2H), 4.08 (q, J = 7.09 Hz, 2H), 3.82 (s, 3H), 3.44 (s, 3H), 2.12 – 2.23 (m, 1H), 1.45 (t, J = 6.85 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.3, 150.5, 148.6, 131.1, 129.8, 125.0, 123.1, 117.2, 114.1, 113.7, 78.4, 73.2, 64.0, 56.7, 55.8, 39.6, 14.6. HRMS for C19H20ClNO5SNa [M + Na+] calculated 432.0643, found 432.0639.</p><!><p>Compound 43 was synthesized using compound 1 (40 mg, 0.11 mmol) and 4-bromo-1-butyne (32k, 11.1 μL, 0.11 mmol) as white solid (3 mg, yield = 7%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.59 (d, J = 8.80 Hz, 2H), 6.98 (s, 1H), 6.90 (d, J = 8.80 Hz, 2H), 6.81 (s, 1H), 4.08 (q, J = 7.09 Hz, 2H), 3.85 (s, 3H), 3.72 (br. s., 2H), 3.35 (s, 3H), 2.42 (dt, J = 2.57, 7.40 Hz, 2H), 1.95 (t, J = 2.57 Hz, 1H), 1.45 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.1, 150.3, 148.6, 131.4, 129.6, 125.2, 123.0, 117.6, 114.0, 113.6, 81.0, 70.0, 63.9, 56.7, 55.5, 48.6, 19.7, 14.6. HRMS for C20H22ClNO5SNa [M + Na+] calculated 446.0799, found 446.0798.</p><!><p>Compound 44 was synthesized using compound 1 (40 mg, 0.11 mmol) and 5-iodopent-1-yne (32l, 13.5 μL, 0.12 mmol) as off-white solid (40 mg, yield = 84%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.59 (d, J = 8.80 Hz, 2H), 6.90 (d, J = 8.80 Hz, 2H), 6.87 (s, 1H), 6.83 (s, 1H), 4.08 (q, J = 7.09 Hz, 2H), 3.84 (s, 3H), 3.64 (br. s., 2H), 3.38 (s, 3H), 2.26 (dt, J = 2.45, 7.21 Hz, 2H), 1.91 (t, J = 2.57 Hz, 1H), 1.67 (quin, J = 7.09 Hz, 2H), 1.44 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.1, 150.7, 148.6, 131.3, 129.6, 125.5, 122.8, 117.0, 114.0, 113.7, 83.4, 68.7, 63.9, 56.8, 55.6, 48.9, 27.8, 15.8, 14.6. HRMS for C21H24ClNO5SNa [M + Na+] calculated 460.0956, found 460.0954.</p><!><p>Compound 45 was synthesized using compound 1 (40 mg, 0.11 mmol) and propargyl-PEG3-bromide (32m, 29.7 μL, 0.12 mmol) as clear oil (48 mg, yield = 82%). 1H NMR (500 MHz, METHANOL-d4) δ 7.59 (d, J = 8.80 Hz, 2H), 7.02 (d, J = 8.80 Hz, 2H), 6.97 (s, 1H), 6.96 (s, 1H), 4.16 (d, J = 2.20 Hz, 2H), 4.11 (q, J = 7.09 Hz, 2H), 3.72 – 3.85 (m, 5H), 3.61 – 3.65 (m, 2H), 3.57 – 3.60 (m, 2H), 3.48 – 3.54 (m, 6H), 3.38 (s, 3H), 2.85 (t, J = 2.32 Hz, 1H), 1.41 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, METHANOL-d4) δ 164.0, 152.4, 150.1, 132.9, 131.0, 127.1, 124.3, 119.1, 115.5, 115.0, 80.7, 76.1, 71.7, 71.5, 71.3, 70.5, 70.2, 65.3, 59.2, 57.4, 56.4, 50.4, 15.1. HRMS for C25H32ClNO8SNa [M + Na+] calculated 564.1429, found 564.1431.</p><!><p>Compound 46 was synthesized using compound 1 (10 mg, 0.03 mmol) and benzyl chloride (32n, 3.4 μL, 0.03 mmol) as white solid (12 mg, yield = 94%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.65 (d, J = 8.80 Hz, 2H), 7.10 – 7.25 (m, 5H), 6.92 (d, J = 8.80 Hz, 2H), 6.75 (s, 1H), 6.63 (s, 1H), 4.74 (br. s., 2H), 4.09 (q, J = 6.85 Hz, 2H), 3.67 (s, 3H), 3.35 (s, 3H), 1.46 (t, J = 6.85 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.1, 150.5, 148.4, 136.5, 131.7, 129.6, 128.8, 128.2, 127.6, 125.1, 122.6, 117.9, 114.0, 113.5, 63.9, 56.6, 55.5, 53.4, 14.6. HRMS for C23H24ClNO5SNa [M + Na+] calculated 484.0956, found 484.0952.</p><!><p>Compound 47 was synthesized using compound 1 (10 mg, 0.03 mmol) and (2-iodoethyl)benzene (32o, 3.4 μL, 0.03 mmol) as off-white solid (12 mg, yield = 95%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.45 – 7.52 (m, 2H), 7.18 (d, J = 7.60 Hz, 2H), 7.12 (t, J = 7.30 Hz, 1H), 7.06 (d, J = 7.60 Hz, 2H), 6.76 – 6.83 (m, J = 8.80 Hz, 2H), 6.72 (s, 1H), 6.66 (s, 1H), 3.99 (q, J = 6.85 Hz, 2H), 3.61 – 3.85 (m, 5H), 3.25 (s, 3H), 2.74 (t, J = 7.70 Hz, 2H), 1.37 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.0, 150.4, 148.5, 138.4, 131.4, 129.5, 128.9, 128.3, 126.4, 125.6, 122.6, 117.5, 113.9, 113.4, 63.9, 56.6, 55.5, 51.2, 35.8, 14.6. HRMS for C24H26ClNO5SNa [M + Na+] calculated 498.1112, found 498.111.</p><!><p>To a solution of compound 1 (10 mg, 0.03 mmol) in anhydrous CH2Cl2 were added, acetyl chloride (32p, 3.8 μL, 0.03 mmol) and triethylamine (15 μL, 0.12 mmol) and the reaction was heated at 45 °C with stirring for 20h upon which more acetyl chloride (3.8 μL, 0.03 mmol) was added to drive the reaction to completion. The solvent was then removed, and the residue was dissolved in EtOAc, washed with water and brine, dried over MgSO4 and concentrated under vacuum to obtain the residue which was purified by silica gel column chromatography (30% EtOAc/hexanes) to obtain compound 48 as white solid (5 mg, yield = 45%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.98 (d, J = 8.80 Hz, 2H), 7.05 (s, 1H), 6.85 – 7.01 (m, 3H), 4.12 (q, J = 7.10 Hz, 2H), 3.91 (s, 3H), 3.67 – 3.76 (m, 3H), 1.86 (s, 3H), 1.46 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 170.1, 163.2, 149.8, 149.3, 131.8, 130.0, 124.9, 123.9, 115.7, 114.1, 113.8, 64.0, 56.9, 56.0, 24.0, 14.6. HRMS for C18H20ClNO6SNa [M + Na+] calculated 436.0592, found 436.0592.</p><!><p>Compound 49 was synthesized using compound 1 (10 mg, 0.03 mmol) and ethylbromoacetate (32q, 3.3 μL, 0.03 mmol) as off-white solid (12 mg, yield = 97%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.60 (d, J = 8.80 Hz, 2H), 7.18 (s, 1H), 6.89 (d, J = 8.80 Hz, 2H), 6.80 (s, 1H), 4.38 (s, 2H), 4.16 (d, J = 7.09 Hz, 2H), 4.07 (d, J = 7.09 Hz, 2H), 3.83 (s, 3H), 3.39 (s, 3H), 1.44 (t, J = 6.97 Hz, 3H), 1.25 (t, J = 7.09 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 169.4, 162.3, 149.8, 148.6, 131.3, 129.7, 125.6, 123.0, 117.8, 114.0, 113.5, 63.9, 61.3, 56.7, 55.7, 51.0, 14.6, 14.2. HRMS for C20H25ClNO7S [M + H+] calculated 458.1035, found 458.1035.</p><!><p>Compound 50 was synthesized using compound 1 (40 mg, 0.11 mmol) and ethyl-4-bromobutyrate (32r, 16.9 μL, 0.12 mmol) as white solid (47 mg, yield = 91%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.58 (d, J = 8.80 Hz, 2H), 6.89 (d, J = 8.80 Hz, 2H), 6.87 (s, 1H), 6.82 (s, 1H), 4.09 (q, J = 7.20 Hz, 2H), 4.07 (q, J = 7.00 Hz, 2H), 3.84 (s, 3H), 3.60 (br. s., 2H), 3.37 (s, 3H), 2.41 (t, J = 7.46 Hz, 2H), 1.74 (quin, J = 7.09 Hz, 2H), 1.44 (t, J = 6.97 Hz, 3H), 1.23 (t, J = 7.21 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 173.1, 162.1, 150.7, 148.6, 131.3, 129.6, 125.3, 122.9, 117.0, 114.0, 113.7, 63.9, 60.4, 56.8, 55.6, 49.0, 31.1, 24.1, 14.6, 14.2. HRMS for C22H28ClNO7SNa [M + Na+] calculated 508.1167, found 508.117.</p><!><p>Compound 51 was synthesized using compound 1 (10 mg, 0.03 mmol) and 3-bromo-1-propanol (32s, 2.7 μL, 0.03 mmol) as off-white solid (5 mg, yield = 42%). 1H NMR (500 MHz, DMSO-d6) δ 7.55 (d, J = 8.80 Hz, 2H), 7.14 (s, 1H), 7.08 (d, J = 8.80 Hz, 2H), 6.76 (s, 1H), 4.43 (t, J = 4.89 Hz, 1H), 4.11 (q, J = 6.85 Hz, 2H), 3.72 (s, 3H), 3.50 (br. s., 2H), 3.42 (s, 3H), 3.33 – 3.36 (m, 2H), 1.46 (quin, J = 6.80 Hz, 2H), 1.34 (t, J = 6.85 Hz, 3H). 13C NMR (126 MHz, DMSO-d6) δ 161.8, 151.1, 147.9, 130.8, 129.5, 125.9, 121.5, 116.4, 114.5, 114.3, 63.8, 58.1, 56.5, 56.1, 47.0, 31.7, 14.5. HRMS for C19H24ClNO6SNa [M + Na+] calculated 452.0905, found 452.0907.</p><!><p>Compound 52 was synthesized using compound 1 (50 mg, 0.13 mmol) and tert-butyl (3-bromopropyl)carbamate (32t, 35.2 μL, 0.15 mmol) as off-white solid (60 mg, yield = 84%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.59 (d, J = 8.80 Hz, 2H), 6.91 (d, J = 8.80 Hz, 2H), 6.84 (s, 1H), 6.84 (s, 1H), 5.00 (br. s., 1H), 4.08 (q, J = 6.85 Hz, 2H), 3.84 (s, 3H), 3.61 (br. s., 2H), 3.38 (s, 3H), 3.27 (br. s., 2H), 1.56 (quin, J = 6.40 Hz, 2H), 1.40 – 1.50 (m, 12H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.1, 156.0, 150.7, 148.7, 131.2, 129.6, 125.2, 123.0, 117.0, 114.0, 113.8, 79.1, 63.9, 56.8, 55.6, 47.0, 37.1, 28.8, 28.4, 14.6. HRMS for C24H33ClN2O7SNa [M + Na+] calculated 551.1589, found 551.1587.</p><!><p>To a solution of compound 50 (38 mg, 0.08 mmol) in MeOH (0.5 mL) was added a solution of lithium hydroxide monohydrate (16.4 mg, 0.39 mmol) in water (0.5 mL) and the reaction was stirred overnight. The solvent was then removed and the residue was dissolved in acidified water (3N aq. HCl) and extracted with EtOAc. The organic layer was dried over MgSO4 and concentrated under vacuum to obtain the residue which was purified by silica gel column chromatography (8% MeOH/CH2Cl2) to obtain compounds 53 as white solid (26 mg, yield = 73%). 1H NMR (500 MHz, DMSO-d6) δ 11.74 – 12.24 (m, 1H), 7.54 (d, J = 9.05 Hz, 2H), 7.14 (s, 1H), 7.07 (d, J = 9.05 Hz, 2H), 6.75 (s, 1H), 4.11 (q, J = 7.09 Hz, 2H), 3.70 (s, 3H), 3.45 – 3.47 (m, 2H), 3.41 (s, 3H), 2.26 (t, J = 7.34 Hz, 2H), 1.50 (quin, J = 7.03 Hz, 2H), 1.34 (t, J = 6.85 Hz, 3H). 13C NMR (126 MHz, DMSO-d6) δ 174.1, 161.8, 151.1, 148.0, 130.7, 129.5, 125.8, 121.6, 116.2, 114.5, 114.3, 63.8, 56.5, 56.1, 48.9, 30.4, 23.5, 14.5. HRMS for C20H23ClNO7S [M − H]− calculated 456.0889, found 456.0889.</p><!><p>To a solution of compound 53 (7 mg, 0.02 mmol) in anhydrous DMF were added, HATU (6.4 mg, 0.02), 2M ethyl amine solution in THF (8.36 μL, 0.02 mmol), and triethylamine (4.3 μL, 0.03 mmol) and the reaction was stirred until completion. The solvent was then removed under vacuum and the residue was purified by silica gel column chromatography (5% MeOH/CH2Cl2) to obtain compound 54 as white solid (6 mg, yield = 81%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.56 (d, J = 8.80 Hz, 2H), 6.91 (d, J = 8.80 Hz, 2H), 6.85 (s, 1H), 6.79 (s, 1H), 5.90 (br. s., 1H), 4.08 (q, J = 7.09 Hz, 2H), 3.83 (s, 3H), 3.58 (br. s., 2H), 3.40 (s, 3H), 3.32 (quin, J = 6.85 Hz, 2H), 2.33 (t, J = 6.85 Hz, 2H), 1.73 (quin, J = 6.48 Hz, 2H), 1.45 (t, J = 6.97 Hz, 3H), 1.18 (t, J = 7.34 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 172.3, 162.2, 150.7, 148.7, 131.1, 129.6, 125.2, 123.1, 116.8, 114.1, 114.0, 63.9, 56.8, 55.7, 48.9, 34.4, 33.3, 24.6, 14.8, 14.6. HRMS for C22H30ClN2O6S [M + H+] calculated 485.1508, found 485.1511.</p><!><p>Compound 52 (48 mg, 0.09 mmol) was dissolved in 4N HCl solution in dioxane (1 mL) and stirred at room temperature for 20h. Solvent was then removed under vacuum, reside was suspended in CH2Cl2 followed by removal of the solvent under vacuum to obtain compound 55 as white solid (37.1 mg, yield = 95%). 1H NMR (500 MHz, DMSO-d6) δ 7.85 (br. s., 3H), 7.56 (d, J = 8.80 Hz, 2H), 7.17 (s, 1H), 7.09 (d, J = 8.80 Hz, 2H), 6.78 (s, 1H), 4.11 (q, J = 6.85 Hz, 2H), 3.71 – 3.75 (m, 3H), 3.53 (br. s., 2H), 3.42 (s, 3H), 2.83 (t, J = 8.10 Hz, 2H), 1.61 (td, J = 7.00, 14.61 Hz, 2H), 1.34 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, DMSO-d6) δ 161.9, 151.0, 148.0, 130.5, 129.5, 125.5, 121.8, 116.3, 114.6, 114.4, 63.8, 56.6, 56.2, 47.1, 36.6, 26.4, 14.5. HRMS for C19H26ClN2O5S [M + H+] calculated 429.1245, found 429.1246.</p><!><p>To a solution of compound 55 (4 mg, 0.01 mmol) in anhydrous DMF were added, triethylamine (3.6 μL, 0.03 mmol) and fluorescein-5(6)-carboxamidocaproic acid N-succinimidyl ester (5(6)-SFX SE, Chemodex, #F0044) (5 mg, 0.01 mmol). After completion of the reaction, solvent was removed, and the residue was purified using silica gel column chromatography (10% MeOH/CH2Cl2) to obtain compound 56 as bright yellow solid (5.4 mg, yield = 70%). 1H NMR (500 MHz, METHANOL-d4) δ 8.12 (d, J = 8.07 Hz, 1H), 8.06 (d, J = 8.10 Hz, 1H), 7.60 (s, 1H), 7.55 (d, J = 8.80 Hz, 2H), 6.99 – 7.03 (m, 2H), 6.97 (s, 1H), 6.87 (s, 1H), 6.65 – 6.73 (m, 2H), 6.60 (br. s., 2H), 6.54 (s, 2H), 4.10 (q, J = 7.09 Hz, 2H), 3.78 (s, 3H), 3.54 – 3.64 (m, 2H), 3.35 – 3.37 (m, 3H), 3.12 – 3.28 (m, 4H), 2.12 (s, 2H), 1.48 – 1.69 (m, 8H), 1.40 (t, J = 6.97 Hz, 3H). HRMS for C46H47ClN3O12S [M + H+] calculated 900.2563, found 900.2563.</p><!><p>To a solution of compound 55 (5 mg, 0.01 mmol) in anhydrous DMF were added, triethylamine (4.5 μL, 0.03 mmol) and 5(6)-carboxy-tetramethylrhodamine succinimidyl ester (NHS-Rhodamine, 6 mg, 0.01 mmol). After completion of the reaction, solvent was removed, and the residue was purified using silica gel column chromatography (20% MeOH/CH2Cl2) to obtain compound 57 as dark purple solid (8.5 mg, yield = 94%). 1H NMR (500 MHz, acetone) δ 8.38 (s, 1H), 7.99 – 8.27 (m, 2H), 7.58 – 7.73 (m, 2H), 7.55 (d, J = 8.80 Hz, 1H), 7.33 (d, J = 7.83 Hz, 1H), 6.93 – 7.11 (m, 4H), 6.52 – 6.69 (m, 5H), 4.11 – 4.19 (m, 2H), 3.86 (s, 2H), 3.68 – 3.83 (m, 3H), 3.59 (s, 2H), 3.47 (s, 2H), 3.37 (s, 1H), 2.97 – 3.12 (m, 12H), 1.78 (quin, J = 6.80 Hz, 1H), 1.59 – 1.68 (m, J = 6.85, 6.85, 6.85, 6.85 Hz, 1H), 1.37 – 1.42 (m, 3H). HRMS for C44H46ClN4O9S [M+] calculated 841.2669, found 841.2659.</p><!><p>To a solution of compound 55 (10 mg, 0.02 mmol) in anhydrous DMF were added, HATU (14.9 mg, 0.04 mmol), biotin (5.8 mg, 0.02 mmol) and triethylamine (11 μL, 0.08 mmol). After stirring overnight, the solvent was removed and the residue was purified using silica gel column chromatography (10% MeOH/CH2Cl2) to obtain compound 58 as off-white solid (14 mg, quantitative yield). 1H NMR (500 MHz, METHANOL-d4) δ 7.57 (d, J = 8.80 Hz, 2H), 7.03 (d, J = 8.80 Hz, 2H), 6.99 (s, 1H), 6.89 (s, 1H), 4.48 (dd, J = 4.89, 7.83 Hz, 1H), 4.30 (dd, J = 4.40, 7.83 Hz, 1H), 4.12 (q, J = 7.09 Hz, 2H), 3.80 (s, 3H), 3.63 (br. s., 2H), 3.39 (s, 3H), 3.17 – 3.29 (m, 3H), 2.92 (dd, J = 4.89, 12.72 Hz, 1H), 2.70 (d, J = 12.72 Hz, 1H), 2.17 (t, J = 7.34 Hz, 2H), 1.57 – 1.75 (m, 6H), 1.36 – 1.48 (m, 5H). 13C NMR (126 MHz, METHANOL-d4) δ 176.2, 166.3, 164.1, 152.6, 150.2, 132.4, 131.0, 126.8, 124.4, 118.5, 115.5, 115.2, 65.3, 63.5, 61.8, 57.3, 57.1, 56.5, 41.2, 37.8, 37.0, 29.9, 29.8, 29.6, 27.1, 15.1. HRMS for C29H40ClN4O7S2 [M + H+] calculated 655.2021, found 655.2025.</p><!><p>To a solution of compound 10 (84 mg, 0.24 mmol) in anhydrous acetonitrile were added, tert-butyl nitrite (37 mg, 0.36 mmol) and azidotrimethylsilane (33 mg, 0.29 mmol) and the reaction was heated at 45 °C for 1h. The solvent was then removed, and the residue was purified using silica gel column chromatography (20% EtOAc/hexanes) under low-light conditions to obtain compound 59 as white solid (52 mg, yield = 57%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.64 (d, J = 8.80 Hz, 2H), 7.19 (s, 1H), 6.81 – 6.89 (m, 3H), 6.36 (s, 1H), 4.04 (q, J = 6.85 Hz, 2H), 3.86 (s, 3H), 3.56 (s, 3H), 1.42 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.5, 146.3, 144.2, 130.1, 129.3, 124.7, 122.9, 114.3, 107.3, 103.8, 63.9, 56.7, 56.3, 14.6. HRMS for C16H18N4O5SNa [M + Na+] calculated 401.089, found 401.0894.</p><!><p>Combine 1,2-dimethylethylenediamine (7.2 μL, 0.069 mmol), CuI (9.2 mg, 0.057 mmol) and sodium ascorbate (8.7 mg, 0.057 mmol) in a microwave reaction vial. Seal and evacuate the vial and add H2O (300 μL). Separately combine compound 25 (50 mg, 0.11 mmol) and NaN3 (41.6 mg, 0.23 mmol) in EtOH (350 μL) and DMF (350 μL) and add to the reaction vial. Fill vial with argon gas and irradiate reaction using microwave at 100 °C for 1h. Water was poured into the reaction mixture and extracted with EtOAc. The organic layer was collected, solvent was removed to obtain the residue which was purified using silica gel column chromatography (30% EtOAc/hexanes) to obtain compound 60 as a tan solid (28 mg, yield = 63%). 1H NMR (500 MHz, METHANOL-d4) δ 7.16 (s, 1H), 7.02 – 7.10 (m, 2H), 6.89 (s, 1H), 6.84 (d, J = 8.56 Hz, 1H), 3.86 (s, 3H), 3.80 (s, 3H), 3.55 (s, 3H). 13C NMR (126 MHz, METHANOL-d4) δ 152.1, 150.5, 147.1, 138.9, 132.7, 127.1, 119.6, 119.0, 114.7, 113.7, 110.4, 109.8, 57.2, 57.2, 56.4. HRMS for C15H18ClN2O5S [M + H+] calculated 373.0619, found 373.062.</p><!><p>To a solution of compound 60 (23 mg, 0.06 mmol) in anhydrous acetonitrile were added, tert-butyl nitrite (10 mg, 0.09 mmol) and azidotrimethylsilane (9 mg, 0.07 mmol) and the reaction was heated at 45 °C for 1h. The solvent was then removed, and the residue was purified using silica gel column chromatography (25% EtOAc/hexanes) under low-light conditions to obtain compound 61 as light yellow solid (22 mg, yield = 88%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.49 (dd, J = 1.96, 8.56 Hz, 1H), 7.38 (d, J = 1.71 Hz, 1H), 7.24 (s, 1H), 6.95 (s, 1H), 6.85 (d, J = 8.56 Hz, 1H), 6.80 (s, 1H), 3.91 (s, 3H), 3.88 (s, 3H), 3.66 (s, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 155.4, 149.3, 143.5, 131.2, 129.2, 125.4, 124.8, 119.2, 118.2, 113.1, 111.2, 106.2, 56.8, 56.4, 56.3. HRMS for C15H15ClN4O5SNa [M + Na+] calculated 421.0344, found 421.0348.</p><!><p>To a solution of compound 61 (10 mg, 0.03 mmol) in anhydrous DMF were added, potassium carbonate (7.0 mg, 0.06 mmol) and propargyl bromide solution in toluene (32j, 3.23 μL, 0.03 mmol). The reaction was heated at 45 °C for 2h, followed by removal of the solvent. The residue was then dissolved in EtOAc and washed by water and brine, dried over sodium sulfate, concentrated under vacuum to obtain the residue which was purified by silica gel column chromatography (28% EtOAc/hexanes) to obtain compound 62 as off-white solid (10 mg, yield = 89%). 1H NMR (500 MHz, CHLOROFORM-d) δ 7.62 (d, J = 8.80 Hz, 2H), 6.81 – 6.97 (m, 3H), 6.41 (s, 1H), 4.43 (br. s., 2H), 4.08 (q, J = 7.10 Hz, 2H), 3.80 (s, 3H), 3.42 (s, 3H), 2.17 (t, J = 2.40 Hz, 1H), 1.44 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, CHLOROFORM-d) δ 162.3, 151.0, 145.6, 131.2, 129.8, 129.2, 122.6, 117.4, 114.1, 104.2, 78.6, 73.1, 63.9, 56.6, 55.7, 39.7, 14.6. HRMS for C19H20N4O5SNa [M + Na+] calculated 439.1047, found 439.105.</p><!><p>To a solution of compound 61 (27 mg, 0.07 mmol) in anhydrous DMF were added, potassium carbonate (19.5 mg, 0.14 mmol) and tert-butyl (3-bromopropyl)carbamate (32t, 22 mg, 0.09 mmol). The reaction was heated at 45 °C for 2h, followed by removal of the solvent. The residue was then dissolved in EtOAc and washed by water and brine, dried over sodium sulfate, concentrated under vacuum to obtain the residue which was purified by silica gel column chromatography (35% EtOAc/hexanes) to obtain N-Boc protected intermediate (14 mg) which was stirred in a solution of 4N HCl in dioxane for 1h. The solvent was then removed to obtain the compound 63 as tan solid (12 mg, yield = 36%). 1H NMR (500 MHz, METHANOL-d4) δ 7.58 (d, J = 8.80 Hz, 2H), 7.04 (d, J = 8.80 Hz, 2H), 6.79 (s, 1H), 6.56 (s, 1H), 4.12 (q, J = 7.09 Hz, 2H), 3.79 (s, 3H), 3.67 – 3.72 (m, 2H), 3.40 (s, 3H), 3.13 (t, J = 7.58 Hz, 2H), 1.77 (quin, J = 7.00 Hz, 2H), 1.42 (t, J = 6.97 Hz, 3H). 13C NMR (126 MHz, METHANOL-d4) δ 164.3, 153.1, 147.8, 132.1, 131.2, 131.0, 124.2, 118.3, 115.6, 106.5, 65.3, 57.5, 56.4, 38.6, 28.0, 15.1. HRMS for C19H26N5O5S [M + H+] calculated 436.1649, found 436.1652.</p><!><p>To a solution of compound 63 (9 mg, 0.02 mmol) in anhydrous DMF were added, HATU (8.6 mg, 0.02 mmol), biotin (4.3 mg, 0.02 mmol), and triethylamine (6.6 μL, 0.05 mmol). Th reaction was stirred overnight, followed by removal of the solvent to obtain the residue which was purified using silica gel column chromatography (10% MeOH/CH2Cl2) to obtain compound 64 as off-white solid (8.4 mg, 67%). 1H NMR (500 MHz, METHANOL-d4) δ 7.91 (br. s., 1H), 7.57 (d, J = 8.80 Hz, 2H), 7.02 (d, J = 8.80 Hz, 2H), 6.84 (s, 1H), 6.51 (s, 1H), 4.48 (dd, J = 5.01, 7.70 Hz, 1H), 4.30 (dd, J = 4.40, 7.83 Hz, 1H), 4.12 (q, J = 6.85 Hz, 2H), 3.80 (s, 3H), 3.61 (br. s., 2H), 3.36 (s, 3H), 3.16 – 3.29 (m, 3H), 2.92 (dd, J = 5.14, 12.72 Hz, 1H), 2.70 (d, J = 12.72 Hz, 1H), 2.17 (t, J = 7.34 Hz, 2H), 1.51 – 1.80 (m, 6H), 1.36 – 1.49 (m, 5H). 13C NMR (126 MHz, METHANOL-d4) δ 176.3, 176.2, 166.3, 164.1, 153.0, 147.6, 132.5, 131.0, 130.7, 124.3, 118.8, 115.5, 106.2, 65.3, 63.5, 61.8, 57.5, 57.1, 56.3, 41.2, 37.8, 37.0, 29.9, 29.7, 29.6, 27.1, 15.1. HRMS for C29H39N7O7S2Na [M + Na+] calculated 684.2245, found 684.2241.</p><!><p>The THP1-Blue™ NF-κB cell line was purchased from Invivogen (San Diego, CA) which contains a stably integrated NF-κB-inducible secreted embryonic alkaline phosphatase (SEAP). ISRE-bla THP-1 cell line was generated by us as described earlier.42 QuantiBlue was purchased from Invivogen, MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-dipheyl tetrazolium bromide) was purchased from Acros Organics, LPS (lps-eb) from Invivogen, and IFN-α from R&D Systems (#11200–2).</p><!><p>THP1-Blue™ NF-κB cells were plated in 96-well plates at 105 cells/well in 100 μl RPMI supplemented with 10% fetal bovine serum (FBS, Omega Scientific, Inc., Tarzana, CA), 100 U/mL penicillin, 100 μg/ml streptomycin (Thermo Fisher Scientific) and Normocin (Invivogen). LPS was prepared in assay medium at a concentration of 20 μg/mL. Tested compounds were dissolved in DMSO at 1 mM as a stock solution and were further diluted in the LPS solution to a final concentration of 10 μM. 100 μL of this solution was then transferred to the plated cells to obtain a final concentration of LPS at 10 μg/mL and compound at 5 μM (0.05% DMSO). The culture supernatants were harvested after a 20h incubation period. SEAP activity in the culture supernatants was determined by a colorimetric assay using QuantiBlue (Invivogen). Plate absorbance was read at 630 nm using a Tecan Infinite M200 plate reader (Männedorf, Switzerland). The SEAP concentration was directly proportional to NF-κB activity, which was 2-point normalized to yield activity of compound 1 + LPS as 200% and activity for LPS as 100%.</p><!><p>ISRE-bla THP-1 cells were plated in 96-well plates at 5×104 cells/well in 50 μl RPMI supplemented with 10% dialyzed FBS (Atlanta Biologicals, Inc., GA), 0.1 mM non-essential amino acids, 1 mM sodium pyruvate, 100 U/mL penicillin and 100 μg/ml streptomycin. Type I IFN-α (R&D Systems, #11200–2) solution was prepared in assay medium at a concentration of 200 U/mL Tested compounds were dissolved in DMSO at 1 mM and were further diluted in the IFN-α solution to a final concentration of 10 μM. 50 μL of this solution was then transferred to the plated cells to obtain a final concentration of IFN-α at 100 U/mL and compound at 5 μM (0.05% DMSO). The cells were incubated for 16h, after which 20 μL of 6xLiveBLAzer™ FRET B/G Substrate (CCF4-AM) mixture (prepared according to the manufacturer's instructions) was added to each well. Plates were incubated at room temperature in the dark for 3h. Fluorescence was measured on a Tecan Infinite M200 plate reader at an excitation wavelength of 405 nm, and emission wavelengths of 465 nm and 535 nm. Background values (cell free wells at the same fluorescence wavelength) were subtracted from the raw fluorescence intensity values and the emission ratios were calculated as the ratio of background subtracted fluorescence intensities at 465 nm to background subtracted fluorescence intensities at 535 nm. The ISRE activity values for these compounds were 2-point normalized to yield activity of compound 1 + IFN-α as 200% and activity for IFN-α as 100%.</p><!><p>THP-1 cells were plated in 96-well plates (105 cells/well) in 100 μL RPMI supplemented with 10% FBS, 100 U/mL penicillin and 100 μg/ml streptomycin. Compounds were dissolved in DMSO at 1 mM stock solution and were further diluted to 10 μM in the assay medium. 100 μL of this solution was added to the cells to obtain a final compound concentration of 5 μM (0.05% DMSO). After 18h incubation, a solution of MTT in assay media (0.5 mg/mL) was added to each well and further incubated for 4 to 6 h, followed by addition of cell lysis buffer (15% w/v SDS and 0.12% v/v 12N HCl aqueous solution), incubated overnight, and then absorbance measured at 570 nm using 650 nm as reference using Tecan Infinite M200 plate reader.</p><!><p>Seven to nine-week-old C57BL/6 (wild-type, WT) mice were purchased from The Jackson Laboratories (Bar Harbor, MA). All animal experiments received prior approval from the UCSD Institutional Animal Care and Use Committee.</p><!><p>WT mice (n=8 per group) were immunized in the gastronemius muscle with ovalbumin (20 μg/animal) mixed with MPLA (10 ng/animal) and compound 1 or 12 or 33 (50 nmol/animal) on days 0 and 21. On day 28, immunized mice were bled and OVA-specific IgG titers were measured by ELISA as previously described.60</p><!><p>Data are represented as mean ± standard error of the mean (SEM). Origin 7 (Origin Lab, Northampton, MA) graphing software was used for figure preparation while Prism 4 (GraphPad, San Diego, CA) software was used for statistical calculations.</p>

PubMed Author Manuscript

A Minimalistic Coumarin Turn-On Probe for Selective Recognition of Parallel G-Quadruplex DNA Structures

G-quadruplex (G4) DNA structures are widespread in the human genome and are implicated in biologically important processes such as telomere maintenance, gene regulation, and DNA replication. Guanine-rich sequences with potential to form G4 structures are prevalent in the promoter regions of oncogenes, and G4 sites are now considered as attractive targets for anticancer therapies. However, there are very few reports of small “druglike” optical G4 reporters that are easily accessible through one-step synthesis and that are capable of discriminating between different G4 topologies. Here, we present a small water-soluble light-up fluorescent probe that features a minimalistic amidinocoumarin-based molecular scaffold that selectively targets parallel G4 structures over antiparallel and non-G4 structures. We showed that this biocompatible ligand is able to selectively stabilize the G4 template resulting in slower DNA synthesis. By tracking individual DNA molecules, we demonstrated that the G4-stabilizing ligand perturbs DNA replication in cancer cells, resulting in decreased cell viability. Moreover, the fast-cellular entry of the probe enabled detection of nucleolar G4 structures in living cells. Finally, insights gained from the structure–activity relationships of the probe suggest the basis for the recognition of parallel G4s, opening up new avenues for the design of new biocompatible G4-specific small molecules for G4-driven theranostic applications.

a_minimalistic_coumarin_turn-on_probe_for_selective_recognition_of_parallel_g-quadruplex_dna_structu

Introduction<!>Examples of Previously Published Coumarin Derivatives Used for G4 DNA and RNA Detection in Vitro and in Cells.19,35,43−47,<!>Introduction<!>Design of the Molecular Probes<!>Characterization of the Compounds by Solvent-Dependent Studies<!><!>Characterization of the Compounds by Solvent-Dependent Studies<!>G4-Sensing Studies<!><!>G4-Sensing Studies<!><!>G4-Sensing Studies<!><!>G4-Sensing Studies<!>2a Stabilizes G4 Structures<!><!>2a Binds by Stacking on the Terminal G-Tetrad<!><!>2a Binds by Stacking on the Terminal G-Tetrad<!>Cellular Imaging Reveals 2a Uptake in Live Cells<!><!>Cellular Imaging Reveals 2a Uptake in Live Cells<!>2a Impaired DNA Replication and Reduced Cell Viability in HeLa Cells<!><!>2a Impaired DNA Replication and Reduced Cell Viability in HeLa Cells<!>Conclusions<!>Materials<!>G4 Folding<!>UV-vis Absorption and Steady-State Emission Measurements<!>Electronic Circular Dichroism (ECD) Measurements<!>Spectrophotometric and Fluorimetric Titrations<!>Nonlinear Global Fitting of the Binding Isotherms<!>Limit of Detection<!>Job’s Plot<!>Nuclear Magnetic Resonance (NMR) Titrations<!>Polyacrylamide Gel Electrophoresis (PAGE)<!>DNA Polymerase Stop Assay<!>Primer 5′-3′<!>c-MYC Pu24T 5′-3′<!>NonG4 5′-3′<!>Cell Viability<!>DNA Fiber Analysis for HeLa Cells<!>Fluorescence Microscopy<!><!>Author Contributions<!>

<p>G-quadruplexes (G4s) are four-stranded non-B DNA helical structures formed by the stacking of four in-plane guanine bases (G-quartets) stabilized through Hoogsteen-type hydrogen bonding and coordinated by a central metal ion (usually K+ or Na+).1−4 Extensive and detailed biophysical and structural studies have highlighted an impressive diversity of G4 topologies (including parallel, antiparallel, and hybrid structures) depending on the number of G-quartets, the strand orientation, the loop composition, and the nature of the stabilizing cation.5,6 Compelling evidence clearly implicates G4 motifs in key biological processes, including telomere maintenance,7,8 translational9−11 and transcriptional regulation,12,13 and DNA replication.14,15 Recent studies using G4-specific monoclonal antibodies16−18 or optical reporters19−23 have provided evidence for G4 formation in cells. Furthermore, bioinformatics24,25 and sequencing26,27 approaches have highlighted the widespread distribution of potential G4-forming sequences in the human genome. G4 motifs are enriched in the promoter regions of oncogenes, at telomeres, and in the untranslated regions (UTRs) of mRNAs.5 Indeed, these sites are now viewed as attractive targets for small molecule therapeutics and diagnostic agents.28,29 For instance, by targeting G4s with small molecules it is possible to control the expression/transcription of otherwise undruggable oncogenic proteins such as c-MYC30−32 and KRAS.33</p><p>A longstanding goal in the G4 field has been to develop topology-specific G4-interactive compounds capable of detecting intracellular G4s located in the nucleus. Different classes of optical probes have been reported to fluoresce upon G4 binding,29,34 but the overwhelming majority of the designed ligands are unable to differentiate between topological classes of G4s or to selectively recognize specific G4 motifs.3 Therefore, achieving selectivity between different G4 topologies still remains one of the most challenging tasks in this field. Some selective compounds have been recently reported such as a coumarin–quinazolinone,35 a quinazoline–quinazolinone,36 a core-extended naphthalene diimide,37 squaraine dyes,38,39 a thiazole peptide,31 and anthracene-based,40 BODIPY-based,41 triarylimidazole-based,42 and bis(quinolinium) pyridodicarboxamide-based43 compounds. However, because of multiple mechanisms of action, chemical scaffolds that fall far outside the "druglike" chemical space, high molecular weights, the induction of conformational changes, and complicated multistep synthetic procedures further strategies are required to rationally design G4-interacting small optical probes as easily accessible diagnostic tools.</p><p>To detect G4s in living cells, the probe should have both high selectivity for specific G4 structures/sequences and good membrane permeability. These basic characteristics, although prerequisites, are rarely simultaneously fulfilled by the same molecular agent. Within the class of G4-probes, coumarin dyes are among the most studied compounds. In previous studies, the coumarin scaffolds were usually extensively modified through the insertion of π-extended heterocyclic aromatic motifs in order to tune both the photophysical properties and G4-binding properties to both DNA and RNA (Scheme 1).19,35,43−47 Even if all these molecular recognition strategies have been successfully implemented in in vitro models and, in some cases, in fixed and live cells, the molecular sizes of some of these newly generated probes are not ideal for membrane permeability and might therefore hinder the detection of G4 DNA in live cells.</p><!><p>The structures of the coumarin analogues reported by us in previous studies35,47 and the structures of the three derivatives used in the present study (2a-c). The coumarin scaffold is marked in red.</p><!><p>The most-selective G4 probes show high signal in the G4-rich nucleoli,15,19,20,35,36 which are highly dense multifunctional domains in which ribosome biogenesis occurs with a high level of transcriptional activity involving both G4 DNA and RNA structures.20</p><p>Inspired by our recently published results on the use of coumarin–quinazolinone,35 coumarin–benzothiazole,47 and quinazoline–quinazolinone15,36 compounds, we synthesized three low-molecular-weight amidinocoumarin derivatives with druglike characteristics that differed only by the nature of their electron-donating substituents that determine their fluorescence properties (Scheme 1). By testing different G4 oligonucleotides, we found that two of these probes exhibited topology-specific G4-binding properties, but only one of them showed promising potential to detect G4 DNA in cells. This molecule displayed negligible background fluorescent signal in its unbound state. However, upon interaction with parallel G4s, a twisted intramolecular charge transfer (TICT) process opened competitive radiative relaxation pathways that led to a marked light-up fluorescence response. The resulting binding events enabled us to discriminate parallel G4s over antiparallel and non-G4 topologies through visible color changes detectable by the naked eye. The structural details of the compound's binding interactions with parallel G4 c-MYC promoter structures were also assessed by 1D 1H NMR titration studies, showing stacking interactions to the terminal G-tetrads. Moreover, we showed that DNA synthesis in cells, studied at single-molecule level, were slowed. The slower DNA synthesis was also confirmed by a biochemical DNA polymerase stop assay that clearly showed the ability of this compound to arrest DNA synthesis prior to the G4 structure. Finally, intracellular studies indicated that this small water-soluble optical probe decrease the viability of cancer cells, and that it is capable of rapid cellular entry and nucleolar localization, thus enabling detection of G4 DNA structures in live cells.</p><!><p>All tested compounds had low molecular weights (<300 Da) and obey Lipinski's "rule of five", which predicts druglike properties. The compounds were synthesized in a single step (see Scheme S1 and Figures S1–S3) through Knoevenagel condensation of commercially available substituted ortho-hydroxyl benzaldehydes (1a-c) with ethyl cyanoacetate in the presence of ammonium acetate via in situ formation of 3-cyanocoumarin, which on subsequent reduction led to the formation of the desired 3-amidinocoumarin derivatives (2a-c) (Scheme 1).</p><!><p>We first determined the spectroscopic properties of the coumarin derivatives 2a–2c and performed solvent-dependent UV/vis absorption and emission measurements in order to determine how different solvent polarities affect the optical properties of the compounds. For these measurements, we used seven different solvents with different polarities (water, methanol (MeOH), ethanol (EtOH), acetonitrile (ACN), dichloromethane (DCM), chloroform (CHCl3), ethyl acetate (EtOAc), and tetrahydrofuran (THF)). By increasing the solvent polarity, the absorption maximum (λmax) was increased from 409 nm to 445 nm (Δλmax = 36 nm) and 432 to 466 nm (Δλmax = 34 nm) for 2a and 2c, respectively (Figures 1 and S4). Emission studies in these solvents demonstrated a shift of the λem to longer wavelengths at increasing solvent polarities. This bathochromic shift of the emission band was 474–486 nm (Δλem = 12 nm) and 487–502 nm (Δλem = 15 nm) for 2a and 2c, respectively (Figures 1 and S4). The plots of solvent polarity versus absorption/emission maximum indicated an overall positive solvatochromic behavior for 2a and 2c (Figure 1), showing that the red-shifted absorption and emission spectra of the compounds are dependent on an increased solvent polarity.</p><!><p>Comparison of polarity-dependent band shifts for 2a (upper panel), 2b (central panel), and 2c (bottom panel) on a normalized solvent polarity scale (ETN).48</p><!><p>For 2b, the solvent-dependent absorption and emission studies both indicated a negative solvatochromic effect of the compound with λmax increasing from 417 nm to 455 nm (Δλmax = 38 nm) and λem ranging from 451 nm to 468 nm (Δλem = 17 nm) when the solvent polarity was decreased (see Figure 1, as well as Figure S4 in the Supporting Information). These findings clearly support a greater stabilization of the first excited state of 2a and 2c, relative to the ground state, which is associated with the increased dipolar character of the molecules in high-polarity solvents.48 Conversely, the negative solvatochromism of 2b is caused by differential stabilization of the ground and the first excited state, with the former being more energetically favorable, compared to the latter.48</p><!><p>Next, we performed G4-binding studies with the different compounds to determine if they can bind and light-up upon binding to G4 DNA. In an aqueous buffer solution of high ionic strength (100 mM K+) without G4 DNA, the UV/vis spectra of 2a, 2b, and 2c revealed well-defined intramolecular charge transfer (ICT) bands centered at 445, 416, and 466 nm, respectively (see Figures 2A and 2B, as well as Figure S5 in the Supporting Information). Upon titration of 2a and 2c with c-MYC Pu22, which is a well-characterized parallel G4 DNA structure (see Table S1 in the Supporting Information),49−52 monotonic hypochromism (H ≈ 26%) of the ICT band, along with a pronounced bathochromic shift of 13 and 14 nm, respectively, was observed (Figures 2A and 2B). Moreover, the appearance of well-defined isosbestic points centered at 370 and 461 nm for 2a-c-MYC Pu22 and 373 and 482 nm for 2c-c-MYC Pu22 indicated the formation of a structured ligand/G4 complex. These spectral modifications can be ascribed to specific short-range interactions between the hydrophobic central coumarin core that is prone to π-stacking and the in-plane G-tetrad scaffold, which decrease the π–π* energy gap and result in a red-shifted λmax.</p><!><p>Absorption and fluorescence titrations of (A) 2a and (B) 2c in complex with c-MYC Pu22. The UV/vis absorption (solid lines) and fluorescence emission spectra (dashed lines) of 2a and 2c upon gradual addition of c-MYC Pu22 (3.0 μM or 1.5 μM 2a/2c, 100 mM KCl, 50 mM Tris-buffer (pH 7.5); λexc(2a) = 461 nm and λexc(2c) = 482 nm). The blue and red solid lines correspond to the UV/vis spectra at (c-MYC Pu22/2a) 0.0 and 11.33 equiv and (c-MYC Pu22/2c) 0.0 and 4.66 equiv, respectively. The blue and red dashed lines correspond to the emission spectra at (c-MYC Pu22/2a) 0.0 and 23.33 equiv and (c-MYC Pu22/2c) 0.0 and 6.66 equiv, respectively. The blue and red arrows show the evolution of the binding profile at the beginning and end of the titration, respectively. The black dashed arrows show the appearance of isosbestic points. The emission intensity was normalized to the absorption maximum. The inset in panel (A) shows the color change derived from the addition of c-MYC Pu22 to 2a as detected by the naked eye, using a 312 nm UV-lamp (3.0 μM 2a, 34 μM c-MYC Pu22).</p><!><p>Under the same experimental conditions as for 2a and 2c, 2b exhibited negligible spectral changes, highlighting the inability of this compound to interact with c-MYC Pu22 (see Figure S5 in the Supporting Information). We hypothesize that the lack of G4 binding response by 2b might arise from either the poor electron-donating character of the hydroxyl group and/or from the reduced overall net positive charge occurring at pH 7.5 (vide infra).</p><p>In contrast to the absorption changes, the steady-state emission spectrum of 2a, which is almost fully quenched in its unbound state (Figure 2A, blue dashed line), showed a marked increase in the fluorescence signal along with a pronounced bathochromic effect of 4 nm upon c-MYC Pu22 titration (Figure 2A, red dashed line). This effect in the presence of c-MYC Pu22 induced a color change that could be detected by the naked eye (inset of Figure 2A).</p><p>The fluorescence intensity of compound 2c was, in contrast to 2a, gradually quenched (27%) and bathochromically shifted by 7 nm by the addition of c-MYC Pu22 (Figure 2B). These spectral changes for 2c may be attributed to the increased hydrophobic interactions between the G4 nucleobases and the flat and rigid julolidine moiety, which might lead to competitive radiationless relaxation pathways. Also, the change in fluorescence in the presence of c-MYC Pu22 was not as dramatic for 2c (∼1.5-fold) as it was for 2a (∼14-fold) (Figure 2). Finally, similar to the UV/vis absorption data, compound 2b did not show any changes in the emission intensity in the presence of c-MYC Pu22, thus showing no ability to detect c-MYC Pu22 (Figure S5).</p><p>The results above prompted us to study the recognition process of our probes toward biologically relevant natural and synthetic G4 structures forming parallel, antiparallel, and hybrid topologies (Table S1 and Figures S6–S8 in the Supporting Information), including those found in the promoter regions of the c-MYC, c-KIT, BCL-2, VAV genes, human telomeres (Tel-22), and TBA.53 As shown in Figures 3A and 3B, as well as Figures S9–S12 in the Supporting Information, both 2a and 2c exhibited a clear-cut binding preference for parallel G4 topologies over antiparallel and non-G4 structures. Both 2a and 2c showed the strongest turn-on and turn-off emission response in the presence of c-MYC Pu22, respectively. Moreover, no binding response was observed for 2b in the presence of parallel, antiparallel, or hybrid G4 morphologies (see Figure S5). Interestingly, the selective light-up properties of 2a toward parallel G4 structures were also detected by the naked eye (Figure S13 in the Supporting Information) or on a native nondenaturing polyacrylamide gel (Figure S14 in the Supporting Information).</p><!><p>Spectrofluorimetric binding isotherms of (A) 2a and (B) 2c complexed with various G4 and non-G4 structures showed a clear-cut preference for parallel/hybrid G4 topologies over nonparallel and non-G4 structures. The dashed black lines correspond to a 1:1 fitting model at λem whose values are reported in Table 1 (1.5 μM 2a/2c, 100 mM, KCl, 50 mM Tris-buffer (pH 7.5); λexc(2a) = 461 nm, λexc(2c) = 482 nm). Parallel G4 in red, hybrid G4 in blue, antiparallel G4 and non-G4 in gray.</p><!><p>The molecular recognition ability of 2a and 2c for parallel G4s can be attributed to their highly accessible G-quartet surfaces that provide better π-stacking platforms for the accommodation of the coumarin core. The different evolutions of fluorescence intensity observed for 2a and 2c in the presence of parallel G4s suggested that rigidification of 2a's chromophoric system upon binding the G4 leads to the increased photoluminescence quantum yield (PLQY) through a potential TICT process (Figure S15), whereas the inherent planarity of 2c results in intense intermolecular chromophore–nucleobase interactions that are responsible for the resulting fluorescence quenching. Finally, to determine the sensitivity of 2a for all four tested parallel G4 structures, we performed limit of detection (LOD) measurements. LOD was calculated by using the measured emission values of 2a in complex with various concentrations of the different parallel G4 structures (see Figure S16 and Table 1). The calculated LOD values ranged from ∼128 nM to 342 nM (Table 1).</p><!><p>Percentage of hypochromic effect (H) on εmax.</p><p>Fitting with a 1:1 binding model was obtained with Bindfit by using multiple global fitting methods (Nelder–Mead method) on the fluorimetric data.</p><p>Coumarin 153 in ethanol (ΦF = 38%) was used as the standard. PLQY are calculated at oligonucleotide/ligand ratio = 5.</p><!><p>Both 2a and 2c showed a net positive charge under our experimental conditions, as shown by structure-based calculations computed with Marvin Sketch software. The microspecies distribution of the compounds in the entire pH range is provided in Figure S17 in the Supporting Information. The number of charges at pH 7.5 was ∼0.5 for 2b and ∼1.0 for 2a and 2c. However, despite the positive charge, no or very low off-target binding was detected with the negatively charged phosphate backbone of the ss- and ds-DNA. These results clearly indicate that the coordination mechanism of both coumarin derivatives toward the parallel G4 templates is not driven by nonspecific electrostatic interactions.</p><p>To further investigate the interplay between the G4 templates and fluorescence signal changes, we performed quantitative binding analysis on the fluorimetric titrations (Figures 3A and 3B and Table 1, as well as Table S2). In all cases, a global nonlinear curve fitting procedure based on a 1:1 (ligand:DNA) stoichiometry model fit well to our experimental fluorescence output data. This 1:1 stoichiometry for the 2a:c-MYC Pu22 system was also supported by a Job's plot where we plotted the integrated emission area as a function of the mole fractions (Figure S18).</p><p>These experiments allowed us to calculate the dissociation constant (Kd) for 2a and 2c with the different DNA structures using both spectrophotometric and fluorimetric data (see Table 1, as well as Table S2). Compound 2c complexed with parallel G4s gave the best Kd values ranging from 0.3 μM to 0.7 μM. In the presence of the hybrid telomeric G4 structure, 2c showed a 5-to-12-fold lower binding affinity (Kd = 3.8 μM). A similar trend was observed for 2a that exhibited Kd values ranging from 9.1 μM to 15.0 μM in the presence of the different parallel G4s. The Kd of 2a in the presence of Tel-22 was 45.2 μM. This value was ∼5-fold higher, compared with the Kd value of the best parallel G4 sequence (c-KIT 2) used in this study.</p><p>These data were fully consistent with the fluorescence response, confirming the preference of the compounds for parallel G4s. The Kd values of the 2a-c-MYC Pu22 and 2c-c-MYC-Pu22 systems calculated using the fluorescence data were also found to be in good agreement with those calculated in our UV-vis titration experiments (see Table S2). No quantitative binding data analysis was performed for the compounds complexed with antiparallel and non-G4 structures, because the optical response of the probes, even in the presence of a large excess of biological templates, did not show a clear binding isotherm characterized by a hyperbolic saturation profile. Overall, these data outlined the ability of 2a and 2c to selectively target parallel G4s over antiparallel and non-G4 structures. Generally, the Kd values obtained for 2a complexed with parallel G4 structures are higher, compared to those reported for other selective parallel G4 ligands (Kd values between 10 and 0.1 μM).31,35,37,40−42 On the other hand, 2c showed excellent parallel G4-interactive binding properties that match some of the best topology selective G4-binders reported so far.31,35,37,40−42</p><p>Because both 2a and 2c featured the same chromophoric system and number of positive charges, we speculate that their different binding strengths are due to the nature of the electron-donating substituents. The higher binding affinities demonstrated for the 2c-G4 DNA complexes compared to the 2a-G4 DNA complexes are likely linked to additional hydrophobic interactions in combination with a reduced entropic loss upon binding of 2c, because the julolidine moiety already is locked in a flat conformation.</p><p>Unfortunately, the fluorescence quenching of 2c bound to parallel G4s may prevent its application as a cellular fluorescent reporter. Therefore, we focused our attention on compound 2a that showed optical properties suited for in-cellulo studies.</p><!><p>To determine if 2a not only selectively binds and lights up G4s, but also stabilizes these structures, we performed a Taq-DNA polymerase stop assay (Figure 4A). In this assay, we determined the stabilization and selectivity by comparing the effect of the compound on DNA synthesis of a non-G4 or c-MYC G4 template. The reaction products were loaded onto a denaturing polyacrylamide gel, and progression and pausing of the DNA polymerase was determined at single nucleotide resolution.15,36 By increasing the 2a concentration, we found increased amounts of replication pausing one nucleotide before the first G-tract on the G4 template as well as decreased amounts of full-length products (Figures 4B and 4C). This showed that 2a inhibits DNA synthesis of Taq-DNA polymerase by selectively stabilizing the c-MYC G4 template in a concentration-dependent manner. The non-G4 template was not affected by 2a, thus supporting the selectivity of our probe for c-MYC G4. The selective stabilization of the G4 structure and pausing before the first G-tract by 2a was confirmed by using the well-known G4-stabilizer Phen-DC3 as a control.</p><!><p>Compound 2a stabilizes the c-MYC Pu24T structure and selectively inhibits DNA synthesis by Taq-DNA polymerase during DNA replication, while DNA replication on the non-G4 DNA template is not affected. (A) Schematic representation of the Taq polymerase stop assay with either a non-G4 template or the c-MYC Pu24T G4 DNA template. (B) Taq-DNA polymerase stop assay with 2a using c-MYC Pu24T and non-G4 DNA templates. The concentration of 2a is indicated above each lane. Phen-DC3 (0.32 μM) was used as the G4-stabilizing reference compound. Black arrows indicate nonextended primer, gray arrows indicate full-length products, and asterisks indicate pausing sites. (C) Quantification of pausing effects of 2a on the c-MYC Pu24T G4 template.</p><!><p>Next, the effect of 2a on the conformation of parallel G4s was investigated by electronic circular dichroism (ECD) measurements. The ECD spectrum of parallel G4s is characterized by typical positive and negative dichroic signals at 265 and 240 nm, respectively (see Figure S19 in the Supporting Information).53 The gradual addition of 2a to the c-MYC Pu22 or c-KIT 2 solution did not induce changes in either the magnitude or the shape of the bands. These data suggest that 2a can exert its topology-specific sensing/stabilization function without inducing topological transitions on the G4 template.</p><p>In order to determine the site localization of our probe on the c-MYC G4 templates (c-MYC Pu22 and c-MYC Pu24T), we performed 1D 1H NMR titration studies (see Figures 5A and 5D). In these experiments, the G4 DNA was titrated with increasing concentrations of 2a and the changes occurring in the imino protons of the G4 templates were analyzed.32,49,54 In the c-MYC Pu22-2a system, a clear shift for all the imino protons belonging to the 5′-quartet was observed, indicating specific interactions between 2a and the 5′-end of the G4 (Figure 5A). No or very weak changes of the imino protons belonging to the central quartet were observed throughout the titration, thus ruling out their involvement in the complexation mechanism. Both the G9 and G18 residues located at the 3′-end displayed evident changes, whereas G13 and G22 remained almost unaffected. An explanation for this observation could be that 2a can only partially access the 3′-end, thus giving rise to specific interactions only with the G9 and G18 sites (Figure 5A).</p><!><p>1H NMR (850 MHz) titrations for (A) c-MYC Pu22 and (D) c-MYC Pu24T complexed with 2a. The initial G4 concentration was 90 μM, and 2a was then added so that the last addition corresponded to a total molar ratio of G4 DNA:2a of 1:5. The guanines involved in the formation of the G4 structure are color coded. G20 is not involved in the G4 structure.54 Graphical representation of (B, C) c-MYC Pu22 (PDB: 5W77) and (E, F)c-MYC Pu24T (PDB: 2MGN) generated in MOE. The top images are viewed from the 5′-side of the G4-structures. The guanines in the quadruplex structures are color-coded, based on how affected their imino protons were throughout the titration experiments. Red: strong changes, blue: moderate changes, and gray: no or very weak changes.</p><!><p>An analogous experiment performed with c-MYC Pu24T, a sequence that features the same 5′-G-tetrad/flanking residue of c-MYC Pu22 but a considerably different 3′-terminal G-tetrad end and flanking sequence,54 showed pronounced structural changes at the 5′-quartet of the G4 template, thus supporting the ability of 2a to coordinate the 5′-end of the c-MYC G4s (Figure 5D). Interestingly, G15 in the 3′ quartet alone was affected as well as G5 or G14 in the central quartet. To better visualize which of the guanines had the major contribution in the complexation event with 2a, a schematic illustration of c-MYC Pu22 (see Figures 5B and 5C) and c-MYC Pu24T (Figures 5E and 5F) was generated using the Molecular Operating Environment (MOE) software. In these illustrations, the guanines in the G4 scaffolds were color-coded, depending on how affected they were throughout the titration experiments.</p><p>To conclude, the chemical shifts changes of the imino protons upon 2a binding suggest that the coumarin derivative coordinates the c-MYC G4 templates through an end-stacking mode that mainly involve both the 5′- and 3′-ends.</p><p>In order to further examine the terminal binding of 2a with the c-MYC G4 template, we performed a fluorescence displacement competition assay in the presence of Phen-DC3. The resolved solution structure of Phen-DC3 bound to the G4 parallel sequence derived from the c-MYC promoter demonstrates the highest-affinity binding site at the 5′-end and a second binding event at the terminal 3′-end, thus showing that Phen-DC3 is a G4 end-stacking molecule.54 Because 2a binds with specificity to parallel G4s, exhibiting association-induced emission, we hypothesized that if competition for the same binding sites occurs between 2a and Phen-DC3, a decrease in the fluorescence intensity should be observed (Figure S20A in the Supporting Information). Indeed, we found that the fluorescence signal of the 2a-c-MYC Pu22 system was efficiently quenched and blue-shifted by the addition of increasing concentrations of Phen-DC3 (Figure S20B in the Supporting Information). These results highlight the strong competitive behavior of the two molecules for the same binding sites and suggest that 2a binds to c-MYC G4 by stacking on the terminal G-tetrad.</p><p>To further examine the terminal stacking binding mode of 2a to parallel G4 structures, we used another G4 structure, c-MYC sG4, which has an identical internal sequence as c-MYC Pu22 except for the 5′ and 3′ terminal flanking regions, but still forms a parallel G4 structure (c-MYC Pu22 = 5′-TGAGGGTGGGTAGGGTGGGTAA-3′; c-MYC sG4 = 5′-GGGTGGGTAGGGTGGG-3′).35,38 By lacking these flanking regions, we hypothesized that c-MYC sG4 would provide better π-stacking possibilities for the planar conformation of the coumarin core (Table S1). Indeed, titration of 2a with c-MYC sG4 induced an ∼20-fold fluorescence enhancement, which was 6-fold higher, compared to that of 2a complexed with c-MYC Pu22, and the Kd value was 12.6 μM (see Figures S21A and S21Bin the Supporting Information). Furthermore, the light-up ability of 2a titrated in the presence of the competitive complementary c-MYC sG4 C-rich sequence (sC4 = 5′-CCCACCCTACCCACCC-3′) and duplex DNA was nearly unchanged (∼15-fold emission enhancement) (Figure S21C). These results highlight the high selectivity of 2a for parallel G4 structures in a complex and highly competitive environment.</p><p>Finally, we also verified the selectivity of 2a for parallel G4s by performing competitive PAGE studies (Figure S21D in the Supporting Information). In this assay, we used c-MYC Pu22 as the target template (20 μM) in the absence or presence of 100 μM of the competitive antiparallel G4 TBA, GC-rich ds-DNA (formed by annealing c-MYC Pu22 with its complementary strand), or self-complementary ds-DNA. 2a was able to impart parallel G4 specificity even in the presence of a 5-fold excess of these nonparallel G4 structures (Figure S21D). Together our in vitro data show that 2a is a selective light-up parallel G4 binder.</p><!><p>To determine if 2a can be used to visualize G4 structures in cells, we used confocal laser scanning microscopy. Fixed HeLa cells treated with 20 μM 2a revealed intense fluorescence signals both in the extranuclear cellular regions and in the subnuclear G4-rich compartments whose appearance is compatible with that of nucleoli (Figure 6A).15,35,36,55−58 The extranuclear signal suggests lysosomal accumulation of 2a, unfortunately a common feature for many G4 probes, even those that operate at single-molecule level.15,23,35,36,57,58</p><!><p>Confocal fluorescence images of HeLa cells. (A) Fixed HeLa cells stained with 2a (20 μM), Hoechst 33342 (1 μM), and corresponding bright-field (BF) and merged images. The white arrows indicate the 2a-mediated nucleoli staining. (B) Fluorescence images of fixed HeLa cells stained with 2a (20 μM) without and with RNase treatment. The BF and fluorescence images for untreated cells show the absence of any autofluorescence signal. (C) Fluorescence displacement assay with 2a (20 μM) in the absence or presence of BRACO-19 (20 μM). The BF and fluorescence images for cells treated only with BRACO-19 (20 μM) show the absence of any BRACO-19-associated fluorescence signal. (D) Fluorescence, BF, and overlay images of live HeLa cells stained with 2a (10 μM) for 10 min. The white arrows indicate the 2a-mediated nucleoli staining in living cells. The fluorescence images for untreated cells show the absence of any autofluorescence signal. Scale bar = 10 μm. Experimental settings (A–D): A 405 nm diode laser was used for Hoechst 33342 (λexc = 405 nm, λem = 410–445 nm), and an argon laser (λexc = 458 nm, λem = 470–700 nm) was used for 2a excitation.</p><!><p>Following the in vitro evidence for 2a's high selectivity for parallel DNA G4 structures, we pretreated the cells with RNase to confirm the nature of the main binding target of the compound (see Figure 6B, as well as Figure S22 in the Supporting Information). RNase treatment did not modify 2a nucleolar staining, thus indicating the ability of 2a to preferentially target DNA G4 structures.35,36 Because DNase treatment does not affect the nucleolar compartments,22,55 we validated the G4-binding ability of 2a in the nucleolar sites through a competitive binding assay using the well-known G4 binder BRACO-1936,59 (Figure 6C). In the presence of BRACO-19, the 2a-associated staining was strongly reduced (Figure S23 in the Supporting Information). These results suggest that BRACO-19 can compete for 2a's binding sites.</p><p>Finally, we asked if 2a can reach the subnuclear G4 compartments of living cells. As shown in Figure 6D, 2a clearly stained the nucleoli after a 10 min incubation, thus confirming the results obtained in fixed cells. The observed signal was specific to 2a, because, under the same experimental conditions used to image 2a, no autofluorescence signal from endogenous cellular chromophores/components was detected.</p><!><p>HeLa cells have increased amounts of G4 DNA structures compared to noncancerous cells.15,16 Therefore, we investigated whether G4 stabilization by 2a affects HeLa cell viability. For these experiments, we treated the cells with either 2a or 2b, which served as a non-G4 control compound, and assessed the metabolic activity of viable cells with the MTT cell viability assay. After treating HeLa cells with 2a for 48 h, there was a concentration-dependent reduction in cell viability with an IC50 ≈ 1.0 μM (Figure 7A). In contrast, treatment of HeLa cells with 2b had little effect on cell viability with an estimated IC50 > 50 μM. These data suggest that the higher toxicity exerted by 2a, compared with that exerted by 2b, might be due to the ability of 2a to bind and stabilize DNA G4s. The effect of 2a on cell viability was dependent on the period of exposure of the cells to the compound (Figure 7A), and the 24 h treatment with 2a showed a reduced cytotoxic effect compared to 48 h. We also monitored the cellular morphology of HeLa cells after 24 h treatment with various concentrations of 2a (Figure 7B). Similar to the results of the MTT viability assay, we observed increased rounding morphology of HeLa cell, which is an indication of cytotoxicity,60 as the concentration of 2a increased.</p><!><p>HeLa cells are sensitive to 2a resulting in impaired DNA replication. (A) Cell viability assay of HeLa cells treated with 2a (24 and 48 h) or 2b (48 h). Data are shown as the mean ± SD, n = 3. (B) Bright-field images of untreated or treated HeLa cells with 0 (mock), 2.5, 5.0, or 10.0 μM 2a for 24 h. (C) Schematic of the DNA fiber analysis. (D) Representative images of replication tracks of different lengths showing IdU labels (white) flanked by CIdU labels (red). (E) Quantification of the DNA fiber length (kb) in mock cells (−) versus treated (2a). Data represent the populations of individual DNA fibers for each condition (109 for control and 100 for treatment). Statistical analysis was performed using the nonparametric Mann–Whitney U test, and medians and p-values are indicated.</p><!><p>One explanation for the impaired cell viability might be altered DNA replication.61 We tested this hypothesis by visualizing and measuring newly replicated DNA molecules by performing DNA fiber analysis in HeLa cells (see Figures 7C–E, as well as Figure S24 in the Supporting Information).15 The mean length of the newly replicated DNA molecules in 2a-treated cells was significantly shorter than that of the untreated cells (p < 0.0001), indicating that 2a affects the rate of DNA synthesis (Figure 7E).15 The decreased replication rate might result from slower replication fork progression, due to 2a's ability to stabilize G4s.</p><!><p>With the goal of developing more accurate and efficient G4-ligands, we investigated the G4-binding ability of three coumarin derivatives having different electron-donating characteristics. By using various biophysical and biochemical methods, we have generated structure–activity relationships that provide valuable information for the design of optical sensors that use distinct fluorescence mechanisms (e.g., TICT) to signal the presence of parallel G4 topologies. In particular, we focused our attention on a small water-soluble fluorescent light-up probe capable of specifically targeting parallel G4 structures over antiparallel and non-G4 structures. This sensor selectively signaled the presence of parallel G4 morphologies via a TICT mechanism. Its striking optical changes enabled naked-eye discrimination between different G4 topologies and non-G4 structures. Furthermore, its recognition ability was very selective for parallel G4 structures even in the presence of highly competitive ds-DNA or complementary C-rich DNA. The structural origin of the compound's binding interactions with parallel G4 c-MYC promoter structures was assessed by 1D 1H NMR titration studies and showed the ability of this ligand to coordinate the G4 structures via an end-stacking binding mode. Besides its enticing quadruplex interacting and optical properties, this fluorescent sensor was also able to selectively stabilize the G4 template and inhibit DNA synthesis in vitro. Confocal fluorescence images of this probe in both fixed and live HeLa cells showed efficient cell permeability and nucleolar DNA G4 binding. Intracellular studies indicated that this compound decreased the viability of cancer cells and reduced DNA replication speed through a possible G4-dependent mechanism. We believe that the low molecular weight and straightforward synthesis of the evaluated compounds combined with the presented findings will be useful for the design of specific bioprobes with optimized optical performances and G4 binding parameters to be used in in vivo models.</p><!><p>All reagents, solvents, chemicals, and biological templates were purchased from Sigma–Aldrich or Eurofins Genomics and used without further modifications unless otherwise stated. The stock solutions of all synthesized G4-binding compounds were prepared in DMSO at a concentration of 0.5 mM unless otherwise stated. Compound 2a was also prepared in Milli-Q water at a concentration of 0.25 mM for cellular studies. The final concentration of DMSO in all the DNA-based assays was kept below 2.0% (v/v).</p><!><p>The oligonucleotides were diluted with ultrapure water to a concentration of 1 mM and stored at 5 °C. The exact oligonucleotide concentration was determined by UV-vis spectroscopy using the molar extinction coefficients (ε260) provided in Table S1 and calculated using the oligo analyzer tool on the IDT Web site. The oligonucleotides were heated at 95 °C for 5 min in the presence of 100 mM KCl and then slowly allowed to reach RT overnight.</p><!><p>UV/vis absorption spectra were recorded on a T90+ UV/vis spectrometer (PG Instruments, Ltd.) with a spectral bandwidth of 1 nm. Steady-state fluorescence spectra were recorded on a Jasco FP-6500 spectrofluorometer equipped with the Jasco Peltier-type temperature controller (Model ETC2736). The slit width of both monochromators was 3 nm. Relative fluorescence quantum yields (ΦF) were determined using Coumarin 153 as the reference (ΦF = 0.38 in EtOH) keeping the optical density (OD) < 0.1. ΦF was calculated by using the excitation wavelengths at 420 or 440 nm according to the following equation (eq 1):1where I is the integrated emission area, OD the optical density at the excitation wavelength, and η the refractive index of the solutions.</p><!><p>ECD spectra were measured with a Jasco Model J-720 spectropolarimeter that was equipped with the Jasco Peltier-type temperature controller (Model PTC-423L) and are presented as the sums of three accumulations. ECD spectra were obtained in the range of 220–490 nm with 2 μM G4 DNA. Appropriate references were subtracted from the obtained ECD spectra. All optical measurements were performed in quartz dual path length cuvettes.</p><!><p>A 3.0 μM and 1.5 μM solution of 2a–2c for UV-vis and fluorescence studies, respectively, was prepared by diluting the stock solution in a suitable amount of buffer (1.2% or 0.6% DMSO, 100 mM KCl, and 50.0 mM Tris-buffer (pH 7.5)). The freshly prepared 2a–2c solutions were titrated with the folded oligonucleotide solution and allowed to equilibrate for several minutes before recording the UV/vis or emission spectra. The concentration in each experiment was optimized to have OD < 0.15 in order to avoid reabsorption of the fluorescence emission. The excitation wavelength was set at the isosbestic point (i.e., λexc2a = 461 nm, λexc2c = 482 nm) to avoid changes in the OD. Compound 2b was excited at λexc2b = 417 nm. All of the emission spectra were baseline-corrected.</p><!><p>All of the data were corrected for the dilution upon titration. Binding constants were obtained with Bindfit using multiple global fitting methods (Nelder–Mead method) with the fluorescence data in the range of 475–555 nm and 490–545 nm for 2a and 2c, respectively.62,63 Dilution corrections were included in the fitting option. In order to ensure that we found the minima in the fitting analyses, all of the fittings were confirmed with three different start values.</p><!><p>LOD experiments were performed by plotting the changes in the emission maximum (λem) as a function of parallel G4 concentration. LOD was calculated according to the following equation:2where sb is the standard deviation calculated out of 10 independent measurements of a blank solution, k is 3 according to IUPAC recommendations, and m is the slope obtained from the linear fitting (I(λem) vs [parallel G4]).</p><!><p>The total concentration of 2a and c-MYC Pu22 was held constant (5 μM) while varying the relative proportions of 2a and c-MYC Pu22. The units on the x axis morph from concentration to mole fraction of 2a and c-MYC Pu22 (χ2a = [2a]/[2a] + [c-MYC Pu22]).</p><!><p>The G4 DNA stock solutions (180 μL) were prepared by folding 100 μM c-MYC Pu24T or c-MYC Pu22 in 10 mM potassium phosphate buffer (pH 7.4) and 35 mM KCl by heating to 95 °C and cooling to ambient temperature on ice. Then, D2O (20 μL) was added to the DNA stock solutions, yielding a final DNA concentration of 90 μM. NMR samples were prepared by sequential addition of 2a from 5 mM DMSO-d6 stock solutions to 200 μL of the DNA solution, which was then transferred to 3 mm NMR tubes. Control samples with c-MYC Pu24T and c-MYC Pu22 with and without 10% DMSO-d6 were also recorded to verify that DMSO did not have a significant effect on the G4 structures. All spectra were recorded at 298 K on a Bruker 850 MHz Avance III HD spectrometer equipped with a 5 mm TCI cryoprobe. Excitation sculpting was used in the 1D 1H NMR experiments, and 256 scans were recorded. Processing of the spectra was performed in MestreNova 10.0.2. available at http://www.chemcomp.com</p><!><p>PAGE was conducted on 20% native gels in TBE buffer supplemented with 100 mM KCl. Oligonucleotides were heated at 95 °C for 5 min in the presence of 100 mM KCl and then slowly allowed to reach RT overnight. The oligonucleotides were then loaded on the gel and electrophoresis was run at 80 V for 120 min at RT. After electrophoresis, the gel was incubated with 2a (5 μM) and, where indicated, also with Thiazole Orange (TO, 5 μM) for 30 min and rinsed with TBE buffer. Visualization was performed on a Typhoon Scanner 9200 (GE Healthcare), using an excitation wavelength of 457 nm.</p><!><p>The DNA polymerase stop assay was performed as described previously.64 Briefly, reaction mixtures containing 40 nM template DNA (G4 or non-G4) annealed to a TET-labeled primer were incubated with 0, 2, 3, 4, or 5 μM 2a in the presence of 50 mM KCl. Phen-DC3 (0.32 μM) was used as the reference G4 compound. Control reactions contained 2% DMSO instead of compound. Reactions were incubated with 0.625 U/μL Taq-DNA polymerase and incubated for 30 min at 50 °C. UV-vis spectroscopy was used to monitor the intrinsic thermal stability of 2a at 95 °C (Figure S25 in the Supporting Information). The oligonucleotide sequences used are listed below:</p><!><p>TET-TGAAAACATTATTAATGGCGTCGAGCGTCCG.</p><!><p>ATATATATATTGAGGGTGGTGAGGGTGGGGAAGGATATATATATCGGACGCTCGACGCCATTAATAATGTTTTCA.</p><!><p>GAGACCATTCAAAAGGATAATGTTTGTCATTTAGTATATGCCCCTGCTCGTCTTCCCTTCTCCGGACGCTCGACGCCATTAATAATGTTTTCA.</p><!><p>HeLa cells (4 × 103 cells/well) were seeded in DMEM high glucose media (Gibco) supplemented with 10% fetal bovine serum and penicillin-streptomycin on 96-well plates the day before the treatment. Compounds (2a or 2b) were dissolved in media at the indicated concentrations and added to the cells. At 24 or 48 h after treatment, MTT (5 mg mL–1 stock) was added to each well and the cells were incubated at 37 °C for three additional hours. DMSO was then added to each well and incubated on a shaker for 15 min. Absorbance at 590 nm was recorded using a Synergy 200 microplate reader. The data were normalized to the control and plotted with means and standard errors. Images of cells after treating with 2a for 24 h were acquired using the EVOS FL Cell Imaging System (Life Technologies).</p><!><p>Asynchronous HeLa cells at 70% confluence were seeded at 1 × 105 cells for 24 h prior to the 24 h treatment with 2a (2.5 μM) or an equivalent volume of water (control cells). Pulse-labeling of cells with IdU and CIdU and subsequent immunostaining of DNA fibers were performed as previously described.15 DNA fibers were visualized using a Leica Thunder Widefield microscope, and images were captured randomly from different fields that contained untangled fibers. Only fibers containing IdU labels flanked by CIdU labels with intact ss-DNA ends were selected for analysis using the LASX (Leica) and ImageJ software packages. A minimum of 100 individual DNA fibers were measured for each experimental condition in two independent experiments. Measurements were made in micrometers and converted to kilobases using a conversion factor for the length of a labeled track of 1 μm corresponding to roughly 2 kb.</p><!><p>HeLa cells (a cervical cancer cell line) were cultured at 37 °C in 5% CO2 in DMEM high glucose media (Gibco) supplemented with 10% fetal bovine serum and penicillin-streptomycin. For live-cell imaging, cells were treated with 2a (10 μM) for 10 min before performing microscopy. For fixed cell imaging, cells were fixed in 2% paraformaldehyde for 10 min and permeabilized with PBST (phosphate-buffered saline supplemented with 0.1% Triton X-100). Fixed cells were treated with 2a (20 μM) for 30 min at RT. The HeLa cell nuclei were visualized with Hoechst 33342 (1 μM). For the fluorescence competition assay, 2a (20 μM) was incubated with BRACO-19 (20 μM) for 30 min at RT. For the RNA degradation assay, 1 mg mL–1 RNase A (Thermo Fisher) was used and samples were preincubated with RNase A for 2 h at 37 °C prior to 2a treatment. Images were acquired on a confocal microscope Leica SP8 FALCON (Fast Life Time Contrast) using an HC PL APO 63x/1.20 Water motCORR CS2 objective. Intensity projections of Z-stack images were used for data presentation. Quantitative data analysis was performed by selecting the regions of interest and measuring the average fluorescence signal from the selected areas. All data were processed with the ImageJ software available at https://imagej.nih.gov/ij/.</p><!><p>Optical studies, G4 characterization, G4-interactive binding studies, limit of detection, structure-based calculations, Job's plot, ECD spectra, displacement assays, G-tetrad selectivity, cellular quantification and synthesis of the compounds (PDF)</p><p>cb1c00134_si_001.pdf</p><!><p>The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.</p><!><p>Work in the E. Chorell's lab was supported by the Kempe foundations (No. SMK-1632) and the Swedish Research Council (No. VR-NT 2017–05235). Work in the N. Sabouri's lab was supported by Knut and Alice Wallenberg Foundation (No. KAW2015-0189), Cancerfonden (No. CAN 2019/126), and the Swedish Research Council (No. VR-MH 2018–02651). M.D. was supported by a fellowship from the MIMS Excellence by Choice Postdoctoral Programme.</p><p>The authors declare no competing financial interest.</p>

PubMed Open Access

Enhanced Charge Transport in 2D materials through Polaritonic States

Strong light-matter interaction of functional materials has opened up an unexplored area of research to effectively utilize the exotic behavior of polaritonic states. Recent experiments suggest that material properties like charge transport can be controlled in strongly coupled systems. Fabry-Perot cavities based on metal mirrors pose a major challenge in reading out the electrical properties of hybrid states. Here, we explored the design of a one-mirror Fabry-Perot cavity and studied charge transport of 2D materials in a field-effect transistor configuration. The optical and electrical signature of strongly coupled few-layer WS2 suggest an increase in electron transport. Charge transport mobility is enhanced by ~50 times under ON reso-nance condition. A clear correlation in the effective mass of the polaritonic state and Schottky barrier height may be indicating a coherent nature of light-matter interaction.

enhanced_charge_transport_in_2d_materials_through_polaritonic_states

<!>Experimental Methods:<!>Electrical characterization:<!>Effective mass and SBH calculations:

<p>2D materials based on transition metal dichalcogenides (TMDCs) are emerging class of materials due to its remarkable optoelectronic properties. 1,2 For example, atomically thin WS2 has very large absorption cross-section, narrow linewidth as well as high electron mobility. 3 These direct band gap materials are nowadays used in many device applications that range from biosensing to optoelectronics. 4 High crystallinity, delocalized electrons and exciton absorption in the visible region are the salient features that make these 2D materials suitable for optoelectronic devices. 5 Along with these properties, TMDs are also used in valleytronics that are polarization sensitive. 6 Strong light-matter coupling was introduced to tailor the energy level that affect material properties from superconductivity to energy transfer. 7,8,9,10 So far, 2D material strong coupling was utilized in polariton lasing, spin selectivity and photoluminescence enhancement. 11,12,13,14,15 WS2 monolayers can be effectively coupled to a Fabry-Perot (FP) cavity due to its high transition dipole moment (˃ 50 Debye) that resulted in the generation of polaritonic states with splitting energy of the order of 100 meV. 16 Multilayer coupling scheme was also adopted to improve the coupling strength in these materials. 17 Other approaches include the usage of plasmonic arrays, distributed Bragg reflectors (DBR) that can also give decent coupling strength to study the nature of hybrid light-matter states. 18,19 So far, all the strong coupling studies in 2D materials are based on optical characterization and understanding the energy delocalization. 20 Here, we used the idea of lightmatter strong coupling to improve the charge transport through polaritonic states in 2D materials.</p><p>Charge transport through the polaritonic state was first demonstrated by Ebbesen and co-workers in 2015. 21 Perylenediimide based organic material was coupled to plasmonic arrays, and the corresponding I-V characteristics was measured using a 2-terminal device configuration.</p><p>Resonance coupling of the active layer to the plasmonic mode resulted in an enhancement of conductivity by an order of magnitude. This may result in the delocalization of molecular orbitals under strong coupling conditions that boost charge transport through the polaritonic states-this concept can be called as polaritronics. 22 There are mainly three mechanisms proposed for this observed phenomenon, that include: (i) effective mass of the polaritonic state, (ii) molecular coherence and (iii) work function modification under strong coupling. 21,23,24,25,26 Both experimental and theoretical studies suggest that either one or all of the above mechanisms may operate together to boost the transport in the medium. 21 Organic electronics is based on hopping mechanism and hence molecular packing has a huge role in controlling their electrical properties. 27,28 Previous observation also points to the fact that crystal packing is a necessary criterion to achieve better electron transport. 29 In the current study, we used highly ordered WS2 multilayers with Wannier-Mott excitons having narrow absorption bandwidth. These features, along with a high absorption cross-section, help us to study charge transport behaviour under strong coupling condition. Here, we used an unconventional configuration of Fabry-Perot (FP) cavity that works in a field-effect transistor (FET) configuration (Figure 1 and 2). Few layer WS2 is effectively coupled to the  cavity mode; both optical and electrical output is measured to understand the nature of transport in the strongly coupled system.</p><p>WS2 monolayer is unique in its optical behaviour that shows the highest absorption cross-section as well as the high quantum yield of fluorescence. 30 Both monolayer and few-layer WS2 shows sharp exciton bands in the visible region (Figure 1). Monolayers are direct bandgap materials, whereas multilayers show indirect bandgap. Due to spin-orbit coupling, the valance band split into two states in K-space, resulting in two optical transitions, shown in Figure 1a, c. A-exciton show powerful absorption features compared to the B-exciton and is utilized for strong coupling experiments. 16 A few-layer WS2 flake is coupled to  mode of the cavity generating hybrid lightmatter states with Rabi splitting energy of 95 meV. Here, the cavity is fabricated using a silicon wafer (with 90 nm SiO2 coating) and a thin metal mirror (Ag; 30 nm) having poly-methyl methacrylate (PMMA) as a tunable spacer (Figure 1b). Introduction of a few-layer (~6 layer) freshly exfoliated WS2 at the anti-node position resulted in the splitting of the energy level, which generate two polaritonic states (P+ and P-). Dispersion experiment gives a clear anti-crossing with the splitting energy of 95 meV as the light-line crosses the A-exciton transition (Figure 1d). In order to understand the nature of the interaction, we have calculated the electromagnetic field distribution using the transfer matrix method (TMM). TMM studies suggest localized modes with a slight penetration of field into the silicon (Figure S1 and S2; Section 1; SI). The cavity modes are dissipative with a Q-factor of ~10, whereas, the splitting energy is approximately the same as compared to the typical FP cavity configuration reported in the literature. 17,31 Further, we explored the effect of thickness of the active layer on the coupling strength. Different cavities are fabricated using monolayer to few-layer WS2. It is reported that the A-exciton peak position moves from 613 nm to 629 nm 17 for monolayer to few-layer system due to Coulomb screening effect in a multilayer structure (Figure 2a). Tuning cavity mode to A-exciton peak position and Rabi splitting energy was measured by placing the active layer precisely at the antinode of the cavity (Figure 2b). The splitting energy varies non-linearly due to change in the effective transition dipole moment of a few-layer WS2 (Figure 2c). In-plane transition dipole moment gradually decreases and saturate for a multilayer, whereas, the in-plane field distribution still enhances the strong coupling contribution in the system. These results are exactly similar to the previous studies reported by Wang et.al, in a nanoparticle array configuration. 17 The electrical properties of the coupled system are measured in 3-terminal, bottom gate FET configuration as shown in Figure 3a. An active layer of WS2 is placed in between the SiO2 and PMMA layer to achieve ON resonance condition. n ++ dopped Si is used as the substrate for the electrical measurements (see experimental methods). SiO2 layer was thermally evaporated to achieve a specific thickness of 90 nm. Ti/Au electrodes (10/30 nm) were deposited on SiO2 to complete the 3-terminal configuration. Few-layer WS2 was exfoliated using scotch tape method and transferred on Au electrode via PDMS substrate. In the next step, Ag mirror (30 nm) was deposited on a PDMS slab followed by spin-coating of PMMA. The thickness of the PMMA is controlled by varying the rotation speed. Now, PMMA coated Ag mirror is bonded on top of the WS2 layer by physical adsorption. This completes the FP cavity in FET configuration; the active layer is now positioned at one of the anti-nodes of the  cavity. Source and drain separation is fixed at a length of L= 5 m, and width of WS2 at ~5.5 m (changes depending on the dimension of the WS2 flake transferred for each device; these parameters are taken for the mobility calculations). AFM section analysis suggested the thickness of the active layer as 6.2 nm, suggesting a 6-layer WS2 device (Figure 3b).</p><p>The above device was then taken into an inert gas glove box and annealed at 200 ̊ C to remove any organic impurities. I-V characteristics are measured in an inert atmosphere, as shown in Figure 3c (black points). ON/OFF resonance data were acquired by introducing the mirror with varying PMMA thickness on the same sample in repeated measurements. Figure 3c is a representative Id-Vd curve (gate voltage; Vg=0 V) for bare, OFF and ON resonance systems using the same WS2 flake as an active layer. Please note that annealing of bare flake is necessary before each measurement, and annealing after the bonding is not recommended. This minimizes the error in the electrical measurements and gets good consistency of the data. In all the structures, we observed a positive gating effect with a non-linear Id-Vd characteristics at different Vg (Figure 3d). It shows Schottky contact between metal/WS2 interface in the device. Mobility of the bare flake is extracted from the transfer characteristic of the device and is approximately 5 x 10 -2 cm 2 V -1 s -1 (Table S1; section 2; SI). The mobility of the bare device is low due to very high Schottky contact and also due to the presence of a dielectric layer (SiO2) creating large number of trap states. Very interestingly, introducing a mirror on top of the device and slowly moving the system from OFF to ON resonance increases the mobility to ~3 cm 2 V -1 s -1 (Figure 3e). Ion/Ioff ratio of the ON resonance cavity is increased by 2 orders of magnitude, suggesting a better performance of the device under strong coupling condition.</p><p>In the next step, mobility was measured while tuning the cavity mode position. A clear enhancement in mobility is observed while the active layer is efficiently coupled to the  cavity mode. In order to understand the strong coupling effect, the ratio of mobility is plotted against the cavity mode position (Figure 4a). Mobility of the active layer is boosted approximately 50 times at ON resonance condition, and the change in mobility follows the line-shape of the A-exciton band of WS2. The effective mass of the polaritonic state (P-) is calculated from the optical detuning experiments. The effective mass change is ~10 -5 at ON resonance condition compared to the bare structure (Figure S4a). In order to understand this large variation in mobility, we extracted the change in Schottky barrier height (SBH) using the thermionic electron transport relation,</p><p>where, Φc and Φb are SBH for cavity and bare structures. Ic and Ib are drain current for cavity and bare device at same bias voltage. mc * and mb * are the effective mass of lower polaritonic state (P-) and bare electron in few-layer WS2, respectively (section 3; SI). SBH decreased by ~45 meV under ON resonance condition (Figure 4b). Change in SBH against the cavity mode position follows the line shape of WS2, indicating a clear signature of strong coupling effect in the charge transport measurements.</p><p>Here, the effective mass calculated from the optical dispersion of the coupled system is comparable to the current density flow in the device. Light-matter strong coupling can boost the delocalization across the Wannier-Mott excitons in WS2, generating an improved flow of electrons through the polaritonic states. Few of the experimental and theoretical measurements of strongly coupled systems suggest lower electron scattering, lattice coherence, and collective nature of the states. 24,32,33,34 Another attempt was to explain the effect of strong coupling to bypass the disordered lattice that improves the charge transport. In general, all the experimental findings are suggesting a better coherence in the coupled system. In the current work, we used the effective mass of lower polaritonic state (optical output) as a handle to incorporate coherence parameter and calculated the change in SBH of the coupled system using measured current density (electrical output). Our results propose a dispersion behavior of SBH in strongly coupled system (Figure 4c). This observation is a clear indication of the hybrid nature of the polaritonic states.</p><p>In conclusion, we studied both optical and electrical properties of 2D materials under strong coupling condition. FP cavity in FET configuration is used here to understand the charge transport properties of the coupled system. Charge transport mobility is enhanced by ~50 times under ON resonance condition. Similarly, strong coupling can improve the Ion/Ioff ratio of the FET device without chemical and mechanical modification of the active layer. Lattice coherence and work function variation are the key factors that control charge transport in strongly coupled system. Our results emphasize the potential of light-matter strong coupling in material science.</p><!><p>Device preparation: WS2 flakes were mechanically exfoliated on PDMS substrate by scotch tape method from bulk crystal (purchased from HQ-Graphene, The Netherlands). Exfoliated WS2 flake of desired thickness on PDMS was selected initially by optical contrast using white light/halogen lamp illumination and later by AFM. n++ doped silicon wafer with 90 nm SiO2 and 10/30 nm Ti/Au electrodes pattern were purchased from Fraunhofer-IPMS, Germany. Next, desired flakes were transferred on the Ti/Au electrode (L= 5 m) by surface adhesion using a home-built transfer stage (dry-transfer). Samples were kept in an inert gas glove box and annealed at 200 ̊ C for 2 hours prior to electrical measurements. FP cavity in the FET configuration was prepared by sputtering (40 mA; BT300, Hind High-Vacuum) Ag mirror on a thick PDMS substrate. Poly-methyl methacrylate (PMMA;120 kDa; Sigma-Aldrich) in anisole was spin-coated (LABSPIN 6, Suss-Microtech) on Ag mirror to achieve different thicknesses. PMMA coated Ag mirror was carefully placed on the bare FET device having active layer WS2.</p><p>Electrical/Optical characterization: All electrical measurements were done in the glove box (1 ppm H2O and 1 ppm O2) using MB 150, Cascade Microtech probe station. Nikon inverted microscope (Eclipse Ti2) was used for optical measurements. Optical measurements are carried out using white light LED (transmission) and halogen lamps (reflection). Transmitted/reflected light was collected through the objective (20X, NA= 0.45) to a spectrometer (SpectraPro HRS-300, Princeton Instruments) coupled to liquid nitrogen cooled at -120 ̊C charge-coupled device</p><!><p>Id-Vd characteristics of bare and OFF resonance cavity device are shown figure S3(a) and S3(b) respectively. Electron transport at room temperature was calculated from the linear regime field effect mobility formula:</p><p>Where, 𝑊 and 𝐿 are channel width and length respectively, 𝐶 𝑜𝑥 is oxide layer capacitance can be calculated from 𝐶 𝑜𝑥 = 𝜀 0 𝜀 𝑟 𝑑 𝑖 . Here 𝜀 𝑟 = 3.9 for SiO2 and 𝑑 𝑖 = 90 𝑛𝑚.</p><!><p>Effective mass for the lower polaritonic state (P-) can be calculated using Hopfield coefficients. Where, mLP is the effective masses of lower polaritonic states, respectively. me≈1.096m0 (electron reduced mass for few-layer WS2), 3 where m0 is the free electron mass and mcav is the cavity-photon effective mass which is given by; Table S1. Calculated mobility for OFF resonance bare and cavity FET device.</p>

ChemRxiv

Design, microwave synthesis, and molecular docking studies of catalpol crotonates as potential neuroprotective agent of diabetic encephalopathy

Catalpol has gained increasing attention for its potential contributions in controlling glycolipid metabolism and diabetic complications, which makes used as a very promising scaffold for seeking new anti-diabetic drug candidates. Acylation derivatives of catalpol crotonate (CCs) were designed as drug ligands of glutathione peroxidase (GSH-Px) based on molecular docking (MD) using Surfex-Docking method. Catalpol hexacrotonate (CC-6) was synthesized using microwave assisted method and characterized by FT-IR, NMR, HPLC and HRMS. The MD results indicate that with the increasing of esterification degree of hydroxyl, the C log P of CCs increased significantly, and the calculated total scores (Total_score) of CCs are all higher than that of catalpol. It shows that CCs maybe served as potential lead compounds for neuroprotective agents. It was found that the maximum Total_score of isomers in one group CCs is often not that the molecule with minimum energy. MD calculations show that there are five hydrogen bonds formed between CC-6 and the surrounding amino acid residues. Molecular dynamics simulation results show that the binding of CC-6 with GSH-Px is stable. CC-6 was screened for SH-SY5Y cells viability by MTT (3-(4, 5-dimethylthiazolyl-2)-2, 5-diphenyltetrazolium bromide) assay, the result indicates CC-6 can effectively reverse SZT induced cells apoptosis with dosedependent manner, which can indirectly show that CC-6 is a potential neuroprotective agent.Catalpol is a kind of iridoid glycoside, which is widely found in several families of Plantaginaceae 1 , Lamiaceae and Bignoniaceae 2,3 . It is also the main active ingredient of Rehmannia glutinosa Libosch. which is one of the most commonly used herbs in traditional Chinese medicine for treatment of diabetes for more than 1000 years 4,5 . Modern pharmacological studies show that catalpol plays an important role in controlling glucose and lipid metabolism and diabetes complications 6,7 . Yang et al. 8 revealed that catalpol had significant protective effect on morphological destruction and apoptosis induced by high glucose of the SH-SY5Y cell. Catalpol (Fig. 1), C 15 H 22 O 10 , is a polar and hydrophilic molecule with the glycoside bond sensitive to acid 9 . Catalpol was reported for its variety pharmacological activities, such as anti-oxidation, anti-inflammatory, antiapoptosis, antidiabetic and other neuroprotective properties [10][11][12][13][14][15] . It also plays a key role in the nerve protection of alzheimer's and Parkinson's disease both in vitro and in vivo models [10][11][12]15 .However, catalpol is rarely reported to be used as a separate drug because of its blood-brain barrier caused by strong hydrophilicity and fast metabolism in vivo 16 . It is necessary to improve its lipophilicity and increase its utilization in vivo. It is known that esterification is a common method used for improving octanol-water partition coefficient (log P), membrane permeability, drug targeting and pharmacokinetics of hydrophilic compounds. Our preliminary studies found that catalpol propionates are better binding ligands with glutathione peroxidase (GSH-Px) than catalpol and show good anti-aging efficacy for its protection of SH-SY5Y cells 17 . There were a series of catalpol derivatives have been synthesized and reported by conventional chemical synthesis 17,18 , and these synthesis process were all time-consuming. Microwave assisted synthesis (MWAS) highlighted its great

design,_microwave_synthesis,_and_molecular_docking_studies_of_catalpol_crotonates_as_potential_neuro

<!>Results and discussion<!>Chemistry.<!>Conclusion<!>Structure design.<!>Molecular dynamics simulation.

<p>potential for synthesis of natural products and therapeutic agents with the advantage of time-saving 19 , but there is no report about MWAS applied for esterification of catalpol.</p><p>In this paper, the docking affinity (Total_score) between acylated derivatives of catalpol crotonic anhydride (CCs) and GSH-Px was calculated by using the Surfex-Docking method with the binding patterns visualized. The derivative catalpol hexacrotonate (CC-6) with the highest Total_score was synthesized using MWAS method and characterized by IR, NMR, UPLC-MS and HPLC. Molecular dynamics simulation was used to evaluated the stability of complex between CC-6 and GSP-Px. SH-SY5Y cells viability of CC-6 was screened for by MTT (3-(4, 5-dimethylthiazolyl-2)-2, 5-diphenyltetrazolium bromide) assay, and it was expected to obtain a potential neuroprotective agent of diabetic encephalopathy.</p><!><p>Molecular docking study of CCs binding with GSH-Px. The structure of catalpol was shown as Fig. 1.</p><p>The position of hydroxyl groups in molecule structure of catalpol is marked as 6, 10, 2′, 3′, 4′, and 6′, respectively.The Total_score and octanol-water partition coefficient (C log P) of the CCs calculated based on Suflux method were shown in Table 1.</p><p>It can be seen from Table 1 that the Total_score and C log P of CCs increased significantly with the increase of the number of crotonate groups increased, and they were all higher than that of catalpol. Furthermore, it also can be seen that the Total_score and C log P reached their maximum values when hydroxyl groups in catalpol were all crotonylated. These results indicate that the crotonylation of hydroxy in catalpol maybe is an effective method to improve CCs' log P and increase their druggability, and CC-6 is the best one among the 71 isomers. Meanwhile, it can be seen that the isomer with the highest Total_score in a group may not be the one with the lowest molecular structure energy, which means that the isomer with better pharmacological activity maybe not the one with more stable chemical structure.</p><p>The interaction pattern between CC-6 and GSH-Px (PDB: 2f8a) amino acid residues was shown in Fig. 2. In Fig. 2a, GSH-Px was represented by a ribbon model, CC-6 was represented by a capped stick, and the active site of the protein was represented as the green surface. In the Fig. 2b, the black ball represents the C atom, the red ball represents the O atom, the blue ball represents the N atom, the red eyelash character represents the hydrophobic interaction between CC-6 and GSH-Px, and the green dashed line is the hydrogen bond formed between the CC-6 and the amino acid residues around CC-6. It can be seen from Fig. 2b that the strong interaction between CC-6 and GSH-Px by the formation of hydrogen bondings and hydrophobic interactions with surrounding amino acid residues. Hydrophobic interactions form primarily via the carbon atoms in the ester part of CC-6 with nearby amino acid residues of Gly(A)48, Leu(A)46, Gly(A)47, Lys(B)186, Cys(B)113, Glu(B)114, Gln(B)78, Thr(B)149, Leu(A)147, Gly(A)80, Asn(B)84. The five hydrogen bonds between CC-6 and GSH-Px were formed by the interaction of oxygen atom, which in the carbonyl of ester group and in the glycosidic bond, with the nearby amino acid residues. The oxygen atom of glycosidic bond in CC-6 forms two hydrogen bonds with Asn(B)77 and Lys(B)112, respectively, the oxygen atom in ester carbonyl group at the 4′ position form two hydrogen bonds with Arg(A)98, the oxygen atom in ester carbonyl group at the 6′ position forms a hydrogen bondwith Gln(A)82.</p><p>Molecular dynamics simulation of CC-6 with GSH-Px. In order to predict the interaction stability of complexes between CC-6 and GSH-Px, the AMBER PACKAGE 20 was used to conduct further dynamic simulation and systematically predict its macroscopic physical properties. Conformation of the system dynamics equilibrium stage were taken as the reference conformation, the root-mean square deviation (RMSD) changes of the GSH-Px skeleton atoms (CA, C, N) and the non-hydrogen atoms of the CC-6 in 40 ns were calculated and shown in Fig. 3. It can be seen from Fig. 3 that the complexes of CC-6 and GSH-Px basically remained stable after 5000 ps, and the RMDS of the complex remained stable at about 3000 ps. These results indicate that the complexes of CC-6 and GSH-Px is stabl.</p><!><p>For purpose of tracking the reaction process more accurately, ESI-HRMS was used for qualitative and semi-quantitative analysis of the products during the process of esterification of catalpol 17,21 . According to the result of ESI-HRMS analysis, it was found that the hydroxyl groups in catalpol were replaced one by one. Furthermore, with the reaction time prolonged, the amount of crotonic anhydride input more, and the reaction temperature increased, it was found that the darker of the color of the reaction mixture becomes. In order to obtain the optimal conditions of the MWAS process, orthogonal experimental design was performed.</p><p>Preparation and structure confirmation of CCs. The MWAS procedure used in this work was shown as Scheme 1. In this scheme, pyridine is not only acts as a solvent, but also acts as an acid binding agent. The molar ratio of reactants, amount of catalyst, reaction temperature and reaction time are the main factors affect the MWAS process. The optimal conditions for the orthogonal experimental design were as following as the molar ratio crotonicanhydride: catalpol = 18:1, the reaction temperature 80 °C, the power of reaction microwave 900 W, and the reaction time 8 h.</p><p>It was found that the relative content of target product in MWAS synthesis process was higher than that of the conventional method 17 , and the reaction time can be shortened by two thirds. CC-6 was purified by silica www.nature.com/scientificreports/ gel column chromatography, solid phase extraction column chromatography (SPE) and preparation-HPLC. It was found that preparation-HPLC was the best method for purification of CC-6.</p><p>Characterization of CC-6. The chemical structure of CC-6 was confirmed by FT-IR, NMR, and UPLC-HRMS.</p><p>The FT-IR spectra of catalpol and CC-6 were shown in Fig. 4a,b. It was found by comparing these two IR spectra that the typical strong absorption peak of the hydroxyl group at 3386.4 cm −1 can be seen in Fig. 4 (Catalpol), however it disappeared in Fig. 4 (CC-6), while the strong absorptive peak of ester carbonyl at 1720.4 cm −1 appeared . These preliminary result show that the six hydroxyl groups of the catalpol were all esterified with crotonoic anhydride.</p><p>The 1 H NMR (400 MHz, CdCl 3 ) and 13 C NMR (100 MHz, CDCl 3 ) of CC-6 were shown as Fig. 5a,b, respectively. In Fig. 5a, δ The UPLC-HRMS of synthesized CC-6 under positive ion mode were shown as Fig. 6. The molecular formula of CC-6 is C 39 H 46 O 16 with molecular weight 770.27, and two main ESI peaks (m/z) [M + H] + 771.29, [M + Na] + 793.27 at three retention times (min) 6.58, 7.24, and 8.09, respectively, are almost the same, this indicates that the six hydroxy groups of catalpol are all be crotonylated and formed CC-6. However, the retention behavior of CC-6 indicates that it is not composed of a single compound, but a mixed components with same molecular weight. . In these assays, streptozotocin (STZ) were used to induced SH-SY5Y cell damage in vitro models, which has been widely used to induce glucose metabolism, neuronal apoptosis and tauopathy through oxidative damage. As demonstrated in Fig. 7, 0.8 mM STZ treatment induced a 29.94% ± 1.48% decrease in cell viability for SH-SY5Y, while coincubation with 5, 10, 20 μM CC-6 effectively reversed STZ-induced reduction of cell viability 19.34% ± 2.27%, 18.57% ± 1.60%. and 12.37% ± 1.95%, with the confidence level 95%, respectively. These indicate that compound CC-6 has protective effects in neurons, and it is dose dependent. Therefore, CC-6 is a potential drug related to encephalopathy.</p><!><p>CC-6 was synthesized using MWAS method with the advantage of time-saving and higher relative yield compared to that the conventional method. Molecular docking and dynamics simulations show that the complex formed by the combination of CC-6 and GSH-Px has good stability. The MTT assay results indicate that CC-6 has protective effects in neurons with dose-dependent manner, and it is a potential drug of diabetic encephalopathy.</p><!><p>It was reported that GSH-Px has neuroprotective activity 23 , so the designed compounds binding with GSH-Px were selected for predict their neuroprotective activities. According to the consistency scoring function of CCs molecular docking with GSH-Px, their neuroprotective activity was predicted. Tripos force field was utilized for energy minimization, and followed by protomol generation 24 . The protomol was created by extracting the original ligand (PDB ID: 2f8a). All crystal water and small molecular ligands in protein crystals were removed. Hydrogen atoms with essential H-bond orientation and charge were added. In Multi-Channel Surface mode, several active pockets were generated and one of the best docking active pockets was selected, and the relevant side chains were repaired. The binding mode of the CCs with GSH-Px was calculated by adopting an empirical scoring function and a patented searching engine. The threshold is set to 0.50, and the expansion coefficient is set to 1 to form a prototype molecule for docking in high-precision mode. The original ligand in the active pocket as a reference was used to carry out the docking operation between the receptor and the ligand sets. Subsequently, a series of isomers of CCs were docked at the virtual active site via Geom X method by considering every ligand 25 . The binding affnity of the ligands is predicted in terms of Total_score which is expressed as − log10 K d , where K d is binding constant.</p><!><p>In order to further verify the rationality and stability of the docking Procedure for synthesis of CCs, characterization and purification of CC-6. The synthesis process of CCs was shown as Scheme 1, in which R1, R2, R3, R4, R5, R6 =CH 3 CH=CHCO-or H-, and the number of all the possible isomers is 71. 100.0 mg (0.276 mmol) catalpol was accurately weighed and put into micro reactor, then it was dissolved with 10 mL pyridine, 766 μL (4.968 mmol) crotonicanhydride was added to the reaction mixture slowly with stirring. Pyridine was used as both solvent and acid-binding agent. The resulting mixture was stirred and reacted at 70 °C assisted with microwave radiation at microwave output frequency 2450 MHz and output power 900 W. ESI-HRMS was used to monitor the synthesis process, and the reaction was terminated when the peak of catalpol was disappeared. The reaction was completed after 8 h, and then removal of pyridine by vacuum concentration using rotary evaporator. The residue concentrated of the reaction was extracted with 20 mL × 3 CHCl 3 , and then the organic layer was washed with 20 mL × 3 of saturated NaHCO 3 solution, and dried over with MgSO 4 overnight. After filtration, the crude product of CC-6 (220 mg) was obtained. CC-6 was purified separately by silica gel column chromatography, SPE-LC and preparation-HPLC. The purity of CC-6 of the purification procedure, silica gel column chromatography eluent: CH 2 Cl 2 : CH 3 OH = 5:1 (V/V) night with a density of 105 cells/well in 100 μL medium. The cells were co-incubated with different concentrations of CC-6 (5, 10, 20 μM) and STZ (0.8 μM) for 24 h. Then, the medium was removed and added 0.5 mg/ μL MTT. After incubation at 37 °C for 4 h, 100 μL dimethyl sulfoxide (DMSO) was added to each well, and the mixture oscillate slowly for 10 min on a shaker to fully dissolve the formazan crystals, followed by the absorbance was measured at 490 nm with a SPECTRA MI3X spectrophotometer (Molecular Devices, USA).</p>

Scientific Reports - Nature

Novel sensor for sequential detection of copper and lactic acid

We report the dual-sensing properties of acyl-thiourea derivative, N-((6-methoxypyridine-2-yl)carbamothioyl)benzamide(NG1) for Cu 2+ ion and lactic acid. The novel sensor can act both as a colorimetric and fluorescence probe for sensing Cu 2+ ion by producing a yellow color and enhanced blue fluorescence in the presence of Cu 2+ ion. The sequential addition of lactic acid on the other hand caused suppression of yellow color produced and also simultaneous quenching of blue fluorescence. The application of NG1 as a selective sensor for Cu 2+ ion and lactic acid has been assessed in detail by UV visible and fluorescence spectroscopy. Further, structural modification of the probe NG1 suggests the crucial role of both pyridine and acylthiourea side chain in the binding of Cu 2+ ion and also an important part of the O-methoxy group

novel_sensor_for_sequential_detection_of_copper_and_lactic_acid

Introduction<!>Results and discussion<!>Scheme 1<!>𝐿𝑂𝐷 = 3𝜎 𝑆𝑙𝑜𝑝𝑒 𝑜𝑓 𝑐𝑎𝑙𝑖𝑏𝑟𝑎𝑡𝑖𝑜𝑛 𝑐𝑢𝑟𝑣𝑒<!>Electrostatic Potential (ESP) Mapped Electron Density Surfaces<!>Self-assembling properties of NG1:<!>Colorimetric Detection of Cu 2+<!>2.4Fluorometricassay for the detection of Cu 2+ ions and lactic acid<!>Computational Details<!>Conclusions

<p>Cu 2+ is an essential micronutrient that is utilized as a cofactor in various enzymatic process and is involved in proper functioning of plethora of biological activities. 1 However, if the concentration of Cu 2+ increases it causes deleterious effects on kidney, liver and gastrointestinal tracts and lead to diseases like Alzheimer's, Parkinson's and Wilson's disease. [2][3][4] Due to the associated toxicity of increased Cu 2+ concentration, the government agencies throughout world have set maximum permissible levels of Cu 2+ in drinking water. U.S. Environment Protection Agency (EPA) and World Health Organization (WHO) regulate maximum permissible levels of copper (II) of less than 20 μM and 31 μM in water, respectively. 5 Hence, to test the permissible concentrations of Cu 2+ , it is imperative to develop novel sensors which could detect Cu 2+ in a highly selective and sensitive manner. There is growing interest to develop novel analytical techniques for Cu 2+ detection and currently the instrumental techniques which are widely used for the detection of Cu 2+ are inductively coupled plasma mass spectrometry (ICP-MS) 6 , atomic absorption spectrometry (AAS) 7 , inductively coupled plasma atomic emission spectrometry (ICP-OES). 8 These instrumental techniques are very costly and complicated to use and require sophisticated training. Hence, optical probes that could detect Cu 2+ with colorimeter and fluorescence are in great demand since they offer simple, cost-effective methods for rapid and efficient detection of Cu 2+ concentrations. Fluorescent chemosensor, for example, oligothiophene-phenylamine based Schiff's bases, 9 ferrocenes based chemosensor, 10 N-benzoyl thioureas, 11 benzimidazole subsidiaries, 12 pyridine acyl thiourea subordinates 13, etc have been accounted for the detection of Cu 2+. Herein, we report novel acyl thiourea derivative N-((6-methoxypyridin-2yl)carbamothioyl)benzamide (1) which could efficiently detect Cu 2+ at very low concentration both by colorimetric and fluorescence methodology. Interestingly, the color and fluorescence are quenched in the presence of lactate. Hence, probe NG1 can also be used for sequential detection of Cu 2+ and lactate. Lactic acid is also very harmful at concentrations above the threshold, and many diseases like mitochondrial diseases, cerebral ischemia, and cancer are associated with high lactate levels in the blood. 14 Combined experimental tools (UV-visible, fluorescence, FTIR)</p><p>and quantum chemical calculations based on density functional theory (DFT) were used for the sequential identification of Cu 2+ ions and lactic acid by probe NG1. 15 Finally, the biocompatibility and application of NG1 for cellular detection of Cu 2+ and lactate were perceived via MTT assay and cell imaging experiments. Notably, it was found that fluorescence of cells was remarkably enhanced in the presence of Cu 2+ while it was quenched when exogenous lactic acid was added to cell culture. Hence the utility of NG1 in sequential detection of Cu 2+ and lactate can also be surmised. (Figure 1) (Bottom) NG1+Cu 2+ +Lactic acid;</p><!><p>The chemical synthesis of NG1 was carried out in one step by condensation reaction between benzyl isothiocyanate and 2-amino-6-methoxy pyridine via Scheme-1. The chemical structure of NG1 was characterized by NMR, mass spectroscopy, LCMS, and its purity were ascertained by HPLC.</p><!><p>There are reports in literature wherein acyl thiourea derivatives have been used as a probe for detection of Cu 2+ ion [16][17][18] Hence, we were motivated to assess the application of NG1 as a sensor for metal ions. Thus, to assess the sensing properties of probe NG1, experiments were performed by co-incubating NG1 with a series of divalent metal ions. Metal-ions binding study of NG1 was carried out following the method described in the experimental section using the metal ions such as Cu + , Na + , K + , Cs + , Ca 2+ , Mg 2+ , Ba 2+ , Cr 3+ , Mn 2+ , Co 2+ , Ni 2+ , Cd 2+ , Zn 2+ , Ag + , Hg 2+ , Pb 2+ , Sr 2+ , Fe 2+ , Fe 3+ and Cu 2+ ions. As can be accessed from Figure 2 there was selective detection of only Cu2+ ion by NG1 and the color of the solution in a vial containing NG1+ Cu 2+ changed from colorless to yellow.</p><p>Further, an interference assay was performed by mixing all metal ions except Cu 2+ with NG1.</p><p>Notably, there was no visible change in the color of the solution even in the presence of all metal ions. However, when Cu 2+ was added to this mixture, the colorless solution changed to yellow.</p><p>As revealed by the colorimetric response NG1 produced yellow color only in the presence of Cu 2+ and was least affected by the presence of other types of common ions (250 ppm: Na + , Cu + , Fe 3+ , Cd 2+ , Mn 2+ , Ag + , Hg 2+ , K + , Mg 2+ , Ca 2+, and Cr 3+ ions). Hence it can be inferred that NG1</p><p>shows optimal activity with negligible interference by other ions (Fig S ). The same results were also evaluated with UV−visible spectroscopy studies. The UV visible spectra of 50 ppm solution of NG1 reveal a broad peak with maxima from 290-320 nm. When 50 ppm Cu 2+ was added to this solution alone sh a peak shift from 310 nm to 339.5 nm could be assessed, There was also slight enhancement of broad peak in the range of 400-450 nm which corresponded well with the yellow color produced in the solution due to interaction between NG1 and Cu 2+ ions, NG1</p><p>selectively interacts with Cu 2+ ions and the new peak appeared in the visible region (400-450 nm) may be assigned to intramolecular charge transfer (ICT) transition (Fig. 1A and 2A). To study the interaction of NG1with Cu 2+ ions in more detail, the absorption spectra of the NG1 probe was studied in the presence of different ppm levels of Cu 2+ ions by UV−visible spectroscopy. As shown in Fig. 2A NG1 exhibited a maximum absorption at ~310 nm. On the gradual addition of Cu 2+ ions, the absorption intensity at ~319 nm decreases whereas an additional absorption appears simultaneously at ~340 nm through an isosbestic point at 320 nm.</p><p>. The intensity of this peak at ~340 nm is increased as the concentration of Cu 2+ ions (0-50 ppm)</p><p>was gradually increased and was ascertained to the formation of aggregated complex state NG1-Cu 2+ complex (Fig. 3B). The interaction is replicated in the visual color change from colorless to yellow.</p><p>To investigate the measurable response of the synthesized probe NG1 towards Cu 2+ ions, different concentrations of NG1: Cu 2+ complex were studied. The plot shows the linear relationship in the concentration range between 2.5 to 50 ppm. The limit of detection (LOD) is detected by UV−visible spectroscopy and is calculated by the following formula.</p><!><p>where the σ is the standard deviation.</p><p>LOD was calculated to be 1.5 ppm from the plot of absorbance vs Cu 2+ ions concentration. In conclusion, the convenient visual detection limit with the naked eye for the newly synthesized probe NG1 is 1.5 ppm for Cu 2+ ions, which is much lower according to the maximal permitted level of Cu 2+ ions. The short response time and high selectivity for the probe NG1 in visual inspection could be accounted for a strong affinity towards specific Cu 2+ ions. (Figure 3C)</p><p>The Job's plots were constructed to demonstrate complexation between Cu 2+ ionsand probe NG1.</p><p>The indicative of 1:1complexation between Cu 2+ ions and probe NG1. Job's plots revealed that 1:1 binding stoichiometry between probe NG1and Cu 2+ ionsand suggests for bis-coordination of Cu 2+ ionsvia nitrogen and Sulphur atom on the probe NG1 ((Figure 3D)). Moreover, the interaction and binding behavior between probe NG1 and Cu 2+ ionswas also evinced via FTIR and was presented in supporting information (Fig. S2). FTIR spectrum supported the change in characteristic peaks of probe NG1 in aromatic region (1450-1650 cm -1 ) due to interaction via nitrogen and sulfur atoms on the probe NG1 with Cu 2+ ions which proved the NG1-Cu 2+ complex formation. Thus, the present study realizes that the probe NG1 can be used for the simultaneous detection of Cu 2+ ions in real water samples as well as it could be potentially used for sensing Cu 2+ ions in mammalian cells and organisms. Once the sensitivity and selectivity of probe NG1 were ascertained by colorimetry, fluorescence assay was performed in the presence of Cu 2+ ions. Interestingly, it could be determined that NG1 also exhibited excellent fluorescence emission properties. When NG1 was excited at 353 nm the emission spectra were obtained with the maxima at 440 nm. When Cu 2+ ionswas added in increasing concentration there was a significant enhancement in the fluorescence intensity and the graph exhibited a redshift towards 470 nm. (Figure 4) In our body, copper mainly exists in the form of copper citrate. However, copper also plays an important role in regulating blood lactate levels and forms copper lactate complex inside the body. Hence. we decided to perform titration studies on the interaction of NG1 with copper nitrate and copper lactate Interestingly while Cu (II) citrate showed a steady increase in fluorescence the Cu (II) lactate salt was not able to bind NG1 and no increase in fluorescence was observed. This study indicated that Cu(II) lactate complex was more stable and hence Cu 2+ ions could not form complex withNG1 and induce a color change. (Figure 4) of lactic acid from 0 to 500 ppm at Wavelength 339.5 nm C) UV spectra of NG1: Cu 2+ complex against the varying concentration of lactic acid from 0 to 500 ppm at wavelength 400 nm Hence, motivated by these results we decided to assess the application of NG1 for sequential detection of both Cu 2+ ions and lactic ions. For sensing of lactate ions, solutions of NG1-Cu 2+ Complex (50ppm) was prepared in 70 % Methanol in water. Then solutions of lactic acid were prepared (5000 ppm stock solution) in Milli -Qx water. Then 0 to 500 ppm Lactic Acid was mixed with of complex solution in a 3 mL in the vial and the reaction mixtures were incubated for 10 minutes. The UV -Vis spectra of all these solutions were recorded after incubation and those were compared to the spectrum of Cu 2+ complex of the same concentration but without any Lactic acid to ascertain the selectivity. The experiment revealed that the NG1-Cu 2+ Complex is selectively interacting with Lactic Acid. Then UV -Vis titration with lactic acid was carried out following a similar procedure as described above for UV -Vis titration of NG1 with Cu 2+ ions, except instead of Cu 2+ ions, Lactic Acid was added into the solution of the Cu 2+ complex.</p><p>(Figure 5)</p><p>Figure 4B and 4C how that decrease in absorbance on increasing concentration of lactic acid in the 50 ppm NG1: Cu 2+ yellow complex from 0 to 500 ppm, which shows the color changes of yellow complex become colorless at higher concentration of lactic acid.</p><p>Further to understand the role of acyl thiourea conjugate in sensing two other analogs structural probes of NG1, NG2 and NG3 were synthesized (Scheme 1, bring old scheme which you draw).</p><p>Interestingly, both probe NG1 and NG2 show selective sensing for Cu 2+ ions, However, probe NG3 showed very little sensitivity for Cu 2+ ions. The order of sensitivity of NG1 for the Further, to perceive a greater insight into the complex formation by NG1 and its structural analog NG2 and NG3 with Cu 2+ and lactic acid we resorted to validate theoretically the experimental results by quantum chemical calculations based on density functional theory (DFT). We have conducted a comparison study on these three probes (NG1, NG2, and NG3) and their complexes with the Cu 2+ ion and Lactic acid. DFT calculations are performed at B3LYP/6-311++G(d,p) level of theory for these probes and their complexes in the gaseous phase. The vibrational frequencies were also calculated for all the studied structures, where the optimization converged successfully to the shallow local minima on the potential energy surface, which was confirmed by the absence of negative/imaginary vibrational frequencies.Fig. 10shows that the ground state optimized geometry of all three probes and their association with Cu 2+ as well as lactic acid. The binding energy (BE) of these complexes are elucidated as:</p><p>This journal is © The Royal Society of Chemistry</p><p>Where 𝐸 𝑐𝑜𝑚𝑝𝑙𝑒𝑥 contains represents the ground state energy of complex and Ʃ𝐸 𝑖𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙 contains indicates the sum of the ground state energies of all the molecules in a particular complex. The BE of all the optimized structures is presented in curly braces in Table 1.</p><p>Table 1. Optimized energy (in Hartree) and binding energy (in kcal/mol) of the Complex NG1 (NG2 and NG3 in ESI)</p><p>The binding energies in kcal/mol are provided in curly braces.</p><p>Initially, the pristine robes i.e. NG1, NG2, and NG3 were optimized using DFT calculations.</p><p>Based on energetic stability, NG1 was found to the most stable ligand. To make a better comparison, along with NG1 other two probes (NG2 and NG3) were also interacted with Cu 2+ and lactic acid (see Fig. 10). There were at least two different possibilities interaction of Cu 2+</p><p>with probes, one with the sulfur atom (represented as probe+Cu( 1))and another one is near the nitrogen atom (denoted as probe +Cu ( 2)) as can be seen in Fig. 10. After the interaction and geometry relaxation process, it is noteworthy that in all the six complexes (probes + Cu(1), Cu</p><p>(2)), Cu 2+ tend to strongly bind with Sulphur atom of all the studied probes (NG1, NG2, and NG3 with Cu (1)) as shown in Fig. 10. This trend is also confirmed from the binding energies which are higher for all the probes + Cu (1)complexes than probes + Cu (2). The BE calculations for all complexes(with Cu) indicate that the complex NG1+ Cu(1)is the most stable complex. Nevertheless, the BE of NG1+ Cu(1) is -25.07 kcal/mol which is slightly higher in comparison to NG2+Cu(1) complex. This small energy difference is considerable here because in experimental observationall the three probes give yellow color after incorporating the Cu but the only key difference is between these three probes were their concentration. For probe NG1, presence of only 1.5 ppm concentration of Cu 2+ causes the occurrence of yellow colour but in case of other probes (NG2 and NG3) yellow colour appears only after the incorporation of high Cu concentration. So,for further studies we have selected NG1+Cu (1) complex on the basis of BE to validate our experimental findings.</p><p>Here, the Cu 2+ strongly interacts with the oxygen and sulfur atoms having a bond length of 1.95Å and 2.18 Å (see Fig. 10). In complexes 2 (all the three probes), Cu 2+ is weakly coordinated with nitrogen and oxygen atoms as bond length increased on the interaction (see Fig. 10). We have also optimized a dimer like structure of NG1 with single Cu 2+ atom (Fig. As experimentally observed, as a result of the interaction of Cu 2+ with NG1a yellow colored complex is obtained but when lactic acid is incorporated with this complex this yellow color quenches. To understand this behavior, we have introduced a molecule of lactic acid with this stable complex of NG1+Cu (1) and it can be seen in Fig. 10 strongly binding with the sulfur and oxygen atom of NG1 through 2.18 Å and 1.95 Å respectively. However, upon complex formation with lactic acid both the bond strengths get weekend and calculated to be 2.26 Å and 1.99 Å respectively. Also, it can be noticed from Fig.</p><!><p>Molecular electrostatic potential (MEP) (also called ESP) of a molecular association is correlated with the corresponding partial charges, dipole moment and chemical reactive sites. The mapping of any molecular system is the way to visualize the relative polarity of the studied molecules.</p><p>The expression for the ESP at any point r in the space near the molecule is given by</p><p>where, ZA is the charge on the nucleus A studied at RA and ρ(r') the electron density. For probe NG1 and complex NG1with Cu and lactic acid, the electron density iso-surface on which the In a three-necked round bottom flask, fitted with a dropping funnel 50ml of dry acetone was filled. 5.0g (27 mM) of 2-Amino 6-Methoxy pyridine was placed followed by dropwise addition of dry acetone under N2 atmosphere during the constant stirring of the reaction mixture. Next, 7.2 g (44mM) of benzoyl isothiocyanate was added and the reaction mixture was then allowed to stir for another 2h at room temperature.The progress of the reaction was monitored by analytical TLC by using the ethyl acetate-hexane (3:7) solvent mixture. After the completion of the reaction, the reaction mixture was then poured carefully with stirring into 500 ml of cold water and the resulting yellow precipitate of (N-((6-methoxypyridin-2-yl)carbamothioyl)benzamide) is separated by suction filtration followed by washing of precipitate with water(3x100 ml). The</p><!><p>The self-assembly property of NG1 assessed by optical microscopy, phase contrast, green and red filter at the 10-uMconcentration, further, to understand the mechanisms of formation of selfassembly at supramolecular state, self-assembly was characterized by Scanning electron microscopy and Atomic force microscopy, UV, FT-IR, NMR. All the study shows that the assembly is formed by hydrogen bonding, π-π stacking, electrostatic interaction, and hydrophobic interaction.</p><!><p>For colorimetric measurements, stock solutions of 500 ppm NG1 and 500 ppmCu(NO3)2. mixing 250 ppm probe NG1 with 100 ppm metals (100 ppm: Cu 2+ , Fe 3+ , Cd 2+ , Mn 2+ , Ag + ,Hg 2+ , Ca 2+ , Cr 3+ , K + , Mg 2+ and Na + ). The LOD of probe NG1was detected by making a complex of 250 ppm:250 ppm NG1: Cu 2++ complex. The stoichiometric ration of NG1: Cu 2+ complex was determined by adding varying concentrations of Cu 2+ (0.5-50 ppm) to 50 ppm NG1.</p><!><p>For fluorometric measurements, 250 ppm of NG1 stock was used and varying concentration of Cu 2+ (0.001-100 ppm) was added. For lactic acid-mediated quenching 250 ppm of lactic acid was added to 250ppm:100 ppm NG1:Cu 2+ complex.</p><!><p>The prediction of certain important molecular properties of the ligandNG1, NG2 and NG3 and their possible complexes with Cu 2+ and lactic acid are calculated using density functional theory (DFT) calculations. All computational calculations were carried out using the Gaussian 16</p><p>programming package in the presented works 15 . GaussView 6 software was used to generate the input molecular structure and to analyze the output of all calculated results 15 . Geometric optimization calculations were performed in the gas phase using DFT by employing the internally stored 6-311++G(d,p) 23,24 and the hybrid density functional theory that incorporates B3LYP correlational functional 25, 26 method. In general, B3LYP is a prominent method and preferable to the HF and MP2 method as it includes Becke's three-parameter exchange with Lee, Yang, and Parr's correlational functional as well as an HF exchange term. DFT calculations using the B3LYP protocol has been provided a meaningful and nice correlation with experimental results [27][28][29][30] . Time-dependent density functional theory (TD-DFT) haswasalso room temperature. The fixed coverslips were visualized for dye uptake using Leica SP8 confocal microscope. The cells were excited using 405 nm with very low laser power (~ 1% laser power) and the images were acquired using a broad width emission spectrum to capture the maximum of the emitted photons.50-60 cells were imaged in every condition using the same laser power and imaging conditions.</p><p>Data analysis: Image analysis and quantification were performed using Image-J software (nih.gov). 50 cells were randomly selected to perform the quantification. The value of the total fluorescence intensity of each cell wasobtained from Image-J. The images were quantified by subtracting thebackground and measuring the cellular intensities by measuring the areas of the cells and the total cellular intensity. The normalized mean value and standard deviation for each were calculated and plotted in the graph using Prism 7 software and the statistical significance was calculated using a one-way Anova test.</p><!><p>The sensitivity and specificity of NG1 towards Cu 2+ ions are tested by various spectroscopic and biophysical assay. Finally, the supramolecular assembly of the probe was studied and it In addition to the experimental study, DFT calculations were also carried out various possible probes NG1, NG2, andNG3 withCu 2+ ions atom and lactic acid to support our experimentally observed results. The comparison of experimental and theoretical results reveals a good agreement between experimental findings and calculated results and shows the same pattern of shifting in absorption towards higher wavelength.</p>

ChemRxiv

Anti-inflammatory and antiproliferative activities of trifolirhizin, a flavonoid from Sophora flavescens roots

Trifolirhizin, a pterocarpan flavonoid, was isolated from the roots of Sophora flavescens, and its chemical structure was confirmed by1H and 13C NMR and MS spectra. Its anti-inflammatory activity was examined in lipopolysaccharide (LPS)-stimulated mouse J774A.1 macrophages. Trifolirhizin not only dose-dependently inhibited LPS-induced expression of pro-inflammatory cytokines including tumor necrosis factor-\xce\xb1 (TNF-\xce\xb1) and interleukin-6 (IL-6), but also inhibited lipopolysaccharide (LPS)-induced expression of cyclooxygenase-2 (COX-2). In addition, trifolirhizin showed in vitro inhibitory effects on the growth of human A2780 ovarian and H23 lung cancer cells. These results suggest that trifolirhizin possesses potential anti-inflammatory and anti-cancer activities.

anti-inflammatory_and_antiproliferative_activities_of_trifolirhizin,_a_flavonoid_from_sophora_flaves

INTRODUCTION<!>General procedures and reagents<!>Plant material<!>Isolation and separation of Trifolirhizin<!>RNA isolation and real-time quantitative PCR<!>Western blot analysis<!>Anti-proliferative activity estimation<!>DPPH\xe2\x80\xa2 scavenging capacity<!>Statistical analysis<!>Isolation and identification of Trifolirhizin<!>Effects of trifolirhizin on TNF-\xce\xb1 and IL-6 mRNA expression in LPS-stimulated mouse J774A.1 macrophage cells<!>Effects of trifolirhizin on COX-2 expression in LPS-stimulated mouse J774A.1 macrophages<!>Anti-proliferative activities of trifolirhizin<!>DPPH\xe2\x80\xa2 scavenging capacity of Trifolirhizin<!>

<p>Inflammation plays an important role in a wide variety of chronic human diseases such as cardiovascular diseases and cancer. It has been demonstrated that pro-inflammatory cytokines, cyclooxygenase-2 (COX-2), and free radical species interact in a complex manner in an inflammation environment (1). For example, tumor necrosis factor-α (TNF-α) has been shown to be one of the major cytokines that mediates many crucial events for the initiation of both acute and chronic inflammation through regulating production of some other cytokines, up-regulation of adhesion molecule expression, and the activation of leukocyte-specific chemotactic cytokines (2). Interleukin-6 (IL-6) is another pro-inflammatory cytokine that promotes inflammatory events through the activation and proliferation of lymphocytes, differentiation of B cells, leukocyte recruitment and the induction of the acute-phase protein response in the liver (3). Pro-inflammatory cytokines such as TNF-α and IL-6 are also interlinked with the production of some small inflammatory mediators such as NO and prostaglandin (PGE2), and thus contribute inflammatory response. In addition, COX-2 has been identified as an important link in the inflammation cascade. Unlike COX-1, COX-2 is selectively induced by pro-inflammatory factors at the site of inflammation, and is responsible for the generation of inflammatory prostaglandins which results in inflammation response and production of pain (4). Inhibition of the expression and production of these powerful mediators by anti-inflammatory components might represent a possible preventive or therapeutic target, and may be used to develop anti-inflammatory nutraceuticals for health promotion and disease prevention.</p><p>The roots of S. flavescens (Leguminosae) have been traditionally used in East Asian countries as herb medicine and functional food ingredient for thousands of years because of its potential health beneficial properties such as improving mental heath, anti-inflammatory, antiashmatic, antithelmintic, free radical scavenging, and antimicrobial activities (5–9). Previous studies have isolated quinolizidine alkaloids, flavonoids, and triterpenoids from the roots of S. flavescens (5, 8, 10). In 2000, flavonoids isolated from S. flavescens showed antiproliferative activities against human myeloid leukemia HL-60 and human hepatocarcinoma HepG2 cells and induced apoptosis in both cell lines (10). Later in 2002, a prenylated flavonoid from this herb was able to down-regulate COX-2 in LPS-treated RAW 264.7 cells and exhibited in vivo anti-inflammatory effect (7). As part of our continuous effect to develop novel nutraceuticals for functional food utilization, this study was conducted to explore the possibility of discover additional natural anti-inflammatory flavonoids from the roots of S. flavescens. The anti-inflammatory activities were examined and estimated as the inhibitory capacity on LPS induced expression of the proinflammaroty cytokines TNF-α and IL-6, and COX-2 in macrophages. The pure flavonoid compound was also evaluated for its antiproliferative activity against A2780 ovarian and H23 lung cancer cells, as well as its scavenging capacity against the stable 2,2-diphenyl-1-picrylhydrazyl DPPH radical.</p><!><p>1H and 13C NMR spectra were obtained on a Bruker-AMX 500 instrument using DMSO-d6 as a solvent. Electrospray ionization (ESI) mass spectra were acquired in the positive ion mode on a LCQ DECA XP instrument (Thermo Finnigan, San Jose, CA, USA) equipped with an ion trap mass analyzer. Column chromatography was carried out on silica gel (200–300 mesh, Fisherscitific, US). Mouse J774A.1 macrophage cell line was obtained from ATCC. RNAqueous total RNA isolation kit was purchased from Ambion (Austin, TX). High-capacity cDNA archive kit and gene expression kit were obtained from Applied Biosystems (Foster City, CA). Bio-Rad protein assay reagent was purchased from Bio-Rad Laboratories (Hercules, CA). Western Lightning Chemiluminescence Reagent Plus was from Perkin-Elmer Life Sciences (Boston, MA).</p><p>2,2-Diphenyl-1-picrylhydrazyl radical (DPPH•) and lipopolysaccharide (LPS) was purchased from Sigma-Aldrich (St. Louis, MO). All other chemicals and solvents were of the highest commercial grade and used without further purification.</p><!><p>Roots of Sophora flavescens was collected from Shanxi Province, China, in October 2006, and authenticated by Dr. Zhihong Cheng.</p><!><p>Air-dried roots of S. flavescens were ground, and refluxed and extracted three times for 4 h with methanol using a dried material/solvent ratio of 1:10 (w/v). The supernatant was collected by filtration, and the solvent was evaporated under reduced pressure to yield a brown solid residue. The residue was subjected to a silica gel column chromatography (CC) eluted with a mixture of chloroform-methanol of increasing polarity to afford three fractions. Fraction III, eluted by a mixture of chloroform-methanol (5:1, v/v), was further separated over silica gel CC eluted with chloroform-methanol (10:1, v/v), followed by recrystalliation in methanol to obtain the pure flavonoid compound, which was identified as trifolirhizin.</p><!><p>Mouse J774A.1 macrophages were pretreated with trifolirhizin (10 or 25 μM) for 2 h, then treated with lipopolysaccharide (LPS) at a final concentration of 0.5 μg/mL for 24 h. Total cellular RNA was isolated using the Ambion RNAqueous kit. Five μg of total RNA was used for first-strand cDNA synthesis using a High-Capacity cDNA Archive Kit. The mRNA levels of TNF-α and IL-6 were quantified using the specific gene expression assay kits for mouse TNF-α and IL-6 on iQ5 Multicolor Real-Time PCR Detection System. The mRNA values for each gene were normalized to internal control, β-actin mRNA. The ratio of normalized mean value for each treatment group to vehicle control group (DMSO) was calculated (11).</p><!><p>Mouse J774A.1 macrophage cells were pretreated with trifolirhizin (100 or 200 μM) for 2 h, then treated with LPS (0.5 μg/mL) for 24 h. Total cell lysates were prepared as previously described (12). The protein concentration was determined using the Bio-Rad Protein Assay reagent. The total cellular proteins (10 μg) were resolved on 10% Bis-Tris gels and transferred to Nitrocellulose membranes. Immunoblots were blocked overnight at 4 °C with 5% non-fat milk in Tris-buffered saline (TBS) and then incubated with antibodies to COX-2, or β-actin. Immunoreactive bands were detected using horseradish peroxidase-conjugated secondary antibody and the Western Lightning Chemiluminescence Reagent Plus. The density of the immunoblot bands was analyzed using Image J computer software (NIH).</p><!><p>The A2780 ovarian cancer or H23 lung cancer cells were plated in 96-well plates with a density of 1 × 104/well. The medium was replaced after 24 h. Following this incubation, trifolirhizin, dissolved in DMSO, was added to the wells in an appropriate series of concentrations. Ten microliters of MTT solution was added to each well. After 24 h incubation at 37 °C in a humidified 5% CO2 atmosphere, the absorbance at 570 nm was recorded using an ELISA plate reader. The control refers to incubations in the presence of vehicle only (DMSO, 0.5%) and was considered as 100% viable cells.</p><!><p>The DPPH• scavenging capacity of trifolirhizin was evaluated using the high throughput assay described previously (13). Briefly, the assay was performed using a Victor3 multilabel plate reader (PerkinElmer, Turku, Finland) and 96-well plates. The reaction mixture contained 100 μL of 0.2 mM DPPH• in ethanol and 100 μL of standards, control, blank, or trifolirhizin. The absorbance of each reaction mixture at 515 nm was measured every minute for 40 min. The level of DPPH• scavenged was calculated as [(A0 − A1/A0)] × 100 (where A0 was the absorbance of the reagent blank, and A1 was the absorbance with trifolirhizin). All the measurements were conducted in triplicate.</p><!><p>All values are expressed as the mean ± SD of three independent determinations. Student's t-test was employed to analyze the differences between sets of data. Statistics were performed using GraphPad Pro (GraphPad, San Diego, CA). Statistical significance was declared at P < 0.05.</p><!><p>The CHCl3-methanol (5:1, v/v) fraction of the methanol extract of S. flavescens roots was further separated by silica gel column chromatography to obtain a pure flavonoid compound. After re-crystallization in methanol, needle crystals with light yellowish were collected. The molecular formula of the compound was established as C22H22O16 by ESI MS (m/z 447 [M + H]+ and 469 [M + Na]+). The structure of the pure compound was identified as trifolirhizin based on the spectroscopic analysis including 1H– and 13C–NMR spectroscopy, and electrospray ionization mass spectrometry (ESI MS). The chemical structure of trifolirhizin is shown in Figure 1. Its 1H and 13C NMR data, listed in Table 1, agreed well with the data reported previously (5, 14).</p><p>Trifolirhizin is a pterocarpan which belongs to a special group of isoflavonoids possessing two contiguous benzofuran and benzopyran rings. It was first isolated from Trifolium pretense L. in 1960 (15) and identified in S. flavescens by Yagi and co-workers in 1989 (5). Trifolirhizin has not been evaluated for its potential anti-inflammatory activity, anti-proliferative property against A2780 ovarian and H23 lung cancer cells, or free radical scavenging capacity, although it has been found to possess antifungal and anti tumor properties (5, 14).</p><p>It is well known that deregulated inflammation leads to the massive production of pro-inflammatory cytokines such as TNF-α, IL-1 and IL-6 by macrophages. Reduced overproduction of these inflammatory cytokines might be a potential target for development of preventive and therapeutic approaches against inflammation-related health problems. In this study, trifolirhizin was for the first time examined for its potential inhibitory effects on LPS-stimulated expression of pro-inflammatory cytokines (IL-6 and TNF-α) and COX-2 in macrophages, along with its antiproliferative activity and the potential to directly react with free radicals.</p><!><p>In the present study, the LPS-stimulated mRNA expression of representative pro-inflammatory molecules including TNF-α and IL-6 in mouse J774A.1 macrophage cells were examined by real-time RT-PCR. As shown in Figures 2A and 2B, trifolirhizin significantly inhibited LPS-induced increase in mRNA expression of TNF-α and IL-6 in a dose-dependent manner. At a concentration of 25 μM, trifolirhizin completely inhibited LPS-induced increase of TNF-α mRNA level.</p><p>The effects of trifolirhizin on the production of TNF-α and Il-6 were also examined by the ELSIA method. LPS-induced cells treated with trifolirhizin showed a significant decrease in TNF-α production in a dose-dependent matter (Figure 2C). This was in agreement with the observation that trifolirhizin dose-dependently suppressed LPS-induced mRNA expression of TNF-α. However, no significant inhibition was observed for IL-6 production under the same experimental conditions (data not shown).</p><p>In 2003, other herbal constituents including apigenin and ginsenoside were found to inhibit the same proinflammatory metabolites (16). It was observed that apigenin at a high dose of 37 μM significantly inhibited only LPS-induced levels of IL-6, without any significant effect on TNF-α concentration (16). It was also reported that ginsenoside Rb1 at a dose of 84 μM completely inhibited both TNF-α and IL-6 induction (16). These treatment concentrations were much higher than the trifolirhizin dose (25 μM) that effected the same observation in TNF-α mRNA levels in the present study. It does appear that trifolirhizin may be a better inhibitor of TNF-α compared with ginsenoside and apigenin. However, it needs to be pointed out that a strict quantitative comparison may lead to inaccurate conclusions due to the different cells and assays used in the different studies.</p><p>Xagorari and others (17) also observed that luteolin and quercetin, two natural flavonoids, possessed strong inhibitory activity against these proinflammatory cytokines. The mechanism by which trifolirhizin inhibits TNF-α expression at both mRNA and protein levels and IL-6 at the mRNA level is unclear at this point, but may be related to its ability to interfere with the transcription factor NF-κB (nuclear factor kappa B). NF-κB is responsible for the expression of these proinflammatory cytokines, and a down-regulation of its activity is a plausible explanation for the observed reduction in IL-6 and TNF-α levels. Further research is required to elucidate the exact mechanism involved in the inhibitory activity of trifolirhizin.</p><!><p>Trifolirhizin was examined for its effect on the expression of COX-2 protein in mouse J774A.1 macrophage cells stimulated with LPS. Trifolirhizin dose-dependently inhibited the LPS-stimulated COX-2 protein expression (Figure 3). At 0.1 and 0.2 mM concentrations, trifolirhizin suppressed LPS-induced COX-2 protein production by 14 and 28%, respectively, based on the density ratio of COX-2 versus internal control β-actin (Figure 3).</p><p>Expression of COX-2 may be induced by proinflammatory cytokines, stress, and growth factors. COX-2 is one of the three cyclooxygenase isozymes responsible for the production of prostaglandins. Prostaglandins are the precursors of series-2 prostanoids which contribute significantly to the inflammatory response. Selective inhibition of COX-2 is a preferred way for controlling inflammation because of the increased risk of peptic ulcers, heart attacks, and thrombosis associated with the inhibition of the other two COX isozymes, COX-1 and COX-3. In addition to its well-established role in inflammation, COX-2 has also been implicated in human carcinogenesis. Inhibition of COX-2 may have the dual effect of controlling both inflammation and cancer, and is the mode of operation of coxibs, a class of NSAIDs. The transcription factor NF-κB is responsible for regulating the expression of COX-2. In the present study, the reduction of the LPS-induced COX-2 expression in macrophages by trifolirhizin may also be mediated through NF- κB, similar to that for the other proinflammatory cytokines studied.</p><!><p>A growing body of evidence suggests a direct link between inflammation and cancer (1, 18). Various steps in tumorigenesis such as cellular transformation, promotion, proliferation, and metastasis have been found to be influenced by chronic inflammation (18). TNF-α has been associated with induction of reactive oxygen in the form of NO, which directly oxidizes DNA and lead to cancer development (19). It is well accepted that anti-inflammatory agents suppressing NF-κB or NF-κB-regulated products including TNF-α, IL-6 and COX-2 may also have a potential in the prevention and treatment of cancer (1, 18).</p><p>In the present study, trifolirhizin dose-dependently suppressed the expression of LPS-stimulated TNF-α, IL-6, and COX-2 in mouse macrophages (Figures 2 and 3), which are widely accepted in vitro models for investigating anti-inflammatory agents. It is interesting to further examine its possible anti-proliferative activities since trifolirhizin suppressed human myeloid leukemia HL-60 and hepatocarcinoma HepG2 cells in a previous study (10). Trifolirhizin was tested against human A2780 ovarian and H23 lung cancer cells using the MTT assay since these two cell lines are commonly used in our laboratory for screening and development of novel anti-proliferative agents. Trifolirhizin dose-dependently suppressed proliferation of both A2780 ovarian and H23 lung cancer cells (Figure 4). When exposed to trifolirhizin with concentrations less than 50 μM, no anti-proliferative activity was observed in either of the two cell lines. For A2780 ovarian cancer cells, significant anti-proliferative (50% growth inhibition) was achieved with concentration up to 100 μM. However, significant anti-proliferative effect was observed only with a trifolirhizin concentration of 250 μM for H23 lung cancer cells. These data demonstrated the potential anti-proliferative activity against caner cells and its possible selectivity.</p><p>Taking into account both anti-proliferative and anti-inflammatory data from this study, trifolirhizin might decrease tumorigenesis through inhibition of the inflammatory mediators. This is consistent with results from other studies that tested anticancer activity of natural anti-inflammatory flavonoids (20–23). Many different mechanisms have been proposed to explain these actions, but they are not mutually exclusive of each other. For example, although the results from this study seem to suggest that the antiproliferative activity of trifolirhizin may be related to its ability to inhibit the expression of some proinflammatory cytokines, another study by Aratanechemuge and others (14) found the antitumor effect of trifolirhizin might be explained by its ability to induce apoptosis in human promyelotic leukemia HL-60 cells. The actual mechanism may very well be a combination of several pathways. Considering in this study that both TNF-α and IL-6 mRNA expressions were suppressed, it is consistent to suggest that trifolirhizin acts at the transcription level.</p><!><p>The capacity of trifolirhizin to directly react with and quench free radicals was evaluated and compared with that of the crud methanol extract of the S. flavescens. The crud extract showed DPPH• scavenging capacity in a dose-dependent manner (Figure 5A), but no significant DPPH• scavenging capacity was observed for pure trifolirhizin at concentrations up to 12 mM (Figure 5B).</p><p>Growing evidence suggests that free radicals and free radical mediated oxidative stress are closely correlated with the development of inflammation by increasing activation of transcription factors important in regulation of pro-inflammatory cytokines (24, 25). As a result, expression of these genes stimulates the secretion of pro-inflammatory cytokines. In the present study, no significant free radical scavenging activity was observed for trifolirhizin with concentrations of up to 12 mM. This observation was in consistent with the previous studies showing that trifolirhizin might not act as an active scavenger of free radicals (6, 8, 9). This may be partially explained by the lack of phenolic hydroxyl groups in trifolirhizin molecule (Figure 1). This result was also supported by another study showing that trifolirhizin did not have detectable ONOO− or DPPH• scavenging activities (9). Taken together these previous observations with our present data, it may be concluded that the anti-inflammatory and antiproliferative effects of trifolirhizin were not mediated through its direct interaction with free radicals.</p><p>In summary, trifolirhizin from S. flavescens roots was not only able to inhibit LPS-induced TNF-α, IL-6 and COX-2 expression in macrophages, but also inhibited cancer cell growth. These findings clearly indicate that trifolirhizin possesses anti-inflammatory and anti-cancer activities and may have application in the prevention and treatment of inflammation and cancer.</p><!><p>Chemical structure of trifolirhizin.</p><p>The effects of trifolirhizin on TNF-α (A), and IL-6 (B) mRNA levels in mouse J774A.1 macrophage cells. Cells (1 × 106/mL) were incubated with either vehicle, LPS (0.5 μg/mL), or LPS plus indicated concentrations of trifolirhizin. The mRNA levels of TNF-α and IL-6 in the culture medium were determined as described in Materials and Methods. Each column represents the mean ± SD of three independent experiments. *P<0.05 indicates a significant difference from the LPS treated control group.</p><p>The effect of trifolirhizin on COX-2 expression in mouse J774A.1 macrophage cells. Cells (1 × 106/mL) were incubated with either vehicle, LPS (0.5 μg/mL), or LPS plus indicated concentrations of trifolirhizin. COX-2 expression was determined by Western blot analysis as described in Materials and Methods.</p><p>Dose effects of trifolirhizin on human cancer cell growth (A) A237 ovarian cancer cell and (B) H27 lung cancer cell. Cells were exposed to serial dilutions of trifolirhizin for 24 h. Each column represents the mean ± SD of three independent experiments.</p><p>Dose and time effects of antioxidants-DPPH• reactions. The final concentration of DPPH• was 100 μM. (A) Reactions with methanol extract of Sophora flavescens; 0.25, 0.20, 0.17, 0.14, 0.11, and 0 represent the concentrations (mg/mL) of the extract in the initial reaction mixture, respectively. (B) Reactions with trifolirhizin; 1.00, 0.67, 0.50, 0.40, 0.33, 0.25, and 0 represent the concentrations of trifolirhizin in the initial reaction in mM.</p><p>NMR spectra data of trifolirhizin (400 MHz, DMSO-d6)</p><p>Data can be exchanged in the column. Values in parentheses are multiplicity and coupling constants (Hz).</p>

PubMed Author Manuscript

Bio-assisted preparation of efficiently architectured nanostructures of γ-Fe2O3 as a molecular recognition platform for simultaneous detection of biomarkers

Magnetic nanoparticles of iron oxide (γ-Fe 2 o 3 ) have been prepared using bio-assisted method and their application in the field of biosensors is demonstrated. Particularly in this work, different nanostructures of γ-Fe 2 o 3 namely nanospheres (NS), nanograsses (NG) and nanowires (NW) are prepared using a bio-surfactant namely Furostanol Saponin (FS) present in Fenugreek seeds extract through co-precipitation method by following "green" route. Three distinct morphologies of iron oxide nanostructures possessing the same crystal structure, magnetic properties, and varied size distribution are prepared and characterized. The resultant materials are analyzed using field emission scanning electron microscopy, transmission electron microscopy, powder X-ray diffraction, X-ray photoelectron spectroscopy, vibrating sample magnetometer and Fourier transform infrared spectroscopy. Moreover, the effect of reaction time and concentration of FS on the resultant morphologies of γ-Fe 2 o 3 nanostructures are systematically investigated. Among different shapes, NWs and NSs of γ-Fe 2 o 3 are found to exhibit better sensing behaviour for both the individual and simultaneous electrochemical detection of most popular biomarkers namely dopamine (DA) and uric acid (UA). Electrochemical studies reveal that γ-Fe 2 o 3 NWs showed better sensing characteristics than γ-Fe 2 o 3 NSs and NGs in terms of distinguishable voltammetric signals for DA and UA with enhanced oxidation current values. Differential pulse voltammetric studies exhibit linear dependence on DA and UA concentrations in the range of 0.15-75 µM and 5 μM -0.15 mM respectively. The detection limit values for DA and UA are determined to be 150 nM and 5 µM. In addition γ-Fe 2 o 3 NWs modified electrode showed higher sensitivity, reduced overpotential along with good selectivity towards the determination of DA and UA even in the presence of other common interferents. Thus the proposed biosensor electrode is very easy to fabricate, eco-friendly, cheaper and possesses higher surface area suggesting the unique structural patterns of γ-Fe 2 o 3 nanostructures to be a promising candidate for electrochemical bio-sensing and biomedical applications.Magnetic nanoparticles attract a great deal of interest among researchers across the globe owing to their unique magnetic properties such as superparamagnetic characteristics, low Curie temperature, negligible coercivity and high magnetic susceptibility that lead to applications in various fields 1,2 . Particularly, different forms of iron oxides such as magnetite (Fe 3 O 4 ), maghemite (γ-Fe 2 O 3 ), hematite (α-Fe 2 O 3 ) and their corresponding nanoparticles are employed for many biological applications since they possess good biocompatibility. Among them, ferri-magnetic material namely, γ-Fe 2 O 3 is very important for technological applications. Nanoscale γ-Fe 2 O 3 has

bio-assisted_preparation_of_efficiently_architectured_nanostructures_of_γ-fe2o3_as_a_molecular_recog

<!>Results and discussion<!>Structural characteristics.<!>optimization of pH. It is found that pH values affect the electrochemical signals of DA and UA at γ-Fe 2 O 3<!>Experimental section<!>Synthesis of iron oxide nanostructures.<!>Characterization of iron oxide nanostructures. Formation of different nanostructures of synthesized<!>Conclusions

<p>www.nature.com/scientificreports/ been used in diverse areas including high-density magnetic recording 3 , ferro-fluids 4 , drug delivery 5 , magnetic resonance imaging 6 , biosensors 7 , bio-probes 8 , hyperthermia treatment of cancer 9 , spintronics 10 , magneto-optics 11 , chemical catalysis 12 and chemical sensing 13 etc. Iron oxide nanoparticles also possess tunable magnetic properties depending upon their size 14,15 , shape [16][17][18] and crystalline phase 19 .</p><p>Recent report suggests that on controlling the diameter and aspect ratio, nanowires (NWs) of γ-Fe 2 O 3 were found to have tunable magnetic and optoelectronic properties 20 . For example, high-aspect ratio γ-Fe 2 O 3 NWs found to possess a larger magnetic coercivity than the corresponding nanoparticles of same volume due to a shape-induced energy barrier which hinders the thermal rotation of magnetic moments 21 . Besides the shape, their crystallinity also has a profound effect on their magnetic and catalytic properties. It is reported that the rotation moment scales up against size of the energy barrier and the effect is directly correlated with the increasing crystallite volume 22 . For a given size of γ-Fe 2 O 3 NWs, single crystal NW has the largest possible crystallite volume through which the energy barrier against thermal rotation moment is maximized. Therefore, it is highly desirable to synthesize single crystal γ-Fe 2 O 3 NWs in order to take full advantage of their large aspect ratio, so that their coercivity values are enhanced for the specified applications. Conventionally, synthesis of γ-Fe 2 O 3 NWs is carried out through the formation of intermediate NWs of other iron oxide phases (like α-Fe 2 O 3 , α-FeO(OH), or Fe 3 O 4 ) using toxic chemicals followed by either reduction, oxidation or transformation processes. These kinds of procedures are usually lengthy, time consuming and difficult to accomplish a good control over the crystallinity and phase purity of the resultant nanostructures.</p><p>To overcome the above stated limitations, Furostanol Saponin (FS) present in the Fenugreek seeds extract is utilized to synthesize iron oxides of size and shape-controlled nanostructures along with control over their growth direction. For the first time, this proposed methodology has been used to produce high yield and high quality of γ-Fe 2 O 3 nanograsses (NGs), nanowires (NWs) and nanospheres (NSs) via biogenic (greener) route 23,24 . The resultant γ-Fe 2 O 3 nanostructures are characterized by using various spectroscopic and microscopic techniques. Further these materials are explored for the simultaneous electrochemical detection of neurotransmitters, the potential biomarkers of many diseases. Simultaneous electrochemical detection of analytes remains still a challenge and we have successfully demonstrated in this work the simultaneous electrochemical sensing of dopamine (DA) and uric acid (UA) using NWs and NSs of γ-Fe 2 O 3 . Such studies are performed by modifying glassy carbon electrode (GCE) with these materials that exhibit increased electrochemical sensing properties towards the detection of DA and UA. Although the basic materials are being self-assembled using chemicals that themselves are biologically unfriendly but the approach of directional self-assembly using a naturally available extract eliminates the need of some very toxic precursors and uneconomic lengthy processes that are otherwise indispensable for obtaining such directional self-assembly of nanomaterials.</p><!><p>Morphological studies. In order to understand the formation process of different nanostructures of iron oxide synthesized using varied volume % of FS extracts, time-dependent evolution of these iron oxide nanostructures was analyzed by varying the time duration associated with the reaction. The samples were collected at different time intervals between 30 min and 3 h from the reaction mixture once the precipitate was formed and their corresponding FESEM images are shown in Fig. 1.</p><p>At the early stage of evolution process (30 min), it can be seen that the product is mainly composed of smaller spherical nanoparticles with an average particle size of about 10, 7 and 5 nm for 2%, 4% and 10% respectively. It is possible to obtain controlled sizes of spherical NSs of γ-Fe 2 O 3 by controlling the volume % of FS extract (biosurfactant) from 2 to 10% at a shorter reaction time of 30 min. When the concentration of FS was reduced from 10 to 2%, the size of the nanocrystals increased from 5 to 10 nm. It has been demonstrated that the spherical NSs are produced in all the volume % of FS (under the same reaction time of 30 min), indicating the homogeneity of the formation of NSs is increased with increasing concentration of FS. The reason is found to be, at a shorter period of reaction time (30 min) the dipolar attraction between these NSs is weaker and the steric repulsion is predominant. At this condition, only smaller particles of NSs are formed with almost no 1D nanostructures are generated.</p><p>To investigate the effect of surfactant concentration and evolution time on the growth of 1D γ-Fe 2 O 3 nanostructures, the ratio of iron precursors to FS concentration was varied at 2:1 ratio comprising of Fe and 2, 4, 10% of FS concentrations by allowing the reaction time of 3 h. Figure 2a-c show the corresponding FESEM images of γ-Fe 2 O 3 grown at 2%, 4% and 10% concentrations for 3 h time duration. As the reaction time is increased, the spherical nanoparticles (observed for 30 min) are completely transformed to 1D nanostructures and found that the oriented aggregation has occurred when the reaction time is increased to 3 h.</p><p>From these images it is clear that the large number of γ-Fe 2 O 3 NSs self-assembled into 1D nanostructures like NGs and NWs in case of 2% and 4% FS reaction medium respectively. At 2% FS concentration, γ-Fe 2 O 3 NGs are formed with an average length and width of 2,500 nm and 100-300 nm. By increasing the concentration of FS from 2 to 4%, NWs are formed with an average length of 2000 nm and the diameter is determined to be 90 nm. The average width of these NWs is narrowed. It is interesting to note that the NW structure is formed when explicit lateral growth of NG is stopped and the longitudinal growth begins. These results revealed that the increased FS extract (more -OH ions) causing the end to close and width to decrease, and most of the saponin (-OH ions) acts as a shape controller to induce anisotropic growth. Therefore, -OH ions play a much crucial role in the formation of 1D nanostructure because of their stronger adsorption effect. Conversely, it can be seen that the obtained products are spherical in shape (monodispersed and non-agglomerated spherical nanoparticles) at higher concentration (10% FS) of surfactant.</p><p>The average particle size of these iron oxide nanospheres is found to be 12 nm. These results demonstrate that the FS concentration significantly influences the morphology of resultant iron oxide nanostructures. In case of www.nature.com/scientificreports/ spherical, crystalline γ-Fe 2 O 3 NSs prepared using 10% FS, the lack of central symmetry due to the distribution of polar facets may result in dipolar attraction along the polar faces of these NSs. For γ-Fe 2 O 3 nanocrystals, (311) plane is polarized which leads to dipolar attraction of γ-Fe 2 O 3 nanoparticles along this axis 25,26 . It is well known that the dipole-dipole attraction favored the oriented attachment, but there is another steric or electrostatic repulsion which prevents them from aggregation. The steric repulsion originates from the stabilizing molecules (such as FS) on the surface of NSs, resulting in the repulsion of spherical particles within each other (10% FS).</p><p>On the other hand, the stabilizing molecules maintain a dynamic equilibrium on the surface of these particles during the adsorption and desorption processes under certain conducive reaction conditions. Increasing the FS concentration in the reaction medium is more favourable for the formation of NSs and to prevent the formation of 1D nanostructures of γ-Fe 2 O 3 . Based on these results, it can be reasonably understood that 1D nanostructures are formed through the aggregation of NSs during the synthesis process, which is usually termed as "oriented attachment". This mechanism involves a two-stage process including the formation of spherical nanoparticles and subsequently the aggregation of a string of these NSs 27 .</p><p>More detailed morphological analysis on the formation of γ-Fe 2 O 3 nanostructures synthesized using various volume % of FS extract for 3 h time duration is illustrated by TEM images along with their corresponding selected area electron diffraction pattern shown in Fig. 3. These TEM images revealed that all the nanostructures synthesized using different % of FS concentrations are in different shapes but uniform in size. The NG and NW like morphology are observed for the iron oxide samples synthesized using 2% and 4% FS medium respectively. In contrast, for 10% FS concentration, non-agglomerated spherical nanoparticles are produced with homogenous dispersion and their respective size is found to be 12 nm. These results are also well matched with the FESEM images of the same nanostructures discussed earlier. The corresponding SAED patterns attained from the large area of particles are shown in Fig. 3b, d, f. Formation of rings in these patterns also reveal the existence of cubic spinel structure of γ-Fe 2 O 3 and well correlated to the hkl (220), (311), (400), (422), ( 511) and (440) planes of X-ray diffraction study discussed later.</p><p>Bio-surfactant assisted formation of 0D and 1D iron oxide nanostructures. The formation mechanism of 1D nanostructured γ-Fe 2 O 3 is proposed mainly based on the chemical structure of FS (capping agent/bio-surfactant) (Supplementary Fig. S1). It seems that the growth of 1D nanostructure in a particular direction is predominantly due to two factors. The first one is the probability of adsorption of FS on the surface of these nanoparticles building blocks rendering a preferential assembly along a certain direction 28 and secondly, the concentration of OH − ions present in FS assists the growth of NGs or NWs of γ-Fe 2 O 3 . It is reported that a slow hydrolysis of the transition metal salts helps to control the branching of the building blocks and leads to a guided growth [29][30][31] . On the other hand, the presence of higher OH − concentration enhances the agglomeration of nanocrystals to a greater extent by ascertaining the metal-oxygen-metal bonds at many different sites and locations which eventually leads to an unguided growth 32,33 .</p><p>Interestingly the inherent anisotropy of crystal structure or crystal surface reactivity is investigated in the previous reports 34,35 , and this has been identified as the driving force for the growth of 1D nanostructures. G. Michael et al. showed that well defined facets of Ag nanoparticles may have different polarizability and reactivity, which leads to the oriented formation of Ag nanowires 36 . In our work, under a certain reaction condition (2% and 4% FS), the adsorption and detachment rate of the stabilizing agent (FS) molecules on the surface of NSs is accelerated. The acceleration of these adsorption and detachment rate of FS molecules reduces the mutual electrostatic repulsion between the nanoparticles. When the attraction force between these NSs dominates over the steric repulsion, the oriented attachment of NSs occurs during the synthesis process. Thus, the product is determined by the repulsion and attractive forces between these nanoparticles. Hence it can be reasonably concluded that the competition between dipolar attraction and steric repulsion influences the growth of these nanocrystals and hence the resultant nanostructures.</p><p>On the other hand, higher concentration (10%) of the bio-surfactant strongly competes with the iron oxide nanostructures and helps to retain the nanospherical structure, where only homogeneous sphere shaped nanoparticles are formed. These bio-surfactant molecules form repulsive forces between the nucleated particles and subsequently prohibiting the further growth. The ability of these micelles to prevent the growth of these particles becomes stronger at higher surfactant concentrations and hence the average particle size decreases, finally leading to the homogeneous dispersal of the spherical nanoparticles. It also means that the distribution of bio-surfactant in various directions on the surface of these nanoparticles is highly isotropic, hence, the better size distribution of the crystal growth in the process could be obtained with more befitting amount of the surfactant used in the reaction 37 . These idiosyncratic properties of the saponin rich bio-surfactants are not only play a vital role in designing the morphology of iron oxide nanostructures and also facilitate the formation of various phases of iron oxides.</p><!><p>Further to confirm the crystalline structure of resultant γ-Fe 2 O 3 samples synthesized at different volume % (2%, 4% and 10%) of FS and for 30 min and then for 3 h time duration, XRD studies are carried out. The corresponding XRD patterns are shown in Fig. 4. Typical XRD characteristics observed for the synthesized iron oxide NSs correlate very well with the cubic inverse spinel phase of γ-Fe 2 O 3 . Figure 4a represents the XRD patterns of NSs of γ-Fe 2 O 3 prepared for 30 min, where all the diffraction peaks appeared at 2θ values of 30.2, 35.4, 43.5, 57.3 and 62.1º corresponding to hkl planes of (220), (311), (400), (511) and (440) that can be indexed to the pure form of γ-Fe 2 O 3 (JCPDS No. 89-5892) as shown in the standard XRD data (Supplementary Fig. S2). Formation of the intense peaks in XRD patterns indicates the pure crystalline nature of resultant γ-Fe 2 O 3 products. In addition, these samples are phase pure, as the spectra did not show any traces of Further XRD results showed that NGs and NWs are preferentially oriented along (311) plane. Moreover, for the increased FS content in the reaction mixture (at 10%) the structure uniformity is maintained but did not form 1D γ-Fe 2 O 3 nanostructures as like in the samples synthesized using 2% (NGs) and 4% (NWs) FS. This can be inferred from the fact that increased number of spheroid-shaped nanoparticles is observed under FESEM and TEM analyses (Figs. 2, 3). This is also reflected from the increased height of particularly maghemite peaks in XRD studies, while the intensity values of the wires and grass shaped maghemite nanoparticles are decreased. The broad nature of these diffraction bands observed in XRD pattern recorded for 10% FS is an indication of the formation of smaller sized particles with an average size of 12 nm. It is also explained and reiterates that the use of bio-surfactant (FS) did not result in the phase change of γ-Fe 2 O 3 , but only controlled the size and shape of the resultant particles 39 . FTIR spectra of pure FS extract and different samples (NGs, NWs and NSs of γ-Fe 2 O 3 ) synthesized using various volume % of FS for 30 min and 3 h time duration have been recorded to confirm the formation of Fe-O bond in these samples. FTIR spectra (Supplementary Figs. S3, S4) of all the iron oxide nanostructures consisting of the characteristic -OH stretching (ν OH) and -HOH bending (δ OH) and vibrational bands in the region between 3,100 and 3,400 cm −1 respectively. This could be attributed to the stretching vibrations of adsorbed moisture and the surface hydroxyl groups present on the surface of iron oxide. The sharp peaks centered at 1,100 cm −1 in all these spectra could be assigned to (δ C-O-H) stretching vibration of saponin molecule present in the FS extract and the peak at ~ 1,620 cm −1 represents pure -CH 2 group vibration of sugar moieties found in FS. Similarly the band appeared at ~ 1,370 cm −1 could be attributed to the deformation vibration of C-H bond of alkane present in the FS extract. Similarly, appearance of a transmittance peak at 604 cm −1 corresponds to the stretching vibration of tetrahedral iron atoms (ν Fe-O).</p><p>Further XPS analysis is carried out to examine the chemical environment of the elements and the oxidation state of iron in these resultant nanostructures (NGs, NWs and NSs) since the core electron lines of both ferrous and ferric ions can be detected and are distinguishable from each other using XPS studies, which is not possible with XRD analysis. The corresponding results are shown in Fig. 5. The binding energy peaks emerged at 710.7 eV and 724.6 eV for γ-Fe 2 O 3 NGs produced from 2% FS; 710.4 eV and 724.5 eV for the NWs prepared from 4% FS and 710.5 eV and 724.8 eV for the NSs synthesized from 10% FS concentrations are the characteristic doublet peak of Fe 2p 3/2 and Fe 2p 1/2 core-level electrons respectively. Formation of these core level peaks indicates the presence of + 3 oxidation state in γ-Fe 2 O 3 . Moreover, the well-resolved satellite peaks appeared at a higher binding energy side of the main doublet peaks are noted at 719.1 eV, 719.3 eV and 719.4 eV for the iron oxide samples synthesized from 2%, 4% and 10% of FS respectively. These peaks represent the fingerprints of the electronic structure of Fe 3+ (indirectly indicates the absence of + 2 ion) 40,41 . Thus XPS results revealed that the charge transfer screening can be solely attributed to the presence of Fe 3+ ions in γ-Fe 2 O 3 . All the above experimental results confirm that the synthesized samples are in pure form of γ-Fe 2 O 3 rather than Fe 3 O 4 phase 42 . Therefore, on the basis of colour, FESEM, TEM, XRD, and XPS analyses, the synthesized nanostructures are identified to be a phase pure form of cubic γ-Fe 2 O 3 with their preferential growth oriented along (311) and (220) planes of the crystal. , where only 15% weight loss is noted. However this higher percentage weight loss observed for NSs suggested the conversion of oxyhydroxide forms to oxide of iron. The second comparatively smaller weight loss of 1.93% can be attributed to the structural arrangement of iron oxide. Consequently, 2nd and 3rd stage weight loss can be attributed to the decomposition of organic moieties present in the FS and phase transformation of γ-Fe 2 O 3 , where higher weight loss for the NSs can be seen because of the more amounts of organic moieties present in the extract (10% FS) adsorbed on the surface of these NSs. Figure 6B represents the broad endothermic peaks at around 120 °C observed in both the DTA curves and is mainly ascribed to the removal of physically bounded water molecules. The above mentioned characteristics arises from the finite size effect, where the surface availability on the nanostructures for the anchoring of water molecules (-OH) play a vital role. On increasing the temperature, the noticeable endothermic peak observed between 600 and 700 °C in the DTA curve is also associated with the transformation of nanosized γ-Fe 2 O 3 phase. This implies that the thermal stability of the synthesized γ-Fe 2 O 3 nanostructure is very high; usually the www.nature.com/scientificreports/ transformation occurs between 450 and 550 °C for the bulk samples. However, the disappearance of obvious exothermic peak at around 640 °C in the DTA curve of both these samples, representing no phase transition occurs from γ-Fe 2 O 3 to α-Fe 2 O 3 43 . Thus, the present bio-surfactant assisted co-precipitation method elucidated its uniqueness by tailoring the stable phase of γ-Fe 2 O 3 nanostructures under simple experimental conditions.</p><p>Investigation of magnetic properties. Furthermore, the effects of size and shape modulation of the resultant nanostructured materials of γ-Fe 2 O 3 on their magnetic properties are investigated for the synthesized samples and their corresponding hysteresis loops are depicted in Fig. 7. The room-temperature hysteresis loops for a series of samples synthesized from various volume % (2%, 4% and 10%) of FS extract for 30 min (Fig. 7a) and for 3 h (Fig. 7b) time duration have been recorded. These curves exhibit neither remanence nor coercivity, regardless of the nanocrystal size and shape and therefore indicate a superparamagnetic behaviour. The saturation magnetization (M s ) values of all these synthesized γ-Fe 2 O 3 samples fall below that of bulk γ-Fe 2 O 3 value of 74 emu/g, which provide further evidence for the phase purity of these materials 44 . The room temperature M s values of different nanostructured materials synthesized at 2%, 4% and 10% of FS using 30 min are measured to be 57, 41 and 34 emu/g respectively. On the other hand, the measured M s values for γ-Fe 2 O 3 NGs, γ-Fe 2 O 3 NWs synthesized using 2% and 4% of FS for 3 h time duration are determined to be 39 and 28 emu/g respectively. The slope of the magnetization is positive even at larger applied fields with an unending hysteresis. This is attributed mainly to the formation of 1D nanostructures that are finite, non-ellipsoidal particles and cannot be saturated by a homogeneous applied field. For γ-Fe 2 O 3 NSs produced from 10% FS, the M s value is further reduced to 16 emu/g as the size decreased nearing to that of quantum dots [45][46][47] .</p><p>These results indicate the single domain magnetic characteristics of the nanoparticles that remained same for the NSs synthesized using 30 min and 3 h (10% FS) time duration. Since the amount of all these samples used for the measurement of magnetic properties was kept constant, the decrease of saturation magnetization is mainly ascribed to the increased amount of bio-surfactant (FS) covered over the maghemite nanoparticles that hinders the magnetic susceptibility. Hence, the notable decrease in the saturation magnetization is most likely due to the existence of FS on the surface of γ-Fe 2 O 3 nanoparticles which may create a magnetically dead layer. With a significant fraction of surface atoms on these nanoparticles and any crystalline disorder within the surface layer may also lead to a significant decrease in the saturation magnetization values of these nanoparticles 46 . Thus the magnetic characteristics of these synthesized iron oxide nanostructures could be influenced by the shape, size and concentration of the bio-surfactant used for the preparation.</p><p>Electrochemical sensing of DA and UA using nanostructures of γ-Fe 2 o 3 . Different nanostructures of γ-Fe 2 O 3 viz., NGs, NWs and NSs are explored further for the electrochemical sensing application for the simultaneous detection of DA and UA. The effects of shape and size of these nanostructures on the electrochemical sensing characteristics are also investigated. Electrocatalytic activity of these γ-Fe 2 O 3 nanostructures was studied by coating these materials onto pre-cleaned glassy carbon electrode (GCE) surface and their electrochemical oxidation capability towards DA and UA was analyzed. These studies are carried out using γ-Fe 2 O 3 / GCE through cyclic voltammetry (CV) technique within the potential range of − 0.2 V to + 0.8 V in aqueous phosphate buffer solution (PBS) having pH = 7.4. Among the different nanostructures studied in this work, NWs and NSs of γ-Fe 2 O 3 found to exhibit better electrochemical sensing capability for DA and UA (Supplementary Fig. S5). Hence the subsequent studies were carried out with these nanostructures. The corresponding cyclic voltammograms recorded for 0.5 mM DA and 0.5 mM UA using NWs and NSs of γ-Fe 2 O 3 are shown in Fig. 8A, B respectively and for comparison similar studies performed using bare GCE were also shown in the figure. The www.nature.com/scientificreports/ individual electrochemical oxidation of both DA (Fig. 8A) and UA (Fig. 8B) on γ-Fe 2 O 3 nanostructures modified GCE exhibited negative shift of anodic potentials with increased current response when compared to bare GCE. The electrocatalytic oxidation of DA is evident from the oxidation peak appeared at 290 mV, 250 mV and 300 mV for NSs and NWs of γ-Fe 2 O 3 modified GCEs along with a bare GCE, respectively. For the electrochemical detection of UA, the oxidation peaks appeared at 350 mV and 330 mV for NSs and NWs of γ-Fe 2 O 3 modified GCEs, whereas in the case of bare GCE the voltammetric response for UA appeared at a higher potential of 520 mV and no obvious cathodic peak potential are observed in all these cases, suggesting the irreversibility of the electrochemical process 36,48,49 . Various electrochemical characteristic parameters are determined for these electrodes from their corresponding CVs (Supplementary Table S1). Interestingly the current response observed for the electrocatalytic activity of DA and UA at γ-Fe 2 O 3 NSs modified GCE is slightly lower than that of γ-Fe 2 O 3 NWs modified GCE. On the other hand, oxidation peak current values noted at γ-Fe 2 O 3 NSs and NWs modified GCEs are greater than that of bare GCE. These results indicate the fast electron transfer kinetics of DA and UA on γ-Fe 2 O 3 nanostructures coated GCEs. The major reason is that the synthesized γ-Fe 2 O 3 NWs can act as a catalyst to increase the rate of electron transfer and lower the overpotential for oxidation of DA and UA. Moreover the fast response resulted from the excess electro-active sites provided by γ-Fe 2 O 3 NWs as well as the good conductivity between the coated film and GCE substrate. The reason for this improved electrocatalytic activity is mainly attributed to the 1D structure of γ-Fe 2 O 3 NW, which acts as an "electron wire", wherein the electron diffusion occurs at a faster rate that aids in increased sensitivity towards the sensing of DA and UA. Such a nanostructure is more favourable for providing a large contact area between the sensing materials and the analyte species than non-hierarchical nanoparticles or bulk Fe 2 O 3 powders. Furthermore, due to the small size of these NWs, the charge distribution on the surface may lead to less resistance to the diffusion of probe ions onto the electrode surface than that of the bulk Fe 2 O 3 .</p><p>Further the material exhibits a facile redox process associated with Fe within this potential window that might result in enhanced conductivity both in terms of ionic and electronic transport phenomena. Such kind of redox process involving Fe(III)/Fe(II) centers of γ-Fe 2 O 3 accelerates the electrocatalytic mechanism behind the sensing of these bio-analytes. Consequently, γ-Fe 2 O 3 NWs modified GCE projected as a sensing platform for the oxidation of DA and UA exhibits more electro-active sites and strong adhesion onto the surface, resulting in the enhanced sensitivity and shows a lower overpotential for the sensor. Finally, from these results it is clear that γ-Fe 2 O 3 NWs aid in more effective transportation and accessibility of bio-analytes and subsequently facilitates the electron transfer process that ultimately leads to less non-faradaic behaviour with pronounced sensitivity 50 .</p><!><p>NWs modified GCEs and hence the optimization studies were carried out over different pH values ranging from 5.4 to 9.4 and the corresponding results are shown in Figs. 9 and 10. These studies are performed to determine the optimum pH value at which the highest electrochemical responses for the oxidation of DA and UA are achieved. From these results it is found that the oxidation peak current of DA reaches the maximum value at pH 7.4 and then drops quickly and finally decreases with increasing pH. Similarly, the peak current of UA also initially goes up gradually with increasing pH (5.4-7.4), and then it decreases with increasing pH (> 7.4). In addition, Figs. 9B and 10B show the variation of peak potentials of both DA and UA oxidation with pH and these plots exhibit a pH dependency. The oxidation peak potential shifted to less negative value when the pH of the solution is increased, which indicates that the electrocatalytic oxidation of DA and UA at these modified electrodes involves equal number of electrons and protons 51,52 . Meanwhile, increasing pH also leads to a sharp decrease in the oxidation current when pH is more than 7.4 (Figs. 9A, 10A). Moreover, the maximum oxidation current response is observed at pH of 7.4 for the determination of DA and UA at γ-Fe 2 O 3 NWs modified electrodes and hence it is adopted as the optimum pH for further experiments. These studies are carried out for a wide range of scan rate varying from 10 mV/s to 100 mV/s using CV within the potential range of − 0.2 V to + 0.8 V. Figures 11 and 12 show the respective CV curves of 0.5 mM DA and 0.5 mM UA recorded using γ-Fe 2 O 3 NWs modified GCEs at different scan rates. It can be seen that the anodic peak current values of DA are proportional to the square root of scan rate in the range from 10 to 100 mV/s with a correlation coefficient of R 2 = 0.9859 (Fig. 11b). Meanwhile, the anodic peak current values of UA also showed an excellent linear relationship with the square root of scan rate in the range from 10 to 100 mV/s with a correlation coefficient of R 2 = 0.9779 (Fig. 12b). As shown in these figures, with increasing scan rate, the redox current values increased and good linear relationships are established between the peak current values and the square root of scan rate, respectively, indicating that the oxidation of DA and UA on γ-Fe 2 O 3 NWs modified GCE is a diffusion controlled process 53 .</p><p>Electrochemical sensing of DA using DPV studies. Differential pulse voltammetry (DPV) technique provides a higher sensitive current and better detection limit when compared to CV and hence the determination of DA was carried out by using DPV. Figure 13a www.nature.com/scientificreports/ concentrations of DA. Interestingly DPV response of γ-Fe 2 O 3 NWs/GCE is linear over the concentration of DA ranging from 0.15 μM to 75 μM (Fig. 13b) with the correlation coefficient of R 2 = 0.9898. This study displayed a highest linear range of detection for DA and the detection limit is determined to be 0.1 μM. The sensor characteristic parameters of the fabricated system have also been compared with the other reported DA sensors and the comparative results are shown in Table 1. The analytical comparison of our proposed sensor clearly reveals that γ-Fe 2 O 3 NWs modified GCE exhibits a lower detection limit and better linear concentration range for the detection of DA on comparison to many other modified electrodes [54][55][56][57][58][59][60][61][62][63][64] . This electrocatalytic effect of DA is mainly attributed to the larger available surface area of the modifying layer due to the nanometer size of the sample and free surface -OH groups of γ-Fe www.nature.com/scientificreports/ current response and the concentration of UA in the range of 5 µM-0.15 mM with a correlation coefficient of R 2 = 0.9909, suggesting a potential application of γ-Fe 2 O 3 NWs modified GCE in the quantitative determination of UA. In comparison with the other values reported in literature, this enzyme-free electrode exhibits a wider linear concentration range than the other modified electrodes (Table 2). Furthermore, a lower detection limit of 0.5 μM UA is determined at a signal-to-noise ratio of 3 and also stipulates the minimum detectable amount of analyte using the developed sensor [67][68][69][70][71][72][73][74][75][76][77] . Thus the experimental results suggested that γ-Fe 2 O 3 NWs can enhance the electron transfer rate and lower the overpotential of UA oxidation. This enhanced electrochemical sensing property is mainly attributed to the larger electro-active sites of the modifying layer due to the large number of NWs on the sample as evident from FESEM and TEM analyses. Furthermore, γ-Fe 2 O 3 layer contains free surface -OH groups that could form hydrogen bonding with UA. It is well known that hydrogen bond acceptor strength of amide group is stronger than the ester group and it facilitates easier oxidation of UA at γ-Fe 2 O 3 NWs modified electrode surface. Hence, the carbonyl group at C-8 of UA forms a hydrogen bonding with surface -OH groups of γ-Fe 2 O 3 NWs. It is also evident that only Fe 2+ /Fe 3+ redox couple is electro-active and hence Fe site plays a vital role in the electrocatalytic oxidation of UA. Recall that the oxidation of UA is irreversible at these electrodes and the oxidation of UA proceeds via 2e − , 2H + process 68 . According to the previous reports, the electrochemical oxidation mechanism of UA at γ-Fe 2 O 3 NWs/GCE can be explained as depicted in Scheme 2.</p><p>As shown in Tables 1 and 2, γ-Fe 2 O 3 NWs modified sensor exhibits a better detection limit and a wider linear concentration range of detection in comparison to most of the other reported sensors. Even though the detection limit of our modified electrode is not as low as few other reported sensors, this particular one displayed acceptable results with a positive gain in the potential shift and sufficient enough to detect the physiologically relevant concentrations of the target analytes. Moreover, the synthesis protocol reported here for the preparation of iron S2). Further the anodic peak potential separation between DA and UA is estimated to be 230 mV and is sufficient enough to identify them as a well-defined two separate peaks. Among different samples, γ-Fe 2 O 3 NWs showed better performance in terms of better peak separation and enhanced current values. Thus the simultaneous oxidation of DA and UA shows that γ-Fe 2 O 3 nanostructures coated GCEs exhibit excellent electrocatalytic behavior of these species. www.nature.com/scientificreports/ clear that no obvious interference is observed, demonstrating the good selectivity of as-fabricated γ-Fe 2 O 3 NWs coated GCE. From these results, it is understood that tenfold physiological and 100-fold common ion interference concentrations did not significantly affect the detection of DA and UA; manifesting a higher selectivity of the proposed biosensor [78][79][80][81] . These results vividly suggest that γ-Fe 2 O 3 NWs modified GCE is highly selective towards DA and UA sensing when compared to the aforementioned interfering compounds. 17a-c. It can be noted from these CVs that mere a 2% decrease in peak current value was observed even after 50 cycles, indicating a good stability of the electrode. Thus, these modified electrodes are successfully used for the simultaneous determination of DA and UA with excellent sensitivity and selectivity. Finally, from these electrochemical results, it can be concluded that γ-Fe 2 O 3 nanostructures modified GCEs can act as a suitable redox mediator and shows potentially efficient electrocatalytic activity for the simultaneous determination of potential biomarkers such as DA and UA. Also, it overcomes the interest of using other film modified GCEs and found as a novel and economically viable one for the simultaneous determination of DA and UA. This study constitutes a simple and versatile protocol which can be used in an effective way to develop electrochemical sensors and biosensors.</p><!><p>Chemicals. Reagent grades of iron (III) chloride hexahydrate (FeCl 3 .6H 2 O), iron (II) sulphate heptahydrate (FeSO 4 •7H 2 O), sodium hydroxide (NaOH), uric acid and dopamine are purchased from E-Merck specialties products, India and were used as such without any further purification. Fenugreek seeds were purchased from the local market in Karaikudi, Tamilnadu, India. Double distilled water was used as a solvent to prepare the extract.</p><p>Preparation of Fenugreek seed extract. Fenugreek (Trigonella foenum-graecum) is basically a spice crop belongs to the family of Fabaceae. It is a semi-arid plant and cultivated worldwide. Generally their pods contain 10 − 20 seeds that are cuboid in shape and yellow to amber in colour. They are mainly used to prepare extracts or powders for cuisines and for medicinal applications. For the preparation of extract, about 10 g of Fenugreek seeds were washed several times with distilled water to remove the surface impurities and then immersed in 50 mL of distilled water for about 12 h at room temperature. These seeds were grinded well and the extract was filtered using filter paper (Whatmann grade no. 10) and then centrifuged at 8,000 rpm for 15 min. Finally, the supernatant was separated and stored at 4 °C for further use. FTIR spectroscopy is used for the analysis of purity of such extract. In a typical experimental procedure, about 0.9, 1.8 and 4.5 mL of FS extracts were used to prepare 2%, 4% and 10% (v/v %) of FS utilized for the synthesis of various nanostructures of iron oxide.</p><!><p>Different nanostructures of iron oxide were prepared by an aqueous co-precipitation method using FeCl 3 •6H 2 O and FeSO 4 •7H 2 O as the primary sources of iron. In a systematic experimental procedure, 30 mL of 0.1 M FeCl 3 •6H 2 O solution was mixed with 15 mL of 0.1 M FeSO 4 .7H 2 O solution in a clean three-necked round bottom flask 25 . Appropriate amounts of saponin rich bio-surfactant, Fenugreek seed extracts were added into the above reaction mixture and allowed to react for an hour. After that, 10 M NaOH aqueous solution was slowly added into the above solution under vigorous stirring to bring the pH upto13 and the stirring was continued for another 2 h for the completion of reaction. Interestingly the colour of this particular solution turned from orange to black, consequently a reddish-brown precipitate was formed. The resulting solid product was removed; washed very well with double distilled water and dried at room temperature in atmospheric air. Following the above procedure, the experiments were carried out under various volume percentage values such as 2%, 4% and 10% of FS extract in the pre-fixed iron precursor's ratio and also at different time intervals. This procedure yields different nanostructures of iron oxide as discussed in the results and discussion part.</p><!><p>γ-Fe 2 O 3 and their structural and morphological characteristics were examined initially using field emission scanning electron microscope (FESEM, Hitachi MODEL S-4800) 18,25 . Similarly, transmission electron microscope (TEM) images were obtained using JEOL TEM 2010 microscope operated at 200 kV 18,25 . X-ray diffraction studies (XPERT-PRO with Cu Kα radiation [λ = 0.154060 nm], PANlytical X'Pert Pro-diffractometer) were carried out to evaluate the crystalline structure, phases and phase purity of the resultant iron oxide samples 25,[39][40][41][42] . Fourier transform infrared spectrometer (FTIR, Nicolet 5700) was used to analyze the surface characteristics and the presence of chemical functionalities in the synthesized γ-Fe 2 O 3 nanostructures. X-ray photoelectron spectroscopy (XPS) was recorded for different nanostructures of γ-Fe 2 O 3 using Kratos ASIS-HS instrument equipped with a standard monochromatic source (Al K α ) operated at 150 W (15 kV, 10 mA) to confirm the presence of elements along with their oxidation states and nature of the formed products 18,25 . Finally the magnetization hysteresis loops of different nanostructures were also characterized using a Lake model 7,300 vibrating sample magnetometer (VSM).</p><p>Fabrication of sensor matrix using iron oxide nanostructures modified GCE. Synthesized γ-Fe 2 O 3 nanostructures by employing different weight % of FS extracts were utilized further for the electrochemical detection of DA and UA by modifying GCEs with these materials. Prior to modification, GCE was polished to a mirror-like surface with progressively decreasing 1 μm, 0.3 μm and 0.05 μm alumina slurries and rinsed with double distilled water thoroughly between each polishing step 25,49,50 . Then it was washed successively www.nature.com/scientificreports/ with double distilled water in an ultrasonic bath and dried in air at room temperature. About 1 mg of a particular nanostructure of γ-Fe 2 O 3 was dispersed in 3 mL of ethanol under ultrasonication. Subsequently γ-Fe 2 O 3 modified GCEs were obtained by drop casting 5 μL of these suspensions onto the surface of pre-cleaned GCE 49,50 . Finally the modified GCE was activated in phosphate buffer solution (PBS) having pH 7.4 by successive cyclic scans between -0.2 and + 0.8 V [54][55][56][57][58][59][60] . Before and after each experiment, the modified GCE was washed well with distilled water and reactivated in PBS as mentioned above.</p><p>Electrochemical sensing of DA and UA using Iron oxide Modified GCEs. Electrochemical sensing studies of DA and UA were performed using different nanostructures of γ-Fe 2 O 3 modified GCEs and bare GCE without any coating (for comparison) in an aqueous solution consisting of 0.5 mM DA and UA. N 2 gas was purged into the freshly prepared DA and UA solutions for about 5 min to eliminate the dissolved oxygen and overflowed to avoid the atmospheric oxygen interference during the electrochemical oxidation of target analyte 49,50,[62][63][64] . All the electrochemical measurements such as cyclic voltammetry (CV), differential pulse voltammetry (DPV), and chronoamperometry (CA) were carried out using CHI 6131D Electrochemical Impedance Analyzer procured from USA using either the modified GCE or bare GCE as a working electrode, a platinum wire was used as a counter electrode, and saturated calomel electrode (SCE) was used as the reference electrode respectively 18,25,[62][63][64]68 . Other details and necessary parameters were mentioned in the respective diagram provided in the results and discussion part.</p><!><p>In conclusion, a distinct type of preparation of zero dimensional and 1D γ-Fe 2 O 3 nanostructures with tunable sizes and shapes using fenugreek seeds extract as a new bio-surfactant has been reported. By careful tuning of the reaction parameters, pure phase of cubic spinel, superparamagnetic structures of γ-Fe 2 O 3 NGs, NWs and NSs have been successfully synthesized using a simple one-pot process. XRD, XPS and FTIR results confirmed that the synthesized nanostructures posses pure crystalline phase of cubic spinel γ-Fe 2 O 3 . FESEM and TEM analyses showed that the obtained morphologies are in 0D and 1D iron oxide nanostructures with an average particle size in the range of 12 nm (0D) and width of 230-250 nm (1D) respectively. Furthermore, γ-Fe 2 O 3 modified GCE shows efficient electrocatalytic activity for the simultaneous determination of DA and UA by significantly decreasing their oxidation overpotential values and enhancing the peak current values when compared to bare GCE without any modification. The proposed electrode is very easy to fabricate and eco-friendly for the simultaneous determination of biomolecules like DA and UA. A large value of peak separation observed between DA and UA also facilitates the simultaneous determination using γ-Fe 2 O 3 coated GCE. Among the different nanostructures of γ-Fe 2 O 3 studied in this work, NWs modified GCE showed an excellent selectivity and higher sensitivity for DA and UA detection, along with higher current response when compared to other nanostructures. It is also found that the other potentially interfering species showed negligible interference towards selective detection of DA and UA revealing a good selectivity of γ-Fe 2 O 3 . The relatively smaller sized unique nanostructures, higher surface area and the presence of surface hydroxyl groups are responsible for the enhanced electrocatalytic activity of γ-Fe 2 O 3 NWs modified GCE. Many outstanding advantages, like appreciable sensitivity, wide linear concentration range, lower detection limit and selectivity of the modified electrodes confirmed that γ-Fe 2 O 3 NWs possesses an excellent analytical performance by making them a promising candidate for electrochemical sensing applications. Our investigation provides an environmentally benign synthetic route for the preparation of iron oxide nanostructures, and the proposed method is eco-friendly, facile and novel that could be used for large scale production.</p>

Scientific Reports - Nature

Electronic Transitions in Different Redox States of Trinuclear 5,6,11,12,17,18‐Hexaazatrinaphthylene‐Bridged Titanium Complexes: Spectroelectrochemistry and Quantum Chemistry

AbstractMultinuclear transition metal complexes bridged by ligands with extended π‐electronic systems show a variety of complex electronic transitions and electron transfer reactions. While a systematic understanding of the photochemistry and electrochemistry has been attained for binuclear complexes, much less is known about trinuclear complexes such as hexaphenyl‐5,6,11,12,17,18‐hexaazatrinaphthylene‐tristitanocene [(Cp2Ti)3HATN(Ph)6]. The voltammogram of [(Cp2Ti)3HATN(Ph)6] shows six oxidation and three reduction waves. Solution spectra of [(Cp2Ti)3HATN(Ph)6] and of the electrochemically formed oxidation products show electronic transitions in the UV, visible and the NIR ranges. Density functional theory (DFT) and linear response time‐dependent DFT show that the three formally titanium(II) centers transfer an electron to the HATN ligand in the ground state. The optically excited transitions occur exclusively between ligand‐centered orbitals. The charged titanium centers only provide an electrostatic frame to the extended π‐electronic system. Complete active self‐consistent field (CASSCF) calculation on a structurally simplified model compound, which considers the multi‐reference character imposed by the three titanium centers, can provide an interpretation of the experimentally observed temperature‐dependent magnetic behavior of the different redox states of the title compound in full consistency with the interpretation of the electronic spectra.

electronic_transitions_in_different_redox_states_of_trinuclear_5,6,11,12,17,18‐hexaazatrinaphthylene

<!>Introduction<!><!>Introduction<!>Electrochemistry of [(Cp2Ti)3HATN(Ph)6] and the Ligand HATN(Ph)6<!><!>Electrochemistry of [(Cp2Ti)3HATN(Ph)6] and the Ligand HATN(Ph)6<!><!>Electrochemistry of [(Cp2Ti)3HATN(Ph)6] and the Ligand HATN(Ph)6<!><!>Spectroelectrochemical Measurements<!><!>Spectroelectrochemical Measurements<!><!>Spectroelectrochemical Measurements<!>Calculation of the Electronic Ground State and Electronic Excitations<!><!>Calculation of the Electronic Ground State and Electronic Excitations<!><!>Calculation of the Electronic Ground State and Electronic Excitations<!><!>Calculation of the Electronic Ground State and Electronic Excitations<!>Calculation of Ground States and Electronic Transition Spectra of Other Redox States<!><!>Calculation of Ground States and Electronic Transition Spectra of Other Redox States<!>Interpretation of Magnetic Properties and Electronic Excitations Using Multireference Methods<!><!>Interpretation of Magnetic Properties and Electronic Excitations Using Multireference Methods<!><!>Interpretation of Magnetic Properties and Electronic Excitations Using Multireference Methods<!><!>Interpretation of Magnetic Properties and Electronic Excitations Using Multireference Methods<!>Conclusions<!>Substances<!>Instrumentation<!>Quantum chemical calculations<!>Conflict of interest<!>

<p>A. Markovic, L. Gerhards, P. Sander, C. Dosche, T. Klüner, R. Beckhaus, G. Wittstock, ChemPhysChem 2020, 21, 2506.</p><!><p>Electron‐transfer (ET) reactions are fundamental for a lot of processes in nature[ 1 , 2 , 3 ] and numerous investigations have been devoted to the study of ET processes in chemical [2] and biological [1] systems. One of the long‐term goals is the desire to understand ET transfer processes in complex molecular systems that eventually will allow to synthesize molecular systems exhibiting control of electronic communication with similar accuracy as in nature. [4] The work of Creutz and Taube [5] on mixed‐valence compounds already illustrated the complexity of processes at binuclear complexes that bear conceptual similarity to ET reactions in natural systems. [3] In particular, polypyridyl complexes of the d6 metals Fe(II), Ru(II), and Os(II) have received considerable attention due to their chemical inertness in a variety of oxidation states. [3] While a systematic understanding of the photochemistry and electrochemistry has been attained for binuclear complexes, [6] much less is known about trinuclear and multinuclear complexes. [7] Trinuclear complexes can be obtained by using hexaazatriphenylene (HAT), [8] hexaazatrinaphthylene (HATN) ligands [9] and backbone substituted derivatives [10] thereof. Here we present investigations of the hexaphenyl substituted isomer (HATN (Ph)6) and the corresponding titanocene complex (Figure 1).[ 11 , 12 ]</p><!><p>Molecular structure of the ligand HAT, HATN, [HATN(Ph)6] and the complex [(Cp2Ti)3HATN(Ph)6].</p><!><p>In extension of the classical concepts of the coordination chemistry, multinuclear derivatives exhibiting bridging N‐heterocyclic ligands are discussed by Haiduc as inverse coordination metal complexes. [13] Trinuclear complexes of early transition metals, are particular interesting if they contain bridging ligands with extended π‐acceptor properties. (Figure 1). [10] Complexes of this type have been extensively studied as attractive supramolecular building blocks because those ligands possess three chelating sites for metal ions. [10] Secondly, the electron‐deficient π‐system of the ligand supports metal‐to‐ligand charge transfer and/or charge transfer between metal centers of the complex. [10] Thirdly, the C3 symmetry axis causes degeneracy of π* orbitals. [10] Fourthly, the ligands offer a variety of interesting chemical/physical properties based on their electronic structures which can be modified by their peripheral substituents. [10]</p><p>HAT and its derivatives also belong to the family of redox‐active ligands, often used as non‐innocent ligands. [14] One of the fundamental properties of these ligands is their ability to act as electron reservoirs, which allows the metal to store electrons on the ligand and/or accept electrons from the ligands. [15] This property is of interest for the design of homogeneous catalysts because many important transformations involve the transfer of multiple electrons between the catalyst and the activated substrate. Such transformations are common for expensive noble metals such as Pd, Pt, Rh, etc., but more difficult to achieve with cheaper and more abundant first‐row transition metals. Non‐innocent ligands actually allow first‐row transition metals to mimic some of the catalytic properties of more noble metals. [16] In addition, the electron transport properties of complexes with non‐innocent ligands have been studied for applications in redox‐flow batteries. [17]</p><p>Because of all the characteristic multifunctionalities of HAT‐type ligands, their metal complexes afford intriguing solid state structures, physicochemical and material properties and have opened up a new field of trinuclear coordination compounds. [10] Complexes with ruthenium, [18] rhenium, [19] cobalt [14] and titanium [7] have been investigated concerning their interesting electrochemical, photophysical and magnetic properties. Among others, the frequent occurrence of mixed‐valence situations have been emphasized.[ 20 , 21 , 22 ] In the seminal review of Kaim and Lahiri, [20] the intervalence charge Transfer (IVCT) is related to spectral features: "Frequently, the intervalence charge‐transfer absorption band have been recognized as the most conspicuous evidence for a mixed valence situation. They arise from intramolecular electronic transitions…" and "The IVCT band is usually observed in the visible or near infrared region of the spectrum and is broad." This notion is commonly accepted and even popularized in Wikipedia. [23] The extend of coupling between redox centers is commonly summarized by the three Robin‐Day classes for negligible (I), weak (II) and strong (III) coupling of the redox centers.[ 21 , 24 ] Spectroscopic methods are recommended for the analysis of those situations. [21]</p><p>In this study, we investigate the electronic transitions in the different redox states of the trinuclear titanium hexaphenyl‐5,6,11,12,17,18‐hexaazanaphtylene. (Figure 1). [12] The results have been obtained by extended voltammetric measurements and spectroelectrochemistry in the ultraviolet (UV), visible (vis) and near infrared (NIR) spectral regions. The initial, preliminary assessment of the NIR spectra as signatures of intramolecular electronic transitions[ 7 , 12 ] with IVCT, ligand‐to‐metal charge transfer (LMCT) and metal‐to‐ligand (MLCT) processes could not provide a contradiction‐free assignment of all spectral features despite its gross agreement with common textbook knowledge about mixed‐valence compounds and their electronic spectra. [23] This became only evident when considering the spectra of the different redox forms accessible by chemical synthesis or by oxidation/reduction in a spectroelectrochemical cell. Density functional theory (DFT) using linear response time‐dependent DFT and complete active space self‐consistent field (CASSCF) provided a surprising, alternative and comprehensive interpretation of the electronic structure, temperature‐dependent magnetic behavior and the resulting electron transition. Those methods consider true multi‐electron transitions rather than the single electron picture underlying the Robin‐Day classification and its extensions.</p><!><p>The cyclic voltammetry (CV) and differential pulse voltammetry (DPV) data of [(Cp2Ti)3HATN(Ph)6] in THF are shown in Figure 2. DPV were started at the open circuit potential (OCP) in the positive direction in order to record the oxidations (red) and in the negative direction in order to record the reductions (blue). The potential range, in which the neutral complex was stable, is colored in lavender, the potential range of the negatively charged species are marked in green and potential ranges in which the positively charged species are stable are marked in red. The DPV data are overlaid with the CV data because DPV data show a better resolution of the redox processes on the potential axis. Please note the separate ordinates for both CV and DPV experiments.</p><!><p>CV (black) and DPV (red, blue) of 0.3 mM [(Cp2Ti)3HATN(Ph)6] at an Au working electrode in 0.2 M TBAPF6 in THF. The shaded areas indicate in which potential range the complex is stable with the indicated total charge.</p><!><p>The voltammogram of [(Cp2Ti)3HATN(Ph)6] shows three reduction waves and six oxidation waves. All steps are one electron processes (Table 1). The six oxidations formally correspond to the removal of 6 valence electrons of the three Ti(II) centers. The rich redox chemistry of this [(Cp2Ti)3HATN(Ph)6] allows the observation of even one more redox transition in the accessible potential range than for the previously reported related compound [7] where 3 reductions and 5 oxidation signals were observed</p><!><p>Formal potentials (in V vs. Fc/Fc+) for electron transfer reactions of [(Cp2Ti)3HATN(Ph)6] from the data in Figure 2.</p><p>Redox pair</p><p>E°'</p><p>−3/−2</p><p>−2.66</p><p>−2/−1</p><p>−2.19</p><p>−1/0</p><p>−1.61</p><p>0/1</p><p>−1.08</p><p>1/2</p><p>−0.49</p><p>2/3</p><p>−0.12</p><p>3/4</p><p>+0.06</p><p>4/5</p><p>+0.24</p><p>5/6</p><p>+0.49</p><!><p>By comparison of the voltammetric data of the complex (Figure 2) with that of the ligand HATN(Ph)6 (Figure 3), the reductions 0→−1 and −1→−2 suggest that they lead to reduction of the ligand. However, such assignment based on voltammetric data alone can never be certain. The third reduction of the complex and −2→−3 cannot associated with a redox transition of the ligand alone.</p><!><p>CV (black) and DPV (blue) of 0.1 mM HATN(Ph)6 (no metal ions coordinated) at an Au disc working electrode in 0.2 M TBAPF6 in a mixture of 50 vol‐% THF and 50 vol‐% MeCN. The shaded areas indicate in which potential range the molecule is stable with the number of accepted electrons compared to the neutral starting state.</p><!><p>The solution spectra of [(Cp2Ti)3HATN(Ph)6] and its oxidation and reduction products [(Cp2Ti)3HATN(Ph)6]3−, [(Cp2Ti)3HATN(Ph)6]2−, [(Cp2Ti)3HATN(Ph)6]1−, [(Cp2Ti)3HATN(Ph)6]1+ and [(Cp2Ti)3HATN(Ph)6]2+ show electronic transitions in the UV, vis and NIR spectral ranges (Figure 4).</p><!><p>The spectra from the spectroelectrochemical measurements of [(Cp2Ti)3HATN(Ph)6] in THF; UV/Vis in blue and NIR in red. The signals are numbered for later reference. Signals marked by (*) are caused by vibrational overtone of molecular vibration of the solvent, which are not fully compensated. They are not relevant for the interpretation of the electronic transitions.</p><!><p>Different types of electronic transitions can be observed in transition metal complexes. In common literature they are usually visualized by molecular orbital (MO) diagrams, where MOs are combined from ligand orbitals (Figure 5a) and metal d orbitals in a ligand field of the Cp ligands (Figure 5b).[ 25 , 26 ] Electronic states in this approach are represented as linear combinations of the occupied MOs. In a single electron picture, transitions between these states can be included in the MO scheme as electron transitions between distinct orbitals. [26] Possible transitions are transitions from ligand‐centered orbitals to d‐type metal orbitals (LMCT), transitions between d‐type orbitals within one metal center (d‐d transitions), transitions from metal centered d‐type orbitals to ligand orbitals (MLCT) and of course transitions between purely ligand centered orbitals (in the visible region mainly π‐π*). In multinuclear complexes transitions between d‐type orbitals of different metal centers in different valence states can be observed as a fifth type of electronic transition, commonly referred to as IVCT. However, an IVCT is only possible in systems with extensive state mixing involving two d orbitals from the metal centers and one ligand orbital which provides strong electronic coupling between the metal centers. [20]</p><!><p>Schematic MO‐diagrams for a) HATN (no metals coordinated) and b) Cp2Ti. HATN has two degenerate HOMOs due to its symmetry (orbital shape is shown in SI‐3, Figure S4). The Cp2Ti complex has two unpaired electrons localized in the dx 2 ‐y 2 and dxy orbitals. For lower lying orbitals a strong mixing between cyclopentadiene and titanium is observed. All calculations were made using PBE0 and Def2‐SVP.</p><!><p>Experimental magnetic moments of the solid compound at elevated temperature suggest the existence of more than two unpaired electrons in the ground state of the neutral complex. [12] Due to the high electron affinity of the HATN(Ph)6 ligand at least one electron is supposed to be transferred from the three Cp2Ti centers to the ligand. The cyclic voltammetric data of HATN(Ph)6 in Figure 3 showed that the two‐electron reduction was irreversible suggesting that the [HATN(Ph)6]2− was unstable in the electrolyte solution. The combined magnetic and electrochemical data made the ground state plausible with a negatively charged ligand, although the exact number of transferred electrons cannot be ascertained with the available methods.</p><p>The measured spectra displayed in Figure 4 exhibit pronounced peaks within the NIR range which in many publications are assigned as an IVCT or LMCT.[ 20 , 21 , 22 ] For an assignment of these peaks and to combine the spectroscopic data with the electrochemical data, we attempted to relate the formal potential of electron transfer reactions (Figure 2, Table 1) and all spectroscopic transitions to an assumed single particle MO scheme for all redox states. From the already published magnetic data, the transfer of one electron to the HATN in the ground state was assumed. This attempt is fully documented and its severe limitations are discussed in the Supporting Information (SI‐2).</p><p>While this assignment seemed plausible at first sight, it can not provide a full explanation for the spectra of the oxidized species, i. e. there is disappearance of the expected CT bands in experimental NIR spectra. For the HATN complexes, this is not surprising because the applicability of the single electron picture and the appropriate definition of the ground state require more detailed quantum chemical calculation as provided in Sections 2.3. to 2.5. of this paper. For [(Cp2Ti)3HATN(Ph)6] studied here, one must suspect appearance of multiple spin states. Those scenarios prevent a meaningful applicability of a single electron picture in addition to the anyway very crude assumptions required to condense electrochemical and spectroscopic data in one picture. Therefore, we strongly discourage to derive theoretical quantities such as orbital energies from experimental data in case high quality quantum chemical calculations are available.</p><!><p>In order to understand the ground state configuration and electronic excitations, the electronic structure of the HATN molecule must first be examined more closely without coordinated metal ions. Due to its D3h symmetry two degenerate HOMOs can be found (Figure 5a). In the literature, a large electron affinity of the HATN ligand is reported. [14] This can also be shown in first approximation by a PBE0/Def2‐SVP/RIJCOSX calculation of the charged species (−1, −2 and −3. From those data, the electron affinities E af in Table 2 were obtained by [Eq. (1)]:(1)EAf=Echarged-Eneutral,</p><!><p>Electron affinities of HATN (no metals coordinated) in different charge states. The energetically lowest lying spin state with no spin contamination were used for comparison.</p><p>Charge Q|Spin S</p><p>Energy [a.u.]</p><p>E af [eV]</p><p>0|0</p><p>−1248.008</p><p>0.00</p><p>−1|1/2</p><p>−1248.070</p><p>−1.69</p><p>−2|0</p><p>−1248.014</p><p>−0.15</p><p>−3|3/2</p><p>−1247.836</p><p>4.69</p><!><p>where E charged and E neutral represent the energies of the charged species and of the neutral molecule, respectively.</p><p>Each Cp2Ti has up to two unpaired electrons that can be transferred while bonding with the HATN molecule (Figure 5b). The electronic structure of Cp2Ti complexes is detailed investigated in literature. [27] The data in Table 2 and the fact that Cp2Ti is known as a strongly reducing agent make it plausible that a charge transfer from Cp2Ti to the HATN ligand is present in the electronic ground state of [(Cp2Ti)3HATN(Ph)6]. Generally, the bent titanocene(II) fragment is often used in the preparative chemistry in order to introduce electron transfer processes as starting point for a broad range of subsequent reactions. [28] In mononuclear 2,2‐bipyridine titanocene complexes, the chelating ligand is well known as redox active. [29] According to our calculations up to three electrons might be transferred. Although, the transfer of three electrons to the isolated ligand is energetically disfavored, stabilization by Coulomb forces from Cp2Ti leaves this transfer as a valid option.</p><p>Calculation of spectra of the charged HATN molecule with no metal coordinated within time‐dependent density functional theory (TD‐DFT) give first insights into the excitations observed in the experimental NIR spectra. When the HATN molecule is charged, the calculation reproduces similar electronic excitations as found for the [(Cp2Ti)3HATN(Ph)6] in the NIR range (Figure 6). For this reason, the excitations observed here may not correspond to an IVCT or an LMCT, but to an internal ligand excitation. The Ti cations may provide only a charged "frame" for the π electronic system of the bridging ligand.</p><!><p>Calculated absorption spectra of HATN ligand on TD‐PBE0/Def2‐SVP level with different charges (Q) and spin (S).</p><!><p>This interpretation has precedence in literature. Moilanen et al. [9] reported about [(HAN){Mg(N,N'‐(2,6‐diisopropylphenyl)‐3,5‐dimethyldiketiminate)}3]⋅toluene complex reduced by [Mg(N,N'‐(2,6‐diisopropylphenyl)‐3,5‐dimethyldiketiminate)2]. In this compound three electrons were transferred to the ligand and caused absorptions at 800 nm and 920 nm. Calculation on the TD−B3LYP level confirmed that the observed spectra resulted from transitions from the highest occupied molecular orbital (HOMO) and a single occupied molecular orbital (SOMO) to the lowest unoccupied molecular orbital (LUMO). In this model compound the assignment was clear because no transition metals and d‐orbitals were involved in the coordination. However, up to six electrons of three TiCp2 complexes are available for a possible charge transfer in [(Cp2Ti)3HATN(Ph)6]. This significantly increases the complexity of the system comared to the system studied by Moilanen et al. [9]</p><p>In order to investigate the ground state of [(Cp2Ti)3HATN(Ph)6] more precisely, four spin multiplicities (S=0, 1, 2, 3) were calculated for the neutral species with PBE0. All multiplicities except S=3 show a strong spin contamination (SI‐3, Table S3). The spin states S={0, 1, 2} are also energetically degenerate. The mixing of higher spin states leads to assumption of a strong multireference character produced by the titanium centers, which is reasonable for such a highly symmetrical system. As reported in the literature, high spin contamination of the ground state wavefunction can lead to even higher spin contamination in the excited states and therefore lead to prediction of unphysical excitations. [30] A single reference method should therefore not be able to describe such a complex accurately.</p><p>However, the experimental spectrum of [(Cp2Ti)3HATN(Ph)6] can still be calculated with sufficient accuracy using TD‐DFT in the spin states S=1 and S=2. Also, the spin states S=0 and S=3 are energetically very close to each other but show no electronic excitations in the NIR range. Table 3 shows that the experimental absorption signals are in exellent agreement with the theoretical simulations from an energetical perspective. Detailed discussion of the spectra will be provided in Section 2.4.</p><!><p>NIR electronic excitations of measured spectrum and TD‐PBE0 excitations of [(Cp2Ti)3HATN(Ph)6].</p><p>Signal</p><p>Absorption energy [eV]</p><p></p><p>Experiment</p><p>TD‐PBE0 S=1</p><p>TD‐PBE0 S=2</p><p>0/I</p><p>0.60</p><p>0.56</p><p>0.55</p><p>0/II</p><p>0.95</p><p>0.92</p><p>0.93</p><p>0/IV</p><p>1.13</p><p>1.04</p><p>1.05</p><p>0/V</p><p>1.27</p><p>1.27</p><p>1.26</p><!><p>A correct description of electronic excitations with a spin‐contaminated wave function has already been observed in the literature. [31] Nevertheless, the results of DFT shown here should be interpreted with reservations due to the high spin contamination of the system.</p><p>Looking at the structure of the orbitals involved in the excitation process (SI‐3, Figure S5), it becomes clear that only ligand‐ligand excitations are observed. As hypothesized above, a charge transfer from the titanium atoms to the ligand takes place in the ground state of [(Cp2Ti)3HATN(Ph)6], i. e. before any electronic excitation. Consequently, the titanium atoms are not involved in the electronic excitation process, but provide a charged frame for the excitation processes on the HATN(PH)6 ligand. Despite the good agreement of the theoretical excitations with experiment, the results are affected by spin degeneracy and spin contamination and therefore should be interpreted with caution. Both attributes indicate an intrinsic multi‐reference character. This character is caused by the titanium atoms and is not required to reproduce the observed NIR spectrum (cf. Section 2.5).</p><!><p>Figure 7 shows, besides the NIR excitation of the charge state 0, the oxidized forms of the complex (+1, +2) in different spin states and the comparison to the experimental results. For all species at least three different multiplicities and their TD‐PBE0 spectra were calculated using the unrestricted Kohn‐Sham ansatz (see SI‐3, Table S3). The most important states were compared with experiment. Due to the high complexity of the system, no anionic species were computed in this work.</p><!><p>Comparison of experimental and theoretical NIR spectra for different charge states (0, +1, +2). Color scheme for multiplicities are orange: S=1/2, blue: S=1, green S=2, violet: S=3/2. Experimental data are referenced on the highest peak in the charge state +1. Theoretical species are normalized separately per redox species for a better comparison with experimental data. Transition dipole moments can be found in supporting information (SI‐3, Tables S3–S5).</p><!><p>As mentioned before, the excitations of the neutral species from the experiment are well reflected by the TD‐DFT calculations. In both spin states (S=1 and S=2) it is possible to simulate the excitations observed in experiment. Only the broad signal around 2100 nm in Figure 7 cannot be reproduced in terms of intensity. Many factors can be considered for this behavior. For instance, the transition dipole moment depends on the selected functional and could be poorly described by PBE0. In addition, the spin contamination is highly significant as mentioned in Section 2.3., and the single‐reference approach cannot completely describe the real system.</p><p>However, a purely energetic consideration of the TD‐PBE0 spectra in comparison with experimental spectra leads to an excellent agreement. Besides that, a comparison of the different spin states shows quasi‐degenerate energies (SI‐3, Table S2). This observation suggests a multi‐reference character of the system, which will be discussed in more detail in Section 2.5. Also, the states S=0 and S=3 are energetically very close to each other but show no electronic excitations in the NIR range.</p><p>An examination of the redox species +1 shows that it is only possible to simulate the measured NIR spectrum if both, the S=1/2 (orange) and the S=3/2 (violet) spin states, are calculated and both spectra are overlaid. Similar to the neutral redox species, the total energies of both spin states are quasi‐degenerate and spin contamination is observed. This behavior also suggests a multi‐reference character. However, the transition energies of the theoretical excitation spectra correspond very well to the experimental data. This clearly shows that the measured results are only accessible by a combination of different spin states within the TD‐DFT framework. Only partial aspects of the electronic properties can be described with the single reference approach, which manifests itself in relevance of different spin states for the explanation of the observed UV/Vis‐spectra. A divergence in intensities, as observed in the neutral species, is also apparent. The charge species +2 shows no significant signals in experiment in Figure 4. No electronic excitation in the NIR range can be found in the theoretical calculations with the S=0 state. Only the spins S=1 and S=2 lead to excitations at ∼2000 nm in the theoretical studies. Similar to the neutral species, the first three spin states are energetically degenerate (SI‐3, Table S2). Again, the system is subject to a strong multi‐reference character. The signals seen in the calculations at about ∼2000 nm of the higher multiplicities cannot be recorded experimentally for two reasons: Firstly, the electronic excitations could be outside the accessible spectral range of the spectrometer if a deviation of 200 nm is assumed between calculated and experimental values. Secondly, another aspect is the poor description of a symmetric open shell system by a single‐reference method.</p><p>Nevertheless, the DFT can qualitatively present the electronic excitations in the NIR range, in which the Ti orbitals are not involved. However, this approach is not sufficient for a more detailed investigation of the system, since the spin degeneracy, which is triggered by a multi‐reference character, cannot be described well. The observations within TD‐DFT suggest that the multi‐reference character and the spin problem are probably located at the titanium atoms, since they are not involved in the excitations. Additionally, further complications can arise from the charged HATN ligand because of degenerate orbitals (Figure 5a). Therefore, the system will be treated within a multi reference framework below.</p><!><p>A key conclusion from Sections 2.3. and 2.4. is that the electronic transitions correspond to electronic excitations of the ligand and neither IVCT nor LMCT play a role in the electronic excitation. This also requires the ligand to accept electrons from the formally divalent TiII centers in the ground state. In order to explain how many electrons are transferred and which magnetic properties will result from such transfer, multi‐reference methods beyond DFT have to be used. In order to reduce the computational effort, a simplified model system (Figure 8) with similar electronic features is used for a CASSCF/NEVPT2 simulation with a CAS(6,7) space and a larger Def2‐TZVP basis set. Here, the Cp ligands are replaced by chlorine atoms and the phenyl groups of HATN(PH)6 are exchanged by hydrogen atoms. This approach was already used successfully in the literature for other systems. [32] In order to validate the justification of the simpler model system a direct comparison of the TD‐DFT spectra with the original complex is shown in SI‐3, Figure S6.</p><!><p>Simplified [(TiCl2)3HATN] complex. Color scheme of atoms is titanium: red, chlorine: green, carbon: brown, nitrogen: blue, hydrogen: white.</p><!><p>Magnetic measurements have shown that up to three unpaired electrons with parallel spin can be present in the system. [12] The calculations in Section 2.3. demonstrate that the charged HATN ligand without coordinated metals can reproduce the NIR electronic excitation provided that up to three electrons are transferred from the metal framework to the ligand in the ground state. For such a transfer to the HATN ligand, six electrons of the three formally titanium(II) centers are available.</p><p>A proper description of the multi‐reference character and the electronic excitation in the NIR range requires at least a (6,7) space for this system. Starting from Q=0, S=3 pbe0 wavefunction, it can be observed that three singly occupied titanium d‐orbitals and three singly occupied π orbitals arise. Therefore, a CAS(6,6) calculation provides a good ansatz for the ground state. According to the D3h point group symmetry of the system, two energetically degenerate states are formed, which consist of a variety of configuration state functions. In order to include possible electronic excitations, another virtual orbital is taken into the active space. A further unoccupied ligand orbital is found to be necessary for the electronic excitation resulting in a CAS(6,7) active space. If the individual states are reduced to the dominant determinants and Löwdin Reduced Orbitals are formed, a formal description can be made in a one‐electron picture with fractional occupation numbers (Figure 9).</p><!><p>Visualization of molecular orbitals in a canonical representation and its occupation within CAS(6,7) for Q=0, S=1 as representative state. Color scheme of atoms is titanium: red, chlorine: green, carbon: brown, nitrogen: blue, hydrogen: white. Orbitals are labeled similar to calculated output.</p><!><p>A total of five orbitals are occupied in the electronic ground state, where three singly occupied orbitals (182, 183, 184) are strongly localized at the titanium atoms. Additionally, one singly occupied (181) and a doubly occupied (180) orbital is found at the HATN ligand. The doubly occupied HATN ligand orbital (180) exhibits a population of 1.68 e− and delivers electron density to a higher lying HATN ligand orbital (185). A comparison of the populated HATN ligand orbitals with the orbitals of a free HATN molecule shows that the LUMO and LUMO+1 orbitals were occupied after the charge‐transfer from the metal centers (SI‐3, Figure S4 and Figure 9). The titanium atoms share three electrons in this picture which are distributed over all three centers. These electrons are responsible for the strong multi‐reference character and lead to a spin degeneracy reminiscent to the single reference calculations (SI‐3, Table S6–S7). These results can also give first insights into the magnetic measurements of the unpaired electrons. [12] Furthermore, it can be concluded that three electrons were donated to the HATN ligand in the ground state (before electronic excitation) and that there are on average three equivalent TiIII centers. This observation could be verified with CAS(12,13) and a fully geometry‐symmetrized model system which leads to similar results even when trying to rotate more Ti‐centered molecular orbitals in the active space (SI‐3, Table S8 and S15).</p><p>The electronic excitation in the framework of the multi‐reference picture displays a similar behavior as the single reference TD‐DFT spectra. With both methods, the NIR region is solely dominated by internal HATN ligand orbitals, and the electrons localized at the titanium atoms are not involved in any excitation. However, a specific transition cannot always be interpreted with a one‐electron picture. For this reason, an attempt is made to use the dominant configuration state functions to obtain an approximate one‐electron image. Table 4 shows the respective excitation energies at CAS(6,7)/NEVPT2 level compared to the TD‐PBE0 energies. In addition, the dominant configurations of the individual states at each transition are shown. Note, that the shown energies for CAS(6,7)/NEVPT2 are provided by the optimized ground state wavefunction and no state averaging or state optimizing of the excited states was performed due to the high complexity of configurations (SI‐5, Table S16).</p><!><p>Electronic excitations with TD‐PBE0, CAS(6.7)/NEVPT2 (Q=0, S=1) and the deviation of both methods. The right columns represent the dominant transitions of the excited states of CAS(6,7) calculations. A full configuration of the excited states is provided in supporting information (SI‐5, Tables S16 and S17).</p><p>Excitation energies [eV]</p><p>Dominant transitions</p><p>TD‐PBE0</p><p>CAS(6,7)/ NEVPT2</p><p>Deviation [eV]</p><p>orbital x</p><p>→</p><p>orbital Y</p><p>0.57</p><p>0.43</p><p>0.14</p><p>180</p><p>→</p><p>181</p><p>0.89</p><p>0.74</p><p>0.15</p><p>180</p><p>→</p><p>185</p><p>1.15</p><p>1.28</p><p>−0.13</p><p>180</p><p>2x→</p><p>185</p><p>1.50</p><p>1.65</p><p>−0.15</p><p>181</p><p>→</p><p>186</p><!><p>Obviously, the excitation energies agree reasonably well between both methods. Three of the most important excitations can be represented by a single transition. The energetically lowest (0.44 eV) can be interpreted as an excitation from the almost doubly occupied orbital 180 (1.68 e−) to the singly occupied orbital 181 (1.00 e−). At 0.77 eV the dominant configuration is a transition from orbital 180 to orbital 185 (0.32 e−), which is already slightly occupied. The energetic largest transition at 1.65 eV can be interpreted as an excitation from orbital 181 (1.00 e−) into the unoccupied orbital 186. The signal at 1.28 eV can no longer be interpreted within a simple one‐electron excitation in the orbital picture, since the dominant configuration state functions (only ∼10 %) represent a double transition from orbital 180 (1.68 e−) to orbital 185 (0.32 e−). The state shows a large multi‐reference character with up to seven dominant determinants with over 5 % weight. The corresponding configuration is listed in SI‐5, Table S16. Please note, that a double excitation cannot be calculated with TD‐DFT and a direct comparison of this excitation can only be made with reservation.</p><!><p>Three reduction waves and six oxidation waves are observed in voltammograms of [(Cp2Ti)3HATN(Ph)6]. All electron transfers are one electron processes clearly separated on the potential scale. The solution spectra of [(Cp2Ti)3HATN(Ph)6] and its oxidation and reduction products [(Cp2Ti)3HATN(Ph)6]3−, [(Cp2Ti)3HATN(Ph)6]2−, [(Cp2Ti)3HATN(Ph)6]1−, [(Cp2Ti)3HATN(Ph)6]1+ and [(Cp2Ti)3HATN(Ph)6]2+ show electronic transitions in the UV, vis and NIR spectral ranges. DFT calculations on TD‐DFT and CASSCF/NEVPT2 levels show that three electrons are transferred from the formally titanium(II) centers to the bridging ligand in the ground state. However, the complete spectra can only be explained when assuming that spin states of different multiplicity form the ground state of [(Cp2Ti)3HATN(Ph)6]. The observed electronic excitations in the UV, Vis and NIR ranges are all transitions between ligand levels. Contrary to common notion, LMCT or IVCT transitions do not play a role. The same assignment rules can also be reproduced by calculating the isolated and appropriately charged HATN ligand. The temperature‐dependent magnetic properties can be explained the hybridization of different electronic configurations of different total spin with one valence electron located at each of the three Ti centers and one electron in a SOMO of the ligand.</p><!><p>[(Cp2Ti)3HATN(Ph)6] was synthesized as reported. [12] All used solutions were freshly prepared from thoroughly dried solvents. The electrolytes were made with tetrabutylammonium hexafluorophosphate (TBAPF6), Sigma Aldrich, Steinheim, Germany), silver perchlorate (AgClO4, Sigma Aldrich). Acetonitrile (MeCN) and tetrahydrofurane (THF) were dried according to standard procedures [33] prior to use and stored under nitrogen. [(Cp2Ti)3HATN(Ph)6] was solved in dry THF to a concentration of 0.3 mM.</p><!><p>Electrochemical and spectroscopic setups have been installed in a glovebox and connected to instruments outside the box via gas‐tight integration of cables and optical fibers into a flange connector (M. Braun). Further details are given in SI‐1. NIR spectra were measured with a fiber‐coupled Matrix‐F FT‐NIR spectrometer (Bruker Optik GmbH, Ettlingen, Germany). UV‐Vis spectra were taken with a GetSpec 2048CCD array spectrometer (GetSpec, Sofia, Bulgaria). Both spectrometers were coupled to a spectroelectrochemical cell (ALS Co., LTD, Tokyo, Japan) located inside an Ar‐filled glove box using 600 μm diameter N227 quartz fibers (Bruker Optik GmbH).</p><p>Cyclic voltammograms (CV) and differential pulse voltammograms (DPV) were recorded at 295 K using a potentiostat (Compactstat, Ivium Technologies, Eindhoven, The Netherlands) with a three‐electrode assembly consisting of an Au disc as working electrode (WE, diameter d=2 mm), a Pt plate as auxiliary electrode (Aux, A=1 cm2) and a Ag/Ag+ reference electrode filled with 0.01 M AgClO4 and 0.1 M TBClO4 in THF.</p><p>The formal potentials E°' were obtained as the arithmetic mean of the anodic E pa and cathodic peak potentials E pc. They are quoted with respect to the ferrocene/ferrocenium redox couple (Fc/Fc+) which was measured in solution with same concentration of supporting electrolyte and the same concentration of the Fc (0.3 mM) as the compound under investigation.</p><!><p>All calculations were performed using the program package ORCA in version 4.2. [34] The basis set Def2‐SVP is taken in all simulations with the [(Cp2Ti)3HATN(Ph)6] complex and the pure HATN ligand. [35] The simplified model system is calculated using a Def2‐TZVP basis set. For the calculation with the single‐reference properties, the hybrid density functional PBE0 an unrestricted open‐shell ansatz was selected. Due to the large number of electrons, the Resolution‐of‐Identity formalism RIJCOSX was used with a Def2/J auxiliary basis set as implemented in ORCA. [36] For the respective redox species, four different spin multiplicities were geometry‐optimized and subsequently TD‐DFT calculations for the UV/Vis spectra were performed. For a verification of these spectra, a CAS(6,7)/NEVPT2 and a CAS(12,13) calculation for the smaller model system was carried out. Due to the large basis set, RIJCOSX with a Def2/JK auxiliary basis set was selected. The calculated wavefunction for CAS(6,7) was state‐averaged over the first two states to generate the correct degenerate ground state. No state optimizations for the excited states were performed due to the high complexity of the system.</p><!><p>The authors declare no conflict of interest.</p><!><p>As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.</p><p>Supplementary</p><p>Click here for additional data file.</p>

PubMed Open Access

Vinblastine 20\xe2\x80\x99 Amides: Synthetic Analogs That Maintain or Improve Potency and Simultaneously Overcome Pgp-derived Efflux and Resistance

A series of 180 vinblastine 20\xe2\x80\x99 amides were prepared in three steps from commercially available starting materials, systematically exploring a typically inaccessible site in the molecule enlisting a powerful functionalization strategy. Clear structure\xe2\x80\x93activity relationships and a structural model were developed in the studies which provided many such 20\xe2\x80\x99 amides that exhibit substantial and some even remarkable enhancements in potency, many that exhibit further improvements in activity against a Pgp overexpressing resistant cancer cell line, and an important subset of the vinblastine analogs that display little or no differential in activity against a matched pair of vinblastine sensitive and resistant (Pgp overexpressing) cell lines. The improvements in potency directly correlated with target tubulin binding affinity and the reduction in differential functional activity against the sensitive and Pgp overexpressing resistant cell lines was found to correlate directly with an impact on Pgp-derived efflux.

vinblastine_20\xe2\x80\x99_amides:_synthetic_analogs_that_maintain_or_improve_potency_and_simultaneo

Introduction<!>Synthesis of Vinblastine C20\xe2\x80\x99 Amides<!>Biological Activity<!>Alkyl 20\xe2\x80\x99 Amides<!>Benzoyl 20\xe2\x80\x99 Amides<!>20\xe2\x80\x99 Acrylamides<!>Polycyclic Benzoyl-like 20\xe2\x80\x99 Amides<!>Monocyclic Heterocyclic 20\xe2\x80\x99 Amides<!>Polycyclic Heterocyclic 20\xe2\x80\x99 Amides<!>C20\xe2\x80\x99 Sulfonamides<!>Key 10\xe2\x80\x99-Fluorovinblastine 20\xe2\x80\x99 Amides<!>Additional Assessments of Compounds 28, 121 and 173<!>Models of 28, 121 and 173 Bound to Tubulin<!>Conclusions<!>General Chemistry Procedures<!>Method 1<!>Method 2<!>Method 3<!>Compound 28<!>Compound 121<!>Compound 173<!>Ritter Reaction used to generate 20\xe2\x80\x99-acetamidoleurosidine (8)<!>Cell Growth Inhibition Assay<!>Tubulin Binding Competition Assay.43<!>Efflux and Permeability Assays

<p>The Vinca alkaloids constitute a family of natural products that continue to have a remarkable impact on anticancer drug discovery and treatment.1–3 Originally isolated as trace constituents of the Madagascar periwinkle plant (Catharanthus roseus (L.) G.Don),4,5 vinblastine (1) and vincristine are the most prominent members of this class and among the first plant-derived natural products used in the clinic for the treatment of cancer (Figure 1). These two compounds and three recent semi-synthetic analogs are integral oncology drugs employed today in highly successful combination drug therapies. Their mode of action, which involves disruption of microtubulin formation and dynamics through tubulin binding, still remains one of the most successful approaches for the development of oncology drugs.6</p><p>Vinblastine and vincristine are superb drugs even by today's standards. A major limitation to their continued use is the observation of clinical resistance mediated by overexpression of the drug efflux pump phosphoglycoprotein (Pgp).7 The identification of vinblastine analogs that might address such resistance, which also results in multidrug resistance (MDR) and is responsible for the majority of all relapses in oncology, has remained a major focus of the field for over 30 years. Not only would the discovery of a vinblastine analog not susceptible to Pgp efflux serve as a potential effective replacement for vinblastine in its current clinical uses or in instances of Pgp-derived vinblastine resistance, but it could also emerge as a new therapeutic option for other Pgp-derived MDR tumor treatments and constitute a major advance for oncology therapeutics. Despite the efforts focused on vinblastine for the past 40 years that have searched for analogs that effectively overcome Pgp-derived vinblastine resistance, little progress has been made.2 Recent advances in the total synthesis of vinblastine, vincristine and related natural products have provided access to analogs of the natural products not previously accessible by semisynthetic modification of the natural products themselves.8–19 The latest of these efforts have provided a powerful approach to access a variety of vinblastine analogs that contain systematic deep-seated modifications within either the lower vindoline-derived20–27 or upper catharanthine-derived11 subunits.28–30 As a result of these developments, we have prepared several series of key analogs, systematically exploring and defining the impact individual structural features and substituents have on tubulin binding affinity and cancer cell growth inhibition.10,19 Complementary to the studies detailed herein, we have systematically probed the impact and role of the vindoline C4 acetate,31,32 C5 ethyl substituent,33 C6–C7 double bond,34–36 and the vindoline core structure itself,36 and have systematically explored the upper catharanthine-derived subunit C20' ethyl substituent,37,38 C16' methyl ester,39 and added C10' or C12' indole substitutions.40 In addition and in preceding studies, we have shown that replacement of the C20'-OH with 20' ureas was possible,41 that substantial42 and even remarkable43 potency enhancements were obtainable with such 20' ureas, and that some exhibited further improvements in activity against vinblastine-resistant cancer cell lines.44 Herein, we report the important extension of these studies to the evaluation of vinblastine 20' amides conducted with the objective of discovering analogs that match or exceed the potency of vinblastine, but that are not subject to Pgp efflux and its derived vinblastine resistance. Not only did the studies provide vinblastine analogs no longer susceptible to Pgp-derived resistance, but they represent the discovery of a site and functionalization strategy for the preparation of now readily accessible vinblastine analogs (3 steps) that improve binding affinity to tubulin (on target affinity) and functional potency in cell-based assays while simultaneously disrupting their efflux by Pgp (off target affinity and source of resistance), offering a powerful approach to discover new, improved, and durable oncology drugs.</p><!><p>The extensive series of nearly 200 vinblastine analogs replacing the C20' alcohol with substituted or functionalized C20' amides was obtained through acylation of 20'-aminovinblastine (6), derived from reduction of the hydroazidation product 5, with either an acid chloride (Method 1, 2 equiv RCOCl, 4 equiv i-Pr2NEt, CH2Cl2, 23 °C, 2 h) or a carboxylic acid (Method 2, 2 equiv RCO2H, 4 equiv EDCI, 0.2 equiv DMAP, DMF, 23 °C, 12 h) (Scheme 1). The former procedure that uses an acid chloride generally provided the better yields and avoids the added purification challenges of removing residual coupling reagents from the reaction products.</p><p>Notably, the precursor azide 5 is available directly in a single step from commercially available vindoline (3) and catharanthine (2) by enlisting first their Fe(III)-promoted coupling (5 equiv FeCl3, 0.1 N aq HCl/CF3CH2OH, 25 °C, 2 h),28 which proceeds by single electron oxidative fragmentation of catharanthine and installs exclusively the C16' natural stereochemistry.29,30 Subsequent in situ Fe(III)-mediated free radical hydrogen atom transfer hydroazidation of anhydrovinblastine (10 equiv Fe2(ox)3, 20 equiv NaBH4, 30 equiv CsN3, aq HCl/CF3CH2OH, 0 °C, 30 min) provided 20'-azidovinblastine (5) directly as a mixture C20' diastereomers, but with exclusive control of the critical C16' stereochemistry.19,41 An X-ray structure determination conducted on the major diastereomer of the reaction revealed that it possessed the unnatural vinblastine C20' stereochemistry (leurosidine stereochemistry) and that the minor diastereomer of the reaction possessed the natural C20' vinblastine stereochemistry.19,41 This powerful hydrogen atom transfer initiated free radical reaction was developed to provide a general method for functionalization of alkenes with use of a wide range of free radical traps beyond O2 (air) used for vinblastine itself and was explored explicitly to provide the late-stage, divergent45 preparation of vinblastine analogs that bear altered C20' functionality at a site previously inaccessible for systematic exploration (Scheme 1). In addition to other free radical traps that were introduced that included azide, the broad alkene substrate scope was defined, the Markovnikov addition regioselectivity was established, the remarkable functional group tolerance was demonstrated, alternative Fe(III) salts and initiating hydride sources were shown to support the reaction, its underlying free radical reaction mechanism was defined, and mild reaction conditions (0–25 °C, 5–30 min) were developed that are remarkably forgiving to the reaction parameters.41,46</p><p>In the course of our examination the 20' amide analogs of vinblastine, we also reexamined a reported Ritter amidation reaction conducted on vinblastine or anhydrovinblastine, which was used to provide a limited series of 20' amides.47 Although this reaction was reported to proceed in very modest conversions (5–10%), our interest was in its disclosure as providing a single diastereomer that possesses the natural 20' vinblastine stereochemistry. By enlisting acetonitrile as the trap of the intermediate carbocation under conditions and in a reaction detailed in this work, we found that the reaction does indeed provide a single 20' diastereomer in low yield (<10%) as described (eq 1).47 However, the product 8 of the reaction from vinblastine (1) was found to possess the unnatural (leurosidine) 20' stereochemistry and was shown to be identical in all respects with an authentic sample of this unnatural 20' acetamide diastereomer. In retrospect, this is not surprising given the now recognized preference for α-face C20' addition, but was originally conducted at a time this knowledge and the powerful modern day characterization techniques were unavailable. The analogous reaction starting with anhydrovinblastine (7) failed to provide any acetamido product. Consequently, the work detailed herein, expanding on the three examples that we disclosed (20' formamide, acetamide and trifluoroacetamide; 9–11) along with the development of the vinblastine hydroazidation reaction,41 represent the only authentic 20' amides disclosed and examined to date.</p><p> (1)</p><!><p>As previously highlighted, a major limitation to the clinical use of vinblastine and vincristine is the observation of resistance mediated by overexpression of the drug efflux pump phosphoglycoprotein (Pgp). The identification of analogs that might address such resistance has remained a major focus for over 40 years and would represent a major advance for oncology therapeutics. With this objective in mind, all analogs we have prepared to date, including the 20' amides detailed herein, were screened simultaneously for growth inhibition activity against HCT116 (human colon cancer cell line) and a matched resistant cell line (HCT116/VM46) that is approximately 100-fold resistant by virtue of the overexpression of Pgp. This well-designed set of functional assays simultaneously provides a direct measure of both functional activity (HCT116) and the analog susceptibility to Pgp efflux (resistance, HCT116/VM46). Key members were assessed for tubulin binding affinity for correlation with functional activity and those that emerge as candidates that avoid Pgp efflux were examined in efflux assays to confirm their behavior toward Pgp and related efflux transporters. The cell growth inhibition activity against the L1210 (mouse leukemia) cell line was also measured and the results were qualitatively and quantitatively (IC50) nearly identical to those observed with the HCT116 cell line. Finally, the HCT116 human tumor cell line was found to accurately reflect activity observed against a larger panel of clinically relevant human tumor-derived cell lines with a class of closely related vinblastine 20' ureas.44</p><p>The results are presented and discussed below in groups of amides that embody related structural characteristics. Notably and importantly, the examined source of vinblastine resistance is not derived from changes in the drug target and their impact on drug binding (tubulin). Rather it is derived from binding and efflux by an off target protein (Pgp). These display distinguishable structure–activity relationships, one impacting potency (tubulin binding) and a second impacting resistance (Pgp binding and efflux). As important and as interesting as the former are in the discussions below, it is the discovery of a small subset of 20' amides that avoid Pgp efflux and overcome its derived resistance, displaying equal activity against both HCT116 and HCT116/VM46, that became a driving focus of our efforts.</p><!><p>Several important trends were observed with simple aliphatic 20' amides that influenced subsequent studies (Figure 2). First, the small series 9–11, examined at the time the 20' azidation reaction was generalized,41 revealed that the simplest of the aliphatic amides (9 and 10) reduced activity approximately 10-fold, that both 9 and 10 displayed an approximate 80-fold resistance with HCT116/VM46 similar to vinblastine, that both 9 and 10 were 10-fold more active than the corresponding free amine 6 (see Figure 1), and that the increased electron-withdrawing properties of the acyl group of the trifluoroacetamide 11 was further and significantly detrimental to compound potency. These trends continued to be observed throughout the expanded and systematic series of aliphatic amides summarized in Figure 2 with some important notable exceptions. Only the sterically most bulky aliphatic amides (14 and 22) were not tolerated and these led to further reductions in activity. Three of the smaller aliphatic amides (17, 18 and 23) uniquely matched the potency of vinblastine in the vinblastine-sensitive L1210 and HCT116 cell lines. These three compounds displayed activity that was improved over both the small or larger aliphatic amides, and two (17 and 18) improved (reduced) the differential activity between the vinblastine sensitive and resistant HCT116 cell lines. Most significantly, a well-defined trend was observed among all the 20' amides against the vinblastine-resistant HCT116/VM46 cell line. The 20' amides with the more hydrophobic substituents displayed a smaller differential in activity between the vinblastine sensitive and resistant cell lines (ratio = HCT116-VM46/HCT116), indicating less effective Pgp efflux in the resistant cell line. Moreover, this differential in activity smoothly and progressively diminishes as the hydrophobic nature of the substituent increased (ratio: R = Me > Et > i-Pr, c-Pr, c-Bu > c-pentyl > c-hexyl > benzyl > n-heptyl). Finally, the acrylamide 23 matched, but did not exceed the activity of vinblastine. It proved to be more active than most, but not all of the simple aliphatic 20'amides and possesses the potential for covalent capture at a tubulin binding site. However, we have found no evidence of such behavior and, as detailed below, have observed improved activity with substituted acrylamides less prone to putative covalent capture.</p><!><p>The initial exploration of aryl versus aliphatic 20' amides led to analogs displaying activity that merited a detailed and systematic examination of such compounds (Figure 3). The parent unsubstituted 20' benzamide 24 proved to be >5-fold more potent than vinblastine and >50-fold more potent than the saturated cyclohexyl counterpart 20. But like 20, compound 24 also displayed a smaller differential in activity between the sensitive and resistant HCT116 cell lines than vinblastine (25-fold vs 88-fold), indicating it also embodied characteristics that make it a less effective substrate for Pgp. Early in the studies detailed herein, we first prepared a small but key series of 4-substituted benzamides to probe the electronic impact of substituents as well as several related analogs to establish sites amenable to substitution. These studies revealed that both 4- and 3-substitution of the phenyl ring were well tolerated and that any given substituent provided nearly equivalent activity when placed at either site (4- (para) > 3- (meta)), but that o-substitution reduced the relative potency by 10-fold or more (potency: p- > m- ≫ o-substitution). In addition, the initial 4-substituted benzamides examined were found to display a well-defined trend in which electron-donating substituents improved potency, whereas electron-withdrawing substituents reduced activity. Plots of the substituent Hammett σp constants versus –log IC50 (nM) revealed a linear relationship with slopes (ρ value) of –1.04 (HCT116) and –1.20 (L1210), indicating a large and remarkably well defined electronic contribution to the behavior of the analogs (Figure 4).</p><p>Throughout our studies, periodic measurements of relative tubulin binding affinities established that the substituent effects on activity correlated with relative target tubulin binding affinities. In initial studies on the 20' benzamide substituent effects detailed herein, three derivatives (24, 54 and 64; R = H, CN, and NMe2, respectively) from the Hammett plot set were examined (Figure 5). Consistent with their relative potencies, the three derivatives displayed the same clear trends in their ability to displace tubulin bound BODIPY-vinblastine (64 > 24 > 54). This direct correlation of functional cell growth inhibition activity with target tubulin binding affinity and the relative magnitude of the effects indicate that the properties of these C20' amides are derived predominately, if not exclusively, from on target effects on tubulin. Retrospective modeling of key analogs bound to tubulin detailed later herein suggest that this pronounced effect arises from the electronic impact of the substituent on the Lewis basicity of the amide carbonyl, enhancing its ability to serve as a H-bond acceptor for the backbone NH of β-tubulin Tyr224. Notably, this complements the requisite H-bond donor property of the secondary 20' amides (tertiary amides are inactive) in which the secondary amide NH mimics the tertiary alcohol of vinblastine itself and H-bonds to the backbone amide carbonyl of Pro222.</p><p>Finally, the early studies revealed that further increasing the hydrophobic character of the benzamide generally reduced the differential in activity between the sensitive and resistant HCT116 cell lines (e.g. compare 24 and 25–29). This paradox of increasing potency through addition of a typically polar electron-donating substituent, enhancing target tubulin binding, while simultaneously disrupting Pgp transport by further decreasing the polarity of the 20' benzamide is chronicled in the subsequent extensive studies detailed in Figure 3.</p><p>Within this series, there are several vinblastine analogs that display stunning enhancements in potencies (e.g. 57, 60–62, 64, 67, 70, 71 and 73), others that display substantially improved potencies (10-fold) and attractive reduced differentials in activity (<10-fold; e.g. 25, 26, 30, 72, 74, 97 and 99), and many that display substantially improved differentials in activity (<10-fold). There are even those that indicate surprisingly large p-substituents are well tolerated (e.g. t-Bu in 38). Many of these would be attractive analogs of vinblastine for further study. For us and the prescribed objective of discovering analogs that match or exceed the potency of vinblastine, but which are not subject to Pgp efflux derived resistance (ratio differential <2-fold), it is the analogs 28, 29, 32, 78, 92, and 98 that met these defined parameters. Of these, it was 28 that was selected for additional study.</p><!><p>A small series of substituted 20' acrylamides was prepared and examined in part for comparison with the unsubstituted acrylamide 23 (Figure 6). Addition of an aryl group to the terminus of the acrylamide was found to provide vinblastine analogs as much as 10-fold more potent than either 23 or vinblastine itself. Although the number of comparisons is small, the differential in activity between the vinblastine-sensitive HCT116 and vinblastine-resistant HCT116/VM46 cell lines decreased with the increased hydrophobic character of the aryl substitute (Pyr > furanyl, thienyl > Ph). This proved consistent with the observations made with the 20' benzoyl amides (see Figure 3) where increased hydrophobic character reduced the activity differential. In fact, the activity of the benzoyl amide 24 proved essentially indistinguishable from the phenyl substituted acrylamide 108, both in terms of their potency and this activity differential. As indicated earlier, we found no evidence of covalent capture at a tubulin binding site with 23 or 108–114 and the improved activity with the substituted acrylamides less prone to putative covalent capture is consistent with enhanced tubulin binding affinity derived simply through non-covalent interactions.</p><!><p>An important series of benzoyl-like 20' amides was examined that contained additional rings fused to the aromatic core (Figure 7). In essence, these represent variations on the 1- or 2-naphthyl amides 115 and 116. Although the 1-naphthyl amide 115 was found to be roughly 10-fold less potent than vinblastine, the 2-naphthyl 20' amide 116 was determined to be approximately 3-fold more potent. Most significantly, the activity of 116 against the resistant HCT116/VM46 cell line was substantially improved such that the differential in activity versus HCT116 was less than 3-fold, indicating that it is no longer effectively subject to Pgp efflux-derived resistance. Although the sensitivity of HCT116 toward 116 was not significantly improved, the improvement in activity against HCT116/VM46 was suggestive that it is no longer a substrate for Pgp efflux. In this series and like the 20' benzamide series, polar electron-donating substituents in conjugation with the amide carbonyl often improved activity (e.g. 117) but did so at the expense of the differential potency against the sensitive and resistant HCT116 cell lines. The amides in which the carbonyl was attached directly to the aryl ring were more effective than the bicyclic systems attached at an aliphatic site (e.g. 121 and 124 vs 122 and 125). In general and consistent with expectations, the derivatives with the greater hydrophobic character led to reduced differentials in activity between the sensitive and resistant HCT116 cell lines. The amides bearing the saturated fused six- or five-membered rings (121 and 124) exhibited the unique combination of slightly improved potency relative to vinblastine (ca. 2-fold) and little differential activity (<2-fold) and proved to be essentially indistinguishable from 28 (3,4-dimethylbenzoylamide). These compounds are similar in activity to the parent benzamide 24, but with an even better improvement in activity against the resistant HCT116/VM116 cell line. As detailed earlier and like 28, 121 exhibited a profile of activity that we were hoping to discover at the start of our studies and both became key compounds that were further profiled.</p><!><p>A systematic series of 20' amides were examined that contain a single heterocyclic ring (Figure 8). In general, the potency of the series against the sensitive cell lines followed trends in which the more hydrophobic and more electron-rich heterocyclic amides displayed the greatest activity (furanyl, thienyl > oxazolyl, thiazoyl, isoxazolyl > imidazoyl, pyrazinyl, pyridazinyl). The exceptions to this generalization are the 4- and 3-pyridyl amides 128 and 129 that proved to be among the most potent analogs in this series despite their polarity and electron-deficient character. Similarly, the differential in activity against the sensitive and resistant HCT116 cell lines also generally increased as the polarity or heteroatom count in the heterocycle increased (furanyl, thienyl < oxazolyl, thiazolyl, isoxazolyl < imidazoyl, pyrazinyl, pyridazinyl, pyridyl). Within this series, impressive potency was observed with the 3-furanyl and 2-thienyl amides (133 and 136), displaying activity (IC50 = 600–700 pM) 10-fold greater than vinblastine and roughly 2-fold better than the 20' benzamide 24 with additional improved reductions in the differential activity (11-fold vs 88-fold and 25-fold) for the sensitive and resistant HCT116 cell lines. Interestingly, the 4- and 3-pyridyl amides 128 and 129 were among the most potent compounds in the series (IC50 = 400–700 pM), whereas the 2-pyridyl amide 130 was among the least potent. However, both 128 and 129 displayed the largest differential in activity against the sensitive and resistant HCT116 cell lines (138-fold and 83-fold, respectively).</p><p>The saturated heterocyclic amides 144–148 proved to be much less potent than vinblastine and less potent than most of the aromatic heterocyclic 20' amides. In the small series examined, the compounds appear to follow trends where the more polar substituents not only led to progressive losses in activity, but also increase the differential in activity between the sensitive and resistant HCT116 cell lines. An instructive comparison is the activity of 144 versus 20 in which a polar oxygen atom is introduced into the all carbon six-membered ring. Although the two compounds proved nearly equipotent against the sensitive cell lines, 144 proved to be much less active in the resistant HCT116 cell line, displaying a differential in activity (100-fold) similar to that of vinblastine (88-fold) and much greater than 20 (11-fold).</p><!><p>An important series of heterocyclic amides that contain two fused aromatic or non-aromatic rings was examined that also provided important insights into structural features that can enhance potency or disrupt Pgp-derived resistance. These are presented in Figure 9. The trends observed with incorporation of a series of polycyclic heterocycles, which constitute benzo-fused versions of the monocyclic 20' heteroaromatic amides (Figure 8), are instructive. First and foremost, the differential in activity between the sensitive and resistant HCT116 cell lines for each benzo-fused heterocycle generally improved (diminished) relative to its companion monocyclic heterocycle. This behavior is analogous to the comparison of benzamide 24 and its benzo-fused counterpart 116 (2-naphthyl amide) and likely reflects the increased hydrophobic character of the benzo-fused heterocycle. For the heterocycles containing a basic nitrogen, acyl linkage at the site adjacent to the basic nitrogen with 152, 156, 169 and 172 (2-quinolyl, 2-quinoxalyl, 2-benzoxazolyl, 2-benzthiazolyl) led to substantial reductions in activity, an observation analogous to that made with the pyridyl series 128–130. It is likely that this adjacent heterocyclic nitrogen participates in an intramolecular H-bond to the proximal amide NH, disrupting the stabilizing intermolecular amide N–H H-bond with the backbone carbonyl of Pro222. Otherwise, the heterocycle acyl linkage site proved remarkably flexible tolerating most possibilities. Notably, derivatives were not generally examined that might represent the distinguishing linkage sites of 2- vs 1-naphthyl where the linkage site is adjacent to a ring fusion center and found to be detrimental. However, the productive activity maintained by the 3-benzofuranyl and 3-benzthiophenyl amides 159 and 164 suggest such linkages should not be ruled out. Finally, the potencies of many of the heterocyclic amides were found to be superb with nearly all exceeding the activity of vinblastine. The potency comparisons of each benzo-fused heterocycle with its companion monocyclic heterocyclic amide was more variable although most were found to maintain or even improve on this activity. These generalizations are perhaps best depicted in comparing the quinolyl and isoquinolyl amide series 149–154 with the pyridyl series 128–130. These maintained the exceptional potency (IC50 = 400–700 pM) observed in the pyridyl series (IC50 = 400–600 pM), also exhibited a substantial loss in activity when the acyl substitution site was ortho to the basic nitrogen (ca. 100-fold, 152), and exhibited a substantially improved (diminished) differential in activity between the sensitive and resistance HCT116 cell lines for the potent variants (ca. 10-fold vs 100-fold). Finally and within this series, the 6-benzfuranyl and 6-benzthiophenyl amides (157 and 162) displayed activity (IC50 = 2–3 nM) 2- to 3-fold greater than vinblastine with additional and superb improved reductions in the differential activity for the sensitive and resistant HCT116 cell lines (3-fold and 2.7-fold vs 88-fold) and proved very comparable to the 2-naphthyl amide 116.</p><p>Even more significant, a series polycyclic heterocyclic 20' amides was examined in which the fused heterocycle was saturated versus aromatic (161 and 173–176). In essence, those examined represent benzoyl amides substituted with a para electron-donating substituent incorporated into a fused ring system, some of which introduce more lipophilic character. In general, such amides displayed the superb potency that accompanies the introduction of a para electron-donating substituent (IC50 = 300–800 pM). Of these, compound 173 emerged as the most attractive vinblastine analog. It is >10-fold more potent than vinblastine against the sensitive cell lines (IC50 = 400–500 pM), >300-fold more active against the resistant HCT116 cell line (IC50 = 1.8 nM), and displays a differential in activity of only 3.8-fold. As a result and like 28 and 121, 173 exhibited a profile of activity that we were hoping to discover at the start of our studies and became a key compound that was further profiled.</p><p>This latter set of compounds also provided us with a series of closely related compounds that display good activity and a favorable activity ratio with which the qualitative correlation of the amide lipophilic character and the differential activity could be retrospectively illustrated. A plot of cLogP versus the fold difference in activity (ratio) that includes compounds 28, 121, and 173, along with a series of compounds that progressively exchange in single heteroatoms, qualitatively highlights the reduction in the activity differential (ratio) that accompanies an increase in the amide lipophilicity (Figure 10). This reduction in activity differential that results from the additional relative improvements in activity against the resistant HCT116/VM46 cell line is suggestive such compounds are less effective substrates for Pgp efflux.</p><!><p>A small series of C20' sulfonamides was also prepared in a single step from 6 (1.5 equiv RSO2Cl, 2 equiv i-Pr2NEt, 0.05 M CH2Cl2, 23 °C, 16 h; Method 3) and examined (Figure 11). No compound in this series approached the potency of vinblastine. In the cases where a comparison C20' amide was prepared, the corresponding sulfonamide (179 vs 24, 180 vs 25, 181 vs 67 and 185 vs 116) proved to be approximately 50–100 times less active.</p><!><p>In recent work, the incorporation of a fluorine atom at the 10' position of vinblastine provided a compound (187) with a nearly 10-fold improvement in activity over vinblastine itself.40 We were interested in determining whether the incorporation of both the 10'-F substituent and a 20' amide would have the same effect of enhancing the potency of the vinblastine 20' amide analog and also establishing whether a 10'-F substituent would impact the improved differential in activity against the matched sensitive and resistant HCT116 cell line. Three key 20' amides also containing a 10'-F substituent were prepared, each of which constitute 10'-F derivatives of 20' amides that displayed superb or improved reductions in the differential activity (Figure 12). In each case, the potency of 20' amide analog was maintained (189 vs 121 and 190 vs 70) or enhanced (188 vs 78) and the improvement in the activity differential observed with the parent 10'-H analogs was maintained (189) or was further improved (188 and 190). The latter compound 190 is notable in that it is exceptionally potent (IC50 = 60–80 pM) against the vinblastine sensitive cell lines and even displays sub-nanomolar activity against the vinblastine-resistant HCT116 cell line (IC50 = 900 pM against HCT116/VM46), representing a >600-fold improvement in this activity.</p><!><p>From the series of 20' amides prepared and examined, three (28, 121 and 173) were chosen for further evaluation. The set consists of a small series of hydrophobic aryl amides that are essentially equipotent with (28 and 121) or more potent (173) than vinblastine against sensitive cancer cell lines (e.g. HCT116) and that maintain this potency against the matched resistant human cancer cell line (HCT116/VM46). The tubulin binding properties of the three compounds were examined alongside vinblastine for their relative ability to displace tubulin bound BODIPY-vinblastine.43 Like the prior comparisons summarized in Figure 5, the cell-based functional activity of the compounds correlated directly with their relative tubulin binding affinities (Figure 13). Thus, the affinities of 28 and 121 were essentially indistinguishable from or perhaps slightly better than that of vinblastine, whereas that of 173 was established to be significantly higher.</p><p>The observations made with the three compounds also indicated that the derivatives are not subject to resistance derived from Pgp overexpression as found in HCT116/VM46. These results suggested they are no longer effective substrates for Pgp efflux and that this type of modification may disrupt binding and/or transport by Pgp. This was confirmed in two widely used secondary assays (Caco-2 bidirectional permeability and stimulated Pgp ATPase activity in membranes).48–50 The results are summarized in Figure 14 where the compounds demonstrate little or no Pgp transport (or efflux) while maintaining the intrinsic permeability of vinblastine. At the start of our studies many years ago now, we viewed this type of result as an initial complete success for the studies – the discovery of vinblastine analogs that matched or exceeded the potency of the clinical drugs, but that would not be subject to clinical resistance derived from Pgp overexpression and efflux.</p><p>Finally, we had the three compounds independently examined alongside vinblastine and taxol, which included their independent assessments against HCT116 and HCT116/VM46 and against a representative and slightly larger panel of additional clinically relevant human cancer cell lines, where they displayed trends analogous to those already highlighted (Figure 15).51 All three compounds exceeded the potency of vinblastine (avg IC50 = 13 nM), matched or exceeded the potency of taxol (avg IC50 = 5.5 nM), and displayed a consistent potency across the slightly larger panel of human cancer cell lines. Compound 173 (avg IC50 = 1.2 nM) proved to be 2–8 fold more potent than 121 and 28, and 28 (avg IC50 = 4.0 nM) was typically found to be slightly more potent than 121 (avg IC50 = 4.6 nM). Each demonstrated a differential in activity (IC50 for HCT116/VM46 vs HCT116) that was small and consistent with our own assessments (2, 2.9 and 7 vs 1.7, 1.8 and 3.8 for 28, 121 and 173), indicating they maintain good activity against the vinblastine-resistant (and taxol-resistant) cell line and are ineffective (28 and 121) or poor (173) substrates for Pgp efflux.</p><!><p>The binding site for vinblastine lies at the head-to-tail tubulin α/β dimer–dimer interface. As depicted in the X-ray co-crystal structures of tubulin bound complexes,52,53 vinblastine is nearly completely buried in the protein binding site. It adopts a T-shaped bound conformation with C3/C4 (bottom of T) lying at the solvent interface and the C20' site (top corner of T) extends deepest into the binding pocket lying at one corner. In contrast to early expectations based on the steric constraints of the tubulin binding site surrounding the vinblastine C20' center, large 20' substituents such as those detailed herein are accommodated. The C20' alcohol extends toward a narrow channel that leads from the buried C20' site to the opposite face of the protein, representing the continuation of the protein–protein interaction defined by the tubulin dimer–dimer interface. Even without adjusting the proteins found in a vinblastine-bound X-ray structure (pdb 5J2T), the modeled 20' amides extend into this narrow channel, continuing along the tubulin α/β dimer–dimer protein–protein interface (Figure 16). The newly introduced vinblastine 20' amide forms two key H-bonds in which the amide N–H serves as a H-bond donor for the backbone carbonyl of Pro222 (2.0 Å) and the amide carbonyl serves as a H-bond acceptor for the backbone amide N–H of Tyr224 (3.0 Å). These serve to anchor the orientation of the 20' amides such that the attached acyl group extends into the adjacent narrow channel. In this orientation, not only are substituents on the benzamides accommodated at either the 3- or 4-position, but such electron-donating substituents that increase the Lewis basicity of the amide carbonyl would be expected to increase the strength of the H-bond with Tyr224, accounting for the enhanced tubulin binding affinity and the resulting increased activity in cell-based functional assays.</p><!><p>A site and powerful functionalization strategy on vinblastine was exploited that provided access to analogs which simultaneously maintain or improve cell-based functional activity, maintain or improve target tubulin binding affinity, and simultaneously disrupt off target activity (Pgp efflux) responsible for resistance. Thus, an extensive and systematic series of synthetic vinblastine 20' amides were prepared in three steps from commercially available material, targeting a site inaccessible to traditional divergent functionalization.54 Many such 20' amides were found to exhibit substantial and some even remarkable potency increases, many exhibited further improvements in activity against a Pgp overexpressing resistant cancer cell line, and an important subset of the vinblastine analogs displayed little or no differential in activity against a matched pair of vinblastine sensitive and resistant (Pgp overexpressing) cell lines. The improvements in potency directly correlated with improvements in target tubulin binding affinity, and the reduction in differential functional activity against the sensitive and resistant cell lines was found to correlate with analogous reductions in Pgp-derived efflux. Well defined structure–activity relationships and a structural model were developed in the studies that confidently account for the structural features that improve functional and target tubulin binding activity and key insights into structural characteristics55 that can be used to simultaneously disrupt off target Pgp binding and/or efflux responsible for drug resistance were obtained. Members of this class of vinblastine 20' amides have the potential of not only serving as vinblastine replacements in the clinic, addressing Pgp-derived clinical resistance limiting its continued use, but may also offer opportunities for the development of powerful new frontline treatment options in instances of other multidrug resistant (MDR) tumors (overexpression of Pgp) refractory to most other chemotherapeutic drugs.7</p><!><p>All commercial reagents were used without further purification unless otherwise noted. All reactions were performed in oven-dried (200 ºC) glassware and under an inert atmosphere of Ar unless otherwise noted. Column chromatography was performed with silica gel 60 (43–60 Å). TLC was performed on</p><p>Whatman silica gel (250 μm) F254 glass plates and spots visualized by UV. PTLC was performed on Whatman silica gel (250 and 500 μm) F254 glass plates. Optical rotations were determined on a Rudolph Research Analytical Autopol III automatic polarimeter using the sodium D line (λ = 589 nm) at room temperature (23 oC) and are reported as follows: [α]D23, concentration (c = g/100 mL), and solvent. FT-IR spectroscopy was conducted on a Nicolet 380 FT-IR instrument. 1H NMR was recorded on a Bruker 600 MHz spectrometer. Chemical shifts are reported in ppm from an internal standard of residual CHCl3 (7.26 for 1H). Proton chemical data are reported as follows: chemical shift (δ), multiplicity (ovlp = overlapping, br = broad, s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet), coupling constant, and integration. High resolution mass spectra were obtained at The Scripps Research Institute Mass Spectrometry Facility on an Agilent ESI-TOF/MS using Agilent ESI-L low concentration tuning mix as internal high resolution calibration standards. The purity of each tested compound (>95%) was determined on an Agilent 1100 LC/MS instrument using a ZORBAX SB-C18 column (3.5 mm, 4.6 mm × 50 mm, with a flow rate of 0.75 mL/min and detection at 220 and 254 nm) with a 10–98% acetonitrile/water/0.1% formic acid gradient (two different gradients). A table of the established purity for each tested compound is provided in the Supporting Information.</p><!><p>A solution of 20'-aminovinblastine41 (6, 3.5 mg, 0.004 mmol) in CH2Cl2 (0.1 mL) was treated with 4 μL of i-Pr2NEt (0.016 mmol) followed by addition of the acid chloride (0.008 mmol). The reaction mixture was stirred for 2 h at room temperature before being quenched with the addition of saturated aqueous NaHCO3 (3 mL). The mixture was extracted with 10% MeOH in CH2Cl2 (3 mL), and washed with saturated aqueous NaCl (3 mL). The organic layer was dried over Na2SO4, and concentrated under reduced pressure. PTLC (SiO2, EtOAc:MeOH:Et3N = 95:5:5) purification provided the pure products; yields (35–98%).</p><!><p>A solution 20'-aminovinblastine41 (6, 4 mg, 0.005 mmol) in DMF (0.1 mL) was treated with EDCI (0.02 mmol), DMAP (20 mol%) and the carboxylic acid (0.01 mmol). The reaction mixture was allowed to stir at room temperature overnight, after which it was diluted with the addition of 10% MeOH in CH2Cl2 (3 mL) and aqueous 10% citric acid or 1 M HCl (3 mL). The aqueous layer was further extracted with 10% MeOH in CH2Cl2, and the combined organic phase was washed with saturated aqueous NaHCO3 (3 mL), and saturated aqueous NaCl (3 mL). The organic layer was dried over Na2SO4, and concentrated under reduced pressure. PTLC (SiO2, EtOAc:MeOH:Et3N = 95:5:5) purification provided the pure products; yields (20–98%).</p><!><p>A solution of 20'-aminovinblastine41 (6, 4 mg, 0.005 mmol) in anhydrous CH2Cl2 (0.1) was treated with i-Pr2NEt (0.01 mmol) and the sulfonyl chloride (0.008 mmol). The resulting mixture was allowed to stir at room temperature overnight, after which it was diluted with the addition of saturated aqueous NaHCO3 (2 mL). The mixture was extracted with 10% MeOH in CH2Cl2, and washed with saturated aqueous NaCl (3 mL). The combined organic extracts were dried over Na2SO4 and concentrated under reduced pressure. PTLC (SiO2, EtOAc:MeOH:Et3N = 97:3:3) purification provided the pure products; yields (41–54%).</p><!><p>Method 1 was followed using 8.0 mg of 20'-aminovinblastine (6, 0.01 mmol) to provide 4.7 mg of 28 as a white solid, yield: 50%. 1H NMR (600 MHz, CDCl3) δ 9.83 (br s, 2H), 8.02 (s, 1H), 7.82–7.76 (m, 2H), 7.44 (d, J = 7.2 Hz, 1H), 7.24 (d, J = 7.7 Hz, 1H), 7.14–7.07 (m, 2H), 6.64 (s, 1H), 6.12 (s, 1H), 6.08 (s, 1H), 5.85 (d, J = 5.7 Hz, 1H), 5.48 (s, 1H), 5.30 (d, J = 10.3 Hz, 1H), 4.00 (br s, 1H), 3.81 (s, 3H), 3.80 (s, 3H), 3.75 (s, 1H), 3.58 (s, 3H), 3.42–3.36 (m, 2H), 3.30 (td, J = 9.5, 4.8 Hz, 1H), 3.22 (t, J = 8.7 Hz, 1H), 3.10 (dd, J = 14.6, 7.2 Hz, 2H), 2.83 (d, J = 16.2 Hz, 1H), 2.72 (s, 3H), 2.67 (s, 1H), 2.45–2.41 (m, 1H), 2.35 (s, 3H), 2.30 (s, 3H), 2.17 (s, 1H), 2.11 (s, 3H), 2.02–1.98 (m, 1H), 1.85–1.78 (m, 2H), 1.62–1.59 (m, 4H), 1.42 (t, J = 7.4 Hz, 1H), 1.37–1.31 (m, 2H), 1.26–1.23 (m, 3H), 0.82 (t, J = 7.0 Hz, 3H), 0.77 (t, J = 6.2 Hz, 3H); HRESI-TOF m/z 942.5012 (C55H67N5O9 + H+, required 942.5011). [α]D23–39 (c 0.027, CHCl3).</p><!><p>Method 2 was followed using 8.4 mg of 20'-aminovinblastine (6, 0.01 mmol) to provide 4.0 mg of 121 as a white solid, yield: 42%. 1H NMR (600 MHz, CDCl3) δ 9.86 (br s, 1H), 8.05 (s, 1H), 7.76 (s, 2H), 7.48–7.46 (m, 1H), 7.19 (d, J = 8.2 Hz, 1H), 7.16 (d, J = 6.7 Hz, 1H), 7.12–7.10 (m, 2H), 6.66 (s, 1H), 6.14 (s, 1H), 6.10 (s, 1H), 5.88–5.87 (m, 1H), 5.50 (s, 1H), 5.33 (d, J = 10.2 Hz, 1H), 4.00 (br s, 1H), 3.84 (s, 3H), 3.82 (s, 3H), 3.78–3.77 (m, 1H), 3.60 (s, 3H), 3.44–3.38 (m, 2H), 3.33 (td, J = 9.5, 4.8 Hz, 1H), 3.26–3.23 (m, 1H), 3.14–3.10 (m, 2H), 2.91–2.87 (m, 2H), 2.84–2.83 (m, 1H), 2.82–2.81 (m, 2H), 2.75 (s, 3H), 2.71–2.69 (m, 1H), 2.48–2.43 (m, 1H), 2.39–2.36 (m, 1H), 2.20 (s, 3H), 2.13 (s, 1H), 1.88–1.85 (m, 1H), 1.83–1.82 (m, 4H), 1.61 (s, 3H), 1.54–1.51 (m, 1H), 1.44 (t, J = 7.4 Hz, 1H), 1.37–1.33 (m, 2H), 1.28 (s, 2H), 0.92–0.89 (m, 1H), 0.84 (t, J = 6.9 Hz, 3H), 0.80–0.78 (m, 3H); HRESI-TOF m/z 968.5164 (C57H69N5O9 + H+, required 968.5168). [α]D23–55 (c 0.069, CHCl3).</p><!><p>Method 2 was followed using 8.0 mg of 20'-aminovinblastine (6, 0.01 mmol) to provide 3.5 mg of 173 as a pale white solid, yield: 37%. 1H NMR (600 MHz, CDCl3) δ 9.84 (br s, 1H), 8.06–8.04 (m, 1H), 7.91 (s, 1H), 7.58–7.50 (m, 2H), 7.30–7.38 (m, 1H), 7.19–7.16 (m, 2H), 6.78–6.77 (m, 2H), 6.53 (s, 1H), 6.11 (s, 1H), 5.88 (s, 1H), 5.47 (s, 1H), 5.32 (d, J = 9.8 Hz, 1H), 4.21 (t, J = 4.6 Hz, 2H), 3.82–3.82 (m, 6H), 3.76 (br s, 1H), 3.63 (s, 3H), 3.51 (s, 1H), 3.42–3.37 (m, 2H), 3.31–3.25 (m, 2H), 3.15–3.10 (m, 2H), 2.91 (s, 1H), 2.82–2.79 (m, 2H), 2.74 (s, 3H), 2.66–2.62 (m, 1H), 2.45–2.41 (m, 1H), 2.32–2.30 (m, 1H), 2.19 (s, 3H), 2.13–2.12 (m, 2H), 2.11–2.09 (m, 1H), 2.01 (br s, 1H), 1.69–1.66 (m, 4H), 1.55–1.52 (m, 2H), 1.33–1.28 (m, 5H), 0.89–0.84 (m, 6H); HRESI-TOF m/z 970.4961 (C56H67N5O10 + H+, required 970.4960). [α]D23–76 (c 0.059, CHCl3).</p><!><p>A solution containing 22 mg of vinblastine sulfate in 0.6 mL of anhydrous acetonitrile was prepared. 60 μL of 18 M sulfuric acid was added. The resulting solution was stirred at ambient temperature for 7 h and then overnight at 0 °C. Next, 318 mg of Na2CO3 and 2 mL of anhydrous MeOH were added. This mixture was stirred for 15 min before 4 mL of a saturated aqueous NaCl was added. The reaction volume was increased to 8 mL by the addition of water. This diluted mixture was stirred for about 15 min, after which time it was extracted four times with an equal volume of benzene. The combined organic layer was dried over Na2SO4, and concentrated under reduced pressure. PTLC (SiO2, EtOAc:MeOH:Et3N = 97:3:3) provided three products. The first product was recovered vinblastine (5.5 mg). The second product was 20'-acetamidoylleurosidine (8, 1.2 mg) and its structure was confirmed by comparison with a sample prepared from authentic 20'-aminoleurosidine. The third product (1.8 mg) was leurosidine. For 20'-acetamidoleurosidine (8): 1H NMR (600 MHz, CDCl3) δ 9.81 (s, 1H), 7.98 (s, 1H), 7.52 (d, J = 7.8 Hz, 1H), 7.21–7.14 (m, 1H), 7.14–7.08 (m, 3H), 6.50 (s, 1H), 6.16 (s, 1H), 6.08 (s, 1H), 5.88–5.80 (m, 1H), 5.45 (s, 1H), 5.28 (d, J = 10.2 Hz, 1H), 3.79 (s, 3H), 3.77 (s, 1H), 3.76 (s, 3H), 3.66–3.61 (m, 1H), 3.59 (s, 3H), 3.40–3.34 (m, 1H), 3.33–3.27 (m, 1H), 3.27–3.21 (m, 2H), 3.15 (t, J = 14.4 Hz, 1H), 3.04 (dd, J = 14.5, 5.9 Hz, 1H), 2.97 (d, J = 10.7 Hz, 1H), 2.91–2.83 (m, 1H), 2.83–2.77 (m, 1H), 2.73 (s, 3H), 2.71–2.64 (m, 2H), 2.63 (s, 1H), 2.47–2.41 (m, 1H), 2.31 (dq, J = 14.8, 7.5 Hz, 1H), 2.27–2.22 (m, 1H), 2.22–2.15 (m, 1H), 2.09 (s, 3H), 1.88 (s, 3H), 1.78 (dt, J = 14.4, 7.4 Hz, 1H), 1.75–1.68 (m, 1H), 1.43 (dq, J = 14.3, 7.2 Hz, 1H), 1.36–1.27 (m, 1H), 1.07–0.99 (m, 1H), 0.96 (d, J = 15.1 Hz, 1H), 0.81 (t, J = 7.4 Hz, 3H), 0.78 (t, J = 7.4 Hz, 3H); 13C NMR (151 MHz, CDCl3) δ 174.5, 171.8, 171.1, 169.8, 158.1, 153.1, 135.0, 130.9, 130.2, 129.4, 124.6, 123.3, 123.0, 122.6, 119.2, 118.3, 117.0, 110.8, 94.5, 83.5, 79.8, 76.6, 65.8, 56.8, 56.0, 54.4, 53.4, 52.6, 52.4, 50.5, 44.7, 43.5, 42.8, 38.6, 37.1, 30.9, 30.7, 30.1, 24.7, 21.3, 8.6, 8.2; IR (film) νmax 3467, 2958, 1738, 1666, 1459, 1229, 1039, 749 cm−1; HRESI-TOF m/z 852.4547 (C48H61N5O9 + H+, required 852.4542). [α]D23 +13 (c 0.31, CHCl3). Identical in all respects with reported data and an authentic sample.41</p><!><p>Compounds were tested for their cell growth inhibition of L1210 (ATCC no. CCL-219, mouse lymphocytic leukemia), HCT116 (ATCC no. CCL-247, human colorectal carcinoma), and HCT116/VM46 (a vinblastine-resistant strain of HCT116) cells in culture. A population of cells (>1 × 106 cells/mL as determined with a hemocytometer) was diluted with Dulbecco's Modified Eagle Medium (DMEM, Gibco) containing 10% fetal bovine serum (FBS, Gibco) to give a final concentration of 30,000 cells/mL. To each well of a 96-well plate (Corning Costar) was added 100 μL of the cell media solution with a multichannel pipet. The cultures were incubated at 37 °C in an atmosphere of 5% CO2 and 95% humidified air for 24 h. The test compounds were then added to the plate as follows: test compounds were diluted in DMSO to a concentration of 1 mM. 10-fold serial dilutions in DMSO were next performed on a separate 96-well plate. Fresh culture media (100 μL) was added to each well of cells resulting in 200 μL of medium per well followed by 2 μL of each test agent. Compounds were tested in duplicate (n = 2–8 times) at six concentrations between 0–1000 nM or 0–10000 nM. Following the addition of compound, cultures were incubated for an additional 72 h. A phosphatase assay was used to establish IC50 values as follows: the media was removed from each well and treated with 100 μL of phosphatase solution (100 mg phosphatase substrate in 30 mL of 0.1 M NaOAc, pH 5.5, 0.1% Triton X-100 buffer). The plates were incubated at 37 °C for 5 min (L1210) or 15 min (HCT116 and HCT116/VM46). After the given incubation time, 50 μL of 0.1 N NaOH was added to each well and the absorption at 405 nm was determined using a 96 well plate reader. Given the absorption is directly proportional to the number of living cells, IC50 values were calculated and reported values represent the average of 4–16 determinations (SD ±10%).</p><!><p>Tubulin (0.1 mg/mL, 0.91 μM) was incubated with BODIPY-vinblastine (BODIPY-VBL, 1.8 μM) for 15 min at 37 °C in PEM buffer containing 850 μM GTP. Subsequently, a competitive ligand (vinblastine, 64, 24, or 54) was added to the solution at a final concentration of 18 μM. After incubation for 60 min at 37°C, 100 μL aliquots from each incubation was measured in a fluorescence microplate reader (FI; ex 480 nm, em 514 nm). Control experiments were performed with BODIPY-VBL in the absence of a competitive ligand (control 1, maximum FI enhancement) and in the absence of tubulin (control 2, no FI enhancement). % BODIPY-VBL displacement was calculated by the formula: (control 1 FI – experiment FI)/(Control 1 FI – Control 2 FI) × 100. Reported values are the average of 5 measurements ± SD.</p><!><p>The amount of drug stimulated Pgp ATPase activity generated by either vinblastine or analog 121 was determined using a MDR1 PREDEASY™ ATPase Kit Assay Protocol (SOLVO Biotechnology, Version Number: 1.2) purchased from Sigma–Aldrich (St. Louis, MO) and following the manufacturer's protocol.48,49 The Caco-2 cell permeability assay was conducted comparing vinblastine, 28, 121, and 173 by following a previously published procedure50 and was conducted by Sekisui XenoTech, LLC (Kansas City, KS). The results are reported in Figure 14 and only the diffusion in the B to A direction in the presence of positive control Pgp inhibitor valspodar (1 μM) impacted vinblastine transport (efflux ratio 2.7) as expected, but did not substantially impact the rate or efflux ratio of 28, 121 and 173.</p>

PubMed Author Manuscript

Docking and Migration of Carbon Monoxide in Nitrogenase: The Case for Gated Pockets from IR Spectroscopy and Molecular Dynamics

Evidence for a CO docking site near the FeMo-cofactor in nitrogenase has been obtained by FT-IR monitored low temperature photolysis. We investigated the possible migration paths for CO from this docking site using molecular dynamics calculations. The simulations support the notion of a gas channel with multiple internal pockets from the active site to the protein exterior. Travel between pockets is gated by motion of protein residues. Implications for the mechanism of nitrogenase reactions with CO and N2 are discussed.

docking_and_migration_of_carbon_monoxide_in_nitrogenase:_the_case_for_gated_pockets_from_ir_spectros

Introduction<!>Photolysis of CO-Nitrogenase<!>Molecular Dynamics<!>Infrared Spectroscopy<!>Molecular Dynamics<!>An Energy Surface<!>Discussion

<p>Nitrogenase (N2ase) is the enzyme responsible for biological nitrogen fixation.1–3 For molybdenum-containing N2ase from Azotobacter vinelandii, which this work will focus on, X-ray diffraction has revealed a unique [Mo-7Fe-9S-C]-homocitrate cluster at the active site of the MoFe protein of N2ase,4, 5 with an interstitial carbide at the center of a prismatic 6 Fe cage.5–7 This cluster, known as the FeMo-cofactor (FeMo-co), is capable of reducing a wide variety of triply bonded substrates, such as N2, C2H2, HCN, as well as protons.8 It has recently been shown that this enzyme can also produce CxHy hydrocarbons from CO9–11 and even CH4 from CO2.12, 13 The migration of substrates, such as: N2 and CO and protons, to the active site, as well as the exit of products such as NH3, various hydrocarbons CxHy, and H2, is clearly a critical part of N2ase reactivity.</p><p>Several different channels have been proposed for access to or from the FeMo-cofactor. 14–21 In 2003, Igarashi and Seefeldt pointed out a candidate substrate channel starting near surface residues α-K209 and α-W205. This mostly hydrophobic pathway (now called the 'IS channel'20) ultimately passes the α-V70, α-H195, and α-R96 region and terminates at the Fe2,3,6,7 face of the FeMo-cofactor.14 At the opposite end of the FeMo-cofactor, a hydrophilic channel extends from a 'water pool'15 proximal to the homocitrate ligand, through the interface between α and β subunits, and finally to the surface. Durrant offered this 'interstitial channel' as an efficient path for diffusion of both dinitrogen and ammonia,16 and his proposition was later tested through site-directed mutagenesis work by Barney and coworkers.17 Dance has noted a similar channel for the egress of ammonia, starting at α-Q191.18 Smith and coworkers have recently proposed a very different dynamic channel that opens and closes on the tens of nanoseconds time scale, starting near surface residues α-R281 and α-H383 and leading to the Fe2,3,6,7 face,19 which may have a role in proton transport.22, 23 Additional substrate/product pathways have most recently been proposed by Morrison and coworkers, based on Caver calculations combined with analysis of binding sites for Xe and small molecules.20 Apart from these small molecule channels, multiple proton relay chains ('proton bays' and 'proton wires'), leading to S2B, S3B, and S5A were noted by Durrant 16 and more recently analyzed by Dance.21</p><p>Pockets and channels for small molecules like CO and O2 have proven important in the study of myoglobin (Mb) dynamics,24 and they have also been found important for hydrogenases,25 cytochrome oxidase,26 carbon monoxide dehydrogenase/acetyl-CoA synthase,27 and many other 'gas-processing' enzymes.28 There is a flourishing literature on how to deduce and evaluate these conduits by computational methods.29–34 On the experimental side, the migration of CO following MbCO photolysis has been followed by infrared spectroscopy, including static,35 picosecond time-resolved,36 and temperature derivative37 methods. Two photolysis product IR bands at 2131 and 2119 cm−1, respectively labeled B1 and B2, are observed for wild-type MbCO.35 These have been taken as evidence for two different CO orientations at the Mb 'docking site',38 and the magnitude of the splitting has allowed calculation of the interior electric field.37 Extensive time-resolved x-ray diffraction studies have allowed observation of CO migration between different pockets within Mb39 and the correlation of CO pockets with those observed under high-pressure Xe.40</p><p>Pockets and channels should certainly be relevant for understanding N2ase. Analogous to Mb, Xe pockets have been identified by x-ray crystallography in Klebsiella pneumoniae41 and Azotobacter vinelandii (Av).42 Recently, a detailed comparison of Xe sites in Av and Clostridium pasteurianum has been made, along with binding sites for CO and other small molecules.43 In this paper, we present FT-IR cryophotolysis data that supports a docking site for CO near the FeMo-cofactor. We then use molecular dynamics calculations to identify a location for the docking site in Av Mo N2ase as well as a channel allowing for escape of CO and for entry of N2.</p><!><p>The N2ase enzyme was prepared from Av and reacted with CO using the same protocol from our previous studies.23, 44 The infrared spectra were collected with a Bruker Vertex 70v FT-IR under cryogenic temperatures. Photolysis was induced by a broadband Sutter Instruments xenon-arc lamp. Samples were held in custom built cells with Teflon spacers to give a pathlength of 70 microns.</p><!><p>Simulations were performed using a customized forcefield in the Gromacs molecular dynamics package.45 The MoFe N2ase-CO α-subunit was first energy minimized, then followed by dynamics at multiple temperatures.</p><p>More detailed information about all experimental methods is available in the supporting information.</p><!><p>As shown in Figure 1, positive product bands appear upon low temperature photolysis of CO-inhibited wild-type N2ase as well as the α-H195Q variant. The 37 cm−1 downshifts with 13CO substitution confirm that these are indeed CO bands. By analogy with the MbCO literature, the absorption bands at 2135 and 2123 cm−1 are consistent with 'free CO' species adopting two orientations at a docking site.38 We label the N2ase free CO bands respectively NB1 and NB2 by analogy to the myoglobin (Mb) CO labels.35 For Mb mutants, additional bands have been observed to range from 2108 to 2152 cm−1, and we cannot rule out the existence of such minor species in N2ase.</p><p>There are differences between wild-type and α-H195Q spectra that suggest that altering the amino acid environment around the FeMo-cofactor has an effect on the photolyzed CO. In the wild type enzyme, the NB2 band at 2123 cm−1 is more intense than the 2135 cm−1 NB1 band. In α-H195Q, the intensity pattern is reversed, and the asymmetry favors the NB1 band, suggestive of two or more unresolved sub-species. Since a small change in side-chain at the 195 position affects the photolysis spectra, these results are at least consistent with locating the 'free CO' close to that position.</p><!><p>Inspired by the experimental evidence for a CO docking site in N2ase, we began a molecular dynamics study of CO migration from that site–analogous to those used for myoglobin and related proteins.15, 46 In this work we took advantage of a force field for the FeMo-cofactor that had been previously validated by comparison with results from nuclear resonance vibrational spectroscopy (NRVS), as well as structural models for CO binding to the FeMo-cofactor from DFT calculations that have been tested against NRVS and EXAFS data.23 We chose to initially place the CO near Fe2. Our previous work has shown the photolabile "Hi-1" CO form has a terminal CO and a formyl-like species.47 Likewise, Fe6 has been implicated as the most reactive site on the FeMo-cofactor,48 therefore we presume Fe6 binds the more reduced formyl-like species and Fe2 binds the terminal CO species that photolyzes to free CO. The entire protein structure was then relaxed to minimize any forces resulting from CO insertion. This yielded an initial Fe2-CO distance of 4.2 Å. The CO was then allowed to migrate through the protein at various temperatures, including 10, 80, 150, 200, 250, and 300 K. The motion of the CO and the protein was followed out to 10 ns (Figure 3).</p><p>The molecular dynamics calculations find a candidate 'docking site' with the CO about 5 Å from Fe2 (Figure 1b and Figure 3). In this location, CO is within H-bonding distance to α-His-195 Nε (3.1 Å). Another residue within H-bonding distance is α-Y229 (O-O distance 4.1 Å). Finally, the critical α-V70 side-chain is in position to constrain the CO motion. This bears some resemblance to the docking site in myoglobin, where photolyzed CO is within H-bonding distance to H64 and is further trapped by aliphatic side-chains of L29 and I107 after photolysis.38, 39, 49 As illustrated by the time courses plotted for different temperatures (Figure 2), at 10K the CO remains in the docking site, but at higher temperatures it rapidly migrates to 3 distinct regions at progressively larger distances from the FeMo-cofactor.</p><p>Upon leaving the docking site, the overall motion was stochastic and bidirectional. However, the path that was most frequently taken was strikingly similar to the IS channel.14 Our result is compared with other proposed channels and pockets in Figure 2. The channel begins near key residues α-H195 and α-V70, and it ends near surface residues α-M78, α-V179, and α-V206 and α-I259 (Figure 2). Although this might only seem a confirmation of the Igarashi channel, from the dynamics calculation we see that the migration of CO is not uniform over time (Figure 2). Rather, just as in myoglobin and other proteins,50 the CO was relatively stable in several distinct sites–the initial 'docking site' and 3 distinct 'pockets'. Passage from the docking site and between pockets is governed by occasional gate openings by key residues along the channel through thermodynamic fluctuations. We now discuss the sites where the CO spent most of the time.</p><p>The first pocket encountered (after leaving the docking site) has the CO center-of-mass (COM) 9.5 Å from the FeMo-cofactor central carbide, with a closest approach to Fe2 of 8.5 Å (Figure 4a). Key side-chains in this pocket include α-M279, α-V71, and α-S278. Unlike the later pockets, this first pocket also includes a tyrosine side-chain, α-Y229, a good candidate for hydrogen-bonding to CO. As the closest pocket to the docking site, we label this the proximal or P-pocket.</p><p>The second ('medial' or 'M') pocket seen in the MD simulations contains α-W253, α-S254, α-I282 (Figure 4b). It is gated from the previous pocket by α-N199, α-M279, α-V71 and α-A198. The four residues sterically prevent CO from returning to the previous pocket, but they can open during thermal fluctuations. When CO is trapped within the M-pocket, the Fe2 to CO distance is about 13.0 Å. We gain confidence in this predicted location for CO because it coincides with the Xe1 pocket seen by x-ray diffraction.20 As shown in Figure 4b, our typical CO position overlaps nicely with the Xe position observed by Rees and coworkers.20,42</p><p>The third ('distal' or 'D') pocket observed in our calculations involves α-I259, α-V206, α-M78, α-V179 (Figure 4c). Access to the D-pocket from the M-pocket is gated by residues α-V202, α-I75, and α-W72. All of the aforementioned side-chains are part of the IS channel proposed by Igarashi and Seefeldt.14 The average Fe2-CO distance from within the D-pocket was 17 Å. We note that a final region for CO occupation was located beyond the D-pocket and within a solvent interaction crater on the surface of the protein. This region had an average Fe2-CO distance of ~21 Å. Penetration into the solvent occurred at a distance of about 25 Å.</p><p>Since N2 is the natural substrate for N2ase, we also performed a simulation at 300K using N2 as the diatomic in motion. As expected, N2 behaved similarly to CO and bidirectionally traversed through the three pockets and into the solvent space (Figure S1). In this simulation, the N2 spent most of its time in the M pocket–just as we observed with CO.</p><!><p>The motion of CO within N2ase appears to involve capture within pockets interlaced with occasional gate openings. We decided to map the potential energy surface for this motion of CO within the protein by performing a potential of mean force (PMF) calculation using GROMACS.51 This involved sampling multiple Fe2-CO distances and evolving them individually in time. The results from these calculations are shown in Figure 4d. The first minimum is a shallow well at 5 Å that corresponds to the proposed docking site. At longer distances there are three distinct potential minima corresponding to each identified pocket in the channel. Beyond these wells we see the solvent interaction crater at 20 Å and solvent space at about 25 Å. An important result from these calculations is that the location identified as the medial pocket is the lowest energy location where CO can reside.</p><p>The overall scale of the PMF is similar to what is seen in previous work for a ligand in the different channel proposed recently,19 however the PMF defined in this work has potential wells with depths greater than what is seen for the that channel. The M pocket has the largest barriers in both direction: 1.9 kcal/mol towards the D-pocket and 7.4 kcal/mol for CO moving towards the P-pocket.</p><!><p>Our calculations are the first use of molecular dynamics to include an experimentally constrained force field for the FeMo-cofactor. By combining these calculations with IR-monitored cryogenic photolysis, we have found the first evidence for a CO docking site in N2ase. We favor this docking site as the most likely position for free CO following cryogenic photolysis. The docking site is adjacent to the FeMo-cofactor and has two possible hydrogen bonding partners for CO–α-H195 and α-Y229. In myoglobin, H-bonding to H64 is considered important for the 12 cm−1 splitting of product bands B1 and B2 at 2131 and 2119 respectively. Substitution of histidine by leucine in the H64L variant in myoglobin causes the loss of that splitting and yields a single product band at 2126 cm−1.38 H-bonding to α-H195 or α-Y229 may play a similar role in splitting the 'free CO' bands of photolyzed N2ase.</p><p>Using the MD calculations, by following the path of CO or N2 starting near this docking site, we discovered a channel for these molecules from the FeMo-cofactor to the protein exterior, in a location consistent with a previous proposal from static CAVER calculations.14 A benefit from the dynamics simulations is that they reveal a novel gating mechanism for CO or N2 migration through this hydrophobic channel–a feature not observable from static calculations. We find that CO spends most of its time in distinct pockets, and travel between pockets is enabled by gating motions of neighboring amino acids.</p><p>Although the exact mechanism of N2ase catalysis is far from understood, all proposed reaction pathways require multiple electron and proton transfers to the FeMo-cofactor region before any ligand binding can occur. In particular, the popular Lowe-Thornley (LT) model for the N2ase reaction pathway52 posits 3 or 4 electron/proton transfers to the FeMo-cofactor and/or its associated ligands before N2 binding. In the absence of substrate binding, the LT model proposes that the FeMo-cofactor will oxidize through H2 evolution.8</p><p>It has been argued that because N2ase is a slow enzyme, it has no need for a hydrophobic tunnel that would allow rapid gas access to the FeMo-cofactor.28 However, the pockets and gates that we observe may play a role in optimizing N2ase catalytic efficiency. By trapping an N2 molecule in the M-pocket, the enzyme would have substrate available for binding within a few nanoseconds of reaching the E3 or E4 levels. This might help minimize 'futile' H2 production that might occur if N2 were not immediately available. These gated "storage" pockets would allow for N2 availability whenever the FeMo-cofactor reaches the appropriate level of reduction. In a similar vein, it has been proposed that large sections of tunnels serve as H2 'gas reservoirs' in hydrogenases. 53, 54</p><p>Weyman and coworkers studied the α-V75I and α-V76I variants in a similar N2ase from Anabaena variabilis (which would correspond to α-V7OI and α-V71I in Av).55 Although as expected, their α-V75I variant showed reduced N2 fixation activity and α-V76I substitution had no effect. Since our proposal for gated access between pockets is dynamic, we can accommodate these findings by simply allowing for comparable gating by isoleucine or valine residues at the Av position 71.</p><p>Some of the other channels illustrated in Figure 2 presumably play a role in the egress of the mandatory H2 co-product as well as NH3 or NH4+. In this area we are agnostic, since Lautier has noted that hydrophobic molecules sometimes travel through hydrophilic channels and vice versa.56 However, it seems unlikely that evolved H2 would exit through our proposed substrate channel, since this might limit N2 influx to the FeMo-cofactor. It seems logical that N2ase would have a mechanism or 'pressure relief' for the active site region to release H2 from the active site without interfering with the catalytic process. Overall, N2ase has to handle the flow of electrons, protons, N2, H2, and NH3, and the nature of 'traffic control' in this remarkable enzyme remains to be completely understood.</p>

PubMed Author Manuscript

Transcription factorIRX5 promotes hepatocellular carcinoma proliferation and inhibits apoptosis by regulating the p53 signalling pathway

Hepatocellular carcinoma (HCC) is the fifth most common cancer worldwide and the third most frequent cause of cancer‐related death. The IRX5 transcription factor plays a different role in multiple cancers and contributes to the development of many tumours. However, little is known about the molecular mechanisms of IRX5 in HCC. In this study, we found that IRX5 was abnormally upregulated in HCC tissues compared with adjacent normal tissues. IRX5 promoted HCC cell proliferation and upregulated the expression of cyclin D1 and knockdown of IRX5 suppressed tumorigenicity in vivo. Furthermore, knockdown of IRX5 increased p53 and Bax expression and decreased Bcl‐2 expression. Thus, IRX5 suppressed apoptosis in HCC cells by inhibiting the p53 signalling pathway, indicating its role as a treatment target for HCC.Significance of the studyOur study demonstrated that IRX5 was abnormally upregulated in HCC tissues compared with adjacent normal tissues. IRX5 promoted HCC cell proliferation and upregulated the expression of cyclin D1, and knockdown of IRX5 suppressed tumorigenicity in vivo. Furthermore, knockdown of IRX5 increased p53 and Bax expression and decreased Bcl‐2 expression. IRX5 suppressed apoptosis in HCC cells by inhibiting the p53 signalling pathway, indicating its role as a treatment target for HCC.

transcription_factorirx5_promotes_hepatocellular_carcinoma_proliferation_and_inhibits_apoptosis_by_r

<!>INTRODUCTION<!>HCC cell lines<!>Human tissues<!>Plasmid construction<!>Cell proliferation assays<!>Reverse transcription and quantitative real‐time polymerase chain reaction<!>Immunohistochemical staining<!>Western blot analysis<!>Cell cycle and cell apoptosis assays<!>Construction of stable cell lines<!>Animal studies<!>Statistical analysis<!>IRX5 was upregulated in HCC tissues and cell lines<!>IRX5 promoted HCC cell proliferation in vitro<!><!>IRX5 induced cell cycle progression by upregulatingcyclinD1 and suppressing apoptosis via inhibiting p53 signalling in HCC cells<!><!>IRX5 induced cell cycle progression by upregulatingcyclinD1 and suppressing apoptosis via inhibiting p53 signalling in HCC cells<!>Knockdown of IRX5 suppressed HCC tumour growth in vivo<!><!>DISCUSSION<!>CONCLUSION<!>CONFLICT OF INTEREST

<p>Zhu L , Dai L , Yang N , et al. Transcription factorIRX5 promotes hepatocellular carcinoma proliferation and inhibits apoptosis by regulating the p53 signalling pathway. Cell Biochem Funct. 2020;38:621–629. 10.1002/cbf.3517</p><p>Liying Zhu and Longguang Dai contributed equally to this work.</p><p>Funding information Education Youth Science and Technology Talents Foundation of Guizhou Province, Grant/Award Number: [2018]176; Yuzhong District Science and Technology Bureau, Chongqing, Grant/Award Number: 20160136; Program for Innovation Team of Higher Education in Chongqing, Grant/Award Number: CXTDX201601015; National Natural Science Foundation of China, Grant/Award Number: 81871653; Doctor Fund from Affiliated Hospital of Guizhou Medical University; Science and Technology Department Academic New Seedling Foundation of Guizhou Province, Grant/Award Number: [2017]5718</p><!><p>Liver cancer was the sixth most commonly diagnosed cancer and the fourth leading cause of cancer‐related death worldwide in 2018,1 with approximately 841 000 new cases and 782 000 deaths annually.2 Hepatocellular carcinoma (HCC) accounts for 75‐85% of primary liver cancers and is one of the most common malignant tumours in the world.3 Surgery is the most effective therapy for HCC at present.4 The vast majority of HCC patients present with advanced‐stage disease that ultimately leads to poor prognosis5 and a 5‐year survival rate of less than 20%4, 6; the recurrence rate is as high as 80%.7 The mechanism of HCC development has not been clarified, particularly the mechanism underlying the regulation of the proliferation and apoptosis of HCC.</p><p>Iroquois homeobox (Irx) genes play pivotal roles in normal embryonic cell patterning and the development of malignancies.8 The Irx family is composed of six genes in humans and mice, including two clusters, IrxA (Irx1, Irx2 and Irx4) located on chromosome 5 and IrxB (Irx3, Irx5 and Irx6) located on chromosome 16.9 IRX5 encodes a transcription factor and is a highly conserved member of the Iroquois homeobox gene family.8 IRX5 plays a different role in multiple cancers, contributing to the development of many tumours by acting as an important transcription factor regulating key regulatory genes that control cell growth, invasion, migration and apoptosis. Recent data suggest that IRX5 is a transcription factor that remarkably promotes tongue squamous cell carcinoma tumour growth by targeting the osteopontin (OPN) promoter and activating the NF‐κB pathway.8 Our previous study showed that CRNDE acted as a tumour oncogene by promoting the oncogenic properties of human HCC and revealed a novel CRNDE‐miR‐136‐5p‐IRX5 regulatory network in HCC.3</p><p>CyclinD1 regulates cell cycle progression, forms complexes with cyclin‐dependent kinase 4 or 6 in the cytoplasm10, 11 and promotes progression from G1 to S‐phase.12 Previous findings have indicated that cyclin D1 is a downstream target of IRX5.9 P53 is a tumour suppressor gene,13 and the p53 protein is a transcription factor that is involved in cell cycle arrest, DNA repair and apoptosis.14 In the p53 signalling pathway, p53 positively regulates the proapoptotic proteins Bax and p53 negatively regulates the transcription of Bcl‐2.15</p><p>In this study, we explored the role ofIRX5 in HCC proliferation and apoptosis. Next, we will discuss the properties of IRX5 in the p53 signalling pathway.</p><!><p>HepG2, SMMC7721, SK‐hep1, Huh7, human immortalized normal human liver cell line (L02) and the embryonic kidney cell line 293 T were obtained from the Chinese Academy of Sciences Cell Bank. They were cultured in Dulbecco's modified Eagle's medium (DMEM) of high glucose with 10% foetal bovine serum (FBS, BI, ISR). All cells were incubated at 37 °C in a humidified incubator with 5% CO2.</p><!><p>Ten pairs of primary HCC and adjacent non‐tumour tissues were obtained from surgical resections of HCC in the Affiliated Hospital of Guizhou Medical University between January 2015 and January 2016. Fresh tissue samples were collected and processed within 10 minutes. Each sample was snap‐frozen in liquid nitrogen and then stored at −196 °C. The data do not contain any information that could identify the patients. All patients provided written informed consent, and ethical consent was granted from the Committees for Ethical Review of Research involving the Affiliated Hospital of Guizhou Medical University (Guizhou, China).</p><!><p>PcDNA3.1‐IRX5 was purchased from GenePharma (Shanghai, China). The short‐hairpin RNA targeting human IRX5 was ligated into the pGreenPuro shRNA vector (SBI, Palo Alto, CA) according to manufacturer's protocol. The constructed sequence was further confirmed by sequencing. Transfections were performed with a Lipofectamine 2000 kit (Invitrogen, Carlsbad, CA), according to manufacturer's instructions. Cells were harvested 48‐72 hours after transfection.</p><!><p>Cells (2000 cells/well) were seeded into 96 well plates after 24 hours of transfection and measured at different time points (0, 24, 48 and 72 hours) using the MTS kit (CellTiter 96 Aqueous One Solution Cell Proliferation Assay, Promega), following the manufacturer's protocol. The absorbance at 490 nm was measured with a spectrophotometer. All experiments were performed in triplicate.</p><!><p>We extracted total RNA from HCC cells using TRIzol reagent (Life Technologies Corporation, Carlsbad, CA, USA) and determined RNA concentration and quality by the 260/280 nm ratio using a Nanodrop Spectrophotometer (ND‐2000, Thermo Fisher Scientific, Waltham, MA, USA). Next, RNA was reverse transcribed into cDNA and used for quantitative real‐time polymerase chain reaction (qRT‐PCR). We used glyceraldehyde 3‐phosphate dehydrogenase (GAPDH) as an internal control and expressed all of the results as the mean ± standard deviation (s.d.) of at least three independent experiments. Comparative quantification was determined using the 2−ΔΔCt method. The primer sequences were as follows: GAPDH‐Forward: GTCTCCTCTGAC TTCAACA and GAPDH‐Reverse: GTGAGGGTCTCTCTCTTCCT; IRX5‐Forward: ACCCAGCGTACCGGAAGAA and IRX5‐Reverse:CGGCGTCCAC GTCATTTTAT.</p><!><p>All tissues were fixed with 10% paraformaldehyde and embedded in paraffin wax. Paraffin sections were placed in incubators kept at 55 °C for 4 hours. The sections were immersed in two consecutive washes of xylol for 20 minutes to remove paraffin. Sections were then hydrated with different concentrations of ethanol including 100%, 95%, 85%, 70% and deionized water. Endogenous peroxidase was quenched with 3% H2O2 in methanol for 10 minutes and washed for 10 minutes in PBS. The sections were immersed in citrate buffer solution (0.01 mol l−1, pH 6.0) and heated to retrieve antigen. Then, 0.5%Triton‐x‐100 was incubated 30 minutes after washing with PBS. The tissues were blocked in 10% goat serum for 1 hour before the addition of the mouse monoclonal antibodies against IRX5 (diluted 1:150, Atlas,USA) and Ki‐67 (1:500, Abcam, Cambridge, UK) at 4 °C overnight. The sections were incubated with HRP‐conjugated goat anti‐mouse secondary antibody (diluted 1:1000, Bioword, USA) for 30 minutes and then with 3,3‐diaminobenzidine (DAB)/H2O2 for 5 minutes. Slides were imaged under a light microscope (Olympus, Japan) at ×400 magnification.</p><!><p>We extracted total proteins from the HCC cells using RIPA buffer and protease inhibitors (Solarbio, Beijing, China) according to manufacturer's protocol, and the protein concentrations were determined using a BCA assay kit (Solarbio, Beijing, China); protein were subjected to sodium dodecyl sulphate polyacrylamide gel electrophoresis and electrophoretically transferred to polyvinylidene fluoride membranes (Merck Millipore, MA, USA). Membranes were incubated in 5% nonfat milk dissolved in Tris‐buffered saline (TBS) containing 0.1% Tween‐20 for 2 hours at room temperature and then incubated with primary antibodies as follows: IRX5 (diluted 1:500, Abcam, Cambridge, MA, USA), GAPDH (diluted 1:2000, ProteinTech, Wuhan, China), Bcl‐2 (diluted 1:500, Affinity Biosciences, Changzhou, China), Bax (diluted1:2000,ProteinTech), cyclinD1 (diluted 1:500, Bioworld Technology, Nanjing, China) and p53 (diluted1:2000, ProteinTech).Then, the membranes were incubated with secondary antibodies (diluted 1:5000, Affinity Biosciences, Changzhou, China) at room temperature for 2 hours. Immunoblots were visualised by enhanced chemiluminescence (ECL kit; Advansta, Menlo Park, CA, USA) and scanned using an ECL chemiluminescence detection system (Pierce Biotechnology, Waltham, MA, USA). All experiments were performed in triplicate.</p><!><p>Cell cycle and cell apoptosis assays were assays by using flow cytometry assays (BD, Franklin Lakes, NJ). The cells were harvested and then washed twice with PBS and resuspended in 100 μL of binding buffer. The cells were fixed in 70% ice‐cold ethanol and after holding overnight at 4 °C, the cells were supplemented with RNaseA (Keygen Biotech) and propidium iodide for 37 °C for 30 minutes. The DNA content of labelled cells was detected using FACS cytometry (BD Biosciences Inc., Franklin Lakes, NJ, USA). Cell apoptosis was assessed by using a flow cytometry assay (BD, Franklin Lakes, NJ). Each experiment was performed in triplicate.</p><!><p>The short‐hairpin RNA targeting human IRX5 was ligated into the pGreenPuro™ shRNA vector (System Biosciences, Palo Alto, CA, USA) according to the manufacturer's protocol. Transfected SMMC7721 cells were selected with puromycin (2 μg ml−1) for 4 weeks. Selected cells were further subcloned for uniform stable cell lines. The stably interfering cell lines were identified using western blot.</p><!><p>Male BALB/c nude mice (4‐6 weeks old) were purchased from the Animal Care Committee of Chongqing Medical College. Ten mice were randomly allocated into two groups. A total of 1 × 107 SMMC7721 cells stably transfected with IRX5‐interference (sh‐IRX5) and negative control shRNA (sh‐NC) were subcutaneously injected into the dorsal flanks of mice. After 28 days, the mice were sacrificed, and tumours were removed and photographed. The expression of IRX5 and Ki‐67 in Xenograft tumours was detected by immunohistochemistry.</p><!><p>SPSS 17.0 software (SPSS Inc., Chicago, IL, USA) and GraphPad software (GraphPad Software, Inc., La Jolla, CA, USA) were used to analyse all data for statistical significance. Two‐tailed Student's t‐test was used for comparisons of two independent groups. A P value <0.05 was considered to indicate statistical significance.</p><!><p>Immunohistochemical (IHC) staining revealed that the IRX5 protein was differentially expressed between tumour tissues and their matched adjacent non‐tumour tissues. The non‐tumour tissues showed weak or no expression of IRX5 (Figure 1A, left panel). However, strong immunoreactivity for IRX5 was detected in the cell nucleus and in the cytoplasm of the tumor tissues (Figure 1A, right panel). Total IHC score of IRX5 in HCC tissues and non‐tumour tissues (n = 10, P < 0.01) (Figure 1B).</p><p>QRT‐PCR analysis and western blot analysis were used to investigate theIRX5 expression in HCC cell lines (HepG2, SMMC7721, SK‐hep1, Huh7) and a human immortalized normal liver cell line (L02).The results showed thatIRX5 expression was markably increased in HCC cells compared with that in L02 cells (Figure 1C,D).</p><!><p>To further understand the correlation between IRX5 expression and the proliferation and apoptosis capacities of HCC cells, we overexpressed IRX5 (pcDNA3.1‐IRX5) and knocked down (sh‐IRX5) via transient transfection in SMMC7721 and HepG2 cells. Western blotting confirmed that transfection with the IRX5 overexpression plasmid significantly increased IRX5 expression (Figure 2A,B). The efficiency of sh‐RNA‐mediated IRX5 knockdown was confirmed (Figure 2A,B).MTS assays showed that the overexpression of IRX5increased the proliferative capacities of SMMC7721 and HepG2 cells compared with control cells containing empty vector, whereas the opposite result was found when IRX5 expression was silenced (Figure 2C). Moreover, IRX5 overexpression enhanced the colony formation capacities of SMMC7721 and HepG2 cells, whereas knockdown of IRX5 reduced them (Figure 2D,E). Together, these data suggest that IRX5 enhanced HCC cell proliferation.</p><!><p>IRX5 was upregulated in HCC tissues and cell lines. (A) Immunohistochemical (IHC) staining of IRX5 in HCC tissues (T) and non‐tumour tissues (N). No. 1 was the representative micrograph. (B) Total IHC score of IRX5 in HCC tissues and non‐tumour tissues (n = 10). **P < 0.01. (C) IRX5 mRNA levels were detected by qRT‐ PCR in HCC cell lines(HepG2,huh7, SK‐hep1 and SMMC7721) and human immortalized, normal liver cell line (L02). *P < 0.05, **P < 0.01.Transcript levels were normal to GAPDH expression. (D) IRX5 protein levels in HCC cell lines(HepG2,huh7, SK‐hep1 and SMMC7721) and L02 were detected by western bloting. GAPDH was used as internal control. *P < 0.05, **P < 0.01</p><p>IRX5 promoted HCC cell proliferation in vitro. (A,B) Overexpression and knock‐down efficiency of IRX5. GAPDH was used as internal control. (C) Proliferation of HCC cells assessed by MTS assays. IRX5 overexpression enhanced HCC cell proliferation, whereas IRX5 interference repressed HCC cell proliferation. (D,E) Clone formation assay of differently treated HCC cells. Representative graphs are shown. The data graphs depict the count number from three independent experiments</p><!><p>To gain insights into the mechanism by which IRX5 enhances HCC cell proliferation and induces apoptosis, fluorescence‐activated cell sorting (FACS) was performed to analyse differences in cell cycle distributions and apoptosis after IRX5 overexpression or silencing. The results revealed that a reduction in the G1 population and an increase in the S and G2/M populations were observed in SMMC7721 cells overexpressing IRX5. Conversely, repressing IRX5 expression mainly led to an accumulation in G1 and decrease in S and G2/M phases (Figure 3A,C).</p><!><p>IRX5 induced cell‐cycle progression and suppressed apoptosis. (A,C) Cell‐cycle analysis of differently treated HCC cells. Representative graphs are shown. The data graphs depict the count number from three independent experiments. (B,D) Cell‐apoptosis analysis of differently treated HCC cells. Representative graphs are shown. The data graphs depict the count number from three independent experiments. (E,F) The expression of CyclinD1, P53, Bcl‐2 and Bax in cells were examined by western blotting analyses. GAPDH was used as internal control. *P < 0.05, **P < 0.01</p><!><p>Apoptosis was measured by FACS‐based annexin‐V/7‐AAD double staining in HCC cells under serum starvation conditions for 48 hours. The results revealed that the percentage of annexin V‐positive cells was lower in IRX5‐overexpressing SMMC7721 cells than in control cells. In contrast, the sh‐IRX5 group had a significantly higher percentage of annexin V‐positive cells than the sh‐NC group (Figure 3B,D).</p><p>Furthermore, we also examined the levels of several key genes involved in cell cycle checkpoints by western blot analysis in SMMC7721 and HepG2 cells stably with silenced IRX5 expression. Knockdown of IRX5 reduced the expression of the oncogenic cell cycle regulator cyclinD1 but increased the expression of the cyclin‐dependent protein kinase inhibitor p53 (Figure 3E,F).</p><p>To investigate the role of IRX5 in the p53 signalling pathway, the expression levels of Bax and Bcl‐2 were detected. As shown in Figure 3E,F, increased Bax and decreased Bcl‐2 expressions were observed following downregulation of IRX5 expression. Taken together, these results indicated that IRX5‐induced cell cycle progression by upregulating cyclinD1 and inhibited apoptosis by inactivating the p53 signalling pathway in HCC cells.</p><!><p>To determine the growth‐enhancing effect of IRX5 in vivo, we injected SMMC7721 cells stably transfected with sh‐IRX5 or sh‐NC subcutaneously into nude mice for xenotransplantation. Mice injected with IRX5‐silenced cells showed significantly decreased tumour growth compared to those injected with cells transfected with sh‐NC (Figure 4A). As assessed by measurements of tumour volume and weight (Figure 4B,C), the knocked down of IRX5 expression significantly inhibited overall tumour growth. The immunohistochemistry analysis of the tumour tissues from xenografts revealed that the expression of the cell proliferation marker, Ki‐67, was significantly weaker in xenografts of sh‐IRX5 cells than in xenografts of sh‐NC cells (Figure 4D,E). These results showed that knockdown of IRX5 suppressed HCC cell proliferation in vivo.</p><!><p>IRX5 silence inhibited HCC growth in vivo. (A) The subcutaneous tumour model of stable IRX5‐interference SMMC7721 cells (n = 5 for both groups). (B,C) Tumour growth and tumour weights curves were analysed. (D,E) Immunohistochemistry analysis of ki‐67and IRX5 were obtained from tumours (magnification, 400×; scale bar, 50 μm)</p><!><p>Our study demonstrates that IRX5 is upregulated in HCC cell lines and HCC tissues. Furthermore, IRX5 promoted cell proliferation in vitro and in vivo. In addition, IRX5 suppressed apoptosis in vitro. IRX5 plays a different role in multiple cancers, contributing to the development of many tumours by acting as an important transcription factor regulating key regulatory genes that control cell growth, invasion, migration and apoptosis. IRX5 activity is also present in the stromal and proliferative late blastemal/early epithelial cells in developing kidneys and Wilms tumours.16 Myrthue et al had revealed that knockdown of IRX5 by RNA interference significantly reduction in LNCaP cell viability, which resulted in increased LNCaP cell apoptosis and was partially mediated by p53.17 However, the role of IRX5 in HCC is unknown. In this study, we provided the first evidence that IRX5 is significantly upregulated in HCC cell lines and tissues. We further identified the effects of IRX5 on the biological behaviours of HCC cells, showing that IRX5 promoted HCC cell proliferation and inhibited apoptosis. The results indicated that IRX5 acts as an oncogene in HCC. Moreover, the in vivo studies also confirmed that knockdown of IRX5 suppressed tumour growth in nude mice, suggesting that IRX5 could potentially be applied in the treatment of HCC. However, the underlying mechanisms by which IRX5 promotes tumour cell proliferation and inhibits apoptosis remain unclear.</p><p>We provided evidence that IRX5 promoted HCC cell proliferation in vitro and in vivo. Furthermore, our study found that a reduction in the G1 population and an increase in the S and G2/M populations were observed in SMMC7721 cells overexpressing IRX5. Mechanistically, IRX5 was shown to induce cell cycle progression by upregulatingcyclinD1. It has been well defined that tumour‐associated cell cycle defects are often mediated by the accumulation of cyclins (CCNs).18 Cyclins are divided into two groups known as the G1/S cyclins, which are essential for the control of the cell cycle at the G1 to S transition, and the G2/M cyclins, which control the cell cycle at the G2 to M transition phase.19 Cyclin D1 is a major regulator of the cell cycle and is responsible for the G1/S‐phase transition.20 Ao et al reported that cyclin D1 was overexpressed in liver cancer cells and promoted migration and invasion by regulating several enzymes.21 Chen et al also reported that HCC tissues and HCC cells exhibited elevated expression levels of cyclin D1 and its expression levels were found to be correlated with tumour size and tumour staging.22 A previous study reported that IRX5 promotes NSCLC cell proliferation by means of regulating the CCND1 promoter.9 In this study, cyclin D1 expression was positively correlated with the expression of IRX5 in HCC cells.</p><p>In addition to enhanced proliferation, resistance to apoptosis is also a hallmark of cancer cells.23 The p53 gene is located on human chromosome 17p13 and is a tumour suppressor and pro‐apoptosis gene. Wild‐type p53 plays a key role in cell gene transcription, cell cycle regulation, apoptosis, proliferation and differentiation.24, 25, 26 The amount of wild‐type p53 protein is low in normal cells.27 The dominant components of p53 signalling, including p53, Bcl‐2 and Bax, have been extensively studied in carcinomas.28, 29 A previous study reported that p53 positively regulates Bax expression30 and negatively regulates the transcription of Bcl‐2.15, 30 Thus, p53 is likely to affect upstream pro‐apoptotic proteins to modulate their functions in the cytoplasm. We also found that the downregulation of IRX5 significantly increased the expression of p53 and Bax and decreased the expression of Bcl‐2 in HCC cells. However, in our study, we did not study the mechanism by whichIRX5 inhibits the p53 signalling pathway. Such mechanisms are left to be investigated in future studies.</p><p>In summary, our study demonstrated that IRX5 is a potential tumour promoter gene in HCC. IRX5 promotes to the promoter of carcinogenesis by facilitating cell proliferation and suppressing cell apoptosis by inhibiting the p53 signalling pathway. As a result, IRX5 might act as a novel molecular target for the detection and treatment of HCC. However, further studies are needed to investigate the effect of IRX5 on other cellular processes in HCC, such as cell adhesion and differentiation.</p><!><p>Our results showed that IRX5 was upregulated in HCC and cells. Overexpression of IRX5 promoted cell proliferation and tumourigenicity. Knockdown of IRX5 promoted cell apoptosis through the p53 signalling pathway in HCC.</p><!><p>The authors have declared that no competing interests exist.</p>

PubMed Open Access

Skeletocutins M–Q: biologically active compounds from the fruiting bodies of the basidiomycete <i>Skeletocutis</i> sp. collected in Africa

During the course of screening for new metabolites from basidiomycetes, we isolated and characterized five previously undescribed secondary metabolites, skeletocutins M-Q (1-5), along with the known metabolite tyromycin A (6) from the fruiting bodies of the polypore Skeletocutis sp. The new compounds did not exhibit any antimicrobial, cytotoxic, or nematicidal activities. However, compound 3 moderately inhibited the biofilm formation of Staphylococcus aureus (S. aureus), while compounds 3 and 4 performed moderately in the ʟ-leucine-7-amido-4-methylcoumarin (ʟ-Leu-AMC) inhibition assay. These compounds represent the first secondary metabolites reported to occur in the fruiting bodies by Skeletocutis. Interestingly, tyromycin A (6) was found to be the only common metabolite in fruiting bodies and mycelial cultures of the fungus, and none of the recently reported skeletocutins from the culture of the same strain were detected in the basidiomes.

skeletocutins_m–q:_biologically_active_compounds_from_the_fruiting_bodies_of_the_basidiomycete_<i>sk

Introduction<!>Results and Discussion<!>Conclusion<!>Experimental General information<!>Fungal material<!>Extraction of the crude extract<!>Isolation of compounds 1-6<!>Cytotoxicity assay<!>Inhibition of biofilm formation<!>Nematicidal activity assay<!>Inhibition of leucine aminopeptidases

<p>Over the past years, we have been studying the secondary metabolites of African Basidiomycota that were collected in rainforests and mountainous areas of Western Kenya. These species were new to science, and proved to be a prolific source of unprecedented natural compounds showing a set of prominent biological activities [1][2][3].</p><p>The present study deals with the comparison of the secondary metabolites located in the basidiomes (fruiting bodies) of another, putatively new species belonging to the genus Skeletocutis, strain MUCL56074. We have recently reported the known metabolite tyromycin A (6), together with 12 unprecedented congeners for which we proposed the trivial names skeletocutins A-L, which were obtained from a liquid culture of the same fungus [4]. A preliminary characterization of the producer organism suggested that it belongs to a new species because neither DNA sequence data in the public domain nor morphological characteristics matched the previously reported species, as compared to the literature. The genus Skeletocutis (of the Polyporaceae) consists of approximately 40 species, which grow as a crust on the surface of collapsing wood [5] and mostly occur in the temperate climate zones.</p><p>In our preceding study, the fungal specimen MUCL56074 has been assigned to the genus Skeletocutis by comparison of morphological features and 5.8S/ITS rDNA sequences, as reported previously [4]. Strain MUCL56074 represents a hitherto undescribed species, which will be formally described in a separate paper in a mycological journal, pending the examination of type material of related species. In view of a potential application of chemotaxonomic methodology, the basidiomes of the fungus were checked for the presence of secondary metabolites for later comparison with herbarium specimens of other species by HPLC-diode array detection (HPLC-DAD)-MS. Surprisingly, we detected further members of the skeletocutin family that were not present in the cultures. The current paper is dedicated to the description of their isolation as well as biological and physicochemical characterization.</p><!><p>The fruiting bodies of the fungal specimen MUCL56074 were extracted with acetone and subsequently purified via preparative HPLC, which led to the isolation of five previously undescribed secondary metabolites, 1-5 (t R = 17.8, 18.8, 15.7, 14.0, and 14.3 min, respectively), and one known compound, namely tyromycin A (6, t R = 16.8 min) [6] (Figure 1).</p><p>Compound 1 (Table 1 and Figure 2), named skeletocutin M, was isolated as a yellow solid. Its molecular formula was determined to be C 28 H 42 O 6 , with eight degrees of unsaturation, by HRESIMS. Signals for m/z = 475.3054, 497.2868, and 971.5839, corresponding to the ions [M + H] + , [M + Na] + , and [2M + Na] + , respectively, were also recorded in the mass spectrum. A singlet resonating at δ = 2.08 ppm for the methyl protons H 3 -6′ and a triplet and quinted resonating at δ = 2.50 and 1.59 ppm, respectively, for the methylene groups, were recorded in the 1 H NMR spectrum. Further, the 13 C NMR spectrum revealed only 14 signals instead of 28, as indicated by the molecular formula, suggesting that the molecule consisted of two identical halves.</p><p>Determination of HMBC correlations between the C-6′ methyl protons (δ = 2.08 ppm) and C-3′/4′/5′ as well as between H 2 -1/18 and C-2′/3′/4′ confirmed the presence of a maleic anhydride moiety in the molecule. Integration of the singlet for the C-6′ methyl group in the 1 H NMR spectrum gave a value of 6, indicating the presence of two maleic anhydride functions. The multiplet at δ = 1.28-1.36 ppm was assigned to the remaining 14 methylene units making up the carbon chain. Integration of this multiplet gave a value of 28, confirming the length of the chain. The connection of this chain to two maleic anhydride units was confirmed by HMBC correlations between the protons H 2 -1/2/17/18 and C-3′ (δ = 144.9 ppm). Additionally, long-range correlations between the protons H 2 -1 and H 3 -6′ were observed in the COSY spectrum. Therefore, the structure of natural product 1 was unambigously concluded to be 1,18bis[4′-methyl-2′,5′-dioxo-3′-furyl]octadecane.</p><p>Compound 2 (skeletocutin N, Table 1 and Figure 2) was obtained as a white solid, with the molecular formula C 30 H 46 O 6 and eight degrees of unsaturation determined from HRESIMS data. The 1D and 2D NMR data for 2 revealed a similar structure to 1, with the difference being the size of the carbon chain in the molecule. A value of 32 obtained from the integration of the C-3-C-18 multiplet (δ = 1.26-1.32 ppm) led to the conclusion that 2 had an icosane chain instead of the octadecane chain elucidated for skeletocutin M (1).</p><p>Compound 3 (skeletocutin O, Table 2 and Figure the olefinic bond between C-3 and C-4 was assigned (E)-configuration. As such, the structure of 3 was concluded to be (E)-2-(19-(4'-methyl-2',5'-dioxo-2',5'-dihydrofuran-3'-yl)nonadecylidene)butanedioic acid.</p><p>Compound 4 (Table 2 and Figure 2), named skeletocutin P, was isolated as a white solid. Compound 5 (skeletocutin Q, Table 2 and Figure 2), with the molecular formula C 29 H 46 O 9 and seven degrees of unsaturation, as established from HRESIMS data, was obtained as a yellow solid. Analysis of 1D and 2D NMR data of 5 indicated a similar structure to 4, with saturation of the olefinic bond between C-4 and C-5. In the 13 C NMR spectrum of 5, the signals that had occurred at δ = 144.7 and 131.9 ppm for compound 4 were missing, and instead, a methylene signal at δ = 28.5 ppm (C-5) and a methine signal at δ = 46.1 ppm (C-4) were recorded. HMBC correlations were observed between H-4 (δ = 2.50 ppm) and C-2/3/23/24 as well as H 2 -5 (δ = 1.56 ppm) and C-4/C6/24. Furthermore, COSY correlations between H-3 and H 2 -2/H-4 as well as H 2 -5 and H-4/H 2 -6 could be recorded. Hence, 5 was concluded to be 21-(4'-methyl-2',5'-dioxo-2',5'dihydrofuran-3'-yl)henicosane-1,2,3-tricarboxylic acid.</p><p>Tyromycin A (6), a closely related compound to the metabolites 1-5, has been reported before, and was isolated from the cultures of the same fungus (i.e., the corresponding mycelial culture of the specimen that was the subject of the present study [4]) and originally from Tyromyces lacteus [6]. In these two cases, 6 was reported to be the major component of the culture extracts. Even though this compound is occurring in fruiting bodies, in this case, skeletocutin M (1) was the major component instead of tyromycin A (6). The two molecules 1 and 6 differ in their chain length, with the former having an 18-carbon chain instead of a 16-carbon chain, as in 6.</p><p>The isolated compounds 1-6 were evaluated for antimicrobial, cytotoxic, and nematicidal activities, as described in the Experimental section. However, compounds 1-5 were devoid of activity in these assays, whereas tyromycin A (6) and skeletocutin A-L had been reported before to be active against several Gram-positive bacteria [4], namely Bacillus subtilis (B. subtilis), S. aureus, methicillin-resistant S. aureus (MRSA), and Micrococcus luteus (M. luteus). In the antimicrobial assay, compounds 3 and 5 were observed to interfere with the forma-tion of biofilms commonly associated with S. aureus. When compounds 3 and 5 were evaluated for biofilm inhibition activity against S. aureus, they showed only weak activity with 20 and 56% inhibition of the biofilm, respectively, at a concentration 256 µg/mL. Tyromycin A ( 6) was previously reported to be an inhibitor of leucine aminopeptidase in HeLa S3 cells [6]. Accordingly, all compounds 1-5 were tested for their inhibition activity against hydrolysis of ʟ-leucine-7-amido-4-methylcoumarin (ʟ-Leu-AMC). Compound 4 exhibited moderate activity, with an IC 50 value of 71.1 µg/mL (Table 3 and Figure 3) when 50 µM of the substrate was used. Compounds 3 and 5 exhibited weak activities, with IC 50 values of >80 µg/mL at 50 and 110 µM substrate concentration. Although tyromycin A ( 6) was previously reported to be active in a similar assay against HeLa S3 cells, with IC 50 values of 31 μg/mL at 50 μM substrate concentration, an IC 50 value >150 μg/mL for this compound was recorded on the HeLa (KB3.1) cells during this cytotoxicity study [6].</p><!><p>In summary, five previously undescribed tyromycin A derivatives 1-5 could be isolated from Skeletocutis sp. fruiting bodies. These metabolites are closely related to the skelotocutins that were previously reported as isolates from liquid cultures. Compounds 3 and 5 were observed to weakly inhibit the biofilm for-mation by S. aureus and constrain inhibitory activity of ʟ-Leu-AMC hydrolysis in KB 3.1 cells. There have been relatively few studies on the production of secondary metabolites in mycelial cultures vs fruiting bodies in higher fungi, but so far, there are only few examples where the same compounds were predominant in both. For instance, in most species hitherto studied of the ascomycete order Xylariales, the fruiting bodies and cultures mostly showed a complementary secondary metabolite production [7]. In the current case, it appears that the basidiomes of Skeletocutis can be used for chemotaxonomic studies. Investigations of herbarium specimens may not even be helpful for the taxonomic revision of the genus but may even lead to the discovery of further, previously undescribed members of the tyromycin/skeletocutin type.</p><!><p>NMR spectra were recorded with a Bruker 500 MHz spectrometer at frequencies of 500.130 ( 1 H NMR) and 125.758 MHz ( 13 C NMR). HRESIMS spectra were recorded after purification with an Agilent 1200 series HPLC-UV system (column size: 2.1 mm⋅50 mm, packing: 1.7 µm, Waters ACQUITY UPLC BEH C18 sorbent, solvent A: H 2 O + 0.1% formic acid, solvent B: acetonitrile + 0.1% formic acid, elution gradient: 5% solvent B for 0.5 min, increasing solvent B to 100% within 19.5 min, 100% solvent B for 5 min, flow rate: 0.6 mL/min, UV-vis detection at λ = 200-600 nm) and ESI-TOF-MS analysis (maXis™ system, Bruker, scan range: 100-2500 m/z, capillary voltage: 4500 V, drying temperature: 200 °C). UV-vis spectra were recorded with a Shimadzu UV-2450 UV-vis spectrophotometer. The chromatogram in Figure 1 was recorded on a Bruker Agilent 1260 Infinity Series equipped with DAD and an ESI ion trap mass spectrometer (amaZon speed ion trap, Bruker).</p><!><p>The fungal specimen was collected by C. Decock and J. C. Matasyoh in the Mount Elgon National Reserve [4]. The dried specimen and corresponding cultures were deposited in the MUCL collection (MUCL accession number: 56074).</p><!><p>A quantity of 9.8 g fruiting bodies were extracted using 500 mL of acetone overnight. Then, the extract was filtered and another 500 mL of acetone were added. This was extracted in an ultrasonic bath for 30 min. The extracts were combined and the solvent evaporated to afford 226 mg of crude extract.</p><!><p>The crude extract (vide supra) was filtered using solid-phase microextraction (SPME) with a Strata™-X 33 µm Polymeric Reversed Phase (RP) cartridge (Phenomenex). The extract was fractionated by preparative RP chromatography (RPC) using a PLC 2020 purification system (Gilson). A VarioPrep (VP) column system packed with NUCLEODUR ® 100-5 C18 ec was used as stationary phase (Machery-Nagel, column size: 25 mm⋅40 mm, packing: 7 µm). Deionized water, obtained from a Milli-Q ® water purification system (Millipore), + 0.05% TFA (solvent A) and acetonitrile + 0.05% TFA (solvent B), respectively, were used as mobile phases (elution gradient</p><!><p>In vitro cytotoxicity, using IC 50 values as a measure, was evaluated against mouse fibroblasts cell line L929 and HeLa (KB3.1) cells and carried out according to our previous reports [8,9].</p><!><p>The assay was performed in Falcon ® 96-well flat bottom plates as previously described [10]. S. aureus DSM1104 was enriched overnight to reach 0.5 McFarland standard turbidity in caseinpeptone soymeal-peptone (CASO) medium containing 4% glucose at pH 7.0 for biofilm formation. Methanol was used as negative control, while tetracycline was used as positive control. All experiments were made in triplicates.</p><!><p>The nematicidal activity of isolated compounds against Caenorhabditis elegans (C. elegans) was performed in 24-well microtiter plates as previously described [11]. Ivermectin was used as positive control and methanol was used as negative control. The results are expressed as LD 90 values.</p><!><p>Hydrolysis of ʟ-Leu-AMC by the surface-bound aminopeptidases of KB3.1 cells was assayed based on the method from Weber and co-workers [6] with slight modifications. KB3.1 cells were grown as monolayer cultures in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal calf serum at 37 °C in 24-well multidishes. After three days, the confluent monolayers were washed twice with phosphate buffered saline (PBS), and the reaction mixture (450 µL Hanks' buffer at pH 7.2 containing 50 or 100 µM substrate ʟ-Leu-AMC, compounds dissolved in 50 µL DMSO) was added. After being incubated at 23 °C for 30 min, 1 mL cold 0.2 M glycine buffer at pH 10.5 was added. The amount of hydrolyzed 7-amino-4methylcoumarin (AMC) was measured in a Tecan Infinite M200 PRO fluorescence spectrophotometer (excitation and emission at λ = 365 and 440 nm, respectively). Bestatin [12] and DMSO were used as positive and negative control, respectively.</p>

Beilstein

Reactivity of Hydrogen-Related Electron Centers in Powders, Layers, and Electrodes Consisting of Anatase TiO2 Nanocrystal Aggregates

Anatase TiO2 nanoparticle aggregates were used as model systems for studying at different water activities the reactivity of electron centers at semiconductor surfaces. The investigated surface conditions evolve from a solid/vacuum interface to a solid/bulk electrolyte interface. Hydrogen-related electron centers were generated either chemically—upon sample exposure to atomic hydrogen at the semiconductor/gas interface—or electrochemically—upon bias-induced charge accumulation at the semiconductor/electrolyte interface. Based on their corresponding spectroscopic and electrochemical fingerprints, we investigated the reactivity of hydrogen-related electron centers as a function of the interfacial condition and at different levels of complexity, that is, (i) for dehydrated and (partially) dehydroxylated oxide surfaces, (ii) for oxide surfaces covered by a thin film of interfacial water, and (iii) for oxide surfaces in contact with a 0.1 M HClO4 aqueous solution. Visible (Vis) and infrared (IR) spectroscopy evidence a chemical equilibrium between hydrogen atoms in the gas phase and—following their dissociation—electron/proton centers in the oxide. The excess electrons are either localized forming (Vis-active) Ti3+ centers or delocalized as (IR-active) free conduction band electrons. The addition of molecular oxygen to chemically reduced anatase TiO2 nanoparticle aggregates leads to a quantitative quenching of Ti3+ centers, while a fraction of ∼10% of hydrogen-derived conduction band electrons remains in the oxide pointing to a persistent hydrogen doping of the semiconductor. Neither trapped electrons (i.e., Ti3+ centers) nor conduction band electrons react with water or its adsorption products at the oxide surface. However, the presence of an interfacial water layer does not impede the electron transfer to molecular oxygen. At the semiconductor/electrolyte interface, inactivity of trapped electrons with regard to water reduction and electron transfer to oxygen were evidenced by cyclic voltammetry.

reactivity_of_hydrogen-related_electron_centers_in_powders,_layers,_and_electrodes_consisting_of_ana

Introduction<!>Chemicals<!>Chemical Vapor Synthesis of Anatase TiO2 Nanoparticle Powders<!>Ensembles of Anatase TiO2 Nanoparticle Aggregates<!>Aggregate Powder<!>Aggregate Layers<!>Aggregate Films/Electrodes<!>Diffuse Reflectance Vis-Spectroscopic Study of Loose Aggregate Powders and Aggregate Films<!>IR Spectroscopy Study of Immobilized Aggregate Layers<!>Chemical Reduction of Loose Aggregate Powders and Aggregate Layers by Atomic Hydrogen<!>Cyclic Voltammetry Study of Immobilized Aggregate Films/Electrodes<!>Chemical Charge Accumulation and Electron Transfer at the Solid/Gas Interface<!>Chemical Reduction of TiO2 Aggregate Powder by Atomic Hydrogen: Vis-Active Ti3+ Centers<!><!>Chemical Reduction of TiO2 Aggregate Powder by Atomic Hydrogen: Vis-Active Ti3+ Centers<!><!>Chemical Reduction of TiO2 Aggregate Powder by Atomic Hydrogen: Vis-Active Ti3+ Centers<!>Reactivity of Ti3+ Centers in the Presence of Interfacial Water<!>Chemical Reduction of TiO2 Aggregate Layers by Atomic Hydrogen: IR-Active Conduction Band Electrons<!><!>Chemical Reduction of TiO2 Aggregate Layers by Atomic Hydrogen: IR-Active Conduction Band Electrons<!><!>Chemical Reduction of TiO2 Aggregate Layers by Atomic Hydrogen: IR-Active Conduction Band Electrons<!>Reactivity of Conduction Band Electrons in the Presence of Interfacial Water<!><!>Reactivity of Conduction Band Electrons in the Presence of Interfacial Water<!>Electrochemical Charge Accumulation and Electron Transfer at the Solid/Electrolyte Interface<!><!>Electrochemical Charge Accumulation and Electron Transfer at the Solid/Electrolyte Interface<!>General Discussion<!>Conclusions<!>

<p>Charge transfer across solid/solid, solid/liquid, or solid/gas interfaces in nanostructured semiconductor materials is exploited in several technologies. The thermodynamic and kinetic details of the charge transfer depend critically on the chemical and structural properties of the interfaces involved. Under application-relevant conditions (e.g., in the presence of a surrounding electrolyte), interfacial conditions of these high surface area materials strongly differ from the situation at solid/vacuum or solid/gas interfaces, which are frequently investigated in model studies. At the same time, the elucidation of the impact of interfacial composition on the energetics and dynamics of interfacial charge transfer is of prime importance for the optimization of materials' functional properties. However, related knowledge gain is challenging and requires the use of well-defined and tunable model systems, the study of relevant model reactions, and the availability of suitable analytical tools.</p><p>The reactivity of electron centers in semiconductor materials is exploited in many technological applications, including photocatalysis and electrocatalysis. Reactive electrons can be generated in a semiconductor nanostructure via different processes. Electron–hole pairs, for instance, are formed in the semiconductor bulk upon supra band gap excitation. This physical process constitutes one prerequisite for a subsequent photocatalytic event at the semiconductor/gas or semiconductor/electrolyte interface.1 Alternatively, charge can be injected into the semiconductor via chemical,2 photochemical,3 or electrochemical processes4 as exploited in sensitized photoelectrochemical cells, batteries, or electrochromic devices.</p><p>Extensive theoretical and experimental efforts have been made to characterize impurity donors such as hydrogen in metal oxides.5−7 In this context, infrared (IR) spectroscopy constitutes a very useful method, as it allows for the detection not only of hydrogen vibrational modes, but also of free and shallow trapped electrons.7−10 A broad signal in the IR range, monotonically increasing toward lower wavenumbers, was observed under high-vacuum conditions after exposure of TiO2 nanoparticles to atomic hydrogen or upon hydrogen dissociation and spillover on Au/TiO2 nanoparticles, respectively.11,12 In both cases, hydrogen atoms were expected to diffuse into the TiO2 bulk and to donate an electron to shallow trapped states just below the conduction band. Electron excitation from these states to the conduction band and inter-conduction-band transitions (Drude-type absorption) were proposed to contribute to the broad IR signal.11,12 Interestingly, such IR signals were detected not only under high-vacuum conditions but also upon band gap excitation of TiO2 in contact with aqueous solutions of hole acceptors13,14 and upon cathodic polarization in acidic aqueous electrolytes.14−17</p><p>In addition to the IR signal, a broad absorption in the visible range is observed upon charge accumulation in TiO2 nanoparticle ensembles and has been interpreted in terms of d-d transitions of Ti3+ centers, that is, electrons localized in band gap states.17−24 Electron paramagnetic resonance spectroscopy has evidenced the presence of Ti3+ species in TiO2 nanoparticles after negative polarization in acidic aqueous solution19 and after reductive treatment with atomic hydrogen.25 In line with these interpretations, calculations have confirmed that the exposure of TiO2 to atomic hydrogen produces Ti3+ species as a result of H atom dissociation into a proton, bound to a lattice oxygen, and an extra electron.5</p><p>Electrochemical accumulation of electrons in semiconductor electrodes in contact with aqueous electrolytes is related to H atom ionization in vacuum insofar as electron localization and proton adsorption/intercalation occur in parallel in both cases.26 In acidic electrolytes, electrochemical electron accumulation is compensated mainly by proton adsorption at the oxide surface271</p><p>Because of the small size of protons, charge compensation may take place also via ion insertion into subsurface regions of the nanocrystals. In such a case, proton diffusion in the oxide bulk is the rate-determining process in both charging and discharging, possibly leading to a transient doping of the semiconductor (electrochemical hydrogen doping).15,16,28</p><p>In the present study, we exploit the donor properties of atomic hydrogen to chemically charge under high-vacuum conditions powders and immobilized layers of TiO2 nanoparticle aggregates and evaluate the chemical reactivity of accumulated electrons toward acceptor species.2,25 In particular, atomic hydrogen is used to generate charged states on clean surfaces while preserving the metal-to-oxygen ratio of the semiconductor. This allows for studying in a systematic way the impact of the interface condition on electron transfer reactions via a stepwise increase of interface complexity starting from a well-defined reference. Aiming at a stepwise build-up of a charged semiconductor/electrolyte interface, we adsorbed water onto the surface of chemically reduced TiO2 aggregates and studied electron transfer reactions in the presence of a thin interfacial water film. Finally, we complemented the study of electron transfer processes at the solid/gas interface by evaluating the reactivity of electrochemically accumulated electron/proton centers. Using hydrogen-related electronic states, we thus probe the electronic properties of defects in semiconductor nanostructures and their reactivity in different environments and at different levels of complexity (including well-defined model conditions, that is, high-vacuum and application-relevant conditions, i.e., aqueous electrolytes). In addition, careful sample synthesis and processing allows us to make the same nanoparticle-based material accessible to analysis by different analytical methods thereby gaining a comprehensive view of the properties of hydrogen-related electron centers in the same material but at very different water activities. The strategy followed in this paper aims at making a step toward bridging the gap between model studies and application. This is a challenging task for all sample types and especially for high surface area, nanoparticle-based materials.</p><!><p>Titanium(IV)isopropoxide (99.999%) and perchloric acid (70% w/w in water) were purchased from Sigma Aldrich and used without further purification. Ultrapure water (18 MΩ cm) was obtained using a Milli-Q water purification system (Merck Millipore) and was cleaned from dissolved gases either using the freeze–pump–thaw method (visible (Vis) and IR-spectroscopy investigation of water adsorption onto TiO2 powders and layers) or by bubbling N2 through the aqueous electrolyte (cyclic voltammetry).</p><!><p>Anatase TiO2 nanocrystals were prepared by metal–organic chemical vapor synthesis (MOCVS) based on the decomposition of titanium(IV) isopropoxide at T = 1073 K in a hot wall reactor system.29,30 For purification, the obtained powder samples were subjected to thermal treatment under high-vacuum conditions (p < 10–5 mbar). First, the powder sample was heated to T = 600 °C using a rate of r ≤ 5 °C min–1. Subsequent oxidation with O2 at this temperature was applied to remove organic remnants from the precursor material and to guarantee the stoichiometric composition of the oxide.</p><!><p>The resulting particle powder was used as the precursor for slurry preparation. The TiO2 nanoparticle powder (0.2 g) was ground in ultrapure water (Millipore, 18.2 MΩ cm, 1.28 mL) in the absence of any additives to avoid the adsorption of organic molecules on the high surface area material. The carbon content of water-treated and subsequently dried anatase TiO2 nanocrystal samples, which result from the adsorption of ubiquitous carbon-containing species during sample handling in ambient air and in water, was estimated in a previous study to correspond to ∼0.3% of a monolayer at the surface of the nanoparticles.31</p><p>Previous studies31,32 have shown that both the size and the crystal structure of vapor-phase grown and thermally processed anatase TiO2 nanoparticles are preserved, if the particle powder is dispersed in pure water, dried at room temperature, and finally annealed at T ≤ 450 °C in air. Aggregate formation was therefore performed for all sample types (i.e., powders, layers, and films/electrodes) at T = 450 °C in air (1 h). Thermal processing at this temperature results in the formation of particle/particle interfaces imparting electronic conductivity to the particle network.32</p><!><p>For the preparation of a loose aggregate powder, the aqueous nanoparticle slurry was placed in a porcelain dish, dried at room temperature, and thermally annealed at T = 450 °C in air. The resulting sample was carefully ground in a mortar to homogenize the aggregate powder.</p><!><p>Supported nanoparticle aggregates were prepared by spreading the aqueous nanoparticle slurry onto a tungsten mesh (Alfa Aesar, tungsten gauze, 100 mesh woven from 0.0509 mm diameter wire). The sample was then dried at room temperature and thermally annealed at T = 450 °C in air. After sintering, the film thickness accounts for 300 ± 50 μm.</p><!><p>The aqueous nanoparticle slurry was spread by doctor blading onto fluorine doped tin oxide (FTO) coated glass (Pilkington TEC 8, resistance 8 Ω/□). The resulting films were dried at room temperature and thermally annealed in air at T = 450 °C. After sintering, the mean film thickness accounts for 10 ± 3 μm. A copper wire was attached to the conducting substrates with silver epoxy. The contact area and the uncovered parts of the substrate were finally sealed by epoxy resin.</p><!><p>The aggregate powder was filled into an alumina ceramic boat and placed in a quartz glass tube. Alternatively, aggregate films deposited on FTO-coated glass (i.e., electrodes) were directly inserted into the quartz glass tube. The tube was connected to a dedicated high-vacuum system, which allows in addition to Vis-spectroscopic experiments in diffuse reflectance mode for (i) thermal sample activation under high-vacuum conditions (p ≤ 10–5 mbar) and in oxygen atmosphere, (ii) chemical reduction of the samples by atomic hydrogen, and (iii) sample exposure to defined gas and vapor atmospheres (Figure S3a).</p><p>The process of aggregate formation as described in Section 2.3 and sample transfer into the high-vacuum reactor are carried out under atmospheric conditions. In order to remove adsorbates from the aggregates' surface, samples were once more subjected to a thermal activation procedure under high-vacuum conditions as well as in a defined oxygen atmosphere. In particular, samples were heated under high-vacuum conditions to T = 450 °C (temperature ramp: 10 °C·min–1) and thermally annealed at this temperature under high-vacuum conditions (30 min) and in an oxygen atmosphere (p[O2] = 100 mbar, 30 min). Afterward, the sample was cooled down to room temperature and then high-vacuum conditions were reestablished. We will refer to a sample, which was subjected to such an activation procedure as an activated sample. Activated samples were used as the reference for all Vis-spectroscopic experiments, that is, all Vis spectra are referenced to the spectrum of an activated aggregate powder or film, respectively. Vis spectra were recorded in diffuse reflectance mode using a fiber optic system consisting of an HR4000 spectrometer (Ocean Optics) and a 150 W Xe lamp (Oriel Instruments). Five scans (integration time: 6 s) were accumulated in order to obtain spectra with a reasonable signal-to-noise ratio resulting in a sampling rate of 2 spectra per minute.</p><!><p>For transmission Fourier-transform infrared spectroscopy, a high-vacuum cell developed by J. T. Yates Jr. and co-workers33 was used and for this purpose aligned in the optical path of the IR beam of a Bruker Tensor 27 spectrometer system and connected to an atomic hydrogen generator unit (Figure S3b). The resolution was 4 cm–1, and 100 interferogram scans were averaged to guarantee a reasonable signal-to-noise ratio.</p><p>IR spectra were recorded at room temperature. Bands between 3000 and 2800 cm–1, which result from organic contamination of the spectrometer's optical components, were observed and represent sample-independent artifacts. Therefore, corresponding data points in this spectral region have been removed and replaced by a dashed line.</p><p>The tungsten mesh carrying the aggregate layer was inserted into the IR spectroscopy reactor. After the establishment of high-vacuum conditions, the supported aggregate layer was activated (following the protocol described in Section 2.4), that is, samples were heated (by resistive heating of the supporting tungsten mesh) under high-vacuum conditions to T = 450 °C (temperature ramp: 10 °C·min–1) and thermally annealed at this temperature under high-vacuum conditions (30 min) and in an oxygen atmosphere (p[O2] = 100 mbar, 30 min). Afterward, the sample was cooled down to room temperature and then high-vacuum conditions were reestablished. Again, we will refer to a sample, which was subjected to such an activation procedure as an activated sample.</p><!><p>Atomic hydrogen was generated in the Vis- and IR-spectroscopy reactors via homolytic splitting of molecular hydrogen at the surface of a coiled tungsten filament at T ∼ 2000 K (hydrogen background pressure p[H2] = 10–3 mbar).33 The sample was optically shielded from the heated tungsten filament to avoid sample heating when operating the tungsten coil. The temperature of aggregate layers was monitored in situ (via a thermocouple connected to the tungsten mesh carrying the sample) and no significant temperature increase was observed upon sample exposure to atomic hydrogen. For all samples reported here, we did not observe any spectral changes (in the Vis and IR) upon sample exposure to molecular hydrogen (i.e., when the tungsten filament was not heated).</p><!><p>Measurements were performed in a standard three electrode electrochemical cell. Electrolytes were purged from O2 by bubbling N2 through the electrolyte (i.e., 0.1 M HClO4 aqueous solution). Alternatively, O2 was bubbled through the electrolyte to maximize the concentration of dissolved oxygen. All potentials were measured against and are referred to a Ag/AgCl/KCl (3 M) electrode (BasInc). A Pt wire was used as a counter electrode. Measurements were performed with a computer-controlled Autolab PGSTAT302N potentiostat. The current densities are given on the basis of the geometric area.</p><!><p>Excess electrons are generated at the surface of anatase TiO2 aggregates upon the dissociation of atomic hydrogen into protons and electrons. The excess electrons can be either localized, forming Ti3+ centers2or delocalized in the conduction band3</p><p>In the following, Vis spectroscopy is used to investigate the generation and reactivity of Ti3+ centers, and conduction band electrons are tracked by IR spectroscopy (Sections 3.1.3−3.1.4).</p><!><p>Sample exposure to atomic hydrogen induces significant changes in the Vis spectrum of an anatase TiO2 aggregate powder (Figure 1a). In particular, a broad absorption in the visible range (380 nm < λ < 800 nm) with a maximum at λ ∼ 780 nm evolves and saturates after ∼150 min (Figure 2a). This signal indicates the formation of Ti3+ centers upon sample reduction by atomic hydrogen (eq 2). Discontinuation of atomic hydrogen exposure and sample storage under high-vacuum conditions (p < 10–5 mbar) for 150 min lead to a decrease of the Vis absorption by ∼60%, while the signal envelope remains virtually unchanged (Figures 1b and 2a).</p><!><p>(a–c) Diffuse reflectance Vis spectra of an activated anatase TiO2 aggregate powder (a) after 150 min of exposure to atomic hydrogen, (b) after 150 min under high-vacuum conditions in the absence of atomic hydrogen and (c) after the addition of 20 mbar O2. (d–f) Diffuse reflectance Vis spectra of an activated anatase TiO2 aggregate powder (d) after 150 min of exposure to atomic hydrogen, (e) after the addition of 0.1 mbar H2O for 3 min and reestablishment of high-vacuum conditions for 150 min and (f) after the addition of 20 mbar O2. Spectra of the activated aggregates (i.e., before chemical reduction) were used as the reference.</p><p>Evolution of the Kubelka–Munk function recorded at 750 nm for an activated anatase TiO2 aggregate powder (a) and of the absorbance recorded at 1300 cm–1 for an activated anatase TiO2 aggregate layer (b) upon exposure to atomic hydrogen, subsequent reestablishment of high-vacuum conditions, and final addition of molecular oxygen.</p><!><p>Molecular oxygen acts as an efficient electron scavenger when added to the gas phase leading to an immediate quenching of the Vis absorption (Figures 1c and 2a).</p><p>The significant decrease of the Ti3+ signal intensity, which is observed, when the surrounding H2/H· atmosphere is replaced by high vacuum (with a rest gas pressure of p < 10–5 mbar) may possibly result from the desorption of molecular hydrogen. Such a process would correspond to the reversal of reactive hydrogen uptake involving the recombination of electron/proton centers at the particle surface according to:4</p><p>The corresponding second order rate law is5</p><p>Integration yields the concentration of electron/proton centers at the semiconductor surface [(e–/H+)s] as a function of time t6</p><p>Here, [(e–/H+)s]0 is the (undetermined) initial concentration of electron/proton centers, and k is the corresponding second order rate constant.</p><p>The (first-order) desorption of hydrogen atoms, on the other hand, would give rise to an exponential decay of electron/proton centers according to7with k' being the corresponding first-order rate constant. Equation 7 would also describe a possible (pseudo-first-order) proton-coupled electron transfer to acceptor molecules (e.g., oxygen) in the gas phase, provided that their concentration stays constant over time. In such a case, k' would correspond to the pseudo-first-order rate constant and its value would depend on the concentration of acceptor molecules.</p><p>Equations 6 and 7 were fitted to the experimental data of the Vis-signal intensity decay as recorded upon the replacement of the H2/H· atmosphere under high-vacuum conditions (Figure 3). While the integrated rate law corresponding to the second order recombination of electron/proton pairs at the semiconductor surface (eq 6) closely resembles the observed intensity decay, experimental data significantly deviate from first-order kinetics (eq 7).</p><!><p>Fitting of the Vis-signal decay (corresponding to the normalized Kubelka–Munk function at 750 nm) observed for a chemically reduced anatase TiO2 aggregate powder upon reestablishment of high-vacuum conditions (compare Figure 2a). Fitting curves corresponding to a first-order decay (blue long-dashed line) and a second order (red short-dashed line) were used to fit the experimental data (gray solid line).</p><!><p>The good fit of the second order decay curve to the experimental data (which cover 60% of the total conversion, Figure 3) points to desorption of molecular hydrogen (eq 4) as the predominant process contributing to the depletion of the Vis-active electron center, which is observed when a H2/H· containing atmosphere is replaced by high-vacuum conditions. The recombination of H+/e– pairs at the semiconductor surface and the subsequent desorption of molecular hydrogen correspond to the reversal of the chemical sample reduction by H atoms. Obviously, there exists a chemical equilibrium between hydrogen atoms in the gas phase and—following their dissociation—Ti3+/proton centers in the oxide. The equilibrium concentration of excess electrons in the semiconductor seems to depend critically on the concentration of atomic hydrogen in the gas phase.</p><!><p>Addition of water vapor to the gas phase results in the formation of a thin water layer at the oxide surface, which persists even upon the subsequent reestablishment of high-vacuum conditions (see Section 3.1.4). Interfacial water, importantly, does not significantly influence the evolution of the Vis spectrum of an aggregate powder previously reduced by atomic hydrogen (Figure 1d). In particular, exposure of the reduced aggregate powder to water vapor (p[H2O] = 0.1 mbar) for 3 min and subsequent sample storage for 150 min under high-vacuum conditions entails a decrease of the signal intensity in the visible by about 40% (Figure 1e). While the decrease of the signal intensity is comparable in the absence (Figure 1b) and in the presence (Figure 1e) of an interfacial water layer, a shift of the absorption maximum from λ ∼ 780 nm to λ ∼ 700 nm is observed in the latter case pointing to a location of Ti3+ species at the oxide surface. While interaction of surface Ti3+ with water dipoles may induce a minor modification of electron transition energies, our results clearly highlight the absence of an interfacial electron transfer from the semiconductor to water or its adsorption products at the oxide surface. However, the addition of oxygen (p[O2] = 20 mbar) leads again to a complete quenching of the Ti3+ signal and the initial reflectance of the powder is restored (Figure 1f). A very similar behavior was observed for anatase TiO2 aggregate films deposited onto FTO-coated glass substrates (Figure S1).</p><!><p>IR spectra of activated anatase TiO2 aggregate layers feature sample-specific bands in two separate spectral regions. At least five overlapping, but clearly distinguishable bands contribute to the spectrum between 3800 and 3500 cm–1 (Figure S2a) and are assigned to the stretching vibration of isolated surface hydroxyl groups (Ti–OH groups).34 The irregular surface of TiO2 nanoparticles gives rise to different local geometries of the Ti–OH groups and, thus, to different vibrational frequencies.</p><p>The spectral range between 1800 and 1300 cm–1 (Figure S2b) features weak bands at 1686, 1621, 1572, 1525, 1462, 1380, and 1360 cm–1. This spectral region contains contributions from antisymmetric νas(COO) and symmetric νs(COO) stretching vibrations as well as δ(CH3) and the δ(HOH) bending vibrations.35−37 The observed bands are thus assigned to carbonate and carboxylate species.</p><p>The IR spectra clearly evidence the presence of surface remnants (chemisorbed water, organic adsorbates) after the activation process. Clearly, higher processing temperatures (i.e., T > 450 °C) under high-vacuum conditions and/or in an oxygen atmosphere would be necessary to desorb and decompose these surface species and generate adsorbate-free (i.e., clean) oxide surfaces. Such elevated processing temperatures, however, may induce significant morphological changes (e.g., particles growth and particle neck formation) and were, therefore, omitted.</p><p>Upon atomic hydrogen exposure, a Drude-type absorption characteristic of free conduction band electrons appears in the difference spectrum of the aggregate layer (Figure 4a). The structureless absorption increases monotonically toward lower wavenumbers between 3600 and 1550 cm–1 with a sharp cutoff near 1250 cm–1.</p><!><p>IR spectra of an activated anatase TiO2 aggregate layer (a) after 60 min of exposure to atomic hydrogen and (b) after storage for 30 min in an oxygen atmosphere (p(O2) = 100 mbar). Spectra are referred to the single channel spectrum of an activated aggregate layer.</p><!><p>The temporal evolution of the signal intensity upon (i) atomic hydrogen exposure, (ii) subsequent reestablishment of high-vacuum conditions, and (iii) final addition of oxygen is tracked at 1300 cm–1 (Figure 2b). The signal intensity saturates after 60 min of sample exposure to atomic hydrogen. Reestablishment of high-vacuum conditions leads to a gradual decrease of the signal intensity. In particular, an intensity decrease of ∼15% is observed after sample storage for 60 min under high-vacuum conditions (Figure 2b). As in the case of Ti3+ centers, we attribute the slow depletion of the IR signal to the recombination of electron/proton centers at the oxide surface and hydrogen desorption.</p><p>Oxygen addition leads to an immediate and significant quenching of the IR signal (Figure 2b). However, ∼10% of the initial intensity remain even after 30 min of sample storage in an oxygen atmosphere (Figure 4b) pointing to an incomplete interfacial electron transfer to oxygen.</p><p>Obviously, a fraction of conduction band electrons remains in the semiconductor even in the presence of the electron acceptor. This unreactive fraction of electrons cannot be detected by Vis spectroscopy where oxygen addition leads to a complete quenching of the Vis absorption corresponding to Ti3+ centers (Figures 1 and 2a). The unreactive fraction of conduction band electrons is assigned to electron/proton pairs in the semiconductor bulk. Because of slow proton diffusion in the oxide bulk, these centers give rise to a transient n-type doping of the semiconductor. Interfacial electron transfer from electron/proton centers located at the surface, in contrast, is not limited by the diffusion of protons or hydrogen atoms in the oxide bulk and is, therefore, fast.</p><p>IR bands assigned to the stretching vibration of OH groups on the surface of activated TiO2 aggregate layers experience a small (∼20%) decrease in intensity upon chemical reduction with atomic hydrogen (Figures 4 and 5a,b). However, the signal envelope in the corresponding wavenumber range remains unchanged. More importantly, upon the addition of oxygen and the associated interfacial electron transfer, the initial intensity of IR bands is restored (Figure 4). This minor (and reversible) change of the intensity of OH bands is attributed to a change of the respective extinction coefficients in the presence of excess electrons. Similar observations were made by Panayotov et al.38 upon the photooxidation of methanol on TiO2 nanoparticles. In that study, the spectral features associated with adsorbed methoxy groups were observed to decrease upon UV exposure, resulting in negative changes in IR absorptivity. At the same time, an increase of the background signal attributed to conduction band electrons was observed. The decrease of the intensity of methoxy-specific IR bands was nearly completely recovered in the absence of UV photons. The effect was assigned to electric field changes within the particles, rather than significant chemical reactions or desorption processes and was attributed to a Stark effect induced by trapped carriers.39</p><!><p>IR spectra of an activated TiO2 aggregate layer (a) before and (b) after chemical reduction with atomic hydrogen (60 min), (c) after subsequent water addition (0.1 mbar H2O, 3 min), and reestablishment of high-vacuum conditions (p = 10–7 mbar, 57 min) and (d) after final storage for 30 min in an oxygen atmosphere (p(O2) = 100 mbar). Spectra are referred to the single channel spectrum, which was recorded under high-vacuum conditions and with the sample/mesh removed from the IR path (empty beam).</p><!><p>Chemical reduction of the aggregate layer by atomic hydrogen does not induce any displacement of IR bands (Figure 5a,b). This is in line with observations made by Yates and Panayotov,34 who did not observe any change in the IR-band positions of isolated OH groups on the surface of TiO2 nanoparticles upon thermal sample reduction (i.e., lattice oxygen removal) and the associated accumulation of excess electrons.</p><p>Finally, it has to be mentioned that the exposure of TiO2 aggregate layers to atomic hydrogen does not lead to an increase of the concentration of OH groups at the oxide surface. Furthermore, there is no indication of the formation of molecular water upon the chemical reduction of the oxide (Figure 5a,b).</p><!><p>The presence of water induces significant changes in the IR spectrum of a chemically reduced anatase TiO2 aggregate layer (Figures 6a,b and 5b,c). In particular, a broad absorption extending from 3700 cm–1 to 2500 cm–1 as well as a narrow band at 1621 cm–1 appear in the spectrum after the addition of water vapor (p [H2O] = 0.1 mbar, 3 min) and the subsequent reestablishment of high-vacuum conditions for 57 min (Figures 6b and 5c).</p><!><p>IR spectra of an activated anatase TiO2 aggregate layer (a) after 60 min of exposure to atomic hydrogen, (b) after subsequent addition of 0.1 mbar H2O for 3 min and reestablishment of high-vacuum conditions for 57 min and (c) after storage for 30 min in an oxygen atmosphere (p(O2) = 100 mbar). Spectra are referred to the single channel spectrum of an activated aggregate layer.</p><!><p>The broad IR signal between 3700 and 2500 cm–1 as well as the narrow band at 1621 cm–1 are attributed to the stretching vibration of hydrogen-bonded OH groups and to the bending mode of molecularly adsorbed water, respectively. A closer inspection of the wavenumber range between 3750 and 3600 cm–1 reveals a broadening of bands corresponding to the stretching vibration of isolated OH groups as well as a change in the spectral envelope upon water adsorption (Figure 5b,c). Hydrogen bonding may involve surface OH groups as well as physisorbed water.</p><p>Apart from the IR band at 1686 cm–1, which experiences a narrowing of the band width, bands detected both for activated as well as for chemically reduced TiO2 aggregate layers between 1800 and 1300 cm–1 (and which are assigned to carbonate and carboxylate species) do not experience significant changes upon water addition (Figure 5b,c). The appearance of the intense band at ∼1621 cm–1 is associated with the physisorption of molecular water.34 While for liquid bulk water, the H–O–H dangling vibration gives rise to a band at 1640 cm–1,40 interfacial water, which is strongly interacting with the TiO2 surface, gives rise to a band at 1621 cm–1.34,41,42 It has to be emphasized that even 60 min after the establishment of high-vacuum conditions (via evacuation of the water vapor at room temperature), molecularly adsorbed water molecules remain adsorbed at the oxide surface, possibly forming a thin water film.</p><p>The monotonic absorption background corresponding to conduction band electrons experiences only a minor decrease in intensity upon water addition (Figure 6b). Obviously, conduction band electrons (as well as Ti3+ centers, compare Figure 1d,e) are unreactive toward water: 80% of the original signal intensity (corresponding to the broad background absorption) persist even 60 min after the addition of water vapor to the chemically reduced sample (Figure 6b). The decrease of the signal intensity by ∼20% is comparable to the intensity loss registered for a chemically reduced sample under high-vacuum conditions (Figure 2b) and is attributed to the desorption of molecular hydrogen.</p><p>The stability of conduction band electrons toward water differs from previous observations by Yates and co-workers.34 Concretely, these authors evidenced the withdrawal of conduction band electrons from thermally reduced TiO2 nanoparticles (Degussa P25, 70% anatase and 30% rutile) upon water addition. Ti–OH surface species resulting from water dissociation at oxygen vacancy sites were assumed to facilitate excess electron depletion. As discussed above, H atom dissociation at the TiO2 surface (as exploited in the present study) allows for electron accumulation upon preservation of the metal-to-oxygen ratio in contrast to thermal reduction of the oxide, which is associated with lattice oxygen removal. Obviously, water adsorption and consecutive charge transfer reactions are critically influenced by the presence of surface defects giving rise to significant differences in the reactivity of excess electrons generated in TiO2 nanoparticle systems by thermal reduction (lattice oxygen removal) or chemical reduction (H atom dissociation), respectively. The elucidation of the impact of different defect types on water chemistry at reduced TiO2 surfaces is highly relevant for applications such as solar water splitting.43 The experimental strategy reported in this paper may contribute to an advancement of analytical methodologies facilitating such an in depth understanding of interfacial reactions on highly dispersed semiconductor oxide systems.44</p><p>A major fraction of conduction band electrons persists in the presence of an interfacial water layer on the surface of chemically reduced TiO2 aggregates (Figure 6b). A similar behavior was found for Ti3+ centers (Figure 1d,e). However, while Ti3+ centers are quantitatively quenched in the presence of oxygen both in the absence and in the presence of interfacial water (Figure 1), a slightly different reactivity is observed for IR-active conduction band electrons. The addition of oxygen (p[O2] = 100 mbar, 30 min) to a chemically reduced aggregate layer in the presence of interfacial water leads to an intensity decrease of the broad background absorption by ∼85%. At the same time, the width of the bands at 1686 and 1621 cm–1 remains unchanged, while their intensity increases (Figures 6c and 5c,d). This resembles the (reversible) decrease of the intensity of OH bands in the presence of conduction band electrons (Figure 4). The broad absorption between 3600 and 2700 cm–1 corresponding to hydrogen-bonded OH groups as well as bands between 3750 and 3650 cm–1 corresponding to isolated surface OH groups remain virtually unchanged upon the addition of oxygen (Figures 6c and 5c,d). The absence of significant changes in the envelope of OH bands differs from findings on reduced TiO2 anatase (101) single crystal surfaces, where the formation of terminal OH groups was observed upon the reaction between water and oxygen.45 This discrepancy may result from the very high water activity in the experiments reported here. Under these conditions, the oxide surface will most probably be fully saturated with OH groups already prior to oxygen addition. Furthermore, very different defect sites are involved in nanoparticle-based systems (chemically reduced by H atoms) and surfaces of (mineral) single crystal surfaces. Notably, it was found that Nb impurities play an important role in the formation of the terminal OH groups.45</p><p>The depletion of the monotonic background absorption (Figure 6c) results from an interfacial electron transfer to oxygen, which is obviously not impeded by the presence of adsorbed water. However, the IR absorption corresponding to conduction band electrons is not completely quenched upon the addition of oxygen. A fraction of ∼15% persists even after sample storage for 30 min in an oxygen atmosphere. This resembles the behavior of chemically reduced TiO2 aggregate layers upon oxygen addition in the absence of physisorbed water (Figure 4) and points to a persistent n-type doping (i.e., hydrogen doping) of the oxide.</p><!><p>Cyclic voltammetry has proven particularly useful for the characterization of electron centers in anatase TiO2 nanoparticle films and for the evaluation of their reactivity at the solid/electrolyte interface.4,27,46 To extend the investigation of the reactivity of electron/proton centers from the solid/gas interface to the solid/electrolyte interface we recorded cyclic voltammograms (CVs) of anatase TiO2 aggregate electrodes in 0.1 M HClO4 aqueous electrolyte (Figure 7).</p><!><p>Cyclic voltammograms for an anatase TiO2 aggregate electrode measured in (a) N2- and (b) O2-purged 0.1 M HClO4 aqueous solution. Scan rate: v = 20 mV·s–1.</p><!><p>If the electrolyte is purged of residual oxygen by bubbling N2 through the solution, reversible currents are observed at potentials EAg/AgCl < −0.2 V (Figure 7a).4,27,46 The symmetrical shape of the CVs is indicative of capacitive processes associated with reversible electron accumulation (in the negative-going scan) and extraction (in the positive-going scan) at the semiconductor/electrolyte interface according to eq 1.</p><p>Despite difficulties associated with the definition and the exact experimental determination of the conduction band edge in nanosized systems,3 it is well-established that anatase TiO2 nanocrystal films feature an exponential surface state distribution just below the conduction band giving rise to capacitive currents in the CVs.3,4,46 The reversibility of these currents in the absence of oxygen (Figure 7a) evidences the absence of any significant interfacial electron transfer (i.e., the absence of Faradaic reactions). Accordingly, electrochemically accumulated electrons are inactive with regard to an interfacial electron transfer to (and thus the reduction of) water or its adsorption products. This inactivity can be rationalized by thermodynamic reasons, namely, the low reducing power of trapped electrons.</p><p>In the presence of appropriate electron acceptors in the electrolyte, however, electron transfer across the semiconductor/electrolyte interface may take place and the resulting Faradaic currents can be tracked by cyclic voltammetry. In this context, the electrochemical reduction of oxygen at the semiconductor/electrolyte interface has been used previously as a model reaction to investigate the reactivity of electrons trapped at band gap states in anatase TiO2 nanoparticle electrodes.47</p><p>Indeed, Faradaic currents are measured for anatase TiO2 aggregate electrodes at potentials EAg/AgCl < −0.2 V in the presence of dissolved oxygen (Figure 7b). Under these experimental conditions, charge transfer across the solid/electrolyte interface occurs in addition to capacitive processes. The observed asymmetrical shape of the CV therefore results from the superposition of a negative current density (both in the negative and positive scan direction) resulting from electron transfer to dissolved oxygen (Faradaic reaction) and of a negative (in the negative scan direction) or positive current density (in the positive scanning direction) resulting from capacitive charging/discharging of the semiconductor.</p><p>The appearance of Faradaic currents clearly demonstrates that the electron transfer from band gap traps in the semiconductor to dissolved oxygen is feasible not only at the semiconductor/gas interface (Section 3.1) but also at the semiconductor/electrolyte interface pointing to the favorable energetic location of band gap traps with respect to the O2/HO2·redox couple.48</p><p>Upon voltammetric cycling (Figure 7), electron accumulation and compensation take place mainly at the oxide surface. However, previous spectroelectrochemical studies have reported the appearance of a Drude-type IR-absorption (characteristic for conduction band electrons) and a broad Vis absorption (characteristic of Ti3+ centers) upon prolonged cathodic polarization of anatase TiO2 electrodes in acidic aqueous electrolytes.17,23,24 While the Vis signal showed a complete reversibility with respect to electrochemical charge extraction at positive potentials, the IR signal partially persisted even after prolonged polarization times.15,16 Based on this observations, Vis-active centers were assigned to localized Ti3+ species at the TiO2 particle surface and IR-active centers to shallow H+/e– traps located at least partially in subsurface regions, giving rise to a persistent electrochemical doping of the electrode.15,16 These conclusions are perfectly in line with the conclusions drawn from our observations at the solid/vacuum interface (Section 3.1). Remarkably, chemical sample reduction by atomic hydrogen at the solid/vacuum interface resembles, at least to some extent, electrochemical sample reduction. A combination of model studies performed at different levels of complexity as proposed in the present study may therefore constitute a valuable tool for the identification of some physical and chemical properties influencing the materials' functional properties under application-relevant conditions.</p><!><p>The chemistry of oxygen and water at the surface of highly dispersed metal oxide semiconductors is at the heart of technologically important processes such as the oxygen reduction reaction49 or water splitting.43,50 Therefore, great efforts are being made to gain molecular insight into underlying reaction steps on the one hand, and to identify the relationship between macroscopic measurables and interfacial properties on the other hand. This task is complicated by the fact that interfaces of working electro- and photocatalytic materials constitute highly complex systems.</p><p>Thermodynamic and kinetic details of the electron transfer from a particular semiconductor oxide particle to a particular acceptor species at the solid/gas or solid/electrolyte interface critically depend, for instance, on the type of exposed crystallographic faces,51,52 the presence of intrinsic53 or extrinsic defects,45,54 the presence of adsorbates55,56 or metal clusters,56 and the protonation state of functional surface groups.57 Furthermore, the experimental decoupling of the cascade beginning with electron generation and ending with the interfacial electron transfer is often challenging or even impossible. Importantly, most reaction steps critically depend on interfacial properties. In this regard, we believe that the methodology reported in this study will allow for investigating the reactivity of hydrogen-related electron centers at very different interfaces featuring partially tunable levels of complexity.</p><!><p>Aggregates consisting of anatase TiO2 nanoparticles were produced by a sequence of (i) MOCVS of isolated nanocrystals (d = 10–20 nm), (ii) purification of the resulting powder by thermal annealing under high-vacuum conditions and in an oxygen atmosphere, (iii) preparation of aqueous colloids, (iv) drying, and (v) final sintering of the samples in air. The resulting aggregates were investigated in the form of loose powders (by Vis spectroscopy), as layers immobilized on a tungsten mesh (by IR spectroscopy) and as electrodes, that is, films deposited onto a transparent conducting substrate (by cyclic voltammetry).</p><p>Powders and immobilized layers consisting of anatase TiO2 nanoparticle aggregates are chemically reduced by atomic hydrogen upon the formation of Vis-active Ti3+ centers and IR-active conduction band electrons. Excess electrons are slowly depleted under high-vacuum conditions. This process is assigned to the recombination of atomic hydrogen at the oxide surface and hydrogen desorption. The presence of an interfacial water layer does not significantly change the rate of excess electron depletion. Obviously, electron transfer from the semiconductor to water and its adsorption products is not feasible. While an interfacial transfer of conduction band electrons to surface water may be hindered kinetically, we attribute the inactivity of trapped electrons to thermodynamic reasons, that is, the low reducing power of trapped electrons. In the presence of molecular oxygen, Vis-active Ti3+ centers are immediately and completely quenched because of interfacial electron transfer reactions both in the presence and in the absence of an interfacial water layer. However, the addition of oxygen leads only to a partial consumption of IR-active conduction band electrons, and 10–15% of the initial IR signal intensity persists after 30 min in oxygen atmosphere pointing to a n-type doping of the sample upon exposure to atomic hydrogen.</p><p>The investigation of anatase TiO2 aggregate films as the electroactive electrode material reveals that electron transfer from the semiconductor to molecular oxygen is feasible not only at the solid/gas interface (as tracked by Vis- and IR-spectroscopy), but also at the solid/electrolyte interface (as tracked by cyclic voltammetry) because of the favorable energetic location of band gap traps with respect to the O2/HO2· redox couple. In the absence of dissolved oxygen, the semiconductor can be charged and discharged reversibly and no Faradaic currents are observed. This highlights the inactivity of electrochemically accumulated electron centers with regard to water reduction.</p><!><p>Additional Vis spectra of anatase TiO2 aggregate electrodes/films; IR spectra of anatase TiO2 aggregate layers; schematic drawings of the high-vacuum reactors; further experimental details (PDF)</p><p>jp1c01580_si_001.pdf</p><p>The authors declare no competing financial interest.</p>

PubMed Open Access

Interaction Profiling Methods to Map Protein and Pathway Targets of\nBioactive Ligands

Recent advances in \xe2\x80\x93omic profiling technologies have ushered in an era where we no longer want to merely measure the presence or absence of a biomolecule of interest, but instead hope to understand its function and interactions within larger signaling networks. Here we review several emerging proteomic technologies capable of detecting protein interaction networks in live cells, as well as their integration to draft holistic maps of proteins that respond to diverse stimuli, including bioactive small molecules. Moreover, we provide a conceptual framework to combine so-called \xe2\x80\x9ctop-down\xe2\x80\x9d and \xe2\x80\x9cbottom-up\xe2\x80\x9d interaction profiling methods and ensuing proteomic profiles to directly identify binding targets of small molecule ligands, as well as unbiased discovery of proteins and pathways that may be directly bound or influenced by those first-responders. The integrated, interaction-based profiling methods discussed here have the potential to provide a unique and dynamic view into cellular signaling networks for both basic and translational biological studies.

interaction_profiling_methods_to_map_protein_and_pathway_targets_of\nbioactive_ligands

Introduction<!>Thermal Protein Profiling<!>Differential Protease Sensitivity and Chemical Accessibility<!>Bottom-Up Interaction Profiling: Intracellular Proximity Labeling<!>Multi-Omic Interaction Profiling: Integrating Top-down and Bottom-Up\nStrategies<!>Conclusion

<p>Bioactive small molecule ligands lie at the center of biological processes, serving as substrates and co-factors in metabolism, diffusible signals, structural components and pharmacologic agents. These roles all originate from direct interactions with other biomolecules within and between cells, tissues and organisms. Therefore, a central challenge in determining the mechanisms of action for a small molecule of interest is to identify the direct biomolecular target(s) with which it physically interacts at pharmacologically relevant concentrations. Beyond providing insights into mechanism of action, for example in the follow-up from high throughput screens, there are myriad examples where target identification has led to discoveries in basic biology and expanded opportunities for the development of new therapeutics. These include classic affinity purification methods, such as those used to identify the protein targets of natural products like FK506[1], Trapoxin[2], and Staurosporine[3]. A similar, contemporary example is the discovery of ubiquitin-ligase recruiting "molecular glues" [4] [5, 6], which has led to a potentially new class of small molecules that promote catalytic degradation of their protein targets [7–10]. Likewise, efforts to identify the targets of endogenous ligands like the "oncometabolite" (R)-2-hydroxyglutarate[11], anti-inflammatory metabolites like itaconate[12, 13], and reactive glycolytic metabolites like methylglyoxal[14, 15] have resulted in the identification of new signaling pathways that operate under normal and pathophysiologic contexts. Beyond those produced in mammalian cells, a recent surge in the discovery of novel microbial metabolites, including short chain fatty acids[16], modified bile acids[17], and diverse natural product classes [18, 19], suggests a rich and broad landscape of small molecule-mediated signaling between microbes and host tissues. While the relevant targets of bioactive small molecules span proteins, RNA, DNA and other biomolecules, the vast majority of annotated small molecules target proteins, and therefore the primary focus of this review will be on the emerging role of proteome-wide interaction profiling methods in network-level protein target identification for bioactive small molecules. We will discuss several new methods that yield "top-down," global views of interactions in the proteome, as well as higher resolution "bottom-up" interaction networks surrounding proteins of interest. Specifically, we will discuss the potential to combine these interaction profiling methods to identify both the direct targets and surrounding pathway networks that are functionally relevant with regard to small molecule-mediated signaling in diverse biological contexts.</p><!><p>The phenomenon of ligand-induced protein stabilization has been used to detect protein-ligand interactions for decades[20, 21]. An important aspect of these assays is the fact that no chemical modification of either the ligand or protein of interest is required, suggesting that it may be a general way to screen for protein binding. In 2013, Nordlund and colleagues introduced a variant of this classic assay, entitled Cellular Thermal Shift Assay (CETSA), which elegantly demonstrated that changes in the stability of a protein of interest could be monitored directly, for example by Western blot, in whole proteome samples after exposure to a ligand of interest[22]. A subsequent expansion on CETSA, Thermal Proteome Profiling (TPP; also referred to as MS-CETSA), integrated this concept with LC-MS/MS detection and quantification of protein unfolding and precipitation after exposure of either whole cell lysate or live cells to parallel, stepped pulses in temperature, typically between room temperature and 65–70° C (Figure 1A). Subsequent isotopic barcoding of whole proteome at each temperature provides a method to quantify the relative amount of each detected protein (measured at the peptide level) present in each temperature aliquot, ultimately yielding protein-specific melting temperatures (Tm's) via curve fitting of isotope abundance across the temperature gradient [23, 24]. TPP datasets provided the first global view of protein stability across the proteome. A main limitation with the original TPP workflow is that it was largely limited to soluble proteins that display prototypical melting behavior. Modification of the sample processing workflow with specific surfactants has proven useful to query membrane-associated proteins[25, 26], ligand dose-responsive changes in protein stability (isothermal dose response-CETSA, ITDR-CETSA), and simultaneous changes in protein abundance and stability (2D-TPP). Also, to relax the requirement for a prototypical sigmoidal melting response, an method named proteome integral solubility alteration (PISA) combines proteome from across the temperature range in mock- and ligand-treated samples and directly compares the integrated ion intensity in one isotope channel, essentially capturing the area under the unfolding curves, regardless of their shape[27]. This modification of TPP should be very useful because it not only allows for comparison of proteins with atypical temperature response, but also increases the number of conditions that can be compared in a single mass spectrometry run by an order of magnitude.</p><p>The primary application of TPP assays has been the interrogation of ligand-induced, protein-specific stabilization or destabilization for thousands of proteins in the same experiment. This was first demonstrated with the pan-kinase inhibitor stauorsporine, which caused significant Tm shifts in dozens of known protein kinases when both drug treatment and temperature pulse exposure was performed in live Jurkat cells[28–30]. An important observation in this study was that both significant destabilization of target proteins (i.e. negative shift in Tm), as well as no shift of known staurosporine targets were present in the dataset. This trend has borne out in subsequent studies, and underscores the fact that each protein is stabilized by a unique set of biophysical interactions and will display equally unique overall changes in stability in response to altered intra- or intermolecular interactions. For this reason, there is significant potential for false negatives in TPP assays. Nevertheless, TPP has subsequently been deployed to detect the direct binding targets of pan-kinase inhibitors [29], PIP4K inhibitors a131 and a166[31], the CDK4/6 inhibitor Palbociclib[32], and the MTH1 inhibitor TH1579[33]. Beyond the mammalian proteome, TPP has been used to profile E. coli, S. cerevisiae, T. thermophilus[34], and P. falciparum, the latter of which led to the identification of nucleoside phosphorylase (PfPNP) as the elusive target of the antimalaria drugs quinine and mefloquine[35]. Because of the unbiased nature of the technique, TPP can capture off-target interactions in families not predicted to be targets of a molecule of interest, with notable examples including ferrochelatase (FECH) for the several kinase inhibitors in clinical use, as well as phenylalanine hydroxylase for the well-known HDAC inhibitor panobinostat[36].</p><p>Beyond pharmacologic small molecules, TPP has been successfully deployed to detect endogenous protein-metabolite and protein-protein interactions in cells and lysates [37, 38]. Several studies have used TPP and ITDR-CETSA to detect the protein targets of nucleotides [26, 39] and nicotinamide adenine dinucleotide coenzymes in cells and in lysates. Focused studies in lysates have drafted proteome-wide maps of ATP binding proteins, rediscovering many high affinity interactors alongside the discovery that the ATP solubilizes a large number of positively charged and intrinsically disordered proteins [38]. Each of these datasets also identified altered stability of proteins that were presumably not direct targets of the metabolites being studied, raising the possibility that secondary effects on protein interaction networks were being detected. Indeed, a recent approach named thermal proximity co-aggregation (TPCA) uses the similarity in protein aggregation behavior to infer protein-protein complex membership[40]. While the potential to detect the downstream effects of biological stimuli could be viewed as a confounding factor in interpreting TPP datasets (i.e. whether Tm shifts are due to direct or indirect interactions with a small molecule), this information has the potential to more completely map complexes and pathways that are affected by small molecule binding. For example, the impact of posttranslational modifications (PTMs) on proteins downstream of small molecule action or other stimuli results in the differentiation of a bulk population of protein into discrete proteoforms that carry out distinct functions in the cell. A recent method, termed Hotspot Thermal Profiling, combines PTM-based enrichment and peptide-level thermal profiling to measure the stability of specific modified proteoforms (modiforms) in live cells (Figure 1B)[30]. The presumption in this approach is that PTMs that significantly perturb a modiform's stability relative to its parent "bulk" protein population may be "hotspot" modification sites that are uniquely functional under specific biological conditions. Several annotated and unannotated phosphorylation sites were explored in this first report, confirming that "hotspot" phosphorylation sites could be identified on the basis of modiform stability changes, and that these sites functionally affect intra- and intermolecular protein-protein interactions, as well as protein-small molecule interactions. In principle, this peptide-level mapping approach could be adapted to interrogate diverse modification types, protein mutants, and protein isoforms in native samples.</p><!><p>Thermal profiling represents just one class of methods that can provide "top-down," global proteome information, however there are significant caveats in translating protein-specific Tm shifts into direct or secondary effects on protein interaction networks. A subset of the proteome does not exhibit prototypical or significant melting behavior, and those proteins that do may not be highly stabilized or destabilized in response to bona fide binding events. In principle, any method that interrogates a generic molecular attribute of protein structure or chemical state could be used to detect altered protein interactions in response to small molecules or other perturbations. Several methods have been developed that take advantage of the increased protease resistance of protein surfaces that are engaged in intermolecular interactions with small molecules or other biomolecules. Drug affinity response sensitivity (DARTS) and limited proteolysis-mass spectrometry (LiP-MS) both compare the kinetics of proteolytic cleavage in whole proteome between conditions to identify proteins that are likely engaged in differential interactions (Fig. 1C)[41, 42]. In a recent study, LiP-MS using the promiscuous protease Proteinase K in E. coli lysate identified more than 1,000 metabolite-protein interactions. Moreover, because the assay results in altered sensitivity at specific sites within proteins (detected by altered daughter peptides from those sites; Figure 1C), this study was able to identify more than 7,000 putative binding sites involved in those interactions[43]. In addition to protease sensitivity, methods that directly probe the reactivity of protein functionalities have been developed to query protein-biomolecule interactions in native conditions. These include susceptibility to protein oxidation (SPROX)[44], protein "painting" [45], and more recent bioconjugation reactions such as labeling of accessible methionine residues in native proteome[46]. Activity-based probes focused on specific amino acid reactivity [47, 48] or photoaffinity chemical probes[49, 50], can also report on the occupancy of small molecules or other biomolecules within subsets of the proteome.</p><!><p>While thermal profiling, protease resistance and related "top-down" methods can provide a global view of the proteins that are impacted by a specific stimulus, there are significant caveats in categorizing direct or indirect targets within protein networks. Therefore, there is also great value in other methods that enable protein-specific interaction profiling in cells in response to small molecule perturbation. This has heralded the emergence of a new class of intracellular proximity profiling platforms. Published proximity profiling approaches invariably involve the expression of genetic constructs encoding an engineered enzyme or receptor fused to a target protein of interest (POI). The resulting fusion can accept endogenous or exogenous substrates and co-factors to generate a reactive molecule in the immediate proximity of the fusion protein (Figure 2A). Direct transfer or diffusion of this reactive tracer results in covalent labeling of proximal proteins with a specific chemical tag for enrichment prior to quantitative profiling by LC-MS/MS. An early example, BioID, employs an engineered biotin ligase-POI fusion, which generates a biotinoyl-5'-AMP product that can covalently label surface amines on proximal proteins[51–53]. An evolved variant termed TurboID, which exhibits significantly faster labeling rates, has been used to map proximal protein interactors in mammalian cells and whole organisms[53], and more recently in plants[54]. Other methods use similarly reactive thioesters within small peptides or proteins as the tagging element, and rely on direct enzymatic transfer to proximal proteins, perhaps providing more specific labeling profiles. These direct proximal transfer methods include the NEDDylator[55], PUP-IT[56], and EXCEL[57] systems (Figure 2A). These ligases efficiently label short consensus motifs or specific amino acids present on proximal proteins both on and inside of cells, and in the case of PUP-IT, for example, can capture even weak, transient protein-protein interactions[58].</p><p>To generate more reactive, diffused labeling elements, genetic fusions of horseradish peroxidase[59] and engineered ascorbic acid peroxidase enzymes (i.e., the APEX system), convert an exogenous biotin phenol probe into a reactive phenoxyl radical[60, 61] in the presence of heme and H2O2 to label proximal proteins (Figure 2B). The phenoxyl radical will react preferentially with surface tyrosines (and less so Trp, His, and Cys), and should provide a spatially-restricted proximity profile due to the short half-life of the radical in a biological environment. Due to these attributes, APEX has proven widely useful in labeling the proteome of sub-cellular compartments like mitochondria[61], membrane-associated proteins[62], and the endoplasmic reticulum[63]. More recently, a light-activated photoproximity protein interaction (PhotoPPI) profiling method has been developed to localize a masked and highly reactive carbene to a POI through a SNAP-Tag fusion and complementary O6-benzylguanine targeting element[64]. Subsequent activation by 365 nm light results in proximal protein labeling and interactome mapping with high spatiotemporal control (Figure 2C). This study focused on identifying steady state and dynamic binding partners of the redox sensor protein KEAP1, which would be challenging to interrogate with affinity purification and likely other proximity labeling approaches. The minimal requirements and modularity of the photoproximity chemical probe in this system, coupled with facile activation by light, may enable proximity profiling in cellular contexts that may not be suitable for other methods. Taken as a group, all of these methods require distinct co-factors to produce uniquely reactive products, and thus provide opportunities to query protein networks in different cellular compartments, under distinct kinetic regimes, and proximity radii.</p><!><p>Given the unique information provided by top-down and bottom-up profiling methods, we posit that integration of these methods in matched experiments could be used to generate high-resolution maps of the proteins and pathways that are impacted by a small molecule or biological signal under investigation. Top-down approaches like TPP can be applied to identify putative target proteins and complexes that are significantly perturbed in response to a specific small molecule or other stimulus. For example, kinetic monitoring of protein Tm shifts could enable proteome-wide tracking of altered biophysical interactions that originate at sites of direct small molecule-target interactions (i.e. early responders), and subsequently flow outward within protein-protein and other protein-biomolecule interaction networks (i.e. neighboring protein complexes and interaction networks; Figure 3). This protein neighborhood-level information should afford greater likelihood of identifying relevant target sites, protein complexes and networks that are involved in small molecule mechanism of action directly in the cells of interest. Taking the next step of determining direct versus indirect target proteins and upstream versus downstream signal propagation events is currently a bottleneck in the interpretation of these global profiles. Generation of matched proximity profiles may represent a general and relatively rapid way to confirm whether the proximal binding partners of a protein of interest are being altered in response to a small molecule perturbation, as well as define the specific protein partners and pathways involved in that network (Figure 3). One avenue toward this would be to perform a kinetic series of top-down interaction profiles that mirror the phenotypic changes caused by a small molecule or other event, identify specific proteins and/or complexes that are differentially affected, and prioritize them for proximity profiling using one of the methods highlighted in Figure 2. Generating a matched kinetic profile around several prioritized "nodes" would generate a high-resolution profile of the protein interactions surrounding each protein, as well as how they change along the kinetic series (Figure 3). Bioinformatic integration of these multi-omic interaction profiles with interaction datasets generated by orthogonal methods, and other -omic datasets that measure the molecules that could be involved in signal propagation (e.g., metabolites or protein PTMs) should enable a systems-level view of the proteins and pathways that are impacted by bioactive small molecules or other biological signals.</p><!><p>The current challenges in defining the molecular makeup of biological systems have shifted away from defining the parts, and more toward understanding the dynamic interactions of those parts that mediate cellular signaling and function. There is no place where this paradigm shift is more relevant than in the study of the proteome, where the cellular locale, proximal protein complexes, posttranslational modification state and small molecule ligand interactions all converge to control function. Here we presented a review of global and local interaction profiling methods that can be combined to uncover the relevant mechanism of action for, among other perturbations, bioactive small molecules. As we have discussed above, we posit that label-free interaction profiling methods operating directly in live cells offer an alternative to traditional approaches because they do not require alteration of the chemical structure of target small molecules or performing assays in lysates. More importantly, they offer a global picture of both the direct and indirect interaction networks affected by a specific perturbation, which provides ample opportunity for basic biological discovery. Integration of protein interaction-based profiling datasets with existing –omic profiling methods can aid in drafting more complete signaling maps within and between pathways. While not explicitly explored previously in the literature, we posit that these approaches could also be applied to interrogate cell-cell and cross-species (i.e. microbe-host) signaling interactions. The integrated, interaction-based profiling methods discussed here should provide a unique and dynamic view into cellular signaling networks for both basic and translational biological studies.</p>

PubMed Author Manuscript

tsRNAs: The Swiss Army Knife for Translational Regulation

tRNA-derived small RNAs (tsRNAs, or tRFs) are a new category of regulatory noncoding RNAs with versatile functions. Recent emerging studies have begun to unveil distinct features of tsRNAs based on their sequence, RNA modifications, and structures that differentially impact their functions towards regulating multiple aspects of translational control and ribosome biogenesis.

tsrnas:_the_swiss_army_knife_for_translational_regulation

The Expanding Functions of tsRNAs<!>Interfering Translational Initiation and the Role of tsRNA Structure and/or Modification<!>AGO-Dependent Translational Inhibition by Targeting Specific mRNAs<!>AGO-Independent Translational Regulation by tsRNAs: Structural Effects<!>Concluding Remarks

<p>tRNAs are a type of highly modified and structured RNA that have a well-defined role in mRNA translation. The fragmentation of tRNAs at different loci gives birth to a new species of small RNAs: tsRNAs (also known as tRNA-derived fragments, tRFs) with unexpected complexity, which is due, in part, to the numerous types of RNA modifications inherited from tRNAs as well as to the RNA interaction potential (e.g., RNAs and proteins) endowed by RNA modifications and novel structures [1]. tsRNAs show diverse functions, ranging from stress response, tumorigenesis, stem cell biology, and epigenetic inheritance [1]. At the molecular level, recent converging studies have begun to provide evidence that different tsRNAs interplay with multifaceted aspects of translational regulation and ribosome biogenesis, which involve their sequence specificity, RNA modifications, and structural effects. Since tsRNAs are at relatively low abundance compared with their corresponding full-length tRNAs, these emerging studies reinforce the idea that tRNA fragmentation in translation interference merely due to tRNA destruction is an oversimplified model, instead indicating a novel layer of regulation repurposed by the generation of various functional tsRNAs.</p><!><p>The function of tsRNA in translational inhibition was documented in early studies by Paul Anderson's group (reviewed in [1]). They found that, under stress, the cleavage of tRNAs at the anticodon by the nuclease angiogenin generates 5′ and 3′ tsRNAs and that the 5′tsRNAs, but not 3′tsRNAs, could inhibit global protein synthesis [1]. Recently, it was further found that two tsRNAs [5′tsRNA-Ala and -Cys, ~30 nucleotides (nt)] with a terminal oligo-G motif (TOG), can form intermolecular RNA G-quadruplexes (RG4), displacing the translation-initiation factor eIF4E/G/A from m7G-capped mRNAs [2]. In addition, TOG-5′ tsRNAs bind to the cold shock domain of Y-Box Binding Protein 1 (YBX1) to facilitate the assembly of stress granules (ribonucleoprotein complexes), resulting in the sequestration of initiation factors and adding to the effect of global translation repression, although YBX1 does not directly displace translation-initiating factors from the m7G-capped mRNAs (Figure 1A,B). However, knocking down YBX1 only partially reverses the translation repression [1], suggesting that other mechanisms are also involved in 5′tsRNA-induced translational inhibition.</p><p>Recently, work by Cristian Bellodi's group further expanded our understanding of 5′tsRNA-mediated translational control by emphasizing the role of RNA modification [4]. They found that the pseudouridine (ψ) synthase PUS7 is enriched in embryonic and/or hematopoietic stem cells, and that it binds to distinct tRNAs and modifies U into ψ at the U8 position (ψ8). PUS7 deletion leads to significantly decreased levels of TOG-5′tsRNAs around 18 nt, which is associated with increased global protein synthesis [4]. Transfecting TOG-5′tsRNAs with ψ8, but not those with U8, can restore the protein synthesis of PUS7-KO hESCs and, thus, impact the process of stem cell commitment.</p><p>Mechanistically, ψ8-containing TOG-5′tsRNA preferentially bind to polyadenylate-binding protein 1 (PABPC1), another initiation factor integral to the formation of the translational initiation complex, resulting in displacement of PABPC1 and eIF4A/G from m7G-capped mRNAs. Moreover, depletion of PABPC1 by small interfering (si)RNA can decrease global protein synthesis in PUS7-KO hESCs, phenocopying the effect of ψ8-containing TOG-5′sRNA. These data demonstrate the novel role of ψ8 in fine-tuning the function of 5′tsRNAs in translational regulation (Figure 1C).</p><p>Notably, the 18-nt U8-TOG-5′tsRNAs show strong binding affinity to YBX1, but cannot displace eIF4A/G, which is distinct from their longer version (30-nt-TOG-5′tsRNAs, which form RG4), as reported previously [3,4]. These results suggest a different secondary RNA structure mediated by tsRNAs length and, thus, a context-dependent binding preference. Given that human embryonic stem cells (hESCs) contain both modified and unmodified TOG-5′tsRNAs (while most bear ψ8), they may function synergistically to exert optimized effects in translational regulation.</p><p>In another example, hypoxic stress induced a specific population of tsRNAs with a distinct motif that can bind to YBX1, displacing YBX1 from the pro-oncogenic, cancer-promoting mRNAs it protects, resulting in suppression of cancer metastasis [5]. This suggests context-dependent translational regulation mechanisms mediated by versatile tsRNA species in a tissue-and cell-dependent manner.</p><!><p>Previous studies on individual 21–22-nt 3′tsRNAs with 3′CCA-end (suggestive of cleavage from mature tRNAs) revealed their miRNA-like behavior in downregulating gene expression. These 3′CCA-tsRNAs show DICER-dependent biogenesis, binding with AGO, and repressing mRNA translation in a sequence-specific manner [6]. The effects of 3′CCA-tsRNAs are primarily at the translational level, because the mRNA of the target genes is not affected [6]. Despite these individual reports, a general understanding of AGO-dependent tsRNA-targeting is still lacking.</p><p>Recently, by analyzing 495 public small RNA-sequencing libraries, combined with mRNA sequencing and ribosome profiling after tsRNA transfection, Jian Lu's group provided further insight into the principles of mRNA targeting under this type of tsRNA-mediated, AGO-dependent translational inhibition via antisense pairing [7]. Their bioinformatic analyses showed that shorter tsRNAs (20–22 nt) preferentially associate with AGO1/2, whereas longer tsRNAs (23–29 nt) associate with AGO3/AUB/PIWI (the absence of longer tsRNAs of 30–34 nt found in mammals may be due to the size limitation (18–30 nt) of library construction in the analyzed data sets, or to species differences). They also found that tsRNAs tend to be derived from 5′ halves of tRNAs in most fly tissues.</p><p>By transfecting S2 cells with 12 tsRNA mimics that have been shown to be abundant in Drosophila (5′tsRNAs and middle-derived tsRNAs), the authors found that the polysome:monosome ratio decreased by 20–50% after transfection, indicating that tsRNAs repressed translational activity. By comparing the mRNA-seq data with ribo-seq data generated by three tsRNA transfections (which show the most prominent effect), they found that the genes with more tsRNA target sites were more likely to be translationally arrested; meanwhile, tsRNAs did not affect the mRNA level of target genes. Notably, 7-mer motifs in tsRNAs can perfectly antisense match conserved target sites in mRNAs; these targeted mRNAs are associated with a reduced translational activity. Interestingly, tsRNA-targeting sites are located not only in the 3′ untranslated regions (UTRs), but also in 5′UTRs or CDSs, consistent with a previous analysis of CLASH data showing the small RNA-mRNA interactome [8].</p><p>Importantly, the mRNA of ribosomal proteins (RPs) and translational initiation or elongation factors (IEFs) have the highest target density of AGO2-bound tsRNAs. Under serum starvation, the levels of some 5′tsRNAs are significantly increased, which are correlated with decreased translational activities of RPs and IEFs. Further AGO2 knockdown experiments indicated that AGO2 is indispensable for tsRNA-mediated translational repression under conditions of serum starvation. Interestingly, tsRNA- and miRNA-mediated gene targets are largely independent [7].</p><p>These data suggest that tsRNAs preferentially repress genes that are essential for ribosome biogenesis (i.e., RPs) and translation regulation (i.e., IEFs), thus enabling the repression of global protein synthesis (Figure 1D). The mechanism for the targeting preference of tsRNAs to RPs and IEFs are not well understood but may result from the long-term coevolution of tsRNAs and target sites.</p><!><p>In addition to AGO-mediated translational repression, recent studies have begun to provide novel modes of action for tsRNA- mediated translational regulation by exerting structural effects on mRNAs or rRNAs, independent of the AGO protein.</p><p>Mark Kay's group recently reported that a 22-nt-3′tsRNA-LeuCAG did not bind to any of the known AGO proteins and could not repress luciferase expression with perfectly complementary target sites. Instead, 3′tsRNA-LeuCAG increased cell viability and its inhibition induced apoptosis in rapidly dividing cells in vitro as well as in a hepatocellular carcinoma mouse model [9]. The effect of 3′tsRNA-LeuCAG was specific, because inhibition of other 3′tsRNAs (3′tsRNA-Asp, 3′tsRNA-Ser, or 3′tsRNA-Met) did not reduce cell viability. In addition, transfection of longer 27-nt-3′tsRNA-LeuCAG rather than 22-nt-3′tsRNA-LeuCAG had no effect on cell viability, suggesting that the effect is related to both sequence specificity and RNA structure.</p><p>Ribosome gradient analysis indicated that the inhibition of 3′tsRNA-LeuCAG decreased the abundance of 40S and 80S ribosomal complexes and increased that of the 60S ribosomal complex, suggesting a reduction in the number of 40S ribosomes and impaired assembly of the 80S ribosome. Additionally, inhibition of 3′tsRNA-LeuCAG resulted in 30S pre-rRNA accumulation and the subsequently decreased level of mature 18S rRNAs. This led to the discovery that 3′tsRNA-LeuCAG inhibition decreased the level of specific ribosomal proteins of the small subunit (RPSs), RPS28 and RPS15, at the translational level, while their mRNA abundance was not affected.</p><p>By using RNA secondary structure prediction, target-site mutation, and an in vivo click selective 2′-hydroxyl acylation and profiling experiment (icSHAPE), the authors convincingly showed that 3′tsRNA-LeuCAG unfolds the duplexed mRNA (RPS15 and RPS28) structure at the target site (Figure 1E). Importantly, 3′tsRNA-LeuCAG cannot increase translation of other RPS mRNAs (RPS9 and RPS14) that have similar target sequence but do not have the secondary RNA structure. These data further support the idea that 3′tsRNA-LeuCAG is involved in unfolding the duplexed secondary structures of RPS mRNAs at the targeting-site, thus facilitating ribosome protein biogenesis.</p><p>Another study by Norbert Polacek's group showed a tsRNA-induced structure effect in regulating translation. They found that stress-dependent 5′tsRNA-Val can bind to the 16S rRNA of the small ribosomal subunit (30S) near the site of mRNA entrance, thus inhibiting global protein synthesis by displacing mRNA from the translational initiation complex [10] (Figure 1F). Similar translational control based on tsRNA-rRNA interactions may also account for other reports showing tsRNA-mediated global translational inhibition without complementary targeting on mRNA [11].</p><!><p>The recent emerging studies discussed above have promoted the idea that various tsRNAs produced by tRNA fragmentation can engender acrobatic ways to regulate multiple aspects of translation machinery. In particular, the function of tsRNAs is augmented by unexpected roles of RNA modifications and RNA secondary structure (see Outstanding Questions). Similarly, recent studies found that mammalian Nsun2-and Dnmt2-mediated m5C in tRNAs can profoundly affect tsRNA biogenesis [12–14] and the structure of the resulting tsRNAs [14], suggesting tsRNA-mediated translational control in stem cell function, embryo development, and intergenerational epigenetic inheritance of specific acquired phenotypes. Recent systematic analyses of tRNA-modifying enzymes in budding yeast also revealed the widespread impact of noncoding RNA modifications in translational regulation and gene expression [15]. Detailed molecular mechanisms involved in these circumstances may go beyond our current knowledge and deserve case-by-case studies in the future.</p>

PubMed Author Manuscript

Integrator: surprisingly diverse functions in gene expression

Summary The discovery of the metazoan-specific Integrator Complex represented a breakthrough in our understanding of noncoding U-rich small nuclear RNA (UsnRNA) maturation and has triggered a reevaluation of their biosynthesis mechanism. In the decade since, significant progress has been made to understand the details of its recruitment, specificity, and assembly. While some discrepancies remain on how it interacts with the carboxy-terminal domain of the RNA polymerase II (RNAPII) and the details of its recruitment to UsnRNA genes, preliminary models have emerged. Recent provocative studies now implicate Integrator in the regulation of protein-coding gene transcription initiation and RNA Polymerase II pause-release thereby broadening the scope of Integrator functions in gene expression regulation. Here, we discuss the implications of these findings while putting them into the context of what is understood about Integrator function at UsnRNA genes.

integrator:_surprisingly_diverse_functions_in_gene_expression

Initial discovery of integrator<!>A long sought UsnRNA 3\xe2\x80\xb2end processing factor<!>Relation between Integrator and the RNAPII CTD<!>Integrator interacts with SPT5 and NELF<!>Role of Integrator in RNAPII transcriptional pause-release<!>Is Integrator a modular complex?<!>Concluding remarks

<p>The Integrator complex (INT) was discovered serendipitously while searching for protein partners of Deleted in Split hand/Split foot protein 1 (DSS1). The initial affinity purification of the complex [1] identified twelve Integrator subunits (INTS1 to INTS12, see Figure 1) and demonstrated its association with the C-terminal domain (CTD see glossary, Box 1) of RPB1, the largest subunit of RNA polymerase II (RNAPII). Subsequent proteomic analyses [2-4], while confirming its composition and association with RNAPII, identified possible additional subunits as well as new potential cofactors. Among these proteins, only C12orf11 (also known as Asunder) and C15orf44 (also known as VWA9 or CG4785) have since been proven to be functionally associated with Integrator and renamed INTS13 and INTS14, respectively [5].</p><p>Unlike the Mediator Complex, a multi-subunit complex required for regulated transcription of most RNAPII dependent genes [6,7], Integrator is restricted to metazoans [8]. Its molecular weight is estimated by size exclusion chromatography to be greater than 1 MDa [1,9]. The size of its subunits, in humans, ranges from 49 kDa for INTS12 to 244 kDa for INTS1, with the majority (eight out of fourteen) possessing molecular weights greater than 100 kDa (Figure 1). Few subunits have identifiable paralogs within the human genome and, despite their number and relatively large size, their sequence is strikingly devoid of readily recognizable domains. The most common predicted motifs within Integrator subunits are alpha-helical repeats such as HEAT, ARM and TPR, or VWA domains [10]. These structures are suggestive of protein-protein interaction surfaces, but fail to provide insight into the function of their respective subunit in the complex or into a potential interaction partner. Contrastingly, two subunits, INTS11 and INTS9 [11], are clearly homologous with CPSF73 and CPSF100, proteins that are involved in the cleavage of pre-messenger RNAs [12] and belong to a large group of zinc-dependent nucleases called the β-CASP family [13] (CPSF, Artemis, SMN1/PSO2, see glossary). This relationship was instrumental in implicating Integrator in the 3′end formation of cellular RNA.</p><!><p>Apart from U6, a typical RNAPIII-dependent transcript whose 3′end is generated by transcription termination driven by a thymidine stretch [14], all UsnRNAs (see glossary) are synthesized by RNAPII. Prior to the discovery of Integrator, extensive work had defined the three requirements governing RNAPII-dependent UsnRNA 3′end formation: i) an UsnRNA-type promoter containing two characteristic elements: a distal sequence element (DSE) that recruits the transcription factors Oct1 and Sp1, and a proximal sequence element (PSE) that is bound by the snRNA activating protein complex (SNAPc, see glossary) [15,16], ii) the CTD of RNAPII [17,18] and iii) a consensus sequence GTTTN0-3AAARNNAGA called the 3′box, which is located 9-19 nucleotides downstream of the 3′end of the UsnRNA [19]. These requirements led to the hypothesis of a co-transcriptional mechanism: a unique factor is recruited to UsnRNA promoters where it associates with the RNAPII CTD and cleaves the nascent pre-UsnRNA once the 3′box is transcribed and recognized.</p><p>Integrator proved to be this long sought factor (Figure 2). Multiple biochemical purifications indicated that it associates with the RNAPII CTD [1,20]. Chromatin immunoprecipitation (ChIP) experiments showed that Integrator is present at the promoter, body and 3′end of the UsnRNA genes in a pattern suggesting that it travels along with the RNAPII as it transcribes the UsnRNA [1,9]. Finally, RNAi-mediated knock-down of various Integrator subunits leads to the accumulation of elongated misprocessed pre-UsnRNA [1,21]. Given that INTS11 is paralogous to CPSF73 [11], and that the overexpression of an INTS11 mutant predicted to be catalytically dead interferes with UsnRNA 3′end processing [1], it is presumed that INTS11 is the enzyme responsible for the pre-UsnRNA cleavage. Integrator knockdown also results in increased RNAPII density downstream of the 3′end cleavage site, indicating that on UsnRNA genes 3′end processing is linked to transcription termination [22]. Whether the termination event is coupled with 3′end cleavage through a mechanism similar to the torpedo model for mRNA genes [23] or if Integrator directly regulates termination remains to be determined.</p><!><p>Of the three functions postulated for Integrator at the UsnRNA genes [recognition of the UsnRNA promoter, specific binding to the RNAPII CTD (Box 1), and cleavage of the nascent UsnRNA, (Figure 2)], the interaction with the RNAPII CTD has been the most thoroughly investigated. It has been clearly established that Integrator shows a strong preference for a ser7P/ser2P dyad (YS2PTS5PS7YS2PTS5PS7) while ser5 phosphorylation appears to be detrimental to Integrator recruitment [20]. Experimental evidence indicated that the RNAPII Associated Protein 2 (RPAP2, homolog of the yeast atypical phosphatase rtr1 [24]) removes the ser5P mark on UsnRNA genes. RPAP2 affinity purified from mammalian cells co-elutes with both Integrator and RNAPII and its interaction with the RNAPII CTD is ser7P-dependent. Moreover, purified human RPAP2 protein exhibits ser5P phosphatase activity in vitro and its knockdown in mammalian cells results in elevated levels of ser5P, decreased Integrator occupancy on the UsnRNA genes and accumulation of misprocessed UsnRNAs [9]. Altogether, these results were coalesced into a model (Figure 3, left pathway) where RNAPII is initially phosphorylated on ser5 and ser7 by TFIIH (Transcription Factor IIH, see glossary) at the UsnRNA promoter [25]. This phosphorylation pattern, in turn, recruits RPAP2 through ser7P and results in the removal of the ser5P mark. Finally, ser2 phosphorylation by p-TEFb (Positive Transcription Elongation Factor b, see glossary) creates the substrate required for optimal Integrator recruitment and subsequent 3′end processing.</p><p>Nevertheless, the role of RPAP2 as well as the exact mechanism by which the ser5P mark is removed has been debated. Recent studies established that RPAP2 phosphatase activity is low [26] (with a turnover rate several orders lower than other known CTD serine phosphatases) and that the enzyme lacks a proper grove or pocket that could fulfill the role of an active site [27]. However, these shortcomings have been mitigated by the recent characterization of two RNAPII CTD binding proteins, Regulation Of Nuclear Pre-MRNA Domain Containing 1A and 1B (RPRD1A and RPRD1B). These proteins form homo- or heterodimers through a coiled-coil domain, bind to the ser2P or ser7P marks on the RNAPII CTD and stimulate RPAP2 phosphatase activity toward ser5P through protein-protein interactions [28]. These findings suggests a possible alternative model (Figure 3, right panel) where an RPRD1A/RPRD1B dimer binds two ser7P marks bracketing a ser5P mark in order to recruit and position RPAP2 optimally toward its substrate.</p><p>Although this model presents a parsimonious solution to explain RPAP2 function in regulating RNAPII CTD phosphorylation, it probably overlooks other roles for RPAP2 in the RNAPII transcription cycle. Recent cellular biology experiments demonstrated that, similar to its yeast homolog rtr1, RPAP2 cellular localization is predominantly cytoplasmic [29]. Considering that two known RPAP2 interacting proteins, GPN-loop GTPase 1 (GPN1) and GPN-loop GTPase 3 (GPN3), play an important role in RNAPII biogenesis and nuclear import [30,31], RPAP2 cytoplasmic localization raises the possibility of a similar role. Moreover it was shown that RPAP2 binds not only the CTD of RPB1 but also to its N-terminal domain [29]. This interaction could correspond to a different function for RPAP2 or could participate in the ser5P mark removal by stabilizing the interaction between RPB1 and RPAP2 to compensate for its slow phosphatase activity.</p><p>Beyond their involvement in Integrator recruitment, there is a general question about ser7P and RPAP2 role in UsnRNA transcription and 3′end processing. Two independent studies using an inducible knockout system in chicken cells investigated the role of the Ssu72 phosphatase and of ser7P in transcription [32,33]. It was found that, similar to RPAP2, knocking out Ssu72 results both in increased ser5P marks on UsnRNA genes and defective UsnRNA processing, indicating a possible redundancy between Ssu72 and RPAP2 [32]. In addition, the substitution of ser7 to alanine in the RNAPII CTD, hence preventing its phosphorylation, showed little effect on UsnRNA processing [33], complicating our interpretation of ser7P and RPAP2 role in UsnRNA processing. The initial work identifying ser7P role in UsnRNA processing [9] relied on α-amanitin resistant RNAPII mutant complementation. Although this method has proven to be a powerful tool to study transcription, prolonged α-amanitin exposure is not without consequences and results obtained through this approach should be interpreted with caution. For example, the elongation factor DSIF (DRB Sensitivity Inducing Factor, see glossary), whose knockdown negatively affects RNAPII recruitment on snRNA genes [34] and that directly interacts with Integrator (see below), is targeted for rapid degradation by α-amanitin even in the presence of the α-amanitin resistant RNAPII mutant [35]. Finally, there is a broader question about ser7P and RPAP2 function in general transcription. Indeed, in eukaryotes ser7P is present on all RNAPII transcribed genes and despite its presumed importance, ser7 to alanine substitution is not lethal in yeast (or chicken cells) and does not result in increased global ser5 phosphorylation as would be predicted [33,36]. Similarly, recruitment of RPAP2 to protein coding genes appears to be independent of ser7 phosphorylation [9].</p><p>Altogether, these data indicate that RPAP2 is most likely involved in removing the RNAPII CTD ser5P mark to facilitate Integrator recruitment, in particular on UsnRNA genes. However, there does not appear to be a direct relation between ser7 phosphorylation, RPAP2 recruitment, and ser5P removal genome-wide; possibly because other protein partners and redundant mechanisms are affecting this relationship. Identifying these factors represents an upcoming challenge to understand how Integrator is temporally and spatially recruited to a specific gene. Conversely, identifying the Integrator subunit(s) involved in the RNAPII CTD recognition will also be critical to answer this question.</p><!><p>A fascinating development in the study of the Integrator biology is the recent discovery of its relationship with the transcription elongation machinery. ChIP experiments revealed that the negative elongation factor NELF (Negative Elongation Factor, see glossary) accumulates at the 3′end of UsnRNA genes. Its knockdown results in increased RNAPII occupancy downstream of the 3′box and in the accumulation of long readthrough transcripts, reflective of a termination defect and of a potential functional interaction with Integrator [34,37,38]. This observation is consistent with a recent affinity purification of SPT5 (see DSIF in glossary) and of the NELF subunit NELF-E that revealed a physical interaction between Integrator and these factors [34]. The interplay between Integrator, SPT5, and NELF on UsnRNA genes is particularly interesting for several reasons. SPT5 appears to function early in the transcription cycle as its knockdown results in a reduction of RNAPII, NELF, and Integrator density at the UsnRNA genes [34]. Conversely, knockdown of NELF-E or Integrator both results in accumulation of RNAPII and SPT5 on the 3′end of the genes and in the accumulation of long misprocessed transcripts [34,37,38]. Therefore, even if both SPT5 and NELF interact with Integrator, their role in relation to the complex seems functionally distinct with SPT5 playing a possible role in transcription initiation and Integrator recruitment while NELF most likely functions in UsnRNA 3′end processing and transcription termination.</p><!><p>The connection between Integrator, SPT5/NELF and elongation revealed its full significance with the recent evidence for Integrator function at mRNA coding genes [39–41]. The initial study of Integrator using conventional ChIP analysis was limited by the small repertoire of high quality antibodies available at the time and failed to demonstrate its presence on protein coding genes [1]. As more antibodies became available, Gardini et al. used ChIP-seq to revisit Integrator occupancy genome-wide and uncovered its association with active mRNA transcription and enrichment at Immediate Early Genes (IEG, see glossary) after epidermal growth factor (EGF) stimulation [39]. Transcriptional regulation of these genes functionally resembles that of the Drosophila Heat Shock gene (HSP70), which is the archetype for RNAPII pause-release (Box 2). Interestingly, the authors observed that under starvation conditions, low levels of Integrator were specifically detected at IEG transcription start sites (TSSs); however after EGF stimulation Integrator occupancy markedly increased at the TSS and within the body of the gene. This localization proved functionally relevant as the knockdown of Integrator subunits (INTS1 and INTS11) abrogates responsiveness of IEGs to EGF stimulation. Importantly, upon Integrator depletion there is a failure of the RNAPII to escape pausing and progress into productive elongation. Mechanistically, the role for Integrator in transcriptional pause-release appears to stem from its capacity to recruit the positive elongation factors p-TEFb and SEC to promoter proximal paused genes upon activation.</p><p>The second study linking Integrator to RNAPII pause-release on mRNA coding genes originates from the investigation of NELF function in Tat-activated transcription of the HIV-1 long terminal repeat (LTR, see glossary) [41]. While purifying the NELF complex from HeLa cells, Stadelmeyer et al. detected low but significant amounts of Integrator, prompting them to explore its function in HIV transcription. They found that Integrator is recruited along with NELF to the HIV-1 TAR element (see glossary) and that knocking down either INTS11 or INTS9 (but not INTS3) resulted in loss of promoter proximal pausing. They then observed that Integrator function is not restricted to the HIV LTR as a common set of genes (>2000) were differentially expressed in response to NELF, INTS3 or INTS11 knockdown in asynchronously growing cells. Consistent with this finding, the analysis of RNAPII, Integrator and NELF ChIP-seq read densities at the TSS of these genes revealed that Integrator and NELF binding closely correlates with the amount of RNAPII pausing at the TSS. Furthermore, in genes bound by NELF and Integrator, knockdown of INTS11 resulted in increased RNAPII occupancy and RNA-seq read density on the gene body, reflective of a promoter proximal pausing defect. Interestingly, INTS3 knockdown had the opposite effect on both LTR and mRNA-coding gene transcription resulting in decreased RNAPII occupancy and RNA-seq read density on the gene body. Whether this effect reflects of antagonistic roles for INTS3 and INTS11 within the complex or of the existence of functionally distinct Integrator subcomplexes remains to be determined.</p><p>Although both studies clearly implicate Integrator in the regulation of pause-release and elongation (Figure 4), an apparent discrepancy exists between the observed phenotypes. In Gardini et al., INTS11 knockdown decreases RNAPII density as well as RNA-seq reads on the body of IEGs while the work by Stadelmeyer et al. describes the opposite behavior. This possibly reflects the dual role of NELF dependent pausing that attenuates transcription under non-induced conditions while at the same time maintaining an active open chromatin state at the promoter. Indeed, the study conducted by Gardini et al. focused on the transcriptional response of IEGs in serum starved cells after EGF induction which is affected mostly by RNAPII pausing and release. In contrast, the work conducted by Stadelmeyer et al. uses asynchronously growing cells and considered a wider range of transcriptional responses, in particular genes whose transcription is stimulated after NELF and Integrator depletion. Regardless of the differences, the data presented in both of these studies indicate that there is a role for Integrator in the transcriptional regulation of protein encoding genes. The details of this function are likely going to depend on the cellular context and the nature of the signal produced to alter gene expression.</p><!><p>While it can be biochemically purified as a single entity, Integrator appears to act as a modular complex on the genome. ChIP experiments conducted on UsnRNA genes in human cells or on the HSP70 gene in fly show that different subunits give distinct occupancy patterns. Human INTS5 shows a predominant occupancy from the promoter region through the 3′end of the U2 snRNA while INTS11 is mostly present at the 3′end of the gene [9]. Similarly, Drosophila INTS12 is present at the HSP70 promoter and peaks at the transcriptional pausing site while INTS9 occupancy is shifted toward the 3′end of the gene with a marked peak in the body of the gene [39]. These observations could indicate a sequential recruitment and different functions during the transcription cycle. Early recruitment of INTS5 or INTS12 could reflect the existence of a module with a primary role in transcription initiation and pausing while later recruitment of INTS9 and INTS11 could identify a module with a role in elongation and 3′end processing. Alternatively, Integrator could exist as a single complex but would be subject to significant conformational remodeling during the transcription cycle resulting in a change in accessibility by ChIP. On mRNA coding genes, the opposing effects of INTS3 and INTS11 knockdown on transcription also suggests the existence of functionally distinct modules [41]. Currently, our understanding of Integrator occupancy throughout the genome is fragmentary because only a limited number of subunits have been fully mapped. Determining the localization of more subunits and the impact of their knockdown on transcription will be essential to clearly identify functional and structural submodules within Integrator. Moreover, such studies might reveal additional unsuspected functions for the Integrator complex.</p><p>Another interesting aspect of Integrator is the association of some of its subunits into functionally unrelated complexes as exemplified by the association of INTS3 and INTS6 with the SOSS (Sensor of single stranded DNA, see glossary) complex [42-46]. We can therefore speculate that the presence of Integrator and potentially SOSS, through its interaction with INTS3 and INTS6, at transcriptional pause sites might also serve a role in maintaining genome integrity. Indeed, regions of the genome with an open chromatin state such as UsnRNA genes or proximal promoter pause sites are more fragile and prone to genome instability [47-49]. UsnRNA genes have been shown to be particularly sensitive to Ad12-induced chromosome instability and display a weak constitutive fragility in cells defective in transcription-coupled nucleotide excision repair [47,49]. This genome instability is transcription-dependent, as inactive UsnRNA pseudo genes do not display such fragility. Similarly, promoter proximal pausing maintains a constitutively open chromatin state that is favorable to the formation of R-loop structures where the nascent RNA transcript falls back on the template DNA strand, leaving the single stranded non-template strand exposed [50,51]. Similar structures also form at the transcriptional pausing site downstream of the polyadenylation signal and help in the transcription termination process [52]. While R-loops can have a positive effect on transcription initiation and termination, they also present a risk for genome stability if not properly resolved as the exposed non-template strand becomes more susceptible to DNA damage. Therefore, the presence of Integrator, through the ribonuclease INTS11 or its SOSS-interaction subunits INTS3 and INTS6, might have a part to play in the prevention of DNA damage induced by R-loops or constitutively open chromatin states.</p><!><p>Altogether, recent biochemical, genomic and functional data elevates the Integrator complex to the status of a primary RNAPII cofactor involved in many steps of the transcription cycle: initiation, pause-release, elongation, 3′end processing and termination. Nevertheless, many aspects of the recruitment of Integrator to the RNAPII remain to be elucidated (see Box 3, Outstanding Questions). Indeed, the model of recruitment of Integrator to UsnRNA versus mRNA genes appears to be in open conflict. The most obvious discrepancy lies in the role played by the RNAPII CTD phosphorylation and the responsible and/or associated kinases, which includes the particularly concerning example of ser2 phosphorylation. It is established that ser2 phosphorylation is necessary in vivo and in vitro for efficient binding of Integrator to the RNAPII CTD. Furthermore, on UsnRNA genes ser2 phosphorylation appears to coincide with INTS11 recruitment, leading to the current model where ser2 phosphorylation by p-TEFb actually triggers the recruitment of Integrator (at least INTS11) leading to efficient UsnRNA 3′end processing. On the contrary, the work conducted on mRNA coding genes tends to demonstrate that INTS11 recruitment precedes and is necessary for the recruitment of p-TEFb and subsequent ser2 phosphorylation. The convenient interpretation, but also the least intellectually satisfactory, is that completely distinct mechanisms govern the recruitment of Integrator, p-TEFb and ser2 phosphorylation on UsnRNA and mRNA coding genes. Due to the short size of the UsnRNA transcription units, ChIP based techniques are probably unable to precisely analyze the interplay between these factors on these genes. Only the precise characterization of how Integrator interacts physically and temporally with the different actors of the transcription cycle (RNAPII, DSIF, NELF, p-TEFb) will bring a clear answer to these essential questions.</p>

PubMed Author Manuscript

Targeting Fluorescent Nanodiamonds to Vascular Endothelial Growth Factor Receptors in Tumor

The increased expression of vascular endothelial growth factor (VEGF) and its receptors is associated with angiogenesis in a growing tumor, presenting potential targets for tumor-selective imaging by way of targeted tracers. Though fluorescent tracers are used for targeted in vivo imaging, the lack of photostability and biocompatibility of many current fluorophores hinder their use in several applications involving long-term, continuous imaging. To address these problems, fluorescent nanodiamonds (FNDs), which exhibit infinite photostability and excellent biocompatibility, were explored as fluorophores in tracers for targeting VEGF receptors in growing tumors. To explore FND utility for imaging tumor VEGF receptors, we used click-chemistry to conjugate multiple copies of an engineered single-chain version of VEGF site-specifically derivatized with trans-cyclooctene (scVEGF-TCO) to 140 nm FND. The resulting targeting conjugates, FND-scVEGF, were then tested for functional activity of the scVEGF moieties through biochemical and tissue culture experiments and for selective tumor uptake in Balb/c mice with induced 4T1 carcinoma. We found that FND-scVEGF conjugates retain high affnity to VEGF receptors in cell culture experiments and observed preferential accumulation of FND-scVEGF in tumors relative to untargeted FND. Microspectroscopy provided unambiguous determination of FND within tissue by way of the unique spectral shape of nitrogen-vacancy induced fluorescence. These results validate and invite the use of targeted FND for diagnostic imaging and encourage further optimization of FND for fluorescence brightness.

targeting_fluorescent_nanodiamonds_to_vascular_endothelial_growth_factor_receptors_in_tumor

INTRODUCTION<!>FND Functionalization for Click Chemistry.<!>Biocompatibility and Functional Activity of Derivatized FND in Tissue Culture.<!>Imaging of FND-scVEGF Accumulation in Tumors.<!>DISCUSSION<!>CONCLUSIONS<!>MATERIALS AND METHODS<!>Nanodiamond Click Chemistry and VEGF Reagent Synthesis.<!>ELISA.<!>In Vitro Cell Viability.<!>Shiga-Like-Toxin VEGF Competition Assay.<!>VEGFR-2 Tyrosine Kinase Autophosphorylation in VEGFR-2 Overexpressing Cells.<!>Labeling of Endothelial Colony-Forming Cells and Fibroblasts.<!>Animal Model and IV Administration of FND.<!>In Vivo and Ex Vivo Imaging.

<p>Though fluorescence imaging plays a powerful role in research development and clinical diagnostics,1–5 the limited photostability and biocompatibility of current fluorophores constrain potential applications (e.g., longitudinal studies of cellular dynamics or continuous, long-term surgical imaging). In contrast, fluorescent nanodiamonds (FNDs) display true infinite photostability,6–9 and their biocompatibility is well-established.10–12 The advantages of FND have been explored in numerous applications, from fluorescence imaging to advanced sensing13–15 and testing in several animal models.8,16–18 However, to date, FNDs have not been explored for targeted tumor imaging utilizing their intrinsic fluorescence.</p><p>Vascular endothelial growth factor A (VEGF) and its two main receptors VEGFR-1 and VEGFR-2 play important roles in normal and pathologic angiogenesis.19 Binding of VEGF to VEGFRs causes dimerization of transmembrane receptor proteins and subsequent activation of their tyrosine kinase activity within the cell, which initiates a number of signaling pathways. This signaling is critical for endothelial cell proliferation, viability, and function, regulating growth of new blood vessels (i.e., vasculogenesis and angiogenesis) and vascular permeability, as well as cell migration, inhibition of apoptosis, and recruitment of progenitor and hematopoietic cells from bone marrow to tumor.20,21 Overexpression of VEGF and its receptors is associated with a number of pathologies, including the growth of primary tumors and metastatic lesions.22–25 Importantly, VEGFR-1 and VEGFR-2 may have different roles in cancer progression,26 whereby VEGFR-2 is a particularly well-characterized marker of tumor angiogenesis,27–31 while VEGFR-1 may be involved in setting protumorigenic microenvironments and contributing to metastatic growth.32–34</p><p>VEGFRs are well-recognized as important therapeutic targets. To this end, many drugs have been and are being developed to inhibit either VEGF binding to the receptors or the tyrosine kinase activity of these receptors.32 In turn, the therapeutic relevance of VEGF receptors motivates development of various tracers for targeted imaging of these receptors.35 Considering the importance of VEGF receptors and potential translational opportunities for effective VEGFR-targeting tracers, we selected these receptors for assessing FND's potential for targeted imaging. We hypothesized that using re-engineered VEGF as a targeting moiety, we could employ FND for selective imaging of VEGF receptors. Herein, a methodology for facile click-chemistry conjugation of VEGFR-targeting ligands, specifically an engineered single chain (sc) VEGF, to FND was developed. The functional activity of the particle-conjugated scVEGF moiety was validated in vitro, and enhanced accumulation of FND-scVEGF versus untargeted FND was observed in a murine tumor model, highlighting the potential for FND for targeted imaging in vivo.</p><!><p>The emission spectrum for the starting FND is shown in Supplementary Figure S1. FND with a poly(glycerol) shell (FND-PG) was prepared as described previously.36–38 This shell increases the particle diameter by ~20 nm (Supplementary Figure S2) and increases the colloidal stability of FND in buffers. To activate FND-PG for the click-chemistry reaction between tetrazine and trans-cyclooctane (Scheme 1), FND-PG were functionalized with methyltetrazine (mTz) amine using carbonyldiimidazole (CDI)-mediated activation of poly-(glycerol) hydroxyl groups in FND-PG to form a carbamate linkage.39 The increase in the concentration of CDI in the reaction mixture led to the higher surface densities of conjugated mTz, which was readily visualized as an enhanced pink color in pelleted FND-mTz (Figure 1a). The chemical reactivity of FND-conjugated mTz moieties was confirmed by click-reaction (Scheme 1) with Cy5-TCO, which is blue in color. The enhanced density of conjugated mTz led to an enhanced blue color in pelleted FND-Cy5 (Figure 1b). UV-vis absorption spectra of resuspended FND-Cy5 and control FND-PG were obtained after removal of background Rayleigh particle scattering by polynomial subtraction (Figure 1c). FND-Cy5 spectra, but not control FND-PG spectra, show an absorption maximum at 650 nm with CDI-dependent intensity and a shoulder at 605 nm that were due to Cy5. Smaller peaks at 560 and 585 nm were observed in samples with higher mTz densities due to absorbance from these ligands. We then used the molar absorptivity of Cy5 at 650 nm to calculate the average density of reactive mTz per particle as a result of the different CDI reaction concentrations used (27, 20, 13, and 7 mg/mL CDI). The most concentrated sample activated with 27 mg/mL CDI exhibited up to 500 mTz per particle; however, at this mTz density, colloidal stability was affected. Thus, particles with lower Cy5 measured densities (~200, 100, and 30 mTz) per particle were used for further validation or in vivo work.</p><!><p>Three preparations of FND-mTz with approximate mTz surface densities of 30, 100, and 200 mTz/FND, named FND-mTz Low, Middle, and High, were prepared as described above and used in tissue culture experiments. Although all FND-mTz formed dense precipitates on the cell surface, 293/KDR cell growth was not affected in the concentration range of 1–350 pM FND-mTz (Figure 2A). This is in agreement with the results previously reported for poly(glycerol) covered FND that were tested with colony forming endothelial cells (ECFCs).38</p><p>For targeting VEGF receptors, preparations of FND-mTz with different surface density of mTz were functionalized with scVEGF-TCO, an engineered single-chain version VEGF121 expressed with N-terminal cysteine-containing tag (Cys-tag), which was site-specifically derivatized with TCO. Click reactions were performed at concentrations of scVEGF-TCO twice higher than calculated mTz concentrations. As a positive control for such reaction, we used agarose-Tz beads (Ag-Tz). After pelleting all FND from click-reaction mixtures, SDS-PAGE of supernatants indicated a significant decrease in the amounts of free scVEGF-TCO relative to the input scVEGF-TCO (compare intensity in lane 1 and lanes 2–5 in Figure 2b). Virtually no scVEGF-TCO was detected in high-salt washes of the corresponding pellets (lanes 6–9 in Figure 2b), indicating that association of scVEGF with FND was not due to a nonspecific binding of scVEGF-TCO to pelleted FND-mTz or Ag-Tz.</p><p>The amount of scVEGF-TCO click-conjugated to FND-mTz with different mTz surface density was estimated in two different assays. First, the upper limits of conjugated scVEGF-TCO was determined from the intensities of the residual scVEGF-TCO bands vs the intensity of the input scVEGF-TCO band and was found to be in the range of 200–300 scVEGF/FND depending on initial mTz surface density in FND-mTz (Table 1). However, considering that not all surface-bound scVEGF-TCO can be simultaneously spatially accessible for interactions with the cellular VEGF receptors, we used a sandwich enzyme-linked immunosorbent assay (ELISA) to assess the fraction of scVEGF in FND-scVEGF that was capable of interacting with other proteins. Using free VEGF to calibrate ELISA, we found that the number of scVEGFs per FND that was responsible for binding of anti-VEGF antibodies to plate-bound FND-scVEGF was at least an order of magnitude lower than the total estimated amount of scVEGF per FND (Table 1).</p><p>Next, we tested FND-scVEGF preparation with the highest scVEGF per FND, as determined by ELISA, in the cell protection assay that was previously developed for characterization of scVEGF and its conjugates. In this 72-h assay, scVEGF-conjugates are tested for their ability to protect VEGFR-2 overexpressing 293/KDR cells from VEGFR-2 mediated cytotoxicity of Shiga-like-toxin (SLT)-VEGF fusion toxin containing Shiga-like toxin enzymatic subunit A genetically fused to VEGF121 isoform.40 The morphology of 293/KDR cells exposed to SLT-VEGF changes dramatically from a normal growing monolayer of cells to few clusters of dying cells (Figure 3a, upper left panel). As expected, untargeted FND-mTz did not rescue cells (Figure 3a, upper right panel). In contrast, we found that FND-scVEGF, but not FND-mTz, protected cells from cytotoxic SLT-VEGF in a dose-dependent manner (Figure 3a, lower panels, and Figure 3b). For dose-dependence analysis, we used the ELISA-based determination of ~40 accessible scVEGF moieties per FND. This surface density of accessible scVEGF led to a calculated EC50 for FND bound scVEGF of 3.5 nM, while EC50 for free scVEGF was 2 nM.</p><p>We then tested the ability of FND-scVEGF prepared with FND-mTz high and middle to activate VEGF-mediated tyrosine autophosphorylation of the VEGFR-2 receptor in 293/KDR cells engineered to overexpress VEGFR-2.40 We found that both types of FND-scVEGF were active in this assay (Figure 3c). Although FND-scVEGF were somewhat less active than free scVEGF-TCO, tyrosine phosphorylation reached saturation at nanomolar concentrations of "accessible" FND-conjugated scVEGF.</p><p>Finally, we tested cellular uptake of FND-scVEGF and control FND-PG in two primary human cell types: human endothelial colony forming cells (ECFCs, #CB002) and human foreskin fibroblasts (CCD1137). Human endothelial cells (e.g., human umbilical vein endothelial cells, human dermal microvascular endothelial cells, and human dermal lymphatic microvascular endothelial cells) express high level of VEGFR-1 and VEGFR-2, while primary fibroblasts express only low levels of VEGFR-1 and no VEGFR-2.42–44 Microscopy revealed readily detectable colocalization of FND-scVEGF, but not untargeted FND-PG with ECFCs, though were not present in nuclei (Figure 4a, Supplementary Figure S3). Although the intensity of FND-scVEGF per cell was variable, virtually all cells showed dose-dependent FND-scVEGF association, with saturation at ~2 pM (Supplementary Figure S4). Interestingly, the association of targeted FND-scVEGF was also readily detectable in experiments with CCD1137 fibroblasts, which express only low level of VEGFR-1 receptors. However, quantitative analysis of fluorescence intensity indicated a higher binding of FND-scVEGF in ECFCs (CB002) as compared to fibroblasts (CCD1137) (Figure 4B). Of note, the differential association of FND-scVEGF by ECFCs and CCD1137 was more prominent at higher FND-scVEGF concentrations (Supplementary Figure S5).</p><!><p>Tumor-bearing mice were injected with targeted FND-scVEGF and untargeted FND-PG. Though whole-body imaging of mice was attempted, significant visualization of FND uptake over time with the particle size studied (140 nm) was not possible. To characterize tracer uptake, tumors were harvested, cryosectioned, and investigated by epifluorescence microscopy, with representative images shown in Figure 5A–C. Tumors harvested from FND-scVEGF injected mice showed increased fluorescence as compared to untargeted FND-PG control (Supplemental Figure S6). Quantitative analysis of cryosections revealed a significantly higher density of FND in tumors from animals injected with targeted FND-scVEGF relative to those from animals injected with untargeted FND-PG (p < 0.0075) and to control (p < 0.0055) (Figure 5D). Note that for analysis, thresholding was performed to discriminate against background and artifact fluorescence; artifact signal not overlapping with tissue was not counted.</p><p>In a separate imaging setup, the FND origin of this fluorescence was then confirmed with microspectroscopy, whereby presence of FND nitrogen-vacancy center signals, specifically readily detectable zero phonon lines at 575 and 648 nm were analyzed (Figure 6). Additionally, red fluorescence was detected under both blue (470/28 nm) and green (517/20 nm) excitation. The long Stokes shift under blue excitation exhibited by red FND fluorescence was useful in discriminating against background fluorescence. Generally, particle clustering was most readily observed toward the periphery of the tumor section, with particles less readily visible toward the center (Supplementary Figure S7). Particles were visible at all explored objectives (10×, 40×, and 100×); however, 40× and 100× gave the most apparent contrast.</p><p>Though there was significantly lower uptake of untargeted FND-PG with fluorescence only slightly above background, it is important to note the presence of untargeted FND-PG could be detected. Sensitivity of detection of untargeted FND-PG uptake was improved by rastering a 532 nm laser focused as a point illumination source. When doing so, homogeneous and fine point sources of FND fluorescence could be observed throughout the tumor in a speckled pattern at 100× magnification. This distribution was entirely different from that of FND-scVEGF and was not present in the control sample. Due to the low contrast and homogeneous visualization of FND-PG, it was difficult to obtain either images or spectra, however in some events it was possible to confirm these were indeed FND (Supplementary Figure S8). This difference in fluorescence contrast agreed with the non-significant increase in signal from untargeted FND as compared to control, as observed in Figure 5D. Interrogation of FND-scVEGF administered samples under 100× objective displayed what appeared to be ordered clustering (Supplementary Figure S9). Finally, tumor sections were imaged via two-photon imaging with 810 nm excitation, again confirming particles present only in the targeted sample (Supplementary Figure S10).</p><!><p>In this study, we report the development of FND that are functionalized with scVEGF-TCO for targeting VEGF receptors. To facilitate FND functionalization through a standardized and facile procedure, we used FND-PG-FND coated with a uniform hydroxylated surface provided by poly(glycerol).36–38 This coating provides not only readily derivatizable hydroxyl groups but also an enhanced colloidal stability in buffers and elevated ionic strength. In turn, FND-PG was used for synthesis of FND-mTz, where methyltetrazine moieties (mTz) provide for rapid, catalyst-free, room temperature click-chemistry reaction with TCO. We found that the surface density of mTz in FND-mTz can be conveniently controlled by changing concentration of CDI used for the activation of hydroxyl groups in FND-PG for reaction with the mTz amino group (Figure 1). Importantly, in 72-h tissue culture cytotoxicity assay, FND-mTz was not toxic to 293/KDR cells up to concentrations as high as 350 pM. These results are consistent with a growing body of literature that support that FND display strong biocompatibility.10–12</p><p>To estimate the surface density of the accessible and reactive mTz, we used a click-reaction between FND-mTz and Cy5-TCO, a cyanine dye with the high molar absorptivity at wavelength longer than 600 nm, where scattering from FND is minimized. Using this method, we could correlate mTz densities up to 500 mTz moieties on the FND-mTz surface based on the CDI concentration used. The highest levels mTz densities, however, showed diminished colloidal stability. Thus, for further validation, particles with densities around 200 mTz per FND and lower were used.</p><p>To assess functionalization of FND-mTz with targeting protein scVEGF-TCO, we used several complementary approaches. SDS-PAGE analysis of the scVEGF-TCO-FND-mTz reaction mixture was used to determine a decrease in free scVEGF-TCO after click-reaction. Since we found that the contribution of nonspecific binding scVEGF-TCO to FND-mTz was negligible, such SDS-PAGE analysis provided an upper limit estimation for the overall amount of scVEGF on FND-scVEGF surface, ~300 scVEGF/particle, which is in accordance to the number found by click-reaction with Cy5-TCO. On the other hand, we have used an ELISA sandwich assay with antibodies against different VEGF epitopes to determine the presence of "accessible" scVEGF on FND-scVEGF surface. We reasoned that since the plate-captured unit is the entire FND-scVEGF particle, only the accessible scVEGF on the side opposite to captured surface (i.e., not all scVEGF on the particle) can be detected with the second antibody. Indeed, using free scVEGF-TCO for calibration, we found that ELISA assay yields approximately 30 scVEGF per particle, or 10-fold lower than estimations based on Cy-TCO reaction or SDS-PAGE analysis of reaction mixture.</p><p>Despite this disagreement, the ELISA result on the "accessible" scVEGF appears to provide a better estimation for the surface scVEGF moieties that are capable of interacting with VEGF receptors. In two different cell-based assays, (i) induction of cellular VEGFR-2 tyrosine autophosphorylation and (ii) protection against SLT-VEGF VEGFR-2 mediated cytotoxicity, dosing based on ELISA results yielded activities per conjugated scVEGF close to that of free scVEGF (Figure 3). This agreement between cell-based and ELISA assay is not surprising, considering that topologically each FND-scVEGF nanoparticle can utilize only a limited number of cell-surface scVEGF moieties for binding to cellular VEGFR-2 receptors. However, we cannot exclude the possibility that in addition to accessibility, surface-bound scVEGF could be functionally inactivated through some additional mechanisms.</p><p>To assess the imaging potential of FND-scVEGF, we tested their ability to selectively accumulate in tumors was explored in a 4T1 mouse tumor model. We found that intravenous injection of targeted FND-scVEGF resulted in their significant tumor uptake (Figures 5 and 6). A positive identification of tissue-captured FND can be made through several methodologies, (i) red fluorescence under blue excitation, (ii) deeper (longer) red fluorescence under green excitation, (iii) pinkish/red coloration under UV illumination, (iv) lack of photobleaching after prolonged illumination, and (v) spectral confirmation with laser excitation. The highest density of FND-scVEGF was found at the periphery of the tumor, in the area known as the angiogenic edge with highest density of angiogenic vasculature and, correspondingly, highest density of VEGF receptors. Similar preferential accumulation of scVEGF-based tracers in angiogenic tumor edge has been reported previously.45 As expected, untargeted FND-PG injected in tumor-bearing mice were also found in tumors. Nanoparticles are known to accumulate in tumors through leaky vasculature via nonspecific enhanced permeability and retention mechanism.46 However, the distribution of FND-PG accumulation was distinct from that of FND-scVEGF. The FND-PG particles that were present were spread homogeneously, in contrast to FND-scVEGF and could only be seen with laser rastering to observe small, individual point sources of light, which also appeared to a greater degree toward the tumor periphery as characterized by leaky angiogenic vasculature. Moreover, unlike FND-scVEGF, untargeted FND-PG did not display any clustering tendency and were difficult to visualize under broad illumination indicating low uptake. Supplementary Figure S8 highlights an example where they could be seen. It is important to note that spectra of FND within tissue (Figure 6) do not resemble exactly spectra alone (Supplementary Figure S1). The contribution from tissue autofluorescence artificially raises the seeming contribution from NV0 (575 nm). Though this does not directly affect standard fluorescent measurements, the presence of spin sensitive NV− allows for signal enhancing possibilities techniques like optically detected magnetic resonance.16,47</p><p>A beneficial outcome of targeting endothelial cells is that particles need not extravasate tissue for successful labeling. The selective accumulation of FND-scVEGF suggest that use in clinical applications for the labeling of tumor margins could be significantly advantageous. Furthermore, photostable characteristics of FND-scVEGF could be particularly useful in longitudinal studies of tumor progression. The field of multiphoton microscopy, which uses high laser power and provides increased penetration depth due to near-infrared excitation source, is a natural field where targeted FNDs' unique photostability would be especially applicable.</p><!><p>Fluorescent nanodiamonds (FND) were functionalized with re-engineered vascular endothelial growth factor scVEGF, which retain its ability to bind to cellular VEGF receptors. Facile click-chemistry functionalization was achieved via using FND coated with poly(glycerol) and derivatized with click-chemistry reagent mTz, while scVEGF was site-specifically derivatized with its click-chemistry partner, TCO. Conjugation of scVEGF was validated in vitro by SDS-PAGE, ELISA, and binding of FND-scVEGF to VEGFR-2 receptors was validated in functional tissue culture assays with VEGFR-2 overexpressing cells. In tumor-bearing mice, targeted FND-scVEGF accumulates in tumor angiogenic edge in readily detectable pattern, distinctive from the lower nonspecific accumulation of untargeted FND. These results validate and invite the use of targeted FND for diagnostic imaging and encourage further optimization of FND for fluorescence brightness.</p><!><p>Materials. The sources of main chemicals were as follows: fluorescent nanodiamond (140 nm) of high-temperature, high-pressure origin (NDNV140nmHi, Adámas Nanotechnologies, Inc.) was produced by irradiation with 3 MeV electrons of starting diamond powder and annealed at 850°C according to reported procedures38 and contains 3 ppm of NV centers according to EPR measurements.48 Their fluorescence spectra and size distribution are provided in Supplementary Figure S1 and S2. Additionally used were dimethylformamide (DMF, Sigma-Aldrich), carbonyldiimidazole (CDI, Sigma-Aldrich), glycidol (Sigma-Aldrich), methyltetrazine-PEG4-amine (mTz-PEG-NH2, Click Chemistry Tools), Cy5-transcyclooctene-(Cy5-TCO, Click Chemistry Tools), tetrazine agarose (Click Chemistry Tools), Dextrose (American Brewmaster Raleigh), SDS-PAGE (SibTech), VEGF enzyme-linked immunosorbent assay (ELISA, R&D Systems), and Cell Titer Assay (Promega). Cell lines used include human foreskin fibroblasts CCD-1137Sk (ATCC CRL-2703), human endothelial colony forming cells (ECFCs, clone #CB002, Creative Scientist, Inc.), and KDR293 cells (SibTech, Inc.).</p><!><p>Methyltetrazine (mTz) functionalized FND (FND-mTz) was synthesized by coupling mTz-PEG-NH2 to previously described poly(glycerol) coated FND (FND-PG).36–38 Briefly, hydroxylated FND, produced by reduction with lithium aluminum hydride,49 was reacted in neat glycidol at 140°C for 2 h followed by washing via centrifugation to produce FND-PG with high colloidal stability in aqueous buffers and solutions of elevated ionic strength. FND-PG was then transferred to DMF by centrifugation and resuspension. Hydroxyl groups of PG were then activated by addition of CDI stocks (10–40 mg/mL in DMF) to 1 mg/mL solutions of FND-PG in anhydrous DMF, producing a range of final CDI concentrations (6.6–26.6 mg/mL) for testing. After 1-h activation at room temperature, FND-PG particles were pelleted and washed once with DMF. Activated FND-PG were then resuspended in an aqueous solution of mTz-PEG-NH2 (0.9 mg/mL, DI) and allowed to react overnight. Particles flocculated by the following next day, and the resulting FND-mTz were pelleted and washed three times with an H2O/DMF mix (3:1). For validation, particles suspended in DI were incubated with excess Cy5-TCO, washed 4x, and analyzed via UV-vis (PerkinElmer Lambda 35). Particles at the highest mTz density showed some degree of instability with time. Thus, to compromise between mTz density and stability, the particles activated with CDI concentration at 17 mg/mL were used for in vivo administration.</p><p>scVEGF (SibTech, Inc.), a 241 amino acid single-chain (sc) variant of vascular endothelial growth factor (VEGF), comprising two 3–112 fragments of VEGF121 cloned head-to-tail and N-terminal 15-aa cysteine-containing tag, was site-specifically derivatized on C4 with TCO-PEG3 -maleimide (Click Chemistry Tools), as described previously for various payloads.40,50 For click-reaction, FND-mTz was concentrated, and a working stock of scVEGF-TCO in PBS/1%DMF was added to FND-mTz to a final concentration 2x relative to the expected mTz concentration (estimated from UV-vis analysis). In addition, a positive control of Agarose-Tz was similarly exposed to scVEGF-TCO. Samples were incubated 30 min at RT with agitation. Particles were spun down (5k rpm, 5 min), and supernatant was analyzed by SDS-PAGE. Particles were then resuspended in 0.5 mL of 0.5 M NaCl in PBS to desorb nonspecifically bound scVEGF-TCO and then spun down (5k rpm, 5 min), and supernatant was analyzed by SDS-PAGE. Finally, pellets were suspended in 100 μL of sterile PBS and stored at 4°C.</p><!><p>A commercial ELISA assay (R&D Systems) for VEGF was used for assessing FND-conjugated scVEGF. FND-scVEGF was captured on a plate coated with the primary anti-VEGF antibody, after which the amount of available scVEGF was determined by incubation with the secondary anti-VEGF antibody. A calibration curve was obtained with a VEGF standard according to manufacturer's instructions. The linear approximation of the calibration curve was used to calculate VEGF concentration in titrated FND-scVEGF samples. In addition to FND-scVEGF, nonderivatized FNDs were assayed alongside to determine level of potential nonspecific antibody binding to FND.</p><!><p>FND-mTz cytotoxicity was tested using 293/KDR cells (SibTech, Inc.), which are VEGFR-2 overexpressing derivatives of human embryonic kidney cells HEK293. Cells were plated on a 96-well plate at 2000 cell/well. After 20 h, 0.25 pmol FND-mTz were resuspended in culture medium (DMEM with 2 mM Glutamine, 10% FBS, antibiotic/antimycotic mixture) and serially diluted in complete culture medium at a range of concentrations from 1–350 pM (1–350 ug/mL) and added to cells. After incubation under normal culture conditions for additional 72 h, the number of viable cells was determined by CellTiter assay and cell viability was expressed as percentage compared to control (nontreated) cells.</p><!><p>Once the absence of intrinsic toxicity of FND to 293 KDR cells was confirmed, VEGF activity of FND-scVEGF was tested using a competition assay between FND-scVEGFs and a cytotoxic VEGF-toxin fusion protein: SLT (Shiga-like toxin)-VEGF. Upon binding to VEGFR-2, SLT-VEGF initiates receptor-mediated endocytosis and enters the cell, killing VEGFR-2 expressing cells after 24–48 h of exposure to 2 nM of SLT-VEGF.51 The competition assay is based on the ability of scVEGF-based constructs to compete with SLT-VEGF for VEGFR-2. The efficiency of protection for scVEGF-based constructs is characterized by an EC50 value, determined from dose-dependent protection experiments. Values of EC50 for derivatized scVEGF are compared to that of parental scVEGF, to assess how derivatization affected scVEGF binding to VEGFR-2.40 Derivatized and nonderivatized FND were pelleted and resuspended in 2 nM SLT-VEGF. After 20 h of plating cells, serially diluted samples were incubated with the culture, in addition to 2 nM SLT-VEGF. After 96 h, cell viability was determined by CellTiter assay and compared to controls. In this experiment, we consider that only scVEGF detected by ELISA would be accessible to the cellular receptors.</p><!><p>The binding of VEGF activates VEGFR-2 tyrosine kinase activity that leads to autophosphorylation detectable by Western blot analysis. The advantage of this assay as that it is short, sensitive, and specific. Five minutes incubation of 293/KDR cells with VEGF is enough to detect phosphorylated tyrosine residues in VEGFR-2. Moreover, low nanomolar VEGF is sufficient to reliably detect its functional activity. Finally, antibodies to every phosphorylated tyrosine in the cytoplasmic tail of VEGFR-2 are commercially available.</p><p>293/KDR cells were plated into 24-well plates, 7.5 × 104 cells/well, and incubated for 20 h at 37°C in 5% CO2. FND-scVEGFs (or scVEGF Standard) were then serially diluted in complete culture medium and added to cells to final scVEGF concentrations of 0.2, 1, and 5 nM. (Concentrations of scVEGF in FND preparations were calculated based on a prior ELISA assay). After 5 min incubation at 37°C, treated and control (no VEGF) cells were lysed and the lysates were immediately loaded on a 10% polyacrylamide gel. Cell lysates were separated by SDS-PAGE and then analyzed by Western blotting with Ph-VEGFR2(pY1175) antibody specific to phosphorylated tyrosine in position 1175 of the cytoplasmic tail of VEGFR-2.</p><!><p>Validation of FND-scVEGF was also performed using human foreskin fibroblasts CCD-1137Sk (ATCC CRL-2703) and human endothelial colony-forming cells (ECFCs, clone #CB002, Creative Scientist, Inc.). ECFCs were isolated from cord blood and propagated according to previously published protocol.52 ECFCs were cultured in VecStem Media (cat. no. VSM01–250 mL, Creative Scientist, Inc.) with 2% human serum supplement (cat. no VS01–5 mL, Creative Scientist, Inc.). CCD1137 fibroblasts were cultured in DMEM (MilliporeSigma) supplemented with 10% FBS (Corning). ECFCs and CCD1137 naturally express different levels of VEGFRs, with ECFCs displaying significantly more VEGFR2 than CCD1137. For FND association studies, cells were plated at 2000/well density in a 384-well plate. The following day, targeted and control FND were added to the cells at final concentrations of 4, 2, 1, 0.5, and 0.25 pM in triplicate wells and incubated for 3 h at 37°C. The cells were washed with HBSS with Ca2+ and Mg2+ (cat. no. 21–023-CV, Corning) and incubated with 1.5 μM 5-Carboxyfluorescein diacetate (cat. no. C4916, MilliporeSigma) and with 1.5 μg/mL Hoechst 33342 (cat. no. H7399, Thermo Scientific) for 15 min at 37°C in HBSS (10 μL/well). After they were washed with HBSS, live cells were imaged using 20× objective in INCA 2200 high content analyzer (GE Healthcare), with four fields of view taken per well. The images were analyzed using INCA 2200 software, and the percentage of FND-positive cells among 5-CFDA/Hoechst positive cell was identified.</p><!><p>Fourteen six-to-eight-week-old ~20 g athymic BALB/c female mice were sourced from Duke Cancer Center Isolation Facility and maintained in standard housing at 22°C with water. A purified, low-fluorescent diet accessed ad libitum was provided by the Duke Animal Care and Use Program. Animal care and procedures were approved by Duke University's Institutional Animal Care and Use Committee. Mice were fed an alfalfa-free diet (AIN-76A, Research Diets, Inc., New Brunswick, NJ) to minimize background fluorescence and were allowed several weeks to adapt to this diet. 4T1 murine mammary carcinoma cells were maintained in Dulbecco's Modified Eagle's Media (Gibco, Gaithersburg, MI) supplemented with 10% heat-inactivated fetal bovine serum (Gibco, Gaithersburg, MI) and 1% antimycotic/antibiotic solution (Gibco, Gaithersburg, MI) in a 37°C and 5% CO2 environment. Mice were injected into the right flank with 2.5 × 105 4T1 cells in 100 μL of 1X PBS. Prior to injection, cells were washed twice with 1× PBS. The study was initiated when tumors were approximately 300 mm3.</p><p>For tail vein injection, all groups were prepared at 10 mM HEPES pH = 7.4 with 5% dextrose. The administered groups were FND-scVEGF (500 μg/mouse, N = 5), FND-PG (500 μg/mouse, N = 5), and control group (only 5% dextrose in 10 mM HEPES, N = 4) with administration volume of 200 μL. (Note: control has one mouse less due to attrition prior to administration.) Mice were sacrificed at the humane end point (tumor volume is 2000 mm3) or the end of study.</p><!><p>Mice were imaged before injection (t = 0), as well as at several time points times post injection, t = 4, 8, 5, 48, 72 h using IVIS kinetic whole-body imaging system (Caliper LifeSciences, PerkinElmer, Hopkinton, MA), with 605 nm excitation and Cy5.5 emission window. IVIS acquisition and analysis were performed in Living Image (v. 4.3.1, Caliper LifeSciences, PerkinElmer, Hopkinton, MA). During acquisition, mice were anesthetized with 1.5% isoflurane (Henry Schein Animal Health, Dublin, OH) in pure oxygen environment and warmed on a heated stage. Mice were sacrificed after 72 h and tissue was flash frozen. Frozen tumors (stored at −280°C) were cut in half and cryotomed (FSE, Thermo Scientific, Waltham, MA) in 10-μm sections in a −25°C environment. Postsectioning, half of the samples were stained with Hoechst DAPI (bisBenzimide H 33342 trihydrochloride, Sigma-Aldrich, St. Louis, MO). Samples were warmed to room temperature in an uncovered humidity rack, then fixed in ice-cold methanol (Sigma-Aldrich, St. Louis, MO) for 10 min. Postfixation, slides were rehydrated in 1× PBS (Gibco, Gaithersburg, MI) for 5 min. Hoechst solution was applied at a concentration of 1 μg/mL and incubated at room temperature for 10 min. Hoechst-stained slides were stored in a dark, room-temperature environment, and unstained slides were stored in a −80°C freezer.</p><p>The sections were imaged using a Zeiss inverted microscope (Zeiss Axio Observer 1.0, Carl Zeiss, Thornwood, NY) with a Hamamatsu camera (ORCA-Flash 4.0, Hamamatsu Photonics, Bridgewater, NJ). Nanodiamond signal was recorded with a standard Cy5 filter set, and a DAPI filter set was used for Hoechst imaging. Imaging and preliminary analysis was performed in Zen (ZEN pro 2012 (blue edition), Carl Zeiss, Thornwood, NY). In this program, images were normalized via their flat field images and exported at their full dynamic range of 16 bit into.tif files. All subsequent analysis was performed in Fiji (v 1.0).53 A brightfield or DAPI image was opened, and a free-form ROI was traced around the outline of the tumor. This was so that any nanodiamond signal could be normalized to the total tumor area and subsequently combined and compared with other subjects. The Cy5 image was then converted to 8 bit and a threshold was applied to decrease autofluorescent contamination. This threshold was kept constant for each image. The thresholded signal area was recorded and normalized to the total tumor area. Averages across groups were taken, and significance was investigated via an ordinary one-way ANOVA in Prism (v 7.0, GraphPad Software, La Jolla, CA). A second analysis was performed on the tumor images: once Cy5 images were thresholded, the particles were counted. Only particles in a circular shape (0.5–1 circularity, ratio of long-short axis) with a size of 5–100 pixels were recorded. These counted pixels were normalized to the tumor's unit area, and animals within a group were averaged. An ordinary one-way ANOVA was performed for statistical analysis followed by Turkey's test with a p-value of less than 0.05 considered significant.</p><p>In addition, microspectroscopy analysis was performed using a previously described home-built system.38 The system, which allows for simultaneous microscopy and spectroscopy, is designed around an Olympus IX71 inverted fluorescence microscope, with one filter turret modified to accommodate laser input (Optically Pumped Semiconductor Laser, Sapphire, 532 nm, 150 mW adjustable drive, Coherent Inc.) A modular USB spectrometer (HR2000, Ocean Optics) was used for spectra collection, while a CCD color camera allowed for color imaging (AmScope, MT5000-CCD-CK). All filters used were by Semrock unless otherwise indicated. Laser power for analysis was kept to <50 mW throughout, through a 532 nm dichroic mirror (LPD02–532RU) and a 532 nm notch filter (NF01–532U). Imaging via broadband illumination was achieved with 100 mV short arc mercury excitation through the following filter sets (i) blue: FF01–470/28 excitation, BLP01–488R emission, and Q505LP dichroic (Chroma) or (ii) green: FF01–517/20 excitation, BLP02–561R emission, and FF552-Di02 dichroic. Additionally, two-photon imaging was performed with a Zeiss LSM 7MP at the UNC Neuroscience Center Microscopy Core</p>

PubMed Author Manuscript

Testing the N-Terminal Velcro Model of CooA Carbon Monoxide Activation

CooAs are dimeric bacterial CO-sensing transcription factors that activate a series of enzymes responsible for CO oxidation. The crystal structure of Rhodospirillum rubrum (rrCooA) shows that the N-terminal Pro from monomer A of the dimer coordinates the heme of monomer B that locks rrCooA in the \xe2\x80\x9coff\xe2\x80\x9d state. When CO binds, it is postulated that the Pro is replaced with CO, resulting in a very large reorientation of the DNA binding domains required for specific binding to DNA. Crystal structures of the closely related CooA from Carboxydothermus hydrogenoformans (chCooA) are available, and in one of these, the CO-bound on-state indicates that the N-terminal region that is displaced when CO binds provides contacts between the heme and DNA binding domains that hold the DNA binding domain in position for DNA binding. This has been termed the N-terminal velcro model of CooA activation. The study presented here tests this hypothesis by generating a disulfide mutant that covalently locks chCooA in the on-state. A simple fluorescence assay was used to measure DNA binding, and the S\xe2\x80\x93S mutant was found to be in the on-state even without CO. We also determined the high-resolution crystal structure of the apo-heme domain, and the resulting structure is very similar to the holo-heme-bound structure. This result shows that the heme binding motif forms a stable structure without heme or the DNA binding domain.

testing_the_n-terminal_velcro_model_of_cooa_carbon_monoxide_activation

<!>MATERIALS AND METHODS<!>Site-Directed Mutagenesis for Q4C and M177C.<!>Expression and Purification of Wild Type chCOOA (WTchCOOA) and N127L/N128L and Q4C/M177C Mutants.<!>Nonreducing PAGE and Determination of Free Thiol Groups by Ellman\xe2\x80\x99s Reagents.<!>Fluorescence Spectrophotometer Studies for DNA Binding.<!>Cloning and Purification of the Heme Domain.<!>Heme Binding.<!>Crystallization, Data Collection, and Structure Solution of the Heme Domain.<!>Purification and Characterization.<!>Mutant Design.<!>Disulfide Bond Formation.<!>Binding of DNA to chCooA.<!>Purification and Spectrophotometric Titration of the Heme Domain.<!>Heme Domain Structure.<!>CONCLUSIONS

<p>The CooA family of proteins consists of prokaryotic CO-sensing transcription factors that regulate the expression of genes involved in the utilization of CO as an energy source.1 These are homodimeric proteins that contain two hemes. Each monomer has an N-terminal heme binding domain and a C-terminal DNA binding domain. The binding of CO to the heme iron switches CooA to a state that binds to 5′ promoter sequences resulting in the transcription2 of genes that enable CO to be utilized as an energy source.3,4 The overall architecture of CooA closely resembles that of the catabolite activator protein (CAP) family of transcription factors that are regulated by cyclic AMP. The available crystal structures of CAP5–7 are in the so-called on-state with the DNA domain in position to bind DNA. The crystal structure of Rhodospirillum rubrum (rrCooA)8 was the first structure of a CAP-like transcription factor in the inactive off-state. A comparison of this off-state structure with the on-states of CAP shows that the DNA binding domain can undergo a very large reorientation between the on- and off-states. The off-state CooA crystal structure shows that the N-terminal Pro residue of monomer A coordinates the heme of monomer B and vice versa (Figure 1). In the presence of CO, the N-terminal Pro ligand is replaced by CO, resulting in a switch to the on-state. Crystal structures of a second CooA, Carboxydothermus hydrogenoformans CooA (chCooA), are available,9,10 and one of these is in the CO-bound on-state. In this case, the N-terminal region is situated between the DNA and heme domains providing a series of nonbonded bridges between the two domains that has been postulated to hold the DNA domain in the active orientation. We have termed this the N-terminal velcro model of CooA activation (Figure 1).</p><p>Here we provide experiments designed to test this model. On the basis of the chCooA crystal structures, a mutant has been designed that should form an intramolecular S–S bond between the DNA and heme domains, thus locking chCooA in the on-state. In addition, a novel fluorescence assay has been developed that provides a simple and rapid method for differentiating between the on- and off-state conformations.</p><!><p>FPLC-purified oligonucleotides were obtained from IDT. Restriction enzymes, T4 DNA ligase, and polymerase were purchased from New England Biolabs (Beverly, MA). Site-directed mutagenesis kits were purchased from Agilent Technologies. Ni-NTA agarose superflow resin and the miniprep plasmid purification kit were from Qiagen (Chatsworth, CA). Dithiothreitol (DTT), 5,5-dithiobis(2-nitrobenzoic acid) (DTNB), imidazole, hemin, and all other chemicals were purchased from Sigma-Aldrich and were of the highest quality. A Cary spectrophotometer (Agilent Technologies) was used for spectroscopic studies. A Hitachi F4500 fluorescence spectrophotometer was used for fluorescence studies.</p><!><p>On the basis of the crystal structure (Protein Data Bank entry 2HKX),10 we designed a mutation to lock chCooA in the on-state. Amino acids Gln4 and Met177 were mutated to Cys by using standard site-directed mutagenesis and polymerase chain reaction. The Q4C/M177C double mutant is designated DMchCooA.</p><!><p>Wild type chCooA, double leucine mutants (N127L/S128L abbreviated as LLchCooA), and cysteine mutant proteins were purified by the same procedure reported previously by Clark et al.,11 with a few modifications. Briefly, protein was overexpressed in Escherichia coli BL21(DE3) cells by induction with 1 mM IPTG at an OD of 0.6 and grown for a further 24 h at 25 °C; 0.4 μM δ-aminolevulinic acid was also added during induction for better heme incorporation. The IPTG-induced cells were harvested, resuspended in 50 mM Tris-HCl (pH 7.8), 500 mM NaCl, and 10 mM histidine (buffer A) and lysed by sonication. The crude lysate was centrifuged at 17000 rpm for 60 min. The supernatant was applied to a Ni2+-NTA column pre-equilibrated with buffer A. Protein was eluted using the same buffer supplemented with 80 mM histidine. Fractions containing protein were pooled and concentrated by using a 10 kDa centricon. Concentrated protein was further purified and buffer exchanged by gel filtration using a Superdex 75 16/60 column equilibrated with 50 mM Tris (7.8) and 500 mM NaCl (buffer B). Protein fractions with a higher heme content (R/Z) were pooled and concentrated by using a 10 kDa cutoff centricon (Amicon). Protein concentrations were determined by using the extinction coefficient reported previously.12 The purity of the protein was confirmed using 12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE).</p><p>Fe(II) and Fe(II)-CO samples were prepared anaerobically. The protein stocks were reduced with 50 mM dithionite in a nitrogen-filled anaerobic chamber, and extra dithionite was removed with a desalting column. For Fe(II)-CO, the Fe(II) CooA is exposed to CO in an anaerobic cuvette with a 1 cm path length.</p><!><p>Disulfide bond formation in the double-cysteine mutant was visualized by using nonreducing SDS–PAGE. Nonreducing gel electrophoresis was conducted using 12% polyacrylamide gels, and samples were loaded by mixing with dye without any reducing agent. Ellman's reagent13 was used to quantitate free thiols; 500 μM Ellman's reagent [5,5-dithiobis(2-nitrobenzoic acid)] was incubated with 5 μM WTchCooA or double-cysteine mutants in 100 mM Tris (7.8), 300 mM NaCl, and 6 M guanidinium chloride for 15 min at room temperature. The absorbance is recorded by using a Cary 3E spectrophotometer at 412 nm, and the concentration of free thiol is calculated with a standard curve using cysteine. To calculate free thiol in proteins reduced by DTT, we used a desalting column (Bio-Rad) to remove excess DTT before recording spectra.</p><!><p>All fluorescence studies utilizing tyrosine emission were performed by using the synthetic oligonucleotide reported previously.14 The sequence of half-sites contains a nucleotide derived from binding sites of one monomer of chCooA. Equimolar concentrations of the two strands were mixed to a final concentration of 1 μM in 50 mM Tris (7.8) and 100 mM NaCl. The strands were annealed by placing a solution at 90 °C and controlled cooling to 4 °C at a speed to 1 °C/min: ssDNA 1, 5′ GCGAAAAGTGTCAC; ssDNA 2, 5′ ATATGTGAC-ACTTTTC.</p><p>chCooA binds to DNA when the heme iron is reduced and bound to CO. To qualitatively analyze DNA binding, we used fluorescence spectroscopy to demonstrate binding of DNA to different states of native and DMchCooA at 25 °C on a Hitachi F4500 fluorescence spectrophotometer. chCooA does not have any tryptophan residue but does contain seven tyrosine residues that were used to monitor DNA binding. Proteins, oligonucleotide, and buffer were degassed and purged with argon to obtain the fluorescence data. Protein (3 μM) was excited at 272 nm, and emission spectra were recorded between 300 and 350 nm to monitor changes in fluorescence upon DNA binding in 50 mM Tris (pH 7.8), 300 mM NaCl, and 5 mM EDTA. Corrections were made by subtracting the spectrum of the appropriate buffer. The pUC19 plasmid is used as a control.</p><!><p>The heme domain (1–137) of chCooA was amplified via polymerase chain reaction from a plasmid containing chCooA and cloned between NcoI and XhoI with a C-terminal His tag in pET28a. Because the level of binding of heme to the heme domain is very low compared to that of the full length protein, 0.4 mM δ-aminolevulinic acid was also used during induction for better heme incorporation. The heme domain was purified by the same method used for purification of wild type and mutant full length protein except a low salt concentration (100 mM) was used during the final step of purification.</p><!><p>Heme binding was monitored by difference spectroscopy in the Soret region of the ultraviolet–visible spectrum at 25 °C. Successive aliquots of freshly prepared 1 mM hemin in 0.1 N NaOH were added to 7.5 μM heme domain and the reference cuvette. Difference spectra were recorded 5 min after the addition of each heme aliquot.</p><!><p>The heme domain was crystallized by the hanging drop method, and the plates were incubated at 22 °C. Apo-heme domain crystals were obtained by mixing 2 μL of 400 μM protein and 1 μL of a reservoir solution containing 1.6–2.0 M ammonium sulfate and 100 mM Tris (8.5). Glycerol (30%) was used as a cryo-protectant, and a set of data were collected from a single crystal at SSRL beamline 7–1. Diffraction images were indexed, integrated, and scaled using Mosflm15 and Aimless.16</p><p>The heme domain (2–143) of rrCooA17 was used as a search model to determine the structure by molecular replacement calculations using Phaser18 implemented in the CCP4 package.19 The transformed model was subjected to refinement using Phenix.20 Phases were improved and extended incrementally to 1.15 Å. The 2Fo – Fc and Fo – Fc electron density maps were visualized, and model building was performed using COOT.21 The crystallographic Rfactor and Rfree were monitored at each stage to avoid any bias. Water molecules were added by the automatic water-picking algorithm of COOT and inspected manually. The refined structure of the heme domain contains two chains and 301 water molecules. Final refinement statistics are listed in Table 1.</p><!><p>The double-Cys mutant, DMchCooA, was overexpressed as a heme protein in the cytosolic fraction like WTchCooA and LLchCooA. Purified protein migrated as a single 26 kDa band via SDS–PAGE. Because imidazole acts as a proximal ligand to heme, we used histidine to elute proteins from the Ni-NTA column. The electronic absorption spectra of chCooA show Soret, α, and β bands at 421, 565, and 538 nm, respectively.</p><!><p>The structure of chCooA we reported10 was that of a mutant called LLchCooA. Using an in vivo selection process, it was found that the conversion of Asn127 and Ser128 to Leu to give LLchCooA can enhance transcription in the absence of CO and hence must be locked in the on-state. As shown in Figure 2, Leu127 on helix F forms nonbonded contacts with Ala2 and Leu7. Comparing the rrCooA off-state and LLchCooA on-state structures shows that the DNA domain can reorient up to 180° (Figure 1). As shown in Figure 2, the DNA binding helix is oriented "down" in the on-state and thus is available for DNA binding, but in the off-state, this helix is oriented "up". We have postulated that the reason the LLchCooA mutant adopts primarily the on-state conformation is that the mutant Leu127 side chain is exposed to solvent in the off-state but buried in the on-state where it can form favorable nonbonded contacts with Leu7 and Ala2. The N-terminal segment thus provides a series of nonbonded contacts between the DNA and heme domains that effectively holds the DNA domain in position for proper DNA binding. By displacing the N-terminal heme ligand, CO frees the N-terminal segment to form the link holding the DNA domain in the active orientation. The LLchCooA mutant shifts the equilibrium to the on-state even in the absence of CO because of the favorable burial of the mutant Leu127 side chain at the DNA–heme domain interface. To further test this view of CooA activation, we designed a mutant that will form an S–S bond between the heme and DNA domains, thus permanently locking chCooA in the on-state. Visual inspection of the LLchCooA structures indicates that mutating Met177 in the DNA domain and Gln4 in the heme domain should form an S–S bond when protein is in the on-state.</p><!><p>To examine disulfide bond formation in the double mutant (DMchCooA), the purified protein was run on an SDS gel with or without a reducing agent (Figure 3). An intramolcular S–S bond generates a more compact structure that can potentially migrate toward lower molecular weights in SDS gels. WTchCooA runs as a single band with or without a reducing agent, while DMchCooA protein produces two bands under oxidizing conditions but only one band under reducing conditions. Treatment of DMchCooA with diamide, an oxidizing agent used to induce disulfide bond formation, increased the amount of the faster-migrating band. These results indicate the band migrating toward a lower molecular weight forms the intramolecular S–S bond.</p><p>The total numbers of free thiol groups in chCooA estimated from DTNB analysis under various treatments are listed in Table 2. These results show that there was only one free thiol group per subunit of purified WTchCooA, while for DMchCooA, the number is greater than one. On the other hand, after treatment of chCooA and DMchCooA with DTT along with 6 M guanidinium chloride, the total thiol count is near three. Taken together, the SDS gels and thiol titration analysis indicate that a large fraction of DMchCooA forms an intramolecular S–S bond.</p><!><p>In seeking a relatively simple method for estimating DNA binding, we found that DNA induces strong quenching of the fluorescence of chCooA only when chCooA is in the CO complex. Because there is no tryptophan present in chCooA, we used a wavelength of 278 nm to excite tyrosine and recorded emission spectra over a range from 290 to 340 nm. As shown in Figure 4, WTchCooA binds to DNA only when the heme is reduced and bound to CO while Fe(II) DMchCooA in the absence of CO showed DNA binding.</p><!><p>The purified heme domain behaves like wild type protein and Fe(II), formed by reducing the Fe(III) form with sodium dithionite, and showed Soret, α, and β bands at 424, 559, and 529 nm, respectively. Treatement of the reduced Fe(II) heme domain with CO revealed the spectral properties with Soret, α, and β bands at 421, 569, and 538 nm, respectively. Compared to the case for WTchCooA, when we purified the heme domain from E. coli it elutes with a very small amount of heme bound. We determined the stoichiometry of binding of heme to the heme domain to determine if the DNA binding domain plays a role in heme binding. Spectrophotometric titration of the heme domain of chCooA with a solution of hemin showed saturation at two hemes bound per dimer (Figure 5).</p><!><p>We attempted to determine the structure of the chCooA heme domain dimer to determine if the N-terminus of molecule A coordinates the heme in molecule B, which is the case in rrCooA. Unfortunately, we were able to crystallize only the heme domain in the apo form without heme, although the heme domain contained a full complement of heme prior to setting up the crystallization drops (Figure 6). Even so, this ultra-high-resolution structure enables us to determine the influence of heme binding on the heme domain structure. A superposition of the heme domain monomer on the holo-chCooA monomer gives a root-mean-square deviation for Cα atoms of 1.6 Å, and most of the deviation is due to differences in surface loop regions. Thus, we can conclude that heme binding or the presence of the DNA binding domain has little influence on the structure of the heme domain. The one significant change relevant to heme binding is the position of the heme ligand, His82, in apo-chCooA. In one subunit of the dimer, His82 is oriented out toward the solvent away from the heme binding pocket. However, in the other monomer, His82 is oriented in and is in position to coordinate heme. Interestingly, the holo-chCooA structure has only one heme bound,10 and in the monomer without heme, His82 also is oriented out toward the solvent.</p><!><p>The available CooA structures and the various structures of CAP provide a reasonably clear picture of the large motions involved in transcription factor activation. The DNA binding domains can reorient by 180° to place the DNA binding helices in the proper position for specific DNA interactions. The two domains are structurally independent given that the isolated heme domain crystal structures with or without heme are the same as in holo-CooA. CooA has provided some advantages over CAP in that a clearer picture has emerged on how the effector, in this case CO, triggers the required structural changes. The N-terminus of molecule A coordinates heme B and vise versa, thus locking the dimer in the off-state. The displacement of the N-terminus by CO frees the N-terminal peptide to provide a bridge between the heme and DNA binding domains. Hence, the N-terminal peptide serves as the "velcro" to hold the DNA domain in the orientation required for DNA binding. These changes also alter the heme binding pocket to favor binding of CO over that of other diatomic ligands, most notably O2. Binding of oxygen to ferrous heme proteins is generally considered to favor ferric-superoxide, Fe(III)-OO−, so H-bonding partners will help to stabilize the oxy complex. In CooA, the bound CO is surrounded by a Val and symmetry-related Leu residues along the dimer interface, thus creating a nonpolar pocket with no potential H-bonding partners.10 This pocket is quite tight and favors a linear Fe–C–O angle rather than the bent angle preferred by O2. The hydrophobic nature of the CO binding pocket and displacement of the Pro ligand are consistent with resonance Raman coupled with mutagenesis studies.22,23 CooA thus provides a fascinating example of how diatomic sensing is coupled to large structural rearrangements of protein domains in addition to providing an example of how heme proteins can control ligand selectivity.</p>

PubMed Author Manuscript

The Proof-of-the Concept of Biochar Floating Cover Influence on Swine Manure pH: Implications for Mitigation of Gaseous Emissions From Area Sources

Mitigation of potentially hazardous and malodor compounds emitted from animal waste is needed to improve the sustainability of livestock agriculture. Bacteria control the generation of these compounds and also depend on the pH of manure. Influencing swine manure pH, especially on the liquid-air interface, may lead to a reduction of emission of odorous and hazardous compounds. The objective of this experiment was to test highly alkaline and porous (HAP) modified biochar with pH = 9.2 and red oak (RO) biochar with pH = 7.5 influence on swine manure pH acquired from the outdoor storage and deep pit storage under a barn. HAP and RO biochars were topically applied on the outdoor-stored (pH = 7.55), and pit (pH = 8.00) manures and spatial pH (every 1 mm of depth) were measured on days 0, 2, and 4. Results showed that HAP biochar increased outdoor-stored manure pH on day 4, particularly within the top 10 mm of depth, where pH ranged from 7.79 to 8.90, while in the case of RO pH ranged between 7.46 and 7.66, i.e., similar to control (7.57–7.64). Both biochars decreased pit-stored manure pH within the top 10 mm of depth (in comparison with the control pH of 8.36–8.47) to 8.19–8.30 (HAP), and 8.18–8.29 (RO) on day 4. However, differences were not considerable. The reason for the insignificant effect of biochars on pit manure was likely due to its higher buffer capacity in comparison with the outdoor-stored manure.

the_proof-of-the_concept_of_biochar_floating_cover_influence_on_swine_manure_ph:_implications_for_mi

Introduction<!><!>Biochar<!>Manure<!>The Determination of Biochar Type Influence on Spatial and Temporal pH Manure<!><!>The Determination of Biochar Type Influence on Spatial and Temporal pH Manure<!><!>The Determination of Biochar Type Influence on Spatial and Temporal pH Manure<!>Results<!><!>Results<!><!>Results<!>Discussion<!><!>Discussion<!><!>Discussion<!>Conclusion<!>Data Availability Statement<!>Author Contributions<!>Conflict of Interest

<p>The increase in livestock production leads to an increase in the volume of manure storage and challenges to its utilization. Manure storage in open lagoons and outdoor storages can be a source of malodor and elevated concentrations of gases such as ammonia (NH3) (Grant and Boehm, 2018; Reyes et al., 2019), hydrogen sulfide (H2S) (Rumsey and Aneja, 2014), and greenhouse gases including methane (CH4), carbon dioxide (CO2), and nitrous oxide (N2O) (Hitaj et al., 2019). Moreover, volatile organic compounds (VOCs) (including sulfur-containing compounds, fatty acids, and phenolics) (Trabue et al., 2019) are also responsible for malodor from stored manure.</p><p>Zahn et al. (2001) reported the emission rate from the deep-pit barn of 2.38 g NH3 m−2 h−1 for ammonia, while hydrogen sulfide had an emission rate of 13.3 mg H2S m−2 h−1. According to the Environmental Protection Agency (EPA), 15% of greenhouse emissions are associated with manure management in the agricultural sector in the United States (The United States Environmental Protection Agency, 2020).</p><p>Solving the environmental problems related to livestock is a challenge for farmers, public, and regulatory agencies. A comprehensive solution that includes not only effectiveness but also practicality and low-cost is in high demand. Iowa State University Extension and Outreach Air Management Practices Assessment Tool summarized 12 different methods of mitigation of gaseous emissions from livestock and manure storage. Application of manure additives can be a practical option in terms of application, logistics, and cost (Iowa State University Extension Outreach, 2020).</p><p>One of the types of manure additives is biochar, which is a solid carbonaceous by-product (char) obtained from pyrolysis, gasification, or torrefaction. It is a carbon-rich, porous, black material. The abundant sources of biochar can be sludge, food waste, agricultural and forestry residues, municipal and animal waste (Białowiec et al., 2018; Dudek et al., 2019; Pulka et al., 2019; Stepien et al., 2019; Swiechowski et al., 2019; Syguła et al., 2019; Wang et al., 2019). A recent review of biochar utilization in crop and livestock agriculture was presented elsewhere (Kalus et al., 2019). Characterizations of biochar such as surface area, porosity, hydrophobicity, pH, cation exchange capacity, and functional groups depend on feedstock and the temperature of treatment (Amin et al., 2016). Maurer et al. (2017) studied the effect of topically applied biochar that floated on swine manure for a month. Observation showed 12.7–22.6% reduction of NH3 emission, 12–30% for H2S, and 8.7–26% for indole. However, due to the complexity of the biochar, the mechanism of emission reduction still needs more investigation.</p><p>According to Zhu (2000), most of the malodor producing bacteria and H2S have pH in the range of 6.5–7.5. Raising the pH of manure by adding high alkaline biochar may cause a decrease in gaseous emissions, especially in VFAs that are considered as major malodor contributors. Mroz et al. (2000) state that decreasing manure pH may help to inhibit ammonium transformation into its volatile form of ammonia. On the other hand, a more basic pH near the manure-air interface can be helpful with mitigating H2S emissions from shallow pit-stored manure (Wi et al., 2019). It is also generally assumed that pH regulates microbial activity in manure, which in turn, is driving odorous compounds generation/utilization.</p><p>Most recently, we have explored the feasibility of using biochars properties to control the pH near the water-air interface in an idealized system where thin layers of biochars were applied to the clean water surface (Meiirkhanuly, 2019; Meiirkhanuly et al., 2019). The results proved that the surficial application of biochar to water was able to change both the pH near the water-air interface and the pH of the solution with time. The pH change was dependent on the biochar pH and water buffer capacity. The biochars had a different floatability as well. These results in Meiirkhanuly et al. (2019) warrant further research into the next logical step, i.e., testing the floatability of biochars on the surface of manures with different pH and other physicochemical properties. Furthermore, the impact of biochars on the manure-air interface pH can be further explored as a factor influencing gaseous emissions of odorous compounds that are sensitive to pH. Results could be used for further development of this technology to mitigate odorous emissions from other area sources such as wastewater treatment basins, landfills, lakes with nutrient imbalance.</p><p>The effect of topically applied biochar on spatial and temporal pH change of swine manure has never been studied. Taking into account that the pH of manure is crucial for generating emissions from manure, the study aims show if topically applied biochar can be used as a treatment in order to mitigate emission from swine manure in further studies. The research on manure pH modification due to biochar application could bring new knowledge for a better understanding of the mechanism of odor emission mitigation from manure by biochar floating covers, previously proven (Maurer et al., 2017).</p><p>We hypothesize that the application of biochar to manure as floating cover will increase the pH value; however, the degree of the influence depends on the depth, biochar alkalinity, initial manure pH, and manure buffering capacity.</p><p>The objective of this study is to test how highly alkaline and porous biochar (HAP) with pH = 9.20 and red oak biochar (RO) with pH = 7.50 (Table 1), applied on top of pit (pH = 8.00) and outdoor-stored (pH = 7.55) manure, can influence on spatial (every 1 mm) and temporal pH of manure, and by that, change their NH3 and H2S dissociation (Figure 1).</p><!><p>Properties of the pit and outdoor-stored manure used in the experiment.</p><p>The ideation of the experiment on testing the spatial and temporal effects of topically applied biochar layer on manure pH.</p><!><p>Description of methods of how properties of biochar were acquired is presented elsewhere (Meiirkhanuly, 2019). The detailed summary of the physicochemical properties of the two biochars HAP and RO was recently shown in Meiirkhanuly et al. (2019). Two different biomass residue corn stover and red oak wood biomass were used to prepare the HAP and RO, respectively. The particle size is <1 mm for both biochars. Briefly, the key differences were the pH and ash content of 9.2 vs. 7.5 and ~47 vs. ~16%, for HAP vs. RO biochars, respectively.</p><!><p>Outdoor-stored manure was acquired from Crawford farm in North Central Iowa, and pit-stored manure was collected from Iowa Select Farms in Mid-West Iowa. Properties of manures that were used for the experiment are given in Table 1.</p><!><p>Three of the food storage glass containers with a volume 1,700 mL (19 × 14.5 × 7.5 cm) were filled with 800 mL of pit manure each, and another three containers were filled with 800 mL of outdoor-stored manure. 6.35 mm thick layer of HAP and RO with 48 g and 58 g of mass, respectively, were applied on day 0, and pH measurement data were collected on days 0, 2, and 4. The matrix of the experiment is represented in Table 2.</p><!><p>The matrix of the experiment.</p><!><p>Thin microelectrode (MI-415 Series Micro-Combination pH Probe, Microelectrodes, Inc., 2020), attached to a laboratory stand, was connected to an Accumet AB 15 pH meter (Fisher Scientific, 2020). A manual lab jack with a container of manure on the top of it was placed under the microelectrode (Figure 2). When the microelectrode penetrated the manure surface, pH measurements for every 1 mm of depth were collected by elevating the lab jack and using a ruler placed next to it.</p><!><p>Experimental design for testing biochar influence on spatial and temporal pH distribution in swine manure.</p><!><p>Buffer capacity of manure was determined by using the titration method. 1 M solution of acetic acid was made by adding 5.742 mL of stock solution to 100 mL of deionized water. After adding a drop of the solution, manure was stirred on a magnetic stirrer for 10 s and the pH of the manure was measured. Following equation was used to estimate the buffer capacity of manure:</p><p>Where the slope is fitted slope of the linear regression line for manure (Costello and Sullivan, 2014).</p><!><p>The biochars used in this study differ in characteristics as exhibited by the proximate analysis. The ash content of HAP is relatively high in comparison to RO, and SiO2 is the main component of the ash content in both biochars. After biochar application on day 0, both HAP and RO were floating on the surface of outdoor-stored manure. Biochars stayed on the top of the manure until day 4, and only a small fraction of biochars sank to the bottom. The HAP biochar had more suspended particles than RO (Figure 3). After biochar application on day 0, both HAP and RO were floating on the top of the pit manure. On day 2, the bottom of the HAP biochar layer crusted, and separation between the biochar layer and manure level occurred while RO biochar was floating on the top of the manure, and only a small fraction of it was suspended. On day 4, the separation between HAP biochar layer and manure became larger, and suspended particles of RO biochar settled on the bottom (Figure 4).</p><!><p>Photos of RO (left) and HAP (right) treated outdoor-stored manure on days 0, 2, and 4. Frames show biochar behavior (red—floating, green—suspended, blue—settled biochar).</p><p>Photos of RO (left) and HAP (right) treated pit manure on days 0, 2, and 4. Frames show biochar behavior (red—floating, green—suspended, blue—settled biochar).</p><!><p>The summary of spatial and temporal change pH in the outdoor-stored and pit-stored manure due to the influence of surficially applied biochar is illustrated in Figure 5 and Supplementary Figure 1. On day 0, the pH range of outdoor-stored manure treated with RO (pH = 7.5) biochar was 7.42–7.37 (p = 0.960) from the surface to bottom whereas control manure had a pH range of 7.52–7.35. On day 2, control manure had pH 7.71–7.59, while RO changed the range of pH to 7.52–7.39 (p < 0.0001) from the surface to the bottom of manure. On day 4, control manure had a pH range of 7.64–7.57 and was similar to the pH of RO treated manure (7.64–7.39) (p = 0.033) from the surface of manure to the bottom of the container.</p><!><p>Spatial and temporal change pH in the outdoor-stored and pit-stored manure due to the influence of surficially applied biochar.</p><!><p>HAP biochar had consistently shifted the pH in the top 10 mm over the course of the experiment, especially for the outdoor-stored manure. On day 0 (40 min after application), HAP biochar increased outdoor-stored manure pH and ranged from 8.42 to 7.60 to (p < 0.0001), surface to the bottom, respectively. However, the greatest change in the pH due to HAP was in the top 10 mm (from 8.42 to 7.60), then it remained nearly the same to the bottom. Similarly, on day 2, the greatest change in the pH due to HAP was in the top 10 mm (from 8.86 to 7.92), with the maximum at ~6 mm depth. On day 4, manure surface pH was 8.9 and dropped to 7.79 at the 10 mm of depth and gradually changed to pH 7.66 at the bottom.</p><p>pH values of pit manure gradually dropped from the surface of manure to the bottom for all treatments. On day 0, the pH of the control manure on the surface was 8.38 and decreased to 7.90 on the bottom. HAP biochar changed manure pH from 8.56 at the surface to 7.91 on the bottom, with a sharp drop in the first 4 mm of depth. Then, below ~20 mm, it was within the range of the control. RO biochar changed the manure pH from 8.27 on the surface of manure, then started dropping up to 8.15 (p = 0.0.026) on 20 mm of depth.</p><p>On day 2, the pH of the control manure on the surface was 8.13 and decreased to 7.93 on the bottom. HAP biochar changed manure pH from 8.30 at the surface to 7.96 on the bottom, with a sharp drop in the first 5 mm of depth. The pH in the RO-treated manure was similar to control. On day 4, the control pH gradually decreased from the 8.47 at the surface to 7.99 at the bottom. The greatest pH changes for biochar-treated manure were observed in the top 10 mm of depth and again the nearest 10 mm to the bottom, the latter likely due to the biochar settling (Figure 4).</p><!><p>Outdoor-stored manure had the most apparent change in pH (~1) due to HAP biochar on day 0 (Figure 5) that continued to be even more noticeable (~1.5) on day 2 and 4 in the top 10 mm of depth (Supplemental Figure 1). On day 2, both biochars showed an apparent change of outdoor-stored manure where pH changed to 9.07 to 7.65 and 7.52–7.39 for HAP and RO, respectively, in comparison with pH of control manure pH 7.71–7.59. On day 4, the change of pH for HAP was still apparent in the top 10 mm of depth, where the drop was from 8.9 at the surface to 7.79 at 10 mm.</p><p>The changes in pH due to HAP treatment of pit-stored manure were not as noticeable as with the outdoor-stored manure. There was still an apparent change in pH in the top 5 to 10 mm near the surface (Supplemental Figure 1). Interestingly, the sharp drop in pH on day 0 at ~5 mm gradually moved to ~10 mm depth with biochar gradual settling on days 2 and 4 (Figures 3, 4).</p><p>The pH changes due to RO treatment were less noticeable over time compared with those associated with the HAP treatment (Supplemental Figure 1). Outdoor-stored manure pH had noticeable change due to both biochars, especially for HAP. The pit manure had a smaller pH changes that were limited to the top ~5 mm on day 0 and 2 (Figure 5) for HAP and RO. Then, on day 4, the HAP treatment caused an ~0.8 shift in the pH in the top ~6 mm. Meiirkhanuly et al. (2019) shown that deionized water (with lower buffer capacity compared with tap water), had an immediate change in pH due to HAP and RO biochars. Similarly, to the controlled experiment with water, the reason for an apparent change in outdoor-stored manure pH in comparison with pit manure was due lower buffer capacity of outdoor-stored manure than pit manure (Figure 6).</p><!><p>Buffer capacity of outdoor-stored and pit-stored swine manure estimated by the amount of acidity needed to drop the pH.</p><!><p>Contrary to the experiment with water described in Meiirkhanuly et al. (2019), HAP biochar was floating much longer on top of outdoor-stored manure from day 0 to day 4. In the case of pit manure, the layer of HAP biochar bridged, and separation between the biochar layer and manure level occurred (Figure 4). RO biochar was floating on top of both outdoor-stored and pit manure (Figure 3).</p><p>It is also important to hypothesize the mechanism of biochar interaction with NH3, a key air pollutant and mild odorant that is commonly considered for mitigation of gaseous emissions from area sources in animal agriculture. Figure 7 presents a model involving several chemical reactions that affect the pH at the manure-air interface. The source of ammonia nitrogen is the ammonification of N-organic compounds occurring in the deeper (anaerobic) zone of the manure storage pit. A fraction of the ammonia is adsorbed to biochar, a fraction is still dissolved in the manure, and some NH3 diffuses through the biofilm covering the surface of the biochar. At the same time, the O2 diffuses from the air-biochar-manure interface. It likely creates a microscale aerobic-anoxic-anaerobic gradient with decreasing redox potential. In the aerobic zone, O2 and ammonia are used for nitrification facilitated by nitrifiers present in the biofilm. Ammonia is then oxidized to nitrates. Next, products of ammonification and organic compounds degradation low weight fatty acids (acetate) are used for the denitrification (as a source of C). Denitrification in the anoxic zone leads to the transformation of nitrate to N2, which is then released to the air.</p><!><p>A preliminary model of the impact of biochar on the manure-air interface on the mitigation of NH3 emissions.</p><!><p>Biochar plays an important buffering role. Biochar releases the hydroxides, which neutralize H+ produced during nitrification. Biochar allows creating the expanded biofilm which is colonized by nitrifying and denitrifying bacteria responsible for the transformation of N compounds. Biochar facilitates creating the microenvironment where different processes of ammonification, nitrification, and denitrification occur. The next function of biochar is the adsorption activity, which bonds the load of ammonia nitrogen. Biochar also plays a role as a buffering agent of pH, which is important for maintaining suitable conditions for the growth of microorganisms. This mechanism still requires further investigation.</p><p>In this paper, we focused on the influence of two types of biochar on pH changes in manure, as a proof-of-the-concept of biochar application as a floating cover for odors emission mitigation. Our research showed that depending on biochar origin and properties, the expected effect may differ. Therefore, the next step in developing solutions should be more advanced research on the adsorption of odors by the biochar-manure system, specifically on the mechanisms potentially mitigating nuisance emissions. Also, the measurements of the concentration of pollutants adsorbed in biochar, for determining the mass balance of the pollutants' fate should be done. An important aspect should also be to observe the biological transformation of odors, especially ammonia and influence of OH−, and H+ transport on nitrification/denitrification (as we proposed the hypothesis by application of abduction reasoning—Figure 7). The next question to be addressed from the practical point of view is to determine the changes of the biochar density and floatability with time after application to manure.</p><p>The present research was a proof-of-the-concept, just an initial study for confirmation of the initial hypotheses; however, the mechanism of how biochar works as a buffering agent of pH still requires further discussion and explanations. The execution of Boehm's titration (Schönherr et al., 2018), FTIR, and XPS to properly evaluate the relationship between pH and biochar properties are crucial. Furthermore, the examination of the ash content influence by the determination of ions content in biochar and leaching should be done.</p><!><p>Biochars applied to manure changed the spatial and temporal distribution of pH. Results showed that HAP biochar increased outdoor-stored manure pH, particularly within the top 10 mm of depth, where pH ranged from 8.90 to 7.79 on day 4. Both biochars increased pit manure pH in comparison with control on day 4. However, differences were not considerable. The reason for the insignificant effect of biochars on pit manure was likely due to its higher buffer capacity in comparison with the outdoor-stored manure.</p><!><p>All datasets generated for this study are included in the article/Supplementary Material.</p><!><p>AB, JK, and ZM designed the experiment. ZM performed the experiment. ZM, JK, AB, and CB analyzed the data and wrote the paper. JK contributed reagents, materials, and analysis tools. ZM, JK, and RB acquired funding. All authors contributed to the article and approved the submitted version.</p><!><p>The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.</p>

PubMed Open Access

The effects of long-term whole-body vibration and aerobic exercise on body composition and bone mineral density in obese middle-aged women

[Purpose]The purpose of this study was to determine the effectiveness of whole-body passive vibration exercise and its differences from aerobic exercise on body composition, bone mineral density (BMD) and bone mineral content (BMC).[Methods]Obese middle-aged women (n=33 out of 45) with 34±3% body fat completed the training protocol. They were randomly assigned into diet (n=9; control group), diet plus whole-body vibration exercise (n=13; vibration group), and diet plus aerobic exercise (n=11; aerobic group) groups and we compared their body composition, BMD, and BMC before and after 9 months of training. There were no significant differences in nutrient intake among groups during the training period.[Results]Relative body fat (%) decreased significantly (p < .05) in all three groups and the exercise groups showed a greater reduction in fat mass than the diet only group. BMD in the whole body, lumbar spine, hip and forearm were not significantly different among the three groups. Total body BMC increased significantly in the vibration group throughout the first 6 months of training.[Conclusion]Results suggest that long- term vibration training when used in conjunction with a diet program is as effective as aerobic exercise with a diet program in improving body composition of obese middle-aged women without compromising BMC or BMD. Thus, it can be considered a novel and effective method for reducing body fat.

the_effects_of_long-term_whole-body_vibration_and_aerobic_exercise_on_body_composition_and_bone_mine

INTRODUCTION<!>Participants<!><!>Diet group<!>Vibration group<!>Vibration training positions<!>Aerobic group<!>Measurements<!>Statistical analysis<!>Nutrition variables<!><!>Body composition and anthropometric variables<!><!>Bone measurements<!><!>DISCUSSION<!>CONCLUSION<!>

<p>Obesity is a severe and growing social-health problem that causes devastating disease and orthopedic complications1-4. The success rate of therapy for obesity is very low. Conventional methods often used to achieve weight loss include nutrition, exercise, and/or behavioral correction interventions5-9. Dieting may be successful in the short term; however, it can cause a reduction of lean body mass and disorders of muscle function. Moreover, it is difficult to maintain many dietary regimens and thus the ensuing weight reduction is limited in the long term 8, 10, 11. Therefore, exercise should be undertaken along with a reduction in caloric intake to prevent decreases in body lean mass, basal metabolism, bone mineral density (BMD), and bone mineral content (BMC), which can result from diet-only therapy12-14.</p><p>The positive effects of exercise on health and weight regulation are well-known7, 9, 10, 15, 16. Specifically, regular long-term aerobic exercise prevents or reduces the progression of diseases in coronary and peripheral arteries by reducing circulating cholesterol, hypertension, blood glucose concentration, and obesity, and enhances the function of heart and blood vessels15, 17-20. The American College of Sports Medicine recommends an energy expenditure of 1,500-2,800 kcal/week by performing aerobic activities at 40-75% of maximum heart rate for 45-60 min, 5~7/week to assist in combating the obesity problem21.</p><p>The BMC and BMD of obese individuals are significantly higher than normal-weight individuals due to their body weight load22-25. Therefore, dynamic weight-bearing exercises are recommended to improve BMC and BMD26. In addition, resistance exercises using body weight or an external load have been shown to ameliorate the age-related decrease in BMD and thus reduce the chance of bone-breakage which worsens the quality of life in older individuals27. Karlsson28 reported that exercise increased the generation of bone and reduced bone loss by increasing the mineral or calcium content in bone, eventually increasing BMD28. Others showed an absence of bone loss with exercise29. Therefore, the ideal way to preserve lean body mass, reduce body fat, and maintain bone health is by a reduction in caloric intake combined with long-term exercise training. In contrast, negative effects of exercise such as damage to ligaments and hip and knee joints, were reported to outweigh the positive results of increased physical strength and reduced body fat30, 31. To prevent injuries due to excessive weight or low muscular strength, exercises with relatively low impact on joints, such as swimming, cycling, or aquarobics, can be effective. However, these activities are difficult to practice regularly because special equipment or space is required. Moreover, these are not as effective in maintaining and enhancing BMD, especially for postmenopausal women with low BMD. Whole body vibration has been shown to safely and effectively improve the skeletal and musclular system, as well as blood circulation and the endocrine system, but the results are still controversial32-37.</p><p>The whole body vibration instrument was developed by Flieger et al.33 in Germany for the purpose of improving quality of life by increasing BMD and balance in the neural system of older people33. It has been widely used at institutes for obesity therapy in Germany, Italy and Korea since 2000. Subsequently, the effect of vibration exercise on muscle strength, balance, average power frequency of electromyogram (EMG), standing high jump, hormones, and blood flow, have been investigated36-47.</p><p>The effect of vibration training on BMD and body composition is lacking consistency in scientific findings. Some researchers found positive effects whereas others did not39, 40, 46, 48. To date, investigations have been dominated by research designs reporting temporary effects subsequent to short-term exercise (10 min), with relatively few investigations of long-term vibration training either alone or in comparison to other aerobic training programs39, 49, 50. Furthermore, there is a paucity of research on the effects of long-term vibration training on BMD. Therefore, the purpose of this study was to compare the physiological effects of caloric restriction and long-term vibration exercise with caloric restriction alone or in combination with aerobic exercise on body composition and BMD in middle-aged obese women.</p><!><p>Participants were selected through two different stages. Forty-five middle-aged obese women (30 to 55 year), who were not taking any medication and with >30 BMI (by Broca's index) were first selected as the subjects of this study. These women were housewives with low levels of activity who had not performed any kind of exercise over the last 6 months. In the second stage where dual energy X-ray absorptiometry was used, among the ones who passed the first filtering, those who were over 30% in percent body weight were selected as participants. Some of the applicants, whose BMI was less than 30, were also selected as final participants, since the ultimate selection criterion in this study was percent body fat. Applicants' menopausal status was not assessed. The participants consented by signature, after sufficient explanation of the experiment and an understanding of the possible adverse effects, and were randomly assigned into a diet group (n=9, control group), diet plus vibration training group (n=13, vibration group), and diet plus aerobic training group (n=11, aerobic group). Thirty-three of the participants completed the study (>95% compliance), thus, only their data were used in the analyses. Data from the remaining 12 subjects were discarded due to medication (n=2), withdrawal (n=5), and noncompliance (n=5). There were no significant differences in physical characteristics among groups before training (Table 1). All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation and with the Helsinki Declaration.</p><!><p>Mean + SD</p><!><p>Five clinical nutritionists calculated the energy needs of each participant based on age, height, weight, and activity, and prescribed the appropriate amount needed to lose weight51. Dietary intake approximated 70% of the recommended daily amount (RDA) for normal adults (~ 1,400 kcal). Participants were requested to record the form of their meal (i.e., type, amount, and ingredients of food) every day both before and during the 9 month intervention period, and nutritionists monitored their records every month to confirm adherence to the dietary treatment. Total caloric intake, relative caloric intake (i.e., %RDA), and amount of each macronutrient were analyzed by selecting the records 2 times per weekday and 1 time per weekend at various time points throughout the study. Nutrition analyses were performed using a Computer Applied Nutrition Analysis Program, according to Korea Nutrition Facts from the Korea Food & Drug Administration in the Ministry for Health, Welfare and Family Affairs, Republic of Korea52.</p><!><p>Participants in this group received the same 9 month dietary treatment as the diet group. Additionally, participants performed vibration training for 33 min which met the suggested guidelines of more than 30 min of training/session for the treatment of obesity21. Five min of stretching was performed before and after vibration exercise. A whole body vibration exercise machine (EOS 6600, MEDIEOS, Korea) was used for training, and the intensity of training was adjusted every 3 months. Participants performed 3 sets of exercise, and each set was composed of 5 min standing, 2 min squatting, 2 min squatting & heel, and 2 min upper body. For the standing position, the frequency of vibration (i.e., the intensity of training), was determined by multiplying the body weight of each participant by a constant of 0.25 for the first 3 months, 0.30 for the next 3 months, and 0.35 for the last 3 months. For squatting and squatting & heel position, the intensity was determined by multiplying the body weight of each participant by 0.40 for the first 3 months, 0.45 for the next 3 months, and 0.50 for the last 3 months. For the sitting position, the intensity was set at 22 Hz for the first 3 months, 24 Hz for the next 3 months, and 26 Hz for the last 3 months. The stimulus of vibration exercise to bones is linearly proportional to the increased ground reaction forces34. The amount of force exerted on the bone will linearly increase along with the reaction force, and amax, the maximal acceleration can be derived from the equation amax = A × ω2 = A×(2πf)2, where A is amplitude and f is frequency. Thus, during a vibration exercise at a frequency of 20Hz and an amplitude of 4 mm, the bone will be subjected to a force six times greater than gravitational acceleration. Therefore, the intensity of vibration based on maximum gravitation acceleration was calculated using 6.8 mm of vibration width, from 16 Hz to 23 Hz in the standing position, and from 28 Hz to 34 Hz in the squatting and squatting & heel positions, and corresponded to 6~29 times the force of gravity. This is similar to the intensity used in the study by Verschueren et al.42. However, the intensity in this study was greater than that of 30 Hz and 0.3g in the study conducted by Vicente et al53. The total duration of training was 9 months. The participants were trained between 10 am and 5 pm, once per day, 5 times/week (Monday through Friday). If a weekday training session was missed, it was made up on a weekend day.</p><!><p>For the standing position, feet were at a shoulder width apart and the body was in an attention position with the back straight. If there was any pain during training, the participants were allowed to bend their knees 5 degrees. For squatting, the position of the upper body was the same as in the standing position, but the angle between the thigh and calf was 120 degrees. Participants were standing in the squatting position with their heels lifted for the squatting & heel position. For the upper body position, participants knelt down behind the machine with their head raised, arms straightened, and hands on the vibration panel. Weight was distributed approximately equally between the hands and knees.</p><!><p>The same 9 month dietary treatment was used as in the diet group. Stretching of the lower body was performed for 5 min before and after aerobic exercise, and the exercise was performed for 33 min/day which was the same duration as in the vibration group. Participants trained 5 days/week by cycling (Combi 75XL, Japan) and treadmill walking (Taeha 6010, KOREA) on alternating days. Cycling intensity was set at a heart rate corresponding to 75% HRmax54 and monitored by sensors attached to the ears. The intensity of the treadmill was also set at a heart rate corresponding to 75% HRmax, measured telemetrically (Polar, Finland), with speed and incline adjusted every 5 min as necessary.</p><!><p>Height was measured as the distance between the bottom of the foot and top of the head using a stadiometer (PKS-1008, JAPAN). Skinfolds were determined as the average of 3 measurements of skin and subcutaneous fat thickness of triceps and thigh (Lange Skinfold Caliper, USA). The measurement was repeated until the difference between measurements was < 2 mm. Circumference of the upper triceps, waist, hip, and thigh were measured twice using a tape measure and values were averaged. The measurement was repeated until the difference between measurements was < 1 cm.</p><p>Weight, lean body mass, body fat mass, relative body fat (%), BMC, and BMD were measured using dual energy X-ray absorptiometry (DEXA) QDR-4500W (Hologic Inc., USA). BMD was measured in the whole body, lumbar spine (L1-L4), proximal femur of the left leg, and left forearm. The instrument was set in medium scan mode, DPX-α scan type, and 0.84 mm collimation. Size of the sample was 0.6 x 0.6 mm. Whole body measurements were made while participants were lying in the center of the table with their feet turned inward. For lumbar spine BMD, participants were in the center of the table with their coxa and knee joint bent. The equipment was set to touch the hip of the participants under their legs. Arms were either above the head or lying naturally on the table. Participants were lying straight with equipment fixed under their legs for the femur measurement. Forearm measurements were made while participants were sitting in a chair with the arm flat on the table and the palm down.</p><p>A 6 in (15 cm) section of the participant's left arm from mid-forearm to the first row of carpal bones was scanned. All the measurements were repeatedly taken by skilled examiners. To enhance quality control all instruments were calibrated prior to taking measurements.</p><!><p>SPSS 22.0 statistical package (IBM Corp., Armonk, USA) was used for the statistical analyses. Repeated measureemnts of two-way ANOVA were performed on data collected at four time points (i.e., pre-training, 3 months, 6 months, and 9 months). Least significant difference (LSD) was used post hoc where appropriate. Statistical significance was accepted at P < 0.05.</p><!><p>As expected, values for the nutritional variables were significantly higher at pre-training compared with 3, 6, and 9 months (p < 0.05). No significant interaction in total caloric intake, relative caloric intake (%RDA), protein, or carbohydrates occurred during the training period. However, there was a significant interaction for fat intake (p < 0.002). Post hoc analysis revealed a significantly higher fat intake at pre-training in the diet group compared with the vibration and aerobic groups. No significant differences in fat intake were found between groups or across time during the training period. Nutrition data are shown in Table 2.</p><!><p>Mean + SD; i, ii, iii: P < 0.05 among time points</p><p>*significantly higher (P < 0.05) than vibration and aerobic groups.</p><!><p>Body composition and anthropometric variables measured throughout the training period are shown in Tables 3 and 4. Body weight and relative body fat changed similarly in all three groups during the training period. Specifically, the women were heavier and had more relative fat at pre-training than at any other time point. They were also heavier and had more relative fat at 3 months than at 6 months or 9 months. Lean body mass was significantly higher at pre-training than at 3, 6, and 9 months. All groups had significantly higher fat mass at pre-training compared with values at 3 months (p < 0.001) and 6 months (p < 0.000), and at 3 months compared to 6 months (p < 0.001) of training. Moreover, fat mass in the two exercise groups, but not in the diet group, was significantly lower at 9 months compared to 3 months and pre-training (p < 0.001). There was a trend for a similar change in relative fat in the exercise groups (p < 0.10), however, it was not statistically significant. Circumferences and skinfold thicknesses were also reduced throughout the training period, with no significant differences among the groups. Specifically, upper triceps skinfold as well as upper triceps and thigh circumferences were significantly higher (p < 0.001) at pre-training than at 3, 6, and 9 months. They were also higher at 3 months than at 6 and 9 months (p < 0.001). Thigh skinfold did not significantly change from pre-training to 3 months, but was significantly lower at 6 months (p < 0.02) with an additional significant decrease at 9 months (p < 0.001). Waist circumference was significantly higher at pre-training and 3 months compared to 6 months (p < 0.001) and 9 months (p < 0.001). Hip circumference was highest at pre-training and decreased significantly (p < 0.001) at each successive time period.</p><!><p>Mean + SD; i, ii, iii: P < 0.05 among time points</p><p>*P < 0.05 versus pre and 3 month</p><p>†P < 0.10 versus pre and 3 month.</p><p>Mean + SD; ii, iii, iv: P < 0.05among time points.</p><!><p>Values for BMD and BMC are shown in Table 5. A significant interaction (p < .05) was detected for whole body BMC, but not for the other variables. Whole body BMC at 3 months and 6 months was significantly higher than at pre-training in the vibration group only (p < 0.001). Additionally, there was a significant time effect such that values at 9 months were significantly lower (p < 0.001) than at all other time points. Significant time effects were also noted for whole body BMD (p < 0.001), lumbar spine BMC (p < 0.04) and BMD (p < 0.04), and hip BMD (p < 0.05). Whole body BMD was significantly higher at 6 months than at pre-training (p < 0.001) or 3 months (p < 0.01), and significantly lower at 9 months than at 3 months (p < 0.02) or 6 months (p < 0.001). Lumbar spine BMC was significantly higher at 9 months (p < 0.04) than at any other time point, whereas lumbar spine BMD was significantly higher at 9 months than at 3 months (p < 0.03). Hip BMD was significantly higher at 6 months versus 9 months (p < 0.01). No significant changes were noted for hip BMC, forearm BMC or forearm BMD.</p><!><p>Mean + SD; WB: whole body</p><p>BMC: bone mineral content</p><p>BMD: bone mineral density</p><p>hip: left hip; forearm: left forearm</p><p>i, ii, iii, iv: P < 0.05 among time points</p><p>*P < 0.05 versus pre-training</p><!><p>Previous research regarding vibration exercise focused on prevention of bone mineral loss, hormones, circulating levels of serum lipids, peripheral circulation, physical strength, reaction, and the adaptive effect on EMG33, 34, 39, 41, 42, 44. However, those studies are limited by the use of short duration vibration exercise in a straight posture. In the current study, we examined changes in body composition, BMD, and BMC in response to long-term vibration training in combination with caloric restriction, and compared them to diet only or diet plus aerobic exercise.</p><p>With the exception of fat mass, body composition changed similarly among the groups during the training period. Specifically, the women lost a significant amount of body weight, lean body mass and relative fat, reduced their skinfold thickness, and had smaller waist and hip circumferences. They also lost a significant amount of fat mass through the first 6 months of training. Women in the two exercise groups also had lower fat mass after 9 months compared to 3 months whereas women in the diet group did not. Therefore, exercise seemed to provide an additional stimulus for reducing fat mass. This was expected for the aerobic group given the additional caloric expenditure from exercise versus the diet only group. Results of the vibration group suggest that vibration exercise is as effective as aerobic exercise in reducing fat mass when combined with dietary intervention, thereby showing promise as an obesity intervention therapy. In contrast, there are contradictory findings on the effect of vibration exercise with men. Di Loreto et al.55 reported that when 10 healthy men were exposed to vibration of 30 Hz for 25 min, glucose levels decreased after 5 min due to muscle contraction and the consequent increased consumption of glucose. However, after 30 min, no changes were noted for growth hormone, insulin, and insulin-like growth factor-1, suggesting that vibration may not be an effective method for obesity treatment. Therefore, the physiological relationship between vibration exercise and the reduction in body fat needs further study.</p><p>The physiological mechanisms associated with vibration exercise can be explained by the 'tonic vibration reflex' (TVR). The physical vibration causes changes in the muscle structure which activates the muscle spindle resulting in a reflective contraction56. Kvorning et al.56 used this mechanism to explain how vibration exercise can improve body structure and strength. In addition, Kasai et al.57 showed that the vibration stimulus not only activates the receptors of muscle spindles directly but also indirectly affected neighboring muscles. Runge et al.58 showed that whole body vibration training for 12 weeks increased muscle strength in the elderly, resulting in an 18% improvement in a chair sit up test. Rittweger et al.34 explained the effect of vibration exercise through the observed reduction in subcutaneous fat through "itching skin" in the lower leg resulting from increased blood circulation. In addition, a study by Figueroa37 showed the positive effects of WBV on the cardiovascular system through a reduction in aortic blood pressure in postmenopausal women with pre hypertension and hypertension. Therefore, it can be suggested that vibration stimulus to the whole body trains both muscular and nervous structures, improves the function of the neuromuscular system and improves circulatory flow resulting in improvement of the whole body structure.</p><p>Analysis of BMC and BMD during the training period suggests that whole body BMC of the women showed different changes depending on their intervention. More specifically, women in the vibration group had significantly greater whole body BMC at 3 months and 6 months of training compared with pre-training. No other group effects or interactions were found for any of the other BMC or BMD measures. Torvinen et al.43 also reported similar results. In their research, 56 adults participated in 8 months of vibration exercise 3-5/week at an intensity of 25-45Hz for 4 min/session. They did not find a significant change in BMD of the lumbar spine after training. The authors suggested that the lack of change was because the vibration stimulus might not have been strong enough to affect the BMD of their young premenopausal women participants. In the present study, more than 70% of the participants were also premenopausal female adults < 45 years of age, so it is possible that the 9 months training program might not have provided a stimulus strong enough to improve BMD.</p><p>In a study by Roelants et al.42, 48 women underwent 6 months of vibration training at an intensity of 35-40 Hz, 5 min/day, 3 days/week. Lean body mass increased significantly while skinfolds, weight, and relative body fat did not change. Compared to a resistive exercise group, knee-extension strength increased significantly, while hip BMD did not change. The authors claimed that the duration of each exercise, rather than the intensity, influences the effect of vibration exercise on body structure. In contrast, others suggest that vibration exercise can positively affect BMD. Flieger et al.33 found in ovariectomized mice that BMD increased by the 5th week of training at 50 Hz and was maintained through week 12. The mice exercised for 30 min/day, 5 days/week for 3 months. Moreover, Verschueren et al.42 conducted a study in which 70 postmenopausal women performed vibration exercise at 35-40 Hz , 3 days/week for 6 months. They found a reduction in total body fat and improved isometric (15.1%) and isotonic (16.4%) quadriceps muscle strength without any negative side effects. In addition, while total body and lumbar spine BMD was not changed, hip BMD significantly increased by 0.93%. This is a different result from participants in the resistance exercise group who showed significant increases in muscle strength, but no change in hip BMD. They also reported the lack of a significant relationship between improvements in isometric and motile muscular strength with changes in BMD and lean body mass in the vibration group, suggesting that reflexive muscular contraction does not affect bone formation.</p><p>One of the limitations in this study is that the study did not fully cover the direct variables for menopausal status, only the BMD and BMC values were directly assessed. In this study, we focused on our first objective- that WBV is an effective aerobic exercise. We decided to work on particular areas only in order to examine BMD and BMC based on thorough review of precedent studies. Future studies are needed to determine how other variables in menopausal women affect vibration training.</p><!><p>Our results suggest that long-term whole body vibration training, when used in conjunction with a reduction in caloric intake, is as effective as combining diet with aerobic exercise in improving body composition of obese middle-aged women without compromising BMC or BMD. Thus, it can be considered a novel and effective method for reducing body fat.</p><!><p>Mihalko WM, Bergin PF, Kelly FB, Canale ST. Obesity, Orthopaedics, and Outcomes. J Am Acad Orthop Surg. 2014;22: 683-90.</p><p>Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of obesity in the United States, 2009-2010. NCHS Data Brief. 2012: 1-8.</p><p>Fruhbeck G, Yumuk V. Obesity: a gateway disease with a rising prevalence. Obes Facts. 2014;7 Suppl 2:33-6.</p><p>Ramachandran A, Chamukuttan S, Shetty SA, Arun N, Susairaj P. Obesity in Asia - is it different from rest of the world. Diabetes Metab Res. 2012;28:47-51.</p><p>Irving BA, Davis CK, Brock DW, Weltman JY, Swift D, Barrett EJ, Gaesser GA, Weltman A. Effect of exercise training intensity on abdominal visceral fat and body composition. Med Sci Sports Exerc. 2008; 40:1863-72.</p><p>Johnston CA, Tyler C, Foreyt JP. Behavioral management of obesity. Curr Atheroscler Rep. 2007;9:448-53.</p><p>Shewmake RA, Huntington MK. Nutritional treatment of obesity. Prim Care. 2009;36:357-77.</p><p>Matarese LE, Pories WJ. Adult Weight Loss Diets: Metabolic Effects and Outcomes. Nutr Clin Pract. 2014;29:759-67.</p><p>Andrade AM, Coutinho SR, Silva MN, Mata J, Vieira PN, Minderico CS, Melanson KJ, Baptista F, Sardinha LB, Teixeira PJ. The effect of physical activity on weight loss is mediated by eating self-regulation. Patient Educ Couns. 2010;79:320-6.</p><p>Redman LM, Rood J, Anton SD, Champagne C, Smith SR, Ravussin E. Calorie restriction and bone health in young, overweight individuals. Arch Intern Med. 2008;168:1859-66.</p><p>Dhurandhar EJ, Kaiser KA, Dawson JA, Alcorn AS, Keating KD, Allison DB. Predicting adult weight change in the real world: a systematic review and meta-analysis accounting for compensatory changes in energy intake or expenditure. Int J Obes (Lond). 2015;39: 1181-7.</p><p>Jensen LB, Quaade F, Sorensen OH. Bone loss accompanying voluntary weight loss in obese humans. J Bone Miner Res. 1994;9:459-63.</p><p>Maahs D, de Serna DG, Kolotkin RL, Ralston S, Sandate J, Qualls C, Schade DS. Randomized, double-blind, placebo-controlled trial of orlistat for weight loss in adolescents. Endocr Pract. 2006;12:18-28.</p><p>Zelissen PM, Stenlof K, Lean ME, Fogteloo J, Keulen ET, Wilding J, Finer N, Rössner S, Lawrence E, Fletcher C, Mc-Camish M; Author Group. Effect of three treatment schedules of recombinant methionyl human leptin on body weight in obese adults: a randomized, placebo-controlled trial. Diabetes Obes Metab. 2005;7:755-61.</p><p>King NA, Hopkins M, Caudwell P, Stubbs RJ, Blundell JE. Beneficial effects of exercise: shifting the focus from body weight to other markers of health. Br J Sports Med. 2009;43:924-7.</p><p>Chaput JP, Klingenberg L, Rosenkilde M, Gilbert JA, Tremblay A, Sjodin A. Physical activity plays an important role in body weight regulation. J Obes. 2011;2011:360257-11</p><p>Alnasir FA, Masuadi EM. The effect of loss of body weight on lipid profile in overweight individuals. Saudi Med J. 2006;27: 687-92.</p><p>Fagard RH. Prescription and results of physical activity. J Cardiovasc Pharmacol. 1995;25:S20-7.</p><p>Ferrara CM, Goldberg AP, Ortmeyer HK, Ryan AS. Effects of aerobic and resistive exercise training on glucose disposal and skeletal muscle metabolism in older men. J Gerontol A Biol Sci Med Sci. 2006; 61:480-7.</p><p>Raz I, Hauser E, Bursztyn M. Moderate exercise improves glucose metabolism in uncontrolled elderly patients with non-insulin-dependent diabetes mellitus. Isr J Med Sci. 1994;30:766-70.</p><p>Medicine ACoS. ACSM's Guidelines for Exercise Testing and Prescription. 7 ed: Lea & Febiger. Philadelphia. ACSM. 2006.p216-9.</p><p>Frost HM. Suggested fundamental concepts in skeletal physiology. Calcif Tissue Int. 1993;52:1-4.</p><p>Hyldstrup L, Andersen T, McNair P, Breum L, Transbol I. Bone metabolism in obesity: changes related to severe overweight and dietary weight reduction. Acta Endocrinol (Copenh). 1993; 129:393-8.</p><p>Lindsay R, Cosman F, Herrington BS, Himmelstein S. Bone mass and body composition in normal women. J Bone Miner Res. 1992;7:55-63.</p><p>Revilla M, Villa LF, Sanchez-Atrio A, Hernandez ER, Rico H. Influence of body mass index on the age-related slope of total and regional bone mineral content. Calcif Tissue Int. 1997;61: 134-8.</p><p>Kelley GA, Kelley KS, Tran ZV. Resistance training and bone mineral density in women: a meta-analysis of controlled trials. Am J Phys Med Rehabil. 2001;80:65-77.</p><p>Daly RM, Dunstan DW, Owen N, Jolley D, Shaw JE, Zimmet PZ. Does high-intensity resistance training maintain bone mass during moderate weight loss in older overweight adults with type 2 diabetes? Osteoporos Int. 2005;16:1703-12.</p><p>Karlsson M. Has exercise an antifracture efficacy in women? Scand J Med Sci Sports. 2004;14:2-15.</p><p>Chubak J, Ulrich CM, Tworoger SS, Sorensen B, Yasui Y, Irwin ML, Stanczyk FZ, Potter JD, McTiernan A. Effect of exercise on bone mineral density and lean mass in postmenopausal women. Med Sci Sports Exerc. 2006;38:1236-44.</p><p>Cooper C, Inskip H, Croft P, Campbell L, Smith G, Mclearn M, Coggon D. Individual Risk factors for Hip Osteoarthritis: Obesity, Hip Injury and Physical Activity. Am J Epidemiol. 1998;147:516-22.</p><p>Manninen P, Riihimaki H, Heliovaara M, Makela P. Overweight, gender and knee osteoarthritis. Int J Obes Relat Metab Disord. 1996;20:595-7.</p><p>Bosco C, Cardinale M, Colli R, Tihanyi J, von Duvillard SP, Viru A. The influence of whole body vibration on jumping performance. Biol Sport. 1998;15:157-64.</p><p>Flieger J, Karachalios T, Khaldi L, Raptou P, Lyritis G. Mechanical stimulation in the form of vibration prevents postmenopausal bone loss in ovariectomized rats. Calcif Tissue Int. 1998;63:510-4.</p><p>Rittweger J, Beller G, Felsenberg D. Acute physiological effects of exhaustive whole-body vibration exercise in man. Clin Physiol. 2000;20:134-42.</p><p>Roelants M, Verschueren SM, Delecluse C, Levin O, Stijnen V. Whole-body-vibration-induced increase in leg muscle activity during different squat exercises. J Strength Cond Res. 2006; 20:124-9.</p><p>Giunta M, Cardinale M, Agosti F, Patrizi A, Compri E, Rigamonti AE, Sartorio A. Growth hormone-releasing effects of whole body vibration alone or combined with squatting plus external load in severely obese female participants. Obes Facts. 2012; 5:567-74.</p><p>Figueroa A, Kalfon R, Madzima TA, Wong A. Effects of whole-body vibration exercise training on aortic wave reflection and muscle strength in postmenopausal women with prehypertension and hypertension. J Hum Hypertens. 2014;28:118-22.</p><p>Bosco C, Colli R, Introini E, Cardinale M, Tsarpela O, Madella A, Tihanyi J, Viru A. Adaptive responses of human skeletal muscle to vibration exposure. Clin Physiol. 1999;19:183-7.</p><p>Bosco C, Iacovelli M, Tsarpela O, Cardinale M, Bonifazi M, Tihanyi J, Viru M, De Lorenzo A, Viru A. Hormonal responses to whole-body vibration in men. Eur J Appl Physiol. 2000;81:449-54.</p><p>Cardinale M, Pope MH. The effects of whole body vibration on humans: Dangerous or advantageous? Acta Physiol Hung. 2003;90:195-206.</p><p>Kerschan-Schindl K, Grampp S, Henk C, Resch H, Preisinger E, Fialka-Moser V, Imhof H. Whole-body vibration exercise leads to alterations in muscle blood volume. Clin Physiol. 2001;21: 377-82.</p><p>Verschueren SM, Roelants M, Delecluse C, Swinnen S, Vanderschueren D, Boonen S. Effect of 6-month whole body vibration training on hip density, muscle strength, and postural control in postmenopausal women: a randomized controlled pilot study. J Bone Miner Res. 2004;19:352-9.</p><p>Torvinen S, Kannus P, Sievanen H, Jarvinen TA, Pasanen M, Kontulainen S, Nenonen A, Järvinen TL, Paakkala T, Järvinen M, Vuori I. Effect of 8-month vertical whole body vibration on bone, muscle performance, and body balance: a randomized controlled study. J Bone Miner Res. 2003;18:876-84.</p><p>Torvinen S, Sievanen H, Jarvinen TA, Pasanen M, Kontulainen S, Kannus P. Effect of 4-min vertical whole body vibration on muscle performance and body balance: a randomized cross-over study. Int J Sports Med. 2002;23:374-9.</p><p>Bemben DA, Palmer IJ, Bemben MG, Knehans AW. Effects of combined whole-body vibration and resistance training on muscular strength and bone metabolism in postmenopausal women. Bone. 2010;47:650-6.</p><p>Cristi-Montero C, Cuevas MJ, Collado PS. Whole-body vibration training as complement to programs aimed at weight loss. Nutr Hosp. 2013;28:1365-71.</p><p>Despina T, George D, George T, Sotiris P, Alessandra DC, George K, Maria R, Stavros K. Short-term effect of whole-body vibration training on balance, flexibility and lower limb explosive strength in elite rhythmic gymnasts. Hum Mov Sci. 2014;33:149-58.</p><p>Wysocki A, Butler M, Shamliyan T, Kane RL. Whole-Body Vibration Therapy for Osteoporosis: State of the Science. Ann Intern Med. 2011;155:680-U60.</p><p>Torvinen S, Kannus P, Sievanen H, Jarvinen TA, Pasanen M, Kontulainen S, Järvinen TL, Järvinen M, Oja P, Vuori I. Effect of four-month vertical whole body vibration on performance and balance. Med Sci Sports Exerc. 2002;34:1523-8.</p><p>Milanese C, Piscitelli F, Zenti MG, Moghetti P, Sandri M, Zancanaro C. Ten-week Whole-body Vibration Training Improves Body Composition and Muscle Strength in Obese Women. Int J Med Sci. 2013;10:307-11.</p><p>Society KN. Recommend Dietary Allowances for Koreans, 7th revision. Seoul. 2000.</p><p>Society KN. Computer Aided Nutritional Analysis Program. Seoul. 1995.</p><p>Gilsanz V, Wren TA, Sanchez M, Dorey F, Judex S, Rubin C. Low-level, high-frequency mechanical signals enhance musculoskeletal development of young women with low BMD. J Bone Miner Res. 2006;21:1464-74.</p><p>Miyashita M, Mutoh Y, Yoshika Y, Sadamoto T. Effects of physical training. Med Sci Sports Exerc. 1985;10:3-5.</p><p>Di Loreto C, Ranchelli A, Lucidi P, Murdolo G, Parlanti N, De Cicco A, Tsarpela O, Annino G, Bosco C, Santeusanio F, Bolli GB, De Feo P. Effects of whole-body vibration exercise on the endocrine system of healthy men. J Endocrinol Invest. 2004; 27:323-7.</p><p>Kvorning T, Bagger M, Caserotti P, Madsen K. Effects of vibration and resistance training on neuromuscular and hormonal measures. Eur J Appl Physiol. 2006;96:615-25.</p><p>Kasai T, Kawanishi M, Yahagi S. The effects of wrist muscle vibration on human voluntary elbow flexion-extension movements. Exp Brain Res. 1992;90:217-20.</p><p>Runge M, Rehfeld G, Resnicek E. Balance training and exercise in geriatric patients. J Musculoskelet Neuronal Interact. 2000;1:61-5.</p>

PubMed Open Access

<i>De novo</i> synthesis of mesoporous photoactive titanium(<scp>iv</scp>)–organic frameworks with MIL-100 topology

Most developments in the chemistry and applications of metal-organic frameworks (MOFs) have been made possible thanks to the value of reticular chemistry in guiding the unlimited combination of organic connectors and secondary building units (SBUs) into targeted architectures. However, the development of new titanium-frameworks still remains limited by the difficulties in controlling the formation of persistent Ti-SBUs with predetermined directionality amenable to the isoreticular approach. Here we report the synthesis of a mesoporous Ti-MOF displaying a MIL-100 topology. MIL-100(Ti) combines excellent chemical stability and mesoporosity, intrinsic to this archetypical family of porous materials, with photoactive Ti 3 (m 3 -O) metal-oxo clusters. By using high-throughput synthetic methodologies, we have confirmed that the formation of this SBU is thermodynamically favored as it is not strictly dependent on the metal precursor of choice and can be regarded as an adequate building block to control the design of new Ti-MOF architectures. We are confident that the addition of a mesoporous solid to the small number of crystalline, porous titanium-frameworks available will be a valuable asset to accelerate the development of new porous photocatalysts without the pore size limitations currently imposed by the microporous materials available.

<i>de_novo</i>_synthesis_of_mesoporous_photoactive_titanium(<scp>iv</scp>)–organic_frameworks_with_m

Introduction<!>Synthesis from pre-formed metal clusters<!>Structure and characterization<!>Optical response and photoactivity<!>Synthesis from other metal precursors<!>Conclusions<!>Chemical and physical characterization<!>Conflicts of interest

<p>Metal-organic frameworks (MOFs) are crystalline, porous materials built from the interlinking of metal-oxo clusters with organic connectors. 1 Their internal porosity can be engineered in size, shape and chemical function to provide a family of crystalline solids with structural and functional diversities beyond comparison. These intrinsic characteristics make MOFs appealing materials for modern storage, separation, delivery and catalysis technologies. [2][3][4][5][6] The problems linked to the development of these applications have stressed the need for producing MOFs with superior chemical stability. 7 Combination of hydrolytic stability with high surface area was rst reported for sodalite-type frameworks built from imidazolate (ZIF) 8 or azolate 9 connectors due to hydrophobic shielding and strong metal-nitrogen bonds. Control over the strength of the coordination bond is also key to the robustness of Zr(IV) and Hf(IV) MOFs.</p><p>Compared to these metals, titanium offers several advantages. It is less toxic, naturally more abundant and more adequate to endow these porous materials with photoactivity. Still, the synthesis of crystalline, porous titanium-organic frameworks remains comparatively underexplored. 10 Though the discovery of the UiO family, based on Zr 6 O 4 (OH) 4 (RCO 2 ) 12 clusters, 11 boosted the synthesis of multiple MOFs from isostructural Zr and Hf metal-oxo clusters, [12][13][14] the controllable synthesis of Ti-MOFs remains quite challenging yet. This is due to the higher reactivity of Ti precursors, more prone to hydrolysis, that facilitates the formation of amorphous titanium oxide under solvothermal conditions. The formation of titanium-oxocarboxylate clusters, capable of directing the formation of high symmetry, periodic architectures, demands stricter control over the hydrolysis-condensation equilibria. These synthetic difficulties are possibly the reason due to which almost all Ti-MOFs available are built from different secondary building units (SBUs). Fig. 1 summarizes the diversity of Ti-SBUs reported for the limited number of MOFs available: Ti 12 O 15 /Ti 6 O 9 titaniumoxo clusters in MIL-177, 15 Ti 8 O 8 in MIL-125, 16 Ti 7 O 4 in PCN-22, 17 Ti 6 O 6 in MOF-901 (ref. 18) and MTM-1, 19 Ti 3 O in COK-69, 20 heterometallic clusters like Ti 8 Zr 2 O 12 in PCN-415 (ref. 21) and Ti 2 M 2 O 2 (M ¼ Ca, Mn) in MUV-10, 22 or simple titanium nodes in NTU-9 (ref. 23) and Ti-CAT-5, 24 conrm the complexity for targeting the formation of specic MOF architectures due to the difficulties in controlling the formation of persistent SBUs.</p><p>This lack of chemical control over the nal material is arguably limiting further development, highlighting the importance of nding alternative synthetic strategies that make Ti-MOFs amenable to the design principles of isoreticular chemistry. One strategy used to gain control is to produce daughter frameworks by incorporation of titanium(IV) into preformed materials by partial post-synthetic replacement of parent metals. This was originally demonstrated for the UiO-66(Zr) 25 and MOF-5(Zn) 26 families. More recent examples describe the use of titanium(III) precursors followed by mild oxidation for partial doping of MOF-74, MIL-100 and PCN-333 with Ti. 27 However, these methodologies do not allow for complete metal replacement and also suffer from poor control over the positioning of titanium atoms in the structure, 28 which can result in spurious deposition of metal oxide coatings. 29 Compared to this post-synthetic strategy, here we report the formation of MIL-100(Ti) by direct synthesis. Systematic exploration of the chemical space, by using high-throughput (HT) methodologies, has enabled identication of the optimal synthetic conditions to produce highly crystalline solids from different titanium precursors. MIL-100(Ti) combines the chemical stability and mesoporosity, intrinsic to the MIL-100 family, with photoactive Ti 3 (m 3 -O) metal-oxo clusters. Our work suggests that the formation of this SBU, classical in the chemistry of MOFs based on trivalent metals, is not strictly dependent on the metal source and might be used for the controlled design of new Ti-MOF architectures.</p><!><p>Our rst approach to the direct synthesis of Ti-MOFs involved the use of titanium-oxo clusters as pre-formed SBUs. In principle, this was expected to favour the formation of crystalline solids under milder conditions than those conventionally used in MOF synthesis. This strategy has been successfully used to produce Zr(IV), 30 Ce(IV), 31 and Fe(III)-MOFs 32 by direct reaction of metal-oxo clusters with polyaromatic carboxylic acids. Compared to simple precursors, we also expected to minimize the precipitation of amorphous oxides due to the higher stability of their cluster against hydrolysis. Accordingly, we reacted the as-made [Ti 6 O 6 (4-tbbz) 6 (O i Pr) 6 ] (tbbz ¼ 4-tert-butylbenzoic acid) clusters (Fig. 2a) with trimesic acid (H 3 btc ¼ benzene-1,3,5-tricarboxylic acid) between 80 and 160 C for multiple combinations of solvents (DMF, DEF, NMP, MeOH, EtOH.), metal to linker ratios, concentration and addition of different modulators like acetic or benzoic acid. We used a FLEX SHAKE high-throughput workstation from Chemspeed© for robotic dispensing of solids and liquids in order to optimize the screening of a broad range of synthetic conditions whilst ensuring reproducibility. All attempts were unsuccessful as we could only isolate the unreacted cluster or amorphous solids for temperatures below or above 100 C, respectively. We argued this was due to the poor solubility of the cluster under these conditions and decided to investigate an alternative mixture of solvents. Aer several tests, we observed that heating of a mixture of [Ti 6 O 6 (4-tbbz) 6 (O i Pr) 6 ] (Ti 6 ) and H 3 btc in acetonitrile : tetrahydrofuran (ACN : THF, 3 : 1 v/v%) led to the formation of a white, microcrystalline solid. In order to optimize the crystallinity of the material we systematically varied the temperature of the reaction, the addition of acetic acid as the modulator, and the metal-to-linker ratio. We observed that, for a given Ti : btc ratio of 1 : 5, only the unreacted cluster was obtained at temperatures below 140 C with no addition of the modulator (Fig. S1-S4 †). However, a crystalline solid is observed in the presence of acetic acid as the modulator (180-550 equiv., 250-750 mL) and at temperatures above 100 C. Note that a secondary crystalline phase is obtained as a side-product at low temperatures and high amounts of acetic acid (i.e. 550 equiv., 750 mL) as evidenced by powder X-ray diffraction (PXRD) (Fig. S2-S5 †), whereas it is not observed for a lower concentration of the modulator. The formation of this secondary phase can be reduced by increasing the temperature of the reaction up to 160 C (Fig. 2b-c, S6 and S7 †). We next varied the Ti : btc ratio at 160 C for a xed amount of acetic acid (i.e. 180 equiv., 250 mL). PXRD of the solids conrms the formation of a highly crystalline MIL-100 phase for Ti : btc ratios of 1 : 2-5 (Fig. S8 †). PXRD of the solids synthesized with increasing temperatures and concentration of the modulator conrms the critical effect of these parameters on the crystallinity of the material (Fig. 2b and c). The optimum conditions for the formation of highly crystalline MIL-100(Ti) corresponds to a Ti : btc ratio of 1 : 5 and addition of 250 mL of acetic acid (180 equiv.) and reaction at 160 C (see ESI † for details). The use of more conventional solvents like DMF or DEF yielded amorphous or poorly crystalline materials under equivalent reaction conditions (Fig. S9 and S10 †). Optical microscopy suggested the formation of small crystals, which was conrmed by scanning electron microscopy (SEM). As shown in Fig. 2e, the bulk is composed of single crystals with an octahedral habit, typical of cubic MIL-100 phases, and sizes oscillating below 5 mm. We also identied small fractions of amorphous material partially covering the surface of the crystals (Fig. S11 †). The solid was recovered by centrifugation and washed with DMF and MeOH prior to drying in a vacuum at room temperature.</p><!><p>PXRD patterns of the solid collected at our laboratory revealed clear similarities to the reported phases of MIL-100(Cr, Fe) (Fig. S16 †). 33,34 For a clearer structural description, we collected high-resolution PXRD data that were rened by using an starting model built from the reported structure of MIL-100(Fe) by using Materials Studio. As shown in Fig. 3a, Rietveld renement of the activated solid converged with excellent residual values (R wp ¼ 5.24%, R exp ¼ 1.93%) for a cubic Fd 3m space group with cell parameter a ¼ 73.5168( 16) Å. The structure of MIL-100(Ti) is built from the interlinking of Ti 3 (m 3 -O) clusters and btc linkers to conform a zeolitic mnt topology that features two types of mesoporous cages with pentagonal and hexagonal pore windows (Fig. 3b).</p><p>The assembly of this structure from the starting Ti 6 O 6 cluster used in the synthesis necessarily involves its degradation for the assembly of a thermodynamically more stable unit under the reaction conditions. This dynamic equilibrium between interconverting species is similar to that controlling the formation of heptameric clusters or Ti-oxo chains from Ti 6 O 6 (abz) 6 (O i Pr) 6 (abz ¼ 4-aminobenzoate) during the synthesis of PCN-22 (ref. 17) or DGIST-1, 35 respectively. By using comparatively milder conditions, the integrity of the starting homo or heterometallic polynuclear Ti(IV) clusters can also be preserved and they can be incorporated into the framework, as demonstrated for MOF-901, 18 MNT-1 (ref. 19) or PCN-405. 21 Compared to other precursors, these precedents seem to conrm the suitability of robust, molecular Ti clusters to either retain their nuclearity or control the formation of alternative SBUs more prone to induce the formation of specic MOF architectures under thermodynamic control. For example, M 3 (m 3 -O) (M ¼ Cr, Al, Fe) SBUs are very common in frameworks produced under relatively harsh conditions like MIL-96 (ref. 36) or MIL-101. 37 The replacement of trivalent metals with Ti(IV) is not directly compatible with the local structure of MIL-100 materials, which features 6-connected [M III 3 (m 3 -O)(X)(H 2 O) 2 (btc) 6 ] (X À ¼ F, Cl, OH) nodes. In our case, the introduction of an excess of one positive charge per metal atom will be counterbalanced by additional anions. We hypothesized that these could be directly coordinated to the Ti sites or occupy the empty space available from the mesoporous cages for neutral or cationic frameworks, respectively. FT-IR spectra of the solid, before and aer Soxhlet washing (Fig. S19 †), 1 H-NMR (Fig. S26 †) and solid-state 13 C-CP-MAS-NMR (Fig. S28 †), all discard the presence of Ti 6 clusters or free trimesate or acetate anions occupying the pores. Hence, we modelled computationally different neutral clusters by modication of the molecules coordinated to the axial position of the octahedral Ti(IV) centers whilst xing the positions corresponding to the btc linkers and the m 3 -O central bridge. We used the same methodology recently reported by De Vos and co-workers for rationalizing the most plausible formula of COK-69, 20 The proposed formulae agree well with the percentage of residual oxide formed from the thermal decomposition of the solid (Fig. S21 †). The experimental 32.2% of TiO 2 is quite close to the calculated value of 33.9%. Permanent porosity was studied using N 2 adsorption-desorption isotherms at 77 K aer activation of the solid under dynamic vacuum (10 À3 mbar) at 150 C overnight. MIL-100(Ti) displays a non-hysteretic, type-I N 2 adsorption with two secondary gas uptakes at P/P 0 ¼ 0.05 and 0.14 (Fig. 4a). The multi-point BET surface area was found to be 1321 m 2 g À1 with a total pore volume of 0.66 cm 3 g À1 (Fig. S22 and Table S3 †). Pore size distribution (PSD) calculated using Non-Linear Density Functional Theory (NLDFT) models reveals two types of mesoporous pores between 1.6-2.4 and 2.6-3.6 nm that correlate well with the theoretical pore size values of 2.4 and 2.9 nm estimated from the structure of MIL-100(Ti). This surface area lies below the values reported for MIL-100(Cr, Fe) phases, 33,34 suggesting ineffective activation or the partial blocking of the pores by molecular precursors with poor solubility. We used other activation protocols more adequate for highly porous materials like solvothermal treatment of the solid or Soxhlet washing with hot EtOH or MeOH. None of them led to a substantial increase in porosity. As commented before, FT-IR, 1 H-NMR and 13 C-CP-MAS-NMR of the activated solid rule out the presence of molecular components occluded in the pores of MIL-100(Ti). Thus, we ascribe the reduction of surface area to the formation of an amorphous fraction of the material that can be inferred from SEM (Fig. S11 †), similar to previous reports for other mesoporous phases like MIL-101-NDC(Cr). 38 CO 2 adsorption studies reveal a modest uptake of 1.70 mmol g À1 (7.5 wt%) at 293 K and 1 bar, with an isosteric heat of adsorption (Q st ) of 34.8 kJ mol À1 (Fig. 4b and S24 †). This value is very similar to the 30-35 kJ mol À1 reported for the Fe(III) phase 39 and agrees well with the presence of vacant metal sites. Water adsorption also reveals clear similarities with other MIL-100 derivatives. As shown in Fig. 4c, the isotherm at 298 K shows an 'S'-shaped prole with hysteretic behaviour, indicative of irreversible capillary condensation. Water adsorption at low relative pressures, which can be ascribed to the presence of open metal sites in the structure, is followed by a steep uptake at a relative humidity between 0.25 and 0.45% with an inection point at P/P 0 ¼ 0.32, characteristic of hydrophilic materials. 40 The maximum water adsorption capacity of MIL-100(Ti) calculated at P/P 0 ¼ 0.9 is 0.52 g g À1 . This gravimetric adsorption is very similar to the value reported for MIL-100(Al) (0.51 g g À1 ) but smaller than those for the Cr and Fe phases ranging from 0.60 to 0.76 g g À1 (Fig. 4c and Table S4 †). 41,42 Thermal and chemical stability One of the main features of MIL-100 materials is their stability in water. For Fe, Al and Sc phases, this stability is thermodynamic in origin and arises from strong coordination bonds with M-O bond dissociation energies of 444, 502 and 671 kJ mol À1 . 43 In turn, the kinetic inertness of Cr(III) ions in MIL-100(Cr) is responsible for its outstanding hydrolytic stability. Accordingly, the incorporation of highly charged Ti(IV) metals into MIL-100 topologies will result in high thermal and chemical stability due to thermodynamic parameters. Compared to the Cr and Fe phases that decompose between 300 and 350 C, 33,34 the thermogravimetric analysis (TGA) of MIL-100(Ti) conrms a signicant increase in the thermal stability of the solid that remains stable up to 450 C (Fig. S21 †). This thermal stability is similar to other Ti-MOFs and agrees well with the higher dissociation energy of O-Ti(IV) bonds of 667 kJ mol À1 . 43 We next evaluated the hydrolytic stability of MIL-100(Ti) by soaking freshly made solids in water under acidic and basic conditions (pH 1-14). PXRD of the solids aer incubation in water for 24 hours conrms that the structure of the solid remains unchanged between pH 2 and 12 (Fig. 4d and S32 †). Stability was also conrmed by gas sorption. N 2 adsorption measurements of the solids aer water treatment show negligible changes in their surface area compared to the as-made material (Fig. 4e, S33 and Table S5 †). These results conrm that MIL-100(Ti) displays excellent hydrolytic stability, comparable to other benchmark materials. 7 For further understanding on the role of Ti 3 (m 3 -O) nodes over the chemical stability of the material we decided to carry out equivalent experiments for direct comparison with the isostructural MIL-100(Fe) 44 and a representative Ti-MOF like MIL-125. 16 PXRD and N 2 sorption conrm that replacement of Fe(III) with Ti(IV) centers does not induce important changes under equivalent conditions. MIL-100(Fe) only shows minor degradation of the structure at pH 12 (Fig. S34 †), whereas its surface area is maintained close to the original value under acidic and basic conditions (Fig. S35 and Table S6 †). This behaviour is different for MIL-125 that retains its structural integrity and porous function almost intact at pH-12 but undergoes signicant degradation at acidic pH (Fig. S36, S37 and Table S7 †). We observe more important differences in the ICP-MS measurements of the supernatants aer incubation in water. As shown in Fig. 4f, titanium leaching in MIL-100(Ti) is almost negligible for acidic and neutral conditions. In turn, the concentration of iron and titanium in solution increases with decreasing pH up to a maximum of 35 and 14 mg mL À1 at pH 2 for MIL-100(Fe) and MIL-125, respectively (see Table S8 † for details). Concerning the isostructural series, we presume that the different degradation rates of the Ti and Fe phases might result in different drug delivery kinetics under simulated biological conditions. Although this study is beyond the scope of our work, we are condent that MIL-100(Ti) might be of interest to the MOF community for biomedical applications.</p><!><p>The applications of Ti-MOFs as photocatalysts for solar fuel production or light-induced organic transformations are pivotal to this family of materials due to the promising combination of high surface areas, chemical stability and photoactivity. 45 The optical band-gap of MIL-100(Ti) obtained by diffuse reectance spectroscopy is 3.40 eV (Fig. S30 †). This value lies between those reported for MIL-125 (3.68 eV) 46 or MUV-10(Ca) (3.10 eV) 22 and is compatible with UV light photoexcitation. This was conrmed by irradiating a suspension of the solid in dry THF with UV-B light (l ¼ 280-315 nm). Aer a few hours, the color of the solid changed from white to dark blue (Fig. S31 †) and remained stable in the absence of oxygen. The electron paramagnetic resonance (EPR) spectra of the solid before/aer irradiation conrms the occurrence of a broad signal at 350 mT, with g-tted parameters g k ¼ 1.927 and g t ¼ 1.952 characteristic of Ti(III) species, only for the irradiated sample (Fig. 5a and S31). This is accompanied by a weak, sharp peak at lower elds, which corresponds to the formation of btc radicals with g z 2.00. 22 This behaviour is consistent with a ligand-to-metal charge transfer (LMCT) mechanism, in which an electron is transferred from the photoexcited state of the linker to the metal sites (Fig. 5b). We also tested the photocatalytic activity of MIL-100(Ti) by irradiating a suspension of the solid in a mixture of H 2 -O : MeOH (4 : 1, v/v%) with a Xe lamp (300 W) and measuring the H 2 produced at different time intervals. As shown in Fig. 5c, MIL-100(Ti) shows an induction period of ca. 6 h for a total H 2 production of 1000 mmol g cat À1 aer 24 hours (see S6 † for experimental details). Stability was conrmed with PXRD and cyclability experiments for 5 consecutive cycles with minimum changes to the total H 2 production measured (Fig. S40 and S41 †). The generation of photoinduced charge separation in MIL-100(Ti), a hallmark feature of semiconductors, was conrmed with transient absorption spectroscopy (TAS). The decay kinetics prole of the solid suspended in ACN, measured at 500 nm, shows two different processes. A rst process that decays during the rst 500 ns, and a second one that remains stable in Ar or in the presence of MeOH for approximately 20 ms aer the light pulse (Fig. 5d). In addition, the longer-lived component is inhibited in the presence of O 2 suggesting that it likely originates from photogenerated Ti(III) species in the framework. This is further conrmed by the TAS spectrum recorded at 60 ns aer the light pulse, which shows an absorption band centred at ca. 640 nm, characteristic of Ti(III) species (Fig. S42 †). The intensity of this band decreases aer 6 ms showing a broad absorption between 400 and 600 nm, likely due to the decay of an important fraction of the photogenerated species.</p><!><p>Previous studies conrm the importance of choosing suitable titanium precursors for directing the synthesis of crystalline titanium frameworks. We are condent that the use of highthroughput synthetic methodologies might help in studying this reactivity more systematically in order to enlarge the pool of precursors at hand. This pushed us to investigate the synthesis of MIL-100(Ti) by using other Ti salts with variable reactivity in water. We replaced Ti 6 with precursors like Ti(O i Pr) 4 , Cp 2 TiCl 2 or CpTiCl 3 and introduced minor changes to the synthesis optimized for the cluster. Our results conrm that MIL-100(Ti) can be also prepared from organometallic precursors. However, Ti(O i Pr) 4 only afforded poorly crystalline solids for all the conditions explored (Fig. 2d), PXRD LeBail renement was used to conrm the phase purity of the solids formed with Cp 2 TiCl 2 and CpTiCl 3 . Full-prole ts yield excellent agreement factors and prole differences for both solids (Fig. S17 and S18 †). SEM analysis also conrms the formation of micrometric particles of sizes below 1 mm for Cp 2 TiCl 2 and CpTiCl 3 compared to the amorphous solid with ill-dened morphologies prepared with Ti(O i Pr) 4 (Fig. 2f-h, S12 and S13 †). Our experiments suggest that organometallic precursors are more adequate than the isopropoxide to induce the formation of crystalline MIL-100(Ti), likely due to their higher resistance against hydrolysis.</p><p>For a better understanding of the impact of the metal precursor over the properties of the MOF, we next analysed the porosity of MIL-100(Ti)-Cp 2 TiCl 2 with N 2 sorption experiments at 77 K. Compared to MIL-100(Ti)-Ti 6 , the solid displays a similar isotherm and pore size distributionindicative of the formation of the same open frameworkbut a substantial decrease in the BET surface area and pore volume to 1130 m 2 g À1 and 0.51 cm 3 g À1 (Fig. S23 and Table S3 †). We argued this could be linked to the presence of bulkier Cp units replacing OH À as capping ligands for a concomitant decrease in porosity. The ability of cyclopentadienyl ligands to induce the formation of Cp-capped titanium clusters was rst anticipated for COK-69. 20 To conrm this point we carried out FT-IR, 1 H-NMR and 13 This systematic exploration of conditions and metal precursors suggests that the formation of trimeric Ti 3 (m 3 -O) clusters units for the formation of MIL-100(Ti) under optimized reaction conditions is thermodynamically favoured not only for pre-formed clusters but also for organometallic complexes. However, the latter are less effective in preventing the formation of amorphous fractions of the material responsible for the observed reduction in porosity in MIL-100(Ti)-Cp 2 TiCl 2 . Overall, our experiments highlight the importance of controlling the reactivity of Ti precursors with water to enable the formation of crystalline, porous titanium-frameworks.</p><!><p>The design of new titanium-frameworks is arguably limited by the high reactivity of Ti(IV) with water and the subsequent difficulties in controlling the formation of persistent SBUs with the rigidity and directionality required to target specic architectures. We have shown how the use of high-throughput synthetic methodologies is useful to identify the conditions required to form Ti 3 (m 3 -O) metal-oxo clusters from different titanium precursors. This classical SBU can be then reticulated into a MIL-100 topology by direct reaction with trimesic acid to produce a mesoporous material with excellent thermal and hydrolytic stability, linked to the incorporation of highly charged Ti(IV) atoms. MIL-100(Ti) also inherits the photoactivity of other titanium-frameworks for the generation of reactive Ti(III) species with light upon LMCT. In this regard, the combination of chemical stability and photoactivity into a mesoporous framework, also amenable to the generation of vacant sites, offers a promising route for the design of advanced porous photocatalysts without the metric limitations imposed by the microporosity of the materials currently available. Based on the preliminary results, we are condent that this MOF will also be compatible with linker functionalization or metal doping for engineering its optical response with visible light. In the context of biomedical applications, the low-toxicity of titanium and the different degradation rates compared to MIL-100(Ti) might also be relevant for approaching targeted release of drugs.</p><!><p>Infrared spectra were recorded on an Agilent Cary 630 FTIR spectrometer equipped with an ATR module. Thermogravimetric analysis was carried out with a Mettler Toledo TGA/SDTA 851e/SF/1100 between 25 and 800 C at a rate of 5 C min À1 with a ow of N 2 : O 2 (4 : 1). PXRD patterns were collected in a PANalytical X'Pert PRO diffractometer using copper radiation (Cu K a ) with an X'Celerator detector, operating at 40 mA and 45 kV. Proles were collected in the 2 < 2q < 90 range with a step size of 0.013. Particle morphologies and dimensions were studied with a Hitachi S-4800 scanning electron microscope operating at 20 kV over metalized samples. Surface area, pore size and volume values were calculated from nitrogen adsorptiondesorption isotherms recorded at 77 K on a Micromeritics 3Flex apparatus. Samples were degassed overnight at 150 C and 10 À6 Torr prior to analysis. Surface areas were estimated according to the BET model and pore size dimensions were calculated using Non-Linear Density Functional Theory (NLDFT) models for the adsorption branch assuming a cylindrical pore model. 13 C-CP-MAS-NMR was carried out on a Bruker Avance III 400 WB Spectrometer. Samples were loaded in a 4 mm zirconia rotor and spun at 8 kHz. 1 H-NMR spectra were run on a Bruker DRX300 spectrometer. See ESI † for additional experimental details on digestion of the solids for NMR study.</p><p>Synthesis of [Ti 6 O 6 (O i Pr) 6 (4-tbbz) 6 ] (Ti 6 )</p><p>The synthesis of the Ti 6 cluster was carried out according to a previously reported procedure. 47 Synthesis of MIL-100(Ti)</p><p>A typical synthesis of MIL-100(Ti) was carried out by adding 7.2 mg of Ti 6 (24 mmol of Ti) and 25.0 mg of H 3 btc (120 mmol) to 3 mL of a mixture of ACN : THF (3 : 1, v/v%) in a glass vial. Subsequently, 250 mL of acetic acid (180 equiv.) were added and the mixture was sonicated to get a homogeneous suspension. The vial was placed in the oven at 160 C for 48 hours. Aer cooling down to room temperature, the white microcrystalline powder was recovered by centrifugation, rinsed with fresh DMF and MeOH and further washed by Soxhlet extraction with hot EtOH or MeOH for several hours. The solid was then allowed to dry under vacuum at room temperature.</p><p>Synthesis of MIL-100(Ti) from other Ti(IV) precursors (Cp 2 TiCl 2, CpTiCl 3 , and Ti(O i Pr) 4 )</p><p>The synthesis of MIL-100(Ti) from simple Ti(IV) precursors was carried out by adding 6.2 mg of Cp 2 TiCl 2 , 5.4 mg of CpTiCl 3 or 7.2 mL of Ti(O i Pr) 4 (24 mmol of Ti(IV)), and 25.0 mg of H 3 btc (120 mmol) to 3 mL of a mixture of ACN : THF (3 : 1, v/v%) in a glass vial. Subsequently, 500 mL of acetic acid were added and the mixture was sonicated to get a homogeneous suspension. The vial was placed in the oven at 160 C for 12 hours. Aer cooling down to room temperature, the white microcrystalline powder was recovered by centrifugation, rinsed with fresh DMF and MeOH and further washed by Soxhlet extraction with hot EtOH or MeOH for several hours. The solid was then allowed to dry under vacuum at room temperature.</p><!><p>There are no conicts to declare.</p>

Royal Society of Chemistry (RSC)

Spatial control of AMPK signaling at subcellular compartments

AMP-activated protein kinase (AMPK) is a master regulator of energy homeostasis that functions to restore the energy balance by phosphorylating its substrates during altered metabolic conditions. AMPK activity is tightly controlled by diverse regulators including its upstream kinases LKB1 and CaMKK2. Recent studies have also identified the localization of AMPK at different intracellular compartments as another key mechanism for regulating AMPK signaling in response to specific stimuli. This review discusses the AMPK signaling associated with different subcellular compartments, including lysosomes, endoplasmic reticulum, mitochondria, Golgi apparatus, nucleus, and cell junctions. Because altered AMPK signaling is associated with various pathologic conditions including cancer, targeting AMPK signaling in different subcellular compartments may present attractive therapeutic approaches for treatment of disease.

spatial_control_of_ampk_signaling_at_subcellular_compartments

Introduction<!>Lysosomes as a hub for glucose-starvation\xe2\x80\x93induced AMPK activation<!>Endoplasmic reticulum as a platform for calcium- and glucose-starvation\xe2\x80\x93induced AMPK activation<!>Regulation of mitochondrial and Golgi dynamics by AMPK<!>AMPK signaling in the nucleus<!>AMPK function at cell junctions<!>Conclusion and perspectives

<p>Sensing nutrient availability is a key feature of virtually all living organisms (Chantranupong et al., 2015; Efeyan et al., 2015). When nutrients are present in abundance, they are sensed, taken up, and utilized by cells to promote energy-consuming anabolic metabolism. In contrast, under conditions of nutrient scarcity, cells remodel their intracellular metabolic network towards energy-generating catabolic metabolism, with the ultimate goal of surviving under these adverse conditions. Maintaining the balance between anabolic and catabolic metabolism depends mainly on the availability of ATP, the cellular energy currency. In eukaryotes, ATP is generated primarily by oxidative phosphorylation and is utilized not only for building cellular macromolecules (such as fatty acids, nucleic acids, and proteins) but also for supporting many other cellular processes (such as nutrient transport and muscle contraction) (Vander Heiden et al., 2009). Correspondingly, eukaryotes have evolved intricate regulatory mechanisms to sense ATP levels and to modulate cellular metabolism in response to fluctuations in intracellular ATP levels, prominent among which is the AMP-activated protein kinase (AMPK) -mediated energy sensing mechanism (Hardie et al., 2012; Hardie et al., 2016; Herzig and Shaw, 2018).</p><p>AMPK is well recognized as a critical energy sensor and a master regulator of energy metabolism in eukaryotes. Being at the centre of many metabolic pathways, AMPK is tightly regulated by cellular energy status, which involves both its allosteric activation and its phosphorylation by upstream kinases. AMPK exists as a heterotrimeric complex consisting of AMPK-α, -β and -γ subunits (Figure 1). Mammalian cells possess two isoforms of AMPK-α, two isoforms of AMPK-β, and three isoforms of AMPK-γ (Ross et al., 2016b). The catalytic activity of AMPK resides in its α subunit, whereas both β and γ subunits serve as the regulatory components (Mihaylova and Shaw, 2011). In response to energy stress, an increase in cellular ADP and AMP levels as a result of ATP consumption triggers AMPK activation through direct binding of ADP or AMP to the AMPK-γ subunit (Oakhill et al., 2011). Although both ADP and AMP can bind to AMPK-γ to facilitate AMPK activation, changes in the AMP/ATP ratio are considered a more sensitive indicator of energy stress than the ADP/ATP ratio because newly generated ADP is rapidly converted to AMP by adenylate kinase; therefore, it is proposed that AMP is a true physiological trigger for AMPK activation during energy stress (Gowans et al., 2013). Based on the current model, AMP promotes AMPK activation by at least three mechanisms, which involve (i) promoting allosteric activation of AMPK, (ii) facilitating the phosphorylation of Thr172 located in the activation loop of the AMPK-α subunit by its upstream kinase liver kinase B1 (LKB1) (Gowans et al., 2013; Hawley et al., 2003; Hong et al., 2003; Shaw et al., 2004), and (iii) preventing AMPK Thr172 de-phosphorylation by protein phosphatases (Gowans et al., 2013; Li et al., 2015). Unlike AMP, ADP has only been shown to prevent AMPK dephosphorylation (Xiao et al., 2011). However to mimic the effects of AMP mediated protection against dephosphorylation, ADP has to be present in 10 fold higher concentration, suggesting AMP is a true metabolic regulator of AMPK (Gowans et al., 2013; Ross et al., 2016a).</p><p>Once activated by energy stress, AMPK then phosphorylates several downstream targets to restore energy balance by shifting from ATP-consuming anabolic processes (such as fatty acid synthesis and protein synthesis) to ATP-generating catabolic processes (such as glycolysis and autophagy) (Garcia and Shaw, 2017) (Figure 1). For example, in response to energy stress, AMPK activates autophagy, a catabolic process that recycles intracellular nutrients to maintain cell survival under nutrient-deprived conditions, through phosphorylating autophagy regulators such as ULK1 (Egan et al., 2011; Kim et al., 2011), while inhibits mechanistic target of rapamycin complex 1 (mTORC1)-mediated protein synthesis by phosphorylating Raptor (an mTORC1 component) and the TSC1–TSC2 complex (an upstream negative regulator of mTORC1) (Gwinn et al., 2008; Inoki et al., 2003).</p><p>In addition to energy-stress–mediated AMPK activation, AMPK can also be activated after a rise in intracellular calcium levels. However, LKB1 does not seem to be required for calcium-induced AMPK activation. Instead, calcium-induced AMPK activation was found to require calcium/calmodulin-dependent protein kinase kinase 2 (CaMKK2)-mediated Thr172 phosphorylation in AMPK without affecting cellular ADP or AMP levels (Hawley et al., 2005; Hurley et al., 2005; Woods et al., 2005). Physiologically, increases in cellular calcium levels and activation of CaMKK2-AMPK signaling have been reported during T-cell activation (Tamas et al., 2006), in hypothalamic neurons to regulate appetite (Anderson et al., 2008), and during the mechanosensitive assembly of actin to control cell adhesion and migration (Tojkander et al., 2018). Pathologically, aberrant CaMKK2-AMPK signaling promotes anoikis resistance and cancer metastasis (Jin et al., 2018; Sundararaman et al., 2016).</p><p>Apart from the canonical regulation of AMPK through allosteric activation and upstream kinase–mediated phosphorylation, AMPK function and its downstream signaling can be regulated by its protein stability, post-translational modifications (reviewed in (Garcia and Shaw, 2017; Jeon, 2016; Zungu et al., 2011)) and non-coding RNAs (Li et al., 2019b; Liu et al., 2016). Notably, recent studies indicate that AMPK localization at different subcellular compartments also serves to fine-tune its activation and downstream signaling activities in response to energy stress and other upstream stimuli. In this review, we will focus on this emerging theme on AMPK signaling, and discuss in detail the regulation of AMPK signaling at different subcellular compartments, including lysosomes, endoplasmic reticulum (ER), mitochondria, Golgi, nucleus, as well as cellular junctions.</p><!><p>Lysosomes, once viewed as the "trash bags" of cells, are now recognized as a critical cellular signaling hub (Appelqvist et al., 2013). Lysosomes are membrane-bound organelles with an acidic luminal environment that contain many enzymes including nucleases, hydrolases, and proteases to facilitate the digestion of macromolecules (Lawrence and Zoncu, 2019). Apart from its established role in macromolecule digestion, lysosomes have vital roles in various other cellular processes such as iron homeostasis, cholesterol transport, immune response, plasma membrane repair, and signaling regulation (Kurz et al., 2011; Lim and Zoncu, 2016). Recent studies have established that, similar to AMPK-mediated energy sensing, lysosomes also modulate cellular metabolism by sensing nutrient availability and triggering appropriate signaling responses (Settembre et al., 2013). For example, stimulation by nutrients, particularly amino acids, triggers a Rag GTPase-dependent signaling event that promotes mTORC1 localization on lysosomes, where mTORC1 is further activated by another GTPase, Rheb (Sabatini, 2017; Saxton and Sabatini, 2017). Once activated, in addition to phosphorylating its substrates involved in protein synthesis and autophagy, mTORC1 also phosphorylates the transcription factor EB (TFEB), a master transcriptional regulator of lysosomal biogenesis (Sardiello et al., 2009; Settembre et al., 2011), resulting in TFEB sequestration on lysosomes and inactivation of its transcriptional activity to regulate lysosomal biogenesis; nutrient starvation leads to mTORC1 inactivation, TFEB de-phosphorylation, and translocation to the nucleus to upregulate TFEB-mediated lysosomal biogenesis (Martina et al., 2012; Settembre et al., 2013; Settembre et al., 2012). Alterations in lysosomal function or biogenesis have been linked with various pathological conditions including lysosomal storage diseases, neurodegenerative diseases, and cancer (Fennelly and Amaravadi, 2017; Platt et al., 2018).</p><p>Recent studies have also linked lysosomes to energy-stress–induced AMPK activation (Carroll and Dunlop, 2017; Zhang et al., 2014; Zhang et al., 2013). The scaffolding protein axis inhibition protein (AXIN) was originally identified as a negative regulator of the Wnt pathway and is well established as a scaffolding component for Wnt signaling (Kikuchi, 1999; Zeng et al., 1997). Surprisingly, liver-specific knockdown of Axin in mouse led to compromised AMPK activation after fasting (Zhang et al., 2013). Subsequently, AXIN was shown to form a complex with AMPK and LKB1, and AXIN depletion prevents the formation of LKB1-AMPK complexes after glucose starvation. In support of the model that AMP is the main physiological regulator of AMPK activation, only the binding of AMP, but not ADP or ATP, to the AMPK-γ subunit promoted the interaction between AXIN and AMPK (Gowans et al., 2013; Zhang et al., 2013). These findings suggest that AMP has a role in assembling an AXIN-AMPK-LKB1 complex on lysosomes for energy-stress–induced AMPK activation (Lin and Hardie, 2018).</p><p>A subsequent study revealed that the v-ATPase-Ragulator complex, which is known to mediate amino-acid–induced mTORC1 activation on lysosomes (Sancak et al., 2010; Zoncu et al., 2011), is also critical for energy-stress–induced AMPK activation. Mechanistically, it was proposed that AMPK is constitutively localized on lysosomes and that, upon glucose starvation, lysosome-localized v-ATPase-Ragulator complex recruits the AXIN-LKB1 complex to the AMPK on lysosomes, resulting in LKB1-mediated phosphorylation and activation of AMPK (Figure 2A-2B) (Zhang et al., 2014). The v-ATPase-Ragulator complex also has critical roles in recruiting mTORC1 to lysosomes via Rag (Sancak et al., 2010; Zoncu et al., 2011). It was further shown that, while promoting AMPK activation in response to energy stress, AXIN binding to the v-ATPase-Ragulator complex releases mTORC1 from lysosomes, resulting in inhibition of mTORC1 activity under energy stress (Zhang et al., 2014). Of note, other studies also support the link between AMPK and lysosomes. For example, studies using a genetically encoded AMPK activity sensor confirmed lysosomal localization of enzymatically active AMPK and further revealed an increase in lysosome-associated AMPK activity after energy stress (Miyamoto et al., 2015b).</p><p>Metformin is a drug widely used to treat type 2 diabetes, and its anti-diabetic activities have been at least partly linked to its ability to activate AMPK (Foretz et al., 2014). However, exactly how metformin activates AMPK remains incompletely understood. Because metformin inhibits respiratory chain complex I in the mitochondria and suppresses ATP synthesis, it has been suggested that metformin activates AMPK via an increase in the cellular AMP/ATP ratio (Foretz et al., 2014). In contrast, others have reported that clinically recommended concentrations of metformin can activate AMPK without affecting cellular energy status, suggesting that metformin can activate AMPK through AMP-independent mechanisms (He and Wondisford, 2015). In support of this latter point, a recent study proposed that metformin can activate AMPK through the lysosome pathway without affecting cellular energy status (Zhang et al., 2016). It was shown that metformin failed to activate AMPK in AXIN-deficient mouse liver, and metformin was found to act on v-ATPase to facilitate tethering of AXIN-LKB1 to the lysosomes, leading to AMPK activation without compromising cellular energy status (Zhang et al., 2016).</p><p>Glucose provides the primary energy source to generate ATP in most cells (Vander Heiden et al., 2009); correspondingly, glucose starvation leads to rapid AMPK activation (Inoki et al., 2003). It is widely accepted that glucose starvation activates AMPK through the increased AMP/ATP ratio. However, short-term glucose deprivation can promote AMPK activation without affecting the cellular AMP/ATP ratio (Zhang et al., 2017). In addition, cells expressing a non-AMP-responsive AMPK-γ subunit (R531G mutant) still exhibited AMPK activation after short-term glucose deprivation, further strengthening an AMP-independent means of activating AMPK in this context (Zhang et al., 2017). As discussed above, AMP promotes the formation of AXIN-AMPK-LKB1 complexes and activates AMPK on lysosome membranes (Zhang et al., 2013). Surprisingly, AMPK activation was found to be compromised in AXIN- or LAMTOR-deficient cells after short-term glucose starvation (despite no obvious change in AMP or ATP levels), suggesting that AXIN-LKB1-mediated AMPK activation can be regulated by sensing glucose availability in addition to AMP (Zhang et al., 2017). This finding led to the hypothesis that either glucose itself or its downstream metabolites might serve as crucial regulators of AMPK activation. Further analyses revealed that fructose-1,6-bisphosphate (FBP), a glycolysis intermediate, has the effect of dissociating AXIN, LKB1, and AMPK from LAMTOR1, leading to compromised AMPK activation. Consistent with the model that FBP is utilized by aldolase during glycolysis, knockdown of aldolase resulted in enhanced AMPK activation even in the presence of glucose. Together, these findings suggest an intriguing model wherein FBP-unoccupied aldolase promotes v-ATPase-Ragulator-AXIN-LKB1 complex formation on lysosomes and AMPK activation in response to acute glucose starvation without obvious changes in the AMP/ATP or ADP/ATP ratio (Figure 2A-2B) (Zhang et al., 2017).</p><p>As AMPK can sense both cellular AMP (or ADP) levels and glucose availability and can also be localized at different subcellular compartments (which will be further discussed below), it has remained unclear how different signalling cues regulate differential activation of AMPK at different subcellular compartments. A recent study provides important insights into this question. It was shown that glucose starvation (before obviously decreasing AMP levels) specifically activates lysosomal but not cytosolic or mitochondrial pool of AMPK via aldolase-v-ATPase-Ragulator-AXIN-LKB1 complex (Figure 2B), resulting in phosphorylation of AMPK downstream substrates such as ACC2, Raptor, TSC2, HDAC4, SREBP1, and TBC1D1 to quickly shift the overall metabolic balance from anabolic processes to catabolic processes (Zong et al., 2019). A moderate increase in cellular AMP levels then promotes both lysosomal and cytosolic activation of AMPK, which was found to be independent of v-ATPase-Ragulator-containing complex on lysosomes but dependent on AXIN-mediated tethering of AMPK and LKB1. In contrast, under severe energy stress, high levels of intracellular AMP caused activation of cytosolic, lysosomal, and mitochondrial pools of AMPK, which was found to be independent of AXIN but dependent of LKB1 (Zong et al., 2019) (Figure 2C). Together, this study suggests that differential degrees of energy stress initiates different metabolic responses through regulating different compartmentalized pools of AMPK, therefore further supporting the importance of AMPK subcellular location in mediating AMPK signaling.</p><p>Notably, the tethering of AMPK to lysosomes is not only crucial for glucose-starvation–induced AMPK activation but also links AMPK to the regulation of lysosomal biogenesis through TFEB (Young et al., 2016). Functional analyses of AMPK in lineage specification of embryonic stem cells (ESCs) revealed that AMPK-deficient ESCs had impaired differentiation due to dysregulated lysosomal function and biogenesis. Mechanistic studies identified an AMPK-mTORC1-TFEB signaling axis wherein, under AMPK inactivation, mTORC1 promotes nuclear exclusion of TFEB and therefore inhibits lysosomal biogenesis (Young et al., 2016). Recent studies further validated the AMPK-TFEB signaling axis in the context of Kras-mediated lung cancer development. While Lkb1 deletion promotes Kras-driven lung cancer growth (Ji et al., 2007), surprisingly, Ampk deletion markedly inhibited Kras-dependent lung cancer in mouse models (Eichner et al., 2019). Further analyses revealed that AMPK promotes lung cancer development at least partly through regulating lysosomal gene expression mediated by Tfe3, another TFEB transcription factor (Eichner et al., 2019). Reciprocally, overexpression of TFEB in muscle cells has been shown to promote AMPK-mediated glucose uptake (Mansueto et al., 2017). In summary, lysosomes have emerged as a crucial hub to mediate AMPK signaling, not only by serving as a platform for AMPK activation in response to metabolic status alterations but also by providing nutrient-sensing capabilities for AMPK.</p><!><p>ER is a dynamic cellular organelle involved in protein folding and secretion (Benham, 2012). Proteins translocated into ER are subjected to correct folding and additional post-translational modifications with the help of ER lumen-localized chaperons (Schwarz and Blower, 2016). Apart from its central role in protein synthesis and trafficking, ER is also involved in numerous other cellular functions including lipid metabolism, calcium signaling, glycosylation, and phagocytosis (Desjardins, 2003; Helenius and Aebi, 2001; Mekahli et al., 2011). Physiological alterations such as imbalances in calcium homeostasis or protein synthesis and folding trigger adaptive responses, collectively called the ER-stress response or unfolded protein response (UPR), which aims to restore cellular homeostasis (Walter and Ron, 2011). Correspondingly, dysregulated UPR is associated with the pathogenesis of various diseases (Ozcan and Tabas, 2012).</p><p>Several previous studies have linked AMPK to ER. Analyses using probes for AMPK activity indicated that enzymatically active AMPK is present on ER (Miyamoto et al., 2015b; Qi et al., 2008). Other studies also suggested that AMPK can protect against ER-stress–induced apoptosis through various mechanisms (Amodio et al., 2018; Yang et al., 2013). AICAR-mediated AMPK activation has been shown to prevent ER stress through FOXO1-mediated upregulation of the ER chaperone ORP150 (Jung et al., 2018; Liu et al., 2018). Similarly, AMPK was also found to inhibit low-density lipoprotein (LDL) -induced ER stress (Dong et al., 2010). Activated AMPK has also been demonstrated to phosphorylate transcription factor SREBP-1c (localized on ER) and inhibit its proteolytic processing and nuclear translocation thereby resulting in the decreased fatty acid and cholesterol biosynthesis (Li et al., 2011). In addition to its role in ER stress, AMPK activity has also been linked to the regulation of ER morphology via phosphorylation of GTPase dynamin-related protein 1 (DRP1), a critical regulator of mitochondrial fission (Wikstrom et al., 2013). However, how ER is involved in regulating AMPK activation by upstream stimuli or kinases had remained unclear.</p><p>As discussed in a previous section, AMPK is activated by energy stress and increased intracellular calcium levels through different upstream kinases: whereas LKB1 is required for energy-stress–induced AMPK activation, CaMKK2 mainly mediates calcium-induced AMPK activation. Intriguingly, analogous to the role of lysosomes in energy-stress–mediated AMPK activation by LKB1 (Zhang et al., 2017; Zhang et al., 2013), recent studies also revealed that ER has a critical role in calcium-induced AMPK activation by CaMKK2, and this regulation requires stromal interaction molecule 2 (STIM2), an ER-resident transmembrane proteins that can potentially sense calcium levels in the ER lumen through their luminal EF-hand motifs (Chauhan et al., 2018). Based on this model, alterations of calcium levels in the ER lumen lead to conformational changes of STIM2 in its cytoplasmic portion, which promotes STIM2 interaction with both CaMKK2 and AMPK and therefore tethers CaMKK2 to phosphorylate AMPK, resulting in calcium-induced AMPK activation (Figure 2D). Interestingly, a recent study has identified both STIM1 and STIM2 as an AMPK substrate to regulate store operated calcium entry (SOCE), suggesting a reciprocal regulation between AMPK and STIM2 (Stein et al., 2019).</p><p>These findings suggest that AMPK uses ER as another docking site to specifically sense cellular calcium fluctuations and regulate its activation and downstream signaling in response to calcium stimulation. Therefore, this model is analogous to the afore-described model wherein energy stress induces AXIN-LKB1-AMPK complex formation on lysosomes, resulting in energy-stress–induced AMPK activation (Figure 2B); in both models, AXIN and STIM2 serve as scaffolding proteins to tether the corresponding kinases (LKB1 or CaMKK2) to phosphorylate AMPK on appropriate intracellular compartments (lysosomes or ER) in response to different stimuli (energy stress or calcium stimulation). The differences between these two models are that, as an ER resident transmembrane protein, STIM2 is sufficient to tether CaMKK2 and AMPK on ER; on the other hand, because AXIN is not a lysosome-resident protein, it requires additional support from lysosome-localized proteins such as v-ATPase and Ragulator to tether LKB1 and AMPK on lysosomes. Considering the versatile role of AMPK in cellular metabolism, it is tempting to speculate that mammalian cells may have evolved these mechanisms to tether AMPK on different intracellular compartments to sense different upstream stimuli and mediate different downstream signaling.</p><p>Interestingly, a recent study revealed that ER-lysosomal contact sites also play important roles in glucose-starvation–induced AMPK activation. As discussed in the previous section, acute glucose deprivation can be sensed by AMPK via interaction between FBP-unoccupied aldolase and the v-ATPase-Ragulator-AXIN-LKB1 complex located on lysosomes, although the exact mechanism regulating this interaction had remained elusive (Zhang et al., 2017). A follow-up study from the same group recently showed that under low glucose conditions, FBP-unoccupied aldolase interacts with the ER-localized transient receptor potential channels (TRPV) to inhibit its calcium release activity at ER-lysosome contact sites. It was proposed that the reduced calcium levels at ER-lysosome contact sites somehow promote the interaction between ER-localized TRPV and lysosome-localized v-ATPase, which subsequently recruits the AXIN-LKB1 complex to activate AMPK on lysosomes independently of AMP sensing (Li et al., 2019a) (Figure 2C). In support of this model, TRPV inactivation inhibited glucose-starvation–induced AMPK activation (Li et al., 2019a). Together, these recent studies suggest that ER plays important roles in both calcium- and glucose-starvation–induced AMPK activation: ER directly regulates calcium-induced AMPK activation, while indirectly controls glucose-starvation–induced AMPK activation through communicating with lysosomes.</p><!><p>In addition to lysosomes and ER, analyses of AMPK activity probes have revealed the association of enzymatically active AMPK with several other intracellular compartments, including mitochondria and Golgi (Miyamoto et al., 2015a; Miyamoto et al., 2015b). In this section, we discuss recent studies highlighting AMPK signaling at these compartments, with a focus on AMPK function in regulating organelle dynamics at these subcellular compartments.</p><p>AMPK has a central role in regulating mitochondrial homeostasis, including mitochondrial biogenesis, mitochondrial fission/fusion, and mitophagy (Herzig and Shaw, 2018). Enhanced mitochondrial biogenesis likely represents another important mechanism to support increased demand for ATP generation during energy stress. AMPK supports mitochondrial biogenesis at least partly by upregulating the expression of mitochondrial proteins via various transcriptional regulators such as peroxisome proliferator-activated receptor-γ (PPARγ) co-activator 1α (PGC-1α) (Jager et al., 2007; Rabinovitch et al., 2017). Mechanistically, AMPK has been shown to both directly phosphorylate PGC-1α, leading to its activation, and upregulate PGC1α expression at least partly by facilitating the nucleosome remodeling via phosphorylating DNA methyltransferase 1 (DNMT1), retinoblastoma binding protein 7 (RBBP7), and histone acetyltransferase 1 (HAT1) (Jager et al., 2007; Marin et al., 2017) (Figure 3).</p><p>Recent studies revealed that AMPK also regulates mitochondrial fission/fusion, a dynamic process involved in the maintenance of mitochondrial homeostasis (Herzig and Shaw, 2018; Youle and van der Bliek, 2012). Damaged or defective mitochondria can be toxic to cells because of their ability to generate reactive oxygen species (ROS) and to interfere with cellular metabolic pathways, and therefore need to be eliminated or repaired (Westermann, 2010; Youle and van der Bliek, 2012). Mitochondrial fission and fusion facilitate the clearance of damaged mitochondria by promoting mitophagy or utilization of its unaltered components by mitochondria fusion (Westermann, 2010). Members of the dynamin family of GTPases such as DRP1, mitofusin (Mfn), and optic atrophy 1 (OPA1) are central regulators of mitochondrial fission and fusion (Lee and Yoon, 2016). AMPK deficiency led to impaired mitochondrial fission in response to environmental insults that cause disruption of mitochondrial respiratory chain complexes, whereas pharmacologic activation of AMPK was sufficient to promote mitochondrial fission. Mechanistically, it was further revealed that under mitochondrial stress, AMPK phosphorylates mitochondria-localized mitochondrial fission factor (MFF) (Ducommun et al., 2015; Toyama et al., 2016), which upon phosphorylation promotes mitochondrial fission by recruiting membrane-remodelling GTPase DRP1 to the outer membrane of mitochondria (Kalia et al., 2018). In addition, a recent phosphoproteomic study identified armadillo repeat-containing protein 10 (ARMC10) as another AMPK substrate that localizes on mitochondria and regulates mitochondrial fission (Chen et al., 2019). Proteomic analysis of ARMC10-interacting proteins revealed that ARMC10 may promote mitochondrial fission by interacting with other proteins involved in regulating mitochondrial fission and mitophagy, including MFF (Chen et al., 2019). Both studies showed that depleting MFF (or ARMC10) or mutating MFF (or ARMC10) phosphorylation by AMPK attenuates AMPK-mediated mitochondrial fission, whereas overexpression of their phosphomimetic mutants is sufficient to promote mitochondrial fission even in the absence of AMPK activation, suggesting that AMPK promotes mitochondrial fission at least partly by phosphorylating MFF, ARMC10, or both (Chen et al., 2019; Toyama et al., 2016). Together, these studies identified MFF and ARMC10 as two downstream effectors of AMPK in regulating mitochondrial dynamics (Figure 3).</p><p>In the study discussed above, mitochondrial dynamics are controlled by AMPK localized outside of mitochondria (or localized on the mitochondrial surface). A recent study also revealed that AMPK can regulate mitochondrial function through its translocation into mitochondria. It was shown that ATP consumption during mitosis activates AMPK and causes its translocation into mitochondria, where AMPK subsequently phosphorylates mitochondrial calcium uniport (MCU), a calcium channel located in the inner mitochondrial membrane i.e. mitochondrial calcium uniport (MCU) (Zhao et al., 2019). Phosphorylated MCU promotes mitochondrial calcium entry from the cytosol, which in turn activates enzymes involved in cellular respiration (Figure 3). It was proposed that this mechanism serves to boost mitochondrial ATP production and restore energy balance to allow mitotic progression (Zhao et al., 2019).</p><p>In addition to mitochondria, AMPK activity has also been reported to regulate the dynamics of the Golgi apparatus. During mitosis, fragmentation of the Golgi apparatus is required for its distribution into daughter cells (Jackson, 2018). Recent studies revealed that AMPK is involved in regulating Golgi fragmentation during mitosis. Both energy stress and mitosis entry promote AMPK-mediated phosphorylation of Golgi-specific brefeldin A resistance factor 1 (GBF1) (Mao et al., 2013; Miyamoto et al., 2008). GBF1 functions as a guanine nucleotide exchange factor (GEF) for Arf GTPases, which regulate protein sorting and maintenance of Golgi apparatus (Bottanelli et al., 2017; Jackson, 2018). AMPK-mediated phosphorylation of GBF1 was shown to result in GBF1 release from Golgi, thereby preventing Arf1 effector recruitment to facilitate Golgi disassembly (Mao et al., 2013). Interestingly, CaMKK2, but not LKB1, has been proposed as the upstream kinase of AMPK to regulate Golgi fragmentation (Jackson, 2018). In summary, increasing evidence indicates that AMPK has an important role in regulating organelle dynamics in response to altered cellular conditions.</p><!><p>While AMPK has been shown to be localized on the membrane surface of various intracellular organelles, AMPK can also translocate into the nucleus. For example, both AMPK-α1 and -α2 isoforms were demonstrated to translocate into the nucleus under different contexts (Lamia et al., 2009; Salt et al., 1998; Vara-Ciruelos et al., 2018). Further analyses revealed that AMPK-α contains both nuclear localization signal in its N-terminal catalytic domain and nuclear export signal in its C-terminal regulatory domain to facilitate signal-dependent shuttling between the cytoplasm and the nucleus (Kazgan et al., 2010; Kodiha et al., 2007; Suzuki et al., 2007). Notably, LKB1, the major upstream kinase of AMPK, also exhibits nucleo-cytoplasmic shuttling (Dorfman and Macara, 2008), although it remains unclear whether nuclear-localized LKB1 plays a role in AMPK activation in the nucleus (or AMPK is first activated by LKB1 in the cytoplasm followed by AMPK translocation into the nucleus).</p><p>Current studies suggest a model that various signaling cues or pathological conditions modulate AMPK nuclear localization and subsequent transcription alterations. For instance, during myogenic differentiation, AMPK-α2-containing AMPK complex translocates into the nucleus to regulate the expression of PGC-1α, cytochrome C, and muscle creatine kinase (MCK) (Okamoto et al., 2019). Similarly, leptin-induced fatty acid oxidation is shown to be partly mediated by the nuclear translocation of AMPK and subsequent upregulation of PPARα (Suzuki et al., 2007). In patients or mice with Huntington's disease, AMPK-α1 exhibited enhanced nuclear localization, which correlated with increased huntingtin protein aggregation, neuronal loss, and brain atrophy (Ju et al., 2011). AMPK nuclear localization was also found to be inversely correlated with the nuclear protein levels of circadian component cryptochrome circadian regulator 1 (CRY1). Mechanistically, it was shown that AMPK-mediated phosphorylation of CRY1 causes its destabilization, therefore regulating circadian clock (Lamia et al., 2009).</p><p>As mentioned previously, in response to metabolic stress, AMPK activates autophagy by directly phosphorylating ULK1 and other autophagy regulators in the cytoplasm. Recent studies showed that AMPK-mediated autophagy activation in response to metabolic stress also involves transcriptional regulation of autophagy genes in the nucleus (Sakamaki et al., 2017). Bromodomain-containing protein 4 (BRD4), an epigenetic reader for acetylated histones, has been demonstrated to repress the transcription of autophagy genes, therefore maintaining basal autophagy under normal conditions. Following nutrient starvation, AMPK activation triggers histone H4K16 deacetylation by SIRT1. This displaces BRD4 from autophagy gene promoters, leading to their transcriptional activation (Sakamaki et al., 2017). Therefore, in response to nutrient stress, AMPK-mediated autophagy activation involves not only acute phosphorylation and activation of autophagy machinery but also chronic transcriptional regulation of autophagy genes. In addition to the nutrient stress, DNA double-strand breaks also promote the nuclear translocation and activation of AMPK. For example, DNA damage caused by the treatment of etoposide was shown to promote the nuclear translocation of AMPK-α1, which requires CaMKK2 but not LKB1 (Vara-Ciruelos et al., 2018).</p><p>Recent studies indicate that AMPK can also rewire the metabolic network by controlling epigenetic regulation of gene transcription in the nucleus. For example, AMPK phosphorylates histone H2B to regulate gene transcription in response to metabolic stress (Bungard et al., 2010). AMPK was also found to phosphorylate EZH2, a histone methyltransferase in the polycomb repressive complex 2 (PRC2), leading to disruption of interactions between EZH2 and other PRC2 components and attenuating PRC2-mediated methylation of histone H3 at Lys27 (Wan et al., 2018). Another recent study revealed that, in response to glucose starvation, AMPK phosphorylates and stabilizes TET2, an epigenetic enzyme that catalyzes the conversion of DNA 5-methylcytosine to 5-hydroxymethylcytosine, and that AMPK-mediated TET2 phosphorylation promotes TET2's tumour suppressive function (Wu et al., 2018). Collectively, these studies suggest a model that aside from its direct action on metabolic enzymes in the cytoplasm, AMPK can also translocate into the nucleus in response to signaling cues and regulate gene transcription through phosphorylating diverse transcriptional regulators (Figure 4).</p><!><p>Although cell junctions, such as tight junction or adherens junction, is clearly different from those classic intracellular organelles we have discussed in previous sections, broadly it can be considered a special subcellular location; therefore, in this section, we will briefly discuss AMPK signaling at cell junctions. LKB1 was initially shown to be able to promote polarization of mammalian intestinal epithelial cells (Baas et al., 2004), which is consistent with the notion that mammalian LKB1 is an orthologue of Par-4 polarity gene in C. elegans. This also raised the question of whether AMPK also plays a role in regulating cell polarity and cell junction assembly. Following studies showed that in polarized epithelial cells, E-cadherin promotes LKB1 localization to adherens junction, leading to subsequent AMPK activation (Sebbagh et al., 2009), although the exact role of AMPK in adherens junction remains less clear. Other studies showed that calcium-induced tight junction assembly and cell polarization in epithelial cells promote AMPK phosphorylation in a LKB1-dependent manner; calcium switch-induced tight junction formation is inhibited upon AMPK inactivation whereas AMPK activation by energy stress promotes tight junction assembly in epithelial cells (Zhang et al., 2006; Zheng and Cantley, 2007). Afadin protein, a component of the nectin-afadin complex involved in tight junction regulation, was later identified as an AMPK substrate, indicating that afadin is a potential downstream effector of AMPK in regulating tight junction assembly (Zhang et al., 2011). AMPK can also phosphorylate cingulin, a structural protein localized at the tight junction. Mechanistically, AMPK-mediated cingulin phosphorylation was shown to regulate cingulin binding to microtubules, and thereby facilitates the association of tight junctions with microtubules (Yano et al., 2013; Yano et al., 2018). Together, current studies indicate that cellular junctions are also important subcellular locations for regulating AMPK activity and function.</p><!><p>Investigations over the past two decades have put AMPK at the center of metabolic regulatory networks at both cellular and organismal levels. To achieve metabolic homeostasis during altered physiological conditions, AMPK acts not only on its direct metabolic targets but is also involved in cross-talks with other signaling pathways including PI3K-AKT, MAPK/ERK, and mTOR signalling (Hardie, 2014; Mihaylova and Shaw, 2011; Zhao et al., 2017). Owing to the central role of AMPK in cellular physiology, mammalian cells have evolved intricate regulatory mechanisms to govern AMPK activation, thereby enabling cells to appropriately sense and respond to diverse metabolic conditions. In addition to its canonical role in sensing energy status and regulating energy homeostasis, AMPK can also sense other signaling molecules such as calcium and mediate corresponding signaling events. It has long remained elusive how AMPK can sense different upstream stimuli and relay diverse signals to downstream pathways without miscommunications among different upstream stimuli and downstream pathways. Emerging evidence indicates that localization of AMPK to different subcellular compartments provides another dimension to regulate AMPK signaling. For example, calcium stimulation tethers CaMKK2 and AMPK on ER through the formation of the STIM2-CaMKK2-AMPK complex, resulting in AMPK phosphorylation and activation by CaMKK2 but not by LKB1. Considering the central role of ER in mediating calcium signaling, it makes perfect sense that this organelle has been selected to mediate AMPK activation by calcium stimulation. In contrast, current studies indicate that different compartmentalized pools of AMPK are involved in mediating cellular responses to glucose starvation and energy stress, depending on the severity of cellular stress: (i) glucose starvation, before resulting in any obvious increase in AMP levels, promotes v-ATPase-Ragulator-AXIN-LKB1 complex formation on lysosomes and AMPK activation through FBP-unoccupied aldolase, which allows AMPK to initiate downstream signaling to adapt to minor metabolic stress; (ii) further energy stress moderately increases AMP levels, leading to cytosolic AMPK activation in an AXIN-independent manner; (iii) severe energy stress dramatically increases AMP levels, resulting in activation of all compartmentalized pools of AMPK, including mitochondrial localized AMPK; (iv) energy stress also promotes AMPK translocation into the nucleus, although currently, it remains unclear whether nuclear translocation of AMPK is also affected by the severity of energy stress. Presumably, this would help cells adapt to long-term energy stress by regulating the transcription of genes involved in metabolic stress adaptation. In summary, current studies suggest that the spatial control of AMPK signalling allows cells to more precisely coordinate with energy and nutritional status and to achieve better metabolic adaption under different stress conditions.</p><p>These intriguing studies also raise several outstanding questions. Because AMPK can exist in possibly 12 complexes with the combinations of different subunit isoforms (including two AMPK-α isoforms, two AMPK-β isoforms, and three AMPK-γ isoforms), whether AMPK functions at different intracellular compartments are mediated by different complexes consisting of combinations of specific subunit isoforms remains an exciting question for future investigation. Addressing this question would require more precise analyses of subcellular localization of all AMPK subunit isoforms. Most current studies have focused on the spatial control of AMPK signaling in response to energy stress (given that AMPK's major function is to mediate energy stress response). Whether and how AMPK activation by other stress conditions (such as hypoxia (Emerling et al., 2009; Liu et al., 2006) can also be subjected to spatial regulation remains unclear and will be an interesting topic for future studies. In addition, how AMPK dynamically translocate among different subcellular compartments as well as cytoplasm in response to specific stimuli remains to be further studied. Likewise, although multiple AMPK substrates have been identified, not all substrates can be phosphorylated by AMPK under any given condition. The mechanisms underlying AMPK substrate preference still remain poorly understood. It is possible that AMPK phosphorylation of certain substrates requires specific bridging/accessory proteins which only come into play under specific metabolic stress conditions. In addition, it is possible that AMPK-mediated phosphorylation of additional substrates would only occur in specific subcellular compartments (for example, AMPK phosphorylation of H2B or TET2 would only occur under conditions in which AMPK localizes in the nucleus). Further studies are needed to understand this fascinating question. Finally, we envision that additional studies will identify additional AMPK substrates or downstream effectors that mediate AMPK function in regulating organelle dynamics. Altered AMPK signalling has been associated with various diseases including cancer and diabetes, making AMPK an attractive therapeutic target in treating these diseases (Cheng et al., 2016; Coughlan et al., 2014; Goodman et al., 2014; Hardie, 2013; Shackelford and Shaw, 2009; Steinberg and Kemp, 2009). A deeper understanding of the spatial control of AMPK signalling at diverse subcellular compartments may therefore reveal new therapeutic opportunities to target AMPK signaling for the treatment of relevant diseases.</p>

PubMed Author Manuscript

From \xe2\x80\x9cCellular\xe2\x80\x9d RNA to \xe2\x80\x9cSmart\xe2\x80\x9d RNA: Multiple Roles of RNA in Genome Stability and Beyond

Coding for proteins has been considered the main function of RNA since the \xe2\x80\x9ccentral dogma\xe2\x80\x9d of biology was proposed. The discovery of noncoding transcripts shed light on additional roles of RNA, ranging from the support of polypeptide synthesis, to the assembly of subnuclear structures, to gene expression modulation. Cellular RNA has therefore been recognized as a central player in often unanticipated biological processes, including genomic stability. This ever-expanding list of functions inspired us to think of RNA as a \xe2\x80\x9csmart\xe2\x80\x9d phone, which has replaced the older obsolete \xe2\x80\x9ccellular\xe2\x80\x9d phone. In this review, we summarize the last two decades of advances in research on the interface between RNA biology and genome stability. We start with an account of the emergence of noncoding RNA, and then we discuss the involvement of RNA in DNA damage signaling and repair, telomere maintenance, and genomic rearrangements. We continue with the depiction of single-molecule RNA detection techniques, and we conclude by illustrating the possibilities of RNA modulation in hopes of creating or improving new therapies. The widespread biological functions of RNA have made this molecule a reoccurring theme in basic and translational research, warranting it the transcendence from classically studied \xe2\x80\x9ccellular\xe2\x80\x9d RNA to \xe2\x80\x9csmart\xe2\x80\x9d RNA.

from_\xe2\x80\x9ccellular\xe2\x80\x9d_rna_to_\xe2\x80\x9csmart\xe2\x80\x9d_rna:_multiple_roles_of_rn

INTRODUCTION<!>Gene Hunting during the Genome Revolution<!>Discovery of RNA Dark Matter<!>Function Does Not Always Mean Protein-Coding<!>RNA INTERFERENCE PATHWAY<!>DICER, DROSHA, and DGCR8: Emerging New Roles in Transcription Regulation<!>DNA Damage Response (DDR) Cascade and the Moonlighting Functions of DDR Factors as RNA Binding Proteins<!>DNA Damage Response to DNA Double-Strand Breaks.<!>KU.<!>MRN.<!>PAR.<!>DNA Damage Mediators: MDC1, 53BP1.<!>DNA Damage Response Small ncRNA (DDRNA).<!>Damage-Induced Long ncRNA (dilncRNA).<!>Splicing, a Brief Overview<!>Reciprocal Interaction between DNA Damage and Splicing<!>RNA-TEMPLATED DNA REPAIR IN YEAST AND MAMMALS<!>Molecular Mechanisms<!>How Does RNA-Templated DSB Repair Work?<!>Involvement of NHEJ Mechanisms<!>Models of DSB Repair Mediated by RNA<!>GENOMIC REARRANGEMENTS AND RNA: LESSONS FROM CILIATES<!>Oxytricha<!>Paramecium and Tetrahymena<!>Epigenetic Inheritance in Ciliates, a Lesson from Plants<!>Consequences of Dysfunctional Telomeres<!>Telomere Repeat-Containing RNA<!>Other Noncoding Telomeric Transcripts<!>Importance of Intracellular RNA Detection in the DNA Damage Response Field<!>Detecting RNA in Fixed Cells<!>Multiplexed Detection.<!>Advances in Signal Amplification.<!>Detecting RNA in Living Cells.<!>Strategies for RNA Secondary Structure-Based Labeling<!>Strategies for Direct RNA Labeling<!>Other Strategies for RNA Detection<!>ANTISENSE OLIGONUCLEOTIDES AS LAB TOOLS AND THERAPEUTIC AGENTS<!>CONCLUSIONS AND PERSPECTIVES

<p>In 1973 the world's first mobile phone call was made, giving birth to the era of cellular phones. Gradually, features such as text messaging, cameras, games, and music were added to the devices, but until recently these additions were considered extras with respect to the original main purpose of voice calling. During the past few decades, the integration of novel technologies and unprecedented connectivity into mobile phones catalyzed the paradigm shift from "cellular" to "smart" phones. Smartphones completely transformed consumer perception of their mobile devices, gradually becoming a virtual toolbox with a solution for almost every need.</p><p>A similar shift in perception has occurred within the scientific community during the last 60 years, surpassing the original view of RNA in Crick's "central dogma" as solely the messenger of genetic information. Initial discoveries of noncoding RNAs (ncRNAs) having a biological function independent from protein coding included tRNA (tRNA), rRNA (rRNA), and spliceosomal RNA. Since then, the list of additional roles assigned to transcripts has grown exponentially. Although the vast majority of the genome is transcribed,1 current estimates indicate that only about 1.5% of it codes for proteins. It is now becoming evident that this nucleic acid is an extremely versatile molecule implicated in many different cellular processes, from structural support, to epigenetic modulation of gene expression, to maintenance of genome integrity. Consequently, numerous links between defects in noncoding RNA and human diseases have been described.2 Thus, RNA has broken free from its original confined role of subordinate messenger for DNA to emerge as an indispensable smart tool for a multitude of cellular needs.</p><p>In this review, we cover the very topical notion that RNA, both coding and noncoding, is involved in the maintenance of genomic stability as an example of "smart RNA". We begin with a historical perspective on the emergence of the noncoding functions of RNA and of the RNA interference machinery. We discuss novel aspects of a recently discovered class of RNAs involved in DNA damage signaling and DNA repair,3-7 as well as RNAs that guide genomic rearrangements8,9 and maintain telomere homeostasis.10-12 We then highlight the importance of using cutting-edge, single-molecule resolution techniques to study the location and biology of low-abundance, highly specialized RNA molecules. Finally, we discuss the exciting potential of targeting such RNAs with antisense tools as a viable therapeutic option. Throughout the article, we guide the reader to additional reviews that describe certain aspects in greater detail that, due to spatial constraints, we only mention.</p><p>Unlike the case of smartphones, the multiple functions of RNA are innate. In this respect, cellular RNA has always been smart, we just had to realize it.</p><!><p>The way we study biology has dramatically changed in the past 20 years due to the genome revolution. Genome sequencing efforts have been paralleled by efforts to identify encoded genes. Although the scientific community has long been aware of the many different functions of RNA (ranging from ribozymes to splicing), the search for novel genes was strongly biased toward the identification of protein-coding elements because of the assumption that they would be the main components of the newly sequenced genomes. Dogmas, such as "one gene, one enzyme"13 or "one gene, one polypeptide", have influenced the community far beyond their original message, long reinforcing a narrow protein-centric view of genetic information.</p><p>The quest for genes ended with the surprising finding that the human genome, as well as the genome of other vertebrates, encodes for only slightly more than 20,000 protein-coding genes.14,15 This small number was a big surprise due to previous estimates suggesting that the human genome would contain more than 100,000 genes,16 commensurate with the expectation for organisms such as humans that have a complex developmental body plan and central nervous system. For instance, the genome of Caenorhabditis elegans contains 19,000 protein-coding genes,17 while the fruit fly Drosophila melanogaster genome has 14,000.18 Thus, the number of protein-coding genes is similar in invertebrate and mammalian genomes. Next, the scientific community postulated that alternative splicing and combinatorial transcriptional control by transcription factors may contribute to human body complexity. However, these two phenomena are not sufficient to explain differences in complexity as they are also found in invertebrates.19 Nonetheless, mainstream genome analysis has continued to focus on protein-coding genes for years, mostly for practical reasons: they are relatively easy to identify given their high expression, long open reading frames, and defined start and termination points.20,21</p><!><p>In the quest for protein-coding genes, efforts have been conducted to identify expressed polyadenylated RNAs.21,22 In retrospect, the selection of cDNA that showed an open reading frame as evidence of protein-coding genes has caused a long delay in the discovery of human non-protein-coding transcripts longer than 200 nt, named long noncoding RNAs (lncRNAs)—this classification simply comes from experimental restrictions inherent to sequencing library preparation.</p><p>The concept that the genomes of organisms with high complexity are largely transcribed and that the main output is comprised of ncRNAs derives from analysis of mouse full-length cDNA collections and whole genome high-density tiling arrays, with a series of parallel studies identifying mounting evidence of widespread transcription. In 2002 (Figure 1), the FANTOM2 project found the first evidence of lncRNA transcription and also identified ~2500 antisense RNAs.20 Subsequently, it became clear that nuclear, nonpolyadenylated lncRNAs are the major output of the genome.23,24 In 2005 (Figure 1), the FANTOM3 project reached the conclusion that at least 63% of the genome is transcribed, most of which is comprised of ncRNAs.1 Additionally, by cap analysis gene expression (CAGE) technology,25,26 it has been shown that at least 73% of loci encoding for protein-coding genes are also transcribed from the antisense strand. This antisense transcription often results in products that regulate the activity of their respective sense mRNAs.27 Furthermore, one of the many early high-throughput RNA sequencing techniques, which are collectively named next generation sequencing, unexpectedly found that even retrotransposon element (RE) expression is tightly regulated in mammalian cells and tissues, producing lncRNAs28 which may in turn contribute to regulate RE expression. These and other studies also suggest that lncRNAs are somehow expressed at lower levels than protein-coding mRNAs, are often localized in the nucleus, and generally display time-, tissue-, or even cell-specific expression.</p><p>Despite initially being met with a large amount of healthy skepticism, the findings that pervasive transcription is one of the most abundant products of the genome have been supported by several prominent studies.29 For instance, in the ENCODE project (Figure 1), next generation CAGE RNA-sequencing confirmed that 62% of the human genome is transcribed and that the main output is noncoding RNAs.30,31 While experimental approaches clearly demonstrated that lncRNAs are indeed transcribed and display specific functions, the interpretation of their biological significance has proven difficult due to their low expression, stability, and conservation, relative to protein-coding RNAs.32 Future studies will be required to disentangle the multiple functions of RNA, which therefore deserves the title "smart" as proposed in our review.</p><p>Interestingly, the early identification of functional small RNAs, or those less than 200 nt long, was readily accepted by the community. This acceptance was so easily given likely due to the previous discovery of the RNA interference (RNAi) pathway (see section 3).</p><p>The discovery of siRNA and miRNA spurred the identification of numerous other classes of small RNAs featuring defined lengths and functions. piRNAs, for example, are 28–29 nt long sncRNAs that associate with Miwi, Mili, and Piwi proteins and function to repress the transcription and mobilization of RE, thus contributing to preserve genome integrity in the germline.89 While there is broad acceptance that piRNAs have a fundamental role in the repression of transposable elements, it is unclear why RE expression is not uniformly silenced, allowing for regulated RE activity during embryonic development,90 in embryonic stem cells, and in induced pluripotent stem cells.91</p><p>Among the vast landscape of small RNAs, some of them do not show a specific length and their interacting partners are less characterized. For example, small RNAs that overlap with transcription starting sites (TSSs), known as promoter associated RNAs (PASRs), and transcription termination sites (TTSs), named termination associated RNAs (TASRs), have been identified but poorly characterized. Some of them have been implicated in regulation of transcription, but much remains to be learned about their biology.92 One standing question regarding PASRs is to which extent they overlap with PROMPTs, a class of unstable RNA degraded by the exosome machinery.93</p><p>A large fraction of the small ncRNAs derives from processing of lncRNA precursors; thus, the regulation of lncRNA transcription and processing plays an important role in many aspects of small RNA biology. A good example is a novel class of small ncRNAs named DNA damage response RNAs, or DDRNAs, involved in the cellular response to DNA damage and in DNA repair (Figure 1).61,94 DDRNAs are DROSHA- and DICER-dependent products of damage-induced lncRNAs, or dilncRNAs, transcribed by RNA polymerase II (RNAP II) at the site of DNA damage95 (see section 4.2.2 for details).</p><p>Due to space constraints, not all known classes of short and long noncoding RNAs could have been discussed here, as the field has dramatically expanded in recent years.</p><!><p>The concept that the protein-coding-centric view cannot explain vertebrate/mammalian complexity came from the pioneering insights of John Mattick (Garvan Institute of Medical Resarch, Australia). By analyzing the noncoding content of all available genomes, Mattick noticed that the fraction of noncoding DNA in a genome increases progressively from ~30% in prokaryotes to a staggering 98.5% in humans. This observation is in line with a role for many noncoding regions, that are in fact largely transcribed, in fine-tuning protein production during development and in participating in tissue homeostasis maintenance in higher eukaryotes and in particular in mammals.19</p><p>In parallel, genetic studies also established that a large fraction of the genetic information falls outside the boundaries of the exons of protein-coding genes. Genome wide association studies (GWAS) have identified a plethora of single nucleotide polymorphism (SNPs) associated with a large variety of human non-Mendelian diseases, the majority of which falls outside protein-coding regions, often mapping to novel promoter elements and enhancers.96 Enhancer regions are sources of another class of lncRNAs called enhancers RNAs (eRNAs). Although their functions are not fully known, in some cases eRNAs are involved in transcription activation through chromatin looping.97 Further, recent studies have identified lncRNAs putatively involved with the molecular cause of some human diseases.98 Altogether, genetic evidence, together with sequence conservation at promoter or exons of lncRNAs, suggests that at least 19,000 human lncRNAs may be functional. A very recent study supports a cell-type specific role for lncRNAs in transcriptional activation.99 In contrast with many small RNAs, such as miRNAs, lncRNAs lack a common, unified function, therefore requiring more intensive investigations to address their activity.</p><p>Many lncRNAs are restricted to the nucleus, where some have structural roles, as in the case of paraspeckles, subnuclear bodies constituted by ncRNAs and proteins.100 Other nuclear lncRNAs are associated with chromatin where they contribute to regulation of the epigenome. A few examples include HOTAIR,101 XIST,102 and lncRNAs associated with imprinted loci. Novel technologies will be essential to map specific RNA-chromatin interactions and to dissect all the functions of chromatin-bound lncRNAs.</p><p>Antisense transcription is another essential source of ncRNA, impacting either positively or negatively on canonical gene expression.103 When antisense RNAs are exported to the cytoplasm, they can also regulate RNA stability27 as well as protein translation. For example, one peculiar class of antisense RNAs that act as translation regulators, SINEUPs, enhance the translation of the mRNAs they overlap through a SINE element embedded in the nonoverlapping part of the antisense.104,105 Curiously, various lncRNAs may act as SINEUPs, independent of their origin, suggesting that RNA structure, rather than its primary sequence, is most important.106</p><p>It is therefore becoming apparent that ncRNAs play diverse and important functions in the cell. An additional layer of complexity is given by the unexpected engagement of ncRNAs, and components of their pathways, in other fundamental mechanisms of the cell, such as the response to DNA damage.</p><!><p>RNAi, the process by which RNAs inhibit gene expression by sequence-specifically base-pairing with other RNAs, was initially described in plants and fungi as a peculiar yet effective mechanism to preserve genome integrity and protect against viruses and transposons.33,34 Later, RNAi was detected in a broad variety of other eukaryotic organisms35-39 and acknowledged as a more general strategy through which cells finely tune gene expression at the post-transcriptional level. By now, RNAi has been used for over two decades as a tool to study and manipulate gene function.</p><p>The first evidence that a long double-stranded RNA (dsRNA) was responsible for triggering sequence-specific silencing of a target gene was provided in 1998 by Andrew Fire and Craig Mello, who coined the term RNA interference.35 Shortly after, other groups proposed a different model for this phenomenon in which small RNAs, released by cleavage of long dsRNA precursors, were the actual effectors of the post-transcriptional gene silencing; hence, they named them short interfering RNAs (siRNAs).40-43</p><p>siRNAs are double-stranded RNA molecules, 20–25 base pairs in length, known to cause the degradation of the perfect complementary target RNA. siRNAs can be produced from RNA transcribed in the nucleus (endogenous siRNAs), or they can be virally derived or experimentally introduced as chemically synthesized dsRNA (exogenous siRNAs). Endogenous siRNAs have been described in plants and in C. elegans, and they can originate from overlapping sense and antisense transcripts44 or from repeat-associated genomic regions.45 To exert their function, siRNAs must be unwound and loaded into the RNA-induced silencing complex (RISC). The RISC complex contains Argonaute (AGO) proteins, which display the endonucleolytic activity responsible for cleavage of the target RNA. In mammals, there are four AGO proteins (1–4) that can participate to the RISC complex, but only AGO2 is catalytically active and functions for the direct degradation of the target mRNA.46 Only one of the two strands of the siRNA duplex (the guide strand) is loaded into the RISC complex, whereas the other strand, known as the passenger strand, is released and degraded.47 Endogenous siRNAs are thought to play an important role in defending genomes against transposable elements, as well as foreign nucleic acids, such as viruses.</p><p>Another class of endogenous small RNAs also capable of eliciting RNAi was discovered by Victor Ambros and colleagues in 1993, and they were later named microRNAs (miRNAs).48-51 miRNAs have been revealed to play important roles in almost every cellular process investigated.52 The biogenesis of most miRNAs requires the RNase III DICER and the Microprocessor complex, which is composed of the other RNase III DROSHA and the dsRNA binding protein DGCR8 (DiGeorge syndrome critical region 8).53 In the canonical pathway for miRNA biogenesis, an RNA polymerase II (RNAP II) dependent, single stranded, and capped primary RNA (pri-miRNA) is first processed by the Microprocessor complex in the nucleus, transforming it into a ~70-nucleotide hairpin-structured precursor RNA (pre-miRNA), which is then exported to the cytoplasm. Interestingly, some pre-miRNAs are produced from very short introns, called mirtrons, as a result of splicing and debranching,54 thereby bypassing the requirement of the Microprocessor complex. In the cytoplasm, cleavage by DICER, that works together with TRBP (transactivation-responsive RNA binding protein) and PACT (protein activator of PKR),55 results in a 20–23 nt miRNA duplex.56 Differently from siRNAs, miRNAs can exert their function either by triggering the degradation of the cognate mRNAs or by preventing their translation. Efficient mRNA targeting requires base-pairing of nucleotides 2 to 8 at the 5′ end of the miRNA, the so-called "seed region", with the target mRNA. The degree of complementarity between the seed region and target mRNA determines if silencing is induced through translational repression, the potential results of imperfect complementarity, or through cleavage, the result of perfect complementarity. As for siRNAs, the guide strand is preferentially incorporated into the RISC complex, which came to be known as "miRISC" following the discovery of its association with miRNA. In some cases, the passenger strand (designated as miRNA*) can also enter the miRISC complex to guide gene silencing. The miRISC complex also contains members of the GW182 (glycine-tryptophan protein of 182 kDa) family, which coordinate translational inhibition and the consequent mRNApoly(A)-tail shortening.57</p><p>Components of RISC are thought to localize and function just in the cytoplasm. However, in human cells RNAi has been demonstrated to mediate repression of target RNAs in the nucleus as well.58</p><!><p>A growing body of evidence has unveiled novel miRNA-independent functions for DICER and the Microprocessor complex, ranging from the maintenance of genome integrity to the modulation of alternative splicing.59-64 Here, we focus on the unanticipated roles played by the Microprocessor complex and DICER in the regulation of transcription with important implications in controlling genome stability, sometimes independently from small RNA generation.</p><p>The function of eukaryotic RNAP II is not limited to faithful copy of the information encoded in the genome, but it takes part in crosstalk with a myriad of other factors involved in the excision, addition, and editing of ribonucleotides in the nascent transcript.65 Among these factors, the Microprocessor complex and DICER turned out to be talkative interlocutors of the RNAP II machinery.</p><p>Microprocessor, initially found to be cotranscriptionally recruited to miRNA-encoding genomic loci,66 has been lately shown to localize at many different non-miRNA genes, including at superenhancers.67-70 In addition, the affinity of Microprocessor for hairpin structures in nascent RNA was demonstrated to be exploited by the cell to promote premature transcription termination of endogenous retroviral genes via stem-loop excision, a process independent from mature miRNA production.70 This Microprocessor-mediated cut inevitably generates an additional 3′-end in the nascent transcript, consequently providing an early potential alternative transcription termination site and thus inhibiting retroviral gene expression. Interestingly, accurate transcription termination of many miRNA-containing lncRNAs relies on Microprocessor endonucleolytic activity rather than the canonical cleavage-and-polyadenylation pathway.71 Moreover, human DROSHA has also been shown to enhance the expression of a subset of coding genes. Intriguingly, while transcriptional regulation exerted at these loci depends on DROSHA ability to interact with RNAP II, its catalytic activity is instead dispensable.67</p><p>It is now well established that many RNAi factors are not relegated to the cytoplasm, as initially proposed, but they are functionally active also in the nucleus of different eukaryotes, whereby they guide transcriptional gene silencing (TGS) through the deposition of repressive chromatin marks at silenced loci.58,72 In recent years, a number of reports revealed that Dicer also plays direct roles in nuclear transcriptional regulation. For example, it has been shown that S. pombe mutants lacking Dicer (Dcr1) failed to remove stalled RNAP II at sites of collision between transcription and DNA replication, resulting in accumulation of recombinogenic DNA–RNA hybrids 73-77 and consequent genome instability. Interestingly, the catalytic-dead Dcr-1 mutant was still able to release RNAP II from these loci, suggesting that in this system the ability of Dicer to promote transcriptional termination is independent from the biogenesis of sncRNAs, similarly to the nuclease-independent function of Drosha in transcriptional regulation described above. Nevertheless, Dcr1-mediated sncRNAs could be detected at these loci in wild type strains, though their physiological role remains elusive.78 The presence of Dicer in the nucleus of mammalian cells is instead a subject of debate. While murine DICER seems to be circumscribed to the cytoplasm,79 several reports demonstrated its presence in the nuclei of human cells.80-87 Hence, the functions played by DICER in human nuclei have started to be elucidated only recently. For example, human nuclear DICER was reported to localize to chromosomal regions with paused RNAP II, specifically in the proximity of transcription start sites and polyadenylation signals (PASs).84-86 Occasionally, these sites were found to be associated with R-loops—three-stranded nucleic acid structures composed of a DNA–RNA hybrid and the displaced single-stranded DNA—that triggered the transcription of antisense RNAs.85 The resulting dsRNA formation in turn could lead to the recruitment of DICER, together with other RNAi factors, and the consequent formation of heterochromatin at RNAP II-paused sites, ultimately enforcing TGS.85,86 The presence of DICER at specific PASs, besides its role in transcriptional repression, suggests it may also control alternative transcription termination since DICER-dependent deposition of repressive chromatin marks surrounding such PASs may decrease RNAP II speed, ultimately imposing altered transcription termination at these sites and the production of alternative RNA variants.84</p><p>The Microprocessor complex and DICER are versatile factors, acting, in addition to their canonical roles, as nuclear transcriptional fine-tuners. Importantly, the ability of DROSHA, DGCR8, and DICER to slow down the transcription rate and mitigate DNA–RNA hybrid accumulation, which represents an intrinsic threat for genome integrity,88 suggests a fail-safe mechanism in genome maintenance.</p><!><p>The integrity of our genome is constantly threatened by endogenous and exogenous agents.107 Cells have evolved a coordinated set of events to recognize the damage and promptly fix it, thus avoiding the replication and perpetuation of a compromised template. The DNA damage response (DDR) cascade is dependent on a broad variety of post-translational modifications such as phosphorylation, ubiquitination, sumoylation, poly(ADP-ribosylation), acetylation, and methylation.107-110 These modifications are recognized by specific protein domains, thereby orchestrating the recruitment of DDR factors to the DNA damage sites and ultimately the spreading of the signal throughout the cell. If the lesion cannot be repaired, persistent DDR activation may induce cell death by apoptosis or a permanent cell-cycle arrest called cellular senescence, both of which are known cellular intrinsic barriers to tumorigenesis.111</p><p>Until recently, the DDR signaling cascade was thought to consist entirely of proteins. The discovery of novel species of small ncRNAs directly implicated in upstream activation of the DDR in 201261,94 and of long ncRNAs induced at the site of DNA breaks in 201795 has radically changed this perspective (see section 4.2 for details). Moreover, other noncoding RNAs have been shown to serve as templates for DNA repair (see section 6 for details) or to guide genomic rearrangements (see section 7 for details). Interestingly, during the past decade, large-scale proteomic analyses and genome-wide screens have revealed that an unexpected proportion of RNA-binding proteins (RBPs) and proteins involved in transcription are involved in the DDR112 and that, conversely, factors originally discovered as guardians of genomic integrity show an affinity for RNA.113 Indeed the number of dual DNA- and RNA-binding proteins has surprisingly grown.114</p><p>Thus, the unexpected relationship between proteins involved in the DNA damage response and RNA may be important for the maintenance of genome stability.</p><!><p>Among the different types of lesions that may threaten our genome, DSBs are the most dangerous since they may lead to loss of genetic materials and chromosomal rearrangements, predisposing cells to malignant transformation. The repair of a DSB relies on either homology-dependent or -independent mechanisms.115,116 Homologous recombination (HR) is a homology-dependent and error-free mechanism that requires a homologous template, usually a sister chromatid, which allows accurate repair of postreplicative DSBs during S and G2 phases of the cell-cycle.117 In contrast, classical nonhomologous end joining (C-NHEJ) is a homology-independent mechanism active throughout the entire cell-cycle; although highly efficient, its very simple mechanism of basic religation, without proof-reading, makes NHEJ amenable to errors and thus to introduce mutations.118 NHEJ involves no or limited processing of DNA ends, while HR requires the formation of 3′ single-stranded overhangs. Thus, a critical step for the cellular choice between the two pathways is the DNA end resection.</p><p>DSBs are powerful activators of two large serine/threonine phosphatidylinositol 3-kinase-related kinases (PIIKKs): ataxia telangiectasia mutated (ATM) and DNA-dependent protein kinase (DNA-PK)119). Exposure of single-stranded DNA is instead recognized by a third PIKK named ataxia telangiectasia and Rad3-related (ATR).119 The recruitment of these apical kinases to the lesions leads to the local phosphorylation in cis of the histone variant H2AX at serine 139 (named γH2AX), a key step in the nucleation of the DDR. Following the first burst of H2AX phosphorylation, the ATM kinase phosphorylates many substrates, including mediator of DNA damage checkpoint 1 (MDC1) and p53 binding protein 1 (53BP1). These phosphorylation events fuel a positive feedback loop that facilitates the recruitment of additional ATM molecules to the DSB site120,121 and the spreading of γH2AX up to megabases away from the lesion. This signal amplification results in the accumulation of numerous copies of various DDR factors at and flanking the DSB, forming cytologically visible foci.</p><p>The localization of DDR factors to DSBs has been described as a two-phase process in which the initial recruitment occurs by the direct recognition of the DNA lesion in a γH2AX-independent manner, followed by accumulation of DDR proteins at the damaged site in a γH2AX-dependent manner.122,123 During the first phase, the DNA ends are promptly recognized by specialized factors: the KU70/KU80 heterodimer (or KU), the MRE11-RAD50-NBS1 (MRN) complex, or poly(ADP-ribose) polymerases (PARPs). A precise distinction of instances in which different DSB sensors are individually engaged, the timing of their recruitment, and whether they cooperate or compete with each other for the same DNA end are all issues that are only now becoming clearer. Likely, the nature of the break, the chromatin environment, the cell-cycle phase during the damaging event, and the cell type are all elements to take into account when attempting to answer these open questions.</p><!><p>The KU70/KU80 heterodimer (KU) is a ring-shaped complex that encloses the DNA ends and recruits the DNA-PK catalytic subunit (DNA-PKcs), which phosphorylates itself, H2AX, and other targets, thereby initiating the classical nonhomologous end joining (C-NHEJ) repair pathway.119 It has been reported that KU binds to the RNA component of telomerase both in yeast124-126 and in human cells127 (see section 8.1 for details about KU and telomere maintenance). KU has been found at promoter regions regulating gene expression128,129 and also in a complex with elongating RNAP II.130 Finally, it has been shown that treating pre-extracted cells with RNase A increases the detection by immunofluorescence of KU and other NHEJ factors at DSBs, indicating that a large fraction of these proteins are bound to RNA in the cell.131</p><!><p>The MRE11-RAD50-NBS1 (MRN) heterotrimeric complex plays important roles in detection and signaling of DSBs, as well as in initial processing of DNA ends prior to repair. The cohesin-like RAD50 protein interacts with MRE11 via its ATPase domains forming the globular head of the complex, and by dimerizing it ensures stable clamping and tethering of the complex to DNA ends.132 NBS1 interacts with MRE11 and is instead responsible for the nuclear localization of the complex and for the local recruitment of ATM to DSBs where it gets activated. MRE11 possesses 3′-to-5′ exonuclease and 5′ overhang endonuclease activities, which, together with the auxiliary endonuclease C-terminal binding protein interacting protein (CtIP), are essential for the initial steps of DNA end resection. In mammalian cells, the MRN complex has not been involved in the C-NHEJ; however, together with CtIP, it regulates alternative NHEJ (alt-NHEJ), which utilizes short microhomologies to direct repair.132</p><p>Given the distinct pathways in which they act, MRN and KU were considered mutually exclusive at DNA ends. However, by high resolution microscopy a certain degree of colocalization between KU and MRN at individual DNA ends was observed.131 Very recently, it has been shown that MRN can indeed access KU-blocked DNA ends by diffusion onto nucleosome-coated DNA.133 Excitingly, two independent studies reported that KU functions as a protein block stimulating yeast MRN-CtIP endonuclease cleavage in vitro.134,135</p><p>So far, there is no evidence suggesting that MRN is capable of binding RNA. However, MRN does appear to be involved in RNAP II transcription following DNA damage. Indeed, RAD50 has been found to interact with RNAP II upon UV-irradiation.136 Additionally, it has been demonstrated that all three components of the complex bind to RNAP II upon ionizing radiations and are important for damage-induced transcription at DSBs95 (see section 4.2.2 for details).</p><!><p>PARylation is the process by which poly(ADP-ribose) polymerases (PARPs) covalently attach (poly)ADP-ribose (PAR) units to Glu, Lys, or Asp residues of acceptor proteins or PARP itself. The activity of the three major PARPs, PARP1, PARP2 and PARP3, is induced by DNA damage: PARP1 is activated by single-strand breaks, DNA cross-links, stalled replication forks, and DSBs; PARP2 recognizes gaps and flap structures; PARP3 selectively responds to DSBs.137,138 PARPs are efficiently and transiently recruited to DSBs, for example during the first 5 min of laser microirradiation, where they have been proposed to orchestrate chromatin decondensation and the subsequent accessibility to the damage sites of a variety of factors, ranging from chromatin remodelers to transcription factors.139 Indeed PARPs are known to promote chromatin decompaction at promoters and to facilitate the loading of RNAP II machinery at transcription start sites.140</p><p>PAR is a nucleic acid-like molecule, and it can be recognized by RNA-binding domains such as the RNA recognition motif (RRM) and the RGG motifs, regions rich in arginines (R) and glycines (G) present in several RNA-binding proteins. Thus, it is not surprising that PAR and RNA can compete for the same RBPs; an example of this is the case of the RNA-binding protein NONO which is recruited to DSBs in a PAR-dependent manner.141 Therefore, changes in the local concentration of either of these molecules may dynamically alter the assembly of protein complexes, which, in turn, may affect cellular processes including the DDR.142</p><p>A good example of a factor involved in the DDR that can promiscuously bind to PAR chains and to RNA molecules is the heterochromatin protein HP1.143,144 HP1 has been shown to rapidly localize to sites of DNA damage in a PAR-dependent manner, subsequently being displaced and then slowly recruited again.145 Interestingly, HP1 requires an RNA component to bind to pericentric heterochromatin,146 and its hinge domain can bind both DNA and RNA.147</p><p>Another example is FUS/TLS (fused in sarcoma/translocated in liposarcoma), a member of the FET family of RNA/DNA binding proteins,148 which is a multifunctional factor with reported roles in splicing, transcription, mRNA export and translation, and the DDR. Initially identified as a fusion oncoprotein, FUS was later implicated in neurodegenerative diseases such as amyotrophic lateral sclerosis and frontotemporal lobar degeneration.149 In response to DNA damage, FUS is rapidly and transiently recruited to DSBs, likely through the interaction between its RGG domain and PAR.150-152 Interestingly, the same domain mediates the recruitment of FET proteins to paraspeckles by direct binding to the lncRNA NEAT1.153 In the absence of FUS, the localization of some DDR factors, such as 53BP1, to the site of damage is reduced and the efficiency of both HR and NHEJ is compromised.150,154 It has been shown that DSBs trigger local ncRNA transcription95 (see sections 4.2.2 for details). Given the ability of FUS to bind RNA, an exciting possibility is that FUS accumulation at DSBs could be modulated by RNA, in synergy or in competition with PAR chains.</p><p>The PAR-dependent localization to sites of DNA damage of several RNA binding factors occurs in two steps: beginning with a transient recruitment and ending with exclusion.141,150,155 It is therefore tempting to speculate that such dynamic behavior can also be mediated by the damage-induced lncRNAs (see section 4.2.2 for details), possibly due to waves of transcription correlating with the bimodal dynamics of chromatin relaxation and compaction at the site of break.</p><!><p>MDC1, which directly binds γH2AX, has surprisingly been found in an RNA interactome capture screen, an unbiased approach to identify protein–polyA RNA direct interactions.156 Interestingly, MOF (orthologue of Drosophila males absent on the first, or MYST1), a histone acetyltransferase shown to be important for MDC1 localization to DSBs, can also bind to ncRNA.157 Moreover, it has been observed that irradiation-induced MDC1 foci are reduced upon treatment with RNase A or in the absence of DICER or DROSHA61,158 (see section 4.2.1 for details).</p><p>Recruitment of 53BP1 to damaged DNA requires the presence of both monoubiquitinated H2A on lysine 15 (H2AK15ub) and dimethylated histone H4 on lysine 20 (H4K20me2).110 53BP1 associates with methylated histones through its tandem Tudor domain, which is usually found in RNA-binding proteins. Indeed, it has been shown that 53BP1 can be immunoprecipitated together with RNA molecules from cell lysates, and RNase A treatment in permeabilized living cells dissociates 53BP1 from IR-induced foci, which can reassemble in an RNA-dependent manner.61,158,159 Moreover, 53BP1, together with components of the C-NHEJ pathway, has been found in a complex with RNAP II in human cells160 (see section 6.3 for details). More recently, 53BP1 has been demonstrated to immunoprecipitate selectively with dilncRNAs and DDRNAs generated at sites of DSBs, in a manner dependent on its tandem Tudor domain95-(see section 4.2.2 for details).</p><p>These examples, though not an exhaustive list, are those that best point to the emerging evidence of an intimate and complex relationship between the DDR factors and RNA, and likely more is yet to come.</p><!><p>Emerging evidence suggests that noncanonical transcription, in the form of damage-induced small noncoding RNAs (sncRNAs), occurs at DNA damage sites.3-5,7,73,75 The existence of ncRNA species induced upon DNA damage and the involvement of RNA in DNA repair processes were originally reported in lower organisms such as yeast and fungi. For example, in Neurospora crassa, quelling and DNA damage-induced small RNAs (qiRNAs) are produced upon treatment with DNA damaging agents. qiRNA biogenesis involves a single-stranded precursor, called aberrant RNA (aRNA),161 which is converted into double-stranded RNA by RNA-dependent RNA polymerase (RdRP) activity and processed into small RNA through the same mechanisms that generate RNAi. qiRNA generation shares the same genetic requirements of the HR pathway; indeed, it depends on on replication protein A (RPA)161 and DNA replication.162 Mutations of genes involved in qiRNA biogenesis sensitize Neurospora strains to DNA damage.163</p><p>qiRNA seem to be mainly induced from repetitive or foreign sequences. Indeed they are transcribed from rDNA, but also of multiple copies of transgenes, acting as transgene-specific endogenous siRNA counteracting the expansion of selfish genetic elements.164 A similar phenomenon has also been described in rice where rDNA exposed to DNA damage locally generates high levels of RecQ DNA helicase- and RdRP-dependent double-stranded sncRNAs, which are required for cell viability after DNA damage exposure.165</p><p>Interestingly, in the yeast Saccharomyces cerevisiae, which lacks RNAi machinery,166 pre-existing RNA can serve as the template for DNA synthesis during repair of a chromosomal DSB and thereby mediate recombination167-169 (see section 6 for additional details). Another link that connects RNA to maintenance of genome integrity in S. cerevisiae lies in the RNA exonucleases Xrn1, Rrp6, and Trf4. These exonucleases were previously implicated in protecting genome stability from DNA–RNA hybrids and transcription-associated hyper-recombination170 but have more recently been shown to control also the activation of Mec1/ATR during DSB-induced DDR.171 While Xrn1 appears to be required for DNA end resection at the initial steps of HR, Rrp6 and Trf4 are dispensable for the resection process itself but are essential for replication protein A (RPA), a ubiquitous single-strand DNA binding protein, loading onto ssDNA.171 RPA affinity for ssDNA is very high; thus, it is interesting that factors related to RNA processing are important for this interaction. However, differently from factors involved in qiRNA production in Neurospora, yeast Xrn1, Rrp6, and Trf4 are not required for completion of later steps of HR repair.171</p><p>In D. melanogaster, it has been shown that although exposed DNA ends of a plasmid are sufficient to induce the generation of sncRNAs,172 they function to repress transcription of adjacent genes, rather than play a role in the DNA repair process.173 However, it was recently proposed that splicing factors may stimulate sncRNA generation at a DSB generated by CRISPR-Cas9 downstream of an intron in cultured Drosophila cells.174 Interestingly, the authors suggest that when RNAP II reaches the DNA end, the cotranscriptional spliceosome triggers a signal for the generation of an antisense transcript, potentially also stimulated by the formation of an R-loop, which then pairs with the sense transcript generating the dsRNA long precursor of the sncRNAs. They also hypothesize that a modification of the RNA polymerase complex may enable a strand switch and therefore allow for synthesis of a long RNA hairpin.174</p><p>Excitingly, different groups have established a direct link between DNA damage and the local generation of sncRNAs in mammalian cells. DDRNAs have the sequence of the damaged locus and are processed by the RNAi (see section 3.1 for details) machinery factors DROSHA and DICER61 (Figure 2). The key difference between DDRNAs and canonical miRNAs is that DDRNAs can carry virtually any genomic sequence, as they are generated where DNA damage occurs. DDRNAs appear to be required for the full activation of DDR signaling158 by mediating DDR foci assembly. In brief, DROSHA or DICER knockdown, but not the silencing of downstream RNAi effectors, impairs MDC1, the activated form of ATM and 53BP1 focal accumulation without affecting phosphorylation of H2AX.158 In agreement with these findings, it has been shown that DICER gets phosphorylated upon DNA damage and translocates to the nucleus where it associates with DSB sites, being necessary for full recruitment of 53BP1 and MDC1.87 A very recent work also confirmed that DROSHA and DICER, but not the silencing of downstream RNAi effectors, are necessary for 53BP1 focal accumulation and that DROSHA is involved in DNA repair by both HR and NHEJ.175</p><p>These results suggest that the focal concentration of diffusible DDR proteins can be regulated by sncRNAs. Indeed, the degradation of RNA by transient treatment with RNase A in a permeabilized living cell dissociates 53BP1, MDC1, and pATM from DNA damage sites.61,158,159 In a system in which a single DSB can be introduced in a traceable locus, RNase A treatment was sufficient to disassemble 53BP1 focus. Strikingly, upon incubation with RNA purified from cells damaged in parallel, but not from parental cells lacking the cleavable site, the 53BP1 focus reassembled. Similarly, incubation with total RNA extracted from cells in which DICER or DROSHA were silenced, or genetically inactivated, did not allow for DDR foci reformation following RNase A treatment.61,158 Together, these data indicate that DDRNAs contain the sequence of the damaged site and that their generation depends on DROSHA and DICER. Indeed, NGS approaches confirmed the DSB-induced production of sequence-specific DDRNAs, displaying a size consistent with DICER and DROSHA products. When chemically synthesized and reintroduced into RNaseA-treated cells, DDRNAs allowed site-specific DDR focus formation, demonstrating that they can function in trans and in the absence of mRNAs.61,158 Notably, DDRNAs are not required for the direct recognition of the DNA lesion, being instead stimulators of DDR foci assembly on γH2AX-decorated chromatin.158 It makes sense, then, that the early DDR step of NBS1 association to sites of DNA damage is not sensitive to global RNA degradation,158 similar to what has been described for the DNA damage sensor KU.131 It thus seems that the modification of chromatin (γH2AX) and the local synthesis of DDRNA are the two events required to form the large structures known as DDR foci.</p><p>Other studies have confirmed the requirement of DICER- and DROSHA-dependent sncRNAs for the recruitment of DDR factors involved in DNA repair, such as RAD51 and BRCA1, together with histone modifier enzymes such as methyltransferase MMSET (WHSC1) and the acetyltransferase Tip60/KAT5.176 Thus, sequence-specific sncRNAs may act as guiding molecules for the localization to and/or the activation of different utilities, such as for instance chromatin remodelers, at broken DNA ends.75</p><p>The existence of a class of similar 21 nt-long small RNAs, named DSB-induced RNAs (diRNAs), has been reported in Arabidopsis thaliana and in mammalian cells.94 diRNAs are induced by DSBs in an ATR-dependent manner, are transcribed from the vicinity of the DSBs by plant RNA polymerase IV, and play a role in the RNA-directed DNA methylation (RdDM) pathway.94 Differently from mammalian DDRNAs, diRNA biogenesis in plants requires not only DICER-like protein but also the activity of an RNA-dependent RNA polymerase, as well as AGO2.94 In human cells, diRNAs generated from the sequence surrounding the DSB were shown to control recruitment of RAD51 to damaged sites via a direct interaction between the diRNA-AGO2 complex and RAD51, thus promoting HR-mediated DNA repair events.177 In the proposed model, the diRNA-AGO2 complex anneals either to homologous broken DNA or to chromatin-bound transcripts originating from the target locus, suggesting that a homing mechanism via DNA:RNA or RNA:RNA paring may mediate the activity of diRNA in trans and influence DNA repair pathway choice. Another study in A. thaliana suggests that diRNAs do not act exclusively in HR-mediated repair but also play a role in NHEJ.178 Consistent with a direct role played by local transcripts in the process of NHEJ, it has been shown that RNAP II and nascent mRNA associate with factors of classical NHEJ and that RNA can serve as template for error-free DNA repair in mammalian cells160 (see section 6.3 for details).</p><p>Recently, the role of diRNAs and AGO2 in DNA repair has been challenged. By the use of CRISPR-Cas9 and TALEN technologies, it was shown that diRNAs are poorly induced upon DSB induction at endogenous genomic regions and that AGO2 inactivation does not affect HR in A. thaliana and in rice.179 These controversial observations indicate that we are far from fully understanding the biogenesis and functions of DNA damage-associated sncRNAs.</p><p>Although they appear to be part of the same phenomenon, DDRNA and diRNA present essential differences both in the process of their biogenesis and in their function. A first difference is that sequencing of diRNAs reveals that they are generated starting from a few hundred bases away from the DNA break.94 Given their above-mentioned involvement in DNA repair by HR, diRNAs might be produced starting where resection stops and dsDNA is left intact. On the other hand, sequencing showed that DDRNAs map very close to DNA ends.61 This difference may also suggest that diRNAs are in fact generated after the initial steps of DDR signaling. Another peculiarity of diRNA biogenesis is the dependency on ATR, which primarily responds to the exposure of single-strand DNA. Because these ATR-activating events occur mainly during resection or replicative stress, a model where diRNAs might be generated after or concomitantly to resection is also supported.</p><p>The abundance of diRNAs in plants appears to be significantly higher than in mammalian cells, possibly due to the presence of RdRP activity.94 Interestingly, high levels of pre-existing transcription of a transgene correlate with the generation of abundant diRNAs upon CRISPR/Cas9-induced DSBs.179 However, these highly abundant diRNAs seem to be dispensable for HR. A possible reconciliation model proposed by the authors is that "primary diRNAs" are low abundant and play a role in DSB repair, while "secondary diRNAs" are more abundant, require active transcription, are not directly involved in DSB repair, but may trigger post transcriptional gene silencing. Indeed, secondary diRNAs may be amplified via a "ping-pong"-like mechanism, where primary diRNAs cleave their complementary long transcripts, which, in turn, are converted by RdRPs into double-stranded RNAs and processed by DICER-like proteins to generate a new pool of diRNAs.</p><p>A similar ping-pong mechanism, by which small RNAs suppress neighboring gene expression, has been proposed in D. melanogaster. According to these findings, endogenous small interfering RNAs (endo-siRNAs) are produced from a transfected linearized plasmid, mimicking DNA ends of a genomic DSB.172,173 These data are in line with a role for break-derived sncRNAs in RNA quality control rather than DNA repair.</p><p>Whether these small RNAs originate from processing of pre-existing transcripts or from de novo transcription at sites of break in mammalian cells has been recently addressed and is discussed in depth in the next section.</p><!><p>Prompted by the discovery of DDRNAs (see section 4.2.1 for details), the group of d'Adda di Fagagna (IFOM, Italy) in collaboration with the group of Nils Walter (University of Michigan, USA) has more recently probed the transcriptional landscape around a DSB in search of DDRNA precursors. By single-molecule fluorescent in situ hybridization (smFISH, see section 9.1 for details) and reverse transcription followed by quantitative PCR (RT-qPCR) techniques, a novel class of lncRNAs named damage-induced lncRNAs (dilncRNAs) transcribed by RNAP II upon damage from and toward the DNA ends was uncovered.95,180 Induction of de novo transcription from DSBs was demonstrated in various mammalian cellular systems. For these experiments, multiple endonucleases were used to generate DSBs at exogenous integrated constructs as well as endogenous genomic loci, in both transcribed and non-transcribed regions. dilncRNAs were generated regardless of transcription state, suggesting independence from pre-existing transcription or canonical promoter and enhancer elements. In a similar study by the same team, dilncRNA induction was also shown at dysfunctional telomeres181 (see section 8.3 for details). In light of these observations, it is tempting to speculate that DSBs can themselves act as promoters; additional experiments, however, are needed to understand if the machinery needed for canonical RNAP II transcription, such as the preinitiation complex,182 is required also for dilncRNA generation.</p><p>The apical DNA damage sensor MRN (see section 4.1.1 for details) has been shown to be required for RNAP II localization to the damaged site and for subsequent dilncRNA transcription.95 Indeed, RNAP II immunoprecipitates with all three components of the MRN complex upon irradiation, although additional studies are needed to understand if the interaction is direct and through which domains and possibly modifications it is mediated. The role of MRN in the production of DDRNA/dilncRNA has been studied by knockdown experiments and by treatment with the small molecule mirin.95 Inhibition of MRN activity by mirin reduces DDR focus reformation when DDRNAs were exogenously added to RNaseA-treated cells.61 This could be because treatment with mirin inhibits dilncRNA synthesis,95 thus reducing the localization of DDRNAs to the site of damage. RNAP II transcription is known to be stimulated by nicks or, more strongly, by a DSB with a 3′-overhang.183,184 Since mirin inhibits both endo- and exonuclease activities of MRN,185 it is tempting to speculate that either, or both, activities are required for RNAP II transcription from the DNA ends. Another possibility is that the reported ability of MRN to unwind DNA ends186 is the step necessary to initiate RNAP II transcription from the DNA ends.</p><p>As discussed above (see section 4.1.1 for details), KU can bind RNA and RNAP II. Given the newly discovered role of MRN in damage-induced transcription, as well as the possibility of KU and MRN coexisting on the same DNA end131,133 and of KU to stimulate MRN activity,134,135 it will be interesting to determine the contribution, if any, of KU and its crosstalk with MRN in the production of ncRNAs at the site of DNA breaks.</p><p>In the proposed model (Figure 2), dilncRNAs divergent from and convergent to the DNA ends have the potential to pair and form a double-stranded RNA, which is processed by DROSHA, and then DICER, to generate DDRNAs. Accordingly, Michelini et al. demonstrated by qRT-PCR analyses that dilncRNAs accumulate in DROSHA-depleted damaged cells, while the products of DROSHA processing, called pre-DDRNAs, accumulate in the absence of DICER.95 In the same samples, DDRNAs are induced upon DNA damage and decrease when DROSHA or DICER are silenced. The latter result is also consistent with the characterization of telomeric DDRNAs181 (see section 8.3 for details).</p><p>The relevance of these ncRNA species in DDR signaling and in DNA repair comes from experiments preventing their transcription or their function. Indeed, a transient inhibition of RNAP II by small molecules, such as alpha-amanitin, prevents global DDR activation downstream of γH2AX and inhibits DNA repair.95 Excitingly, antisense oligonucleotides (ASOs) (see section 10 for details) against dilncRNAs and DDRNAs (Figure 2) are able to reduce 53BP1 accumulation and DNA repair at individual genomic loci with an unprecedented degree of specificity.95</p><p>By intracellular single molecule high resolution localization and counting (iSHiRLoC, see section 9.4 for details), fluorescently labeled DDRNAs localize to the damaged site through base-pairing with unprocessed dilncRNAs emerging from the DSB, and this interaction is fundamental to fully activate the DDR.95</p><p>These events are not unprecedented. Indeed in S. pombe, small RNAs generated by DICER bind to a nascent transcript, which is also their precursor, and together maintain the epigenetic and genetic stability of the centromeric locus.166 According to this so-called "nascent transcript" model, the unstable 2kb-long nascent transcript, synthesized by RNAP II preferably from one strand of the centromeric region,187,188 is converted to dsRNA by RdRPs or by pairing to an antisense transcript. This double-stranded RNA is then processed either by an RNAi pathway-dependent mechanism involving DICER or by RNAi pathway-independent mechanisms, such as the RNA degradation pathway of the Trf4/Air2/Mtr4 (TRAMP), and the exosome complexes.189 The resulting siRNA is loaded into the RITS (RNA-induced transcriptional silencing) complex, where the release of one of the two strands takes place generating mature Ago1-bound single-stranded siRNA.45 The base pairings between the mature small RNAs component of the RITS complex and the nascent transcripts, but not the underlying DNA, are central for the recruitment of enzymes responsible for H3K9 methylation, a repressive histone mark, of the centromeric locus, and enforcement of transcriptional gene silencing.190 Moreover, Ago1 slicing activity may contribute to the production of additional siRNAs and to the exhaustion of pericentromeric RNA in a self-sustaining loop. In this scenario, chromatin-associated nascent transcripts are not just the precursors of siRNAs, but they also act as local platforms for the coordinated assembly of chromatin remodelers guided by the siRNAs to the complementary target regions. A similar "nascent transcript" mechanism may also take place during the formation of a DDR focus. Michelini et al. showed that 53BP1 associates with in situ generated DDRNAs and dilncRNAs in a manner dependent on its Tudor domain.95 The localization of DDRNAs to the damaged site through base-pairing with nascent dilncRNAs may represent one of the mechanisms by which 53BP1 is selectively recruited to DSBs, and potentially a common mechanism for the recruitment of other DDR proteins. However, it will be important to investigate whether this interaction is direct, as well as to assess the possible role of Argonaute proteins in the DDR version of the "nascent transcript" model.</p><p>Several reports have shown that DSBs within a transcriptional unit suppress canonical gene expression,74,191 thus avoiding the transcription of a damaged template. This appears to be in contrast with the observed de novo transcription at DSBs. Once again, the literature on the S. pombe centromeric locus comes in handy. The apparent paradox of yeast cotranscriptional gene silencing, requiring a certain level of transcription to shut down transcription of specific genomic loci, has been recently solved. Indeed, the RNAi-mediated local concentration of chromatin remodelers and their residence time on the target sequence need to be above a certain threshold in order to switch off transcription and maintain the epigenetic marks.190,192 It is therefore possible that damage-induced ncRNA transcription is a similarly tightly regulated mechanism shaping the chromatin surrounding a DSB to induce the suppression of pre-existing gene expression. In S. pombe, it has been demonstrated that de novo transcription is induced at sites of DNA damage and that these newly synthesized RNA molecules anneal with their DNA templates resulting in transient DNA–RNA hybrids required for efficient DSB repair via HR.193 Notably, two very recent studies, exploiting the same endonuclease-based model system, acknowledged the presence of DNA–RNA hybrids at DBSs also in mammalian cells.175,194 In particular, it has been proposed that DROSHA is involved in the accumulation of DNA–RNA hybrids at DSBs175 and that Senataxin, a well characterized R-loop helicase, is recruited at DSBs induced in transcribed genomic regions, where it removes such DNA–RNA hybrids, promotes RAD51 loading, and prevents translocations.194 Future studies on this topic will have to take into account the existence of additional layers of complexity such as the kinetics of the events starting from DNA damage induction in a given cell type and cell-cycle phase. Indeed, the demand for novel approaches taking into account any heterogeneity in the cell population, such as single-cell resolution techniques (see section 9 for details), has become necessary among scientists that deal with quantitative analysis of siRNA-mediated epigenetic silencing.190</p><p>The discovery that each DDR focus relies not only on a common set of shared proteins but also on a set of RNA molecules generated in situ, that individually mark DDR events at distinct genomic loci, represents a leap forward in the understanding of the DDR pathways that may, in the future, be exploited for therapeutic purposes.</p><!><p>Splicing is a complex mechanism by which noncoding intronic sequences are precisely removed from the primary gene transcript (pre-mRNA) to generate a mature mRNA molecule, and its regulation is key in all aspects of cell physiology and pathology. Intron removal is carried out by a large molecular machine, the spliceosome, which is assembled on the pre-mRNA in a stepwise manner and is composed of five small nuclear ribonucleoparticles, named snRNPs U1, U2, U4, U5, and U6, and a large number of proteins.199 The spliceosome recognizes short sequence elements with a loose consensus at exon–intron boundaries (5′ and 3′ splice sites) as well as the branch point located near the 3′ splice site.200</p><p>The weak and dynamic interactions between the spliceosome and the pre-mRNA can be modulated by RNA binding proteins (RBPs), which associate to splicing regulatory sequence elements. These elements are particularly relevant for the selection of splice sites that deviate from the consensus sequences (weak sites) and either stimulate (intronic and exonic enhancers) or repress (intronic and exonic silencers) their recognition, thus affecting the splicing outcome. The list of RBPs involved in this regulation is continuously expanding and includes, but is not limited to, the serine/arginine (SR) family of splicing factors and a group of proteins that bind to heterogeneous nuclear RNA (hnRNP proteins). The partial degeneration of splice site sequences and the possibility to modulate their recognition through protein complexes assembled on enhancers and silencers allow for numerous events of alternative splicing to occur for each transcript. By using various combinations of 5′ and 3′ splice sites, and the respective regulatory proteins that bind them, alternative splicing (AS) is capable of generating different mRNAs from a single pre-mRNA.201 The vast majority (>90%) of human genes display AS events,202 which are modulated not only during development in a cell-type dependent manner but also in response to a wide range of stimuli or stressing conditions, including DNA damage.203-206</p><p>Splicing decisions may be modulated by chromatin organization, in particular nucleosome positioning207 and histone modifications,208-210 and by the elongation rate of RNAP II.201,211 The influence of this latter factor on splicing decisions stems from the fact that the assembly of the spliceosome occurs cotranscriptionally;212 thus, the elongation rate of RNAP II determines the time window available for a weak upstream splice site to interact with splicing factors before a competing stronger downstream splice site is transcribed.213 A key player in coordinating transcription with splicing is the CTD (C-terminal domain) of RNAP II that acts as landing pad for numerous splicing factors.214 The recruitment coupling model suggests that the phosphorylation status of the CTD, which is controlled by numerous factors, including DNA damage, determines the set of RBPs recruited to the transcriptional apparatus.215</p><p>By integrating different levels of regulatory events (chromatin organization, abundance of RBPs, post-translational modifications of RBPs and RNAP II), alternative splicing represents an ideal mechanism to finely tune gene expression in response to cell growth or stressing conditions, including DNA lesions.</p><!><p>It is now becoming apparent that a reciprocal interaction exists between DNA damage generation and the regulation of alternative splicing. The impact of DNA damage on splicing profiles has been addressed in detail in several excellent reviews.203-206,216 We will briefly discuss here only a few recent examples to illustrate how complex the interplay between DNA damage and regulation of AS can be.</p><p>A novel link between AS and the DDR has recently been discovered: detained introns (DIs), a new class of introns that exhibit delayed splicing.217 DI-containing transcripts are usually retained in the cell nucleus and form a reservoir of ready-to-use molecules, that, for example, can be called upon under conditions of impaired transcription. Notably, a subset of DIs, waiting in the nucleus for a signal, is spliced in response to DNA damage. Following DNA damage, a coordinated expression of specific splicing variants occurs, among which it is worth mentioning MDM4 and BCLAF1 that, respectively, control p53 and BRCA1 functions. The impact that DNA damage has on the splicing profile of BCLAF1 transcripts has important biological consequences. Indeed, in response to doxorubicin the pool of nuclear BCLAF1 transcripts containing DIs is halved while the level of protein-coding mRNA is up-regulated.217 Moreover, BCLAF1 protein is excluded, along with the splicing factor THRAP3 and RNAP II, from DNA damage sites in a process that depends on ATM activity.218 As a consequence, BCLAF1 protein is available to form a complex with BRCA1 phosphorylated by ATM. This complex recruits a number of splicing proteins, including Prp8, U2AF65, U2AF35, and SF3B, to a set of genes involved in DNA damage signaling and repair, thus connecting DDR signaling activation with cotranscriptional splicing and mRNA stability.219</p><p>Since DNA damage can control the splicing profile of genes involved in the DDR, it would be expected that splicing inhibition may play a role in the response to DNA damage. This hypothesis has been recently verified by showing that a short period of splicing inhibition prior to irradiation impairs IR-induced DNA damage foci formation.220,221 Furthermore, two natural compounds that affect the assembly of the spliceosome, namely the macrolide pladienolide B, which targets the splicing factor 3B subunit 1 (SF3B1) of the U2 snRNP,222 and the biflavonoid isoginkgetin, which prevents the recruitment of the U4/U6.U5 tri-snRNP,223 reduce ubiquitylation of damaged chromatin which is required for the assembly of DNA repair complexes. In particular, splicing inhibition impairs the recruitment to damaged sites of WRAP53β, RNF168, 53BP1, BRCA1, and RAD51, without affecting γH2AX and MDC1 signals, which are known to be recruited in a ubiquitin independent manner. This effect is due to the reduced expression of the short-lived E3 ubiquitin ligase RNF8,220 rather than a direct effect on DNA damage foci stability. Interestingly, the decreased expression of RNF8 partially explains the defective DNA repair observed after depletion of various splicing factors, thereby demonstrating the importance of splicing factors to genome stability.224</p><p>One of the best characterized examples of a splicing factor with a role in the DDR is the ubiquitin ligase PRP19, also known as Pso4 for Psoralen 4 gene, which is part of a large multiprotein complex comprising six additional subunits.225 PRP19/Pso4 acts at several levels of RNA metabolism (Figure 3): it modifies PRP3, a component of U4 snRNP, with a nonproteolytic ubiquitin chain that enhances protein–protein interactions and stabilizes the U4/U6.U5 complex;226 it interacts with RNAP II and recruits the TREX complex, which is involved in mRNA export, to transcribed genes;227 it forms a complex with U2AF65, which participates in the CTD-dependent coupling of splicing to transcription.</p><p>A large body of data implicates PRP19/Pso4 in the DDR (Figure 3). The PRP19/Pso4 was initially identified as an essential DNA repair factor in S. cerevisiae229 and it is one of the numerous human RBPs implicated in DNA repair.230 Accordingly, its down-regulation increases the sensitivity of human cells to spontaneous DSBs as well as to hydroxyurea or PARP inhibitor treatments. This may be related to the fact that PRP19/Pso4 colocalizes with the replication clamp PCNA both during unperturbed cell-cycle and in response to replication stress inducers such as hydroxyurea or camptothecin.231,232 In addition, PRP19/Pso4 participates in the interstand cross-link DNA repair pathway by interacting with Werner DNA helicase233 and in the transcription-coupled DNA repair pathway through association with Xeroderma pigmentosum group A (XPA) protein.234 PRP19/Pso4 also plays a role in the homologous recombination (HR) pathway by regulating the protein levels of BRCA1 and the generation of single-stranded DNA at DSBs.231 This latter function most likely involves the ability of PRP19/Pso4 to bind RPA-coated single-stranded DNA.232,235 Binding to RPA is required for PRP19/Pso4 localization to sites of DNA damage and for the ensuing RPA ubiquitylation, which facilitates the recruitment of ATRIP and the recovery of stalled replication forks. This mechanism shows strong similarities to what was previously described for DSB repair where ubiquitylation is required for γH2AX to act as a platform for the assembly of DDR complexes.236</p><p>PRP19/Pso4 may be one of the better characterized splicing factors that also plays a role in the DDR, but it is surely not the only one. Indeed, several proteins directly or indirectly involved in splicing associate with sites of DNA damage in a PAR-dependent manner, as mentioned above (see section 4.1.1 for details). In spite of this common feature, these RBPs participate in different DNA repair pathways. Thus, for instance SFPQ and NONO are two multifunctional DNA- and RNA-binding proteins involved in the catalytic step of the splicing reaction, in nuclear retention of defective RNAs and in DNA repair, stimulating NHEJ and repressing HR.141 Another RBP, RBMX/hnRNPG, implicated in tissue-specific regulation of gene transcription and alternative splicing, is a positive regulator of HR.224 However, the involvement of RBMX/hnRNPG in HR does not depend on its recruitment to sites of DNA damage but instead on its ability to control BRCA2 expression. Remarkably, some RBPs influence different steps of the assembly of repair foci. An example is FUS (see section 4.1.1 for details), whose depletion impairs the formation of DNA repair foci after treatment with topoisomerase II poison etoposide.154 Depletion of RBM14, another RBP, stabilizes γH2AX foci237 by reducing the recruitment of the NHEJ factors XRCC4 and XLF to damaged chromatin.238 The transient recruitment of the RBP hnRNPUL1 to DNA damage sites requires both the MRN complex239 and PARP1240 and is necessary for the full activation of the ATR signaling pathway. Moreover, hnRNPUL1 stimulates DSB resection and HR by promoting the association of the BLM helicase to DNA breaks.239 Transient association of RBPs with damaged areas appears to be a common theme. Indeed RBPs, including THRAP3, BCLAF1,218 hnRNPC, and hnRNPK,224 exhibit a prolonged exclusion from irradiated areas. Redistribution of these proteins requires both active transcription and the activity of PIKKs.155 Interestingly, inhibition of PIKK prevents displacement of RBPs from sites of damage and favors the formation of DNA–RNA hybrids, suggesting that the displacement is part of a general mechanism to prevent unwanted DNA–RNA hybrids.155</p><p>Overall, these examples reveal the existence of tight connections between splicing regulation, the assembly of DNA repair complexes, and the activation of checkpoint pathways. RBPs appear to have a central role in the coordination of all these events. However, the underlying molecular mechanisms are still a matter of investigation, and in particular it is unclear whether or not RNA molecules are involved in these dynamic processes.</p><!><p>RNA molecules synthesized during transcription are complementary to the DNA strand that served as their template. Early work demonstrated that RNA could play an indirect role in genome modification and DSB repair if converted into a DNA copy (cDNA) and stitched into damaged sites via NHEJ in yeast and mammalian cells.241-244 Not only can these cDNA molecules be inserted in a nonhomologous manner at sites of DSBs, but cDNA can also function as a homologous donor template to accurately repair DSBs via homologous recombination (HR) in budding yeast.169 However, can an RNA molecule serve directly as a template for repairing/modifying DNA without the need of being converted into cDNA?245,246 Indeed, RNA-containing DNA oligonucleotides can serve as templates for gene editing on plasmid or chromosomal DNA in Escherichia coli.247-249 Similarly, RNA-containing and RNA-only oligonucleotides can serve as RNA donor templates for DSB repair, a phenomenon observed in yeast and human cells.168,248 In addition, artificial long RNA templates injected in ciliate cells can guide genomic rearrangements250 (see section 7 for details). RNA-templated DNA modifications have been proposed to explain the high-frequency non-Mendelian loss of heterozygosity in rice.251 Moreover, cis- and trans-splicing mechanisms of chromosomal translocation suggest that chimeric RNAs generated by intergenic splicing may play a direct role to guide chromosomal rearrangements.252-256 A proof of concept that RNA transcripts are recombinogenic and can directly alter the genetic information in chromosomal DNA derives from experiments performed in budding yeast.169 Given these observations, the importance of RNA-templated repair becomes apparent.</p><!><p>Keskin et al. demonstrated that in S. cerevisiae an endogenous transcript can serve as template for repair of a chromosomal DSB in cis.169 The genetic assay was based on the antisense RNA-dependent repair of a nonfunctional histidine auxotrophic marker gene (his3). Briefly, an artificial intron (AI) is inserted in reverse orientation relative to his3, and antisense transcription is induced (Figure 4). While the AI cannot be spliced out of the sense his3 transcript, it can be spliced out of antisense transcript. Following the generation of a DSB inside the AI, the pre-existing his3 antisense transcripts is used as a template for HR, resulting in a functional HIS3 gene lacking the intronic sequence (Figure 4). While accurate DSB repair of his3 is seen in wild-type yeast cells by the formation of histidine prototrophic (His+) colonies, it is dependent on the reverse transcriptase (RT) activity of yeast retrotransposons, indicating that repair in wild-type cells proceeds through a cDNA intermediate. However, the inability to detect direct RNA-templated DSB repair in wild-type yeast cells may be due to a limitation of the assay used. Indeed, direct RNA-templated DSB repair in wild-type yeast cells is blocked by the function of ribonucleases H (RNase H1 and H2) that cleave the RNA strand of DNA–RNA hybrids. Once the activity of RNase H enzymes is removed, DSB repair is detectable even in the absence of the reverse transcriptase.169 These results demonstrate the existence of direct RNA-templated DSB repair.</p><p>Support for a direct RNA-templated DSB repair mechanism mediated by transcript RNA in cis is provided by the dependence on splicing of the antisense RNA. In fact, removal of the 5′-splice site (Figure 4) eliminates the formation of His+ colonies. Furthermore, sequencing data and Southern blot analysis support the accurate repair by cis-acting RNA, rather than ectopic integration of cDNA transcript from other regions of the yeast genome.169 Interestingly, even in the absence of the DSB, His+ colonies are still detectable.169 This finding suggests that the antisense RNA transcript can even modify DNA without induction of damage, possibly through spontaneous DSBs or nicks in the DNA. Overall, these results demonstrate that RNA can directly transfer genetic information to chromosomal DNA in cis with or without the induction of a DSB, revealing the existence of a mechanism in which genetic information can flow back from RNA to DNA, beyond the special case of reverse transcription postulated by the "central dogma" of molecular biology.243</p><!><p>Since RNA functions in cis as a donor template in DSB repair of his3 in the assay described above, the mechanism of DSB repair by RNA is HR. Instead, the sensitivity to RNase H activity indicates that DNA–RNA hybrids must form to transfer information from RNA to DNA. Previous work showed that the RecA recombinase of Escherichia coli can promote formation of DNA–RNA hybrids.257,258 Yeast RNA-templated DSB repair is strongly dependent on the recombinase Rad52, a fundamental protein in DNA repair by HR.169,259 However, knockout of the RAD52 gene, while reducing the frequency of DSB repair by RNA by a factor of 10, does not eliminate DSB repair by RNA, indicating that Rad52-independent RNA-templated DSB repair mechanisms do exist. These results in yeast are supported by in vitro experiments corroborating the ability of the Rad52 protein to catalyze the annealing of RNA to DNA.169 Recently, it was shown that purified yeast or human Rad52 protein can catalyze an inverse strand-exchange reaction with DNA or with RNA in vitro, a property not observed using the RecA homologue Rad51 recombinase or yeast Rad59, which is important for strand annealing.259 While RPA inhibits inverse strand exchange between two DNA molecules, it stimulates Rad52-mediated inverse strand exchange between DNA and RNA, possibly via protein–protein interaction with Rad52.259 Rad52 also promotes inverse RNA strand exchange with short-tailed or even blunt-ended double-stranded DNA. These results parallel in vivo studies demonstrating that RNA-templated DSB repair is stimulated by the overexpression of either yeast or human Rad52 N-terminal domain (NTD).259 Rad52 NTD retains the catalytic ability to promote inverse RNA strand exchange but lacks the Rad51 and RPA binding domains.259 Furthermore, null mutations of the RAD51 or RAD59 genes increased the frequency of DSB repair by RNA in yeast.169,259 This outcome is thought to occur by curbing the ability of DNA ends to recombine with sister chromatids, funneling repair to an RNA-templated pathway.169 Moreover, impairment of DNA end processing by defects in SAE2, EXO1, or MRE11 genes, which are important for DNA end resection following DSB, either increased or had no change in the frequency of DSB repair by RNA.259 These data support a model in which Rad52 catalyzes inverse strand exchange between RNA and a nonresected, or little-resected end of DNA at the DSB. RNA then guides break repair by bridging the broken DNA ends and is used as a template for DNA synthesis to fill the gap, a mechanism that could be mediated by cellular DNApolymerases.169,259 If resection is long, RNA-templated DNA repair may require reverse transcriptase for more extensive polymerization.</p><!><p>Recently, it has been found that C-NHEJ may play a role in RNA-mediated DSB repair. Following DSBs introduction via bleomycin or ionizing radiation (IR), RNAP II immunoprecipitated with various C-NHEJ and recombination proteins, including LigIV, XRCC4, KU-70, Polμ, DNA-PK, Rad51, and Rad52.160 Differently, alt-NHEJ proteins were absent or far less abundant in RNAP II complexes.160 The authors reasoned that C-NHEJ proteins may have a role in DSB repair in actively transcribed genes and explored this further. However, in this study, little information is provided on the roles of recombination proteins, which have previously been documented to function at DSBs in active genes.260 ChIP and quantitative PCR do indeed support the presence of C-NHEJ proteins (53BP1 and LigIV) at sites of DSBs in actively transcribed genes.160 Importantly, C-NHEJ components were found associated with nascent RNA transcripts by RNA-ChIP and this association significantly decreased following treatment of permeabilized cells with RNase H prior to RNA-ChIP,160 indicating the formation of DNA–RNA hybrids at DSB sites. This led the authors to suggest that C-NHEJ proteins may aid in an RNA-templated DNA repair mechanism. While RNA-donor oligonucleotides could repair a DSB in human cells in trans248 and an actively transcribed DNA could increase the frequency of end joining ligation of a linearized plasmid in human cells either directly or via RNA sequences in trans, it would be important to determine whether nascent pre-mRNA can template DSB repair in cis in mammalian cells. Following up on this possibility, in search of RNA-templated DNA polymerase activity, nuclear extracts of HEK-293 cells strikingly had the capability to copy an RNA template in vitro, independently of the major mammalian retrotransposon long interspersed elements (LINE1).160 This result highlights the possibility that cellular DNA polymerases may have some RT activity, as shown for yeast replicative polymerases,168 bacterial and archaeal polymerases,261 and some mammalian polymerases.262</p><!><p>Overall, these studies unveil an unexpected direct role of RNA in the DSB repair process: RNA may act as a template in repair of DSBs occurring in transcribed DNA.160,169,243,259,263,264 An HR model based on the results of experiments in S. cerevisiae suggests that a DSB occurring in an actively transcribed gene can be repaired in cis by the transcribed RNA as a bridging template for DNA repair. This process is aided by the inverse strand-exchange activity of Rad52 on dsDNA ends that have limited end resection (Figure 5A).</p><p>In cases of extensive resection, RNA-templated DSB repair could proceed with the aid of an RT. In addition, the RNA transcript can mediate DNA modifications in the absence of Rad52.169,264 Remarkably, RNA retains some ability to modify its DNA gene in cis even in the absence of an induced DSB. In this scenario, the RNA partially hybridized to DNA may form an R-loop structure with the intact dsDNA. The failure to remove R-loops from the DNA duplex leads to increase in DNA damage, recombination rates, mutation frequencies, and loss of heterozygosity.76,265 It is generally thought that the majority of R-loop-induced genomic instability stems from encounters between the DNA replication machinery and the altered chromatin environment in the vicinity of an R-loop.266,267 If a spontaneous or induced DSB occurs near the R-loop site, repair by C-NHEJ may occur, with the RNA facilitating end ligation by C-NHEJ proteins through end-bridging (Figure 5B). Thus, the RNA transcript could be a donor in DSB repair either to allow HR or to guide C-NHEJ, possibly depending on the cell-cycle phase, the types of DSB lesions, and the extent of DNA end resection.</p><!><p>Probably the most striking evidence for a physiological role of RNA in controlling genome stability is in ciliates. This is because in the ciliated protozoans sncRNAs have been shown to be involved in the epigenetic transmission of information between maternal nuclei and their derivatives, mediating large-scale genomic rearrangements and elimination or retention of specific DNA sequences.8</p><!><p>All ciliates, including the stichotrich Oxytricha trifallax, are characterized by nuclear dimorphism. These large unicellular ciliated protists contain two separate sources of genetic information: a transcriptionally silent germline micronucleus that is exchanged during matings, and a transcriptionally active somatic macronucleus containing tens of thousands of amplified gene-sized DNA molecules called "nanochromosomes" that are transcribed during asexual growth of the cells.268 These macronuclear nanochromosomes are the smallest known genomic DNA molecules in nature, with an average size of 3.2 kb, and are present at 100–100,000 copies per macronucleus.269,270 The micronuclear genome closely resembles that of a canonical eukaryotic genome with many genes organized on long chromosomes. However, micronuclear genes are typically interrupted by many short nongenic DNA sequences called internally eliminated sequences (IESs). For approximately 3,500 of these genes (~20% of genes in the Oxytricha genome), the macronuclear destined sequences (MDSs) that are connected upon IES removal exist in a nonlinear, scrambled order.271 When a mating occurs under the desired environmental conditions, two ciliates fuse and the process of macronuclear development from a newly acquired diploid micronucleus begins. At this point ciliates undergo a polytene chromosome stage (repeated rounds of micronuclear DNA replication without nuclear division, leading to large, banded chromosomes), eliminate more than 90% of their noncoding micronuclear germline genome (transposable elements, repetitive satellite sequences and IESs), fragment their chromosomes, and then sort and reorder the many thousands of nonlinear macronuclear destined sequences (MDSs) that will form functional genes. Ligation of MDSs, de novo telomere addition, and amplification of macronuclear nanochromosomes to the appropriate high copy number completes the development of a new, functional macronucleus (For general reviews of the process of macronuclear development in ciliates see refs 272-275).</p><p>Previous work has illustrated the roles of RNAs in mediating IES recognition/removal and the unscrambling events that ultimately take place during Oxytricha macronuclear development. Although the junctions of MDSs and IESs contain short direct repeat sequences that are likely involved (so-called "pointers"), they seem to act more as a structural requirement for unscrambling and DNA splicing, rather than for recognition by the necessary protein machinery.276 Instead, maternal guide RNA templates that are transcribed in the maternal macronucleus from the nanochromosomes have been hypothesized to mediate this massive genomic rearrangement process.277, Long noncoding sense and antisense RNA transcripts, corresponding to entire macronuclear DNA molecules, can be detected, peaking at 12–24 h postconjugation, and these are transported to the newly developing macronucleus to provide guide templates for the correct rearrangement, deletion, and sometimes inversion of the micronuclear DNA sequences250 (Figure 6). Microinjection of synthetic double-stranded DNA or RNA versions of alternatively rearranged nanochromosomes into the macronucleus of mating cells leads to changes in the reordering of MDSs, not only in the injected cells, but in offspring as well, suggesting epigenetic inheritance through these RNA templates.250</p><p>Recently, it has also been reported that Oxytricha produce and store RNA copies of whole somatic nanochromosomes during macronuclear development, which are derived from the maternal macronucleus before degradation. More than 60% of Oxytricha nanochromosomes have a corresponding RNA-cached copy, whose levels fluctuate throughout development, suggesting that not all developing macronuclear chromosomes undergo DNA rearrangements simultaneously.278 While extensive studies of the gene expression program during macronuclear development in Oxytricha have implicated hundreds of proteins playing roles during these developmental processes, much remains to be elucidated when it comes to biogenesis, processing, and function of sncRNAs.270,279</p><p>A novel class of macronuclear-derived 27 nt small RNAs, called 27macRNAs, that are highly upregulated after Oxytricha conjugation, peak at 24 h postmixing of complementary mating types.280,281 These 27mers are derived from the parental macronucleus as opposed to the micronucleus, have a strong 5′ U bias, and do not possess a 2′-O-CH3 group modification at their 3′ end, typical of certain classes of small RNAs in other ciliates.280-282 These 27macRNAs have been shown to associate with a PIWI homologue called Otiwi1 and specify which segments of micronuclear DNA will remain protected from degradation throughout macronuclear development. It has been suggested that this may occur through methylation and hydroxymethylation of cytosine residues within the DNA sequences to be eliminated.280,283 Indeed, microinjection experiments of 27 nt RNAs containing a 5′ U, corresponding to IES regions of the genome to be eliminated, lead to their retention after the completion of the macronuclear development program. However, the relationship between the PIWI-associated 27macRNAs and the long noncoding dsRNA "guide templates" implicated in MDS rearrangements remains unknown.</p><p>Ciliates have evolved two extraordinary genomes that demonstrate the complexity of epigenetic inheritance and DNA manipulation in eukaryotes. To date, although the general timing of events involved in macronuclear development has been fairly well characterized, the molecular mechanisms underlying many of these processing events remain poorly understood. Genome-wide studies and high throughput sequencing of mRNAs expressed throughout ciliate macronuclear development have allowed the identification of many factors likely playing roles in the numerous RNA-mediated processes occurring during this time. A disproportionate number of the genes identified as upregulated encode proteins that are involved in DNA and RNA metabolism processes, with the majority of these genes encoding evolutionarily conserved proteins in higher level eukaryotes. A recent study in Oxytricha shows that a striking number of differentially expressed macronuclear development genes in ciliates are preferentially expressed in animal germline cells, illustrating that ciliates possess a highly conserved and primordial set of factors involved in germline and stem cell maintenance.279 Thus, ciliates offer a unique and convenient system to study the influence of noncoding RNAs on genome integrity and transgenerational inheritance.</p><!><p>Macronuclear development has been more extensively studied in the distantly related ciliates Tetrahymena and Paramecium, where it has also been shown that epigenetic information from the parental macronucleus guides the elimination and retention of DNA sequences in the developing macronucleus. During the sexual life cycle of these ciliates, the entire parental micronuclear genome is transcribed bidirectionally to produce long, double-stranded RNAs early on in macronuclear development.284 In Paramecium, these double-stranded RNA precursors are cleaved by DICER-like enzymes DCL2 and DCL3, to produce a class of 25 nt small RNAs, called scan RNAs (scnRNAs),285-288 which are transported to the parental macronucleus where those with homologous macronuclear sequence are degraded. The remaining scnRNAs corresponding to micronuclear-specific sequences survive this filtering step and are transported to the developing macronucleus where, in association with PIWI proteins Ptiwi1/9, they "scan" the genome and mark IESs for excision and elimination.288,289 Notably, this is the opposite of Oxytricha, where PIWI-associated 27 nt piRNAs mark DNA sequences for retention. Although the mechanism of DNA excision and elimination requires further investigation, it has been shown to depend on a "domesticated" piggyBac transposase called PiggyMac.290-293 When IESs are excised from the developing macronuclear chromosomes, they have been shown to circularize, or concatamerize before circularization, depending on their size, to act as templates for the transcription of a second class of small RNAs called iesRNAs.293-295 Precursors of iesRNAs are processed by the DICER-like enzyme DCL5, to produce 22–31 nt small RNAs complementary to the sequence of excised IESs with a bias toward IES ends.288 A second class of sRNAs are also produced later in macronuclear development in Tetrahymena, but it remains unclear whether or not these late-scnRNAs play the same role as iesRNAs in Paramecium.296 While iesRNAs have more variation in length than scnRNAs and peak in expression much later during macronuclear development, both species possess a strong 5′ U bias. IesRNAs have been implicated in genome quality control and help to ensure the complete and precise removal of all remaining IESs matching these sequences from the amplified chromosomes (approximately 800n) in Paramecium, leading to a new, functional macronucleus288 (Figure 7). The general events of this "scnRNA model" also occur during Tetrahymena macronuclear development, although the specific details and associated factors may vary. It is worth noting that while Tetrahymena and Paramecium eliminate IESs during macronuclear development, these ciliates do not possess scrambled micronuclear genes such as the stichotrichs. In addition, their macronuclear chromosomes are much larger, coding for hundreds of genes instead of just one or two, typical of Oxytricha nanochromosomes.268,270</p><!><p>Although epigenetic inheritance has been well characterized and studied in ciliates, many questions still remain in the field. It is still poorly understood how the precursors to the different sRNA classes are initially transcribed and the processing machinery used to target particular genomic regions for either retention or elimination. Using the model organism A. thaliana for comparison, which displays another well studied system of epigenetic inheritance, some inferences can be drawn, although significant differences exist. In addition to the canonical RNAP II, A. thaliana possesses two additional nuclear multisubunit RNA polymerases, named RNA polymerase IV and RNA polymerase V, which play nonredundant roles in RNA-mediated gene-silencing.297 RNA polymerase IV is used to transcribe the precursors to siRNAs, while RNApolymerase V is responsible for transcribing nascent transcripts necessary for AGO-associated siRNA targeting. Like plants, ciliates have undergone whole genome duplications that have led to paralogous transcription subunits and machinery, including paralogs of the first and second largest subunits of RNAP II (RPB1 and RPB2), reminiscent of plant Pol IV and Pol V. In Oxytricha it has been shown that the RNAP II machinery paralogs play a role in development independent from transcribing sRNA precursors, but it is unclear if these separate paralogous RNAP II subunits, also upregulated in Paramecium and Tetrahymena, are associated with general transcriptional machinery.298 In addition, it is unknown how regions of the developing macronuclear genome are targeted for elimination or retention. In the field, this has been hypothesized to depend on production of nascent RNA transcripts at particular loci, but this has not been shown directly. Further studies are necessary to elucidate the underlying mechanisms responsible for genome rearrangements in ciliates, but with the power of next generation sequencing (NGS) of entire genomes and epigenomes, along with reverse genetic approaches, it will be possible to fill in the remaining gaps in our knowledge of these processes.</p><!><p>Telomeres are the distal tips of linear chromosomes, composed of short, guanosine-rich hexameric tandem repeats. In humans the 5′-TTAGGG-3′/3′-CCCTAA-5′ sequence is repeated approximately 2000 times to generate a telomere length between 10 and 15 kb pairs. Telomeres pose a particular conundrum for the cell due to the fact that they resemble the DNA end structure typically present at DSBs.299,300 Unlike DSBs, which must be repaired to ensure cell survival, telomeres actively inhibit DNA repair and DDR signaling. This is achieved by the recruitment of a specific set of proteins, collectively called "Shelterin", which directly inhibit DDR at telomeres, and by assuming a secondary structure, reinforced by Shelterin, called the t-loop, in which the tip of the telomere loops back on itself, thereby hiding the end.10-12,301-303</p><p>Despite the fact that telomeres are refractory to DNA repair activities, KU (see section 4.1.1 for details) is associated with normal telomeres. While C-NHEJ inhibition is achieved by the Shelterin component TRF2 that prevents the recruitment of Ligase IV from telomeres,301,304 the presence of KU seems to be important as a second line of inhibition of HR and alt-NHEJ pathways, in a Shelterin-free environment.305 Beyond DNA repair inhibition, localization of KU to telomeres could also be important for telomere length regulation through its binding to the RNA component of telomerase, TERC.127 This is an additional example of functional RNAs interacting with DNA repair proteins.</p><p>Telomeric "repair"—for instance a fusion of a telomeric chromosome end to another telomere or to an interstitial telomere repeat-containing region—can lead to genomic rearrangements, with consequent potential changes in ploidy, and eventually may contribute to cell transformation.306-309 Proliferating cells not expressing telomerase, or using other mechanisms of telomere elongation, eventually accumulates telomeres that are critically short, or "dysfunctional". Dysfunctional telomeres are akin to exposed DNA ends of DSBs and are promptly recognized by the DDR machinery.111,306,310,311 Telomere shortening can occur gradually over multiple population doublings as a result of the so-called "end-replication problem", in which the cell is unable to replicate all the way to the end of the telomere lagging strand. Shortening can also be due to the abrupt loss of telomeric material via DSB formation, potentially as a consequence of DNA replication stress. Critically short telomeres, as well as DSBs within telomeric repeats, trigger a DNA damage response that cannot support efficient DNA repair,301 provoking a protracted, likely permanent, DNA damage induced-checkpoint that arrests cell-cycle progression. The permanent cell-cycle arrest associated with the protracted DDR caused by unrepaired DNA ends312 is referred to as cellular senescence.111,313</p><p>Some cells can respond to telomeric shortening by de novo telomere elongation through telomerase via catalytic extension of the telomere, or HR, using other sources of telomeric material as templates for extension. Telomerase is a reverse transcriptase (TERT) that carries its own RNA template (TERC), and it is preferentially recruited to the shortest telomeres, presumably due to the absence of telomerase inhibitory proteins, to maintain their length and hence avert senescence.314-316 Telomerase is the primary means of telomere lengthening in stem cells, where it is expressed at low, but detectable, levels.317 Telomerase is also responsible for the maintenance of telomeres in approximately 85–90% of human malignancies.318 In most of the remaining tumor types, the HR-based Alternative Lengthening of Telomeres (ALT) mechanism is activated, which relies on a form of HR between a chromosomal telomere and other telomeric material for elongation. However, some reports have recently challenged the idea that an essential feature of cancer is the acquisition of a telomere maintenance mechanism. Some patient-derived melanoma and neuroblastoma cells do not express telomerase nor activate the ALT mechanism, and indeed, their telomeres shorten during serial passages in culture. The phenotype associated with these cancer cells has been referred to as ever-shorter telomeres.319,320 In support of this notion, bioinformatics analysis of a large cohort of human tumors (18,430 samples) has recently reported that approximately a fifth of the analyzed samples neither expressed telomerase nor harbored alterations in ATRX or DAXX genes, which are commonly mutated in ALT.321 These new findings add an additional layer of complexity for cancer treatment because they suggest that, at least in some cases, prevention of telomere shortening is not required for oncogenesis nor for cancer progression, thus potentially blunting therapeutic approaches targeting telomere maintenance mechanisms.</p><p>Recent studies from multiple laboratories have shown that the regulation of ncRNA transcribed from telomeric regions plays an important role at damaged/shortened telomeres to promote activation of the DDR and, hence, their repair.181,322-326 In addition to chromosome ends, telomeric ncRNA associates with multiple nontelomeric loci to stimulate transcription.327 Therefore, the regulation of ncRNA at telomeres may be a critical determinant with regard to the rate at which a cell enters a state of replicative senescence.</p><!><p>Telomere Repeat-Containing RNA (TERRA) is one example of a lncRNA harboring telomeric repeat sequences.12,328 TERRA transcription is initiated in the subtelomeric region and continues into the telomeric repeats, although it is unlikely to reach chromosome ends.329-332 Recently, there have been important observations describing how TERRA is regulated at different telomeric states.12 TERRA levels are tightly regulated with respect to cell-cycle. In G1 the levels are maintained low, but at the G1/S transition TERRA is transiently upregulated and then subsequently degraded as the cells progress through the S phase and into G2.326,333,334 In the yeast S. cerevisiae, the degradation throughout the S phase is carried out by the 5′ to 3′ RNA exonuclease Rat1 so that TERRA is removed at approximately the time when telomeres are replicated by DNApolymerase and/or extended by telomerase. When telomeres become short, this precise regulation is altered due to the inability of Rat1 to associate with short telomeres326 (Figure 8).</p><p>In S. cerevisiae, TERRA produced from a critically short telomere has the ability to associate with telomerase in the nucleoplasm, which is then recruited specifically to the shortened telomere where the TERRA molecule was produced, presumably to promote telomerase-mediated elongation.324 Consistently, in S. pombe it was demonstrated that polyadenylated TERRA levels increase upon telomere shortening, and these TERRA molecules, in turn, associate with the telomerase enzyme.323 Moreover, in S. pombe experimentally induced expression of TERRA from a single telomere results in telomerase-mediated elongation exclusively at the telomere overproducing TERRA. Similar experiments performed in human cells have also demonstrated that TERRA and telomerase interact in cell extracts.327,335 Taken together it appears that TERRA may function as a S.O.S. signal of sorts at short telomeres, to fetch and direct telomerase to the telomere in need of elongation336 (Figure 8, bottom right).</p><p>Since TERRA degradation is impaired specifically at short telomeres in late S phase (when telomerase acts), this may increase the chances that a productive TERRA-telomerase interaction occurs to facilitate elongation of the short telomere. In future studies, it will be important to interrogate the relationship between TERRA cell-cycle regulation, telomere replication, and telomerase in more depth. Furthermore, it remains unresolved as to how TERRA may help to direct telomerase to the "right" telomere. Is it the subtelomeric sequence information that is important to form DNA–RNA hybrids, or are RNA–protein interactions the key? There are also inconsistencies that remain to be addressed, such as the observation that the overexpression of TERRA in S. pombe leads to telomere elongation323 while in S. cerevisiae this leads to telomere shortening,331,337 although the discrepancy may be due to different expression levels in the different model systems. It is also not understood why the impairment of the RNA exonuclease Rat1 or the NAD-dependent deacetylase Sir2, in budding yeast, does not lead to telomere lengthening, despite high TERRA levels in these mutants. Although these open questions remain, there is ample evidence implicating TERRA as an important intermediate to promote the telomerase-mediated repair of shortened telomeres.</p><p>Another important feature of TERRA is that it can form DNA–RNA hybrids, which likely lead to the formation of R-loops.322,325,338,339 Similar to TERRA levels, telomeric R-loops are regulated in an identical cell-cycle dependent manner. RNase H2, which degrades the RNA moiety of an DNA–RNA hybrid, gets recruited to telomeres, approximately at the time of telomere replication, and promotes R-loop removal.326 In a manner once again reminiscent of the regulation of TERRA levels, TERRA R-loops are no longer degraded in a timely manner when telomeres become critically short, due to the inability of RNase H2 to properly localize to shortened telomeres (Figure 8). This delay in R-loop resolution likely results in an encounter between the replication machinery and R-loops, an event that triggers HR.326 The persistence of TERRA R-loops is an important feature of critically short telomeres in the absence of telomerase. If R-loops are removed, through RNase H1 overexpression, cells enter replicative senescence at an accelerated rate and the HR machinery fails to associate with short telomeres.325,326 In contrast, when R-loops are allowed to accumulate at telomeres, through RNase H2 deletion, the rate of senescence is significantly reduced. Therefore, at normal length telomeres, TERRA and its R-loops are produced at each cell-cycle at the G1/S transition, only to be degraded, which likely facilitates replication passage and does not promote elongation by telomerase or HR. When a telomere is damaged/shortened, the G1/S up-regulation of TERRA occurs in a timely manner but the subsequent degradation is rendered defective, so that TERRA and R-loops persist into late S phase and promote telomerase and HR-mediated elongation, respectively (Figure 8). It will be interesting to determine the significance of G1/S-specific TERRA up-regulation, as it may occur as a precautionary measure in the case of telomere shortening, to regulate telomeric replication origins or to regulate gene expression elsewhere in the genome.327 In human cells TERRA R-loops may perform similar functions, as DNA–RNA hybrids are enriched at telomeres in human Immunodeficiency, Centromeric instability and Facial anomalies (ICF) patient cells, which have extremely short telomeres and are largely responsible for DDR activation at chromosome ends.322</p><p>Telomeric R-loops are also important beyond their role during replicative senescence. Cancer cells that employ the HR-based ALT mechanism show increased TERRA levels and telomeric R-loops. The overexpression of RNase H1 impedes telomere maintenance in these cells, while its depletion causes rapid telomere loss.339,340 It will be important to understand how RNase H1 contributes to HR in ALT cells as well as during replicative senescence. Similarly, the overexpression of RNase H1 in yeast cells prevents the generation of type II survivors, which are considered the yeast ALT equivalent.341 The microbial pathogen Trypanosoma brucei uses telomeric TERRA R-loops to induce HR-dependent antigen switching to evade immune detection and increase pathogenesis.342 Therefore, the use of TERRA R-loops at telomeres to stimulate HR is evolutionary conserved, although exploited for different means.</p><p>In human cells, telomere dysfunction induced by removal of TRF2 leads to increased TERRA levels at all transcribed telomeres.332,343 Furthermore, the TRF2 homodimerization domain, which induces chromatin compaction344 and prevents DDR activation,345 represses TERRA transcription independently of p53 and does not rely on ATM-dependent DDR signaling.332 The UUAGGG-repeat array of TERRA transcripts directly bind to SUV39H1 H3K9 histone methyltransferase, sustaining the accumulation of the heterochromatic mark H3K9me3 at dysfunctional telomeres.332 Similarly, TERRA has been reported to accumulate H3K9me3 at telomeres,346 indicating a functional role of TERRA in heterochromatin reorganization at telomeres. In contrast to the idea of TERRA transcription arising from several individual subtelomeres,332 recent reports propose that TERRA transcription is restricted to one, or two at most, subtelomeres in mouse and human cells.347,348 Further studies are needed to clarify these apparent inconsistencies.</p><p>Although much effort has been focused on understanding TERRA's function at telomeres, it has recently been shown that TERRA also has nontelomeric functions.327 TERRA physically associates with thousands of nontelomeric loci where it frequently positively regulates transcription. TERRA binding sites overlap strikingly with those of ATRX—a chromatin-remodeling protein known to aid deposition of H3K9me3 at telomeres—and the two appear to compete at genomic sites. Indeed, TERRA and ATRX physically interact, and TERRA can displace ATRX from DNA templates, suggesting that it may remove ATRX from chromatin. In agreement, upon TERRA depletion, ATRX foci accumulate in the nucleus, both at telomeres and elsewhere. It is noteworthy that TERRA is upregulated in ALT cancers where ATRX is frequently mutated. In such a scenario, TERRA would be potentially unleashed to activate transcription via removal of ATRX, thereby inhibiting deposition of repressive H3K9me3 chromatin modifications; it will be interesting to determine how R-loops, ATRX, and TERRA-regulated transcription are coordinated. Moreover, gapmer oligonucleotides (see section 10 for details) targeting telomeric repeats have been used to fully deplete TERRA in mouse cells,327 leading to an increase of telomere dysfunction as well as a greater occurrence of other multiple telomeric pathologies, such as loss or duplication of the telomeric repeats and fusions between sister chromatids.</p><p>These findings, together, indicate a functional role for TERRA transcripts in the maintenance of telomere integrity in both mouse and human cells.</p><!><p>TERRA is not the only ncRNA transcript produced from telomeric DNA. In S. pombe a C-rich telomeric transcript (transcribed in a telomere to centromere direction) referred to as ARIA has been reported. Unlike TERRA, ARIA is composed purely of telomeric repeats and does not harbor subtelomeric sequences.349,350 Similar to TERRA, ARIA levels increase when telomeres are rendered dysfunctional by deleting components of the telomere specific Shelterin-like complex in fission yeast. This behavior is reminiscent of the transcriptional induction of both dilncRNA-from and dilncRNA-to species upon DNA damage.95</p><p>An interesting observation comes from mouse embryonic stem cells, where a positive correlation has been found between the heterochromatin status of telomeres and the level of telomere specific small RNAs (tel-sRNAs), that seem not to be TERRA degradation products nor DICER products.351 It has been hypothesized that tel-sRNAs belong to the piRNA family as the chemical features, the size, the resistance to β-elimination treatment, are conserved. tel-sRNA may be involved in the establishment of the heterochromatic state at telomeres. Consistently, they are regulated through epigenetic mechanisms, and their levels are positively correlated with the telomeric heterochromatin status.351 It remains to be determined whether TERRA serves as the precursor molecule for the generation of tel-sRNA.</p><p>As has been demonstrated for DSBs95 (see section 4.2.2 for details), mammalian dysfunctional telomeres, induced via TRF2 removal or functional impairment, also produce dilncRNAs181 (Figure 9). Telomeric dilncRNAs (t-dilncRNAs) arise from the transcription of both G-rich and C-rich strands of deprotected telomeres in mammalian cells. If, as canonical dilncRNAs, t-dilncRNAs are transcribed from the DNA end, they are unlikely to contain subtelomeric elements; thus, they are different from TERRA. Also, t-dilncRNAs are processed by DROSHA and DICER to produce short RNAs, called telomeric DNA damage response RNAs (t-DDRNAs). These sncRNAs are essential to ensure full activation of DDR signaling at dysfunctional telomeres and repair in the form of NHEJ-mediated fusions. It will be interesting to determine whether t-DDRNAs also play a role in the repair of telomeres by HR or in the maintenance of ALT telomeres. Moreover, the interplay between TERRA and t-DDRNAs has not been investigated. Similar to what has been demonstrated for endonuclease-driven DSBs95 (see section 4.2.2 for details), the use of ASOs (see section 10 for more detail) inhibiting the functions of t-dilncRNAs and t-DDRNAs efficiently reduces DDR activation at dysfunctional telomeres both in cultured cells and in mouse models. This suggests that telomeric DDR is amenable to specific control by exogenous agents, that have the potential to be developed as therapeutic agents.</p><!><p>With numerous lines of evidence pointing to RNA's diverse noncoding functions, the imperative to study cellular RNAs in their native context has never been greater. Recently, single molecule fluorescence techniques have been successfully applied to understand and enumerate the intracellular functions and properties of various RNAs and RNA–protein complexes.352-355 Furthermore, the involvement of both long and short noncoding RNAs in guiding the DDR machinery to sites of DNA damage was discovered in part using observations from single molecule fluorescence microscopy95 (see section 4.2.2). Rapid progress in intracellular single molecule detection methods to study RNAs and their binding partners has been driven by their inherent advantages over ensemble-averaging techniques. Most salient among these advantages are their abilities to detect, with great sensitivity, concentrations as low as those of most endogenous pathway components (eliminating the need for artificial overexpression, which bears the risk of overwhelming said pathway) and to reveal critical heterogeneities in the maturation, functionality, and spatiotemporal distribution of macromolecules. For a full historical perspective of the field of single molecule fluorescence-based RNA detection, we refer the reader to Pitchiaya et al. 2014.353 Here we present an overview of key advances in fluorescence-based intracellular RNA detection methods of relevance to the study of the DDR (Figure 10).</p><!><p>In situ hybridization (ISH) techniques localize nucleic acids in fixed tissue samples by detecting bound oligonucleotide probes with sequence complementarity to the target transcript. Since the first demonstration of the technique by Gall et al. in 1969,356 many versions of their protocol, varying in probing strategy and detection method, have been proposed over the years, with the objective of increasing spatial resolution, detection sensitivity, and throughput of RNA transcripts.357</p><p>While fluorescent in situ hybridization (FISH) has been used for the detection of nucleic acids for more than 30 years, its application for single transcript detection and counting has become possible only in the past decade. The premise of single molecule FISH (smFISH) is that multiple fluorophore-labeled nucleic acid probes bind over the length of a target transcript, allowing individual transcripts to be detected as bright spots against a dark background. Automated imaging and counting of these spots then yields information about the subcellular localization of transcripts and cell-to-cell variability in transcript number. Unlike many sequencing technologies, these intracellular transcriptomics methods are not limited to poly adenylated RNAs and are well suited for the study of noncoding RNAs. The sensitivity and ability of these technologies to detect even single transcripts make them particularly suited to studying RNAs involved in the DDR. For example dilncRNAs were detected using smFISH95 (see section 4.2.2).</p><!><p>A major challenge for intracellular RNA detection in the past has been the limited number of species that can be probed simultaneously. In this context, multiplexing refers to the process of detecting multiple significantly distinct molecular species in a biological sample. As with protein detection, multiplexing RNA detection can be achieved by multicolor labeling so that each target species (or probe) is tagged with a fluorophore of different emission maximum.358,359 However, multicolor detection today is limited by the number of colors that can be simultaneously imaged using conventional, single molecule sensitive light microscopy, to at most 7 distinct colors.</p><p>An alternative approach to multiplex RNA detection is sequential barcoding. With this strategy, the same diverse transcripts are repeatedly probed using different probe sequences. Individual transcripts are detected in multiple rounds as fluorescent spots, allowing high confidence detection of each molecule over multiple rounds of probing, without the need for multiple colors. To reprobe the same transcript using a different sequence, hybridized RNA probes can be degraded and washed away between successive rounds of hybridization (seqFISH),360 or the fluorophores on the hybridized probes can be photobleached, allowing the same wavelengths to be used for subsequent rounds of imaging (MERFISH).361 The sequence of hybridization rounds in which a single spot was detected then allows detection errors to be corrected and false positives to be eliminated, thereby increasing the sensitivity of detection (Figure 10A). Finally, suitable image registration and spot detection algorithms allow seqFISH/MERFISH to theoretically detect and locate hundreds of different transcript sequences at the single molecule level, making them powerful tools to study single cell transcriptomes with spatial information. A complementary technique, FISSEQ, employs rolling circle cDNA amplification to sequence transcripts in situ, providing added nucleotide-level resolution to transcripts in cells.362 These tools are poised to transform the field of RNA quantification and sequencing and are well positioned to aid the discovery of novel, rare, noncoding RNA species, such as dilncRNAs.</p><!><p>The primary challenge for single molecule detection is the need for a sufficiently bright spot signal, to enable super-resolution (~10–20 nm) localization of single transcripts. Common smFISH strategies aim to solve this issue by using multiple (>10) labeled probes to decorate the length of the transcript, and/or the use of illumination strategies such as confocal or HILO illumination to decrease background fluorescence.</p><p>One method that has gained attention recently, termed hybridization chain reaction (HCR), achieves signal amplification via self-complementary fluorescent probes that allow single molecule resolution even without specialized illumination schemes. HCR uses fluorescent probes that can self-assemble into long chains or branched structures, where each link is a probe that is bound to two others (Figure 10B). Increased interest in HCR for single transcript detection has led to improved HCR protocols that have been applied to demonstrate RNA detection in whole zebrafish embryos.363 These advances promise to make single molecule methods more accessible to the general research community by reducing the need for specialized microscopy equipment.</p><!><p>Cellular responses to changes in the environment often involve physiological adaptations that occur over time-scales of seconds to minutes. Here, live-cell analysis becomes necessary in order to capture rapid and dynamic physiological processes in real-time. The primary requirements for successful live-cell, real-time RNA visualization strategies are the ability to label transcripts for observation in living cells while retaining their biological functionality, the delivery of labeled RNAs into the cell, and the ability to image the labeled RNAs with high spatial and temporal resolution. The generally lower cellular abundance of RNA transcripts compared to proteins helps to discern closely spaced single molecules, making them easier to study than proteins using these methods.</p><p>Strategies for fluorescently labeling RNAs can be broadly classified into methods that label RNA secondary structures and those that label specific nucleotides within the sequence.</p><!><p>The most widely adopted RNA labeling strategy to detect transcripts in living cells has been the use of RNA-binding viral coat proteins (VCPs).364 This method, first demonstrated by Robert Singer's group, exploits the high specificity and affinity with which VCPs such as the MS2- or PP7-coat proteins bind with their cognate RNA stem-loop structures. These stem-loop sequences are inserted within untranslated regions (usually the 3′UTR) of the transcript of interest, typically in multiple copies (8 to 96). These modified transcripts are then expressed along with fluorescently tagged VCPs (usually expressed as fusions with GFP or mCherry). The fluorescent signal from multiply bound VCPs allows individual RNA molecules to be visualized as single transcripts even without super-resolution techniques (Figure 10C).</p><p>A number of groups have independently developed live-cell single molecule translation reporter systems by combining VCP-based mRNA labeling strategies with intracellular protein immunolabeling methods.365-368 In these methods, intracellular, fluorescent, typically single-chain antibodies (scAb) bind to antigenic sites present on the protein of interest, thereby serving as fluorescent probes for the protein. Classical fluorescent protein tags are of limited utility to study fast processes such as translation elongation when genetically encoded along with the peptide under study, because the maturation of a protein tag itself occurs over longer time scales than those of translation (the fastest maturing GFP variants fold in ~10 min whereas translation occurs over seconds to minutes).369,370 However, moving the tag from the protein under study to an antipeptide scAb probe in these immonolabeling methods allows rapid processes such as peptide elongation to now be studied in real-time at single molecule resolution, where the detection is only limited by diffusion of the probes and antibody–antigen affinity.371,372 Because the binding of each fluorescent antibody is reversible, these detection methods are robust against loss of signal from photobleaching of individual fluorescent tags, making them suitable for time-lapse imaging of proteins, albeit with the caveat of substantial fluorescence background from unbound probe. It is conceivable that these live-cell immunolabeling methods can be used to detect sites of DNA damage or monitor recruitment of protein components with greater ease than classical fluorescent protein tags.</p><p>Another development in the field of protein detection has arisen from work on protein appendages such as Halo-, CLIP-, and SNAP-tags that can be labeled with specific suicide substrates.373 These protein tags are genetically encoded, thereby retaining the specificity of fluorescent proteins, but are more versatile, as they can covalently couple to specific membrane-permeable fluorescent ligands (Figure 10F). The advantage of using these small chemical ligands for labeling in cell culture is that they can be easily added, and the unbound probe easily washed away, making fluorescent labeling amenable to pulse-chase experiments. This technology has been applied to study various aspects of telomere biology374 and protein translation,368,375 and it is a powerful technology for studying single proteins.</p><p>A complementary RNA-detection method using fluorescently tagged, inactivated RNA guided-Cas9 enzyme for intracellular RNA tracking was reported recently,376 further extending the fluorescent toolset using an extant ribonucleoparticle.</p><!><p>Some of the major disadvantages of endogenous labeling—the need for significant sequence modification and protein overexpression—can be overcome by site-specific incorporation of modifiable nucleotides directly into the transcript. Commercial availability of fluorophores with improved photostability and pH tolerance, and development of chemical RNA synthesis and covalent conjugation technologies, allow RNA transcripts to be conveniently labeled in vitro for in vivo visualization.377-379</p><p>NTPs conjugated with fluorescent dyes can be used for sequence-specific or nonspecific cotranscriptional labeling (Figure 10E). The ability to control the number of fluorophores incorporated into a transcript further offers the opportunity of stepwise photobleaching analysis, in which the number of fluorophores, and hence labeled molecules, present in individual fluorescent spots can be counted. The intensity traces from individual spots, representing photobleaching curves, can be analyzed to reveal >10 fluorophores per spot.379</p><p>The primary challenge of labeling by chemical modifications is the delivery of covalently modified RNAs into cells. Diverse delivery strategies in use today range from variable-dosage methods such as vesicle endocytosis (e.g., lipid-based transfection) or delivery via membrane permeabilization (e.g., electroporation, permeabilization by detergents and bacterial pore-forming toxins), to defined-dosage methods such as microinjection, which have been reviewed recently.380 Microinjection-based, defined-dosage iSHiRLoC (intracellular single molecule high resolution localization and counting) has been successfully applied to understand the temporal evolution of miRNA maturation, target-binding, turnover, and subcellular localization in the cell.352,354,381</p><p>iSHiRLoC was applied recently to study the localization of the small DDRNA cleavage products of DROSHA and DICER. In addition to being used to observe localization of DDRNAs to sites of DNA damage, this tool has been used to investigate the functional role of these RNAs in DDR focus formation by virtue of its ability for controlled RNA delivery95 (see section 4.2.2). Together, live-cell single-RNA and -protein visualization technologies present a formidable toolbox that allows novel molecular functions in DDR to be probed in situ, in real-time.</p><!><p>Another class of RNA detection methods relies on the ability of a small molecule ligand, such as a GFP-fluorophore mimic—difluoro-4-hydroxybenzylidene imidazolinone—to emit enhanced fluorescence upon binding to an RNA aptamer, such as the prototypical "Spinach" aptamer355 (Figure 10D). Such aptamer sequences can be inserted into transcripts, and the enhanced fluorophore intensity of the ligand upon binding both reduces background and can be used to read out the concentration level of these RNAs. These aptamer-based methods have been applied to detect toxic RNA aggregates,382 to detect RNA modification activity,383 and as metabolite-sensors.384 Orthogonal to the protein-based RNA detection strategies discussed above, these methods hold great promise for intracellular RNA detection. The discovery of brighter turn-on and higher-affinity dye-aptamer combinations385,386 suggests that this technology may find broader applicability for single molecule detection in the future.</p><p>Multiple developments in the field of fluorescence-based intracellular RNA/ribonucleoprotein detection over the past decade are a testament to the growing awareness of the great importance of spatial and temporal information for understanding cellular RNA biology. The discoveries enabled by the technical advances discussed here continue to underscore the importance of RNA in cellular physiology, ultimately reaffirming the rise of smart RNAs, including in the DDR.</p><!><p>Given the emerging role of several distinct RNAs in multiple physiological processes often of clinical relevance, interfering with RNA functions can be exploited as therapeutic strategies. One of the most powerful sets of tools to achieve this are antisense oligonucleotides (ASOs) which bind to their RNA target directly through Watson–Crick base pairing.</p><p>Based on their mechanism of action, ASOs can be divided into two classes. Gapmers have a central DNA region of 8–12 nucleotides, flanked by 2–3 chemically modified nucleotides on each side. These modifications are designed to increase affinity and stability of binding (see below). When bound to their RNA target, the central part forms a DNA–RNA hybrid, generating a substrate for the activity of cellular RNase H enzymes that degrades the RNA strand.387</p><p>Blockers do not require a specific position of the modified nucleotides and are often referred as "mixmers". Typically modified nucleotides are present every 2–3 deoxynucleotides. Very short ASOs (8 nucleotide-long) containing only modified nucleotides have also been described and shown to be effective.388 Blockers do not activate degradation by nucleases, but they instead impose a steric block, preventing the interaction between their targets and other molecules, such as other nucleic acids or proteins.</p><p>In particular if an impact in vivo is desired, ASOs require chemical modifications which improve their stability, boost their binding to the target, reduce off-target effects, and decrease toxicity.389 The phosphorothioate (PS) backbone is widely used to improve nuclease stability and pharmacokinetics mainly through its increased hydrophobicity (thus increased cell membranes permeability) compared to the natural phosphate group and through its capacity to avidly bind serum proteins such as albumin, thus avoiding clearance by kidneys.390 As alternatives to the PS backbone, two uncharged chemistries are also used: the phosphorodiamidate morpholino oligomer (PMO)391 and the peptide nucleic acid (PNA)392 backbone. These modifications increase the stability and the binding affinity to their target; however, differently from the PS modification, they are not suitable for the RNase H-mediated degradation of the target RNA.</p><p>To further increase the binding affinity and nuclease resistance, sugar modifications have been developed, which are typically inserted at the 2′ position.387 The most commonly used are the 2′-O-methyl (2′-OMe), 2′-O-methoxyethyl (2′-MOE), and 2′-fluoro (2′-F) modifications of RNA. Other sugar modifications include conformationally constrained nucleotides, such as locked nucleic acid (LNA), constrained ethyl (cEt), and tricyclo-DNA (tcDNA). These nucleotide analogues have an extra bridge arising from the 2′ position, reducing the torsional flexibility of the sugar backbone and the entropy of duplex formation, strongly increasing the affinity and the specificity to the target. However, the introduced chirality of the backbone associated with these chemistries remains an unresolved issue.</p><p>Compared to small molecule drugs, ASOs are easier and faster to design, since the only information needed for their development is the sequence of the target RNA. In this regard they are prototypically regarded as the ideal tool to inhibit the functions of "undruggable" targets, typically cell factors lacking an enzymatic activity. In addition, biodistribution, toxicity, and stability are defined mainly by chemical and structural architecture, such as chemical modifications of sugars, bases, and backbone. It is therefore possible to predict to some extent the behavior of different drugs against different targets on the bases of previous studies of similar molecules with different sequences. However, in some cases the sequence itself can influence ASO toxicity, mainly caused by off-target effects, which can also be hybridization-independent. These toxic effects include pro-inflammatory events, immunostimulation, and liver and kidney toxicity.393 This sequence-specific toxicity is in most cases unpredictable; however, it can be reduced by altering the ASO length and/or the position of the chemically modified nucleotides.</p><p>Perhaps the most important challenge for ASO-based therapeutics is the delivery to its target.394 Differently from most drugs, which are small (less than 500 Da) and hydrophobic, ASOs typically weigh a few kDa and contain many negative charges.389</p><p>To be effective in vivo ASOs must first escape circulation and reach the target tissue. Their biodistribution is different in various organs;394 for example, in the liver, endothelium is characterized by fenestrations between the cells, allowing for a more efficient delivery. Some organs are instead completely inaccessible, like the brain. Indeed, ASOs cannot cross the blood-brain barrier. To overcome this issue, ASOs can be administered locally, in the case of brain target by intrathecal injection into the cerebrospinal fluid.395 An extra benefit of this strategy is that the blood brain barrier prevents ASOs to enter the bloodstream and be cleared by kidneys.</p><p>A second obstacle to overcome is the subcellular localization of ASO.396 Unconjugated ASOs are taken up into cells by endocytosis and pinocytosis, in the absence of a delivery agent. But in order to reach their target, ASOs must escape from the endosomes to get into the cytoplasm or the nucleus. This is a slow process, and the efficiency can vary in different cell types.</p><p>Many ASO-based therapies target mature mRNAs to induce gene silencing. This can be achieved through RNase H-mediated mRNA degradation, or translation inhibition. Other ASOs are instead complementary to pre-mRNAs, blocking the donor or the acceptor splice site, thereby preventing the binding of splice factors.397 Both of these strategies can be applied to pathologies caused by overexpression, or a gain of function mutation, of a specific gene, or to correct an irregular splicing event, thus modulating the biosynthesis of different protein isoforms with a possible therapeutic function. Beyond affecting gene expression levels, ASOs targeting mRNA sequences can also prevent the formation of detrimental nuclear structures associated with mutated RNA. For example, in repeat expansion diseases, such as familial amyotrophic lateral sclerosis, Huntington disease, and spinocerebellar ataxias, triplets or hexanucleotides with a high GC content in the transcribed portion of a gene are expanded. The mutated transcript, through multivalent base-pairing forms RNA foci by phase separation that have been proposed to disrupt cellular homeostasis by sequestering various RNA binding proteins.398-400 These aberrant nuclear structures have recently been shown to be effectively disrupted by complementary ASOs.401</p><p>In the last few decades our knowledge about ncRNAs has dramatically increased, and their role in many pathologies has been unveiled. Novel therapeutic approaches based on the use of ASOs could exploit the targeting of ncRNAs, which are virtually undruggable by small molecule inhibitors. Anti-miRs are ASOs complementary to the mature miRNA sequence; blocking a single miRNA could lead to transcription derepression of many different genes,402 while targeting a common seed sequence allows a single ASO to block a family of miRNAs.388 miRNAs have a role in many different diseases, including cancer, diabetes, infections, and cardiovascular diseases.403 In the past years, promising anti-miRs have been designed and successfully tested in vivo in animal models, and some have reached the clinical trials stage. One example is Mirvirasen, an ASO targeting the liver miR-122, which shows a strong antiviral activity in chronic hepatitis C infection.404</p><p>ASO-based therapy may be the best inhibitory method when a target ncRNA acts in the nucleus because, differently from siRNA-driven knock down, ASOs do not require the RISC complex acting in the cytoplasm. This is the case of many lncRNAs. The Metastasis-associated lung adenocarcinoma transcript 1 (MALAT1) is a very abundant nuclear lncRNA, which is upregulated in numerous cancers. ASO-mediated MALAT1 downregulation can reduce tumor progression and metastasis formation in mouse mammary or lung carcinoma models.405,406 Although these results are promising, special attention should be dedicated to the specificity issue and to control experiments for off target effects.407</p><p>A recent example of a novel species of ncRNA targeted by ASOs is the inhibition of dilncRNA and DDRNA functions (as described in sections 4.2.2 and 8.3). In a model of telomere deprotection, DDR signaling and repair at telomeres are suppressed by using ASOs complementary to t-DDRNA sequences.181 Silencing DDR activation at the telomeres could have the potential to treat some pathologies associated with telomere dysfunctions.408 Excitingly, ASOs complementary to dilncRNA and DDRNA transcripts generated at a specific damaged genomic locus inhibit local 53BP1 focus formation, without affecting the DDR activation in other genomic locations within the same cell95 (as already described in section 4.2.2). These data reveal that ASOs can be a suitable tool to interfere with the DNA damage signaling and repair events in a sequence-specific manner, potentially inducing genome instability and cell death specifically only in cells bearing a particular damaged DNA sequence.</p><p>However, despite their simple design and almost 40 years of research and development efforts since the first example of an ASO-based approach was proposed,409 only a few ASO drugs have been approved for use in clinics. The first one was Fomivirsen (Vitravene, Isis Pharmaecuticals/Novartis Ophthalmics), a 21-mer phosphorothioate oligodeoxynucleotide, which was approved by the FDA in 1998 to treat cytomegalovirus retinitis by intraocular administration. It targets the viral mRNA encoding for immediate-early (IE)-2 protein, inhibiting its translation.410 Another ASO approved for use in clinics is Mipomersen (Kynamro, Kastle Therapeutics/Ionis Pharmaceuticals). Mipomersen is systemically delivered411 and has been approved for the treatment of homozygous familial hypercholesterolemia, which is characterized by high plasma concentrations of low-density lipoproteins. It is a 20-mer phosphorothioate 2′-methoxyethoxy (MOE) gapmer that induces RNase H1-mediated degradation of apoB mRNA. Eteplirsen, also known as Exondys 51 (Sarepta Therapeutics), is a 30-mer phosphorodiamidate morpholino oligomer for treatment of Duchenne muscular dystrophy.412 It is designed to induce skipping of exon 51 of the dystrophin protein, generating a shorter mRNA that encodes for a partially active isoform.</p><p>Very recently Nusinersen (Spinraza, Ionis Pharmaceuticals) has been approved. This is a 18-mer phosphorothioate 2′-O-methoxyethoxy oligonucleotide with all cytidines methyl-modified at the 5′-position.413 It is indicated for types 1, 2, and 3 spinal muscular atrophy (SMA) and acts by blocking a splice site in the SMN1 and SMN2 mRNA, causing the inclusion of exon 7.</p><p>The above-reported examples illustrate that RNA molecules are quite promising therapeutic targets and highlight how critical our continuous efforts in understanding RNA systems biology are.</p><!><p>In the introduction to this review we provocatively proposed a comparison between smartphones, objects that have transformed our daily lives with their diverse functions, and RNA molecules, which have increased the complexity of cellular processes as we know them. The obvious caveat is that while cellular phones evolved into smartphones by acquiring novel uses and functions, the multiple activities ascribed to RNA have always been intrinsic to its nature, just awaiting to be discovered.</p><p>The very concept of functionality can be a challenge when referring to an RNA molecule: a fixed length or sequence and interacting protein partners may not be sufficient to separate what is "junk" from what is functional. Essential mechanisms can thus seem invisible even to the eye of skilled and insightful scientists. Indeed, more collaborative research together with the advancement of cutting-edge technologies will help us reveal additional functions of RNA that contribute to the complexity of human life.414 Importantly, this knowledge will be essential to develop novel effective antisense-based therapeutic approaches.</p><p>Developments in the field of fluorescence-based detection of RNA and RNA binding proteins over the past few years are a testament to the growing interest in incorporating spatial and temporal information into the study of RNA and cell biology. These advances are bringing us closer to understanding the numerous protein-coding and noncoding functions played by RNAs in normal and disease physiology. In the future, combinations of the techniques outlined in this review are likely to further increase the amount of information that can be extracted from microscopic observations leading to a new era of fluorescent single molecule imaging in the life sciences.</p><p>In this review, we have discussed how RNA molecules contribute to protecting and repairing the genome, guide genomic rearrangements, regulate telomere homeostasis, and mediate epigenetic transcriptional silencing. The nature of transcription at damaged chromatin is becoming increasingly apparent with emerging evidence involving transcription, splicing, and RNA processing factors and with RBPs being recruited to the sites of DNA damage and being necessary for full DDR activation. How local de novo transcription95 coexists with transcriptional repression of the surrounding chromatin remains unclear and requires further investigation.74,77 The ncRNAs generated at DSBs may be responsible for the transcriptional inhibition of surrounding canonical genes by recruiting chromatin remodeler complexes, a model that is reminiscent of nascent transcripts at centromeric regions in yeast.190</p><p>To ensure efficient signaling and repair of DNA damage, DDR proteins must relocate to the right place at the right time, assembling at DSB sites in a coordinated manner. Although the DNA damage response is an extensively studied pathway, the precise mechanism by which a cell detects and shields DNA lesions is still under debate. An exciting hypothesis places RNA at the apical levels of the DDR cascade. Since RNA is capable of assembling and organizing a compartment in the cell by liquid phase separation,415,416 it is conceivable that it could be key to create a colloidal structure that holds and protects the DNA break and dynamically regulates access of DNA damage signaling and repair factors.</p><p>As proposed by Thomas Kuhn,417 in science, bursts of discoveries on a particular subject are often followed by periods of relative slow, steady progress when every key question seems to have been answered, until a totally unexpected twist occurs. When it comes to smart RNA, the feeling is that the best is yet to come: the burst we are currently experiencing is likely to become a monumental explosion.</p>

PubMed Author Manuscript

Accessing simply-substituted 4-hydroxytetrahydroisoquinolines via Pomeranz–Fritsch–Bobbitt reaction with non-activated and moderately-activated systems

Background: 1,2,3,4-Tetrahydroisoquinolines (THIQs) are common motifs in alkaloids and in medicinal chemistry. Synthetic access to THIQs via the Pomeranz-Fritsch-Bobbit (PFB) methodology using mineral acids for deactivated, electron-poor aromatic systems, is scarcely represented in the literature. Here, the factors controlling the regiochemical outcome of cyclization are evaluated.Results: A double reductive alkylation was telescoped into a one-pot reaction delivering good to excellent yields of desired aminoacetals for cyclization. Cyclization of activated systems proceeded smoothly under standard PFB conditions, but for non-activated systems the use of HClO 4 alone was effective. When cyclization was possible in both paraand ortho-positions to the substituent, 7-substituted derivatives were formed with significant amounts of 5-substituted byproduct. The formation of the 4-hydroxy-THIQs vs the 4-methoxy-THIQ products could be controlled through modification of the reaction concentration. In addition, while a highly-activated system exclusively cyclized to the indole, this seems generally highly disfavored. When competition between 6and 7-ring formation was investigated in non-activated systems, 5,7,8,13-tetrahydro-6,13-methanodibenzo[c,f]azonine was exclusively obtained. Furthermore, selective ring closure in the para-position could be achieved under standard PFB conditions, while a double ring closure could be obtained utilizing HClO 4 .

accessing_simply-substituted_4-hydroxytetrahydroisoquinolines_via_pomeranz–fritsch–bobbitt_reaction_

<!>Results and Discussion<!>Conclusion<!>Experimental Materials and methods<!>General method for the double reductive amination reaction<!>N-Benzyl-N-(2,2-dimethoxyethyl)aniline (9a<!>Supporting Information

<p>1,2,3,4-Tetrahydroisoquinoline (THIQ) motifs are present in many natural alkaloids [1]. THIQ derivatives have also been investigated as potential therapeutics in a wide range of diseases and recent studies have explored their potential as steroidomimetics [2][3][4][5]. Given the success of these authors in designing highly potent non-steroidal chimeric microtubule disruptors based upon decorated THIQ-based mimics of the steroidal AB ring system that possess pendant N-substituents [2], robust routes to direct N-aryl substituted THIQs were targeted for related activities (Figure 1).</p><p>Isoquinolines can be obtained from benzaldehyde and 2,2diethoxyethylamine under Pomeranz-Fritsch (PF) reaction conditions (Scheme 1), as first reported in 1893 [6,7]. A modification of the classic reaction reported by Bobbitt allows access to THIQ analogues (Scheme 1) [8][9][10]. Later, research has principally focused on the asymmetric THIQ synthesis and a substantial number of approaches to this end have been reported in the literature [1,[11][12][13][14]. Notwithstanding this, application of the Bobbitt modification of the Pomeranz-Fritsch reaction, or Pomeranz-Fritsch-Bobbit (PFB), as an approach to simplysubstituted THIQs has been somewhat neglected. In fact, most literature reports of cyclization under PFB conditions have concerned strongly-activated aromatic systems [8][9][10][11][12][15][16][17][18] with a few examples of cyclization of deactivated, electron-deficient, aromatic systems [19] in the presence of mineral acids. Synthetic approaches to electron poor systems more often involve different chemistries [20].</p><p>The PFB reaction proceeds via reduction of an intermediate iminoacetal to provide an aminoacetal that is then cyclized and reduced to deliver the 1,2,3,4-tetrahydroisoquinoline product. Relative to the original Pomeranz-Fritsch conditions, the Bobbitt modification features a reduced acid concentration that advantageously reduces the formation of side products [8][9][10].</p><p>The key cyclization in the PFB synthesis reaction is an electrophilic aromatic substitution that is strongly impacted by the effects of the substituents on the electron density of the aromatic ring in intermediate 4.</p><p>Given our established interest in exploring structural mimetics for the steroid nucleus we were drawn to explore whether the PFB reaction could be used to access libraries of regioisomeric N-aryl-1,2,3,4-tetrahydroisoquinoline derivatives that might facilitate the design of new structural templates for steroidbinding receptors. We thus evaluated the robustness and flexibility of this approach to the THIQ system and considered, in particular, the factors controlling the regiochemical outcome of the reaction to direct synthetic design. Reaction conditions of the PFB methodology and opportune modifications necessary to direct such reactions towards the synthesis of the desired THIQ derivatives are reported here.</p><!><p>To address aforementioned aims, we envisaged that the synthesis set out in Scheme 2 could deliver access to substrates for the PFB reaction such as 9. After an initial, unfruitful, attempt to obtain the aminoacetals 9 from sequential condensation of aniline and alkyl and aryl halides, a double reductive alkylation that could be telescoped into a one-pot reaction was investigat-Scheme 2: Designed synthesis of THIQ. Conditions: (a) NaBH(OAc) 3 , CHCl 3 , rt; (b) 6 M HCl or 70% HClO 4 (see Table 1), rt. ed. This reaction proceeded to deliver good to excellent yields of the desired aminoacetals 9a-g,i-p accompanied by only a small amount of side products (Table 1).</p><p>As an exception to the above, the acetal 9h did not form, which could be attributed to the poor solubility of the intermediate benzylated aniline. In addition, the reaction of 3,4-dimethoxyaniline under the same conditions did not afford 9q and instead 11 was the sole isolable reaction product (62% yield (Scheme 3)). The reaction was reproducible and proceeded even in the absence of a reducing agent (NaBH(OAc) 3 ). The preferential formation of 11 relative to imine reduction, even in the presence of an excess of reducing agent, suggests a rapid intramolecular rearrangement of an iminium derivative 12 to the highly activated ortho-position of the aromatic ring. The possible rearrangement product 13 would then lead to the isolated compound 11 upon hydration, either during the reaction or the work-up. However, the precise mechanism for the formation of 11 has yet to be explored and efforts to unravel it lay outside the scope of this work.</p><p>The addition of 6 M HCl led to the cyclization of the 3-MeO compound 9f, that bears an electron-donating group in a position parato the cyclization point. However, the same conditions were unable to catalyze the cyclization of 4-MeO compound 9b. We thus decided to proceed with an initial screen of acids for cyclization of non-activated compounds (Table 2). As previously mentioned, cyclization with activated systems, such as the one for compound 9f, proceeds smoothly with 6 M HCl, whereas the stronger acids H 2 SO 4 and HClO 4 triggered the degradation of the starting material (Table 2). In contrast, for a non-activated system, e.g., 9b, HClO 4 was the only acid</p><p>HClO 4 (70%) −10.0 c 0 Q SM (starting material), P (product), Q (quantitative); a pK a refers to the acid catalyst used in the reaction; b The reaction was monitored by LC-MS and the same method was used to calculate the percentage of conversion. c pK a measured in water [21]. d Degraded to unknown compound. e Hemiacetal formed. f Reaction was performed in a microwave at 100 °C for 30 min. g The was aldehyde formed.</p><p>able to deliver the desired THIQ 10b. However, in the presence of a deactivating, ortho/para-directing group, such as chlorine (9c) or bromine (9e), HClO 4 afforded the desired cyclized product only in the position directed by the substituent. In addition, even HClO 4 failed to catalyze the cyclization of 2-MeO compound 9n, rendering the target THIQ 10n inaccessible by this approach.</p><p>In certain cases, during scale-up of the reaction to a gram scale, a minor product became clearly identifiable and could sometimes be isolated. Whenever the cyclization was possible in both the paraand the ortho-position to the substituent, as expected, the para-position predominated and the 7-substituted THIQs (e.g., 10e-g, Scheme 4) were obtained as the major product. Nevertheless, it was possible to isolate a usable amount of the 5-substituted THIQ 14e. The ratio of the two regioisomers proved to be fairly constant, ranging from 5:1 to 4:1 for the three compounds considered. However, when more than one substituent was present, i.e., 9l, only compound 10l formed and the alternative product 14l was not observed.</p><p>A kinetic NMR study was performed by running a sample reaction directly in an NMR tube. However, the strongly ionic environment did not allow for a sufficient resolution of the NMR spectra and instead only extremely broad and convoluted peaks were observed. The kinetic study was then repeated by sampling the reaction at regular intervals, showing an almost instant hydrolysis of the acetal 9l to give the corresponding aldehyde followed by its rapid conversion into THIQ 10l within the first ten minutes of the reaction.</p><p>Despite the indication that the reaction proceeded through an initial complete conversion of the acetal to the aldehyde, in certain cases the formation of the ethers 15 was identified (Scheme 5). The formation of the two derivatives could be controlled through changes in the reaction concentration (Table 3). When 9f was dissolved in sufficient 6 M HCl necessary to provide a 1 M solution of the acetal, the main product obtained was the 4-hydroxy-THIQ. However, when 70% HClO 4 was used, the ratio between the 4-hydroxy and 4-methoxy-THIQs varied from ca. 1:1 to approaching 1:0, depending on the concentration of starting material in the reaction mixture. Ultimately, it was postulated that the ratio between formation of the ether and the alcohol is most likely a function of the water content of the reaction mixture. An acetal concentration of 0.3 M proved optimal to minimize the ether formation.</p><p>Scheme 5: Formation of the 4-hydroxy and 4-methoxy-THIQs. Conditions: (a) 6 M HCl or 70% HClO 4 , rt (see Table 3). The PF reaction conditions could possibly lead to a competition between 6-membered and 5-membered ring formation in systems such as 9a,o,p (Scheme 6). For this study, compounds bearing a different number of methoxy groups were selected in order to understand the electronic constrains that might render indole formation competitive. Upon treatment with 70% HClO 4 , acetals 9a and 9o yielded the respective THIQs 10a and 10o. However, the highly-activated acetal 9p exclusively cyclized into the indole 16p. It is thus possible to infer that indole formation is highly disfavored and only the presence of a great number of activating groups can make this outcome competitive.</p><p>We then decided to expand our analysis to include competition between 6-and 7-membered ring formation. Compound 18 (Scheme 7) was generated via a double reductive amination, in analogy to the syntheses of compounds 9a-q. Because none of the aromatic rings contained any activating group, the cyclization was performed with 70% HClO 4 as catalyst. However, in the experimental conditions used it was not possible to isolate either compound 19 or 20 (Scheme 7). Instead, the reaction proceeded to give complete conversion of 18 into the doublycyclized 21. Hence, the experimental conditions used did not allow to discriminate between 6-and 7-membered ring formation.</p><p>Lastly, we investigated potential competition between ring formation in paraor meta-position to an activating group such as a methoxy group, when both were possible. For this, compound 22 (Scheme 8) was synthesized via a double reductive amination, as per synthesis of 9a-q. As expected, when the precursor was treated with 6 M HCl, the cyclization occurred only in the para-position. However, when 70% HClO 4 was used, compound 22 afforded exclusively the double ring-closed product 25 (Scheme 8), analogous to 21. Presumably, in this case the six-membered ring para-cyclization may occur first followed by that directed by the six ring meta-position. Therefore, under our experimental conditions it was possible to obtain selectively the desired mono-cyclized product solely when this contained the activating group in the para-position.</p><!><p>The classical PFB conditions could be used successfully when at least one activating group was present in a position para to the cyclization point. When the substitution position was not para or the group was not an electron-donating group, the PFB conditions were not successful and the target THIQ could be obtained only with the use of 70% HClO 4 as a catalyst. This difference in reactivity could be exploited to control the cyclization point when both conditions were present, e.g., compound 22. In addition, in more simply substituted substrates, e.g., 9e-g, the formation of both 5-and 7-THIQs was observed in a somewhat constant ratio. This was not observed for compounds bearing more than one substituent. We noticed a strong preference for the 6-membered ring formation over the 5-membered ring, and the latter formed only in a greatly activated system. Conversely, no preference between 6-and 7-membered ring formation was observed. In addition, a 4-MeO-THIQ side prod-uct may be identified and its formation could be controlled by modifying the concentration of the starting material in the reaction media. In summary, the reported findings reveal the possibility to select confidently the appropriate conditions to synthesize the desired N-aryl-4-hydroxy-1,2,3,4-tetrahydroisoquinolines and derivatives to be explored as steroidomimetics in medicinal chemistry. Further investigation is currently being conducted and the findings will be reported upon their availability.</p><!><p>All chemicals were purchased from Aldrich Chemical Co. or Alfa Aesar. Organic solvents of A. R. grade were supplied by Fisher Scientific. Thin-layer chromatography (TLC) was performed on precoated plates (Merck TLC aluminum sheets silica gel 60 F254</p><!><p>NaBH(OAc) 3 (3.3 g, 15 mmol) was added to a stirring solution of benzaldehyde (1.0 mL, 10 mmol) and aniline (1.1 mL, 12 mmol) in CHCl 3 (60 mL) and the mixture was stirred at room temperature (rt) for one hour (h). 2,2-Dimethoxyacetaldehyde (30 mmol) was then introduced into the reaction mixture followed by NaBH(OAc) 3 (3.3 g, 15.0 mmol) and the resultant mixture was stirred at rt for further 8 h. The mixture was then quenched with saturated aqueous solution of K 2 CO 3 (60 mL) and the aqueous (aq) layer was extracted with CHCl 3 (2 × 30 mL). The combined organic layers were dried with MgSO 4 , filtered and evaporated to give the crude compound 9a as a pale yellow oil (3.87 g). General method for the PF cyclization with HClO 4 (method A)</p><!><p>Compound 9a (3.0 g, 11.1 mmol) was dissolved in 70% HClO 4 (33 mL) and stirred for 1 h at rt. The mixture was then diluted with water (30 mL) and basified by carefully pouring the mixture over Na 2 CO 3 . The aq layer was then extracted with EtOAc (3 × 30 mL) and the combined organic layers were dried with MgSO 4 , filtered and evaporated to give a brown foam (2.97 g).</p><p>General method for the PF cyclization with HCl (method B)</p><p>Compound 9f (500 mg, 1.66 mmol) was dissolved in 6 M HCl (2 mL) and stirred at rt for 1 h during which time the mixture turned red. The reaction mixture was cooled to 0 °C and then quenched by the slow addition of aq 3 M NaOH (10 mL) (a white suspension with a yellow precipitate formed). The mixture was then extracted with EtOAc (3 × 20 mL). The organic layer was dried with MgSO 4 , filtered and evaporated to give a yellow-brown oil (445 mg).</p><p>2-Phenyl-1,2,3,4-tetrahydroisoquinolin-4-ol (10a). The compound was synthesized according to method A. A sample of the crude compound was purified by column chromatography (eluent: from 0% to 10% EtOAc in pet. ether) to give a yellow oil which showed:</p><!><p>Supporting Information File 1</p><p>Synthetic and purification methodologies and spectroscopic data.</p><p>[http://www.beilstein-journals.org/bjoc/content/ supplementary/1860-5397-13-182-S1.pdf]</p>

Beilstein

Divalent Ion-Specific Outcomes on Stern Layer Structure and Total Surface Potential at the Silica:Water Interface

The second-order nonlinear susceptibility, c (2) , in the Stern layer, and the total interfacial potential drop, F(0)tot, across the oxide:water interface are estimated from SHG amplitude and phase measurements for divalent cations (Mg 2+ , Ca 2+ , Sr 2+ , Ba 2+ ) at the silica:water interface at pH 5.8 and various ionic strengths. We find that interfacial structure and total potential depend strongly on ion valency. We observe statistically significant differences between the experimentally determined χ (2) value for NaCl and that of the alkali earth series, but smaller differences between ions of the same valency in that series. These differences are particularly pronounced at intermediate salt concentrations, which we attribute to the influence of hydration structure in the Stern layer. Furthermore, we corroborate the differences by examining the effects of anion substitution (SO4 2for Cl -). Finally, we identify that hysteresis in measuring the reversibility of ion adsorption and desorption at fused silica in forward and reverse titrations manifests itself both in Stern layer structure and in total interfacial potential for some of the salts, most notable CaCl2 and MgSO4, but less so for BaCl2 and NaCl.

divalent_ion-specific_outcomes_on_stern_layer_structure_and_total_surface_potential_at_the_silica:wa

I. Introduction. Ion specific interactions at charged interfaces have been explored intensely over<!>B.<!>III. Results and Discussion.<!>B. Estimated Trends in<!>D. Hysteresis and Manifestation in 𝚽<!>V. Associated Content

<p>the years [1][2][3][4] but they are challenging to incorporate into models. At mineral/oxide interfaces, mobile ions form an electrical double layer (EDL) that extends from the solid surface into the aqueous bulk, modulating electrostatic interactions and balancing the charges that exist at the interface. Longstanding questions remain about what molecular properties govern EDL structure. [5][6][7] Descriptions of charged interfaces have been commonly based on empirical models such as the Hofmeister series or mean-field theory. 8,9 While these models are adequate for a description of macroscopic behavior, they do not provide a proper fundamental molecular level description of interactions within the EDL, nor a chemical understanding of interfacial electrostatics. 7,9 As a specific example, primitive ion models treat all alkali earth cations as having the same +2 charge, neglecting important ion-specific properties such as hydration environment or hardness/softness. Over the years, much work has sought to fill in those necessary aspects and to provide a detailed description of both hydration structure of ions 5 and its influence on the electrostatic potential at an interface. 10 An important question that has arisen from such studies concerns whether electrolyte valency (z) is a reasonable description of ion correlations at the interface as well as the overall potential that exists at an aqueous interface. Another question that has arisen pertains to the interplay between molecular interactions and electrostatics at the interface. Our previous work has examined these types of questions in the context of monovalent ions and their effects on hydration structure at the silica/water interface. 11 Exploring these questions with divalent ions offers an opportunity to pursue fundamental investigations of ion specific EDL structure and electrostatics at aqueous interfaces.</p><p>Alkali earth cations are common in the environment, play a large role in a variety of from Fisher Scientific (Anhydrous, Catalog # S421-500, ≥99% pure), MgCl2 6H2O from Sigma Aldrich (Part # 499609, ≥99% pure), CaCl2 2H2O from Sigma-Aldrich (Part # 21115, ≥99% pure), SrCl2 from Acros Organics (Catalog # 369740050, 99.9%), and BaCl2 from Mallinckrodt (99.9%).</p><p>Stock solutions were prepared for all salts using ultrapure water (Millipore-Sigma 18.2MΩ) at 1M ionic strength. The solutions were diluted further to the various concentrations and left exposed to laboratory air overnight to reach equilibrium with atmospheric CO2 (pH 5.7 ±0.2) prior to usage in the experiments to avoid formation of any hydroxides with the divalent cation species in solution 24 or complexation on silica substrate in the form of hydrates. 25,26 No further filtration was carried out prior to experimentation.</p><!><p>Experiments. Details of our HD-SHG spectrometer have been previously described. 11,[19][20][21][22] Solutions were flowed through a home-built Teflon sample cell at a rate of 5 mL min -1 for ten minutes for a complete exchange of the sample cell contents. This process was monitored using homodyne-detected SHG. We have previously found the sample cell exchange time to be around three to four minutes. 19 Interference fringes were collected by translating an a-quartz crystal 30 equidistant positions along a 100 mm translational stage, with an acquisition time of approximately 3 minutes per fringe. Sets of five replicate interference fringes were collected for each ionic strength/salt concentration condition, over a period of 20 minutes. We control for phase drift by clamping in the sample cell two to three hours prior to the experiment, after which the spectrometer's phase is stable for up to six hours. 21 The salt titrations were performed in a stepwise fashion, with five replicate measurements acquired for each aqueous phase condition. Each salt titration was repeated once, and our analysis is from the duplicate isotherms. Starting at ultrapure water (Millipore-Sigma 18.2MΩ cm, pH 5.8, strength of 100 mM, after which we return the ionic strength in a stepwise fashion to the starting condition of ultrapure water at 2 µM. Using the procedure described, the phase drift during the duration of the experiment is negligible (<1°).</p><p>C. Data Fitting, Phase Referencing Procedure, and Point Estimates of 𝚽 𝟎 and χ (2) . We utilize our previously published methods to extract the parameters needed for calculating the second order nonlinear susceptibility and surface potential. The generated interference patterns are fit to a sinusoidal function:</p><p>Here, 𝑦 " is the signal intensity offset, the SHG amplitude, A, corresponds to the SHG signal, Esig, f is the periodicity of our spectrometer, x is the position on the translational stage, and the SHG phase, 𝜑 &$' , is obtained from 𝜑 #$% . Both A and 𝜑 #$% are obtained by fitting eqn. 1 to the SHG interference fringes. The 𝜑 #$% obtained at 2 µM pH 5.7 is set to 60 o ± 1 o , which is the previously determined phase difference between that ionic solution condition and the one obtained for 500 mM NaCl at pH 2.5, 20 the measured point of zero charge (PZC) of silica, 1,2,27,28 where the Coulombic contribution to the total interfacial potential is zero. The SHG phase, 𝜑 &$' , at 2 µM and pH 5.7 is then +60 o ± 1 o .</p><p>The non-resonant SHG amplitude and phase obtained from eqn. 1 yields the total secondorder nonlinear susceptibility in our HD-SHG spectrometer according to: 20</p><p>Here, 𝜒 (+) is the second-order nonlinear susceptibility of the interface, which is given by the sum of the second-order nonlinear susceptibilities of the interfacial species (in order of abundance, these are interfacial water molecules, then the surface silanol groups, and then the anions and cations adsorbed to the protonated and deprotonated surface silanol groups, respectively, vide Ma and Geiger 6 infra). The third-order contribution in eqn. 2 is multiplied into the total interfacial potential, F(0)tot, which includes all electrostatic contributions (Coulomb, dipole, and multipolar potentials, vide infra). The third-order term is dominated by the third-order contribution from the water molecules in the EDL, which is given by 𝜒 -.%/0</p><p>(1) (9.6 ± 1.9 × 10 7++ m + V 7+ from off-resonant SHG experiments at the air/water interface or 10.3 × 10 7++ m + V 7+ estimated from the third-order molecular hyperpolarizability obtained through quantum mechanical calculations). 29,30 The thirdorder response from the water molecules in the EDL is modulated by the phase, 𝜑 23,526 , associated with the electrostatic DC field emanating from the charged interface into the bulk. For an exponentially decaying field, the DC phase angle is given by 𝜑 23,526 = 𝑎𝑡𝑎𝑛(∆𝑘 8 𝜆 2 ), where ∆𝑘 8</p><p>is the wavevector mismatch and 𝜆 2 the Debye screening length at a given bulk ionic strength, computed from Debye Hückel theory. For our experimental geometry, ∆𝑘 8 is 1.1 x 10 7 m -1 . The Debye length for each of our experimental conditions of ionic strength is determined using Debye-Hückel theory using bulk water's relative permittivity of 78. We recently reported 20 an additional purely imaginary third-order contribution that may be of quadrupolar nature, 𝑖𝜒 9 (1) , where,</p><p>.</p><p>Dipole and multipolar contributions are included in measurements of the differential capacitance of electrolyte:oxide:semiconductor devices, 28,31,32 as well as XPS signals from silica colloidal jets. [33][34][35] Both methods produce pH-dependent total interfacial potentials for the silica:water interface at various [salt] that agree well with our recently published HD-SHG-derived total potentials. 20 We therefore conclude that measurements of the SHG amplitude and phase provide the total interfacial potential drop across the oxide:water interface. It contains the Coulomb, dipole, quadrupole, and all other contributions, 𝛷(0) $ , to the interfacial potential drop, 36 and is quantified from the measured SHG amplitude, Esig, and phase, jsig, for a given DC phase angle, jDC, according to 19,20 𝛷(0</p><p>For silica substrates, the C/R ratio is 3.6 × 10 7++ 𝑚 + 𝑉 7D in our spectrometer. 20 Interfacial water is an ideal species to probe with nonlinear optics. Even though the nonresonant 2 nd -order hyperpolarizability of water, a (2) , is modest, 37 it is by far the majority species in most aqueous interfacial systems and often aligned in the Stern layer. How an array of water molecules is aligned in the Stern layer is encoded in the second-order susceptibility, a fundamental structural property of matter in noncentrosymmetric environments. 15 It is a measure of how the electrons are distributed in a non-centrosymmetric medium (the interface) and given by the number</p><p>Ni of a given interfacial species i multiplied by the orientational average of the hyperpolarizability, ai (2) . 17, 38, 39 HD-SHG quantifies c (2) from the measured SHG amplitude and phase according to 19, 20</p><p>employing the F(0)tot from eqn. 3 and the same C/R ratio.</p><!><p>A. SHG Amplitudes and Phases. Figures 1A and B show the SHG amplitude and the SHG phase for the range of NaCl, MgCl2, CaCl2, SrCl2, and BaCl2 concentrations indicated. Our earlier homodyne-detected SHG measurements indicate that the cation surface coverage increases with increasing cation concentration [40][41][42] up to an estimated saturation level of approximately 10 12 ions per square cm. This coverage corresponds to the number of negatively charged adsorption sites (SiOgroups) on silica at circumneutral pH reported from XPS measurements. 43 Our current HD-SHG measurements reveal a nonmonotonic trend in the recorded SHG amplitude for the aqueous NaCl solution, as recently reported. 11,19 In contrast, the divalent chloride salts show a less pronounced, weakly non-monotonic trend with increasing ionic strength in the SHG amplitude.</p><p>In our previous work, 21 comparison to NaCl, MgCl2, and SrCl2. We observe similar behavior in the SHG amplitude and phase when comparing Na2SO4 and MgSO4 with their respective chloride salts (Fig. 1C and D).</p><p>The SHG amplitude (Fig. 1C) is similar for both sulfate salts regardless of the cation identity. The SHG phase shifts relative to pure water (Figure 1D) across both sulfate and chloride species are approximately 30 o for all four salts surveyed. F(0) tot and χ (2) Across the Cations. Our recent report shows a possible method of separating the second-and third-order contributions to the SHG signal and therefore estimating interfacial structure and potential using eqn. 2, 20 resulting in eqns. 3 and 4. We now use this method to determine how the second order nonlinear susceptibility (χ (2) ) and the total surface potential, F(0)tot, depend on the chemical identity of the adsorbed ions and their concentrations.</p><!><p>Fig. 2A shows a large difference in χ (2) between NaCl and all the divalent cations, but no statistically significant difference in the χ (2) values among the divalent chloride salts. Fig. 2B reveals differences in χ (2) when sulfate is introduced as an anion (for both Na and the Mg salts).</p><p>The F(0)tot point estimates reveal a statistically significantly larger difference among the alkali earth chlorides we surveyed (bottom halves of Fig. 2). These differences do not follow the ionic radius of each cation species, which matches findings from previous work, including from calorimetric measurements. [44][45][46] Ca 2+ appears to have the largest field screening effect in the cation series, with F(0)tot reaching close to 0 V at 100 mM ionic strength. The larger ions, strontium and barium, come next, while the smallest (and hardest) ion, magnesium, lowers F(0)tot the least relative to NaCl.</p><p>Eqn. 4 shows that χ (2) is a linear combination of the contributions from the individual interfacial constituents. 21,23 These are, in our case, the interfacial water molecules, then the surface silanol groups, and then the anions and cations adsorbed to the protonated and deprotonated surface silanol groups, respectively. The χ (2) value from the interfacial water and surface silanol groups should be close to the one obtained at the lowest ionic strength (ultrapure water, 2 µM ionic strength, pH 5.8) condition. We therefore subtract this χ (2) value from the data shown in Fig. 2 and obtain, at least to leading order, the χ (2) values of the ions bound to the interfacial SiOand SiOH2 + sites (Fig. 3). The results indicate non-monotonic behavior and a maximum in χ (2) around 0.1 mM ionic strength for most of the salts we surveyed. We also find a change in the sign of χ (2) at an ionic strength around 1 mM for the divalent chloride salts we studied (Fig. 3A), and to a lesser extent in the sulfates (Fig. 3B). This observation would be expected in case of a flip in the net orientation of the radiating dipoles that produce the SHG response based on trends documented in previous studies. [47][48][49] We therefore find experimental evidence for a significant change in interfacial structure with increasing surface coverage for some of the ions we surveyed, consistent with reports by others for mica:water [50][51][52][53] and silica:water [47][48][49]54 interfaces. Previous studies of divalent cations, specifically magnesium and calcium, by Gibbs and co-workers, 55,56 have shown that low concentrations of these salts (0.033mM) attenuate the vibrational sum frequency generation (SFG)-resonant water signal in comparison to NaCl at similar concentrations albeit at a higher pH than the conditions studied here. The resonant vibrational SFG experiments attributed these trends to displacement of the hydration layer above the silica surface by ions retaining their centrosymmetric hydration shell. 55 These trends were attributed to close association of Ca 2+ to the interface, 55,[57][58][59][60][61][62] a finding that would be consistent with the non-resonant c (2) estimates reported here (as well as the largest reduction in interfacial potential by CaCl2 relative to NaCl).</p><p>These results are in some ways surprising. Previous studies that have explored hydration structure and the point of zero charge of silica indicate that cation identity has an outsize effect on the electrostatics over structure at the interface. 63 Likewise, previous x-ray reflectivity studies have revealed that hydration shell structures play a role in trends of interfacial potential. 34, 63-65 However, understanding the effects of different multivalent cations on surface charge density remains a point of contention. Potentiometric titrations by Dove and Craven 45 have shown reverse lyotropic effects on surface charge density at the silica/water interface, in the order SrCl2 < BaCl2 < CaCl2 < MgCl2.</p><p>Furthermore, calorimetric titrations by the Kabengi group have also shown slower uptake of M 2+</p><p>ions at the interface relative to M + ions with a positive lyotropic effect on heats of adsorption for increasing ionic radius. 44 In these experiments, the observed trends in ΔHads were strongly correlated with hydration properties for the monovalent ions, but did not hold as strongly for the divalent cations. Adsorption phenomena and changes in that behavior for various alkali earth cations were instead attributed to bare ionic radius and ionic potential (or charge/radius ratio). 44 Our findings may corroborate this scenario given differences between cation species in our experiments decrease with increasing surface coverage. Studies of the muscovite/water interface have also demonstrated that divalent ions with larger electron density such as Sr 2+ can adsorb in both fully and partially hydrated states. 52,66 Different adsorption mechanisms for counterion species with larger electron may explain the larger magnitude surface potential for Sr 2+ and Ba 2+ at 100mM, in spite of similar χ (2) values in comparison to Ca 2+ and Mg 2+ at higher surface coverages.</p><p>With the observed trends in χ (2) among all divalent ions, we compare our findings to previous studies that have examined the effects of increasing ion size on the overall hyperpolarizability in different molecular systems. From our experimental measurements, we find relatively small changes in the second order nonlinear susceptibility between the divalent halide salts. This outcome is surprising given the precedent for relatively large changes in the hydration structure at the interface. 45,55 Simulations have shown, for example, a highly ordered first solvation shell for divalent cations such as Mg 2+ vs a fairly labile hydration structure for Na + . 67 Furthermore, Na + is predicted to form direct contact ion pairs with silanol groups, whereas Mg 2+ is proposed to not bind directly to silanol groups at the surface, but rather form a hydrogen bonded complex through its tightly bonded hydration sphere. 67 Other MD simulations have shown Ca(OH) + can form at the silica surface upon deprotonation of one of the water molecules in the hydration shell of the Ca 2+ cation. 68 Previous theoretical studies of model complexes have indicated that hyperpolarizabilities generally increase with ionic radius, 69,70 but our experiments show little changes between ions to such extent. These studies have also found a threefold increase in the hyperpolarizability in these complexes between substitution of Ca 2+ with Na + . However, when examining the normalized 𝜒 (+) values, Ca 2+ and Mg 2+ represent the smallest change in comparison to Na + observed in these experiments. 68 These are also the hardest cations in the series we studied.</p><p>C. Chloride vs Sulfate. We applied the same analysis to anion identity to corroborate whether cation behavior did indeed play the largest role at the interface. While positively charged experiments (pH 5.8), previous experiments have shown a small number of SiOH + groups present even at neutral pH conditions. 71 In titrations of Na2SO4 and MgSO4, we observe changes in both and Φ " (Fig. 2B) and χ (2) (Fig. 3B) indicating that the presence of sulfate anions may have outsize effects on interactions at the interface, where the establish the largest negative potential at ionic strengths < 1mM. 72,73 Sulfates have been postulated to lead to silica dissolution through salting out effects that become especially pronounced at high temperatures. 26 In spite of silica's overall negative charge at circumneutral pH, 45,71 we observe a reduction in χ (2) and Φ " with the addition of Na2SO4 which demonstrates that the anion species likely plays a role in the electrical double layer 72 even though the "standard" electrical double layer model (e.g. Gouy-Chapman-Stern theory) predicts Na + should be the predominant surface-bound species. 47,74,75 We observe the opposite trend for the 2:2 salt MgSO4 which could indicate that sulfate ions have a notable influence on the magnesium ion coverage. We postulate that these may play a role in the changes in overall structure that facilitate salting out effects (i.e. sulfate-silicic acid structures that form) at higher concentrations. 26 We also note that the Φ " point estimates for chloride and sulfate anions are invariant between each shared cation species at ionic strengths > 1 mM, which may indicate little change to hydration structure in the diffuse layer at higher concentrations.</p><!><p>𝟎 and χ (2) . Another aspect we explore in this study is the dependence of adsorption and desorption reversibility on the cation identity. We perform reverse salt titrations immediately after the forward titration, reducing the salt concentration in a stepwise fashion using the same procedures as the forward titration. The preservation of the timescales from the previous stepwise titration is chosen to determine whether we arrive at the same structural and (approximately three hours per forward titration and the equivalent amount of time for the reverse titration).</p><p>Among all salts studied (Fig. 4 and S8-12) we observe the largest difference in χ (2) values for CaCl2 (Fig. 4A), for which the χ (2) point estimates are ~1.5 times larger in magnitude for the reverse titration in the lower concentration regime. The magnitude of the potential, on the other hand, is smaller for the reverse titration, again in the lower concentration regime. We observe these effects are less pronounced for BaCl2 and SrCl2 (Fig. 4A and S10). The MgSO4 reversibility manifests itself largely in the surface potential (Fig. 4C), while NaCl shows only very minor differences in interfacial structure and surface potential (Fig. 4D). Taken together, we find clear ion-specific outcomes on the structural and electrostatic hysteresis of our system for several of the salts we studied.</p><p>These results can be viewed in light of previous studies that have shown hysteresis in mineral oxide systems depends strongly on water structure as well as surface charge density. 46 Previous calculations of energies of adsorption indicate Ca 2+ likely forms a tighter contact ion pair with the silica surface disrupting the hydration structure at the surface than ions with a larger hydration shell, such as Ba 2+ . 54 These findings may demonstrate why larger hysteretic effects present for the CaCl2 salts compared to the other divalent cations. Surprisingly, we observe the least amount of hysteresis when using NaCl, at least under our present experimental conditions. A previous study we performed on a faster timescale, where the salt concentration was immediately jumped from ultrapure water to 100mM NaCl and vice versa, however, did demonstrate pathdependent effects. 19 The differences between these two outcomes highlight the importance of characterizing each step in surface-specific experiments involving fused silica, water, and salts. Conclusions and Outlook. Divalent cations at the silica/water interface probed with non-resonant HD-SHG spectroscopy measurements reveal that interfacial structure and total potential depend strongly on ion valency. When purely evaluating cation identity, we observe significant differences between the experimentally determined χ (2) value for NaCl and alkali earth cations, but smaller differences between ions of the same valency in that series. These differences are amplified at intermediate salt concentrations, which we attribute to the influence of hydration structure in the Stern layer, which is the origin of χ (2) . Furthermore, we corroborate the differences we report by examining the effects of anion substitution. Finally, we identify that hysteresis in measuring the reversibility of ion adsorption and desorption at fused silica under the stated conditions of our experiments manifests itself both in Stern layer structure, c (2) , and in total interfacial potential, F(0)tot.</p><p>Our estimates of c (2) are directly comparable to atomistic simulations of interfacial structure at aqueous interfaces that readily produce the second-order nonlinear susceptibility for resonant [76][77][78][79][80][81][82] and non-resonant conditions. 19,75 While adding the resonant c (2) values to the c</p><p>3) F(0)tot contribution (eqn. 2) through a simultaneous HD-SHG/HD-SFG experiment has not yet been achieved, the non-resonant c (2) values reported here are obtained directly through eqn. 4 and comparison to atomistic simulations is possible: the first approach calculating c (2) values for a charged interface (oxide:salt solution) was pioneered in 2019 by Chen and Singer, 75 focusing on the polarization of the water molecules. This method was recently expanded by us and the Miller group 19 to include surface hydroxyl groups, the second-most abundant interfacial species besides water at oxide:water interfaces. While that approach distinguished contact ion pairs from solventseparated ion pairs and showed the presence of the former recapitulated the experimentally observed trends in c (2) , ion specific effects were not pursued in that study.</p><p>Our c (2) estimates should be similarly informative for identifying those structural arrangements in the production runs of atomistic simulations that recapitulate the experimental c (2) values at oxide:water interfaces at a given pH (or surface charge), ion identity, and ionic strength.</p><p>Likewise, we expect that the estimates of the F(0)tot drop across the oxide:water interface we report here for the various salts we surveyed provides an experimental benchmark for mean field and/or atomistic models that include dipolar and multipolar contributions to the popular Gouy-Chapman-Stern theory, 83 the "standard model". Ionizing surface potential measurements published in 2021 by Allen and coworkers show that the surface potential ("c potential", 84 no relation to c (2) reported here) of the (nominally uncharged) pure air:water interface is as low as ~ -500 mV, 85 and that it is slightly less negative (~ -400 mV) at the air:electrolyte (1 M NaCl and 1M Na2SO4) interface.</p><p>Dipolar arrays of interfacial water molecules are thought to be the main contributors to this potential. [86][87][88] Multipolar contributions may also be important. 85,[89][90][91] Our own work with the Miller group 19 and the preceding study by Chen and Singer 75 also show surface potentials due to water polarization at negatively charged oxide:electrolyte interfaces estimated from molecular dynamics simulations are negative and in the range of multiple tens of mV, decreasing in magnitude with ionic strength. Those outcomes are consistent with earlier MD reports on structured water at charged aqueous:insulator interfaces, by, for instance, Borguet and Klein and co-workers. 92 New mean-field models beyond the standard model have been introduced that account for non-ideal behavior (short-range ion correlations, surface site availability, etc.) of aqueous electrolytes and ionic liquids. [93][94][95][96][97] A related issue is the spatial variation of the (field-dependent) relative permittivity, er, which the standard model neglects, i.e. the solvent is modeled as a uniform continuum, despite large differences in reported er. [97][98][99][100][101][102][103][104][105] The experimentally determined F(0)tot estimates obtained here include all the contributions to the potential and should thus be an Ma and Geiger 16 appropriate experimental benchmark to which theory must conform. It will also be informative to determine how ion-specific effects manifests themselves on other oxides, such as hematite, 106,107 which is an area we are actively pursuing.</p><!><p>Supporting Information: optical fringe data, c (2) and F(0)tot point estimates for forward and reverse titrations, for all salts studied.</p>

ChemRxiv

Results of VAMAS Survey Regarding Microplastic Issues

Trillions of tiny particles generated by our plastic-reliant society are polluting environments worldwide. An explosion of research has been devoted in the last years to detect, identify and quantify the microplastics, hidden not only in the oceans but also in the world's rivers, lakes, air, soil as well as food and organisms. Thus, we urgently need reliable standards to support the decision-maker to handle many issues related to this question. In this paper, the results of a VAMAS survey that involved 390 experts are presented and discussed. The inter-laboratory studies urgently needed in the next future are proposed.

results_of_vamas_survey_regarding_microplastic_issues

Introduction<!>Researchers of any countries can participate in the projects.<!>The survey<!>General Information<!>MP definition<!>Criteria for defining MPs<!>MP properties<!>Competences declared by the experts<!>MP characterization techniques<!>Inter-laboratory activities<!>Implications

<p>Microplastics (MPs) derive from countless human-made products and are found in every environmental sector 1,2,3 . Below the micron, the term nano-plastics (NPs) should be used, although in general this distinction is not yet applied, at least by current regulations 4 . MPs and NPs have all anthropic origin and consist of mainly six different polymers: polyethylene terephthalate (indicated in the recycling codes as PET or PETE), high or low-density polyethylene (indicated respectively as PE-HD or HDPE and PE-LD or LDPE), polyvinyl chloride (PVC), polypropylene (PP), polystyrene (PS) 5 . They originate from many everyday products (mainly cosmetics, synthetic fabrics, tyres) and are found in the environment principally due to the inadequate disposal of plastic waste and wear phenomena 3,6,7,8,9 .</p><p>The presence of MPs in water bodies worldwide is increasing and scientists, risk-managers, politicians and, above all, the population, consider the related issue to be of great concern. It is now necessary to fully understand the level of risk to human health and ecosystems and, therefore, the necessary prevention and control actions 10 . MPs may reach surface water bodies by different routes, such as leaching of particles in soil or air to water bodies, transport of plastics disposed inadequately and also from untreated sewage drains, due to meteoric events and transported by effluent from sewage treatment plants 11,7,12 .</p><p>Therefore, it is a topic of primary interest and requires fundamental interdisciplinary research. There is a need to implement the necessary environmental policies to deal effectively with the potential environmental problems 13,14 .</p><p>In this context, proper scientific and technical dissemination to the various stakeholders, decisionmakers and the population should promote awareness of the issue, its origin, dissemination and extent, deepening the aspects still under discussion to provide a sound knowledge base and standards 15,16,17,18,19 .</p><p>The risk of MPs for humans and aquatic organisms is still under discussion 20,21,10 . In fact, against the experimental evidence that shows how the MPs, ingested or breathed, can penetrate through the tissues of living organisms, their actual toxicity is still not assessed. Relevant are also potential impacts of the additives such as plasticisers and pigments 22,23,24,25,26 .</p><p>One of the major difficulties is the lack of standardisation of the tests for the evaluation of MPs. An effective strategy can be only achieved thanks to a clear scientific vision of the issues arising from MPs and NPs, allowing the development of common standards, and facilitating the agreement on international regulations.</p><p>MPs issues are addressed in the environmental legislation (with a particular focus on marine protection) and in the legislation dealing with products and product packaging. At a more general political level, in Europe, for instance, it is discussed in the plastic strategy 27 and the European action plan for the circular economy, "Closing the loop" 28 , where, however, MPs and NPs are not explicitly mentioned 4 . Due to the lack of reliable data, the precautionary principle has been part of the basis of the current regulation 29 .</p><p>There are currently no standardised and unambiguous methods for sampling MPs in the various environmental matrices, even if a lot of work has been done 30,31,32,33,34 . In such a complex scenario, it is necessary to know the matrix to be monitored and to carry out the proper sampling. The assessment of MPs is a multi-step process, including sample preparation (such as homogenisation or preconcentration), extraction, purification, identification, and quantification.</p><p>The composition of MPs plays a fundamental role in the precise knowledge and management of the problem of the relative environmental pollution. Unfortunately, any standard method for MPs identification has still been adopted. The characterisation of MPs can be performed by different chemical-physical techniques, each of which has advantages and disadvantages and through which it is possible to estimate different parameters. Among the different techniques, vibrational spectroscopies are attested as the most used techniques, because the analysis is fast, free of chemical reagents, cost-effective, and non-destructive 35 .</p><p>Knowledge of the sources, levels, environmental fate of MPs and the models are based on sampling and analysis, and reliable data are mandatory for the assessment of the management options 36 .</p><p>Another critical aspect of assessing is the ability of MPs and NPs to act as carriers for other environmental contaminants 37,38 . MPs and NPs can be distributed differently from the physical particles in the different abiotic environmental compartments and constitute an additional risk factor for organisms through the diet and the breathing. Further research is needed, which is still limited and scarcely comparable to clarify the transport mechanisms, the fate and potential for bioaccumulation of MPs and NPs in humans and to estimate the actual risks for each compartment.</p><p>The scientific community is rarely involved in the standardisation process since usually this is not considered part of the research. However, scientists should participate in this process when the assessment of experimental procedures is required. Indeed, a technical standard must correspond to the best practice assessed by experts, who must compare the results through inter-laboratory tests and understand the differences among results obtained by different techniques, different instruments, different laboratories. Standards can be used for proficiency tests to guarantee the reliably of the data, a mandatory requirement to any kind of analysis, and to help the operator to improve the laboratory performances. Moreover, the standards are relevant for the decision-makers, and scientists have the responsibility to give a contribution to their validation and assessment.</p><p>The standardisation process for the definition of material properties or new techniques is always based on research results (Figure 1). A new method then must be assessed first by intra-and then interlaboratory test to define the best procedure. After that, the standardisation process inside can begin.</p><p>The standardisation is a mandatory step in the case of issues related to society, in particular in the case of health and environmental normative. 39 . Indeed, these two steps are those often neglected by the scientific community, and VAMAS can internationally promote these actions. The scope of VAMAS includes all the steps necessary to define materials and properties: process, characterisation, and performances.</p><p>Today it involves Australia, Brazil, Canada, China, Chinese Taipei, France, Germany, India, Italy, Japan, Mexico, Republic of Korea, South Africa, UK, USA, Mexico, and European Union.</p><!><p>Why we need reliable standards for micro and nano plastics issues? Before coronavirus pandemic, plastic was considered one of the major environmental issues, together with global warming. Indeed, the production, demand and waste of plastics growth exponentially starting in the last century.</p><p>According to the analysis of European plastics production, demand and waste data 40 , in 2020 more than 368 million tonnes have been produced and this number is expected to double it in the next ten years. The worldwide coronavirus emergency has made make it even worst. Since the pandemic started, there has been a significant increase in plastic waste, such as masks, gloves and gowns. New solutions to handle the waste and to find new materials with more sustainable properties are needed and must be assessed.</p><p>On the other side, researchers are very active in the field of MP issues. During the last 2 years (namely, 2019 and 2020), more than 230 reviews have been published about different topics related to MPs and NPs. As shown in Figure 2a, the number of papers in the last 15 years is growing exponentially.</p><p>The research devoted to nano plastics has the same trend, even if there is an order of magnitude lower.</p><p>There is also a strong correlation between MPs and NPs-related papers published in each country as depicted in Figure 2b. The geographical distribution of the authors mirrors the international degree of the research in this field (Figure 3), confirming recent findings of detailed studies on global trends in MPs research 13,41 .</p><!><p>Last year, a survey was proposed to the scientific community to collect information about experts who work and plan to work on MP issues. 390 experts from 46 countries answered the survey. In this paper, we report a summary of the results. The survey is open and still available 42 .</p><!><p>On the 390 experts answering the survey, about 50% have worked with MPs. In Figure 4, the experts who answered the survey for each country are reported. The black columns indicate the experts who are involved in MP issues, the grey columns those who are planning to work in the field. In the red rectangle, there are countries where no experts who answered are yet involved.</p><p>We had the contribution of experts belonging to universities public research centres, private companies, metrological, and no-profit institutes. In Figure 5, it is shown that the large majority of the experts (75%) belong to university.</p><!><p>When standardisation procedures have to be assessed, definitions have to be clear and clarified 43,44 .</p><p>Regarding microplastics, there are different definitions on the web and in the scientific publications.</p><p>Thus, the first question of the survey regards the definition of MPs.</p><p>We proposed six different definitions 45 :</p><p>A. Small pieces of plastic, less than 5 mm (0.2in) in length; B. According to their origin, primary, if produces to be of microscopic dimensions or secondary if resulting from degradation and fragmentation processes in the environment; C. Plastic particles < 5 mm in diameter, which includes particles in the nano-size range (1nm); D. Lower size limits ranging from 1 to 20 μm; E. Microplastics are any synthetic solid particle or polymeric matrix, with regular or irregular shape and with size ranging from 1 μm to5 mm, of either primary or secondary manufacturing origin, which are insoluble in water; F. "Particles resulting from the degradation of plastic objects" and that "nano plastic exhibit a colloidal behaviour within size ranging from 1 nm to 1 μm".</p><p>As shown in Figure 6, the majority of the expert chose option E as the best definition for MPs.</p><p>Interestingly, the percentage of experts choosing different definition is independent by the country. The ISO technical report published in February 2020 46 has given the definition for MP, large MP and nanoplastics 44 . In Figure 7, the comparison of the survey results and the ISO definition. The term "large microplastic" is introduced for a particle with size between 1 to 5 mm. Remarkably, following the ISO definition, 200 nm particles is defined as nanoplastic, but not as nanoparticle.</p><!><p>The second question is "What are the key criteria for defining microplastic?" Experts can make more than one choice. As expected, the majority chose as key criteria the size. In particular, 255 experts choose only the size. However, 27 experts choose the origin as the only key criterion (Figure 8).</p><!><p>The following question was: "what properties of microplastics are more/less critical for the environment?" The expert can score the properties from 0 (not relevant) to 6 (mandatory).</p><p>Most of the experts chose biodegradability as the most critical properties to be considered, followed by chemical and biological features. Mechanical and thermal properties are those considered less critical from most of the experts. In Figure 9, the scores given by the experts are shown. The black points in Figure 9(a) correspond to experts who declared to have already scientifically involved in MP matters. In Figure 9(b), are shown the choices of the experts who scored as mandatory the thermal prosperities. Interestingly, most of the experts who score 6 for thermal properties, considered also mandatory the definition of all the properties of the MPs. In Figure 10, the correlation map among the scores given to the different properties is shown. Chemical, biological properties and biodegradability are considered the most critical. A strong correlation is clearly shown with the choices of mechanical and thermal properties. Some experts, answering the question gave interesting suggestions regarding other critical properties that have to be considered namely:</p><p>• Capacity to adsorb persistent organic pollutants (POPs);</p><p>• Transportation mechanisms;</p><p>• Biofilm formation around micro-particles;</p><p>• Degradation path;</p><p>• Absorption by living beings, ecotoxicity;</p><p>• Chemical additives in plastic production.</p><!><p>In Figure 11, the competences declared by the experts are reported. Other engineering competencies were added, as manufacturing, advanced water treatment, robotics, image processing. Unfortunately, social and economic expertise are missing. Likely, when reliable data and models are available, also their contribution will be mandatory. In the Supplementary Material, the correlation between the competences of the experts and their scores are shown. There is not a significant correlation among the scores and the declared expertise.</p><p>Based on the survey, experts had developed protocols to analysis MPs in all kinds of matrices, as waste, plants, animals, materials, and air. The majority declared to have developed protocols for the determination of chemical components in water (95) and sediment (67).</p><!><p>As expected by the bibliometric analysis 13 , the most used technique to characterise MPs is the infrared spectroscopy, with 111 experts declared to use it already. However, it has been recently outlined that most studies cannot be replicated due to missing experimental details 47 . Other techniques used by many experts are Raman spectroscopy 48,49 and electron microscopy 50,51 , followed by DSC-TGA 52,53,54 , ICP 55 and Chromatography 56 (see Figure 12).</p><!><p>The survey collected suggestions for inter-laboratory activities urgently needed, namely:</p><p>1. assessment of sampling in different matrices (air, waters and leachates, soils and sands, marine organisms, food) to avoid/limit sample contamination during sampling, storage, extraction and detection 57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72 ; 2. definition of reference standard MPs; 3. definition of protocols for dispersion or homogenisation of the microplastics onto the test media 68 ; 4. detection, identification of the chemical components, classification 73,74,75,76,77,78,79,49,68,56,48,80,50,81,52 ;</p><p>5. quantification and size and mass fraction distribution in different environmental matrices 82,83,84,85,86 ;</p><p>6. biodegradation and degradation and aging 87,88,89,24,90,91,92 ;</p><p>7. biological effect, evaluation of inflammation (and microbiota) in juvenile organisms (fish, mammals, birds, invertebrates), testing the transit time of microplastics through the gastrointestinal tract of marine animals, bioaccumulation in each type of tissue (fats, etc.), toxicity and ecotoxicity tests assessment 14,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118 ;</p><p>8. assessment of sewage and waters treatment 119,120,121,122,123,124,125,126,127,75,128,129,130,131 ;</p><p>9. assessment of fibers released by textiles 132,133,134 ;</p><p>10. transportation models 135,102,123,136,5,137,138 ;</p><p>11. isolation and identification of nano fraction in environmental samples including biota 139,140,141,142,143 ;</p><p>12. assessment of the impacts 144,145,5,75,8,146,147,148,149,150</p><!><p>The scientific community involved in MP and NP issues is growing fast since potential drawbacks for the environment and possible for human health are relevant.</p><p>To have reliable data are mandatory for the society and, in particular, for the decision-makers.</p><p>Inter-laboratory study to assess the protocols for sampling, detection and modelling the phenomena should be a priority to the standardisation. The evaluation of the real dangers should be assessed and declined in legislation acts that should be shared by all the countries to be effective.</p><p>The new technical area of VAMAS has been proposed to give the scientific community an international platform to very and assessed all the scientific topics related to MPs and to support all the national and international projects in developing reliable protocols.</p><p>As a fallout of the survey, a database of international experts with very different competencies who are already involved or would like to work on MP issues is available.</p>

ChemRxiv

HMP Binding Protein ThiY and HMP-P Synthase THI5 are Structural Homologues\xe2\x80\xa0,\xe2\x80\xa1

The ATP-binding cassette transporter system ThiXYZ transports N-formyl-4-amino-5-aminomethyl-2-methylpyrimidine (FAMP), a thiamin salvage pathway intermediate, into cells. FAMP is then converted to 4-amino-5-hydroxymethyl-2-methylpyrimidine (HMP) and recycled into the thiamin biosynthetic pathway. ThiY is the periplasmic substrate binding protein of the ThiXYZ system and delivers the substrate FAMP to the transmembrane domain. We report the crystal structure of Bacillus halodurans ThiY with FAMP bound at 2.4 \xc3\x85 resolution determined by single-wavelength anomalous diffraction phasing. The crystal structure reveals that ThiY belongs to the group II periplasmic binding protein family. The closest structural homologues of ThiY are periplasmic binding proteins involved in alkanesulfonate/nitrate and bicarbonate transport. ThiY is also structurally homologous to thiamin binding protein (TbpA) and to thiaminase-I. THI5 is responsible for the synthesis of 4-amino-5-hydroxymethyl-2-methylpyrimidine phosphate in the thiamin biosynthetic pathway of eukaryotes and shares approximately 25% sequence identity with ThiY. A homology model of Saccharomyces cerevisiae THI5 was generated based on the structure of ThiY. Many features of the thiamin pyrimidine binding site are shared between ThiY and THI5 suggesting a common ancestor.

hmp_binding_protein_thiy_and_hmp-p_synthase_thi5_are_structural_homologues\xe2\x80\xa0,\xe2\x80\xa1

<!>Protein Expression and Purification<!>Crystallization Conditions<!>Data Collection and Processing<!>Structure Determination and Refinement<!>Homology Modeling of THI5<!>Overall Structure<!>Dimeric Structure<!>Ligand Binding Site<!>Homology Modeling of THI5<!>PBPs and ThiY<!>Substrate Specificity of ThiY<!>Comparison of ThiY to Other PBPs<!>Comparison of ThiY with TbpA<!>Similarity Between ThiY and THI5<!>Comparison of ThiY and Thiaminase-I<!>

<p>Thiamin pyrophosphate (ThDP) is an essential cofactor in all living systems and is involved in branched-chain amino acid synthesis and carbohydrate metabolism. ThDP is implicated in the stabilization of an acyl carbanion intermediate formed in a variety of enzymatic reactions. Bacteria and Archaea, as well as certain lower eukaryotes, synthesize thiamin. Higher eukaryotes cannot produce thiamin and require it as a dietary supplement. Deficiency of thiamin in humans causes beriberi and Wernicke-Korsakoff syndrome (1). In all thiamin biosynthetic pathways, the thiazole and pyrimidine moieties of thiamin are synthesized separately and linked to form thiamin monophosphate (ThMP). The thiamin biosynthetic pathway is well studied in prokaryotes and the details of the eukaryotic pathway are beginning to emerge (2-5). In prokaryotes, the thiazole ring is synthesized from 1-deoxy-D-xylulose 5-phosphate, cysteine, and glycine (or tyrosine in Gram negative bacteria) by five gene products. The pyrimidine ring is synthesized by ThiC, a radical SAM enzyme, from 5-aminoimidazole ribonucleotide (4). In eukaryotes, the thiazole ring is formed from nicotinamide adenine dinucleotide, glycine and cysteine by one enzyme, THI4. The mechanistic details for the formation of the thiazole moiety are beginning to emerge and several steps of the reaction have been biochemically and structurally characterized (5, 6). The pyrimidine moiety is synthesized from histidine and pyridoxal 5′-phosphate by THI5 (7-9). The THI5 family of enzymes has a conserved CCCXC motif; however, no mechanistic or structural details are available for this complex reaction.</p><p>In addition to biosynthesis, most organisms have a thiamin uptake system (10-12). In Escherichia coli, thiPBQ encodes an ABC transporter that imports thiamin, thiamin monophosphate (ThMP) and ThDP into the cell (13, 14). The crystal structure of the periplasmic thiamin binding protein (TbpA), with ThMP bound, reveals that the protein belongs to the group II periplasmic binding protein (PBP) family. In addition, the structural similarity of TbpA with thiaminase-I, a thiamin-degrading enzyme, suggests that these proteins evolved from a common ancestor (15).</p><p>Binding proteins in Gram-negative bacteria are found in the periplasm; however, in Gram-positive bacteria, which lack an outer membrane, binding proteins are normally tethered to the cell membrane through acylation of a cysteine residue in a lipoprotein signal sequence (16, 17). ThiY contains a cysteine residue at position 19, which is preceded by a hydrophobic N-terminus. Cys19 is conserved in ThiY orthologs from other Bacillus species and is believed to provide the membrane anchor for ThiY. The hydrophobic lipoprotein signal sequence was removed for structural studies.</p><p>Thiamin is degraded in basic soil to N-formyl-4-amino-5-aminomethyl-2-methylpyrimidine (FAMP) (18). In Bacillus halodurans, FAMP is transported into the cell by the ABC transporter encoded by thiXYZ where it is deformylated and hydrolyzed to HMP, which is then incorporated into the thiamin biosynthetic pathway (Scheme 1). ThiY is the periplasmic FAMP binding component of the ABC transporter. Here we report the 2.4 Å resolution crystal structure of the B. halodurans ThiY/FAMP complex and compare the structure with those of the thiamin related enzymes TbpA and thiaminase-I. We also report a homology model for THI5, the HMP-phosphate (HMP-P) biosynthetic protein from yeast, based on the structure of ThiY. The modeling studies predict that the pyrimidine binding site is conserved in ThiY and THI5, suggesting that these proteins evolved from a common ancestor. The model also provides some preliminary insights into HMP-P biosynthesis in eukaryotes.</p><!><p>The construction of the B. halodurans ThiY overexpression strain was described previously (18). In this construct the N-terminal 20 amino acids were deleted to remove a hydrophobic membrane anchor and increase protein solubility. The gene was cloned into the pET16b plasmid and transformed into a methionine auxotrophic strain B834(DE3) of E. coli. A starter culture (10 mL) was grown overnight at 37 °C in LB medium containing 100 mg/mL ampicillin. The starter culture was harvested by centrifugation at 2000 rpm for 15 min, resuspended in 50 mL of minimal medium, and then transferred to 1 L of minimal medium. The minimal medium contained M9 salts supplemented with 40 mg/mL of each amino acid (except for methionine, which was replaced with L-selenomethionine), 100 mg/mL ampicillin, 0.4 % glucose, 2 mM MgSO4, 25 mg/mL FeSO4·7H2O, 0.1 mM CaCl2 and 1 % MEM vitamin solution. The cells were grown to an OD600 of 0.6 at 37 °C. Protein expression was induced by addition of 1 mM isopropyl-1-thio-β-D-galactopyranoside and the temperature was reduced to 15 °C. Cells were harvested after 15 h and stored at -80 °C.</p><p>The cell pellet was thawed and resuspended in wash buffer (20 mM tris(hydroxymethyl)aminomethane (Tris), pH 7.5, 150 mM NaCl, 10 mM imidazole and 2 mM β-mercaptoethanol). The cells were lysed by sonication and centrifuged at 24000g for 30 min to separate the lysate from the cell debris. The lysate was passed through a Ni-NTA column pre-equilibrated with wash buffer. The column was washed with 150 mL wash buffer followed by 50 mL of wash buffer containing 35 mM imidazole. The protein was eluted with wash buffer containing 200 mM imidazole and passed through a Sephadex G-75 column pre-equilibrated with 20 mM Tris, pH 7.5, 150 mM NaCl and 3 mM dithiothreitol. Fractions containing protein were identified by UV absorbance at 280 nm and were collected, pooled, concentrated to 10 mg/mL, and stored in aliquots at -80 °C.</p><!><p>The protein was thawed and incubated with 3-4 molar excess of FAMP for 2 h prior to crystallization. The crystals were grown using the hanging drop vapor diffusion method at 18 °C in 3 M NaCl, 0.1 M sodium acetate, pH 4.5 and 0.1 M lithium sulfate. Clusters of very thin plates with dimensions of 0.15 mm × 0.1 mm × 0.005 mm grew in a week.</p><!><p>All data were collected at NE-CAT beamline 24-ID-E at the Advanced Photon Source. Crystals were transferred to well solution containing 25% glycerol and flash frozen under liquid nitrogen prior to data collection. The crystals were very thin and were sensitive to radiation damage upon exposure to the X-ray beam. Partial data sets were obtained by exposing each of two crystals multiple times at various spots on the crystal (11 total positions). The partial datasets were merged to obtain a complete, 2.8 Å resolution dataset, which was used for initial structure determination. A second higher resolution (2.4 Å) dataset was obtained later by exposing a larger plate at six different positions. A total of 110° was covered with an exposure time of 1 s per frame and a crystal to detector distance of 300 mm. This dataset was used for final refinement. All data were indexed, integrated and scaled using the HKL2000 program suite (19). Data collection statistics are summarized in Table 1.</p><!><p>Analysis of the unit cell contents suggested two monomers of ThiY in the asymmetric unit, with six methionine residues per monomer excluding the N-terminal methionine residues. Ten initial Se sites were obtained using the program SHELXC (20, 21). The sites were input to the AUTO_SOL program in PHENIX (22); however, the resulting map was of poor quality because of the very thin crystals and because of pseudotranslational symmetry resulting from the twofold noncrystallographic symmetry. Therefore, the Fold and Function Assignment system server (23) was used to identify SsuA, an alkanesulfonate/nitrate binding protein, from Xanthomonas axonopodis (PDB ID 3E4R) as the closest structural homologue of ThiY, and a truncated model of SsuA was used for molecular replacement. The SAD phases and molecular replacement phases were combined using the program PHASER_EP (24) and the program RESOLVE (25) was used for density modification. These phases were used to locate the two remaining Se sites. The 12 Se site SAD phases and molecular replacement phases were combined, improved by density modification and non-crystallographic symmetry averaging, and used to calculate an interpretable electron density map. Model building was performed using the program COOT (26) and the model was refined using the CNS program suite (27). The model obtained from 2.8 Å SAD phasing was refined against the 2.4 Å dataset, which became available after the initial model building. The refinement process included successive rounds of simulated annealing, minimization, B-factor refinement, calculation of composite omit maps, difference Fourier maps and model building. A Fo- Fc difference Fourier map was used to identify FAMP bound to the enzyme. The ligand was added to the model followed by another round of refinement and identification of water molecules. The parameter and the topology files for FAMP were generated using the HIC-UP server (28). The final refinement statistics are given in Table 2.</p><!><p>The sequence alignment of Saccharomyces cerevisiae THI5 and B. halodurans ThiY was obtained using the program ClustalW (29). The sequence alignment showed three main areas of gaps/insertions: (1) THI5 contains an N-terminal deletion of about two dozen residues corresponding to the lipoprotein signal sequence of ThiY, (2) THI5 contains an insertion of nine residues near the conserved CCCXC motif, which is absent in ThiY, and (3) ThiY contains a deletion of about two dozen residues near the C-terminus; however, this deletion is difficult to position because of low sequence identity. The program MODELLER 9v3 was used to generate five homology models followed by energy minimization of each model based on spatial restraints (30). The model with the lowest energy was considered for further analysis. In this model, the conserved CCCXC motif was located in a cavity near the predicted THI5 active site, suggesting a [4Fe-4S] cluster or some other metal center. A [4Fe-4S] cluster was manually placed in the cavity and residues Cys195, Cys197, and Cys199 were reoriented to make covalent bonds between the sulfur and Fe atoms. The resulting structure was energy minimized to a gradient of 0.01 kJ/mol in vacuo using the AMBER* force field in the program Macromodel (31-33). The minimization used a TNCG minimization technique and a distance dependant 4r dielectric treatment (34). All the residues located within 5 Å of the [4Fe-4S] cluster were allowed to move during minimization and the rest of the protein remained frozen.</p><!><p>ThiY crystallized in the space group P212121 with unit cell dimensions a = 59.9 Å, b = 96.5 Å, c = 104.9 Å and one dimer per asymmetric unit. The ThiY protein used for crystallization had an N-terminal 20 amino acid deletion. In addition, residues 21-29 were disordered in the crystal structure. Electron density was present for all other amino acid residues, which were included in the final model. The crystal structure reveals that ThiY belongs to the group II PBP family. The monomer is comprised of two domains. Domain 1 is a three layer αβα sandwich comprised of a mixed β-sheet of five β-strands with topology ↑β4↓β10↑β3↑β1↑β2. The β-sheet is flanked by ten α-helices and one 310-helix. Domain 2 is a three layer αβα sandwich comprised of a mixed β-sheet containing five β-strands with topology ↑β7↓β6↑β8↑β5↑β9. This β sheet is flanked by three α-helices and four 310-helices. The two domains are connected by crossover loops between strands β4 and β5 and between helix η5 and strand β10. A ribbon drawing of the ThiY monomer and the topology diagram showing connections between the secondary structural elements are shown in Figure 1.</p><!><p>The enzyme crystallizes as a tightly packed dimer with monomers related by noncrystallographic twofold symmetry (Figure 2). The dimer interface is primarily hydrophilic and stabilized by extensive hydrogen bonding between residues Thr70-Asp167′, Asn71-Tyr139′, Thr136-Asn268′, Asp167-Phe170′, Arg180-Asn268′, Arg180-Phe170′ (where ′ denotes a residue from the other monomer). The dimer buries 1810 Å2 of surface area and the formation of the dimer seals off the entry/exit path for the ligand.</p><!><p>The ligand binding site of ThiY is located in a cleft between domain 1 and domain 2. FAMP was added during crystallization and was bound in the crystal structure. A stereoview of FAMP in the binding site is shown in Figure 3A and a schematic of interactions of FAMP in the binding site is shown in Figure 3B. The pyrimidine ring of FAMP is sandwiched between the side chains of Trp39 and Tyr188. Residues Phe170 and Phe269 interact edge-to-face with Trp39. Residue Tyr40 interacts edge-to-face with Phe170. The N2 atom of FAMP is hydrogen bonded to Glu192 and the N1 atom is hydrogen bonded to a water molecule. The water molecule is hydrogen bonded to the side chains of Asp38 and Asn42. The side chain of Glu192 is stabilized by hydrogen bonds to the amide group of Tyr188 and a water molecule. The amino group (N3) of FAMP hydrogen bonds to side chain atoms of Asp38 and Tyr90. The carbonyl terminus of FAMP hydrogen bonds to the amide group of Asn145. The hydroxyl group of Tyr188 hydrogen bonds to the N4 atom of FAMP and to the carboxylate group of Glu218. The hydroxyl group of Tyr188 is also at a distance of 2.85 Å from the CE1 atom of Tyr90 in chain B. Residues Trp39 and Phe170 appear to act as a gate covering the entry/exit site for FAMP. Residues Trp39, Phe170, Tyr40 and Phe269 form a hydrophobic pocket shielding the binding site from external solvent.</p><!><p>A homology model of yeast THI5 was generated based on its sequence similarity to ThiY. The sequences are 25% identical and 47% similar. The sequence alignment of ThiY and THI5 shows that the residues involved in binding the pyrimidine ring of FAMP in ThiY are partially conserved in THI5. A superposition of THI5 and ThiY/FAMP suggests that the HMP-P binding residues in THI5 are Trp12, Asn11, Gln165, His18, and Thr15. Residues Asn11 and Trp12 in THI5 align with Asp38 and Trp39 of ThiY, respectively. Residue Gln165 in THI5 aligns with Glu192 of ThiY, which is involved in hydrogen bonding to the N1 atom of FAMP. Residue His45 is hydrogen bonded to Asp38 in ThiY and is also conserved in THI5. These residues are critical to binding FAMP in ThiY and are predicted to correspond to the HMP-P binding site in THI5.</p><p>The homology model of THI5 is similar to the group II PBPs with two distinct domains connected by two crossover loops. The conserved CCCXC motif is located in one of the crossover loops between domain 1 and domain 2 and is located 15 -20 Å away from Trp12. A [4Fe-4S] cluster was built in the cavity between the domains facing the active site. The cavity between Trp12 and the [4Fe-4S] cluster is lined with residues Asn11, Gln164, Gln165, Phe117, His66, Glu161, and Gln121.</p><!><p>ATP-binding cassette (ABC) transporters are responsible for the transportation of a variety of ligands across the cell membrane. The importers and the exporters are believed to have diverged very early from a common ancestor. The importers are generally found in prokaryotes and are comprised of a periplasmic substrate binding domain and two cytosolic ATP binding domains. The exporters consist of a transmembrane domain and an ATP binding domain and recruit their ligands directly from the cytoplasm (35). The role of PBPs is to deliver small molecules and ions to the membrane-bound transporter for entry into the cell.</p><p>The crystal structures of many of the PBPs have been determined and strong structural homology suggests a common ancestor for this family of proteins. The PBPs are comprised of two domains connected by variable linker regions. This two domain architecture, which is believed to arise from gene duplication, is present in all PBPs. The PBPs have evolved into three different classes characterized by the number of crossovers between the two domains and the topological arrangement of the central β-sheet in the domains. Group I PBPs have two α/β type domains with the β sheet topology of β2β1β3β4β5 and three crossovers between the domains. Group II PBPs have two domains with the β sheet topology of β2β1β3βnβ4 with two crossovers between the domains. The strand βn represents a strand immediately after the crossover from the other domain. Group III PBPs have one crossover between the domains and little is known about the topological arrangement of the β-sheets in the domains. The substrate binding site in all PBPs is located between the domains and involves residues from both domains. These domains undergo a hinge-like bending motion ("Venus flytrap mechanism") during ligand binding (36, 37).</p><p>ThiY has two crossover connections between domain 1 and domain 2 revealing that it belongs to the group II PBP family. The topology of the central β sheet in both the domains is in accordance with the group II PBP family. The ThiY construct lacks the first 20 residues (hydrophobic membrane anchor) and an additional nine N-terminal residues were not visible in the electron density. The enzyme binds FAMP in the binding site located between domain 1 and domain 2. FAMP makes extensive stacking and hydrogen bonding interactions with the protein and is shielded from the environment by residues Trp39, Phe170, Tyr40, and Phe269. ThiY crystallizes as a dimer in the asymmetric unit and this dimerization closes the entry/exit path for the ligand. A similar dimeric arrangement was observed in the thiamin binding protein TbpA. In this case the dimerization also occurs in solution and is dependent on the concentration of the ligand (15). The physiological relevance of the TbpA dimerization is unclear. The dimerization of ThiY may be related to the mechanism of delivery of substrate to the ABC transporter or may be an artifact of crystallization.</p><!><p>The ThiXYZ transporter is found in a gene cluster involved in the salvage of FAMP, which is hydrolyzed to HMP for thiamin biosynthesis. The role of ThiY is delivery of FAMP to the ABC transporter and the equilibrium binding constant for FAMP is 200 nM (18). Our structure provides the atomic details of this binding. Tyr188 and Asn145 are involved in binding the formylamino group of FAMP through hydrogen bonding. Tyr188 is also involved in a stacking interaction with the pyrimidine ring. A sequence alignment of diverse ThiY homologues reveals that Tyr188 and Asn145 are not strictly conserved. The tyrosine residue is sometimes replaced by a phenylalanine residue, and the asparagine residue is replaced with a variety of residues such as serine, aspartate and glutamate. Residues Glu192, Trp39, Asp38 and Tyr90 that interact with the pyrimidine ring of FAMP are strictly conserved. The sequence variability in the ligand binding site suggests that the ThiXYZ system may have evolved to transport a variety of different thiamin degradation products with intact pyrimidines into the cell and therefore suggests the existence of other degradation and salvage pathways for thiamin. In addition to ThiXYZ, other transporters such as YkoEDC and CytX are also likely to be involved in the uptake of HMP or HMP precursors (10).</p><!><p>A DALI (38) search identified several structural homologues of ThiY. The closest homologues are SsuAs from X. axonopodis (XaSsuA; PDB ID 3E4R) and E. coli (EcSsuA; PDB ID 2X26) (39), which are alkanesulfonate/nitrate binding proteins, with a Z scores of 25.1 and 24.2, respectively. SsuA belongs to the type II PBP superfamily and XaSsuA has a molecule of 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) bound in the active site. A superposition of ThiY and XaSsuA is shown in Figure 4A. Despite low sequence idenity, XaSsuA was successfully used as a molecular replacement model in combination with SAD during initial phasing. The Cα trace of XaSsuA also served as helpful reference during ThiY model building. Other structural homologues of ThiY are hypothetical phosphate binding protein (PDB ID 1ZBM; Z = 21.4), nitrate binding protein (PDB ID 2G29; Z = 22.6) (40), and bicarbonate binding protein (PDB ID 2I4B; Z = 22.2) (41). Most of the top ThiY structural homologues are group II PBPs involved in transport of small ligands; however, MqnD, which is an enzyme involved in menaquinone biosynthesis, shows up in the top ten (PDB ID 3A3U; Z = 22.7) (42).</p><!><p>Both ThiY and TbpA are transport proteins involved in thiamin salvage. FAMP is the preferred ligand for ThiY, while the preferred ligands for TbpA are thiamin, ThMP and ThDP. A DALI search showed that ThiY and TbpA are structural homologues (PDB ID 2QRY, Z = 10.7) (15). A superposition of ThiY and TbpA (Figure 4C) shows that the ligand binding sites of both proteins are in a similar position. The structure of TbpA was determined with bound ThMP (15) and the binding site residues interacting with ThMP in TbpA are shown in Figure 4D. The thiazole ring stacks between residues Tyr27 and Tyr215 and is positioned edge-to-face from Trp197. The pyrimidine ring is buried in a pocket formed by Trp197, Trp280, Tyr221 and Tyr201 and the phosphate group is hydrogen bonded to Ser161, Trp197, Gly60 and Asp59. A comparison between thiamin and FAMP binding shows that the pyrimidine ring of FAMP superposes on the thiazole moiety of thiamin bound to TbpA and π-stacking is a common feature in the binding of both ligands. The formyl terminus of FAMP is in a similar position to the phosphate of thiamin.</p><!><p>In eukaryotes, HMP-P is synthesized from PLP and histidine in a reaction catalyzed by THI5, the yeast pyrimidine synthase (Scheme 2). The activity of THI5 has not been reconstituted in solution and its mechanism has not been established. Iterated BLAST searches (43) starting with ThiY identified various and proteins annotated as possible periplasmic binding proteins, as well as a few proteins identified as HMP-P synthases. However, all of the proteins identified by this search as HMP-P synthases lacked the conserved CCCXC motif of THI5 and are likely misannotated. Iterated BLAST searches starting with S. cerevisiae THI5 identifed orthologs with the conserved CCCXC motif but also periplasmic binding proteins. The S. cerevisiae THI5 shows 25% sequence identity and 47% sequence similarity to the B. halodurans ThiY, suggesting that ThiY and THI5 are structural homologues. In addition, ThiY binds FAMP, which is structurally similar to HMP-P, the product of THI5. Therefore, we undertook homology modeling in hopes that the THI5 model might provide useful clues for the successful reconstitution of the THI5-catalyzed reaction.</p><p>The sequence alignment of ThiY and THI5 shows that the residues involved in binding the pyrimidine ring in ThiY (Asp38, Trp39, Glu192, and His45) are conserved in THI5. In addition, Ser79, Ser82, Tyr113, Asp155, Asn271, Trp304 and Phe209, conserved in diverse sequences of THI5, are also conserved in ThiY. These residues are outside the active site of ThiY and may be involved in maintaining the structural integrity of the protein.</p><p>THI5 has a conserved CCCXC motif. Related cysteine rich sequences (e.g., CX3CX2C) bind [4Fe-4S] clusters involved in electron transfer or radical reactions (44) and are also known to serve as Lewis acids (45). The CCCXC motif was included in the homology model to evaluate if THI5 might contain a [4Fe-4S] cluster (or other metal center). The location of the HMP binding site was also predicted to identify residues potentially important for binding or catalysis. The homology model shows that the CCCXC motif is located in a deeply buried cleft between domain 1 and domain 2 (Figure 5). In this model, Cys195, Cys197 and Cys199 serve as ligands for the [4Fe-4S] cluster. The active site (identified from the location of FAMP in the superimposed ThiY) is present ∼15 Å away from the cleft. The space between the putative [4Fe-4S] and the HMP-P binding site could be occupied by the histidine and PLP substrates. The [4Fe-4S] cluster is ∼13 Å away from Trp12, which is proposed to stack against the pyrimidine ring of the product. While the [4Fe-4S] cluster has reasonable geometry, its occurrence has not been experimentally verified. It is also possible that the cysteine motif could be involved in binding smaller iron sulfur clusters. The residues lining the proposed active site cavity (Asn11, Gln164, Phe117, His66, Glu161, and Gln121) are strictly conserved in THI5 and may play a role in the reaction mechanism. These predictions will be used in developing a strategy to reconstitute the THI5 catalyzed reaction in vitro.</p><!><p>DALI searches show that ThiY is structurally homologous to thiaminase-I, an enzyme involved in thiamin degradation (PDB ID 4THI) (46), but with a much lower Z score of 5.6 for the full length protein. The homology is comparatively better when only domain I is used with a Z score of 7.1. The superposition of ThiY and thiaminase-I is shown in Figure 4B. The crystal structure of thiaminase-I determined with the mechanism based inhibitor 2,5-dimethylpyrimidin-4-amine shows the pyrimidine moiety covalently attached to Cys113 in the active site (46). The comparison of the structures shows that the binding site for the pyrimidine is in a similar location for both ThiY and thiaminase-I. A closer look at the binding conformations of the substrates reveals that the pyrimidine ring of FAMP is orthogonal to the pyrimidine ring of the inhibitor in thiaminase-I; however, this may result from the covalent attachment to Cys113. The centers of the pyrimidine rings are separated by ∼4 Å in the superposition. TbpA and thiaminase-I are structural homologues and believed to have a common ancestor (15). The structure of ThiY described here demonstrates that the thiamin metabolic proteins ThiY, TbpA, thiaminase-I and THI5 are (1) all members of the group II PBP family, and (2) all have binding sites for a pyrimidine moiety derived from thiamin. These similarities suggest that all four proteins may have evolved from a common ancestor.</p><!><p>Monomeric structure of ThiY. (A) The cartoon diagram of ThiY with the secondary structure elements labeled. The 310 helices are labeled as η. FAMP is shown as sticks. Carbon atoms in FAMP are colored green, nitrogen atoms are colored blue and oxygen atoms are colored red. (B) The topology diagram of ThiY showing the domain organization.</p><p>Dimeric form of ThiY looking down the two-fold axis. The monomers are colored blue and gray. FAMP bound in the active site is shown as sticks.</p><p>FAMP binding site in ThiY. (A) Stereoview of the FAMP binding site. FAMP has carbon atoms colored green. Water molecules are shown as red spheres and hydrogen bonds are shown as dashed lines. (B) Schematic representation of the binding site showing the key hydrogen bonding and stacking interactions.</p><p>Structural homologues for ThiY. Comparisons of ThiY with (A) SsuA, (B) thiaminase-I and (C) TbpA. ThiY carbon atoms are colored green and carbon atoms of the compared structure are colored cyan. The ligands are shown in ball and stick. (D) Superposition of the binding sites of ThiY and TbpA. Hydrophobic residues involved in stacking of TMP in TbpA are shown.</p><p>Homology model of yeast THI5. Stereo ribbon diagram showing the Cα trace of both the domains. The [4Fe-4S] cluster, Trp12, and FAMP (superimposed from ThiY) are shown as sticks. The CCCXC motif binding the cluster is located in the cleft on the other side of the active site.</p><p>Data collection statistics for ThiY</p><p>Values in parenthesis are for the highest resolution shell</p><p>Rsym = ΣΣi| Ii - <I> | /Σ<I>, where <I> is the mean intensity of the N reflections with intensities Ii and common indices h,k,l.</p><p>Refinement statistics for ThiY</p><p>R factor = Σhkl||Fobs|-k|Fcal|/Σhkl|Fobs|, where Fobs and Fcal are observed and calculated structure factors respectively.</p><p>For Rfree, the sum is extended over a subset of reflections (8%) excluded from all stages of refinement.</p>

PubMed Author Manuscript

Electronic structure studies reveal 4f/5d mixing and its effect on bonding characteristics in Ce-imido and -oxo complexes

This study presents the role of 5d orbitals in the bonding, and electronic and magnetic structure of Ce imido and oxo complexes synthesized with a tris(hydroxylaminato 3À) ligand framework, including the reported synthesis and characterization of two new alkali metal-capped Ce oxo species. X-ray spectroscopy measurements reveal that the imido and oxo materials exhibit an intermediate valent ground state of the Ce, displaying hallmark features in the Ce L III absorption of partial f-orbital occupancy that are relatively constant for all measured compounds. These spectra feature a double peak consistent with other formal Ce(IV) compounds. Magnetic susceptibility measurements reveal enhanced levels of temperature-independent paramagnetism (TIP). In contrast to systems with direct bonding to an aromatic ligand, no clear correlation between the level of TIP and f-orbital occupancy is observed. CASSCF calculations defy a conventional van Vleck explanation of the TIP, indicating a single-reference ground state with no low-lying triplet excited state, despite accurately predicting the measured values of f-orbital occupancy. The calculations do, however, predict strong 4f/ 5d hybridization. In fact, within these complexes, despite having similar f-orbital occupancies and therefore levels of 4f/5d hybridization, the d-state distributions vary depending on the bonding motif (Ce]O vs. Ce]N) of the complex, and can also be fine-tuned based on varying alkali metal cation capping species. This system therefore provides a platform for understanding the characteristic nature of Ce multiple bonds and potential impact that the associated d-state distribution may have on resulting reactivity.

electronic_structure_studies_reveal_4f/5d_mixing_and_its_effect_on_bonding_characteristics_in_ce-imi

Introduction<!>Synthesis and X-ray<!>Discussion<!>Conclusion<!>General synthesis methods

<p>The canonical view of lanthanide 4f-orbitals is that they are localized enough that they don't participate in bonding, or, in metals, in conductivity. In addition, their localization imparts local moment magnetism, whereby the orbital angular momentum is not quenched in contrast to typical transition metal d-orbital systems. This view, however, has been shown to be incorrect in that within certain lanthanide intermetallic materials, mixed valence is implicated in low-temperature temperature-independent paramagnetism (TIP) and in heavyfermion conductivity. 1 In such materials, the many-body interactions between the f-orbital and the conduction band (for example, the 5d orbital in elemental Ce metal 2 or the Cu 3d orbital in YbIn 1Àx Ag x Cu 4 3 ) are an important feature in determining the lanthanide (Ln) 4f orbital occupancy and magnetic properties. More recently, various Ce and Yb organometallic molecules, including Ce(C 8 H 8 ) 2 (cerocene) [4][5][6][7][8][9][10] and Cp* 2 Yb(bipy) (bispentamethylcyclopentadienyl ytterbium bipyridyl), 8,11 among others, [12][13][14][15][16][17] have been shown to exhibit similar f-orbital mixing, not with a fully metallic band, but rather with delocalized orbitals from aromatic rings proximal to the metal center. In these materials, single-electron theories such as DFT typically indicate no mixing of the 4f/ligand orbitals, while theories that account for conguration interactions between various magnetic singlets show a mixed ground state with nonintegral occupancy of the 4f orbital, allowing for enhanced TIP behavior from a van Vleck mixing of an excited state triplet. In cerocene, [4][5][6][7][8][9][10] for example, the ground state is primarily a quantum mechanical mixture of f 0 and f 1 congurations (closed-and open-shell singlet congurations, respectively), with a fractional f-occupancy of about 0.89 despite its formal oxidation state assignment as Ce(IV) based on a typical (C 8 H 8 ) 2À , and a van Vleck-induced TIP susceptibility of c 0 ¼ 1.4 Â 10 À4 emu mol À1 . Other formal Ce(IV) molecules have recently been shown to display similar behavior. 12 More recent advances by Sergentu et al. 18 have importantly conrmed that the doublet peak observed in Ce(III)/Ce(IV) systems is due to differences in the core hole interactions between f 1 and f 0 congurations. Additionally, they emphasized the f 2 contribution to what had been assigned as a pure f 1 spectral peak (although less than 10%), and therefore what has been traditionally referred to as the f 1 conguration is more accurately described as f 1,2 .</p><p>In the aforementioned systems, the Ln 5d orbital does not play a crucial role. This is in contrast to what is believed to be true for inorganic compounds such as CeO 2 and Ce 2 O 3 , where magnetism has no clearly measurable contribution from the forbitals, 12 likely due to the stronger delocalization of the felectrons from hybridization with the ligand 2p and metal 5d orbitals. Unfortunately, CAS methods used to elucidate electronic characteristics in organometallic molecules are not currently applicable to extended solids, such as CeO 2 . It would therefore be of interest to consider the behavior of formal Ce(IV) organometallics where Ce forms metal-ligand multiple bonds, which could serve as molecular analogs to CeO 2 . In addition, the notable reactivity patterns observed for transition metal oxo and imido compounds, which stem from the iminyl/oxyl character, [19][20][21][22][23][24][25][26][27][28][29][30][31][32][33] prompt the investigation of electronic structures and spectroscopic properties of lanthanide systems with metalligand multiple bonds. Towards this end, we consider a class of Ce imido and oxo complexes which are designed to enhance 5d participation while de-emphasizing ligand 2p orbital mixing. Recently, several alkali metal-capped Ce(IV) imido compounds [M(solv) n ][Ce]NAr F (TriNO x )] (solv ¼ DME or TMEDA, M ¼ Li + , K +, Rb + , Cs + , Ar F ¼ 3,5-(CF 3 ) 2 -C 6 H 3 , TriNO x 3À ¼ (tBuN(O))</p><p>C 6 H 4 CH 2 } 3 N] 3À , 1-M) and the uncapped congener [Cs(2.2.2 cryptand)][Ce]NAr F (TriNO x )] (1 À ), as well as the parent anilide compound Ce-NHAr F (TriNO x ) (1-H) have been reported (Scheme 1). 34 These compounds feature the TriNO x 3À ligand, which affords stabilization of the Ce(IV) oxidation state, protecting these compounds against reduction. Following isolation of the Ce(IV)-imidos and the uncapped congener, we isolated the cerium oxo complex Rb 4 [Ce]O (TriNO x )] 4 (2-Rb) 4 through an aza-Wittig reaction. 34 With this synthetic foothold established, in the current work, we present a detailed examination of the electronic structures of the cerium(IV)-anilide complex (1-H), select alkali metal-capped cerium imidos, 1-M, complexes and their uncapped congener, 1 À , as well the alkali metal-capped oxo, (2-K) 4 , M ¼ K, Rb, Cs, compounds outlined in Scheme 1. A chief challenge in the study of the cerium congeners is the participation of both the 4f-and 5d-orbitals in the valence electronic structure, and as will be shown below, considering both sets of orbitals is necessary to begin to understand these systems, and, in turn, to gain insight into Ce-ligand multiple bonding characteristics. Here, the 4f-and 5d-orbitals are probed through a union of synthesis, structural characterization, magnetometry, and Ce L III -edge X-ray absorption spectroscopy (XAS) techniques, combined with complete active space selfconsistent eld (CASSCF) calculations. The results of this combined investigation into electronic structure reveal that fractional f-occupancy is less impacted by structural differences between the imido and oxo constructs than d-state distributions controlled by bonding characteristics and alkali metal capping cation species. In contrast to cerocene and previously studied systems that involve direct bonding to an aromatic ligand, these structures present a single-reference ground state with strong 4f/5d hybridization. Ultimately, the insight gained from this study can be applied towards understanding how bonding characteristics can affect the electronic properties of Ln organometallic complexes.</p><!><p>For our expanded electronic structure studies of Ce(IV)-oxo compounds, we report the synthesis of the tetramer oxo complexes</p><p>(2-Cs) 4 . These compounds were prepared through an aza-Wittigtype reaction by reacting the corresponding imido compounds 1-K and 1-Cs with benzophenone and isolated in 59% and 39% yield, respectively. The relatively low yield of (2-Cs) 4 was attributed to its increased solubility compared to (2-K) 4 . Compounds (2-K) 4 and (2-Cs) 4 were crystallized from hot, concentrated THF solutions set undisturbed at RT. Complexes (2-K) 4 and (2-Cs) 4 are poorly soluble in most organic solvents and exhibit limited solubility in THF. In both cases, 1 H NMR spectroscopy of THF-d 8 solutions revealed the formation of a C 3symmetric, cerium-containing, product (Fig. S2 and S5 †). Complex (2-K) 4 indicated a solution magnetic susceptibility of c m ¼ 3.3 Â 10 À3 emu mol À1 following measurement by Evans method, a value that was largely consistent with the solid state magnetic susceptibility, vide infra. Evans method measurement of (2-Cs) 4 indicated a diamagnetic complex in solution. The characteristic doublets of diastereotopic protons of the CH 2 groups originating from the TriNO x -ligand were observed in the 1 H NMR spectra at 4.17 were shorter than in the monomeric, previously reported cerium oxo complex capped with a [Li-12-crown-4] + cation reported by Hayton, where this distance was 1.902(2) Å. 36 X-ray absorption near-edge structure (XANES). Before continuing to present the results of this study, we review the salient points of recent results concerning both single versus multiconguration terminology and using Ln L 3 -edge XANES in the context of determining f-orbital occupancy in hybridized systems, including those governed by conventional covalency and multicongurational systems. It is important to recognize that a system that is describable with a single conguration that is strongly hybridized, i.e. covalently bonded to the lanthanide center, can also be described as multicongurational through a rotation of the active space, although the reverse is not always true (e.g. cerocene). 37 With this fact in mind, although the Ln L 3edge XANES technique has been shown to be accurate in a variety of solid-state contexts (as mentioned above and many others), theoretical determinations have necessarily included assumptions to make these large-system calculations. 38,39 Consequently, concerns about the role of the 2p 3/2 core hole in the nal (excited) state of the XANES measurement have persisted. Recently, Sergentu et al. 18 have made detailed CAS calculations both of the ground state and of the excited state including the 2p 3/2 core hole on both cerocene and CeO 2 . Instructively, they have shown that, as expected, the two-peak feature observed in mixed Ce(III)/Ce(IV) systems (e.g. CeO 2 has an f-occupancy near 1 2 due to covalent mixing of the otherwiseempty 4f orbital and the oxygen 2p orbital) is indeed due to the difference in the core hole excited-state interaction between the f 0 and the f 1 congurations. Both the presence and the relative weight of the spectroscopic features are reective of the ground state properties and the XANES technique is, therefore, a quantitative way of extracting relative orbital occupancies of the ground state. However, while already understood, the work emphasized that such multiple-peak structure can exist either due to conventional covalent (single congurational) mixing or by multicongurational mixing, and is therefore not, by itself, an indication of multicongurational character. Another important point is that any f 2 -orbital contribution (which must be present in mixed valent Ce systems as a consequence of the Brillouin theorem), while typically <10%, is explicitly shown by Sergentu et al. to contribute to the weight of what has typically been considered as "the f 1 peak" and has likely contributed to small errors in f-orbital occupancy values reported in the literature. 18 We therefore have instead chosen to report the f 0orbital fraction, n(f 0 ), here to avoid this potential systematic error. These points are important to the presentation of the results, and so will be repeated and expanded below. XANES spectra (Fig. 2) indicate that both the 1-H starting material as well as 1-M, 1 À and (2-M) 4 samples can all be considered as formal Ce(IV). Similar spectra have been widely used to determine the f-orbital occupancy, although the relationship between the ground state and the actual spectra, which represent a nal state with a 2p 3/2 core hole, has, until recently, not been well understood. 18,40 In all measured samples here, the double peak characteristic to formal Ce(IV) is observed and conrmed by comparisons to a Ce(III) and a Ce(IV) standard (Fig. S12 †). All data were collected at multiple temperatures (50 K and 300 K) and exhibit no temperature-dependent changes in the spectra (Fig. S13 †). Note that in Ce(IV) XANES, the leadingedge peak near 5726 eV indicates primarily the degree of f 1 character due to covalence with the ligand, 18,40,41 since this fraction will screen the 2p 3/2 core hole and shi the 2p 3/2 / 5d 5/ 2 transition to lower energy compared to the peak near 5736 eV that corresponds to the f 0 part of the wavefunction. Since the lower energy peak should also include a small contribution from f 2 character, 11,18 hereaer, we refer to these peaks as the "f 1,2 " and "f 0 " peaks, respectively, recognizing that the predominant contribution to the f 1,2 peak is due to the f 1 conguration. We do, however, emphasize that these peaks are not due to transitions into f states, but rather are due primarily to dipole transitions into empty 5d 5/2 states and this notation refers to energy shis of these states depending on the f-orbital conguration. There is also a small peak at $5718 eV which has been attributed to arising either from a 2p-4f quadrupole excitation 42 or from mixed d-and f-states. 43 This peak is notably larger in the 1-Cs, 1 À and [2-Cs] 4 due to the additional contribution from the Cs L I absorption edge at 5714 eV in these samples, but otherwise appears similar between the other samples. The 1-H data appear to have the greatest ratio of peak intensities between the f 0 peak at higher energy and the f 1,2 peak at lower energy compared with the 1-M, 1 À and (2-M) 4 complexes. The imido spectra also have well-dened f 1,2 and f 0 peaks, but with visually higher peak intensities for the f 1,2 feature compared with the anilide. The oxo spectra, in contrast, while featuring the double-peak characteristic to Ce(IV), also have a at region in the middle of the two characteristic peaks indicative of other contributions. The origin of this feature was investigated using HERFD-XAS and is described below.</p><p>Surprisingly, no signicant differences in the f 1,2 and f 0 peaks were observed in the imido or oxo spectra as a function of alkali metal capping species. The only exception to this trend is 1-Li, which appears to have a slightly lower f 1,2 peak than the other imido counterparts. These trends were further explored through tting the areas under the f 1,2 and f 0 peaks (see Methods) in order to extract the f 0 -conguration fraction, n(f 0 ), within each complex (Table 1, Fig. S11 and Table S1 †). Reporting n(f 0 ) rather than the traditional use of the f-orbital occupation, n f , is necessary because we currently have no way of knowing the signicance of any f 2 conguration contribution to the f 1,2 peak, which has been calculated 18 to be separated by only about 5 eV and is therefore convolved with any 5d ligand eld splitting. Previous estimates of n f took the fraction of the f 1,2 -peak area compared to the total area of the f 0 and f 1,2 peaks with effectively n f ¼ 1 À n(f 0 ), and are therefore in error insofar as the f 2 conguration fraction exceeded $3% (the best estimate of the systematic error using this technique). We therefore follow ref. 18 and report the fraction of the well-resolved f 0 peak to the total area of the f 0 and f 1,2 peaks, which corresponds to the f 0 conguration fraction, n(f 0 ).</p><p>Fitting results from a two-peak model suggest a narrow range for all samples. The anilide has a slightly higher n(f 0 ) value (0.45) than the imidos and oxos, which matches the observable difference in relative peak intensity. All imido and oxo samples have an n(f 0 ) value of $0.4, without any perceivable trend based on alkali metal capping species or imido versus oxo bonding character. This lack of a trend is particularly interesting given the clear differences observed in the spectral features between the imido and oxo samples.</p><p>High energy-resolution uorescence detection (HERFD). HERFD spectroscopy was used for the purpose of highlighting features from Ce L III -edge XANES in higher spectral resolution, given that the resolution in this case is dominated by the 3d 5/2 core-hole lifetime broadening (0.87 eV) rather than by the 2p 3/2 core-hole lifetime broadening (3.2 eV) 44 that dominates conventional L III -edge XANES. This enables observation of features that are not resolvable from XANES and provides additional insight into complex electronic structure. In all HERFD-XAS spectra (Fig. 3, S17 and S19 †), three main peaks are observed. These include the f 1,2 and f 0 peaks observed in conventional XANES as well as the more-resolved pre-edge feature at $5718 eV, without interference of the Cs L I absorption edge present in the conventional XANES measurement. Unfortunately, the better-resolved features require many more t parameters when attempting to t these data, resulting in strong correlations between the parameters that render the t results unreliable. In any case, the peak area of this lowerenergy peak appears to be similar between samples, consistent with the measured n(f 0 ) values obtained within error from conventional XANES, assuming that the feature arises from mixed d-and f-states. 43 With the improved resolution, the "middle peak" feature observed in the oxo XANES spectra is also better resolved. The ability for FDMNES 45 simulations to reproduce this feature (discussed below) provides support that this feature is not due to a separate conguration but rather to ligand eld splitting. This splitting is also observed in the imido and anilide (Fig. S17 †) spectra, but to a lesser extent than with the oxos. The imido and anilide spectra appear similar to each other. The only observable difference is a slightly greater splitting of the f 1,2 peak in 1-H compared with the imidos, in addition to a higher f 0 peak, consistent with XANES. The trends observed in the intensity of the pre-edge peak correlate with those observed in f 1,2 peak splitting, with the imidos having the least intense pre-edge and the oxos the most intense feature. As with XANES, alkali-metal capping of the cation species appears not to appreciably affect the imido spectra. The same is not true for the oxos. Instead, a trend can now be distinguished as a function of cation with increasing Z. The "middle peak" feature appears most intense in (2-Cs) 4 , followed by (2-Rb) 4 , then (2-K) 4 . This trend suggests that the capping cation species can affect the electronic structure of the complexes. This hypothesis was further explored using FDMNES simulations.</p><p>FDMNES simulations of the local density of states (LDOS) provide insight into d-state splitting observed from HERFD-XAS. While FDMNES is limited to features due to the f 1 conguration as only one conguration can be simulated, the individual peak shapes are well-reproduced (Fig. 4, S16-S20 †). Through simulation, the separate contributions from various d orbitals to the density of states can be observed. From HERFD-XAS, a trend is visible in d-state broadening from the imidos (narrowest) to anilide to oxos (broadest). In the imidos, differences in d-state distributions are observed between varying alkali metal capping cations. Specically, a lower energy shoulder, marked with # in Fig. 4, is more pronounced in the 1-Li simulation than for the other imidos. In the oxos, the increased d-state splitting observed in the (2-Cs) 4 measurement compared with those from (2-K) 4 and (2-Rb) 4 in the HERFD-XAS can then be attributed to the d z 2 contribution splitting off to Greater spectral detail can be resolved compared with XANES (Fig. 2).</p><p>higher energy. While in (2-K) 4 , all d orbital distributions appear broad and at similar energy, the d z 2 contribution to the density of states splits to higher energy for (2-Rb) 4 and even higher for (2-Cs) 4 . Magnetism. SQUID magnetometry was performed on all samples and the obtained susceptibility versus temperature curves were t using a Curie-Weiss + constant model to extract values for the Curie constant (C J ), the Curie-Weiss temperature (W CW ) and the level of TIP (c 0 ) (Fig. 5 and S21, † Table 1). Relatively low C J values throughout (for comparison, C 5/2 ¼ 0.807 emu K mol À1 assuming unquenched spins) suggest that any paramagnetic impurity in the samples is of a low amount and the data should otherwise be representative of the sample of interest. While accounting for such impurity "Curie tails," all the complexes exhibit a non-zero c 0 , and hence TIP. Overall, the c 0 values are lowest for 1-H (1.1 AE 0.06 Â 10 À4 emu mol À1 ) and highest for the oxos ($4 Â 10 À4 emu mol À1 ), with the exception of (2-Rb) 4 , which is an outlier in this trend. Trends in c 0 are compared with n(f 0 ) values obtained through XANES tting in Fig. 6. These two quantities are expected to be strongly correlated when a multicongurational ground state singlet mixes via a van Vleck mechanism to a triplet state. 26,28,46 Results, however, show no perceivable relationship between c 0 and n(f 0 ). The results of Evans analysis of 1 H NMR spectra (see ESI †) conrm the paramagnetic nature of the products and are in agreement with the low TIP levels reported from the SQUID data.</p><p>Calculations. Fig. 7 summarizes the results obtained from CASSCF calculations for the ground state and the three lowest excited states of the Ce imido complexes, and Fig. 8 summarizes results for the Ce oxo complexes. Within Fig. 7 and 8, 0 means no electrons and 1 means two electrons, for example, in the notation 11 100 for the singlet ground state of 1-K. When the orbital is half-lled, the spin of the electron is given (alpha or beta). Since there are 5 active orbitals, this corresponds to 5 presented digits, where the lowest energy is on the far le and the highest on the far right. The ground electronic state of 1-K is predicted to be a closed-shell singlet despite the fact that the active orbitals were taken from the restricted open shell Hartree-Fock (ROHF) calculation of the spin triplet state (S ¼ 1). Interestingly, the two f-type orbitals are empty (see ESI † for more information about the MOs used in these calculations) in line with a Ce(IV)-type complex. These CASSCF calculations predict that the rst excited state is a single-reference triplet located 1.51 eV above the ground state. Two multicongurational triplet states are also found 2.39 and 2.85 eV above the ground state.</p><p>The computed ground state for the 1-Rb is also a singlereference closed-shell spin singlet. However, in this case three excited states, two multicongurational singlets and one singlereference triplet are closer in energy to the ground state (0.62, 0.65 and 0.84 eV, respectively) than in 1-K. In the case of the 1-Cs ground state, the CASSCF method predicts a closed-shell spin singlet with three excited states (two nearly single- CASSCF results show single-reference closed-shell singlet ground states for the three alkali metal-capped Ce imido complexes. However, the lack of 4f orbital occupancy in the CAS calculations appears inconsistent with the n(f 0 ) values determined from XANES ($0.4). Therefore, the bonding characteristics of the capped imido complexes were further analyzed using Natural Bonding Analysis (NBO) and Wiberg Indexes (WBI), with nearly the same results for all three of these complexes: in each case, a double bond Ce]N was found. Both s and p bonds are strongly polarized towards nitrogen (82% for s and 85% for p) and involve a hybrid 5d/4f orbital on Ce. Directly consistent with the XANES-derived results, the 5d/4f mixtures in these bonding orbitals are 60% 5f -40% 5d. The associated WBI is 1.76 (close to 2.0) indicating that the bonds are signicantly covalent. Therefore, in a covalent picture, a large fraction of these bonding electrons could be attributed to the Ce (and mainly on the 4f orbital), explaining the n(f 0 ) value obtained experimentally. For the sake of completeness, CASSCF calculations were carried out using the same methodology for the uncapped imido complexes. The ground state is found to be a closed-shell singlet, as for all alkali-capped complexes. The rst excited states are found to be two openshell singlets (0.45 and 0.54 eV) and a triplet (0.74 eV).</p><p>The CASSCF results for the alkali metal-capped Ce oxo complexes (Fig. 8) are quite similar to those determined for the imido case. For all complexes, we found single-reference ground states where the f-type orbitals included in the CAS are unoccupied. For both (2-K) 4 and (2-Rb) 4 , the rst excited state is a single-reference triplet, whereas for the (2-Cs) 4 it is a multi-reference singlet. The energy of the second singlet and second triplet of (2-K) 4 are 2.75 eV and 3.29 eV above the ground state respectively. On the other hand, the next calculated excited electronic states for (2-Rb) 4 (Fig. 8) are single-reference triplets in which one electron with alpha spin occupies one of the f orbitals of the Ce atom. For (2-Cs) 4 , three singlet excited states located 1.52, 1.98 and 2.18 eV above the singlet ground state have been identied. Following a similar procedure as for the imido complexes, the bonding was analyzed further. At the NBO level, only purely donor-acceptor interactions were observed, where the oxygen acts as a donor and the cerium as an acceptor. The associated WBI for each of the oxos is 1.34, indicating that the Ce]O bonds are less covalent than for the imido cases, and therefore having a stronger ionic component. This is reminiscent of the study by Barros et al. 47 regarding the nature of the bonding in uranium oxo and imido complexes. The acceptor orbitals on Ce are again hybrid 5d/4f orbitals that are based on 60% 4f and 40% 5d, just as with the imido complexes. These results are in line with the experimental observation of similar n(f 0 ) values between the imido and oxo complexes. CASSCF calculations for an uncapped oxo complex also predict the ground state to be a closed-shell singlet with the rst excited states being up to 4.0 eV higher in energy.</p><!><p>The overall results described above are surprising due to (1) the feature between the f 1,2 peak and f 0 peak in the L 3 -edge XANES which necessitated more detailed analysis using HERFD; (2) the similarity in n(f 0 ) between the imido and oxo compounds, irrespective of Ce ¼ N or Ce ¼ O motif or alkali metal capping cation species; and (3) the same lack of variability despite large changes in the degree of temperature-independent paramagnetism. In fact, our initial hypothesis was that the capping cation species might enable tunability in n(f 0 ) based on changing the degree of covalency; however, this was not observed.</p><p>More specically, the XAS results show that the anilide, imido and oxo samples can all be considered as formal Ce(IV), yet despite the similarities in n(f 0 ) between the Ce imido and Ce oxo complexes, XANES and HERFD data both exhibit notable differences in the characteristics of the f 1,2 and f 0 peaks. HERFD results conrm that these dissimilarities are not attributable to differences in f-orbital occupancy, but rather to ligand-eld splitting behavior of the 5d states. This result is interesting in that, while the electronic structure between the two classes of complexes is identical with respect to fractional f 0 occupancy, the distribution of the d-states differs depending on the Ce]N vs. Ce]O bonding species, with the splitting greater in the oxo complexes (Fig. 4). In addition, the d z 2 state is notably shied to higher energy in comparison to the imidos. The greater ligand-eld splitting in the oxos is somewhat expected, based on bond length trends, as indicated in Table 1: Ce-L bond lengths (L ¼ ligand) are shortest for the oxos (most splitting) and longer for the imidos (least splitting). Computationally, all alkali metalcapped imido and oxo complexes are determined to have closed-shell singlet ground states with hybrid 5d/4f orbitals that are based on a 60% 4f and 40% 5d mixture, consistent with the n(f 0 ) values extracted from XANES tting. As a general trend, however, the energy gap between the ground and lowest-lying excited state is smaller for the imidos than for the oxos. A different degree of covalent behavior is also observed between the classes of complexes. The WBI parameter corresponds to Ce]N double-bonded character for the imidos, suggesting that the bonding is signicantly covalent. In contrast, the WBI parameter for the oxo case suggests closer to single-bond character with a greater ionic component. Despite the different covalent character, similar n(f 0 ) values result from two different mechanisms. In the imidos, a fraction of the bonding electrons can be attributed to Ce on the 4f orbital, leading to a 5d/4f mixture in the bonding orbitals of 60% 4f and 40% 5d. In the oxos, only pure donor-acceptor behavior is observed, where Ce is the acceptor, with hybrid 5d/4f orbitals also with 60% 4f and 40% 5d. This observation provides an important example of how complexes with signicantly different bonding characteristics can still result in similar n(f 0 ) values.</p><p>We consider also the effects of varying the alkali-metal capping cation species from harder (Li + , K + ) to soer (Cs + ) on the resulting electronic structure of the Ce imido and oxo complexes. Capping-cation-dependent trends observed for the Ce imido complexes are subtle, and best perceived through FDMNES simulations (Fig. 4). The d-state distribution narrows as noted from a lower-energy shoulder going from Li + to Cs + . The opposite trend is observed in the oxo complexes, where dstate splitting is enhanced, resulting in distribution broadening going from K + to Cs + . This trend is visible in experimental HERFD data (Fig. 3) and explained through FDMNES simulations (Fig. 4) as resulting from the out-of-plane d-states (d z 2) shiing to higher energies. The uncapped imido 1 À was synthesized and exhibits very similar spectroscopic behavior to 1-Cs. Computational results provide additional information regarding the excited states of the complexes, and how they are affected by the capping alkali metal cation. As a general trend, for both the imidos and oxos, the energy gap between the ground state and lowest-lying excited state is largest for K + and lowest for Cs + capping species. With the exception of (2-Rb) 4 , all complexes feature multireference excited states within the three lowest energy states, but the nature of these excited states varies from complex to complex in an apparently non-systematic manner. Together, the data suggest that the alkali metal capping species, while having no perceivable effect on n(f 0 ), have the capability to ne-tune the d-state and excited state distributions within the complexes. Moving forward, it would be interesting to consider how this might be used as a tool to provide ner control over complex reactivity.</p><p>Results from our magnetism study as well as CASSCF calculations indicate that the electronic picture within the Ce imido/oxo systems is even more complex than the systematic trends observed from XANES and HERFD suggest. All complexes exhibit enhanced levels of TIP; however, a number of anomalies suggest that TIP is not driven by a simple van Vleck mixing of the ground state singlet and excited state triplet states involving localized f-and ligand orbitals as is believed to be the case for some other lanthanide organometallic systems. [7][8][9][11][12][13][14] For one, in contrast to previous work on Ln coordination complexes where multicongurational, open-shell singlet ground states exist, all these closed-shell Ce imido and oxo complexes exhibit a large difference in c 0 despite having similar n(f 0 ) values. In fact, it should be noted that the c 0 values represent a wider range than has been reported previously even for more structurally diverse formal Ce(IV) complexes, 46 although the amount of available literature data remain limited. Additionally, no observed correlation exists between the ground state and the rst excited state triplet determined from CASSCF results and c 0 , as would be expected for a conventional van Vleck mechanism for TIP. Moreover, while the imidos exhibit generally lower levels of TIP than the oxos, (2-Rb) 4 is a reproducible outlier in this trend. Lastly, within the imido and oxo species with varying alkali metal capping cation, there is no systematic trend from harder to soer cation species regarding the level of TIP observed. While some form of van Vleck mixing must occur to generate TIP behavior, some additional mechanism must be at play than those discussed here to result in the high c 0 values observed. Future work is needed to rationalize the TIP behavior and the lack of energetically close magnetic excited states from calculations.</p><p>An interesting factor that should affect the overall magnetic behavior is that the 5d mixing with the 4f states changes the degree of angular momentum quenching through a combination of the degree of d character and possible variations in the ensuing degree of delocalization of the f-orbital. Measuring the degree of delocalization of Ce 4f orbitals is very challenging. It has been previously noted that a small shi in the white line of the Ce L III -edge between the f 1,2 peak in CeO 2 and Ce(III) compounds is due to a more delocalized conguration of the 4f orbital in CeO 2 . 48 Recent calculations suggest this shi may instead be due to increased f 2 character when intermediate valence increases, which has a similar effect to a delocalized conguration. 18 In any case, similar shis are observed in other formal Ce(IV) compounds, 1 with cerocene having the closest correspondence to the leading f 1,2 peak of a pure Ce(III) compound. 2 In fact, the f 1,2 peak energy is shied from the position of pure Ce(III) (Table S2 †), indicative of delocalized behavior. Notably, the f 1,2 peak is not shied as far as in CeO 2 , suggesting that the 4f orbital in either the imido and oxo samples presented in this study is not as delocalized. However, there are no notable differences between the imido and oxo samples regarding the f 1,2 peak from XANES, the lack of which suggests that if this effect plays a role in the variations in c 0 for the present compounds, the ensuing edge shis are within the resolution of the data.</p><p>CASSCF calculations provide some additional insight into the non-systematic differences observed in the magnetism data for the imido and oxo complexes. We nd that the distribution and nature of these excited states varies from complex to complex. These variations could, in part, explain the lack of trends between the complexes in magnetism data. Although CASSCF results show trends in ground-to-excited state gaps that reect periodic trends in the alkali metal cations, related trends are not apparent in the measured TIP values. It is also worth considering why (2-Rb) 4 remains an outlier with respect to its low level of TIP in comparison to the other oxo complexes. While this exception does not mimic any of the trends observed from our spectroscopic measurements, there is precedence for magnetic behavior to follow non-systematic trends based on associated alkali metal species. 3 We can also consider anomalies in the nature of the excited states of the complexes. In particular, what differs for (2-Rb) 4 compared with the other complexes is both the lack of multireference excited states and the large energy differences between its excited states. Each of the three lowest-lying triplets is fairly discrete in energy (Fig. 8), rather than clustering around a similar energy range as is apparent especially for the imido samples (Fig. 7). Regardless of the exact reason for the anomaly in magnetism trends, it is apparent that the electronic structure of these complexes is complex, and beyond the predictability of periodic trends.</p><!><p>We performed electronic structure studies on previously reported Ce(IV) imido and oxo compounds, as well as two new Ce(IV) oxo species. The electronic behavior of the Ce imido and oxo complexes presented herein contrasts that of Ln organometallics in which delocalized electrons from aromatic groups are directly interacting with the Ln center, and as a result exhibit multicongurational ground states. Rather, by separating the Ln center from the aromatic groups and stabilizing Ce]N or Ce]O moieties through use of an alkali metal capping cation, the ground states are not multicongurational, but rather closed-shell singlet single-reference ground states. We observe that these closed-shell singlet systems are fundamentally different than those with open-shell singlet ground states due to the involvement of the 5d orbital, which we track in detail using XANES and HERFD-XAS. The 4f/5d hybridization within these systems is strong and is what gives rise to the double-peak feature observed in the resulting XANES spectra. Our spectroscopy results conrm that despite having n(f 0 ) values that are essentially the same across the complexes, the energy distribution of d-states within the complexes varies substantially. This distribution is also more subtly affected by the identity of the alkali metal capping cation. While previous studies have shown that small changes in molecular perturbation can result in large f-orbital occupancy changes in lanthanide organometallics, this study shows that small changes in molecular structure can result in relatively xed f-orbital occupancy, with different distributions of underlying states, potentially driving large differences in their paramagnetic states. Moving forward, such complexes provide a platform to consider in future studies how ne-tuning d-state distributions within Ln organometallic complexes such as those reported herein can affect reactivity in a manner that is independent of lanthanide valence.</p><!><p>All reactions and manipulations were performed under an inert atmosphere (N 2 ) using standard Schlenk techniques or in a drybox equipped with a molecular sieves 13X/Q5 Cu-0226S catalyst purier system. Glassware was oven-dried for at least 3 h at 150 C prior to use. 1 H NMR spectra were obtained on a Bruker DMX-300 Fourier transform NMR spectrometer operating at 300 MHz for 1 H, 75.48 MHz for 13 C and 282.2 MHz for 19 F. Chemical shis were recorded in units of parts per million and referenced against residual proteo solvent peaks. Infrared spectra were measured on a PerkinElmer 1600 series spectrometer. UV-vis spectra were recorded on PerkinElmer Lamba 950 spectrometer. Infrared spectra were recorded in the range from 4000 to 500 cm À1 on Bruker Invenio R spectrometer in KBr pellets. Elemental analyses were recorded on a Costech ECS 4010 analyzer. Cyclic Voltammetry (CV) experiments were performed using a CH-Instruments 620D Electrochemical Analyzer/Workstation. Data were processed using CHI soware v9.24. All experiments were performed in an N 2 atmosphere glove-box using electrochemical cells consisting of a 4 mL vial, glassy carbon working electrode, a platinum wire counter electrode, and a silver wire plated with AgCl as a quasi-reference electrode. The working electrode surfaces were polished prior to each set of experiments. THF solutions of the analyzed compound ($1 mM) and supporting electrolyte [ n Pr 4 N][BArF 24 ] (0.1 M) were used for electrochemical studies. Potentials were reported versus the ferrocene/ferrocenium (Fc/Fc + ) couple. Ferrocene was added as an internal standard for calibration at the end of each run. All data were collected in a positivefeedback IR compensation mode. THF and pentane were sparged for 20 min with dry argon and dried using a commercial two-column solvent purication system comprising columns packed with Q5 reactant and neutral alumina respectively (for pentane and hexanes). Deuterated toluene and benzene (Cambridge Isotopes) were stored over molecular sieves (4 Å) overnight prior to use. THF-d 8 (Cambridge Isotopes) was dried over sodium and distilled prior to use. Benzophenone (ACROS organics) was recrystallized from ethanol and vacuum-dried prior to use. Potassium bis(trimethylsilyl)amide (Sigma) was used as received. Cesium bis(trimethylsilyl)amide was prepared according to published reports. 49 Synthesis of (2-K) 4 . To a solution of Ce(TriNO x )[NH(3,5-(CF 3 ) 2 C 6 H 3 )] (400 mg, 0.438 mmol, 1.0 equiv.) in THF (2 mL) was added solid KN(SiMe 3 ) 2 (88 mg, 0.438 mmol, 1.0 equiv.). Aer stirring for 10 minutes, solid benzophenone (80 mg, 0.438 mmol, 1.0 equiv.) was added to the resulting purple solution. A yellow solid gradually precipitated from the solution upon stirring. Aer 2 h, the solid product was collected by decantation and subsequently washed with an additional portion of THF (2 mL). The product was then dried under vacuum to give analytically pure (2-K) 4 . Yield: 187 mg, 58%. 1 7 (300 mg, 0.33 mmol, 1.0 equiv.) in THF (2 mL) was added solid CsN(SiMe 3 ) 2 (96.7 mg, 0.33 mmol, 1.0 equiv.). Aer stirring for 10 minutes, solid benzophenone (60 mg, 0.33 mmol, 1.0 equiv.) was added to the resulting purple solution. A yellow solid gradually precipitated from the solution upon stirring. Aer 2 h, the solid product was collected by decantation and subsequently washed with an additional portion of THF (2 mL). The product was then dried under vacuum to give analytically pure (2-Cs) 4 . Yield: 327 mg, 39%. 1 X-ray crystallography. X-ray intensity data were collected on a Bruker APEX II CCD area detector employing graphitemonochromated Mo-K a radiation (l ¼ 0.71073 Å) at 100(1) K.</p><p>Rotation frames were integrated using SAINT, 50 producing a list of unaveraged F 2 and s(F 2 ) values which were then passed to the SHELXTL program package 51 for further processing and structure solution. The intensity data were corrected for Lorentz and polarization effects and for absorption using SADABS 52 or TWINABS. 53 The structure was solved by direct methods -ShelXS-1997. 54 Renement was by full-matrix least squares based on F 2 using SHELXL-2014. 55 Non-hydrogen atoms were rened anisotropically and hydrogen atoms were rened using a riding model.</p><p>XANES. X-ray absorption near edge structure (XANES) data at the Ce L III absorption edge were collected at the Stanford Synchrotron Radiation Lightsource (SSRL) beamline 11-2 using a Si(220) (4 ¼ 0 ) monochromator detuned to 50% and a Rhcoated harmonic rejection mirror with a cutoff energy set near 10 keV. The vertical slit height was sufficiently narrow that the energy resolution was core-hole lifetime limited. Data were collected in transmission geometry and the monochromator energy was calibrated by dening the energy of the rst inection point at the Ce L III edge of the absorption from a CeO 2 standard to be 5724.0 eV. Data were processed by subtracting a linear pre-edge background and normalizing the edge step to one.</p><p>Powder samples were prepared in an argon-lled drybox for measurement by mixing with dry boron nitride and packed into a slotted aluminum holder with aluminized mylar windows, sealed with crushed indium wire. Since the samples are air sensitive, the sealed holders were kept under argon until measurement, and exposed to air for less than one minute during transfer to vacuum. Samples were measured both at 50 K and 300 K using a liquid-helium cooled cryostat to test for temperature dependence of the resulting spectra. An easilyoxidizable sample, such as cerium tris(tetramethylcyclopentadienyl) (CeCp tet 3 ), was measured along with the samples to ensure that no O 2 had leaked into the sample holder during measurement.</p><p>XANES data were t in order to extract information about the f-orbital occupancy according to previously described methods, 11,12,14 although we report the f 0 -conguration fraction, n(f 0 ), rather than the f-orbital occupancy for reasons that are described in the Results section. Details are provided in ESI. † In particular, the major peaks at approximately 5726 eV and at 5736 eV, which are predominantly due to 2p 3/2 / 5d 5/2 transitions that have been split in the presence of the core hole due to the majority f 1 (with minority f 2 ) or the f 0 conguration fraction, respectively, have been modeled with a single Gaussian function each, even when other contributions are visible. This methodology was chosen so that the data from all the samples could be t with the same model. Consideration of using a more a complex model is provided in the ESI. † Reported errors include uncertainty from this issue.</p><p>HERFD. Ce L III edge high-energy-resolution uorescence detection (HERFD) X-ray absorption spectroscopy measurements were collected at beamline 6-2 of SSRL using a 7-crystal Johann type spectrometer 56 with Ge(331) analyzer crystals and a Si(311) monochromator. The monochromator energy was calibrated by setting the rst infection point of a Ce L III edge CeO 2 spectrum equal to 5724.0 eV. The beam size was conned to a maximum vertical slit size of 200 mm and a horizontal slit size of 400 mm. HERFD spectra were collected at the maximum intensity of the Ce L a emission line (4840.2 eV) at 50 K and 200 K to test for spectral temperature dependence. Sample preparation and cooling methods were identical to those used for XANES (see above). As with the XANES experiments, CeCp tet 3 was measured along with the samples to conrm that O 2 had not leaked into the sample holder. Data were processed by subtracting a constant pre-edge background and normalized.</p><p>HERFD spectra were simulated using the FDMNES code. 45 A convolution over the density of states was calculated for a cluster with a 7 Å radius in multiple scattering mode using a Green's function method on a muffin-tin potential. The partial projection of the density of states, including specic d-state contributions, were extracted. The core level width was reduced to 1.2 eV in order to obtain simulations comparable to the experimental HERFD data. For meaningful orbital contributions, the z-axes of the structures were aligned to be along the Ce]N or Ce]O bond in the complex structure. Throughout, orbitals are listed according to FDMNES convention, where (l,m) ¼ (2,À2) corresponds to d xy , (2,À1) to d yz , (2,0) to d z 2 , (2,1) to d xz and (2,2) to d x 2 Ày 2 .</p><p>Magnetism. Magnetic susceptibility measurements were collected using a liquid-helium cooled 7 T Quantum Design Magnetic Properties Measurement System that uses a superconducting quantum interference device (SQUID). To prepare air-sensitive samples for measurement, samples were sandwiched in a quartz tube between two pieces of quartz wool and sealed under argon. Quartz wool and tubes were baked above 200 C before introduction into a drybox for sample loading to remove any residual organic material. Pure quartz wool was measured (c 0 ¼ (À3.7 AE 0.5) Â 10 À7 emu g À1 ) and its contribution subtracted as background depending on the amount present in the sample by mass. Data were collected at 5 and 40 kG over a temperature range from 2-300 K and a two-eld correction applied in order to remove any ferromagnetic impurity as previously described. 12 Ferromagnetic impurities were minimal (#1 Â 10 À5 emu) and calculated to be <0.01% the moment of magnetite. Diamagnetic corrections were applied to the data using Pascal's constants. 57 TIP values were extracted by tting the backgroundsubtracted and two-eld corrected data to a constant + Curie-Weiss model with c ¼ C J /(T À W CW ) + c 0 , where C J is the Curie constant, W CW is the Curie-Weiss temperature indicating the strength of magnetic interactions, and c 0 is the magnitude of the TIP component to the susceptibility. The Curie-Weiss part of the model is used primarily to subtract small paramagnetic impurity contributions which are generally present for lowsusceptibility samples. Multiple measurements per sample were collected to ensure reproducibility. Error bars in the reported c 0 results were determined by averaging over multiple measurements and samples, and are typically larger than the systematic $10-15% error from the quartz-wool background subtraction combined with diamagnetic correction.</p><p>Measurement of solution magnetic properties. Solution magnetic properties were measured at ambient temperature using Evans method. The analyte solution in THF-d8 was prepared in a J-young tube. A glass capillary with the standard: 1,3,5-tris(triuoromethyl)benzene was placed in the J-young tube. Then, 5 mL of 1,3,5-tris(triuoromethyl)benzene was added to the analyte solution via micro syringe. The observed difference of the chemical shis of the standard peaks D ppm (ppm) on 1 H NMR spectra was recorded. The solution molar magnetic susceptibility c m (emu mol À1 ) was calculated using the equation:</p><p>where c is the analyte concentration in mol ml À1 , F is the measurement frequency in Hz, and Df is the difference between the standard peak shis in solution and in the capillary in Hz.</p><p>Df was calculated using the equation:</p><p>Computational studies. The electronic states of Ce imido and oxo complexes have been carefully analyzed through a systematic CASSCF study. We have followed the procedure reported in our previous studies. 12,58 In order to obtain the optimal geometries of Ce imido and oxo complexes, electronic structure calculations were carried out at the DFT level using the B3PW91 hybrid functional. The optimal geometries of all compounds have been obtained without symmetry restrictions using the hybrid B3PW91 functional. Double-zeta basis set with polarization function (6-31G(d,p)) was used for the hydrogen, carbon, nitrogen and uorine atoms, while the alkali-metals and cerium atoms were treated with Relativistic Effective Core Potentials (RECP) from the Stuttgart-Köln group. 59 Cartesian coordinates of all complexes are reported in the ESI. † Due to the importance of covalency in lanthanide complexes, we have explored the nature of the M-Ce (where M ¼ K, Rb, Cs) bonding through the Wiberg bond indices. Indices were computed using the natural bond orbital (NBO) scheme. 60 CASSCF calculations have been performed distributing six electrons in ve active orbitals. Note that these active orbitals were selected from the molecular orbitals (MO's) sets obtained at Restricted Open Shell (ROHF) level. In Fig. S22, † we present the molecular orbitals used to build the complete active space (CAS) for each Ce imido and oxo complex. All electronic structure calculations reported in this work were carried out using the computational chemistry soware Gaussian09. 61</p>

Royal Society of Chemistry (RSC)

Radical Hydrodifluoromethylation of Unsaturated C−C Bonds via an Electroreductively Triggered Two-pronged Approach

We report here the successful implementation of radical hydrodifluoromethylation of unsaturated C−C bonds via an electroreductively triggered two-pronged approach. Preliminary mechanistic investigations suggest that the key distinction of the present strategy originates from the reconciliation of multiple redox processes under highly reducing electrochemical conditions. The reaction conditions can be chosen based on the electronic properties of the alkenes of interest, highlighting the hydrodifluoromethylation of both unactivated and activated alkenes. Notably, the reaction delivers geminal (bis)difluoromethylated products from alkynes in a single step by consecutive hydrodifluoromethylation, granting access to an underutilized 1,1,3,3-tetrafluoropropan-2-yl functional group. The late-stage hydrodifluoromethylation of densely functionalized pharmaceutical agents is also presented.

radical_hydrodifluoromethylation_of_unsaturated_c−c_bonds_via_an_electroreductively_triggered_two-pr

Introduction<!>Results and Discussion<!>Conclusion

<p>The replacement of hydrogen atoms with fluorides has now become a quintessential approach for new chemical entities to regulate physicochemical and biological properties such as metabolic stability, lipophilicity, hydrogen bonding ability and bioavailability. [1][2][3][4][5][6][7] Particularly, the installation of difluoromethyl (CF2H) functionality instead of CH3 or CF3 groups in biorelevant chemical structures has been an area of intensive research in drug development due to the highly polarized C−H bond of CF2H that serves as a bioisostere of hydrogen bonding donors such as hydroxyl, thiol, and amine groups. [8][9][10][11] In this context, difluoromethylative functionalization, where a difluoromethyl anion, [12][13][14][15][16][17][18][19][20][21] carbene [22][23][24][25][26][27] or radical precursor [28][29][30][31][32][33][34][35][36][37][38][39][40][41][42] is engaged in the direct transfer of the CF2H unit, has been vigorously pursued as a strategy with great promise in organic synthesis. Particularly, radical hydrodifluoromethylation in which CF2H • and H • equivalents add across to unsaturated C−C  bonds has become a central strategy to access a relatively limited class of aliphatic hydrocarbons that contain difluoromethyl group. [43][44][45][46][47][48][49] This includes the use of redox-active CF2H radical precursors with the alkene of interest (Fig 1A, left). For example, CF2H radicals generated under oxidative conditions have frequently been utilized in hydrodifluoromethylation as well as difluoromethylative Heck-type coupling. 38,43 Additionally, several groups independently demonstrated oxidative difluoromethylative radical annulation of alkynes in the presence of aryl groups as the radical trap (Fig 1A , right). [39][40] On the other hand, CF2H radicals generated by reductive photocatalysis [44][45] or photosensitization 47 have also enabled hydrodifluoromethylation of aliphatic or electrondeficient alkenes (Fig 1B). While highly enabling, existing methods often require or involve oxidative chemical species that can presumably hamper the desired reactivity, thus limiting generality of the reaction. For example, a carbon-centered radical intermediate I, which is formed upon addition of CF2H radical into C=C bond can readily be sacrificially oxidized by the quenching cycle of photocatalysis to afford corresponding carbocation (k1), eventually leading to the formation of less-desirable difunctionalization products upon nucleophilic trapping. 10 In addition, inherent transiency of these radicals often led to the decomposition of the reaction intermediates particularly if R groups are aromatic or electron-donating substituents (k2). 47 Furthermore, super-stoichiometric amounts of CF2H radical sources with high molecular weight has been often employed that can significantly limit potential utility of these early precedents. To address these intrinsic limitations in terms of modularity and structural diversity, the development of a mechanistically distinct and more generally valid hydrodifluoromethylation approach remains a key challenge.</p><p>Herein, we describe the successful implementation of radical hydrodifluoromethylation with a wide range of unsaturated C−C bonds via an electroreductively triggered two-pronged approach (Fig 1C). The reaction conditions can be chosen based on electronic properties of the alkenes of interest, highlighting a hydrodifluoromethylation of both unactivated and activated alkenes. Notably, the newly developed protocol showcases unique reactivity towards alkynes, granting access to underutilized 1,1,3,3-tetrafluoropropan-2-yl functional group by a regioselective double hydrodifluoromethylation. To the best of our knowledge, this reactivity represents the first example of multiple difluoromethylation of alkynes. Furthermore, this electrochemical approach is applicable to late-stage functionalization and drug modification with use of commercially available CF2H sources and inexpensive H2O or PhSH as the hydrogen sources.</p><!><p>To invent a less oxidizing hydrodifluoromethylation system, we anticipated an electroreductive reaction conditions using a sacrificial anode (Fig 2). [50][51][52][53][54][55][56][57][58][59][60][61][62][63] A cathodic reduction of A would furnish a CF2H radical (B), which would afford carbon-centered radical D upon addition into alkene substrate. Two mechanistic scenarios are envisioned based upon electronic properties of the employed alkenes. This radical D would be reduced into corresponding carbanion E when the reduction potential of D [Ered(D/E)] is on par with Ered(A/A •-) (path A, if R 1 = EWG or Aryl). A subsequent protonation with water would furnish hydrodifluoromethylation product F, constituting an ECEC-type 52,[64][65] radical-polar crossover mechanism. [66][67][68][69] Alternatively, a carbon-centered radical D would directly perform hydrogenatom transfer (HAT) to form F in the presence of H donor such as thiol, when Ered(D/E) is too negative to be reduced on the cathode (path B, if R 1 = EDG or Alkyl). 72 As anticipated, a control experiment without applied current revealed that electrolytic conditions is necessary for desired reactivity (entry 2). In the absence of either water or piperidine additive, the reaction was prematurely terminated with a significant decrease in yield (entries 3−4). In both cases, we observed the formation of a zinc bridge between the cathode and anode that short-circuited the electrochemical setup. We assume that these additives would facilitate the formation of an electrochemically more stable Zn 2+ complex, thereby preventing unproductive reduction of naked Zn 2+ ion back to Zn 0 on the cathode. 73 Among a variety of amine additives tested, it was found that piperidine was most beneficial to reaction efficiency (entry 5, see Supplementary Information (SI) for full data). Importantly, changing the CF2H radical precursor from 2 to 2-py which exhibits more easily reducible potential (Ered = −1.9 V) 41 was detrimental to desired reactivity (entry 6). We hypothesized that this low reactivity is attributed to the change of reaction potential upon choice of radical precursor under constant current electrolysis. Indeed, the cathodic voltage shifts to a less reducing potential of −1.6 V in the presence of 2-py, which is not capable enough to efficiently reduce a radical intermediate 4 (Fig S5, see SI). Similarly, a variety of CF2H radical sources that are more readily reducible, [41][42] was totally ineffective for the present hydrodifluoromethylation (Table S1). On the other hand, we note that the reaction current was measured to be very high (>20 mA) and quickly drops when the electrolysis was conducted at constant potential of Ecathode = −2.2 V, leading to rapid decomposition of 2 with full recovery of Subsequently, a set of control experiments using deuterated reaction components were conducted to verify our mechanistic hypothesis (Fig 4A ). A significant amount of deuterium incorporation was observed upon employment of D2O in lieu of H2O under optimized reaction conditions, as we envisioned at the outset {equation (1)}. On the contrast, we found that the reaction with piperidine-d11 result in low deuterium incorporation, suggesting that an alternative mechanism where an  C−H bond of amine additive is engaged in H atom transfer is less conceivable {equation (2)}. 74 Further experiments using radical probe substrates were conducted (Fig 4B). Interestingly, vinyl cyclopropane 6a with higher ring opening rate constant (kPh = ~10 8 s −1 ) underwent rupture of the three-membered ring (8a), while the cyclopropyl ring in 6b (kH = ~10 5 s −1 ) remained intact after electrolysis under standard conditions (7b). 72 These results imply that the reduction of the benzylic radical intermediate (Int-8) is sufficiently fast to prevent undesired side reactions, constituting radical-polar crossover mechanism (Fig 2 , path A). Having identified the optimized reaction parameters, we next explored the substrate scope of conjugated alkenes (Table 1). A wide range of terminal styrenes that possess functional groups such as alkyls (10a−b), phenyl (10c), methoxy (10d), halides (10e−f), trifluoromethyl (10g) and cyano (10h) were well tolerated. The methyl groups in vinylmesitylene that are potentially oxidizable remained intact after electrolysis to deliver the product 10i. We found that 2-vinylnaphthalene also afforded the desired hydrodifluoromethylation product 10j in good yield. Additionally, 1,1-disubstituted styrenes such as 9k and 9l were successfully converted into the desired products in good yields. Besides terminal styrenes, more challenging internal styrenes were proved to be compatible with the current electrolytic system with exclusive regioselectivity (10m−p). It was notable to see that the regioselectivity was consistent even with the trisubstituted styrene 9p, furnishing corresponding difluoromethylated quaternary carbon center (10p). Importantly, vinyl heteroarenes such as pyrrole, thiophene and thiazole were all suitable substrates, transforming into the desired products in satisfactory yields (10q−s). Moreover, the reactivity toward biorelevant structures such as tyrosine (10t) and estrone (10u) was successfully illustrated.</p><p>We note that the choice of the radical precursor was important when highly conjugated or electron-withdrawing alkenes were subjected to the reaction. For example, 1-phenyl-1,3butadiene (9v) was efficiently converted to 1,4-hydrodifluorodifluoromethylated product 10v in the presence of more readily reducible 2-py as a radical precursor. We assumed that less negative reaction potential modulated by the use of 2-py was key to the desired polar crossover of highly conjugated radical intermediate Int-10v. Similarly, electron-deficient alkenes were smoothly participated in the reaction to give corresponding hydrodifluoromethylation products (10w−x) under room temperature.</p><p>We further expanded the scope of the current hydrodifluoromethylation reaction with respect to the aliphatic and electron-rich alkenes on the basis of initial mechanistic hypothesis in Fig 2, path B (Table 2). The plausibility of this hypothesis was verified by choosing thiophenol as a hydrogen atom donor in the presence of 2-py as a CF2H radical source. After optimization of the reaction parameters, we found that the desired reaction can be achieved using TBAPF6 as the electrolyte with the application of a constant current of 2 mA in MeCN at room temperature. In addition, the slightly increased amounts of amine (Et3N) and water additives were found to be optimal in preventing a short-circuit caused by a zinc precipitation. We have found that a wide range of terminal aliphatic alkenes was compatible with the reaction conditions (12a−i). The hydrodifluoromethylated products derived from both monosubstituted (12a−b) and 1,1-disubstituted alkenes (12c−d) gave good yields. Importantly, a difluoromethylated alkylboron product 12e could readily be obtained from a masked vinylboronic acid 11e, which can serve as a useful synthon in difluoroalkylative functionalization via cross-coupling. Moreover, the synthetic utility of the current protocol was successfully illustrated by applying it to the derivatization of biorelevant structures (12f−h) and a pharmacophore (Indomethacin derivative, 12i). More importantly, electron rich alkenes that have been previously unexplored in photocatalytic hydrodifluoromethylation such as Nvinylcarbazole (11j), enamide (11k), N-vinyllactams (11l−m) and vinyl ether (11n) underwent smooth conversion to the corresponding products in good to moderate yields under identical reaction conditions. Finally, radical clock substrate 13 afforded cyclized product 14 under standard reaction conditions. This result again highlights the radical intermediacy of the Finally, we showcased our methodology in late-stage drug modification of Ibuprofen, a popular analgesic and antipyretic in which its ameliorative derivatization has attracted constant attention in the pharmaceutical chemistry (Scheme 1). [75][76] The requisite starting material 20 was prepared with good efficiency in three steps from commercially available 2-(4hydroxyphenyl)propanoic acid (17). As anticipated, the desired hydrodifluoromethylation was smoothly proceeded under the standard conditions (21). Hydrolysis of 21 was facile under basic conditions, leading to the formation of difluoromethyl analogue of Ibuprofen (22) in 28% overall yield (5 steps). Notably, the newly developed double hydrodifluoromethylation protocol allowed conversion of alkyne 25 into corresponding geminal bis-difluoromethylation product 26. Upon treatment of 26 with base under aqueous conditions, a bis-difluoromethyl analogue of Ibuprofen (27) could readily be obtained, demonstrating an operationally simple two-track derivatization of a pharmaceutical agent from commercially available starting materials.</p><!><p>In conclusion, we developed a general electroreductive protocol to achieve hydrodifluoromethylation of a wide range of unsaturated C−C bonds by means of a twopronged strategy based upon electronic properties of the employed substrates. A key distinction of the present strategy originates from the reconciliation of multiple redox processes under highly reducing electrochemical conditions. We anticipate that this mechanistically distinct and modular protocol will enhance accessibility of a diverse suite of difluoromethylated hydrocarbons which possess high potential utility in pharmaceutical applications.</p>

ChemRxiv

Liposome destruction by a collapsing cavitation microbubble: A numerical study

Highlights•Cavitation microbubble collapse in vicinity of a liposome is investigated numerically.•Non-attached microbubbles exhibit spherical behavior regardless of vesicle presence.•Three critical modes of vesicle deformation and their driving forces are identified.•Effective non-dimensional distances δ for liposome poration and rupture are given.•A higher potential of larger bubbles for liposome destruction is identified.

liposome_destruction_by_a_collapsing_cavitation_microbubble:_a_numerical_study

Introduction<!>Theoretical background and numerical model<!>Fluid dynamics model<!>Structure dynamics model<!>Material model<!><!>Model coupling<!>Model setup<!><!>Model setup<!><!>Model setup<!><!>Results<!><!>Results<!><!>Results<!><!>Results<!><!>Discussion<!><!>Discussion<!><!>Conclusions<!>CRediT authorship contribution statement<!>Declaration of Competing Interest<!>Determination of DOPC bilayer elastic modulus E‾<!><!>Determination of DOPC bilayer elastic modulus E‾<!>Determination of DOPC bilayer material failure criteria<!>Primary failure criterion εp∗<!>Secondary failure criterion εs∗

<p>Cavitation is a physical phenomenon that can occur in liquids and is accompanied by the appearance of vaporous and gaseous cavities, commonly recognized as bubbles, that grow and collapse due to changes in ambient pressure. At first, cavitation was recognized solely as a nuisance, as it can cause unwanted vibrations, noise, and material erosion in hydraulic machinery [1]. Nevertheless, today cavitation is being utilized in various applications in the fields of chemistry [2], medicine [3], and environmental protection [4]. Hydrodynamic cavitation also poses as a promising new method for wastewater treatment [5], as it has been shown to be able to eradicate bacteria [6], inactivate viruses [7], and destroy other biological structures, such as liposomes [8].</p><p>Liposomes are lipid vesicles, that comprise of a thin spherical envelope and a liquid aqueous interior. Their envelope consists of at least one lipid bilayer, which is most commonly composed of phospholipids. Due to their similarity to a cell membrane, liposomes are used as artificial cells and model systems to study properties and stability of lipid bilayers [9]. In our recent study [8], we have demonstrated, that hydrodynamic cavitation is one of the most effective physico-chemical treatments for destroying giant 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) lipid vesicles, liposomes that measure over 1 μm in diameter and possess an envelope of a single DOPC lipid bilayer. As the effects of hydrodynamic cavitation were comparable to an ultrasound treatment, multiple questions regarding the optimization of hydrodynamic cavitation treatment for destruction of various biological structures remain. For example, different flow conditions are effective against bacteria [6] as they are against viruses [7].</p><p>Numerous potentially damaging mechanisms that accompany hydrodynamic cavitation can be speculated [10], such as strong shear flows [11], jets [12], high local temperatures [13] and pressure changes [14], shock waves [15], and formation of highly reactive free radicals [16]. In addition to this, one could also expect the bubbles to start forming and growing within the bilayer or the vesicle itself, which could lead to local bilayer poration and liposome stretching [17]. However, the contribution of the different mechanisms and their possible synergistic effects in various applications are still being explored [10]. In order to better understand the fundamental physics behind the interaction between bubbles and biological structures, such as here considered liposomes, more research concerning single bubble dynamics in vicinity of freely submerged deformable structures on a micro scale is needed. In the present case, we can distinguish between bubble-liposome interaction on three different spatial scales. Cavitation bubbles can be significantly larger (≫1μm), of a similar size, or smaller (≪1μm) than here considered vesicles (~1μm). As the smaller nano bubbles tend to exhibit a high degree of stability in regard to ambient pressure oscillations [1], we suspect micro and macro bubbles to have the greatest damage potential in the case of hydrodynamic cavitation.</p><p>In the present paper, we numerically address the former – the interaction between a single microbubble in vicinity of a DOPC lipid vesicle of a similar size, as most cavities in water have an initial diameter in the order of a few micrometers [18]. Our main interest is to determine the potentially destructive mechanical mechanisms of microbubble-liposome interaction, which would allow us to better explain the reasons behind the previously observed liposome destruction by the hydrodynamic cavitation treatment [8]. Besides giant liposomes, the chosen spatial scale in the order of 1 to 10 μm is similar to the size of a vast variety of pathogenic and potentially harmful microorganisms, such as most bacteria, certain cyanobacteria, unicellular algae, and yeast cells [19]. Additionally, viruses can reach the diameter of several hundred nanometers [19], which is still relatively close to the spatial scale considered in the present study. However, in all of the given examples, the extent of their structural similarity to DOPC vesicles has to be carefully considered if one attempts to extrapolate the findings of the present study to other biological structures.</p><p>First, we acknowledge the existing experimental and numerical research on the subject of a cavitation bubble interacting with various freely submerged structures on a micro scale. Marmottant and Hilgenfeldt [20] showed that gently oscillating single bubbles (equilibrium radius Req of 10–100 μm) excited by an ultrasound can already result in controlled deformation and lysis of DOPC vesicles of similar sizes. The authors conclude that the acoustic microstreaming, induced by the bubble, plays a key role in vesicle poration as it exerts large enough shear forces on vesicles in the flow to manipulate, deform, and rupture them. In their later work [21], the authors derived analytical predictions of vesicle shape progression and found two possible modes of liposome rupture: a) pore formation at the waist of the vesicle in the case of sufficiently large shear rates, which cause the local tension to exceed the rupture treshold and b) liposome buckling at the poles in the case of sufficient liposome elongation. Although the authors acknowledge that the shear flow is focused on a relatively small volume due to a rapid decay of the acoustic streaming velocity of the bubble, the effective bubble-liposome distance for its destruction is unfortunately not directly reported, as is might not be the most relevant for the case of continuous periodic bubble oscillations. This was addressed later by Zhou et al. [22], who acoustically excited single laser-produced microbubbles (Req=5–12μm) to induce their growth and a subsequent collapse in vicinity of a Xenopus oocyte (cell radius Rcell~ 400 μm). The authors report the effective initial bubble-cell distance for membrane poration to be 1.5Req and the mean pore radius in the order of a tenth of a micrometer. Additionally, they observed a steep decline in membrane disturbance by increasing bubble-cell distance to 2Req and no effect beyond 6Req. Effective distance for cell poration was later also reported by Le Gac et al. [23], who used a single laser-induced cavitation bubble (maximum bubble radius Rmax≈ 40 μm) in a microfluidic confinement to porate suspended human promyelocytic leukemia cells (Rcell~ 6 μm). Cell lysis probability of more than 75% was observed for cells located ⩽0.75Rmax away from the cavitation bubble center, while cells farther away than 4Rmax seemed to have been unaffected. Unfortunately, the utilized definitions of bubble-cell distance differ between both studies, which reduces the reader's ability to directly compare the obtained results. Nevertheless, based on the results of both studies we estimate the critical bubble-cell distance to be similar between both studies. We find their consistency fairly surprising, as the former researchers considered cells much larger in comparison to cavitation bubbles (Rcell/Req≈40), whereas in the latter case their size ratio was the opposite and in the range of Rcell/Rmax≈0.15. This suggests that cell poration is not largely dependent on the bubble-cell size ratio, however it may still play a role on the extent of the cell's membrane disruption – local poration versus its complete destruction. Later, Li et al. [24] reported the poration of single myeloma cells by microbubble jetting, which, while certainly interesting, is unfortunately not directly applicable to the present study, as the reason for the asymmetric bubble collapse was the nearby cell trapping structure, rather than the presence of a myeloma cell itself.</p><p>The considered topic was also addressed numerically in the past. Most of the research is based on the potential flow theory along with boundary element method (BEM) to resolve bubble dynamics, without the consideration of viscous effects and compressibility of the surrounding fluid. Both Gracewski et al. [25] and later Guo et al. [26] considered ultrasonically excited microbubbles in vicinity of a deformable sphere and a red blood cell, respectively. In most cases results showed formation of an axial jet away from the cell and a maximum areal expansion of a cell in the order of 0.1 %, which is well below the rupture threshold of a few percent [27]. On the other hand, the experimental observations and complementary BEM-based simulations of Tandiono et al. [28] show that a single laser induced microbubble (Rmax between 30 and 100 μm) can significantly stretch red blood cells (Rcell~ 4 μm), up to five times their initial size. Noticeable elongation can also occur in cells deformed by a nearby acoustically actuated bubble, which can be employed to characterize cell deformability for various diagnostic and biological purposes [29]. Although viscous and compressibility effects normally play a minor role on bubble dynamics [1], they gain importance when considering micro and nano scale bubbles, strongly collapsing bubbles, emitted shock waves, and bubble's interaction with nearby objects. Our recent numerical study [30] addressed a single collapsing microbubble (Req=1 μm) in vicinity of a freely submerged spherical particle of a similar size. There, a different numerical approach was employed, a finite volume method (FVM) along with the volume of fluid (VOF) method to resolve compressible viscous multiphase flow. The reported results show only slight deviations of bubble shape from the initial spherical for the cases of non-attached bubbles, which indicates that the formation of a strong jet towards a submerged particle on a micro scale is highly unlikely due to the cushioning effects of surface tension and a relatively low impact of the particle's presence on the bubble dynamics itself. In addition, the results show that although the gas inside a collapsing bubble can locally reach several thousand Kelvin, the thermal damage seems to be irrelevant in the cases where a similarly sized particle is not initially in direct contact with the collapsing bubble. This can be explained by the fact that the temperature field inside the collapsing bubble is not uniform and that the thickness of the thermal boundary layer is much smaller than the radius of the collapsing bubble. This further limits the search for the most likely mechanisms of destruction of liposomes and other biological structures during hydrodynamic cavitation treatment.</p><p>As already mentioned above, the present paper numerically addresses interaction between a single cavitation microbubble and a nearby DOPC lipid vesicle of a similar size. Temporal and spatial scale of the considered phenomenon is in the order of 10 ns and 1 μm, respectively. A coupled fluid–structure interaction (FSI) model is employed, which considers the influence of liposome's deformability on the surrounding fluid flow and bubble dynamics, and vice versa. Compressible multiphase flow is resolved using a FVM/VOF approach, whereas the liposome's envelope is modeled as a compliant structure through the finite element method. By choosing a continuum mechanics approach we are able to consider the system's macroscopic properties, such as areal expansion of the bilayer, viscosity, compressibility, and surface tension of fluids, etc. Through this we omit modeling of the actual molecular dynamics on a local, nano and subnanoscale, and neglect certain phenomena such as hydrophobic attraction and hydrophilic repulsion. This can be justified by the fact that even when bubble and liposome are in direct contact, the peak magnitudes of hydrophobic attraction force are expected to not exceed a few tens of Nanonewton [31], [32], [33]. Additionally, the attraction will significantly decay after their separation of only a few nanometers, as is seems to adhere to the exponential trend with the decay length in the order of a nanometer [32], [33]. For reference, the pressure force that causes a microbubble to collapse in the present case is in the order of 0.1 mN.</p><p>As the bubble collapses due to increase in ambient pressure, vesicle deforms according to the temporal development of the surrounding flow field. An emphasis is given on various modes of vesicle's deformation (bilayer stretching and wrinkling) and their corresponding driving mechanisms, from which effective distances for liposome poration and rupture are identified. Results are discussed with respect to vesicle destruction by the hydrodynamic cavitation treatment. Besides the effective bubble-liposome distance, the influence of their size ratio is also discussed.</p><!><p>In this section we present the considered physics and employed models to resolve bubble-liposome interaction. The computational domain is split into two sub-domains: a fluid domain, that consists of a gas bubble in a surrounding liquid – water, and a solid domain, that comprises of a spherical vesicle's envelope. Both sub-domains are coupled together to form the final fluid–structure interaction (FSI) model. Since liposomes contain an aqueous core, we consider it as a compressible and viscous liquid, which is therefore numerically resolved as a part of the fluid domain. It is true that based on the purpose of liposome generation, various drugs, contrast agents, genetic material, etc., can be dissolved in the aqueous solution, which could to some degree affect some of its properties, such as viscosity. However, as this is largely dependent on a specific application, we decided to consider interior as water for the sake of generality.</p><!><p>The compressible multiphase flow is modeled using a finite volume method based solver [34], which was already employed by different authors to model various cases of spherical and non-spherical bubble dynamics [35], [36], [37], [30]. The volume of fluid method is used to resolve multiphase flow, where the interface between both phases, liquid and gas, is tracked by solving a single continuity equation (Eq. (1)) for the volume fraction of water αw.(1)∂αwρw∂t+∇·αwρwVw=0</p><p>Here ρw and Vw denote the density and velocity vector of the liquid phase, i.e., water. Based on the obtained volume fraction field, we can determine the volume-averaged material properties throughout the fluid domain. Following this, a single momentum (Eq. (2)) and energy (Eq. (3)) equation is solved, from which the shared velocity V and temperature T field is obtained based on the already determined material properties.(2)∂∂t(ρV)+∇·ρV⊗V=-∇p+∇·τ+f(3)∂∂t(ρe)+∇·V(ρe+p)=∇·k∇T</p><p>Here p denotes pressure, f body forces, k thermal conductivity, and τ the viscous stress tensor that can be written for Newtonian fluids as(4)τ=μ∇V+∇VT-23∇·VI,where μ is dynamic viscosity and I the unit tensor. Total specific energy e can be written as(5)e=h-pρ+V22,where h is specific enthalpy. Additionally, the effects of surface tension are included in the procedure with a body force in the momentum equation, according to the continuum surface force model [38]:(6)Fvol=γρκg∇αg12ρg+ρw,where γ is surface tension, αg gas volume fraction field, and ρ,ρg,ρw the densities of the mixture, gas, and liquid phase, respectively. κg denotes bubble surface curvature, which is calculated as κg=∇·n|n|, where n is a bubble surface normal, obtained as a gradient of the gas volume fraction field.</p><p>A modified version of the Tait's equation of state is employed to consider the nonlinear compressibility of water:(7)ρρrefn=KKref,where the bulk modulus of water K at pressure p is calculated as K=Kref+n(p-pref). The term n is the density exponent and Kref the reference bulk modulus at the reference pressure pref. For water, we consider the values of n=7.15, and Kref=2.2 GPa, ρref=998.2 kg/m3 at pref=101325 Pa [39]. The bubble contents are modeled as air with the ideal gas law, which states(8)ρ=pRair∗T,where a specific gas constant Rair∗ of 287 J/kgK for dry air is considered. Through this, we neglect the bubble's vapor content and its mass transfer mechanisms – evaporation and condensation. Although vapor pressure is small in comparison to the internal bubble pressure and therefore does not noticeably effect the bubble dynamics in the presently considered case [30], the mass transfer on the other hand could. As the bubble collapses, its contents are compressed, which results in locally elevated temperatures and pressures. In the case of strong bubble compression, i.e. a strong collapse, a fraction of its vapor contents are lost to the ambient liquid through the process of condensation. Even though this does not significantly influence bubble dynamics until the first collapse, the amount of non-condensable gas in the bubble can affect the magnitude of bubble's rebound and its subsequent oscillations [40]. Due to the complex nature of mass transfer mechanisms, their adequate consideration remains one of the challenges up to this day. While it is true that they can be included with empirical models, this approach requires fitting of empirical model parameters to match the obtained numerical results to the experimental data for each separate case, which limits its applicability for smaller micro and nano bubbles. Even though that the present paper considers inertially collapsing bubbles, the intensity of their collapse is relatively weak (Rmax/Rmin~10) in comparison to laser induced bubbles, where the ratio between the maximum and minimum bubble ratio can exceed one hundred. Based on all this, we see the use of ideal gas law as a fair approximation for the presently considered phenomenon.</p><p>For all calculations the Pressure-implicit with splitting of operators (PISO) pressure–velocity coupling algorithm [41] was employed, along with a first order implicit temporal discretization. Pressure staggering option (PRESTO!) scheme [42] for the spatial discretization of pressure was used, while density, momentum, and energy were discretized using the second order upwind scheme. A Piecewise linear interface calculation (PLIC) geometric reconstruction scheme [43] was used as a numerical implementation of the VOF method to capture the water-bubble interface. Boundary conditions at the end of the computational domain were set to wave non-reflecting pressure outlet, which was placed reasonably far away from the bubble (~100Rmax) to minimize its influence on the bubble-liposome interaction. For a further insight into the numerical model readers are referred to [30], where the considered theoretical background and a numerical model description is given in more detail.</p><p>Additionally, we utilize a spring-based dynamic mesh smoothing method to adapt the numerical grid of the fluid domain to the bilayer's movement during each FSI coupling iteration step. In this method the edges between mesh nodes in the fluid domain are represented as a network of linearly elastic springs that obey the Hooke's law. Displacements of the internal mesh nodes are calculated according to the user specified spring constant factor and the obtained displacements of the lipid bilayer from the structural model.</p><!><p>The envelope of a liposome is modeled as a thin spherical shell structure, as the ratio of bilayer thickness τ to liposome radius RL can be neglected in comparison with unity for giant unilamellar vesicles (τ/RL<1/100). Through this we are able to consider the envelope's macroscopic properties, such as areal expansion and bending stiffness, but omit modeling of the actual molecular dynamics on a local, nano and subnanoscale. The dynamic response of a shell structure to the bubble-induced loads is resolved using a nonlinear finite element method based transient structural solver [44]. The time-varying displacements, strains, and stresses of the envelope are obtained by solving the following equation of motion(9)Mu¨+Cu˙+Ku=f,where M,C, and K represent the corresponding mass, damping, and stiffness matrices of the structure, respectively. f and u denote the load and nodal displacement vectors, whereas on overdot represents the derivative with respect to time. Large deflections are considered and true stresses and strains are considered in the model as vesicles are expected to exert a high level of compliance.</p><p>The displacement vector u can be obtained from u=x-X, where x and X correspond to the nodal position vectors in the deformed and undeformed state, respectively. From this the deformation gradient tensor F can be obtained as(10)F=I+∂uX,where I denotes the identity matrix. The deformation gradient is a second-order tensor, which can be decomposed into a product of rotation R and right stretch tensor U. Logarithmic strain tensor ε (also known as true or Hencky strain) is defined as(11)ε=lnU,and can be calculated at the locations of the element integration points through the spectral decomposition of U.</p><p>The numerical model uses a generalized Hilber-Hughes-Taylor-α method [45] for implicit time integration and a full Newton–Raphson method in which the stiffness matrix is updated at every equilibrium iteration [46]. Both methods come as one of the standard options in the utilized structure dynamics solver [44]. The shell itself was geometrically defined with a surface through its mid-plane and was discretized with second order shell elements with four in-plane integration points (element SHELL281). Three integration points through the thickness of the shell were considered, which correspond to its mid, bottom, and top-surface. The last two are offset in the normal direction of the mid-surface for the half of the shell's thickness (± τ/2).</p><!><p>The DOPC bilayer is modeled as a linearly elastic material with the equivalent elastic modulus of E‾=53.3 MPa and Poisson's ratio of ν=0.485 [47]. The considered value of elastic modulus E‾ is obtained over the whole area of interest on the σ-ε curve, bounded by the material failure criterion of εp∗ (see Appendix A). Through this, we also indirectly account for stress-softening of the material at larger strains. The bilayer thickness τ0 is taken to be 4 nm in its undeformed state. Material damping and viscoelasticity in the present case are not considered since viscous dissipation in the adjacent aqueous phase dominates the dynamic response of vesicles to macroscopic shear deformations [48]. Additionally, according to Wu et al. [49] the viscoelastic relaxation parameter for giant lipid bilayer vesicles is in the order of 0.1 s, which by several orders of magnitude exceeds the duration of herein considered phenomenon. The material density of a DOPC bilayer is set to 1009 kg/m3 [50], which is ~1.01ρw, as common for phospholipids.</p><p>Material failure and pore formation within the bilayer are not directly modeled but estimated by comparing the obtained stresses and strains in the envelope with the reported bilayer rupture thresholds from the literature. Bilayer rupture is generally considered to occur at tensions γ from 1 to 25 mN/m [51], [52], [53], which corresponds to areal strains in the order of 2 to 5%. Nevertheless, Evans et al. [54] showed that the ultimate membrane tension before its rupture is not a static material property, as it can for DOPC bilayers vary from 6 mN/m to 13 mN/m for loading rates of 0.07 mN/m/s and 25 mN/m/s, respectively. This data suggests that membrane rupture tension largely depends on the loading rate.</p><p>Additionally, the authors [54] identified two different dynamic regimes of membrane strength, a low-strength cavitation-limited and a high-strength defect-limited regime, with a transition at loading rates around 10 mN/m/s for DOPC bilayers. Due to the significantly shorter time scale of the considered phenomena in the present case (t~ 10 ns), here encountered loading rates exceed the experimentally achievable values by several orders of magnitude (present loading rate ~109-1010 mN/m/s versus the peak experimental loading rates ~102 mN/m/s [54], [55]). According to the defect-limited kinetic model for membrane failure [54], the rupture strength rises logarithmically with the loading rate, and should be between 80 and 95 mN/m for the present case. To be more precise, the model predicts a critical membrane tension of 92 mN/m for a loading time of 10 ns. Although this value is obtained for a loading rate that highly surpasses the scope of experimental observations, it fits surprisingly well with the obtained value of 90 mN/m from molecular dynamics simulations for a liquid-phase DPPC bilayer [56]. The reported critical tension corresponds to the lateral membrane stress of σ∗=20 MPa, which was later also shown by Xie et al. [57] to result in bilayer rupture in a matter of a few nanoseconds, regardless of the loading regime. Additionally, Leontiadou et al. [56] also reported that uniform lateral loading of 5 MPa is already enough to cause unstable growth of pre-existing meta-stable pores, which could be thought of as a secondary failure criterion.</p><p>Based on these values, we consider two membrane rupture criterions (see Appendix B). The primary failure criterion is related to the creation of a defect in the case of heavy lateral loading and is set at linear strain of εp∗=0.45, whereas the secondary criterion, connected to the expansion of pre-existing pores, is identified at εs∗=0.035. The likelihood of vesicle destruction can be thus estimated by the phenomenological criterion of maximum strain, where maximum principal strain in the bilayer is compared to both material failure thresholds, εp∗ and εs∗.</p><!><p>the structural solver resolves the bilayer's response to the recieved loads,</p><p>incremental displacements of the bilayer are transferred from the structural to the fluid solver,</p><p>computational mesh of the fluid domain is updated according to the received displacements,</p><p>the fluid solver computes the corresponding solution,</p><p>normal and shear forces acting on the bilayer are transferred back to the structural solver as external loads.</p><!><p>The given procedure is repeated until the desired level of data transfer convergence in reached, which is then followed by the advance of the whole system to the next time step. More precisely, root mean square convergence is monitored for both data transfers, force and incremental displacement, at both FSI participants, fluid and structural dynamics solver. The exchange of data, i.e., loads and displacements at the FSI interface, is achieved through mapping, which establishes a link between both coupling participants at the beginning of the simulation. The profile-preserving and conservative mapping procedure was used for the exchange of displacements and forces, respectively. In the former case, the mapping weights were determined through the use of shape functions, whereas in the latter case the intersect-scatter–gather algorithm was used [58].</p><!><p>As already mentioned, we consider a single cavitation bubble of radius R0=Rmax=1μm and a nearby liposome of radius RL=1μm, both at equilibrium with an initial ambient pressure of one atmosphere p∞,0=101325 Pa and an ambient temperature of 20°C. According to the Young–Laplace equation the corresponding internal bubble pressure amounts to 2.47×105 Pa. Inertial bubble collapse is induced with a sudden ambient pressure increase to p∞=107 Pa, which is a typical value one could expect to occur on here considered spatial (~1 μm) and temporal (~10 ns) scales in the case of hydrodynamic cavitation. To further clarify, this value does not represent the operating pressure of a given hydraulic system, but rather one that can locally occur within a cavitating flow, e.g. the ambient pressure of a microbubble increases due to a nearby or surrounding bubble cloud collapse [15].</p><!><p>A schematic representation of the considered setup – an initially stable bubble with radius R0 (left) in vicinity of a freely submerged spherical liposome with radius RL (right). In addition, the main regions of the liposome's envelope are also marked: proximal and distal pole (also tip, proximal and distal refer to the position in regard to the bubble) and the equator.</p><!><p>Results are reported for twelve cases of bubble-liposome interaction, corresponding to the following values of bubble-liposome distance parameter: δ = 1.15, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.75, 2, 2.25, 2.5, and 3. Additionally, cases with δ between 1 and 1.15 were also attempted, but resulted in termination even before the first bubble collapse, due to severe wrinkling and deformations of the bilayer, which led to numerical instabilities and a failure in convergence of the structural solver. Surface tension and viscosity of water were set to γw=72.8×10-2 N/m and μw=10-3 Pa·s, whereas viscosity of air is also considered with μg=1.8×10-5 Pa·s. Thermal conductivity of water and air are considered as kw=0.6 W/mK and kg=0.0242 W/mK, respectively. Boundary conditions at the end of the computational domain were set to wave non-reflecting pressure outlet with a static pressure of 107 Pa, temperature of 20°C, and volume fraction of water αw set to unity. No slip boundary condition is considered at the vesicle's shell.</p><!><p>(a) A schematic representation of the utilized wedge geometry and the whole fluid computational domain. The wedge revolves around the axis of symmetry (dashed black line) with the thickness of one computational cell. The fluid domain is divided into two sections: a static mesh region (blue) and dynamic mesh smoothing region (red). Please note that the dimensions of the computational domain on are not directly proportional to the actual geometry for the sake of figure readability. (b) The computational mesh in direct vicinity of the FSI region showing mesh of the fluid domain (blue and red fill) and the solid domain (green fill). Please note that only a fraction of computational cells is shown and the solid domain is offset along the black dashed lines for the sake of visibility.</p><!><p>The shell itself was geometrically defined with a surface through its mid-plane, which in the present case resembles a thin slice of a sphere with radius of 1 μm. It was discretized with quadrilateral second order shell elements with four in-plane integration points (element SHELL281) in the structural solver and was meshed to be conformal with the mesh of the fluid–solid interface in the fluid domain. Two triangular elements were also used, one at each pole, where the FSI interface meets the axis of symmetry. The total number of computational cells in the fluid domain ranged between 330 and 450 thousand, with 190 to 275 thousand of them being included in the dynamic remeshing process. A constant computational time step was set to 4 ps, which resulted in 6250 time steps for each case. This time step was chosen with respect to the preliminary simulations, from which the discretization errors were estimated. The corresponding results are given in the beginning of Section 3.</p><p>Approximately four coupling iterations per time step were performed on average to reach data transfer convergence target of 10-3. Convergence criteria for the fluid dynamics solver were set with the values of scaled residuals for continuity and momentum equations to 10-6, and energy equation to 10-9. In addition, custom convergence criteria were also set to 10-6 for bubble radius, along with the integrals of pressure and shear stresses over the coupling interface, i.e., both sides of the liposome's envelope. Convergence criteria in the structural solver were set as default vector norm checks (L2 norm for force, moment, rotation and infinity norm for displacement) with the specified tolerance of 10-6. Computational times varied from 7 to 11 days per case, with each being computed on a 24 core HPC cluster node. Cases with larger values of δ required longer computational times due to the larger extent of their mesh adaptation region in the fluid domain, which turned out to be the limiting factor for the use of even finer computational meshes without increasing the utilized processing power.</p><!><p>Estimation of discretization errors and Richardson extrapolation of the normalized minimum bubble radius Rmin/R0 and peak membrane εmaxm and bending strains εmaxtb of the bilayer for the case with δ=1.2. The results in the top row correspond to the finally chosen spatio-temporal resolution, whereas the rest serve as a means to estimate the magnitude of discretization errors.</p><p>∗ Based on the estimation of discretization errors from Zevnik and Dular [30].</p><p>∗∗ Estimated values according to the Richardson's extrapolation of the obtained results.</p><!><p>The results show convergent behavior towards the grid-independent solution, however it is clear that the rate of convergence is much higher for the bilayer's response, which implies that the driving process of bubble dynamics primarily determines the required spatio-temporal resolution in the present case. As the peak velocities of the bubble's wall reach the order of ~500 m/s, the finally chosen resolution of Δx=5nm and Δt=4ps ensures that the maximum Courant number in the fluid domain will not exceed 0.4 at any point of the simulation.</p><!><p>The obtained temporal progression of bubble radius R for the case with δ=1.15. The curves from all other cases overlay the one shown here, which indicates that the presence of a nearby liposome does not affect the dynamics of an unattached bubble with δ⩾1.15.</p><p>A more detailed insight into the bubble-liposome interaction for the case with δ=1.2, at the time (a) before the onset of bilayer wrinkling, (b) when bubble reaches half of its initial size, and (c) during the bubble collapse, when peak bilayer wrinkling occurs (magnified envelope showing wrinkling of the proximal liposome tip is given in the top right corner of the corresponding contour plot). Left column shows the contours of the pressure (top) and velocity (bottom) fields in the fluid along with the bubble wall (left) and liposome's envelope (right). The right column shows the spatial distribution of envelope's midpoint strains εϕm and εθm in local directions of ϕ and θ, that also correspond to both principal directions.</p><!><p>Fig. 4 (a) shows a state at the time of 4 ns. Due to a sudden increase of ambient pressure the bubble starts to shrink, which results in a sink-type velocity field in the ambient liquid. A relatively continuous pressure and velocity fields can be observed in the direct vicinity of the envelope and the boundary layer can be barely identified, as the vesicle moves and deforms with the surrounding liquid. Since velocity field decays approximately with the square of a distance from the bubble wall, a nearby vesicle is exposed to a velocity gradient field which induces spatially uneven envelope stretching in the axial direction and contraction perpendicularly to it. The latter causes compressive strains εθm throughout the whole envelope, with the peak at the proximal pole (εθm=-0.035 at ϕ=0). On the other hand, three zones can be identified for strains εϕm in the tangential direction ϕ - a compressive zone at both poles and a tensile zone at the waist.</p><p>As the bilayer at both poles experiences lateral compression, this can lead to buckling related instabilities in form of local bilayer wrinkling. We use the following first-order approximation to estimate the critical loads for the onset of buckling in thin elastic spherical shells [60]: pcr=2E‾/3(1-ν2)τ0/RL2, from which the critical strain can be derived as εcr≈(ν-1)/3(1-ν2)τ0/RL. With ν=0.485,τ0=4 nm, and RL=1μm, the critical strain is approximated at εcr=-1.4×10-3. One can notice, that compressive strains at the proximal pole (εϕm≈εθm=-0.035) are far beyond the estimated critical value of εcr=-1.4×10-3, which results in a gradual development of wrinkles through time. This can be also seen in Fig. 6 (a), which shows the temporal progression of peak shell strains at midpoint εm and top/bottom εtb through its thickness. The onset of wrinkling can be identified as a sudden split between curves εm and εtb at the time around 6 ns, which marks the occurrence of severe bilayer bending.</p><p>Fig. 4 (b) shows the state of the system at 8.6 ns, when the bubble shrinks to half of its initial size. The pressure and velocity fields still show a strong resemblance to ones in the case of an unbounded bubble with the development of a ridge-like pressure field just outside the bubble wall. The only directly visible disturbance can be observed at the proximal pole of the vesicle, where the pressure field is locally altered due to changes in the bilayer curvature and the occurrence of wrinkling.The latter also affects the spatial distribution of midpoint strains, which exhibit strong oscillations at the proximal pole. Regardless of that, the vesicle's waist continues to experience severe stretching in the circumferential direction with the peak εϕm of 0.1 appearing near the point at ϕ=π/4. Similar strain distribution can be seen at the time of 9.4 ns (Fig. 4 (c)), when overall maximum strains in the envelope occur at the proximal pole. This can be also seen on Fig. 6 (a) as a peak of εϕtb=0.31, which exceeds peak midpoint strains εϕm by almost threefold. Again, this indicates the occurrence of severe local wrinkling of the bilayer, which can be also seen in the magnified box in the top right corner of the attached contour plots. Please note, that digital magnification of subfigure (c) is required to observe wrinkling in the non-magnified instance of the proximal liposome tip due to the small wavelength of ~25 nm.</p><!><p>A more detailed insight into the bubble-liposome interaction for the case with δ=1.2, at the time (a) shortly after the bubble collapse, when shock wave reaches the liposome, (b) during shock wave propagation through the vesicle, and (c) during the second bubble collapse. Left column shows the contours of the pressure (top) and velocity (bottom) fields in the fluid along with the bubble wall (left) and liposome's envelope (right). The right column shows the spatial distribution of envelope's midpoint strains εϕm and εθm in local directions of ϕ and θ, that also correspond to both principal directions.</p><!><p>As the shock wave propagates through the envelope (Fig. 5 (b)) it causes a second local maximum of liposome stretching at the waist. This can be seen in Fig. 6, that shows temporal progression of peak εϕm at the time of ≈ 11 and 12 ns for cases (a) δ=1.2 and (b) δ=1.75, respectively. Additionally, on subfigure (a) one can also notice a drop in top/bottom shell strains εtb after bubble collapses, which implies a gradual smoothening of the wrinkles. This is to be expected, since after shock wave propagation the envelope's movement is reversed towards its initial undeformed state. Consequently, compressive strains at the proximal tip slowly decay, indicating a transition to a tensile phase. The expanding bubble reaches the maximum size of the rebound phase with R=0.523 μm at t=15.2 ns, after the pressure wave has already propagated past the vesicle. This is soon followed by the less intense second bubble collapse at tc,2=21.4 ns with Rmin,2=0.227 μm. The second collapse also results in the emission of a shock wave, although weaker in magnitude and more spatially flattened due to the slower process of the collapse. After all, a large part of the bubble's energy has been already lost with the emission of the first shock wave.</p><p>Fig. 5 (c) shows a state during the second bubble collapse. Here, a boundary layer at the vesicle's envelope is more prominent because the elastic effects of the bilayer come to effect after the first collapse. This can be also seen from the directions of velocity vectors at the bilayer, which are not directly aligned with the sink-type surrounding flow, implying the presence of vesicle's own elastic oscillations. Looking at the corresponding spatial distribution of shell midpoint strains, we can observe that the proximal tip is under uniform stretching with εm=0.034. Time-wise, it's peak magnitudes occur during the time of second bubble contraction, which corresponds to local maximums of εθm (Fig. 6) at 16.2 and 18.1 ns for cases (a) δ=1.2 and (b) δ=1.75, respectively. On the other hand, the distal pole of the vesicle remains in uniform compression for the whole simulated time of 25 ns and the values for the most part exceed the approximated critical value of εcr=-1.4×10-3. Although we noticed minor bending of this region upon a more thorough inspection of the results, it remains invisible to the naked eye and negligible in terms of through-thickness distribution of shell strains.</p><!><p>Temporal progression of peak shell strains at midpoint εm and top/bottom εtb for the case with (a) δ=1.2 and (b) δ=1.75. The directions of both principal shell strains ε1 and ε2 correspond to the local element directions of ϕ and θ, respectively.</p><p>Peak (a) shell midpoint strains εm and (b) shell strains at top/bottom εtb in both local directions ϕ and θ, marked by the subscript. Values are given in relation to the initial bubble-liposome distance δ. Curves in subfigure (a) represent the power law fits (R2>0.995) of the obtained results from the simulations (hollow circles).</p><!><p>The magnitude of bubble's liposome destruction potential can be estimated upon comparison of the obtained results with both previously determined bilayer rupture thresholds εp∗=0.45 and εs∗=0.035 (see Section 2.3 and Appendix B). First, we can conclude that none of the numerically evaluated cases with a bubble-liposome distance parameter δ between 1.15 and 3 exceeds the primary failure criterion εp∗=0.45, which is related to the creation of a defect and a subsequent membrane rupture due to heavy lateral loading. On the other hand, for small enough values of δ, we can notice that both liposome stretching (εm) and wrinkling (εtb) can surpass the secondary failure criterion εs∗=0.035 by as much as ten-fold. In the case of pre-existing pores in the bilayer can we therefore expect the damage to occur at the waist of the vesicle for δ below 1.9 and at the vesicle's pole for δ under 1.25. Additionally, bilayer wrinkling at the nearer pole could also aggravate existing defects in cases with δ⩽1.6.</p><!><p>Peak values of local bilayer extension at the vesicle's tip εtip and waist εwaist in relation to the initial bubble-liposome stand-off distance δ. Curves represent the power law fits of the obtained results from the simulations (hollow circles) with R2>0.995.</p><!><p>Both extrapolated curves meet the value of εp∗=0.45 at δ~0.75, which is in remarkably good agreement with the values reported from previous experimental studies addressing single bubble-cell interaction of Le Gac et al. [23]. They used single laser-induced cavitation microbubbles to porate suspended human promyelocytic leukemia cells and observed cell lysis probability of more than 75% for δ⩽0.75. As already mentioned in Section 1, similar effective distance for cell membrane poration was later also reported by Zhou et al. [22], who acoustically excited single laser-induced microbubbles in vicinity of a Xenopus oocyte. Despite the good match with both mentioned experimental studies, we acknowledge that the actual bubble dynamics and the deformation process of liposomes could be qualitatively different for small values of δ. This also holds for the cases where the bubble undergoes initial expansion before it collapses, which largely depends on the boundary conditions, type of cavitation, size of bubbles, etc. In this scenario, values of δ below 1.0 are not uncommon. For this reason we limit the extrapolation of data to δ>1.0 and conclude that liposomes with equilibrated envelopes, i.e., no pores are present in the bilayer, are not expected to be structurally compromised in cases with δ>1.0, when a nearby collapsing bubble is not in their direct contact. As an approximation, this could be also extended to other liposome-like biological structures, such as single suspended cells (e.g. leukocytes [23], erythrocytes [28]) or cell's organelles, although a special care would have to be given to the extent of their structural similarity to here considered giant unilamellar DOPC vesicles.</p><p>However, from here obtained results we can expect local bilayer rupture in the case of pre-existing semi-stable defects, which could remain from previous mechanical stresses. The limiting values of δ are obtained from the intersections (red asterisk) of both curves representing peak bilayer strains, εtip and εwaist, with the secondary failure criterion εs∗=0.035 at δ=1.25 and 1.9, respectively. In other words, liposomes are expected to be unaffected for δ>1.9, regardless of the existence of past bilayer defects.</p><p>As already mentioned in the end of previous section, bilayer wrinkling at the pole could also aggravate existing defects in cases with δ under 1.6, although it remains unclear whether it could also cause rupture of previously undamaged membranes. The reason for this is, that we were unable to extrapolate the obtained peak bending strains from Fig. 7 (b) below δ=1.15 with a sufficient degree of confidence. After all, a relatively small region of 1.15⩽δ⩽1.5 is available for extrapolation of a polynomial-like trend, which can yield vastly different results. Therefore this remains as one of the challenges for our further investigations, where encountered numerical instabilities in the cases with δ<1.15 should be addressed more in-depth.</p><p>Qualitatively, the presented results are also in good agreement with the findings of Marmottant and Hilgenfeldt [20], who experimentally showed that gently oscillating single bubbles excited by an ultrasound can already result in controlled deformation and lysis of DOPC vesicles of similar sizes. In their later work, [21] derived analytical predictions of vesicle shape progression and found two possible modes of liposome damage: a) pore formation at vesicle's waist in the case of sufficiently large shear rates and b) liposome buckling at the poles in the case of sufficient liposome elongation. These predictions are further supported by the present numerical investigations. In addition to this, we also identified a third relevant mode of liposome damage in the case of an inertial bubble collapse – membrane poration at the liposome's tip, which could occur during the contraction phase of a rebounding bubble.</p><p>At this point, it might be worth mentioning again, that the reported values of effective distances for liposome poration are given solely for here considered mechanical effects that result from a single bubble collapse. It is known, that strong bubble collapses are also linked to chemical effects, which are caused by the homolysis of vaporous water molecules. This leads to formation of reactive oxygen species, namely .OH and .H [16]. The formation of reactive oxigen species can affect biological structures chemically, via oxidation, although the effective distances for the poration of various cells and bio membranes in the case of a single bubble collapse are not yet known [10].</p><!><p>(a) A linear relation (R2> 0.999) between peak liposome length strains ε‾L and peak local bilayer strains at the waist εwaist among all evaluated cases. (b) Peak length strains according to surrogate fluid particle movement simulations and actual FSI simulations of bubble-liposome interaction show a good level of agreement with maximum discrepancy of less than 3%.</p><!><p>At this point, it might be worth mentioning again, that we are here considering an initially stable microbubble in vicinity of an undeformed DOPC liposome of a similar size. Vesicle stretching and deformation is driven by an inertially collapsing bubble, which occurs due to a sudden increase in ambient pressure, e.g. collapse of a nearby bubble cluster results in emission of a shock wave, which propagates past the bubble-liposome pair and causes the bubble to collapse. Nonetheless, when one considers the phenomenon of hydrodynamic cavitation, many other scenarios of bubble behavior are to be expected, depending on the development of the cavitating flow. For example, many bubbles experience significant growth due to an initial ambient pressure decrease and collapse only after they have reached a region of increased ambient pressure. Maximal size of those bubbles can thus exceed the size of nearby liposomes by a few orders of magnitude. Additionally, more than one bubble can be present in the vesicle's vicinity, which could amplify or even dampen the loads exerted on the bilayer. Certainly, a plethora of questions arises when one considers the whole range of possible scenarios. Although advanced numerical simulations can prove as an invaluable tool, especially when considering phenomena on very small spatial and temporal scales, their feasibility can be limited due to numerous constraints. For example, here utilized methodology is currently limited to the axially symmetric scenarios, as a full 3D model would be simply too computationally demanding. We face similar problems when considering much larger bubbles (Rmax≫RL) in vicinity of a micrometer-sized vesicle, as the needed spatial resolution does not scale with the bubble's size due to a nearby liposome. After all, as already mentioned in Section 2.5, computational times for presently considered scenarios already amounted to between 7 and 11 days per case, with each being computed on a 24 core HPC cluster node.</p><p>Having said that, we are still able to utilize much simpler and computationally less demanding models to estimate whether the collapse of larger bubbles also carries a potential for liposome's destruction. To achieve this, we consider the observation from the previous section, that the liposomes's envelope movement closely follows the movement of the surrounding liquid until the first collapse, which implies its inertial movement with negligible elastic effects (witch an exception of bilayer wrinkling at the proximal pole). The reason for fluid-like behavior of the vesicle can be found in a relatively high compliance of the bilayer, which is due to its inherent material characteristics, small thickness (~4 nm), and similar density to water. As peak magnitudes of liposome stretching εwaist clearly correlate with peak length strains ε‾L (Fig. 9 (a)) and both occur at the time of the first bubble collapse, we can thus predict peak values of ε‾L by only resolving pathlines of two fluid particles corresponding to the location of both liposome's poles. Through this we can omit the modeling of a full FSI system and only resolve a simpler case of spherical bubble collapse, since the presence of a nearby liposome does not seem to significantly affect the dynamics of a collapsing bubble of a similar size (see Section 3). This significantly simplifies the considered problem at hand and shortens the required computational times per case by more than a ten-fold.</p><p>A comparison of the obtained results between the actual FSI and surrogate simulations is given in Fig. 9 (b). A very good agreement between both curves can be observed, with maximum relative discrepancy of less than 3%, which occurs at the smallest considered value of δ=1.15. For the most part, we attribute the difference to the emergence of bilayer wrinkling, which cannot be predicted by only resolving fluid flow. Further simplifications were attempted by resolving the Rayleigh-Plesset equation, without consideration of the water's compressibility and emission of shock waves. This even further reduced the required computational times to a matter of seconds per case. Although it resulted in surprisingly good agreement with the obtained results from FSI simulations for δ⩽1.5 (discrepancies within few percents), the relative error began to increase with larger values of δ due to the neglection of compressibility effect in form of continued "liposome" stretching during the time of shock wave propagation from the collapsing bubble to the liposome's proximal tip.</p><!><p>Estimated values of peak liposome length strains ε‾L with respect to the non-dimensional liposome-bubble stand-off distance δ and their size ratio RL/R0, which imply that larger bubbles carry a higher potential for causing stretching-induced liposome destruction. The corresponding results from FSI simulations are included with solid spherical markers.</p><!><p>Hydrodynamic cavitation poses as a promising new method for wastewater treatment as it has been shown to be able to eradicate bacteria, inactivate viruses, and destroy other biological structures, such as liposomes. Although engineers are already commercializing devices that employ cavitation, we are still not able to answer the fundamental question: What precisely are the mechanisms of how bubbles can clean, disinfect, kill bacteria and enhance chemical activity?.</p><p>The aim of the present paper was to research the dynamics of a single cavitation microbubble (Req=1μm) in vicinity of a DOPC lipid vesicle of a similar size, which allowed for a better explanation of the mechanisms behind the recently observed liposome destruction by the hydrodynamic cavitation treatment [8]. Due to small spatial (~1μm) and temporal (~10 ns) scales of the considered phenomenon a purely numerical approach was used. A coupled fluid–structure interaction model was employed, which considered the influence of liposome's deformability on the surrounding fluid flow and bubble dynamics, and vice versa. Compressible multiphase flow was resolved using a finite volume/volume of fluid method approach, whereas the liposome's envelope was modeled as a compliant structure through the finite element method. Simulations were carried out for various cases of bubble-liposome standoff distance δ between 1.15 and 3. The required computational times varied between 4000 and 6500 core-hours, where cases with larger values of δ required longer computational times due to the larger extent of their mesh adaption region in the fluid domain.</p><p>Regardless of the nearby liposome, the results show spherical bubble behavior, which points towards the negligible effect of vesicle's presence on the dynamics of a nearby unattached (δ>1) and similarly sized cavitation bubble. As the bubble collapses due to increase in ambient pressure, vesicle deformation is driven according to the temporal development of the surrounding flow field. Three critical modes of vesicle deformation were identified and temporally placed in relation to their corresponding driving mechanisms: (a) unilateral bilayer stretching at the waist of the liposome during the first bubble collapse and subsequent shock wave propagation, (b) local wrinkling at the tip of the liposome until the bubble rebounds, and (c) bilateral bilayer stretching at the tip of the liposome during the phase of a second bubble contraction. Here, unilateral and bilateral stretching refer to the local in-plane extension of the bilayer in one and both principal directions, respectively.</p><p>Based on the obtained results, effective distances for liposome poration and rupture were identified, which are in good agreement with previous bubble-cell interaction studies. Liposomes with equilibrated envelopes, i.e., no pores are present in the bilayer, are not expected to be structurally compromised in cases with δ>1.0, when a nearby collapsing bubble is not in their direct contact. However, the critical dimensionless distance for vesicle poration and rupture is identified at δ=1.9 for the case of an envelope with pre-existing pores. In other words, liposomes are expected to be unaffected for δ>1.9, regardless of the existence of past bilayer defects. Results were further discussed with respect to vesicle destruction by the hydrodynamic cavitation treatment, where the influence of bubble-liposome size ratio was also addressed. A higher potential of larger bubbles for causing stretching-induced liposome destruction was identified, which can be also used to explain previously observed efficiency of supercavitation for eradication of bacteria [6].</p><!><p>Jure Zevnik: Conceptualization, Formal analysis, Writing - original draft, Writing - review & editing, Visualization. Matevž Dular: Conceptualization, Writing - review & editing, Supervision, Funding acquisition.</p><p>Jure Zevnik: Conceptualization, Formal analysis, Writing - original draft, Writing - review & editing, Visualization. Matevž Dular: Conceptualization, Writing - review & editing, Supervision, Funding acquisition.</p><!><p>The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.</p><!><p>Membrane tension in lipid bilayers can be generally written as [61]:(12)γ=2Π01-a0al3,where Π0 is a tension-free surface pressure, which amounts to 50 mN/m for DOPC bilayers. Here, a0=0.68nm2 denotes the tension-free area per lipid molecule and al area per molecule corresponding to a given value of membrane tension γ. In the present case, it is more practical to express membrane tension in relation to areal strain εA=al/a0-1:(13)γ(εA)=2Π01-11+εA3.</p><p>Membrane tension can be also obtained by integration of average membrane stresses in the lateral direction σ throughout its thickness τ, which yields(14)γ=(σ1+σ2)2τ0(1+ετ),where τ0=4 nm is the initial bilayer thickness and ετ its strain corresponding to εA. Additionally, σ1 and σ2 represent both in-plane principal stresses. For the case of uniaxial stretching, where σ1=σu and σ2=0, the areal strain εA and strain throughout the bilayer's thickness ετ can be written as(15)εAu=(1+εu)(1-νεu)-1andετ=-2ν1-νεu,where ν=0.485 denotes the Poisson's ratio of the bilayer [47]. Considering this alongside Eq. (13), (14), we can obtain the final relation between linear strain εu and the corresponding stress σu for the case an uniaxially loaded bilayer:(16)σu=4Π0τ0(1-νεu)1-1(1+εu)(1-νεu)3.</p><p>The material elastic modulus E can be obtained as the initial slope of the obtained σu-εu curve, which amounts to 77.3 MPa. This is equivalent to the area stretch modulus kA=Eτ0/2(1-ν)=300 mN/m, which is reported for a full range of accessible tensions in the existing literature and is also applicable to the cases of fast hyper-stretching of the DOPC bilayer [62], [61]. One can also notice that the relation in Eq. (16) shows the inherent material nonlinearity of the DOPC bilayer, which can be seen through stress-softening at large strains.</p><!><p>The obtained σu-εu curve for the DOPC bilayer and the corresponding linearized material model E‾∊u.</p><!><p>Considering the primary failure criterion of εp∗=0.45 (see Appendix B) as the upper bound of ε∗, we obtain the equivalent elastic modulus of E‾=53.3 MPa.</p><!><p>Relevant failure criteria are established based on the critical lateral stresses and the corresponding surface tensions from the existing literature. The reported values are translated to the case of uniaxial loading. This allows for the use of phenomenological failure criteria in a general loading scenario, which does not necessarily result in uniform membrane tension.</p><!><p>According to the primary bilayer failure criterion, bilayer rupture is limited by the creation of a defect due to heavy lateral loading. For an equibiaxial loading case, where σ1=σ2, the critical value for lipid bilayers is in the order of 20 MPa [56], [54], [57]. As defect formation and subsequent membrane rupture is related to the strength of intramolecular forces within the bilayer, we consider the same value of σp∗=20 MPa as the ultimate tensile strength on a temporal scale of ~10 ns. The corresponding ultimate tensile strain from Eq. (16) amounts to εp∗=0.45.</p><!><p>In this case, material failure is limited by the expansion of pre-existing meta-stable pores that can emerge from prior mechanical stresses. When pores and defects are already present in a bilayer, much lesser loads are needed to cause their uncontrollable expansion and subsequent membrane rupture. For the case of equibiaxial extension, the critical value is in the order of 5 MPa [56], which corresponds to the linear strain of 3.5%. As material failure is dependent on the expansion of the pore itself, we consider the corresponding strain of εs∗=0.035 as the secondary failure criterion on a temporal scale of ~10 ns. According to Eq. (16) the critical strain corresponds to σs∗=2.57 MPa.</p>

PubMed Open Access

Degradation of Soluble and Fibrillar Amyloid \xce\xb2-Protein by Matrix Metalloproteinase (MT1-MMP) in-Vitro\xe2\x80\xa0

The progressive accumulation of \xce\xb2-amyloid (A\xce\xb2) in senile plaques and in the cerebral vasculature is the hallmark of Alzheimer\xe2\x80\x99s disease and related disorders. Degradation of A\xce\xb2 by specific proteolytic enzymes is an important process that regulates its levels in brain. Matrix metalloproteinase 2 (MMP2) was shown to be expressed in reactive astrocytes surrounding amyloid plaques and may contribute to A\xce\xb2 degradation. Membrane type-1 (MT1)-MMP is the physiological activator for the zymogen pro-MMP2. Here, we show that in addition to MMP2, its activator MT1-MMP is also expressed in reactive astrocytes in regions with amyloid deposits in transgenic mice. Using a Cos-1 cell expression system we demonstrated that MT1-MMP can degrade exogenous A\xce\xb240 and A\xce\xb242. A purified soluble form of MT1-MMP degraded both soluble and fibrillar A\xce\xb2 peptides in a time dependent manner yielding specific degradation products. Mass spectrometry analysis identified multiple MT1-MMP cleavage sites on soluble A\xce\xb240 and A\xce\xb242. MT1-MMP-mediated A\xce\xb2 degradation was inhibited with the general MMP inhibitor GM6001 or the specific MT1-MMP inhibitor tissue inhibitor of metalloproteinases-2. Furthermore, in situ experiments showed that purified MT1-MMP degraded parenchymal fibrillar amyloid plaques that form in the brains of A\xce\xb2 precursor protein transgenic mice. Together, these findings indicate that MT1-MMP possesses A\xce\xb2 degrading activity in vitro.

degradation_of_soluble_and_fibrillar_amyloid_\xce\xb2-protein_by_matrix_metalloproteinase_(mt1-mmp)_

<!>Reagents and Chemicals<!>A\xce\xb2PP Transgenic Mice<!>Immunofluorescent Labeling<!>Gelatin Substrate Zymography<!>A\xce\xb2 Degradation in Cos-1 Cells Expressing Human MT1-MMP<!>Quantitative Immunoblotting<!>A\xce\xb2 ELISA Analysis<!>Purification of Soluble MT1dTM Protein<!>In-vitro Soluble A\xce\xb2 Degradation<!>Mass Spectrometry<!>In-vitro Fibril A\xce\xb2 Degradation<!>Transmission Electron Microscopy<!>In-Situ Fibrillar Amyloid Plaque Degradation<!>Statistical Analysis<!>MT1-MMP and MMP2 are expressed in brain regions with prominent cerebral microvascular fibrillar A\xce\xb2 deposits in Tg-SwDI mice<!>MT1-MMP and MMP2 are selectively expressed in reactive astrocytes in brain regions with microvascular fibrillar amyloid deposits in aged Tg-SwDI mice<!>Exogenous A\xce\xb240 and A\xce\xb242 degradation in Cos-1 cells expressing human MT1-MMP<!>In-vitro A\xce\xb240 and A\xce\xb242 degradation by purified soluble MT1-MMP<!>In-vitro fibrillar A\xce\xb2 degradation by purified soluble MT1-MMP<!>MT1dTM degrades parenchymal fibrillar A\xce\xb2 plaques in situ<!>DISCUSSION<!>MT1-MMP and MMP2 are expressed in regions of fibrillar A\xce\xb2 accumulation in Tg-SwDI mouse brain<!>MT1-MMP and MMP2 are expressed in reactive astrocytes near fibrillar microvascular amyloid deposits in Tg-SwDI mouse brain<!>Activation of pro-MMP2 by MT1-MMP expressed in Cos-1 cells<!>A\xce\xb240 and A\xce\xb242 are degraded by MT1-MMP expressed in Cos-1 cells<!>A\xce\xb240 and A\xce\xb242 degradation by soluble MT1dTM<!>A\xce\xb240 degradation by soluble MT1dTM is inhibited by GM6001 and TIMP2<!>Analysis of MT1dTM mediated A\xce\xb2 cleavage fragments on acid/urea gels<!>MALDI-TOF MS analysis of A\xce\xb2 fragments released from by purified MT1dTM<!>Fibrillar A\xce\xb2 degradation by soluble MT1dTM<!>In Situ brain fibrillar amyloid plaque degradation by soluble MT1dTM<!>

<p>A key pathological feature of Alzheimer's disease (AD)1 is the progressive accumulation of β-amyloid (Aβ) in senile plaques and the cerebral vasculature. Aβ is derived from amyloidogenic processing of the amyloid precursor protein (AβPP), which involves sequential cleavage by β-secretase and γ-secretase (1–2). The steady-state level of Aβ peptides in the brain is controlled by a balance between production and clearance (3). Impaired clearance of Aβ peptides is likely important in the pathogenesis of AD, especially in the more common sporadic form. Several major pathways for Aβ clearance have been identified including receptor-mediated cellular uptake, blood-brain barrier transport (4–5), and direct proteolytic degradation.</p><p>Several proteinases/peptidases which can degrade Aβ have been reported including neprilysin (6–7), insulin-degrading enzyme (8), plasmin (9), endothelin-converting enzyme (10), angiotensin-converting enzyme (11), myelin basic protein (12), matrix metalloproteinase (MMP) 2 (13–14), and MMP9 (15). Regarding MMP2, it has been reported to cleave Aβ peptides at several sites (14). MMP2 expression and activity is induced in cultured human cerebrovascular smooth muscle cells in response to pathogenic Aβ (16). Also, in astrocytes the activity of MMP2 is increased in the presence of Aβ (17–20). Reactive astrocytes are found in regions with fibrillar amyloid deposits in brain tissue of human AD subjects and of APPsw (Tg-2576) transgenic mice, and have been shown to participate in the Aβ degradation in the extracellular space (20–23).</p><p>MMP2 is released in a latent form (pro-MMP2) that requires activation by membrane-type 1 (MT1)-MMP (24). MT1-MMP was the first MMP to be identified as an integral membrane protein with a single transmembrane domain and a short cytoplasmic C-terminal tail (25). MT1-MMP is inhibited by the endogenous tissue inhibitor of MMPs 2 (TIMP-2) and recruits pro-MMP2 forming a ternary complex. Then, adjacent uninhibited MT1-MMP cleaves the tethered pro-MMP2 (26). MT1-MMP is expressed in a variety of tissues including brain (27). In addition to activating pro-MMP2, MT1-MMP is involved in the breakdown of various extracellular matrix components including collagens, laminins, fibronectin, and proteoglycans (28). This function enables it to participate in numerous normal biological processes, such as reproduction, embryonic development, wound healing, angiogenesis, and apoptosis (29) or in pathological processes, such as rheumatoid arthritis, cardiovascular disease, tumor invasion and metastasis (30).</p><p>Expression of MT1-MMP can be induced in human glioma cells and human cerebrovascular smooth muscle cells in response to Aβ (23). It was reported that MT-MMPs induce cleavage and shedding of the AβPP ectodomain and that one of these cleavage sites is within the Aβ peptide region (31). However, any role for MT1-MMP in the degradation of Aβ peptides and the pathology of AD is unknown. In the present study, we show that, like MMP2, MT1-MMP is expressed in reactive astrocytes in regions with fibrillar microvascular amyloid deposits in a human AβPP transgenic mouse model. Subsequently, we show that MT1-MMP expressed in Cos-1 cells is capable of degrading soluble Aβ40 and Aβ42 peptides. A purified soluble truncated form of MT1-MMP also degraded soluble and fibrillar Aβ in-vitro. Mass spectrometry analysis identified multiple MT1-MMP cleavage sites on soluble Aβ40 and Aβ42. Furthermore, in situ experiments show that purified soluble MT1-MMP can degrade parenchymal fibrillar amyloid plaques that form in the brains of AβPP transgenic mice. Together, these data indicate that MT1-MMP possesses Aβ degrading activity.</p><!><p>Synthetic Aβ40 and Aβ42 peptides were synthesized by solid-phase Fmoc (9-fluorenylmethoxycarbonyl) amino acid chemistry, purified by reverse phase high performance liquid chromatography, and structurally characterized as previously described (32). Thioflavin-S (Th-S), Thioflavin-T (Th-T) and TIMP2 were purchased from Sigma-Aldrich (St. Louis, MO). The general MMP inhibitor GM6001 was purchased from Calbiochem (La Jolla, CA).</p><!><p>Generation of Tg-SwDI transgenic mice on a pure C57BL/6 background was previously described (33). These mice express low levels of human Swedish/Dutch/Iowa mutant AβPP in neurons under control of the mouse Thy1.2 promoter. Tg-SwDI mice accumulate extensive cerebral microvascular fibrillar amyloid. Brain tissues from homozygous 24 months old Tg-SwDI and similarly aged control non-transgenic mice were used in this study. In other experiments brain tissues from 18 months old Tg2576 mice, a model of AD-like parenchymal fibrillar amyloid plaques, were used (44).</p><!><p>Immunofluorescent stainings were performed on paraffin sections as recently described (33). The following primary antibodies were used for immunostaining: monoclonal antibody 66.1 (1:300), which recognizes residues 1 to 5 of human Aβ (34); rabbit polyclonal antibody to collagen type IV (1:100; Research Diagnostics Inc., Flanders, NJ); mouse monoclonal antibody to glial fibrillary acidic protein (GFAP) for identification of astrocytes (1:300, Chemicon); mouse monoclonal anti-keratan sulfate antibody for the detection of activated microglia (clone: 5D4, 1:200, Seikagaku Corporation, Japan); rabbit polyclonal antibody to MT1-MMP (1:100; Triple Point Biologics Inc., Forest Grove, OR); rabbit polyclonal antibody to MMP2 (1:100; Sigma). Primary antibodies were detected with goat anti-rabbit IgG (Alex 594; 1:2500; Molecular Probes Inc., Eugene, OR) or/and donkey anti-mouse IgG (Alex 488; 1:2500; Molecular Probes, Inc., Eugene, OR). Th-S staining for fibrillar amyloid was performed as described (35).</p><!><p>Cos-1 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (Gemini Bio-Products, Woodland, CA). Full-length MT1-MMP and pro-MMP2 in pcDNA3.1 plasmids were the kind gifts of Dr. Jian Cao (Department of Medicine, Stony Brook University, NY, USA). Triplicate near confluent cultures were transfected with plasmids for expression of pro-MMP2, MT1-MMP or both pro-MMP2 and MT1-MMP using FuGENE 6 (Roche, Indianapolis, IN). Transfected cells were incubated in serum-free culture media and 72 h. The conditioned culture media samples were collected and aliquots were electrophoresed on 8% SDS-polyacrylamide gels containing 0.1% gelatin at 100 V for 2 h at 22°C. The gels were removed and incubated in rinse buffer (50 mM Tris-HCl, pH 7.5, 200 mM NaCl, 5 mM CaCl2, 2.5% Triton-X 100) for 3 h with several changes, washed 3 × 10 min with ddH2O, then incubated in assay buffer (50 mM Tris-HCl, pH 7.5, 200 mM NaCl, 5 mM CaCl2) overnight at 37°C, washed 3 × 10 min with ddH2O, stained with 0.25% Coomassie Brilliant Blue R-250 and then destained. Gelatinolytic MMP activity was observed as clear zones of lysis in the gels.</p><!><p>Triplicate near confluent cultures were transfected with purified empty pcDNA3.1 plasmid DNA or full-length MT1-MMP in pcDNA3.1 DNA by FuGENE 6 treatment (Roche, Indianapolis, IN), followed by addition of 2 μg/ml of Aβ40 or Aβ42 in serum-free media for 48 h. The culture media samples were collected and cell lysates were prepared. Aβ in the cell culture media samples was quantitatively analyzed by immunoblotting and sandwich ELISA analysis as described above.</p><!><p>Samples containing MT1-MMP or Aβ were added directly into SDS-PAGE sample buffer, and stored at −70°C. Aliquots were loaded onto 12% or 10–20% polyacrylamide gels, electrophoresed and transferred onto Hybond-ECL nitrocellulose membranes (Amersham, Arlinton Heights, IL) at 100 V for 1.5 hour at RT. Membranes were blocked in 5% milk/PBS/0.05% Tween20 (PBS-T) for 1 h at RT. Primary antibodies were added (RP1-MMP14 for MT1-MMP; mAb20.1 for Aβ) for 1 h at RT, washed 3 × min with PBS-T. Horseradish peroxidase-conjugated mouse sheep anti-rabbit or anti-mouse IgG (1:5000 Amersham-Pharmacia, Piscataway, NJ), and washed 3 ×5 min with PBS-T. Bands were visualized using the ECL detection method (Amersham-Pharmacia, Piscataway, NJ). Quantitation of MT1-MMP or Aβ bands was performed using a VersaDoc Imaging System (BioRad, Hercules, CA) and the manufacturer's Quantity Oneton software.</p><!><p>The levels of soluble Aβ40 and Aβ42 peptides were measured using a quantitative sandwich ELISA as previously described (33).</p><!><p>The cDNA for a soluble, truncated form of MT1-MMP encoding residues Met1-Gly535 that lack the carboxyl-terminal transmembrane and cytosolic domains of full length MT1-MMP (MT1dTM) in pSG5 expression vector was the kind gift of Dr. Jian Cao (Department of Medicine, Stony Brook University, NY, USA). Two hundred ml of serum-free conditioned media from Cos-1 cells overexpressing soluble MT1dTM were passed through gelatin-agarose (Sigma-Aldrich, St. Louis, MO) to remove any gelatinases and then concentrated using an Amicon ultrafiltration unit (NMWL 5000 membrane) (Millipore, Bedford, MA). The enzymatic activity of purified MT1dTM was determined using the specific substrate- Mca-Pro-Leu-Gly-Leu-Dpa-Ala-Arg (Bachem, California, CA).</p><!><p>Synthetic Aβ40 or Aβ42 were first dissolved in DMSO to a concentration of 1 mg/ml. 40 nM of purified MT1dTM were incubated with 1 μM of synthetic Aβ40 or Aβ42 in zymogen buffer (50 mM Tris-HCl, pH 7.5, 200 mM NaCl, 5 mM CaCl2) at 37 °C for specific lengths of time. The Aβ levels were measured in the samples by SDS-PAGE on 10–20% polyacrylamide Tris-Tricine gels and subsequent quantitative immunoblotting (as described above). In some experiments the selective MMP inhibitors GM6001 (10 μM; Calbiochem) or TIMP2 (40 nM; Sigma) were added.</p><p>To visualize MT1-MMP generated Aβ cleavage products N-terminal biotin labeled Aβ40 or Aβ42 were incubated with purified MT1dTM for 24 h at 37 C. The samples were diluted in the sample buffer containing 9M urea/5% acetic acid and methyl green. For analysis an acid/urea 22% polyacrylamide gel was prepared and pre run anode to cathode at 250 V for 30 min at 4°C in 5% glacial acetic acid running buffer (36). Following the pre run, the samples were loaded on the gel and electrophoresed at 4 ° from anode to cathode with increasing the voltage every 15 min as follows: 25, 50, 100, 200 volts and then 275 volts for 15 h until the end of the run. Prior to transfer, the acid/urea gel was neutralized in a glass tray by washing 5x with Tris-HCl/glycine transfer buffer on a rocking platform for 15 min. Then the gel was transferred to PVDF membranes be electroblotting for 2.5 hr (80V) at 4°C. After transfer, the membrane was boiled in PBS for 5 min in a glass dish and was cooled down in PBS. The membrane was blocked in 5% milk/PBS/0.05% Tween20 (PBS-T) for 1 h at RT. The membrane was incubated with streptavidin-horseradish peroxidase (1:5000 dilution) for 1 h at RT, and washed 3 × 5 min with PBS-T. Bands were visualized using the ECL detection method as described above.</p><!><p>40 nM purified MT1dTM was incubated with 1 μM synthetic Aβ40 or Aβ42 in zymogen buffer (50 mM Tris-HCl, pH 7.5, 200 mM NaCl, 5 mM CaCl2) at 37°C for 2 days. After incubation, the samples were dried in a rotary evaporator (Savant, Farmingdale NY), suspended in 20 μl of 0.1% TFA, ZipTipped using μC18 tips (Millipore, Milford, MA) and then eluted to the target. The addition of 1 μl of matrix consisting of acetonitrile/0.1% trifluoroacetic acid containing ∞-cyano-4-hydroxy cinammic acid (CHCA, 5 mg/ml) was dried on the sample plate. Samples were run on a Voyager-DE STR (Applied Biosystems, Framingham, MA) using a matrix assisted laser desorption ionization - time of flight (MALDI-TOF) mass spectrometer system operated in the reflector mode unless otherwise indicated. The mass scale (m/z 500–5000) was calibrated with a mixture of peptides or internal calibration was performed using a matrix ion at m/z 568.1330 and Aβ42 peptide amino acid 1–13 m/z 1561.6672. For samples acquired in the linear mode, 1 μl was dissolved in 10 μl of a 50% solution of acetonitrile/0.3% trifluoroacetic acid containing sinapinic acid (10 mg/ml) and dried on the sample plate. The mass scale (m/z 1000–25000) was calibrated with myoglobin (400 fM/μl).</p><!><p>To prepare amyloid fibrils, 5 mM Aβ42 in DMSO was diluted in PBS to 100 μM, vortexed for 30 sec, and incubated at 37 °C for 5 days (37). Triplicate samples of 10 μM of aged fibrillar Aβ was then incubated with 100 nM of purified MT1dTM in absence or presence of 100 μM GM6001 at 37 °C for 5 days. After digestion, the remaining fibrillar Aβ was quantitated using a Th-T fluorescence binding assay. Briefly, 5 μl of 100 μM Th-T was added to 100 μl of sample, mixed and incubated at RT in the dark for 10 min. Th-T fluorescence, an indicator of fibril Aβ binding, was measured at λex of 446 nm and λem of 490 nm.</p><!><p>Sample aliquots were deposited onto carbon-coated copper mesh grids. Sample grids were allowed to stand for 60 sec, and excess solution was wicked away. Sample grids were then negatively stained with 2% (w/v) uranyl acetate and allowed to dry. The samples were viewed with an FEI Tecnai 12 BioTwin transmission electron microscope at 80 kV, and digital images were taken with an Advanced Microscopy Techniques camera.</p><!><p>For this analysis the well-characterized Tg2576 (APPsw) mouse model of AD that develops abundant fibrillar amyloid pathology (38) was used. Brains were removed from anesthetized 18 months old Tg2576 mice after perfusion with cold saline and snap-frozen on dry ice. Five-μm cryostat sections will be collected on slides. Every other section was flipped 180° so that identical faces of adjacent sections were exposed (37). Paired adjacent sections (one incubated with zymogen buffer, the other with 100 nM purified MT1dTM in absence or presence of 100 μM GM6001) in triplicate were incubated at 37°C for 5 days, stained with thioflavin-S (ThS), and then imaged with fluorescence microspcopy. The parenchymal plaque amyloid area of ThS fluorescence was determined using image analysis software (Image J). Fractional area was compared between paired sections.</p><!><p>Data were analyzed by Student's t-test at the 0.05 significance level.</p><!><p>Previously, MMP2 was found increased in reactive astrocytes adjacent to parenchymal amyloid plaques in aged AβPP transgenic mouse brain (39). MMP2 expressed by reactive astrocytes is implicated in extracelluar Aβ catabolism (39). We have generated the Tg-SwDI mouse model, which develops early-onset and progressive accumulation of regional cerebral microvascular fibrillar amyloid deposition (33). Th-S staining of brain sections of aged Tg-SwDI mice revealed extensive fibrillar Aβ accumulation in the microvessels of the thalamus, but not in the cortex (Fig. 1A,B). MMP2 is also expressed in cells in the thalamus where fibrillar microvascular Aβ accumulates, but not the cortex (Fig. 1C,D). MMP2 is expressed as an inactive zymogen (pro-MMP2) requiring proteolytic activation by MT1-MMP. Labeling for MT1-MMP also showed strong expression by cells in the thalamus where fibrillar microvascular Aβ accumulates, but not the cortex (Fig. 1E,F).</p><!><p>To identify the cell type that expresses MMP2 and MT1-MMP near microvascular amyloid deposits the Tg-SwDI mouse brain sections were double immunolabeled for GFAP to detect reactive astrocytes and either MMP2 or MT1-MMP (Fig. 2). Immunolabeling for MMP2 and its activator MT1-MMP strongly co-localized with GFAP-positive cells. Immunolabeling for activated microglia in these tissue sections failed to show co-localization with MMP2 or MT1-MMP (data not shown). These data show that like MMP2, its activator MT1-MMP is selectively expressed in reactive astrocytes near cerebral microvascular fibrillar amyloid deposits in aged Tg-SwDI mouse brain.</p><!><p>We next determined if MT1-MMP, like MMP2, could play a role in Aβ degradation using a cell culture expression system. Cos-1 cells were chosen since they do not normally express either MT1-MMP or MMP2. Therefore, Cos-1 cells were transfected to express pro-MMP2 alone, MT1-MMP alone or both pro-MMP2 and MT1-MMP. Post transfection, the cells were incubated with serum-free media for an additional 48 h. The cell lysates were collected and analyzed by immunoblotting using an anti-MT1-MMP antibody demonstrating protein expression in the cells transfected with the MT1-MMP plasmid (Fig. 3A). The culture media samples were collected and subjected to gelatin zymography to assay for MMP2 activities (Fig. 3B). The zymography assayed showed that pro-MMP2 was only expressed in the Cos-1 cells transfected with the pro-MMP2 plasmid. Whereas pro-MMP2 alone migrated at ≈72 kDa the co-transfection with pro-MMP2 and MT1-MMP exhibited activated MMP2 which migrated as a doublet at ≈66 kDa. These experiments demonstrated that MT1-MMP expressed in Cos-1 cells was enzymatically active.</p><p>To determine if MT1-MMP expressed in Cos-1 cells could degrade Aβ, the cells were transfected with either empty plasmid vector (pcDNA3.1) or the MT1-MMP plasmid vector. Post transfection; the cells were incubated with 2 μg/ml of freshly prepared soluble Aβ40 or Aβ42 in serum-free media for an additional 48 h. The cell lysates were collected and analyzed by immunoblotting using the anti-MT1-MMP antibody confirming MT1-MMP protein expression in the transfected cells (Fig. 4A). Although small amounts of Aβ peptides were found associated with the cells present in the cell lysates there was no difference in the amounts between control and MT1-MMP expressing Cos-1 cells (data not shown). The media samples were collected and analyzed for Aβ40 and Aβ42 peptide levels by immunoblotting using monoclonal anti-Aβ (Fig. 4B,D, respectively) and by quantitative ELISA measurements (Fig. 4C,E, respectively). These results indicate that both Aβ40 and Aβ42 were strongly reduced by about 50% and 70%, respectively, in MT1-MMP transfected Cos-1 cells.</p><!><p>MT1-MMP is normally expressed as a membrane bound enzyme. However, a soluble transmembrane domain-lacking form of MT1-MMP (MT1dTM) can be used to study the proteolytic function of the enzyme in solution. Therefore, we used purified MT1dTM protein to investigate if Aβ peptides could be degraded in-vitro. Aβ40 or Aβ42 (1 μM) was incubated at 37°C in the presence or absence of purified MT1dTM (40 nM) up to 24 h. At designated time points, samples were collected and analyzed for Aβ levels by quantitative immunoblotting using the anti-Aβ mAb. As shown in Fig. 5, Aβ40 and Aβ42 were degraded by purified MT1dTM in-vitro in a time-dependent manner with ≈40% reduction in the levels of both peptides in 24 h.</p><p>To confirm that the enzymatic activity of purified MT1dTM was required for the Aβ degradation in-vitro we used the general MMP inhibitor GM6001 and specific MT1-MMP inhibitor TIMP2, Aβ40 was incubated with purified MT1dTM at 37°C in the presence or absence of GM6001 or TIMP2 for 24 h. The immunoblotting data presented in Fig. 6 shows that MT1dTM mediated Aβ40 degradation was blocked by GM6001 and TIMP2 indicating that the proteolytic activity of MT1dTM was responsible for the observed Aβ degradation.</p><p>The data above demonstrates that MT1-MMP exhibits proteolytic activity towards Aβ40 or Aβ42 in-vitro or in Cos-1 cells expressing MT1-MMP. However, these analyses only show loss of intact Aβ peptides based on immunoblotting or ELISA analysis. To identify MT1-MMP mediated Aβ cleavage products we performed acid/urea gel analysis, a technique that can resolve low molecular mass peptides. For this analysis soluble amino-terminal, biotinylated Aβ40 or Aβ42 peptides were incubated with purified MT1dTM for 48 h. Following incubation, the samples were electrophoresed on a 22% polyacrylamide acid/urea gels, transferred to membranes, and analyzed for biotin-labeled intact Aβ and amino-terminal fragments using a streptavidin-horseradish peroxidase conjugate. As shown in Fig. 7, the levels of intact biotin-labeled Aβ40 and Aβ42 were markedly reduced by digestion with MT1dTM and several biotin-labeled amino-terminal fragments of each Aβ peptide were observed. These data further confirm that MT1dTM degrades soluble Aβ in-vitro.</p><p>To identify specific cleavage products, synthetic Aβ40 or Aβ42 was digested with purified MT1dTM and analyzed by MALDI-TOF mass spectrometry (Fig. 8). The major fragments generated from proteolytic cleavage of Aβ40 were similar to those generated from Aβ42. Several cleavage sites were identified mainly around V12 through L17, generating major fragments of 1–14 to 1–17, which were consistent with the amino terminal major cleavage products shown in the acid/urea gels (Fig. 7).</p><!><p>Aβ peptides largely accumulate in the AD brain in the form of fibrillar amyloid deposits. To determine whether MT1-MMP could degrade fibrillar Aβ, we prepared aged fibrillar Aβ42, subsequently incubated it with purified MT1dTM at 37° for 5 days and measured the remaining fibrillar Aβ using a Th-T fluorescence binding assay. Fig. 9A shows a >50% reduction in the Th-T fluorescence signal in the fibrillar Aβ sample treated with MT1dTM. Importantly, the MMP inhibitor GM6001 largely blocked MT1dTM mediated fibrillar Aβ degradation indicating the loss of fibrillar Aβ was dependent on the enzymatic activity of MT1dTM. To further confirm this finding at the ultrastructural level fibrillar Aβ was incubated in the absence or presence of purified MT1dTM for 5 days and then TEM analysis was performed to visualize the extent fibrillar Aβ structure (Fig. 9B). Fibrillar Aβ incubated with MT1dTM showed a marked reduction in the number and length of amyloid fibrils. Together, these data indicate that MT1dTM is also capable of degrading the assembled fibrillar form of Aβ.</p><!><p>The above data showed that purified MT1dTM was capable of degrading soluble and fibrillar synthetic Aβ peptides in vitro. We next determined if purified MT1dTM could degrade actual amyloid deposits that form in the brains of APP transgenic mice. To do this, adjacent brain slices of aged Tg2576 mice, which contain abundant amyloid plaques, were incubated at 37°C for 5 days with buffer alone or purified MT1dTM in the presence or absence of the MMP inhibitor GM6001. After incubation the sections were stained with Th-S and the area of fluorescence between matching fibrillar plaque deposits from adjacent sections was measured. The area of Th-S fluorescence of adjacent brain sections did not show a difference when incubated with buffer alone, while the area of parenchymal amyloid plaque deposits was significantly decreased (p < 0.001) in the brain sections incubated with purified MT1dTM (Fig. 10). Importantly, amyloid plaque degradation by purified MT1dTM was effectively blocked with the MMP inhibitor GM6001. These results suggest that purified MT1dTM is capable of degrading fibrillar amyloid plaques in brain tissue.</p><!><p>In the present study we show that MT1-MMP, the physiological activator of pro-MMP2, can degrade soluble and fibrillar forms of Aβ in vitro. The MMP superfamily consists of secreted and membrane types of metalloproteinases largely involved in degradation and remodeling of the extracellular matrix. It was previously shown that MMP2 and MMP9 are produced by reactive astrocytes surrounding amyloid plaques in aged human AβPP transgenic mice (17–20, 39). Similarly, we found that MMP2 and its activator MT1-MMP are expressed in reactive astrocytes in brain regions with microvascular amyloid deposits in aged Tg-SwDI mice (Figs. 1 and 2). Previously, we reported that pathogenic Aβ stimulates the expression and activation of MT1-MMP and MMP2 in the cultured human cerebrovascular smooth muscle cells (16, 40). Consistent with these earlier in vitro findings, in aged Tg-SwDI mice we also found MMP2 and MT1-MMP expression in the smooth muscle cell medial layer of meningeal vessels that occasionally contained fibrillar Aβ deposits (data not shown).</p><p>MT1-MMP was shown to degrade Aβ peptides in both its natural transmembrane form expressed in Cos-1 cells and as a purified soluble form lacking the carboxyl-terminal transmembrane region (MT1dTM). However, more robust degradation of Aβ peptides was observed when MT1-MMP was expressed in Cos-1 cells compared to using the soluble MT1dTM form in vitro. This disparity may reflect different levels of MT1-MMP present in each type of experiment or, more likely, is a consequence of the soluble MT1dTM exhibiting less enzymatic activity than its natural transmembrane counterpart (41). In any case, soluble MT1dTM provided a useful tool to demonstrate degradation of soluble and fibrillar Aβ in in vitro and in situ experiments.</p><p>Most of the well-known Aβ-degrading enzymes such as endothlein converting enzyme, insulin-degrading enzyme, and neprilysin largely show degradative activity towards soluble forms of Aβ, but not fibrillar Aβ. However, plasmin and MMP9 are two Aβ-degrading enzymes shown to be capable of degrading fibrillar Aβ in-vitro (37, 42). In the initial experiments of the present study we mainly focused on the degradation of the monomer form of soluble Aβ in the in-vitro assay or in Cos-1 cells. However, in Fig. 4 above the prominent Aβ monomer a faint Aβ dimer band was observed which was also degraded in Cos-1 cells expressing MT1-MMP. Although we did not investigate the specific degradation of other soluble forms of Aβ such as trimers, tetramers, or higher order oligomers our experiments showed that fibrillar Aβ was degraded by soluble MT1dTM. Based on this latter finding we predict that other soluble oligomeric forms of Aβ are likely degraded by MT1-MMP although this will need to be confirmed.</p><p>Several structural models of amyloid fibrils have been proposed (43). The common feature is the β-pleated sheet structure perpendicular to the fibril axis with a hairpin loop at the C-terminus. The conversion of soluble Aβ to fibrillar amyloid is accompanied by an increased resistance to proteolytic degradation (44). In this regard it is noteworthy that purified soluble MT1dTM can similarly degrade fibrillar Aβ in-vitro and fibrillar amyloid deposits in brain tissue sections of human AβPP transgenic mice. MMP9, like MT1-MMP, cleaves between residues A30-I31 (37). This site is exposed on the surface of Aβ fibrils allowing access for cleavage by MMP9 and MT1-MMP (37). In contrast, Aβ fibrils were observed to be more resistant to degradation by MMP2. It was proposed that the major MMP2 cleavage site of Leu34-Met35 within the hydrophobic domain of Aβ would be inaccessible within an amyloid structure (44). Collectively, these findings suggest that the various Aβ-degrading enzymes likely work at different sites in the brain for Aβ catabolism. For example, secreted Aβ-degrading enzymes such as IDE, MMP2, and MMP9 may effectively target soluble forms of Aβ in interstitial fluid whereas membrane-bound Aβ-degrading enzymes such as neprilysin and MT1-MMP are better suited for deposited fibrillar Aβ or Aβ associated with cell surfaces.</p><p>MT1-MMP appears to be highly expressed in brain regions exhibiting amyloid pathology and neuroinflammation (Figs. 1 and 2). On the other hand, in normal brain or in the absence of amyloid pathology little, if any, expression of MT1-MMP is observed. This suggests that under normal conditions MT1-MMP likely has little involvement in regulating basal brain Aβ levels compared with other Aβ-degrading enzymes that are constitutively expressed. However, when amyloid deposition and neuroinflammation occur, as in AD, reactive astrocytes and vascular smooth muscle cells markedly increase their expression of MT1-MMP which may then play a significant role degrading soluble and deposited Aβ peptides. This increased expression in response to amyloid deposition implies that MT1-MMP may be an opportunistic Aβ-degrading enzyme. Future experimentation will be needed to determine if MT1-MMP does indeed contribute to Aβ degradation in vivo under pathological conditions when it is likely expressed.</p><p>Various members of the MMP superfamily may play some role regulating the levels of Aβ in the CNS. For example, MMP2, MMP9, and MT1-MMP possess Aβ-degrading activity (15, 45). MMP2, MMP3, MMP9 and MT1-MMP exhibit increased expression in response to Aβ (16, 23, 46). However, the protein levels and activity of MMP2, MMP3, and MMP9 showed no difference in the frontal cortex of AD patients compared with control patients (47). This may reflect a very limited, focal expression in specific cells that was not discerned in this study. It was reported that MT1-MMP, MT3-MMP and MT5-MMP have α secretase-like shedding activity on AβPP which would preclude Aβ formation (31). More specifically, recombinant MT3-MMP showed multiple cleavage sites on AβPP within the Aβ domain. Since the shedding pattern for MT1-MMP and MT3-MMP are very similar, MT1-MMP may also cleave AβPP within the same sites. Here, our mass spectrometry data showed an MT1-MMP cleavage site at the H14-Q15, which is the same as an MT3-MMP shedding site on AβPP (30). However, Aβ peptide was not degraded by recombinant MT3-MMP or by cells expressing MT3-MMP. Therefore, regarding MT-MMPs the Aβ degradation activity appears specific to MT1-MMP.</p><p>In conclusion, we have demonstrated that MT1-MMP is selectively expressed in reactive astrocytes near fibrillar amyloid deposits in human AβPP transgenic mouse brain. MT1-MMP was found to degrade soluble Aβ40 and Aβ42 as well as fibrillar amyloid. Together, our data suggest MT1-MMP could function as an opportunistic Aβ degrading enzyme when expressed by reactive astrocytes adjacent to fibrillar amyloid deposits. Future in vivo studies are needed to determine its role in relation to other identified Aβ degrading enzymes in regulating Aβ levels in brain.</p><!><p>Brain sections from twenty four month old Tg-SwDI mice were labeled for fibrillar Aβ using Th-S (green) showing that the cortex (A) lacks appreciable fibrillar amyloid whereas the thalamic region (B) contains extensive microvascular amyloid accumulations. Immunolabeling for MMP2 or MT1-MMP (red) in adjacent brains sections shows weak expression in the cortex (C and E, respectively) but strong expression in the thalamic region (D and F, respectively) containing abundant microvascular amyloid. Scale bars = 50 μm.</p><!><p>Brain sections from twenty four month old Tg-SwDI mice were double immunolabeled for GFAP to identify astrocytes (green) and MMP2 or MT1-MMP (red). The thalamic region, which contains extensive microvascular fibrillar amyloid, is shown. Numerous reactive astrocytes were observed (A and E) as well as strong immunolabeling for MMP2 (B) and MT1-MMP (F). Merging of the images showed strong co-localization of GFAP and MMP2 (C) or MT1-MMP (G). Scale bars = 50 μm. Higher magnifications of the merged images are shown in (D) and (H), respectively. Scale bars = 10 μm.</p><!><p>Triplicate cultures of Cos-1 cells were transfected with empty plasmid vector (pcDNA3.1), Pro-MMP2 vector alone, MT1-MMP vector alone, or both Pro-MMP2 vector and MT1-MMP vector. Twenty four hours post transfection, the cells were incubated in serum-free medium for an additional 48 h. (A) The cell lysates were collected and analyzed by immunoblotting by using anti-MT1-MMP. (B) The culture media samples were collected and analyzed by gelatin zymography. Co-expression of Pro-MMP2 and MT1-MMP led to conversion of pro-MMP2 to MMP2 demonstrating the MT1-MMP was proteolytically active.</p><!><p>Triplicate cultures of Cos-1 cells were transfected with empty plasmid vector (pcDNA3.1) or MT1-MMP vector. Twenty four hours post transfection, the cells were incubated with 2 μg/ml of freshly solublized Aβ40 or Aβ42 in serum-free media for an additional 48 h. (A) The cell lysates were collected and analyzed by immunoblotting using anti-MT1-MMP. The culture media samples were collected and analyzed for Aβ40 and Aβ42 peptides levels by immunoblotting using anti-Aβ (B and D, respectively) and by ELISA (C and E, respectively). The data shown are the mean ± S.D. (n=3). *, p < 0.05; **, p < 0.01.</p><!><p>Aβ40 (A and B) or Aβ42 (C and D) was incubated at 37°C in the presence or absence of purified 40 nM of MT1dTM. At each time point, samples were collected and analyzed for Aβ level by quantitative immunoblotting using anti-Aβ mAb. The data shown are the mean ± SD of three separate determinations.</p><!><p>Aβ40 was incubated with purified soluble MTdTM at 37°C in the presence or absence of the general MMP inhibitior GM6001 (10 μM) or the specific MT1-MMP inhibitor TIMP2 (40 nM) for 24 h. Following incubation, the samples were collected and analyzed for Aβ levels by quantitative immunoblotting using anti-Aβ (A). The data shown are the mean ± SD of three separate determinations (B). *, p < 0.05, paired t test.</p><!><p>Soluble amino terminal, biotinylated Aβ40 or Aβ42 were incubated with purified soluble MT1dTM for 48 h. Following incubation, samples were separated on 22% polyacrylamide acid/urea gels, transferred to membranes, and analyzed for Aβ products by using a streptavidin-horseradish peroxidase conjugate to detect biotinylated peptides and fragments. Lane 1, biotinylated-Aβ40; lane 2, biotinylated-Aβ40 + MT1dTM; lane 3, biotinylated-Aβ42; and lane 4, biotinylated-Aβ42 + MT1dTM. The brackets denote amino terminal cleavage products common to Aβ40 and Aβ42.</p><!><p>Synthetic Aβ40 alone (A), Aβ42 alone (C), Aβ40 and purified MT1dTM (B) or Aβ42 and purified MTdTM (D) were incubated with 37°C for 2 days. After incubation the samples were analyzed by MALDI-TOF mass spectrometry. Comparing with Aβ40 or Aβ42 alone, several specific peaks were identified as Aβ fragments (reflector mode). (E) Summary of the MT1dTM cleavage sites on Aβ (▼).</p><!><p>(A) Fibrillar Aβ42 was incubated alone or with MT1dTM in the presence or absence of the MMP inhibitor GM6001 for 37°C for 5 days. The remaining fibrillar Aβ was quantitated using a Th-T binding fluorescence assay. The data shown are the mean ± SD of three separate determinations. (B) Fibrillar Aβ42 was incubated alone or with purified MT1dTM at 37°C for 5 days. The samples were collected an analyzed by TEM. Scale bars = 100 nm.</p><!><p>Adjacent 5 μm fresh frozen brain sections from 18 months old Tg2576 mice were incubated alone (A,B,C,E) or with purified MTIdTM (D) or GM6001-treated MT1dTM (F) at 37°C for 5 days. The sections were then fixed, and stained with Th-S. Insets show parallel representative plaques enlarged. Scale bars = 50 μm. (G) The parallel cortical fibrillar amyloid plaque areas were quantified in the treated and untreated sections and expressed as percent remaining Th-S area. The data presented are the mean ± S.D. of n = 25 plaques (buffer alone); n = 37 plaques (incubated with MT1dTM); n = 27 plaques (incubated with GM6001-treated MT1dTM). *, p < 0.001, paired t test.</p><!><p>Abbreviations used are: AD, Alzheimer's disease; Aβ, amyloid β-protein; MMP, matrix metalloproteinase; MT1-MMP, membrane type 1 matrix metalloproteinase; Th-T, thioflavin T; Th-S, thioflavin S; DMSO, dimethyl sulfoxide; SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis; PBS, phosphate-buffered saline; TIMP2, tissue inhibitor of metalloproteinase-2, MT1dTM: soluble transmembrane domain-lacking form of MT1-MMP; GFAP, glial fibrillary acidic protein.</p><p>This work was supported by National Institutes of Health grants HL72553 and NS55118. We thank Dr. Mahiuddin Ahmed and Dr. Steven Smith for performing the TEM analysis of fibrillar Aβ. ELISA antibody reagents were generously provided by Eli Lilly Laboratories.</p>

PubMed Author Manuscript

Retinal Degeneration 3 (RD3) Protein Inhibits Catalytic Activity of Retinal Membrane Guanylyl Cyclase (RetGC) and Its Stimulation by Activating Proteins

Retinal membrane guanylyl cyclase (RetGC)1 in the outer segments of vertebrate photoreceptors is controlled by guanylyl cyclase activating proteins (GCAPs), responding to light-dependent changes of the intracellular Ca2+ concentrations. We present evidence that a different RetGC binding protein, retinal degeneration 3 protein (RD3), is a high-affinity allosteric modulator of the cyclase which inhibits RetGC activity at submicromolar concentrations. It suppresses the basal activity of RetGC in the absence of GCAPs in a non-competitive manner and it inhibits the GCAP-stimulated RetGC at low intracellular Ca2+ levels. RD3 opposes the allosteric activation of the cyclase by GCAP, but does not significantly change Ca2+ sensitivity of the GCAP-dependent regulation. We have tested a number of mutations in RD3 implicated in human retinal degenerative disorders and have found that several mutations prevent the stable expression of RD3 in HEK293 cells and decrease the affinity of RD3 for RetGC1. The RD3 mutant lacking the carboxy-terminal half of the protein and associated with Leber congenital amaurosis type 12 (LCA12) is unable to suppress the activity of the RetGC1/GCAP complex. Furthermore, the inhibitory activity of the G57V mutant implicated in cone-rod degeneration is strongly reduced. Our results suggest that inhibition of RetGC by RD3 may be utilized by photoreceptors to block RetGC activity during its maturation and/or incorporation into the photoreceptor outer segment rather than participate in dynamic regulation of the cyclase by Ca2+ and GCAPs.

retinal_degeneration_3_(rd3)_protein_inhibits_catalytic_activity_of_retinal_membrane_guanylyl_cyclas

INTRODUCTION<!>Antibodies<!>Human recombinant RD3 expression in E. coli<!>Expression of a full-length human RD3 in HEK293 cells used in RetGC activity assays<!>GCAP1 and GCAP2 expression<!>RetGC activity<!>Ca2+ buffers<!>Co-immunoprecipitation (co-IP)<!>Inhibition of the basal and the GCAP-stimulated RetGC activity by RD3<!>RD3 affects the activation of RetGC rather than Ca2+ sensitivity of the GCAP-dependent regulation<!>Mutations in human RD3 affect its expression in HEK293 cells and RetGC activity in vitro<!>DISCUSSION

<p>Cyclic GMP in vertebrate photoreceptors couples rhodopsin in disk membranes to cGMP-gated cation channels in the plasma membrane of rod outer segments (ROS) through a photoransduction cascade (reviewed in (1)). Following photoexcitation, cGMP in photoreceptors is first hydrolyzed by a light-activated phosphodiesterase (PDE6) and then replenished by a retinal membrane guanylyl cyclase (RetGC) (2–4). In addition to its role as a key enzyme in photoreceptor physiology, RetGC has been linked to multiple cases of congenital blinding disorders in human patients including Leber congenital amaurosis type 1 (LCA1) and dominant cone-rod dystrophy (5–7). RetGC activity in rods and cones is controlled by Ca2+-sensitive guanylyl cyclase activating proteins, GCAPs (8–11), via a Ca2+ feedback mechanism. When cGMP-gated channels close in response to light, the reduced influx of Ca2+ through the channels causes GCAP to switch from a Ca2+-bound (inhibitor) state to a Mg2+-bound (RetGC activator) state (reviewed in (12)). This accelerates the re-synthesis of cGMP for timely recovery of photoreceptors to their dark state or establishing light adaptation (13–14).</p><p>More recently, it has been shown that RD3, a 23-kDa protein unrelated to GCAPs and linked to Leber congenital amaurosis type12 (LCA12) and rapid retinal degeneration in the rd3 mouse (15), co-immunoprecipitates with RetGC and is essential for the normal expression of RetGC in rod and cone photoreceptor cells (16). The RD3 gene transcripts are highly abundant in the retina and concentrated in photoreceptors but undetectable in other mouse tissues (15). An anti-RD3 polyclonal antibody was observed to stain the inner and outer segments of mouse rod and cone photoreceptor cells by immunofluorescence microscopy (16). RD3 also co-localized with RetGC in intracellular vesicles when both proteins were co-expressed in HEK293 cells (16), thus leading to a hypothesis that at least one function of RD3 in photoreceptor cells is to participate in the intracellular trafficking of the cyclase.</p><p>In the present study, we have investigated whether the association of RetGC with RD3 can affect the catalytic function of RetGC. We find that RD3 acts as a high-affinity allosteric inhibitor of RetGC, capable of both effectively competing with GCAPs and suppressing the catalytic activity of the cyclase. Some mutations in the human RD3 gene found in patients with congenital retinal disorders strongly affect the inhibitory activity of the human recombinant RD3 protein in vitro.</p><!><p>Rabbit polyclonal antibody RD3 #497 was raised against the isolated full-length human recombinant RD3 expressed in E. coli as described below and purified from the serum by immunoaffinity chromatography on RD3 coupled to CNBr-activated Sepharose CL-4B (GE Heath Sciences). The Rho-1D4 mouse monoclonal antibody (17) was against the TETSQVAPA peptide used as a C-terminal tag in some RD3 constructs, and the anti-RD3 mouse monoclonal RD3-9D12 antibody was produced against the C-terminal peptide of RD3 (16). The rabbit polyclonal anti-RetGC1 antibody was produced against the catalytic domain of human RetGC1 (18) and the mouse monoclonal GC-8A5 antibody was raised against the C-terminus of mouse RetGC1.</p><!><p>Human RD3 cDNA was amplified from a pCMV-SPORT6/MHS1010-9206149 cDNA clone (Open Biosystems/Thermo Scientific) using high-fidelity Phusion Flash DNA polymerase (Finnzymes), subcloned into the NcoI/BamHI sites of pET11d vector (Novagen/Calbiochem), sequenced, and expressed in a BL21(DE3)CodonPlus (Agilent Technologies) E. coli strain in the presence of isopropyl β-D-1-thiogalactopyranoside for 2 hours. The protein which accumulated in inclusion bodies was purified by series of sonication and centrifugation cycles described for GCAP purification (19), solubilized in 10 mM Tris-HCl (pH 7.5) buffer containing 2 mM EDTA, 8 M urea, and 14 mM 2-mercaptoethanol, and dialyzed against 2 × 300 volumes of 10 mM Tris-HCl (pH 7.5) buffer containing 0.1 mM EDTA and 14 mM 2-mercaptoethanol at 4°C. Insoluble protein was removed by centrifugation at 10,000 × g for 10 min at 4°C, and the supernatant containing RD3 (typically 80 – 90% purity by SDS PAGE) was collected and used either immediately or after storage at −70°C with 50% v/v glycerol. RD3 has a tendency to precipitate under normal storage conditions. For expression of RD3 mutants (15) in E. coli, the corresponding mutations were introduced into the RD3 cDNA using Phusion Flash DNA polymerase (Finnzymes) by a conventional "splicing by overlap extension" method and the mutated cDNA was verified by sequencing the entire coding region of the resulting plasmid.</p><!><p>The RD3 cDNA was inserted into the HindIII/XbaI sites of a modified pRCCMV vector (Invitrogen), transfected into a 50 to 80%-confluent cell culture (ca. 20 μg of DNA per 100-mm dish) using the calcium phosphate DNA precipitation protocol and expressed for 24 – 36 hours. The soluble fraction containing RD3 was extracted from the harvested cells which were subjected to a hypotonic shock on ice in a buffer solution containing 10 mM Tris-HCl (pH 7.5), 0.1 mM EGTA, 14 mM 2-mercaptoethanol, 60 μg/ml aprotinin, and 30 μg/ml leupeptin and further disrupted by 10–15 strokes in a Dounce glass-to-glass homogenizer. The soluble fraction was clarified by centrifugation at 12,000 × g for 30 min, 4°C, and concentrated in a Millpore Amicon Ultra-4 MW10000 cartridge. A 10-μl sample was subjected to SDS PAGE on a 15% gel. Expression of RD3 was confirmed by immunoblotting probed with the rabbit polyclonal anti-RD3 antibody #497 and visualized using a SuperSignal chemiluminescence peroxidase substrate kit (Pierce/ThermoFisher) in a FotoDyne Luminous FX imaging system.</p><p>For comparative testing of the RD3 mutant expression in cultured HEK293T cells, wild type and the mutant RD3 cDNA constructs generated using the QuikChange Mutagenesis kit (Agilent Technologies) all contained an extension that coded for the C-terminal Rho-1D4 peptide, TETSQVAPA. HEK293T cells were transfected with 10 μg of the RD3 plasmid DNA per a 100-mm culture dish using the calcium phosphate precipitation protocol. Forty-eight hours after transfection, the cells from 2 plates were solubilized in 0.8 ml of 1% Triton X-100 in PBS containing Complete Protease Inhibitor cocktail (Roche) for 15 min at 4°C and the insoluble material was removed by centrifugation. Each sample was resolved on a 10% SDS polyacrylamide gel and analyzed on immunoblots sequentially probed with the Rho-1D4 mouse monoclonal antibody and secondary goat anti-mouse antibody conjugated to IR dye 680 for imaging on a LI-COR Odyssey infrared imaging system.</p><!><p>GCAP1 and GCAP2 were expressed from pET11d vector in a BLR(DE3) E. coli strain harboring yeast N-myristoyl transferase as previously described (19, 20).</p><!><p>RetGC cyclase activity in ROS fraction isolated from mouse retinas using Optiprep density gradient centrifugation as described in (21) was assayed in the dark using infrared viewers as described previously (14, 22–23). The activity of the recombinant human RetGC1 expressed from pRCCMV vector in HEK293 cells transfected by calcium-phosphate precipitation method was assayed under normal illumination as described previously (22). The assay mixture (25 μL) contained (unless indicated otherwise) 30 mM MOPS – KOH (pH 7.2), 60 mM KCl, 4 mM NaCl, 1mM DTT, 2 mM Ca2+/EGTA buffer, 1 mM free Mg2+, 0.3 mM ATP, 4 mM cGMP, 1 mM GTP, 10 mM creatine phosphate, 0.5 unit of creatine phosphokinase, 1 μCi of [α-32P]GTP, 0.1 μCi of [8-3H]cGMP (Perkin Elmer), PDE6 inhibitors zaprinast and dipyridamole, and variable concentrations of GCAPs, RD3, and membranes containing RetGC1. The reaction continued at a linear time-course for 12 min at 30°C (40 min for the recombinant RetGC1) and subsequently heat-inactivated at 95°C for 2 min. The [32P]cGMP was quantified using a thin-layer chromatography on polyethylenimine cellulose plates. Statistical analysis of the data, where applicable, was performed using KaleidaGraph (Synergy Software).</p><!><p>Free Ca2+ and Mg2+ concentrations in the RetGC assay were maintained using a series of Ca2+/EGTA mixtures prepared using Tsien and Pozzan method (24), and their free Ca2+ and Mg2+ concentrations were calculated using the Marks and Maxfield algorithm in Bound and Determined and MaxChelator software which corrects for the effects of other components of the assay mixture. The calculated Ca2+/EGTA buffering was verified by fluorescent indicator dyes as previously described in detail (20, 25).</p><!><p>For the co-IP studies, the HEK293T cell cultures were transfected with 10 μg of the RD3 plasmid and 10 μg of human RetGC1 plasmid DNA per a 100-mm diameter culture dish. Twenty-four hours after the transfection, the cells were solubilized in 800 μl 1% Trition X-100 in PBS containing Complete Protease Inhibitor cocktail (Roche). After removal of the insoluble material by centrifugation, the supernatant was applied to an Rd3-9D12-Sepharose immunoaffinity column. After 1 hour the column was washed several times with 0.1% Triton X-100 in PBS and the fraction containing bound proteins was eluted with 80 μl of 3% SDS.</p><!><p>Recombinant RD3 strongly inhibits the activity of the native RetGC1/GCAP1 complex in ROS membranes and the recombinant RetGC1 expressed in HEK293 membranes and reconstituted with purified GCAP1 (Fig. 1A–C). The suppression of the RetGC activity was observed by RD3 expressed and purified from E. coli (Fig. 1A, B) and RD3 present in a soluble extract from the HEK293 cells expressing recombinant protein (Fig. 1C). The experiment presented in Fig. 1D also demonstrates that RD3 suppresses the catalytic activity of RetGC in the absence of GCAPs. Two isoforms of GCAPs, GCAP1 and GCAP2, are the products of neighboring genes, which are both eliminated by a single knockout construct in GCAP1,2−/− mice (26). GCAP1 and GCAP 2 are the only two GCAP isoforms present in the mouse genome. Therefore, RetGC in the GCAP1,2−/− remains insensitive to Ca2+ (13, 26) and its basal cyclase activity is much lower than the GCAP-stimulated RetGC (21), even though the level of RetGC expression in the double knockout retinas remains normal (21, 26). Since RD3 inhibits RetGC activity in ROS membranes from the GCAP1,2−/− retinas (Fig. 1D), it must directly inhibit the catalytic activity of the cyclase per se.</p><p>Classical inhibitory analysis of Michaelis kinetics is presented in Fig. 1E–F for basal RetGC activity. There is no significant difference between the KmGTP values (mean ± SD from 3 independent measurements) in the absence (1.31 ± 0.17 mM) or in the presence of 30 nM (1.57 ± 0.25 mM) and 60 nM (0.94 ± 0.41 mM) RD3, whereas the Vmax was reduced nearly 2.5-fold and 5-fold – from 3.5 ± 0.7 to 1.4 ± 0.4 and 0.7 ± 0.4 nmol min−1 mg Rh−1, respectively (P<0.05). This argues that the inhibition of the basal catalytic activity of RetGC by RD3 in the absence of GCAPs is mostly non-competitive (Ki ~ 20 nM). Based on this, it is unlikely that RD3 directly binds to the catalytic pocket of the enzyme and affects substrate binding. Since the RetGC is active only as a homodimer (7, 27–29), one possible mechanism for the RD3 affecting RetGC catalytic activity is by altering the conformation of the cyclase dimer. This possible mechanism requires further study.</p><p>In addition to the basal activity, RD3 also inhibits the activity of RetGC reconstituted with GCAP as well as the RetGC activity in ROS fraction isolated from wild type mice (Fig. 1A). This suggests that the GCAP-stimulated activity of RetGC can also be affected by RD3, perhaps through its competition with GCAPs. Indeed, the data presented in Fig. 2 directly support the possibility that RD3 strongly competes with GCAP1 for the recombinant human RetGC1 when both regulator proteins are present in the assay. When RetGC1 is co-expressed with RD3, the concentration-dependence of its activation by GCAP measured in HEK293 cell homogenates shifts to a much higher range compared to the RetGC1 expressed in the absence of RD3 (Fig. 2A). Likewise, the competition between RD3 and GCAPs for the cyclase is also evident from the experiments when RetGC1 expressed in HEK293 cells is activated by purified GCAP1 (Fig. 2C) or GCAP2 (Fig. 2D) in the presence of different concentrations of purified recombinant RD3. While the maximal level of the cyclase activity stimulated by GCAPs at saturation is only slightly affected by RD3, the concentration-dependence of the activation by each GCAP shifts toward much higher EC50 values. The latter can only be interpreted as a competitive blocking of the GCAP-dependent stimulation of RetGC by RD3. We emphasize that in these experiments RD3 competes with the GCAPs regardless of what expression system is used to produce the RD3 protein, i.e. HEK 293 cells (Fig. 2A) or E. coli (Fig. 2C–D). Therefore, the effect we observe cannot be attributed to a non-specific artifact(s) of a particular RD3 expression system.</p><!><p>The effect of RD3 on Ca2+ and Mg2+- dependence of RetGC regulation by GCAPs was tested using two different preparations – the native ROS fraction obtained from dark-adapted mouse retinas by density gradient centrifugation (Fig. 3A–C) and a membrane fraction from the HEK293 cells expressing RetGC1 and reconstituted with the recombinant GCAP1 (Fig. 3D–F). When RetGC activity partially suppressed by RD3 is measured as a function of Ca2+ concentration, its sensitivity to Ca2+ remains similar to that in the absence of RD3. The sensitivity of cyclase activity to Mg2+ is directly relevant to its sensitivity to Ca2+. This is because Mg2+ is required for maintaining the activator conformation of GCAP1 in the light (12, 20, 26, 30) and competes with Ca2+ binding in the EF hand domains of GCAP (20) thus shifting the Ca2+ sensitivity of the cyclase regulation by GCAPs (23). This shift appears in the presence of RD3 when the free Mg2+ level in the assay is increased from a near-physiological 1 mM to 6 mM and is similar to the control samples assayed in the absence of RD3 (Fig. 3C, F).</p><!><p>A nonsense mutation in rd3 mice that terminates RD3 synthesis after Glu106 and a mutation in LCA12 patients that alters a splicing signal and also generates a stop codon after Arg99 both delete the C-terminal half of the protein (15). In addition to the deletion mutation directly linked to LCA12, a number of missense mutations were detected in patients with other retinal disorders (15), although their roles in development of the disease remain to be established. We have tested the effect of several such mutations on human RD3 expression in HEK293 cells and their ability to to associate with RetGC in co-IP expertiments (Fig. 4). The F100ter mutation mimicking the LCA-linked truncation of RD3 as well as the G57V mutant failed to accumulate in transfected HEK293T cells at detectable levels (Fig. 4A, B). This is in agreement with the very low expression of the related mouse RD3 truncated protein observed in transfected COS-1 culture cells (15). Other tested RD3 mutants were expressed in HEK293 cells and co-immunoprecipitated with RetGC1, indicating that they retain their general ability to interact with the cyclase (Fig. 4C). We then evaluated the inhibitory activity of different mutants expressed and purified from E. coli (Fig. 5). Although the expression of the F100ter and G57V mutants was undetectable in HEK293 cells, their expression in E. coli is fairly robust. All RD3 mutants expressed in bacteria were purified in quantities sufficient for the in vitro RetGC assay (Fig. 5A). When reconstituted with the recombinant RetGC1 activated by GCAP1, the purified F100ter RD3 did not compete with the GCAP and was unable to suppress the RetGC activity (Fig. 5B). All other mutants are able to inhibit the activity of the GCAP1/RetGC1 complex, but to varying degrees. The W6R/E23D and the G35R mutants both inhibited RetGC1 activity in a manner nearly indistinguishable from the wild type protein, whereas the R68W, K130M, and G57V mutants were significantly less effective in suppressing cyclase activation by GCAP. This is particularly true for the G57V mutant, which has an IC50 value that is more than 10-fold higher than wild type RD3 (Fig. 5B).</p><!><p>It was recently found that a retina-specific protein RD3 linked to LCA12 blindness in humans and rapid retinal degeneration in rd3 mouse line (15) was expressed in rods and cones and was capable of associating with the photoreceptor-specific guanylyl cyclase RetGC (16). Our present data argue that this association has a strong functional consequence and that RD3 is a novel potent inhibitor of RetGC, capable of suppressing both the basal and GCAP-stimulated activity of the cyclase. A diagram summarizing the RD3 effects on the cyclase activity is presented in Fig. 6. In the absence of GCAPs RD3 suppresses the basal catalytic activity of the RetGC acting as a noncompetitive inhibitor (see also Fig. 1E–F). It also counteracts the GCAP-dependent activation of the cyclase acting as a negative allosteric modulator (see also Fig. 2). It should be emphasized that stimulation by GCAP increases the Vmax for RetGC in mouse ROS up to 20-fold (21) and up to 100-fold in HEK293 membranes (reference 31 and Fig. 2). Therefore, compared to the non-competitive inhibition of the basal RetGC activity in GCAPs−/− ROS (Fig. 1E–F), a direct competition with the stimulating effect of GCAPs clearly dominates the overall effect of RD3 on the cyclase regulation in a RetGC/GCAP complex (Fig. 2). Indeed, once RD3 partially or completely displaces GCAP from the complex with the cyclase, the activity of the recombinant RetGC would fall as much as 100-fold since the cyclase is no longer activated by GCAP. The additional suppression of the basal cyclase activity by RD3 would also contribute to the overall inhibition, but relatively less than the displacement of GCAP from the cyclase. As a result, the non-competitive inhibition of the catalytic activity evident in Fig. 1D–F becomes masked in experiments presented in Fig. 2 by the more robust allosteric modulation, i.e. RD3 competition with GCAPs. These results would be consistent with a model in which RD3 does not strongly inhibit the catalytic activity of the RetGC in complex with GCAP until after GCAP is displaced from the complex (Fig. 6).</p><p>The exact binding sites for RD3 in RetGC1 have not yet been identified. Deletion of a 48-amino acid fragment from the C-terminus of RetGC hampered RD3 binding in co-immunoprecipitation experiments (16), but at this point it is unclear whether or not this is the only critically important region for RD3 binding and whether or not the binding sites for GCAP and RD3 in RetGC overlap. Multiple fragments of RetGC primary structure have been proposed by different groups to participate in binding of GCAP, including a portion of the cyclase catalytic domain (32–35). Yet, the fragments of the catalytic domain that may participate in GCAP binding (or at least be in a close proximity to the GCAP binding site, ref. 33–35) are thus far located upstream from the C-terminal fragment whose deletion affects co-immunoprecipitation with RD3. However, it is also important to note that although a competition for an overlapping epitope(s) in RetGC appears to be the simplest and the most likely possibility, RD3 and GCAP do not necessarily have to bind to exactly the same peptide epitopes – the effect of their mutual exclusion can even be potentially achieved through changing RetGC dimer conformation upon binding of the two allosteric protein regulators in different places.</p><p>At the same time, we see no evidence that RD3 is able to critically affect Ca2+-sensitivity of RetGC regulation by GCAPs either in the native ROS fraction or as a recombinant cyclase (Fig. 3). This is generally consistent with the model presented in Fig. 6, where RD3 mainly competes with GCAPs rather than alters the activity of the GCAP/RetGC complex. Two different physiological forms of a metal-bound GCAP1 have been characterized – a Mg2+-bound activator form to which GCAP converts in the light and a Ca2+-bound inhibitor form which suppresses the activity of RetGC in the dark when the free Ca2+ rises (reviewed in 12). Competition between Ca2+ and Mg2+ for EF-hands in GCAP1 and GCAP2 therefore strongly affects Ca2+ sensitivity of RetGC regulation by GCAPs resulting in a prominent right-shift of the Ca2+ sensitivity curve as Mg2+ concentrations increase (23, 31). This same shift remains in the presence of RD3 (Fig. 3). These results argue that RD3 does not significantly affect metal binding to GCAP and is unlikely to alter the dynamic regulation of the cyclase in response to light induced changes in Ca2+ concentrations in photoreceptor cells (12,13).</p><p>What then is the possible role for the inhibitory activity of RD3 on RetGC activity in a photoreceptor cell? Evidently, the biochemical mechanism of RetGC inhibition by RD3 itself does not require the environment of the photoreceptor cell (Fig. 1, 2), yet it should be specific for the photoreceptors since both RD3 and RetGC (16) are both highly expressed in rod and cone cells and largely absent in other cells of the body (15). RD3 displays a much stronger apparent affinity than GCAP for RetGC since it inhibits the cyclase in a nanomolar range, whereas GCAPs activate RetGC at micromolar concentrations (Fig. 2). Furthermore, unlike GCAP/RetGC complex, the RD3/RetGC complex is sufficiently stable in detergent to sustain the co-IP procedures (Fig. 4 and reference 16). The precise binding stoichiometry of RD3 and RetGC1 remains unknown at this point, but the concentration of RD3 could hamper the GCAP-dependent stimulation based on the following estimate. Since nearly half of RetGC monomers may be co-immunoprecipitated with RD3 from the detergent-solubilized retinal membranes as an irreversible complex (16) and the concentration of RetGC dimer in mammalian ROS lies in a micromolar range (35, 36) reaching ~3.6 μM in mouse ROS (or 7.2 μM RetGC monomer)(21), the total concentration of RD3 could approach ~ 3.6 μM (assuming for simplicity that one RD3 binds per RetGC monomer). The total concentration of GCAPs can reach to as high as ~ 17 μM in the mouse ROS cytosol, nearly 5 times higher than RD3 (calculated based on ref. 21 and 35). However, RD3 exerts its allosteric modulation of the cyclase at concentrations nearly 100 times lower than GCAPs (Fig. 2, 3, 5). Therefore, even by a conservative estimate, RD3 has the potential for suppressing a large fraction of the cyclase activity. Even though we do not know at this time the fine mechanism of RD3/RetGC complex formation and quantification of RD3 in photoreceptors needs further investigation, we would not anticipate that RD3-dependent inhibition of the cyclase activity or its high-affinity competition with GCAP contributes to the dynamic regulation of RetGC by Ca2+ and GCAPs. Furthermore, based on our data it would also very likely require additional mechanisms to reduce or prevent the RD3 inhibitory effects on RetGC at the cellular level to allow the dynamic regulation of RetGC in the outer segment.</p><p>A possible functional role of RD3 in photoreceptor physiology is to suppress the basal and GCAP activated RetGC activity in the inner segment during its transport to the outer segment. Another possibility is that RD3 itself can actively participate in the process of the transport (16) and/or assembly of the cyclase complex into the proper structure of the outer segment, during which its ability to strongly compete with GCAPs can prevent GCAP from prematurely interfering with this photoreceptor-specific process.</p><p>An apparent involvement of mutated RD3 in congenital retinal diseases corroborates its possible physiological role in photoreceptors. The 195-amino acid product of Rd3 (C1of36) gene originally identified by Friedman and co-workers (15) has been linked to a recessive retinal degeneration in the rd3 mice (37, 38) and LCA12 patients (15). Although RD3 effectively competes with GCAPs for the target enzyme, it fails to do so when truncated by the F100ter mutation linked to LCA12 (Fig. 5). A similar shortening of RD3 in rd3 mice (15) also leads to the loss of interaction between RD3 and RetGC and dramatically suppresses the levels of RetGC1 and RetGC2 expression in mouse photoreceptors in vivo (16). A variety of other mutations in the RD3 gene were reported in patients with retinal disorders (15). While the importance of those mutations for the development of congenital diseases remains to be determined, we find that some of these mutations can affect RD3 expression in human cultured cells and reduce the inhibitor activity of RetGC (Fig. 4 and 5).</p><p>One could expect that a possible consequence of mutations hampering RD3 binding to the cyclase would be to prevent RetGC from reaching the outer segment resulting in its accumulation in the inner segment. Evidently, this is not the case in rd3 mice (16). When RD3 is truncated similar to the F100ter mutant found in LCA12 patients and is unable to bind RetGC, the photoreceptors suppress production of RetGC rather than allow it to accumulate in the inner segment. It is likely that the photoreceptors possess a safety check capability that prevents the accumulation of RetGC in the absence of RD3. It is therefore tempting to speculate that failure of the mutant RD3 to properly bind the cyclase could trigger the mislocalization and rapid destruction of the cyclase as a protection against unsuitable activation of cGMP synthesis. Depletion of RetGC in ROS would inevitably suppress cGMP production resulting in a drop in intracellular Ca2+ to a level which could cause photoreceptor degeneration (16). The consequence of the altered activity of several RD3 mutants shown in this study requires further in-depth analysis of the mutants in vivo.</p>

PubMed Author Manuscript

Intracellular Delivery of Virus-Like Particles Using a Sheddable Linker

Intracellular targeting is an important aspect of the efficient delivery of drugs and nanotherapeutics. Cytosolic transport of nanomaterials is often an essential requirement for therapeutic delivery into cells but remains a challenge owing to the endosomal trap and eventual lysosomal degradation of cargo. To address this, we designed a functional carrier that escapes the endosome and delivers biological materials into the cell's cytoplasm. For this purpose, we synthesized a glutathionesensitive linker that connects the well-known mitochondria targeting lipophilic triphenylphosphonium cation (TPP) to the surface of a proteinaceous nanoparticle based on the engineered virus-like particle (VLP) Qβ. Once in the cytosol, the thiol sensitive linker severs the TPP from the nanoparticle, halting its trafficking to the mitochondria, and marooning it in the cytosol. We demonstrate the successful in vitro cytosolic delivery of a VLP loaded with Green Fluorescent Protein, where evenly distributed fluorescence is observed in A549 lung cancer cells after four hours. We further demonstrate successful cytosolic delivery by showing that encapsulating siRNA inside the VLP promotes luminescence silencing in luciferase expressing HeLa cells more efficiently than VLPs that lack our sheddable TPP linker.

intracellular_delivery_of_virus-like_particles_using_a_sheddable_linker

Introduction<!>Result and Discussion<!>Cellular Uptake, Cytotoxicity, and Cytosolic Delivery of Qβ-M-TPP<!>Cytoplasmic Delivery of siRNA to Efficiently Silence Luciferase in vitro.<!>Conclusion

<p>Effective medicine for diagnosing and treating diseases including cancer, genetic disorders, Alzheimer's, and Parkinson's requires access to the cell cytoplasm. [1][2][3][4] Cytosolic delivery of nanomaterials and biomacromolecules is fraught with biological barriers-the largest being endosomal escape. 5 There are few routes of entry for large or charged macromolecules aside from endocytosis, and once inside the cell endosome, the most likely outcome is eventual lysosomal degradation resulting in low dose of delivery, poor bioavailability, and limited therapeutic efficiency. This limitation has presented itself as a significant challenge, specifically in cytosolic delivery of large and charged therapeutic cargos and targeted delivery of drugs into the intracellular compartment. Therefore, a delivery system capable of escaping the endosome before lysosomal degradation is needed. Some available delivery systems, including hydrogels, 6 lipid nanoparticles, 7 cell-penetrating peptides, 8 and metal-organic frameworks, 9 have shown potential in promoting cellular uptake, protecting cargo from enzymatic hydrolysis, and enhancing site-specific delivery. However, these systems suffer from various limitations including, leakage of encapsulated cargo, toxicity, and inefficient delivery that results in low bioavailability. The small lipophilic cationic molecule, triphenylphosphonium (TPP), has been long exploited as a mitochondrial targeting moiety for small and medium-sized molecules. 10,11 Cargo tagged with a TPP moiety can escape the endosome and preferentially bind to the negatively charged mitochondrial matrix. [12][13][14] We wondered if we could promote endosomal escape yet avoid mitochondrial targeting by installing a linker between the nanoparticle and the TPP moiety that degrades once the nanoparticle enters the cytoplasm. TPP is an easily synthesized small molecule with low toxicity and can be chemically functionalized to a wide range of nanoparticle platforms. This provides our approach with a distinct advantage over peptides, which are more costly, more complicated to attach to surfaces, and susceptible to enzymatic degradation. To demonstrate our approach, we employed a proteinaceous nanoparticle called a virus-like particle (VLP). VLPs have emerged as a promising platform for various therapeutic applications ranging from imaging, gene delivery, and drug delivery. [15][16][17][18] They are noninfectious, biocompatible, biodegradable, monodisperse, and robust platforms that can carry small molecules, 19,20 polymers, 21 and/or intact proteins either by supramolecular entrapment within their interior or by chemical conjugation to their surface. 22 The bacteriophage Qβ is a 28 nm icosahedral engineered VLP with 180 identical coat proteins that self-assembles around random mRNA during expression in E. coli. [23][24][25][26] We, and others, have shown that the coat proteins of Qβ can be disassembled, the random mRNA discarded, and then reassembled around new cargo or genetic material. 24,25,27 Furthermore, Qβ possesses functionalizable primary amine groups on three surface-exposed lysine residues (K2, K13, and K16) and the N-terminus (Scheme 1) as well as 180 solvent-exposed disulfide groups that crosslink either six or five proteins to form hexametric (Scheme 1A) and pentameric subunits, respectively. These disulfides can be reduced (Scheme 1B), and the free sulfhydryl groups are effectively "rebridged" in quantitative yields using a maleimide crosslinker (Scheme 1C). 19,28 This work demonstrates a "sheddable" cytosolic delivery approach by attaching a TPP containing moiety through a disulfide-bridging maleimide that forms a stable two-carbon bridge between the sulfurs. The maleimide on the VLP readily undergoes retro-Michael additions in the presence of glutathione (Scheme 1D), which is present in millimolar concentrations (10 -3 M) in the cytosol but micromolar concentrations (10 -6 M) outside the cell. 29,30 In vitro, the TPP-functionalized virus undergoes endocytosis, escapes the endosome, and then sheds the TPP-functionalized maleimide linker through a thiol exchange with glutathione in the cytosol. Moreover, we have developed a single-pot approach that allows us to simultaneously reassemble the virus around genetic cargo with a functionalized dibromomaleimide serving to 'sew up' the capsid. Finally, as proof of principle, we demonstrate the cytosolic delivery of Qβ loaded with Green Fluorescent Protein (GFP) and, in a second example, the cytosolic delivery of siRNA that stops expression of luciferase.</p><p>Scheme 1. Structure of Qβ VLP (exposed nitrogens are labeled in blue, and cysteines are labeled yellow). Hexametric and pentameric structures are colored green and orange, respectively. A) Close-up of one of the hexametric disulfide-lined pores. B) The disulfides can be reduced quantitatively to produce free sulfhydryl groups. C) These can be crosslinked using a dibromomaleimide reagent. D) The maleimide crosslinkers come off in the presence of glutathione to produce 12 free thiols.</p><!><p>Synthesis and characterization of Qβ-M-TPP TPP has a delocalized positive charge over large hydrophobic phenyl rings, which is known to permeate lipid bilayers and cross into the mitochondrial matrix through non-carrier-mediated transport because of the significant mitochondrial membrane potential. 11,[31][32][33] Our approach involves chemically modifying the surface of the engineered VLP Qβ using a synthetic linker that separates the TPP from the viral surface once the capsid enters the cytosol. We prepared a dibromomaleimide-triphenylphosphonium (DB-TPP) linker (Figure 1A&B) via direct EDC coupling. The dibromomaleimide moiety reacts with thiol groups in a two-step reaction, 34 each step having a half-life of seconds. 28,35 Qβ contains 180 disulfide bonds, each of which can be a potential site for functionalization via this approach. The attachment of DB-TPP to Qβ forms a covalent twocarbon "bridge" between the free thiol groups in reduced Qβ (Figure 1B). 35,36 First, Qβ is reduced using tris(2-carboxyethyl) phosphine (TCEP), which is confirmed by electrophoretic mobility in nonreducing sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The starting Qβ pentameric and hexametric subunit bands are shown in Figure 1C, which are converted almost exclusively into monomer bands following reduction. Ellman's assay further confirms reduction against a cysteine standard curve (Figure S1), where approximately 95% of the disulfides on Qβ were found to be reduced into free thiols. The bioconjugation was then completed following the addition of the DB-TPP molecule, and within a few minutes, conjugation was confirmed by a significant visible increase in yellow-green fluorescence under UV light. Electrophoretic mobility using non-reducing SDS-PAGE (Figure 1C) shows the reformation of the hexameric and pentameric subunits and a slight upward migration of these bands compared to unreduced Qβ. These bands are also fluorescent under UV light, providing further evidence for the successful conjugation of the M-TPP linker to Qβ. Native agarose electrophoresis of the conjugate, which is visualized by UV and Coomassie staining, shows less migration toward the positive electrode than unfunctionalized Qβ, which is attributed to the additional positive charge from the TPP moiety. Intensity from the newly formed Qβ maleimide-TPP conjugate (Qβ-M-TPP) can be measured by fluorescence spectroscopy at 540 nm (excitation 400 nm) (Figure 1D). 28 This emergence of yellow fluorescence following the successful displacement of the bromides and formation of the dithiolated conjugate has been attributed to lower self-quenching and a decrease in the frequency of emission-decreasing collisional events with solvent molecules following conjugation. 28,37 We found no changes in the size of Qβ-M-TPP compared to Qβ as determined by size exclusion chromatography (SEC) and dynamic light scattering (DLS) analyses (Figures 1E and 1F) -Qβ and Qβ-M-TPP have hydrodynamic radii (Rh) of 31.92 ±10.76 and 31.22 ±10.36 nm respectively. Transmission electron microscopy (TEM) confirms that the morphology of conjugated Qβ-M-TPP is unchanged (Figure 1G). ζ-potential measurements show an increase in positive charge on Qβ after conjugation, which arises from the cationic nature of TPP (Figure S2). Lastly, the total number of conjugated linkers per capsid was determined to be approximately 140 linkers (78% of surface disulfide bonds) per Qβ by Ellman's assay (Figure S1). Glutathione (GSH) tripeptide is the most abundant thiol species in the cytoplasm of living cells and acts as a biological reducing agent. The intracellular concentration of GSH is (1-10 mM), whereas the concentration drops to about 1-10 μM in extracellular matrices. 38,39 Therefore, the cytoplasm of mammalian cells contains 100-1000 times the amount of GSH compared to the extracellular compartment creating a thiol-rich environment. This environment presents a unique opportunity for the specific and selective cytosolic release of cargo via a thiol exchange reaction with our disulfide-linker on Qβ. We hypothesized that the Qβ-M-TPP formulation, with its thiol-maleimide bonds, will undergo retro-Michael additions with the abundant GSH, separating most of the TPP from the VLP surface, and forming GSH-M-TPP. To verify this supposition, we labeled dibromomaleimide with the small fluorescent molecule FITC (Figure S3) and did an ex-vitro thiol-exchange experiment by subjecting the Qβ-M-FITC conjugate to conditions that would approximate the cytoplasm (20 mM HEPES, 100 mM KCl, 1 mM MgCl2, 1 mM EDTA, 1 mM glutathione, pH 7.4, 37 °C). 36,40 Scrambling of M-FITC linker from Qβ onto GSH was analyzed using SEC. As expected, data shown in Figure S4 indicates cleavage of the M-FITC linker from Qβ and attachment to GSH. The starting retention time of Qβ-M-FITC is 16.8 min in the SEC trace. After 24 h, however, we observed the formation of new FITC-labeled oligomers of GSH with lower MW compared to Qβ, having retention times of 21.3 and 27.9 min. We note that there is still Qβ-M-FITC after 24 h, though the peak height decreased about 61% over the course of 24 h. This observation suggests such conjugates have the potential to cleave in the cytoplasm of cells; however, to determine if the amount of cleavage was sufficient, we moved to in-vitro experiments.</p><!><p>We next moved to in vitro studies of the conjugates. A549 human lung cancer cells were employed to evaluate the cell viability, cellular uptake, and cytosolic delivery following treatment with the designed formulation. To visualize the protein trafficking into and throughout the cell, we used fluorescently engineered Qβ that contains Green Fluorescent Protein within the viral capsid-Qβ(GFP). 41 The viability of A549 cells following different treatments with Qβ(GFP) and Qβ(GFP)-M-TPP was measured by 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) assay and compared to that of untreated controls. As shown in Figure 2A, we found exceptionally low toxicity even at high concentrations, with cell viabilities greater than 97% following exposure to each formulation at a 2 mg/ml concentration after a 4 h incubation. These results indicate that the Qβ particles decorated with M-TPP should have high biocompatibility and low toxicity. Next, we treated cells with Qβ(GFP) or Qβ(GFP)-M-TPP and quantified cellular uptake with flow cytometry after fixing the cells. We observed a slight increase in uptake in cells treated with Qβ(GFP)-M-TPP, as compared to Qβ(GFP), which suggests modifying Qβ with the linker does not significantly affect particle uptake (Figure 2B). We next used fluorescence microscopy to visualize uptake and cytosolic delivery of Qβ(GFP)-M-TPP in fixed A549 cells. The obtained images revealed that the delivered Qβ(GFP)-M-TPP was evenly distributed as green fluorescence throughout the cytoplasm (Figure 2D) while Qβ(GFP) shows punctate fluorescence dots that indicate endosomal entrapment (Figure 2C). We also tested the cytosolic delivery of Qβ(GFP)-M-TPP vs Qβ(GFP) in live A549 cells as shown in Figure S5A. In the live-cell imaging studies, we observed the same results as in fixed cells. The Qβ(GFP)-M-TPP green fluorescence was diffuse in the cells, proving the ability of Qβ(GFP)-M-TPP to escape the endosome and reach the cytoplasm; however, we saw very little green fluorescence in live cells treated with Qβ(GFP) because it was quenched from the acidic environment of the endosome. 27,42 The live cells were also stained with Lysotracker and nuclear stain Hoescht dye as shown in Figure S6 and S7. Furthermore, to emphasize that TPP does not deliver Qβ(GFP) to the mitochondria because of cytosolic cleavage, the treated and fixed cells were stained with MitoTracker™ Deep Red FM. As shown in Figure 2E To further prove the role of the intermediate linker in thiol exchange cleavage and cytosolic delivery, we labeled the lysine residues of Qβ with TPP using a non-cleavable linker. 43 We expected that functionalizing the Qβ surface with TPP without a cleavable linkage would traffic our carrier, Qβ, to the mitochondria. To test this, we covalently attached the TPP to the free amine groups on the surface of Qβ through NHS ester chemistry (Figure 3A). Curiously, unlike the attachment at the disulfides, this functionalization resulted in a sensitivity of the Qβ towards ions in the buffer and resulted in poor colloidal stability and precipitation when preparing the sample in cell media (DMEM) for in vitro studies. Efforts to control this by reducing the number of TPP molecules to a relative minimum were only modestly successful. We previously found that attaching lipophilic drug molecules to the Qβ lysines creates problems for colloidal stability, and by attaching polyethylene glycol (PEG) linkers across the disulfides, we could significantly increase the solubility and stability of the Qβ. 44 Here, we applied the same strategy and first decorated Qβ with dibromomaleimide-PEG at the reduced disulfide bonds followed by attachment of TPP moieties to the free lysines. Bioconjugation conditions were optimized, and particle size, surface charge, and morphology before and after conjugation were characterized, as shown in Figure 3B-D. The mitochondria-targeting of Qβ-M-TPP was assessed in vitro by confocal fluorescence microscopy. Cells treated with Qβ(GFP)-M-TPP and stained with MitoTracker Deep Red FM were used to determine the colocalization of the green fluorescence of Qβ(GFP) conjugates with the mitochondria (Figure 3E). Colocalization was calculated using Pearson's coefficient (ρ) and NIH ImageJ. From the results, the targeted conjugates show greater colocalization with MitoTracker (ρ = 0.56) than the non-targeted conjugates (ρ = 0.28), suggesting modest colocalization within the mitochondria. Notably, a significant number of particles were still observed in the endosome/lysosome (punctate dots), but there was no diffuse fluorescence indicative of cytosolic delivery.</p><!><p>To further verify cytosolic entry and check the applicability of our delivery system, we used Qβ-M-TPP to deliver siLuc, a siRNA (smallinterfering RNA) probe capable of silencing luciferase expression in HeLa luciferase cells. 45 siRNA is a powerful therapeutic tool that inhibits specific messenger RNA (mRNA) expression in the cytosol and effectively downregulates the gene expression processes. 46 However, siRNA is highly anionic, has a relatively large molecular weight, and quickly degrades in cell media. Naked siRNA cannot readily penetrate cell membranes and reach the cytoplasm; therefore, we hypothesized that encapsulation of siRNA in Qβ-M-TPP could enhance cytosolic delivery and improve gene silencing compared to free siRNA or Qβ that lacked cleavable TPP groups. Qβ capsid formation occurs in the recombinant E. coli expression system through charge-mediated interactions between the negatively charged random E. coli RNA and positively charged Qβ coat protein subunits. 27,47 Disulfide formation in the biological synthesis of Qβ is occurring after the capsid has self-assembled likely following exposure to molecular oxygen in the bacteria or following cell lysis. Given the highly negative charge of siRNA, 48 we hypothesized siLuc could promote assembly of Qβ capsids and addition of the DB-TPP could "sew" the capsid up in a one-pot process. Specifically, we anticipated that we could reduce the Qβ with DTT and disassemble the capsid in a salt solution. The addition of siLuc could then promote capsid reassembly, and DB-TPP could seal the capsid in situ as the last step (Figure 4A).</p><p>First, purified Qβ VLPs were disassembled into coat proteins through a salt-controlled disassembly method 15,27 by adding the reducing agent 1,4-dithioerithrol (DTT) and magnesium chloride (MgCl2), which facilitates the recovery and precipitation of packed RNA. Next, the obtained coat protein (CP) was purified using dialysis and centrifugation. SEC, agarose, and SDS gel electrophoresis were used to verify the E. Coli RNA was removed and the Qβ disassembled into CPs (Figure S8). Data show the presence of CPs through a single band in SDS-PAGE and positive charge migration through agarose gel electrophoresis with a change in capsid retention time by SEC. Reassembly proceeded in the presence of siRNA under DTT reducing conditions. We found that a siRNA to CP ratio of 1:4 was sufficient to promote complete capsid formation. Finally, we proceeded with the surface installation of TPP directly in the reassembly buffer by adding 50 eq of DB-TPP to the solution. Absorbance at A260/280 ratio was used to confirm complete siLuc packing. We were happy to find that the A260/280 ratio increased from 0.87 (pre-siLuc packing) to 2.0 (post-siLuc packing). Direct observation of the VLP assemblies was done by TEM. Figure S9 shows well formed VLPs after the encapsulation of the siRNA, which is further confirmed by agarose and SDS gel electrophoresis with the hexametric and pentameric subunits of siRNA@Qβ-M-TPP exhibiting similar integrity to native Qβ (Figure S9). To assess the in vitro cytosolic delivery of siRNA using siRNA@Qβ-M-TPP, we compared the silencing ability of naked siRNA, siRNA loaded in Qβ without the M-TPP linker, and siRNA loaded in Qβ-M-TPP in luciferase-expressing HeLa cells. The cell viability and luciferase expression were measured using standard One-Glo Tox luciferase and cell viability assay. It was found that after 24 h, all treatment groups had high cell viability (Figure 4B) but the siRNA@Qβ-M-TPP treated cells showed a significantly lower luciferase expression (47%) as compared to free siRNA treated cells (93%) (Figure 4C). Considering uptake of Qβ and Qβ-M-TPP are comparable, this result further proves the cytosolic delivery of our formulation and demonstrates -as a proof-ofprinciple -that we can enhance cytosolic siRNA delivery via this approach.</p><!><p>Most nanoparticle-based delivery systems must escape the endosome and lysosome to reach the cytosol for efficient therapeutic and diagnostic action. In this work, we chemically modify the surface of an engineered protein model, Qβ, using a glutathionesensitive linker attached to a lipophilic cation to overcome endosomal entrapment and achieved cytosolic delivery. As proof-ofprinciple, we successfully demonstrate the applicability of our synthetic bioconjugation strategy in the cytosolic delivery of a VLP carrying GFP and siRNA in vitro. Our ''sheddable linker" strategy is further confirmed because no cytosolic delivery is observed without a GSH cleavable linker. We believe our synthetic strategy addresses critical challenges for the intracellular delivery of macromolecular therapeutics.</p>

ChemRxiv

Rapid in situ synthesis of polymer-metal nanocomposite films in several seconds using a CO2 laser

We demonstrate the rapid in situ synthesis of polymer-metal nanocomposite films using a CO 2 laser at 10.6 μm. The mechanism of our method is that the precursor of the metal nanoparticles, i.e., the metallic ions, is very rapidly reduced in the laser-heated polymer matrix without any reducing agent. Unlike other known laser-induced reduction methods using UV lasers, which produce radicals to promote reduction, the CO 2 laser energy is mainly absorbed by the glass substrate, and the laser-heated substrate heats the polymer matrix through heat diffusion to promote reduction. The superiority of the use of CO 2 lasers over nanosecond visible~UV lasers is also demonstrated in terms of the damage to the film. The developed method can be a new alternative to quickly synthesize a variety of polymer-metal nanocomposite films.In recent years, the synthesis of various kinds of nanoparticles (NPs) and their applications have garnered great interest, as shown, for example, in ref. 1 . For the efficient and reliable use of NPs, a uniform dispersion of NPs is key. If NPs are used in a solution, then a surfactant can be conveniently introduced to ensure the uniform dispersion of NPs without aggregation. If NPs are dispersed in inorganic or organic matrices, such materials are called nanocomposites [2][3][4][5] , which have been of recent interest since the introduction of the various kinds of filler into the inorganic or organic matrices can result in the improvement of the mechanical, electrical, and optical properties of the matrix itself by the appropriate choice of the filler [6][7][8][9] .There are many ways to synthesize organic (polymer) nanocomposites with metallic NPs as a filler 3,4 . The most straightforward way is to directly disperse metallic NPs into the polymer solution 10 . Another approach is to disperse NPs into the monomer solution and then induce polymerization 2 . An alternative method is to mix a monomer solution with a solution, which contains a precursor of the NPs, and then simultaneously induce reduction and polymerization to obtain NPs dispersed in a polymer solution 2,11 . After preparing a polymer solution with NPs, a nanocomposite film is easily obtained by spin-coating or casting methods. In addition to the above three processes, it is also possible to synthesize a polymer-metal nanocomposite film after making a polymer film, which contains a precursor of NPs, and then induce chemical 12,13 , photoinduced [14][15][16][17][18] , microwave 19 , or thermal reduction [20][21][22][23] to produce NPs in the polymer matrix. These methods are called in situ reduction methods. Note that these in situ reduction methods require a relatively long time (typically tens of minutes to tens of hours) for the synthesis of nanocomposite films. Obviously, there is still room for us to seek a rapid and scalable method to synthesize in situ polymer-metal nanocomposite films.In this paper, we demonstrate a rapid and scalable synthesis of polymer-metal nanocomposite films with a mid-infrared laser at a modest laser power. More specifically, we employ a CO 2 laser at 10.6 μm and irradiate a spin-coated polymer (polyvinyl alcohol (PVA) or polyethylene glycol (PEG)) film containing the precursor of metal (Ag) NPs on a glass substrate, as shown in Fig. 1. Unlike the well-known photoinduced methods with a UV lamp or UV laser [14][15][16][17][18] where photoexcited polymers or radicals of the additive are used to serve as a reducing agent, most of the CO 2 laser energy is absorbed by the glass substrate rather than the polymer film itself, mainly because the glass substrate (thickness~0.15 mm) has a much larger absorbance than the spin-coated polymer film (thickness ~sub-μm) on it. Indeed, we measure the transmittance of the CO 2 laser through the bare glass substrate and the free-standing polymer film to find that it is 0 and ~1, respectively. Thus, the polymer film on the laser-heated glass substrate effectively undergoes thermal annealing. In this paper, the synthesis of the Ag-PVA film is completed in only 10 sec, while the Ag-PEG film takes 10-40 sec, which may be compared with the other rapid synthesis methods developed under different contexts 24,25 . Moreover, we show that the synthesized Ag-PVA film can be made into a free-standing form [26][27][28] by the removal of the film from the glass substrate. We would like

rapid_in_situ_synthesis_of_polymer-metal_nanocomposite_films_in_several_seconds_using_a_co2_laser

<!>Experimental<!>Lasers.<!>Fabrication of the free-standing Ag-PVA film.<!>Synthesis of the Ag-PEG film.<!>Characterization of the nanocomposite films.<!>Results and Discussions<!>Conclusions

<p>to emphasize that, compared with the UV laser-based method 15 , the CO 2 laser-based method we report in this work is an interesting alternative since it is suitable for the rapid and large-area synthesis of the nanocomposite films, whether they are on the glass substrates or in free-standing form.</p><!><p>Materials. PVA (molecular weight (MW) ~60,000) and PEG (MW ~60,000 and ~500,000) are purchased from Sigma-Aldrich. Silver nitrate (AgNO 3 ) and polystylene (PS) are purchased from Wako. All the chemicals are of reagent grade and used as purchased without any further purification.</p><!><p>For most of the experiments to be presented in this work we employ a CO 2 laser at 10.6 μm (AL30P, Access Laser Co., peak power 60 W, pulse duration 100-400 μs depending on the laser power, repetition rate 2.5 kHz). Since the pulse duration is comparable to the pulse interval, which is 400 μs, it is nearly in the quasi-continuous-wave (CW) mode. To fabricate free-standing films we employ not only the CO 2 laser but also the second and third harmonic of an Nd:YAG laser (INDI 30, Spectra Physics, maximum pulse energy 80 mJ at 532 nm and 70 mJ at 355 nm, pulse duration 8 ns, repetition rate 10 Hz) for comparison. The CO 2 laser power is measured with a power metre (Pronto-250, Gentec-EO Co.), while the pulse energy of the Nd:YAG laser is measured with a thermal sensor (30A-P-SH-V1, Ophir) at the position of the polymer film. The laser beam diameter is ~8 mm (FWHM) with a Gaussian spatial profile at the position of the film for both lasers. As a result, the irradiated laser power on the film is different at different positions, which influences the film properties. Although it is possible to convert the Gaussian beam profile to a flat-top one using a beam shaper, we do not do this in this work. All the analyses at the film position are performed where the irradiated laser power is at its maximum.</p><p>Synthesis of the Ag-PVA film. PVA (0.125 g) is mixed with 2 mL of highly purified water at room temperature under continuous stirring for 20 min. Then, the solution is heated to 95 °C for 45 min to completely dissolve the PVA. The PVA solution is mixed with a separately prepared solution, which contains 0.16 g of silver nitrate and 1 mL of water. Then, the mixed AgNO 3 -PVA solution is spin-coated on a microscope cover glass (22 × 22 × 0.15 mm) at 500 rpm for 5 sec followed by 4000 rpm for 10 sec. The AgNO 3 -PVA film is dried in air at room temperature for 30 min, and then it is irradiated with the CO 2 laser or Nd:YAG laser at 355 or 532 nm under the chosen laser powers and durations.</p><!><p>To fabricate a free-standing Ag-PVA film, a PS solution is prepared from 0.375 g of PS and 3 mL of toluene to spin-coat the cover glass at 1000 rpm for 10 sec as a sacrificial layer 4 . Then, the AgNO 3 -PVA solution is spin-coated on top of it. After the drying process, the AgNO 3 -PVA film with a sacrificial PS layer is irradiated with the CO 2 laser at the chosen laser power and duration. After the CO 2 laser irradiation, the synthesized Ag-PVA films with a PS layer are peeled off from the cover glass and dipped in a toluene solution for several seconds to dissolve the PS layer. Finally, using a wire ring of ~15 mm diameter, a free-standing Ag-PVA film is obtained.</p><!><p>The process to fabricate an Ag-PEG film is almost the same as that of Ag-PVA described above. We employ PEG with two different MWs, MW~60,000 and 500,000. One gram of PEG is mixed with 2 mL and 4 mL of water, respectively. The respective PEG solutions are mixed with a separately prepared solution, which contains 0.16 g of silver nitrate and 1 mL of water. The AgNO 3 -PEG solutions are spin-coated on a cover glass, dried in air, and then irradiated with the CO 2 laser.</p><!><p>To characterize the fabricated polymer-metal nanocomposite films, we employ a compact CCD spectrometer (USB2000+, Ocean Optics), scanning electron microscope (SEM) (JSM-6500FE, JEOL) at 5 kV, and X-ray diffraction (XRD) (RINT-TTR III, Rigaku). For XRD, the diffraction angle is scanned with a speed of 5 °/min using a 10 kV micro-X-ray source obtained from the Co rotor target.</p><!><p>Film temperature. To investigate the temperature change in the film during CO 2 laser irradiation, we measure the film temperature. Since our CO 2 laser is in the quasi-CW mode (i.e., pulse duration ~ pulse interval), we can conveniently use a thermocouple with a small head (~1 mm diameter) to measure the film temperature as a function of time. The results are presented in Fig. 2 at laser powers of 1, 1.5, and 2 W for an irradiation time of 120 sec. Upon laser irradiation, the film temperature rapidly increases and reaches a nearly steady-state temperature in 30 sec regardless of the incident laser power. Naturally, the steady-state temperature is higher for the higher laser power. When we turn off the laser, the film temperature rapidly cools down to the room temperature (~25 °C). Since the boiling and thermal decomposition temperatures of bulk PVA are approximately 230 and 300 °C, respectively, we can say that the appropriate range of CO 2 laser power is 1-1.5 W.</p><p>Ag-PVA films. We now irradiate the AgNO 3 -PVA film with a laser power of 1 W. Although the film is practically transparent before CO 2 laser irradiation, the irradiated area of the film gradually turns yellow, as shown in Fig. 3a. The corresponding optical absorption spectra in Fig. 3b clearly indicate that the formation of Ag NPs in the PVA film takes place within a few seconds, and after irradiation for only 5 sec at 1 W, an SPR of the Ag NP is clearly observed. The SPR grows during the first several seconds of irradiation, and then from 10 to 20 sec, it stays nearly the same height and width (not shown here). After that, the height of the SPR gradually decreases and the tail extends to the longer wavelength side, and after the irradiation of 40 sec there is no clear tail on the long wavelength side. This change in the shape of the SPR implies the coalescence of NPs during the longer irradiation time, and this interpretation is confirmed by the SEM images (Fig. 3c-f) and the size distribution of the Ag NPs (Fig. 3g-i). The XRD spectra presented in Fig. 4 indicate additional evidence for the successful rapid in situ synthesis of Ag-PVA nanocomposite films. The crystalline size of the Ag NPs estimated from the peak width at 44° is 18.8 nm for the case of 40 sec irradiation at 1 W, which compares well with that of the 20 nm Ag-PVA film fabricated by the chemical reduction method 12 . (For the cases of 5 and 10 sec irradiation the peak heights in the XRD spectra are too small to reliably obtain the crystalline size.) Next, we fix the irradiation time to 10 sec and vary the laser power. The results are summarized in Fig. 5. By comparing the optical absorption spectra for laser powers of 1, 1.5, and 2 W (Fig. 5a), we can say that the 1 W laser power is sufficient for the 10 sec irradiation time in the sense that the use of higher laser power results in coalescence, as implied by the broader SPRs with long tails on the long wavelength side for laser powers of 1.5 and 2 W. This interpretation is again confirmed by looking into the SEM images (Fig. 5c-e) and the size distribution of the Ag NPs (Fig. 5f-h). During the 40 sec irradiation at 2 W the film temperature reaches 250 °C, and accordingly the property of the polymer matrix may change to some extent due to the partial thermal decomposition. This change is not a very serious problem, however, because we have chosen to employ a laser power of 1 W with the CO 2 laser to synthesize the nanocomposite films.</p><p>Free-standing Ag-PVA films. Now, we synthesize Ag-PVA films in a free-standing form with the procedure described above. After the CO 2 laser irradiation onto the PS + AgNO 3 -PVA film for 10 sec at 1 W, we measure the optical absorption spectra to confirm the formation of Ag NPs in the PVA film, which can be alternatively and most conveniently confirmed by the change in the film's colour, as shown in Fig. 6a. Then, we peel off the synthesized PS + Ag-PVA film from the glass substrate and dip it into a toluene solution 4 . The PS layer is dissolved into the toluene solution in a few seconds, leaving the Ag-PVA film alone. Finally, we capture the Ag-PVA film in the toluene solution with a wire ring of approximately 15 mm diameter to obtain the free-standing Ag-PVA film, as shown in Fig. 6b.</p><p>Ag-PVA films by 355 and 532 nm lasers. For comparison, we employ the third harmonic (355 nm) and second harmonic (532 nm) of an Nd:YAG laser to synthesize Ag-PVA films because it is one of the most commonly used lasers for the processing of various materials, and irradiate the AgNO 3 -PVA films at 10 Hz. Optical absorption spectra of the films after the irradiation of 100-6000 laser shots at the fluence of 100 mJ/cm 2 are shown in Fig. 7. From the fact that the SPR appear by the irradiation of 355 and 532 nm lasers and it grows as the number of laser pulses increases, we can confirm that the Ag-PVA film is successfully synthesized with the 355 and 532 nm lasers even without any reducing agent 14,15 . However, the height of the SPR in the sample produced with the 355 nm laser is much smaller than that produced by the CO 2 laser (Fig. 3b), and when the 532 nm laser is employed, the SPR is even smaller. This difference arises from the different photoabsorption mechanisms induced by the 355 and 532 nm lasers, which have a duration at 10 Hz of a few nanoseconds, and the CO 2 laser with a duration of a few hundred microseconds at 2.5 KHz: The film irradiated by the quasi-CW CO 2 laser stays at a high temperature for much longer time, which efficiently promotes the photothermal annealing needed to form Ag NPs. In contrast, the nanosecond 355 and 532 nm lasers can hardly induce the photothermal process and can only induce the photoexcitation of the polymers during the laser pulse, for which the irradiation of the 532 nm laser is less efficient than that of the 355 nm laser because the photon energy at 532 nm is too small to induce the relevant photoexcitation of the polymer molecules. A further increase in the laser fluence does not help since it damages the film. Actually, even at a fluence of 100 mJ/cm 2 , the film on the glass substrate is already damaged after a few hundred laser shots (inset photos of Fig. 7), and we cannot make a free-standing Ag-PVA film. We note that this kind of damage does not occur at all when we employ the CO 2 laser at 1~2 W, probably because the pulse duration of the CO 2 laser is very long and hence the heating is very mild. This finding demonstrates that the use of a CO 2 laser is much more suitable than a pulsed Nd:YAG laser for fabricating metal-polymer nanocomposite films without damage. Ag-PEG films. Before closing this paper, we demonstrate that we can also synthesize Ag-PEG films using the technique we have developed in this work. Figure 8 shows the optical absorption spectra of the synthesized Ag-PEG films using PEG with MW~60,000 (Fig. 8a) and the SEM images before (Fig. 8b) and after laser irradiation (Fig. 8c-e). Similar results are presented in Fig. 8f-j for the Ag-PEG films using PEG with MW~500,000. Notably, the MW of PEG we employ in this work is much larger than that of PEG (MW~200 to 6,000) employed in the earlier works 29,30 where the reduction of silver nitrate occurs in the PEG solution. For the Ag-PEG films using PEG with MW~60,000, SPR of Ag starts to appear due to the irradiation of CO 2 laser at 1 W after only several seconds, and it grows as the irradiation time increases (Fig. 8a). Note, however, that it is still much smaller than that of the Ag-PVA film (Fig. 3b) after 120 sec of irradiation. After the longer irradiation times of 40 and 120 sec, a secondary peak is observed at 350 nm. This is a plasmon resonance of the bulk Ag. The spectrum after 40 sec of irradiation has a plateau region in the range of 400-560 nm, and this may suggest that the Ag NPs produced under this condition have strong polydispersity. For the optical absorption spectra of the Ag-PEG films using PEG with MW~500,000 (Fig. 8f), 10 sec of irradiation is sufficient to obtain the eminent SPR at 450 nm. For longer irradiation times, the tail in the range of >600 nm becomes almost flat, implying that many coalescence of Ag NPs occurs.</p><p>As the successful synthesis of Ag-PEG films with our method suggests, our method should be applicable to a variety of polymer films with a metal precursor. For instance, this technique could be used to synthesize Au-PVA 15 , Ag-PEG, Au-PEG 31 , Ag-PVP ((poly)vinylpyrolidone) 32,33 , Ag-PEDOT/PSS, and so on.</p><!><p>We have demonstrated the rapid in situ synthesis of metal-polymer nanocomposite films in several seconds using a CO 2 laser. The rapid formation of nanoparticles in the polymer matrix is confirmed by the optical absorption, SEM, and X-ray diffraction measurements. The role of the CO 2 laser is to heat the glass substrate, which indirectly heats polymers in the film through thermal diffusion. Although the polymers (polyvinyl alcohol and polyethylene glycol) employed in this work have very small reducing power at room temperature, the heated polymers rapidly reduce Ag ions, and eventually Ag nanoparticles form in the polymer matrix over the course of several seconds. Therefore, our method is more similar to thermal annealing rather than UV laser annealing. The advantage of our method is that, without introducing any reducing agent, we can synthesize nanocomposite films in several seconds at a modest laser power density of ~1 W/cm 2 with a compact commercial CO 2 laser, and the process is scalable since a high-power CO 2 laser for industrial purposes is readily available for the large-area synthesis of nanocomposite films. We have also shown that the use of a CO 2 laser is very suitable for the synthesis of free-standing nanocomposite films. The new fabrication method of nanocomposite films developed in this work can be a new alternative for quickly synthesizing a variety of polymer-metal nanocomposite films.</p>

Scientific Reports - Nature

Alkynes as Electrophilic or Nucleophilic Allylmetal Precursors in Transition Metal Catalysis

Diverse late transition metal catalysts convert terminal or internal alkynes to transient allylmetal species that display electrophilic or nucleophilic properties. Whereas classical methods for the generation of allylmetal species often mandate formation of stoichiometric byproducts, recent use of alkynes as allylmetal precursors enables completely atom-efficient catalytic processes, including enantioselective transformations.

alkynes_as_electrophilic_or_nucleophilic_allylmetal_precursors_in_transition_metal_catalysis

1. Introduction<!>2. Alkynes as Electrophilic \xcf\x80-Allyl Precursors<!>2.1. C-O Bond-Forming Reactions<!>2.2. C-S Bond-Forming Reactions<!>2.3. C-N Bond-Forming Reactions<!>2.2. C-C Bond-Forming Reactions<!>3. Alkynes as Nucleophilic \xcf\x80-Allyl Precursors<!>4. Summary and Outlook

<p>As exemplified by the Tsuji-Trost reaction[1] and carbonyl allylation,[2] allylmetal species may display electrophilic or nucleophilic properties,[3,4] respectively. Consequently, allylmetal complexes are found to mediate an especially diverse range of catalytic or stoichiometric transformations.[3] Classical methods for the generation of discrete or transient allylmetal species often exploit precursors that mandate generation of stoichiometric byproducts, for example, allylic carboxylates.[1–3] It has been shown that allenes[5,6] or dienes[7,8] provide completely atom-efficient access to allylmetal intermediates. Although the stoichiometric conversion of alkynes to allenes[9] and π-allylmetal complexes[10] has been documented, the use of alkynes as allylmetal precursors is far less developed. However, the use of alkynes as precursors to both electrophilic or nucleophilic allylmetal species recently has witnessed enormous progress. In the context of electrophilic allylation, that is, alkyne-mediated Tsuji-Trost reactions, catalysts based on palladium[11–15] and rhodium[16–19] have been developed. Even more recently, alkyne-mediated nucleophilic allylations of carbonyl compounds under the conditions of iridium[20] or ruthenium[21] catalysis have been reported. In both electrophilic or nucleophilic modes, completely atom-efficient catalytic processes are enabled, and in several cases enantioselective transformations have been achieved.[11–21] Many of these processes represent examples of "tandem catalysis," as a single metal complex will catalyze the isomerization of an alkyne to an allene and, in a separate catalytic process, transform the allene to a product of electrophilic or nucleophilic allylation. In this mini-review, we offer an exhaustive account of this emerging area of research.</p><!><p>Allylic substitution and oxidation reactions to generate linear or branched adducts is an important research topic in modern organic chemistry, especially due to the versatility of the allyl moiety in terms of further functionalization. However, from the perspective of modern chemical synthesis, one can recognize that a major drawback of these methods is the requirement of preinstalled leaving groups and consequent generation of stoichiometric byproducts, which diminish atom-economy[22] and synthetic efficiency[23] (Scheme 1). Hence, the byproduct-free addition of pronucleophiles to allenes and alkynes is an attractive alternative.</p><!><p>Trost and co-workers were the first to demonstrate the concept of "tandem catalysis" embodied by alkyne-mediated allylic substitutions. In 1992, under the conditions of palladium catalysis using carboxylic acid pronucleophiles, propargyl acetates were isomerized to the corresponding allenes. Allene hydropalladation forms a π-allylpalladium intermediate, which upon addition of acetate delivers the indicated gem-diacetates (Scheme 2).[11a,12] A variety of propargyl acetates were converted to the corresponding gem-diacetates in good yields with exclusive olefin (E)-stereoselectivity. Beyond intermolecular additions of acetic acid, macrocycle formation also occurs in excellent yield with good olefin (E:Z)-stereoselectivity and good diastereocontrol.</p><p>In 2001, Yamamoto and co-workers reported the first in a series of studies involving the use of internal alkynes as π-allyl precursors. Using a palladium catalyst in combination with substoichiometric benzoic acid, aryl propynes react with alcohols to form linear products of hydroalkoxylation as single regioisomers (Scheme 3).[11b] Application of these conditions to acetylenic alcohols, enabled formation of tetrahydrofurans and tetrahydropyrans, albeit in modest yields.</p><p>The proposed mechanism for this transformation, which encompasses the tandem action of two discrete catalytic cycles, is shown in Scheme 4. In the first cycle, the palladium-hydride (A), generated from benzoic acid, hydropalladates the alkyne to form the vinyl-palladium species (B), which upon β-hydride elimination releases the allene (C). Then, in a second catalytic cycle, allene hydropalladation delivers a π-allylpalladium species, which upon regioselective alcohol addition releases the linear allylic ether along with palladium(0).[5]</p><p>An alternative class of oxygen pro-nucleophiles, carboxylic acids, were found to be suitable reaction partners in alkyne-mediated Tsuji-Trost reactions, as reported independently by Zhang[11c] and Yamamoto.[11d] Using a palladium catalyst, a range of different aromatic and aliphatic carboxylic acids are converted to the linear allylic esters with complete alkene (E)-stereoselectivity (Scheme 3). Later in 2006, Yamamoto reported an enantioselective variant of his previously described cyclization of acetylenic alcohols.[11e] Using a chiral palladium catalyst modified by (R,R)-renorphos, tetrahydrofurans, tetrahydropyrans and isochromanes were generated with modest levels of enantiomeric enrichment (Scheme 5). This reaction is limited to internal aryl-substituted alkynes. One year later, a related palladium catalyzed cyclization of acetylenic carboxylic acid to form racemic γ- and δ-lactones was achieved (Scheme 5).[11f]</p><p>In 1987, Werner described the stoichiometric reaction of the indicated rhodium-2-butyne complex with Brønsted acids possessing non-coordinating counterions to form the π-allyl rhodium complex A (Scheme 6).[10b] This transformation was postulated to involve initial rhodium hydride formation, alkyne hydrometalation and subsequent β-hydride elimination to deliver the allene complex B, which could be detected by NMR spectroscopy. Upon warming, allene complex B isomerizes to form π-allyl complex A. This observation suggested the feasibility of utilizing rhodium complexes as catalysts for alkyne-based Tsuji-Trost reactions.</p><p>In 2011, Breit reported the first alkyne-based Tsuji-Trost reaction catalyzed by rhodium (Scheme 7).[16a] A broad range of alkyl-substituted terminal alkynes react efficiently with various aliphatic, aromatic and α,β-unsaturated carboxylic acids to form the branched allylic esters as single regioisomers. Such branched regioselectivity, which also is evident in conventional rhodium-catalyzed allylic substitutions[24] and related allene hydrofunctionalizations,[5] complements the linear regioselectivity observed in corresponding palladium catalyzed reactions. This transformation proved very sensitive to the nature of the ligand. Whereas rhodium-DPEPhos provides the branched allylic esters, the P/N-ligand DPPMP provides the (Z)-enol esters as the major product in good yields and selectivities (Scheme 7).[16b,c]</p><p>An intramolecular variant of this process enabled an atom economic macrolactone synthesis (Scheme 8).[16d] Remarkably, the macrolactonization does not require high dilution or syringe pump addition to avoid the formation of diolides or higher oligomers. The synthesis of γ- and δ-lactones was possible, although attempted formation of medium-sized lactones resulted in diolide generation. In 2017, Breit applied this method as a key step in the total synthesis of the 16-membered macrolactone epothilone D (Scheme 8).[16e] The rhodium-catalyzed macrolactonization proceeded in a diastereoselective fashion, albeit in moderate yield with high catalyst loading. This synthesis also utilizes several other methods developed in the Breit laboratory, including the stereospecific zinc-catalyzed cross-coupling of α-hydroxy ester triflates with Grignard reagents[25] and hydroboration-magnesium exchange.[26]</p><p>Mechanistic studies involving both experimental and computational methods corroborate the indicated catalytic cycle for the rhodium catalyzed alkyne-based Tsuji-Trost reaction (Scheme 9).[16f] The rhodium chloride dimer modified by DPEphos dissociates to form a monomer, which in the presence of the terminal alkyne forms the η2-alkyne complex C. The carboxylic acid binds to the rhodium center, forming a rather stable hydrogen-bond to the terminal sp-carbon. Complete proton transfer to this carbon delivers the vinylrhodium carboxylate E. Reductive elimination from this species accounts for the formation of enol-ester side-products. β-Hydride elimination from E leads to the hydridorhodium allene complex F. Isotopic labeling experiments corroborate fast and reversible interconversion of E and F. Allene hydrometalation furnishes the π-allyl complex G, which is in equilibrium with the thermodynamically more stable σ-allyl complex. As established by in situ IR, NMR and MS studies, the σ-allyl complex is the resting state of the catalytic cycle. Indeed, the σ-allyl complex H could be isolated and was characterized by X-ray diffraction analysis. DFT calculations indicate a reisomerization from H to the π-allyl G followed by an inner sphere reductive elimination process to form the allylic C-O bond.</p><p>Catalytic enantioselective formation of branched allylic esters from alkynes was achieved by Breit in 2015 using chiral rhodium catalysts modified by DIOP-family phosphine ligands (Scheme 10).[16g] Although the reaction products are themselves potential π-allyl precursors, erosion of enantiomeric enrichment was not observed. This methodology enabled protecting group-free syntheses of trans-whiskey and cognac lactones. In 2016, Breit reported that aliphatic alcohols participate in enantioselective alkyne-based Tsuji-Trost allylation using chiral rhodium catalysts modified by (R,R)-DTBM-Garphos (Scheme 10).[16h] The use of diphenylphosphate as Brønsted acid cocatalyst was essential in terms of promoting alkyne-to-allene isomerization.</p><!><p>Despite their significance in medicinal chemistry,[27] S-nucleophiles are not frequently employed in metal catalyzed C-S bond formation, perhaps due to catalyst poisoning.[28] In 2014, Breit reported the use of sulfonyl hydrazides as precursors to sulfinic anions in rhodium catalyzed alkyne-based Tsuji-Trost allylations (Scheme 11).[17] Although a Brønsted acid cocatalyst, benzoic acid, is required to promote alkyne-to-allene isomerization, the more reactive S-nucleophile captures the resulting π-allyl. The allylic sulfones are formed in good yield with complete branched regioselectivity. Later, in 2016, Lu and Jin developed a related palladium catalyzed process that delivers the linear regioisomers (Scheme 11).[13]</p><!><p>Nitrogen-containing compounds are ubiquitous in medicinal and agricultural chemistry.[29] As over 25 of the 150 top-selling small-molecule drugs incorporate stereogenic C-N bonds,[30] atom-efficient methods that convert π-unsaturated feedstocks to amines, enamines or imines remain in high demand.[22,31,32] Under the conditions of palladium catalysis, Yamamoto developed alkyne-based Tsuji-Trost aminations of aliphatic secondary amines or aniline-derivatives (Scheme 12).[14a,b] In both cases, benzoic acid serves as a co-catalyst and the allylic amines are generated with complete alkene (E)-stereoselectivity. Use of aliphatic primary amines led to significant quantities of N,N-diallylated product.</p><p>Intramolecular variants of this reaction deliver pyrrolidines, piperidines, tetrahydroisoquinolines and tetrahydroquinolines (Scheme 13).[14c-f] Although rather high catalyst loadings are required, chiral palladium catalysts modified by (R,R)-renorphos provide access to enantiomerically enriched heterocycles. To highlight the utility of this chemistry, Yamamoto applied the palladium-catalyzed alkyne hydroamination reaction as a key step for the construction of indolizidine alkaloid (−)-209D (10) (Scheme 13).[14g] The hydroamination precursor 9 was prepared in five steps from the (L)-proline derived aldehyde 8. Highly diastereoselective palladium catalyzed alkyne hydroamination followed by alkene hydrogenation delivered indolizidine (−)-209D (10). Furthermore, N-Tosyl amides also participate in cyclization reactions to furnish racemic δ-lactams.[14h]</p><p>In 2015, Dong reported the rhodium catalyzed hydroamination of internal aromatic alkynes with indolines (Scheme 14).[18a] Remarkably, upon variation of the acidic additive, the formation of either linear or branched regioisomers could be achieved. Using the chiral rhodium catalyst modified by (S,S)-bdpp in combination with xylylic acid, enantiomerically enriched branched adducts were formed with moderate to good selectivities and yields. Conversely, using an achiral rhodium catalyst with phthalic acid, linear adducts are formed.</p><p>In a contemporaneous independent study, Breit showed that the chiral rhodium catalyst modified by JoSPOphos converts internal methyl-substituted or terminal alkynes to branched N-allylated pyrazoles (Scheme 15).[18b] Using pyridinium p-toluene sulfonic acid (PPTS) as cocatalyst, good yields were accompanied by excellent regioselectivity and good to excellent levels of enantioselectivity. Notably, nonsymmetric pyrazoles engage in highly position-selective N-functionalization.</p><!><p>The first examples of C-nucleophiles in alkyne-based Tsuji-Trost reactions were reported by Yamamoto in 1998 and 2004 (Scheme 16).[15a,b] Using a palladium catalyst and benzoic acid as cocatalyst, internal alkynes react with malonitrile and malonic ester derivatives to form linear adducts in good to excellent yield. Using α-branched or cyclic β-dicarbonyl pronucleophiles, racemic all-carbon quaternary stereocenters were formed in an efficient manner (Scheme 16).[15c] Improvements to these conditions were subsequently disclosed. Using DavePhos as ligand, these transformations could be conducted at significantly lower temperatures (50°C).[15d] A solvent free, microwave assisted version decreased reaction time to only a few minutes.[15e] Additionally, intramolecular variants[11e] and the use of activated aldehyde pronucleophiles were reported.[14b]</p><p>Based on this work, in 2016 Lin developed a cooperative palladium/proline-catalyzed process wherein unactivated ketones or aldehydes are coupled to aryl propynes to form racemic γ,δ-unsaturated ketones and aldehydes in moderate to good yields (Scheme 17).[15f] The structure of the amino acid cocatalyst was decisive. Only (L)-proline enforced high levels of efficiency, which was rationalized by postulating intervention of the indicated 7-membered palladacycle. In a concurrent study, Lin and Yao reported the dearomatizing allylic alkylation of indoles with aryl propynes to furnish indolines bearing C3-quaternary stereocenters. Again, carboxylic acid cocatalysts were required as additives (Scheme 17).[15g]</p><p>The first rhodium-catalyzed addition of C-pronucleophiles to alkynes was reported by Breit in 2016 (Scheme 18).[19a] A diverse range of 1,3-dicarbonyl partners couple efficiently with terminal alkynes with complete levels of branch-regioselectivity. Again, the Brønsted acid cocatalyst played a critical role. Electron poor carboxylic acids were required to obtain the resulting γ,δ-unsaturated diketones in good to excellent yields. To demonstrate the utility of this methodology, the allylated adducts were converted to high-value building blocks. For example, treatment with ethanolic potassium hydroxide promotes retro-Claisen condensation to form the γ,δ-unsaturated ketones or trisubstituted dihydropyrans. Related reactions of internal alkynes were subsequently developed and applied to the synthesis of heterocycles (Scheme 18).[19b] In 2016, the laboratories of Dong and Breit simultaneously demonstrated that direct access to γ,δ-unsaturated ketones may be achieved in decarboxylative couplings of alkynes with β-keto acids (Scheme 19).[19c,19d]</p><p>More recently in 2017, Dong employed two chiral co-catalysts to promote diastereo- and enantioselective alkyne-based allylations of α-aryl aldehydes (Scheme 20).[19e] A chiral amine is used to generate a transient enamine from the aldehyde pronucleophile, and an (R)-BINAP-modified rhodium-catalyst is used to generate a chiral allylrhodium intermediate. Depending on the choice of chiral amine two of the four stereoisomers of the product can be obtained selectively.</p><!><p>Unlike related alkyne-based Tsuji-Trost allylations, the use of alkynes as nucleophilic π-allyl precursors is far less developed. To date, this mode of reactivity has only been observed in hydrogen transfer-mediated carbonyl allylations of alcohol proelectrophiles.[2j,33] In 2009, following reports by Krische on related allylative processes involving allenes,[6] dienes[8] and allylic carboxylates,[34] Obora and Ishii reported the first examples of alkyne-alcohol C-C coupling to form products of carbonyl allylation.[20a,b] Specifically, using the iridium catalyst derived from [Ir(OH)(cod)]2 and P(nOct)3 primary alcohols react with aryl propynes to provide racemic products of (α-aryl)allylation with complete levels of anti-diastereoselectivity. Although mechanistic studies were not undertaken, previously reported alcohol-allene C-C couplings deliver products of carbonyl allylation,[6] suggesting a catalytic cycle involving alkyne-to-allene isomerization is operative (Scheme 21).</p><p>In ruthenium catalyzed alkyne-alcohol C-C couplings developed by Krische, subtle changes in reaction conditions can result in widely different outcomes (Scheme 22). For example, using the cationic ruthenium catalyst generated through the acid-base reaction of [H2Ru(CO)(PPh3)3] and 2,4,6-(2-Pr)3PhSO3H, alkynes isomerize to form transient allenes. The cationic ruthenium complex exists in equilibrium with a ruthenium(0) species that promotes allene-aldehyde oxidative coupling to form an oxaruthenacycle, which upon primary alcohol-mediated transfer hydrogenolysis delivers the (Z)-homoallylic alcohols.[21a] Oxidative coupling pathways are suppressed upon introduction of iodide ion and a chelating phosphine ligand, Josiphos SL-J009-1, yet alkyne-to-allene isomerization pathways are maintained. Under these conditions, alcohol dehydrogenation triggers allene hydrometalation to form chiral allylruthenium-aldehyde pairs that deliver enantiomerically enriched branched homoallylic alcohols as single diastereomers.[21b] For both reaction types, an extensive series of deuterium labelling studies were undertaken to corroborate the proposed mechanisms (Scheme 22).</p><p>Krische reports that a third distinct mechanism for alkyne-alcohol C-C coupling is operative when the preceding conditions (Scheme 22)[21b] are applied to the propargyl ether, TIPSOCH2C ≡CMe (TIPS = triisopropylsilyl). The n-σ* interaction between the silyl ether oxygen atom and the propargylic C-H bond promotes a 1,2-hydride shift that converts the metal-bound alkyne to a vinyl carbene, which upon protonation forms the indicated siloxy-π-allylruthenium nucleophile. Carbonyl addition occurs from the σ-allylruthenium haptomer where ruthenium resides at the oxygen-bearing carbon, presumably due to the negative inductive effect of oxygen. Using a Josiphos (SL-J009-1) modified ruthenium(II) catalyst, the resulting products of siloxy-crotylation form as single regioisomers with complete levels of anti-diastereoselectivity and high levels of enantioselectivity.[21c] Although mixtures of enol geometrical isomers are produced, the (E/Z)-selectivity is inconsequential as fluoride assisted cleavage of the enol in the presence of NaBH4 converts both isomers to the same 1,4-diol (Scheme 23).</p><p>This pathway for hydride shift enabled π-allyl formation is general and transferrable to other metal catalysts. Using the chiral iridium complex formed in situ from [Ir(cod)Cl]2 and (R)-H8-BINAP, Krische reports the direct enantioselective C-C coupling of the simple propargyl ether, TIPSOCH2C≡CH, with primary alcohols to form products of (Z)-siloxyallylation. Uniformly high levels of enantioselectivity are accompanied by complete alkene (Z)-stereoselectivity (Scheme 24).[20c]</p><!><p>New reactivity is the foremost basis for methodological innovation in the field of chemical synthesis. The recent ability of metal catalysts to transform alkynes into π-allylmetal species streamlines methods for electrophilic or nucleophilic allylation. In alkyne-based Tsuji-Trost reactions, one circumvents generation of stoichiometric byproducts associated with the use of allylic carboxylates, the traditional electrophilic π-allyl precursors. Similarly, in alkyne-based nucleophilic allylations via hydrogen auto-transfer, one bypasses the requirement of discrete allylmetal reagents, enabling completely atom-efficient carbonyl addition. In both polarity modes, highly enantioselective processes have been established and hitherto inaccessible extensions in scope have been achieved. It is the authors hope that the present review of this bourgeoning area will accelerate further progress toward allylative transformations of alkynes.</p>

PubMed Author Manuscript

Cathepsin B-deficient mice as source of monoclonal anti-cathepsin B antibodies

Cathepsin B has been demonstrated to be involved in several proteolytic processes that support tumor progression and metastasis and neurodegeneration. To further clarify its role, defined monoclonal antibodies are needed. As the primary structure of human cathepsin B is almost identical to that of the mouse, cathepsin B-deficient mice were used in a novel approach for generating such antibodies, providing the chance of an increased immune response to the antigen, human cathepsin B. Thirty clones were found to produce cathepsin B-specific antibodies. Seven of these antibodies were used to detect cathepsin B in MCF10-DCIS human breast cancer cells by immunocytochemistry and immunoblotting. Five different binding sites were identified by epitope mapping giving the opportunity to combine these antibodies in oligoclonal antibody mixtures for an improved detection of cathepsin B.

cathepsin_b-deficient_mice_as_source_of_monoclonal_anti-cathepsin_b_antibodies

<p>The lysosomal cysteine proteinases of the papain family – the cathepsins B, C, F, H, K, L, O, S, V, W and X/Z – are involved in a variety of physiological and pathological processes. Most of the cysteine cathepsins are endopeptidases and eight of them were shown to contribute to tumor development and progression (Gocheva et al., 2006 ; Watson and Kreuzaler, 2009 ; Reiser et al., 2010 ; Mullins et al., 2012). Cathepsin B (EC 3.4.22.1), which also has a peptidyl-dipeptidase activity, is constitutively expressed in normal cells and overexpressed in many human malignancies by tumor cells and tumor-associated cells at the mRNA and protein levels (Podgorski and Sloane, 2003 ; Mohamed and Sloane, 2006 ; Vasiljeva et al., 2006 ; Andl et al., 2010). Cathepsin B has been linked to apoptosis, tumor-associated inflammation, angiogenesis and tumor progression and metastasis by contributing to the altered intracellular protein metabolism of cancer cells and to proteolytic cascades in the microenvironment of tumors. In cancer cells, lysosomes are redistributed from the perinuclear area to the cellular periphery, where they can release cathepsins or be secreted into the extracellular space to contribute to matrix degradation and tumor cell invasion. Cathepsin B is a prognostic marker in several types of cancer and its increased expression by tumor cells is correlated with poor outcome, e.g., in breast cancer (Podgorski and Sloane, 2003 ; Joyce et al., 2004 ; Sloane et al., 2005 ; Nagaraj et al., 2006 ; Fehrenbacher et al., 2008 ; Malla et al., 2011 ; Sevenich et al., 2011 ; Gopinathan et al., 2012 ; Rafn and Kallunki, 2012, Rafn et al., 2012).</p><p>There is growing evidence that cathepsin B may have the potential to be a therapeutic target for reducing the malignant progression of tumor cells and for treating some kinds of metastatic cancer because ablation or inhibition of cathepsin B in tumor models decreased or delayed metastasis (Mohanam et al., 2001 ; Bervar et al., 2003 ; Fehrenbacher and Jäättelä, 2005 ; Bell-McGuinn et al., 2007 ; Vasiljeva et al., 2008 ; Gopinath et al., 2010 ; Victor et al., 2011 ; Reinheckel et al., 2012 ; Withana et al., 2012 ; Rothberg et al., 2013). The depletion of cathepsins B and L is able to completely reverse the invasive phenotype of MCF7 cells and HER2-expressing SKBR-3 and MDA-MB-453 cells (Rafn et al., 2012). The overexpression of mouse mammary tumor virus-polyoma middle T antigen (PyMT) in mouse mammary gland epithelium results in higher cathepsin B levels and increased metastasis (Vasiljeva et al., 2006 ; Sevenich et al., 2011 ; Bengsch et al., 2013).</p><p>Cathepsin B has also been shown to participate in the production of brain pyroglutamate amyloid-beta, thus contributing to the development of Alzheimer's disease (Hook et al., 2014).</p><p>To elucidate its role in these processes, the cathepsin B protein must be efficiently and thoroughly detected, e.g., by specific antibodies. In order to get specific and high affinity mouse anti-human cathepsin B monoclonal antibodies we tried a novel approach, i.e., cathepsin B-knockout mice as the basis for generating antibodies to human cathepsin B. As the sequence of human cathepsin B differs from that of the mouse in only a few amino acids, the chance of human cathepsin B being recognized as a foreign protein by the mouse immune system is low. We, therefore, tried to provoke an immune response in cathepsin B-deficient knockout mice as the basis for the generation of anti-cathepsin B monoclonal antibodies. We also used active human cathepsin B for immunization because recombinant cathepsin B had failed in normal mice in several previous efforts to result in high affinity antibodies against native cathepsin B. Cathepsin B was purified from the supernatants of the human non-small cell lung cancer cell line 32M1 according to a novel isolation protocol (Figure 1).</p><p>The antibodies were generated by a modified version of the Köhler and Milstein (1975) method. Three-month old cathepsin B-deficient mice were immunized with 20 μg of single-chain human cathepsin B emulsified in 300 μl of complete Freund's adjuvant (CFA). The intraperitoneal immunization was repeated 48 and 117 days after the first injection and 1 day before hybridization on day 121 of the immunization process with 20 μg of the antigen in 300 μl of incomplete Freund's adjuvant (IFA).</p><p>Mouse spleen cells (2 × 107) and cells (5 × 107) of the mouse myeloma cell line P3X63Ag8.653 were thoroughly mixed and cell fusion was mediated by polyethylene glycol 1500 (Boehringer, Mannheim, Germany). Fused cells were cultured in RPMI-based HAT selection medium (Invitrogen, Karlsruhe, Germany; 100 μM hypoxanthine, 4 mM aminopterine, 160 μM thymidine) for 3 weeks and another week in HT-medium (Invitrogen, Karlsruhe, Germany; 100 μM hypoxanthine, 160 μM thymidine). Surviving hybridomas were grown in RPMI 1640 medium supplemented with 10% of an ultra low IgG fetal calf serum (Life Technologies, Germany), HEPES (4.77 g/l), glucose (2.5 g/l), 2-mercaptoethanol (1 mM), L-glutamine (2 mM), insulin (Roche, Mannheim, Germany; 5 mg/l) and gentamycin (Serva, Heidelberg, Germany; 80 mg/l). Hybridomas were screened for the production of cathepsin B-specific antibodies in an indirect ELISA.</p><p>Thirty different anti-cathepsin B antibody-producing clones were identified among approximately 650 hybridomas. The specificity of these clones and their functionality were further examined by immunoblotting against purified human cathepsin B. Nine of these antibodies were selected for further characterization by determination of the epitopes (Figures 2 and 3 ; Table 1) they recognize, and seven of them by their performance in detecting cathepsin B in human MCF10-DCIS breast cancer cells by immunoblotting (Figure 4) and immunocytochemistry (Figure 5). These antibodies did not cross-react with cathepsins H, L and S.</p><p>The PepSpot technique was used for identifying the epitopes detected by the antibodies. The sequence of procathepsin B was synthesized as a library of 110 spots of 13mer peptides onto a cellulose membrane (Jerini, Germany). The sequences of the peptides of neighboring spots were overlapping to ensure the continuous presentation of the complete procathepsin B sequence. After washing the membrane with a casein-based blocking buffer, it was incubated with the primary antibody (1 μg/ml) and after washing with Tween-TBS buffer with HRP-conjugated secondary antibody. The chemiluminescent light detection of the recognized spots was performed with a FujiFilm LAS4000 luminescent image analyzer directly from the blots.</p><p>A direct comparison of the performance of the obtained antibodies in immunoblotting and immunocytochemistry revealed only minor differences in their effectiveness in detecting cathepsin B. They all recognize cathepsin B. Based on immunoblotting, mAb 3E4 clearly is the most effective, followed by 1H1, 1G5, 6D5 and 3C3. MAb 8F10 and 4H9 gave the weakest signal and had the lowest efficacy in recognizing cathepsin B in immunoblots, whereas in immunofluorescence mAB 4H9 and 8F10 are the most effective in detecting the protein. Their signals in the images were bright, and punctuate staining was well-defined in the periphery of all cells. All other mAbs, i.e., 3C3, 6D5, 1H1, 1G5 and 3E4, also detected cathepsin B.</p><p>With this novel approach we were able to generate a panel of monoclonal anti-human cathepsin B antibodies that proved to be able to detect cathepsin B in standard techniques like immunoblotting and immunofluorescence. The determination of the binding sites in the cathepsin B molecule revealed that these selected antibodies recognized five different epitopes (Figure 3). This allows the combination of single monoclonal antibodies in oligoclonal mixtures in order to enhance the intensity of the analytical signals in immunoblotting or immunocytochemistry or both. We saw that the signals were more intense when several combinations of these antibodies were used, clearly improving the detection of cathepsin B.</p><p>SDS-PAGE (12% gels) under reducing conditions of purified single-chain human cathepsin B.</p><p>Silver staining. Left lane: marker proteins (11, 17, 26, 34, 43, 55 kDa). Right lane: single-chain cathepsin B (0.75 μg) of a molecular mass of 31 kDa. Human procathepsin/cathepsin B was purified from the conditioned medium of the human non-small cell lung cancer cell line EPLC-32M1 according to a novel protocol. Cells (1 × 108) were cultured in RPMI-1640 medium supplemented with 10% FCS. After dialyzing the supernatants against 0.1 M glycine-buffer (pH 3.0) to degrade bovine serum albumin of the medium by activating procathepsin D, also secreted by the EPLC-32M1 cells, the supernatants were equilibrated with 50 mM Tris-HCl (pH 7.3) containing KCl (100 mM) and run on a Blue-Sepharose (GE Healthcare Europe) column to remove remnants of albumin. Procathepsin B containing fractions were collected, dialyzed against tri-ethanolamine (20 mM, pH 7.4) and fractionated with a NaCl gradient (0 to 0.15 m) on a DEAE-Sephacel (GE Healthcare Europe) column equilibrated in the same buffer. During this fractionation step procathepsin B was processed to active single-chain cathepsin B (Weber, unpublished data). Its activity was determined with Z-Arg-Arg-NHMec as substrate (10 μM Z-Arg-Arg-NHMec, 50 mM phosphate buffer, 2.5 mM DTT, 2.5 mM EDTA, pH 5.5) and its specificity with the highly selective cathepsin B inhibitor CA-074 (10 μM), which completely blocked the activity (Murata et al., 1991). The catalytic rate constant, kcat, was 172 (S−1) and the Michaelis constant, Km, was 175 μM. 3 mg purified cathepsin B were obtained from 1200 ml conditioned medium.</p><p>Epitope mapping.</p><p>Spot 43: AHVSVE VSAEDLL</p><p>Spot 44: SVE VSAEDLL TCC</p><p>Spot 45: VSAEDLL TCCGSM</p><p>Detected sequence: V133 SAEDLL139(B) mAb 3E4; (C) mAb 1H1; (D) mAb 8D7.</p><p>Schematic representation of the epitopes of 9 of the anti-cat B monoclonal antibodies in cathepsin B.</p><p>The sequence C211 EPGYSP is located on the surface of cathepsin B, based on the crystal structure of human cathepsin B (PDB file 1CSB-2dcd).</p><p>Immunoblot analysis demonstrating that the selected mAbs detect cathepsin B in lysates (0.68 μg DNA/sample) of MCF10-DCIS human breast cancer cells.</p><p>Lysates were run under reducing conditions on 12% SDS-PAGE gels at 50 mA and the separated proteins were transferred to nitrocellulose membranes. The membranes were incubated with the mAbs (1:500) for 2 h at room temperature and a secondary anti-mouse antibody (1:10 000). Detected cathepsin B was visualized through Western Lightening Plus ECL – Western Chemiluminescence Substrate. The 25/26 kDa band represents the heavy chain of the active two-chain form of cathepsin B and the 31 kDa band the active single-chain form of cathepsin B. Second step controls were negative and are not shown.</p><p>Immunofluorescence staining with the antibodies 8F10, 4H9, 6D5 and 3C3 for intracellular cathepsin B (green) in methanol fixed (−20° C) MCF10-DCIS cells cultured on glass coverslips for 48 h. When cell cultures reached 80% confluency, cells were fixed using cold methanol (−20° C) for 5 min. For blocking, non-specific binding sites cells were incubated with 0.2% BSA/PBS at room temperature for 45 min and then stained with the primary antibodies diluted 1:40 in 0.1% saponin/PBS overnight at 4° C. Excess primary antibodies were removed by washing with 0.1% saponin/PBS and the secondary antibody (1:1000) was then added for 1 h at room temperature under dark conditions. Finally, cells were washed with aqua dest. and left to dry. Coverslips were mounted on microscope slides. A 2 μl aliquot of an Antifade solution was added and clear nail polish was used to seal the edges. The cells were observed with a Zeiss LSM 310 microscope – epitome setting with optical sectioning and DIC phase contrast under oil immersion at an original magnification of 40x. Bar, 10 μM.</p><p>List of monoclonal antibodies and the cathepsin B epitopes they are detecting as determined by the PepSpot technique.</p><p>All antibodies are IgG1-type antibodies.</p>

PubMed Author Manuscript

Non-neural surface ectodermal rosette formation and F-actin dynamics drive mammalian neural tube closure

The mechanisms underlying mammalian neural tube closure remain poorly understood. We report a unique cellular process involving multicellular rosette formation, convergent cellular protrusions, and F-actin cable network of the non-neural surface ectodermal cells encircling the closure site of the posterior neural pore, which are demonstrated by scanning electron microscopy and genetic fate mapping analyses during mouse spinal neurulation. These unique cellular structures are severely disrupted in the surface ectodermal transcription factor Grhl3 mutants that exhibit fully penetrant spina bifida. We propose a novel model of mammalian neural tube closure driven by surface ectodermal dynamics, which is computationally visualized.

non-neural_surface_ectodermal_rosette_formation_and_f-actin_dynamics_drive_mammalian_neural_tube_clo

Introduction<!>Animals<!>Scanning electron microscopy (SEM)<!>Wholemount immunofluorescence, F-actin labeling, and confocal microscopy<!>Dynamic visual modeling of neural tube closure at single-cell resolution<!>Scanning electron microscopy reveals multicellular structures of non-neural surface ectodermal cells during posterior neural tube closure<!>The rosette formation and convergent F-actin protrusions are disrupted in the surface ectodermal Grhl3 mutants with fully penetrant spina bifida<!>Genetic fate mapping and immunofluorescence confocal microscopy demonstrate the surface ectodermal nature of the rosette-forming cells that generate the F-actin protrusions and cable network during posterior neural tube closure<!>Computational visual model of the dynamic neural tube closure process driven by non-neural surface ectodermal cells and F-actin dynamics<!>Discussion<!>

<p>The neural tube is the embryonic precursor structure of the brain and spinal cord. During mammalian neurulation, neural tube closure initiates at the border region of the future brain and spinal cord, and the closure spreads rapidly towards both anterior and posterior directions to seal up the brain and spinal cord [1,2]. Defective closure of the neural tube may cause neural tube closure defects (NTDs), which is among the commonest and severest structural birth defects in humans [3–6]. Among various anatomically distinct NTDs, spina bifida, resulted from defective caudal neural tube closure, is the commonest NTDs affecting numerous liveborn infants [7,8]. Mutant mice have been long and widely used to address the cause and prevention of human neural tube defects [9–14]. Thus, understanding the basic mechanisms of normal and defective neural tube closure, particularly, the spinal neurulation in the mouse model, is both scientifically and clinically significant.</p><p>The neural tube arises from the neural plate flanked by the non-neural surface ectoderm. Numerous studies focused on the apical constriction of the neuroepithelial cells and neural plate bending [15–17], while little is known about the role of the non-neural surface ectodermal cells in neural tube closure. It has been long known that cellular protrusions (referred as ruffles or lamellipodia) present at the dorsal tips of mouse neural folds [18,19]. A recent study shows that these cellular protrusions are regulated by small GTPase Rac1 and may be extended from the surface ectodermal cells during mouse spinal neurulation [20]. Nevertheless, it remains unknown how these cellular protrusions are generated and whether these cellular protrusions and related surface ectodermal cells play an active role in neural tube closure.</p><p>In this study, we found unique cellular structures of the non-neural surface ectodermal cells that form multicellular rosette-like structures and convergent F-actin protrusions and cable network encircling the closure site during posterior neural tube closure, which are severely disrupted in the mutant mouse models with fully penetrant spina bifida. These results provide evidence for a novel neural tube closure model driven by the non-neural surface ectodermal cell dynamics, which is computationally visualized.</p><!><p>The Grhl3Cre knock-in mouse line [21] was acquired through the Mutant Mouse Resource & Research Centers (MMRRC) at UC Davis. Rosa26-mT/mG reporter mice [22] were acquired through the Jackson Laboratory. All research procedures using laboratory mice were approved by UC Davis Animal Care and Use Committee and conform to NIH guidelines. Pregnant, timed mated mice were euthanized prior to cesarean section. The noon of the conception day was designated as E0.5.</p><!><p>SEM was carried out as previously described [23]. Briefly, mouse embryos were fixed in 2.5% glutaraldehyde and 2% paraformaldehyde (PFA) in PBS and dehydrated in a graded ethanol (200 proof) series. Hexamethyldisilazane (HMDS) was used for the final drying of samples. The treated embryos were mounted on aluminum stubs and then sputter coated with gold (Pelco Auto Sputter Coater, Ted Pella) for scanning electron microscopy (Phillips XL30 TMP, F.E.I. Company, Hillsboro, OR, USA).</p><!><p>Mouse embryos were dissected into ice-cold PBS and fixed in 4% PFA in PBS overnight in the cold room. Wholemount immunofluorescence was performed according to standard protocols with slightly modifications. Briefly, fixed embryos were washed three times for 30 minutes each time in PBS with 0.1% TritonX-100, blocked in PBS with 0.1% TritonX-100 and 5% donkey serum for 2 hours, incubated in diluted primary antibodies in PBS with 0.1% TritonX-100 and 1% donkey serum overnight in cold room with gentle rocking, washed three times for 1 hour each time in PBS with 0.1% TritonX-100, incubated in diluted secondary antibodies in PBS 0.1% TritonX-100 and 1% donkey serum for 4 hours, washed three times for 30 minutes each time in PBS with 0.1% TritonX-100 by gentle rocking, counter-stained in PBS with 1μg/ml DAPI for 20 minutes with gentle rocking. The antibodies and dilutions were as following: Rabbit anti ß-tubulin (1:100, Cell Signaling), Rat anti-E-cadherin ((1:300, EMD Millipore), and Alexa 488 donkey anti-rabbit or rat IgG (1:200, Invitrogen). Alexa Fluor Phalloidin (1:100, Invitrogen) was used for F-actin staining. The stained embryos were transferred to a Petri dish with 4% agarose (for positioning embryos) filled with PBS for confocal imaging (Nikon A1 confocal laser microscope with NIS-Elements C software).</p><!><p>Using our scanning electron microscopy and immunofluorescence confocal micrographs as references, a dynamic closure model of the mouse caudal neural tube was created by a digital sculpting software Pixologic ZBrush. This 3D model was imported into a computer animation & modeling software Autodesk Maya and a custom animation rig was developed to visualize the dynamic closure process at single-cell resolution. In addition, convergent F-actin protrusions and F-actin cables were modeled and rigged separately in Maya. This dynamic model was rendered into a series of images from Maya and imported into Adobe After Effects for compositing and creation of the final video.</p><!><p>Using scanning electron microscopy (SEM), we found a large rosette-shaped multicellular structure that is formed by surface ectodermal cells and is converged with clustered cellular protrusions at the leading midline fusion site of the mouse posterior neural pore (PNP) around embryonic day (E) 9.5 (around 24-somite stage (ss24)) (Fig. 1A and E). This large midline rosette has partial opening with the pending closure gap towards the posterior direction. Around ss28 (Fig. 1B and F), two large midline rosettes with convergent protrusions are found at both anterior and posterior leading fusion sites, consisting of 10 to 14 visible cells (no. 1 & no. 7 in Fig. 1F). Between these two midline rosettes, there are 5 paired partial rosettes/protrusions (no. 2 to no. 6 in Fig. 1F) located at the opposing sides of the pending closure gap of the PNP. Along the pending closure edge, one or two visible and elongated cells contribute to two adjacent rosettes and protrusions. Around ss32 (Fig. 1C and G), the last two midline rosettes/protrusions (labeled no. 3 and no. 4 in Fig. 1G, presumably equivalent to the same numbers for the regions with the largest pending closing gap in Fig. 1F) merged through a shortened surface ectodermal cell on each side of the final midline epithelial fusion site. These SEM observations strongly suggest that each rosette-forming cell generates cellular protrusions that are intermingled and converged together to mediate surface ectodermal cell fusion at the leading midline closure site, to seal up the pending closure gap, and to complete the final midline closure process. These observations indicate that the surface ectodermal rosette formation and convergent cellular protrusions may play a previously unrecognized active role in neural tube closure.</p><!><p>To demonstrate this speculation, we examined these unique cellular processes in the surface ectodermal mutant mice that exhibit fully penetrant spinal NTDs. Indeed, the surface ectodermal rosette formation and convergent protrusions are severely disrupted in the mutant PNPs of Grhl3Cre/Cre (Grhl3-KO) [21] mouse embryos at E9.5 (Fig. 2). The grainyhead-like transcription factor Grhl3 is predominantly expressed in the surface ectodermal cells and is required for epidermal development and neural tube closure (particularly, spina bifida) [24–27]. In addition to the missing midline rosette and protrusions at the fusion site, small rosettes centered with visible protrusions are abnormally located away from the closure edge in the Grhl3-KO PNPs (Fig. 2A and B). Together, these results demonstrate that a local signaling cascade mediated by transcription factor Grhl3 is required for the surface ectodermal rosette formation and convergent cellular protrusions to promote PNP closure.</p><!><p>To further demonstrate the role of surface ectodermal cells and the nature of the cellular protrusions in neural tube closure, we conducted genetic fate mapping by crossing Grhl3Cre/+ mice with the Cre reporter RosamT/mG mice [22], in combination with F-actin labeling by fluorescence-conjugated phalloidin (Fig. 3). We found that the rosette-forming cells are Grhl3-expressing lineage cells that wrap the opposing dorsal tips of the pending closing PNP (Fig. 3A and A1). These cells generate the F-actin-based cellular protrusions (Fig. 3B and B1), evident by the complete overlapping of the fate-mapped membrane EGFP and F-actin labeling (Fig. 3C and C1). At the leading closure site, the EGFP/F-actin colabeled cellular protrusions merge at the midline connecting the opposing tips of the dorsal PNP (Fig. 3A2–C2). At the newly closed dorsal midline, the EGFP-labeled cells attached or fused to each other, bridging the dorsal neuroepithelial cells on each side of the PNP (Fig. 3A3–C3). Notably, the rosette-forming cells also generate a large and tangentially oriented F-actin cable network with a cable running anteriorly from the leading fusing site along the recently closed midline, and two cables separated from the leading fusion site running posteriorly along the edges of the pending closing gaps (Fig. 3B,B3,C,C3). The posteriorly running F-actin cables have recently been shown to form a cable ring connecting cellular protrusions around the pending closure site and to play a role in biomechanical coupling for PNP closure at late somite stages [28]. The surface ectodermal nature of these rosette-forming cells wrapping along the dorsal edges of the PNP is also demonstrated by the immunolabeling of ß-tubulin or the surface ectoderm adherens junction marker E-cadherin (Fig. 4A and B). Although E-cadherin immunolabeling also shows some signals in the neuroepithelial cells, it has been used to support that the non-neural surface ectoderm wraps around neural ectoderm right before the neural folds meet during mouse cranial neural tube closure [29]. Our genetic fate mapping and immunolabeling results provide conclusive evidence for the non-neural surface ectodermal nature of the rosette-forming cells (Fig. 4C) that wrap around the pending closure tips and fuse at the dorsal midline by merging the convergent cellular protrusions during posterior neural tube closure.</p><!><p>Based our SEM, genetic fate mapping, and immunolabeling results, we propose a dynamic model of PNP closure driven by surface ectodermal rosette formation and F-actin protrusion/cable network (Fig. 4D & Movie 1). Each rosette-forming cell generates polarized F-actin protrusion that converges at the center of the rosette to form a large cluster of F-actin protrusions. There are one or two elongated cells running along the F-actin cable. Both ends of these cells generate F-actin protrusions connecting two adjacent rosettes. Thus, the convergent protrusions of the adjacent rosettes generate the force to tighten the F-actin cables between them and to pull the paired rosettes toward the midline. When the first pair (next to the leading fusion site) of rosette protrusions meet and fuse at the midline, a transiently existing full rosette will form at the midline, resulting in one-cell length closure. Sequentially, it will start the next full rosette formation when the next paired rosettes meet at the midline, and so on, advancing the PNP closure (Fig. 4D & Movie 1).</p><!><p>The current study revealed novel mechanisms of mouse neural tube closure driven by the non-neural surface ectoderm cells. Neural tube closure involves dynamic morphogenetic movements, such as convergent extension, neural plate bending, and apical constriction of neuroepithelial cells. However, the role of the non-neural surface ectodermal cells in neural tube closure remained poorly understood until recently. Meanwhile, the dynamics of mammalian neural tube closure has been understudied due to the challenge of live imaging experiments for in utero development [30]. The Niswander group has pioneered ex vivo live imaging of mouse neurulation and revealed several important findings in cranial neural tube closure, including the filopodia- and lamellipodia-like cellular projections from the non-neural ectoderm and membrane shuttling during cranial neural tube closure [29,31,32]. These findings suggest an active role of the non-neural surface ectodermal cells in mammalian neural tube closure. However, it remains unknown how these cellular projections are generated and whether they are essential for cranial neural tube closure. It also remains unclear if similar cellular dynamics may occur during spinal neurulation, which should be addressed in future live imaging studies.</p><p>Our SEM, genetic fate-mapping, and gene-targeting results clearly demonstrate that the non-neural surface ectodermal cells form rosette-like multicellular structures that generate the F-actin-based convergent protrusions for dorsal midline fusion during spinal neurulation, which are disrupted in the surface ectodermal mutants of Grhl3-KO mice that exhibit fully penetrant spina bifida [24]. The surface ectodermal role of Grhl3 in PNP closure is further demonstrated by its expression timing and conditional ablation of Grhl3 in the neuroepithelium using Nkx1–2Cre, which generated no spina bifida [33]. Together, these results demonstrate that Grhl3 is required for non-neural surface ectodermal rosette formation and convergent cellular protrusions to fuse the dorsal midline during PNP closure. Intriguingly, conditional ablation of Grhl3 in the hindgut using Sox17Cre generated partial penetrant spina bifida, supporting the hypothesis that diminished Grhl3 expression in the hindgut maybe the cause of spina bifida in the curly tail mice with a hypomorphic Grhl3 allele [33]. Increased expression of Grhl3 can rescue spina bifida in the curly tail mice [27], but transgenic overexpression of Grhl3 also causes spina bifida [34]. It would be important to examine whether non-neural ectodermal rosette formation and F-actin dynamics are intact or disrupted in Grhl3fl/fl;Sox17Cre (hindgut ablation), curly tail (hypomorphic Grhl3), or homozygous TgGrhl3 (gain-of-function) mouse embryos during spinal neurulation.</p><p>We have previously demonstrated that canonical Wnt/beta-catenin signaling regulates transcription factors Pax3 and Cdx2 to promote spinal neural tube closure in mice [35]. Nevertheless, it remains unknown if Pax3 and Cdx2 also play a role in the surface ectodermal dynamics during neural tube closure. Grhl3 has been suggested to act downstream of canonical Wnt signaling to specify the neural plate border cells during mouse neural tube closure [36]. Therefore, it is highly possible that Wnt/beta-catenin signaling also regulates Grhl3 to promote surface ectodermal morphogenesis and F-actin dynamics for neural tube closure. Significantly, the Drosophila homologue Grainyhead (grh) is required for dorsal closure that is an analog event of neural tube closure and wound healing in vertebrates [37], suggesting an evolutionally conserved role of Grhl transcription factors in epithelial closure processes, with potentially varied cellular mechanisms among different anatomical regions and species.</p><!><p>Movie 1. Dynamic visual modeling of the stepwise closure process of the mouse spinal neural fold driven by surface ectodermal rosette formation (light green) and the convergent F-actin protrusion and cable network (red). This is a dorsal view of the PNP that closes towards the caudal (bottom) direction.</p>

PubMed Author Manuscript

Walker, et al. -Proton Transfer in HPTS -Page 1 Proton Transfer from a Photoacid to Water: First Principles Simulations and Fast Fluorescence Spectroscopy

Proton transfer reactions are ubiquitous in chemistry, especially in aqueous solutions. We investigate photo-induced proton transfer between the photoacid 8-hydroxypyrene-1,3,6trisulfonate (HPTS) and water using fast fluorescence spectroscopy and ab initio molecular dynamics simulations. Photo-excitation causes rapid proton release from the HPTS hydroxyl. Previous experiments on HPTS/water described the progress from photoexcitation to proton diffusion using kinetic equations with two time constants. The shortest time constant has been interpreted as protonated and photoexcited HPTS evolving into an "associated" state, where the proton is "shared" between the HPTS hydroxyl and an originally hydrogen bonded water. The longer time constant has been interpreted as indicating evolution to a "solvent separated" state where the shared proton undergoes long distance diffusion. In this work, we refine the previous experimental results using very pure HPTS. We then use excited state ab initio molecular dynamics to elucidate the detailed molecular mechanism of aqueous excited state proton transfer in HPTS. We find that the initial excitation results in rapid rearrangement of water, forming a strong hydrogen bonded network (a "water wire") around HPTS. HPTS then deprotonates in ≤3 ps, resulting in a proton that migrates back and forth along the wire before localizing on a single water molecule. We find a near linear relationship between emission wavelength and proton-HPTS distance over the simulated time scale, suggesting that emission wavelength can be used as a ruler for proton distance. Our simulations reveal that the "associated" state corresponds to a water wire with a mobile proton and that the diffusion of the proton away from this water wire (to a generalized "solvent-separated" state) corresponds to the longest experimental time constant.

walker,_et_al._-proton_transfer_in_hpts_-page_1_proton_transfer_from_a_photoacid_to_water:_first_pri

I. Introduction<!>II. Experimental Procedures<!>III. Experimental Results and Discussions<!>IV. Ab Initio Molecular Dynamics Simulations Methodology<!>V. Ab Initio Microsolvated Geometry Optimizations<!>VI. Collective Coordinate Methods<!>Dynamics<!>Optimizations<!>VI. Concluding Remarks<!>Supporting Information

<p>Proton transfer reactions in aqueous solutions are important in many physical, chemical, and biological processes, including water oxidation, 1 tautomerization of bases in DNA, 2 ATP activities in living cells, 3 and proton diffusion in water. [4][5] Water's ability to form complex hydrogen bonding networks, particularly in the presence of ions, gives rise to various suggested mechanisms for accelerated proton transfer, including stepwise hopping and collective deprotonation within a water wire. [6][7][8] A useful approach to study such processes is to monitor the photo-induced proton transfer between an excited photoacid and water. [9][10] A photoacid is a particular type of molecule that becomes a stronger acid after promotion to an electronic excited state. A common photoacid is 8-hydroxypyrene-1,3,6-trisulfonate (pyranine, or HPTS), which has a pKa near 7 in the ground state but drops to ~0 upon photoexcitation. 11 The proton transfer between HPTS and water has been extensively studied with time resolved fluorescence, transient IR/visible absorption, and stimulated Raman spectroscopies. [12][13][14][15][16][17][18][19] Upon the photo-excitation of an HPTS molecule in water, several fast and slow processes have been observed by transient absorption and time resolved fluorescence spectroscopies. 13,[18][19][20][21] The fastest process is the Stokes shift, which takes place between a few hundred femtoseconds and a picosecond. 13,17 This Stokes shift is not directly related to proton transfer. Following the Stokes shift, the observed excited state dynamics have been analyzed using kinetic equations. 13 Three spectroscopically identifiable excited HPTS states have been discussed in relation to the kinetic model. The first proposed state is the "protonated state." This is the short-lived state that is initially formed after excitation. The HPTS is excited but the system of HPTS and water still essentially has its ground state configuration, including the hydroxyl proton. The hydroxyl proton in the ground state is assumed to be hydrogen bonded to an oxygen of a water molecule. The second proposed state is referred to as the "associated state." In this state the proton is assumed to have Walker, et al. -Proton Transfer in HPTS -Page 3 moved some distance from the hydroxyl oxygen towards the oxygen of the hydrogen bonded water molecule, i.e. the proton is shared between HPTS and the water molecule. The time constant of the transition from protonated state to associated state is ~3 ps. The third proposed state is the deprotonated or solvent separated state. In this state, a water molecule is hydrogen bonded to the Oof deprotonated HPTS. The state of the proton is not defined other than it is no longer shared between the HPTS and the oxygen of the initially hydrogen bonded water. The time constant of the transition from the associated to deprotonated state is ~90 ps. In this work, we present computational evidence demonstrating that the proposed associated state is more likely a deprotonation event where the proton remains in a water wire emanating from the HPTS molecule.</p><p>Different states of HPTS display different fluorescence spectra, where the peak maximum can be shifted by up to 80 nm. 21 In addition to these relatively short times, there is a much slower process associated with long range proton diffusion followed by proton recombination, which reforms the associated state. The recombination occurs over nanoseconds to tens of nanoseconds.</p><p>By repopulating the associated state to a small extent, the fluorescence intensity of the "associated" peak decays non-exponentially with an ~90 ps time constant. 12,21 It has been shown theoretically and experimentally that the recombination causes the final complete decay of the associated state to occur as a power law, t -α , where α = ~1.5. [22][23] This process has been discussed in detail using a Smoluchowski type diffusion-reaction theory. 12 Bulk water forms and reforms hydrogen bonds continually, with fast hydrogen bond randomization on the order of 2 ps in bulk water. 24 This can change at interfaces and around solutes, which can organize the water differently than bulk water. In the case of reorganizing around an excited dye, such as HPTS, water generally rearranges on the order of a single picosecond. 6 A number of theoretical studies have shown that a positive ion in water can have varying rates of diffusion based on the relative orientation of waters and solutes, and complex collective motions through tightly bound water wires. [6][7] Rates of proton diffusion are correlated to this relative organization, with various outcomes depending on the orientations of the water molecules through the bulk. 25 The focus of this paper is to understand the nature of the associated state and what deprotonation (solvent separation) means in terms of the experimental spectral observables on the quantum molecular level. To this end we have performed ab initio molecular dynamics simulations to study the fastest dynamical events. The simulations calculate the electronic excited state energies and dynamic changes associated with the different stages of deprotonation. These are compared to the results of time dependent fast fluorescence experiments monitoring the fluorescence spectra of HPTS in aqueous solution following photo-excitation. The experiments use time correlated single photon counting, as has been described previously. [26][27] The emphasis of the experiments presented here is to obtain very accurate time dependent results. In past experiments, the spectra have been somewhat contaminated by emission from a fluorescent impurity that is present in commercial HPTS. The impurity fluorescence spectrum needed to be characterized and removed from the HPTS spectra, which is difficult as its molecular identity is unknown. It needed to be modeled and subtracted. Here we avoid this step by using HPTS synthesized by Dr. Ernst Koller using a new procedure that does not produce the impurity.</p><p>For the first time, the ab initio molecular dynamics simulations (including as many as 100 explicit quantum mechanical water molecules capable of proton transfer) address the details of the proton dynamics as a proton leaves the HPTS in pure water. The initial excitation results in immediate (within 100 fs) strengthening (or in some cases formation) of a hydrogen bond between the HPTS hydroxyl group and the nearest water molecule. This is quickly followed (200-2000 fs) by water rearrangement to form a water wire. Previous theoretical work on excited HPTS in both aqueous 28 and aqueous/acetate 29 solution also observed an initial tightening of the hydrogen bond between HPTS and a water molecule coordinating to the HPTS hydroxyl group. However, the formation of a water wire was not reported, likely because most or all of the surrounding water molecules were treated with an empirical force field that restricted proton transfer. The water wire network we observe generally comprises 3-6 waters, and extends from the HPTS hydroxyl group to the nearest sulfate group that it is pointing towards. Within 3 ps, HPTS deprotonates and the proton rapidly moves along the water wire. The overall dynamics of the proton transfer are similar to previously reported theoretical work for protons transferring through bulk water-the movement of the proton through the wire is the result of a complex collective motion with rest and burst phases, and varying degrees of proton sharing throughout the wire prior to the burst deprotonation event. 7 After this burst of activity, the proton localizes on one water molecule. The emission wavelength in these stages is linear in the proton-HPTS distance, suggesting that the spectral evolution occurring on a 3 ps timescale (after an initial ultrafast Stokes shift) corresponds to deprotonation of HPTS. Geometry optimizations show that the final experimental spectral shift is within 0.02 eV of the shift computed with large proton solvent separation, indicating that the 78 ps experimental spectral shift is due to diffusion of the proton away from HPTS, rather than a single deprotonation event.</p><!><p>The HPTS (8-hydroxypyrene-1,3,6-trisulfonate) used in the experiments was provided by Dr. Ernst Koller at Lambda Probes & Diagnostics. Commercially available HPTS is contaminated by an unknown fluorescent impurity that causes problems in the analysis of the experimental data.</p><p>The samples used here are free from contamination. MPTS (8-methoxypyrene-1,3,6-trisulfonate) was purchased from Sigma Aldrich. All chemicals were used as received. In the experiments, 1 mm path length cuvettes were filled with H2O or D2O solutions of 10 -4 M HPTS or MPTS for the fluorescence measurements. Time resolved fluorescence decays were measured using a time-correlated single photon counting (TCSPC) setup described previously. 30 Approximately 100 fs laser pulses centered at 730 nm were generated by a Ti:Sapphire oscillator (Spectra Physics MaiTai). An acousto-optic single pulse selector was used to select pulses from the 80 MHz laser train at 5 MHz. The selected pulses were doubled in a barium borate crystal, yielding 365 nm light to excite HPTS and MPTS. The samples were excited and measured from the front surface in a near-normal geometry. Excitation occurred through a hole in a large lens, which was used to collect and collimate the emitted fluorescence. Another lens focuses the fluorescence on into the monochromator slit. The emission was detected at the magic angle polarization. The frequency of the fluorescence was resolved by the monochromator, and the single photons were detected by a multichannel plate (MCP) detector.</p><p>The fluorescence decays were collected from 400 nm to 600 nm in 2 nm increments by stepping the monochromator. The wavelength scan was repeated multiple times. Entire fluorescence decays were obtained at each wavelength, and time dependent spectra were assembled from the time decays at each wavelength. The instrument response function (IRF) was collected using a weakly scattering suspension in water in the same cuvette used to measure the time dependent fluorescence.</p><!><p>Figure 1 displays time dependent fluorescence spectra of excited HPTS in H2O and D2O from 30 ps to >25 ns. The data were taken as described above. A decay curve was acquired at each wavelength (2 nm steps), and then the intensities at each wavelength at a particular time were plotted. Figure 1A shows the data for HPTS in H2O; Figure 1B shows HPTS in D2O; and Figure 1C shows MPTS data in H2O. The initial HPTS excited protonated state decays into the associated state in 3.5 ps, 13 which is faster than the 30ps (with deconvolution) time resolution of the TCSPC system. In Figures 1A and B, the states labeled "associated" and "deprotonated" are at higher/lower frequency, respectively. The curves are for the times given in Figure 1A. In both H2O and D2O, the associated peak decays and the deprotonated peak grows in. The process is considerably slower in D2O than in H2O. At very long time, the associated state maintains some amplitude. As discussed in the introduction, following deprotonation, proton (hydronium) diffusion repopulates the associated state to a small extent. The decay of the associated state into the deprotonated state is exponential followed by a low amplitude power law that arises from the diffusive repopulation of the associated state. [12][13] Figure 1C shows the time dependent spectrum of MPTS. MPTS is identical to HPTS except that the HPTS hydroxyl is replaced by a methoxy. Therefore, it does not deprotonate in the excited state. The spectra were taken under identical conditions as the HPTS data. The lifetime decay of the amplitude was removed from the data so that it can be seen that the band is independent of time because there is no proton dissociation. The peak position is at higher frequency than the associated peak, and is more akin to the initial protonated state of the HPTS.</p><p>Figure 2 shows several of the HPTS spectra from Figure 1 with fits to the band shapes. At long time (>5 ns) the spectrum is essentially only the deprotonated band. The long time deprotonated band shape was fixed and only its amplitude was allowed to vary. The associated band shape and amplitude were allowed to vary, but its shape was held constant over time. In each panel, the black curve is the experimental spectrum. The red curve is the fit. It overlaps so well with the black curve that the black curve is substantially obscured. The blue and green curves are the band shapes of the associated and protonated states. The peak positions of the bands are: H2O associated, 446 ± 2 nm; D2O associated, 444 ± 2 nm; H2O deprotonated, 510 ± 2 nm; and D2O deprotonated, 512 ± 2 nm. These types of fits allow the areas of the bands to be obtained as a function of time. The total area, i. e., the sum of the areas of the two bands, decays with the fluorescence lifetime, 4.55 ns. 12 Figure 3 displays the decay of the associated state for HPTS in H2O (A) and in D2O (B).</p><p>The data are the time dependent areas of the associated state divided by the total area of the total spectrum (associated plus deprotonated bands) at each time. Dividing by the total area removes the lifetime decay from the data. Therefore, the data represents only the decay of the associated state into the deprotonated state. The decay is an exponential at short time and a power law at long time. As the deprotonated state is formed, the protons diffuse away from the HPTS. A small fraction returns to HPTS to recreate the associated state, producing a small associated state population for times very long compared to the exponential decay of the associated state into the deprotonated state. 23 The reformed associated state again decays into the deprotonated state, and the proton diffuses away. As time increases, protons diffuse out of the vicinity of HPTS, and fewer recombine. This process leads to the power law decay. 22 At very long time, the power, t -α , will asymptotically approach α = 1.5. [22][23] At short time, theory indicates that α = 1.4, which has been observed experimentally. 12,21 The data in Figure 3 were fit in the following manner. The functional form starts as an exponential and then makes a smooth transition into the power law, which grows in as the exponential decays. The curve is not the sum of an exponential and a power law as recombination, which gives rise to the power law, cannot occur until the deprotonated state becomes populated.</p><p>In this work, we are primarily interested in the short time dynamics where the exponential behavior dominates. This is in contrast to prior experiments where the power law was the focus and data were taken to much longer time. 12,21,23 As the power exponent was found to be 1.4 from both experiment and theory, in the fits, α was fixed at 1.4. The exponential decay constants are 78 ± 2 ps and 219 ± 2 ps in H2O and D2O, respectively.</p><p>In the following sections, ab initio molecular dynamics simulations are applied to examine the HPTS proton transfer dynamics. The initial processes immediately following excitation, the nature of the associated state, and the final proton dissociation are explicated. Although the simulations agree with the observed time constants, the molecular picture of the structures involved in the "associated" and "deprotonated" states are found to be more complex than the simple picture inferred from experiments and described above.</p><!><p>In order to investigate this system, the computational work comprised several different steps, following our previous work. 30 We first performed classical molecular dynamics simulations on the ground state and extracted equilibrated structures. These were then used to investigate the excited state surface using quantum mechanics/molecular mechanics (QM/MM) Born-Oppenheimer molecular dynamics (BOMD). From these trajectories, we extracted microsolvated clusters to compute stationary points using QM methods, and computed barriers along minimum energy paths connecting these structures using the improved dimer method, 31 as described in the SI.</p><p>The HPTS force field parameters used in ground state equilibration were taken from previous work. 30 In brief, the partial charges were obtained with RESP fitting, 32 and most of the geometric parameters were taken from the general Amber force field (GAFF). 33 The sulfate group S-O parameters for HPTS were taken from the literature. 34 The water model used was TIP3P, with relevant ion parameters for Na + . 35 The HPTS parameters used are provided in the SI. The classical empirical force field trajectory had a single HPTS molecule coordinated by 3 Na + ions and solvated in a water box of 4277 waters (50 Å) using periodic boundary conditions.</p><p>An initial structure of solvated HPTS was created with tleap. The structure was minimized, equilibrated with NPT at 1 atm/10 K for 40 ps, slowly heated to 300 K over 100 ps in NVT, and then equilibrated for 40 ps (1 atm/300 K/NPT). This was followed by 100 ns of production NPT simulation at 1 atm and 300 K. These were run in duplicate, using pmemd.cuda from Amber. [36][37] A Langevin thermostat with 5 ps -1 friction parameter was used to maintain the temperature. 38 Snapshots from the ground state dynamics were clustered according to the hydroxyl O-H bond distance and the distance from the H atom of the HPTS hydroxyl to the O atom of the closest water. Outlying waters were removed to create a sphere of ~18 Å (1000 waters), to be used for excited state QM/MM AIMD. Several QM regions were tested-4 waters were insufficient to see deprotonation in dynamics, while two solvation shells (~100 waters) around the HPTS molecule was prohibitively expensive for the desired 3 ps run time. The final QM region chosen included two solvation shells around the hydroxyl moiety, spanning the distance between the OH and the closest HPTS sulfate which the OH pointed toward (~30 waters). This was sufficient to see deprotonation within 3 ps or less of BOMD dynamics, and enough to converge the charge description of the HPTS hydroxyl moiety as described below.</p><p>All electronic structure calculations were performed with TeraChem, [39][40] interfaced with OpenMM 41 to describe the MM region. For geometry optimizations, we used the ωPBE method with range separation parameter ω=0.35 and the 6-31g** basis set, as in previous work. 30 The range separated functional avoids spurious low-lying charge transfer excited states (which would otherwise be ubiquitous because of the presence of the Na cations). Polarizable continuum implicit Walker, et al. -Proton Transfer in HPTS -Page 11 solvation [42][43][44] was also used for a set of geometry optimizations and some of the transition state optimizations. For the QM/MM BOMD, an additional d3 dispersion correction was added. [45][46] QM/MM dynamics (NVE with a 1 fs time step) were performed on S1, with 5 initial conditions taken from the classical MD trajectory. An additional 6 th trajectory on S0 was run to ensure that no deprotonation occurs on the ground state within 4 ps. Each snapshot was cut into a sphere as described above and simulated using spherical boundary conditions. The classical trajectories were clustered based on the HPTS hydroxyl O-H distance and the equatorial/axial disposition of the OH group (as discussed in previous work, the axial conformer has the OH pointing along the short axis of the HPTS molecule and the equatorial conformer has the OH pointing along the long axis of HPTS). 30 The centroid snapshot for each cluster was extracted, and the radial distribution function (RDF) (water to the H atom of the HPTS hydroxyl group) was computed for each. Five snapshots from this pool were chosen to cover the range of O-H distance and O-H orientation (axial vs. equatorial). Each snapshot was run for 10 fs on S0 to minimize artifacts from changing the description from MM to QM/MM, and then run for 600 fs to 2.5 ps in the NVE ensemble on S1. Trajectories were stopped a short time after the proton localized to within 1 water of the QM/MM boundary for at least 100 fs. We track the proton through the wire using a collective coordinate previously described by König et al. 47 and originally suggested by Chakrabarti et al., 48 as discussed in detail below.</p><!><p>The ground state and excited state dynamics calculations are accompanied by structural optimizations of microsolvated cluster geometries in gas phase and implicit solvent. These optimizations start with geometries taken from the snapshots from the ground state QM/MM-MD trajectories. We found that 6-8 explicit QM water molecules are needed in order to stabilize the deprotonated structures. When these snapshots are minimized on S0, the retained water molecules do not rearrange significantly. However, minimization on S1 leads to a large-scale rearrangement of the waters to form a wire. This rearrangement is consistent for different starting structures.</p><p>Therefore, the reported S0 minimum is obtained from an optimization starting from one of these S1 structures, minimizing ground/excited-state differences in water rearrangement of the microsolvated systems. Individual sodium atoms were positioned to coordinate with the HPTS sulfate groups prior to minimization. Further details for the transition state optimization can be found in the SI.</p><!><p>Collective coordinates were computed by determining the proportion of proton localization as a function of distance, computing the coordination number of each water in the wire based on those proportions, and relating it back to the overall length of the water wire as shown in the literature. 47 To determine a coordination number, we first use the following switching function fsw(r):</p><p>which drops off quickly for distances greater than 1.4 Å. The excess proton coordination number for each water molecule w O i ( ) is then given by:</p><p>where r O i H j is the distance between the ith oxygen atom and the jth hydrogen atom, NH is the total number of hydrogen atoms in the system, and N H ref is the number of protons which are attached to the oxygen in its reference state (2 for water molecules and 0 for the HPTS OH group). With this definition, water molecules with no excess proton character have w = 0 and an idealized hydronium has w = 1. We compute the excess proton coordination number for each QM oxygen atom in the simulation after 1 ps in order to determine which water molecules are involved in the water wire (these have w>0). Once the water molecules in the wire are known, we compute the collective coordinate v as:</p><p>where r z O i is the sum of the O-O distances along the wire to the ith oxygen atom. This gives the location of the excess proton as a function of distance along the wire, while taking into account the overall expansion and contraction of the wire that can occur in the simulation. It is important to note that this is a distance along the wire; if the wire is curved this distance will be longer than a straight line drawn between the location of the proton and the HPTS hydroxyl moiety.</p><!><p>We extracted five structures from the ground state dynamics to explore the first excited state surface (S1) with BOMD. These five snapshots were chosen from cluster analysis as described above to cover variations in O-H distance and equatorial/axial disposition of the hydroxyl group.</p><p>A representative structure can be seen in Figure 4.</p><p>We investigated several sizes of the QM region to balance computational cost while retaining the ability to describe deprotonation and the initial stages of the proton shuttling. Single point energies were computed for a representative snapshot with progressively larger numbers of surrounding explicit QM water molecules. Figure S1 shows the evolution of the excitation energy and the Mulliken charges for the HPTS OH group as a function of increasing numbers of QM waters. We find that the OH charge stabilizes at around 17 waters, i.e., including all water molecules within 3 Å of the O atom in the HPTS hydroxyl group (measured by O-O distance). The excitation energy is converged to better than 0.1 eV after inclusion of 17 water molecules. We therefore used two solvation shells around the HPTS hydroxyl (30 QM water molecules) for subsequent dynamics calculations. Exploratory calculations revealed that including four QM water molecules surrounding the HPTS hydroxyl was insufficient to observe any deprotonation within 3 ps of simulation and also that the dynamics observed (up to 600 fs) when including 100 QM water molecules was very similar to that observed with 30 QM water molecules. Figure 5 shows the QM region of a representative trajectory at varying timescales, indicating a consistent area of QM waters remaining throughout the trajectory and reorganization within the QM region, without intrusion of MM water molecules. RDFs were computed for each trajectory to confirm this, showing that MM waters never get closer than 3.8 Å to the HPTS hydroxyl, most likely due to the comparatively short timescale of the simulations. (Figure S2).</p><p>In general, each dynamics trajectory follows a similar path (on varying timescales), depending on the specific initial snapshot selected. Upon excitation, the hydrogen bond between the closest water molecule and the HPTS hydroxyl group is formed and/or strengthened within ~100 fs (as indicated by shortening of the O-O distance). In cases where there is already a wellformed H-bond between HPTS and a water molecule on the ground state, the O-O distance decreases to 2.5 Å or less. The next 200-2000 fs, depending on the specific trajectory, involve a rapid and large rearrangement of the QM water (Figure 5). Water molecules up to 6 Å away from the hydroxyl group rearrange to form a water wire. This initial reorganization results in a small red shift in emission energy of less than 0.05 eV. Experiments observed an emission red shift of 0.2eV, occurring with a 3 ps time constant. Our results indicate that this shift is not due to the rearrangement/tightening of hydrogen bonds in the network, which occurs much faster than the observed 3 ps time constant. Therefore, we do not attribute this to an "associated" state geometry.</p><p>The form of the resulting water wires depends on the equatorial/axial disposition of the HPTS hydroxyl group. When the OH group is equatorial, the wire extends from the OH group to the sulfate which is closest along the long axis of HPTS (typically involving six water molecules), and when the OH group is axial, it extends from the OH group to the sulfate which is closest along the short axis (typically involving three water molecules). Water wire formation and organization from nearby charged functional groups has been observed in other simulations of excited state dyes and for proton transfer in organized environments, such as proteins, as discussed in the Introduction. Unlike the more randomized formation of these networks in bulk water, the negatively charged sulfate group appears to favor a consistent formation of this water wire across different trajectories. Once this wire forms, HPTS deprotonates and the proton is briefly shared within the network (Figure 6A/6C) before localizing on a specific water. A representative structure is shown in Figure 5B, where the proton localizes on a water molecule midway along the water wire (labeled "Water 2" in Figure 6). This general pattern is followed for all the trajectories tested that include deprotonation events (four out of five of the initial conditions led to deprotonation on the 3 ps time scale of the simulation). The binned emission energies over time for all five trajectories are given in Figure 6D, showing a Stokes shift of almost 0.2 eV by 2500 fs. We compute a collective coordinate as described above to relate these wavelength shifts to the location of the proton within the water wire, and further to its relative distance from the anionic oxygen of deprotonated HPTS (Figure 7).</p><p>For the same trajectory shown in Figure 6, we find that the proton has rest phases where it oscillates between the HPTS and the first water in the wire, and burst phases where the proton is highly delocalized and then localizes on one of the water molecules in the wire. This is quite similar to previous theoretical findings for ground state excess proton dynamics in bulk water. 7 In Figure 7A, the presence of a tightly bound third hydrogen on the relevant water is shown in yellow. Regions where two or more waters have considerable excess proton character (e.g. around 975 fs in Figure 7A) indicate an overall contraction of the water wire and strong sharing between the waters. Figure 7B shows the excess proton distance along the water wire over time, with an inset of the water wire geometries during the burst phase. This data suggests that if a traditional Zundel cation is involved, its lifetime is very short.</p><p>Another initial condition snapshot has preorganized waters and deprotonates within 100 fs.</p><p>This trajectory was run both with 30 and 100 QM water molecules, yielding qualitatively similar results, as mentioned above. Figure 8 shows the trajectory with 100 QM waters. Interestingly, despite the much shorter timescale, Figure 8A is quite similar in character to the burst phase in Figure 7A (compare the 600 fs shown in Figure 8A with the last 600 fs shown in Figure 7A). Such similarities are also observed when comparing Figures 8B and 8C to Figures 6A and 6B. This suggests that much of the variation in the relative timescales of deprotonation is due to latency in the water reorganization to form a water wire. Corroborating this is a particular trajectory (one out of the five that were modeled) that shows no deprotonation events, and also has little to no hydrogen bonding character that would assist with water wire formation (see SI for details).</p><p>The excited state natural orbitals for HPTS remain essentially unchanged over time and after deprotonation occurs (Figure 9), and do not involve surrounding water molecules to any significant degree at any stage. The excitation energies also do not drop below 2.7 eV in any trajectory. This indicates that the excited state character is confined to the chromophore, suggesting that our observations may be generalizable to ground state proton transfer.</p><p>Taken together, this is strong evidence that the initial shift in emission wavelength over the first 3 ps after photoexcitation does not arise from transition to an associated state where the proton is shared between the HPTS OH group and the nearest water molecule. Instead, it corresponds to localization of the excess proton on a water molecule located in the water wire, but not directly Hbonded to the HPTS OH group.</p><!><p>To examine the origin of the process observed experimentally with a 78 ps time constant, we turned to microsolvated geometry optimizations, since the timescale was too long to directly observe with QMMM BOMD. Geometry optimizations were performed on single HPTS molecules neutralized with sodium ions and with 6-8 waters in a microsolvated cluster. These optimizations, similar to the dynamics trajectories, demonstrate an interesting dependence of emission wavelength on distance the proton has traveled from the HPTS oxygen (Figure 10A). If the proton has localized 9.1 Å (about 3.5 waters) away, the emission wavelength redshifts by 0.30 eV. This is very similar to the redshift observed experimentally on the 3 ps timescale (0.28 eV). If the proton is moved one water further, the emission wavelength redshifts by 0.47 eV. Finally, if the proton is removed to an infinite distance away, the emission wavelength redshifts by 0.54 eV.</p><p>This supports the assignment of the deprotonated state (experimental emission redshift of 0.45 eV)</p><p>to proton diffusion away from the initial water wire.</p><p>We can compare these shifts to those obtained from the averaged dynamics trajectories and their collective coordinates for proton transfer, as discussed above (Figure 10B). The dynamics show strong qualitative similarity to the microsolvated optimizations, showing a redshift of ~0.25</p><p>eV when the proton is localized 7 Å away. Figure 10B shows a very nearly linear relationship between emission energy and proton distance, and therefore provides a "ruler" with which one can determine approximately how far the proton has traveled by its emission wavelength.</p><p>Walker, et al. -Proton Transfer in HPTS -Page 18</p><p>We investigated proton transfer between HPTS and its neighboring water with the most minimal possible rearrangement of waters in order to separate the question of whether the deprotonation process is barrierless from the mechanism of water reorganization. To this end, we optimized the corresponding transition state on S1. We included a water cluster consisting of eight H2O molecules and surrounding implicit solvent, as described in the SI. The potential energy along the corresponding intrinsic reaction coordinate (IRC) is shown in Figure 11. The reaction has a small potential energy barrier of 2.2 kcal/mol. Therefore, we can qualitatively confirm that the timescale for the initial deprotonation is far below the 78 ps timescale previously assigned to the deprotonated state.</p><p>Figure 11 furthermore shows that the vertical excitation energy shifts by -0.1 eV during the minimal deprotonation event. This supports the conclusion from the dynamics, that the shift for the "associated" structure is actually an initial short-range deprotonation event. It also implies that, rather than a completely stepwise mechanism where the proton hops one water at a time, it is more likely that the first step involves a proton hopping 2-3 waters away in a concerted motion over the course of 100-200 fs. This is further supported by the experimental redshift of 0.28 eV after 3 ps as compared to the theoretical redshift from dynamics and optimizations (0.15 eV/0.30 eV shift for proton moving two/three waters away from HPTS along the wire).</p><!><p>In this work, we presented the first QM/MM simulations of photoexcitation dynamics in the solvated photoacid HPTS including sufficient QM water molecules to capture the proton transfer events. The strongly negative sulfate groups provide anchoring points for the formation of to protons and protein residues to have a strong impact on the relative speed of the proton shuttling, with sub-mechanisms associated with faster and slower proton shuttles depending on the organization of solvent around the deprotonation site. [49][50][51][52] We have similar findings, with "preorganized" water leading to deprotonation within hundreds of femtoseconds, and less organized water leading to deprotonation on a 3 ps timescale. Interestingly, the SO3 − moieties on HPTS appear to perform a similar function to negatively charged protein residues close to deprotonation sites, assisting the water to form wires in a relatively predictable fashion. Simulations of small molecule dyes, similar to HPTS, also show that the speed of proton shuttling depends on water organization at the moment of excitation. 53 Furthermore, these same authors find that long water wires containing more than four water molecules are not very stable, and that concerted proton hopping mechanisms are usually limited to this range. 53 Our results align well with this previous work.</p><p>We find a strong correlation between the emission wavelength and the distance between the HPTS oxygen atom and the excess proton in the water wire. This nearly linear relationship could be used to track proton motion in the wire with the measured emission wavelength. That said, to be conclusive one would need to test more initial conditions than we have in this work.</p><p>Our work provides a clear molecular picture of the intermediates involved in the excited state proton transfer of aqueous HPTS. The "intermediate" time constant of 2-3 ps observed in numerous experiments, 13,[18][19] previously assigned to a "shared" proton between HPTS and a coordinating water molecule, has been revealed to instead correspond to deprotonation of the HPTS hydroxyl group and rattling of the excess proton within a water wire. We also showed that</p><!><p>The supporting information is available free of charge at XX. Details about QM/MM region selection, collective coordinate chosen for QM/MM dynamics, and transition state optimization methodology (PDF).</p><p>Optimized geometries for critical point structures and force field parameters for HPTS (ZIP).</p>

ChemRxiv

Pyroligneous Acids of Differently Pretreated Hybrid Aspen Biomass: Herbicide and Fungicide Performance

The pyroligneous acids (PAs) of woody biomass produced by torrefaction have pesticidal properties. Thus, PAs are potential alternatives to synthetic plant protection chemicals. Although woody biomass is a renewable feedstock, its use must be efficient. The efficiency of biomass utilization can be improved by applying a cascading use principle. This study is novel because we evaluate for the first time the pesticidal potential of PAs derived from the bark of hybrid aspen (Populus tremula L. × Populus tremuloides Michx.) and examine simultaneously how the production of the PAs can be interlinked with the cascade processing of hybrid aspen biomass. Hybrid aspen bark contains valuable extractives that can be separated before the hemicellulose is thermochemically converted into plant protection chemicals. We developed a cascade processing scheme, where these extractives were first extracted from the bark with hot water (HWE) or with hot water and alkaline alcohol (HWE+AAE) prior to their conversion into PAs by torrefaction. The herbicidal performance of PAs was tested using Brassica rapa as the test species, and the fungicidal performance was proven using Fusarium culmorum. The pesticidal activities were compared to those of the PAs of debarked wood and of commercial pesticides. According to the results, extractives can be separated from the bark without overtly diminishing the weed and fungal growth inhibitor performance of the produced PAs. The HWE of the bark before its conversion into PAs appeared to have an enhancing effect on the herbicidal activity. In contrast, HWE+AAE lowered the growth inhibition performance of PAs against both the weeds and fungi. This study shows that hybrid aspen is a viable feedstock for the production of herbicidal and fungicidal active chemicals, and it is possible to utilize biomass according to the cascading use principle.

pyroligneous_acids_of_differently_pretreated_hybrid_aspen_biomass:_herbicide_and_fungicide_performan

Introduction<!><!>Introduction<!>Samples<!><!>Samples<!>Lipophilic and Hydrophilic Extractive Contents<!>Determination of Bark Suberinic Acid<!>Determination of Lignin Content<!>Determination of Cellulose Content<!>Determination of Hemicelluloses Content<!>Preparation of a Hot Water-Extracted Bark Sample for Torrefaction<!>Preparation of the Alkaline Alcohol-Extracted Bark Sample for Torrefaction<!>Torrefaction of Wood, Bark, and Pre-Extracted Bark Biomass<!>Pyroligneous Acid Titratable Acidity<!>Water and Organic Matter Contents of Condensates<!>Compositional Analysis of Organic Matter<!>Herbicidal Activity Test<!>Fungicidal Activity Test<!>Statistical Analysis<!>Results and Discussion<!>Basic Composition of Hybrid Aspen Biomass Fractions<!><!>Basic Composition of Hybrid Aspen Biomass Fractions<!><!>Torrefaction Product Yields<!><!>Torrefaction Product Yields<!><!>Torrefaction Product Yields<!>Pyroligneous Acid Composition<!><!>Pyroligneous Acid Composition<!><!>Herbicidal Activity<!><!>Fungicidal Activity<!><!>Current State of Pyroligneous Biopesticide Usage and Research<!>Conclusion

<p>One possible way to produce acidic chemical products from biomass is thermochemical conversion, e.g., torrefaction (Fagernäs et al., 2015; Grewal et al., 2018). Typically, the technique is used as a biomass pretreatment process, such as for gasification or combustion (Van der Stelt et al., 2011; Cahyanti et al., 2020). Torrefaction is a technique similar to slow pyrolysis, where conversion is performed at a slow heating rate under anaerobic conditions, but is conducted at lower temperatures. When using lower process temperatures, the hemicelluloses of lignocellulosic biomass are the most essential precursors to produce pyroligneous acids (PAs) (Collard and Blin, 2014; Chen et al., 2019). PAs are known to have pesticidal activity (Oramahi and Yoshimura, 2013; Theapparat et al., 2015; Oramahi et al., 2018; Hagner et al., 2020); therefore, the method appears to be a viable way to produce bio-based pesticides as a substitute for synthetic chemicals (Tiilikkala et al., 2010). Wood material is a potential and renewable natural resource for the production of pesticide-active chemicals, but the sustainable use of forest resources must be taken into account (Cambero and Sowlati, 2014). This can be promoted by studying the use of raw materials according to the cascade principle and by investigating the effects of the processes on the chemical products.</p><p>In recent years, the utilization of biomass in chemical production has increased. Recovery techniques are mainly conversion processes, such as catalytic, biological, and thermochemical processes. So far, the sources of bio-based raw materials for biochemical production are largely those that compete with other sectors, such as the food industry. In the cultivation of forest as a feedstock source, such a competitive situation is more easily evaded (Cherubini, 2010). Currently, the high market potential for bio-based chemicals includes biofuels, biomaterials such as degradable plastics, and a variety of biochemical applications such as adhesives and packaging coatings (Cherubini, 2010; Hassan et al., 2019). The number of bio-based products on the market is clearly growing, but the utilization of by-products is still low (Cherubini, 2010). As the chemical industry aims to electrify processes with renewable energy, produce biochemicals by converting from forest biomass, and reduce the need for fossil-derived reagents in biochemical production (Toivanen et al., 2021), pyrolysis techniques such as torrefaction are essential methods to be explored. The method also has the advantage of being able to generate side streams for recovery (Figure 1).</p><!><p>Utilization of hybrid aspen biomasses in the production of biopesticides according to the cascade principle.</p><!><p>There are still numerous encounters with the enhanced use of forest biomass, such as availability, price, and sustainability (Toivanen et al., 2021). One possible solution could be to cultivate fast-growing tree species as a source of raw materials for chemical production. By cultivating clonal trees, the raw material can be tailored (Korkalo et al., 2020) to serve better the needs of the chemical industry, locate the biomass source closer to chemical production, and accelerate biomass production (Beuker et al., 2016).</p><p>At the same time, as the goal is to increase the utilization of forest biomass in the chemical industry, the adequacy and sustainable consumption of forest resources must be taken into account (Azapagic, 2014). This means more efficient and versatile use of raw materials, where cascade utilization plays a key role. The term "cascade utilization" describes an enhanced use of biomass that produces multiple products before the biomass ends up as a waste residue, energy production, or other end-products. There is no consistent description of the concept of the cascade principle, but the definition of the term is met if the utilization of biomass involves linear utilization, product side streams, recycling, and other activities that increase the life cycle of biomass or maximizes its value in use (Keegan et al., 2013; Sokka et al., 2015).</p><p>Cultivated hybrid aspen is an interesting raw material alternative for various bio-based conversion products due to its rapid biomass yield potential. In Finland, the tree species reaches its felling age in the first growth cycle in 20–25 years and in the second growth cycle in 15 years. In northern growth areas, the growth rate compared to other commercially important tree species can be 40% faster than birch and 50% faster than spruce. Thanks to its rapid growth, the biomass yield of cultured hybrid aspen can reach up to 20 m3 ha−1 year−1 in 25-year cycles (Beuker et al., 2016).</p><p>This study investigated the use of hybrid aspen (Populus tremula L. × Populus tremuloides Michx) as a feedstock and the cascade utilization of the biomass in the production of herbicidal and fungicidal active chemicals (Figure 1). Due to the low extractive content of the debarked wood, the wood biomass is led directly to thermochemical conversion. However, the bark is known to be rich in extractives (Korkalo et al., 2020), so the effects of bark raw material pre-extraction on the pesticidal activity of PAs were studied.</p><p>Hot water extraction (HWE) was chosen as the first pre-extraction step for bark cascade processing (Figure 1) since the method can be used to separate soluble hydrophilic extractives into value-added products (Rasi et al., 2019). Moreover, HWE at moderate temperature and normal atmospheric pressure is not expected to significantly affect bark hemicellulose contents. The soluble hydrophilic compounds consist of tannins and strongly antioxidant phenolic substances (Korkalo et al., 2020). These groups of compounds are of interest for several practical applications, e.g., tannins in rigid foam forming (Tondi and Pizzi, 2009; Varila et al., 2020) and phenolic compounds as antioxidants in various end uses (Reuter et al., 2010).</p><p>Alkaline alcohol extraction (AAE), which can be used to separate insoluble bark compounds such as suberinic acids, was chosen as the second pre-extraction step to be tested (Figure 1). Suberin-derived fatty acids have been found to have promising properties to act, e.g., as a coating material for paper, fabrics, and packaging (Korpinen et al., 2019). The pretreatment of the bark with AAE prior to conversion to the biopesticide has been considered, even though hemicellulose structures are known to be soluble under alkaline conditions (Stoklosa and Hodge, 2012). Since hemicelluloses are the essential precursors for PA formation, the effect of bark AAE pretreatment on the pesticidal activity of PAs was analyzed in this study.</p><p>The hypothesis of this study is that several valuable products can be separated from hybrid aspen biomass without losing all the potential of PAs to be suitable for herbicide and fungicide applications. The herbicidal activity of PAs produced from wood, bark, and pre-extracted bark was tested using Brassica rapa as the model plant and the fungicidal activity tested using F usarium culmorum. The activity tests examined how well the bio-based pesticides compete with the efficacy of commercial products and whether the biomass can be utilized in the production of biopesticides according to the cascade principle.</p><!><p>Six sample trees of hybrid aspen clones (P. tremula L. × P. tremuloides Michx.) were randomly selected from a designed field trial of the Natural Resources Institute Finland (Luke) near Lohja, Southern Finland (60°12′ N, 23°55′ E). For chemical analyses of hybrid aspen biomass, a sample disc was cut from each tree at breast height (1.3 m). Bark mass, including both inner and outer layers, was peeled from the wood, after which both separated sample fractions were freeze dried. Dried bark and wood samples were ground with a 0.75 mm milling blade (rotor mill PULVERISETTE; FRITSCH, Idar-Oberstein, Germany) and stored at −20°C until analysis. Chemical assays made from wood and bark were used to characterize the wood chemical composition of hybrid aspen biomass and to examine the differences between the separated biomass fractions. Chemical characterizations were determined by performing similar chemical analyses for both sample fractions, except for suberinic acids which were determined only from bark samples (Table 1).</p><!><p>Basic information on hybrid aspen tree samples, chemical composition analysis for the wood and bark biomass fractions, number of biomass samples prepared for torrefaction, and chemical characterization of pyroligneous acids</p><p>HWE, hot water extraction; AAE, alkaline alcohol extraction; GC-FID, gas chromatography with a flame ionization detector; FT-ICR, Fourier transform ion cyclotron resonance</p><p>The national list of approved basic forest reproductive material, kept by the Finnish Food Authority (Finnish Food Authority, 2021).</p><!><p>For the torrefaction and the bark cascade processing experiments, wood, bark, hot water-extracted bark (HWE bark), and hot water and alkaline alcohol-extracted bark (HWE+AAE bark) samples were prepared by pooling the biomass of six sample trees into representative bulk samples of each feedstock. Feedstock samples were prepared for the torrefaction experiments freshly after collection. For the pesticide (i.e., herbicide and fungicide) experiments, one PA sample was produced from each of the feedstocks. The collected PA samples were used for pesticide experiments and for acid strength analysis shortly after the conversion. Information on the hybrid aspen tree, sampling data, and the list of chemical analyses are shown in Table 1.</p><!><p>Lipophilic and hydrophilic extractives of wood (n = 6) and bark (n = 6) were determined from freeze-dried and finely ground samples using the accelerated solvent extraction (ASE-350, Dionex, Sunnyvale, CA, USA) method. Lipophilic compounds were extracted with hexane at 90°C three times for 15 min, followed by extraction of the hydrophilic compounds with 95% acetone (aq) at 100°C three times for 15 min. Each separated lipophilic and hydrophilic extractive solution was adjusted to a final volume of 50 ml with used extractants, after which samples of each solution were dried in a nitrogen evaporator at 40°C to a dry matter residue to determine the extractive yields. Extract-free wood and bark samples were used for the determination of hemicellulose, cellulose, lignin, and suberinic acid contents of the bark. The method was modified from Willför et al. (2003).</p><!><p>The suberinic acid content of the bark was determined from extractive-free samples (n = 6) with a modified method adapted from Krogell et al. (2012). Bark samples were weighed into seal tight test tubes, followed by the addition of 3% potassium hydroxide–ethanol solution (KOH, w/v). The bark samples were extracted for 2 h at 70°C. The solution containing suberinic acids was separated and collected by vacuum filtration. The bark residue was dried overnight at 105°C and saved for lignin assays. Suberinic acid solution samples were measured into sample tubes and diluted with water. Bromocresol green was added to the samples as a pH indicator, after which the sample solutions were acidified with 0.25 M aqueous sulfuric acid. After pH adjustment, an internal standard mix of C21:0 and betulinol in tert-butyl methyl ether (MTBE) was added to the samples. Suberinic acids, including the internal standard, were separated from aqueous solution using liquid–liquid extraction with the MTBE solvent. The liquid–liquid extraction was repeated a total of three times. Finally, the collected suberinic acid and the internal standard containing the MTBE solution was washed with water before drying in a nitrogen evaporator at 40°C. The dry residues of the samples were silylated by adding a reagent mix of pyridine–BSTFA–TMCS in a 1:4:1 ratio and allowed to react at 70°C for 45 min. The clear phase of the silylated samples was collected and the content of suberinic acids was quantitated against the internal standard of C21:0 using gas chromatography–mass spectrometry (GC-MS).</p><!><p>The lignin content of the hybrid aspen samples was determined for wood from the extractive-free samples (n = 6) and for the bark from samples previously prepared for the determination of suberinic acids (n = 6). The total lignin content of the samples was determined as the sum of acid-soluble and insoluble (Klason) lignin. The prepared bark and wood samples were dried at 105°C overnight and weighed in duplicate for analysis. Of sulfuric acid, 72% was added to the sample vessels and mixed thoroughly, after which the samples were incubated for 1 h at 30°C. After incubation, the acid content of the samples was diluted to 4% with water, and the solution was transferred into a glass vessel with a screw cap. The samples were then placed in an autoclave at 120–125°C for 1 h, after which they were removed from the autoclave and allowed to cool to room temperature. Acid-insoluble lignin was separated from the samples using a vacuum filter and finally dried overnight at 105°C for weighing. The acid-soluble lignin separated by vacuum filtration was analyzed spectrophotometrically at 240 nm. The lignin content was analyzed and calculated according to the laboratory analytical procedure (LAP) (Sluiter et al., 2012).</p><!><p>The cellulose content of bark (n = 6) and wood (n = 6) was determined from the extract-free samples with the acid hydrolysis method adapted from Krogell et al. (2012). Samples and cellulose standards were weighed into sealable glass tubes, followed by the addition of 72% aqueous sulfuric acid. The strong acid solution was allowed to take effect for 2 h, after which it was diluted with water and allowed to react for the next 4 h. The acid content of the samples was further reduced by the addition of water and the samples left to stand overnight at room temperature. The following day, the samples were placed in an autoclave for 1 h at 120–125°C. Bromocresol green was added as a pH indicator, and the acidity of the solutions was neutralized with barium carbonate. d-Sorbitol was added into the wood, bark, and standard samples as an internal standard, after which the samples were mixed thoroughly and centrifuged. The separated clear phase was collected and evaporated to dryness in a nitrogen evaporator at 40°C, and finally in a vacuum oven at 40°C. Samples were silylated by the addition of pyridine, hexamethyldisilazane (HMDS), and trimethylchlorosilane (TMCS) and left to react overnight. The next day, the samples were analyzed by gas chromatography with a flame ionization detector (GC-FID). The cellulose contents of the samples were calculated by means of an inner standard.</p><!><p>The acid methanolysis method for the determination of bark (n = 6) and wood (n = 6) hemicelluloses content was adapted from Sundberg et al. (1996). Extract free bark and wood samples, as well as standard samples (monosaccharide mix of arabinose (Ara), glucose (Glc), glucuronic acid (GlcA), galactose (Gal), galacturonic acid (GalA), mannose (Man), rhamnose (Rha), 4-O-methylglucuronic acid (4-O-Me-GlcA), and xylose (Xyl); 1.0 mg/ml of each) were measured in sealable pressure-resistant glass containers. Methanolysis reagent (2 M HCl in anhydrous methanol) was added to the samples and placed in a 100°C oven for 5 h. After cooling the samples, the acid content of the samples was neutralized with pyridine, and a mixture of sorbitol and resorcinol in methanol was added to the samples as an internal standard. The separated clear phase was collected from the samples and dried in a nitrogen evaporator at 40°C to dryness. Samples were silylated by the addition of pyridine, HMDS, and TMCS and left to react overnight at room temperature. The samples were then analyzed with GC-FID to identify hemicellulose sugar units and to calculate the concentrations by using internal standards to determine the hemicellulose content.</p><!><p>To prepare the HWE bark feedstock sample (n = 1), 2 kg of bark mass was pooled from six sample trees. For the extraction, the water-to-dry matter ratio was adjusted to 6.1:1 (v/m) by checking the dry matter (d.m.) content of the bark sample with a moisture analyzer (MLB 50-3N, KERN & Sohn GmbH, Balingen, Germany) before adding the required amount of water. Bark mass was extracted in a steel boiler at normal atmospheric pressure at 90°C for 2 h. After extraction, the HWE bark sample was divided into two: one half was saved for torrefaction and the other half saved for the following alkaline alcohol extraction (Figure 1).</p><!><p>The hot water and alkaline alcohol-extracted bark (HWE+AAE bark, n = 1) was prepared from the previously prepared HWE bark mass. The AAE method was adapted from Rižikovs et al. (2014) and Korpinen et al. (2019). The AAE extractant mixture was prepared from crystalline sodium hydroxide (NaOH), water, and IPA (2-propanol) (Merck KGaA, Darmstadt, Germany) in a ratio of 1,200 ml IPA:300 ml H2O:20 g NaOH:100 g HWE bark d.m. The amounts of NaOH, water, and IPA required for the extractant mixture were calculated based on the d.m. content of the HWE bark sample, which was checked with a moisture analyzer (MLB 50-3N, KERN & Sohn GmbH, Balingen, Germany). The extraction solvent was prepared by dissolving NaOH in water before mixing with IPA and finally adding with HWE bark into a glass laboratory flask. The sample was extracted at normal atmospheric pressure by heating the extraction vessel in a hot water bath at 85°C for 3 h. After extraction, the HWE+AAE bark sample was rinsed with water until the pH of the rinsing water decreased.</p><!><p>The biomass conversion process was carried out in a bench-scale slow pyrolysis batch reactor under anaerobic conditions. The oxygen in the system was removed by purging the pyrolysis vessel with nitrogen gas before heating. The pyrolysis equipment consisted of a temperature control unit, an indirectly heating furnace, a gas-tight pyrolysis vessel with an internal thermometer (TCC-K250-6.0-KY), a water-cooled condenser with a condensate collection vessel, and a gas flow meter (drum-type gas meter, TG1/5, RITTER, Bochum, Germany). Hybrid aspen wood, bark, HWE bark, and HWE+AAE bark feedstocks were dried at 37°C before being placed into the pyrolysis apparatus. The sample sizes and d.m. contents (in weight percent), respectively, of dried biomass were 440.1 g, 97.7 wt.% for wood; 440.0 g, 96.2 wt.% for bark; 440.0 g, 96.0 wt.% for HWE bark; and 231.0 g, 96.2 wt.% for HWE+AAE bark. The torrefaction heating steps were programmed to ramp up from the pre-drying phase to the target temperature of 280°C. The actual measured heating rate of the pyrolysis vessel was 1.7°C/min from an initial temperature range of 30–250°C, after which the heating rate was decreased before reaching the temperature of 279 ± 2°C. The level of the target temperature was maintained for 40–50 min. The yields of solids and PAs formed during torrefaction were weighed and the total mass of gas formed calculated by subtracting the total mass of the solid and PA products from the weights of the feedstock samples loaded. All product yields were determined as mass percentages (weight percent) of the amount of samples loaded into the pyrolysis vessel.</p><!><p>The titratable acid strength of the PAs was determined by titrating the solutions with 1 M aqueous NaOH solution and calculating the result as acetic acid equivalent. Titration pH change was measured with a pH meter (SevenExcellence, Mettler-Toledo, Columbus, OH, USA) and electrode (Mettler-Toledo InLab Expert Pro-2m-ISM). The acetic acid equivalence of the condensates was determined by plotting the change in pH with the consumption of NaOH, verifying the titration equivalence point with the first (ΔpH/ΔV) and second (ΔpH2/Δ2 V) derivatives, and finally calculating the titratable acid content of the solutions as acetic acid CH3COOH (% m/v) equivalence.</p><!><p>The water contents of the condensed liquids were determined with Karl Fisher titration using the volumetric ASTM E203-08 method. The measurements were made with the Metrohm 870KF Titrino Plus titrator equipped with double Pt-wire electrode (Metrohm AG, Herisau, Switzerland). The titrant was the commercial Hydranal Composite 5K reagent (Sigma-Aldrich, St. Louis, MO, USA). A mixture of chloroform and methanol (3:1, v/v) was used as the solvent. The result of each sample was reported as the average of three parallel measurements. The proportion of organic matter in pyroligneous acids was determined by subtracting the water and the titratable acid contents from the total mass of the liquid.</p><!><p>A further compositional analysis of the organic matter in the conversion distillates was performed using ultrahigh-resolution direct-infusion mass spectrometry, which allows a non-targeted complex mixture characterization without chromatographic separation. All the measurements were performed on a 12-T solariX XR Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer (Bruker Daltonics GmbH, Bremen, Germany) using negative-ion electrospray ionization (ESI). This instrument has been described in detail elsewhere (Zhao et al., 2020). The distillates were diluted with methanol at a ratio of 1:1,000 (v/v) and directly infused into the ion source with a syringe pump, operating at a flow rate of 2 μl/min. Dry nitrogen was used as the drying (200°C, flow rate of 4 L/min) and nebulizing gas (1 bar). The calibration of the mass spectra was done externally using sodium trifluoroacetate clusters. The ions were detected within a mass range of m/z 98–1,000. Three hundred time–domain transients were co-added for each spectrum with a data size of 4 MWord. Bruker Compass ftmsControl 2.1 software was used for the instrument control and data acquisition, and Bruker DataAnalysis 5.1 software was used for data post-processing. To improve mass accuracy, the mass spectra were further internally recalibrated with selected analyte ions. The signal-to-noise ratio was set at 5.0 for the peak picking, and the relative intensity threshold was 0.01%. The following parameters were used for the molecular formula assignment: elemental formula, 12C1–100 1H1–2000 14N0–2 6O0–25 32S0–1; double bond equivalent (DBE), 0–80; and mass error, ≤0.8 ppm. Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) and CERES viewer 1.82 were used for the data sorting and visualization.</p><!><p>In the herbicide tests, turnip rape (B. rapa; Apollo, Boreal Plant Breeding Ltd.) was used as the test species due to its recommended use in phytotoxicity assessments (OECD, 2006). The herbicidal effect of the PAs was assessed according to Hagner et al. (2020) using a Jacobsen germinator (Rubart Apparate GmbH, Laatzen, Germany) that consists of a germination plate being temperature-conditioned by means of a water basin below. Germination spirals (filter papers, Ø = 6 cm) equipped with a paper wick were placed on the germination plate. The wick was led through slots in the germination plate into the water basin below, thus supplying the required humidity and temperature (20°C) to the germination spiral. Turnip rape seeds (20 pieces) were placed on the germination spirals, which were covered with a transparent glass dome to provide the air humidity required for germination. A little hole in the dome ensured minimum evaporation. Seeds achieved 16-h light a day. After 7 days, the covers were removed, the germinated seedlings were counted, and the germination spirals with seedlings were gently transferred into Petri dishes for herbicide treatments. An even deposit of the test liquid over the seedlings was achieved using a Potter precision laboratory spray tower spraying 1.0 ml of the selected liquid per filter paper. Tap water was used as an inert control and BioNeko (120 g/L acetic acid) as the active control. The concentrations of the PAs tested were 50%, 25%, 12.5%, and 5% (v/v), each with four replicates. The treated germination spirals were returned into the Jacobsen germinator and the cover domes were set on their places. The number of living seedlings was counted after 7 days.</p><!><p>The fungicidal properties of the tested PAs were estimated using F. culmorum as the test species. The utilized F. culmorum was separated in 2005 from barley (Hordeum vulgare strain: Presitge) grown in Sotkamo, Finland, and stored in liquid nitrogen (−190°C) since then. Each test liquid was examined in the concentrations of 100%, 50%, 25%, 10%, and 5% (v/v). Tap water was used as an inert control and 1.5 g/L Switch (active substances: 375 g/kg cyprodinil and 250 g/kg fludioxonil) as an active control. Each treatment had three replicates. The test was conducted in Petri dishes containing 25 ml of potato dextrose agar (PDA) as growth media for fungus. Four filter papers (6 mm diameter) treated with 100 µm of the test liquid were placed on the growth medium of each Petri dish. The fungus was inoculated in the middle of the Petri dish and the lids closed. The dishes were incubated at 20°C. After 2, 4, and 7 days of incubation, the fungal-covered area was measured.</p><!><p>All dependent variables (percentage of dead B. rapa, the F. culmorum-covered area, and the development of fungal coverage) were analyzed using generalized linear models (GLMs) with treatment (BioNeko/Switch, wood, bark, HWE bark, HWE+AAE bark, and water control), concentration (from 5% to 50%), and their interaction as fixed effects. The concentrations of the first model were 5%, 12.5%, and 20%, and those of the second were 5%, 10%, 25%, and 50%. In the latter model, a concentration of 25% was chosen and the measurement day (2, 4, and 7 days) was used as a fixed effect instead of concentration. Four replicates were used without randomization.</p><p>The assumption of beta distribution with a logit link was used for percentages (model 1), and for coverage areas, the assumptions of log-normal distribution with an identity link (model 2) and gamma distribution with a log link (model 3) were used because the distributions of areas were highly skewed. The restricted maximum likelihood (REML) estimation method was used for the second model, maximum likelihood (ML) for the first model, and residual pseudo-likelihood (REPL) for the third model.</p><p>Heterogeneous variances for treatments were allowed in the second model. In the third model, correlations between measurement days within the same experimental unit were taken into account using the unstructured covariance structure, which estimates unique correlations for each pair of time points. Both solutions were based on lower values of the information criterion (corrected Akaike information criterion, AICc), although a likelihood test was also used for the decision of heterogeneous variances for treatments.</p><p>All treatments were compared within each concentration (models 1 and 2) or measurement days (model 3). The step-down method of Westfall, which is known to be one of the most effective when the design is balanced, was used for the pairwise comparisons of means with a significance level of 0.05 (Westfall, 1997). The method of Kenward and Roger was used for calculating the degrees of freedom for models 2 and 3 and the residual method used for the first model (Kenward and Roger, 2009).</p><p>The analyses were performed using the GLIMMIX procedure of the SAS Enterprise Guide 7.15 (SAS Institute Inc., Cary, NC, USA).</p><!><p>We hypothesized that several valuable products can be separated from hybrid aspen biomass without losing all the potential of PAs suitable for herbicide and fungicide applications. To prove this, the herbicidal and fungicidal activities of PAs produced from 1) wood, 2) untreated bark, 3) HWE-treated bark, and 4) HWE+AAE-treated bark were examined. The activity tests showed that aspen wood and bark can be used as raw materials for bio-based pesticides as the herbicidal and fungicidal activities of undiluted PA were comparable to those of commercial products. In addition, the bark biomass can be utilized in the production of biopesticides according to the cascade principle as HWE pretreatment even improved the herbicidal activity of bark PA.</p><!><p>The chemical characterization of the bark and wood of hybrid aspen trees was used to evaluate the potential of the raw material in terms of cascade processing and to consider the composition of the biomass in relation to the yield of PAs. In the chemical compositions of wood and bark (Table 2), the most significant difference in terms of cascade use can be found in the extractive contents. The bark was found to have a significantly higher content of hydrophilic extractives than wood, bark biomass thus having more potential as feedstock for cascade utilization. Due to its low hydrophilic extractive content, wood is best suited for direct conversion to PAs without pre-extractions. The content of lipophilic extractives in both biomass fractions was low. Thus, the separation of lipophilic compounds from wood or bark before conversion to PAs does not seem a promising way to increase the value of biomass. Variations in the cellulose and lignin contents were also found between wood and bark, but the differences were not considered essential as pesticidal active PAs are converted at ≤280°C, i.e., below the temperature where cellulose and lignin are thermally degraded (Collard and Blin, 2014).</p><!><p>Basic wood chemical composition of the wood and bark biomass fractions (mean ± SD, n = 6) of Populus tremula L. × Populus tremuloides Michx.</p><p>aTotal lignin content is the sum of acid-insoluble and acid-soluble lignin.</p><!><p>In the torrefaction of lignocellulosic biomass at ≤280°C, hemicelluloses are the most important precursors in PA production (Collard and Blin, 2014). The hemicelluloses of hardwood species consist mainly of xylan and, to a lesser extent, glucomannan and pectin (Sjöström, 1993; Willför et al., 2005). The chain structure of xylan consists of xylose (Xyl) units with an O-acetyl group attached to 7 units out of 10 and 4-O-Me-GlcA to 1 unit out of 10 in ratios (Sjöström, 1993). The O-acetyl groups of xylans are precursors of the acetic acid formed in thermochemical conversion; therefore, a proportion of xylan is the most decisive hemicellulose characteristic of biomass (Chen et al., 2019). The xylan content of the wood and bark of hybrid aspen was estimated from the concentrations of the Xyl and 4-O-Me-GlcA sugar units (Figure 2). The bark was found to have a higher total amount of hemicellulose (Table 2), but a lower xylan content compared to the wood fraction (Figure 2). Due to its higher xylan content, wood can be considered as a more promising feedstock for the production of PAs with a stronger acid concentration.</p><!><p>Mean (±SD) hemicellulose sugar units of hybrid aspen wood and bark per biomass dry weight (d.w.).</p><!><p>The formation of torrefaction products was used to examine the behavior of wood, bark, and pre-extracted bark during the conversion process. In torrefaction, the reactive temperature range for lignocellulosic biomass conversion begins at 150–175°C, and the actual conversion of hemicellulose occurs between 200°C and 300°C (Basu, 2013; Collard and Blin, 2014). The graph of the biomass gas formation rate during conversion showed that the reactive phase of hybrid aspen bark, wood, and pre-extracted bark biomasses began above 220°C and was at its most active phase above 250°C (Figure 3). Bark pre-extractions did not appear to have a significant effect on the reactive temperature range as all biomasses produced gas over the same temperature range. Although there were differences in the gas formation rates between the different biomasses, the total amounts of gases produced were only a few percentage points apart and did not appear to follow a clear trend relative to bark cascading pretreatments. For all feedstocks, gas formation ceased during the temperature maintenance phase at 280°C. Thus, it can be assumed that the biomass conversion reactions reached their end point and that the maximum yields of solid, liquid, and gas products were achieved.</p><!><p>Gas production of hybrid aspen biomass fractions and pre-extracted bark masses (HWE, hot water extraction; AAE, alkaline alcohol extraction) during the torrefaction process. The maximum gas formation rate for each feedstock at different temperatures is highlighted in the diagram.</p><!><p>Figure 4 shows the weight percent (wt.%) yields of the conversion products produced from wood, bark, and pre-extracted bark biomasses. The wood raw material was found to form the highest amount of PAs relative to the dry matter mass, although the total content of hemicelluloses was found to be lower than that in the bark. The higher wood PA yield can be expected as due to the higher xylan ratio of wood hemicelluloses (Prins et al., 2006). Pre-extractions of bark did not appear to have a notable effect on the total yields of PAs.</p><!><p>Solid pyroligneous acid (PA) and gaseous product yields (weight percent of feedstock dry weights) produced by torrefaction from hybrid aspen wood, bark, and pre-extracted bark masses.</p><!><p>Wood biomass had the highest mass loss compared to bark and pre-treated bark biomasses during conversion (Figure 4). In bark torrefaction, HWE and HWE+AAE pretreatments appeared to decrease mass loss, although the differences were small (Figure 4). The effect of AAE treatment on bark mass loss was not expected, as various studies have found that alkali metals bound to organic sites in lignocellulosic biomass can significantly increase mass loss during torrefaction (Shoulaifar et al., 2016a; Shoulaifar et al., 2016b; Macedo et al., 2018; Zhang et al., 2018). The observed opposite result is interesting since bark AAE pretreatment is likely to cause Na residues in the biomass and would be expected to cause accelerated mass loss.</p><!><p>PAs contain precipitate (tar) and liquid phases, both of which were summed to give the total yield of PAs. Tar was separated from the solutions by centrifugation, and only the separated liquids were included in the subsequent herbicide and fungicide experiments to avoid the polyaromatic hydrocarbons (PAHs) in the tar fraction (Fagernäs et al., 2012). PAs obtained from HWE-treated bark contained the highest amount of tar (8.75 wt.%), followed by bark PA (4.09 wt.%), wood PA (1.85 wt.%), and HWE+AAE bark PA (1.29 wt.%).</p><p>Tar-free samples of bark PA were found to differ significantly from wood PA in their basic chemical composition (Figure 5). Bark pre-extractions also had a notable effect on the basic composition of PAs (bark PA vs. HWE bark PA vs. HWE+AAE bark PA) (Figure 5). The basic chemical composition of the PAs was determined by dividing the chemical fractions into groups of water content, acid content (CH3COOH w/v percent equivalent), and other contents of organic matter (Figure 5). Wood PA was found to contain a higher amount of acid, as well as organic matter, compared to the bark PA product. The higher acid content is explained by the higher xylan content of the wood (Prins et al., 2006). In terms of the bark feedstock pre-extraction processing, the most interesting finding emerged from the effect of HWE on the chemical composition of the produced PAs. The acid strength of PAs, as well as the proportion of organic matter, appeared to have increased because of HWE of the bark (bark PA vs. HWE bark PA) (Figure 5). However, following AAE, a significant negative effect on the acid content of PAs was shown, but the proportions of other soluble organic substances were clearly increased (HWE bark PA vs. HWE+AAE bark PA) (Figure 5). The decreased acid content of PAs can be explained by the alkaline extraction conditions that likely affected the hemicellulose xylan (Borrega et al., 2013).</p><!><p>Basic chemical composition of pyroligneous acids (PAs) produced from wood, bark, hot water-extracted (HWE) bark or HWE+alkaline alcohol-extracted (AAE) bark of hybrid aspen.</p><!><p>The composition of the organic matter in the PA samples was further analyzed with negative-ion ESI-FT-ICR mass spectrometry, which specifically targets less volatile organic compounds present in thermochemical wood distillates (Zhao et al., 2020; Zhao et al., 2021), such as organic acids, phenolic compounds, and carbohydrates (sugars and anhydro sugars). Zhao et al. have previously shown that extractives, some phenolics, and hydrocarbons are enriched in the tar (i.e., water-insoluble) fractions, while the more polar oxygenates are enriched in the PA (i.e., water-soluble) fractions (Zhao et al., 2021). The van Krevelen diagrams for all the detected compounds in each sample are presented in Figure 6, which allow the overall chemical compositions to be compared. The compositions of the tar-free PA samples were similar, mainly comprising different phenolic compounds and carbohydrates (mainly hemicellulose-derived mono- and disaccharides) (Figure 6), consistent with the previous work of Zhao et al. (2021). However, the HWE+AAE bark PA had clearly lower amounts of phenolics and monosaccharides than the other samples. Due to the complex chemical nature of the samples, individual compounds were not further identified in this work. The smallest organic acids, alcohols, or furanic compounds were not efficiently ionized by ESI and needed to be analyzed by other means, e.g., GC-MS or photoionization combined with FT-ICR MS. These further analyses remain as a topic for future studies. The negative-ion ESI-FT-ICR mass spectra of the PAs are shown in the Supplementary Material (Supplementary Figure S1).</p><!><p>Van Krevelen diagrams for the compounds detected in different conversion distillates with negative-ion ion electrospray ionization Fourier transform ion cyclotron resonance (ESI-FT-ICR) mass spectrometry. The dot size/color represents the relative abundance of each detected compound.</p><!><p>The herbicidal activity of the PAs produced from hybrid aspen fractions and pretreated bark masses was compared with that of the acetic acid-based commercial herbicide (BioNeko) and that of water. Seven days after the spraying, all PA solutions with 20% or higher concentrations eliminated 100% of the B. rapa seedlings, being as effective as the commercial herbicide (BioNeko), but each differing significantly from water (p < 0.05). At lower dilutions, differences between the PAs produced from various feedstocks were revealed (Figure 7): as 12.5% solutions, wood PA, bark PA, and HWE bark PA were as herbicidal active as BioNeko, but HWE+AAE bark PA was found to have a significantly reduced activity (HWE+AAE bark PA vs. all others, p < 0.05). At 5% dilutions (Figure 7), wood PA still retained its herbicidal activity, being as effective as the commercial herbicide (wood PA vs. BioNeko, p = 0.847). However, the herbicidal activity of bark PA was significantly lower than that of wood PA (5% bark PA vs. 5% wood PA, p < 0.001). HWE of the bark had a significant positive effect on the herbicidal activity of the PAs produced from the bark, as the performance of HWE bark PA was not statistically different from that of wood PA (HWE bark PA vs. wood PA, p = 0.919) or that of BioNeko (HWE bark PA vs. BioNeko, p = 0.274).</p><!><p>Herbicidal effectivity of pyroligneous acids (PAs) produced from wood and variously treated (HWE, hot water extracted; AAE, alkaline alcohol extraction) bark products (mean ± SD, n = 4) against Brassica rapa seedlings 7 days after the spraying compared to that of a commercial product (BioNeko) and water controls.</p><!><p>To investigate the fungicidal (F. culmorum) activity of PAs produced from hybrid aspen biomasses, we compared the antifungal efficiency of 50%, 25%, 10%, and 5% PA dilutions (Figure 8). In addition, the antifungal effect of 25% dilution of PAs produced from different biomasses was monitored for 7 days (Figure 9). Wood PA, bark PA, HWE bark PA, and HWE+AAE bark PA corresponded in 50% solutions to the commercial product Switch during the 7-day observation period, preventing F. culmorum from spreading almost completely (Figure 8). Dilutions of 25% of bark PA and pre-extracted bark PA showed a reduced antifungal effect, but wood PA still corresponded to Switch (wood PA vs. Switch, p = 0.866). It can be seen from the 25% dilutions in Figure 8 that the bark pre-extractions were found to have a negative effect on the fungicidal performance of the PAs produced. The growth area of F. culmorum was shown to be increased when comparing the growth inhibition performance of bark PA to that of HWE bark PA (p = 0.006). Comparison of the fungicidal performance between HWE bark PA vs. HWE+AAE bark PA showed that the effect appeared to have decreased further, although statistical difference was not found (p = 0.167). All PA solutions significantly lost their fungicidal activity against F. culmorum at 10% dilutions, and the activities no longer differed from water at 5% dilutions (all 5% PAs vs. water, p > 0.05). Figure 9 shows the F. culmorum growth inhibitory ability of 25% PAs over a 7-day follow-up period. After 4 days from treatments, differences in the fungicidal activity of PAs slightly appeared, but on day 7, only wood PA corresponded to the Switch product. Ascending growth trends suggest that the fungicidal performance of all tested treatment agents appeared to have decreased over the follow-up period, but this should be verified by extending the observation period.</p><!><p>Fungicide potency of pyroligneous acids (PAs) produced from aspen tree wood, bark, or variously extracted bark (HWE, hot water extraction; AAE, alkaline alcohol extraction), commercial herbicide (Switch), or water measured via the inhibition of fungal (Fusarium culmorum) growth 7 days after the treatments (mean ± SD), n = 4.</p><p>Fungal coverage area (in square millimeters) at 2, 4, and 7 days after treating the growth media with pyrolysis acids produced from wood, bark, hot water-extracted (HWE) bark, or HWE+alkaline alcohol-extracted (AAE) bark (mean ± SD, n = 4).</p><!><p>The use of PAs as herbicides, insecticides, and fungicides has a long tradition in many Asian countries (Yatagai et al., 2002; Tiilikkala et al., 2010; Hossain et al., 2015). PAs produced by slow pyrolysis have been proven to be effective against a wide number of pests (Yatagai et al., 2002; Lindqvist et al., 2010; Hossain et al., 2015), to have antifungal activity against pathogenic fungi and yeasts (Ibrahim et al., 2013; Oramahi and Yoshimura, 2013; Mashuni et al., 2020), and to induce systemic resistance to fungal diseases in plants (Kårlund et al., 2014). The results of our study agree with those of previous studies and extend the results to also concern PAs produced by torrefaction. However, we are not aware of any studies concerning the utilization of biomasses according to the cascade principle, which allows the separation of specific soluble and insoluble compounds before pyrolysis and biopesticide production. As HWE pretreatment even improved the herbicidal activity of bark PA, the cascading use, i.e., the separation and utilization of hot water-extractable chemicals before the production of PAs for herbicidal purposes, is possible. However, the pretreatments used decreased the fungicidal properties of the PAs. Consequently, it is the target organism that determines the usability of various pretreatments to produce biopesticides from wood biomass by torrefaction.</p><!><p>The results of the study showed that both wood and bark biomasses of hybrid aspen (Populus tremula L. × P. tremuloides Michx.) are promising raw materials for the production of pesticide-active PAs on weeds and fungal diseases. PAs converted from wood, bark, HWE bark, and HWE+AAE bark were equally herbicidal active solutions at >20% and fungicidal active solutions at >50% with commercial products. The results are novel, as we showed for the first time that, along with the production of biopesticides, several other valuable products can also be separated from the bark, if the solution strengths of the herbicides and fungicides used are at least as described above. We further observed that HWE pretreatment improved the herbicidal activity of bark PA, which also supports the cascade utilization potential of the bark. No similar benefit was observed with alkaline alcohol extraction, but despite pretreatment, the PA did not completely lose its biopesticide potency. The results also showed that the fungicidal activity of PAs produced from hybrid aspen biomasses was lower compared to its herbicidal activity. This is evident in the differences of the activity of the PA dilution ratios tested. It can be summarized that utilization according to the cascade principle is possible by separating the tree biomass fractions into their own raw material sources and by using different separation extractions to form value side streams. Still, further research is needed to determine the market value and market entry potential of the products. Also, more detailed characterization of the chemical products produced from hybrid aspen biomasses awaits further research.</p>

PubMed Open Access

Exposing Plasmids as the Achilles\xe2\x80\x99 Heel of Drug-Resistant Bacteria

Many multi-drug resistant bacterial pathogens harbor large plasmids that encode proteins conferring resistance to antibiotics. While the acquisition of these plasmids often enables bacteria to survive in the presence of antibiotics, it is possible that plasmids also represent a vulnerability that can be exploited in tailored antibacterial therapy. This review highlights three recently described strategies designed to specifically combat bacteria harboring such plasmids: Inhibition of plasmid conjugation, inhibition of plasmid replication, and exploitation of plasmid-encoded toxin-antitoxin systems.

exposing_plasmids_as_the_achilles\xe2\x80\x99_heel_of_drug-resistant_bacteria

Introduction<!>Plasmids as mobile genetic elements that mediate drug resistance<!>Inhibition of plasmid conjugation<!>Inhibition of plasmid replication by mimicking plasmid incompatibility<!>Toxin-antitoxin systems<!>Chromosomally-encoded toxin-antitoxin systems<!>Plasmid-encoded toxin-antitoxin systems<!>Targeting toxin-antitoxin systems<!>Analysis, Summary, and Future Directions

<p>Bacterial resistance to antibiotics is a worldwide health crisis [1]. Resistance to multiple antibiotics has been reported in nearly all pathogenic bacteria, with vancomycin-resistant enterococci (VRE), methicillin-resistant Staphylococcus aureus (MRSA), multi-drug resistant (MDR) Pseudomonas aeruginosa, extensively-drug resistant (XDR) Mycobacterium tuberculosis, MDR Acinetobacter baumannii, β-lactam-resistant Enterobacteriaceae, and penicillin-resistant Streptococcus pneumoniae (PRSP) being particularly notorious [1-3]. Resistance typically occurs as a result of chromosomal mutation or acquisition of a mobile genetic element, such as a plasmid, that harbors resistance-mediating genes. The looming threat of a "post-antibiotic" era where untreatable bacterial infections are common is exacerbated by the shift of research programs in the pharmaceutical industry away from the development of novel antibacterials [4,5]. New strategies to combat drug-resistant bacteria are necessary to keep pace with ever-evolving bacterial resistance.</p><!><p>Lateral transfer of mobile genetic elements between diverse bacteria leads to a rapid dissemination of genes encoding resistance to antibiotics. These mobile genetic elements include plasmids, which are extrachromosomal DNA that transfer horizontally within and across bacterial genera and species by conjugation. Plasmids can also serve as vehicles for transposons and integrons; thus, through plasmid conjugation bacteria are exposed to a wide array of genes from the mobile gene pool. Plasmid-encoded resistance to multiple antibiotics, including β-lactams, aminoglycosides, tetracyclines, macrolides, and glycopeptides is prevalent in a plethora of pathogenic bacteria including VRE and MRSA [6,7]. In fact, recent analyses of >100 VRE isolates from humans, animals, and food show that vanA, the gene cluster encoding vancomycin resistance, resides in the Tn1546 transposon carried on plasmids [8,9]. In addition, plasmid-encoded virulence and antibiotic resistance contribute to the pathogenicity of biowarfare agents such as Bacillus anthracis and Yersinia pestis [10-12]. Most frightening is the recently observed transfer of plasmids from VRE to MRSA, resulting in the virtually untreatable vancomycin resistant S. aureus (VRSA) [13,14].</p><p>However, the very nature of their importance to the antibiotic resistant phenotype may expose plasmids as the Achilles' heel of drug-resistant bacteria. Indeed, creative strategies have recently been devised to prevent the transfer of plasmids between bacteria, to inhibit plasmid replication and hence induce the elimination of plasmids from bacteria, and to exploit plasmid maintenance systems to directly and selectively induce death in drug-resistant bacteria (Figure 1). [15-24]. Although compounds based on these approaches have not yet progressed to clinical trials, the well-documented prevalence of plasmids within the most problematic drug-resistant bacteria makes the targeting of plasmid-encoded elements an intriguing antibacterial option. This Current Opinion focuses on these recent efforts to exploit plasmids in antibacterial therapy.</p><!><p>To prevent the transfer and dissemination of resistance-mediating plasmids, the inhibition of plasmid conjugation has been postulated as a prophylactic strategy [15,25]. Using a cell-based assay involving the transfer of a plasmid containing the lux gene (encoding luciferase) from a donor strain to a recipient strain (Figure 2A), chemical libraries and bacterial/fungal extracts were screened for inhibitors of plasmid conjugation [15,16]. Through these screens dehydrocrepenynic acid (DHCA) and linoleic acid were identified as conjugation inhibitors (Figure 2B). The compounds depicted in Figure 2B were found to only inhibit the transfer of plasmids with similar DNA replication and transfer machinery and did not inhibit the proteins involved in the mating bridge (mating pair formation). Secondary assays ruled out general inhibitory effects of these unsaturated fatty acids, suggesting that these compounds may act through a conjugation-specific mechanism [15].</p><p>A subsequent study on conjugation inhibition also used the same fluorometric, cell-based assay to identify intrabodies that specifically inhibit conjugation [16]. Intrabodies are intracellularly-expressed antibodies that have been used to inactivate proteins in yeast [26,27], plants [28,29], mammals [30-32], and bacteria [33-35]. The relaxase enzyme, which catalyzes the cleaving and religating of plasmid DNA, is an essential component of plasmid conjugation systems (Figure 3A). Recognizing the critical importance of relaxases to plasmid conjugation, Garcillan-Barcia and co-workers expressed intrabodies in the recipient cell to inactivate the TrwC relaxase enzyme encoded by plasmid R388 in a proof-of-concept study [16]. Mice were immunized with the TrwC relaxase domain (the N-terminal 293 amino acids (N293)), and single chain Fv antibody clone libraries were created from splenocytes. Screening of the intrabody libraries for their binding to TrwC-N293 and for their inhibition of conjugation using the aforementioned fluorescence-based assay yielded two conjugation inhibitors, scFv-P4.E7 and scFv-P1.F2. Whereas scFv-P4.E7 recognizes a region of TrwC not known to be involved in catalysis, scFv-P1.F2 binds to the conserved motif 1 of the MOBF relaxase family, which is a mobile loop containing the catalytic tyrosine-26 [36,37]. TrwC relaxase function depends on two catalytic tyrosines: Y18 carries out the initial cleavage event at oriT and Y26 is thought to catalyze a transesterification, which recircularizes the T-DNA, the DNA that is transferred, in the recipient cell [38]. The observed 20-fold conjugation inhibition of scFv-P1.F2 matches the reduction in activity observed by the TrwC-Y26F mutant [38], suggesting that the binding of scFv-P1.F2 to the mobile Y26-containing loop may prevent the transesterification and recircularization of T-DNA in the recipient cell. Another intriguing result is that mutant TrwC-Y18F but not wild-type TrwC could partially rescue the reduced conjugation of TrwC-Y26F, suggesting different roles for each tyrosine and possible different conformations of TrwC during conjugative DNA processing. Using a target-based approach to study conjugation, these results confirm previous evidence that TrwC is active in the recipient cell and suggests relaxase inhibition is a viable strategy for preventing plasmid conjugation. However, because these intrabodies do not actually induce cell death, this type of prophylactic strategy would only be useful for preventing the dissemination of genes that mediate antibiotic resistance.</p><p>Although the use of relaxase-targeting intrabodies validated the notion that interference with relaxase function could inhibit plasmid conjugation, the therapeutic application of intrabodies will be difficult due to the biological stability, cell permeability, and pharmacokinetic problems faced by any macromolecular drug. In a recent study by Lujan and co-workers, however, a series of small molecule relaxase inhibitors were identified and shown to prevent plasmid conjugation [17]. Through X-ray crystallographic analysis of the F plasmid TraI-N300 relaxase domain, it was hypothesized that simple bisphosphonates could interact with the active site Mg2+ ion and two catalytic tyrosine residues to inhibit relaxase. Thus, an enzymatic assay was developed that measured the relaxase-catalyzed cleavage of a fluoroescently-labeled ssDNA containing the F plasmid oriT sequence. This assay showed that nanomolar concentrations of imidobisphosphate (PNP) (Figure 3B) inhibited relaxase-catalyzed cleavage of oriT ssDNA. The crystal structure of PNP-bound relaxase revealed a phosphate of PNP within 3.7 Å of the Mg2+ metal center. Of twelve bisphosphonates tested in the kinetic assay, six compounds (shown in Figure 3B) were found to potently inhibit relaxase.</p><p>A conventional conjugation assay showed that PNP inhibited transfer of the F plasmid between two E. coli cells with an EC50 of 10 μM. Surprisingly, PNP was found to selectively kill F+ E. coli expressing the TraI relaxase but had no effect on strains containing TraI relaxase but no F plasmid, F plasmid but no TraI relaxase, or F plasmid in which all four relaxase active site tyrosines were mutated to phenylalanines. These data suggests that PNP inhibits conjugation and produces a bactericidal effect dependent on the presence of active relaxase and F plasmid. The exact mechanism behind this relaxase-dependent antibacterial activity of bisphosphonates is unknown. All six bisphosphonates in Figure 3B inhibited conjugation and displayed F plasmid specific killing in the nanomolar-to-low-micromolar range, making them significantly more selective over cells lacking the F plasmid. Two of these potent bisphosphonates, Clodronate and Etidronate, are clinically approved for the treatment of bone disease. These compounds are promising candidates for use in combination with current antibiotics to prevent dissemination of plasmid-encoded antibiotic resistance in the gastrointestinal tract, and may have potential as single entity antibacterial agents against bacteria harboring plasmid-encoded relaxases. Before either of the relaxase-targeting strategies described above can be broadly utilized, there will need to be a demonstration that homologous relaxases are present and active in clinically significant bacterial pathogens.</p><!><p>Another novel approach to combat bacteria harboring plasmid-encoded resistance genes is the use of small molecules to inhibit plasmid replication and hence eliminate the plasmid from the bacterial population. Plasmid incompatibility is a natural phenomenon for plasmid elimination; two plasmids of the same incompatibility group will not stably cosegregate to a daughter cell. Studies by DeNap et al. [18] and Thomas et al. [19] exploit this natural mechanism in the identification of small molecule mimics of plasmid incompatibility, "antiplasmid" compounds that eliminate plasmids from the bacterial population and re-sensitize the bacteria to antibiotics.</p><p>Plasmid incompatibility is determined by the plasmid replication machinery, which has been extensively studied in the incompatibility group B (IncB) plasmids [39-46]. IncB plasmid replication is tightly controlled by the levels of the phophodiesterase RepA [47,48], the translation of which is controlled by a small, untranslated RNA, called RNA I (Figure 4A). RepA mRNA forms an intramolecular pseudoknot between stem-loop I (SLI) and stem-loop III (SLIII), which allows RepA translation and hence plasmid replication [43]. RepA translation is shut down when the countertranscript RNA I binds SLI [39,44-46]. In an effort to mimic this process with a small molecule, antiplasmid compounds that bound to SLI were sought. A variety of aminoglycosides were tested for their ability to bind SLI RNA, and apramycin (Figure 4B) was found to bind with a Kd of 93 nM. Through mutagenesis studies it was determined that bases A22 and A23 on SLI were essential for apramycin binding, as this binding event was completely abolished in the SLI-A22G/A23G double mutant. Plasmid stability assays showed that the IncB plasmid was almost completely eliminated after 250 generations, and a general correlation between SLI binding affinity and plasmid loss was observed [19]. In contrast, when the SLI-A22G/A23G mutations were created on the same IncB plasmid (abolishing the apramycin binding site), this plasmid was not eliminated by apramycin. These studies demonstrate that plasmids can be eliminated from bacterial cells in a mechanistically distinct fashion, that is, through the identification of compounds that bind tightly to RNAs essential to plasmid replication control. For this approach to find clinical utility, the homology of the RNAs that mediate plasmid replication control in pathogenic bacteria will need to be investigated. The little information that is available does indicate that some plasmids do indeed have homologous regions in these key countertranscript RNAs [49,50]. Furthermore, identification of compounds that cause more rapid plasmid loss will improve this strategy and increase its potential in antibacterial therapy.</p><!><p>Proteic toxin-antitoxin systems, found on both bacterial plasmids and chromosomes, produce a stable toxic protein and a labile antitoxin protein. The possibility of exploiting toxin-antitoxin (TA) systems as a novel antibacterial strategy with a compound that activates the latent toxin through one of two pathways (Figure 5), has been proposed [6,20-22,51]; although the end result (toxin-induced cell death) is the same, the two strategies depicted in Figure 5 differ mechanistically. In pathway 1, a compound acts at either the transcriptional or translational level to prevent the synthesis of new antitoxin. Thus, when the highly labile pre-existing antitoxin molecules are degraded, the stable toxin is freed to kill the cell. The second mechanism for the exploitation of TA systems as antibacterial targets involves the direct disruption of the toxin-antitoxin protein-protein interaction, freeing the toxin to induce cell death (pathway 2 in Figure 5) [6,20,21,51]. When considering TA systems as an antibacterial target, it is helpful to make a distinction between TA systems found on chromosomes and those found on plasmids, although compounds acting through either mechanism should be effective against plasmid- and chromosomally-encoded TA systems alike.</p><!><p>Although the genes for toxin-antitoxin proteins have been found on bacterial and archaeal chromosomes, the function of chromosomally-encoded TA systems remains elusive. Data from several studies indicate that these systems function to halt bacterial growth during times of stress (Figure 6A). For example, the mazEF TA system has been described as a suicide module that causes programmed cell death (PCD) in response to extreme amino acid starvation. In this scenario relA synthesizes the stringent response molecule guanosine 3′,5′-bispyrophosphate (ppGpp), inhibiting mazEF transcription, activating MazF, and ultimately leading to cell death [52-54]. Furthermore, addition of antibacterials that inhibit transcription (rifampicin), translation (chloramphenicol and spectinomycin) or that cause thymine starvation (trimethoprim and sulfonamide) cause mazEF-dependent cell death [23,55,56]. Based on these studies it has been proposed that a new class of antibacterials could be developed that would stress the cells such that the toxin protein(s) are activated, causing cell death [22,23]. The recent discovery of a short peptide that appears to induce bacterial cell death in a MazF-dependent fashion in E. coli bolsters the argument that chromosomally-encoded TA systems are a tractable antibacterial target [24,57].</p><p>However, other studies offer conflicting evidence, including a recent report in which the genes for several chromosomally-encoded TA systems were systematically knocked out in E. coli, and the resulting bacterial strain had no obvious change in phenotype in response to the cellular stresses that were tested [58]. Given this contradictory evidence, a variety of potential functions for chromosomally-encoded TA systems have been postulated, including the possibility that they have no function [59]. TA systems have also been reported as modulators of the persister cell phenotype, in which cells neither grow nor die in the presence of bactericidal antibiotics, resulting in multi-drug tolerance (MDT) [60-62]. HipA, of the TA operon hipBA, was the first validated persister-MDT protein; knocking out hipA significantly reduces the occurrence of persister cells [61]. However, knocking out other TA systems shown to be involved in producing the persister cells in E. coli resulted in no phenotype, thus suggesting that persister genes are redundant [62].</p><p>Although several genomic studies have revealed the presence of TA genes on the chromosomes of a variety of different bacteria [63-65] and their absence in obligate host-associated organisms, definitive evidence showing that chromosomally-encoded TA genes are functional in clinical isolates of pathogenic bacteria will be required before disruption of chromosomally-encoded TA systems can be considered a viable antibacterial strategy. Furthermore, understanding the role of chromosomally-encoded TA systems in the formation of persister cells will help to further evaluate TA systems as an antibacterial target.</p><!><p>The role of plasmid-encoded TA systems is to function as post-segregational killers (PSKs) (Figure 6B) [66-68]. Proteic TA systems are utilized by plasmids to ensure that only those daughter cells that inherit the plasmid survive after cell division. When both proteins are present, the antitoxin binds to the toxin, preventing its toxic activity. However, if during cell division a plasmid-free daughter cell arises, the labile antitoxin is quickly degraded (and not replenished), freeing the toxin to induce cell death. Because of this indelible link between plasmid maintenance and bacterial life, TA systems have been termed 'plasmid addiction systems' [69].</p><p>Before the search for toxin activators could commence, it was necessary to know if the genes for TA systems were present on plasmids isolated from major drug-resistant bacterial pathogens, if a certain TA systems was more prevalent than others (making it a more attractive antibacterial target), and if these plasmid-encoded TA genes were functional in the drug-resistant bacteria. A recent epidemiological survey of VRE isolates provided answers to these questions [21]. In this survey, plasmids were purified from 75 different VRE clinical isolates, and then probed by PCR for the presence of TA genes. Surprisingly, the genes for TA systems were found to be ubiquitous on plasmids from VRE and physically linked to the vanA gene cluster; 75 out of 75 VRE isolates contained plasmids harboring genes for TA systems. Certain TA systems were indeed more prevalent than others, with mazEF (75 out of 75), axe-txe (56 out of 75), relBE (35 out of 75), and ω-ε-ζ (33 out of 75) being the most common. RT-PCR showed that the mazEF transcripts are produced in the VRE isolates. Furthermore, plasmid pS345RF, which contains mazEF as the only detectable TA system, was shown to be highly stable in the absence of selection. Finally, the cloning of mazEF and its native promoter into the unstable enterococcal vector pAM401 was shown to impart a significant increase in plasmid stability, thus suggesting that TA systems are functional in VRE [21].</p><!><p>The discovery that certain TA systems are ubiquitous in clinical isolates of difficult-to-treat drug-resistant pathogens suggests the exciting possibility that disruption of TA systems could be a viable target for tailored antibacterial therapy. The next challenge is to develop high-throughput screens and use them to identify compounds that induce toxin-dependent death. In this vein, a continuous fluorometric assay that follows the ribonuclease activity of MazF was recently developed [70]. This assay employs a short oligonucleotide containing the MazF cleavage sequence 5′-labeled with 6-carboxyfluorescein (6-FAM) and 3′-labeled with a black-hole quencher (BHQI). Cleavage of the oligonucleotide releases the fluorophore from the quencher, resulting in a large increase in fluorescence emission of 6-FAM. This in vitro assay could be used to screen compounds for their ability to induce activation of MazF activity from the MazE-MazF complex. Cell-based assays to identify compounds that selectively restrict growth of toxin-antitoxin producing bacteria can also be envisioned.</p><p>It is possible that targeting plasmid-encoded TA systems could have advantages over traditional antibiotics. One could envision resistance to such toxin-activating compounds arising through an inactivating mutation in the toxin protein, or through mutation of the target of the toxic protein. However, both situations are problematic from the bacteria's perspective. If the toxin protein is mutated and inactivated, a compound that released the toxin would indeed no longer be an effective antibacterial. On the other hand, this mutation would also eradicate the plasmid stabilization system, and hence the plasmid (containing the drug-resistance genes) would be eliminated from the bacterial population, re-sensitizing the bacteria to conventional antibiotics. Mutation of the target would be equally complicated as a resistance mechanism. It is difficult to foresee how toxic RNase activity (such as in MazF) could be abolished through target mutation. For the toxins that inhibit DNA gyrase (such as CcdB), mutants of this enzyme could indeed arise that are resistant to the toxin proteins. However, once again this sort of mutation would eliminate the natural function of the TA systems, that of plasmid stabilization; gyrase mutants would be resistant to the post-segregational killing effect, and plasmid-free daughter cells (that are sensitive to the effect of antibiotics) would likely arise quickly.</p><!><p>The very fact that plasmids are responsible for large swaths of drug-resistance in bacteria makes them attractive antibacterial targets. It would seem that prophylactic strategies, for example those that are designed to stop the spread of drug-resistance genes through inhibition of plasmid conjugation, are less attractive and less practical then those that directly induce bacteria cell death. However, as shown by the creative work of Lujan and co-workers [17], the prevention of plasmid conjugation through the inhibition of relaxase can indeed directly induce cell death, a surprising and welcome discovery. While strategies that rely on compounds to induce plasmid elimination may have some utility, the heterogeneity of the plasmid replication elements and the number of generations required for elimination will need to be defined in clinical isolates before this approach can be implemented. The direct induction of cell death through the disruption of toxin-antitoxin systems appears to hold considerable promise, given the ubiquity of certain TA systems on plasmids isolated from VRE and the strong toxicity of the various toxic proteins. Although the lack of plasmids in XDR M. tuberculosis presents a limitation for the proposed plasmid conjugation and replication inhibition antibacterial strategies, toxin-antitoxin systems have been shown to reside on the M. tuberculosis chromosome [63,71].</p><p>As these approaches toward utilizing mobile genetic elements against bacteria are further explored and exploited, a major effort will need to be made to move studies past proof-of-concept work in E. coli with model plasmids and into demonstrations in actual clinical isolates. Given the wide array of naturally occurring plasmids, target heterogeneity will be a major question for any strategy seeking to exploit a plasmid-encoded trait; an ideal target would be one that is fully conserved throughout a variety of difficult-to-treat bacteria. Because plasmids often harbor the genes for resistance-mediating enzymes, standard mechanisms of resistance may not be as applicable to compounds that target plasmid-encoded elements. Just as dipping Achilles into the river Styx gave him overall strength but left his heel vulnerable, so too does the very resistance conferred on bacteria by a plasmid make them susceptible to plasmid-targeting strategies.</p>

PubMed Author Manuscript

A Targetable Fluorescent Sensor Reveals that Copper-Deficient SCO1 and SCO2 Patient Cells Prioritize Mitochondrial Copper Homeostasis

We present the design, synthesis, spectroscopy, and biological applications of Mitochondrial Coppersensor-1 (Mito-CS1), a new type of targetable fluorescent sensor for imaging exchangeable mitochondrial copper pools in living cells. Mito-CS1 is a bifunctional reporter that combines a Cu+-responsive fluorescent platform with a mitochondrial-targeting triphenylphosphonium moiety for localizing the probe to this organelle. Molecular imaging with Mito-CS1 establishes that this new chemical tool can detect changes in labile mitochondrial Cu+ in a model HEK 293T cell line as well as in human fibroblasts. Moreover, we utilized Mito-CS1 in a combined imaging and biochemical study in fibroblasts derived from patients with mutations in the two synthesis of cytochrome c oxidase 1 and 2 proteins (SCO1 and SCO2), each of which is required for assembly and metallation of functionally active cytochrome c oxidase (COX). Interestingly, we observe that although defects in these mitochondrial metallochaperones lead to a global copper deficiency at the whole cell level, total copper and exchangeable mitochondrial Cu+ pools in SCO1 and SCO2 patient fibroblasts are largely unaltered relative to wildtype controls. Our findings reveal that the cell maintains copper homeostasis in mitochondria even in situations of copper deficiency and mitochondrial metallochaperone malfunction, illustrating the importance of regulating copper stores in this energy-producing organelle.

a_targetable_fluorescent_sensor_reveals_that_copper-deficient_sco1_and_sco2_patient_cells_prioritize

Introduction<!>Design and Synthesis of a Targetable Coppersensor Platform and Preparation of Mito-CS1<!>Spectroscopic Properties of Mito-CS1<!>Fluorescence Detection of a Labile Mitochondrial Cu+ Pool in HEK 293T Cells Using Mito-CS1<!>Mito-CS1 Imaging and ICP-OES Experiments Reveal that Mitochondrial Cu+ Pools in SCO1 and SCO2 Patient Fibroblasts Are Comparable to Wildtype Counterparts<!>Concluding Remarks<!>Synthetic Materials and Methods<!>4-Hydroxy-2,6-dimethylbenzaldehyde (1)<!>Methyl 2-(4-formyl-3,5-dimethylphenoxy)acetate (2)<!>Methyl 2-(4-(di(1H-pyrrol-2-yl)methyl)-3,5-dimethylphenoxy)acetate (3)<!>(Z)-Methyl 2-(4-((5-chloro-1H-pyrrol-2-yl)(5-chloro-2H-pyrrol-2-ylidene)methyl)-3,5-dimethylphenoxy)acetate (4)<!>3,7-Dichloro-5,5-difluoro-10-(4-(2-methoxy-2-oxoethoxy)-2,6-dimethylphenyl)-5H-dipyrrolo[1,2-c:2\xe2\x80\xb2,1\xe2\x80\xb2-f][1,3,2]diazaborinin-4-ium-5-uide (5)<!>3-Chloro-5,5-difluoro-7-methoxy-10-(4-(2-methoxy-2-oxoethoxy)-2,6-dimethylphenyl)-5H-dipyrrolo[1,2-c:2\xe2\x80\xb2,1\xe2\x80\xb2-f][1,3,2]diazaborinin-4-ium-5-uide (6)<!>3-(Bis(2-((2-(ethylthio)ethyl)thio)ethyl)amino)-5,5-difluoro-7-methoxy-10-(4-(2-methoxy-2-oxoethoxy)-2,6-dimethylphenyl)-5H-dipyrrolo[1,2-c:2\xe2\x80\xb2,1\xe2\x80\xb2-f][1,3,2]diazaborinin-4-ium-5-uide (8)<!>3-(Bis(2-((2-(ethylthio)ethyl)thio)ethyl)amino)-10-(4-(carboxymethoxy)-2,6-dimethylphenyl)-5,5-difluoro-7-methoxy-5H-dipyrrolo[1,2-c:2\xe2\x80\xb2,1\xe2\x80\xb2-f][1,3,2]diazaborinin-4-ium-5-uide (Carboxy-CS1 9)<!>(2-Aminoethyl)triphenylphosphonium bromide (10)<!>Mitochondrial Coppersensor-1 (Mito-CS1, 11)<!>Rhodamine 123<!>Spectroscopic Materials and Methods<!>Preparation of Cell Cultures<!>Cell Staining<!>Fluorescence Imaging Experiments<!>ICP-OES analysis.36<!>Miscellaneous Procedures

<p>Copper is a required element for life, and regulating its uptake, efflux, and compartmentalization at the cellular level is vital for maintaining normal physiology.1-3 As such, cells have evolved intricate mechanisms that coordinate the activity of transporters, chaperones, and small-molecule ligands to dynamically control the distribution of copper within discrete subcellular spaces.4-24 In this context, mitochondria are important reservoirs for cellular copper owing to the essential role of this metal ion as a cofactor for the respiratory enzyme cytochrome c oxidase (COX), which reduces oxygen to water in the terminal step of aerobic respiration.25-30 In addition to the copper present in COX, mitochondria utilize this metal ion as a cofactor for superoxide dismutase (SOD1) contained within the mitochondrial intermembrane space (IMS), and a matrix-localized pool of copper is relocalized to the IMS to metallate both of these enzymes.31 Assembly of active COX in mitochondria requires a host of proteins, including the two synthesis of cytochrome c oxidase genes (SCO1 and SCO2) that collectively mediate a series of metal- and redox-dependent events crucial for COX metallation and holoenzyme maturation.32-44 Mutations in either of the SCO chaperones result in severe, tissue-specific clinical phenotypes that are caused by both a failure to mature functionally active COX for aerobic respiration, and an inability to appropriately regulate cellular copper homeostasis.36,45-59</p><p>Because regulating mitochondrial copper homeostasis is critical to maintaining central oxygen metabolism in the cell, new chemical tools that allow direct, real-time visualization of exchangeable mitochondrial copper pools in living samples are potentially powerful reagents with which to directly investigate the spatiotemporal distribution of this redox-active metal in both healthy and disease states. Towards this end, we have initiated a broad-based program to create fluorescent and MRI agents for monitoring labile copper stores in living systems.60-68 Previous work from our laboratory on fluorescent sensors for live-cell copper imaging include intensity and ratiometric probes Coppersensor-1 (CS1)60,61, Coppersensor-3 (CS3)68, and Ratio-Coppersensor-1 (RCS1)67, that are capable of selectively tracking labile cellular Cu+ stores with metal and/or redox stimulation. However, these first-generation sensors and other elegant examples of small molecule or protein-based copper sensors69-71 are not preferentially directed to mitochondria or other organelles, which offers an opportunity to devise new probes that can be targeted to discrete subcellular compartments for imaging local changes in the abundance/availability of exchangeable Cu+.</p><p>In this report we present the design, synthesis, spectroscopy, and biological applications of Mitochondrial Coppersensor-1 (Mito-CS1), a new type of fluorophore for imaging dynamic mitochondrial copper stores in living cells with metal ion and spatial specificity (Scheme 1). Mito-CS1 can reversibly detect endogenous, exchangeable mitochondrial Cu+ pools in a model HEK 293T cell line as well as in cultured human fibroblasts. Moreover, we apply Mito-CS1 in a combined molecular imaging and biochemical study to investigate mitochondrial copper homeostasis in fibroblasts derived from patients with SCO1 and SCO2 mutations. Interestingly, the data establish that although loss of function mutations in SCO1 and SCO2 cause a severe, global copper deficiency at the whole cell level, both the exchangeable Cu+ and total mitochondrial copper pools are only mildly perturbed in SCO1 and SCO2 patient fibroblasts compared to wildtype congeners. Our results show that the cell maintains the homeostatic regulation of copper within mitochondria, even when faced with a globally severe state of copper deficiency, and underscore the primary importance of this redox-active metal in central metabolism and cellular energy homeostasis.72</p><!><p>In order to monitor labile mitochondrial copper pools in living cells by real-time fluorescence imaging, we designed a bifunctional BODIPY dye that contains both a fluorescence-responsive Cu+-binding domain and a mitochondrial-targeting moiety. For the latter, we exploited phosphonium head groups that have been pioneered by Murphy73-76 and subsequently utilized by ourselves77,78 and others79-84 to selectively direct attached cargo to mitochondria via their proton gradients. In addition, we designed a synthetic pathway that builds up to one key intermediate, Carboxy-CS1, that combines a BODIPY chromophore, a thioether-rich receptor for selective recognition and binding of Cu+, and a carboxylic acid synthetic handle (Scheme 1, compound 9) for facile introduction of a triphenylphosphonium tag or any other desired targeting functionality. We note that modification of the BODIPY chromophore to incorporate additional functionalities has been explored in the literature.85,86</p><p>Scheme 1 outlines the synthesis of Mito-CS1. Briefly, alkylation of aldehyde 1 with methyl bromoacetate furnishes the methyl ester 2 in 58% yield. Condensation of 2 with excess pyrrole and catalytic trifluoroacetic acid affords dipyrromethane 3 in 19% yield. Subsequent chlorination followed by oxidation with p-chloranil gives the dichloro dipyrromethene 4 in 44% yield over two steps; because isolation of the putative dichloro dipyrromethane intermediate is low yielding, we found that a halogenation/oxidation protocol was more synthetically tractable. Boron insertion with BF3•OEt2 generates the dichloro BODIPY 5 in 75% yield. Nucleophilic displacement with sodium methoxide yields 6, which is then coupled to the azatetrathia receptor 7 to generate 8 in 56% yield. Ester hydrolysis of 8 under basic conditions yields Carboxy-CS1 9. Peptide coupling of 9 and the mitochondrial-targeted tag 10 with HATU delivers Mito-CS1 (11) in 48% yield. We note that these mild and robust coupling conditions allow for the potential introduction of other functional tags onto Carboxy-CS1, including ones that can be used to direct fluorescent copper sensors to other organelles and other subcellular targets and that the synthetic pathway towards Mito-CS1 generates a versatile set of precursor materials (compounds 3, 5, 6, and 7).</p><!><p>The spectroscopic properties of Mito-CS1 were evaluated in aqueous media buffered to physiological pH (PBS, pH 7.4). Apo Mito-CS1 features one prominent optical band at 555 nm (2.8 × 104 M−1 cm−1) with a shoulder at 520 nm and a corresponding emission maximum at 569 nm with weak fluorescence (Φ = 0.009). Upon addition of Cu+, the absorption spectrum of Cu+-bound Mito-CS1 displays a single major visible absorption band at 550 nm (2.6 × 104 M−1 cm−1). The fluorescence intensity of Mito-CS1 increases by ca. 10-fold (Φ = 0.05, Figure 1a) with a slight blue shift of the emission maximum to 558 nm with 1 equivalent of Cu+ added (Figure 1a, inset), similar in photochemical properties to the previously reported turn-on fluorescent sensor CS160 (Cu+-bound Φ = 0.13, 10-fold turn-on response) but with the added functionality of a targeting phosphonium moiety. Binding analysis using the method of continuous variations (Job's plot) indicates a 1:1 Cu+:dye complex is responsible for the observed fluorescence enhancement (Figure 1b). These results demonstrate that Mito-CS1 can dynamically respond to changes in Cu+ levels in aqueous media. The apparent Kd for the Mito-CS1:Cu+ complex is 7.2(3) × 10−12 M in PBS buffer at pH 7.4 (Figure 2a). Even in the presence of a lipophilic phosphonium cation, Mito-CS1 maintains its high selectivity for Cu+ over other biologically relevant metal ions (Figure 2b). The fluorescence response of apo or Cu+-bound Mito-CS1 is not affected by the presence of physiologically relevant concentrations of Ca2+, Mg2+, and Zn2+. Moreover, other bioavailable divalent metal ions (Mn2+, Fe2+, Co2+, Ni2+, Cu2+) do not induce a change in the emission intensity of the apo probe and do not interfere with the Cu+ response. Finally, Mito-CS1 is selective for Cu+ over Cu2+, showing that this probe has metal and redox specificity.</p><!><p>With spectroscopic data establishing that Mito-CS1 can selectively respond to Cu+ in aqueous solution, we turned our attention to evaluate Mito-CS1 in live-cell imaging assays using HEK 293T as a model cell line. First, we tested Mito-CS1 for its ability to localize to mitochondria. Accordingly, HEK 293T cells stained with 500 nM Mito-CS1 for 15 min at 37 °C show measurable levels of fluorescence in discrete subcellular locations as determined by confocal microscopy (Figure 3a, Figure S1a). Co-staining experiments with MitoTracker Deep Red, a commercially available mitochondrial tracker (Figure 3b), BODIPY FL C5-ceramide, a marker for the trans-Golgi (Figure 3c), and LysoTracker Green DND-26, a lysosomal marker (Figure S1b), establish that the observed fluorescence from Mito-CS1 is localized to mitochondria in these live cells. Furthermore, nuclear staining with Hoechst 33342 indicates that the cells are viable throughout the imaging experiments (Figure 3d, Figure S1c).</p><p>We next tested whether Mito-CS1 was sensitive enough to detect basal levels of labile mitochondrial Cu+ and whether it could respond to increases and/or decreases in the size of this pool. First, HEK 293T cells were cultured in growth media only or growth media supplemented with 300 μM CuCl2 for 18 h to globally elevate intracellular copper stores and subsequently imaged by confocal microscopy (Figure 4a, 4b). As visualized by Mito-CS1, we observe that exchangeable mitochondrial Cu+ levels rise by 34% with copper supplementation relative to control, indicating that Mito-CS1 can detect expansions in the mitochondrial Cu+ pool (Figure 4d).31 To evaluate whether Mito-CS1 can also report on decreases in exchangeable mitochondrial Cu+ levels, HEK 293T cells were cultured in growth media only or growth media supplemented with 100 μM of the membrane-impermeable Cu+ chelator bathocuproine disulfonate (BCS) for 18 h to globally deplete intracellular copper stores,87 and each set of cells were then stained with Mito-CS1 and imaged by confocal microscopy. Upon addition of BCS, labile mitochondrial Cu+ levels decrease by 36% relative to basal Cu+ levels (Figure 4a, 4c), showing that Mito-CS1 can image basal levels of mitochondrial Cu+ and detect depletions in this pool. In addition, nuclear staining with Hoechst 33342 confirms that the viability of HEK 293T cells is unaffected by the manipulation of cellular copper status (Figure 4a, 4b, 4c).</p><p>Finally, as an additional set of controls to corroborate that Mito-CS1 is detecting copper-dependent events, we performed analogous imaging studies with Rhodamine 123, a well established fluorescent marker for assaying mitochondrial membrane potential 88,89, under basal, copper-supplemented, and copper-depleted conditions. As described above, endogenous copper pools of HEK 293T cells were elevated with supplementation or depleted with chelation and then stained with 100 nM Rhodamine 123 (Figure 4e, 4f, 4g). The membrane potential as measured by Rhodamine 123 does not show statistically significant changes with either 300 μM CuCl2 or 100 μM BCS treatment relative to control HEK 293T cells (Figure 4h), providing further evidence that Mito-CS1 is sensing changes in labile mitochondrial copper pools rather than reporting on mitochondrial membrane potential.</p><!><p>In our first application of Mito-CS1 as an analytical tool, we investigated whether the labile and total mitochondrial copper pools are affected by mutations in SCO1 and SCO2, which are known to cause a severe copper deficiency at the cellular level in affected tissues and cell types.36 We first evaluated the ability of Mito-CS1 to localize to mitochondria in human fibroblasts. As seen with HEK 293T cells, confocal microscopy of control fibroblasts stained with 5 μM Mito-CS1 for 15 min at 37 °C shows measurable levels of fluorescence in discrete subcellular locations (Figure 5a, Figure S2a). Co-staining experiments with MitoTracker Deep Red, a commercially available mitochondrial tracker (Figure 5b), BODIPY FL C5-ceramide, a marker for the trans-Golgi (Figure 5c), and LysoTracker Green DND-26, a lysosomal maker (Figure S2b), establish that the observed fluorescence from Mito-CS1 is localized to mitochondria of these cells. Furthermore, nuclear staining with Hoechst 33342 indicates that the cells are viable throughout the imaging experiments (Figure 5d, Figure S2c).</p><p>Analogous to the HEK 293T cells, we next tested how changes in global copper status affect the exchangeable mitochondrial copper pool in control fibroblasts. Control fibroblasts were grown in media supplemented with 300 μM and 500 μM CuCl2 for 18 h and then stained with 5 μM Mito-CS1 for 15 min at 37 °C. The labile mitochondrial Cu+ pool as visualized by Mito-CS1 is significantly expanded upon co-culture of fibroblasts with 500 μM CuCl2 (Figure S3a, S3b, S3c). Staining with Rhodamine 123 does not show a statistically significant difference in the membrane potential between untreated and copper-supplemented fibroblasts (Figure S3d, S3e, S3f). Control fibroblasts treated with 100 μM BCS for 12 h and then stained with 5 μM Mito-CS1 for 15 min at 37 °C show a 45% decrease in mitochondrially-localized fluorescence intensity compared to their untreated counterparts (Figure 6a, 6b, 6c). The observed decrease in fluoresence intensity of Mito-CS1 in BCS-treated fibroblasts demonstrates that these cells possess a dynamic mitochondrial Cu+ pool that is affected by changes in global copper status. In a set of parallel control experiments, Rhodamine 123 staining of control fibroblasts grown in normal media or media with 100 μM BCS for 12 h reveals no differences in membrane potential (Figure 6d, 6e, 6f) and nuclear staining with Hoescht 33342 shows that the cells are viable throughout the BCS treatment and imaging experiments (Figure 6a, 6b, 6c, 6d). COX activity is mildly reduced with 100 μM BCS treatment for 24 h (Figure 6g) but not sufficiently to affect the membrane potential (Figure 6e). Taken together, the data further establish that Mito-CS1 directly detects and responds to changes in endogenous mitochondrial Cu+ levels in cultured human fibroblasts.</p><p>After establishing that Mito-CS1 can monitor the labile mitochondrial Cu+ pool in control fibroblasts and sense dynamic changes in its size with alterations in copper status, we used this new chemical tool to characterize mitochondrial Cu+ homeostasis in fibroblasts derived from patients with mutations in SCO1 and SCO2. We also included fibroblasts derived from a patient with mutations in the copper exporter protein ATP7A in our analyses, for loss of ATP7A function in this cell type results in significant increases in total cellular copper content.90,91 Control, SCO1, SCO2, and ATP7A patient fibroblasts were cultured for 48 h in growth medium, stained with 5 μM Mito-CS1 in DMEM, and imaged live by scanning confocal microscopy. Consistent with our findings using control fibroblasts, mutations in ATP7A, SCO1 and SCO2 do not affect the mitochondrial localization of Mito-CS1 or overall cell viability (Figure 7a, 7b, 7c, 7d). Interestingly, quantification of the Mito-CS1 fluorescence intensities reveals that the size of the labile mitochondrial Cu+ pool is comparable in control, SCO1 and SCO2 patient fibroblasts, while it is significantly expanded in ATP7A patient fibroblasts (Figure 7e). We next used ICP-OES to measure total Cu at the whole cell and mitochondrial levels of organization in all four genetic backgrounds (Figure 7f). As expected, total cellular copper content is elevated in ATP7A patient fibroblasts and decreased in SCO1 and SCO2 patient fibroblasts when compared to control fibroblasts. In contrast, the total mitochondrial copper pool is only modestly altered in the patient backgrounds (Figure 7g), illustrating that this organelle more tightly controls its total and labile copper pools relative to the whole cell. Because the majority of mitochondrial copper is housed in the matrix and previous studies with competitor proteins, in vitro titrations and targeted chelators31,69 argue that this pool is exchangeable, Mito-CS1 is likely visualizing this store within living cells. We performed two sets of additional control experiments to further support that Mito-CS1 is specifically reporting on a Cu+-sensitive phenomenon in these patient fibroblasts. First, we used Rhodamine 123 staining to demonstrate that the mitochondrial membrane potential is maintained in both SCO1 and SCO2 patient backgrounds relative to the control (Figure 8a, 8b, 8c, 8d, 8e) Second, we measured COX activity in fibroblasts from all four genetic backgrounds (Figure 8f) to confirm the presence of a severe COX deficiency in fibroblasts lacking functional SCO1 or SCO2. The collective data show that the observed changes in Mito-CS1 fluorescence are copper-dependent and are not an indirect consequence or non-specific phenomenon of differences across the cell lines in COX activity and/or membrane potential.</p><p>Finally, to support that the data obtained from SCO1 and SCO2 patient fibroblasts are representative of the observed in vivo phenotypes, we measured total copper content in crude mitochondria isolated from control, SCO1 and SCO2 patient livers by ICP-OES (Figure S4). As in cultured fibroblasts, the data show that the mitochondrial copper pool is not significantly altered in liver as a result of mutations in SCO1 or SCO2 (Figure S3a). While there was considerable variability in mitochondrial copper content across control liver samples, this observation can be explained by differences in total hepatocyte copper content (Figure S3b). Comparable copper levels in crude and highly purified mitochondria isolated from HEK 293T cells further suggest that the variation in mitochondrial copper levels observed across control liver samples is not simply a consequence of the isolation procedure (Figure S3a). Therefore, consistent with previously reported results36 these data further suggest that fibroblasts are a competent disease model in terms of copper misregulation. Taken together, the combination of Mito-CS1 imaging and ICP-OES measurements patently shows that the mitochondrial copper store is a tightly regulated metal ion pool, as we observe little to no perturbation in the labile mitochondrial Cu+ and total mitochondrial copper pools in ATP7A, SCO1, and SCO2 patient fibroblasts compared to control congeners.</p><!><p>In this report, we have described the synthesis, properties, and biological applications of Mito-CS1, a new targetable fluorescent probe that can selectively detect labile Cu+ in mitochondria of living cells. Mito-CS1 is a unique Cu+-specific small-molecule fluorescent indicator that features visible excitation and emission profiles, a turn-on response, and selectivity for Cu+ over other abundant mitochondrial metal ions, including Fe2+, Cu2+, and Zn2+. Confocal microscopy experiments with a model HEK 293T cell line and human fibroblasts establish that Mito-CS1 is chemically and spatially specific within living cells for mitochondrial Cu+. Furthermore, we used Mito-CS1 in conjunction with biochemical and ICP metal analyses to monitor mitochondrial copper pools in SCO and ATP7A patient fibroblasts, cell lines that exhibit profound alterations in total cellular copper levels that mirror those observed in affected patient tissues in vivo. Interestingly, these experiments show that the mutations in question do not dramatically alter labile mitochondrial Cu+ or total mitochondrial Cu pools relative to control cells, suggesting that cells maintain homeostatic control over mitochondria regulating copper homeostasis within a narrow window, to protect the main oxygen-consuming and energy-producing organelle relative to other areas of the cell. We anticipate that the Mito-CS1 reagent will find utility for future interrogations of how discrete copper handling pathways are organized into dynamic networks at the cell and systems levels of organization. In addition to applying Mito-CS1 for studies of mitochondrial copper biology, we are exploiting the modular probe scaffold to create new multifunctional probes for detecting Cu+ and other biologically relevant analytes in discrete subcellular locales.</p><!><p>All reactions were carried out under a dry nitrogen atmosphere with flame-dried glassware. Silica gel P60 (SiliCycle) was used for column chromatography. Analytical thin layer chromatography was performed using SiliCycle 60 F254 silica gel (precoated sheets, 0.25 mm thick). Compound 192, compound 1093, and Rhodamine 12394 were synthesized according to modified literature procedures. Compound 760,61 was prepared according to previously reported procedures. HATU was purchased from ChemPep Incorporated (Wellington, FL). MitoTracker Deep Red, BODIPY FL C5-ceramide, LystoTracker DND-26, and Hoechst 33342 were purchased from Invitrogen (Carlsbad, CA). All other chemicals were purchased from Sigma-Aldrich (St. Louis, MO) and were used as received. 1H and 13C NMR spectra were collected in CDCl3, CD3OD, and (CD3)2SO (Cambridge Isotope Laboratories, Cambridge, MA) at 25 °C on a Bruker AVB-400 spectrometer at the College of Chemistry NMR Facility at the University of California, Berkeley. All chemical shifts are reported in the standard notation of parts per million using the peak of residual proton signals of CDCl3 as an internal reference. Low-resolution mass spectral analyses were carried out using a 6130 quadrupole LC/MS 1200 Series (Aglient Technologies, Santa Clara, CA). High-resolution mass spectral analyses were carried out at the College of Chemistry Mass Spectrometry Facility at the University of California, Berkeley.</p><!><p>Compound 1 was prepared according to a previously reported procedure with the following modifications: CH3Cl3 was passed over alumina and then added dropwise over 30 min to the reaction mixture. Characterization was consistent with that previously reported in the literature.92</p><!><p>Methyl 2-bromoacetate (3.8 mL, 40 mmol) in anhydrous MeCN (60 mL) was added dropwise to a suspension of compound 1 (5.0 g, 33.3 mmol) and K2CO3 (9.19 g, 66.6 mmol) in MeCN (50 mL) at 60 °C. The resulting solution was heated at 60 °C for two days. The reaction mixture was cooled to room temperature and concentrated in vacuo. The crude reaction mixture was partitioned between water (100 mL) and EtOAc (3 × 100 mL). The organics were combined, dried over Na2SO4, filtered, and dried in vacuo. Diethyl ether (50 mL) was added to the crude solid and the resulting solid was collected via vacuum filtration and washed with diethyl ether (50 mL) to furnish compound 2 as an off-white solid (4.31 g, 58%). The reaction was repeated to generate more material for the subsequent reactions. 1H NMR (400 MHz, CDCl3): δ 2.59 (6H, s), 3.81 (3H, s), 6.58 (2H, s), 10.46 (1H, s). 13C NMR (100 MHz, CDCl3): δ 21.1, 52.5, 64.8, 115.3, 126.9, 144.6, 160.7, 168.8, 191.7. LRESI-MS calculated for [MH+] 223.1, found 223.1.</p><!><p>A solution of 2 (6.38 g, 28.7 mmol) in freshly distilled pyrrole (50.0 mL, 721 mmol) was wrapped in foil to protect it from light and purged with a stream of N2 for five minutes. TFA (240 μL, 5.4 mmol) was then added dropwise, and the solution was stirred at room temperature for 1 h. After 1 h, TEA (1.0 mL, 7.2 mmol) was added and the reaction mixture continued to stir for 15 min. The reaction mixture was poured into toluene (150 mL) and washed with brine (2 × 100 mL). The organics were combined, dried over Na2SO4, filtered, concentrated to dryness and purified by flash chromatography two times (1st column: silica CHCl3 to 0.1% MeOH in CHCl3, 2nd column: silica, 0.5% EtOAc in CH2Cl2) to give a crude solid. Diethyl ether was added (3 × 20 mL), decanted, and the solid was collected via vacuum filtration and washed with diethyl ether (40 mL) to yield 3 as an off-white solid (1.81 g, 19%). 1H NMR (400 MHz, CDCl3): δ 2.08 (6H, s), 3.83 (3H, s), 4.62 (2H, s), 5.89 (1H, s), 5.99 (2H, s), 6.18 (2H, q, J = 2.8 Hz), 6.60 (2H, s), 6.67 (2H, br), 7.97 (2H, s). 13C NMR (100 MHz, CDCl3): δ 21.0, 38.1, 52.3, 65.1, 106.6, 108.7, 115.4, 116.3, 131.1, 131.2, 139.5, 156.3, 169.6. LRESI-MS calculated for [MH+] 339.2, found 339.2.</p><!><p>A solution of 3 (1.76 g, 5.2 mmol) in anhydrous THF (67 mL) was wrapped in foil to protect it from light and cooled to −78 °C. N-Chlorosuccinimide (1.53 g, 11.5 mmol) in anhydrous THF (20 mL) was wrapped in foil to protect it from light and was added dropwise in two portions via an addition funnel over the course of 15 min. The reaction was stirred for an additional 2 h at −78 °C, capped, and then placed in the freezer (−20 °C) overnight. The crude reaction mixture was poured into water (100 mL) at 0 °C and then extracted with CH2Cl2 (3 × 100 mL). The organics were combined, dried over Na2SO4, filtered, concentrated to dryness, and purified by flash chromatography (1st column: silica, 5% hexanes in CH2Cl2 to 5% MeOH in CH2Cl2, 2nd column: silica, 20% hexanes in CH2Cl2 to CH2Cl2) to provide the dichloro dipyrromethane as a crude red solid. This was carried onto the next step without further purification. p-Chloranil (1.30 g, 5.3 mmol) was added to a solution of the crude intermediate (1.05 g, 2.6 mmol) in CH2Cl2 (25 mL) and the reaction mixture continued to stir overnight at room temperature. The red-orange reaction mixture was concentrated to dryness and the residue was purified by flash chromatography (silica, 20% hexanes in CH2Cl2) afforded 5 as a dark orange solid (924 mg, 44% over two steps). 1H NMR (400 MHz, CDCl3): δ 2.08 (6H, s), 3.84 (3H, s), 4.67 (2H, s), 6.18 (2H, d, J = 4 Hz), 6.31 (2H, d, J = 4.4 Hz), 6.64 (2H, s). 13C NMR (100 MHz, CDCl3): δ 20.3, 52.4, 65.2, 113.3, 117.2, 128.1, 128.7, 138.2, 138.6, 138.9, 141.5, 157.6, 169.5. LRESI-MS calculated for [MH+] 405.1, found 405.1.</p><!><p>Distilled DIEA (1.78 mL, 10.2 mmol) was added dropwise to a solution of 4 (924 mg, 2.3 mmol) in anhydrous CH2Cl2 (25 mL) and the resulting solution was allowed to stir for an additional 20 min. BF3•OEt2 (2.6 mL, 20.5 mmol) was then added dropwise over a period of 5 min, and the resulting solution was allowed to stir overnight. The reaction mixture was quenched with water (100 mL) and then extracted with CH2Cl2 (3 × 25 mL). The organics were combined, dried over Na2SO4, filtered, and dried in vacuo. Purification by flash chromatography (silica, 10% hexanes in CH2Cl2) gave 5 as a red-orange solid with a green luster (771 mg, 75%). 1H NMR (400 MHz, CDCl3): δ 2.06 (6H, s), 3.80 (3H, s), 4.65 (2H, s), 6.34 (2H, d, J = 4 Hz), 6.58 (2H, d, J = 4.4 Hz), 6.65 (2H, s). 13C NMR (100 MHz, CDCl3): δ 20.2, 52.3, 64.9, 113.6, 119.0, 123.3, 130.3, 134.2, 138.5, 143.1, 144.8, 158.2, 169.1. LRESI-MS calculated for [MH+] 453.1, found 453.0.</p><!><p>A solution of 5 (771 mg, 17 mmol) in anhydrous THF (40 mL) was cooled to 0 °C and 25 wt% sodium methoxide (398 mg, 18.4 mmol) in anhydrous MeOH (90 mL) was added dropwise in two portions over 2.5 h. The reaction mixture was maintained at 0 °C for 4 h until the reaction was complete by TLC analysis. The solvent was removed in vacuo, and the resulting solid was partitioned between water (50 mL) and CH2Cl2 (3 × 50 mL). The organics were combined, dried over Na2SO4, filtered, and dried in vacuo. Purification by flash chromatography (silica, CH2Cl2) delivered 6 as a red-orange solid (640 mg, 84%). 1H NMR (400 MHz, CDCl3): δ 2.08 (6H, s), 3.83 (3H, s), 4.13 (3H, s), 4.65 (2H, s), 6.10 (1H, d, J = 4.8 Hz), 6.20 (1H, d, J = 4.0 Hz), 6.31 (1H, d, J = 4 Hz), 6.65 (2H, s), 6.70 (1H, d, J = 4.4 Hz). 13C NMR (100 MHz, CDCl3): δ 20.3, 52.4, 59.3, 65.1, 104.9, 113.5, 115.5, 124.8, 125.2, 130.7, 132.4, 133.9, 137.0, 139.0, 139.2, 158.0, 169.3, 169.6. LRESI-MS calculated for [MH+] 449.1, found 449.1.</p><!><p>Anhydrous MeCN (10 mL) was added to 6 (185 mg, 413 μmol) and 7 (575 mg, 1.8 mmol). The reaction mixture was purged with a stream of N2 for five minutes and continued to stir at 45 °C in the dark for 72 h. The reaction mixture was concentrated to dryness. Purification by column chromatography (silica, 0.5% toluene in CH2Cl2 to 0.25% MeOH in CH2Cl2 to 0.5% MeOH in CH2Cl2) provided 8 as a red oil (168 mg, 56%). 1H NMR (400 MHz, CDCl3): δ 1.20 (6H, t, J = 7.4 Hz), 2.05 (6H, s), 2.53 (4H, q, J = 7.4 Hz), 2.70—2.74 (4H, m), 2.77—2.81 (4H, m), 2.86 (4H, t, J = 7.2 Hz), 3.79 (3H, s), 3.89 (4H, t, J = 7.4 Hz), 3.93 (3H, s), 4.62 (2H, s), 5.60 (1H, d, J = 4 Hz), 5.94 (1H, d, J = 4.8 Hz), 6.09 (1H, d, J = 3.6 Hz), 6.41 (1H, d, J = 4.8 Hz), 6.60 (2H, s). 13C NMR (100 MHz, CDCl3): δ 14.8, 20.2, 25.6, 30.1, 31.8, 32.2, 52.3, 53.3, 58.0, 65.0, 94.9, 110.0, 113.1, 121.7, 125.5, 126.8, 131.5, 133.0, 139.3, 157.3, 159.7, 161.6, 169.4. LRESI-MS calculated for [MH+] 726.3, found 726.3.</p><!><p>Powdered LiOH (20 mg, 833 μmol) was added to a solution of 8 (168 mg, 231 μmol) in anhydrous MeOH (3 mL) and anhydrous THF (3 mL) and the reaction mixture continued to stir overnight at room temperature. The reaction mixture was concentrated in vacuo, dissolved in water (15 mL), and the pH was adjusted to pH 1—2 with 0.5 M HCl. The reaction mixture was extracted with CHCl3 (2 × 25 mL), the organics were combined, dried over Na2SO4, filtered, and dried in vacuo. Purification by column chromatography (silica, 1.5% MeOH in CH2Cl2 yielded 9 as a red oil (98 mg, 60%). 1H NMR (400 MHz, CDCl3): δ 1.24 (6H, t, J = 7.2 Hz), 2.10 (6H, s), 2.57 (4H, q, J = 7.2 Hz), 2.73—2.78 (4H, m), 2.81—2.85 (4H, m), 2.89 (4H, t, J = 7.2 Hz, J = 7.6 Hz), 3.93 (4H, t, J = 7.4 Hz), 3.97 (3H, s), 4.71 (2H, s), 5.64 (1H, d, J = 4.4 Hz), 5.96 (1H, d, J = 4.8 Hz), 6.13 (1H, d, J = 4 Hz), 6.45 (1H, d, J = 4.8 Hz), 6.66 (2H, s). 13C NMR (100 MHz, CDCl3): δ 14.8, 20.3, 26.0, 30.3, 31.7, 31.9, 32.2, 32.5, 53.3, 58.1, 64.7, 95.0, 110.1, 113.2, 121.8, 125.6, 127.2, 131.4, 131.6, 133.0, 139.6, 157.0, 159.8, 161.7, 173.6. LRESI-MS calculated for [MH+] 712.2, found 712.2.</p><!><p>Compound 10 was prepared according to a previously reported procedure with the following modifications: the crude solid was triturated with diethyl ether as reported but it was not further recrystallized from ethanol/diethyl ether as reported. The characterization data were consistent with results previously reported in the literature.93</p><!><p>Compound 9 (98 mg, 137 μmol) and HATU (59 mg, 155 μmol) were dissolved in anhydrous CH2Cl2 (6 mL) and purged with a stream of N2 for five minutes. After 20 min, 10 (59 mg, 153 μmol) was added in one portion and the reaction mixture continued to stir for 20 min. Distilled DIEA was added after 20 min and the reaction stirred at room temperature overnight. The reaction mixture was concentrated in vacuo and purified by column chromatography (1st column: silica, 0.25% EtOAc and 0.25% MeOH in CH2Cl2, 2nd column: silica, 0.5% EtOAc and 0. 5% MeOH in CH2Cl2, 3rd column: silica, 0.25% EtOAc and 0.25% MeOH in CH2Cl2 to 1% MeOH in CH2Cl2) to furnish 11 as a red semi-solid (67 mg, 48%). 1H NMR (400 MHz, CDCl3): δ 1.23 (6H, t, J = 7.2 Hz, J = 7.6 Hz), 2.07 (6H, s), 2.56 (4H, q, J = 7.4 Hz), 2.72—2.76 (4H, m), 2.80—2.84 (4H, m), 2.88 (4H, t, J = 7.4 Hz), 3.52 (2H, p, J = 7.2 Hz), 3.74 (2H, p, J = 6.4 Hz, J = 7.2 Hz), 3.91 (4H, t, J = 7.4 Hz), 3.95 (3H, s), 4.34 (2H, s), 5.62 (1H, d, J = 4 Hz), 5.95 (1H, d, J = 4.8 Hz), 6.11 (1H, d, 4 Hz), 6.45 (1H, d, J = 4.8 Hz), 6.71 (2H, s), 7.67—7.83 (15H, m). 13C NMR (100 MHz, CDCl3): δ 14.8, 20.2, 26.0, 30.3, 31.9, 32.4, 33.3, 53.3, 58.0, 66.6 95.0, 110.1, 113.4, 117.1, 117.9, 121.9, 125.7, 126.9, 130.7, 130.8, 133.4, 133.5, 135.5, 139.3, 156.9, 159.7, 169.5. HRESI-MS calculated for [M+] 999.3576, found 999.3585.</p><!><p>Rhodamine 123 was prepared according to a previously reported procedure with the following modifications: the crude solid was triturated with chloroform and the resulting solid was collected via vacuum filtration and washed with chloroform. Characterization was consistent with that previously reported in the literature.94</p><!><p>Millipore water was used to prepare all aqueous solutions. All spectroscopic measurements were performed in phosphate buffered saline (1X PBS, pH 7.4, Invitrogen, Carlsbad, CA). Absorption spectra were recorded on a Varian Cary 50 spectrophotometer (Walnut Creek, CA) and fluorescence spectra were recorded using a Photon Technology International Quanta Master 4 L-format scanning spectrofluorometer (Lawrenceville, NJ) equipped with an LPS-220B 75-W xenon lamp and power supply, A-1010B lamp housing with integrated igniter, switchable 814 photon-counting/analog multiplier detection unit, and MD5020 motor driver. Samples for absorption and emission measurements were contained in 1-cm × 1-cm path length quartz cuvettes (1.4-mL volume, Starna, Atascadero, CA). Fluorescence quantum yields were determined by reference to rhodamine 101 inner salt in methanol (Φ = 1.0).95 The binding affinity of Cu+ to Mito-CS1 was measured using thiourea as a competitive ligand to provide a buffered Cu+ solution. Stability constants for thiourea binding were taken from the literature b12 = 2.0 × 1012, b13 = 2.0 × 1014, b14 = 3.4 × 1015.96. Thiourea was delivered from an aqueous stock solution (200 mM, 500 mM). Cu+ was delivered in the form of [Cu(MeCN)4[PF6]] from an acetonitrile stock solution (1 mM, 5 mM). Excitation was provided at 540 nm and collected emission was integrated from 550 to 700 nm after 1 min or 2 min after the addition of dye or metal analyte. The apparent dissociation constant (Kd) was determined using the following equation: (F − Fmin)/(Fmax − Fmin) = [Cu+]/(Kd + [Cu+]), where F is the observed fluorescence, Fmax is the fluoresence for the Cu+:Mito-CS1 complex, Fmin is the fluorescence for free Mito-CS1, and [Cu+] is the 'free' Cu+ available for complexation, which was calculated using the stability constants for thiourea and standard competition equilibrium expressions. All other metal ions tested for metal ion selectivity studies with the exception of Fe2+ were from their chloride salts as aqueous solutions. Ammonium iron(II) sulfate hexahydrate was used as a source of Fe2+ and the salt was dissolved in degassed water.</p><!><p>Cells were grown in the Tissue Culture Facility at the University of California, Berkeley with expert technical assistance from Ann Fischer and Michelle Yasukawa. HEK 293T cells were cultured in Dulbecco's Modified Eagle Medium (DMEM, Invitrogen, Carlsbad, CA) containing high glucose and without phenol red supplemented with GlutaMAX (Invitrogen, Carlsbad, CA) and 10% Fetal Bovine Serum (FBS, Hyclone, Logan, UT). Two days before imaging, cells were passed and plated on 12-mm glass coverslips coated with poly-L-lysine (50 mg/mL, Sigma, St. Louis, MO). Control, ATP7A, SCO1, and SCO2 fibroblasts were cultured in Dulbecco's Modified Eagle Medium (DMEM, Invitrogen, Carlsbad, CA) containing high glucose and without phenol red supplemented with GlutaMAX (Invitrogen, Carlsbad, CA), 1 mM sodium pyruvate, and 10% Fetal Bovine Serum (FBS, Hyclone, Logan, UT). Two days before imaging, cells were passed and plated on 12-mm glass coverslips or 4-well chambered coverglass slides.</p><!><p>For all experiments, solutions of Mito-CS1 (from 1 mM or 5 mM stocks in DMSO), CS3 (from 1 mM stocks in DMSO), Rhodamine 123 (from 50 μM stocks in DMSO), MitoTracker Deep Red (from a 50 μM stocks in DMSO), BODIPY FL C5-ceramide-BSA complex (from 1 mM stocks in water), and Hoechst 33342 (from a 5 mM stock in DMSO) were made in Dulbecco's Phosphate Buffered Saline with calcium chloride and magnesium chloride (DPBS, Invitrogen, Carlsbad, CA) or Dulbecco's Modified Eagle Medium without phenol red (DMEM, Invitrogen, Carlsbad, CA). For colocalization experiments control HEK 293T cells were incubated with 500 nM Mito-CS1 2.25 μM BODIPY FL C5-ceramide-BSA complex, 50 nM Mitotracker Deep Red, and 5 μM Hoechst 33342 for 15 min at 37 °C, 5% CO2 in DPBS. Coverslips were then transferred to fresh DPBS for imaging. A similar procedure was followed for imaging with 500 nM Lysotracker Green DND-26. For copper and BCS treatments, 300 μM CuCl2 or 100 μM BCS was added to the cells from a 0.1 mM aqueous stock solution one day prior to imaging. Cells were then incubated at 37 °C, 5% CO2. After 18 hours, the media was exchanged for DPBS with 500 nM Mito-CS1 or 100 nM Rhodamine 123 and Hoechst 33342 and incubated for 15 min at 37 °C, 5% CO2. Coverslips were then transferred to fresh DPBS for imaging. For colocalization experiments control fibroblasts were incubated with 5 μM Mito-CS1, 2.25 μM BODIPY FL C5-ceramide-BSA complex, and 5 μM Hoechst 33342 for 15 min at 37 °C, 5% CO2 in DPBS. A similar procedure was followed for imaging with 250 nm Lysotracker Green DND-26 in DMEM. Coverslips were then transferred to fresh DPBS for imaging. MitoTracker Deep Red (50 nM in DPBS) was added on stage and then imaged. For BCS treatment, 100 μM BCS was added to the cells from a 0.1 mM aqueous stock solution one day prior to imaging. After 12 hours, the media was exchanged for DMEM with 5 μM Mito-CS1 or 100 nM Rhodamine 123 and 5 μM Hoechst 3342 and incubated for 15 min at 37 °C, 5% CO2. Coverslips were then transferred to fresh DPBS for imaging. Similar procedures were followed for imaging changes in mitochondrial Cu+ in the wild type, ATP7A, SCO1, and SCO2 fibroblasts. For copper treatment, 300 μM or 500 μM CuCl2 was added to the cells from a 0.1 mM aqueous stock solution one day prior to imaging. After 18 hours, the media in the chambered coverglass slides was exchanged for DMEM with 5 μM Mito-CS1 or 100 nM Rhodamine 123 and 5 μM Hoechst 3342 and incubated for 15 min at 37 °C, 5% CO2, washed with fresh DPBS, and imaged in fresh DPBS. Control and ATP7A patient fibroblasts were stained with 2 μM CS3 and incubated for 10 min at 37 °C, 5% CO2 in DPBS, washed with fresh DPBS, and imaged in fresh DPBS.</p><!><p>Confocal fluorescence images were acquired at the Molecular Imaging Center at the University of California, Berkeley. Imaging experiments were performed with a Zeiss LSM510 META NLO Axioplan 2 laser-scanning microscope, a Zeiss 510NL META AxioIMAGER laser-scanning microscope, and a Zeiss LSM 710 laser-scanning microscope with a 40x or 63x water-immersion objective lens. Excitation of Mito-CS1 or CS3 loaded cells at 543 nm was carried out with a HeNe laser, and emission was collected using a META detector between 554—650 nm. Excitation of Lysotracker Green DND-26 at 488 nm was carried out with an Ar laser, and emission was collected using a META detector between 501—533 nm. Excitation of BODIPY FL C5-ceramide-BSA complex at 488 nm was carried out with an Ar laser, and emission was collected using a META detector between 498—511 nm. Excitation of Rhodamine 123 at 488 nm was carried out with an Ar laser, and emission was collected using a META detector between 501—576 nm. Excitation of MitoTracker Deep Red at 633 nm was carried out with a HeNe laser, and emission was collected using a META detector between 640—704 nm. Excitation of Hoechst 33342 was carried out using a MaiTai two photon laser at 780-nm pulses or a 405 nm diode laser and emission was collected between 469—522 nm with Rhodamine 123 and 447—533 nm for all other dyes. ImageJ from the National Institutes of Health was used for analysis of the images. Specifically, the threshold for a field of cells was adjusted to select the pixels and was kept consistent in a given experiment. The selected pixels were analyzed for the median value. The mean of the median value for n fields of cells with standard error is reported and statistical analyses were performed with a two-tailed Student's t-test in Microsoft Excel.</p><!><p>Mitochondria and whole cells were digested in 40% nitric acid by boiling for 1 hr in capped, acid-washed tubes; samples were then diluted in ultra-pure, metal-free water; and analyzed by ICP-OES (PerkinElmer, Optima 3100XL). Acid-washed blanks were used as controls. Concentrations were determined from a standard curve constructed with serial dilutions of commercially available mixed metal standards (Optima). Error bars cannot be added to the data points for mitochondria isolated from SCO1 and SCO2 patient liver, owing to the scarcity of the material; the total tissue sample only consists of milligram quantities originating from a metabolic autopsy, thus precluding additional, large-scale analyses. Furthermore, we did not have access to liver samples from deceased ATP7A patients. To minimize any potential effects associated with experimental variability, mitochondria were isolated on the same day using identical buffers and reagents from each patient sample. Each sample was then split into two aliquots, and total mitochondrial copper content quantified in duplicate by ICP-OES in the same run.</p><!><p>COX and citrate synthase activities were measured in HEK 293T and fibroblast cell extracts as described elsewhere.32,97 Protein concentration was measured by the Bradford method.98</p>

PubMed Author Manuscript

Exposure–response relationship of ramucirumab in patients with advanced second-line colorectal cancer: exploratory analysis of the RAISE trial

PurposeTo characterize ramucirumab exposure–response relationships for efficacy and safety in patients with metastatic colorectal cancer (mCRC) using data from the RAISE study.MethodsSparse pharmacokinetic samples were collected; a population pharmacokinetic analysis was conducted. Univariate and multivariate Cox proportional hazards models analyzed the relationship between predicted ramucirumab minimum trough concentration at steady state (C min,ss) and survival. Kaplan–Meier analysis was used to evaluate survival from patients in the ramucirumab plus folinic acid, 5-fluorouracil, and irinotecan (FOLFIRI) treatment arm stratified by C min,ss quartiles (Q). An ordered categorical model analyzed the relationship between C min,ss and safety outcomes.ResultsPharmacokinetic samples from 906 patients were included in exposure–efficacy analyses; samples from 905 patients were included in exposure–safety analyses. A significant association was identified between C min,ss and overall survival (OS) and progression-free survival (PFS) (p < 0.0001 for both). This association remained significant after adjusting for baseline factors associated with OS or PFS (p < 0.0001 for both). Median OS was 11.5, 12.9, 16.4, and 16.7, and 12.4 months for ramucirumab C min,ss Q1, Q2, Q3, Q4, and placebo group, respectively. Median PFS was 5.4, 4.6, 6.8, 8.5, and 5.2 months for ramucirumab C min,ss Q1, Q2, Q3, Q4, and placebo group, respectively. The risk of Grade ≥3 neutropenia was associated with an increase in ramucirumab exposure.ConclusionsExploratory exposure–response analyses suggested a positive relationship between efficacy and ramucirumab exposure with manageable toxicities in patients from the RAISE study with mCRC over the ranges of exposures achieved by a dose of 8 mg/kg every 2 weeks in combination with FOLFIRI.Electronic supplementary materialThe online version of this article (doi:10.1007/s00280-017-3380-z) contains supplementary material, which is available to authorized users.

exposure–response_relationship_of_ramucirumab_in_patients_with_advanced_second-line_colorectal_cance

Introduction<!><!>Materials and methods<!>Exposure–efficacy analysis<!>Exposure–safety analysis<!>Results<!><!>Results<!><!>Discussion<!>

<p>Colorectal carcinoma (CRC) is the third leading cause of cancer worldwide [1] and ranks fourth among leading causes of cancer deaths worldwide [2]. Conventional systemic therapy for CRC includes fluoropyrimidine-based regimens alone or in combination with irinotecan or oxaliplatin [3–7]. The development of agents targeting the epidermal growth factor receptor (EGFR) and angiogenic pathways has provided additional treatment options. Vascular endothelial growth factor (VEGF) and VEGF receptor-2 (VEGFR-2)-mediated signaling and angiogenesis are important in CRC tumor growth and are established therapeutic targets. Ramucirumab is a human IgG1 monoclonal antibody that specifically binds to the extracellular domain of VEGFR-2 with high affinity, preventing binding of VEGF-A, C, and D ligands and receptor activation [8]. The safety and efficacy of ramucirumab in combination with second-line folinic acid, 5 fluorouracil, and irinotecan (FOLFIRI) in patients with metastatic CRC that progressed during or after first-line therapy with bevacizumab, oxaliplatin, and a fluoropyrimidine were evaluated in a randomized, double-blind, placebo-controlled phase III trial (RAISE) [9]. On Day 1 of each 2-week cycle, patients received either 8 mg/kg ramucirumab or placebo as a 60 min intravenous infusion, followed by the FOLFIRI regimen. The RAISE trial demonstrated a statistically significant survival benefit for patients treated with ramucirumab plus FOLFIRI versus placebo plus FOLFIRI with a median overall survival (OS) of 13.3 months (95% confidence interval [CI] 12.4–14.5) for patients in the ramucirumab group versus 11.7 months (95% CI 10.8–12.7) for the placebo group (hazard ratio [HR] 0.844, 95% CI 0.730–0.976; log-rank p = 0.0219). Ramucirumab plus FOLFIRI was well tolerated and the adverse events were considered manageable [9].</p><p>All drugs have dose effect curves, a threshold concentration below which they are ineffective, a concentration where effect has reached a maximum plateau and between these extremes a range where increasing exposure increases effectiveness. The 'exposure–response' phenomenon occurs in the range of concentrations (exposure) where increasing exposure correlates with increasing effect. The phenomenon of exposure–response is seen with many antibodies in the treatment of cancer, including ipilimumab in melanoma [10], trastuzumab emtansine in breast cancer [11], rituximab in the treatment of low-grade B cell malignancies, and rilotumumab in gastric cancer [12]. Analyses of the exposure–response relationship of ramucirumab in patients with gastric cancer (REGARD and RAINBOW trials) and non-small cell lung cancer (REVEL trial) have previously been reported [13, 14]. In both RAINBOW and REGARD, higher exposure to ramucirumab was associated with longer OS and progression-free survival (PFS) for gastric cancer patients. In REVEL, higher exposure to ramucirumab was associated with longer OS and PFS for non-small cell lung cancer patients.</p><p>The objective of this exploratory analysis was to determine whether there is an exposure–response relationship for ramucirumab in patients with advanced CRC enrolled in the RAISE trial.</p><!><p>RAISE study design aIrinotecan: 180 mg/m2; Folinic acid: 400 mg/m2; 5-flurouracil: 400 mg/m2 bolus followed by 400 mg/m2 given as a continuous infusion over 48 hours FOLFIRI, folinic acid, 5-fluorouracil, and irinotecan; IG immunogenicity; IV intravenous; ORR objective response rate; PFS progression-free survival; PK pharmacokinetics; PRO patient-reported outcomes; RAM ramucirumab</p><!><p>The primary endpoint was OS, defined as the time from randomization to death from any cause. Key secondary endpoints included PFS (defined as time from randomization to progressive disease or death, whichever occurred first), the proportion of patients who achieved an objective response (defined as complete response or partial response), pharmacokinetic (PK) parameters of ramucirumab, and safety.</p><p>Pre-dose and 1-h post-infusion PK samples to determine ramucirumab serum concentration were collected at the following timepoints per study protocol: Day 1 of Cycles 3, 5, 9, 13, and 17. Due to the timing of PK sample collection, only patients from both treatment arms who had non-missing concentration data and who did not die or discontinue prior to Day 1 Cycle 3 were included in the exposure–response analysis. The patients included in the exposure–response analyses are a non-random subset of the intent-to-treat (ITT) population.</p><p>Serum ramucirumab concentration was determined using a validated enzyme-linked immunosorbent assay (ELISA) at Intertek Pharmaceutical Services (San Diego, CA, USA). A population PK model was developed using a nonlinear mixed-effect modeling approach (NONMEM 7.3 [Icon, Ellicott City, MD]) and in accordance with the U.S. Food and Drug Administration (FDA) Guidance for Industry on Population Pharmacokinetics [15, 16]. Population PK model-predicted minimum concentration at steady state (C min,ss) was used to assess the exposure–response relationship.</p><!><p>The exposure–efficacy analysis was conducted using SAS® version 9.1.2 or higher. Univariate and multivariate Cox proportional hazards models were used to evaluate the relationship between ramucirumab exposure and efficacy endpoints (OS and PFS). Data for the ramucirumab plus FOLFIRI treatment arm were separated into four quartiles (Q) based on the exposure parameter of interest, C min,ss. Kaplan–Meier analyses were performed for OS and PFS with data from patients in the ramucirumab plus FOLFIRI arm stratified according to C min,ss quartile; each quartile was compared with the data from patients in the placebo plus FOLFIRI (control) arm. A multivariate Cox model adjusted for baseline covariates was used to estimate the HR for each quartile versus the control arm. Stepwise Cox regression, with entry p value = 0.05 and exit p value = 0.10, selected the baseline factors prognostic for OS or PFS. These significant factors were used to adjust the final model for evaluating exposure–efficacy relationships. An additional matched case–control analysis for OS was explored to adjust for the potential imbalances in important prognostic factors between treatment arms within each exposure quartile group [17]. In this analysis, the case groups are the four exposure quartiles of C min,ss in the ramucirumab plus FOLFIRI arm. For every patient in each case group, a matched control patient was identified from all patients receiving placebo plus FOLFIRI, through a matching scheme based on the six significant potential prognostic factors identified in the stepwise Cox regression analysis and two additional covariates with the largest baseline imbalance in the subset of patients for this analysis (sex and prior bevacizumab use—composite subgroup). The two additional factors were selected based on a stepwise logistic regression of treatment arm assignment on the same pool of covariates in the stepwise Cox regression, and stepwise selection using entry and exit p values of 0.2 (since randomization is supposed to make treatment assignment to be independent of baseline variables, a larger significance level was used to pick up any imbalanced factors due to chance). Missing values in any of the matching factors excluded the patients from the matched case–control study.</p><!><p>Ordered categorical and logistic regression models were developed to explore the relationship between ramucirumab exposure (C min,ss) and the safety endpoints. Safety endpoints for exposure–safety analysis were the three most common Grade ≥3 treatment-emergent adverse events (TEAEs) in the ITT population occurring in ≥5% of patients in the ramucirumab plus FOLFIRI arm, with a difference in incidence rate between the ramucirumab arm and the placebo arm of ≥2%. Neutropenia, hypertension, and fatigue were the selected endpoints for exposure–safety analysis based on these criteria. For this analysis, neutropenia and fatigue are consolidated terms (composite terms consisting of multiple related preferred terms) based on Standardized MedDRA® Queries (SMQ) and medical review. Diarrhea was included as an adverse event of interest as it is one of the most frequent TEAE for the FOLFIRI regimen [18, 19]. Safety endpoints were graded per the NCI-CTCAE v4.0.</p><!><p>Data from a total of 425 patients from the ramucirumab plus FOLFIRI arm and 481 patients from the placebo plus FOLFIRI arm were included in the exposure–efficacy analyses; data from a total of 425 patients from the ramucirumab plus FOLFIRI arm and 480 patients from the placebo plus FOLFIRI arm were included in the exposure–safety analyses. One patient randomized to the placebo group received ramucirumab as the first dose and was included in the placebo plus FOLFIRI arm for exposure–efficacy analysis and excluded from the placebo plus FOLFIRI arm for exposure–safety analysis.</p><p>An exploratory Cox regression analysis identified a statistically significant positive association between OS and C min,ss in a univariate analysis (p < 0.0001). A multivariate Cox regression analysis was used to adjust for the factors that were significantly associated with OS: time to progression after beginning first-line therapy, KRAS status, ECOG PS, number of metastatic sites, liver only metastasis, and carcinoembryonic antigen (CEA). After adjusting for these baseline factors, the association between OS and C min,ss remained statistically significant (p < 0.0001). Similar to OS, a statistically significant positive association was identified between PFS and C min,ss in a univariate Cox regression analysis (p < 0.0001). A multivariate Cox regression analysis was used to adjust for the factors that were significantly associated with PFS: ECOG PS, number of metastatic sites, liver only metastasis, CEA, and prior bevacizumab use (composite subgroup). A similar association between ramucirumab exposure and PFS was observed after adjusting for these significant baseline factors (p < 0.0001).</p><!><p>Exposure–response population baseline demographics and disease characteristics by ramucirumab C min,ss quartile</p><p>Note Patients in each exposure quartile group were a non-randomized subset of the ITT population and potential imbalances in prognostic factors between the placebo arm and the quartile groups may be generated due to the loss of randomization. However, the multivariate Cox regression analysis was adjusted for all prognostic factors significantly associated with OS or PFS</p><p>C min,ss minimum concentration at steady state, ECOG PS Eastern Oncology Cooperative Group performance status, FOLFIRI folinic acid, 5-fluorouracil, and irinotecan; ITT intent-to-treat, OS overall survival, PFS progression-free survival, Q quartile</p><p>a RAISE overall survival by ramucirumab Cmin,ss exposure quartile. Ramucirumab Cmin,ss quartile concentrations: Q1 (<25%) <49.7 µg/mL, Q2 (25% to <50%) 49.7 to <62.6 µg/mL, Q3 (50% to <75%) 62.6 to <81.1 µg/mL, Q4 (≥75%) ≥81.1 µg/mL. b RAISE progression-free survival by ramucirumab Cmin,ss exposure quartile. Ramucirumab Cmin,ss quartile sconcentrations: Q1 (<25%) <49.7 µg/mL, Q2 (25% to <50%) 49.7 to <62.6 µg/mL, Q3 (50% to <75%) 62.6 to <81.1 µg/mL, Q4 (≥75%) ≥81.1 µg/mL Cmin,ss, minimum concentration at steady state; PBO placebo; Q quartile; RAM ramucirumab</p><p>RAISE overall survival and progression-free survival by ramucirumab exposure quartile</p><p>C min,ss minimum concentration at steady state, ECOG PS Eastern Oncology Cooperative Group performance status, FOLFIRI folinic acid, 5-fluorouracil, and irinotecan, OS overall survival, PBO placebo, PFS progression-free survival, Q quartile, RAM ramucirumab</p><p>aAdjusted for time to progression after beginning first-line therapy, KRAS status, ECOG PS, number of metastatic sites, liver only metastasis, and carcinoembryonic antigen</p><p>bAdjusted for ECOG PS, number of metastatic sites, liver only metastasis, carcinoembryonic antigen, and prior bevacizumab use</p><p>cAdjusted for significant prognostic factors relative to PBO + FOLFIRI in RAISE</p><p>dWald's test of RAM quartile versus PBO + FOLFIRI</p><p>eMedian OS and PFS for PBO + FOLFIRI differ from those reported in Tabernero et al. [9], because patients in the PBO arm who dropped out prior to the third dose were excluded from the exposure–efficacy analyses</p><!><p>The Kaplan–Meier plots of PFS curves were similar to those for OS (Fig. 2b). The higher exposure groups were associated with longer PFS. Median PFS was 5.4, 4.6, 6.8, and 8.5 months for the ramucirumab C min,ss Q1, Q2, Q3, and Q4 groups, respectively (Table 2). Median PFS in the placebo plus FOLFIRI arm was 5.2 months (Table 2). The PFS hazard ratios were adjusted for baseline factors and decreased as the concentration of predicted ramucirumab C min,ss increased by quartile. Similar to the results for OS, PFS for Q1 and Q2 were not significantly different from the placebo group, but Q3 and Q4 demonstrated progressively decreasing HRs that were significantly different from the placebo group.</p><p>A matched case–control analysis for OS was explored to evaluate the exposure–OS relationship and adjust for imbalances between the C min,ss quartiles. As previously described, there were eight matching factors for which outcomes were to be adjusted: time to progression after beginning first-line therapy, KRAS status, ECOG PS, number of metastatic sites, liver only metastasis, CEA, gender, and combined prior bevacizumab use (composite subgroup). Patients from both ramucirumab and placebo arms were included in the exposure–response analysis only if they did not die or discontinue treatment prior to Day 1 Cycle 3 and had exposure data (C min,ss) available. The matching was performed separately for each of the four Cmin,ss exposure quartiles (Q1–Q4) from the ramucirumab plus FOLFIRI arm (Supplemental Table 1).</p><p>To compare the two treatment arms in each of the four matched case–control groups, Kaplan–Meier curves for OS in each group are shown in Supplemental Fig. 1. Clear separation of OS curves was observed in matched Q3 and Q4, but not Q1 and Q2. As compared with matched control patients, patients in Q3 and Q4 groups showed longer survival relative to patients in Q1 and Q2 groups. This is consistent with the exposure–response association as observed earlier. In addition, the steep dose–effect relationship depicted in Supplemental Table 3 shows that the risk of death or disease progression was reduced by approximately 40% or 30%, respectively, when C min,ss was doubled.</p><!><p>Dose omission/dose modification by exposure-response quartile 5-FU, 5-fluorouracil, Q quartile</p><p>Incidence of Grade ≥3 treatment-emergent adverse events by ramucirumab Cmin,ss exposure quartile. There was only one reported Grade 4 hypertension event, nine reported Grade 4 diarrhea events, and no Grade 4 fatigue events. A total of 9.4% patients reported Grade 4 neutropenia. There were no Grade 5 events for all four safety endpoints. Treatment-emergent adverse events were graded by NCI-CTCAE v4.0. Neutropenia and fatigue are consolidated terms, meaning they are a composite term consisting of multiple related preferred terms based on Standardized Medical Dictionary for Regulatory Activities (MedDRA) Queries and medical review. Cmin,ss, minimum concentration at steady state; FOLFIRI, folinic acid, 5-fluorouracil, and irinotecan; NCI-CTCAE National Cancer Institute-Common Terminology Criteria for Adverse Events; PBO placebo; Q quartile; RAM ramucirumab</p><!><p>In the RAISE trial, median OS and PFS were significantly greater for patients receiving ramucirumab plus FOLFIRI when compared to placebo plus FOLFIRI treatment [9]. Exploratory exposure–response analyses presented here demonstrate that longer median OS and PFS may be associated with increasing ramucirumab exposure as seen in patients in higher exposure quartiles (Q3 and Q4). This relationship was demonstrated with both an unadjusted analysis and a matched pair analysis. These analyses suggest that the observed exposure–efficacy relationship is independent of a patient's baseline characteristics. Although incidence of ≥Grade 3 neutropenia was found to be significantly correlated with predicted ramucirumab concentration, severity of neutropenia (Grade 3 vs Grade 4) was independent of exposure. Grade 3 or greater febrile neutropenia in the ITT population was low (ramucirumab: 18 patients, 3%; placebo: 13 patients, 2%) [9] and could not be evaluated by quartile due to the low incidence. There was no statistically significant relationship identified between ramucirumab exposure and hypertension, fatigue, or diarrhea. Thus, the analysis suggests that some of the toxicities seem to reach maximal intensity in the Q1 population and do not get worse with increasing exposure.</p><p>Of note, for both PFS and OS in multivariate Cox regression and the matched case–control analysis, patients in the two lowest exposure quartiles had no apparent benefit from adding ramucirumab to FOLFIRI; the benefit was seen in patients with higher exposure (Q3 and Q4). This steep dose–effect relationship shows that the risk of death or disease progression was reduced when C min,ss was doubled.</p><p>Although patients in the higher exposure quartiles appeared to have greater incidences of ramucirumab dose delay, there was no apparent relationship between ramucirumab exposure and dose reduction, dose omission, or dose discontinuation of ramucirumab. Ramucirumab dose intensity was consistent across all four quartiles, demonstrating that a ramucirumab 8 mg/kg every-2-week dosing regimen is safe. Higher incidences of dose delay, dose reduction, dose omission, or dose discontinuation of all components of FOLFIRI were observed for higher exposure quartiles. However, despite these dose reductions and delays in the Q4 group, they have superior PFS and OS compared to lower quartiles.</p><p>The development and introduction of monoclonal antibodies targeting EGFR and angiogenic pathways have expanded treatment options for cancer patients [20–30]. In most cases, the dosing strategy is based on the titration of the drug to a tolerable dose. However, clinical pharmacokinetics have been used to confirm the rationale for the recommended dose of antiangiogenic agents [31–33]. The current results with ramucirumab are also consistent with exposure–response (efficacy and safety) relationships observed for agents targeting tyrosine kinase inhibitors (TKIs). Exposure–response analyses of sunitinib, an oral multi-targeted receptor TKI, demonstrated that increased exposure is associated with longer OS, greater antitumor response, longer time to tumor progression, and some increased risk of adverse effects in patients with advanced tumors [34]. Similarly, greater C min of nilotinib, a selective break point cluster-Abelson (BCR-ABL) TKI, was associated with shorter time to complete cytogenetic response, shorter time to major molecular response, longer time to progression, and a trend toward higher response rates in patients with chronic myeloid leukemia [35]. These results are also consistent with exposure–response relationships observed for other agents (ipilimumab, ado-trastuzumab, and rituximab) targeting pathways in cancer. A higher C min of ipilimumab, a fully human IgG1 monoclonal antibody that blocks cytotoxic T-lymphocyte antigen-4, was associated with increased tumor responses and longer survival in patients with advanced melanoma [10]. An association between C min levels and OS, PFS, and objective response rate was observed for ado-trastuzumab, a human epidermal growth factor 2 (HER2)–directed antibody–drug conjugate, in HER-2 positive metastatic breast cancer [11]. In addition, increased clearance of bevacizumab, a humanized monoclonal IgG1 antibody that targets VEGF-A, was also associated with poorer prognosis for gastric cancer patients [36].</p><p>Previous analyses have demonstrated that higher predicted ramucirumab exposure was associated with longer OS and PFS, smaller hazard ratios, and increased but manageable toxicity in patients with previously-treated advanced gastric or gastroesophageal cancer as well as in patients with metastatic non-small cell lung cancer [13, 14]. The ramucirumab dose of 8 mg/kg every 2 weeks regimen is a clinically effective and safe dose in the CRC indication and offers a favorable benefit-risk profile in patients with CRC. The present exposure–response analysis shows a positive relationship between efficacy and ramucirumab exposure, seen particularly in the upper two quartiles. This opens the question of whether an increased dose of ramucirumab could achieve higher exposure in more patients, thus increasing the efficacy of ramucirumab treatment while maintaining a tolerable safety profile. Such clinical trials are ongoing in gastric and gastroesophageal junction cancers: with monotherapy (NCT02443883) and in the combination therapy with paclitaxel setting (NCT02514551).</p><!><p>Supplementary material 1 (DOCX 197 kb)</p><p>Electronic supplementary material</p><p>The online version of this article (doi:10.1007/s00280-017-3380-z) contains supplementary material, which is available to authorized users.</p>

PubMed Open Access

Theoretical Study of Shallow Distance Dependence of Proton-Coupled Electron Transfer in Oligoproline Peptides

Long-range electron transfer is coupled to proton transfer in a wide range of chemically and biologically important processes. Recently the proton-coupled electron transfer (PCET) rate constants for a series of biomimetic oligoproline peptides linking Ru(bpy)32+ to tyrosine were shown to exhibit a substantially shallower dependence on the number of proline spacers compared to the analogous electron transfer (ET) systems. The experiments implicated a concerted PCET mechanism involving intramolecular electron transfer from tyrosine to Ru(bpy)33+ and proton transfer from tyrosine to a hydrogen phosphate dianion. Herein these PCET systems, as well as the analogous ET systems, are studied with microsecond molecular dynamics, and the ET and PCET rate constants are calculated with the corresponding nonadiabatic theories. The molecular dynamics simulations illustrate that smaller ET distances are sampled by the PCET systems than by the analogous ET systems. The shallower dependence of the PCET rate constant on the ET donor-acceptor distance is explained in terms of an additional positive, distance-dependent electrostatic term in the PCET driving force, which attenuates the rate constant at smaller distances. This electrostatic term depends on the change in the electrostatic interaction between the charges on each end of the bridge and can be modified by altering these charges. On the basis of these insights, this theory predicted a less shallow distance dependence of the PCET rate constant when imidazole rather than hydrogen phosphate serves as the proton acceptor, even though their pKa values are similar. This theoretical prediction was subsequently validated experimentally, illustrating that long-range electron transfer processes can be tuned by modifying the nature of the proton acceptor in concerted PCET processes. This level of control has broad implications for the design of more effective charge-transfer systems.

theoretical_study_of_shallow_distance_dependence_of_proton-coupled_electron_transfer_in_oligoproline

INTRODUCTION<!>Molecular dynamics simulations<!>Theoretical modeling of the ET and PCET reactions<!>Probability distributions of ET distances<!>Experimental and calculated rate constants<!>Impact of distance-dependent electrostatic term and validation of theoretical prediction<!>Impact of probability distributions of ET distances<!>Interplay between electrostatic term and probability distribution of ET distances<!>Impact of coupling attenuation parameter<!>CONCLUSIONS

<p>Long-range electron transfer (ET) plays a significant role in a wide range of chemical and biological systems.1–4 Biomimetic model systems, in which the electron donor and acceptor are connected by molecular bridges, have been designed to investigate the fundamental physical principles underlying such long-range ET reactions. A variety of experimental and theoretical studies have been performed to elucidate the origin of the exponential dependence of ET rate constants on the length of the bridge between the electron donor and acceptor sites.1, 5–21 The nature of the bridges and their conformational flexibility can play significant roles in determining the distance dependence of the ET rate constants.17, 19, 22–26 Moreover, dependence of the driving force and reorganization energy on the distance may also influence the distance dependence of the ET rate constants.21, 27–30</p><p>Many long-range ET reactions in biological systems are also coupled to proton transfer (PT) reactions. These proton-coupled electron transfer (PCET) reactions play critical roles in biological processes such as nitrogen fixation,31 photosynthesis,32 respiration,33 and DNA synthesis.34 A significant amount of attention has been directed toward the dependence of PCET reactions on the proton donor-acceptor distance.35–43 However, only a relatively small number of studies have focused on the dependence of PCET reactions on the ET donor-acceptor distance.44–46 Recently, studies of PCET model systems with phenylene or p-xylene bridges44–46 indicated that PCET rate constants depend on the ET donor-acceptor distances in a similar manner as do the analogous ET systems with the same bridges.12, 47–49</p><p>Recently, a series of oligoproline peptides connecting Ru(bpy)32+ and tyrosine was designed and studied experimentally with a flash-quench transient absorption method.50 In these experiments, photogenerated Ru(bpy)33+ oxidized tyrosine by a concerted PCET process involving intramolecular ET from tyrosine to Ru(bpy)33+ and proton transfer from tyrosine to the hydrogen phosphate buffer in aqueous solution (Figure 1). The concerted PCET mechanism is supported by the phosphate buffer concentration dependence of the kinetics and the kinetic isotope effects. Moreover, the dependence of the PCET rate constant on the ET donor-acceptor distance was studied by varying the number of oligoproline linkers from n = 1 to 4. The experimentally measured ET distance dependence was found to be much shallower than the corresponding distance dependence measured for analogous bimetallic ET systems with oligoproline bridges 13, 51–53 (Figure 1). The shallow distance dependence for the PCET systems relative to the ET systems was proposed to be related to the necessity of employing stronger oxidants for tyrosine, thereby leading to lower tunneling barrier heights within the McConnell superexchange model.1, 54</p><p>Herein we use a combination of microsecond molecular dynamics (MD) simulations and vibronically nonadiabatic PCET theory55–57 to further elucidate the ET distance dependence of these oligoproline metallopeptides. This PCET theory has been used to investigate the dependence of the PCET rate constant on the proton donor-acceptor distance for a wide range of systems,35–36, 40–41 but the dependence on the electron donor-acceptor distance has not been studied as extensively. We perform microsecond MD simulations of the solvated ET and PCET oligoproline systems with bridges of different lengths, namely one to four proline linkers. We also calculate the ET and PCET rate constants for these systems using nonadiabatic ET and PCET theories, respectively. Analysis of the results provides a physical explanation for the shallower dependence of the PCET rate constant on the ET distance in terms of the net charges on the redox species terminating the bridge, producing an attractive electrostatic interaction that is smaller for the product than for the reactant, thereby decreasing the driving force. This effect becomes more pronounced as the electron donor-acceptor distance decreases. Moreover, our theoretical model predicts that this distance dependence of the PCET rate constant will increase when imidazole rather than hydrogen phosphate is the proton acceptor, even though the relevant pKa values are similar. This theoretical prediction is subsequently validated experimentally, illustrating how the nature of the proton acceptor can significantly impact the dependence of the PCET rate constant on the ET distance.</p><!><p>The ET and PCET systems studied in the present work are shown in Figure 1. The Cambridge Structural Database (CSD) entries TIRTIY58 and SOYRIJ59 were used to prepare the structures of the metal-containing terminal groups, and the CSD entry SOWJUL60 was used to build the proline linkage. For the MD simulations, the AMBER ff14SB force field61 was employed to model the proline linkage and the terminal Tyr group, the general AMBER force field (GAFF)62 was used to model the metal-containing terminal groups, and the TIP3P water force field63 was used to model the solvent. The partial charges of the terminal Tyr group were obtained through a two-stage RESP fitting procedure for two different configurations,64–65 and the parameters for the metal sites were generated with the MCPB.py program.66 Each of the four PCET systems was solvated in 0.1 M [HPO4]2− buffer solution and neutralized by the addition of K+ ions. Each of the four ET systems was solvated in a water box neutralized by the addition of Cl− ions. The long-range electrostatic interactions were treated with the particle mesh Ewald method.67 After a careful equilibration procedure, a 1 μs trajectory was propagated using the pmemd.cuda program68 for each of the eight systems, with configurations saved every 10 ps. These configurations were used for the analysis of the distributions of the ET donor-acceptor distance. Further details of the force field parameters and MD simulations are provided in the Supporting Informaton (SI).</p><!><p>For each ET system, the rate constant was calculated by thermal averaging of the standard nonadiabatic ET rate constant over the ET donor-acceptor distance:69 (1)kET=∫kET(RET)P(RET)dRET (2)kET(RET)=1ℏ|Vel0|2exp[−βET(RET−RET0)]πλkBTexp[−(ΔGET+λ)24λkBT] Here RET is the distance between the two Ru ions in the bimetallic ET system and P(RET) is the probability distribution obtained from the conformational sampling in the MD simulations. Moreover, kB is the Boltzmann constant, T is the temperature, and Vel0 is the electronic coupling between the ET donor and acceptor sites with the ET distance RET0 corresponding to a single bridge unit (i.e., n = 1). The specific value of Vel0 is not needed for this study because it simply corresponds to an additive constant to ln(kET). The coupling attenuation parameter βET was set to 1.4 Å−1 according to the ET study of the oligoproline bimetallic systems.13 The reorganization energy λ for each of the four ET systems was evaluated using a Marcus two-sphere model70 and averaging the resulting values according to the distribution P(RET) (see SI). The ET reaction free energy ΔGET was determined to be −26.9 kcal/mol on the basis of the relevant reduction potentials (see SI).</p><p>In the vibronically nonadiabatic PCET theory,55–57 the reaction is described in terms of nonadiabatic transitions between reactant and product electron-proton vibronic states. When fluctuations in the environment lead to the degeneracy between a pair of reactant and product vibronic states, the electron and proton tunnel simultaneously with a probability proportional to the square of the vibronic coupling. For each PCET system, the calculation of the rate constants required thermal averaging over both the ET donor-acceptor distance RET and the PT donor-acceptor distance RPT. The thermal averaging over RET was analogous to that described for the ET systems, where kET is replaced by kPCET in Eq. (1), and in this case the electron transfer distance is defined as the distance between the Ru3+ ion and the geometrical center of the six carbon atoms of the Tyr benzene ring. The probability distribution P(RET) was assumed to be the same for the PCET systems with either hydrogen phosphate or imidazole as the proton acceptor.</p><p>The nonadiabatic PCET rate constant kPCET(RET) at each fixed ET distance is given by (3)kPCET(RET)=1ℏ|Vel0|2exp[−βET(RET−RET0)]πλkBT×∑μPμ∑v〈|Sμv(RPT)|2〉exp[−(ΔGμv+λ)24λkBT] This PCET rate constant expression accounts for all pairs of reactant and product proton vibrational states μ and ν obtained by solving the one-dimensional Schrödinger equation for a proton moving in the reactant and product electronically diabatic proton potentials, respectively. The rate constant is calculated by averaging over the reactant proton vibrational states μ with Boltzmann populations Pμ and summing over the product proton vibrational states ν. The vibronic coupling is the product of the electronic coupling and the overlap integral Sμν(RPT) between the reactant and product proton vibrational wavefunctions. The angular brackets denote averaging over the proton donor-acceptor distance RPT with the Gaussian distribution function: (4)P(RPT)=fPT2πkBTexp[−fPT(RPT−RPT0)22kBT] Note that this model neglects the dependence of the electronically diabatic proton potentials, as well as the reorganization energy, on RPT, thereby enabling the thermal averaging over RPT to be restricted to the overlap integrals.</p><p>In the present work, the electronically diabatic proton potentials were represented by Morse potentials corresponding to O-H or N-H bonds (see SI), and the overlap integrals were calculated analytically.71 Furthermore, RPT0, which is the equilibrium proton donor-acceptor distance in the reactant state, and the force constant fPT, which is related to the second derivative of the energy profile along RPT, were obtained from density functional theory (DFT) calculations (Table S4). As for the ET systems, the reorganization energy for each of the PCET systems was estimated by averaging the RET-dependent value obtained from the Marcus two-sphere model according to the distribution P(RET) (see SI). For the PCET systems, Vel0 is defined analogously as the corresponding parameter for the ET system but is expected to assume a different value, which is not needed to compute the relevant slopes. For consistency, we used the same electronic coupling attenuation parameter βET for both the ET and PCET systems, although βET may be different for these two types of systems because of their different thermodynamic properties. To explore this possibility, we also investigated the impact of using different values of βET for the ET and PCET systems, but the qualitative conclusions did not change.</p><p>The free energy difference ΔGμν in the PCET rate constant is given by (5)ΔGμv=ΔGPCET+εv−εμ+ΔEele where ΔGPCET is the reaction free energy of the PCET reaction, and εμ and εν are the proton vibrational energy levels for the states μ and ν relative to the lowest vibrational states of the corresponding diabatic proton potentials. ΔGPCET was determined to be −10.0 kcal/mol from the pKa values and redox potentials of the relevant species (see SI). The additional electrostatic term ΔEele, which depends on the distance between the metal center (Ru) and the proton acceptor (PA) group, accounts for the difference in electrostatic interactions between the charges localized on these sites in the reactant and product states in the PCET reaction and is given by (6)ΔEele =Δ(QRuQPA)εeffRRu-PA Here RRu-PA, which is the distance between the metal center and the proton acceptor group, is approximated by RET. In this expression, Δ(QRuQPA) is the difference between the product of the charges on the metal center and the proton acceptor group after and before the PCET reaction. For example, when [HPO4]2− serves as the proton acceptor, QRuQPA is (+3e)×(‒2e) before and (+2e)×(‒1e) after the PCET reaction, resulting in Δ(QRuQPA) = 4e2. In this case, because the attractive electrostatic interaction is smaller for the product than for the reactant, this difference is positive and therefore decreases the driving force, particularly at shorter distances. Finally, εeff is an empirical factor that accounts for the screening effects of the molecular and solvent environment as well as delocalization of the charges and thus depends on factors such as the solvent, ionic strength, ion pairing, and electronic structure of the donor-bridge-acceptor system. As discussed further below, this factor reflects an average over electrostatic interactions through the solution and/or the molecular bridge. In the present study, this factor was set to 5.1 to reproduce the slope of the experimentally measured ln(kPCET) versus the computed average ET distances for the system with hydrogen phosphate serving as the proton acceptor. We used the same value of this factor to model the system with imidazole as the proton acceptor and found good agreement with the experimentally measured slope for this system. We emphasize that ΔEele adjusts the driving force of the oligoproline peptides by including the electrostatic interactions between the electron donor and acceptor because ΔGPCET is computed from the thermodynamic properties of the isolated donor and acceptor species.</p><!><p>To characterize the ET donor-acceptor distance fluctuations in the ET and PCET reactions, we analyzed the probability distributions of the ET distances obtained from microsecond MD trajectories. For the ET systems, RET was defined as the distance between the two Ru ions, and for the PCET systems, RET was defined as the distance between the Ru3+ ion and the geometrical center of the six carbon atoms of the Tyr benzene ring. As shown in Figure 2, the probability distribution for each ET system resembles a single Gaussian. In contrast, the probability distribution for each PCET system is broader with a shoulder at smaller distances, indicating that multiple conformations are sampled on the microsecond timescale. The average RET values, denoted R¯ET, for the PCET systems are slightly smaller than those for the ET systems with the same number of proline residues and exhibit larger fluctuations (Table 1). These differences may be due to the flexibility of the neutral terminal Tyr electron donor group in the PCET systems, which enables sampling of conformations with smaller distances with respect to the positively charged Ru3+ center. In contrast, the metal centers in the ET systems are relatively bulky with the bound ligands and are both positively charged, preventing effective sampling of conformations with smaller distances. To understand the role of the counterions in the ET systems, we calculated the radial distribution function (RDF) between the Ru2+ and Ru3+ metal centers and the Cl− ions for each of the four systems (Figure S4). For all of these systems, the Cl− ions were found closer to the Ru3+ than to the Ru2+ because of its greater positive charge and smaller ligands.</p><p>To further analyze the probability distributions for the PCET systems, we fit the distribution for each system to a linear combination of three Gaussians (Table S6 and Figures S5 and S6). We also examined representative conformations corresponding to the distance associated with the maximum of the probability distribution function for the ET systems and the distance associated with each Gaussian component for the PCET systems, as depicted in Figures S7 and S8. The probability distribution function for the PCET system with n = 1 has an additional peak at short distances compared to the probability distribution functions for the longer peptide bridges. The conformation associated with this Gaussian component with RET ~7.2 Å (Figure S8) illustrates that the single proline linkage allows the Tyr to curl around in a manner that decreases its distance from the Ru3+ center, although it also samples more extended conformations. In contrast, the longer peptide bridges do not have the flexibility to curl in this manner, although they also sample a wide range of extended as well as bent conformations (Figure S8), leading to the shoulders exhibited in Figure 2B.</p><!><p>The magnitudes of the slopes estimated from the dependence of the experimentally measured ln(k) values on the calculated R¯ET values for the ET and PCET systems studied are provided in the last column of Table 2. The experimentally determined slope for the PCET reactions of the Ru/Tyr system, with hydrogen phosphate as the proton acceptor, is significantly smaller in magnitude than the slope for the ET reactions of the related Ru/Ru system. Note that the ET donor-acceptor distances can also be estimated on the basis of experimental data from related systems or using other strategies (see SI). We found that the slopes obtained with these different options for determining the ET distances do not vary significantly (Table S8 and Figures S9 and S10). For consistency, we present all of the experimental and calculated kinetic data in the main paper in terms of the calculated R¯ET values.</p><p>To elucidate the factors that determine the shallow distance dependence of the PCET reaction rate constant with hydrogen phosphate as the proton acceptor, we calculated kET and kPCET for the eight systems using the nonadiabatic rate constant expressions and parameters given above. The slopes associated with the distance dependence of the rate constants were obtained from linear fits of the calculated values of ln(k) versus the corresponding average distances R¯ET for the ET systems and the PCET systems with hydrogen phosphate as the proton acceptor (Figure 3). The calculated values of ln(k) versus R¯ET were found to behave linearly with slopes that agree well with the corresponding slopes obtained experimentally (Table 2). Note that this agreement was obtained for the PCET systems by fitting the empirical parameter εeff to the experimentally determined slope (Figure 3B). As discussed below, however, the same value for εeff was used to successfully predict the increased magnitude of the slope for the PCET system with imidazole as the proton acceptor (Figure 3C).</p><p>The parameter εeff should be viewed as an empirical factor that accounts for many different effects that are present in the experimental systems but are not included explicitly in our theoretical model. For example, this simple electrostatic model does not include the effects of charge delocalization, ion pairing, or the detailed electronic structure of the molecular system. To examine the fundamental basis for this electrostatic term, we performed constrained DFT calculations on three conformations of the n = 1 PCET system sampled during the MD simulations (Figure S8). The relative energies of the reactant and product states for the PCET reaction were computed by moving the proton from the donor to the acceptor and constraining the charge distribution using constrained DFT. As shown in Table S9, these relative reaction energies exhibit the trend that would be followed using the electrostatic term with εeff = 1, as expected for gas phase reactions. The conformations depicted in Figure S8 illustrate that the charges associated with the ends of the bridge interact through the solution or the molecular bridge or a combination thereof. Thus, the value of ~5 should be viewed as the result of averaging over all of these types of interactions, as well as accounting for the other effects mentioned above.</p><!><p>To understand the physical basis for the significantly different slopes observed for the ET systems and the PCET systems with hydrogen phosphate as the proton acceptor, we investigated the influence of the electrostatic term ΔEele on kPCET. We found that decreasing Δ(QRuQPA) with all other parameters, including εeff, held constant leads to a significant increase in the magnitude of the slope of ln(kPCET) versus R¯ET. Specifically, this slope magnitude is 0.59 Å−1 for Δ(QRuQPA) = 4e2 and is 1.74 Å−1 when Δ(QRuQPA) = 0, illustrating that the slope becomes similar to that of the ET system when the ΔEele term in Eq. (5) is omitted. Moreover, the theoretical model predicts an intermediate slope magnitude of 0.96 Å−1 when Δ(QRuQPA) = 2e2 and all other parameters remain the same. This analysis also assumed that the electronic coupling attenuation parameter βET is the same for the ET and PCET systems. This assumption will be removed in additional calculations discussed below but does not change these qualitative trends.</p><p>To test this theoretical prediction, the PCET rate constants were experimentally measured for the same oligoproline metallopeptides with imidazole as the proton acceptor. The conjugate acid of imidazole has a pKa similar to the pKa of [H2PO4]− but is cationic rather than anionic, resulting in Δ(QRuQPA) = 2e2 in the numerator of the electrostatic contribution to the driving force. The magnitude of the slope of the experimental ln(kPCET) versus R¯ET was found to be 0.87 Å−1, which is in qualitative agreement with our calculated value of 0.95 Å−1 for the system with imidazole as the proton acceptor. Note that using the pKa value of imidazole instead of phosphate and the Morse potential associated with an N-H rather than an O-H bond impacts the slope by only 0.01 Å−1. The agreement between the experimental measurement and our theoretical prediction provides validation for the theoretical model. It also supports our hypothesis that the electrostatic term ΔEele plays an important role in determining the shallow distance dependence of the PCET reaction rate constant with hydrogen phosphate as the proton acceptor.</p><p>To separate the effects of the electrostatic term and the probability distribution, we computed the slopes for both ET and PCET with two different electrostatic terms added to the driving force. If we set the electrostatic term ΔEele to 0 and 4e2/(εeffRET), we obtain different responses for the ET and PCET distance dependences. Specifically, the magnitude of the slope for the ET rate constant is 1.37 Å−1 and 1.14 Å−1, while the magnitude of the slope for the PCET rate constant with hydrogen phosphate as the proton acceptor is 1.74 Å−1 and 0.59 Å−1, respectively, for these two values of the electrostatic term. Thus, the electrostatic term exerts a much greater impact on the slope for the PCET reaction than for the ET reaction. As discussed below, our analysis suggests that this difference is due to the qualitative differences in the ET donor-acceptor distance probability distributions P(RET) for the ET and PCET systems shown in Figure 2.</p><!><p>To further analyze the impact of P(RET), we fit the probability distributions obtained from the microsecond MD simulations to a linear combination of three Gaussians for each of the four PCET systems (Table S6 and Figures S5 and S6). To determine the contributions from these Gaussian components to the distance dependence of the PCET rate constants, we calculated the slopes for each of the four systems using only the third (major) Gaussian component, the sum of the second and third Gaussian components, or the sum of all three Gaussian components (Table 3). The magnitude of the slope calculated using only the third Gaussian component is larger than that calculated using the sum of the second and third Gaussian components, which in turn is larger than the magnitude of the slope calculated using the sum of all three Gaussian components. Thus, conformational sampling of the shorter ET distances effectively decreases the magnitude of the slope, leading to a shallower distance dependence. In contrast, the shorter distances are not as accessible for the ET systems, as indicated by the absence of the shoulders in the probability distributions (Figure 2). Note that when ΔEele = 0, this effect is not observed.</p><!><p>To investigate the interplay between the electrostatic term ΔEele and the probability distribution P(RET) in determining the distance dependence of the PCET rate constant, we analyzed the dependence of the various terms in the PCET rate constant kPCET(RET) on RET (Figure 4). This analysis includes only the term in Eq. (3) corresponding to the lowest reactant and product proton vibrational states, which was found to be the dominant contributor to the overall PCET rate constant (Table S10). This analysis used a constant reorganization energy corresponding to the value computed for the PCET system with n = 1 and hydrogen phosphate as the proton acceptor. The activation term ΔG00‡=exp[−(ΔG00+λ)24λkBT] increases as RET increases (Figure 4A) because ΔEele decreases as this distance increases. However, the squared electronic coupling, which is proportional to exp[−βET(RET−RET0)], exhibits the opposite trend (Figure 4B). The resulting ln[kPCET(RET)] exhibits a shallower distance dependence at RET distances in the range 12 – 16 Å and provides an explanation for the observation that these shorter distances are mainly responsible for the shallow distance dependence of the PCET rate constant (Figure 4C and Table 2). Thus, both the electrostatic term and the RET probability distribution play significant roles in determining the shallow distance dependence of the PCET rate constant. Changing ΔEele alters the slope in Figure 4A and therefore the curve shape in Figure 4C which, combined with the broadening of P(RET) toward shorter distances, leads to the shallow distance dependence of the overall PCET rate constant. This analysis indicates that the dependence of the rate constant on the ET distance can be influenced not only by the electronic coupling prefactor but also by the free energy barrier appearing in the exponential.</p><p>The relatively flat curve for ln[kPCET(RET)] versus RET in Figure 4C at small RET predicts similar rate constants for the PCET systems with no proline linkage and with a single proline linkage, assuming the same value for Vel0. This behavior is consistent with experimental data indicating that the PCET system without a proline linkage has a rate constant of 3×106 M−1s−1,74 which is similar to the rate constant of 1.4×106 M−1s−1 measured experimentally50 for the PCET system with one proline linkage. As mentioned in the Introduction, only a few studies of the dependence of the PCET rate constant on the ET distance have been reported.12, 44, 47 In the study of PCET in iron carboxytetraphenylporphyrin complexes, a relatively small slope magnitude of 0.23 Å−1 for the overall distance dependence of the rate constant was measured,44 compared to the value of 0.4–0.46 Å−1 measured for ET reactions with the same bridge unit.12, 47 In this system, the PCET reaction has an electrostatic term with Δ(QRuQPA) = e2, while the ET reaction has an electrostatic term with Δ(QRuQPA) = −e2 or zero. These different electrostatic terms may account for the lower magnitude of the slope for PCET compared to ET in these systems. In general, these two factors, namely the broad distribution of ET distances and the distance-dependent electrostatic contributions to the driving force, could impact the distance dependence of the rate constants for ET systems as well as PCET systems.</p><!><p>Previous work50 has suggested that the coupling attenuation parameter βET may be different for the ET and PCET systems, based on the different donor-bridge-acceptor energetics. To investigate this possibility, we repeated these calculations with βET = 1.2 Å−1 rather than 1.4 Å−1 for the PCET systems. We found that the experimentally determined slopes for the PCET systems with the two different proton acceptors, phosphate and imidazole, could also be reproduced after readjusting the empirical factor εeff in Eq. (6) (Table S11). However, a model in which the value of βET was different for the ET and PCET systems, without including the electrostatic term for PCET, cannot reproduce the larger magnitude of the slope with imidazole as the proton acceptor. Thus, the electrostatic term appears to be essential for describing the experimental data, but the different values of βET may also be a factor in describing the ET and PCET systems.</p><!><p>We investigated ET reactions in oligoproline bimetallic systems and PCET reactions in oligoproline metallopeptides using theoretical modeling. We performed microsecond MD simulations for the solvated ET and PCET systems and observed significant differences in the ET donor-acceptor distance probability distributions in these two systems. Specifically, the probability distributions for the ET systems are relatively narrow and symmetric, whereas those for the PCET systems are broader with shoulders at smaller distances, allowing the conformational sampling of these smaller distances on the microsecond timescale of the experiments. Another important distinction between the ET and PCET systems is the contribution to the driving force arising from the change in the electrostatic interaction between the charges at each end of the bridge. Although this electrostatic term is zero for the ET systems, it is positive and increases as the electron donor-acceptor distance decreases for the PCET systems. We emphasize that these effects are not related to fundamental differences between ET and PCET processes but rather are specific to these particular systems.</p><p>The ET and PCET rate constants computed with nonadiabatic theories reproduced the experimentally measured dependence of these rate constants on the average ET donor-acceptor distance for varying bridge lengths. The shallower distance dependence of the PCET rate constant with hydrogen phosphate as the proton acceptor was explained in terms of an additional positive, distance-dependent electrostatic term in the driving force, which attenuates the rate constant at smaller distances, in conjunction with the enhanced sampling of smaller distances observed in the MD simulations. The theoretical model predicted an increased magnitude of the slope for the PCET systems when imidazole is the proton acceptor because, although the conjugate acid of imidazole has a similar pKa value as dihydrogen phosphate, its cationic rather than anionic charge leads to a smaller positive electrostatic contribution to the driving force. This theoretical prediction of an increased magnitude of the slope was subsequently validated by experimental measurements.</p><p>This work highlights the impact of the nature of the proton acceptor, particularly its charge, on the ET distance dependence of the PCET rate constant. Because the ET distance dependence of the rate constant can be influenced by not only the conventional β parameter in the electronic coupling prefactor but also by a term in the free energy barrier, the slope associated with the distance dependence assumes a more complex interpretation. These insights into how short-range proton transfer reactions can be used to tune long-range electron transfer reactions have broad implications for designing charge-transfer systems relevant to solar cells, sensors, and other types of devices.</p>

PubMed Author Manuscript

Adsorption Free Energy Predicts Amyloid Protein Nucleation Rates

Primary nucleation is the fundamental event that initiates the conversion of proteins from their normal physiological forms into pathological amyloid aggregates associated with the onset and development of disorders including systemic amyloidosis, as well as Significance StatementProtein malfunction and misfolding has long been associated with the onset and development of neurodegenerative conditions such as Alzheimer's and Parkinson's diseases.Protein misfolding/aggregation proceeds via primary nucleation whereby proteins can convert from their normal physiological forms into pathological amyloid aggregates.The presence of interfaces plays an important role in this aggregation phenomenon and can dramatically modulate nucleation, either by accelerating or inhibiting aggregation.Through coarse-grained computer simulations, kinetic theory, and by measurement of the Gibbs free energy of adsorption, we show that systems which have high affinity to adsorb to an interface display delayed kinetics, whereas systems that have low affinities have similar behaviour, suggesting a maximum between the two regimes (non-monotonic behaviour) where the aggregation rates are highest.

adsorption_free_energy_predicts_amyloid_protein_nucleation_rates

<!>Introduction<!>Surface-to-volume ratio affects protein aggregation by modulating the rate of heterogeneous primary nucleation<!>Heterogeneous nucleation at interfaces is bypassed in the presence of preformed seed fibrils<!>Morphological changes due to varying interface chemistry<!>Molecular mechanism of heterogeneous amyloid nucleation<!>Conclusions<!>Protein solution preparation<!>Silk fibroin preparation and purification<!>Aggregation kinetics at different interfaces<!>Preparation of lipid interface<!>Analysis of kinetic data<!>Atomic Force Microscopy (AFM)<!>Interfacial tension (IFT) measurement<!>Simulation method<!>Fitting procedure

<p>the neurodegenerative conditions Alzheimer's and Parkinson's diseases. It has become apparent that the presence of surfaces can dramatically modulate nucleation. However, the underlying physico-chemical parameters governing this process have been challenging to elucidate, with interfaces in some cases having been found to accelerate aggregation, while in others they can inhibit the kinetics of this process. Here, we show through kinetic analysis that for three different fibril-forming proteins, interfaces affect the aggregation reaction mainly through modulating the primary nucleation step. Moreover, we show through direct measurements of the Gibbs free energy of adsorption, combined with theory and coarse-grained computer simulations, that overall nucleation rates are suppressed at high and at low surface interaction strengths, but significantly enhanced at intermediate strengths, and we verify these regimes experimentally. Taken together, these results provide a quantitative description of the fundamental process which triggers amyloid formation and shed light on the key factors that control this process.</p><!><p>In order to fulfil their biological roles, proteins typically remain soluble in the cellular environment. Under certain conditions, however, proteins can transition from these physiological forms into pathological amyloid aggregates. [1][2][3][4][5] During this aggregation process, small oligomeric species are formed. These intermediates are toxic and have been associated with membrane disruption and synaptic dysfunction. 6 Such structures play a crucial role in the development of many neurodegenerative diseases, such as Alzheimer's and Parkinson's diseases, which represent a major threat to healthcare systems in an ageing society. 4,5 The process of amyloid formation is triggered by primary nucleation events that generate seed aggregates capable of sequestering further protein molecules through rapid growth. [7][8][9][10][11][12][13][14] The mechanisms which underpin this fundamental nucleation step, and the factors which control its rate have remained challenging to elucidate.</p><p>Interfaces and surfaces are ubiquitous in nature and play an important role as a platform for many biological reactions. Hydrophobic-hydrophilic interfaces are essential in biomolecular interactions, and can act as catalysts in the self-assembly of micelles and bio-membranes 15,16 or for regulating protein folding. 17,18 Proteins and peptides are among the most common amphiphilic molecules in biological environments; and it has been shown that they can display different behaviours, depending on the interface involved. [19][20][21][22][23][24][25][26][27][28][29] More importantly, it has been shown that surfaces can accelerate or retard protein aggregation processes. [30][31][32] To study the effects on protein assembly, introducing well-defined model interfaces into the protein solutions is critical. To date, a mechanistic approach describing the effect of interfaces on protein aggregation has not been well described and it has remained challenging to determine the molecular mechanisms by which surfaces modulate protein aggregation.</p><p>Here, we investigate the mechanistic basis by which surfaces can modulate protein aggregation. To this effect, we first explored how different interfaces inhibit or by contrast promote amyloid self-assembly and compare our results with the behaviour of a β-sheet rich non-amyloid system, silk. 33,34 We find that increasing the surface-to-volume ratio accelerated protein self-assembly, while the addition of pre-formed seed aggregates largely negates the influence that interfaces have on the kinetics of amyloid formation, demonstrating that surfaces largely control primary nucleation. Finally, by determining the Gibbs free energy of adsorption for different interface-systems we relate the interface affinity to the changes in nucleation rates, providing a quantitative description of how interfaces modulate amyloid formation.</p><!><p>To determine the contribution of interfaces relative to that of bulk solution to the kinetics of amyloid formation, we first investigated the effect of changes in the surface-to-volume (S/V) ratio, as shown in the schematic in Fig. 1a. To this effect, kinetic experiments were conducted in systems with different volumes; by altering the volume of the protein solution in a well from 80 µL to 140 µL, the S/V ratios in the range from 0.218 mm −1 to 0.132 mm −1 were explored as shown in Fig 1b . We first focused on the peptide hormone insulin and monitored the aggregation kinetics through fluorescence spectroscopy. The resulting kinetic curves for the buildup of fibril mass, which display a characteristic sigmoidal shape, show that for the range of interfaces explored a higher S/V results in faster aggregation kinetics (Fig. 1b). To quantify this effect, we employed a chemical kinetics framework [36][37][38] that allows us to interpret the aggregation profiles in terms of the rate constants of the underlying microscopic steps of aggregation, including primary nucleation, fibril elongation and secondary processes such as surface catalysed secondary nucleation and fragmentation. The fitting was done first for one S/V ratio and all parameters other than k + k n were kept constant thereafter. The elongation and secondary nucleation product, k + k 2 , was obtained through this analysis and was found to be 4.04 × 10 10 with n c = n 2 = 2; where k n , k + and k 2 are the primary nucleation rate, the elongation rate, and the secondary nucleation rate constant respectively, while n c is the reaction order of the primary process and n 2 is the reaction order of the secondary pathway.</p><p>In particular, this analysis (Fig. 1c) reveals that the rate constants for elongation and secondary nucleation are independent of the S/V ratio, while the primary nucleation rate constant varies from 1.16 × 10 5 to 1.23 × 10 4 M −2 h −2 . Thus changing the surfaceto-volume ratio by less than a factor of two, results in an order of magnitude change in the primary nucleation rate constant. This finding therefore suggests that interfaces can strongly modulate amyloid formation through controlling the rate of primary nucleation.</p><p>To further establish whether this behaviour is concentration dependent, insulin solutions ranging from 0.05 to 5 mg/mL were prepared and monitored (Fig. 1d). These data show that the same trend can be observed across the concentration range tested. The kinetic analysis (Fig. 1e) revealed that the primary rate constants varied across all concentration ranges tested. We next repeated this experiment with the protein lysozyme.</p><p>The data shown in Fig. S1a demonstrate that the interfacial effects found for insulin are also characteristic of the kinetics of amyloid formation for lysozyme.</p><!><p>In order to verify the conclusion obtained from kinetic analysis suggesting that interfaces modulate primary nucleation, we sought to bypass this step in the aggregation reaction through addition of pre-formed protein seeds to the protein solutions in 96-well plates (top panels of Fig. 2a-c). If the formation of nuclei predominantly occurs at an interface, the addition of seeds effectively negates any surface dependence on fibrillar growth as the required nuclei would already be in solution, and consequently there should be minimal differences in kinetics when changing the S/V ratio. This prediction is verified in the aggregation kinetic curves for insulin where independently of the S/V ratio similar aggregation kinetics are observed upon the addition of seed fibrils, in stark contrast with the behaviour observed in the absence of seeds (Fig. 2a-c and Fig 2d). Repeating this experiment for the proteins lysozyme and silk fibroin, revealed that also for these systems, the addition of pre-formed fibril seeds eliminated the dependence of aggregation kinetics on the S/V ratio (Fig. 2e-f). However, as shown in Fig. 2f, RSF did not perform in a similar manner, and surface-driven aggregation was not observed. This is due to the extraction method in which silk fibroin is obtained, where, during the purification process, small aggregates are formed. These aggregates can consequently act as seeds, thus accelerating self-assembly processes.</p><p>In order to investigate how interfaces affect amyloid formation, we varied the nature of the interface by using immiscible phases such as mineral oil (MO) or silicon oil (SO) on top of each protein solution in the 96-well plate (Fig. 3a top panel). Thus we had water-mineral oil (W-MO) and water-silicon oil (W-SO). This ensured that the oil fully covered the aqueous phase and thus eliminated the water-air (W-A) interface. Taken together, these data further corroborate the idea that the principal effect of interfaces for amyloid formation is to introduce a heterogeneous nucleation step into the pathway which can substantially modify the effective rate of primary nucleation, and that interfaces have a more minor effect on the other processes in the system (Fig. 3). It is remarkable that homogeneous nucleation through secondary processes is not affected by interfaces but rather through primary processes, revealing the much more effective nature of monomer-fibril interactions compared to monomer-monomer interactions. As previously mentioned, due to the formation of nuclei in the purification process for RSF, no significant interface dependence is observed, and the addition of seeds only reduces the overall aggregation time.</p><p>In order to probe whether the oil affects the fibrillisation process, FTIR measurements were conducted. We incubated the monomeric protein solution with the oil at room temperature for 10 hours and took the FTIR spectra of both that and of a monomeric protein solution which was only in contact with an air interface. The spectra of both these samples looked identical and had amide I and II bands at around 1650 cm −1 and 1540 cm −1 respectively, which are characteristic of random coil structure. We then incubated the protein solution that was in contact with the oil at 50 • C in order to promote aggregation and took the FTIR spectrum of this sample. The protein seems to have aggregated with the characteristic amide I and II bands shifting to lower wavenumber, indicating a transition to beta-sheet formation. This shows that the oil interface does not inhibit fibrilisation and it is molecular adsorption which plays a critical role in defining aggregation kinetics (Fig. S1c). no seeds 0.002% w/w seeds 0.02% w/w seeds</p><!><p>The finding that nucleation takes place predominantly at interfaces suggests that the nature of the interface could modulate the morphology of the structures formed. To explore this effect, atomic force micrographs (AFM) of fibrils nucleated on different interfaces were acquired. As can be seen in Fig. 4a-c, the AFM results revealed that the samples where there is a water-air interface formed mature fibrils, whereas the samples where a water-oil interface is present mostly consisted of protofibrils and oligomers.</p><p>Additionally, AFM micrographs were acquired for protein fibrils grown from pre-formed seeds. The AFM (Fig. 4d-f) images show that under such conditions the surface has a much smaller effect on the morphology of the structures formed and rather that seeded samples give structure to the final fibrils formed.</p><!><p>We next explored the origin of the effect of interfaces on amyloid formation by varying the nature of the interface. Again, mineral oil (MO) and silicon oil (SO) were used. For all concentrations tested in this study, we found that the aggregation reaction was most rapid in the presence of an air-water interface, followed by the water-SO and finally by the water-MO (Fig. 5a). This behaviour was also observed for lysozyme (Fig. S1b). We next varied the S/V ratio for the three different interfaces at a fixed insulin concentration.</p><p>The data shown in Fig. 5b demonstrate that a variation in the aggregation kinetics is observed with changes in the S/V ratio for both water-air and water-SO, and while this trend is also observed for water-MO, the S/V ratio plays a smaller role on the half-time. For lysozyme, a similar trend was observed for both the water-air and water-MO interfaces. Interestingly, however, for the case of water-SO, amyloid formation was independent of the S/V ratio (Fig. 5c). Moreover, k + k n was plotted as a function of the S/V ratio for both proteins. Again, the same kinetic analysis approach was employed as in Fig. 1 and k + k 2 was found to be 2.83 × 10 10 and 2.08 × 10 10 for Fig. 5d and Fig.</p><p>5e respectively, while n c = n 2 = 2 for both plots. Moreover, in order to exclude any effects due to evaporation, we conducted a series of experiments with 2 different protein concentrations and with all three interfacial systems, in order to determine whether the surface to volume ratio is the contributing factor behind the difference in aggregation kinetics. By conducting insulin aggregation kinetics at 40 • C rather than at 50 • C, we found that even at lower temperatures, the S/V ratio and not evaporation is the key factor behind the difference in the aggregation kinetics (Fig. S2a-e).</p><p>Taken together, the experimental data describing how the S/V ratio affects protein aggregation, coupled with the effect that seeds have on the different systems,</p><p>show that protein self-assembly has a strong dependence on both the surface area but also on the nature of the interface used. Specifically, when the S/V ratio is increased, the rate of nucleation of protein aggregates is affected (Fig. 1), indicating that nuclei form on the interface and therefore increasing the surface-to-volume ratio promotes this effect. These surface effects must, therefore, originate from the interaction of the protein molecules with the interface. To probe these interactions, we next</p><p>Surface-to-volume (mm -1 ) Insulin (0.5 mg/mL)</p><p>Surface-to-volume (mm -1 ) 0.12 0.14 0.16 0.18 0.2 0.22 0.24 10</p><p>, where Γ(∞) = (V /S) s tot is the maximal surface excess, K 1 is the equilibrium adsorption constant of monomers to the interface and s tot is the maximal surface concentration of adsorbed molecules. Interfacial tension (IFT) measurements for different insulin concentrations were conducted using pendant drop tensiometry for the eight different interface systems (see methods). The resulting data, shown in Fig. 6a, were fitted to the Langmuir-Szyszkowski equation to yield, for each interface system, the equilibrium adsorption constant K 1 and therefore the standard free energy of adsorption ∆G ads = −RT ln(K 1 ). From this analysis (Fig. 6a-b) it can be seen that the W-MO and W-MO(lipids) systems have the largest adsorption free energies, indicating a higher tendency to adhere to the interface, while the W-MO(2%AbilEM90) system has the smallest adsorption energy. The other two systems, lie in between these extremes. This observed trend, follows that of increasing hydrophobicity of the nonaqueous phase, suggesting a major role for hydrophobic interactions in modulating the surface adsorption process. Strikingly, however, a plot of k + k n against adsorption free energy (Fig. 6d), shows that stronger adsorption at the interface does not necessarily yield a faster nucleation rate. Indeed, even though the W-MO and W-MO(lipids) systems promote strongly protein adsorption at the interface, as shown by the surface tension measurements, the acceleration of the nucleation process by these interfaces is less effective than for systems with a lower affinity such as W-SO and W-Air. More generally, the trend between adsorption free energy and nucleation rate is non-monotonic, displaying a maximum at intermediate values of the strength of adsorption (Fig. 6d). To describe quantitatively these effects, we consider a kinetic model of heterogeneous nucleation (Fig. 6c) in order to extend the classical theory of nucleated polymerisation to include heterogeneous nucleation. 39 It is known that a plot of rate against adsorption energy (Volcano plot) displays a non-monotonic behaviour for surface catalysed reactions.</p><p>Here, we extend these ideas to heterogeneous nucleated polymerization. Monomers in bulk initially adsorb to the interface, where they can form surface-bound nuclei. Adsorption at the interface occurs with adsorption constant K 1 = e ε , which in turn is determined by the (dimensionless) free energy of adsorption ε = −∆G ads /(RT ). The oligomerisation step at the surface is associated with the equilibrium constant K 2 . Surface-bound nuclei can undergo structural conversions into growth-competent fibrillar structures, which then detach from the interface, with rate constant k c , and can elongate into mature fibrils. We note that during this conversion/detachment process, bound proteins lose contact with the surface. Therefore, when the affinity for the surface ε is high, this conversion/detachment step becomes energetically unfavourable at high surface affinities.</p><p>As such, the conversion/detachment of surface-bound nuclei has a dependence on the adsorption energy of the form k c ∝ e −ξ•ε (see methods). In addition to transforming into fibrillar structures, surface-bound nuclei are also in equilibrium with nuclei formed in bulk directly from monomers. The oligomerisation step in bulk has equilibrium constant K 3 , while the adsorption constant for the detachment of surface-bound nuclei is denoted with K 4 . Fibril formation can also proceed by the direct conversion of bulk nuclei with rate constant k c . The dynamics of this heterogeneous nucleated polymerization process can be obtained in closed form (see methods) and reveal that the rate of heterogeneous amyloid fibril nucleation is crucially controlled by the parameter K 1 , which describes the propensity of monomers to adhere to the interface and is related to the Gibbs surface excess Γ. The rate of heterogeneous nucleation increases with ε as r ∝ e (n−ξ)ε in limit of weak adsorption, where n is the reaction order for nucleation, but decreases with increasing ε as r ∝ e −ξ•ε for strong adsorption. This non-monotonic behavior (Sabatier's principle) of the nucleation rate is the result of two competing factors at play: the adsorption of molecules to the interface can accelerate aggregation through heterogeneous nucleation, but a very strong adsorption prevents effective release of the formed nuclei into solution. 31,40 To provide a mechanistic interpretation of the results, we extended a previously developed a coarse-grained computer simulation model of heterogeneous nucleation as summarised Fig. 7. Proteins were modelled as hard spherocylinders, which can exist in two distinct states: a soluble ('s') and a fibril-forming 'β' conformation, as in our previous work. 41 These states represent the different conformational ensembles of protein molecules in their non-β-sheet states both in solution and bound to the surface, and the corresponding β-prone states driving amyloid fibrillation.</p><p>Non-specific attractive interactions between proteins in the 's'-state of strength ε ss are mediated by a patch at the cap of the spherocylinder allowing for the formation of micellar-like oligomers. Additionally, the s-state has an affinity for the rigid surface in the simulation, which is set by the parameter ε. Proteins in the β-state interact via an attractive side patch, the strength is set by the parameter ε ββ , which favours parallel packing required for fibril assembly. The strength of interactions between the s-and βstate is set by ε sβ . Importantly, in the simulations proteins have a small probability per unit time to transition from the soluble to the fibril-forming state. This is implemented by attempting the conformation switch from 's' to 'β' conformation, accompanied by a free energy penalty ∆ sβ , for the conformational change. This free energy penalty reflects the loss of conformational entropy associated to the transition to more ordered β-rich conformations. The rigid substrate is modelled as an infinite plane that can bind the attractive caps of the soluble proteins with the interaction strength ε (units of thermal energy k B T ).</p><p>We observe that throughout the simulation, proteins deposit on the rigid surface and oligomer formation can occur. Similarly to what has been previously observed in solution, 41 surface-bound oligomers then take the role of pre-nucleation clusters which facilitate the crossing of the free energy barrier of conversion. Subsequently, a structural transition can occur because the β-sheet-prone state is stabilised sufficiently by interac-tions within the oligomer, as illustrated in Fig. 7b. Further transitions can take place due to favourable interactions leading to the growth of fibrils, which detach from the surface. To quantify the amyloid formation kinetics, we take the formation rate of a β-state dimer as a proxy for the primary nucleation rate.</p><p>As the surface affinity of proteins ε is increased we observe that the nucleation rate is a non-monotonic function of ε (Sabatier's principle), as can be seen in Fig. 7c. Since the protein surface affinity increases, more proteins are deposited increasing the probability to form oligomers which are large enough to serve as pre-nucleation clusters. The reaction order associated to this effect is measured as n ≈ 5 in our simulations. This value is higher than the reaction order 2 obtained from the experimental data. However, it should be noted that the reaction order in simulations depends on the specific choice of interaction parameters (the conversion free energy barrier and the interactions strengths between proteins) that, however, were not designed to describe a particular experimental system. It is rather our intention to explore the molecular mechanisms behind the general behaviour of surface-catalysed amyloid nucleation processes. Coarse-grained simulations allow us to do so in an orthogonal way to kinetic theory, as they assume discrete molecular building-blocks and their interactions energies, as opposed to bulk reaction rates.</p><p>Interestingly, increasing the protein-surface affinity beyond a threshold value causes the nucleation rate to decay, as also observed in our experimental results shown in Fig. 6d. In this high affinity regime, the average oligomer size is typically large enough to allow for a fast first conversion within the oligomer and subsequent fibril nucleation. However, instead of the oligomerisation in the low affinity regime, the rate-limiting step now is the detachment of the protein associated to the conformation conversion, in agreement with the analytical theory results shown in Fig. 6d. As the surface-protein affinity is increased the nucleation reaction is gradually inhibited leading to the non-monotonic behaviour shown in Fig. 7b. Analogous behaviour has been observed in simulations for the nucleation on lipid membranes and protein fibrils themselves. 31,40 The reaction order of the conversion/detachment step from both simulation and experimental data turns out to be 0.5.</p><!><p>We have revealed the mechanism of nucleation in the presence of a surface and explored the role of interfaces in the context of aggregation kinetics. By designing a robust experimental procedure for monitoring aggregation kinetics of three representative proteins (insulin, lysozyme and reconstituted silk fibroin) with different interfaces and surfaceto-volume (S/V) ratios, we have found that two regimes exist. Systems that have high affinity to adsorb to an interface display delayed kinetics, whereas systems that have low affinities have similar behaviour, suggesting a minimum between the two regimes where the rate of protein aggregation is highest. Moreover, the experimental results are corroborated by combining theory and coarse-grained simulations, which yield that at low and high surface interaction strengths, aggregation rates are suppressed, whereas at intermediate strengths kinetics are augmented.</p><p>It was found that by increasing the S/V ratio, protein aggregation is augmented, indicating that nuclei formation initiates at the interface. Moreover, our analysis revealed that the rate constants for elongation and secondary nucleation were independent of the S/V ratio, while the primary nucleation rate varied by an order of magnitude, Fig. 1.</p><p>Furthermore, the addition of pre-formed aggregates to the system bypasses the crucial interfacial nuclei formation step, and subsequently negates any surface effects. More importantly, when enough seeds are added to the solution, the rate of aggregation is the same regardless of the S/V ratio, which was confirmed for all three proteins tested (Figs. 2-3). Moreover, as seen in Fig. 4, AFM micrographs revealed that the samples where there is a water-air interface formed mature fibrils, whereas the samples where a water-oil interface is present mostly consisted of protofibrils and oligomers, clearly indicating that the choice of interface modulates the morphology of the structures formed. Finally, we determined that the tendency of monomeric protein molecules to adsorb at an interface varies. This was done through the combination of the Gibbs adsorption isotherm with the Langmuir adsorption isotherm in order to yield the Langmuir-Szyszkowski equation, where it was found that the larger the propensity of a molecule to adsorb, the slower the self-assembly process for that system (Fig. 6). Conversely, if there is poor adsorption to the interface, then the aggregation half-time is also increased, indicating that a minimum between these two regimes exists. This was attributed to the fact that fibrillar formation occurs predominantly through a heterogeneous nucleation pathway and thus an interface is necessary, however, too high of an adsorbance does not allow for the final kinetic step, i.e. the conversion of nuclei to their fibrillar state which is in solution (Fig. 6). This effect was also corroborated through coarse-grained simulations (Fig. 7), where a maximum on the nucleation rate as a function of the adsorption free energy was measured. We believe that this systematic and biologically relevant study can help with gaining better mechanistic insights into protein self-assembly at the interface in vivo.</p><!><p>To investigate the effect of surface area on protein aggregation, three different protein stock solutions were prepared. Fresh hydrochloric acid solution at pH 1.3 and pH 1.6 was prepared every time to dissolve insulin (from bovine pancreas, Sigma Aldrich) and lysozyme (from chicken hen egg white, Sigma Aldrich), respectively. Lysozyme was dialyzed against Mili-Q water for 2 days at 4 • C prior to use. Insulin was used without any purification. The acid solution was filtered using 0.22 µm filter paper to remove any impurities which might affect the aggregation kinetics. After that, insulin and lysozyme were dissolved into acid solution at the concentration of 2 mg/mL and 200 mg/mL respectively. For reconstituted silk fibroin (purification process mentioned below), the purified batch was diluted into 10 mg/mL by using Mili-Q water. Concentration of these solution was checked with NanoDrop (Thermo Scientific) and used as stock solutions.</p><p>Additionally, in order to determine the effect an interface may have on protein selfassembly, Mineral oil (BioReagent, light oil, Sigma Aldrich), which is an oil that is comprised of a mixture of alkanes, and Silicon oil (viscosity 50 cSt (25 • C), Sigma Aldrich) were added to the protein solutions via the experimental procedure mentioned below.</p><!><p>Bombyx mori silk cocoons (Mindsets (UK) Limited) were used to extract the silk fibroin protein by a well-established protocol. 42 Initially, the cocoons were cut into pieces and placed in a beaker containing a solution of 0.02 M sodium carbonate. This was then boiled for 30 minutes, ensuring that the sericin that is present within the silk fibres, dissolved, while the insoluble fibroin remained in the beaker. The fibroin was then removed from the beaker, rinsed with cold water three times and left overnight to dry out. A 9.3 M lithium bromide solution was prepared and added to the dried silk fibroin in a 1:4 ratio of silk fibroin to lithium bromide. The mixture was heated to 60 • C and left for 4 hours, resulting in the silk fibroin dissolving in the lithium bromide. LiBr was removed from the solution by placing the mixture in a 3 kDa dialysis tube. This in turn was placed in a beaker containing ultrapure water, while the use of a large magnetic stir bar with a magnetic stir-plate was employed to ensure mixing. The water was changed a total of 6 times in 48 hours.</p><p>Finally, the silk fibroin solution was removed from the dialysis tube and placed in Eppendorf tubes. These were then centrifuged at 9000 r.p.m. at 4 • C for 20 minutes in order to remove small impurities. The process was repeated twice and the final solution was stored at 4 • C. All experiments were conducted within 2 weeks of extracting and purifying the silk fibroin to ensure no gelation had occurred.</p><!><p>For the experiments involving the investigation of surface-to-volume ratio on protein aggregation kinetics the following procedure was established and followed. As each of the wells in the 96-well plate (Corning 3881, Half-area) used was cylindrical, the crosssectional surface area remains the same throughout the well. Consequently, by pipetting different volumes of protein solution into the 96-well plate (80, 100, 120 and 140 µL), the surface-to-volume ratio was decreased respectively. This is schematically represented in the top panels of Fig. 1a. The plates used in all experiments are specifically coated so that proteins do not adsorb to the walls of the wells, and thus the interactions of the walls are considered to be negligible in this study. Additionally, in order to monitor the selfassembly process (i.e. the transition of protein monomers to β-sheet structures), 40 µM Thioflavin T (ThT), which is a molecule that is known to fluoresce only when in contact with β-sheets, was added to the aqueous solution. ThT fluorescent increase was thus monitored as a function of time using a microplate reader (FLUOstar-BMG labtech), where the intensity directly correlates to the amount of fibrils present in the solution.</p><p>The temperature during the measurement was kept at 50 • C for insulin (except where otherwise stated), 60 • C for lysozyme and 37 • C for silk fibroin and all measurements were performed without agitation. Furthermore, for the experiments involving the investigation of protein aggregation in the presence of a liquid-liquid interface, the same procedure as described above was employed, with a minor difference. 30 µL of an oil phase, which was either mineral oil or silicon oil, was carefully added on top of the protein solution. Again, different aqueous volumes were used (80, 100 and 120 µL respectively), which resulted in a change of the surface-to-volume ratio. This is schematically represented in the top panels of Fig. 1d. 40 µM ThT was also added to the aqueous phase to monitor protein aggregation by observing the increase of fluorescent intensity as a function of time. Moreover, in order to determine whether the meniscus of the different interfacial systems could affect the surface area, a picture of the three different interface systems was takes (Fig. S2f). As seen in this image, it is evident that the curvature of the meniscus is quite similar for all three interface systems and so any effect that this may have on the surface area can be neglected from our current setup. The variation in meniscus area between the water-air and water-MO system is 1.19%, while for the water-air and water-SO systems it was found to be 1.05%.</p><p>Finally, different concentration of seeds (obtained from incubating 10% w/v protein solutions for one month at 65 • C) were added to the 4 volumes of protein solution (80, 100, 120 and 140 µL), as is schematically depicted in the top panels of Fig. 1ad. Following the addition of 40 µM ThT, aggregation kinetics were monitored using a microplate reader.</p><!><p>A concentration of 1 mg/mL 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) was used in all lipid based experiments. DOPC was dissolved in mineral oil (MO) by means of a variant to the film hydration method. In brief, 1 mg/mL of lipid was initially dissolved in chloroform and was left to evaporate overnight in a fume hood. Following this, 1 mL of MO was added to the dried film and stirred for 2-3 minutes. This lipid-MO solution was now used, rather than plain MO, when added on top of the protein solution in the 96-well plate. A 2 mg/mL solution of insulin was prepared and the kinetics of the W-MO(lipids) interface was investigated.</p><!><p>The kinetic data obtained from the protocol was analyzed using Amylofit. 35 All data were normalized with the minimum and maximum fluorescence intensity, and half time was calculated from the time at which half the protein that is present, initially as monomer, has aggregated, i.e. the time at which the normalized intensity reaches at 0.5. Each experiment was repeated four times and then averaged before fitting, while a secondary nucleation dominated model was assumed.</p><!><p>To characterise the aggregates formed by incubation at different interfaces, protein solutions were collected from the 96-well microplate after incubation and diluted 100 times. 20 µL aliquot of the diluted solution was then placed on a freshly cleaved mica and incubated at room temperature for 30 min. The mica was rinsed with Mili-Q water 3 times and dried with compressed nitrogen gas. AFM images were taken with Park NX10 (Park Systems) using non-contact mode.</p><!><p>To measure interfacial tension (IFT) for different interfaces, pendant drop tensiometry (FTA1000B, First Ten Angstroms) was used. Insulin monomer solution with varying concentrations was prepared and filled in a dispensing system consisting of a 100 µL glass syringe and a 27 gauge needle. The drop was generated in either air or an oil bath by carefully pushing the syringe. 20 pictures of droplet were taken at 0.5 seconds interval and the images were analyzed with a FTA32 (First Ten Angstroms) software in order to calculate the IFT. The water-air interface was stabilized for 1,000 seconds before image acquisition. For water-oil interfaces, a quartz cuvette was used to surround the droplet with oil, and droplets were stabilized at the oil interface for at least 1 hour before image acquisition.</p><!><p>In the coarse-grained simulation model employed here, proteins are represented as hard spherocylinders of diameter σ = 2nm and length 4σ. The spherocylinders are equipped with interaction patches allowing them to self-assemble into clusters of various morphologies. Two different conformational states are considered: the soluble state ('s') and the β-sheet-rich fibril-forming state ('β'). The 's' state represents the conformational ensemble of protein molecules in their soluble states, which have a propensity to self-assemble into micellar oligomers. The pair-wise interaction potential between tips of two s-proteins is given by V ss (r) = ε ss r −6 with cutoff r c = 1.5σ, where in this work ε ss = 6k B T . These oligomers play the role of pre-nucleation clusters in the two-step nucleation pathways explored with this model. The β-state serves as a low entropy conformational state with a high propensity to assemble into well-defined and thermodynamically fibrillar structures.</p><p>Fibril forming proteins are modelled to interact via an aisotropic patch interaction of strength ε ββ = 30k B T , which drives parallel alignment and fibril formation. The cross state interaction between soluble and fibril-forming proteins is also anisotropic and has strength ε sβ = ε ss + 1k B T in order to modulate the surface coverage.</p><p>. Consequently, if free monomers absorb at the surface with adsorption constant K 1 ∝ e ε , where ε = −∆G ads /(RT ) is the free energy of adsorption (in units of thermal energy RT ), then detachment of nuclei depends on the adsorption energy ε as K 4 ∝ e −nε . The conversion/detachment constant depends on (dimensionless) free energy of conversion ε conv and detachment ε detach = ξ • ε, ξ is the absolute ratio between nuclei detachment free energy and monomer adsorption free energy, implies K c ∝ e −(εconv+ε detach ) ∝ e −ε detach = e −ξ•ε . Detachment from the surface is therefore hindered at high ε. The rate of nucleation is</p><p>We work in a regime when there is an excess of monomers in bulk, such that we can set m m tot throughout, where m tot is the total protein concentration. We then impose conservation of the total concentration s tot of surface sites, yielding</p><p>Eq. ( 2) can be solved numerically to yield s in terms of s tot ; here we seek for an approximate analytical solution which is obtained by focusing on two limits. In the limit of low surface coverage (K 1 m tot 1), the first term on the right-hand side of Eq. ( 2)</p><p>dominates over the nonlinear one, yielding</p><p>In the limit when the surface is saturated with monomers (K 1 m tot 1), the dominant term in Eq. ( 2) is the nonlinear term, yielding</p><p>The true solution to Eq. ( 2) interpolates between these limiting values. We therefore obtain the following expressions for the rate of nucleation</p><p>where</p><p>is the surface coverage. In the limit k c = 0 (only bulk nuclei convert into fibrils) we recover, as expected, the rate of homogeneous nucleation</p><p>In the opposite limit k c = 0 (only surface-bound nuclei convert into fibrils) we obtain</p><p>Using K 1 = e ε and k c ∝ e −ξ•ε we find</p><p>In the limit of weak adsorption (K 1 m tot 1), the rate of nucleation increases with increasing ε. Therefore, increasing the adsorption free energy causes more monomers to adsorb at the interface, resulting in overall faster heterogeneous nucleation. In the case of strong adsorption (K 1 m tot 1), the rate of heterogeneous nucleation r decreases with increasing ε as r ∝ e −ξ•ε . Even though the interface is fully covered with monomers, the conversion/detachment step is hindered, resulting in an overall smaller rate of heterogeneous nucleation.</p><!><p>Experimental and simulations data in Figs. 6d and 7c are fitted to the model in Eqs. ( 1) and ( 2) with the fitting parameters α, β, γ defined as</p><p>where k 0 is defined through k c = k 0 e −ξ•ε . In terms of these parameters Eqs. ( 1) and ( 2)</p><p>become</p><p>where y = s/s tot . In the limits K 1 m tot = αe ε 1 and K 1 m tot = αe ε 1 the solution to ( 11) is</p><p>In a logarithmic plot, we obtain two straight lines</p><p>where a = ln(α), b = ln(β), c = ln(γ). The fitting parameters for Figs. 6d and 7c are summarised in Table 1.</p><p>Table 1: Fitting parameters for Figs. 6d and 7c.</p><p>Parameter Experiments (Fig. 6d) Simulations (Fig. 7c</p><p>While Eq. ( 16) is a very useful working definition for determining Γ experimentally, to relate Γ to adsorption energy we need another formula for Γ that accounts explicitly for the amount of protein adsorbed at the interface. We consider a binary system consisting of the solvent, denoted as 1, and protein, denoted 2. The bulk concentrations of the species 1 and 2 are c 1 , c 2 (lower bulk phase) and c 1 , c 2 (upper bulk phase). The number of moles of species i at the surface, denoted n σ i , is obtained by subtracting from the total number of moles in the system, n i , the amount of moles in the lower (n i ) and upper (n i ) bulk phases:</p><p>where V is the volume of the lower bulk phase and V is the volume of the upper bulk phase. Using V + V = V and eliminating V we obtain</p><p>Dividing by the area of the interface S we obtain the surface excess of species 2 relative to the solvent 1:</p><p>Assuming the concentrations of solvent and solute to be negligible in the outer phase, c 1 c 2 0, Eq. ( 20) simplifies to</p><p>The second term in the brackets is negligible if the number of moles of solvent adsorbed is negligible relative to total number of moles of the solvent. In this limit Eq. ( 21) simplifies to:</p><p>We see that the surface excess is the difference between the total protein concentration, c 2 , and the bulk protein concentration c 2 after adsorption has been established. For our experimental system, we have c 2 = m tot , while c 2 can be calculated using the Langmuir adsorption isotherm as:</p><p>where s tot is the total concentration of surface sites and Θ = K 1 m tot /(1 + K 1 m tot ) is the surface coverage. Therefore</p><p>where Γ(∞) = (V /S) s tot is the maximal surface excess and S/V is the surface-to-volume ratio. Combining Eq. ( 24) with Eq. ( 16) yields the following formula for the change of surface tension γ with added protein concentration γ = γ 0 − Γ(∞)RT log(1 + K 1 m tot ), (25) which is known as the Langmuir-Szyszkowski equation.</p>

ChemRxiv

UPR-Mediated Membrane Biogenesis in B Cells

The unfolded protein response (UPR) can coordinate the regulation of gene transcription and protein translation to balance the load of client proteins with the protein folding and degradative capacities of the ER. Increasing evidence also implicates the UPR in the regulation of lipid synthesis and membrane biogenesis. The differentiation of B lymphocytes into antibody-secreting cells is marked by significant expansion of the ER, the site for antibody synthesis and assembly. In activated B cells, the demand for membrane protein and lipid components leads to activation of the UPR transcriptional activator XBP1(S) which, in turn, initiates a cascade of biochemical events that enhance supplies of phospholipid precursors and build machinery for the synthesis, maturation, and transport of secretory proteins. The alterations in lipid metabolism that occur during this developmental transition and the impact of membrane phospholipid restriction on B cell secretory characteristics are discussed in this paper.

upr-mediated_membrane_biogenesis_in_b_cells

1. Introduction<!>2. Lipid Supply and Demand<!>3. Phosphatidylcholine Synthesis<!>4. A “Physiologic” UPR<!>5. XBP1(S), Lipid Synthesis, and ER Biogenesis<!>6. Phosphatidylcholine Synthesis and UPR Signaling

<p>Activated B lymphocytes proliferate and proceed along distinct developmental pathways that determine their function and fate. Specifically, responding B cells can rapidly differentiate in extrafollicular sites into short-lived antibody-secreting cells that predominantly secrete IgM antibodies [1]. Alternatively, responding B cells can enter germinal centers, undergo somatic hypermutation and isotype switching, and then become memory B cells or long-lived antibody-secreting cells [2]. Extrinsic factors, including the nature of the antigen and T cell help in the form of membrane-bound molecules and soluble cytokines, play key roles in regulating B cell responses. However, intrinsic signals are also pivotal in directing the fate of responding B cells as evidenced by the critical role of the unfolded protein response (UPR) transcription factor XBP1(S) in driving the differentiation of antibody-secreting cells [3, 4], the effectors of humoral immunity. Here, we discuss the current understanding of the relationship between the UPR, lipid biosynthesis and organelle biogenesis in activated B cells.</p><!><p>B lymphocytes proliferate and differentiate into antibody-secreting cells upon interaction with specific antigen or certain Toll-like receptor (TLR) ligands. When B cells are stimulated to enter the cell cycle and proliferate, the mechanisms that control the membrane phospholipid supply in rapidly dividing cells are engaged. The division of one cell into two daughter cells requires a doubling of membrane content during cell cycle progression [5]. Phosphatidylcholine (PtdCho) is the major membrane phospholipid in mammalian cells and is a precursor to the two other most abundant membrane phospholipids, sphingomyelin (SM) [6] and phosphatidylethanolamine (PtdEtn) [7]. PtdCho and the other phospholipids accumulate in a periodic manner during S phase, coincident with DNA synthesis. The net increase in membrane PtdCho results from an interaction between cell cycle-dependent oscillations in the rates of PtdCho biosynthesis and degradation. PtdCho synthesis is stimulated very early during G1 phase [8–10], but is accompanied by rapid PtdCho turnover. Two phospholipases have been implicated in the PtdCho turnover associated with cell cycle progression, the group VIA calcium-independent phospholipase A2 [11] and the neuropathy target esterase [12]. Near the G1/S transition, PtdCho turnover is diminished substantially, yielding a net increase in membrane PtdCho. Toward the latter part of the cell cycle, prior to cytokinesis, PtdCho synthesis is downregulated [5]. This cyclic variation in the supply of membrane phospholipid for cell proliferation is maintained in the absence of differentiation.</p><p>B cells are unique, however, and in addition to proliferation also undergo a subcellular membrane expansion as they differentiate into antibody-secreting cells after stimulation. There is a major increase in synthesis and secretion of immunoglobulin (Ig) heavy (H) and light (L) chains [13]. Nascent Ig chains are cotranslationally translocated into the endoplasmic reticulum (ER), an oxidizing, calcium-rich environment containing many resident molecular chaperones and folding enzymes [14]. Within this specialized protein folding compartment, H and L chains are assembled into functional antibodies. Induction of high-rate Ig synthesis during the differentiation process is accompanied by expansion of the rough ER membrane, at least 3- to 4-fold in surface area and volume [15, 16]. Thus, both proliferation and differentiation require an increased supply of phospholipids to fuel membrane and organelle biogenesis. To meet this demand, the synthesis of phospholipids, particularly PtdCho, increases when B cells are activated [15, 17].</p><!><p>The predominant means for PtdCho biosynthesis in mammalian cells proceeds via the three steps of the cytidine diphosphocholine (CDP-choline) pathway [18] (Figure 1). First, choline kinase (CK) phosphorylates choline in the presence of ATP to yield phosphocholine. CKα and CKβ are two isoforms which are soluble proteins found in the cytosol [19, 20]. Second, choline cytidylyltransferase (CCT) converts phosphocholine to CDP-choline in the presence of CTP, and this is the rate-limiting step in the pathway [21]. In every cell type examined thus far, including B cells [17], CCT catalyzes the slow step in the pathway and thereby determines the rate of PtdCho formation. Comparatively small amounts of CDP-choline are found in cells, in relation to other phospholipid precursors, as CDP-choline is utilized almost immediately after it is made. CCT, including all mammalian isoforms, transiently associates with the ER membrane and the lipid composition of the ER membrane governs CCT association and activity [22]. Elevated expression of CCT stimulates PtdCho synthesis but often does not result in an increased amount of cellular PtdCho in most proliferating cells due to compensatory elevation of PtdCho turnover mediated by phospholipases [23, 24]. Third, the phosphocholine moiety of CDP-choline is transferred to diacylglycerol (DAG), producing PtdCho. This final step can be catalyzed by either cholinephosphotransferase (CPT1) or choline/ethanolaminephosphotransferase (CEPT1), a bifunctional enzyme that can synthesize both choline- and ethanolamine-containing phospholipids. The CPT enzymes are integral membrane proteins, and the CPT1 is found with the Golgi apparatus while the CEPT1 associates with the ER [25, 26]. Here, we refer to the activities of CPT1 [27] and CEPT1 [28] collectively as CPT activity. The locations of the CPT enzymes designate the subcellular sites of membrane biogenesis; however, enforced overexpression of CPT activity does not enhance PtdCho synthesis [29, 30]. Rather, the supply of CDP-choline and DAG determine the amount of PtdCho. Thus, elevated expression of the CPT enzymes can be considered as a marker for Golgi and/or ER membrane expansion, but not necessarily as a driver of membrane phospholipid synthesis.</p><p>In lipopolysaccharide- (LPS-) stimulated splenic B cells, CK activity remains fairly constant, CCT activity modestly increases ≈2-fold, and CPT activity increases ≈6-fold [15]. These modulations of the CDP-choline pathway enzymes in LPS-stimulated splenic B cells correlate with a 6- to 7-fold increase in PtdCho synthesis [15, 31]. Our studies using the CH12 B cell lymphoma indicate that increased CCT activity is pivotal for enhanced flux through the CDP-choline pathway in LPS-stimulated B cells [17]. In this system, the CCT expression and enzyme specific activity do not increase when assayed under optimal in vitro conditions following LPS stimulation. However, radiolabeling experiments of stimulated cells demonstrate that the formation of CDP-choline is substantially enhanced, indicating allosteric activation of CCT by membrane lipids. Indeed, microsomal lipids isolated from stimulated cells contain an elevated amount of DAG and significantly stimulate the activity of purified recombinant CCT, compared to lipids isolated from unstimulated cells. Thus, in this case, the formation of DAG is key to stimulation of PtdCho synthesis: first, by activating CCT, and second, by providing substrate for the CPT enzymes. The CCT, in turn, governs the fate of the DAG as DAG is incorporated either into phospholipid under permissive CCT conditions or into triacylglycerol (TAG) when the CCT activity is reduced [32] (Figure 1).</p><!><p>ER stress occurs when the load of client proteins exceeds the folding capacity of the ER, a condition that can be catastrophic if unresolved. To rebalance load with capacity in the ER, thereby relieving ER stress, the UPR can slow the flow of nascent polypeptides into the ER lumen and enhance the ER machinery needed for folding and/or disposal of client proteins [33, 34]. The mammalian UPR is orchestrated by a trio of signaling pathways that are separately initiated by three ubiquitously expressed ER transmembrane proteins: PERK (PKR-like ER kinase) [35, 36], ATF6 (activating transcription factor 6) α and β [37, 38], and IRE1 (first identified in a yeast mutant with inositol requiring phenotype) α and β [39, 40]. The activation status and role of each UPR pathway has been examined during the differentiation of antibody-secreting B cells.</p><p>The PERK protein possesses a serine/threonine kinase domain in its cytoplasmic region through which it mediates translational attenuation [35, 36]. Upon activation, PERK phosphorylates the α subunit of eIF-2 (eukaryotic initiation factor-2) on serine 51, thereby impeding formation of translation initiation complexes and slowing the flow of nascent polypeptides into the ER [41, 42]. PERK does not appear to be activated during the differentiation of antibody-secreting B cells [43, 44]. In support of this concept, studies of gene-targeted mice reveal that the PERK pathway is dispensable for antibody secretion [43].</p><p>ATF6α and ATF6β are type II ER transmembrane proteins [37, 38]. Upon UPR activation, ATF6 traffics from the ER to the Golgi complex where it is clipped by the Site-1 and Site-2 proteases [45, 46]. Once liberated from the membrane by this process of intramembrane proteolysis, the cytosolic N-terminal domain of ATF6 moves into the nucleus where it functions as a transcriptional activator of genes encoding ER resident molecular chaperones, folding enzymes and components involved in ER-associated degradation (ERAD) of misfolded proteins [37, 38, 47–49]. While ATF6α and β are both functional, only ATF6α appears essential for induction of ER stress responsive genes and survival of cells subjected to ER stress conditions [48, 49]. Overexpression of active ATF6α is sufficient to drive synthesis of fatty acids and phospholipids and to induce expansion of rough ER [50], suggesting that this UPR pathway might participate in the differentiation of antibody-secreting B cells. Indeed, ATF6α is activated in LPS-stimulated B cells [43, 51, 52]. However, recent studies of ATF6α-deficient mice indicate that ATF6α, like PERK, is dispensable for the differentiation of antibody-secreting B cells (Brewer et al., manuscript in preparation).</p><p>The IRE1 proteins contain a serine-threonine kinase module and a C-terminal endoribonuclease domain in their cytoplasmic regions [39, 40]. Upon activation, IRE1 executes site-specific cleavage of Xbp1 (X-box binding protein 1) mRNA. A 26-nt intron is excised and an undefined mechanism then ligates the resulting 5′ and 3′ fragments, yielding a spliced Xbp1 mRNA with an altered reading frame [53–55]. Both unspliced and UPR-spliced Xbp1 transcripts encode basic leucine zipper (bZIP) transcription factors, XBP1(U) and XBP1(S), respectively. The XBP1(S) factor exhibits enhanced transactivating capacity and greater stability as compared to XBP1(U) [53–56]. Like ATF6α, XBP1(S) is sufficient to upregulate synthesis of fatty acids and phospholipids and to drive expansion of rough ER [30, 50]. Xbp1 is essential for optimal induction of genes encoding proteins that function throughout the secretory pathway and for proper development of the ER in a variety of specialized secretory cell types [57, 58]. When B cells are stimulated to secrete antibody, Xbp1 mRNA increases and undergoes UPR-mediated splicing to yield XBP1(S) [3, 52, 53], a factor required for the generation of antibody-secreting B cells [3, 4]. Thus, the physiologic UPR of activated B cells features the IRE1/XBP1 pathway.</p><!><p>Xbp1 is required for embryonic development [59]; thus, the role of this UPR transcription factor in lymphocytes was first investigated using the Rag-2 complementation system [4]. Those experiments revealed that XBP1-deficient B cells are markedly defective in antibody secretion in vivo in response to immunization and in vitro in response to LPS. Importantly, it was shown that XBP1(S), but not XBP1(U), effectively restores the ability of XBP1-deficient B cells to secrete antibody in response to LPS in vitro [3] and is sufficient to drive ER expansion [30, 58]. More recently, the Cre-loxP system has been employed for selective deletion of Xbp1 in B cells and studies using this system have corroborated the earlier findings [60, 61]. Using this system, the abundance of PtdCho was shown to increase in LPS-stimulated XBP1-deficient B cells, but to a lesser degree than in wild-type cells [62]. The levels of PtdCho, SM, and phosphatidylinositol were significantly reduced in activated XBP1-deficient B cells, but PtdEtn, phosphatidylserine, and phosphatidylglycerol were similar to corresponding amounts in wild-type activated B cells. In addition, a meager, but discernible, expansion of the rough ER was observed in LPS-stimulated XBP1-deficient B cells [62].</p><p>PtdCho is most drastically affected by XBP1 deficiency because it is the most abundant phospholipid of the ER membranes. SM is derived directly from PtdCho, where the phosphocholine headgroup of PtdCho is transferred to ceramide by the SM synthase [63] (Figure 1). Thus, a reduction in PtdCho availability would be reflected by a reduction in SM. The pathway for PtdCho conversion to PtdEtn is not as direct, however, and a second pathway of PtdEtn synthesis via CDP-ethanolamine can bypass a deficiency in PtdCho [64]. Thus, the amount of PtdEtn is less affected following activation of XBP1-deficient B cells and PtdEtn increases to almost the same extent as in activated wild-type B cells. On the other hand, the enforced expression of XBP1(S) in NIH-3T3 fibroblasts leads to a substantial increase in PtdEtn [30], augmenting the XBP1(S)-independent mechanism(s) of lipogenesis. The de novo synthesis of ceramides, key precursors in SM production, is upregulated upon LPS stimulation [65] and contributes to the increase in SM. Inhibition of ceramide formation impairs ER expansion and protein glycosylation in the ER lumen [65], suggesting a link among these processes. These data establish that XBP1 is required for maximal increases in PtdCho, SM, and rough ER in LPS-stimulated B cells, but the mechanisms by which XBP1 mediates these events remain to be elucidated. The scheme in Figure 1 shows a cascade of biochemical events which illustrates how XBP1(S) stimulation of fatty acid synthesis [50] is a key feature that drives membrane phospholipid expansion in B cells [17]. Furthermore, these data suggest that XBP1-independent mechanisms, as yet undefined, must also contribute to the regulation of PtdCho synthesis and ER biogenesis during the differentiation process.</p><p>It has been proposed that the escalation of Ig synthesis in differentiating B cells taxes the protein folding machinery of the ER and, consequently, triggers the UPR [3]. This model was supported by an experiment showing reduced induction of XBP1(S) in B cells that had undergone ex vivo Cre-mediated deletion of IgH chain prior to LPS stimulation [3]. In contrast, recent studies have shown strong induction of XBP1(S) in μ s −/− B cells stimulated with LPS [60, 62], indicating that increased synthesis of soluble μ H chains is not a prerequisite for UPR activation. In keeping with these data, we previously showed that synthesis of XBP1(S) precedes induction of maximal Ig translation in LPS-stimulated CH12 B cells [52], indicating that the IRE1/XBP1 pathway is activated at an earlier stage of the differentiation process. What then is the signal(s) for UPR activation in stimulated B cells? This remains a fundamental question, and its answer is integral to understanding the mechanisms that drive development of antibody-secreting B cells.</p><!><p>Mammals express three CCT isoforms that are similar in enzymatic activity and regulation. CCTα is encoded by the Pcyt1a gene whereas CCTβ2 and CCTβ3 are encoded by alternatively spliced transcripts from the Pcyt1b gene [66]. CCTα is predominantly expressed in most tissues, including B cells [17], and is required for early embryonic development [67]. Tissue-specific deletion of the Pcyt1a gene using the Cre-loxP system has revealed critical roles for CCTα in specialized secretory cells, including surfactant lipid production and secretion by alveolar epithelial cells [68], assembly and secretion of lipoproteins by hepatocytes [69], and cytokine secretion by activated macrophages [70]. We recently showed that selective deletion of CCTα significantly hampers the ability of B cells to upregulate PtdCho synthesis upon stimulation, and interestingly, this correlates with heightened induction of the IRE1/XBP1 branch of the UPR [31].</p><p>When challenged with a T cell-dependent protein antigen, the animals harboring CCTα-deficient B cells were unable to produce normal levels of IgG but secreted hyperlevels of IgM [31]. The correlation between the reduced PtdCho synthesis and elevated IgM secretion in the CCTα-deficient B cells was counterintuitive, however, based on the implied need for membrane PtdCho expansion during plasma cell differentiation. Investigation of the UPR components revealed that the impaired production of PtdCho triggers IRE-mediated splicing of Xbp1 mRNA early after activation, thereby promoting differentiation of IgM-secreting cells. The inability of CCTα-deficient B cells to undergo isotype switching correlates with a proliferation defect. However, blocking proliferation by a different mechanism did not elicit XBP1(S) activation, supporting the idea that the early and potent induction of XBP1(S) by PtdCho deficiency in CCTα-deficient B cells accelerates and augments the transition into antibody secretion. From these observations, we propose that the IRE1/XBP1 branch of the UPR responds to increased demand for phospholipids as well as increased demand on the protein folding capacity of the ER (Figure 2). In agreement, restriction of either PtdCho [71] or fatty acid synthesis [72] has been shown to elicit activation of UPR components in other systems. It is intriguing to speculate that lipid supply might function as a metabolic cue for induction of the IRE1/XBP1 pathway in activated B cells.</p>

PubMed Open Access

Textile sensors platform for the selective and simultaneous detection of chloride ion and pH in sweat

The development of wearable sensors, in particular fully-textile ones, is one of the most interesting open challenges in bioelectronics. Several and significant steps forward have been taken in the last decade in order to achieve a compact, lightweight, cost-effective, and easy to wear platform for healthcare and sport activities real-time monitoring. We have developed a fully textile, multi-thread biosensing platform that can detect different bioanalytes simultaneously without interference, and, as an example, we propose it for testing chloride ions (Cl − ) concentration and pH level. The textile sensors are simple threads, based on natural and synthetic fibers, coated with the conducting polymer poly(3,4-ethylenedioxythiophene):poly(styrene-sulfonate) (PEDOT:PSS) and properly functionalized with either a nano-composite material or a chemical sensitive dye to obtain Cl − and pH selective sensing functionality, respectively. The single-thread sensors show excellent sensitivity, reproducibility, selectivity, long term stability and the ability to work with small volumes of solution. The performance of the developed textile devices is demonstrated both in buffer solution and in artificial human perspiration to perform on-demand and point-of-care epidermal fluids analysis. The possibility to easily knit or sew the thread sensors into fabrics opens up a new vision for a textile wearable multi-sensing platform achievable in the near future.

textile_sensors_platform_for_the_selective_and_simultaneous_detection_of_chloride_ion_and_ph_in_swea

<!>Material and methods<!>Results and discussion<!>Artificial sweat characterization.<!>Conclusion

<p>Wearable sensing technologies are attracting a growing academic and industrial interest thanks to the driving force of market demand and the prospective large impact on real life. Personalized healthcare and point-of-care medical assistance, together with fitness, represent the main fields of applications meeting the highest request for wearable biosensors.</p><p>The convergence of technical advancements in complementary research areas is making possible the development of wearable devices such as: (i) energy-harvesting devices [1][2][3] , (ii) energy-storage devices [3][4][5] and (iii) sensors 6,7 .</p><p>On one hand, wearable physical sensors exhibit a high readiness level and include strain sensors to monitor physical movement 8 , motion 9,10 , heartbeat 11 or respiratory rate evaluation 12 and pressure sensors based on memory systems 13 , textile devices 14,15 or skin-based architectures 16 . Moreover, bio-potential sensitive sensors can record skin potential that allows, for example, electrocardiograms [17][18][19] and electromyography 20 .</p><p>On the other hand, wearable chemical sensors are at an early stage of development even if they represent a powerful tool to monitor human physiological parameters in a real-time, non-invasive and accurate manner. A number of new chemical sensors are proposed in the literature for the detection of biochemical markers like dopamine 21 , adrenaline 22,23 , cortisol 24 , glucose 25 , lactate 26 , phenolic compounds 27,28 , and electrolytes [29][30][31][32] with the aim of providing new tools for monitoring human health, physical exertion, fatigue and mental accuracy. An interesting approach has been proposed by Gao et al. based on a fully integrated and mechanically flexible sensor array for multiplexed sweat analysis, which selectively and simultaneously measures perspiration electrolytes (sodium and potassium ions) and metabolites (glucose and lactate) 33 . A fully textile sensor device [34][35][36][37][38][39] embodies the most advanced technological frontier to achieve complete flexibility, portability, non-invasiveness and lightweight towards continuous human body monitoring. Their wide spread use, in addition to a well-established OPEN 1 Department of Physics and Astronomy, University of Bologna, Viale Berti Pichat 6/2, 40127 Bologna, Italy. 2 Department of Industrial Chemistry, University of Bologna, Viale Risorgimento 4, 40136 Bologna, Italy. * email: luca.possanzini2@unibo.it</p><!><p>Reagents and materials. CLEVIOS PH 1000 suspension (PEDOT:PSS) was purchased by Haraeus (Hanau, Germany). (3-glycidyloxypropyl)trimethoxysilane (GOPS), potassium nitrate, silver nitrate, potassium hydroxide, acetic acid, boric acid, 85% phosphoric acid, phosphate buffered saline, sodium dihydrogen phosphate, L-histidine, sodium chloride, potassium chloride, 3,4-ethylenedioxythiophene (EDOT), bromothymol blue (BTB) were purchased from Sigma-Aldrich (St. Louis, USA). Ethylene glycol (EG) and polyethylene glycol were obtained from Carlo Erba (Cornaredo, Italy). All chemicals were of reagent grade. The Universal Buffer solution was prepared with 0.01 M H 3 BO 3 , 0.01 M H 3 PO 4 , and 0.01 M CH 3 COOH in 0.1 M KNO 3 . The artificial sweat formulation (ISO pH 5.5) was made up with 0.05% w/v L-histidine, 0.22% w/v NaH 2 PO 4 , and 0.5% w/v NaCl in distilled water.</p><p>Apparatus. Scanning electron microscope (SEM, Cambrige Stereoscan 360) images have been acquired to investigate the microstructure, morphology, composition and coating features of the thread, both before and after the functionalization. The same instrument, coupled with an energy-dispersive spectrometer, was used to carry out the Energy-Dispersive X-ray Spectroscopy to determine the presence and distribution of the chemical elements in our samples. The accelerating voltage was 20 kV.</p><p>The electrochemical depositions were done using the Metrohm Autolab (Origgio, Italy) potentiostat in a three-electrodes cell, using the PEDOT:PSS-coated threads, a Pt wire and an aqueous saturated calomel electrode (SCE) as working, counter and reference electrodes, respectively. The electrical resistance of the threads was measured using a four-probe setup with the Keithley (Cleveland, USA) 2400 whereas, for the sensor characterization, the Keysight (Santa Rosa, USA) B2912A has been used to apply the required potentials. The accurate pH level has been measured using a commercial pH meter equipped with a pH glass electrode (XS Instruments pH 7). All the measurements were carried out in air and at room temperature.</p><p>Fabrication of thread sensors. A conductive solution was prepared mixing CLEVIOS PH1000 suspension, ethylene glycol (secondary dopant), 3-glycidyloxypropyltrimethoxysilane (cross linker) and polyethylene glycol in the following volumetric ratio 77.1:13.8:0.9:8.2. After mixing by sonication CLEVIOS PH1000 suspension, ethylene glycol (secondary dopant) and 3-glycidyloxypropyltrimethoxysilna (cross linker) for 10 min, polyethylene glycol was added in the solution and kept under vigorous stirring for 20 min. The conductive ink was then warmed up in an oven at 70° aiming to increase the solution viscosity. Three different kinds of thread, that differ in the basic material composition, were employed as substrate: both natural (cotton: COT, and silk: SILK), and synthetic (polyester: POL). In addition, we introduced a second cotton yarn, named here COT2, to investigate the possible effects of a different fibers arrangement. It results in a tighter thread with a Metric Count of 2/30 instead of 2/50. The Metric cotton Count represents an indirect measure of the linear density and it is the number of threads hanks long 1 km to reach the mass of 1 kg. In order to deposit the polymeric mixture, the yarns were washed with neutral soap and distilled water and then immerged in the viscous solution and rolled around a stick with a constant velocity. The procedure was repeated until the thread surface was evenly covered. The modified thread electrodes were left to completely dry in an oven before proceeding with the next fabrication steps. The total amount of conductive polymer deposited on 2 cm long thread is about (0.5 ± 0.2) mg and it might slightly vary for different fibers. A proper functionalization process has the purpose to make the conductive textile thread highly sensitive and selective towards the target biocompound.</p><p>The chloride ion thread sensor was realized by electrochemically depositing Ag/AgCl nanoparticles using a three-electrode cell unit. Two consecutive steps compose the procedure: the Ag NPs were deposited dipping the 2 cm long conductive thread, in 0.1 M AgNO 3 solution and applying a potential of − 0.2 V for 60 s. The device was rinsed in distilled water, dried and then immersed in 1 M KCl where the Ag/AgCl composite particle have been formed by applying a bias of + 0.6 V for 60 s. According to our previous work 45 , the Ag/AgCl NPs are composed by a core of silver, deposited during the first step, and an external shell of silver chloride, created throughout the second deposition step.</p><p>The polymerization solutions (PS) used to deposit PEDOT:BTB on conductive threads was prepared mixing 10 mM EDOT, 1 mM PBS with pH 7.0 and, 1 mM of BTB in 0.1 M KNO 3 aqueous solution following a previously reported method 37 . The solution was stirred for 20 min. The thread was immersed in 25 ml of the described solution and cyclic voltammetry was carried out, sweeping the potential from 0 to 1 V applied vs SCE with a scan rate of 0.1 V/s for 10 cycles. The samples/threads were finally rinsed in DI water.</p><!><p>Electrical and morphological characterization. Figure 1a shows a schematic drawing of fabrication and functionalization steps of the sensitive threads for chloride anions and pH. Briefly, the commercial threads were covered with a PEDOT:PSS layer by physical deposition, in order to fabricate yarns that exhibit a semiconductor behaviour. Then, the transducing materials (Ag/AgCl for Cl − and PEDOT:BTB for pH) were electrochemically deposited on the conductive polymer. The threads are purchased in local market and can be sewed into a fabric with a standard sewing machine or an industrial textile system thanks to their completely textile nature. This easy and low-cost fabrication process offers both high handling feature and conductive behaviour. After the PEDOT:PSS coating, the electrical resistances of the threads for the different fibers were measured using a four-point probe set-up, placing the probes one centimetre apart from each other. Four different types of threads have been tested: COT (pure cotton), POL (polyester) and SILK (pure silk) threads show comparable values (Fig. 1b) of (56 ± 9) Ω/cm, (90 ± 10) Ω/cm and, (85 ± 7) Ω/cm, respectively, while COT2 presents a resistance value of (89 ± 5) Ω/cm (see paragraph above for more details about fiber composition). Two bundles twisted together compose all the threads. Two different cotton threads were studied to investigate the possible effects due to the different numbers of fibers that compose the threads. We hypothesize that more PEDOT:PSS is present on COT thread since it fills also the spaces among the fibres on the surface (see the next paragraph for more details), causing a lower resistance. In order to check the success of the fabrication procedure previously described, Scanning Electron Microscopy was employed to obtain images of the four yarns coated with the polymeric mixture and after the electrochemical deposition of Ag/AgCl NPs and PEDOT:BTB. One cotton fiber is reported in Fig. 1c while the others in the Supplementary Fig. S1. No substantial difference was found among the four coated yarns, where the semiconducting polymer PEDOT:PSS mainly covers the outside of the thread with an average thickness of (12 ± 2) µm, as highlighted from the cross-sections in the inset of the left image. This is confirmed by Energy-Dispersive X-ray Spectroscopy (EDX) on cotton (bottom left image in Fig. 1c), which shows the presence of sulphur only in the outer shell of the thread (yellow), while carbon is present everywhere in the sample. In the pictures of PEDOT:PSS modified with Ag/AgCl, NPs are uniformly deposited on the surface of the polymer-coated threads having various size ranging from 300 nm to 30 µm. EDX confirmed the results of our previous work 45 , highlighting that the inner part of the nanoparticles is composed by silver (in green), while the outer shell is made of AgCl (chloride is represented in red in the image), thus being present in lower amount. The PEDOT:BTB electrodeposition resulted similar for all the fibers, only slightly changing their surface that looks rougher than before electrodeposition. The EDX map shows that bromine is uniform on the surface (in orange), suggesting that PEDOT:PSS is covered by a continuous layer of PEDOT:BTB. The SEM/EDX investigation demonstrates that modification was accomplished on all substrates.</p><p>Working principles. The single yarn sensors work thanks to the potentials that are generated by spontaneous and reversible electrochemical reactions involving the sensing materials and the analytes. It is well known that an Ag wire covered by AgCl exhibits a reproducible and well-defined electrochemical potential (E) with respect to a reference electrode, which depends on the concentration of chloride ions in the solution. The Nernst equation describes this behaviour:</p><p>Vol:.( 1234567890 where E o , R, T, n and F are the standard potential of the redox couple, the gas constant, the temperature, the number of exchanged electrons and the Faraday constant, respectively. a Cl − is the activity of chloride ion, but it can be replaced for practical purposes by [Cl − ] as long as the experimental set-up ensures a constant activity coefficient. Since the Ag/AgCl NPs are in intimate contact with the semiconducting polymer, they play as nano-gate elements and the generated potential modulates the conductivity of the PEDOT:PSS. When the Cl − concentration increases or decreases, the Ag/AgCl NPs potential must change in agreement with Nernst equation. Since the NPs are in electrical contact with PEDOT:PSS, a current will flow between the two materials until they reach the same potential value. This phenomenon changes the charge carrier concentration in the conductive polymer and thus its conductivity. The Fig. 1d shows a schematic representation of the working principle of the sensing threads. This process can be addressed as an electrochemical gating that was thoroughly described in our previous work 45 . When BTB is used as the counterion, the acid-base equilibria involving its protonation and deprotonation impact on the stabilisation of the conducting form of PEDOT 37 . Overall, a redox equilibrium is established between the PEDOT + /PEDOT couple that depends on the solution pH, thanks to the pH dye doping agent:</p><p>Also in this case the potential is described by Nerst equation:</p><p>where [2PEDOT + :BTB 2 ], [PEDOT] and [PEDOT + + BTB − ] are the concentration of 2PEDOT + :BTB 2 , PEDOT and PEDOT + + BTB − inside the polymer, respectively. The transduction mechanism, originating the two terminal sensor response was discussed in a recent work 54 . Briefly, the electrochemical potential that is spontaneously generated at the PEDOT:BTB/electrolyte interface can be used to modulate the conductivity of the underlying PEDOT:PSS layer, in analogy with the electrochemical gating described for the Ag/AgCl NPs. The two terminal sensors enable to perform a reliable, sensitive and quantitative measurement, with the great advantage of using only two terminals connected to a single yarn to extract a significant measure of pH or chloride concentration.</p><p>Single and multi-thread detection. As a first example of the sensor calibration, a long-term measure was carried out dipping the threads in 10 mL of 0.1 M KNO 3 for the different yarn types (Fig. 2a). We changed the chloride anions concentration by adding selected volumes of 2 M KCl under a soft stirring. Each different fiber was tested in a Cl − concentration range from 0.1 to 120 mM, in order to match the Cl − content in human sweat, which is 10-120 mM 40,41 . In all cases, a linear fit between the normalized current variation (ΔI/I 0 ) and the logarithm of [Cl − ] was obtained in the physiological relevant range. All the calibration plots are linear (R 2 > 0.97) in agreement with a potentiometric-like transduction, which is described by Nernst equation, as reported in the previous section. Figure 2b shows the data collected, with the value of the current at 0.1 mM [Cl − ] used as I 0 . The points are reported without vertical error bars since the relative error is less than 1%. The table in Fig. 2c shows the extracted sensitivity values in which the error is the one associated to the slope derived from the linear fit. To prove the ability of our sensors to have reliable and repeatable sensitivities in different volumes, we tested the threads in different volume amounts of 0.1 M KNO 3 . As an example, Fig. 2d shows performance of COT, reporting a perfect superposition of the normalized currents for increasing volumes of the solution. This fact is also highlighted by the complete agreement among the sensitivities, as reported in Fig. 2e and in the table of Fig. 2f, the robustness and the high repeatability of the measure. In order to quantitatively assess and compare the performance of single thread sensors in the detection of Cl − concentration and pH, we have separately evaluated them in 10 mL of Universal Buffer (UB), which allows stabilizing the pH values and to obtain well-controlled variations due to the additions of an acid or base.</p><p>The two-terminal Cl − sensor operates at a bias of 0.1 V and the chloride ion concentration was varied in the physiological relevant range of human sweat using 2 M KCl solution. This potential value is used to characterize the Cl − sensors throughout this study. The UB presents an initial concentration of 0.1 mM Cl − used as normalizing value for the current signal. Figure 3a shows the current flowing through an Ag/AgCl NPs functionalized cotton thread, as function of time, with the arrows highlighting the variation, upon increasing additions, of Cl − concentration.</p><p>On the other hand, the two-terminal pH sensors were operated at a bias of − 0.2 V and the current value recorded at the initial pH 4.2 was used to normalize the signal. This potential value is used to characterize the pH sensors throughout this study. The pH level was studied in the physiologically relevant range from 4 to 7, adding 1 M KOH to the buffered solution. Figure 3b shows the current flowing through a PEDOT:BTB functionalized cotton thread, as function of time, with the arrows underlining the variation, upon KOH additions, of pH level. The observed sensitivities vary with the fiber substrate and are likely affected by the chemical nature and surface treatment of the commercial yarns.</p><p>It is worthy to note that we have not observed any kind of chemical contamination between the functionalized textile-thread sensors and the overall solution that can affect the sensing of the target analytes. In order to avoid any possible contamination of the sample solution, the sensors were always thoroughly rinsed with doubly distilled water after the electrodeposition step. For the BTB-functionalized thread sensor, we did not observe any coloration of the solution under investigation after several measuring hours, thus confirming that the release of BTB dye is negligible, if even present. In addition, the Ag/AgCl is a well-known and wide spread material approved for the use on skin in various medical application, such as ECG 55 and EEG 56 electrodes, and it can be considered non-toxic 57 in the framework of the final application of these sensors. Furthermore, according to the solubility equilibrium of silver chloride in waterbased solution, the eventually released of nanoparticles would not influence the Cl − sensor signal. Figure S2, in the Supplementary Information, presents the response of all the different fibers. Excellent sensing performances were found for both sensors types and Fig. 3c,d show the linear relationship between the normalised current variation and log [Cl − ] or pH, respectively. The table reported in Fig. 3e shows the sensitivities found for the different sensors and yarns. In detail, it highlights the accuracy and the comparable behaviour of the sensors despite the threads types. Moreover, POL and COT based sensors show the highest and lowest sensitivities, respectively, for both chloride and pH detection. The average response time (assessed as the time required to reach the 90% of the final signal) is equal to (19 ± 6) s for a 115 mM chloride addition and (190 ± 30) s for a KHO addition to reach the 7.1 pH level. These values are in agreement with the previous reported results 22,58 . As regards the two www.nature.com/scientificreports/ types of cotton threads, despite the sensitivity to chloride is comparable for both COT and COT2, COT shows a considerably lower sensitivity to pH than COT2. Even though a complete investigation was beyond the purposes of this work and will be the subject of further studies, we hypothesized that the reason may be found in the different ways PEDOT:PSS is adsorbed by the threads and fills inter-fibers spaces as showed on Supplementary Fig. S3. The COT thread presents higher conductivity (G) than COT2 but exhibits, for pH sensors, a lower sensitivity (S). We suggest that this behaviour www.nature.com/scientificreports/ is related to the different linear density of the two threads. The Numeric Count, which is a standard parameter to express the size of a cotton yarn, was ρ L = 20 g/km and ρ L = 33 g/km, for COT and COT2, respectively. Assuming the same section for every fiber and the same volumetric density for both yarns (made of the same material), COT presents fewer fibers than COT2. The schematic and simplified representation is reported in Fig. S3a. For this reason, a higher quantity of conducting polymer PEDOT:PSS coats the COT thread, still without wetting the "bulk" of the thread. Accordingly, a lower resistance is reported for COT than COT2. On the other hand, Ag/AgCl Nanoparticles (Fig. S3b) and PEDOT:BTB (Fig. S3c) differently functionalize the PEDOT:PSS coated threads. Ag/AgCl NPs, owing to the small diameter, succeed in functionalizing almost all the PEDOT:PSS, whereas the PEDOT:BTB acts as a surface coverage 54 , leaving the polymeric material between the fibres in its pristine state (see the inset in Fig. S3c). Taking into account the Universal Buffer, Cl − sensors report closer sensitivities for the two yarns, while pH sensors based on COT show halved sensitivities than COT2, possibly due to the unfunctionalized PEDOT:PSS present in the spaces among the outer separated fibres.</p><p>Repeatability and reproducibility of the different threads, for chloride and pH detection, were assessed showing promising results.</p><p>As an example, Fig. 4a,b show repeatability studies carried out using a cotton thread, for chloride and pH measurement. To this aim, the same thread was successfully used to make three successive and consecutive experiments, giving almost superimposable response.</p><p>As far as the reproducibility, Fig. 4c,d show the results of the experiments conducted over three different POL thread for Cl − and pH detection. Moreover, the reproducibility is very high as pointed out by the low standard deviation associated to the slopes of the calibration plots.</p><p>Since the target of this work is the development of next-gen textile sensors, we assessed the ability of the here reported two-terminal Cl − and pH sensors to work in parallel for simultaneous and selective data collection from the same analysed solution. Their selectivity, chemical interference and cross-talk effects are crucial features for wearable sensors that have to deal with simultaneous variation of different analytes and metabolites in the same www.nature.com/scientificreports/ medium. Since sweat is a complex mixture of several compounds, the ability to discriminate different stimuli is a fundamental property to define the performance of a sensor. To investigate the potential interference of our multi-textile thread platform, the current responses of two single thread two-terminal sensors were measured simultaneously (Fig. 5a,b). Both 1 M KOH and 2 M KCl were alternately added to 10 mL of UB. Figure 5a shows the response of two sensors realized with the same fiber (COT), while Fig. 5b proves the effectiveness of employing sensors based on different fibers together (i.e. SILK and POL for the Cl − concentration and pH monitoring, respectively).</p><p>As shown in the Supplementary Fig. S4 and reported in Fig. 5, when the single thread sensors work in parallel and in the same Universal Buffer solution, they perform the same way in terms of stability, selectivity and sensitivity, compared to when they are characterized separately with no cross-talking effects and interference.</p><p>Therefore, multiplexed and quantitative detection of the two analytes can be carried out simultaneously and with no need to perform output signal processing or correction. These results assess the possibility to integrate the sensors directly into a fabric, choosing the fiber substrate according to the desired specifications in terms of sensitivity and material compatibility. It is worth noting that both natural and synthetic fibers can be exploited to realize a complete textile platform to monitor chloride ion concentration and pH level after a proper functionalization procedure.</p><p>Both types of sensors, fabricated on COT yarns, were further tested to examine the long-term stability by evaluating the eventual current drift. As reported in Fig. 5d, the sensors show a very low drift in the signal, pointing out the good long-term monitoring capability of the Cl − and pH sensors.</p><!><p>A further relevant demonstration of the reliability of the here reported Cl − and pH textile sensors was achieved by assessing their performance in artificial sweat solution, which excellently simulates the real-life sweat composition (ISO pH 5.5). Figure 6a shows the results of the single thread sensors characterization in 10 mL of artificial sweat in response to Cl − concentration and pH variation. Figure S5, in Supplementary Information, summarizes the sensor electrical responses. The current acquired at 40 mM [Cl − ] is used as I 0 to normalize the extracted values. To be consistent with the previous cases, the performance of sensors based on COT were widely tested. Figure 6b shows the results of the interference test using two different sensors www.nature.com/scientificreports/ for simultaneous detection of Cl − and pH in the same solution. In this case, the sensitivities of the Cl − and pH sensors are only slightly higher than the one obtained in the single thread characterization, since KOH or KCl additions barely affect the Cl − and pH sensors behaviour, respectively. Indeed, their values in artificial sweat are (141 ± 8) 10 -3 dec −1 and (18 ± 2) 10 -3 pH unit −1 for Cl − and pH sensors, respectively. Moreover, Supplementary Figure S6 reports the calibration curves of different threads sensors obtained from two separate interference tests. In both situations, no interference was observed and the sensitivities for Cl − and pH sensors are in perfect agreement, highlighting the device reproducibility.</p><p>Noteworthy, the multi-sensor textile platform composed by the two functionalized threads maintained high selectivity upon varying not only the Cl − concentration or pH value, but also the concentration of compounds that are usually present in human perspiration. Figure S7 (see Supplementary Information) reports the addition of urea, ethanol and lactate at their typical concentration in human sweat 41,[59][60][61] . In all the studied cases, no interference occurred in the determination of Cl − concentration.</p><p>To ensure the reliability of the sensors, we randomly dipped the Cl − thread sensor in 10 mL of artificial sweat with different Cl − content to extract a calibration curve, which was comparable to the results obtained with the standard characterization. The sensor response is shown in Fig. 6c while the its calibration plot and sensitivity are reported in Fig. 6d. The current dramatically increases when the thread is removed from the solution and left in air, returning to its initial value. After each immersion, the current value decreases and rapidly stabilizes accordingly to the specific chloride ion concentration. This test allows us to assert that hysteresis and memory-like effects can be neglected in our sensors. The feasibility and wearability of the Cl − thread sensor was demonstrated using a portable wireless data-reader connected via Bluetooth with a smartphone. The custom application, shown in Fig. 6e, allows monitoring and recording in real-time the current signal. Figure 6f reports the characterization plot of the single cotton thread showing the reversibility of the response after variation of Cl − concentration mimicking a real experiment in liquid environment. In addition, different frames of a continuous Cl − concentration monitoring are reported in the Supplementary Fig. S8 in which the recovery ability and reliability of the sensor is proved.</p><p>Herein, the two types of sensing threads have been sewn into the neck of a T-shirt (Fig. 6h) to perform an in-situ sweat analysis of Cl − concentration and pH level. We provide the proof of concept and an applicative example that the threads can be sewn in a fabric and continuously monitor in real-time analytes concentration. Since we demonstrated that the sensors response does not depend on volume, our sensors platform is also able to work with very low amount of human perspiration. In this case, we soak the COT threads with 1 mL of artificial sweat and we add independently 1 M KOH to change the pH from 5.5 to 7.7 and 3 M KCl to cover the chloride concentration from 30 to 100 mM. In this first stage, using a hydrophobic T-shirt allow us to work without a microfluidics system that will be required for further developments. Figure 6g-i report the acquired signal, the calibration plot using normalised current and the sensitivities values of the pH and Cl − sensors, respectively. Even in these conditions, more similar to the real one, the sensors show good performance comparable to the one obtained in laboratory environment with a sensitivity for the Cl − and pH sensor of (19 ± 1) 10 -3 pH unit −1 and (78 ± 5) 10 -3 dec −1 , respectively.</p><p>Finally, the here reported study shows the possibility to realize textile multi-sensor devices able to simultaneously measure the pH and Cl − concentration in the same solution without interferences or the need of shielding to avoid sample cross-talk.</p><p>In the last years, several electrochemical textile devices, able to monitor only the pH of sweat in a linear range physiologically relevant for human biofluids, have been proposed (see Table 1). Similarly to other cited studies, our two-terminal sensors keep the robustness of the potentiometric-like transduction as explained in our previous works 37,54 .</p><p>To the best of our knowledge, the here reported Cl − thread sensors is the first example of a selective and fullytextile thread device that does not need a reference electrode and is able to simultaneously operate 2 sensors immersed in the same solution, without interferences . We reported in Table 1 the performance of up-to-date wearable chemical sensors. In particular, in 2019 Xu et al. 62 , reported a electrochemically wearable Cl − sensors that could be used in parallel with a Ca 2+ sensor patch showing a sub-nerstian sensitivity and a limit of detection slightly lower than ours (see Fig. S9 in Supplementary). www.nature.com/scientificreports/ Kim et al. 63 described a two-terminal PEDOT:PSS thread that can be sewn into a fabric and non-selectively sense total cations concentration. Bujes-Garrido et al. 64 reported voltammetric screen-printed three-terminal Cl − sensors with AgNPs deposited onto the working electrode, employing both PET and Gore-Tex as substrate. Despite the low limit of detection of 5 µM, it exhibits a narrow operation range 5-60 mM, thus limiting its operation to applications where small concentration variations are expected.</p><!><p>The development of textile multi-sensor arrays is an advanced technological frontier in the design of tools for non-invasive monitoring of physiological parameters in humans. By exploiting a new concept of electrochemical transduction, we have here reported a novel example of multi-analyte textile platform for the continuous detection of pH and chloride in human perspiration. Each sensor is a two-terminal device composed by a single thread, working in aqueous solution without the need of a reference or gate electrode, with a significant simplification of the final device with respect to amperometric, potentiometric or transistor configuration. The sensitive yarns are realized using commercial textile fibres, both natural and synthetic, coated with a semiconducting polymer, properly functionalized to selectively detect chloride ion concentration and pH level in sweat. The reliability of the sensors was demonstrated first in universal buffer and then in artificial sweat, in order to mimic a real-life application environment. The sensors show stability, reproducibility and repeatability for all of the studied fibers and there are not significant difference between the fibres in terms of the overall performance, drawing the attention on the possibility to use the proper threads according to final product. Their ability to work in parallel, without interfering or change sensitivity, paves the ways for simultaneous and real-time Cl − concentration and pH level monitoring in body fluid, such as sweat.</p><p>This work provides the technological basis for the fabrication of single-thread sensitive elements and describes their assembly in a multi-platform biosensor system. Nevertheless, it still needs further improvement for the use in real-life applications. On one hand, a reliable sampling system should be employed to regenerate sweat on sensors surface and to ensure a very continuous monitoring. On the other hand, new functionalization should be designed to widen the bio-compounds that can be detected with this technological approach.</p>

Scientific Reports - Nature

ABHD4 regulates multiple classes of N-acyl phospholipids in the mammalian central nervous system

N-acyl phospholipids are atypical components of cell membranes that bear three acyl chains and serve as potential biosynthetic precursors for lipid mediators such as endocannabinoids. Biochemical studies have implicated ABHD4 as a brain N-acyl phosphatidylethanolamine (NAPE) lipase, but in vivo evidence for this functional assignment is lacking. Here, we describe ABHD4\xe2\x88\x92/\xe2\x88\x92 mice and their characterization using untargeted lipidomics to discover that ABHD4 regulates multiple classes of brain N-acyl phospholipids. In addition to showing reductions in brain glycerophospho-NAEs (GP-NAEs) and plasmalogen-based lyso-NAPEs (lyso-pNAPEs), ABHD4\xe2\x88\x92/\xe2\x88\x92 mice exhibited decreases in a distinct set of brain lipids that were structurally characterized as N-acyl lysophosphatidylserines (lyso-NAPSs). Biochemical assays confirmed that NAPS lipids are direct substrates of ABHD4. These findings, taken together, designate ABHD4 as a principal regulator of N-acyl phospholipid metabolism in the mammalian nervous system.

abhd4_regulates_multiple_classes_of_n-acyl_phospholipids_in_the_mammalian_central_nervous_system

<!>Materials<!>Generation of ABHD4\xe2\x88\x92/\xe2\x88\x92 Mice<!>Preparation of Mouse Tissue Proteomes<!>Gel-based ABPP Analysis<!>Mass Spectrometry (MS)-based Proteomics<!>Cell Culture and Transfection<!>Plasmids<!>In vitro Enzyme Assays<!>Untargeted Metabolomics Analysis<!>Tandem MS Fragmentation Experiments<!>Targeted Metabolite Measurements<!>Synthesis of 1-stearoyl-2-hydroxy-sn-glycero-3-phospho (N-palmitoyl) serine<!>Synthesis of 1-heptadecenoyl-L-serine (N-17:1 NAS)<!>Generation and Initial Characterization of ABHD4\xe2\x88\x92/\xe2\x88\x92 Mice<!>Deregulated Brain NAPE Metabolism in ABHD4\xe2\x88\x92/\xe2\x88\x92 Mice<!>Lipidomics Identifies Deregulated NAPS Metabolism in ABHD4\xe2\x88\x92/\xe2\x88\x92 Brains<!>Deregulated Brain NAPS Metabolism in ABHD4\xe2\x88\x92/\xe2\x88\x92 Mice<!>NAPS is a Direct Substrate of ABHD4<!>CONCLUSIONS<!>

<p>Lipids serve several fundamental roles in biology, including providing the chemical matter for cell membrane structures and acting as signaling molecules that bind to protein receptors to regulate diverse physiological processes.1 Phospholipids are a large class of lipids that contain a glycerol backbone esterified with two fatty acyl chains and conjugated through its third hydroxyl to a phosphate head group. The phosphate in most phospholipids is further modified with a polar substituent, such as those found in phosphatidylcholine (choline), phosphatidylethanolamine (ethanolamine), and phosphatidylserine (serine). N-acyl phospholipids, however, represent an unusual class of phospholipids that contain head groups modified with fatty acids.2, 3 Two major types of N-acyl phospholipids have been discovered in mammals – N-acyl phosphatidylethanolamines (NAPEs) and N-acyl phosphatidylserines (NAPSs) (Figure 1). The presence of a third fatty acid group grants N-acyl phospholipids with substantially altered physicochemical properties compared to other classes of phospholipids. Although the initial structural characterization of N-acyl phospholipids as endogenous constituents of mammalian tissues occurred several decades ago,4, 5 we still today have only a limited understanding of the biological functions and routes for enzymatic metabolism of this class of lipids.</p><p>NAPEs are considered rare phospholipids and, under physiological conditions, account for ~0.01% of total phospholipids of animal membranes.2, 3 However, NAPEs can accumulate under conditions of injury or stress.2, 3 Owing to the presence of a third acyl chain, N-acyl phospholipids impart stabilizing properties on cell membranes,6 as well as displaying fusogenic potential.7 Beyond these structural roles, some NAPEs may function as signaling molecules, as N-palmitoyl phosphatidylethanolamine has been shown to regulate food intake in mice8 and inflammatory responses.9 Arguably, however, NAPEs are most well recognized as precursors to the N-acyl ethanolamine (NAE) class of signaling lipids, which includes the endocannabinoid anandamide (C20:4 NAE).10 The conversion of NAPEs to NAEs can occur through one of multiple enzymatic routes, including direct phospholipase D-mediated catalysis by an NAPE-PLD11–14 and multi-step pathways that proceed through phospholipase A1/215–17 and C18, 19 enzymes. Plasmalogen-based NAPEs (pNAPEs) are a major subclass of NAPEs and can also be catabolized by NAPE-PLD, as well as the combined action of a pNAPE lipase and lyso-pNAPE phospholipase D.14 The biosynthesis of NAPEs appears to involve both calcium-dependent and –independent transacylase enzymes that transfer the fatty acid groups from phospholipids onto the amine of phosphatidylethanolamine.20–24 While the calcium-dependent transacylase remains molecularly uncharacterized, HRAS-like suppressor family proteins have been shown to function as calcium-independent transacylases that produce NAPEs.25–32 The NAPS class of N-acyl phospholipids, which has been isolated from human brain, animal tissues and cells, as well as yeast33, 34, is less well-characterized in terms of its biological functions and metabolism.</p><p>NAPE-PLD−/− mice exhibit massive increases in brain NAPE and lyso-NAPE content (~10-fold), but only modest reductions in major NAEs (~two-fold or less), including C16:0, C18:0, C18:1, and C20:4 NAEs.13, 14 These findings pointed to the existence of additional pathways for metabolizing NAPEs. One candidate pathway is mediated by PLA1/2 enzymes to yield lysoNAPE and glycerophospho (GP)-NAE intermediates, which would then be metabolized further to NAEs.10 The serine hydrolase ABHD4 has been identified as an NAPE and lyso-NAPE lipase with strong expression in brain.17 The function of ABHD4 in vivo, however, remains unknown. In this manuscript, we describe the generation and characterization of ABHD4−/− mice. Lipidomic analysis of brain tissue from these animals uncovered significant reductions in not only glycerophospho-NAEs (GP-NAEs) and lyso-pNAPEs, but also lyso-NAPSs. Recombinant ABHD4 was found to hydrolyze NAPS. These data thus indicate that ABHD4 is a principal enzyme responsible for the hydrolytic metabolism of both the NAPE and NAPS classes of N-acyl phospholipids in vivo.</p><!><p>d4-anandamide and d8-anandamide were purchased from Cayman Chemicals. 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (PE), 1-oleoyl-2-hydroxy-sn-glycero-3-phosphoethanolamine (lysoPE), 1-O-1'-(Z)-octadecenyl-2-oleoyl-sn-glycero-3-phosphoethanolamine (plasmenyl PE), 1-O-1'-(Z)-octadecenyl-2-hydroxy-sn-glycero-3-phosphoethanolamine (plasmenyl lysoPE), 1-stearoyl-2-oleoyl-sn-glycero-3-phosphoserine (PS), 1-stearoyl-2-hydroxy-sn-glycero-3-phosphoserine (lysoPS), 1-hexadecanoyl-L-serine, and 1-(5Z, 8Z, 11Z, 14Z–eicosatetraenoyl)-L-serine were purchased from Avanti Polar Lipids (Alabaster, AL). Heptadecenoic acid, pentadecenoyl chloride, palmitoyl chloride, heptadecenoyl chloride, nonadecenoyl chloride, and arachidonoyl chloride were purchased from NuCheck Prep (Elysian, MN). 1,2-dioleoyl-sn-glycero-3-phospho (N-nonadecenoyl) ethanolamine, 1,2-dioleoyl-sn-glycero-3-phospho (N-arachidonoyl) ethanolamine, 1-O-1'-(Z)-octadecenyl-2-oleoyl-sn-glycero-3-phospho (N-nonadecenoyl) ethanolamine, 1-oleoyl-2-hydroxy-sn-glycero-3-phospho (N-pentadecenoyl) ethanolamine, 1-oleoyl-2-hydroxy-sn-glycero-3-phospho (N-palmitoyl) ethanolamine, 1-oleoyl-2-hydroxy-sn-glycero-3-phospho (N-heptadecenoyl) ethanolamine, 1-O-1'-(Z)-octadecenyl-2-hydroxy-sn-glycero-3-phospho (N-heptadecenoyl) ethanolamine were synthesized as described.17, 35 Briefly, acid chlorides were reacted with an excess of PE, plasmenyl PE, lysoPE, or plasmenyl lysoPE and allowed to react for 1 h in CH2Cl2 with a catalytic amount of triethylamine. N-acylated lipids were then purified by silica gel flash column chromatography or preparative TLC. 1,2-dihydroxy-sn-glycero-3-phospho (N-pentadecenoyl) ethanolamine was synthesized by base hydrolysis of 1-oleoyl-2-hydroxy-sn-glycero-3-phospho (N-pentadecenoyl) ethanolamine as described.17, 35 The synthetic phospholipid standards were evaluated by LC-MS to assess purity and quantified by the Bartlett assay.36 More detailed synthetic methods are described below.</p><!><p>Construction of mice bearing a disruption in the Abhd4 gene was achieved using a standard targeting strategy in CMTI-1 embryonic stem cells from the 129S6/SvEvTac strain. Genomic DNA corresponding to a region including exons 3 and 4 of Abhd4 was amplified from a BAC clone isolated from the RPCI-22 library. Identification of neomycin-resistant ES cell clones bearing homologous recombination events was achieved via southern blot using a probe that hybridizes downstream of the targeting construct. Recombinant clones were injected into blastocysts from C57BL/6 mice and germline-transmission was determined by coat-color. Tail DNA from brown mice was subjected to southern-blotting using a probe that hybridizes upstream of the targeting construct, thereby confirming true homologous recombination. These mice were back-crossed into the C57BL/6 background for ten generations prior to biochemical and lipidomic analyses.</p><!><p>Mice were anesthetized with isoflurane and killed by decapitation. Brains were immediately removed and snap-frozen in liquid N2. Frozen brains were dounce-homogenized on ice in PBS (pH 7.4) and centrifuged at 1,000 × g for 10 minutes to remove debris. The resulting supernatant was further centrifuged at 100,000 × g for 45 minutes to provide the soluble fraction in the supernatant and the membrane fraction as a pellet. The pellet was washed and resuspended in PBS by brief sonication. Protein concentrations were determined using the DC Protein Assay (Bio-Rad), and samples were stored at −80 °C until use.</p><!><p>Brain soluble and membrane proteomes (50 µg in 50 µL of PBS) were prepared from 2-week-old ABHD4+/+ and ABHD4−/− females and incubated with 1 µM FP-rhodamine for 30 min at 37 °C. After 30 minutes reactions were quenched with 4x SDS/PAGE loading buffer (reducing), separated by SDS/PAGE [10% (wt/vol) acrylamide] and visualized by in-gel fluorescence scanning (Hiatchi FMBio IIe, MiraBio). Rhodamine fluorescence is shown in gray scale.</p><!><p>Serine hydrolase enrichment was performed essentially as described.37 Briefly, brain proteomes were adjusted to a final protein concentration of 2 mg/mL and treated with FP-biotin (500 µL total reaction volume, 10 µM final concentration) for 2 hours at room temperature. Excess probe was removed by protein precipitation using 4:1 (vol/vol) methanol:chloroform and proteins were then dissolved in 6 M urea in 25 mM ammonium bicarbonate. Proteins were reduced with 10 mM DTT, alkylated with 40 mM iodoacetamide, and diluted to a final concentration of 2 M urea with 25 mM ammonium bicarbonate. Biotinylated proteins were then enriched with avidin beads (SIGMA; #A9207) by incubation for 2 h at room temperature in 0.2% SDS in DPBS. The beads were washed three times with 1% SDS in DPBS, then three additional times in DPBS, then resuspended in 25 mM ammonium bicarbonate with 2 M urea. On-bead digestion with 2 µg trypsin (Promega; #V511A) was performed overnight at 37 °C in the presence of 1 mM CaCl2. Tryptic digests were acidified with 5% (vol/vol) formic acid, and aliquots were frozen at −80 °C until use. Multidimensional liquid chromatography tandem mass spectrometry (MudPIT) analysis was performed as described previously38 and peptides were eluted directly into a Velos Orbitrap mass spectrometer (Thermo Fisher) essentially as previously described.37 See Table S1 for data complete list of serine hydrolases detected in these experiments.</p><!><p>COS-7 cells were grown at 37 °C and 5% CO2 to ~70% confluence in Dulbecco's modified Eagle's medium containing 10% fetal calf serum in 10 cm dishes and transfected with 5 µg of plasmid DNA (or empty vector control) using polyethyleneimine (Polysciences, Warrington, PA, USA). After 24 h, the cells were washed twice with PBS, scraped, resuspended in PBS, and sonicated to lyse. The lysates were spun at 100,000 x g for 45 min to isolate the cytosolic fraction.</p><!><p>Full-length mouse Abhd4 was amplified by PCR from mouse brain cDNA with primers Abhd4 forward (5′-TGGTGGAATTCGCCACCATGGGCTGGCTCAGCTCGAC-3′) and reverse (5′-ACCTATCTAGAGTCAACTGAGTTGCAGATCT-3′) and was cloned into the pcDNA3.1/Myc-His vector with a C-terminal Myc-His tag using EcoRI and XbaI sites.</p><!><p>Detection of (lyso)phospholipid hydrolysis was accomplished via mass-spectrometric detection of the release of oleic acid or lyso-NAPE from substrates. Enzyme assays were performed in PBS in a total volume of 100 µl using 0.5 mg/ml protein. To avoid contaminating signals from endogenous lipids, soluble extracts of brain or cells were assayed. Reactions were incubated at 37 °C for 1 h with 100 µM substrate, 2 mM EDTA, with or without 0.1% Triton X-100. Reactions were stopped by the addition of 500 µl of MeOH and 1 nmol of heptadecenoic acid or 100 pmol of 1-oleoyl-2-hydroxy-sn-glycero-3-phospho (N-heptadecenoyl) ethanolamine was added as an internal standard. Lipids were extracted by the Bligh and Dyer method.39 A portion of the extracted lipid was injected onto an Agilent 6520 series quadrupole-time-of-flight (Q-TOF) MS. Chromatography was performed on a 50 × 4.60 mm 5 µm Gemini C18 column (Phenomenex) coupled to a guard column (Gemini; C18; 4 × 3.0 mm; Phenomenex SecurityGuard cartridge). The LC method consisted of 0.1 ml/min of 100% buffer A [95:5 (vol/vol) H2O:MeOH plus 0.1% (vol/vol) of 28% ammonium hydroxide] for 1.5 min, 0.5 ml/min linear gradient to 100% buffer B [60:35:5 (vol/vol) iPrOH:MeOH:H2O plus 0.1% (vol/vol) of 28% ammonium hydroxide] over 5 min, 0.5 ml/min 100% buffer B for 5.5 min, and equilibration with 0.5 ml/min 100% buffer A for 3 min (15 min total run time). MS analysis was performed in negative scanning mode with an electrospray ionization (ESI) source. The capillary voltage was 4.0 kV, the fragmentor voltage was 100 V, the drying gas temperature was 350 °C, the drying gas flow rate was 11 l/min, and the nebulizer pressure was 45 psi. Oleic acid release or lyso-NAPE release was quantified by measuring the area under the peak and comparing to the heptadecenoic acid or lyso-NAPE standard, respectively.</p><!><p>Discovery metabolite profiling (DMP) was performed as described previously.40 ABHD4+/+ and ABHD4−/− mice from 2 to 4 months of age (n = 4 per genotype) were anesthetized with isoflurane and killed by decapitation. Brains were harvested, laterally sectioned, and immediately submerged in liquid N2. One frozen brain hemisphere per mouse was weighed and immediately Dounce-homogenized in 8 ml of 2:1:1 (vol/vol/vol) CHCl3:MeOH:50 mM Tris pH8.0 with heptadecenoic acid and d8-anandamide added as internal standards for negative- and positive-mode analysis, respectively. Homogenates were centrifuged for 10 min at 1,400 × g. The organic (lower) phase was transferred to a clean vial and dried under a stream of N2. The metabolomes were resolubilized in 2:1 vol/vol CHCl3:MeOH (120 µl), and 30 µl was injected onto an Agilent 6520 series quadrupole-time-of-flight (Q-TOF) MS. LC separation was achieved using the same solid and mobile phases described above for substrate assays. To assist in ion formation, 0.1% (vol/vol) of 28% ammonium hydroxide or 0.1% (vol/vol) formic acid was added to the buffers for negative or positive ionization mode, respectively. The LC method consisted of 0.1 ml/min 0% buffer B for 5 min, a 0.4 ml/min linear gradient over 40 min to 100% buffer B, 0.5 ml/min 100% buffer B for 10 min, and 0.4 ml/min equilibration with 0% buffer B for 5 min, for an overall run time of 60 min. MS analysis was performed with an ESI source in scanning mode from m/z = 50–1,200. The capillary voltage was set to 4.0 kV, and the fragmentor voltage was set to 100 V. The drying gas temperature was 350 °C, the drying gas flow rate was 11 l/min, and the nebulizer pressure was 45 psi. Analysis of the LC-MS data were performed with the XCMS software,41, 42 which identifies, matches, aligns, and integrates chromatographic peaks and identifies m/z values that are significantly altered in control vs. experimental datasets. XCMS results from two independently performed experiments were compared and filtered for m/z values that appeared in both datasets with greater than three-fold change between genotypes, greater than 10,000 average peak-integration area, LC elution time during the linear gradient, and P < 0.01. Obvious isotopic peaks were manually excluded. Results are presented as the ratio of the average peak area measured in the ABHD4+/+ vs. ABHD4−/− brain metabolomes. Putative assignments were confirmed by coelution with synthetic standards and/or fragmentation analysis as described below. Relative abundance of known lipid species was determined by manually extracting the mass corresponding to the [M – H]− or [M + H]+ parent ion (in negative- or positive-ionization mode, respectively), integrating the area under the peak and normalizing this value for the tissue weight and internal standard area peak. Statistical significance was determined by unpaired, two-tailed Welch's t test.</p><!><p>MS/MS analysis was performed on an Agilent 6520 series quadrupole-time-of-flight (Q-TOF) instrument, using the same LC separation and buffers as described above for untargeted metabolomics. MS and MS/MS data collection, both in scanning mode from m/z = 50–1200 and a rate of 1.0 spectra/s, was performed with an electrospray ionization (ESI) source. The capillary voltage was set to 4.0 kV and the fragmentor voltage was set to 100 V. The drying gas temperature was 350°C, the drying gas flow rate was 11 l/min, and the nebulizer pressure was 45 psi. The collision energy was set to 20 V for lyso-NAPS and 15 V for NAS.</p><!><p>NAPEs, pNAPEs, lyso-NAPEs, lyso-pNAPEs, GP-NAEs, NAEs, NAPSs, and lyso-NAPSs in brains from ABHD4+/+ and ABHD4−/− mice 2 months of age were measured by multiple reaction monitoring (MRM) methods. Brains were harvested and the metabolomes were extracted as described above. 30 µl of lipid extract was injected onto an Agilent 6460 series triple-quadrupole MS connected to an Agilent 1290 Infinity HPLC system. LC separation was achieved using the solid and mobile phases described above for untargeted metabolomics and the following method: 0.1 ml/min 0% buffer B for 5 min, a 0.4 ml/min linear gradient over 15 min to 100% buffer B, 0.5 ml/min 100% buffer B for 8 min, and 0.4 ml/min equilibration with 0% buffer B for 5 min (33 min total run time per sample). For measurements in negative mode, 0.1% (vol/vol) of 28% ammonium hydroxide was added to the buffers. For measurements in positive mode, 0.1% (vol/vol) of formic acid or 10 mM ammonium formate was added to the buffers. For NAPEs and pNAPEs, MRM transition from [M – H]− to the fragment ion [R2COO]− with collision energy 45 V was used. Absolute abundance of NAPEs and pNAPEs were estimated by comparison to the N-19:1 NAPE (1,2-dioleoyl-sn-glycero-3-phospho (N-nonadecenoyl) ethanolamine) standard. For pNAPEs, the extraction and ionization efficiencies of N-19:1 pNAPE (1-O-1'-(Z)-octadecenyl-2-oleoyl-sn-glycero-3-phospho (N-nonadecenoyl) ethanolamine) were compared to those of the N-19:1 NAPE standard, and the levels of pNAPEs were corrected accordingly. For lyso-NAPEs, MRM transition from [M – H]− to the fragment ion [R1COO]− or [R2COO]− with collision energy 35 V was used. Absolute abundance of lyso-NAPEs was estimated by comparison to the N-17:1 lyso-NAPE (1-oleoyl-2-hydroxy-sn-glycero-3-phospho (N-heptadecenoyl) ethanolamine) standard. For GP-NAEs, MRM transition from [M -H]− to the fragment ion [deprotonated glycerophosphate]− at m/z 171 with collision energy 24 V was used. Absolute abundance of GP-NAEs was estimated by comparison to the N-15:1 GP-NAE (1,2-dihydroxy-sn-glycero-3-phospho (N-pentadecenoyl) ethanolamine) standard. Extraction efficiencies for N-15:1, N-16:0, N-18:0, and N-18:1 GP-NAEs under these conditions were 2.3, 10.5, 36.6, and 22.0%, respectively, and endogenous GP-NAE levels were corrected accordingly. For certain GP-NAE measurements (Figure S1C), 1% formic acid was used for the aqueous phase of the lipid extraction, rather than Tris pH 8. Under these conditions, extraction efficiencies for N-15:1, N-16:0, N-18:0, and N-18:1 GP-NAEs were 62.2, 78.2, 83.4, and 78.9%, respectively, and measurements of endogenous species were corrected accordingly. For NAPSs, MRM transition from [M – H]− to the fragment ion [phosphatidic acid]−(loss of N-acyl serine) with collision energy 35 V was used. Absolute abundance of NAPSs was estimated by comparison to the N-17:1 NAPS (1-stearoyl-2-oleoyl-sn-glycero-3-phospho (N-heptadecenoyl) serine) standard. For lyso-NAPSs, MRM transition from [M – H]− to the fragment ion [lyso phosphatidic acid]− (loss of N-acyl serine) with collision energy 20 V was used. Abundance of lyso-NAPSs was compared to the internal N-17:1 lyso-NAPS (1-stearoyl-2-hydroxy-sn-glycero-3-phospho (N-heptadecenoyl) serine) standard, however the amount of synthesized N-17:1 lyso-NAPS standard was too low to permit absolute quantitation so the values are reported as arbitrary units. For NASs, MRM transition from [M – H]− to the fragment ion [C2H4NO2]− with collision energy 15 V was used. Absolute abundance of NASs was estimated by comparison to the N-17:1 NAS (1-heptadecenoyl-L-serine) standard. For NAEs, formic acid was included in the mobile phase and MRM transition from [M + H]+ to m/z 62 (ethanolamine fragment) with collision energy 11 V was used, except for the d4-anandamide standard which monitored the transition from [M + H]+ to m/z 66 (d4 ethanolamine fragment). NAEs were quantified by comparison to the d4-anandamide standard. For lyso pNAPEs ammonium formate was added to the buffers and MRM transition from [M + H]+ to [CH2CH2NHCORN]+ with collision energy 15 V was used. Absolute abundance of lyso pNAPEs was estimated by comparison to the N-17:1 lyso-NAPE standard. The extraction and ionization efficiencies of N-17:1 lyso pNAPE (1-O-1'-(Z)-octadecenyl-2-hydroxy-sn-glycero-3-phospho (N-heptadecenoyl) ethanolamine) were compared to those of the N-17:1 lyso-NAPE standard, and the levels of lyso pNAPEs were corrected accordingly. MS analyses were performed by multiple reaction monitoring (MRM) with an ESI source. The capillary voltage was set to 3.5 kV, and the fragmentor voltage was set to 100 V. The drying gas temperature was 350 °C, the drying gas flow rate was 9 l/min, and the nebulizer pressure was 45 psi. The dwell time for each lipid was set to 100 ms.</p><!><p>A solution of 1-stearoyl-2-hydroxy-sn-glycero-3-phosphoserine (3.0 mg, 5.7 µmol, 1.0 equiv.) in CH2Cl2 (1 mL) and triethylamine (12.5 µL) was treated with palmitoyl chloride (1.4 mg, 5.1 µmol, 0.9 equiv.), and the reaction mixture was stirred for 1 h at ambient temperature. The reaction was concentrated under a stream of N2, and the remaining residue was purified by preparative TLC (CHCl3:MeOH:NH4OH:H2O = 65:35:5:1) providing the title compound. HRMS (ESI-TOF-) m/z calc'd for C40H77NO10P [M – H]−: 762.5290, found 762.5297. 1-stearoyl-2-oleoyl-sn-glycero-3-phospho (N-palmitoyl) serine (HRMS (ESI-TOF-) m/z calc'd for C58H109NO11P [M – H]−: 1026.7743, found 1026.7746.), 1-stearoyl-2-oleoyl-sn-glycero-3-phospho (N-heptadecenoyl) serine (HRMS (ESI-TOF-) m/z calc'd for C59H109NO11P [M – H]−: 1038.7743, found 1038.7743.), and 1-stearoyl-2-hydroxy-sn-glycero-3-phospho (N-heptadecenoyl) serine (HRMS (ESI-TOF-) m/z calc'd for C41H77NO10P [M – H]−: 774.5290, found 774.5286.) were synthesized from their corresponding PS and lyso-PS analogs and characterized in the same manner.</p><!><p>L-Serine (27.4 mg, 0.26 mmol, 1.5 equiv.) was dissolved in 2 M NaOH (0.4 mL, 4.6 equiv.), and THF (0.4 mL) was added. The mixture was stirred on ice for 10 min, and heptadecenoyl chloride dissolved in 0.2 mL of THF (49.9 mg, 0.174 mmol, 1.0 equiv.) was added drop-wise. The mixture was stirred for 1 h at ambient temperature, and then poured into ice-cold HCl. The product was extracted with EtOAc and the combined organic layers were dried over anhydrous Na2SO4 and concentrated under reduced pressure. The residue was purified by preparative TLC (CHCl3:MeOH:NH4OH:H2O = 65:35:5:1) providing the title compound as a white solid (25.0 mg, 40%): 1H NMR (600 MHz, CDCl3) δ 5.36 (s, 2H), 4.34 (s, 1H), 3.99 (s, 1H), 3.76 (s, 1H), 1.21–2.37 (m, 26H), 0.91 (m, 3H). HRMS (ESI-TOF-) m/z calc'd for C20H36NO4 [M – H]−: 354.2649, found 354.2648.</p><!><p>ABHD4−/− mice were generated by replacing exons three and four of the Abhd4 gene (exon three encodes the catalytic serine) with a Neo cassette using traditional gene-targeting techniques (Figure 2A). Disruption of the genomic locus was confirmed by Southern blotting (Figure 2B and C) and PCR genotyping (Figure 2D). Loss of Abhd4 mRNA expression and ABHD4 protein activity were confirmed by RT-PCR (Figure 2E) and activity-based protein profiling [ABPP; gel-based (Figure 2F) and mass spectrometry (MS)-based (Figure 2G and Table S1)]. ABHD4−/− mice were born at the expected Mendelian frequency, were viable and healthy, and showed no overt differences in their cage behavior compared to wild-type littermates. Brain homogenates from ABHD4−/− mice displayed significant, but incomplete reductions in NAPE- and lyso-NAPE-lipase activity compared to brain homogenates from ABHD4+/+ mice (Figure 1H), indicating that ABHD4 is a principal, but not sole enzyme in the nervous system that hydrolyzes NAPEs and lyso-NAPEs.</p><!><p>We previously showed that ABHD4 exhibits both NAPE- and lysoNAPE-lipase activity in vitro to sequentially produce lysoNAPEs and GP-NAEs,17 the latter of which can be processed further to NAEs by glycerophosphodiesterase such as GDE1.35, 43 Based on these biochemical assignments, we measured the major classes of lipids along the NAPE metabolic pathway in brain tissue from ABHD4+/+ and ABHD4−/− mice. No changes in NAPE or NAE content were observed in ABHD4−/− brains (Figure S1). In contrast, the concentrations of multiple GP-NAEs were significantly reduced in ABHD4−/− brains compared to wild-type brains, with the decrease in N-16:0 GP-NAE being the most prominent (Figure 3A and Figure S1). The absolute concentrations of GP-NAEs were affected by the pH of the aqueous phase of the lipid extraction (see below), but, regardless, significant reductions were observed with or without acidification of the aqueous phase (Figure 3A and Figure S1). Modest, but significant reductions in some lysoNAPEs, including N-16:0, O-18:0 and N-16:0, O-20:4 were also detected in brains from ABHD4−/− mice (Figure S1).</p><p>There are two primary classes of ethanolamine glycerophospholipids – diacyl-PEs and 1-alkenyl 2-acyl PEs (plasmalogens). Plasmalogen-type NAPE and lyso-NAPE (pNAPE and lyso-pNAPE, respectively) are also present in mouse brain and NAEs can be formed from these lipids through both NAPE-PLD-dependent and -independent pathways.14 In contrast to the aforementioned minor reductions in lyso-NAPEs, the lyso-pNAPE content of ABHD4−/− brains was markedly decreased (Figure 3B). We found that recombinant ABHD4 expressed by transient transfection in COS-7 cells showed robust pNAPE-lipase activity (Figure 3C) and that brain lysates from ABHD4−/− mice exhibited significantly decreased pNAPE-lipase activity (Figure 3D). Recombinant ABHD4 displayed, on the other hand, a very limited capacity to hydrolyze plasmenyl-PE (Figure 3C) and the hydrolytic activity for this substrate was unaltered in ABHD4−/− brain tissue (Figure 3D). Despite substantial reductions in lyso-pNAPEs, the concentrations of pNAPEs were unaltered in brain tissue from ABHD4−/− mice (Figure S1).</p><p>As mentioned above, acidification of the aqueous phase during lipid extraction resulted in higher absolute concentrations of GP-NAEs and a more prominent decrease in these concentrations in ABHD4−/− brains that resembled the magnitude of reductions observed in brain lyso-pNAPEs. Based on the known vulnerability of alkenyl bonds to acid cleavage,44 we speculate that acid-catalyzed breakdown of the sn-1 alkenyl bond of lyso-pNAPEs may contribute to accumulated GP-NAE signals under acid extraction conditions.</p><p>These data, taken together, demonstrate that ABHD4 regulates multiple, but not all components of the NAPE pathway. The substantial reductions in GP-NAEs and lyso-pNAPEs in ABHD4−/− brains support a physiological role for ABHD4 as a principal (lyso)-NAPE lipase in the mammalian nervous system. The more modest reductions in lyso-NAPEs could be explained by a dual role for ABHD4 in contributing to both the biosynthesis and degradation of this class of lipids (through NAPE and lyso-NAPE hydrolysis, respectively), which is not a complication for lyso-pNAPEs, since these ether lipids cannot be further hydrolyzed by ABHD4. That neither brain NAPE nor pNAPE were substantially affected by ABHD4 disruption (Figures S1C and S1D), could reflect the involvement of other metabolic pathways that control the steady-state concentrations of these lipids, such as their direct conversion to NAEs by NAPE-PLD,13, 14 or a consequence of the much larger concentrations of (p)NAPEs compared to their downstream lyso-(p)NAPE and GP-NAE products.</p><!><p>We more fully characterized the metabolic changes that occur in brains from ABHD4−/− mice by performing untargeted MS-based lipidomic experiments. Briefly, lipid extracts were prepared using a modified Folch extraction45 in which frozen tissue was homogenized in 2:1:1 chloroform:methanol:buffer and the lipid-containing organic layer evaporated to dryness. Lipid extracts were analyzed by LC-MS using reverse-phase chromatography and high-resolution MS in both positive and negative modes across a mass range from 50 – 1200 m/z units. The XCMS software42 was employed to align chromatograms across multiple independent LC-MS runs and identify substantial (> three-fold) and statistically significant (P < 0.01) differences in m/z value peaks between the ABHD4+/+ and ABHD4−/− brain tissues. Using this approach, a major difference was observed in the negative ion mode with mass-to-charge (m/z) ratios of 380.8 and 762.5 (likely representing the [M – 2H]2− and [M – H]− ions of the same metabolite) (Figure 4A). This metabolite was substantially reduced in brains from ABHD4−/− mice (Figure 4A and Table S2). Several candidate structures for this metabolite were considered including phospholipids (PE and PS), lyso-NAPEs, and lyso-NAPSs, among others. Chemical synthesis of N-16:0/O-22:6 lyso-NAPE and N-16:0/O-18:0 lyso-NAPS (Figure 4B), which have the same nominal mass of 762 provided material for comparison with the endogenous m/z 762.5 metabolite. These analyses revealed that the endogenous m/z 762.5 metabolite co-eluted with N-16:0/O-18:0 lyso-NAPS (Figure 4C), whereas the retention times of the isobaric PE, lyso-NAPE, and PS species were clearly distinct (data not shown). The endogenous m/z 762.5 metabolite also gave an identical tandem MS fragmentation pattern compared to the synthetic N-16:0/O-18:0 lyso-NAPS (Figure 4D).</p><p>Taken together, these lipidomic and follow-up analytical studies indicate that ABHD4 regulates N-16:0/O-18:0 lyso-NAPS content in mouse brain.</p><!><p>We next performed a broader analysis of lyso-NAPS species in ABHD4+/+ and ABHD4−/− brain tissues by targeted (multiple reaction monitoring) MS. Nearly all of the O-16:0 and O-18:0 lyso-NAPS species were dramatically reduced in brain tissue from ABHD4−/− mice, including the most abundant detected species N-16:0, O-18:0 and N-16:0, O-18:1 (Figure 5A). In contrast, lyso-NAPS species with unsaturated O-fatty acyl groups on the glycerol backbone were unchanged in ABHD4−/− brain tissue. Conversely, modest, but significant elevations were observed in several NAPS lipids in ABHD4−/− brain tissue (Figure 5B). Considering that unsaturated fatty acids are generally attached at sn-2 position of phospholipids, these lipid profiles suggest that ABHD4 hydrolyzes the sn-2 fatty acyl chain of N-acyl phosphatidylserines (NAPSs) to produce sn-1-O-saturated lyso-NAPSs.</p><p>NAPSs are potential precursors of N-acyl serines (NASs), a class of bioactive lipids implicated in inflammation,46 neuroprotection,47, 48 and bone maintenance.49 We did not, however, observe differences in NAS content in ABHD4+/+ and ABHD4−/− brain tissues (Figure S2).</p><!><p>To determine whether ABHD4 acted as an NAPS lipase, we next measured the NAPS hydrolytic activity of lysates from mock- and ABHD4-transfected COS-7 cells and found that the latter samples showed significantly elevated NAPS lipase activity when assayed either in the presence or absence of Triton-X-100 as a detergent (Figures 6A and S3). Inclusion of Triton X-100 as a detergent proportionally increased the NAPS hydrolytic activities of both mock- and ABHD4-transfected lysates (Figures 6A and S3). No differences were detected in the PS lipase activity of mock- and ABHD4-transfected COS-7 cell lysates in the presence or absence of Triton X-100 (Figure 6A). The NAPS lipase activity of brain tissue was also substantially increased in the presence of Triton X-100, and, under these conditions, a significant reduction in NAPS (Figures 6B and S3) and (lyso)-NAPE hydrolysis (Figure S3) was detected in ABHD4−/− brain lysates. These data, taken together, indicate that ABHD4 serves as a general N-acyl phospholipid hydrolase capable of accepting both NAPEs and NAPSs as substrates. That brain tissue from ABHD4−/− mice still exhibits considerable N-acyl phospholipase activity (Figures 2H, 3D, S3, and 6B), however, points to the existence of additional lipases that may contribute to NAPE and NAPS metabolism in vivo.</p><!><p>Our studies provide strong evidence that ABHD4 is a principal enzyme responsible for N-acyl phospholipid metabolism in the nervous system. The contribution of ABHD4 appears complementary to that of NAPE-PLD in that disruption of ABHD4 primarily affects lyso-(p)NAPE and lyso-NAPS content, while disruption of NAPE-PLD causes dramatic elevations in lyso-(p)NAPEs13, 14. Ablation of either enzyme alone is insufficient to produce large changes in major NAEs (e.g., C16:0, C18:0, C18:1) or anandamide, possibly indicating that these pathways can compensate for another or that additional enzymes also participate in the biosynthesis of NAEs. Future studies with NAPE-PLD−/−/ABHD4−/− mice should help to address this question.</p><p>Several studies have implicated the direct involvement of NAPEs in cellular and physiological processes,2, 3 including feeding,8 inflammation,9 and biomembrane stability.6 Lyso-(p)NAPEs are less well characterized. NAPSs were only recently identified as natural brain constituents,33, 50 and to our knowledge, lysoNAPSs have not previously been reported as endogenous metabolites in mammalian systems. Knockdown of ABHD4 has been shown to confer resistance to anoikis (or cell-detachment-induced apoptosis) in RWPE-1 prostate cells.51 Whether this effect is mediated by N-acyl phospholipids, however, remains unknown. Future studies using ABHD4−/− mice, as well as the development of ABHD4-selective inhibitors, should strengthen our understanding of the functions of this enzyme and its N-acyl phospholipid substrates and products in mammalian biology and disease.</p><!><p> Associated Content </p><p>Supporting Information.</p><p>Supporting tables and figures. This material is available free of charge via the Internet at http://pubs.acs.org.</p>

PubMed Author Manuscript

Evaluation of a Cell Penetrating Prenylated Peptide Lacking an Intrinsic Fluorophore via in situ Click Reaction

Protein prenylation involves the addition of either a farnesyl (C15) or geranylgeranyl (C20) isoprenoid moiety onto the C-terminus of many proteins. This natural modification serves to direct a protein to the plasma membrane of the cell. A recently discovered application of prenylated peptides is that they have inherent cell-penetrating ability, and are hence termed cell penetrating prenylated peptides. These peptides are able to efficiently cross the cell membrane in an ATP independent, non-endocytotic manner and it was found that the sequence of the peptide does not affect uptake, so long as the geranylgeranyl group is still present. The present study investigates the effect of removing the fluorophore from the peptides and investigating the uptake by confocal microsopy and flow cytometry. Our results show that the fluorophore is not necessary for uptake of these peptides. This information is significant because it indicates that the prenyl group is the major determinant in allowing these peptides to enter cells; the hydrophobic fluorophore has little effect. Moreover, these studies demonstrate the utility of the Cu-catalyzed click reaction for monitoring the entry of nonfluorescent peptides into cells.

evaluation_of_a_cell_penetrating_prenylated_peptide_lacking_an_intrinsic_fluorophore_via_in_situ_cli

<p>One particularly prevalent post-translational modification is known as protein prenylation, which occurs on approximately 2% of all mammalian proteins.[1] This modification involves the addition of a C15 (farnesyl) or C20 (geranylgeranyl) isoprenoid moiety onto a cysteine residue near the C-terminus of proteins that bear a 'CAAX' box motif, catalyzed by either the farnesyltransferase or geranylgeranyltransferase enzyme.[2-4] In the 'CAAX' box, A represents any aliphatic amino acid and X represents the residue that controls farnesylation or geranylgeranylation.[5, 6] The prenylation modification serves to direct membrane association of many proteins including members of the Ras superfamily of proteins.[7] It has been previously shown that the K-Ras protein is involved in approximately 30% of all human cancers and inhibition of the farnesylation of Ras abolishes its oncogenic activity.[8-10] Hence, considerable work has been aimed at developing farnesyltransferase inhibitors, many of which are in clinical trials.[11-13]</p><p>Several important molecules for therapeutic consideration are not able to naturally traverse the cell membrane and are thus being attached to cell penetrating peptides (CPPs) to facilitate membrane crossing.[14] The majority of the current CPPs range from 5-25 amino acids and often contain several charged amino acid residues.[15-17] The drawback of most current CPPs is that they are taken up into cells by an energy-dependent endocytotic process, which leads to a majority of the cargo being transported to endosomes and eventually lysosomes for degradation.[18] Designing CPPs that are able to escape lysosomal degradation has become a high priority.</p><p>Our group has previously developed a series of peptides that are based on the C-terminus of the naturally geranylgeranylated protein CDC42. These peptides were functionalized with a fluorophore and a geranylgeranyl group and found to have intrinsic cell-penetrating ability, by a mechanism that was energy independent.[19] This study also demonstrated that without the geranylgeranyl group, the peptides were unable to enter cells; related experiments demonstrated that farnesylated peptides derived from CDC42 can also enter a wide variety of cell types.[20] A further study examining the role of the positively charged amino acids in the sequence found that these amino acids were not necessary for uptake, and that a peptide as short as two amino acids was able to penetrate cells, provided the geranylgeranyl group was present.[21] Our current study serves to elucidate the role of the 5-carboxyfluorescein (5-Fam) fluorophore on the uptake of these peptides. The benefit of this class of cell penetrating prenylated peptides is that they enter cells by a non-endocytotic, ATP independent mechanism that allows accumulation into the cytosol of cells.[19, 21] These peptides may prove useful in cell-penetrating applications in which the cargo can be easily released in the cytosol.</p><p>For this study, the peptides (Figure 1) were synthesized using standard Fmoc coupling conditions on an automated synthesizer with Rink-amide resin to afford the C-terminal amide peptides. While still on resin, the N-terminal lysine side chain (ε amine) was acetylated, followed by acylation with an alkyne-containing acid (4-pentynoic acid) on the amino group to give the resulting N-terminal alkyne. Geranylgeranylation of that peptide was performed in solution using acidic Zn(OAc)2 coupling conditions[22] to afford peptide 1. Peptide 2 was prepared in an analogous manner except that the ε-amino group of the N-terminal lysine was acylated with 5-Fam. Thus peptide 1 contains an alkyne but no fluorophore whereas peptide 2 contains both an alkyne and a fluorophore.</p><p>The ability of both of these peptides to enter cells was first established using confocal laser scanning microscopy (CLSM). Because peptide 2 contains the 5-Fam fluorophore, visualization after uptake in HeLa cells is easily accomplished after a 2 h incubation of the peptide at 1 μM (Figure 2B). To investigate whether the 5-Fam fluorophore is necessary for uptake, peptide 1 was synthesized lacking the fluorophore. Following cellular incubation of 1 at 1 μM for 2 h, HeLa cells were rinsed three times with PBS and fixed with 4% paraformaldehyde, followed by permeabilization with 0.1 % Triton X-100. A copper catalyzed click reaction was then conducted using tetramethylrhodamine azide (TAMRA-N3) to form a covalent triazole linkage between the peptide-alkyne and the tetramethylrhodamine (TAMRA) fluorophore; when 1 enters cells, it can be visualized through the TAMRA fluorophore using CLSM (Figure 2C). Thus, it can be seen that the peptide lacking a fluorophore is still able to efficiently cross the membrane of HeLa cells. While control experiments in which HeLa cells were treated with TAMRA-N3 (without prior peptide treatment) manifested no background labeling (Figure 2A), we wanted to compare the localization pattern obtained by direct visualization of 5-Fam with that observed from the TAMRA-N3 labeling. Incubation of 2 with HeLa cells and subsequent click reaction to TAMRA-N3 resulted in strong co-localized fluorescence in the cells (yellow color, Figure 2D); this observation also confirms that little background reaction occurs in the TAMRA-N3 labeling process since the red TAMRA fluorophore is only present where the green 5-Fam of the peptide is observed. The Mander's overlap coefficient of the cells is 0.901 when the nucleus is included in the analysis and 0.960 when it is excluded, indicating there is significant overlap of green and red fluorescence, even when the nucleus is considered (a Mander's coefficient of 1 indicates perfect overlap[23]). It is important to note that it appears there is some nuclear labeling when HeLa cells are incubated with 1. To explore this issue, we incubated peptide 1 with HeLa cells, followed by a subsequent click reaction to either TAMRA-N3 or 5-Fam-PEG-N3. This will help establish if the nuclear localization is an artifact of the fixation or of the fluorophore used. Regardless of the fluorophore used, nuclear staining is observed when incubating HeLa cells with 1 (Figure 3). This effect is absent when performing the click reaction on cells that were not treated with alkyne peptide (see Figure 2A), and is also absent in cells treated with 2 (see Figure 2B). This suggests that peptide 1 that lacks the fluorophore partitions differently within cells and is able to enter the nucleus whereas peptide 2 does not; however, it should be noted that the amount of nuclear localization is small compared to the amount of peptide distributed throughout the remainder of the cell. Overall, the localization patterns of 1 and 2 are quite similar.</p><p>Having visually established that peptide 1 can enter HeLa cells, we next sought to quantify the differences in uptake between 1 and 2. Accordingly, a reagent containing 5-Fam linked to an azide moiety was synthesized for this purpose. This reagent was designed so that the fluorescence of the two peptides could be more directly compared since both peptides would be labeled by the same fluorophore. A succinimidyl ester of 5-Fam was reacted with 11-Azido-3,6,9-trioxaundecan-1-amine in the presence of diisopropylethylamine overnight at room temperature. Following purification by reverse phase HPLC, 5-Fam-PEG-N3 was isolated and used as the subsequent azide for click chemistry. HeLa cells were incubated with 1 or 2 at various concentrations for 1 h. Both sets of cells were rinsed, fixed and permeabilized. Cells treated with peptide 1 were subjected to the click reaction with 5-Fam-PEG-N3 for 1 h followed by several rinses, and subsequent flow cytometry analysis. Cells treated with peptide 2 were analyzed directly by flow cytometry without click reaction. Based on those experiments, peptide 1 appears to enter the cells to a lesser extent (2-3-fold) than 2 (Figure 4) at all concentrations tested. This could indicate that 1 is indeed not as efficient at penetrating the cellular membrane compared to 2 since it lacks the fluorophore. Alternatively, this result could be an artifact of the analysis; the click reaction may not quantitatively label all of the available alkyne-functionalized peptide during the course of the reaction. It is also plausible that if 1 enters the cells and localizes to an internal membrane (endoplasmic reticulum localization is natural for geranylgeranylated proteins[24]) that the N-terminal alkyne becomes buried in the membrane and is inaccessible for the click reaction. These latter two possibilities would both lead to an artificially lower mean fluorescence value. Regardless of which explanation is correct for the reduced fluorescence of cells treated with 1 versus 2, the important conclusion remains – peptide 1 lacking a fluorophore is able to efficiently enter HeLa cells. This further substantiates the importance of the hydrophobic geranylgeranyl group for peptide uptake; without that moiety, the peptide is unable to enter cells.[19]</p><p>In previous work, we have shown that peptide sequences containing a geranylgeranyl group are efficiently taken into cells in an energy-independent manner, regardless of positive charge in the sequence. The findings reported here demonstrate that the presence of the hydrophobic fluorophore 5-Fam in such peptides has minimal effect on their cell penetrating ability, further underscoring the importance of the isoprenoid moiety. This information may be useful for applications utilizing cell-penetrating peptides to deliver cargo across membranes. Using the smallest molecule for a CPP has the benefit of facile synthesis as well as minimum disturbance of the cell during cargo delivery. The non-endocytotic mechanism that functions in the uptake of these peptides may also prove to be useful since it avoids potential endosomal localization/degradation of cargo. For instance, Medintz and coworkers attached a CPP to a quantum dot and a fluorescent protein to study the internalization of cargo, finding that microinjection of the conjugate into cells was necessary to avoid endosomal uptake.[25] Finally, we note the novel method for visualizing cell penetrating peptides described herein. To date, most studies of cell penetrating peptides and related materials have employed fluorophores linked to the compounds themselves to report on cellular entry. The incorporation of bulky, hydrophobic fluorescent groups inevitably perturbs the chemical and physical properties of the molecules under study and hence complicates structure/function analysis. For example, the addition of the 5-Fam fluorophore to a lysine residue alters the calculated partition coefficient (CLogP) by 3 units, illustrating a large change in the hydrophobic properties of the parent molecule (Figure S3). The method reported here provides a simple solution to this problem. The incorporation of a small alkyne-containing moiety into the peptide results in minimal alteration of the properties of the parent peptide. The CLogP value of Lys(5-Fam) with either an acetylated N-terminus or an N-terminal alkyne differs by only approximately 0.5 units, indicating the addition of the alkyne is a rather benign change to the hydrophobicity of the peptide (see Figure S4). However, the presence of the alkyne allows for facile visualization via click-mediated fluorescent labeling when desired. Furthermore, the alkyne group may be used in a variety of click reactions beyond fluorophore attachment. This approach should be quite useful for a variety of studies of cell penetrating molecules.</p>

PubMed Author Manuscript

Cyclobutane dication, (CH₂)₄²⁺: a model for a two-electron four-center (2e-4c) Woodward–Hoffmann frozen transition state

The structures of the elusive cyclobutane dication, (CH 2 ) 4 2+ , were investigated at the MP2/cc-pVTZ and CCSD(T)/cc-pVTZ levels.Calculations show that the two-electron four-center (2e-4c) bonded structure 1 involving four carbon atoms is a minimum. The structure contains formally two tetracoordinate and two pentacoordinate carbons. The non-classical σ-delocalized structure can be considered as a prototype for a 2e-4c Woodward-Hoffmann frozen transition state. The planar rectangular shaped structure 2 with a 2e-4c bond was found not to be a minimum.

cyclobutane_dication,_(ch₂)₄²⁺:_a_model_for_a_two-electron_four-cen

Introduction<!>Calculations<!>Results and Discussion<!>Conclusion

<p>The protonated hydrogen cation (H 3 + , i) is the simplest known structure involving a two-electron three-center (2e-3c) bond (Scheme 1). The first spectroscopic detection of the H 3 + ion was reported by Oka in 1980 [1]. Similarly, the structure of five coordinated protonated methane (CH 5 + , ii), a prototype of nonclassical carbocations, could not be interpreted by the classical tetravalency bonding concept [2,3] also requiring the engagement of a 2e-3c bond as suggested by Olah in 1969 [4,5].</p><p>A compelling number and huge array of carbocations involving higher coordinate carbon is by now realized by experimental studies.</p><p>Hypercarbon chemistry covers in addition to carbocations also carboranes, carbon-bridged organometallics, carbonyl clusters, along with others. The rapidly evolving field has been extensively surveyed [6]. In comparison, structures of the carbocations involving a two-electron four-center (2e-4c) bond are rare. This type of bonding could occur in rigid frameworks such as in 1,3-dehydro-5,7-adamantanediyl dication (iii) [7] and pagodane dication (iv) [8]. The question is can a 2e-4c bond exist in a molecular structure involving four atoms without any rigid frameworks such as in the diprotonated hydrogen (H [8,10] using semiempirical and ab initio methods and later by Herges, von Ragué Schleyer, Schindler and Fessner [11] using an ab initio method. The various levels of calculations including MP2/6-31G* indicated that the structure vi was not a minimum [11]. The 1,3-dehydro-5,7-adamantanediyl dication (iii) [7] with 2e-4c bonding was generated and identified by 13 C NMR spectroscopic and theoretical methods corresponding to a three dimensional aromaticity. The four bridge-head p-orbitals overlap inward in a tetrahedral fashion involving two electrons.</p><p>We have now extended our study to obtain information on the structure, stabilities and possible rearrangement pathways of the elusive cyclobutane dication. The species is an example of the simplest carbodication containing a 2e-4c bond and can be considered as a prototype for a frozen Woodward-Hoffmann transition state analog [12,13]. The 2e-4c delocalized σ-bishomoaromatic system is representative of a 2e-aromatic pericycliclic species. This type of system may be considered as the transition state of the allowed cycloaddition of ethylene to ethylene dication. Electron delocalizations take place in the plane of the conjugated systems unlike cyclobutadiene dication (vi), where delocalization takes place through conventional p-type orbitals. GIAO-CCSD(T) derived 13 C NMR chemical shifts of the structures were also computed to probe the nature and extent of the positive charge delocalization.</p><!><p>The Gaussian 09 program [14] was employed for geometry optimizations and frequency calculations. Vibrational frequencies at the MP2/cc-pVTZ//MP2/cc-pVTZ level were used to charac-terize stationary points as minima (NIMAG (number of imaginary frequencies) = 0 or transition state NIMAG = 1) and to compute zero point vibrational energies (ZPE), which were scaled by a factor of 0.96 [15]. CCSD(T)/cc-pVTZ optimizations and GIAO-CCSD(T) 13 C NMR chemical shifts calculations by the GIAO (Gauge Invariant Atomic Orbitals) method [16][17][18][19] using tzp and qzp basis sets, which were optimized by Schäfer, Horn and Ahlrichs [20,21], have been performed with the CFOUR program [22,23]. The 13 C NMR chemical shifts were computed using TMS (calculated absolute shift, i.e., σ(C) = 197.9 (GIAO-CCSD(T)) as a reference.</p><!><p>Structures of 1-4 were optimized at the MP2/cc-pVTZ and CCSD(T)/cc-PVTZ levels. CCSD(T)/cc-PVTZ structures will be discussed throughout unless otherwise stated. Dication 1 is expected to form by removal of two electrons from cyclobutane. At both MP2/cc-pVTZ and CCSD(T)/cc-PVTZ levels, the C 2 symmetric form 1 (Figure 1) was found to be a minimum for (CH 2 ) 4 2+ . This is confirmed by frequency calculations at the corresponding levels. The ring of the structure 1 embraces a puckered conformation with a puckering angle (the angle between the two three-membered rings) of 136° as shown in Figure 1. The structure contains conventionally two tetracoordinate and two pentacoordinate carbons. It resembles a complex between two ethylene radical cations, (CH 2 ) 4 2+ , culminating in the formation of a 2e-4c bond. Total electron density, HOMO and LUMO of the dication 1 are depicted in Figure 2. In such a small ring doubly charged system, charge-charge repulsion is certainly strong, but the bonding interactions as well as charge delocalization are good enough to counter this repulsion. The C1-C4 and C2-C3 bonds (1.968 Å) were calculated to be significantly longer than the C1-C2 and C3-C4 bonds (1.430 Å). The D 2h symmetric structure 2 was also computed for comparison with the structure 1. Computed vibrational frequencies at the MP2/cc-pVTZ//MP2/cc-pVTZ level indicated that the rectangular shaped structure 2 could not be a minimum as it a The 13 C NMR chemical shifts were referenced to TMS, for numbering scheme please see Figure 1.</p><p>contains two imaginary frequencies (NIMAG = 2, corresponding to the vibrations of the two ethylene units in plane and out of plane).</p><p>Moreover, structure 2 was also found to be notably disfavored over the structure 1 by 14.3 kcal/mol at the CCSD(T)/cc-PVTZ + ZPE level (Figure 1, Table S9 in Supporting Information File 1). Attempts to find a minimum for the open 1,4-butanediyl dication, bearing two primary carbenium centers failed due to automatic transformation to the thermodynamically more stable hydrogen-bridged structure 3 (Figure 1).</p><p>The twisted angle between the planes of the hydrogen-bridged units in 3 was found to be 91.4°. Expectedly, the structure 3 was found to be favored over the structure 1 by 11.4 kcal/mol (Figure 1, Table S9 in Supporting Information File 1). Planar parallelogram-shaped structure 4 was identified as the transition state for the transformation of 1 to 3 (Figure 1). The structure 4 lies 4.1 kcal/mol above 1. Dissociation of 1 leading to two ethylene radical cations (CH 2 =CH 2 •+ ) was also considered.</p><p>The process was found to be exothermic by 12.8 kcal/mol at the CCSD(T)/cc-PVTZ + ZPE level. We tried, but could not locate a transition state for the dissociation process at the CCSD(T)/ cc-PVTZ level.</p><p>The 13 C NMR chemical shifts of the structures 1-3 were computed by employing GIAO-MP2 and GIAO-CCSD(T) methods using CCSD(T)/cc-pVTZ geometries. GIAO-CCSD(T) calculated 13 C NMR chemical shifts for 1 show that the C1 and C2 carbons are deshielded at δ 216.6 and 40.2 ppm, respectively, indicating a nonclassical nature of the ion in accord with its tetra-and pentacoordinate nature. Computed 13 C NMR chemical shifts of the structure 1 are given in Table 1. Vibrational frequencies of the structure 1 are given in Table S10 (see Supporting Information File 1).</p><!><p>The present study at the MP2/cc-pVTZ and CCSD(T)/cc-pVTZ levels shows that the cyclobutane dication, (CH 2 ) 4 2+ (1), a prototype for a 2e-4c Woodward-Hoffmann frozen transition state, is a viable minimum on its potential energy surface. The structure contains formally two tetracoordinate and two pentacoordinate carbons. It resembles a complex between two ethylene radical ions, (C 2 H 4 ) 2•+ , culminating in the formation of a 2e-4c bond involving four carbon atoms. The planar rectangular shaped structure 2 with a 2e-4c bond was found to be not a minimum.</p>

Beilstein

trans-Symmetric Dynamic Covalent Systems: Connected Transamination and Transimination Reactions

The development of chemical transaminations as a new type of dynamic covalent reaction is described. The key 1,3-proton shift is under complete catalytic control and can be conducted orthogonally to, or simultaneous with, transimination in the presence of an amine to rapidly yield two-dimensional dynamic systems with a high degree of complexity evolution. The transamination–transimination systems are proven to be fully reversible, stable over several days, compatible with a range of functional groups, and highly tunable. Kinetic studies show transamination to be the rate-limiting reaction in the network. Furthermore, it was discovered that readily available quinuclidine is a highly potent catalyst for aldimine transaminations. This study demonstrates how connected dynamic reactions give rise to significantly larger systems than the unconnected counterparts, and shows how reversible isomerizations can be utilized as an effective diversity-generating element.

trans-symmetric_dynamic_covalent_systems:_connected_transamination_and_transimination_reactions

Introduction<!><!>Introduction<!>Theoretical analysis of dynamic exchange<!><!>Theoretical analysis of dynamic exchange<!>Transamination catalysis<!><!>Transamination catalysis<!><!>Transamination catalysis<!><!>Transamination catalysis<!><!>Transamination catalysis<!>Connected TATI<!><!>Connected TATI<!><!>Kinetic studies<!><!>Dynamic pathway control<!><!>Substrate scope and synthetic considerations<!><!>Substrate scope and synthetic considerations<!><!>Substrate scope and synthetic considerations<!>Conclusion<!>General TATI procedure<!>One-pot protocol for imine condensation–TATI<!>General procedure for imine synthesis<!>Isolation protocol<!>Hydrolysis of the imine system

<p>Constitutional dynamic chemistry (CDC) involves studies of chemical systems that respond to stimuli and adapt to external or internal pressure.[1] In this context, the systemic properties that emerge through interactions between components in a dynamic compound mixture can be very different from those of the isolated individual members. The systems are also fundamentally under thermodynamic control, in which reversible covalent bonds and noncovalent interactions allow access to the most energetically stable state, from which adaptation can occur. The area has expanded rapidly during recent years, and many applications for dynamic systems within ligand/receptor interaction studies, materials chemistry, catalysis, and molecular sensors have been developed.[2] Furthermore, CDC provides a relevant framework for the emerging field of systems chemistry.[3]</p><p>However, a challenge in CDC has been the relatively low number of suitable reversible covalent bonds that partake in dynamic systems. To generate large, diverse dynamic systems for different applications, it is of high importance to develop more reversible covalent connections capable of exchange under mild conditions. Only a few dynamic covalent-bond functionalities, mainly imine/acyl hydrazone[4] and disulfide exchange,[5] are typically used in dynamic systems, and despite recent developments of reversible transformations, such as dynamic thiol ester exchange,[6] the Strecker reaction,[7] nitrone exchange,[8] the nitroaldol reaction,[9] transamidation,[10] and alkyne metathesis,[11] there is still a growing demand for new reversible bonds that can be utilized in the construction of complex networks and systems.</p><p>Dynamic covalent bonds that can operate orthogonally to, or simultaneous with, other reversible functionalities are, in this respect, of particular interest. Compared with systems based on single dynamic reactions, such multidimensional arrays lead to significantly larger systems with higher diversity, potentially covering more chemical space. A range of examples of orthogonal covalent and noncovalent reactions, mostly based on metal coordination, acyl hydrazones, and disulfide chemistry, have also been reported.[7, 12]</p><p>In this context, we envisaged that the coupling of reversible isomerization to intermolecular dynamic exchange would provide rapid entry to highly complex dynamic systems. However, isomerization reactions constitute an underexplored area of CDC; the only example involves fluxional systems of bullvalene derivatives.[13]</p><p>The reversibility of azomethine transamination is well known.[14] In biological systems, transaminases catalyze both transamination and transimination of amino acids to and from α-ketoacids with pyridoxal-5′-phosphate as a cofactor (Scheme 1). The transformation is of high industrial interest, and biocatalytic equilibrium control is also heavily pursued for the production of chiral amines.[15]</p><!><p>Reversible transamination of α-ketoacids under transaminase catalysis.</p><!><p>Renewed interest has also emerged in nonbiological systems, for which transamination strategies have been developed in, for example, asymmetric organocatalysis, synthesis of fluorinated amines, and selective N-terminal functionalization of peptides.[16] The synthetic challenge with transaminations resides in often unfavorable equilibria, with typical equilibrium constants near unity. From a CDC viewpoint, however, this challenge is instead an advantage, since it allows a predictable constituent expression in the systems.</p><p>Herein, we report the development of reversible transamination of aromatic imines and orthogonal coupling to transimination, which yields double dynamic imine systems. These transamination–transimination (TATI) systems exhibit interesting properties because each of the two individual dynamic reactions can be toggled on or off by the addition or removal of the respective catalyst for each process. Since both dynamic reactions manipulate the same functional group, combining the two dynamic exchange processes leads to connected dynamic systems with unusually large complexity evolution.</p><!><p>Figure 1 a and b illustrates the two main types of dynamic covalent bonds used in CDC. For symmetric dynamic bonds, both exchange partners belong to the same class of functional groups, of which disulfide exchange and alkene metathesis are two prominent examples.1</p><!><p>Illustration of the main classes of dynamic bonds in CDC. a) Symmetric dynamic exchange; b) unsymmetric dynamic exchange; c) trans-symmetric dynamic exchange.</p><!><p>For such a system, the number (N) of generated system constituents upon addition of n different types of initial monofunctional components is given by Equation (1):</p><p>(1)</p><p>For unsymmetric exchange, the two partners belong to different functional groups, such as aldehydes and amines in imine exchange chemistry. With n monofunctional components of one class and m of the other, the number (N) of constituents becomes that given by Equation (2):</p><p>(2)</p><p>In this study, conditions to dynamically alter the nature of the components of an unsymmetric reversible connection have been developed, so that the two functional groups interconvert into each other (Figure 1 c). This scenario, which we term trans-symmetric exchange, will in most instances give rise to a number (N) of system constituents predicted by Equation (3):[17]</p><p>(3)</p><p>From inspection of Equations (1)–(3), it is clear that a TATI system will generate systems that are at least (for n=m) four times larger than the analogous unsymmetric imine system. Thus, the diversity of screening collections generated by trans-symmetric exchange will be considerably higher than that for other exchange types.</p><!><p>Imine 1 was initially investigated in an isolated transamination reaction (Table 1), in which many of the previously reported conditions proved unsuitable. For example, the use of strong heating or strong bases, such as tBuOK and 1,5-diazabicyclo[5.4.0]undec-5-ene (DBU), instigated rapid decomposition, which generated complex mixtures well before the transamination equilibrium could be reached. Finally, successful clean transamination could be achieved by using the conditions developed by Soloshonok and co-workers, with a large excess of NEt3 in MeCN at 50 °C.[18] After 24 h, about 66 % of imine 1 had undergone transamination to form isomer 2. Extending the reaction time to 30 h led to arrival at the equilibrium point, with a product distribution of 25/75 in favor of transamination product 2. This result is in agreement with theory because the equilibrium position of benzylic aldimine/aldimine transaminations can be correlated with the Hammett substituent parameters; more electron-rich systems are favored to exhibit the imine functionality.[14]</p><!><p>Initial optimization of the transamination reaction[a]</p><p>[a] Conditions: imine 1 (0.25 mmol), NEt3 (1.0 mmol), 3 Å molecular sieves (MS; 10 mg), anhydrous solvent (0.25 mL), 24 h, N2. [b] Analyzed by 1H NMR spectroscopy; [c] n.d.=not determined.</p><!><p>A reversible covalent reaction for CDC needs to display long-term stability, mild conditions, and rapid kinetics. Thus, the use of a large excess of NEt3, along with the relatively long equilibrium time, highlighted the need for an improved protocol. The initial screening (Table 1) indicated that the highest rates occurred in the presence of NEt3 in polar aprotic solvents; anhydrous MeCN, DMSO, and DMF provided access to equilibrium compositions in slightly longer than 24 h. Since the use of the last two solvents led to complications in product isolation and reaction monitoring, MeCN was adopted in the next optimization step.</p><p>The use of less NEt3 resulted in significant retardation of the equilibration rate (Table 2, entry 2). To be able to decrease the catalyst loading, a more suitable Brønsted base was thus deemed necessary.2</p><!><p>Base optimization studies for reversible transamination[a]</p><p>[a] Conditions: imine 1 (0.25 mmol), base (n equivalents), 3 Å MS (10 mg), anhydrous MeCN (0.25 mL), 50 °C, N2, 24 h. NMI=N-methylimidazole, NMM=N-methylmorpholine, NMP=N-methylpyrrolidine, DIPEA=diisopropylethylamine, DMAP=4-dimethylaminopyridine, DABCO=1,4-diazabicyclo[2.2.2]octane, TMG=N,N,N′,N′-tetramethyl-1,3-propanediamine; for the structures of C1–C5 (see Figure 2). [b] Analyzed by 1H NMR spectroscopy. [c] Conversion towards the equilibrium position, 25/75 for 1/2. [d] Low solubility.</p><!><p>Mechanistic studies have indicated that many base-catalyzed 1,3-proton shifts may proceed through a concerted proton shuffling mechanism (Scheme 2).[18] The transamination rate should therefore not only be dependent on the base strength, but also on steric congestion around the basic site. A more accessible nitrogen center should be able to provide more efficient proton shuffling.</p><!><p>Proposed concerted proton shuffling during the chemical transamination of aldimines.[18]</p><!><p>Thus, a range of bases with less steric bias were screened. As expected, the position of the equilibrium did not change significantly upon variation of base or base loading. Weak bases, such as NMI and NMM, showed low activity (Table 2, entries 3 and 4), as did the stronger base NMP, which somewhat surprisingly provided only a modest equilibration rate (Table 2, entry 5). However, the use of 4 equivalents of quinuclidine C1, 3-hydroxyquinuclidine C2, or DABCO led to efficient transamination, reaching equilibria within 24 h. Drastically lowering the loadings of these three catalysts to 0.2 equivalents still led to decent rates, with C1 providing the best performance with an equilibrium reached in around 20 h. As little as 5 mol % C1 could be utilized for the transformation, albeit at the cost of equilibration time.</p><p>Due to their successful application in the asymmetric transamination of α-ketoesters and the high similarity of the active basic site to that of C1, cinchona alkaloids were also evaluated as catalysts (Figure 2).[16b] However, quinine C3 was almost completely inactive, possibly due to low solubility (Table 2, entry 14). Alkylation of the hydroxyl functionality yielded soluble catalyst C4, but no increase in activity could be observed (Table 2, entry 15). Compound C5 was furthermore evaluated, since demethylation of the quinoline methoxy group could reveal a stabilizing hydrogen-bond-donor functionality, but this was also found to be completely inactive (Table 2, entry 16). Because cinchona alkaloids are significantly bulkier around the quinuclidine nitrogen than C1, these results again indicate that the transamination reaction is strongly dependent on the steric environment around the basic site.2</p><!><p>Quinuclidine-based catalysts evaluated for transamination activity.</p><!><p>Because system stability is of utmost importance in dynamic chemistry applications, it was gratifying to observe that C1 catalysis did not induce any degradation, even six days after equilibrium had been reached. Given the general sluggishness of aldimine transamination, catalyst C1 seems to be a remarkably effective and mild catalyst, even from a synthetic perspective.</p><!><p>With conditions for the reversible transamination reaction at hand, transimination was next investigated.[19] A range of Lewis acids were evaluated as catalysts (see the Supporting Information), with most catalysts providing complete transimination of all tested substrates within 10 min. The most robust transimination conditions were achieved with ZnBr2,[7] although transimination also worked well with primary amine catalysis. The time to reach equilibrium did, however, increase to about 1 h in this case, compared with a few minutes for ZnBr2.</p><p>Connected TATI systems could be generated simply through the addition of both catalysts to the same mixture, as shown in Figure 3. Mixing imine 1 with benzylamine A1, ZnBr2, and base C1 in MeCN, in the presence of 3 Å MS, led to the generation of four imines, as evidenced by both 1H NMR spectroscopy and GC/FID analysis. After around 30 h, no further system composition changes could be observed.</p><!><p>TATI equilibrium perturbation experiments. The relative system composition was dependent on the amine added: a) 0.1 equivalents of amine A1, b) 0.5 equivalents of amine A1, c) 1.0 equivalents of amine A1, d) 0.5 equivalents of amine A2. The system composition was analyzed by both 1H NMR spectroscopy and GC. Conditions: imine 1 (0.25 mmol), amine, C1 (0.05 mmol), ZnBr2 (0.0125 mmol), 3 Å MS (10 mg), MeCN (0.25 mL).</p><!><p>The system distribution at equilibrium could be easily tuned through amine addition. The use of 0.1 equivalents of benzylamine A1 yielded a system of constituents 1–4 (Figure 3 a). Here, compound 2 constituted nearly 43 % of the total imine content, which indicated a preferred systemic expression of 2 over the other constituents under these conditions. In general, the system indicated a preference for releasing the less-basic free amine A2 and incorporating higher amounts of more basic amine A1 into the imine system. Increasing the amount of A1 to 0.5 and 1.0 equivalents led to a clear, gradual increase in the proportion of imine 4 up to 50 % of the total imine content (Figure 3 b and c). However, by omitting amine A1 and instead adding 0.5 equivalents of the trifluoromethyl-substituted benzylamine A2, the relative amount of compound 4 was drastically reduced and the expression of the doubly trifluoromethyl-substituted imine 3 instead increased (Figure 3 d). Curiously, the ratios of compounds 1 and 2 were relatively stable during these system manipulations, only varying slightly despite drastic changes in the overall system composition. This indicates that homosubstituted imines 3 and 4 are acting as "sinks" that modulate and buffer the concentration of imines 1 and 2 by preferential incorporation of newly added amine.</p><p>The high level of adaptability strongly indicates that the connected system was under thermodynamic control. However, further support that the TATI system had reached equilibrium was provided by dual-entry-point analysis. Two identical systems were thus generated, starting from either compound 1 or 2, as displayed in Figure 4. Upon addition of 0.5 equivalents of amine A1 to a solution of imine in MeCN with C1 and ZnBr2 as catalysts, both systems resulted in identical distributions, which remained stable over several days.4</p><!><p>Results obtained from dual-entry-point equilibration analysis. The double dynamic systems were identical when generated from either direction. See the Supporting Information for further details and more control experiments.</p><!><p>The kinetic profile of the dynamic system generation was next evaluated (Figure 5). Since the system based on imines 1–4 required GC analysis for full quantification due to overlapping signals in the NMR spectra, compound 5 and benzylamine A3 were instead subjected to the TATI conditions. Careful reaction monitoring by NMR spectroscopy during the equilibration process revealed that transamination of compound 5 was the limiting step of the reaction system, in which equilibrium was reached after around 48 h. The concentration profile of compound 5 indicated an exponential decrease, which was symptomatic of first-order behavior with respect to the reagent. Immediately upon transamination, the new heterosubstituted imine 6 was formed. Transimination with amine A3 immediately occurred to form compound 7, releasing benzylamine A2, which, in turn, underwent transimination with starting compound 5 to yield the second homosubstituted imine 3. Interestingly, the concentration profile of direct transamination product 6 turned out to be almost sigmoidal in shape. During a long induction period, the system seemed to settle into local transimination equilibria, in which compound 6 was an under-expressed constituent. Only once a certain ratio of free amines A2 and A3 had been formed did compound 6 start to accumulate, eventually reaching approximately the same concentration as imine 3. NMR spectroscopy analysis confirmed that the free amine ratio stayed relatively constant after 18 h, which again provided support for a model in which the homosubstituted imines acted as amine sinks during equilibration. This unexpected effect is an example of the intricate kinetic behavior of interconnected chemical systems and highlights the need for dynamic reaction networks to be analyzed from a broader systems chemistry perspective.</p><!><p>Kinetic profile for the four different imines in the TATI system (5 (⧫); 6 (▪); 7 (▴); 3 (○)), as measured by 1H NMR spectroscopy. Conditions: imine 1 (0.25 mmol), amine A3 (0.025 mmol), catalyst C1 (0.05 mmol), ZnBr2 (0.0125 mmol), 3 Å MS (10 mg), MeCN (0.25 mL). Data were obtained from duplicate experiments.</p><!><p>The individual transimination and transamination equilibria of model compound 5 were easily accessible by selection of the reagent, and represent two orthogonal reversible reaction pathways in the dynamic covalent reaction network (Figure 6). However, the key diversity generation originates from a combination of the two pathways. When mixing compound 5 with transamination catalyst C1 and benzylamine A1 (1 equiv), a system of nine imine compounds and three amines was efficiently generated (Figure 6). This experiment verifies the trans-symmetric exchange mode of the system and illustrates the potential in the TATI protocol for rapid systemic complexity generation. The system size growth relative to the amount of starting compounds is thus demonstrated to be significantly higher than that in non-trans-symmetrical protocols utilizing monofunctional exchange partners.6</p><!><p>1H NMR spectra illustrating selective access to each equilibration mode of imine 5: a) starting material 5; b) transimination equilibrium (15 min) with imine 5 (1.0 equiv), benzylamine A1 (1.0 equiv), ZnBr2 (0.05 equiv); c) transamination equilibrium (24 h) between imines 5 and 6, with C1 (0.2 equiv); d) coupled transamination and transimination equilibrium (45 h) with A1 (1.0 equiv), 5 (1.0 equiv), C2 (0.2 equiv), and ZnBr2 (0.05 equiv). e) Enlarged view of the characteristic imine resonance region. Resonances corresponding to compounds 1 and 2 overlap.</p><!><p>Finally, the substrate scope of the TATI system formation was investigated in detail. As displayed in Table 3, a range of substrates were compatible with the optimized reaction conditions. Generally, the system needed at least one electron- withdrawing component on one of the aromatic rings for acceptable equilibration times. This can be understood when considering the reaction mechanism, in which the initial transamination proton shuffling is the rate-limiting step. Because electron-withdrawing groups on the aromatic rings increase the acidity of the benzylic protons, more electron-poor systems should lead to faster transamination.3</p><!><p>Substrate scope of TATI equilibration[a]</p><p>[a] Conditions: imine (0.25 mmol), amine (0.125 mmol), C1 (0.05 mmol), ZnBr2 (0.0125 mmol), 3 Å MS (10 mg), MeCN (0.25 mL). [b] Approximate time for the system to reach equilibrium. [c] At RT with four equivalents of NEt3 instead of compound C1. [d] With 0.1 equivalents of amine, no ZnBr2. [e] With 0.5 equivalents of compound C1.</p><!><p>Cyano, trifluoromethyl, chloro, and bromo substituents all worked with the system, and different ortho-, meta-, and para-substituted compounds were all well tolerated. The strongly electron-withdrawing nitro-substituted imine 14 decomposed under the reaction conditions (Table 3, entry 6). Problematic behavior of nitro-substituted compounds in aldimine transaminations has been observed before.[14] Reverting to the use of less reactive NEt3 at room temperature led to quick equilibration, although the system decomposed over the next 48 h (Table 3, entry 7). Heteroaromatic systems generally performed well, albeit without the addition of ZnBr2 and with a lower amount of free amine with pyridine-based structures due to complexation and degradation of the picolylamines formed (Table 3, entries 8 and 9). With some compounds, such as imidazole imine 17 and furyl imine 18, small amounts of unidentified side products appeared during equilibration.</p><p>Generally, compounds with only electron-donating substituents underwent very slow transamination (Table 3, entries 12–14). An exception was imine 12, which reacted smoothly to provide a full TATI system in a short time (Table 3, entry 4). One reason for this reactivity could be intramolecular hydrogen bonding between the phenolic proton and the imine nitrogen.</p><p>Furthermore, benzylic imines with substituents on both aromatic rings were compatible with the system (Table 3, entries 16–19). Even methyl- and methoxy-substituted aromatics participated readily in the exchange processes, as long as at least one other component had a sufficiently electron-withdrawing functional group attached. The high compatibility range of these model substrates indicates that TATI exchange can be utilized with a range of substituted aromatic and heteroaromatic exchange partners, covering a wide part of the relevant chemical space in terms of aromatic compounds.</p><p>An advantage of TATI systems from a practical standpoint is the dual nature of the involved functional groups. If a particularly interesting aldehyde or amine is commercially unavailable or difficult to access synthetically, the reverse compound can instead be employed and the desired compound created for the screening collection in situ. This is, for example, the case for entries 10 and 15 in Table 3, for which very expensive benzylamines are created in situ from cheap, widely available aldehydes. It is also straightforward to form the benzylic imine in the flask before equilibration to generate a one-pot protocol that omits the extra imine formation step. Simply condensing the aldehyde and benzylamine in the presence of 3 Å MS, as shown in Scheme 3, followed by addition of the catalysts, led to the evolution of an identical system to that observed with preformed imine 1.</p><!><p>One-pot protocol for the direct creation of TATI dynamic systems from aldehydes and amines.</p><!><p>Furthermore, simple aqueous washing led to the removal of all catalysts and benzylamines, which allowed the isolation of imine mixtures in high purity and yield. The imines can also be readily hydrolyzed, leading to aldehyde mixtures. This protocol could thus in theory also be of use for the preparation of nondynamic imine or aldehyde systems for high-throughput screening, although this would not utilize the advantages for in situ screening that a system provides.</p><!><p>We have developed the first example of trans-symmetric reversible exchange in dynamic systems. The results showed that reversible transamination and transimination reactions could be carried out orthogonally or connected to provide access to catalyst-controlled dynamic systems with high level of complexities from simple starting materials. New conditions for aldimine transaminations were also developed, with which quinuclidine was found to be an effective and mild catalyst for the transformation. The TATI system furthermore proved compatible with a wide range of different functional groups, which were well tolerated under the reaction conditions.</p><p>The generated dynamic systems could potentially be used for in situ dynamic screening or as starting point systems for combinatorial multicomponent reactions. Efforts to expand the scope of TATI systems for metal sensing and double dynamic materials are currently underway.</p><!><p>ZnBr2 (2.8 mg, 0.0125 mmol) was added to a dry 2 mL screw-cap vial with activated 3 Å MS (ground, 10 mg) and the mixture was left under N2 for 1 h. Benzylic imine (0.25 mmol) was dissolved in anhydrous MeCN (0.25 mL), and C1 (5.6 mg, 0.05 mmol) was added. The resulting mixture was transferred to the vial and amine was added. The solution was stirred under N2 at 50 °C in a sand bath. Reaction monitoring was performed by removing an aliquot (10 μL) of the reaction mixture, filtering through a pad of cotton, and diluting with CDCl3, after which an NMR spectrum was recorded.</p><!><p>4-Trifluoromethylbenzaldehyde (66.8 μL, 87.1 mg, 0.5 mmol) was dissolved in anhydrous MeCN (0.5 mL) in a dry 2 mL screw-cap vial with activated 3 Å MS (ground, 100 mg), and benzylamine (81.9 μL, 80.4 mg, 0.75 mmol) was added. The mixture was stirred at RT under N2 for 24 h, after which time C1 (11.1 mg, 0.1 mmol) and ZnBr2 (5.6 mg, 0.025 mmol) were added in one batch. The system was subsequently generated in about 24 h, and analyzed as described in the general procedure.</p><!><p>Aldehyde (3.0 mmol) was dissolved in anhydrous CH2Cl2 (30 mL) in a dry round-bottomed flask in the presence of activated 4 Å MS (whole beads, ca. 3.0 g). Amine (3.0 mmol) was added dropwise under N2, and the solution was stirred slowly under N2. The reaction was monitored by 1H NMR spectroscopy sampling, and upon consumption of starting materials the reaction mixture was filtered through a pad of Celite and concentrated in vacuo to directly obtain the imine in a typical purity of 98–99 %.</p><!><p>A dynamic imine system was equilibrated from compound 25 (88.3 mg, 0.25 mmol) and 3,4,5-trimethoxybenzylamine (4.3 μL, 4.9 mg, 0.025 mmol) for 48 h under the conditions described in the general procedure. After equilibrium was reached, the reaction solution was diluted with diethyl ether (2 mL) and the organic phase was washed with a saturated aqueous NH4Cl solution (2×1 mL) and brine (1 mL). The colorless solution was dried with MgSO4, filtered, and concentrated to afford a clean mixture of the four imines (79.1 mg, 90 % mass recovery). No amine or quinuclidine was present in the sample according to NMR spectroscopy analysis, and hydrolysis of the imines under the workup conditions was less than 1 %.</p><!><p>The imine systems could also be hydrolyzed to provide mixtures of aldehydes. The imine mixture (0.25 mmol imine content, either crude or purified according to the procedure outlined above) was dissolved in MeOH (1 mL) and a 1 m aqueous solution of HCl (1 mL) was added dropwise. The resulting solution was stirred at RT under air for 2 h, and the aldehydes were subsequently extracted with diethyl ether (3×2 mL). Drying with MgSO4, filtering, and concentration yielded the pure aldehydes without any side products in typically 95–99 % mass balance.</p>

PubMed Open Access

Uranium removal from seawater by means of polymeric fabrics grafted with diallyl oxalate through a single-step, solvent-free process

In order to test the effectiveness of oxalate-based polymeric adsorbents in the recovery of uranium from seawater, diallyl oxalate (DAOx) was grafted onto nylon 6 fabrics by exposing the fabric, immersed in pure liquid DAOx or in a surfactant-stabilized dispersion of DAOx in water, to electron beam or gamma radiation. Following drying and weighing to determine the degree of grafting (DoG), the presence of oxalate in the fabrics was verified using XPS. Zeta potential measurements showed the fabric surfaces to be negatively charged. The fabrics were tested by rotating them for 7 days in a rotary agitator with actual seawater spiked with 0.2 or 1.0 mg\xe2\x88\x99L\xe2\x88\x921 uranium. The fraction of uranium in the solution which was removed due to uptake on the fabrics was found to rise with increasing DoG at both uranium concentrations. EDS measurements were used to map the distribution of adsorbed uranium on the polymeric fibers.

uranium_removal_from_seawater_by_means_of_polymeric_fabrics_grafted_with_diallyl_oxalate_through_a_s

Introduction<!>Materials<!>Preparation of DAOx Grafted Fabrics<!>X-ray Photoelectron Spectroscopy<!>Microparticle Formulation and Zeta Potential Analysis<!>Adsorption Experiments<!>Scanning Electron Microscopy<!>Results and Discussion<!>Radiation grafting in the absence of solvents<!>Radiation grafting in aqueous solutions and in the presence of a surfactant<!>Surface characterization using XPS and Zeta Potential<!>Uranium Removal from Spiked Seawater<!>SEM/EDS Results<!>Conclusions

<p>Most of the work performed in recent decades on the extraction of uranium from seawater has been focused on polymer fabrics grafted with amidoxime groups1,2. However, the problems associated with the use of such polymer fabrics have not yet been completely resolved. One such problem is the necessity for pre-treatment of the amidoxime-grafted polymer fabrics with alkali in order to make them hydrophilic. This results in reduced chemical and mechanical stability and prevents the adsorbent fabrics from being effectively regenerated after use without significant loss of capacity. In view of these problems, other active groups, such as phosphates3, have been considered as alternatives, but their effectiveness with respect to uranium adsorption within the pH range of seawater has been found to be limited.</p><p>The present study was intended to examine the possibility of using oxalates attached to solid surfaces as sorbents for uranium. The oxalate ion forms a variety of complexes with the uranyl ion4,5. Early studies of the complexation constants showed that logarithm of the stability constant of [(UO2)(C2O4)] (logβ11) is approximately 6.36. This value significantly increases with increasing ionic strength but is only slightly affected by the pH5. Values of 6.77 and 12.00 were reported for the logarithms of the stability constants (β) in mildly acidic solutions for [(UO2)(C2O4)]0 and [(UO2)(C2O4)2]2−, respectively7. At a temperature of 25 °C and ionic strength of 3.0 M, the logβ values of the complexes [(UO2)(C2O4)]0, [(UO2)(C2O4)2]2−, [(UO2)(C2O4)3]4−, [(UO2)2(C2O4)3]2− and −, [(UO2)2(C2O4)5]6− were found to be 6.31, 11.21, 13.8, 18.5, and 28.5, respectively8. At an ionic strength of 0.05 M, the logβ value of [(UO2)(C2O4)]0 was reported to be 5.719. An extensive study including both uranyl oxalate and uranyl oxalate hydroxide complexes at 25 °C and an ionic strength of 1.0 M, yielded, for the complexes [(UO2)(C2O4)]0, [(UO2)(C2O4)2]2− and [(UO2)(C2O4)3]4−, logβ values of 5.87, 190.484 and 12.61, respectively, and for the complexes [(UO2)(C2O4)3]OH]−, [(UO2)(C2O4)2(OH)2]2−, [(UO2)(C2O4)2OH]3− and [(UO2)(C2O4)3OH]5−, logβ values of 0.62, −6.25, 3.93 and 5.32, respectively10. Based on these values, at pH 8, an ionic strength of 1.0 M, a temperature of 25 °C, a uranyl ion concentration of 5×10−4 M, and a much higher oxalate concentration of 0.1 M, the majority species is UO2(C2O4)(OH)22− (about 65%), followed by (UO2)3 (OH)7− (about 20%), UO2(C2O4)2OH3− (about 10%) and UO2(C2O4)OH− (about 5%). A roughly similar distribution is obtained under these conditions when the concentration of uranyl ion is lowered to 1×10−4 M and the concentration of oxalate to 3×10−4 M10.</p><p>Crystallographic measurements on UO2(C2O4)(H2O)3 have shown that each uranium atom exists as a linear (O-U-O)2+ ion with five secondary oxygen atoms coordinated to it in a perpendicular plane11. According to this study, the oxalate groups have a tetradentate nature, with each oxalate group acting as a bridge between two uranyl ions using all four oxygen atoms for coordination. The oxalate groups occupy centrosymmetric positions and are planar. Further studies confirmed that the uranium atom has pentagonal bipyramidal coordination12. The water molecules are hydrogen-bonded into zigzg chains. The resulting structure consists of [UO2(C2O4)(H2O)] chains12,13 with each third oxygen atom of the chain formed of water molecules coordinated to the uranium atom, with the uranyl oxalate chains linked into [UO2(C2O4)(H2O)]·2H2O layers13. The structural reported by the aforementioned papers, including the existence of hydrogen-bonded chains, have been supported by XRD and IR studies14. Furthermore, according to molecular dynamics simulations, the most stable form of the UO2(C2O4)(H2O)3 complex is a five-coordinate chelate15. It is reasonable to expect that the complex UO2(C2O4)(OH)22−, which is probably the dominant form of uranyl under the conditions prevailing near the surface of an oxalate-grafted polymeric surface at pH 8 (see above) may have a similar structure to the five-coordinated UO2(C2O4)(H2O)3 with two of the water molecules replaced by hydroxide groups. This is compatible with the structural models described above, which represent the UO2(C2O4)(H2O)3 complex in the form of [UO2(C2O4)(H2O)]∙2H2O, where two of the three water molecules being more amenable to change. Another finding that is compatible with this model is the observation that upon heating UO2(C2O4)(H2O)3 two of the three water molecules are lost at a much lower temperature than the third one6.</p><p>The present study was aimed at studying the effectiveness of polymer-grafted diallyl oxalate in removing uranium from seawater environments. Initial measurements performed with ammonium oxalate and dimethyl oxalate absorbed on active carbon showed these compounds to be largely ineffective in adsorbing uranium. However, diallyl oxalate (DAOx) was observed to have significant activity. In addition, the presence of double bonds in DAOx makes it possible to graft it onto polymeric substrates exposed to ionizing radiation. Accordingly, the study focused on measuring the uptake of uranium from seawater by polymeric fabrics grafted with diallyl oxalate under various conditions. A key objective of the study was to investigate the effects of grafting (direct or indirect), the medium (an aqueous or organic solution of DAOx or pure liquid DAOx), the radiation parameters (dose and dose rate), the dissolved gases present, etc., on the degree of grafting of the fabrics and on the effectiveness of these fabrics in removing uranium from seawater. Following the grafting process, the study was aimed at investigating the relationship between the degree of grafting and uranium uptake from seawater, as well as at characterization of the changes in chemical composition and morphology that take place during the grafting and the subsequent contact with seawater.</p><!><p>Diallyl oxalate (DAOx) was purchased from Monomer-Polymer and Dajac Labs. Milli-Q (Millipore, Billerica, MA, USA) Type 3 water was used in all experiments. TWEEN® 20 (Sigma), Arsenazo III (Sigma-Aldrich), ethanol (Sigma-Aldrich and Pharmco-AAPER), argon gas (Ar, industrial grade), and nitrous oxide (N2O, Grade 6.0) were used as received. Winged Fibers™ composed of nylon 6 were purchased from Allasso Industries®, Inc. For extraction experiments, seawater was collected from the Atlantic Ocean (34° 42′ N, 76° 43′ W). Uranyl acetate dihydrate (Fisher Scientific Company) was used in the preparation of U-spiked simulated seawater. The rotary agitator used to contact the fabrics with seawater was Model 3740-4-BRE, produced by Associated Design and Manufacturing Company.</p><!><p>Figure 1 describes the grafting polymerization of DAOx on nylon 6 and the subsequent extraction testing. Pieces of Nylon 6 Winged Fiber fabric were cut into square pieces, each weighing approximately 20 mg. These pieces were then placed in vials along with either 10 mL of TWEEN® 20 food additive surfactant and DAOx solution or with 1-2 mL of pure DAOx monomer. The combined TWEEN® 20 and DAOx solutions were prepared by mixing and stirring quantities of TWEEN® 20, DAOx, and deionized water for at least one hour. Each vial was then purged for 20 minutes using either N2O or Ar, respectively for the surfactant and pure monomer samples, in order to eliminate oxygen from the system and, in the case of N2O, also to convert hydrated electrons produced by the irradiation into hydroxyl radicals. Finally, the vials were wrapped with paraffin film around the caps in order the limit the diffusion of oxygen into the vials.</p><p>The DAOx monomer was grafted to the surface of the nylon 6 fabric through the use of electron beam radiation at an energy of about 10.5 MeV, generated by the Medical-Industrial Radiation Facility at the National Institute of Standards and Technology. Up to 16 samples could be irradiated at one time using a mechanical turntable. Irradiation doses and dose rates were initially calibrated using NIST-traceable alanine dosimetry films on two sides of multiple vials on the turntable. The total dose was calculated using the average dose of the two sides, wherein a linear dose-depth profile was assumed, in conjunction with the counts obtained from a Faraday cup positioned behind the rotating samples. Following the irradiation, some samples were heat treated by placing them in an oven at 50°C for one week. Finally, samples were washed to remove excess monomer and homopolymer from the sample. Pure DAOx solution samples were rinsed with ethanol, placed in ethanol, and sonicated for about ten minutes. DAOx and surfactant solutions were rinsed with DI water, placed in a solution of TWEEN® 20 and water, and sonicated for about ten minutes. Following this initial wash, both types of samples were rinsed and washed at least two more times with DI water with a ten minute sonication between each wash. Finally, the fabric samples were rinsed with DI water and dried overnight in an oven at 60 °C. These dried samples were then weighed to determine their degree of grafting (DoG), which is calculated from [(mf − mi)/mi] × 100, where mi and mf represent the initial and final fabric masses, respectively.</p><!><p>X-ray photoelectron spectroscopy (XPS) measurements were carried out using a Kratos Axis 165 spectrometer using monochromatic Al radiation (at a power of 280 W) with a vacuum level at or below 5 × 10−8 torr throughout the data collection process. Due to the insulating nature of the nylon 6 samples, charge neutralization was required to minimize sample charging. All survey spectra were collected with a pass energy of 160 eV and all high resolution spectra with a pass energy of 20 eV. All spectra were calibrated to C-C/C-H bonding at 284.8 eV.</p><!><p>Microparticle solutions were prepared using pieces of nylon 6 grafted fabric via the "solvent in water" precipitation method. Samples were first dissolved in acetic acid at 80°C, at a concentration of 20 mg/mL to form the diffusing phase (organic phase). This phase was then added drop by drop into filtered deionized water (DI), which is the dispersing phase (aqueous phase), under moderate magnetic stirring. The formation of microparticles was instantaneous and the solution was kept under mild agitation for 4 h to allow for particle stabilization. Small aliquots from each microparticle solution were diluted in DI water and titrated to pH 8.3, pH 4.5, and pH 3 using sodium hydroxide (NaOH) and hydrochloric acid (HCl) solutions prior to surface charge characterization. The zeta potential of nylon 6 microparticles was assessed by means of electrophoretic light scattering (Zetasizer nano-ZS90; Malvern instruments; Westborough, MA).</p><!><p>Test solutions of uranium in seawater were prepared by dissolving a suitable quantity of uranyl acetate dihydrate in Atlantic Ocean Seawater. In each test, the adsorbent sample was added to a desired volume of a solution of uranium in seawater and the combination of test solution and solid adsorbent was rotated for a desired period of time at 30 rpm in a rotating agitator. At the end of this period, the solution was separated from the adsorbent and analyzed to determine the amount of uranium remaining in the solution by means of a spectrophotometric method based on the use of Arsenazo III as a color-forming reagent16. Arsenazo III forms a complex with uranium with an absorption peak at 651 nm. Each test was carried out 4 times under identical conditions. The fraction of the uranium taken up by the adsorbent was calculated from the difference in concentration of uranium in the test solution resulting from the contact with the adsorbent. The remaining solid adsorbent was dried in air and set aside for characterization by means of spectroscopic techniques such as XPS and SEM-EDS.</p><!><p>A scanning electron microscope (SEM) equipped with an energy dispersive spectroscopy (EDS) detector was used to characterize the morphology of the adsorbent fabric and to identify the chemical species extracted by it. Adsorbent samples that have been exposed to Atlantic Ocean Seawater seawater during the adsorption experiments were dried overnight in a vacuum dessicator and mounted on a SME aluminum stub for EDS analysis. The EDS detector was part of a Hitachi S-3400 variable pressure SEM.</p><!><p>Direct radiation grafting of DAOx onto nylon 6 was performed in the absence of oxygen in neat liquid DAOx and also in aqueous solutions in the presence of a surfactant.</p><!><p>In these experiments, the oxygen-free mixtures contain only DAOx and nylon 6. Figure 2 shows that as the dose increases, the degree of grafting increases. The irradiation was carried out at room temperature at varying dose rates. As shown in Scheme 1, radiation induces the formation of the C-centered free radicals of nylon● and DAOx●. These free radicals undergo various desired reactions (grafting through the formation of C-C bonds between DAOx and the nylon 6), as well as undesired reactions, consisting of DAOx homopolymerization and the crosslinking reactions of nylon●.</p><p>Therefore, under our experimental conditions, the decay rates of the DAOx● and nylon● can be expressed as follows: −d[DAOx•]dt=k1[nylon•][DAOx•]+2k2[DAOx•]2+k3[DAOx•][DAOx]+k4[nylon−(DAOx)n−DAOx•][DAOx•] −d[nylon•]dt=k1[nylon•][DAOx•]+k5[nylon•][DAOx]+2k6[nylon•]2where k1, k2, k3, k4, k5, k6 are the rate constants of the desired grafting reaction, the undesired termination reaction of the DAOx● radicals, the undesired homopolymerization reaction of the DAOx, the termination reaction of the growing grafted DAOx● radicals, the desired grafting-addition reaction of the DAOx on nylon, and the undesired crosslinking reaction of the nylon, respectively.</p><p>Figure 2 also shows that after receiving a dose of more than approximately 175 kGy, the grafting density increases much more sharply as a function of dose. This can be explained by the fact that as the viscosity of the medium increases with increasing dose, the diffusion of the DAOx● radicals is slowed down, thus hindering the homopolymerization reactions and enhancing the grafting on the nylon surface.</p><!><p>Degrees of grafting as high as 25 % have been reached in the aqueous, N2O-saturated mixtures containing 0.11 M DAOx and 4.5 × 10−3 M TWEEN® 20. Under these experimental conditions, water absorbed most of the electrons from the electron accelerator resulting in the formation of the following active species with their radiation chemical yields in micromole per joule: G( ●OH)=G(eaq−)=G(H3O+)=0.28,G( ●H)=0.062,G(H2)=0.042,G(H2O2)=0.082</p><p>Hydroxyl radicals (●OH) constitute a powerful oxidant, and they are highly reactive (through addition, abstraction or electron transfer). The ●OH radicals are responsible for initiating the grafting polymerization and other reactions in this system through the production of DAOx● and nylon● radicals upon reacting with nylon and DAOx. On the other hand, hydrated electrons (eaq−) are very strong reducing radicals and can be converted to ●OH radicals through the following reactions: N2O+eaq−+H2O→•OH+N2+OH−</p><p>The above reaction is very fast, having a reaction rate constant of k= 8 ×109 M−1s−1 17. Hence, saturating the system with N2O prior to irradiation, would double the ●OH yield to G(●OH) = 0.56 micromole per joules. In addition to ●OH, H-atoms (●H) with G(●H)= 0.062 micromole per joule, also react with nylon and DAOx to produce DAOx● and nylon●.</p><p>Under these irradiation conditions and in the absence of oxygen, the radiolytically produced ●OH radicals and H-atoms add to the unsaturation site of DAOx and abstract H-atom from the backbone of the polymer substrate (nylon) producing OH—DAOx●, and H-DAOx● radicals, and ●nylon (-H) radicals, respectively. It should be mentioned that ●OH and ●H are also scavenged by the TWEEN surfactant, since the mixture contains, 4.2 × 10−3 M TWEEN, leading to a decrease in the concentrations of OH—DAOx●, H-DAOx●, and ●nylon (-H) radicals. This also leads to the possibility of TWEEN being grafted on nylon 6. This can dramatically decrease the number of sites available for uranium adsorption.</p><!><p>Figure 3 shows the XPS spectra of the ungrafted nylon and of the nylon radiation-grafted with DAOx. The XPS results demonstrate the presence of C-C/CH, oxalate, amide, C-N, and ester groups.</p><p>In order to reinforce the hypothesis that the uranyl ion is binding to the negative oxalate group attached to the nylon 6 fabric, zeta potential measurements were performed on grafted and ungrafted nylon 6 fabric that had been chemically transformed into microparticles. The zeta potentials for both the grafted and ungrafted nylon 6 fabric were measured and the results are summarized in Table 1. These results show that at a pH of about 8 the surface of the oxalate-grafted fabric is negatively charged and thus suitable for extraction of UO22+ from the seawater.</p><!><p>The results obtained for the removal of uranium from spiked seawater by means of nylon 6 fabrics grafted with neat diallyl oxalate are shown in Figure 4. The level of spiking was either 1.0 mg/L or 0.2 mg/L U (introduced as uranyl acetate). In each test, a sample weighing approximately 30 mg was rotated with 10 mL of the spiked seawater at 30 rpm for 7 days. The results show that significant removal of uranium from these solutions (>5%) took place when the degree of grafting exceeded approximately 18%, and, in general, the percent removal of uranium from both the 1.0-mg/L and 0.2-mg/L U solutions increased with increasing degree of grafting. The large scatter in the data can be ascribed to the non-uniform distribution of the grafted material on the nylon 6 fibers as observed by SEM (see below). Another reason for the scatter is associated with the fact that some of the adsorbent samples were subjected, following the radiation induced grafting, to heat treatment at 50°C for 7 days. The percent removal of uranium from the test solutions observed with these samples was generally higher than the corresponding percent removal observed with samples which did not undergo heat treatment. It should also be noted that the amounts of uranium in the test solutions were very small (0.01 or 0.002 mg, respectively) so that the results in Figure 4 cannot provide a realistic estimate of the maximum amount of uranium that can be removed by the fabric from large volumes of seawater.</p><p>On the other hand, the extent of extraction of uranyl ion from seawater by means of fibers grafted with DAOx in an aqueous environment in the presence of TWEEN was very low (<5%). This may be due to the grafting of TWEEN onto nylon, which hinders the grafting of DAOx onto nylon, since a large fraction of the ●OH radicals may react with TWEEN, producing TWEEN● radicals.</p><!><p>The SEM observations showed a non-uniform distribution of the grafted material on the fibers, with some sections of the fibers coated with significant amounts of grafted material and other sections free of such coating. However, the EDS results (see Table 2) obtained for the fabrics following contact with the uranium-spiked seawater were remarkably similar for different regions of the fabrics, showing that the relative affinities of the adsorbent for the various ionic solutes in the seawater were consistent across the entire structure of the fabric.</p><!><p>In conclusion, our results demonstrate that radiation grafting polymerization of DAOx on nylon 6 through a solvent-free, single-step-direct process can be accomplished in the absence of oxygen. The desired reaction between the radiolytically produced nylon● and DAOx● C-centered radicals to form the grafting C-C bonds occurs despite the strong competition from the DAOx homo-polymerization reaction. These results also show that at radiation doses up to around 175 kGy, the undesirable homo-polymerization is the predominant reaction. However, as the viscosity increases due to the homo-polymerization reaction, the diffusion of the DAOx● C-centered radicals is slowed down. This hinders the homo-polymerization reaction and enhances the local grafting reaction, and this allows reaching a grafting density of 140% at a dose level of 250 kGy.</p><p>A further series of experiments employed direct radiation grafting of DAOx onto nylon 6 using N2O-saturated aqueous solutions containing DAOx and the surfactant TWEEN. These experiments showed that a degree of grafting 25% can be achieved at dose of 250 kGy. This relatively low grafting density may be explained by the fact that the surfactant TWEEN also scavenges the ●OH radicals, causing a decrease in the radiolytic yields of nylon● and DAOx● C-centered radicals, and thus its presence causes a decrease in the grafting density.</p><p>As expected, Figure 4 shows that as the grafting density increases, the extraction of uranyl from spiked seawater increase. The scattering of the results of the percentage extraction can be related to the non-uniformity of the grafting within the samples. This non-uniformity in the grafting is the principal disadvantage of the solvent-free grafting. Notwithstanding this disadvantage, removal of as much as 50% of the uranium from the test solution was achieved using adsorbent fabrics produced using this method. In general, it was observed that the measured extent of uranium uptake from the seawater, whether spiked with 1 mg/L or 0.2 mg/L of uranium, increased with the degree of grafting of oxalate on the polymeric fabric.</p>

PubMed Author Manuscript

Synthesis of novel tricyclic chromenone-based inhibitors of IRE-1 RNase activity

Inositol-requiring enzyme 1 (IRE-1) is a kinase/RNase ER stress sensor that is activated in response to excessive accumulation of unfolded proteins, hypoxic conditions, calcium imbalance, and other stress stimuli. Activation of IRE-1 RNase function exerts a cytoprotective effect and has been implicated in the progression of cancer via increased expression of the transcription factor XBP-1s. Here, we describe the synthesis and biological evaluation of novel chromenone-based covalent inhibitors of IRE-1. Preparation of a family of 8-formyl-tetrahydrochromeno[3,4-c]pyridines was achieved via a Duff formylation that is attended by an unusual cyclization reaction. Biological evaluation in vitro and in whole cells led to the identification of 30 as a potent inhibitor of IRE-1 RNase activity and XBP-1s expression in wild type B cells and human mantle cell lymphoma cell lines.

synthesis_of_novel_tricyclic_chromenone-based_inhibitors_of_ire-1_rnase_activity

INTRODUCTION<!>FRET-suppression assay of potential IRE-1 inhibitors<!>Synthesis of tricyclic chromenones<!>Structure-activity relationships<!>Inhibition of XBP-1s expression in whole cells<!>CONCLUSION<!>General synthesis notes<!>Procedure for synthesis of \xce\xb2-ketoesters 18a\xe2\x80\x93d<!>Methyl 5-(((allyloxy)carbonyl)amino)-3-oxobutanoate (18a)<!>Methyl 5-(((allyloxy)carbonyl)amino)-3-oxopentanoate (18b)<!>Allyl 2-(2-methoxy-2-oxoethylidene)pyrrolidine-1-carboxylate (18c)<!>Methyl 7-(((allyloxy)carbonyl)amino)-3-oxoheptanoate (18d)<!>General procedure for synthesis of coumarins 19a\xe2\x80\x93d<!>Allyl (2-(7-hydroxy-2-oxo-2H-chromen-4-yl)methyl)carbamate (19a)<!>Allyl (2-(7-hydroxy-2-oxo-2H-chromen-4-yl)ethyl)carbamate (19b)<!>Allyl (2-(7-hydroxy-2-oxo-2H-chromen-4-yl)propyl)carbamate (19c)<!>Allyl (2-(7-hydroxy-2-oxo-2H-chromen-4-yl)butyl)carbamate (19d)<!>Duff reaction condition A<!>Duff reaction condition B<!>Duff reaction condition C<!>Allyl (2-(8-formyl-7-hydroxy-2-oxo-2H-chromen-4-yl)methyl)carbamate (20a)<!>Allyl (2-(8-formyl-7-hydroxy-2-oxo-2H-chromen-4-yl)ethyl)carbamate (20b)<!>Allyl (2-(8-formyl-7-hydroxy-2-oxo-2H-chromen-4-yl)propyl)carbamate (20c)<!>Allyl (2-(8-formyl-7-hydroxy-2-oxo-2H-chromen-4-yl)butyl)carbamate (20d)<!>Allyl 7-formyl-8-hydroxy-5-oxo-4,5-dihydro-1H-chromeno[3,4-c]pyridine-3(2H)-carboxylate (21b)<!>Allyl 8-formyl-9-hydroxy-6-oxo-2,3,5,6-tetrahydrochromeno[3,4-c]azepine-4(1H)-carboxylate (21c)<!>Allyl 9-formyl-10-hydroxy-7-oxo-2,3,5,6-tetrahydro-1H-chromeno[3,4-d]azocine-4(7H)-carboxylate (21d)<!>Allyl 8-hydroxy-7-(hydroxymethyl)-5-oxo-4,5-dihydro-1H-chromeno[3,4-c]pyridine-3(2H)-carboxylate (22)<!>Allyl 8-hydroxy-7-chloro-5-oxo-4,5-dihydro-1H-chromeno[3,4-c]pyridine-3(2H)-carboxylate (23)<!>Allyl 7-(1,3-dioxan-2-yl)-8-hydroxy-5-oxo-4,5-dihydro-1H-chromeno[3,4-c]pyridine-3(2H)-carboxylate (24)<!>Allyl 7-(1,3-dithian-2-yl)-8-hydroxy-5-oxo-4,5-dihydro-1H-chromeno[3,4-c]pyridine-3(2H)-carboxylate (25)<!>Allyl 7-formyl-8-methoxy-5-oxo-4,5-dihydro-1H-chromeno[3,4-c]pyridine-3(2H)-carboxylate (26)<!>Allyl 7-formyl-8-benzyloxy-5-oxo-4,5-dihydro-1H-chromeno[3,4-c]pyridine-3(2H)-carboxylate (27)<!>7-(1,3-dioxan-2-yl)-8-hydroxy-3,4-dihydro-1H-chromeno[3,4-c]pyridin-5(2H)-one (28)<!>3-Acetyl-8-hydroxy-5-oxo-2,3,4,5-tetrahydro-1H-chromeno[3,4-c]pyridine-7-carbaldehyde (29)<!>8-Hydroxy-3-methyl-5-oxo-2,3,4,5-tetrahydro-1H-chromeno[3,4-c]pyridine-7-carbaldehyde (30)<!>3-Benzyl-8-hydroxy-5-oxo-2,3,4,5-tetrahydro-1H-chromeno[3,4-c]pyridine-7-carbaldehyde (31)<!>3-(4-Fluorobenzyl)-8-hydroxy-5-oxo-2,3,4,5-tetrahydro-1H-chromeno[3,4-c]pyridine-7-carbaldehyde (32)<!>8-Hydroxy-3-(2-methylallyl)-5-oxo-2,3,4,5-tetrahydro-1H-chromeno[3,4-c]pyridine-7-carbaldehyde (33)<!>7-Formyl-8-hydroxy-5-oxo-4,5-dihydro-1H-chromeno[3,4-c]pyridine-3(2H)-carboximidamide (34)<!>Allyl 7-formyl-8-(methoxymethoxy)-5-oxo-4,5-dihydro-1H-chromeno[3,4-c]pyridine-3(2H)-carboxylate (35)<!>Allyl 7-acetyl-8-hydroxy-5-oxo-4,5-dihydro-1H-chromeno[3,4-c]pyridine-3(2H)-carboxylate (36)<!>Allyl 7-(3-ethoxy-3-oxoprop-1-en-1-yl)-8-hydroxy-5-oxo-4,5-dihydro-1H-chromeno[3,4-c]pyridine-3(2H)-carboxylate (37)<!>Allyl 8-hydroxy-7-(2-(methylsulfonyl)vinyl)-5-oxo-4,5-dihydro-1H-chromeno[3,4-c]pyridine-3(2H)-carboxylate (38)<!>3-((Allyloxy)carbonyl)-8-(methoxymethoxy)-5-oxo-2,3,4,5-tetrahydro-1H-chromeno[3,4-c]pyridine-7-carboxylic acid (39)<!>3-((Allyloxy)carbonyl)-8-hydroxy-5-oxo-2,3,4,5-tetrahydro-1H-chromeno[3,4-c]pyridine-7-carboxylic acid (40)<!>3-Allyl 7-methyl 8-hydroxy-5-oxo-4,5-dihydro-1H-chromeno[3,4-c]pyridine-3,7(2H)-dicarboxylate (41)<!>Allyl 8-hydroxy-7-(methoxy(methyl)carbamoyl)-5-oxo-4,5-dihydro-1H-chromeno[3,4-c]pyridine-3(2H)-carboxylate (42)<!>Recombinant human IRE-1 expression and purification<!>In vitro IRE-1 RNase FRET-suppression assay<!>Antibodies and reagents<!>Cell culture<!>Protein isolation and immunoblotting<!>Cell proliferation XTT assays<!>

<p>The endoplasmic reticulum (ER) stress response is a cytoprotective mechanism activated in response to proteotoxic burden and is crucial for homeostatic regulation.1,2 Disruption in the stoichiometric balance, transport, or processing of intracellular proteins leads to the activation of 3 distinct pathways mediated by the ER stress sensor proteins IRE-1, ATF6, and PERK. IRE-1 is unique in that it contains a stress sensor domain in the lumen of the ER and a cytosolic serine/threonine kinase domain linked to an RNase domain. Multiple stress conditions can cause IRE-1 to oligomerize. Oligomerization brings the IRE-1 cytoplasmic kinase domains into close proximity, allowing for autophosphorylation and activation IRE-1 RNase activity. The IRE-1 RNase domain is critical for the function of IRE-1, because it splices 26 nucleotides from the mRNA of X-box binding protein 1 (XBP-1), causing a frame shift in translation.3–5 The spliced XBP-1 mRNA encodes a functional 54-kDa XBP-1s protein in mammalian cells, which is a transcription factor that translocates into the nucleus and regulates ER stress response genes.</p><p>Since gene copy number amplifications and aberrant protein expression are hallmarks of cancer, many human tumors rely on a robust ER stress response for growth and survival.6,7 As a result, IRE-1 and related stress sensors have emerged as potential therapeutic targets for the treatment of cancer.8,9,10 IRE-1-mediated activation of XBP-1 has also been implicated in the evasion of virus-induced cytotoxicity11 as well as in the development of inflammatory arthritis.12,13 Small molecules capable of modulating IRE-1 RNase activity and XBP-1s transcription thus represent useful chemical tools and potential therapeutic agents.</p><p>Efforts to identify inhibitors of IRE-1 RNase function have relied primarily on high-throughput screening of large chemical libraries. This has led to the discovery of various salicylaldehydes with in vitro activity against IRE-1-mediated mRNA splicing.14–17 A limited number of non-electrophilic inhibitors of IRE-1 RNase activity have also been reported (Figure 1).18,19 While aldehydes and related functional groups are generally considered undesirable with respect to chemical probe development, the recent FDA-approval of various electrophilic drugs has renewed interest in covalent inhibitors.20 The importance of the aldehyde moiety for potent IRE-1 RNase inhibition by 5 (4µ8C)15 and related compounds has been rationalized by the formation of an unusually stable Schiff base with lysine 907 in the IRE-1 endonuclease domain.21 Although IRE-1 contains 25 lysine residues in its cytosolic domain, only covalent modification at K907 (and in some cases K599) is observed in vivo.15 This selectivity has been attributed to specific perturbation of the K907 ε-amino group pKa, resulting in enhanced nucleophilicity, increased rate of Schiff base formation with aldehyde inhibitors, and slow off-rate.</p><p>Here, we report the synthesis and biological evaluation of novel chromenone-based inhibitors of IRE-1 RNase activity. A tandem Duff formylation/annelation reaction en route to candidate inhibtors gave rise to fused tricyclic chromenopiperidine, chromenoazepane, and chromenoazecane scaffolds. Selected analogs based on a tetrahydrochromeno[3,4-c]pyridine core structure potently inhibit XBP-1 splicing in vitro and block the expression of XBP-1s in whole cells, making them useful compounds for interrogating IRE-1 RNase activity in biological systems.</p><!><p>To assess the in vitro activity of potential IRE-1 RNase inhibitors, we carried out the expression and purification of recombinant human IRE-1 for use in an in vitro FRET-suppression assay.17 The cytoplasmic kinase/RNase domain (aa. 547–977) of human IRE-1 was expressed as a soluble puritin-His-tagged 59 kD fusion protein in SF21 cells and purified by Ni-NTA affinity chromatography. To confirm that hIRE-1 exhibited a functional RNase domain, we evaluated its activity in vitro using a synthetic mRNA stem-loop corresponding to the XBP-1 substrate sequence. This stem-loop incorporates a Cy5 fluorophore on its 5' end and the black hole quencher (BHQ) on its 3' end, resulting in fluorescence only upon site-specific cleavage by the protein. Protein (5 nM) was incubated in a 96-well plate at room temperature with different concentrations of the XPB-1 stem loop for up to 2h, and fluorescence was measured upon excitation and emission at 620 and 680 nm, respectively. Recombinant hIRE-1 exhibited functional RNase activity a Km value of 45 nM (see Supporting Information).</p><p>We first evaluated a small set of known IRE-1 inhibitors, synthetic analogs, and selected commercially available salicyladehyde derivatives using the FRET-suppression assay (Figure 2). We recently reported the in vivo characterization of naphthaldehyde derivative 2 (A-I06), which was postulated to be the bioactive breakdown product of the known IRE-1 inhibitor 1 (STF-038010).22 When evaluated in our assay, 1 and 2 exhibit similar IC50 values (9.94 and 9.73 µM, respectively), while decomposition product 8 and reduced derivative 9 showed no appreciable inhibition at 20 µM. Interestingly, the salicylaldehyde moiety alone was not sufficient for IRE-1 RNase inhibition, as evidenced by the weak activity (>20 µM IC50) of compounds 10–13. Modification of the aldehyde or phenol functionalities also resulted in inactive compounds (14–16). Coumarin derivative 5, recently identified in a high-throughput screening effort,15 exhibited significantly enhanced potency against IRE-1 RNase function with an IC50 value of 206 nM in our FRET-suppression assay.</p><!><p>In an effort toward functionalized derivatives of 5 for use in covalent tagging and pulldown experiments, we synthesized analogs 20a–d in 4 steps from the appropriate amino acids (Figure 3). Installation of the aldehyde moiety in each case relied on a Duff formylation carried out using hexamethylenetetramine (HMTA) in refluxing glacial acetic acid. Interestingly, when the reaction was carried out in refluxing TFA using intermediate 19b as a starting material, formylation was attended by an annulation reaction involving the pendant carbamate nitrogen to give tetrahydrochromeno[3,4-c]pyridine 21b as the sole product. The structure and connectivity of this tricyclic scaffold was confirmed by HMBC NMR. As is typically the case for Duff formylations,23,24 complete consumption of 19 still resulted in low yields of 20 and 21 due to significant decomposition. However, the yield of 21b improved to 41% when the reaction was preceded by acetylation of the o-hydroxyl group.</p><p>A plausible mechanism for the formation of 21b involves electrophilic aromatic substitution at position 3 of the chromenone core (Scheme 1). The reaction of electron rich aromatics with HMTA in organic acid occasionally results in aminomethylation in addition to formylation via decomposition of intermediates such as B.23,24 In the case of 21b, this decomposition is likely precluded by attack of the carbamate nitrogen onto the electrophilic methylene group in C. The interrupted Duff reaction at position 3 presumably occurs prior to formylation at position 8, as the use of only 1 equivalent of HMTA in refluxing TFA afforded intermediate D as the major product from 19b. The concomitant annulation reaction was not observed in the case of substrate 19a under any of the conditions listed in Figure 2. However, hexahydrochromeno[3,4-c]azepine 21c and hexahydrochromeno[3,4-c]azocine 21d were isolated as the sole products from 19c and 19d when TFA was used as the solvent.</p><!><p>When evaluated in the FRET-suppression assay, bicyclic derivatives 19a–d, exhibited inhibitory activities in the 100–500 nM range (Figure 4). The constrained tricyclic derivative 21b consistently showed enhanced activity against IRE-1 RNase activity relative to the bicyclic compounds 20b and 5 in side-by-side experiments. Given the optimal in vitro potency and chemical yield of 21b, we carried out the synthesis of a family of analogs to assess importance of the hydroxyl group and the distal N-substituent (Scheme 2). A potential covalent irreversible inhibitor 23 was obtained by chlorination of the reduced derivative 22. Compounds 24 and 25 were prepared by acid-catalyzed protection of the aldehyde in 21b as the 1,3-dioxane or dithiane derivative. Analogs 26 and 27 were prepared by O-alkylation of 24, followed by acidic hydrolysis of the dioxane. Compounds 29–34 were synthesized by reaction of intermediate 28 with various acylating or alkylating reagents, followed by acidolysis.</p><p>The presumed importance of the aldehyde functionality for IRE-1 RNase inhibition also prompted us to explore alternative electrophilic groups at the 8 position of the chromenone core. Scheme 3 depicts the synthesis of analogs 36–42 from compound 21b. Formation of the ketone in 36 via oxidation of the Grignard product required prior protection of the o-hydroxyl as methoxymethyl ether 35. Olefination of 35 and acetal hydrolysis afforded electrophilic analogs 37 and 38. Oxidized variants 40–42 were synthesized via Pinnick oxidation of 35.</p><p>All compounds were evaluated by FRET-suppression assay in side-by-side experiments using 21b as a control inhibitor (Table 1). As anticipated, protection of the aldehyde group in 21b as the 1,3-dioxane or dithiane acetal (24 and 25) resulted in weaker IRE-1 inhibitory activity. Alkylation of the phenol oxygen (compounds 26, 27, and 35) resulted in a complete loss of potency below 20 µM. The N-acyl derivative 29 exhibited an IC50 value of 312 nM while N-alkyl analogs 30–33 were found to be slightly more potent. Interestingly, N-benzyl analog 31 was almost 3-fold more active than the corresponding fluorinated derivative 32. Guanidinylation to give 34 resulted in a notable increase in potency (IC50 = 47 nM) relative to the parent compound, though solubility significantly decreased. Ketone 36, vinyl sulfone 38, and Weinreb amide 42 showed no significant IRE-1 RNase inhibitory activity below 20 µM. However, electrophilic compounds 37, 40, and 41 displayed moderate potency (1–5 µM) in vitro. Also of note, 1,3-dioxane derivative 24 exhibited an in vitro IC50 of 3.1 µM, whereas the corresponding 1,3-dithiane analog 25 displayed more than 5-fold weaker activity. To confirm that the enhanced inhibitory activity of 24 is not simply a function of a labile aldehyde masking group, we carried out stability studies in assay buffer and observed no significant decomposition of the 1,3-dioxane moiety over 12 hours (see Supporting Information).</p><!><p>In order to determine whether our inhibitors could block the expression of XBP-1s in whole cells, we incubated LPS-stimulated B cells from the spleens of wild-type mice with 20 µM of selected compounds for 24 hours, lysed the cells, and analyzed the lysates for the expression of XBP-1s by immunoblots. Compounds 29 and 30 potently suppress the expression of XBP-1s at 20 µM in wild-type mouse B cells (Figure 5A). In addition, 5, 21b, and 24 exhibit strong inhibition of XBP-1s, as does treatment with 50 µM of 2. Despite their activity in the FRET-suppression assay, compounds 31–34 did not effectively inhibit XBP-1s expression in whole cells, presumably due to poor cell permeability and solubility. Compounds 37, 40, and 41, which feature alternative electrophilic functional groups, similarly showed little to no inhibitory effect on XBP-1s expression in B cells at 20 µM. Consistent with previous results showing upregulation of IRE-1 in response to XBP-1s deficiency25 and suppression,22 we observed an inverse correlation between pharmacological inhibition of XBP-1s and expression level of IRE-1 (Figure 5A).</p><p>The IRE-1/XBP-1 pathway is known to be critical for the survival multiple myeloma, malignancies derives from plasma cells.14,26 However, the functional role of the ER stress response in leukemia or lymphoma derived from mature B cells has been largely overlooked because leukemia and lymphoma cells do not expand their ER like that of multiple myeloma cells. We recently showed that chronic lymphocytic leukemia (CLL) growth and survival is highly dependent on the IRE-1/XBP-1 pathway and is inhibited by small molecules targeting IRE-1 RNase activity.22 Mantle cell lymphoma (MCL) is an incurable non-Hodgkin's lymphoma developed from mantle zone-resident B cells. Since the role of the IRE-1/XBP-1 pathway in MCL is completely unknown, we examined the MCL cell lines Mino and Jeko for the expression of XBP-1s, and discovered that XBP-1s is constitutively expressed by both. A subset of inhibitors was examined for inhibition of XBP-1s in these human MCL cell lines. As with wild-type mouse B cells, compounds 21b, 29, and 30 potently suppress the expression of XBP-1s and induce upregulation of IRE-1 in Mino and Jeko cells. N-isobutenyl derivative 33 also exhibits significant activity at 20 µM (Figure 5B and 5C).</p><p>To establish the dependency of XBP-1s expression on inhibitor concentration we used MCL cells to determine the whole cell IC50 values for 21b, 29, and 30, in comparison to 5, by immunoblots and densitometry (Figure 5D–G). Compound 30 proved to be the most potent inhibitor of XBP-1s expression in both Mino and Jeko cell lines (IC50 = 0.57 and 0.98 µM, respectively).</p><p>Lastly, we carried out XTT dose-response experiments to determine approximate GI50 concentrations for 30, our most potent inhibitor of XBP-1s expression. After 48 h treatment, 30 exhibited GI50 values of 34 and 19 µM in Mino and Jeko cells, respectively (Figure 6A). Total growth inhibition by 30 was achieved between 55 and 66 µM for these cell lines. We confirmed that growth inhibition is the result of apoptosis by treating Mino and Jeko cells with 30 for 72 h and analyzing cell lysates for cleaved PARP. Consistent with its superior potency in the suppression of XBP-1s, compound 30 induced PARP cleavage more strongly than either 21b or 5 at 50 µM (Figure 6B). We also determined a GI50 value of ~34 µM in LPS-stimulated wild-type mouse B cells after treatment with 30 for 72 hours (see Supporting Information). As expected, this result suggests that the growth of antibody-secreting plasma cells are also sensitive to inhibition of IRE-1 RNase activity.</p><!><p>We have described the synthesis and biological characterization of novel inhibitors of IRE-1. Although various salicylaldehydes have been reported to inhibit IRE-1 RNase activity in vitro, our results confirm that the presence of an o-hydroxy aromatic aldehyde is not sufficient for biological activity. In an effort toward functionalized derivatives of potent chromenone-based inhibitors, we prepared a series of carbamate substituted 2H-chromene-2-ones for further derivatization. Duff formylation of these substrates resulted in a tandem annelation reaction, giving rise to novel fused tricyclic scaffolds. Tetrahydrochromeno[3,4-c]pyridine 21b served as a lead compound for the synthesis of a family of analogs.</p><p>Although replacement of the critical aldehyde group in 21b with electrophilic surrogates diminished potency, some compounds retained weak to moderate inhibitory activity in vitro. Modifications to the phenol group in 21b had a deleterious effect on potency in the FRET suppression assay, while changes at the distal N substituent were generally well tolerated. The ability of selected compounds to inhibit XBP-1s expression in wild-type B cells and human MCL cell lines highlights the importance of cell-based assays for this class of inhibitors, as a number of compounds with low- to mid-nanomolar activity in the FRET-suppression assay did not significantly reduce XBP-1s expression in whole cells. The N-methyl analog 30 displayed an in vitro IRE-1 RNase IC50 value of 200 nM and potently inhibited the expression of XBP-1s in Mino and Jeko cells (IC50 = 0.57 and 0.98 µM, respectively). Compared to 21b, compound 30 is also more effective at inducing apoptosis in MCL cells. The described tricyclic chromenones thus represent useful tool compounds for suppressing IRE-1 RNase activity in whole cells and for probing the importance of the IRE-1/XBP-1 pathway of the ER stress response in biological systems.</p><!><p>Unless stated otherwise, reactions were performed in flame-dried glassware under a positive pressure of argon or nitrogen gas using dry solvents. Commercial grade reagents and solvents were used without further purification except where noted. Diethyl ether, toluene, dimethylformamide dichloromethane, and tetrahydrofuran were purified by a Glass Contour column-based solvent purification system. Other anhydrous solvents were purchased directly from chemical suppliers. Thin-layer chromatography (TLC) was performed using silica gel 60 F254 pre-coated plates (0.25 mm). Flash chromatography was performed using silica gel (60 µm particle size). The purity of all compounds was judged by TLC analysis (single spot/two solvent systems) using a UV lamp, CAM (ceric ammonium molybdate), ninhydrin, or basic KMnO4 stain(s) for detection purposes. 1D and 2D NMR spectra were recorded on a Varian 400 MHz spectrometer. Proton chemical shifts are reported as δ values relative to residual signals from deuterated solvents (CDCl3, CD3OD, or DMSO-d6). The purity of all assayed compounds was determined by RP-HPLC using an analytical C18 column with MeCN/water (0.1% formic acid) as eluent (4 × 150mm column, 1 mL/min flow rate). All final compounds were determined to be between 95 and 98% pure. Compounds 2, 5, 8, 10–12, and 14-were purchased from commercial sources. Compounds 1 and 9 were synthesized as described previously.22</p><!><p>A solution of the appropriate (N-Alloc) amino acid 17 (23.9 mmol) in 100 mL of DCM at 0 °C was treated with 2,2-dimethyl-1,3-dioxane-4,6-dione (4.47 g, 31.0 mmol), 4-dimethylaminopyridine (2.92 g, 23.9 mmol), and diisopropylcarbodiimide (3.70 mL, 23.9 mmol). The reaction was stirred from 0 °C to rt over 4 h, then washed with 10% aq. KHSO4 followed by brine. The organic layer was dried over Na2SO4 and concentrated. The resulting colorless liquid was dissolved in 50 mL of a 10:1 MeOH:toluene mixture and stirred at reflux for 15 h. After cooling, the reaction was concentrated under reduced pressure. Purification by flash column chromatography over silica gel (25%–60% EtOAc/hexanes) afforded 18a, 18b, and 18d as colorless oils. Alkylidene pyrrolidine 18c was obtained as a white solid.</p><!><p>Obtained in 64% yield from 17a. 1H NMR (400 MHz, CDCl3) δ 5.88 (ddt, J = 16.2, 10.7, 5.6 Hz, 1H), 5.49 (s, 1H), 5.29 (d, J = 17.2 Hz, 1H), 5.20 (d, J = 10.5 Hz, 1H), 4.56 (d, J = 5.5 Hz, 2H), 4.18 (d, J = 5.1 Hz, 2H), 3.72 (s, 3H), 3.50 (s, 2H); 13C NMR (101 MHz, CDCl3) δ 198.2. 167.0, 156.1, 132.5, 117.9, 66.0, 52.6, 50.8, 46.2; HRMS (ESI-TOF) m/z [M + H]+ calcd for C9H14NO5 216.0867, found 216.0862.</p><!><p>Obtained in 94% yield from 17b. 1H NMR (400 MHz, CDCl3) δ 5.97 – 5.82 (m, 1H), 5.37 – 5.12 (m, 3H), 4.53 (d, J = 5.6 Hz, 2H), 3.73 (s, 3H), 3.50 – 3.37 (m, 4H), 2.80 (t, J = 5.7 Hz, 2H); 13C NMR (101 MHz, CDCl3) δ 202.2, 167.3, 156.2, 132.8, 132.8, 117.6, 117.5, 65.4, 52.4, 52.4, 48.9, 42.8, 35.3; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C10H16NO5 230.10285, found 230.10297.</p><!><p>Obtained in 56% yield from 17c. 1H NMR (400 MHz, CDCl3) δ 6.52 (s, 1H), 5.94 (ddt, J = 17.2, 10.5, 5.7 Hz, 1H), 5.33 (d, J = 17.2 Hz, 1H), 5.25 (d, J = 10.4 Hz, 1H), 4.66 (d, J = 5.7 Hz, 2H), 3.73 (t, J = 7.2 Hz, 2H), 3.65 (s, 3H), 3.17 (t, J = 7.7 Hz, 2H), 1.91 (p, J = 7.5 Hz, 2H); 13C NMR (101 MHz, CDCl3) δ 169.2, 157.3, 152.6, 131.9, 118.5, 96.4, 66.6, 50.8, 49.5, 31.6, 21.1; HRMS (ESI-TOF) m/z [M + H]+ calcd for C11H16NO4 226.1074, found 226.1068.</p><!><p>Obtained in 65% yield from 17d. 1H NMR (400 MHz, CDCl3) δ 5.89 (ddt, J = 16.2, 10.7, 5.4 Hz, 1H), 5.28 (dd, J = 17.2, 1.5 Hz, 1H), 5.19 (dd, J = 10.4, 1.1 Hz, 1H), 4.82 (s, 1H), 4.53 (d, J = 5.5 Hz, 2H), 3.72 (s, 3H), 3.43 (s, 2H), 3.16 (dd, J = 12.9, 6.5 Hz, 2H), 2.56 (t, J = 7.1 Hz, 2H), 1.68 – 1.57 (m, 2H), 1.56 – 1.43 (m, 2H); 13C NMR (101 MHz, CDCl3) δ 202.4, 167.6, 156.3, 132.9, 117.6, 65.4, 52.4, 49.0, 42.4, 40.5, 29.1, 20.2; HRMS (ESI-TOF) m/z [M + H]+ calcd for C12H20NO5 258.1336, found 258.1326.</p><!><p>A solution of the appropriate β-keto ester 18 (10.1 mmol) in 50 mL of methanesulfonic acid at 0 °C was treated with resorcinol (1.11 g, 10.1 mmol) and stirred for 3.5 h. The mixture was poured into ice cold water and the resulting yellow mixture was filtered. The filtrate was extracted with EtOAc and combined with the solids. The combined organic layer was concentrated and purified by flash chromatography over silica gel (0–20% MeOH/CHCl3) to afford the pure coumarin derivatives 19a–d.</p><!><p>Obtained in 36% yield from 18a. 1H NMR (400 MHz, DMSO-d6) δ 10.60 (s, 1H), 7.88 (t, J = 5.9 Hz, 1H), 7.64 (d, J = 8.7 Hz, 1H), 6.78 (d, J = 8.7 Hz, 1H), 6.73 (d, J = 2.3 Hz, 1H), 5.99 (s, 1H), 5.92 (ddt, J = 17.0, 10.6, 5.4 Hz, 1H), 5.29 (dd, J = 17.2, 1.6 Hz, 1H), 5.18 (d, J = 10.5 Hz, 1H), 4.52 (d, J = 5.3 Hz, 2H), 4.37 (d, J = 5.8 Hz, 2H); 13C NMR (101 MHz, DMSO-d6) δ 161.7, 160.8, 156.6, 155.4, 154.2, 134.0, 126.2, 117.6, 113.4, 110.3, 107.9, 102.8, 65.2, 41.0; HRMS (ESI-TOF) m/z [M + H]+ calcd for C13H14NO5 276.0867, found 276.0863.</p><!><p>Obtained in 88% yield from 18b. 1H NMR (400 MHz, DMSO-d6) δ 10.55 (s, 1H), 7.68 (d, J = 8.8 Hz, 1H), 7.40 (m, 1H), 6.80 (dd, J = 8.7, 2.3 Hz, 1H), 6.71 (d, J = 2.3 Hz, 1H), 6.07 (s, 1H), 5.99 – 5.78 (m, 1H), 5.24 (m, 1H), 5.15 (m, 1H), 4.45 (m, 2H), 3.29 (m, 2H), 2.87 (t, J = 6.7 Hz, 2H); 13C NMR (101 MHz, DMSO-d6) s 161.1, 160.3, 156.0, 155.2, 154.2, 133.8, 133.7, 126.3, 116.9, 113.0, 111.3, 110.5, 110.4, 102.5, 102.4, 64.3, 31.5, 23.4; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C16H16NO5 302.10285, found 302.10305.</p><!><p>Obtained in 88% yield yield from 18c. 1H NMR (400 MHz, DMSO-d6) δ 10.53 (s, 1H), 7.61 (d, J = 8.8 Hz, 1H), 7.33 (t, J = 5.5 Hz, 1H), 6.78 (d, J = 8.7, 1H), 6.69 (d, J = 2.4 Hz, 1H), 6.10 (s, 1H), 5.89 (ddt, J = 17.0, 10.6, 5.4 Hz, 1H), 5.25 (dd, J = 17.2, 1.6 Hz, 1H), 5.15 (d, J = 10.4 Hz, 1H), 4.45 (d, J = 5.3 Hz, 2H), 3.07 (q, J = 6.6 Hz, 2H), 2.72 (t, J = 7.6 Hz, 2H), 1.96 – 1.63 (m, 2H); 13C NMR (101 MHz, DMSO- d6) δ 161.5, 160.8, 157.0, 156.4, 155.6, 134.3, 126.7, 117.3, 113.3, 111.6, 109.9, 102.9, 64.6, 40.2, 28.7, 28.6; HRMS (ESI-TOF) m/z [M + H]+ calcd for C16H18NO5 304.1180, found 304.1172.</p><!><p>Obtained in 84% yield from 18d. 1H NMR (400 MHz, DMSO- d6) δ 10.50 (s, 1H), 7.60 (d, J = 8.8 Hz, 1H), 7.21 (t, J = 5.7 Hz, 1H), 6.76 (d, J = 8.7 Hz, 1H), 6.67 (d, J = 2.4 Hz, 1H), 6.05 (s, 1H), 5.86 (ddt, J = 17.2, 10.5, 5.3 Hz, 1H), 5.22 (dd, J = 17.2, 1.7 Hz, 1H), 5.11 (dd, J = 10.4, 1.6 Hz, 1H), 4.42 (d, J = 5.3 Hz, 2H), 3.00 (d, J = 6.1 Hz, 2H), 2.69 (t, J = 7.4 Hz, 2H), 1.62 – 1.51 (m, 2H), 1.51 – 1.42 (m, 2H); 13C NMR (101 MHz, DMSO- d6) δ 161.5, 160.9, 157.5, 156.4, 155.5, 134.3, 126.8, 117.2, 113.3, 111.6, 109.7, 102.8, 64.5, 40.2, 31.0, 29.5, 25.8; HRMS (ESI-TOF) m/z [M + H]+ calcd for C17H20NO5 318.1336, found 318.1339.</p><!><p>The appropriate coumarin derivative 19 (0.73 mmol) in 9 mL of AcOH was treated with HMTA (255 mg, 1.82 mmol) and stirred for 18 h at 95 °C. The reaction mixture was concentrated and the resulting slurry was dissolved in 12 mL of a 1:1 1M aq. HCl:EtOAc solution and stirred at 60 °C for 2 h. The organic layer was separated and the aqueous layer was extracted with EtOAc. The combined organic layers were washed with water, dried with MgSO4, and concentrated. Purification by silica gel flash column chromatography (EtOAc/hexane) afforded the desired bicyclic formyl derivatives 20a–d.</p><!><p>The appropriate coumarin derivative 19 (0.73 mmol) in 3 mL of TFA was treated with HMTA (255 mg, 1.82 mmol) and stirred for 18 h at 75 °C. The reaction mixture was concentrated and the resulting slurry was dissolved in 12 mL of a 1:1 1M aq. HCl:EtOAc solution and stirred at 60 °C for 2 h. The organic layer was separated and the aqueous layer was extracted with EtOAc. The combined organic layers were washed with water, dried with MgSO4, and concentrated. Purification by silica gel flash column chromatography (EtOAc/Hexane) afforded the desired bicyclic and tricyclic formyl derivatives.</p><!><p>The appropriate coumarin derivative 19 (0.47 mmol) in 15 mL of MeCN was treated with pyridine (18.5 mg, 0.23 mmol) and acetic anhydride (239 mg, 2.35 mmol). After stirring for 6 hours at rt, the reaction was diluted with brine and extracted with EtOAc. The organic layer was dried with MgSO4 and concentrated. The resulting crude product was dissolved in 2 mL of TFA was treated with HMTA (164 mg, 1.17 mmol) and stirred for 18 h at 95 °C. The reaction mixture was concentrated and the resulting slurry was dissolved in 12 mL of a 1:1 1M aq. HCl:EtOAc solution and stirred at 60 °C for 2 h. The organic layer was separated and the aqueous layer was extracted with EtOAc. The combined organic layers were washed with water, dried with MgSO4, and concentrated. Purification by silica gel flash column chromatography (EtOAc/Hexane) afforded the desired bicyclic and tricyclic formyl derivatives.</p><!><p>Obtained in 4% yield (Methods A, B, and C) from 19a. 1H NMR (400 MHz, CDCl3) δ 12.24 (s, 1H), 10.60 (s, 1H), 7.73 (d, J = 9.0 Hz, 1H), 6.91 (d, J = 9.0 Hz, 1H), 6.32 (s, 1H), 5.94 (ddt, J = 16.5, 11.1, 5.8 Hz, 1H), 5.34 (d, J = 17.2 Hz, 1H), 5.27 (d, J = 10.3 Hz, 1H), 5.19 (t, J = 5.6 Hz, 1H), 4.64 (dt, J = 5.7, 1.4 Hz, 2H), 4.54 (d, J = 6.3 Hz, 2H); 13C NMR (101 MHz, CDCl3) δ 193.3, 165.4, 159.1, 156.3, 156.1, 152.1, 132.2, 131.8, 118.5, 114.7, 109.74, 109.71, 108.8, 66.4, 41.3; HRMS (ESI-TOF) m/z [M + H]+ calcd for C15H14NO6 304.0816, found 304.0820.</p><!><p>Obtained in 10% yield (Method A) from 19b. 1H NMR (400 MHz, CDCl3) δ 12.24 (s, 1H), 10.60 (s, 1H), 7.92 (d, J = 9.1 Hz, 1H), 6.93 (d, J = 9.0 Hz, 1H), 6.19 (s, 1H), 5.90 (m, 1H), 5.39 – 5.15 (m, 2H), 5.03 (bs, 1H), 4.58 (m, 2H), 3.49 (m, 2H), 2.99 (t, J = 7.2 Hz, 2H); 13C NMR (101 MHz, CDCl3) δ 193.5, 193.4, 165.5, 159.2, 156.6, 156.5, 153.4, 133.1, 132.6, 118.2, 114.8, 112.2, 112.1, 111.1, 109.0, 66.0, 40.1, 32.8; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C16H16NO6 318.09777, found 318.09746.</p><!><p>Obtained in 13% yield (Method A) from 19c. 1H NMR (400 MHz, CDCl3) δ 12.20 (s, 1H), 10.58 (s, 1H), 7.72 (d, J = 9.0 Hz, 1H), 6.88 (d, J = 9.0 Hz, 1H), 6.19 (s, 1H), 5.90 (ddt, J = 16.8, 11.1, 5.6 Hz, 1H), 5.29 (dd, J = 17.2, 1.5 Hz, 1H), 5.20 (dd, J = 10.4, 1.2 Hz, 1H), 4.98 (t, J = 5.2 Hz, 1H), 4.56 (d, J = 5.4 Hz, 2H), 3.33 (q, J = 6.5 Hz, 2H), 2.98 – 2.59 (m, 2H), 1.90 (tt, J = 13.7, 6.9 Hz, 2H); 13C NMR (101 MHz, CDCl3) δ 193.4, 165.2, 159.3, 156.4, 156.3, 155.6, 132.7, 132.5, 117.9, 114.4, 111.0, 110.9, 108.8, 65.7, 40.4, 29.1, 28.5; HRMS (ESI-TOF) m/z [M + H]+ calcd for C17H18NO6 332.1129, found 332.1128.</p><!><p>Obtained in 15% yield (Method A) from 19d. 1H NMR (400 MHz, CDCl3) δ 12.22 (s, 1H), 10.60 (s, 1H), 7.74 (d, J = 9.0 Hz, 1H), 6.89 (d, J = 9.0 Hz, 1H), 6.17 (s, 1H), 5.90 (ddt, J = 16.1, 10.8, 5.7 Hz, 1H), 5.29 (dd, J = 17.2, 1.6 Hz, 1H), 5.20 (dd, J = 10.4, 1.3 Hz, 1H), 4.80 (s, 1H), 4.55 (d, J = 5.6 Hz, 2H) 3.26 (q, J = 6.4 Hz, 2H), 2.93 – 2.56 (m, 2H), 1.78 – 1.56 (m, 4H); 13C NMR (101 MHz, CDCl3) δ 139.4, 165.2, 159.4, 156.38, 156.37, 156.1, 132.8, 132.7, 117.8, 114.4, 111.1, 111.0, 109.8, 65.6, 40.3, 31.5, 29.9, 25.2; HRMS (ESI-TOF) m/z [M + H]+ calcd for C18H20NO6 346.1285, found 346.1288.</p><!><p>Obtained in 22% (Method B) and 41% (Method C) yield from 19b. 1H NMR (400 MHz, CDCl3) δ 12.15 (s, 1H), 10.61 (s, 1H), 7.68 (d, J = 8.4 Hz, 1H), 6.92 (d, J = 9.0 Hz, 1H), 5.94 (m, 1H), 5.33 (m, 1H), 5.24 (m, 1H), 4.64 (d, J = 5.7 Hz, 2H), 4.47 (m, 2H), 3.81 (t, J = 5.8 Hz, 2H), 2.86 (m, 2H); 13C NMR (101 MHz, CDCl3) δ 193.3, 164.9, 158.4, 155.2, 154.7, 146.4, 132.7, 131.8, 118.3, 117.2, 114.8, 111.2, 108.7, 66.7, 41.9, 39.2, 24.9; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C17H16NO6 330.09721, found 330.09624.</p><!><p>Obtained 18% (Method B) and 17% (Method C) yield from 19c. 1H NMR (400 MHz, CDCl3) δ 12.17 (s, 1H), 10.61 (s, 1H), 7.79 (d, J = 9.2 Hz, 1H), 6.90 (d, J = 8.9 Hz, 1H), 5.87 (ddt, J = 16.3, 10.8, 5.3 Hz, 1H), 5.28 (dd, J = 17.2, 1.5 Hz, 1H), 5.16 (d, J = 10.8 Hz, 1H), 4.65 (s, 2H), 4.55 (d, J = 5.1 Hz, 2H), 3.98 – 3.57 (m, 2H), 3.08 – 2.96 (m, 2H), 2.13 – 2.00 (m, 2H); 13C NMR (101 MHz, CDCl3) δ 193.4, 164.8, 159.3, 155.7, 155.0, 152.2, 132.54, 132.50, 122.0, 117.3, 114.4, 111.9, 108.6, 66.3, 47.8, 42.9, 27.6, 24.6; HRMS (ESI-TOF) m/z [M + H]+ calcd for C18H18NO6 344.1129, found 344.1137.</p><!><p>Obtained in 3% (Method B) and 9% (Method C) yield from 19d. 1H NMR (400 MHz, CDCl3) δ 12.18 (s, 1H), 10.61 (s, 1H), 7.78 (d, J = 8.6 Hz, 1H), 6.91 (d, J = 8.9 Hz, 1H), 5.95 (ddt, J = 16.9, 10.8, 5.6 Hz, 1H), 5.32 (d, J = 17.1 Hz, 1H), 5.21 (d, J = 10.3 Hz, 1H), 4.66 (s, 2H), 4.64 (d, J = 4.7 Hz, 2H), 3.62 – 3.49 (m, 2H), 3.07 – 2.96 (m, 2H), 1.92 – 1.80 (m, 2H), 1.80 – 1.69 (m, 2H); 13C NMR (101 MHz, CDCl3) δ 139.3, 164.8, 159.1, 155.8, 155.5, 152.0, 132.9, 132.6, 119.5, 117.6, 114.5, 111.3, 108.7, 66.5, 46.2, 44.4, 26.1, 25.7, 25.1; HRMS (ESI-TOF) m/z [M + H]+ calcd for C19H20NO6 358.1285, found 358.1290.</p><!><p>Compound 21b (24 mg, 73 µmol) in 2 mL of MeOH at 0 °C was treated with sodium borohydride (3.0 mg, 73 µmol) and stirred for 40 min. The reaction was quenched with 1M aq. HCl and extracted with EtOAc. The organic layers were dried over Na2SO4 and concentrated. Purification by flash column chromatography over silica gel (40% –50% EtOAc/hexane) afforded 22 as white foam (16 mg, 66%); 1H NMR (400 MHz, CDCl3) δ 9.65 (bs, 1H), 7.35 (d, J = 8.4 Hz, 1H), 6.85 (d, J = 8.7 Hz, 1H), 5.94 (m, J = 11.1, 5.6 Hz, 1H), 5.33 (m, 3H), 5.24 (m, 1H), 4.64 (m, 2H), 4.40 (s, 2H), 3.77 (t, J = 5.7 Hz, 2H), 2.84 (m, 2H); 13C NMR (101 MHz, CDCl3) δ 180.6, 160.4, 149.8, 147.8, 132.7, 123.5, 118.2, 114.7, 111.8, 111.1, 66.7, 59.0, 41.8, 39.3, 24.9; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C17H18NO6 332.11342, found 332.11473.</p><!><p>Compound 22 (17 mg, 51 µmol) in 2 mL of DCM at rt was treated with thionyl chloride (19 µL, 257 µmol) and stirred for 5.5 h. The reaction was diluted with DCM and washed with sat. aq. NH4Cl, dried over Na2SO4, and concentrated under reduced pressure. The resulting white solid 23 was suffiently pure by NMR and HPLC analysis for further use (12 mg, 67%). 1H NMR (400 MHz, CDCl3) δ 9.42 (m, 0.5H), 7.69 (m, 0.5H), 7.38 (m, 1H), 6.89 (m, 1H), 5.95 (m, 1H), 5.30 (m, 3H), 4.91 (s, 1H), 4.65 (m, 2H), 4.44 (d, J = 17.7 Hz, 2H), 3.79 (m, 2H), 2.85 (m, 2H); 13C NMR (101 MHz, CDCl3) δ 160.3, 157.9, 151.5, 149.9, 132.7, 124.7, 123.6, 118.3, 114.5, 113.1, 112.7, 112.3, 111.9, 111.2, 66.8, 58.9, 42.0, 39.4, 34.3, 29.8, 24.8; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C17H16ClNO5 350.07898, found 346.12850 (observed mass corresponds to the 7-methoxymethyl derivative, resulting from displacement of the chloride with methanol during LCMS).</p><!><p>A solution of 21b in (150 mg, 455 µmol) in 4 mL of benzene was treated with 1,3-propanediol (99.0 µL, 1.40 mmol) and p-toluenesulfonic acid monohydrate (4.3 mg, 23 µmol) and stirred for 2 h. The reaction was quenched with 2 drops of NEt3, diluted with EtOAc, and washed with brine. The organic layer was dried over Na2SO4 and concentrated. Purification by flash column chromatography over silica gel (30%–50% EtOAc/hexanes eluent) afforded 24 as a yellow solid (157 mg, 89%). 1H NMR (400 MHz, CDCl3) δ 8.82 (s, 1H), 7.36 (d, J = 8.2 Hz, 1H), 6.79 (d, J = 8.8 Hz, 1H), 6.28 (s, 1H), 5.91 (m, 1H), 5.30 (m, 1H), 5.20 (m, 1H), 4.61 (d, J = 5.6 Hz, 2H), 4.39 (s, 2H), 4.28 (dd, J = 11.6, J = 4.6 Hz, 2H), 4.09 (m, 2H), 3.74 (t, J = 5.8 Hz, 2H), 2.79 (m, 2H), 2.26 (m, 1H), 1.53 (m, 1H); 13C NMR (101 MHz, CDCl3) δ 159.5, 159.3, 155.2, 150.5, 146.6, 132.8, 125.3, 118.0, 116.3, 114.5, 111.8, 109.9, 98.1, 67.9, 66.5, 41.8, 39.3, 25.8, 24.7; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C20H22NO7 388.13908, found 388.13810.</p><!><p>Compound 21b (39.0 mg, 118 µmol) and 1,3-propanedithiol (13.0 µL, 130 µmol) in 2.5 mL of DCM at rt was treated with BF3·OEt2 (6.0 µL, 47 µmol) and stirred for 17 h. The reaction was quenched with sat. aq. NaHCO3 and the aqueous layer was extracted with EtOAc. The combined organic layers were dried over Na2SO4 and concentrated under reduced pressure. Purification by flash column chromatography over silica gel (25–50% EtOAc/hexane) afforded 25 as a white foam (39 mg, 79%). 1H NMR (400 MHz, CDCl3) δ 7.55 (s, 1H), 7.43 (d, J = 8.4 Hz, 1H), 6.91 (d, J = 8.9 Hz, 1H), 6.26 (s, 1H), 5.95 (m, 1H), 5.33 (m, 1H), 5.24 (m, 1H), 4.64 (m, 2H), 4.47 (m, 2H), 3.78 (t, J = 5.8 Hz, 2H), 3.17 (m, 2H), 2.91 (m, 4H), 2.24 (m, 1H), 1.94 (m, 1H); 13C NMR (101 MHz, CDCl3) δ 159.5, 155.3, 149.8, 146.9, 132.8, 124.8, 118.1, 116.7, 114.9, 112.5, 111.1, 110.6, 77.5, 77.2, 76.8, 66.6, 42.0, 39.2, 37.4, 31.3, 24.9, 24.7, 23.0, 14.3, 14.3. HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C20H22NO5S2 420.09399, found 420.09248.</p><!><p>A solution of 24 (20 mg, 52 µmol) in 1 mL DMF was treated with K2CO3 (36 mg, 258 µmol) followed by iodomethane (10 µL, 155 µmol). After stirring at rt for 18 h, the mixture was diluted with sat. aq. NH4Cl, extracted with DCM, and concentrated to dryness. The residue was taken up in 500 µL of dioxane, treated with 2 mL of 4M aq. HCl, and stirred at rt for 30 min. The mixture was diluted with water and extracted with DCM. The combined organic layers were dried over Na2SO4 and concentrated under reduced pressure. Purification by flash column chromatography over silica gel (0–10% MeOH/CHCl3) afforded 26 as a white powder (12 mg, 67%). 1H NMR (400 MHz, CDCl3) δ 10.68 (s, 1H), 7.71 (d, J = 8.7 Hz, 1H), 6.98 (d, J = 9.0 Hz, 1H), 5.95 (m, 1H), 5.33 (m, 1H), 5.24 (m, 1H), 4.65 (m, 2H), 4.48 (s, 2H), 4.01 (s, 3H), 3.82 (t, J = 5.8 Hz, 2H), 2.87 (m, 2H); 13C NMR (101 MHz, CDCl3) δ 187.2, 162.6, 158.6, 157.2, 155.3, 145.7, 132.8, 132.7, 132.7, 129.7, 118.3, 118.2, 112.9, 112.7, 108.2, 66.7, 56.8, 42.0, 39.3, 29.9, 24.9; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C18H18NO6 344.11341, found 344.11432.</p><!><p>A solution of 24 (20 mg, 52 µmol) in 1 mL DMF was treated with K2CO3 (36 mg, 260 µmol) followed by benzyl bromide (9.0 µL, 78 µmol). After stirring at rt for 18 h, the mixture was diluted with sat. aq. NH4Cl, extracted with DCM, and concentrated to dryness. The residue was taken up in 500 µL of dioxane, treated with 2 mL of 4N aq. HCl, and stirred at rt 30 min. The mixture was diluted with water and extracted with DCM. The combined organic layers were dried over Na2SO4 and concentrated under reduced pressure. Purification by flash column chromatography over silica gel (0–10% MeOH/CHCl3) afforded 27 as a white powder (18 mg, 72%). 1H NMR (400 MHz, CDCl3) δ 10.72 (d, J = 5.4 Hz, 1H), 7.67 (d, J = 8.6 Hz, 1H), 7.58 – 7.31 (m, 5H), 7.01 (d, J = 9.0 Hz, 1H), 5.95 (m, 1H), 5.35 (m, 0.5H), 5.30 (m, 2.5H), 5.24 (m, 1H), 4.65 (m, 2H), 4.47 (s, 2H), 3.81 (t, J = 5.8 Hz, 2H), 2.87 (m, 2H); 13C NMR (101 MHz, CDCl3) δ 187.1, 161.7, 158.7, 154.0, 145.6, 135.4, 132.7, 129.5, 128.9, 128.5, 127.0, 118.3, 113.2, 113.1, 109.6, 71.2, 66.7, 51.3, 42.0, 39.2, 29.8, 24.8; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C24H22NO6 420.14471, found 420.14529.</p><!><p>A solution of 24 (70 mg, 180 µmol) in 4 mL of DCM at rt was treated with phenylsilane (67 mg, 540 µmol) and tetrakis(triphenylphosphine)palladium(0) (10 mg, 9.0 µmol), and stirred at rt for 25 min. The reaction was concentrated and the residue purified by flash chromatography over silica gel (0%–10% MeOH/CHCl3) to afford 28 as a yellow solid (54 mg, 98%). 1H NMR (400 MHz, CDCl3) δ 7.35 (d, J = 8.8 Hz, 1H), 6.78 (d, J = 8.8 Hz, 1H), 6.28 (s, 1H), 4.24 (m, 2H), 4.06 (m, 2H), 3.75 (m, 2H), 3.11 (t, J = 5.8 Hz, 2H), 2.70 (m, 2H), 2.36 – 2.11 (m, 1H), 1.92 (bs, 1H), 1.50 (m, 1H); 13C NMR (101 MHz, CDCl3) δ 160.2, 159.0, 150.6, 146.8, 135.0, 125.1, 119.0, 114.3, 112.5, 109.9, 98.3, 68.0, 43.4, 42.0, 25.9, 25.3; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C16H18NO5 304.11795, found 304.11782.</p><!><p>A solution of 28 (20 mg, 66 µmol) in 1 mL of DCM was treated with pyridine (11 µL, 130 µmol) and acetyl chloride (7.0 µL, 99 µmol), then stirred at rt for 20 min. After concentration under reduced pressure, the residue was taken up in 500 µL of dioxane, treated with 2 mL of 4M aq. HCl, and stirred at rt 30 min. The mixture was diluted with water and extracted with DCM. The combined organic layers were dried over Na2SO4 and concentrated under reduced pressure. Purification by flash column chromatography over silica gel (0–10% MeOH/CHCl3) afforded 29 as a white powder (17 mg, 90%). 1H NMR (400 MHz, DMSO-d6) δ 11.84 (s, 1H), 10.46 (s, 1H), 7.90 (m (rotomer), 1H), 7.00 (d, J = 8.9 Hz, 1H), 4.32 (m, 2H), 3.73 (t, J = 5.7 Hz, 2H), 2.96 (m, 2H), 2.83 (m, 1H), 2.10 (m (rotomer), 3H); 13C NMR (101 MHz, DMSO-d6) δ 191.1, 191.0, 168.9, 163.2, 163.1, 158.2, 153.5, 147.0, 146.8, 132.2, 116.7, 116.5, 113.9, 111.1, 109.1, 104.6, 43.2, 41.4, 36.3, 25.0, 24.3, 21.8, 21.3; HRMS (ESI-TOF) (m/z) [M+H]+ calcd C15H14NO5 288.08720, found 288.08654.</p><!><p>A solution of 28 (50.0 mg, 165 µM) in 2 mL of 1:1 dioxane:THF was treated with 37% aq. formaldehyde (27.0 µL, 330 µM), 10% Pd/C (40 mg), placed under H2 atmosphere, and stirred at rt for 3 h. The reaction was filtered through celite with MeOH rinsing and concentrated to afford the crude methyl amine. The residue was taken up in 500 µL of dioxane, treated with 2 mL of 4M aq. HCl, and stirred at rt 30 min. The mixture was diluted with water and extracted with DCM. The combined organic layers were dried over Na2SO4 and concentrated under reduced pressure. Purification by flash column chromatography over silica gel (0–10% MeOH/CHCl3) afforded 30 as a white powder (35 mg, 67%). 1H NMR (400 MHz, CDCl3) δ 12.15 (s, 1H), 10.61 (s, 1H), 7.66 (d, J = 9.0 Hz, 1H), 6.92 (d, J = 9.0 Hz, 1H), 3.59 (s, 2H), 3.03 – 2.97 (m, 2H), 2.97 – 2.90 (m, 2H), 2.64 (s, 3H); 13C NMR (101 MHz, CDCl3) δ 193.2, 164.8, 158.4, 154.6, 145.5, 131.7, 117.3, 114.6, 111.0, 108.6, 51.6, 50.2, 45.0, 25.3; HRMS (ESI-TOF) m/z [M + H]+ calcd for C14H14NO4 260.0917, found 260.0915.</p><!><p>A solution of 28 (20 mg, 66 µmol) in 1.5 mL DMF at rt was treated with NEt3 (10 mg, 99 µmol) and benzylbromide (12 mg, 73 µmol). After stirring 5 hours, the reaction was concentrated and treated with 4 mL of 4M aq. HCl and stirred for 1 hour. The reaction was adjusted to pH 7 with 10% aq. Na2CO3, extracted with DCM, dried over MgSO4, and concentrated. Purification by silica gel flash column chromatography (MeOH/CHCl3) 31 as a white solid (15.4 mg, 70%). 1H NMR (400 MHz, CDCl3) δ 12.15 (s, 1H), 10.61 (s, 1H), 7.65 (d, J = 9.0 Hz, 1H), 7.42 – 7.30 (m, 5H), 6.90 (d, J = 9.0 Hz, 1H), 3.87 (s, 2H), 3.59 (s, 2H), 2.94 (s, 4H); 13C NMR (101 MHz, CDCl3) δ 193.3, 164.5, 158.8, 154.5, 146.2, 137.1, 131.7, 129.2, 128.5, 127.6, 119.0, 114.3, 111.5, 108.5, 62.3, 50.1, 48.0, 26.0; HRMS (ESI-TOF) m/z [M + H]+ calcd for C20H18NO4 336.1230, found 336.1224.</p><!><p>A solution of 28 (20 mg, 66 µmol) in 1.5 mL DMF at rt was treated with NEt3 (10 mg, 99 µmol) and 4-fluorobenzyl bromide (12 mg, 73 µmol). After stirring 5 h, the reaction was concentrated and treated with 4 mL of 4M aq. HCl and stirred for 1 h. The reaction was adjusted to pH 7 with 10% aq. Na2CO3, extracted with DCM, dried over MgSO4, and concentrated. Purification by silica gel flash column chromatography (MeOH/CHCl3) gave 32 as a pale yellow solid (12 mg, 49%). 1H NMR (400 MHz, CDCl3) δ 12.15 (s, 1H), 10.62 (s, 1H), 7.66 (d, J = 9.0 Hz, 1H), 7.36 (m, 2H), 7.06 (m, 2H), 6.86 (m, 1H), 3.70 (bs, 2H), 3.51 (m, 2H), 2.86 (m, 4H); 13C NMR (101 MHz, CDCl3) δ 193.4, 164.7, 158.8, 156.3, 154.7, 146.2, 143.8, 131.9, 131.0, 127.9, 125.4, 115.7, 115.5, 114.6, 114.0, 111.5, 108.7, 68.7, 68.0, 61.5, 50.4, 48.1, 31.2, 29.9, 26.0; HRMS (ESI-TOF) m/z [M + H]+ calcd for C20H17FNO4 354.11416, found 354.11438.</p><!><p>A solution of 28 (20 mg, 67 µmol) in 1.5 mL DMF at rt was treated with NEt3 (10 mg, 99 µmol) and 3-bromo-2-methylpropene (9.9 mg, 74 µmol). After stirring 5 h, the reaction was concentrated and treated with 4 mL of 4M aq. HCl and stirred for 1 h. The reaction was adjusted to pH 7 with 10% aq. Na2CO3, extracted with DCM, dried over MgSO4, and concentrated. Purification by silica gel flash column chromatography (MeOH/CHCl3) gave 33 as a yellow solid (15 mg, 75%). 1H NMR (400 MHz, CDCl3) δ 12.14 (s, 1H), 10.61 (s, 1H), 7.66 (d, J = 9.0 Hz, 1H), 6.90 (d, J = 9.0 Hz, 1H), 5.01 (s, 2H), 3.52 (s, 2H), 3.22 (s, 2H), 2.94 (s, 2H), 2.87 (s, 2H), 1.81 (s, 3H); 13C NMR (101 MHz, CDCl3) δ 193.2, 164.6, 158.6, 154.6, 146.0, 140.5, 131.7, 115.3, 114.5, 111.3, 108.5, 105.0, 64.4, 50.2, 48.0, 25.6, 20.8; HRMS (ESI-TOF) m/z [M + H]+ calcd for C17H18NO4 300.1230, found 300.1223.</p><!><p>A solution of 28 (20 mg, 66 µmol) in 1 mL of DCM was treated with NEt3 (28 µL, 198 µmol) followed by 1,3-di-Boc-2-(trifluoromethylsulfonyl)guanidine (58 mg, 146 µmol) and stirred at rt for 18 h. The reaction was diluted with sat. aq. NH4Cl and extracted with DCM. The combined organic layers were dried over Na2SO4 and concentrated under reduced pressure. Purification by flash column chromatography over silica gel (40% EtOAc/hexanes) gave the guanidinylated intermediate as a glassy solid. The material was then treated with 2 mL of a 1:1 TFA:DCM solution and stirred at rt for 4 h. The reaction was concentrated to remove TFA and the resulting solid was washed with 3 portions of DCM. Drying of the solid under vacuum afforded 34 (12 mg, 63%), which was pure by NMR. 1H NMR (400 MHz, DMSO-d6) δ 11.93 (s, 1H), 10.46 (s, 1H), 7.94 (d, J = 9.0 Hz, 1H), 7.64 (m, 3H), 7.02 (d, J = 9.0 Hz, 1H), 4.31 (s, 2H), 3.71 (t, J = 5.7 Hz, 2H), 2.99 (m, 2H); 13C NMR (101 MHz, DMSO-d6) δ 190.8, 163.5, 158.1, 156.3, 153.5, 146.7, 132.3, 115.3, 114.1, 110.9, 109.3, 104.7, 43.2, 41.0, 24.2; HRMS (ESI-TOF) m/z [M + H]+ calcd for C14H14N3O4 288.09843, found 288.09881.</p><!><p>Compound 21b (301 mg, 914 µmol) in 5 mL of DCM at 0 °C was treated with DIEA (790 µL, 4.57 mmol) and chloromethyl methyl ether (347 µL, 4.57 mmol). The reaction was stirred for 30 h, quenched with sat. aq. NH4Cl, and the organic layer washed with sat. aq. NH4Cl. The organic layer was dried over Na2SO4 and concentrated. Purification by flash column chromatography over silica gel (35%–70% EtOAc/hexanes) afforded 35 as a white solid (226 mg, 67%). 1H NMR (400 MHz, CDCl3) δ 10.68 (s, 1H), 7.67 (d, J = 8.9 Hz, 1H), 7.20 (d, J = 9.0 Hz, 1H), 5.95 (m, 1H), 5.35 (m, 2.5H), 5.30 (m, 0.5H), 5.24 (m, 1H), 4.64 (d, J = 5.7 Hz, 2H), 4.48 (s, 2H), 3.81 (t, J = 5.8 Hz, 2H), 3.53 (s, 3H), 2.87 (m, 2H); 13C NMR (101 MHz, CDCl3) δ 186.8, 160.2, 158.4, 155.0, 153.4, 145.7, 132.6, 129.4, 118.2, 117.9, 113.5, 113.3, 111.4, 94.9, 66.4, 56.8, 41.7, 39.1, 24.7; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C19H20NO7 374.12343, found 374.12310.</p><!><p>Compound 35 (50.0 mg, 134 µmol) in 2 mL of THF at −78 °C under Ar was treated with 3M MeMgBr in Et2O (134 µL, 402 µmol). After 3h at −78 °C, the reaction was carefully quenched then diluted with sat. aq. NH4Cl, warmed to rt, and partitioned with EtOAc. The organic layer was dried over Na2SO4 and concentrated under reduced pressure to give the crude alcohol as an oil.</p><p>The above alcohol was dissolved in 3 mL of DCM and treated with Dess-Martin periodinane (123 mg, 291 µmol) and stirred at rt for 3 h. the reaction was quenched with 10% aq. Na2S2O3 and washed with brine. The organic layer was dried over Na2SO4, concentrated, and the residue purified by flash column chromatography over silica gel (35–70% EtOAc/Hexane) to give the intermediate ketone as a gum (34 mg, 66%, 2 steps). 1H NMR (400 MHz, CDCl3) δ 7.50 (d, J = 8.8 Hz, 1H), 7.13 (d, J = 8.9 Hz, 1H), 5.95 (m, 1H), 5.34 (m, 1H), 5.30 (m, 1H), 5.25 (m, 2.5H), 5.21 (m, 0.5H), 4.64 (d, J = 5.7 Hz, 2H), 4.45 (m, 2H), 3.80 (t, J = 5.8 Hz, 2H), 3.48 (s, 3H), 2.86 (m, 2H), 2.61 (s, 3H); 13C NMR (101 MHz, CDCl3) δ 199.1, 158.9, 155.6, 149.3, 145.8, 133.5, 132.7, 125.0, 120.6, 118.2, 114.0, 111.2, 108.6, 94.8, 66.7, 56.7, 42.0, 39.3, 32.7, 29.8, 24.8; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C20H22NO7 388.13908, found 388.13945.</p><p>The ketone above (9.0 mg, 23 µmol) in 1.5 mL of 33% TFA/DCM solution was stirred for 1.5 h at rt. The reaction was concentrated under reduced pressure and the resulting residue was purified by flash column chromatography over silica gel (30% EtOAc/Hexane) to afford 36 as a white foam (6.0 mg, 75%). 1H NMR (400 MHz, CDCl3) δ 13.54 (s, 1H), 7.63 (d, J = 8.8 Hz, 1H), 6.95 (d, J = 9.0 Hz, 1H), 5.96 (m, 1H), 5.33 (m, 2H), 5.24 (ddd, J = 10.4, 2.5, 1.2 Hz, 1H), 4.65 (dt, J = 5.7, 1.3 Hz, 2H), 4.47 (m, 2H), 3.81 (t, J = 5.8 Hz, 2H), 2.98 (s, 3H), 2.87 (m, 2H); 13C NMR (101 MHz, CDCl3) δ 204.4, 166.3, 158.6, 155.3, 153.8, 132.8, 130.2, 118.2, 116.6, 115.7, 111.3, 109.5, 66.7, 41.8, 39.3, 34.2, 25.1; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C18H18NO6 344.11286, found 344.11116.</p><!><p>Compound 35 (40.0 mg, 107 µmol) in 2 mL of DCM at rt was treated with triethylphosphonoacetate (48.0 mg, 139 µmol) and stirred for 20 h. the reaction was concentrated under reduced pressure. Purification by flash column chromatography over silica gel (20%–40% EtOAc/hexanes) afforded the intermediate ethyl enoate as a white solid (46 mg, 96%). 1H NMR (400 MHz, CDCl3) δ 8.11 (d, J = 16.4 Hz, 1H), 7.46 (m, 1H), 7.16 (d, J = 9.0 Hz, 1H), 7.05 (d, J = 16.4 Hz, 1H), 5.96 (m, 1H), 5.34 (m, 2.5H), 5.30 (m, 0.5H), 5.23 (m, 1H), 4.64 (dt, J = 5.6, 1.2 Hz, 2H), 4.46 (s, 2H), 4.28 (q, J = 7.1 Hz, 2H), 3.79 (t, J = 5.8 Hz, 2H), 3.50 (s, 3H), 2.85 (m, 2H), 1.35 (t, J = 7.1 Hz, 3H); 13C NMR (101 MHz, CDCl3) δ 167.8, 158.5, 151.8, 132.8, 132.7, 132.2, 132.1, 132.1, 132.0, 128.7, 128.5, 125.2, 124.4, 118.1, 113.8, 112.4, 110.9, 94.8, 66.6, 60.7, 56.8, 56.8, 42.0, 39.3, 24.9, 14.5; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C23H26NO8 444.16529, found 444.16576.</p><p>The above ethyl enoate (20 mg, 45 µmol) in 2 mL MeOH:CHCl3 (3:1) at rt was treated with 2 mL of 4N aq. HCl and stirred for 18 h. The reaction was extracted with EtOAc, dried over Na2SO4, and concentrated under reduced pressure. Purification by flash column chromatography over silica gel (40%–70% EtOAc/hexanes) afforded 37 as a white solid (15 mg, 83%). 1H NMR (400 MHz, DMSO-d6) δ 11.57 (m, 1H), 8.01 (d, J = 16.3 Hz, 1H), 7.62 (d, J = 8.8 Hz, 1H), 6.99 (m, 2H), 5.96 (ddt, J = 17.2, 10.6, 5.3 Hz, 1H), 5.31 (m, 1H), 5.22 (m, 1H), 4.59 (dt, J = 5.3, 1.4 Hz, 2H), 4.34 – 4.12 (m, 4H), 3.68 (m, 2H), 2.87 (m, 2H), 1.27 (t, J = 7.1 Hz, 3H); 13C NMR (101 MHz, DMSO-d6) δ 167.0, 160.3, 158.7, 154.3, 151.7, 146.9, 133.3, 133.2, 126.7, 121.5, 117.4, 112.9, 111.3, 108.2, 65.6, 60.1, 41.3, 24.4, 14.3; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C21H22NO7 400.13908, found 400.13992.</p><!><p>LiCl (8.0 mg, 20 µmol) and diethyl(methylsulfonylmethyl) phosphonate (54.0 g, 236 µmol) in 2.5 mL of acetonitrile at rt was treated with DBU (24.0 µL, 157 µmol) and stirred for 10 min. Compound 35 (43.0 mg, 131 µmol) in 2 mL of acetonitrile was cannulated into the mixture and stirred for 2h. The reaction was quenched with sat. aq. NH4Cl, and extracted with EtOAc. Purification by flash column chromatography over silica gel (0–5% MeOH/CHCl3) afforded the intermediate vinyl sulfone as a white solid (41 mg, 79%). 1H NMR (400 MHz, CDCl3) δ 8.09 (d, J = 15.8 Hz, 1H), 7.68 (d, J = 15.8 Hz, 1H), 7.56 (d, J = 8.9 Hz, 1H), 7.20 (d, J = 9.0 Hz, 1H), 5.95 (m, 1H), 5.34 (m, 2.5H), 5.30 m, 0.5H), 5.23 (m, 1H), 4.64 (m, 2H), 4.46 (s, 2H), 3.80 (t, J = 5.8 Hz, 2H), 3.51 (s, 3H), 3.06 (s, 3H), 2.86 (m, 2H); 13C NMR (101 MHz, CDCl3) δ 158.8, 158.6, 155.3, 152.2, 146.1, 132.7, 131.7, 131.5, 126.8, 118.2, 113.8, 110.9, 109.9, 95.1, 66.6, 57.0, 43.3, 41.9, 39.3, 24.9; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C21H24NO8S 450.12172, found 450.12390.</p><p>The above vinyl sulfone (38 mg, 84 µmol) in 2.5 mL of acetonitrile:CHCl3 (2:1) was treated with 2.5 mL of 4N aq. HCl and stirred for 20 h at rt. the reaction was concentrated under reduced pressure. The resulting white solid was washed with DCM/Et2O and the resulting solid dried to afford pure 38 (32 mg, 92%). 1H NMR (400 MHz, DMSO-d6) δ 11.78 (s, 1H), 7.87 (d, J = 15.7 Hz, 1H), 7.68 (d, J = 8.8 Hz, 2H), 7.66 (d, J = 15.7 Hz, 2H), 7.01 (d, J = 8.9 Hz, 1H), 5.97 (m, 1H), 5.31 (m, 1H), 5.21 (m, 1H), 4.59 (dt, J = 5.3, 1.5 Hz, 2H), 4.29 (s, 2H), 3.84 (s, 3H), 3.15 (s, 3H), 2.88 (s, 2H); 13C NMR (101 MHz, DMSO-d6) δ 160.5, 158.6, 154.3, 151.8, 146.9, 133.3, 130.8, 130.1, 127.6, 117.4, 112.9, 111.3, 106.5, 65.6, 42.6, 41.3, 24.2; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C19H20NO7S 406.09550, found 406.09500.</p><!><p>Compound 35 (80.0 mg, 214 µmol) and 2-methyl-2-butene (272 µL, 2.57 mmol) in 3.5 mL of t-BuOH:H2O:CH3CN (3:3:1) at 0 °C was treated with a solution of sodium chlorite (145 mg, 1.29 mmol) and sodium monophosphate (265 mg, 1.93 mmol) in water, dropwise. After 30 min the reaction was quenched with 5% aq. Na2S2O3. The pH of the solution was adjusted to 6 and extracted with EtOAc. The organic layer was dried over Na2SO4 and concentrated. Purification by flash column chromatography over silica gel (70%–100% EtOAc/hexanes) afforded 39 as a thick oil (64 mg, 77%). 1H NMR (400 MHz, CDCl3) δ 7.57 (d, J = 8.2 Hz, 1H), 7.19 (d, J = 9.0 Hz, 1H), 6.07 (bs,12H), 5.96 (m, 1H), 5.33 (m, 2.5H), 5.24 (m, 1H), 4.65 (m, 2H), 4.48 (s, 2H), 3.82 (t, J = 5.7 Hz, 2H), 3.52 (s, 3H), 2.90 (m, 2H); 13C NMR (101 MHz, CDCl3) δ 166.5, 163.5, 162.5, 159.9, 156.9, 156.4, 155.4, 150.3, 149.7, 146.7, 132.7, 125.5, 118.3, 113.7, 113.5, 111.5, 111.1, 99.9, 95.0, 94.7, 91.8, 66.8, 56.9, 56.7, 41.9, 41.7, 39.3, 24.8; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C19H20NO8 390.11835, found 390.11750.</p><!><p>Compound 39 (32 mg, 82 µmol) in 1 mL of MeOH was treated with 2 mL of 4N aq. HCl and stirred for 20 h at rt. The reaction was concentrated under reduced pressure. Purification by semi-preparative RP-HPLC (C18 column, 0%–70% MeCN/H2O gradient over 20 min) and subsequent lyophilization afforded compound 40 as a white solid (12 mg, 56%). 1H NMR (400 MHz, CD3CN) δ 12.39 – 11.57 (m, 1H), 7.75 (d, J = 9.0 Hz, 0.7H), 7.69 (rotamer: d, J = 8.9 Hz, 0.3H), 6.95 (dd, J = 9.0, 1.0 Hz, 0.7H), 6.85 (rotamer: d, J = 8.9 Hz, 0.3H), 5.99 (m, 1H), 5.32 (m, 1H), 5.21 (m, 1H), 4.61 (m, 2H), 4.31 (s, 2H), 3.73 (m, 2H), 2.86 (m, 2H); 13C NMR (101 MHz, CD3CN) δ 171.4, 165.9, 159.7, 153.7, 148.9, 147.7, 134.3, 131.4, 130.4, 117.6, 116.1, 115.0, 112.9, 102.5, 66.8, 42.4, 42.3, 25.6; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C17H16NO7 346.09213, found 346.09198.</p><!><p>Compound 39 (20 mg, 51 µmol) in 2 mL of acetone at rt was treated with potassium carbonate (10 mg, 77 µmol) and methyl iodide (5.0 µL, 77 µmol) and stirred for 24 h. The reaction was diluted with EtOAc, washed with brine, and dried over Na2SO4. Purification by flash column chromatography over silica gel (50%–70% EtOAc/hexanes) afforded the intermediate methyl ester as a thick oil (10 mg, 48%). 1H NMR (400 MHz, CDCl3) δ 7.51 (d, J = 8.2 Hz, 1H), 7.13 (d, J = 9.0 Hz, 1H), 5.94 (m, 1H), 5.37 – 5.18 (m, 4H), 4.63 (m, 2H), 4.44 (s, 2H), 3.98 (s, 3H), 3.78 (t, J = 5.8 Hz, 2H), 3.48 (m, 3H), 2.85 (m, 2H); 13C NMR (101 MHz, CDCl3) δ 164.4, 163.5, 158.8, 156.3, 150.1, 145.6, 132.8, 125.6, 125.4, 118.2, 113.9, 113.4, 111.2, 94.8, 91.9, 66.7, 56.7, 53.2, 42.2, 39.3, 24.8; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C20H22NO8 404.13399, found 404.13465.</p><p>The above ester (10 mg, 25 µmol) was treated with 1.5 mL of 33% TFA/DCM at rt and stirred for 1 h. The excess TFA was removed under reduced pressure to afford 41 as semi-solid (8.0 mg, 90%). 1H NMR (400 MHz, CDCl3) δ 11.96 (bs, 1H), 7.63 (d, J = 8.5 Hz, 1H), 7.01 (d, 8.9 Hz, 1H), 5.95 (m, 1H), 5.33 (m, 1H), 5.25 (m, 1H), 4.66 (m, 2H), 4.49 (s, 2H), 4.08 (s, 3H), 3.80 (m, 2H), 2.88 (m, 2H); 13C NMR (101 MHz, CDCl3) δ 170.4, 165.3, 152.8, 152.6, 146.8, 132.5, 129.5, 118.5, 115.1, 111.9, 102.2, 101.0, 67.0, 53.5, 41.8, 39.5, 25.0; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C18H18NO7 360.10778, found 360.10759.</p><!><p>Compound 39 (91.0 mg, 276 µmol) and 2-methyl-2-butene (350 µL, 3.31 mmol) in 3.5 mL of CH3CN:H2O (1:1) at 0 °C was treated with a solution of sodium chlorite (187 mg, 1.66 mmol) and sodium monophosphate (343 mg, 2.48 mmol) in water, dropwise. After 1h stirring, the reaction was quenched with 5 % aq. Na2S2O3 solution in water. The pH of the solution was adjusted to 6 and extracted with EtOAc. The organic layer was dried over Na2SO4 and concentrated under reduced pressure.</p><p>The resulting thick oil was dissolved in 4 mL of DCM and treated with 4-N-methyl morpholine (60 µL, 540 µmol), N,O-dimethylhydroxylamine hydrochloride (27 mg, 280 µmol), and EDC (53 mg, 280 µmol). The reaction was stirred for 20 h at rt, diluted with DCM, and washed with 1M aq. HCl. The organic layer was dried over Na2SO4 and concentrated under reduced pressure. Purification by flash column chromatography over silica gel (3–6% MeOH/CHCl3) to give the intermediate Weinreb amide as a gum (61 mg, 51%, 2 steps) 1H NMR (400 MHz, CDCl3) δ 7.50 (d, J = 8.6 Hz, 1H), 7.14 (t, J = 7.8 Hz, 1H), 5.94 (m, 1H), 5.28 (m, 4H), 4.63 (d, J = 5.6 Hz, 2H), 4.42 (m, 2H), 3.96 (s, 0.5H), 3.73 (m, 2H), 3.48 (m, 5.5H), 3.43 (m, 2.5H), 3.14 (s, 0.5H), 2.87 (s, 2H); 13C NMR (101 MHz, CDCl3) δ 164.3, 159.1, 155.8, 155.3, 149.5, 145.8, 132.8, 125.4, 124.7, 118.2, 115.0, 113.8, 111.1, 94.7, 66.6, 61.8, 61.2, 56.7, 42.1, 39.3, 35.8, 32.4, 24.8; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C21H25N2O8 433.16054, found 433.15886.</p><p>The above amide (15 mg, 35 µmol) was treated with 1.5 mL of 33% TFA/DCM at rt and stirred for 2 h. The excess TFA was removed under reduced pressure to afford pure 42 as a semi-solid (13 mg, 96%). 1H NMR (400 MHz, CDCl3) δ 7.49 (bs, 1H), 6.96 (d, J = 7.2 Hz, 1H), 6.88 – 6.35 (bs 1H), 5.94 (m, 1H), 5.33 (m, 1H), 5.24 (m, 1H), 4.65 (m, 2H), 4.45 (m, 2H), 3.80 (m, 2H), 3.75 – 3.50 (bs, 3H), 3.39 (s, 3H), 2.87 (m, 2H); 13C NMR (101 MHz, CDCl3) δ 159.2, 158.9, 155.4, 150.0, 146.6, 132.6, 126.4, 118.3, 116.9, 114.2, 112.0, 108.8, 76.6, 76.5, 66.8, 61.8, 41.9, 39.4, 24.9; HRMS (ESI-TOF) (m/z) [M+H]+ calcd for C19H21N2O7 389.13433, found 389.13365.</p><!><p>Expression of 59.2 kD polyhistidine-tagged puritin-hIRE-1 fusion protein was carried out in SF21 cells using the Bac to Bac expression system (Invitrogen) according to manufacturer specifications. An 8X-His-puritin sequence was fused to the N-terminal end of the cytoplasmic kinase/RNase domain of human IRE-1 (aa. 547–977) in the pFastbacDual-PBL expression vector and included a PreScission protease cleavage site in the linker. Frozen insect cell paste (1g) was suspended in 8 mL lysis buffer (50 mM Tris/HCl pH 8.0, 300 mM NaCl, 5 mM βME, 10 mM imidazole) containing one protease inhibitor tablet and lysed using sonication. After removal of the cell debris via centrifugation, the supernatant was applied to a Ni(NTA) column (5 mL). After washing untagged protein by flushing with 10 column volumes of lysis buffer, the target protein was eluted using a linear imidazole gradient (15 column volumes, 10–300 mM). Fractions were analyzed via SDS-PAGE. Pooled protein-containing fractions were concentrated and rebuffered into 50 mM Tris, pH 8.0, 150 mM NaCl, 1 mM DTT via ultrafiltration. Typically, 1L of insect cell culture yielded 3 mg of recombinant 8x-His-puritin-hIRE-1 following Ni(NTA) column purification.</p><!><p>The endoribonuclease activity of recombinant hIRE-1 was assayed by incubation of 50 µL of 10 nM hIRE-1 and 50 µL of various concentrations (0.01–1 µM) of fluorescently tagged XBP-1 RNA stem loop (5'-Cy5-CAGUCCGCAGCACUG-BHQ-3', obtained from Sigma-Aldrich Co.) in assay buffer (20 mM HEPES, pH 7.5, 50 mM KOAc, 0.5 mM MgCl2, 3 mM DTT, 0.4% PEG, and 5% DMSO) for up to 2 hours at room temperature in a black 96-well plate. Fluorescence was read at various time points using a Biotek Synergy H1 plate reader with excitation and emission at 620 nm and 680 nm, respectively. The Km of purified recombinant hIRE-1 was determined to be 45 nM using the Michaelis-Menten kinetic model. Inhibition of RNA cleavage by small molecules was determined by pre-incubation of 40 µL of 15 nM hIRE-1 with various concentrations of compounds (40 µL) in assay buffer for 30 min at room temperature. A 150 nM solution of fluorescent XBP-1 RNA (40 µL) was then added to each well and the reaction allowed to proceed for 2 hours before reading fluorescence as described above. Final concentrations of hIRE-1 and XBP-1 RNA were 5 nM and 50 nM, respectively. All fluorescence readings were corrected using background values from wells containing only 120 µL of 50 nM XBP-1 RNA. Dose-response experiments were carried out a minimum of 4 times on different days and IC50 values calculated from the mean inhibition value at each concentration.</p><!><p>Antibodies against IRE-1 (Cell Signaling), PARP (Cell Signaling), XBP-1s (Santa Cruz), p97 (Fitzgerald), and actin (Sigma), were obtained commercially.</p><!><p>Primary B cells were purified from wild-type mouse spleens by negative selection using anti-CD43 magnetic beads (Miltenyi Biotech). These cells as well as the human mantle cell lymphoma (MCL) cell lines Mino and Jeko were cultured in RPMI 1640 media (Gibco) supplemented with 10% heat-inactivated fetal bovine serum (FBS), 2 mM L-glutamine, 100 U/ml penicillin G sodium, 100 Sg/ml streptomycin sulfate, 1 mM sodium pyruvate, 0.1 mM non-essential amino acids, and 0.1 mM β-mercaptoethanol (β-ME).</p><!><p>Cells were lysed using RIPA buffer (10 mM Tris-HCl, pH 7.4; 150 mM NaCl; 1% NP-40; 0.5% sodium deoxycholate; 0.1% SDS; 1 mM EDTA) supplemented with protease inhibitors (Roche). Protein concentrations were determined by BCA assays (Pierce). Samples were boiled in SDS-PAGE sample buffer (62.5 mM Tris-HCl, pH 6.8; 2% SDS; 10% glycerol; 0.1% bromophenol blue) with β-ME and analyzed by SDS-PAGE. Proteins were transferred to nitrocellulose membranes, blocked in 5% non-fat milk (wt/vol in PBS), and immunoblotted with indicated primary antibodies and appropriate horseradish peroxidase-conjugated secondary antibodies. Immunoblots were developed using Western Lighting Chemiluminescence Reagent (Perkin-Elmer).</p><!><p>Appropriate numbers of cells were suspended in phenol red-free culture media, seeded in 96-well cell culture plates, and treated with indicated IRE-1 inhibitors. After 48 h, cells were spun down and proliferation was assessed by XTT assays (Roche) according to the manufacturer's instructions. Briefly, 50 µl XTT labeling reagent, 1 µl electron-coupling reagent and 100 µl phenol red-free culture media were combined and applied to each well of the 96-well plates. Cells were then incubated for 4 h in a CO2 incubator to allow for the yellow tetrazolium salt XTT to be cleaved by mitochondrial dehydrogenases of metabolic active cells to form the orange formazan dye, which can be quantified at 492 nm using a BioTek Synergy H1 MicroPlate Reader.</p><!><p> ASSOCIATED CONTENT </p><p>Supporting Information. Copies of 1H NMR spectra for assayed inhibitors, kinetic RNase activity data for recombinant IRE-1, FRET-suppression assay dose response curves for all active inhibitors, stability data for compound 24, and XTT assay with 30 in the presence of mouse B cells. These materials are available free of charge via the Internet at http://pubs.acs.org.</p>

PubMed Author Manuscript

Electrostatic Unfolding and Interactions of Albumin Driven by pH Changes: A Molecular Dynamics Study

A better understanding of protein aggregation is bound to translate into critical advances in several areas, including the treatment of misfolded protein disorders and the development of self-assembling biomaterials for novel commercial applications. Because of its ubiquity and clinical potential, albumin is one of the best-characterized models in protein aggregation research; but its properties in different conditions are not completely understood. Here, we carried out all-atom molecular dynamics simulations of albumin to understand how electrostatics can affect the conformation of a single albumin molecule just prior to self-assembly. We then analyzed the tertiary structure and solvent accessible surface area of albumin after electrostatically triggered partial denaturation. The data obtained from these single protein simulations allowed us to investigate the effect of electrostatic interactions between two proteins. The results of these simulations suggested that hydrophobic attractions and counterion binding may be strong enough to effectively overcome the electrostatic repulsions between the highly charged monomers. This work contributes to our general understanding of protein aggregation mechanisms, the importance of explicit consideration of free ions in protein solutions, provides critical new insights about the equilibrium conformation of albumin in its partially denatured state at low pH, and may spur significant progress in our efforts to develop biocompatible protein hydrogels driven by electrostatic partial denaturation.

electrostatic_unfolding_and_interactions_of_albumin_driven_by_ph_changes:_a_molecular_dynamics_study

1. INTRODUCTION<!>Atomistic BSA Model Simulations<!>Circular Dichroism Experiments<!>Electrostatic Potential Calculations<!>3. RESULTS AND DISCUSSION<!>4. CONCLUSIONS

<p>The study of protein aggregation is critically important for understanding the etiology of many misfolded protein disorders such as Alzheimer's disease, Parkinson's disease, type 2 diabetes, and sickle cell anemia.1 In addition to a deeper understanding of these conditions, lessons learned from protein aggregation studies have led to the fabrication of several commercially interesting self-assembling biomaterials.1 Not surprisingly, given their intrinsic biocompatibility and similarities to the extracellular matrix of certain tissues,2,3 biological hydrogels are used extensively in medical applications. These biological hydrogels have been synthesized from a variety of biomacromolecules by forming intermolecular cross-links via thermal or chemical methods.2,3 Proteins are one such type of complex biomacromolecules with a well-established hierarchical structure, from the primary sequence of amino acid residues to multiprotein assemblies at the quaternary level. The structural complexity of a single protein's three-dimensional structure (tertiary level) depends on the delicate interplay between electrostatic, hydrophobic, hydrogen bonding, and other interactions, whose modifications can result in significant conformational changes.4 Evolutionary optimization of these interactions in physiological environments has resulted in protein conformations that are functionally operational.5 However, by altering the electrostatic charges on protein surfaces in a targeted manner, we can unlock the original protein structure leading to the exposure of buried hydrophobic regions. These regions could ideally drive new quaternary assemblies that promote hydrogel formation while preserving some of the original protein functionality in unchanged domains. Earlier studies have utilized similar approaches to probe the mechanisms behind the formation of functional6 and diseased7 amyloid structures.</p><p>The simulations in the present study explore the origins of intermolecular aggregation of albumin at the level of atomistic interactions prior to hydrogel formation. The results are likely to shed critical light into the fundamental understanding of the competition between folding and assembly of macromolecules and also for the understanding of their influence in many clinically relevant phenomena, such as amyloid formation and the slow releasing properties of drug carrier systems that exploit albumin's natural drug binding capacity.8</p><p>Albumin is a 66 kDa water-soluble, monomeric protein, and the most abundant protein in blood plasma (40–50 mg/mL). It has three primary domains that are arranged in a heart shape with 17 disulfide bond linkages that stabilize the domains.9 It serves as the primary carrier of various solutes in plasma, including cations, bilirubin, fatty acids, and therapeutic drugs.9 There is extensive literature regarding serum albumin's affinities to various compounds,10–13 denaturation conditions,4,14–17 gelation mechanisms,18–26 and current or potential medical uses.27–33 Albumin hydrogels formed by thermal or chemical cross-linking (e.g., glutaraldehyde) or by polymer-albumin conjugated methods have been reported.9,25,32,34–36 However, these hydrogel systems typically require either the physical/ chemical modification of the albumin or the incorporation of synthetic components into the hydrogel network.</p><p>Albumin can reversibly and drastically change its conformation when exposed to changes in solution pH (transitions occurring at pH 2.7, 4.3, 8, and 10).4,9 For example, at pH 7.4, albumin has a normal heartlike structure (N isoform), while at pH 3.5 it has a partially expanded cigarlike shape (F isoform).37 Below its transition point at pH 2.7, albumin denatures into its fully expanded E isoform. During the N-F isoform transition, bovine serum albumin (BSA) passes through its isoelectric point at pH 4.7 and the net charge on the protein changes from −16 at pH 7.4 to +100 at pH 3.5.9 Low solution pH also shifts the denaturation temperature of BSA from 62 °C (at pH 7.4) to 46.8 °C (at pH 3.5).38 In concentrated solutions, we have observed that BSA proteins in the F isoform can self-assemble into a solid hydrogel network within 24 h at room temperature (RT; 25 °C) or in 30 min at 37 °C but do not form networks in the E isoform.39 In contrast, pure N isoform BSA solutions do not exhibit this gelation behavior unless the temperature rises above 62 °C when it triggers thermal denaturation of the N isoform.35 These findings suggest that the self-assembly of albumin hydrogels at RT hinges on the presence of a specific set of physicochemical features that are strongly favored in the F isoform. This raises the important question of what other interactions might be recruited in order to overcome the highly charged nature of the F isoform. To answer this question, we used atomistic molecular dynamics simulations to calculate the conformational changes in BSA due to the changes in pH from the N isoform structure and studied the interaction between two partially denatured proteins.</p><!><p>We performed a series of molecular dynamics (MD) simulations of bovine serum albumin (BSA) to develop a model of the protein at pH 3.5 and then used the model to investigate intermolecular interactions between two proteins. Fully atomistic MD simulations were performed using the GROMACS 4.5.4 simulation package.40–43 The tertiary structure of BSA was first obtained by submitting the BSA primary sequence (GenBank: CAA76847.1) to a protein homology modeling server (CPHmodels 3.0).44,45 CPHmodels identified HSA as the closest existing protein structure to BSA and the result matches well (RMSD = 1.39 Å) with recent crystallographic BSA structures.46 The resulting output file was used as the basis for all subsequent atomistic simulations of BSA. The all-atom optimized potential for liquid simulations (OPLS/AA) force field parameters47 were used to describe interactions among the atoms. FAMBE-pH, a program that calculates the total solvation free energies of proteins as a function of pH, was used to calculate the ionization state of titratable residues (ASP, GLU, HIS, LYS, ARG) on BSA at pH 7.4 and 3.5.48 The 1:1 salt effect is included, indirectly, in the FAMBE-pH method as was done for the salt-dependent generalized Born method.48 Protonation states were fixed to the model for each pH and the model was then energetically stabilized by steepest descent algorithm, followed by an equilibration for at least 1 ns in water at 300K. The protein was then immersed in a 17 × 7 × 7 nm3 box of SPC water molecules49 to allow room for protein expansion along the long axis of the simulation box, and a simulation was run for production for 64 ns in canonical (NVT) ensemble at constant temperature 300 K with Nose-Hoover temperature coupling method.40,47 Analysis of protein secondary structures was performed by the STRIDE webserver.50</p><!><p>Dilute solutions (0.005 wt %) of essentially fatty acid free bovine serum albumin (A6003, Sigma, St. Louis, MO) in deionized water were titrated to different pH levels near the N-F transition (3.5, 4, 4.5) with HCl. Solutions were loaded into triple rinsed quartz cuvettes and evaluated by Circular Dichroism spectrography (J-815, JASCO Inc., Easton, MD) with a wavelength scan from 190 to 260 nm in triplicate. Internal heating elements in the J-815 were used to thermally denature dilute albumin solutions (0.005 wt %) at pH 7.4 to 60 and 80 °C.</p><!><p>A Python script was written to compute the electrostatic potential explicitly (including all water and counterion molecules) at each point along the Connolly surface of the protein with the following equation: ∑i14πε0Qiri where the sum runs over all atoms i that are within 3 nm from the point of the Connolly surface, Qi is the charge of atom i, and ri is the distance between the charge i and the Connolly surface. The Connolly surface was computed using the built in GROMACS g_sas command with settings identical to those used by the program APBS51 embedded in Chimera52 that was used to generate the Poisson-Boltzmann potential surface. The radius of the solvent probe was 1.4 Å with 20 dots per sphere on the surface. A 3 nm radius cutoff was used in calculating the electrostatic potential contribution of every atom near each mesh point. Visualizations of molecular structures are performed with the VMD 1.9.1 software package.53</p><!><p>While the atomic structure of human albumin at pH 7.4 has been determined at a 2.5 Å resolution,54 only low-resolution 3D models based on X-ray scattering (SAXS) data exist for the F isoform.37 Therefore, to study protein aggregation of F isoform bovine serum albumin at pH 3.5, we first needed to generate an accurate model of the F isoform albumin. Recent advances in computational power, MD software,41 and theoretical methods to calculate titration states of residues in large proteins48 now enable us to simulate these conformational changes from first principles. Since the size of the simulations required to model pH atomistically in this system remains prohibitively expensive, we utilized a program called FAMBE-pH to calculate the total solvation free energies of proteins as a function of pH.48 Briefly, this program employs a combination of approaches to calculate these free energies and involves (i) solving the Poisson equation with a fast adaptive multigrid boundary element method (FAMBE); (ii) calculating electrostatic free energies of ionizable residues at neutral and charged states; (iii) defining a precise dielectric surface interface; (iv) tessellating the dielectric surface with multisized boundary elements; and (v) including 1:1 salt effects.48 The computation of the free energy of solvation by FAMBE-pH includes the following terms: (1) the free energy of creation of a molecular cavity in the water; (2) the free energy of van der Waals interactions between the protein and the water solvent; (3) the free energy of polarization of the water solvent by the protein; and (4) the free energy of equilibrium titration of protein for a given pH and conformation.55 Since the number of ionizable groups in albumin (198) is more than ~20–25, the Tanford-Schellman integral was used to calculate the equilibrium proton binding/release.48 With this program, we calculated the ionization state of titratable residues (ASP, GLU, HIS, LYS, ARG) at pH 3.5 (Figure 1a). Residues with carboxylic acid groups that increase in charge state from −1 to 0 (ASP and GLU) between pH 7.4 and pH 3.5 are shown in red while residues with primary and secondary amines (LYS, ARG, and HIS) that increase in charge from 0 to +1 are shown in blue (Figure 1). At pH 7.4, FAMBE-pH correctly predicted the deprotonation of all ASP and GLU residues: the protonation of all ARG and LYS residues, and a balance of protonated and deprotonated HIS residues were consistent with an expected overall net charge of −9 (Figure 1b). At pH 3.5, FAMBE-pH predicted that all ASP and GLU residues become protonated and that the remaining HIS residues also be protonated, while LYS and ARG residues remain unchanged (Figure 1b). The locations of these residues on BSA are distributed uniformly over the tertiary structure (Figure 1b) and represent the ionization state of BSA at pH 3.5 (net charge of +100). While the pH of the protein was effectively set to 3.5, the conformational structure was still that of the N isoform. This predicted net charge was higher than the net charge (+65) and effective charge (+13) for albumin molecules at pH 3.5, as determined by experimental titration and electrophoresis NMR experiments,56,57 but this may be due to the fact that the structure of the protein was not yet in its ideal conformation. This difference can also be explained by the fact that any observable measurement should be computed from an ensemble of structures via a Boltzmann average, however, this is not feasible with the existent computational resources.</p><p>To produce the conformational changes induced by the change in the number of charges upon pH change, we added 100 neutralizing counterion charges, a large water box, and ran a large molecular dynamics simulation with ~300 000 atoms. We observed that the electrostatic repulsions between the three domains in the protein induced a conformational transition from the N isoform to an F-type isoform as shown in the simulation snapshots for the time evolution of this process in the presence of neutralizing counterions (Figure 2a). Within tens of nanoseconds, the distance between domains 1 (orange) and 3 (purple) has increased, with the area between domain 2 (green) and domain 3 acting as a hinge for the expansion as predicted in the literature.58,59 After the initial expansion within this time, the conformation remained stable for up to 64 ns without significant conformational change (Figure 2b). To quantify the simulated expansion, we measured the interdomain distances between center of mass of domain 1 and domain 3 (Figure 2c). Consistent with the simulation snapshots, the initial rate of protein expansion was ~1.2 nm/ns. Final interdomain spacings of albumin was found to increase from 3.47 ± 0.12 nm (N isoform) to 7.26 ± 0.32 nm (F isoform) (Figure 2c).</p><p>In addition to tertiary structural changes, the partial denaturation also resulted in a net loss of alpha helical secondary structure, from 62.9% ± 2.9% in the N isoform to 53.2% ± 2.2% in the F isoform (Figure 3a). When resolved by domain, differences in the degree of preservation emerged. Domain 1 was the most preserved with a nonsignificant (p > 0.05) decrease in alpha helical content from 58.6% ± 3.8% in the N isoform to 55.8% ± 2.4% in the F isoform. In contrast, both domains 2 and 3 had significant (p < 0.01) decreases in helical content (domain 2: N = 69.3% ± 3.8% to F = 57.7% ± 3.7% and domain 3: N = 61.0% ± 2.8% to F = 46.6% ± 3.9%). Alpha helical signatures calculated from simulations were consistent with the presence of alpha helical signatures measured experimentally via circular dichroism spectroscopy at different pH values (3.5, 4, 4.5) in the F isoform range (Figure 3b). In contrast, circular dichroism data for thermally denatured albumin near the limit (60 °C) and above (80 °C) albumin's denaturation temperature, reveals complete or near complete loss of all native secondary structures (Figure 3b). Persistent secondary structural content in pH denatured albumin supports the notion that this partial denaturation pathway does not require disruption of the entire protein as in the case for thermally denatured albumin. The predicted and observed preservation of secondary structures further supports the use of hydrogels formed by the electrostatic triggering method of partial denaturation for drug delivery applications, particularly for drugs that utilize binding sites in domain 1.</p><p>Having evaluated both tertiary and secondary structural changes during the N-F conformational transition, we then analyzed the effects of this transition at the individual residue level. Specifically, we focused on the change in solvent exposure of hydrophobic residues to investigate whether hydrophobic attractions might be present that could help explain the observed protein aggregation. We calculated the solvent accessible surface (SAS) area for each residue and categorized all residues as hydrophobic or hydrophilic as determined by the Serada et al. scale (Figure 4).60 We normalized measured SAS areas by the number of atoms contained within each category (domain 1/2/3 and hydrophobic/hydrophilic) for each of the (N and F) isoforms and report the absolute values (Figure 4a). The hydrophobic SAS for domains 1, 2, and 3 in the F isoform was 0.0362 ± 0.0007 nm2/atom, 0.0406 ± 0.0009 nm2/atom, and 0.0388 ± 0.0011 nm2/atom, respectively. In the N isoform, the hydrophobic SAS for domains 1, 2, and 3 was 0.0333 ± 0.0008 nm2/atom, 0.0361 ± 0.0008 nm2/atom, and 0.0355 ± 0.0007 nm2/atom, respectively. The analysis shows that all three domains have a statistically significant (p < 0.0001) increase in the SAS area of hydrophobic residues during the N-F transition. The differences between these absolute SAS area values (domain 1: 0.0028 ± 0.0016 nm2/atom; domain 2: 0.0045 ± 0.0018 nm2/atom; domain 3: 0.0033 ± 0.0019 nm2/ atom) during the N-F transition reiterate the increase in hydrophobic SAS area for each domain (Figure 4b). In contrast, the SAS area of hydrophilic residues decreased significantly during the N-F transition. Hydrophilic SAS for domains 1, 2, and 3 in the F isoform was 0.0550 ± 0.0010 nm2/ atom, 0.0549 ± 0.0012 nm2/atom, and 0.0524 ± 0.0011 nm2/ atom respectively. In the N isoform, hydrophilic SAS for domains 1, 2, and 3 was 0.0576 ± 0.0010 nm2/atom, 0.0548 ± 0.0010 nm2/atom, and 0.0760 ± 0.0009 nm2/atom respectively. From a physical point of view of the entire protein, the hydrophobicity increases by 16% and the hydrophilicity decreases by 13%.</p><p>The total SAS area measurements when both hydrophobic and hydrophilic residues are taken together can be used to infer whether the individual domains are expanding or collapsing (Figure 4a). Although all of the N-F differences were different, the difference in domain 1 was modest (N = 0.0429 ± 0.0007 nm2/atom, F = 0.0436 ± 0.0007 nm2/atom). This small change is consistent with the earlier result that the change in alpha helical content was not significantly different between the two isoforms. However, the domain 2 expanded (N = 0.0435 ± 0.0007 nm2/atom, F = 0.0462 ± 0.0009 nm2/atom) and domain 3 collapsed (N = 0.0524 ± 0.0006 nm2/atom, F = 0.0441 ± 0.0009 nm2/atom) to a greater degree during the transition.</p><p>The large decrease in hydrophilic SAS area measured for domain 3 is worth noting. This effect is likely due to several reasons; first is the fact that ASP and GLU residues are protonated at pH 3.5 and thus, less hydrophilic, and second is the greater loss of secondary structure in domain 3. Taken together, these two effects allow ASP and GLU residues to become buried, reducing their SAS area contribution (Figure 5). While ASP and GLU residue SAS areas decrease in every domain, they are disproportionately represented in domain 3, making these effects more noticeable. On the whole, the protein is more hydrophobic in the F isoform than in the N isoform. The increases in hydrophobic SAS area and decreases in hydrophilic SAS area suggest that aggregation of F isoform BSA molecules in high concentrations may be due to intermolecular hydrophobic interactions.</p><p>To test this hypothesis, we investigated the interactions between two proteins using our new F isoform albumin models. We placed two of these configurations in contact such that their newly exposed hydrophobic surfaces, as determined by the increase in local hydrophobic SAS, were facing each other. With this arrangement, the effective concentration of albumin in water in this simulation was ~7 mg/mL, substantially lower than the experimentally observed threshold for gelation (15 mg/mL) but sufficient for examining the interaction between two proteins. We run two types of simulations, one with explicit counterions and the other without them. The absence of counterions, while unphysical, results in a tremendous speed up of the simulations and the aim was to check whether physical insightful results could be obtained. However, in the absence of counterions, large electrostatic repulsions between the proteins forced them to move away from each other soon after overcoming the initial contact attraction (Figure 6a), leading to a result that is qualitatively wrong, as shown next, demonstrating the importance of appropriately counting for the explicit counterions.</p><p>In the presence of counterions necessary to maintain system electroneutrality (200 Cl−), the two proteins stayed within 0.25 nm of each other, as indicated by the minimum distance measured between the two proteins (Figure 6a). The persistent point of contact between the two proteins was located in domain 2 but this may be an artifact of the initial protein placement (Figure 6b). Interestingly, after 36 ns, the two proteins separated from each other. This suggests that the attraction observed between the two proteins may be a result of a local minimum in the free energy as a result of the increased hydrophobicity but would need to be corroborated with additional simulations.</p><p>Calculation of the electrostatic surface energy potential provides an additional method to evaluate the intermolecular interactions. The usual way to determine electrostatic potentials in proteins is by solving the Poisson-Boltzmann (PB) equation. However, it is not clear how good the mean-field approximation would be in a system with such larger number of charges. Therefore, we performed PB calculations and explicit determination of the electrostatic potentials from the findings of the positions of all the molecules, including the ions, from the simulations. Explicit electrostatic potential calculations that factor the contribution of counterions in the system results in surface potentials that are more negative when compared to the result from PB (Figure 7a). Particularly interesting is that, in the scale shown in Figure 7a, the PB results show an almost constant, relatively high, positive potential that directly reflects the charge on the proteins, that is, the +100 that result from the low pH. In sharp contrast, the explicit calculations demonstrate relatively large, variation of the electrostatic potential across the protein surface, showing that the explicit positions of the counterions plays a dramatic role in determining the structure and interactions of proteins. This is very important since the PB calculations would suggest strong attractive interactions between the protein (anywhere on its surface!) and negatively charged molecules, or surfaces. On the other hand, the full calculations show a much more complex surface that could lead to a variety of possible interactions.</p><p>While in many cases the PB calculation is sufficient, it misses many important details regarding the effect of individual counterions in highly charged systems. For example, at residue ARG 208 (Figure 7a, green arrow), PB predicts the nitrogen atom to have a positive electrostatic potential. In fact, the explicit calculation indicates the potential is negative due to the attraction of a neighboring Cl− counterion (Figure 7c, right). A histogram of the distances to the nearest Cl− ion for the charged N+ atom on ARG 208 demonstrates that this residue is typically bound to a counterion (Figure 7c, left).</p><p>In the case of two proteins interacting with each other, we observe similar effects of the electrostatic surface potential calculation as in the single protein case (Figure 7b). To underscore the important contribution of these counterions on the interpretation of the electrostatic potential, we have additionally computed the explicit electrostatic potential while ignoring the counterions present (Figure 7b center). This results in a relatively high, positive potential similar to the one calculated by PB (Figure 7b right). We also show the potentials calculated at the Connolly surface (0.14 nm) and the SAS surface (1.4 nm) to demonstrate how the potential becomes more negative as we move away from the positive charges on the protein. This detail is largely lost in the PB calculation where the effects of numerous positive surface charges persist for greater distances.</p><p>Four additional residues at the point of contact between the two proteins are highlighted for further analysis (Figure 7d). All four (and two in particular, LYS 350 and LYS 474) rarely had any associated counterions in the single protein case. But, when brought in contact with another highly charged protein, all four residues were substantially more likely to have counterions present (Figure 7e). Both LYS 350 and LYS 474 were rarely seen without a counterion present after dimerization. In the case of GLU 478 and LYS 350, a chlorine ion was found close to both proteins (orange and purple). While there are many other positively charged surface residues on both proteins, they do not all recruit counterions to them as in the case of LYS 350 and LYS 474. This is due to the inherent entropic cost of binding every free counterion with every positively charged residue but it becomes more likely when the proteins are dimerized (Figure 7e). The increased likelihood of finding nearby counterions in the dimerized state suggests the attraction of these counterions is necessary to neutralize residue charges and promote protein aggregation.</p><p>Thus far, these observations support the hypothesis that hydrophobic interactions from the protein core and counterion association to charged residues at the proteins point of contact drives the self-assembly of the hydrogel network. Importantly, the electrostatically driven denaturation observed in these fully atomistic BSA simulations captures the conformational structures predicted by others in the literature4,37 but with a much greater accuracy.</p><!><p>Our results provide insights into what are the interactions necessary to overcome the highly charged nature of the F isoform in forming protein aggregates. When the individual proteins are highly charged, strong intramolecular electrostatic repulsions trigger a partial denaturation of the protein. We used FAMBE-pH to calculate the total solvation free energy of the protein as a function of pH and determined the probability of residue ionization on albumin. Lowering the solution pH to 3.5 and simulating with molecular dynamics enables albumin to make the N to F isoform transition in a manner that is driven by electrostatic repulsions, and that results in the exposure of core hydrophobic regions. These hydrophobic regions are critically involved in the aggregation of the proteins despite the electrostatic repulsions still present between proteins. Inter-protein electrostatic repulsions are mitigated by the attraction of counterions to charged residues at the point of contact. Extended simulations after 36 ns showed separation of two proteins, suggesting a local free energy minimum for the aggregated state with two proteins at subthreshold concentrations. Larger simulations with more than four interacting proteins would be necessary to meet the threshold concentration but are computationally demanding to perform. An explicit counterion calculation of electrostatic surface potentials resulted in new insights that were missed by conventional PB calculations. Solving the electrostatic surface potential with explicit consideration of counterions may be a useful approach in other protein and drug binding studies. Analysis of the protein conformation reveals that alpha helical structures in domain 1 are preserved and that the total secondary structural content is more preserved when compared to thermally denatured albumin gels. Future studies will explore whether such preserved structures, particularly in domain 1, can retain the binding capacity of N isoform albumin for use in drug delivery or toxin removal applications. Building on this improved understanding of partially denatured albumin conformations, puts us in a better position to harness these electrostatically triggered hydrophobically self-assembled protein gelation mechanisms to reveal new solutions to longstanding problems in drug delivery and unwanted protein self-assembly, for example, amyloid formation.</p>

PubMed Author Manuscript

Hedonic sensitivity in adolescent and adult rats: Taste reactivity and voluntary sucrose consumption

Adolescents have been hypothesized to exhibit an age-related partial anhedonia that may lead them to seek out natural and drug rewards to compensate for this attenuated hedonic sensitivity. In the present series of experiments, taste reactivity (TR) and 2 bottle choice tests were used to assess hedonic reactions to sucrose. In Exp 1, total positive taste responses to 10% sucrose solution were significantly higher in adolescent than adult rats during the infusion period. In Exp 2, adolescent animals exhibited a concentration-effect shift that was consistent with a greater hedonic sensitivity compared to adults. Conversely, adolescents exhibited fewer negative responses to quinine. Using a shortened infusion period, adolescents in Exp 3 exhibited a trend for greater positive TR than adults in response to 10 and 34% sucrose. Consistent with the TR results of Exp 1\xe2\x80\x933, adolescents consumed significantly more sucrose solution (ml/kg) than adults, although no significant age difference in sucrose preference rates emerged. The results of the current series of experiments do not support the hypothesis that adolescents exhibit an age-related, partial anhedonia, with adolescent animals, under a number of test circumstances showing greater positive taste reactivity and reduced negative responding.

hedonic_sensitivity_in_adolescent_and_adult_rats:_taste_reactivity_and_voluntary_sucrose_consumption

1. Introduction<!>2.1. Subjects<!>2.2.1. Surgery<!>2.2.2. Apparatus<!>2.2.3. Behavioral measures<!>2.2.4. Experiment 1<!>2.2.5. Experiment 2<!>2.2.6. Experiment 3<!>2.3.1. Experiment 4<!>3. Discussion

<p>Adolescence is a developmental period marked by both neural and behavioral changes. Adolescent-associated behavioral changes are surprisingly well conserved across species and include elevations in social interactions with peers, novelty seeking, and risk taking (Csikszentmihalyi et al., 1977; Primus and Kellogg, 1989; Pellis and Pellis, 1990; Pellis and Pellis, 1997; Trimpop et al., 1998). Adolescents exhibit both hyperdipsia and hyperphagia, with humans and rats exhibiting higher caloric intake relative to their body weight than any other time in the life span (Post and Kemper, 1993; Nance, 1983). In humans this developmental period is also the time during which drug use is typically initiated. The physiological mechanisms responsible for increases in consumption of natural reinforcers during adolescence might contribute to the propensity of adolescents to consume drug reinforcers as well, given that natural and drug reinforcer are thought to share common reward pathways (Di Chiara, 1999; Berridge and Robinson, 1998; Wise, 1989).</p><p>It is currently unclear if this adolescent-typical increase in the seeking and consumption of appetitive stimuli is related to increases or decreases in the incentive value attributed to rewarding stimuli. On one hand, adolescents might avidly seek out natural and drug rewards in an attempt to compensate for an attenuated sensitivity to hedonic stimuli (Spear, 2000). Some of the limited human data available supporting this hypothesis include the observation that human adolescents demonstrate less ventral striatal activation in response to reward cues and report a greater incidence of depressed mood compared to adults (Bjork et al., 2004; Petersen et al., 1993). Conversely, adolescent-typical increases in consummatory behaviors may be driven by increases in the hedonic value assigned to reinforcers (Ernst et al., 2006; Chambers et al., 2003). Supporting this notion, other functional magnetic resonance imaging (fMRI) work has shown adolescents to exhibit an exaggerated magnitude of activation in the nucleus accumbens (NAc) in response to visual stimuli associated with a monetary reward when compared with younger and older subjects (Galvan et al., 2006).</p><p>Traditionally, the consumption of appetitive stimuli has been used to index the hedonic value attributed to those stimuli, effects thought to be mediated through changes in mesolimbic dopamine function/sensitivity (Wise et al., 1978; Wise and Bozarth, 1982). The preference for mildly sweet sucrose solution (~1.0%) is one of the most extensively used behavioral measures of anhedonia, with anhedonia interpreted as a reduction in choice of the sucrose solution over water in two-bottle choice test (Willner et al., 1987).</p><p>More recently, Robinson and Berridge have proposed an incentive salience theory of reward that separates reward conceptually and functionally into two component psychological processes mediated by different neural mechanisms: "wanting" and "liking" (Berridge, 1996; Berridge and Robinson, 1998). Wanting refers to the motivational or craving element of reward, and is a reward component found to sensitize following repeated exposure to drugs of abuse. Liking, on the other hand refers to the hedonic or pleasure component of reward and is not thought to sensitize. Within this conceptual framework, dopamine is thought to be involved in assigning incentive motivational value to rewarding stimuli (i.e. wanting) (Berridge. 1996; Berridge and Robinson, 1998). Hedonic "liking" properties, on the other hand, are thought to be mediated by a hierarchical system involving opioid, cannabinoid, and to some extent GABA systems (Peciña et al., 2006).</p><p>One model that has been extensively employed to examine hedonic sensitivity to appetitive tastants is the taste reactivity (TR) test, which is thought to be a direct measure of the hedonic value attributed to stimuli. In the TR test, a solution is presented to the subject and the oral facial reactions to that tastant are proposed to reflect the palatability of the solutions. Palatable solutions, such as sucrose, elicit appetitive TR behaviors such as rhythmic and lateral tongue protrusions, whereas aversive solutions, such as quinine, elicit aversive TR behaviors (including gapes) (see methods for more details). Both appetitive and aversive taste reactions are highly conserved, with similar oral facial responses to appetitive and aversive stimuli seen in human infants as in mature primates and rodents (Steiner et al., 2001). Across-species differences emerging primarily in the timing at which particular TR behaviors are emitted, with humans and gorillas expressing behaviors much slower than rodents (Steiner et al., 2001).</p><p>The current series of experiments used the TR test, as developed for rats by Grill and Norgren (1978), to assess developmental differences in hedonic sensitivity to appetitive and aversive stimuli in adolescent and adult rats, with TR assessed both during and immediately following the infusion period. Developmental differences in voluntary consumption of an appetitive solution were also assessed using a two-bottle choice test.</p><!><p>Male Sprague–Dawley rats (Taconic Farms) bred in our colony were used. All animals were maintained in a temperature-controlled (22 °C) vivarium on a 14-h/10-h light cycle (lights on 0700) with ad libitum access to food (Purina Rat Chow, Lowell, MA) and water. On postnatal day 1 (P1), litters were culled to 7–10 pups. Animals were weaned on P21 and pair-housed with same-sex littermates until the time of the experimental procedures. No more than one animal from a given litter was used in any experimental group. At all times, rats used in these experiments were maintained and treated in accordance with guidelines for animal care established by the National Institutes of Health (1986).</p><!><p>At the onset of each experiment, beginning on P 28–29 for adolescents and P 69–70 for adults, animals were anesthetized with isoflurane prior to the implantation of an oral cannula. Each cannula consisted of polyethylene tubing (PE-50) with a heat-flanged tip. A guide needle was used to insert the cannula through the cheek at the level of the 1st maxillary molar. Another guide needle was then used to tunnel the cannula subcutaneously to the dorsal skull surface where the cannula was secured with surgical glue. Following surgery, animals were singly housed to avoid across-animal damage to the cannulas.</p><!><p>Testing was conducted in a Plexiglas cylinder (adolescent – height = 16.0 in, diameter = 8.0 in; adult – height = 16.0 in, diameter = 10.0 in). A mirror was placed at a 45° angle below the Plexiglas floor, providing a ventral view of the rat for video recording. Solutions were infused into animals' cannulas via connection to a calibrated syringe pump controlling rate, duration and volume of the infusion.</p><!><p>All taste reactivity behaviors during the infusion period and 30 s post-infusion were scored frame-by-frame. Scoring criteria for each behavior, and rationale for the classification of components into hedonic or aversive categories were based on Berridge and Grill (1983). Positive hedonic reactions included: (a) rhythmic tongue protrusions — the anterior tip of the tongue visibly emerges directly on the midline, covering the upper incisors and is then retracted — movements repeated at an approximate rate of 8.8 cycles/s; (b) lateral tongue protrusions — the tongue protrudes (nonrhythmic) from the side of the mouth followed by a forward extension, generally occurring on alternating sides of the mouth or in one single protrusion; (c) paw licks — any instance of the rat licking its paws, with the exception of paw licking occurring within a grooming sequence; (d) lateral tongue movements — the tongue emerges on the side of the mouth, extending the upper lip/cheek laterally without protruding from the mouth. Aversive reaction patterns included; (a) gapes — the mandible lowers, while the corners of the mouth contract, forming a triangle shaped mouth opening that is held for approximately 83 ms; (b) chin rubbing — the animal rubs its chin on the floor while projecting the body forward; face washing — single or several wipes of the face with the forepaws; head shake — a bout (typically with a duration of <1 s) of rapid (>60 cycles/s) side to side movements of the head; (e) forelimb flails — a brief bout (<1 s) of shaking of the forelimbs at a rate greater than 60 cycles/s; (f) paw treads — planting of the forelimbs on the floor and alternating forceful strokes forward and back. Unless otherwise noted, scoring criterion used was based on those of Berridge (Berridge, 1996, 2000). Discrete actions such as lateral tongue protrusions, gapes, and chin rubbing as well as bouts of head shaking, forelimb flails and paw treads were scored as a single count for each occurrence. Behaviors that typically persist for >1 s were recorded as follows: paw licks, rhythmic mouth movements, grooming and face washing were recorded in 5-s bins (any occurrence of these behaviors within each 5-s time bin was scored as a single count). Rhythmic tongue protrusions were scored similarly, except 2-s bins were used. The final total positive hedonic score was composed of the sum of all rhythmic tongue protrusions, lateral tongue protrusions, paw licking, and lateral tongue movements, whereas total negative hedonic scores were composed of the sum of all gapes, chin rubs, face washing, head shakes, forelimb flails and paw treads.</p><p>The experimenter was present during the test sessions to videotape, using a digital camera (JVS, HDD), a close-up view of the oral region during the infusion period and for 30 sec thereafter. An investigator blind to both solution type and age later analyzed the sessions frame-by-frame.</p><!><p>Following surgery (Experiment [Exp] Day 0), all animals were given 3 days to recover, after which the taste reactivity test was conducted (Exp Day 3—P 33–34 for adolescents (n = 5); P 72–73 for adults (n = 6)). On test day, following a 15 min habituation to the chamber, each animal received a 45 s continuous infusion of 10% sucrose solution at a rate of 1 ml/min. Total number of positive and negative responses during the 45 s infusion period and for 30 s thereafter were recorded. Number of negative responses to this appetitive tastant was negligible at both ages and assessment periods. Consequently, only positive response data were analyzed.</p><p>Total positive taste responses (i.e., the sum of paw licks, rhythmic tongue protrusions, lateral tongue protrusions and lateral tongue movements) to the 10% sucrose solution were significantly higher in adolescent than adult rats during the 45-s infusion period (t = 2.6; df = 9; p<0.028)(seeFig. l).</p><p>Total positive taste responses during the 30-s post-infusion period were relatively low and did not differ significantly with age, despite a seeming trend for adults to exhibit more responses compared to adolescents (adults: 1.3 ±0.4; adolescents: 0.6±0.3) (t = 0.9; df = 9; p = 037).</p><!><p>To determine if the findings in Exp 1 reflected a shift in the concentration–effect curve for appetitive taste stimuli, 3 different concentrations of sucrose solutions as well as a water control were examined in this experiment. Taste reactivity to a quinine solution was also assessed to determine whether age-related differences in affective responses observed in Exp 1 would also be evident in response to aversive stimuli.</p><p>A 2 age (adolescent and adult) × 5 solution (water, 0.34%, 3.4% or 34.0% sucrose, and quinine) within subjects design was used. The procedure was similar to Exp 1 with the exception that animals were given a 1-day recovery period following surgery. On Exp Day 2 (P 30 for adolescents; P 72 for adults), all animals received a water trial. Exp Days 3–5 consisted of 1 sucrose trial per day, with the order of concentration presentation varied randomly among animals. A quinine (3.0 × l0−4 M) trial was conducted on the last day of testing for all animals. Each infusion was given over 45 s and was infused at a rate of 1 ml/min. Oral responses to each tastant were examined during the 45 s infusion period and for 30 s thereafter.</p><p>Adult animals exhibited significantly greater baseline responding (i.e., more positive taste responses to water) during the infusion period than did adolescents (adults: 4.25 ± 1.2; adolescents: 1.5 ± 1.1). Given this age difference in responding to the baseline test stimulus, adolescent and adult data were analyzed separately. In these analyses, positive taste responses to water and sucrose (0.34, 3.4 & 34.0%) were examined at each age via a repeated measures ANOVA (n = 10). with responses to quinine versus water examined separately (n = 8). Repeated measures ANOVA of positive taste responses observed during the 45-s infusion period again revealed a significant effect of solution in both adolescents [F(3,27) = 17.58, p<0.05] and adults [F(3,24) = 14.32, p<0.05]. Post hoc tests showed that the 3.4 and 34.0% sucrose concentrations produced significantly elevated levels of appetitive taste reactions in adolescent animals, whereas 34.0% was the only concentration that produced an increase in responding among adults. These data are shown in Fig. 2.</p><p>The repeated measures ANOVAs of positive responses observed during the 30-s post-infusion period again revealed a main effect of concentration in both adolescents [F(3,24) = 5.24, p<0.05] and adults [F(3,24) = 12.7, p<0.05]. Post hoc tests in both the adolescents' and adults' analyses indicated that positive taste reactions in response to 34% sucrose were significantly greater than those exhibited to water (Fig. 3).</p><p>An age difference in negative taste responses to water was also observed, with adult animals exhibiting significantly fewer negative responses to water during the infusion period than adolescents (adult 4.4 ± 0.9; adolescent 7.3 ± 1.0) (t = 2.1; df = 14; p<0.05). Given this age difference in baseline responding, adolescent and adult data were analyzed separately. In these analyses, negative taste responses to water and quinine were examined at each age via repeated measures ANOVAs. Adults exhibited significantly higher negative taste responses to quinine relative to their water baseline during the 45-s infusion period [F(l,6) = 19.3, p<0.05], whereas negative responding to quinine among adolescents did not differ from their water baseline [F(l,8) = 0.4, p>0.05]. A similar pattern was seen, during the 30-s post-infusion period, with adults exhibiting higher negative responses to quinine than water [F(1,6) = 16.1, p<0.05], while the adolescents' TR to quinine did not differ from their response to water [F(l,8) = 1.0, p>0.05] (see Fig. 4).</p><!><p>Previous studies have suggested that under certain experimental conditions post-infusion responding can be more sensitive to manipulation than responding during the infusion period (Grill et al., 1996). Unfortunately, in Exp 1 and 2, levels of responding during the post-infusion period were generally low, and hence possible floor effects constrain interpretation of these data. Given that shorter infusion periods have previously been associated with greater post-infusion responding (Grill et al., 1996), in this experiment the duration of each infusion period was shortened to 30 s (Fig. 5).</p><p>The design of this study was a 2 age (adolescent and adult) × 3 solution (water, 10% and 34% sucrose) factorial (n = 8), with the last factor consisting of a repeated measure. To reduce the number of animals lost to cannula occlusion, animals were implanted with bilateral cheek cannulas. As in Exp 2, animals were given a 1-day recovery period following surgery prior to testing. Animals received one trial with each tastant, with all trials conducted on the same day. A 2 h interval separated each trial, during which animals were returned to their home cage. Water was presented first, followed by the sucrose solutions in ascending concentration. For each test, animals were placed into the test chamber and given a 15 min habituation period, followed by a 30 s continuous infusion of the solution presented at a rate of 1 ml/min. Behaviors were initially scored as described in General Methods, and then were rescored using criteria designed to minimize differential weighing of behaviors to the composite positive TR score. With this rescoring, behaviors previously analyzed as counts per 2 or 5 s bins were rescored into 1 s bins.</p><p>When the data were analyzed using the original scoring criteria, no age differences emerged during the water infusion period. The ANOVA of adolescent and adult total positive responses during the 30-s infusion period revealed a main effect of solution, with both concentrations of sucrose producing a higher number of responses than to water [F(2, 492) = 24, p<0.05] (see Fig. 6a). There was no significant main effect or interaction involving age. A main effect of solution also was evident in the analysis of the 30-s post-infusion period, with both concentrations of sucrose producing significantly more responses than water [F(2. 206) = 8, p<0.05] (see Fig. 6b). Again, no age effects were evident.</p><p>Examination of individual TR responses revealed age differences in some of the component behaviors to specific taste stimuli (see Fig. 7). For instance, adolescents exhibited primarily rhythmic tongue protrusions during water infusion (1.9 ± 0.9), whereas the adults' response consisted mainly of lateral tongue protrusions (1.8 ± 1.8). The pattern of behavioral responses to 10% sucrose was relatively similar between adolescents and adults, with rhythmic tongue protrusions comprising the major component responses at both ages at this concentration (adolescent: 10.9 ± 1.6; adult: 9.3 ± 1.8) and in response to the 34% sucrose concentration (adolescent: 7.4 ± 2.1; adult 7.8 ± 2.0). During the infusion of 34% sucrose, however, paw licking emerged as a prominent component of adolescents' TR behavior (2.0 ± 1.0), whereas this behavior was rare in adults (0.3 ± 0.3). In contrast, lateral tongue protrusions continue to be a notable component of adults' response to 34% sucrose (3.6 ± 0.8), as it was also in response to water and 10% sucrose.</p><p>Given the different patterns of specific TR responses observed across age, composite data summed over differentially weighted responses are difficult to interpret. Consequently, data from Experiment 3 were rescored using criteria designed to minimize differential weighing of behaviors to the composite positive taste reactivity score. Specifically, all behaviors were scored either as single occurrences or in 1 s bins.</p><p>The repeated measures ANOVA of total positive responses as scored by the revised scoring method during the 30-s infusion period revealed a main effect of solution, with both concentrations of sucrose eliciting greater total positive TR than water [F(2,28) = 36.07, p<0.05] (see Fig. 8). There was no significant main effect or interaction involving age, despite a trend for adolescents to exhibit greater positive TR in response to both concentrations of sucrose than adults [F(2, 28) = 0.65, p>0.05]. A main effect of solution was again observed in the analysis of the 30-s post-infusion period, with both concentrations of sucrose producing higher responding than water [F(2,28) = 11.26, p<0.05], and adolescents again tended to show greater overall positive responding than adults, although the age effect did not reach significance [F(l,14) = 0.82, p>0.05].</p><p>Chi square tests comparing component behaviors across age with each sucrose solution and test period revealed that individual responses composing the total positive TR to the infusion of 34% sucrose differed significantly between adolescents and adults [X2(3, n = 16) = 8.32, p<0.05]. As in the analysis of the data from the original scoring method, adolescents given 34% sucrose exhibited significantly more paw licks than adults (adolescent: 8.9 ± 4.2; adult: 0.8 ± 0.8), whereas lateral tongue protrusions were exhibited with significantly higher frequency among adults than adolescents (adolescent: 0.9 ± 0.6; adult: 3.6±0.8). There was no difference in individual responses to 10% sucrose during the infusion or post-infusion period or the 34% sucrose post-infusion period.</p><!><p>All animals were single housed (Exp Day 0) and given a 24 h acclimation period prior to initiation of testing (n = 8). On the following 14 days (Exp Day 1–14 — P 28–42 for adolescents; P 65–79 for adults) animals were given 24 h access to one bottle containing water and one bottle containing 1.0% sucrose solutions, with the location of the two bottles alternated daily. Food was continuously available ad libitum. Consumption of food, water and sucrose solution was recorded daily.</p><p>As expected, a t-test revealed that adolescents' average daily food consumption (g/kg) across Exp Day 1–14 was significantly greater than their adult counterparts [t = 1.5; p<0.05] (see Table 1). A repeated measures ANOVA of solution (ml/kg) consumed revealed a significant solution by age interaction [F(l, 14) = 14.9, p<0.05]. Post hoc tests indicated that adolescents consumed significantly more sucrose solution than adults did and when compared with their own intake of water, whereas adult sucrose intake did not differ from water (Fig. 9). Despite adolescents' elevated sucrose intake, no significant age difference emerged in the analysis of the sucrose preference scores (sucrose intake / (sucrose intake + water intake) × 100) across Exp Day 1–14 [t = 1.32; df = 14; p>0.05].</p><!><p>The purpose of the present series of experiments was to examine developmental differences in hedonic sensitivity in adolescent and adult rats as measured by both taste reactivity and voluntary consumption. The TR experiments suggest that adolescent rats exhibit a developmental difference in hedonic sensitivity to both appetitive and aversive stimuli. In general adolescents were more sensitive to the hedonic properties of appetitive stimuli than adults, findings that are inconsistent with the hypothesis that adolescence is characterized by an age-related partial anhedonia. Conversely, adolescents were less sensitive to the aversive properties of quinine. This finding was also unexpected, given reports that human adolescents may show a propensity to exhibit greater negative affect to various experiences relative to other aged individuals (e.g. Larson and Asmussen, 1991).</p><p>In Exp 1, adolescents demonstrated nearly 3 times more positive TR in response to a 45 s infusion of 10% sucrose than did adult animals, although there was a trend for adults to exhibit more TR during the post-infusion period. Consistent with these findings, adolescent rats in Exp 2 required a lower concentration of sucrose (3.4%) to exhibit an increase in positive TR relative to water compared to adults (34.0%). Although an age effect was not observed during the post-infusion period, adults again exhibited a trend for more pronounced positive TR in response to 34.0% sucrose when compared with their adolescent counterparts. The shorter infusion period used in Exp 3 resulted in similar levels of total positive responding to 10 and 34% sucrose solutions across age during both the infusion and post-infusion periods. An examination of the individual positive responses in this later experiment, however revealed some age differences in the component behaviors contributing to the overall positive TR responses. When the results were rescored using a revised scoring method designed to minimize differential weighing of component behaviors, adolescents showed a trend for greater total responding to sucrose at both concentrations tested.</p><p>It was only during the infusion of 34% sucrose that paw licks emerged, with the incidence of this response to the most concentrated sweet solution more prominent in adolescents than adults. As can be seen in Fig. 8, both ages exhibited different patterns of responding across the two observation periods, with rhythmic tongue protrusion being the primary response during the infusion period, while lateral tongue protrusions occurred more frequently during the post-infusion period.</p><p>In Exp 4, when sucrose and water intake were examined in two bottle choice tests, adolescents consumed significantly more sucrose than adults and than their own intake of water, while adult intake did not differ across solutions. However, no significant age difference in sucrose preference ratios emerged. Adolescents' elevated consumption of sucrose solution is reminiscent of the findings of Exp 1–3, where adolescents were observed to be more sensitive to appetitive properties of sucrose. Unlike the TR test, the two-bottle choice test does not discriminate wanting from liking, given that voluntary consumption tests inherently contain an element of instrumental behavior. Traditionally, hedonic aspects of reward have been inferred from tests that do not differentiate between liking and wanting aspects of reward (i.e. preference (choice), consumption (intake), or instrumental behavior (bar pressing or approach), based on the assumption that the two processes are mutually inclusive. In direct contradiction of this assumption are the findings that pharmacological manipulations of the dopamine system decrease both food intake and instrumental measures without altering liking as measured by TR (Peciña et al., 1997; Treit and Berridge 1990). Additionally, 6-hydroxydopamine lesions are known to produce aphagia and adipsia, but do not alter hedonic or aversive taste reactions (Berridge et al., 1989).</p><p>A hierarchical system involving opioid, cannabinoid, and to some extent GABA systems is thought to mediate hedonic properties of rewarding stimuli (Peciña et al., 2006), although little is known about these systems during adolescence. There is some evidence for ontogenetic difference in expression of type 1 endocannabinoid receptors (CB1), with numbers peaking during adolescence in striatum. limbic forebrain, and ventral mesencephalon. and lower levels in other brain areas including the NAc, hippocampus and cortex as compared to adults (Rodríguez de Fonseca et al., 1993; Romero et al., 1997). Likewise, the limited behavioral evidence available suggests that adolescent rats may be more sensitive to CB agonists under some conditions than adults (Cha et al., 2007; Cha et al., 2006), an increased sensitivity postulated to reflect compensatory upregulation of CB1 receptors in response to lower endogenous CB "tone" in areas such as the hippocampus of adolescents (Kang-Park et al., 2007). To the extent that adolescence is characterized by low CB tone and compensatory CB receptor upregulation within hedonic hotspots, the endocannabinoid system may partially mediate the increased hedonic sensitivity observed in adolescent animals in the present series of experiments. Opiate receptor systems could contribute to enhanced hedonic sensitivity during adolescence as well, with for instance stimulatory effects of mu opioid receptor agonist DAMGO on social facilitation peaking in early adolescence and declining dramatically thereafter (Varlinskaya and Spear 2008).</p><p>Although the traditional scoring method was revised to reduce differential weighing of individual behaviors by scoring all behaviors as single occurrences or in 1 s bins, age differences in the expression of component behaviors continue to complicate interpretation of hedonic sensitivity. For example, adolescents primarily exhibit behaviors scored in time bins (e.g., rhythmic tongue protrusions and paw licks) creating an inherent ceiling effect, whereas adults exhibit more behaviors measured in counts (e.g., lateral tongue protrusions), presenting challenges for data interpretation.</p><p>For instance, is it reasonable to conclude that adolescents demonstrate a greater hedonic response to sucrose than adults from data showing that a behavior emitted only in response to the sweetest solution (i.e., paw licks) is emitted with greater frequency among adolescents? Additionally, ontogenic differences in the individual behaviors comprising patterns of TR across concentrations of sucrose make it difficult to determine how individual behaviors relate to intensity of the hedonic response.</p><p>One possible contributor to the ontogenic differences in taste reactivity seen here is the decision to equate rate and duration of administration across age, hence delivering the same total volume of fluid at each age, despite differences in body size between adolescents and adults. This strategy was chosen based on preliminary work showing no significant effects of variation in volume on positive TR at either age. Ontogenetic differences in the current experiments certainly do not appear to be related in any simple fashion to differences in solution volume to body size ratio, given that adolescents demonstrated greater positive TR than adults in response to sucrose in Exp 1, but diminished negative TR in response to quinine in Exp 2.</p><p>The results of the current series of experiments do not support the hypothesis that adolescents exhibit an age-related, partial anhedonia that might contribute to their propensity to "consume" natural and drug reward more avidly (e.g. see Spear, 2000). Indeed, the present experiments suggest adolescents often express enhanced sensitivity to the hedonic properties of appetitive tastants. The possibility remains that taste reactivity data might not extrapolate to other reinforcers. For example, when hedonic affect is measured via ultrasonic vocalizations (USV) during social interactions, adolescent animals emit fewer USVs compared to adults, suggesting adolescents may be less sensitive to the hedonic properties of social interaction (Willey et al., in press). It may be the case that age differences in liking are dependent on the reinforcer examined. Future studies will need to address this possibility by examining a number of natural rewards.</p>

PubMed Author Manuscript

Targeting KIT in Melanoma: A Paradigm of Molecular Medicine and Targeted Therapeutics

Despite multiple clinical trials utilizing a spectrum of therapeutic modalities, melanoma remains a disease with dismal outcomes in patients with advanced disease. However, it is now clear that melanoma is not a single entity, but can be molecularly divided into subtypes that generally correspond to the anatomical location of the primary melanoma. Melanomas from acral lentiginous, mucosal, and chronic sun-damaged sites frequently harbor activating mutations and/or increased copy number in the KIT tyrosine kinase receptor gene, which are very rare in the more common cutaneous tumors. Multiple case reports and early observations from clinical trials suggest that targeting mutant KIT with tyrosine kinase inhibitors is efficacious in KIT mutant melanoma. This review recounts what is known about the role of KIT in melanocyte maturation, our current understanding of KIT genetic aberrations in melanoma, and how this knowledge is being translated into clinical oncology.

targeting_kit_in_melanoma:_a_paradigm_of_molecular_medicine_and_targeted_therapeutics

1. The KIT Receptor Tyrosine Kinase<!>2. The Identification of KIT Mutations and Amplifications in Melanoma<!>3. Melanoma KIT Mutant Cell in vitro Experiments<!>4. The Biology of KIT Mutations in Melanoma<!>5. Therapeutic Interventions into KIT Mutant Melanoma<!>6. Conclusion

<p>The KIT receptor tyrosine kinase gene (c-kit) was first identified in 1987 based on sequence similarity to the acute transforming Hardy-Zuckerman 4 feline sarcoma virus (v-kit) [1, 2]. KIT (a.k.a., CD117) is a type III receptor tyrosine kinase characterized by a glycosylated extracellular ligand binding domain containing five immunoglobulin-like repeats, a single hydrophobic transmembrane domain, and an intracellular segment containing a juxtamembrane inhibitory domain, and two tyrosine kinase domains separated by a kinase insert region (Figure 1)[3]. Alternative splicing of KIT can result in the loss of a GNNK amino acid sequence at the 5' end of the extracellular domain and/or the loss of a serine amino acid residue in the kinase region of the intracellular domain[4, 5]. Stem Cell Factor (SCF, a.k.a., kit ligand, steel factor, or mast cell growth factor), the ligand for KIT, is also a glycosylated transmembrane protein. Alternative splicing results in the presence or absence of a proteolytic cleavage site within the SCF protein[6]. SCF that harbors the cleavage site is released as the soluble form whereas SCF without the cleavage site remains on the cell surface. Either form of SCF is capable of binding to KIT resulting in receptor dimerization, autophosporylation, and activation of the intracellular tyrosine kinase domain, although the ultimate signaling effects generated by the soluble versus membrane-bound SCF differ. Binding of the soluble form of SCF causes KIT activation, internalization, and degradation, whereas binding of the membrane-bound form of SCF results in prolonged KIT activation[7]. Activated KIT has been shown to initiate multiple downstream signaling pathways (e.g., MAPK/MEK, PI3K/AKT, JAK/STAT) that can vary depending on the cellular context in which KIT is activated[3, 8-10].</p><p>The importance of both KIT and SCF for proper melanogenesis, proliferation, migration and survival has been clearly demonstrated by the phenotypes of mice and humans that harbor genetic alterations in these genes. Loss of function mutations in mouse KIT or SCF result in a white color spotting of fur, and loss of function mutations in KIT in humans results in lack of the pigmentation of skin/hair (viz, piebaldism)[11-13]. The pattern of these phenotypes indicates that developing melanocytes with loss of function mutations cannot migrate to distant sites from the neural crest. In addition, other traits associated with proper KIT activity (e.g., gametogenesis, hematopoeisis, mast cell and interstitial cells of Cajal function) are affected by loss of function KIT mutations[10, 14-16]. Despite its essential role in melanogenesis, KIT was not thought to have a major role in promoting melanomas, as KIT protein expression is frequently lost during local melanoma growth/invasion, and overexpression of KIT in metastatic melanoma cells resulted in reduced tumor growth[17, 18].</p><!><p>The first report of a KIT mutation in melanoma came from a 2004 publication by Went et al., who used tissue microarray to screen different tumor types for KIT protein expression level followed by mutation analysis on a small subset of tumors with high KIT protein expression[19]. Fourteen of 39 (36%) primary malignant melanomas showed KIT expression by immunohistochemistry (IHC). Extrapolating from the location of KIT mutations in other tumors, exons 2, 8, 9, 11, 13, and 17 were sequenced (Table 1). Of the two melanoma tumors selected for sequencing analysis, one had a point mutation in KIT. In 2005, Willmore-Payne et al., reported screening 100 melanomas (84 metastatic, 12 primary cutaneous, and 4 in situ, no primary mucosal) for KIT protein expression. Twenty-nine samples (29%) showed KIT expression by IHC, two of which harbored mutations in KIT as analyzed by high resolution amplicon melting[20]. Both of these samples were metastatic samples with high KIT expression and did not have BRAF mutations. In a follow-up paper in 2006 WIllmore-Payne et al. screened an additional 53 cases of melanoma finding KIT protein expression in 6 cases (11%)[21]. Mutation analysis demonstrated one of these samples to harbor a KIT mutation. Flourescent in situ hybridization (FISH) indicated a slight increase in KIT/CEP4 ratio in one of three of the tumors with a KIT mutation.</p><p>In 2005, the group of Dr. Boris Bastian published a seminal paper describing DNA copy number changes in melanomas arising from different anatomic sites[22]. The study included a total of 126 melanomas: 30 from skin with chronic sun-damage, 40 from skin without chronic sun-damage, 36 from acral sites, and 20 from mucosal sites. Comparative genome hybridization (CGH) analysis demonstrated that while certain genetic alterations were shared, there were many marked differences between the tumors originating from these four sites.</p><p>In particular, the non-cutaneous tumors had a much higher frequency of amplification events than the cutaneous melanomas, and they involved distinct areas. In addition, analysis of the prevalence of mutations in BRAF and NRAS, the most common activating somatic mutations in melanoma, also showed marked differences between the groups. Mutations in BRAF (59%) and NRAS (22%) were very common in tumors arising from skin without chronic sun-damage. BRAF and NRAS mutations were much less common in melanomas with chronic sun damage (11% and 15%), acral melanomas (23% and 10%), and mucosal melanomas (11% and 5%). Subsequent to these findings, Bastian's group performed an in-depth analysis of chromosomal region 4q12, which had evidence of frequent copy number gain in the mucosal, acral, and chronic sun-damaged melanomas, but not in the non-sun-damaged cutaneous tumors. In addition to this interesting distribution, the reqion was also of interest as it harbored several genes that could be utilized as therapeutic targets. Immunohistochemical and mutational analysis of the tumors with copy gain in this region found both a correlation with stong expression of the KIT protein, and point mutations in the KIT gene. This prompted dedicated analysis of the KIT gene in the full set of tumors, and identified KIT point mutations in 21% of the mucosal, 11% of the acral, 17% of chronic sun-damaged cutaneous, and 0% of non-chronic sun-damaged melanomas[23]. This was the first study to show that KIT genetic aberrations are enriched in specific subtypes of melanoma and are mostly mutually exclusive with BRAF or NRAS mutations within melanoma tumors.</p><p>The observations of Bastian significantly enriched the understanding of different types of melanoma. Traditionally, melanomas have been categorized by anatomical location and histological features -- melanomas localized to the skin being designated as superficial spreading, nodular, acral lentiginous, and lentigo maligna, and non-cutaneous melanomas being comprised of uveal and mucosal subtypes. These categorizations are still of importance for clinical-pathological purposes, as there are important differences among these subtypes (Table 2).</p><p>The KIT mutations identified in these subtypes of melanoma differ from those found in gastrointestinal stromal tumors (GIST). Somatic mutations in KIT have been identified in approximately 80% of GISTs. The majority of these mutations are deletions or insertions in the gene,[9] whereas all the melanoma KIT mutations in this study where substitution mutations. Another significant difference in melanoma was that mutations in KIT were not present in exon 9, which is the location of about 15% of KIT mutations in GIST. In addition, the Bastian study identified KIT gene amplification in 7% of acral lentiginous, 8% of mucosal, and 6% of chronic sun-damaged melanomas (Table 3). A substantial number of acral lentiginous and mucosal melanomas also had a KIT copy number that was increased, but did not meet the definition of amplification. In contrast, KIT amplification is rarely observed in GIST tumors. In reviewing all reports to date, about 30% of melanoma samples with KIT mutations also show increased copy number/amplification of KIT [21, 24-26].</p><p>Multiple groups have now reported the KIT genetic aberration status in additional cohorts of melanoma tumor samples. Antonescu et al., examined 20 primary anal mucosal melanoma samples and discovered a 15% frequency of KIT point mutations, with 1 of 3 KIT mutant samples also showing increased copy number[24]. Of 15 evaluable primary oral mucosal melanoma tumors examined by Rivera et al, 27% had KIT point mutations[27]. KIT copy number and exon 17 mutation status was not determined in this study. Although the total number of patients assessed to date is modest, the frequency of KIT mutation genitourinaryanorectal melanoma is nearly twice that of head and neck mucosal melanoma (Table 4). In studies that separated the genitourinary from anorectal mucosal melanomas, KIT mutations were more common in the former. Whether these differences in the frequency of KIT mutation in different subtypes of mucosal melanoma will be maintained awaits assessment of greater sample sizes.</p><p>Beading et al., screened a large number of melanoma subtypes from mostly primary tumors, including conjunctival and choroidal melanomas, for KIT mutations and amplification[26]. They found one of thirteen conjunctival melanomas to harbor a KIT mutation, but none of the sixty choroidal melanomas tested had a KIT mutation. The study showed 23% of acral lentiginous, 15.6% of mucosal, and 2% of cutaneous (sun-damaged status not provided) melanomas to harbor KIT mutations. This was the first study to report non-point KIT mutations in melanoma. One acral melanoma sample had an in-frame deletion mutation and one rectal mucosal melanoma sample was shown to have an insertion/duplication; both of these gene alterations were in exon 11 of KIT. Quantitative PCR was used to assess KIT copy number in these samples, and showed 27.3% of acral lentiginous, 26.3% of mucosal, and 6.7% cutaneous melanoma samples to have an increased KIT copy number compared to a GAPDH control.</p><p>Satzger et al., examined 37 mucosal melanoma samples for which DNA was available for KIT mutations, and found 11% of head and neck, 30% of genital tract, 12.5% of anal/rectal mucosal melanomas to harbor mutations in KIT[28]. Of 26 evaluable acral lentiginous melanoma samples screened, Ashida et al. found two with KIT point mutations, one of which also showed an increase in copy number as determined by quantitative PCR. There were no KIT mutations found in the three mucosal samples they evaluated. Consistent with other studies, BRAF mutations were in low abundance in acral melanomas and mutually exclusive with KIT mutant containing samples.</p><p>In a recent report from our group at the M. D. Anderson Cancer Center, Torres-Cabala et., showed 12% of acral lentiginous and 17% of mucosal melanoma samples to have mutations in KIT [29]. Two of these mutations were insertions in exon 11 and one primary vulva sample demonstrated KIT point mutations in both exons 13 and 17. In three cases, both primary and metastatic samples from the same patient were available for mutation analysis and the same KIT mutation was demonstrated in both samples.</p><!><p>To date, there are only two reports of cultured cells being generated from patients with either acral lentiginous or mucosal melanoma. Jiang, et al., recently analyzed three low passage primary mucosal cell cultures. One of the three cell cultures demonstrated a highly amplified KIT (exon 11 V559D) mutation without evidence of a wild-type allele by sequencing[30]. The other two cell cultures had wild-type KIT without significant changes in copy number. The mutant/amplified KIT cells showed marked KIT phosphorylation at baseline, consistent with constitutive kinase activity, whereas the wild-type/non-amplified KIT cells did not demonstrate baseline KIT phosphorylation. Imatinib (a.k.a., Gleevec, Novartis Pharma AG) treatment of the mutant/amplified KIT cells resulted in G1 cell cycle arrest, induction of apoptosis and a significant reduction in cell proliferation at nanomolar concentrations. The activity of multiple downstream mediators (p42/44, AKT, MTOR, STAT1, STAT3, P70S6K, and S6K) of KIT was markedly reduced after imatinib treatment. Wold-type/non-amplified KIT cells failed to show any of these changes after imatinib treatment.</p><p>Ashida et al, recently published the analysis of six acral lentiginous cell lines. One of these cell lines was shown to harbor a non-amplified KIT (exon 17 D820Y) mutation[25]. The remaining acral lentiginous cell lines had wild-type/non-amplified KIT genes with varying degrees of protein expression. The KIT D820Y cell line was the only one to demonstrate KIT phosphorylation in the absence of the KIT ligand, consistent with constitutive activity. KIT D820Y is an imatinib resistant mutation, usually arising as a secondary mutation in the setting of imatinib therapy. Treatment of the KIT D820Y cell line with sunitinib (a.k.a., Sutent, Pfizer), which has greater binding affinity for KIT exon 17 mutations, resulted in a modest reduction in cell proliferation, not seen in KIT wild-type acral cell lines when treated with 1 uM of sunitinib.</p><!><p>Although there are an abundance of published reports on the biology of GIST, mastocytosis and leukemia cells with different KIT mutations, little is known about the behavior of genetically altered KIT in melanoma cells.</p><p>Alexeev et al., genetically engineered immortalized mouse melanocytes to express an endogenous, constitutively active KIT D814Y[31]. These cells migrated at a far greater rate in in vitro dual chamber experiments, and when injected into the hypodermis migrated through the dermis to the epidermis. Cells that did not have the KIT D814Y mutation did not demonstrate the same migratory capacity. Curiously, KIT D814Y melanoma cells had reduced cell cycling times and appeared to be less differentiated. Thus the same KIT mutation that drives proliferation in mastocytosis cells, exhibits a migratory phenotype when expressed in melanocytes.</p><p>A recent report by Monsel et al., showed that transfection of immortalized mouse melanocytes with KIT mutants failed to result in transformation of these cells[32]. However, when KIT mutants were expressed in the setting of hypoxia or co-expressed with HIF-1α, the melanocytes were transformed, indicated by colony formation in soft agar. Hypoxia resulted in the marked activation of the MAPK pathway in cells expressing mutant KIT, but not in cells expressing wild-type KIT. Imatinib markedly decreased the MAPK phosphorylation and cell proliferation in cells stably transfected with HIF-1α and KIT K642E or KIT 576del mutants, but not in cells transfected with HIF-1α and wild-type KIT. Also, of note, neither mutant BRAF nor NRAS required hypoxic conditions in order to transform cells and there ability to do so was not enhanced by hypoxia, suggesting very different mechanism of cellular transformation between these known melanoma gene mutations.</p><p>Bougherara et al., used green fluorescent protein KIT mutant chimeras to track the cellular localization of mutant KIT in CHO cells[33]. KIT mutant proteins were phosphorylated, but exhibited an immature glycosylation pattern (high mannose type) and were retained intracellularly. Imatinib treatment of cells expressing KIT V560G resulted in the loss of mutant KIT phosphorylation, conversion from the high mannose type to the mature complex glycosylated form and redistribution to the cell membrane.</p><p>These studies show that the phenotypic expression of KIT mutations may vary under particular cellular and microenvironmental conditions. Although the ongoing clinical trials will provide the response of KIT mutant tumors to TKI treatment, they will not necessarily provide the explanation for that response. As in GIST, cell line generation has been difficult. There is a need to create in vitro or ex vivo systems that will provide workable models for research. Engineering mouse models of mutant KIT melanoma will also likely be important to advance our understanding of the biology of these tumors and their mechanisms of resistance to treatment.</p><!><p>Up to this point, patients with metastatic acral lentiginous and mucosal melanoma have generally been treated with the same regimens used to treat patients with superficial spreading cutaneous melanoma, including high dose bolus interleukin-2 (HD IL-2), chemotherapy, and biochemotherapy[34]. In contrast, small molecule tyrosine kinase inhibitors (TKIs) are the standard of care for GIST, where they produce clinical benefit in the overwhelming majority of patients. Three phase II clinical trials examining the efficacy of imatinib in melanoma were performed in the early 2000s, prior to the identification of KIT mutations in subsets of this disease[35-37]. All three trials failed to show significant responsiveness of metastatic melanomas to imatinib treatment. Of 63 patients treated in these studies, only one was reported to have a clinical response. The responding patient had metastatic acral lentiginous melanoma whose tumor had very high KIT protein expression, but did not demonstrate a KIT mutation in exons 9, 11, 13, 15, or 17. KIT copy number status was not determined. As all these studies enrolled patients without regard to melanoma type, they were highly enriched for patients with the most prevalent form of melanoma, non-chronic sun-damaged, which does not harbor KIT mutations. Even if acral, mucosal, or chronic sun-damaged melanoma patients were enrolled, we now know that only about 15% of these patients would have had KIT mutations. If a KIT genetic aberration is necessary for imatinib response in melanoma, then it is likely that the significant lack of response among the patients treated in these studies is due to their lack of having tumors with KIT genetic aberrations.</p><p>As these early clinical studies with imatinib were coming to completion, Curtin et al, published the identification of KIT genetic aberrations (mutations +/− increased copy number) in acral, mucosal, or CSD melanomas. This observation prompted researchers to examine the tumors of patients with these subtypes and in some cases to treat patients with TKIs that target KIT. Lutzky et al., chronicled the dramatic clinical course of a patient with anal mucosal melanoma with positive inguinal lymph nodes bilaterally[38]. The patient's melanoma tumor was demonstrated to harbor an amplified KIT K642E mutation. The patient underwent wide local excision of the tumor and bilateral lymph node dissection, followed by adjuvant radiotherapy. Shortly after adjuvant treatment, the patient developed multiple subcutaneous nodules in the anogenital/inguinal areas. After four weeks of imatinib treatment there was a complete resolution of the subcutaneous melanoma metastasis. A recurrence of subcutaneous nodules emerged about six months later following a dose reduction of imatinib due to neutropenic fever. A complete clinical response was again achieved after the imatinib dose was increased and endured for 8 months and was present at the time of the article's publication. At about the same time, our group at M. D. Anderson Cancer Center reported a patient with KIT V560D anal melanoma with isolated lung metastases who had a complete response to a temozolomide/sorafenib (Nexavar, Bayer) regimen[39].</p><p>Cases have also been reported in which dramatic responses were achieved in patients with more extensive disease. Hodi, et al. reported a significant clinical response in a patient with a KIT PYDHKWE duplication rectal melanoma that had metastasized to multiple sites[40]. The size and FDG-avidity of pulmonary, epicardial, suprarenal and pelvic metastases were markedly reduced after only 4 weeks of imatinib treatment. Woodman et al, also reported a dramatic reduction in metabolic activity in a KIT L576P vaginal mucosal melanoma that had extensive metastases in the pelvis and inguinal lymph nodes when treated with dasatinib (a.k.a., Sprycel, Brisyol-Myers Squibb) treatment[41]. These case reports suggest that small molecule KIT inhibition has efficacy when used in melanoma patients with KIT mutations.</p><p>There are currently multiple ongoing clinical trials prospectively testing TKIs that target KIT in patients with acral lentiginous, mucosal or chronic sun-damaged skin melanoma. Carvajal, et al., reported interim results at the American Society of Clinical Oncology in October 2009 (abstract ID 9001) from a multi-institutional phase II study of imatinib in stage III or IV patients with somatic alterations in KIT[42]. Of the 12 evaluable patients presented, 2 demonstrated a complete response and 2 showed a partial response. All but two of the remaining patients had stable disease on imatinib. Of note, the two patients who achieved a complete response were the only patients to have both amplification and mutation of KIT, whereas the two patients whose disease progressed despite imatinib treatment had KIT mutations known to be resistant to imatinib in gastrointestinal stromal tumor.</p><p>Hodi, et al. presented an update on a multi-institutional phase II clinical trial of imatinib in melanoma patients with mucosal, acral/lentiginous or chronically sun damaged skin at the International Melanoma Congress in November 1-4, 2009[43]. Of 20 evaluable patients presented, 0 of 10 patients who had wild-type/amplified KIT showed a clinical response, although two of these patients had stable disease for 6-7 months. No complete responses were achieved, but five of 10 patients with KIT mutations demonstrated a partial response to imatinib treatment, three of whom also had amplified KIT.</p><p>The oncology community awaits the final results of these and other trials to better understand the efficacy of TKIs in melanoma subtypes with KIT genetic aberrations.</p><!><p>Recent basic and clinical research has generated great excitement in the melanoma research community. The discovery of oncogenes within subtypes of melanoma has provided promising targets for therapy. Particularly, the identification of KIT genetic aberrations in acral lentiginous, mucosal, and CSD melanoma tumors has allowed for trials to be enriched with patients that have a KIT mutation and/or amplification. The availability of FDA-approved TKIs that inhibit KIT has accelerated the pace with which the prospective trials could be performed. After multiple prior negative clinical trials in unselected melanoma patients, the recent case reports and early results from clinical trials suggest that the currently FDA-approved KIT inhibitors have activity in "KIT-driven" melanoma. While these early observations are very encouraging, definitive answers to many key questions await the maturation of the clinical trials and analysis of additional tumors.</p><p>The efficacy of treating KIT mutant tumors is best described in GIST, where at least 80% of the tumors harbor a KIT mutation. As most of the KIT-mutant tumors respond to KIT inhibition in GIST, it is easy to make the inference that KIT inhibition will be equally efficacious in KIT mutant melanoma. However, KIT mutations in melanoma differ from those in GIST in several respects. First, KIT mutations in melanoma are almost exclusively point mutations, whereas KIT mutations in GIST are predominantly deletion or insertion mutations. Although more data needs to accrue, the first reports testing TKIs on primary human melanoma cells with a KIT V559D and D820Y mutations in vitro show these cells to respond similarly to GIST tumors with these mutations. Second, exon 9 KIT mutations account for 15% of KIT mutations in GIST, but are very rare in melanoma. In contrast, mutations in imatinib resistant residues in exons 13, 17 and 18 account for up to 15% of KIT mutations in melanoma versus less than 1% in GIST. This difference may affect the ultimate percentage of KIT mutant melanomas that respond to TKI therapy, as to date the TKIs in clinical use do not inhibit these mutations. Third, approximately 30% of mutant KIT and wild-type KIT genes in melanoma show increased copy number/amplification, which is a very rare event in GIST. The early clinical trial data suggests that tumors with amplified wild-type KIT are not very sensitive to imatinib treatment. Intriguingly, the only three patients with complete responses with imatinib therapy reported to date had a KIT mutation that was amplified. It is tempting to speculate that tumors that have both a mutant and amplified KIT are exquisitely oncogene addicted and may exhibit the best responses to KIT inhibition. However, definitive conclusions will require the accrual of additional patients. Finally, although secondary KIT mutations are a clear and common mechanism of treatment resistance in GIST, and by inference are likely to occur in melanoma, they have not yet been reported in melanoma. It will be very interesting to evaluate pre- and post-treatment tumor specimens to evaluate the mechanisms of resistance in these patients to see if they are akin to those observed in GIST. How the differences in KIT genetic aberration in melanoma versus GIST will ultimately unfold clinically is an empirical question for which the melanoma research community anxiously awaits the answer.</p><p>Apart from the differences in the KIT gene itself, there may also be significant differences in the cellular milieus of melanoma tumors that alter the behavior of KIT mutant melanoma cells and their response to KIT inhibition. Multiple studies have shown that KIT mutant proteins have different signaling pathways depending on the cellular context. The report by Alexeev et al., showing a KIT mutation to generate a motogenic phenotype in melanoma versus a mitogenic phenotype in mastocytosis cells suggests that the cellular environment in which the KIT mutation occurs can markedly effects its function. Comparing genomic and proteomic profiles between KIT mediated tumor types may provide for a better understanding of common and disparate signaling networks in different KIT mutant tumor types and the mechanisms by which cells are either sensitive or resistant to KIT inhibition.</p><p>KIT mutant tumor progression on imatinib treatment has already been reported and may be inevitable in most patients. To date, studies that have been enriched for KIT mutant melanoma patients are testing TKIs directed against the kinase domain of the molecule. Currently, the TKIs that are FDA-approved target multiple kinases, including KIT, and are approved for Philadelphia positive CML, KIT mutant positive GIST and some types of renal cancers. No TKI is yet FDA-approved for KIT-mutant melanoma patients. It is will be interesting to see if the TKIs that are being tested in KIT-mutant melanomas – imatinib, dasatinib, sunitinib, and nilotinib (a.k.a., Tasigna, Novartis) – will have significant differences in efficacy in this disease, or if the activity of the agents will mostly reflect the domain in which the KIT mutation occurs (like in GIST). Other possible therapies should be explored as well, including KIT antibody directed treatment, KIT RNA interference strategies, and combinatorial therapies which couple targeted therapies to KIT with chemo and/or immuno-therapies.</p><p>It is an exciting time to be a clinician and researcher in melanoma. The era of being able to molecularly categorize melanoma patients and align them with therapies that more directly treat the underlying mechanism of their tumor pathology is upon us. Hopefully such personalized approaches will lead to rational and more effective treatments that improve outcomes in this challenging disease.</p>

PubMed Author Manuscript

Histone Deacetylase 1/2 Mediates Proliferation of Renal Interstitial Fibroblasts and Expression of Cell Cycle Proteins

We recently reported that the histone deacetylase (HDAC) activity is required for activation of renal interstitial fibroblasts. In this study, we further examined the role of HDACs, in particular, HDAC1 and HDAC2, in proliferation of cultured rat renal interstitial fibroblasts (NRK-49F) and expression of cell cycle proteins. Inhibition of HDAC activity with trichostatin A (TSA), blocked cell proliferation, decreased expression of cyclin D1, a positive cell cycle regulator, and increased expression of p27 and p57, two negative cell cycle regulators. Silencing either HDAC1 or HDAC2 with siRNA also significantly inhibited cell proliferation, decreased expression of cyclin D1, and increased expression of p57. However, down-regulation of HDAC2, but not HDAC1 resulted in increased expression of p27. Furthermore, HDAC1 and HDAC2 downregulaton was associated with dephosphorylation and hyperacetylation of STAT3 (signal transducer and activator of transcription 3). Blockade of STAT3 with S3I-201 or siRNA decreased renal fibroblast proliferation. Finally, mouse embryonic fibroblasts (MEFs) lacking STAT3 reduced the inhibitory effect of TSA on cell proliferation, add-back of wild type STAT3 to STAT3\xe2\x88\x92/\xe2\x88\x92 MEFs restored the effect of TSA. Collectively, our results reveal an important role of HDAC1 and HDAC2 in regulating proliferation of renal interstitial fibroblasts, expression of cell cycle proteins and activation of STAT3. Further, STAT3 mediates the proliferative action of HDACs.

histone_deacetylase_1/2_mediates_proliferation_of_renal_interstitial_fibroblasts_and_expression_of_c

Introduction<!>Materials<!>Cell culture<!>Cell proliferation<!>Transfection of siRNA or plasmisds into cells<!>Western blot analysis<!>Immunofluorescent analysis<!>Statistical analysis<!>HDAC activity is required for the proliferation of renal interstitial fibroblasts and the expression of cell cycle proteins<!>Expression of HDAC1 and HDAC2 in renal interstitial fibroblasts<!>HDAC1 and HDAC2 knockdown decreases the proliferation of renal interstitial fibroblasts<!>HDAC2 and HDAC1 knockdown differentially regulates the expression of cycle proteins<!>Serum-induced STAT3 phosphorylation is regulated by HDAC1 and HDAC2<!>STAT3 mediates proliferation of renal interstitial fibroblasts<!>STAT3 is required for HADC-elicited proliferation of renal interstitial fibroblasts<!>Discussion

<p>Chronic kidney disease (CKD) is characterized by accumulation of a large population of myofibroblasts and excessive amount of extracellular matrix (ECM) [Deelman and Sharma, 2009; Wynn, 2008]. As the myofibroblast is a principal effector cell type responsible for production of ECM protein under various kidney diseases associated with renal interstitial fibrosis, Identification of the major mediator(s) that controls cell proliferation would offer new therapeutic approaches for halting progression of renal disease.</p><p>Increasing evidence indicates that activation of histone deacetylases (HDACs) is required for the regulation of gene expression and cell proliferation [Mai et al., 2005]. HDACs are a group of enzymes that catalyze the deacetylation of lysine residues in histones and in a number of non-histone proteins associated with cell proliferation such as tyrosine kinase and transcriptional factor 3 (STAT3) [Yuan et al., 2005]. In mammals, there are 18 HDACs, which are grouped into four classes. Class I HDACs include HDAC1, −2, −3, and −8; class II HDACs include HDAC4, −5, −6, −7, −9, and −10; class III HDACs are known as sirtuins; and class IV HDAC is HDAC11 [Beumer and Tawbi,; Bieliauskas and Pflum, 2008]. Among HDACs, HDAC1 and HDAC2 have been reported to require for organ development and proliferation of embryonic stem cells and embryonic fibroblasts [Zimmermann et al., 2007; Zupkovitz et al., 2006]. HDAC inhibitors, such as trichostatin A (TSA), inhibit the activity of both class I and class II HDACs and are effective in suppressing cell cycle progression and cell proliferation in numerous cell types [Bieliauskas and Pflum, 2008]. Recent studies showed that treatment with SK-7041, a selective inhibitors of class I HDACs [Hinkle et al., 2004], had a similar anti-proliferative effect as the class I/II HDAC inhibitors did [Lee et al., 2006], suggesting that class I HDACs play a dominant role in the progression of fibrosis.</p><p>Cellular proliferation is an ordered, tightly regulated process involving growth factor-controlled expression of multiple cell cycle proteins. Of them, cyclins and their catalytic partners cyclin-dependent kinases (CDKs) are positive regulators [Golias et al., 2004] and the cyclin kinase inhibitors (CKIs) such as p27 and p57 are negative regulators [Golias et al., 2004]. Whereas cyclin D1 is critically involved in promotion of G1-S phase cell cycle transition, p27 and p57 inhibit the cell cycle at multiple checkpoints through inactivation of cyclin-CDK complexes [Golias et al., 2004]. As such, increased expression of cyclins and decreased expression of p27 and p57 are associated with cell proliferation. It has been documented that inhibition of the HDAC activity with TSA decreased expression of cyclin D1 [Alao et al., 2004] and increased expression of p57 and results in cell cycle arrest [Yamaguchi et al.]. Furthermore, HADC1 and HDAC2 can be directly bound to the promoter regions of the p57 gene and suppresses their expression [Yamaguchi et al.], suggesting that HDAC1 and HDAC2 can directly regulate expression of some CKIs. While it is unknown whether HDACs directly regulate the expression of cyclin D1, our recent studies showed that inactivation of the HDAC activity with TSA blocked phosphorylation/activation of STAT3, a transcriptional factor that drives expression of cyclin D1 [Pang et al., 2009], suggesting that HDACs may regulate cyclin D1 through an indirect mechanism involving activation of STAT3.</p><p>STAT3 belongs to a family of tyrosine kinase and transcriptional factors that consists of seven members (STAT1-4, 5a/b, 6) [Bowman et al., 2000; Song and Grandis, 2000]. In response to a variety of growth factors and hormones, STATs are phosphorylated, dimerized and then translocated to the nucleus, where they binds to the DNA promoter region of target genes to drive expression of many genes [Rossi et al., 2007]. In addition to tyrosine-phosphorylation, STAT3 and some other STAT proteins such as STAT1 are also subjected to regulation by acetylation [Pang et al., 2009; Yuan et al., 2005]. While the functional significance of STAT3 acetylation remains poorly understood, it was reported that acetylation of STAT1 counteracts its activity by forming a complex with T-cell protein tyrosine phosphatase (TCP45) [Kramer et al., 2009].</p><p>Currently, little is known about the role of HDACs in the regulation of fibroblast proliferation and expression of cell cycle regulators. The purpose of this study aimed to determine the role of the class I HDACs, in particular, HDAC1 and HDAC2 in these processes. Furthermore, we examine the role of STAT3 in HDAC-mediated cell proliferation.</p><!><p>Trichostatin A was purchased from Biomol (Plymouth Meeting, PA). Antibodies to phospho-STAT3 (Tyr-705), STAT3 and cyclin D1 were purchased from CellSignaling Technology (Danvers, MA). Antibodies to HDAC1, HDAC2, p27, p57 and GAPDH were obtained from Santa Cruz Biotechnology,Inc. (Santa Cruz, CA). Pre-designed siRNAs targeting HDAC1 and HDAC2 and silencer negative control siRNA were purchased from Invitrogen. pcDNA3 empty vector, STAT3 plasmid (pcDNA3-STAT3) and anti-acetyl-STAT3 (K685) were obtained from Dr. Eugene Chin (Brown University). Antibodies to α-actin and all other chemicals and reagents werepurchased from Sigma (St. Louis, MO).</p><!><p>Normal rat renal interstitial fibroblasts (NRK-49F) and STAT3−/− mouse embryonic fibroblasts (MEFs) were grown in DMEM (Sigma-Aldrich, St. Louis, MO) containing 5% fetal bovine serum (FBS), 0.5% penicillin and streptomycin in an atmosphere of 5% CO2 air at 37°C. For proliferation study, cells were seeded in 12-well plates. 30–40% confluent cells were starved for 24 h in DMEM with 0.5% FBS prior to receiving stimulation with 5% FBS. If necessary, various pharmaceutical inhibitors were added to the culture and then incubated for an additional 24 or 48 h. Control cells were treated with an equivalent amount of vehicle.</p><!><p>Cell proliferation was assessed by counting cells, the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay and BrdU incorporation assay. For counting cell numbers, at least five pictures from each well of a 12-well plate were taken and cell numbers in five wells were counted for a sample under the microscope. For the MTT assay, MTT was added (final concentration, 0.5 mg/ml) to individual cultures for 1 h at the end of experiments. Tetrazolium was released by dimethyl sulfoxide, and the optical density was determined with a Spectramax M5 plate reader (Molecular Devices Corporation, Sunnyvale, CA) at 570 nm. For the BrdU incorporation assay, a colorimetric BrdU cell proliferation enzyme-linked immunosorbent assay kit (Roche Applied Science, Penzberg, Germany) was used according to our previous protocols [Xing et al., 2008]. Briefly, BrdU labeling solution was added to cells and then incubated for 4 h. At the end of incubation, the labeling medium was removed, the cells were fixed, and the DNA was denatured. After addition of the anti-BrdU-peroxidase conjugate, the immune complexes were detected by subsequent reaction with tetramethylbenzidine as substrate for 20 min. The reaction product was quantified using a Spectramax M5 plate reader.</p><!><p>The siRNA oligonucleotides targeted specifically to HDAC1 or HDAC2 (200 pmol) were transfected into NRK-49F cells (1×106) using the Amaxa Cell Line Nucleofector Kit T (Lonza Cologne AG, Cologne, Germany) and the Amaxa Nucleofector device according to the manufacturer's instructions (Gaithersburg, MD). In parallel, 200 pmol silencer negative control siRNA was used for off-target changed in NRK-49F cells. pcDNA3 empty vector and STAT3 plasmid (pcDNA3-STAT3) were transfected by lipofectamine 2000. After transfection, cells were plated and cultured for 48 h in DMEM before cell lysates were prepared for immunoblot analysis.</p><!><p>Cell samples were prepared in lysisbuffer (Cell Signaling Technology, Dancers, MA) containing protease inhibitors cocktail (Roche Diagnostic Co. Indianapolis, IN), 1 mM phenylmethanesulphonylfluoride, and phosphatase inhibitors (2 mM Na3VO4). After sonication and centrifugation, supernatants were collected, and the protein concentration was determined using the BCA protein assay kit (Pierce, Rockford, IL). Twenty micrograms of proteinfrom each sample was separated by SDS-PAGE and transferred to a polyvinylidene difluoride (PVDF) membrane (Millipore, Bedford, MA). After treatment with 5% skim milk at 4°C overnight, membranes were incubated with various antibodies for 1 h or longer time and then incubated with an appropriate horseradish peroxidase-conjugated secondary antibody (Santa Cruz Biotechnology Inc., Santa Cruz, CA). Bound antibodies were visualized following chemiluminescence detection on autoradiographic film. The densitometry analysis of immunoblot results was conducted using Image J software (http://rsb.info.nih.gov/ij) based on the manufacture's instruction. Briefly, after development, the film was scanned to obtain the digital image. The quantification of immunoblot is based on the intensity (density) of band, which is calculated by area and pixel value of the band. The quantification data are given as ratio between target protein and loading control.</p><!><p>Cells cultured on cover slips were washed with phosphate buffered saline (PBS), fixed with 4% paraformaldehyde, permeabilized with 0.1% (vol/vol) TritonX-100 and 0.1 mM glycine, and then incubated 30 min in PBS containing 5% serum. Cells were then treated with primary antibodies at room temperature for 1 hr. After washing with PBS, cells were incubated with a mixture of FITC-labeled goat anti-rabbit IgG antibody for 1 h at room temperature. Morphological analysis was performed by using light and fluorescent microscopy.</p><!><p>Data are presented as means ± SD and were subjected to one-way ANOVA. Multiple means were compared by using Tukey's test. The differences between two groups were determined by Student t-test. P < 0.05 was considered as statistically significant difference between mean values.</p><!><p>To determine the role of HDACs in proliferation of renal interstitial fibroblasts, we treated NRK49-F cells with TSA, a potent and reversible inhibitor of HDACs [Monneret, 2005], in the complete medium containing 5% FBS. Cell proliferation was assessed by directly counting the number of cells and measuring by the MTT as well as BrdU incorporation assays. Treatment with TSA significantly decreased proliferation of NRK49-F (Figure 1A, B, D). This inhibitory effect of TSA occurred in a dose dependent manner with the significant effect at 25 nM and the maximum inhibition at 100 nM (Figure 1C). These data, together with our recent results showing that 25–100 nM TSA can inhibit the deacetylase activity of HDAC [Pang et al., 2009], suggest that the HDAC activity is required for the proliferation of renal interstitial fibroblasts.</p><p>Cell proliferation is controlled by expression of cell cycle proteins. Up-regulation of cycle D1 and down-regulation CDK inhibitors such as p27 and p57 are associated with the progression of the cell cycle [Golias et al., 2004]. To understand whether activation of HDACs contributes to the expression of those proteins, NRK-49F were incubated with 100 nM TSA for 12, 24 and 48 h, and then harvested for immunoblot analysis of cyclin D1, p27 and p57. As shown in Figure 1E, the expression level of cyclin D1 was seen at 12 h following TSA treatment. At 48 h, cyclin D1 was no longer detectable. In contrast, p27 and p57 expression was increased over time with the maximum at 48 h. These data illustrated that all of those cycle proteins are subjected to regulation by HDACs in renal interstitial fibroblasts.</p><!><p>Recent studies showed that SK-7041, a selective inhibitor of class I HDACs, exhibited an anti-proliferative effect similar to that of the class I/II HDAC inhibitors did [Lee et al., 2006], and that HDAC1 and HDAC2 are required for organ development and proliferation of embryonic stem cells and embryonic fibroblasts [Zimmermann et al., 2007; Zupkovitz et al., 2006]. As a first step towards understanding the role of HDAC1 and HDAC2 in renal fibroblast proliferation, we examined HDAC1 and HDAC2 expression in NRK-49F cells. Immunoblot analysis indicated that HDAC1 and HDAC2 were abundantly expressed in renal interstitial fibroblasts (Figure 2A). Immunofluorescent staining demonstrated a nuclear localization for HDAC1 and both nuclear and cytosolic distribution for HDAC 2 in serum-starved NRK-49F cells. After treatment with 5% FBS, both HDAC1 and HDAC2 were only observed in the nucleus (Figure 2B). These data suggest that HDAC1 and HDAC2 are expressed in renal interstitial fibroblasts and that stimulation with mitogens can induce translocation of HDAC2 from the cytosol to the nucleus.</p><!><p>To determine the role of HDAC1 and HDAC2 in renal fibroblast proliferation, the siRNA approach was used. NRK-49Fcells were transfected with siRNA specifically targeting HDAC 1 and HDAC2. At 48 h after transfection, cell proliferation was analyzed by both cell counting and the MTT assay. Figure 3A and 3B showed that knockdown of either HDAC1 or HDAC2 significantly inhibited proliferation of NRK-49F. Immunoblot analysis indicated that expression levels of HDAC 1 and HDAC2 are inhibited by 70%, 75%, respectively, in cells transfected with HDAC 1 or HDAC2 siRNA relative to those transfected with scrambled siRNA. These data suggest that both HDAC1 and HDAC2 are involved in the regulation of renal fibroblast proliferation.</p><!><p>We further examined the effect of HDAC1 and HDAC2 down-regulation on the expression of cyclin D1, p27, and p57 using siRNA approach. Figure 4 showed that knockdown of either HDAC1 or HDAC2 decreased the expression of cyclin D1 and increased expression of p57. However, increased protein abundance of p27 was only observed in cells with silencing of HDAC2. Therefore, we suggest that expression of cyclin D1 and p57 are under control by both HDAC1 and HDAC2, and expression of p27 is only regulated by HDAC2.</p><!><p>Our recent studies showed that inhibition of HDACs by TSA inhibits STAT3 phosphorylation [Pang et al., 2009]. Here we further examined the effect of HDAC 1 and HDAC2 down-regulation on the phosporylation of STAT3 at Tyr705, an active form of STAT3. As shown in Figure 5, knockdown of either HDAC 1 or HDAC2 reduced STAT3 phosphorylation and increased its acetylation levels. Interestingly, down-regulation of these two HDAC isozymes also decreased expression of total STAT3. However, the densitometry analysis on immunoblots showed that cells transfected with either HDAC1 or HDAC2 siRNA have a lower ratio of p-STAT3/total STAT3 compared with that in the scramble siRNA transfected cells, suggesting that p-STAT3 reduction in cells with knocking down of HDAC1 or HDAC2 siRNA is at least in part due to alteration in the rate of phosphorylation. Collectively, these data suggest that both HDAC1 and HDAC2 are required for regulation of STAT3 phosphorylation, acetylation, and expression in renal interstitial fibroblasts.</p><!><p>To determine the role of STAT3 in proliferation of renal fibroblasts, we first examined the effect of S3I-201, a specific inhibitor of STAT3 on the proliferation of NRK-49F. S3I-201 inhibited proliferation of NRK-49F in dose and time dependent manners with initial inhibition at 25 µM and maximal inhibition at 100 µM (Figure 6A, B). Our recent studies have demonstrated that S3I-201 completely blocked STAT3 phosphorylation at 100 µM [Pang et al.]. Next, we examined the effect of siRNA-mediated STAT3 knockdown on the proliferation of renal interstitial fibroblasts. Transfection of STAT3 siRNA led to a significant decrease in the expression of STAT3 (Figure 6E and F) and proliferation of NRK-49F (Figure 6C and D). However, expression levels of HDAC1 and HDAC2 were not affected by STAT3 down-regulation (Figure 6C and D). These results suggest that STAT3 may mediate the proliferation of renal interstitial fibroblasts by acting downstream of HDAC1 and HDAC2.</p><!><p>If STAT3 acts downstream of HDACs to mediate renal fibroblast proliferation, the inhibitory effect of TSA on renal fibroblasts should be decreased in cells lacking STAT3 and add-back of STAT3 should restore the inhibitory ability of TSA. To test this hypothesis, we treated the wild type (STAT3+/+) or STAT3 deficient (STAT3−/−) mouse embryonic fibroblasts (MEFs) with TSA and then determined their proliferation by the MTT assay. TSA decreased cell proliferation by 52% in STAT3+/+ MEFs. In contrast, TSA only decreased proliferation by 25% in STAT3−/− cells (Figure 7A). As expected, re-introduction of wild type STAT3 to STAT3−/− cells led to the restoration of the ability of TSA to inhibit the proliferation of MEFs (Figure 7B). Similar results were obtained when cell numbers were counted (data not show). The expression level of total STAT3 in STAT3−/− MEFs after STAT3 transfection was the similar to that in STAT3+/+ MEF (Figure 7C and D). These data indicated that TSA-mediated inhibition of proliferation depends upon the presence of STAT3 in fibroblasts.</p><!><p>In this report, we demonstrated that either inhibition of the HDAC activity with TSA or specific silencing of major class 1 HDAC1 and HDAC2 enzymes with siRNA decreased proliferation of cultured renal interstitial fibroblasts and expression of cell cycle proteins. Moreover, we showed that both HDAC1 and HDAC2 mediates activation of STAT3 and that STAT3 plays an essential role in mediating the action of HDACs. These results, together with our recent studies showing the requirement of the HDAC activity in activation of renal interstitial fibroblasts [Pang et al., 2009], suggest that HDACs, particularly of HDAC1 and HDAC2, are involved in the development of renal interstitial fibrosis.</p><p>Although all the class I and II HDAC isoforms are sensitive to TSA, HDAC1 and HDAC2 are reported to be the major HDACs implicated in tissue development and cell proliferation. Genetic studies revealed that gene disruption of HDAC1 led to severe developmental defects and reduced proliferation both in the mouse embryo and in embryonic stem (ES) cells [Zupkovitz et al., 2006]. Inactivation of HDAC2 also delayed growth in mice, leading to smaller and less fertile animals than wild-type and heterozygous littermates [Zimmermann et al., 2007]. Recently, Yamaguchi et al., demonstrated that both HADC1 and HADC2 are required for the proliferation of mouse embryonic fibroblasts [Yamaguchi et al.]. Our studies indicated that HDAC1 and HDAC2 are abundantly expressed in cultured renal fibroblasts and that silencing either of them significantly decreases proliferation of NRK-49F, which supports involvement of these two HDAC isoforms in renal interstitial fibroblasts.</p><p>HDAC-mediated proliferation of renal fibroblasts is associated with down-regulation of CKIs (i.e. p27 and p57) and up-regulation of cyclin D1. In this study, we examined the effect of HDAC1 and HDAC2 down-regulation on the expression of cyclin D1, p27 and p57 in renal fibroblasts. Although knockdown of either HDAC1 or HDAC2 significantly increased expression of p57, sliencing of HDAC2, but not HDAC1 resulted in an increased protein level of p27. In addition to CKD inhibitors, siRNA-mediated silencing of HDAC1 and HDAC2 decreased protein abundance of cyclin D1, a key regulator that promotes S phase entry in the cell cycle. In line with our observation, HDAC1 and HDAC2 have been reported to promote the G1-to-S-phase transition by inhibiting the expression of p57 in MEFs [Yamaguchi et al.], and inhibition of the HDAC activity with TSA decreased the expression of p27 in rat mesangial cells [Freidkin et al.]. Since cyclin D1 is a positive regulator of the cell cycle whereas p27 and p57 are negative regulators of the cell cycle [Golias et al., 2004], the coordinate regulation of expression of those proteins by HDAC1 and HDAC2 may accelerate the cell cycle progression in renal fibroblasts.</p><p>Currently, the mechanism by which HDAC1 and HDAC2 differently regulates p27 expression remains unclear. A previous study showed that the enforcement of elevated p27 protein levels by HDAC-inhibitors is not due to changes in the level of p27 transcript but associated with the post-translational mechanism, mainly p27 stability [Chen and Faller, 2005]. As p27 stability is typically regulated by proteolytic degradation through the ubiquitin-proteasome pathway [Freytag and Geddes, 1992; Hengst and Reed, 1996; Shirane et al., 1999; Vlach et al., 1997], it is possible that HDAC2, but not HDAC1, plays a role in regulating activation of this pathway. This hypothesis will be tested in our future studies.</p><p>An interesting observation is that knocking down of either HDAC1 or HDAC2 reduces STAT3 phosphorylation at Tyr 705 in renal fibroblasts. As STAT3 phosphorylation at this residue is required for its dimerization and activation, we suggest that these HDAC isoforms play an important role in regulation of STAT3 activation. Furthermore, as cyclin1 is one of the STAT3 transcriptional targets, it is possible that HDACs-mediated STAT3 activation is functionally linked to proliferation of renal fibroblasts. This hypothesis is supported by our observations that inhibition of STAT3 with S3I-201 or specific siRNA decreased proliferation of renal interstitial fibroblasts and that deletion of STAT3 alleviated the inhibitory effect of TSA on cell proliferation in MEFs whereas re-introduction of STAT3 to STAT3−/− MEFs restored the ability of TSA to inhibit fibroblast proliferation. Therefore, it seems that STAT3 acts as an intermediary molecule in HDACs regulation of cyclin D1 expression. Nevertheless, the STAT3 signaling pathway may not be the sole one that mediates the expression of cyclin1 by HDACs. Previous studies have indicated that TSA–mediated suppression of cyclin D1 expression is involved in the decreased NF-kappaB/DNA binding [Hu and Colburn, 2005] and ubiquitin-dependent degradation [Alao et al., 2006]. Further studies are necessary to elucidate the molecular basis by which STAT3 coordinates with NF-kappaB/p65 and ubiquitins to regulate cyclin D1 expression.</p><p>The detailed mechanism by which HDAC1/2 inhibition reduces STAT3 tyrosine phosphorylation remains poorly understood, but may be associated with acetylation. Our previous and current studies showed that dephosphorylation of STAT3 by TSA or siRNAs targeting HDAC1 or HDAC2 was accompanied by STAT3 hyperacetylation. Similar reciprocal alterations in tyrosine phosphorylation and acetylation were also observed in STAT1 after inhibition of the HDAC activity in a variety of cell types [Klampfer et al., 2004; Nusinzon and Horvath, 2003]. Recently, Krämer et al. reported that STAT1 acetylation can induce its binding to T-cell protein tyrosine phosphatase (TCP45), which catalyzed STAT1 dephosphorylation [Kramer et al., 2009]. On this basis, they proposed a model to explain the mechanism by which STAT1 acetylation counteracts its phosphorylation: Acetylation of STAT1 leads to the formation of a complex with the T-TCP45 and the highly active PTP TCP45 acts as a "transmission control protein" docking to and inhibiting previously activated STAT1 [Kramer et al., 2009]. Whether this functional phospho-acetyl switch, regulated by an acetylation/deacetylation balance, also modulates STAT3 phosphorylation needs further investigation.</p><p>In summary, we demonstrated that HDAC1 and HDAC 2 play an important role in regulating proliferation of renal interstitial fibroblasts and expression of multiple cell cycle proteins. Further, HDAC1 and HDAC 2 contribute to modulation of STAT3 tyrosine phosphorylation/activation and STAT3 mediates the proliferative effect of HDACs. Inhibition of HDAC1/2 promotes STAT3 acetylation. Acetylated STAT3 leads to its dephosphorylation/inactivation and subsequently antagonizes transcriptional regulation of the target genes associated with cell cycle regulation and proliferation (Figure 8). These findings suggest that targeting HDAC1/2 may be a promising strategy to attenuate progression of kidney diseases associated with renal fibrosis.</p>

PubMed Author Manuscript

On the Mechanism of Electrochemical Generation and Decomposition of Phthalimide N-oxyl (PINO)

Phthalimide N-oxyl (PINO) is a potent hydrogen atom transfer (HAT) catalyst that can be generated electrochemically from N-hydroxyphthalimide (NHPI). However, catalyst decomposition has limited its application. This paper details mechanistic studies of the generation and decomposition of PINO under electrochemical conditions. Voltammetric data, observations from bulk electrolysis, and computational studies suggest two primary aspects. First, base-promoted formation of PINO from NHPI occurs via multiple-site concerted proton-electron transfer (MS-CPET). Second, PINO decomposition occurs by at least two second-order paths, one of which is greatly enhanced by base. Optimal catalytic efficiency in PINO-catalyzed oxidations occurs in the presence of bases whose corresponding conjugate acids have pKas in the range of 12-15, which strike a balance between promoting PINO formation and minimizing its decay.

on_the_mechanism_of_electrochemical_generation_and_decomposition_of_phthalimide_n-oxyl_(pino)

■ INTRODUCTION<!>■ RESULTS<!>■ DISCUSSION<!>■ CONCLUSION

<p>Electrocatalysts that promote anodic half reactions have been specifically developed for oxidative organic transformations due to longstanding industrial interest. 1 The use of oxidative electrocatalysts, shuttling electrons between molecular substrates and anode interfaces, have realized chemical transfromations in an environmentally friendly manner, as electrons are collected by electrode interfaces instead of stoichiometric chemical oxidants. 2 For example, the electrochemically-generated N-oxoammonium ion (TEMPO + ) derived from TEMPO serves as an active catalytic species and readily reacts with alcohols to produce ketones. 3 Both TEMPO and TEMPO + salt are isolable, which facilitates mechanistically-guided development of this and related redox mediators for selective and efficient oxidations. 3 PINO is another promising electrocatalyst generated by the facile oxidation of its precursor NHPI (Figure 1A). 3 PINO can abstract allylic 4 and benzylic [5][6][7] hydrogen atoms, as well as other relatively weak C-H bonds, 8 making it a potent HAT electrocatalyst. However, PINO is short-lived, which has caused a limited operational utility for methodology development due to PINO decomposition. 8 The practical consequence of this decomposition is the need to use NHPI at high catalyst loading (10-20 mol%) to achieve efficient reactivity, as exemplified in the application to the oxidation of β-O-4 linkages in lignin. 6 Furthermore, the short lifetime of PINO has made its mechanistic studies challenging. 3 Despite the interest in the NHPI/PINO couple as a valuable redox mediator, 3 a full assessment of off-cycle (i.e., PINO decomposition) pathways has not been performed. PINO decomposition under electrochemical conditions was first reported by Masui and co-workers in 1987. 9 A trimeric species (hereafter trimer) was isolated and characterized as the major PINO decomposition product after bulk electrolysis of NHPI in the presence of pyridine (Figure 1B). The formation of the trimer by nucleophilic attack of NHPI upon its oxoammonium ion was suggested (Figure 1C, path a). Later, Pedulli and co-workers proposed a separate radical pathway involving coupling between PINO and the acyl radical arising from C-N bond fragmentation of another PINO (Figure 1C, path b). 10 Both decomposition pathways have been presumed as relevant in the literature without detailed scrutinity. Advancing the NHPI/PINO redox mediator as an efficient catalytic system requires disambiguation of the PINO decomposition mechanisms.</p><p>To enable development of new strategies for selective electrochemical oxidation using NHPI or similar species, we have examined elementary aspects of PINO generation, and ascertained how it participates in oxidation of benzylic alcohols using electroanalytical and computational methods. The data indicate a multiple-site concerted proton-electron transfer (MS-CPET) mechanism of PINO generation from NHPI, where the proton and electron move to different locations. 11 Moreover, it suggests that in lieu of the two aforementioned mechanistic proposals for the decomposition of PINO, electrochemically-generated PINO undergoes two distinct second-order decay processes, one of which is promoted by base. The influence of base strength on the catalytic efficiency of benzylic alcohol oxidation is quantitively analyzed. Taken together, this study provides a foundation for the development of more effective N-oxyl electrocatalysts.</p><!><p>Methods and Materials. The study was conducted in the sequence of analyzing the redox response of NHPI with and without bases, then NHPI-base complex, and ascertaining the base activity on substrate oxidation, finally simulating PINO decomposition mechanisms computationally. All experiments were performed in anhydrous acetonitrile (MeCN) due to its high dielectric constant, inertness towards NHPI and PINO, and comparatively high solubility of NHPI relative to other solvents. The energetics associated with NHPI complexation, electrochemical oxidation and possible mechanisms of PINO decomposition were assessed using density functional theory. Unless otherwise indicated, the calculations were carried out using the B3LYP functional with 6-311+G(d) basis set and implicit solvation (IEFPCM) parameterized for MeCN.</p><p>Redox Response of NHPI with/without Bases. The redox response of NHPI at a glassy carbon electrode was measured in anhydrous MeCN with recrystallized tetrabutylammonium hexafluorophosphate (NBu4PF6). 12 Figure 2A shows the voltammetric response of NHPI without base in the electrolyte. In contrast to numerous previous reports, including our own, 4-7,13-14 a reversible voltammetric response was not observed. Instead, NHPI could not be readily oxidized at 0.1 Vs -1 in this potential range under anhydrous conditions. Moreover, the lack of cathodic current on the return sweep suggested oxidized NHPI (NHPI •+ ) was unstable on this timescale. The redox response of NHPI oxidation speaks aganist a common observation that NHPI can be reversibly oxidized both with and without a Brønsted base. [4][5][6][7][13][14] Variation in the electrolyte did not affect the voltammetric response for NHPI unless care was not taken to exclude water (Supporting Information, SI). Adventitious and/or intentional addition of water to the electrolyte caused the voltammetric response to appear quasi-reversible, implicating water as an active agent during the oxidation of NHPI and suggesting similar wet conditions were likely used in prior reports of electrochemical oxidation of NHPI without added base. [4][5][6][7][13][14] In fact, we discovered that leakage of water (H2O) from an aqueous reference electrode was sufficient to cause water contamination. The formal standard potential for NHPI oxidation in the presence of added H2O, as determined from the midpoint of the redox waves (E 0' ≈ E1/2), was a function of the concentration of added water (SI). When one equivalent of water was used, E1/2 = +1.04 V vs E 0' (Fc + /Fc) (Figure 2A). 15 Cyclic voltammetric responses for the oxidation of NHPI in the presence of 1 equivalent of 2,6-lutidine were collected from 0.01 Vs -1 to 0.5 Vs -1 (Figure 2B). Normalizing the current densities in Figure 2C by √𝑣 highlighted the fact that the voltammetric response shape was sensitive to the experimental timescale, indicating the oxidation of NHPI was coupled with at least one additional chemical reaction. At scan rates ≤ 0.01 Vs -1 , the voltammetric response became sigmoidal, attaining a steady-state current density (Figure 2C) that suggested redox active species were being generated at the electrode by follow-up reactions in solution. A steady-state current density with the same magnitude was observed consistently for bases with pKa values < 15 and was insensitive to base concentration. (SI)</p><p>Voltammetric measurements were performed with anodically activated glassy carbon 16 to examine the possiblecontribution of inner-sphere effects in the apparent redox responses. However, this deliberate change in electrode surface chemistry did not impact the primary voltammetric features, as the observed peak splitting was invariant with the nature of the glassy carbon surface (SI).</p><p>NHPI-Base Complex. Identification of any possible reactions between NHPI and base was performed first, as most of PINO-catalyzed oxidations were operated in the presence of bases. When one equivalent of 2,6-lutidine was used as base, a quasi-reversible voltammogram was obtained with E1/2 = +0.39 V vs E 0' (Fc + /Fc) (Figure 2A). The presence of base decreased the standard potential and increased the reversability of NHPI oxidation. The reported pKa values of NHPI (23.5) 11 and the 2,6-lutidinium ion (14.1) 17 in MeCN suggest that proton exchange is highly unfavourable. Nevertheless, the plausitibility of a hydrogen bond formation between NHPI and pyridine derivative was assessed through IR spectroscopy. NHPI in MeCN shows a characteristic broad O-H stretch at 3210 cm -1 (Figure 3A). Addition of 2,6-lutidine caused a decrease in this absorption and the appearance of a very broad absorption at ~2450 cm -1 , corresponding to the O-H stretch of the hydrogen-bonded complex. 18 The cumulative spectra were consistent with an equilibrium constant Keq = 11.8 M -1 for a hydrogen-bonded complex (Figure 3A).</p><p>To determine the stoichiometry of the hydrogen-bonded complex, cyclic voltammetric titration experiments were conducted at 0.1 Vs -1 (Figure 3B). A solution of NHPI in MeCN was titrated with 2,6-lutidine. The quasi-reversible voltammetric response at ~+0.4 V vs E 0' (Fc + /Fc) corresponding to the oxidation event of NHPI-lutidine complex grew in magnitude as the concentration of 2,6-lutidine increased, with a concomitant disappearance of the voltammetric response observed in the absence of a base. The anodic peak current density for the oxidation of the complex was saturated until one equivalent of 2,6-lutidine was added, suggesting NHPI and 2,6-lutidine form a 1:1 complex in solution (Figure 3B). The E1/2 was dependent upon the A set of calculations on the energies for both an ion pair of deprotonated NHPI and pyridinium and for a hydrogenbonded complex was performed (Figure 3C). The free energy change regarding the formation of the 1:1 hydrogenbonded complex between NHPI and pyridine was calculated to be -2.3 kcal/mol, rendering a formation constant (Kf) of 47.9. In reasonable agreement with this computation, Abraham's hydrogen-bonding parameters determined for pyridine (𝛽 = 0.62) 19 and NHPI (𝛼 = 0.37) 20 yield an estimate a Kf of 3.9. In contrast, the Brønsted acid-base reaction (i.e. deprotonation) was predicted to be highly endergonic (ΔG = +10.6 kcal/mol). With 4-dimethylamino-pyridine (DMAP), deprotonation was expectedly more favourable compared to pyridine, but still calculated to be endergonic (ΔG = +1.9 kcal/mol). Similarly, the formation of a NHPI-DMAP H-bonded complex is more thermodynamically favoured (ΔG = -4.1 kcal/mol) than its NHPI-pyridine counterpart. The smaller ΔG between full deprotonation and NHPI-DMAP formation (+6.0 kcal/mol) is consistent with the higher pKa of DMAP's conjugate acid (17.7). 17 Base Effect. To further probe the role of base in PINO generation and decomposition, a series of experiments were carried out with organic bases whose conjugated acids have well-defined pKa values in MeCN. 17 Cyclic voltammograms for the oxidation of NHPI in the presence of one equivalent of different organic bases were collected at 0.1 Vs -1 (selected examples presented in Figure 4A, the others in SI). In the presence of stronger bases, NHPI was oxidized at less positive potentials. The dependence of the E1/2 values on pKa are summarized in Figure 4B (blue plots). Across a pKa range of 2 to 19, the linear least squares fit yielded a slope of 56±1 mV per pKa unit, consistent with the Nernst factor (2.303×RT/nF = 59 mV/pKa) for proton-coupled electron transfer (PCET). 11 A separate analysis of the apparent electron transfer rate constant (k 0 ) for each voltammogram in Figure S7 did not yield an obvious trend with pKa. A likely factor was the differences in steric crowding around the pyridine nitrogen atom. For example, voltammetry with bipyridine yielded a much smaller k 0 value in comparison to pyridine, despite both voltammograms showing similar E1/2.</p><p>Besides the potential shift, we also observed that the magnitude of the cathodic peak current varied with base strength (Figure 4A). Across these organic bases, the ratio of the anodic and cathodic peak current (ipa/ipc) was a function of the pKa of the base, approaching 1 as the pKa increased to 15 and then increasing at larger pKa values (Figure 4B, red plots). In the presence of stronger bases, e.g. DMAP and quinuclidine, the oxidation was irreversible. While the role of base as a proton acceptor is generally recognized, [4][5][6][7][8]11 the reason for the erosion of the reversibility of the electrochemical oxidation is not immediately apparent. Based on the lack of reversability above a certain range of pKa, it would seem that at least under electrochemical conditions, the decomposition of PINO is accelerated under increasingly basic conditions.</p><p>In an effort to delineate the effects of individual bases, the redox response of the conjugate base of NHPI (i.e., PINO -) was determined. An NHPI salt, tetrabutylammonium phthalimide-N-oxide, was prepared 21 and its cyclic voltammogram was collected at 0.1 Vs -1 in MeCN. Taking E1/2 to be diagnostic of the formal potential, these data implied E 0' (PINO/PINO -) = +0.07 V vs E 0' (Fc + /Fc) (Figure 4A). The small cathodic peak further suggested PINO itself is unstable on the timescale of the voltammetry. 22 Based on the cumulative E1/2 vs pKa data, the voltammetry indicate a pKa of ~19 for NHPI in MeCN. For additional insight on the effect of base on electrochemical oxidation, we expanded the computations in Figure 3C to compare oxidation of NHPI, the H-bonded complexes thereof, and PINO -(Figure 4A). Exepctedly, NHPI was the most difficult species to oxidize, followed by the hydrogen-bonded complexes, and finally PINO -. Importantly, the H-bonded complexes of NHPI and either pyridine or DMAP were predicted to decrease the ionization potential by 24.4 kcal/mol and 30.4 kcal/mol, respectively relative to the NHPI-water complex. The decrease in ionization potential (IP) parallels the trend of decreasing standard potential for oxidation of NHPI observed experimentally, though a lack of quantitative agreement is not surprising given the absence of implicit solvation, which is presumably particularly relevant in the case of water. Notably, the complexes resulting from oxidation (i.e., PINO---H-base + ) were found to favor dissociation (ΔG = -4.5 and -4.7 kcal/mol, respectively), rendering pyridinium or dimethylaminopyridinium and free PINO. Hence, the increased reversibility of NHPI oxidation in the presence of pyridine does not appear to be due to product complexes which increase the persistence of PINO.</p><p>A test substrate, 1-phenylethanol, was used to assess whether the base affects the catalytic efficiency. A catalytic current enhancement was observed when the substrate was added to a solution of NHPI and organic bases (pKa < 15) (SI). The catalytic current was independent of the pKa of the base (SI). Importantly, no current enhancement was observed when stronger bases (i.e., DMAP, NEt3 or quinuclidine) were used. Comparing the catalytic currents measured for 1-phenylethanol and 1-phenylethanol-d1 yielded a kinetic isotope effect of 1.8, suggesting that the abstraction of the benzylic hydrogen atom is rate-determining in the electrochemical oxidation (SI).</p><p>Electrolysis of 1-phenylethanol catalyzed by NHPI were performed in the presence of different bases at a constant potential for 6 hours (Figure 5). The potential was set +0.2 V relative to the anodic peak potential for each redox couple. The resulting solutions were analyzed by gas chromatography coupled to a mass spectrometer (GC-MS). The best catalytic efficiency was observed for bases with a pKa within the range of ~12 to 15, with full conversion within the 6 hours. A substituent effect was noted, as the current magnitude and time to reach full conversion was dependent upon the steric bulk of the base. The less sterically hindered bases promoted higher currents and shorter reaction times. The correlation of reaction time and steric hinderance was in line with the following premise: Weaker bases (pKa < 10) displayed a current drop during the reaction and low conversions were obtained. On the other hand, stronger bases (pKa > 17) were ineffective for catalysis, resulting in no desired oxidation. Consistent with the foregoing, electrolysis of 1-phenylethanol in the presence of 1 equivalent of the NHPI salt resulted in >90% recovery of the starting material.</p><p>Mechanistic Insights on PINO decomposition. To understand the causes of catalyst decomposition in order to develop more efficient catalytic systems, PINO decomposition was investigated computationally to interrogate both the existing mechanistic proposals as well as others which we considered plausible. Initially, we examined the second-order decay originally proposed by Masui (Figure 1C, path a). 9 This decomposition path proceeds first by disproportionation of two molecules of PINO as the (presumed) rate determining step. The resulting oxoammonium PINO + (2) could then undergo acyl substitution by either NHPI or its conjugate base PINOto yield the putative dimeric intermediate, which could subsequently lead to the isolated trimer (vide infra). Unlike the oxoammonium ion derived from TEMPO, PINO + has not yet been reported in the literature. Indeed, the computed free energy difference between two molecules of PINO and PINO -+ PINO + (ΔG = +51.9 kcal/mol, Figure 6, path a) suggests that disproportionation is highly endergonic. Furthermore, the IP of PINO was computed to be highly similar to that of NHPI (ΔG = +158.6 vs +159.9 kcal/mol), suggesting that the potential required for anodic oxidation would be well beyond the potential at which oxidation of the NHPI-base complex occurs. On this basis, the formation of a discrete intermediate 2, as originally depicted by Masui, is unlikely to occur.</p><p>The previously suggested C-N fragmentation was also examined (Figure 1C, path b). 10 The only productive mode of C-N cleavage identified by computational means was through the rotation of one of the C(aryl)-C(acyl) bonds (Figure 6, path a), which led to a high energy charge-separated activated complex 3 with acylium cation and amide anion N-oxyl moieties (ΔG = +53.9 kcal/mol) -not an acyl radical. Intrinsic reaction coordinate (IRC) calculations revealed that this transition state collapses to phthalic anhydride monoimine-N-oxyl 4, obviously a consequence of the conformational restriction imposed on the reactive moieties by the aryl ring to which they are attached. To provide some insight on the intrinsic strength of the C-N bond in question, we determined the energy difference between maleimide-N-oxyl 5 and the (E)-isomer of its C-N bond fragmentation product 6 (Figure 6, path b). This transformation has essentially the same free energy cost as formation of the activated complex of PINO heterolysis (ΔG = +54.1 kcal/mol). The results imply that the suggested C-N bond cleavage is also unlikely to occur under relevant conditions. In addition, the concerted fragmentation of both C-N bonds analogous to the reverse reaction of quinone dimethidebased nitric oxide cheletropic traps (NOCTs) was considered (Figure 6, path c). 23 Although a transition state was not found, considering the endergonicity of this reaction (ΔG = +59.3 kcal/mol), it is also unlikely to contribute to the decay of PINO. Admittedly, this point is perhaps not surprising NOCT reactions have not been reported as being reversible.</p><p>Considering the modes by which two molecules of PINO may dimerize, a reaction pathway similar to Masui's original disproportionation proposal was found (Figure 7). An activated complex corresponding to the combination of PINOand PINO + was identified with ΔG = +36.5 kcal/mol relative to two isolated PINO molecules -much lower in energy than the sum computed from the energies of the discrete PINOand PINO + ions (ΔG = +51.9 kcal/mol, Figure 6A). IRC calculations suggest that this transition state connects Masui's aforementioned dimeric intermediate to a PINO contact ion pair at ΔG = +16.8 kcal/mol relative to two isolated PINO molecules. It seems reasonable to suggest that this contact ion pair results from charge transfer from a corresponding PINO radical pair to avoid formation of the higher energy separated PINOand PINO + ions. Indeed, the corresponding PINO (singlet) radical pair is 7.0 kcal/mol lower in free energy than the contact ion pair. Suspecting that the free energy difference between the PINO radical pair and two separated PINO radicals (9.8 kcal/mol) was overestimated on account of DFT's failure to account for dispersion interactions, we incorporated Grimme's empirical dispersion correction (GD3) 24 in our calculations. Doing so led to a decrease in the free energy change for association of the PINO radical pair (to +5.7 kcal/mol), the contact ion pair (to +10.3 kcal/mol) and overall barrier for substitution (to +30.7 kcal/mol) -see Supporting Information for further details.</p><p>An additional bimolecular pathway involving acyl substitution was assessed. Given the the lack of reversibility observed when the tetrabutylammonium salt of NHPI is oxidized or when stronger bases are present, we wondered if PINO  may attack PINO directly. Indeed, a comparatively low energy transition state structure was readily identified for this substitution (Figure 7). The transition state was characterized by significant O-C bond formation and C-N bond cleavage. IRC calculations revealed that this structure connected a PINO/PINO  complex and the substitution product independent of a tetrahedral intermediate. Again, inclusion of the GD3 empirical dispersion correction systematically decreased the overall computed free-energy barrier from +17.9 to +12.5 kcal/mol. Although, as in the case of the reaction of two PINO molecules, the reaction is endergonic (ΔG = +10.7 kcal/mol), the resultant amide anion N-oxyl should oxidize readily since its ionization potential is essentially equivalent to that calculated for PINO -(ΔG = 106.3 vs 107.1 kcal/mol, respectively). 25 This oxidation would generate Masui's proposed dimeric intermediate.</p><p>The optimized structure of Masui's proposed dimer was characterized by a rather conspicuously long C-N bond between the carbonyl carbon and nitroso nitrogen atoms (1.59 Å), implying it to be quite weak. Indeed, the calculated bond dissociation free energy (BDFE) is +10.2 kcal/mol, implying the homolysis would occur readily under the reaction conditions, releasing nitric oxide and yielding the corresponding acyl radical. The incipient acyl radical could be trapped by PINO, in a reaction that is predicted to be 42.6 kcal/mol downhill. 26 It should be noted that, in absolute terms, the B3LYP/6-311+G(d) method used for the foregoing calculations is likely to underestimate these bond strengths. Therefore to provide further insight, we carried out high accuracy CBS-QB3 calculations on a model model wherein the 'spectator' PINO moiety is replaced with a hydroxyl group, which yielded corresponding free energies of +15.1 and -58.7 kcal/mol, respectively). 27 Nevertheless, the results support a transitory 'dimeric' intermediate and a stable 'trimeric' product in line with the observational evidence accrued to date.</p><p>Considering the PINO decomposition mechanism is base promoted, the decomposition rate is base dependent. A series of low scan rate cyclic voltammetric measurement was performed at 0.01 Vs -1 as in Figure 2c. Among these investigated bases, the steady-state currents were observed consistently (SI). With weak bases (pKa < 15), the steady-state current was independent of base strength and base concentration. When stronger bases (pKa > 15) were used, the steady-state current was proportional to the base concentration. Taking this information together, we posit that the observed steady state current indicated that PINO decomposition produces a new redox active species that is easier to oxidize. These voltametric observations are in agreement with aforementioned computional studies.</p><!><p>The presented data speak to the following three points. First, PINO-catalyzed electrochemical oxidation occurs by a MS-CPET mechanism. Second, the reversibility for electrochemical oxidation of NHPI is dependent on the base strength, and consequently the base strength affects the conversion efficiency of the electrocatalytic oxidation of 1phenylethanol. Third, PINO decomposition is promoted by base via nucleophilic substitution by PINOon a carbonyl of a second PINO followed by single electron oxidation of the resultant complex.</p><p>The linear dependence of E1/2 on pKa with a slope of 56±1 mV/pKa, observed for the oxidation of NHPI in the presence of base, is a hallmark of PCET reactions. 11 Indeed, Mayer and co-workers have suggested MS-CPET as a plausible mechanism for the base-mediated formation of PINO from NHPI, in analogy to their study on base mediated oxidation of N- hydroxy-2,2,6,6-tetramethyl-piperidine (TEMPO-H). 18 The E1/2 vs pKa correlation addresses two further aspects. First, these data indicate that E 0' (NHPI •+ /NHPI) is ≥ +1.05 V vs E 0' (Fc + /Fc), in agreement with prior predictions from theory. 11 Second, these data indicate that E 0' (PINO/PINO -) is closer to +0.07 V vs E 0' (Fc + /Fc) -which is noticeably more positive than prior theoretical predictions of E 0' (PINO/ PINO -) = -0.1 V vs E 0' (Fc + /Fc). 11 Similarly, the cumulative voltammetric measurements shown here indicate that the pKa of NHPI in MeCN is 19 (rather than 23.5). 11 The subject of PINO decomposition has been a matter of interest since the seminal kinetic studies conducted by Masui, 9 although the details have remained hitherto nebulous. Both first 28 and second order decay, 9,29 with respect to PINO, have been reported leading to a bifurcation of proposed decay processes that have been presumed to lead to the same trimer decomposition product. 8 This paper argues that ascribing the observed first-order decay 28 of PINO to a unimolecular phenomenon perhaps takes too much for granted. The underlying assumption is that the principal product of the process is the same trimer previously observed by Masui. Additionally, as pointed out by Pedulli, the kinetics of decay are quite dependent on solvent composition, 28a an observation which seems inconsistent with unimolecular decay being rate-determining. However, without more information on the actual product(s) formed as a result of the first-order decay process(es), or the prerequisites for its occurrence, searching for the rate defining reaction in silico may be futile. Suffice to say, the oft-invoked fragmentation of the C-N bond in PINO to form a transient acyl radical which can be captured appears unlikely.</p><p>The presented data suggest that the strength of the base alters the catalytic efficiency influencing PINO decomposition. The voltammetric measurements and electrolysis experiments suggest an optimal base strength in the pKa range of ~12-15. Specifically, we posit that PINO has two operative decomposition pathways in MeCN (Figure 7 and 8). One decomposition involves the dimerization of two molecules of PINO via a charge-transfer complex. In the presence of weak bases (pKa < 15), this pathway is dominant. In the presence of comparatively strong bases (pKa > 15), a second significantly faster decomposition pathway consisting of the reaction between PINO  and PINO is operative. The resulting intermediate is rapidly oxidized at the anode to produce the same dimer. This oxidation event is observable at the slow scan rate voltammetry. The dimer undergoes C-N cleavage to release nitric oxide and form an acyl radical which reacts with another molecule of PINO to form trimer.</p><p>Our expansion on early observations regarding base-promoted oxidations and elucidation of a reasonable PINO decomposition mechanism provides guidance to advance PINO-catalyzed electrochemical oxidations. Previous efforts on designing PINO-derived HAT catalysts have largely focused on adjusting BDFEs to improve reactivity. In contrast, an approach to improving catalysis through extension of the effective catalytic lifetime of PINO by obviating degradation pathways has not been a focus of research efforts. On the basis of the results presented herein, we suggest that both of these aspects must be considered in the design of more effective N-oxyl catalysts for C-H oxidation reactions.</p><!><p>In summary, we have described a comprehensive electroanalytical study of the NHPI/PINO redox mediator, and conducted computational studies to uncover the mechanisms by which PINO degrades as a function of the identity of the base. These data provide a road map for developing more effective N-oxyl catalysts for C-H oxidation reactions.</p>

ChemRxiv

Biotin tethered homotryptamine derivatives: high affinity probes of the human serotonin transporter (hSERT)

Quantum dot conjugates of compounds capable of inhibiting the serotonin transporter (SERT) could form the basis of fluorescent probes for live cell imaging of membrane bound SERT. Additionally, quantum dot-SERT antagonist conjugates may be amenable to fluorescence-based, high-throughput assays for this transporter. This paper describes the synthesis of SERT-selective ligands amenable to conjugation to quantum dots via a biotin-streptavidin binding interaction. SERT selectivity and affinity were incorporated into the ligand via a tetrahydropyridine or cyclohexylamine derivative and the affinity of these compounds for SERT was measured by their ability to produce SERT-dependent currents in Xenopus laveis oocytes.

biotin_tethered_homotryptamine_derivatives:_high_affinity_probes_of_the_human_serotonin_transporter_

<p>Neurotransmitters such as dopamine, norepinephrine and 5-hydroxytryptamine (serotonin, 5-HT) modulate neuronal signaling in response to a variety of sensory and physiological stimuli. Synaptic concentrations of these neurotransmitters are regulated by presynaptic transporters, specifically the dopamine transporter (DAT), the norepinephrine transporter (NET) and the serotonin transporter (SERT).1,2 Methodologies used to study the location and temporal dynamics of these transporters within the membrane may provide important insights to their dysfunction, which is suspected to contribute to disorders such as autism, OCD, depression, ADHD and addiction. Our efforts currently center on the development of fluorescent probes based upon DAT and SERT antagonists that can be conjugated to quantum dots.3-6 Ultimately, these probes will be used to study the expression, location and dynamics of these transporters within the presynaptic membrane7-9 and to identify abnormalities and therapeutics pertinent to transporter-related illnesses.</p><p>In earlier studies, we identified 3-(1,2,3,6-tetrahydropyrindin-4-yl)-1H-indole (2) as a high affinity SERT antagonist that may be conjugated to quantum dots.8,9 A linker arm may be attached to this tetrahydropyridyl nitrogen atom, this biological activity is retained (Figure 1, Table 1). The ability of these derivatives to inhibit the uptake of tritiated serotonin (5-HT) in human SERT (hSERT) expressing HEK cells was measured. After conjugation to quantum dots, the uptake inhibition by these conjugates was measured and IC50 values were calculated following titration of conjugated dot concentrations (Table 1). Studies of non-specific binding of quantum dots to cellular membranes suggested a long polyethylene glycol chain on the surface of the dot was necessary to reduce non-specific adsorption of the quantum dots to cellular membranes.10 Such compounds were shown to retain antagonist activity.</p><p>Although we have demonstrated our ligands have a high affinity for hSERT in uptake inhibition assays, we sought further evidence that ligand-conjugated dots bind tightly to hSERT in intact cells in real-time physiological assays. Using hSERT expressing oocytes, it is possible to detect antagonist suppressed hSERT leak currents (which appear as outward currents above a pretreatment baseline), from these measurements it is possible to gain insights regarding the relative potency of hSERT ligands. In this report, the synthesis of a series of high-affinity hSERT ligands (Figure 3.) and their ability to suppress hSERT leak currents in hSERT expressing oocytes was studied.</p><p>Derivatives of both (2) and cyclohexenyl amine were synthesized with an alkyl spacer attached to biotin-PEG5000 (Figure 2). The alkyl spacer length was changed to study the effect on binding (compounds (8), (10) and (11)). The effect of substituting on the indole ring was studied by synthesizing compounds (8), (9) and (12). Compounds (13) and the intermediate (14) were synthesized extending the basic nitrogen from the tetrahydropyridine ring to study the effects of this modification on binding. Finally, a methyl caped PEG ligand (15) was synthesized and used as a negative control.</p><p>Compounds (8)-(12) were synthesized as shown in Scheme 1. Initially (2), the methoxy, and the cyano analogs were synthesized using the method described by Guillaume et al11; these were then attached to 2-(bromoalkyl)isoindoline-1,3-dione (15a), (15b) or (15c), resulting in (16a)-(16f) and were converted to the amines (17a)-(17f)9, after coupling to Biotin-PEG5000-NHS compounds (8)-(12) were obtained.12 The synthesis of (16a) and (16b) are previously described by Tomlinson et al.9</p><p>Compound (13) was synthesized as outlined in Scheme 2. The mono ethyl ketal of commercially available cyclohexane dione was converted to N-methyl-1,4-dioxaspiro[4.5]decan-8-amine (19) by a reductive amination with sodium triacetoxy borohydride and methylamine. The ketal was hydrolysed to yield 4-(methylamino)cyclohexanone (20) and this was reacted with indole to yield 4-(1H-indol-3-yl)-N-methylcyclohex-3-enamine (21) in a 20% overall yield.13 The 11 carbon spacer was attached to this using the same methodology as previously described to give (22) in a 64% yield.14 The phthalimide protecting group was removed by stirring in ethanol in the presence of hydrazine mono hydrate using the method previously described9 to give (14) in a 93% yield. This was coupled to Biotin-PEG5000-NHS to give compound (13).15 Compound (15) was obtained by reacting Biotin-PEG5000-NHS with methylamine.16</p><p>The ability of compounds (8)-(15) to suppress hSERT-dependent leak currents in Xenopus oocytes was measured by perfusing oocytes with a 1μM solution of (8)-(14) in buffer. Electrophysiological recordings for each ligand were obtained using a low-pass filter at 10 Hz and digitized at 20 Hz. All analyses were performed using Origin 7 (OriginLab, Northampton, MA). The magnitude of the leak currents for (8)-(14) was obtained using 5 replicates and the magnitudes of the leak currents for compounds (8)-(13) are shown in Figure 3. An example raw trace of leak suppression and 5-HT induced currents is shown in Figure 4. Compound (15) was a control compound and induced no suppression of leak current indicating that the indole derivative is necessary for biological activity.</p><p>Figure 4 shows a raw data trace obtained for the intermediate IDT373 (14).</p><p>In addition to measuring leak current suppression, the ability of the ligands to inhibit the uptake of tritiated 5-HT in hSERT expressing Xenopus oocytes was also measured and IC50 values were obtained using 5 replicates. Figure 5 shows a correlation of the IC50 value obtained using this oocyte system with the observed leak current suppression.</p><p>Compounds (8)-(14) showed leak currents that are typical of SERT antagonists. It is crucial to understand how tightly our ligands are binding to the SERT since we intend to utilize them in a variety of biological assays. Ligands that bind tightly and do not diffuse into the surrounding media during an assay may be required for assays to track the location and temporal dynamics of SERT within live cell membranes. On the other hand, ligands that do not bind as tightly may be more suitable for displacement assays which could be the basis of a quantum dot based high throughput (HTS) assay. The magnitude of the measured leak currents and IC50 value will enable the selection of the appropriate ligand.</p><p>The magnitude of the leak current is proportional to the potency of the ligand for hSERT interactions. IDT374 (13) exhibited the greatest activity while IDT317 (11) exhibited the lowest leak current suppression, and consequently, the weakest binding to hSERT. These data were mirrored by the measured IC50 values for SERT uptake inhibition. The relative potencies of our SERT ligands appears to be determined by three factors. The 1st of these is the length of the alkyl spacer in the linker arm; when the alkyl spacer is increased in length from 2 carbon atoms in IDT317 (11) to 6 carbon atoms in IDT366 (10), an increase in leak current suppression can be observed. This trend continues when the alkyl spacer is increased to 11 carbon atoms (IDT357 (8)), indicating that the alkyl spacer in these molecules participates in binding. One hypothesis is that hydrophobic interactions within SERT itself support ligand interactions, or another possibility may be that hydrophobic interactions between the alkyl spacer and the lipid membrane increase the tightness of binding. Substitution on the indole ring has a less significant effect on the magnitude of the leak current. Derivatives including electron withdrawing substituents on the 5-position of the indole ring, such as the cyano derivative (IDT361 (12)), have a slightly larger leak current suppression than derivatives where an electron donating substituent, such as a methoxy derivative is on the ring (IDT318 (9)). However these differences do not seem to correlate well with IC50 values since it has been reported that the cyano substituted parent drug has a potency that is approximately 20 fold greater than the methoxy substituted parent drug.17 A large increase in the magnitude of the leak current suppression was obtained by moving the basic nitrogen out of the tetrahydropyridyl ring to give IDT374 (13), suggesting this basic nitrogen may be interacting with the binding site in hSERT. The basic nitrogen may be acting as a hydrogen bond acceptor and the binding strength is likely dependent upon the orientation of the lone pair of electrons on the hydrogen bond acceptor. Consequently, it is not surprising any modification in the structure of the homotryptamine, which changes the orientation of this nitrogen atom and its lone pair of electrons, will have a significant impact upon the tightness of binding to hSERT if it is acting as a hydrogen bond acceptor.</p><p>In conclusion, the relative potency for interactions with hSERT for our ligands follows the order: IDT374 (13)∼IDT373 (14) > IDT361 (12) ∼ IDT357 (8) = IDT318 (9) > IDT366 (10) > IDT317(11). In addition to these leak current experiments, we also demonstrated that IDT317 (11) binds relatively weakly to hSERT since it can be easily washed off, restoring a normal 5-HT induced influx current (data not shown). The leak current revealed by incubating IDT317 with hSERT expressing oocytes was similar in magnitude to the leak current obtained with fluoxetine, while the antagonists with longer alkyl spacers had leak currents of similar magnitude to the higher affinity hSERT antagonist paroxetine. These data suggest that the other ligands have a higher affinity for hSERT, bind tighter than IDT317, and one or more of these compounds may be amenable to the development of quantum dot-based fluorescent assays using hSERT-expressing mammalian cells (preliminary data not shown).</p><p>High affinity SERT ligands for conjugation to quantum dots.</p><p>The ratio of antagonist revealed leak current to 5-HT induced current for the analogs used in this study.</p><p>A raw data trace showing a representative leak current obtained for the intermediate IDT373 (14).</p><p>Leak current correlations with 5-HT uptake inhibition IC50 in oocytes obtained for compounds (8)-(14).</p><p>(i) Et3N; (ii) Hydrazine monohydrate; (iii) Biotin-PEG5000-NHS</p><p>(i) Methylamine, Na(AcO)3BH; (ii) TFA; (iii) Indole, NaOMe; (iv) Et3N; (v) Hydrazine monohydrate; (vi) Biotin-PEG5000-NHS</p><p>Literature value11</p>

PubMed Author Manuscript

Geochemical compositional controls on DNA strand breaks induced in in vitro cell-free assays by crushed rock powders from the Panasqueira mine area, Portugal

DNA strand breaks are a common form of DNA damage that can contribute to chromosomal instability or gene mutations. Such strand breaks may be caused by exposure to heavy metals. The aim of this study was to assess the level of DNA strand breaks caused by µm-scale solid particles of known chemical composition with elevated heavy metals/metalloids, notably arsenic, using an in vitro cell-free DNA plasmid scission assay. These samples were incubated with and without H2O2 to see whether damage occurs directly or indirectly through the Fenton reaction. Levels of DNA damage in the absence of H2O2 were < 10%, but in the presence of H2O2, all samples showed higher levels of damage ranging from 10 to 100% suggesting that damage was being incurred through the Fenton reaction. Using bivariate correlation analysis and multiple linear regression, manganese oxide (MnO), sulphur (S), copper (Cu), and zinc (Zn) concentrations in the particulates were found to be the most significant predictors of DNA damage. The mechanism of this DNA damage formation has yet to be thoroughly investigated but is hypothesised to be due to reactive oxygen species formation. Further work is required to assess the extent of contribution of reactive oxygen species to this DNA damage, but this study highlights the potential role of chemistry and/or mineralogy to the extent and/or nature of DNA damage caused by particulates.

geochemical_compositional_controls_on_dna_strand_breaks_induced_in_in_vitro_cell-free_assays_by_crus

Introduction<!>Materials and methods<!>Sample collection<!>XRD analysis<!>Particle size distribution<!>Plasmid scission assay<!>Statistical analysis<!>XRD analysis<!>XRF analysis<!>Particle size distribution<!><!>Discussion<!>Conclusion<!>

<p>Heavy metals and metalloids are natural elements characterised by their high densities, atomic weights, or atomic numbers (Koller and Saleh 2018). Our natural environment contains a large number of heavy metals and metalloids, such as arsenic, cadmium, chromium, and nickel, that become sources of exposure to humans as a result of natural or anthropogenic processes (Alloway 2013; Tchounwou et al. 2012; Bhavani and Sujatha 2014).</p><p>It is well established that exposure to many heavy metals and metalloids causes adverse health effects in humans. Many heavy metals and metalloids are classified as human carcinogens by the International Agency for Research on Cancer (IARC 2018). A variety of signalling and cellular regulatory proteins that are involved in important processes such apoptosis, cell cycle regulation, DNA repair, DNA methylation, cell growth, and differentiation are affected by exposure to heavy metals and metalloids (Kim et al. 2015; Engwa et al. 2019). Any disruptions to these processes can lead to cancer (Engwa et al. 2019). The main mechanism of inducing these disruptions is oxidative stress. Certain heavy metals and metalloids, such as arsenic, iron, copper, chromium, cobalt, and vanadium, are known for their ability to produce reactive oxygen species (ROS) such as superoxide ion, hydrogen peroxide, and hydroxyl radical by utilising the Fenton chemistry/Haber–Weiss reaction (Jaishankar et al. 2014; Manoj and Padhy 2013; Szivák et al. 2009). Their production results in oxidative stress, a state where cells have elevated levels of reactive oxygen species (ROS), which causes damage to proteins (e.g. protein fragmentation), lipids (e.g. lipid peroxidation), and DNA (e.g. DNA strand breaks) (Schieber and Chandel 2014; Barrera 2012; Rehman et al. 2018; Engwa et al. 2019).</p><p>Given the known effects of exposure to heavy metals and metalloids, this study will focus on the effect of samples with known mineralogical, chemical, and physical characteristics on DNA strand breaks. The samples were collected from inside and around the Panasqueira mine area in Portugal and selected for the wide range of heavy metals/metalloids and major oxide composition. The overall objectives of this study were to (1) determine the level of DNA damage induced in the presence and absence of H2O2 using the plasmid scission assay and (2) identify the main determinants of DNA damage formation using bivariate correlation analysis and multiple linear regression.</p><!><p>Rock samples were collected from in and around the Panasqueira mine, Portugal. After crushing, the resultant crushed rock powders (CRPs) were analysed by means of X-ray diffraction (XRD), X-ray fluorescence (XRF), and particle size analyser to investigate their mineralogical, chemical, and particle size characteristics, respectively. The ability of CRPs to cause DNA damage was investigated using an in vitro cell-free plasmid scission assay. A more detailed description of each method is provided below.</p><!><p>Whole-rock samples ranging from 0.5 to 1 kg in weight were collected from the Panasqueira mine area in Portugal in February/March 1984. Every sample was broken up with a carbide splitter and reduced to millimetre-sized particles in a jaw crusher. A portion of the crushed material was then placed in a Cr-V stainless steel Tema Mill (TEMA Machinery Ltd., Woodford Halse, Northants, UK) and further crushed to < 50 µm powders. Pressed powder pellets were then prepared by standard techniques as outlined in (Polya 1987, 1988). Throughout the whole process, every piece of equipment was thoroughly cleaned after each sample treatment. Around 250, crushed rock powders (CRPs) were eventually obtained and subsequently stored in sealed individual zip-bags at room temperature. For the purpose of this study, a subset of 24 samples were selected on the basis of their chemical compositional variability, particularly with respect to arsenic, to test their association with toxic effects.</p><!><p>Sample preparation involved grinding ~ 0.1 g of crushed rock powder, mixing with ~ 1 ml of amyl acetate, using an agate pestle and mortar, transferring the resultant slurries to a glass microscope slide and air drying. Measurements were carried out on a Bruker D8 Advance diffractometer, equipped with a Göbel Mirror and a Lynxeye detector. The X-ray tube had a copper source, providing CuKα1 X-rays with a wavelength of 1.5406 Å. Samples were scanned from 5–70° to 2θ, with a step size of 0.02°–2θ and a count time of 0.2 s per step. The resultant XRD patterns were evaluated using EVA version 4, which compares experimental data to standards from the ICDD (International Centre for Diffraction Data) Database.</p><!><p>Particle size analysis was conducted at the British Geological Survey (BGS) in Keyworth by Thomas Walker (Walker, unpublished work). Each sample was weighed out into 2 vials of 0.25 g and suspended into 10 ml solution of Calgon (25% sodium hexametaphosphate). The samples were then shaken and mixed for 30 s using a vortex mixer at 2500 rpm before being analysed using a Beckman Coulter LS 13 320 Particle Sizing Analyser. Each sample was analysed twice. Each resultant particle size distribution was characterised by five parameters [P1 (% 100 µm peak), P2 (% 10 µm peak), and P3 (% < 1 µm), D10, D50,] chosen based on their ability to describe the sample as a whole or describe the largest fraction peak (100 µm), the modal fraction (10 µm), and the nanoparticle fraction (less than one µm). D10 is the diameter at which 10% of a sample's mass is comprised of smaller particles, while D50 is the diameter at which 50% of a sample's mass is comprised of smaller particles. Both D10 and D50 values were calculated from the distribution, and statistical analysis was conducted using Gradistat© software on Microsoft Excel (Walker, unpublished).</p><!><p>The ability of CRPs to cause DNA strand breaks was investigated using the plasmid scission assay as described previously (Dumax-Vorzet et al. 2015) with minor modifications. When plasmid DNA runs through an agarose gel, three bands are observed. Supercoiled plasmid DNA is the native form (covalently closed circular DNA) where there are no strand breaks. When one DNA strand is cut, the resulting nicked or relaxed plasmid DNA will have a floppy open circle structure. When both strands of the plasmid are cut, the result is linear plasmid DNA. These three forms have different migration speeds where supercoiled plasmid DNA is the fastest as it does not have any strand breaks and its compactness sustains less friction against the agarose gel. Linear plasmid DNA runs through the gel slower than supercoiled plasmid DNA but faster than nicked or relaxed plasmid DNA. Nicked or relaxed plasmid DNA is the slowest due to its large floppy circular nature. In brief, pchAT plasmid DNA (kindly provided by Prof. Geoff Margison, purified from E.coli in lab using Miniprep (Qiagen, The Netherlands)) (5 ng) was diluted to 20 μl in an elution buffer (10 mM Tris–HCl pH 8.5) with different levels of CRPs and H2O2. Samples were incubated for 1–5 h at 37 °C. The reaction was stopped by adding loading buffer (Promega blue/orange 6 × loading dye) and the whole reaction mixture loaded onto 0.6% TBE-agarose gel. Electrophoresis was conducted at 90–100 V for 45 min–2 h in 1 × TBE buffer. The different forms of plasmid were visualised on a Typhoon 9200 variable mode imager. The intensity of the different forms of plasmid in each lane was analysed using ImageQuantTL (GE Healthcare Life Sciences), and the level of damaged plasmid in each sample was calculated as shown in Eq. (1).1\documentclass[12pt]{minimal} sepackage{amsmath} sepackage{wasysym} sepackage{amsfonts} sepackage{amssymb} sepackage{amsbsy} sepackage{mathrsfs} sepackage{upgreek} \setlength{\oddsidemargin}{-69pt} egin{document}$${ ext{DP}} \left( \% ight) = rac{R + L}{R + L + S} imes 100$$\end{document}DP%=R+LR+L+S×100where DP is the percentage of DNA damage, R is the relaxed form of plasmid DNA, L is the linear form of plasmid DNA, and S is the supercoiled form of plasmid DNA.</p><p>In each experiment, positive and negative controls were added. The positive control was H2O2 (3.5 mM), pchAT plasmid DNA (5 ng), and FeSO4 (25 µM) in elution buffer. The negative control was H2O2 (3.5 mM) and pchAT plasmid DNA (5 ng) diluted in elution buffer.</p><!><p>Data obtained from each plasmid scission assay were described using the mean, standard deviation, minimum, and maximum values. A one-way analysis of variance was used to compare each group individually to determine whether the levels of DNA strand breaks varied significantly between samples. A bivariate correlation analysis was then conducted to examine possible associations between DNA strand breaks and the physiochemical composition of the samples. All the significant variables from this analysis were then plotted against the percentage of DNA strand breaks to examine the correlations and entered into a backward stepwise multiple linear regression model. The least significant variable was eliminated step by step, and all the models were compared using Bayesian Information Criterion (BIC) to choose the best explanatory model. Statistical analyses were performed using SPSS Statistics version 22. Graphs and scatterplots were created using Microsoft Excel 2010.</p><!><p>The crystalline minerals identified by XRD in the CRPs were mostly silicates with minor sulphides. These included quartz, muscovite 2M1, dravite tourmaline, and albite (Tables 1, 2 in supplementary material). The most abundant crystalline minerals found were quartz (Modal abundance = 40%, SD = 17%), muscovite 2M1 (M = 32%, SD = 19%), and dravite tourmaline (M = 11%, SD = 16%) with lesser amounts of albite (M = 9%, SD = 12%), phlogopite 1 M mica (M = 4%, SD = 8%), clinochlore II2b (M = 3%, SD = 5%), and traces of microcline intermediate 1, magnetite, and pyrite. (Tables 1, 2 in supplementary material).</p><!><p>The chemical compositions of the CRPs are summarised in Tables 3, 4, 5 and 6 of Supplementary Material. The compositions of these largely lower Greenschist facies meta-silstones and meta-sandstones are dominated by SiO2 (M = 63%, SD = 8%), Al2O3 (M = 19%, SD = 5%), and Fe2O3 (M = 7%, SD = 2%). Notable traces included S (M = 1600 µg/g, SD = 3800 µg/g), Ba (M = 560 µg/g, SD = 250 µg/g), and As (M = 380 µg/g, SD = 650 µg/g).</p><!><p>The analysis showed that samples were very similar in terms of their size distribution (Table 7 in supplementary material). The model fraction P2 (% 10 µm peak) was the most abundant fraction (M = 25.6, SD = 5). The largest fraction P1 (% 100 µm peak) was the second most abundant (M = 13.4, SD = 5.2). The nanoparticle fraction P3 (% < 1 µm) was only present in small amounts (M = 0.8, SD = 0.8). (Table 8 in supplementary material).</p><!><p>Production of DNA strand breaks by H2O2 and FeSO4. a electrophoresis results, b  %DNA damage induced by increasing amounts of H2O2)—pchAT plasmid DNA was incubated at 37 °C for 1 h with 0-200 μM H2O2 in elution buffer (10 mM Tris–HCl pH 8.5) (20 μl). c Electrophoresis results, d  % DNA damage induced by increasing amounts of FeSO4 in the presence of H2O2)—pchAT plasmid DNA was incubated at 37 °C for 1 h with 3.5 mM H2O2 and increasing concentration of FeSO4 (0–25 µM). Both reactions were stopped by the addition of loading buffer (Promega blue/orange 6 × loading dye). The samples were separated on 0.6% TBE-agarose gel at 90 V for 2 h. The different forms of plasmid were visualised on a Typhoon 9200 variable mode imager. The intensity of the different forms of plasmid (relaxed, linear, and supercoiled) in each lane was analysed using ImageQuantTM, and the level of damaged plasmid in each sample was calculated. Error bars represent the standard deviation (SD) of three independent experiments</p><p>Production of DNA strand breaks by increasing amounts of CRPs (100–1250 µg/ml). a–c % DNA damage induced by increasing amounts of CRPs in the presence of H2O2 for three representative samples E, G, and K). Samples were suspended in distilled water (5 mg/ml) by sonication for a total of 3 min. Sonication was performed at 80% amplitude. The samples were used directly after being suspended without centrifugation. Plasmid DNA (5 ng) was incubated at 37 °C for 5 h with 3.5 mM H2O2 and increasing concentrations (100 μg/ml–1250 μg/ml) of the CRPs. (d  % DNA damage induced by negative and positive controls) positive [FeSO4 25 µM, plasmid DNA (5 ng), H2O2 3.5 mM, and elution buffer (10 mM Tris–HCl pH 8.5)] and negative [plasmid DNA (5 ng), H2O2 3.5 mM, and elution buffer (10 mM Tris–HCl pH 8.5)] controls were also added. The reaction was stopped by the addition of loading buffer (Promega blue/orange 6 × loading dye). The samples were separated on 0.6% TBE-agarose gel at 100 V for 45 min. Three independents were carried out, and the mean for each sample was calculated. Error bars represent the standard deviation (SD) of three independent experiments</p><p>Production of DNA strand breaks by CRPs. (a % DNA damage induced by 1250 μg/ml of all CRPs in the presence of H2O2). Samples were suspended in distilled water (5 mg/ml) by sonication for a total of 3 min. Sonication was performed at 80% amplitude. The samples were used directly after being suspended without centrifugation. pchAT Plasmid DNA was incubated at 37 °C for 5 h with and without H2O2 at the highest concentration of the sample. (b  % DNA damage induced by negative and positive controls) positive [FeSO4 25 µM, plasmid DNA (5 ng), H2O2 3.5 mM, and elution buffer (10 mM Tris–HCl pH 8.5)] and negative [plasmid DNA (5 ng), H2O2 3.5 mM, and elution buffer (10 mM Tris–HCl pH 8.5)] controls were also added. The reaction was stopped by the addition of loading buffer. The samples were separated on 0.6% TBE-agarose gel at 100 V for 45 min. Six independent experiments were carried out, and the mean for each sample was calculated. Error bars represent the standard deviation (SD) of six independent experiments</p><p>Correlation between significant components and DNA damage (r and p values shown were obtained from bivariate correlation analysis)</p><!><p>All samples collected from inside and around the Panasqueira mine area were able to induce DNA damage when incubated as a CRP with plasmid DNA in the presence of H2O2. The percentage of plasmid DNA damage varied significantly with MnO, S, Cu, and Zn, these chemical components also being significant predictors of DNA damage in a multivariate model. To the best of knowledge, this is the first study to determine direct apparent effects of MnO, S, Cu, and Zn in CRPs in a cell-free DNA scission assay, although it is noted that the large degree of covariance of these compositional parameters with other compositional parameters means that the DNA damage cannot be uniquely ascribed to each of these components and this represents a fundamental limitation of such toxicological studies involving real multicomponent geological materials. Previous studies have reported associations between these chemical components and DNA damage in cell-based studies (Alarifi et al. 2017; Frick et al. 2011; Hoffman et al. 2012; Linder 2012; Cervantes–Cervantes et al. 2005; Arciello et al. 2005; Zyba et al. 2016; Sharif et al. 2012; Ho and Ames 2002; Ho et al. 2003; Wysokinski et al. 2012). Manganese oxide nanoparticles have been associated with DNA strand breaks in human neuronal cells (Alarifi et al. 2017), and type-II alveolar epithelial cells (Frick et al. 2011). Sulphur (Hoffman et al. 2012) and copper (Linder 2012; Cervantes–Cervantes et al. 2005; Arciello et al. 2005) have been associated with superoxide and hydroxyl radicals which can result in oxidative stress and can lead to DNA damage. An increase in dietary zinc is actually known to reduce DNA damage (Zyba et al. 2016). Zinc deficiency on the other hand induces oxidative stress which leads to DNA damage (Sharif et al. 2012; Ho et al. 2003; Ho and Ames 2002). However, it has been reported before that zinc behaves differently in normal cells and cancer cells. Wysokinski et al. (2012) found that cancer cells exhibited higher levels of DNA damage in the presence of zinc, while in normal lymphocytes, such an effect was not found (Wysokinski et al. 2012). In this study, zinc levels were associated with an increase in DNA damage.</p><p>It has been previously reported that arsenic causes DNA strand breaks in mouse lungs (Yamanaka and Okada 1994), human fibroblasts (Mourón et al. 2006), and human HeLa S3 cells (Schwerdtle et al. 2003). However, in our study, we found no significant linear association between arsenic concentrations and DNA strand breaks, notwithstanding that the CRPs contained up to 3000 µg/g As. The lack of association between arsenic and DNA damage could have been attributed to its insolubility in a wide range of pH conditions (Flora 2014). Moreover, certain contaminants need to be converted from their original form by enzymes first in the human body to show any adverse effects. That especially applies for arsenic as its metabolism is a critical determinant of its toxic effects (Navas-Acien and Guallar 2008; Hughes et al. 2011; Jomova et al. 2011). This could be why arsenic was not found in this study to be associated with DNA damage as the assay was cell-free, while in the other previously mentioned studies, the assays were cell-based.</p><!><p>Crushed rock powders of known chemical composition have been shown to induce variable levels of DNA damage in a cell-free assay. MnO, S, Cu, and Zn were significant predictors of this DNA damage. Further work is required to characterise the mechanism of DNA damage formation and to determine to what extent these cell-free studies correlate with cellular studies. In particular, the perhaps surprising lack of association of DNA damage with arsenic concentration in the crushed rock powders highlights how cell-free assays may not be representative of toxicity behaviour in human cells, but the assays nevertheless confirm that the toxicity of µm-scale particles may be strongly dependent upon their chemical and mineralogical composition.</p><!><p>Mineralogical composition of crushed rock powders from Panasqueira; as determined by XRD</p><p>*ND not detected</p><p>Descriptive statistics for abundance (wt%) of crystalline minerals identified by XRD in crushed rock powders from Panasqueira</p><p>Chemical composition (major oxides) of crushed rock powders from Panasqueira; determined by XRF (Polya 1987, 1988)</p><p>ND not detected; 0.0 indicates < 0.05</p><p>Chemical composition (trace elements) of crushed rock powders from Panasqueira; as determined by XRF (Polya 1987, 1988)</p><p>ND not detected</p><p>Descriptive statistics for chemical composition (major oxides) of Panasqueira crushed rock powders (n = 24) determined by XRF</p><p>Descriptive statistics for chemical composition (trace elements) of Panasqueira crushed rock powders (n = 24) determined by XRF</p><p>Zero used as default to indicate below detection limit</p><p>Particle size distribution of crushed rock powders from Panasqueira; as determined by Beckman Coulter LS 13 320 Particle Sizing Analyser</p><p>D10 is the diameter at which 10% of a sample's mass is comprised of smaller particles. D50 is the diameter at which 50% of a sample's mass is comprised of smaller particles. P1 (% 100 µm peak) describes the largest fraction peak (100 µm). P2 (% 10 µm peak) describes the modal fraction peak (10 µm). *P3 (% < 1 µm) describes the sub-micron particle fraction peak (< 1 µm)</p><p>Descriptive statistics for particle size distribution parameters for crushed rock powders from Panasqueira</p><p>A comparison of the level of DNA strand breaks (mean/standard deviation of six independent experiments) between powdered rock samples from Panasqueira (n = 24) incubated with and without H2O2 using a paired samples t test</p><p>*There was a significant difference for all samples when incubated with and without H2O2. All p values for the t test were < 0.001</p><p>*(−) values could not be calculated</p><p>Publisher's Note</p><p>Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.</p>

PubMed Open Access

Structural Characterization of Mutations at the Oxygen Activation Site in Monomeric Sarcosine Oxidase\xe2\x80\xa0\xe2\x80\xa1

Oxygen reduction and sarcosine oxidation in monomeric sarcosine oxidase (MSOX) occur at separate sites above the si- and re-face, respectively, of the flavin ring. Mutagenesis studies implicate Lys265 as the oxygen activation site. Substitution of Lys265 by a neutral (Met, Gln, Ala) or basic (Arg) residue results in a ~104-fold or 250-fold decrease, respectively, in reaction rate. The overall structure of MSOX and residue conformation in the sarcosine binding cavity are unaffected by replacing Lys265 with Met or Arg. The side chain of Met265 exhibits the same configuration in each molecule of Lys265Met crystals and is nearly congruent with Lys265 in wild-type MSOX. The side chain of Arg265 is, however, dramatically shifted (~4 to 5 A) compared with Lys265, points in the opposite direction, and exhibits significant conformational variability between molecules of the same crystal. The major species in solutions of Lys265Arg is likely to contain a \xe2\x80\x9cflipped-out\xe2\x80\x9d Arg265 and exhibit negligible oxygen activation, similar to Lys265Met. The 400-fold higher oxygen reactivity observed with Lys265Arg is attributed to a minor (< 1%) \xe2\x80\x9cflipped-in\xe2\x80\x9d Arg265 conformer whose oxygen reactivity is similar to wild-type MSOX. A structural water (WAT 1), found above the si-face of the flavin ring in all previously determined MSOX structures, is part of an apparent proton relay system that extends from FAD:N(5) to bulk solvent. WAT1 is strikingly absent in Lys265Met and Lys265Arg, a feature that may account for the apparent kinetic stabilization of a reductive half-reaction intermediate that is detectable with the mutants but not wild-type MSOX.

structural_characterization_of_mutations_at_the_oxygen_activation_site_in_monomeric_sarcosine_oxidas

<!>Enzyme Expression and Purification<!>Crystallization and Data collection<!>Structure Determination and Refinement<!>Structural Calculations and Drawings<!>Overall Structure of the Mutant Enzymes and Comparison with Wild-type MSOX<!>Does Mutation of Lys265 Affect the Structure of the Sarcosine Oxidation Site?<!>Does Mutation of Lys265 Cause Structural Changes Above the Si-face of the Flavin Ring?<!>DISCUSSION<!>Concluding Remarks<!>

<p>The ability to activate the reduction of molecular oxygen underpins all aerobic biology. Flavoprotein oxidases catalyze the reduction of oxygen to hydrogen peroxide, a highly reactive molecule that, analogous to nitric oxide, is both a cytotoxin and a cell signaling molecule. The mechanism of oxygen activation by flavoprotein oxidases is poorly understood and an area of considerable current interest (1-3). Recently, we initiated studies on the mechanism of oxygen activation by monomeric sarcosine oxidase (MSOX)1 (4). The enzyme catalyzes the oxidation of sarcosine (N-methylglycine) to an imine (CH2=NH+CH2CO2−) that is subsequently hydrolyzed to produce glycine and formaldehyde. MSOX is a 44 kDa two-domain protein that contains covalently bound FAD (8α-S-cysteinyl-FAD). The flavin ring of FAD is located at the interface between the flavin and catalytic domains (5-8). MSOX is a member of a family of monomeric amino acid oxidases (N-methyltryptophan oxidase, nikD, pipecolate oxidase, fructosyl amino acid oxidase) that contain covalently bound FAD (9-13). MSOX also exhibits structural and sequence homology with the 44 kDa β-subunit of heterotetrameric sarcosine oxidase (TSOX) (14, 15). The TSOX β-subunit contains two flavins: i) a noncovalently bound FAD that does not react with oxygen but is the site of sarcosine oxidation and is structurally equivalent to the covalent FAD in MSOX; (ii) a covalently bound FMN (8α-(N3-histidyl)FMN) that is attached to the surface of the β-subunit and the site of oxygen activation (16, 17).</p><p>The active site for sarcosine oxidation in MSOX or TSOX is located at highly similar sites above the re-face of the flavin ring of FAD (8, 14). Recent mutagenesis studies show that Lys265 is the site of oxygen activation in MSOX and is entirely responsible for the rate acceleration observed with wild-type enzyme (4). Lys265 is located above the si-face of the flavin ring and is hydrogen bonded to the N(5) position of FAD via a bridging water molecule (Figure 1). The existence of separate sites for sarcosine oxidation and oxygen reduction on opposite faces of the flavin ring is likely to facilitate oxygen access to the reduced flavin by avoiding the sterically crowded region above the re-face in the reduced enzyme·imine complex, a catalytically significant intermediate (7). Lys265 is absent from the homologous but oxygen-unreactive FAD site in TSOX but is conserved in members of the MSOX family of amino acid oxidases (10, 18, 19) with the notable exception of nikD. Significantly, ligand-free reduced nikD exhibits low reactivity with oxygen, unlike MSOX. Instead, oxygen activation by nikD is triggered by the presence of bound substrate or product (20). Lysine residues hydrogen bonded to flavin N(5) via a bridging water are found in a number of other flavoprotein oxidases (monoamine oxidase B, polyamine oxidase, monoamine oxidase A, L-amino acid oxidase, lysine-specific histone demethylase) (21-25). The possible role of these lysines in oxygen activation, however, remains to be determined.</p><p>The 2-electron reduction of oxygen to hydrogen peroxide by free reduced flavin is thermodynamically favorable but spin-forbidden. Instead, the reaction proceeds via an initial 1-electron transfer step that generates a flavin radical-superoxide anion radical pair in a spin-allowed but energetically unfavorable rate-determining step (Scheme 1) (26). The rate acceleration observed for the reduction of oxygen by reduced MSOX must clearly be achieved by decreasing the activation energy (ΔG‡) for the initial 1-electron transfer step. ΔG‡ will depend on the free energy change (ΔG) and the reorganization energy (λ) of the reaction. The kinetics observed for the self-exchange reaction between oxygen and superoxide anion indicate that the reoganization energy required to change the configuration of the surrounding medium (λout) constitutes the major energy barrier in the 1-electron reduction of oxygen (27). The positively charged ε-amino group of Lys265 and the adjacent pocket occupied by WAT 1 and/or WAT 2 (see Figure 1) might define a pre-organized binding site for superoxide anion that could accelerate the 1-electron reduction of oxygen by lowering λout. Consistent with this scenario, mutation of Lys265 to a neutral residue (Met, Gln, Ala) results in a ~104-fold decrease in the observed rate of oxygen reduction. Unexpectedly, a chemically conservative substitution of Lys265 by Arg results in a relatively modest but still substantial decrease (250-fold) in the rate of the oxidative half-reaction (4). Rapid reaction kinetic studies show that mutation of Lys265 to Met or Arg hardly affects the actual rate of sarcosine oxidation. However, a novel spectral intermediate is observed during the anaerobic reaction of the mutant enzymes with sarcosine. Although the intermediate is not observed with wild-type MSOX, virtually identical spectral properties are observed for substrate-reduced wild-type or mutant enzyme (4).</p><p>In this paper, we report crystal structures of Lys265Met and Lys265Arg. The results provide considerable insight regarding the observed catalytic properties of the mutant enzymes.</p><!><p>The Lys265Met and Lys265Arg mutants were expressed and purified as previously described (4). The isolated preparations were largely (Lys265Arg) or totally (Lys265Met) devoid of covalently bound FAD. The mutant apoenzymes were reconstituted with FAD to yield preparations containing covalently bound FAD, as previously described (4).</p><!><p>Crystals of the Lys265Arg MSOX mutant were grown under two conditions ("phosphate" and "PEG") and of the Lys265Met mutant under a single condition ("phosphate") by the hanging drop method as described previously (8). Equal volumes of 5 μL each of protein solution (10 mg/ml in 20 mM Tris-HCL, pH 8.0) and reservoir solution (1.7 M Na/K phosphate buffer, pH 7.0 for "phosphate" and ~20% PEG4000, 200 mM sodium acetate, 100 mM Tris, pH 8.5 for "PEG") were mixed and allowed to equilibrate. X-ray data were recorded from a single crystal from each condition at 100 °K, using 15% glycerol as a cryoprotectant for the Lys265Met "phosphate" and the Lys265Arg "PEG" crystals and paratone oil for the Lys265Arg "phosphate" crystal, on an ADSC Quantum-315 CCD detector at Biocars Beamline 14-BM-C at the Advanced Photon Source, Argonne IL. Spot integration and data scaling were carried out using HKL2000 (28, 29). The space group symmetry, unit cell parameters, contents of the asymmetric units and the data collection statistics are summarized in Table 1.</p><!><p>The initial coordinates of both of the "phosphate" crystals were obtained by direct refinement of the isomorphous wild type MSOX structure (pdb code 2GB0) while those for the "PEG" crystal were obtained by molecular replacement using MOLREP from the ccp4 package (29), using the same wild type MSOX structure as the search molecule.</p><p>The refinement and electron density map calculations were carried out using REFMAC (30) and 5% of the reflections were selected randomly and set aside as a test set for cross validation (31). Reflections from 40 A to the diffraction limit recorded for each data set were included in the refinements and a bulk solvent correction was applied (32). Model building and analysis of the structures were carried out using COOT (33). Rigid body refinement followed by several cycles of positional and temperature factor refinement and solvent placement with manual examination, were carried out, also using COOT. This procedure utilized electron density difference maps calculated with Fourier coefficients (2Fo−Fc) and (Fo−Fc), where Fo and Fc are the observed and calculated structure factors, respectively. The quality of the refined structures and the resulting electron density maps of both structures is high. The final refinement statistics are shown in Table 1.</p><!><p>Structural diagrams were rendered using PYMOL (http://www.pymol.org).</p><!><p>The Lys265Met mutant was crystallized under precipitating conditions of high salt to produce "phosphate" crystals that diffract to 1.6 A and contain two molecules in the asymmetric unit. When the two molecules in the crystals are aligned, the only significant differences in backbone structure are found in an external loop (Ile185-Tyr192) and an active site loop (Tyr55-Tyr61) (Figure 2). Previous studies with "phosphate" crystals of other MSOX preparations show that the configuration of the external loop is sensitive to the crystal packing environment and is consistently different between molecule 1 (A configuration) and molecule 2 (B configuration) in the asymmetric unit (34). The same pattern is observed with "phosphate" crystals of Lys265Met (Table 2). The active site loop controls access to a solvent-filled active site cavity above the re-face of the flavin ring. The active site loop is found in the closed configuration in complexes of wild-type enzyme with various sarcosine analogs (6, 8). In contrast, the active site loop is mobile in ligand-free MSOX, as judged by the open and closed configurations observed in molecules 1 and 2, respectively, in "phosphate" crystals of wild-type enzyme or Lys265Met (Table 2). Importantly, no significant difference in backbone structure is detected when molecule 1 or 2 in "phosphate" Lys265Met crystals is compared with the corresponding molecule in "phosphate" wild-type crystals (data not shown) (see Figure S1 of the Supporting Information).</p><p>The Lys265Arg mutant was crystallized in the presence of high salt ("phosphate" crystals) to produce crystals that diffract to 1.6 A and contain two molecules in the asymmetric unit. Molecules 1 and 2 in "phosphate" Lys265Arg crystals exhibit configurations of the external and active site loops that are identical to those observed for wild-type enzyme (Table 2). Very similar backbone structures are apparent upon comparison of molecule 1 or 2 in the mutant crystals with the corresponding molecule in the wild-type crystal (data not shown) [see Figure S2 (top and middle panels) of the Supporting Information].</p><p>The Lys265Arg mutant was also crystallized in the presence of high PEG ("PEG" crystals) to produce crystals that diffract to 2.1 A and contain four molecules in the asymmetric unit. The four molecules in the "PEG" crystals of Lys265Arg exhibit an A or A-like configuration of the external loop, similar to that observed with "PEG" crystals of Arg49Lys (34). The results are consistent with an altered packing arrangement in "PEG" crystals as compared to "phosphate" crystals. The active site loop is in the open configuration in each of the four molecules in "PEG" Lys265Arg crystals whereas "PEG" crystals of Arg49Lys exhibit both open and closed configurations (Table 2). Except for minor variations in the external loop, no significant difference in backbone structure is detected upon comparison of molecule 1, 2, 3, or 4 in "PEG" Lys265Arg crystals with molecule 1 in "phosphate" wild-type crystals (data not shown) [see Figure S2 (bottom panel) of the Supporting Information].</p><!><p>The active site cavity for sarcosine oxidation is located above the re-face of the flavin ring. When the active site loop is in the open configuration, the cavity in ligand-free wild-type MSOX contains eight water molecules (WAT 1 to 8) and Arg52 is found in the "out" position (Figure 3). WAT 1 to 5 occupy the sarcosine binding site and are displaced upon binding of substrate analogs (35). Complex formation also results in closure of the active site loop, movement of Arg52 from the "out" to the "in" position, and displacement of two additional waters (WAT 6 and 7), probably caused by the accompanying motion of Arg52. Only a single water molecule (WAT 8) is retained in MSOX complexes with substrate analogs. The structures observed for these complexes indicate that the substrate carboxylate group is bound to Lys348 and Arg52. The carbonyl oxygen of Gly344 forms a hydrogen bond to NH in sarcosine (6, 8). His269 and Tyr317 are important for optimizing the orientation of the bound substrate (36, 37).</p><p>Comparison of molecule 1 in "phosphate" Lys265Met crystals with molecule 1 in the corresponding wild-type crystals indicates that the active site cavity is scarcely affected by the mutation. Two changes are, however, worth noting: (i) WAT 7 is not detected in the mutant structure; (ii) there is a 2.4 A shift in the position of Met245:CE (Figure 3, top panel). The active site cavity in Lys265Arg is also largely unperturbed, as can be seen by the structure observed for molecule 1 in "phosphate" crystals (Figure 3, bottom panel). However, in this case, two additional water molecules are absent in the mutant crystals (WAT 5 and 8) and an additional side chain atom, Glu57:OE2, is shifted by 1.8 A. Similar results are obtained for molecules 1 to 4 in "PEG" Lys265Arg crystals (data not shown).</p><!><p>Oxygen activation in wild-type MSOX occurs at Lys265, a residue located above the si-face of the flavin ring. Lys265:NZ is hydrogen bonded to FAD (N5) via a bridging water molecule (WAT 1). WAT 1 is also hydrogen bonded to Arg49:NH1, Thr48:O and a second water molecule (WAT 2). Arg49 is in van der Waals contact with the si-face of the flavin ring (Figure 4). Thr48 is part of a putative proton relay system extending from FAD N(5) to bulk solvent (see Figure 1). WAT 1 is present in all previously determined MSOX structures, including both molecules found in "phosphate" crystals of wild-type enzyme (2GBO), His269Asn (1L9C), Arg49Lys (3BHF) or complexes of wild-type MSOX with various inhibitors (1EL9, 1EL1, 2GF3, 1EL5) and both molecules in "PEG" crystals of Arg49Lys (3BHF). Five of the eight crystals also contain WAT 2 in both molecules (2GBO, 1L9C, 2GF3, 3BHF, 1EL5) whereas the others (1EL9, 1EL1, 3BHK) contain WAT 2 in one of the two molecules.</p><p>Molecules 1 and 2 in "phosphate" crystals of Lys265Met exhibit identical conformations for residues above the si-face of FAD. Importantly, the side chain of Met265 is nearly congruent to that of Lys265 in wild-type MSOX, except for a 1.6 A difference between the position of the sulfur atom in Met265 and the corresponding carbon atom in Lys265. In wild-type MSOX two alternate conformations for Asp47 are observed in molecule 1 (A and B). Only the A conformation of Asp47 is found in molecule 2 of wild-type MSOX or in both molecules of "phosphate" crystals of Lys265Met. Thr48 and Arg49 occupy identical positions in wild-type and mutant enzyme but WAT 1 and WAT 2 are strikingly absent in both molecules of the mutant crystals. Aside from this difference, the region above the si-face of the flavin is virtually unaffected by mutation of Lys265 to Met (Figure 4, top panel). Figure S3 (top panel) shows (2Fo−Fc) and (Fo−Fc) electron density maps for selected residues above the si-face in molecule 1 of "phosphate" crystals of Lys265Met.</p><p>A dramatic effect on the conformation and flexibility of the side chain of residue 265 is observed upon mutation of Lys265 to Arg, as judged by results obtained for the two molecules in "phosphate" crystals or the four molecules in "PEG" crystals of Lys265Met (Figure 4, middle and bottom panels, respectively). The side chain of Arg265 has moved about 4 to 5 A away from FAD and points in a significantly different direction, as compared with Lys265 in wild-type MSOX. Unlike Lys265 or Met265, the side chain of Arg265 exhibits substantial mobility, as judged by a 1 to 2 A distance observed between equivalent atoms when molecules in the same crystal are aligned with each other. Two alternate conformations for Asp47 are observed in molecule 1 (A-like and B-like) or molecule 2 (A-like and B-like) in "phosphate" crystals of Lys265Arg. Only the B conformation of Asp47 is found in molecules 1-4 in "PEG" crystals of Lys265Arg. Mutation of Lys265 to Arg does not affect the conformation of Arg49 or Thr48 but does result in the apparent loss of the two si-face waters (WAT 1 and 2) that are found in crystals of wild-type MSOX but not Lys265Met (Figure 4). Figure S3 (bottom panel) shows (2Fo−Fc) and (Fo−Fc) electron density maps for selected residues above the si-face in molecule 1 of "phosphate" crystals of Lys265Arg. The corresponding data for molecules 1 and 3 of "PEG" crystals of Lys265Arg are shown in Figure S4.</p><!><p>Mutation of Lys265 to Met results in a ~104-fold decrease in the rate of the reaction of reduced MSOX with oxygen (4). The mutation does not affect the overall structure of the enzyme. The conformation of residues at the active site for sarcosine oxidation and the proposed oxygen activation site on the opposite face of the flavin ring is nearly identical in Lys265Met and wild-type MSOX. The results provide compelling evidence that the enormous decrease in oxygen reactivity observed with Lys265Met is attributable to the loss of a basic residue at position 265 that is likely to provide a pre-organized binding site for superoxide anion. A chemically conservative mutation of Lys265 to Arg results in a relatively modest but still substantial decrease (250-fold) in the rate of oxygen reduction (4). The overall structure of Lys265Arg and the configuration of residues at its two active sites are virtually identical to wild-type MSOX with the notable exception of the mutated residue. In Lys265Met, the side chain of residue 265 is nearly congruent with that of Lys265 in wild-type enzyme and exhibits the same configuration in each molecule of the mutant crystals. In sharp contrast, substitution of Lys265 with Arg has a profound effect on the position and mobility of the mutated residue. The side chain of Arg265 is shifted by about 4 to 5 A as compared with Lys265, points away from the flavin ring, and exhibits significant conformational variability in different molecules within the same crystal. The accessible surface area of an arginine residue in a peptide (225 A2) is about 10% larger than for lysine (200 A2) (38). The observed local structural perturbation suggests that the side chain of Arg265 is not readily accommodated in the space occupied by Lys265 in wild-type MSOX. The structures observed for "phosphate" or "PEG" Lys265Arg crystals strongly suggest that molecules containing a "flipped-out" Arg265 will comprise the major species in solution. These molecules are likely to exhibit negligible oxygen activation, similar to that observed upon mutation of Lys265 to a neutral residue. We postulate that the 400-fold higher oxygen reactivity observed with Lys265Arg compared with mutants lacking a basic residue at position 265 is attributable to a minor (< 1%) "flipped-in" conformer of Arg265 that exhibits oxygen reactivity similar to wild-type MSOX.</p><p>Mutation of Lys265 to Met or Arg causes only a very modest decrease (~ 5-fold) in the rate of sarcosine oxidation (Scheme 2, kred) (4), consistent with the observed structural integrity of the active site cavity above the re-face of the flavin ring. However, sarcosine oxidation by the mutant enzymes results in the formation of a novel intermediate (B) that exhibits an absorption spectrum significantly different from that expected for a 2-electron reduced flavin (Figure 5, curve 2). Intermediate B undergoes an apparent first-order conversion to a final reduced species (C) (Scheme 2, k2) that exhibits a typical fully reduced flavin spectrum (Figure 5, curve 3) (4). Intermediate B is not detected during reduction of wild-type MSOX (35). The absorption spectrum of substrate-reduced wild-type enzyme is, however, virtually identical to that observed for species C with the Lys265 mutants. The results strongly suggest that intermediate B is also formed during reduction of wild-type MSOX but is not detectable owing to a much faster conversion of the intermediate to species C (i.e., k2 ≥ 20kred). The value obtained for k2 with Lys265Met or Lys265Arg is about 25-fold slower than the value observed for kred with wild-type MSOX (4, 35). This analysis indicates that the Lys265 mutations probably cause at least a 500-fold decrease in the rate of conversion of intermediate B to species C.</p><p>MSOX oxidizes the anionic form of sarcosine (CH3NHCH2CO2−) (37). Transfer of a hydride equivalent from the substrate methyl group to N(5) of FAD will generate a reduced enzyme complex with the protonated form of sarcosine imine (EFADH−·CH2=NH+CH2CO2−). Charge-transfer interaction between the electron-rich 1,5-dihydroflavin anion and the positively charged imimium group could account for the "atypical" absorption spectrum observed for intermediate B. Interestingly, a spectrally similar intermediate has been observed with the wild-type forms of at least two other flavoprotein oxidases (dimethylglycine oxidase, alditol oxidase) and analogously attributed to a reduced enzyme·product complex (39, 40).</p><p>Release of the protonated imine from intermediate B would eliminate charge-transfer interaction and might account for the spectral properties observed for species C. Two sets of data, however, indicate that species C cannot be free reduced enzyme. (i) Species C is formed at a catalytically significant rate during reduction of wild-type MSOX (35). Steady-state kinetics studies indicate that a reduced enzyme·imine complex is the species that reacts with oxygen (7). (ii) Conversion of intermediate B to species C is estimated to be at least 500-fold slower with the mutant enzymes. If this step involved product release, the mutant enzymes should exhibit a large decrease in the binding affinity for substrate or substrate analogs. The latter is not expected based on the observed structural integrity of the sarcosine binding site nor is it consistent with the very modest decrease (< 5-fold) in stability observed for complexes of the mutant enzymes with sarcosine or methylthioacetate (4).</p><p>Mutation of Lys265 to Met or Arg results in the apparent loss of two water molecules on the si-face of the flavin ring. WAT 1 forms part of an apparent proton relay system that extends from the N(5) position of FAD to bulk solvent (see Figure 1). WAT 1 is found in each molecule of all previously determined structures of wild-type MSOX and two other mutants. We propose that conversion of intermediate B to species C involves ionization of the protonated sarcosine imine (Scheme 2), a step that would eliminate charge-transfer interaction between the reduced flavin and the imine. Importantly, proton release is likely to be at least 100-fold slower in the Lys265 mutants, as judged by results obtained upon disruption of a proton relay network in dihydroorotate dehydrogenase (41). Experiments to evaluate this structure-based mechanism are a subject for future studies.</p><!><p>Mutation of Lys265 to Met or Arg results in the elimination or displacement of a positive charge from a pocket above the si-face of the flavin ring. This results in the loss of a charge-stabilized preformed binding site for the superoxide anion, an intermediate in the oxidative half-reaction of MSOX, and a greatly reduced rate of oxygen reduction. At the same time, the mutations disrupt a water relay system near the flavin N(5) atom that probably catalyzes the ionization of a protonated sarcosine intermediate, leading to a transient accumulation of a charge-transfer intermediate along the pathway of product release from the reoxidized enzyme.</p><!><p>View of the region above the si-face of the flavin ring in wild-type MSOX (PDB entry code 2GBO). Carbons are white, oxygens are red and nitrogens are blue. Waters are shown as balls. Waters 3-5 are in contact with bulk solvent. Hydrogen bonds are indicated by dashed lines.</p><p>Stereo ribbon drawing comparing the two molecules in "phosphate" crystals of Lys265Met. Molecule 1 is shown as a green ribbon, except for white active site and external loops. Molecule 2 is shown as a cyan ribbon, except for magenta active site and external loops. FAD is shown in spacefill with carbon, nitrogen and oxygen atoms colored yellow, blue and red, respectively. The active site and external loops are indicated by red and black arrows, respectively.</p><p>Stereoview comparison of the sarcosine oxidation site above the re-face of the flavin ring in Lys265Met or Lys265Arg with wild-type MSOX (PDB entry code 2GBO). Oxygen and nitrogen atoms are colored red and blue, respectively, except for water molecules in wild-type MSOX (WAT 1-8) which are shown as cyan balls. The top panel compares molecule 1 in Lys265Met "phosphate" crystals (green carbons) with molecule 1 in wild-type MSOX "phosphate" crystals (yellow carbons). The bottom panel compares molecule 1 in Lys265Arg "phosphate" crystals (green carbons) with molecule 1 in wild-type MSOX "phosphate" crystals (yellow carbons).</p><p>Stereoview comparison of the region above the si-face of the flavin ring in Lys265Met or Lys265Arg with wild-type MSOX (PDB entry code 2GBO). Two water molecules, shown as red balls in each panel, are found in wild-type MSOX but not in the two Lys265 mutants. The two alternate conformations of Asp47 (A and B) found in wild-type MSOX molecule 1 are shown in the bottom panel; only the A conformation is shown in the top and middle panels. Selected hydrogen bonds are indicated by dashed lines. The top panel compares molecules 1 and 2 in "phosphate" Lys265Met crystals (green and cyan carbon atoms, respectively) with molecule 1 in wild-type MSOX (yellow carbons). The middle panel compares molecules 1 and 2 in "phosphate" Lys265Arg crystals (green and cyan carbon atoms, respectively) with molecule 1 in wild-type MSOX (yellow carbons). For clarity, only the A and A-like conformations of Asp47 are shown for molecules 1 and 2, respectively, in the mutant enzyme. The bottom panel compares molecules 1 to 4 in "PEG" Lys265Arg crystals (green, magenta, cyan and white carbons, respectively) with molecule 1 in wild-type MSOX (yellow carbons).</p><p>The absorption spectrum of oxidized Lys265Met is shown in curve 1. Curve 3 is the final absorption spectrum of reduced Lys265Met (species C) observed after anaerobic reaction with 140 mM sarcosine in 100 mM potassium phosphate buffer, pH 8.0, at 25 °C. Curve 2 is the calculated absorption spectrum of intermediate B. The data were taken from Zhao et al.(4). Similar results are obtained with Lys265Arg (4).</p><p>Reduction of molecular oxygen by reduced flavin. Following the initial rate-determining step (rds), conversion of the radical pair to oxidized flavin and hydrogen peroxide may or may not proceed via a 4a-peroxyflavin intermediate (not shown). This intermediate is not detected during oxidative half-reaction studies with MSOX (4).</p><p>Proposed mechanism for the reduction of MSOX by sarcosine (S = sarcosine, ImH+ = protonated sarcosine imine; Im = unprotonated sarcosine imine).</p><p>Summary of Data Collection and Refinement for the Lys265Met and Lys265Arg Mutants of MSOX</p><p>Rmerge = ΣhΣi_I(h)−Ii(h) iff; | ΣhΣiIi(h), where Ii(h) and I(h) are the ith and mean measurements of reflection h.</p><p>I/σ(I) is the average signal to noise ratio for merged reflection intensities.</p><p>R = Σh_Fo−Fc_/Σh_Fo_, where Fo and Fc are the observed and calculated structure factor amplitudes of reflection h.</p><p>Rfree is the test reflection data set, about 5 % selected randomly for cross validation during crystallographic refinement (31).</p><p>Root-mean-squared deviation (Rmsd) from ideal bond lengths and angles and Rmsd in B-factors of bonded atoms.</p><p>mm, main chain to main chain; ms, main chain to side chain, ss, side chain to side chain.</p><p>For Lys265Met ("phosphate") the residues in alternate conformations are AGlu141, AArg154, ASer191, and BGlu141; for Lys265Arg ("phosphate") the residues in an alternate conformation are AAsp47, AArg154, ASer358 and BAsp47.</p><p>Comparison of Lys265Met or Lys265Arg Crystals with Wild-type MSOX or Arg49Lys Crystals</p><p>Comparison is made between molecules in mutant and wild-type crystals that exhibit the same configuration for the active site and external loops. If the latter is not possible, the molecule in the mutant crystal is compared with each of the two molecules in wild-type crystals.</p><p>PDB entry code 2GBO.</p><p>PDB entry code 3BHK.</p><p>PDB entry code 3BHF.</p>

PubMed Author Manuscript

Fatty acid composition of lipids in pot marigold (Calendula officinalis L.) seed genotypes

BackgroundCalendula officinalis L. (pot marigold) is an annual aromatic herb with yellow or golden-orange flowers, native to the Mediterranean climate areas. Their seeds contain significant amounts of oil (around 20%), of which about 60% is calendic acid. For these reasons, in Europe concentrated research efforts have been directed towards the development of pot marigold as an oilseed crop for industrial purposes.ResultsThe oil content and fatty acid composition of major lipid fractions in seeds from eleven genotypes of pot marigold (Calendula officinalis L.) were determined. The lipid content of seeds varied between 13.6 and 21.7 g oil/100 g seeds. The calendic and linoleic acids were the two dominant fatty acids in total lipid (51.4 to 57.6% and 28.5 to 31.9%) and triacylglycerol (45.7 to 54.7% and 22.6 to 29.2%) fractions. Polar lipids were also characterised by higher unsaturation ratios (with the PUFAs content between 60.4 and 66.4%), while saturates (consisted mainly of palmitic and very long-chain saturated fatty acids) were found in higher amounts in sterol esters (ranging between 49.3 and 55.7% of total fatty acids).ConclusionsAll the pot marigold seed oils investigated contain high levels of calendic acid (more than 50% of total fatty acids), making them favorable for industrial use. The compositional differences between the genotypes should be considered when breeding and exploiting the pot marigold seeds for nutraceutical and pharmacological purposes.

fatty_acid_composition_of_lipids_in_pot_marigold_(calendula_officinalis_l.)_seed_genotypes

Background<!>Oil contents<!><!>Fatty acid composition<!><!>TL fatty acids<!><!>TL fatty acids<!><!>TAG fatty acids<!>PL and SE fatty acids<!>Conclusions<!>Seeds and chemicals<!><!>Seeds and chemicals<!>Oil extraction and fractionation<!>Fatty acid analysis<!>Statistics<!>Abbreviations<!>Competing interests<!>Authors’ contributions<!>Acknowledgements

<p>Calendula officinalis L. (pot marigold), a member of the Asteraceae family, is an annual aromatic herb with yellow or golden-orange flowers, native to the Mediterranean climate areas, being also successfully cultivated in temperate regions of the Earth for ornamental and medicinal purposes [1]. The species have been reported to contain a variety of phytochemicals, including carbohydrates, lipids, phenolic compounds, steroids, terpenoids, tocopherols, carotenoids and quinones [2-5] with potential health benefits [1,6-10].</p><p>Besides the usual fatty acids, a few plants are capable to biosynthesize some unusual fatty acids, with special chemical structure. Usually these fatty acids accumulate in storage tissues, while in green organs they are absent or present in very small amounts. The presence of unusual fatty acids is genetically determined and they are highly significant indicators of phylogenetic relationships [11,12]. The seeds of pot marigold have a significant oil content (around 20%), of which about 60% is the unusual calendic acid (8 t, 10 t, 12c-18:3) [13-16]. Several studies demonstrated that calendic acid is synthesized in Calendula seeds via desaturation of linoleic acid [17-21]. Due to its special structure – with three conjugated double bonds – calendic acid and Calendula seeds oil exhibit interesting chemical and physiological properties.</p><p>The seed oils such of Calendula officinalis L., Momordica charantia L. or Aleurites fordii Hemsl., rich in conjugated linolenic acids (CLNAs) have a high rate of oxidation and are used as raw materials in paints and coatings industry, and have applications in the manufacture of cosmetics and some industrial polymers [19,22-24]. For these reasons, in the last few years, a concentrated research effort in Europe has been directed towards the development of Calendula officinalis L. as an oilseed crop for industrial purposes [25] and for the engineering of transgenic plants containing the metabolic route for the conjugated fatty acids biosynthesis [26,27].</p><p>The increasing interest for plants producing conjugated fatty acids is also motivated by the recent findings related to their biological effects. It has been shown that CLNAs have an important body fat-lowering effect [28] and possess anti-carcinogenic properties, exhibiting apoptotic activity against a wide variety of tumor cells, such as the U-937 human leukemic cancer cell line and the colon cancer cells (Caco-2) [24,29,30]. Bhaskar et al. [31] observed that the trans CLNAs exhibited stronger growth inhibition and more DNA fragmentation in human colon cancer cells than corresponding cis CLNA isomers.</p><p>To our knowledge, all the studies, excepting two short reports of Ul'chenko et al. [32] and Pintea et al. [33], respectively, conducted on marigold seed oils determined the fatty acid contents by analyzing only the total lipid matrix.</p><p>Therefore, the aim of the present investigation was to compare the oil content and fatty acid compositions of total lipids (TLs), triacylglycerols (TAGs), polar lipids (PLs) and sterol esters (SEs) in seeds of eleven pot marigold genotypes from six different locations in Europe, grown in the Transylvanian region (Romania). The information obtained is helpful to identify suitable genotypes for use in breeding programs of Calendula officinalis.</p><!><p>The oil (total lipids) contents in eleven genotypes of pot marigold (Calendula officinalis L.) (CO) seeds are presented in Figure 1.The values were found to vary between 13.6- 21.7 (g oil/100 g seeds). There were no significant differences (p < 0.05) among genotypes, except for oil contents of samples CO4 and CO6 versus CO9. The highest amounts of oils were found in the CO4 (21.7 g/100 g), CO6 (21.5 g/100 g) and CO11 (21.3 g/100 g), whereas the genotypes CO1 (15.5 g/100 g), CO5 (15.3 g/100 g) and CO9 (13.6 g/100 g), exhibited the lowest contents of the TLs. These values were similar to those reported by Cromack and Smith [25] but much higher than those observed by Ozgul- Yucel (5.9% oil in Turkish Calendula seeds) [34] and Angelini et al. (5.4% oil in Italian CO seed crops from 1994) [35]. The TLs content of the analyzed CO seeds in this study were also comparable with those of some non-conventional vegetable oil sources with unique phytochemical compositions, such as bitter gourd (21%), cherry laurel (18.3%), pomegranate (18.1%), blackthorn (16.5%), linseed dodder (15.5-20.7%), and coriander (12.7-18%) seeds [34,35].</p><!><p>The oil content of Pot marigold (Calendula officinalis L.) seeds. CO1- CO11, pot marigold (Calendula officinalis L.) genotypes. Results are given as mean ± SD (n = 3); * - significant difference, p < 0.05 (using "Kruskal-Wallis non-parametric test" followed by "Dunn's Multiple Comparison Test").</p><!><p>The total lipid fatty acid composition as well as the fatty acid composition of TAGs, PLs and SEs of the analyzed pot marigold seed oils is presented in Tables 1 and 2.</p><!><p>Fatty acid composition (%) in total lipids and individual lipid classes of different genotypes of pot marigold seed oils</p><p>The values represent the means of three samples, analyzed individually in triplicate (n = 3x3).</p><p>CO1- CO11, pot marigold (Calendula officinalis L.) genotypes.</p><p>TL- total lipids, TAG- triacylglycerols, PL- polar lipids, SE- sterol esters, nd- not detected, tr- trace.</p><p>The composition (%) of fatty acid classes in total lipids and major lipid fractions from different genotypes of pot marigold seed oils</p><p>Values are given as mean ± SD of three samples, analyzed individually in triplicate (n = 3x3).</p><p>Means in the same column followed by different subscript letters indicate significant differences (p < 0.05) among lipid classes of each genotype of pot marigold (ANOVA "Tukey's Multiple Comparison Test"). TL - total lipids, TAG- triacylglycerols, PL- polar lipids, SE- sterol esters. SFAs- saturated fatty acids, MUFAs- monounsaturated fatty acids, PUFAs- polyunsaturated fatty acids, VLCSFAs- very long chain saturated fatty acids, ∑CLNAs [18:3 (8trans, 10trans, 12cis) + 18:3 (8trans,10trans,12trans)] - conjugated linolenic acids.</p><!><p>Nineteen fatty acids were identified in the studied pot marigold seed oils (Figure 2), including very low amounts of a hydroxy fatty acid, namely 9- hydroxy- trans-10, cis-12 octadecadienic-acid (9-HODE).</p><!><p>GC-MS chromatogram of FAMEs in the TLs of Calendula officinalis L. (CO2: cv. Prolifera nr. 214) seeds analysed with a BPx- 70 capillary column. Peaks: (1) lauric (12:0), (2) myristic (14:0), (3) pentadecanoic (15:0), (4) cis-7 hexadecenoic [16:1 (n-9)], (5) palmitoleic [16:1 (n-7)], (6) palmitic (16:0), (7) margaric (17:0), (8) linoleic [18:2 (n-6)], (9) oleic [18:1 (n-9)], (10) elaidic [18:1 (9 t) (n-9)], (11) linoelaidicic [18:2 (9 t,12 t) (n-6)], (12) stearic (18:0), (13) α- linolenic [18:3 (n-3)], (14) calendic [18:3 (8 t, 10 t, 12c) (n-6)], (15) gondoic [20:1 (n-9)], (16) β- calendic [18:3 (8 t, 10 t, 12 t) (n-6)], (17) arachidic (20:0), (18) 9- hydroxy- trans-10, cis-12 octadecadienic (9- HODE), (19) behenic (22:0) acids.</p><!><p>As expected, calendic acid [18:3 (8 t, 10 t, 12c) (n-6)] was the predominant polyunsaturated fatty acid (PUFA) in all TL extracts, and its composition varied between 51.47% (in CO8) and 57.63% of total fatty acids (in CO4). The next most abundant fatty acid was linoleic acid [18:2 (n-6)] (28.50 to 31.86%), followed by oleic [18:1 (n-9)] (4.44 to 6.25%) and palmitic acids (16:0) (3.86 to 4.55%). Small and very small (or trace) amounts (<2%) of stearic (18:0), β- calendic [18:3 (8 t, 10 t, 12 t) (n-6)], elaidic [18:1 (9 t) (n-9)], arachidic (20:0), behenic (22:0), gondoic [20:1 (n-9)], α- linolenic [18:3 (n-3)], linoelaidicic [18:2 (9 t,12 t) (n-6)], cis-7 hexadecenoic [16:1 (n-9)], palmitoleic [16:1 (n-7)], lauric (12:0), myristic (14:0), pentadecanoic (15:0), and margaric (17:0) acids were also determined. Similar results for the calendic acid content (over 50%) were reported by Cromack and Smith [25] for two of nine hybrids of pot marigold seeds grown in England, as well as by Cahoon et al. [26]. Ozgul- Yucel concluded that Turkish calendula seed oil is characterized by high concentration of linoleic acid (43.5%) and low content of CLNAs (calendic acid (18.3%) + β- calendic (11.2%)) [34]. Moreover, the calendic acid levels reported here are considerably higher than those reported previously by Suzuki et al. [29] (33.4%) and Angelini et al. [35] (16- 46%- in the Italian pot marigold seed oils, crops from 1993).</p><p>The available literature shows that the fatty acid composition of oil seeds varies strongly according to their origin/genotype, and geographical/climatic conditions of the growth areas [25,36]. It was also found that the maturity stage of the seeds is an important factor that influences the accumulation of calendic acid in calendula seeds oil. Pintea et al. [33] showed that during the maturation period of the pot marigold seeds (0–2 weeks after flower drops) the concentration of calendic acid increased sharply and steadily (from 8.62% to 53%), accompanied by a decrease in the amounts of linoleic and oleic acids. These observations are in agreement with the presence of the specific conjugase which is able to convert linoleic acid into calendic acid in Calendula seeds [18,26]. The stereospecific analysis of TAG proved that calendic acid preferentially esterifies the sn-2 position of TAG [26,37].</p><p>The analysis of fatty acids classes showed statistically significant differences (p < 0.05) with the exception of PUFAs (Figure 3). The highest value of saturated fatty acid (SFAs) (p < 0.05) was registered in the TLs of Czech genotypes (CO11) (7.34%), whereas CO5, CO7 and CO8 were the richest sources of monounsaturated fatty acids (MUFAs) (Figure 3A). On the other hand, small variations (p < 0.05) were found in CLNAs contents (Figure 3B), with the highest proportions in CO4 (58.54%) and the lowest in CO8 (51.95%), respectively. As shown in Table 2, the levels of the PUFAs/SFAs (saturated fatty acids) ratios were significantly higher (p < 0.05) in TLs (due to the high values of 18:3 and 18:2 fatty acids) than in the lipid fractions (TAGs, PLs and SEs) of each pot marigold genotypes.</p><!><p>Comparative representation of fatty acid classes from total lipids of different genotypes of pot marigold (Calendula officinalis L.) seed oils. CO1- CO11, pot marigold (Calendula officinalis L.) genotypes. Values are mean ± SD of three samples, analyzed individually in triplicate (n = 3x3). Values with different letters (a-e) are significantly different (p < 0.05), using ANOVA "Tukey's Multiple Comparison Test". SFAs- saturated fatty acids, MUFAs- monounsaturated fatty acids, PUFAs- polyunsaturated fatty acids, CLNAs- conjugated linolenic acids.</p><!><p>The fatty acid profiles of the TAGs were similar to that of the profiles of the TL fractions, due to the dominance of the PUFAs (18:3 and 18:2 (n-6) fatty acids) in their compositions (see Tables 1 and 2) and due to the fact that TAG are major components of the seeds oil.</p><!><p>The fatty acid composition of the PLs and SEs was different from that of the TL and TAG fractions in all the pot marigold genotypes analyzed (Tables 1 and 2).</p><p>The PL fractions were highly unsaturated, with the linoleic acid content ranging from 55.02% (CO2) to 61.51% (CO10) of total fatty acids. Ul'chenko et al. [32] studied the fatty acid compositions of the lipids from seeds, leaves and flowers of Calendula officinalis L. and reported lower value of linoleic acid (24.5%) in the phospholipids of seeds, than those determined in the present work.</p><p>With four exceptions (samples CO1-4), the calendic acid content in the PL fractions was lower than 3% (Table 1). This conjugated fatty acid was found to be below 1% in the phosphatidylcholine (PC) of Calendula officinalis seeds oil [26] or was not detected [32]. The differences between the reported data and our data can be explained by the fact that we have investigated the total polar lipids fraction which includes phospholipids and glycolipids. Transgenic soybean and Arabidopsis seeds engineered to synthesize calendic acid (by cloning of the fatty acid conjugase from Calendula) accumulated moderate level of conjugated fatty acids. Calendic acid was found at comparable levels in PC and TAG fractions (85% in the sn-2 position of PC) proving that complex mechanisms involving both desaturation and transacylation processes are involved in the biosynthesis of rich CLNAs enriched TAG [26]. Same authors showed that accumulation of conjugated fatty acids in PC of transgenic plants (soybean and Arabidopsis) negatively affected the appearance and the germination rate of seeds due to the special chemical and physical properties of CLNAs. In consequence, the selection of valuable genotypes of Calendula which are able to produce large amounts of oil enriched in CLNAs still has an economical importance.</p><p>The levels of SFAs in SEs were significantly higher (p < 0.05) than in the corresponding lipid fractions of each genotype (Table 2). The amounts of saturated consisted mainly of palmitic (16:0) acid, very long-chain saturated fatty acids (VLCSFAs) (more than 20 carbon atoms) and stearic (18:0) acid, respectively, and varied between 49.31% (in CO1) and 55.74% (in CO8) of total fatty acids from SEs (Tables 1 and 2). These observations are in agreement with the data reported by Zlatanov [38], Kallio et al. [39] and Yang et al. [40] about the fatty acid composition of the phospholipids and the SE fractions of other non-conventional seed oils.</p><p>In plant tissues, the very long-chain fatty acids (≥20 carbon atoms) are precursors for the synthesis of lipids, such as cuticular waxes (on the aerial plant surfaces), suberin (embedded in the cell walls of plant-environment interfaces), triacylglycerols (in seeds), and ceramides (in the cell membranes) [41,42].</p><p>The TL, TAG and PL fractions of all analyzed pot marigold genotypes exhibited very low proportions of VLCSFAs (<1.50% of total fatty acids), whereas the SEs showed significantly higher (p < 0.05) amounts of this type of fatty acids (from 20.59% (CO4) to 24.78% (CO11)) (Table 2).</p><p>As shown in Table 2, in all extracts of pot marigold seeds, the PUFAs/SFAs ratios were significantly lower (p < 0.05) in SE and PL fractions than in the corresponding TLs or TAGs. A comprehensive study of the Diabetes and Nutrition Study Group of the Spanish Diabetes Association showed that a dietary PUFAs/SFAs ratio > 0.4 can greatly reduce the risk of onset of diabetic complications [43]. Moreover, in some earlier reports, the authors indicate that the values of this ratio comprised between 1.0 and 1.5, are optimal to reduce the risk of cardiovascular diseases [44,45]. Thus, the results of the present study show that the Calendula officinalis oil, whatever the genotype analyzed in this paper, may reduce the risk of cardiovascular diseases because both TLs and TAG presented PUFAs/SFAs ratios values are closed to the recommended PUFA/SFA intake by nutrition scientists.</p><!><p>In the present paper, seeds of eleven genotypes of Calendula officinalis L. originating from six different locations in Europe, cultivated in Romania (Transylvanian) were analyzed with respect to oil yields and fatty acid contents. To the best of our knowledge, data about detailed fatty acid composition of main lipid fractions in pot marigold seeds investigated in this study are not available in literature.</p><p>The oil content observed in most of the calendula seed samples studied was noted to range between 18 and 22 g oil/100 g seeds. The oil TAGs were similar in fatty acid composition to the TLs, containing substantial amounts of calendic and linoleic acids, making them excellent dietary sources of PUFAs, especially of CLNAs. The PL fractions were highly unsaturated, due to the dominance of the linoleic acid in their structures. A clear characteristic of the SEs from the pot marigold seed oils analyzed were the significantly high levels of SFAs, with considerable amounts of VLCSFAs.</p><p>The compositional differences between the genotypes should be considered when breeding and exploiting the calendula seeds for industrial, nutraceutical or pharmacological purposes.</p><!><p>Eleven genotypes of Calendula officinalis L. originating from six different locations in Europe (botanical gardens and institutes) (Table 3) were cultivated on experimental fields of the University of Agricultural Sciences and Veterinary Medicine of Cluj- Napoca (Romania). The crops were established in the first half of May 2011, to a target population of 40 plants m-2. Plot area was prepared before (autumn of 2010) by fertilization with animal manure. Nitrogen- based fertilizers were applied during the vegetation period. The seeds were harvested manually at full maturity (end of September-beginning of October).</p><!><p>Genotypes of Calendula officinalis L. (CO) evaluated</p><!><p>All reagents (used for the oil extraction, fractionation and fatty acid methyl esters (FAMEs) preparation) and lipid standards (used for identification of the lipid class) were of chromatographic grade (Sigma–Aldrich (St. Louis, MO, USA)). The thin layer chromatography (TLC) plates (silica gel 60 F254, 20 × 20 cm) were purchased from Merck (Darmstadt, Germany).The FAMEs standard (37 component FAME Mix, SUPELCO, catalog No: 47885-U) were purchased from Supelco (Bellefonte, PA, USA).</p><!><p>The oils were extracted from 5 g of seeds, using a methanol/chloroform extraction procedure, according to Yang et al. [36] and Dulf et al. [46]. The sample was homogenized in 50 mL methanol for 1 min using a homogeniser (MICCRA D-9, Germany), 100 mL chloroform was added, and homogenization was continued for further 2 min. The mixture was filtered under vacuum through a Buchner funnel and the solid residue was resuspended in 150 mL of chloroform: methanol (2:1, v/v) and homogenized for another 3 min. The mixture was filtered again and washed with 150 mL chloroform: methanol (2:1, v/v). The filtrates were combined and cleaned with 0.88% potassium chloride water solution and methanol: water (1:1, v/v) solution. The bottom layer (with the purified lipids) was filtered before the solvent was rotary evaporated. The total lipids recovered were transferred to vials with 4 mL chloroform (stock solution), and stored at −18°C for further analysis.</p><p>Neutral and polar lipid fractions were separated by TLC [47]. Lipid aliquots (0.2 ml of stock solution) were applied on the TLC plates and then developed in a mixture of petroleum ether: diethyl ether: acetic acid (85:15:1, v/v/v), sprayed with 2', 7'-dichlorofluoroscein/methanol (0.1% w/v) and viewed under UV light (254 nm) [48]. The lipid classes were identified using commercial standards and then scraped from the TLC plates. The first band (at the origin of the plates), corresponding to the PLs was eluted from silica layer with methanol: chloroform (1:1, v/v), and the upper two major bands of TAGs and SEs respectively were eluted with chloroform. The samples were filtered, the solvent was removed and the dry residue was subjected to transesterification and gas chromatographic (GC) analysis.</p><!><p>The total lipid, PL, TAG and SE fractions were derivatized by sodium methoxide catalysis [49]. The FAMEs were determined by gas chromatography–mass spectrometry (GC-MS), using a PerkinElmer Clarus 600 T GC-MS (PerkinElmer, Inc., Shelton, U.S.A.) equipped with a, BPx- 70 capillary column (60 m × 0.25 mm i.d., 0.25 μm film; SGE, Ringwood, Australia). The initial oven temperature was 140°C, increased to 220°C with a rate of 2°C/min and then held at this temperature for 25 min. Flow rate of the carrier gas He and the split ratio were 0.8 ml/min and 1:24, respectively. The injector temperature was 210°C. The positive ion electron impact (EI) mass spectra was recorded at an ionization energy of 70 eV and a trap current of 100 μA with a source temperature of 150°C. The mass scans were performed within the range of m/z: 22–395 at a rate of 0.14 scan/s with an intermediate time of 0.02 s between the scans. The injected volume was 0.5 μl. Identification of FAMEs was achieved by comparing their retention times with those of known standards (37component FAME Mix, SUPELCO # 47885-U) and the resulting mass spectra to those in our database (NIST MS Search 2.0).</p><!><p>Three different samples of Calendula seeds for each genotype were assayed. The analytical results reported for the fatty acid compositions, are the average of triplicate measurements of three independent oils (n = 3x3). The assumptions of equality of variances and normal distribution of errors were checked for the tested response variables. Since the assumptions were satisfied, data were subjected to one-way ANOVA (repeated measures ANOVA) and Tukey's post hoc test. Statistical differences among oil samples were estimated using: "Kruskal-Wallis non-parametric test" followed by "Dunn's Multiple Comparison Test" (Graph Pad Prism Version 4.0, Graph Pad Software Inc., San Diego CA). A probability value of p < 0.05 was considered to be statistical significant.</p><!><p>CLNAs: Conjugated linolenic acids; TLs: Total lipids; TAGs: Triacylglycerols; PLs: Polar lipids; SEs: Sterol esters; PUFAs: Polyunsaturated fatty acids; SFAs: Saturated fatty acids; MUFAs: Monounsaturated fatty acids; VLCSFAs: Very long-chain saturated fatty acids; PC: Phosphatidylcholine; CO: Calendula officinalis; FAMEs: Fatty acid methyl esters; TLC: Thin layer chromatography; GC-MS: Gas chromatography–mass spectrometry.</p><!><p>The authors declare that they have no competing interests.</p><!><p>FVD and DP carried out the experimental design, interpretation of results and preparation of the paper. ADB contributed to the extraction of lipids. AP contributed to the separation, identification and quantification of the lipid fractions and fatty acids from the samples. All authors read and approved the final manuscript.</p><!><p>This work was supported by the POSDRU/89/1.5/S/52432 project ("Organising the National Interest Postdoctoral School of Applied Biotechnologies with Impact on Romanian Bioeconomy") and by a grant of the Romanian National Authority for Scientific Research, CNCS – UEFISCDI, project number PN-II-ID-PCE-2011-3-0721.</p>

PubMed Open Access

Flexibility vs Preorganization: Direct Comparison of Binding Kinetics\nfor a Disordered Peptide and Its Exact Preorganized Analogues

Many intrinsically disordered proteins, which are prevalent in nature, fold only upon binding their structured partner proteins. Such proteins have been hypothesized to have a kinetic advantage over their folded, preorganized analogues in binding their partner proteins. Here we determined the effects of ligand preorganization on the kon for a biomedically important system: an intrinsically disordered p53 peptide ligand and the MDM2 protein receptor. Based on direct simulations of binding pathways, computed kon values for fully disordered and preorganized p53 peptide analogues were within error of each other, indicating little if any kinetic advantage to being disordered or preorganized for binding the MDM2 protein. We also examined the effects of increasing the concentration of MDM2 on the extent to which its mechanism of binding to the p53 peptide is induced fit vs conformational selection. Results predict that the mechanism is solely induced fit if the unfolded state of the peptide is more stable than its folded state; otherwise, the mechanism shifts from being dominated by conformational selection at low MDM2 concentration to induced fit at high MDM2 concentration. Taken together, our results are relevant to any protein binding process that involves a disordered peptide of a similar length that forms a single \xce\xb1-helix upon binding a partner protein. Such disorder-to-helix transitions are common among protein interactions of disordered proteins and are therefore of fundamental biological interest.

flexibility_vs_preorganization:_direct_comparison_of_binding_kinetics\nfor_a_disordered_peptide_and_

INTRODUCTION<!>METHODS<!>The Protein Model<!>Weighted Ensemble (WE) Simulations<!>Propagation of Dynamics<!>Calculation of Bimolecular Rate Constants<!>Calculation of the Percentage of Productive Collisions<!>RESULTS<!>Is There a Kinetic Advantage to Being Disordered vs Preorganized?<!>Effect of Including Hydrodynamic Interactions (HIs)<!>Effect of Increasing Receptor Concentration<!>DISCUSSION<!>CONCLUSIONS<!>

<p>Many proteins that are either partially or completely unfolded in their unbound states1,2 fold only upon binding their structured partner proteins. Such "intrinsically disordered" proteins (IDPs) have been proposed to have a kinetic advantage over their preorganized, folded analogues for binding their partners,3,4 which challenges the long-standing assumption that the preorganization of a ligand to its receptor-bound conformation results in a faster association rate constant (kon). Potential mechanisms by which this kinetic advantage might be achieved are (i) the "fly-casting" mechanism, in which the IDP collides more rapidly with the partner receptor due to a larger "capture" radius,3 and (ii) the "dock-and-coalesce" mechanism for IDPs with two or more segments in which the initial docking of one segment results in a more rapid, pseudointramolecular docking of the remaining segments.4 Throughout this work, the term "ligand" refers to a molecule (e.g., small molecule, peptide, or protein) that binds to a larger molecule that serves as the target receptor.</p><p>While experimental studies have provided informative insights about the effects of preorganization on the binding kinetics of IDP ligands,5–10 these studies have not been able to provide definitive proof of a kinetic advantage (or lack thereof) to being disordered vs preorganized. Existing experimental studies indicate differing results on the effect of ligand preorganization on binding kinetics. For example, preorganization has resulted in faster binding for certain IDPs (ACTR and Y507A mutant of the E3 rRNase domain),5,6 and no significant effect on the binding kinetics for other IDPs (PUMA and c-Myb).7,8 In addition, an unfolded variant of the Fyn SH3 domain that was engineered via truncation of only four residues has achieved the same kon as the full-length, folded domain for a high-affinity peptide,9 and the preorganization of the disordered monomers of an engineered GCN4-p1 leucine zipper variant has resulted in slower dimerization.10 Ideally, the effect of ligand preorganization on binding kinetics would be assessed by engineering peptide analogues that differ only in their degree of preorganization without altering the chemical structures, which is not possible in experiments.</p><p>Molecular simulations provide the only practical means to compute kon values for both IDPs and their exact preorganized analogues—which have been engineered in silico—by directly generating the corresponding binding pathways. Furthermore, while experiments can typically measure only the kon, simulations can be used to directly compute the rate constants of individual steps. However, due to the relatively long time scales of protein binding processes, only one simulation study has reported atomistic binding pathways along with the kon for an IDP ligand (p53 peptide) and its protein receptor (MDM2), and these simulations did not sample fully disordered analogues of the ligand.11 Both atomistic and residue-level models have been used to characterize solely the late stages of binding, i.e., after the IDPs have collided with their partner proteins.12–14 Residue-level simulation studies of binding pathways for IDPs have been reported,15,16 including the only study that has determined the effects of preorganization on the binding kinetics of an IDP, focusing on the intrinsically disordered, phosphorylated KID (pKID) domain and its folding into a pair of linked-together α-helices upon binding the KIX protein.16</p><p>Here, we focused on an IDP ligand that adopts a single α-helix upon binding its folded protein receptor: the intrinsically disordered, N-terminal peptide fragment of tumor suppressor p53 and MDM2 protein. We determined the effects of ligand preorganization on the kon by directly simulating binding pathways of the disordered p53 peptide and several of its exact analogues with various extents of preorganization. In addition, we used the computed kon values to predict the effect of increasing the concentration of MDM2 on the extent to which the binding mechanism proceeds through induced fit and conformational selection. Based on atomistic simulations, the binding mechanism of the MDM2 receptor and p53 peptide ligand is predicted to shift from being dominated by conformational selection at low receptor concentration to induced fit at high receptor concentration.17 Likewise, based on experimental rate constants, this shift in mechanism is expected to occur upon increasing the ligand concentration for systems involving disordered protein receptors and their small organic ligands.18,19 Given the prevalence of single α-helix binding motifs among protein–ligand interactions,20 the mechanism of MDM2–p53 binding is not only of biomedical importance21 but fundamental to biology.</p><!><p>Key features of our simulation strategy are the following. First, we employed minimal residue-level models (Cα models) along with a Gō-type potential energy function,22,23 which enables tuning of the extent of preorganization of the IDP (in our case, the p53 peptide) from fully disordered to fully preorganized. Second, dynamics were propagated using a Brownian dynamics algorithm with the inclusion of appropriately parametrized hydrodynamic interactions (HIs) between protein residues to yield realistic diffusion properties.24 Third, we applied the weighted ensemble (WE) path sampling strategy,25–27 which has been demonstrated to be orders of magnitude more efficient than standard Brownian dynamics simulations in generating pathways and rate constants for protein binding processes.28 Full details of the protein model, simulations, and analysis are below.</p><!><p>Residue-level protein models were used in which each residue was represented by a single pseudoatom at its Cα position, yielding 85 pseudoatoms for the MDM2 protein (residues 25–109) and 13 pseudoatoms for the p53 peptide (residues 17–29). Coordinates for the unbound and bound conformations of MDM2 and p53 peptide were taken from the crystal structure of the MDM2–p53 peptide complex (PDB code: 1YCR).29</p><p>A Gō-type potential energy function22,23 was used to govern the conformational dynamics of the protein model. In this energy function, bonded interactions between pseudoatoms are modeled by standard molecular mechanics terms for bonds, angles, and dihedrals; and nonbonded interactions between pseudoresidues separated by four or more pseudobonds were treated as either native or non-native contacts. A native contact was defined as a residue–residue contact in which the heavy atoms of the two residues are within 5.5 Å of each other in the crystal structure of the native complex. In addition to 57 intermolecular native contacts between p53 and MDM2, the p53 peptide and MDM2 consisted of 10 and 266 intramolecular contacts, respectively.</p><p>The protein model was parametrized by focusing separately on the following three contributions to the total energy function: (1)Etotal=Ep53+EMDM2+EMDM2/p53 where Ep53 and EMDM2 correspond to intramolecular contributions from p53 and MDM2, respectively, and EMDM2/p53 corresponds to the intermolecular MDM2/p53 contributions.</p><p>As others have done,30 we tuned the degree of structure and backbone flexibility of the IDP (in our case, the p53 peptide) by applying a single scaling factor α to the pseudoangle, pseudodihedral, and intramolecular nonbonded terms of the energy function involving solely the IDP: (2)Ep53=∑bondskbond(r-req)2+α{∑angleskangle(θ-θeq)2+∑dihedralsV1[1+cos(φ-φ1)]+V3[1+cos(3φ-φ3)]+∑i<j-4,non-nativep53εnon-native(σijnon-nativerij)12+∑i<j-4,nativep53εnative[5(σijnativerij)12-6(σijnativerij)6]} where r, θ, φ are pseudo bond lengths, pseudoangles, and pseudodihedrals, respectively; V1 and V3 are potential barriers for the dihedral terms; εnative is the energy well depth for native contacts, rij is interatomic distance between pseudoatoms i and j during simulation, and σijnative is the corresponding distance in the crystal structure; σijnon-native and εnon-native for non-native contacts were set to 4.0 Å and 1 kcal/mol, respectively. Equilibrium bond lengths (req), angles (θeq), and dihedral phase angles (φ1 and φ3) were taken from the crystal structure. The force constants, kbond and kangle, were set to 100 kcal/mol/Å and 20 kcal/mol/rad, respectively, and V1 and V3 were set to 1 and 0.5 kcal/mol, respectively. The scaling factor α was set to 0.1, 0.5, 1.0, and 2.0 to model analogues of the p53 peptide that exhibit, on average, a fraction of native contacts (Qp53) of 0.25, 0.5, 0.85, and 0.99, respectively, based on 10 μs standard simulations of the isolated peptide (Figures S1 and S2). Thus, α values of 0.1 and 2.0 represent the fully disordered and fully preorganized versions of the p53 peptide, respectively.</p><p>The same potential function was used for MDM2 (EMDM2) and nonbonded MDM2–p53 interactions (EMDM2/p53), except for the omission of the scaling factor α. An εnative of 1.0 kcal/mol was used for intramolecular native contacts of MDM2, yielding a fraction of native contacts QMDM2 > 0.8 based on five 10 μs simulations. To ensure that the fully disordered p53 peptide folds upon binding MDM2, the εnative for native MDM2–p53 interactions was set to the minimum value (2.0 kcal/mol) required to ensure that the peptide folds upon binding MDM2 (Qp53 > 0.7 throughout a 10 μs standard simulation (no WE sampling); Figure S3). Following others,16 the same εnative value for intermolecular contacts (in our case, MDM2–p53 contacts) was used for all analogues of the IDP (the p53 peptide). The same εnative was also used for native contacts within the fully preorganized p53 peptide.</p><!><p>To generate MDM2–p53 peptide binding pathways, we applied the WE path sampling strategy,31 as implemented in the WESTPA software package (https://westpa.github.io/westpa),32 to orchestrate a large set of Brownian dynamics trajectories that were carried out using the framework of the Northrup–Allison–McCammon (NAM) method.33 In this hybrid WE/NAM approach, two concentric spherical surfaces are first defined with radii b and q that correspond to separation distances between MDM2 and the p53 peptide. The inner sphere, or b surface, represents the initial unbound state, and the outer sphere, or q surface, represents a much larger separation distance (q ≫ b) at which trajectories are terminated to avoid wasting computing time sampling any indefinite drifting apart of the binding partners. The next step of the WE/NAM approach is to define a progress coordinate between the unbound and bound states and to divide this coordinate into bins with the goal of populating each bin with N trajectories, each of which is assigned a statistical weight. Starting from N trajectories in the initial unbound state, the dynamics of each trajectory are simultaneously propagated in parallel and occasionally coupled by replication and combination events at fixed time intervals τ based on their progress toward the target state (e.g., the bound state), splitting and combining the statistical weights, respectively, such that no bias is introduced into the dynamics.31 To maintain steady-state conditions, any trajectory that reaches the q surface is "recycled" by terminating the trajectory and starting a new trajectory from an initial, unbound state with the same statistical weight.</p><p>In our WE simulations, the radii b and q were set to 35 and 50 Å, respectively; as required for the WE/NAM approach, b is sufficiently large such that the intermolecular forces between the binding partners can be assumed to be isotropic (as mentioned above, only short-range residue–residue interactions were modeled in our simulations). Initial unbound states were generated by randomly reorienting the binding partners with respect to each other at a separation of 35 Å using their corresponding conformations from the crystal structure of MDM2–p53 complex.29 For the progress coordinate, we used the Cα RMSD of the p53 peptide after alignment of MDM2 ranging from 0 to 100 Å. This progress coordinate was evenly divided into 29 bins with a target number of 6 trajectories/bin, yielding a maximum total of 390 trajectories at any point in the WE simulation. The fixed time interval τ for each WE iteration was set to 100 ps, which allowed for at least one trajectory to advance to the next bin after each WE iteration.</p><p>For each p53 peptide analogue (each α value), 10 independent WE simulations of the MDM2–p53 binding process were carried out under pseudoequilibrium conditions in which trajectories were recycled at the q surface, but not the bound state, to allow for refinement of the bound-state definition after completion of the simulations. Once this was refined, we effectively recycled trajectories that reached the refined definition of the bound state by removing the trajectories from subsequent analysis with proper renormalization of the remaining probabilities. This renormalization was straightforward given that no trajectories in the reverse, unbinding direction were generated in our Gō-type simulations. Each WE simulation was carried out for a maximum trajectory length of 200 ns (2000 WE iterations), which was sufficiently long for obtaining converged estimates of the kon (Figure S5). Conformations were sampled every 1 ps for analysis.</p><!><p>The dynamics of our WE simulations were propagated using a standard Brownian dynamics algorithm34 with the inclusion of hydrodynamic interactions (HI),24 as implemented in the UIOWA_BD software.24,35 Hydrodynamic radii were set to 5.3 Å, which has been found to reproduce the translational and rotational diffusion coefficients of all-atom models of folded proteins when using the residue-level models of this study.24 The solvent viscosity was set to 0.89 cP to represent water at 25 °C. To enable the use of a 50 fs time step, all pseudobonds between residues were constrained to their native bond lengths by applying the LINCS algorithm.36</p><!><p>All bimolecular rate constants k were calculated using the Northrup–Allison–McCammon (NAM) equation:33</p><p> (3)k=kD(b)β1-(1-β)kD(b)/kD(q) where kD(r) is the diffusion rate constant for the two binding partners achieving a separation distance r, and β is the probability that a simulation starting from the unbound state with a separation distance of b (35 Å) reaches the target state before drifting apart to a separation distance of q (50 Å). To calculate the rate constant k1, the target state is the encounter complex; likewise, to calculate kon, the target state is the native, bound state (see definitions in Results).</p><p>Assuming that the motions of the two binding partners are isotropic, the diffusion rate constants were calculated using the Smoluchowski equation: kD = 4πDr, where D is the relative translational diffusion coefficient of the two partners (i.e., the sum of their corresponding diffusion coefficients). Therefore, eq 3 reduces to</p><p>The translational diffusion coefficient of MDM2 was calculated using five 10 μs standard simulations of isolated MDM2, and the translational diffusion coefficient for each analogue of the p53 peptide was calculated using conformations sampled every 100 ps from a single 10 μs standard simulation of the corresponding isolated p53 peptide. The β value was estimated using the following equation:37</p><p> (5)β=fSStargetfSStarget+fSSqsurf where fSStarget is the steady-state flux into the target state (encounter complex or bound state) and fSSqsurf is the steady-state flux across the q surface in the WE simulation. All rate constants were calculated from each of 10 independent WE simulations, and then averaged. Uncertainties in the averaged rate constants represent two standard errors of the mean (SEM).</p><!><p>The percentage of productive collisions (i.e., encounter complexes that succeed in rearranging to the bound state) was calculated according to the following equation: (6)%productivecollisions=fSSnativefSSencounter where fSSnative is the steady-state flux into the native, bound state and fSSencounter is the steady-state flux into the encounter complex; both fluxes were evaluated only after an approximate steady state was achieved (Figure S5). Reported percentages of productive collisions are averages over 10 independent WE simulations with uncertainties representing two SEM.</p><!><p>The goals of this study were to determine (i) the effects of preorganizing the p53 peptide ligand on its kon for binding the MDM2 protein receptor and (ii) the effect of increasing the concentration of the MDM2 receptor on the binding mechanism. As shown in Figure 1A, the extent of preorganization in the p53 peptide was tuned by applying a scaling factor α to the components of the energy function that involve solely the p53 peptide (see Methods) and setting the α values to 0.1 (fully disordered), 0.5, 1.0, and 2.0 (fully preorganized). To enable the calculation of statistically robust rate constants, we applied the WE path sampling strategy25,26 in conjunction with molecular simulations to enhance the sampling of binding events while maintaining rigorous kinetics. For each p53 peptide analogue (i.e., each α value), a set of 10 independent WE simulations were carried out, yielding >3000 binding events per simulation to achieve highly precise rate constants with relative errors of ≤16%, which amounts to a ≤ 0.1 kcal/mol difference in the corresponding free energy barrier at 25 °C as estimated by −RT ln(1/1.16). The simulations required one month to complete using 128 CPU cores of 2.3 GHz AMD Interlagos processors.</p><!><p>To directly compare the binding kinetics of the fully disordered p53 peptide relative to the other more preorganized analogues, it was essential to ensure that the fully disordered peptide was able to fold into an α-helical conformation upon binding MDM2. As shown by Figure 1B, all of the p53 peptide analogues are folded when bound to the MDM2 protein. By construction, our model of the fully disordered peptide (α = 0.1) results in an induced fit (folding-after-binding) mechanism38 in which the peptide folds only upon binding MDM2 in our simulations; likewise, the fully preorganized peptide (α = 2.0) results in a conformational selection (binding-after-folding) mechanism in which the peptide is fully folded before binding MDM2 in our simulations (Figure 1B).</p><p>For all of the p53 peptide analogues, ranging from fully disordered to fully preorganized, our simulations reveal that the mechanism of binding to the MDM2 receptor involves a two-step process in which diffusive collisions of the binding partners first form a metastable "encounter" complex followed by rearrangement of the encounter complex to the native, bound state (Figure 1; Figure S3): p53peptide+MDM2⇌k-1k1encountercomplex⇌k-2k2boundstate where k1 is the rate constant for formation of the encounter complex, k−1 is the rate constant for the dissociation of the encounter complex to the unbound state, k2 is the rate constant for rearrangement of the encounter complex to the bound state, and k−2 is the rate constant for rearrangement of the bound state to the encounter complex.</p><p>For our calculations of rate constants, we used the most stringent definitions of the encounter complex and bound state that encompassed the corresponding basins in the probability distributions of both the fully disordered and preorganized p53 peptides in Figure 2. The encounter complex was defined as those conformations satisfying the following criteria: (i) the binding partners are within van der Waals contact (<6 Å), (ii) the Cα RMSD for the p53 peptide after alignment of MDM2 is >2 Å, and (iii) at least one MDM2–p53 native contact is formed. The bound state was defined as having the binding partners within van der Waals contact and a Cα RMSD ≤ 2 Å of the p53 peptide after alignment of MDM2.</p><p>To assess whether there is a kinetic advantage to the peptide ligand being disordered or preorganized, we computed the kon values of the exact ordered and disordered analogues using the NAM framework in conjunction with WE simulations (see Methods). As shown in Table 1, the ratio of the kon for the fully disordered peptide relative to that of the fully preorganized peptide is 0.9 ± 0.2 (uncertainties represent two SEM), with a percent uncertainty that amounts to only a 0.1 kcal/mol difference in the corresponding free energy barrier as estimated by -RTln(konα=2.0/konα=0.1). Thus, given the high precision of these computed values, any kinetic advantage to being disordered (or preorganized) is very small.</p><p>We next examined the extent to which ligand preorganization influences the individual steps of the binding process. The computed bimolecular rate constant for formation of the encounter complex, k1, of the fully disordered p53 peptide is within error of that of its fully preorganized analogue with a ratio of 1.0 ± 0.1, indicating that being disordered (or preorganized) did not enable more rapid initial collisions. Given that native contacts are rewarded and non-native contacts are penalized in our simulation model (a Gō-type model), k−2 ≪ k2 such that the expression for the overall association rate constant is kon = (k1k2/(k−1 + k2). Since kon and k1 are within error of each other for all of the peptide analogues [e.g., for the fully disordered peptide, the kon and k1 are (5.7 ± 0.6) × 107 M−1 s−1 and (6.1 ± 0.5) × 107 M−1 s−1, respectively], the kinetics of the binding processes must be close to the limiting case where k−1 ≪ k2, such that kon = (k1k2/(k−1 + k2) ≅ k1.38 The formation of the encounter complex is therefore rate-limiting for all of the p53 peptide analogues (k2 was not computed since the hybrid WE/NAM approach permits calculation of bimolecular rate constants, but not unimolecular rate constants). Interestingly, the most preorganized peptide analogues (α = 1.0 and α = 2.0) undergo partial loss of structure upon forming the encounter complex (Figure 1B). This result suggests that the MDM2 receptor might aid the process of binding by disrupting preformed interactions within the p53 peptide that hinder rearrangement of the encounter complex to the bound state.</p><p>To gain further insight into the similarity in the kon values among all of the p53 peptide analogues, we calculated the percentage of productive collisions (i.e., those collisions that eventually reach the bound state) and the lifetime of the encounter complex. As shown in Table 1, the percentage of productive collisions for the fully disordered and fully preorganized p53 peptides are within error of each other (a ratio of 1.0 ± 0.1 for the percentage of productive collisions of the fully disordered peptide relative to that of the fully preorganized peptide) as are the lifetimes of the encounter complex (ratio of 1.0 ± 0.4). The high percentages of productive collisions (65 ± 3% and 66 ± 2% for the fully disordered and fully preorganized peptides, respectively) are consistent with our conclusion above that k−1 ≪ k2. Given that our simulations were carried out under steady-state conditions, generating pathways in only the binding direction, it was possible to obtain statistically robust estimates of non-equilibrium observables (e.g., rate constants and percentage of productive collisions), but not equilibrium observables (e.g., populations and lifetimes of the encounter complex), which would require sampling of unbinding as well as binding pathways. Nonetheless, since both the percentage productive collisions and lifetimes of the encounter complex are similar for the fully disordered and fully preorganized peptides, k−1 as well as k2 must be similar for the peptides. Thus, the folding of the fully disordered p53 peptide upon binding MDM2 does not appear to affect k2 relative to that of the fully preorganized peptide. It is worth noting that the k2 step may be slower in all-atom simulations due to attractive non-native interactions that are missing in our Gō-type simulations and that such nonnative interactions would likely result in additional benefits to the p53 peptide being preorganized relative to being disordered.</p><p>Our computed kon values are within error of the computed kon from atomistic simulations [(7 ± 4) × 107 M−1 s−1]11 and 6× faster than the experimental value (9.2 × 106 M−1 s−1).39 Thus, while the use of the Gō-type potential energy function22,23 would be expected to artificially accelerate the dynamics,40,41 the inclusion of appropriately parametrized HIs yields realistic rate constants.24 In particular, the computed relative translational diffusion coefficients for MDM2 and the p53 peptide for all of the peptide analogues are in excellent agreement with that predicted for the corresponding all-atom models by the hydrodynamics program HYDROPRO,42 3.7 × 10−6 cm2/s. As others have shown,24 the translational diffusion coefficients of proteins are underestimated in molecular simulations that neglect HIs—in our case, by 10× (Table 1; Table S1)—underscoring the importance of including HIs in simulations that lack explicit solvent.24 Interestingly, the extent of structure in the p53 peptide has no significant effect on the relative translational diffusion coefficient for the p53 peptide and MDM2 protein.</p><!><p>The inclusion of HIs in our simulations increased the kon by 30× (Table 1; Table S1). This result may appear at odds with previous simulation studies of protein–protein associations in which the inclusion of HIs was found to slow down the approach of the proteins.28,43 However, our results are in fact consistent with these studies since the effect of including HIs on the kon depends on the extent to which the intramolecular and intermolecular HIs have opposing effects on the diffusion of the binding partners. Whereas intramolecular HIs speed up the diffusion of binding partners that have no interactions with each other, yielding larger translational diffusion coefficients, intermolecular HIs slow down the diffusion of the binding partners when they are close to one another and have the tendency to move together. Our results involving the MDM2–p53 system reveal that the net effect of including both intramolecular and intermolecular HIs is a faster k1 as well as slower dissociation of the encounter complex (k−1), the latter being evident from longer lifetimes of the encounter complex and a greater percentage of productive collisions.</p><!><p>As demonstrated by previous experimental studies, the mechanism by which a small organic ligand binds a disordered protein receptor shifts from conformational selection to induced fit with increasing ligand concentration.18,19 Here, we examined the effects of increasing the concentration of a protein receptor (MDM2) on its mechanism of binding to a disordered peptide (p53 peptide), i.e., the relative fluxes through conformational selection and induced-fit mechanisms (Figure 3A).</p><p>Given that the computed kon ≅ k1 for all of the p53 peptide analogues in this study, the binding mechanism for the MDM2/p53 peptide system can be approximated as a two-step mechanism with a very fast second step (the k2 step; Figure S6) such that the fractional flux can be calculated using the following equation: (7)FCSFCS+FIF≅kf(ku+kf)+kon[R] where kon is set to an order-of-magnitude estimate (107 M−1 s−1) since the computed kon values are essentially the same for the fully disordered and fully preorganized p53 peptides; FCS and FIF are the fluxes through the conformational selection and induced-fit mechanisms, respectively; [R] is concentration of the folded receptor (MDM2); as shown in Figure 3A, kf is the rate constant for folding of the ligand (p53 peptide) from the fully disordered, unfolded (U) state, and ku is the rate constant for unfolding of the ligand from the fully preorganized, native folded (N) state. Thus, in this scenario, the fractional flux through conformational selection depends only on the concentration of the receptor and is therefore independent of ligand concentration. A detailed derivation of eq 7 can be found in the Supporting Information.</p><p>Since the equilibrium constant Keq (ratio of kf/ku) for the folding of the isolated p53 peptide is not known, we tested three different scenarios: (i) Keq = 1 for equally stable unfolded and folded states, (ii) Keq = 100 for an unfolded state that is much less stable than the folded state, and (iii) Keq = 0.01 for an unfolded state that is much more stable than the folded state (Figure 3B). When the folded state is much less stable than the unfolded state (Keq = 0.01), the mechanism of binding would be solely induced fit, regardless of MDM2 concentration. Substantial flux through conformational selection would be expected only when the folded state is equal or greater in stability to the unfolded state (Keq ≥ 1). For example, if Keq = 1, ~10% flux through conformational selection would be expected at the MDM2 concentration (1 μM) in binding kinetics experiments.39 In the regime where Keq ≥ 1, the mechanism of binding is predicted to shift from being dominated by conformational selection to induced fit with increasing MDM2 concentration (Figure 3B). These results are consistent with those from atomistic simulations in which a Markov state model44,45 was constructed to estimate rate constants for the MDM2–p53 peptide binding process and relative fluxes through conformational selection and induced fit were estimated (i) using a mechanism consisting of four instead of the three states used here and (ii) for various extent of helical content of the p53 peptide, which is analogous to varying Keq values for the unfolding/folding equilibrium of the peptide.17 In particular, the dominant binding mechanism becomes induced fit as the concentration of MDM2 increases and the extent of helical content decreases (or Keq decreases).</p><!><p>To our knowledge, the only other study that has directly compared the binding kinetics of an IDP relative to its exact preorganized analogue is a simulation study that focused on the binding of the disordered pKID domain to its partner protein, KIX.16 In this study, the disordered pKID domain was found to have a modest kinetic advantage (~2.5×) for binding relative to the preorganized analogue due to a more rapid k2 step, which corresponds to the rearrangement of the encounter complex to the native, bound state. In contrast, our study yielded similar computed kon values for the disordered and preorganized analogues of the p53 peptide in binding the MDM2 protein, revealing that the folding of the disordered p53 peptide upon binding MDM2 is very fast such that the k2 step is just as rapid as that of the preorganized analogue.</p><p>As noted above, the pKID domain is significantly larger than the p53 peptide: upon binding its partner protein, the pKID domain adopts two α-helices while the p53 peptide adopts only a single α-helix. Given its larger size, the folding of the fully disordered pKID domain is slower and may therefore have a more significant influence on k2. In particular, since the fully disordered pKID consists of two segments, the folding of the domain can take advantage of a dock-and-coalesce mechanism4 in which the docking of one segment facilitates the folding process in the k2 step.</p><p>The fact that our computed k1 values for the formation of the encounter complex are the same for the disordered and preorganized p53 peptides indicates that the MDM2–p53 binding process does not involve the "fly-casting" mechanism in which the disordered peptide would be predicted to collide more rapidly with its partner protein due to a greater capture radius.3 The lack of a fly-casting effect in our molecular simulations is underscored by our use of a Gō-type potential, which creates the optimal scenario for capturing the effect, i.e., the fully disordered p53 peptide folds only upon binding (forming ≥70% of intramolecular p53 native contacts only upon forming ≥98% of intermolecular MDM2–p53 native contacts; Figure 1B). Furthermore, we observed no differences in the capture radius of the fully disordered p53 peptide relative to its fully preorganized analogue as quantified by the radius of gyration Rg (most probable values of 7.7 and 7.3 Å, respectively) as well as a more sensitive metric, the maximum principal axis radius RM (6.6 and 6.9 Å, respectively; see Figure S5), despite the fact that the disordered conformations were generated with no rewarding of native contacts. The lack of differences in the capture radius and therefore the hydrodynamic radius is consistent with the fact that the computed translational diffusion coefficients of the fully disordered and fully preorganized p53 peptides are indistinguishable from each other (Table 1). Regardless, based on the Stokes–Einstein equation in which the translational diffusion coefficient is inversely proportional to the hydrodynamics radius, any kinetic advantage that could result from a larger capture radius (and therefore hydrodynamics radius) of the disordered peptide relative to its preorganized analogue might be canceled out by the effects of a slower translational diffusion coefficient.</p><!><p>We have determined the effects of preorganization of the intrinsically disordered, N-terminal p53 peptide on the kinetics of binding its partner protein, MDM2, using molecular simulations. In particular, our application of the WE strategy enabled the generation of >3000 of binding events, yielding statistically robust kon values for the fully disordered p53 peptide and exact analogues of the peptide that have been preorganized to various extents.</p><p>The resulting computed kon values are in reasonable agreement with experiment. Notably, the kon for the fully disordered p53 peptide is within error of that for its fully preorganized analogue, indicating no kinetic advantage to being disordered or preorganized for binding MDM2. Given that the rate constant k1 for formation of the encounter complex is essentially the same for the fully disordered and fully preorganized peptides, fly-casting is not a significant effect in our simulations of the MDM2–p53 peptide system, even though the ideal scenario for this effect was modeled, i.e., using a Gō-type potential that ensured folding of the fully disordered peptide only upon binding MDM2. Furthermore, since the percentages of productive collisions and lifetimes of the encounter complex are similar for the fully disordered and preorganized p53 peptides, the rate constant k2 for rearrangement of the encounter complex to the bound state must also be similar. Thus, folding of the fully disordered p53 peptide upon binding MDM2 during the k2 step must be very rapid. In contrast, the slower folding of larger IDPs may have a more significant effect on k2 relative to that for their fully preorganized analogues, as predicted for the pKID domain16 and by the dock-and-coalesce mechanism.4 Interestingly, the two most preorganized p53 peptide analogues undergo partial loss of structure upon forming the encounter complex, implying that the MDM2 receptor might "erase" preformed interactions within the p53 peptide that hamper the k2 step.</p><p>Finally, based on our kon values, we determined the effect of increasing the concentration of MDM2 on its mechanism of binding to the disordered p53 peptide ligand. When the unfolded state is much less stable than the folded state of the isolated p53 peptide, the mechanism for the binding of the MDM2 receptor to the disordered p53 peptide is predicted to switch from being dominated by conformational selection to induced-fit with increasing concentration of MDM2. On the other hand, when the unfolded state is either equal to or much greater in stability than the folded state, the mechanism of binding is solely induced fit, regardless of the MDM2 concentration. These results are consistent with those from recent atomistic simulations of the binding process involving the MDM2 receptor and p53 peptide ligand.17</p><p>Given the general features of our residue-level simulation models, results from our molecular simulations are relevant to any protein binding process involving a disordered peptide of a similar length to the p53 peptide that folds into a single α-helix upon binding its partner protein. Such disorder-to-helix transitions are common among molecular recognition events, including protein interactions of IDPs that play crucial cellular roles.2,20,46 Our results provide a valuable set of simulation data for testing future hypotheses that might be proposed for the binding mechanisms of IDPs and their preorganized analogues.</p><!><p>FIGURE S1 Probability distributions of the fraction of native contacts of the p53 peptide (Qp53) in the absence of MDM2, ranging from fully disordered (α = 0.1) to fully preorganized (α = 2.0). Distributions for each value of the scaling factor α were generated using conformations sampled every 100 ps from a single 10-μs standard simulation starting from the MDM2-bound conformation.</p><p>FIGURE S2 Average fraction of native contacts in p53 (Qp53) as a function of the εnative for MDM2-p53 native contacts. For each εnative value, the average Qp53 was calculated using conformations sampled every 100 ps from a single 10-μs standard simulation of the MDM2-p53 peptide complex with the fully disordered peptide (α = 0.1) starting from the MDM2-bound conformation.</p><p>FIGURE S3 Full view of the free energy landscape of the MDM2-p53 binding process as a function of the Cα RMSD of the p53 peptide after alignment of MDM2 from the crystal structure of the MDM2-p53 peptide complex 1 and the minimum MDM2-p53 distance for the fully disordered p53 peptide (α = 0.1). Data shown is based on conformations sampled every 1 ps from 10 independent WE simulations under steady-state conditions. Contour lines represent intervals of 0.5 kcal/mol.</p><p>FIGURE S4 Computed kon as a function of WE iteration for p53 peptide analogues with various extents of structure, ranging from fully disordered (α = 0.1) to fully preorganized (α = 2.0). The molecular time is defined as Nτ where N is the number of WE iterations and τ is the fixed time interval of each iteration.</p><p>FIGURE S5 Probability distributions of the "capture" radius for the p53 peptide with various extents of structure, ranging from fully disordered (α = 0.1) to fully preorganized (α = 2.0), as monitored by the radius of gyration Rg, and maximum principal axis radius RM. For each α value, the probability distribution was calculated from a 10-μs BF simulation of the isolated peptide. Based on the Rg metric, it may appear that the fully disordered p53 peptide achieves a significantly larger maximum value than that of the fully preorganized peptide (11.6 Å vs. 8.0 Å). However, the more sensitive RM metric reveals that the fully disordered p53 peptide not only assumes more expanded conformations (maximum value of 10.2 Å), but also more contracted conformations (minimum value of 3.3 Å).</p><p>Figure S6: Approximation of the binding mechanism of the disordered p53 peptide ligand to the folded MDM2 protein receptor when kon ≅ k1. States are defined as in Fig. 3A.</p><p>Table S1. Computed kon, k1 for formation of the encounter complex, lifetime of the encounter complex, % productive collisions, and relative translational diffusion coefficients for the MDM2-p53 binding process and various analogues of the p53 peptide, ranging from fully disordered (α = 0.1) to fully preorganized (α = 2.0) in the absence of hydrodynamic interactions (HIs). Data shown are averages from 10 independent WE simulations; uncertainties represent 95% confidence intervals.</p>

PubMed Author Manuscript

Shape-Assisted Self-Assembly

Self-assembly and molecular recognition are critical processes both in life and material sciences. They usually depend on strong, directional non-covalent interactions to gain specificity and to make long-range organization possible. Most supramolecular constructs are also at least partially governed by topography, whose role is hard to disentangle. This makes it nearly impossible to discern the potential of shape and motion in the creation of complexity. Here, we demonstrate that long-range order in supramolecular constructs can be driven by the topography of the individual units even in the absence of directional interactions. Here, molecular units of remarkable simplicity self-assemble in solution to give homogeneous singlemolecule thin two-dimensional supramolecular polymers of defined boundaries. This dramatic example spotlights the critical function that topography can have in molecular assembly and paves the path to rationally designed systems of increasing sophistication.

shape-assisted_self-assembly

Main Text:

<p>Shape plays a critical role in natural molecular recognition processes such as enzyme catalysis. As early as 1894, Emil Fischer proposed with his 'Lock and Key' model that a substrate must possess a profile complementary to that of the enzymatic cleft; otherwise, there is poor association between the two (1). Within this understanding, one of the simplest topographical elements that can allow for (molecular) recognition is curvature. The notion of curvature has an underlying and accepted role in creating order within supramolecular systems but is overshadowed by the driving forces for (self-)assembly (2,3). Despite its relevance, the role of topography, in the absence of any other contributing factors, as an element of recognition in supramolecular assembly processes has long eluded experimental proof, and little is known about the unique advantages exploiting it would bring forward.</p><p>Over three decades ago, the first reports on supramolecular polymers (4,5) catapulted the interest in soft dynamic materials and, particularly, hydrogen-bonded linear assemblies (6)(7)(8)(9). Together with other highly directional non-covalent interactions (10)(11)(12)(13) they remain the classical choice for designing self-assembling systems due to their reversible and self-healing properties (14,15). Less directional non-covalent interactions between π-surfaces or hydrophobic contacts seldom result in assemblies (16)(17)(18). In the rare cases when they do, assistance is required from another effect in tandem (19)(20)(21) because of the limited long-range control. To a certain extent, curvature has been recognized as a feature for assembly-proficient monomers. For instance, Aida and co-workers exploited a bowl-shaped corannulene for permitting the creation of dormant supramolecular species and thus the development of living-supramolecular polymerization (22). Moreover, they recently have used saddle-shaped molecules to report the first example of alternating heterochiral supramolecular copolymerization (23). Additionally, Itami demonstrated that curved nanographenes can associate to form nanofibers purely based on dipolar π-π stacking (24). These rare examples offer a small glimpse of the scope that this approach towards developing functional materials has to offer.</p><p>Herein, we report that in the absence of directional non-covalent interactions, the shape of a molecular unit can become the governing ordering force to assemble micrometer-long stacks. Those columns then self-assemble into two-dimensional (2D) sheets of single-molecular thickness and highly defined boundaries. These nanostructures are formed by assembly in solution of a negatively curved molecule and held together primarily by dipolar π-π interactions. The simplicity of this approach towards 2D materials is unique and contrasts strongly with existing strategies for which a representative overview is presented in Fig. S1. Notably, this approach is complementary to on-surface self-assembly (25) and surface nucleated thin films (26) because the nanostructures exist and are persistent in solution. The uniqueness of this process, which we term 'shape-assisted self-assembly', demonstrates the significance of shape in the self-assembly process, provides a crucial conceptual tool and brings us closer to control and rational design of complex bottom-up assemblies in nanotechnology in general.</p><p>Flat molecules are frequently observed in supramolecular polymerizations due to ease of access to these structures synthetically (8,27,28). Bowl shapes (29,30), and more so saddles (24,31,32), are far less commonly used topographies for the opposite reason. Yet, to introduce order in an assembly, negatively curved systems stand out as preferred candidates to induce eclipsed stacking because transverse rigidity is increased on both molecular and macroscopic scales (23,24,33). Asymmetry in saddle design, in addition to replacing a uniform surface with a framework scaffold, heightens the entropic barriers further due to unequal axes of principal curvature (Fig. 1). An energetic minimum can then be reached with eclipsing negatively curved frames such that small deviations from overlap demand a high energetic penalty and, consequently, enforce order. Depending on the asymmetry of the saddle, their assemblies likely present intrinsically differentiated lateral interactions permitting the formation of tertiary structures. Müllen and co-workers (34) first reported a small saddle-shaped molecule that possesses these characteristics within the aryl frame. Remanufacturing of this core with different functional groups could serve as a monomer suitable for such polymerization (Fig. 1).</p><p>As a first approach, peripheral alkyl substitution of four of the six available sites was proposed to (i) enhance solubility, and (ii) favor eclipsing π-surfaces. The addition of hydrogen bonding motifs was deliberately avoided because of their superior capability to order systems (23). As reference systems, we included the unfunctionalized porphyrinoid to investigate the effect of a core without sidechains and the macrocycle bearing tert-butyl groups due to its reduced propensity to self-assemble (34).</p><p>Fig. 1 Design principle, structural scheme and assembly process. (left) Shape as a design principle to restrict rotational and translational freedom. (right) Generic chemical structure of carpyridines with R group modifications to the core. (bottom) Multi-stage assembly process, where (i) units assemble into linear stacks guided by the shape of their core and driven by entropy before (ii) assembling into defined nanosheets.</p><p>Based on Miyaura borylation of dibromocarbazoles prior to macrocyclization, we developed a unifying synthetic protocol. Four-fold Suzuki-Miyaura cross-coupling with a commercial pyridine gave cyclic products (up to 13% yield; 60% yield per coupling) that we name 'carpyridine' as a consequence of their interlinked constituents. Although all carpyridines could be synthesized using this route, preparation of the unsubstituted derivative 2H-Car-H resulted in inferior yields (< 1%). By subjecting 2H-Car-tBu to a reverse Friedel-Crafts alkylation reaction, we could efficiently remove the tert-butyl groups from the rim of the macrocycle (75% yield). Full characterization of the synthesized compounds and their intermediates is described in the SI.</p><p>All compounds showed excellent solubility and stability in common organic solvents such as chlorinated solvents or THF. In apolar solvents like toluene or methylcyclohexane (MCH) the solubility was visibly reduced. This was corroborated by variable temperature 1 H NMR which displayed an entirely different behavior as a function of the solvent used. 2H-Car-C6 (7.5 mM) in deuterated 1,1,2,2-tetrachloroethane (TCE-d2) in a temperature range from 343 K to 233 K showed a marked broadening of the aromatic signals together with an almost negligible downfield shift. However, in toluene-d8 (9.2 mM), full coalescence of the pyridine NMR signals was observed at 253 K and both increasing or decreasing temperatures resulted in new environments and/or broadening (Fig. 2 and Fig. S19 to S21). While broadening at lower temperature is expected for classical supramolecular polymers, broadening or splitting at higher temperatures is an indication of an entropy-driven process (35,36). Further spectroscopic evidence could be obtained from VT UV-vis and fluorescence spectroscopies (Fig. S27 and S28A). The former shows a decrease in intensity with only minor spectral changes, in agreement with other entropy-driven examples (35,36), and the latter a decrease in fluorescence quantum yield at higher temperatures due to aggregation-induced fluorescence quenching. Key evidence of the carpyridine ordering and shape-assisted self-assembly was obtained by microscopy. By (scanning) transmission electron microscopy (TEM; STEM) we were to observe extremely well-defined, micrometer-long 2D sheets when imaging samples prepared from 2H-Car-C6 (1 mM in toluene) (Fig. 3D). These structures were only observed for 2H-Car-C6 bearing hexyl side chains and when assembled in toluene. Longer side chains (2H-Car-C12), or the absence of them (2H-Car-H), resulted in significantly less defined aggregates (Fig. S30). The results observed with 2H-Car-C12 imply that too long sidechains prevent assemblies of significant order and that interactions at the periphery of the cycles are not the dominant driving force. On the other hand, the observations for 2H-Car-H mean that the carpyridine core is insufficient to solely drive the assembly. Higher resolution images and topologies of the 2D sheets were obtained by atomic force microscopy (AFM) as seen in Fig. 3A. Individual sheets can be seen isolated or stacked on top of one another, with highly uniform edges allowing identification of each distinct assembly (Fig. 3, B and C). In addition to single-layer sheets (Fig. S36), holes in one multi-layer sheet allowed the measurement of the thickness of a single layer, determined to be 2 nm; a distance that is equivalent to the width of a single carpyridine molecule (Fig. 3, E and F). From the AFM images and the clearly defined edges, we propose that the long dimension of the assembled 2D sheets is formed by ordered stacks of carpyridines, and the side edges are mainly aliphatic contacts. Such a construction fully explains the critical interplay between both driving forces resulting in highly ordered aggregates. To confirm this hypothesis, we used the TEM sample for selected area electron diffraction (SAED) on ensembles of 2D sheets due to the rapid degradation of the organic material under the electron beam. The diffraction patterns obtained (Fig. 3G, and Fig. S34) clearly show long-range order within the structures as observed by defined diffraction spots on top of diffuse halos. We measured a principal characteristic distance of about 4 Å, which can be readily assigned to the π-π distance between the carpyridines cores as observed by crystallography and predicted by density functional theory (DFT) calculations (Fig. 4). The observed halo pattern further emphasizes the soft-material nature of the self-assembled 2D sheets. Additional evidence for the proposed molecular arrangement within the 2D-sheets was obtained by single-crystal X-ray diffraction and DFT calculations. As expected, obtaining defined crystals of sufficient size in all dimensions was challenging but ultimately, single crystals of 2H-Car-C6 were obtained from a toluene/methanol mixture albeit of poor quality. Despite the heavily disordered hexyl sidechains, the obtained structures clearly show that three independent carpyridine molecules in the asymmetric unit have aligned principal axes of curvature in a 'slipped' stack. Although this arrangement is thought to be uncharacteristic of the reported 2Dsheets, it indicates that by increasing the barriers to translation and rotation of stack formation, the curvature of the saddle-shaped molecules drives the order of the assembly as postulated. Our theoretical calculations by density functional theory (DFT) using various functionals (B3LYP, wB97XD, PBE0 and the 6-31(d,p) basis set) predict that the preferred assembled configuration of the carpyridine core is an alternated stacked conformation (Fig. S42 to S44). The displaced saddle conformation as observed in the crystal structure is another local minimum but at a slightly higher energy (∆EDFT ~ 3 kJ•mol −1 ). Although saddles are translationally displaced by 1.6 Å from centroid to centroid, all carpyridines retain the same orientation along the principal axis of the stacks. In both arrangements, the predicted inter-carpyridine distance between centroids is 3.8 Å and within the expected range of π-π stacking effects (37,38). Our DFT studies were complemented by semi-empirical calculations using the PM6 method, showing that the addition of the hexyl chains (stacks of six 2H-Car-C6 units were studied) leads to the preferred stable linear alternated stacked structure proposed above (Fig. S45).</p><p>Further observations of 2H-Car-C6 in the solid-state were made with thermal analysis using differential scanning calorimetry (DSC). This showed several phase transitions and notably at least two distinct crystallization events on the cooling traces. By polarized optical microscopy, we observe at least two different (semi)crystalline phases (Fig. S39). Notably, when the cooling rate of the melt is carefully controlled (less than 5 K•min -1 ), the last crystallization transition yields large, highly uniform, sheet and needle-like crystalline regions of >100 µm in length, in strong support of a highly directional supramolecular interaction.</p><p>The simplicity of the carpyridine supramolecular units, which results in the self-assembly of large 2D sheets of single molecular thickness is unique. It is only possible because its topological elements enforce shape-assisted directional assemblyremarkably achieved in absence of directional interactionsbut also enough flexibility to prevent the system from collapsing in solution like most rigid 2D materials. The 2D sheets themselves can only be assembled as a consequence of the delicate equilibrium between entropic gains and competition of both dipolar π-π and hydrophobic interactions. This first and exceptional example is a direct demonstration of the power of shape in the assembly of molecular nanostructures, yet the chemical space of shapeassisting systems is largely unexplored. As the obtained assembly is a direct consequence of the underlying topography, shape-assisted self-assembly holds substantial promise as a design concept for new soft-matter materials and its implementations in nanotechnological applications such as sensing and nanofabrication.</p>

ChemRxiv

Refinement of variant selection for the LDL-C genetic risk score in the diagnosis of the polygenic form of clinical Familial Hypercholesterolemia and replication in samples from six countries

Background Familial Hypercholesterolemia (FH) is an autosomal-dominant disorder caused by mutations in one of three genes. In the 60% of patients who are mutation-negative we have recently shown that the clinical phenotype can be associated with an accumulation of common small-effect LDL-C-raising alleles using a 12-SNP score. The aims of the study were to improve the selection of SNPs, and to replicate the results in additional samples. Methods Receiver-operating characteristic curves were used to determine the optimum number of LDL-C SNPs. For replication analysis, we genotyped patients with a clinical diagnosis of FH from six countries for six LDL-C-associated alleles. We compared the weighted SNP score among patients with no confirmed mutation (FH/M-), those with a mutation (FH/M+), and controls from an UK population sample (WHII). Results Increasing the number of SNPs to 33 did not improve the ability of the score to discriminate between FH/M- and controls, while sequential removal of SNPs with smaller effects/lower frequency showed a weighted score of six SNPs performed as well as the 12-SNP score. Meta-analysis of the weighted 6-SNP score, based on polymorphisms in CELSR2, APOB, ABCG5/8, LDLR and APOE loci, in the independent FH/M- cohorts showed a consistently higher score in comparison to the WHII population (P<2.2\xc3\x9710-16). Modeling in individuals with a 6-SNP score in the top three quarters of the score distribution, indicated a >95% likelihood of a polygenic explanation of their increased LDL-C. Conclusion A 6-SNP LDL-C score consistently distinguishes FH/M- patients from healthy subjects. The hypercholesterolemia in 88% of mutation-negative patients is likely to have a polygenic basis.

refinement_of_variant_selection_for_the_ldl-c_genetic_risk_score_in_the_diagnosis_of_the_polygenic_f

Introduction<!>LDL-C genetic risk score SNPs selection and genotyping<!>Patient cohorts<!>SNP score calculations in replication cohorts<!>Estimating the probability of a polygenic cause<!>Patients\xe2\x80\x99 baseline characteristics<!>Variant selection<!>LDL-C SNP score<!>Estimation of the proportion of FH/M- subjects likely to be polygenic by SNP score<!>Discussion

<p>Familial Hypercholesterolemia (FH), in its classical form, appears to be autosomal co-dominant disorder, characterized by increased plasma concentrations of LDL-cholesterol (LDL-C) and premature symptoms of Coronary Heart Disease (CHD) (1). The prevalence of heterozygous FH is between 0.2% - 0.5% (1, 2) with a higher prevalence in some populations due to founder effects (2). Worldwide between 14 to 34 million people are thought to be affected with heterozygous FH, of whom at least 95% are undiagnosed (2). Clinical diagnostic systems for FH have been developed in the UK (3), USA (4) and the Netherlands (5), based on the degree of increase in LDL-cholesterol concentrations (typically >189 mg/dL or >4.9mmol/L in adults), and a family history of early CHD and/or increased cholesterol concentrations, such patients are given a diagnosis of Possible FH (PFH). The additional presence of clinical features such as tendon xanthomas, results in a diagnosis of Definite FH (DFH). When patients carry variants deemed to be pathogenic, they also receive "DFH" as diagnosis. The usefulness of a molecular test to provide an unequivocal diagnosis is becoming increasingly appreciated, in particular to enhance unambiguous identification of affected relatives (6). Early identification of at-risk individuals allows changes in lifestyle including dietary intervention, and drug treatment, usually one of the statin class of lipid-lowering agents, which have been shown to significantly reduce coronary atherosclerosis (7) and to improve life expectancy (8, 9).</p><p>Since 2008, several guidelines for the identification and management of patients with FH have been published (10–13). Although differing in detail and emphasis, there are several common threads (reviewed in (14)), including the utility of genetic testing to confirm the diagnosis and apply it in "cascade" testing of the relatives, which is a cost effective approach to find new cases (15–17). Cascade testing has been used extensively in several countries in Europe, most notably in the Netherlands (6), where it has resulted in the identification of 67% of FH patients with an assumption of 1:450 prevalence (18), which is probably underestimated (19, 20). The UK guidelines (10) state that cascade testing of first degree relatives of every FH proband should be carried out where a mutation has been identified in the proband, or if no mutation can be identified, based on LDL-C measures. However in the Netherlands, cascade testing is carried out only in families where a mutation has been identified (6), and this approach is also being adopted in Wales (21).</p><p>FH is caused by mutations in the LDLR, APOB or PCSK9 genes (1, 2). The most common class of genetic defect is a mutation in LDLR, and currently over 1200 mutations have been reported world-wide (http://www.ucl.ac.uk/fh, (22)). Even when exhaustively screened, in a small proportion of DFH subjects (10-15%) with tendon xanthomas, and a larger proportion of PFH patients (60-75%), no mutation can be found (e.g. (23)). This may be due to several reasons, for instance failure to detect all DNA changes present using current methods, or because the mutations are in genetic regions that are not currently covered (e.g. introns), or the mutations are in genes that are yet-to-be identified as FH-causing. However, the most likely reason would be because of the inclusion of non-FH patients (i.e. a clinical false positive diagnosis).</p><p>In 2010, meta-analysis of GWAS data identified 95 loci involved in determining lipid concentrations (24), and we have used a 12-SNP LDL-C genetic risk score (the weighted sum of the LDL-C-raising alleles, where weights are the effect sizes from GWAS), as an unbiased genetic instrument for Mendelian randomization studies (25). Compared to over 3000 subjects from the UK population-based Whitehall II (WHII) study, the weighted LDL-C-raising SNP score frequency distribution among UK FH patients with no identified mutation (FH/M-) was significantly higher (P=4.5×10-16), an effect which was confirmed in a cohort of similar patients from Belgium. This strongly suggests that a substantial proportion of FH/M- patients (up to 20%) are likely to have a polygenic cause of the increase in LDL-C rather than an, as yet, unknown single gene mutation. Cascade testing is likely to be less effective in such cases, since fewer than the predicted 50% of first degree relatives will have inherited enough of the "polygenes" to have concentrations of LDL-C above the diagnostic threshold (26).</p><p>In the current paper we have examined the possibility of using additional SNPs to improve discrimination and fewer SNPs to reduce genotyping costs, we have examined the utility of the LDL-C SNPs score in an additional seven cohorts (from six countries) of patients with a clinical diagnosis of FH, and have estimated the likelihood of having a polygenic (as opposed to a monogenic) cause of hypercholesterolemia.</p><!><p>The LDL-C genetic score variant selection analysis was performed using the WHII cohort (25) and the FH/M- patients (n=175) (Oxford familial hypercholesterolemia study (27) and (28)). 21 additional meta-GWAS LDL-C-associated SNPs (Supplementary Table S1) (24) were added to the original 12-SNP score. Genotypes were obtained using the Metabochip (Illumina Inc) genotyping array. All SNPs were included in regression models for LDL-C, and the best fitting sets of SNPs determined by the lowest Akaike information criterion (AIC) and Bayesian information criterion (BIC) were used to construct two additional scores. The discriminatory ability of these scores was assessed using the area under the receiver operating curve (ROC). The original 12 SNPs in the score were ranked by their frequency and effect size and top six SNPs were selected for score calculations. ROC curves were used to evaluate the sensitivity/specificity of using a 12-SNP vs. 6-SNP score in discriminating between a general population and FH patients with no mutation (performed on cohorts from the original study (25)).</p><p>Genotyping was performed using KASPar PCR TaqMan assays (Life Technologies) and genotype calling carried out using an automated system, the results of which were checked manually by study personnel using SNPviewer software. One SNP (rs4299376) could not be genotyped by TaqMan, and a proxy was used instead (rs6544731).</p><!><p>Seven independent cohorts of patients diagnosed with FH were collected. Informed written consent was obtained from all subjects and the study was approved by ethics committees in each county. The biggest cohort comprised Dutch adults, which included 66 mutation positive (FH/M+) and 572 mutation negative (FH/M-) patients. Other cohorts included Greek children (68 FH/M+, 60 FH/M-), 22 Dutch children (all mutation FH/M-), adults from Canada (39 mutation FH/M+, 37 FH/M-), Italy (144 FH/M+, 58 FH/M-), Poland (14 FH/M+, 15 FH/M-), and Israel (20 FH/M+, 43 FH/M-). All individuals were of Caucasian background. All subjects had an autosomal dominant mode of inheritance of hypercholesterolemia in the family; the presence of primary hypercholesterolemia TC ≥290 mg/dL (≥7.5 mmol/L) (or TC ≥259 mg/dL or ≥6.7 mmol/L for children under 16 years of age) in the proband or proband's first degree relative; plasma or serum LDL-C ≥189 mg/dL or ≥4.9 mmol/L; and family history of coronary artery disease at <55 years for men and <60 years for women in a first degree relative. In addition, some subjects had a personal or a family history of tendon and cutaneous xanthomas. Patients from Israel were clinically diagnosed using the Make Early Diagnosis Prevent Early Death (MED-PED) criteria (4). Informed consent was obtained from all the subjects investigated. The FH mutation detection methods varied slightly, however, they all included screening of the entire coding region of LDLR. The samples were also tested for the APOB p.R3527Q (apart from the Greek cohort, since the mutation has never been found in Greece), and the PCSK9 p.D374Y mutation.</p><!><p>The LDL-C SNP score was calculated using weighted sums for the six selected SNPs. A group of 3,020 healthy individuals (participants of the UK Whitehall II study (29)) was used for comparison (baseline characteristics of WHII are shown in Supplementary Table S2).</p><!><p>Given an individual who is diagnosed with FH but for whom no causal mutation has been found in the known FH genes, we assume that their LDL-C is greater than 189 mg/dL or 4.9 mmol/L either because of an unknown single gene mutation or a polygenic cause. For such individuals we can use the equation below to calculate the probability of a polygenic cause (explained further in Supplementary Methods):P(x=−ve|LDL>189,g,m=−ve)=P(LDL>189|x=−ve,g).P(x=−ve|m=−ve)ΣxP(LDL>189|x,g).P(x|m=−ve)</p><p>Given an individual who is diagnosed with FH but for whom no causal mutation has been found in the known FH genes, we assume that their LDL-C is greater than 189 mg/dL (4.9 mmol/L) either because of an unknown single gene mutation or a polygenic cause. The relative probability of these two causes depends on the frequency of unknown single gene mutations and the probability distribution of the polygenic effects. Given these, it is straightforward to work out the probability of a polygenic cause given an individual's mutational status at the known FH genes, LDL-C measurement and polygenic score. However, we do not know either the true polygenic score, since not all the LDL-C genes have been found, or the frequency of unknown single gene mutations (by definition). Here we approximate the polygenic term using the effects of the 6-SNP score in WHII individuals, and calculate the probability for several different unknown mutation frequencies (0, 0.001, 0.005, 0.01). Note that, if we assume the frequency of confirmed FH is 1/500 = 0.002, when we have found all of the LDLR/APOB/PCSK9 mutations, the prevalence of undetected monogenic mutations must be lower than 0.002. Also note that use of the 6-SNP genetic risk score underestimates the role of the polygenic component and so will underestimate the probability of a polygenic cause.</p><!><p>The baseline characteristics of the FH patients included in this study are shown in Supplementary Table S3. Overall, in all cohorts where data was available, FH/M+ patients had higher pre-treatment TC and LDL-C than FH/M- patients from the same cohort.</p><!><p>We first attempted to improve the performance of the SNP score by including 21 additional SNPs (Supplementary Table S1), previously identified by the GLGC GWAS meta-analysis as influencing LDL-C (24). To maintain a high specificity for LDL-C, we had originally included SNPs where the only or major effect of the SNP was on LDL-C and not on another lipid trait, but for this analysis the additional genes (e.g. CETP), which were included, affected lipid traits other than LDL-C.</p><p>Addition of these 21 LDL-C-raising SNPs did not significantly improve the ability of the SNP score to discriminate between FH/M- and healthy subjects (AROC=0.673 (95%CI: 0.632-0.715), P=0.98) (Supplementary Figure S1). BIC analysis selected 13 SNPs for the score (six SNPs from the original 12-SNP score and seven GWAS SNPs out of the additional 21) (Supplementary Table S4). AIC analysis selected a 25-SNP set (composed of 8 SNPs from the original 12 SNPs and 17 SNPs from the additional 21) (Supplementary Table S5). Neither BIC nor AIC SNP selections improved the performance of the 12-SNP score (Supplementary Figure S1). Following this, the sequential removal of SNPs of smaller effects and/or lower minor allele frequencies showed that a weighted score of six SNPs performed as well as the 12-SNP score (P=0.16, Figure 1). Thus, to improve the cost-efficiency of the study, the SNP score calculations in the replication cohorts were based on genotypes of six SNPs (nearby gene): rs629301 (CELSR2), rs1367117 (APOB), rs6544713 (proxy of rs4299376, ABCG5/8), rs6511720 (LDLR), rs429358 (APOE), and rs7412 (APOE), summarized in Supplementary Table S6. Genotypes for the 6-SNP score were available in a total of 351 FH/M+ and 807 FH/M- patients.</p><!><p>Overall, the FH/M- patients group had the highest mean LDL-C SNP score (0.708), followed by the FH patients with an identified mutation (FH/M+) (0.656). The control WHII cohort had the lowest weighted score (0.632), which was significantly lower than the FH/M- (P < 2.2×10-16), and the FH/M+ (P = 0.04) cohorts (Supplementary Figure S2). Among the FH/M- patient cohorts, the highest LDL-C SNP score was observed in Dutch children (0.782) followed by Greek children (0.731) (Supplementary Table S7). 707 (88%) of FH/M- patients had a score above the first quartile, of whom 288 (36% of the whole FH/M- cohort) had a score that fell within the top quartile of the WHII LDL-C SNP score distribution.</p><p>The FH/M+ patients were divided into LDLR mutation carriers (n=323), APOB p.R3527Q (n=13) carriers and PCSK9 p.D374Y carriers (n=2). Patients who had the APOB p.R3527Q mutation had significantly lower LDL-C SNP score than patients with other mutations (0.521 vs. 0.661, P = 0.05) (Figure 2).</p><p>LDL-C SNP score results for each of the seven cohorts genotyped in this study were combined with two large cohorts (from the UK Simon Broome register and from Belgium) analysed in the original study (25), for a meta-analysis, shown in Figure 3. Again the effect in all cohorts was highly consistent and the overall standardised mean difference (SMD) for all FH/M- groups compared to the WHII sample was 0.381 (95% CI: 0.328-0.433).</p><!><p>For clinical utility, it would be valuable to estimate the probability that the increased LDL-C seen in an FH/M- individual can be explained by their weighted 6-SNP score. The first estimate needed for this calculation is the underlying rate of undetected monogenic mutations in FH/M- subjects. Based on the lack of novel genes causing FH reported to date, and in our whole exome Next Generation Sequencing data of 70 FH no-mutation patients, which also failed to identify a novel common FH-causing gene (30), this is a reasonable estimate. By contrast, if we have identified only 75% of all mutations to be found, the frequency of the remaining undetected mutations would be 0.0005, and this seems likely to be the upper limit of undetected mutations. At an undetected mutation frequency of 0.0005, our analysis, shown in Figure 4, suggests that the probability of a polygenic cause for LDL>189 mg/dL (4.9mmol/L) in all the assessed FH/M- individuals is >95%, and it goes down when the frequency of undetected monogenic cause increases (Supplementary Figure S3).</p><!><p>The LDL-C SNP score analysis in seven independent cohorts consistently confirmed the findings reported by Talmud et al. (25), that patients with clinical diagnosis of FH but with no identified mutation (FH/M-) have a significantly higher mean LDL-C-raising SNPs score than individuals from the general population (combined sample P < 2.2×10-16), which suggests that their high plasma LDL-C concentrations are considerably influenced by 'polygenes'. In addition, as previously reported, FH patients who carry an FH-causing mutation (FH/M+) also had a higher mean LDL-C SNP score than the WHII cohort (P = 0.04), which confirms results from (25) and suggests that in at least some cases the FH phenotype is being caused by the combination of a single mutation of large effect and several LDL-C-raising alleles of modest effect. This result could help to explain the variability in penetrance of certain FH mutations in the relatives of FH probands. When analyzing the mutation-positive patients by the mutated gene, patients with the defective APOB (due to the p.R3527Q mutation) had the lowest SNP score (0.521) among all studied groups. This suggests that the APOB mutation is highly penetrant, which is contradictory to what has been shown previously (31), and may reflect sample bias in this selected group of FH patients. Another explanation is that not all LDLR variants identified in the FH/M+ group are truly pathogenic, which leads to misclassification. The highest LDL-C SNP score was observed in the two mutation negative hypercholesterolemic children cohorts (one from the Netherlands and one from Greece), showing for the first time that the SNP score discriminates well in children as well as adults. In general, the mutation detection rate in children with a clinical diagnosis of FH is higher than in adults (32), and this is because, when comparing the LDL-C distribution in FH patients and their unaffected siblings, the false positive and false negative rate is smaller in children than in adulthood (26), where secondary "environmental" causes for high LDL-C concentrations may be having an influence. Our data suggest that in a child, once a single gene cause for having highly increased LDL-C is ruled out, a polygenic cause is highly likely. However, to confirm this result the children cohorts should be compared against a country-matched homogeneous lifestyle background children control, which is currently unavailable and remains a limitation to this study.</p><p>One of the limitations in the 6-SNP score we have used here is that it does not contain all of the information on the genetic determinants of LDL-C concentrations available following the recent GWAS studies. If patient samples are being tested using Next Generation Sequencing approaches it is technically and financially feasible to include all 12 SNPs and indeed to include all SNPs that have been associated with LDL-C even if they also influence other lipid traits. From a diagnostic point of view we have shown that the 6-SNP score is as good at discriminating between FH/M- and the WHII control subjects as the 12-SNP score and a smaller number of SNPs would clearly have cost benefits. While we show here that the 6-SNP score discriminates well in FH patients from an additional 6 countries, all samples are from Caucasian patients and we currently have no data to allow us to extrapolate the utility of this score to patients from other ethnic backgrounds, where the minor allele frequency will differ considerably, and where the raising effect of the SNPs on LDL-C may not be consistent. Another limitation may be that the probability of having LDL-C greater than 189 mg/dL (4.9 mmol/L) given the LDL-C genetic risk score was estimated in the FH patients based on a model using observed LDL-C concentrations in the WHII cohort. However, given the high mean LDL-C concentration in WHII, the estimated probability of having LDL-C >189 mg/dL (>4.9 mmol/L) is likely to be higher than that estimated in a younger, healthier sample. This will translate into a lower probability of a polygenic cause, especially for those in the lowest quartile of the genetic risk score.</p><p>The question remains as to whether or not the mutation negative patients do indeed carry an unidentified FH causing mutation and if so what proportion this represents. While we accept that this is a possibility we believe it will be a very rare event. The prior probability that a patient with a clinical diagnosis of FH has a mutation in one of the three known FH causing genes is ~80% (i.e. in those with the clinical diagnosis of definite FH this is the mutation detection rate previously reported (3)). To date there have been no reports of any identified fourth gene where mutations cause autosomal dominant FH. Once the presence of a mutation in known genes is ruled out by comprehensively molecular genetic diagnostic methods, the second most likely probable cause, as we show here, is a polygenic inheritance. Our analysis here indicates that, assuming an undetected mutation frequency of 0.0005, the probability of a polygenic cause for LDL-C>189 mg/dL (4.9mmol/L) in the assessed FH/M- individuals is >95%. There is also a possibility that FH/M- patients who have the LDL-C genetic risk score in the lower quartiles of the score distribution have an intermediate phenotype between FH and Familial Combined Hyperlipidemia (FCHL), hence the slightly higher TG in FH/M- patients. This could be due to inheritance of higher number of LDL-C- and TG-raising alleles. Therefore as before (25), we believe that the clinical diagnosis of FH should be used only for patients with a DNA identified genetic cause.</p><p>All recent guidelines for the diagnosis and cascade testing of FH, except in the US, have recommended the utility of DNA testing when the family mutation is known and the use of LDL-C measures where the mutation is not available (14). Based on the data we present here, only in those probands with a confirmed monogenic cause will cascade testing be cost effective, because in the remainder there is most likely to be a polygenic cause. In countries where DNA testing is not available for reasons of availability or willingness to fund such genetic tests, cascade testing, based on the use of LDL-C measures, will prove to be less effective than it could be. These data support the approach taken in the Netherlands and Wales of only utilizing costly cascade testing resources in the families where the proband has an identified mutation, since in the majority (at least 75% i.e. the top three quartiles of the 6-SNP score) of the no-mutation patients, the most likely explanation for their clinical diagnosis of FH is a polygenic cause. In individuals with a clinical diagnosis of FH with a SNP score in the lowest quartile, it is however unlikely that there is a polygenic cause and while a mutation in one of the three known FH genes may have been missed for technical reasons, research to identify whether the individuals have a mutation in a yet to be identified gene would be valuable.</p>

PubMed Author Manuscript

Putting Synthesis into Biology \xe2\x80\x93 A Viral View of Genetic Engineering Through de novo Gene and Genome synthesis

The rapid improvements in DNA synthesis technology hold the potential to revolutionize biosciences in the near future. Traditional genetic engineering methods are template dependent and make extensive but laborious use of site-directed mutagenesis to explore the impact of small variations on an existing sequence \xe2\x80\x9ctheme\xe2\x80\x9d. De novo gene and genome synthesis frees the investigator from the restrictions of the pre-existing template and allows for the rational design of any conceivable new sequence theme. Viruses, being amongst the simplest replicating entities, have been at the forefront of the advancing biosciences since the dawn of molecular biology. Viral genomes, especially those of RNA viruses, are relatively short, often less than 10,000 bases long, making them amenable to whole genome synthesis with the currently available technology. For this reason viruses are once again poised to lead the way in the budding field of synthetic biology \xe2\x80\x93 for better or worse.

putting_synthesis_into_biology_\xe2\x80\x93_a_viral_view_of_genetic_engineering_through_de_novo_gene

A brief history of DNA synthesis<!>Methods for the assembly of long synthetic DNA<!><!>Assembly of synthetic genes and genomes<!>Assembly PCR<!>Ligase chain reaction (LCR) followed by fusion PCR with flanking primers<!>Limitations of current oligo-based DNA synthesis methods<!><!>Limitations of current oligo-based DNA synthesis methods<!>Codon Optimization<!>Creating new chassis for protein engineering<!>Viral Gene and Genome Synthesis<!>Exploiting the intrinsic sequence biases of the human genome for the generation of synthetic virus vaccines<!>General requirements for the application of SAVE to a virus system<!><!>General requirements for the application of SAVE to a virus system<!>Societal implications of synthetic biology<!><!>Societal implications of synthetic biology<!>

<p>The chemical synthesis of nucleotide chains took its first infant steps soon after the discovery of the DNA double helix. The race to elucidate the genetic code was driven by the use of triplet sequences of ribonucleotides synthesized by liquid-phase chemistry. Depending on their sequence these triplets selectively interacted with amino-acylated tRNA (the codon:anticodon recognition)(Nirenberg and Leder, 1964; Soll et al., 1965), which led to the assignment of codons to their respective amino acids, and to a much deserved Nobel Prizes for these heroic efforts in these earliest days of synthetic biology. Khorana's group "raced" to synthesize the first DNA copy of the 75 base pair long tRNAAla in 1970 (Agarwal et al., 1970) a monumental task requiring 20 man-years of labor, only to be outclassed by himself in 1979 by a 207 bp DNA cassette containing the tyrosine suppressor tRNA gene (Khorana, 1979).</p><p>The innovations of synthesizing DNA oligonucleotides ("oligos") on solid supports (Letsinger and Mahadevan, 1965) combined with new activated phosphoramidite nucleosiodes (Caruthers et al., 1987) led to steady improvements in the availability of quality oligos up to 100 bases long. This resulted in a boost in gene synthesis activity throughout the 1990's that continues unabatedly today. Some of the most notable synthesis achievements are summarized in Figure 1 (Agarwal et al., 1970; Becker et al., 2008; Blight, Kolykhalov, and Rice, 2000; Cello, Paul, and Wimmer, 2002; Chan, Kosuri, and Endy, 2005; Edge et al., 1981; Ferretti et al., 1986; Gibson et al., 2008; Gupta et al., 1968; Kalman et al., 1990; Khorana, 1979; Kodumal et al., 2004; Nirenberg and Leder, 1964; Pan et al., 1999; Soll et al., 1965; Stemmer et al., 1995; Tian et al., 2004). Significant landmarks include the synthesis of an entire 2.7 kb plasmid sequence by Stemmer et al. (Stemmer et al., 1995), the 4.9 kb MSP-1 gene of Plasmodium (Pan et al., 1999), the 7.5 kb of the poliovirus genome as the first synthetic self replicating organism (Cello, Paul, and Wimmer, 2002), and the 32 kb polyketide synthase gene cluster (Kodumal et al., 2004). The trend has culminated in the recent synthesis of 582,970 base pairs corresponding to the first artificial bacterial genome by the group of Craig Venter (Gibson et al., 2008). Starting with 101 prefabricated segments of 5–7 kb in length (purchased from commercial vendors), Gibson et al. used state of the art methods and brute force to assemble larger and larger DNA pieces, at first by recombination in bacteria, and finally in yeast (Gibson et al., 2008). Alas, the synthetic genome was not, or could not, be "booted" to life, by transplanting the genome into an "empty" chassis as the group has shown previously with a natural genome (Lartigue et al., 2007). Therefore, the first synthetic autonomous life form is still just below the horizon.</p><!><p>It is not yet possible to synthesize entire genes as long continuous strands of DNA from scratch. Rather, all synthetic genes are assembled from short custom made single stranded DNA oligonucleotides or "oligos", which are literally strings of a few nucleotides. Oligos are by-and-large still synthesized the same way as they were 15 or 20 years ago. Through incremental improvements in instrumentation and higher throughput, oligos have become a cheap commodity for use in standard recombinant DNA technologies. But, more than anything else, great demand and even greater competition by manufacturers has driven the oligo prices down by about 10-fold over the past 15 years (Figure 2). In comparison, the prices of finished, sequence confirmed, gene synthesis by commercial gene foundries have plummeted 50 fold in only 10 years (Figure 2). As a reference point, at the outset of the poliovirus synthesis project [15] in 1999 commercial gene synthesis was simply unheard of. As recently as 2000, after much searching, we found a vendor who agreed to synthesize parts of the genome by special arrangement at a price of $12/bp (Cello, Paul, and Wimmer, 2002).</p><p>In the ideal world, an efficient and economical de novo gene synthesis platform would combine cheap error-free oligo synthesis with accurate assembly methods. Neither one are currently available. There are two dramatically different methods of synthesiszing oligos. In the traditional, time-proven, method of solid-phase oligo synthesis each oligo is synthesized individually, on a separate small column or a well on a multiwell plate. The method is high yielding but costly ($ 0.10–0.20 per nucleotide synthesis cost), which is a critical aspect if the oligos are needed for the assembly of long DNA sequences. The price given above translates into an oligonucleotide cost of approximately $ 200–400 for a 1kb DNA sequence and that's for the raw material only.</p><p>The development of optical deprotection chemistries heralded a new era of parallel synthesis methods on micro biochips (Fodor et al., 1991) that can be used for both oligo or peptide synthesis. Depending on the chip platform being used, several thousands to hundreds of thousands of distinct oligonucleotides can theoretically be synthesized on a single chip.</p><p>In an ingenious extension Tian and collegues (Tian et al., 2004) mated the light-induced deprotection chemistry with microfluidic technology that allows the programmable synthesis of thousands individual oligonucleotides on a tiny chip (Figure 3A). At the heart of this method is the Digital Light Processing technology (DLP) that was developed for digital projectors and High Definition Projection TV sets. On a microfluidic chip containing a labyrinth of thousands of connected tiny reaction chambers (Figure 3C), each chamber is computer-addressable by a light beam generated on a digital micromirror device (Singh-Gasson et al., 1999) (akin to the individual color light spots making up the projection-TV picture). A DNA synthesis mixture containing the first nucleotide (A, for instance) is pumped through the system. Here, A only "sticks" to the chambers which call for an A at the specific position in their sequence, which are the ones that are being illuminated at that time (Figure 3A). Although all chambers receive the same synthesis mixture at any given times, no reaction occurs in the chambers that are "left in the dark" (in the example above, the ones that need a C, G, or T at their corresponding position). After the first reaction, the A-mix is washed out and the next reaction mix, containing the next nucleotide is pumped in and the process is repeated, four times in total. After all four nucleotide reaction mixes have gone through the chip, in each chamber the oligonucleotide chain has now grown by at least one nucleotide of the desired sequence.</p><p>At the end of the reaction the oligonucleotides are eluted from the chambers as a single pool. Each of the oligo sequences is only present in minute quantities. This may present a challenge in further increasing the throughput by increasing the number of reaction chambers per chip, while decreasing chip size. Tian et al. demonstrated the potential power of this technology for the synthesis of large numbers of oligonucleotides to be used in synthetic gene assembly (Tian et al., 2004).</p><p>Companies already offer parallel on-chip-synthesized custom oligo mixtures that are amenable for gene synthesis (LC Sciences, Houston Texas). Currently the price of a pool of 3,912 90-mers is approximately $1000. This technology is still very much in the exploratory stage. One inherent difficulty of the method is that all oligos are released from the chip as a mixture. The low yields of oligos that come off the chip (107 –108 molecules per sequence) are insufficient to drive a gene assembly reaction, which mandates a post-synthesis PCR amplification step before oligos can be used. For this purpose each oligo is synthesized with two flanking generic adaptor sequences, which allows amplification of all oligos in parallel in a single PCR reaction using the corresponding adaptor primer pair (Figure 4) (Tian et al., 2004). Using distinct sets of adaptors on distinct subsets of oligos in the same chip-synthesis reaction allows the subsequent selective amplification of a desired subset of oligos, for instance a set necessary for the assembly of one particular gene. Therefore, it is possible that in a separate reaction a different set of oligos can be amplified from the same chip-eluted oligo mix. Thus fractioning the entire oligo pool into gene-specific subsets will reduce complexity of the mixture, increase concentration of each specific oligo, and reduce potential interference or cross-hybridization from other oligos in the pool. This will be especially useful as the number of individual sequences synthesized on the chip increases. The higher the number of discrete oligo sequences synthesized per chip, the lower the absolute yield per oligonucleotide (sub fmole range) because the total yield of DNA is a direct function of the total reaction surface on the chip. With more distinct oligos the potential for unwanted cross-hybridizations during the gene assembly step also increases.</p><p>The second drawback of the chip-based oligo synthesis is that the PCR amplified oligos are now in a double stranded form. The presence of a perfectly matched antisense strand may reduce the efficiency in the subsequent assembly of these oligos into larger genes. The assembly reaction depends on the complementarity of the overlapping "construction" oligos, those designed to build the gene, and the antisense oligos are likely to compete more effectively for the same hybridization partner. To overcome this problem the desired single stranded construction oligos can be selectively enriched by specific hybridization to antisense selection-oligos affixed to a column and subsequent elution (Tian et al., 2004). When done under stringent enough conditions this procedure also contributes to a significant elimination of error-containing oligos, as they produce mismatches with the selection oligo and consequently elute from the column at a lower temperature. On the downside, this method requires twice the amount of selection oligos than there are contruction oligos. In other words, to produce one chip's worth of oligos one needs two additional chips's worth of selection oligos, tripling the cost of synthesis (Tian et al., 2004). This brings the current "rock-bottom" cost of the final construction oligos before the gene assembly to about $0.03/bp.</p><p>While these new multiplex synthesis systems are technically feasible it is our understanding that the major suppliers of large synthetic DNA for now continue to assemble genes from individually synthesized overlapping oligonucleotides by traditional methods.</p><p>The sheer number of different oligonucleotides synthesized on a chip mandates the use of new software programs to handle the complexity of possible interactions of the various oligo sequences in the mix (Czar et al., 2009). Several software programs are freely available to design optimal sets of assembly oligonucleotides. The basic tasks that successful software needs to perform are:</p><!><p>Breaking down the target sequences to be synthesized into suitable overlapping oligos.</p><p>Designing hybridization units, the overlapping portion between two oligos, with the same melting temperature.</p><p>Ensuring hybridization specificity of each oligo pair to eliminate potential cross-hybridization by choosing the best possible breaking points between oligos for a particular gene, and by altering synonymous codons.</p><!><p>There are two basic methods available for assembling long DNA sequences, such as virus genomes, from short overlapping synthetic oligonucleotides, direct assembly PCR and ligase chain reaction (LCR) followed by fusion PCR with flanking primers.</p><!><p>Assembly PCR is based on the principle of generating stepwise elongation of the amplicon, a piece of DNA formed in an amplification event, by one oligonucleotide at each end of the growing amplicon with each PCR cycle (Stemmer et al., 1995), and on the possibility of intermediate products to act as overlapping megaprimers to assemble even larger amplicons (Figure 4). Theoretically, the reaction continues until the two outermost oligos are incoproated to give the full length product. The full length product is subsequently amplified with an excess of the two flanking PCR primers. Practically, obtaining large DNA fragments in a single assembly reaction is exceedingly difficult. For this reason, and for error-management purposes, it is generally necessary to first synthesize, clone and verify the sequence of several intermediate size sub-fragments (500–1000 bp). These can then be linked by fusion PCR to form larger genes or by standard cloning methods.</p><!><p>The ligase chain reaction (LCR) is similar in that it uses overlapping oligos. But unlike with PCR assembly, oligos for LCR have to be designed to anneal without gaps between them, head to toe, forming annealed stretches of DNA which are then ligated using a thermostable DNA ligase (Barany, 1991). In contrast to PCR assembly where a single oligo is added at each end of a synthon in each cycle, during LCR several overlapping oligos can be ligated to one another. Owing to the thermostability of the ligase, LCR can be cycled similar to a PCR reaction, leading to assembly of longer and longer chains, but no net amplification. The desired product is finally amplified by PCR using gene-flanking primers.</p><!><p>Regardless of the many variations on the theme of how to assemble a large synthetic DNA, at the core of all current methods are chemically synthesized oligonucleotides. The downward price trend for oligos has slowed significantly over the past 5 years and appears to be bottoming out (currently in the $0.10–0.20/base range). As the price gap, and therefore the profit margin, between finished synthetic genes and their oligo building blocks is narrowing, it can be expected that oligo-based gene synthesis prices will soon follow. For long DNA synthesis to become economical, radically new technologies need to be developed that either reduce the errors in run-of-the-mill oligos by orders of magnitude, or allow de novo gene synthesis independent of the error-prone oligonucleotide chemistry, perhaps by developing enzyme based synthesis of long accurate polynucleotides. Barring such breakthrough, the routine synthesis of bacterial or larger genomes will likely remain prohibitively expensive for some time to come. As a case in point the recent synthesis of the Mycoplasma genome (Gibson et al., 2008) cost an estimated $ 10 million (Herper, M. 2007, http://www.forbes.com/2007/06/28/venter-synthetic-bacteria-tech-science-cx_mh_0628venter.html). At the research level on the other hand, once gene synthesis hits the $0.10–0.20/bp price range synthesis will very likely replace the traditional recombinant DNA methods for many smaller scale cloning projects within the next few years.</p><p>A major problem with genes assembled from overlapping oligos is the inherent error rate of about 1% during the chemical synthesis of the oligos themselves. The most frequent error is the failure to incorporate bases due to less than perfect deprotection of the reactive groups or incorporation of the incoming nucleotide. It appears that there is a rather hard limit for improving the oligo accuracy during the synthesis step much beyond the 1/100. Therefore several techniques are being employed, often in combination, to improve the accuracy of oligos and the assembled DNA intermediates.</p><!><p>Keeping the oligos and the overlapping regions between them short (40–50 bases) not only reduces the relative error rate per nucleotide in the oligo but also increases the disruptive effect of mismatches between annealed oligos. Using stringent hybridization conditions thus reduces the chance of incorrect oligos to partake in the assembly reaction (Young and Dong, 2004).</p><p>A common approach is to gel purify oligos before the assembly reaction, which helps eliminate many of the shorter aberrant oligo species. This reduces the error rate to about 1 in 500. At this error rate, short, several hundred base pairs long, intermediate assembly products are cloned by traditional recombinant DNA methods and sequence verified. The vetted sequence segments are then either combined by further rounds of cloning, or by assembly PCR. The need for gel purification is another reason to keep oligo length limited, as oligos that are too long can no longer be effectively separated from the most troublesome offender, the N–1-mer. If all construction oligos for one specific synthesis project are kept the same length, the gel purification can be done by combining all oligos in one sample, much reducing time and cost (Smith et al., 2003).</p><p>Another approach relies on the selective hybridization of the construction oligos to a column of immobilized selection oligos (Tian et al., 2004), as noted above.</p><p>Finally, a second tier of error correction can be implemented after the LCR or PCR assembly of gene fragments. It is based on the enzymatic activity of T7 endonuclease, which recognizes and specifically cleaves dsDNA at mismatched nucleotide pairs (Picksley et al., 1990; Young and Dong, 2004). Following the final PCR amplification the DNA amplicon is heat denatured and re-annealed. Since mutations in the original construction oligo sequences are distributed randomly the probability of two hybridizing strands to carry a mutation on one and the corresponding compensatory mutation on the other oligo is miniscule. It can therefore be expected that virtually every mutation in every oligo that participates in the assembly reaction will create a mismatch. Similarly, error correction by mismatch binding proteins, such as MutS of Thermus aquaticus, can be employed, facilitating the separation of the MutS bound mismatched DNA from the correct DNA by gel electrophoresis (Carr et al., 2004).</p><!><p>The quality of the oligos critically determines the practical size of the synthesis intermediates that need to be cloned and sequence verified (Carr et al., 2004). If sequence errors follow a normal Gaussian distribution along the length of the DNA an error rate of 1 in 600 would make it impractical to assemble a DNA longer than 1–2 kb in a single reaction without intermediate sequence verification (Figure 5).</p><!><p>In many cases it is desirable to express a gene of interest (often a human gene) in a heterologous, more economical, expression system, such as bacteria or yeast. All too often, however, the codon usage within the gene is at odds with the codon usage of the new host species. As a result the gene expresses poorly. Thus, the need for "codon optimization" was born (Itakura et al., 1977). During codon optimization the codon usage of the gene is altered to reflect that of the host species by replacing suboptimal codons with preferred synonymous codons. Since this often involves many simultaneous sequence changes, it is best done by de novo gene synthesis. Probably the best known example of codon optimization is the "humanization" of the Green Fluorescent Protein (GFP) of the jellyfish A. Victoria (Zolotukhin et al., 1996). Codon optimization is currently still the most prevalent reason for de novo gene synthesis (Gustafsson, Govindarajan, and Minshull, 2004).</p><p>In some instances gene synthesis has been used to recreate a DNA sequence from a publicly available sequence database in an effort to sidestep licensing, patenting or material transfer issues.</p><!><p>It is theoretically possible to synthesize a bacterial genome in which the redundancy of the genetic code is eliminated, such that each amino acid in every bacterial protein is represented by exactly one codon only. Thus, only 20 codons plus one Stop codon would be needed to synthesize all the bacteria's own genes. At the same time, the remaining 43 "orphaned" codons could be freed up to specify non-natural amino acids. Bacteria with such an expanded genetic code could one day become a powerful chassis for the production of artificial proteins (2006; Carr and Isaacs, 2006).</p><!><p>Viruses are amongst the simplest replicating genetic systems. For this reason they have been at the forefront of the advancing biosciences since the dawn of molecular biology. Their small genome sizes (most RNA virus genomes are 10+/−5 kb) makes them amenable to whole genome synthesis with the currently available technology. For this reason viruses are poised to lead the way in the budding field of synthetic biology.</p><p>A significant use for genome synthesis consists in the recreation of viruses or perhaps other organisms in the future, for which no intact natural template is available. The synthesis of the 1918 flu virus was accomplished by piecing together sequence fragments recovered from victims buried in the Alaskan permafrost and archived tissue samples (Tumpey et al., 2005). The creation of bat SARS coronavirus (Becker et al., 2008) and HIV from Chimpanzee feces (Takehisa et al., 2007) also fall into this category. A clever extension of this idea has been the resurrection of live infectious retroviruses assembled from a consensus of ancient remnants that are endogenous to the human genome, and which have perhaps been inactive for millions of years (Dewannieux et al., 2006; Lee and Bieniasz, 2007). Once the stuff of science fiction movies, these "Jurassic Parkesque" projects are likely to be just the teaser trailers of the coming attractions in the budding synthetic technology.</p><p>Through the process of natural selection, evolution favors systems that work, especially those that work better than their direct predecessors and competitors. This selection process however does not follow what humans would consider a logical design process. Evolutionary changes are small and incremental following a one-directional ratchet that does not move backward. There is no "reset" button that allows evolution to jump back to an earlier version and try again. De novo gene and genome synthesis provides this virtual reset button by allowing the creation of any conceivable genome at will and at once, no matter how different from its predecessor.</p><p>One recurring theme in viral genomes is the evolution of overlapping reading frames. This space saving measure allows a virus to encode portions of two proteins on the same stretch of genome sequence, but in two different reading frames. Studying individual genes and proteins of such a virus genetically and biochemically poses a problem for the experimenter, since manipulating one protein inadvertently changes the other. To simplify these interdependencies in the genome Chan and colleagues redesigned and synthesized parts of the bacteriophage T7 genome, eliminating the overlapping reading frames (Chan, Kosuri, and Endy, 2005). In the resulting virus, the individual genes could be then manipulated and studied independently, a process they called "refactoring" in analogy to the process of redesigning and improving computer code, while retaining it's basic function.</p><!><p>The basic mechanism of mRNA translation is preserved from the simplest virus to the most complex organism. Viruses, just like human cells need to produce mRNA molecules, which are used to convert their genetic information into proteins. Different viruses have devised different strategies to accomplish this, and have different ways to store this genetic information in their genome. Invariably, however, viruses need to divert the host's cellular machinery for the translation of their proteins, as they themselves cannot execute this function. The degeneracy in the genetic code (several synonymous codons specify the same amino acid) gives an organism the flexibility to encode a given protein sequence in its genome in an unimaginably large number of ways. The poliovirus polyprotein, for instance, could be encoded by a staggering 101100 different mRNA sequences, all of them specifying the same protein sequence (for comparison, the number of atoms in the observable universe is estimated to be on the order of 1080). This raises the question to what extend the natural encoding of a gene is optimal or special. The cell's preference of one synonymous codon over another to specify the same amino acid is termed "codon bias". It is thought that codon bias is correlated with the abundance of the corresponding cognate tRNAs in the cell. Consequently, rare codons are associated with a suboptimal translation of an mRNA. In addition, the frequencies of which two codons occur next to one another in the genome are not what is statistically expected from the frequencies of the two codons that make up the pair - a phenomenon called the "codon-pair bias". There are codon-pair combinations that are statistically greatly underrepresented while others are greatly overrepresented. The significance of codon pair bias has been largely unknown and underappreciated. We have recently shown that it is possible to exploit the codon-pair bias phenomenon for the synthesis of novel live attenuated forms of viruses with incredible properties (Coleman et al., 2008). By large-scale computer-aided redesign of the viral genome we engineered hundreds of silent mutations into poliovirus. These mutations were targeted to introduce a maximum number of unfavorable synonymous codon-pairs, without changing codon bias or protein sequence. By forcing a virus to "make do" with this heavily biased synthetic genome we showed that viral protein translation is greatly reduced. Thus, codon-pair deoptimized viruses cannot reproduce their genetic information as quickly as their wild type cousins which puts them at a decisive disadvantage against the host's innate and immune defences. One of the major benefits of the whole-genome deoptimization strategy is that the resulting attenuated viruses are phenotypically and genotypically extremely stable. The attenuation (att) phenotype is dependent on many hundreds, even thousands, of silent mutations, each by themselves virtually inconsequential, or "death by a thousand cuts". Therefore, the fitness gain from reverting individual mutations appears to be too small to drive genetic selection, and thus, reversion apparently does not occur (Coleman et al., 2008). We termed this process of perturbing intrinsic viral genome biases by synthetic genome re-design SAVE for Synthetic Attenuated Virus Engineering (Figure 6).</p><p>SAVE attacks a virus at one of the most fundamental processes common to all living systems, the translation of protein, for which viruses depend on the host cell's machinery. Thus it should be predicted that SAVE may work on most if not any virus. The rational genetic changes imposed on SAVE designed viral genomes are completely independent of protein sequence. The viral protein sequences, and therefore their function remain 100% preserved in the recoding process. Therefore an understanding of the proteins function is not necessary, sidestepping the need of most of classic virology in order to produce an attenuated vaccine candidate in a very short time with a predictable degree of attenuation in virtually any virus system. Viruses live lives of genetic austerity, and therefore don't usually carry unnecessary genes around. By that rationale most viral genes product can be considered essential. Depending upon the virus system, interfering with the synthesis of several of those genes just a little bit turns out to pack a great punch against the overall fitness of the virus (Coleman et al., 2008; Mueller et al., 2006)</p><p>Using the SAVE method we can profit from these genomic biases that have arisen over evolutionary time-scales and turn them upside down and inside out, undoing eons of viral evolution. If we think of evolution as "walking" along a dirt path, SAVE allows us to "leap" across the evolutionary universe at warp speed. Since it is evident that many viruses have actively selected against the occurrence of certain sequence features, such as unfavorable codons, codon-pairs, as well as other sequences motifs, the whole genome recoding approach by de novo synthesis will very likely have a profound effect on any virus.</p><!><p>Since SAVE targets a virus at the level of protein translation, a function elementary to all viruses, we believe this approach is applicable to many virus systems for which the following basic requirements are met:</p><!><p>A target virus has a known genome sequence, preferably available online.</p><p>The desired de-optimized genome sequence are prepared by computer aided redesign using the SAVE algorithm</p><p>De novo synthesis of the artificial viral genome is performed according to the design specifications, usually outsourced to a commercial vendor.</p><p>A reverse genetics system is employed to boot the artificial genome to life and make a virus. This is decidedly simple for many human viruses. Often a genome length copy of the DNA itself or an RNA transcript of that DNA is infectious upon transfection into susceptible cells.</p><p>A method to screen for viruses of desired phenotype has to be available. An initial screen in susceptible cell culture will yield valuable information as to the viability of various deoptimized virus designs. Clearly the virus still must be able to replicate at least at a low level in order to be useful as a live vaccine.</p><p>A suitable animal model to test attenuation and immune response is required.</p><!><p>Provided above requirements are met, SAVE strategy can successfully be employed for redesign and synthesis of viruses.</p><p>Synthetic virology, i.e. the redesign and synthesis of custom-tailored whole virus genomes, has become economically feasible with recent rapid improvements in DNA synthesis technology. This holds the potential to revolutionize the way virology and vaccinology is done. Viral genomes, especially of RNA viruses and retroviruses are short enough to make them amenable to whole genome synthesis with currently available technology. Such freedom of design could provide tremendous power to perform large-scale redesign of DNA/RNA coding sequences, to study the impact of large-scale changes in codon bias, codon-pair bias, dinucleotide biases, GC content, RNA secondary structures, and other sequence signatures, on viral fitness, with the aim to develop a new platform for vaccine design and genetic engineering.</p><!><p>What is synthetic biology? It is neither a field in its own right, nor a separate science. It is perhaps best described as an improvement of existing enabling technologies that are beginning to penetrate mainstream sciences, as they become more and more economical. This has led to an "organized" crossover of different scientific fields (e.g. biology, chemistry, mathematics, engineering etc.) that promises to yield organisms with useful biochemical pathways never seen before.</p><p>The new reality of synthetic genes and genomes calls for a fundamental revision of the ways biology is taught to students. The Johns Hopkins University has already embraced these cutting-edge developments, and is now offering an undergraduate course, in which the students collaboratively work toward synthesizing the yeast genome. Impressively, within only one year this unified effort resulted in the synthesis of hundreds of 750bp cassettes amounting to the 280kb of the yeast chromosome III (Dymond et al., 2009). An equally imaginative and playful introduction to engineering of biological systems is fostered by the International Genetically Engineered Machine Competition (iGEM; http://www.igem.org) organized by synthetic biologists at MIT. Here undergraduate teams compete in designing and building genetic circuits and systems from an ever expanding toolkit of standard genetic parts, or "BioBricks™" (Goodman, 2008)</p><p>However, although the excitement about synthetic biology is substantial enough, it faces equally big scepticism and "fear of the new" in our society. A disservice to their own science is perhaps the tendency of some researchers in the "synthetic biology field" to overvalue its novelty and uniqueness. The most commonly cited public concerns with regard to synthetic biology are probably the ethical implications connected with the creation of "new life forms" and the fear of synthetic "killer viruses". These sentiments are often picked up and fuelled by the media potentiating the perceived fear of the uncertain.</p><p>Virtually every organism ever modified in molecular or genetic research is by definition a new life form. This definition could be expanded to all naturally occurring organisms that genetically differ from their parent, in other words: all the living creatures. Why would an organism created by synthetic methods be qualitatively different? The question presents itself: "Why do we, as a society, worry more about the possibility of a synthetic designer pathogen, when some of the worst pathogens known to mankind are still raging?" Measles virus, as a case in point, is one of the most contagious viruses to humans. As recently as in 2000, approximately 777,000 people died per year from measles, and in third world countries with poor health care systems the fatality rate can be as high as 28% (Perry and Halsey, 2004). Annually, 250,000 – 500,000 people die from complications of the flu (WHO, 2003). Additionally, only a few critical mutations in the H5N1 bird flu virus separate us from a virus that can easily spread amongst humans and lead to an influenza pandemic. The AIDS pandemic, caused by primate viruses that jumped the species barrier to humans, claims approximately 2 million lives annually (http://www.avert.org/worldstats.htm). In 2003, the world barely escaped a pandemic by a SARS-coronavirus now thought to have jumped from bats to humans ((Becker et al., 2008) and references therein).</p><p>Although in theory at least, we have the capacity to generate any genetic sequence that we can conceive, what we can do with this capacity is in fact quite limited. While it's easy to think up fantastic and scary scenarios of a synthetic killer viruses wiping out mankind, bio-terrorists and the brightest scientific thinkers alike would be hard pressed to say what such a designer super-pathogen would look like. In reality, all that can basically be accomplish via synthesis for now and for some time to come, is to emulate, copy and recreate what mother nature has brought forth and thrown at us incessantly throughout our history on this planet. It is possible to produce variations on an existing theme. It is not possible, as yet, to design from scratch a qualitatively new pathogen, that is completely different from any organism that exists now or has existed in the past. The level of abstraction required to "piece together" qualitatively new lifeforms form defined of the shelf parts (genes), is far from being realized (Goler, Bramlett, and Peccoud, 2008). It is probably this misconception, trumpeted by the media, which strikes a cord of fear in the general population. Cases in point:</p><!><p>The 2002 poliovirus synthesis (Cello, Paul, and Wimmer, 2002), the first synthesis of a pathogen, caught the world off guard and ignited a heated debate in its aftermath. All we had done was to recreate an exact synthetic copy of the poliovirus genome, except for some genetic "watermarks" to prove the authenticity of the synthetic genome. The resulting virus was at the protein level 100% identical to the wild type virus used in countless laboratories around the world, a virus that even now naturally circulates in several countries and that is available for purchase at repositories such as the American Type Culture Collection (ATCC). Being an exact antigenic match to the currently available poliovirus vaccine, an overwhelming proportion of the world population is immune against this virus. Worldwide vaccine coverage against poliovirus is arguably the greatest of any vaccine preventable disease. This is hardly a blueprint for an imminent bioterrorist attack. But it was suddenly becoming clear that viruses can never be regarded as extinct, as long as their genome sequence information is preserved, be it on a government-sponsored online database, a 29 year old Nature journal (Kitamura et al., 1980) gathering dust in libraries across the world, or just written down on a smudgy piece of paper forgotten in a desk drawer... It is sufficient to re-create a virus at any point, even long after any traces of it's natural presence have vanished. It is this uncomfortable realization that brought about the level of public discussion that the original poliovirus synthesis had. The publication was intended not only to herald a new era in the study of organisms but also as a "wake-up call" for dual use technology.</p><p>The recreation of the highly pathogenic 1918 flu virus (Tumpey et al., 2005) out of sequences extracted from influenza victims preserved in the northern permafrost also met with criticism, although no one had maligned the publication of the genome sequence as much as 8 years earlier (Taubenberger et al., 1997). In fact, the synthesis the 1918 virus brought critical new insight into the pathogenesis of the influenza and it is a prerequisite for the production of an adequate vaccine should such a need ever arise. Isn't society in the long run much better off with this knowledge than without it, understanding 1918 flu virus in detail rather than hoping that something like the 1918 flu will never happen again? This sentiment is even more inappropriate with the looming threat of the H5N1 bird flu pandemic.</p><p>Over 30 years of random, "unenlightened" genetic manipulation of viral genomes through recombinant DNA technology by countless laboratories around the world has not shown any evidence, that researchers would accidentally and unbeknownst to them create a human super-virus. Whole genome synthesis will be no different.</p><p>The adapation of a human pathogen to an experimental animal species by repeated passaging through that species (a decidedly "pre-synthetic era" method) has been employed ever since viruses were discovered. It leads to the increased pathogenicity in the new species compared to the wild type virus. These host-adapted models have greatly facilitated the study of viruses and the diseases they cause. Equally important these experiments resulted in the development of some of the most successful vaccines ever produced (polio, measles, mumps, rubella, and smallpox). As it turned out passaging these viruses through diverse animal species lead to the mitigation of their disease-causing potential for humans – a process termed "attenuation".</p><!><p>All the above considerations notwithstanding, de novo genome synthesis, like many technologies in the past, does hold a potential for dual use. And unlike many technologies before it, nuclear proliferation for instance, which require immense resources that cannot escape detection, the intentional misuse of genome synthesis technologies will become increasingly undetectable. It seems next to impossible that genome synthesis can ever be government-regulated effectively. The technology and its components are too ubiquitous already, and too easy to jury-rig from off-the-shelf parts. The nature of genome synthesis is such that in the very near future pathogens can, and perhaps will, be synthesized in the proverbial hobbyist's basement, high school Science lab or by a bio-terrorist organization. These possibilities are not an academic's hyperbole either. In fact the grass roots "bio-hacker" culture is already flourishing, outside the realm of academia, industry and government oversight (Nair, 2009). When considering these issues our society would be prudent to shift focus from prevention of such dual use proliferation to preparing for it. The latter may include the development of new vaccines and/or the stockpiling of available vaccines against the most likely bio-terrorist agents.</p><!><p>Pushing the limits - a historical progression of notable achievements in gene synthesis with references. Each point represents a report of an individual gene synthesis accomplishment with respect to the length of the synthetic sequence and the year it was first reported.</p><p>Price development of oligonucleotide synthesis and de novo gene synthesis. Shown are the approximate end user prices per base for oligonucleotides (desalted, non-purified) or per base pair for synthetic genes (below 3kb, sequence guaranteed). The data was compiled from a "look back" of vendor invoices, and a survey among colleagues. While by no means comprehensive, the prices shown here are representative of what the typical research laboratory paid for these services at the time.</p><p>Microfluidic chip technology coupled with light activated chemistries hold great promise for the massive parallel synthesis of oligonucleotides. (A) On an array of tiny flippable mirrors, each mirror can be separately computer-controlled (flipped to an "ON" or "OFF" position). Mirrors in the "ON" position reflect light onto their corresponding reaction chamber on a micro-fluidic chip (bright blue spots), leading to the incorporation of the nucleotide currently loaded on the chip (here, A-mix). While all chambers receive the same nucleotide mix at any one time, no reaction occurs in the dark chambers (black spots). The process is repeated with the next nucleotide mix and a new light pattern, which specifies the chambers to incorporate the new nucleotide. After the last nucleotides are incorporated, the finished oligos are released from the chip and collected as a pool (B) actual size of a microfluidic chip holding 4000 sequence features. (C) A magnified view of the interconnected microscopic reaction chambers on a Atactic microfluidic chip. (B–C) Reproduced with permission by LC Sciences, LLC, Houston, Texas</p><p>Assembly of gene sequences from chip-synthesized oligonucleotides. The pool of overlapping oligos in minute amounts is released from the microchip, followed by PCR amplification with universal adapter primers. Double strand copies produced in this way are subjected to type II restriction enzymes to remove the adapter sequence. Construction oligos are purified by stringent hybridization to immobilized selection oligos. This leads to the elimination of the unwanted antisense oligos and reduces the error frequency in the construction oligos. Next, the eluted construction oligos are heat denatured and reannealed, and subjected to PCR cycling to produce intermediate or final DNA products. The reaction is driven by excess concentration of a gene flanking primer pair</p><p>The impact of oligonucleotide error rate on the accuracy of assembled synthetic genes. The various curves assume error rates in the construction oligonucleotides typically achieved after different error correction methods used to assemble a target sequence are 1/600 (red; using gel purified oligos), 1/1400 (blue; using hybridization selected oligos) and 1/10,000 (black; using mismatch specific endonucleases). Adapted from Carr et al., 2004.</p><p>Recoding of viral genomes according to the SAVE method (Synthetic Attenuated Virus Engineering). A. Example of the level of sequence alteration after codon reassignment of the poliovirus capsid gene (Mueller et al., 2006). PV(M), part of the wild type capsid coding sequence; PV-AB, the same amino acid sequence encoded by rare codons; PV-SD, the same amino acid sequence encoded by random shuffling of synonymous codons present in the wild type sequence. Note, that the amino acid sequence encoded by all three sequences reamins the same. B. Codon pair bias after "SAVE"-mediated codon reassignment of viral genes. The codon pair bias (CPB) score for each of 14,795 confirmed annotated human genes was calculated. Each red dot represents the calculated CPB score of one human gene plotted against its amino acid length. Predominant use of under-represented codon pairs yields negative CPB scores. The codon-pair scores of three wild type viral genes fall within the bulk of the human genes. After computer-aided recoding and de novo synthesis of the viral genome according to the SAVE algorithm the new genes ("Min" for minimized CPB) have extremely unfavorable CPB, unlike any gene the cellular translation machinery has ever encountered. Note, that the amino acid sequence of all proteins remains unchanged during this process. By analogy to other virus systems a decreasing CPB leads to reducded translatability of the mRNA and increased attenuation of the virus. After Coleman et al., 2008.</p><p>This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.</p>

PubMed Author Manuscript

Stimulus-responsive Controlled Release System by Covalent Immobilization of an Enzyme into Mesoporous Silica Nanoparticles

Mesoporous silica nanoparticles (MSN) have emerged as an attractive class of drug delivery carriers for therapeutic agents. Herein, we explored the covalent immobilization of proteins into MSN to generate a stimulus-responsive controlled release system. First, MSN were functionalized with thiol groups using (mercaptopropyl)-trimethoxysilane (MPTMS). Functionalization was verified by X-ray photoelectron spectroscopy (XP), Fourier-transform infrared (FTIR) spectroscopy, and dynamic light scattering. The model enzyme carbonic anhydrase (CA) was coupled to sulfosuccinimidyl 6-[3\'(2-pyridyldithio)-propionamido]hexanoate (Sulfo-LC-SPDP) at a low ratio of 1:1 to prevent enzyme inactivation and subsequently covalently immobilized into MSN via thiol-disulfide interchange. The enzyme could be released from MSN with 10 mM glutathione which represents intra-cellular redox conditions while it remained bound to the MSN at extra-cellular redox conditions represented by 1 \xce\xbcM glutathione. The activity of the released enzyme was >80% demonstrating that the enzyme was still largely functional and active after immobilization and release. Human cervical cancer (HeLa) cells were incubated with the MSN-CA bioconjugates at various concentrations for 24 h and the data show good biocompatibility. In summary, we demonstrate the potential of MSN as potential drug delivery systems for proteins.

stimulus-responsive_controlled_release_system_by_covalent_immobilization_of_an_enzyme_into_mesoporou

INTRODUCTION<!>EXPERIMENTAL PROCEDURES<!>Thiol-functionalized Mesoporous Silica Nanoparticles (MSN-SH)<!>X-ray Photoelectron Spectroscopy (XPS)<!>FTIR Spectroscopy<!>Dynamic Light Scattering (DLS)<!>Carbonic Anhydrase Modification with Sulfo-LC-SDPD<!>Degree of Protein Modification<!>Circular Dichroism (CD) Spectroscopy<!>Carbonic Anhydrase Activity<!>Covalent Immobilization of CA into MSN<!>Cytotoxicity Assay<!>Residual Activity of Released CA from MSN<!>Release of Carbonic Anhydrase from MSN in the Presence of Glutathione<!>Statistical Data Treatment<!>MSN Activation and Characterization<!>CA Activation and CA-SPDP Characterization<!>MSN-CA Bioconjugate and its Characterization<!>Controlled Release of CA from Nanoparticles<!>Cell Viability Studies<!>CONCLUSION

<p>There has been a growing interest in utilizing nanotechnology in the development of drug delivery systems.1 Nanomaterials, such as polymers, mesoporous silica nanoparticles and nanotubes have emerged as potent drug delivery devices.2–4 Likewise, the employment of proteins including enzymes as therapeutic agents has increased due to their selectivity and catalytic efficiency.5 Due to recent advances in nanotechnology, the integration of nanosized materials (e.g., nanoparticles) and therapeutic agents (e.g., enzymes) is now conceivable.</p><p>Drug delivery systems provide important tools for enhancing the efficacy of therapeutic agents thus reducing unwanted side effects. For example, locally higher doses of chemotherapeutic agents can be delivered using nanoparticles with 100–300 nm in diameter because they accumulate preferentially in tumors.6,7 However, challenges remain in the design of nanosized delivery systems. For example, it is highly desirable that drug delivery systems show zero-premature release until the target is reached and that the release of the drug is controlled by a stimulus-responsive design.8,9 Stimuli can include pH,2 electric fields,10 and photoirradiation5 and have been utilized for drug release.2, 9, 11 In general, stimulus-responsive release systems allow a smart release of drugs by responding to endogenous or exogenous activation.4,12–17 Recently, specific chemical reactions, such as disulfide reduction, have emerged as alternative mechanisms for drug release.16 The main driving force of the drug release in the cell is by reduced glutathione for such systems.18</p><p>Herein we selected silica as the material for the development of a nanosized delivery system. Once injected, MSN degrade into biocompatible monomeric silicic acid (Si(OH)4). Si is an essential human nutrient and been implicated in protecting against the toxic effects of aluminum and in promoting calcification.3 In recent years, the use of silica nanoparticles has been extended to biomedical and biotechnological fields, such as biosensors, DNA deliver, drug delivery, and enzyme immobilization.19 It has been demonstrated that MSN can be endocytosed by cells and thus can serve as carriers for the controlled intracellular release of drugs.20–23 In addition, MSN are capable to escape the endosomal entrapment6, 21, 24–26 and can potentially protect the delivered protein drug from exposure to proteases. In addition, the large pore volume that MSN possess allow for loading of large protein amounts into the nanoparticles.21, 24–26 Slowing et al. encapsulated cytochrome c (Cyt-c) into MSN and realized studies related to the uptake and release of Cyt-c by mammalian cells.21 They demonstrated that the Cyt-c remained active after its release from MSN. However, as far as we know, there are no reports on conjugating enzymes into MSN via stimulus-responsive bonds, which would allow for the smart release under physiological conditions solely in the presence of a targeted stimulus. This was the goal of our work.</p><p>Herein, we report and demonstrate the use of chemically functionalized MSN to form conjugates with an enzyme via formation of stimulus-responsive covalent bonds (Figure 1). Carbonic anhydrase (CA) was selected as the model biological molecule since it is a well-characterized enzyme with easily measurable bioactivity, which makes it an ideal candidate for the proof-of-concept of our system.27,28 CA (EC 4.2.1.1) in this study was from bovine erythrocytes and thus belongs to the group of α-CAs, Zn2+ containing metalloenzymes. CAs catalyze the conversion of CO2 and H2O to H2CO3, essential to various biological processes including maintaining the acid-base balance in blood and tissues, enabling efficient transport of carbon dioxide out of tissues, and increasing the CO2 concentration in chloroplasts.29 We demonstrate that CA can be released from the silica nanoparticles under conditions emulating intra- but not extra-cellular redox conditions with more than 80% of residual activity. The system developed is the first step in the direction of a smart protein delivery system which could be used to treat various diseases.</p><!><p>Mesoporous silica nanoparticles (product number 643637), (3-mercaptopropyl)trimethoxysilane (MPTMS), 4-nitrophenyl acetate (p-NPA), L-glutathione, dithiothreitol (DTT), toluene (99.9%) and acetone (99.9%) were from Sigma-Aldrich (St. Louis, MO). Carbonic anhydrase (CA, EC 4.2.1.1) from bovine erythrocytes was from Worthington Biochemical Corporation (Lakewood, NJ). Sulfosuccinimidyl 6-(3'-[2-pyridyldithio]-propionamido)hexanoate (Sulfo-LC-SPDP) was from Proteochem (Denver, CO). The CellTiter 96® AQueous non-radioactive cell proliferation assay was from Promega (Madison, WI). HeLa cells were purchased from ATCC (Manassas, VA). Dialysis membranes were obtained from Spectrum Laboratories (Rancho Dominguez, CA). All reagents were used as supplied without further purification.</p><!><p>MSN were activated by functionalization with thiol groups. The activation was done by the addition of 3.75 ml of (mercaptopropyl)-trimethoxysilane (MPTMS, 95%) to 3.75 ml of a stirred suspension of 125 mg of MSN in refluxing toluene at 110°C for 1 h. After slow cooling of the mixture, the solid phase was recovered by filtration and washed with 200 ml of toluene and 500 ml of acetone. The product was dried overnight under vacuum.</p><!><p>XPS was performed using a PHI 5600ci spectrometer with an Al Kα monochromatic source (350 W, 15 kV) and at a takeoff angle of 45°. Survey spectra were obtained at 187.85 eV.</p><!><p>FTIR studies were conducted with a ThermoNicolet Nexus 470 FTIR spectrometer. Silica nanoparticles and the conjugates were measured as KBr pellets (0.5 mg sample per 200 mg of KBr). Each sample was measured at least five times.</p><!><p>The size of particles was determined by dynamic light scattering (DLS) using a DynaPro Titan.</p><!><p>The protein was dissolved in 50 mM phosphate buffered saline (PBS) with 0.15 M NaCl and 10 mM EDTA at pH 7.2 to a final concentration of 2 mg/ml. Sulfosuccinimidyl 6-[3'(2-pyridyldithio)-propionamido] hexanoate (Sulfo-LC-SPDP, 8.5 mg) was added directly to the reaction flask. The mixture was reacted for 30 min at room temperature under gently stirring and dialyzed thrice against nanopure water at 4°C using cellulose ester membranes with a 10 kDa cut-off. A volume ratio of 1:100 (sample-tonanopure water) was used during dialysis. Modified carbonic anhydrase (CA-SPDP) was subsequently lyophilized for 48 h using a 6-L lyophilizer (model 77530, Labconco, Kansas City, MO) at a condenser temperature of −45°C and a pressure of <60 m Hg. The lyophilized powder was stored until use at −20°C.</p><!><p>The extent of enzyme modification (i.e., the number of amine groups coupled to the cross-linker) was determined by measurement of the release of pyridine-2-thione and the 2,4,6-trinitrobenzene sulfonic acid (TNBSA) assay. For the pyridine-2-thione release assay modified carbonic anhydrase was dissolved in 50 mM PBS with 0.15 M NaCl and 10 mM EDTA at pH 7.2 to achieve a protein concentration of 0.5 mg/ml. The absorbance of the samples was measured and recorded at 343 nm. Then, 10 μl of 15 mg/ml dithiothreitol (DTT) solution was added to each sample. After 15 min, the absorbance of the samples was measured and recorded at 343 against a blank treated as above. The molar ratio of SPDP to protein was calculated. For the TNBSA assay modified carbonic anhydrase was dissolved in 0.1 M sodium bicarbonate buffer, pH 8.5 at a concentration of 0.05–0.2 mg/ml. Then 0.25 ml of 0.01% (w/v) of a TNBSA solution was added to 0.5 ml of sample solution. The solutions were incubated at 37°C for 2 h. After incubation 0.25 ml of 10% SDS solution and 0.125 ml of 1 N HCl were added to each sample. The absorbance of each sample was measured at 335 nm against a blank treated as above. The amount of SPDP bound to the protein was obtained from a calibration curve. For both assays the samples were prepared in triplicate.</p><!><p>CD spectra were acquired with an OLIS DSM-10 UV-Vis CD spectrophotometer at 20°C in the near-UV region (250–320 nm) at 0.5 nm spectral resolution using a 10 mm quartz cell with a protein concentration of 0.6 mg/ml in 15 mM Trissulfate buffer at pH 7.6 and 25°C. Each spectrum was obtained by averaging six scans. Spectra of buffer blanks were measured prior to the samples and were digitally subtracted from the sample CD spectra.</p><!><p>Carbonic anhydrase (1.2 μM) activity was determined at 25°C by monitoring the hydrolysis of p-nitrophenol acetate (p-NPA) in 15 mM tris-sulfate buffer, pH 7.6. The hydrolysis was monitored at 400 nm using a Shimadzu 2450 UV/Vis spectrophotometer. The assay was performed in 1 ml cuvette. The reaction was started by mixing 700 μl of 15 mM tris-sulfate buffer, pH 7.6, 100 μl of 16.56 mM p-NPA dissolved in acetonitrile, and 200 μl of enzyme solution. The residual activity was calculated with respect to the specific activity of native carbonic anhydrase.30,31</p><!><p>2.0 mg of MSN-SH was sonicated for 3 min in 50 mM PBS, 0.15 M NaCl, and 10 mM EDTA, pH 7.2 in Eppendorf vials to create a suspension. Different volumes of a carbonic anhydrase stock solution were added to each MSN-SH suspension. The mixtures were gently stirred overnight at 4°C. Then the samples were centrifuged at 14,000 rpm for 20 min. To remove the unreacted enzyme three cycles of washing/centrifugation were performed. The washing steps were performed with 50 mM PBS containing 0.15 M NaCl and 10 mM EDTA at pH 7.2. The amount of immobilized enzyme was determined by depletion by measuring the protein concentration in the supernatants obtained during washing and compare it to the total protein amount employed. Protein concentration in the supernatant was determined from the absorbance at 280 nm using a calibration curve generated with native carbonic anhydrase.</p><!><p>Experiments were performed using a human cervical cancer cells line (HeLa). The cells were cultured as recommended by ATCC in minimum essential medium (MEM) with L-glutamine supplemented with 10% fetal bovine serum (FBS) and 1% penicillin. HeLa cells were incubated at 37°C under 5% CO2 and used before 25 passages. Mitochondrial function was assessed using the CellTiter 96 aqueous non-radioactive cell proliferation assay (Promega, Madison, WI). About 40,000 cells/well were seeded into 96-well microtiter plates in 100 μL of penicillin free culture medium with 10% FBS. After 24 h, culture medium was replaced with culture medium containing serial dilutions of the MSN-CA bioconjugate suspensions. The cells were incubated with the suspended bioconjugate for 24 h. The MSN suspension settled down over time and consequently the cells were in direct contact with the nanoparticles during incubation. Then, 20 μL of 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt (MTS) was added to each well. After 2 h, the optical intensity of each well was measured spectrophotometrically at 490 nm in a microplate reader (Thermo Electron Corporation, Multiskan Ascent).32,33 The spectrophotometer baseline was obtained using culture medium without cells. Experiments were performed 8 times.</p><p>Controls: HeLa cells treated with 2 μM staurosporin for 6 h were used as positive control and cells without any treatment were used as negative control.34 The statistical analysis for the cell viability results indicated that under the tested conditions the cytotoxicity obtained for a CA-SPDP concentration ≥ 30.87 μg/mL is highly significant from the obtained for the negative control (p < 0.001). Consequently the concentration range used for the experiments was 0.3 – 30.87 μg/mL.</p><!><p>The MSN-CA bioconjugates were suspended in 1 ml of 50 mM PBS containing 10 mM DTT. The samples were incubated overnight at 37°C under constant shaking at 125 rpm. Then, the samples were centrifuged at 14,000 g for 20 min to precipitate the MSN. DTT was removed from the supernatant using Centricon devices with an exclusion size of 10 kDa and the released enzyme obtained in the supernatant. The specific activity of modified carbonic anhydrase was obtained using these supernatants and the residual activity determined.</p><!><p>The release of CA from MSN was measured by preparing a suspension of 5 mg of MSN-CA conjugates in 1 ml of 50 mM PBS with 1 mM EDTA at pH 7.4 and glutathione (GHS) concentrations of 0, 0.001, and 10 mM. The chemical stability of thiols is affected by parameters such as temperature, pH, and the presence of metal ions. Addition of a chelating agent, i.e., EDTA, greatly improves the chemical stability of thiols.35 Incubation was performed for 24 h at 37°C and the MSN were pelleted by centrifugation at 14,000 rpm for 20 min. The supernatant was removed and used to determine the concentration of released CA and the pelleted MSN were resuspended in fresh GHS-PBS buffer. The solubility of CA Carbonic is >10 mg/ml according to the manufacturer. The MSN-CA conjugate used in the release experiments contained a total of 1.1 mg of immobilized CA. Therefore, all the release experiments were performed under sink conditions.</p><!><p>All data were obtained at least in triplicate, the data averaged and standard deviations calculated. Statistical significance for comparisons of multiple groups was established using one-way multiple Tukey comparison post-test ANOVA. All statistical analysis were performed with InStat 3.06 (GraphPad Software Inc., San Diego, CA, USA).</p><!><p>The goal of our investigation was to immobilize CA in MSN by introducing a redox-sensitive covalent SS-bond to accomplish release under intracellular redox conditions. To accomplish this it was necessary to first chemically decorate MSN with thiol groups. This was accomplished using (3-mercaptopropyl)trimethoxysilane (MPTMS). MPTMS reacts with the hydroxyl groups of silica forming a covalent bond. Thiol groups are introduced in this way into the material and presented terminal of the propyl groups. XPS was used to investigate the MSN prior to and after modification to confirm the introduction of the thiol groups (Figure 2). The introduction of SH groups into the MSN was evident from photoemission peaks of the 2s and 2p orbitals of S in the MSN-SH spectrum, which were absent in the spectrum of MSN. FTIR spectroscopy was also employed to characterize the product. A broad band centered at 3442 cm−1 was observed in the MSN-SH spectrum corresponding to hydroxyl stretching vibrations (Figure 3). Peaks at lower frequencies are due to various vibrational modes of the silicate (Si-OH and Si-O-Si). Bands at 2922 and 2861 cm−1 are due to partial hydrolyzation of the material used to synthesize the silica nanoparticles.36 However, importantly the spectrum of MSN-SH shows a band at 2571 cm−1 consistent with SH-stretching vibrations which is missing in the spectrum of MSN.</p><p>Dynamic light scattering (DLS) was employed to determine if the chemical modification process had an effect on the particle size or would cause aggregation of the suspended silica nanoparticles (Table 1). In principle introduction of SH-groups could cause formation of interparticle disulfide bonds and thus covalent aggregation. DLS data revealed two different hydrodynamic particle size distributions for both samples. Furthermore, somewhat in contrast to our expectations, the data revealed a reduction in the hydrodynamic radius upon chemical modification of the particles (Table 1). SEM micrographs were obtained to evaluate the effect of the modification process on the morphology of the MSN. No substantial morphology changes were detected in the micrographs (for details see Supporting Information). The characteristics of the particles after modification were still in line with the end application in mind, namely passive targeted delivery. It should be noted, however, that for clinical applications nanoparticles with a diameter of over 400 nm should be removed from the preparation because these particles will be effectively cleared by the mononuclear phagocytic system (MPS).</p><!><p>We selected carbonic anhydrase (EC 4.2.1.1, CA) as the model enzyme in our studies because it is a well-studied, accessible, and relatively inexpensive enzyme.27 First we introduced a SH-group at the surface of the enzyme in order to immobilize it into the thiol functionalized MSN. This was accomplished by utilizing the reagent sulfo-LC-SPDP which reacts with amines through the succinimidyl group. The SPDP molecule also has a disulfide-containing linkage that can be cleaved with reducing agents such as DTT and glutathione.37 After performing the reaction with the enzyme, the extent of enzyme modification was assessed using two different assays. In the first the release of pyridine-2-thione was determined, which is the product of the reaction of Sulfo-LC-SPDP with a thiol containing reagent.36 The results were corroborated by using the 2,4,6-trinitrobenzene sulfonic acid (TNBSA) assay, which determines the amount of amino groups on the protein surface (largely the ε-amino groups of lysines).38 Under the conditions employed in this work, both assays demonstrated that the protein was modified with the cross-linker at a linker-to-carbonic anhydrase molar ratio of 1.0 ± 0.5, which represents a ca. 5% modification of the available enzyme surface amino groups. The enzyme activity was reduced to 80.0 ± 0.7% by the chemical modification. [It should be noted that higher levels of modification caused larger enzyme inactivation (for details see Supporting Information).] Near UV-circular dichroism (CD) spectroscopy was employed to investigate if the enzyme tertiary structure was affected by the modification. The spectra revealed that no significant changes in tertiary structure of SPDP-CA occurred after attaching the cross-linker (Figure 4).</p><!><p>SPDP-CA was covalently immobilized into the thiol-functionalized MSN. The strategy was that the 2-pyridyl disulphide group of the immobilized cross linker on the enzyme surface would react with the thiol groups introduced into MSN by thiol-disulfide interchange to form aliphatic disulphides.37 To establish the amount of SPDP-CA that could be immobilized into the MSN-SH we reacted different amounts of SPDP-CA with 2 mg of MSN-SH. The amount of immobilized enzyme increased proportionally to the amount of enzyme added into the MSN-SH suspension (Figure 5). The maximum immobilization was determined to be 577 ± 65 mg of carbonic anhydrase per 1.0 g of MSN-SH under our conditions.</p><p>The bioconjugate was characterized by XPS and FTIR to confirm protein immobilization. The XPS spectrum of MSN-CA presents a new N (1s) signal confirming enzyme immobilization (Figure 2). The FTIR spectrum of the MSN-CA bio-conjugate shows the appearance of two strong bands at 1645 cm−1 and 1542 cm−1 due to the protein amide I and II vibrational modes. Thus, these data unequivocally demonstrate immobilization of the enzyme into the functionalized MSN. However, these data do not distinguish adsorption from covalent immobilization.</p><p>However, proof was obtained by being able to remove the CA from the MSN using 10 mM DTT but not with buffer. The activity of the released enzyme was measured and was compared with that obtained for CA-SPDP (see Materials and Methods section for details). The enzyme retained 79±11% of activity after immobilization and release. This demonstrates some detrimental impact of the procedure on enzyme integrity. While some inactivation of the enzyme is probably inevitable using such strategies, future experiments are being designed to minimize this by increasing protein thermodynamic stability and minimizing protein-surface interactions. To identify potential changes in the morphology of the MSN due to enzyme immobilization we employed SEM and dynamic light scattering. The SEM micrographs show that no major morphology changes (size and porosity) occurred due to enzyme immobilization (for details see Supporting Information). Dynamic light scattering showed that the size of the MSN-CA bioconjugates was similar to that of non-modified MSN (Table 1).</p><!><p>In vitro release studies were conducted to assess the release of immobilized CA from the nanoparticles under somewhat physiological conditions in PBS at pH 7.4 and 37°C. Intracellular glutathione concentrations (1–10 mM) are sufficient to cleave disulphide bonds and thus should afford release of the immobilized enzyme while the concentration of 1 μM glutathione representing extracellular plasma conditions should have no effect.16 However, release could also be promoted by other cell-produced antioxidants, such as NADH and dihydrolipoic acid.39 It is in consequence possible that the intra-cellular release in vivo could be faster than in our in vitro assay. Fresh glutathione had to be added daily to each sample during release due to limited glutathione stability in the in vitro release assays. We found that only a small amount of 1% and 3% of CA was released using no or 1 μM concentrations of glutathione (Figure 6). In contrast, when exposing the CA-MSN to 10 mM glutathione, CA was released from the nanoparticles (Figure 6). The fact that carbonic anhydrase release only occurred in the presence of glutathione proves that the enzyme was indeed covalently immobilized and not absorbed at the surface in agreement with our spectroscopic evidence.</p><p>The protein was completely released from the MSN under reducing conditions in 480 h. In contrast, in systems using adsorption for protein immobilized typically only 45–55% of the immobilized protein was released.21 Figure 6 also shows that the CA release profile at intracellular glutathione concentrations is sigmoidal which is characteristic of an energy-dependent release process. This type of system is characterized by slower release at initial stages followed by increased release at later stage.40</p><p>In summary, our results demonstrate the ability of our system to release the enzyme under intracellular conditions but not at extracellular conditions.</p><!><p>Cell viability is essential when creating a new drug delivery system to avoid non-selective cytotoxic events and was thus studied herein. Several publications have already established the non-cytotoxity of MSN.19,23,41 However, the cytotoxity of MSN-CA bioconjugates has not been investigated. Cell viability was studied via the measurement of cell metabolic activity. The mitochondrial function was measured using the MTS assay after incubating HeLa cells with different concentrations of MSN-CA bioconjugate for 24 h. Viable cells convert tetrazolium salts (MTS) to formazan dyes that are measured spectrophotometrically.33 The assay was performed varying the concentrations of the silica and the enzyme. Figure 7a shows the obtained results of the assay performed in terms of varying silica concentration. In general, the cytotoxic effect of the MSN-CA conjugate is lower than that of the MSN-SH alone (except for the 17.15 μg/mL concentration). Cell viability decreased from 97% to 63% at increasing CA concentration compared with 100% to 80% at increasing MSN-CA conjugate concentration, respectively (Figure 7b). However, statistical analysis by ANOVA demonstrated that there was no significant difference between the two groups (p >0.05).</p><!><p>In summary, we have explored the covalent immobilization of a model protein into MSN to generate a stimulus-response controlled release. We demonstrate that the release of the immobilized enzyme occurs under intra-cellular but not extra-cellular redox conditions. The released enzyme was still functional and active but some activity decrease was encountered. Cell viability studies establish good biocompatibility of the generated bioconjugate. Future experiments include studies related to endocytosis and endosomal escape of the bioconjugate. We envision that this bioconjugate could serve as a platform for the creation of a drug delivery system utilizing therapeutic proteins.</p>

PubMed Author Manuscript

Specificity of Herbivore Defense Responses in a Woody Plant, Black Poplar (Populus nigra)

The specificity of woody plant defense responses to different attacking herbivores is poorly known. We investigated the responses of black poplar (Populus nigra) to leaf feeding by three lepidopteran species (Lymantria dispar, Laothoe populi and Amata mogadorensis) and two leaf beetle species (Phratora vulgatissima and Chrysomela populi). Of the direct defenses monitored, increases in trypsin protease inhibitor activity and the salicinoid salicin were triggered by herbivore damage, but this was not herbivore-specific. Moreover, the majority of leaf salicinoid content was present constitutively and not induced by herbivory. On the other hand, volatile emission profiles did vary among herbivore species, especially between coleopterans and lepidopterans. Monoterpenes and sesquiterpenes were induced in damaged and adjacent undamaged leaves, while the emission of green leaf volatiles, aromatic and nitrogen-containing compounds (known to attract herbivore enemies) was restricted to damaged leaves. In conclusion, indirect defenses appear to show more specific responses to attacking herbivores than direct defenses in this woody plant.Electronic supplementary materialThe online version of this article (10.1007/s10886-019-01050-y) contains supplementary material, which is available to authorized users.

specificity_of_herbivore_defense_responses_in_a_woody_plant,_black_poplar_(populus_nigra)

Introduction<!>Plants and Insects<!><!>Plant Harvest and Quantification of Experimental Leaf Damage<!>Defense Hormone Analysis<!>Trypsin Protease Inhibitor Activity<!>Salicinoid Analysis<!>VOC Collection and Analysis<!>Statistical Analyses<!><!>Defense Hormones<!><!>Defense Hormones<!><!>Salicinoid Concentrations<!>Volatile Organic Compounds<!><!>Volatile Organic Compounds<!>Discussion<!>

<p>Plant chemical defenses of many types are well known to be induced upon attack by insect herbivores. Such induction is sometimes thought to be specifically tailored to the attacking herbivore species giving rise to terms such as the specificity of plant responses (Karban and Baldwin 1997), the specificity of elicitation (Stout et al. 1998) and the specificity of induced resistance (Agrawal 2000). However, the causes and mechanisms of insect herbivore-specific responses in plants are not yet fully understood. Recent studies have investigated whether a plant responds in an herbivore-specific manner may dependent on the feeding guild of the insect, the level of feeding specialization (reviews by Ali and Agrawal 2012; Bonaventure 2014; Heidel-Fischer et al. 2014 and references therein) or salivary cues (Erb et al. 2012) and herbivore-associated microbe communities (Acevedo et al. 2015).</p><p>However, most investigations of the specificity of plant response to different insect attackers have focused on only a single defensive compound or compound group. For example, Van Zandt and Agrawal (2004) reported that the volume of pressurized latex, a putative anti-herbivore defense in milkweed, was differentially induced after herbivory when comparing monarchs (Danaus plexippus) to swamp milkweed beetles (Labidomera clivicollis). Silva et al. (2017) observed differences in the profiles of tomato volatiles when comparing plants infested by the whitefly Bemisia tabaci to plants infested by the leaf miner Tuta absoluta. Studies of multiple classes of defenses and defense signals are uncommon.</p><p>Research on the specificity of plant defense induction has also concentrated on herbaceous rather than woody plant species, which are less studied due to the methodological problems accompanying their size and longevity (Lämke & Unsicker 2018). However, these characteristics of woody plants may lead to different responses to herbivores. First, throughout their longer lives woody plants may repeatedly encounter the same herbivore species. Second, they may be under constant attack during the growing season, from leaf flush (van Asch and Visser 2007) to senescence (White 2015). Third, among aboveground organs, the larger amount of biomass concentrated in stems may make losses of leaves to herbivores less critical. This might lead to a defense strategy where damaged tissue is sacrificed while defenses are concentrated in surrounding tissue. In light of these possibilities, both specific and non-specific defense responses might be viable strategies for woody plants under different conditions. While non-specific defenses are effective against more herbivores, a specific defense tailored to a single herbivore may be less costly (Onkokesung et al. 2016). However, there are comparatively few studies investigating the specificity of woody plant defense responses after herbivory, such as those conducted in willow (Fields and Orians 2006) and birch (Hartley and Lawton 1987). Yet these studies focus on a narrow set of defensive compounds, while investigations on a broad set of chemical compounds in combination with measurements of defense hormones are still missing.</p><p>The aim of this study was to investigate the defense responses of black poplar (Populus nigra) towards feeding by five different leaf-chewing insect herbivore species (two coleopterans and three lepidopterans) commonly occurring on poplar. We investigated herbivore-species-specific changes in the defense-related phytohormones salicylic acid (SA) and jasmonic acid (JA), salicinoids, trypsin protease inhibitor activity, and volatile organic compounds in black poplar to obtain a more complete picture about defense responses in woody plants. Salicinoids, a group of phenolic glycosides highly abundant in poplar trees (Boeckler et al. 2011) negatively affect generalist herbivore performance (Hemming and Lindroth 1995; Lindroth and Peterson 1988; Osier and Lindroth 2001). Protease inhibitors (Bradshaw et al. 1990; Haruta et al. 2001) and certain classes of volatiles (Clavijo McCormick et al. 2014b; Unsicker et al. 2015) are also typical poplar compounds reported to be active in defense against herbivores.</p><p>Among the herbivores, the two beetle species, Chrysomela populi (poplar leaf beetle) and Phratora vulgatissima (blue willow leaf beetle), and one of the lepidopteran caterpillar species Laothoe populi (poplar hawk moth) used in this study are specialist feeders, according to the classification by Ali and Agrawal (2012), because they feed on only a narrow range of tree species within the Salicaceae. In contrast the lepidopteran caterpillar species Amata mogadorensis and Lymantria dispar are true generalists, accepting host plant species from different plant families. Most of these herbivores may occur together on P. nigra, especially at the end of the season. In this study, we expected to find marked variations in the specificity of poplar defense responses both among various herbivore species and among different classes of defense metabolites. Since defoliation by chewing herbivores typically has only minor effects on the salicinoid concentrations of poplar (Boeckler et al. 2013, Osier and Lindroth 2001), we hypothesized that the induction of these phenolics would be weak and not herbivore species-specific. However, we expected protease inhibitor activity to be differentially induced, especially when comparing lepidopteran with coleopteran herbivores. Differential induction of protease inhibitors between these taxa has been described before (Chung and Felton 2011). It was also reported that the spectrum of volatile organic compounds induced by herbivores depends on feeding mode, the level of feeding specialization (Danner et al. 2018, Rowen and Kaplan 2016) and the composition of their oral secretions (Acevedo et al. 2015). We therefore hypothesized that volatile emission in black poplar would vary depending on the species identity of the attacker. In order to test these hypotheses, we investigated defense responses in both herbivore-damaged and nearby undamaged foliage.</p><!><p>Populus nigra saplings were grown from cuttings of young trees made in the summer. All genotypes were originally taken from a natural black poplar population located in a floodplain forest on the Oder River of northeastern Germany (52°34′1" N, 14°38′3″ E). The trees were reared in the greenhouse under summer conditions (24 °C; 60% relative humidity; 16 hr/8 hr, light/dark) in 2-L pots filled with a 1:1 mixture of sand and soil. The experiments were carried out in a controlled environment chamber (20 °C/18 °C, day/night: 60% relative humidity; 16 hr/8 hr, light/dark) to which trees were transferred 24 hr before the start of the experiments. All trees were regularly fertilized and watered once per day.</p><p>Lymantria dispar caterpillars were hatched from eggs obtained from the US Department of Agriculture (Buzzards Bay, MA, USA), reared on artificial diet (MP Biomedicals LLC, Illkirch, France) in a climate chamber (23 °C, 60% relative humidity, 14 hr/10 hr, light/dark) and used in experiments as 3rd instar larvae. Laothoe populi caterpillars were obtained in 1st instar from a commercial provider (The World of Butterflies and Moths, UK, http://www.wobam.co.uk) and reared on black poplar foliage under laboratory conditions until they were used in experiments as 4th instar larvae. Amata mogadorensis caterpillars were hatched from eggs obtained from a private breeder (https://www.entomologenportal.de) and reared on black poplar foliage under laboratory conditions until they were used in experiments as 3rd instar larvae. The two beetle species Chrysomela populi and Phratora vulgatissima were reared from egg clutches collected in old-growth black poplar trees in the field.</p><!><p>Experimental design (a), typical herbivore feeding pattern (b) and amount of herbivore damage (c) for experiments in which each of three different herbivores was tested on leaves of single black poplar genotypes. The ten full-sized leaves of each sapling were divided into two groups of five leaves each. Each group was wrapped in a polyethylene terephthalate (PET) bag attached to the saplings with cable binders at both ends and supported with a constant flow of charcoal-purified air. The herbivores were caged on the lower leaves and allowed to feed for 44 hr. Lower leaves (from inside the cage) were harvested as "damaged" leaves. Upper leaves from the same sapling were sampled as "adjacent undamaged" leaves. Comparable leaves harvested in the lower and upper leaf pool of non-damaged trees (control) functioned as controls. Differences in the extent of herbivory were analyzed using the non-parametric Kruskal-Wallis test</p><p>Experimental design (a), typical herbivore feeding pattern (b) and amount of herbivore damage (c) for the experiment testing the effect of four different herbivore species on volatile emission of black poplar saplings of five different genotypes. The approximately ten full-sized leaves of each sapling were divided into two leaf pools containing five leaves each. Each leaf pool was surrounded by polyethylene terephthalate (PET) foil and supported with a constant flow of charcoal-purified air. The lower leaf pool was exposed to four different herbivore treatments, and leaves harvested as described in the Fig. 1 legend. Caterpillars of Lymantria dispar, Laothoe populi, and adults of Phratora vulgatissima and Chrysomela populi were used as herbivores. While Lymantria dispar, Laothoe populi and C. populi have a biting-chewing feeding mode, P. vulgatissima feeds in a piercing-chewing style and has therefore a different feeding pattern. All herbivores were allowed to feed for 44 hr. Differences in the extent of herbivory were analyzed using the non-parametric Kruskal-Wallis test. Pairwise comparisons were made using the Dunn's post hoc test. Black dots represent outliers</p><!><p>Right after the experiments, all damaged and adjacent undamaged leaves from all treated trees were harvested and photographed after being spread out on a white board with a reference area. After the midribs were removed (due to difficulties in consistently grinding them to a powder), leaves were flash-frozen in liquid nitrogen and then stored in 5 ml plastic vials at −80 °C until further processing. In addition, the equivalent leaves of non-damaged control trees were separately frozen. All leaf material was lyophilized (ALPHA 1–4 LDplus, Christ, Germany) and ground to a fine powder using a paint shaker (Scandex, Pforzheim, Germany) and five stainless steel balls (diameter 3 mm). Experimental leaf area loss in the different herbivore treatments was determined by analyzing the digital images of the leaves with Adobe Photoshop (Version 15.0.0, Adobe Systems Incorporated, San Francisco, USA) following the method described in Boeckler et al. (2013).</p><!><p>Defense hormones were extracted from an aliquot of 10 mg ground lyophilized leaf material. The aliquot was dissolved in 1 mL of pre-cooled methanol (MeOH) containing the following internal standards [D6-abscisic acid (Santa Cruz Biotechnology, Dallas, TX, USA; 40 ng ml−1), D4-salicylic acid (Santa Cruz Biotechnology; 40 ng ml−1), D6-jasmonic acid (HPC Standards GmbH, Cunnersdorf, Germany; 40 ng ml−1), 13C-jasmonoyl-isoleucine (synthesis described in (Kramell et al. 1988), using 13C-Ile, Sigma Aldrich; 8 ng ml−1)]. The samples were shaken for 30 sec with a paint shaker. Then they were centrifuged at 2000 g for 5 min, and 400 μL of the supernatant were transferred into a new tube. The rest of the supernatant was carefully removed from the solid phase using a pipette. Another 200 μL portion of the supernatant was used for salicinoid analysis. Subsequently, 1 mL of fresh MeOH (without labeled standards) was added to the solid phase before repeating the extraction procedure (shaker + centrifuge). Again, 400 μL (and 200 μL for salicinoids) of the supernatant was collected and combined with the supernatant of the first extraction. The extracts were stored at −20 °C until measurement.</p><p>Defense hormones were analyzed using high performance liquid chromatography (Agilent 1100 Varian ELSD, Varian, USA) coupled to a mass spectrometer (API 5000 LC/MS/MS System, AB Sciex, Framingham, MA, USA). The analytes were separated on a C18 column (XDB-C18, 50 × 4.6 mm × 1.8 μm, Agilent, Santa Clara, CA, USA) using a formic acid (0.05% in water) / acetonitrile gradient (flow: 1.1 ml min−1) and detected via multiple reactions monitoring (MRM) in negative ionization mode (ion spray at −4500 eV at 700 °C) as described in Vadassery et al. (2012). Data were processed using Analyst 1.5.2 (Applied Biosystems, Foster City, CA, USA), and hormones were quantified relative to the peak area of their corresponding standard.</p><!><p>Protease inhibitor activity was analyzed via a radial diffusion assay (Jongsma et al. 1993). Samples of 10 mg of freeze-dried leaf material were dissolved in 400 μL of extraction buffer (25 mM Hepes, pH 7.2, adjusted with KOH, 3% PVPP, 2% PVP, 1 mM EDTA). After the addition of one steel ball (diameter 3 mm) and homogenization using a paint shaker (2 × 4 min), the samples were centrifuged at 4 °C and 2000 g for 10 min. A 200 μL portion of the supernatant was transferred into a 1 mL centrifuge tube and kept on ice until the analysis. An agar gel (1.8%) was prepared containing 2 μL/mL of fresh trypsin (Merck, Germany) dissolved in 25 mM Hepes-KOH buffer (pH 7.2). After pouring the gel solution onto a square petri dish, the gel was solidified for 3 hr at 4 °C. Subsequently, 5 mm-diameter wells were punched into the gel with a distance of 2 cm to each other using a hollow metal cork-borer. Along with the samples, a standard dilution series of bovine serum albumin (BSA) was added as reference. The gel was then incubated at 4 °C for 22 hr. After the gel was rinsed once with the extraction buffer (Hepes-KOH buffer) containing 10 mM CaCl2 and stained with a solution of 72 mg Fast Blue B Salt in 90 mL Hepes buffer (25 mM, pH 7.2, pre-warmed to 37 °C), a 60 mg portion of N-acetyl-DL-phenylalanine beta-naphthyl ester (APNE) dissolved in 10 mL N, N-dimethylformamide was added before pouring the solution on the agar plate (pre-warmed to 37 °C as well). Incubation time was 90 min before the staining solution was decanted and the gel was rinsed with water, and a reference curve with BSA was created following the protocol of Bradford (1976) with assays run in triplicate. Before usage, the BSA was reconstituted by mixing with deionized water.</p><!><p>Salicinoids were extracted during the procedure for the extraction of phytohormones (see above) with the addition of 0.8 mg/mL phenyl-β-glucopyranoside as an internal standard. The 2 × 200 μL extracts were combined and 400 μL of milli-Q-purified water was added before measuring the analytes via high performance liquid chromatography (HPLC). Analytes were injected onto a chromatographic column (EC 250 × 4.6 mm NUCLEODUR Sphinx RP, 5 μm, Macherey Nagel, Düren, Germany) connected to a precolumn (C18, 5 μm, 4 × 3 mm, Phenomenex). The temperature of the column oven was set to 25 °C. The mobile phase consisted of two solvents, solvent A (Milli-Q water) and solvent B (acetonitrile), from which solvent B was used in a gradient mode with time/concentration (min/%) of: 0:00/0; 19:00/52; 19:10/100; 21:00/100; 21:10/14; 26:00/14). The flow rate was set to 1 mL/min and injection volume to 20 μL. The signal was detected using photodiode array and evaporative light scattering detectors (Varian, Palo Alto, CA, USA). Using these settings and components, salicin eluted at a retention time of about 5.1 min, salicortin at about 10.2 min and homaloside D at about 15.2 min. The compounds were detected by absorption at 200 nm and identified by comparison of retention time in relation to those of standards isolated from previous work (Boeckler et al. 2013). Quantities were calculated on the basis of peak areas using standard curves prepared with pure standards corrected by the recovery of the internal standard.</p><!><p>VOCs released from various treatments were collected over a 4 hr period (9:00–13:00 hr) 40–44 hr after the insects were released on basal leaves of treated trees. VOCs in all treatments and leaf pools were trapped on five PDMS (polydimethylsiloxane) tubes (length: 5 mm) attached to 15 cm pieces of acetone cleaned aluminum wire hung inside each bag. PDMS tubes were prepared as described in Kallenbach et al. (2014). After the experiment tubes from each treatment and leaf pool were separately collected in glass vials (VWR International, Darmstadt, Germany) and frozen at −20 °C until further analysis.</p><p>Volatile analysis was performed with gas chromatography-mass spectrometry using the Ultra Thermo desorption unit TD20 connected to a quadrupole GC-MS-QP2010Ultra (Shimadzu, Kyoto, Japan). The PDMS tubes were placed in 89 mm glass TD tubes (Supelco, Sigma-Aldrich, Munich, Germany). After desorption in He with a flow rate of 60 mL/min at 200 °C for 8 min, the substances were cyro-focused onto a Tenax® adsorbent trap at −20 °C. The trap was then heated to 230 °C in 10 sec and the sample was injected into an Rtx-5MS column with a length/diameter of 30 m/0.25 mm and a film thickness of 0.25 μm (Restek, Bellefonte, PA, USA). Helium was used as carrier gas with a constant linear velocity of 44.3 cm/s. The TD-GC interface was held at 250 °C. The oven was set to 45 °C for 3 min, raised to 185 °C with an increase of 6 °C/min and subsequently to 320 °C at 100 °C/min with a 15 min hold. Electron impact (EI) mass spectra were recorded at 70 eV in scan mode from 33 to 350 m/z at a scan speed of 1666 Da/s. The ion source was held at 230 °C. Compounds were identified by comparison of mass spectra and retention times to those of authentic standards and spectra in Wiley and National Institute of Standards and Technology (NIST) libraries.</p><!><p>Analyses were carried out using SPSS Statistics version 20.0 (IBM, New York, USA). For the volatile analysis using different poplar genotypes, genotypes with more than one replicate were analyzed as one genotypic replicate by taking the mean of the replicates. If necessary the dataset was log-transformed before statistical analysis. To analyze differences in the leaf chemical composition between all treatments, including the control plants, ANOVAs were used. To analyze differences in the leaf chemical composition only between the herbivore treatments ANCOVAs were used with the herbivore damage (leaf area loss) as a co-variable (compound ~ herbivore damage*treatment). Both the ANOVA and ANCOVA models were checked for homoscedasticity, outliers and normal distribution of residuals. For some compounds, the assumptions were violated and could not be rescued with data transformation. Here the treatment was analyzed using the non-parametric Kruskal-Wallis rank sum test. Posthoc comparisons were performed using the Tukey-Kramer post hoc test (for ANOVA) and Dunn's post hoc test (for non-parametric Kruskal-Wallis test). For the analysis, the experiment block was left out as well (even if its importance as a factor was significant) because the importance was based on the herbivore damage, which differed between the experiment blocks. Principal component analyses were performed using the online platform MetaboAnalyst (https://www.metaboanalyst.ca). Data were scaled, (mean-centered and divided by the standard deviation of each variable) and transformed using generalized logarithm transformation.</p><!><p>Effect of damage by three herbivore species on the concentrations of two defense hormones, salicylic acid and jasmonic acid, in the damaged and adjacent undamaged leaves of young Populus nigra trees as compared to equivalent leaves from non-infested control trees. Samples were collected 44 hr after infestation with caterpillars of Amata mogadorensis and Lymantria dispar, and adults of the beetle Phratora vulgatissima, and from undamaged control plants. The boxplots depict medians ±1.5 x interquartile range of n = 10 tree replicates. Pairwise comparisons were conducted using Tukey's post hoc test (ANOVA) and Dunn's post hoc test (Kruskal-Wallis) and are indicated by small letters. Circles indicate outliers. Statistical results comparing only the herbivore treatments are given in Table 1</p><!><p>In the adjacent undamaged leaves, SA concentrations did not differ between the four treatments (Fig. 3). However, there were significant differences between the four treatments in the JA concentrations of the adjacent undamaged leaves. Pairwise comparisons revealed that JA levels were significantly higher in the P. vulgatissima-infested trees (Tukey-Kramer post hoc test: P = 0.015) compared to the controls and also to the Lymantria dispar-infested trees (Tukey-Kramer post hoc test: P = 0.012). In contrast, the JA levels of A. mogadorensis-infested trees were not different from the controls and from the JA levels of the other herbivore-infested trees (Fig. 3). JA concentrations in the adjacent undamaged leaves were generally lower compared to those of damaged leaves.</p><!><p>Effect of herbivore damage level and herbivore identity on defense metabolites of young Populus nigra trees in damaged and adjacent undamaged leaves</p><p>Non-parametric Kruskal-Wallis and ANCOVA tests were employed to determine the significance of changes in the concentrations of the phytohormones salicylic acid and jasmonic acid, concentrations of salicinoids, levels of trypsin proteinase inhibitor activity, and emission of major groups of volatiles. The number of replicates was n = 10 trees for phytohormones, salicinoids and trypsin protease inhibitor activity and n = 5 trees for volatile organic compounds. The tests were performed on the same dataset shown in the graphs, but excluding the control treatment to check for differences only between the plants infested by the different herbivore species. Whenever the assumptions for ANCOVA were met, % leaf area loss (damage) was integrated as a covariate. When ANCOVA assumptions were not met, non-parametric Kruskal-Wallis tests were performed (marked by the letter "a"). Bold numbers indicate significant results</p><p>aKruskal-Wallis H-Test</p><!><p>In the adjacent undamaged leaves, JA concentrations were significantly affected by herbivore damage levels and herbivore species identity (Table 1). For SA, herbivore species identity had no effect on the concentration. The effect of herbivore damage levels on SA concentrations in the adjacent undamaged leaves could not be tested as the statistical assumptions for ANCOVA were not met.</p><!><p>Effect of damage by three herbivore species on the trypsin protease inhibitor activity in the damaged and adjacent undamaged leaves of young Populus nigra trees as compared to equivalent leaves from non-infested control trees. Samples were collected 44 hr after infestation with caterpillars of the two lepidopteran species Amata mogadorensis and Lymantria dispar, adults of the coleopteran species Phratora vulgatissima, and untreated control plants. The boxplots represent the median ± 1.5 x interquartile range of n = 10 tree replicates. Pairwise comparisons were conducted using Dunn's post hoc test (Kruskal-Wallis) and are indicated by small letters. Circles indicate outliers and asterisks indicate extreme outliers. The results of statistical analyses comparing only the herbivore treatments are given in Table 1</p><p>Effect of damage by three herbivore species on the salicinoid concentrations in damaged and adjacent undamaged leaves of young Populus nigra trees infested by three different herbivore species as compared to equivalent leaves from non-infested control trees. Samples were collected 44 hr after infestation with caterpillars of two lepidopterans, Amata mogadorensis and Lymantria dispar, adults of one coleopteran, Phratora vulgatissima, and undamaged control plants. The box plots represent median ± 1.5 x interquartile range for n = 10 tree replicates. Pairwise comparisons were conducted using Tukey's post hoc test (ANOVA) and Dunn's post hoc test (Kruskal-Wallis) and are indicated by small letters. Circles indicate outliers. Statistical results comparing only the herbivore treatments are given in Table 1</p><!><p>In the adjacent undamaged leaves, salicin levels were also significantly different in all herbivore-infested trees when compared to non-damaged control trees (Dunn's post hoc test: A. mogadorensis P = 0.024, Lymantria dispar P = 0.008, P. vulgatissima P = 0.012), and there were no significant differences among the herbivore treatments. The levels of salicortin and homaloside D were not significantly different when all treatments were compared (Fig. 5).</p><p>In the damaged leaves, herbivore identity did not significantly affect the concentration of salicin although a trend was observed. However, the influence of herbivore damage levels on salicin concentration could not be tested because ANCOVA assumptions were not met. Salicortin and homaloside D levels were not affected by herbivore damage levels or by herbivore identity (Table 1). Also, in the adjacent undamaged leaves the influence of herbivore damage level could not be tested as statistical assumptions were not met. However, herbivore species identity did not significantly affect the concentrations of the three salicinoids measured. (Table 1).</p><!><p>To determine if different herbivore species cause different volatile responses in black poplar we set up a second experiment using multiple black poplar genotypes and a somewhat different set of herbivore species (Fig. 2). Phytohormone patterns in response to this set of insect herbivores were similar to the patterns observed in the first experiment (Fig. 3, Fig. S2). Altogether 86 volatile organic compounds were measured in this experiment, of which 69 could be (tentatively) identified (Table S1). A PCA performed with all identified volatiles measured in the headspace of the different treatments showed some separations between the herbivore treatments and the control treatment (Fig. S3). The volatile blends were further classified as monoterpenoids, sesquiterpenoids, green leaf volatiles (GLVs), aromatic compounds, nitrogenous compounds and "other volatiles" (compounds that did not fall into any of the chemical classes listed above), as we know from previous studies that certain volatile groups such as GLVs and nitrogenous compounds play essential roles in direct and indirect poplar defense.</p><!><p>Effect of damage by four herbivore species on the relative amounts of major groups of volatile organic compounds emitted from damaged (lower row) and adjacent undamaged leaves (upper row) of young Populus nigra trees as compared to equivalent leaves from non-infested control trees. Samples were collected 44 hr after infestation with caterpillars of two lepidopteran species, Lymantria dispar and Laothoe populi, adults of two coleopteran species, Phratora vulgatissima and Chrysomela populi, and untreated control plants. The box plots represent median ± 1.5 x interquartile range for n = 5 tree replicates. Pairwise comparisons were conducted using the Tukey-Kramer post hoc test (ANOVA) and Dunn's post hoc test (Kruskal-Wallis) and are indicated by small letters. Circles indicate outliers and asterisks indicate extreme outliers. Statistical results comparing only the herbivore treatments are given in Table 1</p><!><p>From the adjacent undamaged leaves, monoterpene emission differed significantly among the treatments (Fig. 6, upper row). While the two caterpillar species (Lymantria dispar and Laothoe populi) did not significantly induce monoterpene emission as compared to the equivalent leaves on non-damaged control trees, the two beetle species did (Dunn's post hoc test: P. vulgatissima P = 0.008, C. populi P = 0.006). Trends towards differences in monoterpene emission were also observed between Laothoe populi- and both beetle-infested trees (Dunn's post hoc test: P. vulgatissima P = 0.062, C. populi P = 0.060). Sesquiterpene emission in the adjacent undamaged leaves differed significantly between trees infested by beetles in the basal leaves and equivalent leaves on control trees (Dunn's post hoc test: P. vulgatissima P = 0.002, C. populi P = 0.005). In contrast, the two caterpillar species did not significantly induce sesquiterpene emission from undamaged leaves (Fig. 6, upper row). Differences in sesquiterpene emission were also observed between Lymantria dispar- and P. vulgatissima-infested trees (Dunn's post hoc test: P = 0.032). For aromatic and nitrogenous volatiles as well as for green leaf volatiles and other volatiles there were no significant differences among the treatments (Fig. 6 upper row, Fig. S1).</p><p>In the damaged leaves the emission of nitrogenous volatiles was significantly affected by the herbivore damage level and herbivore species identity, while sesquiterpene emission from damaged leaves was influenced by herbivore species identity (Table 1). In adjacent undamaged leaves none of the classified volatile groups was significantly affected by herbivore damage level and herbivore identity.</p><!><p>In this study we found that young black poplar trees damaged by the three different leaf-chewing herbivores tested in the single genotype experiment showed increases in the defense hormone jasmonic acid (JA), the salicinoid salicin and trypsin protease inhibitor activity. This was mainly observed in the damaged foliage, but in case of JA, also in the adjacent undamaged foliage. Additionally, all four herbivores tested in the second experiment induced different volatile organic compounds in the damaged as well as the adjacent undamaged foliage. While there was no herbivore-species-specificity for elicitation of the direct defenses surveyed, black poplar did display herbivore-specific emission of several classes of volatiles, in particular sesquiterpenes and nitrogenous compounds. In the case of sesquiterpenes the specificity of elicitation was also visible systemically in undamaged foliage adjacent to the attacked leaves.</p><p>When analyzing the two major defense-related phytohormones JA and SA, we found JA to be induced by the two leaf-chewing herbivores Lymantria dispar and P. vulgatissima, but not by A. mogadorensis. (Figure 3, Fig. S2). Local JA induction upon herbivore damage is a common phenomenon in herbaceous and woody plant species (Erb et al. 2012; Singh et al. 2016; Irmisch et al. 2014). The fact that SA was not induced by most of the herbivores investigated is in agreement with the literature. It is well documented that SA is mainly triggered by piercing-sucking insects like aphids (Li et al. 2016; Thaler et al. 2012) or infections by biotrophic pathogens (Kunkel and Brooks 2002). The general lack of SA induction by most of the herbivore species tested and induction of JA suggest a lack of specificity of defense signaling. The only exception was the specialist Laothoe populi that triggered the induction of SA in damaged leaves (Fig. S2). We also found that SA levels in damaged leaves were significantly affected by the amount of herbivore damage inflicted (ANCOVA, Table 1), even though there were no differences in SA concentrations between the different herbivore treatments (Fig. 3, Fig. S2). This result differs from other studies, where SA was not significantly influenced by chewing herbivores (Kawazu et al. 2012; Niveyro et al. 2013; Soler et al. 2012) although increasing and decreasing concentrations are also reported (Agrawal et al. 2014; Diezel et al. 2009). These observations demonstrate the complexity of the perception network involved in the recognition of herbivores by plants. This probably involves not only salivary cues, regurgitants and feces of herbivores, but also the associated herbivore microbiota. Investigations about the interaction of plants, herbivores and herbivore-associated microbes are just beginning and general models are hard to establish (Acevedo et al. 2015). The results obtained here and in other studies show that SA levels do respond to herbivory in a more subtle way than usually appreciated. The effects of the resulting signaling processes on the deployment of defenses are not known. Specificity might also be revealed by measurements of other hormones, such as ABA, ethylene and cytokinins (Erb et al. 2012), which were not quantified here.</p><p>Feeding by the generalist caterpillar species Lymantria dispar and one specialized leaf beetle, P. vulgatissima, increased the activity of trypsin protease inhibitors in damaged leaves (Table 1, Fig. 4). Also A. mogadorensis visibly increased the activity, although the differences were non-significant. The increased activity of protease inhibitors after wounding is a well-known inducible defense mechanism of plants (Jongsma and Bolter 1997). Since the production of protease inhibitors is associated with significant fitness costs (Zavala et al. 2004), their formation only in response to damage rather than being constitutively produced is understandable. Green and Ryan (1972) found the induction of protease inhibitors to be dependent on the number of wounding sites and the time after wounding. Although there were no significant differences in trypsin protease inhibitor activity in the leaves damaged by the different herbivore species, we observed a trend towards differential inductions (Table 1), which was probably caused by the higher numbers of wound sites from P. vulgatissima herbivory.</p><p>In contrast to most other black poplar metabolites measured, the major salicinoids, salicortin and homaloside D, were not induced by any of the herbivore species. A significant induction by leaf chewing caterpillars and beetles was only observed in the case of salicin (Fig. 5). Although there is little doubt about the role of salicinoids as defense compounds of Salicaceae plants (Boeckler et al. 2011), their induction patterns after herbivore attack are highly variable. While inductions of salicinoids are evident in some studies (Clausen et al. 1989; Fields and Orians 2006; Rubert-Nason et al. 2015; Stevens and Lindroth 2005) this is not always the case (Boeckler et al. 2013). The variability of herbivore-triggered salicinoid induction may arise because the levels of these phenolic compounds are influenced by many other factors. The most prominent factor is the genotype, which has been observed in many studies to cause much larger variation in salicinoid concentration than defoliation by herbivores (Osier and Lindroth 2001; Rubert-Nason et al. 2015). Other factors are the availability of nutrients and water (Hale et al. 2005) as well as organ, developmental and seasonal variation (Boeckler et al. 2011). Furthermore, individual salicinoids may be differentially induced after herbivory. The lower concentration of salicin compared to the other salicinoids measured does not necessarily mean that its defensive role is less important (Boeckler et al. 2016). In other species, inducible anti-herbivore metabolites with comparatively low concentrations but high impact on herbivores are known, such as indolic glucosinolates (Jeschke et al. 2016; Tian et al. 2005). Future studies should aim to investigate the toxicity and deterrency of herbivore-inducible salicin in comparison to the other less-inducible salicinoids.</p><p>When volatiles were measured, herbivory by two lepidopteran species and two leaf beetle species led to significant inductions of almost all major volatile groups (Fig. 6). The inducibility of plant volatiles after herbivory has been shown in both herbaceous (e.g. Fontana et al. 2009; Kigathi et al. 2013; Piesik et al. 2016; Skoczek et al. 2017) and woody plants (e.g. Courtois et al. 2016; Giacomuzzi et al. 2017; Maja et al. 2015) including poplar trees (Clavijo McCormick et al. 2014a; Philippe and Bohlmann 2007). In black poplar, nitrogenous volatiles released upon herbivory have been the focus of attention because they play a major role in attracting natural enemies of herbivores (Clavijo McCormick et al. 2014a). In other plant systems, terpenoids and GLVs are well-known to be involved in the attraction of natural enemies of herbivores (Turlings and Erb 2018). The induction of most of the groups of black poplar volatiles measured has been reported to be associated with JA signaling (Luck et al. 2016; Martin et al. 2003; Semiz et al. 2012). Herbivore-induced increases in protease inhibitor activity have also been connected with elevated jasmonate levels (Haruta et al. 2001; Lomate and Hivrale 2012). These reports are consistent with the JA induction measured in this study where we showed that an assortment of leaf-chewing herbivores all trigger increases in JA.</p><p>Elevated JA levels were found both in herbivore damaged leaves and in adjacent undamaged leaves (Fig. 3, Fig. S2) and the effect in adjacent undamaged leaves was dependent on the identity of the attacking herbivore species (Table 1). The systemic induction of JA in adjacent undamaged leaves after herbivory is a known phenomenon in herbs (Singh et al. 2016), but woody plants such as poplar have not always given consistent results. While herbivory by Lymantria dispar caused JA inductions exclusively in damaged poplar leaves (Clavijo McCormick et al. 2014b), other studies found JA also increased in the adjacent undamaged leaves (Babst et al. 2009; Boeckler et al. 2013). In the present study, Lymantria dispar feeding also led to significantly increased JA levels only in damaged leaves. A trend for higher JA levels in damaged leaves was also visible after feeding by the other generalist caterpillar species, A. mogadorensis. However, feeding by the beetle P. vulgatissima resulted in significantly higher amounts of JA in both the damaged and adjacent undamaged leaves (Fig. 3). We also observed significant systemic induction of salicin in the adjacent undamaged leaves of black poplar and of monoterpenes and sesquiterpenes as has been reported previously for this species (Clavijo McCormick et al. 2014b; Unsicker et al. 2015). In contrast the most prominent compounds induced only in herbivore-damaged leaves were the trypsin protease inhibitors and the nitrogen-containing volatiles (Fig. 4, Fig. 6). In herbaceous plants, herbivory commonly increases protease inhibitor activity significantly in both damaged and adjacent undamaged leaves (Arce et al. 2017; Bozorov et al. 2017; Lomate and Hivrale 2012). This is not true for poplar where induction in adjacent undamaged leaves (Bradshaw et al. 1990) has been reported to be much weaker and delayed compared to the induction in herbivore-damaged leaves (Haruta et al. 2001). Nitrogen-containing volatiles have previously been reported to be emitted only from herbivore-damaged foliage of black poplar and not systemically (Clavijo McCormick et al. 2014a; Unsicker et al. 2015). This may explain their use by herbivore predators and parasitoids as reliable cues to locate prey and hosts (Clavijo McCormick et al. 2014b).</p><p>The volatile bouquets released from black poplar upon herbivore damage differed between the lepidopteran and coleopteran species used in this experiment (Fig. 6, Fig. S3), especially for terpenoids, which were more abundant after coleopteran damage. Similar emission profiles of black poplar have been shown previously (Clavijo McCormick et al. 2014a, 2014b; Unsicker et al. 2015), even though a different volatile collection method was used here. In herbaceous plants, the emission of specific volatile patterns by different herbivore species is known (Cai et al. 2014; Danner et al. 2018; Hare and Sun 2011; Pinto-Zevallos et al. 2018; Turlings et al. 1998) and the pattern of stronger volatile induction after beetle herbivory was also observed (Hare and Sun 2011). At least one other woody plant also showed stronger induction of terpene emission after attack by coleopteran compared to lepidopteran herbivores (Moreira et al. 2013). Several possibilities might be responsible for herbivore species-specific defense responses in plants, including the type of damage and presence of specific elicitors (Ali and Agrawal 2012; Cai et al. 2014; Dicke et al. 2009; Rowen and Kaplan 2016). Specialist herbivores are thought to induce more total volatiles than generalists, although these patterns are not the same for each chemical class (Rowen and Kaplan 2016). One of the herbivore species employed in the present study can be classified as a generalist (Lymantria dispar) and the other three are specialists (Laothoe populi, P. vulgatissima, C. populi). However, the volatile pattern observed after herbivory differed more based on taxonomic grounds between lepidopterans and coleopterans than based on the degree of specialization.</p><p>In summary, our investigation demonstrated that both direct and indirect defenses are induced in black popular by a range of different herbivores. However, the induction of protease inhibitor activity (only in damaged leaves) and salicin (in both damaged and adjacent undamaged leaves) is not specific to the attacking herbivore species. Moreover, the bulk of salicinoids are constitutively present and do not change in concentration with attack. In contrast, the induced volatiles, of which some are known to play a role in indirect defense, do show specific responses to herbivores. The emission pattern from damaged and adjacent undamaged leaves differs between lepidopteran and coleopteran herbivores. Whether this pattern is characteristic of other woody plants requires further investigation.</p><!><p>(DOCX 1.89 MB)</p>

PubMed Open Access

UV-Induced DNA Damage and Mutagenesis in Chromatin\xe2\x80\xa0

UV radiation induces photolesions that distort the DNA double helix and, if not repaired, can cause severe biological consequences, including mutagenesis or cell death. In eukaryotes, both the formation and repair of UV damage occur in the context of chromatin, in which genomic DNA is packaged with histones into nucleosomes and higher-order chromatin structures. Here, we review how chromatin impacts the formation of UV photoproducts in eukaryotic cells. We describe the initial discovery that nucleosomes and other DNA-binding proteins induce characteristic \xe2\x80\x98photofootprints\xe2\x80\x99 during the formation of UV photoproducts. We also describe recent progress in genome-wide methods for mapping UV damage, which echoes early biochemical studies, and highlights the role of nucleosomes and transcription factors in UV damage formation and repair at unprecedented resolution. Finally, we discuss our current understanding of how the distribution and repair of UV-induced DNA damage influence mutagenesis in human skin cancers.

uv-induced_dna_damage_and_mutagenesis_in_chromatin\xe2\x80\xa0

Introduction<!>Chromatin Structure Primer<!>Bulk Nucleosomes (Random DNA Sequences)<!>Designed Nucleosomes (Single DNA Sequences)<!>UV Damage at Transcription Factor Binding Sites<!>UV Damage to DNA in Chromatin (Post-Genomic Era)<!>Genomic Approaches to Monitor UV Damage and Repair in Chromatin<!>Nucleosomes in the Yeast Genome Modulate CPD Formation and Repair<!>Transcription Factor Binding Significantly Alters CPD Formation<!>Chromatin effects on UV-Induced Mutagenesis in Cancer<!>Conclusions

<p>It has long been recognized that the target of DNA damaging agents and the 'landscape' of DNA repair enzymes in eukaryotes is the compact and dynamic structure of chromatin. Understanding variations in the distribution of DNA damage and repair in chromatin, together with their effects on chromatin structure, are crucial features of genomic instability. A prototype environmental agent used extensively in studies examining DNA damage and repair in chromatin has been ultraviolet (UV) radiation. UV radiation is a ubiquitous threat to the genomes of essentially all terrestrial organisms and the etiological agent underlying the development of several hereditary diseases and skin cancers. In this report, we review the historical development of our current understanding of UV-induced DNA damage in chromatin, and highlight recent developments exploring DNA damage and repair in chromatin at the genome-wide level. For brevity, we refer the reader to reviews wherever possible, which contain the seminal references for specific concepts.</p><!><p>In all eukaryotic cells, chromatin fibers are heterogeneous and dynamic. The basic repeating unit of chromatin fibers, the nucleosome, consists of the nucleosome core particle (NCP) and varying lengths of 'linker DNA' (from 20 bp to >100 bp, depending on the organism) stretching between these units (1). NCPs and linker DNA are often complexed with linker histones (H1 and associated variants), which promote compaction of the nucleosome arrays into chromatin fibers. The NCP contains an octamer of the well-conserved core histones H2A, H2B, H3, and H4, and ~147 bp of DNA coiled in ~1.7 left-handed turns around the wedge-shaped histone octamer surface (2). In the octameric complex, histones H3 and H4 form a tetramer that is flanked by two dimers of histones H2A and H2B. In addition, histone variants (primarily of histones H2A and H3) are present in subsets of nucleosomes, and play important roles in specific chromatin functions (3).</p><p>Positioning of histone octamers on DNA is greatly influenced by histone–DNA interactions in the NCP (4). These contacts are mainly electrostatic, where the minor groove of DNA interacts primarily with 'sprocket' arginine residues at regular ~ 10 bp intervals on each side of the pseudo-two-fold dyad axis of symmetry (5, 6). The H3–H4 tetramer strongly interacts with the central ~60 bp of NCP DNA, while the H2A–H2B dimers interact more weakly with the flanking ~30 bp of NCP DNA. The H3 αN helix specifically interacts with the ends of the NCP DNA, and thus plays an important role in regulating NCP DNA unwrapping (see below). This variation in binding strength along the DNA yields intrinsic differences in the accessibility of different DNA regions within NCPs. The reader is referred to a recent review for more detailed discussion of the nucleosome structure (2).</p><p>The stability of NCPs can be regulated in several different ways, including (a) exchange of canonical histones with histone variants (3), (b) ATP-dependent chromatin remodeling complexes (7) and (c) post-translational modifications of the core histones (8). In addition, the sequence of NCP DNA is a strong determinant of nucleosome stability (4). This latter feature of nucleosomes has been exploited for the generation of synthetic DNA sequences with strong positioning strengths (e.g., the Widom 601 positioning sequence) that bind histone octamers with high affinity (9). Use of these DNA sequences has allowed the assembly of near homogenous populations of NCPs that maintain a single rotational and translational setting on the DNA molecule. [Here, 'rotational setting' refers to orientations of DNA relative to the histone surface (e.g., 'in' rotational orientations are regions of the DNA phosphate backbone oriented toward the histone octamer and 'out' are regions oriented toward the solvent) and 'translational setting' refers to the DNA sequence positions relative to the center of symmetry (or dyad axis) of NCPs.]</p><p>Natural DNA sequences exhibit a wide range in nucleosome positioning power, where histone octamer binding affinity for different DNA sequences can vary by as much as 5000-fold (10). These differences dramatically affect the accessibility of occluded DNA sites in NCPs. Moreover, both the rotational and translational settings of DNA sites within NCPs significantly influence the accessibility of these sites. Restriction enzyme cleavage activity and Förster resonance energy transfer (FRET) have shown that regions of NCP DNA transiently unwrap from the histone surface (11). These studies revealed there is a progressive decrease in site exposure from translational settings near the ends of the DNA toward the 'dyad center' of NCPs (12). This dynamic equilibrium of unwrapping and rewrapping is dependent on DNA sequence. Indeed, stopped-flow FRET and fluorescence correlation spectroscopy studies have shown that the DNA unwrapping time varies by almost 10,000-fold going from the DNA ends to the dyad center (13). Thus, at the primary level of chromatin structure, are multiple, interdependent mechanisms capable of changing accessibility to DNA through the regulation of nucleosome positioning and stability.</p><!><p>The location of DNA damage sites in chromatin plays an important role in their accessibility to repair enzymes. The distributions of a number of different DNA lesions in bulk (or 'mixed sequence') chromatin have been reported over the past four decades [e.g., see (14, 15)]. These studies helped form the basis for our understanding of the 'efficiency' of DNA excision repair [both nucleotide excision repair (NER) and base excision repair (BER)] in chromatin (16, 17). Early on it was clear that some classes of DNA lesions (e.g., bulky chemicals) form preferentially in nucleosome linker DNA (14). Interestingly, the major UV photoproduct in DNA [cis-syn-cyclobutane pyrimidine dimer (or CPD)] forms almost randomly between linker and core regions of chromatin (on a unit DNA basis). In contrast, the second most prevalent UV photoproduct, pyrimidine (6-4) pyrimidone dimer [or (6-4)PD], has a much stronger bias for linker DNA (14, 18).</p><p>The distribution of UV photoproducts within NCPs has also been mapped (18, 15). Using a T4 polymerase-exonuclease blockage assay, the distributions of CPDs and (6-4)PD were mapped at nucleotide resolution in NCPs from UV-irradiated cells, isolated chromatin, and isolated NCPs. These studies showed that the level of CPDs among NCPs in bulk chromatin has a striking periodic pattern, with an average periodicity of 10.3 ± 0.1 bases (Figure 1). This 'photofootprint' reflects the rotational setting of DNA on the histone surface, where the highest levels of CPD formation occur where the DNA phosphate backbone is farthest from the histone surface (i.e., an 'out' rotational setting) (19). The peak separation in the photofootprint pattern varies by as much as 1.3 bases near the nucleosome dyad relative to the average periodicity (20). In contrast, (6-4)PDs form much more randomly within nucleosome cores (Figure 1a; (21)). Thus, formation of UV photoproducts in chromatin is very sensitive to DNA-histone interactions on the octamer surface in nucleosomes.</p><p>The UV photofootprint results from the bending of DNA around histone octamers, reflecting the structural constraints imposed on DNA in nucleosomes [Figure 2, and (22)]. Moreover, the first crystal structure of a CPD-containing NCP was recently reported and indicates that CPDs do not significantly alter the DNA structure at the damage site (23). Thus, the variation in mobility of DNA on the histone surface may explain the UV photofootprint (Figure 2). After absorption of a UV photon, the [2+2] cycloaddition that occurs during the excited state of the 5–6 double bond in a pyrimidine base (24) will be more probable when adjacent pyrimidines are frequently aligned. In NCPs, the DNA mobility is highly variable, being minimal near histone–DNA contacts, where the minor groove faces toward the histone octamer, and maximal for the bases with phosphate groups on the outside of the particle (Figure 2). The increased mobility of dipyrimidines with their minor groove facing away from the histone octamer should make these regions the most favorable sites for CPD formation in NCPs.</p><p>Finally, although (6-4)PDs are less prevalent than CPDs in chromatin, the yield of these photoproducts at specific sites can be much higher than in free DNA (18). This can alter their impact on UV-induced mutagenesis at specific sites in mammalian cell chromatin (25). As these UV photoproducts are distributed more randomly than CPDs within NCPs (Figure 1a) and have a much stronger bias for linker regions (18), the distribution of (6-4)PDs in chromatin differs markedly from CPDs, possibly reflecting the different structures of these photoproducts in NCPs. Indeed, the recently reported crystal structure of a (6-4)PD-containing NCP indicates that the damaged region of NCP DNA is flexibly disordered, particularly in the solvent exposed strand (26). Thus, it is reasonable to assume that the more constrained NCP DNA is less capable of forming such structures, as compared to the linker DNA in chromatin. This feature of the (6-4)PD distribution undoubtedly plays a role in the rapid NER of these UV lesions relative to CPDs (27).</p><!><p>Reconstitution of defined DNA sequences into nucleosomes (or 'designed nucleosomes') allows analysis of DNA damage effects on preexisting nucleosome structures and on nucleosome formation during assembly. Most notably, designed nucleosomes allow detailed studies on repair of DNA damage in specific nucleosome locations. However, results for defined DNA sequences depend strongly on (a) sequence specificity of damage formation, (b) location of target sequences relative to the histone octamer, (c) local flexibility of DNA, and (d) stability of the designed nucleosome structure.</p><p>Several nucleosome-positioning sequences have been used to create designed nucleosomes, and many of these differ in nucleosome 'positioning power' (9). A favorite sequence used in early studies was the sea urchin 5S rRNA gene (28), a sequence with moderate positioning power (several-fold) over the bulk of the DNA sequences in chromatin (Table 1). On the other hand, the Widom 601 clone, having a 'free energy of reconstitution' (ΔΔG) of ~ 2.9 kcal/mol lower than the 5S rDNA sequence, has an ~100-fold higher positioning power than 5S rDNA (Table 1). These differences in positioning power can significantly affect the 'occupancy time' of DNA in preferred rotational and translational settings of designed NCPs (Figure 3) (29, 30). Thus, when comparing levels of DNA damage sites and their accessibility to repair enzymes in designed nucleosomes with different sequences, variation can arise from the differences in positioning power of these sequences.</p><p>Reports on the impact of nucleosome positioning on the distribution of UV-induced photoproducts in defined sequences started appearing in the 1990's (14, 18). UV irradiation of a yeast nucleosome positioning sequence (called HISAT) yielded a strong modulation of CPD formation in NCP DNA (31). This sequence contains a 40 bp polypyrimidine stretch, including several long T-tracts, allowing measurement of CPD formation over three helical turns of DNA in a defined nucleosome. As with mixed-sequence nucleosomes, the pattern of CPD formation in HISAT nucleosomes is significantly different than in free HISAT DNA. Interestingly, the distribution of CPDs in HISAT nucleosomes only partly resembled that of random-sequence nucleosomes. Specifically, CPD levels were high at two contact points between the minor grooves of the HISAT sequence and the histone surface, a region where the random-sequence DNA predicts low levels of CPDs (Figure 1). This most likely reflects an altered T-tract structure in nucleosomes, as compared to random DNA, as these tracts have a non-B-form structure in solution (32). We note that charge neutralization of the DNA phosphates occurs essentially on one side of the helix in nucleosomes (i.e. facing the histones), which produces DNA bending around the histone octamer and formation of kinks at regular intervals in the DNA (33). Thus, the natural bending of certain DNA sequences will impact the formation of nucleosomes, being favorable or unfavorable depending on the location of these bends. Therefore, the consequence of DNA damage-induced bending on nucleosome reconstitution was studied using the 5S rDNA nucleosome (34). The results showed that damage induced by a bulky chemical carcinogen (benzo[a]pyrene diol epoxide) enhanced nucleosome formation on 5S rDNA, while UV-induced photoproducts repressed nucleosome formation (Table 1; see ΔΔG∘ values for UV and BPDE damaged 5S rDNA).</p><p>The effects of nucleosome folding on formation of UV photoproducts were also analyzed in the 5S rDNA nucleosome (35). Surprisingly, it was found that even after a fairly high amount of UV damage (~ 0.8 CPDs/NCP on average), there was no affect on either the average rotational or translational settings of the 5S rDNA. Additionally, it was found that, although nucleosome folding had little effect on CPD formation in the transcribed strand of rDNA, formation of these photoproducts was significantly repressed in the non-transcribed strand in an ~ 30 bp region around the nucleosome dyad center. These results presumably reflect constrained rotation and bending of the DNA bases, and the tight binding of the H3-H4 tetramer, near the center of NCPs (35). These authors also investigated the affect of nucleosome formation on CPD yields in a 14 base pyrimidine tract in the 5S rDNA, as it spans a complete turn of the DNA helix (35). It was found that CPD formation in the pyrimidine tract was either elevated or reduced at sites where the minor groove faced 'out' or 'in' toward the histone surface, respectively, agreeing with the CPD modulation predicted by random sequence NCPs (Figure 1).</p><p>In the late 1990's, Kosmoski and Smerdon exploited the conformational stability of tightly positioned nucleosomes to allow incorporation of a single UV photoproduct at a specific site in NCPs (36). Using a synthetic nucleosome positioning sequence (37) to bracket a short (30 bp) DNA sequence containing a chemically synthesized CPD, these authors assembled nucleosomes specifically damaged at only one site and one structural orientation, with the DNA minor groove at the CPD site facing out from the histone surface ~5 bases from the nucleosome dyad center. Competitive gel-shift analysis showed there was only a small increase in histone binding energy required to form a NCP (ΔΔG = ~0.15 kcal/mol), which allowed complete nucleosome loading on the damaged DNA (36). Thus, it was now possible to 'design' specifically damaged nucleosomes for in vitro studies on how nucleosome structure influences DNA damage formation, and how DNA repair enzymes detect and repair damage that is occluded by the histone octamer. For example, it was found that NER in Xenopus oocyte extracts can effectively repair the single UV lesion in this designed nucleosome, and nucleosome reassembly occurs on the newly repaired DNA shortly after NER in these extracts (38). Many reports followed these studies using designed nucleosomes to monitor everything from the effect of rotational orientation of DNA damage to the effect of specific histone modifications on the activities of DNA repair enzymes in vitro. Although these studies are beyond the scope of the present review, the reader is referred to recent reviews that cover these topics [e.g., (16, 15, 17)].</p><p>Finally, as discussed in the Chromatin Structure Primer section, nucleosomes are dynamic and exist in a conformational equilibrium, where a portion of the NCP DNA spontaneously unwraps and rewraps around the histone octamer (Figure 4a). Förster Resonance Energy Transfer (FRET) and restriction enzyme accessibility have been used to examine how UV-induced photoproducts in nucleosome DNA affect this conformational equilibrium (39). The 147-bp Widom 601 nucleosome positioning sequence was pre-labeled with 'donor' and 'accepter' fluorophores on opposite DNA strands and used for NCP reconstitution with purified Xenopus histones (Figure 4b). This system is a sensitive detector of changes in the distance between the fluorescent dyes (Figure 4c). It was shown that UV photoproducts in the 601 sequence decreased FRET efficiency (Figure 4d) and increased restriction enzyme accessibility in the 601 NCP, indicating that UV photoproducts promote DNA unwrapping from the histone octamer (39). This increased DNA unwrapping was even detected when a UV photoproduct was inserted at a single site of the NCP. Thus, increased exposure of nucleosome DNA following DNA damage formation may play an important role in the accessibility of DNA repair enzymes in chromatin. Indeed, it was observed that rapid repair of UV-induced lesions by UV photolyase, a repair enzyme whose activity is quantitatively blocked by nucleosomes, occurs in nucleosome-loaded DNA of intact yeast cells (40). These rapid repair kinetics are consistent with the rapid kinetics of nucleosome unwrapping-rewrapping in the cell, rather than the much slower kinetics expected for histone exchange, chromatin remodeling or modification of the core histones (41).</p><!><p>The first report on protein-DNA interactions modulating UV photoproduct formation was on the lac repressor complex of Escherichia coli lac operator DNA (42). These authors used a series of chemical reactions to cleave DNA at UV photoproducts, and it was found that both increased and decreased photoproduct formation occurs at specific sites of the lac repressor binding sequence. This method was later employed to demonstrate transcription-dependent changes in the control region of the GAL1 and GAL10 genes in yeast cells (43).</p><p>These studies were followed by the use of CPD-specific endonuclease 'T4 endo V' (44) to detect CPD formation profiles at the single nucleotide level in protein-DNA complexes. Indeed, a number of high-resolution methods were even developed to measure UV damage at the single nucleotide level in intact cells, as the sensitivity of this assay was greatly enhanced by amplification of T4 endo V cleavage fragments in DNA and radioactive labeling (45). One such method involved amplification through ligation-mediated polymerase chain reaction, or LMPCR (46). This technique was used to show that modulation of UV photoproducts occurs in promoter regions during induction of several genes in intact human cells, including JUN, FOS, and PCNA (47). Another method used to detect DNA damage directly (i.e., not involving amplification of fragments created by strand cleavage) at the single nucleotide level excluded some potential problems with amplification methods (45). With this method, genomic DNA is digested with one or more restriction enzymes to release fragments cleaved at DNA damage sites and annealed to sequence-specific biotinylated oligonucleotides (Figure 5a). The annealed fragments can then be separated from the mix with streptavidin magnetic beads, and end-labeled by extension of the 3′ overhang of the fragment consisting of a poly(dT) tract (Figure 5a). Although less sensitive than the amplification methods, this method provides a linear representation of the actual number of DNA damage sites and has been very useful for monitoring UV damage and repair in organisms with small genomes, such as bacteria and yeast (Figure 5b). Thus, the modulation of UV photoproducts in transcription factor (TF) binding sites and positioned nucleosomes may be a general phenomenon, a premise that was recently confirmed by genome-wide mapping of UV photoproducts in intact cells (see the section on UV Damage to DNA in Chromatin (Post-Genomic Era)).</p><p>A useful model system for studying UV photoproduct formation in protein-DNA complexes in vitro has been the 5S rRNA gene sequence bound to transcription factor IIIA [TFIIIA; (48, 49)]. Binding of TFIIIA to 5S rDNA is the first step in RNA polymerase III-directed synthesis of 5S rRNA (50). This ~40-kDa protein contains nine zinc finger motifs, and binds tightly to the internal control region (ICR) of 5S rDNA, an ~50 bp sequence located inside the transcription unit (e.g., see 49–51). The dependence of the yield of UV-induced photoproducts on TFIIIA binding has been studied in the Xenopus borealis 5S rRNA gene (49, 51). TFIIIA binding was found to modulate UV photoproduct formation mainly in the template strand of 5S rDNA, where the strongest contacts with this protein are observed (52, 53). The modulation pattern was not uniform, however, and strong inhibition of CPD formation was observed at several sites in the ICR [reviewed in (15)]. Interestingly, CPD formation was also enhanced at one site in the template strand following TFIIIA binding. This region of the ICR with enhanced CPD formation binds the middle three zinc fingers of TFIIIA, and this binding differs from that of the remaining zinc fingers in TFIIIA (52, 53). The zinc finger 'cassettes' on the N- and the C-terminal of TFIIIA wrap along the major groove of 5S rDNA, while the three middle fingers bind nearly parallel to the helix axis in the center of the complex, and these three zinc fingers of TFIIIA may cause bending in the 5S rDNA that enhances CPD formation (51).</p><p>Finally, in addition to the modulation of UV damage in DNA by bound TFs, UV lesions themselves can alter TF binding. Tommasi et al (54) incorporated site-specific CTDs [cis-syn-cyclobutane thymine dimers] into oligonucleotides containing the recognition sequences of five different TFs (E2F, NF-Y, AP-1, NFκB, and p53). In each case, the presence of a single CTD strongly inhibited binding (11- to 60-fold) of the respective TF in vitro. More recently, Kwon and Smerdon studied the binding of TFIIIA to the ICR in 5S rDNA containing either site-specific CTDs (55) or CPDs throughout the ICR in randomly damaged 5S rDNA (51). These authors also examined the relationship between TFIIIA binding and the efficiency of DNA repair in Xenopus oocyte nuclear extracts. These studies demonstrated that (a) CTDs in the ICR of the 5 S rRNA gene can increase or decrease binding affinity of DNA for TFIIIA depending on their position, (b) modulation of TFIIIA binding correlates with the modulation in dissociation rate (koff) of the complex, (c) the decrease in binding affinity is accompanied by the loss of TFIIIA-DNA contacts near CTD sites, and (d) NER of CTDs in the TFIIIA-5S rDNA complex is extremely sensitive to changes in the dissociation rates of the complex. Thus, this simple model system has provided a detailed view of how strong protein-DNA interactions can have major effects on the yield of UV photoproducts at specific sites in the binding domain and how TF binding can be modulated by the presence of UV photoproducts.</p><!><p>Although chromatin is uniformly comprised of nucleosomes, recent studies have highlighted the diversity of distinct types of chromatin in eukaryotic chromosomes, distinguished by differences in covalent histone modifications, histone variants, and the degree of ongoing histone exchange, nucleosome mobility, and higher-order chromatin compaction (3, 56). While much has been learned from studies of defined sequences or in bulk chromatin, it is important to elucidate how this diverse spectrum of chromatin states impact UV damage formation and subsequent repair in eukaryotic genomes. To this end, a number of different genomic approaches have been developed to map UV-induced lesions across the yeast and human genomes (57).</p><!><p>Initial genomic studies adapted the chromatin immunoprecipitation-microarray (ChIP-chip) method (58, 59), using anti-CPD antibodies to immunoprecipitate lesion-containing DNA fragments, which were then detected using tiling microarrays (60–62, 57). These studies yielded insights into how DNA sequence influences UV damage formation across the genome. For example, our labs showed that UV damage formation is enriched adjacent to specific classes of repeat elements in the human genome, particularly those containing poly-T tracts (61). More recently, a microarray-based method was used to investigate how global genomic repair of CPD lesions is organized across the yeast genome by a sequence-specific DNA binding factor (i.e., Abf1) and regulated by histone acetylation (63). However, microarray-based methods can only map DNA lesions at low resolution, due to both the relatively low density of microarray probes along chromosomes and the large size of the sonicated DNA fragments [e.g., ~300–400 bp (60, 61)]. This limits their utility for characterizing how chromatin, at least at the level of individual nucleosomes, modulates UV damage formation and repair.</p><p>More recently, next generation sequencing (NGS) methods have been employed to map UV damage at higher resolution across the genome. The first of these methods was called Excision-seq (64). Excision-seq utilized DNA repair enzymes (in this case UV DNA Endonuclease [UVDE]) to generate single strand breaks (SSBs) immediately upstream of either CPDs or 6-4PDs. However, the application of Excision-seq is constrained to sequences where the SSBs are converted to DNA double strand breaks (DSBs), such as two adjacent SSBs occurring on opposing DNA strands. For this reason, very high UV doses (~10,000 J/m2) were required, in order to increase the probability that cleavable photoproducts would form in close proximity on opposing strands. Excision-seq was used to map UV damage formation across the yeast genome at single nucleotide resolution, thus revealing the expected DNA sequence preferences for CPD and 6-4PP formation (64). However, the influence of chromatin on UV damage formation was not explored in this study. A second method, known as excision repair sequencing (XR-seq), sequenced the ~25–30 nucleotide fragments that are excised during NER of UV photoproducts (65). This has proven to be a powerful method to map the repair of UV lesions across the genome (65, 66), but does not provide direct information about initial UV damage formation.</p><p>Recently, Mao et al. reported a new method named CPD-seq to map both initial CPD damage formation and subsequent NER activity in the yeast genome at single nucleotide resolution (67). CPD-seq was adapted from a method used to map ribonucleotide incorporation during DNA replication (68, 69). In CPD-seq, UV-irradiated DNA is sonicated into small DNA fragments, ligated to a double stranded DNA adapter (trP1, colored brown in Figure 6A), and treated with terminal transferase [and dideoxy-ATP (ddATP)], resulting in DNA fragments lacking free 3′-OH groups (Figure 6A). The DNA is then sequentially digested with T4 endo V and AP-endonuclease (APE1) to generate new 3′-OH groups immediately upstream of the CPD lesion (Figure 6A). These fragments are then ligated to a second adaptor DNA (A adaptor, colored orange in Figure 6A), with a biotin label on one strand to allow purification of the ligated fragments. The generated CPD-seq library is briefly amplified with primers complementary to trP1 and A adaptors and subjected for next generation DNA sequencing. Because the only 3′-OHs available for A adaptor ligation are derived from CPD cleavage by T4 endo V and APE1, the genomic location of CPD lesions is precisely determined from the corresponding DNA sequencing read, which initiates within the A adaptor (Figure 6A). Hence, using the CPD-seq strategy, one can map initial CPD formation across the genome at single nucleotide resolution. CPD maps can also be generated at different repair times to investigate the genome-wide time course of CPD removal (67).</p><!><p>The high resolution of CPD-seq data provided a unique opportunity to examine how chromatin influences CPD formation across the genome. Overlaying the CPD-seq data onto a well-defined map of yeast nucleosome positions (70) revealed that yeast nucleosomes induce a strong UV photofootprint. The peaks of CPD formation (after normalizing for dipyrimidine content) coincided with outward rotational settings in the nucleosome, exhibiting a striking periodicity of ~10 bp (67), closely mirroring the UV photofootprint previously identified in mammalian chromatin [Figures 1 and 6B (19)]. Notably, the UV photofootprint was most apparent among strongly positioned nucleosomes in yeast (~10,000 nucleosomes), but was barely detectable among weakly positioned nucleosomes (~7500 nucleosomes) (67). Presumably, the lack of a uniform rotational setting among weakly positioned nucleosomes masks the UV photofooprint at these locations.</p><p>Nucleosomal DNA shows clear sequence biases: A-T rich sequences tend to adopt 'in' rotational settings, while G-C rich sequences tend to adopt 'out' rotational settings (4). For this reason, TT dinucleotides, which are most prone to forming CPD lesions, tend to be positioned at 'in' rotational settings [e.g., (70, 67)]. The variations of TT frequency in nucleosomal DNA are mirrored by CPD levels in UV-irradiated naked DNA, indicating that in the absence of nucleosomes, the highest CPD yields occur at locations corresponding to 'in' rotational settings in NCPs, and the lowest yields at locations corresponding to 'out' rotational settings (67). However, the exact opposite pattern occurs in vivo when DNA is packaged into nucleosomes. The impact of nucleosomes on UV damage formation is highlighted by directly comparing CPD-seq data of UV damage formation in vivo, in intact cells, with CPD-seq data of UV damage formation in vitro, in naked DNA (Figure 6B) (67). The packaging of DNA into nucleosomes suppresses CPD formation at 'in' rotational settings relative to naked DNA; however, CPD formation in nucleosomes is markedly higher at 'out' rotational settings (Figure 6B). By placing TT-rich DNA sequences at 'in' rotational settings, nucleosomes essentially protect (or 'shade') these intrinsically vulnerable sequences from UV damage. We hypothesize that this mechanism operates in all eukaryotes, and may be an important modifier of UV-induced mutagenesis (see below).</p><p>Although the rotational setting of DNA in nucleosomes modulates CPD formation, analysis of CPD removal indicates that it does not generally impact subsequent NER activity in yeast. Analysis of CPD-seq data indicates that higher CPD levels are retained at 'out' rotational settings during repair, even though overall CPD levels are decreased, regardless of the rotational positioning (67). Interestingly, the translational positioning significantly regulates CPD removal in strongly positioned nucleosomes. CPDs located in the distal regions of nucleosomal DNA, where spontaneous unwrapping occurs more rapidly (see the Chromatin Structure Primer section above), are repaired more efficiently than CPDs located near the nucleosome dyad (67). It is not known to what extent altering nucleosomes by incorporating histone variants or histone post-translational modifications will affect initial UV damage formation and repair on a genome-wide scale, but previous studies (e.g., (71–74)) suggest this is an important avenue for future research.</p><!><p>The high resolution of the CPD-seq data also permitted investigating how TF binding to DNA affects UV damage formation. It was previously reported that TF binding to DNA can modulate CPD formation (see the section on UV Damage at Transcription Factor Binding Sites), suppressing CPD formation at some locations within the binding site, and enhancing CPD formation at other locations [reviewed in (47)]. Genome-wide analysis of CPD formation at the binding sites of yeast TFs Abf1 and Reb1 demonstrated that these DNA-bound TFs generally inhibited CPD formation in vivo, although a single CPD 'hotspot' was detected at a specific location in Abf1 binding sites (67). TF binding appears to be critical for modulating CPD formation, as Reb1 binding sites with low Reb1 occupancy in vivo had relatively little affect on CPD formation, while high occupancy Reb1 binding sites had a striking affect on CPD formation (67). CPD formation was as much as 3-fold less frequent at Abf1 or Reb1 binding sites in vivo relative to unbound naked DNA, indicating that the magnitude (or fold-change) of the TF photofootprint is significantly greater than that of the nucleosome photofootprint. This difference in the magnitude of UV photofootprinting may be due in part to the higher intrinsic level of DNA dynamics within nucleosomes in vivo (i.e., due to DNA unwrapping, etc.) relative to TF-DNA complexes.</p><p>It will be important to investigate how different families of TFs affect UV damage formation. For example, it is possible that other families of TFs may primarily enhance CPD (or 6-4PD) formation, as has been previously reported for the TFs that bind the CCAAT-box and serum response elements in mammalian promoters (47), which could be an important contributor to mutagenesis in human cancers (see the section on Chromatin Effects on UV-induced Mutagenesis in Cancer below). Moreover, CPD-seq data could be exploited as an experimental tool to identify unknown TF binding sites across the genome based on their UV photofootprint signature.</p><!><p>The role of UV light in inducing mutations that underlie skin cancers has long been understood (75). However, the ability to relate the genomic profile of DNA damage and lesion repair distributions to mutagenesis in human cells, model organisms, and clinical tumor samples has the potential to expand our knowledge of the important factors that govern where disease causing mutations are prone to occur (57). In the case of UV mutagenesis, the analysis of whole genome sequenced human melanomas has taken a primary focus. One of the first whole genome sequences of a human cancer was that of a melanoma cell line, where the mechanistic knowledge of Transcription Coupled Nucleotide Excision Repair (TC-NER) removal of UV lesions from the transcribed DNA strand was utilized as a means to support the observed strand bias of mutations occurring specifically within genes (76). Since then, the large number of UV-induced mutations in melanoma genomes has made this cancer type particularly well-suited for assessing how mutations are distributed across chromosomes.</p><p>Initial analyses of melanoma genomes, as well as genomes from several other cancer types, determined that mutation densities, like lesion densities (see the UV Damage to DNA in Chromatin (Post-Genomic Era) section), are highly heterogeneous, changing drastically with chromosome position (77). A comparison of the distribution of melanoma mutations to various aspects of chromosome structure detailed by the Encyclopedia of DNA Elements (ENCODE) project indicated that in addition to the expected constraints for UV-induced mutations to TC and CC dinucleotides (25), chromatin likely also impacts where these mutations occur. Specifically, melanoma mutations were found to be elevated in regions with higher nucleosome occupancy as well as in late replicating regions of the genome (78, 79). The correlation of mutation density with both of these chromosome features suggests that an impaired ability for NER to remove UV lesions within more compact chromatin prior to DNA replication is one key factor dictating the location of UV-induced mutations. The impacts of nucleosome structure on restricting NER activity have been well established both biochemically and genetically (15–17), while the realization that late replicating genome regions are generally gene poor and heterochromatic (80) suggests that neither TC-NER nor global-NER would function efficiently in these regions. Further support for chromatin accessibility impacting the location of UV-induced melanoma mutations have steadily accumulated as regions of high mutation density have been shown to be associated with histone modifications that traditionally indicate heterochromatic chromosome structure (81), while decreased mutation densities have been found within open chromatin domains marked by high DNase I accessibility (82, 83). This decreased abundance of UV-induced mutations in DNase I hypersensitive sites compared to heterochromatic regions appears to be a function of NER activity as tumors containing mutations in key NER components (83) or from XPC-deficient individuals (84) lack this phenomenon. Sancar and colleagues' development of XR-seq (see the UV Damage to DNA in Chromatin (Post-Genomic Era) section) (65), which enables monitoring of NER activity across the genome, has provided direct evidence that globally, melanoma mutations are enriched in regions where NER occurs more slowly (66). More fine scale analyses of the distribution of NER and UV-induced mutation in active promoters indicate that DNA-binding proteins other than nucleosomes likely influence mutation positions during melanoma development (Figure 7). Specifically, TF binding sites were shown to be prone to both decreased NER activity (as measured by XR-seq) and increased mutation density (85, 86), consistent with prior biochemical work indicating that TFs can occlude lesions from NER (see the section on UV Damage at Transcription Factor Binding Sites above). However, these same analyses indicated the presence of regions adjacent to promoters that contain low relative levels of both NER activity and mutation density (Figure 7). Thus, the higher mutation densities in these regions cannot solely be explained by a difference in NER efficiency and suggests other factors, such as differences in initial UV damage levels, likely contribute to mutational heterogeneity in human tumors.</p><p>In addition to impacting NER efficiency, chromatin could also impact melanoma mutation distributions by affecting where lesions are most likely to form. As discussed earlier, a large amount of biochemical data indicates that the structure of the nucleosome core particle, as well as TF binding to promoters, results in photofootprints after UV exposure. Despite this knowledge, most current genomic analyses of NER and mutation distributions have assumed that UV lesions form relatively evenly across chromosomes, and little information currently exists relating the formation of UV lesions to chromatin features on a genomic scale (66). The recent development of the CPD-seq method to map UV-lesions at single nucleotide resolution (see the UV Damage to DNA in Chromatin (Post-Genomic Era) section above), indicates that at least in yeast, photofootprints of the nucleosome core particle, as well as Abf1 and a Reb1 TFs, can be detected using genomic methods (67). As UV-lesion formation is a fundamental pre-requisite for both NER activity and eventual UV-induced mutations, UV photofootprints likely have a significant effect on the distribution of mutations observed in clinical melanoma samples and may help explain why certain genome regions contain both low levels of mutation and NER activity.</p><!><p>Efforts during the past four decades have led to significant progress in understanding UV damage formation and NER in the context of chromatin. Studies conducted in mammalian cells demonstrated the strong modulation of UV damage formation by DNA packaging into chromatin (19). Formation of UV damage in nucleosomes, on the other hand, alters the nucleosome unwrapping dynamics (40), which may serve as an important mechanism to recruit damage recognition factors for repair. Recent progress in mapping UV damage across the genome not only confirmed the early observations made in bulk chromatin, but also provides a unique opportunity to closely investigate the heterogeneous distribution of UV damage and variable repair dynamics in different eukaryotic chromatin domains. Considering the largely heterogeneous feature of mutation density in melanoma genomes, data generated from recent genome-wide damage maps provide new insights into mechanisms underlying variable mutation formation in the cancer genome (i.e., 82, 83). Similar mapping methods will likely be developed to map other types of DNA damage and their corresponding DNA repair pathways, which undoubtedly will strengthen our understanding of the important mechanisms that connect chromatin structure, DNA damage formation and DNA repair with mutagenesis in human cancers.</p>

PubMed Author Manuscript

Thermally and Magnetically Robust Triplet Ground State Diradical.

High spin (S = 1) organic diradicals may offer enhanced properties with respect to several emerging technologies, but typically exhibit low singlet triplet energy gaps and possess limited thermal stability. We report triplet ground state diradical 2 with a large singlet-triplet energy gap, \xce\x94EST \xe2\x89\xa5 1.7 kcal mol\xe2\x88\x921, leading to nearly exclusive population of triplet ground state at room temperature, and good thermal stability with onset of decomposition at ~160 \xc2\xb0C under inert atmosphere. Magnetic properties of 2 and the previously prepared diradical 1 are characterized by SQUID magnetometry of polycrystalline powders, in polystyrene glass, and in other matrices. Polycrystalline diradical 2 forms a novel one-dimensional (1D) spin-1 (S = 1) chain of organic radicals with intrachain antiferromagnetic coupling of J\xe2\x80\xb2/k = \xe2\x88\x9214 K, which is associated with the N\xc2\xb7\xc2\xb7\xc2\xb7N and N\xc2\xb7\xc2\xb7\xc2\xb7O intermolecular contacts. The intrachain antiferromagnetic coupling in 2 is by far strongest among all studied 1D S = 1 chains of organic radicals, which also makes 1D S = 1 chains of 2 most isotropic, and therefore an excellent system for studies of low-dimensional magnetism. In polystyrene glass and in frozen benzene or dibutyl phthalate solution, both 1 and 2 are monomeric. Diradical 2 is thermally robust and is evaporated under ultra-high vacuum to form thin films of intact diradicals on silicon substrate, as demonstrated by X-ray photoelectron spectroscopy. Based on C-K NEXAFS spectra and AFM images of the ~1.5-nm thick films, the diradical molecules form islands on the substrate with molecules stacked approximately along the crystallographic a-axis. The films are stable under ultra-high vacuum for at least 60 h but show signs of decomposition when exposed to ambient conditions for 7 h.

thermally_and_magnetically_robust_triplet_ground_state_diradical.

INTRODUCTION<!>Synthesis of 2.<!>X-ray crystallography.<!>EPR spectroscopy.<!>SQUID magnetometry: triplet ground states of 1 and 2.<!>SQUID magnetometry: 1D antiferromagnetic S = 1 chain for polycrystalline diradical 2.<!>Stability.<!>Thin films of 2 on SiO2/Si(111) substrate.<!>CONCLUSION<!>EXPERIMENTAL SECTION<!>X-ray crystallography.<!>Synthesis of 2.<!>Blatter radical 4.<!>Diradical 2.<!>Computational details.<!>Thin film growth and XPS measurements.

<p>Open-shell organic molecules with high-spin ground states and large energy gaps between the high-spin ground state and low-spin excited states possess unique, intriguing characteristics that are not only of fundamental interest but also have significant potential for numerous advanced technological applications. Notably, these molecules have long been considered the holy grail of purely organic magnets,1–5 and recently have emerged as promising building blocks for organic spintronics,6 spin filters,7 sensors,8 memory devices,9 and probing quantum interference effects in molecular conductance.10 Although the design principle has been well laid out,11,12 such high-spin molecules with strong ferromagnetic interactions between unpaired electrons and persistence at room temperature remain highly uncommon,13–16 and triplet ground state diradicals with robust thermal stabilities are especially rare.17–19 The advancement in the design and synthesis of these exotic molecules will be crucial to the development of advanced organic magnetic materials and devices.</p><p>Recently, we reported triplet ground state diradicals with robust thermal stabilities, such as tetraazacyclophane diradical dication (TDD) and diradical 1 (Figure 1).17,18 To our knowledge, these are the only two such diradicals that are well characterized with respect to stability by thermogravimetric analysis (TGA). Both TDD and 1 start decomposing at approximately 180 °C under nitrogen atmosphere and neutral diradical 1 could be sublimed under high vacuum at 140 °C without decomposition.</p><p>While these high-spin diradicals possess remarkable thermal properties, they are not fully populated in the high-spin state at room temperature. The singlet triplet energy gap, ΔEST, of both diradicals is only about 0.5 kcal mol−1, as determined by quantitative EPR spectroscopy and SQUID magnetometry, which is similar to thermal energy at room temperature (RT ≈ 0.6 kcal mol−1).17,18 In such a case, there is a significant depopulation of the triplet ground state at room temperature and above, and thus the unique magnetic properties of the high-spin state with extra-large magnetic moment are compromised at ambient temperatures.</p><p>To rectify this deficiency, we designed diradical 2 and preliminary computed its magnetic properties. The DFT calculations suggested an increased ΔEST by a factor of ~2.5, compared to 1, and an estimated ~95+% occupancy of the triplet ground state at room temperature.17 We anticipated diradical 2 to possess both superior magnetic and excellent thermal properties, providing a novel high-spin diradical with robust high temperature stability and near-full occupation of the triplet ground state at ambient temperatures.</p><p>The potential of organic radicals in the development of organic electronics hinges upon their processability. The capacity of a molecule to form contacts or to evaporate onto a substrate without degradation is a critical prerequisite for device fabrication. In this regard, it is important to test the robustness of diradical 2 toward evaporation. Controlled evaporation of a diradical is considered very challenging, and to our knowledge, it has not been reported in literature. Achieving the first thin film of a high-spin diradical with nearly full-occupation of the triplet ground state at ambient temperature would be a significant step forward, providing exciting new avenues for the development of organic electronics.</p><p>Here we report the synthesis and study of high spin diradical 2 (Figure 1). As predicted, 2 has a large ΔEST of ≥1.7 kcal mol−1, much larger than the thermal energy at room temperature, thus possessing a triplet ground state that is nearly exclusively populated (98+%) at room temperature. For comparison, we characterize both 1 and 2 by SQUID magnetometry in dilute matrices and as polycrystalline powders. Notably, polycrystalline diradical 2 forms a novel one-dimensional (1D) spin-1 (S = 1) chain consisting of close contacts between the heteroatoms (oxygens and nitrogens) of radical moieties with the largest spin densities. The 1D chain is distinctly different from those previously reported,18,20 and it is the first observed 1D system consisting of nitronyl nitroxide and Blatter radicals. Importantly, the observed intra-chain exchange coupling constant of J′/k = −14 K is much larger than the previously studied 1D S = 1 chains, with the next strongest J′/k = –5.4 K found in TDD.18Diradical 2 is thermally robust, with an onset of decomposition at ~160 °C under inert atmosphere and is thermally evaporated under ultra-high vacuum to form thin films on SiO2/Si(111) wafers, with X-ray photoelectron spectroscopy suggesting the presence of intact 2. The C-K NEXAFS spectra and AFM images of the films indicates the diradical molecules form islands on the substrate with molecules stacked approximately along the crystallographic a-axis. Diradical 2 possesses an unprecedented combination of a triplet ground state that is nearly exclusively populated at room temperature and a novel 1D S = 1 antiferromagnetic chain, with remarkable thermal stability and suitability for thin film deposition via thermal evaporation. The films are stable under ultra-high vacuum for at least 60 h but show signs of decomposition when exposed to ambient conditions for 7 h. We present here the preparation and characterization of the first thin film of high spin organic diradical.</p><!><p>Our synthetic approach to 2, follows closely the synthesis of diradical 1 and it takes advantage of the unusual stability of the Blatter radical, such that it may tolerate many common reaction conditions (Scheme 1).21 Cyano-Blatter radical 3 is synthesized by adopting procedures available in the literature, as outlined in detail in the Supporting Information.17,22–25 Treatment of 3 with DIBAL-H followed by hydrolysis of the imine group provides formyl-Blatter radical 4. Radical 4 is initially reduced to the corresponding leuco-amine-aldehyde, or alternatively, it is directly condensed with 2,3-bis(hydroxyamino)-2,3-dimethylbutane;26 the resultant adduct is oxidized in air to provide diradical 2, which is purified by normal phase chromatography (silica gel) at ambient conditions.</p><!><p>The structure of diradical 2 consists of two non-equivalent molecules, A and B, and one molecule of solvent (CH2Cl2). In molecule A, the nitronyl nitroxide radical moiety is nearly co-planar with the Blatter radical π-system with the corresponding N-C8A-C1A-C torsions in the –13 – (–15)° range, while in molecule B, there is considerably greater out-of-plane twisting with the corresponding torsions in the 28 – 30° range (Fig. S1, SI). In the crystal, molecules A and B pack in an alternating fashion into one-dimensional chains (along the crystallographic a-axis) with close intermolecular N∙∙∙N and O∙∙∙N contacts (Figure 2, bottom plot).</p><p>Within the π-conjugated pathway in the diradical, there are two dihedral angles of ~ 49° and 30° connecting the 1,2,4-benzotriazinyl and nitronyl nitroxide moieties in 1,17 as opposed to one dihedral angle (or one torsion) in 2. The radicals in 2 are then more coplanar than in 1. The DFT geometry optimizations using UB3LYP/6–31G(d) level of theory27 and starting from either conformation as observed in molecule A or B give more co-planar structures that are similar to molecule A, with N-C8-C1-C torsions in the ±13.7 – (±13.8)° ranges (SI, Table S6).</p><!><p>The EPR spectrum of 2 in glassy matrices shows exclusively a triplet diradical, without a trace of monoradical impurities (Figure 3, Table 1). As expected, the spectral simulation28 reveals much greater spectral width, 2|D/hc| = 1.616 × 10−2 for 2 vs. 2|D/hc| = 4.64 × 10−3 for 1,17 because of relative proximity of the unpaired electrons and the nearly co-planar nitronyl nitroxide and benzotriazinyl moieties in 2. The B3LYP/EPR-II calculations estimate the relative values of D as D/hc = +1.20 × 10−2 cm–1 for 2 and D/hc = −5.47 × 10−3 cm–1 for 1.29 The computed D-tensor components are not only overestimated, as typical for these type of diradicals, but also the positive sign of D/hc in 2 is inconsistent with its experimental EPR spectrum (Supporting Information).30,31</p><!><p>We estimate singlet triplet energy gaps in polycrystalline 1 and 2 by fitting the χT vs. T data in the T = 1.8 – 320 K and 70 – 320 K ranges, respectively, to the diradical model (eq. 1).4a (1)χT=(1.118T/H)N{2sinh(a)/[1+2cosh(a)+exp((–2J/k)/T)]}</p><p>a = 1.345(H/(T – θ))</p><p>For 1, an average of three fits (Figure 4 and SI, Fig. S8) gives 2J/k = 252 ± 10 K, corresponding to ΔEST = 0.50 ± 0.02 kcal mol−1 (Table 1), which is in good agreement with 2J/k = 234 ± 36 K, obtained previously by quantitative EPR spectroscopy in dilute solutions/matrices.17 For 2, a much larger 2J/k = 876 ± 36 K is obtained as an average of four fits (Figure 5 and SI: Fig. S9 and Table S2), corresponding to ΔEST = 1.74 ± 0.07 kcal mol−1 (Table 1). Although the measured values of χT > 0.75 emu K mol−1 and fitted values of ΔEST > 0 would suggest triplet ground states for 1 and 2, these results are dependent on an accurate weight of the SQUID sample. Also, the fitting of χT vs. T for polycrystalline 1 and 2 require relatively large absolute values of negative mean-field parameters, θ ≈ –6 and –14 K, thus suggesting significant antiferromagnetic interactions between the S = 1 diradicals.</p><p>With such large values of |θ |, magnetization data at low temperatures could not be fit adequately to Brillouin functions, thus sample-weight independent evidence for triplet ground state could not be obtained. Therefore, we prepare dilute diradicals in glassy matrices by dispersing 1 and 2 in polystyrene.</p><p>For the dilute sample of 1 and 2 in polystyrene, two-parameter fits to χT vs. T data in the T = 1.8 – 370 and 1.8 – 360 K ranges give somewhat lower values of 2J/k = 165 ± 18 K and 2J/k = 838 ± 78 K (vs. polycrystalline diradicals), as an average of two and three fits, respectively (Figure 6 and SI: Figs. S11 and S12). Most importantly, these fits indicate that the value of θ is very small and practically negligible.</p><p>The magnetization (M) versus magnetic field (H) data, that is M versus H/(T − θ), at low temperatures (T = 1.8 – 5 K) provide excellent fits to the Brillouin functions with a small negative mean-field parameter, |θ| ≤ 0.05 K. Such fits have two variable parameters: total spin (S) and magnetization at saturation (Msat); the mean-field parameter θ is adjusted until the M/Msat versus H/(T − θ) plots overlap at all temperatures. The values of S ≈ 1.0 (0.978−1.012), determined from the curvature of the Brillouin plots, unequivocally indicate the triplet (S = 1) ground state for both diradicals (Figure 6 and SI: Figs. S11 and S12). Since the values of Msat = 0.74 and 0.925 μB for 1 and 2, respectively (μB = Bohr magneton) match well the corresponding values of N = 0.73 and 0.93 obtained from fits of χT vs. T data to the diradical model, this implies that both diradicals are pure and N < 1.00 is obtained because of a "weighing error" of sub-milligram amounts of diradicals. The values of θ < 0 K and |θ| ≈ 0 K imply nearly negligible, and almost certainly intermolecular, antiferromagnetic coupling.</p><p>We also investigate dilute diradicals in benzene and dibutylphthalate (DBP) matrices. Similar results are obtained for 1 and 2 in benzene and 2 in DBP; however, the fits to the Brillouin functions are less satisfactory because of much larger |θ| = 0.6 – 0.8 K observed in these matrices (SI, Table S3, Figs. S13 – S15).</p><!><p>For polycrystalline 1, the χ vs. T plot shows continuously increasing χ with decreasing T. For 2, a broad maximum at about 19 K in χ vs. T data is observed, thus suggesting relatively strong intermolecular coupling between S =1 diradicals 2 (Figures 4 and 5). Two limiting models for such antiferromagnetic coupling are considered: (1) one-dimensional (1-D) Heisenberg chains of S = 1 diradicals (spin-1 chain) (eq. 2) and (2) pairs of S = 1 diradicals (dimer) (eq. S2, SI).32,33 The numerical fits to these models are obtained at low temperatures, T = 1.8–70 K, to ensure that diradical 2 is almost completely in its S = 1 ground state at the highest temperature (70 K), which is significantly below 2J/k ≈ 880 K.</p><p>Initially, we fit χT vs. T data in the low temperature range, T = 1.8 – 70 K, using Eq. 2 and Eq. S2B (Table 2, Figure 5). 3-Parameter fits with the following variable parameters, intermolecular Heisenberg exchange coupling constant, J'/k, weight factor, N, and weight factor for isolated S = 1 diradical, Nimp, are in excellent agreement with 1D-chain model (coefficient of determination, R2 = 0.9999 and standard error of estimate, SEE = 0.0028 or 0.0032); fits to an S = 1 dimer model are less satisfactory (R2 = 0.9998 or 0.9994 and SEE = 0.0043 or 0.0067). Even larger differences in fit quality between the two models are observed when the variable parameter Nimp is replaced with the mean-field parameter θ > 0 (Table 2).</p><p>The χ versus T data provide a more sensitive measure of fit quality for different models. The fits to a 1D-chain (eq. 2) provide SEE = 0.0002 – 0.0003, while the S = 1 dimer fits (SI, eq. S2A) are much less satisfactory, with much larger SEE = 0.0017 – 0.0089, (Table 2, Figure 5).</p><p>a1 = 0.0194, a2 = 0.777, b1 = 4.346, b2 = 3.232, b3 = 5.634,</p><p>K = –J'/kT and a = 1.345(H/(T – θ))</p><p>The 1D antiferromagnetic chain of weakly coupled S = 1 diradicals in polycrystalline 2 is consistent with the crystal packing, as discussed above (Figure 2). Two types of short N2A∙∙∙N3B = 3.509 Å (dimer I) and O4B∙∙∙N1A = 3.335 (dimer II) contacts between molecules A and B are identified. Because all nitrogens and oxygens in diradical 2 bear large positive spin densities, such contacts are anticipated to give rise to intermolecular antiferromagnetic coupling between pairs of molecules A and B within 1D chain</p><p>We carried out DFT calculations, based upon the broken symmetry approach, at the fixed X-ray geometry for molecules A and B forming dimers I (AB) and II (BA) as well as trimers (ABA and BAB),18,34 to determine the values of J'/k. The computed J'/k are summarized in Table 3.</p><p>The agreement between the computed values of J'/k and the experimental J'/k = –14 K is reasonable, especially when accuracy of the DFT computations is considered.35 Also, while the alternating chain of J'/k may not be excluded, it is predicted by computation that the degree of alternation is relatively small and in agreement with the experiment.</p><!><p>Diradical 2 possesses excellent stability at ambient conditions, not only in air-saturated solution at ambient conditions (Fig. S4, SI) but also on silica gel. Notably, solid 2 shows no signs of decomposition after storing on air at –20 °C for more than 2 years. In addition, solid 2 shows remarkable stability under vacuum annealing at 100 °C for 24 h, and we found that such condition effectively removes the solvent of crystallization (CH2Cl2) from the crystal lattice (SI, Fig. S6). Thus, the annealed polycrystalline 2 is obtained for the EPR spectroscopy (Figure 3) and other studies, including TGA (Figure 7), as well as for thin films (Figure 8). Thermogravimetric analysis data suggest that thermal decomposition of 2 starts at 160 °C, at a temperature that is about 15 °C lower than that for diradical 1 (Figure 7).</p><!><p>We test the robustness of diradical 2 toward evaporation. This aspect is important in view of the potential applications of this diradical in electronics. The ability to attach a molecule to a contact or to evaporate it onto a substrate, without degradation, is a requirement for device fabrication. Controlled evaporation of diradicals is considered very challenging, and not yet reported in literature. The presence of two radical sites, in particular the cross-conjugated diradicals with significant spin density within the π-system connecting two radical sites, such as in 2, could potentially increase their instability during evaporation.36</p><p>We prepare thin films of diradical 2 on SiO2/Si(111) wafers by using organic molecular beam deposition (OMBD) that allows for controlled evaporation and the consequent deposition of molecules onto a substrate, tuning the preparation conditions in ultra-high vacuum (UHV).37 We investigate the films by using X-ray photoelectron spectroscopy (XPS). XPS is an effective and powerful tool for investigation of organic and organic radical thin films.38 Besides providing insight into the occupied states, it is element-sensitive, and it can also deliver quantitative information on the stoichiometric composition of the films,39 due to the high sensitivity of the signal to the concentration of the emitting atoms. In addition, the features contributing to the spectroscopic lines are sensitive to the different chemical environment of the atoms of the same element. These assets are the basis for our analysis of core level spectra of a multilayer of 2 (Figure 8).</p><p>We focus on the C 1s and N 1s spectra, because the O 1s spectrum is a convolution of the substrate and the molecule signal making the analysis less reliable. The film C 1s spectrum is characterized by a main line at around 285 eV due to photoelectrons emitted from the atoms in the aromatic ring and the carbon atoms bound to hydrogen atoms (C-C, C-H and CH3). The shoulder at higher binding energy is due to contributions from the electrons emitted from carbon atoms bound also to nitrogen (C-N). Nitrogen atoms, because of their higher electronegativity, shift the electronic cloud. Thus, the carbon atoms bound to nitrogen atoms have smaller electron density and, consequently, the electrons are emitted with lower kinetic energy, i.e., higher binding energy. The N 1s core level spectrum shows contributions due to five nitrogen atoms: the three nitrogen atoms belonging to the Blatter radical have different chemical environment,40,41 while the two nitrogen atoms belonging to the nitronyl nitroxide (NN) radical have an equivalent chemical environment. These differences are mirrored in the spectrum by the presence of two broad features, showing the highest intensity at around 402 eV. This binding energy corresponds to the line expected in the NN radical N 1s core level spectrum.42</p><p>A best fit procedure allows identifying the contributions from different atomic sites having slightly different binding energies due to variations in the chemical environment.43 In calculating the best fit, we applied several constraints based on electronegativity, and bond strength42,43 (see also Supporting Information for details). We used Voigt profiles, with fixed constant Lorentzian width (0.08 and 0.10 eV, for C 1s and N 1s curves, respectively).42–44 The Voigt profile takes into account both the finite core-hole lifetime (Lorentzian profile) and the broadening due to the finite experimental resolution and various inhomogeneities, e.g., molecular packing and local morphology43,45 (Gaussian profile). To calculate the stoichiometry of the films, we also took into account the intensity of the satellites43,44 typical features in photoemission that appear as an effect of the relaxation processes due to the creation of a core-hole.46 Based on comparison of the film fit results (SI: Tables S4 – S6) and the molecule stoichiometry, we can conclude that there was no degradation of the diradical molecules under our controlled evaporation condition.</p><p>This result is further supported by the XPS investigations performed on the powder samples, i.e., on molecules that did not undergo evaporation (Figure 8, bottom plots). Apart from a broadening of the lines and small energy shifts, due to typical charging effects occurring in organic crystals,47 the film spectra are fully concomitant with the powder spectra. Thus, it is evident that diradical 2 is stable and robust to be evaporated under controlled conditions to form films of intact diradical molecules.</p><p>To shed light on the growth mode of diradical 2 under this preparation conditions, we follow the XPS core level signal of the substrate (Si 2p) by looking at its attenuation upon film deposition (Figure 9, top panel). The curve is characterised by a very slow decay. This intensity trend hints at a Volmer-Weber (VW) growth mode, i.e., island growth.48 This result is consistent with the atomic force microscopy (AFM) ex-situ images obtained on diradical 2 films (Figure 9 middle panel) clearly showing a film morphology dominated by islands. The VW growth mode occurs when the the interaction between the deposited molecules is much stronger than between the molecules and the substrate. The VW growth of diradical 2 signifies the interacting molecules of diradical 2, possibly resembling the 1D spin-1 chain in the solid state, on the inert SiO2/Si(111) surfaces. The line profile (Figure 9, bottom panel), obtained averaging the AFM signal over all rows, evidences the formation of islands of different lateral size that for big assemblies ranges between 80 and 300 nm.</p><p>Near edge X-ray absorption fine structure (NEXAFS) spectroscopy offers the advantage to investigate in-situ not only the electronic structure of a material, namely the unoccupied states, but also the structural properties of very thin films. Thus, we investigate diradical 2 thin films to determine the molecular orientation with respect to the substrate, by using two different polarization directions of the incident light, giving rise to NEXAFS dichroism (Figure 10).</p><p>We focus on the C-K edge spectra. In analogy with the NEXAFS spectra of carbon-based molecules, the spectra in Figure 10 are characterized by two main regions: the π* region up to around 290 eV and the σ* region in the photon energy range above 290 eV.50 Several features are typically expected in the 286.0–287.4 eV photon range in the spectra of N-substituted aromatic carbon:51 indeed, a strong resonance is visible at 286.2 eV. This resonance is due to transitions from C 1s levels, belonging to carbon atoms bound to nitrogen.52–54 We also observe a very pronounced shoulder with two small knees at 284.4 and 285.2 eV. They have a C=C character, with the first shoulder mainly due to excitations along the molecular backbone and the intensity at around 285 eV within the phenyl groups.50,55–57 The features show a strong dichroic behavior with the signal intensity at around 285 eV quenched and the peak at 286.2 eV losing intensity in normal incidence. First, this clearly indicate that the island aggregation is not amorphous because in that case the signal for the two polarization directions would overlap. Second, this kind of NEXAFS dichroism agrees with an average orientation of the molecules in the film (note that the NEXAFS signal is averaged on the area spotted by the photon beam) similar to the one adopted in the single crystals and with the crystallographic a-axis of the unit cell almost perpendicular to the substrate (see Figure 2).</p><p>Finally, we monitor the stability of the films in UHV (base pressure 2 × 10−10 mbar) by using XPS, focusing on the N 1s core level spectra that represent nitronyl nitroxide and Blatter radicals.40–42 We observe no major changes in the spectra of the films after their exposure to UHV at room temperature for 17 and 60 h (Fig. S20, SI). However, the films are much less robust in air, as we observed major changes in their XPS after 7 h of air exposure (Fig. S21, SI). We note that the previously studied films of nitronyl nitroxide and the Blatter monoradical derivatives showed similar changes in their XPS after the films were kept for several weeks and several months at ambient conditions, respectively.40–42,58 While we have demonstrated that it is possible to evaporate diradicals and deposit their thin films under controlled conditions without degradation, our results indicate that the diradical films are less stable when compared to the films of their monoradical analogues.</p><!><p>We have synthesized an organic diradical 2, which, at room temperature, exists nearly exclusively in its high-spin, S = 1, ground state and it possesses a remarkable thermal stability to permit fabrication of intact diradical thin films on silicon substrate via evaporation under ultra-high vacuum. The diradical molecules form islands on the substrate with molecules stacked approximately along the crystallographic a-axis. The diradical films were found to stable under ultra-high vacuum for at least 60 h, however, within few hours of exposure to air, XPS of the films showed major changes. While we have demonstrated that it is possible to evaporate diradicals and deposit their thin films under controlled conditions without degradation, our results indicate that the diradical films are less stable when compared to the films of nitronyl nitroxide or Blatter monoradicals. Polycrystalline diradical 2 consists of nearly isotropic 1-D antiferromagnetic S = 1 Heisenberg chains at low temperature. Notably, 2 possesses record intra-chain antiferromagnetic coupling, J'/k = –14 K, among all to date studied S = 1 chains of organic radicals,18,20 with a Haldane gap of 0.41 × 2|J′/k| ≈ 11.5 K. The 1D chain of 2 is also most isotropic, with very weak local anisotropy, |D/2J′| ≈ 4 × 10–4,59 and thus is potentially an excellent system for studies of low dimensional magnetism.60 Such diradical with an unprecedented combination of novel magnetic and thermal properties, suitable for thin film fabrication under ultra-high vacuum, could facilitate the development of purely organic magnetic and electronic materials.</p><!><p>Frozen solution EPR spectra were obtained using a Bruker EMX or EMX-plus X-band spectrometer and simulated with the EasySpin software.28 The TGA/DSC or TGA instrument (TA Instruments TGA 550) was run either without or with IR attachment (Thermo NICOLET Is50 NIR). Variable temperature (from 1.8 K to up to 370 K) magnetic susceptibility measurements of 1 and 2 were performed using a Quantum Design SQUID magnetometer with applied magnetic fields of 30 000, 5000, and 500 Oe. Variable field (0 – 50,000 Oe) magnetization studies were carried out at temperatures of 1.8 – 5 K. Sample tubes for SQUID studies in dilute matrices are described in the SI.61</p><!><p>Crystals of 2 for X-ray studies was prepared by slow evaporation from solution in DCM/cyclohexane. Data collection was performed at the Advanced Photon Source, Argonne National Laboratory using λ = 0.41328 Å synchrotron radiation (silicon monochromators). Final cell constants were calculated from the xyz centroids of 9989 strong reflections from the actual data collection after integration (SAINT).62 The intensity data were corrected for absorption (SADABS).63 The space group P-1 was determined based on intensity statistics and the lack of systematic absences. The structure was solved and refined using the SHELX suite of programs.64 All non-hydrogen atoms were refined with anisotropic displacement parameters. The hydrogen atoms were placed in ideal positions and refined as riding atoms with relative isotropic displacement parameters. Crystal and structure refinement data for 2 are in the Supporting Information and the accompanying file in CIF format.</p><!><p>Standard techniques for synthesis under inert atmosphere (argon or nitrogen), using custom-made Schlenk glassware, custom-made double manifold high vacuum lines, argon-filled Vacuum Atmospheres gloveboxes, and nitrogen-filled glovebags. Chromatographic separations were carried out using normal phase silica gel. Multi-step, efficient synthesis and characterization of the starting Blatter radical 3 is outlined in the Supporting Information (Scheme S1).</p><!><p>Starting 7-cyano Blatter radical 3 (1.605 g, 5.19 mmol) was dissolved in dichloromethane (100 mL) and cooled to –78 °C under a light N2 flow. DIBAL-H (1 M in hexane, 12.0 mL, 12.0 mmol) was then added to the solution at –78 °C. The reaction was stirred at –78 °C for one hour and then warmed to room temperature with stirring for one hour. Then, 1 M HCl (100 mL) was added and the bilayer was stirred at room temperature for about 20 min. This caused a sizeable amount of precipitate to collect on the walls of the round bottom flask. The bilayer was decanted, separated, and the organic layer was shaken vigorously with aqueous KOH. The organic layer was then dried over Na2SO4 and evaporated (0.634 g); TLC indicated this solid to be radical 4 with only very minor impurities. The precipitate that was formed after HCl addition was then exposed to concentrated KOH and dichloromethane, causing it to dissolve in the organic layer upon mixing. TLC indicated a sizeable amount of target material 4, but with significantly more impurities. The solvent was evaporated and this residue purified on silica (dichloromethane eluent) to yield an additional 0.199 g of pure radical, to provide total of 0.833 g of 4 (52% yield). IR (powder, cm−1): 3072, 3018, 2825, 2756, 2727, 1682, 1572, 1487, 1386, 1311, 1184, 1114, 1026, 829, 768. EPR (X-band, 9.65 GHz, benzene): g = 2.0035, aN1 = 0.77 mT, aN2 = 0.48 mT, aN3 = 0.46 mT. HR-ESI: 313.1229, 100%, [M+H]+, calculated for [M+H]+: 313.1215, also: 312.1136, 77%, M+, calculated for M+: 312.1137. M.p. (DSC, 5 °C/min): 212–217 °C. To further characterize radical 4 by diamagnetic NMR spectroscopy, 4 (~4 mg) was dissolved in DMSO-d6 (~0.5 mL), and then an excess of sodium dithionite was added to the NMR sample. Gently heating the NMR tube (to dissolve enough sodium dithionite to reduce the radical to leuco-triazine) caused a color change to the characteristic yellow of the reduced radical. This allowed for acquisition of 1H NMR and 13C NMR spectra for the leuco-triazine. 1H NMR (400 MHz, DMSO-d6): 9.56 (s, 1H), 9.34 (s, 1H), 7.82 (dd, 2H, J1 = 7.8 Hz, J2 = 1.4 Hz), 7.43–7.51 (m, 7H), 7.33 (dd, J1 = 7.8 Hz J2 = 1.4 Hz), 7.22–7.18 (m, 1H), 6.86 (d, 1H, 7.6 Hz), 6.68 (s, 1H). 13C NMR (DMSO-d6): δ = 190.6, 146.4, 143.3, 140.5, 135.1, 132.2, 130.7, 130.4, 129.53, 129.34, 128.5, 125.9, 124.6, 122.2, 113.1, 107.9</p><!><p>Note: in this procedure, the first step of reduction of radical 4 with Na2S2O4 was omitted, that is, the second step (condensation with bis-hydroxyamine) was run directly on the radical. Blatter radical 4 (0.626 g, 0.46 mmol) was added to a Schlenk vessel followed by 2,3-bis(hydroxyamino)-2,3-dimethylbutane (0.503 g, 3.34 mmol). After purging the Schlenk vessel with nitrogen gas, MeOH (20 mL) was added. The suspension was heated to 70 °C in the Schlenk vessel overnight, during which time the mixture became homogenous. Then the solution was cooled and diluted approximately fourfold with ethyl acetate, and subsequently washed twice with brine, dried, and evaporated. The solid obtained was dissolved in dichloromethane (300 mL), and then triethylamine (2.0 mL, 14.28 mmol) was added. The solution was stirred overnight with a light air flow bubbling through the reaction. The resulatant purple/red colored solution was evaporated. The diradical was purified on silica eluting with 4:1 dichloromethane/EtOAc. Then the solid diradical was washed sequentially with 10 mL pentane, 10 mL Et2O, and finally 10 mL MeOH; the solid was then dried under high vacuum at 100 °C in a chamber overnight, to remove any co-crystallized dichloromethane and other residual solvents (0.332 g, 38% yield). IR (powder, cm−1): 3101, 3049, 2982, 2914, 1585, 1483, 1388, 1361, 1315, 1269, 1136, 823, 777, 733. HR-ESI: 439.2010, 100%, M +, calculated for M +: 439.2008.</p><!><p>All geometry optimizations were carried out at the UB3LYP/6–31G(d) level of theory, with obtained minima confirmed by frequency calculations. The broken-symmetry approach was applied for open-shell singlet calculations and spin contamination errors were corrected by approximate spin-projection method.65 Computations of an intra-dimer coupling constant J'/k were carried out using dimers and trimers of 2 at X-ray geometry, using broken symmetry approach at the UB3LYP/6–31G(d) or UB3LYP/6–311++G(d,p) levels of theory.34,66 All calculations were performed with the Gaussian 09 program suite.27</p><!><p>Thin film growth and XPS measurements were performed in an UHV system comprising a substrate preparation chamber and a dedicated OMDB chamber connected to an analysis chamber (base pressure 2 × 10−10 mbar) equipped with a monochromatic Al Kα source (SPECS Focus 500) and a SPECS Phoibos 150 hemispherical electron analyser. As a substrate, native SiO2 grown on single-side-polished n-Si(111) wafers was used. The substrate was cleaned in an ultrasonic bath in acetone and ethanol (one hour each consecutive bath) and then annealed at around 500 K for 15 hours. The cleanness was verified by XPS. Thin films of 2 were grown in-situ by OMBD using a Knudsen cell keeping the substrate at room temperature. Powder samples were obtained embedding the powder in a passivated indium foil. The nominal thickness was determined by using the attenuation of the Si 2p XPS signal of the substrate. The spectra were measured at 20 eV pass energy, and the binding energy calibrated to the Si 2p signal at 99.8 eV. Because of the radiation-sensitivity of the diradical, beam exposure was minimized and a freshly prepared film was used for each set of spectra to prevent radiation damage. For the XPS measurements probing stability, the set of spectra was measured on the same films upon UHV or air exposure minimizing the acquisition time. The spectra have a slightly worse signal-to noise ratio to preserve the intactness of the molecules in the films.</p><p>NEXAFS measurements were performed at the third-generation synchrotron radiation source Bessy II (Berlin) at the LowDose PES end-station, installed at the PM4 beamline (E/ΔE = 6000 at 400 eV) that included substrate preparation facilities like those described above for the XPS station. The measurements were carried out in multibunch hybrid mode (ring current in top up mode = 250 mA, cff = 1.6, 100-μm exit slit). The NEXAFS spectra, measured in total electron yield, were normalized by using the clean substrate signal and the ring current into account, and then scaling all spectra to give an equal edge jump.50,57,67 Atomic force microscopy (AFM) studies were performed under ambient conditions in tapping mode with a Digital Instruments Nanoscope III Multimode AFM. No beam-induced degradation of the samples was observed on the time scale of all discussed experiments.</p>

PubMed Author Manuscript

A PAL for Schistosoma mansoni PHM

Parasitic helminth neuromuscular function is a proven target for chemotherapeutic control. Although neuropeptide signaling plays a key role in helminth motor function, it has not yet provided targets for known anthelmintics. The majority of biologically active neuropeptides display a C-terminal amide (NH2) motif, generated exclusively by the sequential action of two enzymes, peptidylglycine \xce\xb1-hydroxylating monooxygenase (PHM) and peptidylglycine \xce\xb1-amidating lyase (PAL). Further to our previous description of a monofunctional PHM enzyme (SmPHM) from the human blood fluke Schistosoma mansoni, here we describe a cDNA encoding S. mansoni PAL (SmPAL). SmPAL is a monofunctional enzyme which, following heterologous expression, we find to have functionally similar catalytic activity and optimal pH values, but key catalytic core amino acid substitutions, when compared to other known PALs including those found in humans. We have used in situ hybridization to demonstrate that in adult schistosomes, SmPAL mRNA (Sm-pal-1) is expressed in neuronal cell bodies of the central nervous system, consistent with a role for amidated neuropeptides in S. mansoni neuromuscular function. In order to validate SmPAL as a putative drug target we applied published RNA interference (RNAi) methods in efforts to trigger knockdown of Sm-pal-1 transcript in larval schistosomula. Although transcript knock-down was recorded on several occasions, silencing was variable and inconsistent and did not associate with any observable aberrant phenotype. The inconsistent outcomes of RNAi suggest that there may be tissue-specific differences in the applicability of RNAi methods for S. mansoni, with neuronal targets proving more difficult or refractory to knockdown. The key role played by schistosome amidating enzymes in neuropeptide maturation make them appealing as drug targets; their validation as such will depend on the development of more robust reverse genetic tools to facilitate efficient neuronal gene function studies.

a_pal_for_schistosoma_mansoni_phm

1. Introduction<!>2.1. Schistosoma mansoni culture<!>2.2. Bioinformatics<!>2.3. RNA extraction and PCR analyses<!>2.4. Functional SmPAL expression<!>2.5. Cell culture, transfection, and PAL enzyme assay<!>2.6. Sm-pal-1 probe generation and WISH<!>2.7. dsRNA generation<!>2.8. dsRNA delivery<!>2.9. Reverse-transcriptase PCR (RT-PCR)<!>3.1. Identification and characterisation of schistosome PAL (SmPAL)<!>3.2. Functional analysis of SmPAL<!>3.3. Sm-pal-1 transcript expression<!>3.4. Sm-pal-1 RNAi<!>4. Discussion<!>

<p>Diseases caused by helminth parasites remain one of the most significant public health problems facing society. The most prevalent and chronic flatworm infection of humans, resulting in a major, poverty-related problem in the developing world is schistosomiasis. Schistosoma mansoni, designated a Neglected Tropical Disease, is a major cause of human schistosomiasis, having a global prevalence of over 200 million in an estimated 74 developing countries [1]. The difficult situation facing health organisations involved in the reduction of morbidity due to schistosomiasis is aggravated by over-reliance on a single drug, praziquantel (PZQ). Although resistance does not appear to be widespread, it has been generated in the laboratory [2], and has occurred transiently in field situations [3, 4]. These observations, coupled with increasing distribution and usage of PZQ, provide circumstances known to foster the development of drug resistance. As a result of this dependence and the consequent risk of inadvertently selecting for drug-resistant parasites, there is a pressing need to exploit and develop new drug targets and to develop novel chemotherapeutic agents for the treatment and control of schistosomiasis [5].</p><p>The neuropeptidergic system is central to flatworm motor control - neuropeptide signalling molecules appear to be involved in the modulation of a range of biological processes throughout the phylum Platyhelminthes [6–11]. So far, we have identified 18 distinct amidated peptides in S. mansoni [12, 13], several of which have been localised to the CNS and PNS of larval and adult schistosomes [13–19]. Some of these peptides have also been shown to exert potent myoexcitatory effects on isolated muscle fibres from adult schistosomes, and as such are thought to be directly involved in muscle function [20]. Consequently, the neuropeptidergic component of the S. mansoni nervous system represents an attractive repository of novel drug targets.</p><p>The majority of neuropeptides require a carboxy-terminal amide motif to confer biological activity, as demonstrated by the reduced activity of most non-amidated glycine-extended precursors (<10 %) in comparison to their α-amidated counterparts [21–24]. Two key enzymes act sequentially in the only known mechanism for secretory peptide amidation: peptidylglycine α-hydroxylating monooxygenase (PHM), a copper, oxygen and ascorbate dependant enzyme which catalyses hydroxylation of the C-terminal glycine present on a peptide precursor; and peptidylglycine α-amidating lyase (PAL), a zinc dependant enzyme which catalyses dealkylation of the hydroxylated intermediate to form the α-amidated neuropeptide and glyoxylate [21, 23]. This process occurs throughout the Metazoa, the importance of which is demonstrated in larval Drosophila melanogaster mutants, where the deletion of the PHM gene results in embryonic death [25]. In higher organisms PHM and PAL are commonly encoded by the same transcript, generating a bifunctional protein designated peptidylglycine α-amidating monooxygenase (PAM) [24]. In stark contrast, the genomes of a number of invertebrates including the flatworm Dugesia japonica, encode single copies of monofunctional PHM and PAL [26], while others possess multiple monofunctional enzymes which can be membrane associated, soluble or even inactive [24, 27]. Crucially, adult S. mansoni express monofunctional PHM (SmPHM), which is functionally divergent from the mammalian homologue [28]. These organisational and functional differences between the schistosome and mammalian amidating enzymes encourage the characterisation of schistosome amidating enzymes and their exploration as potential drug targets.</p><p>Even though neuropeptide signalling is a core component of schistosome neural function and neuromuscular modulation, it has not been a major focus for novel drug target discovery and validation in schistosomes, largely hindered by the absence of known neuropeptides and their cognate receptors. Recent proliferations in EST datasets and publication of the S. mansoni genome sequence [29] have fuelled the recent discovery of SmPHM and eighteen schistosome neuropeptides, making neuropeptide signalling ripe for more detailed interrogation and validation as a source of suitable targets for schistosome control [13, 28, 29]. Further, genome level bioinformatics has uncovered a set of at least 24 putative schistosome neuropeptide receptors [29], which are appealing because of the proven druggability of many known G-protein coupled receptors (GPCRs) and their suitability for high throughput screens. We hypothesise that neuropeptide activation/processing provides an alternative and potentially powerful set of drug targets. Since the combined activities of PHM and PAL are the only known conduit to neuropeptide amidation/activation and we already know schistosomes possess multiple amidated peptide messengers, these enzymes have much appeal as drug target candidates. Previously we demonstrated that one half of the neuropeptide amidating pathway in S. mansoni is functionally distinct from that seen in vertebrates and, indeed, other invertebrates, such that we now switch focus to the second half of the amidation pathway, PAL. Here we report the identification, localization, cloning and functional characterisation of a S. mansoni cDNA encoding a novel, monofunctional PAL enzyme. Further, we attempt to validate its candidature as a drug target through the application of RNA interference (RNAi).</p><!><p>Schistosome infected mice supplied by Dr Fred Lewis, Biomedical Research Institute, Rockville MD, USA, were maintained and processed at Iowa State University (ISU). Adult schistosomes recovered from infected mice were either processed for whole-mount in situ hybridisation (WISH) experiments as previously described [30], or stored in RNAlater to allow subsequent RNA extraction, before being shipped to Queen's University Belfast (QUB) on dry ice. Schistosomes used in the RNA interference (RNAi) experiments were maintained at QUB in Biomphalaria glabrata snails; cercariae were shed from infected snails by photostimulus, and mechanically transformed to schistosomula by vortexing, schistosome bodies were isolated from tails by centrifugation over a solution of 30% Percoll in water (500 g for 15 min at 4°C), protocol adapted from [53]. Schistosomula were maintained in vitro in complete RPMI media [31] containing 20% serum, at 37°C in a 5% CO2 atmosphere.</p><!><p>BLAST methodology [32] was employed to uncover novel putative PAL encoding transcripts from S. mansoni EST and genomic datasets. BLASTn and tBLASTn tools were used at the National Centre for Biotechnology Information (NCBI) BLAST server (http://blast.ncbi.nlm.nih.gov/Blast) as well as the Schistosoma mansoni genome database, SmGeneDB (http://www.genedb.org/genedb/smansoni). Searches were performed on the 'est_others' database, had an expect value of >1000, and were limited by the query 'Schistosoma mansoni'. Known PAL genes [25] were used as query sequences. Returns were translated in all six reading frames (http://www.expasy.org/tools/dna.html), and examined for both the presence of PAL specific motifs and similarity to other PAL genes/proteins (http://www.ebi.ac.uk/Tools/InterProScan) [33]. Sequences were also analysed for the presence of an N-terminal secretory signal peptide, and putative N-glycosylation sites, using the online SignalP 3.0 (http://www.cbs.dtu.dk/services/SignalP/) [34], and NetNGlyc servers (http://www.cbs.dtu.dk/services/NetNGlyc/) respectively.</p><!><p>Messenger RNA was extracted from between 10–50 mg of S. mansoni tissue using Dynabeads mRNA Direct™ kit (Invitrogen) according to the manufacturer's instructions. mRNA eluted in Tris-HCl was quantified using a NanoDrop™ 1000 spectrophotometer. Separate populations of 5′ and 3′ RACE-ready cDNAs were generated using the SMART™ RACE cDNA Amplification kit (Clontech/Takara), using ≥1 μg mRNA per synthesis according to manufacturer's instructions. RACE cDNA was stored at −20°C until use.</p><p>Gene specific primers (GSP) designed against the putative S. mansoni PAL (SmPAL) ESTs (GenBank accession numbers AM047261 and AM043268) were used in 50 μl PCR reactions (94°C for 2 min; 40 cycles of : 94°C 1 min, 55°C 1 min, and 72°C 1 min; 72°C for 7 min) to confirm the presence of SmPAL in the following reaction mixture: 5 μl 10x PCR buffer (Invitrogen), 3 μl MgCl2 (50 mM, Invitrogen), 1 μl dNTP mix (10 mM, Promega), 1 μl of primer SmPAL-F1 and SmPAL-R1 (20 μM; Table 1), 1 μl cDNA template, 0.3 μl Platinum® Taq DNA Polymerase (5 U/μl, Invitrogen), and ddH2O to 50 μl. Reaction products were TOPO-TA cloned (Invitrogen) and at least 3 plasmids were sequence verified.</p><!><p>The SmPAL open reading frame (1236 bp) was PCR-amplified using sense primer SmPAL-Xba1-F and antisense primer SmPAL-Bgl2-R (see Table 1) incorporating 5′ Xba1 and Bgl2 restriction sites respectively. Products were TOPO-TA cloned (Invitrogen) and sequence verified. Using Xba1, Bgl2, and pBS.rhodopsin, the rhodopsin tag (11 amino acids) was appended to the C-terminus of SmPAL. pCIS.SmPAL-rhod was then created using Xba1 and Hpa1. pBS.SmPAL-rhod was used as a template to PCR-amplify SmPAL-rhod without the native S. mansoni signal peptide (aa 1–18) using primers SmPAL-Nhe1-F and SmPAL-Hpa1-R (see Table 1). The SmPAL-rhod reaction product was placed after the rat PAM signal peptide and a poly histidine tag in pCIS.sig-His6 yielding pCIS.sig-His6-SmPAL-rhod. pCIS.SmPAL-rhod was used in functional expression assays.</p><!><p>Functional expression and characterisation of SmPAL enzymatic activity were carried out as described [35]. SmPAL, a control vector encoding cytosolic Enhanced Green Fluorescent Protein (EGFP; pEGFP-N2, Clontech, Palo Alto, CA, USA), and Drosophila PAL (dPAL2) [35] were expressed in pEAK RAPID cells (a derivative of hEK-293 cells, Edge Biosystems, Gaithersburg, MD, USA). Cells were transiently transfected using Lipofectamine 2000 (2 μg/30 mm well). 24 h after transfection, medium was replaced with serum-free medium, and the cells were incubated for a further 24 h before cells and spent media were harvested for analysis.</p><p>Cell extracts and spent medium were assayed for catalytic activity and Western blots were carried out to evaluate secretion efficiency; protease inhibitor cocktail and phenylmethanesulfonylfluoride (PMSF) were added to cell extracts and to spent medium prior to analysis. For comparison, cells and spent medium containing expressed dPAL2 [34] were analysed under the same experimental conditions as SmPAL. pEAK RAPID cells expressing SmPAL were extracted in a low ionic strength buffer with detergent and protease inhibitors. Spent medium from pEAK RAPID cells was centrifuged to remove debris. For Western blot analysis, media were denatured by boiling in Laemmli sample buffer. For enzyme assays, media were diluted into assay diluents (20 mM Na TES (pH 7.4), 10 mM mannitol, 1.0 mg/mL bovine serum albumin, and1% Triton X-100) [39].</p><p>Aliquots of cell extract and spent medium were fractionated on 4–15% polyacrylamide gels (Invitrogen), transferred to PVDF membranes and subjected to Western blot analysis using a monoclonal rhodopsin antibody. Secreted, rhodopsin tagged SmPAL was assessed for catalytic activity, using 0.5 μM αN-acetyl-Tyr-Val-α-hydroxyglycine (mixed with a trace amount of [125I]-labelled peptide), prepared using recombinant PHM. SmPAL was diluted in 20 mM Na TES (pH 7.4, 100mM Tris, 100mM EDTA and 2% SES), 10 mM mannitol, 1.0 mg/ml bovine serum albumin, and 1% Triton X-100 (Pierce). The reaction volume (40 μl) typically contained 0.4 μl of spent medium (diluted to 5 μl), 100–150 mM Na MES (pH 4.5), 1 mM CdCl2 and 0.05% Thesit (Boehringer Mannheim). The reaction mixture was incubated for 30 min to 1 h at 37 °C, following which the substrate and product were separated using ethyl acetate phase separation [39]. SmPAL dependence upon pH was analysed using the catalytic activity assay conditions described above and altering pH using 150 mM Na MES buffers that ranged in pH from 3.0 to 7.5. In all activity assays and Western blots, pEGFP-transfected cells were used as a blank.</p><!><p>Sm-pal-1 WISH probe templates were generated by PCR (94 °C 2 min; 40 cycles of: 94 °C 1 min, 55 °C 1 min, 72 °C 1 min; 72 °C for 7 min) using GSPs flanked by T7 RNA polymerase promoter sequences (sense labelled T7 template: sense primer SmPAL-ISH-T7-F1 and antisense primer SmPAL-ISH-R1; antisense T7 labelled template: sense primer SmPAL-ISH-F2 and antisense primer SmPAL-ISH-T7-R2; see Table 1) in a 50 μl reaction mixture as described above. Probe templates were cloned into pCR2.1-TOPO and at least 3 plasmids were sequence verified. Sm-pal-1 digoxigenin (DIG)-labelled single stranded RNA (ssRNA) probes with sense and antisense polarity were generated from their cDNA templates, and WISH was carried out according to the methods described [30, 36]. Hybridised probes were detected with substrate (5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium tablet; BCIP/NBT, Sigma-Aldrich) [30]. Specimens were photographed using a Leica DFC300FX camera and Leica FW4000 V 1.2 software with a Leica DMR light microscope.</p><!><p>A 226 bp fragment of Sm-pal-1 was generated by PCR (94 °C 2 min; 40 cycles of: 94 °C 1 min, 55 °C 1 min, 72 °C 1 min; 72 °C for 7 min) using GSP flanked by T7 RNA polymerase promoter sequences (sense T7 labelled template: sense primer SmPAL-RNAi-T7-F1 and antisense primer SmPAL-RNAi-R1; antisense T7 labelled template: sense primer SmPAL-RNAi-F2 and antisense primer SmPAL-RNAi-T7-R2; see Table 1), in a 50 μl reaction mixture as described above. In addition a 224 bp fragment of the SmPHM transcript [28] was amplified by PCR (94 °C 2 min; 40 cycles of: 94 °C 1 min, 55 °C 1 min, 72 °C 1 min; 72 °C for 7 min) using GSP flanked by T7 RNA polymerase promoter sequences (sense T7 labelled template: sense primer SmPHM-RNAi-T7-F1 and antisense primer SmPHM-RNAi-R1; antisense T7 labelled template: sense primer SmPHM-RNAi-F2 and antisense primer SmPHM-RNAi-T7-R2; see Table 1), in a 50 μl reaction mixture. Sm-phm-1 dsRNA was used in Sm-pal-1 RNAi experiments as a non-target control dsRNA. Products were TOPO-TA cloned and at least 3 plasmids were sequence verified. The Sm-pal-1 construct displayed no similarity to any other S. mansoni sequence as shown by BLASTp and tBLASTn searches of schistosome sequences on NCBI BLAST server.</p><p>Separate populations of ssRNA were generated from T7 labelled templates with sense or antisense polarity, with the MEGAshortscript™ High Yield Transcription kit (Ambion) at 37 °C for 4 h according to the manufacturer's instructions. 1 μl of DNase I (New England Biolabs) was incubated with ssRNA populations at 37 °C for 15 min, equimolar amounts of sense and antisense ssRNA were mixed and annealed at 75 °C for 5 min and cooled at room temperature for 30 min to generate dsRNA. 4 μl of RNase If (New England Biolabs) was added to dsRNA and incubated at 37 °C for 20 min to remove any ssRNA. dsRNA was recovered by phenol/chloroform extraction and ethanol precipitation, re-suspended in 100 μl H2O, quantified using the NanoDrop 1000 spectrophotometer, and stored at −80 °C until use.</p><!><p>Sm-pal-1 and Sm-phm-1 dsRNA delivery by electroporation was carried out as described [37]. 200, 10 day old schistosomula were electroporated in 50 μl of serum-free RPMI culture medium supplemented with 1 mg/ml Sm-pal-1 dsRNA. During each experiment 200 schistosomula were electroporated in serum-free RPMI without the addition of dsRNA (−dsRNA), as a negative control. Schistosomula were observed immediately following electroporation to assess damage, in all cases worms appeared normal following exposure. Schistosomula were cultured overnight in 100 μl complete RPMI culture medium, which was replaced with fresh complete RPMI (100 μl) each day for two days after which gene expression was measured by reverse-transcriptase (RT)-PCR. Schistosomula were observed once a day, throughout the post-dsRNA delivery culture period, for the appearance of any aberrant phenotypes associated with altered movement. dsRNA delivery by soaking was also attempted in this study. Following the 10 day transformation period, approximately 200 schistosomula were transferred into 100 μl of serum-free RPMI supplemented with 0.1 mg/ml Sm-pal-1 and Sm-phm-1 dsRNA. Each day 50 μl of RPMI was replaced with fresh RPMI supplemented with 0.1 mg/ml dsRNA. One batch of schistosomula was cultured for the duration of each soaking experiment in serum-free RPMI in the absence of dsRNA (−dsRNA), as a negative control. Soaking was originally carried out for 7 days; however, a high death rate after 5–6 days forced a shorter soaking time such that experiments were reduced to 5 days. Each day, dsRNA treated and untreated schistosomula were observed visually for the presence of any aberrant phenotypes, including differences in general appearance, morphology and movement (number and speed of contractions). After 5 days gene expression was measured by RT-PCR.</p><!><p>mRNA was extracted from schistosomula using Dynabeads mRNA Direct™ kit (Invitrogen) according to the manufacturer's instructions. Any remaining DNA was removed by DNase digestion using DNase I (New England Biolabs), following which the mRNA was purified using RNeasy Mini Elute Cleanup kit (Qiagen), and quantified using the NanoDrop 1000 spectrophotometer. Specific gene products were amplified from 12 ng of mRNA per RT-PCR reaction using GSP (Table 1) at a final concentration of 0.6 μM, and the QIAGEN OneStep RT-PCR kit. In addition to GSPs designed against the experimental target gene Sm-pal-1, GSPs designed against a 98 bp fragment of the S. mansoni alkaline phosphatase gene [31], were employed in RT-PCR reactions to enable the comparison of the relative expression levels of Sm-pal-1 and Sm-ap. Prior to use in RT-PCR reactions, all primer sets were used in RT-PCR optimisation experiments to ensure that they amplified specific gene products consistently. Each primer set (SmPAL and SmAP) was used in RT-PCR reactions on mRNA extracted from schistosomula treated with Sm-pal-1 dsRNA, Sm-phm-1 dsRNA (non-target control dsRNA) or buffer without dsRNA (negative control). Cycling conditions were as follows: 50 °C for 30 min, 95 °C for 15 min, followed by 40 cycles of 94 °C 30 sec, 50 °C 45 sec, and 72 °C for 2 min, and a final extension step of 72 °C for 10 min. RT-PCR products were visualised by agarose gel electrophoresis (as previously described), images were taken, and transcript abundance quantified relative to standardised DNA size markers using Quantity One 1-D Analysis software (BioRad).</p><!><p>Interrogation of the schistosome EST database with the Drosophila PAL1 sequence (GenBank accession number ACJ13179) identified two ESTs encoding a putative SmPAL (GenBank accession numbers AM047261 and AM043286). Subsequent computational analysis showed that these ESTs could represent a single SmPAL enzyme with significant similarity to the PAL domain of PAM proteins and to monofunctional PAL enzymes. Highest homology (identified via tBLASTn searches) was to the partial ORF of a putative Macrostomum lignano PAL (GenBank accession number EG951302, 34.0 % identity). Primers designed for PCR confirmation of the corresponding full length SmPAL open reading frame generated a 1236 nucleotide cDNA sequence (GenBank accession number FJ668386), encoding a 412 amino acid protein. SmPAL has a predicted molecular mass of 48 kDa, and includes an 18 amino acid (aa) signal peptide [38]. The coding sequence of SmPAL is flanked by a 333 bp 5′ untranslated region (UTR) and a 107 bp 3′ UTR with a polyadenylation signal (AATATA) 25 bp upstream of the poly(A)+ tail. The sequence of SmPAL suggests that adult schistosomes express a gene encoding an active monofunctional PAL enzyme.</p><p>A number of amino acids common to all eukaryotic PAL proteins and essential for catalysis are conserved [39–41] (Fig. 1). The six-bladed β-propeller structure of the rat PAL catalytic core (rat PALcc) includes a catalytic Zn2+ and a structural Ca2+. The three His residues that bind the catalytic Zn2+ (His585, His690, His786 in rat PALcc) are conserved, suggesting that SmPAL and rat PAL may bind Zn2+ in a similar manner. In rat PALcc, the Ca2+ is bound by two main chain carbonyls (Val520, Leu587) and a carboxylate (Asp787). While Leu587 is conserved, Val520 is replaced by Ala and Asp787 is replaced by Ser in SmPAL; although the interactions of Ca2+ with the main chain carbonyls would be conserved in SmPAL, its interaction with Ca2+ may be altered. The four cysteine residues that form disulfide bonds positioning key active site residues in rat PALcc are conserved in SmPAL, which has four additional cysteine residues not common to rat PALcc; three of these additional cysteine residues are located in what is predicted to form the sixth β-propeller of SmPAL. All of the key active site residues in rat PALcc (Arg533, Tyr654 and Arg706) are completely conserved [41].</p><!><p>In order to assess the catalytic activity of SmPAL, it was expressed transiently in mammalian cells. To ensure efficient secretion and detection, the signal sequence of SmPAL was replaced by the signal sequence of rat PAM; further, an epitope tag (rhodopsin) was appended to the C-terminus of SmPAL. Drosophila PAL (dPAL2) was previously analyzed in a similar manner [35] and was expressed at the same time for comparison. Spent medium and cell extracts were analysed by Western blot, revealing the presence of rhodopsin-tagged SmPAL in cell extracts and in spent medium (Fig. 2A); transient expression of EGFP served as a control. While the SmPAL and dPAL2 present in cell extracts were similar in mass (51 kDa), the SmPAL recovered from spent medium was 59 kDa larger than secreted dPAL2. The predicted mass of His6-SmPAL.rho is 48 kDa, the SmPAL sequence includes four potential N-linked glycosylation sites (N-X-S/T) (Fig. 1). The potential N-glycosylation sites closest to the N- and C-termini of SmPAL fall outside of the β-propeller structure observed for rat PALcc and, therefore, should be accessible to oligosaccharide transferase [41]; the remaining sites may also be accessible since they would be expected to occur in the loops connecting the β1/β2 strands of propeller 1 and the β3/β4 strands of propeller 2. The mass of the rhodopsin-tagged SmPAL protein in cell extracts (51 kDa) makes it unlikely that more than one or two N-linked oligosaccharide chains are present. The size increase observed upon secretion of SmPAL suggests the presence of additional modifications such as the addition of sialic acid residues to N-linked oligosaccharides or O-linked glycosylation.</p><p>Rhodopsin-tagged SmPAL and dPAL2 were both secreted; the amount of medium analyzed represents the amount secreted in approximately 5 h. Equivalent amounts of cell extract and spent medium were analyzed for dPAL2 and SmPAL. Since misfolded proteins are not efficiently secreted, this result indicates that SmPAL, like dPAL2, folds properly in mammalian cells. Although SmPAL and dPAL2 were expressed at similar levels (Fig. 2A), SmPAL was secreted less efficiently. In order to avoid any contribution from newly synthesized, immature enzyme, activity assays were carried out on spent medium. Cells transiently expressing EGFP were analyzed as a control and the low level of activity detected in their medium was subtracted as background. SmPAL enzyme activity was readily detected in the spent medium (Fig. 2A). The enzymatic activity of recombinant rhodopsin-tagged dPAL2 was previously characterised [35], and activity was readily detected in the spent medium. Since both proteins were detected using antibody to rhodopsin, their specific activities can be compared. Under the assay conditions used, the specific activity of SmPAL is substantially higher than that of dPAL2.</p><p>In mammalian cells, PAL must function in the acidic environment of the secretory granule. To assess the effect of assay pH on SmPAL activity, aliquots of spent medium from cells expressing SmPAL were diluted into buffers ranging in pH from 3.5 to 7.5. Optimal activity was detected at pH 4.5, with a second clear peak of SmPAL activity apparent at pH 6 (Figure 2B). No activity was detectable above pH 7.0. Based on the crystal structure of rat PALcc, a single residue contacts the amino acid that is to be amidated; the Met784 that serves this function in rat PALcc is replaced by Asp in SmPAL (Fig. 1). If titration of the side chain carboxylate affects the binding of substrate to SmPAL, this may explain the presence of two distinct pH optima.</p><!><p>Sm-pal-1 in situ hybridisation experiments were carried out on whole mount preparations of a mixed sex population of adult S. mansoni. Antisense probes were used as experimental probes to identify the distribution of Sm-pal-1 transcript, visualised by a purple/brown staining pattern, while sense probes were used in time- and condition-matched experiments as a negative control. Colour deposition could be visualised within 2 h of incubation with the BCIP/NBT substrate and continued to develop overnight.</p><p>Sm-pal-1 gene expression was localised to neuronal cell bodies in the CNS of adult schistosomes. Female schistosomes display a marked deposition of Sm-pal-1 staining in the cerebral ganglia and longitudinal nerve cords of the CNS in the anterior fore-body (Fig. 3A and B). Transcript localisation is notable in the longitudinal nerve cords surrounding the oral sucker, and running towards the posterior, parallel to the oesophagus and the bifurcated gut (Fig. 3A and B). The oesophagus and gut intestinal caecae appear dark brown in colour due to non-specific interaction of BCIP/NBT with endogenous phosphatses (also evident in sense-strand treated control worms, where no CNS staining was observed, see Fig. 3). The longitudinal nerve cords in the mid body of the female worm also demonstrate Sm-pal-1 expression, where they can be seen in close proximity to the uterus (Fig. 3C). Interestingly, the longitudinal nerve cords display Sm-pal-1 gene expression in a distinct, but increasingly more diffuse pattern than noted in the bilobed cerebral ganglia. This irregular staining pattern was also observed in the tail of male worms, where Sm-pal-1 gene expression is evident in several discrete accumulations of staining (Fig. 3D). There was no evidence of Sm-pal-1 gene expression in the PNS – likely due to the limited access of antibodies and staining reagents in the absence of flat-fixing.</p><p>Sense probes were employed in negative control experiments to identify any non-specific staining. Specimens showed no observable staining in control experiments. Note that dark brown staining of the oesophagus and intestinal caecae was observed in female negative control specimens as in experimental worms, demonstrating the non-specific staining of these elements by the BCIP/NBT substrate; no CNS staining was evident (Fig. 3E).</p><!><p>Following SmPAL dsRNA delivery by electroporation, and 2 day post-dsRNA delivery schistosomula culture, RT-PCRs revealed no measurable changes in Sm-pal-1 transcript levels (data not shown). dsRNA delivery by electroporation was attempted twice with two separate batches of dsRNA, and on both occasions was unsuccessful, thus longer-term soaking as a means of dsRNA delivery was explored.</p><p>In total, six soaking experiments were carried out, of which only two experiments revealed diminished transcript abundance of Sm-pal-1. In experiments where transcript knock-down was detected, dsRNA delivery by soaking induced a marked and comparable reduction in Sm-pal-1 transcript levels as measured by endpoint detection (71–90 %; Fig. 4). Levels of Sm-pal-1 were unaffected in Sm-phm-1 dsRNA treated worms (Fig. 4). The alkaline phosphatase control gene Sm-ap showed no measurable difference in expression levels in dsRNA treated worms (Fig. 4). These observations suggest a specific reduction in the quantity of Sm-pal-1 transcript. All RT-PCR no template controls were negative.</p><p>On each of four further attempts to diminish Sm-pal-1 transcript levels in schistosomula, RT-PCR did not show any measurable difference in transcript abundance. In all experiments, whether or not Sm-pal-1 transcript levels were detectably reduced, schistosomula appeared normal and did not display any observable aberrant phenotypes from untreated control experiments.</p><!><p>Carboxy-terminal α-amidation is an essential post-translational modification in the generation of biologically active neuropeptides. Various studies, involving the amidating enzymes of eukaryotes, report emerging disparities in amidating enzyme gene organisation [see 24 for review]. For the most part, PHM and PAL are expressed as separate domains of a bifunctional PAM protein in higher organisms such as Rattus norvegicus [42], and Xenopus laevis [43]. In contrast, an increasingly complicated picture of PHM and PAL expression is emerging from invertebrates, as a number of species including Lymnaea stagnalis [44], Aplysia californica [45], Calliactis parasitica [46], Drosophila [25, 35], Caenorhabditis elegans [47], and the planarian Dugesia japonica [26] encode single or multiple copies of one or both enzymes, expressed as bifunctional or monofunctional proteins.</p><p>PHM has previously been characterised from S. mansoni and identified as a monofunctional enzyme [28]. In the same study the use of degenerate primers and BLAST analysis of the GenBank EST dataset, failed to identify an S. mansoni PAL cDNA [28]. Capitalizing on more recent S. mansoni EST datasets, we identified a cDNA encoding a monofunctional schistosome PAL. Interestingly, extended interrogation of schistosome EST and genome datasets confirmed the absence of an EST representing a bifunctional PAM enzyme in S. mansoni.</p><p>Monofunctional SmPAL is a soluble, secreted enzyme, similar in sequence to other PALs such as the secreted Drosophila enzyme, dPAL2 [35], retaining four conserved cysteines for the maintenance of secondary structure, plus four cysteine residues unique to SmPAL, and a tyrosine residue known to be essential for functionality [39, 40]. Yet there are key amino acid substitutions in SmPAL, in the Ca2+ binding site, and where the amino acid preceding the α-hydroxyglycine and the PAL active site interact, which set it apart from the host enzyme. The latter of the two appears particularly significant as it is the Met784 in rat PALcc, substituted by Asp in SmPAL (Fig. 1), which recognizes the carbonyl group of the peptide bond linking the amino acid to be amidated to α-hydroxyglycine [41].</p><p>Rat PAL is completely N-glycosylated at a single consensus site located in the loop connecting β3/β4 in propeller five [48]. Although it is unclear what dictates the efficiency with which N-glycosylation sites are used [49], it is probable that only two of the four candidate N-linked glycosylation sites within SmPAL are glycosylated. The most N- and most C-terminal sites are not included in the β-propeller structure; while both sites may be accessible, the termination of translation may limit N-glycosylation of the most C-terminal site. The other two possible sites are in loops and could also be used (Figure 1). The size increase observed upon secretion of SmPAL may be a result of further post-translational modifications such as the addition of sialic acid residues or O-linked glycosylation, which occur at a later stage in protein processing [49].</p><p>The conversion of α-N-acetyl-Tyr-Val-α-hydroxyglycine, to α-N-acetyl-Tyr-Val-α-NH2, a reaction mechanism specific to PAL enzymes, revealed SmPAL catalytic properties within the range of other eukaryote enzymes. The pH optimum of SmPAL coincides with the reported characteristics of other PALs [35, 44, 50, 51], and as expected, is less acidic than the pH optimum of SmPHM [28]. The low pH optimum and high Km (44 μM) of SmPHM, which deviate from the catalytic properties of the human host enzyme, make SmPHM a potential novel drug target [28]. Although the functional characteristics of SmPAL revealed in this study do not display the same degree of divergence from the schistosome host enzyme as SmPHM [51], it is the degree of structural divergence evident from the crystallography of rat PALcc [41] which may provide opportunity for chemotherapeutic intervention. Regardless, the expression of a catalytically active monofunctional SmPAL, unequivocally confirms the existence of monofunctional amidating enzymes in schistosomes which act independently of each other, in contrast to the mammalian host enzyme.</p><p>Although SmPAL transcript localisation is consistent with the immunolocalisation patterns of the amidating co-enzyme SmPHM in the schistosome CNS, SmPHM is also localised extensively in the PNS of adult schistosomes [28]. Although currently there is no molecular or bioinformatic evidence supporting the possibility of a second SmPAL enzyme, the enzyme complement of other eukaryotes [24], and the restricted Sm-pal-1 expression pattern presented here, does not totally diminish the possibility of another active schistosome amidating lyase. One possible reason for the absence of SmPAL staining in the PNS is that transcript abundance was below our detection limit. Another reason could be that most of the PNS cell bodies reside close to the CNS such that transcript does not physically associate with much of the PNS and as such is identified as CNS staining by in situ hybridisation. Clearly, more reliable comparisons could be made with immunocytochemical approaches using a SmPAL antiserum.</p><p>The expression of Sm-pal-1 in the CNS of adult schistosomes, coupled with the extensive immunolocalisation of SmPHM throughout the nervous system of adult schistosomes [28] is consistent with the hypothesis that SmPAL and SmPHM are neuronally expressed enzymes, involved in neuropeptide biosynthesis. The widespread localisation of neuropeptide amidating enzymes in the schistosome nervous system is supported by a recent study which exploited available transcriptomic and genomic datasets for flatworms, including S. mansoni, and reported seventeen novel amidated neuropeptides from S. mansoni [13]. A transcript encoding schistosome NPFF (PQRFamide motif), a peptide similar to vertebrate neuropeptide FF and not previously reported in invertebrates, was identified, along with several RFamides, an additional NPF, and novel L/Iamides and PWamides [13]. Thus our recent bioinformatic data go some way to support and explain the localisation pattern of sm-pal-1 and the extensive staining patterns reported for SmPHM.</p><p>RNAi has quite recently become popular for the functional analysis of schistosome genes and a growing number of publications report successful gene silencing, of non-neuronal transcripts, in larval parasites following dsRNA/siRNA delivery by electroporation and soaking [31, 37, 52–64]. Our attempts to knockdown schistosomula Sm-pal-1 gene transcripts were somewhat less successful than previous publications of schistosomula gene knockdown. Importantly however, none of the genes previously targeted in schistosomes have been exclusively involved in neuronal processes, and are most commonly genes such as the cathepsins B and D, and alkaline phosphatase, which are highly expressed in tissues relatively more accessible to external dsRNA than neuronal cells [31, 37, 53, 57, 60].</p><p>In the studies mentioned above, electroporation has been reported to induce >85 % knockdown of some non-neuronal transcripts in schistosomula [31]. However, in this study electroporation did not induce measurable gene knockdown of our PAL transcript on either of two attempts. Some reports detail concentration dependant responses to dsRNA [55, 57] such that it is likely, in the case of Sm-pal-1, that future electroporation experiments may benefit from the optimisation of dsRNA concentration to produce measurable and consistent gene silencing.</p><p>DsRNA delivery by electroporation is reported to produce much more dramatic gene suppression (>10 fold) than dsRNA delivery by soaking [31, 57]. In this study however, soaking did successfully and specifically reduce Sm-pal-1 transcript abundance by up to 90 % on two occasions out of six. While these results do indicate that some, if not all, schistosome neuronal genes are susceptible to RNAi, the question remains as to why neuronal gene silencing in this case does not appear to be consistently reproducible. Tissue types are known to differ in their degree of penetrability by dsRNA constructs, resulting in varying degrees of gene suppression between different tissues [65]. This may present one possible reason for Sm-pal-1 knockdown unreliability, as dsRNA constructs may have limited access to deeper tissues such as neuronal cell bodies, where target transcripts are expressed.</p><p>Further, dsRNA delivery by soaking may have proven more successful if larval parasites were exposed to dsRNA during cercarial transformation. It is thought that schistosome intra-molluscan life stages are more susceptible to RNAi during transformation when tegumental membranes are undergoing reorganisation [52]; some reports suggest that this may also be the case during cercarial transformation [53]. In contrast, other studies involving dsRNA delivery during cercarial transformation disagree, suggesting that dsRNA uptake is no more efficient when external membranes are in flux, than when schistosomula are fully formed [57]. The exploration of alternative dsRNA delivery methods, or the application of smaller siRNAs may prove beneficial to the induction of neuronal RNAi in schistosomes.</p><p>The recent discovery of a rich flatworm neuropeptide complement provides strong support for the importance of neuropeptide signaling in flatworm neurobiology. One particularly appealing feature of these discoveries is the fact that the majority of putative neuropeptides discovered in the flatworm datasets are novel peptides with no obvious homologues in other animal phyla. Clearly, such differences are a preferred feature of parasite drug targets and underscore the possibility of exploiting flatworm neuropeptide signaling processes for parasite control. The mechanisms and processes associated with neuropeptide signaling such as peptide amidation represent attractive targets for chemotherapeutic intervention. The presence of a widely expressed PHM enzyme in the nervous system of adult schistosomes [28], and the identification by this study of a functionally active, monofunctional schistosome PAL enzyme provides target enzymes for consideration as drug target candidates. Unfortunately, RNAi has yet to be optimised for neuronally-expressed transcripts in schistosomes such that validating neuronal gene products as drug targets remains a challenge. Although we only observed Sm-pal-1 RNAi knockdown intermittently, the fact that knockdown occurred on a couple of occasions gives hope that further optimization of RNAi methods could facilitate robust neuronal RNAi.</p><!><p>ClustalW alignment of schistosome PAL (SmPAL) with the catalytic core of other eukaryotic PAL domains/proteins. Text boxed in yellow denotes perfectly conserved residues, text boxed in blue denotes partially conserved residues. Four conserved cysteine residues are marked by black circles, while four additional SmPAL specific cysteine residues are highlighted in red. Three conserved His residues involved in binding Zn2+ are marked by black triangles. Residues in rat PALcc that bind Ca2+ (Val520, Leu587, Asp787) are not perfectly conserved and are marked by black boxes. Conserved key active site residues in rat PALcc (Arg533, Tyr654, Arg706) are marked by asterisks. Rat PALcc Met784, which interacts with the peptide substrate, and is replaced by Asp in SmPAL is marked by a black rectangle. Four possible SmPAL N-glycosylation sites are highlighted in magenta. GenBank accession numbers: Rat PAL (RatPALcc, P14925), Human PAL (HsPALcc, AAA36414), Drosophila melanogaster PAL 1 (dPAL1cc, ACJ13179), Drosophila melanogaster PAL 2 (dPAL2cc, AAF47043), Lymnaea stagnalis PAL (LsPALcc, AAD42259), and Caenorhabditis elegans (CePALcc, 172108).</p><p>In vitro expression of SmPAL. Western blot analysis of the steady-state distribution of transfected EGFP (control), dPAL2 and SmPAL. Cell extracts (C; 4% of total amount cell extract) and spent medium (M; 0.8% of total amount spent medium) are shown; monoclonal antibody to rhodopsin was used to visualize the transiently expressed proteins. PAL activity was quantified in aliquots of spent medium; the total amount of activity loaded into the lanes containing medium is indicated below lanes M4 and M6. The apparent molecular masses of SmPAL and dPAL2 were compared; Western blot analysis yielded an apparent molecular mass of 51 kDa for cellular SmPAL (C5) and 59 kDa for secreted SmPAL (M6), dPAL2 is shown for comparison (C3 and M4). (B) Graph showing analysis of the pH dependence of SmPAL activity, indicating optimal activity at pH 4.5, with an additional peak of activity at pH 6.</p><p>Light microscopy images of whole mount in situ hybridisation in adult Schistosoma mansoni showing Sm-pal-1 gene expression. (A) A female S. mansoni showing Sm-pal-1 localisation in the cerebral ganglia (CG) and longitudinal nerve cords (LNC). The muscular oral sucker (OS) and acetabulum (AC) are also labelled. (B) A female worm with Sm-pal-1 gene expression in the cerebral ganglia (asterisk, *) and LNC running in an anterior and posterior direction. The oesophagus (OE) and intestinal caecae (IC) are also shown. (C) A female specimen showing Sm-pal-1 gene expression in the LNCs running parallel to the uterus (U). (D) The tail of a male S. mansoni showing Sm-pal-1 transcript distribution (arrow heads) close to the posterior end of the gynaecophoric canal (GC). (E) A S. mansoni negative control, treated with Sm-pal-1 sense probes. An adult male and female pair can be seen to lack Sm-pal-1 staining. The oral sucker (OS), acetabulum (AC) and intestinal caecae (IC) of the female worm are shown, with the OS of the male positioned below the female anterior fore-body. Scale bars: A=60 μm; B=20 μm; C=40 μm; D=27 μm; E= 50 μm.</p><p>The effects of RNA interference on the expression of Schistosoma mansoni PAL (Sm-pal-1) and S. mansoni alkaline phosphatase (Sm-ap) in larval schistosomula. Representative RT-PCR results showing marked reduction in Sm-pal-1 transcript levels (~71 %, lane 1) following treatment with double stranded (ds) RNA for Sm-pal-1. Transcript levels of Sm-ap are unaffected (lanes 5, 6 and 7). (Lanes: 1, SmPAL-dsRNA treated; 2, SmPHM-dsRNA treated; 3, untreated (− dsRNA) control; 4, no template control; 5, SmPAL-dsRNA treated; 6, SmPHM-dsRNA treated; 7, - dsRNA control; 8, no template control).</p><p>Primer sequences</p><p>T7 RNA polymerase promoter sequences are underlined</p>

PubMed Author Manuscript

A photo zipper locked DNA nanomachine with an internal standard for precise miRNA imaging in living cells

DNA nanomachines are capable of converting tiny triggers into autonomous accelerated cascade hybridization reactions and they have been used as a signal amplification strategy for intracellular imaging. However, the "always active" property of most DNA nanomachines with an "absolute intensitydependent" signal acquisition mode results in "false positive signal amplification" by extracellular analytes and impairs detection accuracy. Here we design a photo zipper locked miRNA responsive DNA nanomachine (PZ-DNA nanomachine) based on upconversion nanoparticles (UCNPs) with a photocleavable DNA strand to block the miRNA recognition region, which provided sufficient protection to the DNA nanomachine against nonspecific extracellular activation and allowed satisfactory signal amplification for sensitive miRNA imaging after intracellular photoactivation. Multiple emissions from the UCNPs were also utilized as an internal standard to self-calibrate the intracellular miRNA responsive fluorescence signal. The presented PZ-DNA nanomachine demonstrated the sensitive imaging of intracellular miRNA from different cell lines, which resulted in good accordance with qRT-PCR measurements, providing a universal platform for precise imaging in living cells with high spatialtemporal specificity.

a_photo_zipper_locked_dna_nanomachine_with_an_internal_standard_for_precise_mirna_imaging_in_living_

Introduction<!>Principle of the PZ-DNA nanomachine with an internal standard for miRNA imaging<!>Conclusions<!>Conflicts of interest

<p>DNA nanomachines are articially designed DNA self-assembly structures based on sequence specic interactions. [1][2][3][4] With DNA reaction strands compacted in the synthesized nano-structures, [5][6][7][8] DNA nanomachines accelerate cascade hybridization reactions by converting tiny triggers like nucleic acids, 9-12 proteins 5,13 and pH [14][15][16] into autonomous mechanical motions and resulting outputs, such as the conformational change of DNA assemblies. Self-quenched DNA nanomachines have been constructed by conjugating both uorescent molecules and quenchers to DNA strands with close distance in between, [17][18][19] and the machine motion triggered by a specic target resulted in the continuous conguration change of the DNA strands with amplied uorescence recovery in a short time, which efficiently enhanced the target signal and was applied as a signal amplication strategy for bioanalysis and imaging. [20][21][22][23] Considering the low expression level of miRNA per cell 18 and the complex intracellular environment, 3,[24][25][26] DNA nanomachines have been used as valuable uorescence signal amplication tools for intracellular miRNA imaging. 17,[27][28][29][30] Conjugating with a DNAzyme, a DNA walking machine was activated by target miRNA and automatically cleaved substrate DNA strands in a continuous fashion for sensitive intracellular imaging. [31][32][33] Though being capable of amplifying the existence of a low amount miRNA to an appreciable uorescence signal in a short time, the further application of DNA nanomachines for precise intracellular miRNA imaging still faces the challenges of: (1) the "always active" design of most DNA nanomachines makes them susceptible to extracellular target miRNA in the tumor microenvironment or serum, which would cause nonspecic uorescence signal amplication before intracellular delivery and result in a false positive signal; 18,[33][34][35][36][37][38] (2) the uptake efficiency variation for different cells would also contribute to intracellular signal differences due to the "absolute intensity-dependent" mode with a single luminance channel for signal acquisition. [39][40][41] Both of these factors impair the detection accuracy. DNA nanomachines with controllable activation and an internal standard are needed for precise miRNA intracellular imaging. Taking advantage of multiple luminescence emissions from upconversion nanoparticles (UCNPs) under near infrared (NIR) light excitation, [42][43][44] here we present a photo zipper locked DNA nanomachine (PZ-DNA nanomachine) with an internal standard for precise intracellular miRNA imaging. The PZ-DNA nanomachine was composed of UCNPs with surface functionalization of a target miRNA responsive DNA walker and its corresponding substrate DNA strands labelled with quencher BHQ2. Dye Cy3 was also co-immobilized on the UCNP surface to facilitate energy transfer from the interior UCNP emission at 540 nm under NIR irradiation to BHQ2 labelled at the terminus of the substrate DNA strands. To protect the DNA nanomachine against unwanted activation before it is located inside tumor cells, a UV cleavable DNA strand was used to block the recognition of the DNA nanomachine by the extracellular target miRNA. Aer internalization and subsequent photo-activation, the PZ-DNA nanomachine was operated by intracellular miRNA and continuously cleaved the BHQ2 labelled substrate DNA strands, with corresponding Cy3 uorescence recovery for intracellular miRNA imaging. The concentration of the imaging dye Cy3 on the UCNP surface beneted intracellular imaging. It not only contributed to high signal output compared with cytoplasm dispersed imaging dyes, but also realized uorescence emission upon NIR irradiation. The UCNP emission at 658 nm under NIR light excitation remained stable during the PZ-DNA nanomachine operation process, which served as the internal standard for self-correction of the Cy3 uorescence to improve the detection accuracy. This PZ-DNA nanomachine will provide a versatile strategy for precise intracellular imaging of low amounts of biomarkers.</p><!><p>To prepare the PZ-DNA nanomachine, core-shell structured UCNPs NaYF 4 :Yb,Er,Gd@NaYF 4 were synthesized, ligand exchanged with alendronic acid (ADA) to functionalize the surface with amine groups, and covalently coupled with a photo zipper locked DNAzyme walker (P-DNA walker) as well as its substrate DNA strand with the quencher BHQ2 labelled at the terminus (S-DNA-BHQ2). To prepare the P-DNA walker, the DNAzyme walker was hybridized with the "photo zipper" DNA strand to block the catalytic reaction of the DNAzyme towards S-DNA-BHQ2, and equipped with a 40 nt polyT spacer strand for walking. The photo zipper was composed of three regions: a 6 nt DNAzyme block region (c) that hybridized with the DNAzyme anchor position; a 22 nt miRNA responsive region (a* + b) with complementary sequence to the target miRNA; and a 15 nt miRNA block region (a + h) that partially hybridized with the miRNA responsive region. Pc-linker, a UV light cleavable molecule as the photocleavable linker, was embedded between h and a*, which locked the recognition of the miRNA responsive region (a* + b) to miRNA in the absence of light to protect the P-DNA walker from extracellular nonspecic activation during its delivery process. Dye Cy3 was also coimmobilized on the UCNP surface, which bridged the two continuous luminance resonance energy transfer (LRET) processes of concentrating the UCNP 540 nm luminance under NIR light irradiation from the interior of the nanoparticle to the UCNP surface (intra-LRET) and the subsequent quenching of the Cy3 luminance by BHQ2 at the DNA strand S-DNA-BHQ2 terminus (external LRET) (Scheme 1a). Aer internalization of the PZ-DNA nanomachine, the miRNA block region of the photo zipper (a + h) was cleaved off via UV irradiation to activate the miRNA responsive region (a* + b), and the corresponding hybridized intracellular target miRNA zippered off the photo zipper to activate the DNAzyme walker for S-DNA-BHQ2 cleavage in the presence of Mn 2+ . Aer the cleavage of S-DNA-BHQ2, the DNAzyme walker was liberated because of insufficient binding base pairs and moved to the next S-DNA-BHQ2. The successive cleavage of S-DNA-BHQ2 by the autonomous walking of the DNA nanomachine resulted in continuous release of BHQ2-containing DNA fragments from the UCNP surface with Cy3 uorescence recovery at 580 nm under NIR light irradiation for precisely amplied intracellular miRNA imaging. The operation of the PZ-DNA nanomachine only utilized UCNP emission at 540 nm under NIR light irradiation, while another UCNP emission at 658 nm wasn't affected, therefore serving as an internal standard, and the intensity of Cy3 luminance at 580 nm was ratioed over the UCNP luminance at 658 nm for miRNA imaging self-correction (Scheme 1b).</p><p>Preparation and characterization of the PZ-DNA nanomachine ß-NaYF 4 crystalline structures co-doped with Er 3+ (2%), Yb 3+ (18%) and Gd 3+ (10%) were synthesized according to a previously reported solvothermal method 45 and coated with NaYF 4 to prevent surface quenching and enhance the upconversion luminance. Gd 3+ was doped to control the size and shape of the UCNPs. 46 Compared with UCNPs NaYF 4 :Er,Yb,Gd, the asprepared core-shell structured UCNPs NaYF 4 :Er,Yb,Gd@NaYF 4 Scheme 1 Schematic illustrations of (a) the preparation and (b) activation of the photo zipper locked DNA nanomachine, as well as its operation in response to miRNA in living cells.</p><p>exhibited increased particle size from 20.3 AE 0.7 nm (Fig. S1a †) to 22.6 AE 1.1 nm (Fig. 1a) and showed much enhanced luminance intensities for the emission peaks at 520 nm, 540 nm, and 658 nm corresponding to the 2 1b). Oleic acid (OA), the surface ligand of the UCNPs, was replaced with ADA, a bisphosphine ligand with amine terminal groups due to the high binding affinity between bidentate phosphates and lanthanide ions, which improved the dispersion of the UCNPs in aqueous solution and facilitated the subsequent functionalization. The successful surface ligand exchange process was conrmed by FTIR, which showed the IR absorption of OA at 2920 cm À1 and 2850 cm À1 for the OA functionalized UCNPs (UCNPs-OA) due to the asymmetric and symmetric stretching vibrations of the -CH 2group, respectively (line 1, Fig. 1c). Aer ligand exchange with ADA, the characteristic peaks for the -CH 2group were strongly suppressed, corresponding with the appearance of an absorption peak at 1050 cm À1 attributed to the stretching vibration of P-O for ADA (line 2, Fig. 1c). The asprepared ADA stabilized UCNPs (UCNPs-ADA) exhibited an increased zeta potential of 35.7 AE 1.2 mV compared with the bare UCNPs and a hydrodynamic size of 28.4 AE 2.3 nm similar to that of the bare UCNPs (Fig. 1d).</p><p>Cy3, with its maximum absorption well overlapped with the UCNP emission at 540 nm (Fig. 1e), was chosen as the bridge dye connecting two continuous LRET processes and covalently immobilized on the UCNP surface via amidation to facilitate luminance energy transfer from the interior of the UCNPs to the surface extended DNA strands that were subsequently modied. NHS-PEG-Mal was co-immobilized on the UCNP surface with Cy3 to facilitate subsequent DNA strand immobilization, and the molar ratio of Cy3 over NHS-PEG-Mal was optimized to maximize the UCNP energy transfer efficiency to Cy3, which achieved the highest Cy3 luminance intensity at 580 nm with Cy3/NHS-PEG-Mal of 1 : 7 (Fig. S1b and c †). To quantify the amount of Cy3 immobilized on each UCNP, Cy3 uorescence from UCNPs-Cy3/PEG-Mal was measured at 545 nm and compared with the Cy3 standard calibration curve to get 170 AE 16 Cy3 per UCNP (Fig. S1d and e †). The efficient energy transfer from the interior of the UCNPs to Cy3 was conrmed by the obvious uorescence intensity decrease at 520 nm and 540 nm and increase at 580 nm (Fig. S1b † and 1f). The successful synthesis of UCNPs-Cy3/PEG-Mal was also conrmed by the characteristic IR absorption peaks at 1108 cm À1 and 1170 cm À1 attributed to the stretching and asymmetric stretching vibrations of C-O-C groups, respectively (line3, Fig. 1c), and the decrease of the zeta potential to 12.9 AE 0.6 mV due to PEG conjugation with an increase in hydrodynamic size to 33.6 AE 2.3 nm (Fig. 1d).</p><p>Before modication of DNA strands on the UCNP surface, the performance of the catalytic reaction for the DNAzyme walker in solution was veried via gel electrophoresis with a polyT linked DNAzyme walker and substrate DNA (S-DNA). In the presence of Mn 2+ , S-DNA was cleaved into two fragments (F1, F2) by the DNAzyme walker (Fig. S2a †), and gave two distinct bands at their individual corresponding positions in the gel electrophoresis images (lane 5, Fig. S2b †). A thiol terminus P-DNA walker was then modied on UCNPs-Cy3/PEG-Mal with S-DNA-BHQ2 to complete the PZ-DNA nanomachine, which showed DNA characteristic absorption peaks at 260 nm in the UV-Vis absorption spectrum (Fig. S1f †) and 1430 cm À1 in the FTIR spectra with a series of peaks below 900 cm À1 attributed to the bending vibrations of the heterocycle in DNA 48 (line 4, Fig. 1c). The resulting PZ-DNA nanomachine demonstrated decreased zeta potential of À14.0 AE 1.2 mV due to the negative charges of the DNA strands with a highly increased hydrodynamic size of 57.6 AE 3.1 nm (Fig. 1d).</p><p>The maximum absorption of BHQ2 was located at 572 nm, overlapped with the Cy3 emission peak at 585 nm (Fig. 1e) and the PZ-DNA nanomachine with S-DNA-BHQ2 functionalization showed a complete suppression of Cy3 luminance compared with UCNPs-Cy3/PEG-Mal, while the UCNP luminance at 658 nm remained unchanged (Fig. 1f), indicating the efficient energy transfer from UCNP luminance at 540 nm to the S-DNA-BHQ2 terminus BHQ2 via bridge dye Cy3, which facilitated two continuous LRET processes of intra-LRET from the interior of the UCNPs to the UCNP surface and external LRET from the UCNP surface to BHQ2. In comparison, considering the inefficient long distance energy transfer from the interior of the UCNPs to the surface extended DNA strand, there was little miRNAs play important roles in various cellular processes and the progression of many diseases, therefore, it is signicant to determine their expression levels in living cells. As a potential diagnostic biomarker, 49 miRNA 21 is critical to a lot of biological functions and disease progressions, and was chosen here as the model target to trigger the operation of the PZ-DNA nanomachine. 28 To prepare the P-DNA walker, a photo-cleavable hairpin structured photo zipper was hybridized with the DNAzyme walker to block its anchoring region with S-DNA-BHQ2.</p><p>The feasibility of photo-responsive activation of the P-DNA walker was rst veried in homogeneous solution. 5 min UV (7 mW cm À2 ) exposure cleaved off the miRNA block region a + h, and generated a cleaved DNA walker (C-DNA walker) with partially exposed target miRNA 21 recognition region a* + b (Fig. 2a). The efficient photo cleavage of the miRNA block region was conrmed by the complete disappearance of the P-DNA walker band, accompanied by the clear appearance of the C-DNA walker band with higher mobility when the P-DNA walker was exposed to UV light (lane 2, Fig. 2b). The dropped fragment (a + h) with short strand length did not appear in the gel electrophoresis images. The subsequent miRNA 21 recognition completely unzipped the C-DNA walker and liberated the DNAzyme for the enzymatic catalytic reaction (Fig. 2a). Incubation of the P-DNA walker with miRNA 21 aer photo activation resulted in two new bands with much higher mobility, which were at the same positions of the DNAzyme walker band and hybridized duplex miRNA 21/AL band, respectively (lane 4, Fig. 2b), indicating the successful activation of the P-DNA walker in the presence of miRNA 21 aer photo irradiation. On the contrary, miRNA 21 didn't react with the P-DNA walker in the absence of photo irradiation, and individual miRNA 21 and P-DNA walker bands were observed in gel electrophoresis (lane 3, Fig. 2b). S-DNA-BHQ2 and the P-DNA walker were then conjugated to UCNPs-Cy3/PEG-Mal with a molar ratio of 10 : 1 to prepare the PZ-DNA nanomachine, and the intensity ratio of Cy3 uorescence recovery over UCNP luminance at 658 nm (I 580 /U 658 ) was measured according to time to conrm the photo activation of the PZ-DNA nanomachine and its response to miRNA 21. With UV exposure, the mixture of 2 nM miRNA 21 with the PZ-DNA nanomachine resulted in an extensive increase of I 580 /U 658 according to time and the signal increase quickly saturated at 80 min, indicating the high reaction efficiency of the PZ-DNA nanomachine, while little increase of I 580 /U 658 was observed either in the absence of light irradiation or miRNA 21 (Fig. 2c). Aer photo activation, the PZ-DNA nanomachine was challenged with miRNA 21 of different concentrations from 0 to 10 nM, which resulted in a gradual increase of the Cy3 uorescence peak at 580 nm, while the UCNP luminance at 658 nm remained unchanged (Fig. 2d). A linear correlation between I 580 / U 658 and miRNA 21 concentration was obtained in the range from 0.05 to 3.00 nM with a limit of detection (LOD) of 3.71 pM (Fig. 2e), which is much improved compared with many previous reported UCNP-based miRNA detection probes. 50 The introduction of the DNA nanomachine on the UCNP surface contributed to impressive enhancement of the detection sensitivity. 23,51,52 To further demonstrate the amplication effect of the PZ-DNA nanomachine, a FAM labelled P-DNA walker was prepared by conjugating the dye FAM to the photo zipper, and immobilized on the UCNP surface with S-DNA at a molar ratio of 1 : 10 to prepare a PF-DNA nanomachine (Fig. S4a †). Aer photo exposure, the operation of the PF-DNA nanomachine in response to 2 nM miRNA 21 released hybridized duplex miRNA 21/AL-FAM, and the FAM uorescence intensity from the supernatant was measured. Since FAM dye was labelled on the P-DNA walker instead of S-DNA, the operation of the PF-DNA nanomachine with continuous cleavage of S-DNA did not contribute to uorescence amplication, and only resulted in very limited FAM uorescence intensity aer the reaction with miRNA 21 (Fig. S4b †). In comparison, the SF-DNA nanomachine with FAM labelled S-DNA (Fig. S4a †) was also prepared with S-DNA-FAM and the P-DNA walker immobilized on the UCNPs at a molar ratio of 10 : 1. In response to photo exposure and 2 nM miRNA 21, the SF-DNA walker continuously cleaved the FAM labelled F1 segment (F1-FAM) off the UCNP surface, and resulted in a substantial increase in the FAM uorescence recovery compared with the PF-DNA nanomachine due to the catalytic amplication reaction of the DNAzyme walker (Fig. S4c †).</p><p>The reaction specicity of the PZ-DNA nanomachine to miRNA 21 was also investigated with nonspecic miRNAs of three base mismatch, single base mismatch, miRNA 141 and miRNA 199-a (Fig. S4d †). The three base mismatched miRNA and nonspecic miRNAs all showed very low ratio of luminance intensities for Cy3 at 580 nm over UCNPs at 658 nm (I 580 /U 658 ) close to the blank control, indicating impressive reaction specicity of the PZ-DNA nanomachine in response to miRNA 21. The I 580 /U 658 for single base mismatched miRNA 21 was slightly higher due to approximate thermodynamic energy with miRNA 21. 11 Live cell imaging of miRNA 21 with the PZ-DNA nanomachine HeLa cells, with high miRNA 21 expression prole, were chosen as the model cells to demonstrate the feasibility of the PZ-DNA nanomachine for in situ visualization of intracellular miRNA.</p><p>Aer internalization, the PZ-DNA nanomachine accumulated inside the endosomes at rst and then escaped from the endosomes. The PZ-DNA nanomachine in the absence of BHQ2 was incubated with HeLa cells to demonstrate the lysosome escape process, which demonstrated progressive separation of Cy3 uorescence (red) and Lysotracker uorescence (green) in the cytoplasm (Fig. S5 †). Considering the steric hindrance from the nanomaterial to prevent the access of nucleases and protect the nanomaterial-assembled nucleic acids from degradation, [53][54][55] the PZ-DNA nanomachine retained its integrity in the lysosome, which was indicated by the complete overlap of the DNA conjugated Cy5 uorescence (blue) and UCNP uorescence (red) (Fig. S6 †). Aer the lysosome escape process, the HeLa cells were then exposed to UV irradiation (7 mW cm À2 ) for 5 min to activate the photo zipper, and the de-protected PZ-DNA nanomachine was subsequently activated by intracellular miRNA 21. Considering the low level of Mn 2+ in native cells, the required Mn 2+ concentration for the intracellular operation of the PZ-DNA nanomachine was achieved by incubating the cells with Mn 2+ solution. 33,56 Intracellular Cy3 uorescence recovery began show up in the TCS SP5 confocal laser scanning microscope (CLSM) at 0.5 h, and the intensity of Cy3 uorescence recovery gradually increased with incubation time and saturated at 2 h (Fig. S7 †). Instead of dropping from the nanocarrier and diffusing in the cytoplasm aer the DNAzyme catalytic reaction, the imaging dyes were concentrated on the UCNP surface aer reacting with the target miRNA. This improved the signal output and allowed efficient emission upon NIR irradiation. To conrm the complete protection of the PZ-DNA nanomachine with the photo zipper, as well as its efficient unzipping intracellularly, HeLa cells were incubated with the PZ-DNA nanomachine (functionalized with P-DNA walker) and an unprotected DNA nanomachine (functionalized with C-DNA walker) respectively, and there was little Cy3 uorescence recovery from the PZ-DNA nanomachine incubated HeLa cells in the absence of UV irradiation, indicating good protection from the photo zipper (Fig. S8a †). Aer 5 min of UV activation, the activated PZ-DNA nanomachine incubated HeLa cells demonstrated similar intensity of intracellular Cy3 uorescence recovery compared with that of the unprotected DNA nanomachine incubated HeLa cells (Fig. S8b and c †). To demonstrate photo zipper protection of the PZ-DNA nanomachine against unwanted extracellular activation, a miRNA 21 inhibitor, a synthetic 22 nt RNA strand with complementary sequence to miRNA 21, was used to treat HeLa cells to suppress intracellular miRNA 21 expression. The unprotected DNA nanomachine (functionalized with C-DNA walker) and PZ-DNA nanomachine (functionalized with P-DNA walker) were pretreated with 2 nM miRNA 21 for 9 h in vitro to simulate the nonspecic activation during the systemic circulation process, and were subsequently incubated with inhibitor treated HeLa cells. Strong UCNP luminance at 658 nm was observed from all the cells, indicating the efficient intracellular delivery of the PZ-DNA nanomachine and unprotected DNA nanomachine. Cy3 uorescence recovery was barely observed from both the unprotected DNA nanomachine (Fig. 3a) and PZ-DNA nanomachine (Fig. 3d) incubated HeLa cells due to the suppressed expression of miRNA intracellularly, while obvious Cy3 uorescence was observed from the HeLa cells incubated with a miRNA 21 pretreated unprotected DNA nanomachine (Fig. 3b); the "false positive" uorescence signal came from the activation of the unprotected DNA nanomachine before endocytosis. In contrast, the miRNA 21 pre-treated PZ-DNA nanomachine demonstrated little uorescence recovery intracellularly (Fig. 3c), indicating the effective protection of the DNA nanomachine from the photo zipper against nonspecic activation and the impressive suppression of the background uorescence signal.</p><p>The expression levels of identical miRNA change at different periods of tumorigenesis, 57 therefore the accurate imaging of the miRNA expression level in living cells is of great signicance in disease diagnosis and evaluation of therapeutic effects. The potential of the PZ-DNA nanomachine for quantitative evaluation of the relative expression levels of miRNA in living cells was further conrmed. miRNA 21 mimic, a synthetic RNA mimicking miRNA 21, and miRNA 21 inhibitor were incubated with HeLa cells separately to adjust the intracellular miRNA 21 expression to mimic the natural change of miRNA expression upon biological stimulus. Aer being incubated with the PZ-DNA nanomachine and irradiated with UV light (7 mW cm À2 ) for 5 min, the intracellular Cy3 uorescence recovery was imaged via CLSM and the uorescence intensity was compared with the UCNP luminance at 658 nm to evaluate the intracellular miRNA 21 expression level. Both the miRNA 21 mimic and inhibitor treated HeLa cells demonstrated similar uorescence intensity for the UCNP luminance at 658 nm, indicating the efficient endocytosis process of the PZ-DNA nanomachine (Fig. 4a). Compared with the Cy3 uorescence recovery from untreated HeLa cells, the miRNA 21 mimic treated HeLa cells demonstrated obviously enhanced Cy3 uorescence, while Cy3 uorescence recovery was barely observed from the miRNA 21 inhibitor treated HeLa cells due to the decrease of intracellular miRNA 21 expression, providing the specic and quantitative response of the PZ-DNA nanomachine to intracellular miRNA 21. The luminance intensities for Cy3 at 580 nm and UCNPs at 658 nm as an internal standard were calculated from CLSM images, and I 580 /U 658 for the miRNA 21 mimic and inhibitor treated HeLa cells were 197.6% and 23.6%, respectively, compared with that of the untreated HeLa cells (Fig. 4b). Meanwhile, real time PCR (qRT-PCR) was performed to verify the accuracy of intracellular imaging by means of a miRNA 21 calibration curve with a linear range of 1 pM to 10 nM (Fig. S9 †), and the miRNA 21 copy number for the miRNA mimic-treated HeLa cells was 2.2 fold that of the untreated HeLa cells, while the miRNA 21 copy number for the miRNA inhibitor treated HeLa cells was 20.5% that of the untreated HeLa cells, which demonstrated a similar tendency to the result from CLSM imaging, conrming the accuracy of the PZ-DNA nanomachine for intracellular miRNA 21 imaging.</p><p>The cytotoxicity of the PZ-DNA nanomachine was evaluated with a standard 3-(4,5-dimethylthiazol-2-yl)-2diphenyltetrazolium bromide (MTT) assay at a series of PZ-DNA nanomachine concentrations (Fig. S10a †). Even at a very high concentration of PZ-DNA nanomachine, the HeLa cells still kept 91.4% viability. In addition, 5 min of UV irradiation (7 mW cm À2 ) didn't inuence the viability for both the HeLa cells and PZ-DNA nanomachine loaded HeLa cells (Fig. S10b †). The good biocompatibility of the PZ-DNA nanomachine and the duration of cells over short time UV irradiation guarantee its application in intracellular cell imaging. Though little inuence on cell viability and satisfactory photo control for precise cellular imaging have been demonstrated, the involvement of UV irradiation might limit the application of the PZ-DNA nanomachine in the living body considering its low penetration depth. Our ongoing follow up project will utilize NIR irradiation for activation of the DNA nanomachine, which hopefully will extend its in vivo applications.</p><p>The PZ-DNA nanomachine was further applied for miRNA 21 imaging of different kinds of cell lines, including MCF-7 cells and MDA-MB-231 cells with high miRNA 21 expression and HEK-293 cells with low miRNA 21 expression. Aer incubation with the PZ-DNA nanomachine and 5 min light irradiation, strong Cy3 uorescence recovery was observed from the MCF-7 cells and MDA-MB-231 cells, while weak Cy3 uorescence recovery was observed from HEK-293 cells with similar UCNP luminance intensity at 658 nm for all three cells lines in the CLSM images (Fig. 5a). The relative uorescence signals obtained aer normalization with internal standard corresponded with the copy numbers calculated from qRT-PCR (Fig. 5b) and previously reported results, 36 indicating the applicability of the PZ-DNA nanomachine for precise intracellular miRNA imaging.</p><!><p>A PZ-DNA nanomachine with an internal standard based on UCNPs was presented in this work for accurate intracellular miRNA imaging. A miRNA responsive DNAzyme walker was immobilized on the UCNP surface with its corresponding substrate DNA strands, and the miRNA responsive region of the DNAzyme was blocked by a photo-cleavable DNA strand, which well protected the unwanted activation of the DNA nanomachine before intracellular delivery and effectively avoided "false positive" signals for miRNA imaging in living cells. Aer photoactivation, the PZ-DNA nanomachine was operated by intracellular miRNA, and gave impressive luminance signal amplication for miRNA imaging in living cells. The concentration of the imaging dye on the UCNP surface improved the signal output and achieved intracellular imaging upon NIR irradiation. Multiple emissions from the UCNPs were used as an internal standard to enhance the detection accuracy. Meanwhile, the design allowed the presented PZ-DNA nanomachine to demonstrate high detection sensitivity for miRNA imaging from different cell lines, and the results were in good accordance with qRT-PCR measurements, providing a universal platform for the precise imaging of biomarkers in living cells.</p><!><p>The authors declare no competing nancial interests.</p>

Royal Society of Chemistry (RSC)

Copper Redistribution in Atox1-deficient Mouse Fibroblast Cells

Quantitative SXRF imaging of adherent mouse fibroblast cells deficient in Atox1, a metallochaperone protein responsible for delivering copper (Cu) to cuproenzymes in the trans-Golgi network, revealed striking differences in the subcellular Cu distribution compared to wildtype cells. While the latter showed a pronounced perinculear localization of Cu, the Atox1-deficient cells displayed a mostly unstructured and diffuse distribution throughout the entire cell body. Comparison of the SXRF elemental maps for Zn and Fe of the same samples showed no marked differences between the two cell lines. The data underscore the importance of Atox1, not only as a metallochaperone for delivering Cu to cuproenzymes, but also as a key player in maintaining the proper distribution and organization of Cu at the cellular level.

copper_redistribution_in_atox1-deficient_mouse_fibroblast_cells

Introduction<!>Reagents and Materials<!>Sample Preparation<!>Synchrotron Radiation X-ray Fluorescence Microscopy (SXRF)<!>Data Analysis<!>Results and Discussion<!>Basal Growth Conditions<!>Effect of Cu(II) Supplementation<!>Conclusions<!>

<p>Copper (Cu) is a trace metal essential for the maintenance of human health. Serving as a cofactor for a number of enzymes, Cu plays an important role in a broad range of biological processes including cellular respiration, free radical defense, iron mobilization and uptake, formation of connective tissue, pigmentation, blood clotting, and the synthesis of neurotransmitters [1]. Conversely, cellular Cu accumulation to excessive levels is detrimental as it catalyzes the formation of reactive oxygen species (ROS) that may damage DNA, proteins, and other biomolecules, thus resulting in increased oxidative stress and ultimately cell death [2,3]. Given this dualistic nature, cells have evolved an intricate machinery for transport, storage and regulation of intracellular Cu, such that sufficient amounts are available to fulfill its essential roles while avoiding accumulation to potentially damaging levels [4–9]. Instances of Cu imbalance or homeostatic dysfunction have been directly associated with a number of neurological diseases, including Menkes disease [10,11], Wilson's disease [12–14], Alzheimer's disease [15–19], Parkinson's disease [16–18], and amyotrophic lateral sclerosis [17,20,21], further highlighting the necessity for tightly controlled levels of intracellular Cu [22–24].</p><p>The discovery of genes coding for proteins involved in Cu-uptake, transport and regulation led to the identification of a host of proteins involved in Cu homeostasis. On a cellular level, Cu is imported across the plasma membrane via the high affinity Cu transporter Ctr1, and subsequently partitioned into distinct trafficking pathways that involve specialized metallochaperones [25–29]. For example, the Cu chaperone Ccs is responsible to escort and incorporate Cu into cytoplasmic superoxide dismutase (SOD) [30,31]. Likewise, the three chaperone proteins Cox17, Cox11, and Sco1 deliver Cu to mitochondrial cytochrome c oxidase [32–35], and the copper chaperone Atox1 is critical for transport of Cu to either the Menkes (ATP7a) or Wilson's (ATP7b) disease ATPases situated in the trans-Golgi network (TGN) [36–38], where it is delivered either to cuproenzymes such as ceruloplasmin and lysyl oxidase, or trafficked through the secretory pathway for extracellular release [39].</p><p>While significant progress has been made towards the mechanistic understanding of cellular Cu homeostasis at the molecular level, important questions remain regarding the maintenance of Cu at the cellular level. For example, Cu uptake and release occur with a surprisingly rapid kinetics [40], suggesting that a portion of the total cellular Cu is present in a kinetically labile form and thus readily available for distribution and uptake into cuproenzymes; however, the nature of this labile pool, its subcellular localization, and potential alterations in diseases associated with defects in Cu trafficking are poorly understood. In view of the potential toxicity of Cu, a detailed understanding of the redistribution and mislocalization of labile Cu in these diseases is of particular importance.</p><p>Several cell lines have been created that may serve as model systems to investigate the potential Cu redistribution associated with defects in Cu trafficking pathways. In this report, we focused on studying the subcellular Cu distribution in Atox1−/− fibroblasts, an embryonic mouse cell line that is deficient in the chaperone protein Atox1 responsible for trafficking of Cu to the TGN network [36]. The protein Atox1 was originally discovered through its yeast homologue Atx1, which showed a protective function against superoxide and peroxide toxicity [41]. The Atox1−/− cell line was created by disruption of the Atox1 locus through gene-trap insertion of a β-galactosidase-neomycin marker in mouse embryonic stem cells [37]. Mice deficient for Atox1 display a severe phenotype characterized by growth retardation, perinatal mortality, and congenital malformations, much like the phenotype observed in Menkes' disease [37]. At the cellular level, significant differences in the Cu-mediated trafficking of the Menkes ATPase from the TGN to cytosolic vesicular compartments were observed between Atox1+/+ and Atox1−/− cells [36]. While Atox1−/− cells still showed a Cu-dependent translocation of the Menkes protein out of the TGN compartments, the movement was significantly impaired in a dose and time-dependent manner. Most recently, Atox1 has been also implied as a Cu-dependent transcription factor that mediates Cu-induced cell proliferation [42]. In view of the essential role of Atox1 in cellular Cu trafficking and homeostasis, the elucidation of differences in the subcellular Cu distribution between Atox1+/+ and Atox1−/− cells was of particular interest.</p><p>To quantify the distribution of Cu at the subcellular level, highly sensitive microanalytical techniques are required. While the sensitivity of traditional methods, such as inductively coupled plasma mass spectrometry (ICP-MS), atomic emission spectroscopy, (AE) or x-ray fluorescence analysis (XRF) is insufficient for analyzing the elemental contents of a single cell, several more recently developed techniques, notably secondary ion mass spectrometry (SIMS), nuclear microprobes (PIXE or PIGE), and synchrotron x-ray fluorescence microscopy (SXRF, SRIXE, or microXRF) offer sufficient sensitivity to obtain spatially resolved elemental maps at the subcellular level [43–47]. These new imaging techniques have already provided important insights into the localization of Cu pools within individual cells. For example, an SXRF study of Cu-loaded NIH 3T3 cells yielded detailed topographical elemental maps and suggested the presence of a labile Cu pool in the TGN region and mitochondria [48]. High-resolution SXRF imaging of human microvascular endothelial cells revealed an intriguing relocalization of Cu from intracellular compartments towards the tips of filopodia, thus highlighting the importance of endogenous Cu during angiogenesis [49]. A similar Cu accumulation was recently described in thin neurites formed upon nerve growth factor (NGF) stimulated differentiation of rat pheochromocytoma (PC12) cells, which were used as an in vitro model of dopaminergic cells [50]. Given these early successes in imaging the subcellular distribution of Cu with great detail, SXRF imaging seemed ideally poised for studying the effects of Cu relocalization in cells with altered trafficking pathways.</p><!><p>The wildtype Atox1+/+ and Atox1−/− embryonic mouse fibroblast cell lines were a generous gift from Jonathan D. Gitlin [36]. The cells were cultured in Dulbecco's modified Eagle's medium (DMEM, GIBCO-BRL, Gaithersburg, USA) supplemented with 10% bovine serum (GIBCO), 200 mM L-glutamine (GIBCO), 250 μg/mL geneticin (GIBCO), penicillin (100 IU/mL) and streptomycin (100 mg/L) at 37°C under an atmosphere of humidified air containing 5% CO2. The culture medium was sterilized by filtration through 0.2 μm filters.</p><!><p>Cells were grown to 50–80% confluency on silicon nitride windows (2 × 2 × 0.0005 mm, Silson Ltd., UK) that were pre-treated for 30 min with 0.01% poly-L-lysine solution (Sigma-Aldrich). For preparations in basal medium, cells were directly seeded in 6-well culture plates containing the silicon to reach the desired confluency. For experiments that nitride windows and grown at 37°C/5% CO2 involved supplementation of the culture medium with Cu(II), cells were first grown in basal medium to 50% confluency as described above, then switched for 24 h to basal medium containing 200 μM bathocuproine disulphonate (BCS, Sigma-Aldrich), thoroughly washed with 1X-PBS (pH 7.2, pre-warmed to 37°C), and finally incubated in full growth media supplemented with 50 μM CuCl2 (Sigma-Aldrich) for 4 h at 37°C.</p><p>Following the growth/incubation conditions described above, cells were further prepared for SXRF experiments as previously described [51]. Briefly, cells were washed with PBS (pre-warmed at 37°C), and fixed for 10 min at room temperature with 3.7% paraformaldehyde (Sigma-Aldrich, freshly prepared solution in PBS). After thoroughly washing with PBS, the samples were rinsed twice with sterile dH2O followed by two brief washes with isotonic ammonium acetate (Sigma-Aldrich, 0.1 M solution in sterile dH2O). Finally, samples were air-dried overnight in a covered sterile cell culture dish.</p><!><p>Synchrotron radiation x-ray fluorescence (SXRF) microscopy was performed at the 2-ID-D beamline of the Advanced Photon Source located at Argonne National Laboratory (IL, USA). The air-dried cells grown on silicon nitride windows were placed onto a kinematic specimen holder suitable for both optical and x-ray fluorescence microscopy. The holder was mounted on a light microscope (Leica DMXRE) and target cells were located on the grid relative to a pre-determined reference point using a motorized x/y stage (Ludl Bioprecision). Coordinates were recorded and used to precisely locate the target cell(s) once the grid was transferred to the microprobe. For XRF excitation, a monochromatic X-ray beam generated by an undulator source was focused to a spot size of 0.5×0.5 μm2 on the specimen by using a Fresnel zone plate. An excitation energy of 10 keV was chosen to ensure excitation of all first row transition elements as well as Ca and K, although to a lesser extent. The sample was subsequently raster scanned through the beam at 298K under a helium atmosphere. The pixel step size was set to 0.5 μm and the entire X-ray spectrum was recorded for each pixel using an energy dispersive germanium detector (Canberra LEGe detector).</p><!><p>Elemental maps were created by spectral filtering, using spectral regions of interest matched to characteristic x-ray emission lines to determine the fluorescence signal for each element. Calibration to elemental area densities (μg/cm2) was done by comparison of x-ray fluorescence signal strength from the sample to fluorescence from thin film standards NBS-1832 and NBS-1833 from the National Bureau of Standards (NBS/NIST, Gaithersburg, MD) using MAPS software [52]. The elemental content was calculated by fitting of individual spectra of the acquired fluorescence datasets, and comparing fitted fluorescence signal strength with that resulting from fitting of NBS 1832/33 standard spectra.</p><!><p>To choose growth conditions that would maximize the differences in Cu distribution and thus best reveal the impact of altered Cu trafficking pathways, we utilized earlier studies on the Cu-dependent relocalization of the Menkes ATPase as guidelines [36]. Specifically, the following findings were taken into consideration: (1) the Menkes protein localization and total cellular Cu were identical for both Atox1−/− and wildtype cells under Cu-limiting conditions, where cells were grown in low-serum media supplemented with 200 μM BCS as extracellular Cu(I)-chelator, (2) differences in Cu-mediated trafficking out of the TGN were most pronounced when the medium was supplemented with 10 μM CuCl2 following BCS treatment, and (3) the impaired Menkes protein trafficking in Atox1−/− cells could be rescued by supplementation with ≥ 100 μM CuCl2 and extended incubation periods (> 4 h) [36]. Based on these data we anticipated that differences in Cu distribution between Atox1−/− and wildtype cells would be most pronounced when cells were pretreated with 200 μM BCS for 48 hours followed by supplementation with 10 but less than 100 μM CuCl2 in full media over a period of 4 hours. Although no data were provided for the intermediated concentration range, supplementation with 50 μM CuCl2 appeared to strike the best balance between improving the signal-to-noise ratio for SXRF detection while still maintaining the delayed Menkes trafficking in Atox1−/− cells. As a control, both cells lines were also cultured under basal conditions in full media without supplementation. Adherent cells were directly grown on a x-ray compatible substrate (silicon nitride), chemically fixed, and dried in air. Individual cells were raster scanned with excitation at 10.0 keV with 0.5 μm step size, yielding 2D maps for all biologically relevant first row transition elements. The considerable size of adherent fibroblast cells combined with the necessity to obtain elemental maps with high spatial resolution resulted in long data acquisition times (2–4 hours per cell), and therefore restricted the sample size to only three repeats for each growth condition and cell type. Quantitative elemental maps for Cu, Zn, and Fe were analyzed in terms of the total cellular content as well as the nuclear and cytoplasmic portions (Table 1). To account for differences in cell size that might also contribute to differences in trace metal content, the data are listed as densities in units of pmol/cm2. For ease of comparison, the density data compiled in Table 1 are also illustrated in a set of three bar graphs (Figure 1).</p><!><p>Due to the low cellular Cu content under these growth conditions, the signal-to-noise ratios of the resulting SXRF maps were insufficient to accurately assess differences in the subcellular Cu distribution between the two cell lines. Nevertheless, integration of the Cu signal in the nucleus compared to cytoplasm revealed significant differences between Atox1−/− and wildtype control cells (Table 1). Consistent with previous findings based on atomic absorption spectroscopy of bulk samples [36], Atox1−/− cells significantly accumulated Cu due to impaired Cu efflux under these conditions. Furthermore, a recent study with a metallothionein (MT)-knockout cell line showed also increased intracellular Cu levels upon siRNA-mediated knockdown of Atox1 [53]. In wildtype cells, approximately 23% of the total cellular Cu was localized in the nuclear region (Table 1). Interestingly, Atox1−/− cells consistently showed an elevated nuclear Cu content compared to wildtype. In contrast, no significant changes were observed for the Zn and Fe content, neither in the whole cell analysis nor the nuclear/cytoplasmic distribution ratios (Table 1 and Figure 1).</p><!><p>Wildtype cells incubated with 50 μM CuCl2 for 4 hours displayed a marked increase of the cellular Cu content (Table 1). Interestingly, Atox1−/− cells treated under the same conditions revealed a similar increase of intracellular Cu (Table 1) [36], thus contrasting the observations under basal growth conditions. Nevertheless, these findings are in agreement with the data on bulk samples reported in the literature [36]. While the impaired Cu efflux mechanism in Atox1−/− cells eventually leads to a substantial increase in the intracellular Cu content [37], at early time points the response to elevated extracellular Cu(II) is indistinguishable compared to wildtype cells. Likewise, the ratio of nuclear to total Cu was identical within experimental error for both cell lines and similar to wildtype cells grown under basal conditions. A comparison of the Fe content showed no apparent differences between the two cells lines in response to Cu(II) supplementation, although the ratio of nuclear Fe appeared to be reduced in both cases. Similarly, analysis of the Zn content and relative distribution between nucleus and cytoplasm showed no significant differences and appears to be independent of the cell type and growth conditions with Cu(II) supplementation.</p><p>Despite the similarities of the total Cu content, the subcellular distribution, as revealed by the quantitative SXRF maps shown in Figure 2, showed striking differences between Atox1−/− and Atox1+/+ cells. The wildtype cell line consistently displayed a strong Cu localization in the perinuclear region with a slightly less pronounced abundance throughout the cytoplasm and within the cell nucleus (Figure 2, top row). Conversely, the Cu distribution in Atox1−/− cells revealed no obvious compartmentalization or subcellular accumulation, but rather a diffuse distribution throughout the entire cell body (Figure 2, bottom row). While both cell lines showed a similar ratio of nuclear to total cellular Cu, the distribution in Atox1−/− cells appeared entirely unstructured and lacked areas of localization as observed in the nucleus of wildtype cells. A comparison with the Zn elemental maps obtained from the same set of cells showed no marked differences between Atox1−/− and wildtype cells (Figure 3).</p><p>Previous studies demonstrated that supplementation with 100 μM CuCl2 over a time period of 4 hours resulted in almost complete relocalization of the Menkes ATPase out of the TGN to cytoplasmic compartments [36]. Given the comparable incubation conditions and time course of our SXRF experiments, we can assume that a similar relocalization of the Menkes protein occurred; however, a significant fraction of Cu remained localized in the perinuclear region (Figure 2, top row). While the Cu- dependent trafficking of the Menkes protein out of the TGN is well established [36,54], it still remains to be determined what fraction of cellular Cu is actually relocalized in this process. The SXRF data imply that Cu trafficking and distribution is not exclusively linked with Menkes trafficking at early time points following stimulation with elevated extracellular Cu(II). Taking into account that the requirement for Atox1 in Cu-mediated trafficking of the Menkes protein can be bypassed with excess Cu [36], we cannot exclude the possibility that the diffuse Cu distribution in Atox1−/− cells is a consequence of Cu delivery to Menkes protein containing compartments, either through Atox1-independent loading prior to or after trafficking out of the TGN. Such a scenario would imply, however, that alternative Atox1-dependent trafficking pathways exist that lead to compartmentalization of Cu in the perinuclear region of the wildtype cells. Alternatively, the diffuse Cu distribution pattern in Atox1−/− cells might be the result of a random redistribution process due to the absence of the regular trafficking pathway.</p><!><p>Spatially well resolved SXRF elemental maps of individual adherent mouse fibroblast cells revealed intriguing differences in the Cu distribution of Atox1−/− cells compared to the corresponding wildtype. While the latter cells showed a distinct perinculear Cu localization, the distribution in Atox1−/− cells was reproducibly unstructured and diffuse throughout the entire cell. Although the SXRF elemental maps cannot reveal the nature of the associated cellular structure or organelles in absence of a xenobiotic label, the characteristic perinuclear Cu distribution pattern might point towards the involvement of the Golgi apparatus, late endosomes or mitochondria, all of which are typically found in the vicinity of the nuclear envelope. The data highlight the importance of Atox1, not only as a metallochaperone for delivering Cu to cuproenzymes, but as a key player in maintaining the proper distribution and organization of Cu at the cellular level.</p><!><p>Trace metal contents of Atox1+/+ and Atox1−/− cells cultured in basal medium and medium supplemented with 50 μM CuCl2 for 4 h. The SXRF data were quantitatively analyzed for the content of Cu (a), Fe (b) and Zn (c). The corresponding numerical values are compiled in Table 1.</p><p>False-color micrographs showing the subcellular distribution of Cu in Atox1-deficeint (Atox1−/−) and wildtype (Atox1+/+) cells visualized by SXRF microscopy. Cells were grown in media supplemented with 50 μM CuCl2 for 4 h after pretreatment with 200 μM BCS for 48 hours. The density interval indicated above each scan refers to the full dynamic range of the false-color scale depicted in the bottom row.</p><p>False-color micrographs showing the subcellular distribution of Zn in Atox1-deficeint (Atox1−/−) and wildtype (Atox1+/+) cells visualized by SXRF microscopy. Cells were grown in media supplemented with 50 μM CuCl2 for 4 h after pretreatment with 200 μM BCS for 48 hours. The density interval indicated above each scan refers to the full dynamic range of the false-color scale depicted in the bottom row.</p><p>Trace metal contents of Atox1+/+ and Atox1−/− cells cultured in basal medium and medium supplemented with 50 μM CuCl2 for 4 h.</p><p>DMEM, 10% bovine serum, 200 mM glutamine;</p><p>cells cultured basal medium and switched to medium supplemented with 50 μM CuCl2 for 4 h.</p><p>ROI = region of interest indicated in each row.</p><p>average metal content based on three individual cells.</p><p>percent ratio of the integrated nuclear metal and total cellular metal content.</p>

PubMed Author Manuscript

Extension of the Tryptophan \xcf\x872,1 Dihedral Angle - W3 Band Frequency Relationship to a Full Rotation: Correlations and Caveats\xc2\xa5

The correlation of the UVRR \xce\xbdW3 mode with the tryptophan \xcf\x872,1 dihedral angle (1-3) has been extended to a full, 360\xc2\xb0 rotation. The three-fold periodicity of the relationship (cos 3\xcf\x872,1) over 360\xc2\xb0 results in up to six dihedral angles for a given \xce\xbdW3. Consideration of a Newman plot of dihedral angles for proteinaceous tryptophans taken from the Protein Data Bank shows that sterically hindered ranges of dihedral angle reduce the possible \xcf\x872,1 to one or two. However, not all proteinaceous tryptophans follow the \xce\xbdW3-\xcf\x872,1 relationship. Hydrogen bonding at the indole amine, weaker, electrostatic cation-\xcf\x80 and anion quadrapole interactions and environmental hydrophobicity are examined as possible contributing factors to noncompliance with the relationship. This evaluation suggests that cumulative weak electrostatic and nonpolar interactions characterize the environment of tryptophans that obey the \xce\xbdW3-\xcf\x872,1 relationship, matching that of the crystalline tryptophan derivatives used to derive the relationship. In the absence of methods to quantify these weak interactions, measurement of the full width half maximum bandwidth (FWHM) of the W3 band is suggested as a primary screen for evaluating the applicability of the \xce\xbdW3 - \xcf\x87 2,1 relationship.

extension_of_the_tryptophan_\xcf\x872,1_dihedral_angle_-_w3_band_frequency_relationship_to_a_full_ro

<!>Materials and Methods<!>UV Resonance Raman Spectroscopy<!>Crystallographic Measurements<!>Raman Data Simulations<!>Results and Discussion<!>The W3 Mode (1542 \xe2\x80\x93 60 cm-1) and the Dihedral Angle, \xcf\x872,1 for Trp-168 of TIM<!>The W3 Mode - Dihedral Angle, \xcf\x872,1 Relationship Extends over a Full Rotation<!>Modification of the \xce\xbdWd3-\xcf\x872,1 Relationship<!>Steric Hindrance result in Excluded Values for the Tryptophan \xcf\x872,1 Dihedral Angle<!>Exceptions to the Rule: Not all Protein Tryptophans Manifest the \xce\xbdW3 -- \xcf\x872,1 Relationship<!>Indole Bond Lengths with Respect to the \xce\xbdW3-\xcf\x872,1 Relationship<!>DFT Calculations Applied to the Discrepancy between the \xcf\x872,1 Torsion Angle Predicted by \xce\xbdW3 versus X-ray Crystallography for Ligated TIM<!>Noncovalent Interactions at the Trp-168 Indole - Hydrogen Bonding<!>Noncovalent Interactions at the Trp-168 Indole: Anion-Quadrapole and Cation-\xcf\x80 Interactions<!>Noncovalent Interactions at the Trp-168 Indole: Hydrophobicity<!>Noncovalent Interactions at the Trp-168 Indole: Steric Hindrance<!>Tryptophans Compliant with the \xce\xbdW3 - \xcf\x872,1 Relationship<!>Tryptophans Noncompliant with the \xce\xbdW3 - \xcf\x872,1 Relationship<!>W3 Bandwidth as a Predictor of Compliance with the \xce\xbdW3-\xcf\x872,1 Relationship<!>Conclusion<!>

<p>The utility of spectroscopy with respect to protein structure lies in its predictive value in the absence of crystallographic structure. In UV resonance Raman spectroscopy, tryptophan residues are structural bellwethers because they can be selectively probed at 229 nm. Tryptophan vibrational band are good markers for hydrophobic interaction (3-6), hydrogen bonding (5, 6) and conformation (1, 2, 6-8). The intensity ratio, I1360/I1340, of the tryptophan Fermi 'doublet' (also known as the W7 mode) serves as a good marker for hydrophobic interactions (2, 4, 9). A rise in the ratio indicates an increasingly hydrophobic environment for the indole ring. Several tryptophan modes, W2, W4, W6, and W17 are sensitive to hydrogen bonding. The stronger the hydrogen bonding the higher is νW2, νW4 and νW6 (2) and the lower is νW17 (8). One mode, the W3 mode, is of great interest because it tracts the conformation of tryptophan. The W3 mode is a hybrid mode consisting of indole C2=C3 stretching (50%), N1-C2 (23%) stretching and C2-H (19%) bending vibrations (2, 6), yet its frequency depends on the χ2,1 dihedral or torsion angle about the Cα–Cβ–C3=C2 linkage (2). A systematic study of the W3 mode frequency (νW3) for several crystalline tryptophan analogs revealed that νW3 varies with the χ2,1 torsion angle according to the empirically derived equation (1):</p><p>Deuteration of the tryptophan ring d5 carbon affects the relationship of νW3 with the χ2,1 torsion angle (8), and a similar equation can be written (1, 7):</p><p>These equations are derived using the absolute value of the torsion angle, |χ2,1|, within a range of 60-120° (3). It is apparent that the angular dependence for νWd3 is greater than for νW3 as the angular coefficient for νWd3 is larger.</p><p>In this work, we extend and modify the relationship, νW3(χ2,1), to cover a full 360° rotation since all dihedral angles are plotted in the same rotational direction. We show that the relationship is not single-valued---and therefore not a function---and has a three-fold periodicity over 360° due to the three-fold steric overlap of the indole residue with moieties bound to Cα. Thus, a single νW3 corresponds to as many as six dihedral angles. This follows intuitively from a consideration of a Newman diagram for dihedral steric clashes as discussed below. νW3-χ2,1 data points for single tryptophan wild-type and mutant proteins, multiple tryptophan proteins where the W3 band has been resolved into components, fibrillar peptides (fd and Pf3 virion coat peptides) and two model tryptophan-containing peptides, the bee venom peptide, melittin, and the caged tryptophan peptide, exendin-4 TC5b, have been added to the original νW3 plot for crystalline tryptophan analogs (1, 7). This plotting suggests that three ranges of dihedral angle are poorly populated, attributable to steric interference with the protein backbone.</p><p>At the same time, the νW3-χ2,1 data points for the proteins and model peptides show that the cosine-based relationship can only provide a rough guide to the tryptophan dihedral angle, even for tryptophans in a hydrophobic, constrained environment similar to that of the crystalline tryptophan derivatives used to derive the νW3-χ2,1 relationship. For some of the peptides and proteins, the UV resonance-Raman determined νW3, coupled with a χ2,1 dihedral angle taken from an x-ray crystal structure, constitutes a data point that lies off the cosine-dependent curve. In particular, we examine the influence of hydrogen bonding at the indole amine on νW3 compliance with the cosine-dependent relationship through density function theory (DFT) calculations. For these calculations, we focus on the S. cerevisiae mutant enzyme, TIM Trp90Tyr Trp157Phe, bound to the ligand, 2-phosphoglycolate (PGA), because all other structural information gleaned from the UVRR data correlates with previously published results for TIM. DFT calculations based on model systems are utilized to understand the deviation of TIM νW3 from Eq. (1).</p><p>We also examine the weak, noncovalent cation-π and anion-quadrapole interactions of proteinaceous tryptophans as the source of noncompliance with the νW3-χ2,1 relationship. Tryptophan dihedral angles for these proteins are taken from x-ray crystal or NMR structures available in the Protein Data Bank while νW3 values are taken from the literature or have been measured. We compare the indole C2=C3, Cβ–C3, and C2–N1 bond lengths for the νW3-χ2,1 relationship-compliant versus noncompliant tryptophans. This information is also gleaned from structures in the Protein Data Bank, and is relevant because the νW3 depends upon C2=C3 and C2–N1 stretching, and the Cβ–C3 bond length has been found to vary inversely with |χ2,1| (1, 2). Hydrophobicity of the tryptophan environment is indicated by measure of the W7 1360 cm-1:1340 cm-1 band ratio (4, 5). As the crystalline tryptophan derivatives used to formulate the νW3-χ2,1 relationship are expected to reside in a hydrophobic environment, the W7 ratio was examined for model peptides and relationship-compliant and noncompliant proteins.</p><p>The purpose of this analysis is to expose the specific tryptophan interactions and indole structural factors that contribute to deviations of νW3-χ2,1 data points from the relationship given in Eq. (1). We find that numerous weak electrostatic and nonpolar interactions accumulate to create an environment about the indole of tryptophan that is both constrained and hydrophobic, like the environment of the crystalline tryptophan derivatives from which the νW3 - χ 2,1 relationship was derived. In the absence of methods to quantify these weak interactions, measurement of the FWHM bandwidth of the W3 band is suggested as a primary screen for evaluating the applicability of the νW3 - χ 2,1 relationship.</p><!><p>All reagents used were purchased from either Sigma-Aldrich Co. (St. Louis, MO) or Fluka (Milwaukee, WI) with the exception of glycerol-3-phosphate dehydrogenase, which was purchased from Boehringer Mannheim Ltd. (Indianapolis, IN). The model peptide, exendin-4 TC5b, was synthesized according to the protocol given in (10). S. cerevisiae TIM (Trp90Tyr Trp157Phe; hereafter referred to as the TIM mutant or simply TIM) was expressed and purified as described before (11). The kinetic parameters of the mutant are not distinguishable from those of the wild type TIM (12). Enzymatic activity was determined by the conversion of GAP to DHAP in the presence of TIM and glycerol 3-phosphate dehydrogenase as described by Putman and Knowles (13).</p><p>TIM enzyme samples employed in UV resonance Raman spectroscopy were at a concentration of 0.3 mM TIM in 50 mM Tris-HCl, 50 mM NaCl buffer, pH 7.8. Four samples were measured: unbound TIM and TIM bound to PGA (5 mM) in the aqueous buffer given above, and deuterium solutions of TIM bound to Pi (50 mM) and G3P (10 mM) with buffer conditions as above. Exendin-4 TC5b (1 mM) was spun cast in a silica gel matrix. Following gelation, the gel was bathed in 15 mM phosphate buffer, pH7.0. The method of silica gel preparation is given in (14). The melittin peptide was dissolved in 20 mM Tris buffer, pH 7.4 to a concentration of 2 mM.</p><!><p>All UVRR spectra were acquired with CW 229 nm excitation at an incident power of 1.8 mW. The UVRR instrument has been described elsewhere (15). For each sample, three acquisitions of three minutes, collected over an 805-1680 cm-1 frequency window, were summed. The sample-containing quartz tube was rastered vertically, spun and chilled to 16°C to minimize protein exposure to potentially denaturing UV radiation. UVRR peak positions were calibrated against the known peaks of indene. The resulting calibration-limited accuracy is ± 1 cm-1. Curvefitting analysis was carried out via the GRAMS/AI software package. All curvefits were carried out with the minimal number of Gaussian-Lorentzian peaks (optimally 50 -70 % Lorentzian) necessary to yield a χ2< 3.</p><!><p>Dihedral angles, bond lengths, and intermolecular distances were measured from the indicated Protein Data Bank structures using the UCSF Chimera package from the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco (supported by NIH P41 RR-01081) (16).</p><!><p>A series of calculations were performed using DFT-based simulations with Gaussian03W (17). A DFT approximation implementing the Becke's three-parameter exchange functional (18). in combination with the Lee, Yang, and Parr correlation (19) function (or B3LYP) was used. As a compromise between accuracy and applicability to large molecules, the 6-31G(d) basis set was used. Smaller basis sets are enough in DFT based calculations because the basis functions do not have to describe correlating orbitals (20, 21). For all systems the geometry was first optimized before the frequency calculations. All calculated frequencies were scaled by a factor of 0.963 (20, 22). GaussView (23) was used to visualize the atomic displacements correlating to each of the simulated Raman peaks. Using normal coordinate analysis, the displacement vectors associated with each different vibrational mode of skatole (6), a tryptophan analog, and skatole-d5 (1) were previously assigned. To identify the W3 or Wd3 mode in the simulated Raman spectra, we compared the visualized atomic displacements to the displacement vector-W3 mode relationship for skatole and skatole-d5. A series of calculation were made to prove that the assignment was correct. First, a full geometry optimization and frequency calculation at fixed |χ2,1| torsion angles of 60.6 °, 78.0 °, 83.5 °, 88.0 °, 98.7 °, 105.1 ° and 116.5° were carried out. These χ2,1 torsion angles were chosen because they correspond to the values used to generate Eq. (1). Second, we made a similar set of calculations using Trp-d5 to generate a relationship similar to Eq. (2). Finally, we performed unconstrained simulations on a Trp molecule, a Trp-d5, and a Trp analog where the indole N had been replaced by an isotope, 15N. As defined, the W3 mode should have very minimal contribution from the indole N (2, 6).</p><!><p>Four 229 nm.-excited UVRR spectra for the S. cerevisiae mutant enzyme, TIM (Trp90Tyr Trp157Phe), in the 800 – 1550 cm-1 range are shown in Figure 1. These include results for the ligand-free enzyme (Fig. 1a, trace 1), and the enzyme bound to the tight-binding transition state analog, PGA (Fig. 1a, trace 2). UVRR results for deuterium solution are also given for TIM ligated to G3P (Fig. 1b, trace 1) and phosphate (Pi)-ligated TIM (Fig. 1b, trace 2). The many peaks represent resonantly-enhanced vibrational modes for the several tyrosines in the TIM sequence, and most importantly, the single tryptophan residue, Trp-168, which is critically situated at the substrate binding loop hinge. The Fig. 1 inset shows a close-up view of the W3 bands for these spectra.</p><p>The UVRR data (Figure 1a) present a picture of the immediate environment of Trp-168 in TIM. The intensity ratio, I1360/I1340, of the W7 mode peaks has been linked to the hydrophobicity of the tryptophan environment, where a value less than one for the ratio corresponds to a hydrophilic environment, and conversely, a value greater than one indicates a hydrophobic one (4). The value of the νW17 (876 cm-1) for TIM in both the bound and unbound conformations indicates weak hydrogen bonding at the indole amine while the W6 band position at 1420 cm-1 corresponds to a non-hydrogen bonded state (2). Like W3, the W1 mode, located at 1620 cm-1, depends upon indole ring stretching, but νW1 is buried beneath the intense tyrosine Y8a/b bands at ∼1614 cm-1 (not shown) and is therefore undecipherable. Concomitantly, the significance of tyrosine band changes accompanying binding loop motion---such as those for the Y9a mode at 1173 cm-1---are difficult to interpret due to the presence of nine tyrosine residues in TIM. However, the importance of the Y208 residue to enzyme activity has been realized (12, 24).</p><!><p>While x-ray crystal studies of the bound and unbound binding loop conformers of TIM show that a backbone hydrogen bond network enforces a rigid loop movement between the two binding states (25), subsequent solid-state TIM NMR studies (11, 26) and T-jump studies of solution phase TIM (27) show that Trp -168 has some local mobility. The Trp-168 W3 mode (∼1542 – 1560 cm-1) is expected to report on the local dihedral angle changes of Trp-168 that accompanying loop motion according to the νW3-χ2,1 relationship discussed above. The x-ray crystal structure, 1YPI, of apo TIM with Trp-168 in the loop unbound position reveals the χ2,1 torsion angle to be -79.65° (28). When TIM is bound to the PGA transition state analog, x-ray crystallography reveals (PDB file: 7TIM), the χ2,1 torsion angle of Trp-168 to be -99.61° (25).</p><!><p>A problem arises when we try to reconcile the observed W3 frequencies and corresponding χ2,1 torsion angles as derived from Eq. (1) (2) with that obtained from the x-ray crystal structures. Binding of the transition state analogue, PGA, to TIM downshifts the W3 mode from 1547 cm-1 to 1544 cm-1 (Fig. 1a, trace 4). According to Eq. (1), this νW3 corresponds to a change in |χ2,1| from 83° to 78°, a -5 ° dihedral angle change in going from the apo to holo enzyme conformation. The x-ray diffraction derived dihedral angle for apo TIM is -79.6° (PDB:1YPI; (28)), which is within the error bounds defined by Eq. (1). However, for PGA-bound TIM, the x-ray diffraction-obtained value for χ2,1 is -99.6° (PDB:7TIM; (25)), far from the value of ±77.6° predicted by Eq. (1). The resulting apo-to-holo dihedral angle change predicted from the crystal structures is -20 °. While the magnitude of the dihedral angle change predicted from crystal structures is four-fold higher, the sign of the change is the same as for the UVRR data.</p><p>These discrepancies between χ2,1 values predicted by the relationship given in Eq. (1) and those available from crystallographic structures suggest that plotting νW3 against x-ray crystallographic dihedral angles for proteins would reveal the general validity of Eq. (1). This is of interest because the relationships reproduced in Eqs. (1) and (2) are often used to predict tryptophan dihedral angles from νW3 results for proteins. Table 1 lists several tryptophan-bearing peptides and proteins with their corresponding νW3 and crystal structure-derived χ2,1 dihedral angles. According to the convention used in calculating these angles, values are reported within the range of ± 0 - 180°. These dihedral angle values have been recalculated so they can be plotted in the same rotational direction. Thus a negative value of χ2,1 is recalculated as 360° + χ2,1. Furthermore, given the three-fold redundancy---3 cos χ2,1---of the relationship given in Eq. 1 over a full rotation, the angular values of 1-120°, 121-240° and 241-360° are degenerate, and all dihedral angles have been reduced to a value between 0-120 ° (Table 1).</p><p>The reduced χ2,1 dihedral angles given in Table 1 are plotted against the corresponding UVRR-determined νW3 in Figure 2a (solid triangles). Also included are data points (open squares) for the crystalline tryptophan analogs from which Eq. 1 was derived (1), and data points (open diamonds) for model tryptophan-containing peptides, melittin conformers A and B (labeled 5 and , respectively) and the truncated exendin-4 TC5b W/P cage peptide (labeled 4). Data points for Trp-bearing proteins that lie far from the fitted curve are also numbered as follows: 1. fd virion coat protein, 2. TIM mutant Trp90Tyr Trp157Phe ligated to PGA, 3. Pf3 virion coat protein. The relationship given by Eq. (1) has been modified (see discussion below) to fit the experimental data points given in Table 1, and is also plotted in Fig. 2a as a solid line. The overlay of the three degenerate χ2,1 dihedral angle ranges means that data points for proteins (Table 1) define the relationship over one full rotation (Fig. 2a). This has not been previously realized. Due to the χ2,1 angular degeneracy and the fact that the fit relationship is not single valued, one νW3 can map to as many as six χ2,1 dihedral angles, as can be seen from Fig. 2a. Therefore, νW3 would seem to be a poor predictor of the tryptophan χ2,1 dihedral angle. However, as discussed below, large ranges of dihedral angle values for tryptophan are generally avoided in protein structures, greatly reducing the possible number of dihedral angles for any one νW3.</p><p>The νW3-χ2,1 relationship was previously extended to deuterated tryptophan in UV Raman spectroscopic studies of deuterated, solid state tryptophan derivatives (1). The νWd3 - dihedral angle data pairs originally used to define this relationship are reproduced in Fig. 2b (solid circles). Additionally, a single data pair for deuterated, phosphate and G3P ligated TIM is shown (star) as the same dihedral angle as used for aqueous, ligated TIM (Table 1) is used here. A plot of a modified νWd3 - dihedral angle relationship, fit to the data pairs for deuterated tryptophans, is also shown in Fig. 2b (solid line). As for G3P-ligated TIM in aqueous buffer (Table 1 and Fig. 2a), the experimentally determined νWd3 does not place the corresponding crystallographic dihedral angle value squarely on the modified curve fit discussed below. Possible reasons for deviations of protein νW3 and νWd3 - dihedral angle data points (Fig. 2 a, b) from the modified Eqs. (1) and (2) are discussed below.</p><!><p>In the first paper to discuss the νWd3-χ2,1 relationship, the data was fit to an unspecified third order spline function (2). Subsequently, Eqs. (1) and (2) were introduced as empirical fits to the νW3 and νWd3 - dihedral angle data points (1). Expansion of the relationship given in Eq. (1) to cover a full rotation, and using the same rotational sense for all dihedral angles, χ2,1, yields</p><p>Where for all negative values of dihedral angle, χ2,1</p><p>And a and b are empirically-determined parameters that are determined from fitting to the νW3-χ2,1 data pairs. The physical significance of the cos 3 χ3602,1 dependence is now discussed.</p><!><p>A circular plot of the crystallographically determined tryptophan dihedral angles for the proteins given in Table 1 reveals that certain ranges of dihedral angle are favored in protein structures. This plot is given in Figure 3. χ3602,1 is plotted clockwise and all dihedral angles are plotted as positive angles in the 0 - 360° range. Dihedral angle values for tryptophans found in proteins in water solution (triangles; see Table 1) and those for crystalline tryptophan derivatives (diamonds) (2) are plotted. Dihedral angle values for two model peptides, melittin and exendin-4 TC5b (Table 1), are indicated by crosses. The three allowed tryptophan dihedral angles for an α helix, as determined from a dataset of 61 globular proteins in the Protein Data Bank, are also plotted (stars) (29). Concentric with this plot is a Newman diagram looking down the tryptophan C3---Cβ bond axis. The C3== C2 double bond is part of the indole pentagonal ring as is the fourth bond of C3. The χ3602,1 dihedral angle plot clearly shows that some angular ranges are preferred. Cα, and the protein backbone extending away from it (not shown), are a steric barrier to the free rotation of the bulky tryptophanyl indole, whose plane is roughly perpendicular to that formed by the protein backbone. Dihedral angles that line up the indole plane with the Cβ— Cα bond will likely lead to steric clashes as the bulky indole encounters atoms attached to the backbone and the backbone itself. These sterically hindered dihedral angle ranges appear to extend over the ranges of approximately 0° - 75°, 150° -240°, and 300° – 360° (Fig. 3). Complementary to these are two sterically 'favored' dihedral angle ranges: 75° – 150° and 240° – 300°, resulting in the elimination of possible dihedral angles for any given νW3. Likely dihedral angles for any one νW3 are now reduced to one or two choices, as can be ascertained from the plot of the νW3-χ2,1 relationship in Fig. 2a, suggesting the continued utility of the νW3 -- χ2,1 relationship as given by Eqs. (3) and (4).</p><p>Consideration of the Newman diagram makes plain the three fold cosine-dependency of the relationship given in Eq. (3). The highest νW3 is encountered whenever the indole plane coincides with one of three Cβ bonds or when χ2,1 assumes values of 0°, 120° or 240°. It is precisely at these angles that steric hindrance for the indole is greatest. Since the cosine depends on 3χ2,1, the latter three angles become equivalent to 360 ° = 0 °, where the cosine value of these is unity, and the value of the relationship assumes its maximum value. The addition of one to the cosine 3χ2,1 value is necessary to prevent subtractions from the minimal νW3 value given by the parameter, a (See Eq. (3)). The parameter, b, and the exponent value of 1.2 are empirically determined by fitting to the set of available experimental νW3 - χ2,1 data pairs.</p><!><p>Scrutiny of the νW3 - χ2,1 dihedral angle data pairs given in Table 1 and plotted in Fig. 2a shows that not all νW3 for proteinaceous tryptophans conform to the νW3-χ2,1 relationship of Eq. (3). In particular, tryptophans found in the fibrillar viral coat proteins of fd (point 1, Fig. 2a) and Pf3 (point 3, Fig. 2a), are far from the curve defined by Eq. (3). The data point for the loop hinge Trp-168 in PGA-ligated TIM (point 2, Fig. 2a) also falls into this set. Even with an estimated uncertainty of ±10° (1), these data pairs would not lie on the curve defined by Eq. (3). Indeed, the νW3 for the fd virion coat protein (1560 cm-1) has a value that is outside the range of the νW3-χ2,1 relationship (1542-1558 cm-1). Clearly, factors besides the χ2,1 dihedral angle influence νW3 in both fibrillar and globular proteins. We now consider several parameters as a suitable criterion for reliable application of the νW3-χ2,1 relationship.</p><!><p>Maruyama and Takeuchi have shown that the bond length of C2-C3 for the indole pentagonal ring increases with increases in |χ2,1| for crystalline tryptophan derivatives while the bond length of Cβ-C3 decreases under these conditions (1). Taking this cue, C2-C3, C2-N1 and Cβ-C3 bond lengths of tryptophanyl indoles that greatly deviate from the 3(cosine) function were measured, and compared to bond lengths for tryptophans whose data pairs lie on the curve at the same dihedral angle (Fig. 2a) to look for trends in bond length with respect to data points that fell above or below the νW3-χ2,1 curve. Specifically, the bond lengths for the viral coat protein, fd (Fig. 2a, point 1), are compared to those for melittin W19B and the crystalline tryptophan derivative, N-acetyl-L-tryptophan methyl ester (Ac-L-TrpME), whose data points are on the νW3-χ2,1 curve at roughly the same dihedral angle (Fig.2a). The bond lengths for the tryptophans in PGA ligated TIM (Fig. 2a, point 2) and the viral coat protein, Pf3 (Fig. 2a, point 3) are compared to those for the tryptophan in gyrase B with clorobiocin, whose data point in Fig. 2a is just above the νW3-χ2,1 curve at 22°, 1555 cm-1. These bond lengths are given in Table 2. For this limited data set, the C2-C3 bond lengths show no apparent trend for tryptophans whose data points deviate from the νW3-χ2,1 relationship The C3--Cβ bond length of the fd virion coat protein tryptophan is not significantly different from those of melittin or Ac-L-Trp ME while the C3--Cβ bond lengths of PGA-ligated TIM and Pf3 are shorter than those of gyrase B. The C2-N1 bond lengths of fd and PGA-ligated TIM are not significantly different from those of curve compliant proteins. Only the C2-N1 bond length of Pf3 shows any appreciable difference from that of curve compliant gyrase B. Indole bond lengths do not appear to be diagnostic of rule-breaking tryptophans. We next consider a specific case of disagreement between the crystallographic-determined tryptophan dihedral angle and that predicted by W3 band measurements, namely that for the PGA-ligated TIM mutant. DFT-based simulations are applied to explain the discrepancy with the expectation that these explanations may be generalized to other proteinaceous tryptophans that do not follow the νW3-χ2,1 relationship.</p><!><p>To explain the low νW3 value obtained for PGA-ligated TIM (Fig. 2A, point 2), we first calculate the vibrational modes of skatole, a tryptophan analog, in order to verify the νW3-χ2,1 relationship, and then proceed with molecular simulations of Trp-168 noncovalent interactions in the liganded and nonliganded forms of TIM. The results of the vibrational calculations indicate (Figure 4 and Table 3) that indeed we have assigned the correct simulated peak for the W3 mode. The simulated Raman data yield the following equation for νW3 and νWd3, respectively:</p><p>Where the errors associated with the intercept and coefficient values are the fitting errors. Not only have we reproduced Eqs. (1) and (2), but we have also shown that that the isotopic substitution on the indole nitrogen had very little effect on the νW3. Values of 1539.9 cm-1 for the natural Trp and 1539.4 cm-1 for 15N-Trp were obtained (Table 4). On the other hand, the unconstrained calculations for Trp-d5 yielded a νW3 downshifted from that of the natural Trp by 30 cm-1 (Table 4). This peak shift is consistent with was previously reported (1). The simulated peak positions corresponding to the W6 and W7 tryptophan modes (Table 4) were also identified using the unconstrained calculations involving Trp, 15N-Trp, and Trp-d5.</p><!><p>With the assignment of the W3 mode ascertained, we attempt to reconcile the νW3 and x-ray crystallographic dihedral angle obtained for Trp-168 in PGA-ligated TIM. We examine specifically the hydrogen bonding and around Trp-168 in TIM, followed by more generalized considerations of weaker electrostatic interactions, steric hindrance and hydrophobic interactions.</p><p>To simulate hydrogen bonding in PGA-ligated TIM, we carried out calculations involving tryptophan and three different hydrogen bonding partners. These are: 1. a phenol, to mimic the interaction between Trp-168 and Tyr-164, 2. a carboxyl, to mimic the interaction of Trp-168 with Glu-129, and 3. a water molecule. The distances between the indole N–H and the hydrogen bonding partner were varied and fixed for each set of calculations. Varying the distance simulates the change in the extent of hydrogen bonding: the closer the distance, the greater is the hydrogen bonding. The angle of the hydrogen bond was set to 180°. No other constraints were used.</p><p>The set of Trp-H2O simulations indicates that as the distance between the water molecule and the indole N–H narrows from 4 to 2 Å, the values of νW3 and χ2,1 hardly change (Table 5), and are near the values obtained for the unconstrained Trp simulation (Table 3). Values of νW3 are around 1536.7 cm-1 while the χ2,1 are about 42.4°. The upshift in νW6 as the water is placed increasingly closer to the indole N-H group indicates that the extent of hydrogen bonding increases with closer approach (8). The insensitivity of νW3 to hydrogen bond length indicates that the extent of hydrogen bonding cannot explain the observed shifts in νW3. Similar results were observed when either phenol or a carboxylic group was used as the hydrogen bonding partner. This set of observations, and the frequency positions of the hydrogen bond markers, W6 and W17 (876 cm-1, Fig. 1A) (2, 8), rule out strong hydrogen bonding as an explanation for the misfit of the PGA-bound TIM νW3 - χ2,1 dihedral angle pair to Eq. (3). This supports the earlier finding of Maruyama and Takeuchi (1) that there is no direct correlation between νNH and νW3 for the crystalline tryptophan derivatives.</p><!><p>Here, we explore the possibility that weaker, attractive electrostatic interactions influence the W3 mode. The anion-quadrapole interaction involves the electrostatic attraction between an anion and the partial positive charge surrounding an aromatic ring, here the indole ring (30, 31) while the cation-π interaction involves the attractive force between a cation and the π electron cloud on either face of the indole ring (32, 33). Typically, the anion in a proteinaceous anion-quadrapole interaction is a glutamate or aspartate while the cation in proteinaceous cation-π interactions is either arginine or lysine. As this weaker interaction is difficult to quantify in the presence of stronger hydrogen bond interactions, we look for the influence of these aromatic interactions on tryptophan discrepancies with the νW3-χ2,1 relationship by examining the x-ray crystal structures. The results for the proteinaceous tryptophans discussed above are given in Table 6. We see that for proteins with tryptophans that follow the νW3-χ2,1 relationship such as bacteriorhodopsin, gyrase B with cholorbiocin and melittin, electrostatic interactions are common and multiple. It is noteworthy that for the enzymes, TIM and PTPase, the ligand state which best conforms to the νW3-χ2,1 relationship has more than one such electrostatic interaction. While weak, the cumulative effect of these interactions can be significant (31). On the other hand, the relationship-compliant tryptophan in the model peptide, exendin-4 TC5 with a caged tryptophan, has no such electrostatic interactions, while the noncompliant tryptophan in the Pf3 virion coat peptide does. As the evidence accumulates, it becomes clearer that no one factor is the bellwether for compliance with the νW3-χ2,1 relationship. Rather, compliance appears to be based on a complex web of weak interactions that are difficult to quantify individually.</p><!><p>The homogeneous composition of the crystalline tryptophan derivatives used to formulate the νW3-χ2,1 relationship would seem to indicate a hydrophobic environment for the constituent indole rings. Hydrophobicity can be spectoscopically estimated from the intensity ratio, 1360 cm-1:1340 cm-1, which are νW7 bands, since a hydrophobic environment yields a band ratio greater than or approximately equal to one (4, 5). Estimate of this simple intensity ratio for the crystalline tryptophan derivatives is not straightforward because the W7 bands are not split into doublets but into triplets, quartets or even more component bands (2). Estimate or measurement of this W7 band ratio for the set of peptides and proteins under discussion here is somewhat more straightforward because W7 doublets are present or readily discernible from the UVRR results. W7 1360 cm-1:1340 cm-1 band ratios are given in Table 7. The ratios range in value from 0.8 to 1.7, with the fd virion peptide taking the maximum value. As the W7 band ratio for the ligated forms of both enzymes, TIM and PTPase, and horse heart cytochrome C, is lowest in value, the least hydrophobic environment for their respective tryptophans is indicated. W7 band ratios for skatole in solvents of varying polarity show that a band ratio of 0.17 is obtained in dimethyl formamide and 1.0 in benzene (4). Even at a 1360 cm-1:1340 cm-1 band ratio of 0.8, enzymatic tryptophans would seem to be in a fairly hydrophobic environment. The highest values in W7 band ratio are found for the virion coat peptides, fd and Pf3 (1.7 and 1.5, respectively), but the ratio for PGA-ligated TIM is 0.8, the minimum value found. As the νW3 - χ2,1 data points for these peptides and protein all do not fall on the νW3 - χ2,1 curve, hydrophobicity as measured by the W7 1360 cm-1:1340 cm-1 band ratio is not a good indicator of compliance with the derived relationship. Consideration of the W7 band ratio for the relationship compliant model peptides and proteins leads to the same conclusion.</p><!><p>No one molecular interaction appears to account for the noncompliance of a given tryptophan with the νW3 - χ2,1 relationship. The sum total of all weak interactions will determine whether the νW3 - χ2,1 data point for any single tryptophan will follow the relationship. It should come as no surprise, therefore, that tryptophans with an environment similar to those of the crystalline tryptophan derivatives used to define the νW3 - χ2,1 relationship will comply with it; that is, tryptophans in a constrained environment.</p><!><p>NMR study of the tryptophan in the model truncated peptide, exendin-4 TC5b (point 4, Fig. 2a), shows that Trp-26 is encased by an extensive hydrophobic network of methyl groups and prolines on both faces of the indole and about its edge (34). The conformation of Trp-26 is also stabilized by hydrogen bonds both at the indole amine and backbone carbonyl and amine (34). This environment is very much like the environment of the crystalline tryptophan derivatives used to define the νW3 - χ2,1 relationship: hydrophobic, constrained and in most cases, with hydrogen bonds to either the indole amine or the backbone carbonyl or amine. The environment of Trps-19A and B of the model peptide, melittin (points 6 and 6, Fig. 2a), is also hydrophobic and constrained. The tryptophans are aligned on the apolar face of the bent rod of 26 amino acids (35). Four peptide rods pack together in a bilayer sandwich, two rods per layer, with the hydrophobic faces packed together. This arrangement persists in the venom sack of the insect and in solution where the positive face of each rod accounts for aqueous solubility and prevents protein aggregation (35). All subunit contacts are hydrophobic, with tight interhelical packing of valine, leucine, isoleucine and tryptophan residues. Lys-23 in the same and an adjacent helix stabilize Trp-19 with cation-π interactions. Considering Trp-182 in retinal-bound bacteriorhodopsin, which also follows the νW3 - χ2,1 relationship (data point 36°, 1550 cm-1, Fig. 2a), the retinal cofactor both creates an hydrophobic environment for the indole ring and sterically hinders it via cation-π interactions (Table 6) (36).</p><!><p>The fibrillar structure of the virion coat peptides, fd (point 1, Fig. 2a) and Pf3 (point 3, Fig. 2a), suggests a constrained, crystal-like, environment for their tryptophans. However, the interpretation of fiber x-ray structures of both of these coat proteins is clouded by the lack of resolution in the data (37). Welsh et al. (38) construct several possible structural models for Pf3 and Marvin and coworkers (37) refer to the Trp-26 dihedral angle values provided by Raman studies of fd in constructing their structure (39, 40). In models of Pf3, Trp-38 is directed to the central core of the fiber, which is occupied by viral DNA. There, multiple copies of Trp-38 are thought to help neutralize the negative charge of the DNA phosphates via anion-quadrapole electrostatic interactions, or by cation-π interactions bridged by a cation (38). Several of the models for Trp-38 place it in a polar environment, which would help explain the noncompliance of Trp-38 with the νW3 - χ2,1 relationship. An earlier x-ray structure for the fd virion coat protein provided a dihedral angle of 290.1° for Trp-26 (41). With νW3 = 1560 cm-1, the data point for Trp-26 would lie very far from the νW3 - χ2,1 curve. Subsequent crystallographic study of fd (37) provides a dihedral angle of 87 °, bringing the data point for Trp-26 closer to the νW3 - χ2,1 curve (point 1, Fig. 2a). As this angle was determined by reference to the dihedral angle values provided by Raman studies of fd (39, 40), there is uncertainty in the x-ray structural position of Trp-26. More to the point, however, is the value of νW3 for Trp-26, 1560 cm-1, which is outside the possible value range of the νW3 - χ2,1 relationship given in Eq. (3), namely 1542-1558 cm-1. This narrow value range---only 16 cm-1---suggests that νW3 bandwidth could be a useful guide in empirically determining steric hindrance for tryptophan, and therefore the utility of the νW3 - χ2,1 relationship for predicting the χ2,1 dihedral angle.</p><!><p>νW3 bandwidths for several of the peptide and protein data points shown in Fig. 2a are given in Table 7. νW3 bandwidths have been equated with the FWHM value for each peak. As published spectra were only available for FWHM measurements in some cases (virion peptides and horse heart cytochrome C), the values given are only a first order approximation. The FWHM value was also estimated from published spectra for each of the crystalline tryptophan derivatives used to construct the νW3-χ2,1 relationship (2), and these are summarily reported as a range of values, 9-15 cm-1. For these tryptophan derivatives, the widest νW3 bandwidth would seem to lie just inside the maximum value of the νW3-χ2,1 relationship. Estimated FWHM values for relationship-compliant proteins, such as horse heart cytochrome C (42), fall just within the νW3-χ2,1 relationship bandwidth boundaries while those for noncompliant fd and Pf3 virion coat proteins (43) do not. More valuable are FWHM measurements that can be made on the original data, as for exendin-4 TC5b peptide and melittin. Here, the FWHM bandwidths are 16 cm-1 and 15 cm-1, respectively (Table 7). Perhaps the most instructive measurements are those made on two different states of the same protein where one state follows the νW3-χ2,1 relationship and the other does not. This is true for the liganded and nonliganded states of the enzymes, TIM and PTPase (44). The FWHM data for these enzymes provide a nice contrast because in one case, for TIM, the nonliganded state is compliant with the νW3-χ2,1 relationship while for PTPase, the liganded state is relationship compliant. The FWHM measurements in Table 7 are in agreement with the νW3-χ2,1 plot results. Data points for enzymatic states with the narrower FHWM bandwidths (nonligated TIM, FWHM =13 cm-1 and ligated PTPase, FWHM=11 cm-1) follow the relationship while states with wider FWHM bandwidths (ligated TIM, FWHM=16 cm-1 and nonligated PTPase, FWHM=17 cm-1) do not. Thus, FWHM measurements of νW3 provide an estimate of how constrained the tryptophan is with respect to the constrained environment of the crystalline tryptophan derivatives used to construct the νW3-χ2,1 relationship, and thus can provide a rough guide to the utility of the relationship in predicting the χ2,1 dihedral angle for any one tryptophan. It should come as no surprise that constraint on the motion of the indole ring of tryptophan, as measured by the FWHM of νW3, is relevant here. As discussed above and illustrated by the Newman projection in Fig. 3, the 3cosine dependence of νW3 stems from the three-fold steric hindrance encountered by the indole from moieties bonded to Cα. Clearly, this relationship would find great utility in Raman crystallographic studies where residue motion is minimized. Even with this spectroscopic guideline, the experimental value of νW3 for the fd viral coat protein still lies outside the boundaries of the νW3-χ2,1 relationship. Science is an ongoing process, and undoubtedly refinements to this relationship will be made by others, advancing the utility of Raman spectroscopy for making structural predictions.</p><!><p>The νW3 - χ 2,1 relationship (1-3) has been extended to a full, 360° rotation by plotting all dihedral angles in the same rotational direction, i.e., clockwise. As the period of the relationship given in Eq. (3) is three-fold (3cos χ 2,1) the χ 2,1 dihedral angle ranges 0° – 120°, 120° – 240°, 240° – 360°, are superimposable. It is clear from this plotting that as many as six dihedral angles correlate to a single νW3. A Newman plot of tryptophan dihedral angles gleaned from protein and tryptophan analog crystallographic data shows, however, that the χ 2,1 dihedral angle ranges, 0° - 75°, 150° -240°, and 300° – 360°, are generally avoided because of steric clashes between the indole rings and the protein backbone. This elimination of dihedral angle ranges reduces the number of possible dihedral angles for any νW3 to two or even a single angle, suggesting a general utility for the νW3 - χ 2,1 relationship. At the same time, νW3 - χ 2,1 data points for some proteinaceous tryptophans do not conform to the relationship given in Eq. (3). For one of these proteins, the mutant enzyme TIM (Trp90Tyr Trp157Phe), DFT-based calculations and simulations are used to explore the possibility that hydrogen bonding, a strong, noncovalent interaction, at the indole amine is responsible for noncompliance of the ligated TIM νW3 - χ 2,1 data point with the νW3 - χ 2,1 relationship. In agreement with Maruyama and Takeuchi (1), our simulation results show no dependence of νW3 on hydrogen bonding at the indole amine. The χ2,1 dependence on C2-C3, C2-N1 and Cβ-C3 bond lengths of tryptophanyl indoles, suggested earlier (1), is not found here for proteinaceous tryptophans that obey the νW3 - χ 2,1 relationship. The effect of weaker anion-quadrapole and cation-π interactions on νW3 is evaluated through examination of x-ray crystal structures for several proteins for which νW3 is available. This evaluation suggests that multiple, weak electrostatic anion-quadrapole, cation-π, and aromatic face-edge interactions coupled with van der Waals interactions act in a cumulative fashion to stabilize the indole side chain of tryptophan. In the absence of methods to quantify or model the cumulative effect of these interactions, the environment about the crystalline tryptophan derivatives used to formulate the νW3 - χ 2,1 relationship is considered. That environment is both constrained and hydrophobic. The hydrophobicity of the tryptophan environment can be evaluated from the W7 band ratio, 1360 cm-1:1340 cm-1, where a value greater than one indicates a hydrophobic environment (5), but band ratio measurements for the set of peptides and proteins considered here showed no correlation with the νW3 - χ 2,1 relationship. Estimation of environmental constraint for tryptophan as measured by FWHM of the W3 band showed greater success in predicting compliance with the νW3 - χ 2,1 relationship. For model peptides where the indole is stabilized through numerous noncovalent interactions and other relationship-compliant proteins, the νW3 - χ 2,1 data point conforms to the νW3 - χ 2,1 relationship. As the νW3 bandwidth of the νW3 - χ 2,1 relationship is only 16 cm-1, the FWHM of νW3 for the tryptophan under study should be no greater than this. Given this constraint, application of νW3 - χ 2,1 relationship for predicting the tryptophan χ 2,1 angle has some chance of success. Clearly, where a spectral bandwidth nearly matches the full range of the relationship used to predict a structural parameters, the error bars are large.</p><!><p>UVRR spectra for TIM in aqueous and deuterium solution with and without ligands. a. Aqueous solution. 1. ApoTIM 2. TIM ligated to PGA. Inset displays the W3 mode for apoTIM and PGA bound TIM. b. Deuterium solution 1. TIM ligated to G3P 2. TIM ligated to phosphate.</p><p>Relationship between νW3 and the dihedral angle over a full rotation. a. Data points are for crystalline tryptophan derivatives (open squares) (2), proteinaceous tryptophan residues (solid triangles, see Table 1), and model peptides (open diamonds, see Table 1) for which νW3 is available (See Table 1). χ2,1360° is given by Eq. (3) where a= 1542 cm-1 and b=7.0. As in reference (1), these parameters are empirically adjusted. Numbered points correspond to: 1. fd virion coat protein, 2. TIM mutant Trp90Tyr Trp157Phe ligated to PGA, 3. Pf3 virion coat protein, 4. the truncated exendin-4 TC5b peptide, 5, 6. melittin, two data points corresponding to dihedral angles W19A (point 6) and W19B (point 5) from crystal structure, 2MLT. b. Data points are for deuterated tryptophan derivatives (solid circles) (1), and the Pi/G3P ligated, deuterated TIM proteins (cross), which share the ligated, aqueous TIM dihedral angle given in Table 1. νWd3 taken from Fig. 1. The curve fit to this data is given by Eq. (3) where a= 1514 cm-1 and b=6.8. As in reference (1), these parameters are empirically adjusted.</p><p>Newman plot of tryptophan χ2,1 dihedral angles for proteinaceous tryptophans (triangles) and model peptides (crosses), as given in Table 1, and for crystalline tryptophan derivatives (diamonds) from (2). The three allowed (stars) tryptophan dihedral angles for an α helix, as determined from a dataset of 61 globular proteins in the Protein Data Bank, are also plotted (29). All negative dihedral angles are plotted as χ2,1360° = 360° + χ2,1. Concentric with the plot is a Newman diagram showing the disposition of the tryptophan indole C2==C3 bond with respect to the Cβ ---Cα bond along the Cα–Cβ–C3=C2 linkage.</p><p>Simulated νW3 = (1540.2 ± 1.1) + (6.245 ± 0.853)(cos 3|χ2,1|+1)1,2 (Table 2, νW3calc scaled; broken line trace and solid circles) and νWd3 = (1510.1 ± 1.1) + (6.041 ± 0.806)(cos 3|χ2,1|+1)1,2 (Table 2, νWd3calc scaled; broken line trace and solid squares) as a function of the χ2,1 torsional angles and where the errors associated with the intercept and coefficient values are fitting errors. Results of the simulations are very similar to the published data on the νW3 (solid line trace and open circles, from (2)) and the νWd3 (solid line trace and open squares, from (1)). All frequencies reported are calculated using DFT/B3LYP and 6-31g(d) with a scaling factor = 0.963 (20, 22).</p><p>W3 mode frequencies and χ2,1 torsional angle for Tryptophan in Peptides and Proteins</p><p>Austin et al., 1989 (45)</p><p>Couling et al., 1998 (46)</p><p>Hashimoto et al., 1997 (36)</p><p>Wen and Thomas Jr., 2000 (43)</p><p>Wen and Thomas Jr., 2000 (43)</p><p>This paper</p><p>Juszczak et al., 1997 (44)</p><p>Rodgers et al., 1992 (47)</p><p>Jordan et al., 1995 (42)</p><p>Skarzynski et al., 1987 (48)</p><p>Lamour et al., 2002 (49)</p><p>Lafitte et al., 2002 (50)</p><p>Grigorieff et al., 1996 (51)</p><p>Welsh et al., 1998 (38)</p><p>Marvin et al., 2006 (37)</p><p>Lolis et al., 1990 (28)</p><p>Lolis and Petsko, 1990 (25)</p><p>Stuckey et. al, 1994 (52)</p><p>Fauman et al., 1996 (53)</p><p>Safo et al., 2002 (54)</p><p>Fermi, G. et al., 1984 (55)</p><p>Bushnell et al., 1990 (56)</p><p>Terwilliger et al., 1982 (35)</p><p>Neidigh et al., 2001 (34)</p><p>Indole bond lengths for Tryptophan-containing Proteins and Derivatives</p><p>All bond lengths measured from corresponding Protein Data Bank file; see Table 1 for corresponding file; pm=picometer.</p><p>Observed and calculated W3 mode frequency for Trp (νW3) and Trp-d5 (νWd3) as a function of the |χ2,1| torsional angle.</p><p>From (2)</p><p>Based on DFT/B3LYP and 6-31g(d) calculations.</p><p>Scaling factor = 0.963 (20, 22)</p><p>From (1)</p><p>Isotopic dependence of the calculated W3 mode frequency for Trp (νW3) and χ2,1 torsional angle.†</p><p>All values presented are based DFT/B3LYP and 6-31g(d) and scaling factor = 0.963 (20, 22).</p><p>Calculated frequencies of some tryptophan modes and χ2,1 torsional angle for the Trp-water complex. The effect of hydrogen bonding and steric strain. †</p><p>All values presented are based DFT/B3LYP and 6-31g(d) and scaling factor = 0.963 (20, 22).</p><p>Noncovalent Interactions for Tryptophan in Several Proteins</p><p>(57)</p><p>W3 Bandwidth for Tryptophans in Peptides and Proteins</p><p>density functional theory</p><p>dihydroxy acetone phosphate</p><p>full width half maximum</p><p>glyceraldehyde phosphate</p><p>glycerol 3-phosphate</p><p>protein tyrosine phosphatase</p><p>2-phosphoglycolate</p><p>phosphate</p><p>Protein Data Bank</p><p>triose phosphate isomerase</p><p>ultraviolet resonance Raman spectroscopy</p><p>This work was supported by National Institutes of Health grants GM08153 (RZBD), 5P01GM068036, EB01958, R01 EB-00296 and the W. M. Keck Foundation.</p>

PubMed Author Manuscript

Expanded ensemble methods can be used to accurately predict protein-ligand relative binding free energies

Alchemical free energy methods have become indispensable in computational drug discovery for their ability to calculate highly accurate estimates of protein-ligand a nities. Expanded ensemble (EE) methods, which involve single simulations visiting all of the alchemical intermediates, have some key advantages for alchemical free energy calculation. However, there have been relatively few examples published in the literature of using expanded ensemble simulations for free energies of protein-ligand binding.In this paper, as a test of expanded ensemble methods, we computed relative binding free energies using the Open Force Field Initiative force field (codename "Parsley") for twenty-four pairs of Tyk2 inhibitors derived from a congeneric series of 16 compounds.The EE predictions agree well with the experimental values (RMSE of 0.94 ± 0.13 kcal mol 1 and MUE of 0.75 ± 0.12 kcal mol 1 ). We find that while increasing the 1 number of alchemical intermediates can improve the phase space overlap, faster convergence can be obtained with fewer intermediates, as long as the acceptance rates are su cient. We find that convergence can be improved using more aggressive updating of the biases, and that estimates can be improved by performing multiple independent EE calculations. This work demonstrates that EE is a viable option for alchemical free energy calculation. We discuss the implications of these findings for rational drug design, as well as future directions for improvement.

expanded_ensemble_methods_can_be_used_to_accurately_predict_protein-ligand_relative_binding_free_ene

Introduction<!>Methods<!>Preparation of structures and force field parameters for Tyk2 and its inhibitors<!>Molecular simulation protocol<!>Results<!>Accurate predictions of RBFEs for Tyk2 inhibitors<!>Averaging multiple independent EE estimates is more accurate than a single estimate<!>Discussion<!>Conclusion<!>Code and data availability

<p>Over the last decade, alchemical free energy methods have increased in accuracy and computational e ciency to become the dominant modeling approach for computing high-quality estimates of ligand binding free energy. [1][2][3][4] Particularly popular have been relative binding free energy methods, which can typically deliver accuracies within 1 kcal mol 1 . [5][6][7] Unlike absolute free energy methods, which require the complete decoupling of a molecule's nonbonded interactions, relative free energy methods only require alchemically transforming the set of atoms that di↵er between two molecules, resulting in better phase space overlap between alchemical intermediates. This e cient approach is especially useful for structurebased computational lead optimization, where one frequently makes make small changes to a single sca↵old, and many tools are now widely available to automate and perform these calculations. [8][9][10][11][12][13] The expanded ensemble (EE) method, similar to simulated tempering, is an algorithm in which multiple thermodynamic ensembles are adaptively sampled in the same simulation. 14 Specific variants of this adaptive approach include self-adjusted mixture sampling (SAMS) 15,16 and the accelerated weight histogram method. 17 Most other free energy approaches rely on multiple parallel or coupled simulations performed for each alchemical intermediate, as in Hamiltonian replica exchange, 18,19 Therefore, the ability to sample multiple ensembles in a single simulation replica is attractive for many applications, such as enabling large-scale virtual screening on distributed parallel computing platforms. [20][21][22] EE methods, however, have not been widely adopted for estimating ligand binding free energies per se, although much work has been done towards improving adaptive expanded ensemble estimates for calculating free energies from perturbed Hamiltonians. [23][24][25][26] Of the handful of studies that have previously used EE for ligand binding, all have focused on the problem of computing absolute binding free energies (ABFE). 22,27,28 In the SAMPL4 host-guest challenge, Monroe et al. showed that EE yields estimates comparable to other methods, in a system where molecular flexibility and multiple binding modalities are important. 27 Rizzi et al. report similarly results in the recent SAMPL6 host-guest challenge. 28 To our knowledge, no group has published an example of relative binding free energy (RBFE) estimates predicted using EE. In theory, expanded ensemble simulations are particularly well-suited for absolute binding free energies, as the ligand is able to freely bind and unbind, sampling di↵erent binding modes. However, even with relative binding free energies, the available phase space changes between ligands, meaning there is likely to be some sampling advantage even here. Given the better thermodynamic overlap of relative vs. absolute methods, we hypothesized that EE calculations would perform well at this task.</p><p>To see if this was the case, we examined the performance of EE in estimating RBFE for a set of twenty-four pairs of Tyk2 inhibitors (Figure 1) selected from a congeneric series of 16 compounds designed by Liang et al. 29,30 Below, we present our methodology and results, which suggest that EE can predict RBFEs accurately and e ciently, given appropriately chosen protocols. We explore the sensitivity of prediction accuracy and adaptive convergence to a number of di↵erent parameters and analysis choices, and discuss some possible improvements of the algorithm for the future. Tyk2 is shown in complex with ligand ejm 31 in Table 1. 29</p><!><p>The expanded ensemble method Consider a set of N thermodynamic ensembles parameterized by , where ranges between 0 and 1. For a set of i values indexed by i = 1, ..., N, each ensemble is defined by a reduced potential energy function u i (x) = U (x| i ), where = (k B T ) 1 . The goal of the expanded ensemble (EE) method is to use Monte Carlo sampling to perform a random walk in -space, throughout a single simulation, where all thermodynamic states defined by di↵erent values of are uniformly sampled, or alternately, satisfy some other desired distribution as a function of . This is achieved through the use of configuration-independent bias potentials fi , that modify each potential energy function as In this study, to adaptively refine estimates of fi , the Wang-Landau (WL) flat-histogram algorithm is used. 31 This algorithm periodically evaluates a histogram of values h i storing the number of times state i is visited. At each iteration t, the histogram and the free energy estimates fi are updated:</p><p>The quantity is called in this study Wang-Landau (WL) increment, and it is initially set to a large value (for example, 10 k B T ). This has the e↵ect of penalizing visits to state i so that subsequent MC moves to other states will more likely be accepted. When the histogram is su ciently flat, the WL increment is scaled by a factor ↵ < 1, and the histogram counts h i are reset to zero. The histogram is su ciently flat if the ratio N ratio of all histogram values h i to the mean value h = 1 N P i h i is su ciently close to 1. This is determined by ensuring that both N ratio > ⌘ and 1/N ratio > ⌘ for all values of i, for some value of ⌘ (for example, 0.8) which we will call the Wang-Landau (WL) ratio. [32][33][34] Relative binding free energy calculations In RBFE calculations, two separate EE simulations are needed to predict relative free energy of binding, G for a pair of ligands L and L ⇤ . In one simulation, the ligand L is alchemically transformed to L ⇤ in aqueous solution to obtain G L . In the other simulation, a receptorbound ligand RL is transformed to RL ⇤ to obtain G RL . The RBFE can then be obtained as G = G RL G L (Figure 2).</p><!><p>Structure preparation and initial coordinates. The Tyk2 dataset is part of the Schrödinger's "JACS set" 5 and was used in several RBFE benchmark studies. 5,6,35,36 The receptor model is based on the X-ray structure of PDB ID 4GIH (resolution 2.00 Å, Open-Eye Iridium score 0.5, highly trustworthy). 29 The preparation and initial coordinates of the protein are the same as in Gapsys et al. 6 , where it was prepared with the Protein Prepara-tion Wizard 37 using default settings: missing atoms, sidechains, and loops were modelled, protein protonation states were assigned with PROPKA at pH 7.0, the hydrogen bonding network was optimized. To relieve local clashes, a restrained minimization was performed with a 0.5 Å heavy-atom RMSD displacement cut-o↵. The inhibitors were modelled in their neutral form according to their protonation state at pH 7. For the construction of their initial coordinates, the coordinates of the crystallographic ligand ejm 46 (see Table 1) were used and all other ligands were flexibly aligned to the reference ligand for improving the 3D-overlay of their Bemis-Murcko sca↵olds.</p><p>Generation of Force Field Parameters. The receptor was parameterized using the AMBER ↵99sb*ILDN force field parameters [38][39][40] using the GROMACS gmx tool pdb2gmx</p><p>and the AMBER ↵99sb*ILDN parameter files available in the pmx distribution. 41 The ligands were parameterized and prepared using a workflow 42 based on the Open Force Field toolkit, 43 the pmx toolkit 41,44 and GROMACS gmx program suite. The parameter set employed was the Open Force Field version 1.0.0 (codenamed "Parsley"). 45 Hybrid structures and topologies for the ligand pairs were generated using pmx 41,44 following a single topology approach. The workflow establishes a mapping between atoms of the two ligands based on the maximum common substructure and conformational alignment. Polar hydrogens are not mapped to each other to decrease the influence of the perturbation on the hydrogen bond network. The mass of atoms in the non-interacting state ("dummies") was set to 12u.</p><p>Covalent force field parameters between atoms in the dummy state are not changed during the perturbation. The mappings are illustrated in the Supplementary Information Figure The RL boxes had a diameter of 9.6 nm.</p><!><p>Simulations were performed on the Owlsnest and CB2RR High-Performance Computing clusters at Temple University, and TACC Stampede (XSEDE). Molecular dynamics production runs were performed using the expanded ensemble functionality of GROMACS 5.1.4 46 (GROMACS/EE). Solvated systems (L and RL) were energy minimized using GROMACS with 50,000 steepest descent steps. Equilibration was performed Verlet integration was performed in the NPT ensemble at 300K using a 1 fs timestep. For each system, L or RL, 2.5 ns NPT equilbirium simulation was generated. The pressure was kept at 1 bar using the Parrinello-Rahman barostat. Long-range electrostatic interactions were handled by Particle Mesh Ewald (PME).</p><p>EE perturbation calculations were performed at 300 K in the NVT ensemble using Verlet integration with a 1 fs timestep and a velocity-rescaling thermostat. Long-range electrostatics were modeled using PME, and long-range dispersion correction was used.</p><p>GROMACS/EE parameters. The Metropolized Gibbs algorithm 32 implemented in GRO-MACS/EE was used to perform Monte Carlo sampling, with moves proposed every 500 time steps. This method proposes moves from the current state i to all states j 6 = i with Gibbs probability exp( u j )/ P k exp( u k ), and with a rejection step to satisfy detailed balance. Metropolized Gibbs sampling has been proven to enhance the mixing rate. 32 The initial WL increment was set to 10.0 k B T . The WL free energy estimates fi were updated every Monte Carlo move. Two values were explored for the WL scaling factor ↵: 0.8 (default) and 0.5. Modification of weights was set to be discontinued when the WL increment reached a value of of 10 5 k B T , with the assumption that they weights are su ciently close to equilibrium at that point, however, none of the simulations in this study reach this limit.</p><p>Data for free energy estimation was collected after the WL increment fell below 0.01 k B T . We chose this threshold based on the empirical observation that free energy estimates begin to converge beyond this value (see Results). Final free energy estimates were made by the averaging the values of f1N = fN f1 collected after the WL increment fell below 0.01 k B T . 47 Our default protocol was to collect an aggregate simulation time of 400 ns for each solvated ligand transformation (L ! L ⇤ ), and 100 ns for each receptor-bound complex (RL ! RL ⇤ ). We chose these trajectory length to achieve ⇠80% of the sampling with a WL increment below 0.01 k B T .</p><p>The modified potential used for each alchemical intermediate</p><p>, where i ranges from 1 = 0 to N = 1. The U 0 (x) term represents potential energy terms not coupled to the ligand, while U L (x| i ) and U L ⇤ (x| i ) are potentials energy terms coupled to ligands L and L ⇤ , respectively, which depend on i through the use of soft-core potentials (sc-alpha = 0.5, sc-power = 1, sc-sigma = 0.3).</p><p>Several numbers of alchemical intermediates were chosen for comparison:</p><p>Table 1: Twenty-four pairs of Tyk2 inhibitors. The numbering of the inhibitors corresponds to the numbering in References. 29,30 The naming convention "ejm" denotes they are from Reference 29 and "jmc" denoted they are from Reference.</p><!><p>EE convergence times vary with WL increment scaling and numbers of alchemical intermediates.</p><p>As an illustrative example, we show how the EE algorithm converges for L and RL calculations of the ejm 31 ! ejm 45 alchemical transformation (Figure 3).</p><p>Inspection of the instantaneous estimates of free energy estimates for G L and G RL over time show occasional sudden deviations and corrections that occur as the ligand undergoes slow conformational changes. In the example (Figure 3), we can trace this behavior to slow interconversion between torsional states, which temporarily "switches" the system to a new e↵ective free energy landscape (Figure S6). This behavior is observed in all simulations, regardless of the number of alchemical intermediates. Although enhanced sampling over these slow sampling barriers is not specifically addressed by the EE algorithm, we nevertheless observe adequate convergence of RBFE calculations for all of the systems considered in this study. Using fewer alchemical intermediates can generally converge EE estimates faster, because the flat-histogram criteria can be achieved in a smaller amount of simulation time.</p><p>Another parameter that a↵ects convergence is the WL scaling factor. We performed tests on seven di↵erent alchemical transformations (the first seven pairs of inhibitors in Table 1), using two di↵erent values for the scaling factor: 0.5 (more aggressive), versus 0.8</p><p>(the GROMACS default). These tests used 400 ns simulation data used for each L ! L ⇤ transformation, and 150 ns for each RL ! RL ⇤ transformation. The only exception was ejm 31 (RL) ! jmc 28 (RL ⇤ ), which required ⇠190 ns to converge.</p><p>The results show that while the accuracy of the predicted G values using a scaling factor of 0.5 is statistically indistinguishable from results using 0.8 (Figure 4), using a scaling factor of 0.5 decreases the convergence time (the time it takes for the WL increment to reach 0.01 kT ) by about a factor of 2, with ligand-only simulations converging within an average time of 52 ns. Interestingly, despite the more complex protein environment, receptor-ligand simulations converged within an shorter average time of 36 ns (Figure 4B and 4D). These results suggest -at least for these RBFE calculations and with many intermediates -that more aggressive WL scaling may be more e cient.</p><!><p>To test the convergence and performance of our EE protocol, we calculated RBFE estimates from 400 ns of ligand-only trajectories, and 100 ns of receptor-ligand trajectories, for all 24 alchemical transformations of Tyk2 ligands. Informed by the above results, we used a WL scaling factor of 0.5 in all simulations. The predictions agree well with the experimental values, achieving an RMSE of 0.95 ± 0.17 kcal mol 1 and MUE of 0.72 ± 0.13 kcal mol 1</p><p>(RMSE and MUE uncertainties from 1000 bootstrapped samples of the set of ligands), The convergence profiles of the simulations, however, suggest that shorter trajectories might su ce for making accurate predictions. We calculated "normalized" convergence profiles of G L and G RL estimates as a function of simulation time T , as</p><p>The normalized profiles all converge to unity, enabling objective comparison. The profiles appear to equilibrate by around 150 ns (Figure 5). On average, ligand-only systems reach WL increments of 0.01 kT in about 100 ns (Figure 5B), while receptor-ligand systems reach this value on average in about 40 ns (Figure 5D).</p><p>To assess how the length of EE trajectory a↵ects the accuracy of the predictions, we and 400 ns, while using 100 ns of receptor-ligand trajectory data. We find that using 100 ns of trajectory data increases the RMSE to 0.98 kcal mol 1 and MUE to about 0.77 kcal mol 1 , but beyond this, estimates approach those using all 400 ns of the trajectory data (Figure 6, blue lines).</p><p>When making free energy estimates from simulation trajectory data, it is imperative to properly consider the extent of time correlation in the trajectory to make accurate estimates of uncertainty. 49,50 Since EE methods make inherently history-dependent estimates, subsampling trajectory data to remove time correlation may be an important consideration.</p><p>To determine a statistically optimal subsampling interval, we computed the autocorrelation time ⌧ c of the estimate f1N (t) as ⌧ c = R T T 0 g(⌧ )d⌧ where S1).</p><p>As expected, we find that while subsampling increases the values of computed uncertainties, making these uncertainties more accurate because time-correlated data is removed, it does not statistically impact the overall accuracy of estimated RBFEs (Figure 7). Using 150 ns of ligand-only trajectory data and 100 ns of receptor-ligand trajectory data, we find that across all 24 alchemical transformations of Tyk2 inhibitors, using unsubsampled data results in an RMSE of 0.93 ± 0.14 kcal mol 1 and MUE of 0.72 ± 0.12 kcal mol 1 , while subsampled data results in an RMSE of 0.94 ± 0.13 kcal mol 1 and MUE of 0.75 ± 0.12 kcal mol 1 .</p><p>We also checked to see whether there was any significant conformational change in the receptor during RL simulations. To assess this, we computed the RMSD of the protein backbone coordinates with those of the first frame, over time. We find that all RMSD values are less than 0.3 nm, suggesting the protein receptor is relatively stable throughout the simulation (Figure S8). We also assessed whether backbone restraints had any e↵ect on the convergence time or the accuracy of the RBFE predictions. Therefore, we repeated all 24 RL ! RL ⇤ transformations using 100 ns trajectories, in the presence of backbone restraints with a force constant of 1000 kJ mol 1 nm 2 (Figure S7). While the calculated free energies for restrained vs.</p><p>unrestrained simulations are statistically indistinguishable, we find that backbone restraints decrease the convergence time, from an average convergence time of 32.8 ns (unrestrained), to an average convergence time of 20.2 ns (restrained).</p><!><p>There are several reasons why independent EE simulations will give varying estimates of free energies. The first is uncertainty due to finite sampling as occurs for all sampling algorithms. The second is that the Wang-Landau flat-histogram algorithm, as implemented in GROMACS with geometric scaling of the WL increment, can result in saturation of the error, where the WL increment may become vanishingly small but the biases nevertheless converged to the incorrect value in the limit of infinite sampling. 51,52 While there are ways to avoid saturation of the error, such as using "1/t"-scaling of the WL increment, 51 this algorithm is not implemented in GROMACS, and we do not pursue the issue here.</p><p>To examine the reproducibility of RBFE estimates from EE, we performed five independent EE simulations for each of the 24 alchemical transformations of Tyk2 inhibitors. We found that the standard deviation of the estimates varies between 0.14 and 0.83 kcal mol 1 , suggesting that EE estimates for at least these systems are reproducible (Figure 8).</p><p>An interesting question is whether the uncertainty estimates i from each single EE simulation (estimated from the variation in f1N ) are able to capture the run-to-run variation in G estimates observed for multiple calculations. If so, such uncertainties could be trusted to provide estimates of the overall uncertainty in EE predictions. To see if this was the case, we compared the uncertainties in G estimated using error propagation from single replicas ( P E = q</p><p>i ) to the uncertainties estimated from the standard error of the mean across the five replicas, computed as the corrected sample standard deviation, 2 , where N = 5, and G is the mean of G estimates across all replicas (Figure S10). We find that, across the 24 transformations, the computed uncertainties P E tend to be slightly larger than those estimated from the standard error of the mean, SEM , by about 0.22 kcal/mol (Figure S11). This suggests that the uncertainty estimates from single EE replicas do a reasonable if not perfect job of estimating the overall uncertainty in G estimates, and can be prudently used for this purpose.</p><p>If we make the assumption that the force field is su ciently accurate, then improved sampling will improve accuracy. We find in this case that prediction accuracy is improved when RBFE estimates are made by averaging the results of multiple replicates. Averaging five replicas achieves an RMSE of 0.83 ± 0.14 kcal mol 1 and a MUE of 0.66 ± 0.11 kcal mol 1 ; these values are smaller than any of the single-replicate EE estimates. (Figure 9)</p><p>To assess the statistical significance of this change in accuracy, we computed a nonparametric null distribution for the RMSE and MUE statistics in each of the above cases, Figure 8: Comparisons of RBFEs calculated from five independent replicates, for all 24 alchemical transformations of Tyk2 inhibitors. Shown are five independent estimates and their uncertainties (blue, orange, green, red, purple circles), and the average of five replicas (brown circles). The magneta horizontal lines represent the experimental values.</p><p>using 1000 permutations of the experimental labels. In all cases, we find p-values less or equal to 0.006 (Figure S9), which suggests that these improved accuracy results are statistically significant.</p><p>To further explore the apparent accuracy gains from averaging multiple replicas, we computed predicted RBFEs for a test set of 16 Tyk2 inhibitor alchemical transformations (those with ligand-only simulations that converged within 100 ns), using n replicas of length (400 ns)/n for L simulations and (160 ns)/n for RL simulation, for n = 1, 2, 3 and 4 (Figure 10) . The results of these calculations suggests that, given a fixed amount of available simulation time, lower RMSE and MUE can be achieved by averaging the results of n replicas, versus performing one simulation n times as long, in the case when trajectories are long enough to reach convergence.</p><!><p>In this work, we have shown that expanded ensemble methods, coupled with the latest Open Force Field Initiative force field, can accurately predict relative binding free energy predictions for Tyk2 inhibitors. The main computational expense of these calculations is the simulation time required for convergence of the relative binding free energies, which depends on the number of alchemical intermediates used. Compared to a typical FEP+ protocol 5 using 20 alchemical intermediates and 5 ns of sampling per intermediate (100 ns of total simulation time each for L and RL), the EE approach used here is only slightly more expensive, but can be run in fully independent runs. In many situations, we expect the advantages of performing such simulations without parallelizations across replicas may outweigh the disadvantages of computational cost.</p><p>Most of the results we have presented in this work use a large number of alchemical intermediate (109), despite the fact that we have demonstrated that fewer alchemical intermediates (e.g. 21) leads to faster convergence. Why use so many intermediates? One reason is that it is very hard to predict the optimal schedule of i values without performing preliminary simulations. In practice, using more values is simple and e↵ective. A disadvantage to this strategy is the longer simulation time it takes to achieve flat histograms. As we show above, however, more aggressive WL increment scaling (using a scaling factor of 0.5 rather than 0.8) can help accelerate convergence in this case.</p><p>Another reason to use a large number of alchemical intermediates is to avoid the risk of sampling bottlenecks. When the MC acceptance probability between two intermediates becomes low enough, we have observed that a kind of hysteresis can develop, where the more time spent in a given intermediate, the less probable it is to make outgoing transitions. The net e↵ect is that EE sampling appears to "get stuck" intermittently at particular values of i , for increasingly longer periods of time. In future work we are studying this phenomenon and its relation to the saturation of the error that is known to occur with the current geometric scaling of the WL increment. 51,52 It should be possible to determine conditions under which such hysteresis will occur, so steps can be taken to avoid it.</p><p>There are several future directions to explore that might help improve sampling and convergence beyond the EE approach we have pursued here. Algorithms such as the accel-erated weight histogram (AWH) method 17 and self-adjusted mixture sampling (SAMS), 15 which modify biases for multiple intermediates simultaneously using more sophisticated estimators, may lead to better convergence. Versions of these algorithms have recently been implemented in GROMACS and can be readily applied. More frequent swapping attempts (every 50 steps, perhaps) might also lead to faster convergence. Moreover, it is clear that "hidden barriers", which arise due to rare conformational transitions (see Figures 3 and S6), are responsible for slow convergence. Future work should focus on automatically identifying the conformational degrees of freedom corresponding to these slow motions, and employing novel alchemical intermediates to help overcome these barriers.</p><p>A key lesson from this study is that averaging EE free energy estimates over multiple short simulations gives more accurate predictions than a single long simulations, presumably due to better sampling. This finding suggests strong benefits for using EE to perform massively parallel virtual screening e↵orts on cloud computing and other distributed platforms on which tight-coupling algorithms like Hamiltonian replica exchange are unfeasible. EE approaches are feasible on such platforms, with the additional benefit that better accuracy may be achieved by performing more simulations with shorter trajectory lengths.</p><!><p>In this work we use an expanded ensemble (EE) method, implemented in GROMACS, 46 along with the current Open Force Field potential, 45 to accurately predict the relative binding free energies (RBFEs) of twenty-four Tyk2 inhibitors (RMSE of 0.94 ± 0.13 kcal mol 1 and MUE of 0.75 ± 0.12 kcal mol 1 ). We found that EE convergence times can be accelerated by decreasing the number of alchemical intermediates (provided su cient MC acceptance can be maintained) and by using a more aggressive Wang-Landau scaling factor of 0.5. We also find a statistically significant benefit in estimating RBFEs as the average over multiple independent EE replicates. These results suggest that while EE methods may be currently underutilized for RBFE estimation, they are poised to play a bigger role in virtual screening especially on large-scale cloud computing platforms.</p><!><p>Input files, mdp files and examples of ligand and receptor-ligand EE simulations can be found on GitHub at https://github.com/Sizhang92190/RBFE_EE_TYK2</p>

ChemRxiv