A Multiplexing Activity-Based Protein-Profiling Platform for Dissection of a Native Bacterial Xyloglucan-Degrading System

Bacteria and yeasts grow on biomass polysaccharides by expressing and excreting a complex array of glycoside hydrolase (GH) enzymes. Identification and annotation of such GH pools, which are valuable commodities for sustainable energy and chemistries, by conventional means (genomics, proteomics) are complicated, as primary sequence or secondary structure alignment with known active enzymes is not always predictive for new ones. Here we report a “low-tech”, easy-to-use, and sensitive multiplexing activity-based protein-profiling platform to characterize the xyloglucan-degrading GH system excreted by the soil saprophyte, Cellvibrio japonicus, when grown on xyloglucan. A suite of activity-based probes bearing orthogonal fluorophores allows for the visualization of accessory exo-acting glycosidases, which are then identified using biotin-bearing probes. Substrate specificity of xyloglucanases is directly revealed by imbuing xyloglucan structural elements into bespoke activity-based probes. Our ABPP platform provides a highly useful tool to dissect xyloglucan-degrading systems from various sources and to rapidly select potentially useful ones. The observed specificity of the probes moreover bodes well for the study of other biomass polysaccharide-degrading systems, by modeling probe structures to those of desired substrates.

Cellvibrio japonicus is a saprotrophic bacterium which possesses the remarkable ability to grow in isolation using a variety of hemicellulosic polysaccharides as sole carbon sources. 18- 20 t produces a diverse collection of GHs, assembling enzyme systems that can degrade cellulose, xylans, mannans, and xyloglucans, among others. 21,22In contrast to polysaccharide utilization loci found in many gut bacteria 5 , C. japonicus polysaccharide-degrading systems are not tightly organized into complete substrate-specific gene clusters, complicating the enumeration of the components of a complete enzyme system.Recent transcriptomic work identified a cluster of four genes within the C. japonicus genome, three of which are essential for xyloglucan oligosaccharide (XyGO) saccharification. 21However, this gene cluster does not include any apparent xyloglucanase or exo-β-glucosidase, or any other possible accessory activities.Recombinant production and characterization of homology-selected putative xyloglucanases in the C. japonicus genome identified Cel5D, Cel5E, Cel5F, and CjGH74A as specific xyloglucanases, but the roles of each of these remain unclear. 23,24Among exo-βglucosidases, Cel3D was found to be xyloglucan oligosaccharide-specific. 20Yet knocking out these genes only generated a mild growth phenotype, suggesting the presence of additional compensating enzymes.
https://doi.org/10.26434/chemrxiv-2023-lxghmORCID: https://orcid.org/0000-0002-7343-776XContent not peer-reviewed by ChemRxiv.License: CC BY-NC 4.0 With the aim to annotate these compensating exo-glycosidases, and to allow for rapid discrimination between cellulose-and xyloglucan acting endo-glycosidases, we developed a multiplexing activity-based protein profiling (ABPP) assay, the results of which are presented here.0][31][32][33][34][35][36][37][38] More so than other GH-directed probe designs, which are often limited in activity and/or GH selectivity, cyclophellitol-based ABPs are viable tools to assess polysaccharide-induced microbial secretomes for desirable activities, which can then be selected for further annotation.Besides targeting a single GH within a biological system, ABPP assays can be executed in a multiplexing format 37,38 , allowing dissection of complex enzyme systems such as that of the xyloglucan degradome of C. japonicus studied here.The work presented here comprises the design and validation of trisaccharidic xyloglucan ("XyG")-type cyclophellitol probes and their validation as bona fide, predictive tools for the identification of xyloglucanase activities and their discrimination from cellulase activities within a xyloglucan-elicited C. japonicus degradome.The XyG probes complement our previously described [29][30][31][32][33][34][35][36][37][38] suite of exo-and endo-GH probes, which we combined to investigate the time-dynamic and substrate concentration-dependent expression of xyloglucanases, cellulases and retaining -exo-glucosidases in secretomes obtained from C. japonicus grown on xyloglucan, xyloglucan oligomers and other polysaccharide food sources.In this way and using both in-gel detection (with fluorescent probes) and proteomics annotation (with biotinylated probes), we revealed Cel5D and Cel5F to be the exclusive specific retaining xyloglucanases (inverting glycosidases cannot be detected with cyclophellitols), filling distinct functional niches, with Cel5C being a cellulase and Cel3A, Cel3B, and Cel3D acting as -exoglucosidases.Utilization of -L-arabinofuranose, -galactopyranose, -xylopyranose and xylopyranose configured cyclophellitol probes allowed further in-depth dissection of C. japonicus secretomes and lysates.Our results provide a blueprint for designing multiplexing ABPP assays for the rapid profiling of secretomes of microorganisms grown on specific polysaccharide materials and in which the probes are designed to represent structural elements of the carbohydrate source, and therefore activities of the corresponding retaining GHs.

Results
Assembly and validation of the suite of activity-based probes.Xyloglucan (XyG), the primary carbohydrate source used in this study, contains α-(1,6)-xylose branches at the +2 and/or +3 glucose residues of linear tetraglucose stretches.Other branching sugars include -(1,2)galactopyranose, α-L-(1,2)-fucopyranose, α-L-(1,2)-arabinofuranose and, in rare circumstances, β-(1,2)-xylopyranose 39 .These structures (with the exception of fucose) are captured in the set of mono-, di-and trisaccharidic cyclophellitol probes as depicted in Figure 1A.These ABPs react with their target GH in a mechanism-based fashion to form a covalent and irreversible enzyme-inhibitor adduct as depicted for inactivation and tagging of retaining -exo-glucosidases in Figure 1B.Variation in configuration and substitution pattern yields probes designed to target xyloglucanases (denoted as ABP-XyG), cellulases (ABP-Cel) 38 , retaining -exo-glucosidases (ABP-Glc) 30 , retaining -exo-galactosidases (ABP-Gal) 40 , retaining -exo-xylosidases (ABP-Xyl) 37 , retaining -exo-xylosidases (ABP-Xyl) and retaining -L-exo-arabinofuranosidases (ABP-Araf). 31The probes were prepared in fluorescent form bearing either a Cy5 dye (denoted with the extension Cy5), a Cy3 dye (Cy3) or a Bodipy-FL dye (FL) to allow for in-gel multiplexing ABPP detection.For the purpose of kinetic measurements, probes were prepared with simple azide (N3) tags.For the purpose of target GH identification by pull-down/mass spectrometry proteomics, all probes were also prepared in biotinylated form (Bio).With the exception of the ABP-XyG and ABP-Xyl probes, the synthesis and labeling efficacy of all probes on recombinant and/or isolated GHs, as well as detection of these in complex biological samples, have been reported previously. 30,31,37,38,40e full structures of all probes and the synthesis of the ABP-XyG and ABP-Xyl probes is given in the Supporting Information.In order to validate the ABP-XyG probes for profiling xyloglucanases in complex biological samples we first established the potency and mode of action of ABP-XyG-N3 as inhibitor of various previously characterized recombinant xyloglucanases.Incubation of pure recombinant Bacteroides ovatus BoGH5A 41 , Paenibacillus pabuli PpXG5 13 , and Cellvibrio japonicus CjCel5D 23 with 100 μM ABP-XyG-N3 for 1 hour under optimal activity conditions gave near-quantitative labeling as assessed by intact protein mass spectrometry analysis, while identical treatment with ABP-Cel-N3 gave minimal labeling (Figure 1A, Supplemental Figure 1-3).Pre-treatment of ABP-XyG-N3 and ABP-XyG-Cy5 with 0.1 mg/mL BoGH31 αxylosidase for 1 hour at 37 °C caused no significant change in labeling behavior and no detectable formation of a dexylosylated species by LC-MS, indicating that ABP-XyG-Cy5 and ABP-XyG-N3 are resistant to exo-hydrolase activity (Supplemental Figure 4).This is consistent with the known recognition mode of xyloglucan-specific α-xylosidases, requiring an unsubstituted non-reducing chain terminus. 42,43ay diffraction data of the complex between CjCel5D and ABP-XyG-N3 shows near-perfect mimicry of the known glycosyl enzyme intermediate state in the -1 and -2 subsites (Figure 2B).Irreversible inhibition kinetics, measured using bespoke 4-methylumbelliferyl (4MU) and 6-chloro-4-methylumbelliferyl (6C4MU) XXXG fluorogenic substrates (Figure 2C, D and Supplemental Figure 5-25; see supporting information for synthetic details and Tuomivaara et al. for oligosaccharide nomenclature) 9 , showed probe selectivity values ((ki,XyG-N3/KI,XyG-N3)/(ki,Cel-N3/KI,Cel-N3)) ranging from >17 for CjCel5D to 0.13 and <0.0026 for HiCel7B 44 and BaCel5A 45 , two well-known cellulases, respectively (Table 1).Table 1.Kinetic parameters for covalent inhibition of endo-glucanases by ABP-Cel-N3 and ABP-XyG-N3.Where it was not possible to obtain distinct kinact and KI parameters at the inhibitor concentrations tested, and the combined kinact/KI parameter is shown for these cases.ND: not determined.Specificity as determined from the ((kinact,XyG-N3/KI, ABP-XyG-N3)/(kinact,Cel-N3/KI, Cel-N3) values.To dissect the native xyloglucan-degrading system of C. japonicus, we prepared "primed" cells by growth on glucose to carbon-limited saturation in MOPS minimal medium (see supplemental methods for details). Sbsequent dilution into medium containing glucose, cellobiose, tamarind xyloglucan, or wheat arabinoxylan was hypothesized to reveal substratespecific responses.Secretome, intact cell, and lysate samples from each culture were treated with a triplex probe mixture containing ABP-Glc-FL, ABP-Cel-Cy3 and ABP-XyG-Cy5 (Figure 3A, and Supplemental Figure 26 for Coomassie stain).We observed strong and uniquely xyloglucan-induced production of a ~65 kDa ABP-XyG-Cy5-selective outer membraneassociated enzyme.ABP-Glc-FL treatment revealed two xyloglucan up-regulated glucosidase bands at ~120 and ~58 kDa, and ABP-Cel-Cy3 treatment shows a cellulase band at ~40 kDa also observed in the cellobiose culture.Surprisingly, a major ~75 kDa secreted band from growth on arabinoxylan reacted with both ABP-Cel-Cy3 and ABP-XyG-Cy5 (major yellow band).This band was observed at a much lower intensity in samples from growth on glucose and was not observed in samples from growth on xyloglucan.Pulldowns from saturation cultures using ABP-βGlc-Bio, ABP-Cel-Bio, and ABP-XyG-Bio unambiguously identified Cel5D as the only exclusively ABP-XyG-Bio-reactive band, Cel5C as the exclusively ABP-Cel-Bio-reactive band, and Cel3A, Cel3B, and Cel3D as the ABP-βGlcreactive bands (Supplemental File 1).The secreted ABP-Cel-and ABP-XyG-reactive band in the arabinoxylan secretome (the intense band in Figure 3A, last lane) was identified as Cel5B.Cel5B and Cel5D both ran ~15 kDa heavier on SDS-PAGE than would be expected from their amino acid sequences.To investigate the origin of this discrepancy, ~50 μg of native Cel5B was partially purified from 200 mL of secretome collected from growth of C. japonicus on arabinoxylan to carbon-limited saturation via ultrafiltration and anion-exchange chromatography (Supplemental Figure 27). SD-PAGE of the purified protein followed by staining with the Pro-Q™ Emerald glycoprotein gel stain kit (Invitrogen) yielded a strong glycoprotein signal at the band position of Cel5B (Supplemental Figure 28). Exending this analysis to xyloglucan-grown cell lysate yielded a complex pattern of apparent glycoproteins, including a band at the position of Cel5D.Intact mass of the purified Cel5B measured via denaturing LC-ESI-MS gave a protein peak with a series of deconvoluted mass values from 71-75 kDa with spacing of 162 Da, indicating heavy glycosylation with variable hexose content (Supplemental Figure 29).Acid hydrolysis of the Cel5B sample followed by HPAEC-PAD analysis of the resulting monosaccharides revealed a complex mixture, including peaks that match D-mannose, D-glucose, D-galactose, and L-arabinose standards (Supplemental Figure https://doi.org/10.26434/chemrxiv-2023-lxghmORCID: https://orcid.org/0000-0002-7343-776XContent not peer-reviewed by ChemRxiv.License: CC BY-NC 4.0 30).L-Arabinose and D-xylose may be derived from the arabinoxylan substrate, but glucose, galactose, and mannose must have been synthesized by C. japonicus, underscoring the versatility of this species to grow -and derive the building blocks it needs -from such welldefined, single food stocks as used here.Considering the intact MS and monosaccharide composition, we propose that the underlying glycan structure is a galactoglucomannan Oglycan.Peptide LC-MS/MS analysis of native Cel5B digested with ProAlanase (Promega) yielded no detectable peptides from the serin-rich linker between the N-terminal catalytic domain and the C-terminal domain (Supplemental Figure 31).Having identified the core components of the C. japonicus xyloglucan-degrading system, we investigated the time-dynamics and substrate concentration-dependence of xyloglucanase expression.We diluted primed C. japonicus cells 10-fold into medium containing either xyloglucan or xyloglucan oligosaccharides.ABPP using ABP-XyG-Cy5 on cells harvested during the early induction with xyloglucan revealed two bands, the lower, sharper band running at the expected molecular weight of Cel5E or Cel5F and a higher, more diffuse Cel5D band (Figure 3B, C, Supplemental Figure 32).Notably, the lower band was primarily induced by xyloglucan oligosaccharides while Cel5D was primarily induced during growth on xyloglucan.To identify the putative xyloglucanase, primed cells were collected by centrifugation and resuspended in 100 mL of fresh medium containing 150 μg/mL of xyloglucan oligosaccharides.After 2 hours of incubation, cells and secretome were separated by centrifugation and tested for ABP-XyG-Cy5-reactive bands.The band of interest was found exclusively in the secretome while Cel5D was found in the cell fraction, so the secretome was collected and concentrated 50-fold by ultrafiltration prior to pulldown using ABP-XyG-Bio.This identified Cel5D, Cel5E, Cel5F and Cel5B (Supplemental File 1) as probe-reactive components.To our surprise, both xyloglucan and xyloglucan oligosaccharides induced xyloglucanase expression more efficiently at low (0.05-0.15 mg/mL) concentrations (Figure 3B, Supplemental Figure 33).This may suggest that C. japonicus can adapt to grow on persistently low levels of xyloglucan, such as those reported in soil samples near root tips.46 Sampling cultures grown in 0.1% xyloglucan or xyloglucan oligosaccharides over three hours showed that induction by xyloglucan oligosaccharides occurs rapidly, with a xyloglucanase band detectable after only 30 minutes (Figure 3C, Supplemental Figure 33).We also observed that growth on glucose resulted in low-level expression of Cel5B, while growth on xyloglucan resulted in expression of Cel5D correlating with a decrease in observed Cel5B activity, suggesting that Cel5B is acting as a "sensing" enzyme that is repressed by the detection of xyloglucan.Cel3A/B and Cel5C showed no change in expression under any condition tested, indicating that these are constitutively expressed.Interestingly, Cel5D was more strongly expressed in the presence of xyloglucan than xyloglucan oligosaccharides but expression of Cel5D in the presence of xyloglucan occurred with a lag.This may be explained by a period of time required to generate small, inducing fragments from large xyloglucan molecules.Cel5F/Cel5E and Cel3D expression appeared to be driven primarily by xyloglucan oligosaccharides, indicating that they are differentially regulated from Cel5D.Thus, in spite of being secreted, Cel5F/Cel5E do not appear to be "sensing" enzymes since their expression is dependent on induction by xyloglucan fragments.We speculate that it is instead acting as a "booster" enzyme, aiding with the solubilization of xyloglucan.Not having been previously functionally or structurally characterized in detail, we produced and purified Cel5B and Cel5C recombinantly in E. coli to assess further the correlation between probe reactivity and enzyme specificity.We found that Cel5B and Cel5C were both cellulases, efficiently degrading carboxymethylcellulose and mixed-linkage β-glucan (Supplemental Table 3).Cel5B showed only weak activity towards tamarind xyloglucan while Cel5C had weak activity towards carob galactomannan and no detectable xyloglucanase activity.Measurements of irreversible inhibition kinetics showed strong selectivity of Cel5C for ABP-Cel-N3 over ABP-XyG-N3 and only weak selectivity of Cel5B for ABP-Cel-N3 over ABP-XyG-N3, in line with in-gel fluorescence results (Table 1).To determine the molecular basis for the reactivity of ABP-XyG-Cy5 with CjCel5B, but not CjCel5C we crystallized both enzymes and solved their structures by molecular replacement in both unliganded and ABP-bound forms (Supplemental Figure 34).Cel inhibitor bound to CjCel5C displaying torsion angles (Φ,Ψ) of (-83°, 94°) between the non-reducing β-D-glucose and cyclophellitol moiety in ABP-Cel (Supplemental Figure 34F).O6' is recognized in the -2 position by both H87 and Y137 and O2' is recognized by the backbone carbonyl of S311.In contrast, CjCel5B recognizes ABP-XyG-N3 with (-83°, 133°) torsion angles (Supplemental Figure 34C).The consequent twist in the glucose backbone positions the α-(1,6)-xylose residue above W28 and W33, forming a hydrogen bond between O4 and D64.Notably, the active site cleft of CjCel5B is significantly more open beyond the -2 subsite, so we hypothesized that cellulase-specificity in Cel5B is dictated primarily by an inability to accommodate α-(1,6)-xylose residues in the positive subsites.To test this, we synthesized 4MU-XXXG and 6C4MU-XXXG as fluorogenic xyloglucanase substrates (see supporting information for synthetic details).Kinetics for the hydrolysis of 4MU-XXXG and commercially available 4MU-cellotetraose were measured to isolate contributions to specificity from the negative subsites.Cel5B showed a ~14-fold preference for 4MU-GGGG over 4MU-XXXG (Supplemental Table 2), roughly in line with its 7-fold preference for ABP-Cel-N3 over ABP-XyG-N3 (Table 1), but highly divergent from its 3,000-fold specificity towards carboxymethyl cellulose (CMC) over xyloglucan (Supplemental Table 3), supporting our hypothesis.The detection of the putative xylosidase Xyl39A using ABP-βGlc-FL (the band around 55 kDa in Figure 3A) was particularly interesting, since this enzyme, having 45% identity to the Xanthomonas citri XynB 47 , is adjacent to Cel5D in the genome.Staining C. japonicus lysates with the beta-xylose configured probe, ABP-βXyl-FL and ABP-βGlc-Cy5 confirmed the xyloglucan-dependent expression of Xyl39A and also revealed it has specificity towards ABP-βXyl-FL (Supplemental Figure 36).Chemical proteomics confirmed that Xyl39A was found in xyloglucan-grown cells and could be pulled down with ABP-βXyl-Bio and ABP-βGlc-Bio (Supplementary File 1) To investigate the specificity of CjXyl39A further, we produced the enzyme recombinantly in E. coli.Activity measurements against a variety of 4methylumbelliferyl (4MU) glycosides showed specific recognition of β-D-xylose over other glycosides (Supplemental Table 4).Functionalization of α-(1,6)-xylose branches with β-(1,2)xylose has been reported in xyloglucan extracted from the leaves 39 (but not fruit 48 ) of argan trees.We speculate that the co-expression of Xyl39A and Cel5D during growth on xyloglucan suggests an evolved adaptation of C. japonicus towards degradation of β-xylosylated xyloglucan, however we were not able to obtain a suitable sample of β-xylosylated xyloglucan for testing.Having dissected the endo-β-glucanase, exo-β-glucosidase, and exo-β-xylosidase components of the native C. japonicus xyloglucan-degrading system, we turned to the essential exo-αxylosidase, exo-α-L-arabinofuranosidase and exo-β-galactosidase activities as potential handles for characterizing xyloglucan-degrading systems using ABPP.Staining with ABP-αAraf-Cy5 showed the presence of Abf51A in C. japonicus under all growth conditions.Abf51A staining was less intense in samples grown on glucose and more intense in samples grown on arabinoxylan (Supplemental Figure 35).We conclude from this that Abf51A displays similar regulatory logic to the E. coli araBAD operon 49 and is not co-regulated with xyloglucandegrading machinery.Attempts of using ABP-αXyl-Cy5 to detect CjXyl31A in the xyloglucan-grown C. japonicus lysate during induction by xyloglucan oligosaccharides revealed the emergence of a band at the expected ~115 kDa, but identification of the band was hindered by weak reactivity and significant non-specific labeling (Supplemental Figure 37).We attribute the poor potency and selectivity of this probe to a lack of binding in the positive subsites known to be important for substrate recognition in this enzyme class.43 Indeed, the ABP-αXyl-Cy5 probe performs well on purified recombinant CjXyl31A, but is less effective in doped lysates, consistent with its poor performance on C. japonicus lysates (Supplemental Figure 38).Finally, comparing the reactivity of ABP-βGal-Cy5 and ABP-βGlc-Cy5 in C. japonicus cell lysates showed strong, and clearly orthogonal, labeling of putative β-glucosidases and β-galactosidases (Supplemental Figure 39).A pulldown from the xyloglucan-grown lysate using ABP-βGal-Bio revealed the presence of Bgl35A, the known xyloglucan oligosaccharide-specific β-galactosidase, and Bgl2A, an uncharacterized putative β-galactosidase which was previously reported not to be up-regulated in response to growth on xyloglucan (Supplemental File 1).In summary, we have developed a platform with which native bacterial xyloglucan-degrading systems can be sensitively detected and functionally interrogated.Dissection of native proteomes derived from C. japonicus grown on various polysaccharide food sources using these tools reveals features not previously observed, including the production of β-xylosidase during growth on xyloglucan, low-level secretion of Cel5B which we conclude to be a cellulase during growth on glucose, the occurrence of significant enzyme glycosylation, and the different expression and secretion behaviors of the vanguard xyloglucanases, Cel5D, Cel5E, and Cel5F.The ability to detect xyloglucanases with such high sensitivity and throughput enabled measurement of the concentration-dependence and time-dependence of xyloglucanase expression in response to different inducers, demonstrating surprisingly sensitive xyloglucan detection by C. japonicus.Building on these developing capabilities, we envision the assembly of different polysaccharide-specific toolkits to enable the characterization of native component enzymes from diverse microbial polysaccharidedegrading systems.Importantly, known or putative glycosidase products, which are the result of enzyme recognition and processing of specific polysaccharide substructural stretches, can be imbued in mechanism-based probe designs, as shown here for linear and branched hemicellulose structures (cellobiose versus xyloglucan).This allows establishment of enzyme specificities also in situations where such preferences cannot be gleaned from genomic data alone, in a low-throughput, gel-based assay.The exquisite specificity of our cyclophellitolbased probe designs compares well to alternative probe designs 50,51 , leading to relatively simple gel images, with fluorescent bands pointing to probe-reactive proteins that in all likelihood feature substrate specifics correlating with that of the probe structure.It should be noted that the suite of probes presented here -indeed probes based on the cyclophellitol scaffold -are reactive towards retaining glycosidases only, excluding inverting glycosidases for identification using our platform.This caveat aside, designing probes targeting exo-and endo-glycosidases produced to digest different biomass polysaccharides are expected to shed light also in other microbial digestive systems.As well, and as was demonstrated recently, bespoke probes can also be used in machine learning assisted, de novo glycosidase design.52

Methods
Detailed methods are available in the online supplementary information.

Data availability
The data that support the findings of this study are openly available in the protein databank at https://www.rcsb.org/,reference numbers 8BQA, 8BQB, 8BQC, 8BN7, and 8OZ1.References 1. Helbert

Figure 1 .
Figure 1.A) Strategy of the multiplexing activity-based protein profiling (ABPP) platform subject of the here-presented study.Activity-based probes (ABPs) bearing orthogonal fluorophores are designed to emulate xyloglucan/cellulose structural elements.Treatment of secretomes of microbes from various sources grown on xyloglucan followed by SDS-PAGE resolution will yield color-coded fingerprints of both exo-acting and endo-acting GHs.B) Mode of action of a retaining -exo-glucosidase and its mechanism-based, covalent and irreversible inhibition, thereby labeling, by cyclophellitol aziridine ABPs.

Figure 2 .
Figure 2. Labeling of xyloglucanases with inhibitors and probes.A) Intact MS of CjGH5D xyloglucanase treated with 100 µM ABP-Cel-N3, 100 µM ABP-XyG-N3, or vehicle control for 1 hour.B) Crystal structure of CjCel5D labelled with ABP-XyG-N3 (purple).2Fo-Fc density is shown for the ligand and catalytic residues as a grey mesh contoured at 2σ.The complex between CjCel5D and 2-fluoro-XXXG (PDB 6HAA) is superimposed in teal.C) Residual activity kinetics of CjCel5D inhibited by different concentrations of ABP-XyG-N3.The model fits are shown as dashed lines.6C4MU-XXXG was used as substrate.D) kapp vs. inhibitor concentration for CjCel5D interacting with ABP-XyG-N3.The model fit is shown as a dashed line.

Figure 3 .
Figure 3. ABPP analysis of C. japonicus cultures.A) Laser-scanning fluorescence image of a 4-20% SDS-PAGE separation of C. japonicus proteins following treatment of intact cells (C), lysate (L), or supernatant (S) with a mixture of ABP-Glc-FL, ABP-Cel-Cy3, and ABP-XyG-Cy5.The carbon source on which the cells were grown is noted above each set of three lanes.B) Representative gel image from ABPP analysis of C. japonicus lysate following 2 hours of growth in the presence of increasing concentrations of xyloglucan or xyloglucan oligosaccharides.C) Representative gel image from ABPP analysis of C. japonicus lysate collected over time from growth in the presence of 0.1% xyloglucan oligosaccharides, xyloglucan, or glucose (Glc).