JBC Papers in Press. Published on January 23, 2012 as Manuscript M111.314286 The latest version is at http://www.jbc.org/cgi/doi/10.1074/jbc.M111.314286 1   Functional  association  of  the  catalytic  and  ancillary  modules  dictates   enzymatic  activity  in  a  glycoside  hydrolase  family-­‐43  β-­‐xylosidase*       Sarah Moraïsa,b, Orly Salama-Albera, Yoav Baraka,c, Yitzhak Hadarb, David B. Wilsond, Raphael Lamede, Yuval Shohamf and Edward A. Bayera a Department of Biological Chemistry, The Weizmann Institute of Science, Rehovot, Israel; b Faculty of Agricultural, Food and Environmental Quality Sciences, The Hebrew University of Jerusalem, P.O. Box 12, Rehovot, 76100, Israel; c Chemical Research Support, The Weizmann Institute of Science, Rehovot 76100 Israel; d Department of Molecular Biology and Genetics, Cornell University, Ithaca, NY14853, USA D o w e Department of Molecular Microbiology and Biotechnology, Tel Aviv University, Ramat Aviv 69978 n lo a d Israel; e d fro f Department of Biotechnology and Food Engineering, Technion-Israel Institute of Technology, Haifa m h ttp ://w w   w *Running title: Modular association in a family-GH43 β-xylosidase .jb c .o rg b/ y To whom correspondence should be addressed: Edward A. Bayer, Department of Biological gu e s Chemistry, The Weizmann Institute of Science, Rehovot, Israel. Tel: (+972)-8-934-2373. Fax: (+972)- t o n A 8-946-8256. Email: [email protected] p ril 5, 2 0 Keywords: GH43, hemicellulase, Thermobifida fusca. 19 Copyright 2012 by The American Society for Biochemistry and Molecular Biology, Inc. 2   Background: Thermobifida fusca β-xylosidase INTRODUCTION Xyl43A contains a catalytic and an ancillary module. Xylan is the major constituent of Results: Enzymatic activity is lost when the hemicellulose, which represents 20 to 30% of modules are expressed independently, but restored lignocellulosic plant biomass (1) and represents an upon specific non-covalent reassociation. important industrial target for degradation [e.g. Conclusion: The catalytic and ancillary modules bleaching in the pulp and paper industry (2)]. of Xyl43A behave as a single functional entity. Xylan is a linear polysaccharide, consisting of β- Significance: Two phylogenetically distinct 1,4 linked D-xylose units with a large variety of modular components of an enzyme evolved side-chain substituents. Consequently, the together to form an enzymatically active unit. contribution of multiple hemicellulases is required for complete hydrolysis of xylan by synergistically complementary enzymes. The well-studied aerobic thermophilic soil SUMMARY bacterium, Thermobifida fusca, can use cellulose and xylan as sole carbon sources (3,4). For this β-Xylosidases are hemicellulases that purpose, the bacterium produces a set of six D o hydrolyze short xylo-oligosaccharides into cellulases and several hemicellulases such as w n xylose units, thus complementing endoxylanase xylanases, a β-xylosidase, an α-L- loa d degradation of the hemicellulose component of arabinofuranosidase and an acetyl xylan esterase ed lignocellulosic substrates. Here, we describe the (5,6). Two T. fusca endoxylanases (Xyn11A and fro m cloning, characterization and kinetic analysis of Xyn10B) have been expressed and fully h ttp a glycoside hydrolase family-43 β-xylosidase characterized (7,8). A third endoxylanase, ://w (Xyl43A) from the aerobic cellulolytic Xyn10A, was characterized from a related species w w bacterium, Thermobifida fusca. Temperature Thermomonospora alba (9), and a nearly identical .jb c and pH optima of 55-60°C and 5.5 to 6, gene coding for this enzyme was found in T. fusca. .org respectively, were determined. The apparent Due to their potential industrial applications, the b/ y Km value was 0.55 mM, using p-nitrophenyl cloning and characterization of other thermostable gu e xylopyranoside as substrate, and the catalytic xylanases deserve particular attention. We have st o n constant (k ) was 6.72 s-1. T. fusca Xyl43A therefore investigated in this communication the A cat p contains a catalytic module at the N-terminus putative hemicellulase encoded by the T. fusca ril 5 and an ancillary module (termed herein gene Tfu 1616 which was related in sequence to , 2 0 Module-A) of undefined function at the C- the family-GH43 glycoside hydrolases (GH43), 19 terminus. We expressed the two recombinant the members of which are known to possess β- modules independently in Escherichia coli and xylosidase (EC 3.2.1.37), β-1,3-xylosidase (EC examined their remaining catalytic activity and 3.2.1.72), α-L-arabinofuranosidase (EC 3.2.1.55), binding properties. The separation of the two arabinanase (EC 3.2.1.99), xylanase (EC 3.2.1.8), Xyl43A modules caused the complete loss of or galactan 1,3-β-galactosidase (EC 3.2.1.145) enzymatic activity, whereas potent binding to activities (http://www.cazy.org/) (10). To date, xylan was fully maintained in the catalytic over 1500 bacterial genes have been associated module and partially in the ancillary Module-A. with this family. Over fifty bacterial enzymes from Non-denaturing gel electrophoresis revealed a this family have been examined for their specific non-covalent coupling of the two enzymatic activities, but only a part of them have modules, thereby restoring enzymatic activity been fully characterized, the results of which were to 66.7% (relative to the wild-type enzyme). published in the primary literature. Among them, Module-A contributes a phenylalanine residue β-xylosidases (11-16) endoxylanases (17,18), that functions as an essential part of the active arabinanases (19-22), arabinofuranosidases (23- site, and the two juxtaposed modules function 25) and galactanases (26) were described. as a single functional entity. 3   The present work reports the cloning of the Tfu (ordered from Syntezza Bioscience Ltd., 1616 gene of T. fusca and the characterization of Jerusalem), and PCR reactions were performed the protein. The gene was expressed in E. coli, and using Pfu UltraII DNA polymerase (Aligent the protein was purified to near homogeneity. Data Technologies, Santa Clara, CA). on physico-chemical parameters and enzyme Protein expression and purification— activity of the intact Xyl43A are provided. The Plasmids containing genes coding fo rthe full- two modules of the protein (a GH43 catalytic length Xyl43A enzyme, the GH43 modul,e module and an associated ancilalry module — ancillary Module-A, and mutants Xyl43A(F518A) Module-A) were produced as separate protein and Module-A(F518A) were expressed in E. coli entities; the catalytic and binding activities were BL21 (lDE3) pLysS cells, and the expressed His- studied for both the separated modules as well as tagged proteins were purified on a Ni-NTA their in-vitro combined form. column (Qiagen), as reported earlier (27). A gel filtration purification step was conducted for Experimental procedures association of the GH43 module and the Module- A, using an AKTA-prime System and a Hiload Cloning—Xyl43A (Tfu 1616, GenBank 16/60 Superdex 75 column (GE Healthcare). SDS- accession number: AAZ555651) was cloned from PAGE was employed to test the purity of the Thermobifida fusca YX genomic DNA. Primers recombinant proteins (12% acrylamide gels). The D o were designed with the program Oligo Primer w fractions with the highest degree of purity were n Analysis Software version 5.1 and ordered at the lo pooled, and the concentrations of the recombinant ad Weizmann Institute of Sciences facility (N- ed terminal primer: 5’- proteins were estimated by absorbance at 280 nm fro based on their amino acid composition using the m TCATGACATATGCACCATCAC h Protparam program ttp CATCACCATACTTCTCCCCAAGTCACGTCC (http://web.expasy.org/protparam/). Proteins were ://w T-3’; C-terminal primer 5’- w stored in 50% (v/v) glycerol at -20°C. w TGATTGCTCGAGTTAGGAGGGGGACTGAG Substrates—Microcrystalline cellulose .jbc GCCGGTA-3’ (NdeI and XhoI sites in bold). The .o (Avicel) was purchased from FMC Biopolymer rg PCR product was inserted into and ligated into b/ (Philadelphia, PA, USA) and was used for the y linearized pET21a to form pXyl43A. g preparation of phosphoric acid swollen cellulose ue GH43 was cloned using Xyl43A WT forward (PASC 7.5 mg ml-1 pH 7). Insoluble xylan was st o primer and 5’- prepared by boiling oat-spelt xylan (Sigma Chem. n A p TGAGCGG-3C’T CasT aC GreAveGrsCeT pAriCmGeGr C(XChAoIC GsiGteG iTn GbColGd) Canod., rSetc. oLvoeuriinsg, MthOe ) pfeolrl e3t0 bmyi nc einnt rdiifsutgilaletido nw; attheirs ril 5, 2 0 and Module-A using 5’- 1 was followed by 3 cycles of washes with distilled 9 TTAAGCCATATGCAGCCGTCAGAGACCGA water to remove soluble sugars, and determination CCACTTCGACGA-3’ and 5’- of dry weight (28). Birchwood xylan, beechwood TTATGTCTCGAGGGAGGGGGACTGAGGCC xylan, chitin, p-nitrophenyl-β-D-glucopyranoside, GGT-3’ (NdeI and XhoI sites in bold). p-nitrophenyl-α-L-arabinofuranoside, p- PCR reactions were performed using ABgene nitrophenyl-β-D-cellobioside, p-nitrophenyl-β-D- Reddymix x2 (Advanced Biotechnologies Ltd., xylopyranoside (pNPX) and xylobiose were Epsom, UK). DNA samples were purified using a HiYieldTM Gel/PCR Fragments Extraction Kit purchased from Sigma Chemical Co. (St. Louis MO). Debranched linear arabinan was purchased (Real Biotech Corporation, RBC, Banqiao City, from Megazyme International, Ltd. (Wicklow, Taiwan). Ireland). Hatched wheat straw (0.2–0.8 mm) was PCR mutagenesis was performed for the provided by Valagro (Poitiers, France) and treated preparation of Xyl43A(F518A) and Module- as described previously (29,30). A(F518A) plasmids using phosphorylated primers pH studies—The β-xylosidase activity over a 5’- pH range of 3.5 to 10 was investigated using GGTGCCACGGGAGCGTTCCTCGGCCTGTG citrate buffer from 3.5 to 8, Tris buffer from 7 to 9, GG-3’ and 5’-CCACATGATGGGGTCGTTGC-3’ and glycine/NaOH buffer from 9 to 10. Xyl43A 4   was assayed at 50°C for 7 min in a 700 µl reaction buffer pH 6.0 containing 12 mM CaCl and 2 mM 2 mixture containing 0.05 µM enzyme, 100 mM EDTA were mixed with 0.5 mg of insoluble xylan buffer, 1 mg/ml bovine serum albumin (BSA) and (or 10 mg of microcrystalline cellulose). Tubes 5 mM pNPX. After 7 min, the tubes were placed were mixed gently at 4 °C for 1 h and centrifuged on ice, the reaction was terminated by adding 200 at 14000 rpm for 1 min. The supernatant fluids µl of 1 M NaOH, and the optical density was (containing the unbound fraction of the proteins) measured at 420 nm. were transferred to new tubes and supplemented with SDS-containing buffer. The pellets Temperature dependence and stability—The (containing the bound fraction of the proteins) influence of temperature upon activity was assayed were washed once by resuspension in 100 µl of the as follows: the release of p-nitrophenol by 0.05 above-described buffer, centrifuged, and µM Xyl43A at temperatures ranging from 30 to resuspended in 60 µl of SDS-containing buffer. 80°C was measured at 420 nm after 7 min reaction Boiled samples were subjected to SDS-PAGE and 15 min cooling on ice-water in 700 µl final (12% gels) for analysis. BSA served as a negative volume (100 mM buffer citrate pH 6, 1 mg/ml control for binding specificity. BSA, 5 mM pNPX). The effect of temperature on the stability of Xyl43A was determined by Thin Layer Chromatography (TLC)—The incubating the enzyme without substrate over a 15- hydrolysis of xylobiose by Xyl43A was monitored min period for 50 h at 50, 60 and 70°C at pH 6. by TLC analysis. The xylobiose hydrolysis Do w Similarly, the purified GH43 catalytic module, and reaction included 5 mg xylobiose, 0.2 mM Xyl43A nlo Module-A were incubated at 50°C over the same and 100 mM citrate buffer, pH 6. The reaction was ad e d time period, either before or after combining the performed at 50°C, at different time points (5 min, fro GH43:Module-A at 1:0.8 molar ratio. After adding 15 min, 30 min and 2 h), samples were taken, and m h substrates, 7-min reactions (as specified above) the reaction was terminated by adding Hg2+ to a ttp were carried out at 50°C. The optical density at final concentration of 1 mM (which completely ://w w 420 nm was then measured after 15-min inhibits the enzyme). Chromatography was carried w incubation on ice. out on silica gel 60 plates (Merck, Germany), .jbc .o using butanol/ethanol/water (3:2:2) as a rg Kinetic studies—Temperature-equilibrated b/ reactions of 180 µl final volume (100 mM buffer developing solvent. Spots were visualized by y g u charring with 0.2% orcinol (in methanol and 20% e c5i0t°raCte) pwHe r6e, 1in mitiga/tmedl BbSyA ,a d0d tion g1 51 m0 Mµ lp NoPf Xt haet phosphoric acid in methanol, 1:1 vol/vol) and st on A eanbzsoyrmbaen c(fei nwala sc omnocnenittorraetido na t o4f 200 .0n5m µ aMt )1. 4-Tshece heatiMngu flotirp l2e mainm aint o1-5a0c°iCd . sequence alignment— pril 5, 2 0 intervals for 10 min, using an ELISA plate reader The Clustal W2 program 1 9 (BioTek Synergy HT, Winooski, VT USA). (http://www.ebi.ac.uk/Tools/msa/clustalw2/) was Absorbancies (420 nm) were then converted to employed to analyze the amino-acid sequence- molarities using an extinction coefficient of 3365 based alignment between the characterized β- M-1·cm-1 for the absorbance of p-nitrophenol in xylosidases in GH family 43 (11-16) and T. fusca citrate buffer pH 6 at 50 °C. Xyl43A. This alignment served to identify the homologous conserved residues between Effect of salts and chemicals—Various salts and chemicals were added at a final concentration Geobacillus stearothermophilus and T. fusca β- of 1 mM to the standard enzymatic reaction xylosidases that appear to be involved in the mixtures in order to check potential inhibitory interface between the catalytic and the C-terminal effects on enzymatic activity. Module-A of Xyl43A. Binding to insoluble polysaccharides—The binding of the proteins to insoluble xylan and RESULTS microcrystalline cellulose was determined qualitatively using SDS-PAGE. In a final volume Sequence analysis and production of Xyl43A of 100 µl, 10 µg (5 µg for cellulose binding (Tfu1616). The Xyl43A gene is part of the hemicellulolytic system of Thermobifida fusca. It experiments) of pure protein in 50 mM citrate 5   encodes a 550 amino-acid protein with a Pro 334 and Gln 335, where there seemed to be calculated molecular weight of 61686 Da. The little or no interactions between the linker and the amino-acid sequence of the Xyl43A protein was modules in the predicted model. The two modules subjected to BLAST analysis (blastp, nr database, were therefore cloned separately (each with an N- http://blast.ncbi.nlm.nih.gov/) and showed terminal His tag), expressed and purified by metal significant homology (from 44% to 59% sequence ion affinity chromatography. The resultant GH43 identity) to β -xylosidases and α-L- module and ancillary Module-A exhibited their arabinofuranosidases that belong to GH family 43. expected molecular masses (35.8 kDa and 25.9 The ORF has a G+C content of 68.4%, not kDa, respectively) in the SDS-PAGE gel and significantly different from that (67%) of the purity above 95% was estimated (data not shown). whole genome of T. fusca, which is typical of A schematic representation of the recombinant many thermophilic organisms (31). The Xyl43A proteins expressed in this study is shown in Fig. 2. gene does not include any recognizable signal Substrate specificity. The activity of the peptide as determined by the SignalP server recombinant enzyme was tested on a variety of (http://www.cbs.dtu.dk/services/SignalP/). Thus, substrates. No enzymatic activity was detected on the protein is assumed to remain intracellularly in xylans (from birch wood, beech wood and oat- T. fusca. SDS-PAGE analysis of the purified spelt, both soluble and insoluble), cellulosic Xyl43A protein revealed a single protein band of substrates (Avicel and PASC), chitin, wheat straw, D o about 60 kDa, which is in good agreement with the arabinan p-nitrophenyl-β-D-glucopyranoside, p- w n calculated molecular mass. A typical Xyl43A nitrophenyl-β-D-cellobioside and p-nitrophenyl-α- loa d preparation yielded 31.2 mg/L of purified protein. L-arabinofuranoside (pNPA). Detectable levels of ed Bioinformatic analysis of the protein revealed hydrolysis were only observed for p-nitrophenyl- fro m the presence of a GH43 catalytic module located β-D-xylopyranoside (pNPX) and xylobiose, http between residues 16 and 319 and a second module confirming that the enzyme belongs to the β- ://w of undefined function (ancillary module, Module- xylosidase class. Residual activity (kcat/Km of ww A) located between residues  338  and  550  in the 0.75 -s1·mM-1 versus 12.25 for pNPX) was .jb c C-terminal part of the protein. In order to evaluate observed with pNPA, due to the structural .org the binding properties and the enzymatic activity similarity between β-D-xylopyranoside and by/ of both N-terminal (catalytic core) and C-terminal α-L-arabinofuranoside. These results gue (ancillary) modules, we produced two distinct support the premise that the conserved arginine st o n chimaeras, GH43 and Module-A. Since there is (position 288 in T. fusca Xyl43A) allows the A p 45% identity and 60% similarity between the distinction between β-xylosidase and ril 5 previously characterized GH43 from G. arabinofuranosidase classes (32). This arginine is , 20 stearothermophilus (15,32), its structure was used 1 believed to be conserved only in β-xylosidases. 9 and compared to the predicted structure of T. fusca Neither the GH43 nor Module-A alone had Xyl43A, obtained using Swiss Model measurable activities on any substrates (including (http://swissmodel.expasy.org/) (Fig. 1). The xylans and arabinoxylan). quality of the model was estimated using the pH studies. The optimal pH range of Xyl43A ProSA-web site was 5.5 to 6 for β-xylosidase activity (Fig. 3A) (https://prosa.services.came.sbg.ac.at/prosa.php). with more than 70% activity retained at pH 4.5 and The Z-score of the model was -6.57, which is well pH 7. Xyl43A is more stable at alkaline pH than at within the range of scores typically found for acidic pH and is drastically inactivated at pH 4. It native proteins of similar size. is indeed known that changes in pH affects ionic The model was especially helpful for or electric charge of active site amino acids (that determining the linker region of T. fusca Xyl43A. participate in substrate binding and catalysis) and In G. stearothermophilus, the linker extends from either enhances and stabilizes interactions with the Val 319 to Asp 329; by analogy, we deduced that substrate or breaks intra- and intermolecular the linker region in T. fusca Xyl43A would extend bonds, changing the shape of the enzyme and, similarly from Leu 328 to Glu 338. The most therefore, affecting its efficacy (33). Activity appropriate juncture between the two modules was profiles of glycosidases versus pH are typically determined to be the bond between the adjacent 6   bell-shaped, reflecting the ionization state of, at length Xyl43A exhibited strong binding capacity least, two essential amino-acids residues. The for insoluble xylan, whereas weak binding to expected pK values of the general base and microcrystalline cellulose was observed (Fig. 4). a general-acid ionizable groups in the enzyme As expected, the bulk of the BSA negative control (derived from the pH-dependency graph) are 4.5 was found in the unbound fraction, and the and 8, respectively, close to the values obtained for respective insoluble xylan and microcrystalline G. stearothermophilus XynB3 (15). cellulose positive controls were located in the bound fractions. The GH43 catalytic module alone Thermostability. The hydrolysis rate demonstrated full binding capacity for insoluble increased from 30°C to 60°C, and the temperature xylan whereas Module-A was found in bound and coefficient (Q10) was calculated to be 1.3. An unbound fractions in almost equal proportions. Arrhenius plot exhibited a straight line from 30°C Thus it seems that xylan-binding capacity of to 60°C, typical of a single rate-limited thermally Xyl43A is due mainly to the residues present in activated process, indicating an increase in enzyme the catalytic module and the ancillary Module-A activity with increasing temperature. The plays a minor role in the binding interaction (Fig. activation energy and the pre-exponential factor 4). were determined to be 18.9 KJ.  mol−1 and 390, Re-association of GH43 and Module-A. Non- respectively. Under the reaction conditions the covalent association of the separately expressed temperature optimum was found to be 55-60°C for D o and purified GH43 and Module-A components at w β-xylosidase activity (Fig. 3B); however the n room temperature was analyzed by non-denaturing lo enzyme underwent rapid inactivation at 60°C (Fig. ad PAGE. The catalytic module was mixed with ed 3C). Purified Xyl43A demonstrated moderate increasing amounts of the helper module in molar fro thermostability, even after 50 h at 50°C, 70% of m ratios ranging from 1:0.2 to 1:1.8, and the mobility h the β-xylosidase activity remained. Incubation at ttp 60°C resulted in about 50% loss of activities after pattern of each mixture was compared with that of ://w each module alone and with that of the full-length w only 2 h and almost no activity after 24 h. w Xyl43A enzyme. Mixtures of GH43 and Module- .jb Treatment at 70°C resulted in complete c A resulted in complex formation (Fig. 5A). The .o inactivation of the enzyme in less than 15 min rg mobility pattern of the complex was nearly b/ (Fig. 3C). identical to that of wild-type Xyl43A, revealing a y g u e Kinetic constants. The kinetic constants of specific interaction between the catalytic and st o Xyl43A on pNPX were determined by running the ancillary modules. The intensity of the band, n A p r3eDac).t ioAnt a5t 0d°iCff eraenndt sautb sptHra te6 ,c otnhcee nrterlaetaiosen s o(fF ipg-. cMoorrdeuslpeo-And, ingi ntcor eathseed intuernatcilt iona of 1:G0.H84 3 raatnido ril 5, 2 0 nitrophenol was linear with time and proportional (GH43:Module-A), whereas the intensity of the 1 9 to enzyme concentration. The apparent Km value bands of each of the individual modules decreased. was 0.55 mM pNPX while the k was 6.72 ·s-1. At this ratio, the Module-A band interacted cat Effect of various chemicals. The effects of completely with GH43. A relatively small part of different chemicals on Xyl43A activity are shown the latter fails to form the complex, probably due in Table 1. The enzyme retained almost full to incorrect folding of a portion of the GH43 activity in the presence of 1 mM Li+, Na2+, K+, fraction. Up to 25% of the wild-type β-xylosidase Mg2+, Ca2+, Fe3+, Mn2+, Co2+, Cu2+, and the reagent activity was thus recovered upon association of the SDS. The addition of EDTA had no effect on the two modules (Fig. 5B). activity, suggesting that no metals are needed for Gel filtration was used to purify the the enzyme reactions. Group IIb metals such as reassociated enzyme from the two modules at a Hg2+ and Ag2+ gave stronger inhibition [mercury 1:0.8 ratio, and the kinetics parameters were inhibition is indeed well known for most enzymes determined for the purest fraction obtained. The including cellulases and xylanases, indicating the apparent Km value was 0.6 mM pNPX (similar to existence of essential thiol groups (7)]. that of the wild-type enzyme) while the k was cat Binding to insoluble polysaccharides. The 4.83 s-1. Up to 71.8% of the wild-type β-xylosidase binding capacity of Xyl43A to insoluble xylan and activity was thus recovered upon association of the microcrystalline cellulose was investigated. Full- two modules. 7   The moderate thermostability characteristics Phe506 therein. This phenylalanine is the only observed for the wild-type enzyme (Fig. 3C) raised residue in Module-A, which participates in the the question as to whether the re-associated active site and presumably plays a role in substrate Xyl43A would also be stable to heat treatment binding and specificity. In order to further under the same conditions. Moreover, it was of investigate the role of this amino acid in the interest to determine whether exposure of the two catalytic activity of T. fusca Xyl43A, the constitutive modules to heat would result in conserved Phe518 was replaced with Ala. reduced stability. We therefore subjected the two The kinetic constants of Xyl43A(F518A) for isolated modules to heat treatment at 50°C both pNPX hydrolysis were determined by interaction prior to and after re-association. Heat treatment of of the mutant enzyme at different substrate the two modules separately before re-association concentrations as performed for the wild-type led to a marked loss of stability (Fig. 5C). About enzyme (data not shown). At 50°C and pH 6, the 60% of the residual activity was obtained after 2 h release of p-nitrophenol was linear with time and of heat treatment and only 15% remained after 5 h. proportional to enzyme concentration. The In contrast, re-association of the two modules apparent Km value was 0.98 mM pNPX while the before heat treatment resulted in increased k was 0.059 s-1. The replacement of the cat thermostability: about 80% relative activity phenylalanine by alanine affected the activity of remained after 5 h of heat treatment and 50% after the enzyme dramatically, the activity of the mutant D o 48 h (Fig. 5C). The results indicated that the representing less than 1% of the wild-type activity w n individual modules exhibit decreased stability in on pNPX (0.89%). lo a d solution but upon re-association form a stable Re-association of GH43 and mutated Module- ed complex. A(F518A). In order to further examine the role of fro m the Phe518 residue in the catalysis, Module-A was h In order to gain insight into the molecular ttp basis of re-association of the GH43 and Module-A, replaced with alanine in a manner similar to that of ://w the wild-type enzyme. As described for the wild- w we employed the Protein Interaction Calculator w (PIC) server (34) to analyze the known 3D type modules, non-covalent association of GH43 .jbc and the mutated Module-A(F518A) was .o structure of the family-GH43 XynB3 β-xylosidase rg performed. A strong interaction between the two b/ from G. stearothermophilus. Using this server, the y modules was observed (data not shown), g various interactions involved in the interface ue between the catalytic and the C-terminal Module- suggesting that Phe518 does not play a dominant st o role in the strong association between the two n A A were determined, and the homologous p icdoennsteirfvieedd. r eTsiadbulees i2n Tsh. ofuwssc at hXey l4m3aAjo rw errees itdhueens , mmoodduulleess., βN-xeyvleorsthidealesses ,a cutipvoitny wasasso cnioatt iorenc oovfe rtehde, ril 5, 20 suggesting that the role of this particular amino 1 predicted to be involved in the intramodular 9 acid is directly related to catalysis per se. interface. The data suggest strong reciprocal binding interaction between the two modules of T. fusca Xyl43A and provide an explanation for their DISCUSSION observed re-association. A number of GH43 family members (as well Production of mutated Xyl43A(F518A). as GH3, GH51 and GH54 families) have been According to the crystal structure of the family reported to exhibit simultaneously both β- GH43 β-xylosidase (XynB3) from G. xylosidase and α-L-arabinofuranosidase activities, stearothermophilus T-6 (32), the active site of the probably due to the active site geometry, that does enzyme possesses a pocket topology, mainly not allow the enzyme to distinguish between the constructed from β-propeller domain residues. In two saccharides. Indeed, both activities seem to addition, a conserved phenylalanine residue occur in the same active site (35,36). The (Phe518 of T. fusca Xyl43A, homologous to Thermobifida fusca Xyl43A, characterized here, Phe506 of G. stearothermophilus XynB3) is exhibited primarily β-xylosidase activity and only located on a loop of Module-A in the latter nominal levels of α-L-arabinofuranosidase structure. The loop serves to close one side of a activity. cleft, formed by the catalytic module, and inserts 8   In former research (37), Asp and Glu residues the active site in G. stearothermophilus T-6 of family GH43 enzymes have been directly arabinanase (AbnB) resembles a groove with a implicated in substrate hydrolysis, in line with the narrow bridge formed by Phe104 (conserved in classical general acid (proton donor) and family GH43 arabinanases) and Cys222. A nucleophile/base that characterize enzymatic neighboring cysteine (Cys221) bridges with the hydrolysis of the glycosidic bond (38). Multiple- Cys222 and covers the cleft formed in the active sequence alignment of glycoside hydrolase (GH) site. The Phe and both Cys residues are absent families 32, 43, 62, and 68 revealed three from β-xylosidases in general, and T. fusca β- conserved blocks, each containing an acidic xylosidase in particular. Moreover according to residue at an equivalent position in all of the Brux, et al. (32), in the β-xylosidases, an extended enzymes; site-directed mutagenesis studies have arabinopyranose ligand would clash with the shown that Asp/Glu (block I), Asp (block II), and hydrophobic area generated by conserved Glu (block III) are essential for catalytic activity of xylosidase residues, i.e., Trp74, Phe127, Phe32, GH43. This indicates direct involvement of the and Phe506 of G. stearothermophilus XynB3. In conserved residues in substrate binding and the case of galactopyranose, there is also a steric hydrolysis, and substitutions of these specific clash caused by the additional hydroxymethyl residues inactivate the enzyme or considerably group, which encounters Phe506 (1.4 Å). These reduce its activity. These three strictly conserved criteria explain the inability of the T. fusca D o key acidic residues are located in a funnel-shaped catalytic GH43 module to degrade arabinan, and w n active site that comprises two subsites with a underscore the sophisticated divergence of the loa d single route for access by ligands in a β-propeller conserved GH43 arabinanases and β-xylosidases. ed architecture, and the three residues operate with Physical separation and reassociation of the from the canonical reaction mechanism of inversion of two modular components of Xyl43A (GH43 and http anomeric configuration (32,39). More recently, the Module-A) enabled us to further analyze the ://w crystal structures of Cellvibrio japonicus enzymatic and binding properties of this enzyme. w w arabinanase 43A and Geobacillus Separating the two modules led to a complete loss .jb c stearothermophilus β-xylosidase further in enzymatic activity of the N-terminal module .org demonstrated the role of these three carboxylates GH43 but did not affect its ability to bind xylan, b/ y in catalytic activity (19,31). Similar blocks and suggesting that the active site would thus need the gu e essential residues were found in T. fusca Xyl43A combined conformation of the complete protein in st o by sequence comparison: block I Y29PDPS33, order to degrade its substrate correctly. Indeed, a n A p block II G142FDPS146 and block III V201TEAPH206. strong non-covalent interaction occurs between the ril 5 In the GH43 family, several modular two modules of T. fusca Xyl43A. This physical , 2 0 architectures are observed (40). One group of interaction led to substantial recovery of β- 19 GH43 enzymes is composed only of a single xylosidase activity (71.8%), demonstrating the role GH43 module and lacks the C-terminal module of the Module-A in enzymatic degradation of its (Module-A). These enzymes belong to the substrate. Based on the solved structure of G. arabinanase class (20-22,41). Another group of stearothermophilus (32,43), several residues play family GH43 enzymes comprising both modules is an essential role in formation of the strong defined as either β-xylosidases or α-L- interface between the two modules of the enzyme arabinofuranosidases (32,42). One possibility is (Table 2). that the C-terminal module in family GH43 would The re-association of GH43 and Module-A in be responsible for switching substrate specificity T. fusca Xyl43A provided a key functional residue from arabinan to xylobiose. However, in the for β-xylosidase activity. The phenylalanine present work, the T. fusca GH43 module alone was located at position 518 of Module-A was thus unable to degrade arabinan and no binding ability demonstrated to be inserted in a critical position of either of the two Xyl43A modules towards and essential for Xyl43A enzymatic activity. The arabinan was observed, thus suggesting that the F518A replacement, either in the intact enzyme difference in specificities is a function of the Xyl43A or in Module-A, resulted in the near- topography and composition of the active site. complete loss of enzymatic activity. According to According to Alhassid, et al. (22), the topology of 9   Brux et al. (32), this conserved phenylalanine unclear, since it is phylogenetically unconnected emerges from a loop originated from the β- with any known protein in the databases other than sandwich domain and contributes to substrate those associated with the family-GH43 enzymes. binding by a stacking interaction with the xylose Perhaps an ancestral form of Module-A was unit at the -1 subsite. Interestingly, the amino-acid originally a CBM, since it possesses reminiscent replacement does not appear to affect the overall characteristics of a CBM (e.g., β-sandwich folding of the enzyme according to the re- structure); its binding role would presumably have association of the separated GH43 and mutated been lost during evolvement of its singular Module-A. catalytic contribution to the GH43 xylosidases. In a recent article (44), physical re-association Moreover, since Xyl43A does not contain any of a family-GH9 endoglucanase catalytic module signal peptide, the enzyme appears to remain in and its ancillary family-3c CBM was also the intracellular milieu in T. fusca and would not demonstrated with attendant regain of enzyme require a CBM to target the enzyme to its activity. The latter study, together with the current substrate. Indeed, xylobiose is small enough to work, promotes the original description of the enter the cell, presumably via an appropriate combined S-protein and S-peptide to generate an transport apparatus, and can be degraded internally active form of ribonuclease A (45). In the current into xylose units, so a CBM would be irrelevant studies on the glycoside hydrolases, however, the for the enzyme. Consequently, as representative of D o two interacting modules are clearly of distinct the GH43 β-xylosidases, the two linked modules w n evolutionary origin. of T. fusca Xyl43A should thus be considered as loa d Indeed, the two separate modules of Xyl43A an intact, synchronized functional unit, rather than ed fold on their own, suggesting that originally they a fortuitous pair of independent modules. fro m may have had distinct functions. The evolutionary h ttp source of the C-terminal Module-A, however, is ://w w w .jb c .o rg b/ y g u e s t o n A p ril 5 , 2 0 1 9 10   REFERENCES 1. Kulkarni, N., Shendye, A., and Rao, M. (1999) FEMS Microbiol Rev 23, 411-456 2. Beg, Q. K., Kapoor, M., Mahajan, L., and Hoondal, G. S. (2001) Appl Microbiol Biotechnol 56, 326-338 3. Wilson, D. B. (1992) Crit. Rev. Biotechnol. 12, 45-63 4. Wilson, D. B. (2004) Chem. Rec. 4, 72-82 5. Bachmann, S. L., and McCarthy, A. J. (1991) Appl. Environ. Microbiol. 57, 2121-2130 6. Ghangas, G. S., Hu, Y. J., and Wilson, D. B. (1989) J. Bacteriol. 171, 2963-2969 7. Irwin, D., Jung, E. D., and Wilson, D. B. (1994) Appl. Environ. Microbiol. 60, 763-770 8. Kim, J. H., Irwin, D., and Wilson, D. B. (2004) Can. J. Microbiol. 50, 835-843 9. Blanco, J., Coque, J. J., Velasco, J., and Martin, J. F. (1997) Appl Microbiol Biotechnol 48, 208-217 10. Cantarel, B. L., Coutinho, P. M., Rancurel, C., Bernard, T., Lombard, V., and Henrissat, B. (2009) Nucleic Acids Research 37, D233-238 11. Moriyama, H., Fukusaki, E., Cabrera Crespo, J., Shinmyo, A., and Okada, H. (1987) Eur J Biochem 166, 539-545 12. La Grange, D. C., Pretorius, I. S., and van Zyl, W. H. (1997) Appl Microbiol Biotechnol 47, 262-266 13. Chun, Y. C., Jung, K. H., Lee, J. C., Park, S. H., Chung, H. K., and Yoon, K. H. (1998) J. Microbiol. Biotechnol. 8, 28-33 14. Whitehead, T. R., and Cotta, M. A. (2001) Curr Microbiol 43, 293-298 D o w 15. Shallom, D., Leon, M., Bravman, T., Ben-David, A., Zaide, G., Belakhov, V., Shoham, G., Schomburg, n lo D., Baasov, T., and Shoham, Y. (2005) Biochemistry 44, 387-397 ad e d 16. Smaali, I., Remond, C., and O'Donohue, M. J. (2006) Appl Microbiol Biotechnol 73, 582-590 fro 17. Gosalbes, M. J., Perez-Gonzalez, J. A., Gonzalez, R., and Navarro, A. (1991) J Bacteriol 173, 7705- m h 7710 ttp 18. Bourgois, T. M., Van Craeyveld, V., Van Campenhout, S., Courtin, C. M., Delcour, J. A., Robben, J., ://w w and Volckaert, G. (2007) Appl Microbiol Biotechnol 75, 1309-1317 w 19. McKie, V. A., Black, G. W., Millward-Sadler, S. J., Hazlewood, G. P., Laurie, J. I., and Gilbert, H. J. .jbc .o (1997) Biochem J 323 ( Pt 2), 547-555 rg b/ 20. Nurizzo, D., Turkenburg, J. P., Charnock, S. J., Roberts, S. M., Dodson, E. J., McKie, V. A., Taylor, E. y g J., Gilbert, H. J., and Davies, G. J. (2002) Nat Struct Biol 9, 665-668 ue s 21. Leal, T. F., and de Sa-Nogueira, I. (2004) FEMS Microbiol Lett 241, 41-48 t o n 22. Alhassid, A., Ben-David, A., Tabachnikov, O., Libster, D., Naveh, E., Zolotnitsky, G., Shoham, Y., and A p Shoham, G. (2009) Biochem J 422, 73-82 ril 5 23. Matsuo, N., Kaneko, S., Kuno, A., Kobayashi, H., and Kusakabe, I. (2000) Biochem J 346 Pt 1, 9-15 , 20 1 24. van den Broek, L. A., Lloyd, R. M., Beldman, G., Verdoes, J. C., McCleary, B. V., and Voragen, A. G. 9 (2005) Appl Microbiol Biotechnol 67, 641-647 25. Vandermarliere, E., Bourgois, T. M., Van Campenhout, S., Strelkov, S. V., Volckaert, G., Delcour, J. A., Courtin, C. M., and Rabijns, A. (2007) Acta Crystallogr Sect F Struct Biol Cryst Commun 63, 692-694 26. Ichinose, H., Yoshida, M., Kotake, T., Kuno, A., Igarashi, K., Tsumuraya, Y., Samejima, M., Hirabayashi, J., Kobayashi, H., and Kaneko, S. (2005) J Biol Chem 280, 25820-25829 27. Caspi, J., Irwin, D., Lamed, R., Shoham, Y., Fierobe, H.-P., Wilson, D. B., and Bayer, E. A. (2006) Biocat. Biotransform. 24, 3-12 28. Kluepfel, D., Vats-Mehta, S., Aumont, F., Shareck, F., and Morosoli, R. (1990) Biochem J 267, 45-50 29. Fierobe, H.-P., Mingardon, F., Mechaly, A., Belaich, A., Rincon, M. T., Lamed, R., Tardif, C., Belaich, J.-P., and Bayer, E. A. (2005) J. Biol. Chem. 280, 16325-16334 30. Tabka, M. G., Herpoël-Gimbert, I., Monod, F., Asther, M., and Sigoillot, J. C. (2006) Enzyme Microb. Technol. 39, 897-902 31. Kukolya, J., Nagy, I., Laday, M., Toth, E., Oravecz, O., Marialigeti, K., and Hornok, L. (2002) Int J Syst Evol Microbiol 52, 1193-1199 32. Brux, C., Ben-David, A., Shallom-Shezifi, D., Leon, M., Niefind, K., Shoham, G., Shoham, Y., and Schomburg, D. (2006) J Mol Biol 359, 97-109