|Home | About | Journals | Submit | Contact Us | Français|
Xylanases are utilized in a variety of industries for the breakdown of plant materials. Most native and engineered bifunctional/multifunctional xylanases have separate catalytic domains within the same polypeptide chain. Here we report a new bifunctional xylanase (XynBE18) produced by Paenibacillus sp. E18 with xylanase and β-1,3-1,4-glucanase activities derived from the same active center by substrate competition assays and site-directed mutagenesis of xylanase catalytic Glu residues (E129A and E236A). The gene consists of 981 bp, encodes 327 amino acids, and comprises only one catalytic domain that is highly homologous to the glycoside hydrolase family 10 xylanase catalytic domain. Recombinant XynBE18 purified from Escherichia coli BL21(DE3) showed specificity toward oat spelt xylan and birchwood xylan and β-1,3-1,4-glucan (barley β-glucan and lichenin). Homology modeling and molecular dynamic simulation were used to explore structure differences between XynBE18 and the monofunctional xylanase XynE2, which has enzymatic properties similar to those of XynBE18 but does not hydrolyze β-1,3-1,4-glucan. The cleft containing the active site of XynBE18 is larger than that of XynE2, suggesting that XynBE18 is able to bind larger substrates such as barley β-glucan and lichenin. Further molecular docking studies revealed that XynBE18 can accommodate xylan and β-1,3-1,4-glucan, but XynE2 is only accessible to xylan. These results indicate a previously unidentified structure-function relationship for substrate specificities among family 10 xylanases.
Cellulose, hemicellulose, and lignin are the major components of plant cell walls. Hemicellulose and lignin provide a protective barrier against enzymatic attack of cellulose (15). Xylan is the major component of hemicellulose, the complete degradation of which requires a multistep process involving xylanases and various xylan debranching enzymes, such as β-xylosidase, acetylxylan-esterase, α-glucuronidase, and α-arabinofuranosidase (5, 28).
Xylanases, in conjunction with other enzymes (14), such as cellulases, glucanases, and proteases, are widely used in animal feed, brewing, food processing, and waste treatment, as well as in the pulp and paper industries (25, 32). For example, combined application of xylanase and β-1,3-1,4-glucanase can reduce the intestinal viscosity of feed for higher nutrition availability and improve the filtration rate and extraction yield in the brewing industry (16, 21). The natural diversity of enzymes provides these industries with candidates having bifunctional activity, such as the xylanases from Aspergillus niger (12) and Marasmius sp. (26) that have xylanase and cellulase activities and the bifunctional xylanase-lichenase from Ruminococcus flavefaciens 17 (7). On the other hand, many artificial bifunctional xylanases have been synthesized for more efficient biodegradation of plant fiber (19). Except for the bifunctional xylanase-glucanase from A. niger A-25, which has not been subjected to sequence and structural analysis but is conjectured to have one catalytic domain only based on molecular weight and kinetic analysis (4), all other bi- or multifunctional enzymes have separate catalytic domains with distinct substrate specificities.
Corn straw consists mainly of cellulose, hemicellulose, and lignin, and thus, the microorganisms present in corn straw often produce various enzymes such as endoglucanase, cellobiohydrolase, β-glucosidase, xylanase, and so on (33). In the present study, we selected corn ensilage as the microbial source for the isolation of multifunctional xylanases. A Paenibacillus sp. strain with high xylanolytic activity was isolated, and the xylanase gene was cloned. Analysis of the sequence and enzyme properties and kinetics revealed that the xylanase gene encodes a bifunctional xylanase-glucanase with a single catalytic domain. Further homology modeling and molecular dynamic (MD) studies confirmed that the bifunctional protein has a substrate binding cleft large enough to accommodate xylan or β-1,3-1,4-glucan. These results suggest that this enzyme is a new glycoside hydrolase (GH) family 10 (http://www.cazy.org/) bifunctional xylanase-glucanase with a single catalytic domain (3).
The medium for expansion and isolation of xylan-degrading strains from corn ensilage was used as described by Lv et al. (20). Corn ensilage (about 5 g) was inoculated into 300 ml isolation medium in a 500-ml Erlenmeyer flask and incubated at 50°C for 4 days. The mixed suspension was serially diluted with sterile water, spread onto xylan agar plates (2.5 g l−1 xylan), and incubated at 30°C for 48 h. The plates were stained with 0.5% (wt/vol) Congo red solution for 20 min and then destained with 1 M NaCl. Strains with clear zones around the colonies were defined as xylan-degrading microorganisms. One strain, designated E18, showing the highest xylanolytic activity was identified and used for further analysis.
Escherichia coli JM109 (TaKaRa) and BL21(DE3) (Novagen) were maintained in Luria-Bertani (LB) broth or on agar plates at 37°C for recombinant plasmid amplification and protein expression, respectively. Vectors pGEM-T Easy (Promega) and pET-28a(+) (Novagen) were used for plasmid preparation and gene cloning and for gene expression, respectively. Oat spelt xylan, birchwood xylan, laminarin, lichenan, carboxymethyl cellulose (CMC), and Avicel were purchased from Sigma. The standard xylooligosaccharides, xylose, xylobiose, xylotriose, and xylotetraose, were purchased from Wako Pure Chemicals. All of the other chemicals used were of analytical grade and are commercially available.
Genomic DNA was isolated from strain E18 and subjected to PCR amplification using degenerate primers specific for the GH 10 xylanase genes (xyl10-F, 5′-TACGACTGGGAYGTNGDIAAYGA-3′; xyl10-R, 5′-GTGACTCTGGAWICCNAYICCRT-3′) and used to amplify xylanase gene fragments by PCR. The PCR products were purified and ligated into the pGEM-T Easy vector for sequencing. Based on the partial sequence, the 5′ and 3′ flanking regions were isolated by thermal asymmetric interlaced (TAIL)-PCR (18) using two sets of specific primers and the Genome Walking kit (TaKaRa) according to the manufacturer's instructions.
The gene fragment encoding mature XynBE18 was amplified from strain E18 genomic DNA by PCR using Pfu DNA polymerase and two synthetic primers (XynBE18 EF, 5′-GGGCATATGAGGTTGCGTGAAGCGTTTAAAAAG-3′ [NdeI site underlined]; XynBE18 ER, 5′-GGGAAGCTTTCATATGGTTTCAAGCACCTTCCAATAAGAC-3′ [HindIII site underlined]). The PCR product was cloned into the NdeI and HindIII sites of pET-28a(+) and then transformed into E. coli BL21(DE3). Positive transformants, identified by the Congo red method and confirmed by DNA sequencing, were cultured in liquid LB broth at 37°C to A600 of ~0.6. Protein expression was induced at 30°C for 6 h by addition of isopropyl-β-d-thiogalactopyranoside (IPTG) to a final concentration of 0.6 mM. The presence and activity of recombinant xylanase in the culture supernatant were determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and enzyme activity assays, respectively.
To purify recombinant XynBE18, the induced cells were harvested by centrifugation (12,000 × g, 4°C, 5 min) and sonicated on ice. After removal of the cell debris by centrifugation, the concentrated supernatant was loaded onto an Ni-nitrilotriacetic acid chelating column for purification with a linear imidazole gradient of 20 to 300 mM in Tris-HCl buffer (20 mM Tris-HCl, 500 mM NaCl, 10% glycerol, pH 7.6). The homogeneity of the purified enzyme was evaluated by SDS-PAGE. The protein concentration was determined by the Bradford method using bovine serum albumin as the standard (2).
The standard assays for xylanase and β-1,3-1,4-glucanase activities were performed at 50°C for 10 min in 10 mM McIlvaine buffer (pH 6.0) in the presence of 1.0% (wt/vol) substrate (19). The reducing sugar released was determined by the 3,5-dinitrosalicylic acid (DNS) method (22).
The optimal pH for recombinant XynBE18 activity was determined at 37°C in buffer containing 1.0% (wt/vol) oat spelt xylan or barley β-glucan at pH 5.0 to 12.0. The optimal temperature for XynBE18 activity was determined by performing the standard assay between 30°C and 70°C at the optimal pH for each substrate. With oat spelt xylan as the substrate, the pH stability of XynBE18 was estimated by preincubating the enzyme at 37°C for 1 h in the appropriate buffer (pH 2.0 to 12.0) and then measuring residual activity under standard conditions. The buffers for determining pH stability were McIlvaine buffer, pH 2.0 to 7.0; 0.1 M Tris-HCl, pH 7.0 to 9.0; and 0.1 M glycine-NaOH, pH 9.0 to 12.0. Thermal stability was monitored by assessing the residual xylanase activity under standard conditions after incubating the enzyme at 40°C or 50°C for 60 min without substrate.
Recombinant XynBE18 activity was also analyzed in the absence or presence of 1 mM or 10 mM Na+, K+, Li+, Mg2+, Ca2+, Hg2+, Ag+, Zn2+, Fe3+, Ni2+, Cu2+, SDS, EDTA, and β-mercaptoethanol.
The substrate specificities of recombinant XynBE18 were determined by measuring the enzyme activity after incubation in McIlvaine buffer containing 1.0% (wt/vol) substrate (birchwood xylan, oat spelt xylan, barley β-glucan, lichenan, CMC, or Avicel) at 50°C for 10 min by the DNS method.
The apparent Km and Vmax values for the xylanase or glucanase activity of XynBE18 were determined in McIlvaine buffer containing 1 to 10 mg ml−1 oat spelt xylan or barley β-glucan as the substrate, respectively. The data were determined from a Lineweaver-Burk plot using the nonlinear regression computer program GraFit. To determine whether recombinant XynBE18 utilizes the same active center to hydrolyze xylan and barley β-glucan, substrate competition assays were performed as described previously (4).
To further verify whether the xylanase and β-1,3-1,4-glucanase activities of XynBE18 were derived from the same active center, site-directed mutagenesis of the catalytic residues was performed using the QuikChange method (Stratagene). Point mutations Glu129Asp and Glu236Asp were generated in the wild-type XynBE18 DNA in the expression plasmid. The resulting mutant plasmids were transformed into E. coli BL21(DE3) cells and confirmed by DNA sequencing. Expression and purification of the mutant xylanase followed the same procedure as for wild-type XynBE18.
A family 10 intracellular xylanase (IXT6) from Geobacillus stearothermophilus (27) (Protein Data Bank entry 2Q8X; resolution, 1.45 Å) showing 63.6% sequence identity with XynBE18 was adopted as the template for three-dimensional structure modeling of XynBE18. The structure of XynBE18 was predicted and evaluated using Accelrys Discovery Studio software (DS 2.5). The family 10 intracellular xylanase from Anoxybacillus sp. E2 (XynE2) (31), which shows 61.1% identity with XynBE18 but only exhibits xylanase activity, was subjected to homology modeling with the IXT6 template for structure comparison analysis.
MD simulation was performed using the Gromacs program 4.0.5 (30) with Gromos 96.1 (53A6) force field (23, 24). XynBE18, XynE2, and IXT6 were each solvated in cubic boxes with a 90-Å side length using simple point charge water molecules, which were then replaced with counterions for electroneutrality. The system was then energy minimized with the steepest-descent algorithm for a 20-ps MD simulation at 300 K with uranium and calcium atoms fixed sequentially and then a 5-ns MD simulation at 300 K. All bond lengths were constrained using the LINCS algorithm. The cutoff value for van der Waals interactions was set to 1.0 nm, and electrostatic interactions were calculated using the particle-mesh Ewald algorithm. The time step of the simulation was set to 2 fs, and the coordinates were saved every 1 ps. Differences in the trajectories were analyzed using the Gromacs 4.0.5 package.
The structures of the XynBE18 and XynE2 protein clefts containing the active sites were calculated from the equilibrium ensemble of the last 20 ps using the Gromacs g_rmsf utility (17, 30). An automated feature-based docking simulation of xylan and β-glucan in the active site of the receptor was performed using the DS2.5 docking program combined with CHARMM for accurate receptor sampling. Residues Gln83, Asn175, Phe244, and Trp286 of XynBE18 and residues Gln85, Asn177, Phe246, and Trp288 of XynE2 were set as flexible residues. The generated protein and ligand conformations were 100 and 255, respectively. The root mean square deviation (RMSD) tolerance for any given match was set to 0.5 Å, and other parameters were set to default.
The nucleotide sequences of the Paenibacillus sp. E18 16S rRNA and xylanase gene (xynBE18) were deposited in GenBank under accession numbers FJ899682 and FJ899683, respectively.
Five bacterial strains isolated from corn ensilage had xylanolytic activity. Among them, strain E18 exhibited the most significant activity (1.5 U ml−1) in LB medium containing 1% (wt/vol) birchwood xylan. Strain E18 was classified into the genus Paenibacillus based on 16S rRNA gene sequences, which were 99% identical between strain E18 and Paenibacillus sp. FR1 (GenBank accession no. DQ912930).
A xylanase gene fragment (258 bp) was amplified by PCR using degenerate primers. The 5′ and 3′ flanking regions were amplified by TAIL-PCR and assembled with the known partial sequence. The full-length xylanase gene, xynBE18, contains 981 bp and encodes a protein of 327 amino acids with a calculated molecular mass of 37.8 kDa. The deduced amino acid sequence encoded by xynBE18 showed the greatest identity (85%) with a putative xylanase from Paenibacillus sp. HPL-001 and showed 65% identity with intracellular xylanase IXT6 from G. stearothermophilus (27). Amino acid sequence analysis indicated that XynBE18 has only one catalytic domain that contains two glutamate residues—Glu129 and Glu236—that are highly conserved among family 10 members (8).
After induction with 0.6 mM IPTG at 30°C for 6 h, a positive transformant with pET-28a(+) containing xynBE18 had a xylanase activity of 3.1 U ml−1. The recombinant xylanase was purified to apparent homogeneity using one-step affinity chromatography. The specific activity of the purified xylanase was 59.6 U mg−1 after 10.6-fold purification, with a final activity yield of 9.2%. The purified enzyme was composed of a single polypeptide with an apparent molecular mass of 39.0 kDa, as determined by SDS-PAGE analysis.
The optimal pHs for the xylanase and β-1,3-1,4-glucanase activities of recombinant XynBE18 were 7.5 to 9.0 and 6.5, respectively (Fig. (Fig.1A).1A). The optimum temperature for both enzymatic reactions was about 50°C (Fig. (Fig.1B1B).
Recombinant XynBE18 was stable over a pH range of 6.0 to 10.0, and it retained more than 50% of the xylanase activity after incubation for 1 h at pH 12.0. The enzyme was stable at 40°C but lost activity rapidly above 50°C.
Addition of 1 mM or 10 mM Na+, K+, Li+, Mg2+, Ca2+, or EDTA had no significant effect on the xylanase activity of recombinant XynBE18 (data not shown). However, the activity was strongly inhibited in the presence of 1 mM Hg2+ or Ag+ and partially inhibited by certain metal ions or chemicals at a 10 mM concentration in the following order: Zn2+ > Fe3+ > Ni2+ > SDS > Cu2+. Addition of 1 mM or 10 mM β-mercaptoethanol enhanced the activity ~1.2- or 1.5-fold, respectively (data not shown).
Substrate hydrolysis was determined using the DNS method. Purified recombinant XynBE18 exhibited the highest xylanase activity when birchwood xylan or oat spelt xylan was used as the substrate (Table (Table1),1), with relatively lower activities being measured for barley β-glucan (42.0%) and lichenan (33.2%). No activity was detected in the presence of CMC or Avicel.
The Km and Vmax values of XynBE18 were 0.51 mg ml−1 and 357.93 μmol min−1 ml−1, respectively, for oat spelt xylan and 4.42 mg ml−1 and 132.59 μmol min−1 ml−1, respectively, for barley β-glucan.
Substrate competition between oat spelt xylan and barley β-glucan is shown in Table Table2.2. The observed values for the overall rates of hydrolysis agreed well with the theoretical values (V1) calculated assuming a single dual-function catalytic site, suggesting that XynBE18 hydrolyzes both substrates at the same active center.
Two catalytic glutamic acid residues conserved in GH 10 xylanases were replaced with aspartic acid in mutant XynBE18. No xylanase and β-1,3-1,4-glucanase activity was detected.
The structures of XynBE18 and XynE2 were predicted based on the X-ray crystal structure of IXT6 (2Q8X). Using the Gromacs program, energy minimization calculations for the XynBE18 and XynE2 structures modeled using the IXT6 template indicated that the total energy of XynBE18 and XynE2 decreased substantially in the first 1 ns of MD simulation and then stabilized after a 2-ns equilibration (data not shown). Compared to the starting coordinates, the RMSD of the heavy atoms increased in the first 1 ns and then reached a plateau in the subsequent simulation time (data not shown). All of these properties converged after a 2-ns MD simulation, indicating that the models were stable and were sufficiently accurate for subsequent studies.
Based on the MD simulation calculation, the cleft of XynBE18 containing the active site is larger than that of XynE2 or IXT6 (Fig. (Fig.2A),2A), suggesting that XynBE18 could accept larger substrates such as β-1,3-1,4-glucan. To test this hypothesis, the substrates xylan and β-glucan were docked to XynBE18 and XynE2 using the flexible docking program in DS 2.5. Xylan and β-glucan were easily docked to the active site of XynBE18 (Fig. (Fig.2B),2B), confirming that XynBE18 is a bifunctional enzyme with only one catalytic domain. The active site of XynE2 was accessible to xylan but not to β-glucan (data not shown).
Microbial xylanases are critical for a variety of industrial applications. Substrate specificities of GH 10 xylanases have been reported. In addition to xylan hydrolysis, most of these enzymes catalyze the hydrolysis of glycosidic bonds in β-1,4-linked glucose-derived substrates. For example, Cex from Cellulomonas fimi hydrolyzes aryl-β-cellobiosides but Xyl10A from Streptomyces lividans has little activity against glucose-based polymers (6). In this study, XynEB18 exhibited substantial activity against β-1,3-1,4-glucan. Based on amino acid sequence analyses, XynBE18 was determined to be a bifunctional xylanase-glucanase with only one catalytic domain.
The catalytic mechanisms of bifunctional or multifunctional GH 10 xylanases remain poorly understood (1, 6, 10, 11). Both GH 10 xylanases and GH 17 β-glucanases have two glutamic acid residues conserved in GH-A enzymes that drive substrate catalysis (9) and belong to the same superfamily of enzymes with a triosephosphate isomerase (TIM)-like barrel structure. It is possible that organisms containing these enzymes have a common ancestor. The replacement of Glu129 or Glu236 with Asp resulted in a complete loss of both xylanase and β-1,3-1,4-glucanase activities, indicating that XynBE18 catalyzes the hydrolysis of xylan and β-1,3-1,4-glucan at the same active site within a single catalytic domain. The results of competition between oat spelt xylan and barley β-glucan also demonstrated that the hydrolysis of both substrates took place at the same active center. This suggests that the structural modifications of GH 10 xylanases which have a TIM-like barrel structure may give rise to the observed wide variation in substrate specificity (9).
Homology modeling, MD simulation, and molecular docking methods have been successfully used to study protein function (29). Also, molecular docking is important for the prediction of protein-ligand interactions and for understanding protein function (13, 29). In the present study, a three-dimensional model of XynBE18 was constructed by the homology modeling approach, followed by MD simulation to refine the initial model; models of XynBE18 binding to xylan and β-1,3-1,4-glucan were generated. Importantly, the homology-modeled structure of bifunctional XynBE18 revealed a substrate binding cleft larger than that of XynE2, which only exhibits xylanase activity. Of endo-acting glycosyl hydrolases with a TIM-like barrel structure, such as endoglucanases, chitinases, α-amylases, xylanases, and β-1,3-1,4-glucanases, a larger cleft or groove allows the binding of several sugar units and makes the enzyme active on a variety of substrates (9). The exact role of the large cleft in specific enzyme-substrate interactions in the case of XynBE18 requires further crystallization analysis of the substrate-enzyme complex. Furthermore, molecular docking studies suggested that XynBE18 binds xylan and β-1,3-1,4-glucan with high affinity but that XynE2 only binds xylan with high affinity (data not shown). These results were consistent with the experimental data and further supported the notion that XynBE18 is a bifunctional xylanase-glucanase containing a single catalytic domain. Future detailed analysis of the XynBE18-glucan complex structure is required to verify the substrate binding sites and catalytic residues of the enzyme. The discovery of this bifunctional xylanase-glucanase is seminal to future analyses of the catalytic mechanisms of glycosyl hydrolases that contain a TIM-like barrel structure.
This research was supported by the Chinese Key Program of Transgenic Plant Breeding (2008ZX08003-020B), the Earmarked Fund for Modern Agro-Industry Technology Research System (NYCYTX-42-G2-05), and the National High Technology Research and Development Program of China (863 Program, grant 2007AA100601).
Published ahead of print on 9 April 2010.