Improving Thermostability of GH11 Xylanase XynASP by the Design of Loop Region

Abstract

Highly flexible loop regions can affect the structure-function relationship of enzymes. In this study, to reveal the effect of the loop on the thermal stability of GH11 xylanase XynASP from Aspergillus saccharolyticus JOP 1030-1, three mutants (T41V, A79Y, T81Q) located in the loop region were predicted by the FireProt web server and constructed by site-directed mutagenesis. Heat tolerance experiments were performed in the mutants and wild-type XynASP, as well as for previously reported mutant T41W, indicating that the thermostability of enzymes from weak to strong was wild-type XynASP, T41V, T81Q, A79Y and T41W. Novel hydrophobic contacts and hydrogen bonds in the single mutants were found and explained for enhanced thermostability. T41W, A79Y and T81Q were combined by site-directed mutagenesis. The combined double mutants T41W/A79Y and A79Y/T81Q displayed significantly higher thermostability than that of the wild-type, yet lower than that of the robust mutant T41W after 30 min of incubation at 35–60 °C. But the triple mutant T41W/A79Y/T81Q displayed a slight improvement in thermal stability compared to T41W after 30 min of incubation at 35–60 °C. Meanwhile, T41W/A79Y/T81Q exhibited substantially improved thermostability with a half-life of inactivation enhanced from 20.7 min to 127.0 min at 45 °C. Furthermore, the optimum temperature increased by 10 °C compared to the wild-type XynASP. The pH stability of T41W/A79Y/T81Q at pH 3.0-8.0 all obtained more than 88% residual activities, displaying a significant improvement compared to the wild-type XynASP (40–70% residual activities) at the same conditions. This study confirms that designing GH11xylanases in the loop region is an effective strategy for obtaining thermostabilizing xylanases.

1. Introduction

Xylanases (EC 3.2.1.8) can hydrolyze the β-1,4-glycoside bond of xylan, which is considered the key enzyme for xylan degradation [1]. Xylanases have great potential for application in paper making, feed processing, and food processing [2,3]. For example, xylanases can be used as additives in the pulp bleaching industry, which can effectively reduce the use of chemicals and environmental pollution [4]. The extensive applications in various industrial fields have inspired researchers to develop xylanases with excellent performance [5]. Based on the structure, sequence, and catalytic domain differences, xylanases are classified into the glycoside hydrolase (GH) family 5, 7, 8, 10, 11, 30 and 43. Among them, GH10 and GH11 xylanases were focused on for further research [6,7].

GH10 xylanases’ structure is composed mainly of eight pairs of (β/α)₈ TIM barrel fold with a shallow active-site cleft [8]. GH11 xylanases are composed of two twisted anti-parallel β-sheets and one single α-helix that form the structure resembling the shape of a partially closed right hand, which employ the double exchange catalytic mechanism using two glutamates (Glu) as the catalytic residues [9,10]. Compared to GH10 xylanases, more attention is paid to GH11 xylanases due to their strict substrate selectivity, simple structure, and high catalytic efficiency [11,12]. However, most GH11 xylanases have poor thermal stability, which limits the industrial application of the enzyme [13]. Therefore, developing novel thermostabilizing GH11 xylanases by protein engineering is an effective approach to promote the industrial application of enzymes.

Several studies have been carried out to enhance the thermal stability of xylanases with protein engineering, which contained different strategies, such as directed evolution approaches and rational/semi-rational design [14,15,16]. Based on the above strategies, fruitful works have been carried out around the main secondary structure of GH11 xylanases, such as β-folding and α-helix structures. Many studies have found that factors such as hydrophobic interactions, disulfide bonds, salt bridges, hydrogen bonds, and amino acid ratios in these secondary structures significantly affect the thermal stability of xylanases [13]. Recent studies have revealed that reducing the flexibility and reinforcing the rigidity of enzymes improve thermal stability [17]. The loop of GH11 xylanases shows high flexibility, which is easily influenced by the external environment (e.g., pH, temperature, interaction with the substrate, etc.), thereby affecting its enzymatic properties (enzyme activity, stability, etc.) [18,19]. Therefore, enhancing the rigidity of loop to stabilize the protein structure and improve the thermal stability has become a new challenge and excited great interest for researchers [20].

Numerous studies showed that several in silico tools were available and effective for designing the thermostable mutants of enzymes. Many xylanases have been engineered to improve heat resistance by in silico strategy [21,22]. The FireProt web server combining energy- and evolution-based strategies is an effective tool for the prediction and design of thermostabilizing mutants, which could design single-point mutants, as well as predict multiple-point mutants with a rational combination strategy of single-point variants (Scheme 1) [23]. As a powerful strategy, the FireProt web server integrating multiple computer-aided design tools can rapidly calculate and analyze the results while designing robust mutants. This server has been successfully used to improve the thermostability of monooxygenase [24], serine protease [25], nitrile hydratase [26] as well as ketoreductase [27].

In a previous study, we mined a mesophilic GH11 xylanase XynASP (Geneback accession No. XM_025579666) deriving from Aspergillus saccharolyticus JOP 1030-1 by using genomic data mining technology. Based on structure and sequence analysis, the aromatic amino acid tryptophan was introduced at the 41^st site of the loop of XynASP to design the variant T41W, which exhibited a substantially improved thermostability. It proved that loop might play a crucial role in improving the thermal stability of xylanase XynASP [28]. In this study, based on the FireProt web server, a loop was designed to obtain thermostable mutants of XynASP. Thr41, Ala79 and Thr81 in the loop are identified as the key amino acid residues for the thermal stability of wild-type XynASP. Subsequently, three single mutants (T41V, A79Y, T81Q) were obtained by site-directed mutagenesis. Three single mutants all displayed higher thermostability than that of the wild-type. Moreover, the thermostability mechanism was explored by structural analysis. To explore the combined effect of multiple-site mutants, the single mutants were combined into three double mutants and one triple mutant. Multiple-site mutants also exhibited higher thermostability than that of the wild-type. Facilitated by the FireProt web server, the successfully thermostable mutants were obtained by design in the loop region, which provides new insights into engineering thermostable xylanase.

2. Results

2.1. Identification of The Thermal Stability Key Residues

A three-dimensional model of XynASP was constructed using the endo-1 4-beta-xylanase 11C (PDB No. 3WP3.1.A) as the template, which shared 64.21% identity with XynASP at the amino acid level. Quality analysis of the model revealed a GMQE score of 0.83 and a QMEAN score of 0.64, which were within the confidence intervals, indicating that the model was of high quality [28,29]. The model of the wild-type is shown in Figure 1. The wild-type XynASP mainly consists of two twisted anti-parallel β-sheets and one single α-helix, exhibiting a right-handed semi-grip structure and belonging to GH11 family xylanases. The catalytic residues of the wild-type XynASP are Glu110 and Glu201 (Figure 1).

Based on the model of the wild-type XynASP, the FireProt web server predicted the thermal stability-related sites from the energy and evolution strategy; eighteen mutant sites were obtained, respectively (Supplementary Material, Table S1). Among them are seven mutant sites, Thr41, Ser66, Ala79, Thr81, Ser94, Ala125 and Gly126, located in the loop region of wild-type XynASP (Figure 1). According to the principle that the lower the folding free energy (ΔΔG^fold), the more stable the structure is [30], three predicted sites with the smallest FoldX values, namely T41V, A79Y and T81Q, were selected based on the change of folding free energy of single mutants.

Numerous studies have shown that aromatic amino acids play a crucial role in the thermal stability of GH11 family xylanases [28,31]. In our previous study, a mutant T41W was constructed and characterized, exhibiting higher thermal stability than the wild-type XynASP [28]. Interestingly, Thr41 also was predicted by the FireProt as a potential beneficial site, substituting Thr41 with Val. Therefore, the effect of T41W and T41V on thermostability was compared in this study.

2.2. Construction and Characterization of Three Single Mutants

Based on the FireProt web server, mutants were constructed by site-directed mutagenesis. The recombinant xylanases were expressed into E.coli BL21(DE3) and purified to electrophoretic homogeneity as delineated already. Every single mutant and the wild-type XynASP exhibited similar molecular mass (~27 kDa) by SDS-PAGE (Figure 2).

2.3. Effects of Temperature on the Single Mutants

The optimal temperature of the mutants T41W, A79Y and T81Q was 55 °C, 55 °C and 50 °C, respectively. The optimal temperature of the wild-type was 45 °C, which indicated that the wild-type XynASP acquired a thermophilic property after mutation of these three sites in the loop region (Figure 3A). Unfortunately, the mutant T41V was similar to the wild type and retained the optimum temperature of 45 °C. The residual activities of four single mutants were tested after incubation in the range of 35–60 °C for 30 min to determine their respective thermostability. As shown in Figure 3B, after incubating 30 min at 40 °C, the residual activities of A79Y and T81Q were 67.9% and 55.6%, respectively, which showed significantly higher than that of wild-type (46.9%). When the incubation of temperature ≥45 °C, the residual activities of A79Y and T81Q were still slightly higher than that of the wild-type (Figure 3B). However, the activity of T41V was inactivated nearly. Meanwhile, after incubation at 50–60 °C for 30 min. Nevertheless, after incubation in the range of 35–60 °C for 30 min, the mutant T41W displayed higher thermal stability than the wild-type XynASP. T41W still retained 19.3% of activity after incubating 30 min at 60 °C, while the activity of XynASP decreased to 5.1% at the same condition. To sum up, T41V demonstrated slightly lower thermal stability than T41W. Therefore, considering the thermal stability of the T41V and T41W, three single mutants T41W, A79Y and T81Q, were selected for further combined multiple-site mutations.

2.4. Enhanced Thermal Stability of the Single Mutants Explored by Structural Simulations

To understand the thermostability mechanisms of the substitutions, the three-dimensional structure of four single mutants was constructed using the wild-type as the template. Mutations with enhanced heat resistance may mainly be attributed to hydrogen bonds, improvement of hydrophobic packing, stabilization of flexible loops and formation of disulfide bridges [32]. To gain insight into the enhanced thermostability, the intramolecular interactions were monitored in the mutated residues by the DynaMut web server [33]. The server can assess the effect of mutations on protein dynamics and stability resulting from vibrational entropy changes, clarifying that the T41V mutation could form new hydrophobic contacts with Phe 17 and Val 26 Figure 4A vs. Figure 4B). The T41W mutation might form more hydrophobic contacts and hydrogen bonds with the surrounding residues (Figure 4A vs. Figure 4C). Meanwhile, the hydrophobic contacts of Trp41 with the surrounding residues were significantly more than Val41, which might explain why T41W had higher thermostability than that of T41V (Figure 4B vs. Figure 4C). The A79Y mutation adds new hydrogen bonds with Thr154 despite the reduction of ionic interactions (Figure 4D vs. Figure 4E). The T81Q mutation added new weak hydrophobic contact with Thr151, compared to the wild-type (Figure 4F vs. Figure 4G). Surface-charged groups, such as stability, play key roles in protein structure and function. To further decipher the enhancement of thermostability for four single mutants, the electrostatic characteristic was calculated by APBS (Figure 5). The electrostatic potential surface of all the single mutants was found to be similar to the wild-type XynASP, exhibiting fewer positive charges. This result shows that the mutations appear to not affect the electrostatic potential surface of XynASP.

2.5. Temperature Characterization of Combined Multiple-Site Mutants

In the thermal stability assay of single mutants, the mutants T41W, A79Y and T81Q displayed higher thermostability than the wild-type. To explore the combined effect of multiple-site mutants, the single mutants (T41W, A79Y and T81Q) were combined by site-directed mutagenesis. Three double mutants (T41W/A79Y, A79Y/T81Q and T41W/T81Q) and one triple mutant (T41W/A79Y/T81Q) were constructed by site-directed mutagenesis.

The optimal temperature of combined mutants was all 50 or 55 °C, which was 5 or 10 °C higher than that of wild-type, yet similar to that of the single mutants (Figure 6A). Even so, thermostability was evaluated by measuring residual activities of the multiple-site mutants and wild-type at 35–60 °C for 30 min of incubation. It was found that double-site mutants (T41W/A79Y and A79Y/T81Q) showed greater residual activity than the wild-type. By comparison, T41W/T81Q displayed poor thermostability, of which residual activity was similar to wild-type but much lower than that of the single mutant T41W at the incubated temperature ≥45 °C (Figure 6B). Significantly, after 30 min of incubation at 60 °C, the activity of T41W/A79Y and A79Y/T81Q retained 22.3% and 26.4%, respectively. This was significantly higher than that of wild-type (8.5%) under the same conditions. In addition, an additive effect was observed in A79Y/T81Q since the residual activities of A79Y and T81Q, which composes the above double-site mutation, was lower than that of A79Y/T81Q at 35–60 °C for 30 min of incubation. Nevertheless, an additive effect was not observed in the T41W/A79Y and T41W/T81Q since residual activities of the above double mutants were lower than that of T41W at the same condition. After 30 min of incubation at 45 °C, the residual activity of T41W/A79Y/T81Q was 82.9%, which was significantly higher than the wild-type XynASP (29.8%), T41W (71.0%), T41W/T81Q (29.9%), T41W/A79Y (78.4%) and A79Y/T81Q (44.4%). In addition, after 30 min of incubation at 40–55 °C, the residual activity of T41W/A79Y/T81Q was higher than the wild-type XynASP and other mutants (Figure 6B). This result showed that the triple mutant T41W/A79Y/T81Q displayed higher thermal stability than single mutants, double mutants, and wild-type XynASP.

2.6. Thermostability and pH Stability of the Triple mutant T41W/A79Y/T81Q

All the combined mutants exhibited high thermostability compared to that of the wild-type. Among them, the triple mutant T41W/A79Y/T81Q exhibited the highest thermal stability. To further explore the thermostability of three-site mutates, residual activities of T41W/A79Y/T81Q and wild-type thermostability were evaluated after incubating at 40 °C and 45 °C for 2 h (Figure 7a). It was found that the mutant T41W/A79Y/T81Q showed greater residual activity than the wild-type XynASP. After 2 h of incubation at 40 °C, T41W/A79Y/T81Q retained 60.9% activity, while the wild-type was inactivated rapidly and remained at 25.1% activity. Specifically, T41W/A79Y/T81Q retained more than 50% activity after incubation for 2 h at 45 °C, whereas the residual activity of the wild-type decreased rapidly and lost more than 80% of the original activity within 2 h. The half-lives of thermal inactivation (t_1/2)were estimated at 40 °C and 45 °C. The half-life of T41W/A79Y/T81Q at 40 °C was evaluated to be 177.8 min, 4.7-fold higher than that of wild-type (37.8 min). The half-life of T41W/A79Y/T81Q at 45 °C also represented an impressive result, which was enhanced from 20.7 min to 127.0 min, a 6.1-fold increase over the wild-type. After exploring the thermostability of the mutant T41W/A79Y/T81Q, the pH effect of T41W/A79Y/T81Q on the activity at various pH (from3.0 to 8.0) was checked (Figure 7b). Compared to the wild-type, the optimal pH of T41W/A79Y/T81Q showed a slight change, which declined from 6.0 to 5.0, and more than 50% of activity was observed at a pH range between 4.0 and 6.5. According to the pH stability assays, T41W/A79Y/T81Q showed a significant improvement after incubation 1 h at 35 °C, which displayed more than 88% residual activities at pH 3.0–8.0. Whereas the residual activities of wild-type were distributed in 40–70% at the same pH condition (Figure 7c). These results indicate that T41W/A79Y/T81Q has more pH resistance than the wild-type.

3. Materials and Methods

3.1. Materials

Fast site-Directed Mutagenesis Kit was purchased from Vazyme (Nanjing, China). Beechwood xylan was obtained from Sigma (Saint Louis, MO, USA). Isopropyl β-D-thiogalactoside (IPTG), 1kb DNA Ladder Marker, kanamycin (Kan), protein marker and dNTP were purchased from TaKaRa (Otsu, Japan). FastDigest Dpn I was purchased from New England Biolabs (County Road, Ipswich, MA, USA). 3,5-dinitrosalicylic acid, D-(+)-xylose, Ni-NTA sefinose (TM) resin kit, Bradford protein assay kit and plasmid DNA extraction mini kit were purchased from Sangon Biotech Co., Ltd. (Shanghai, China). All other chemicals were of analytical grade reagents.

The XynASP gene was obtained from Aspergillus saccharolyticus JOP 1030-1. E. coli DH5α was used to clone and sequence the recombinant vector. E. coli BL21 (DE3) cells (Invitrogen, USA) and expression plasmid pET-28a(+) (Novagen, Darmstadt, Germany) were used for expression experiments. The recombinant plasmid pET-28a-xynASP was constructed and maintained in our laboratory. The strains grew in low-salt Luria-Bertani (LB) medium containing 0.5% (w/v) yeast extract, 1% (w/v) peptone, and 1% (w/v) NaCl.

3.2. Homology Modeling and Prediction of Thermal Stability Sites

The amino acid sequence of xynASP without signal peptide was submitted to the online website SWISS-MODEL (https://swissmodel.expasy.org) for homology modeling [34]. The most reliable model, basis on the C-score, was selected and further optimized and evaluated with Discovery Studio 3.0 Client [35]. The model of the wild-type XynASP was submitted to the FireProt web server to predict and design thermostable mutants using the default parameters [23]. The effects of variants on protein stability were analyzed using DynaMut with the default parameters. The structure of xylanases was visualized and analyzed with PyMOL.

3.3. Site-Directed Mutagenesis

Three single mutants (T41V, A79Y, and T81Q), Three double mutants (T41W/A79Y, T41W/T81Q, A79Y/T81Q), one triple mutant (T41W/A79Y/T81Q) were constructed by introducing mutations to the wild-type XynASP, using Fast site-Directed Mutagenesis System according to the manufacturer’s instructions. Sequences of mutated primers are listed in Table S2. PCR cycling conditions consisted of an initial step of 3 min at 98 °C, followed by 25 cycles of denaturation at 98 °C for 30 s, annealing at 55 °C for 45 s, and elongation at 72 °C for 2 min; an extra elongation at 72 °C for 10 min. The PCR products were digested using Dpn I at 37 °C for 30 min, then transformed into E. coli DH5α. Finally, the positive clones were screened on Luria-Bertani (LB) medium plates containing Kan (50 ug/mL) at 37 °C for overnight culture. DNA sequencing was performed to verify the mutants by Sangon Biotech Co., Ltd.

3.4. Protein Expression and Purification

Expression vector pET-28a introduced the wild-type XynASP, and mutant genes were transformed into E. coli BL21 (DE3) for the expression of recombination xylanases. The transformants were inoculated in 2 mL LB medium containing Kan at 37 °C overnight. Afterward, 1 mL of the culture suspension was transferred into a fresh 100 mL LB medium containing Kan at 37 °C. The cultures were grown at 37 °C with 180 rpm until optical density (OD₆₀₀) reached 0.6–0.8. After that, IPTG (2 mM) was added to the culture suspension and incubated at 30 °C over 2 h for protein expression. Cells were then harvested by centrifugation (3000× g, 10 min, 4 °C) and resuspended with sodium dihydrogen phosphate-citrate buffer (pH 6.0). To obtain the crude enzyme solution of recombinant xylanases, supernatants were harvested after the cells were disrupted by sonication and centrifugation (12,000× g, 20 min, 4 °C). The recombinant xylanases containing the N-terminal His-tagged were purified using Ni-NTA Sefinose (TM) Resin Kit (Sangon Biotech), according to the manufacturer’s instructions. The cell-free extract was applied to a metal chelate affinity column charged with Ni²⁺ and equilibrated with binding buffer contained with 20 mM imidazole, then eluted by the elution buffer supplemented with 250 mM imidazole, and finally dialyzed against the same buffer without containing imidazole. SDS-PAGE determined the purified protein, and the concentration of enzyme was determined by the Bradford protein assay kit.

3.5. Enzyme Assays

The activity was analyzed by determining the release of reducing sugars (xylose) hydrolyzed from beechwood xylan using the 3,5-dinitrosalicylicacid (DNS) method [36]. The reaction mixture (sodium dihydrogen phosphate-citrate buffer, pH 6.0) containing 1.5 mL 0.5% (w/v) beechwood xylan and 1 mL diluted pure enzyme was incubated at 40 °C for 15 min. After that, the reaction was terminated by adding 2.5 mL DNS, then boiled for 7 min, and the absorption was measured at 540 nm. One unit (U) of xylanase activity was defined as the quantity required to produce 1 μmol of reduced sugar from beechwood xylan per minute at the above conditions. All assays in this work were carried out in triplicate.

3.6. Measurements of Temperature and pH Characteristics in the Wild-Type and Mutants

To determine the optimum temperature, the activity of purified enzymes was examined at a range of temperatures (35–60 °C) using 0.5% beechwood xylan as the substrate, as described above. The thermal stability of enzymes was measured by incubating the enzymes without substrate at a temperature ranging between 35 °C and 60 °C for 30 min. The half-lives (t_1/2) of enzymes were calculated according to reference [37]. Similarly, the optimum pH of enzymes under the optimum temperature was examined at a pH range (3.0–8.0), respectively. To determine the pH stability, the enzymes were incubated in various buffer systems at pH ranging from 3.0–8.0 for 1 h at 35 °C. The residual activity was determined using the standard assay condition.

3.7. Electrostatic Property Analysis

Electrostatic potential of the wild-type (XynASP) and mutants were calculated by the Adaptive Poisson-Boltzmann Solver (APBS) [38]. The model of mutants was built by SWISS-MODEL using the optimized model of XynASP as the template. The results of the calculations for the electrostatic potential molecular surface of enzymes were exhibited using the PyMOL software.

4. Discussion

Previous studies have clarified that the design of mutations at the β-sheets of the N-terminal region of GH11 xylanases as an effective strategy could enhance the thermostability of enzymes. This was ascribed to the thermally mediated unfolding of GH11 xylanases beginning in this region [30,39,40]. It was found that structural flexibility was an unfavorable factor for the thermal stability of enzymes [17]. In the GH11 xylanases, the loop is characterized by the irregular crimp, which has high flexibility and is subject to conformational changes due to environmental factors (pH, temperature, interaction with the substrate, etc.), thus affecting the thermal stability of xylanases. In this study, to gain insight into the influence of the loop on the thermal stability of GH11 xylanase XynASP, the FireProt web server calculated the key residues of thermostability in the loop regions. Three single mutants (T41V, A79Y, T81Q) were predicted and constructed by site-directed mutagenesis. Among them, A79Y and T81Q slightly improved thermostability than the wild-type XynASP. However, the previously reported T41W exhibited higher thermal stability than the wild-type. Therefore, electrostatic property and structural analysis explored the heat-resistant mechanisms of four mutants (T41V, T41W, A79Y, T81Q). In addition, the mutant T41W had higher heat resistance than the mutant T41V. Therefore, the mutants T41W, A79Y and T81Q were subsequently performed to combine mutations. An additive effect was observed in the combined double-site (A79Y/T81Q) and triple-site mutants (T41W/A79Y/T81Q). Therefore, there are several points worth discussing.

Firstly, the FireProt web server is a promising tool that predicts protein thermal stability related sites using energy- and evolution-based strategies. Energy strategy depends on the crystal structure information of protein, while the evolution strategy is based on the evolutionary information of amino acid sequence. The most important features of the FireProt web server were the predictability of multi-site mutants, eliminating some conservative and highly correlated amino acid sites, filtering some potential deleterious mutations and superposing single-site mutants, thus designing the multi-site mutants in one step. In fact, FoldX and Rosetta, used by FireProt to calculate the free energy of mutation, can predict the thermal stability of mutant. The algorithm focuses on the local stability of protein but fails to investigate the mutated sites for a dynamic influence of the overall structural protein. Meanwhile, the accuracy of the energy-based prediction depends on the quality of the protein structure provided. A low-resolution structure or homology model may affect the prediction accuracy of beneficial mutants. Thus, for not yet resolved the protein crystal structure, the prediction of thermostable mutants has a certain risk using the FireProt. Although the wild-type XynASP obtained a high-quality structural model by homology modeling, this wild-type model still varies somewhat from the actually resolved crystal structure, which may cause the mutant T41V predicted by FireProt to be a non-beneficial mutation. Experiments also verified that the thermal stability of T41V was significantly worse than that of T41W and XynASP.

Secondly, structures of thermophilic proteins are often associated with higher degrees of rigidity. The flexible region is not conducive to the stability of the protein [41]. Choosing the residues in the flexible regions of enzymes as a strategy designed for the thermostable mutants to enhance the rigidity of residues may get the desired result. In this study, based on the structural model of wild-type, the beneficial sites Thr41, Ala79 and Thr81 identified by FireProt are all located in the flexible loop region, which may contribute more to the overall stability.

Electrostatic characteristic analysis indicated that the electrostatic properties of four single mutants were not significantly altered compared to the wild-type. However, less positively and highly negative attaching to the surface of wild-type and mutants could display better thermostability, confirmed in one surface charge engineering study [42].

Thirdly, the intramolecular interactions may affect the thermal stability of enzymes. In our study, novel hydrophobic contacts and hydrogen bonds were discovered in four single mutants, enhancing the stability of xylanases. In addition, an interesting phenomenon was observed: the additive effect of mutants is incomplete in the combined multiple-site mutants. After 30 min of incubation at 35–60 °C, the residual activities of double mutants (T41W/A79Y, T41W/T81Q) were higher than that of the single A79Y and T81Q but lower than that of T41W. Nevertheless, the double mutants A79Y/T81Q displayed a slight additive effect, which exhibited higher thermal stability than A79Y and T81Q. Moreover, the triple mutant (T41W/A79Y/T81Q) also exhibited higher thermal stability than the single mutants (T41W, A79Y and T81Q). The additive effects of combined multiple-site mutants were not routinely in our study, which may be ascribed to the fact that the mutated sites are all located in the highly flexible loop regions. If we substitute with other regions, such as β-sheets, the combined effects for thermal stability may be altered. One study showed that combined multiple-site mutants located in the β-sheets clusters of GH11 xylanases displayed an additive effect [34]. Even so, our result leads to the conclusion that a loop can affect the thermostability of GH11 xylanases, in which a single-point mutation can improve the thermal stability of the enzyme. However, the mechanism for the additive effect of multiple-site mutants in the loop region needs to be verified in the future.

5. Conclusions

In this study, we successfully identified residues Thr41, Ala79 and Thr81 located in the loop region of wild-type XynASP for the key sites of thermal stability and constructed the single mutants (T41V, A79Y, T81Q). Experiments exhibited that A79Y and T81Q showed higher thermostability compared to the wild-type. However, T41V showed the least thermostability among all the three single mutants. Since we previously reported that the mutant T41W has higher thermal stability compared to T41V, heat resistant mechanisms of four single mutants (T41W, T41V, A79Y, T81Q) were explored by structural analysis simultaneously. Novel hydrophobic contacts and hydrogen bonds formed with the mutated residue lead to improved thermal stability. Combined double and triple mutants were also performed, which all displayed higher thermal stability than the wild-type but exhibited unconventional additive effects due to the mutation sites located in the flexible loop region. The triple mutant T41W/A79Y/T81Q displayed a substantially improved half-life at 45 °C, which was a 6.1-fold increase over the wild-type. The pH stability assays showed that T41W/A79Y/T81Q also significantly increased pH resistance compared to the wild-type. This study demonstrated that the loop region was crucial to the thermostability of GH11 xylanases.