Improve thermostability of Bacillus sp. TS chitosanase through structure-based alignment

Chitosanases can catalyze the release of chitooligosaccharides which have a number of medical applications. Therefore, Chitosanases are good candidates for large-scale enzymatic synthesis due to their favorable thermostability properties and high catalytic efficiency. To further improve the thermostability of a chitosanase from Bacillus sp. TS, which has a half-life of 5.32 min, we mutated specific serine residues that we identified as potentially relevant through structure comparison with thermophilic CelA from Clostridium thermocellum. Out of a total of 15 mutants, three, namely S265G, S276A, and S347G, show higher thermostability. Their half-lives at 60 °C were calculated as 34.57 min, 36.79 min and 7.2 min. The Km values of S265G, S276A and S347G mutants show substrate binding ability comparable to that of the wild-type enzyme, while the S265G mutant displays a significant decrease of enzymatic activities. Additionally, we studied the synergistic effects of combined mutations, observing that all double mutants and the triple mutant are more stable than the wild-type enzyme and single mutants. Finally, we investigated the mechanisms which might give a reasonable explanation for the improved thermostability via comparative analysis of the resulting 3D structures.

Site-directed mutagenesis. The previously reported CsnTS gene was cloned into pET29a (+) vector between the unique restriction sites NdeI and XhoI to make the expression plasmid. We used this plasmid as template in site-directed mutagenesis PCR. PCR-amplification was performed to get the mutagenesis with the Fast Mutagenesis System kit from Transgen Biotech Co. LTD using primers that are listed in Table 1. The PCR Table 1. List of primers. Mutated nucleotides are in bold.

Name
Sequence Description   S51G_F  TTG AAA AAT GAT TTA GGT TCT TTA CCT  S51G   S51G_R  CCA CCA GGT AAA GAACC TAA ATC ATTT  S51G   S52G_F  AAA AAT GAT TTA TCT GGT TTA CCT GGT  S52G   S52G_R  TAA CCA CCA GGT AAACC AGA TAA ATCA  S52G   S112G_F  AAC TTT TAAA GGC TCT CAA AAT CCT A  S112G   S112G_R  TTT TGA GAGCC TTT AAA AGT TCT TGC  S112G   S126G_F TGG GTT GTC GCA GAT GGT AAA AAA GC S126G S126G_R CTT GTG CTT TTT TACC ATC TGC GACA S126G S135A_F CAA GGT CAT TTT GAT GCT GCT ACT GA S135A S135A_R CCC CAT CAG TAG CAGC ATC AAA ATGA S135A  T  S347G   S347G_R  TGA TTA TTGCC ATT TGT TAT ACT TGC  S347G   S355G_F  GTG GGT AAAT GGC GGT TGG GAT TGG  S355G   S355G_R  TCC CAA CCGCC ATT TAC CCA CTT TTG  S355G   S369G_F  AGG CTA TTTT GGT GAT AGT TAT AAT T  S369G   S369G_R  TAA CTA TCACC AAA ATA GCC TTC TCT  Enzymatic activity assay. Chitosanase activity was evaluated by the method described below. 1 g of chitosan was dissolved in reaction buffer (100 mM sodium acetate, pH 5.0) to make substrate solution. Chitosanase was added to the substrate solution to start the hydrolysis reaction. The hydrolysis of chitosan reaction was carried out at 50 °C for a quarter. The released reducing sugar was quantified by the dinitrosalicylic acid method 5 . In one minutes, the amount of chitosanase required to release 1 μmol reducing sugar was defined as 1 unit enzyme. D-glucosamine was used to make calibration curve. The reaction mixture without enzyme was measured in a control experiment.
Screening for mutants with increased thermostability. All plasmids to expression proteins mutants were transformed into expression host and spread on LB plates with 25 µg mL −1 of kanamycin. Fresh monocolonies were picked out and cultured in 50 mL of LB medium with according antibiotics at 37 °C with rotatory shaker. The induction was started by putting 0.1 mM IPTG at the culture OD 600 of 0.5-0.8. The cultures were centrifuged at 3000g for 15 min. The cell pellets were suspended with HT buffer that contains 10 mM MgCl 2 and 100 mM KCl dissolved in 50 mM HEPES-KOH (pH 7.6) and then disrupted by sonication. The crude enzyme solutions were obtained by centrifugation at 20,000g for 45 min at 4 °C. Crude enzyme solutions were then incubated at 55 °C for 30 min. 100 µL of each sample was used for enzymatic activity measurements.
Protein expression and purification. The protein expression was similar to the process described above but scale up the culture volume to 500 mL. Crude enzyme solutions were also obtained by sonification of E. coli pellets.
The over-expressed proteins were purified by NTA-Ni affinity chromatography resin using GE commercial His-Trap FF columns as reported previously 5,25 . The purified proteins were stored in a HT buffer at − 20 °C. The protein purity was determined to 98% by SDS-PAGE (12% gel), and the protein concentrations were evaluated using a Bradford protein assay kit.
Biochemical characteristics of the enzymes. The thermostabilities of purified wild-type CsnTS and its variants were determined by pre-incubating the proteins at 60 °C for various periods and then measuring the residual enzymatic activities under the standard assay conditions. The enzyme's half-life (t 1/2 ) was used to evaluate its thermostability.
To determine the optimal temperature, enzymatic activity was investigated at temperatures from 40 to 80 °C with 5 °C increments. For each protein, the maximal enzymatic activity under optimal temperature was defined to 100%, relative activity was the ratio of the enzymatic activity at an appointed temperature to the maximal activity.
The optimal pH was also investigated over a pH range of 3.5-7.5. pH 3.5-5.8 range was achieved using buffer with 100 mM sodium acetate and pH 6.0-7.5 range was achieved using 100 mM sodium phosphate buffer. All the reactions were performed at 50 °C. Like calculation in optimal temperature, the maximal activity at the optimal pH was defined to 100%, then relative activity was the ratio of the enzyme activity to the maximal activity at an indicated pH.
The reaction to obtain kinetic parameters of the wild-type CsnTS and its variants were carried out under standard enzymatic assay conditions. In the reactions, the concentration of enzyme was 0.5 U/mL enzymes, while chitosan concentrations ranged from 0.1 to 10 mg/mL. The experiment results were analyzed by fitting data sets to the enzyme kinetic equation of the Michaelis-Menten model using GraphPad Prism.
Thermal unfolding of the enzymes. The T m values of proteins was determined using Differential Scanning Calorimetry (DSC) measurements with a Nano DSC scanning microcalorimeter (Model 5100, Calorimetry Science Corporation, Utah, USA). The concentration of each protein sample was 0.2 mg/mL in HT buffer. And increased the temperature from 40 to 80 °C at a heating rate of 1 °C per min, with pressure set to 3.0 atm. The data were analyzed using the self-contained software NanoAnalyze.

Results
Identification of mutants with increased thermostability. Clostridium thermocellum is a thermophilic bacterium. Its protein CelA possesses an optimal enzymatic temperature of 75 °C 23,27 . The structure alignment of CsnTS and CelA from Clostridium thermocellum was performed in order to determine potential target residues for subsequent engineering of mutations (Fig. 1). 15 serine residues were selected and mutated to alanine or glycine. The mutants were overexpressed in E. coli BL21 (DE3) strain. The harvested cells were crushed by ultra-sonication and lysates were used for activity measurement. The activities of the wild-type CsnTS and 15 mutations were evaluated with or without heat treatment. The retention of enzymatic activity was measured after incubating the samples for 30 min at 55 °C. The wild-type CsnTS retained approximately 20% of its activity after heating treatment. As shown in Fig. 2, the preliminary screening results showed that mutants S265G, S276A and S347G displayed enhanced thermostability. Consequently, these three mutants were purified for further studies.  Fig. 3a. All mutants display a wider range of temperature stability than the wild-type CsnTS. Notably, the mutant S276A exhibits an enzymatic activity of 70%, even at a high temperature of 80 °C while the wild-type enzyme has completely lost its activity at this temperature. The optimal temperature of S347G is 55 °C, which is lower than that of the wild-type enzyme. To study the relationship between the thermal stability and protein conformation, a DSC assay was applied to measure the T m values of the wild-type and all mutated enzymes. The results are shown in Table 2 and Figure S1. The T m values of mutants S265G and S276A are increased by 3 °C and 2 °C, respectively, as compared to the T m of 62 °C of the wild-type enzyme, while S347G displays the decreased T m value. The optimal pH values of the mutated enzymes are identical to that of the wild-type CsnTS (data not shown). The specific activities of mutants S276A and S347G are 456 and 602 U/mg and thus similar to the specific activity of wild-type CsnTS of 566 U/mg, whereas the specific activity of mutant S265G is 279 U/mg and hence significantly lower. The K m values of all three mutants are slightly increased. The K cat of mutant S276A is increased while the K cat of mutant S265G is significantly decreased. Lastly, the S347G mutant has a K cat value similar to that of wild-type CsnTS ( Table 3).

Combinations of beneficial mutations.
Recent studies have shown that combining two or more beneficial mutations can further improve the thermostability of proteins 7,10,28 . Therefore, we constructed mutant combinations from the identified three mutants to further improve the thermostability of CsnTS.   (Table 2), respectively. Except for the S265G/S347G mutant, all other combined mutants display a significantly higher thermostability than those of the single mutants. The T m values of these combined mutants are about 6 °C higher than the T m of wild-type CsnTS. In addition, all combined mutants display stability over a much wider temperature range (Fig. 3b); they all remain enzymatically active on a 80% level for temperatures between 45 and 75 °C. Remarkably, mutant S276A/S347G performed best by exhibiting enzymatic activity between 40 and 80 °C, with an optimal temperature between 50 and 65 °C.
Moreover, the optimal pH values and pH stabilities of the double and triple mutated enzymes are similar to those of wild-type CsnTS and single-site mutants (data not shown). Although the S265G single-site mutant has a significantly reduced enzymatic activity, all other multi-site mutants that include S265G display catalytic abilities similar to that of wild-type CsnTS. Table 2. Activity and thermostability of csnTS and its mutants: specific activity, t 1/2 (60 °C) and optimum temperature.

Enzyme
Specific activity (U/mg) t 1/2 (60 °C) (min) T m (°C)  www.nature.com/scientificreports/ We anticipate our thermostable mutations will become very useful for industrial applications, and may also be further improved by using a methodology similar as presented herein.

Structural interpretation for increased thermostability in mutants.
To better understand the factors that affect the thermostability of these mutants, we compared the generated structures of CsnTS and mutants. As has been previously confirmed, Ser265 forms hydrogen bonds with nearby residues in the endoglucanase CelA 23 . With Gly replacing Ser265, the mutated enzyme has now a half-life of 34.57, which is ~ 7 times longer than the half-life of wild-type chitosanase. On the other hand, the specific activity and catalytic ability of the S265G mutant decreased compared to that of the wild type. Ser265 is on the flexible loop adjacent to the substrate binding position (Fig. 1). Intramolecular interaction analysis revealed that the interaction between S265 and Y267 in wild-type CsnTS is replaced by the interaction between G265 and T268 in the S265G mutant (Fig. 4, Table S1). This change in interactions may stabilize the loop. Moreover, the number of intramolecular interactions increased in the S265G mutant as compared to that in the wild-type enzyme. These interactions reduce the flexibility of the loop, thus improving the mutant's thermostability while resulting in a loss of its catalytic ability. This is in line with previous observations where site-directed mutagenesis of proteins failed to simultaneously improve thermostability and enzymatic activity 29,30 .
The Ala276-Phe335 interaction in the S276A mutant substitutes the Ser276-Tyr273 interaction in the wildtype chitosanase (Fig. 5, Table S3). The Ser/Ala276 residue is on the α-helix adhered to the enzymatic activity site, Tyr273 is on the loop at the end of the same α-helix, and Phe335 is nearby but on another α-helix. The substituted intramolecular interaction stabilizes the tertiary structure of the protein. Additionally, it also has been reported that alanine residues are beneficial for the α-helix stability 31 .
Lastly, our analysis of the S265G/S276A/S347G mutant revealed that its intramolecular interactions show many re-arrangements (Table S4). Notably, far more interactions disappeared as compared with those that were created because of the mutation. This could indicate that proper structure arrangements rather than individual intramolecular interactions stabilize the proteins. Table 3. Kinetic parameters of wild-type and mutated csnTS.

Discussion
The thermostability of proteins can be determined by many important structural features 32 , including hydrogen bonds 33 , salt bridges 19,34 , aromatic π-π interactions, and cation-π interactions. Thermostability can be improved by a number of methods 6 . One such method compares the sequence of a protein with that of a thermostable and highly homologous counterpart. This can suggest residues that are potentially influencing thermostability and may be modified in the original protein. By engineering and screening of enzymes with mutations at such residues, mutants with improved thermostability have been identified for several enzymes already 14,15,35,36 .
In this study, we successfully adopted this method and engineered CsnTS mutants with improved thermostability. We selected several serines as mutation targets through structure comparison of CsnTS and CelA from Clostridium thermocellum and simultaneously considering statistical data revealing that serine residues are not favorable for thermostability 23,24 . By screening a total of 15 mutants with a single mutation in one serine residue, we identified three mutations, namely S265G, S276A and S347G, which are particularly beneficial for our purpose. Our study shows that all mutants, with single-site mutations and with combined mutations, exhibit a higher thermostability than the wild-type chitosanase. Moreover, all combined mutants outperformed the single-site mutants.
Statistical analysis of structural distribution of amino acids between thermophilic and mesophilic proteins revealed that serine was observed at low frequency at the surface 21,22,37 . In single serine mutants, the increased intramolecular interactions were found to be the main factors that could enhance thermal stability 38 . However, the number of intramolecular interactions of the multiple mutant S265G/S276A/S347G, is lower than that of the wild-type CsnTS. Thus, we hypothesize the re-arrangements in intramolecular interactions may contribute most toward the enhanced thermostability of the multiple mutant S265G/S276A/S347G.
In summary, we engineered mutants of chitosanase from Bacillus sp. TS following structure comparison with CelA from Clostridium thermocellum. Mutants S276A and S347G were identified as those that exhibit a higher thermostability in comparison to the wild-type chitosanase without losing catalytic activity. In addition, most of combinations of these mutations enhanced the thermostability of the enzyme further, making these mutants more enzymatically active during hydrolysis of chitosan than wild-type chitosanase. We presented a promising avenue for rational design and anticipate our recombinant mutants will have great potential for industrial applications.

Data availability
The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.