Improved activity of β-cyclodextrin glycosyltransferase from Bacillus sp. N-227 via mutagenesis of the conserved residues.

β-Cyclodextrin glycosyltransferase (β-CGTase) belongs to the α-amylase family of enzymes and converts starch to cyclic oligosaccharides called β-cyclodextrins (β-CD). The β-CGTase from alkalophilic Bacillus sp. N-227 was separately mutagenized to give three site-directed β-CGTase mutants, Y127F, R254F and D355R, that showed enhanced cyclization activity towards a starch substrate from 1.64 to 2.1-folds. Kinetic studies indicate that the mutants had higher affinity towards the substrate than the wild type β-CGTase. The Y127F mutant had the highest affinity which was indicated by the lowest K m of 15.30 mM and the highest catalytic activity. Increasing hydrophobicity around the catalytic center appeared to favor the cyclization activity of the mutants. The β-CGTase and the three mutants showed optimal enzyme activity at 60 °C and pH 6.0. All the enzymes were stable for at least 60 min across a wide pH range (5.0-7.0).


Introduction
Cyclodextrin glycosyltransferase (EC 2.4.1.19,CGTase) belongs to the a-amylase family of enzymes (Bautista et al. 2012;Janecek and Sevcik 1999;MacGregor et al. 2001), which catalyzes three type reactions towards starch, namely cyclization to yield cyclodextrins (CDs), hydrolysis, and disproportionation (van der Veen et al. 2000b, c). Catalysis of cyclization is the primary function of CGTase . The well-characterized forms of CDs are a-, band c-CD that consist of six, seven and eight-D glucose units, respectively (Goh et al. 2009). Among them, the b-CD harbors a hydrophobic internal cavity and a hydrophilic outer surface, which endows b-CD with the water-soluble capacity (Szejtli 1990).
These properties make b-CD highly attractive for diverse applications in the field of industry, cavitation, food, medicine, as well as cosmetic (Astray et al. 2010;Slominska et al. 2002;Stella and He 2008;Szente and Szejtli 2004). The widespread use of b-CD suggests the importance for controlling the activity of b-CGTase. Previous studies have shown that mutations in the amino residues of b-CGTase influence the enzyme activity of b-CGTase at different levels (Leemhuis et al. 2002a(Leemhuis et al. , 2003b(Leemhuis et al. , 2004aPenninga et al. 1995;van der Veen et al. 2000a). W652G amino acid substitution of b-CGTase in Bacillus firmus var. alkalophilus enhanced the cyclization activity and decreased the hydrolytic activity of b-CGTase (Hyun-Dong et al. 2000). Contrarily, mutant A230V CGTase in Bacillus circulans strain 251 improved the hydrolytic activity but reduced cyclization activity (Leemhuis et al. 2003a). Additionally, Dijkhuizen and collages uncovered that mutation of tyrosine 195, which located in the central active site cleft of CGTase, significantly suppressed the cyclodextrin formation (Penninga et al. 1995). The amino acid side chain at conserved Phe260 residues also regulated the hydrolytic activity of CGTase. Mutation of Phe260 converts CGTase from a transglycosylase into a starch hydrolase (Leemhuis et al. 2002a). Beyond these findings, the conserved residue Asp135 was reported to be required for the proper conformation of amino residues in the catalytic sites and activity (Leemhuis et al. 2003b). Results from the Leemhuis.H. demonstrated that substrate-binding site mutations of CGTase including Y167F, G179L, G180L, N193G and N193L exhibited decreased b-CD production, disproportionation as well as coupling reactions (Leemhuis et al. 2002b). All these findings suggested that amino acid mutations in the conserved and critical sites regulate the activity of CGTase. Therefore, it is desirable to prepare, characterize and study other novel mutants that may improve the activity of b-CGTase.
The structure of b-CGTase consists of five domains (Klein et al. 1992). Domains A and B are responsible for the catalytic activity of CGTase, while domains C and E mediate the binding to raw starch (Lawson et al. 1994;Penninga et al. 1996). The function of domain D remains to be illustrated. It was reported that Arg204 plus the three catalytic residues Asp206, Glu230 as well as Asp297 are totally conserved in the a-amylase family (Janecek 2002). Additionally, seven conserved sequence regions covering strand beta 2, beta 3, beta 4, beta 5, beta 7 and beta 8 of the catalytic (beta/alpha)(8)-barrel as well as the C-terminus of domain B have been identified (Janecek 2002). These conserved regions contain the catalytic, specificity and substrate-binding residues (Janecek 2002). The crystal structure of b-CGTase from Bacillus circulans strain 251 has been resolved (Lawson et al. 1994). The b-CGTase from Bacillus sp. N-227 in this study has identical amino acid sequence as the b-CGTase from Bacillus circulans. A Swiss Model structure based on a template (PDB ID: 1-pamB) is available. The three-dimensional structure of the b-CGTase revealed a 'pocket' or substrate-binding cleft which is the activity center of the enzyme. We report here three mutants of this b-CGTase, Y127F, R254F and D355R corresponding to variant positions located in Domain A, a part of the catalytic center, and their improved catalytic activity towards cyclization of starch.

Strain and culture condition
The wild type b-CGTase from Bacillus sp. N-227 (Gen-Bank accession DQ631916) and Escherichia coli strain BL21 (DE3) was provided by the Technology Center of Bright Dairy & Food Co. Ltd. (Shanghai, China). The b-CGTase expression plasmid was based on the pET-28b vector and also generated by the Technology Center (The b-CGTase bacterial expression plasmid (pET-28b-b-CGTase-6xHis) which expresses the b-CGTase-6xHis fusion protein was purchased from the Technology Center).

Culture conditions
The seed medium components, 1% tryptone, 1% NaCl, and 0.5% yeast extract, were dissolved in distilled water in a total volume of 1 L, and then sterilized by autoclaving at 118°C for 20 min.

Introduction of mutations by using overlap extension PCR
Overlap extension PCR is a well-established PCR amplification method that can be used to introduce and recover mutations in specific DNA sequences without the requirement for restriction enzyme sites (Angelaccio and di Patti 2002).

Primers
Based on the homologous nucleotide sequences of the wild type b-CGTase from Bacillus sp. N-227 (GenBank accession DQ631916), four pairs of primers were designed using Primer Premier 6.0 and Oligo 6.0 software. Three of the 4 pairs of primers had the desired substitution (Table 1).

Overlap extension PCR analysis
Overlap extension PCR was performed using the ABI Step One Plus system (Applied Biosystems) followed by melting curve analysis with the cycling programs described below. For the first round of PCR, pET-28b (pET-28b-b-CGTase-6xHis) was used as the template for production of ORF-1 and ORF-2. Initial denaturation was at 94°C for 5 min, followed by 35 cycles of 94°C for 45 s, 55°C for 45 s and 72°C for 90 s. A final extension of 72°C for 10 min was performed, followed by storage of products at 4°C. A mixture of ORF-1 and ORF-2 was used as the template to produce the mutated gene fragment during the second round of PCR. The procedure involved initial denaturation at 94°C for 5 min, followed by 35 cycles of denaturation at 94°C for 45 s, 55°C for 45 s, and 72°C for 10 min. Taq polymerase was then added into the system, and the reaction was continued at 72°C for a further 30 min. The samples were stored at 4°C.

In vitro protein synthesis
A gel extraction kit (TIGEN Biotech) was used to isolate the PCR fragments which were then ligated into the pMD18-T. Simple vector E. coli DH5a was transformed with these plasmids using the heat shock method. Bluewhite selection was conducted on flat LB plates containing ampicillin (100 mg/mL) to select recombinant colonies for expansion in liquid culture. Recovered plasmids with the correct restriction pattern were sent to Sangon Biotech for gene sequencing. The sequence analysis confirmed that cDNAs from pMD18-T were subcloned to pET-28b and introduced into E.coli BL21 (DE3) for protein expression.

Determination of a mutation point of the b-CGTase gene
The three bacterial strains expressing the b-CGTase mutants were named Y127F, R254F and D355R. Recombinant proteins were purified and subjected to western blotting using mouse anti-6-His monoclonal antibody. The medium containing the b-CGTase expressing strains was collected after shake cultivation for 8 h, and then 1% lactose was added overnight for induction of protein expression. After shake cultivation for an additional 8 h, a solution containing 0.5% Triton X-100 and 1% glycine was added, before centrifugation at 60009g, 4°C for 10 min. The supernatant was dialyzed against 4 L of a 20 mM phosphate buffer, pH 7.0 at 4°C for 24 h and then loaded onto a His-Tag affinity column (General Electrics) for purification. The expression was confirmed by SDS-PAGE and western blotting.

Enzyme activity assays
The enzyme activity assays were conducted with soluble corn starch as a substrate and measured spectroscopically. The reaction medium (0.01 mL of the enzyme solution, 0.2 mL of 0.2% starch solution, 0.2 mL of a 0.2 M glycine-NaOH buffer pH 9.0) was incubated in a water bath at 40°C for 10 min (Rimphanitchayakit et al. 2005). 100 lL of 4 mM iodine in 30 mM potassium iodide was added to the reaction and diluted to 10 mL with water. The starch iodine complex formation was quantified by measuring absorbance at 700 nm. One unit of enzyme activity was defined as the amount that elicited a 10% reduction in absorbance (Gawande et al. 2003). In the formula below, 'a' is the optical density (OD/min) of the reaction from the control group, and 'b' is the OD of the reaction from the sample.

Activity-pH profile
The optimum pH of the purified enzymes was determined by replacing 0.2 M glycine-NaOH buffer (pH 9.0) with either 0.2 M sodium acetate buffer (pH 4-5), 0.2 M sodium phosphate buffer (pH 6-7), or glycine-NaOH buffer (pH 8-10). Reactions were performed following the previous procedures described for the enzyme activity assay.

Activity-temperature profile
The optimum temperature of the purified enzymes was determined by incubating the reaction mixture at different temperatures, ranging from 30 to 80°C for 10 min. Reactions were performed following the previous procedures described for the enzyme activity assay.

pH stability profile
The pH stability of the enzymes was measured by incubating 0.1 mL pure enzyme with 0.2 mL of 0.2 M sodium acetate buffer (pH 4-5), 0.2 M sodium phosphate buffer (pH 6-7) or glycine-NaOH buffer (pH 8-10), at 60°C, and without substrate for 60 min. The CGTase assay procedure described above was followed to determine the residual activity of the enzyme.

Thermal stability profile
The thermal stability of the enzymes was measured by incubating 0.1 mL pure enzyme with 0.2 mL buffer (0.2 M sodium phosphate buffer, pH 6.0) without substrate at different temperatures (30-80°C) for 60 min. The CTTCATCGACAATCATcgcATGGAGCGTTTCCACAC GTGTGGAAACGCTCCATgcgATGATTGTCGATGAAG D355R standard CGTase activity assays described above was used to determine the residual activity of each enzyme (Jeang et al. 2005).

Kinetic assays
The K m and V max values for the purified enzymes were determined by incubating 100 lL of enzyme (0.5 lg) in 200 lL of 0.2 M phosphate buffer (pH 6.0) with soluble starch solution (0.4-6.0 mg/mL) at 60°C for 10 min. The kinetic parameters of K m and V max were obtained by a nonlinear least-square fitting procedure using the Michaelis-Menten equation and the curve fitting software (Origin 8.0).

Generation of the b-CGTase mutants
The overlap extension PCR was used for site-directed mutagenesis of the generation of the b-CGTase mutants, and three mutagenic primers were designed at three key positions of b-CGTase. As shown in Fig. 1a, the first-round PCR using primers F1 and R1 yielded a pattern containing 6 different bands following agarose gel electrophoresis. Lanes 1 and 2 were the two fragments of mutant gene of R254F, the fragments length was, respectively, 762 and 1377 bp; lanes 3 and 4 were the two fragments of mutant gene of D355R, the fragments length was, respectively, 1065 and 1074 bp; lanes 5 and 6 were the two fragments of mutant gene of Y127F, the fragments length was, respectively, 381 and 1758 bp. The second round extension PCR using primers F2 and R2 yielded complete mutant genes. Agarose electrophoresis analysis of the mutation genes showed that there were three specific bands of 2.1 kbp (from lanes 2 to 4), respectively. The size of the mutant genes was similar with genes of b-CGTase (lane 1) (Fig. 1b). We prepared the amino sequence alignment of b-CGTase from Bacillus sp. (142676), Bacillus circulans (39420) and Paenibacillus sp. Xw-6-66 (452182092). The location of each mutant within the primary amino acid sequence of the b-CGTase is shown in Fig. 2, where the 5 domains of b-CGTase and the mutation sites were highlighted.
We examined the gene products of the cloned b-CGTases fragment and its mutants by in vitro protein synthesis using an E. coli extract system. The recombinant expression plasmids were transformed into E. coli BL21 (DE3). Following induction of protein expression overnight using 1% lactose, we lysed the bacteria and obtained the mutant proteins, performed SDS-PAGE analysis of the proteins and the purifications (Fig. 3a-c). A band of about 70 kDa for the wild type b-CGTase (lane 1, Fig. 3a, b) and all mutant constructs (lanes 2-4, Fig. 3a, b) were observed. The novel combination of the modern molecular approaches used in this study for generation of b-CGTase mutants seems to be suitable for quick, reliable and simple. Mutant b-CGTase was constructed into the pET-28a vector that harbors His-tag. To detect the expression of mutant b-CGTase, Western blot analysis was performed with anti-His antibody. Western blotting for expression of the fusion His-tagged recombinant protein was shown in Fig. 3c. A band at about 70 kDa was observed for the wild type b-CGTase (lane 3) and the mutants (lanes 4-6), while the lane 1 and lane 2 were obtained with E. coli strain BL21 (DE3) and pET-28b.

Enzyme activity assays
We tested relative enzyme activity of the wild type b-CGTase and Y127F, R254F, and D355R mutant b-CGTases and the results are listed in Tables 2 and 3. The data showed that cyclization activity of the mutant b-CGTases was increased to 2130-1660 U/mL compared to 1012 U/mL for the wild type b-CGTase, indicating a 1.6 to 2.1-fold activity enhancement.
The kinetic parameters for the b-CGTase and its mutants Y127F, R254F, and D355R were measured using the starch substrate and they are tabulated in Table 3. The V max values in the range of 22.02 and 21.41 lg/min for the mutant proteins were higher than 21.00 lg/min for the parent wild type b-CGTase. Meanwhile, the observed k cat values varying between 73.40 and 71.37/s for the mutant b-CGTases were also higher than the number of 70.00/s for the wild type b-CGTase. The ratio of k cat /K m for Y127F, R254F, and D355R variants was 4.80, 4.04, and 4.10/s/ mM, respectively, compared to 3.65/s/mM for the wild type b-CGTase. The ratio suggested how efficiently a catalytic enzyme converts a substrate into products. A higher ratio indicates a higher efficiency of the enzyme.
The increased ratio of 1.15, 0.39, and 0.45 for the mutants compared to the wild type b-CGTase indicates enhanced catalytic activity for the mutants which was consistent with the enzyme activity assay.
The Michaelis constant K m is an indicator of the substrate's affinity for the enzyme. As shown in Table 3, the observed K m was lowered by 3.90, 1.40, and 1.80 mM for the Y127F, R254F, and D355R mutants to 15.30, 17.80, and 17.40 mM from 19.20 mM for the wild type b-CGTase. This observation indicates that all the three mutants had higher affinity towards the starch substrate than the parent b-CGTase.

Determination of pH optima using a starch substrate
The activity of the wild type and mutant forms of the b-CGTase as a function of pH was examined over a pH range of 4-8 in a universal buffer using the starch substrate (Fig. 4a). The activity-pH profiles for the Y127F, R254F, and D355R mutants were similar to that for the wild type b-CGTase. The highest activity was observed at pH 6.0 for all the four enzymes, indicating limited influence of the single site mutation on the activity-pH relationship. The activity of each enzyme without pre-incubation in the buffer at pH 6.0 was defined as 100%. All the enzymes, including the wild type, were relatively stable at pH 4-7 with residual activity of more than 90%. However, they all became unstable at higher pH as indicated by the residual activity dropping from 85% at pH 8.0 to less than 30% at pH 10. The stability as a function of pH for the wild type and mutant forms of the b-CGTase was tested between pH 4 and pH 10 in a universal buffer (Fig. 4b). It is important to note that all mutants yielded stability  1660.75 ± 1.720* 1.64 * The mean difference is significant at the P \ 0.01 level a The ratio of total activity to mutants and b-CGTase. Experiments were performed in triplicate curves which were very similar to that of the wild type enzyme (P [ 0.05).

Activity-temperature profile
The optimal temperatures and thermostability of mutants and were compared to b-CGTase (Fig. 5a, b). The effect of temperature on the b-CGTase activity was investigated using the starch substrate at temperatures between 30 and 80°C for 10 min at the optimal pH 6.0 for each enzyme (Fig. 5a). The optimal temperature for all the enzymes including the wild type b-CGTase was 60°C. The activity-temperature profiles for the wild type b-CGTase and the mutants were very similar. They were both optimally active at 30-60°C, at temperatures from 30 to 60°C, all the four b-CGTases retained at least 90% of the original activity for at least 1 h (Fig. 5b). Enzymatic activity gradually decreased when temperature exceeded 60°C for all the enzymes. The observed residual activity at 70°C was about 40%, and the activity decreased to 25% at 80°C. Thus, the mutation did not have significant negative effect on stability when compared to the wild type b-CGTase (P [ 0.05); the cyclization assays were performed as described (Penninga et al. 1995).

Discussion
b-CGTase is one of the important a-amylase family, and CGTase is the first enzyme of all the transglycosylases. In general, the cyclization activity of CGTases is much higher than the disproportionation and hydrolysis activities.
In this work, we engineered the mutants of b-CGTase by site-directed mutagenesis with the attempt to improve catalytic cyclization activity of b-CGTase. The three mutants (Y127F, R254F, and D355R) showed enhanced activity of b-CGTase. Interestingly, these mutant residues were all involved in the activity center of the enzyme, and they were located at the bottom of the 'pocket' in the N-terminal of the b-CGTase. The NH 2 -terminal region of CGTase was important for cyclization characteristics, so the mutations would improve the activity of b-CGTase (Penninga et al. 1995). Fujiwara et al. (1992a, b) constructed the CGTase gene from Bacilus stearothermophilus Each of these mutations was predicted to alter the hydrophobicity of a conserved domain within the b-CGTase by Support Vector Machine. Since single site mutations of the b-CGTase occurred at three different sites (127, 254, and 355) and their relative positions in the substrate-binding cleft (pocket) were expected to have different influence on the incoming starch substrate, it is difficult to discuss the activity-structure relationship. However, analysis of the observed cyclization activity of each of the three mutants and their wild type b-CGTase displayed a trend that increasing hydrophobicity of the catalytic center domain appeared to favor the cyclization activity of the mutants. The D355R mutant b-CGTase replacing a hydrophilic aspartic acid (D, hydropathy index = HI = -3.5) with a hydrophilic arginine (R, HI = -4.5) enhanced the enzyme activity to 1.6-folds. More significant enzyme activity enhancements (1.9 and 2.1-fold) were observed for the Y127F and R254F mutant b-CGTases where a hydrophilic tyrosine (Y, HI = -1.3) or arginine (R) was replaced with a hydrophobic phenylalanine (F, HI = 2.8), respectively. It was assumed that the cyclization was taken by nucleophilic attack, and the key amino acid residues in the cyclization reaction were aspartic acid 229 in the cyclization reaction (Wind et al. 1998). When aspartic acid 229 nucleophilic attack C1 as shown in Fig. 6, the carboxyl ionization, the arginine with positive charge took the place of the negatively charged aspartic acid 355, the existence of positive charge was more advantageous to ASP229 nucleophilic attack C1, and conducive to Asp229 nucleophilic attack C1, and conducive to the substrate cyclization (Chung et al. 1998;Kragh et al. 2010).
In mutant of R254F, phenylalanine with a pair of conjugate lone pair electrons was advantageous to the Fig. 6 Schematic representation of the reactions catalyzed by CGTase. After bond cleavage a covalently bound reaction enzyme glucosyl intermediate is formed. In the second step of the reaction the reaction intermediate is transferred to an acceptor molecule. In the cyclization reaction, the terminal OH -4 group of the covalently linked oligosaccharide is used as acceptor, whereas water or a second sugar is used as acceptors in the hydrolysis and disproportionation reactions, respectively. This figure has been adapted from Ref. Leemhuis et al. (2004a) hydrophobic stacking effect and conducive to the stability of the seventh sugar molecules cyclization. For the mutant b-CGTase-Y127F, the hydrophobic phenylalanine blocked the entering H 2 O into the active center, which restrained the hydrolysis and enhanced the cyclization. This result is similar with previous report (Leemhuis et al. 2003c).
The mutant proteins were successfully expressed and purified. The molecular weight of the purified mutant proteins was about 70 kDa, which was consistent with the expected values as reported by Nomoto et al. (1986). These enzymes present common properties with the majority of the reported CGTases which are monomeric with a molecular mass ranging from 60 to 110 kDa (Biwer et al. 2002;Nomoto et al. 1986).
The optimum pH of the mutants was pH 6.0, which was in accord with some other CGTases from alkalophilic Bacillus sp. G1 (Cao et al. 2005) and B. stearothermophilus ET1 (Sian et al. 2005). However, the mutants showed lower activity at pH 4.0 and 8.0, suggested that the mutants and wild type of b-CGTase required a near-neutral pH range to perform its reaction. Extreme pH values were not suitable for the enzyme to carry out cyclization activity. Most of the reported CGTases exhibited optimum pH range from 5.0 to 8.0 (Bovetto et al. 1992;Chung et al. 1998;Tachibana et al. 1999;Sian et al. 2005), but the enzyme from Brevi-bacterium sp. no. 9605 possessed a higher optimum pH value at 10.0 (Cao et al. 2005).
The pH stability for cyclization is broad, in the range of pH 4.5-9.0, indicating that the mutants have not changed the pH-dependent activity of b-CGTase. A similar observation implying that the mutants of b-CGTase were related to pH stability has been also reported by Kimura et al. who constructed mutants by deleting the C-terminus of the CGTase from the alkalophilic Bacillus sp. #1011 (Mori et al. 1994).
The optimal temperature for all the enzymes including the wild type b-CGTase was 60°C, and they were both optimally active at 30-60°C. The activity-temperature profiles for the wild type b-CGTase and the mutants were very similar. Studies by other researchers on CGTase from B. autolyticus 11149, B. stearothermophilus (Tomita et al. 1993) andB. circulans E 192 (Chung et al. 1998) also found 60°C was the optimum temperature, which is in agreement with mutants. And the enzymes had not a higher temperature stability compared to CGTase from alkalophilic Bacillus sp. G1 (Cao et al. 2005).
In vitro site-directed mutagenesis was performed to introduce Y127, R254 and D355 mutations in a conserved region of the b-CGTase enzyme. The mutant proteins were successfully expressed and purified. The Y127F, R254F, and D355R mutant b-CGTases exhibited a similar optimal temperature, optimal pH, thermal stability, and pH stability as the wild type b-CGTase. However, all the three mutants displayed a higher activity towards the corn starch substrate than the wild type b-CGTase. Increasing hydrophobicity around the substrate-binding cleft of the catalytic center appeared to enhance the cyclization activity of the enzymes.