Study on Transglucosylation Properties of Amylosucrase from Xanthomonas campestris pv. Campestris and Its Application in the Production of α-Arbutin

α-Arbutin (4-hydroquinone-α-D-glucopyranoside), an effective skin-lightening agent due to its considerable inhibitory effect on human tyrosinase activity, is widely used in the pharmaceutical and cosmetic industries. Recently, α-arbutin was prepared through transglucosylation of hydroquinone using microbial glycosyltransferases as catalysts. However, the low yield and prolonged reaction time of the biotransformation process of α-arbutin production limited its industrial application. In this work, an amylosucrase (ASase) from Xanthomonas campestris pv. campestris str. ATCC 33913 (XcAS) was expressed efficiently in Escherichia coli JM109. The catalytic property of the purified XcAS for the synthesis of α-arbutin was tested. The recombinant strain was applied for highly efficient synthesis of α-arbutin using sucrose and hydroquinone as glucosyl donor and acceptor, respectively. By optimizing the biotransformation conditions and applying a fed-batch strategy, the final production yield and conversion rate of α-arbutin reached 60.9 g/L and 95.5%, respectively, which is the highest reported yield by engineered strains. Compared to the highest reported value (<1.4 g/L/h), our productivity (7.6 g/L/h) was improved more than five-fold. This work represents an efficient and rapid method for α-arbutin production with potential industrial applications.


Introduction
α-Arbutin, a glycosylated hydroquinone, is a prominent natural compound used as a skin-whitening agent due to its antimelanogenesis effect without any cytotoxicity [1].It has been reported that the inhibitory function of α-arbutin against tyrosinase is 10 times greater than that of its isomer, β-arbutin [2].Specifically, α-arbutin only inhibits mammalian tyrosinase, while β-arbutin reduces tyrosinase activity by both melanoma and somatic cells [3].Moreover, a study on cultured human melanoma cells and three-dimensional skin models indicates that α-arbutin does not inhibit cell viability and exhibits outstanding performance in reducing melanin synthesis [2].α-Arbutin also shows antioxidative, antimicrobial, and anti-inflammatory functions.Therefore, the use of α-arbutin as a safe and effective skin-whitening agent has received increasing attention in recent years.
Amylosucrase (ASase), belonging to the glycoside hydrolase 13 family, is a versatile enzyme catalyzing multiple reactions [17].With sucrose and extra glucosyl acceptors as substrates, ASase has glucosyltransferase activity and can transfer glucose molecules from sucrose to glucosyl acceptors with α-1,4 glycoside linkages only [18,19].This enzyme utilizes the energy derived from the cleavage of glucosidic bonds in sucrose to synthesize other glycosides without involving any primer or nucleotide-activated sugars [20].Recently, the ASase gene has been cloned from Deinococcus geothermalis and Cellulomonas carboniz for in vitro biosynthesis of α-arbutin by using sucrose as a glucosyl donor and hydroquinone as acceptor [11,12].In these studies, the high conversion rate (90%) and low reaction time (2 h) revealed that amylosucrase has remarkably higher transglucosylation capacity, showing good potential for commercial production of α-arbutin.As compared to the other transglucosidases, amylosucrase demonstrates higher efficiency in the synthesis of α-arbutin.However, to date, its maximum production yields are still low (5.8 g/L and 0.6 g/L for ASase from Deinococcus geothermalis and Cellulomonas carboniz, respectively).Thus, it is necessary to establish a simple and cost-effective method to produce α-arbutin.
In this study, we first employed an ASase from X. campestris pv.campestris (XcAS) as the catalyst to synthesize α-arbutin, and the catalytic property of the purified XcAS was clarified.In addition, α-arbutin was synthesized by a whole-cell recombinant E. coli JM109/pet28a-xcas biocatalyst system using sucrose as glucoside donor and HQ as acceptor.Furthermore, a fed-batch strategy was employed to increase the production yield of α-arbutin.Thus, this work represents a highly efficient method for α-arbutin production with potential industrial applications.

XcAS Heterologous Expression and Protein Purification
The XcAS encoding gene, locus XCC3359 from the X. campestris pv.campestris str.ATCC 33913 genome, consists of 1914 base pairs encoding 637 amino acid proteins.The target gene was synthesized and inserted into the pET28a (+) vector with 6 × histidine tagged at its carboxyl terminus, resulting in expression of the pET28a-xcas plasmid.Through regulated expression of the xcas gene under the control of a T7 promoter and lac repressor, it was transformed into E. coli JM109 [DE3].As shown in Supplementary Figure S1A, the initial expression result of XcAS was almost manifested as inclusion bodies (IBs).To improve the production of soluble amylosucrase, the process conditions were optimized.As shown in Supplementary Figure S2, the results suggest that 0.1 mmol/L isopropyl β-D-1-thiogalactopyranoside (IPTG) and 25 • C is the optimal inducement condition for XcAS production, and almost all of the target proteins were expressed in soluble form (Supplementary Figure S1B).The soluble fraction of the expressed protein was purified as described in Materials and Methods and analyzed by SDS-PAGE.As shown in Figure 1, the purified protein had a molecular mass of approximately 66.2 kDa.

Bioconversion of HQ to α-Arbutin by Purified XcAS
The transglucosylation reaction was performed using sucrose as a glucosyl donor and 10 mM hydroquinone as an acceptor in 50 mM phosphate buffer (Supplementary Figure S3A).It is known that benzoquinone, produced by the oxidation of hydroquinone, can inhibit the enzyme activity by chemically modifying and reducing its stability [9].Thus, ascorbic acid (VC) was added to the reaction system to inhibit the oxidation of hydroquinone and improve the bioconversion rate of α-arbutin.After the reaction, the sample was analyzed by HPLC.As shown in Supplementary Figure S3B, there were two peaks at retention times (tR) of 4.5 min and 5.8 min, corresponding to α-arbutin and HQ, respectively.

Catalysis Characterization of Purified XcAS on Transglucoside Reaction
The influence of pH and temperature on the transglucoside activity of recombinant XcAS was determined by using sucrose as a glycosyl donor and hydroquinone as an acceptor.As shown in Figure 2A, the enzyme had optimal activity between pH values of 6.0 and 7.5, with specific activity of 0.9 U/mg.Moreover, in a pH stability assay (Figure 2B), XcAS transglucoside activity showed a slight reduction in pH in the range of 5.5-8.0 after being incubated in different pH at 4 • C. As shown in Figure 2C, the maximal transglucoside activity of the recombinant XcAS was at 30 • C. Furthermore, thermal stability analysis indicated that the enzyme was stable from 20 to 30 • C.However, the activity of XcAS decreased sharply when the temperature increased above 35 • C (Figure 2D).The kinetics of the XcAS-catalyzed transglucoside reaction was tested at the optimal pH and temperature using sucrose as a glucosyl donor and hydroquinone as an acceptor.As shown in Figure 3, the K m and V max values of the XcAS-catalyzed reaction were 11.29 ± 0.43 mM and 2.49 ± 0.04 U/mg, respectively.

Effects of pH and Temperature on α-Arbutin Production by Whole-Cell Biocatalysis
Bioconversion of α-arbutin from HQ by whole-cell biocatalysis is a favorable approach with many advantages [7,21].Thus far, α-arbutin production yield is limited by the low level of enzyme activity, and research on high substrate concentration for α-arbutin production is still lacking.In the present study, E. coli JM109/pet28a-xcas with a high activity of amylosucrase, as described above, was used for the bioconversion of HQ to α-arbutin.Furthermore, the biotransformation conditions were optimized.
To optimize the efficiency of whole-cell biocatalysis on α-arbutin production, the effects of pH and temperature were investigated.As illustrated in Figure 4A, a reaction mixture containing 20 g/L wet cell weight (WCW) of the culture, 167 mM of HQ, and 1 M of sucrose in 10 mL of phosphate buffer (50 mM, pH 7.0) was incubated for 12 h under a broad range of temperatures (20 to 40 • C).The results showed that the highest conversion rate of HQ was 80.4% at 30 • C, which can be used as the optimal temperature for whole-cell biocatalysis.As the temperature increased from 30 to 40 • C, the conversion rate reduced dramatically, suggesting that XcAS activity was low and unstable at temperatures higher than 30 • C. Additionally, the effect of pH (range 4-10) was examined on whole-cell biocatalysis.All of the experiments were carried out at 30 • C. At pH 7.0, the maximum conversion rate was obtained, while the production of α-arbutin sharply decreased with a slight divergence from the optimal pH (Figure 4B).Thus, the optimum pH value of 7.0 and temperature of 30 • C were chosen for further experiments.
To improve α-arbutin production, the effect of TritonX-100 concentration was further investigated.The results demonstrated that the conversion rate of HQ increased, attaining the highest value of 0.4% (w/v) with TritonX-100 induction (Figure 5B).However, the conversion rate decreased rapidly when a higher concentration of surfactant was added.Therefore, the optimal TritonX-100 concentration of 0.4% (w/v) was used for α-arbutin production, with a high conversion rate of 94.5%.

Effects of Substrate Ratio and Concentration on α-Arbutin Production
Enzymatic synthesis of α-arbutin from sucrose and HQ by E. coli JM109/pet28a-xcas is a one-step stereospecific reaction and its theoretical molar ratio is 1:1 [11].The effect of the molar ratio of sucrose to HQ on α-arbutin production was investigated with 165 mM HQ in this study.When both substrates were in the same ratio (1:1), the conversion rate was only 49.7% (Figure 5C).By increasing the sucrose:HQ ratio to 4:1, the conversion rate increased to 93.5%.However, with further increases of the molar ratio to 5:1 and 6:1, the conversion rates remained approximately the same, indicating that the amylosucrase in E. coli JM109/pet28a-xcas was probably saturated with sucrose.
HQ can induce apoptosis in vivo through a change in cellular redox status, especially by reducing the endogenous thiol level and increasing the reactive oxygen species level [7,23].If the concentration of HQ exceeds a certain value, the production of α-arbutin may drop sharply due to cell apoptosis.In this study, a series of HQ concentrations were tested and the time courses for these reactions were evaluated at the substrate ratio of 4:1.As shown in Figure 5D, 125 mM of HQ was completely converted within 2 h and the production of α-arbutin attained 34.1 g/L.Notably, the molar production yields of α-arbutin gradually decreased and the reaction time was extended with increasing concentration of HQ.In particular, the conversion rate decreased to 68% within 6 h after 208 mM of HQ was added.These results clearly demonstrate that the maximum HQ tolerance of E. coli JM109/pet28a-xcas was 208 mM.Furthermore, all the conversion rates retained their maximum values within 12 h.

Efficient Production of α-Arbutin through a Batch and Fed-Batch Strategy
Based on the aforementioned optimal conditions, a total yield of 33.8 g/L α-arbutin was obtained from 13.7 g/L of HQ within 2 h using the whole-cell E. coli JM109/pet28a-xcas biocatalyst system (Figure 6A).The productivity and conversion yield were 16.9 g/L/h and 99.4%, respectively.No byproduct was detected during the bioconversion process, indicating that XcAS displayed region-selective activity.A fed-batch strategy can improve the final concentration of the desired products [24].A gradient decreasing batch was used to attain higher α-arbutin production for the sake of cell apoptosis induced by high HQ concentration.The initial concentration of HQ was 13.8 g/L (125 mM), and decreasing amounts of HQ (i.e., 62.5 mM, 31.25 mM, and 15.62 mM) and sucrose (molar ratio of HQ:sucrose was 1:4) were added to the conversion system every 2 h.After 8 h of conversion, the final α-arbutin yield reached 60.9 g/L, 180% higher than before (Figure 6B).More importantly, the final molar yield of α-arbutin reached 95.5% with a productivity of 7.6 g/L/h, which was the highest compared to other bioconversion methods (Table 1).Biocatalysis is the transformation carried out using whole cells; enzymatic catalysis is the transformation carried out by purified enzyme; fermentation is the transformation carried out with fermentative strains.

Discussion
Amylosucrase is a multifunctional enzyme from the glycoside hydrolase family 13 that has powerful transglucosylation activity to generate α-1,4-linked glucose-transfer products from sucrose as glucoside donor [17].To date, ASase from Deinococcus geothermalis (DG-ASase) [11,25] and Cellulomonas carbonis (CC-ASase) [12] has been reported to synthesize α-arbutin via the transglucosylation of inexpensive sucrose as glucosyl donor and HQ as glucosyl receptor.Compared with other glycosyltransferases, amylosucrase does not require the addition of expensive activated sugars and exhibits predominant biotransformation ability and higher efficiency in the glycosylation of HQ (Table 1).In this study, we first employed an ASase from X. campestris pv.campestris (XcAS) as the catalyst to synthesize α-arbutin.The commercial expression vector pet28a+, containing an induced expression system, was used for heterologous expression of XcAS in E. coli JM109 (DE3).The transcriptional expression was controlled by the lac repressor, thus required IPTG as an inducer.The initial recombinant protein was mainly found in IBs (Supplementary Figure S1A) and displayed low enzymatic activity.Previous studies have reported that the heterologous protein, with a high molecular weight and complicated structure, further impedes the capacity of host cells to fold the protein intermediates, and results in the formation of IBs [26].As a consequence, we optimized the temperature and IPTG concentrations for amylosucrase expression.The highest intracellular hydrolysis activity of 5.0 U/mg was obtained at a low temperature (25 • C) and IPTG concentration of 0.1 mM, and almost no IBs were seen in the expression results (Supplementary Figures S1B and S2).These phenomena can be explained by the reduced IB formation due to the proper folding of precursor proteins under low temperature and low inducer concentration [27,28].
The catalytic property of XcAS in the transglucoside reaction was studied by using sucrose as the glucoside donor and HQ as the receptor.The optimal pH of XcAS transglucosylation activity was in the range of 6.0-7.5 (Figure 2A) and the enzyme maintained stability between pH 5.5 and 8.0 (Figure 2B).It is assumed that the optimum working condition of this enzyme is neutral, which is similar to that of ASase from Deinococcus geothermalis (pH 7.0) [11], Cellulomonas carboni (pH 7.0) [12], and Truepera radiovictrix (pH 7.5) [29].The maximal transglucosylation activity was obtained at 30 • C (Figure 2C) and dropped rapidly when the temperature was higher than 35 • C (Figure 2D).It appeared that the enzyme obtained in this study showed low thermal stability, which is a little different from CC-ASase (40 • C) and DG-ASase (45 • C).Furthermore, the apparent K m and V max values of the XcAS-catalyzed transglucoside reaction using HQ as acceptor were 11.29 ± 0.43 mM and 2.49 ± 0.04 U/mg, respectively (Figure 3).This suggests that XcAS has potential application in α-arbutin biosynthesis.
Recently, whole-cell biocatalyst has been widely used in the industrial production of glycosides, due to its efficient biotransformation process and simplified downstream purification procedure [30].In our work, the resting cells of recombinant E. coli JM109/pet28a-xcas were prepared as a whole-cell biocatalyst in order to synthesize α-arbutin from HQ.By optimizing the biocatalysis conditions with the addition of TritonX-100 surfactant and using an HQ concentration of 125 mM, the production yield, conversion rate, and productivity of α-arbutin were found to be 33.8 g/L, 99.4%, and 16.9 g/L/h, respectively.Compared with other surfactants, TritonX-100 can obviously increase the conversion rate of α-arbutin, while SDS and CTAB have an inhibitory effect on whole-cell biocatalyst (Figure 5A).It appeared that much of this loss of amylosucrase enzymatic activity caused by ionic surfactants was due to the destruction of XcAS, while nonionic surfactants stimulated the reaction by increasing cell membrane permeability without affecting enzyme activity.In addition, the biocatalytic ability of E. coli JM109/pet28a-xcas cells was inhibited by a high HQ concentration (208 mM), which is consistent with previous studies [6,7,11,12].A high concentration of HQ may induce cell apoptosis by changing intracellular redox status through decreasing the thiol level and increasing the reactive oxygen species level [31,32].To prevent the toxicity of high HQ concentration while retaining α-arbutin production yield, a gradient decreasing fed-batch strategy was applied.As a result, the final production yield and molar conversion yield of α-arbutin reached 60.9 g/L and 95.5%, respectively, with an HQ concentration of 234.4 mM (Figure 6B).Furthermore, the highest productivity of α-arbutin (7.6 g/L/h) was obtained, which was five-fold higher than the previously reported value (<1.4 g/L/h).
Overall, our findings suggest that recombinant E. coli JM109/pet28a-xcas may be used as whole-cell biocatalyst system to produce α-arbutin from HQ. Furthermore, a high conversion rate in this system can simplify the downstream processing of α-arbutin purification.Hence, the proposed method may have potential applications for α-arbutin synthesis in the pharmaceutical and cosmetic industries.

Protein Purification
To obtain the purified XcAS, the supernatant was filtered through a 0.45 µm filter and loaded into an Ni 2+ -chelating Sepharose Fast Flow column (Uppsala, Sweden).The column was balanced by 50 mM sodium phosphate buffer (pH 7.0) containing 500 mM NaCl and washed with the same buffer containing 50 mM imidazole after protein loading.The bound protein was eluted with 50 mM sodium phosphate buffer (pH 7.0) containing 500 mM imidazole and 500 mM NaCl.The active fractions were collected and concentrated using Amicon Ultra-15 (Millipore) with a pore size of 30 kDa.All of the purification processes were performed at 4 • C [33].

Enzymatic Activity Assay and SDS-PAGE
Determination of ASase hydrolysis activity was carried out on the basis of the hydrolysis reaction of sucrose.The fructose and glucose concentrations were determined by the 3,5-dinitrosalicylic (DNS) acid method [20].The reaction was initiated by adding 50 µL of the enzyme solution in a reaction mixture with 100 µL of 200 mM sucrose, then incubated for 10 min at 37 • C. Subsequently, the reaction was terminated by adding 200 µL DNS solution and then incubated at 100 • C for 5 min.The resultant mixture was measured at 545 nm to determine the total reducing sugar, and fructose was used as a standard.One unit of enzymatic hydrolysis activity was defined as the amount of enzyme required to hydrolyze 1 µmol of sucrose per minute in standard assay conditions.
ASase transglucosylation activity was assayed using sucrose as glucosyl donor and HQ as acceptor.The 1 mL mixture containing 70 µg/mL XcAS, 50 mM sucrose, 1 mM VC, and 10 mM HQ in 50 mM phosphate buffer was incubated at 30 • C for 2h [12].One unit of transglucosylation activity was defined as the amount of enzyme required to produce 1 µmol of α-arbutin per minute in the assay conditions.Meanwhile, the soluble expression of ASase in E. coli JM109*[DE3] cells was analyzed with SDS-PAGE [24].

Enzyme Properties of Purified XcAS on Transglucosylation
The optimal pH of XcAS was determined in citrate/phosphate buffer from 4.0 to 9.0 at 30 • C. To measure the pH stability, the purified XcAS was preincubated at various levels (pH 4.5-9.0)for 24 h at 4 • C before enzyme activity was measured.The optimal temperature was measured from 20 to 50 • C at pH 7.0.To determine the thermostability of the purified XcAS, the enzyme was incubated at different temperatures for 1 h at pH 7.0 buffer [34].
Kinetic assays of XcAS transglucosylation were performed under the condition of optimal temperature and pH with sucrose (25-400 mM) and HQ (5-80 mM) as substrates.The kinetic parameters, K m and V max , were determined by the Lineweaver-Burk plotting method [33].The equation was as follows: 1 where V indicates the reaction rate, K m is the Michaelis-Menten constant, V max is the maximum reaction rate, and [S] is the substrate concentration.

Bioconversion of HQ to α-Arbutin by Whole-Cell Biocatalysis
Recombinant E. coli JM109*[DE3]/pET28a-xcas strain was grown in 50 mL TB medium at different temperatures, and induced with IPTG for 12 h.The cells were harvested by centrifugation at 7000 rpm for 20 min, and washed twice with 50 mM PBS (pH 7.0).Then, the cells were resuspended with 10 mL of 50 mM PBS (pH 7.0) containing HQ (glucosyl acceptor) and sucrose (glucosyl donor).To increase production yield, ascorbic acid was added to the reaction mixture at a final concentration of 1 mM to prevent oxidation of hydroquinone [11].Whole-cell biocatalysis was performed on a rotary shaker (200 rpm) at 30 • C, unless otherwise mentioned.The biocatalysis product was harvested by centrifugation for 10 min at 12,000 rpm, followed by HPLC analysis.

Analytical Methods
The amount of α-arbutin was determined by HPLC on an Eclipse XDB-C18 column (4.6 × 250 mm; 45 • C) connected to an SPD-10A UV detector (280 nm) and SPD-LC10 pump.The separation of α-arbutin was carried out using methanol and water (9:1 v/v) as the mobile phase at a flow rate of 0.8 mL min −1 with a column temperature of 45 • C [11].The conversion rate of α-arbutin is given by: where M 1 is the mole of HQ before transformation and M 2 is the mole of α-arbutin after transformation.

Figure 2 .
Figure 2. Effects of pH and temperature on purified XcAS transglucoside activity and stability.(A) Specific activity of XcAS at different external pH.Assays were performed at 30 • C in 50 mM citrate buffer (pH 4.0-6.0)and 50 mM sodium phosphate buffer (pH 6.0-9.0).(B) pH stability of XcAS.Purified XcAS was preincubated at pH varying from 4.5 to 9.0 for 24 h at 4 • C before enzyme activity was measured.(C) Specific activity of XcAS at different temperatures.Tests were performed at pH 7.0 in 50 mM sodium phosphate buffer.(D) Enzyme stability of XcAS at different temperatures.Assays were performed after preincubating XcAS for 1 h at different temperatures.

Figure 3 .
Figure 3. Determination of kinetic parameters of XcAS-catalyzed transglucoside reaction: (A) Michaelis-Menten plot and (B) double-reciprocal plot.Reactions were carried out at pH 7.0 and 30 • C by adding various amount of sucrose and hydroquinone (HQ) (ratio 10:1) to 50 mM sodium phosphate buffer.

Figure 4 .
Figure 4. Effects of (A) temperature and (B) pH on α-arbutin conversion rate.All assays were performed in triplicate in three independent experiments.Standard deviations of biological replicates are represented by error bars.

Figure 6 .
Figure 6.Time course for (A) batch and (B) fed-batch bioconversion of HQ to α-arbutin.Filled triangles represent concentration of α-arbutin; filled circles represent concentration of HQ.

Supplementary Materials:
The following are available online at http://www.mdpi.com/2073-4344/9/1/5/s1,FigureS1: SDS-PAGE analysis of XcAS expression in E. coli JM109[DE3] (A) before and (B) after process condition optimization, Figure S2: Effects of (A) temperature and (B) IPTG concentration on E. coli cell growth and XcAS hydrolysis activity in E. coli JM109/pet28a-xcas, Figure S3: Structural determination of the transglucosylated product.Author Contributions: Conceptualization, J.S. and B.Z.; Data curation, C.Y. and R.Z.; Formal analysis, C.Y. and B.Z.; Funding acquisition, B.Z.; Investigation, C.Y. and W.F.; Methodology, C.Y. and R.Z.; Project administration, W.F. and J.S.; Writing-original draft, C.Y.; Writing-review and editing, Z.K.-J.and B.Z. Funding: This research was funded by the grants from the State Key Project of Research and Development Plan (grant number: 2017YFE0112700).This work was partly supported by the National Natural Science Foundation of China (grant number: 41106125).

Table 1 .
Comparison of different biocatalysts or fermentative strains for α-arbutin production.