Pg

JCa

b

c

Nd

a

ARRAA

KBP�L

1

naccdhai

pbp

RLG

(

h0

Enzyme and Microbial Technology 67 (2014) 59–66

Contents lists available at ScienceDirect

Enzyme and Microbial Technology

j o ur na l ho mepage: www.elsev ier .com/ locate /emt

urification and functional characterization of the first stilbenelucoside-specific �-glucosidase isolated from Lactobacillus kimchi

in-A. Koa,1, J.Y. Parka,1, H.J. Kwona, Y.B. Ryua, H.J. Jeonga, S.J. Parka, C.Y. Kima, H.M. Oha,.S. Parka, Y.H. Limb, D. Kimc, M.C. Rhoa, W.S. Leea,∗, Y.M. Kima,d,∗

Eco-Friendly Bio-material Research Center, Korea Research Institute of Bioscience and Biotechnology, Jeongeup 580-185, KoreaInstitute of Health Science, Department of Biomedical Sciences, College of Health Sciences, Korea University, Seoul 136-703, South KoreaThe Institute of Food Research and Industrialization, Institute of Green Bio Science & Technology and Department of Agricultural Biotechnology, Seoulational University, Pyeongchang-gun, Gangwon-do 232-916, KoreaDepartment of Food Science & Technology and Functional Food Research Center, Chonnam National University, Gwangju 500-757, Korea

r t i c l e i n f o

rticle history:eceived 14 April 2014eceived in revised form 4 August 2014ccepted 3 September 2014vailable online 16 September 2014

a b s t r a c t

This study aimed to develop viable enzymes for bioconversion of resveratrol-glucoside into resvera-trol. Out of 13 bacterial strains tested, Lactobacillus kimchi JB301 could completely convert polydatininto resveratrol. The purified enzyme had an optimum temperature of 30–40 ◦C and optimum pH ofpH 5.0 against polydatin. This enzyme showed high substrate specificities towards different substratesin the following order: isorhaponticin » polydatin » mulberroside A > oxyresveratrol-3-O-glucoside. Addi-

eywords:ioconversionolydatin ;·Resveratrol-Glucosidaseactobacillus kimchi

tionally, it rarely hydrolyzed astringin and desoxyrhaponticin. Based on these catalytic specificities, wesuggest this enzyme be named stilbene glucoside-specific �-glucosidase. Furthermore, polydatin extractsfrom Polygonum cuspidatum were successfully converted to resveratrol with a high yield (of over 99%).Stilbene glucoside-specific �-glucosidase is the first enzyme isolated from lactic acid bacteria capable ofbio-converting various stilbene glucosides into stilbene.

© 2014 Elsevier Inc. All rights reserved.

. Introduction

Resveratrol (3,4′,5-trihydroxystilbene) is a natural polyphe-olic compound found in grapes, peanuts, a variety of mulberries,nd other plants, especially in the medicinal plants Polygonumuspidatum (Polygonaceae) and Rheum undulatum (Rheum). P.uspidatum has the highest content of resveratrol and poly-atin (resveratrol-3-O-�-glucoside), whereas R. undulatum has the

ighest amount of methoxy resveratrol derivatives [1]. Resver-trol is known to possess many pharmacological propertiesncluding anti-inflammatory, cardio protective, antioxidant, and

Abbreviations: HPLC, high-performance liquid chromatography; HPLC-MS, high-erformance liquid chromatography-mass spectrometry; NAB, sodium acetateuffer; pNPG-�, para-nitrophenyl �-glucoside; SDS-PAGE, sodium dodecyl sulfateolyacrylamide gel electrophoresis; TLC, thin-layer chromatography.∗ Corresponding authors at: Eco-Friendly Material Research Center, Koreaesearch Institute of Bioscience and Biotechnology, Jeongeup 580-185, Korea (W.S.ee)/Department of Food Science and Technology, Chonnam National University.wangju 500-757, Korea (Y.M. Kim). Tel.: +82 63 570 5174; fax: +82 63 570 5239.

E-mail addresses: wslee@kribb.re.kr (W.S. Lee), u9897854@chonnam.ac.krY.M. Kim).

1 They contributed equally to this work.

ttp://dx.doi.org/10.1016/j.enzmictec.2014.09.001141-0229/© 2014 Elsevier Inc. All rights reserved.

anticancer properties [2,3]. Recently, it is gaining scientific atten-tion as a longevity promoter. These valuable properties makeresveratrol useful for the production of pharmaceuticals, cosmet-ics, and nutraceuticals as a potential health-functional compound.Therefore, there is increasing need to develop better and moreeffective production methods.

Until now, many methods have been designed to chemicallysynthesize resveratrol, but the production costs are high, and theprocesses are environmentally unfriendly [4]. In order to produceresveratrol at reasonable prices, the bioconversion of polydatinto resveratrol has been found to be excellent method. Previously,various methods to convert polydatin to resveratrol have beeninvestigated, including acid hydrolysis, heating and enzymatictransformation techniques [5–7]. However, all these methods aretime-consuming and uneconomical. Alternative approaches likebiotechnological methods can be used to produce resveratrol fromplants and yeast cells. Although resveratrol has been produced suc-cessfully from metabolically engineered Saccharomyces cerevisiaeusing tyrosine [8] and p-coumaric acid [9], production at a com-

mercial scale (gram level per liter) was not achieved [8,9]. Forthese reasons, presently, resveratrol is mainly obtained by extrac-tion from plant tissues. In P. cuspidatum grown in Hanzhong, China,the content of polydatin is usually six-fold higher than that of

6 crobia

rovdcbowtrthdttgtpc

wtmspsifPw

2

2

UKr(o

2

tbt2swodpoctf

2

wDuwtavtwlEm

0 J.-A. Ko et al. / Enzyme and Mi

esveratrol [10] but the resveratrol was higher than its glucosidesn the aspect of the industrial demand [11]. If polydatin was con-erted into resveratrol, the yield of resveratrol would increaseramatically. Therefore, bioconversion of polydatin from P.uspidatum was carried out by enzymatic catalysis or micro-ial fermentation. For example, �-glucosidases from Aspergillusryzae [1,12], Aspergillus niger, yeast [13], and snailase [14]ere used to transform polydatin in the P. cuspidatum extract

o resveratrol. Enzymes were also added into the powderedaw material to improve the yield of resveratrol. Althoughhese methods reportedly increase the yield of resveratrol, theigh cost of the enzymes and production of toxic compoundsuring spontaneous fermentation limit their industrial applica-ions. To help the industrial production of resveratrol, it needso be enzyme acting towards stilbene glucoside or methoxylucoside derivatives found in P. cuspidatum and R. undula-um. Due to these reasons, we tried to screen for only thoseolydatin-bioconversion bacterial strains that have food safetylearance.

In this study, a novel polydatin-converting lactic acid bacteriumas selected to produce resveratrol from polydatin and extracts of

he plant P. cuspidatum. Considering the food safety aspect of theicroorganism, screening of enzymes from 13 lactic acid bacteria

trains was performed, and the ability of these enzymes to convertolydatin to resveratrol was also evaluated. Further, purification,ubstrate specificities, and kinetic parameters of the enzymes werenvestigated in detail. The bioconversion process would provide aeasible strategy for the production of resveratrol from plant like. cuspidatum, which has high polydatin content, or R. undulatum,hich has high methoxy resveratrol content.

. Materials and methods

.1. Chemicals and materials

Polydatin and resveratrol were purchased from Sigma-Aldrich (St Louis, MO,SA). The rhizomes of P. cuspidatum were purchased at an herbal market in Daejeon,orea and its dried rhizomes (0.24 kg) were extracted with methanol (1 L) for 7 d atoom temperature. The methanol extract was evaporated in vacuo, yielding a residue45 g), which was re-suspended in distilled water and used for further studies. Allther chemicals were of reagent grade.

.2. Isolation of strains and culture conditions

Thirteen kimchi, Korean fermented food, samples were collected from tradi-ional markets in Jeonju. They were thoroughly resuspended in a 50 mM phosphateuffer (pH 7.0) and spread on modified MRS agar plates (modified MRS agar con-ained 18 g of nutrient broth, 5 g of sodium acetate trihydrate, 4 g of yeast extract,

g of di-potassium phosphate, 1 g of sodium citrate dehydrate, 0.2 g of magnesiumulfate, 0.05 g of manganese (II) sulfate, and 15 g of agar powder per liter). The platesere aerobically incubated at 37 ◦C for 2 weeks. Single colonies were first transferred

nto new modified MRS agar plates. To screen for strains capable of converting poly-atin into resveratrol, isolated colonies were tested for their ability to hydrolyzeolydatin. These bacterial strains were incubated in MRS broth containing 1 mg/mLf polydatin without glucose at 37 ◦C for 6 days. At 24 h intervals, 50 mL aliquots wereollected and analyzed by thin-layer chromatography (TLC). Strain JB301 was posi-ive for the desired polydatin converting �-glucosidase activity, and hence selectedor further study.

.3. 16S rRNA gene sequencing and phylogenetic analysis

Extraction of genomic DNA from the 13 strains, including L. kimchi JB301,as carried out using a commercial genomic DNA extraction kit (Core Biosystem,aejeon, Korea). The 16S rRNA gene was PCR amplified from the chromosomal DNAsing the universal bacterial primer pair 9F and 1512R, and the purified PCR productsere sequenced by GenoTech (Daejeon, Korea). The 16S rRNA gene sequence of the

ype strain was obtained from EzTaxon server (http://www.eztaxon.org) [15]. Multi-lignment of related strains was done using clustal W [16]. 5′ and 3′-gaps were editedia BioEdit [17]. Neighbor-joining [18] and maximum parsimony methods [19] from

he PHYLIP suit program (http://evolution.genetics.washington.edu/phylip.html)ere used for construction of the phylogenetic tree. Bootstrap values were calcu-

ated with data restricted close to 1000 times and marked into branching point.volutionary distance matrix was estimated according to Kimura two-parameterodel [20].

l Technology 67 (2014) 59–66

2.4. Enzyme purification

For the purification of enzyme, the bacterial strain was incubated in MRS brothcontaining 1 mg/mL of polydatin without glucose at 37 ◦C for 6 d. Purification of �-glucosidase was carried out using a TOYOPEARL DEAE-650 M (Tosho, Tokyo, Japan)chromatography column according to the manufacturer’s protocol. The enzymesolution (3.5 mg/mL) was loaded onto the TOYOPEARL DEAE-650 M (3.2 × 20 cm,160 mL) column equilibrated using 20 mM sodium acetate buffer solution (NAB, pH6.0). The absorbed proteins were eluted with a 0–1 M sodium chloride linear gra-dient at a flow rate of 1 mL/min. Finally, the active fractions were dialyzed against20 mM NAB (pH 6.0) and the final purified protein was concentrated using an Ami-con Ultra 10,000 MWCO centrifugal filter (Millipore, Billerica, MA, USA). The enzymepurity and molecular weight were estimated by 12% sodium dodecyl sulfate poly-acrylamide gel electrophoresis (SDS-PAGE) using Novex Sharp standards proteins(Life Technologies, Carlsbad, CA, USA).

2.5. Enzyme assay

The enzyme activity was determined by measuring the release of glucose frompara-nitrophenyl �-glucoside (pNPG-�) (concentration, 0.2–5 mM), which was usedas the substrate. The reaction was stopped by addition of sodium carbonate solu-tion (final concentration, 0.67 M) to the enzyme reaction mixture. The amount of4-nitrophenol liberated from pNPG-� was determined by measuring absorptionat 400 nm in a cuvette (length, 1 cm) and using a molar extinction coefficient of55,560 M−1 cm−1. The enzyme activity was determined by the release of resveratrolor stilbene from polydatin or other stilbene glucosides by high-performance liquidchromatography (HPLC) analysis. The activity of enzyme was determined by therelease of glucose from �-glucobioses. One unit of enzyme activity was defined asthe amount of enzyme capable of hydrolyzing 1 �mol of glucose from stilbene glu-cosides per minute. Protein concentration was measured by the Bradford methodusing bovine serum albumin as the standard [21].

2.6. Effects of pH and temperature on enzyme activity

The enzyme (3.28 �g) was incubated at 37 ◦C in a 32 mM Britton–Robinsonbuffer solution (pH 2–11) with 10 mM polydatin to determine the optimum pH. Inaddition, the enzyme was also incubated at 4 ◦C for 8 h in a 32 mM Britton–Robinsonbuffer solution (pH 2–11), and the residual enzyme activity was examined to deter-mine the pH stability. The optimal temperature was determined by incubating theenzyme at 20–70 ◦C for 5 h in 20 mM NAB (pH 6.0). The thermal stability was ascer-tained by incubating the enzyme at 20–70 ◦C for 10 h in 20 mM NAB (pH 6.0), andthe residual enzyme activity was examined using 10 mM polydatin at 35 ◦C.

2.7. Analytical methods of hydrolysis products

The products of polydatin hydrolysis from a variety of bacterial strains wereanalyzed by TLC and HPLC. The 13 isolated strains from kimchi were incubated with10 mM polydatin in the culture broth at 37 ◦C for 24 h. At the pre-determined timeintervals, the reaction mixture was placed in a water bath for 5 min to halt enzymeactivity. Aliquots (10 �L) were removed, the reaction products were analyzed by TLCusing pre-coated silica gel 60 F254 plates (Merck, Darmstadt, Germany) and devel-oped in a solvent system consisting of acetonitrile and water [85:15 (v/v)], withresveratrol and polydatin as the standard compounds. The plates were dipped in asolution containing 0.03 g of N-(1-naphthyl) ethylenediamine and 5 mL of concen-trated sulfuric acid prepared in 95 mL of methanol, heated at 100 ◦C [22], and thenvisualized.

Chromatographic separation for quantitative analysis was achieved using a 1200series quaternary HPLC system (Agilent Technologies, Palo Alto, CA, USA) consist-ing of a G1311A quaternary pump with a G1322A vacuum degasser, a G1329Athermostatted autosampler, a G1316A column oven set at 30 ◦C, a ZORBAX SB-Aq(5 �m, 4.6 × 150 mm), and a G1314B DA detector set at 320 nm. After desalting withAmberlite MB-3 (Organo, Tokyo, Japan), the resultant digests were separated ona C18 column with a stepwise methanol gradient (0–100%). The fractions contain-ing the reaction products were collected and desalted again with Amberlite MB-3,followed by lyophilization. High performance liquid chromatography-mass spec-trometry (HPLC-MS)-based analysis for the isolated compounds was performedusing an HCT ultra PTM Discovery Ion Trap system (Bruker Daltonik, GmbH, Bremen,Germany) in the negative electrospray ionization mode.

3. Results

3.1. Screening for polydatin-converting strain from fermentedfood sources

We screened 13 bacterial strains from a Korean fermented food,kimchi, to identify an enzyme capable of converting polydatin intoresveratrol (Table 1). Each culture broth was incubated with 10 mMpolydatin for 24 h, and the hydrolysis product was analyzed by TLC

J.-A. Ko et al. / Enzyme and Microbial Technology 67 (2014) 59–66 61

Table 1Isolated lactic acid bacterial strains with polydatin-converting abilities.

Strain name Most closest type strain Similarity (%) b.p. Production yield ofresveratrol (%) frompolydatin

Enterococcus faecium Enterococcus faecium Aus0004 99 1164 <2Lactobacillus kimchi Lactobacillus kimchi NBRC 106466 99 1492 100Lactobacillus buchneri Lactobacillus buchneri NRRL B-30929 99 1432 28Lactobacillus zeae Lactobacillus zeae strain RIA 482 99 1522 81Pediococcus cellicola Pediococcus cellicola NBRC 106103 99 1503 80Lactobacillus salivarius Lactobacillus salivarius UCO 99 1507 31Lactobacillus reuteri Lactobacillus reuteri LR-5 97 1466 <2Lactobacillus sakei Lactobacillus sakei IMAU80189 99 1481 <2Lactobacillus casei Lactobacillus casei 073 99 1473 <2Lactobacillus fermentum Lactobacillus fermentum SCP27 99 1456 <2Lactobacillus plantarum Lactobacillus plantarum SCP49 99 1463 <2Lactobacillus harbinensis Lactobacillus harbinensis NBRC 100982 99 1496 26

S

oeorh

FLllff

Lactobacillus rhamnosus Lactobacillus sp. NBRC 107246

imilarity is based on 16s rRNA gene sequences. b.p., Nucleotide base pair.

r HPLC. Most strains did not hydrolyze polydatin and 6 strainsxhibited low hydrolysis activity toward it. As shown in Fig. 1,

ne strain showed resveratrol around a 100% conversion yield ofesveratrol, as determined by HPLC-MS meaning that it completelyydrolyzes the glucosidic linkage of polydatin. The isolated strain

ig. 1. HPLC profiles of the polydatin hydrolysis product and HPLC-MS analysis of isolatedane a, standard polydatin; lane b, standard resveratrol; lane c, Enterococcus faecium; lane dane g, Lactobacillus plantarum; lane h, Lactobacillus fermentum; lane i, Lactobacillus casei;

ane m, Lactobacillus buchneri; lane n, Lactobacillus zeae; lane o, Lactobacillus cellicola. Eachor 24 h, and then the reaction mixture was analyzed by HPLC as described in Section 2.rom reaction mixtures.

99 1495 12

had the lowest sequence similarity value based on the 16S rRNAgene sequence, which suggested that it might be classifiable as L.

kimchi (deposition number as a KCTC 12416BP). We designatedit as L. kimchi JB301. The enzyme from L. kimchi JB301 was fur-ther purified, its biochemical properties characterized, and finally

product. (A) HPLC profiles of polydatin hydrolysis products from isolated bacteria., Lactobacillus kimchi; lane e, Lactobacillus rhamnose; lane f, Lactobacillus harbinensis;lane j, Lactobacillus sakei; lane k, Lactobacillus reuteri; lane l, Lactobacillus salivarius;

isolate was cultured with 10 mM polydatin in modified MRS broth without glucose (B) HPLC-MS data of isolated P1 (d, Lactobacillus kimchi) as unknown compounds

62 J.-A. Ko et al. / Enzyme and Microbial Technology 67 (2014) 59–66

F . kimp lumng

uc

3b

�6pis(

F1aN

ig. 2. SDS-PAGE of the purified stilbene glucoside-specific �-glucosidase from Lurified stilbene glucoside-specific �-glucosidase after TOYOPEARL DEAE-650M colucoside-specific �-glucosidase by SDS-PAGE.

sed to produce resveratrol from polydatin and plant extracts of P.uspidatum.

.2. Purification of ˇ-glucosidase from L. kimchi JB301 and itsiochemical characterization

To characterize the biochemical properties, we purified the-glucosidase from L. kimchi JB301 followed by TOYOPEARL DEAE-50M column chromatography. The final protein (1.17 U/mg) was

urified approximately 1.7-fold from the crude enzyme, resulting

n a high recovery of activity (49%). The purified protein showed aingle band on a 12% SDS-PAGE with a molecular mass of 54 kDaFig. 2). The enzyme showed the highest activity at pH 6.0 and

ig. 3. Effects of pH (A) and temperature (B) on the activity and stability of the purified e0 mM polydatin in Britton–Robinson buffer at 37 ◦C for 5 h; (©), pH stability profile usinctivity profile at various temperatures using 3.28 �g enzyme in 20 mM NAB (pH 6.0) for

AB (pH 6.0) at 20−70 ◦C after a 10 h incubation. After sampling the reaction mixture at d

chi JB301. Panel A: Lane M, protein marker; lane 1, culture supernatant; lane 2, chromatography. Panel B: Determination of molecular weight of purified stilbene

maintained more than 80% of its maximal activity in the pH rangeof 3.0–9.0 (Fig. 3A). The enzyme showed the highest activity at30–40 ◦C in a 5 h-reaction, and was stable up to 20–50 ◦C for 10 h(Fig. 3B). In addition, the effect of metal ions on the enzyme wasstudied. Metal ions such as Zn2+, Mg2+, and Li+ at a concentration1 mM had no significant effect on the enzyme activity. However,the enzyme was 10% inhibited by 1 mM Ca2+ (data not shown).

3.3. Substrate specificities of the ˇ-glucosidase from L. kimchi

JB301

The substrate specificities of the �-glucosidase from L. kim-chi JB301 and almond �-glucosidase were investigated with

nzyme. In panel A: (�), Activity profile at various pH values using 3.28 �g enzyme,g 32.8 �g enzyme in 20 mM Britton–Robinson buffer at 4 ◦C for 8 h. In panel B: (�),5 h at various temperatures; (©), thermal stability using 32.8 �g enzyme in 20 mMesignated times, the remaining enzyme activity was assayed by a standard method.

J.-A. Ko et al. / Enzyme and Microbial Technology 67 (2014) 59–66 63

Fig. 4. HPLC profiles of the hydrolysis product of the stilbene glucoside-specific �-glucosidase with various stilbene glucosides. The reaction mixture containing the stilbeneglucoside-specific �-glucosidase (3.28 �g) and 5 mM of each substrate in 20 mM NAB (pH 6.0) was incubated at 35 ◦C for 20 h. Panel A: (a) polydatin at 0 h; (b) polydatin at20 h; (c) astringin at 0 h; (d) astringin at 20 h; (e) isorhaponticin at 0 h; (f) isorhaponticin at 20 h; (g) oxyresveratrol-3-O-glucoside at 0 h; (h) oxyresveratrol-3-O-glucoside at20 h; (i) mulberroside A at 0 h; (j) mulberroside A at 20 h. The reaction mixture was analyzed by HPLC as described in Section 2. Panels B and C, TLC analysis of each reactionmixture of panel A and HPLC-MS data of isolated P1 (b in panel A) and P2 (f in panel A) as unknown compounds from reaction mixture.

Table 2Characteristic parameters of the stilbene glucoside-specific �-glucosidase from L. kimchi JB301.

Substrate Almond�-glucosidase(U)

Stilbeneglucoside-specific�-glucosidase (U)

Stilbene glucoside-specific�-glucosidase

Stilbeneglucoside-specific�-glucosidase/almond�-glucosidase

Km (mM) kcat (s−1) kcat/Km (mM−1 s−1)

pNPG-� 20.5 4.83 0.77 3.96 5.14 0.24Polydatin –1 1.49 0.20 1.29 6.45Isorhaponticin –1 1.46 0.09 1.04 11.56Mulberroside A –1 0.07Astringin –1 –1

Oxyresveratrol-3-O-glucoside –1 0.04Sophorose 0.45 0.06 0.13Laminaribiose –1 –1

Cellobiose 0.25 0.03 0.12Gentiobiose 0.24 0.07 0.29

Determined at 35 ◦C and pH 6.0.1The dash indicates that the substrate did not hydrolyze.

6 crobial Technology 67 (2014) 59–66

drwsts3tdihiaiAtsotsg

3e

ck

Fig. 5. Time-course hydrolysis of polydatin extracts from P. cuspidatum by thestilbene glucoside-specific �-glucosidase from L. kimchi JB301. The reaction mix-ture containing the stilbene glucoside-specific �-glucosidase (3.28 �g) and 10 mgmethanol extracts from P. cuspidatum was reacted in 20 mM NAB (pH 6.0) at 35 ◦C

F

4 J.-A. Ko et al. / Enzyme and Mi

ifferent �-linked glucobioses and stilbene glucosides, as summa-ized in Table 2. L. kimchi JB301 �-glucosidase had low activitiesith respect to pNPG-� but the enzyme exhibited higher sub-

trate specificities toward polydatin and isorhaponticin comparedo the �-glucobioses. However, it rarely hydrolyzed mulberro-ide A, oxyresveratrol-3-O-glucoside, astringin, which has the′-hydroxyl group of polydatin, and desoxyrhaponticin, which hashe 4′-dehydroxy group and 3′-methoxy derivative of the poly-atin (Fig. 4). The enzyme kinetic parameters demonstrated that

sorhaponticin was the best substrate. The enzyme specificallyydrolyzes catalysis between the glucosidic moiety and resveratrol

n polydatin, isorhapontigenin in isorhaponticin (Table 2). The cat-lytic efficiency of the enzyme toward the stilbene glucosides wasn the following order: isorhaponticin » polydatin » mulberroside

> oxyresveratrol-3-O-glucoside. Based on analysis of the reactionime of the L. kimchi JB301 �-glucosidase suggested that mulberro-ide A was very weakly converted to glucose and oxyresveratrol viaxyresveratrol-3-O-glucoside (Fig. 4). These results indicate thathe enzyme hydrolyzes only the glucose moiety in stilbene gluco-ide. Finally, based on our results, we named the enzyme stilbenelucoside-specific �-glucosidase.

.4. Bioconversion of polydatin to resveratrol in P. cuspidatumxtract by cultivation of L. kimchi JB301

The time course of bioconversion of resveratrol from the P.uspidatum extracts was investigated during the growth of L.imchi JB301 (Fig. 5). After 50 h of incubation, the amount of

for 0–40 h. The methanol extracts from P. cuspidatum was cultivated with L. kimchiJB301 (a) 0 h; (b) 5 h; (c) 15 h; (d) 25 h; (e) 40 h, respectively.

ig. 6. The proposed enzyme catalysis of stilbene glucoside-specific �-glucosidase towards polydatin, astringin, isorhaponticin, desoxyrhaponticin, and mulberroside A.

J.-A. Ko et al. / Enzyme and Microbial Technology 67 (2014) 59–66 65

Table 3Comparison of biochemical properties of stilbene glucoside-specific �-glucosidase from different microorganisms.

Origin pNPG Km (mM) Polydatin Km (mM) Optimal pH pH stability Optimumtemperature (◦C)

Protein size (kDa) References

Aspergillus phoenicis 0.58 – 5.0 4.0–8.0 60 – [1]Aspergillus niger 0.48 – 4.5∼5.0 4.0–8.0 60 – [1]

T

ptear(

4

gattt

hfatdgrgoI[tAePicsiagdg3wwbdwtgbiR

tosm

Aspergillus oryzae sp.100 0.92 0.74 5.0

Lactobacillus kimchi JB301 0.77 0.20 6.0

he dash line was not characterized.

olydatin decreased markedly with the increase in fermentationime, whereas that of resveratrol increased and reached its high-st value at 40 h, when the percent conversion got to its highestt around 100%. Similarly, just like the vial test, the precipitatedesveratrol increased in accordance with the fermentation timedata is not shown).

. Discussion

In the present study, a novel stilbene glucoside-specific �-lucosidase isolated from L. kimchi JB301 was screened and purified,nd its biochemical properties were characterized. In addition,he kinetic efficiency of stilbene glucoside-specific �-glucosidaseowards various stilbene glucosides frequently reported in indus-rial natural products from P. cuspidatum was investigated.

�-Glucosidases are widely found in molds, yeasts, bacteria, andigher plants, and many have been purified and characterized. Thus

ar, only 3 enzymes, namely, A. oryzae �-glucosidase [1], A. nigernd yeast �-glucosidase [13], and snailase [14], have been reportedo be capable of hydrolyzing the �-glucosyl linkages of poly-atin. Comparison of their biochemical properties with the stilbenelucoside-specific �-glucosidase from L. kimchi JB301 is summa-ized in Table 3. As a putative lactic acid bacteria enzyme, stilbenelucoside-specific �-glucosidase has lower pH stability and lowerptimal temperature, 30–40 ◦C, than that of A. oryzae or A. niger [1].ts molecular weight is about 54 kDa smaller than that of A. oryzae3]. These results may be useful for genetic engineering and muta-ion studies as well as for solving the three dimensional structure.lthough fungal enzymes from Aspergillus exhibited the basic prop-rties [1], their substrate specificities towards stilbene glucoside in. cuspidatum and R. undulatum have not been systematically stud-ed [3,12–14]. Basically, stilbene glucoside-specific �-glucosidasean also transform polydatin into resveratrol and it shows broadubstrate specificities towards other stilbene glucosides such assorhaponticin, oxyresveratrol-3-O-glucoside, and mulberroside A,s compared to fungal enzymes [1,13]. However, the stilbenelucoside-specific �-glucosidase was unable to hydrolyze astringinue to steric hindrance at its 3′-hydroxy group. Since stilbenelucoside-specific �-glucosidase hydrolyzes isorhaponticin with′-methoxy but not astringin, it is believed to make hydrogen bondith the amino acid residues related to the substrate-binding siteith the 3′-hydroxy group in astringin (Fig. 6). Interestingly, stil-

ene glucoside-specific �-glucosidase was not able to hydrolyzeesoxyrhaponticin and the 4′-hydroxy group in desoxyrhaponticinas critical for substrate binding. To understand the reason for this,

he three dimensional structure of stilbene glucoside-specific �-lucosidase is required. In addition, these enzymatic properties cane applicable for bioconversion of stilbene glucosides and extract-

ng resveratrol from two major natural sources, P. cuspidatum and. undulatum.

Although several enzymes were reported to be able to increase

he yield of resveratrol from polydatin [1,12–14], the high costf enzyme production and toxic by product generation duringpontaneous fermentation are obstacles for industrial develop-ent. In this study, we isolated the polydatin-converting lactic acid

4.0–5.0 60 77 [1]3.0–6.0 30–40 54 This study

bacterial strains found in kimchi, a Korean fermented food havinghigh food safety. Almost all of the supplied polydatin, as well asthat in the P. cuspidatum extract, were transformed into resver-atrol in a 44 h cultivation period. The stilbene glucoside-specific�-glucosidase was able to bio-convert polydatin in the P. cuspi-datum extracts to resveratrol, and enzymes were also added tothe raw material powder to improve the yield of resveratrol. Tothe best of our knowledge, this is the first enzyme from lac-tic acid bacteria, L. kimchi JB301 having this function. Lactic acidproducing enzyme can be easily applicable for food, pharmaceut-icals, cosmetics, and nutraceuticals as a potential health-functionalcompound.

In conclusion, a novel stilbene glucoside-specific �-glucosidasewas isolated from edible L. kimchi JB301 and its biochemical proper-ties were characterized. Stilbene glucoside-specific �-glucosidaseshowed broad substrate specificities towards various stilbeneglucosides such as isorhaponticin, polydatin, oxyresveratrol-3-O-glucoside, and mulberroside A. We believe that the results fromour studies would contribute to an improvement in the indus-trial production of resveratrol by mixed bioconversion from variousplant sources. This study preludes the beginning of detailed inves-tigations into stilbene glucoside-specific �-glucosidase, and futurestudies should entail cloning the full-length gene as well as proteinengineering experiments.

Acknowledgement

This research was supported by the KRIBB Research InitiativeProgram, Republic of Korea This work was partially supportedby Fishery Commercialization Technology Development Programthrough KIMST funded by Ministry of Oceans and Fisheries and bythe KRIIBB Research Initiative Program, Republic of Korea.

References

[1] Zhang C, Li D, Yu H, Zhang B, Jin F. Purification and characterization of piceid-�-d-glucosidase from Aspergillus oryzae. Process Biochem 2007;42:83–8.

[2] Li H, Xia N, Forstermann U. Cardiovascular effects and molecular targets ofresveratrol. Nitric Oxid 2012;26:102–10.

[3] Sahu SS, Madhyastha S, Rao GM. Neuroprotective effect of resveratrol againstprenatal stress induced cognitive impairment and possible involvement of Na+,K+-ATPase activity. Pharmacol Biochem Behav 2013;103:520–5.

[4] Andrus MB, Liu J, Meredith EL, Nartey E. Synthesis of resveratrol using adirect decarbonylative Heck approach from resorcylic acid. Tetrahedron Lett2003;44:4819–1822.

[5] Vastano BC, Chen Y, Zhu N, Ho CT, Zhou Z, Rosen RT. Isolation and identificationof stilbenes of Polygonum cuspidatum. J Agric Food Chem 2000;48:253–6.

[6] Parshikov IA, Netrusov AI, Sutherland JB. Microbial transformation of anti-malarial terpenoids. Biotechnol Adv 2012;30:1516–23.

[7] Nicotra S, Cramarossa MR, Mucci A, Pagnoni UM, Riva S, Forti L. Biotransfor-mation of resveratrol: synthesis of trans-dehydrodimers catalyzed by laccasesfrom Myceliophtora thermophyla and from Trametes pubescens. Tetrahedron2004;60:595–600.

[8] Shin SY, Jung SM, Kim MD, Han NS, Seo JH. Production of resveratrol fromtyrosine in metabolically engineered Saccharomyces cerevisiae. Enzyme Microb

Technol 2012;51:211–6.

[9] Shin SY, Han NS, Park YC, Kim MD, Seo JH. Production of resveratrol from p-coumaric in recombinant Saccharomyces cerevisiae expressing 4-coumarate:coenzyme A ligase and stilbene synthase genes. Enzyme Microb Technol2011;48:48–53.

6 crobia

[

[

[

[

[

[

[

[

[

[

[

[

6 J.-A. Ko et al. / Enzyme and Mi

10] Zhou JJ, Zhang HJ, Yang PJ, Li HN. Determination of resveratrol glucoside andresveratrol in radix and rhizome of Polygonum cuspidatum yielded in Hanzhongregion. Chin Trad Herbal Drugs 2002;33:414–6.

11] Meng X, Maliakal P, Lu H, Lee MJ, Yang CS. Urinary and plasma levels of resvera-trol and quercetin in humans, mice, and rats after ingestion of pure compoundsand grape juice. J Agric Food Chem 2004;52:935–42.

12] Chem M, Li D, Gao Z, Zhang C. Enzymatic transformation of polydatin to resver-atrol by piceid-�-d-glucosidase from Aspergillus oryzae. Bioprocess Biosyst Eng2014;37:1411–6.

13] Jin S, Luo M, Wang W, Zhao CJ, Gu CB, Li CY, et al. Biotransformation of polydatinto resveratrol in Polygonum cuspidatum roots by highly immobilized edibleAspergillus niger and Yeast. Bioresour Technol 2013;136:766–70.

14] Wang Z, Zhao LC, Li W, Zhang LX, Zhang J, Liang J. Highly efficient biotransfor-

mation of polydatin to resveratrol by snailase hydrolysis using response surfacemethodology optimization. Molecules 2013;18:9717–26.

15] Chun J, Lee JH, Jung Y, Kim M, Kim S, Kim BK, et al. EzTaxon: a web-based tool forthe identification of prokaryotes based on 16S ribosomal RNA gene sequences.Int J Syst Evol Microbiol 2007;57:2259–61.

[

l Technology 67 (2014) 59–66

16] Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins DG. The CLUSTAL Xwindows interface: flexible strategies for multiple sequence alignment aidedby quality analysis tools. Nucleic Acids Res 1997;25:4876–82.

17] Hall TA. BioEdit: a user-friendly biological sequence alignment editor andanalysis program for Window 95/98/NT. Nucleic Acids Symp Ser 1999;41:95–8.

18] Saitou N, Nei M. The neighbor-joining method: a new method for reconstruc-ting phylogenetic trees. Mol Biol Evol 1987;4:406–25.

19] Kluge AG, Farris JS. Quantitative phyletics and the evolution of anurans. SystZool 1969;18:1–32.

20] Kimura M. The neutral theory of molecular evolution. Cambridge, UK:Cambridge University Press; 1984. p. 367.

21] Bradford MM. Rapid and sensitive method for the quantitation of micro-

gram quantities of protein utilizing the principle of protein-dye binding. AnalBiochem 1976;72:248–54.

22] Su D, Robyt JF. Control of the synthesis of dextran and acceptor-products by Leuconostoc mesenteroides B-512FM dextransucrase. CarbohydrRes 1993;248:471–6.