FEBS Letters 586 (2012) 4344–4350

journal homepage: www.FEBSLetters.org

In situ UDP-glucose regeneration unravels diverse functions of plant secondary ⇑ Kazuyoshi Terasaka a, Yuki Mizutani a, Akito Nagatsu b, Hajime Mizukami a, a Graduate School of Pharmaceutical Sciences, Nagoya City University, Japan b School of Pharmacy, Kinjyogakuin University, Japan article info abstract

Article history: The catalytic function of plant secondary product glycosyltransferases (PSPGs) was investigated by Received 29 September 2012 coupling the activities of recombinant flavonoid having different regiospecific- Revised 31 October 2012 ities with sucrose synthase from . In the present system, UDP, a product inhib- Accepted 31 October 2012 itor of PSPGs, was removed from the reaction mixture and used for regeneration of UDP-glucose by Available online 15 November 2012 AtSUS1. The in situ UDP-glucose regeneration system not only enhanced the glucosylation efficiency Edited by Ulf-Ingo Flügge but also unraveled the novel regioselectivity of PSPGs. The effect of the system was shown to be because of the removal of UDP from the reaction system and not because of the additional supply of UDP-glucose. Keywords: Plant secondary product Ó 2012 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. Glucosyl conjugation Function Regiospecificity In situ UDP-glucose regeneration

1. Introduction higher value [5]. A wide array of cDNA encoding PSPGs has been isolated from various plant sources, and the catalytic functions A unique feature of plant secondary metabolism is the capabil- have been extensively characterized using the recombinant en- ity to produce and accumulate structurally diverse glycosides, as zymes in vitro [6,7]. However, key structural factors determining exemplified in grapes, in which more than 200 different glycosides the specificities, particularly regiospecificities, of PSPGs are still have been identified [1]. Plant secondary product glycosyltransfer- poorly understood [8]. ases (PSPGs) catalyze the glycosylation of secondary products by When PSPG activity is assayed with UDP-sugar and an acceptor transferring sugar molecules from activated sugar donors, typically in vitro, UDP accumulates in the reaction mixture as a UDP-sugars to acceptor molecules. Consistent with the structural reaction product. Because of the structural similarity between diversity of plant glycosides, a large number of genes encoding the substrate UDP-sugar and the product UDP, the latter molecule PSPGs are present in the plant genome. In fact, more than 100 PSPG is generally a potent inhibitor of PSPGs [9], and PSPG activity de- genes have been identified in Arabidopsis thaliana [2]. clines as UDP accumulates in the reaction mixture. In contrast, Glycosylation, particularly glucosyl conjugation, of low-molecu- UDP produced by the glycosylation reaction may be rapidly con- lar-weight lipophilic compounds has a wide range of effects on verted to other metabolites such as UMP [10], and product inhibi- their physicochemical and physiological properties, including in- tion is unlikely to occur in the cellular environment. Therefore, creased water solubility, improved chemical stability, reduced measurement of the recombinant PSPG activity in vitro may not chemical toxicity, and altered pharmacokinetic and biological completely reflect functions of PSPGs in plant cells, and may pro- activities [3,4]. There is considerable potential for using PSPGs as vide limited information on the characteristics of the , novel biocatalysts for regio-, enantio-, and chemospecific glycosyl- including regiospecificity. ation of small molecules, leading to production of glycosides of Previously, we established a one-pot two- system for efficient glucosylation of small molecules by coupling the activity of two recombinant enzymes, PSPGs and sucrose synthase [11]. Abbreviations: MCS, multicloning site; PSPG, plant secondary product In this system, UDP produced from UDP-glucose by PSPGs is used glycosyltransferase. for UDP-glucose regeneration and is eliminated from the reaction ⇑ Corresponding author. Address: Graduate School of Pharmaceutical Sciences, Nagoya City University, Mizuho-ku, Nagoya 467-8603, Japan. Fax: +81 52 836 3415. system by sucrose synthase; thus, it mimics the enzymatic glu- E-mail address: [email protected] (H. Mizukami). cosylation of secondary metabolites in the cellular environment.

0014-5793/$36.00 Ó 2012 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.febslet.2012.10.045 K. Terasaka et al. / FEBS Letters 586 (2012) 4344–4350 4345

In the present study, we compared the glucosylation products of was performed on a reversed-phase column (COSMOSIL 5C18-ARII, recombinant flavonoid glucosyltransferases with and without the Nacalai Tesque, Kyoto, Japan), and the eluates were monitored at UDP-glucose regeneration system, and demonstrated that PSPGs 340 nm. The solvent conditions were as follows (flow rate, 1 ml/ may have more diverse function in cells than that suggested by min): 0–10 min, 10% methanol in 0.1% formic acid; 10–20 min, in vitro assays. 10%–50% methanol in 0.1% formic acid; 20–25 min, 50%–100% methanol in 0.1% formic acid; 25–27 min, 100% methanol. For anal- 2. Materials and methods ysis of resveratrol glucosides the following solvent conditions were used (flow rate, 1 ml/min): 0–5 min, 5% acetonitril in water; 5– 2.1. Glucosyltransferases and sucrose synthase 20 min, 5–35% acetonitril in water; 20–22.5 min, 35–100% acetoni- tril in water; 22.5–27.5 min, 100% acetonitril, and the elution was UGT73A15 cDNA was previously isolated from Catharanthus ro- traces at 300 nm. For quantitative determination of quercetin glu- seus cell cultures in our laboratory [12]. F7GT cDNA isolated from cosides the products were quantified on the basis of calibration Scutellaria baicalensis[13] was a generous gift of Dr. T. Yoshikawa, curves prepared using the respective glucosides. Professor Emeritus of Kitasato University. UGT71C1 and UGT71C3 cDNA was isolated from A. thaliana by reverse RT-PCR on the basis 2.6. Isolation and identification of glucosylation products of the sequence retrieved from the DNA database [accession nos.: AC005496 (At2g29750) for UGT71C1 and AC067971 (At1g07260) For preparative-scale glucosylation in order to elucidate the for UGT71C3] [14]. The open reading frames (ORFs) of these cDNA structure of the products, the reaction was performed in 100 ml molecules were subcloned into pQE-30 vectors. Isolation of A. tha- of reaction mixture containing 1 mM acceptor substrate, 2.5 mM liana sucrose synthase 1 (AtSUS1) cDNA was described earlier [11]. UDP-glucose, 0.3 M sucrose, crude AtSUS1, and the respective crude at 30 °C overnight. The reaction was ter- 2.2. Chemicals minated by the addition of 200 ml methanol to the mixture, fol- lowed by centrifugation at 10,000g for 10 min. The supernatant The co-expression vector pRSFDuet-1 was purchased from was directly applied to a DIAION HP20 column (Mitsubishi Chem- Merck KGaA (Darmstadt, Germany). The expression vector pQE- ical Corporation, Tokyo, Japan), which was then washed with water 30 was from Qiagen (Valencia, CA, USA). All other chemicals used to remove sucrose. The products were eluted with methanol and were of commercial reagent-grade quality, unless otherwise each product was further purified by preparative HPLC. For isola- stated. tion of the glucosides with similar retention times, shallower gra- dient elution was used than that described in Section 2.5. The 1 13 2.3. Construction of the co-expression vector pRSFDuet-UGT73A15/ isolated products were analyzed by H and C NMR spectroscopy AtSUS1 using a Brucker ABANCE 600 (Brucker biospin, Bremen, Germany).

A pRSFDuet-1 vector was used for the co-expression of 2.7. UGT73A15 and AtSUS1 in Escherichia coli. These recombinant en- zymes were expressed as fusion proteins with a His6 tag at the To determine apparent Km and Ki values for UDP-glucose and N-terminus for purification. ORFs of UGT73A15 and AtSUS1 sub- UDP, respectively, enzyme assays were performed in duplicate cloned into pQE-30 were amplified by PCR with KOD FX DNA poly- using curcumin (UGT73A15) or kaempferol (UGT71C1) as a gluco- merase (Toyobo, Tokyo, Japan). PCR products were subcloned into syl acceptor substrate at a fixed concentration of 0.5 mM with 2 mg pRSFDuet-1 through the appropriate restriction sites. Addition of of each crude enzyme at 30 °C for 5 or 10 min. The UDP-glucose the sequence encoding a His6 tag to AtSUS1 was per- concentrations used were 0.1–2.5 mM. UDP was added to the reac- formed by inverse PCR. Self-ligation of the PCR products was per- tion mixture at a concentration of 0.2 mM. The initial velocity data formed using T4 polynucleotide kinase and . Successful were visualized with Lineweaver–Burk plots, and kinetic parame- introduction of the His6 tag was confirmed by nucleotide sequenc- ters were calculated on the basis of linear regression analysis using ing of the entire AtSUS1 gene. The E. coli strain Rosetta-gami2 Excel 2007 (Microsoft Japan, Tokyo, Japan). HPLC conditions for (Merck KGaA) was used as the host for expression. determining curcumin glucosides were described previously [16].

2.4. Preparation of recombinant enzymes 3. Results and discussion The transformed bacteria were cultured at 30 °C in Luria– Bertani medium containing 50 lg/ml carbenicillin for 1 day and 3.1. Glucosylation of quercetin by UGT73A15 in the presence of an then harvested by centrifugation and stored at 80 °C until use. in situ UDP-glucose regeneration system The bacterial pellet was homogenized in 50 mM sodium phosphate buffer (pH 8.0). The slurry was centrifuged at 12,000g for 15 min The coding regions of UGT73A15 and AtSUS1 cDNA with His6- at 4 °C, and the supernatant was used as a crude enzyme prepara- tag sequences were inserted into the multicloning sites (MCS) 1 tion. The protein concentration of the supernatant was determined and 2, respectively, of pDuet-1. The two N-terminal His-tagged according to the method described by Bradford [15]. polypeptides were coexpressed as soluble enzymes in E. coli (Fig. S1). For the enzyme assay, the crude protein preparation con- 2.5. Enzyme assay taining both UGT73A15 and AtSUS1 was added to the reaction mixture comprising quercetin as a sugar acceptor and UDP-glucose The standard reaction mixture (100 ll) contained 50 mM Tris– as a sugar donor. UGT73A15 was first isolated from cultured cells HCl (pH 8.0), 2.5 mM UDP-glucose, 1 mM acceptor substrate, and of C. roseus as UDP-glucose:curcumin glucosyltransferase; how- the crude enzyme mixture containing glucosyltransferase and su- ever, it exhibits glucosyl transfer activity on various phenolic com- crose synthase with or without 0.3 M sucrose. The reaction was pounds, including flavonols [12]. Sucrose synthase 1 was isolated incubated at 30 °C for an appropriate time and terminated by the from A. thaliana. Sucrose synthase (SUS; E.C. 2.4.2.13.) is capable addition of 200 ll methanol. After centrifugation at 12,000g for of catalyzing both synthesis and cleavage of sucrose. Although 5 min, the reaction products were analyzed by HPLC. HPLC analysis the reaction is reversible, its biological function is considered to 4346 K. Terasaka et al. / FEBS Letters 586 (2012) 4344–4350

Fig. 1. HPLC analysis of glucosylation products of quercetin by UGT73A15 with sucrose synthase (AtSUS1). Quercetin was incubated with UGT73A15 and AtSUS1 (a) in the absence or (b) in the presence of sucrose for 1, 6 h, and overnight, and the assay mixtures were subjected to HPLC analysis. The elution was monitored at 340 nm. Chemical structures of glucosylation products (c) were determined based on the 1H and 13C NMR spectra of the isolated products.

be the synthesis of nucleotide diphosphate derivatives of glucose mixture, whereas without the addition of sucrose, it accumulates in the presence of high concentrations of sucrose [17]. Accordingly, in the mixture. when sucrose is added at a high concentration (0.3 M) to the reac- Three major glucosylation products (Qg1–Qg3) were detected tion mixture, UDP produced by the glucosylation reaction is con- when quercetin and UDP-glucose were incubated with the verted to UDP-glucose and thus eliminated from the reaction UGT73A15/AtSUS1 crude enzyme mixture for 1 h in the absence K. Terasaka et al. / FEBS Letters 586 (2012) 4344–4350 4347

Fig. 2. Glucosylation of trans-resveratrol by UGT73A15 with sucrose synthase (AtSUS1). Resveratrol was incubated with UGT73A15 and AtSUS1 (a) in the absence or (b) in the presence of sucrose for 1 h and overnight, and the assay mixtures were subjected to HPLC analysis. The elution was monitored at 300 nm. Chemical structures of Rg1 and Rg2 (c) were estimated by comparing the Rt values with those described in Ref. [14]. of sucrose. No further products were detected even after prolonged glucosylation products (Rg1 and Rg2) by the UGT73A15/AtSUS1 incubation (Fig. 1a). In contrast, when sucrose was added to the crude enzyme mixture in 1 h and two additional products Rg3 reaction mixture, quercetin was completely converted to the three and Rg4 were detected after overnight incubation (Fig. 2a). In the glucosylation products Qg1, Qg2, and Qg3 within 1 h of incubation. presence of sucrose, formation of Rg3 and Rg4 was detected within Further incubation led to the formation of five additional products, 1 h and overnight incubation led to formation of two additional Qg4–Qg8, while the amounts of Qg1, Qg2, and Qg3 decreased products Rg5 and Rg6 which were not formed in the absence of su- (Fig. 1b). The eight products Qg1–Qg8 were isolated from the reac- crose even after prolonged incubation (Fig. 2c). Rg1 and Rg2 were tion mixture by DIAION HP-20 chromatography, followed bypre- presumably identified as resveratrol 3-O-glucoside and resveratrol parative HPLC. Based on the 1H and 13C NMR spectra, Qg1–Qg8 4’-O-glucoside, respectively, by comparing their Rt values with were identified as quercetin 40-O-glucoside (Qg1), quercetin 3-O- those reported in Ref. [19] (Fig. 2c). Rg3 to Rg6 are supposed to glucoside (Qg2), quercetin 3,40-O-diglucoside (Qg3), quercetin be resveratrol 5-O-glucoside and di- and/or triglucosides of resve- 3,30-O-diglucoside (Qg4), quercetin 7,40-O-diglucoside (Qg5), quer- ratrol although their structural identification awaits further cetin 3,7-O-diglucoside (Qg6), quercetin 3,7,30-O-triglucoside investigation. (Qg7), and quercetin 3,7,40-O-triglucoside (Qg8) (Fig. 1c). NMR data A time course of changes in quercetin glucosylation was com- of these products are shown in Supplementary material. pared for incubation with and without sucrose. When quercetin The results clearly showed that in vitro assay of recombinant was incubated with UGT73A15/AtSUS1 without the addition of su- UGT73A15 without coupling of the UDP-regeneration system crose, the quantity of quercetin monoglucosides (Qg1 plus Qg2) in- may lead to a conclusion that this enzyme exhibits regiospecificity creased until 4 h and then gradually decreased. The formation of of glucosyl conjugation of the 3- and 40-OH groups of quercetin. diglucoside (Qg3) increased after a 1-h lag, reached a maximum However, once UDP is removed from the reaction mixture by con- after 6 h, and then gradually decreased (Fig. 3a). The overall yield version to UDP-glucose by sucrose synthase, UGT73A15 is seen to of quercetin glucosides was 20% of the initially added quercetin have glucosyl conjugation activity on all the phenolic hydroxyl when calculated at the highest time point (6-h incubation). In groups, except for 5-OH, and a wide variety of flavonol glucosides sharp contrast to this pattern, when sucrose was included in the up to triglucosides can be produced enzymatically. Of the eight reaction mixture, the diglucosides (Qg3–Qg6) appeared after glucosylation products, quercetin 3,7,30-O-glucoside (Qg7) has 15 min of incubation, increased up to 6 h, and then gradually de- not, to our knowledge, been found in nature, whereas quercetin creased. Furthermore, the formation of triglucoside (Qg7 and 3,7,40-O-glucoside (Qg8) has been found as a biotransformation Qg8) was first detected after 4-h incubation and then increased lin- product in Vitis cell cultures exogenously supplemented with quer- early until 24 h. Almost 100% of the initially added quercetin was cetin [18]. converted to glucosides after 24-h incubation (Fig. 3b). To clarify To confirm that coupling the UDP-glucose regeneration system whether the striking effect of UDP-glucose regeneration from with UGT73A15 is applicable to other highly substituted acceptor UDP was caused by the additional supply of UDP-glucose to the substrates in addition to flavonoids, we compared the glucosyla- reaction system or the removal of UDP from the reaction system, tion of trans-resveratrol between with and without the UDP-glu- we added UDP-glucose (Fig. 3c) or sucrose (Fig. 3d) to the glucosy- cose regeneration system. Resveratrol was converted to two lation mixture 4 h after the reaction was started without sucrose. 4348 K. Terasaka et al. / FEBS Letters 586 (2012) 4344–4350

Fig. 3. Time course of changes in quercetin glucosylation catalyzed by UGT73A15 with AtSUS1. Assays were performed at 30 °C for 0, 10 30, 60, and 120 min in 1.5 ml of 50 mM Tris–HCl buffer (pH 8.0) comprising (a) 2.5 lmol quercetin, 4 lmol UDP-glucose, and 5 mg crude protein containing UGT73A15 and AtSUS1 or (b) 0.3 mmol sucrose, 2.5 lmol quercetin, 4 lmol UDP-glucose, and 5 mg of crude enzyme mixture. (c) UDP-glucose (2 lmol) or (d) sucrose (0.24 mmol) was added to the reaction mixture 4 h after the reaction was started without sucrose. Closed circles, total amount of quercetin monoglucoside; closed squares, total amount of quercetin diglucoside; closed triangles, total amount of quercetin triglucoside; open circles, total amounts of quercetin mono-, di-, and triglucosides.

The addition of UDP-glucose led to a slight increase in the forma- appears in the combination of AtSUS1 with PSPGs other than tion of monoglucosides and diglucosides, whereas the addition of UGT73A15, we examined the glucosylation of apigenin by flavo- sucrose induced a drastic increase in the formation of diglucosides. noid 7-O-glucosyltransferase (F7GT) from S. baicalensis and querce- Quercetin triglucosides began to accumulate 4 h after the addition tin by UGT71C1 and UGT71C3 from A. thaliana. F7GT was first of sucrose, whereas monoglucosides decreased. Little triglucoside isolated as wogonin 7-O-glucosyltransferase from root cultures of formation was detected when the reaction mixture was addition- S. baicalensis[13]. UGT71C1 exhibited regiospecificity toward the ally supplemented with UDP-glucose. 7- and 30-OH groups of quercetin, whereas UGT71C3 produced only These results unambiguously reveal that glucosylation effi- quercetin 3-O-glucoside (isoquercitrin) when assayed in vitro ciency can be drastically enhanced by coupling the UDP-glucose using their recombinant enzymes [14]. In the experiments de- regeneration system with UGT73A15 and that the effect is not be- scribed in this section, the recombinant enzymes (glucosyltransfe- cause of the additional supply of UDP-glucose to the reaction sys- rases and AtSUS1) were individually expressed in E. coli, and the tem but because of the removal of UDP accumulating in the crude enzyme preparations from each bacterial culture were added reaction mixture. to the reaction mixture. First, glucosylation of apigenin by F7GT with or without UDP- 3.2. Glucosylation of flavonoids by various glucosyltransferases in the glucose regeneration was examined. Two product peaks corre- presence of the in situ UDP-glucose regeneration system sponding to apigenin 7-O-glucoside (Ag1) and 7,40-O-diglucoside (Ag2) were detected by HPLC analysis after 2-h incubation in the UDP elimination from the reaction mixture by linking UDP-glu- absence of sucrose (Fig. 4a). Further incubation led to an increase cose regeneration with glucosylation by UGT73A15 not only led to in 7,40-O-diglucoside (Ag2) with a decrease in 7-O-glucoside. When a drastic increase in the glucosylation efficiency but also unraveled sucrose was included in the reaction mixture, as shown in Fig. 4b, a novel regiospecificity of the enzyme that was masked by the the third product corresponding to apigenin 5,7-O-diglucoside product inhibition of UDP. To investigate whether such an effect (Ag3) appeared after 2-h incubation. Overnight incubation resulted K. Terasaka et al. / FEBS Letters 586 (2012) 4344–4350 4349

Fig. 4. Glucosylation of apigenin by F7GT with AtSUS1. Apigenin was incubated with F7GT and AtSUS1 (a) in the absence or (b) in the presence of sucrose for 2 h and overnight, and the assay mixture was subjected to HPLC analysis. Elution was monitored at 340 nm. Chemical structures of glucosylation products (c) were determined on the basis of the 1H and 13C NMR spectra of the isolated products. in the formation of 5,7,40-O-triglucoside (Ag4). The chemical struc- tures of Ag1–Ag4, as shown in Fig. 3c, were confirmed by compar- ison of the 1H NMR spectra with those reported previously [20,21]. The 1H NMR data of Ag1–Ag4 are shown in Supplementary mate- rial. Although F7GT was first characterized as a flavanone 7-O-glu- cosyltransferase, the results clearly demonstrate that F7GT is capable of conjugating a glucose molecule to all the hydroxyl groups of apigenin once the reaction is coupled with the UDP-glu- cose regeneration system by AtSUS1. Regiospecificity of UGT71C3 was then examined using querce- tin as an acceptor substrate. A single glucosyl conjugation product corresponding to quercetin 3-O-glucoside (Qg2) was detected by HPLC analysis (Fig. 5a), consistent with the previous report describing UGT73C3 as a flavonol 3-O-glucosyltransferase [14]. However, when the UDP-glucose regeneration system was coupled with the glucosyl transfer reaction by the addition of sucrose to the reaction mixture, three additional products, quercetin 3,30-O-dig- lucoside (Qg3), 3,40-O-diglucoside (Qg4), and 3,7-O-diglucoside (Qg6), were produced, as shown in Fig. 5b. Thus, UGT71C3 is capa- Fig. 5. Glucosylation of quercetin by UGT71C3 with AtSUS1. Quercetin was ble of conjugating glucose to not only the 3-OH but also the 7-, 30-, incubated with F7GT and AtSUS1 (a) in the absence or (b) presence of sucrose 0 overnight, and the assay mixture was subjected to HPLC analysis. Elution was and 4 -OH groups of quercetin when UDP has not accumulated in monitored at 340 nm. For the products Qg2, Qg3, Qg4, and Qg6, see Fig. 1c. the reaction mixture. Finally, we compared glucosyl conjugation by an enzyme mix- ture of UGT71C1 and AtSUS1 with and without sucrose. UGT71C1 converted quercetin to three glucosylation products corresponding diglucoside were detected even after prolonged incubation (data to quercetin 30-O-glucoside, quercetin 7-O-glucoside, and querce- not shown). The results indicated that UGT71C1 performs posi- tin 7,30-O-diglucoside. When the glucosyl transfer reaction was tion-specific glucosylation of the 7- and 30-OH of quercetin even coupled with UDP-glucose regeneration by the addition of sucrose if the reaction was coupled with the UDP-glucose regeneration to the reaction mixture, no other products than quercetin 7,30-O- reaction. 4350 K. Terasaka et al. / FEBS Letters 586 (2012) 4344–4350

3.3. Kinetic analysis of the glucosyl transfer reaction catalyzed by References UGT73A15 and UGT71C1 [1] Sefton, M.A., Francis, I.L. and Williams, P.J. (1994) Free and bound volatils secondary metabolites of Vitis vinifera grape cv. Souvignon Blanc. J. Food Sci. The in situ UDP-glucose regeneration system unraveled a broad 59, 142–147. regiospecificity of UGT73A15, whereas no such novel function ap- [2] Li, Y., Baldauf, S., Lim, E.K. and Bowles, D.J. (2001) Phylogenetic analysis of the peared for UGT71C1 even in the presence of the UGT-regeneration UDP-glycosyltransferase multigene family of Arabidopsis thaliana. J. Biol. system, although glucosylation efficiency was increased. UDP Chem. 278, 43910–43918. [3] Jones, P. and Vogt, T. (2001) Glycosyltransferases in secondary plant inhibits PSPGs in general; however, sensitivity to UDP inhibition metabolism: tranquilizers and stimulant controllers. Planta 213, 164–174. varies from one enzyme to another. For example, flavanone 7-O- [4] Makino, T., Shimizu, R., Kanemaru, M., Suzuki, Y., Moriwaki, M. and Mizukami, glucosyltransferase from Petunia hybrida was strongly inhibited H. (2009) Enzymatically modified isoquercitrin, a-oligoglucosyl quercetin 3- O-glucoside, is absorbed more easily than other quercetin glycosides or by UDP with a Ki of 0.009 mM [22], whereas flavanone 7-O-hydrox- aglycone after oral administration in rats. Biol. Pharm. Bull. 32, 2034–2040. ylase of grapefruit was more than 10,000-fold less sensitive to UDP [5] Lim, E.K. (2005) Plant glycosyltransferases: their potential as novel [23]. To investigate whether such a difference in sensitivity to inhi- biocatalysts. Chem. Eur. J. 11, 5486–5494. [6] Bowles, D., Lim, E.K., Poppenberger, B. and Vaistij, F.E. (2006) bition by UDP causes the difference between UGT73A15 and Glycosyltransferases of lipophilic small molecules. Annu. Rev. Plant Biol. 8, UGT71C1 in response to the UDP-regeneration system, we deter- 254–263. [7] Malik, V. and Black, G.W. (2012) Structural, functional and mutagenesis mined the apparent Km values for UDP-glucose and Ki values for studies of UDP-glycosyltransferases. Adv. Protein Chem. Struct. Biol. 87, 87– UDP, with respect to UDP-glucose, of UGT73A15 and UGT71C1 115. on the basis of pseudo-single substrate kinetics, in which glucose [8] Cartwright, A.M., Lim, E.K., Kleanthous, C. and Bowles, D.J. (2008) A kinetic acceptor substrates were maintained at a constant concentration. analysis of regiospecific glucosylation by two glycosyltransferases of Arabidopsis thaliana. J. Biol. Chem. 283, 15724–15731. Km values for UDP-glucose did not greatly differ between the two [9] Terragni, J., Bitinaite, J., Zheng, Y. and Pradhan, S. (2012) Biochemical PSPGs (0.28 ± 0.07 mM for UGT73A15 and 0.69 ± 0.08 mM for characterization of recombinant b-glucosyltransferase and analysis of global UGT71C1: an average with a standard deviation from three inde- 5-hydroxymethylcytosine in unique genomes. Biochemistry 51, 1009–1019. pendent measurements); however, the K value for UDP of [10] D’Allsio, C., Trombetta, E.S. and Parodi, A.J. (2003) Nucleotide diphosphatase i and glycosyltransferase activities can localize to different subcellular UGT73A15 (0.019 ± 0.001 mM) was 25-fold less than that of compartments in Schzosaccharomyces pombe. J. Biol. Chem. 278, 22379–22387. UGT71C1 (0.51 ± 0.34 mM). The strong sensitivity of UGT73A15 [11] Masada, S., Kawase, Y., Nagatoshi, M., Oguchi, Y., Terasaka, K. and Mizukami, to UDP inhibition may lead to a drastic effect of the UDP-regener- H. (2007) An efficient chemoenzymatic production of small molecule glucosides with in situ UDP-glucose regeneration. FEBS Lett. 581, 2562–2566. ation system on enzyme function. Lineweaver–Burk plots revealed [12] Kaminaga, Y., Sahin, F.P. and Mizukami, H. (2004) Molecular cloning and that the inhibition by UDP was competitive for both GTs, as ex- characterization of a glucosyltransferase catalyzing glucosylation of curcumin pected (data not shown). in cultured Catharanthus roseus cells. FEBS Lett. 567, 197–202. [13] Hirotani, M., Kuroda, R., Suzuki, H. and Yoshikawa, T. (2000) Cloning and expression of UDP-glucose: flavonoid 7-O-glucosyltransferase from hairy root 4. Conclusion cultures of Scutellaria baicalensis. Planta 210, 1013–1063. [14] Lim, E.K., Ashford, D.A., Hou, B., Jackson, R.G. and Bowles, D.J. (2004) Arabidopsis glucosyltransferases as biocatalysts in fermentation for In in vitro assays of recombinant PSPGs, UDP, one of the reac- regioselective synthesis of diverse quercetin glucosides. Biotechnol. Bioeng. tion products accumulating in the reaction system, may obscure 87, 623–631. our understanding of the PSPG function in cellular environments. [15] Bradford, M.M. (1976) A rapid and sensitive method for quantitation of Linking the in situ UDP-glucose regeneration system to glucosyl microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. transfer reactions is a powerful tool for unraveling previously [16] Kaminaga, Y., Nagatsu, A., Akiyama, T., Sugimoto, N., Yamazaki, T., Maitani, T. unanticipated regiospecificity of the recombinant PSPGs, particu- and Mizukami, H. (2003) Production of unnatural glucosides of curcumin with larly when the enzymes are sensitive to the inhibition by UDP. drastically enhanced water solubility by cell suspension cultures of Catharanthus roseus. FEBS Lett. 555, 311–316. Notably, the method is applicable to not only flavonoid acceptors [17] Dennis, D.T. and Blakeley, S.D. (2000) Carbohydrate Metabolism in: but structurally diverse acceptor substrates, and to not only UDP- Biochemistry & Molecular Biology of Plants (Buchanan, B.B., Gruissem, W.J. glucose glucosyltransferase but also PSPGs using various UDP-sug- and Jones, R.L., Eds.), pp. 630–675, American Society of Plant Physiologists, Rockville. ars as sugar acceptors, given that UDP is eliminated while regener- [18] Kokubo, T., Nakamura, M., Yamakawa, T., Noguchi, H. and Kodama, T. (1991) ated UDP-glucose is not utilized as a sugar donor. Quercetin 3,7,40-triglucoside formation from quercetin by Vitis hybrid cell cultures. Phytochemistry 30, 829–831. Acknowledgments [19] Ozaki, S., Imai, H., Iwakiri, T., Sato, T., Shimoda, K., Nakayama, T. and Hamada, H. (2012) Regioselective glucosidation of trans-resveratrol in Escherichia coli expressing glucosyltransferase from Phytolacca americana. Biotechnol. Lett. The authors wish to thank Dr. T. Yoshikawa, Professor Emeritus 34, 475–481. 1 of Kitasato University, for the generous gift of F7GT cDNA clone. [20] Markham, K.R. and Geiger, H. (1993) H nuclear magnetic resonance spectroscopy of flavonoids and their glycosides in The present study was supported by a Grant-in-Aid for Scientific hexadeuterodimethylsulfoxide in: The Flavonoids. Advances in Research Research and a Grant-in-Aid for Young Scientists from the Ministry Since 1986 (Harborne, J.B., Ed.), pp. 441–497, Chapman & Hall, London. of Education, Culture, Sports, Science and Technology of Japan. [21] Agrawal, P.K. and Bansal, M.C. (1989) Flavonoid glycosides in: Carbon-13 NMR of Flavonoids (Agrawal, P.K., Ed.), pp. 283–364, Elsevier, Amsterdam, Oxford, New York, Tokyo. Appendix A. Supplementary data [22] Durren, R.L. and McIntosh, C.A. (1999) Flavanone 7-O-glucosyltransferase activity from Petunia hybrida. Phytochemistry 52, 793–798. [23] McIntosh, C.A. and Mansell, R.D. (1990) Biosynthesis of naringenin in Citrus Supplementary data associated with this article can be found, in paradisi: UDP-glucosyltransferase activity in grapefruit seedlings. theonlineversion,athttp://dx.doi.org/10.1016/j.febslet.2012.10.045. Phytochemistry 29, 1533–1538.