Substrate specificity of the human UDP-glucuronosyltransferase UGT2B4 and UGT2B7 Identification of a critical aromatic amino acid residue at position 33 Lydia Barre 1 , Sylvie Fournel-Gigleux 1 , Moshe Finel 2 , Patrick Netter 1 , Jacques Magdalou 1 and Mohamed Ouzzine 1 1 UMR 7561 CNRS, Universite ´ Henri Poincare ´ – Nancy I, Faculte ´ de Me ´ decine, Vandoeuvre-le ` s-Nancy, France 2 Drug Discovery and Development Technology Center (DDTC), Faculty of Pharmacy, University of Helsinki, Finland UDP-glucuronosyltransferases (UGT) constitute a super- family of enzymes that are involved in the phase II detoxification pathway of many drugs, pollutants pre- sent in our environment and numerous exogenous compounds [1]. They catalyze the formation of glu- curonides by the transfer of glucuronic acid, from the high energy donor UDP-glucuronic acid, to hydroxyl, carboxyl or amine groups of structurally diverse mole- cules. The hydrophilic glucuronides are readily excreted from the body via urine and bile. Endogenous compounds, such as bilirubin, fatty acids, steroids and retinoic acid are also substrates of UGTs. Thus, these enzymes that are expressed in several tissues, such as liver, lung, brain, kidney and gastro-intestinal tract, play a major role in both physiological and toxicologi- cal processes [2]. UGTs have been classified into two main sub- families, UGT1A and UGT2B, based on similarities between their amino acid sequences and gene organiza- tion. Molecular cloning of cDNAs has identified to date up to 16 human UGT isoforms, most of which have been extensively characterized in terms of sub- strate specificity upon heterologous expression [3]. Determination of their activity towards series of sub- stances led to the conclusion that most of them present distinct, but frequently overlapping substrate specifici- ties [4]. Interestingly, this redundancy provides an effi- cient protection against toxicity of drugs, pollutants Keywords site-directed mutagenesis; substrate specificity; UDP-glucuronosyltransferase; UGT2B4; UGT2B7 Correspondence M. Ouzzine, UMR 7561 CNRS-UHP-Nancy I, Faculte ´ de Me ´ decine, BP 184, F-54505 Vandoeuvre-le ` s-Nancy cedex, France Fax: +33 3 83683959 Tel: +33 3 83683972 E-mail: ouzzine@medecine.uhp-nancy.fr (Received 10 November 2006, revised 21 December 2006, accepted 22 December 2006) doi:10.1111/j.1742-4658.2007.05670.x The human UDP-glucuronosyltransferase (UGT) isoforms UGT2B4 and UGT2B7 play a major role in the detoxification of bile acids, steroids and phenols. These two isoforms present distinct but overlapping substrate spe- cificity, sharing common substrates such as the bile acid hyodeoxycholic acid (HDCA) and catechol-estrogens. Here, we show that in UGT2B4, sub- stitution of phenylalanine 33 by leucine suppressed the activity towards HDCA, and impaired the glucuronidation of several substrates, including 4-hydroxyestrone and 17-epiestriol. On the other hand, the substrate speci- ficity of the mutant UGT2B4F33Y, in which phenylalanine was replaced by tyrosine, as found at position 33 of UGT2B7, was similar to wild-type UGT2B4. In the case of UGT2B7, replacement of tyrosine 33 by leucine strongly reduced the activity towards all the tested substrates, with the exception of 17-epiestriol. In contrast, mutation of tyrosine 33 by phenyl- alanine exhibited similar or even somewhat higher activities than wild- type UGT2B7. Hence, the results strongly indicated that the presence of an aromatic residue at position 33 is important for the activity and substrate specificity of both UGT2B4 and UGT2B7. Abbreviations HDCA, hyodeoxycholic acid; UGT, UDP-glucuronosyltransferase. 1256 FEBS Journal 274 (2007) 1256–1264 ª 2007 The Authors Journal compilation ª 2007 FEBS and harmful endogenous compounds. When the activ- ity of one isoform is impaired by mutations or upon inhibition, other UGTs can often act as a relay to overcome the deficiency. Such redundancy in substrate specificity is clearly observed for the human UGT2B4 and UGT2B7. UGT2B4 is mainly involved in the glucuronidation of the bile acid, hyodeoxycholic acid (HDCA) [5] and catechol-estrogens, such as 17-epiestriol and 4-hydroxy- estrone [6]. In addition to the substrates accepted by UGT2B4, UGT2B7 is able to glucuronidate various steroid hormones (androsterone, epitestosterone) and fatty acids [7]. UGT2B4 and UGT2B7 therefore play a key role in the detoxification of cholestatic bile acids and may prevent the formation of proximal carcino- gens such as quinone estrogens. In addition, UGT2B7 is also able to conjugate major classes of drugs such as analgesics (morphine), carboxylic nonsteroidal anti- inflammatory drugs (ketoprofen) and anticarcinogens (all-trans retinoic acid). However, the molecular basis of the overlapping substrate specificity of these enzymes remains to be elucidated. Several studies have highlighted the role of the N-ter- minal domain of UGTs in substrate specificity, and many lines of evidence indicated that it may contain the major structural determinants for substrate recognition. The organization of the UGT1A complex locus suggests that the N-terminal part encoded by separate exons 1 governs the individual substrate specificity of each iso- form, whereas the identical C-terminal halves, encoded by exons 2–5, would interact with the common co-sub- strate, UDP-glucuronic acid [8]. In addition, Mackenzie [9] showed that exchanging the N-terminal half between two rat UGT2B isoforms, UGT2B2 and UGT2B3, resulted in a switch-over of their respective substrate selectivity. In agreement, Li et al. [10] showed that replacement of the C-terminal part of rabbit UGT2B16 with its counterpart in UGT2B13 did not change the specificity of this isoform. The aim of this study was to identify amino acid res- idues that are involved in substrate specificity of UGTs 2B4 and 2B7 in order to better understand the molecular basis of substrate recognition and catalysis by these enzymes. Attention was paid to amino acids at the N-terminal end of these UGTs, as this region is believed to interact with the substrates, although the contribution of the C-terminal part cannot be totally excluded. Mutation of phenylalanine at position 33 at the N-terminus of UGT2B4 was specifically carried out, as we have discovered that this residue was substi- tuted by leucine, in a UGT2B4 variant cDNA that was previously described by Jin et al. [11] to encode a UGT2B4 deficient in HDCA glucuronidation activity. As the phenylalanine residue at position 33 in the UGT2B4 isoform was replaced by tyrosine in UGT2B7, the mutation of this residue into leucine in UGT2B7 was also performed. We also mutated the phenylalanine 33 residue of UGT2B4 into the tyrosine residue found at the same position in UGT2B7 and carried out the corresponding mutations in UGT2B7, namely UGT2B7Y33L and UGT2B7Y33F. The results demonstrated the critical importance of an aromatic amino acid at position 33 for the activity and substrate specificity of both UGT2B4 and UGT2B7. Results The phenylalanine residue at position 33 of UGT2B4 is important for substrate specificity of the enzyme towards HDCA. Investigation of the deficiency in HDCA glucuronidation by the UGT2B4 variant described by Jin et al. [11] led to the discovery of the previously unreported mutation of phenylalanine resi- due 33 to leucine. Sequence alignment showed that all UGT2B members contained either a phenylalanine or tyrosine residue at this position (Fig. 1). In order to determine the effect of phenylalanine at position 33 on HDCA glucuronidation, this residue was replaced by leucine, creating the UGT2B4F33L mutant and expressed in baculovirus-infected insect cells. As illustra- ted in Fig. 2, immunoblot analysis of the membrane fraction of these cells showed that the full-length protein was produced. The expression level of each UGT in the current set of recombinant enzymes, mutants as well as wild-types, was determined by dot-blot analyses using monoclonal antibodies, as previously described [19]. The substrate specificity of UGT2B4F33L mutant was evaluated towards HDCA and a range of steroids and phenolic compounds in addition to carboxylic acids and was compared to that of the wild-type UGT2B4 (Fig. 3). The results confirmed that both 17-epiestriol Fig. 1. Sequence alignment of the region encompassing residue 33 of several UGTs of subfamily 2B. The alignment was performed using the program resident in GCG DNA and Protein Analysis Package (Promega, Madison, WI, USA). The fully conserved amino acids in this alignment are indicated by bold font. L. Barre et al. Substrate specificity in UGT2B4 and UGT2B7 FEBS Journal 274 (2007) 1256–1264 ª 2007 The Authors Journal compilation ª 2007 FEBS 1257 and HDCA were efficiently glucuronidated by this iso- form (Fig. 3A). In addition, we show here that UGT2B4 could also glucuronidate bulky and planar phenols (eugenol, 4-hydroxybiphenyl and 1-naphthol). In contrast, other steroids such as testosterone and 17a-ethynylestradiol were not accepted. The carboxylic nonsteroidal anti-inflammatory drug ketoprofen or the anti-HIV drug 3¢-azido-3¢-deoxythymidine were conju- gated at a very low rate (Fig. 3A). Altogether, the results of this substrate screening indicated that UGT2B4 is able to transfer glucuronic acid onto struc- turally diverse substrates, with a marked preference for 17-epiestriol, HDCA and phenolic substrates. The activity profile of the UGT2B4F33L mutant showed a selective change in substrate preference (Fig. 3B). Indeed, the mutant was unable to glucuroni- date HDCA, and its activity towards phenolic sub- strates, as well as the steroids 4-hydroxyestrone and 17-epiestriol was strongly affected (Fig. 3B). Apparent kinetic constants of the wild-type UGT2B4 and of the mutant were evaluated and V max values were normal- ized according to the level of protein expression (Table 1). In the case of 4-hydroxyestrone and 17-epi- estriol, the K m values of the mutant enzymes were increased by six- and two-fold, respectively, compared Fig. 2. Western blot analyses of the enzymes included in this study. The gels were loaded with 2, 10 or 100 lg of membrane proteins for UGT2B4, UGT2B4F33L and UGT2B4F33Y, respectively (A), or 12, 15 or 15 lg of membrane proteins for UGT2B7, UGT2- B7Y33L and UGT2B7Y33F, respectively (B). The UGTs were probed with primary antibody directed against the His-tag, and the second antibodies were horseradish-peroxidase-conjugated anti-mouse Ig. The blot was then developed using LumiGLO TM . A Enzyme activity (pmol/min/mg protein) 0 25 50 75 100 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 UGT2B4 Substrates 1 2 3 4 5 6 7 8 9 1011121314 1516 B Normalized enzyme activity (pmol/min/mg protein) UGT2B4F33L Substrates 0 0.5 1.0 1.5 2.0 123456789 10 11 12 13 14 15 16 C Substrates 0 25 50 75 100 Normalized enzyme activity (pmol/min/mg protein) UGT2B4F33Y Fig. 3. Glucuronidation activity of UGT2B4 (A) and UGT2B4 mutants (B, C) for the probe substrates. 1, 4-Methylumbelliferone; 2, euge- nol; 3, hyodeoxycholic acid (HDCA); 4, androsterone; 5, testoster- one; 6, epitestosterone; 7, b-estradiol; 8, 17a-ethynylestradiol; 9, 4- hydroxyestrone; 10, 17-epiestriol; 11, 4-hydroxybiphenyl; 12, 4-iso- propylphenol; 13, 4-nitrophenol; 14, 1-naphthol; 15, RS-ketoprofen; 16, 3¢-azido-3¢deoxytymidine. The enzyme reaction was carried out with 50 lg protein and was incubated with 0.02 m M UDP-glucuron- ic acid containing 0.1 lCi UDP-[ 14 C]glucuronic acid and 0.5 mM sub- strate as indicated in Experimental procedures. The glucuronides were separated by thin layer chromatography, visualized by auto- radiography (shown in insert) and quantitated by liquid scintillation counting. The rate values are the mean of three experiments. The film was exposed for four days in the case of UGT2B4 and UGT2B4F33Y and for one week in the case of UGT2B4Y33L. Substrate specificity in UGT2B4 and UGT2B7 L. Barre et al. 1258 FEBS Journal 274 (2007) 1256–1264 ª 2007 The Authors Journal compilation ª 2007 FEBS with wild-type UGT2B4. The V max values showed a decrease of 19- and 34-fold for 4-hydroxyestrone and 17a-epiestriol, respectively (Table 1). The primary structure of UGT2B7 is 87% identical to that of UGT2B4 with 76 differences out of 528 amino acids, including 55 differences in the first 300 amino acids of the N-terminus. Both enzymes share common substrates, including HDCA [6]. This led us to compare the N-terminal amino acid sequence of UGT2B4 and UGT2B7, predicted from its cDNA, in the region encompassing residue 33 (Fig. 1). The ana- lysis revealed that residue F33 of UGT2B4 was replaced by Y33 in UGT2B7. Therefore, we have also constructed and expressed in Sf9 cells a UGT2B4 mutant in which F33 was replaced by tyrosine, gener- ating the mutant UGT2B4F33Y (Fig. 2). Analysis of the glucuronidation activity of this mutant showed an activity profile similar to the wild-type UGT2B4. Moreover, HDCA and 4-hydroxyestrone were even more efficiently glucuronidated by the mutant (Fig. 3C, Table 1). Kinetic analysis indicated that the K m and V max values of UGT2B4F33Y towards HDCA and 4-hydroxyestrone were increased by 3.5- and two- fold, and by 4.6- and three-fold, respectively, com- pared with UGT2B4 (Table 1). These results led us to hypothesize that the aromatic tyrosine residue at posi- tion 33 in UGT2B4 may play an important role in the substrate specificity of the isoform. Importance of amino acid residue tyrosine 33 in the substrate specificity of UGT2B7 The wild-type UGT2B7 efficiently glucuronidates 17-epiestriol and eugenol and, in comparison with UGT2B4, it exhibited a marked preference for 4-hydroxyestrone and HDCA (Table 1). In addition, UGT2B7 efficiently glucuronidated androsterone and epitestosterone (Fig. 4A). To investigate whether the tyrosine residue at position 33 in UGT2B7 plays a role in HDCA glucuronidation and substrate specifi- city, we substituted this residue by leucine, as found in the HDCA-deficient UGT2B4 variant, and expressed the mutant in insect cells (Fig. 2). Analysis of the activity of the UGT2B7Y33L mutant towards various substrates showed that replacement of Y33 by leucine resulted in a dramatic change in activity and substrate specificity of UGT2B7 (Fig. 4, compare parts A and B). Indeed, the mutation abolished glucu- ronidation of several substrates including phenols such as 1-naphthol and steroids such as androsterone and b-estradiol (Fig. 4B), and greatly reduced the activity towards HDCA and 4-hydroxyestrone (Fig. 4B, Table 1). In addition, glucuronidation of bulky phen- ols, 4-hydroxybiphenyl and 4-isopropylphenol, and the steroid epitestosterone was dramatically decreased. On the other hand, the activity towards 17-epiestriol was increased by the Y33L mutation in UGT2B7 (Table 1). These data showed that the presence of a leucine resi- due at position 33, instead of tyrosine, led to an enzyme with restricted and somewhat modified specific- ity. Further kinetic characterization of this mutant indicated that the K m values towards HDCA and 17-epiestriol were in the same range as that of the wild- type. However, the K m value towards 4-hydroxyestrone was decreased by six-fold (Table 1). Furthermore, the V max values underwent major changes, with 20- and 25-fold decrease for HDCA and 4-hydroxyestrone, respectively, and 1.2-fold increase for 17-epiestriol. In contrast to leucine residue, replacement of tyro- sine by phenylalanine at position 33 of UGT2B7 had Table 1. Apparent K m and normalized V max values for glucuronidation of selected substrates by wild-type UGT2B4 and UGT2B7 and mutants. Kinetic parameters were evaluated from initial velocity values of the reaction performed in triplicates using varying concentrations of substrates (0–1 m M) at a constant concentration of UDP-glucuronic acid (0.5 mM). Expression of wild-type and mutants was evaluated as described in the Experimental procedures and expressed relative to UGT2B4 or UGT2B7. ND, not determined, due to lack of detectable activity. UGT HDCA 4-Hydroxy- estrone 17-Epiestriol Relative protein expression (%) V max (pmolÆmin )1 Æ mg )1 Æprotein) K m lM V max (pmolÆmin )1 Æ mg )1 Æprotein) K m lM V max (pmolÆmin )1 Æ mg )1 Æprotein) K m lM 2B4 26 ± 1 25 ± 3 19 ± 1 28 ± 5 276 ± 7 42 ± 5 100 2B4F33 L ND ND 1 173 ± 29 8 93 ± 12 44 2B4F33Y 48 ± 1 91 ± 10 57 ± 4 131 ± 26 182 ± 5 71 ± 7 4 2B7 1164 ± 36 21 ± 4 2365 ± 55 81 ± 8 570 ± 10 52 ± 4 100 2B7Y33 L 56 ± 2 29 ± 6 94 ± 3 12 ± 2 670 ± 17 45 ± 5 63 2B7Y33F 2156 ± 114 39 ± 9 1260 ± 33 117 ± 11 3283 ± 56 159 ± 8 60 L. Barre et al. Substrate specificity in UGT2B4 and UGT2B7 FEBS Journal 274 (2007) 1256–1264 ª 2007 The Authors Journal compilation ª 2007 FEBS 1259 only a minor effect on activity and substrate specifi- city. The mutant UGT2B7Y33F exhibited similar sub- strate specificity as wild-type UGT2B7 (Fig. 4C) and kinetic analysis indicated that the K m values towards HDCA and 17-epiestriol were increased by about two- and three-fold, respectively. The V max value towards 4-hydroxyestrone was decreased by two-fold and it was increased by two- and six-fold for HDCA and 17-epiestriol, respectively (Table 1). These experiments highlighted the importance of an aromatic residue at position 33 in the capacity of UGT2B7 to glucuroni- date a broad range of aglycone substrates. Discussion A major property of the UGTs is their large and over- lapping substrate specificity, which confers to glucuroni- dation a significant role in the detoxification processes. This characteristic feature is typically illustrated from comparison of the activity of UGT2B4 and UGT2B7, which are both able to glucuronidate HDCA and cate- chol-estrogens as well as xenobiotics, as shown in this and other studies [5]. However, UGT2B7 has a broader specificity than UGT2B4 and it is able to accommodate various steroids such as androsterone and epitestoster- one. The molecular basis of the substrate selectivity of these enzymes is difficult to understand because no com- mon structural features between the substrates glucuro- nidated by each isoform were thus far found [12]. This general assessment prompted us to identify amino acids that may account for the substrate specific- ity of these UGTs. The high sequence homology between UGT2B4 and UGT2B7, in combination with a marked difference in substrate specificity, especially towards steroid substrates, was favorable for attempt- ing to pinpoint the amino acid residues that are critical for the substrate specificity. In the current study, we have shown that the presence of an aromatic residue at position 33 of UGT2B4 and UGT2B7 is important in that respect. This conclusion is based on the following lines of evidence: (a) the UGT2B4F33L mutant exhib- ited a strong decrease in HDCA glucuronidation; (b) the UGT2B4F33Y mutant was able to sustain the glucuronidation of both HDCA and 4-hydroxyestrone; (c) mutation of residue Y33 of UGT2B7 to leucine led to an enzyme with a restricted substrate specificity; and (d) the mutant UGT2B7Y33F exhibited similar activity and substrate specificity to those of UGT2B7. Interest- ingly, Villeneuve et al. [13] recently reported a novel polymorphism of the UGT1A9 isoform, whose muta- tion M33T (corresponding to position 31 in UGT2B4) was responsible for a large decrease in the activity (by 96%) of the glucuronidation of the anticancer drug, SN-38. In contrast, the activity measured with flavo- piridol was unaffected, indicating that, similar to our findings, a single mutation can affect enzyme activity for a subset of aglycones substrates. The above study by Villeneuve et al. [13] and our work emphasize the crucial role of the region encompassing residue at posi- tion 33 in the substrate specificity of UGT isoforms. A Substrates 2 3 4 5 6 7 8 9 10 11 1213 1415 161 0 150 300 600 750 450 UGT2B7 Enzyme activity (pmol/min/mg protein) B Substrates UGT2B7Y33L 162 3 4 5 6 7 8 9 10 11 121314151 0 50 75 100 25 Normalized enzyme activity (pmol/min/mg protein) C Normalized enzyme activity (pmol/min/mg protein) UGT2B7Y33F 2 3 4 5 6 7 8 9 10 11 12 13 14 15 161 Substrates 0 600 1500 300 750 450 150 Fig. 4. Glucuronidation activity of UGT2B7 (A) and UGT2B7Y33L mutant (B) for probe substrates. Numbers refer to substrates as in Fig. 3. The insert shows the glucuronides separated by thin layer chromatography and visualized by autoradiography. (All the films were exposed for 4 days.) Activities were measured as indicated in the legend to Fig. 3 and are the mean of three experiments. Substrate specificity in UGT2B4 and UGT2B7 L. Barre et al. 1260 FEBS Journal 274 (2007) 1256–1264 ª 2007 The Authors Journal compilation ª 2007 FEBS The changes in specificity observed for the different mutants were characterized further by kinetic analyses. The results with UGT2B4F33L revealed that the impairment in 4-hydroxyestrone and 17-epiestriol glucu- ronidation efficacy resulted from a large increase in K m values, along with a decrease in the V max values. These data suggest that the mutations primarily affect binding of the substrates, but they do not rule out the possibility of a reduced access of the substrate to the catalytic site upon mutation. On the other hand, replacement of F33 by tyrosine led to mutant UGT2B4F33Y with similar substrate specificity as the wild-type enzyme support- ing the idea that a tyrosine can substitute to the wild-type phenylalanine residue. Moreover, mutant UGT2B4F33Y exhibited enhanced glucuronidation towards HDCA and 4-hydroxyestrone compared with wild-type. The kinetic parameters of the mutant indica- ted an increase in both V max and K m values (Table 1). In the case of UGT2B7, substitution of Y33 to leu- cine led to a severe restriction in aglycones accepted by the enzyme. In fact, the effects of replacing the aroma- tic residue at position 33 by leucine on the substrate specificity of UGT2B7 were even more dramatic than in UGT2B4. Only three out of the 12 compounds pre- viously glucuronidated by UGT2B7 remained effi- ciently glucuronidated by the UGT2B7Y33L mutant. Nonetheless, the K m value for HDCA was not signifi- cantly different from that obtained for the wild-type enzyme, suggesting that the affinity of the enzyme for HDCA was not largely altered by the mutation. In the case of 4-hydroxyestrone glucuronidation, the K m indi- cated an enhanced apparent affinity of the mutant. For both substrates, the mutation decreased the V max values. On the other hand, the V max of the mutant towards 17-epiestriol was slightly increased and the K m was not significantly modified. Replacement of the Y33 residue of UGT2B7 by phe- nylalanine led to a mutant, UGT2B7Y33F, with even more enhanced glucuronidation activity towards HDCA and 17-epiestriol compared with the wild-type. Analyses of the kinetic parameters of the UGT2- B7Y33F mutant indicated enhanced V max and K m values, except for 4-hydroxyestrone, which showed a two-fold decrease in the V m value (Table 1). Taken together, the results of this study are consis- tent with the notion that residue 33 is involved in the interactions of the enzyme with the substrate in both UGT2B4 and UGT2B7. In contrast to the F33L mutation, which reduces the activity of UGT2B4 and UGT2B7, exchanging F33 for tyrosine sustained the enzyme activity and specificity. Although a leucine residue can establish hydrophobic interactions, it will produce more steric hindrance than an aromatic residue such as phenylalanine or tyrosine. In agreement with this proposal, a tyrosine residue at position 33 in UGT2B4 was able to support glucuroni- dation of HDCA, thus suggesting that p-stacking interactions and ⁄ or steric hindrance conferred by an aromatic residue are critical for access or recognition of this substrate. Steric hindrance by a critical residue has been proposed as an underlying principle that can regulate substrate and ⁄ or product specificities of enzymes catalyzing the metabolism of hydrophobic substrates. For example, the phenylalanine residue at position 87 of cytochrome P450 BM-3 was suggested to act through steric hindrance to determine the regio- and stereospecificity of the arachidonic acid epoxy- genase activity [14]. Such a situation is also exemplified in the case of estrogen sulfotransferase, which posses- ses two critical aromatic residues forming a gate-like structure that was suggested to confer estrogen specifi- city to this enzyme [15]. The involvement of several residues in determining the substrate specificity probably also stands true for the UGTs. Coffman et al. [16] reported the important role of the aspartic residue at position 99 of UGT2B7 in the binding of morphine. When this charged amino acid was substituted with alanine, a dramatic decrease in activity was observed. In agreement, the structure– function analysis of UGT2B15 and UGT2B17 sugges- ted that a set of residues (including residue 121) is implicated in the steroid specificity of these isoenzymes [17]. These studies, along with our work, indicate that substitution of a single amino acid can substantially affect substrate recognition, but multiple differences between two related isoforms probably contribute to their individual specificity. In conclusion, this study shows, for the first time, that an aromatic residue at position 33 is critical for the sub- strate specificity of UGT2B4 and UGT2B7. The data provide the basis with which to modulate the substrate specificity of human UGT isoforms by protein engineering. Experimental procedures Chemicals and reagents 4-Methylumbelliferone (free acid), 1-naphthol, 4-nitro- phenol, 4-hydroxybiphenyl, 4-hydroxyestrone, 17a-ethinyl- estradiol, testosterone, eugenol, HDCA, androsterone, epitestosterone, b-estradiol, 17-epiestriol, isopropylphenol, ketoprofen, 3¢-azido-3¢-deoxythymidine and UDP-glucuro- nic acid (sodium salt) were purchased from Sigma (L’Isle d’Abeau, St. Quentin Fallavier, France). UDP-[U- 14 C]- glucuronic acid (418 mCiÆmmol )1 ) was obtained from NEN (Perkin Elmer, Courtaboeuf, France). Restriction enzymes L. Barre et al. Substrate specificity in UGT2B4 and UGT2B7 FEBS Journal 274 (2007) 1256–1264 ª 2007 The Authors Journal compilation ª 2007 FEBS 1261 were provided by New England Biolabs (Hitchin, UK). The QuikChange site-directed mutagenesis kit was from Strata- gene (La Jolla, CA, USA), LumiGLO TM was from Cell Signaling (Beverly, MA, USA), and AdvantageÒ 2 poly- merase mix was from Clontech (Palo Alto, CA, USA). All other reagents were of the best quality and commercially available. Expression vectors constructions Expression vectors used to express human UGT2B4 and UGT2B7 with an apparent molecular mass of about 53 kDa in baculovirus-infected insect cells were previously described [18]. The short C-terminal extension, including a His-tag, was added by subcloning the respective cDNAs into the modified shuttle vector pFBXHA to generate 2B4- XHA and 2B7-XHA expression vectors [18]. Site-directed mutagenesis Construction of amino acid substituted mutants of UGT2B4 and UGT2B7 were performed using the Quik- Change site-directed mutagenesis kit according to the recommendations of the manufacturer. 2B4-XHA and 2B7- XHA expression vectors were used as a template. The sequence of the sense and antisense mutation primers is indicated in Table 2. Full-length mutated cDNAs were sys- tematically checked by DNA sequencing. Heterologous expression in insect Sf9 cells Wild-type UGT2B4 and UGT2B7 and mutants expression vectors were transfected in the Escherichia coli strain DH10Bac for the generation of recombinant ‘bacmids’ that, in turn, were employed for the production of recom- binant baculovirus stocks according to the Bac-to-Bac procedure (Invitrogen, Cergy Pontoise, France). The pro- duction of recombinant proteins was carried out following optimization trials in which the suitable amount of virus from the new stocks for the infection of insect Sf9 cells was estimated. The relative expression level of each UGT in microsomal membranes was evaluated by immunodetection using the monoclonal anti-His-tag anti- body Tetra-His (Qiagen, Hilden, Germany) as described previously [19]. Western blot analysis was performed by loading onto the gel 2, 10 and 100 lg of membrane proteins for UGT2B4, UGT2B4F33L and UGT2B4F33Y, respectively, and 12, 15 and 15 lg of membrane proteins for UGT2B7, UGT2- B7Y33L and UGT2B7Y33F, respectively. The proteins were separated in 10% SDS ⁄ PAGE gels, transferred to a polyvinylidene difluoride membrane (Millipore, Eschborn, Germany), and subsequently blocked in Tris-buffer saline- Tween 20 containing 5% nonfat milk. Membranes were incubated overnight with monoclonal anti-His-tag antibody Tetra-His directed against His-tag followed by incubation with horseradish-peroxidase-conjugated secondary antibod- ies. The blot was then developed using LumiGLO TM according to the instructions of the manufacturer (Cell Signaling, Danvers, MA, USA). Analysis of glucuronidation activity Protein concentration was measured as previously described [20] with the Bio-Radä reagent (Bio-Rad, Hercules, CA, USA). The activity of the recombinant wild-type and mutant UGT2B4 and UGT2B7 towards several substrates was determined as described [21]. Briefly, incubation in Eppendorf tubes (total volume 40 lL) consisted of 50 lgof microsomal proteins for UGT2B7, UGT2B7Y33L, UGT2- B7Y33F and UGT2B4 and 200 l g for UGT2B4F33Y and UGT2B4F33L in 100 mm Tris ⁄ HCl buffer (pH 7.4), 10 mm MgCl 2 containing 0.02 mm UDP-glucuronic acid and 0.1 lCi UDP-[U- 14 C]glucuronic acid. The reaction was star- ted by addition of substrate (0.5 mm final concentration) dissolved in 2 lL dimethylsulfoxide. A control was per- formed in which the substrate was omitted and dimethyl- sulfoxide added. After incubation for 1 h at 37 ° C, the proteins were precipitated by 40 lL ethanol in ice, and removed by centrifugation at 4000 g for 10 min at 4 °C. The supernatant was loaded onto thin layer chromato- graphy plates (LK6DF silica gel, 250 lm; Whatman, Clif- ton, NJ, USA). The plates were developed with n-butanol, acetone, acetic acid, aqueous ammoniac (28%), water Table 2. Sequence of the primers used for site-directed mutagenesis. Mutant amino acid codons are underlined. Mutant Primer Sequence (5’- to 3’) 2B4F33 L Sense CTGGTGTGGCCCACAGAA CTCAGCCACTGGATGAATATAAAG Antisense CTTTATATTCATCCAGTGGCT GAGTTCTGTGGGCCACACCAG 2B4F33Y Sense CTGGTGTGGCCCACAGAA TACAGCCACTGGATGAATATAAAG Antisense CTTTATATTCATCCAGTGGCT GTATTCTGTGGGCCACACCAG 2B7Y33 L Sense CTGGTGTGGGCAGCAGAA CTCAGCCATTGGATGAATATAAAG Antisense CTTTATATTCATCCAATGGCT GAGTTCTGCTGCCCACACCAG 2B7Y33F Sense CTGGTGTGGGCAGCAGAA TACAGCCATTGGATGAATATAAAG Antisense CTTTATATTCATCCAATGGCT GTATTCTGCTGCCCACACCAG Substrate specificity in UGT2B4 and UGT2B7 L. Barre et al. 1262 FEBS Journal 274 (2007) 1256–1264 ª 2007 The Authors Journal compilation ª 2007 FEBS (70 : 50 : 18 : 1.5 : 60 v ⁄ v). They were dried and sprayed with 1% (v ⁄ v) 2-(4-t-butylphenyl)-5() 4-biphenyl)-1,3, 4-oxadiazole in toluene. The radioactivity associated with the glucuronide was visualized by autoradiography with X-Omat Kodak films (Sigma) for 3 days at )20 °C. The sil- ica gel areas of the glucuronides were scraped off and the associated radioactivity was quantified on a LKB spectro- meter using Fluoran Safe Ultima Gold scintillant cocktail (Packard, Rungis, France). The decomposition per min value in a given sample was considered significant when it was at least two-fold of that of the blank sample. Kinetic analysis of the data Kinetic parameters were evaluated from initial velocity val- ues of the reaction performed as described above. Varying concentrations of the substrates (0–1 mm) at a constant concentration of UDP-GlcA (0.5 mm) were used. K m and V max values for HDCA, 4-hydroxyestrone and 17a-epiestriol were determined using nonlinear least square analysis of the data fitted to Michaelis-Menten rate equation (v ¼ V max · [S] ⁄ K m + [S]), where S is the substrate and v is the velocity, using the curve-fitter program sigmaplot 9.0 [22]. Acknowledgements This work was supported by grants from Ligue Contre le Cancer Re ´ gion Lorraine, Agence Nationale de la Recherche (ANR number NT05-3_42251) and Re ´ gion Lorraine, as well as the Academy of Finland (Project 210933). We thank J, Mosorin for excellent technical assistance and PI. Mackenzie (Flinders University, Adelaide, Australia) for kindly providing the UGT2B4 variant cDNA. References 1 Clarke DJ & Burchell B (1994) The Uridine Diphosphate Glucuronosyltransferase Multigene Family: Function and Regulation Springer-Verlag, Berlin, Heidelberg, New York. 2 Tukey RH & Strassburg CP (2000) Human UDP- glucuronosyltransferases: metabolism, expression, and disease. Annu Rev Pharmacol Toxicol 40 , 581–616. 3 Guengerich FP, Parikh A, Johnson EF, Richardson TH, von Wachenfeldt C, Cosme J, Jung F, Strassburg CP, Manns MP, Tukey RH et al. (1997) Heterologous expression of human drug-metabolizing enzymes. Drug Metab Dispos 25, 1234–1241. 4 Radominska-Pandya A, Czernik PJ, Little JM, Battaglia E & Mackenzie PI (1999) Structural and functional stu- dies of UDP-glucuronosyltransferases. Drug Metab Rev 31, 817–899. 5 Fournel-Gigleux S, Jackson MR, Wooster R & Burchell B (1989) Expression of a human liver cDNA encoding a UDP-glucuronosyltransferase catalysing the glucuroni- dation of hyodeoxycholic acid in cell culture. FEBS Lett 243, 119–122. 6 Ritter JK, Chen F, Sheen YY, Lubert RA & Owens IS (1992) Two human liver cDNAs encode UDP-glucurono- syltransferases with 2 log differences in activity toward parallel substrates including hyodeoxycholic acid and certain estrogen derivatives. Biochemistry 31, 3409–3414. 7 Radominska-Pandya A, Little JM & Czernik PJ (2001) Human UDP-glucuronosyltransferase 2B7. Curr Drug Metab 2, 283–298. 8 Ritter JK, Chen F, Sheen YY, Tran HM, Kimura S, Yeatman MT & Owen IS (1992) A novel complex locus UGT1 encodes human bilirubin, phenol, and other UDP-glucuronosyltransferase isozymes with identical carboxyl termini. J Biol Chem 267, 3257–3261. 9 Mackenzie PI (1990) Expression of chimeric cDNAs in cell culture defines a region of UDP-glucuronosyltrans- ferase involved in substrate selection. J Biol Chem 265, 3432–3435. 10 Li Q, Lou X, Peyronneau MA, Obermayer-Straub P & Tukey RH (1997) Expression and functional domains of rabbit liver UDP-glucuronosyltransferase 2B16 and 2B13. J Biol Chem 272, 3272–3279. 11 Jin C, Miners JO, Lillywhite KJ & Mackenzie PI (1993) Complementary deoxyribonucleic acid cloning and expression of human liver uridine diphosphate-glucuro- nosyltransferase glucuronidating carboxylic acid-con- taining drugs. J Pharm Exp Ther 264, 475–479. 12 Jin CJ, Mackenzie PI & Miners JO (1997) The regio- and stereo-selectivity of C19 and C21 hydroxysteroid glucuronidation by UGT2B7 and UGT2B11. Arch Bio- chem Biophys 341, 207–211. 13 Villeneuve L, Girard H, Fortier LC, Gagne JF & Guil- lemette C (2003) Novel functional polymorphisms in the UGT1A7 and UGT1A9 glucuronidating enzymes in Caucasian and African-American subjects and their impact on the metabolism of 7-ethyl-10-hydroxycamp- tothecin and flavopiridol anticancer drugs. J Pharmacol Exp Ther 307, 117–128. 14 Graham-Lorence S, Truan G, Peterson JA, Falck JR, Wei S, Helvig C & Capdevila JH (1997) An active site substitution, F87V, converts cytochrome P450 BM-3 into a regio- and stereoselective (14S,15R)-arachidonic acid epoxygenase. J Biol Chem 272, 1127–1135. 15 Petrotchenko EV, Doerflein ME, Kakuta Y, Pedersen LC & Negishi M (1999) Substrate gating confers steroid specificity to estrogen sulfotransferase. J Biol Chem 274, 30019–30022. 16 Coffman BL, Kearney WR, Goldsmith S, Knosp BM & Tephly TR (2003) Opioids bind to the amino acids 84–118 of UDP-glucuronosyltransferase UGT2B7. Mol Pharmacol 63, 283–288. L. Barre et al. Substrate specificity in UGT2B4 and UGT2B7 FEBS Journal 274 (2007) 1256–1264 ª 2007 The Authors Journal compilation ª 2007 FEBS 1263 17 Dubois SG, Beaulieu M, Levesque E, Hum DW & Belanger A (1999) Alteration of human UDP-glucuro- nosyltransferase UGT2B17 regio-specificity by a single amino acid substitution. J Mol Biol 289, 29–39. 18 Kurkela M, Garcia-Horsman JA, Luukkanen L, Morsky S, Taskinen J, Baumann M, Kostiainen R, Hirvonen J & Finel M (2003) Expression and character- ization of recombinant human UDP-glucuronosyltrans- ferases (UGTs). UGT1A9 is more resistant to detergent inhibition than other UGTs and was purified as an active dimeric enzyme. J Biol Chem 278, 3536–3544. 19 Kurkela M, Hirvonen J, Kostiainen R & Finel M (2004) The interactions between the N-terminal and C-terminal domains of the human UDP-glucuronosyl- transferases are partly isoform-specific, and may involve both monomers. Biochem Pharmacol 68, 2443–2450. 20 Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein util- izing the principle of protein-dye binding. Anal Biochem 72, 248–254. 21 Bansal SK & Gessner T (1980) A unified method for the assay of uridine diphospho-glucuronyltransferase activity toward various aglycones using uridine diphospho[U- 14 C]glucuronic acid. Anal Biochem 109, 321–329. 22 Segel IH (1975) Enzyme Kinetics. John Wiley & Sons, New York, NY. Substrate specificity in UGT2B4 and UGT2B7 L. Barre et al. 1264 FEBS Journal 274 (2007) 1256–1264 ª 2007 The Authors Journal compilation ª 2007 FEBS . CTTTATATTCATCCAGTGGCT GAGTTCTGTGGGCCACACCAG 2B4F33Y Sense CTGGTGTGGCCCACAGAA TACAGCCACTGGATGAATATAAAG Antisense CTTTATATTCATCCAGTGGCT GTATTCTGTGGGCCACACCAG 2B7Y33. CTTTATATTCATCCAGTGGCT GTATTCTGTGGGCCACACCAG 2B7Y33 L Sense CTGGTGTGGGCAGCAGAA CTCAGCCATTGGATGAATATAAAG Antisense CTTTATATTCATCCAATGGCT GAGTTCTGCTGCCCACACCAG 2B7Y33F Sense CTGGTGTGGGCAGCAGAA TACAGCCATTGGATGAATATAAAG Antisense