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

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:
(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-

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.

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-

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-

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

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

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

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)

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

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–

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’

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

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,

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.

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