The human UDP-glucuronosyltransferase UGT1A3 is highly selective towards N2 in the tetrazole ring of losartan, candesartan, and zolarsartan

Article abstract of DOI:10.1016/j.bcp.2008.07.006

Losartan, candesartan, and zolarsartan are AT₁ receptor antagonists that inhibit the effect of angiotensin II. We have examined their glucuronidation by liver microsomes from several animals and by recombinant human UDP-glucuronosyltransferases (UGTs). Large differences in the production of different glucuronide regioisomers of the three sartans were observed among liver microsomes from human (HLM), rabbit, rat, pig, moose, and bovine. However, all the liver microsomes produced one or two N-glucuronides in which either N1 or N2 of the tetrazole ring were conjugated. O-Glucuronides were also detected, including acyl glucuronides of zolarsartan and candesartan. Examination of individual human UGTs of subfamilies 1A and 2B revealed that N-glucuronidation activity is widespread, along with variable regioselectivity with respect to the tetrazole nitrogens of these sartans. Interestingly, UGT1A3 exhibited a strong regioselectivity towards the N2 position of the tetrazole ring in all three sartans. Moreover, the tetrazole-N2 of zolarsartan was only conjugated by UGT1A3, whereas the tetrazole-N1 of this aglycone was accessible to other enzymes, including UGT1A5. Zolarsartan O-glucuronide was mainly produced by UGTs 1A10 and 2B7. UGT2B7, alongside UGT1A3, glucuronidated candesartan at the tetrazole-N2 position, whereas UGTs 1A7-1A10 mainly yielded candesartan O-glucuronide. In the case of losartan, no O-glucuronide was generated by any tested human enzyme. Nevertheless, UGTs 1A1, 1A3, 1A10, 2B7, and 2B17 glucuronidated losartan at the tetrazole-N2, while UGT1A10 also yielded the respective N1-glucuronide. Kinetic analyses revealed that the main contributors to losartan glucuronidation in HLM are UGT1A1 and UGT2B7. The results provide ample new data on substrate specificity in drug glucuronidation.

Full text of DOI:10.1016/j.bcp.2008.07.006

b i o c h e m i c a l p h a r m a c o l o g y 7 6 ( 2 0 0 8 ) 7 6 3 – 7 7 2
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/biochempharm
The human UDP-glucuronosyltransferase UGT1A3 is highly
selective towards N2 in the tetrazole ring of losartan,
candesartan, and zolarsartan
a
Anna Alonen a, Moshe Finel b, , Risto Kostiainen
*
a Faculty of Pharmacy, Division of Pharmaceutical Chemistry, P.O. Box 56 (Viikinkaari 5), FIN-00014 University of Helsinki, Finland
b Faculty of Pharmacy, Center for Drug Research (CDR), P.O. Box 56 (Viikinkaari 5), FIN-00014 University of Helsinki, Finland
a r t i c l e i n f o
a b s t r a c t
Article history:
Losartan, candesartan, and zolarsartan are AT1 receptor antagonists that inhibit the effect of
angiotensin II. We have examined their glucuronidation by liver microsomes from several
animals and by recombinant human UDP-glucuronosyltransferases (UGTs). Large differ-
ences in the production of different glucuronide regioisomers of the three sartans were
observed among liver microsomes from human (HLM), rabbit, rat, pig, moose, and bovine.
However, all the liver microsomes produced one or two N-glucuronides in which either N1 or
N2 of the tetrazole ring were conjugated. O-Glucuronides were also detected, including acyl
glucuronides of zolarsartan and candesartan. Examination of individual human UGTs of
subfamilies 1A and 2B revealed that N-glucuronidation activity is widespread, along with
variable regioselectivity with respect to the tetrazole nitrogens of these sartans. Interest-
ingly, UGT1A3 exhibited a strong regioselectivity towards the N2 position of the tetrazole
ring in all three sartans. Moreover, the tetrazole-N2 of zolarsartan was only conjugated by
UGT1A3, whereas the tetrazole-N1 of this aglycone was accessible to other enzymes,
including UGT1A5. Zolarsartan O-glucuronide was mainly produced by UGTs 1A10 and
2B7. UGT2B7, alongside UGT1A3, glucuronidated candesartan at the tetrazole-N2 position,
whereas UGTs 1A7–1A10 mainly yielded candesartan O-glucuronide. In the case of losartan,
no O-glucuronide was generated by any tested human enzyme. Nevertheless, UGTs 1A1,
1A3, 1A10, 2B7, and 2B17 glucuronidated losartan at the tetrazole-N2, while UGT1A10 also
yielded the respective N1-glucuronide. Kinetic analyses revealed that the main contributors
to losartan glucuronidation in HLM are UGT1A1 and UGT2B7. The results provide ample new
data on substrate specificity in drug glucuronidation.
Received 12 May 2008
Accepted 1 July 2008
Keywords:
Losartan
Candesartan
Zolarsartan (GR117289)
UDP-glucuronosyltransferases
Substrate specificity
Tetrazole
# 2008 Elsevier Inc. All rights reserved.
1.
Introduction
tan is also an active compound both in vitro [5] and in vivo [6].
Losartan itself is not as active as its carboxylic acid metabolite,
Losartan, candesartan, and zolarsartan (GR117289, also called
zolasartan) inhibit the effect of angiotensin II by acting as AT1
receptor antagonists. Losartan and candesartan are in clinical
use for the treatment of hypertension. In addition, it was
recently reported that losartan might be effective in treating
Marfan syndrome and perhaps other illnesses [1–4]. Zolarsar-
which is also longer acting. Candesartan is administered as a
prodrug, candesartan cilexetil, which has better bioavailability
than candesartan. Candesartan cilexetil is hydrolyzed to the
active candesartan during the absorption process. The
molecular structures of losartan, candesartan, and zolarsar-
tan include
a tetrazole moiety, while candesartan and
* Corresponding author. Tel.: +358 9 19159193; fax: +358 9 19159556.
E-mail address: moshe.finel@helsinki.fi (M. Finel).
0006-2952/$ – see front matter # 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.bcp.2008.07.006

764
b i o c h e m i c a l p h a r m a c o l o g y 7 6 ( 2 0 0 8 ) 7 6 3 – 7 7 2
Fig. 1 – Structures of losartan, candesartan, and zolarsartan. The arrows indicate glucuronidation sites and the abbreviation
for the regioisomer.
zolarsartan also have a carboxylic acid group (Fig. 1). Losartan
is metabolized by three main routes: oxidation of the alcohol
to carboxylic acid, hydroxylation, and glucuronidation [7].
Candesartan is mostly excreted unchanged [8], but glucur-
onide conjugates have been found from rats and dogs [9,10].
Rats and dogs also produce zolarsartan glucuronides [11].
Glucuronidation is an important phase II metabolic
reaction. UDP-glucuronosyltransferases (UGTs) are mem-
brane bound enzymes that catalyze this conjugation reaction
[12–16]. Glucuronidation is an SN2 reaction where the
nucleophilic substrate attacks the uridine-50-diphospho-a-D-
glucuronic acid (UDPGA). The most common glucuronidation
products are O- and N-glucuronides formed from hydroxyl,
carboxyl, or amino functional groups, but there are also
reports of S- and C-glucuronides [12,16,17].
glucuronides may be formed [20] (Fig. 1). The O-glucuronide
and one of the N-glucuronides, tetrazole-N2-glucuronide,
were found to be formed during biotransformation and
tetrazole-N1-glucuronide has been synthesized both chemi-
cally [20] and enzymatically [21]. The glucuronidation of
losartan has been previously studied using liver slices [7,20]
and liver microsomes [22] from various species, as well as rat
intestine [23]. Monkeys and rats produced both O-glucuronide
and tetrazole-N2-glucuronide whereas humans and dogs
produced only the tetrazole-N2-glucuronide of losartan. The
metabolism of candesartan cilexetil has been studied in rats
and dogs, yielding both the O-glucuronide (acyl glucuronide)
and the tetrazole-N2-glucuronide of candesartan [9,10] (Fig. 1).
Tetrazole-N1-glucuronides of candesartan have not thus far
been found. Zolarsartan can be glucuronidated to O-glucur-
onide in rats, while dogs, in addition to acyl glucuronide, also
produce tetrazole-N2-glucuronide [11] (Fig. 1). Zolarsartan N1-
glucuronide has been synthesized enzymatically using rat
liver microsomes as a catalyst [21].
The human UGTs are divided into three subfamilies,
UGT1A, UGT2A, and UGT2B, based on their amino acid
sequences and gene structures [14,18]. UGTs catalyze the
glucuronidation of a wide range of endogenous and exogenous
compounds, including steroid hormones, dietary constitu-
ents, and drugs. The substrate specificity of most UGTs is wide
and often partly overlapping. Most human UGTs catalyze the
glucuronidation of hydroxyl or carboxyl groups in the form of
ether or acyl O-glucuronides, whereas N-glucuronides, formed
from various amines, are produced by much fewer UGTs
[12,16,19].
The aim of this study was to examine the glucuronidation
of the aglycones losartan, candesartan, and zolarsartan by the
human UGTs of subfamilies 1A and 2B. We have selected the
three substrates on the basis of similarity of structure, i.e. all
contain a tetrazole, and in addition both O- and N-glucur-
onides can be formed without a phase I metabolism reaction.
Among other things, we wanted to find out if the tetrazole
group, common to all three substrates, was glucuronidated by
the same UGT isoform(s). In addition, liver microsomes from
Losartan can be glucuronidated at three different positions
so that one O-glucuronide and two different tetrazole-N-

b i o c h e m i c a l p h a r m a c o l o g y 7 6 ( 2 0 0 8 ) 7 6 3 – 7 7 2
765
human, bovine, moose, pig, rabbit, and rat were studied to
explore species differences in glucuronidation of sartans.
respect to both parameters. Otherwise the reactions were
carried out similarly to the screening assays described above.
The kinetic parameters were obtained by fitting the observed
velocities to kinetic models using GraphPad Prism 4.02 for
Windows (GraphPad Software Inc., San Diego, CA). The kinetic
models were Michaelis–Menten (v ¼ Vmax½Sꢀ=ðKm þ ½SꢀÞ) and
2.
Materials and methods
2.1.
Reagents
substrate
inhibition
equation
(v ¼ Vmax=ð1 þ ðKm=½SꢀÞþ
ð½Sꢀ=KsiÞÞ).
Losartan potassium (98% purity), candesartan (purity not
reported), and zolarsartan (purity not reported) were obtained
2.3.
Analytics
from Merck (Rahway, NJ), AstraZeneca (Molndal, Sweden), and
¨
GlaxoSmithKline (Hertfordshire, UK), respectively. Glucuro-
nide regioisomers of losartan, candesartan, and zolarsartan
were synthesized previously in our laboratory and their
structures were characterized by nuclear magnetic resonance
spectroscopy (NMR) [21]. Saccharic acid 1,4-lactone and
UDPGA (trisodium salt, 99.8% purity) were purchased from
Sigma (St. Louis, MO). Analytical or high-performance liquid
chromatography (HPLC) grade reagents and solvents were
used and they were purchased from J.T. Baker (Deventer,
Holland; acetic acid), Fluka (Steinheim, Germany; disodium
hydrogen phosphate dihydrate), Merck (Darmstadt, Germany;
magnesium chloride hexahydrate and perchloric acid), Rath-
Reactions containing losartan as the aglycone were termi-
nated by adding 25 mL cold 4 M perchloric acid and transferring
the samples to ice for 10 min. The denatured proteins were
then removed by centrifugation (10 min, 16,100 ꢁ g). The
reactions with candesartan and zolarsartan were terminated
by transfer to ice for 10 min (perchloric acid was not used since
the compounds would have precipitated under acidic condi-
tions) followed by centrifugation (10 min, 16,100 ꢁ g) and
subsequent ultrafiltration of 200 mL of the supernatants
(Ultrafree-MC 30,000 NMWL, Millipore, Bedford, MA; 20 min,
5000 ꢁ g) to remove residual proteins. The supernatants of the
losartan reactions and filtrates of candesartan and zolarsartan
reactions were analyzed by an Agilent 1100 Series HPLC
equipped with an autosampler and the Agilent 1100 Series
fluorescence or UV multiwavelength detector (Waldbronn,
Germany). Details of the analytical HPLC conditions are given
in Table 1.
burn (Walkerburn, UK; acetonitrile, ACN), Riedel-de Haen
¨
(Seelze, Germany; ammonium acetate, dimethyl sulfoxide,
formic acid, and potassium dihydrogen phosphate). Water
was purified with a Milli-Q water purification system (Milli-
pore, Molsheim, France). The human recombinant UGTs 1A1,
1A3–1A10, 2B4, 2B7, 2B10, 2B15, 2B17, and 2B28 were expressed
in baculovirus-infected insect cells [24]. Pooled human liver
microsomes (HLMs) and male New Zealand White rabbit liver
microsomes were purchased from BD Gentest (Woburn, MA)
and In Vitro Technologies (Baltimore, MD), respectively. Rat
liver microsomes were prepared from Aroclor 1254-induced
male Sprague–Dawley rats, and pig, moose, and bovine liver
microsomes were prepared from untreated animals, as
previously described [25]. A BCA Protein Assay Kit (Pierce
Chemical, Rockford, IL) was used to measure the protein
concentrations of UGTs and microsomes. The relative expres-
sion level of each recombinant UGT was determined by
dot-blot analyses using anti-His-tag antibodies, as detailed
elsewhere [24].
The HPLC methods were validated with respect to
specificity and repeatability. Method specificity was deter-
mined by comparing incubation samples to three different
controls prepared either without UDPGA, protein, or aglycone.
Repeatability of the HPLC was assessed by injecting incubation
samples, which were diluted to two different concentrations,
four times for both concentrations. The detection limit in the
screening assays and the quantitation limit in the kinetic
analyses were based on a signal-to-noise ratio of 3 and 9,
respectively. In kinetic studies the quantitation was per-
formed using an authentic standard, losartan N2-glucuronide.
The retention times of the different glucuronide regioisomers
were identified with glucuronide standards that were recently
synthesized in our laboratory [21], as well as by LC–MS. The
mass spectrometric analysis of losartan glucuronides was
carried out on a PerkinElmerSciex API 3000 triple quadrupole
mass spectrometer (MDS Sciex, Toronto, ON, Canada).
Glucuronides of candesartan and zolarsartan were analyzed
by a quadrupole time-of-flight mass spectrometer (Q-Tof
Micro, Waters/Micromass, Manchester, UK). Both apparatuses
were equipped with an ionspray source. Analyses were
performed in positive ion mode with a capillary voltage of
5000 V for losartan glucuronides and 3000 V for candesartan
and zolarsartan glucuronides.
2.2.
Glucuronidation assays
Screening assays of recombinant UGTs and liver microsomes
were carried out in 250 mL reactions containing 1 mM
substrate (losartan, candesartan, or zolarsartan), 5 mM sac-
charic acid 1,4-lactone, 5 mM UDPGA, 5 mM MgCl2, 50 mM
phosphate buffer (pH 7.4), and 625 mg protein (recombinant
UGT membrane or microsome). Substrates were added as
dimethyl sulfoxide solutions so that the final solvent
concentration was 2% of the reaction volume. All samples
were prepared in duplicate except respective controls without
UDPGA. Samples were incubated at +37 8C for 4 h.
3.
Results
Kinetic analyses were carried out, in triplicates, with HLM
and the three most active UGTs in the formation of losartan
tetrazole-N2-glucuronide, UGT1A1, UGT1A3, and UGT2B7. The
samples contained 2–600 mM losartan. The protein concentra-
tion and incubation time were within the linear range with
We have recently biosynthesized and characterized by NMR
the different glucuronides of the three sartans that are
included in the current study [21]. The glucuronidation assays
used here were carried out using a validated HPLC method.

766
b i o c h e m i c a l p h a r m a c o l o g y 7 6 ( 2 0 0 8 ) 7 6 3 – 7 7 2
Table 1 – HPLC methods for analysis of losartan, candesartan, zolarsartan, and their glucuronidesa
Losartan
Candesartan
Zolarsartan
Column
HP Hypersil BDS C18 5 mm
150 mm ꢁ 4.6 mm
HP Hypersil BDS C18 5 mm
250 mm ꢁ 4.0 mm
HP Hypersil BDS C18 5 mm
250 mm ꢁ 4.0 mm
Eluents A and B
Eluent A:B
20 mM NH4Ac pH 4.5—ACN
1% CH3COOH/H2O—ACN
0.1% HCOOH/H2O—ACN
73:27
Gradient (% B)
Gradient (% B)
0–7 min 28%, 7–25 min 28–55%
25–30 min 55%
0–9 min 36%, 9–14 min 36–40%
14–21 min 40%, 21–23 min 40–80%
23–26 min 80%, 26–27 min 80–36%
27–42 min 36%
30–31 min 55–28%
31–46 min 28%
Flow rate (mL/min)
Injection volume (mL)
Detector
1 (40 8C)
40
1 (25 8C)
1 (40 8C)
30
40
UV 256 nm
12.0 min
Fluorescence Ex 260 nm, Em 395 nm
17.4 min
Fluorescence Ex 306 nm, Em 428 nm
24.7 min
Retention time of
aglycone
Retention time of
glucuronides
3.6 min (LOG),
4.1 min (LN1G),
8.5 min (LN2G)
10.4 min (COG), 15.2 min (CN2G)
10.6 min (ZN1G), 12.9 min (ZOG),
19.1 min (ZN2G)
a
LOG, losartan O-glucuronide; LN1G, losartan N1-glucuronide; LN2G, losartan N2-glucuronide; COG, candesartan O-glucuronide; CN2G,
candesartan N2-glucuronide; ZN1G, zolarsartan N1-glucuronide; ZOG, zolarsartan O-glucuronide; ZN2G, zolarsartan N2-glucuronide.
The results of the repeatability experiments are presented in
Table 2, and they demonstrated that this method gave reliable
analyses of the samples. In particular, the losartan N2-
glucuronide standard curve that was used for quantitation
in the kinetic assays, was linear between 0.5 and 20 mM
(R = 0.99971). The LC–MS analyses confirmed the analyte peaks
to be sartan glucuronides, since the m/z was 176 Da higher
than for the corresponding aglycones. The identification of
glucuronide regioisomers was based on similar retention
times to the glucuronide standards that were earlier char-
acterized by NMR [21]. Chromatograms of losartan, cande-
sartan, and their glucuronides are shown in Fig. 2.
and 2B17 (Fig. 3). UGT1A10 also catalyzed the formation of
tetrazole-N1-glucuronide at significant rates. Very small
amounts of tetrazole-N1-glucuronide were detected in the
reactions of UGTs 1A1, 1A3, 1A7, and 1A8, whereas very small
amounts of tetrazole-N2-glucuronide were found in the
reactions of UGTs 1A4, 1A6–1A9, and 2B4 (Fig. 3).
Glucuronidation of candesartan by human UGTs yielded
two metabolites, identified as O-glucuronide and tetrazole-N2-
glucuronide. The main candesartan glucuronide produced by
UGTs 1A7–1A10 was the O-glucuronide, while 1A3 and 2B7
produced primarily tetrazole-N2-glucuronide (Fig. 3). In addi-
tion, the O-glucuronide of candesartan was produced by 1A1,
1A3, 1A4, and 2B7, as was tetrazole-N2-glucuronide by 1A1,
1A7–1A10, 2B4, and 2B17.
Glucuronidation of losartan, candesartan, and zolarsartan
was studied using recombinant human UGTs and liver
microsomes from various species. Analyses of losartan
glucuronidation revealed that of the three potential glucur-
onides of the drug, human UGTs produced only two, tetrazole-
N1-glucuronide and tetrazole-N2-glucuronide. The main
metabolite was tetrazole-N2-glucuronide of losartan and the
most active human UGT isoforms were 1A3, 2B7, 1A10, 1A1,
In the case of zolarsartan, human UGTs produced three
different glucuronides: O-glucuronide, tetrazole-N1-glucuro-
nide, and tetrazole-N2-glucuronide. UGT1A3 mainly catalyzed
the formation of tetrazole-N2-glucuronide, as well as smaller
amounts of O-glucuronide (Fig. 3). UGT1A10 produced mainly
O-glucuronide plus tetrazole-N1-glucuronide. Interestingly,
Table 2 – Repeatability of the HPLC methods for the analysis of sartan glucuronidesa
Repeatability (%R.S.D.)
Concentration 1 (n = 4)
Concentration 2 (n = 4)
Retention time
Area
Retention time
Area
LOG
0.7
0.6
0.1
2.5
2.3
2.8
0.3
2.8
0.1
0.1
0.03
0.4
0.1
0.2
0.2
0.1
0.4
1.3
0.1
1.2
0.8
0.7
0.8
1.8
0.1
0.1
0.04
0.1
0.04
0.1
0.1
0.1
LN1G
LN2G
COG
CN2G
ZOG
ZN1G
ZN2G
a
LOG, losartan O-glucuronide; LN1G, losartan N1-glucuronide; LN2G, losartan N2-glucuronide; COG, candesartan O-glucuronide; CN2G,
candesartan N2-glucuronide; ZOG, zolarsartan O-glucuronide; ZN1G, zolarsartan N1-glucuronide; ZN2G, zolarsartan N2-glucuronide.

b i o c h e m i c a l p h a r m a c o l o g y 7 6 ( 2 0 0 8 ) 7 6 3 – 7 7 2
767
Fig. 2 – Chromatographic separation of losartan glucuronides (A), candesartan glucuronides (B), and zolarsartan
glucuronides (C). The chromatograms are from screening assays in which the substrate concentration was 1 mM and the
protein concentration was 2.5 mg/mL. The enzyme sources were rat liver microsomes in (A), recombinant human UGT1A3
in (B), and pig liver microsomes in (C) zolarsartan.
UGTs 1A5 and 2B4 selectively and fairly effectively produced
tetrazole-N1-glucuronide of zolarsartan, whereas UGT2B7 did
the same for O-glucuronide. Traces of tetrazole-N1-glucur-
onide of zolarsartan were detected in the reactions of 1A1, 1A6,
1A7, and 1A8. UGTs 1A1, 1A3, 1A7, and 1A8 formed small
amounts of zolarsartan O-glucuronide.
amounts of losartan O-glucuronide were produced by rabbit
and bovine liver microsomes. Moose liver microsomes were
exceptionally active, yielding mainly tetrazole-N2-glucuro-
nide from losartan, as well as selectively producing O-
glucuronide from zolarsartan. In the case of candesartan,
moose liver microsomes catalyzed the formation of both
glucuronide regioisomers almost equally. Candesartan glu-
curonides were produced almost equally by all of the tested
liver microsomes, except rat liver microsomes that generated
the O-glucuronide of candesartan at much higher rates. Rat
liver microsomes were also exceptional in the glucuronidation
Liver microsomes from human, bovine, moose, pig, rabbit,
and rat produced three glucuronide regioisomers from
losartan, two from candesartan, and three from zolarsartan
(Fig. 4). All the tested liver microsomes glucuronidated
losartan to tetrazole-N2-glucuronide. In addition, significant

768
b i o c h e m i c a l p h a r m a c o l o g y 7 6 ( 2 0 0 8 ) 7 6 3 – 7 7 2
Fig. 3 – Screening the glucuronidation activity of recombinant human UGTs towards losartan, candesartan, and zolarsartan.
The normalized activity (activity corrected for expression level of individual UGTs) is presented and relative expression
levels of the different recombinant UGTs are given in parenthesis under the enzyme name. The different glucuronide
regioisomers are indicated by the column colour, O-glucuronide (white bar), tetrazole-N1-glucuronide (gray bar), and
tetrazole-N2-glucuronide (black bar). The values are averages of duplicate samples in which the substrate and protein
concentrations were 1 mM and 2.5 mg/mL, respectively, and the incubation time was 4 h.
of zolarsartan, since their main product was tetrazole-N1-
glucuronide rather than the O-glucuronide. Rabbit liver
microsomes were highly selective, producing only O-glucur-
onide of zolarsartan. Tetrazole-N2-glucuronide of zolarsartan
was only obtained at low rates in the reactions catalyzed by
human and pig liver microsomes.
(Fig. 3). The results of the kinetic analyses are presented in
Fig. 5. Losartan glucuronidation by human liver microsomes
followed Michaelis–Menten kinetics, as did the reactions
catalyzed by UGTs 1A1 and 2B7 (Fig. 5B and D). In contrast,
the catalysis of losartan N2-glucuronide by UGT1A3 is
governed by substrate inhibition kinetics (Fig. 5C). The kinetic
parameters for HLM and the three UGTs are presented in
Table 3.
Losartan is one of the oldest AT1 receptor antagonists and is
thus probably the most widely used drug of the three sartans
that were included in this study. Glucuronidation is an
important metabolic pathway for losartan and to get deeper
insight the pharmacokinetics of this drug, we have examined
the kinetics of its glucuronidation by HLM and by the three
hepatic human UGTs that exhibited the highest activity
towards this compound, UGT1A3, UGT1A1, and UGT2B7
4.
Discussion
The glucuronidation of losartan, candesartan, and zolarsartan
by different animal liver microsomes and a large set of

b i o c h e m i c a l p h a r m a c o l o g y 7 6 ( 2 0 0 8 ) 7 6 3 – 7 7 2
769
Fig. 4 – Screening the activity of liver microsomes from various species towards losartan, candesartan, and zolarsartan. The
type of glucuronides and the incubation conditions are the same as described for the recombinant UGTs in the legend to
Fig. 3.
recombinant human UGTs was studied. An interesting aspect
of these sartans is the presence of several potential sites for
both N- and O-glucuronidations within each aglycone (Fig. 1).
We have recently characterized the structures of the N- and O-
glucuronides that are produced from these sartans by NMR
[21], providing a tool to study the regioselectivity of individual
UGTs in such reactions.
Many of the human UGTs were active in the glucuronida-
tion of one or more of the three sartans, at one site or more
(Fig. 3). Among them, the activity of UGT1A3 was perhaps the
main finding of this study, particularly due to its high
regioselectivity towards the tetrazole-N2 of the three agly-
cones (Fig. 3). These results agree with the previously
published study in which UGT1A3 catalyzed the formation
Table 3 – Enzyme kinetic parameters (WS.E.) for the formation of losartan N2-glucuronide
HLMa
UGT1A1
UGT1A3
UGT2B7
Km (mM)
100.2 ꢂ 11.7
85.7 ꢂ 3.0
21.5 ꢂ 1.7
32.2 ꢂ 0.6
35.9 ꢂ 3.9b
162.3 ꢂ 20.7
16.1 ꢂ 0.8
Vmax (pmol/min/mg)
225.7 ꢂ 12.2
a
Human liver microsomes.
b
Substrate inhibition interaction Ksi = 400.4mM ꢂ 48.6.

770
b i o c h e m i c a l p h a r m a c o l o g y 7 6 ( 2 0 0 8 ) 7 6 3 – 7 7 2
Fig. 5 – Kinetics of losartan glucuronidation by human liver microsomes (HLM, A), and recombinant human UGTs 1A1 (B),
1A3 (C), and 2B7 (D). The data points, averages of triplicate measurements W S.E.M., were fitted to the Michaelis–Menten (A,
B, and D) or the substrate inhibition (C) equations.
of tetrazole-N2-glucuronide of RG 12525, a chemical entity
evaluated for the treatment of type II diabetes [26]. UGT1A3
was previously shown to be active in several N-glucuronida-
tion reactions, not merely with tetrazoles, but in those cases it
was often less active than UGT1A4 [12,16,19,27–29].
Losartan, like zolarsartan, turned out to be a useful
aglycone for exposing N-glucuronidation activity of unex-
pected enzymes, in this case UGTs 1A1, 1A10, 2B7, and 2B17. In
contrast to zolarsartan, however, nearly all the enzymes that
were active towards losartan, glucuronidated it at the same
site, the tetrazole-N2. The only exception was UGT1A10 that
also catalyzed losartan conjugation at the tetrazole-N1, even if
at a lower rate (Fig. 3). The versatility of UGT1A10 in this case
may reflect its activity in dobutamine glucuronidation, where
it was found to be the only human UGT that catalyzed high
glucuronidation of both catechol hydroxyls [36].
The results of the present study revealed that many UGTs
are able to catalyze N-glucuronidation reactions, even if the
enzyme most frequently associated with N-glucuronidation,
UGT1A4 [12,16,19,28–32], exhibited no significant activity
towards any of the three sartans studied here (Fig. 3). A
possible reason for this is that tetrazole is acidic, with a pKa
value of 4.9. The pKa value is similar to those of carboxylic
acids and therefore tetrazoles have been used as non-classical
isosteres for carboxylic acid moieties in biologically active
molecules [33,34]. In comparison to carboxylic acids, tetra-
zoles are more lipophilic, which could increase their affinity to
the UGTs.
Candesartan glucuronidation may be considered in terms
of both pharmacokinetics and human UGT substrate speci-
ficity. Candesartan differed from the two other sartans,
particularly from losartan, in being readily conjugated to O-
glucuronide, giving rise to candesartan acyl glucuronide
(Fig. 3). Interestingly, the three UGTs that exhibited the
highest O-glucuronidation of candesartan, 1A7, 1A8, and
1A10, are mainly extrahepatic enzymes and could account
for the low bioavailability of candesartan.
Zolarsartan was an interesting aglycone for several
different reasons. Since only UGT1A3 was the only human
UGT that catalyzed the formation of zolarsartan N2-glucur-
onide (Fig. 3), this reaction could be used as a specific probe for
the presence of active UGT1A3 in human tissue samples. In
contrast to N2, the conjugation of zolarsartan N1-tetrazole, at
least at low rates, was catalyzed by several different UGTs,
including 1A5, 1A10, 2B4, and perhaps also 1A6–1A8 (Fig. 3).
This is somewhat surprising since none of these human UGTs
is known for catalyzing N-glucuronidation. Incidentally,
zolarsartan appears to be a reasonably good substrate for
recombinant UGT1A5, an enzyme that is highly homologous
to both UGT1A3 and UGT1A4, but which was thus far exhibited
only marginal activities [35].
Comparing the glucuronides produced from zolarsartan by
human liver microsomes and individual UGTs (Figs. 3 and 4)
may be useful for assessing the contribution of UGT1A3 to
hepatic glucuronidation in humans. While UGT1A3 readily
and specifically produced zolarsartan tetrazole-N2-glucuro-
nide (Fig. 3), the fraction of zolarsartan tetrazole-N2-glucur-
onide in the HLM reaction, although detectable, was small
(Fig. 4). The main product in the latter reaction was the
zolarsartan O-glucuronide, even though the enzyme most
efficient in this reaction, UGT1A10 (Fig. 3), is extrahepatic [14].
Nevertheless, UGT2B7, an important liver enzyme, also

b i o c h e m i c a l p h a r m a c o l o g y 7 6 ( 2 0 0 8 ) 7 6 3 – 7 7 2
771
catalyzed relatively high rates of zolarsartan O-glucuronida-
tion (Fig. 3), probably accounting for most of the glucuronide
produced by the liver microsome sample (Fig. 4). UGT1A1,
another important liver enzyme, was also found to catalyze
zolarsartan O-glucuronidation, but at a low rate (Fig. 3).
Nevertheless, if the contribution of UGT1A1 for zolarsartan
was significant, one would have also expected to detect
zolarsartan tetrazole-N1-glucuronide in the liver microsome
reaction (Fig. 4). Since the latter metabolite was not detected in
this case, UGT1A1 was probably not significantly involved. It
may thus be suggested that mutations that lower the
expression level or activity of UGT1A1 should not significantly
affect the pharmacokinetics of zolarsartan.
tionactivitybetweenlivermicrosomesofdifferentanimals.The
new knowledge should be highly useful for understanding
substrate specificity of the UGTs, the formation of acyl
glucuronides and the suitability or unsuitability of using animal
liver microsomes for metabolic studies on drugs for humans.
Acknowledgments
We thank Merck, AstraZeneca, and GlaxoSmithKline for
generously providing losartan, candesartan, and zolarsartan
(GR117289), respectively. Katriina Itaaho, Sirkku Kallonen,
¨
Mika Kurkela, Johanna Mosorin, and Sanna Sistonen are
acknowledged for helpful discussions, help with GraphPad
Prism and LC–MS analyses, as well as for determination of
expression levels of UGTs and other technical assistance. This
study was financially supported by the Graduate School in
Pharmacy (AA) and by the Academy of Finland (project no.
210933).
The kinetics of losartan glucuronidation was examined in
order to obtain a deeper insight into its metabolism within the
human liver. Losartan glucuronidation by HLM followed
Michaelis–Menten kinetics, even if two enzyme kinetics could
not be excluded in this case (Fig. 5A). In trying to assess the
contribution of UGTs 1A1, 1A3, and 2B7 to the losartan
glucuronidation activity of HLM, we have examined their
reactions as well. UGT1A3 exhibited the highest losartan
glucuronidation activity in the screening assays (Fig. 3) and its
apparent affinity for this compound is high (Table 3). Inter-
estingly, UGT1A3 exhibited clear substrate inhibition kinetics
(Fig. 5C), a phenomenon not seen in HLM (Fig. 5A). It can thus
be concluded that the contribution of UGT1A3 to the losartan
glucuronidation activity of HLM is low, regardless the high
activity of this enzyme. This finding is in full agreement with
our conclusion about the minor involvement of UGT1A3 in the
zolarsartan glucuronidation activity of HLM and it strongly
implies that the concentration of UGT1A3 in HLM is low.
The kinetic analyses of UGT1A1 and UGT2B7, two enzymes
that play major roles in hepatic glucuronidation, suggest that
both of them contribute to losartan glucuronidation by HLM.
UGT1A1, as well as UGT2B7, exhibited Michaelis–Menten
kinetics without significant substrate inhibition (Fig. 5B and
D). Nevertheless, there is a clear difference in the Km values of
these UGTs for losartan with UGT1A1 having a significantly
lower Km for this substrate (Table 3). Interestingly, the Km of
HLM for losartan, about 100 mM, is higher than the respective
valueinUGT1A1, about22 mM,butlowerthantheKm ofUGT2B7,
about 162 mM (Table 3). These findings, together with the shape
of the kinetic curve and the above conclusion on the only minor
contribution of UGT1A3, suggest that both UGTs 1A1 and 2B7
contribute significantly, and rather similarly, to the losartan
glucuronidation activity of human liver microsomes.
r e f e r e n c e s
[1] Habashi JP, Judge DP, Holm TM, Cohn RD, Loeys BL, Cooper
TK, et al. Losartan, an AT1 antagonist, prevents aortic
aneurysm in a mouse model of Marfan syndrome. Science
2006;312:117–21.
[2] Matt P, Habashi J, Carrel T, Cameron DE, Van Eyk JE, Dietz
HC. Recent advances in understanding Marfan syndrome:
should we now treat surgical patients with losartan? J
Thorac Cardiovasc Surg 2008;135:389–94.
[3] Travis J. Medicine. Old drug, new hope for Marfan
syndrome. Science 2006;312:36–7.
[4] Lacro RV, Dietz HC, Wruck LM, Bradley TJ, Colan SD,
Devereux RB, et al. Rationale and design of a randomized
clinical trial of beta-blocker therapy (atenolol) versus
angiotensin II receptor blocker therapy (losartan) in
individuals with Marfan syndrome. Am Heart J
2007;154:624–31.
[5] Robertson MJ, Barnes JC, Drew GM, Clark KL, Marshall FH,
Michel A, et al. Pharmacological profile of GR117289 in vitro:
a novel, potent and specific non-peptide angiotensin AT1
receptor antagonist. Br J Pharmacol 1992;107:1173–80.
[6] Hilditch A, Hunt AA, Gardner CJ, Twissell DJ, Polley J,
Travers A, et al. Cardiovascular effects of GR117289, a novel
angiotensin AT1 receptor antagonist. Br J Pharmacol
1994;111:137–44.
[7] Stearns RA, Miller RR, Doss GA, Chakravarty PK, Rosegay A,
Gatto GJ, et al. The metabolism of DuP 753, a nonpeptide
angiotensin II receptor antagonist, by rat, monkey, and
human liver slices. Drug Metab Dispos 1992;20:281–7.
[8] Fenton C, Scott LJ. Candesartan cilexetil: a review of its use
in the management of chronic heart failure. Drugs
2005;65:537–58.
[9] Kondo T, Yoshida K, Yoshimura Y, Motohashi M,
Tanayama S. Disposition of the new angiotensin II receptor
antagonist candesartan cilexetil in rats and dogs.
Arzneimittelforschung 1996;46:594–600.
The analyses of the glucuronidation activity of liver
microsomes from different animals revealed, once again, that
there are major differences in glucuronidation activity
between different species (Fig. 4). The differences are both
qualitative, in the type of metabolites formed, and quantita-
tive. Since acyl glucuronides may be toxic [37–40], it is
important to note that rat liver microsomes catalyze the
formation of candesartan acyl glucuronides faster than any
other microsomes sample, certainly much faster than the
human liver microsomes (Fig. 4).
[10] Kondo T, Yoshida K, Yoshimura Y, Motohashi M,
Tanayama S. Characterization of conjugated metabolites of
a new angiotensin II receptor antagonist, candesartan
cilexetil, in rats by liquid chromatography/electrospray
tandem mass spectrometry following chemical
derivatization. J Mass Spectrom 1996;31:873–8.
In conclusion, this study of losartan, candesartan, and
zolarsartan glucuronidation produced interesting findings
regarding N-glucuronidation, provided a model substrate for
UGT1A3, and demonstrated large differences in glucuronida-

772
b i o c h e m i c a l p h a r m a c o l o g y 7 6 ( 2 0 0 8 ) 7 6 3 – 7 7 2
[11] Bowers GD, Eddershaw PJ, Hughes SY, Manchee GR, Oxford
J. Characterization of glucuronic acid conjugates of a novel
angiotensin receptor antagonist. Rapid Commun Mass
Spectrom 1994;8:217–21.
[12] King CD, Rios GR, Green MD, Tephly TR. UDP-
glucuronosyltransferases. Curr Drug Metab 2000;1:143–61.
[13] Ritter JK. Roles of glucuronidation and UDP-
glucuronosyltransferases in xenobiotic bioactivation
reactions. Chem Biol Interact 2000;129:171–93.
N-glucuronidation by in vitro and in vivo approaches. Drug
Metab Dispos 2001;29:289–95.
[27] Green MD, King CD, Mojarrabi B, Mackenzie PI, Tephly TR.
Glucuronidation of amines and other xenobiotics catalyzed
by expressed human UDP-glucuronosyltransferase 1A3.
Drug Metab Dispos 1998;26:507–12.
[28] Green MD, Tephly TR. Glucuronidation of amine substrates
by purified and expressed UDP-glucuronosyltransferase
proteins. Drug Metab Dispos 1998;26:860–7.
[14] Tukey RH, Strassburg CP. Human UDP-
[29] Kaji H, Kume T. Characterization of afloqualone N-
glucuronidation: species differences and identification of
human UDP-glucuronosyltransferase isoform(s). Drug
Metab Dispos 2005;33:60–7.
glucuronosyltransferases: metabolism, expression, and
disease. Annu Rev Pharmacol Toxicol 2000;40:581–616.
[15] Ouzzine M, Barre L, Netter P, Magdalou J, Fournel-Gigleux S.
The human UDP-glucuronosyltransferases: structural
aspects and drug glucuronidation. Drug Metab Rev
2003;35:287–303.
[30] Vashishtha SC, Hawes EM, McKay G, McCann DJ.
Quaternary ammonium-linked glucuronidation of 1-
substituted imidazoles: studies of human UDP-
glucuronosyltransferases involved and substrate
specificities. Drug Metab Dispos 2001;29:1290–5.
[31] Nakajima M, Tanaka E, Kobayashi T, Ohashi N, Kume T,
Yokoi T. Imipramine N-glucuronidation in human liver
microsomes: biphasic kinetics and characterization of
UDP-glucuronosyltransferase isoforms. Drug Metab Dispos
2002;30:636–42.
[32] Wiener D, Doerge DR, Fang JL, Upadhyaya P, Lazarus P.
Characterization of N-glucuronidation of the lung
carcinogen 4-(methylnitrosamino)-1-(3-pyridyl)-1-
butanol(NNAL) in human liver: importance of UDP-
glucuronosyltransferase 1A4. Drug Metab Dispos
2004;32:72–9.
[33] Herr RJ. 5-Substituted-1H-tetrazoles as carboxylic acid
isosteres: medicinal chemistry and synthetic methods.
Bioorg Med Chem 2002;10:3379–93.
[34] Myznikov LV, Hrabalek A, Koldobskii GI. Drugs in the
tetrazole series (review). Chem Heterocycl Compd
2007;43:1–9.
[35] Finel M, Li X, Gardner-Stephen D, Bratton S, Mackenzie PI,
Radominska-Pandya A. Human UDP-
glucuronosyltransferase 1A5: identification, expression,
and activity. J Pharmacol Exp Ther 2005;315:1143–9.
[36] Alonen A, Aitio O, Hakala K, Luukkanen L, Finel M,
Kostiainen R. Biosynthesis of dobutamine
[16] Kiang TK, Ensom MH, Chang TK. UDP-
glucuronosyltransferases and clinical drug–drug
interactions. Pharmacol Ther 2005;106:97–132.
[17] Guillemette C. Pharmacogenomics of human UDP-
glucuronosyltransferase enzymes. Pharmacogenom J
2003;3:136–58.
[18] Mackenzie PI, Walter Bock K, Burchell B, Guillemette C,
Ikushiro S, Iyanagi T, et al. Nomenclature update for the
mammalian UDP glycosyltransferase (UGT) gene
superfamily. Pharmacogenet Genom 2005;15:677–85.
[19] Radominska-Pandya A, Czernik PJ, Little JM, Battaglia E,
Mackenzie PI. Structural and functional studies of UDP-
glucuronosyltransferases. Drug Metab Rev 1999;31:817–99.
[20] Stearns RA, Doss GA, Miller RR, Chiu SH. Synthesis and
identification of a novel tetrazole metabolite of the
angiotensin II receptor antagonist DuP 753. Drug Metab
Dispos 1991;19:1160–2.
[21] Alonen A, Jansson J, Kallonen S, Kiriazis A, Aitio O, Finel M,
et al. Enzyme-assisted synthesis and structure
characterization of glucuronic acid conjugates of losartan,
candesartan, and zolarsartan. Bioorg Chem 2008;36:148–55.
[22] Huskey SW, Miller RR, Chiu SHL. N-Glucuronidation
reaction. I. Tetrazole N-glucuronidation of selected
angiotensin II receptor antagonists in hepatic microsomes
from rats, dogs, monkeys, and humans. Drug Metab Dispos
1993;21:792–9.
monoglucuronides and glucuronidation of dobutamine by
recombinant human UDP-glucuronosyltransferases. Drug
Metab Dispos 2005;33:657–63.
[23] Krieter PA, Colletti AE, Miller RR, Stearns RA. Absorption
and glucuronidation of the angiotensin II receptor
antagonist losartan by the rat intestine. J Pharmacol Exp
Ther 1995;273:816–22.
[37] Stachulski AV, Harding JR, Lindon JC, Maggs JL, Park BK,
Wilson ID. Acyl glucuronides: biological activity, chemical
reactivity, and chemical synthesis. J Med Chem
2006;49:6931–45.
[38] Bailey MJ, Dickinson RG. Acyl glucuronide reactivity in
perspective: biological consequences. Chem Biol Interact
2003;145:117–37.
[39] Sallustio BC, Sabordo L, Evans AM, Nation RL. Hepatic
disposition of electrophilic acyl glucuronide conjugates.
Curr Drug Metab 2000;1:163–80.
[40] Shipkova M, Armstrong VW, Oellerich M, Wieland E. Acyl
glucuronide drug metabolites: toxicological and analytical
implications. Ther Drug Monit 2003;25:1–16.
[24] Kurkela M, Patana AS, Mackenzie PI, Court MH, Tate CG,
Hirvonen J, et al. Interactions with other human UDP-
glucuronosyltransferases attenuate the consequences of
the Y485D mutation on the activity and substrate affinity of
UGT1A6. Pharmacogenet Genom 2007;17:115–26.
[25] Luukkanen L, Elovaara E, Lautala P, Taskinen J, Vainio H.
Characterization of 1-hydroxypyrene as a novel marker
substrate of 3-methylcholanthrene-inducible phenol UDP-
glucuronosyltransferase(s). Pharmacol Toxicol 1997;80:152–8.
[26] Stevens JC, Fayer JL, Cassidy KC. Characterization of 2-[[4-[[2-
(1H-tetrazol-5-ylmethyl)phenyl]methoxy]methyl]quinoline