Pyrazole-O-glucosides as novel Na+-glucose cotransporter (SGLT) inhibitors

Article abstract of DOI:10.1016/S0960-894X(03)00466-9

O-glucuronides and O-glucosides of a series of pyrazoles analogues were synthesized and evaluated for their SGLT inhibitory activity in brush border membrane vehicles (BBMVs) of rat kidney. O-glucosides of certain pyrazole analogues inhibited the transport of [¹⁴C]-glucose in BBMVs, and induced glucosuria in Wistar rats by intravenous injection.

Full text of DOI:10.1016/S0960-894X(03)00466-9

Bioorganic& Medicinal Chemistry Letters 13 (2003) 2269–2272
Pyrazole-O-Glucosides as Novel Na+-Glucose Cotransporter
(SGLT) Inhibitors
Koji Ohsumi,* Hiroyuki Matsueda, Toshihiro Hatanaka, Ryusuke Hirama,
Takashi Umemura, Akiko Oonuki, Nozomu Ishida, Yoko Kageyama, Katsumi Maezono
and Nobuo Kondo
Pharmaceutical Research Laboratories, Pharmaceutical Company, Ajinomoto Co., Inc., 1-1, Suzuki-cho, Kawasaki-ku,
Kawasaki, 210-8681, Japan
Received 24 January 2003; accepted 2 May 2003
Abstract—O-glucuronides and O-glucosides of a series of pyrazoles analogues were synthesized and evaluated for their SGLT
inhibitory activity in brush border membrane vehicles (BBMVs) of rat kidney. O-glucosides of certain pyrazole analogues inhibited
the transport of [14C]-glucose in BBMVs, and induced glucosuria in Wistar rats by intravenous injection.
# 2003 Elsevier Science Ltd. All rights reserved.
Introduction
the increase of urinary sugar excretion and decrease of
blood glucose level. Thus, SGLT inhibitors are thought
to be promising candidates for diabetes treatment.
Na+-glucose cotransporter (SGLT) is a membrane pro-
tein that plays an important role in the reabsorption of
glucose in the kidney. It has been reported that there are at
least three isoforms of SGLTs (SGLT1, SGLT2 and
SGLT3).1ꢀ3 SGLT2 mainly exists in renal uriniferous
tubule and SGLT1 exists in the kidney and the intetestine.
Glucose that is filtered through glomeruli is reabsorbed at
the renal uriniferous tubules via SGLT1 and SGLT2. It is
expected that the inhibition of SGLT at the kidney could
decrease glucose reabsorption and this could result in
Tsujihara et al. reported a SGLT inhibitor T-1095A (1)
and its prodrug T-1095 (2) as anti-diabeticagents
(Fig. 1).4ꢀ6 T-1095A lowers blood glucose level by inhib-
iting SGLT2 in the kidney, leading to increase urinary
glucose excretion.
We were interested in this mechanism of action and
started searching for new anti-diabetic agents that have
Figure 1.
*Corresponding author. E-mail: koji_ohsumi@ajinomoto.com
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00466-9

2270
K. Ohsumi et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2269–2272
SGLT2 inhibitory activity. By database searching,
we found WAY-123783 (3), which was reported to
increases glucosuria and lower blood glucose levels in
db/db mice. (Fig. 1).7,8 The authors suggested that
WAY-123783 was a SGLT2 inhibitor in their report.
nizes glucose as a substrate, we thought it is possible
that sugar analogues such as the glucuronide and glu-
coside of WAY-123783 could inhibit SGLT2. There-
fore, we synthesized these compounds and tested their
SGLT inhibitory activity in vitro and in vivo.
To clarify the profile of WAY123783 (3) and related
compounds, we synthesized 3 and WAY-2-12 (7) and
tested their activity in several assays. Oral administra-
tion of 3 increased urinary glucose excretion in mice but
did not induce the same effect in rats. In addition,
compound 3 did not inhibit the transport of [14C]-glu-
cose in BBMVs of rat kidney at all.5
Synthesis
We attempted to synthesize the glucuronide and gluco-
side of WAY-123783 by the Mitsunobu procedure.11
PhenolicOH and a nitrogen atom in the pyrazole ring
of WAY-123783 were thought to be the reaction site.
First, we examined the regioselectivity of the Mitsunobu
reaction. Treatment of a THF solution of WAY-123783
and isopropanol with ethyl azodicarboxylate and tri-
phenylphosphine provided single product 4 in 68%. To
elucidate the position of isopropyl group, we used NOE
method. Cross peak was obtained between the proton
on the benzene ring (Ha) and the proton on the isopro-
pyl group (Hb). Compound 4 obtained was confirmed
as an O-isopropyl analogue (Scheme 1).
From these results, we hypothesized that WAY-123783
was converted to a metabolite after oral administration
in mice and this metabolite induced glucosuria in vivo.
Several polar metabolites were observed in mice plasma
after oral administration of WAY-123783 by HPLC
analysis. From mass spectrum, one of the metabolites
showed the molecular weight of a glucuronide.
It has been shown that phenolic compounds can be
metabolized to give the corresponding glucuronides.9
Furthermore, metabolic glucosidation of phenolic com-
pounds has also been reported.10 Since SGLT2 recog-
Synthesis of a glucuronide of WAY-123783 is shown in
Scheme 2. By using the Mitsunobu procedure, glucur-
onide 5 was obtained in 33%. Debenzylation of 5 with
Scheme 1. Reaction of WAY-123783 and isopropanol by Mistunobu reaction. (a) Isopropanol, 40% DEAD/toluene, PPh3.
Scheme 2. Synthesis of O-glucuronide of WAY-123783. (a) 2,3,4-O-tribenzyl-d-glucopyranosideuronic acid benzylester, PPh3, 40% DEAD/toluene
THF; (b) 20% Pd (OH)2, MeOH-EtOAc.
Figure 2.

K. Ohsumi et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2269–2272
2271
Scheme 3. Synthesis of O-glucuronide and O-glycoside of WAY-2-12. (a) 2,3,4-O-tribenzyl-d-glucopyranosideuronic acid benzylester, 40% DEAD/
toluene, PPh3, THF; (b) 2,3,4,6-tetrabenzyl-d-glucopyranose, 40% DEAD/toluene, PPh3, THF; (c) 20% Pd (OH)2 MeOH–EtOAc.
Table 2. Effect of pyrazole analogues on urinary glucose excretion in rat
Table 1. Effect of pyrazole analogues on rat kidney SGLT inhibitory
activity
Compd
Glucosuria (mg/24 h)
iv (3 mg/kg)
Compd
Inhibition rate (%)
9
11
30% (100 mM)
84% (10 mM)
90% (10 mM)
9
11
2.6 mg
63 mg
300 mg
0 mg
T-1095A (1)
WAY-123783 (3)
T-1095A (1)
WAY-123783 (3)
Not active (100 mM)
Inhibition of [14C]-glucose uptake in BBMV’s of rat kidney.
Each value represents the mean of three experiments.
Each value represents the mean of three experiments.
20%Pd (OH)2 was sluggish, and did not give the desired
product 6. The thiomethyl group on the benzene ring
was thought to diminish the activity of the palladium
catalyst, so we used WAY-2-12 (7) instead of WAY-
123783 (3). WAY-2-12 has a 4-ethyl group in place of
the thiomethyl group of WAY-123783. In Wyeth-Ayer-
st’s report, WAY-2-12 showed as potent an anti-diabetic
effect as WAY-123783 (Fig. 2).8
9 showed weak inhibitory activity against rat kidney
BBMVs (30% inhibition at 100 mM). On the other
hand, O-glucoside 11 showed potent inhibitory activity.
(84% inhibition at 10 mM). T-1095A, an active form of
T-1095 showed potent inhibitory activity (90% inhibi-
tion at 10 mM) (Table 1).
Next, the effect on urinary glucose excretion activity of
O-glucuronide 9 and O-glucoside 11 were evaluated in
Wistar rats by intravenous injection.6,15 3 mg/kg of
O-glucuronide 9 induced only marginal urinary sugar
excretion (2.6 mg). On the other hand, 3 mg/kg of
O-glucoside 11 induced 63 mg of urinary sugar excre-
tion. At the same dose, T-1095A induced 300 mg of
glucosuria in Wistar rats (Table 2).
WAY-2-12 was glucuronidated by the Mitsunobu pro-
cedure to give benzyl-protected glucuronide derivative 8
in 44%. The stereochemistry of 8 at the anomericcenter
was assigned as beta by NMR studies. (Janomeric H=6.0
Hz). Glucuronide 8 obtained was successfully debenzy-
lated with 20%Pd(OH)2 under a hydrogen atmosphere
to give compound 9 in 74%.12
In a similar fashion, WAY-2-12 was converted to ben-
zyl-protected glucoside derivative 10 in 72%. The ste-
reochemistry of 10 at the anomericcenter was assigned
as beta by NMR studies. (Janomeric H=6.6 Hz). This
glucoside 10 was debenzylated with 20%Pd (OH)2 to
give the desired glucoside 11 in 92% (Scheme 3).13
Conclusion
We have synthesized O-glucoside and O-glucuronide of
WAY-123783 analogue based on the hypothesis that
WAY-123783 shows SGLT2 inhibitory activity after meta-
bolism in mice. We found that glucoside 11 was a potent
inhibitor of SGLT in rat kidney BBMVs. Intravenous
injection of glucoside analogue 11 increased urinary
sugar excretion in Wistar rats. This glucoside is thought
to be a promising candidate as an anti-diabetic agent.
Although, glucoside 11 showed inhibitory activity as
potent as T-1095A in rat kidney BBMVs, it showed
only one-fifth the effect of T-1095A on urinary glucose
Biological Result
The SGLT inhibitory activity of O-glucuronide 9 and
O-glucoside 11 against [14C]-glucose transport in
BBMVs of rat kidney was evaluated.4,14 O-glucuronide

2272
K. Ohsumi et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2269–2272
excretion in rats. This discrepancy was thought to be due to
the difference in stability between glucoside 11 andT-1095A
in rats. Optimization of O-glucoside 11 is now underway to
improve stability and selectivity against SGLTs.
J=11.4), 4.69 (1H, d, J=11.1), 4.54 (1H, d, J=11.1), 4.49
(1H, d, J=11.1), 4.46 (1H, d, J=11.1), 4.12 (1H, d, J=9.6),
3.98 (1H, t, J=8.4), 3.80 (2H, d, J=3.0), 3.76-3.84 (2H, brd),
3.71 (1H, t, J=7.5), 3.67 (1H, t, J=6.9), 2.49 (2H, q, J=7.5),
1.12 (3H, t, J=7.5). Compound 8 (122 mg, 0.15 mmol) was
hydrogenated over 20% Pd(OH)2 (200 mg) in EtOAc(4 mL)-
methanol (4 mL) for 8 h. The catalyst was removed by fil-
tration,. The filtrate was purified by SepPack column (10%
MeOH/H2O to 100% MeOH) to give desired product 9 as
During the course of this work, compound 11 was
claimed for SGLT inhibitors (WO 0116147). To date,
details of the method of discovery have not been pub-
lished.
1
white solid (22 mg, 33%). H NMR (300 MHz, DMSO-d6) d
7.09 (2H, d, J=8.4 Hz), 7.06 (2H, d, J=8.4 Hz), 5.00–5.20
(1H, br), 3.70–3.90 (3H, m), 3.52–3.65 (1H, m), 3.35–3.51 (2H,
m), 2.58 (2H, q, J=7.5 Hz), 1.19 (3H, t, J=7.5 Hz), ESI-
MS(m/z) 445[(M-H)ꢀ], 447[(M+H)+].
References and Notes
13. 4-(4-Ethylphenyl)methyl-5-trifluoromethyl-1H-pyrazol-3-yl-
O-ꢀ-D-glucopyranoside 11. A mixture of 2,3,4,6-O-tetrabenzyl-
d-glucopyranose (107 mg, 0.39 mmol) 1,2-dihydro-4-(ethyl-
phenyl)methyl-5-trifluoromethyl-3H-pyrazol-3-one 7 (214 mg,
0.39 mmol) triphenylphosphine (102 mg, 0.39 mmol) and
anhydrous THF (5 mL) were stirred in ice-water bath. To this
solution, 40% Ethyl azodicarboxylate/toluene (0.175 mL, 0.39
mmol) was added dropwise. After stirring for 24 h, the reac-
tion mixture was purified by silica-gel chromatography (25%
EtOAc–hexane) to give 40-(40-ethylphenyl)methyl-50-trifluoro-
methyl-1H-pyrazol-30 -yl-O-2,3,4,6-O-tetrabenzyl-b-d-gluco-
pyranoside 10 (223 mg, 72%) as oil. 1H NMR (300 MHz,
CDCl3) d 6.98–7.34 (24H, m), 5.09 (1H, anomeric, d, J=6.6),
4.46–4.88 (8H, m), 3.81 (2H, s), 3.56–3.74 (6H, m), 2.53 (2H,
q, J=7.5), 1.14 (3H, t, J=7.5). ESI-MS (m/z) 791 [(M–H)ꢀ],
793[(M+H)+].
Compound 10 (220 mg, 0.28 mmol) was hydrogenated with
20% Pd(OH)2 (100 mg) in EtOAc(4 mL)-methanol (4 mL) for
1.5 h. The catalyst was removed by filtration, and the filtrate
was evaporated. The residue was purified by silica gel chro-
matography (10% MeOH–CH2Cl2) to give desired product 11
as white solid (110 mg, 92%). 1H NMR (300 MHz, DMSO-d6)
d 7.08 (4H, s), 5.11(1H, brs), 5.02(1H, brs), 4.89 (1H, anomeric
d, J=8.7), 3.74 (2H, s), 3.66 (1H, d, J=12.0), 3.48 (1H, dd,
J=5.1, 12.0), 3.10-3.24 (4H, br), 2.53 (2H, q, J=7.5), 1.13
(3H, t, J=7.5). ESI-MS (m/z) 431 [(M–H)ꢀ], 863 [(2M–H)ꢀ].
14. Rat kidney BBMVs assay method
1. Lee, W.-S.; Kanai, Y.; Wells, R. G.; Hediger, M. A. J. Biol.
Chem. 1994, 269, 12032.
2. You, G.; Lee, W.-S.; Barros, E. J. G.; Kanai, Y.; Huo, T.-
L.; Khawaja, S.; Wells, R. G.; Nigam, S. K.; Heidiger, M. A.
J. Biol. Chem. 1995, 270, 29365.
3. Mackenzie, B.; Panayotova-Heiermann, M.; Loo, D. D. F.;
Lever, J. E.; Wright, E. M. J. Biol. Chem. 1994, 269, 22488.
4. Tsujihara, K.; Hongu, M.; Saito, K.; Inamasu, M.; Ara-
kawa, K.; Oku, A.; Matsumoto, M. Chem. Pharm. Bull. 1996,
44, 1174.
5. Tsujihara, K.; Hongu, M.; Saito, K.; Kawanishi, H.; Kur-
iyama, K.; Matsumoto, M.; Oku, A.; Ueta, K.; Tsuda, M.;
Saito, A. J. Med. Chem. 1999, 42, 5311.
6. Oku, A.; Ueta, K.; Arakawa, K.; Ishihara, T.; Nawano,
M.; Kuronuma, Y.; Matsumoto, M.; Saito, A.; Tsujihara, K.;
Anai, M.; Asano, T.; Kanai, Y.; Endou, H. Diabetes 1999, 48,
1794.
7. Kees, K. L.; Caggiano, T. J.; Steiner, K. E.; Fitzgerald, J. J.;
Kates, M. J.; Christos, T. E.; Kulishoff, J. M.; Moore, R. D.;
McCaleb, M. L. J. Med. Chem. 1995, 38, 617.
8. Kees, K. L.; Fitzgerald, J. J.; Steiner, K. E.; Mattes, J. F.;
Mihan, B.; Tosi, T.; Mondoro, D.; McCaleb, M. L. J. Med.
Chem. 1996, 39, 3920.
9. Dodge, J. A.; Lugar, C. W.; Cho, S.; Osborne, J. J.; Phil-
lips, D. L.; Glasebrook, A. L.; Frolik, C. A. Bioorg. Med.
Chem. Lett. 1997, 7, 993.
10. Tang, B.-K. Pharmac. Ther. 1990, 46, 53.
11. Vaccaro, W. D.; Sher, R.; Davis, H. R. Bioorg. Med.
Chem. Lett. 1998, 8, 35.
Brush border membrane vesicles (BBMVs) were prepared
by the Ca2+ precipitation method from kidney of male Wistar
rats (8–10 weeks old). The test compounds, d-glucose (0.2
mM), [14C]-d-glucose (0.1 mCi) and NaCl (100 mM) were
added and incubated with BBMV’s for one min. The glucose
uptake reaction was terminated by addition of 1mM phlorizin
in 40 mM Hepes–Tris (pH 7.4). The vehicles were immediately
centrifuged and filtered, then the radioactivity on the mem-
brane was measured by a liquid scintillation counter.
12. 4-(4-Ethylphenyl)methyl-5-trifluoromethyl-1H-pyrazol-3-yl-
ꢀ-D-glucopyranoside uronic acid 9 A mixture of 2,3,4-tri-O-
benzyl-d-glucopyranoside uronic acid benzyl ester (199 mg,
0.36
mmol),1,2-dihydro-4-(ethylphenyl)methyl-5-trifluoro-
methyl-3H-pyrazol-3-one 7 (99 mg, 0.37 mmol), triphenyl-
phosphine (109 mg, 0.42 mmol) and anhydrous THF (0.5 mL)
were stirred in ice-water bath. To this solution, 40% ethyl
azodicarboxylate/toluene (0.18 mL, 0.40 mmol) was added
dropwise. After 1.5 h, the reaction mixture was purified by
silica-gel chromatography (20% EtOAc–hexane) to give ben-
zyl 40-(40-ethylphenyl) methyl-50-trifluoromethyl-1H-pyrazol-
30-yl-2,3,4-O-tribenzyl-b-d-glucopyranouronate (127mg, 44%)
8 as yellow oil. 1H NMR (300 MHz, CDCl3 ) d 7.22–7.36
(16H, m), 7.08–7.18 (4H, m), 7.07 (2H, d, J=8.7), 7.00 (2H, d,
J=8.7), 5.24 (1H, anomericd, J=6.0), 5.15 (1H, d, J=12.0),
5.10 (1H, d, J=12.0), 4.76 (1H, d, J=11.4), 4.70 (1H, d,
15. Urinary glucose excretion in rats
Test compounds were dissolved in DMSO and diluted by
saline. The compounds were intravenously administered to
male Wistar rats (4 week old, Charles River Japan, Tokyo,
Japan) housed in metaboliccages. Twenty-four hours after the
administration, urine was collected and urine volume was
measured. Glucose concentration in urine was measured using
an auto-analyzer (FUJI DRI-CHEM, Fuji Photo Film Co.,
Ltd, Tokyo, Japan).