Design, synthesis and in vivo anti-hyperglycemic activity of gem-dimethyl-bearing C-glucosides as SGLT2 inhibitors

Article abstract of DOI:10.1016/j.cclet.2011.05.022

A series of gem-dimethyl-bearing C-glucosides were designed and synthesized as SGLT2 inhibitors, with anhydrous aluminum chloride-mediated Friedel-Crafts alkylation to construct the gem-dimethyl functionality being the key step. The in vivo anti-hyperglycemic activity was evaluated with mice oral glucose tolerance test (OGTT), and all the synthesized compounds showed significant but less potent anti-hyperglycemic activity than the positive control dapagliflozin.

Full text of DOI:10.1016/j.cclet.2011.05.022

Available online at www.sciencedirect.com
Chinese Chemical Letters 22 (2011) 1215–1218
www.elsevier.com/locate/cclet
Design, synthesis and in vivo anti-hyperglycemic activity of
gem-dimethyl-bearing C-glucosides as SGLT2 inhibitors
b
Wen Jing Zhao a, Yong Heng Shi b,c, Gui Long Zhao b, , Yu Li Wang ,
*
Hua Shao b, Li Da Tang b, Jian Wu Wang a,
*
a School of Chemistry and Chemical Engineering, Shandong University, Jinan 250100, China
b Tianjin Key Laboratory of Molecular Design and Drug Discovery, Tianjin Institute of Pharmaceutical Research, Tianjin 300193, China
c School of Pharmacy, Tianjin Medical University, Tianjin 300070, China
Available online 18 July 2011
Abstract
A series of gem-dimethyl-bearing C-glucosides were designed and synthesized as SGLT2 inhibitors, with anhydrous aluminum
chloride-mediated Friedel-Crafts alkylation to construct the gem-dimethyl functionality being the key step. The in vivo anti-
hyperglycemic activity was evaluated with mice oral glucose tolerance test (OGTT), and all the synthesized compounds showed
significant but less potent anti-hyperglycemic activity than the positive control dapagliflozin.
# 2011 Gui Long Zhao. Published by Elsevier B.V. on behalf of Chinese Chemical Society. All rights reserved.
Keywords: Synthesis; C-Glucoside; SGLT2 inhibitor; Anti-hyperglycemic activity; gem-Dimethyl
Diabetes is a chronic metabolic disease which is characterized by hyperglycemia. If not being controlled
appropriately, hyperglycemia would lead to a variety of severe diabetic complications.
Over 99% of the plasma glucose that is filtered in the renal glomerulus is reabsorbed into the blood mainly by
sodium-glucose co-transporter 2 (SGLT2) in the renal proximal tubule [1]. Therefore, inhibition of SGLT2 would
suppress the reabsorption of glucose from glomerular filtrate, thus lowering the blood glucose levels. SGLT2 inhibitors
have become a promising class of hypoglycemic agents for the treatment of type 2 diabetes.
The origin of the discovery of SGLT2 inhibitors was the natural product phlorizin (Fig. 1), which has been
discovered as a glycosuric agent more than several decades ago. More recently, phlorizin was found to be a SGLT2
inhibitor, but it was non-selective against SGLT2/SGLT1 and was liable to the degradation by b-glycosidase in the
small intestine [2]. Encouraged by these observations, a number of O-glucosides were discovered as SGLT2 inhibitors
based on the molecular structure of phlorizin in the earlier stage (Fig. 1). However, like phlorizin, they were later found
to be still liable to the degradation by b-glycosidase [3], whose pre-clinical or clinical trials were all discontinued.
The C-glucosides were subsequently designed, and they were found to be robust to the degradation by b-
glycosidase (Fig. 2). Among the C-glucosides discovered as SGLT2 inhibitors so far, dapagliflozin was the most
advanced one, which is now in phase III clinical trials [1].
The benzylic methylene group between the two benzene rings of dapagliflozin was liable to be hydroxylated in in
vivo metabolism, and the metabolite thus formed was not a SGLT2 inhibitor [4]. Inspired by these observations, a
* Corresponding author.
E-mail addresses: zhao_guilong@126.com (G.L. Zhao), jwwang@sdu.edu.cn (J.W. Wang).
1001-8417/$ – see front matter # 2011 Gui Long Zhao. Published by Elsevier B.V. on behalf of Chinese Chemical Society. All rights reserved.
doi:10.1016/j.cclet.2011.05.022

1216
W.J. Zhao et al. / Chinese Chemical Letters 22 (2011) 1215–1218
O
HO
OMe
OR
O
OR
OH
OR
O
O
O
O
HO
HO
O
HO
HO
O
O
O
HO
HO
HO
OH
O
O
OH
OH
OH
HO
N
N
OH
OH
T-1095A, R = H
T-1095, R = CO2Me
R = H
sergliflozin, R = CO Et
R = H;
OH
phlorizin
remogliflozin, R = CO Et
2
2
Fig. 1. Molecular structures of some well-established O-glucosides as SGLT2 inhibitors in the earlier stage.
variety of gem-dimethyl-bearing C-glucosides have been designed and evaluated in vivo in our laboratories, which
were found to be potent SGLT2 inhibitors (Fig. 3) [5]. In order to further investigate the structure-activity relationship
of this novel class of gem-dimethyl-bearing SGLT2 inhibitors, a new series of C-glucosides were further designed here
to explore the effect of the substituents at the different positions of benzene ring adjacent to the glucose moiety on the
anti-hyperglycemic activity (Scheme 1).
The synthetic route to the title compounds was outlined in Scheme 1. D-Glucolactone 2 was pertrimethylsilylated
according to a known procedure [1,6]. Benzoic acids 4a–d were converted to their ethyl esters 5a–d following a
standard procedure. Esters 5a–d were treated with excess MeMgCl in THF to smoothly furnish the tertiary alcohols
6a–d, which were coupled with phenetol to give rise to 7a–d by AlCl3-mediated Friedel-Crafts alkylation.
Bromides 7a–d were treated with n-BuLi at À78 8C to give the corresponding aryl lithiums, which were trapped in
situ with 3 to afford 8a–d. The intermediates 8a–d formed were treated in situ with methanol in the presence of
MsOH to yield 9a–d. The anomeric methoxy group in 9a–d was reductively removed with Et3SiH in the presence of
BF3ÁEt2O to furnish C-glucosides 10a–d as anomeric mixtures. Peracetylation of 10a–d to 11a–d with Ac2O was
achieved in refluxing acetic acid in the presence of anhydrous NaOAc. Tetraacetates 11a–d, which were still in the
form of anomeric mixtures, were subjected to column chromatography to separate the desired b-anomers 12a–d.
Treatment of tetraacetates 12a–d with MeONa in methanol at room temperature smoothly cleavaged all the acetyl
groups to afford the desired products 1a–d [8] after neutralizing the MeONa in the reaction mixture with strongly
acidic resin in H+ form.
F
OH
OH
MeO
S
OEt
OEt
Cl
OEt
OH
S
O
HO
HO
HO
HO
O
HO
HO
OH
OH
OH
TS071
Dapagliflozin
F
Canagliflozin
YM543
OH
OH
Cl
HO
O
O
O
O
HO
HO
HO
HO
HO
HO
S
OH
OH
OH
ASP1941
LX4211
Fig. 2. Molecular structures of some SGLT2 inhibitors that are now in clinical trials.
OH
R
OEt
O
HO
HO
R = Cl, F, Me
OH
Fig. 3. The SGLT2 inhibitors designed previously in our laboratories.

W.J. Zhao et al. / Chinese Chemical Letters 22 (2011) 1215–1218
1217
OH
OTMS
O
(i)
O
HO
HO
TMSO
TMSO
O
OH
O
OTMS
2
3
R
R
R
R
OEt
(iv)
(ii)
O
(iii)
OH
Br
CO H
Br
Br
Br
CO Et
2
2
4a-4d
5a-5d
6a-6d
7a-7d
OH
OH
R
R
OEt
OEt
O
(vi)
(v)
(vii)
HO
HO
HO
HO
OH
OH
OH
OMe
8a-8d
9a-9d
R
OH
OAc
R
OEt
OEt
OEt
O
O
(ix)
HO
HO
AcO
AcO
(viii)
(x)
OH
OAc
OAc
6
10a-10d
11a-11d
R
R
OH
5
OEt
4
O
O
AcO
AcO
HO
HO
1a: R = 2-Me
1b: R = 4-H
1c: R = 5-Br
1d: R = 6-Me
OAc
OH
2
12a-12d
1a-1d
Scheme 1. Reagents and conditions: (i) TMSCl (6 eq), NMM (8 eq), THF, 0–35 8C; (ii) CH2SO4, EtOH, PhH, reflux; (iii) 2MeMgCl, THF, À5 8C to
0 8C; (iv) AlCl3, PhOEt, CH2Cl2, 0 8C – rt; (v) n-BuLi, THF, À78 8C; then, 3 in PhMe, À78 8C; (vi) MsOH, MeOH, rt; (vii) Et3SiH, BF3ÁEt2O,
CH2Cl2, 0–5 8C and then rt; (viii) NaOAc, Ac2O, AcOH, reflux; (ix) column chromatography; (x) NaOMe, MeOH, rt; then strongly acidic resin (H+
form), rt.
The anti-hyperglycemic activities of 1a–d were evaluated in vivo using mice oral glucose tolerance test (OGTT)
according to a known procedure [5,7], and the results were summarized in Fig. 4 and Table 1. As shown in Table 1, the
anti-hyperglycemic activities of 1a–d were all significant but less potent than the positive control dapagliflozin, and
the inhibition rates were in the following order, 1b > 1d > 1c > 1a, demonstrating that substituents at the 4-postion of
the benzene ring adjacent to the glucose moiety were the most preferred and the order of preference for positions on
this benzene ring is 4-position > 6-position > 5-position > 2-position, which is consistent with the SAR obtained
from analogous structures [1].
In conclusion, a series of gem-dimethyl C-glucosides were designed and synthesized as SGLT2 inhibitors, and the
in vivo anti-hyperglycemic activities were evaluated with mice oral glucose tolerance test, which revealed that
substituents at the 4-position of the benzene adjacent to the glucose moiety were most preferred.
Fig. 4. Changes of mean blood glucose levels following administration of compounds 1a-c.

1218
W.J. Zhao et al. / Chinese Chemical Letters 22 (2011) 1215–1218
Table 1
In vivo inhibition of blood glucose levels of 1a–c.
Compounds
Dapagliflozin
79
1a
1b
1c
1d
Inhibition rates/%
57
72a
65
68b
a
p < 0.05 (vs dapagliflozin).
p < 0.05 (vs dapagliflozin).
b
Acknowledgment
The authors are very grateful to Key Projects of Tianjin Science and Technology Support Plan (No.
10ZCKFSH01300) for financial support.
References
[1] M. Meng, B.A. Ellsworth, A.A. Nirschl, et al. J. Med. Chem. 51 (2008) 1145.
[2] J.T. Link, B.K. Sorensen, Tetrahedron Lett. 41 (2000) 9213.
[3] W.N. Washburn, J. Med. Chem. 52 (2009) 1785.
[4] M. Obermeier, M. Yao, A. Khanna, et al. Drug Metab. Disp. 38 (2010) 405.
[5] Y.H. Shi, G.L. Zhao, Y.Y. Lou, et al. Chin. J. Chem. 29 (6) (2011) 1192.
[6] H. Shao, G.L. Zhao, W. Liu, et al. Chin. J. Synth. Chem 18 (2010) 389.
[7] Y.L. Gao, G.L. Zhao, W. Liu, et al. Indian J. Chem 49B (2010) 1499.
[8] 1H NMR data for 1a–d. 1a, white foam, 1H NMR (DMSO-d6, 400 MHz): d 7.45 (d, 1H, J = 8.0 Hz), 7.28 (d, 1H, J = 7.6 Hz), 7.19 (t, 1H,
J = 7.8 Hz), 7.00 (d, 2H, J = 8.8 Hz), 6.79 (d, 2H, J = 8.4 Hz), 4.88 (d, 1H, J = 4.4 Hz), 4.85 (d, 1H, J = 4.8 Hz), 4.68 (d, 1H, J = 5.6 Hz), 4.38 (t,
1H, J = 5.8 Hz), 4.21 (d, 1H, J = 9.2 Hz), 3.96 (q, 2H, J = 6.9 Hz), 3.64 (dd, 1H, J = 5.4 Hz and 11.4 Hz), 3.31–3.40 (m, 2H), 3.22–3.27 (m, 1H),
3.13–3.17 (m, 2H), 1.80 (s, 3H), 1.58 (s, 3H), 1.57 (s, 3H), 1.17 (t, 3H, J = 7.0 Hz); 1b, white foam, 1H NMR (DMSO-d6, 400 MHz): d 7.04–7.22
(m, 6H), 6.80 (d, 2H, J = 8.8 Hz), 4.89 (t, 2H, J = 4.8 Hz), 4.70 (d, 1H, J = 5.6 Hz), 4.41 (t, 1H, J = 5.8 Hz), 3.94–3.99 (m, 3H), 3.67–3.72 (m,
1H), 3.40–3.46 (m, 1H), 3.23–3.29 (m, 1H), 3.12–3.21 (m, 3H), 1.60 (s, 6H), 1.30 (t, 3H, J = 7.0 Hz); 1c, white foam, 1H NMR (DMSO-d6, 400
MHz): d 7.33 (s, 1H), 7.21 (s, 1H), 7.16 (t, 1H, J = 1.6 Hz), 7.11 (d, 2H, J = 8.8 Hz), 6.81 (d, 2H, J = 8.8 Hz), 4.91 (bs, 2H), 4.82 (bs, 1H), 4.44
(bs, 1H), 3.94–4.00 (m, 3H), 3.68 (d, 1H, J = 11.6 Hz), 3.42–3.46 (m, 1H), 3.14–3.26 (m, 3H), 3.08 (t, 1H, J = 5.6 Hz), 1.58 (s, 6H), 1.30 (t, 3H,
J = 6.8 Hz); 1d, white foam, 1H NMR (DMSO-d6, 400 MHz): d 7.27 (s, 1H), 7.09 (d, 2H, J = 8.8 Hz), 7.00 (d, 1H, J = 8.0 Hz), 6.90 (d, 1H,
J = 8.0 Hz), 6.79 (d, 2H, J = 8.4 Hz), 4.88–4.91 (m, 2H), 4.74 (d, 1H, J = 5.6 Hz), 4.41 (t, 1H, J = 5.6 Hz), 4.21 (d, 1H, J = 9.6 Hz), 3.97 (q, 2H,
J = 6.9 Hz), 3.69 (dd,1H, J = 5.6 Hz and 11.2 Hz), 3.35–3.43 (m, 2H), 3.27–3.30 (m, 1H), 3.20–3.23 (m, 1H), 3.10–3.16 (m, 1H), 2.28 (s, 3H),
1.58 (s, 6H), 1.30 (t, 3H, J = 6.8 Hz).