Design, synthesis and biological activity of cyclohexane-bearing C-glucoside derivatives as SGLT2 inhibitors

Article abstract of DOI:10.1016/j.cclet.2013.03.026

Seven cyclohexane-bearing C-glucoside derivatives (7, 9, 12, 13 and 17-19) were designed and synthesized as SGLT2 inhibitors starting from a potent SGLT2 inhibitor we discovered in earlier work, (1S)-1-deoxy-1-[4-methoxy-3-(trans-n- propylcyclohexyl)methylphenyl]-d-glucose (1). The in vitro and in vivo biological activities were evaluated by hSGLT2/hSGLT1 inhibition and urinary glucose excretion (UGE), respectively. Among the synthesized compounds 12, the 6-deoxy derivative of 1 was the most active and selective SGLT2 inhibitor (IC₅₀ = 1.4 nmol/L against hSGLT2; selectivity = 1576). Compound 12 was a potent SGLT2 inhibitor, which could induce more urinary glucose than 1 and dapagliflozin in UGE.

Full text of DOI:10.1016/j.cclet.2013.03.026

Chinese Chemical Letters 24 (2013) 429–432
Contents lists available at SciVerse ScienceDirect
Chinese Chemical Letters
journal homepage: www.elsevier.com/locate/cclet
Original article
Design, synthesis and biological activity of cyclohexane-bearing C-glucoside
derivatives as SGLT2 inhibitors
Shuo Zhang a,b, Yu-Li Wang b, Qun-Chao Wei b, Wei-Ren Xu b, Li-Da Tang b, Gui-Long Zhao 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
A R T I C L E I N F O
A B S T R A C T
Article history:
Seven cyclohexane-bearing C-glucoside derivatives (7, 9, 12, 13 and 17–19) were designed and
Received 22 January 2013
Received in revised form 1 March 2013
Accepted 11 March 2013
Available online 11 April 2013
synthesized as SGLT2 inhibitors starting from a potent SGLT2 inhibitor we discovered in earlier work,
(1S)-1-deoxy-1-[4-methoxy-3-(trans-n-propylcyclohexyl)methylphenyl]-D-glucose (1). The in vitro and
in vivo biological activities were evaluated by hSGLT2/hSGLT1 inhibition and urinary glucose excretion
(UGE), respectively. Among the synthesized compounds 12, the 6-deoxy derivative of 1 was the most
active and selective SGLT2 inhibitor (IC50 = 1.4 nmol/L against hSGLT2; selectivity = 1576). Compound
12 was a potent SGLT2 inhibitor, which could induce more urinary glucose than 1 and dapagliflozin in
UGE.
Keywords:
Synthesis
C-Glucoside
ß 2013 Gui-Long Zhao, Jian-Wu Wang. Published by Elsevier B.V. on behalf of Chinese Chemical
Society. All rights reserved.
SGLT2 inhibitor
Urinary glucose excretion
Cyclohexane-bearing
1. Introduction
In an earlier work, we discovered a class of novel SGLT2
inhibitors which incorporated
molecules, among which a potent SGLT2 inhibitor, (1S)-1-deoxy-
1-[4-methoxy-3-(trans-n-propylcyclohexyl)methylphenyl]- -glu-
cose (1), was discovered (Fig. 2) [3,4]. Encouraged by these
promising results, we moved on to further study this class of novel
SGLT2 inhibitors in anticipation of discovering more potent SGLT2
inhibitors, and herein would like to report the study on the
compounds resulting from the derivatizations of the 6-OH of the
sugar ring in 1 (Schemes 1–3).
a cyclohexane moiety in the
Diabetes is
a kind of chronic metabolic disease that is
characterized by hyperglycemia, either because the pancreas does
not produce enough insulin and/or because the tissue cells poorly
respond to the insulin that is produced. Diabetes without adequate
treatment can cause a number of severe complications. Although
there are many anti-diabetic drugs currently available, the
hyperglycemia still can not be controlled in many cases, indicating
that new anti-diabetic drugs with novel action mechanisms are
urgently needed.
D
Most 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 [1,2].
Therefore, inhibition of SGLT2 is able to suppress the reabsorption
of glucose from the glomerular filtrate into blood, which will lower
2. Experimental
1H NMR spectra were recorded on a Bruker AV400 spectrometer
with DMSO-d6 as solvent and TMS as internal standard. The HR-MS
data were obtained on an Agilent Q-TOF 6510 mass spectrometer
using electrospray ionization (ESI) technique. The synthetic routes
of target compounds, 7, 9, 12, 13, 17–19, are depicted in Schemes
1–3 and the steps involved are described in ‘‘Results and
discussion’’.
A 50-mL dried flask was charged with 5 mL of dried MeOH and
0.10 g (4 mmol) of sodium, and the mixture was stirred at room
temperature until all the sodium disappeared. Compounds 6, 10,
11 and 14–16 (1 mmol) were added to the solution, individually.
The stirring was continued at room temperature until all the
starting compounds were consumed completely (typical within
the blood glucose levels. SGLT2 inhibitors have become
a
promising class of hypoglycemic agents for the treatment of
type-2 diabetes, and many SGLT2 inhibitors are now in clinical
trials. One of the SGLT2 inhibitors, dapagliflozin, has been
approved recently in EU (Fig. 1).
* Corresponding authors.
E-mail addresses: zhao_guilong@126.com (G.-L. Zhao), jwwang@sdu.edu.cn
(J.-W. Wang).
1001-8417/$ – see front matter ß 2013 Gui-Long Zhao, Jian-Wu Wang. Published by Elsevier B.V. on behalf of Chinese Chemical Society. All rights reserved.
http://dx.doi.org/10.1016/j.cclet.2013.03.026

[
S. Zhang et al. / Chinese Chemical Letters 24 (2013) 429–432
OH
OH
OH
Cl
O
Cl
OEt
F
F
O
O
O
O
HO
HO
HO
HO
HO
HO
S
OH
OH
OH
Empagliflozin
Dapagliflozin
Ipragliflozin
OH
Cl
OH
OH
O
MeO
S
OEt
S
HO
HO
O
HO
HO
HO
HO
OH
O
OH
Tofogliflozin
Canagliflozin
Cl
OH
TS071
OEt
Cl
OEt
O
O
O
O
HO
HO
HO
HO
OH
OH
LX4211
PF-04971729
OH
Fig. 1. Molecular structures of some SGLT2 inhibitors that are now in clinical trials or launched.
[(Fig._2)TD$FIG]
OH
OH
OMe
OMe
R
O
O
HO
HO
HO
HO
OH
OH
1
R = H, Me, Et, n-Pr, i-Pr, n-Bu, t-Bu, n-C5H11
Fig. 2. The cyclohexane-bearing SGLT2 inhibitors discovered previously in our laboratories.
1 h) as demonstrated by TLC. On completion, 2 g of dried acidic
resin was added to the reaction mixture and the stirring was
continued at room temperature until the pH 7. The reaction
mixture was filtered, and the filtrate was evaporated on a rotary
evaporator to afford a white residue as crude products, which were
further dried by vacuum oil pump to yield the pure products 7, 13,
12 and 17–19, respectively.
100 mL of water. The aqueous mixture was adjusted to pH 7 with
concentrated hydrochloric acid and extracted with CH2Cl2
(15 mL ꢀ 3). The combined extracts were washed with saturated
brine, dried over Na2SO4 and evaporated on a rotary evaporator to
afford a residue, which was further dried in vacuo to yield the pure
product 9.
7: White foam, 1H NMR (400 MHz, DMSO-d6):
d 7.10 (dd, 1 H,
A 50-mL flask was charged with 8 (1 mmol) and 10 mL of EtOH.
The mixture was stirred at room temperature, followed by addition
of 2 mL of 30% aqueous NaOH. The mixture thus obtained was
refluxed for 10 min, cooled to room temperature and poured to
J = 2.0 Hz and 8.4 Hz), 7.02 (d, 1H, J = 2.0 Hz), 6.85 (d, 1H,
J = 8.4 Hz), 5.15 (d, 1H, J = 4.8 Hz), 4.97 (d, 1H, J = 4.8 Hz), 4.76
(d, 1H, J = 5.6 Hz), 4.01 (d, 1H, J = 9.6 Hz), 3.73 (s, 3H), 3.52
(dd, 1H, J = 2.0 Hz and 13.2 Hz), 3.42–3.44 (m, 1H), 3.21–3.37
[(Scheme_1)TD$FIG]
OH
OTBDMS
O
OTBDMS
OMe
OMe
OMe
OMe
OMe
iii
ii
i
1
O
O
AcO
AcO
AcO
AcO
HO
HO
OAc
OAc
OH
3
6
4
iv
2
N3
N3
OMs
OMe
vi
v
O
O
O
HO
HO
AcO
AcO
AcO
AcO
OH
OAc
OAc
7
5
vii
OH
OH
N
N
N
N
N
N
OMe
OMe
viii
O
O
HO
HO
AcO
AcO
OH
OAc
8
9
Scheme 1. Synthetic route of target molecule 7 and 9. Reaction conditions: i) TBDMSC l (1.1 eq), imidazole, dried DMF, 0-r.t.; ii) Ac2O, DMAP, pyridine, 0-r.t.; iii) 90% AcOH,
45 8C; iv) MsCl, EtN, CH2Cl2, 0-r.t.; v) NaN3, DMF, 100 8C; vi) MeONa, MeOH, r.t., then strongly acidic ion exchange resin (H+ form); vii) propargyl alcohol, Cu(I), DMF, r.t.; viii) 3
0% NaOH, EtOH, reflux.

[
S. Zhang et al. / Chinese Chemical Letters 24 (2013) 429–432
431
I
OMe
OMe
i
ii
4
O
O
AcO
AcO
AcO
AcO
OAc
OAc
10
11
iii
iii
OMe
I
OMe
O
HO
HO
O
HO
HO
OH
OH
12
13
Scheme 2. Synthetic route of target molecule 12 and 13. Reaction conditions: i) I2, PPh3, imidazole, CH2Cl2, r.t.; ii) n-Bu3SnH, AIBN, PhMe, r.t.; iii) MeONa, MeOH, r.t., then
strongly acidic ion exchange resin (H+ form).
[(Scheme_3)TD$FIG]
R2
R1
OMe
OMe
i
ii
O
4
O
HO
HO
AcO
AcO
OH
OAc
17: R2 = SH
14: R1 = SAc
18: R2 = OPh
19: R2 = SPh
15: R1 = OPh
16: R1 = SPh
Scheme 3. Synthetic route of target molecule 17–19. Reaction conditions: i) R1H, PPh3, DEAD, THF, r.t.; ii) MeONa, MeOH, then strongly acidic ion exchange resin (H+ form).
(m, 3H), 3.11–3.13 (m, 1H), 2.38–2.39 (m, 2H), 1.59–1.66 (m, 4H),
1.40 (bs, 1H), 1.23–1.28 (m, 2H), 1.08–1.11 (m, 3H), 0.74–0.92 (m,
7H); HR-ESI-MS, calcd. for C23H39N4O5 ([M+NH4]+) 451.2920,
found 451.2915.
4.72 (d, 1H, J = 5.6 Hz), 4.25 (d, 1H, J = 10.0 Hz), 3.98–4.04 (m, 2H),
3.72 (s, 3H), 3.55–3.59 (m, 1H), 3.27–3.32 (m, 3H), 3.15–3.21 (m,
1H), 2.44 (dd, 1H, J = 6.8 Hz and 12.8 Hz), 2.34 (dd, 1H, J = 7.0 Hz
and 13.0 Hz), 1.59–1.66 (m, 4H), 1.39 (s, 1H), 1.21–1.28 (m, 2H),
1.06–1.11 (m, 3H), 0.78–0.96 (m, 7H); HR-ESI-MS, calcd. for
9: White foam, 1H NMR (400 MHz, DMSO-d6):
d 7.71 (s, 1H),
7.04–7.07 (m, 1H), 6.95 (s, 1H), 6.84 (d, 1H, J = 8.4 Hz), 5.34 (d, 1H,
J = 5.2 Hz), 5.09 (t, 1H, J = 5.6 Hz), 5.02 (d, 1H, J = 4.8 Hz), 4.77 (d,
1H, J = 6.0 Hz), 4.69 (d, 1H, J = 12.0 Hz), 4.45–4.46 (m, 2H), 4.46 (d,
2H, J = 5.6 Hz), 3.95 (d, 1H, J = 9.6 Hz), 3.73 (s, 3H), 3.59–3.62 (m,
1H), 3.03–3.09 (m, 2H), 1.58–1.67 (m, 4H), 1.39 (s, 1H), 1,23–1,29
(m, 2H), 1.09–1.12 (m, 3H), 0.76–0.93 (m, 7H); HR-ESI-MS, calcd.
for C26H40N3O6 ([M+H]+) 490.2917, found 490.2910.
C
29H44NO6 ([M+NH4]+) 502.3169, found 502.3170.
19: White foam, 1H NMR (400 MHz, DMSO-d6):
d 7.31 (d, 2H,
J = 7.6 Hz), 7.25 (t, 2H, J = 7.6 Hz), 7.13 (t, 1H, J = 7.2 Hz), 7.05 (dd,
1H, J = 1.6 Hz and 8.4 Hz), 6.97 (d, 1H, J = 1.6 Hz), 6.84 (d, 1H,
J = 8.4 Hz), 5.21 (d, 1H, J = 4.4 Hz), 4.96 (s, 1H), 4.72 (d, 1H,
J = 5.6 Hz), 3.95 (d, 1H, J = 9.6 Hz), 3.73 (s, 3H), 3.42–3.45 (m, 2H),
3.22–3.24 (m, 2H), 3.11–3.17 (m, 1H), 3.02 (dd, 1H, J = 8.0 Hz and
14.0 Hz), 2.44 (dd, 1H, J = 6.8 Hz and 12.8 Hz), 2.33 (dd, 1H,
J = 7.0 Hz and 13.0 Hz), 1.60–1.67 (m, 4H), 1.39 (s, 1H), 1.21–1.30
(m, 3H), 1.07–1.12 (m, 3H), 0.81–0.96 (6H); HR-ESI-MS, calcd. for
12: White foam, 1H NMR (400 MHz, DMSO-d6):
d 7.07 (dd, 1H,
J = 2.0 Hz and 8.4 Hz), 6.97 (d, 1H, J = 2.0 Hz), 6.84 (d, 1H,
J = 8.4 Hz), 4.91 (d, 1H, J = 5.6 Hz), 4.84 (d, 1H, J = 4.4 Hz), 4.62
(d, 1H, J = 5.2 Hz), 3.91 (d, 1H, J = 9.2 Hz), 3.73 (s, 3H), 3.12–3.28 (m,
3H), 2.89–2.94 (m, 1H), 2.44 (dd, 1H, J = 6.8 Hz and 12.8 Hz), 2.35
(dd, 1H, J = 7.2 Hz and 12.8 Hz), 1.59–1.67 (m, 4H), 1.39 (s, 1H),
1.23–1.29 (m, 3H), 1.07–1.15 (m, 5H), 0.88–0.96 (m, 7H); HR-ESI-
MS, calcd. for C23H40NO5 ([M+NH4]+) 410.2906, found 410.2916.
C
29H44NO5S ([M+NH4]+) 518.2940, found 518.2926.
3. Results and discussion
As shown in Scheme 1, the starting material 1 [3] was treated
with 1.1 eq. of TBDMSCl (t-butyldimethylsilyl chloride) in the
presence of imidazole in DMF to regioselectively protect its 6-OH
with TBDMS to give 2, which was further peracetylated to afford
triacetate 3 by treatment with Ac2O in pyridine in the presence of
DMAP (4-dimethylaminopyridine) at room temperature [5]. The
TBDMS in 3 was then cleaved by 90% aqueous AcOH at 45 8C to
yield 4. Conversion of 4 to its mesylate 5 was followed by SN2
substitution of mesylate with azide to furnish the azide 6.
Deacetylation of 6 by treatment with MeONa afforded 7. Click
reaction of azide 7 and propargy alcohol in the presence of Cu(I)
generated by in situ reduction of CuSO4 with ascorbic acid
produced the 1,2,3-triazole 8 [6], which afforded 9 on saponifica-
tion with aqueous NaOH in refluxing EtOH. Compound 4 was
converted to iodide 10 by treatment with I2/PPh3/imidazole.
Reduction of iodide 10 with n-Bu3SnH in the presence of AIBN
(azobisisobutyronitrile) in toluene gave 11, which was deacety-
lated with MeONa to yield 12. Direct deacetylation of 10 with
MeONa afforded 13. Mitsunobu reactions of 4 and AcSH, PhOH and
PhSH gave 14–16, respectively. Compounds 14–16 were deacety-
lated with MeONa to yield 17–19, respectively.
13: White foam, 1H NMR (400 MHz, DMSO-d6):
d 7.11 (dd, 1H,
J = 2.0 Hz and 8.4 Hz), 7.02 (d, 1H, J = 2.0 Hz), 6.87 (d, 1H,
J = 8.4 Hz), 5.16 (bs, 1H), 4.78 (bs, 2H), 4.04 (d, 1H, J = 9.2 Hz),
3.74 (s, 3H), 3.52 (dd, 1H, J = 2.6 Hz and 10.6 Hz), 3.39 (dd, 1H,
J = 5.2 Hz and 10.4 Hz), 3.31 (t, 1H, J = 8.8 Hz), 3.08–3.15 (m, 2H),
2.95–2.99 (m, 1H), 2.34–2.46 (m, 2H), 1.64–1.67 (m, 4H), 1.41 (s,
1H), 1.21–1.30 (m, 2H), 1.07–1.12 (m, 3H), 0.83–0.97 (m, 7H); HR-
ESI-MS, calcd. for
541.1431.
C
23H35INaO5 ([M+Na]+) 541.1427, found
17: White foam, 1H NMR (400 MHz, DMSO-d6):
d
7.10 (dd, 1H,
J = 1.8 Hz and 8.2 Hz), 7.00 (d, 1H, J = 1.6 Hz), 6.86 (d, 1H,
J = 7.6 Hz), 5.06 (d, 1H, J = 5.2 Hz), 4.92 (d, 1H, J = 4.0 Hz), 4.70
(d, 1H, J = 6.0 Hz), 3.96 (d, 1H, J = 9.6 Hz), 3.11–3.29 (m, 4H), 2.85–
2.91 (m, 1H), 2.56–2.63 (m, 1H), 2.34–2.47 (m, 2H), 2.00 (t, 1H,
J = 7.6 Hz), 1.61–1.67 (m, 4H), 1.38–1.42 (m, 1H), 1.21–1.30 (m,
2H), 1.07–1.18 (m, 3H), 0.76–0.98 (m, 7H); HR-ESI-MS, calcd. for
C
23H40NO5S ([M + NH4]+) 442.2627, found 442.2621.
18: White foam, 1H NMR (400 MHz, DMSO-d6):
7.23–7.27 (m,
d
2H), 7.09 (dd, 1H, J = 2.0 Hz and 8.4 Hz), 6.98 (d, 1H, J = 2.0 Hz),
6.84–6.93 (m, 4H), 5.18 (d, 1H, J = 4.8 Hz), 4.98 (d, 1H, J = 4.0 Hz),

432
S. Zhang et al. / Chinese Chemical Letters 24 (2013) 429–432
Table 1
and SPh, were not well tolerated. Noteworthy is that compound 12
could induce more urinary glucose excretion than dapagliflozin,
although their IC50s against hSGLT2 were comparable (1.4 nmol/L
vs 1.3 nmol/L), presumably because 12 possessed better pharma-
cokinetic properties than dapagliflozin as determined by in vivo
evaluation.
hSGLT1 and hSGLT2 inhibitory potency and UGE results for the synthesized SGLT2
inhibitors.
Compoundsa
IC50 (nmol/L)
hSGLT2
Selectivity
Urinary glucose
(mg/200 g)c
b
hSGLT1
Vehicle
–
1.3
–
1289
3834
4766
7812
2206
8331
2987
9947
9701
–
992
609
31
13 ꢁ 2
1612 ꢁ 102
1633 ꢁ 87
620 ꢁ 76
541 ꢁ 69
1844 ꢁ 71
673 ꢁ 107
994 ꢁ 97
319 ꢁ 45
336 ꢁ 42
Dapagliflozin
4. Conclusion
1
6.3
7
155
456
1.4
In conclusion, modifications at the 6-OH position in the sugar
ring of (1S)-1-deoxy-1-[4-methoxy-3-(trans-n-propylcyclohexyl)-
methylphenyl]-D-glucose (1), a potent SGLT2 inhibitor discovered
earlier in our laboratories, resulted in the discovery of a much more
9
140
1576
37
12
13
17
18
19
223
79
38
207
358
48
potent SGLT2 inhibitor 12, which was the 6-deoxy derivative of 1.
27
a
Dose: 10 mg/kg (po).
b
c
Acknowledgment
IC50 (hSGLT1)/IC50 (hSGLT2).
The urinary glucose excreted 24 h post-challenge (mean ꢁ SD) is expressed as
mg/200 g of rat weight (n = 10).
The authors are very grateful to Key Projects of Tianjin Science
and Technology Support Plan (No. 10ZCKFSH01300) for financial
support.
The in vitro (hSGLT2 and hSGLT1 inhibitory activity) and in vivo
hypoglycemic activity (UGE, urinary glucose excretion) of the
synthesized compounds were evaluated according to known
methods [1] and summarized in Table 1. As shown in Table 1,
all the synthesized compounds were selective SGLT2 inhibitors (all
selectivities >30), with 12 being the most selective one (1576),
which was even more selective than 1 (609) and dapagliflozin
(992). 12 was most active against hSGLT2 (1.4 nmol/L), which was
more active than 1 (1.4 nmol/L vs 6.3 nmol/L) and comparable with
dapagliflozin (1.4 nmol/L vs 1.3 nmol/L), indicating that the 6-
deoxylation could be well tolerated in this case. 12 was also the
most potent one in in vivo evaluation, which was significantly more
potent than 1 (1844 vs 1633, P < 0.05) and dapagliflozin (1844 vs
1612, P < 0.05). The other synthesized compounds were much less
potent, indicating that except for 6-deoxylation, other modifica-
tions at 6-OH, including introduction of small functional groups,
such as N3, SH and I and bulky groups, such as 1,2,3-triazole, OPh
References
[1] M. Meng, B.A. Ellsworth, A.A. Nirschl, et al., Discovery of dapagliflozin: a potent,
selective renal sodium-dependent glucose cotransporter 2 (SGLT2) inhibitor for
the treatment of type 2 diabetes, J. Med. Chem. 51 (2008) 1145–1149.
[2] W.N. Washburn, Development of the renal glucose reabsorption inhibitors: a new
mechanism for the pharmacotherapy of diabetes mellitus type 2, J. Med. Chem. 52
(2009) 1785–1794.
[3] H. Shao, Y.L. Gao, Y.Y. Lou, et al., Design, synthesis and antihyperglycemic activity of
trans-cyclohexane-bearing C-glucosides as SGLT2 inhibitors, Chin. J. Org. Chem. 31
(2011) 836–842.
[4] W.J. Zhao, X.L. Yu, H. Shao, et al., Synthesis and anti-hyperglycemic activities of
trans-cyclohexane-bearing C-glucosides as SGLT2 inhibitors II, Chin. J. Synth.
Chem. 20 (2012) 527–536.
[5] K.J. Liao, X.F. Jin, X.B. Meng, et al., Synthesis of an antimetastatic tetrasaccharide b-
D-Gal-(1 ! 4)-b-D-GlcpNAc-(1 ! 6)-a-D-Manp-(1 ! 6)-b-D-Manp-OMe,
Chin.
Chem. Lett. 23 (2012) 1371–1374.
[6] V.V. Rostovtsev, L.G. Green, V.V. Fokin, et al., A stepwise huisgen cycloaddition
process: copper(I)-catalyzed regioselective ‘‘ligation’’ of azides and terminal
alkynes, Angew. Chem. Int. Ed. Engl. 41 (2002) 2596–2599.