Modification on the O-glucoside of Sergliflozin-A: A new strategy for SGLT2 inhibitor design

Article abstract of DOI:10.1016/j.bmcl.2016.03.065

Poor pharmacokinetic stability is one of the issues of O-glucoside SGLT2 inhibitors in clinical trials, hence C-glucoside inhibitors have been developed and extensively applied. Herein, we provided an alternative approach to improve the pharmacokinetic st

Full text of DOI:10.1016/j.bmcl.2016.03.065

Bioorganic & Medicinal Chemistry Letters xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Modification on the O-glucoside of Sergliflozin-A: A new strategy for
SGLT2 inhibitor design
a,
Xuefeng Cao a,y, Wenpeng Zhang b,y, Xu Yan a, Zhi Huang a, Zhenqing Zhang b, Peng Wang a, , Jie Shen
⇑
⇑
a State Key Laboratory of Medicinal Chemical Biology, College of Pharmacy and Tianjin Key Laboratory of Molecular Drug Research, Nankai University, Tianjin 300071, PR China
b State Key Laboratory of Toxicology and Medical Countermeasures, Beijing Institute of Pharmacology and Toxicology, Beijing 100850, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Poor pharmacokinetic stability is one of the issues of O-glucoside SGLT2 inhibitors in clinical trials, hence
C-glucoside inhibitors have been developed and extensively applied. Herein, we provided an alternative
approach to improve the pharmacokinetic stability of such inhibitors. Nine derivatives of Sergliflozin-A
with modifications on the O-glucoside fragment were prepared, among which the 4-O-methyl derivative
exhibited similar pharmacodynamics potency in excreted glucose urine test. Most attractively, signifi-
cantly increased pharmacokinetic stability was observed for 4-O-methyl derivative of O-glucosides.
This work proved that modification on the O-glucoside fragment could be a promising approach to the
future SGLT2 inhibitor design.
Received 18 February 2016
Revised 14 March 2016
Accepted 16 March 2016
Available online xxxx
Keywords:
SGLT2 inhibitor
Sergliflozin-A
Ó 2016 Elsevier Ltd. All rights reserved.
O-Glucoside
Pharmacokinetic stability
During the past few years, great efforts have been devoted to
develop sodium glucose co-transporter 2 (SGLT2) inhibitors as a
new class of drugs for the treatment of diabetes.1–8 These com-
pounds (Fig. 1) could effectively relieve diabetic hyperglycemia
via lowering the renal glucose reabsorption.9–11 Phlorizin, isolated
from the bark of apple roots, was the first O-glucoside compound
which can inhibit both SGLT1 and SGLT2.12,13 Other O-glucoside
inhibitors with high selectivity towards SGLT2 were also devel-
oped. However, most of O-glucoside inhibitors, for example, T-
109514 and Sergliflozin,15,16 were suspended. Currently, majority
of the launched SGLT2 inhibitors, e.g., Dapagliflozin,4
Empagliflozin6,17 and Canagliflozin,7,18 are C-glucosides. This could
at least partially be attributed to the poor metabolic stability of O-
glucosides, arising from the glucosidase mediated hydrolysis.
Alternatively, it is understandable that modification on the O-
glucoside fragment may affect the substrate-glucosidase binding,
thus could also help to improve in vivo stability. In this work, tak-
ing Sergliflozin-A (active form of Sergliflozin) as a parent structure,
a series of derivatives with modification on hydroxyl groups (e.g.,
methylated, fluorine-substituted, deoxidized) of its O-glycoside
were synthesized (Fig. 2) and evaluated.
mide donor A0 was coupled with the aglycon acceptor S219 to
afford the intermediate A1 which was deacetylated to give the
compound A, Sergliflozin-A. Promoted by boron trifluoride ether-
ate, the glycosyl donor B019 was coupled with S2 to provide the
intermediate B1, which was sequentially deacetylated and methy-
lated at its O2 position to afford the 2-O-methyl 3,4,6-tri-O-benzyl
derivative B3. Pd-catalyzed debenzylation of B3 gave the com-
pound B. The 2-fluorinated glucosyl Schimidt donor C019 was cou-
pled with S2 to provide the intermediate C1 as an unseparable a/b
mixture.19 After the sequential debenzylation, acetylation and flash
chromatography separation, C1 was converted to the intermediate
C3 as a pure b isomer, which was converted to the compound C by
the global deacetylation. The compounds D, E, I were prepared
from the 1-acetylated donors D0,19 E019 and I019 respectively via
a similar two-step procedure, including a glycosylation reaction
followed by a global deacetylation. The compound F was prepared
from compound A in 5 steps. Specifically, after the 4,6-benzylide-
nation and 2,3-benzylation reactions, A was converted to the inter-
mediate F2. Regioselective and reductive ring opening of the 4,6-
benzylidene of F2 gave the intermediate F3 whose 4-hydroxyl
group was then methylated with MeI. Then the obtained interme-
diate was converted to the product F via global debenzylation.
Regioselective benzoylation of the thioglycoside G019 gave the
intermediate G1 whose 4-hydroxyl group was thiocarbonylated
to give the intermediate G2. Barton–McCombie radical reduction
was then performed to convert G2 to the 4-deoxy intermediate,
which was converted to 1-acetylated donor compound G3. Then
G3 was coupled with S2 to afford the glycosylation product G4,
The synthesis of Sergliflozin-A and its nine derivatives were
illustrated in Scheme 1. Briefly, the peracetylated
a-glucosyl bro-
⇑
Corresponding authors.
E-mail addresses: pwang@nankai.edu.cn (P. Wang), jieshen@nankai.edu.cn (J.
Shen).
y
Authors contributed equally to this work.
http://dx.doi.org/10.1016/j.bmcl.2016.03.065
0960-894X/Ó 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Cao, X.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.03.065

2
X. Cao et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
OH
OH
O
O
HO
O
OH
O
OMe
O
O
O
O
O
RO
HO
RO
HO
HO
OH
OH
OH
HO
OH
R=-H,
OH
R=-H,
OH
Phlorizin
Sergliflozin-A
T-1095A
R=CH3OCO-, T-1095
R=CH3CH2OCO-, Sergliflozin
Cl
O
Cl
O
O
O
O
O
HO
HO
HO
HO
HO
S
OH
OH
OH
HO
F
OH
OH
OH
Dapagliflozin
Canagliflozin
Empagliflozin
Figure 1. Important SGLT2 inhibitors.
OH
O
OH
O
OH
O
OH
O
HO
HO
HO
HO
HO
HO
HO
H3CO
OR
OR
OR
OR
OR
OH
OCH3
F
OH
A
B
C
D
(2-OMe)
(3-OMe)
(Sergliflozin-A)
(2-F)
OH
O
OH
O
OH
O
OCH3
O
HO
H3CO
H
HO
HO
OR
OR
OR
H
HO
HO
OH
OH
OH
OH
E
G
F
H
(3-deoxy)
(6-OMe)
(4-deoxy)
(4-OMe)
H
O
F
O
OMe
HO
HO
HO
HO
OR
OR
R =
OH
I
OH
J
(6-deoxy)
(6-F)
Figure 2. Sergliflozin-A and its 9 derivatives.
which was converted to the compound G via the global de-esteri-
fication. Syntheses of the last two compounds H and J shared the
same intermediated H1, which was prepared from A via regioselec-
tive protection of 6-hydrxoyl group with trityl chloride. For the
synthesis of H, H1 was per-benzylated to afford the intermediate
H2, whose trityl group at O6 was replaced with a methyl group
in two steps. The therefore obtained intermediated H4 was con-
verted to the compound H via global debenzylation. For the syn-
thesis of J, H1 was per-acetylated to afford the intermediate J2.
After the trityl group of J2 was removed, DAST reagent was utilized
to accomplish the fluorine substitution reaction at C6, so as to con-
vert J3 to J4, whose global deacetylation led to the product J.
As for the glycosylation reactions that promoted by boron triflu-
oride etherate to give products B1, D1, E1, I1 and G4, product iso-
merization at 0 °C was observed for B1 which is an armed
glycoside. As the kinetically favored b-isomer, B1 was formed as
cation at these two hydroxyl groups (compounds B, C, D, E)
almost completely abolished the hypoglycemic effect. In contrast,
4-methylated derivative F and the 6-deoxy derivative H are the
two most potent derivatives, which maintained 73% and 31%
in vivo potency compare with the parent compound A. These
observations were in good consistence with structural data, where
both 2- and 3-hydroxyl groups formed at least two hydrogen
bonds with the bacterial vSGLT (PDB# 3DH4)20,21 or human
SGLT222 in their substrate recognition pocket. On the contrary,
the 4- and 6-hydroxyl groups bound much loosely to the pocket
and each of them formed only one hydrogen bond in both
structures.
We also envisioned that pharmacokinetic factors might also
benefit the in vivo potency of the compound F. To compare the
metabolic stability of glucoside and its 4-methylated derivative,
two fluorescent compounds K and L were prepared and treated
in rat liver homogenate for 8 h. Whereas over 80% glucoside K
was hydrolyzed, neglected amount of hydrolysis product could
be detected for the corresponding 4-methylated compound L
(Fig. 4). It should be addressed that the poor metabolic stability
was one of the reasons for the failure of O-glucoside SGLT2 inhibi-
tors in clinical trials.
the major product initially with little amount of
ever, it was gradually converted to its more stable
a
-isomer. How-
a-isomer. There-
fore whereas the glycosylation of the disarmed D1 was performed
at 0 °C, the other the other four reactions were performed at
ꢀ20 °C.
Rat urinary glucose excretion experiments were then per-
formed to evaluate the in vivo hypoglycemic effects of the ten
compounds A–J (Fig. 3). The results clearly indicated that the 2-
hydroxyl group and 3-hydroxyl group were essential to the inhibi-
tion potency of the compound A against SGLT2, since any modifi-
It is now well accepted that the topological polar surface area
(tPSA) of an orally administrated drug should be less than
140 Å2. Considering the tPSA value of a glucoside fragment is
around 100 Å2, it is of no wonder that the glycons of all SGLT2 inhi-
Please cite this article in press as: Cao, X.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.03.065

X. Cao et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
3
OAc
O
OAc
O
OH
OMe
OMe
O
AcO
AcO
AcO
HO
1
( )
a
OR
b
OR
AcO
R =
HO
AcO
Br
OAc
OH
OH
S2
A0
1
A
A
OBn
O
OBn
O
OBn
O
OH
OBn
O
O
BnO
c
b
BnO
HO
HO
BnO
BnO
d
e
2
( )
BnO
BnO
OR
OAc
OR
OR
OR
BnO
BnO
OH
OMe
OAc
B1
OMe
B3
OAc
B0
B
B2
NH
OBn
O
OH
OH
O
OBn
O
OAc
O
O
3
HO
HO
HO
HO
( ) BnO
BnO
BnO
AcO
AcO
g
f
OR
e
OR
OR
OR
b
O
CCl3
BnO
F
F
F
F
F
C
C3
C0
C1
C2
R2
O
R2
O
R2
O
AcO
R1
HO
R1
4
( )
AcO
E1,I1
D1
c for
h for
OAc
OR
OR
R1
b
OAc
OH
OAc
1
2
1
2
1
2
D
D0
D1
R =-OMe,R =-OH,
R1=-H,R2=-OH,E
R1=-OH,R2=-H,I
R =-OMe,R =-OAc,
R1=-H,R2=-OAc,E0
R1=-OAc,R2=-H,I0
R =-OMe,R =-OAc,
R1=-H,R2=-OAc, E1
R1=-OAc, R2=-H, I1
OH
OH
OBn
O
Ph
Ph
O
O
BnO
O
O
O
O
HO
O
O
MeO
HO
HO
(5)
(6)
j
i
OR
OR
OR
OR
k
OR d,e
HO
BnO
HO
OH
OBn
F2
OH
F
OH
OBn
A
F1
F3
S
OBz
OBz
O
OH
O
OBz
O
OBz
O
O
HO
H
HO
AcO
O
H
PhO
l
STol
OR
OAc
STol
STol
OR
n, o
c
AcO
m
AcO
AcO
AcO
OAc
OAc
G3
OAc
G0
OAc
G2
OAc
G1
G4
OH
O
H
b
HO
OH
G
R4
O
OTrt
OH
O
OH
O
OTrt
R3O
R3O
R3O
R3O
O
O
HO
HO
R3O
HO
HO
H2
J2
j for
g for
H4
J4
p
OR
q
d for
r for
OR
OR
OR
R3O
OR
(7)
OR3
H3
OH
OH
OR3
H2
OR3
3
3
3
4
R =-OBn,
R3=-OAc,J2
R =-OBn,
A
H1
H4
R =-OBn, R =-OMe,
R3=-OAc,J3
R3=-OAc, R4=-F,J4
R4
O
HO
H
J
e for
b for
OR
HO
OH
4
H
R =-OMe,
R4=-F,J
Scheme 1. Chemical synthesis route of compound A–J. Reagents and conditions: (a) S2, NaOH, TBAB, 53%; (b) NaOMe, MeOH, A 87%, B2 53% 2 steps from B0, C 88%, D 79% 2
steps from D0, E 63% 2 steps from E0, I 85%, G 58% 3 steps from G2, J 78%; (c) S2, BF3–Et2O, ꢀ20 °C for 24 h. I1 47%, B1, E1, G4 without purification; (d) MeI, NaH, B3 93%, H4
96%, F4 90%; (e) Pd/C, H2, B 83%, C2 without purification, F 80%, H 81%; (f) S2, BF3–Et2O, ꢀ40 °C to rt without purification; (g) Ac2O, pyridine, C3 66% 3 steps from C0, J2 91% 2
steps from J0; (h) S2, BF3–Et2O, 0 °C to rt without purification; (i) PhCH(OMe)2, CSA, 80%; (j) NaH, BnBr, F2 95%, H2 96%; (k) Et3SiH, CF3COOH, 80%; (l) BzCl, pyridine, 80%; (m)
carbonochloridothioic acid, O-phenyl ester, DMAP, 85%; (n) Bu3SnO, AIBN, without purification; (o) NIS, AcOH without purification; (p) TrtCl, pyridine, 68%; (q) 33% HBr in
HOAc, H3 85%, J3 74%; (r) DAST, s-collidine, 63%.
bitors in clinical trials have no more than three N and O atoms in
sum. Therefore, depolarization modification on the glucoside
might provide a chance to introduce a polar functional group at
the aglycon fragment without affecting intestinal absorption. As
expected, The logP values of A and F were experimentally deter-
mined to be 1.32 and 1.83, and their tPSA were calculated to be
109 and 98 Å2,23 indicating that 4-methylated derivative F more
suitable to accept a polar functional group onto its aglycon. The
current work therefore provided a new approach to new SGLT2
inhibitors.
Taken together, nine derivatives with modifications on O-glu-
coside fragment of Sergliflozin-A were prepared and the excreted
glucose urine experiments were performed. The results showed
the 2-OH and 3-OH are essential for SGLT2 inhibition potency,
and modifications on the two hydroxyl groups completely abol-
ished the inhibition. In contrast, 4-O-methyl compound F showed
Figure 3. Excreted glucose urine results of ten compounds.
almost equal SGLT2 inhibition compared to Sergliflozin-A. It should
Please cite this article in press as: Cao, X.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.03.065

4
X. Cao et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
References and notes
1. Tsujihara, K.; Hongu, M.; Saito, K.; Kawanishi, H.; Kuriyama, K.; Matsumoto, M.;
Oku, A.; Ueta, K.; Tsuda, M.; Saito, A. J. Med. Chem. 1999, 42, 5311.
2. Ehrenkranz, J. R.; Lewis, N. G.; Ronald Kahn, C.; Roth, J. Diab. Metab. Res. Rev.
2005, 21, 31.
3. Katsuno, K.; Fujimori, Y.; Takemura, Y.; Hiratochi, M.; Itoh, F.; Komatsu, Y.;
Fujikura, H.; Isaji, M. J. Pharmacol. Exp. Ther. 2007, 320, 323.
4. Han, S.; Hagan, D. L.; Taylor, J. R.; Xin, L.; Meng, W.; Biller, S. A.; Wetterau, J. R.;
Washburn, W. N.; Whaley, J. M. Diabetes 2008, 57, 1723.
5. Meng, W.; Ellsworth, B. A.; Nirschl, A. A.; McCann, P. J.; Patel, M.; Girotra, R. N.;
Wu, G.; Sher, P. M.; Morrison, E. P.; Biller, S. A.; Zahler, R. J. Med. Chem. 2008, 51,
1145.
6. Grempler, R.; Thomas, L.; Eckhardt, M.; Himmelsbach, F.; Sauer, A.; Sharp, D. E.;
Bakker, R. A.; Mark, M.; Klein, T.; Eickelmann, P. Diab. Obes. Metab. 2012, 14, 83.
7. Lamos, E. M.; Younk, L. M.; Davis, S. N. Expert Opin. Drug Metab. Toxicol. 2013, 9,
763.
8. Miao, Z.; Nucci, G.; Amin, N.; Sharma, R.; Mascitti, V.; Tugnait, M.; Vaz, A. D.;
Callegari, E.; Kalgutkar, A. S. Drug Metab. Dispos. 2013, 41, 445.
9. Ahmed, D.; Sharma, M.; Kumar, V.; Subhashchandra, Y. P. J. Diab. Metab. 2014,
5, 358.
10. Idris, I.; Donnelly, R. Diab. Obes. Metab. 2009, 11, 79.
11. Verspohl, E. J. Pharmacol. Rev. 2012, 64, 188.
12. Panayotova-Heiermann, M.; Loo, D. D.; Wright, E. M. J. Biol. Chem. 1995, 270,
27099.
13. Kanai, Y.; Lee, W. S.; You, G.; Brown, D.; Hediger, M. A. J. Clin. Invest. 1994, 93,
397.
14. Isaji, M. Curr. Opin. Investig. Drugs 2007, 8, 285.
Figure 4. The pharmacokinetic stability of compound K and compound L.
15. Hussey, E. K.; Clark, R. V.; Amin, D. M.; Kipnes, M. S.; O’Connor-Semmes, R. L.;
O’Driscoll, E. C.; Leong, J.; Murray, S. C.; Dobbins, R. L.; Nunez, D. J. Diabetes
2007, 56, A189.
16. Dobbins, R. L.; O’Connor-Semmes, R.; Kapur, A.; Kapitza, C.; Golor, G.;
Mikoshiba, I.; Tao, W.; Hussey, E. K. Diab. Obes. Metab. 2012, 14, 15.
17. Luippold, G.; Klein, T.; Mark, M.; Grempler, R. Diab. Obes. Metab. 2012, 14, 601.
18. Nomura, S.; Sakamaki, S.; Hongu, M.; Kawanishi, E.; Koga, Y.; Sakamoto, T.;
Yamamoto, Y.; Ueta, K.; Kimata, H.; Nakayama, K.; Tsuda-Tsukimoto, M. J. Med.
Chem. 2010, 53, 6355.
be emphasized that the stability of 4-O-methyl compound L was
significantly improved. Meanwhile, due to the decreased molecular
polarity through such a modification, it provided a chance to
explore SGLT2 inhibitors with a polar functional group at its agly-
con in the future.
19. For preparation details see Supporting information.
20. Watanabe, A.; Choe, S.; Chaptal, V.; Rosenberg, J. M.; Wright, E. M.; Grabe, M.;
Abramson, J. Nature 2010, 468, 988.
Acknowledgment
21. Enkavi, G.; Li, J.; Wen, P.; Thangapandian, S.; Moradi, M.; Jiang, T.; Han, W.;
Tajkhorshid, E. Annu. Rep. Comput. Chem. 2014, 10, 77.
22. Xu, J.; Yuan, H.; Ran, T.; Zhang, Y.; Liu, H.; Lu, S.; Xiong, X.; Xu, A.; Jiang, Y.; Lu,
T.; Chen, Y. J. Mol. Recognit. 2015, 28, 467.
This study was supported by National Natural Science Founda-
tion of China (No. 21272127).
23. Predicted by ChemDraw 12.0.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2016.03.
065.
Please cite this article in press as: Cao, X.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.03.065