In silico design, synthesis and evaluation of 3′-O-benzylated analogs of salacinol, a potent α-glucosidase inhibitor isolated from an Ayurvedic traditional medicine salacia

Article abstract of DOI:10.1039/c2cc34144a

With the aid of an in silico method, α-glucosidase inhibitors with far more potent activities than salacinol (1), a potent natural α-glucosidase inhibitor isolated from an Ayurvedic traditional medicine Salacia reticulata, have been developed.

Full text of DOI:10.1039/c2cc34144a

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 8646–8648
www.rsc.org/chemcomm
COMMUNICATION
In silico design, synthesis and evaluation of 30-O-benzylated analogs of
salacinol, a potent a-glucosidase inhibitor isolated from an Ayurvedic
traditional medicine ‘‘Salacia’’w
Genzoh Tanabe,a Shinya Nakamura,a Nozomi Tsutsui,a Gorre Balakishan,a Weijia Xie,b
Satoshi Tsuchiya,a Junji Akaki,c Toshio Morikawa,c Kiyofumi Ninomiya,c Isao Nakanishi,a
Masayuki Yoshikawad and Osamu Muraoka*a
Received 9th June 2012, Accepted 6th July 2012
DOI: 10.1039/c2cc34144a
With the aid of an in silico method, a-glucosidase inhibitors with
far more potent activities than salacinol (1), a potent natural
a-glucosidase inhibitor isolated from an Ayurvedic traditional
medicine Salacia reticulata, have been developed.
(hNtMGAM) in complex with 2 and 5 and/or in silico docking
studies9b of 1 with hNtMGAM indicated that the 30-O-sulfate
anion of 1 was constrained by the hydrophobic residues of the
enzyme, and made no hydrogen bonding interactions with
them. Actually, the introduction of a hydrophobic group
at the 30 position in 1 enhanced the activity to some extent.9c
In the present study, with the aid of an in silico docking study,
a series of 30-O-benzylated analogs (7b–7m), in which the
characteristic sulfate moiety of 1 was replaced by mono-
substituted benzyl groups, were designed as more potent
inhibitors. The synthesis and evaluation led to a compound
(7k) ca. forty times as potent as the natural inhibitor (1)
(Fig. 1).
In the late 1990s we isolated the highly potent a-glucosidase
inhibitor, salacinol (1), from the roots and stems of Salacia
reticulata, which has usually been used for the treatment of
diabetes in traditional Ayurvedic medicine. The a-glucosidase
inhibitory activity of 1 was revealed to be as potent as those of
voglibose and acarbose, which have been widely used clinically.1
After the discovery of 1, the related sulfonium sulfates,
kotalanol2 (2) and ponkoranol3 (3), and their desulfonated
analogs, neosalacinol4 (4), neokotalanol5 (5), and neoponkoranol6
(6) were subsequently isolated from the same genus plant as
the compounds responsible for the antidiabetic activity. On this
basis, human clinical trials on patients with type-2 diabetes,
conducted with the extract of Salacia reticulata, have shown
effective treatment of type-2 diabetes with minimal side effects.7
On the other hand, owing to both the high inhibitory activity and
the intriguing structure of the constituents (1–6), much attention
has been focused on them, and intensive structure–activity
relationship (SAR) studies on this new class of a-glucosidase
family have been conducted. Through these studies several
inhibitors with measurably better activity have been developed.8
Recent intensive X-ray crystallographic studies9a on the
human N-terminal catalytic domain of maltase-glucoamylase
From the intensive in silico docking studies on salacinol
derivatives, it was predicted that the introduction of chloro or
nitro substituents onto the benzyl moiety of 30-O-benzyl-
neosalacinol (7a) would increase the inhibitory activity [see
Table 1, Ebind (kcal molꢀ1) of the predicted binding affinities
calculated by the MM/GBVI method].10 In this study, to
examine the effect of the position of the substitution in the
a School of Pharmacy, Kinki University, 3-4-1 Kowakae, Higashi-osaka,
Osaka 577-8502, Japan. E-mail: muraoka@phar.kindai.ac.jp
b Department of Medicinal Chemistry, China Pharmaceutical
University, 24 Tongjia Xiang, Nanjing, Jiang su 210009, PR China.
Fax: +86-25-83242366; Tel: +86-25-83271445
c Pharmaceutical Research and Technology Institute, Kinki University,
3-4-1 Kowakae, Higashi-osaka, Osaka 577-8502, Japan.
Fax: +81-6-6721-2505; Tel: +81-6-6721-2332
d Kyoto Pharmaceutical University, 1 Shichono-cho, Misasagi,
Yamashina-ku, Kyoto 607-8412, Japan. Fax: +81-75-595-4768;
Tel: +81-75-595-4633
w Electronic supplementary information (ESI) available: Experimental
procedures for 7k and 13k, and their 1H and 13C NMR spectra. See
DOI: 10.1039/c2cc34144a
Fig. 1 Cyclic sulfoniums as a new class of a-glucosidase inhibitors.
c
8646 Chem. Commun., 2012, 48, 8646–8648
This journal is The Royal Society of Chemistry 2012

View Article Online
Table 1 Ebind (kcal molꢀ1) of 7a–7m, 1, and 4 to hNtMGAM, and
their IC50 (mM) values against rat intestinal disaccharidases
conditions gave the desired epoxides (8b–8m) in 88–75%
yields.
The obtained epoxides 8b–8m (1.2 eq.) were treated with
thiosugar 9 (1.0 eq.) in the presence of HBF4ꢁMe2O or
HBF4ꢁEt2O (1.3 eq.) at ꢀ60 1C to give the corresponding
coupling products (13b–13m) in good yields. In all cases, the
1H NMR spectra showed the formation of a smal1 amount
of diastereoisomers at the sulfonium center (dr Bca. 8/1),
which was consistent with our previous works.4c The anti
relation between the side chain and the hydroxymethyl moiety
on C4 of the major isomer (a-13) was confirmed by means of
nuclear Overhauser effect (NOE) experiments as shown in
Scheme 1.
a
Entry
Compound
Ebind
Maltase
Sucrase
Isomaltase
1
2
3
4
5
6
7
8
7a (H)
ꢀ37.2
ꢀ37.2
ꢀ36.7
ꢀ35.5
ꢀ41.6
ꢀ42.6
ꢀ40.0
ꢀ37.5
ꢀ35.4
ꢀ40.5
ꢀ38.9
ꢀ37.4
ꢀ42.2
ꢀ37.0b
ꢀ36.4b
0.32c
0.66
0.84
0.86
0.31
0.53
0.89
0.33
0.98
0.98
0.13
0.94
0.68
5.2d
8.0d
1.2
0.44c
0.41
1.3
0.14c
0.48
0.35
0.68
0.26
0.31
0.48
0.19
0.25
0.38
0.21
0.23
0.23
1.3d
0.3d
2.1
7b (o-CH3)
7c (m-CH3)
7d (p-CH3)
7e (o-Cl)
7f (m-Cl)
7g (p-Cl)
7h(o-CF3)
7i (m-CF3)
7j (p-CF3)
7k (o-NO2)
7l (m-NO2)
7m (p-NO2)
1
1.1
0.09
0.80
0.72
0.15
0.82
0.72
0.042
0.49
0.38
1.6d
1.3d
0.2
9
10
11
12
13
14
15
16
17
Then the PMB groups of 13b–13m were removed by aqueous
trifluoroacetic acidꢀat rt to give the corresponding sulfonium
salts with the BF4 anion, which was finally exchanged with
the Clꢀ anion (7b–7m) by treating with the ion exchange resin
IRA-400J (Clꢀ form).
4
Voglibose
Acarbosec
1.7e
1.5e
646e
a
Calculation has been performed using the reported scheme.9b
e
Ref. 9b. Ref. 9c. Ref. 3. Ref. 17.
b
c
d
a-Glucosidase inhibitory activities of 7b–7m were tested for
rat small intestinal a-glucosidases in vitro, and compared with
those of salacinol (1), neosalacinol (4) and currently used
antidiabetics (voglibose and acarbose), as shown in Table 1.
All the sulfonium salts (7b–7m) showed superior inhibitory
activities regardless of the substitution position and of the
substituent species. It is interesting to note that the inhibitory
activities of ortho-substituted compounds (7b, 7e, 7h, and 7k)
against maltase and sucrase were stronger than those of the
corresponding meta- and para-substituted analogs, although
the results could not be predicted by the in silico docking
studies. Among the twelve derivatives evaluated in this study,
7k was found to be the most potent, and against maltase,
the strongest inhibitor with approximately a 40-fold activity
relative to 1.
benzene ring and/or the electronic effects of the substituents on
the benzene ring, both of which were difficult to be precisely
evaluated by simulation, three regioisomers (ortho, meta, and
para) with respect to the four kinds of substituents (CH3, Cl,
CF3, and NO2) were synthesized and evaluated to confirm the
computational SAR.
Syntheses of the twelve derivatives (7b–7m) were carried out
by applying the regioselective ring-opening reaction of the
appropriate epoxides 8b–8m with a thiosugar, 2,3,5-tri-O-
(p-methoxybenzyl)-1,4-dideoxy-1,4-epithio-D-arabinitol11 (9).
Firstly, the primary hydroxyl in 1012 was selectively protected
with p-methoxybenzyl (PMB) chloride to give the corre-
sponding PMB ether (11) in 91% yield. The preparation of
methyl-, chloro-, and trifluoromethyl benzyl derivatives
(12b–12j) was accomplished in the usual manner, whereas
nitrobenzyl ethers (12k–12m) were obtained by employing
NaOH as the base.13 The deacetalization of 12b–12m followed
by epoxidation of the resulting glycols under the Mitsunobu
The differences between the computational prediction and
the experimental results in this study might be caused by the
different origin of the enzymes, although rat intestinal maltase
is known to be homologous to hNtMGAM.14 The contri-
bution of other catalytic domains such as C-terminal maltase-
glucoamylase (CtMGAM) and/or sucrase-isomaltase (CtSI),
Scheme 1 Synthesis of compounds 7b–7m.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 8646–8648 8647

View Article Online
We thank MEXT (Ministry of Education, Culture, Sports,
Science and Technology) and Japan Society for the Promotion
of Science (JSPS) for their financial support of this work
through ‘‘High-Tech Research Center’’ Project for Private
Universities: matching fund subsidy from 2007–2011, and
grant-in aid for scientific research, respectively.
Notes and references
1 (a) M. Yoshikawa, T. Murakami, H. Shimada, H. Matsuda,
J. Yamahara, G. Tanabe and O. Muraoka, Tetrahedron Lett.,
1997, 38, 8367; (b) M. Yoshikawa, T. Morikawa, H. Matsuda,
G. Tanabe and O. Muraoka, Bioorg. Med. Chem., 2002, 10,
1547.
2 M. Yoshikawa, T. Murakami, K. Yashiro and H. Matsuda, Chem.
Pharm. Bull., 1998, 46, 1339.
3 M. Yoshikawa, F. Xu, S. Nakamura, T. Wang, H. Matsuda,
G. Tanabe and O. Muraoka, Heterocycles, 2008, 75, 1397.
4 (a) Y. Minami, C. Kuriyama, K. Ikeda, A. Kato, K. Takebayashi,
I. Adachi, W. J. G. Fleet, A. Kettawan, T. Okamoto and
N. Asano, Bioorg. Med. Chem., 2008, 16, 2734; (b) G. Tanabe,
K. Yoshikai, T. Hatanaka, M. Yamamoto, Y. Shao, T. Minematsu,
O. Muraoka, T. Wang, H. Matsuda and M. Yoshikawa, Bioorg.
Med. Chem., 2007, 15, 3926; (c) G. Tanabe, W. Xie, A. Ogawa,
C. Cao, T. Minematsu, M. Yoshikawa and O. Muraoka, Bioorg.
Med. Chem. Lett., 2009, 19, 2195.
Fig. 2 Superposition of 7k in the hNtMGAM active site. Dotted lines
show hydrogen bonding (gray) and salt bridge (red). Double-headed
green arrows show the van der Waals interactions of the phenyl ring with
the amino acid residues (distances of a: 3.42 A, b: 4.04 A, c: 4.28 A).
Van der Waals interactions between an oxygen atom of the nitro
group of 7k and the amino acid residues are shown with the pink
arrows (e: 3.69 A, f: 4.05 A). Distance between the p-planes shown
with the purple dotted line was calculated to be 4.42 A.
5 (a) S. Ozaki, H. Oe and S. Kitamura, J. Nat. Prod., 2008, 71,
981; (b) O. Muraoka, W. Xie, G. Tanabe, F. A. M. Amer,
T. Minematsu and M. Yoshikawa, Tetrahedron Lett., 2008,
49, 7315.
which have been proven to affect the hydrolysis of maltose,15
would be another reason for these discrepancies.
6 W. Xie, G. Tanabe, J. Akaki, T. Morikawa, K. Ninomiya,
T. Minematsu, M. Yoshikawa, X. Wu and O. Muraoka, Bioorg.
Med. Chem., 2011, 19, 2015.
By the hitherto conducted docking studies on salacinol (1)
with hNtMGAM, it has been revealed that efficient hydrogen
bonding interactions of C20-OH and C40-OH with Asp203
caused the strong inhibitory activity, which was essential for
these inhibitors to exert their activity.9b It is interesting to note
that in the present in silico docking study of 7k with the
enzyme, another binding mode different from that of 1 has
been revealed as shown in Fig. 2, where instead of the
hydrogen bonding interactions of C40-OH of 1 with Asp203,
van der Waals interactions were detected between the phenyl
ring at C30 and the hydrophobic parts of Asp203, Phe450 and
Lys480 around the phenyl group (Fig. 2, green arrows a, b,
and c). In addition, the phenyl ring of Phe450 and the
C30-phenyl were stacked almost in parallel, and the distance
between these two p-planes was calculated to be ca. 4.4 A, a
sufficient distance for effective p–p interaction16 (purple dotted
line d). The nitro group introduced at the ortho-position on the
benzene ring anchored in a concave pocket of the enzyme, and
one of the oxygens of the nitro group affected the van der
Waals interactions with the surrounding methylene chains of
Asp203 and Lys480 (Fig. 2, pink arrows e and f), the distances
of which were 3.7 and 4.1 A, respectively. This anchoring effect
on the binding could have been more effective than expected in
the docking studies, and might be contributing to the inhibi-
tory activity.
7 M. H. S. Jayawardena, N. M. W. de Alwis, V. Hettigoda and
D. J. S. Fernando, J. Ethnopharmacol., 2005, 97, 215.
8 (a) R. Eskandari, K. Jones, D. R. Rose and B. M. Pinto, Chem.
Commun., 2011, 47, 9134; (b) S. Mohan and B. M. Pinto, Nat.
Prod. Rep., 2010, 27, 481; (c) R. Eskandari, K. Jones, D. R. Rose
and B. M. Pinto, Bioorg. Med. Chem. Lett., 2010, 20, 5686;
(d) S. Mohan and B. M. Pinto, Carbohydr. Res., 2007, 342, 1551
and references cited therein.
9 (a) L. Sim, K. Jayakanthan, S. Mohan, R. Nasi, B. D. Johnston,
B. M. Pinto and D. R. Rose, Biochemistry, 2010, 49, 443;
(b) S. Nakamura, K. Takahira, G. Tanabe, T. Morikawa,
M. Sakano, K. Ninomiya, M. Yoshikawa, O. Muraoka and
I. Nakanishi, Bioorg. Med. Chem. Lett., 2010, 20, 4420;
(c) G. Tanabe, T. Otani, W. Cong, T. Minematsu, K. Ninomiya,
M. Yoshikawa and O. Muraoka, Bioorg. Med. Chem. Lett., 2011,
21, 3159.
10 P. Labute, J. Comput. Chem., 2008, 29, 1693.
11 A. Ghavami, K. S. Sadalapure, B. D. Johnston, M. Lobera,
B. B. Snider and B. M. Pinto, Synlett, 2003, 1259.
12 E. Abushanab, P. Vemishetti, R. W. Leiby, H. K. Singh,
A. B. Mikkilineni, D. C.-J. Wu, R. Saibaba and R. P. Panzica,
J. Org. Chem., 1988, 53, 2598.
13 J. Wu, S. Zhang, Q. Meng, H. Cao, Z. Li, X. Li, S. Shi, D. H. Kim,
L. Bi, N. J. Turro and J. Ju, Proc. Natl. Acad. Sci. U. S. A., 2007,
104, 16462.
14 Genome net (http://www.genome.jp/dbget-bin/www_bget?ko:K12047).
15 (a) L. Sim, C. Willemsma, S. Mohan, Y. H. Naim, B. M. Pinto and
D. R. Rose, J. Biol. Chem., 2010, 285, 17763; (b) K. Jones, L. Sim,
S. Mohan, J. Kumarasamy, H. Liu, S. Avery, H. Y. Naim,
R. Quezada-Calvillo, B. L. Nichols, B. M. Pinto and
D. R. Rose, Bioorg. Med. Chem., 2011, 19, 3929; (c) L. Ren,
X. Cao, P. Geng, F. Bai and G. Bai, Carbohydr. Res., 2011,
346, 2688.
16 (a) G. B. McGaughey, M. Gagne and A. K. Rappe, J. Biol. Chem.,
´ ´
1998, 273, 15458; (b) M. O. Sinnokrot and C. D. Sherril, J. Phys.
Chem. A, 2006, 110, 10656.
17 O. Muraoka, T. Morikawa, S. Miyake, J. Akaki, K. Ninomiya,
Y. Pongpiriyadacha and M. Yoshikawa, J. Nat. Med., 2011,
65, 142.
Thus, with the aid of the in silico method, more potent
a-glucosidase inhibitors (7b–7m) than the seed compound
salacinol (1) were effectively designed and developed. Finally,
we comment that the sulfonium salt 7k was found to be the
most potent among the cyclic sulfonium family, a new class of
a-glucosidase inhibitors, synthesized so far. On the basis of the
present findings further studies to develop the more active and
safe candidate for the new type of a-glucosidases are in progress.
c
8648 Chem. Commun., 2012, 48, 8646–8648
This journal is The Royal Society of Chemistry 2012