全 文 ： 676 Chin J Nat Med Nov. 2013 Vol. 11 No. 6 2013 年 11 月 第 11 卷 第 6 期

Chinese Journal of Natural Medicines 2013, 11(6): 0676−0683
doi: 10.3724/SP.J.1009.2013.00676
Chinese
Journal of
Natural
Medicines







Total synthesis of neokotalanol, a potent α-glucosidase
inhibitor isolated from Salacia reticulata
XIE Wei-Jia 1, TANABE Genzoh 2, TSUTSUI Nozomi 2, WU Xiao-Ming 1, MURAOKA Osamu 2*
1 Department of Medicinal Chemistry, China Pharmaceutical University, Nanjing 210009, China;
2
School of Pharmacy, Kinki University, 3-4-1 Kowakae, Higashi-osaka, Osaka 577-8502, Japan
Available online 20 Nov. 2013
[ABSTRACT] Neokotalanol, a potent α-glucosidase inhibitor isolated from Salacia reticulata, was synthesized through a key cou-
pling reaction between a perbenzylated thiosugar and an appropriately protected perseitol triflate derived from D-mannose. This key
step was found to be quite temperature dependent, and a simultaneous cyclization of the triflate leading to a characteristic
2,4,7-trioxabicyclo[4.2.1]nonane system was detected.
[KEY WORDS] Neokotalanol; Total synthesis; Salacia reticulata; α-glucosidase inhibitor
[CLC Number] R284.1 [Document code] A [Article ID] 1672-3651(2013)06-0676-08

1 Introduction
Glucosidases play a fundamental role in catalyzing the
hydrolysis of complex carbohydrates, which is closely related
to various physiological and biological processes in living
systems. The inhibition of glucosidases has long been con-
sidered a promising approach to treat a variety of diseases,
such as diabetes, obesity, glycosphingolipid lysosomal stor-
age disease, HIV infections, cancer, and Gaucher’s disease
[1-10]. Thus, identifying and designing small molecules with
potent glucosidase inhibitory activity has been of great inter-
est for medicinal chemists for a long time.
In the late 1990’s, salacinol (1) was isolated from Salacia
reticulata, a large woody climbing plant widespread in Sri
Lanka and South India. The aqueous extracts of this plant have

[Received on] 10-May-2013
[Research Funding] This project was supported by the High-Tech
Research Center Project for Private Universities: matching fund
subsidy from MEXT (Ministry of Education, Culture, Sports. Sci-
ence and Technology), 2007–2011, Fundamental Research Funds for
the Central Universities (JKZ2011003), the National Natural Science
Foundation of China (No. 81202409), Natural Science Foundation of
Jiangsu Province (SBK201240392), Scientific Research Foundation
for the Returned Overseas Chinese Scholars, Ministry of Education
(2013) and Technology Foundation for Selected Overseas Chinese
Scholar, Ministry of Personnel of China (2013).
[*Corresponding author] MURAOKA Osamu: Prof., E-mail: mu-
raoka@phar.kindai.ac.jp
These authors have no conflict of interest to declare.
traditionally been used for the treatment of the type II diabetes
in the Ayurvedic system of Indian traditional medicine [11-12].
The structure of salacinol is quite unique, bearing a permanent
positive charge as the thiosugar sulfonium sulfate inner salt
comprised of 1-deoxy-4-thio-D-arabinofranosyl cation and a
3’-sulfate anion, as shown in Fig. 1. The α-glucosidase in-
hibitory activity of 1 was tested in vitro, and it was revealed
to be as potent as voglibose and acarbose, which are widely
used in the clinic as antidiabetic agents these days [11-12].
Since the discovery of salacinol (1), related sulfonium sul-
fates, kotalanol [13] (2), ponkoranol [14] (3) salaprinol [14] (4),
as well as their de-O-sulfonated sulfonium analogs, neosa-



Fig. 1 Sulfonium salts isolated from Salacia species as a
new class of α-glucosidase inhibitors
XIE Wei-Jia, et al. /Chinese Journal of Natural Medicines 2013, 11(6):676−683
2013 年 11 月 第 11 卷 第 6 期 Chin J Nat Med Nov. 2013 Vol. 11 No. 6 677



Fig. 2 Attempted synthesis of kotalanol (2) and the strategy to synthesize neokotalanol (6)

lacinol [15-16] (5), neokotalanol [17-19] (6), neoponkoranol [20-21]
(7), and neosalaprinol [20-21] (8) were subsequently isolated
from the same genus (Fig. 1). Other than 4 and 8, these sul-
fonium salts showed potent α-glucosidase inhibitory activi-
ties, composing a new class of naturally occurring
α-glycosidase inhibitors, and much attention has been fo-
cused on the synthesis and structure-activity relationships
(SAR) of these sulfonium salts [22].
In this series of natural inhibitors, kotalanol (2) and
neokotalanol (6) were the most attractive targets, because the
exact stereostructure of their long side chains had not been
clarified until recently, when it was elucidated through the
total synthesis by Jayakanthan and co-workers [23], and our
degradation study of natural kotalanol (2) [24]. Interestingly, in
the course of a synthetic study on kotalanol, Jayakanthan and
co-workers designed a well-protected cyclic sulfate of per-
perseitol (10) as the key precursor of the coupling reaction [23].
The subsequent deprotection of the resulted sulfonium salt
(11) unexpectedly caused desulfonation, and led to neokota-
lanol (6, Fig. 2).
At about the same time, neokotalanol (6) was revealed to
be the most potent inhibitor among these naturally occurring
sulfonium salts [17-19]. In addition, substitution of the hydro-
philic sulfate moiety on the 3’-position of salacinol (1) to
hydrophobic groups was found to be an efficient way to
substantially increase the α-glycosidase inhibitory activity
[25-26]. For further SAR studies on neokotalanol (6), it is nec-
essary to develop a coupling strategy which could directly
format the de-O-sulfonated sulfonium salt structure (A, Fig.
2). In 2011, we succeeded in the total synthesis of neopon-
koranol (7) through the coupling reaction between the per-
benzylated thiosugar and a triflate of protected glucose [20].


Scheme 1 Reagents and conditions: (a) (Ph3P)3RhCl, iPr2NEt, EtOH, reflux, then HgO, HgCl2, acetone/H2O, 9/1, r.t.; (b)
Ph3PCH3Br, nBuLi THF, 0 °C, then 45 °C; (c) CH2Br2, tBu4N+Br–, 50% aqueous NaOH, 60 °C; (d) AD-mix-, tBuOH, H2O, 0 °C;
(e) BnBr, NaH, DMF, 0 °C–r.t.; (f) BCl3, CH2Cl2, –78–0 °C; (g) TBAF, THF–H2O, 0 °C–r.t.; (h) Tf2O, 2,6-lutidine, –200 °C
XIE Wei-Jia, et al. /Chinese Journal of Natural Medicines 2013, 11(6): 676−683
678 Chin J Nat Med Nov. 2013 Vol. 11 No. 6 2013 年 11 月 第 11 卷 第 6 期

As a continuing synthetic study on this series of compounds,
this paper describes the direct synthesis of neokotalanol (6)
through the coupling reaction of the protected thiosugar and a
perseitol triflate derived from D-mannose.
2 Results and Discussion
In the course of synthetic studies on kotalanol (2) and
neokotalanol (6), which contain the 7-carbon side chain, the
authors have encountered numerous obstacles. The biggest
problem addressed the stability of the polyprotected perseitol
as a precursor of the coupling reaction. Due to the long side
chain structure, most of the designed precursors easily un-
derwent decomposition or intramolecular cyclization. As the
1, 3-dioxepane structure of 10 designed by Pinto and
co-workers showed good stability under the coupling reaction
conditions, this structure was used as the basic scaffold in the
present work.
3 Preparation of Triflate (20)
In light of the elucidated side chain stereostructure of
neokotalanol (6), D-mannose was selected as the starting
material, and was converted to known allylmannnoside, al-
lyl -3, 4-dibenzyl-6-O-tert-butyldiphenylsilyl-α-and-β-D-ma-
nnopyranoside [27-28] (12) in good yield. The allyl group was
removed by isomerization of the terminal olefin moiety fol-
lowed by hydrolysis of the resulting vinyl ether with HgO
and HgCl2 to give an anomeric mixture of 3, 4-dibenzyl-6-O-
tert-butyldiphenylsilyl-D-mannopyranose (13) in 90% yield.
The mixture was then subjected to the Wittig reaction with
methyltriphenylphosphonium bromide (Ph3P+CH3Br–) to give
a terminal olefin, 1-O-tert-butyldiphenylsilyl-3, 4-di-O- ben-
zyl-D-manno-hept-6-enitol (14), in 74% yield. The olefin (14)
was treated with dibromomethane (CH2Br2) in the presence
of aqueous sodium hydroxide and tetra-n-butylammonium
bromide (tBu4N+Br–) as a phase transfer catalyst to give a 1,
3-dioxepane derivative, 1-O-tert-butyldiphenylsilyl-3, 4-di-
O-benzyl-2, 5-O-methylene-D-manno-hept-6-enitol (15), in
54% yield. The acetal (15) was then subjected to the Shar-
pless asymmetric dihydroxylation using AD-mix-β to give a
difficult to separate mixture of 3, 6-O-methylene-4, 5-di-O- ben-
zyl-7-O-tert-butyldiphenylsilyl-D-glycero-D-galacto-heptitol
(16), and its 2’-epimer (2-epi-16) in 74% yield (16/2-epi-16,
ca. 8/1). After treatment of the epimeric mixture with benzyl
bromide in the presence of sodium hydroxide, the major epi-
mer,1,2,4,5-tetra-O-benzyl-3,6-O-methylene-7-O-tert-butyl-
diphenylsilyl-D-glycero-D-galacto-heptitol (17) was isolated
in pure form with a yield of 53% in two steps. The stereo-
chemistry of 17 was confirmed after leading to the known
compound, D-glycero-D-galacto-heptitol [29-31] (18). Selective
deprotection of the silyl group in compound 17 was accom-
plished by using TBAF in THF to give 1,2,4,5-
etra-O-benzyl-3, 6-O-methylene-D-glycero-D-galacto-heptitol
(19) in 86% yield. Triflation of the primary hydroxyl of 19
with trifluoromethanesulfonic anhydride (Tf2O) in the pres-
ence of 2, 6-lutidine gave 1, 2, 4, 5-tetra-O-benzyl-3, 6-O-
methylene-7-O-trifluoromethanesulfonyl-D-glycero-D-galacto-
heptitol (20) in 83% yield. Triflate 20 was found to be unsta-
ble even at r.t., and was subjected to the next coupling reac-
tion immediately (Scheme 1).
Coupling reaction between triflate (20) and thiosugar (9b)
The mode of the coupling reaction between triflate 20
and thiosugar (9b) was found to be quite temperature depend-
ent. At room temperature, no formation of the desired sulfo-
nium salt, 2, 3, 5-tri-O-benzyl-1, 4-dideoxy-[(R)-7-deoxy-1, 2,
4, 5-tetra-O-benyl-3, 6-O-methylene-D-glycero-D-galacto-
heptitol-1yl]-D-arabinitol trifluoromethanesulfonate (21), was
detected at all, while the starting material 20 was totally
converted to a bicyclic compound, 4,7-anhydro-1,2, 5-tri-
O-benzyl-3, 6-O-methylene-D-glycero-D-galacto-heptitol (22),
accompanied by a formation of the S-benzylated thiosugar, 2,
3, 5-tri-O-benzyl-1, 4-dideoxy-1, 4-[(R)-benzyl episulfoni-
umylidene]-D-arabinitoltrifluoromethanesulfonate (23), in
less than one hour. At lower temperature (0 °C), the coupling
reaction progressed slowly, and it took one week for the
starting material (20) to be totally consumed. The desired
sulfonium salt (21, 24% from 20) was obtained together with
bicyclic compound (22, 70% from 20) and sulfonium salt (23,
70% from 20) (Scheme 2). When the reaction temperature
was set below –20 °C, the reaction did not proceed, and the
reactants were totally recovered. A plausible mechanism of
formation of the bicyclic compound 22 is presented in Fig. 3.
In this reaction, two different modes (routes a and b, in Fig-
ure 3) of nucleophilic attack were proposed to take place. At
0 °C (route a), direct attack of the sulfur atom of the thi-
osugar 9b to the methylene carbon at C-7 bearing the leaving
group (TfO–) took place through the thermodynamically
stable conformer 20-A (–390 kcal·mol −1) [32] providing the
desired sulfonium salt 21, while at room temperature, a ki-
netically advantageous intramolecular cyclization of 20,
which was triggered by the attack of the sulfur atom of 9b on
the methylene carbon of the benzyloxy moiety at C-4 of
compound 20 (route b), mainly proceeded to provide bicyclic
compound 22. The formation of product 23 well supported
the reaction mechanism. Thus, it is of interest to note that the
chemoselectivity in the reaction between 20 and 9b entirely
varied depending on the temperature.
The FAB mass spectrum of 22, run in a positive ion
mode, showed a peak at m/z 499 corresponding to the sodium
adduct ion of the molecule [M + Na]+, which lost one benzyl
moiety from reactant 20. A NOE correlation between H-7b
and the endo-proton on the acetal moiety, as well as HMBC
correlations between positions 4 and 7, supported the struc-
ture 22. On the other hand, the FAB mass spectrum of the
sulfonium salt 23 in a positive ion mode showed a peak at
m/z 511, which indicated the introduction of one more benzyl
moiety to reactant 9b. The negative-ion FAB mass spectrum
of 23 showed a peak at m/z 149, corresponding to the triflate
anion moiety.
XIE Wei-Jia, et al. /Chinese Journal of Natural Medicines 2013, 11(6):676−683
2013 年 11 月 第 11 卷 第 6 期 Chin J Nat Med Nov. 2013 Vol. 11 No. 6 679



Scheme 2 Reagents and conditions: thiosugar (9b), THF, –5–0 °C, 1 week


Fig. 3 Plausible mechanism of the coupling reaction between triflate 20 and thiosugar 9b



Scheme 3 Reagents and conditions: (a) BCl3, CH2Cl2, –78 °C; (b) IRA 400 J (Cl– form), MeOH, r.t.

4 Deprotection of Sulfonium Salt (21)
The resulted sulfonium salt (21) was then treated with
BCl3 in CH2Cl2 to remove all the protecting groups, and sub-
sequent treatment of the product with ion exchange resin IRA
400 J (Cl– form) to change any anions to chlorine gave the
target compound (6) in 60% yield in two steps (Scheme 3).
The 1H and 13C NMR spectra of 6 were in good accord with
those of an authentic specimen isolated from Salacia reticu-
lata [17-19].
5 Conclusions
Neokotalanol (6), the de-O-sulfonated sulfonium salt of
kotalanol (2) bearing a 7-carbon-polyhydroxylated side chain,
was synthesized by the direct coupling reaction between the
perbenzylated thiosugar (9b) and a perseitol triflate (20) as
the key reaction. It is interesting to note that the coupling
reaction between these two substrates proceeded via different
two conformations of 20, depending on the reaction tem-
perature. A characteristic intramolecular cyclization via an
unstable conformation of 20 leading to the trioxabicyclic
compound (22) competed with the desired direct coupling
between 20 and 9b, even at 0 °C. The above coupling strat-
egy to construct the 3’-de-O-sulfonated sulfonium salt is of
great importance for SAR studies on this type of
α-glucosidase inhibitor. Further optimization of the thiosugar
- triflate coupling reaction, and the structure modification of
neokotalanol using this strategy as the key reaction, are now
in progress.
6 Experimental
Mps were determined on a Yanagimoto MP-3S micro-
melting point apparatus, and mps and bps are uncorrected.
IR spectra were measured on a Shimadzu FTIR-8600PC
spectrophotometer. NMR spectra were recorded on a JEOL
JNM-ECA 500 (500 MHz 1H, 125 MHz 13C) or a JEOL
JNM-ECA 600 (600 MHz 1H, 150 MHz 13C) or a JEOL
JNM-ECA 700 (700 MHz 1H, 175 MHz 13C) spectrometer.
Chemical shifts (δ) and coupling constants (J) are given in
XIE Wei-Jia, et al. /Chinese Journal of Natural Medicines 2013, 11(6): 676−683
680 Chin J Nat Med Nov. 2013 Vol. 11 No. 6 2013 年 11 月 第 11 卷 第 6 期

ppm and Hz, respectively. Low-resolution and high- resolu-
tion mass spectra were recorded on a JEOL JMS-HX 100
spectrometer. Optical rotations were determined with a
JASCODIP-370 digital polarimeter. Column chromatography
was effected over Fuji Silysia silica gel BW-200. All the or-
ganic extracts were dried over anhydrous sodium sulfate prior
to evaporation.
3, 4-Dibenzyl-6-O-tert-butyldiphenylsilyl-α-and -β-D-
mannopyranose (13)
To a solution of 12 [27-28] (7.0 g, 11.0 mmol) in EtOH
(150 mL) was added (Ph3P)3RhCl (0.55 g) and iPr2NEt (17
mL), and the mixture was heated under reflux for 1.5 h. After
removal of the solvent under reduced pressure, the residue
was diluted with CH2Cl2 (200 mL), and the resulting mixture
was washed with brine, and concentrated to give a brown oil
(6.7 g), which was then treated with HgO (4.5 g) and HgCl2
(4.5 g) in a mixture of acetone/water (9/1, 400 mL) at r.t. for
12 h. The reaction mixture was evaporated under reduced
pressure, and the residue was diluted with water (200 mL).
The resulting mixture was extracted with ether. The extract
was washed with brine and condensed under reduced pres-
sure to give a pale brown oil (6.8 g), which on column chro-
matography (n-hexane/AcOEt, 15/1) gave the title compound
(13, 5.9 g, 90%) as a pale yellow oil.
1-tert-Butyldiphenylsilyl-3, 4-di-O-benzyl-D-manno-
hept-6-enitol (14)
To a solution of methyltriphenylphosphonium bromide
(Ph3PCH3Br, 17.6 g, 49.3 mmol) in THF (160 mL) was added
1.6 mol·L−1 nBuLi (44 mL, 70 mmol) at 0 °C. The mixture
was stirred at r.t. for 2.5 h. To the resulting mixture was
added dropwise a solution of 13 (4.9 g, 8.2 mmol) in THF
(100 mL) at 0 °C, and the mixture was stirred at 45 °C for 30
min. After the reaction was quenched by the addition of ace-
tone, the resulting mixture was diluted with water, and con-
centrated under reduced pressure. The aqueous residue was
extracted with ether, washed with brine, and condensed to
give a crude (5.6 g) as a yellow oil, which, on column chro-
matography (n-hexane/AcOEt, 15/1), gave the title com-
pound (14, 3.6 g, 74% yield) as a pale yellow oil. [α]25D
+8.4 (c 1.48, CHCl3). IR (neat): 3 435, 2 928, 2 359, 1 471, 1
456, 1 429, 1 213, 1 096, 1 074 cm–1. 1H NMR (700 MHz,
CDCl3) δ: 1.06 [9H, s, C(CH3)3], 3.70 (1H, dd J = 5.2, 3.0,
H-4), 3.77 (1H, dd J = 10.2, 5.0, H-1a), 3.81 (1H, dd J = 10.2,
5.5, H-1b), 3.93 (1H, dd J = 6.5, 3.0, H-3), 4.01 (1H, ddd-like,
J = 6.5, 5.5, 5.0, H-2), 4.48 (1H, dddd, J = 5.2, 5.2, 1.6, 1.6,
H-5), 4.52/4.67 (each 1H, d, J = 11.3, PhCH2), 4.56 (2H, s,
PhCH2), 5.23 (1H, ddd, J = 10.8, 1.6, 1.6, H-7a), 5.41 (1H,
dd J =17.0, 1.6, 1.6, H-7b), 5.91(1H, ddd, J = 17.0, 10.8, 5.2,
H-6), 7.18–7.65 (20H, m, arom.). 13C NMR (175 MHz,
CDCl3) δ: 19.2 [C(CH3)3], 26.9 [C(CH3)3], 64.8 (C-1), 71.1
(C-2), 71.8 (C-5), 72.6/73.0 (PhCH2), 78.0 (C-3), 80.1 (C-4),
116.2 (C-7), 137.8 (C-6), 127.8/127.9/128.3/128.35/128.41/
128.43/129.9/135.57/ 135.62 (d, arom.), 132.98/133.04/
137.45/137.51 (s, arom.). FABMS m/z 619, [M + Na]+ (pos.).
1-tert-Butyldiphenylsilyl-2, 5-O-methylene-3, 4-di-O-
benzyl-D-manno-hept-6-enitol (15)
To a solution of 14 (3.6 g, 6.0 mmol) and nBu4N+Br-
(250 mg) in CH2Br2 (60 mL) was added 50% aqueous NaOH
solution (130 g), and the mixture was stirred at 60 °C for 45
min. After addition of CH2Cl2 (150 mL), the resulting mixture
was diluted with water (150 mL), and washed with brine. The
organic layer was condensed under reduced pressure to give a
yellow oil (4.2 g), which, on column chromatography, gave
the title compound (15, 2.0 g, 54%) as a pale yellow oil. [α]
25
D −29.7 (c 0.7, CHCl3). IR (neat): 3 069, 2 930, 2 888, 2
357, 1 472, 1 456, 1 427, 1 213, 1 096, 1 073 cm–1. 1H NMR
(700 MHz, CDCl3) δ: 1.05 [9H, s, C(CH3)3], 3.42 (1H, dd, =
9.5, 7.5, H-4), 3.71 (1H, ddd, J = 9.8, 4.5, 2.0, H-2), 3.73 (1H,
dd, J = 9.8, 7.5, H-3), 3.90 (1H, dd, J = 10.0, 4.5, H-1a), 3.95
(1H, dd, J = 10.0, 2.0 H-1b), 4.19 (1H, dddd-like, J = 9.5, 6.0,
1.5, 1.5, H-5), 4.67/4.73 (each 1H, d, J = 10.5, PhCH2),
4.69/4.82 (each 1H, d, J = 10.9, PhCH2), 4.80/4.87 (each 1H,
d, J = 4.4, OCH2O), 5.28 (1H, ddd, J = 10.5, 1.5, 1.5, H-7a),
5.44 (1H, ddd, J = 17.0, 1.5, 1.5, H-7b), 6.07 (1H, ddd, J =
17.0, 10.5, 6.0, H-6), 7.23–7.73 (20H, m, arom.). 13C NMR
(175 MHz, CDCl3) δ: 19.3 [C(CH3)3], 26.7 [C(CH3)3],
63.8(C-1), 74.9/75.1 (PhCH2), 75.0 (C-5), 75.6 (C-2), 81.8
(C-3), 86.7 (C-4), 92.7 (OCH2O), 116.9 (C-7), 135.8 (C-6),
127.5/127.6/127.69/127.73/127.9/128.3/128.4/129.6/135.6/
135.9 (d, arom), 133.3/133.7/ 138.2/138.3 (s, arom.). FABMS
m/z 631, [M + Na]+ (pos.).
7 Dihydroxylation of Terminal Olefin (15)
To a solution of 15 (1.4 g, 2.3 mmol) in a mixture of
tert-butyl alcohol and water (1/1, 10 mL) was added
AD-mix-β (3.5 g). The reaction mixture was stirred vigor-
ously at 0 °C for 7 days. The reaction was quenched by the
addition of solid sodium sulfite (4 g), and the resulting mix-
ture was extracted with AcOEt (3 × 100 mL). The extract was
washed successively with water (50 mL) and brine (50 mL),
and condensed under reduced pressure to give a pale yellow
oil (1.56 g), which, on column chromatography,
(n-hexane/AcOEt, 10/1) gave a ca. 8/1 mixture of (16/2-epi-
16, 1.1 g, 74%) as a pale yellow oil.
Benzylation of diols 16/2-epi-16
To a mixture of sodium hydride (NaH, 375 mg, 9.4
mmol, 60% in liquid paraffin) and benzyl bromide (BnBr,
1.36 mL, 11.2 mmol), and dry DMF (8 mL) was added a
solution of 16/2-epi-16 (1.0 g) in dry DMF (5 mL) at 0 °C.
After stirring at rt for 2 h, the reaction mixture was poured
into ice-water, and extracted with ether. The extract was
washed with brine, and condensed under reduced pressure to
give a yellow oil (1.4 g), which, on column chromatography
(n-hexane/AcOEt, 20/1), gave 7-O-tert-butyldiphenylsilyl-1,
2,4,5-tetra-O-benzyl-3,5-O-methylene-D-glycero-D-galacto-
heptitol (17, 0.91 g, 53% from 15) and a ca. 1/1 mixture of
(17/2-epi-17, 0.26 g, 15% from 15).
XIE Wei-Jia, et al. /Chinese Journal of Natural Medicines 2013, 11(6):676−683
2013 年 11 月 第 11 卷 第 6 期 Chin J Nat Med Nov. 2013 Vol. 11 No. 6 681

Compound 17: Colorless oil. [α]25D −3.7 (c 0.82, CHCl3).
IR (neat): 1 497, 1 454, 1 427, 1 397, 1 096, 1 090, 1 076, 1
043 cm-1. 1H NMR (700 MHz, CDCl3) δ: 1.05 [9H, s,
C(CH3)3],3.70 (1H, dd, J = 9.2, 6.8, H-1a), 3.74 (1H, ddd, J =
9.8, 4.3, 2.0, H-6), 3.76 (1H, dd, J = 9.8, 6.7, H-5), 3.80(1H,
dd, J = 9.2, 5.2, H-1b), 3.91 (1H, dd, J = 10.8, 4.3, H-7a),
3.90-3.93 (2H, m, H-3 and H-4), 3.94 (1H, dd, J = 10.8, 2.0,
H-7b), 4.20 (1H, ddd, J = 6.8, 5.2, 1.0, H-2), 4.50/4.78 (each
1H, d, J = 11.6, PhCH2), 4.51/4.56 (each 1H, d, J = 11.8,
PhCH2), 4.59/4.88 (each 1H, d, J = 11.4, PhCH2),4.64/4.80
(each 1H, d, J = 4.2, OCH2O), 4.70/4.73 (each 1H, d, J =
10.8, PhCH2), 7.05–7.73 (20H, m, arom.). 13C NMR (175
MHz, CDCl3) δ: 19.3 [C(CH3)3], 26.9 [C(CH3)3], 63.7(C-7),
69.2 (C-1), 72.3/73.5/74.2/74.5 (PhCH2), 73.3 (C-3), 75.4
(C-6), 76.1 (C-2), 81.5 (C-4), 82.8 (C-5), 93.4 (OCH2O),
127.0/127.36/127.45/127.52/127.55/127.62/127.66/127.70/12
7.75/127.81/128.27/128.36/128.44/129.5/129.6/135.6/135.9
(d, arom.), 133.3/133.7/138.0/138.2/138.4/138.8 (s, arom.).
FABMS m/z 845, [M + Na]+ (pos.).
1, 2, 4, 5-Tetra-O-benzyl-3, 6-O-methylene-D-glyc-
ero-D-galacto-heptitol (19)
To a solution of 17 (0.9 g, 1.1 mmol) in THF (10 mL)
was added dropwise 1.0 mol·L−1 TBAF in THF (10 mL) at 0
°C. After being stirred at r.t. for 12 h, the reaction mixture
was condensed under reduced pressure. The residue was di-
luted with water, and extracted with AcOEt. The extract was
washed with brine and condensed under reduced pressure to
give a pale yellow oil (0.85 g), which, on column chroma-
tography, gave the title compound (19, 0.55 g, 86%) as a pale
yellow oil. [α]25D −17.5 (c 0.4, CHCl3). IR (neat): 3 030, 2
959, 1 497, 1 454, 1 395, 1 362, 1 261, 1 121, 1 092, 1 082, 1
069, 1 045 cm–1. 1H NMR (700 MHz, CDCl3) δ: 3.66 (1H, dd,
J = 8.4, 7.9, H-5), 3.68 (1H, dd, J = 9.1, 7.0, H-1a),
3.73–3.76 (2H, m, H-6 and H-7a), 3.77 (1H, dd, J = 9.1, 5.2,
H-1b), 3.88 (1H, dd, J = 9.8, 1.6, H-3), 3.88-3.91 (1H, m,
H-7b), 3.93 (1H, dd, J = 9.8, 7.9, H-4), 4.16 (1H, ddd, J = 7.0,
5.2, 1.6, H-2), 4.48/4.55 (each 1H, d, J = 11.8, PhCH2),
4.49/4.76 (each 1H, d, J = 11.6, PhCH2), 4.59/4.89 (each 1H,
d, J = 11.4, PhCH2), 4.62/4.82 (each 1H, d, J = 4.3, OCH2O),
4.69/4.78 (each 1H, d, J = 10.8, PhCH2), 7.22–7.73 (20H, m,
arom.). 13C NMR (175 MHz, CDCl3): δ: 62.8 (C-7), 68.8
(C-1), 72.4/73.5/ 74.1/74.8 (PhCH2), 73.7 (C-3), 74.3 (C-6),
75.9 (C-2), 81.1 (C-4), 83.3 (C-5), 93.4 (OCH2O),
127.5/127.6/127.7/127.8/127.9/128.0/128.1/128.2/128.3/128.
5 (d, arom.), 137.9/138.0/138.3/138.7 (s, arom.). FABMS m/z
607, [M + Na]+ (pos.).
Coupling reaction between triflate (20) and thiosugar
(9b)
Under an argon atmosphere, Tf2O (0.22 mL, 1.3 mmol)
was added to a solution of 2, 6-lutidine (0.15 mL, 1.3 mmol)
in CH2Cl2 (10 mL) at –20 °C, and the mixture was stirred at
that temperature for 5 min. To the resulting mixture was
added dropwise a solution of 19 (0.5 g, 0.86 mmol) in
CH2Cl2 (10 mL) at –20 °C. After being stirred at 0°C for 30
min, the reaction mixture was poured into ice-cooled water
and extracted with CH2Cl2. The extract was condensed under
reduced pressure to give a pale yellow oil (0.61 g), which, on
column chromatography (n-hexane/AcOEt, 20/1), gave the
triflate (20, 0.51 g, 83%) as a pale yellow oil. Owing to the
instability of 20, even at rt, it was subjected to the coupling
reaction with thiosugar (9b) immediately. Thus, under an
argon atmosphere, to a solution of triflate (20, 0.45 g, 0.62
mmol) in THF (3 mL) was added a solution of thiosugar (9b,
0.78 g, 1.86 mmol) in THF (2 mL) at –15 °C, and the mixture
was stirred at 0 °C for 7 days. The reaction mixture was con-
densed under reduced pressure to give a pale yellow oil (1.20
g), which, on column chromatography (CHCl3/MeOH, 200/1)
gave 2,3,5-tri-O-benzyl-1,4-dideoxy-[(R)-7-deoxy- 1,2,4,5-
tetra-O-benzyl-3,6-O-methylene-D-glycero-D-galacto-heptito
l-1yl]-D-arabinitol trifluoromethanesulfonate (21, 0.17 g,
24% from 20), together with the bicyclic compound, 4,
7-anhydro-1, 2, 5-tri-O-benzyl-3,6-O-methylene-D-glycero-
D-galacto-heptitol (22, 210 mg, 70% from 20) and 1,
4-dideoxy-1, 4-[(R)-benzylepisulfoniumylidene]-D-arabini-
toltrifluoromethanesulfonate (23, 290 mg,70% from 20) as
pale yellow oils.
Compound 21: [α]25D −1.8 (c 0.65, CHCl3). IR (neat): 1
497, 1 454, 1 360, 1 279, 1 263, 1 147, 1 117, 1 088, 1 030
cm–1. 1H NMR (700 MHz, CDCl3) δ: 3.57–3.63 (2H, m,
H-5a/H-5b), 3.58 (1H, dd, J = 9.5, 8.0, H-3’), 3,61 (1H, dd, J
= 9.0, 5.5, H-7’a), 3.64 (1H, dd, J = 12.5, 9.5, H-1’a), 3.69
(1H, dd-like, J = 12.5, 3.6, H-1a), 3.70 (1H, dd, J = 12.5, 2.5,
H-1’b), 3.75 (1H, dd, J = 9.0, 5.5, H-7’b), 3.83 (1H, dd. J =
9.6, 1.5, H-5’), 3.91 (1H, dd, J = 9.6, 8.0, H-4’), 3.98 (1H,
ddd, J = 9.5, 9.5, 2.5, H-2’), 4.09 (1H, ddd, J = 5.5, 5.5, 1.5,
H-6’), 4.11 (1H, br d-like, J = 12.5, H-1b), 4.12 (1H, br m,
H-3), 4.44–4.83 (14H, m, PhCH2), 4.47/4.66 (each 1H, d, J =
4.4, OCH2O), 4.56 (1H, br m, H-2), 7.13–7.82 (35H, m,
arom.). 13C NMR (175 MHz, CDCl3) δ: 48.4 (C-1), 49.4
(C-1’), 66.4 (C-5), 66.8 (C-4), 68.5 (C-7’), 70.1 (C-2’),
72.17/72.22/72.6/73.5/73.7/73.9/75.1(PhCH2), 74.6 (C-5’),
75.7 (C-6’), 81.1 (C-4’), 82.2 (C-2), 82.7 (C-3), 84.2 (C-3’),
93.3 (OCH2O), 126.9/127.45/127.53/127.59/127.64/127.7/
127.85/127.87/127.96/127.99/128.07/128.12/128.25/128.28/1
28.31/128.38/128.43/128.46/128.49/128.68/128.72/128.86/13
3.4/133.5/135.06/135.08(d, arom.), 135.8/135.9/136.3/ 137.4/
137.7/138.0/138.1(s, arom.). FABMS m/z 987 [M –
CF3SO3–]+ (pos.), 149 [CF3SO3]– (neg.).
Compound 22: [α]25D −5.67 (c 1.50, CHCl3). IR (neat): 1
497, 1 456, 1 369, 1 340, 1 215, 1 157, 1 144, 1 086, 1 024
cm–1. 1H NMR (700 MHz, CDCl3) δ: 3.51 (1H, dd, J = 10.8,
5.2, H-1a), 3.65 (1H, dd, J = 10.8, 4.0, H-1b), 3.91 (1H, dd,
10.9, 4.0, H-7a), 3.92 (1H, ddd-like, J = 7.8, 5.2, 4.0, H-2),
4.06 (1H, d, J = 7.8, H-3), 4.14 (1H,d-like, J = 3.5, H-4), 4.15
(1H, dd, J = 10.9, 4.0, H-7b), 4.43/4.46 (each 1H, d, J = 11.6,
PhCH2), 4.44 (1H, dd, J = 4.0, 4.0, H-6), 4.47 (1H, d, J = 3.5,
H-5). 4.51/4.72 (each 1H, d, J = 12.0, PhCH2), 4.66/4.76
(each 1H, d, J = 11.6, PhCH2), 5.08/5.21 (each 1H, d, J = 8.0,
OCH2O), 7.23–7.34 (15H, m, arom.). 13C NMR (175 MHz,
XIE Wei-Jia, et al. /Chinese Journal of Natural Medicines 2013, 11(6): 676−683
682 Chin J Nat Med Nov. 2013 Vol. 11 No. 6 2013 年 11 月 第 11 卷 第 6 期

CDCl3) δ: 69.3 (C-7), 71.0 (C-1), 72.3/73.2/73.5 (PhCH2),
74.0 (C-6), 76.6 (C-5), 77.9 (C-2), 81.0 (C-4), 86.3 (C-3),
89.7 (OCH2O), 127.4/127.5/127.6/127.9/128.2/ 128.3/ 128.4
(d, arom.), 137.6/138.3/138.8 (s, arom.). FABMS m/z 499,
[M + Na]+ (pos.).
Compound 23: Colorless oil, [α]25D –1.6 (c 1.26, CHCl3).
IR (neat): 1 497, 1 454, 1 366, 1 339, 1 273, 1 215, 1 154, 1
088, 1 026 cm−1. 1H NMR (700 MHz, CDCl3) δ: 3.58 (1H, dd,
J = 10.2, 10.2, H-5a), 3.66 (1H, dd, J = 13.4, 3.8, H-1a), 3.67
(1H, dd, J = 10.2, 5.8, H-5b), 3.89 (1H, br dd, J = 10.2, 5.8,
H-4), 4.17 (1H, dd-like, J = ca. 1.5, 1.5, H-3), 4.28 (1H, dd, J
= 13.4, 1.4, H-1b), 4.30/4.34 (each 1H, d, J = 11.8, PhCH2),
4.41/4.58 (each 1H, d, J = 11.8, PhCH2O), 4.48–4.50 (1H, m,
H-2), 4.49/4.55 (each 1H, d, J = 11.8, PhCH2O), 4.63/5.04
(each 1H, d, J = 12.7, PhCH2S+), 7.04–7.39 (20H, m, arom.).
13C NMR (175 MHz, CDCl3) δ: 46.0 (C-1), 49.3 (PhCH2S+),
63.2 (C-4), 66.8 (C-5), 72.0/ 72.6/73.4 (PhCH2O), 82.1 (C-2),
83.2 (C-3), 120.7 (q, J = 320, CF3SO3–), 127.7/128.1/128.50/
128.54/128.7/128.8/128.9/129.7/130.3/130.7 (d, arom.),
127.6/135.7/135.9/136.5 (s, arom.). FABMS m/z 511 [M –
CF3SO3–]+ (pos.), 149 [CF3SO3]– (neg.).
Deprotection of sulfonium salt (21)
To a solution of compound 21 (0.05 g, 0.04 mmol) in
CH2Cl2 (2 mL) was added 1.0 mol·L1 BCl3 in CH2Cl2 (1.0
mL) at –78 °C. The mixture was allowed to reach room tem-
perature, and stirred for 12 h. The reaction was quenched by
the addition of MeOH, and the resulting mixture was con-
densed under reduced pressure. The residue was dissolved in
water (2 mL), and the resulting solution was washed with
CH2Cl2 (3 × 2 mL). The water layer was condensed under
reduced pressure to give a colorless oil, which was treated
with ion exchange resin IRA 400J (Cl– form, 100 mg) at r.t.
for 3 h. The resins were filtered off and washed with metha-
nol. The combined filtrate and the washings were condensed
to give a colorless oil, which on column chromatography
(CHCl3/MeOH/water, 20/4/1) gave neokotalanol (6, 10 mg,
60%), the 1H and 13CNMR spectroscopic properties of which
were in accord with those reported [17-19].
References
[1] Heightman TD, Vasella AT. Recent insights into inhibition,
structure, and mechanism of configuration-retaining glyco-
sidases [J]. Angew Chem Int Ed, 1999, 38 (6): 750-770.
[2] Krasikov VV, Karelov DV, Firsov LM. α-Glucosidases [J].
Biochemistry, (Moscow) 2001, 66 (3): 267-281.
[3] Lillelund VH, Jensen HH, Liang X, et al. Recent develop-
ments of transition-state analogue glycosidase inhibitors of
non-natural product origin [J]. Chem Rev, 2002, 102(2):
515-553.
[4] Herscovics A, Pinto BM, Barton DHR, et al. Glycosidases of
the asparagine-linked oligosaccharide processing pathway
[J]. Compr Nat Prod Chem, 1999, 3: 13-35.
[5] Hollander P, Pi-Sunyer X, Coniff RE. Acarbose in the treat-
ment of type I diabetes [J]. Diabetes Care, 1997, 20 (3):
248-253.
[6] Scott L, Spencer CM. Miglitol: a review of its therapeutic
potential in type 2 diabetes mellitus [J]. Drugs, 2000, 59 (3):
521-549.
[7] Cheng AYY, Josse RG. Intestinal absorption inhibitors for
type 2 diabetes mellitus: prevention and treatment [J]. Drug
Discov Today, 2004, 1 (2): 201-206.
[8] Kordik CP, Reitz AB. Pharmacological treatment of obesity:
therapeutic strategies [J]. J Med Chem, 1999, 42 (2):
181-201.
[9] Takeuchi M, Kamata K, Yoshida M, et al. Inhibitory effect of
pseudo-aminosugars on oligosaccharide glucosidases I and II
and on lysosomal α-glucosidase from rat liver [J]. J Biochem,
1990, 108 (1): 42-46.
[10] Butters TD, Dwek RA, Platt FM. Imino sugar inhibitors for
treating the lysosomal glycosphingolipidoses [J]. Glycobiol-
ogy, 2005, 15(10): 43-52.
[11] Yoshikawa M, Murakami T, Shimada H, et al. Salacinol, po-
tent antidiabetic principle with unique thiosugar sulfonium
sulfate structure from the Ayurvedic traditional medicine
Salacia reticulata in Sri Lanka and India [J]. Tetrahedron
Lett, 1997, 38 (48): 8367-8370.
[12] Yoshikawa M, Morikawa T, Matsuda H, et al. Absolute ste-
reostructure of potent α-glucosidase inhibitor, Salacinol,
with unique thiosugar sulfonium sulfate inner salt structure
from Salacia reticulate [J]. Bioorg Med Chem, 2002, 10 (5):
1547-1554.
[13] Yoshikawa M, Murakami T, Yashiro K, et al. Kotalanol, a
potent α-glucosidase inhibitor with thiosugar sulfonium sul-
fate structure, from antidiabetic Ayurvedic medicine Salacia
reticulata [J]. Chem Pharm Bull, 1998, 46 (8): 1339-1340.
[14] Yoshikawa M, Xu F, Nakamura S, et al. Salaprinol and pon-
koranol with thiosugar sulfonium sulfate structure from Sa-
lacia prinoides and α-glucosidase inhibitory activity of
ponkoranol and kotalanol desulfate [J]. Heterocycles, 2008,
75 (6): 1397-1405.
[15] Minami Y, Kuriyama C, Ikeda K, et al. Effect of
five-membered sugar mimics on mammalian glycogende-
grading enzymes and various glucosidases [J]. Bioorg Med
Chem, 2008, 16 (6): 2734-2740.
[16] Tanabe G, Xie W, Ogawa A, et al. Facile synthesis of
de-O-sulfated salacinols: Revision of the structure of neosa-
lacinol, a potent α-glucosidase inhibitor [J]. Bioorg Med
Chem Lett, 2009, 19 (8): 2195-2198.
[17] Ozaki S, Oe H, Kitamura SJ. α-Glucosidase inhibitor from
Kothala-himbutu (Salacia reticulata Wight) [J]. Nat Prod,
2008, 71 (6): 981-984.
[18] Oe H, Ozaki S. Hypoglycemic effect of 13-membered ring
thiocyclitol, a novel α-glucosidase inhibitor from
Kothala-himbutu (Salacia reticulata) [J]. Biosci Biotechnol
Biochem, 2008, 72 (7): 1962-1964.
[19] Muraoka O, Xie W, Tanabe G, et al. On the structure of the
bioactive constituent from Ayurvedic medicine Salacia re-
ticulata: revision of the literature [J]. Tetrahedron Lett, 2008,
49 (51): 7315-7317.
[20] Xie W, Tanabe G, Akaki J, et al. Isolation, structure identifi-
cation and SAR studies on thiosugar sulfonium salts,
neosalaprinol and neoponkoranol, as potent α-glucosidase
inhibitors [J]. Bioorg Med Chem, 2011, 19 (21): 2015-2022.
[21] Muraoka O, Tanabe G. WO2011093471, 2011.
[22] Xie W, Tanabe G, Xu J, et al. Research progress of synthesis
XIE Wei-Jia, et al. /Chinese Journal of Natural Medicines 2013, 11(6):676−683
2013 年 11 月 第 11 卷 第 6 期 Chin J Nat Med Nov. 2013 Vol. 11 No. 6 683

and structure-activity relationship studies on sulfonium-type
α-glucosidase inhibitors isolated from Salacia genus plants
[J]. Mini-Rev Org Chem, 2013, 10: 141-159.
[23] Jayakanthan K, Mohan S, Pinto BM, et al. Structure proof
and synthesis of kotalanol and de-O-sulfonated kotalanol,
glycosidase inhibitors isolated from an herbal remedy for the
treatment of type-2 diabetes [J]. J Am Chem Soc, 2009, 131
(15): 5621-5626.
[24] Muraoka O, Xie W, Osaki S, et al. Characteristic alkaline
catalyzed degradation of kotalanol, a potent α-glucosidase
inhibitor isolated from Ayurvedic traditional medicine Sala-
cia reticulata, leading to anhydroheptitols: another structural
proof [J]. Tetrahedron, 2010, 66 (21): 3717-3722.
[25] Tanabe G, Otani T, Cong W, et al. Biological evaluation of
3’-O-alkylated analogs of salacinol, the role of hydrophobic
alkyl group at 3’-position in the side chain on the
α-glucosidase inhibitory activity [J]. Bioorg Med Chem Lett,
2011, 21 (11): 3159-3162.
[26] Tanabe G, Nakamura S, Tsutsui N, et al. In silico design,
synthesis and evaluation of 3’-O-benzylated analogs of sa-
lacinol, a potent α-glucosidase inhibitor from Ayurvedic tra-
ditional medicine “Salacia” [J]. Chem Commun, 2012, 48
(69): 8646-8648.
[27] Grathwohl M, Schmidt RR. Solid phase syntheses of oligo-
mannosides and of a lactosamine containing milk trisaccha-
ride using a benzoate linker [J]. Synthesis, 2001, 15:
2263-2272.
[28] Mayer TG, Schmidt RR. Glycosylphosphatidylinositol (GPI)
anchor synthesis based on versatile building blocks. Total
synthesis of a GPI anchor of yeast [J]. Eur J Org Chem,
1999, 1999 (5): 1153-1165.
[29] Angyal SJ. Conformations of acyclic sugar derivatives, part
VI. The carbon-13 NMR spectra and the conformations of
heptitols in solution [J]. Carbohydr Res, 1984, 126 (1):
15-26.
[30] Lewis D. Proton nuclear magnetic resonance spectra and
conformations of six heptitols in deuterium oxide [J]. J
Chem Soc Perkin Trans 2, 1986: 467-470.
[31] Ishizu T, Winarno H, Tsujino E, et al. Indonesian medicinal
plants. XXIV. Stereochemical structure of perseitol·K+ com-
plex isolated from the leaves of Scurrula fusca (Lorantha-
ceae) [J]. Chem Pharm Bull, 2002, 50 (4): 489-492.
[32] Molecular energy was calculated by semi-empirical molecu-
lar orbital AM1 method in MOPAC package.

五层龙属植物中提取的具有α-葡萄糖苷酶抑制活性的天然产物
neokotalanol的全合成
谢唯佳 1，田边元三 2，筒井望 2，吴晓明 1，村冈修 2*
1中国药科大学药学院，南京 210009，中国;
2近畿大学药学部，大阪府东大阪市小若江 3-4-1, 577-5802，日本
【摘 要】 以苄基保护的硫糖和由 D-甘露糖为起始原料, 合成得到的三氟甲磺酸酯间的偶合反应作为关键反应合成了从五
层龙属植物根茎部提取的具有很强α-葡萄糖苷酶抑制活性的天然产物 neokotalanol。研究发现该关键的偶合反应的产物是由反应
温度直接决定的，反应除了得到目标锍糖化合物的同时, 还分离出了通过经典的分子内成环反应所得到 2, 4, 7-三氧代二环壬烷
类衍生物。
【关键词】 Neokotalanol; 全合成; 五层龙; α-葡萄糖苷酶抑制剂

【基金项目】 中央高校基本科研业务费(JKZ2011003), 国家自然科学基金青年基金(No. 81202409), 江苏省自然科学基金青年基
金 (SBK201240392), 教育部留学回国人员启动经费（2013）, 人事部留学人员科技活动项目择优资助经费（2013）