Diastereoselective Synthesis of Salacinol-Type α-Glucosidase Inhibitors

Article abstract of DOI:10.1021/acs.joc.7b02566

A facile and highly diastereoselective approach toward the synthesis of potent salacinol-type α-glucosidase inhibitors, originally isolated from plants of the genus "Salacia", was developed using the S-alkylation of thiosugars with epoxides in HFIP (～90%, dr, α/β = ～ 26/1). The dr ratio of the product was significantly improved by the protocol as compared to that of the conventional S-alkylation of thiosugars (dr, α/β = ～ 8/1). The protocol could be used for gram scale synthesis of the desired compounds. The 3′-O-benzylated salacinol analogs, which are the most potent in vitro inhibitors to date, were synthesized and evaluated in vivo; all analogs suppressed blood glucose levels in maltose-loaded mice, at levels comparable to those of the antidiabetic agent, voglibose.

Full text of DOI:10.1021/acs.joc.7b02566

Subscriber access provided by Columbia University Libraries
Article
Diastereoselective synthesis of salacinol-type #-glucosidase inhibitors
Fumihiro Ishikawa, Kazumi Jinno, Eri Kinouchi, Kiyofumi Ninomiya, Shinsuke
Marumoto, Weijia Xie, Osamu Muraoka, Toshio Morikawa, and Genzoh Tanabe
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.7b02566 • Publication Date (Web): 30 Nov 2017
Downloaded from http://pubs.acs.org on November 30, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
The Journal of Organic Chemistry is published by the American Chemical Society.
1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 28
The Journal of Organic Chemistry
1
2
3
Diastereoselective synthesis of salacinol-type α-glucosidase inhibitors
4
5
6
7
8
9
Fumihiro Ishikawa,† Kazumi Jinno,† Eri Kinouchi,‡ Kiyofumi Ninomiya,‡ Shinsuke Marumoto,§ Weijia
Xie,¶ Osamu Muraoka,‡ Toshio Morikawa,‡ and Genzoh Tanabe,†,‡,*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
†Faculty of Pharmacy, ‡Pharmaceutical Research and Technology Institute, §Joint Research Center,
Kindai University, 3ꢀ4ꢀ1 Kowakae, Higashiꢀosaka, Osaka 577ꢀ8502, Japan
¶State Key Laboratory of Natural Medicines and Department of Medicinal Chemistry, China
Pharmaceutical University, Nanjing 210009, P. R. China
Table of Contents
OH
R'
O
S
, HFIP
R'
deprotection
RO
Cl-
S+
reflux
RO
R = Bn or PMB
OR
HO
(~89%, dr, α/β = ~26/1)
OBn
O
HO
OH
OPMB
9 : R' =
Salacinol-type
α-glucosidase
16 : R' =
OC4H6X
O
X = CH3, Cl, CF3, NO2
O
1
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 28
1
2
3
Abstract
4
5
6
7
8
A facile and highly diastereoselective approach towards the synthesis of potent salacinolꢀtype αꢀ
glucosidase inhibitors, originally isolated from plants of the genus “Salacia”, was developed using the
Sꢀalkylation of thiosugars with epoxides in HFIP (~90%, dr, α/β = ~26/1). The dr ratio of the product
was significantly improved by the protocol as compared to that of the conventional Sꢀalkylation of
thiosugars (dr, α/β = ~8/1). The protocol could be used for gram scale synthesis of the desired
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
compounds. The 3'ꢀOꢀbenzylated salacinol analogs, which are the most potent in vitro inhibitors to date,
were synthesized and evaluated in vivo; all analogs suppressed blood glucose levels in maltose–loaded
mice, at levels comparable to those of the antiꢀdiabetic agent, voglibose.
2
ACS Paragon Plus Environment

Page 3 of 28
The Journal of Organic Chemistry
1
2
3
Introduction
4
5
6
7
8
9
At the end of 1990s, salacinol (1) was isolated from the roots and stems of Salacia reticulata, a
plant that has been used for treating diabetes in Ayurvedic medicine. The αꢀglucosidase inhibitory
activity of 1 was as strong as those of current antidiabetics, such as voglibose and acarbose.1 Following
the discovery of 1, the side chain extended homologs, kotalanol (2)2 and ponkoranol (3)3 as well as
their deꢀOꢀsulfonated versions neosalacinol (4),4 neokotalanol (5),5 and neoponkoranol (6)6 were
isolated from plants in the same genus in the form of sulfonium salts, and these compounds also
exhibited antidiabetic activities (Fig. 1).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
OH OH
OH OH OH
OH
OH
3'
OH
4'
1'
2'
OH
S+ OR OH
S+ OR OH
S+ OR
HO
HO
HO
HO
ponkoranol (3) : R = SO3
neokotalanol (5) : R = H neoponkoranol (6) : R = H
OH
HO
OH
HO
OH
-
-
-
kotalanol (2) : R = SO3
salacinol (1) : R = SO3
neosalaciniol (4) : R = H
Figure 1. A new class of natural αꢀglucosidase inhibitors
Towards that end, human clinical trials revealed the efficacy of extracts from Salacia reticulata as a
potential treatments for type 2 diabetes mellitus with minimal side effects.7 Owing to the high
inhibitory activities and intriguing structures of 1–6, they have garnered considerable attention, and
their total syntheses8 as well as intensive structureꢀactivity relationship (SAR) studies9 have been
carried out worldwide on this class of αꢀglucosidase inhibitors; numerous analogs have been
synthesized to date, and structural determinants of the glycosidase inhibitory activities have been
clarified to some extent. However, no candidates with superior inhibitory activities to those of the
original sulfonium salts 1–6 were developed until a series of neosalacinols (7a–7m) with 3'ꢀOꢀ(monoꢀ
substituted benzyl) groups were synthesized recently.9b Among them, orthoꢀsubstituted derivatives 7b,
7e, 7h, and 7k were found to be stronger inhibitors than the corresponding metaꢀ and paraꢀsubstituted
ones, and their in vitro inhibitory activities (IC50 = 0.13–0.66 ꢀM against rat intestinal maltase) were
3
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 28
1
2
3
significantly superior to those of parent compounds 1 and 4 [IC50 (ꢀM): 1, 5.2, 4, 8.0], indicating that
4
5
they are the most potent compounds among cyclic sulfonium species synthesized to date (Scheme 1).9b
6
7
8
9
Scheme 1. Previous synthetic protocol for 7a–7m
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
OH
1) IRA-400J (Cl- form)
CH3OH, rt
2) 80% aq. TFA
CHCl3, rt
OPMB
OH
X
O
9
X= a: H
b: o-CH3, c: m-CH3, d: p-CH3,
e: o-Cl, f: m-Cl, g: p-Cl
-
OPMB
OH
BF4
S
Cl-
S+
O
X
S+
O
X
PMBO
PMBO
PMBO
PMBO
HBF4 Et2O
CH2Cl2, -60°C
HO
h: o-CF3, i: m-CF3, j: p-CF3,
k: o-NO2, l: m-NO2, m: p-NO2
OPMB
(~ 82%)
OPMB
HO
OH
8a
(~ 81%, dr, α/β = ~ 8/1)
7
+ β-isomer (β-10)
α-10
Despite their potent inhibitory activities, synthetic issues associated with the originally reported
synthesis of 7a–7m severely limited access to these molecules, precluding the assessment of their
effects in vivo. Specifically, the original synthesis of 7a–7m utilized the acid mediated Sꢀalkylation of
thiosugars (8a) with epoxides (9a–9m) as a key reaction. However, the reaction exhibited a low
diastereoselectivity, albeit good yields, and the diastereomeric ratio of the products 10 never exceed
α/β = ~8/1. Separation of the αꢀ and βꢀstereoisomers of 10 was quite difficult because of their
extremely similar polarity, and repeated separations were required to obtain sufficient amounts of the
desired αꢀisomer. As shown in Scheme 2, poor diastereoselectivity in the Sꢀalkylation process in the
conventional synthesis of salacinol analogs was also reported,8c,10 in which both αꢀ and βꢀstereoisomers
were obtained as either a poorly separable or an inseparable mixture. This is a significant drawback of
the thiosugarꢀSꢀalkylationꢀbased approach toward salacinolꢀtype sulfonium salts. Thus, a facile and
effective protocol is required for these compounds, for which the conventional protocol is inefficient.
In the course of our independent efforts to synthesize natural αꢀglucosidase inhibitor neoponkoranol (6),
we discovered the unique reactivity of 8b with epoxide 16 in 1,1,1,3,3,3ꢀhexafluoroisopropanol (HFIP),
in which the key sulfonium salt α-17 was formed with excellent diastereoselectivity [dr, α/β = ~26/1,
Scheme 3]. Herein, we describe a protocol that highlights the surprises associated with the highly
4
ACS Paragon Plus Environment

Page 5 of 28
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
diastereoselective synthesis of αꢀisomer of sulfonium salt in HFIP. Additionally, the protocol
successfully yielded potent in vitro inhibitors 7b, 7e, 7h, and 7k in good yields. Based on the
evaluation of their activity in mice, 7b, 7e, 7h, and 7k were revealed to be stronger in vivo than natural
seeds 1–6.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. The key reaction in conventional synthesis of salacinol-analogues
CH3
I-
S+
CH3I, DMF, 45°C,12 h
BnO
(69%, dr, α/β = 7/1)
BnO
OBn
13
OBn
O
OBn
OBn
OBn
O
S
O
O
O
OBn
-
O
S
, HFIP, 70°C, 42 h
OBn
11
S+
OSO3
BnO
(50%, dr, α/β =~ 4/1)
BnO
BnO
OBn
8b
BnO
OBn
14
OBn
OCH3
OBn
OCH3
OCH3
O
O
O
O
O
R = perbenzylated
maltose
O
OCH
3, DCM, rt, 24 h
OTf OR
S+
12
OR
BnO
(95%, dr, α/β = 5/1)
15
BnO
OBn
Results and Discussion
Strong Brønsted acids such as CF3SO3H and HBF4 are known to be good proton sources for the Sꢀ
alkylation of sulfides with epoxides, and the anions function as counter anions for the sulfonium
cation.11 Due to the excellent protonꢀdonating ability, the reaction rapidly (~2 h) proceeded to give the
βꢀhydroxylated sulfonium salts in good yields. However, HFIP has not been used as an acid in Sꢀ
alkylation to date because of its weak acidity (pKa = 9.3). Therefore, we set out to determine whether
HFIP would function as an acid in the Sꢀalkylation, and if the resultant sulfonium salts with
–
–
(CF3)2CHO–, which is a significantly stronger base than CF3SO3 or BF4 , could be isolated. Thus, the
Sꢀalkylation of 8b (1.3 eq.) with 16 (1.0 eq.) was conducted in HFIP (Scheme 3). The reaction
proceeded slowly under reflux; after 18 h, the nearly complete consumption of 16 was observed by
TLC. After the removal of HFIP in vacuo, a mixture of α- and β-17 (~26:1 ratio) was observed, as
1
judged by the integration of the individual C6′ methine doublets in the H NMR spectrum of the crude
13
reaction mixture [Supporting Information (SI)ꢀFig. 1]. The C NMR spectrum of the reaction mixture
5
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 28
1
2
3
4
5
6
7
8
9
showed 11 resonances likely related to the 11 carbons (C1–C5 and C1′–C6′) of the sulfonium moiety of
α-17, along with two peaks corresponding to the 2 carbons of (CF3)2CHO–, indicating the successful
sulfonium ion formation (Table 1). Thus, HFIP was confirmed to function as an acid for the Sꢀ
alkylation of 8b and the conjugate anion of HFIP [(CF3)2CHO–] could play the role as the counter
anion for sulfonium ions; however, sulfonium salts 17 free from HFIP gradually decomposed into
starting materials 8b and 16 in CDCl3 upon standing at room temperature, according to the NMR
analysis. This suggested that purification of the sulfonium salts 17 by column chromatography would
be impossible. As expected, attempts to purify the sulfonium salts resulted in unsuccessful recovery of
the starting materials 8b and 16. The lability of intermediates 17 could be attributed to the proton
abstraction from the hydroxyl at C2' by (CF3)2CHO–, because the corresponding sulfonium salts with
Cl– are known to be stable. Thus, we abandoned the purification of 17 and tried to exchange
(CF3)2CHO– with Cl–. An attempt to exchange (CF3)2CHO– with Cl– by IRAꢀ400J (Cl– form) resin led
to the complete decomposition to the starting materials, whereas (CF3)2CHO– could be replaced with
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Cl– by in situ treatment with methanolic HCl. Following the neutralization of excess HCl with NaHCO3,
separable chlorides 18 were obtained in good yields without significant loss in the diastereomeric ratio
(dr, α/β = ~23/1). Notably, the analogous reaction between 8b and 16 in the refluxing 2ꢀpropanol over
18 h left the starting materials unchanged, instead of yielding the desired product. Thereby, the acidity
of HFIP was suggested to play a critical role in promoting this reaction. The FABꢀMS spectrum of α-
18 obtained in positive mode showed peaks at m/z 713 due to the sulfonium cation moiety, and the
compound exhibited a positive reaction in the Beilstein test, indicating the presence of chlorine in the
13
compound. As all of the C NMR data for α-18 and α-17 agreed well with each other (Table 1), the
structure of α-18 was confirmed by direct comparison with an authentic sample synthesized according
to the procedure reported previously.8f Briefly, 8b was subjected to HBF4•(C2H5)2O mediated Sꢀ
alkylation with 16 to give a ~7/1 diastereomeric mixture of α- and β-19 in 85% combined yield. After
6
ACS Paragon Plus Environment

Page 7 of 28
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
–
purification, the BF4 anion of α-19 was exchanged with Cl– using IRAꢀ400J (Cl– form) resin to
produce α-18 in 96% yield. Finally, the stereoconfiguration of the side chain of α-18 relative to the
BnOCH2 moiety at Cꢀ4 was confirmed to be anti on the basis of NOE spectroscopy experiments
(Scheme 3), indicating that αꢀfacial Sꢀalkylation exclusively took place on the sulfur atom.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 3. Synthesis of neoponkoranol (6) via the S-alkylation of 8b with 16
BnO
HO
2'
HO
1'
H2C
OBn
4'
OBn
4'
NOE
2'
O
O
O
HFIP, reflux
1'
(CF3)2CHO-
HCl, MeOH, 0 °C (in situ),
then NaHCO3
O
O
Cl-
S+
S+
O
O
O
16
+
(dr, α/β = ~26/1)
H
6'
1
1
6'
6
O
O
BnO
BnO
(89%, dr, α/β = ~23/1)
IRA-400J (Cl- form)
CH3OH, rt
3
6
3
BnO
OBn
BnO
OBn
+ β-isomer (β-18)
8b
+ β-isomer (β-17)
α-17
α-18
HO
OBn
1) H2, Pd-C, 20% aq TFA
dioxane, 50 °C
2) NaBH4, MeOH, 0 °C
-
BF4
BnO
O
IRA-400J (Cl- form), CH3OH, rt
(96%)
S+
HBF4 Et2O, CH2Cl2, -60°C
O
(60%)
O
(85%, dr, α/β = ~7/1)
neoponkoranol (6)
BnO
OBn
+ β-isomer (β-19)
α-19
Table 1. 13C NMR data (125 MHz, CDCl3) for eleven carbons (C1–C5 and C1’–C6’) of sulfonium
salts α-17, α-18, and α-19
C1
C2
C3
C4
C5
C1’
51.6
52.1
51.3
C2’
65.0
64.9
65.4
C3’
82.4
82.1
81.9
C4’
81.3
81.0
81.0
C5’
82.1
82.8
82.4
C6’
α-17a
α-18
α-19
48.3
48.3
48.0
82.3
82.3
82.4
81.9
82.2
82.2
66.9
66.5
67.1
66.5
66.8
66.6
105.1
105.0
105.1
aTwo signal due to two carbons of (CF3)2CHO– moiety were also observed at
28 Hz) (see SIꢀFig. 1ꢀ2).
δC 122.7 (q, J = 283 Hz) and 69.5 (hept, J =
Next, we attempted to elucidate the mechanistic details behind the high diastereoselectivity in HFIP.
The reaction between 8b (1.3 eq.) and 16 (1.0 eq.) was conducted in HFIPꢀd2 at 30 °C in a sealed tube.
1H NMR indicated that the reaction was complete in ~42 h, giving a ~9/1 diastereomeric mixture of α-
and β-17. The ratio of α- and β-17 in the mixture was nearly unchanged at 30 °C after 6 h (SIꢀFig. 2, 3),
whereas the ratio dramatically increased and reached ~24/1 upon heating at 55–60 °C for 13 h. Notably,
the reꢀformation of a small amount of 16 was detected after heating for 3 h, and to our surprise, the
amount of reꢀformed 16 was nearly unchanged at around 2% during the last 6 h (SIꢀFig. 4, 5). This
7
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 28
1
2
3
4
5
6
7
suggested that the Sꢀalkylation and retroꢀreaction were at equilibrium at this temperature. On the basis
of these results, we surmised that three pathways contributed to the highly diastereoselective formation
of α-17 in HFIP (Scheme 4): (1) Sꢀalkylation would proceed with ~9/1 faceꢀselectivity, mainly giving
8
9
α-17; (2) the retroꢀreaction of β-17 to starting materials 8b and 16 was approximately four times faster
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
than that of α-17 (SIꢀFig. 1). The reaction would lead to the hydrogen abstraction of the hydroxyl at the
C2′ position by (CF3)2CHO–, and the resulting alkoxide at C2′ would participate in an intramolecular
SN2 attack on the C1′ carbon to give 8b and 16, which would subsequently react with each other. (3)
The isomerization of the sulfonium center12 from thermodynamically less stable β-17 (–89.2 kcal/mol)
to more stable α-17 (–93.3 kcal/mol) was accelerated upon heating at 55–60 °C, increasing the α/β
ratio. The proposed stability of the sulfonium structure was supported by the isomerization experiment
of the corresponding chloride β-18 to α-18, in which the ratio of α/β = ~9/1 increased to ~24/1 after
heating at 55–60 °C in HFIPꢀd2 for 14 h (SIꢀFig. 6, 7). Thus, the reaction cycle consisting of the
reversible Sꢀalkylation and isomerization would be repeated until the highly diastereoselective
formation of α-17 was complete.
Scheme 4. Plausible mechanism for the diastereoselective formation of α-17 in HFIP
H
(CF3)2CHO-
O
2'
OBn
4'
kinetically favored α facial
attack of 8b
O
S+
O
6'
O
BnO
slow retro-reaction
(CF3)2CHO
H
(-93.3 kcal/mol)
BnO
OBn
slow
α-17
O
OBn
O
O
fast
thermal isomerization
O
S
16
BnO
H
(CF3)2CHO-
fast retro-reaction
OBn
4'
O
OBn
BnO
8b
2'
O
S+
O
6'
O
BnO
kinetically unfavored β facial
attack of 8b
BnO
OBn
β-17 (-89.2 kcal/mol)
8
ACS Paragon Plus Environment

Page 9 of 28
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Given the reactivity observed in HFIPꢀd2, we set out to develop a rapid and highly
diastereoselective approach toward α-17. Thus, the reaction between 8b (1.3 eq.) and 16 (1.0 eq.) in
HFIPꢀd2 was monitored by NMR spectroscopy. As shown in Fig. 2, after heating at 55–60 °C for 42 h,
~97% of 16 was consumed, and α-17 (α/β = ~27/1) was formed exclusively. The conversion of 16 was
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
nearly unchanged upon heating for another 23 h. On the contrary, the α/β ratio dramatically increased
upon heating for 24 h (30 min: ~8/1→24 h: ~25/1), and was nearly unchanged after 41 h (65 h: ~28/1)
(SIꢀFig. 8, 9). Finally, in order to accelerate the Sꢀalkylation rate, a mixture of 16 was treated with five
equivalents of 8b in nonꢀdeuterated HFIP. Heating the reaction mixture at reflux for 6 h followed by
the usual workꢀup gave a ~23/1 mixture of α- and β-18 in 89% yield. After chromatographic separation,
the benzyl moieties of α-18 were removed by hydrogenolysis, and the resulting crude hemiacetal was
reduced with NaBH4 to give neoponkoranol (6) in 60% yield (Scheme 3).
Figure 2. (A) Conversion % of 16 in the S-alkylation in HFIP-d2 at 55–60°C, and (B)
diastereomeric ratio of α- and β-isomer. Conversion % and dr ratio (α/β) were estimated by 500
MHz 1H NMR spectrum (Time interval NMR spectra were given in SI-Fig. 8, 9)
With the conditions for the oneꢀpot highly diastereoselective synthesis of the αꢀisomer in hand, we
applied the reaction to the gramꢀscale production of intermediates for 7b, 7e, 7h, and 7k in order to
evaluate the in vivo activity (Scheme 5). From epoxides 9b, 9e, 9h, and 9k (1.0–1.3 g), α-20b, α-20e,
9
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 28
1
2
3
α-20h, and α-20k were obtained predominantly with (CF3)2CHO– as the counter anion; the counter
4
5
–
anion was then exchanged to BF4 by treating in situ with HBF4•(C2H5)2O at 0 °C to give the
6
7
8
9
corresponding sulfonium salts α-10b, α-10e, α-10h, and α-10k in good yields and good
diastereoselectivities. After purification, they were converted to 7b, 7e, 7h, and 7k in around 80%
yields (~800–950 mg) according to a previously reported method.9b
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 5. Synthesis of 7a, 7b, 7e, 7h and 7k via the S-alkylation in HFIP
OH
9, HFIP
reflux
(CF3)2CHO-
OPMB
HBF4 Et2O, 0 °C (in situ),
then NaHCO3
S+
O
X
8a
7b, 7e, 7h and 7k
α-10b, 10e, 10h and 10k
PMBO
PMBO
(~88%, dr, α/β = ~ 18/1)
lit.9b
OPMB
X= b: o-CH3, e: o-Cl,
h: o-CF3 k: o-NO2
α-20
+ β-isomer (α-20)
With suitable amounts of 7b, 7e, 7h, and 7k in hand, their αꢀglucosidase inhibitory activities were
tested against human small intestinal maltase in vitro, and their potencies as inhibitors were compared
with those of several reference standards, such as salacinol (1), neosalacinol (4), and three currently
used antidiabetics (voglibose, acarbose, and miglitol). The synthesized analogs inhibited human small
intestinal maltase more potently than 1, 4, and three commonly used antidiabetics (Table 2).
Table 2. Ebind (kcal/mol) to hNtMGAM, and IC50 and Ki values (ꢀM) values against human
intestinal maltase
a
Entry
Compound
7b (oꢀCH3)
7e (oꢀCl)
Ebind
–37.2
–41.6
–37.5
–38.9
–37.0
–36.4
–
IC50
0.58
0.11
0.22
0.15
4.9
Ki
1
2
0.073
0.023
0.035
0.015
0.44
1.2
3
7h(oꢀCF3)
7k (oꢀNO2)
Salacinol (1)
Neosalacinol (4)
Voglibose
Acarbose
4
5
6
9.0
7
1.3
0.17
2.6
8
–
16.7
3.7
9
alit9b.
miglitol
–
0.57
10
ACS Paragon Plus Environment

Page 11 of 28
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Among them, 7e, 7h, and 7k showed strong in vitro inhibitory activities (IC50: 0.11ꢀ0.22 ꢀM), and
were approximately 20–45 times more active than salacinol (1, IC50: 4.9 ꢀM), although their potency
was weaker than the previously reported Ebind values9b of these molecules to hNtMGAM. The
discrepancies between the predicted and obtained experimental results obtained in this study might be
related to the existence of other catalytic domains such as sucraseꢀisomaltase (CtSI) and/or Cꢀterminal
maltaseꢀglucoamylase (CtMGAM), which have been proven to affect the maltose hydrolysis.13 On the
contrary, the inhibition mode of these analogs was revealed to be competitive by LineweaverꢀBurk
plots of inhibition kinetics against human small intestinal maltase (SIꢀFig. 10). The Ki values were in
the range of 15–73 nM, indicating that all 3′ꢀOꢀbenzylated analogs could function as superior
competitive inhibitors against human small intestinal maltase. Next, the activities of the analogs were
examined in vivo in mice. Following the oral administration of maltose, mice in the control group
showed a maximum blood glucose levels after 30 min. On the contrary, oral administration (0.3 mg/kg)
of 7b, 7e, 7h, and 7k prior to maltose loading significantly suppressed the blood glucose levels,
indicating that these compounds are good in vivo αꢀglucosidase inhibitors. The degree of suppression
by 7b, 7e, 7h, and 7k was superior to that of salacinol (1) (Table 3). Notably, 7b, which was a weaker
inhibitor than 7e, 7h, and 7k in vitro, showed a nearly equal suppression of blood glucose levels in vivo
to that of 7e, 7h, and 7k. On the contrary, examination of doseꢀdependent effects indicated that the
elevation in blood glucose levels in maltoseꢀloaded mice was effectively suppressed by 0.1 mg/kg dose
of 7k and the potency was nearly equivalent to that of the antidiabetic, voglibose (Table 4). Thus, the
3’ꢀOꢀsulfonium salts 7b, 7e, 7h, and 7k were revealed, for the first time, to be highly potent in vivo αꢀ
glucosidase inhibitors.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
11
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 28
1
2
3
Table 3. Inhibitory effects of 7b, 7e, 7h, 7k and salacinol (1) on blood glucose levels in mice
Treatment
Dose
N
Blood glucose (mg/dL)
4
5
6
7
8
9
(mg/kg, po)
0 min
77.8±5.0
70.8±5.5
74.0±5.6
72.7±3.7
70.0±3.2
74.7±4.2
77.8±3.0
15 min
30 min
60 min
120 min
90.3±5.1
100.6±4.3
95.8±4.6
102.5±4.1
96.2±4.9
107.5±5.9
106.8±6.7
180 min
Normal
—
—
4
8
6
6
6
6
6
99.0±6.2##
195.5±10.0
158.0±9.3**
124.8±3.2**
118.3±4.4**
119.2±6.4**
116.5±2.7**
99.0±2.1##
196.3±10.7
165.8±5.8*
128.2±3.2**
130.3±4.9**
130.5±9.0**
127.0±6.4**
98.3±2.8##
158.1±9.8
143.0±8.1
127.5±6.6*
123.0±6.5**
134.0±8.8
125.8±4.3*
82.8±4.1
90.4±6.1
84.0±2.8
91.5±4.1
82.2±5.4
82.3±3.5
96.5±7.3
Control
Salacinol (1)
7b (oꢀCH3)
7e (oꢀCl)
0.3
0.3
0.3
0.3
0.3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7h (oꢀCF3)
7k (oꢀNO2)
Each value represents the mean±SEM. Significantly different from the control:*p<0.05,**p<0.01,#p<0.05,##p<0.01. Normal group:
administrated a distilled water; Control group: administrated a 10% (w/v) maltose solution (1 g/kg).
Table 4. Dose-dependent effect of 7k, salacinol (1) and voglibose on blood glucose levels in mice
Treatment
Dose
(mg/kg, po)
—
N
Blood glucose (mg/dL)
0 min
53.3±3.6
59.4±2.0
63.2±4.8
61.7±3.8
63.5±5.4
62.5±2.8
62.8±7.9
59.0±5.7
15 min
30 min
60 min
120 min
65.8±8.0#
85.5±4.5
80.3±3.3
82.2±5.5
93.5±11.1
81.3±6.1
85.8±4.0
71.3±5.0
180 min
63.0±5.4
82.3±3.3
73.0±5.0
74.2±3.4
79.2±4.3
74.5±3.6
79.2±4.4
67.8±3.9
Normal
4
8
6
6
6
6
6
6
71.0±3.7##
166.5±6.4
130.8±4.3**
108.5±7.1**
137.8±11.8*
161.8±5.9
138.3±4.6*
107.2±6.9**
73.3±7.3##
163.4±6.6
135.3±8.9
119.2±10.4**
155.5±8.9
162.3±8.2
142.5±7.9
98.5±7.1**
80.3±8.4##
135.6±8.0
124.0±9.6
113.5±6.6
140.8±12.4
131.0±6.2
127.5±3.5
94.5±6.4**
Control
—
Voglibose
0.1
0.3
Salacinol (1)
7k (oꢀNO2)
0.3
0.03
0.1
0.3
Each value represents the mean±SEM. Significantly different from the control:*p<0.05,**p<0.01,#p<0.05,##p<0.01. Normal group:
administrated a distilled water; Control group: administrated a 10% (w/v) maltose solution (1 g/kg).
In conclusion, we have developed an easy and highly diastereoselective synthetic protocol toward
salacinolꢀtype αꢀglucosidase inhibitors via the Sꢀalkylation of thiosugars with epoxides in HFIP.
Compared to the conventional protocol9 for the Sꢀalkylation of thiosugars, this method proved far more
amenable to scaleꢀup, enabling the preparation of large amounts of the key intermediates required to
produce the desired inhibitors. During these studies, a cooperative mechanism consisting of reversible
Sꢀalkylation and thermal isomerization of the sulfonium salts was found to contribute to the highly
diastereoselective formation of the intermediates. We have also established that sulfonium salts 7 are
the most potent in vivo inhibitors among salacinol analogs synthesized to date. We have also proved
that a series of sulfonium salts 7 are more potent in vivo inhibitors than the parent sulfonium salts 1,
and the potency was nearly equivalent to that of the antidiabetic agent voglibose.
12
ACS Paragon Plus Environment

Page 13 of 28
1
The Journal of Organic Chemistry
2
3
EXPERIMENTAL SECTION
4
5
6
General Experimental Details. IR spectra were measured on a FTꢀIR spectrophotometer. NMR
7
8
spectra were recorded on a FTꢀNMR spectrometers (1H, 500 or 800 MHz; 13C, 125 or 200 MHz).
9
Chemical shifts (δ) and coupling constants (J) are given in ppm and Hz, respectively. TMS was used
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
as an internal standard in the measurement of NMR spectra in CDCl3, CD3OD or HFIPꢀd2. 1D NMR
peak assignments were confirmed by COSY and HSQC spectra. Lowꢀresolution and highꢀresolution
mass spectra were recorded on a doubleꢀfocusing mass spectrometer (FAB). Optical rotations were
determined with a digital polarimeter. Column chromatography was performed over silica gel (45–106
ꢀM). HPLC was performed on a DAISOPAKꢀSP120ꢀ5ꢀODSꢀBP (20x250 mm) with a refractive index
detector. All the organic extracts were dried over anhydrous Na2SO4 prior to evaporation.
Conformational energies of sulfonium structure of α- and β-17 were calculated using MOPAC7 with
AM1 Hamiltonian.
S-Alkylation of Thiosugar (8b, 1.3 eq.) with Epoxide (16) in HFIP. A mixture of thiosugar
(8b, 374 mg, 0.89 mmol), epoxide (16, 200 mg, 0.68 mmol), and HFIP (4 mL) was heated under reflux
for 12 h. Removal of the solvent in vacuo left a colorless oil (580 mg). 1H NMR spectrum of the crude
mixture, which was immediately measured in CDCl3 after evaporation of HFIP, indicated that the Sꢀ
alkylation gave a ca. ca. 26/1 mixture of 1,4ꢀdideoxyꢀ1,4ꢀ[(R)ꢀ(6ꢀdeoxyꢀ3ꢀOꢀbenzylꢀ1,2ꢀOꢀ
isopropylideneꢀαꢀDꢀglucofuranosꢀ6ꢀyl)episulfoniumylidene]ꢀ2,3,5ꢀtriꢀOꢀbenzylꢀDꢀarabinitol
1,1,1,3,3,3ꢀhexafluoroisopropoxide (α-17) and its βꢀisomer (β-17). After the crude mixture was
allowed to stand at room temperature for 12 h in CDCl3 in the NMR tube, the NMR spectrum of the
resulting mixture indicated that the ratio of two isomers α-17/β-17 changed to ca. 39/1 by
1
decomposition of α-17 and β-17 to starting materials 8b and 16. Hꢀ and 13CꢀNMR spectra were
13
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 28
1
2
3
depicted in Supporing information Figures [(SI)ꢀFigures 1ꢀ1 and 1ꢀ2]. 13C NMR data for the major
4
5
isomer α-18 extracted from the spectrum of the crude mixture in CDCl3 was summarized in Table 1.
6
7
8
NMR Monitoring Experiment.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
S-Alkylation of 8b with 16 in HFIP-d2 at 30 °C. A mixture of 8b (93.6 mg, 0.22 mmol), 16 (50
mmol, 0.17 mmol) and HFIPꢀd2 (0.1 mL) was kept on standing at 30 °C in a test tube with tight cap.
The reaction mixture, which was picked up (5 ꢀL) at certain intervals (12 h, 24 h, 36 h, 42 h, 48 h),
1
was diluted with HFIPꢀd2 (0.6 mL), and H NMR spectrum of the resulting solution was immediately
1
measured by a 500 MHz NMR spectrometer. Time interval HꢀNMR spectra were depicted in SIꢀ
Figures 2 and 3.
Thermal Isomerisation of Sulfonium 1,1,1,3,3,3-Hexafluoroisopropoxide-d1 (17) in HFIP-d2. In
a sealed NMR test tube, a ca. 9/1 mixture of α-17 and β-17, which was obtained by the
aforementioned Sꢀalkylation of 8b with 16 (after 48 h at 30 °C in HFIPꢀd2), was heated at 55–60 °C,
1
and time interval H NMR spectra (1 h, 3 h, 5 h, 7 h, 9 h, 13 h) were measured by a 500 MHz NMR
spectrometer. Time interval 1HꢀNMR spectra were depicted in SIꢀFigures 4 and 5.
Thermal isomerisation of Sulfonium Cloride (18) in HFIP-d2. In a sealed NMR test tube, a ca. 9/1
mixture of 1,4ꢀdideoxyꢀ1,4ꢀ[(R)ꢀ(6ꢀdeoxyꢀ3ꢀOꢀbenzylꢀ1,2ꢀOꢀisopropylideneꢀαꢀ
D
ꢀglucofuranosꢀ6ꢀ
yl)episulfoniumylidene]ꢀ2,3,5ꢀtriꢀOꢀbenzylꢀ ꢀarabinitol chloride (α-18 and β-18, 55 mg) and HFIPꢀd2
D
(0.6 mL) was heated at 55–60 °C, and time interval 1H NMR spectra (2 h, 4 h, 5 h, 6 h, 8h, 11 h, 14 h)
were measured by a 500 MHz NMR spectrometer. Time interval 1HꢀNMR spectra were depicted in SIꢀ
Figures 6 and 7.
S-Alkylation of 8b with 16 in HFIP-d2 at 60 °C. In a sealed NMR test tube, a mixture of 8b (46.8
mg, 0.11 mmol), 16 (25 mg, 0.085 mmol) and HFIPꢀd2 (0.6 mL) was heated at 55–60 °C, and time
14
ACS Paragon Plus Environment

Page 15 of 28
The Journal of Organic Chemistry
1
2
3
4
5
interval 1H NMR spectra (0.5 h, 1 h, 2 h, 3 h, 5 h, 9 h, 12 h, 15 h, 18 h, 21 h, 24 h, 27 h, 33 h, 42 h, 65
1
h) were measured by a 500 MHz NMR spectrometer. Time interval HꢀNMR spectra were depicted in
6
7
SIꢀFigures 8 and 9.
8
9
Synthesis of Sulfonium Chloride (18) through S-Alkylation of Thiosugar (8b, 5 eq) with
Epoxide (16) in HFIP. A mixture of thiosugar (8b, 1.44 g, 3.43 mmol), epoxide (16, 200 mg, 0.68
mmol), and HFIP (5 mL) was heated under reflux for 6 h. After being cooled with cold water, the
reaction mixture was acidified with 5% methanolic hydrogen chloride to pH 3. The mixture was
immediately neutralized with NaHCO3 with ice cooling. The resulting suspension was filtered by
suction, and the filter cake was washed with CH2Cl2. The combined filtrate and washings were
condensed in vacuo to give a colorless oil (1.85 g), which on column chromatography (CHCl3/MeOH
= 100/1→30/1→10/1) gave α-18 (391 mg, 76%), a ca. 2.5:1 mixture of α-18 and β-18 (67 mg, 13%),
and thiosugar (8b, 1.1 g).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Major isomer
α
-18: colorless amorphous. [α]24 –38.2 (c = 1.17, CHCl3). IR (neat): 3379, 1076,
D
1
1115 cmꢀ1. HꢀNMR (500 MHz, CDCl3)
δ: 1.29/1.47 [each 3H, s, C(CH3)2], 3.74 (2H, dꢀlike, J = ca.
8.0, Hꢀ5a and Hꢀ5b), 4.04 (1H, dd, J = 12.6, 3.7, Hꢀ1’a), 4.08 (1H, dd, J = 12.6, 8.0 Hz, Hꢀ1’b), 4.17
(1H, br tꢀlike, J = ca. 8.0, Hꢀ4), 4.19 (1H, br sꢀlike, Hꢀ3), 4.20 (1H, d, J = 3.2 Hz, Hꢀ4’), 4.21 (1H, ddꢀ
like, J = ca. 13.0, 3.5, Hꢀ1a), 4.33 (1H, dd, J = 8.3, 3.2, Hꢀ3’), 4.39 (1H, dd d, J = 13.0, 1.8, Hꢀ1b),
4.43/4.55 (each 1H, d, J = 11.7, CH2Ph), 4.45 (1H, br sꢀlike, Hꢀ2), 4.46/4.58 (each 1H, d, J = 12.0,
CH2Ph), 4.56 (1H, d, J = 3.7, Hꢀ5’), 4.58 (2H, s, CH2Ph), 4.62/4.75 (each 1H, d, J = 11.5, CH2Ph),
4.63 (1H, dddꢀlike, J = ca. 8.3, 8.0, 3.7, Hꢀ2’), 5.66 (1H, br s, OH), 5.84 (1H, d, J = 3.7, Hꢀ6’), 7.13–
13
7.42 (20H, m, arom.). CꢀNMR (125 MHz, CHCl3)
δ: 26.2/26.8 [C(CH3)2], 48.3 (Cꢀ1), 52.1 (Cꢀ1’),
64.9 (Cꢀ2’), 66.5 (Cꢀ4), 66.8 (Cꢀ5), 71.8/72.3/73.1/73.5 (CH2Ph), 81.0 (Cꢀ4’), 82.1 (Cꢀ3’), 82.2 (Cꢀ3),
82.3
(Cꢀ2),
82.8
(Cꢀ5’),
105.0
(Cꢀ6’),
112.1
(d,
[C(CH3)2],
arom.),
15
127.79/127.88/127.93/128.08/128.14/128.2/128.4/128.5/128.6/128.7
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 28
1
2
3
135.7/135.8/136.6/137.4 (s, arom.). LRMS (FAB) m/z: 713 [M–Cl]+. HRMS (FAB) m/z: [M–Cl]+
4
5
Calcd for C42H49O8S 713.3148; Found 713.3123.
6
7
8
9
Minor isomer
β
-18 (NMR data extracted from the spectrum of a ca. 2.5:1 mixture): 1HꢀNMR
: 1.28/1.47 [each 3H, s, C(CH3)2], 3.72–3.80 (2H, m, Hꢀ1a and Hꢀ5a), 3.76–3.88
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(500 MHz, CDCl3)
δ
(2H, m, Hꢀ1a’a and Hꢀ5a), 4.13–4.24 (4H, m, Hꢀ1’b, Hꢀ3, Hꢀ4 and Hꢀ4’), 4.29 (1H, dd, J = 7.2, 2.9, Hꢀ
3’), 4.37–4.39 (1H, m, Hꢀ2), 4.47–4.49 (1H, m, Hꢀ5’), 4.70–4.63 (1H, m, Hꢀ2’), 4.76–4.79 (1H, m, Hꢀ
1b), 5.71 (1H, d, J = 3.7, Hꢀ6’), 4.37–4.64 (8H, m, CH2Ph), 7.14–7.43 (20H, m, arom). 13CꢀNMR (125
MHz, CHCl3) δ: 26.2/29.6 [C(CH3)2], 43.9 (Cꢀ1’), 45.4 (Cꢀ1), 61.0 (Cꢀ4), 63.3 (Cꢀ2’), 65.2 (Cꢀ5),
72.3(2C)/73.1/73.4 (CH2Ph), 81.3 (Cꢀ4’), 82.0 (Cꢀ3’), 82.6 (Cꢀ2), 82.7 (Cꢀ5’), 84.6 (Cꢀ3), 105.9 (Cꢀ
6’), 112.0 [(CH3)2C], 127.9/128.4/128.6/128.74/other signals due to aromatic methine carbons
overlapped with those of α-18 (d, arom.), 135.8/136.0/136.3/137.5 (s, arom.).
Synthesis of Sulfonium Tetrafluoroborate (19) through S-Alkylation of Thiosugar (8b)
with Epoxide (16) in the presence of HBF4•Et2O. To a mixture of 8b (780 mg, 1.86 mmol), 16
(652 mg, 2.23 mmol) and CH2Cl2 (22 mL) was added tetrafluoroboric acid diethyl ether complex
(HBF4•Et2O, 331 ꢀL, 2.4 mmol) at –60 °C, and the mixture was stirred at –60 °C for 30 min. After the
reaction was quenched with NaHCO3, the resulting suspension was filtered by suction, and the filter
cake was washed with CH2Cl2. The combined filtrate and washings were condensed in vacuo to give a
colorless oil (1.53 g), which on column chromatography (CHCl3/MeOH = 100/1→30/1→10/1) gave
8b (1.1 g) and a mixture of 1,4ꢀdideoxyꢀ1,4ꢀ{(R)ꢀ[6ꢀdeoxyꢀ3ꢀOꢀbenzylꢀ1,2ꢀOꢀisopropylideneꢀαꢀ
glucofuranoseꢀ6ꢀyl]episulfoꢀniumylidene}ꢀ2,3,5ꢀtriꢀOꢀbenzylꢀ ꢀarabinitol tetrafluoroborate (α-19, 650
mg, 44%) and a ca. 3:1 mixture of α-19 and β-19 (617 mg, 41%).
Dꢀ
D
Major isomer
1. 1HꢀNMR (500 MHz, CDCl3)
α
-19: colorless amorphous. [α]23D –17.5 (c = 0.8, CHCl3). IR (neat): 3480, 1072 cmꢀ
δ: 1.29/1.47 [each 3H, s, C(CH3)2], 3.66 (1H, dd, J = 10.0, 10.0, Hꢀ5a),
16
ACS Paragon Plus Environment

Page 17 of 28
1
The Journal of Organic Chemistry
2
3
4
5
6
7
8
9
3.73 (1H, dd, J = 10.0, 6.0, Hꢀ5b), 3.76 (1H, ddꢀlike, J = ca. 13.5, 3.5, Hꢀ1a), 3.77 (1H, dd, J = 13.2,
4.0, Hꢀ1’a), 3.83 (1H, dd, J = 13.2, 3.5, Hꢀ1’b), 4.00 (1H, br dd, J = 10.0, 6.0, Hꢀ4), 4.12 (1H, br d, J =
ca. 13.5, Hꢀ1b), 4.13 (1H, d, J = 3.2, Hꢀ4’), 4.18 (1H, dd, J = 7.7, 3.2, Hꢀ3’), 4.20 (1H, br sꢀlike, Hꢀ3),
4.39/4.51 (each 1H, d, J = 11.7, CH2Ph), 4.45/4.52 (each 1H, d, J = 12.0, CH2Ph), 4.47 (1H, br sꢀlike,
Hꢀ2), 4.50 (1H, dd, J = 7.7, 4.0, Hꢀ2’), 4.54/4.56 (each 1H, dꢀlike, J = 12.0, OCH2Ph), 4.57 (1H, d, J =
3.7, Hꢀ5’), 4.60 (2H, s, CH2Ph), 5.84 (1H, d, J = 3.7, Hꢀ6’), 7.13–7.36 (20H, m, arom.). 13CꢀNMR
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(125 MHz, CHCl3) δ: 26.2/26.8 [C(CH3)2], 48.0 (Cꢀ1), 51.3 (Cꢀ1’), 65.4 (Cꢀ2’), 66.6 (Cꢀ5), 67.1 (Cꢀ4),
71.9/72.4/72.7/73.6 (CH2Ph), 81.0 (Cꢀ4’), 81.9 (Cꢀ3’), 82.2 (Cꢀ3), 82.36 (Cꢀ5’), 82.41 (Cꢀ2), 105.1 (Cꢀ
6’), 112.3 [C(CH3)2], 127.95/128.06/128.10/128.2/128.3/128.4/128.55/128.58/128.7/128.8 (d, arom.),
135.8/135.9/136.6/137.1 (s, arom.). LRMS (FAB) m/z: 713 [M–Cl]+. HRMS (FAB) m/z: [M–Cl]+
Calcd for C42H49O8S 713.3148; Found 713.3139.
Minor isomer
β
-19 (NMR data extracted from the spectrum of a ca. 3:1 mixture): 1HꢀNMR
: 1.29/1.47 [each 3H, s, C(CH3)2], 3.70–3.85 (4H, m, Hꢀ1a, Hꢀ1’a, Hꢀ5a, and Hꢀ
(500 MHz, CDCl3)
δ
5b), 3.94 (1H, dd, J = 12.9, 3.4 Hz, Hꢀ1’b), 4.00–4.04 (1H, m, Hꢀ1b), 4.13–4.23 (4H, m, Hꢀ3, Hꢀ3’, Hꢀ
4, and Hꢀ4’), 4.39–4.41 (1H, m, Hꢀ2), 4.52–4.53 (1H, m, Hꢀ5’), 4.60–4.64 (1H, m, Hꢀ2’), 5.74 (1H, d,
J = 3.5 Hz, Hꢀ6’), 4.37–4.64 (8H, m, CH2Ph), 7.13–7.36 (20H, m, arom). 13CꢀNMR (125 MHz,
CHCl3)
δ: 26.2/26.8 [C(CH3)2], 43.2 (Cꢀ1’), 44.9 (Cꢀ1), 61.7 (Cꢀ4), 64.2 (Cꢀ2’), 64.9 (Cꢀ5),
72.26/72.33/72.8/73.4 (CH2Ph), 81.1 (Cꢀ4’), 81.9 (Cꢀ3’), 82.3 (Cꢀ2), 82.7 (Cꢀ5’), 84.1 (Cꢀ3), 105.2 (Cꢀ
6’), 112.1 [(CH3)2C], 128.2/128.6/other signals due to aromatic methine carbons overlapped with those
of α-19 (d, arom.), 135.9/136.2/137.0/ a signal due to aromatic quaternary carbon overlapped with
those of α-19 (s, arom.).
Ion Exchange Reaction of Tetrafluoroborate (α-19) to Chloride (α-18). A mixture of α-19
(625 mg, 0.78 mmol), ion exchange resin [IRAꢀ400J (Cl– form), 15 g], and MeOH (15 mL) was stirred
17
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 28
1
2
3
4
5
6
7
at room temperature for 2 h. The reaction mixture was filtered by suction, and the resins were washed
with MeOH. The combined filtrate and washings were condensed in vacuo to give a colorless
amorphous (590 mg), which on column chromatography (CHCl3/MeOH = 30/1→10/1) gave α-18
8
9
1
13
(560 mg, 96%) as a colorless amorphous. The H and C NMR data agreed well with those obtained
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
by the Sꢀalkylation of 8b with 16 in HFIP.
1,4-Dideoxy-1,4-{(R)-[(2S,3S,4R,5S)-2,3,4,5,6-pentahydroxyhexyl]episulfoniumylidene}-
D
-arabinitol Chloride [neoponkoranol (7)]. A suspension of 10% PdꢀC (500 mg) in 80%
aqueous AcOH (4 mL) was preꢀequilibrated with hydrogen. To the suspension was added a mixture of
a solution of α-18 (532 mg, 0.71 mmol) in 80% aqueous AcOH (6 mL). The resulted mixture was
hydrogenated at 50°C under atmospheric pressure until uptake of hydrogen ceased. The catalyst was
filtered off and washed with water. The combined filtrate and the washings were condensed in vacuo
to give a ca. 1:1 amomeric mixture of 1,4ꢀdideoxyꢀ1,4ꢀ[(S)ꢀ(6ꢀdeoxyꢀ1ꢀ
Dꢀglucopyranosꢀ6ꢀ
yl)episulfoniumylidene]ꢀ2,3,5ꢀtriꢀOꢀbenzylꢀ ꢀarabinitol chloride containing a small amount of their
D
partially acetylated compounds as a pale yellow amorphous. The amorphous was then treated with 5%
hydrochloric acid (5 mL) at 50°C for 4 h, and the reaction mixture was condensed in vacuo to give a
pale yellow amorphous (214 mg), which was used for the next NaBH4 reduction without further
purification.
To a solution of the amorphous (214 mg) in water (5 mL) was added NaBH4 (130 mg, 3.4 mmol) at 0
°C, and the mixture was stirred at 0 °C for 5 h. The reaction mixture was acidified with 1 M HCl to pH
4 at 0 °C. After the reaction mixture was condensed in vacuo, the residue was triturated with methanol.
The MeOH insoluble material was filtered off, and washed with methanol. The combined filtrate and
washings were condensed in vacuo to give a colorless solid (278 mg), which on column
chromatography (CHCl3/MeOH = 5/1→3/1→3/2) gave a colorless solid (220 mg). Reꢀpurification by
18
ACS Paragon Plus Environment

Page 19 of 28
The Journal of Organic Chemistry
1
2
3
4
5
HPLC (H2O) gave the title compound 7 (129 mg, 52% from α-18) as colorless solid. 1H and 13C NMR
data of 7 agreed well with those reported.6
6
7
8
S-Alkylation of Thiosugar (8a) with Epoxides (9b, 9e, 9h, and 9k) in HFIP.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1,4-Dideoxy-1,4-{(R)-[4-deoxy-1-O-(p-methoxybenzyl)-2-O-(o-methylbenzyl)-
D
-erythritol-4-
-10b).
yl]episulfoniumylidene}-2,3,5-tri-O-(p-methoxybenzyl)- -arabinitol Tetrafluoroborate (
D
α
According to the method used for the Sꢀalkylation of 8b with 16 in HFIP, a mixture of 8a (10.2 g, 20
mmol), 9b9b (1.31 g 4.0 mmol), and HFIP (20 mL) was heated under reflux for 4 h. After being cooled
with cold water, the reaction was quenched with HBF4•Et2O (660 ꢀL, 4.8 mmol), and the resulting
mixture was immediately neutralized with NaHCO3 with ice cooling. The resulting suspension was
filtered by suction, and the filter cake was washed with CH2Cl2. The combined filtrate and washings
were condensed in vacuo to give a colorless oil (12.5 g), which on column chromatography
(CHCl3→CHCl3/MeOH=100/1→30/1→10/1) gave α-10b (2.44 g, 66%) and a ca. 3:1 mixture of α-10
and β-10 (737 mg, 20%).
Compound
1. 1H NMR (500 MHz, CDCl3)
α
-10b: colorless oil. [α]25D –8.96 (c = 1.15, CHCl3). IR (neat): 3499, 1612, 1080 cmꢀ
δ: 2.30 (3H, s, C6H4CH3), 3.58 (1H, dd, J = 13.2, 2.0, Hꢀ1a), 3.60 (1H,
dd, J = 10.9, 3.5, Hꢀ4’a), 3.61 (1H, dd, J = 10.3, 8.0, Hꢀ5a), 3.64 (1H, d, J = 13.2, 3.8, Hꢀ1b), 3.66
(1H, dd, J = 10.3, 6.9, Hꢀ5b), 3.70 (1H, dd, J = 10.9, 4.1, Hꢀ4’b), 3.73 (1H, dd, J = 14.0, 7.5, Hꢀ1’a),
3.74–3.77 (1H, m, Hꢀ3’), 3.77/3.78/3.79/3.82 (each 3H, s, OCH3), 3.82 (1H, dd, J = 14.0, 3.8, Hꢀ1’b),
3.97 (1H, br ddꢀlike, J = ca. 8.0, 6.9, Hꢀ4), 4.11 (1H, br ddꢀlike, J = ca. 2.0, 1.2, Hꢀ3), 4.23 (1H, dddꢀ
like, J = ca. 3.8, 2.0, 2.0, Hꢀ2), 4.21/4.28 (each 1H, d, J = 11.5 Hz, OCH2Ar), 4.23 (1H, br ddꢀlike, J =
ca. 7.5, 3.8, Hꢀ2’), 4.36–4.47 (6H, m, OCH2Ar), 4.57/4.66 (each 1H, d, J = 11.2, OCH2Ar), 6.81–6.88
(8H, m, arom.), 7.02–7.27 (12H, m, arom.). 13C NMR (125 MHz, CDCl3)
1), 51.1 (Cꢀ1’), 55.5 (OCH3), 66.2 (Cꢀ4), 66.4 (Cꢀ5), 68.4 (Cꢀ4’), 68.7 (Cꢀ2’), 71.2/71.4/71.7/73.1/73.3
19
δ: 18.8 (C6H4CH3), 48.1 (Cꢀ
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 28
1
2
3
(OCH2Ar),
79.6
(Cꢀ3’),
82.0
(Cꢀ3),
82.1
(Cꢀ2),
4
5
113.85/113.93/114.0/114.1/125.9/128.3/129.4/129.6/130.0/130.3 (d, arom.), 127.9/128.0/128.9/
6
7
135.6/137.1/159.3/ 159.5/159.7/159.8 (s, arom.). FABMS (pos.) m/z: 839 [M–BF4]+.
8
9
1,4-Dideoxy-1,4-{(R)-[4-deoxy-1-O-(p-methoxybenzyl)-2-O-(o-chlorobenzyl)-
D
-erythritol-4-
-10e).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
yl]episulfoniumylidene}-2,3,5-tri-O-(p-methoxybenzyl)- -arabinitol Tetrafluoroborate (
D
α
In a similar manner described above, a mixture of 8a (7.93 g, 15.5 mmol), 9e9b (1.08 g, 3.1 mmol), and
HFIP (20 mL) was heated under reflux for 4 h. Workꢀup gave a colorless oil (10.0 g), which on
column chromatography gave α-10e (2.02 g, 69%) and a ca. 2:1 mixture of α-10e and β-10e (463 mg,
16%).
Compound
1. 1H NMR (500 MHz, CDCl3)
α
-10e: colorless oil. [α]23D –3.57 (c = 1.26, CHCl3). IR (neat): 3499, 1612, 1072 cmꢀ
δ: 3.636 (1H, dd, J = 10.3, 8.9, Hꢀ5a), 3.644 (1H, dd, J = 10.3, 3.8, Hꢀ
4’a), 3.68 (1H, dd, J = 10.3, 6.9, Hꢀ5b), 3.71 (1H, dd, J = 13.0, 3.0, Hꢀ1a), 3.72 (1H, dd, J = 10.3, 3.8,
Hꢀ4’b), 3.74 (1H, dddꢀlike, J = ca. 4.5, 3.8, 3.8, Hꢀ3’), 3.77 (3H, s, OCH3), 3.78 (1H, ddꢀlike, J = 13.2,
7.0, Hꢀ1’a), 3.79 (6H, s, OCH3), 3.81 (1H, ddꢀlike, J = ca. 13.0, 4.0, Hꢀ1b), 3.81 (3H, s, OCH3), 3.87
(1H, dd, J = 13.2, 3.4, Hꢀ1’b), 4.00 (1H, br ddꢀlike, J= ca. 8.9, 6.9, Hꢀ4), 4.12 (1H, br ddꢀlike, J = ca.
2.0, 1.2, Hꢀ3), 4.27/4.35 (each 1H, d, J = 11.5, OCH2Ar), 4.30 (1H, dddꢀlike, J = ca. 4.0, 3.0, 2.0, Hꢀ
2), 4.30–4.35 (1H, m, Hꢀ2’), 4.38–4.48 (6H, m, OCH2Ar), 4.66/4.73 (each 1H, d, J = 11.7, OCH2Ar),
6.81–6.88 (8H, m, arom.), 7.03–7.43 (12H, m, arom.). 13C NMR (125 MHz, CDCl3)
50.8 (Cꢀ1’), 55.26/55.29 (OCH3), 66.35 (Cꢀ4), 66.44 (Cꢀ5), 68.3 (Cꢀ4’), 68.6 (Cꢀ2’),
δ: 48.1 (Cꢀ1),
70.0/71.5/71.8/73.2/73.3
(OCH2Ar),
79.7
(Cꢀ3’),
81.9
(Cꢀ3),
82.1
(d,
(Cꢀ2),
113.9/114.0/114.08/114.13/127.1/129.4/129.5/129.6/129.65/129.70/130.0/130.7
arom.),
127.8/128.0/128.8/133.6/135.1/159.3/159.6/159.77/159.80 (s, arom.). FABMS (pos.) m/z: 859 [M–
BF4]+.
20
ACS Paragon Plus Environment

Page 21 of 28
1
The Journal of Organic Chemistry
2
3
1,4-Dideoxy-1,4-{(R)-[4-deoxy-1-O-(p-methoxybenzyl)-2-O-(o-trifluorobenzyl)-
D
-erythritol-4-
4
5
yl]episulfoniumylidene}-2,3,5-tri-O-(p-methoxybenzyl)-D-arabinitol Tetrafluoroborate (α-10h).
6
7
8
9
In a similar manner described above, a mixture of 8a (8.65 g, 17.0 mmol), 9h9b (1.30 g, 3.4 mmol),
and HFIP (20 mL) was heated under reflux for 4 h. Workꢀup gave a colorless oil (10.0 g), which on
column chromatography gave α-10h (2.47 g, 74%) and a ca. 2:1 mixture of α-10h and β-10h (462 mg,
14%).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compound
1H NMR (500 MHz, CDCl3)
α
-10h: colorless oil. [α]24D –8.1 (c = 0.98, CHCl3). IR (neat): 3406, 1612, 1083 cmꢀ1.
δ: 3.62 (1H, dd, J = 10.9, 3.7, Hꢀ4’a), 3.64 (1H, dd, J = 10.6, 9.0, Hꢀ5a),
3.69 (1H, dd, J = 10.6, 6.9, Hꢀ5b), 3.71 (1H, dd, J = 10.9, 4.5, Hꢀ4’b), 3.72 (1H, dd, J = 13.0, 3.8, Hꢀ
1a), 3.75–3.83 (3H, m, Hꢀ1b, Hꢀ1’a and Hꢀ3’), 3.85 (1H, dd, J = 13.2, 3.7, Hꢀ1’b), 3.76/3.77/3.78/3.80
(each 3H, s, OCH3), 4.01 (1H, br ddꢀlike, J = ca. 9.0, 6.9, Hꢀ4), 4.13 (1H, br sꢀlike, Hꢀ3), 4.26/4.34
(each 1H, d, J = 11.5, OCH2Ar), 4.30 (1H, br m, Hꢀ2), 4.32–4.34 (1H, m, Hꢀ2’), 4.37–4.47 (6H, m,
OCH2Ar), 4.74/4.82 (each 1H, d, J = 12.4, OCH2Ar), 6.80–7.63 (20H, m, arom.). 13C NMR (125
MHz, CDCl3) δ: 48.1 (Cꢀ1), 50.7 (Cꢀ1’), 55.2/55.3 (OCH3), 66.3 (Cꢀ4), 66.4 (Cꢀ5), 68.2 (Cꢀ4’), 68.5
(Cꢀ2’), 68.8/71.5/71.7/73.1/73.2 (OCH2Ar), 80.2 (Cꢀ3’), 81.9 (Cꢀ3), 82.1 (Cꢀ2),
113.8/113.9/114.0/114.1/127.9/129.57/129.61/ 129.7/129.9/130.7/132.2 (d, arom.), 124.3 [q, J = 272
Hz, CF3,], 125.8 [q, J = 5.0 Hz, CorthoꢀCF3,], 127.7 [q, J = 31.0 Hz, CipsoꢀCF3,],
128.1/128.8/130.6/132.2/135.9/159.3/159.5/159.71/159.74 (s, arom.). FABMS (pos.) m/z: 893 [M–
BF4]+.
1,4-Dideoxy-1,4-{(R)-[4-deoxy-1-O-(p-methoxybenzyl)-2-O-(o-nitrobenzyl)-
D
-erythritol-4-
-10k).
yl]episulfoniumylidene}-2,3,5-tri-O-(p-methoxybenzyl)- -arabinitol Tetrafluoroborate (
D
α
In a similar manner described above, a mixture of 8a (6.83 g, 13.4 mmol), 9kb (1.30 g, 2.7 mmol), and
HFIP (20 mL) was heated under reflux for 4 h. Workꢀup gave a colorless oil (10.0 g), which on
21
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 28
1
2
3
4
5
column chromatography gave α-10k (1.75 g, 67%) and a ca. 1:1.2 mixture of α-10k and β-10k (220
mg, 9%). 1H and 13C NMR data of α-10k agreed well with those reported.9b
6
7
8
Synthesis of Sulfonium Salts (7b, 7e, 7h, and 7k).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1,4-Dideoxy-1,4-{(R)-[4-deoxy-2-O-(o-methylbenzyl)-
D
-erythritol-4-yl]episulfoniumylidene}- -
D
arabinitol Chloride (7b). A mixture of α-10b (2.43 g, 2.6 mmol), 80% aqueous TFA (30 mL), and
CHCl3 (15 mL) was stirred at room temperature for 5 h. After the reaction mixture was condensed in
vacuo, the residue was triturated with methanol. The MeOH insoluble material was filtered off, and
washed with methanol. The combined filtrate and washings were condensed in vacuo to give a
colorless oil (1.19 g), which was subsequently stirred with ion exchange resin IRAꢀ400J (50 g) in
methanol (100 mL) for 12 h. The resin was filtered off, and washed with methanol. The combined
filtrate and washings were condensed in vacuo to give colorless oil (888 mg), which on column
chromatography (CHCl3/MeOH = 10/1→3/1) gave 7b (817 mg, 79%) as colorless oil: [α]24D +11.9 (c
= 1.40, CH3OH). IR (neat): 3302, 1605, 1076 cmꢀ1. 1H NMR (500 MHz, CD3OD)
δ: 2.37 (3H, s, CH3),
3.56 (1H, dt, J = 5.5, 4.3, Hꢀ3’), 3.67 (1H, dd, J = 13.2, 8.9, Hꢀ1’a), 3.69 (1H, br dꢀlike, J = ca. 12.6,
Hꢀ1a), 3.72 (1H, dd, J = 11.7, 4.3 Hz, Hꢀ4’a), 3.75 (1H, dd, J = 12.6, 3.4, Hꢀ1b), 3.79 (1H, dd, J =
13.2, 3.5, Hꢀ1’b), 3.84 (1H, dd, J = 11.7, 4.3, Hꢀ4’b), 3.89 (1H, dd, J = 10.1, 9.8, Hꢀ5a), 3.93 (1H, br
ddꢀlike, J = ca. 9.8, 3.8, Hꢀ4), 4.02 (1H, dd, J = 10.1, 3.8, Hꢀ5b), 4.25 (1H, ddd, J = 8.9, 5.5, 3.5, Hꢀ
2’), 4.34 (1H, br dꢀlike, J = ca. 1.5, Hꢀ3), 4.57 (1H, br ddꢀlike, J = ca. 3.4, 1.5, Hꢀ2), 4.63/4.82 (each
1H, d, J = 11.5, OCH2Ar), 7.13–7.22 (3H, m, arom.), 7.35 (1H, d, J = 7.2, arom.). 13C NMR (125
MHz, CD3OD) δ: 19.1 (CH3), 51.8 (Cꢀ1’), 52.2 (Cꢀ1), 60.9 (Cꢀ4’), 61.0 (Cꢀ5), 68.8 (Cꢀ2’), 71.9
(OCH2Ar), 73.7 (Cꢀ4), 79.4 (Cꢀ2), 79.6 (Cꢀ3), 83.0 (Cꢀ3’), 126.9/129.3/130.4/131.3 (d, arom.),
137.2/138.2 (s, arom.). LRMS (FAB) m/z: 359 [M–Cl]+. HRMS (FAB) m/z: [M–Cl]+ Calcd for
C17H27O6S 359.1528; Found 359.1560.
22
ACS Paragon Plus Environment

Page 23 of 28
1
The Journal of Organic Chemistry
2
3
1,4-Dideoxy-1,4-{(R)-[4-deoxy-2-O-(o-chlorobenzyl)-D-erythritol-4-yl]episulfoniumylidene}-D-
4
5
arabinitol Chloride (7e). In a similar manner described above, from α-20e (2.01 g, 2.12 mmol) was
6
7
8
9
preparerd 7e (704 mg, 80%) as colorless oil: [α]21 +12.6 (c = 1.00, CH3OH). IR (neat): 3302, 1597,
D
1
1084 cmꢀ1. H NMR (500 MHz, CD3OD)
δ: 3.58 (1H, dt, J = 5.8, 4.0, Hꢀ3’), 3.74 (1H, dd, J = 13.2,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9.2, Hꢀ1’a), 3.75 (1H, dd, J = 12.0, 4.0, Hꢀ4’a), 3.78 (1H, dd, J = 12.9, 2.0, Hꢀ1a), 3.81 (1H, dd, J =
12.9, 3.2, Hꢀ1b), 3.87 (1H, dd, J = 13.2, 3.5, Hꢀ1’b), 3.88 (1H, dd, J = 12.0, 4.0, Hꢀ4’b), 3.90 (1H, dd,
J = 10.9, 9.2 Hz, Hꢀ5a), 3.97 (1H, br ddꢀlike, J = 9.2, 4.9, Hꢀ4), 4.03 (1H, dd, J = 10.9, 4.9, Hꢀ5b),
4.28 (1H, ddd, J = 9.2, 5.8, 3.5, Hꢀ2’), 4.36 (1H, br dꢀlike, J = ca. 1.2, Hꢀ3), 4.60 (1H, dddꢀlike, J = ca.
3.2, 2.0, 1.2, Hꢀ2), 4.75/4.85 (each 1H, d, J = 12.3, OCH2Ar), 7.29 (1H, td, J = 7.5, 2.3, arom.), 7.31
(1H, td, J = 7.5, 2.3, arom.), 7.39 (1H, ddꢀlike, J = 7.5, 2.3, arom.), 7.57 (1H, ddꢀlike, J = 7.5, 2.3,
arom.). 13C NMR (125 MHz, CD3OD)
δ: 51.9 (Cꢀ1’), 52.2 (Cꢀ1), 60.7 (Cꢀ4’), 61.0 (Cꢀ5), 68.7 (Cꢀ2’),
70.7 (OCH2Ar), 73.8 (Cꢀ4), 79.46 (Cꢀ2), 79.53 (Cꢀ3), 83.5 (Cꢀ3’), 128.2/130.4/130.5/131.4 (d, arom.),
134.4/137.0 (s, arom.). LRMS (FAB) m/z: 379 [M–Cl]+. HRMS (FAB) m/z: [M–Cl]+ Calcd for
C16H24ClO6S 379.0982; Found 379.0989.
1,4-Dideoxy-1,4-{(R)-[4-deoxy-2-O-(o-trifluorobenzyl)-D-erythritol-4-yl]episulfoniumylidene}-D-
arabinitol Chloride (7h). In a similar manner described above, from α-20h (2.46 g, 2.5 mmol) was
24
preparerd 7h (918 mg, 82%) as colorless oil: [α]D +12.3 (c = 2.0, CH3OH). IR (neat): 3286, 1612,
1
1115 cmꢀ1. H NMR (800 MHz, CD3OD)
δ: 3.62 (1H, ddd, J = 4.8, 4.0, 4.0, Hꢀ3’), 3.74 (1H, dd, J =
12.0, 4.0, Hꢀ4’a), 3.77 (1H, dd, J = 13.6, 8.8 Hꢀ1’a), 3.79 (1H, dd, J = 12.8, 1.6, Hꢀ1a), 3.81 (1H, dd, J
= 12.8, 3.2, Hꢀ1b), 3.85 (1H, dd, J = 12.0, 4.8, Hꢀ4’b), 3.87 (1H, dd, J = 13.6, 3.2, Hꢀ1’b), 3.91 (1H,
dd, J = 11.2, 9.6, Hꢀ5a), 3.98 (1H, br dd, J = ca. 9.6, 4.8, Hꢀ4), 4.03 (1H, dd, J = 11.2, 4.8, Hꢀ5b), 4.30
(1H, ddd, J = 8.8, 4.0, 3.2, Hꢀ2’), 4.36 (1H, ddꢀlike, J = ca. 2.0, 1.6, Hꢀ3), 4.59 (1H, dddꢀlike, J = 3.2,
2.0, 1.6, Hꢀ2), 4.86/4.96 (each 1H, d, J = 12.8, OCH2Ar), 7.46/7.63 (each 1H, br t, J = ca. 8.0, arom.),
7.67/7.82 (each 1H, br d, J = ca. 8.0, arom.). 13C NMR (200 MHz, CD3OD)
δ: 51.6 (Cꢀ1’), 52.3 (Cꢀ1),
23
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 28
1
2
3
4
5
6
7
8
9
60.9 (Cꢀ4’), 61.0 (Cꢀ5), 68.9 (Cꢀ2’), 69.7 [br qꢀlike, J = 1.9 Hz, CH2C6H4ꢀ(oꢀCF3)], 73.7 (Cꢀ4), 79.6
(Cꢀ2), 79.6 (Cꢀ3), 84.0 (Cꢀ3’), 125.9 [q, J = 272 Hz, CF3,], 126.7 [q, J = 5.8 Hz, CorthoꢀCF3,], 128.6 [q,
J = 31.0 Hz, CipsoꢀCF3,], 129.0/131.2/133.4 (d, arom.), 138.0 (s, arom.). LRMS (FAB) m/z: 413 [M–
Cl]+. HRMS (FAB) m/z: [M–Cl]+ Calcd for C17H24O6F3S 413.1246; Found 413.1220.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1,4-Dideoxy-1,4-{(R)-[4-deoxy-2-O-(o-nitrobenzyl)-D-erythritol-4-yl]episulfoniumylidene}-D-
arabinitol chloride (7k). In a similar manner described above, from α-20k (1.69 g, 1.8 mmol) was
preparerd 7k (586 mg, 78%). 1H and 13C NMR data of 7k agreed well with those reported.9b
Bioassay.
Effects on Human Intestinal
α-Glucosidase. The experiment was performed according to the
method as described in our previous report.14 A human small intestinal microsome (batch MIC318017,
purchased from BIOPREDIC International, Rennes, France) in 0.1 M maleate buffer (pH 6.0) was
used to determine the activity of maltase, a small intestinal αꢀglucosidase. A test sample was dissolved
in dimethyl sulfoxide (DMSO) and the resulting solution was diluted with 0.1 M maleate buffer to
prepare the test sample solution (concentration of DMSO: 10%). A substrate solution in the maleate
buffer (maltose: 74 mM, 50 ꢀL), the test sample solution (25 ꢀL), and the enzyme solution (25 ꢀL)
were mixed at 37°C for 30 min and then immediately heated in boiling water for 2 min to stop the
reaction. The glucose concentrations were determined using the glucoseꢀoxidase method. The final
concentration of DMSO in the test solution was 2.5% and no influence of DMSO on the inhibitory
activity was detected. The IC50 was determined graphically (N = 2–4). The intestinal αꢀglucosidase
inhibitors (acarbose, voglibose, and miglitol) were used as reference compounds. For kinetic analysis
of the inhibitory activity on human maltase by salacinol (1), neosalacinol (4), and the 3'ꢀOꢀbenzylated
analogues (7b, 7e, 7h, and 7k), the enzyme solution (25 ꢀL) and each test sample solution (1: 0.50,
1.0, and 2.0 ꢀM; 4: 0.50, 1.0, and 2.0 ꢀM; 7b: 0.05, 0.10, and 0.20 ꢀM; 7e: 0.01, 0.04, and 0.08 ꢀM;
24
ACS Paragon Plus Environment

Page 25 of 28
The Journal of Organic Chemistry
1
2
3
7h: 0.02, 0.04, and 0.08 ꢀM; and 7k: 0.01, 0.02, and 0.04 ꢀM, 25 ꢀL) were incubated with increasing
4
5
6
7
8
9
concentrations of the substrate solution (1.5–5.3 mM, 50 ꢀL). The typical competitive inhibitors,
acarbose (1.0, 2.0, and 4.0 ꢀM, 25 ꢀL), voglibose (0.10, 0.20, and 0.40 ꢀM, 25 ꢀL), and miglitol
(0.25, 0.50, and 1.0 ꢀM, 25 ꢀL), were used as reference compounds. Lineweaver–Burk plots of the
inhibition of human intestinal maltase activities by compounds 7b, 7e, 7h, 7k, salacinol (1),
neosalacinol (4), voglibose, acarbose, and miglitol were depicted in SIꢀFigure 10.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Animals. Male ddY mice (6ꢀweeksꢀold) were purchased from Kiwa Laboratory Animal Co., Ltd.,
Wakayama, Japan. The animals were housed at a constant temperature of 23 ± 2°C, and were then fed
a standard laboratory chow (MF, Oriental Yeast Co., Ltd., Tokyo, Japan). The animals were fasted for
20–24 h prior to the beginning of the experiment, but were allowed free access to tap water. All of the
experiments were performed on conscious mice unless otherwise noted. The experimental protocol
was approved by the Experimental Animal Research Committee at Kindai University.
Effects on Blood Glucose Levels in Maltose-loaded Mice. The experiments were performed
according to the method as described in our previous reports with a slight modification.14 After fasting,
the mice were orally administrated a 10% (w/v) maltose solution (1 g/kg) with or without a test
sample. At 0, 15, 30, 60, 120, and 180 min after the administration, blood samples were taken from the
tail vein and immediately subjected to the measurement of blood glucose using the glucose oxidase
method. As a baseline, distilled water was administrated to rats as a normal group. An intestinal αꢀ
glucosidase inhibitor voglibose was used as a reference compound.
Statistics. Values are expressed as the mean ± S.E.M. Oneꢀway analysis of variance (ANOVA)
followed by Dunnett’s test was used for statistical analyses. Probability (p) values less than 0.05 were
considered significant.
25
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 28
1
2
3
4
5
ASSOCIATED CONTENT
Supporting Information
6
7
8
The Supporting Information is available free of charge on the ACS Publications website at DOI:
9
A new synthetic protocol for the known epoxide 16, time interval NMR spectra in HFIPꢀd2, 1H and 13C
NMR spectra for isolated compounds, and Lineweaver–Burk plots of the inhibition of human intestinal
maltase activities by compounds 7b, 7c, 7d, 7e, salacinol (1) (PDF).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AUTHOR INFORMATION
Corresponding Author
*Eꢀmail: gꢀtanabe@phar.kindai.ac.jp
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
This study was supported by a grant from the Houansha Foundation, 2017–2020 and by a GrantꢀinꢀAid
for scientific research from the Japan Society for the Promotion of Science [JSPS (KAKENHI),
15K08008, 15K08009, 16K08313, 17K08377 and 17H05438]. We are also thankful for the financial
support extended by the MEXTꢀSupported Program for the Strategic Research Foundation at Private
Universities, 2014–2018, the Antiaging Project for Private Universities, and Kobayashi International
Scholarship Foundation. We thank Prof. Isao Nakanishi (Kindai University) for kindly calculating the
conformational energy of reaction intermediates.
REFERENCES
(1) (a) Yoshikawa, M.; Murakami, T.; Shimada, H.; Matsuda, H.; Yamahara, J.; Tanabe G.; Muraoka,
O. Tetrahedron Lett. 1997, 38, 8367. (b) Yoshikawa, M.; Morikawa, T.; Matsuda, H.; Tanabe G.;
Muraoka, O. Bioorg. Med. Chem. 2002, 10, 1547.
26
ACS Paragon Plus Environment

Page 27 of 28
1
The Journal of Organic Chemistry
2
3
4
5
6
7
(2) Yoshikawa, M.; Murakami, T.; Tashiro K.; Matsuda, H. Chem. Pharm. Bull. 1998, 46, 1339.
(3) Yoshikawa, M.; Xu, F.; Nakamura, S.; Wang, T.; Matsuda, H.; Tanabe G.; Muraoka, O.
Heterocycles 2008, 75, 1397.
8
9
(4) Minami, Y.; Kuriyama, C.; Ikeda, K.; Kato, A.; Takebayashi, K.; Adachi, I.; Fleet, W. J. G.;
Kettawan, A.; Okamoto, T.; Asano, N. Bioorg. Med. Chem. 2008, 16, 2734.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(5) (a) Ozaki, S.; Oe, H.; Kitamura, S. J. Nat. Prod. 2008, 71, 981. (b) Muraoka, O.; Xie, W.; Tanabe,
G.; Amer, F. A. M.; Minematsu, T.; Yoshikawa, M. Tetrahedron Lett. 2008, 49, 7315.
(6) Xie, W.; Tanabe, G.; Akaki, J.; Morikawa, T.; Ninomiya, K.; Minematsu, T.; Yoshikawa, M.; Wu,
X.; Muraoka, O. Bioorg. Med. Chem. 2011, 19, 2015.
(7) (a) Williams, J. A.; Choe, Y. S.; Noss, M. J.; Baumgartner C. J.; Mustad, V. A. Am. J. Clin. Nutr.
2007, 86, 124. (b) Jayawardena, M. H. S.; de Alwis, N. M. W.; Hettigoda V.; Fernando, D. J. S. J.
Ethnopharmacol. 2005, 97, 215. (c) Kajimoto, O.; Kawamori, S.; Shimoda, H.; Kawahara, Y.; Hirata
H.; Takahashi, T. Nippon Eiyo Shokuryo Gakkaishi 2000, 53, 199. (d) Shimoda, H.; Fujimura, T.;
Makino, K.; Yoshijima, K.; Naito, K.; Ihota H.; Miwa, Y. J. Food Hyg. Soc. Jpn. 1999, 40, 198.
(8) (a) Tanabe, G.; Matsuda, Y.; Oka, M.; Kunikata, Y.; Tsutsui, N.; Xie, W.; Balakishan, G.; Amer,
M. F. A.; Marumoto, S.; Muraoka, O. J. Org. Chem. 2016, 81, 3407. (b) Jayakanthan, K.; Mohan, S.;
Pinto, B. M. J. Am. Chem. Soc. 2009, 131, 5621. (c) Johnston, B. D.; Jensen, H. H.; Pinto, B. M. J.
Org. Chem. 2006, 71, 1111. (d) Eskandari, R.; Kuntz, D. A.; Rose, D. R.; Pinto, B. M. Org Lett. 2010,
12, 1632. (e) Tanabe, G.; Xie, W.; Ogawa, A.; Cao, C.; Minematsu, T.; Yoshikawa, M.; Muraoka, O.
Bioorg. Med. Chem. Lett. 2009, 19, 2195. (f) Ghavami, A.; Sadalapure, K. S.; Johnston, B. D.; Lobera,
M.; Snider B. B.; Pinto, B. M. Synlett 2003, 1259 and references cited therein.
(9) (a) Mohan, S.; Eskandari, R. Pinto, B. M. Acc. Chem. Res. 2014, 47, 211. (b) Tanabe, G.;
Nakamura, S.; Tsutsui, N.; Balakishan, G.; Xie, W.; Tsuchiya, S.; Akaki, J.; Morikawa, T.; Ninomiya,
K.; Nakanishi, I.; Yoshikawa, M.; Muraoka, O. Chem. Commun. 2012, 48, 8646 and references cited
therein.
27
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 28
1
2
3
4
5
6
7
(10) (a) Yuasa, H.; Takada, J.; Hashimoto, H. Bioorg. Med. Chem., 2001, 11, 1137; (b) Eskandari, R.;
Jones, K.; Reddy, K. R.; Jayakanthan, K.; Chaudet, M.; Rose, D. R.; Pinto, B. M. Chem. Eur. J., 2011,
17, 14817.
8
9
(11) (a) Cimetiere, B.; Jacob, L.; Julia, M. Tetrahedron Lett. 1986, 27, 6329. (b) Cimetiere, B.; Jacob,
L.; Julia, M. Bull. Soc. Chim. Fr., 1991, 128, 926. (c) Okuma, K.; Nakamura, S.; Ohta, H. Heterocycles,
1987, 26, 2343.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(12) Brinkmann, T;. Uzar, H. C. J. Chem. Soc. Perkin Trans. 2, 2000, 527.
(13) (a) Sim, L.; Willemsma, C.; Mohan, S.; Naim, Y. H.; Pinto, B. M.; Rose, D. R. J. Biol. Chem.
2010, 285, 17763. (b) Jones, K.; Sim, L.; Mohan, S.; Kumarasamy, J.; Liu, H.; Avery, S.; Naim, H. Y.;
QuezadaꢀCalvillo, R.; Nichols, B. L.; Pinto B. M.; Rose, D. R. Bioorg. Med. Chem. 2011, 19, 3929. (c)
Ren, L.; Cao, X.; Geng, P.; Bai F.; Bai, G. Carbohydr. Res. 2011, 346, 2688.
(14) (a) Morikawa, T.; Akaki, J.; Ninomiya, K.; Kinouchi, E.; Tanabe, G.; Pongpiriyadacha, Y.;
Yoshikawa, M.; Muraoka, O. Nutrients 2015, 7, 1480. (b) Morikawa, T.; Ninomiya, K.; Akaki, J.;
Kakihara, N.; Kuramoto, H.; Matsumoto, Y.; Hayakawa, T.; Muraoka, O.; Wang, L.ꢀB.; Wu, L.ꢀJ.;
Nakamura, S.; Yoshikawa, M.; Matsuda, H. J. Nat. Med. 2015, 69, 494.
28
ACS Paragon Plus Environment