Synthesis and glycosidase inhibitory activity of some N-substituted 6-deoxy-5a-carba-β-dl- and L-galactopyranosylamines

Article abstract of DOI:10.1016/S0960-894X(03)00742-X

Chemical modification of 5a-carba-β-DL-fucopyranosylamine (3) generated six N-substituted derivatives 9a-f, among which N-octyl 9b, decyl 9c, and phenylbutyl ones 9f were found to be very strong β-galactosidase as well as β-glucosidase inhibitors. The inh

Full text of DOI:10.1016/S0960-894X(03)00742-X

Bioorganic & Medicinal Chemistry Letters 13 (2003) 3461–3463
Synthesis and Glycosidase Inhibitory Activity of Some
N-Substituted 6-Deoxy-5a-carba-ꢀ-DL- and
L-galactopyranosylamines
Seiichiro Ogawa,* Shigeo Fujieda, Yuko Sakata, Masahiro Ishizaki, Seiichi Hisamatsu
and Kensuke Okazaki
Department of Biosciences and Informatics, Faculty of Science and Technology, Keio University, Hiyoshi, Kohoku-ku,
Yokohama 223-8522, Japan
Received 14May 2003; accepted 14July 2003
Abstract—Chemical modification of 5a-carba-b-dl-fucopyranosylamine (3) generated six N-substituted derivatives 9a–f, among
which N-octyl 9b, decyl 9c, and phenylbutyl ones 9f were found to be very strong b-galactosidase as well as b-glucosidase inhibitors.
The inhibitory activity appeared attributable to d-enantiomers from biological assays of prepared l-enantiomers. Therefore,
6
-deoxy-5a-carba-b-d-galactopyranosylamine (D-3) might be a promising lead compound for further design of new carba sugar-
type b-galactosidase inhibitors.
2003 Elsevier Ltd. All rights reserved.
#
1
Recently, the N-octyl derivative 2 of 5a-carba-a-dl-
fucopyranosylamine (1) has been demonstrated to be
was performed to hopefully allow estimation of the
approximate enantiomeric contribution to the activity,
based on comparison of data observed for d- and/or l-
enantiomers. The antipode of 5a-carba-b-l-fucopyr-
anosylamine should be 6-deoxy-5a-carba-b-d-galacto-
pyranosylamine. Therefore, inhibitory activity of the
racemic modification of 5a-carbahexopyranosylamine
should be a sum of those of its components. This con-
sideration is based on the assumption that each enan-
tiomer does not interfere with the others competitive
inhibitory action. For example, with racemic dl-9c,
2
very strong a-l-fucosidase inhibitor (bovine kidney)
with
p-nitrophenyl-a-l-fucopyranoside,
essentially
comparable with deoxyfuconojirimycin (DFJ) in this
regard. Although the oxocarbenium ion transition-state
analogues 4a,b of 1 have been shown to possess a weak
activity, contrary to expectation, it is of interest to note
3
that the b-anomer 3 has high activity against a-l-fuco-
sidase. These results are in contrast with the a-glucosi-
dase inhibitory activity relationship observed between
4
a ground-state mimic of 5a-carba-a-d-glucopyrano-
sylamine, validamine, and a transition-state mimic of its
unsaturated derivative, valienamine (Fig. 1, Scheme 1).
In the present communication, chemical modification of
the b-anomer 3 was carried out by incorporation of
hydrophobic alkyl and phenylalkyl portions into the
amino group by direct treatment of a carba-sugar
epoxide 7, a newly prepared versatile precursor, with
alkyl and phenylalkyl amines. For convenience, rough
screening of the activity of several racemic compounds
*
0
Corresponding author. Tel.: +81-45-563-1141; fax: +81-45-563-
446; e-mail: ogawa@bio.keio.ac.jp
Figure 1. 5a-Carba-a and b-fucopyranosylamines and deoxyfuco-
nojirimycin (DFJ).
0
960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00742-X

3462
S. Ogawa et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3461–3463
with m-chloroperbenzoic acid gave selectively the b-
epoxide 7 (90%). Reaction of 7 with alkyl amines or
5
phenylalkylamines proceeded almost regio-selectively to
afford the respective individual N-substituted carbaglyco-
6
sylamines 8a–f with a b-galacto configuration in ꢀ85%
yield. Both selectivity of epoxidation and regioselectivity
of cleavage of the epoxide ring appeared to be improved
by incorporation of a cyclohexylidene group at C-1 and
7
-
2. The products were deprotected with aqueous acetic
acid and the amine acetates obtained were purified by
passage through an acidic resin column with 1% aqueous
8
ammonia, giving the free bases 9a–f quantitatively.
Enzyme Inhibitory Activity
Four N-alkyl and two N-phenylalkyl derivatives 9a–f
of 3 were assayed for activity against seven glycosidases:
a-galactosidase (green coffee beans), b-galactosidase
(bovine liver), a-glucosidase (Baker’s yeast), b-glucosi-
dase (almond), b-glucosaminidase (bovine kidney),
a-mannosidase (Jack beans), and a-fucosidase (bovine
kidney) in a standard manner. All compounds were
Scheme 1. For convenience, the formulas depict only one enatiomer
(L-series) of the respective racemates. Reagents and conditions: (a)
9
.
,1-dimethoxycyclohexane (4molar equiv), TsOH H O (0.2 molar
1
equiv), DMF, rt; (b) mCPBA (1.3 molar equiv), phosphate buffer,
2
shown not to be inhibitors of glycosidases other than
b-galactosidase, b-glucosidase, and a-fucosidase, as lis-
ted in Table 1. Compared to the corresponding deriva-
ꢁ
CH Cl
panol, 120 C; (d) 80% aq AcOH, 80 C; Dowex 50Wꢂ2 (H ) resin,
2
2
, 0 C; (c) a molar equiv of alkyl or phenylalkyl amine, 2-pro-
ꢁ
ꢁ
+
1
% aq NH
3
.
1
tives of the a-anomers, they possessed about one tenth
of the inhibitory activity against a-fucosidase and the
potential was thought to be mainly due to b-l-fucose-
mimicking enantiomers, as verified by assaying some
prepared l-enantiomers.
activity (IC50 2.7 mM) against l-fucosidase should be
largely due to the L-enantiomer, and high inhibition
toward b-galactosidase (IC50 0.2 mM) and b-glucosidase
(
IC50 2.0 mM) would be probably attributable to the d-
enantiomer (Table 1). This conclusion is partially sup-
ported by data obtained for newly synthesized optically
active l-9c.
It is interesting to note that all the tested compounds
possess very strong activity toward both b-galactosidase
and b-glucosiadase. Their high potential could be
demonstrated to be largely due to the respective d-enan-
tiomers, that is, N-alkyl-6-deoxy-5a-carba-b-d-galacto-
pyranosylamines. In fact, the l-enantiomers, N-alkyl-5a-
carba-b-l-fucopyranosylamines had decreased activity
toward these two enzymes as shown for 9b, 9c, and 9f in
Table 1. Compound dl-9e has been demonstrated to be
As indicated by N-substitution of the a-anomer 1, che-
mical modification of the b-anomer 3 might be expected
to increase its potential. Protection of (1RS,2RS)-6-
4
methyl-3-cyclohexene-1,2-diol (5) with a cyclohexyl-
idene group (!6, 85%) and subsequent epoxidation
a
Table 1. Inhibitory activity of some N-alkyl and -phenylalkyl derivatives 9a–f of 6-deoxy-5a-carba-b-dl- and l-galactopyranosylamines
Compd
n
X
Inhibitory activity, IC50
K
i
(mM)
a-Fucosidase (bovine kidney)
b-Galactosidase (bovine liver)
b-Glucosidase (almond)
dl
l
dl
l
dl
l
9a
9b
9c
9d
9e
9f
5
7
9
11
2
4Ph
Me
Me
Me
Me
Ph
8.2
5.5
2.7
NT
7.5
NT
1.8
0.7
1.2
NT
3.7
NT
3.7
1.0
3.0
NT
0.73
NT
14.7
16.6
3.6
0.7 (0.11)
0.2 (0.009)
NT
1.5 (3.2)
2.0 (0.4)
NT
5.7
0.38 (0.057)
NT
5.1
1.2
0 .69 ) (0.04
2.4 (1.2 1) .4
10.2
NT, not tested.
a
Compounds 9a and 9e are racemic, and compound 9d is an l-enantiomer.

S. Ogawa et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3461–3463
3463
2
4
ꢁ
1
very strong inhibitor possessing characteristic pH-
dependent activity against b-glucosidase (almond): Ki
.39 mM at pH 5.5; K 0.057 mM at pH 6.8. These results
6. Data for compound 7: [a]
(300 MHz, CDCl ): d 4.31 (br d, 1H, J3,4=5.6 Hz, H-3), 3.98
br d, 1H, J3,4=5.6 Hz, H-4), 3.23 (br s, H-1), 2.97 (br d, 1H,
D 3
+19 (c 1, CHCl ); H NMR
3
(
0
i
J
1
7
2,3=3.2 Hz, H-2), 1.84[m, 3H, H-5, H-5a(ax), 5a(eq)], 1.78–
.25 (m, 10H, C H ), 1.04(d, 3H, J5,6=4.4 Hz, Me).
. Diequatorial cleavage of the epoxide ring of 7 proceeded
would indicate that optically active D-9e is a very
potent b-glucosidase inhibitor fully compatible with
11
6
10
1
0
isofagomine ) (K 0.11 mM, at pH 6.8) and calystegine
i
very slowly but regioselectively due to steric hindrance of the
(
K 0.75 mM, pH independent). Therefore, 6-deoxy-5a-
i
bulky 1,2-O-cyclohexylidene group.
carba-b- d-galactopyransylamine might be a promising
lead compound for development of new carba sugar-
type b-galactosidase inhibitors. Preliminary experi-
24
ꢁ
. For example, data for compound 9b: [a] +21 (c 1.4,
D
1
8
MeOH); H NMR (300 MHz, CD
3.59 (dd, 1H, J1,2=J2,3=9.5 Hz, H-2), 3.21 (br d, 1H,
2,3=9.5 Hz, H-3), 2.75 (dt, 1H, J=7.1 Hz, Jgem=11.0 Hz,
NHCH ), 2.39–2.51 (m, 2H, H-1, NHCH ), 1.59–1.70 [m, 2H,
H-5, H-5a(eq)], 1.29–1.36 [m, 7H, (CH CH , H-5a(ax)], 1.04
d, 3H, J5,6=6.5 Hz, H-6), 0.88 (t, 3H, J=6.6 Hz, CH CH );
3
OD): d 3.80 (br s, 1H, H-4),
1
2
ments suggest that N-alkyl-5a-carba-b-d-galactopyr-
anosylamines are moderate b-galactosidase inhibitors,
so that the hydrophobic area conferred in by the 5-
methyl branching on the cyclohexane ring is likely to
enhance binding potential at the active site of enzymes.
J
2
2
2
)
3
3
(
2
3
2
4
ꢁ
1
for 9c: [a] +1.4 (c 0.8, MeOH); H NMR (300 MHz,
CD OD): d 3.61 (br s, 1H, H-4), 3.45 (dd, 1H, J1,2=9.5 Hz,
D
3
J
1
1
1
2,3=9.3 Hz, H-2), 3.21 (br d, 1H, J2,3=9.3 Hz, H-3), 2.73 (dt,
2
H, J=7.3 Hz, Jgem=11.5 Hz, NHCH ), 2.46–2.60 (m, 2H, H-
Acknowledgements
, NHCH ), 1.45–1.57 [m, 4H, NHCH CH H-5, H-5a(eq)],
2
2
2,
2 7 3
.21–1.34[m, 15H, (C H ) CH , H-5a(ax)], 1.05 (d, 3H,
1
The authors sincerely thank Drs. Atsushi Takahashi
and Akihiro Tomoda (Hokko Chemical Industry, Co.
Ltd., Atsugi, Japan) for the biological assays, Prof.
Naoki Asano (Hokuriku University, Kanazawa, Japan)
for providing us with a sample of calystegine, Prof.
Yoshiyuki Suzuki (International University of Health
and Welfare, Otawara, Japan) for helpful discussions,
and Ms. Miki Kanto for her assistance in preparing this
manuscript.
J5,6=6.3 Hz, Me); for 9e: H NMR (300 MHz, CD OD): d
7.16–7.31 (m, 5H, Ph), 3.69 (br s, 1H, H-4), 3.50 (dd, 1H,
3
J
1,2=9.7 Hz, J2,3=9.6 Hz, H-2), 3.30 (br d, 1H, J2,3=9.6 Hz,
H-3), 2.82–3.06 [m, 4H, NH(CH ], 2.53 [ddd, 1H, J1,2=9.7
Hz, J1,5a(ax)=11.9 Hz, J1,5a(eq)=2.2 Hz, H-1], 1.59–1.68 [m,
H, H-5, H-5a(eq)], 1.39 [m, 1H, H-5a(ax)], 1.02 (d, 3H,
2 2
)
2
2
4
ꢁ
1
J5,6=6.6 Hz, H-6); for 9f: [a] +0.4 (c 1.5, MeOH); H NMR
D
(
300 MHz, CDCl
3
): d 7.04–7.17 (m, 5H, Ph), 3.60 (br s, 1H, H-
4
1
1
J
J
9
), 3.38 (dd, 1H, J1,2=9.3 Hz, J2,3=9.5 Hz, H-2), 3.18 (br d,
H, J2,3=9.5 Hz, H-3), 2.29–2.69 [m, 5H, NH(CH ) , H-1],
.45–1.62 [m, 6H, (CH ) Ph, H-5, H-5a(eq)], 1.24[br dd,
2
1,5(aax)=8.7 Hz, Jgem=12.6 Hz, H-5a(ax)], 0.92 (d, 3H,
5,6=6.5 Hz, H-6).
. Dr. Akihiro Tomoda (Hokko Chemical Industry, Co. Ltd.)
2 2
2
References and Notes
1
. Ogawa, S.; Mori, M.; Takeuchi, G.; Doi, F.; Watanabe,
M.; Sakata, Y. Bioorg. Med. Chem. Lett. 2002, 12, 2811.
. Ogawa, S.; Sekura, R.; Maruyama, A.; Yuasa, H.; Hashi-
moto, H. Eur. J. Org. Chem. 2000, 2089.
. Ogawa, S.; Watanabe, M.; Maruyama, A.; Hisamatsu, S.
Bioorg. Med. Chem. Lett. 2002, 12, 749.
. Takeuchi, M.; Kamata, K.; Yoshida, M.; Kameda, Y.;
carried out the biological assays applying
approach.
10. (a) Jespersen, T. M.; Dong, W.; Sierks, M. R.; Skrydstrup,
T.; Lundt, I.; Bols, M. Angew. Chem., Int. Ed. Engl. 1994, 33,
1778. (b) Ichikawa, Y.; Igarashi, Y.; Ichikawa, M.; Suhara, Y.
J. Am. Chem. Soc. 1998, 120, 3007.
11. Asano, N.; Kato, A.; Oseki, K.; Kizu, H.; Matsui, K. Eur.
J. Biochem. 1995, 229, 369.
12. N-Octyl and decyl derivatives have been shown to be
moderate b-galactosidase (bovine liver) inhibitors: IC50 2.3
and 22 mM, respectively: Ogawa, S.; Sakata, Y.; Fujieda, S.
Unpublished results.
a standard
2
3
4
Matsui, K. J. Biochem. 1990, 108, 42. Kameda, Y.; Takai, N.;
Asano, N.; Matsui, K. Chem. Pharm. Bull. 1990, 38, 1970.
5. Other than compound 6 the di-O-benzyl and 1,2-O-iso-
propylidene derivatives were subjected to conventional condi-
tions for epoxidation. Exclusive selectivity was obtained for 6.