T. Suzuki et al. / Bioorg. Med. Chem. Lett. 22 (2012) 4246–4248
4247
Table 1
1H and 13C NMR data for 1 and 2 (in D2O)
Position
1
2
1H
13C
1H
13C
(Multiplicity, J in Hz)
d
(Multiplicity, J in Hz)
d
Aglycone
2
3.57 (m)
4.02 (m)
2.33 (m)
67.8
3.59 (m)
4.05 (m)
2.45 (m)
68.3
3
40.6
87.8
39.9
87.8
3a
3b
4
5
6
131.2
124.9
121.6
130.9
109.6
148.5
99.7
130.2
125.3
121.1
131.3
108.2
148.7
98.4
7.33 (d, 7.3)
7.39 (dd, 7.6, 0.9)
6.93 (dd, 7,6, 7.3)
7.29 (ddd, 7.3, 7.9, 0.9)
6.78 (d, 7.9)
6.92 (dd, 7.3, 7.3)
7.25 (dd 7.3, 7.6)
6.83 (d, 7.6)
7
7a
8a
Sugar
10
5.59 (s)
5.51 (s)
4.88 (d, 9.2)
86.2
71.3
77.0
69.8
77.9
61.0
4.89 (d, 9.2)
85.5
70.1
78.0
69.9
78.0
61.2
20
3.68 (dd, 9.2, 9.5)
3.53 (dd, 9.5, 8.9)
3.39 (dd, 8.9, 9.5)
3.42 (m)
3.61 (dd, 12.2, 4.8)
3.73 (dd, 12.2, 1.6)
3.78 (dd, 9.2, 8.9)
3.59 (dd, 8.9, 8.8)
3.48 (dd, 8.8, 9.8)
3.54 (ddd, 9.9, 4.9, 2.1)
3.73 (dd, 12.5, 4.9)
3.83 (dd, 12.5, 2.1)
30
40
Scheme 2. Synthesis of 1 and 2 by N-glucosidation of 3.
50
60
the N-glucosidation of 3 with D- and L
-glucoses11–13 in quantitative
yields, respectively (Scheme 2). The NMR data and the specific
rotation (½a 2D1
ꢁ
ꢀ45, H2O, c 0.07)3 of synthetic 1 were completely
identical with those of natural 1, indicating that the absolute con-
figuration of 1 was as shown.
The effects of the compounds on C6 cell viability were ana-
lyzed.14 The natural product (1) showed no effect. However,
100
at 24 h (112.0 9.8%) and 48 h (114.2 11.0%) treatment. On the
other hand, -glucoside (2) significantly reduced the cell number
at 10 M (92.4 9.1%) and 100 M (83.9 8.0%) in 24 h treated
cells, and at 100 M (87.9 11.1%) in 48 h treated cells. Since dead
cells were hardly observed by the treatment with 2 (data not
shown), the reducing effect of the compound is thought to be
growth retardation. Glycosylated natural products such as sapo-
nins or glycoalkaloids have been reported to show anti-prolifera-
tive activity or apoptotic effect in mammalian cells, especially
cancer cells.15 Although they are expected to be utilized as anti-
cancer agents, some of them show cytotoxicity even in normal
cells. Tomato saponin called as tomatine, has been reported to be
cytotoxic, and its deglycosylation abolished the cytotoxicity.16 It
raises the possibility that each compound used in this study show
distinct effect on mammalian cells.
lM of the aglycone (3) significantly stimulated the cell growth
L
9.5 Hz, H-20), 3.53 (dd, J = 9.5, 8.9 Hz, H-30), 3.39 (dd, Hz, J = 8.9,
9.5 Hz, H-40), 3.42 (m, H-50), 3.61 (dd, J = 12.2, 4.8 Hz, H-6a0), 3.73
(dd, J = 12.2, 1.6 Hz, H-6b0). The presence of a 1,2-substituted phe-
nyl group (C-3b to C-7a) in the aglycon was also suggested by the
NMR data: dC 131.2 (C-3b), 124.9 (C-4), 121.6 (C-5), 130.9 (C-6),
109.6 (C-7), 148.5 (C-7a); dH 7.33 (d, J = 7.3 Hz, H-4), 6.92 (dd,
J = 7.3, 7.3 Hz, H-5), 7.25 (dd J = 7.3, 7.6 Hz, H-6), 6.83 (d,
J = 7.6 Hz, H-7). The structure of the other part, the condensed ring
part, in the aglycon and the linkage position with the phenyl were
elucidated by the correlations in the HMBC data (Fig. 1) and the
molecular formula. The N-glucosidic bond and its position between
the sugar and the aglycon were determined by the HMBC correla-
tions (H-8a/H-10, H-10/H-7a, H-10/H-8a, Fig. 1) and the relatively
high field chemical shift at C-10 (dC 86.2) compared with those of
O-glycosides.4,5
l
l
l
Acknowledgment
To confirm the planar structure and determine the whole abso-
lute configuration of 1, an optically active D-glucoside and L-gluco-
side, 1 and 2,6 were synthesized chemically. Previously, we
completed total synthesis of madindoline A, involving an asym-
metric oxidative ring-closure reaction of tryptophol to give 3 by
modified Sharpless epoxidation [Ti(O-i-Pr)4, (+)-DIPT, t-BuOOH,
CH2Cl2, ꢀ20 to ꢀ40 °C (Scheme 1).7,8 Interest in the synthetic
applications of 3 led us to develop to the asymmetric total synthe-
sis of other optically active indole alkaloid, (ꢀ)-physovenine,9 in a
concise manner.10 Two diastereomers, 1 and 2, were prepared by
We thank V.K. Deo (Shizuoka University) for valuable
discussion.
References and notes
1. Kawagishi, H.; Hota, K.; Masuda, K.; Yamaguchi, K.; Yazawa, K.; Shibata, K.;
Uzuka, N.; Matahira, Y. Biosci. Biotechnol. Biochem. 2006, 70, 2800.
2. The material, Makomotake, was purchased from local market in Shizuoka,
Japan. Fresh Makomotake (16.7 kg) was extracted with EtOH (22 L, 3 times)
and then with acetone (10 L, once). After the solutions had been combined and
concentrated under reduced pressure, the concentrate was partitioned
between CH2Cl2 and H2O, and then the H2O soluble part was partitioned
between EtOAc and H2O. The residue (20.0 g) obtained after removing EtOAc
was fractionated by silica gel flash column chromatography (80% CH2Cl2/
acetone, 50% CH2Cl2/MeOH, MeOH, each 2 L) to obtain seven fractions. Fraction
2 (eluted with 80% CH2Cl2/acetone, 1.28 g) was further separated by reversed-
phase HPLC (Develosil C30-UG, 15% MeOH, 5 mL/tube), and nine fractions were
obtained. Fraction 2–5 (65.7 mg) was further fractionated by reversed-phase
HPLC (Cosmosil cholester, 10% MeOH) to afford compound
1
(1.8 mg; Rt
77.6 min). Compound 1. mp 154–156 °C; ½a D28
ꢁ
ꢀ48 (c 0.30, H2O); IR (neat);
1486, 1608, 3367 cmꢀ1 1H and 13C NMR, see Table 1; ESIMS m/z 362 [M+Na]+;
;
Scheme 1. The enantioselective preparation of 3.
HRESIMS m/z 362.1191 [M+Na]+ (Calcd for C16H21NO7Na, 362.1216).