The first direct synthesis of α-mangostin, a potent inhibitor of the acidic sphingomyelinase

Article abstract of DOI:10.1016/S0040-4039(01)02137-2

A total synthesis of α-mangostin 1a has been achieved. The key cyclization reaction to construct the xanthone framework was undertaken by employing the PPh₃-CCl₄ conditions. The inhibitory activities of 1a and the benzophenone intermediate 16 against the acidic sphingomyelinase were discussed.

Full text of DOI:10.1016/S0040-4039(01)02137-2

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 291–293
The first direct synthesis of a-mangostin, a potent inhibitor of
the acidic sphingomyelinase
Kazuhiko Iikubo,a Yuichi Ishikawa,a Noritaka Ando,b Kazuo Umezawab and Shigeru Nishiyamaa,*
aDepartment of Chemistry, Faculty of Science and Technology, Keio University, Hiyoshi 3-14-1, Kohoku-ku,
223-8522 Yokohama, Japan
bDepartment of Applied Chemistry, Faculty of Science and Technology, Keio University, Hiyoshi 3-14-1, Kohoku-ku,
223-8522 Yokohama, Japan
Received 22 October 2001; revised 5 November 2001; accepted 9 November 2001
Abstract—A total synthesis of a-mangostin 1a has been achieved. The key cyclization reaction to construct the xanthone
framework was undertaken by employing the PPh3–CCl4 conditions. The inhibitory activities of 1a and the benzophenone
intermediate 16 against the acidic sphingomyelinase were discussed. © 2002 Elsevier Science Ltd. All rights reserved.
From among extensive investigations to acquire new
active substances in cancer chemotherapy, inhibitory
activity against the acidic sphingomyelinase closely
related to apoptosis-induction was found in a-man-
gostin 1a,1 known as a component of mangosteen
Garcinia mangostana L. (Guttiferae).2 While such bio-
logical activities were confirmed as a competitive antag-
onist of the histamine H1 receptor,3a inhibition of
topoisomerases I and II,3b antibacterial activity against
Helicobacter pylori,3c anti-inflammatory activities3d and
inhibition of oxidative damage of human LDL,3e no
synthesis of 1a carrying the intriguing highly-function-
alized xanthone structure has been published, with the
exception of the formal process by Lee:4 b-mangostin
1b and dimethylmangostin 1d were synthesized from
chroman derivatives. Since conversion of 1d into three
other mangostins 1a–c was accomplished, the synthesis
of 1d constituted the total synthesis of 1a.5 Accordingly,
an effective synthetic route to 1a and related deriva-
tives, would be required to understand the detailed
mode of action of their inhibitory activities. This back-
ground prompted us to initiate a synthetic investigation
of 1a.
cleaved at the diaryl ether moiety, and the resulted
benzophenone would be divided into two aromatic
derivatives (A, B). Upon employing this strategy, the
following problems would be solved: (1) the specific
phenol groups should be effectively activated and/or
protected to construct the diaryl ether moiety in a
regioselective manner. (2) Since the prenyl groups tend
to cyclize leading to the corresponding chroman deriva-
tives,4 all of the reaction steps should be manipulated
under mild reaction conditions.
O
O
OH
O
O
OH
MeO
HO
R1O
R2O
OH
OR3
R1 = R3 = Me, R2 = H
R1 = R2 = R3 = H
R1 = R2 = R3 = Me
β-mangostin 1b
γ-mangostin 1c
dimethylmangostin 1d
α-mangostin 1a
OP
According to the retrosynthetic analysis (Fig. 1), it
might be reasonable that a-mangostin 1a would be
MeO
PO
Br OHC
α-mangostin 1a
P = protective group
+
OP
PO
OP
fragment B
fragment A
Keywords: Garcinia mangostana L.; a-mangostin; acidic sphin-
gomyelinase; apoptosis.
* Corresponding author. Tel./fax: +81-45-566-1717; e-mail: nisiyama
@chem.keio.ac.jp
Figure 1. Mangostins 1a–d, and the retrosynthetic analysis of
1a.
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)02137-2

292
K. Iikubo et al. / Tetrahedron Letters 43 (2002) 291–293
Synthesis of fragment 6. The synthesis was commenced
by protection of 2,4-dihydroxybenzaldehyde 2 by ben-
zyl groups, followed by Baeyer–Villiger oxidation and
acid hydrolysis to provide the corresponding phenol 3
(Scheme 1). Compound 3 was subsequently submitted
to bromination and allylation leading to 4 in high
overall yield. Upon heating of 4 at 160°C, the expected
Claisen rearrangement was effected to produce an allyl-
benzene, and the resulted phenol was protected as a
methyl ether, leading to 5. The Lemieux–Johnson oxi-
dation of 5, followed by the Wittig reaction provided
the desired bromobenzene 6.
OHC
HO
HO
O
Br
a, b
c, d
OH
BnO
OBn
BnO
OBn
2
3
4
MeO
BnO
Br
OBn
MeO
BnO
Br
OBn
e, f
g, h
5
6
Scheme 1. (a) BnBr, K2CO3, DMF, rt, 96%; (b) mCPBA,
CH2Cl2, rt; 6 M HCl, MeOH, rt 95% in two steps; (c) Br2,
CHCl3, rt, 84%; (d) allyl bromide, K2CO3, DMF, rt, 80%; (e)
160°C, 73%; (f) MeI, K2CO3, DMF, rt, 87%; (g) OsO4,
NaIO4, Et2O/H2O (1/1), rt, 95%; (h) iPrPh3P+I−, nBuLi,
THF, 0°C, 72%.
Synthesis of fragment 13. Another fragment 13 was
synthesized from 1,3,5-trihydroxybenzene (phloroglu-
cinol) 7 (Scheme 2). Thus, 7 was exhaustively protected
with MOM groups, and the following prenylation gave
8 in high overall yield. After introduction of ethoxycar-
bonyl groups to 8, the resulting 9 was subsequently
submitted to acid-catalyzed methanolysis, TBS-protec-
tion, then DIBAL-reduction to give benzyl alcohol 10.
Among such oxidation methods attempted as MnO2,
PCC, Swern and Dess–Martin oxidations, only IBX6
effected the desired conversion of 10 into the aldehyde
11, fortunately with concomitant removal of the TBS
group adjacent to the prenyl group.7 The position of
the phenol generated was confirmed by the NOE experi-
ments of 12, as depicted in Scheme 2. For utilization in
the following regioselective cyclization, this phenol was
protected as an MOM ether to 12, and the following
manipulation involving the exchange of the TBS to
benzyl groups, provided 13. With the two benzalde-
hydes 12, 13 in hand, both compounds were submitted
to the coupling reaction with 6.
OH
OMOM
OMOM
c
a, b
EtOOC
MOMO
HO
OH
MOMO
OMOM
OMOM
7
8
9
OTBS
OR
H
OHC
d - f
HO
g
TBSO
OTBS
TBSO
OTBS
NOE
10
NOE
11: R = H
h
OMOM
12: R = MOM
OHC
BnO
i, j
OBn
13
Coupling reaction to the target molecule. Coupling of 13
with an anion generated from 6 provided the corre-
sponding alcohol 14 in moderate yield, although the
case employing 12 was unsuccessful, probably owing to
the steric hindrance of the TBS group (Scheme 3).
Compound 14 was converted in two steps into the
benzophenone 15.8 Although cyclization reactions
under the usual acidic and basic conditions were unsuc-
Scheme 2. (a) NaH, MOMCl, DMF, rt, 96%; (b) nBuLi;
prenyl bromide, THF, 0°C, 89%; (c) nBuLi; (EtO)2CO, THF,
0°C, 95%; (d) CSA, MeOH, 60°C, 100%; (e) TBSCl, DMAP,
Et3N, DMF, rt, 100%; (f) DIBAL-H, toluene, −78°C, 78%;
(g) IBX, toluene/DMSO (1/1), rt, 76%; (h) NaH, MOMCl,
CH2Cl2, rt, 65%; (i) TBAF, THF, 0°C, 100%; (j) BnBr,
K2CO3, DMF, rt, 98%.
OH OMOM
O
OMOM
MeO
MeO
HO
a
b, c
+ 13
6
BnO
OBn
OH
OBn OBn
OH OH
14
15
e
d
O
OH
O
O
OH
MeO
HO
MeO
HO
OH
OH
OH OH
1a
16
Scheme 3. (a) sBuLi, THF, −78°C, 49%; (b) IBX, toluene/DMSO (1/1), rt, 76%; (c) 10% Pd/C, HCO2NH4, acetone, rt, 63%; (d)
PPh3, CCl4, THF, rt, then silica gel, 43%; (e) CSA, MeOH, rt, 76%.

K. Iikubo et al. / Tetrahedron Letters 43 (2002) 291–293
293
cessful, the PPh3–CCl4 protocol9 with the following
silica gel treatment, enabled the desired cyclization, as
well as removal of the MOM group directly to give
a-mangostin 1a. This method was developed in this
laboratory for conversion of b-triketide into the corre-
sponding g-pyrones under mild acidic conditions.
Transformation of 15 into 1a might be initiated by
phosphorylation of the phenols, and the following
abstraction of triphenylphosphine oxide effected the
cyclization. Deprotection of the MOM group was
simultaneously conducted with the plausible active spe-
2. Isolation: Schmid, W. Liebigs Ann. Chem. 1855, 93, 83–
88; Dragendorff, O. Liebigs Ann. Chem. 1930, 482, 280–
301. Structural determination: Yates, P.; Stout, G. H. J.
Am. Chem. Soc. 1958, 80, 1691–1700; Scheinmann, F.
Chem. Commun. 1967, 1015–1017; Stout, G. H.; Krahn,
M. M.; Yates, P.; Bhat, H. B. Chem. Commun. 1968,
211–212.
3. (a) Chairungsrilerd, N.; Furukawa, K.; Ohta, T.; Nozoe,
S.; Ohizumi, Y. Eur. J. Pharmacol. 1996, 314, 351–356;
(b) Tosa, H.; Iinuma, M.; Tanaka, T.; Nozaki, H.;
Ikeda, S.; Tsutsui, K.; Tsutsui, K.; Yamada, M.; Fuji-
mori, S. Chem. Pharm. Bull. 1997, 45, 418–420; (c)
Hasegawa, H.; Sasaki, S.; Aimi, N.; Takayama, H.;
Koyana, T. Jpn. Patent 1996, 08231396; (d)
Shankarayan, D.; Gopalakrishnan, C.; Kameswaran, L.
Arch. Int. Pharmacodyn. Ther. 1979, 239, 257–269; (e)
Williams, P.; Ongsakul, M.; Proudfoot, J.; Croft, K.;
Beilin, L. Free Rad. Res. 1995, 23, 175–184.
4. Lee, H.-H. J. Chem. Soc., Perkin 1 1981, 3205–3213.
5. Jefferson, A.; Quillinan, A. J.; Scheinmann, F.; Sim, K.
Y. Aust. J. Chem. 1970, 23, 2539–2543.
6. (a) Frigerio, M.; Santagostino, M.; Sputore, S. J. Org.
Chem. 1999, 64, 4537–4538; (b) Nicolaou, K. C.; Zhong,
Y.-L.; Baran, P. S. J. Am. Chem. Soc. 2000, 122, 7596–
7597; (c) Wirth, T. Angew. Chem., Int. Ed. 2001, 40,
2812–2814.
cies (Ph3P+Cl·CCl3 ). The spectroscopic data of the
−
synthetic sample was identical with those of the natural
product.1
Biological activity. The biological assessment of the
benzophenone derivative (e.g. 16) possessing similar
functions to 1a, has not been reported, in contrast to
tetrahydro- and diacetyl-xanthones 17 and 18 derived
from 1a: the hydrophylic phenols were essential to
exhibit the activity (18), whereas the olefinic moiety
gave no critical effect (17) (Fig. 2). Accordingly, com-
pound 15 was converted by acid hydrolysis into 16 to
confirm its activity as part of our biochemical investiga-
tion of 1a. In the biological assay in the presence of
acidic sphingomyelinase and NBD–sphingomyelin,1 the
benzophenone 1610 exhibited the comparable inhibitory
activity to that of 1a (1a: IC50 4.21 mg/ml; 16: IC50
16.56 mg/ml). Since lower cytotoxicity than that of 1a
was observed, 16 will provide new information towards
understanding the mechanism of the acidic sphingo-
myelinase-regulated apoptosis. Extensive investigation
of the structure–activity relationship is in progress.
7. Steric hindrance of the TBS group interfered with the
prenylation of the tri-TBS derivative of 7. Accordingly,
a stepwise process was required as described in Scheme
2.
8. When using the tri-MOM derivative of 13, efforts to
remove the protective groups from the corresponding
benzophenone product were unsuccessful. Upon reacting
even under CSA in MeOH (30°C) conditions, by-prod-
ucts possessing
a chroman-framework with MOM
O
O
OH
groups were observed, cf. Mahabusarakam, W.;
Pakawatchai, C.; Wiriyachitra, P.; Taylor, W. C.; Skel-
ton, B. W.; White, A. H. Aust. J. Chem. 1998, 51,
249–254.
O
O
OH
MeO
AcO
MeO
HO
OAc
OH
18: inactive
17: active
9. (a) Arimoto, H.; Nishiyama, S.; Yamamura, S. Tetra-
hedron Lett. 1990, 31, 5619–5620; (b) Yamamura, S.;
Nishiyama, S. Bull. Chem. Soc. Jpn. 1997, 70, 2025–
2037.
Figure 2. a-Mangostin derivatives 17, 18 and their inhibitory
activities against the acidic sphingomyelinase (Ref. 1).
10. 16: lH (CDCl3) 1.39 (3H, s), 1.50 (3H, s), 1.75 (3H, s),
1.80 (3H, s), 3.32–3.34 (4H, complex), 3.76 (3H, s), 4.97
(1H, m), 5.23 (1H, m), 5.91 (1H, s) and 6.50 (1H, s); lC
(CDCl3) 17.5, 17.9, 21.5, 25.4, 25.8, 25.9, 29.1, 61.9,
95.8, 102.8, 106.4, 106.8, 118.4, 121.1, 121.8, 134.1,
134.2, 139.7, 150.8, 152.1, 160.0, 161.4, 163.2 and 196.1.
References
1. Okudaira, C.; Ikeda, Y.; Komdo, S.; Furuya, S.; Hira-
bayashi, Y.; Koyano, T.; Saito, Y.; Umezawa, K. J.
Enzym. Inhib. 2000, 15, 129–138.