Synthetic studies on the sulfur-cross-linked core of antitumor marine alkaloid, discorhabdins: Total synthesis of discorhabdin A

Article abstract of DOI:10.1002/1521-3773(20020118)41:2<348::AID-ANIE348>3.0.CO;2-I

A concise and efficient method to construct the characteristic sulfur-cross-linked core of an unprecedented antitumor marine alkaloid, discorhabdins, has been developed. By utilizing this methodology the first total synthesis of (±)-discorhabdin A (1) has

Full text of DOI:10.1002/1521-3773(20020118)41:2<348::AID-ANIE348>3.0.CO;2-I

COMMUNICATIONS
[14] Factors influencing configurational stability of binaphthyl systems are
not entirely predictable: A. K. Yudin, J. P. Martyn, S. Pandiaraju, J.
Zheng, A. Lough, Org. Lett. 2000, 2, 41.
Synthetic Studies on the Sulfur-Cross-Linked
Core of Antitumor Marine Alkaloid,
Discorhabdins: Total Synthesis of
Discorhabdin A**
[15] Data were corrected for Lorentz and polarization effects, and
equivalent reflections were averaged using the Bruker SAINT
software as well as utility programs from the XTEL library. The
structure was solved using SHELXTL and Fourier techniques, and
Hirofumi Tohma, Yuu Harayama, Miki Hashizume,
Minako Iwata, Masahiro Egi, and Yasuyuki Kita*
refined by least squares against jF 2 j using
a combination of
SHELXTL and the XTEL library. All hydrogen atoms were included
as fixed isotropic contributors in the final cycles of refinement.
Crystallographic data (excluding structure factors) for the structures
reported in this paper have been deposited with the Cambridge
Crystallographic Data Centre as supplementary publication no.
CCDC-173110 (rac-2a) and CCDC-173109 (rac-2b). Copies of the
data can be obtained free of charge on application to CCDC, 12 Union
Road, Cambridge CB21EZ, UK (fax: (‡44)1223-336-033; e-mail:
deposit@ccdc.cam.ac.uk).
Discorhabdins and prianosins have been isolated from
marine sponges such as New Zealand sponges of the genus
Latrunculia, Okinawan sponge Prianos melanos, and Fijian
sponge Zyzzya cf. Marsailis. Among the various discorhab-
dins (A R) isolated, discorhabdins A (1),[1a,b,d] B,[1b] D,[1c]
Q,[1e] and R[1f] have a unique sulfur-containing fused-ring
system incorporating an azacarbocyclic spirocyclohexadi-
enone and a pyrroloiminoquinone system (Scheme 1), and
show potent antitumor activity.[2] The discorhabdins have
[16] Crystal data for rac-2a: crystal dimensions 0.30 Â 0.10 Â 0.03 mm,
monoclinic, space group C2c, a ˆ 21.404(4), b ˆ 12.253(3), c ˆ
18.103(4) ä,
b ˆ 96.27(3)8;
Vˆ 4719.58 ä3,
Z ˆ 4,
1 ˆ
1.160 mgmmÀ3; m ˆ 0.71073mmÀ1, F(000) ˆ 1720, Bruker Smart6000
CCD diffractometer, 1.92 < q < 26.378, MoKa radiation, l ˆ 0.71073 ä,
W scans, Tˆ 112(2) K, 36868 reflections measured, 4813 unique, 1668
with I > 2s(I), À26 ꢀ l ꢀ 25, À15 ꢀ k ꢀ 14, À22 ꢀ l ꢀ 22; R ˆ 0.0667,
O
O
Br
N
Br
N
17
wR ˆ 0.0644, GOF ˆ 1.565, D1max ˆ 0.31 eäÀ3
.
16
[17] G. M. Diamond, S. Rodewald, R. F. Jordan, Organometallics 1995, 14,
5.
S
H
S
H
[18] For an example of metal-associated ligand atropisomerization, see:
a) M. T. Ashby, G. N. Govindan, A. K. Grafton, J. Am. Chem. Soc.
N
H
N
H
N
H
N
H
O
O
¬
1994, 116, 4801; b) M. D. Tudor, J. J. Becker, P. S. White, M. R. Gagne,
Discorhabdin B
Discorhabdin Q
(16,17-dehydro)
Discorhabdin A (1)
Organometallics 2000, 19, 4376.
(Prianosin A)
[19] Rac-1 (14.6 mg, 74.6 mmol) and lithium dimethylamide (3.8 mg,
74.5 mmol) were dissolved in toluene and allowed to stir at room
temperature for 30 min. The volatiles were removed in vacuo from the
violet solution to leave a residue that was redissolved in toluene,
treated with ZrCl4 (8.7mg, 37.3 mmol), and stirred at room temper-
ature. An orange precipitate formed over the course of 12 hours, from
which single crystals could be retrieved for analysis.
O
O
R
+
N
O
N
O
[20] Crystal data for rac-2b: crystal dimensions 0.15 Â 0.12 Â 0.10 mm,
S
S
monoclinic, space group P21/c, a ˆ 20.4685(12), b ˆ 17.3030(10), c ˆ
Vˆ 6608.52 ä3,
Z ˆ 8,
1 ˆ
N
H
N
H
N
H
21.1452(13) ä,
b ˆ 118.062(2)8;
H
N
H
H
1.465 mgmmÀ3; m ˆ 0.71073mmÀ1, F(000) ˆ 2976, Bruker Smart6000
CCD diffractometer, 2.24 < q < 27.508, MoKa radiation, l ˆ 0.71073 ä,
W scans, Tˆ 111 K, 48532 reflections measured, 15184 unique, 6638
with I > 2s(I), À24 ꢀ l ꢀ 26, À21 ꢀ k ꢀ 22, À27 ꢀ l ꢀ 23; R ˆ 0.0322,
O
R=OH:
R=H: Discorhabdin D
Prianosin C
Discorhabdin R
Scheme 1. Sulfur-containing discorhabdins.
wR ˆ 0.0241, GOF ˆ 0.627, D1max ˆ 0.36 eäÀ3
.
[21] A full discussion of the solid-state structure will be provided else-
where.
attracted the synthetic interest of several groups including
ours because of their cytotoxicity and unusual ring struc-
tures.[3] However, to the best of our knowledge, the total
syntheses of sulfur-containing discorhabdins have not yet
been reported because construction of the labile and highly
strained N,S-acetal (sulfur-cross-linked) core was difficult.
Furthermore, the timing and insertion point for the introduc-
[22] L. Gade, Chem. Commun. 2000, 173.
[*] Prof. Dr. Y. Kita, Dr. H. Tohma, Y. Harayama, M. Hashizume,
M. Iwata, M. Egi
Graduate School of Pharmaceutical Sciences
Osaka University
1
6 Yamada-oka, Suita, Osaka 565-0871 (Japan)
Fax : (‡81)6-6879-8229
E-mail: kita@phs.osaka-u.ac.jp
[**] This work was supported by a Grant-in-Aid for Scientific Research (S)
(No. 13853010), for Scientific Research on Priority Area (A)
(No. 13029062), and for the Encouragement of Young Scientists
(No. 13771331) from the Ministry of Education, Science, Sports, and
Culture, Japan. H.T. also thanks the Hoh-ansha Foundation for a
research fellowship. We thank Professor Murray H. G. Munro (Uni-
versity of Canterbury) for providing a copy of spectroscopic data of
natural discorhabdin A.
348
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
1433-7851/02/4102-0348 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 2

COMMUNICATIONS
tion of sulfur in discorhabdins have not yet been clarified
biosynthetically.[4]
We report herein synthetic studies on the sulfur-cross-
linked spirodienone core of discorhabdin alkaloids, which
results in the first total synthesis of (Æ)-discorhabdin A (1).
First, on the basis of a plausible hypothesis by Munro and
co-workers,[4] we examined the biosynthetically potential
route from makaluvamine F (2)[1d, 3h] using our previously
developed oxidative spirocycliza-
Br
HO
tion reaction with phenyliodine(iii)
bis(trifluoroacetate) (PIFA).[3d, 5]
N
The spirocyclization of 2 using
PIFA under various conditions
yielded a complex mixture. We also
examined the formation of a spiro-
dienone from 2 using the CuCl2/
NEt3/O2 system developed by Au-
S
N
H
N
H
O
Makaluvamine F (2)
bart and Heathcock,[3i] but obtained a complex mixture.
Therefore, we altered the synthetic strategy. Retrosynthetic
analysis of the highly strained core of 1, namely, the sulfur-
cross-linked spiro-fused ring system 3, is outlined in Scheme 2.
Key elements of our strategy include the iodine(iii)-induced
oxidative spirocyclization of 4 and the final introduction of the
sulfur group leading to the cross-linked core.
Scheme 3. Construction of sulfur-cross-linked core 3. a) PIFA-MK 10,
CF3CH2OH, 0.5 h, 42%; b) 6m aq HCl, 1,4-dioxane, 608C, 2.5 h, 76%;
c) À 2e, NaOMe (5 mol%), MeOH, 60% (Pb(OAc)4, CH2Cl2-MeOH,
45%; PhI(OAc)2, MeOH, 22%; PIFA, MeOH, 29%); d) AcSK-BF3 ¥ Et2O,
THF, À78 !48C, 24 h, 78%; e) 2m NH3/EtOH, 12 h, 51%.
OH
O
O
retro-
retro-
sulfur
spiro-
PhCOSH, and tBuSH, which may convert the methoxy group
of 8 into a sulfur group, the thioacetyl group was introduced
efficiently using AcSK in the presence of BF3 ¥ Et2O to give 9
in 78% yield. Several thiol groups could also be substituted in
low to moderate yields. Treatment of 9 with 2m NH3 in EtOH
yielded the sulfur-cross-linked compound (Æ)-3 in 51% yield
(Scheme 3). The structural features of 3 were supported by
spectroscopic data[8] and was compared with the data reported
for 1 by Munro and co-workers.[1b] The three-dimensional
structure of 3 was deduced from difference NOE spectra
recorded in (CD3)2SO. The observed enhancements
(Scheme 3) were in complete agreement with structure 3.
Thus, we applied this model study to the total synthesis of
discorhabdin A. We first examined the same path as that
towards the naphthoquinone model compound 3. Tritylation
of 5 followed by monobromination with NBS yielded 10 in
65% yield (2 steps; Scheme 4). A coupling reaction of 10 with
pyrroloiminoquinone 11a, which was prepared by our pre-
viously developed PIFA-induced pyrroloiminoquinone for-
mation,[9] provided 12 in 46% yield. We then examined the
oxidative spirocyclization reaction of 12 using PIFA. Al-
though various reaction conditions were tested, this reaction
did not give the corresponding spirodienone, but unexpect-
edly yielded a complex mixture. Thus, we modified the
synthetic strategy as follows: reduction of 10 with DIBAH
followed by silylation of the resulting alcohol with TBSCl
gave the bissilylated compound 13. Selective desilylation of 13
with TBAF in THF, followed by a coupling reaction with
1-tosylated pyrroloiminoquinone derivative 11b yielded 14.
Spirodienone formation using PIFA proceeded effectively in
the presence of MK10 to give 15 as a mixture of two
diastereomers in 45% yield.
O
O
cyclization
insertion
S
X
N
H
X
N
H
N
H
4
3
Scheme 2. Retrosynthetic analysis of sulfur-cross-linked core (3) of
discorhabdin A.
We set out to explore the feasibility of constructing the
sulfur-cross-linked core of discorhabdins using aminonaph-
thoquinone 6 as a model substrate. Compound 6 was readily
prepared from commercially available (l)-tyrosine methyl-
ester (5) and 1,4-naphthoquinone (Scheme 3). Oxidative
spirocyclization of
6
using PIFA/montmorillonite K10
(MK10), followed by acid hydrolysis (6n HCl/dioxane-H2O)
yielded the corresponding spirodienone carboxylic acid 7. The
use of MK10[6] improved the yield of this spirocyclization step
compared to our previous procedure.[3d, 5] We then attempted
the direct introduction of a sulfur functional group by
oxidative decarboxylation of 7 in the presence of several
thiols or AcSH. However, 7 was mostly recovered because of
the high reactivity of sulfur nucleophiles towards oxidants and
anodic oxidation. Thus, we examined an alternative route via
N,O-acetal 8, which could be readily prepared by oxidative
decarboxylation of 7. Consequently, the anodic oxidation[7]
(graphite anode/graphite cathode system in a nondivided cell)
of 7 proceeded smoothly in MeOH in the presence of 5 mol%
NaOMe to give 8 in 60% yield. In contrast, chemical
oxidation reactions using Pb(OAc)4, PhI(OAc)2 (PIDA), or
PIFA afforded 8 only in low to moderate yields. After an
extensive survey of sulfur nucleophiles, such as AcSR (R ˆ H,
TMS, K), EtOCS2 K, M2S (M ˆ Li, Na, Me3Si), TrSH,
Angew. Chem. Int. Ed. 2002, 41, No. 2
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
1433-7851/02/4102-0349 $ 17.50+.50/0
349

COMMUNICATIONS
used p-methoxybenzylthiol in 30% HBr-AcOH followed by
treatment with aqueous MeNH2 and gave 18[10] as well as the
undesired debrominated compounds 19 and 20, which were
formed by 1,4-addition to the bromo-enone side. Removal of
the Ts group of 18 with NaOMe provided (Æ)-discorhabdin A
(1) in 65% yield (Scheme 4).
The spectral data of synthetic 1 was identical to those
reported for the natural product.[1a,b] Further improvement of
the overall yield and an investigation into the total synthesis
of optically active discorhabdin A are now underway.
OH
OH
Br
Br
N
O
c) HCl, then
PIFA
a) TrCl
b) NBS
complex
mixture
5
N
10
NH
CO2Me
NHTr
N
H
MeO
N
d) DIBAH
e) TBSCl
H
O
11a
OH
12
CO2Me
O
OTBS
Br
Br
N
N
Br
f) TBAF
g) HCl, then
h) PIFA
N
N
NH
Ts
NHTr
N
H
N
O
MeO
N
Ts
O
14
Ts
TBSO
O
O
Received: September 17, 2001 [Z17910]
OTBS
OTBS
15
13
11b
O
Br
Br
N
N
[1] a) J. Kobayashi, J.-F. Cheng, M. Ishibashi, H. Nakamura, Y. Ohizumi,
Y. Hirata, T. Sasaki, H. Lu, J. Clardy, Tetrahedron Lett. 1987, 28, 4939;
b) N. B. Perry, J. W. Blunt, M. H. G. Munro, Tetrahedron 1988, 44,
1727; c) N. B. Perry, J. W. Blunt, M. H. G. Munro, T. Higa, R. Sakai, J.
Org. Chem. 1988, 53, 4127; d) D. C. Radisky, E. S. Radisky, L. R.
Barrows, B. R. Copp, R. A. Kramer, C. M. Ireland, J. Am. Chem. Soc.
1993, 115, 1632; e) M.-G. Dijoux, W. R. Gamble, Y. F. Hallock, J. H.
Cardellina, R. van Soest, M. R. Boyd, J. Nat. Prod. 1999, 62, 636; f) J.
Ford, R. J. Capon, J. Nat. Prod. 2000, 63, 1527, and references therein.
[2] a) L. R. Barrows, D. C. Radisky, B. R. Copp, D. S. Swaffar, R. A.
Kramer, R. L. Warters, C. M. Ireland, Anti-Cancer Drug Des. 1993, 8,
333; b) Q. Ding, K. Chichak, J. W. Lown, Curr. Med. Chem. 1999, 6, 1.
[3] a) G. G. Kubiak, P. N. Confalone, Tetrahedron Lett. 1990, 31, 3845;
b) H.-J. Knˆlker, K. Hartmann, Synlett 1991, 428; c) S. Nishiyama, J.-F.
Cheng, X. L. Tao, S. Yamamura, Tetrahedron Lett. 1991, 32, 4151; d) Y.
Kita, H. Tohma, M. Inagaki, K. Hatanaka, T. Yakura, J. Am. Chem.
Soc. 1992, 114, 2175; e) J. D. White, K. M. Yager, T. Yakura, J. Am.
Chem. Soc. 1994, 116, 1831; f) E. V. Sadanandan, S. K. Pillai, M. V.
Lakshmikantham, A. D. Billimoria, J. S. Culpepper, M. P. Cava, J.
Org. Chem. 1995, 60, 1800; g) D. Roberts, J. A. Joule, M. A. Bros, M.
Alvarez, J. Org. Chem. 1997, 62, 568; h) Y. Kita, M. Egi, T. Takada, H.
Tohma, Synthesis 1999, 885; i) K. M. Aubart, C. H. Heathcock, J. Org.
Chem. 1999, 64, 16, and references therein.
k) R1SH
i) BF3•Et2O
j) Pb(OAc)4,
MeOH
R1S
N
N
N
N
MeO
H
Ts
H
Ts
O
O
17
16
R':p-MeOC6H4CH2
O
O
O
Br
N
Br
N
N
S
H
S
+
+
S
H
N
N
H
N
N
N
R2
N
H
R2
H
H
O
R2
O
O
(±)-18: R2 = Ts
R
2 = Ts
(±)-20
(±)-19
l) NaOMe
(±)-1: R2=H
(discorhabdin A)
Scheme 4. Total synthesis of discorhabdin A (1). a) TrCl, Et3N, DMF,
quant.; b) NBS, DMF, 65%; c) 0.1n HCl/MeOH, then 11a, MeOH, 20 h,
46%; d) DIBAH, CH2Cl2, À788C !RT, 5 h, 96%; e) TBSCl, DBU,
CH2Cl2, 08C, 1.5 h, 87%; f) TBAF, THF, 0 8C, 0.5 h, quant.; g) 0.1n HCl/
MeOH, then 11b, MeOH, 16 h, 54%; h) PIFA-MK 10, CF3CH2OH, 0.5 h,
45%; i) BF3 ¥ Et2O, CH2Cl2, 08C !RT, 7h, 90%; j) Pb(OAc) 4, CH2Cl2/
MeOH (2/1), 08C, 1.5 h, 88%; k) p-MeOC6H4CH2SH, 30% HBr-AcOH,
CH2Cl2, À78 !48C, 15 h, 18(22%), 19(19%), 20(13%); l) NaOMe, THF-
MeOH, 08C, 1 h, 65%. Tr ˆ triphenylmethyl; NBS ˆ N-bromosuccinimide;
DIBAH ˆ diisobutylaluminum hydride; TBS ˆ tert-butyldimethylsilyl;
DBU ˆ 1,8-diazabicyclo[5.4.0]undec-7-ene; TBAF ˆ tetrabutylammonium
fluoride; Ts ˆ toluene-4-sulfonyl.
[4] R. E. Lill, D. A. Major, J. W. Blunt, M. H. G. Munro, C. N. Battershill,
M. G. McLean, R. L. Baxter, J. Nat. Prod. 1995, 58, 306.
[5] Y. Kita, T. Takada, M. Ibaraki, M. Gyoten, S. Mihara, S. Fujita, H.
Tohma, J. Org. Chem. 1996, 61, 223.
[6] In the absence of MK10, the spirodienone was obtained in 30% yield.
[7] H. Yamazaki, H. Horikawa, T. Nishitani, T. Iwasaki, Chem. Pharm.
Bull. 1990, 38, 2024.
The diastereomeric mixture 15 was desilylated, and then
converted into the methoxy compound 16 by oxidative
dealkylation with Pb(OAc)4. The remaining challenge was
to transform N,O-acetal 16 to N,S-acetal 17 because of the
instability of 16 towards both acidic and basic reaction
conditions. We first examined the introduction of a thioacetyl
group, and applied the best method for the preparation of a
model compound 9 to compound 16. However, 17 (R1 ˆ Ac)
was not obtained at all, but instead a complex mixture was
yielded. Thus, we re-investigated various sulfur nucleophiles,
such as Na2S, (R3Si)2S, TrSH, (p-MeO)BnSH, and tBuSH to
obtain 17. As a result, a p-methoxybenzylthiol group was
introduced efficiently in the presence of BF3 ¥ Et2O to give 17.
Debenzylation of 17, a labile and highly functionalized
compound, also required the mildest possible reaction
conditions. Our initial strategy was to perform a mild
debenzylation on the p-methoxybenzylsulfonium salt formed
by 1,4-addition of a sulfide group. Accordingly, we treated 17
with 30% HBr-AcOH followed by workup with NaHCO3 but
obtained only a trace amount of N-tosylated discorhabdin A
(18). Ultimately, we found an efficient one-pot transformation
procedure yielding 18 in 22% yield from 16. The procedure
[8] Characterization of (Æ)-3: orange solid; Rf ˆ 0.20 (silica gel, n-hexane/
AcOEt 2/1); m.p. 255 2578C (from AcOEt); 1H NMR (300 MHz,
CDCl3, 258C, TMS): d ˆ 8.07(d, J ˆ 8.0 Hz, 1H), 8.02 (d, J ˆ 8.0 Hz,
1H), 7.75 (t, J ˆ 8.0 Hz, 1H), 7.64 (t, J ˆ 8.0 Hz, 1H), 7.05 (d, J ˆ
10.5 Hz, 1H), 6.49 (brs, 1H), 6.19 (d, J ˆ 10.5 Hz, 1H), 5.29 5.33
(m, 1H), 4.60 (dd, J ˆ 12.0, 7.5 Hz, 1H), 2.61 2.83 (m, 4H);13C NMR
(75.5 MHz, [D6]DMSO, 258C): d ˆ 195.7, 180.6, 178.4, 154.9, 144.9,
135.0, 133.0, 132.4, 129.8, 125.7, 125.4, 124.9, 118.8, 59.5, 55.7, 46.8, 45.3,
39.4; IR (KBr): nÄ ˆ 1680, 1675, 1595, 1565, 1495 cmÀ1
; UV/Vis
(MeOH): lmax (e) ˆ 473 (1300), 270 (10900), 245 nm (10000); HR-
‡
MS: calcd for C18H13NO3S [M ]: 323.0616, found: 323.0621.
[9] Y. Kita, H. Watanabe, M. Egi, T. Saiki, Y. Fukuoka, H. Tohma, J.
Chem. Soc. Perkin Trans. 1 1998, 635.
[10] Characterization of (Æ)-18: red solid; Rf ˆ 0.70 (silica gel, CH2Cl2/
MeOH 40/1); m.p. > 3008C; 1H NMR (300 MHz, CDCl3, 258C,
TMS): d ˆ 8.02 (d, J ˆ 8.4 Hz, 2H), 7.49 (s, 1H), 7.43 (s, 1H), 7.33 (d,
J ˆ 8.4 Hz, 2H), 5.93 (d, J ˆ 3.9 Hz, 1H), 5.28 (t, J ˆ 3.9 Hz, 1H), 4.62
(dd, J ˆ 11.7, 7.8 Hz, 1H), 4.29 (dt,J ˆ 18.3, 6.0 Hz, 1H), 3.89 (dt, J ˆ
18.3, 9.0 Hz, 1H), 2.80 2.86 (m, 2H), 2.61 2.77 (m, 4H), 2.42 (s, 3H);
13C NMR (125 MHz, CDCl3, 258C): d ˆ 188.0, 168.3, 154.8, 152.8,
146.0, 140.6, 134.5, 129.8, 128.7, 126.1, 125.1, 121.9, 119.6, 118.4, 116.2,
61.3, 55.9, 50.0, 50.0, 45.5, 39.9, 21.7, 17.7 ; IR (KBr):nÄ ˆ 3365, 1680,
1660, 1610, 1595, 1565, 1525, 1455, 1375 cmÀ1; UV/Vis (CHCl3): lmax
(e) ˆ 483 (1400), 326 (12200), 265 (14400), 242 nm (18200); HR-FAB-
‡
MS: calcd for C25H21BrN3O4S2 [M‡H ]: 570.0157, found: 570.0171.
350
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
1433-7851/02/4102-0350 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 2