Enantiodivergent synthesis of both enantiomers of marine alkaloids haliclorensin and isohaliclorensin, a constituent of halitulin

Article abstract of DOI:10.1021/ol049887k

Starting from (3R)-5-benzotriazolyl-3-phenylperhydropyrido[2,1-b][1,3] oxazole 9, the enantiodivergent syntheses of both enantiomers of the marine alkaloids haliclorensin 1 and isohaliclorensin 3 have been achieved. Our syntheses feature ring-expansion re

Full text of DOI:10.1021/ol049887k

ORGANIC
LETTERS
2004
Vol. 6, No. 7
1139-1142
Enantiodivergent Synthesis of Both
Enantiomers of Marine Alkaloids
Haliclorensin and Isohaliclorensin,
a Constituent of Halitulin†
Jian-Feng Zheng, Li-Ren Jin, and Pei-Qiang Huang*
Department of Chemistry and The Key Laboratory for Chemical Biology of
Fujian ProVince, Xiamen UniVersity, Xiamen, Fujian 361005, P. R. China
pqhuang@xmu.edu.cn
Received January 18, 2004
ABSTRACT
Starting from (3R)-5-benzotriazolyl-3-phenylperhydropyrido[2,1-b][1,3]oxazole 9, the enantiodivergent syntheses of both enantiomers of the
marine alkaloids haliclorensin 1 and isohaliclorensin 3 have been achieved. Our syntheses feature ring-expansion reactions for the formation
of the aza-macrocycle ring system of 3 and sequential ring-expansion reactions (aza-Claisen rearrangement and Zip reaction) for the formation
of the aza-macrocycle ring system of 1.
Haliclorensin (1)1 and halitulin (2)2 are two unique alkaloids
isolated from the marine sponge Haliclona tulearensis, which
was collected in Sodwana Bay, Durban, South Africa. The
strong cytotoxicity of haliclorensin against P-388 mouse
leukemia cells and that of halitulin against several tumor cell
lines have stimulated studies toward the total syntheses of
both molecules. Steglich3 and Banwell’s4 syntheses of
haliclorensin allowed the revision of its structure to (-)-
(S)-1,3 and the initially assigned structure (3) for haliclorensin
was subsequently renamed isohaliclorensin.3 A recent report5
on the first total synthesis of halitulin also confirmed the
previously assigned structure (2) and allowed determination
of its absolute configuration (15S).
† Dedicated to Professor Dr. Khi-Rui Tsai on the occasion of his 90th
birthday.
(1) Koren-Goldshlager, G.; Kashman, Y.; Schleyer, M. J. Nat. Prod.
1998, 61, 282.
(2) Kashman, Y.; Koren-Goldshlager, G.; Gravalos, M. D. G.; Schleyer,
M. Tetrahedron Lett. 1999, 40, 997.
(3) (a) Heinrich, M. R.; Steglich, W. Tetrahedron Lett. 2001, 42, 3287.
(b) Heinrich, M. R.; Kashman, Y.; Spiteller, P.; Steglich, W. Tetrahedron
2001, 57, 9973.
(4) Banwell, M. G.; Bray, A. M.; Edwards, A. J.; Wong, D. J. New J.
Chem. 2001, 25, 1347.
(5) Heinrich, M. R.; Steglich, W.; Banwell, M. G.; Kashman, Y.
Tetrahedron 2003, 59, 9239.
Retrosynthetic analysis of 1 and 3 suggested 4 as a
common precursor (Scheme 1). We envisioned forming the
ring system of haliclorensin 1 via a ring expansion reaction8
on the amino-lactam derived from 4, and this, in turn, could
potentially be derived from 6. The latter could itself be
accessed from 7 by an aza-Claisen rearrangement,9,10 and 7
10.1021/ol049887k CCC: $27.50 © 2004 American Chemical Society
Published on Web 03/09/2004

oxazolidines by Grignard reagents is a well-known pro-
cedure,11-13 little was known about the stereochemical
behavior of simple bicyclic oxazolidines such as 814,15 in this
type of reaction. Treatment of 8 with 4 molar equiv of
vinylmagnesium bromide led to the vinylated products 11a/
11b in 80:20 ratio. The two diastereomers were separable
by column chromatography. To determine the stereochem-
istry of the major diastereomer 11a, the acetate 12b, prepared
from 11b, was hydrogenated to give a known piperidine 13b
(Scheme 3).16 Comparing the spectral data and optical
Scheme 1
could possibly be prepared from Husson’s oxazolopiperidine
811 or from Katritsky’s benzotriazolyl oxazolopiperidine 912
(Scheme 1).
Scheme 3
As shown in Scheme 2, we started the synthesis of 11a
Scheme 2
rotation value of 13b {[R]20 -65.1 (c 2.0, CHCl3)} with
D
those reported {lit.16b [R]23 -65.2 (c 2.3, CHCl3) for the
D
(2R,2′R)-enantiomer} allowed us to determine its absolute
configuration as 2R,2′R. Thus, the absolute configuration of
11b was 2S,2′R, and that of 11a was 2R,2′R.
Removal of the benzylic chiral auxiliary from 11a, without
also affecting the vinyl group, was not a trivial task. Agami’s
nonreductive procedure17 was adopted for this purpose. Thus,
stirring a thionyl chloride solution of alcohol 11a at room
temperature for 1.5 h, followed by treatment of the resulting
chloride with KCN in DMSO-THF, furnished, in one pot,
the desired 2-vinylpiperidine (R)-14. The crude (R)-14,
without purification, was allowed to react with propionyl
chloride (Scheme 4). In this way, (R)-N-propionyl-2-vinyl-
and 11b by condensing (R)-phenylglycinol with glutaralde-
hyde and benzotriazole11,12 to obtain 9 as a regio- (Bt1 and
Bt2) and diastereomeric mixture in high yield (Scheme 2).
To convert 9 to 8, a chemoselective reductive debenzotri-
azolation was required. We attempted Husson’s conditions,
namely, Raney nickel in refluxing THF for 20 h.11a However,
this led to complex mixtures of products. Fortunately, when
9 was treated with an excess of freshly prepared Raney-Ni
at room temperature for 7∼8 h, 811a was obtained as a 9:1
diastereomeric mixture in 57% yield.
Scheme 4
With compound 8 in hand, we proceeded to study the
nucleophilic alkylation of 8. Although the ring opening of
(6) Banwell, M. G.; Bray, A. M.; Edwards, A. J.; Wong, D. J. J. Chem.
Soc., Perkin Trans. 1 2002, 1340.
(7) Usuki, Y.; Hirakawa, H.; Goto, K.; Iio, H. Tetrahedron: Asymmetry
2001, 12, 3293.
(8) For a treatise on the ring-enlarging reactions, see: Hesse, M. Ring
Enlargement in Organic Chemistry; VCH Publishers: New York, 1991.
(9) (a) Suh, Y. G.; Lee, J. Y.; Kim, S. A.; Jung, J. K. Synth. Commun.
1996, 26, 1675. (b) For an enantioselective aza-Claisen rearrangement of
an analogue of 8, see: Suh, Y. G.; Kim, S. A.; Jung, J. K.; Shin, D. Y.;
Min, K. A.; Koo, B. A.; Kim, H. S. Angew. Chem., Int. Ed. 1999, 38, 3545.
(10) For related ring expansion reactions, see: (a) Edstrom, E. D. J. Am.
Chem. Soc. 1991, 113, 6690. (b) Diederich, M.; Nubbemeyer, U. Angew.
Chem., Int. Ed. 1996, 35, 1026. (c) Sudau, A.; Nubbemeyer, U. Angew.
Chem., Int. Ed. 1998, 37, 1141. (d) Sudau, A.; Munch, W.; Unbbemeyer,
U. J. Org. Chem. 2000, 65, 1710.
(11) (a) Francolis, D.; Poupon, E.; Lallemand, M. C.; Kunesch, N.;
Husson, H.-P. J. Org. Chem. 2000, 65, 3209. (b) Poupon, E.; Kunesch, N.;
Husson, H.-P. Angew. Chem., Int. Ed. 2000, 39, 1493.
(12) Katritzky, A. R.; Qiu, G.; Yang, B.; Steel, P. J. J. Org. Chem. 1998,
63, 6699.
piperidine (7) {[R]20D +53 (c 1.1, CHCl3)} was obtained in
an overall yield of 51% from 11a. The intermediate 15 was
also isolated in a yield of 10%.
We next addressed the key ring expansion reaction. When
a toluene solution of 7 was heated in the presence of LHMDS
(13) For a review, see: Husson, H.-P.; Royer, J. Chem. Soc. ReV. 1999,
28, 383.
(14) Poerwono, H.; Higashiyama, K.; Yamauchi, T.; Kubo, H.; Ohmiya,
S.; Takahashi, H. Tetrahedron 1998, 54, 13955 and refs cited therein.
(15) For the nucleophilic alkylation of a related ring system, see: (a)
Yamazaki, N.; Kibayashi, C. Tetrahedron Lett. 1997, 38, 4623. (b) Pandey,
G.; Das, P. Tetrahedron Lett. 1997, 38, 9073.
1140
Org. Lett., Vol. 6, No. 7, 2004

for 10 min, the enolate 7a was generated and the desired
aza-Claisen rearrangement9a took place, leading to the
formation of (-)-azacyclodec-5-en-2-one 6 {white solid, mp
148-149 °C, [R]D -48.8 (c 1.0, CHCl3)} in a yield of
75% (Scheme 5). Since we had predicted that the aza-Claisen
synthetic isohaliclorensin (3) as R, which further confirmed
the 2R,2′R stereochemistry assigned for 2-phenyl-2-(2-vinyl-
piperidin-1-yl)-ethanol (11a) (vide supra).
Next, we turned our attention to the asymmetric synthesis
of haliclorensin (1). Chemoselective reduction of the nitrile
group of 4 (H2, PtO2, Norit, H2SO4, 96 h)18 under acidic
conditions provided the desired amido-amine 16 in 87% yield
(Scheme 7). Treatment of 16 with 0.95 molar equiv of
20
Scheme 5
Scheme 7
rearrangement would proceed via the conformer 7a,9a this
would establish an (R)-configuration in 6. The (R)-stereo-
chemistry was confirmed by our synthesis of both (R)-
isohaliclorensin (3) and (R)-haliclorensin (1) (vide infra).
Saturation of olefinic double bond in 6 (H2, 10%Pd/C,
MeOH) furnished (+)-azacyclodecan-2-one 5 {white solid,
mp 152-154 °C, [R]20D +20.3 (c 1.0, CHCl3)} in quantitative
yield.
LHMDS in toluene at reflux led to the desired ring-
expanded19 product 1720 {mp 129-130 °C, [R]20 -5.6 (c
D
0.8, CHCl3)} as a white solid in 43% yield (77% based on
Treatment of lactam 5 with a catalytic amount of n-
butyllithium at -78 °C followed by addition of acrylonitrile
led to the addition product 4 {[R]20D -59.2 (c 1.05, CHCl3);
yield 65%} along with recovered starting material 5 (33%)
(Scheme 6). Finally, reduction of both the amide carbonyl
recovered starting material).
The final transformation of 17 to haliclorensin (1) turned
out to be problematic. Attempts to reduce 17 with lithium
aluminum hydride led to complex mixtures of products, with
the desired product 1 only being isolated in low yield. Finally,
it was found that the reduction of amide 17 with borane
generated in situ from a NaBH4-I2 system21 (THF, reflux,
16 h) provided the desired (R)-haliclorensin (R-1) {[R]20
D
Scheme 6
19.4 (c 0.8, MeOH); lit.1 [R]D -2.2 (c 1.3, MeOH) for natural
1; lit.3b [R]D -18.5 (c 0.6, MeOH) for (S)-1; lit.3b [R]20D 20
(c 2.0, MeOH) for (R)-1} in 75% yield.
Since the natural haliclorensin (1) was shown to consist
of (R)- and (S)-enantiomers in a 1:3 ratio, with the (S)-
enantiomer being predominant,3b we decided to pursue the
synthesis of the (S)-enantiomers of haliclorensin (1) and
isohaliclorensin 3. Toward this end, Katritzky’s method12 was
adopted for the synthesis of 11b. Thus, treatment of 9 with
1.0 molar equiv of vinylmagnesium bromide at -78 °C,
followed by reduction of crude 18 with NaBH4 provided 11b
as the major diastereomer (dr ) 89:11, overall yield, 54%)
and the cyano groups of 4 with an excess of borane dimethyl
sulfide complex at 60 °C for 15 h provided isohaliclorensin
(3) in 55% yield. Comparing the specific optical rotation of
(18) Kramer, U.; Guggisberg, A.; Hesse, M.; Schmid, H. HelV Chim.
Acta 1978, 61, 1342.
(19) For some other transamidation reactions, see: (a) Wasserman, H.
H.; Berger, G. D.; Cho, K. R. Tetrahedron Lett. 1982, 23, 465. (b) Crombie,
L.; Jones, R. C. F.; Osborne, S.; Mat-Zin, A. R. J. Chem. Soc., Chem.
Commun. 1983, 959. (c) Bienz, S.; Guggisberg, A.; Walchli, R.; Hesse, M.
HelV. Chim. Acta 1979, 62, 1932. (d) Crombie, L.; Jones, R. C. F.; Haigh,
D. Tetrahedron Lett. 1986, 27, 5147. (e) Crombie, L.; Jones, R. C. F.; Haigh,
D. Tetrahedron Lett. 1986, 27, 5151. (f) Wasserman, H. H.; Robinson, R.
P.; Matsuyama, H. Tetrahedron Lett. 1980, 21, 3493. (g) Manhas, M. S.;
Amin, S. G.; Bose, A. K. Heterocycles 1976, 5, 669.
(20) Attempts to determine the enantiomeric excess of 17 by HPLC with
several types of chiral columns were unsuccessful.
(21) Bhanu Prasad, A. S.; Bhaskar Kanth, J. V.; Periasamy, M.
Tetrahedron 1992, 48, 4623.
our synthetic isohaliclorensin (3) {[R]20 +70 (c 0.6,
D
MeOH)} with the reported values {lit.3a [R]D -70 (c 0.9,
MeOH) for (S)-3; lit.7 [R]20D +74.6 (c 0.9, MeOH) for (R)-
3} allowed us to confirm the absolute configuration of our
(16) (a) Munchhof, M. J.; Meyers, A. I. J. Org. Chem. 1995, 60, 7084.
(b) Andres, J. M.; Herraiz-Sierra, I.; Pedrosa, R.; Perez-Encabo, A. Eur. J.
Org. Chem. 2000, 1719. (c) For a related method, see: Guerrier, L.; Royer,
J.; Grierson, D. S.; Husson, H.-P. J. Am. Chem. Soc. 1983, 105, 7754.
(17) Agami, C.; Couty, F.; Evano, G. Tetrahedron Lett. 1999, 40, 3709.
Org. Lett., Vol. 6, No. 7, 2004
1141

were obtained in overall yields comparable to those for (R)-3
and (R)-1.
Scheme 8
To summarize, starting from (3′R)-9, the first enantio-
divergent syntheses of both enantiomers of isohaliclorensin
(3) and haliclorensin (1) have been achieved. Notably, good
agreement of the specific rotation values of both enantiomers
of 1 and 3 with reported data implies that, under controlled
conditions, racemization can be minimized during the ring
expansion reactions.
Acknowledgment. The authors are grateful to the Na-
tional Science Fund for Distinguished Young Investigators,
the NNSF of China (29832020; 20072031; 20272048;
203900505), the Ministry of Education (Key Project 104201),
and the Specialized Research Fund for the Doctoral Program
of Higher Education (20020384004) for financial support.
(Scheme 8). Compound 11b was then converted to (S)-3 and
(S)-1 using the procedures described for (R)-3 and (R)-1,
respectively (vide supra). In this way, (S)-isohaliclorensin
Supporting Information Available: Experimental pro-
cedures and spectral data for compounds 1, 3-5, 7, 8, 11,
16, and 17. This material is available free of charge via the
Internet at http://pubs.acs.org.
(S-3) {[R]20 -69 (c 0.5, MeOH); lit.3a [R]D -70 (c 0.9,
D
MeOH) for (S)-3} and (S)-haliclorensin (S-1) {[R]20D -18.2
(c 0.4, MeOH); lit.3b [R]D -18.5 (c 0.6, MeOH) for (S)-1}
OL049887K
1142
Org. Lett., Vol. 6, No. 7, 2004