Synthetic study of (-)-norzoanthamine: Construction of the ABC ring moiety

Article abstract of DOI:10.1002/anie.200804544

Throw your hat in the ring: A highly diastereoselective synthesis of the ABC rings of (-)-norzoanthamine has been achieved starting from the (-)-Hajos-Parrish ketone (see scheme). Three asymmetric quaternary carbon centers on the C ring were constructed b

Full text of DOI:10.1002/anie.200804544

Communications
DOI: 10.1002/anie.200804544
Natural Products Synthesis (1)
Synthetic Study of (ꢀ)-Norzoanthamine: Construction of
the ABC Ring Moiety**
Yoshihisa Murata, Daisuke Yamashita, Katsushi Kitahara, Yohei Minasako,
Atsuo Nakazaki, and Susumu Kobayashi*
Angewandte
Chemie
1400
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 1400 –1403

Angewandte
Chemie
Zoanthamines are members of a marine alkaloid family of
compounds that display potent biological activities. In
particular, (ꢀ)-norzoanthamine (1) is considered to be a
promising anti-osteoporotic drug candidate because it sup-
presses the loss of bone weight and strength in ovariectomized
mouse model (Scheme 1).[1] In addition to the significant
Scheme 1. Selected zoanthamine alkaloids.
Scheme 2. Retrosynthetic analysis. MOM=methoxymethyl, TBS=tert-
butyldimethylsilyl, TES=triethylsilyl.
biological activity, zoanthamines have attracted a great deal
of attention from a synthetic point of view owing to their
complex and unique heptacyclic structure. Indeed, a number
of research groups, including ourselves, have been investigat-
ing synthetic routes to the zoanthamine family of com-
pounds.[2–4] So far, only the research group of Miyashita has
succeeded in the total synthesis of (ꢀ)-norzoanthamine.[4]
The characteristic structural features of (ꢀ)-norzoanth-
amine are: 1) a bisaminal skeleton in the CDEFG ring
moiety, 2) a trans-decalin motif in the AB rings, and 3) a
stereochemically dense C ring. As the bisaminal skeleton has
been reported to be rather unstable,[1b,c] we reasoned that
total synthesis could be achieved by forming this moiety at a
late stage of the synthesis. In this regard, we initially
developed the formation of the bisaminal skeleton. Indeed,
we have already reported an excellent methodology by
demonstrating a tandem cyclization to give a simplified
model of zoanthamines.[3] With this methodology established,
we embarked on the total synthesis of (ꢀ)-norzoanthamine.
Herein we focus on the construction of the ABC ring system
by demonstrating the synthesis of 4 from the enantiopure (ꢀ)-
Hajos–Parrish ketone 9 (Scheme 2),[5] and the following paper
describes the completion of the total synthesis of (ꢀ)-
norzoanthamine.[6]
(ꢀ)-norzoanthamine. Consequently, most of the previous
investigations have mainly focused on this moiety. The most
challenging synthetic problem is the synthesis of the stereo-
chemically dense C ring, which possesses three asymmetric
quaternary carbon centers (C9, C12, and C22). Contrary to
the impressive total synthesis by Miyashita and co-workers,[4]
our basic strategy involves an initial formation of the C ring
moiety with appropriate substituents that would be necessary
for the construction of the AB and DEFG ring moieties. The
key intermediate 6 might be derived from ketone 7, and the
quaternary chiral center at C12 could be constructed by a 1,4-
addition of vinyl cuprate reagent to enone 8. The methyl
group at C22 could be introduced in a syn fashion by 1,4-
addition to the (ꢀ)-Hajos–Parrish ketone 9, in which one of
the three quaternary methyl groups of the C ring is already
installed. Although there is no precedent concerning the 1,4-
addition to 9, we expected it to proceed by analogy to the
Wieland–Miescher ketone reaction.[7] We envisaged that the
AB ring moiety could be constructed in the desired trans
fashion by an intramolecular Diels–Alder reaction (IMDA).[8]
Furthermore, the requisite diene unit could be introduced by
the addition of a pentadienyl anion species onto the aldehyde
6.[9] The cyclopentanol moiety in 6 might serve as a handle for
annelating the DEFG ring system after oxidative cleavage.
Our first task was a stereoselective construction of the
asymmetric quaternary carbon center at C22 (Scheme 3).
Treatment of the (ꢀ)-Hajos–Parrish ketone 9 (> 99% ee)
with Gilman reagent, and subsequent trapping of the enolate
in situ by the addition of TMSCl and Et3N, provided the silyl
enol ether as a single diastereomer. The 1,4-addition occurred
from the less hindered b face in a stereoselective manner.
b-Selective reduction of the remaining carbonyl group with
LiAlH4, then quenching with aqueous HCl, and subsequent
protection as the TBS ether, afforded ketone 10 in 90%
overall yield from 9. The relative stereochemistry was
determined by NOE experiments.[10] Regioselective forma-
As mentioned earlier, construction of the ABC ring
system is one of the major obstacles to the total synthesis of
[*] Dr. Y. Murata, D. Yamashita, K. Kitahara, Y. Minasako,
Dr. A. Nakazaki, Prof. Dr. S. Kobayashi
Faculty of Pharmaceutical Sciences, Tokyo University of Science
2641 Yamazaki, Noda-shi, Chiba 278-8510 (Japan)
Fax: (+81)4-7121-3671
E-mail: kobayash@rs.noda.tus.ac.jp
[**] This research was supported in part by a Grant-in-Aid for Scientific
Research (B; KAKENHI No. 18390010) from the Japan Society for
the Promotion of Science.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200804544.
Angew. Chem. Int. Ed. 2009, 48, 1400 –1403
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1401

Communications
Scheme 3. Synthesis of primary alcohol 14. Reagents and conditions:
a) Me2CuLi, Et2O, ꢀ78!08C; then NEt3, TMSCl, 08C; b) LiAlH4, THF,
ꢀ788C; then aq. HCl (1n); c) TBSCl, imidazole, DMAP, DMF, RT,
90% (over 3 steps); d) TMSOTf, NEt3, CH2Cl2, 08C; e) cat. Pd(OAc)2,
O2, DMSO, 608C, 92% (over 2 steps); f) LDA, paraformaldehyde,
THF, DMF, ꢀ788C!RT, 87%; g) PhMe2SiLi, CuCN, THF, ꢀ788C; then
SiO2, RT, 95%; h) HBF4·OEt2, CH2Cl2, 08C; i) MOMCl, iPr2NEt, DMAP,
DCE, 608C; j) aq. H2O2, NaHCO3, KF, THF, MeOH, 08C!RT, 93%
(over 3 steps). DCE=1,2-dichloroethane, DMAP=4-dimethylaminopyr-
idine, DMF=N,N-dimethylformamide, DMSO=dimethyl sulfoxide,
LDA=lithium diisopropylamide, Tf=trifluoromethanesulfonyl,
TMS=trimethylsilyl.
Scheme 4. Synthesis of aldehyde 6. Reagents and conditions: a) TESCl,
imidazole, DMAP, CH2Cl2, 08C, quant.; b) MeLi, THF, ꢀ788C, 96%;
c) PDC, benzene, reflux, 39%; d) vinyl lithium, CuCN, Et2O, ꢀ78!
08C, 83% (d.r. 13:1); e) DIBAL-H, CH2Cl2, ꢀ788C; f) TBSOTf, 2,6-
lutidine, CH2Cl2, 08C, quant. (over 2 steps); g) PPTS, MeOH, RT;
h) DMP, CH2Cl2, 08C, 95% (over 2 steps). DIBAL-H=diisobutylalumi-
num hydride, DMP=Dess–Martin periodinane, PDC=pyridinium
dichromate, PPTS=pyridinium p-toluenesulfonate
which might be attributed to steric hindrance at the concave
face.
tion of silyl enol ether from 10 and subsequent Saegusa
oxidation gave enone 11.[11]
To construct the third quaternary carbon center at C12,
1,4-addition of enone 8 with vinyl cuprate was then per-
formed. The reaction of 8 with the corresponding lithium
cyanocuprate resulted in the formation of the desired ketone
7, which bears the three requisite quaternary carbon centers
on the C ring, in 83% yield with high diastereoselectivity
(d.r. 13:1). As expected, this conjugate addition occurred
from the more accessible convex face of the cis-fused ring
system. Face selective reduction of the carbonyl group in
ketone 7 and subsequent protection of the resulting hydroxy
group as the TBS ether afforded 16 as a single diastereomer.
After the selective cleavage of the primary TES group, the
resulting alcohol was subjected to a Dess–Martin periodinane
oxidation[14] to give aldehyde 6—a key intermediate used for
the preparation of the precursor of the intramolecular Diels–
Alder reaction.
The synthesis of an appropriately functionalized
ABC ring system was carried out by using an IMDA reaction
as a key step (Scheme 5). To unite aldehyde 6 and the siloxy
diene moiety, we utilized the nucleophilic addition of the allyl
lithium generated from unsymmetrical siloxy diene 17. It was
hoped that this reaction would furnish the requisite precursor
of the Diels–Alder reaction as employed by Oppolzer et al.[15]
Unfortunately, all attempts at reactions involving the lithium
species derived from 17 were unsuccessful and only gave a
mixture of regioisomers. Based on these results, we decided to
use allyl lithium generated from the symmetrical siloxy diene
18. The missing methyl group at C26 could be introduced at a
later stage of the synthesis. Treatment of aldehyde 6 with allyl
lithium derived from 18 in THFat ꢀ788C afforded the desired
hydroxy diene 19 in excellent yield as a single diastereomer.
The hydroxy group at C20, which generates a chiral center by
nucleophilic addition, eventually becomes a carbonyl carbon.
However, we were aware that the stereochemistry of the
hydroxy group would affect the Diels–Alder reaction. The
We next examined the introduction of a hydroxymethyl
group at C21 of enone 11. Initial attempts at the direct aldol
reaction with paraformaldehyde did not stop at the aldol
stage, but resulted in the formation of dehydrated exo olefin
12 in high yield. Other methods, such as the Yb(OTf)3-
mediated aldol reaction in aqueous solution,[12] were also
unsuccessful owing to low yields as well as low stereoselec-
tivities. We then considered an alternative route involving
enone 12. Exposure of enone 12 to (PhMe2Si)2Cu(CN)Li2 and
subsequent addition of silica gel led to g-silyl ketone 13 as a
single diastereomer. Highly diastereoselective protonation of
the resulting enolate proceeded from the less hindered
convex face to control the requisite R configuration at the
C21 position. The configuration was determined by NOESY
experiments that showed a correlation between the C21
hydrogen atom and the two angular methyl groups.[10] The
phenyldimethylsilyl group in 13 was converted into a fluo-
rodimethylsilyl group by treatment with HBF4·OEt2. During
this operation, concomitant deprotection of the silyl group
afforded the secondary alcohol, which was protected as the
MOM ether. Finally, oxidation with H2O2 gave the requisite
hydroxymethylated enone 14.[13] Although introduction of a
hydroxymethyl group into 11 required five steps, 14 was
obtained in a high overall yield (77%) with complete
stereoselectivity.
Scheme 4 shows the final stage of the synthesis of the
C ring. This most critical step secured the construction of the
third quaternary carbon center at C12. Protection of the
primary hydroxy group in enone 14 and subsequent 1,2-
addition of the resulting enone afforded tertiary alcohol 15 in
excellent yield as a single diastereomer. Oxidative rearrange-
ment of 15 using PDC provided the desired enone 8 in 39%
yield along with a substantial amount of the dehydrated diene,
1402
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 1400 –1403

Angewandte
Chemie
decalin scaffold with complete diastereoselectivity. The cyclo-
pentanol moiety in 4 might serve as a handle for annelating
the DEFG ring in the total synthesis of (ꢀ)-norzoanthamine,
which will be described in the following communication.
Received: September 16, 2008
Published online: November 26, 2008
Keywords: diastereoselectivity · Diels–Alder reaction ·
.
Michael addition · natural products · total synthesis
[1] a) S. Fukuzawa, Y. Hayashi, D. Uemura, A. Nagatsu, K.
Yamada, Y. Ijuin, Heterocycl. Commun. 1995, 1, 207 – 214;
b) M. Kuramoto, K. Hayashi, Y. Fujitani, K. Yamaguchi, T. Tsuji,
K. Yamada, Y. Ijuin, D. Uemura, Tetrahedron Lett. 1997, 38,
5683 – 5686; c) M. Kuramoto, K. Hayashi, K. Yamaguchi, M.
Yada, T. Tsuji, D. Uemura, Bull. Chem. Soc. Jpn. 1998, 71, 771 –
779.
[2] For a review, see: D. C. Behenna, J. L. Stockdill, B. M. Stoltz,
Angew. Chem. 2008, 120, 2400 – 2421; Angew. Chem. Int. Ed.
2008, 47, 2365 – 2386.
[3] a) N. Hikage, H. Furukawa, K.-i. Takao, S. Kobayashi, Tetrahe-
dron Lett. 1998, 39, 6237 – 6240; b) N. Hikage, H. Furukawa, K.-i.
Takao, S. Kobayashi, Tetrahedron Lett. 1998, 39, 6241 – 6244;
c) N. Hikage, H. Furukawa, K.-i. Takao, S. Kobayashi, Chem.
Pharm. Bull. 2000, 48, 1370 – 1372.
Scheme 5. Construction of the AB rings by an IMDA reaction. Reagents
and conditions: a) 18, sBuLi, THF, ꢀ788C, 93%; b) TBSOTf, 2,6-
lutidine, CH2Cl2, 08C, 91%; c) pyridine, toluene, 2108C; d) cat. Pd-
(OAc)2, DMSO, 708C, 69% (over 2 steps); e) MeLi, CuBr·SMe2,
TMSCl, HMPA, THF, ꢀ78!ꢀ428C; f) Pd(OAc)2, CH3CN, 608C, 78%
(over 2 steps). HMPA=hexamethylphosphoramide.
[4] M. Miyashita, M. Sasaki, I. Hattori, M. Sakai, K. Tanino, Science
2004, 305, 495 – 499.
[5] a) Z. G. Hajos, D. R. Parrish, J. Org. Chem. 1974, 39, 1615 – 1621;
b) Z. G. Hajos, D. R. Parrish, Org. Synth. 1984, 63, 26 – 36.
[6] D. Yamashita, Y. Murata, N. Hikage, K.-i. Takao, A. Nakazaki, S.
Kobayashi, Angew. Chem. 2008, DOI:10.102/ange.200804546;
Angew. Chem. Int. Ed. 2008, DOI:10.102/ange.200804546.
[7] R. A. J. Smith, D. J. Hannah, Tetrahedron 1979, 35, 1183 – 1189.
[8] For a review of IMDA reactions, see: a) W. R. Roush in
Comprehensive Organic Synthesis Vol. 5 (Eds.: B. M. Trost, I.
Fleming, L. A. Paquette), Pergamon, Oxford, 1991, pp. 513 –
550; b) A. G. Fallis, Acc. Chem. Res. 1999, 32, 464 – 474;
c) B. R. Bear, S. M. Sparks, K. J. Shea, Angew. Chem. 2001,
113, 864 – 894; Angew. Chem. Int. Ed. 2001, 40, 820 – 849; d) Y.
Suzuki, T. Murata, K.-i. Takao, K.-i. Tadano, J. Synth. Org.
Chem. Jpn. 2002, 60, 679 – 690; e) K.-i. Takao, R. Munakata, K.-i.
Tadano, Chem. Rev. 2005, 105, 4779 – 4807.
R configuration as shown was determined after an IMDA
reaction. After protection of 19 as the TBS ether, we were
able to examine the crucial IMDA process. To our delight, the
IMDA reaction of triene 20 proceeded at 2108C in toluene, in
the presence of a trace amount of pyridine as an acid
scavenger, and gave the desired silyl enol ether 21.[16] The silyl
enol ether 21 was found to be relatively unstable, and was
subsequently subjected to Saegusa oxidation under modified
reaction conditions to obtain enone 22 in 69% yield (over 2
steps) as a single diastereomer. The desired trans-decalin
structure was confirmed by NOE and NOESYexperiments,[10]
which revealed that the key IMDA reaction might proceed
through an exclusive exo-chair transition state, as anticipated.
NMR experiments also lead to the determination of the
R configuration at C20, inferring that the siloxy group at C20
occupies a favorable pseudoequatorial position in the tran-
sition state of the IMDA process. Finally, introduction of the
methyl group at C26 by means of 1,4-addition and subsequent
Saegusa oxidation afforded a,b-unsaturated ketone 4 bearing
the appropriately functionalized ABC rings in good overall
yield.
[9] 4-Siloxy-2,5-heptadiene could be a potential precursor for the
synthesis of zoanthamine (2).
[10] See the Supporting Information.
[11] a) Y. Ito, T. Hirao, T. Saegusa, J. Org. Chem. 1978, 43, 1011 –
1013; b) R. C. Larock, T. R. Hightower, G. A. Kraus, P. Hahn, D.
Zheng, Tetrahedron Lett. 1995, 36, 2423 – 2426.
[12] S. Kobayashi, Chem. Lett. 1991, 2087 – 2090.
[13] a) K. Tamao, T. Kakui, M. Akita, T. Iwahara, R. Kanatani, J.
Yoshida, M. Kumada, Tetrahedron 1983, 39, 983 – 990; b) I.
Fleming, R. Henning, H. Plaut, J. Chem. Soc. Chem. Commun.
1984, 29 – 31; c) K. Tamao, N. Ishida, J. Organomet. Chem. 1984,
269, c37 – c39; d) I. Fleming, P. E. J. Sanderson, Tetrahedron Lett.
1987, 28, 4229 – 4232.
[14] D. B. Dess, J. C. Martin, J. Org. Chem. 1983, 48, 4155 – 4156.
[15] W. Oppolzer, R. L. Snowden, S. P. Simmons, Helv. Chim. Acta
1981, 64, 2002 – 2021.
In conclusion, we have synthesized an ABC ring moiety of
(ꢀ)-norzoanthamine. The present approach highlights the
following important points: 1) all stereogenic centers in the
C ring were constructed with complete stereoselectivity
starting from the (ꢀ)-Hajos–Parrish ketone, (2) the intra-
molecular Diels–Alder reaction provided a requisite trans-
[16] Prolonged reaction time or higher reaction temperature resulted
in the undesirable isomerization of 21 to give the tetrasubsti-
tuted silyl enol ether.
Angew. Chem. Int. Ed. 2009, 48, 1400 –1403
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1403