Total synthesis of (-)-amphidinolide K

Full text of DOI:10.1002/anie.200805266

Communications
DOI: 10.1002/anie.200805266
Natural Product Synthesis
Total Synthesis of (À)-Amphidinolide K**
Haye Min Ko, Chung Whan Lee, Hyung Kyoo Kwon, Hea Seung Chung, Soo Young Choi,
Young Keun Chung, and Eun Lee*
(À)-Amphidinolide K (1; see Scheme 1) is a member of the
cytotoxic macrolides that were isolated by Kobayashi and co-
workers from the laboratory-cultured dinoflagellates Amphi-
dinium sp., which are symbionts of the marine flatworms
Amphiscolops sp. found in Okinawan.[1] Amphidinolide K (l)
is known to possess cytotoxic activity against L1210 (IC50
=
1.65 mgmLÀ1) and KB (IC50 = 2.9 mgmLÀ1) cancer cells in vi-
tro. The total synthesis of (+)-amphidinolide K reported by
Williams and Meyer[2] has clarified the problems concerning
configurational ambiguities at C2, C4, and C18, and as a result
(À)-amphidinolide K (1) was found to be the natural product.
The unique structural features and potent bioactivity of 1
elicited considerable interest in the synthetic community,[3]
and herein we wish to report the results of our recent efforts
on the synthesis of this intriguing molecule.
In the retrosynthetic analysis, epoxidation of the allylic
alcohol intermediate A (R = H) was envisaged as the final
step (Scheme 1). The macrolide intermediate A would be
obtained by lactonization of the seco acid B, which would in
turn be synthesized through the Julia–Kocienski reaction[4] of
aldehyde C and sulfone D. Fragment C would be obtained
from fragment E (Y= B(pinacol)) by a Suzuki coupling
reaction. Fragment E would be the product of the enyne
cross-metathesis reaction[5] of alkynyl boronate F and alkene
G. In another key step, a radical cyclization reaction[6] of the
homopropargylic b-alkoxyacrylate I would provide the meth-
ylidene-substituted oxolane intermediate H en route to frag-
ment D.
Scheme 1. Retrosynthetic analysis of (À)-amphidinolide K (1). Ar=aro-
matic.
In practice, b-alkoxyacrylate 4 was prepared from the
known homopropargylic alcohol 3[7] (Scheme 2). Next, the
radical cyclization reaction[6] of 4 in the presence of tributyl-
stannane and triethylborane proceeded efficiently and gave
mainly (16:1) the cis-2,5-disubstituted oxolane intermediate 5
after acidic destannylation. The corresponding aldehyde was
converted into the homologous aldehyde 6, which was then
treated with alcohol 7[8] to produce the homoallylic alcohol 8.
After protection of the alcohol as the benzoate derivative, 1-
phenyl-1H-tetrazolyl sulfone 9 was prepared by using the
standard procedures as outlined.
Olefin 12 was prepared from the known alcohol 11[9]
through mesylation and subsequent reduction (Scheme 3).
Alkyne 14 was prepared from (R)-glycidol (13) through
protection of the alcohol with THP, then treatment with
lithium TMS-acetylide, protection with a TBS group, and
hydrolytic cleavage of the alkynyl TMS group. Then alkynyl
boronate 15 was prepared from alkyne 14 under standard
reaction conditions. Subsequently, the enyne cross-metathesis
reaction[5,10] of 15 with olefin 12 proceeded effectively in the
presence of the second-generation Grubbs catalyst to yield a
mixture (7.5:1) that favored the desired E isomer 16. The use
of alkynyl boronate 15 in the enyne cross-metathesis reaction
was important; the use of methyl-substituted alkynes did not
yield useful amount of the cross-metathesis products. For the
synthesis of diene 17 from vinyl boronate 16, a Suzuki–
[*] H. M. Ko, C. W. Lee, H. K. Kwon, H. S. Chung, S. Y. Choi,
Prof. Dr. Y. K. Chung, Prof. Dr. E. Lee
Department of Chemistry, College of Natural Sciences
Seoul National University, Seoul 151-747 (Korea)
Fax: (+82)2-889-1568
E-mail: eunlee@snu.ac.kr
[**] This work was supported by a grant from the Marine Biotechnology
Program funded by the Ministry of Land, Transport, and Maritime
Affairs, Republic of Korea, and a grant from the Korea Research
Foundation (MOEHRD, Basic Research Promotion Fund; KRF-
2008-314-C00198). BK21 graduate fellowship grants to H.M.K.,
C.W.L., H.K.K., and H.S.C., and a Seoul Science Fellowship grant to
H.M.K. are gratefully acknowledged.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200805266.
2364
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 2364 –2366

Angewandte
Chemie
Scheme 3. Synthesis of the fragment C. a) TBDPSCl; b) LiBH4;
c) SO3·pyridine, TEA, DMSO; d) (E)-MeCHCHCH2B(dIpc)2; e) MsCl,
TEA/CH2Cl2 (1:2); f) LiAlH4, diethyl ether, reflux; g) DHP; h) LiCC-
SiMe3; i) TBSOTf; j) K2CO3, MeOH; k) nBuLi, (pinacol)B(OiPr), THF,
À788C; HCl, RT; l) 12, [(H2IMes2)(P(Cy)3)RuCl2CHPh] (15 mol%),
CH2Cl2, reflux; m) [Pd(Ph3P)4] (20 mol%), MeI, TlOEt, THF/H2O (3:1).
Cy =cyclohexyl, DHP=3,4-dihydro-2H-pyran, DMSO=dimethyl sulfox-
ide, Ipc=isopinocampheyl, Mes=2,4,6-trimethylphenyl, Ms=meth-
anesulfonyl, TBS=tert-butyldimethylsilyl, TEA=triethylamine, THP=
tetrahydropyranyl.
Scheme 2. Synthesis of the fragment D. a) TBDPSCl; b) LiCCSiMe3;
c) K2CO3, MeOH; d) CHCCO2Me, NMM, CH2Cl2; e) nBu3SnH, Et3B,
toluene; f) pTsOH, CH2Cl2; g) DIBAL, CH2Cl2, À788C;
h) (Ph3P+CH2OMe)ClÀ, tBuOK, THF; Hg(OAc)2, THF/H2O (10:1),
08C; i) 7, TFA, CH2Cl2 (0.03m); j) BzCl, DMAP, CH2Cl2; k) TBAF, THF;
l) PTSH, DIAD, Ph3P, THF; m) (NH4)6[Mo7O24]·4H2O, H2O2, EtOH.
Bz=benzoyl, DIAD=diisopropyl azodicarboxylate, DIBAL=diisobutyl-
aluminum hydride, DMAP=4-dimethylaminopyridine, NMM=N-meth-
ylmorpholine, PT=1-phenyl-1H-tetrazolyl, TBAF=tetra-n-butylammo-
nium fluoride, TBDPS=tert-butyldiphenylsilyl, TFA=trifluoroacetic
acid, THF=tetrahydrofuran, Ts=4-toluenesulfonyl.
Miyaura reaction was envisioned. Suzuki–Miyaura reaction of
16 using iodomethane did not proceed under the known
reaction conditions;[11] however, in the presence of thallium
ethoxide[12] the diene 17 was obtained in 74% yield.
Selective removal of the TBDPS protecting group of 17,
oxidation, and treatment with TMS diazomethane led to the
methyl ester 18, which was further converted into aldehyde 19
by selective removal of the THP protecting group[13] and
oxidation (Scheme 4). A Julia–Kocienski reaction[4] between
aldehyde 19 and sulfone 9 in the presence of potassium
hexamethyldisilazide proceeded stereoselectively in DMF to
yield the E olefin 20. The seco acid was obtained from 20
through hydrolysis, and it was converted into the correspond-
ing lactone under modified Yamaguchi reaction conditions.[14]
Subsequent removal of the TBS protecting group provided
the allylic alcohol 21. (À)-Amphidinolide K (1)[15] was
prepared in high yield by asymmetric epoxidation of 21 in
the presence of (+)-diethyl tartrate.
Scheme 4. Synthesis of (À)-amphidinolide K (1). a) TBAF, THF, 08C;
b) IBX, DMSO/THF (1:1); NaClO2, NaH2PO4, 2-methyl-2-butene/
tBuOH/H2O (1:1:1); c) TMSCHN2, MeOH; d) BF3·OEt2, EtSH/CH2Cl2
(1:5), À308C; e) DMP, CH2Cl2; f) 9, KHMDS, DMF, À788C, 30 min;
RT, 1 h; g) NaOH, MeOH/H2O (4:1); h) 2,4,6-Cl3C6H2COCl, TEA,
DMAP, toluene, reflux; i) TBAF, THF; j) (+)-DET, Ti(OiPr)4, TBHP,
M.S. (4 ꢀ), CH2Cl2, À208C. DET=diethyl tartrate, DMF=N,N-dime-
thylformamide, DMP=Dess–Martin periodinane, HMDS=1,1,1,3,3,3-
hexamethyldisilazane, IBX=o-iodoxybenzoic acid, M.S.=molecular
sieves, TBHP=tert-butyl hydroperoxide.
Received: October 28, 2008
Revised: January 23, 2009
Published online: February 18, 2009
The present synthesis represents a highly convergent
route to (À)-amphidinolide K (1) requiring 18 steps in the
longest linear sequence (6.8% total yield) from (S)-glycidol
(2). This synthesis presents another successful example of
stereoselective radical cyclization reactions of b-alkoxyacry-
lates.
Keywords: anticancer agents · macrolides · natural products ·
.
radical cyclization · total synthesis
Angew. Chem. Int. Ed. 2009, 48, 2364 –2366
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2365

Communications
[8] J. Nokami, M. Ohga, H. Nakamoto, T. Matsubara, I. Hussain, K.
Kataoka, J. Am. Chem. Soc. 2001, 123, 9168 – 9169.
[9] S. V. Ley, D. S. Brown, J. A. Clase, A. J. Fairbanks, I. C. Lennon,
H. M. I. Osborn, E. S. E. Stokes, D. J. Wadsworth, J. Chem. Soc.
Perkin Trans. 1 1998, 2259 – 2276.
[10] The analogous ring-closing metathesis reactions of the enyne
intermediates for production of macrolides like A failed, and
therefore the synthesis was abandoned.
[11] a) P. Kraft, W. Tochtermann, Liebigs Ann. Chem. 1994, 827 –
830; b) J. S. Clark, R. P. Freeman, M. Cacho, A. W. Thomas, S.
Swallow, C. Wilson, Tetrahedron Lett. 2004, 45, 8639 – 8642.
[12] For use of thallium ethoxide in Suzuki reactions, see: a) S. A.
Frank, H. Chen, R. K. Kunz, M. J. Schnaderbeck, W. R. Roush,
Org. Lett. 2000, 2, 2691 – 2694.
[1] a) M. Ishibashi, M. Sato, J. Kobayashi, J. Org. Chem. 1993, 58,
6928 – 6929; for recent reviews on amphidinolides, see: b) J.
Kobayashi, M. Tsuda, Nat. Prod. Rep. 2004, 21, 77 – 93; c) J.
Kobayashi, T. Kubota, J. Nat. Prod. 2007, 70, 451 – 460.
[2] D. R. Williams, K. G. Meyer, J. Am. Chem. Soc. 2001, 123, 765 –
766.
[3] a) D. R. Williams, K. G. Meyer, Org. Lett. 1999, 1, 1303 – 1305;
b) G. Mas, L. Gonzꢀlez, J. Vilarrasa, Tetrahedron Lett. 2003, 44,
8805 – 8809; c) T. Andreou, A. M. Costa, L. Esteban, L. Gonzꢀ-
lez, G. Mas, J. Vilarrasa, Org. Lett. 2005, 7, 4083 – 4086.
[4] a) P. R. Blakemore, W. J. Cole, P. J. Kocienski, A. Morley, Synlett
1998, 26 – 28; b) P. Liu, E. N. Jacobsen, J. Am. Chem. Soc. 2001,
123, 10772 – 10773.
[5] M. Kim, D. Lee, Org. Lett. 2005, 7, 1865 – 1868; for a recent
review on enyne metathesis reactions, see: S. T. Diver, A. J.
Giessert, Chem. Rev. 2004, 104, 1317 – 1382.
[6] a) E. Lee, J. S. Tae, C. Lee, C. M. Park, Tetrahedron Lett. 1993,
34, 4831 – 4834; b) M. A. Leeuwenburgh, R. E. J. N. Litjens,
J. D. C. Codꢁe, H. S. Overkleeft, G. A. van der Marel, J. H.
van Boom, Org. Lett. 2000, 2, 1275 – 1277.
[13] K. P. Nambiar, A. Mitra, Tetrahedron Lett. 1994, 35, 3033 – 3036.
[14] For use of the modified Yamaguchi reaction conditions, see:
D. A. Evans, H. P. Ng, D. L. Rieger, J. Am. Chem. Soc. 1993, 115,
11446 – 11475.
[15] [a]2D8 = À75.2 degcm3 gÀ1 dmÀ1
(c = 0.0006 gcmÀ3
,
MeOH)
([a]2D1 = À71 degcm3 gÀ1 dmÀ1 (c = 0.0005 gcmÀ3, MeOH)[1a]). It
should be noted that the NMR spectra of (+)-amphidinolide K
reported by Williams and Meyer[2] were actually recorded in
C6D6, and not in CDCl3 as reported.
[7] B. M. Trost, H. Yang, G. Wuitschik, Org. Lett. 2005, 7, 4761 –
4764.
2366 www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 2364 –2366