Synthesis of (-)-octalactin A by a strategic vanadium-catalyzed oxidative kinetic resolution

Article abstract of DOI:10.1002/anie.200800554

(Chemical Equation Presented) Both products of the kinetic resolution were used in the resolution/recombination approach shown in the scheme for the enantioselective total synthesis of (-)-octalactin A. Bn=benzyl, TBS=tert-butyldimethylsilyl.

Full text of DOI:10.1002/anie.200800554

Angewandte
Chemie
DOI: 10.1002/anie.200800554
Kinetic Resolution
Synthesis of (ꢀ)-Octalactin A by a Strategic Vanadium-Catalyzed
Oxidative Kinetic Resolution**
Alexander T. Radosevich, Vincent S. Chan, Hui-Wen Shih, and F. Dean Toste*
The kinetic resolution of racemic material is a well-estab-
lished approach used to prepare a wide range of enantiomer-
ically enriched compounds.[1] However, the overall efficiency
of a standard kinetic resolution is limited to a maximum
theoretical yield of 50% of the enriched chiron, with the
balance of undesired material being discarded in most
applications.[2] As a result of this inherent overall efficiency,[3]
kinetic resolution is often deemed unacceptable on a prep-
arative scale except for the most simple and inexpensive of
substrates. Consequently, kinetic resolution methods, espe-
cially in the context of complex molecule synthesis, have
largely been relegated to the preparation of small chiral
building blocks at an early stage of the synthesis.
Having recently developed a vanadium-catalyzed asym-
metric aerobic oxidation of a hydroxycarbonyl compounds,[4]
we became interested in exploring the potential of this kinetic
resolution methodology, to move beyond the preparation of
simple building blocks toward a more meaningful, strategic
synthetic function. The observation that asymmetric oxida-
tion of substrates bearing multiple stereocenters by catalytic
amounts of [VO(OiPr)3] and (S)-2-(3,5-di-tert-butylsalicyli-
deneamino)-tert-butyl-1-ethanol (1) resulted in the isolation
of both enantioenriched alcohol and ketone products moti-
vated us to consider a synthetic strategy (Figure 1) wherein
point in the synthetic sequence and the overall efficiency of
the resolution is improved beyond 50%. Herein, we report
the successful implementation of this type of resolution/
recombination synthetic strategy[5] to the total synthesis of the
bioactive marine metabolite octalactin A.[6]
Octalactin A (2) belongs to a small class of medium-ring
lactones isolated from marine microorganisms. In vitro bio-
assays have shown octalactin A to possess significant cyto-
toxicity toward B-16-F10 murine melanoma and HCT-116
human colon tumor cell lines.[6,7] As a result, numerous
synthetic investigations, including several total syntheses,[8]
have been directed towards its preparation.[9] Our approach
to octalactin A begins in an antithetic sense with disconnec-
tion of the C10–C15side chain and retrolactonization of the
eight-membered core, thus leading to the simplified frag-
ments 3 and 4 (Scheme 1). From here, the basis of our strategy
Figure 1. Resolution/recombination synthetic pathway.
both components of the resolution might be leveraged toward
a single synthetic target. By doing so, the kinetic resolution
serves as the key structural and stereochemical branching
Scheme 1. Retrosynthetic analysis of octalactin A. Bn=benzyl,
TBS=tert-butyldimethylsilyl, PG=protecting group.
stems from the identification of an element of pseudosym-
metry in the seco-acid component 3. Specifically, we envi-
sioned that the repeating anti-1,2-vic-hydroxymethyl stereo-
diad could be derived from the enantiomers of racemic anti-
1,2-vic-hydroxymethyl stereoconstruct (ꢁ )-5. As such, oxi-
dative resolution of (ꢁ )-5 and subsequent head-to-tail
recombination of elaborated fragments would allow rapid
access to 3. The C10–C15side chain 4 could be obtained by
asymmetric addition of an isopropyl group to a suitably
functionalized aldehyde, wherein the trisubstituted olefin
geometry might be secured by stereospecific cuprate-induced
dyotropic rearrangement of 2,3-dihydrofuran.
[*] A. T. Radosevich, V. S. Chan, H.-W. Shih, Prof. F. D. Toste
Department of Chemistry
University of California, Berkeley
Berkeley, CA94720-1460 (USA)
Fax: (+1)510-643-9480
E-mail: fdtoste@berkeley.edu
[**] We gratefully acknowledge the University of California, Berkeley,
NIHGMS (grant no. R01 GM074774), Merck Research Laboratories,
Bristol-Myers Squibb, Amgen Inc., and Novartis for financial
support. We thank Prof. E. Vedejs and Dr. S. T. Staben for helpful
discussions.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2008, 47, 3755 –3758
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3755

Communications
Preparation of (ꢁ )-5 began with a
Darzens condensation[10] of benzyl chloro-
acetate (6) and acetone followed by an
acid-catalyzed rearrangement of the inter-
mediate glycidic ester to give hydroxy
ester 7 in good yield (Scheme 2). Treat-
ment of 7 with 9-BBN resulted in a
diastereoselective (d.r. > 10:1) hydrobo-
ration of the double bond of 7 to afford
the desired anti stereodiad, in accord with
the model proposed by Still and Bar-
rish.[11,12] Following selective monosilyla-
tion of the unstable intermediate diol and
chromatographic removal of the minor
Scheme 3. Elaboration and recombination of resolution fragments (2S,3S)-5 and (R)-8.
=
Reagents and conditions: a) 2.5 equiv PMBO(C NH)CCl3, 1 mol% Ph3CBF4, Et2O, 08C, 87%;
b) AcOH/THF/H2O (3:1:1), RT, 97%; c) 1.5 equiv Dess–Martin periodinane, CH2Cl2, RT, 91%;
d) 1 atm H2, Pd/C, EtOAc, RT, quant.; e) 1. 1.1 equiv PhIO, THF, RT; 2. EtO(CO)Cl; then Et3N,
THF, RT; 3. LiCH2P(O)(OMe)2, 55% (over 3 steps); f) 0.6 equiv BaO, 1.2 equiv H2O, Et2O,
syn diastereomer, key building block 08C, 82% (>20:1 E/Z). PMB=para-methoxybenzyl.
(ꢁ)-5 was obtained.
At this juncture, our resolution/recombination strategy
called for the merger of aldehyde 9 and ketophosphonate 10.
A modification of the barium hydroxide promoted method[16]
proved effective in promoting the olefination reaction with-
out undesired saponification of the product benzyl ester. In
this way, alkene 11 was obtained in 82% yield with high E/Z
selectivity (> 20:1) and no observable epimerization of the
stereocenters.
Having now recombined the resolution fragments, we set
about advancing toward the lactone 16 (Scheme 4). A reagent
controlled reduction (with (+)-Ipc2BCl)[17] of ketone 11
delivered the desired anti diastereomer 12 in good yield
(80%) and diastereoselectivity (d.r. 83:17).[18] Heterogeneous
hydrogenation with an ethylenediamine-modified palladium
Scheme 2. Synthesis and resolution of (ꢁ)-5. Reagents and condi-
tions: a) tBuOK, acetone, THF, ꢀ788C, 95%; b) 10-camphorsulfonic
acid, toluene, 1108C, 90%; c) 9-BBN, THF, RT; then mCPBA, 67%;
d) TBSCl, imidazole, CH2Cl2, 85%; e) 5.5 mol% (R)-1, 5 mol% [VO-
(OiPr)3], 1 atm O2, acetone, 358C. 9-BBN=9-borabicyclo[3.3.1]nonane,
mCPBA=meta-chloroperbenzoic acid.
With multigram quantities of (ꢁ )-5 in hand, the stage was
set for the pivotal asymmetric oxidation. The use of a solution
of ligand (R)-1 in acetone at 408C under one atmosphere of
oxygen resulted in the asymmetric oxidation proceeding to
approximately 50% conversion over the course of 24 h and
gave high levels of enantiocontrol at all three stereocenters.[13]
Good mass recovery of optically active alcohol (2S,3S)-5 and
ketone (R)-8 was obtained after column chromatography. The
absolute configuration of the alcohol component (2S,3S)-5
was readily determined by analysis of the Mosher ester
derivative,[14] which confirmed the configuration required for
naturally occurring octalactin A.
Both the alcohol ((2S,3S)-5) and the ketone ((R)-8)
components could now be advanced independently along
parallel paths toward fragment coupling partners 9 and 10
(Scheme 3). Specifically, (2S,3S)-5 was subjected to a conven-
tional three-step protection/oxidation sequence to provide
the aldehyde 9 in good yield. With respect to (R)-8, after
hydrogenolysis of the benzyl ester, conversion of the resulting
keto acid into ketophosphonate 10 was effected by a one-pot
procedure, which involved excision of the carboxylate by
iodosobenzene-promoted oxidative decarboxylation,[15] for-
mation of the ethyl chloroformate derived mixed anhydride,
and addition of lithio dimethylmethanephosphonate.
Scheme 4. Formation of lactone 16. Reagents and conditions: a) (+)-
Ipc2BCl, Et2O, ꢀ208C, 80% (d.r. 83:17); b) 1 atm H2, Pd/C(en),
MeOH, quant.; c) TESCl, Et3N, DMAP, CH2Cl2, RT, 84%; d) EtO-
(CO)Cl, Et3N, CH2Cl2, ꢀ788C; then CH2N2, Et2O, RT, 74%; e) AcOH,
MeOH, RT, 87%; f) 5 equiv AgOBz, 10 equiv DMAP, THF (0.005m),
RT, 26%; g) 1. THF, H2O, hn=254 nm; 2. Bz2O, DMAP, CH2Cl2
(0.002m), RT, 67% (over 2 steps); h) TBAF, AcOH, THF, RT, 90%;
i) Dess–Martin periodinane, CH2Cl2, RT, quant. Bz=benzoyl, DMAP=
4-dimethylaminopyridine, en=ethylenediamine, Ipc=isopinocam-
pheyl, TBAF=tetra-n-butylammonium fluoride, TES=triethylsilyl.
3756
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 3755 –3758

Angewandte
Chemie
catalyst[19] resulted in clean saturation of the C5 C6 double
ꢀ
bond with concomitant removal of the benzyl ester group and
gave a quantitative yield.
In preparation for installation of the final C2 methylene
unit of the octalactin lactone core by way of an Arndt–Eistert
homologation,[20] acid 13 was converted into the a diazoke-
tone 14 by a conventional three-step sequence. With the
homologation precursor in hand, we considered that the
Wolff rearrangement[21] of diazoketone 14 might be conven-
iently coupled to the lactonization process if the transient
ketene intermediate were directly engaged by the free C7
hydroxy group. Indeed, treatment of diazoketone 14 with
silver benzoate and DMAP in dilute tetrahydrofuran afforded
lactone 15 directly, albeit in a modest 26% yield. Alterna-
tively, the transformation (14!15) can be efficiently accom-
plished in two synthetic steps: photolytic Wolff rearrange-
ment of 14 in aqueous tetrahydrofuran provided ready access
to seco-acid 3, which can then be treated with benzoic
anhydride to induce lactonization.[8e,22] Desilylation and
oxidation of lactone 15 gave aldehyde 16, the functionalized
octalactin core.
Scheme 6. Completion of the (ꢀ)-octalactin Asynthesis. Reagents and
conditions: a) CrCl2, NiCl2, DMSO, RT, 70% (d.r. 1.05:1); b) 25 mol%
[VO(acac)2], tBuOOH, benzene, RT, 84%; c) Dess–Martin periodinane,
CH2Cl2, RT, 69%; d) DDQ, CH2Cl2, H2O, RT, quant. acac=acetylaceto-
nate, DDQ=2,3-dichloro-5,6-dicyano-1,4-benzoquinone, DMSO=di-
methyl sulfoxide.
This report illustrates the application of vanadium-cata-
lyzed oxidative kinetic resolution to the total synthesis of
octalactin A. Contrary to the typical application of kinetic
resolution at the early stages of total syntheses, the vanadium-
catalyzed methodology plays a significant strategic role in the
overall design and execution of this synthesis. Moreover,
whereas standard kinetic resolutions are limited to a 50%
yield of the enriched chiron, the resolution/recombination
strategy employed here effectively uses both products of the
asymmetric operation. We note that insofar as any asymmet-
ric operation performed on racemic starting material might
With respect to the C10–C15side chain, iodoaldehyde
19[23] (prepared in three steps from dihydrofuran by using 1,2-
metalate rearrangement methodology;[24] Scheme 5) was
serve
a similar function, this resolution/recombination
approach could in principle be applicable to a wide range of
synthetic targets.
Received: February 2, 2008
Published online: April 10, 2008
Keywords: alcohol oxidation · kinetic resolution · lactones ·
.
total synthesis · vanadium
[1] For general discussions on kinetic resolution, see: a) H. B.
Kagan, J. C. Fiaud in Topics in Stereochemistry, Vol. 18 (Ed.:
E. L. Eliel), Wiley, New York, 1988, pp. 249 – 330; b) J. M. Kieth,
J. F. Larrow, E. N. Jacobsen, Adv. Synth. Catal. 2001, 343, 5 – 26;
c) A. H. Hoveyda, M. T. Didiuk, Curr. Org. Chem. 1998, 2, 489 –
526.
Scheme 5. Synthesis of side chain 20. Reagents and conditions:
a) 1. tBuLi; 2. Li2[Bu3Sn(Bu)CuCN]; 3. MeI, THF, 69%; b) I2, Et2O, RT,
97%; c) Dess–Martin periodinane, CH2Cl2, RT, 94%; d) 0.75 equiv
diisopropenylzinc, 1.5 equiv Et2Zn, 15 mol% 21, toluene, ꢀ408C, 68%
=
(83% ee); e) 1000 psi H2, [RhCl(Ph3P)3], toluene, 75%; f) PMBO(C
NH)CCl3, 15 mol% CSA, CH2Cl2, 70%. CSA=camphorsulfonic acid.
[2] More specialized kinetic resolutions fall outside the purview of
this restriction. For a review of dynamic kinetic resolutions, see:
H. Pellisier Tetrahedron 2003, 59, 8291 – 8327)
identified as a suitably functionalized intermediate. Installa-
tion of the terminal asymmetric isopropyl unit was achieved
by catalytic isopropenylation of 19 with diisopropenylzinc[25]
and ligand 21,[26] followed by chemoselective hydrogenation
of the newly installed 1,1-disubstituted olefin with Wilkin-
sonꢀs catalyst. Protection of the free alcohol set the stage for
subsequent manipulation.
[3] For an insightful discussion of efficiency as it pertains to kinetic
resolution, see: E. Vedejs, M. Jure, Angew. Chem. 2005, 117,
4040 – 4069; Angew. Chem. Int. Ed. 2005, 44, 3974 – 4001.
[4] a) A. T. Radosevich, C. Musich, F. D. Toste, J. Am. Chem. Soc.
2005, 127, 1090 – 1091; b) A. Blanc, F. D. Toste, Angew. Chem.
2006, 118, 2150 – 2153; Angew. Chem. Int. Ed. 2006, 45, 2096 –
2099; for a related catalysts system, see: c) S.-S. Weng, M.-W.
Shen, J.-Q. Kao, Y. S. Munot, C.-T. Chen, Proc. Natl. Acad. Sci.
USA 2006, 103, 3522 – 3527.
[5] For a related resolution approach to leucascandrolide A, see: Q.
Su, L. A. Dakin, J. S. Panek, J. Org. Chem. 2007, 72, 2 – 24.
[6] D. M. Tapiolas, M. Roman, W. Fenical, T. J. Stout, J. Clardy, J.
Am. Chem. Soc. 1991, 113, 4682 – 4683.
[7] J. P. Perchellet, E. M. Perchellet, S. W. Newell, J. A. Freeman,
J. B. Ladesich, Y. M. Jeong, N. Sato, K. Buszek, Anticancer Res.
1998, 18, 97 – 106.
[8] a) K. R. Buszek, N. Sato, Y. M. Jeong, J. Am. Chem. Soc. 1994,
116, 5511 – 5512; b) J. C. McWilliams, J. Clardy, J. Am. Chem.
Soc. 1994, 116, 8378 – 8379; c) K. R. Buszek, N. Sato, Y. Jeong,
In accord with previous synthetic reports,[8a,d] we effected
the union of aldehyde 16 and vinyl iodide 20 by using Nozaki–
Hiyama–Kishi coupling (Scheme 6). Although the diastereo-
selectivity of this reaction was disappointing, the diastereo-
meric alcohols could readily be separated by chromatography
on silica gel. Vanadium-catalyzed diastereoselective epoxida-
ꢀ
tion of a-22 secured the desired C10 C11 configuration. Final
oxidation of the C9 alcohol and cleavage of the PMB ether
groups completed the total synthesis to yield (ꢀ)-octalactin A
(2).[27]
Angew. Chem. Int. Ed. 2008, 47, 3755 –3758
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3757

Communications
Tetrahedron Lett. 2002, 43, 181 – 184; d) P. T. OꢀSullivan, W.
[18] Luche conditions (NaBH4, CeCl3, MeOH, ꢀ788C) gave good
yield but with poor diastereoselectivity (ca. 1.7:1) in favor of the
undesired anti-Felkin–Anh product (7-epi-12).
[19] H. Sajiki, K. Hattori, K. Hirota, J. Org. Chem. 1998, 63, 7990 –
7992.
[20] W. E. Bachmann, W. S. Struve, Org. React. 1942, 1, 38 – 62.
[21] a) G. B Gill in Comprehensive Organic Synthesis, Vol. 3 (Eds.:
B. M. Trost, I. Fleming), Pergamon, New York, 1991, pp. 887 –
912; b) W. Kirmse, Eur. J. Org. Chem. 2002, 2193 – 2256.
[22] I. Shiina, M. Kubota, H. Oshiumi, M. Hashizume, J. Org. Chem.
2004, 69, 1822 – 1830.
[23] a) L. Commeiras, J. Bourdron, S. Douillard, P. Barbier, N.
Vanthuyne, V. Peyrot, J.-L. Parrain, Synthesis 2006, 166 – 181;
b) L. Commeiras, M. Santelli, J.-L. Parrain, Org. Lett. 2001, 3,
1713 – 1715.
[24] a) P. Kocienski, S. Wadman, K. Cooper, J. Am. Chem. Soc. 1989,
111, 2363 – 2365; b) P. Kocienski in Organic Synthesis via
Organometallics (Eds.: D. Enders, H.-J. Gais, W. Klein),
Vieweg, Wiesbaden, 1993, pp. 203 – 223; c) T. Fujisawa, Y.
Kurita, M. Kawashima, T. Sato, Chem. Lett. 1982, 1641 – 1642.
[25] T. Shibata, K. Nakatsui, K. Soai, Inorg. Chim. Acta 1999, 296,
33 – 36.
[26] J. X. Ji, L. Q. Qiu, C. W. Yip, A. S. C. Chan, J. Org. Chem. 2002,
68, 1589 – 1590.
[27] The spectroscopic data for synthetic (ꢀ)-octalactin A are in full
agreement with those of the natural material. See the Supporting
Information for complete procedural and analytical details for
all compounds.
Buhr, M. A. M. Fuhry, J. R. Harrison, J. E. Davies, N. Feeder,
D. R. Marshall, J. W. Burton, A. B. Holmes, J. Am. Chem. Soc.
2004, 126, 2194 – 2207; e) I. Shiina, M. Hashizume, Y. Yamai, H.
Oshiumi, T. Shimazaki, Y. Takasuna, R. Ibuka, Chem. Eur. J.
2005, 11, 6601 – 6608.
[9] For additional synthetic studies on octalactin, including formal
syntheses, see: I. Shiina, Chem. Rev. 2007, 107, 239 – 273.
[10] a) G. Darzens, C. R. Hebd. Seances Acad. Sci. 1911, 151, 883 –
884; b) T. Rosen in Comprehensive Organic Synthesis, Vol. 2
(Eds.: B. M. Trost, I. Fleming), Pergamon, New York, 1991,
pp. 409 – 441.
[11] W. C. Still, J. C. Barrish, J. Am. Chem. Soc. 1983, 105, 2487 –
2489.
[12] Optimum conditions for the oxidation of the intermediate
alkylborane involve a peracid oxidation; use of more traditional
oxidation protocols (H2O2, NaOH, or NaBO3) resulted in the
formation of an undesired g lactone.
[13] See the Supporting Information for details.
[14] I. Ohtani, T. Kusumi, Y. Kashman, H. Kakisawa, J. Am. Chem.
Soc. 1991, 113, 4092 – 4096.
[15] R. M. Moriarty, S. C. Gupta, H. Hu, D. R. Berenschot, K. B.
White, J. Am. Chem. Soc. 1981, 103, 686 – 688.
[16] a) I. Paterson, K.-S. Yeung, J. B. Smaill, Synlett 1993, 774 – 776;
b) C. A. Ibarra, S. Arias, M. J. Fernµndez, J. V. Sinisterra, J.
Chem. Soc. Perkin Trans. 2 1989, 503 – 508.
[17] a) H. C. Brown, P. V. Ramachandran, Acc. Chem. Res. 1992, 25,
16 – 24; b) M. M. Midland, Chem. Rev. 1989, 89, 1553 – 1561.
3758
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 3755 –3758