Total synthesis of (-)-callystatin A

Article abstract of DOI:10.1021/ol048664r

(Chemical Equation Presented) A convergent enantioselective synthesis of the natural product (-)-callystatin A (1) is described. Key features of the synthesis include a lipase-mediated kinetic resolution to install the C5 lactone stereochemistry, a hydrozirconation-based approach to the C8-C9 trisubstituted (Z)-olefin, and a stereoselective cross-coupling of a vinyl dibromide to install the C14-C15 trisubstituted (E)-olefin.

Full text of DOI:10.1021/ol048664r

ORGANIC
LETTERS
2004
Vol. 6, No. 18
3203-3206
Total Synthesis of (−)-Callystatin A
Neil F. Langille and James S. Panek*
Department of Chemistry and Center for Chemical Methodology and Library
DeVelopment, Metcalf Center for Science and Engineering, Boston UniVersity,
590 Commonwealth AVenue, Boston, Massachusetts 02215
panek@chem.bu.edu
Received July 12, 2004
ABSTRACT
A convergent enantioselective synthesis of the natural product (−)-callystatin A (1) is described. Key features of the synthesis include a
lipase-mediated kinetic resolution to install the C5 lactone stereochemistry, a hydrozirconation-based approach to the C8−C9 trisubstituted
(Z)-olefin, and a stereoselective cross-coupling of a vinyl dibromide to install the C14−C15 trisubstituted (E)-olefin.
(-)-Callystatin A (1) is a polyketide-based natural product
isolated in 1997 by Kobayashi and co-workers from the
marine sponge Callyspongia truncata.1 Initial biological
testing1,2 revealed remarkable antiproliferative properties
(IC50 values 10 and 20 pg/mL vs KB and L1210 cell lines,
respectively), derived from (-)-callystatin A’s ability to
interfere with nuclear export signal (NES) dependent protein
transport. Structure-activity relationship analysis indicates
that the δ-lactone is the principle pharmacophore, although
no enhancement of biological activity through analogue
synthesis has been reported.3 (-)-Callystatin A’s potent
cytotoxicity, along with its challenging structure, has stimu-
lated much attention, culminating in seven total syntheses
to date.4,5
transition-metal-mediated cross-coupling reactions would
enable assembly from three advanced intermediates at a late
stage in the synthesis (Scheme 1). Fragment 3 possesses both
a vinyl iodide equivalent (vinyl TMS) and a vinylstannane,
thereby allowing carbon-carbon bond formation at C13-
C14 and C7-C8 in a predetermined sequence. A hydrozir-
conation-iodination protocol developed in our laboratory
facilitates stereoselective installation of the C8-ethyl sub-
stituent in 3, where alkyne 6 serves as the appropriate starting
material.6 Diene RCM provides access to lactol 2, with C7-
C8 bond formation planned through a tandem hydrozircona-
tion-Negishi coupling protocol.7 Chiral organosilane-
mediated bond construction8 enables stereoselective synthesis
Our retrosynthetic analysis of (-)-callystatin A (1) relies
on bond disconnections at the two conjugated dienes, where
(4) (a) Murakami, N.; Wang, W.; Aoki, M.; Tsutsui, Y.; Sugimoto, M.;
Kobayashi, M. Tetrahedron Lett. 1998, 39, 2349-2352. (b) Crimmins, M.
T.; King, B. W. J. Am. Chem. Soc. 1998, 120, 9084-9085. (c) Smith, A.
B., III; Brandt, B. M. Org. Lett. 2001, 3, 1685-1688. (d) Kalesse, M.;
Quitschalle, M.; Khandavalli, C. P.; Saeed, A. Org. Lett. 2001, 3, 3107-
3109. (e) Marshall, J. A.; Bourbeau, M. P. J. Org. Chem. 2002, 67, 2751-
2754. (f) Marshall, J. A.; Bourbeau, M. P. Org. Lett. 2002, 4, 3931-3934.
(g) Vicario, J. L.; Job, A.; Wolberg, M.; Mu¨ller, M.; Enders, D. Org. Lett.
2002, 4, 1023-1026. (h) Lautens, M.; Stammers, T. A. Synthesis 2002,
1993-2012.
(5) For additional studies toward fragment synthesis, see: (a) Dias, L.
C.; Meira, P. R. R. Tetrahedron Lett. 2002, 43, 185-187. (b) Dias, L. C.;
Meira, P. R. R. Tetrahedron Lett. 2002, 43, 1593. (c) Dias, L. C.; Meira,
P. R. R. Tetrahedron Lett. 2002, 43, 8883-8885.
(6) Arefolov, A.; Langille, N. F.; Panek, J. S. Org. Lett. 2001, 3, 3281-
3284.
(1) Kobayashi, M.; Higuchi, K.; Murakami, N.; Tajima, H.; Aoki, S.
Tetrahedron Lett. 1997, 38, 2859-2862.
(2) (a) Murakami, N.; Sugimoto, N.; Tsutsui, Y.; Nakajima, T.; Higuchi,
K.; Aoki, S.; Yoshida, M.; Kudo, N.; Kobayashi, M. Abstracts of Papers,
41st Symposium on the Chemistry of Natural Products, Nagoya, October
1999, pp 229-234. (b) For a discussion of the biological activity of
(-)-callystatin A, leptomycin, and related compounds, see: Kalesse, M.;
Christmann, M. Synthesis 2002, 981-1003.
(3) (a) Murakami, N.; Sugimoto, M.; Nakajima, T.; Kawanishi, M.;
Tsutsui, Y.; Kobayashi, M. Bioorg. Med. Chem. 2000, 8, 2651-2661. (b)
Murakami, N.; Sugimoto, M.; Kobayashi, M. Bioorg. Med. Chem. 2001,
9, 57-67.
10.1021/ol048664r CCC: $27.50 © 2004 American Chemical Society
Published on Web 08/11/2004

in the presence of the transesterifying agent vinyl acetate
(Figure 1). As anticipated, 10 reacted under these conditions
Scheme 1. Retrosynthetic Analysis of (-)-Callystatin A (1)
Figure 1. Lipase Pseudomonas AK mediated kinetic resolution
of 13 (eq 1, ref 11) and 10 in the presence of vinyl acetate (eq 2)
and vinyl acrylate (eq 3).
of 4, with a Stille reaction proposed to form the C13-C14
bond.9 It is worthy to note that unlike other completed
syntheses of (-)-callystatin A, our approach does not include
a Still-Gennari reaction for the formation of the trisubsti-
tuted C8-C9 (Z)-alkene4a-h or a phosphorus-based olefina-
tion for the trisubstituted C14-C15 (E)-alkene.4a-c,g-h
Grignard reaction between allylmagnesiumbromide and 910
initiated the synthesis of fragment 2 (Scheme 2), providing
racemic propargyl alcohol 10 in excellent yield. Structural
similarities between 10 and the known allylic alcohol 1311
suggested that this substrate might participate in an enantio-
selective kinetic resolution with lipase Pseudomonas AK,12
to produce (R)-acetate 15 in 48% yield and >95% ee.13 With
this result in hand, we hypothesized that a similar reaction
in the presence of vinyl acrylate might resolve 10 with
concomitant installation of the requisite acrylate ester on the
desired (R)-enantiomer. Gratifyingly, this biocatalytic resolu-
tion proceeded with excellent stereoselectivity, providing
(R)-5 directly in a one-pot process in 46% yield and >95%
ee.13,14 Following chromatographic separation of 5 and (-)-
10, a Mitsunobu reaction transformed unreacted (-)-10 to
5 with complete stereochemical inversion. Ruthenium-
catalyzed RCM,15 lactone protection as the isopropoxy acetal,
and removal of the silicon protecting group completed the
synthesis of 2 in a six-step sequence in 51% overall yield.
The synthesis of fragment 3 from aldehyde 1616 com-
menced with a four-step sequence featuring a hydrozircona-
Scheme 2. Synthesis of Lactol 2a
(7) (a) Negishi, E. In Metal-Catalyzed Cross-Coupling Reactions;
Diederich, F., Stang, P. J., Eds.; Wiley-VCH: New York, 1998; Chapter
1. (b) For a related transformation, see: Jacobsen, E. N.; Chavez, D. E.
Angew. Chem., Int. Ed. 2001, 40, 3667-3670.
(8) For a review, see: Masse, C. E.; Panek, J. S. Chem. ReV. 1995, 95,
1293-1316.
(9) Shen, W.; Wang, L. J. Org. Chem. 1999, 64, 8873-8879.
(10) Journet, M.; Cai, D.; DiMichele, L. M.; Larsen, R. D. Tetrahedron
Lett. 1998, 39, 6427-6428.
(11) Sparks, M. A.; Panek, J. S. Tetrahedron Lett. 1991, 32, 4085-
4088.
(12) Pseudomonas AK lipase was purchased from Amano International
Enzyme Co.
(13) The ee of recovered (S)-10 was >95%. See the Supporting
Information for full characterization and the stereochemical assignments
of 5 and 15.
(14) For an example of lipase-mediated acryloyl transfer, see: Bisht, K.
S.; Gross, R. A.; Kaplan, D. L. J. Org. Chem. 1999, 64, 780-789.
(15) Grubbs, R. H.; Miller, S. J.; Fu, G. C. Acc. Chem. Res. 1995, 28,
446-452.
(16) Myers, A. G.; Yang, B. H.; Chen, H.; McKinstry, L.; Kopecky, D.
J.; Gleason, J. L. J. Am. Chem. Soc. 1997, 119, 6496-6511.
a Reagents, conditions, and yields: (a) allylmagnesium bromide,
THF, -20 °C, >99%; (b) vinyl acrylate, lipase AK, hexanes, 7
days, rt, 44% (-)-10, 46% 5; (c) DIAD, acrylic acid, PPh3, THF,
0 °C to rt, 86%; (d) Grubbs I, CH2Cl2, 40 °C, 83%; (e) DIBAL-H,
-78 °C, CH2Cl2; (f) i-PrOH, PPTS, PhH, 80 °C, 82% (two steps);
(g) 1.3:1 AcOH/TBAF, THF, rt, 91%.
3204
Org. Lett., Vol. 6, No. 18, 2004

tion-iodination method6 for the installation of the requisite
C8-ethyl substituent in the trans configuration (Scheme 3).
Scheme 4. Synthesis of Vinyl Dibromide 4a
Scheme 3. Synthesis of Vinyl Stannane 3a
a Reagents, conditions, and yields: (a) CBr4, PPh3, 2,6-lutidine,
CH2Cl2, 0 °C to rt, 96%; (b) n-BuLi, THF, -78 °C, then TMSCl,
-78 °C to rt, 98%; (c) Cp2ZrHCl, THF, 50 °C, 1 h, then I2, THF,
rt, 89%, >20:1 crude dr; (d) EtZnX, Pd(PPh3)4, THF, rt, 96%; (e)
DDQ, CH2Cl2/H2O, rt, 83%; (f) (COCl)2, DMSO, Et3N, CH2Cl2,
-78 °C; 92%; (g) CrCl2, Bu3SnCHI2, DMF, 0 °C to rt, 68%, E/Z
>20:1.
a Reagents, conditions, and yields: (a) (R)-7, TiCl4, CH2Cl2, -30
°C, 84%, 10:1 crude dr; (b) TBSOTf, 2,6-lutidine, CH2Cl2, 0 °C to
rt, 92%; (c) Ac2O, pyr, DMAP, CH2Cl2, rt, 99%; (d) O3, pyr,
MeOH/CH2Cl2, -78 °C, then Me2S; (e) (R)-7, TiCl4, CH2Cl2, -30
°C, 21a to 22a: 0%; 21b to 22b: 68% (two steps), >20:1 crude
dr; (f) CBr4, PPh3, 2,6-lutidine, CH2Cl2, 0 °C to rt, 82% (two steps);
(g) K2CO3, MeOH, rt, 99%; (h) TIPSOTf, 2,6-lutidine, CH2Cl2, 0
°C, 90%; (i) PCC, CH2Cl2, rt, 85%.
We have shown that upon exposure to an electrophilic iodine
source (I2, NIS), iododesilylation of 18 occurred with
retention of alkene configuration. However, we felt that
elaboration of the benzyl ether moiety into the corresponding
vinylstannane, prior to iodide formation and cross-coupling,
would represent a more convergent approach. Accordingly,
treatment of 18 with DDQ followed by Swern oxidation17
yielded aldehyde 19, which participated in a chromium(II)-
mediated vinylstannation18 with freshly prepared Bu3SnCHI2
to complete the preparation of 3 in seven steps from aldehyde
16.
Organosilane reagents such as 7 provide rapid access to
polypropionate fragments with high levels of selectivity. At
first glance, the all-syn C16-C20 stereochemistry in (-)-
callystatin A is a readily accessible target, with the (R)-
configuration of 7 directing the C16 and C18 methyl
stereocenters. Accordingly, our synthesis began with treat-
ment of aldehyde 819 with (R)-7 in the presence of TiCl4 to
yield homoallylic alcohol 20 as the syn-syn product in 84%
yield (Scheme 4). Protection of 20 as a silyl ether and
oxidative cleavage afforded aldehyde 21a, setting the stage
for a second crotylation reaction. However, exposure of 21a
to (R)-7 in the presence of TiCl4 produced homoallylic
alcohol 20 in >80% yield, with no trace of desired alcohol
22a. We rationalize that a Lewis acid promoted deprotec-
tion-retroaldol-crotylation sequence led to exclusive for-
mation of undesired 20. Indeed, the analogous reactions with
other acid-sensitive ether protecting groups at C19 (OBn,
OTES) yielded similar results.20 To alleviate this difficulty,
the acetate analogue 21b was synthesized and subjected to
crotylation conditions. Although 21b proved remarkably
unreactive, prolonged reaction time (-30 °C, 2 d) resulted
in formation of the desired homologated alcohol 22b in good
yield. Ozonolytic cleavage and dibromoolefination21 of the
free alcohol22 directly provided 23, possessing the fragment’s
full C14-C22 carbon skeleton. Concerned with removal of
the C19 acetate at a late stage, we carried out a two-step
protecting group exchange sequence to generate 25. Interest-
ingly, TIPS protection of diol 24 proceeded with complete
regioselectivity at C19, suggesting the different steric
environments of the two alcohol functionalities. Oxidation
of 25 using PCC completed the advanced fragment 4 in nine
steps and 32% overall yield from 8.
With access to the three advanced fragments, we envi-
sioned the rapid assembly of (-)-callystatin A beginning with
C13-C14 carbon-carbon bond formation (Scheme 5). A
Pd2dba3-mediated cross-coupling9 between stannane 3 and
dibromide 4 yielded vinyl bromide 26, with the desired (E,Z)-
diene as the only observed product isomer. Installation of
the C14-methyl group by the Negishi protocol23 proceeded
smoothly without epimerization or nucleophilic addition to
the ketone. Treatment of 27 with the electrophilic iodine
source NIS,24 however, resulted in undesired regioselective
iodination of the C14-C15 olefin. Close inspection of the
(17) Omura, K.; Swern, D. Tetrahedron 1978, 34, 1651-1660.
(18) (a) Hodgson, D. M.; Boulton, L. T.; Maw, G. N. Tetrahedron 1995,
51, 3713-3724. (b) Hodgson, D. M.; Foley, A. M.; Boulton, L. T.; Lovell,
P. J.; Maw, G. N. J. Chem. Soc., Perkin Trans. 1 1999, 2911-2922.
(19) White, J. D.; Bolton, D. L.; Dantanarayana, A. P.; Fox, C. M. J.;
Hiner, R. N.; Jackson, R. W.; Sakuma, K.; Warrier, U. S. J. Am. Chem.
Soc. 1995, 117, 1908-1939.
(21) Corey, E. J.; Fuchs, P. L. Tetrahedron Lett. 1972, 13, 3769-3772.
(22) Unusual reactivity of similar intermediates has been reported. See
refs 4a,e,f and 5b for details.
(23) For application in related ene-yne systems, see: Shi, J.; Zeng, X.;
Negishi, E. Org. Lett. 2003, 5, 1825-1828.
(24) Stamos, D. P.; Taylor, A. G.; Kishi, Y. Tetrahedron Lett. 1996, 37,
8647-8650.
(20) Hydroxy aldehyde (21, R ) H) displayed similar reactivity.
Org. Lett., Vol. 6, No. 18, 2004
3205

Scheme 5. Assembly and Completion of (-)-Callystatin Aa
Figure 2. 1H NMR chemical shifts (in ppm) as a measure of
relative alkene electronegativity in 27 and 26. Upon exposure to
NIS, highlighted alkenes undergo regioselective electrophilic io-
dination.
produced a vinylzinc species that participated in a Pd-
catalyzed Negishi coupling with 29 to afford the protected
natural product 30 in 51% yield.26 Completion of the
synthesis proceeded through AcOH-promoted lactol depro-
tection, oxidation, and fluoride-mediated removal of the silyl
ether (65% over three steps). Synthetic (-)-callystatin A
possesses spectroscopic properties identical to those reported
for the natural product.1,4
In summary, we have developed an effective synthetic
strategy for the total synthesis of (-)-callystatin A, utilizing
cross-coupling reactions for the union of three highly
functionalized fragments. The approach demonstrates the
effectiveness of selective dibromide cross-coupling reactions
and features a complimentary hydrozirconation-iodination
approach for the stereoselective synthesis of trisubstituted
alkenes. These methods should find utility in the synthesis
of other related natural and unnatural molecules of biological
importance.
a Reagents, conditions, and yields: (a) Pd2dba3, P(2-fur)3, PhMe,
100 °C, 92%; (b) Me2Zn, Pd(t-Bu3P)2, THF, 0 °C, 26 to 27: 96%;
28 to 29: 93%; (c) NIS, EtCN, 26 to 28: 84%; (d) (i) Cp2ZrHCl,
THF, rt, (ii) ZnCl2, THF, (iii) 29, Pd(PPh3)4, rt, 51%; (e) AcOH,
wet THF, rt; (f) PDC, CH2Cl2, rt, 74% (two steps); (g) HF/pyr,
THF, rt, 88%.
1H NMR spectrum of 27 (Figure 2) shows the electron-rich
nature of the C14-C15 alkene relative to the C8-C9 olefin,
explained by the strong donor capability of the C14-methyl
substituent. Comparison to the 1H NMR spectrum of bromide
26, however, revealed the possibility of successful iodo-
desilylation on this alternate substrate. Indeed, exposure of
26 to NIS resulted in regioselective addition of I+ to the C8
carbon, with retention of alkene stereochemistry. Iodo
bromide 28 underwent regioselective bromide cross-coupling
upon treatment with Me2Zn in the presence of Pd catalyst
to form 29. We postulate that the electron-donating effect
of the C8-ethyl group may deactivate the vinyl iodide,
providing entry into oxidative addition of the relatively
electron-deficient C14-bromide.25 In summary, this sequence
provided fully functionalized iodide 29 in three steps from
stannane 3 and vinyl dibromide 4. Subjection of 2 to
Schwartz’s reagent followed by in situ treatment with ZnCl2
Acknowledgment. This work has been financially sup-
ported by the NIH (GM55740). J.S.P. is grateful to Johnson
& Johnson, Merck Co., Amgen, Inc., Novartis, Pfizer, and
GlaxoSmithKline for financial support.
Supporting Information Available: Spectroscopic data
and experimental procedures for all new compounds; key
spectra. This material is available free of charge via the
Internet at http://pubs.acs.org.
OL048664R
(26) Reproducibly, iodide 29 was recovered in 20-30% yield from this
reaction. Attempts to increase iodide conversion to desired coupling product
were unsuccessful.
(25) For a discussion of electronic effects in Pd-mediated cross-couplings,
see: Grushin, V. V.; Alper, H. Chem. ReV. 1994, 94, 1047-1062.
3206
Org. Lett., Vol. 6, No. 18, 2004