Synthesis of the spirastrellolide A trioxadispiroketal

Article abstract of DOI:10.1021/ol0609457

The core trioxadispiroketal system representing C26-C40 of spirastrellolide A was assembled using a sequenced double-intramolecular hetero-Michael addition process.

Full text of DOI:10.1021/ol0609457

ORGANIC
LETTERS
2006
Vol. 8, No. 14
2997-3000
Synthesis of the Spirastrellolide A
Trioxadispiroketal
Ce Wang and Craig J. Forsyth*
Department of Chemistry, UniVersity of Minnesota, 207 Pleasant Street SE,
Minneapolis, Minnesota 55414
forsyth@chem.umn.edu
Received April 19, 2006
ABSTRACT
The core trioxadispiroketal system representing C26−C40 of spirastrellolide A was assembled using a sequenced double-intramolecular hetero-
Michael addition process.
Spirastrellolide A (Figure 1) is a polyketide isolated from
extracts of the Caribbean sponge Spirastrella coccinea.1 This
arrest. The methyl ester of spirastrellolide A also exhibited
potent tubulin-independent activity in a cell-based antimitotic
assay.1a As a structurally novel entrant to the okadaic acid
class of phosphatase inhibitors,4 spirastrellolide A and its
derivatives may have therapeutic potential.
The complete stereochemistry of spirastrellolide A is
undefined, but the relative configurations of three separate
portions were reported.1b Several individual fragments have
also been synthesized.5 We were initially interested in
exploring the double-intramolecular hetero-Michael addition
(DIHMA) process6 for the synthesis of spirastrellolide A’s
trioxadispiroketal domain 1 (Scheme 1).7 This would rep-
resent an extension of this synthetic methodology following
its success in generating the azaspiracid trioxadispiroketal
domain.8
Trioxadispiroketal 1 (Scheme 1) was targeted in an
arbitrary absolute configuration. Establishment of both C31
Figure 1. Two possible diastereomers of spirastrellolide A.
(3) Boger, D. L.; Lewy, D. S.; Gauss, C. M.; Soenen, D. R. Curr. Med.
Chem. 2002, 9, 917.
(4) Sassa, T.; Richter, W. W.; Uda, N.; Suganuma, M.; Suguri, H.;
Yoshizawa, S.; Hirota, M.; Fujiki, H. Biochem. Biophys. Res. Comm. 1989,
159, 939.
(5) (a) Liu, J.; Hsung, R. P. Org. Lett. 2005, 7, 2273. (b) Paterson, I.;
Anderson, E. A.; Dalby, S. M.; Loiseleur, O. Org. Lett. 2005, 7, 4121. (c)
Paterson, I.; Anderson, E. A.; Dalby, S. M.; Loiseleur, O. Org. Lett. 2005,
7, 4125. (d) Pan, Y.; De Brabander, J. K. Synth. Lett. 2006, 6, 853.
(6) Forsyth, C. J.; Hao, J.; Aguaide, J. Angew. Chem., Int. Ed. 2001, 40,
3663.
(7) Wang, C.; Forsyth, C. J. Abstracts of Papers, 229th National Meeting
of the American Chemical Society, San Diego, Mar 13-17, 2005; American
Chemical Society: Washington, DC, 2005; ORGN-414.
natural product inhibits protein phosphatases 1 and 2A with
IC50 values of 50 and 1 nM, respectively. The activities of
spirastrellolide A1b are similar to those of known protein
phosphatase inhibitors, including okadaic acid2 and fostrie-
cin,3 which induce premature entry into mitosis and mitotic
(1) (a) Williams, D. E.; Roberge, M.; Van Soest, R.; Andersen, R. J. J.
Am. Chem. Soc. 2003, 125, 5296. (b) Williams, D. E.; Lapawa, M.; Feng,
X.; Tarling, T.; Roberge, M.; Andersen, R. J. Org. Lett. 2004, 6, 2607.
(2) Dounay, A. B., Forsyth, C. J. Curr. Med. Chem. 2002, 9, 1951.
10.1021/ol0609457 CCC: $33.50
© 2006 American Chemical Society
Published on Web 06/15/2006

Scheme 1
Scheme 2a
and C35 spiroketals in 1 under thermodynamic control was
expected to deliver the natural product’s configurational and
conformational array due to double-anomeric stablization9
and the predominant equatorial orientation of its appendages.
Further, the C28 chlorine would be installed at a late stage
onto the assembled trioxadispiroketal 2 by selective R-keto
chlorination. Bis-spiroketal 2 would arise from linear ynone
3 via a DIHMA process.6 Ynone 3, in turn, would be derived
from the C37-C40 aldehyde 4, C30-C36 methyl ketone 5,
and the C26-C29 aldehyde 6. These units could be
combined sequentially to construct the linear ynone 3 by
Mukaiyama aldol10 and NHK11 coupling reactions.
The synthesis began with ketone 5 (Scheme 2).12 Kinetic
enolization followed by trapping with TMSCl provided the
corresponding silyl enol ether. This was coupled with the
L-malic acid-derived aldehyde 4 by a chelation-controlled
Mukaiyama aldol reaction10 to yield 7. This was originally
protected as a TBS ether; however, the silyloxy group
underwent â-elimination at later stages. Therefore, benzyl
ether formation led to 8 with subsequent manipulations.
Silver trifluoroacetate and N-iodosuccinimide were applied
to convert alkyne 8 into iodide 9.13 NHK coupling11 of 9
with L-malic acid-derived aldehyde 6 gave a propargylic
alcohol that was oxidized to ynone 3.
a DMP ) Dess-Martin periodinane.
cyclize derivatives of 3 under acidic reaction conditions
resulted in â-elimination of the C37 oxygen functionalities.
Various Bro¨nsted and Lewis acids were examined with
different hydroxyl protecting groups at the C27 and C37
alcohols. A benzyl ether at C37 seemed to be the least labile
toward elimination.
It was apparent that the rates of elimination needed to be
suppressed, while the rates of conjugate addition needed to
be enhanced, to accomplish the desired trioxadispiroketal
formation with this type of labile substrate. Accordingly, the
C27 and C38 hydroxyls were liberated from 3 at low
temperature, and the resultant acid labile diol 10 was
subjected to alkyne activation using various metal salts.
Several silver, copper, mercury, and palladium salts were
Our initial intent was to liberate the C27 and C38
hydroxyls from 3 under acidic conditions, whereupon bis-
conjugate addition might ensue to form the trioxadispiroketal.
This would involve successive 6-exo and 6-endo ring closures
of the C35 hemiketal and C27 hydroxyls, respectively, upon
the C29-C31 Michael acceptor.6 However, all attempts to
Scheme 3
(8) Geisler, L. K.; Nguyen, S.; Forsyth, C. J. Org. Lett. 2004, 6, 4159.
(9) Kirby, A. J. The Anomeric Effect and Related Stereoelectronic Effects
at Oxygen; Springer-Verlag: New York, 1983.
(10) Fujisawa, H.; Sasaki, Y.; Mukaiyama, T. Chem. Lett. 2001, 3, 190.
(11) (a) Jin, H.; Uenishi, J. I.; Christ, W. J.; Kishi, Y. J. Am. Chem. Soc.
1986, 108, 5644. (b) Takai, K.; Tagashira, M.; Kuroda, T.; Oshima, K.;
Utimoto, K.; Nozaki, H. J. Am. Chem. Soc. 1986, 108, 6084.
(12) Paquette, L. A.; Duan, M.; Konetzki, I.; Kempmann, C. J. Am. Chem.
Soc. 2002, 124, 4257.
(13) Balog, A.; Meng, D.; Kamenecka, T.; Bertinato, P.; Su, D. Angew.
Chem. 1996, 108, 2976.
2998
Org. Lett., Vol. 8, No. 14, 2006

Scheme 4
surveyed, and the results were mixed. Either a single hetero-
desired R-chloro aldehyde 14. A dimethoxybenzyl group was
strategically installed at the C27 hydroxyl to facilitate
subsequent selective removal.16
Michael addition was accompanied by â-elimination (Scheme
3) and furan formation (11), or no reaction occurred at all.
Next, basic reaction conditions were examined to cyclize
diol 10 (Scheme 4). Upon treatment of 10 with potassium
tert-butoxide, the C38-35 hemiketal conjugatively added to
the C29-C31 ynone to form spiroketal 12. This occurred
without elimination of either C27 or C37 oxygen substituents.
Efficient formation of the second spiroketal required acidic
reaction conditions. Hence, treatment of enone 12 with CSA
in benzene afforded bis-spiroketal 2. It was presumed that
the thermodynamically favored spiroketal configurations at
C31 and C35 emerged from the sequenced base- and acid-
induced assembly of the trioxadispiroketal 2. Purification of
ketone 2 was challenging due to its instability on silica gel.
A retro-hetero-Michael reaction occurred during chroma-
tography. As a remedy, crude ketone 2 was simply reduced
to alcohol 13, which could be isolated in moderate overall
yield. As anticipated, NOE studies confirmed the configura-
tions and conformation of trioxadispiroketal 13 which
reflected the thermodynamically favored array, including
double anomeric stabilization at C31 and C35.
Because tricyclic ketone 2 proved to be unstable, the
original plan to install the C28 chlorine R to the ketone after
trioxadispiroketal formation was revised. Instead, an early
installation of the chlorine was investigated.
This revision was simple to initiate, as it required only a
C26-C29 R-chloro aldehyde 14 instead of the original 6.
The synthesis of aldehyde 14 commenced with trans-2-
butene-1,4-diol (15). Monoprotection followed by Sharpless
asymmetric epoxidation14 afforded epoxide 16 (Scheme 5).
Chloride opening of trimethyl borate-chelated epoxy alcohol
16 yielded the anticipated chlorohydrin in >20:1 regio-
selectivity.15 Simple protecting group manipulations followed
by mild Dess-Martin periodinane oxidation then gave the
Scheme 5a
a DMB ) 3,4-dimethoxybenzyl.
The previously successful NHK coupling between alde-
hyde 6 and iodoalkyne 9 (Scheme 2) became problematic
with R-chloro aldehyde 14. Only trace amounts of the desired
coupling product were observed, whereas dechlorinated 14
was returned as the major product. This reaction17 could not
be avoided after several attempts at optimization. Thus, the
corresponding lithium acetylide derivative was employed.
Alkyne 7 was desilylated and then converted into alcohol
(15) Tomata, Y.; Sasaki, M.; Tanino, K.; Miyashita, M. Tetrahedron Lett.
2003, 44, 8975.
(16) Horita, K.; Yoshioka, T.; Tanaka, T.; Oikawa, Y.; Yonemitsu, O.
Tetrahedron. 1986, 42, 3021.
(14) Katsuki, T.; Sharpless, B. K. J. Am. Chem. Soc. 1980, 102, 5974.
Org. Lett., Vol. 8, No. 14, 2006
(17) Fu¨rstner, A. Chem. ReV. 1999, 99, 991.
2999

induce spirocyclizations did not lead to the desired trioxa-
dispiroketal. Under basic reaction conditions, 21 gave a
C31-C38 spiroketal, analogous to that formed from 10 under
similar conditions (Scheme 4), but with concomitant conver-
sion of vicinal C27,28 chlorohydrin into the corresponding
epoxide. In contrast, the use of acidic conditions or metal
salts to induce trioxadispiroketal formation led to elimination
products while retaining the chlorine atom. Thus, we have
yet to define conditions that allow efficient incorporation of
the vicinal C28,29 chloromethoxy functionalization into the
intact trioxadispiroketal system of spirastrellolide A via the
sequenced DIHMA process. To date, these approaches seem
best suited for the generation of des-chloro analogues of
spirastrellolide A.
Scheme 6
In summary, a sequential conjugate addition cascade
assembly of the spirastrellolide A trioxadispiroketal moiety
has been developed, but incorporation of the C28 chloride
was problematic. This approach is readily amenable to the
synthesis of either enantiomer of the spirastrellolide A C26-
C40 domain, which may accommodate either absolute
configurational assignment. Extension of this synthetic entry
to the generation of spirastrellolide analogues and evaluation
of their biological activities are ongoing.
Acknowledgment. This publication was made possible
by Grant No. ES10615 from the National Institute of
Environmental Health Sciences (NIEHS), NIH. Its contents
are solely the responsibility of the authors and do not
necessarily represent the official views of NIEHS, NIH. We
thank the University of Minnesota scientists L. Weyer and
J. Nielson for helpful comments on the manuscript and T.
Murray, B. Wang, and J. Xu for helpful discussions.
19 in a stepwise process (Scheme 6). Treatment of 19 with
n-BuLi generated the alkoxide/acetylide dianion. This was
added to aldehyde 14 to give a diol that was then oxidized
to ketone 20.
Removal of the C27 and C37 hydroxyl protecting groups
from 20 gave chlorohydrin diol 21 (Scheme 6). Submission
of 21 to acidic or basic reaction conditions in attempts to
Supporting Information Available: Experimental pro-
cedures and spectroscopic data for new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
OL0609457
3000
Org. Lett., Vol. 8, No. 14, 2006