Studies toward a synthesis of epothilone A: Stereocontrolled assembly of the acyl region and models for macrocyclization

Full text of DOI:10.1021/jo961674o

8000
J . Org. Chem. 1996, 61, 8000-8001
Sch em e 1. Con ver gen t Str a tegy for a Tota l
Syn th esis of Ep oth ilon e A (1)
Stu d ies tow a r d a Syn th esis of Ep oth ilon e
A: Ster eocon tr olled Assem bly of th e Acyl
Region a n d Mod els for Ma cr ocycliza tion
Peter Bertinato,† Erik J . Sorensen,†
Dongfang Meng,†,‡ and Samuel J . Danishefsky*,†,‡
Laboratory for Bioorganic Chemistry, Sloan-Kettering
Institute for Cancer Research, 1275 York Avenue,
New York, New York 10021, and Department of Chemistry,
Columbia University, Havemeyer Hall,
New York, New York 10027
Received August 30, 1996
In our previous paper, we described a synthesis of the
“alkoxy” segment of epothilone A 1 (see compound 2,
Scheme 1) encompassing carbons 10-21.1 In this paper,
we address the synthesis of another fragment encoding
the stereochemical information of acyl section carbons
3-9. It was envisioned that the aldehydo center (C3) of
the formal target 3 would serve as an attachment site to
a nucleophilic construct derived from compound 2 (re-
quiring placement of a two-carbon insert, as suggested
in Scheme 1), through either inter- or intramolecular
means. In such a context, it would be necessary to deal
independently with the stereochemistry of the secondary
alcohol center eventually required at C3. One of the
interesting features of system 3 is the presence of
geminal methyl groups at carbon 4 (epothilone number-
ing). It was our hope to again use a dihydropyran
strategy to assemble a cyclic matrix corresponding, after
appropriate disassembly, to a viable equivalent of system
3. We hoped to expand upon our dihydropyran paradigm
to include the synthesis of gem dimethyl containing cyclic
and acyclic fragments. The particular reaction type we
had in mind for this purpose is generalized under the
heading of transformation of 4 f 5 (see Scheme 2). At
this juncture, we deliberately avoid commitment as to
the nature of the electrophile, E. Accordingly, we leave
for the moment unaddressed the question as to whether
a reduction would or would not be necessary in going
from structure type 5 to reach the intended generalized
target 3.
Once again, our opening step consisted of a stereo-
chemically tunable version of the diene-aldehyde cyclo-
condensation reaction2 (Scheme 3)sin this instance
drawing upon chelation control in the merger of the
readily available enantiomerically homogeneous aldehyde
6 with the known diene 7.3 Indeed, as precedent would
have it, under the influence of titanium tetrachloride
there was produced substantially a single isomer shown
as compound 8.4 In the usual and stereochemically
reliable way,5 the dihydropyrone was reduced to the
corresponding glycal 9. At this point, we utilized a
directed Simmons-Smith reaction for the conversion of
glycal 9 to cyclopropane 10.6 This compound is indeed
an interesting structure in that it corresponds in one
Sch em e 2. Glyca l Cyclop r op a n e Solvolysis
Str a tegy for th e In tr od u ction of Gem in a l Meth yl
Gr ou p s
sense to a cyclopropano version of a C-glycoside. At the
same time, the cyclopropane is part of a cyclopropylcarbi-
nyl alcohol system with attendant possibilities for rear-
rangement.7 It was our intention to cleave the C-glyco-
sidic bond of the cyclopropane in a fashion that would
elaborate the geminal methyl groups, leaving in its wake
a solvent-derived glycoside with the desired aldehyde
oxidation state at C-3 (see hypothesized transformation
4 f 5, Scheme 2). In early efforts, the nonoxidative
version of the projected reaction (i.e., E+ ) H+) could not
be reduced to practice. Instead, products clearly at-
tributable to the ring-expanded system 118 were identi-
fied.
Fortunately, however, the desired sense of cyclopro-
pane opening, under the influence of the ring oxygen, was
achieved by subjecting compound 10 to oxidative opening
with N-iodosuccinimide.9 The intermediate iodomethyl
compound, obtained as a methyl glycoside 12, when
exposed to the action of tri-n-butyltin hydride, gave rise
to pyran 13 containing the geminal methyl groups.
Protection of this alcohol (see 13 f 14), followed by
cleavage of the glycosidic bond, revealed the acyclic
dithiane derivative 15 which can serve as a functional
version of the hypothetical aldehyde 3.
We have also begun to explore possible ways of com-
bining fragments relating to 2 and 3 in a fashion to reach
epothilone and congeners thereof. Mindful of the pio-
(6) (a) Winstein, S.; Sonnenberg. J . J . Am. Chem. Soc, 1961, 83,
3235. (b) Dauben, W. G.; Berezin, G. H. J . Am. Chem. Soc. 1963, 85,
468. (c) Furukawa, J .; Kawabata, N.; Nishimura, J . Tetrahedron 1968,
24, 53. For selected examples, see: (d) Boeckman, R. K., J r.; Charette,
A. B.; Asberom, T.; J ohnston, B. H. J . Am. Chem. Soc. 1991, 113, 5337.
(e) Hoberg, J . O.; Bozell, J . J . Tetrahedron Lett. 1995, 36, 6831.
(7) Wenkert, E.; Mueller, R. A.; Reardon, E. J ., J r.; Sathe, S. S.;
Scharf, D. J .; Tosi, G. J . Am. Chem. Soc. 1970, 92, 7428.
† Sloan-Kettering Institute for Cancer Research.
‡ Columbia University.
(1) Meng, D.; Sorensen, E. J .; Bertinato, P.; Danishefsky, S. J . J .
Org. Chem. 1996, 61, 7998-7999.
(2) Danishefsky, S. J . Aldrichim. Acta 1986, 19, 59.
(3) Danishefsky, S. J .; Yan, C.-F.; Singh, R. K. Gammill, R. B.;
McCurry, P. M., J r.; Fritsch, N.; Clardy, J . J . Am. Chem. Soc. 1979,
101, 7001.
(4) (a) Danishefsky, S. J .; Pearson, W. H.; Harvey, D. F.; Maring,
C. J .; Springer, J . P. J . Am. Chem. Soc. 1985, 107, 1256. (b)
Danishefsky, S. J .; Myles, D. C.; Harvey, D. F. J . Am. Chem. Soc. 1987,
109, 862.
(8) For example, exposure of 10 to acidic methanol gave rise to an
epimeric mixture of seven- membered mixed acetals, presumably
through the addition of methanol to oxocarbenium ion 11. This most
interesting transformation is under active study in our laboratory.
(9) For interesting Hg(II)-induced solvolyses of cyclopropanes that
are conceptually similar to the conversion of 10 to 12, see: (a) Collum,
D. B.; Still, W. C.; Mohamadi, F. J . Am. Chem. Soc. 1986, 108, 2094.
(b) Collum, D. B.; Mohamadi, F.; Hallock, J . S. J . Am. Chem. Soc. 1983,
105, 6882. Following this precedent, we did, in fact, accomplish a Hg-
(II)-induced solvolysis of cyclopropane 10, although this transformation
proved to be less efficient than the reaction shown in Scheme 3.
(5) Danishefsky, S. J . Chemtracts Org. Chem. 1989, 2, 273.
S0022-3263(96)01674-X CCC: $12.00 © 1996 American Chemical Society

Communications
J . Org. Chem., Vol. 61, No. 23, 1996 8001
Sch em e 3. En a n tioselective Syn th esis of Com p ou n d 15
Sch em e 4. Con str u ction of Ep oth ilon e Mod el
System s 20-22 by Rin g-Closin g Olefin Meta th esis
methyl groups corresponding to their placement at C4
of epothilone A. Once again, olefin metathesis occurred,
this time curiously producing olefin 21 as a single entity
in 70% yield (stereochemisty tentatively assigned as Z).
Substantially identical results were obtained through the
use of Schrock’s molybdenum alkylidene metathesis
catalyst.
Having shown that olefin metathesis is equal, in
principle, to the challenge of constructing the 16-
membered ring containing both the required epoxy and
thiazolyl functions of our target system, we have started
to project a synthesis of epothilone A itself. Clearly, with
these fragments in hand, a variety of strategies for their
combination, culminating in either carbon-carbon bond
formation or macrolactonization, can be entertained, and
these are being evaluated. At the present writing,
however, it is appropriate to point out that no successful
olefin metathesis reaction has yet been realized from seco-
systems bearing a full compliment of functionality re-
quired to reach epothilone. These negative outcomes may
merely reflect a failure to identify, as yet, a suitable
functional group constraint pattern appropriate for mac-
rocylization.13 Many possibilities remain to be screened.
Accordingly, intramolecular olefin metathesis is still
included in a variety of ring-forming options currently
being evaluated for reaching epothilone A.
neering studies of Schrock10a and Grubbs10b and the
recent ground-breaking disclosure of Hoveyda,11 we
wondered about the possibility of realizing such an
approach en route to our goal.12 The matter was first
examined with two model ω-unsaturated acids 16 and
17 that were used to acylate alcohol 2 to provide esters
18 and 19, respectively (see Scheme 4). These com-
pounds did indeed undergo olefin metathesis macrocy-
clization in the desired manner under the conditions
shown. In the case of substrate 18, 20 was obtained as
a mixture of E- and Z- stereoisomers (ca. 1:1). Diimide
reduction of 20 was then conducted to provide homoge-
neous 22 in 50% yield. The olefin metathesis reaction
was also extended to compound 19 bearing geminal
Ack n ow led gm en t. This work was supported by
NIH Grant No. CA 28824. Postdoctoral Fellowships are
gratefully acknowledged by E.J .S. (NSF, CHE-9504805)
and P.B. (NIH, CA 62948). We gratefully acknowledge
Dr. George Sukenick (NMR Core Facility, Sloan-Ket-
tering Institute) for NMR and mass spectral analyses.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures and spectroscopic data for the compounds illustrated
in the schemes (compounds 8-10, 12-16, and 18-22) (6
pages).
J O961674O
(10) (a) Schrock, R. R.; Murdzek, J . S.; Bazan, G. C.; Robbins, J .;
DiMare, M.; O’Regan, M. J . Am. Chem. Soc. 1990, 112, 3875. (b)
Schwab, P.; France, M. B.; Ziller, J . W.; Grubbs, R. H. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 2039. For reviews of ring-closing metathesis,
see: (c) Grubbs, R. H.; Miller, S. J .; Fu, G. C. Acc. Chem. Res. 1995,
28, 446. (d) Schmalz, H.-G. Angew. Chem., Int. Ed. Engl. 1995, 34,
1833.
(11) Houri, A. F.; Xu, Z.; Cogan, D. A.; Hoveyda, A. H. J . Am. Chem.
Soc. 1995, 117, 2943.
(12) For recent examples of ring-closing metathesis, see: (a) Martin,
S. F.; Chen, H.-J .; Courtney, A. K.; Liao, Y.; Pa¨tzel, M.; Ramser, M
N.; Wagman, A. S. Tetrahedron 1996, 52, 7251. (b) Fu¨rstner, A.;
Langemann, K. J . Org. Chem. 1996, 61, 3942.
(13) Substrates containing the full complement of oxygenated
functionality, including the trisubstituted olefin and thiazolyl moiety,
were screened for ring-closing olefin metathesis. In an effort to favor
ring closure through the rigidification of the carbon backbone, a seco
structure possessing a cyclic isopropylidene ketal bridging a C3-C5
diol relationship was prepared and subjected to ring-closing metathesis.
In one instance, we also screened a substrate containing functionality
that would lead to the C12-C13 epoxide, but lacking this function,
per se. In spite of these setbacks, efforts to fashion the macrolide of
epothilone A through ring-closing metathesis are continuing A full
account of these studies will be disclosed in due course.