The total synthesis of (±)-rishirilide B [11]

Full text of DOI:10.1021/ja003272a

J. Am. Chem. Soc. 2001, 123, 351-352
351
Scheme 1
The Total Synthesis of (()-Rishirilide B
John G. Allen1 and Samuel J. Danishefsky1,2
Sloan-Kettering Institute
New York, New York 10021
Department of Chemistry
Columbia UniVersity
New York, New York 10028
ReceiVed September 5, 2000
Rishirilides B and A were isolated from Streptomyces rishir-
iensis OFR-1056 in 1984 by Naki and co-workers.3 They exhibit
antithrombotic activity4 through selective R2-macroglobulin in-
hibition, thereby leading to the activation of plasmin. Rishirilide
B is substantially more potent than A in this assay. The structure
of rishirilide A (although not its absolute configuration) was
established, by crystallographic means, to be 2. The assignment
of structure 1 to rishirilide B was not supported by crystallographic
data, but was rendered under the assumption of its biogenetic
connectivity to 2. In addition to their novel mechanism of action,
impinging on a crucial biological cascade, the structures of the
rishirilides interested us as focusing targets for total synthesis.
Recently, we have described the use of systems 3 as viable
equivalents of quinodimethides (4) for intermolecular cycload-
dition reactions with a range of dienophiles (Scheme 1).5 With
peri-substituents (R1 * H), as in 3b, 5a, or 5b, the rate of
cycloaddition is significantly reduced. Although substituted
cyclohexenones failed to react usefully with 5, we nonetheless
proposed the synthetic route to rishirilide B shown in Scheme 1
(vide infra).6
Central to the success of the proposal was the need to deal
with the serious retardation effect of peri-substituents required
to reach the C6 phenolic hydroxyl of 1. In particular, we sought
to exploit a discovery of Masamune,7 wherein a strategically
placed hydroxyl group could enhance the dienophilicity of an
acyclic R,â-unsaturated ketone, presumably by internal hydrogen
bonding. We wondered whether the Masamune effect could be
realized with an R′-hydroxylated cyclohexenone, to the extent
that it would react with quinodimethide precursors such as 5. This
line of conjecture led to the selection of 66 to serve as a putative
dienophile. In this modeling phase, we used the readily prepared8
5a as the presumptive quinodimethide precursor.
Scheme 2a
a (a) 160 °C, toluene-d8 15 h; (b) CSA, MeOH, 65%, 2 steps; (c)
isoamylmagnesium bromide (1.08 M), THF, 70%.
Thus, the tertiary alcohols of 6 and 7 exerted two critical
directivity functions. Reaching 1 from 8 required oxidation at
C1 (see arrow). However, various attempts to accomplish this
transformation failed, thereby prompting us to a rather more
interesting solution.
Toward this end, we synthesized the enedione 12.6 Condensa-
tion of 2-(trimethylsilyl)ethyl acetoacetate 99 with crotonaldehyde
gave 10.10 Formation of the silyl dienol ether, followed by
selective Rubottom-type oxidation using dimethyl dioxirane,11
presumably under guidance of the proximal secondary methyl
function, provided 11. Allylic bromination was found to be the
only effective method for functionalizing the γ-position. The
resulting bromide was smoothly converted to enedione 12
(Scheme 3).
In the event, reaction of 5a and 6 did occur at 160 °C over 15
h (Scheme 2). The crude cycloadduct was treated with camphor-
sulfonic acid in methanol under reflux, providing a 65% yield of
7. In the next step, the â-disposed hydroxyl group of 7 cleanly
directed the reaction of isoamylmagnesium bromide to the â face
of the ketone to afford 8 (mp 128-129 °C) in 70% yield. The
structure of this compound was verified by X-ray crystallography.
It was expected that enedione 12 would be a much more
reactive dienophile than was 6.12 Furthermore, it was envisioned
that through hydrogen bonding, the C4 ketone would be differ-
entially activated relative to that at C1.7c,d If this bias were
complementary to the bias introduced by the silyloxy peri-
substituent in 5b, there seemed to be a chance of gaining control
over the regiochemistry issue.
All expectations were realized in the Diels-Alder-like cy-
cloaddition of 12 and 5b. The predominant product (13, 90%
yield) was shown by extensive NMR measurements to be the
result of endo addition, anti to the Me and CO2TSE functions
with the desired regiochemistry, as shown. At this stage, however,
the stereochemistry at C3, relative to that at carbons 2 and 4a, in
the proposed 13 could not be rigorously established. Dehydration
and aromatization were accomplished through the action of CSA
to give 14. Reaction of 14 with excess isoamylmagnesium
(1) Bioorganic Chemistry Laboratory, Sloan-Kettering Institute.
(2) Department of Chemistry, Columbia University.
(3) Iwaki, H.; Nakayama, Y.; Takahashi, M.; Uetsuki, S.; Kido, M.;
Fukuyama, Y. J. Antibiot. 1984, 37, 1091.
(4) Plasmin, an enzyme important in fibrin degradation, is inhibited by
R2-macroglobulin. Agents such as the rishirilides that inhibit R2-macroglobulin
may be useful in the treatment and prevention of thrombosis by fibrinolytic
accentuation.
(5) (a) Allen, J. G.; Hentemann, M. F.; Danishefsky, S. J. J. Am. Chem.
Soc. 2000, 122, 571. (b) Hentemann, M. F.; Allen, J. G.; Danishefsky, S. J.
Angew. Chem., Int. Ed. 2000, 39, 1937.
(6) See Supporting Information.
(7) (a) Masamune, S.; Reed, L. A., III; Davis, J. T.; Choy, W. J. Org.
Chem. 1983, 48, 4441. (b) Choy, W.; Reed, L. A., III; Masamune, S. J. Org.
Chem. 1983, 48, 1137. Other examples include (c) Kelly, T. R.; Fu, Y.; Sieglen,
J. T., Jr.; De Silva, H. Org. Lett. 2000, 2, 2351, and (d) Tripathy, R.; Carroll,
P. J.; Thornton, E. R. J. Am. Chem. Soc. 1990, 112, 6743.
(8) (a) Bill, J. C.; Tarbell, D. S. Org. Synth. 1954, 34, 82. (b) Arnold, B.
J.; Sammes, P. G.; Wallace, T. W. J. Chem. Soc., Perkin Trans. 1 1974, 415.
(c) Nozaki, H.; Noyori, R.; Kozaki, N. Tetrahedron 1964, 20, 641.
(9) Ueda, Y.; Roberge, G.; Vinet, V. Can. J. Chem. 1984, 62, 2936.
(10) (a) Hauser, F. M.; Pogany, S. A. Synthesis 1980, 815. (b) Bohlmann,
F.; Prezewowsky, K. Chem. Ber. 1964, 97, 1176.
(11) Use of wet DMDO avoided silyl transfer, cf. (a) Chenault, H. K.;
Danishefsky, S. J. J. Org. Chem. 1989, 54, 4249. (b) Rubottom, G. M.;
Marrero, R. J. Org. Chem. 1975, 40, 3783.
(12) Sauer, J.; Wiest, H.; Mielert, A. Chem. Ber. 1964, 97, 3183.
10.1021/ja003272a CCC: $20.00 © 2001 American Chemical Society
Published on Web 12/21/2000

352 J. Am. Chem. Soc., Vol. 123, No. 2, 2001
Communications to the Editor
Scheme 3a
methylation of our synthetic 1 with diazomethane. We were also
1
able to obtain3,16 a copy of the high-field H NMR spectrum of
17 derived from naturally occurring 1. This spectrum was
superimposable with that of 17 derived by bis-methylation of our
synthetic 1. Accordingly, the structure of rishirilide B is now
rigorously established to be 1.17,18
Two further experiments serve to support the key role of the
hydroxyl group in directing these Diels-Alder-like reactions.
Benzocyclobutene 5a failed to react with the TBS derivative 18.
Moreover, in reaction with 3b, 19 provided a highly activated
dienophile, but with the C3 alcohol protected, a total loss of
regiocontrol resulted.6
a TSE ) 2-(trimethylsilyl)ethyl; (a) 1. TSEOH, ether, Na; 2. HCl(g)
,
34%; (b) NaH, DMF, TBSOTf; (c) DMDO (0.061 M), 76%, 2 steps; (d)
NBS, AIBN, CCl4, 91%; (e) Ag2CO3, acetone:H2O (8:1), 63%; (f) Dess-
Martin, CH2Cl2, 92%; (g) toluene-d8, 90 °C 12 h; (h) CSA, pyridine,
MeOH; 72% 2 steps; (i) isoamylmagnesium bromide (0.73 M), THF; (j)
TAS-F, THF, 65%, 2 steps; (k) CH2N2, Et2O, MeOH, 15% 16, 43% 17,
31% 1.
It seems likely that the chemistry and insights developed for
this total synthesis will find wider application.
Acknowledgment. This paper is dedicated to Professor Satoru
Masamune of the Massachusetts Institute for his discovery of the hydrogen
bonding effect in Diels-Alder reactions. This work was supported by
the National Institutes of Health: AI 16943/CA 28824 (S.J.D.) and CA-
80356 (J.G.A.). We thank Brian M. Bridgewater and David G. Churchill
for providing X-ray data. We thank Dr. George Sukenick for mass spectral
analyses.
bromide resulted in monoalkylation13 cis to the neighboring
hydroxyl group (see compound 15). Desilylation via TAS-F
reagent14 afforded (()-rishirilide B (1) in 47% overall yield from
dienophile 12.
At this point we still could not rigorously assert the stereo-
chemisty at C3. There was indeed some concern, since the 13C
NMR data obtained from our synthetic material, presumed to be
1, did not match those tabulated for the natural product in the
original publication.3,15 Furthermore, no specimen samples of 1
or any derivative thereof were available to us for direct compari-
son. Fortunately, we were able to obtain crystallographic verifica-
tion of the structure of 16 (mp 94-96 °C) obtained from partial
Supporting Information Available: Experimental, spectral and
crystallographic information (PDF). This material is available free of
charge via the Internet at http://pubs.acs.org.
JA003272A
(16) Personal communication from Professor Yoshiyasu Fukuyama, Faculty
of Pharmaceutical Sciences, Tokushima Bunri University.
(13) No dialkylation was observed, even with excess Grignard reagent,
possibly due to intramolecular lactol or lactolate formation.
(14) (a) Noyori, R.; Nishida, I.; Sakata, J.; Nishizawa, M. J. Am. Chem.
Soc. 1980, 102, 1223. (b) Scheidt, K. A.; Chen, H.; Follows, B. C.; Chemler,
S. R.; Coffey, D. S.; Roush, W. R. J. Org. Chem. 1998, 63, 1215.
(15) 1H NMR data for 1 was in excellent agreement with those previously
published. IR data was not available for the natural product.
(17) Presumably the discrepancy in the 13C NMR of synthetic 1 with the
tabulated data reported for the natural product reflects the concentration, pH,
or other experimental inconsistencies in the measurements.
(18) Since the 1H NMR spectra of 17 obtained from both synthetic and
natural sources differ markedly from that of the same terminal product of a
previously claimed synthesis by Hauser (Hauser, F. M.; Xu, Y. Org. Lett.
1999, 2, 335), reformulation of that work will be necessary.