Prins desymmetrization of a C2-symmetric diol: Application to the synthesis of 17-deoxyroflamycoin

Full text of DOI:10.1021/jo970380f

3022
J . Org. Chem. 1997, 62, 3022-3023
Communications
P r in s Desym m etr iza tion of a C2-Sym m etr ic
Diol: Ap p lica tion to th e Syn th esis of
17-Deoxyr ofla m ycoin
Sch em e 1
Scott D. Rychnovsky,* Guang Yang, Yueqing Hu, and
Uday R. Khire
Department of Chemistry, University of California,
Irvine, California 92697-2025
Received February 28, 1997
Sch em e 2
Roflamycoin is an ion-channel-forming polyene mac-
rolide antibiotic that shows significant antifungal activ-
ity.1,2 We recently reported the first total synthesis of
roflamycoin.3 One of the principle stumbling blocks in
that synthesis was the introduction of the acid-labile
hemiacetal ring. We were interested in preparing ana-
logs of roflamycoin to evaluate the structural require-
ments for ion-channel activity, and it was clear that the
presence of the hemiacetal ring would make analog
synthesis more difficult. Is the hemiacetal ring really
necessary for biological activity? Amphotericin B analogs
in which the hemiacetal position is modified range from
equipotent to less active by a factor of 16 in a fungistatic
assay.4 Presumably, the hemiacetal group in roflamycoin
plays a structural role by defining the conformation in
one of the two turn regions of the macrocycle, but that
role could also be played by an appropriately substituted
tetrahydropyran ring. The tetrahydropyran analog of
roflamycoin would not show the acid sensitivity of the
natural product and would have a clearly defined struc-
ture as there could be no possible equilibration between
the ketone form and the two possible hemiacetal forms
that is possible in the case of roflamycoin. Thus, 17-
deoxyroflamycoin was selected as the first roflamycoin
analog for synthesis and biological evaluation.
Natural roflamycoin was assembled from three seg-
ments, a C11-C26 bromide, C23-C26 cyanohydrin ac-
etonide 3, and C27-C35 cyanohydrin acetonide 12.3 The
two latter components could be reused in a synthesis of
17-deoxyroflamycoin, and a new C11-C26 tetrahydro-
pyran segment (9) would be substituted for the original
bromide. The TIPS-protected cyanohydrin acetonide 3
is a key building block in the synthesis in that it is first
alkylated with the C11-C26 bromide and then converted
into C26-iodide 11 for subsequent alkylation of nitrile 12.
Thus, compound 3 is a four-carbon syn-1,3-diol synthon
that can be used repeatedly to build up a polyol chain
and is the linchpin in the convergent roflamycoin strat-
egy.5 Cyanohydrin acetonide 3 was prepared as a single
enantiomer from methyl (S)-3,4-dihydroxybutyrate, 1,
itself available in two steps from (S)-malic acid (Scheme
1).6 Sequential silylations and reduction gave aldehyde
2, which was treated with TMSCN and KCN/18-crown-6
catalysis followed in one pot by acetonide protection to
give cyanohydrin acetonide 3. Preparation of compound
3 required six steps and proceeded in 32% overall yield
from (S)-malic acid.
Synthesis of 17-deoxyroflamycoin required substituting
C11-C22 tetrahydropyran segment 9 for the protected
C17-ketone segment used in the synthesis of natural
roflamycoin.3 The synthesis of tetrahydropyran 9 is
designed around an unusual Prins cyclization-desym-
metrization reaction outlined in Scheme 2.7 Desymme-
trization of a C2-symmetric substrate requires selective
monofunctionalization, and an intramolecular Prins cy-
clization automatically generates a monofunctionalized
product.8 Diepoxide 4 was reacted with vinyl Grignard
and catalytic CuI to give (4R,6R)-nona-1,8-diene-4,6-diol.9
Acetal exchange with 5 gave the cyclic acetal 6 in 80%
overall yield. Intramolecular Prins cyclization of 6 under
conditions optimized for acetate trapping10 gave tetrahy-
dropyran 7 in 42-51% yield. The relatively low yield in
the cyclization was due to the presence of the benzyl
ether: the corresponding acetaldehyde acetal cyclized in
80% yield under the same conditions.11 The final ste-
reogenic centers were introduced by Sharpless asym-
metric dihydroxylation, which gave a ca. 8:1 mixture of
stereoisomers.12,13 Monobromide formation using Mof-
fatt’s reagent,14 followed by standard reprotection chem-
(1) (a) Schlegel, R.; Thrum, H. Experientia 1968, 24, 11-12. (b)
Schlegel, R.; Thrum, H. J . Antibiot. 1971, 24, 368-74. (c) Schlegel, R.;
Thrum, H. J . Antibiot. 1971, 24, 360-7. (d) Schlegel, R.; Thrum, H.;
Zielinski, J .; Borowski, E. J . Antibiot. 1981, 34, 122-3.
(2) (a) Schlegel, R.; Grigorjev, P. A.; Thrum, H. Stud. Biophys. 1982,
92, 135-40. (b) Grigorjev, P.; Schlegel, R.; Thrum, H.; Ermishkin, L.
Biochim. Biophys. Acta 1985, 821, 297-304.
(3) Rychnovsky, S. D.; Khire, U. R.; Yang, G. J . Am. Chem. Soc.
1997, 119, 2058-2059.
(4) Taylor, A. W.; Costello, B. J .; Hunter, P. A.; MacLachlan, W. S.;
Shanks, C. T. J . Antibiot. 1993, 46, 486-93.
(7) Snider, B. B. In Comprehensive Organic Synthesis; Trost, B. M.,
Fleming, I., Heathcock, C. H., Ed.; Pergamon Press: New York, 1991;
Vol. 2, pp 527-561.
(8) Schreiber, S. L. Chem. Scr. 1987, 27, 563-566.
(9) Rychnovsky, S. D.; Griesgraber, G.; Zeller, S.; Skalitzky, D. J .
J . Org. Chem. 1991, 56, 5161-5169.
(5) For a related synthon see: Rychnovsky, S. D.; Griesgraber, G.
J . Org. Chem. 1992, 57, 1559-1563.
(6) Saito, S.; Hasegawa, T.; Inaba, M.; Nishida, R.; Fujii, T.; Nomizu,
S.; Moriwake, T. Chem. Lett. 1984, 1389-1392.
S0022-3263(97)00380-0 CCC: $14.00 © 1997 American Chemical Society

Communications
J . Org. Chem., Vol. 62, No. 10, 1997 3023
Sch em e 3
istry, gave the tetrahydropyran 9 in good overall yield.13
The key Prins desymmetrization reaction allowed tet-
rahydropyran 7 to be prepared as a single stereoisomer
in just three steps from diepoxide 4, and the C11-C22
building block 9 was prepared in only eight steps from
diepoxide 4.
mycoin were very similar to those of natural roflamycoin
and showed the expected deviations due to the absence
of the C17 oxygen atom. MS data, HPLC mobility, and
the UV spectrum of 17-deoxyroflamycoin were all con-
sistent with the assigned structure.
Roflamycoin shows antifungal activity with minimum
inhibitory concentrations (MIC) of 32 and 64 µg/mL
against Candida albicans and Cryptococcus neoformans,
respectively, in disk-diffusion assays.17 Synthetic 17-
deoxyroflamycoin was not active against either of these
organisms at 1280 µg/mL. Although the conformations
of roflamycoin and 17-deoxyroflamycoin are very similar
as judged by molecular modeling and the similarity
between their NMR spectra, the latter compound is
essentially inactive. The absence of the C17 oxygen shuts
down fungistatic activity, but the question still remains
whether the loss of biological activity is due to a loss of
ion-channel activity. We are currently investigating this
point. The C17 oxygen is crucial for antifungal activity,
and future roflamycoin analogs will incorporate this
functionality.
The synthesis of 17-deoxyroflamycoin is illustrated in
Scheme 3. The synthetic plan closely follows that used
in the total synthesis of natural roflamycoin.3 The anion
of nitrile 3 (2.1 equiv) was alkylated with bromide 9 to
give 10 in almost quantitative yield. Deprotection and
iodination15 of 10 gave the C26 iodide 11 that was ready
for the second alkylation reaction. Alkylation of 2.6 equiv
of the anion from 12 with 11 gave dinitrile 13 in 90%
yield, and reductive decyanation removed the nitrile and
benzyl groups to give protected polyol 14 as a single
stereoisomer. Phosphonopropionate ester synthesis fol-
lowed by oxidation gave the aldehyde 15, and elaboration
using two iterations of Wollenberg’s procedure gave the
tetraenal 16.3,16 Roush-Masamune cyclization gave an
excellent yield of the pentaene macrocycle 17, and a
simple acid-catalyzed deprotection gave 17-deoxyrofla-
mycoin in 80% yield. The NMR spectra of 17-deoxyrofla-
Ack n ow led gm en t. Support has been provided by
the National Institutes of Health, the University of
California, Irvine, and Pfizer Inc. We would like to
thank Professor Michael Selsted and Mr. Bruce Rogers
for evaluating the antifungal activity of roflamycoin and
17-deoxyroflamycoin.
(10) Hu, Y.; Skalitzky, D. S.; Rychnovsky, S. D. Tetrahedron Lett.
1996, in press.
(11) Overman encountered a similar problem with â-heteroatom
substitution in intramolecular oxonium ion cyclizations: Overman, L.
E.; Thompson, A. S. J . Am. Chem. Soc. 1988, 110, 2248-2256.
(12) Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem.
Rev. 1994, 94, 2483-547.
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
(13) The C21 configuration of the major isomer of 9 was determined
by 13C acetonide analysis that confirmed the C19-C21 anti relation-
ship: acetal, 100.9; methyl groups, 24.8, 24.6 ppm. Rychnovsky, S. D.;
Skalitzky, D. J . Tetrahedron Lett. 1990, 31, 945-8.
(14) Greenberg, S.; Moffatt, J . G. J . Am. Chem. Soc. 1973, 95, 4016-
4025.
(15) Garegg, p. J .; Samuelsson, B. J . Chem. Soc., Chem. Commun.
1979, 978-980.
(16) Wollenberg, R. H. Tetrahedron Lett. 1978, 717-720.
tails for the preparation of 17-deoxyroflamycoin (14 pages).
J O970380F
(17) (a) Lehrer, R. I.; Rosenman, M.; Harwig, S. S. S. L.; J ackson,
R.; Eisenhauer, P. J . Immunol. Methods 1991, 137, 167-173. (b) Van
Abel, R. J .; Tang, Y.-Q.; Rao, V. S. V.; Dobbs, C. H.; Tran, D.; Barany,
G.; Selsted, M. E. Int. J . Peptide Protein Res. 1995, 45, 401-409.