Synthesis of rimocidinolide methyl ester, the aglycone of (+)-rimocidin

Article abstract of DOI:10.1002/anie.200453697

An efficient, convergent route based on cyanohydrin acetonide couplings was used for the synthesis of the aglycone of rimocidin, rimocidinolide methyl ester (1). Cyanohydrin acetonides were used as β-hydroxyketone synthetic equivalents in the assembly of the polyol chain.

Full text of DOI:10.1002/anie.200453697

Communications
Natural Products Synthesis
Synthesis of Rimocidinolide Methyl Ester, the
Aglycone of (+)-Rimocidin**
Garrick K. Packard, Yueqing Hu, Andrea Vescovi, and
Scott D. Rychnovsky*
Polyene macrolide antibiotics have found widespread use as
antifungal agents.[1] The family of polyene macrolides incor-
porating a mycosamine sugar represent the most common and
widely used members of the class of natural products, and
include amphotericin B, nystatin, candidin, pimaricin, and
rimocidin.[2] These stereochemically complex, highly oxy-
genated structures have attracted the attention of synthetic
chemists.[3] Amphotericin B was synthesized by Nicolaou and
co-workers,[4] and the aglycone of amphotericin B, amphoter-
onolide B, was prepared by Masamune and co-workers.[5,6] We
have been interested in the synthesis of polyene macrolides,
but most of our work has focused on the oxopolyene
subclass.[7] We report herein the first synthesis of the
rimocidin aglycone, rimocidinolide methyl ester (2).
Rimocidin (1) is a 28-membered polyene macrolide which
was isolated from Streptomyces rimosus in 1951.[8] Cope et al.
established the carbon skeleton in the mid-1960s,[9] and its flat
structure was determined independently by the groups of
Borowski and Rinehart in the 1970s.[10] The stereochemistry
of rimocidin was determined by Borowski and co-workers
group in 1995 through 2D NMR analysis.[11] Rimocidin
includes 14 stereogenic centers, a conjugated tetraene, a
tetrahydropyran hemiketal, and a macrocyclic lactone and is
b-linked to mycosamine. These functionalities make rimoci-
din sensitive to both strongly acidic and strongly basic
reaction conditions. Smith et al. recently reported a synthesis
of the C1–C18 segment of rimocidin[12] through a dithiane
[*] Dr. G. K. Packard, Dr. Y. Hu, Dr. A. Vescovi, Prof. S. D. Rychnovsky
Department of Chemistry
University of California, Irvine
516 Rowland Hall, Irvine, CA 92697–2025 (USA)
Fax: (+1)949-824-6379
E-mail: srychnov@uci.edu
[**] This work was supported by the National Institutes of Health (GM-
43854) and by a DAAD postdoctoral fellowship and a European
Merck Postdoctoral Fellowship to A.V.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
2822
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200453697
Angew. Chem. Int. Ed. 2004, 43, 2822 –2826

Angewandte
Chemie
linchpin strategy.[13] Our strategy is based on cyanohydrin
acetonide couplings to facilitate the convergent assembly of
the rimocidin aglycone.
Cyanohydrin acetonides have been used extensively in
our laboratories as synthons for syn-1,3-diols,[14] and this
strategy has been applied, for example, to syntheses of
roxaticin[15] and dermostatin.[16] The structure of rimocidin,
however, does not contain any syn-1,3-diol segments. In our
approach to rimocidin, cyanohydrin acetonides would be used
as synthons for b-hydroxyketones rather than for syn-1,3-diol
acetonides. Thus the cyanohydrin would revert to its earlier
application as an acyl anion equivalent,[17] and the strategy is
comparable to the widely used dithiane disconnection.[18]
A retrosynthetic analysis of rimocidinolide methyl ester
(2) is presented in Scheme 1. The macrolide ring would be
Scheme 2. Reagents and conditions. a) Bu2BOTf, Et3N, CH2Cl2, then 6,
À788C; b) MeONHMe·HCl, Me3Al, CH2Cl2, 53% over two steps;
c) AllylMgBr, THF, À108C, 92%; d) Me4N(AcO)3BH, CH3CN/HOAc,
À208C, 94%; e) 2,2-DMP, acetone, CSA, reflux, repeat; f) TBAF, THF,
99% over two steps; g) I2, PPh3, imidazole, benzene/Et2O, 99%. TIP-
S=triisopropylsilyl, Tf=trifluoromethanesulfonyl, 2,2-DMP=2,2-dime-
thoxypropane, CSA=Æ-camphorsulfonic acid, TBAF=tetrabutylammo-
nium fluoride.
put. Cleavage of the TIPS group and introduction of the
iodide substituent with PPh3 and I2 completed the synthesis of
10[22] in an overall yield of 44% in seven steps.
Two cyanohydrin acetonides, 16 and 20, were prepared for
the synthesis. The synthesis of 16 is presented in Scheme 3 and
also uses an Evans aldol coupling.[19] Aldol coupling between
11 and aldehyde 12 led to the expected adduct. The optimal
procedure for producing diol 13 involved hydrolytic removal
of the auxiliary and reduction of the acid with LAH. Direct
reduction of the adduct led to complexmitxures. Several
approaches were investigated for the selective benzylation of
Scheme 1. Retrosynthetic analysis of rimocidinolide methyl ester 2.
prepared from two segments, an unsaturated aldehyde 3 and
the complexphosphonate 4. Two bonds would be formed
between 3 and 4: an ester linkage and an alkene. The ester and
alkene could be formed in either order, but in general the
Horner–Emmons cyclization route has better precedent and
would be investigated first. The phosphonate 4 would be
prepared from protected polyol 5. We planned to assemble
the polyol 5 through cyanohydrin acetonide couplings.
The most complexsegment of the polyol chain is the C12–
C17 segment 10, the precursor to the hemiacetal ring. Its
synthesis, based on an enantioselective aldol reaction, is
outlined in Scheme 2. An Evans aldol reaction[19] between 7
and aldehyde 6 gave the expected product contaminated with
the deconjugated crotonate. The aldol adduct was most
conveniently isolated after conversion into its Weinreb amide
8.[20] Treatment with allyl magnesium bromide and subsequent
anti-selective reduction with the Evans triacetoxyborohy-
dride[21] generated diol 9 as a single diastereomer. Acetonide
formation did not proceed to completion, and the recovered
starting material was recycled to improve material through-
Scheme 3. Reagents and conditions. a) Bu2BOTf, Et3N, CH2Cl2, then
12, À788C, 92%; b) LiOH, H2O2; c) LAH, THF, reflux, 74% over two
steps; d) NaH, BnBr, DMF, À508C, 70%; e) HCl (1n), EtOH, reflux,
98%; f) TEMPO, KBr, NaOCl; g) TMSCN, KCN·18[crown]-6, 92% over
two steps; h) 2,2-DMP, acetone, PPTS, reflux, 97% (3:2 ratio).
PMB=p-methoxybenzyl, LAH=lithium aluminum hydride,
DMF=N,N-dimethylformamide, TEMPO=2,2,6,6-tetramethylpiperi-
din-1-yloxyl, TMS=trimethylsilyl, PPTS=pyridinium toluene-p-sulfo-
nate.
Angew. Chem. Int. Ed. 2004, 43, 2822 –2826
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2823

Communications
the primary alcohol 13,[23] but the most effective method used
excess NaH and 1 equivalent of BnBr.[24] The PMB group was
cleaved selectively by treatment with TMSCl, SnCl2, and
anisole,[25] but the deprotection step was more conveniently
carried out by treatment with HCl (1n) in refluxing ethanol.
Conversion of the 1,3-diol into a cyanohydrin acetonide made
use of the selective oxidation of primary alcohols to aldehydes
with TEMPO.[26] Oxidation of 14 catalyzed by TEMPO led to
a solution of the b-hydroxyaldehyde in CH2Cl2. The solution
was treated directly with TMSCN and KCN/[18]crown-6 to
generate the cyanohydrins 15. Experience has shown that b-
hydroxyaldehydes rapidly dimerize in the absence of solvent,
but solutions can be handled without trouble.[27] Acetonide
formation catalyzed by PPTS gave the cyanohydrin acetonide
16 as an inconsequential 3:2 mixture of isomers at the
cyanohydrin center. We consider this strategy for the
conversion of 1,3-diol 14 into cyanohydrin acetonide 16 to
be the most reliable and convenient method for the prepa-
ration of cyanohydrin acetonides.
Synthesis of the second cyanohydrin is presented in
Scheme 4. Enantioselective allylation of aldehyde 17, fol-
lowed by ozonolysis and reduction with NaBH4, produced the
diol 18 with 89% ee. Oxidation with TEMPO, cyanohydrin
formation, and acetonide formation completed the five-step
synthesis of cyanohydrin acetonide 20.
The cyanohydrin acetonide segments and iodide 10 were
combined as illustrated in Scheme 5. In the past, we alkylated
cyanohydrins by forming the anion and then adding the
electrophile.[14] Several years ago we reported that deproto-
nation was compatible with most alkyl halide electrophiles.[28]
The alkylation of cyanohydrin 20 was most conveniently and
reliably conducted through an in situ deprotonation method.
The less precious cyanohydrin 20 was used in twofold excess
relative to the iodide 10.[29] The two segments were combined
in THF at À408C, and a solution of LDA was added. If the
Scheme 4. Reagents and conditions. a) 1. (À)-Ipc2BAllyl, THF,
À1008C; 2. H2O2, NaOH (1n), MeOH; b) 1. O3, CH2Cl2/MeOH,
À788C; 2. NaBH4, À788C, 82% over two steps; c) TEMPO, KBr,
NaOCl; d) TMSCN, KCN·[18]crown-6, 94% over two steps; e) 2,2-
DMP, acetone, PPTS, reflux, 88% (1:1 ratio). Ipc=isopinocampheyl.
reaction were incomplete by TLC analysis, more LDA
solution was added and the process was repeated until no
starting iodide remained. This approach is robust and works
well on both small (< 100 mg) and large (ca. 10 g) scales.
The hemiacetal was introduced by deprotection of the
cyanohydrin acetonide. Treatment of 21 with acidic Dowex
resin in wet methanol liberated the alcohols, and exposure of
the solution to Et3N led to spontaneous generation of the
hemiacetal 23. Reprotection of 23 began with methyl acetal
formation and protection of the secondary alcohols as TBS
ethers. Reductive debenzylation and iodide formation gave
iodide 24. Cyanohydrin 16 (2 equiv) was combined with
iodide 24 by using the in situ deprotonation method to yield
the alkylated product 5. Both cyanohydrins 5 and 21 were
formed as single isomers in the alkylation steps, apparently
owing to a kinetic anomeric effect.[14] Thus the C1–C17
protected polyol segment 5 was prepared in seven steps and in
55% overall yield from iodide 10.
Oxidation of the diene 5 to a dialdehyde was much more
difficult than we had anticipated, and we opted for a selective
Scheme 5. Reagents and conditions. a) LDA, THF, DMPU, À408C, 92%; b) 1. wetMeOH, Dowex 50X2–100; 2. Et 3N, 87%; c) HC(OMe)3, MeOH,
PPTS, 89%; d) TBSCl, imidazole, DMAP, DMF, 97%; e) Li/NH3, THF, À788C, 93%; f) I2, PPh3, imidazole, benzene, 92%; g) LDA, THF, DMPU,
À408C, 93%. LDA=lithium diisopropylamide, DMPU=N,N’-dimethyl-N,N’-propyleneurea, TBS=tert-butyldimethylsilyl, DMAP=4-dimethylami-
nopyridine.
2824
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 2822 –2826

Angewandte
Chemie
Scheme 6. Reagents and conditions. a) OsO4, NMO, 90%; b) Pb(OAc)4, 91%; c) (EtO)2P(O)CH3, BuLi, THF; d) O3, MeOH/CH2Cl2, À788C, 95%;
e) TPAP, NMO, 87%; f) LiOMe, MeOH, 98%; g) Dess–Martin periodinane, CH2Cl2, 95%; h) Pd(OH)2, H2, EtOAc, 96%; i) Dess–Martin periodi-
nane, CH2Cl2; j) NaClO2, 2-methyl-2-butene, tBuOH, 99% over two steps; k) Yamaguchi esterification, 79%; l) K2CO3, [18]crown-6, toluene, 608C,
5 h, 74%; m) HF/pyridine, pyridine, THF, 238C, 30 min; n) HCl (6n), THF/H2O, 238C, 5 h; o) Et3N, CH2Cl2, 238C, 10 min, 55% over three steps.
NMO=N-methylmorpholine-N-oxide, TPAP=tetra-n-propylammonium perruthenate.
oxidation route. Osmium tetraoxide selectively oxidized the
C17 alkene in preference to the C14 alkene (Scheme 6).
Presumably the C14 alkene is hindered by the flanking
equatorial substituents. Cleavage of the diol gave the C17
aldehyde 28, and the phosphonate was introduced by using
lithiated diethyl methylphosphonate. Lactol 29 was produced
upon ozonolysis. The lactol was converted into the methyl
ester by oxidation with TPAP–NMO[30] followed by treatment
with lithium methoxide in methanol. Synthesis of phospho-
nate 4 was completed by oxidation of the alcohol, hydro-
genation to cleave the benzyl ether, and then a two-step
oxidation of the alcohol function at C1 to an acid. The
described conversion of polyol 5 into phosphonate 4 is the
Cleavage of the acetonide and hydrolysis of the methyl acetal
took place upon treatment with HCl (6n) in THF. Finally, the
C7 ketone was unmasked by brief exposure to Et3N. Thus,
rimocidinolide methyl ester (2) was synthesized in 29 steps in
the longest linear sequence and in 4.3% overall yield from
aldehyde 6. In practice, a number of these steps were
combined for convenience. The synthetic rimocidinolide 2
was characterized by NMR and IR spectroscopy as well as
HRMS and CD.[36]
The synthesis of the rimocidinolide and in particular of the
polyol segment 5 illustrates the usefulness of cyanohydrin
acetonides as b-hydroxyketone synthons for the convergent
synthesis of polyol chains.[14] Segment 5 was assembled in a 14-
step linear sequence in nearly 25% overall yield from
aldehyde 6. The cyanohydrin acetonides 16 and 20 were
prepared in high yield from terminal 1,3-diols by using a new
protocol. Alkylation of these synthons with alkyl iodides 24
and 10 proceeded in 93% and 92% yields, respectively,
through an in situ deprotonation technique. Ketones can be
liberated from cyanohydrin acetonides by standard depro-
tection methods. Cyanohydrin acetonide alkylations repre-
sent a very efficient strategy for assembling ketone-containing
chains, and this approach complements our previous use of
these materials and syn-1,3-diol synthons. Polyol syntheses
based on cyanohydrin acetonide couplings are therefore a
competitive alternative to the more common dithiane strat-
egies.[18]
result of many different attempts using
a variety of
approaches. Functional-group interactions thwarted most of
the routes investigated. The final route is reliable and avoids
the common pitfalls.
The synthesis of rimocidinolide 2 was completed as
illustrated in Scheme 6. The acid 4 was esterified with alcohol
3
[31] by using the Yamaguchi protocol.[32] (Alternatives such as
DCC, HATU, and PyBOP produced the ester but in lower
yields.[33]) An attempted cyclization of the resulting ester with
LiCl/DBU resulted in elimination of acid 4, which was
reisolated in 75% yield.[34] On the other hand, Horner–Wad-
sworth–Emmons macrocyclization with K2CO3 and
18[crown]-6 in toluene produced the macrolide 31 in 74%
yield.[35] Deprotection of macrolide 31 began with removal of
the TBS ethers with HF–pyridine in pyridine/THF. The
methyl acetal was partially removed under these conditions.
Received: January 8, 2004 [Z53697]
Angew. Chem. Int. Ed. 2004, 43, 2822 –2826
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2825

Communications
À
[25] T. Akiyama, H. Shima, S. Ozaki, Synlett 1992, 415 – 416.
[26] P. L. Anelli, C. Biffi, F. Montanari, S. Quici, J. Org. Chem. 1987,
52, 2559 – 2562.
[27] S. D. Rychnovsky, D. J. Skalitzky, J. Org. Chem. 1992, 57, 4336 –
4339.
Keywords: C C coupling · cyanohydrins · macrocycles ·
natural products · total synthesis
.
[1] S. B. Zotchev, Current Med. Chem. 2003, 10, 211 – 223.
[2] S. Omura, H. Tanaka in Macrolide Antibiotics: Chemistry,
Biology, and Practice (Ed.: S. Omura), Academic Press, New
York, 1984, pp. 351 – 404.
[3] a) T. Nakata in Macrolide Antibiotics: Chemistry, Biology, and
Practice, 2nd ed. (Ed.: S. Omura), Academic Press, San Diego,
CA, 2002, pp. 181 – 284; b) C. Schneider, Angew. Chem. 1998,
110, 1449 – 1451; Angew. Chem. Int. Ed. 1998, 37, 1375 – 1378;
c) J. M. Beau in Recent Progress in the Chemical Synthesis of
Antibiotics (Eds.: G. Lukacs, M. Ohno), Springer, Berlin 1990,
pp. 135 – 182.
[4] a) K. C. Nicolaou, R. A. Daines, W. S. Uenishi, W. S. Li, D. P.
Papahatjis, T. K. Chakraborty, J. Am. Chem. Soc. 1988, 110,
4672 – 4685; b) K. C. Nicolaou, R. A. Daines, W. S. Uenishi,
W. S. Li, D. P. Papahatjis, T. K. Chakraborty, J. Am. Chem. Soc.
1988, 110, 4685 – 4696; c) K. C. Nicolaou, R. A. Daines, Y.
Ogawa, T. K. Chakraborty, J. Am. Chem. Soc. 1988, 110, 4696 –
4705.
[5] a) R. M. Kennedy, A. Abiko, T. Takemasa, H. Okumoto, S.
Masamune, Tetrahedron Lett. 1988, 29, 451 – 454; b) S. Masa-
mune, Ann. N. Y. Acad. Sci. 1988, 544, 168 – 179.
[28] S. D. Rychnovsky, S. S. Swenson, J. Org. Chem. 1997, 62, 1333 –
1340.
[29] When the ratio of nitrile 20 to iodide 10 was reduced to 1.5:1, the
product 21 was isolated in 85% yield. In practice, the relative
value of the nitrile and iodide dictated a higher ratio of the two
reagents.
[30] S. V. Ley, J. Norman, W. P. Griffith, S. P. Marsden, Synthesis
1994, 639 – 666.
[31] The synthesis of aldehyde 3 is included in the Supporting
Information.
[32] J. Inanaga, K. Hirata, H. Saeki, T. Katsuki, M. Yamaguchi, Bull.
Chem. Soc. Jpn. 1979, 52, 1989 – 1993.
[33] DCC = 1,3-dicyclohexylcarbodiimide; HATU = O-(7-azabenzo-
triazol-1-yl)-N,N’,N’’,N’’’-tetramethyluronium; PyBOP = 1-ben-
zotriazolyloxy-tris(pyrollidino)phosphonium.
[34] M. A. Blanchette, W. Choy, J. T. Davis, A. P. Essenfeld, S.
Masamune, W. R. Roush, T. Sakai, Tetrahedron Lett. 1984, 25,
2183 – 2186.
[35] a) G. Stork, E. Nakamura, J. Org. Chem. 1979, 44, 4010 – 4011;
b) K. C. Nicolaou, S. P. Seitz, M. R. Pavia, N. A. Petasis, J. Org.
Chem. 1979, 44, 4011 – 4013.
1
[6] For a synthesis of pimaronolide, see: A. J. Duplantier, S.
Masamune, J. Am. Chem. Soc. 1990, 112, 7079 – 7081.
[7] S. D. Rychnovsky, Chem. Rev. 1995, 95, 2021 – 2040.
[8] J. W. Davisson, F. W. Tanner, A. C. Finlay, I. A. Solomons,
Antibiot. Chemother. 1951, 1, 289 – 290.
[36] The complete assignment of the H NMR spectrum of rimoci-
dinolide 2 is listed in the Supporting Information.
[9] A. C. Cope, E. P. Burrows, M. E. Derieg, S. Moon, W. D. Wirth,
J. Am. Chem. Soc. 1965, 87, 5452 – 5460.
[10] a) L. Falkowski, J. Golik, J. Zielinski, E. Borowski, J. Antibiot.
1976, 29, 197 – 198; b) R. C. Pandey, K. L. J. Rinehart, J. Anti-
biot. 1977, 30, 146 – 157.
[11] P. Sowinski, J. Pawlak, E. Borowski, P. Gariboldi, J. Antibiot.
1995, 48, 1288 – 1291.
[12] A. B. Smith III, S. M. Pitram, M. J. Fuertes, Org. Lett. 2003, 5,
2751 – 2754.
[13] a) A. B. Smith III, A. M. Boldi, J. Am. Chem. Soc. 1997, 119,
6925 – 6926; b) A. B. Smith III, S. M. Pitram, A. M. Boldi, M. J.
Gaunt, C. Sfouggatakis, W. H. Moser, J. Am. Chem. Soc. 2003,
125, 14435 – 14445.
[14] C. J. Sinz, S. D. Rychnovsky, Top. Curr. Chem. 2001, 216, 51 – 92.
[15] S. D. Rychnovsky, R. C. Hoye, J. Am. Chem. Soc. 1994, 116,
1753 – 1765.
[16] C. J. Sinz, S. D. Rychnovsky, Angew. Chem. 2001, 113, 3324 –
3327; Angew. Chem. Int. Ed. 2001, 40, 3224 – 3227.
[17] a) G. Stork, L. Maldonado, J. Am. Chem. Soc. 1971, 93, 5286 –
5287; b) J. D. Albright, Tetrahedron 1983, 39, 3207 – 3233.
[18] M. Yus, C. Najera, F. Foubelo, Tetrahedron 2003, 59, 6147 – 6212.
[19] D. A. Evans, J. Bartroli, T. L. Shih, J. Am. Chem. Soc. 1981, 103,
2127 – 2129.
[20] S. Nahm, S. M. Weinreb, Tetrahedron Lett. 1981, 22, 3815 – 3818.
[21] D. A. Evans, K. T. Chapman, E. M. Carreira, J. Am. Chem. Soc.
1988, 110, 3560 – 3578.
[22] B. Classon, Z. Liu, B. Samuelsson, J. Org. Chem. 1988, 53, 6126 –
6130.
[23] a) M. Carmen Cruzado, M. Martin-Lomas, Carbohydr. Res.
1988, 175, 193 – 199; b) D. A. Leigh, R. P. Martin, J. P. Smart,
A. M. Truscello, J. Chem. Soc. Chem. Commun. 1994, 1373 –
1374; c) D. L. Boger, J. A. McKie, T. Nishi, T. Ogiku, J. Am.
Chem. Soc. 1996, 118, 2301 – 2302; d) K. C. Nicolaou, F. L.
van Delft, S. R. Conley, H. J. Mitchell, Z. Jin, R. M. Rodriguez,
J. Am. Chem. Soc. 1997, 119, 9057 – 9058.
[24] A. Fukuzawa, H. Sato, T. Masamune, Tetrahedron Lett. 1987, 28,
4303 – 4306.
2826
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 2822 –2826