Total synthesis of myxalamide A

Article abstract of DOI:10.1021/jo9813742

The polyene antibiotic myxalamide A (1) has been prepared by total synthesis. The synthesis illustrates a useful strategy for synthesis in which the high 1,2-stereocontrol achievable with the aldol reaction can be parlayed by other stereoselective process

Full text of DOI:10.1021/jo9813742

J . Org. Chem. 1999, 64, 23-27
23
Tota l Syn th esis of Myxa la m id e A
Anna K. Mapp and Clayton H. Heathcock*
Department of Chemistry, University of California, Berkeley, California 94720
Received J uly 14, 1998
The polyene antibiotic myxalamide A (1) has been prepared by total synthesis. The synthesis
illustrates a useful strategy for synthesis in which the high 1,2-stereocontrol achievable with the
aldol reaction can be parlayed by other stereoselective processes so as to give compounds having
two or more stereocenters with remote relationships. Application of the Evans asymmetric aldol
reaction to aldehyde 13 gives the â-hydroxy imide 17. Because the substrate is an R,â-unsaturated
aldehyde, the alcohol is allylic. After suitable functional group manipulation, this allylic alcohol
is subjected to enolate Claisen rearrangement (as propionate 22) to give allylsulfide 23, having
three stereocenters with a 1,4,5-relationship. Further functional group manipulation and one-
carbon homologation converts this intermediate into 26, which is oxidized and subjected to Evans-
Mislow allylsulfoxide rearrangement to obtain 27, having three stereocenters with a 1,2,5-
relationship. The synthesis of myxalamide A was completed by converting aldehyde 30 into dienyne
40. Alkyne 40 was hydroborated with catechol borane, and the resulting E-vinylborane was
subjected to Suzuki coupling with the Z-iodo triene 9 to provide myxalamide A (1).
As a result of a great amount of research in the 1970s
and 1980s, the addition of preformed enolates to alde-
hydes to obtain R-substituted-â-hydroxy carbonyl com-
pounds is a well-established synthetic procedure.1 As
part of our own work in this area, we showed that one
can parlay the high 1,2-stereocontrol of the aldol reaction
into relative control of more remote stereocenters. For
example, by carrying out the aldol reaction on an R,â-
unsaturated aldehyde and following with an Ireland ester
enolate Claisen rearrangement, one can achieve 1,4- and
1,5-stereocontrol.2 When the starting aldehyde is a
â-alkylthio-R,â-unsaturated aldehyde, one can couple the
stereocontrolled aldol reaction with an ester enolate
Claisen rearrangement and an Evans-Mislow allylsul-
foxide rearrangement to prepare compounds having three
stereocenters with a 1,2,5-relationship.3 In this article,
heart submitochondrial particles with an IC50 of 170 pm/
mg of protein. The phenalamides were isolated more
recently from Myxococcus stipiatus.5 Phenalamide A1
(stipiamide) (2) exhibits antifungal and antiviral proper-
ties as well as the ability to reverse P-glycoprotein-
mediated multidrug resistance.5,6
The myxalamides and the phenalamides have at-
tracted some synthetic attention. A partial synthesis of
myxalamide D (5) was reported by Cox and Whiting in
which an anti aldol reaction was used to set the C12/
C13 stereochemistry.7 In addition, a total synthesis of
phenalamide A1 (2) was recently described.6 In this
approach, an asymmetric alkylation set the distal ste-
reocenter (C16), while an asymmetric crotylboration
created the anti stereorelationship at C12/C13. In
we report a further application of this strategy for the
total synthesis of myxalamide A (1).
Myxalamide A is one of a growing number of polyene
antibiotics, which includes the phenalamides as well as
the myxalamides. The four myxalamides were isolated
from the gliding bacteria Myxococcus xanthus, and the
most abundant of the group, myxalamide B (3), was found
to be a potent electron-transport inhibitor and exhibited
antibiotic and antifungal activity.4 Myxalamide B was
shown to inhibit NADH oxidation at complex I in beef
(1) Heathcock, C. H. In Asymmetric Synthesis; Morrison, J . D., Ed.;
Academic Press: New York, 1984, Vol. 3.
(2) (a) Heathcock, C. H.; J arvi, E. T. Tetrahedron Lett. 1982, 23,
2825. (b) Heathcock, C. H.; J arvi, E. T.; Rosen, T. Tetrahedron Lett.
1984, 25, 243. (c) Heathcock, C. H.; Finkelstein, B. L. J . Chem. Soc.,
Chem. Commun. 1983, 919. (d) Heathcock, C. H.; Finkelstein, B. L.;
Aoki, T.; Poulter, C. D. Science (Washington, D.C.) 1985, 229, 862. (e)
Heathcock, C. H.; Radel, P. A. J . Org. Chem. 1986, 51, 4323. (f)
Heathcock, C. H.; Finklestein, B. L.; J arvi, E. T.; Radel, P. A.; Hadley,
C. R. J . Org. Chem. 1987, 53, 1922.
(3) Munchhof, M. J .; Heathcock, C. H. J . Org. Chem. 1994, 59, 7566.
(4) (a) J ansen, R.; Reifenstahl, G.; Gerth, K.; Reichenbach, H.; Ho¨fle,
G. Liebigs Ann. Chem. 1983, 1081. (b) Gerth, K.; J ansen, R.; Reifen-
stahl, G.; Ho¨fle, G.; Irschik, H.; Kunze, B.; Reichenbach, H.; Thierbach.
G. J . Antibiot. 1983, 36, 1150. (c) J ansen, R.; Sheldrick, W. S.; Ho¨fle,
G. Liebigs Ann. Chem. 1984, 78.
(5) (a) Kim, Y. J .; Furihata, K.; Yamanaka, S.; Fudo, R.; Seto, H. J .
Antibiot. 1991, 553. (b) Trowitzsch-Kienast, W.; Forche, E.; Wray, V.;
Riechenbach, H.; J unsmann, G.; Ho¨fle, G. Liebigs Ann. Chem. 1992,
659.
(6) (a) Andrus, M. B.; Lepore, S. D. J . Am. Chem. Soc. 1997, 119,
2327. (b) Andrus, M. B.; Lepore, S. D.; Turner, T. M. J . Am. Chem.
Soc. 1997, 119, 12159.
(7) (a) Cox, C. M.; Whiting, D. A. J . Chem. Soc., Perkin Trans. I
1991, 1901. (b) Cox, C. M.; Whiting, D. A. J . Chem. Soc., Perkin Trans.
I, 1991, 1907. (c) Cox, C. M.; Whiting, D. A. J . Chem. Soc., Perkin I,
1991, 660.
10.1021/jo9813742 CCC: $18.00 © 1999 American Chemical Society
Published on Web 12/11/1998

24 J . Org. Chem., Vol. 64, No. 1, 1999
Mapp and Heathcock
Sch em e 1
contrast to the other synthetic efforts, our approach to
myxalamide A (1) would parlay a single asymmetric
reaction early in the synthesis into the anti C12/C13
relationship as well as set the distal C16 stereocenter.
Retrosynthetically, formation of the pentaene late in
the synthesis seemed desirable given the reported sen-
sitivity of the natural product to light and oxygen.4 Our
early efforts utilized a Pd-mediated coupling of vinyl
bromide 6 and dienyne 7 followed by partial reduction
of the internal triple bond to install the sensitive pen-
taene system.
The initial syn 1,2-stereorelationship was to be set by
an Evans asymmetric aldol reaction (Scheme 1).10 To-
ward this end, aldehyde 13 was prepared from acrylate
1011 by treatment with tert-butyl thiolate to yield ester
11 followed by a reduction/oxidation sequence. Early
attempts at the aldol reaction employed the conditions
developed during the course of the ACRL Toxin IIIb
synthesis for substrates containing readily oxidizable
functional groups.3 However, while this procedure had
worked well for a less substituted aldehyde, only a
complex mixture of products was isolated from the aldol
reaction with aldehyde 13. As a result, the less sterically
encumbered aldehyde 15 was prepared,12,13 but analogous
results were obtained. Examination of the 1H NMR
spectra of the crude reaction mixtures from both aldol
reactions revealed additional resonances in the alkyl
region, thus suggesting that incomplete removal of di-
alkylboronates from the product aldols (17 or 18) at the
conclusion of the reaction could be the culprit. Treatment
of the crude reaction mixture with sodium perborate14
However, initial studies of this approach revealed that,
while the coupling proceeded smoothly, the reduction of
the alkyne was problematic.8 Therefore, attention turned
to the formation of the cis C6/C7 olefin prior to combining
the two halves of the natural product. Given the recent
report of Geˆnet and co-workers on the use of Suzuki
couplings in aqueous media to produce stereodefined
polyenes,9 a Suzuki coupling appeared to be an excellent
choice for the final step of the synthesis. Retrosyntheti-
cally, this led back to fragments 8 and 9. Alcohol 8 would
be prepared using the aldol-Claisen-Evans-Mislow
strategy, while triene 9 was viewed as being accessible
through a combination of metal-mediated couplings and
Wittig reactions.
(10) Gage, J . R.; Evans, D. A. Org. Synth. 1989, 68, 83.
(11) (a) Ford, M. C.; Waters, W. A. J . Chem. Soc. 1951, 1851. (b)
Aberhart, D. J .; Tann, C.-H. J . Chem. Soc., Perkin Trans. 1 1979, 939.
(12) For preparation of bromide 14, see: (a) Caubere, P. Bull. Soc.
Chim. Fr. 1964, 45, 3584. (b) Brande, E. A.; Evans, E. A. J . Chem.
Soc. 1955, 3324. (c) Fischetti, W.; Mak, K. T.; Stakem, G.; Kim, J .-I.;
Rheinhold, A. L.; Heck, R. F. J . Org. Chem. 1983, 48, 948.
(13) For oxidation of alcohol 14 using TPAP, see: Griffith, W. P.;
Ley, S. V. Aldrichimica Acta 1990, 23, 13.
(8) For details of this approach, see: Mapp, A. K. Ph.D. Thesis,
University of California, Berkeley, 1997.
(9) Geneˆt, J . P.; Linquist, A.; Blart, E.; Mouries, V.; Savignac, M.
Tetrahedron Lett. 1995, 36, 1443.
(14) Kabalka, G. W.; Shoup, T. M.; Goudgaon, N. M. J . Org. Chem.
1989, 54, 5930.

Total Synthesis of Myxalamide A
J . Org. Chem., Vol. 64, No. 1, 1999 25
Sch em e 2
did remove the alkyl boronates, but the yields were
variable due to oxidation of the vinyl sulfide moiety.
However, the exchange resin IRA-74315 cleanly and
effectively removed all dialkyl boronates.16,17 Because the
product aldol was somewhat sensitive to silica gel chro-
matography, consistently higher yields were obtained
when the crude reaction mixture was directly submitted
to LiBH4 reduction conditions.18 This resulted in a 90%
overall yield for the production of diol 19 from the two-
step procedure. At this stage, the desired syn 1,2-
stereorelationship had been obtained. Although either
vinyl sulfide (n-propyl or tert-butyl) could be used in the
subsequent chemistry, efforts focused on the tert-butyl
sulfide due to its enhanced stability to varied reaction
conditions.
In preparation for a Claisen rearrangement,19 the
primary alcohol of diol 19 was protected as the TBS ether
and the secondary alcohol was then acylated using
propionyl chloride to provide ester 22 (Scheme 2). Treat-
ment of ester 22 with LDA followed by TBSCl produced
the E-silyl ketene acetal, which upon warming to room
temperature underwent the desired rearrangement. The
crude silyl ester was hydrolyzed, and the resulting
carboxylate was treated with diazomethane to yield
methyl ester 23 with the appropriate 1,2- and 1,5-
stereorelationships. In addition, the allylic thioether was
positioned for the anticipated Evans-Mislow rearrange-
ment to set the final 1,2-stereorelationship.
of the reaction mixture to room temperature and addition
of tosylate 25, 76% of the desired product was isolated
along with significant amounts of alcohol 24, presumably
resulting from attack on the tosylate. After some ex-
perimentation with reaction conditions, it was found that
when the cuprate was formed at a higher temperature
(0 °C instead of -78 °C) the yield increased to 94% with
little or no formation of alcohol 24. The stage was then
set for the Evans-Mislow rearrangement to produce the
final anti 1,2-stereorelationship.20 Oxidation of sulfide
26 with m-CPBA afforded a mixture of diastereomeric
sulfoxides, which were then treated with the thiophile
P(OMe)3 to provide alcohol 27 in good overall yield. At
this juncture, the necessary stereorelationships had been
successfully created by the aldol-Claisen-Evans-Mis-
low strategy in 12 steps and 32% overall yield from
acrylate 10.
With the stereocenters in place, it only remained to
elaborate the primary alcohol of 27 to an enyne. It
seemed possible to install the enyne in one step using a
phosphonate such as 31. Related phosphonates have
been used to produce disubstituted E-enynes in good
yields and high selectivity.21 In preparation for the
transformation, the secondary alcohol in 27 was protected
as a TIPS ether followed by selective cleavage of the TBS
silyl ether under acidic conditions. Alcohol 29 was then
oxidized using TPAP/NMO22 in high yield. Aldehyde 30
was added to a solution of the lithium anion of phospho-
nate 31,23 and the reaction proceeded smoothly to produce
a 1:1 mixture of E/Z enynes (Scheme 3). The use of a
bulkier phosphonate (32) changed the selectivity slightly
but in the undesired direction (1:1.5, E/Z).
Prior to exploring the Evans-Mislow rearrangement,
it was necessary to extend the carbon framework. To-
ward this end, methyl ester 23 was reduced to alcohol
24, which was then treated with tosyl chloride. The
displacement of the tosylate proved to be quite sensitive
to reaction conditions. When methyllithium was added
to a slurry of CuCN cooled to -78 °C followed by warming
Although formation of the enyne had not proven to be
selective, it appeared likely that treatment of aldehyde
30 with a stabilized ylide or phosphonate would produce
the trisubstituted C10-C11 bond with higher selectivity
and subsequent functional group manipulations could
(15) Hicks, K. B.; Simpson, G. L.; Bradbury, A. G. W. Carbohydr.
Res. 1986, 147, 39.
(16) Evans, D. A.; Ng, H. P.; Clark, S. J .; Rieger, D. L. Tetrahedron
1992, 48, 2127.
(20) (a) Tang, R.; Mislow, K. J . Am. Chem. Soc. 1970, 92, 2100. (b)
Evans, D. A.; Andrews, G. C.; Sims, C. L. J . Am. Chem. Soc. 1971, 93,
4956. (c) Evans, D. A.; Andrews, G. C. J . Am. Chem. Soc. 1972, 94,
3672.
(21) For example, see: (a) Nicolaou, K. C.; Veale, C. A.; Webber, S.
E. J . Am. Chem. Soc. 1985, 107, 7515. (b) Gibson, A. W.; Humphrey,
G. R.; Kennedy, D. J .; Wright, S. H. B. Synthesis 1991, 414.
(22) Griffith, W. P.; Ley, S. V.; Whitcombe, G. P.; White, A. D. J .
Chem. Soc., Chem. Commun. 1987, 1625.
(23) Phosphonate 31 was prepared by heating trimethyl phosphite
and 3-(trimethylsilyl)propynyl bromide followed by deprotonation with
BuLi and subsequent treatment with MeI. Phosphonate 32 was
prepared in analogous fashion starting with triisopropyl phosphite.
(17) Alternatively, reasonable yields of aldols 17 and 18 could be
obtained by using TiCl4 as the Lewis acid followed by a basic (NaHCO3)
workup, but careful chromatography of the crude product was neces-
sary due to the presence of undesirable diastereomers. Evans, D. A.;
Rieger, D. L.; Bilodeau, M. T.; Urpi, F. J . Am. Chem. Soc. 1991, 113,
1047.
(18) Penning, T. D.; Djuric, S. W.; Haack, R. A.; Kalish, V. J .;
Miyashiro, J . M.; Rowell, B. W.; Yu, S. S. Synth. Comm. 1990, 20, 307.
(19) (a) Ireland, R. E.; Mueller, R. H.; Willard, A. K. J . Org. Chem.
1976, 41, 986. (b) Ireland, R. E.; Mueller, R. H.; Willard, A. K. J . Am.
Chem. Soc. 1976, 98, 2868.

26 J . Org. Chem., Vol. 64, No. 1, 1999
Mapp and Heathcock
Sch em e 3
Sch em e 4
Sch em e 5
then be used to install the triple bond. As anticipated,
reaction of aldehyde 30 with tributylphosphonium ylide
3424 produced the unsaturated ester in an improved 2:1
(E/Z) ratio and good yield (56% of isomerically pure 35).
Similar results were obtained by using the anion of
triethyl-2-phosphonopropionate. In an attempt to further
increase selectivity, diisopropyl-2-phosphonopropionic
acid ethyl ester was prepared following the procedure of
Kishi25 and the anion was added to aldehyde 30. While
the ratio did improve to 4:1 E/Z, the reaction was sluggish
and proton transfer was a significant competing reaction.
This is not surprising given the hindered nature of both
the nucleophile and the electrophile. The unsaturated
ester 35 was converted into enyne 39 by reduction with
DIBALH to allylic alcohol 37, oxidation to the aldehyde,
and treatment with the Gilbert-Seyferth phosphonate.26
Removal of the silyl protecting group was then achieved
by treatment with TBAF in THF to yield 40 (Scheme 4).
With enyne 40 in hand, attention turned to the
synthesis of the requisite iodotriene 9. Diene 41 ap-
peared to be a good starting point for the synthesis, since
oxidation of the allylic alcohol followed by a Wittig
reaction would provide the desired cis vinyl iodide.
Initially, we attempted to prepare diene 41 by a Heck
reaction between allyl alcohol and bromide 10 as previ-
ously reported in the literature (Scheme 5);27 however,
we consistently isolated an inseparable mixture of iso-
mers from the reaction. This was easily circumvented
by hydroboration of propargyl ether 4228 followed by a
Suzuki coupling with bromide 10 to produce the desired
diene 43 as one stereoisomer. The trityl group was
removed by treatment with Amberlyst-15. Oxidation of
the allylic alcohol proceeded smoothly and the vinyl
iodide was then installed by a Wittig reaction to provide
44 (5:1, Ε,E,Z/E,E,E) using the conditions of Stork.29
While this was certainly an acceptable ratio, separation
of the isomers by a variety of purification methods (flash
chromatography, HPLC, crystallization) proved difficult.
For this reason, an alternate approach to iodotriene 9
was explored.
Our revised approach incorporated conversion of a cis
vinyl stannane, readily available by reduction of the
corresponding acetylenic stannane, to the vinyl iodide by
treatment with iodine (Scheme 6). Silylation of 2-penten-
4-yn-1-ol proceeded in a straightforward manner.30 The
alkyne was then deprotonated and treated with tribu-
tyltin chloride to furnish the alkynyl stannane. The
crude product was treated with Cp2ZrHCl followed by
water31 and then desilylated using TBAF to provide
alcohol 47 in good overall yield.
Prior to installation of the trisubstituted olefin, allylic
alcohol 47 was oxidized using MnO2. Treatment with the
anion of triethyl-2-phosphonopropionate provided the
desired triene 48 in high yield. Following the procedure
of Andrus,6 the ester was hydrolyzed and the carboxylate
was activated and treated with (S)-2-amino-1-propanol
to produce amide 49 as one isomer. Upon treatment with
iodine, tin-halogen exchange occurred to give a 2:1
mixture of isomers of triene 9. This was unexpected but
(24) (a) Trippett, S.; Walker, D. M. J . Am. Chem. Soc. 1961, 83, 6.
(b) Garner, P.; Ramakanth, S. J . Org. Chem. 1987, 52, 2631.
(25) Nagaoka, H.; Kishi, Y. Tetrahedron 1981, 37, 3873.
(26) (a) Seyferth, D.; Marmor, R. S.; Hilbert, P. J . Org. Chem. 1971,
36, 1379. (b) Gilbert, J . C.; Weerasooriya, U. J . Org. Chem. 1979, 44,
4997.
(27) Kao, L.-C.; Stakem, F. G.; Patel, B. A.; Heck, R. A. J . Org. Chem.
1982, 47, 1267.
(28) (a) Zeile, K.; Meyer, H. Chem. Ber. 1942, 75, 356. (b) Zhou, Z.-
L.; Huang, Y.-Z.; Shi, L.-L.; Hu, J . J . Org. Chem. 1992, 57, 6598.
(29) Stork, G.; Zhao, K. Tetrahedron Lett. 1989, 30, 2173.
(30) Crevisy, C.; Couturier, M.; Dugave, C.; Dory, Y. L.; Deslong-
champs, P. Bull. Chim. Soc. Fr. 1995, 132, 360.
(31) Lipshutz, B. H.; Keil, R.; Barton, J . C. Tetrahedron Lett. 1992,
33, 5861.

Total Synthesis of Myxalamide A
J . Org. Chem., Vol. 64, No. 1, 1999 27
Sch em e 6
Sch em e 7
completion of the hydroboration, combined with iodotri-
ene 9 and catalytic palladium acetate. After 6 h at room
temperature, the reaction appeared to be complete.
1
Analysis of the crude material by H NMR spectroscopy
revealed that it was a mixture of myxalamide A along
with the all-trans isomer and triene 9. There was some
concern that the separation of myxalamide A from its all-
trans isomer would be difficult given the inability of
Andrus and co-workers to separate phenalamide A1 (2)
from its all-trans isomer.6 However, after flash chroma-
tography on silica gel, the all-trans isomer of 1 had been
removed, and further purification by HPLC removed the
remaining impurities to yield myxalamide A in 44% yield
for the two steps (hydroboration and coupling). The
synthetic myxalamide A corresponded spectroscopically
to the natural material (1H, 13C, IR, optical rotation,
HRMS) (Scheme 7).33
In conclusion, the total synthesis of myxalamide A was
accomplished in 22 steps (longest linear sequence) and
4.9% overall yield from bromide 10. The completion of
the synthesis not only expanded the utility of the aldol-
Claisen-Evans-Mislow strategy but also emphasized
the usefulness of the Suzuki coupling for the preparation
of polyene-containing natural products. Given these
results, the synthetic strategy would be useful for the
preparation of other members of the polyene natural
product family as well as related diastereoisomers for
biological evaluation.
in retrospect not surprising. If the mechanism of the
reaction includes an intermediate with cationic character,
the extended conjugation present in triene 49 could
stabilize the intermediate long enough to allow isomer-
ization. Thus, the reaction was repeated using ester 48
as a substrate, and indeed the isomeric ratio increased
to 6:1. The course of the exchange reaction was moni-
tored by 1H NMR, and the results indicated that the
isomerization occurred during the tin-halogen exchange
as opposed to before or after the reaction, since the
starting material remained as one isomer throughout the
course of the reaction and the product ratio did not alter
upon extended treatment with excess iodine. Further-
more, the treatment of diene 47 with iodine yielded a
1
greater than 15:1 ratio of isomers as detectable by H
NMR.
Ack n ow led gm en t. This work was supported by a
research grant from the United States Public Health
Service (AI 15027).
A similar sequence of reactions was then used to
prepare the target iodotriene. Diene 47 was transformed
into iodotriene 44 by treatment with iodine, oxidation
using MnO2, and olefination. Hydrolysis of the ester then
proceeded smoothly. The crude reaction mixture was
filtered through a plug of silica gel, and the isolated
material was treated with pivaloyl chloride followed by
(S)-2-amino-1-propanol to yield the desired amide 9.
Although great care was taken to avoid light during
handling of the compounds, some isomerization inevita-
bly occurred. Therefore, the final iodotriene was gener-
ally isolated as a mixture of isomers (9:1 to 15:1), which
was stored at low temperature and handled in the
absence of light to prevent further isomerization.
In preparation for the final coupling reaction, enyne
40 was treated with 2 equiv of catecholborane32 and, upon
Su p p or tin g In for m a tion Ava ila ble: Full experimental
details (15 pages). This material is contained in libraries on
microfiche, immediately follows this article in the microfilm
version of the journal, and can be ordered from the ACS; see
any current masthead page for ordering information.
J O9813742
(32) Suseela, Y.; Prasad, A. S. B.; Periasamy, M. J . Chem. Soc.,
Chem. Commun. 1990, 446.
(33) One resonance in the 13C NMR spectrum of synthetic 1 did not
match that reported in the original isolation paper (ref 4a). However,
this discrepancy proved to be the result of a typographical error in
the isolation paper as confirmed by Dr. R. J ansen by personal
communication.