Enantioselective synthesis of (-)-terpestacin and (-)-fusaproliferin: Clarification of optical rotational measurements and absolute configurational assignments establishes a homochiral structural series

Article abstract of DOI:10.1021/ja020072l

An enantioselective synthesis of the syncytium formation inhibitor (-)-terpestacin (1, 19 steps, 5.8% yield from the allylation product of (R,R)-pseudoephedrine propionamide, 3) and the fungal metabolite (-)-fusaproliferin (2, 21 steps, 5.3% yield from 3) in their natural configurations is described. The route employs a series of stereoselective enolate alkylation reactions to establish the initial stereogenic center, set the quaternary carbon configuration, close the 15-membered ring, and introduce the side-chain residue with proper stereocontrol. Careful analysis of our synthetic materials alongside natural samples has revealed that several errors were made in the earlier measurements of optical rotation or in the absolute stereochemical assignments of these natural products. Clarifying all discrepancies, we show here that natural terpestacin (1) is levorotatory, not dextrorotatory as originally described, but was correctly assigned as the (1S,11S,15R,23S)-enantiomer. Fusaproliferin (2) is levorotatory, as reported, but is in fact the (1S,11S,15R,23S)-enantiomer and not the antipodal configuration originally assigned. Copyright

Full text of DOI:10.1021/ja020072l

Published on Web 03/23/2002
Enantioselective Synthesis of (-)-Terpestacin and (-)-Fusaproliferin:
Clarification of Optical Rotational Measurements and Absolute
Configurational Assignments Establishes a Homochiral Structural Series
Andrew G. Myers,* Michael Siu, and Feng Ren
Department of Chemistry and Chemical Biology, HarVard UniVersity, 12 Oxford Street,
Cambridge, Massachusetts 02138
Received January 15, 2002
Terpestacin (1) is a fungal metabolite that inhibits the formation
of syncytia (IC50 ) 0.46 mg/mL),1 multinucleated cell bodies that
are part of the pathology of AIDS infection.2 Relative stereochem-
ical assignments within terpestacin (1) were determined by X-ray
crystallographic analysis; absolute stereochemistry was assigned
by two different methods (exciton chirality method, 1H NMR
analysis of diastereomeric O-methylmandelate ester derivatives).1
Subsequently, an enantioselective synthesis of 1 proceeding in ∼38
steps from tri-O-acetyl-D-galactal was described, apparently con-
firming the absolute stereochemical assignment (synthetic: [R]D
liferin (2) is levorotatory, as reported, but is in fact the (1S,11S,-
15R,23S)-enantiomer,6 not the antipodal configuration originally
assigned.
Our synthetic route to 1 and 2 begins with the allylation product
3 (Scheme 1), formed from (R,R)-pseudoephedrine propionamide
in 93% yield and 99% de (25-g scale).7 Iodolactonization of 3
afforded the trans-iodolactone 4 selectively (trans:cis, 12:1, 86%
yield of pure trans-isomer after flash column chromatography).8
The iodolactonization reaction served to cleave the pseudoephedrine
amide bond8b,9 and also as a device to control the configuration of
the quaternary center of terpestacin (vide infra). Toward this end,
the iodide 4 was transformed into the triisopropylsilyl ether 5, in
an efficient two-step sequence, via an alcohol of previously proven
) +27, c 0.22, CHCl3, natural: [R]22 ) +26, c 0.5, CHCl3).3
D
Contemporaneous with reports of the isolation of 1, the natural
product fusaproliferin, an acetic acid ester of 1, was identified in
the fermentation broth of the maize pathogen Fusarium proliferatum.4a
A tentative assignment of its structure was made and later twice
redefined to include the assignment of absolute stereochemistry.4b,c
The structure finally proposed was enantiomeric with structure 2
shown below. Most recently, a third natural product was isolated
from cultures of Ulocladium that inhibited syncytium formation in
cells infected with respiratory syncytial virus.5 This natural product
was spectroscopically identical with terpestacin (1) but differed in
the sign (and magnitude) of its optical rotation, ([R]2D5 ) -16.5, c
0.29, CDCl3), on which basis it was assigned as the enantiomer of
terpestacin.5
absolute configuration ([R]2D6 ) -38.6, c 0.32, EtOH, lit.10 [R]2D0
)
-35.7, c 0.3, EtOH). For alkylative coupling with 5, the biselec-
trophile 7 was prepared from the known epoxy alcohol 6,11 as shown
in Scheme 1. Coupling of the components 5 and 7 and formation
of the quaternary carbon center of 1 was achieved diastereoselec-
tively (8.8:1) by the addition of 7 (1 equiv) to the potassium enolate
derived from 5 (1.5 equiv) at -78 °C. The trans-diastereomer (8)
was obtained in pure form by flash column chromatography (86%
yield, two steps). Alkaline hydrolysis of the lactone 8 followed by
acidification (pH ) 5), immediate esterification of the resulting
hydroxy acid, and oxidation with the Dess-Martin periodinane12
furnished the keto ester 9 in 82% yield. Dieckmann-like cyclization
then occurred upon slow addition of 9 to a suspension of sodium
hydride in N,N-dimethylformamide at 24 °C. Subsequent addition
of triethylamine, N,N-dimethylcarbamoyl chloride, and 4-(dimethyl-
amino)pyridine to the cyclization reaction at 0 °C led to the
exclusive acylation of the cyclization product to form the enol
carbamate 10 in 81% yield from 9. Reduction of 10 with
diisobutylaluminum hydride in tetrahydrofuran (THF) at -78 °C
formed a 1:1 mixture of diastereomeric allylic alcohols (85% yield);
addition of methyllithium (-78f0 °C) to this mixture led to
cleavage of the enol carbamate group of both isomers and, upon
workup, elimination of water to give a single product, the
cyclopentenone 11 (88% yield). To establish the stereochemistry
of the remaining trisubstituted olefin of terpestacin (C12-C13) and,
at the same time, provide a suitable substrate for macrocyclization,
the allylic epoxide 11 was transformed stereoselectively into the
allylic iodide 12. In the optimized procedure, treatment of 11 with
lithium iodide (5 equiv) in the presence of scandium trifluo-
romethanesulfonate (1 equiv, hydrate) in THF (-78f-25 °C)
followed by hydroxyl protection with tert-butyldimethylsilyl triflate
gave 12 in 89% yield. Deprotonation and macrocyclization occurred
upon treatment of a dilute solution of 12 (0.002 M, THF, 0 °C)
with Masamune’s base (lithium bis(dimethylphenylsilyl)amide) to
Here, we describe an enantioselective synthesis of terpestacin
(1, 19 steps, 5.8% yield from the pseudoephedrine amide 3) and
fusaproliferin (2, 21 steps, 5.3% yield from 3) in their natural
configurations, by a route that employs a series of stereoselec-
tive enolate alkylation reactions to establish the initial stereogenic
center, set the quaternary carbon configuration, close the 15-
membered ring, and introduce the side-chain residue with proper
stereochemistry. Careful analysis of our synthetic materials along-
side natural samples has revealed that several errors were made in
the earlier measurements of optical rotation or in the absolute
stereochemical assignments of these natural products. Clarifying
all discrepancies, we show here that natural terpestacin (1) is
levorotatory, not dextrorotatory as originally described, but was
correctly assigned as the (1S,11S,15R,23S)-enantiomer. Fusapro-
9
4230
J. AM. CHEM. SOC. 2002, 124, 4230-4232
10.1021/ja020072l CCC: $22.00 © 2002 American Chemical Society

C O M M U N I C A T I O N S
Scheme 1a
Scheme 2a
a Reaction conditions: (a) CH3CH2CO2t-Bu, LDA, THF, -78 °C, 94%;
(b) i. Red-Al, THF, -78f0 °C; HOAc; ii. Red-Al, toluene, -78f24 °C,
75%; (c) Martin sulfurane, CH2Cl2, -78 °C, 89%; (d) DMDO, acetone,
-24 °C; (e) CF3CO2H, Et2O; Et2NH; (f) K2CO3, MeOH, 73%; (g) 1 N
HCl, THF, 94%; (h) AcCl, i-Pr2NEt, CH2Cl2, -78 °C, 95%; (i) n-PrNH2,
toluene, 0 °C, 96%.
a Reaction conditions: (a) I2, THF, H2O, 96% (trans:cis 12:1); (b)
CsO2CCF3, DMF, 90 °C; Et2NH, 95%; (c) TIPSCl, imidazole, DMF, 98%;
(d) (COCl)2, DMSO, CH2Cl2; Et3N, -78f0 °C; (e) NaHMDS, Ph3PCH3Br,
THF, 93% (2 steps); (f) K2CO3, MeOH, 99%; (g) MsCl, Et3N, THF, -45
°C; LiBr, 0 °C; (h) KHMDS, THF, -78 °C; 86% (2 steps); (i) KOH, EtOH;
CH2N2; Dess-Martin periodinane, pyridine, CH2Cl2, 82%; (j) NaH, DMF;
(CH3)2NCOCl, DMAP, Et3N, 81%; (k) DIBAL, THF, -78 °C, 85%; (l)
CH3Li, Et2O, -78f0 °C, 88%; (m) LiI, Sc(OTf)3, THF, -25 °C, 92%;
TBSOTf, 2,6-lutidine, THF, -78 °C, 97%; (n) LiN(Si(CH3)2Ph)2, THF, 0
°C, 53% (trans:cis 4.8:1).
ester group of 14 proceeded with concomitant cleavage of the
triisopropylsilyl enol ether, providing the tricyclic hemiketal 15
(75% yield). Completion of the synthesis of 1 was then ac-
complished by the following sequence, developed after extensive
experimentation. Treatment of the hemiketal 15 with the Martin
sulfurane produced the cyclic enol ether 16 in 89% yield. Epoxi-
dation of 16 with dimethyldioxirane in acetone at 0 °C then
produced an unstable epoxide intermediate; epoxide opening with
trifluoroacetic acid at 0 °C, followed by treatment of the resulting
trifluoroacetate with excess diethylamine gave the presumed
intermediate 17. Base-mediated isomerization-dehydration of 17
(potassium carbonate, MeOH, 24 °C, 36 h) followed by acidic
cleavage of the tert-butyldimethylsilyl group then gave synthetic 1
in 69% yield from 16, as a white amorphous powder. Crystallization
from methanol provided a single-crystal suitable for X-ray analysis,
confirming all relative stereochemical assignments. Surprisingly,
however, the specific rotation of our synthetic sample deviated both
in sign and magnitude from values reported in both the initial
isolation study1 and the prior synthetic route3b (Table 1) but closely
matched data from the second isolation report5 of an “enantiomeric”
terpestacin. Analysis of an authentic sample of terpestacin obtained
from Arthrinium sp. FA1744 (original source) showed that, in fact,
the natural material was levorotatory, not dextrorotatory as originally
reported (the small sample size precluded accurate determination
of the magnitude of the rotation). Furthermore, synthetic and natural
afford the trans-fused alkylation product 13 selectively (53%,
diastereomeric ratio: 4.8:1).
With successful construction of the trans-fused [13.3.0]-bicyclic
ring system of terpestacin, all that remained to complete the
synthesis of 1 and 2 was to introduce the three-carbon side-chain
residue with proper stereochemistry, followed by formal oxidative
1,3-transposition within the five-membered ring. The first objective
was readily attained by 1,2-addition of the (Z)-enolate derived from
tert-butyl propionate and lithium diisopropylamide to the enone
13, giving rise to a single aldol addition product, the hydroxy ester
14, in 94% yield (Scheme 2). â-Face addition, as indicated in
structure 14, was presumed on the basis of steric considerations
and was supported by nOe data for 15 (vide infra); the more
important issue of the stereochemistry of the C23 center was proven
conclusively by the ultimate conversion of 14 to 1 (confirmed by
independent X-ray analysis). Two-step reduction of the tert-butyl
9
J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002 4231

C O M M U N I C A T I O N S
Table 1. Optical Rotational Measurements of Terpestacin and
Fusaproliferin
On the basis of the unambiguous stereo- and enantioselective
synthesis of terpestacin and fusaproliferin described herein and
comparisons of synthetic and natural materials, it is clear that these
natural products form a homochiral structural series. The terpest-
acins isolated from Arthrinium and Ulocladium fungal sources are
thus shown to be the same, and the absolute configuration of
fusaproliferin is revised as shown in structure 2.
Acknowledgment. Financial support from the National Science
Foundation is gratefully acknowledged. We thank Professors T.
Oki and U. Gra¨fe for kindly providing samples of natural terpesta-
cin. We also express our gratitude to Dr. A. Ritieni for a sample of
natural fusaproliferin and Dr. R. J. Staples for X-ray analysis of
our synthetic terpestacin. M.S. acknowledges a Glaxo summer
research fellowship.
terpestacin were found to provide identical circular dichroism spec-
tra (and were identical by all other spectroscopic and chromato-
graphic comparisons). A sample of terpestacin from Ulocladium
was also found to provide an identical CD spectrum. It is thus clear
that our synthetic 1 and 1 derived from both natural sources are
identical in all respects, including absolute stereochemistry.
A plausible explanation for what must be erroneous optical
rotational measurements reported in the original isolation work1
and the earlier synthesis3b was suggested when we discovered that
exposure of synthetic or natural 1 to certain lots of laboratory
chloroform (the solvent used for optical rotational measurements)
led to the formation of a chlorine-containing byproduct, tentatively
assigned as the tetrahydrofuran derivative 18 (further chlorination
of 18 occurred upon extended exposure).13 It was noted that
chloroform solutions producing 18 had been stored over granular
potassium carbonate. These solutions also tested positive for
chlorine in a semiquantitative spot test. Interestingly, identical lots
of chloroform that had been stored instead over 4 Å molecular
sieves did not test positive for chlorine, and gave stable solutions
of 1. Solutions of the pure chlorination product 18 in chloroform
were found to be dextrorotatory and the magnitude of the rotation
was found to exceed that of pure terpestacin (Table 1).
Supporting Information Available: Listings of spectral data (PDF)
and X-ray crystallographic file (CIF) for synthetic 1. This material is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) (a) Oka, M.; Iimura, S.; Tenmyo, O.; Yosuke, S.; Sugawara, M.; Ohkusa,
N.; Yamamoto, H.; Kawano, K.; Hu, S.-L.; Fukagawa, Y.; Oki, T. J.
Antibiot. 1993, 46, 367. (b) Iimura, S.; Osa, M.; Narita, Y.; Konishi, M.;
Kakisawa, H.; Gao, H.; Oki, T. Tetrahedron Lett. 1993, 34, 493. (c) Oka,
M.; Iimura, S.; Narita, Y.; Furumai, T.; Konishi, M.; Oki, T.; Gao, Q.;
Kakisawa, H. J. Org. Chem. 1993, 58, 1875.
(2) Levy, J. A. HIV and the Pathogenesis of AIDS; American Society for
Microbiology: Washington, DC, 1998 pp 121-125 and references therein.
(3) Racemic synthesis: (a) Tatsuta, K.; Masuda, N.; Nishida, H. Tetrahedron
Lett. 1998, 39, 83. Enantioselective synthesis: (b) Tatsuta, K.; Masuda,
N. J. Antibiot. 1998, 51, 602. Model studies related to the synthesis of 1:
(c) Takeda, K.; Nakajima, A.; Yoshii, E. Synlett 1995, 249. (d) Mermet-
Mouttet, M.-P.; Gabriel, K.; Heissler, D. Tetrahedron Lett. 1999, 40, 843.
(4) (a) Randazzo, G.; Fogliano, V.; Ritieni, A.; Mannina, L.; Rossi, E.;
Scarallo, A.; Segre, A. L. Tetrahedron 1993, 49, 10883 (Fusaproliferin
was originally named proliferin). (b) Manetti, C.; Fogliano, V.; Ritieni,
A.; Santini, A.; Randazzo, G.; Logrieco, A.; Mannina, L.; Segre, A. L.
Struct. Chem. 1995, 6, 183. (c) Santini, A.; Ritieni, A.; Fogliano, V.;
Randazzo, G.; Mannina, L.; Logrieco, A.; Benedetti, E. J. Nat. Prod. 1996,
59, 109.
(5) Schlegel, B.; Schmidtke, M.; Do¨rfelt, H.; Kleinwa¨chter, P.; Gra¨fe, U. J.
Basic Microbiol. 2001, 41, 179.
(6) Terpestacin numbering.
(7) (a) Myers, A. G.; Yang, B. H.; Chen, H.; McKinstry, L.; Kopecky, D. J.;
Gleason, J. L. J. Am. Chem. Soc. 1997, 119, 6496. (b) Myers, A. G.;
Yang, B. H.; Chen, H.; Gleason, J. L. J. Am. Chem. Soc. 1994, 116, 9361.
(8) Stereoselective halolactonization of (()-R-substituted-γ,δ-unsaturated
amides: (a) Tamura, Y.; Mizutani, M.; Furukawa, Y.; Kawamura, S.;
Yoshida, Z.; Yanagi, K.; Minobe, M. J. Am. Chem. Soc. 1984, 106, 1079.
An asymmetric alkylation-halolactonization cleavage sequence was previ-
ously demonstrated using amides of chiral amines other than pseudoephe-
drine: (b) Moon, H.; Eisenberg, S. W. E.; Wilson, M. E.; Schore, N. E.;
Kurth, M. J. J. Org. Chem. 1994, 59, 6504.
(9) In work contemporaneous with this early stage in the development of our
synthetic route, bromolactonization of pseudoephedrine amides was shown
to provide a viable cleavage method: (a) Dragovich, P. S.; Prins, T. J.;
Zhou, R. J. Org. Chem. 1997, 62, 7872. (b) Sandham, D. A.; Taylor, R.
J.; Carey, J. S.; Fassler, A. Tetrahedron Lett. 2000, 41, 10091.
(10) Herdeis, C.; Lutsch, K. Tetrahedron: Asymmetry 1993, 4, 121.
(11) (a) Arm, C.; Pfander, H. HelV. Chim. Acta 1984, 67, 1540. (b) Lan, J.;
Liu, Z.; Yuan, H.; Peng, L.; Li, W.-D. Z.; Li, Y.; Chan, A. S. C.
Tetrahedron Lett. 2000, 41, 2181.
Acetylation of terpestacin (1) to form fusaproliferin (2) could
not be achieved directly but was accomplished by bisacetylation
followed by selective (enol) acetate cleavage (Scheme 2). Synthetic
2 was found to be identical in all respects to a sample of authentic
fusaproliferin, including circular dichroism measurements. Thus,
the absolute configuration of fusaproliferin must be revised as
shown in structure 2. Like 1, 2 was observed to undergo chlorinative
cyclization in chlorine-positive chloroform solutions, forming the
acetate ester corresponding to 18.
(12) Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277.
(13) Assignment of 18 based on 1H NMR, 1H homonuclear decoupling,
NOESY, and HRMS.
JA020072L
9
4232 J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002