Feigrisolide C: Structural revision and synthesis

Full text of DOI:10.1021/jo051107d

Feigrisolide C: Structural Revision and
Synthesis
Woo Han Kim, Jae Hoon Jung, and Eun Lee*
Department of Chemistry, College of Natural Sciences,
Seoul National University, Seoul 151-747, Korea
eunlee@snu.ac.kr
Received June 2, 2005
FIGURE 1. Structures proposed for feigrisolide C.
SCHEME 1. Chemical Transformation of
Feigrisolide C
The original macrodiolide structure proposed for feigrisolide
C was incorrect. The true structure of feigrisolide C was
identified as (2′S,3′S,6′R,8′R)-homononactoyl (2R,3R,6S,8S)-
nonactic acid, which was confirmed by total synthesis.
acid with the characteristic ABX signals from C1 meth-
ylene group (δ 3.96 and 4.20, ∆δ g 0.2). Diacetates from
erythro nonactic acids should exhibit ABX signals differ-
ent from those of C1 methylene protons; ∆δ 0.1.4
(+)-Methyl homononactate (12) was synthesized from
methyl (R)-3-hydroxypentanoate (6)6 following the scheme
used for previous synthesis of (+)-methyl nonactate (14)7
(Scheme 2). The diacetate 13 obtained from 12 after
LiBH4 reduction/acetylation exhibited NMR spectra iden-
tical to those of B.
Feigrisolide C was isolated from the culture broth of
Streptomyces griseus (strain GT 051022)1 and was also
claimed to be a metabolite of Streptomyces sp. 6167 of
marine origin.2 The structure was proposed1 to be a 15-
membered macrodiolide 1 (originally intended to be 2)
incorporating 8-epi-nonactic acid (Figure 1). Both these
compounds were synthesized in our laboratory and found
to be different from feigrisolide C.3
A small sample of natural feigrisolide C was subjected
to exhaustive reduction with excess amount of lithium
aluminum hydride, and the crude products were acetyl-
ated4 to furnish two products A and B (Scheme 1).
If one assumes that the true structure of feigrisolide
C is epimeric to structure 1 or 2, then diacetate 3 should
represent A, and a tetraacetate epimeric to 4 or 5 should
correspond to B. Surprisingly, examination of the NMR
spectra of the major product5 B clearly indicated that it
was a diacetate originated from a threo homononactic
Examination of the NMR spectra8 of A indicated that
it was also a diacetate from a threo nonactic acid. The
diacetate 15, prepared from (+)-methyl nonactate (14),
furnished NMR spectra identical to those of A. For
comparison, the alternative diacetate 3 was prepared
from (+)-methyl 8-epi-nonactate (16),3 and it was clearly
different from A.
Four candidates for feigrisolide C then emerged: (2′S,
3′S,6′R,8′R)-homononactoyl (2S,3S,6R,8R)-nonactic acid
(17), ent-17, (2′S,3′S,6′R,8′R)-homononactoyl (2R,3R,
6S,8S)-nonactic acid (18), or ent-18 (Figure 2).
Spectroscopic differences between 17 and 18 were
anticipated to be small, and it was decided to prepare
both isomers for comparison. The benzyl ether 19 of
(2′S,3′S,6′R,8′R)-homononactic acid was prepared from
the methyl ester 11. It was reacted with methyl (2S,3S,
6R,8R)-nonactate (14) under Yamaguchi conditions to
furnish the diester 20, which was converted into 17 via
21. Using 19 and ent-14, we also prepared the other
* Corresponding author. Phone: +82-2-880-6646. Fax: +82-2-889-
1568.
(1) Tang, Y.-Q.; Sattler, I.; Thiericke, R.; Grabley, S.; Feng, X.-Z. J.
Antibiot. 2000, 53, 934-943.
(2) Sobolevskaya, M. P.; Fotso, S.; Havash, U.; Denisenko, V. A.;
Helmke, E.; Prokofeva, N. G.; Kuznetsova, T. A.; Laatsch, H.; Elyakov,
G. B. Chem. Nat. Compd. 2004, 40, 282-285.
(3) Kim, W. H.; Jung, J. H.; Sung, L. T.; Lim, S. M.; Lee, E. Org.
Lett. 2005, 7, 1085-1087.
(4) This strategy was used previously in our synthesis of pamamy-
cin-607 to check epimerization problems at C2 and C2′ of seco acids:
Jeong, E. J.; Kang, E. J.; Sung, L. T.; Hong, S. K.; Lee, E. J. Am. Chem.
Soc. 2002, 124, 14655-14662.
(6) Noyori, R.; Ohkuma, T.; Kitamura, M.; Takaya, H.; Sayo, N.;
Kumobayashi, H.; Akutagawa, S. J. Am. Chem. Soc. 1987, 109, 5856-
5858.
(7) Lee, E.; Choi, S. J. Org. Lett. 1999, 1, 1127-1128.
(8) We failed to get clean NMR spectra of A due to the lack of
material, but the characteristic features were easily recognized.
(5) The yield of A was much lower than that of B, and A was initially
overlooked as TLC separation was difficult.
10.1021/jo051107d CCC: $30.25 © 2005 American Chemical Society
Published on Web 09/01/2005
8190
J. Org. Chem. 2005, 70, 8190-8192

SCHEME 3. Synthesis of the Possible Structures
of Feigrisolide C
FIGURE 2. Possible structures of feigrisolide C.
SCHEME 2. Preparation of Diacetates 3, 13, and
15
water sediment sample, [R]25D 0 (c 0.84, CH2Cl2)), has the
same gross structure as that of feigrisolide C, but it was
claimed to consist of a racemic mixture of (()-nonactic
acid and (()-homononactic acid.11 A nonactoyl homonon-
actic acid of unknown stereochemistry was also recently
claimed to be isolated from a strain of Streptomyces
globisporus.12 Determination of the exact structure of
each of these compounds is quite difficult, and the present
studies suggest expedient ways of solving the problems.
Experimental Section
Ester 22. To a mixture of carboxylic acid 19 (63 mg, 0.21
mmol) and TEA (0.065 mL, 0.48 mmol) in THF (2.5 mL) was
added 2,4,6-trichlorobenzoyl chloride (0.05 mL, 0.35 mmol). The
reaction mixture was stirred for 2 h at room temperature. The
white precipitate that formed was removed by filtration under
N2 via cannula transfer to a glass pipet equipped with a septum
and a plug of glass wool. The THF filtrate was evaporated by a
stream of N2. The residue was diluted with benzene (5 mL), and
DMAP (75 mg, 0.63 mmol) was added. To this mixture was
added alcohol ent-14 (65 mg, 0.30 mmol) in benzene (2.5 mL).
The reaction mixture was stirred for 2 h at room temperature,
and the reaction was quenched with saturated NH4Cl solution
(8 mL). The reaction mixture was extracted with DCM (20 mL
× 3), and the organic extracts were dried over MgSO4, filtered,
and concentrated. Purification by flash column chromatography
(hexane-EtOAc, 2:1) gave ester 22 (98 mg (oil), 94%). Rf 0.50
candidate 18 via 22 and 23. Careful comparison studies
revealed that 18 exhibited spectroscopic characteristics
identical to those of feigrisolide C. The synthetic sample
of 18 had the same optical rotation value ([R]25D +16.2 (c
0.2, MeOH)) as that reported for feigrisolide C ([R]20
D
+17.2 (c 0.4, MeOH)), and the true structure of fei-
grisolide C was finally identified as 18 (Scheme 3). A
small sample of feigrisolide C was converted into the
methyl ester via reaction with diazomethane, and it
exhibited spectroscopic properties identical to those of
23,9 confirming the conclusion.
A literature search revealed that bonactin10 (isolated
from a Streptomyces sp. BD21-2 cultured from a shallow
(10) Schumacher, R. W.; Talmage, S. C.; Miller, S. A.; Sarris, K. E.;
Davidson, B. S.; Goldberg, A. J. Nat. Prod. 2003, 66, 1291-1293.
(11) This claim is probably erroneous if bonactin is
a single
(9) In 1H NMR spectra, the features around δ 1.7 were diagnostic
in differentiating between 17 and 18 and also their methyl esters 21
and 23.
compound.
(12) Rˇ ezanka, T.; Sp´ızˇek, J.; Prˇikrylova´, V.; Prell, A.; Dembitsky,
V. M. Tetrahedron 2004, 60, 4781-4787.
J. Org. Chem, Vol. 70, No. 20, 2005 8191

(hexane-EtOAc, 2:1). 1H NMR (300 MHz, CDCl3): δ 7.36-7.27
(m, 5H), 5.04-4.93 (m, 1H), 4.54 and 4.48 (AB q, 2H, J ) 11.3
Hz), 4.07-3.94 (m, 3H), 3.92-3.83 (m, 1H), 3.67 (s, 3H), 3.58-
3.51 (m, 1H), 2.55-2.43 (m, 2H), 2.02-1.92 (m, 4H), 1.77-1.72
(m, 2H), 1.69-1.42 (m, 8H), 1.23 (d, 3H J ) 6.3 Hz), 1.09 (d,
6H, J ) 7.0 Hz), 0.92 (t, 3H, J ) 7.4 Hz); 13C NMR (75 MHz,
CDCl3): δ 175.2, 174.2, 139.0, 128.2, 127.7, 127.3, 80.3, 80.1,
78.1, 76.6, 76.5, 71.5, 69.2, 51.5, 45.7, 45.3, 42.5, 40.8, 31.5, 31.3,
28.4, 28.3, 27.1, 20.5, 13.2, 13.1, 9.2. IR (neat): νmax ) 3450, 2972,
2877, 1738, 1497, 1456, 1377, 1259, 1196, 1167, 1061 cm-1. MS
m/z (FAB, relative intensity): 505 (M+ + 1, 49), 397 (10), 307
(24), 199 (100), 154 (64), 136 (46), 91 (79). HRMS (FAB) calcd
for C29H45O7 (M+ + 1) 505.3165, found 505.3169. [R]25D -15.0 (c
2.75, CHCl3).
of saturated NaHCO3 (0.15 mL) solution and water (0.11 mL).
The reaction mixture was extracted with CHCl3 (0.11 mL × 6).
The aqueous phase was acidified to pH 1 with 2 N HCl and
extracted with CHCl3 (0.22 mL × 8). The organic extracts were
dried over MgSO4, filtered, and concentrated. Purification by
flash column chromatography (CHCl3-MeOH, 9:1) gave (2′S,3′S,
6′R,8′R)-homononactoyl (2R,3R,6S,8S)-nonactic acid (18, 21 mg
(oil), 76%). Rf 0.40 (CHCl3-MeOH, 9:1). 1H NMR (300 MHz,
CDCl3): δ 5.09-5.03 (m, 1H), 4.17-4.11 (m, 1H), 4.05-3.91 (m,
3H), 3.80-3.72 (m, 1H), 2.52-2.46 (m, 2H), 2.04-1.97 (m, 4H),
1.83-1.46 (m, 11H), 1.25 (d, 3H J ) 6.3 Hz), 1.17 (d, 3H, J )
7.0 Hz), 1.11 (d, 3H, J ) 7.0 Hz), 0.93 (t, 3H, J ) 7.4 Hz); 13C
NMR (75 MHz, CDCl3): δ 176.8, 174.3, 81.1, 80.5, 77.2, 76.7,
70.4, 68.9, 45.4, 44.9, 42.2, 40.3, 31.0, 30.4, 29.8, 29.2, 28.8, 20.3,
13.7, 13.5, 10.1. IR (neat): νmax ) 3444, 2970, 2933, 2873, 1730,
1460, 1379, 1263, 1198, 1061 cm-1. MS m/z (FAB, relative
intensity): 401 (M+ + 1, 47), 307 (20), 289 (10), 199 (25), 154
Alcohol 23. Ester 22 (38 mg, 0.073 mmol) was dissolved in
MeOH (3.2 mL), and palladium on activated carbon (10% w/w,
16 mg) was added. The mixture was stirred under hydrogen
atmosphere for 30 min and then filtered through a filter paper.
After solvent evaporation, flash column chromatography (hex-
ane-EtOAc, 2:1) yielded alcohol 23 (28 mg (oil), 92%). Rf 0.41
(hexane-EtOAc, 1:3). 1H NMR (300 MHz, CDCl3): δ 5.05-4.94
(m, 1H), 4.16-4.12 (m, 1H), 4.02-3.95 (m, 2H), 3.93-3.84 (m,
1H), 3.78-3.67 (m, 1H), 3.69 (s, 3H), 2.79 (d, 1H J ) 4.8 Hz),
2.56-2.44 (m, 2H), 2.04-1.94 (m, 4H), 1.80-1.43 (m, 10H), 1.24
(d, 3H J ) 6.2 Hz), 1.10 (d, 6H, J ) 7.0 Hz), 0.93 (t, 3H, J ) 7.4
Hz); 13C NMR (75 MHz, CDCl3): δ 175.3, 174.2, 80.8, 80.4, 76.6,
76.4, 70.3, 69.4, 51.6, 45.4, 45.3, 42.4, 40.7, 31.3, 30.6, 30.1, 28.6,
28.5, 20.5, 13.3, 13.2, 10.1. IR (neat): νmax ) 2974, 1732, 1462,
1379, 1263, 1198, 1063 cm-1. MS m/z (FAB, relative intensity):
415 (M+ + 1, 64), 391 (27), 307 (19), 282 (68), 199 (54), 154 (100),
136 (71), 107 (25). HRMS (FAB) calcd for C22H39O7 (M+ + 1)
(100), 136 (76), 107 (33). HRMS (FAB) calcd for C21H37O7 (M+
+
1) 401.2539, found 401.2543. [R]25 +16.2 (c 0.24, CHCl3).
D
Acknowledgment. This work was supported by a
grant from MarineBio21, Ministry of Maritime Affairs
and Fisheries, Korea, and a grant from Center for
Bioactive Molecular Hybrids (Yonsei University and
KOSEF). Brain Korea 21 graduate fellowship grants to
W.H.K. are gratefully acknowledged. We also thank Dr.
I. Sattler (Hans-Kno¨ll-Institut fu¨r Naturstoff-Fors-
chung) for a sample of feigrisolide C.
Supporting Information Available: Experimental pro-
cedures, 1H NMR spectra of A, B, 3, 13, 15, 21, 23, feigrisolide
C, and feigrisolide C methyl ester, and 13C NMR spectra of
the intermediates. This material is available free of charge
via the Internet at http://pubs.acs.org.
415.2696, found 415.2706. [R]25 +6.0 (c 0.53, CHCl3).
D
(2′S,3′S,6′R,8′R)-Homononactoyl (2R,3R,6S,8S)-Nonactic
Acid (18). To a solution of alcohol 23 (28 mg, 0.067 mmol) in
HMPA (0.27 mL) was added lithium n-propyl mercaptide (0.48
M in HMPA, 0.027 mL). The reaction mixture was stirred for 2
h at room temperature. The reaction was quenched by addition
JO051107D
8192 J. Org. Chem., Vol. 70, No. 20, 2005