Enantioselective synthesis of 15-epi-haterumalide NA methyl ester and revised structure of haterumalide NA

Article abstract of DOI:10.1021/ol0341804

(Matrix presented). The enantioselective synthesis of the enantiomer of the haterumalide NA methyl ester, a cytotoxic macrolide from an Okinawan sponge, was achieved from the threitol derivative in 26 steps. The key steps are the stereoselective construct

Full text of DOI:10.1021/ol0341804

ORGANIC
LETTERS
2003
Vol. 5, No. 6
957-960
Enantioselective Synthesis of
15-epi-Haterumalide NA Methyl Ester
and Revised Structure of Haterumalide
NA
Hideo Kigoshi,* Masaki Kita, Seiji Ogawa, Masahiro Itoh, and Daisuke Uemura
Department of Chemistry, UniVersity of Tsukuba, Tsukuba, Ibaraki 305-8571, Japan,
and Research Center for Materials Science and Department of Chemistry,
Graduate School of Science, Nagoya UniVersity, Chikusa, Nagoya 464-8602, Japan
kigoshi@chem.tsukuba.ac.jp
Received February 3, 2003
ABSTRACT
The enantioselective synthesis of the enantiomer of the haterumalide NA methyl ester, a cytotoxic macrolide from an Okinawan sponge, was
achieved from the threitol derivative in 26 steps. The key steps are the stereoselective construction of a chloroolefin unit and the intramolecular
Reformatsky-type reaction. This synthesis revised the absolute stereochemistry of haterumalide NA.
Haterumalide NA is a macrolide isolated from the Okinawan
sponge Ircinia sp. This compound exhibited a cytotoxicity
against P388 cells with an IC50 of 0.32 µg/mL.1 The gross
structure and stereochemistry were elucidated by the spec-
troscopic analysis and the modified Mosher’s method as
structural formula 1 (R ) H). The structural features of this
compound are a 14-membered macrolide involving a trans-
disubstituted tetrahydrofuran ring, a Z-chloroolefin, and a
â,γ-unsaturated acid moiety. The structurally related
haterumalide B2 and oocydin A3 were isolated from an
Okinawan ascidian and a South American epiphyte, respec-
tively, and their stereostructures have not been fully estab-
lished. It is noteworthy that haterumalide NA was recently
isolated from a soil bacterium.4 We describe herein the
enantioselective synthesis of the ent-haterumalide NA methyl
ester, which revises the initially assigned stereostructure.
The haterumalide NA methyl ester (1: R ) Me) can be
logically divided into the macrolide unit 3 and the side chain
unit 2 (Scheme 1). The side chain unit 2 can be easily
prepared from the corresponding carboxylic acid 30 (see
Scheme 4).5 The macrocyclic structure of 3 can be estab-
lished by lactonization of the seco acid 5 or by the
intramolecular Reformatsky-type reaction of the bromo ester
derivative 4. These precursors, 4 and 5, can be synthesized
from a common intermediate 6, which can be prepared from
the tetrahydrofuran unit 8 by a coupling reaction with 7.
Scheme 2 summarizes the synthesis of the tetrahydrofuran
unit 8. The mono-MPM ether 9 was synthesized from
commercially available (+)-2,3-O-isopropylidene-L-threitol.6
After transformation into the corresponding iodide, C1
homologation was effected by using the FAMSO7 carbanion
* Address correspondence to this author at the University of Tsukuba.
(1) Takada, N.; Sato, H.; Suenaga, K.; Arimoto, H.; Yamada, K.; Ueda,
K.; Uemura, D. Tetrahedron Lett. 1999, 40, 6309.
(2) Ueda, K.; Hu, Y. Tetrahedron Lett. 1999, 40, 6305.
(3) Strobel, G.; Li, J.-Y.; Sugawara, F.; Koshino, H.; Harper, J.; Hess,
W. M. Microbiology 1999, 145, 3557.
(4) Thaning, C.; Welch, C. J.; Borowicz, J. J.; Hedman, R.; Gerhardson,
B. Soil Biol. Biochem. 2001, 33, 1817.
(5) Thibonnet, J.; Launay, V.; Abarbri, M.; Duchene, A.; Parrain, J.-L.
Tetrahedron Lett. 1998, 39, 4277. We synthesized 30 from 3-butyn-1-ol in
2 steps: (1) Cp2ZrCl2, Me3Al then I2 and (2) Jones oxidation.
(6) Yadav, J. S.; Deshpande, P. K.; Sharma, G. V. M. Tetrahedron 1990,
46, 7033.
10.1021/ol0341804 CCC: $25.00 © 2003 American Chemical Society
Published on Web 02/26/2003

42%).9 The desired trans-tetrahydrofuran 14 was quantita-
tively converted into the bromide 8 in three steps.
Scheme 1. Retrosynthetic Analysis of Haterumalide NA
3-Butyn-1-ol (16) was transformed into the E-alkenylsilane
18 (67%) following a reported procedure10 (Scheme 3). The
Methyl Ester
Scheme 3. Synthesis of Seco Acid 5a
to afford sulfoxide 10 (81%). Sequential acidic methanolysis8
and hydrolysis afforded the hemiacetal 12 in 57% yield.
Wittig reaction of 12 and cyclization provided the 5.3:1
diastereomeric mixture of tetrahydrofurans, which could be
separated after silylation to afford the desired trans-tetra-
hydrofuran 14 (70%) and the cis-isomer. The latter could
be isomerized into the 1:1 mixture (isolation yield of 14:
a Reagents and conditions: (a) DHP, p-TsOH (93%). (b) TMSCl,
n-BuLi, ether-hexanes (74%). (c) (i) DIBAL, ether-hexanes; (ii)
pyridine, ether; (iii) Br2, CH2Cl2 (91%). (d) hν, Br2, pyridine,
CH2Cl2 (99%). (e) (i) 7, s-BuLi, THF-hexanes; (ii) 8, HMPA, THF
(68%). (f) NCS, H2O, DMF (45%). (g) AcOH, THF-H2O (80%).
(h) Dess-Martin periodinane, CH2Cl2. (i) 28, KHMDS, 18-crown-
6, THF-toluene (75%, 2 steps). (j) DIBAL, toluene (100%). (k)
Dess-Martin periodinane, CH2Cl2. (l) 29, t-BuMgCl, THF (57%,
3S:3R ) 19:1, 2 steps). (m) Al-Hg, THF-H2O (86%). (n) Ac2O,
pyridine (99%). (o) HF-py, pyridine, THF (93%). (p) TMSOTf,
2,6-lutidine, CH2Cl2 (90%). (q) TBDPSCl, DMAP, Et3N, CH2Cl2.
(r) AcOH, THF-H2O (47%, 2 steps).
Scheme 2. Synthesis of trans-2,5-Dialkyl Tetrahydrofuran
Unit 8a
E-alkenylsilane 18 was photochemically isomerized to the
Z-isomer 7 in 99% yield. The coupling reaction between the
tetrahydrofuran unit 8 and the carbanion generated from the
Z-alkenylsilane 7 and sec-butyllithium was accomplished to
give compound 6 in 68% yield.
There are only a few published procedures for the
stereoselective preparation of chloroolefins. We modified the
a Reagents and conditions: (a) p-TsCl, pyridine. (b) NaI, CaCO3,
acetone (93%, 2 steps). (c) FAMSO, n-BuLi, THF-hexanes (87%).
(d) concentrated HCl, MeOH (8:92) (58%). (e) 1 M HCl aq, THF
(1:1) (99%). (f) Ph3PdCHCO2Me, MeCN. (g) NaOMe, MeOH.
(h) TBSCl, imidazole (70%, 3 steps). (i) LiAlH4, THF (100%). (j)
p-TsCl, pyridine. (k) LiBr, DMF (100%, 2 steps).
(7) Ogura, K.; Tsuchihashi, G. Tetrahedron Lett. 1971, 34, 3151.
(8) Dondoni, A.; Fantin, G.; Foganolo, M.; Merino, P. Tetrahedron Lett.
1990, 31, 4513.
(9) The stereochemistry was determined by the coupling constants from
1
their H NMR data and the NOE experiments.
(10) (a) Miller, R. B.; Al-Hassan, M. I. J. Org. Chem. 1983, 48, 4113.
(b) Zweifel, G.; Lewis, W. J. Org. Chem. 1978, 43, 2739.
958
Org. Lett., Vol. 5, No. 6, 2003

procedure11 for conversion of an alkenylsilane to a bromo-
olefin for preparation of the chloroolefin 19. After several
attempts, we found that the addition of a catalytic amount
of water was important for the reaction to be reproducible.
Acidic hydrolysis of 19 gave 20 and subsequent Dess-
Martin oxidation afforded a labile aldehyde, which was
converted into the Z-conjugated ester 21 by using the Still-
modified Horner-Emmons reaction12 (60%, three steps). The
DIBAL reduction of 21 gave the allylic alcohol 22 (100%),
which was oxidized to a conjugated aldehyde. The asym-
metric aldol reaction13 with Corey’s sulfoxide 2914 provided
a hydroxysulfoxide,15 amalgam reduction of which gave the
desired hydroxy ester 23 (49%, three steps). The absolute
stereochemistry of the C-3 hydroxy group in 23 was
established by the modified Mosher’s method.16 After
acetylation, the protecting groups were removed to give the
dihydroxy acid 26 (92%, two steps), the primary hydroxy
group of which was protected as the TBDPS ether to afford
the seco acid 5 (47%, two steps). Thus, the precursors for
the macrolide unit 3 were in hand. However, all attempts at
macrolactonization of the dihydroxy acid 26 or the seco acid
5 to 3 failed under the Yamaguchi, Keck, and Mukaiyama-
Corey conditions.
which was oxidized to afford the conjugated aldehyde 4,
a precursor of the intramolecular Reformatsky-type reac-
tion, in 93% yield. Attempts toward the intramolecular
Reformatsky-type reaction are summarized in Table 1. The
Table 1. Intramolecular Reformatsky-Type Reaction of
Bromoester 4
(3S,4Z)/(3R,4Z)/(4E)
entry
conditions
SmI2, 0 °C
SmI2, -100 °C
Zn, B(OMe)3, rt
Zn, CuBr, Et2AlCl, -20 °C
Et2Zn-RhCl(PPh3)3
-20 °C
product
(yield)
1
2
3
4
5
37
37
37
37
37
-/-/100 (86%)
-/-/100 (7%)
b
-
-
-
c
b
6a
Et2Zn-RhCl(PPh3)3
-20 °C
35
20/72/8 (45%)
We next tried to cyclize the 14-membered ring using the
intramolecular Reformatsky-type reaction (Scheme 4). The
a After 4 h, Ac2O was added to trap the reactive product 37. b Complex
mixture. c Noncyclized reduced products were obtained: debromo-4 (65%)
and the corresponding debromo-allyl alcohol (23%).
Scheme 4. Synthesis of Haterumalide NA Methyl Ester 36a
17
cyclization with SmI2 provided the cyclic compounds in
good yields (86%, 3S:3R ) 1:1); however, the stereochem-
istry of the C-4 double bond was totally isomerized into trans
(entry 1). The reaction at lower temperature also gave the
same trans-products in a lower yield (entry 2). The molecular
mechanics calculation indicated that the desired cis-
compound 35 was less stable (7.5 kJ/mol) than the trans-
compound.18 This isomerization might be due to the allylic
radical nature of the transition state and/or the reactive
intermediates. Therefore, we investigated the cyclization with
zinc reagents apt to effect the two-electron reduction. The
reactions under the standard conditions19,20 afforded no
cyclized compounds (entries 3 and 4). The reaction under
Honda’s conditions21 with Et2Zn-RhCl(PPh3)3 resulted in
(11) Tamao, K.; Akita, M.; Maeda, K.; Kumada, M. J. Org. Chem. 1987,
52, 1100.
(12) Still, W. C.; Gennari, C. Tetrahedron Lett. 1983, 24, 4405.
(13) (a) Corey, E. J.; Weigel, L. O.; Chamberlin, R.; Cho, H.; Hua, D.
H. J. Am. Chem. Soc. 1980, 102, 6613. (b) Mioskowski, C.; Solladie, G.
Tetrahedron 1980, 36, 227. (c) Solladie, G.; Moghadam, F.-M. J. Org.
Chem. 1982, 47, 91.
(14) Mioskowski, C.; Solladie, G. Tetrahedron Lett. 1975, 3341.
(15) Diastereoselectivity of this aldol reaction was 95:5, and the mixture
could be chromatographically separated.
(16) Ohtani, I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. J. Am. Chem.
Soc. 1991, 113, 4092.
a Reagents and conditions: (a) TMSCHN2, hexane-benzene-
MeOH (74%). (b) MMTrCl, pyridine (100%). (c) TBAF, THF
(99%). (d) BrCH2COBr, pyridine, CH2Cl2. (e) AcOH, THF-H2O
(82% in 2 steps). (f) Dess-Martin periodinane, CH2Cl2 (93%). (g)
(i) Et2Zn, RhCl(PPh3)3, THF-hexane; (ii) Ac2O (9%). (h) DDQ,
CH2Cl2-phosphate buffer (pH 5.9) (88%). (i) Dess-Martin perio-
dinane, CH2Cl2. (j) 2, CrCl2, NiCl2, DMSO (57%, 15S:15R ) 11:
1, 2 steps).
(17) (a) Tabuchi, T.; Kawamura, K.; Inagawa, J.; Yamaguchi, M.
Tetrahedron Lett. 1986, 27, 3889. (b) Molander, G. A.; Etter, J. B.; Harring,
L. S.; Thorel, P.-J. J. Am. Chem. Soc. 1991, 113, 8036.
(18) The calculations were executed by MacroModel (Version 6.0) with
the MM2* force field.
(19) Rathke, M. W.; Lindert, A. J. Org. Chem. 1970, 35, 3966.
(20) Maruoka, K.; Hashimoto, S.; Kitagawa, Y.; Yamamoto, H.; Nozaki,
H. Bull. Chem. Soc. Jpn. 1980, 53, 3301.
hydroxy group of the allylic alcohol 22 was protected as an
MMTr ether to quantitatively afford compound 31. The silyl
group in 31 was removed, and the resulting alcohol 32 (99%)
was converted into the bromo ester 33. The MMTr group
was removed to give allylic alcohol 34 (82%, two steps),
Org. Lett., Vol. 5, No. 6, 2003
959

the decomposition of the starting material; however, we
found by TLC monitoring, the generation of an intermediate,
the â-hydroxy lactone 37, which decomposed upon workup
(entry 5). The addition of Ac2O to trap the reactive products
allowed us to isolate the desired cyclized product (3S,4Z)-
35 in 9% yield along with the (3R,4Z)- and (4E)-isomers
(entry 6).22
(R ) Me), we postulated the folded conformation of the side
chain in the Mosher esters of 1, which led us to the wrong
conclusion that the stereochemistry at C14-C15 was threo
(Figure 1). The generally accepted zigzag conformation of
The MPM group in (3S,4Z)-35 was removed to give the
alcohol 3 in 88% yield, which was oxidized with the Dess-
Martin periodinane to afford an unstable aldehyde. The
Nozaki-Hiyama-Kishi coupling reaction23 of the aldehyde
and iodide 2, prepared from 30,5 afforded the coupling
product 36 (57%, S:R ) 11:1), and the diastereomers were
separated with HPLC. The major isomer, (15S)-36, was
found to be identical with the naturally occurring sample
upon comparison of their spectral data and chromatographic
behavior except for the sign of the CD spectrum (see
Supporting Information). Because the stereochemistry of the
C-3, C-11, C-13, and C-14 in synthetic 36 was undoubtedly
constructed by the organic synthetic method, we unambigu-
ously determined the absolute stereochemistry of C-3, C-11,
C-13, and C-14 in natural 1 to be R, R, R, and R. On the
other hand, since the absolute stereochemistry of C-15 in
natural 1 was determined to be R by the modified Mosher’s
method, the total absolute stereochemistry of haterumalide
NA was revealed, which revised the previously reported
structure.1 To confirm these results, 36 was converted into
the (R)-MTPA ester, which was found to be the enantiomer
of the (S)-MTPA ester of the natural haterumalide NA methyl
Figure 1. Relative stereochemistry of the C14-C15 part of
haterumalide NA (1, R ) Me).
the side chain in the Mosher esters of 1 is consistent with
the revised stereostructure, in which the stereochemistry at
C14-C15 is erythro, from the viewpoint of the coupling
constants and the NOESY correlations of 1.
In summary, the enantioselective synthesis of 15-epi-
haterumalide NA methyl ester (36) has been achieved from
the threitol derivative 9 in 26 steps. This synthesis re-
vises the absolute stereochemistry of haterumalide NA.
Further biological studies and the synthesis of natural (-)-
haterumalide NA are now in progress.
1
Acknowledgment. This work was supported in part by
a Grant-in-Aid for Priority Area (Targeted Pursuit of
Challenging Bioactive Molecules) from the Ministry of
Education, Culture, Sports, Science, and Technology of
Japan, Suntory Institute for Bioorganic Research, Yamada
Science Foundation, the Fujisawa Foundation, and Wako
Pure Chemical Industries, Ltd.
ester on comparison of their H NMR spectra.
In a previous paper,1 in light of the abnormal ∆δ values
in the experiments of the modified Mosher’s method for 1
(21) Kanai, K.; Wakabayashi, H.; Honda, T. Org. Lett. 2000, 2, 2549.
(22) The stereochemistry at the C-4 double bond was easily determined
by the NOE experiments. On the other hand, the stereochemistry at C-3
was determined by the modified Mosher’s method. The minor isomer,
(3S,4Z)-35, could not be transformed into the corresponding MTPA esters
because of the instability during the methanolysis of the acetyl group.
However, the modified Mosher’s method could be applied to (3R,4Z)-37
that was obtained by methanolysis of the major isomer, (3R,4Z)-35,
establishing that the major isomer possessed the undesired stereochemistry
3R, i.e., the minor isomer was the desired (3S)-compound.
Supporting Information Available: Spectral data and
experimental procedures of key compounds 6, 7, 8, 19, 35,
and 36. This material is available free of charge via the
Internet at http://pubs.acs.org.
(23) Takai, K.; Tagashira, M.; Kuroda, T.; Oshima, K.; Utimoto, K.;
Nozaki, H. J. Am. Chem. Soc. 1986, 108, 6048.
OL0341804
960
Org. Lett., Vol. 5, No. 6, 2003