Second-generation total synthesis of haterumalide NA using B-alkyl Suzuki-Miyaura coupling

Article abstract of DOI:10.1021/ol800554f

Second-generation total synthesis of haterumalide NA, a potent cytotoxic marine macrolide, was achieved by using B-alkyl Suzuki-Miyaura coupling and Nozaki-Hiyama-Kishi coupling as key steps (1.2% in 33 steps). Compared to our first-generation approach, the second-generation synthesis is much improved in the yield of key intermediate.

Full text of DOI:10.1021/ol800554f

ORGANIC
LETTERS
2008
Vol. 10, No. 9
1859-1862
Second-Generation Total Synthesis of
Haterumalide NA Using B-Alkyl
Suzuki–Miyaura Coupling
Ichiro Hayakawa, Mitsuru Ueda, Masashi Yamaura, Yoichi Ikeda, Yuta Suzuki,
Kensaku Yoshizato, and Hideo Kigoshi*
Department of Chemistry, Graduate School of Pure and Applied Sciences, and
Center for Tsukuba AdVanced Research Alliance, UniVersity of Tsukuba,
Ibaraki 305-8571, Japan
kigoshi@chem.tsukuba.ac.jp
Received March 10, 2008 (Revised Manuscript Received March 21, 2008)
ABSTRACT
Second-generation total synthesis of haterumalide NA, a potent cytotoxic marine macrolide, was achieved by using B-alkyl Suzuki–Miyaura
coupling and Nozaki–Hiyama–Kishi coupling as key steps (1.2% in 33 steps). Compared to our first-generation approach, the second-generation
synthesis is much improved in the yield of key intermediate.
Haterumalide NA (1) is a macrolide isolated from the
Okinawan sponge Iricinia sp. by Uemura and co-workers
(Figure 1).1 Haterumalide NA (1) exhibits cytotoxicity
ascidian Didemnidae sp.2a,b We compared the cytotoxicity
of haterumalide NA (1), haterumalide NA methyl ester (2),
biselides A (3), B (4), and C (5), and found that haterumalide
NA (1), haterumalide NA methyl ester (2), biselides A (3)
and B (4) showed stronger cytotoxicity than did anticancer
drug cisplatin against human breast cancer MDA-MB-231
and human non-small cell lung cancer NCI-H460.2b Interest-
ingly, haterumalide NA (1) showed strong toxicity against
brine shrimp, with an LD50 of 0.6 µg/mL, while haterumalide
NA methyl ester (2) and biselide A (3) are less toxic. These
results encouraged us to search for novel anticancer drugs
based on these unique lead compounds.
The unique structures of haterumalides and biselides, along
with their potent biological activity, have made them
attractive targets for synthesis.3 In 2003, we reported the first
synthesis of ent-haterumalide NA methyl ester (2).4 This
synthesis revised the stereochemistry of haterumalide NA
Figure 1. Structures of haterumalides and biselides.
against P388 cells and moderate acute toxicity against mice.
On the other hand, we isolated biselides A–E, which are
oxygenated analogues of haterumalides, from the Okinawan
(2) (a) Teruya, T.; Shimogawa, H.; Suenaga, K.; Kigoshi, H. Chem.
Lett. 2004, 33, 1184. (b) Teruya, T.; Suenaga, K.; Maruyama, S.; Kurotaki,
M.; Kigoshi, H. Tetrahedron 2005, 61, 6561.
(3) Kigoshi, H.; Hayakawa, I. Chem. Rec. 2007, 7, 254.
(4) Kigoshi, H.; Kita, M.; Ogawa, S.; Itoh, M.; Uemura, D. Org. Lett.
2003, 5, 957.
(1) Takada, N.; Sato, H.; Suenaga, K.; Arimoto, H.; Yamada, K.; Ueda,
K.; Uemura, D. Tetrahedron Lett. 1999, 40, 6309.
10.1021/ol800554f CCC: $40.75
Published on Web 04/09/2008
2008 American Chemical Society

(1) and determined its absolute configulation. Snider et al.5a
and Hoye et al.5b synthesized ent-haterumalide NA methyl
ester (2) and haterumalide NA (1) itself, respectively.
Because our previous route included low-yield steps, we
planned to develop an efficient method for the second-
generation synthesis of haterumalides, biselides, and their
derivatives, which will provide a practical supply for further
biological studies.
Scheme 2
.
Synthesis of the Precursor of B-Alkyl Suzuki–
Miyaura Coupling
Our retrosynthetic analysis of haterumalide NA (1) is
shown in Scheme 1. Haterumalide NA (1) can be logically
Scheme 1. Retrosynthetic Analysis of Haterumalide NA (1)
carried out by using NaOMe in MeOH, but the yield and
stereoselectivity were not so high (56%, trans/cis ) 5.3:1).
In this work, the intramolecular oxy-Michael cyclization of
16 was attempted by using Triton B in MeOH to provide
tetrahydrofuran 17 as the sole product.10 This cyclization
improved the stereoselectivity and enhanced the reaction rate.
Reduction of methyl ester 17 by LiAlH4 gave alcohol 18.
Alcohol 18 was converted to iodide 19, a precursor of the
requisite boranate. On the other hand, alcohol 18 was
converted to the terminal olefin 20 via a seleno ether.11
With iodide 19 and terminal olefin 20 in hand, we
attempted B-alkyl Suzuki–Miyaura coupling,6 as depicted in
Table 1. Boranate 21, which was prepared from 19 by
divided into the macrolactone 6 and the appropriately
protected side-chain unit 7. The macrolactone 6 can be
established by the lactonization of the seco acid 8. Seco
acid 8 can be synthesized from a common intermediate 9
for haterumalides and biselides. We planned the practical
synthesis of the common intermediate 9 by B-alkyl
Suzuki–Miyaura coupling6 between the alkenylsilane seg-
ment 117 and alkylborane 12 and subsequent stereoselective
construction of a chloroolefin part from alkenylsilane 10.
Synthesis of the common intermediate 9 started from the
known glycal 13 (Scheme 2).8 The hydroxyl group of 13
was protected as the DMPM ether 14. The DMPM ether 14
was converted to hemiacetal 15 by the oxymercuration–re-
duction sequence.9 The Wittig reaction of hemiacetal 15
afforded the R,ꢀ-unsaturated ester 16. We next tried stereo-
selective construction of the tetrahydrofuran part by using
intramolecular oxy-Michael cyclization. In our previous
reports,4 similar intramolecular oxy-Michael cyclization was
Table 1. Study of B-Alkyl Suzuki–Miyaura Coupling
entry
precursor
reagent
yield (%)
1
2
3
19
20
20
t-BuLi, 9-BBN-OMe
9-BBN
9-BBN dimer
32
-
quant
lithiation and transmetalation,12 participated in the cross-
coupling reaction with alkenylsilane 117 to provide desired
coupling compound 10, but the yield was low (32%) (entry
(5) (a) Gu, Y.; Snider, B. B. Org. Lett. 2003, 5, 4385. (b) Hoye and
Wang first achieved the total synthesis of haterumalide NA (1) itself. Hoye,
T. R.; Wang, J. J. Am. Chem. Soc. 2005, 127, 6950.
(6) Miyaura, N.; Suzuki, A. Chem. ReV. 1995, 95, 2457.
(7) Miura, K.; Hondo, T.; Okajima, S.; Nakagawa, T.; Takahashi, T.;
Hosomi, A. J. Org. Chem. 2002, 67, 6082.
(10) (a) Ko, S.; Klein, L.; Pfaff, K.; Kishi, Y. Tetrahedron Lett. 1982,
23, 4415. (b) The stereochemistry was determined by NOESY correlations.
(11) Grieco, P. A.; Gilman, S.; Nishizawa, M. J. Org. Chem. 1976, 41,
1485.
(8) Ireland, R.; Thaisrivongs, S.; Vanier, N.; Wilcox, C. J. Org. Chem.
1979, 45, 48.
(9) Oishi, T.; Ando, K.; Inomiya, K.; Sato, H.; Iida, M.; Chida, N. Bull.
Chem. Soc. Jpn. 2002, 75, 1927.
(12) Marshall, J. A.; Johns, B. A. J. Org. Chem. 1998, 63, 7885.
1860
Org. Lett., Vol. 10, No. 9, 2008

1). We next investigated B-alkyl Suzuki–Miyaura coupling6
with 20 and 11. Although hydroboration of the terminal olefin
20 with 9-BBN was followed by the addition of PdCl2(dppf)
and alkenylsilane 11,7 the desired compound 10 could not
be obtained (entry 2). However, the use of 9-BBN dimer
instead of 9-BBN gave the best result: a quantitative yield
of 10 maybe due to the concentrated conditions (entry 3).
We next tried to stereoselectively construct a chloroolefin
part from alkenylsilane 10 (Scheme 3). In our previous
Scheme 4. Synthesis of Seco Acid 26
Scheme 3
.
Synthesis of Common Intermediate 9 of
Haterumalides and Biselides
macrolactone produced low yields (17%) under the Yamagu-
chi conditions.17 On the other hand, Snider and Gu achieved
satisfactory macrolactonization of a similar seco acid under
the same condition.5a This suggested that steric hindrance
of the acetonide group in our seco acid 26 interfered with
macrolactonization. Therefore, we next tried macrolacton-
ization using seco acid without an acetonide group.
report,4 a chloroolefin part was stereoselectively constructed
from an alkenylsilane by a modification of Tamao’s proce-
dure.13 We reported that the addition of a catalytic amount
of water was important for the reaction to be reproducible.
In this study, we attempted this reaction condition to
alkenylsilane 10 but achieved a low and irreproducible yield,
thus prompting us to reexamine the reaction conditions.
Extensive examination of this reaction led us to find
satisfactory conditions, i.e., NCS (2.0 equiv) in DMF at 50
°C in the presence of K2CO3 (0.5 equiv) as a base.14 This
modification increased the yield of the desired chloroolefin
to 58% reproducibly.15 Treatment of chloroolefin under
acidic conditions gave a triol, and the resulting 1,2-diol group
was reprotected as an acetonide group to afford the common
intermediate 9 for haterumalides and biselides. The overall
sequence proceeded in 13 steps from D-mannose and in 32%
overall yield, and thus, the common intermediate 9 could
be synthesized in multigram quantities.
The acetonide group in 25 was removed by using CSA to
give a diol (Scheme 5). Oxidative cleavage of the diol by
Scheme 5. Synthesis of Key Intermediate 30
Next, we tried to synthesize haterumalide NA (1) from
the common intermediate 9 (Scheme 4). Oxidation of 9 by
Dess–Martin periodinane afforded a labile aldehyde, which
was converted into the Z-conjugated ester 23 by using Ando’s
modified Horner–Wadsworth–Emmons reaction.16 The DIBAL
reduction of 23 gave an allylic alcohol, which was oxidized
to the conjugated aldehyde 24. The aldol reaction of 24 with
isopropyl acetate provided a ꢀ-hydroxy ester as a diastere-
omeric mixture of the hydroxyl group at C-3. The resulting
secondary hydroxyl group was protected as a TBS ether to
afford compound 25. The DMPM group in 25 was removed,
and hydrolysis of the isopropyl ester afforded seco acid 26,
a precursor of the macrolactonization.
NaIO4 followed by reductive workup with NaBH4 afforded
an alcohol. The primary hydroxyl group was protected as a
trityl group, and the DMPM group was removed. Hydrolysis
of the isopropyl ester afforded seco acid 27. The macrolac-
tonization of 27 was accomplished by the Yamaguchi
conditions17 to give the desired lactone 28 along with the
dimer (6%).18 After the TBS group in 28 was removed by
TBAF, the C-3 isomers 29a and 29b were separated by silica
Thus, the precursor for the macrolactonization was in hand.
However, attempts to macrolactonize seco acid 26 to
(13) Tamao, K.; Akita, M.; Maeda, K.; Kumada, M. J. Org. Chem. 1987,
52, 1100.
(14) Addition of KF showed little improvement, and addition of CaCO3
had no effect in this case.
(15) We could not recover alkenylsilane 10 under the previous condi-
tions.
(17) Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M. Bull.
Chem. Soc. Jpn. 1979, 52, 1989.
(16) Ando, K. J. Org. Chem. 1998, 63, 8411.
Org. Lett., Vol. 10, No. 9, 2008
1861

gel chromatography. The undesired isomer 29a was trans-
formed into the desired isomer 29b by oxidation and Luche
reduction.19 To convert 29b to 30, we followed the procedure
reported by Snider and Gu.5a Acetylation of the hydroxyl
group at C-3 and removal of the trityl group gave alcohol
30,20 which is the enantiomer of the key intermediate of our
previous total synthesis of ent-haterumalide NA methyl ester
(2).4
synthesis via 2,4-dimethoxybenzyl ester 35. Nozaki–Hiya-
ma–Kishi coupling21 with 2,4-dimethoxybenzyl 33 afforded
the coupling product 35.22 The 2,4-dimethoxybenzyl ester
in 35 was cleaved with TFA and anisole to afford hateru-
malide NA (1).23 Synthetic haterumalide NA (1) gave
spectral data (1H NMR, 13C NMR, HRMS, and CD24) in
full agreement with those of the natural one,1 completing
the total synthesis.
To convert 30 into haterumalide NA (1), we followed our
first-generation synthesis with modification by Hoye (Scheme
6).4,5b Oxidation with Dess–Martin periodinane and Noza-
In conclusion, we have achieved the second-generation
total synthesis of haterumalide NA (1). Practical synthesis
of the common intermediate 9 for haterumalides and biselides
has been achieved on the basis of B-alkyl Suzuki–Miyaura
coupling6 as a key step in multigram quantities. Compared
to our first-generation approach,4 which required 25 steps
(longest linear sequence) and proceeded in 0.22% overall
yield, the second-generation synthesis is much improved in
the yield of key intermediate. Also, we achieved total
synthesis of haterumalide NA (1) itself (1.2% in 33 steps)
by using Nozaki–Hiyama–Kishi coupling21 with a modifica-
tion of our first-generation procedure. This strategy is now
being applied to the synthesis of other haterumalides,
biselides, and their derivatives, and further structure-activity
relationship studies are in progress.
Scheme 6. Total Synthesis of Haterumalide NA (1)
Acknowledgment. We thank Nippon Kayaku Co., Ltd.,
for its financial support. This work was supported in part by
Grants-in-Aid for Young Scientists (B), Scientific Research
on Priority Area “Creation of Biologically Functional
Molecules”; and by the 21st COE program from the Ministry
of Education, Culture, Sports, Science and Technology
(MEXT), Japan. We would like to thank Professors Akira
Sekiguchi, Tatsuya Nabeshima, Masaaki Ichinohe, and
Shigehisa Akine (University of Tsukuba) for APCI mass
analysis and CD spectrum analysis.
ki–Hiyama–Kishi coupling21 with iodide 32, prepared from
31,4 afforded haterumalide NA MPM ester (34). However,
the MPM ester in 34 could not be successfully cleaved under
reported conditions (TFA, Et3SiH).5b We next tried the total
Supporting Information Available: Detailed experimen-
tal procedures and spectroscopic data. This material is
available free of charge via the Internet at http://pubs.acs.org.
OL800554F
(18) Another macrolactonization, the Shiina conditions, gave only
undesired C-3 hydroxyl isomer (12%), see: Shiina, I.; Kubota, M.; Ibuka,
R. Tetrahedron Lett. 2002, 43, 7535.
(22) Due to the small reaction scale of Nozaki–Hiyama–Kishi coupling,
we could not isolate the minor isomer at C-15.
(19) Gemal, A. L.; Luche, J-L. J. Am. Chem. Soc. 1981, 103, 5454.
(20) The alcohol 30 gave spectral data (1H NMR, 13C NMR and HRMS)
in full agreement with the authentic sample. The optical rotation of our
sample 30 corresponded to the reported values (-11.7 compared with +10.7
for ent-305a and -16.0 for 305b).
(23) Kobayashi et al. have reported removal of the 2,4-dimethoxybenzyl
group in similar esters. Kobayashi, M.; Sato, K.; Yoshimura, S.; Yamaoka,
M.; Takase, S.; Ohkubo, M.; Fujii, T.; Nakajima, H. J. Antibiot. 2005, 58,
648.
(24) Comparison of the CD spectra of synthetic and natural samples
identified absolute configuration. The CD spectral data of synthetic sample,
CD (MeOH) λext 220 nm, ∆ꢀ +0.12, was the same sign as natural sample
[CD (MeOH) λext 220 nm, ∆ꢀ +0.10].
(21) (a) Takai, K.; Kimura, K.; Kuroda, T.; Hiyama, T.; Nozaki, H.
Tetrahedron Lett. 1983, 24, 5281. (b) Jin, H.; Uenishi, J.; Christ, W. J.;
Kishi, Y. J. Am. Chem. Soc. 1986, 108, 5644.
1862
Org. Lett., Vol. 10, No. 9, 2008