Total synthesis of (+)-bongkrekic acid

Article abstract of DOI:10.1016/j.tetlet.2004.09.162

Total synthesis of bongkrekic acid, an important apoptosis inhibitor, has been accomplished. The strategy includes inexpensive starting materials, asymmetric alkylation, anionic allyl coupling and oxidative manipulations. This process would provide a sufficient amount of bongkrekic acid and its analogues.

Full text of DOI:10.1016/j.tetlet.2004.09.162

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 8863–8866
Total synthesis of (+)-bongkrekic acid
Mitsuru Shindo,* Tomoyuki Sugioka, Yuko Umaba and Kozo Shishido*
Institute for Medicinal Resources, University of Tokushima, Sho-machi 1, Tokushima 770-8505, Japan
Received 31 August 2004; revised 22 September 2004; accepted 29 September 2004
Available online 14 October 2004
Abstract—Total synthesis of bongkrekic acid, an important apoptosis inhibitor, has been accomplished. The strategy includes inex-
pensive starting materials, asymmetric alkylation, anionic allyl coupling and oxidative manipulations. This process would provide a
sufficient amount of bongkrekic acid and its analogues.
ꢀ 2004 Elsevier Ltd. All rights reserved.
Bongkrekic acid (1)1 is a natural toxic antibiotic pro-
duced by the microorganism Pseudomonas cocovenenans.
The high toxicity of 1 has been attributed to its affinity
for the ATP/ADP translocator protein residing in the
mitochondrial inner membrane, which prevents oxida-
tive phosphorylation (IC50 = 2 · 10ꢀ8 M).2 Further-
more, by binding to the protein to prevent opening of
mitochondrial permeability transition pores, bongkrekic
acid 1 has been found to be an inhibitor of apoptosis,3
making it a significant biochemical tool for investiga-
tion. However, due to the difficulty in obtaining 1 by
fermentation, an efficient chemical synthesis of 1 is still
required. Since Coreyꢀs first total synthesis in 1984,4
including optical resolution steps, there have been no re-
ports on the synthetic investigation of 1.5 We report
herein the stereocontrolled, asymmetric synthesis of
(+)-bongkrekic acid.
Bongkrekic acid (1), a polyene–tricarboxylic fatty acid,
has three pairs of conjugated dienes and two allylic stereo-
genic centres. The efficiency of the synthesis depends
upon the stereocontrolled construction of this character-
istic polyene skeleton, especially the C2–C3 and C18–
C19 trisubstituted (Z)-alkenes and oxidation to give
the terminal carboxylic acids, since the polyene unit
might be unstable under harsh conditions. Our synthetic
strategy is outlined in Scheme 1. Since a convergent syn-
thesis would be useful for the preparation of bongkrekic
Scheme 1. Retrosynthesis of bongkrekic acid 1.
C10) and B (C11–C22), which would be connected by
acid analogues for biochemical investigation, 1 was di-
allylic coupling. Segment A would be synthesized via
the Suzuki coupling6 of the (E)-vinyl borane D with
vided at the C10 and C11 bond into segments A (C1–
the (Z)-iodoalkene C. For the preparation of segment
B, the construction of the trisubstituted (Z)-alkene
(C18–C19) next to the chiral carbon (C17) would consti-
tute the key step, which would be accomplished using
the chiral butenolide (Scheme 2).
Keywords: Apoptosis; Carboxylic acid; Oxidation; Natural product;
Total synthesis.
*
Corresponding authors. Tel./fax: +81 88 633 7294; e-mail: shindo@
ph.tokushima-u.ac.jp
0040-4039/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.09.162

8864
M. Shindo et al. / Tetrahedron Letters 45 (2004) 8863–8866
Scheme 2. Reagents and conditions. (a) LDA, THF, 3, 85%, >95% de; (b) LiAlH4, THF, 88%; (c) (COCl)2, DMSO, Et3N, CH2Cl2; (d) 5, CrCl2
(0.35equiv), LiI, Mn, TMSCl, THF; 74% for two steps; (e) TBAF, THF, 92%; (f) 7, PdCl2 (PPh3)2, Et3N, MeOH, 92%; (g) MsCl, iPr2EtN, 4-DMAP,
CH2Cl2; (h) PhSO2Na, DMF, 81% for two steps.
Synthesis of the (E)-vinyl borane D commenced with
asymmetric alkylation of Evansꢀ oxazolidinone 27 with
(E)-1-tert-butyldiphenylsiloxy-4-iodo-2-butene (3)8 to
provide 4 with excellent diastereoselectivity in good
yield. After reductive removal of the chiral auxiliary,
the resulting alcohol was oxidized to furnish the alde-
hyde, which was subjected to CrCl2 catalyzed borylalk-
enylation9 with dichloromethylboronic ester 510 to
produce the (E)-alkenylboronic ester 6 after deprotec-
tion. Suzuki coupling of the boronic ester 6 with the
(Z)-iodoalkene 7, prepared from 3-butyn-1-ol,11 pro-
vided the desired (2Z,4E,8E)-triene alcohol 8 in 92%
yield without any detected stereoisomers. The alcohol
8 was converted to the unstable 9 as segment A, which
was immediately subjected to the next coupling reaction.
The synthesis of segment B was completed as shown in
Scheme 3. The (S)-glycidyl ether 10, prepared by silyla-
tion of the readily available (S)-glycidol,12 was reacted
with the lithium acetylide of 3-(p-methoxyphenylmeth-
oxy)propyne in the presence of BF3ÆOEt2,13 followed
by protection of the resulting hydroxyl group, to afford
12. The alkyne was hydrogenated using the Lindlar cat-
alyst to give the cis-alkenyl alcohol 13 after selective
deprotection of TBS ether. Swern oxidation, followed
by methylation of the resulting aldehyde, furnished the
Scheme 3. Reagents and conditions. (a) 11, BuLi, BF3ÆOEt2, THF, 75%; (b) TIPSOTf, 2,6-lutidine, CH2Cl2, 88%; (c) Pd–CaCO3–Pb(OAc)2, H2,
quinoline, hexane, 93%; (d) 3M HCl aq, THF, 91%; (e) (COCl)2, DMSO, Et3N, CH2Cl2; (f) MeMgBr, THF, 87% for two steps; (g) (COCl)2, DMSO,
Et3N, CH2Cl2, 96%; (h) AcOEt, LDA, THF; (i) TBAF, THF, 80% for two steps; (j) MsCl, Et3N, CH2Cl2, 99%; (k) DDQ, CH2Cl2, H2O, 92%; (l)
t
MnO2, CH2Cl2; Ph3P@CHCO2 Bu, 87%; (m) BF3ÆOEt2, CH2Cl2; (n) ClCO2Et, Et3N, THF; NaBH4, 97% for two steps; (o) CBr4, PPh3, CH2Cl2,
99%.

M. Shindo et al. / Tetrahedron Letters 45 (2004) 8863–8866
8865
Scheme 4. Reagents and conditions. (a) BuLi, HMPA, THF; 17, 86%; (b) DIBAH, THF; (c) NaBH4, THF, H2O, 70% for two steps; (d) 5% Na–Hg,
t
Na2 HPO4, MeOH, 93%; (e) BuCOCl, Et3N, CH2Cl2, 96%; (f) MeOTf, 2,6-di-tert-butylpyridine, benzene, 83%; (g) DIBAH, THF, 81%; (h) Dess–
t
Martin reagent, benzene–CH2Cl2 (2:1); (i) Ph3P@C(CH3)CO2 Bu, benzene, 80% for two steps; (j) 3M HCl aq, THF, 82%; (k) MnO2, benzene–
CH2Cl2 (2:1); (l) NaClO2, NaH2PO4, 2-methyl-2-butene, THF, tBuOH, H2O; (m) ClCH2OTBDPS, Et3N, THF, 40% for three steps; (n) Dess–Martin
t
reagent, benzene–CH2Cl2 (2:1); (o) NaClO2, NaH2PO4, 2-methyl-2-butene, THF, BuOH, H2O; (p) 3M HCl, MeOH, 50% for three steps.
(R)-a-siloxy methyl ketone 14. The lithium enolate of
ethyl acetate efficiently added to the ketone 14 to give
an adduct, which was treated with TBAF followed by
dehydration, to provide the butenolide 15 in excellent
yield. After deprotection of the MPM group, the result-
ing allylic alcohol was oxidized to the aldehyde, which
was subjected to the Wittig reaction to give the unsatu-
rated t-butyl ester 16 bearing the (12E,14Z,18Z)-triene
unit of segment B with no detectable stereoisomers.
The tert-butyl ester was converted to 17 as segment B.
of C-1 carboxylic acid, the oxidation of C-24 was found
to be extremely poor. To improve the yields in the final
sequence, we set the triester of 1 as the initial target.
Although the yield of oxidative functionalization was
improved, the hydrolysis at C-1 and C-24 methyl esters
was poor. After investigation on suitable esters of 1, we
found that the 2-tert-butyldiphenylsiloxymethyl ester14
was easily removed under mild conditions without isom-
erization and survived oxidative conditions. The diol 22
was oxidized by MnO2, followed by NaClO2, to give the
C-1 carboxylic acid, which was esterified by 2-tert-but-
yldiphenylsiloxymethyl chloride with triethylamine to
give 23. The C-24 oxidation was performed using the
Dess–Martin reagent, followed by NaClO2, to afford
24. Final hydrolysis of 24 was accomplished by HCl in
methanol at room temperature to furnish 1 in 50% yield
for three steps from 23. Synthetic bongkrekic acid was
identical with an authentic sample by HPLC. The corre-
sponding trimethyl ester of 1,15 prepared by esterifica-
tion using diazomethane, was spectroscopically
indistinguishable from the trimethyl ester derived from
natural sample (Scheme 4).
With both segments A and B in hand, the task of cou-
pling them was addressed. Attempted coupling of the
lithiated 9 with 18 as segment B led to the desired prod-
uct in only 42% yield. After attempts to improve the effi-
ciency of the coupling, compound 17 bearing the
butenolide function was found to be a much better cou-
pling partner and 19 was isolated in 86% yield. The lac-
tone 19 was reduced to afford the diol 20.
Desulfonylation was carried out with Na–Hg, then pro-
tection of the primary alcohol, methylation of the sec-
ondary alcohol and deprotection were successively
employed to afford 21 in good overall yield. Oxidation
of 21, followed by Wittig olefination, afforded 22 with
excellent E-selectivity, after removal of the TBS groups.
In conclusion, a convergent asymmetric synthesis of
bongkrekic acid has been achieved. This process pro-
vides a sufficient amount of bongkrekic acid and also
constitutes a route to a wide variety of analogues for
apoptosis research.
Acknowledgements
Our final concern in the synthesis was oxidation of the
terminal hydroxyl groups to the tricarboxylic acid in
the presence of the conjugated dienes. In the presence
This work was supported in part by a Grant-in-Aid for
Scientific Research from the Ministry of Education,
Culture, Sports, Science and Technology, Japan and

8866
M. Shindo et al. / Tetrahedron Letters 45 (2004) 8863–8866
9. Takai, K.; Hikasa, T.; Ichiguchi, T.; Sumino, N. Synlett
1999, 1769–1771.
10. Wuts, P. G. M.; Thompson, P. A. J. Organomet. Chem.
1982, 234, 137–141.
11. Esumi, T.; Iwabuchi, Y.; Irie, H.; Hatakeyama, S.
Tetrahedron Lett. 1998, 39, 877–880.
Suntory Institute of Bioorganic Research. We thank
Professor Y. Shinohara and Professor H. Terada for
providing natural bongkrekic acid and Daiso Co. Ltd
for supplying (S)-3-chloro-1,2-propandiol.
12. (S)-Glycidol was prepared by Sharpless asymmetric epox-
idation of propenol, see: Katsuki, T.; Martin, V. S. Org.
React. 1996, 48, 1. The commercially available (S)-3-
chloro-1,2-propanediol can be easily converted to (S)-
glycidol.
References and notes
1. (a) Lijmbach, G. W. M.; Cox, H. C.; Berends, W.
Tetrahedron 1970, 26, 5993–5999; (b) Zyblir, J.; Gaude-
mer, F.; Gaudemer, A. Experientia 1973, 29, 648–649.
2. Stubbs, M. Pharmacol. Ther. 1979, 7, 329–350.
3. (a) Marchetti, P.; Castedo, M.; Susin, S. A.; Zamzami, N.;
Hirsch, T.; Macho, A.; Haeffner, A.; Hirsch, F.; Geuskens,
M.; Kroemer, G. J. Exp. Med. 1996, 184, 1155–1160; (b)
Marchetti, P.; Hirsch, T.; Zamzami, N.; Castedo, M.;
Decaudin, D.; Susin, S. A.; Masse, B.; Kroemer, G. J.
Immunol. 1996, 157, 4830–4836.
4. Corey, E. J.; Tramontano, A. J. Am. Chem. Soc. 1984,
106, 462–463.
5. Analogue synthesis, see: Pei, Y.; Carroll, A. K.; Anderson,
C. M.; Moos, W. H.; Ghosh, S. S. Synthesis 2003, 1717–
1721.
13. Yamaguchi, M.; Hirao, I. Tetrahedron Lett. 1983, 24, 391–
394.
14. Sawada, D.; Ito, Y. Tetrahedron Lett. 2001, 42, 2501–
2504.
26
15. Trimethyl ester of 1: ½aꢁD +80 (c 0.15, CHCl3) (the
26
naturally derived sample: ½aꢁD +84 (c 0.24, CHCl3);
;
IR(CDCl3): 1740, 1714, 1635, 1603cmꢀ1 1H NMR
(400MHz, CDCl3) d: 1.02(3H, d, J = 7.2Hz), 1.83 (3H,
s), 1.94 (3H, s), 1.95–2.16 (5H, m), 2.33–2.39 (3H, m), 2.56
(1H, ddd, J = 7.2, 7.2, 14.0Hz), 3.21 (3H, s), 3.32 (2H, s),
3.67 (3H, s), 3.71 (3H, s), 3.75 (3H, s), 4.34 (1H, t,
J = 7.2Hz), 5.22 (1H, m), 5.31–5.44 (2H, m), 5.67 (1H, dt,
J = 6.0, 15.0Hz), 5.69 (1H, s), 5.99 (1H, m), 6.04 (1H, m),
6.26 (1H, dd, J = 11.2, 15.6Hz), 6.36 (1H, d, J = 11.2Hz),
7.49–7.53 (2H, m); 13C NMR (100MHz, CDCl3): d: 170.5,
168.9, 166.1, 147.0, 145.3, 144.7, 134.8, 132.0, 131.3, 130.4,
128.0, 126.2, 125.4, 124.8, 124.5, 124.1, 118.1, 78.2, 56.3,
52.1, 51.7, 51.1, 40.3, 39.6, 37.5, 32.8, 32.3, 32.0, 19.2, 18.6,
12.3. MS (EI) m/z 528 (M+).
6. Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457–
2483.
7. Evans, D. A.; Ennis, M. D.; Mathre, D. J. J. Am. Chem.
Soc. 1982, 104, 1737–1739.
8. Roush, W. R.; Straub, J. A.; Nieuwenhze, M. S. J. Org.
Chem. 1991, 56, 1636–1648.