Total synthesis of (+)-belactosin A.

Article abstract of DOI:10.1039/b316142k

A concise first total synthesis of the antitumour antibiotic belactosin A is reported, involving coupling of beta-lactone carboxylic acid 3 with N-Ala-aminocyclopropyl alanine 11.

Full text of DOI:10.1039/b316142k

Total synthesis of (+)-belactosin A†
Alan Armstrong* and James N. Scutt
Department of Chemistry, Imperial College London, South Kensington, London, UK SW7 2AZ.
E-mail: A.Armstrong@imperial.ac.uk; Fax: +44 (0) 20 75945804
Received (in Cambridge, UK) 11th December 2003, Accepted 14th January 2004
First published as an Advance Article on the web 27th January 2004
A concise first total synthesis of the antitumour antibiotic
belactosin A is reported, involving coupling of b-lactone
carboxylic acid 3 with N-Ala-aminocyclopropyl alanine 11.
employed. All remaining acid-labile protecting groups were then
removed with TFA/CH2Cl2 to furnish the amine-TFA salt 11 in
high yield.
Encouraged by literature precedent, we hoped that synthesis of
b-lactone 3 would be possible via stereoselective chlorination of the
monosubstituted succinate 187,8 followed by cyclisation. Synthesis
Small molecules which modulate the cell cycle offer promise for
the control of proliferative diseases.1 The recently isolated natural
product belactosin A 1 arrests cell-cycle progression at the G2/M
phase,2 and recent reports3 indicate that this compound and
analogues effect 20S proteasome inhibition, an important target for
the control of cancer and auto-immune diseases.4 We have
embarked upon a programme aimed at the total synthesis of 1
allowing access to appropriate derivatives for probing its precise
of 18 commenced with hydrodeamination of -isoleucine with a
L
slight modification to the literature procedure (Scheme 3),9
followed by conversion to the oxazolidinone derivative 16 and a
highly diastereoselective alkylation using tert-butyl bromoacetate
to afford 17 in 82% yield and 93 : 7 d.r.7 Recrystallisation followed
by hydrolysis using LiOH/H2O2 then furnished the desired acid 18
in 92% yield and > 95 : 5 d.r. The key chlorination step was
achieved by prior conversion to the dianion using LiHMDS
followed by treatment with 1 equivalent of carbon tetrachloride.8
Lactonisation then proceeded as described by Barlaam through
exposure to a biphasic ether/NaHCO3 mixture,8 required in order to
remove chloride from the ether phase and prevent SN2 ring opening
at C2.10 This two-step sequence was further optimised to a one-pot
protocol providing b-lactone 19 in 55% yield and > 95 : 5 d.r. from
18. The exceptional diastereoselectivity presumably arises via a
cyclic transition state presenting the less-hindered face of the
role in various cellular processes. We envisaged that coupling of -
L
alanine and the b-lactone carboxylic acid 3 to the intriguing and
unique central (2S,1AR,2AS)-3-(trans-2-aminocyclopropyl)alanine
core 25 would constitute an effective and flexible synthetic strategy
(Scheme 1). Recently, we reported a novel stereocontrolled route to
ent-25b (vide infra). In this paper, we report application of this
chemistry to protected 2 itself, stereocontrolled synthesis of the b-
lactone 3, and coupling of these fragments to complete the first total
synthesis of 1.
We first prepared 3-(trans-2-aminocyclopropyl)alanine 8 in 7
steps from (R)-glycidol benzyl ether 4 (Scheme 2). As in our
reported synthesis of ent-25b the key steps included “Wadsworth–
Emmons cyclopropanation” to convert the glycidol derivative 4
into the required enantiomerically pure cyclopropane 5, followed
by Curtius rearrangement to generate the cyclopropylamine 6.
Conversion to 7 and alkylation with a glycine enolate derivative in
the presence of the cinchonidinium phase-transfer catalyst 12
(O(9)-allyl-N-(9-anthracenylmethyl)cinchonidinium bromide),6
then allowed installation of the C2-stereocentre, affording 8 in 67%
yield and 93 : 7 diastereomeric ratio. Recrystallisation to diaster-
eomeric purity was followed by selective deprotection of the imine
using 15% citric acid, setting the scene for coupling to N-CBz-
alanine. Under standard DCC/HOBt/CH2Cl2 conditions, 10 was
obtained in 88% yield, but this was accompanied by partial
epimerisation at the alanine stereocentre (89 : 11 d.r.). We
attributed this to the limited solubility of HOBt in CH2Cl2, and
indeed after switching to DMF as the reaction solvent epimerisation
was completely suppressed, to yield 10 in 47% yield ( > 95 : 5 d.r.),
or 100% yield ( > 95 : 5 d.r.) when 2 eq of N-CBz-Ala were
1
enolate to the chlorinating agent. H NMR and crystallographic
analysis of tert-butyl ester 19 confirmed the relative ster-
eochemistry to be in agreement with that found for the natural
product (Fig. 1). Lactone 19 was then converted to its acid 3 with
anhydrous TFA/CH2Cl2 in high yield, maintaining the integrity of
the b-lactone ring.
Scheme 2 (a) Triethyl phosphonoacetate, NaH, toluene, 110 °C, 14 h, 63%;
t
(b) NaOH (aq), EtOH, 96%; (c) DPPA, BuOH, NEt3, reflux, 53%; (d)
Scheme 1 Retrosynthetic analysis of belactosin A.
Boc2O, MeCN, DMAP, 95%; (e) Pd/C, H2, cat. AcOH, THF, 98%; (f)
Bu4NI, DDQ, PPh3, CHCl3, rt; (g) 2 eq N-(diphenylmethylene)glycine tert-
butyl ester, 20 mol% 12, 10 eq CsOH·H2O (s), toluene/CH2Cl2 (1 : 1), 240
°C, 40 h; 67%; (h) 15% citric acid/THF, 84%; (i) 2 eq N-CBz-Ala, DCC/
HOBt/DMF, 100%; (j) TFA/CH2Cl2, 15 °C, 20 h, 90%.
† Electronic Supplementary Information (ESI) available: Data/procedures
for 1, 3, 8–11, 19, 20; spectra for 1, 19; X-ray data for 19 (CCDC 226865;
cif format). See http://www.rsc.org/suppdata/cc/b3/b316142k/
510
C h e m . C o m m u n . , 2 0 0 4 , 5 1 0 – 5 1 1
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4

View Article Online
DMF mixture. Under these conditions any unreacted EDCI/HOBt
remains in the aqueous phase, thus allowing the use of excess
activating agent without incompatibility with the coupling partner
11. Amide 20 was obtained in 50% yield as a single diastereoisomer
after purification.
After some experimentation, final deprotection of 20 was
achieved by hydrogenation in the presence of Pd/C in THF under
TFA activation, yielding the TFA salt of belactosin A. Although
this material could be taken to its isoelectric pH (as determined by
1H NMR spectroscopy using 5% NaHCO3/D2O), subsequent
purification to remove sodium trifluoroacetate proved troublesome.
Further consideration suggested that the free amino acid could be
generated directly if we used a volatile acid catalyst with a higher
pKa than that of the carboxylate group in 20. Pleasingly, when the
reaction was carried out using H2 and Pd/C in a 3 : 2 THF/HCO2H
solvent mixture the desired amino acid 1 was produced in 96%
yield. The synthetic sample (m.p. 186–187 °C, [a]2D1+4.8 (c 0.84,
H2O) (lit.2a m.p 184–185 °C, [a]D27+4.8 (c 0.37, H2O)) displayed
satisfactory HRMS data, and its TLC Rf value (0.5, butanol : acetic
1
acid : water (71 : 14 : 15 v/v/v)), H and 13C NMR spectra were
identical to those reported for the natural product.2a
In conclusion, we have completed the first total synthesis of
belactosin A 1. The synthetic strategy, particularly the knowledge
gained from the synthesis of the central 3-(trans-2-aminocyclopro-
pyl)alanine core 8 and the conditions for coupling 11 to the
sensitive b-lactone unit 3, will now facilitate the preparation of a
wide range of synthetic probes of some significant biological
pathways. Studies along these lines are currently underway.
We thank the EPSRC (studentship to JNS) for support of this
work. We are grateful to Pfizer, Merck Sharp and Dohme, and
Bristol-Myers Squibb for unrestricted support of our research
programmes. We thank Dr. A. Asai of Kyowa Hakko Kogyo Co.
Ltd. for supplying the 1H NMR spectrum of 1 and Dr. A. J. P. White
for X-ray structure determination.
Scheme 3 (a) H2NOSO3H/KOH (aq), 0 °C to rt, 74%; (b) (COCl)2, CH2Cl2,
0 °C to rt, 80%; (c) (4R)-benzyl-2-oxazolidinone, nBuLi 278 °C, 79%; (d)
tert-butyl bromoacetate, NaHMDS, 278 °C, 82%; (e) LiOH (aq), H2O2, 0
°C to rt, 92%; (f) 2 eq LiHMDS, CCl4, 278 °C to rt; then ether/NaHCO3,
55%; (g) TFA/CH2Cl2, 0 °C, 20 h, 90%.
Notes and references
Fig. 1 X-Ray structure of b-lactone 19.
1 C. M. Crews and R. Mohan, Curr. Opin. Chem. Biol., 2000, 4, 47.
2 (a) T. Mizukami, A. Asai, Y. Yamashita, R. Katahira, A. Hasegawa, K.
Ochiai and S. Akinaga, Eur. Patent 768317, 1997 (Chem. Abstr., 1997,
126, 338840) (b) A. Asai, A. Hasegawa, K. Ochiai, Y. Yamashita and T.
Mizukami, J. Antibiot., 2000, 53, 81; (c) K. Yasuki, I. Atsushi, T.
Yoshiichi, S. Shiro and A. Shiro, Jpn. Patent 2002047202, 2002 (Chem.
Abstr., 2002, 136, 172760).
3 A. Asai, T. Tsujita, S. V. Sharma, Y. Yamashita, S. Akinaga, M.
Funakoshi, H. Kobayashi and T. Mizukami, Biochem. Pharmacol.,
2004, 67, 227.
4 C. Garcia-Echeverria, Mini-Rev. Med. Chem., 2002, 2, 247.
5 For synthetic studies, see: (a) M. Brandl, S. I. Kozhushkov, K. Loscha,
O. V. Kokoreva, D. S. Yufit, J. A. K. Howard and A. de Meijere, Synlett,
2000, 1741; (b) A. Armstrong and J. N. Scutt, Org. Lett., 2003, 5, 2331;
(c) D. Diez, P. Garcia, I. S. Marcos, N. M. Garrido, P. Basabe, H. B.
Broughton and J. G. Urones, Org. Lett., 2003, 5, 3687; (d) R. P. Jain and
J. C. Vederas, Org. Lett., 2003, 5, 4669; (e) O. V. Larionov, S. I.
Kozhushkov, M. Brandl and A. de Meijere, Mendeleev Commun., 2003,
199.
With both fragments 11 and 3 in hand we were in a position to
attempt the crucial coupling, with the ambitious aim of avoiding
additional protection steps by selective activation of the b-lactone
carboxylic acid 3 for direct coupling to the amino acid 11.
However, preliminary experiments in which 3 was mixed with 1 eq
DCC/2 eq HOBt/DMF prior to addition to a solution of 11 in
EtNiPr2/DMF gave only 25% of the coupling product. We
suspected that conversion of 3 to its active ester was incomplete,
but in this case use of excess DCC was not possible since any
unreacted activating agent was likely to consume amino acid 11.
This problem was solved by exploiting biphasic conditions in
which 3 was converted to the active ester through brief exposure to
2 eq EDCI/4 eq HOBt in CH2Cl2/H2O at 0 °C (Scheme 4),11
followed by transfer of the organic phase to a cooled 11/EtNiPr2/
6 (a) E. J. Corey, F. Xu and M. C. Noe, J. Am. Chem. Soc., 1997, 119,
12414; (b) B. Lygo and P. G. Wainwright, Tetrahedron Lett., 1997, 38,
8595.
7 R. P. Beckett, M. J. Crimmin, M. H. Davis and Z. Spavold, Synlett,
1993, 137.
8 B. Barlaam, G. T. Bird, C. Lambert-van der Brempt, D. Campbell, S. J.
Foster and R. Maciewicz, J. Med. Chem., 1999, 42, 4890.
9 G. A. Doldouras and J. Kollonitsch, J. Am. Chem. Soc., 1978, 1, 341.
10 J. S. Bajwa and M. J. Miller, J. Org. Chem., 1983, 48, 1114.
11 This two-phase system has previously been used for direct amide
formation without isolation of the active ester. G-J. Ho, K. M. Emerson,
D. J. Mathre, R. F. Shuman and E. J. J. Grabowski, J. Org. Chem., 1995,
60, 3569.
Scheme 4 (a) EDCI/HOBt in CH2Cl2/H2O, 0 °C; then 11/EtNiPr2/DMF, 0
°C, 50%; (b) Pd/C, H2, THF/HCO2H (3 : 2), 96%.
C h e m . C o m m u n . , 2 0 0 4 , 5 1 0 – 5 1 1
511