Synthesis of the potent antimalarials calothrixin A and B

Article abstract of DOI:10.1021/ol006649q

(matrix presented) A concise synthesis of calothrixins A (1) and B (2) that confirms their assigned structures and affords straightforward synthetic access to them is reported.

Full text of DOI:10.1021/ol006649q

ORGANIC
LETTERS
2000
Vol. 2, No. 23
3735-3737
Synthesis of the Potent Antimalarials
Calothrixin A and B
T. Ross Kelly,* Yajun Zhao, Marta Cavero, and Mercedes Torneiro
E. F. Merkert Chemistry Center, Boston College, Chestnut Hill, Massachussets 02467
ross.kelly@bc.edu
Received September 25, 2000
ABSTRACT
A concise synthesis of calothrixins A (1) and B (2) that confirms their assigned structures and affords straightforward synthetic access to
them is reported.
In 1999 Rickards, Smith, and colleagues reported1 the
isolation and structure determination of calothrixin A (1) and
B (2). Both compounds have exceptional (nanomolar)
antimalarial activity, and their utility in that regard is
currently under evaluation.
These considerations, taken with our continuing interest2 in
the synthesis of heteroaromatic natural products and the
minor reservations expressed above about the structure of
(2) See: Kelly, T. R.; Fu, Y.; Sieglen, J. T.; De Silva H. Org. Lett. 2000,
2, 2351 and references therein.
(3) (a) Kelly, T. R.; Echavarren, A.; Chandrakumar, N. S.; Ko¨ksal, Y.
Tetrahedron Lett. 1984, 25, 2127. (b) Kelly, T. R.; Bell, S. H.; Ohashi, N.;
Armstrong-Chong, R. J. J. Am. Chem. Soc. 1988, 110, 6471. (c) Kelly, T.
R.; Jagoe, C. T.; Li, Q. J. Am. Chem. Soc. 1989, 111, 4522. (d) Kelly, T.
R.; Kim, M. H. J. Org. Chem. 1992, 57, 1595. (e) Kelly, T. R.; Walsh, J.
J. J. Org. Chem. 1992, 57, 6657. (f) Kelly, T. R.; Xu, W.; Sundaresan, J.
Tetrahedron Lett. 1993, 34, 6173. (g) Kelly, T. R.; Xu, W.; Ma, Z.; Li, Q.;
Bhushan, V. J. Am. Chem. Soc. 1993, 115, 5843. (h) Kelly, T. R.; Xie, R.
L. J. Org. Chem. 1998, 63, 8045.
(4) For reviews, see: (a) Gschwend, H. W.; Rodriguez, H. R. Org. React.
1979, 26, 1. (b) Snieckus, V. Chem. ReV. 1990, 90, 879.
The structure assigned to 1 is based on an X-ray crystal-
lographic determination. The structure assignment for 2 is
based on a careful consideration of the similarities and
differences of the spectra of 1 and 2, but the two have not
been interconverted. While two natural products that both
have a C19N2 core and co-occur are likely to have the same
skeleton, it is not necessary.
(5) Watanabe, M.; Snieckus, V. J. Am. Chem. Soc. 1980, 102, 1457.
(6) Comins, D. L.; Killpack, M. O. J. Org. Chem. 1987, 52, 104.
(7) Work, T. S. J. Chem. Soc. 1942, 429, 431.
(8) Fraser, R. R.; Mansour, T. S. J. Org. Chem. 1984, 49, 3442.
(9) See Supporting Information.
1
(10) The H NMR and UV-vis spectra of synthetic 2 are in excellent
agreement with those reported for the natural product. Our9 13C NMR data
are in good, but not perfect, agreement with those reported. We attribute
the minor (all less than 1 ppm) differences to variations in the water content
(or pH) of our DMSO-d6 13C NMR solution and that of Rickards et al. A
13C NMR spectrum of synthetic 2 subsequently recorded in Prof. Rickards’s
laboratory is in much better agreement with the original1 data. Direct TLC
comparisons (including cospotting) of synthetic 2 with natural 2 showed
the two are indistinguishable (we thank Professor Rickards for recording
the 13C NMR spectrum and carrying out TLC comparisons; NMR spiking
experiments were precluded by the paucity of the natural material available).
To our knowledge the calothrixins are the only known
naturally occurring members of this pentacyclic ring system.
(1) Rickards, R. W.; Rothschild, J. M.; Willis, A. C.; de Chazal, N. M.;
Kirk, J.; Kirk, K.; Saliba, K. J.; Smith, G. D. Tetrahedron 1999, 55, 13513.
10.1021/ol006649q CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/21/2000

2, led us to undertake its synthesis. We now report the first
total synthesis of 2, the demonstration that the structure of
naturally occurring 2 is correctly assigned, and the conversion
of 2 to 1.
Scheme 1. Synthesis of Calothrixins A and B
Our past success3 with ortho lithiation-based strategies for
total synthesis encouraged us to consider such an approach
to 2. The two retrosyntheses suggested in Figure 1 appeared
Figure 1. Retrosynthetic analysis of 3.
particularly attractive because they offered the prospect of
achieving a highly convergent synthesis from two relatively
simple precursors. In fact, we now report a total synthesis
of 2 in which the longest linear sequence is only two steps
from known compounds (four steps from commercially
available materials).
In pursuing the retrosyntheses in Figure 1, a search of the
literature4 led to a report in which Watanabe and Snieckus5
demonstrated that the strategy implicit in retrosynthesis b is
a viable one, at least when applied to simpler targets. In the
present instance, however, the reaction conditions reported
by Watanabe and Snieckus totally failed to give any of 3.
An extensive examination of reaction conditions was required
before the synthesis of 3 was reduced to practice.
The synthesis is summarized in Scheme 1. The known 66
and 7,7 each prepared in a one-pot operation from com-
mercially available materials 4 and 5, were coupled, again
in a one-pot operation, to give 13 (= 3) directly. None of
the putative intermediates 8, 10, 11, or 12 shown in Scheme
1 was isolated, but the choice of lithium tetramethylpiperidide
(LiTMP) as base was crucial. The base used to metalate the
amide in the original report (s-BuLi) failed with 7; t-BuLi
was equally useless under a variety of conditions. Attempts
to deprotonate 7 with lithium hexamethyldisilazide (pKa of
amine ) 29.5)8 led to no detectable reaction with 6. The
more basic lithium diisopropylamide (LDA, pKa of amine
) 35.7)8 resulted in the production of small amounts of 13.
The still more basic (pKa of amine ) 37.3)8 LiTMP was
most effective. The reaction of 7 with LDA and LiTMP
behaved as if an equilibrium was being established [quench-
ing of ostensibly 8, prepared using either LDA or LiTMP,
with TMSCl gave good yields of the TMS quinoline amide
9, but this normally near-instantaneous reaction was slow,
requiring several hours for consumption of 7/8, as judged
by thin-layer chromatography (any 8 present would have
been detected as 7)]. Accordingly, use of excess rather than
stoichiometric LiTMP in the deprotonation of 7 was exam-
ined. The optimized conditions9 enlisted 4 equiv of LiTMP
in the deprotonation of 7, without the need to add an
additional equivalent in the subsequent step involving the
deprotonation of 10.
Even with all the optimization, the yield of 13 from 7 is
only 12% [26% if one uses 0.4 equiv of 6 and calculates the
yield based on 6 (using 1 equiv of 6 does not improve the
yield based on 7)]. However, the route is extremely concise,
and the main step is amenable to operation on at least a gram
scale. Furthermore, after removal of the MOM protecting
group (74%), the sequence serves to confirm the structure
assigned to 2.10 In addition, selective oxidation of the pyridine
nitrogen of synthetic 2 with m-CPBA affords 1 (71%), whose
spectra are in excellent agreement with those reported for
natural 1.
In conclusion, we provide very short syntheses of calo-
3736
Org. Lett., Vol. 2, No. 23, 2000

thrixins A and B that confirm the assigned structures and
afford straightforward synthetic access to them and, in
principle, their analogues.
Supporting Information Available: Experimental pro-
cedures and characterization data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Acknowledgment. We thank Professor Rickards1 for
providing spectra of natural 1 and 2 and for the TLC and
NMR comparison of natural 2 with synthetic 2.
OL006649Q
Org. Lett., Vol. 2, No. 23, 2000
3737