Stereochemistry of the core structure of the mycolactones

Full text of DOI:10.1021/ja0105414

5128
J. Am. Chem. Soc. 2001, 123, 5128-5129
Stereochemistry of the Core Structure of the
Mycolactones
Andrew B. Benowitz,† Steve Fidanze,† P. L. C. Small,‡ and
Yoshito Kishi*,†
Department of Chemistry and Chemical Biology
HarVard UniVersity, Cambridge, Massachusetts 02138
Department of Microbiology
UniVersity of Tennessee, KnoxVille, Tennessee 37996
Figure 1. Mycolactone A, Z-∆4′,5′; mycolactone B, E-∆4′,5′
.
ReceiVed February 28, 2001
The mycolactones (Figure 1) were isolated in 1999 by Small
and co-workers1 from Mycobacterium ulcerans, the causative
pathogen of Buruli ulcer. This disease is characterized by the
formation of large, painless, necrotic ulcers, and the lack of an
acute inflammatory response. Evidence from animal studies
suggests that the mycolactones are directly responsible for the
observed pathology,1a and they have attracted considerable
attention for their highly potent apoptotic activity2 as well as for
being the first examples of polyketide macrolides to be isolated
from a human pathogen.3 The gross structure of these compounds
was elucidated through 2-D NMR experiments; however, the
stereochemistry remained undetermined.
We recognized that the concepts recently advanced in our
laboratory related to the structural elucidation of natural products4,5
were uniquely suited to address the stereochemical assignment
of the mycolactones. To this end, the use of a combined approach
employing both an NMR database and the preparation of model
compounds to determine the relative and absolute configuration
of the mycolactone core structure is herein reported.
It was envisioned that the stereochemical elucidation of the
mycolactones would begin by preparing an NMR database of
model compounds corresponding to C.13-C.20. Comparison of
this database with the reported NMR data for the mycolactones
was expected to allow us to predict the relative stereochemistry
at C.16, C.17, and C.19. This database was prepared beginning
with methyl (R)-3-hydroxybutyrate to set the C.19 stereocenter.
The C.16 and C.17 stereocenters were subsequently installed via
the Brown crotylboration protocol6 employing separately each
enantiomer of Ipc2BOMe with each of Z- and E-2-butene.
Subtraction of the chemical shift values of this database (i.e., 1a-
d) from the reported NMR data for the mycolactones (Figure 2)
clearly demonstrates the relative configuration of C.16, C.17, and
C.19 to be all-syn (i.e., 1a).
Figure 2. Graphs representing chemical shift difference (y-axis, ppm)
of 1H (500 MHz, left column) and 13C (125 MHz, right column) spectra
between each corresponding database compound 1a-d atom (x-axis) and
the reported values for the mycolactones (CD3COCD3).
Figure 3. Four subgroups of C.1 to C.25 diastereomers.
preparing a series of C.1-C.25 synthetic diastereomers for
comparison with the mycolactones.
Having predicted the relative stereochemistry at C.16, C.17,
and C.19, we next sought to determine the relative stereochemistry
at C.5, C.6, and C.11, C.12. However, we speculated that, in
contrast to the desertomycins/oasomycins,5 a database approach
that treated these two stereochemical clusters independently was
not likely to be successful as three of these four stereocenters
were likely to interact with each other within the 12-membered
lactone. We therefore explored the possibility of logically
Several comments should be made regarding this strategy. First,
in principle, a series of C.1-C.14 model compounds should be
sufficient to determine the relative stereochemistry at the C.5-
C.12 cluster. However, we opted to employ diastereomers of the
entire core structure both to correlate the relative stereochemistry
of the C.5-C.12 cluster to the C.16-C.19 cluster and also to
determine the absolute configuration of these stereocenters through
correlation with the natural core structure. Second, eight C.1-
C.25 diastereomers should be required to determine the relative
stereochemistry at the C.5-C.12 cluster, plus one additional
diastereomer to correlate this cluster to the C.16-C.19 cluster.
However, the amount of required synthetic work can be reduced
by assuming that the interactions within the ring will be manifest
primarily at C.5, C.6, and C.11. Therefore, the eight possible
diastereomers can be logically divided into four subgroups
consisting of the four possible C.5, C.6, and C.11 diastereomers
(holding C.5 constant); the two diastereomers contained in each
subgroup will differ only in the configuration at C.12 (Figure 3).
Holding the C.16, C.17, and C.19 stereocenters constant, we
expect that the preparation of one member of each subgroup would
yield a compound with an NMR spectrum that should correlate
† Harvard University.
‡ University of Tennessee.
(1) (a) George, K. M.; Chatterjee, D.; Gunawardana, G.; Welty, D.;
Hayman, J.; Lee, R.; Small, P. L. C. Science 1999, 283, 854. (b) Gunawardana,
G.; Chatterjee, D.; George, K. M.; Brennan, P.; Whittern, D.; Small, P. L. C.
J. Am. Chem. Soc. 1999, 121, 6092.
(2) George, K. M.; Pascopella, L.; Welty, D. M.; Small, P. L. C. Infect.
Immun. 2000, 68, 877.
(3) Rohr, J. Angew. Chem., Int. Ed. 2000, 39, 2847.
(4) AAL toxins and fumonisins: (a) Boyle, C. D.; Harmange, J.-C.; Kishi,
Y. J. Am. Chem. Soc. 1994, 116, 4995. (b) Boyle, C. D.; Kishi, Y. Tetrahedron
Lett. 1995, 36, 5695 and references therein.
(5) Desertomycins/oasomycins: (a) Kobayashi, Y.; Lee, J.; Tezuka, K.;
Kishi, Y. Org. Lett. 1999, 1, 2177. (b) Lee, J.; Kobayashi, Y.; Tezuka, K.;
Kishi, Y. Org. Lett. 1999, 1, 2181. (c) Kobayashi, Y.; Tan, C.-H.; Kishi, Y.
J. Am. Chem. Soc. 2001, 123, 2076 and references therein.
(6) Brown, H. C.; Bhat, K. S. J. Am. Chem. Soc. 1986, 108, 5919.
10.1021/ja0105414 CCC: $20.00 © 2001 American Chemical Society
Published on Web 05/05/2001

Communications to the Editor
J. Am. Chem. Soc., Vol. 123, No. 21, 2001 5129
Scheme 1a
1
Figure 4. (a) H NMR (500 MHz, CD3OD) of the tri-(S)-Mosher ester
of the core structure of the natural mycolactones. (b) 1H NMR (500 MHz,
CD3OD) of the tri-(S)-Mosher ester of 4b.
a Conditions: (a) t-BuLi (3 equiv), ZnCl2, Pd(Ph3P)4, THF, 60%; (b)
(1) CH2Cl2-H2O-TFA (8:2:0.5), 77%; (2) PivCl, pyr., 99%; (3) TESCl,
imid., CH2Cl2, 91%; (4) DiBAl-H, CH2Cl2, -78 °C, 98%; (5) I2, Ph3P,
imid., Et2O-MeCN (3:1), 91%; (c) t-BuLi (3 equiv), ZnCl2, Pd(Ph3P)4,
THF, 50%; (d) (1) HF‚pyr.-pyr.-THF (1:1:4), THF, 72%; (2) TEMPO,
NCS, Bu4NCl, CH2Cl2-pH 8.6 buffer (1:1), 95%; (3) NaClO2, NaH2PO4,
m-(MeO)2-C6H4, DMSO-t-BuOH (1:1), 94%; (e) (1) Cl3C6H2COCl,
i-Pr2NEt, PhH; DMAP, PhH, 70%; (2) CH2Cl2-H2O-TFA (8:2:0.5),
62%; (3) HF‚pyr., MeCN, 77%.
1
and reagents for each diastereomer. H NMR comparison (500
MHz, CD3OD) of each core structure diastereomer with an
authentic sample prepared via hydrolysis of the natural myco-
lactones (K2CO3, MeOH) revealed that only 4b corresponded
exactly to the natural material; the other diastereomers were
significantly different. As expected, preparation and NMR
comparison of the other member of this subgroup (i.e., 4a) showed
that this diastereomer corresponded very closely, but not exactly,
to the natural core structure.
closely to that of the core structure of the mycolactones. Finally,
the C.16,C.17,C.19 epimer of the closest matching diastereomer
would be prepared to determine the relative stereochemistry of
the mycolactone core structure.
We envisioned preparing each core structure diastereomer via
a convergent strategy employing fragments corresponding to C.1-
C.7, C.8-C.13, and C.14-C.20 (i.e., 7, 6, and 10, Scheme 1).
This approach affords an extremely flexible method for the
preparation of each diastereomer through the appropriate choice
of stereochemistry in each fragment.
In the event, preparation of diastereomer 4b was accomplished
(Scheme 1) by coupling the C.8-C.13 vinyl iodide 6 with the
C.1-C.7 iodide 7 employing a modification7 of Negishi’s
conditions8 to give 8. This compound was next subjected to a
five-step sequence to generate iodide 9,9 which was then coupled
with the C.14-C.20 vinyl iodide fragment 10 via a second
modified Negishi coupling to give 11.7, 8
Both the TES ether and the primary TBS ether were then
selectively removed with buffered HF‚pyr., and the primary
alcohol was selectively oxidized first to the aldehyde with
TEMPO,10 and then to the carboxylic acid with buffered NaClO2
to give 12. Yamaguchi macrolactonization11 proceeded smoothly
to give the lactone, and a two-step deprotection protocol afforded
the C.1-C.25 diastereomer 4b.
To determine the relative stereochemical relationship between
the C.5-C.12 cluster and the C.16-C.19 cluster, the (C.16,C.17,
1
C.19)-epi-4b diastereomer was prepared; as expected, the H
NMR spectrum (500 MHz, CD3OD) of this compound was almost
identical to that of the natural core structure. Therefore, the tri-
(S)- and tri-(R)-Mosher esters of both 4b and (C.16,C.17,C.19)-
epi-4b were prepared and compared to the tri-(S)-Mosher ester
prepared from the natural material. 1H NMR comparison showed
that only the tri-(S)-Mosher ester prepared from 4b was identical
to the tri-(S)-Mosher ester prepared from the natural mycolactone
core structure (Figure 4), thereby proving that 4b represents both
the relative and absolute configuration of the core structure of
the natural mycolactones.
In summary, we have employed a combined NMR database
and model compound synthesis approach to determine that 4b
represents the relative and absolute stereochemistry of the core
structure of the mycolactones. Additionally, the NMR database
approach has been employed to determine the relative stereo-
chemistry at C.12′, C.13′, C.15′ as â, R, R,12 and full details of
these results will be reported in due course.
Preparation of one member of each additional subgroup (i.e.,
2b, 3a, and 5a) proceeded in an analogous manner as the synthesis
of 4b, employing the appropriate asymmetric starting materials
Acknowledgment. We are grateful to the National Institutes of Health
for generous financial support (CA 22215). A.B.B. gratefully acknowl-
edges support in the form of an American Cancer Society postdoctoral
fellowship (PF-99-117-01-CDD, partially funded by the National Fisheries
Institute).
(7) Smith, A. B., III; Qiu, Y.; Jones, D. R.; Kobayashi, K. J. Am. Chem.
Soc. 1995, 117, 12011.
(8) Negishi, E.; Okukado, N.; King, A. O.; Van Horn, D. E.; Spiegel, B.
I. J. Am. Chem. Soc. 1978, 100, 2254.
(9) Choice of solvent was found to be critical for the conversion of the
alcohol to the iodide: Strunz, G. M.; Finlay, H. J. Can. J. Chem. 1996, 74,
419.
(10) Einhorn, J.; Einhorn, C.; Ratajczak, F.; Pierre, J.-L. J. Org. Chem.
1996, 61, 7452.
Supporting Information Available: Complete experimental details
for the synthesis of 4b as well as 1H NMR comparisons (PDF). This
material is available free of charge via the Internet at http://pubs.acs.org.
JA0105414
(11) (a) Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M. Bull.
Chem. Soc. Jpn. 1979, 52, 1989. (b) Hikota, M.; Sakurai, Y.; Horita, K.;
Yonemitsu, O. Tetrahedron Lett. 1990, 31, 6367.
(12) Szlosek-Pinaud, M., Kishi, Y. Unpublished results.