sequential derivatization of the resulting C7-secondary hy-
droxy and the C13-carboxy groups with Mosher’s acid
chloride and PGME, respectively.12,13 However, both acid
and basic hydrolysis conditions afforded compound 2, which
undergoes further addition reactions to the aldehyde func-
tionality, as indicated by the presence of several minor
analogues in the crude reaction products (see SI for a
mechanistic proposal on the formation of 2). Furthermore,
the expected C1-C12 t-butyl hydroxy acid derived from 1
was not isolated even under more strenuous reaction condi-
tions, suggesting the possibility of dehydration at C7 (the
donated from S-adenosyl-L-methionine (SAM).16 Both the
C19 and C20 methyl groups likely also come from SAM,
whereas the C12 methyl resides at a predicted carbonyl site
in the nascent polyketide and, thus, should be produced by
a ꢀ-branch mechanism involving an HMGCoA synthase
cassette.17 A predicted NRPS module next adds glycine that
is ketide extended, reduced at the ꢀ-carbonyl, dehydrated,
and double bond isomerized from C16-C17 to form the
C17-C18 enamide. Cyclization to 1 likely occurs coincident
with offloading from this final NRPS module but must
overcome the extreme steric congestion that is present about
this ester functionality.18
1
δ4.88 oxymethine disappears in the H NMR spectra of
reaction crudes).
Pure palmyrolide A (1) showed significant inhibition of
Ca2+ oscillations in murine cerebrocortical neurons, with an
IC50 value of 3.70 µM (2.29-5.98 µM, 95% CI) (see SI).
Additionally, in an assay designed to detect sodium channel
blocking activity using mouse neuroblastoma (neuro2a) cells,
compound 1 suppressed the veratridine- and ouabain-induced
sodium overload that leads to cytotoxicity, with an IC50 of
5.2 µM. Finally, palmyrolide A was not cytotoxic when
tested against H-460 human lung adenocarcinoma cells up
to 20 µM. These results suggest that palmyrolide A (1) may
function as a VGSC antagonist to suppress spontaneous Ca2+
oscillations and to protect against veratridine-induced sodium
influx, making it an intriguing candidate for further phar-
macological exploration.
The unexpected stability of this ester, clearly due to a
combination of steric and electronic effects provided by the
contiguous t-butyl substituent at C7, and perhaps assisted
by the C19 methyl group, leads us to speculate that
cyanobacteria might produce secondary metabolites exhibit-
ing this architecture to prevent cleavage of the lactone ester
bond and preserve the bioactive macrocycle under a wide
variety of conceivable environmental conditions. Our specu-
lation that the t-butyl group in palmyrolide A protects against
lactone hydrolysis is consistent with known principles in
medicinal chemistry (e.g., pivalic acid esters) and represents
an interesting chemical biology concept that warrants further
exploration.
The configuration of the stereocenter at C14 was deter-
mined by ozonolysis and oxidative workup of palmyrolide
A (1), followed by acid hydrolysis of the resulting crude
reaction product to afford 2-methylglutaric acid (3), which
is commercially available as its pure (R)-(-)- and (S)-(+)-
enantiomers (route B). Further derivatization with (S)-(+)-
2-octanol and acetyl chloride afforded the amiable diester 4
which was more easily detected and purified than its diacid
precursor. Comparison of NMR data and specific rotation
values measured for 4 and the similarly derivatized standards
(see SI) led to assignment of the C14 chiral center as R.
Unfortunately, this procedure failed to produce a suitable
C1-C12 derivative, hindering our efforts to assign the
absolute configuration at C5 and C7 (relative configuration
shown in Figure 1A).
Palmyrolide A (1) is structurally related to the laingolides,
initially reported from Papua New Guinea specimens of
Lyngbya bouillonii.14 However, given the morphological
similarities between Lyngbya and Oscillatoria cyanobacteria,
and the fact that the laingolides producer was described solely
from morphological characters rather than with the inclusion
of phylogenetic analysis, we speculate that the laingolides
may have been actually isolated from an Oscillatoria sp.
Biogenetically,15 1 appears to derive from a combination of
PKS and NRPS pathways, with its t-butyl appendage likely
assembled from malonyl-CoA and three methyl groups
Acknowledgment. We thank A. Jansma (UCSD) and Y.
Su (UCSD) for assistance with NMR and HRMS data
acquisition, respectively. Support was provided by NIH
NS053398 (W.H.G and T.F.M), NSF CHE-0741968 (NMR,
UCSD), and Consejo Nacional de Ciencia y Tecnologia of
Mexico (I.S.M).
Supporting Information Available: Experimental, full
NMR data of 1, 2, and 4 (all stereoisomers), bioassay data,
and taxonomic characterization. This material is available
OL101752N
(11) Williamson, R. T.; Marquez, B. L.; Gerwick, W. H.; Kover, K. E.
Magn. Reson. Chem. 2000, 38, 265–273.
(12) Ohtani, I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. J. Am. Chem.
Soc. 1991, 113, 4092–4096.
(13) Nagai, Y.; Kusumi, T. Tetrahedron Lett. 1995, 36, 1853–1856.
(14) (a) Klein, D.; Braeckman, J. C.; Daloze, D. Tetrahedron Lett. 1996,
37, 7519–7520. (b) Klein, D.; Braeckman, J. C.; Daloze, D.; Hoffmann,
L.; Castillo, G.; Demoulin, V. J. Nat. Prod. 1999, 62, 934–936. (c) Matthew,
S.; Salvador, L. A.; Schupp, P. J.; Paul, V. J.; Luesch, H. J. Nat. Prod.
[online early access]. DOI: 10.1021/np1004032. Published online: August
12, 2010.
(15) This biogenetic dissection is also applicable to the laingolides which
are of related structure to 1.
(16) Grindberg, R. V.; Ishoey, T.; Brinza, D.; Esquenazi, E.; Liu, W.;
Coates, R. C.; Gerwick, L.; Dorrestein, P.; Pevzner, P.; Lasken, R.; Gerwick,
W. H., in preparation.
(17) Geders, T. W.; Gu, L.; Mowers, J. C.; Liu, H.; Gerwick, W. H.;
Hakansson, K.; Sherman, D. H.; Smith, J. L. J. Biol. Chem. 2007, 282,
35954–35963.
(10) Uhrin, D.; Batta, G.; Hruby, V. J.; Barlow, P. N.; Kover, K. E. J.
Magn. Reson. 1998, 130, 155–161.
(18) Ramaswamy, A. V.; Sorrels, C. M.; Gerwick, W. H. J. Nat. Prod.
2007, 70, 1977–1986.
Org. Lett., Vol. 12, No. 20, 2010
4493