Total synthesis of the putative structure of lucentamycin A

Article abstract of DOI:10.1021/ol902251c

A rapid and stereoselective enolate-Claisen rearrangement provides access to the 4-ethylidene-3-methylproline (Emp) subunit of lucentamycin A. Synthesis of the putative structure of the cytotoxic natural product suggests the need for structural revision.

Full text of DOI:10.1021/ol902251c

ORGANIC
LETTERS
2009
Vol. 11, No. 22
5298-5301
Total Synthesis of the Putative Structure
of Lucentamycin A
Ujjwal Pal,† Sujeewa Ranatunga,‡ Yamuna Ariyarathna,‡ and Juan R. Del Valle*,†
Drug DiscoVery Department, H. Lee Moffitt Cancer Center and Research Institute,
Tampa, Florida 33612, and Department of Chemistry and Biochemistry, New Mexico
State UniVersity, Las Cruces, New Mexico 88003
juan.delValle@moffitt.org
Received September 29, 2009
ABSTRACT
A rapid and stereoselective enolate-Claisen rearrangement provides access to the 4-ethylidene-3-methylproline (Emp) subunit of lucentamycin
A. Synthesis of the putative structure of the cytotoxic natural product suggests the need for structural revision.
Marine-derived bacteria are a prolific source of bioactive
natural products featuring complex amino acids.1 In 2007,
Fenical and co-workers reported the isolation of lucentamy-
cins A-D, a new family of cytotoxic metabolites from the
fermentation broth of Nocardiopsis lucentensis (Figure 1).2
Structural elucidation based on extensive NMR and degrada-
tion studies revealed unique tripeptides composed of an
N-acylated homoarginine (Har), a C-terminal leucine or
tryptophan, and a central 4-ethylidene-3-methylproline (Emp)
residue unprecedented in the natural product literature. Of
the four closely related structures, lucentamycin A showed
Figure 1. Proposed structures of lucentamycins A (1), B (2), C
(3), and D (4).
potent in Vitro cytotoxicity against human colon carcinoma
† H. Lee Moffitt Cancer Center.
‡ New Mexico State University.
(1) For selected reviews, see: (a) Mauger, A. B. J. Nat. Prod. 1996, 59,
1205–1211. (b) Fenical, W.; Jensen, P. R. Nat. Chem. Biol. 2006, 2, 666–
673. (c) Jones, A. C.; Gu, L.; Sorrels, C. M.; Sherman, D. H.; Gerwick,
W. H. Curr. Opin. Chem. Biol. 2009, 13, 216–223. (d) Moore, R. E.
J. Indust. Microbiol. Biotechnol. 1996, 16, 134–143. (e) Burja, A. M.;
Banaigs, B.; Abou-Mansour, E.; Grant Burgess, J.; Wright, P. C. Tetrahe-
dron 2001, 57, 9347–9377. (f) Luesch, H.; Harrigan, G. G.; Goetz, G.;
Horgen, F. D. Curr. Med. Chem. 2002, 9, 1791. (g) Ersmark, K.; Del Valle,
J. R.; Hanessian, S. Angew. Chem., Int. Ed. 2008, 47, 1202–1223.
(2) Cho, J. Y.; Williams, P. G.; Kwon, H. C.; Jensen, P. R.; Fenical,
W. J. Nat. Prod. 2007, 70, 1321–1328.
cells (HCT-116 IC50 ) 0.20 µM), and lucentamycin B
exibited weaker activity against the same cell line (IC50
)
11 µM).
As the first step in our investigation of lucentamycin
structure-activity relationships, we targeted 1 for synthesis
and sought to devise a versatile strategy for construction of
the 4-alkylidene-3-alkylproline subunit common to this
family. We were intrigued to find that a derivative of the
10.1021/ol902251c CCC: $40.75
Published on Web 10/28/2009
 2009 American Chemical Society

Scheme 1. Synthesis of Glycine Ester 10
of 5 with N-phenyltriflimide afforded the Z isomer of vinyl
triflate 6 as the major product in 78% yield. Subsequent Heck
carboxymethylation in DMF/MeOH led to significant erosion
of alkene isomeric purity. We found that the integrity of the
Z alkene could be largely preserved by using MeOH as the
sole solvent. Thus, treatment of 6 with Pd(OAc)2, PPh3,
DIEA, and MeOH under a CO atmosphere gave the expected
enoates in a 13:1 ratio and 87% overall yield.7 After
separation by column chromatography, we isolated pure 7
in 79% yield and proceeded with HF-mediated cleavage of
the silyl ether and ester reduction to provide allylic diol 8.
Selective TBDPS-protection of the primary alcohol was then
followed by condensation with N-Boc-glycine to give rear-
rangement substrate 10.
We had initially planned to employ Kazmeier’s chelation
conditions8 for the ensuing [3,3] sigmatropic shift, as model
substrates afforded good yields in our hands. However, we
were unable to find precendent for a glycine enolate-Claisen
rearrangement carried out on a substrate with the alkene
substitution pattern of 10. The silyloxymethyl substituent did
in fact pose an obstacle, as evidenced by low conversions
in the presence of LDA, LiHMDS, or KHMDS, with or
without ZnCl2. Interestingly, addition of 10 into 2.2 equiv
of NaHMDS9 in THF at -78 °C, followed by warming to
rt, gave 63% yield of alcohol 9 and minor amounts of the
desired carboxylic acid. Reducing the amount of NaHMDS
to 1.1 equiv resulted in near-quantitative recovery of 9.10
We reasoned that this unusual ester cleavage is favored over
Figure 2. Retrosynthesis of Emp and proposed [3,3] rearrangement
transition state.
Emp residue had been prepared by Gais and co-workers as
part of a study on the synthetic utility of chiral vinyl
aminosulfoxonium salts.3 Our interest in chimeric prolines
prompted us to explore an alternate route employing lactic
acid as a chiral progenitor and a stereoselective enolate-
Claisen rearrangement as the key step. Here, we report the
first total synthesis of 1, which also calls into question the
proposed structure of lucentamycin A.4
We envisioned a general approach toward 3,4-disubstituted
prolines by way of the linear precursor shown in Figure 2.
In turn, this polysubstituted allylglycine substrate could be
efficiently accessed via a stereoselective [3,3] sigmatropic
rearrangement. On the basis of chair transition state models
proposed for similar enolate-Claisen rearrangements,5 it
became apparent that the appropriate allyl ester substrate
should possess an S configuration and E alkene geometry to
afford the desired 2S,3R,4Z Emp configuration. The structure
of the requisite chiral alcohol was retrosynthetically traced
back to (-)-ethyl lactate as a readily available and inexpen-
sive starting material.
Synthesis of the key allyl ester commenced with ketone
5, conveniently obtained on large scale over three steps
according to precedent.6 Trapping of the potassium enolate
(3) (a) Tiwari, S. K.; Gais, H. J.; Lindenmaier, A.; Babu, G. S.; Raabe,
G.; Reddy, L. R.; Kohler, F.; Gunter, M.; Koep, S.; Iska, V. B. R. J. Am.
Chem. Soc. 2006, 128, 7360–7373. (b) Tiwari, S. K.; Schneider, A.; Koep,
S.; Gais, H. J. Tetrahedron Lett. 2004, 45, 8343–8346.
(6) (a) Ferrero´, M.; Galobardes, M.; Mart´ın, R.; Montes, T.; Romea,
P.; Rovira, R.; Urp´ı, F.; Vilarrasa, J. Synthesis 2000, 2000, 1608–1614. (b)
Paterson, I.; Wallance, D. J.; Cowden, C. J. Synthesis 1998, 639–652.
(7) The alkene geometry was confirmed after treatment of 7 with HF
and comparison with spectroscopic data for the racemic enoate: Cha, J. H.;
Pae, A. M.; Choi, K. I. I.; Cho, Y. S.; Koh, H. Y.; Lee, E. J. Chem. Soc.,
Perkin Trans. 1 2001, 2079–2081.
(4) Although we were unable to find a detailed account of a previous
synthesis of 1, the Lindsley group recently disclosed their efforts towards
lucentamycin A at a national ACS meeting: Daniels, N. R.; Lindsley, C. W.
Abstracts of Papers, 235th National Meeting of the American Chemical
Society, New Orleans, LA, April 6-10, 2008; American Chemical Society:
Washington, DC, 2008; ORGN 698.
(8) (a) Kazmaier, U. Angew. Chem., Int. Ed. 1994, 33, 998–999. (b)
Kazmaier, U.; Krebs, A. Angew. Chem., Int. Ed. 1995, 34, 2012–2014.
(9) Looper, R. E.; Runnegar, M. T. C.; Williams, R. M. Tetrahedron
2006, 62, 4549–4562.
(5) (a) Piero, V.; Hans-Ju¨rgen, H.; Hans, S. HelV. Chim. Acta 1975, 58,
1293–1309. (b) Bartlett, P. A.; Barstow, J. F. J. Org. Chem. 2002, 47, 3933–
3941. (c) Kazmaier, U. In The Claisen Rearrangement; Hiersemmann, M.,
Nubbemeyer, U., Eds.; Wiley-VCH Verlag Gmbh & Co.: Weinheim,
Germany, 2007; pp 233-299.
(10) Based on pKa considerations, our results suggest an intramolecular
fragmentation induced by attack of the nitrogen anion onto the carbonyl
carbon.
Org. Lett., Vol. 11, No. 22, 2009
5299

Scheme 2. Synthesis of 1
thermally induced [3,3] sigmatropic rearrangement at low
temperatures. Indeed, when 10 was added to a solution of
2.5 equiv of NaHMDS in THF at rt, consumption of the
starting material was complete within 1 min, and we obtained
11 in 76% yield, along with 23% of fragmentation product
9 (Scheme 2). On gram-scale, 9 was routinely recycled to
provide 11 in >90% overall yield.
With suitable rearrangement conditions in hand, we
proceeded to esterify the carboxyl group of 11 and prepare
the allylic alcohol for activation and displacement. Unfor-
tunately, treatment of the intermediate methyl ester with HF/
pyridine or TBAF/THF led to silyl ether cleavage and rapid
cyclization to afford lactone 12. To circumvent this, we
coupled 11 directly to L-leucine tert-butyl ester in the
presence of PyBOP and NEt3. Analysis of the crude product
by HPLC and identification of minor diastereomers by MS
and UV indicated a 59.6:2.8:1 dr, implying >23:1 er (and
17:1 dr) for 11.11 Dipeptide 13 was then purified to provide
the major isomer and treated with HF/pyridine to unmask
the primary alcohol in 86% yield.
desired reaction was formally an attempt at a disfavored
5-endo-trig cyclization (SN1) under these conditions.14 As
an alternative, acidic cleavage of the N-Boc group followed
directly by treatement with K2CO3 in acetone led to rapid
SN2 cyclization to give pyrrolidine 15 in 56% yield over 3
steps.
Figure 3 depicts the key NOE enhancements observed
about the pyrrolidine ring of the Emp-Leu subunit 15.
Although the geometry of the alkene could be easily deduced,
there was only indirect evidence to suggest a cis relationship
for H2 and H3. Both protons exhibited correlation to H5R
(and not H5ꢀ), but a direct NOE between H2 and H3 was
conspicuously weak relative to other signals. To obtain more
Efforts to promote one-pot activation/cyclization to form
the pyrrolidine ring were unsuccessful under a variety of
conditions (PPh3, I2; PPh3, DIAD; MsCl, pyridine; DAST).
Notably, attempted mesylation of 14 afforded a high yield
of the allylic chloride,12 indicating a reluctance toward
intramolecular displacement despite the reactivity of the
intermediate allylic mesylate. Since carbamate nitrogens have
been shown to participate in similar displacements through
planar allyl cation mechanisms,13 we presumed that our
(11) See Supporting Information for details.
(12) For examples of allylic chlorination with methanesulfonyl chloride,
see: (a) Artman, G. D.; Grubbs, A. W.; Williams, R. M. J. Am. Chem. Soc.
2007, 129, 6336–6342. (b) Li, F.; Castle, S. L. Org. Lett. 2007, 9, 4033–
4036.
Figure 3. Confirmation of 2S,3R,4Z Emp configuration.
Org. Lett., Vol. 11, No. 22, 2009
5300

definitive evidence we carried out crystallization trials on a
number of intermediates and derivatives. Suitable diffraction
quality crystals were obtained from compounds 20 and 21
(derived from 14 and 12), and X-ray structures provided
unambiguous confirmation of our stereochemical assign-
ments.11
concentration effect, while variable temperature and pH
experiments also failed to induce shifts in C-H proton
resonances. In addition, the specific rotation we obtained for
1 ([R]25D -35.2, c 0.500 in MeOH) was significantly different
from that reported for lucentamycin A ([R]25D -6.3, c 0.175
in MeOH).
The final assembly of 1 employed homoarginine dervative
17, available in one step from Fmoc-Lys-OH using Good-
man’s guanidinylation reagent (Scheme 2).15 Condensation
of 17 and 15 in the presence of DEPBT15 afforded tripeptide
18 in good yield. Fmoc-deprotection of the N-terminus was
then carried out by treatment with excess diethylamine,
followed by benzoylation.17 Finally, global deprotection of
19 with 90:5:5 TFA/TES/DCM and subsequent purification
by RP-HPLC completed the total synthesis of presumed
lucentamycin A.18
Given our structural confirmation of intermediates 20 and
21 and the fact that most signals attributable to the Emp
residue in 1 exhbit an NMR discrepancy (in position or
splitting) relative to the natural product, we believe that a
stereochemical revision is warranted. While the exact nature
of this correction is not currently known, key NOE correla-
tions observed for natural lucentamycin A strongly support
the 2,3-cis substitution and Z alkene configuration originally
proposed for Emp.2 It is therefore likely that the absolute
stereochemistry of the novel proline residue should be
revised.
1
Although the H NMR of 1 and natural lucentamycin A
were qualitatively very similar, we noticed a number of
chemical shift discrepancies (0.05-0.12 ppm) that could not
be attributed to experimental error.19 Clear differences were
also evident upon comparison of 13C NMR spectra, with 6
out of 8 Emp signals exhbiting deviations >0.5 ppm. Serial
dilution NMR studies of 1 in DMSO-d6 ruled out a possible
We are currently applying our rearrangement strategy to
the synthesis of other polysubstituted prolines and pipecolic
acids, as well as to the suspected natural diastereomer of 1.
Results from these studies will be reported in due course.
Acknowledgment. We thank Profs. William Fenical
(University of California-San Diego) and Philip Williams
(University of Hawaii) for providing raw NMR data for
natural lucentamycin A and Prof. Bill Baker (University of
South Florida) for helpful discussions. We also gratefully
acknowledge Dr. Eileen Duesler (University of New Mexico)
for X-ray diffraction studies on compounds 20 and 21.
Funding for this work was provided, in part, by the Moffitt
Cancer Center.
(13) (a) Gosselin, F.; Lubell, W. D. J. Org. Chem. 2000, 65, 2163–
2171. (b) Vaswani, R. G.; Chamberlin, A. R. J. Org. Chem. 2008, 73, 1661–
1681.
(14) In contrast to the examples in ref 13, the incipient electrophilic sp2
center in 14 is formally within the cyclic transition state (endo), leading to
a strained attack geometry. See: Johnson, C. D. Acc. Chem. Res. 2002, 26,
476–482.
(15) Feichtinger, K.; Zapf, C.; Sings, H. L.; Goodman, M. J. Org. Chem.
1998, 63, 3804–3805.
(16) Li, H. T.; Jiang, X. H.; Ye, Y. H.; Fan, C. X.; Romoff, T.; Goodman,
M. Org. Lett. 1999, 1, 91–93.
(17) As expected on the basis of Young’s test considerations, attempts
to couple 15 directly to N-Bz-Har(Boc)2-OH resulted in moderate yields
of coupled product and significant racemization of the Har residue.
(18) Semi-preparative RP-HPLC in the presence and absence of formic
acid as an aqueous additive gave products with identical spectroscopic
properties. We presume that purified 1 exists in the neutral zwitterionic
form.
Supporting Information Available: Experimental data
and copies of NMR spectra for all new compounds; HPLC
and UV spectra for 13 and 1; comparative NMR spectra for
1 and natural lucentamycin A; X-ray diffraction data for
compounds 20 and 21. This material is available free of
charge via the Internet at http://pubs.acs.org.
(19) Although an authentic sample of lucentamycin A was unavailable
for HPLC co-injection studies, analysis of original NMR data files
graciously provided by Prof. Fenical clearly show chemical shift
discrepancies. Superimposed NMR expansions are included in the
Supporting Information.
OL902251C
Org. Lett., Vol. 11, No. 22, 2009
5301