Absolute configuration and total synthesis of a novel antimalarial lipopeptide by the de novo preparation of chiral nonproteinogenic amino acids

Article abstract of DOI:10.1021/ol300293a

The absolute configuration (via degradation and Marfey's derivatization studies) and the total synthesis of a novel antimalarial lipid-peptide isolated from Streptomyces sp. (IC₅₀ = 0.8 μM, Plasmodium falciparum 3D7) is disclosed. To this end, versatile stereocontrolled routes to nonproteinogenic amino acids (via catalytic Mannich, Sharpless methods) and enantiomeric trans fatty acids (via Evans alkylation, Kocienski-Julia olefination) have been developed.

Full text of DOI:10.1021/ol300293a

ORGANIC
LETTERS
2012
Vol. 14, No. 6
1560–1563
Absolute Configuration and Total
Synthesis of a Novel Antimalarial
Lipopeptide by the de Novo Preparation of
Chiral Nonproteinogenic Amino Acids
Shibaji K. Ghosh,† Brinda Somanadhan,‡ Kevin S.-W. Tan,§ Mark S. Butler,†,‡, and
Martin J. Lear*,†
Department of Chemistry, Faculty of Science, and Medicinal Chemistry Program of the
Life Sciences Institute, National University of Singapore, 3 Science Drive 3, Singapore
117543, Singapore, MerLion Pharmaceuticals, The Capricorn #05-01, Singapore
Science Park II, Singapore 117528, and Department of Microbiology, National
University of Singapore, 5 Science Drive 2, Singapore 117597, Singapore
Martin.Lear@nus.edu.sg
Received February 6, 2012
ABSTRACT
The absolute configuration (via degradation and Marfey’s derivatization studies) and the total synthesis of a novel antimalarial lipid-peptide
isolated from Streptomyces sp. (IC50 = 0.8 μM, Plasmodium falciparum 3D7) is disclosed. To this end, versatile stereocontrolled routes to
nonproteinogenic amino acids (via catalytic Mannich, Sharpless methods) and enantiomeric trans fatty acids (via Evans alkylation,
KocienskiꢀJulia olefination) have been developed.
Without multiple modes of attack, even our best and
latest antimalarial agent will face drug resistance in the
field. New lead structures with atypical bioactivities are
therefore of high interest. In our efforts to understand drug
resistance and synthesize lipid conjugates,1 we became
interested in elucidating the novel structure and function
of a Streptomyces-derived lipidated peptide metabolite
(1, Figure 1). This was originally disclosed in a 2004 patent
application by the Yamanouchi Pharmaceutical Co. (now
Astella Pharma) to inhibit the growth of Plasmodium
falciparum.2 Independent to this disclosure, MerLion
Pharmaceuticals isolated the same metabolite (ꢀ)-1 from
various Streptomyces species and confirmed its planar
structure by NMR and MS analysis. Similar to the patent,2
they found the lipopeptide (ꢀ)-1 to have little or no activity
against a panel of mammalian, fungal, and Gram positive
bacteria cell lines. This is curious since antimalarial pep-
tides, although poorly understood, can often possess anti-
bacterial activity via membrane disruption.3 In view of
providing primary antimalarial data and material for
mode of action studies, we communicate the stereochemi-
cal determination and first total synthesis of (ꢀ)-1. This
required the nontrivial absolute stereogenic identification
and denovo synthesisof two nonproteinogenic amino acids
(3 and 4) and the synthesis of both enantiomers of the
methyl-branched trans fatty acid 5 (Figure 1).
† Department of Chemistry, National University of Singapore.
‡ MerLion Pharmaceuticals.
§ Department of Microbiology, National University of Singapore.
Current address: Institute for Molecular Bioscience, University of
Queensland, St Lucia, Queensland 4072, Australia.
Our studies began by determining the absolute config-
uration of each constituent 2ꢀ5 (Figure 1). A sample of the
natural product (ꢀ)-1 (5 mg) was refluxed in 6 M HCl for
(1) (a) Ch’ng, J. H.; Kotturi, S. R.; Chong, A. L.; Lear, M. J.; Tan,
K. W. Cell Death Dis. 2010, 1, e26. (b) Ali, A.; Wenk, M. R.; Lear, M. J.
Tetrahedron Lett. 2009, 50, 5664–5666.
(2) Nakagawa, A.; Tanaka, K.; Yamaguchi, H.; Nagai, K.; Kamigiri,
K.; Takeda, Y.; Suzumura, K.; Takahashi, S.; (Yamanouchi Pharma-
ceutical Co., Ltd., Japan). Application: WO, 2004/111076.
(3) Bell, A. Curr. Pharm. Des. 2011, 17, 2719–2731.
r
10.1021/ol300293a
Published on Web 03/05/2012
2012 American Chemical Society

Scheme 1. Organocatalytic Synthesis of R-Amino-γ-lactone 11
solution and Boc N-protection cleanly afforded the desired
amino ester 10. Refluxing 10 in 6 M HCl for 24 h then
produced the expectedγ-lactone 11. Further derivatization
with Marfey’s reagent, and analysis with naturally derived
material, indicated 3 to be (2S,3R).
Figure 1. Structures of Marfey’s reagent, amino acid, and fatty
acid components of the natural product including the two
nonproteinogenic amino acids 3 and 4.
Due to the practical limitations and inefficiencies of
reported routes to selected diastereomers of the amino
acid 4,9 we designed a general strategy to provide all four
diastereomers of 4 via a stereoselective dihydroxylation of
a cis or trans olefin (cf. 15).10 The specific methodology
developed to obtain the naturally occurring isomer in 1 is
shown in Scheme 2. Thus the propargylic PMB ether 12
was activated with methyl chloroformate 13 to generate
1411 for a carefully controlled Michael addition of Me2-
CuLi to afford the Z-olefin 1512 exclusively. Sharpless
asymmetric dihydroxylation of 15 then gave the diol 16
in 82% ee.13 Cyclic sulfate formation to 17 (using SOCl2
and then RuCl3ꢀNaIO4 oxidation)14 followed by regiose-
lective azide opening and acidic hydrolysis afforded the
azido alcohol 18.
24 h, and the hydrolyzed amino acid mixture was deriva-
tized with either UV-active CbzCl or Marfey’s reagent.4
The lipid portion 5 did not survive the 6 M HCl treatment,
and attemptstohydrolyzethe acidunder milder conditions
were unsuccessful. However, the aspartic acid (2) in the
natural product was recovered and readily assigned as the
L-isomer (S). This was achieved by comparison of the
HPLC retention time and MS of the Marfey’s derivatized
hydrolysis products with authentic standards of L- and D-
aspartic acid (see Supporting Information).
Our initial analysis of the LC-MS data of the hydrolysis
mixture failed to identifiy derivatives of 3. However, NMR
and ESI-MS analysis of a Cbz N-protected isolate indi-
cated that the carboxylic acid 3 had cyclized to a γ-lactone
during hydrolysis. The lactone 11 was thus isolated after
basifying the acidic aqueous phase and extracting with
CHCl3. This conveniently allowed a trans relative stereo-
chemistry to be assigned to the isolated lactone 11 by 1H
NMR and NOE experiments (1JHR/β 12 Hz).5 A Marfey’s
derivative was then formed from naturally isolated 11.
To determine the absolute stereochemistry of 11 and
provide further evidence for the cyclization process, an
expedient synthetic route was designed (Scheme 1). Thus,
the L-proline catalyzed Mannich reaction6 between butan-
2-one (6) and the imine 7 generated the syn-adduct 8 in
99% ee. Although Wittig-type reactions failed, presum-
ably due to the sterics of R-methylation, standard Tebbe
treatment7 of the keto-ester 8 produced the olefin 9 chemo-
selectively in 45% yield.
After extensive studies to convert 18 to 20, we found that
the tertiary alcohol of 18 first needed TES protection
before oxidation of the PMB-protected alcohol. However,
the PMB deprotection step also needed care and DDQ
oxidative removal to 19 required buffered pH 7.5.15 In
both cases, undesired lactone formation could be cir-
cumvented, and the best oxidation sequence for 19 was
16
achieved by using TEMPO combined with PhI(OAc)2
(9) Wehbe, J.; Kassem, T.; Rolland, V.; Rolland, M.; Tabcheh, M.;
Roumestant, M. L.; Martinez, J. Org. Biomol. Chem. 2003, 1, 1938–
1942.
(10) Stereodivergent synthesis of all diastereomers of 16 will be
reported elsewhere.
(11) Takemura, A.; Fujiwara, K.; Shimawaki, K.; Murai, A.; Kawai,
H.; Suzuki, T. Tetrahedron 2005, 61, 7392–7419.
(12) Leonard, M. S.; Carroll, P. J.; Joullie, M. M. J. Org. Chem. 2004,
69, 2526–2531.
Next, deprotection of the amino p-methoxyphenyl
(PMP)groupwasfound problematic, butreverseaddition8
of compound 9 to a ceric ammonium nitrate (CAN)
(13) Absolute stereochemistry of 16 was confirmed by reduction and
PMB cleavage to its known tetraol: Moen, A. R.; Ruud, K.; Anthonsen,
T. Eur. J. Org. Chem. 2007, 2007, 1262–1266.
(4) Marfey, P. Carlsberg Res. Commun. 1984, 49, 591–596.
(5) Jaime, C.; Segura, C.; Dinares, I.; Font, J. J. Org. Chem. 1993, 58,
154–158.
(14) Shao, H.; Goodman, M. J. Org. Chem. 1996, 61, 2582–2583.
(15) Ley, S. V.; Tackett, M. N.; Maddess, M. L.; Anderson, J. C.;
Brennan, P. E.; Cappi, M. W.; Heer, J. P.; Helgen, C.; Kori, M.;
Kouklovsky, C.; Marsden, S. P.; Norman, J.; Osborn, D. P.; Palomero,
M. A.; Pavey, J. B. J.; Pinel, C.; Robinson, L. A.; Schnaubelt, J.; Scott,
J. S.; Spilling, C. D.; Watanabe, H.; Wesson, K. E.; Willis, M. C.
ꢀ
(6) Cοrdova, A.; Notz, W.; Zhong, G.; Betancort, J. M.; Barbas,
~
C. F., III. J. Am. Chem. Soc. 2002, 124, 1842–1843.
(7) Tebbe, F. N.; Parshall, G. W.; Reddy, G. S. J. Am. Chem. Soc.
1978, 100, 3611–3613.
Chem.;Eur. J. 2009, 15, 2874–2914.
(8) Overman, L. E.; Owen, C. E.; Pavan, M. M.; Richards, C. J. Org.
Lett. 2003, 5, 1809–1812.
(16) De Mico, A.; Margarita, R.; Parlanti, L.; Vescovi, A.; Piancatelli,
G. J. Org. Chem. 1997, 62, 6974–6977.
Org. Lett., Vol. 14, No. 6, 2012
1561

Scheme 2. Stereocontrolled Synthesis of Monomethyl Ester of Amino Acid 4
followedbyPinnick oxidation17 toaffordthe monoester20
in 63% yield over two steps.
Scheme 3. Synthesis of Chiral Trans Fatty Acids 25a and 25b
To confirm the absolute and relative stereochemistry,
the azide ester 20 was reduced and hydrolyzed with
6 M HCl to form the amino acid 4 (see Supporting
Information). Marfey’s derivatization again matched syn-
thetic material data with those from the natural product by
HPLCand LCMS, confirming the (2S,3S) stereochemistry
for 4.
After determining the absolute configuration of each
amino acid2ꢀ4, we focused onsynthesizing the fatty acid5
(Scheme 3). Having no isolated natural derivative of 5 to
hand, a total synthesis of diastereomeric 1 with both
enantiomers was undertaken. To this end, the Evans-based
oxazolidinones 21a/b were prepared18 and convertedtothe
thiotetrazoles 22 after reduction and Mitsunobu reaction.
Oxidation to the sulfone 23 and KocienskiꢀJulia olefina-
tion to 24 gave high selectivity (E/Z = 92:8, by GCMS)
with the use of KHMDS.19 Deprotection of TBDPS
followed by Jones oxidation then gave the fatty acids 25a
(5S) and 25b (5R).
With all synthetic fragments available, the Boc-N-
protected amino ester 10 was hydrolyzed with LiOH in
THF/MeOH to give the acid 26, which was coupled with the
(S)-aspartate methyl ester 27 under HATU conditions to
produce the dipeptide 28 (Scheme 4). After Boc deprotec-
tion, the TFA salt from 28 was coupled with the azido
methyl ester 20 to afford the azido tripeptide 29. Although
other phospines failed, the azide functionality was success-
fully reduced with Me3P20 and the amine formed coupled
with either lipid 25a or 25b. This provided the trimethyl
ester stereoisomers 30 and 31 of the targeted lipopeptide
with [R]25D values of þ18.8 (c = 0.25) and ꢀ23.1 (c = 0.22),
respectively.
Although MeOD revealed minor NMR differences
between (þ)-30 and (ꢀ)-31, comparison in CDCl3 revealed
significant differences in the C-2 and C-3 positions
(proximal to the C5 chiral center). The NMR data and
optical rotation of (þ)-30 correlated closely to the tri-
methylated natural product, [R]25 = þ31.6 (c = 0.18),
D
as prepared by reacting (ꢀ)-1 with TMSCHN2.21 This
indicated a C-5 S configuration for the fatty acid 5. Further-
more, data from hydrolysis of (þ)-30 to its triacid matched
the NMR and optical rotation data of the natural product:
synthetic (ꢀ)-1 [R]25 = ꢀ20.0 (c = 0.1), natural (ꢀ)-1
D
[R]25D = ꢀ23.7 (c = 0.3) (see Supporting Information).
In summary, we applied the parallel use of structure
elucidation methods with total synthesis to establish the
absolute stereochemistry of each amino acid fragment 3, 4,
(17) Bal, B. S.; Childers, W. E., Jr; Pinnick, H. W. Tetrahedron 1981,
37, 2091–2096.
(18) Garcia-Ruiz, V.; Woodward, S. Tetrahedron: Asymmetry 2002,
13, 2177–2180.
(19) Blakemore, P. R.; Cole, W. J.; Kocienski, P. J.; Morley, A.
Synlett 1998, 26–28.
(20) Nicolaou, K. C.; Nevalainen, M.; Zak, M.; Bulat, S.; Bella, M.;
(21) Francais, A.; Leyva, A.; Etxebarria-Jardi, G.; Ley, S. V. Org.
Lett. 2010, 12, 340–343.
Safina, B. S. Angew. Chem., Int. Ed. 2003, 42, 3418–3424.
1562
Org. Lett., Vol. 14, No. 6, 2012

Scheme 4. Total Synthesis of Lipopeptide (ꢀ)-1 by Peptide Coupling and Lipid Conjugation
and 5. Assignment of the final, lipid component 5 was then
confirmed via total synthesis of 1. Noteworthy steps
include the following: (1) the L-proline organocatalytic
formation of a syn-Mannich product (10) in 99% ee; (2)
the stereospecific formation of a cis-olefin (15) by cuprate
conjugate addition onto an alkynoate; (3) the highly
selective KocienskiꢀBlakemore Julia formation of a trans
R-methylated olefin (24); (4) the Sharpless-based synthesis
of an amino acid precursor (20) bearing a quaternary
center and an azido group for a reductive-lipidation
sequence to the natural product (ꢀ)-1.
and A. niger, show (ꢀ)-1 to be inactive (<50 μM). Mode of
action studies of this rather compact, novel lipoprotein lead
structure are thus anticipated to provide interesting insights
and differences in microbial and mammalian cell biology.3
Acknowledgment. We thank the National University of
Singapore for a Graduate Scholarship (to S.K.G.) and
start-up funding (to M.J.L.). This work was funded latterly
by the National Research Foundation, Competitive Re-
search Program (NRF-G-CRP 2007-04). We are grateful
to G. Subramanium (MerLion Pharmaceuticals) for the
initial isolation of (ꢀ)-1.
Interestingly, it is likely that both the carboxylic acid
motifs and the olefins of 1 play a key role for bioactivity.
For example, the trimethyl ester derivative (þ)-30 and
tetrahydrogenated version of the lipopeptide are devoid
of antimalarial activity. In contrast, (ꢀ)-1 exhibits signifi-
cant in vitro activity against P. falciparum (IC50 = 0.8 μM
against 3D7 strain). Moreover, primary whole cell screenings
against HeLa, B. subtilis, S. aureus, C. albicans, S. cerervisae,
Supporting Information Available. Experimental pro-
cedures, compound characterization, NMR spectra, and
LCMS. This material is available free of charge via the
Internet at http://pubs.acs.org.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 6, 2012
1563