Total synthesis and assignment of the side chain stereochemistry of LI-F04a: An antimicrobial cyclic depsipeptide

Article abstract of DOI:10.1021/ol101254m

(Equation Presented). The total synthesis of the potent antifungal and antibiotic cyclic depsipeptide LI-F04a and its side chain epimer was accomplished using macrolactonization to assemble the cyclic peptide core, followed by attachment of the 15-guanidino-3-hydroxypentadecanoyl (GHPD) side chain. The side chain was assembled by Yamaguchi-Hirao alkylation of both enantiomers of a chiral epoxide to provide a pair of enantiomeric side chains. The attachment of both these chains to the cyclic peptide allowed the absolute configuration of the side chain hydroxyl group in LI-F04a to be assigned as (R).

Full text of DOI:10.1021/ol101254m

ORGANIC
LETTERS
2010
Vol. 12, No. 15
3394-3397
Total Synthesis and Assignment of the
Side Chain Stereochemistry of LI-F04a:
An Antimicrobial Cyclic Depsipeptide
James R. Cochrane, Christopher S. P. McErlean, and Katrina A. Jolliffe*
School of Chemistry, The UniVersity of Sydney, 2006, NSW, Australia
kate.jolliffe@sydney.edu.au
Received May 31, 2010
ABSTRACT
The total synthesis of the potent antifungal and antibiotic cyclic depsipeptide LI-F04a and its side chain epimer was accomplished using
macrolactonization to assemble the cyclic peptide core, followed by attachment of the 15-guanidino-3-hydroxypentadecanoyl (GHPD) side
chain. The side chain was assembled by Yamaguchi-Hirao alkylation of both enantiomers of a chiral epoxide to provide a pair of enantiomeric
side chains. The attachment of both these chains to the cyclic peptide allowed the absolute configuration of the side chain hydroxyl group
in LI-F04a to be assigned as (R).
The increasing incidence of invasive fungal mycoses and the
development of problematic multidrug resistant bacterial infec-
tions, such as hospital associated methicillin resistant Staphy-
lococcus aureus (HA-MRSA), are emerging as serious threats
to public health, and there is an increasingly urgent need for
the discovery of new antimicrobial agents active against these
pathogens.1-7 LI-F04a is a cyclic depsipeptide produced as
the major component of a family of closely related cyclic
depsipeptides (collectively labeled as the LI-Fs) isolated from
the L-1129 strain of Paenibacillius polymyxa (formerly
Bacillus polymyxa).8,9 These compounds are analogous to
the fuscaricidins which were originally isolated from the
KT-8 strain of Paenibacillus polymyxa.10-14 The LI-F
peptides have been found to exhibit antifungal activity against
a wide range of fungi including the clinically relevant species
Candida albicans and Cryptoccoccus neoformans.8-14 They
also exhibit activity against Gram-positive bacteria such as
strains of Staphylococcus aureus and Micrococcus luteus8,11
and have been shown to have low acute toxicity in mice.14
The LI-F peptides contain a cyclic depsihexapeptide core with
a unique 15-guanidino-3-hydroxypentadecanoyl (GHPD) side
chain attached to the nitrogen atom of an L-threonine residue
(Scheme 1). Three amino acids, L-Thr, D-allo-Thr, and D-Ala,
(1) Brickner, S. J.; Mobashery, S. Curr. Opin. Microbiol. 2007, 10, 425
(2) Lipsitch, M. Trends Microbiol. 2004, 8, 515
(3) Barrett, J. F. Expert Opin. Ther. Targets 2005, 9, 253
(4) Lamagni, T. L.; Evans, B. G.; Shigematsu, M.; Johnson, E. M.
Epidemiol. Infect. 2001, 126, 397
(5) Pfaller, M. A. Clin. Infect. Dis. 1998, 27, 11481150.13
(6) Marr, K. A.; Carter, R. A.; Crippa, F.; Wald, A.; Corey, L. Clin.
Infect. Dis. 2002, 34, 909
(7) Norby, S. R.; Nord, C. E.; Finch, R. Lancet Infect. Dis. 2005, 5, 115
(8) Kaneda, M.; Kajimura, Y. Yakugaku Zasshi 2002, 122, 651
.
.
(9) Kuroda, J.; Fukai, T.; Konishi, M.; Uno, J.; Kurusu, K.; Nomura, T.
.
Heterocycles 2000, 53, 1533
(10) Choi, S. K.; Park, S. Y.; Kim, R.; Lee, C. H.; Kim, J. F.; Park,
S. H. Biochem. Biophys. Res. Commun. 2008, 365, 89
.
.
.
.
(11) Kajimura, Y.; Kaneda, M. J. Antibiot. 1996, 49, 129
(12) Kajimura, Y.; Kaneda, M. J. Antibiot. 1997, 50, 220
.
.
.
(13) Kajimura, Y.; Sugiyama, M.; Kaneda, M. J. Antibiot. 1995, 48, 1095
.
.
(14) Kurusu, K.; Ohba, K.; Arai, T.; Fukushima, K. J. Antibiot. 1987,
40, 1506
.
.
10.1021/ol101254m  2010 American Chemical Society
Published on Web 07/07/2010

that the chiral alcohol functionality at C3 of 3 could be
introduced by a Yamaguchi-Hirao alkylation16 of the known
chiral epoxides (S)-4 and (R)-4 with alkyne 5, thereby
providing both side chain enantiomers for attachment to the
cyclic peptide core 2.17,18 While there are a number of
possible macrocyclization sites possible for the synthesis of
2, we anticipated that Yamaguchi macrolactonization19,20 of
linear peptide 6 would yield the required cyclic depsipeptide
and enable rapid access to libraries of LI-F04a analogues
(including the other peptides in the LI-F family) for future
biological studies, by standard Fmoc solid phase peptide
synthesis of the linear precursors.
Scheme 1. Retrosynthesis
Synthesis of the (R)-enantiomer of the GHPD side chain,
(R)-3, began with the enantiomerically enriched epoxide
(S)-4 which was obtained in 32% yield (>99% ee) by
hydrolytic kinetic resolution (HKR) of (()-4 in the
presence of Jacobsen’s (S,S)-(salen)Co(III) catalyst.17
Nucleophilic opening of epoxide (S)-4 with the lithio-
acetylide of alkyne 521 via a BF3·OEt2-promoted alkyla-
tion16 at -78 °C gave the alcohol (R)-7 in 69% yield
(Scheme 2). Protection of the secondary alcohol as a
methoxy methyl ether, followed by desilylation (TBAF),
gave the primary alcohol (R)-8 in 77% yield over 2 steps.
Reaction of (R)-8 with di(tert-butoxycarbonyl)guanidine
under Mitsunobu conditions proceeded smoothly to give
(R)-9 in 90% yield.22 After optimization of both the
catalyst and solvent, debenzylation and concomitant
reduction of the internal alkyne were achieved upon
treatment with H2 in the presence of Pd(OH)2/C to give
(R)-10 in 79% yield. The use of the basic catalyst and
mild conditions was necessary as the guanidine Boc-
protecting groups were very acid labile and prone to
cleavage under more forcing conditions. Finally, alcohol
(R)-10 was subjected to a ruthenium tetroxide catalyzed
oxidation23-25 to give the protected GHPD fragment (R)-3
in 63% yield. The synthesis of this fragment was thus
achieved in 6 steps and 23% overall yield from (S)-4. The
(S)-GHPD enantiomer (S)-3 was prepared in an identical
manner starting from the chiral epoxide (R)-4, which was
obtained in 40% yield (>99% ee) upon HKR of (()-4 in the
presence of Jacobsen’s (R,R)-(salen)Co(III) catalyst (Scheme
3).
are conserved throughout the LI-F series, while there are slight
variations in the other three amino acids present. In LI-F04a,
these are D-Asn, L-Val, and D-Val. The GHPD side chain is
conserved among the LI-F series of antifungal cyclic peptides,
but to the best of our knowledge, the absolute stereochemistry
of the 3-hydroxyl group has not previously been confirmed.9,11
This, together with the biological activity and limited
availability of isolated LI-F04a, makes this compound an
attractive target for total synthesis. The ability to synthesize
individual members of the LI-F peptide family will allow
the structural basis for the biological activity of these cyclic
peptides to be determined in the future.
With both enantiomers of the GHPD side chain in hand,
our attention turned to the synthesis of the cyclic peptide
core of LI-F04a. Thus, linear peptide precursor 6 was
The synthesis of a simplified analogue of LI-F04a, in
which the side chain 3-hydroxy group was omitted, has
previously been reported.15 We report here a total synthesis
of both LI-F04a (1) and its GHPD side chain epimer. The
synthesis of both compounds allowed the unambiguous
assignment of the absolute stereochemistry of the alcohol in
the natural product as the (R)-isomer.
Our synthetic strategy was based upon the retrosynthetic
analysis presented in Scheme 1. The late-stage coupling of
the cyclic peptide 2 with the GHPD side chain 3 would allow
ready access to both side chain epimers of 1. It was envisaged
(16) Yamaguchi, M.; Hirao, I. Tetrahedron Lett. 1983, 24, 391.
(17) Schaus, S. E.; Brandes, B. D.; Larrow, J. F.; Tokunaga, M.; Hansen,
K. B.; Gould, A. E.; Furrow, M. E.; Jacobsen, E. N. J. Am. Chem. Soc.
2002, 124, 1307.
(18) Liang, Q.; Sun, Y.; Yu, B.; She, X.; Pan, X. J. Org. Chem. 2007,
72, 9846.
(19) Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M. Bull.
Chem. Soc. Jpn. 1979, 52, 1989.
(20) Sarabia, F.; Chammaa, S.; Lopez-Herrera, F. J. Tetrahedron Lett.
2002, 43, 2961.
(21) Carpita, A.; Mannocci, L.; Rossi, R. Eur. J. Org. Chem. 2005, 1859.
(22) Dodd, D. S.; Kozikowski, A. P. Tetrahedron Lett. 1994, 35, 977.
(23) Djerassi, C.; Engle, R. R. J. Am. Chem. Soc. 1953, 75, 3838.
(24) Carlsen, P. H. J.; Katsuki, T.; Martin, V. S.; Sharpless, K. B. J.
Org. Chem. 1981, 46, 3936.
(25) Joo, J. E.; Pham, V. T.; Tian, Y. S.; Chung, Y. S.; Oh, C. Y.; Lee,
K. Y.; Ham, W. H. Org. Biomol. Chem. 2008, 6, 1498.
(15) Stawikowski, M.; Cudic, P. Tetrahedron Lett. 2006, 47, 8587.
Org. Lett., Vol. 12, No. 15, 2010
3395

hexafluoroisopropanol,26 leaving the N-terminal and side
chain protecting groups intact (Scheme 4).
Scheme 2. Synthesis of (R)-3
Scheme 4. Peptide Synthesis
The key macrolactonization step was then attempted under
standard Yamaguchi macrolactonization conditions.19,20 Initial
experiments provided cyclized peptide in good yield (>60%).
However, analysis of this material by HPLC indicated that
it was a 3:1 mixture of diastereoisomers, presumably as a
result of epimerization of the C-terminal D-Ala residue.
Epimerization of the C-terminal amino acid prior to mac-
rocyclization is frequently observed in the synthesis of small
cyclic peptides.27,28 Fortunately, use of the modified Yone-
mitsu conditions,29 which involved slow addition of the acid
to a solution of DMAP, 2,4,6-trichlorobenzoyl chloride, and
triethylamine in toluene at room temperature, provided the
required cyclic peptide 11 in 58% yield, without significant
epimerization (<5% epimer observed by HPLC) (Scheme 5).
Hydrogenolysis of the Cbz protecting group on 11 was
unexpectedly problematic. Numerous conditions were em-
ployed; however, in all cases the reactions were sluggish,
and mixtures of starting material and over-reduced side
products were frequently obtained, despite reports that trityl
protecting groups on amides are stable to hydrogenolysis.30
prepared by standard Fmoc solid phase peptide protocols
using PyBOP/Hu¨nig’s base as the activation reagent and the
2-chlorotritylchloride resin as the solid support. Cbz-L-Thr-
OH was added as the N-terminal amino acid, and then the
peptide was cleaved from the resin upon treatment with
Scheme 3. Synthesis of (S)-3
(26) Bollhagen, R.; Schiedberger, M.; Barlos, K.; Grell, E. J. Chem.
Soc., Chem. Commun. 1994, 2559.
(27) Kopple, K. D. J. Pharm. Sci. 1972, 61, 1345
(28) Ehrlich, A.; Heyne, H.-U.; Winter, R.; Beyermann, M.; Haber, H.;
Carpino, L. A.; Bienert, M. J. Org. Chem. 1996, 61, 8831
.
.
(29) Hikota, M.; Sakurai, Y.; Horita, K.; Yonemitsu, O. Tetrahedron
Lett. 1990, 31, 6367.
´
(30) Isidro-Llobet, A.; Alvarez, M.; Albericio, F. Chem. ReV. 2009, 109,
2455.
3396
Org. Lett., Vol. 12, No. 15, 2010

Scheme 5. Total Synthesis of LI-F04a and epi-LI-F04a
The optimal conditions were found to be H2, Pd/C in THF,
providing the free amine 2 in 58% yield.
showed a closer correlation with those of the natural product
(maximum ∆δ of 0.02 ppm) than that of (S)-1 (maximum ∆δ of
0.14 ppm). Similarly, while the 13C NMR spectra of (R)-1 and
(S)-1 were similar, a comparison of chemical shifts for all signals
indicated that the spectrum of (R)-1 matched that of the natural
product more closely than that of (S)-1. Taken together, the
excellent correlation of the 1H and 13C spectra of (R)-1 with those
of the natural product and the similar optical rotations of the natural
material and synthetic (R)-1 enabled us to assign the stereogenic
center at C3 of the side chain of naturally occurring LI-F04a as
(R)-configured.
In summary, we have completed a total synthesis of LI-
F04a as well as its side chain epimer. This has allowed the
complete stereochemical assignment of the natural product.
The synthetic route we have developed is general, is efficient,
and will enable the rapid synthesis of other members of the
LI-F family. It will also be applicable to the synthesis of the
related circulocin family of cyclic peptides, which have a
similar cyclic peptide core bearing ω-guanidino-3-hydroxy
side chains of varying length.31
The final key step in the synthesis of LI-F04a involved coupling
of the cyclic peptide 12 to both enantiomers of the GHPD side
chain 3. This was achieved using HATU/Hu¨nig’s base as the
coupling reagent. The coupled products were immediately sub-
jected to global deprotection upon treatment with TFA/CH2Cl2/
H2O (90/5/5 v/v/v), providing both side chain epimers of LI-F04a
in unoptimized yields of 33-40% over 2 steps.
With both compounds available with known stereochemistry
at the side chain hydroxy group, we were able to compare their
spectroscopic and physical properties with those of natural LI-
F04a to establish the absolute configuration of the side chain
hydroxy group. First, comparison of the optical rotations of the
synthetic material with the (R)-configured side chain (+20°)
and that of the (S)-configured side chain (-10°) with those
reported for the natural product (+24°,9 +12.8°11) suggested
that the natural product may bear a side chain with (R)-
stereochemistry at C3. Second, the 1H and 13C spectra of both
synthetic epimers of 1 were fully assigned using a combination
of 1- and 2-D NMR experiments and compared with the data
reported for natural LI-F04a. While both side chain epimers
had similar 1H and 13C NMR spectra, there were some subtle
differences. Notably, both the chemical shift and coupling
constant for one of the methylene protons on the carbon (C2)
adjacent to the side chain hydroxy substituent differ significantly
in the two epimers. In (S)-1, the signal attributable to this proton
appears at 2.27 ppm and couples to the proton on C3 with a
coupling constant of 3.7 Hz, while the signal for the same proton
in (R)-1 is observed at 2.36 ppm with a coupling constant of
6.6 Hz, which is identical to the data reported for the natural
product. A comparison of chemical shifts for all protons (excluding
the NH and OH signals) indicated that the spectrum of (R)-1
Acknowledgment. We thank Dr. Ian Luck (USyd) and Dr.
Kelvin Picker (USyd) for assistance with NMR and HPLC,
respectively. We thank the University of Sydney for financial
support and the award of a postgraduate scholarship to JRC.
Supporting Information Available: Experimental proce-
dures and spectroscopic data for all compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
OL101254M
(31) He, H. Y.; Shen, B.; Korshalla, J.; Carter, G. T. Tetrahedron 2001,
57, 1189.
Org. Lett., Vol. 12, No. 15, 2010
3397