Asymmetric total synthesis and stereochemical revision of gymnangiamide

Article abstract of DOI:10.1021/ol900184d

The asymmetric total synthesis of the originally proposed structure of gymnangiamide, a cytotoxic pentapeptide isolated from the marine hydroid Gymnangium regae Jaderholm, has been achieved. Key to the synthesis was the use of asymmetric hydrogenation of

Full text of DOI:10.1021/ol900184d

ORGANIC
LETTERS
2009
Vol. 11, No. 9
1995-1997
Asymmetric Total Synthesis and
Stereochemical Revision of
Gymnangiamide
Hitoshi Tone,† Marie Buchotte,† Ce´line Mordant,† Eric Guittet,‡ Tahar Ayad,† and
Virginie Ratovelomanana-Vidal*,†
Laboratoire Charles Friedel, ENSCP Chimie ParisTech, UMR 7223, 11 rue P. et M. Curie,
F-75231 Paris Cedex 05, France, and Laboratoire de Chimie et Biologie Structurales, Institut
de Chimie des Substances Naturelles, UPR 2301 91190 Gif-sur-YVette, France
Virginie-Vidal@enscp.fr
Received January 28, 2009
ABSTRACT
The asymmetric total synthesis of the originally proposed structure of gymnangiamide, a cytotoxic pentapeptide isolated from the marine
hydroid Gymnangium regae Jaderholm, has been achieved. Key to the synthesis was the use of asymmetric hydrogenation of r-substituted
ꢀ-ketoesters through dynamic kinetic resolution for the preparation of nonproteinogenic chiral amino acids. The disparity of the NMR spectra
between the synthetic material containing the L-serine residue and the natural product required a revision of the proposed structure.
Gymnangiamide 1 is a cytotoxic pentapeptide isolated by
Gustafson and co-workers1 in 2004 from the marine hydroid
Gymnangium regae Jaderholm, collected in the Philippines.
On the basis of spectroscopic investigations, chemical
degradation, and derivatization studies, structure 1 depicted
in Figure 1 was assigned as gymnangiamide. In addition to
the amino acids (2S,3S)-isoleucine (Ile) and (2S,3R)-phe-
nylserine (Pser), this pentapeptide contains three nonstandard
amino acid residues, N-desmethyl-dolaisoleuine (Ddil), O-
desmethyl-dolaproine (Ddap), and a rare R-guanidino-L-
serine (Gser), which has not previously been reported in a
natural product. Furthermore, and to the best of our
Figure 1. Originally proposed (1) and revised structure (2-TFA)
of gymnangiamide.
knowledge, there is no report on the total synthesis of
gymnangiamide.
† ENSCP.
In this paper, we describe the asymmetric total synthesis
of gymnangiamide and in turn clarify its configuration since
we also report that structure 1 (34S) does not correspond to
‡ Institut de Chimie des Substances Naturelles.
(1) Milanowski, D. J.; Gustafson, K. R.; Rashid, M. A.; Pannell, L. K.;
McMahon, J. B.; Boyd, M. R. J. Org. Chem. 2004, 69, 3036.
10.1021/ol900184d CCC: $40.75
Published on Web 04/08/2009
 2009 American Chemical Society

The diastereochemically pure unusual O-desmethyl-dola-
proine (Ddap)4 residue 8 was prepared on multigram scale
from N-Boc-(S)-proline through DKR using ruthenium-
catalyzed asymmetric hydrogenation as a key step (Scheme
2).
Scheme 1. Synthesis of the Pser Fragment 7
The synthesis of the N-desmethyl-dolaisoleuine (Ddil)
residue 10 was readily obtained from N-Boc-L-isoleucine as
a single diastereoisomer in 44% overall yield through a four-
step sequence involving Masamune reaction,5 ketone reduc-
tion, selective O-methylation,6 and hydrogenolysis.
Having established a scalable access to the three amino
acids derivatives Pser 7, Ddap 8, and Ddil 10, the stage
was now set for their assembly and elaboration into
gymnangiamide 1 using iterative DEPC-mediated cou-
plings. The (2S,3R)-phenylserine (Pser) fragment 7 was
coupled with the Ddap fragment 8 to provide the dipeptide
9 in 94% yield. Cleavage of the N-terminal Boc protective
group of 9 followed by coupling with Ddil fragment 10
gave the desired tripeptide 11 in 89% isolated yield. The
peptide chain was further elongated by coupling 11 with
the dipeptide N-Boc-L-Ser-Ile-OH in the presence of DCC-
HOBt as condensating agent to give the linear pentapeptide
in 85% yield (Scheme 2). Finally, treatment of the latter
under mild acidic conditions followed by guanidinylation7
of the resulting free amine and subsequent hydrogenolysis
of both the Cbz and benzyl ester protective groups
afforded fully synthetic free pentapeptide 1 in stereopure
form and good yield after recrystallization from MeOH.
Surprisingly, the spectroscopic and physical properties of
our synthetic sample8 were distinctively different from
those of the natural gymnangiamide 1.
natural gymnangiamide. We give synthetic evidence that
compound 2-TFA (34R), incorporating a D-serine instead
of L-serine amino acid residue, corresponds to the natural
product. The approach we have chosen for assembling the
five units of the peptide involves the construction of the
subunits (2S,3R)-phenylserine (Pser) and O-desmethyl-dola-
proine (Ddap) in diastereochemically pure form using
ruthenium-promoted asymmetric hydrogenation through dy-
namic kinetic resolution (DKR)2 and the coupling with other
fragments.
Initially, we synthesized the putative gymnangiamide 1
(Figure 1). The synthesis started with the valuable (2S,3R)-
phenylserine (Pser) fragment 7, which was prepared through
DKR of racemic 3 using 1 mol % of [(RuCl((R)-SYN-
PHOS)2(µ-Cl)3][NH2Me2]3 catalyst, at 60 °C in dichlo-
romethane under 120 bar. Under these optimized reaction
conditions, the two chiral centers of Pser were established
in one step, providing after recrystallization from ethyl
acetate the optically pure compound (2S,3R)-4 in 85%
isolated yield. Treatment of 4 in refluxing 6 N HCl gave the
(2S,3R)-phenylserine HCl salt 5, which was converted into
the N-Boc derivative 6 followed by esterification and
deprotection of the N-Boc protective group providing the
(2S,3R)-phenylserine (Pser) fragment 7 in 68% overall yield
(Scheme 1).
Careful examination of the original isolation procedure
revealed that the reported data were most likely the data
of the corresponding trifluoroacetic salt.9 Thus, pentapep-
tide 1 was converted into its corresponding TFA salt
1
1-TFA (Scheme 2). However, the H and 13C NMR still
unexpectedly showed some different chemical shifts from
those reported in the literature for the natural product,
although the patterns of the peaks were quite similar.10
We then directed our efforts toward ascertaining the nature
(4) (a) Mordant, C.; Ratovelomanana-Vidal, V.; Genet, J.-P. Org. Lett.
2001, 12, 1909. (b) Mordant, C.; Reymond, S.; Ratovelomanana-Vidal, V.;
Genet, J.-P. Tetrahedron 2004, 60, 9715.
(5) Brooks, D.; Lu, L. D. L.; Masamune, S. Angew. Chem., Int. Ed.
Engl. 1979, 18, 72.
(2) For reviews, see: (a) Noyori, R.; Tokunaga, M.; Kitamura, M. Bull.
Chem. Soc. Jpn. 1995, 68, 36. (b) Ratovelomanana-Vidal, V.; Genet, J.-P.
Can. J. Chem. 2000, 78, 846. (c) Pellissier, H. Tetrahedron 2008, 64, 1563.
(d) Hamada, Y.; Makino, K. J. Synth. Org. Chem. 2008, 66, 1057. For
some applications in organic synthesis, see: (e) Scalone, M.; Waldmeier,
P. Org. Process Res. DeV. 2003, 7, 418. (f) Eustache, F.; Dalko, P. I.; Cossy,
J. J. Org. Chem. 2003, 68, 9994. (g) Mordant, C.; Du¨nkelman, P.;
Ratovelomanana-Vidal, V.; Genet, J.-P. Chem. Commun 2004, 1296. (h)
Labeeuw, O.; Phansavath, P.; Genet, J.-P. Tetrahedron: Asymmetry 2004,
15, 1899. (i) Makino, K.; Goto, T.; Hiroki, Y.; Hamada, Y. Angew. Chem.,
Int. Ed 2004, 43, 882. (j) Ito, M.; Kitahara, S.; Ikariya, T. J. Am. Chem.
Soc. 2005, 127, 6172. (k) Mordant, C.; Reymond, S.; Tone, H.; Lavergne,
D.; Touati, R.; Ben Hassine, B.; Ratovelomanana-Vidal, V.; Genet, J.-P.
Tetrahedron 2007, 63, 6115.
(6) (a) Finch, N.; Fitt, J. J.; Hsu, I. H. S. J. Org. Chem. 1975, 40, 206.
(b) Greene, A. E.; Le Drian, C.; Crabbe, P. J. Am. Chem. Soc. 1980, 102,
7583.
(7) (a) Feichtinger, K.; Zapf, C.; Sings, H. L.; Goodman, M. J. Org.
Chem. 1998, 63, 3804. (b) Feichtinger, K.; Sings, H. L.; Baker, T. J.;
Matthews, K.; Goodman, M. J. Org. Chem. 1998, 63, 8432.
(8) The optical rotation of synthetic 1, [R]27 -18.5 (c 0.22, MeOH),
D
did not match the value [R]27D -32.5 (c 0.24, MeOH) of the natural product.
See Supporting Informationfor NMR data.
(9) The natural product was isolated by preparative HPLC using CH3CN/
H2O with 0.1% TFA as eluent. Generally, peptides tend to retain variable
amounts of water and acids (as counterions or residual acid) resulting from
the final stage of purification or isolation. Sewald, N., Jakubbe, H.-D., Eds.
Peptides: Chemistry and Biology; Wiley-VCH: Weinheim, Germany, 2002;
Chapter 2, pp 5-59.
(3) (a) Jeulin, S.; Champion, N.; Dellis, P.; Ratovelomanana-Vidal, V.;
Genet, J.-P. Synthesis 2005, 3666. (b) Duprat de Paule, S.; Jeulin, S.;
Ratovelomanana-Vidal, V.; Genet, J.-P.; Champion, N.; Dellis, P. Eur. J.
Org. Chem. 2003, 1931. (c) Jeulin, S.; Duprat de Paule, S.; Ratovelomanana-
Vidal, V.; Genet, J.-P.; Champion, N.; Dellis, P. Proc. Natl. Acad. Sci.
U.S.A. 2004, 101, 5799.
(10) The NH region was especially demonstrative of the solved
discrepancies between the two isomers, with a strict superimposition between
the published 1H spectrum of the natural product and that of synthetic
2-TFA. [R]27 -32.5 (c 0.16, MeOH), which matched the value [R]27
D
D
-32.5 (c 0.24, MeOH) of the natural product.
1996
Org. Lett., Vol. 11, No. 9, 2009

Scheme 2. Completion of the Synthesis of the Originally Proposed Structure 1 for Gymnangiamide
of the discrepancy. After careful reinspection of the
original published procedure used by Gustafson et al.1 for
the assignment of the absolute stereochemistry of 1, we
suspected that the stereochemical assignment of the
R-guanidino serine residue might be incorrect and we
further favored a revision to the TFA salt of 1 from a
L-serine to D-serine. Working as described for the synthesis
of 1-TFA but coupling the tripeptide 11 with N-Boc-D-
Ser-Ile-OH and following the sequence depicted in Scheme
3, we synthesized 2-TFA, which matches all available
spectroscopic data of the natural gymnangiamide. Thus,
we conclude that the R-guanidino serine residue in the
natural gymnangiamide has the R configuration.
Scheme 3. Synthesis of the Revised Structure 2-TFA for
Gymnangiamide
In conclusion, the asymmetric total synthesis of gym-
nangiamide (1) has been accomplished through successful
Ru-SYNPHOS asymmetric hydrogenation and has implied
the stereochemical reassignment of the natural product
from 1 to 2-TFA. In addition to the structural revision,
the described chemistry renders the natural substance
readily available and opens the way to design analogues
in amounts suitable for a thorough investigation of their
biological properties.
Acknowledgment. Thanks are due to the CNRS/DGA and
Ecole Polytechnique for financial support.
Supporting Information Available: Experimental pro-
cedures and full analyses of products 1, 1-TFA, and 2-TFA.
This material is available free of charge via the Internet at
http://pubs.acs.org.
OL900184D
Org. Lett., Vol. 11, No. 9, 2009
1997