Total synthesis of halicylindramide A

Article abstract of DOI:10.1021/jo802213q

A rapid and efficient Fmoc solid-phase synthesis of halicy-lindramide A is described. The strategy comprises resin attachment of the first amino acid via the side chain of aspartic acid, stepwise solid-phase synthesis of the linear peptide analog up to th

Full text of DOI:10.1021/jo802213q

Total Synthesis of Halicylindramide A
Hyunju Seo and Dongyeol Lim*
Department of Chemistry, Sejong UniVersity,
Seoul 143-747, South Korea
dylim@sejong.ac.kr
ReceiVed October 4, 2008
FIGURE 1. Structure of halicylindramide.
These cyclic depsipeptides share many unusual amino acids,
such as D-amino acids, N-methyl amino acids, and cysteic acid,
and contain a formyl group at the N-terminus. These peptides
possess various antimicrobial, anticancer, and anti HIV-1
activities and interact with proteins by exhibiting inhibitory
activity toward phospholipase A2 or agonistic activity at a G
protein coupled receptor (GPCR). To date, none of these
depsipeptides have been synthesized, nor have their conforma-
tions been elucidated. Because these peptides are composed of
six hydrophobic amino acid residues at the N-terminus and two
hydrophilic amino acids in the middle of the sequence, their
activities might be related to their association with membranes
due to their amphiphilic character. The plasma membrane
permeabilization of Ca2+ and ATP by discodermin A was
reported in 2001,8 but no other mechanism or structural
information was obtained. Therefore, a total synthesis of these
products would be valuable to explore their potential therapeutic
applications and to understand the mechanism of the membrane
interactions of these novel peptides. Here, we describe an
efficient solid-phase synthesis of halicylindramide A.
A rapid and efficient Fmoc solid-phase synthesis of halicy-
lindramide A is described. The strategy comprises resin
attachment of the first amino acid via the side chain of
aspartic acid, stepwise solid-phase synthesis of the linear
peptide analog up to the cysteine residue, on-resin head-to-
tail cyclization and linear peptide synthesis of the N-terminal
region, and finally cysteine oxidation and formylation. The
stereochemistry of the halicylindramide A was confirmed by
comparison of NMR and RP-HPLC data with the natural
molecule. A distinctive conformational change was observed
from the CD spectra of the halicylindramide A in sodium
dodecyl sulfate.
In the retrosynthetic analysis of cyclic depsipeptides, the sites
of resin attachment and ring disconnection are crucial for
successful synthesis. The selection of the point of cyclization
is critical because cyclization often proceeds very slowly,
causing side reactions such as dimerization and/or epimerization
of the C-terminal residue. In our synthesis, we decided to anchor
the side chain of the Asp residue to Rink amide resin and to
cyclize the precursor via macrolactamization between the Sar
and Asn residues on the resin. In this way, cyclization and
coupling of the N-terminal amino acids could be performed
while the peptide remained anchored to the resin.
The synthesis of halicylindramide A begins with the attach-
ment of C-terminal Fmoc-D-Asp-OAllyl via the side chain to
the Rink amide resin using a PyBOP/HOBt/DIPEA procedure
(Scheme 1). Standard Fmoc chemistry was used throughout.
To introduce the N-methylglutamine residue into the ring, Fmoc-
NR-methylglutamine was prepared from commercially available
Fmoc-Glu(Trt)-OH according to the published procedure, in
86% yield.9 To couple the D-Phe and (N-Me)Gln residues, 1 h
In 1995, Fusetani and co-workers reported the isolation and
structure of halicylindramides A-C (Figure 1), which are cyclic
depsipeptides isolated from the Japanese marine sponge Hali-
chondria cylindrata.1 A subsequent paper described the struc-
tures of halicylindramide D, a tridecapeptide lactone, and
halicylindramide E, a truncated linear peptide.2 Other sponge
peptides, including discodermins A-H,3 polydiscamide A,4 and
microspinosamide,5 constitute a family of structurally related
trideca- and tetradecapeptides. Recently, corticiamide A6 and
polydiscamides B-D7 from sponges have also been reported.
(1) Li, H.-y.; Matsunaga, S.; Fusetani, N. J. Med. Chem. 1995, 38, 338.
(2) Li, H.-y.; Matsunaga, S.; Fusetani, N. J. Nat. Prod 1996, 59, 163.
(3) (a) Matsunaga, S.; Fusetani, N.; Konosu, S. Tetrahedron Lett. 1984, 25,
5165. (b) Matsunaga, S.; Fusetani, N.; Konosu, S. J. Nat. Prod. 1985, 48, 236.
(c) Matsunaga, S.; Fusetani, N.; Konosu, S. Tetrahedron Lett. 1985, 26, 855.
(d) Ryu, G.; Matsunaga, S.; Fusetani, N. Tetrahedron Lett. 1994, 35, 8251. (e)
Ryu, G.; Matsunaga, S.; Fusetani, N. Tetrahedron 1994, 50, 13409.
(4) Gulavita, N. K.; Gunasekera, S. P.; Pomponi, S. A.; Robinson, E. V. J.
Org. Chem. 1992, 57, 1767.
(5) Rashid, M. A.; Gustafson, K. R.; Cartner, L. K.; Shigematsu, N.; Pannell,
L. K.; Boyd, M. R. J. Nat. Prod. 2001, 64, 117.
(6) Laird, D. W.; LaBarbera, D. V.; Feng, X.; Bugni, T. S.; Harper, M. K.;
Ireland, C. M. J. Nat. Prod. 2007, 70, 741.
(7) Feng, Y.; Carroll, A. R.; Pass, D. M.; Archbold, J. K.; Avery, V. M.;
Quinn, R. J. J. Nat. Prod. 2008, 71, 8.
(8) Sato, K.; Horibe, K.; Amano, K.-i.; Mitusi-Saito, M.; Hori, M.; Matsunaga,
S.; Fusetani, N.; Ozaki, H.; Karaki, H. Toxicon 2001, 39, 259.
(9) (a) Freidinger, R. M.; Hinkle, J. S.; Perlow, D. S. J. Org. Chem. 1983,
48, 77. (b) Aurelio, L.; Box, J. S.; Brownlee, R. T. C.; Hughes, A. B.; Sleebs,
M. M. J. Org. Chem. 2003, 68, 2652.
906 J. Org. Chem. 2009, 74, 906–909
10.1021/jo802213q CCC: $40.75 2009 American Chemical Society
Published on Web 12/02/2008

SCHEME 1. Synthesis of Halicylindramide A
SCHEME 2. OfN Acyl Shifted Byproduct
of shaking with PyBOP/HOAt/DIPEA was sufficient to com-
plete the reaction. However, repeated coupling (three times) was
required for the attachment of Fmoc-Thr to the N-Me-Gln
residue, providing the resin-bound linear peptide 3. Coupling
of the D-Cys analog was performed while the side chain of the
Cys residue was protected with an acetamidomethyl (Acm)
group, which can later be removed selectively and oxidized to
cysteic acid. To build a cycle on the resin, Alloc-Sar-OH was
coupled to the hydroxyl group of the Thr residue via an
anhydride method using the DIC/DMAP procedure, and then
the Alloc and allyl protecting groups at the Sar and Asn residues
were removed selectively by treatment with Pd(Ph3P)4.10 The
deprotected peptide was cyclized between the Sar and Asn
residues with an excess of a HATU/HOAt/NMM mixture,
providing the resin-bound depsipeptide 5. The cyclization
reaction was assessed using the chloranil test to detect the
absence of the secondary amine of the sarcosine residue.11 The
remaining N-terminal amino acids were coupled one by one
using PyBOP/HOAt/DIPEA conditions to afford compound 6.
When we tried to synthesize a cyclic peptide without attaching
the Cys analog, we obtained the cyclic peptide 8 as a main
product after resin cleavage (Scheme 2). The formation of a
byproduct via an OfN acyl shift was anticipated12 and may
be formed during the removal of the Fmoc group from the Thr
residue using piperidine. Therefore, the coupling of one more
amino acid to the Thr residue prior to depsipeptide cyclization
was necessary to prevent an undesired OfN acyl shift
rearrangement.
The remaining steps of the complete synthesis are the
oxidation of Cys and formylation of the N-terminal Ala residue.
Because Trp residues in peptides or proteins can be easily
oxidized to form various byproducts during oxidative pro-
cesses,13 a protection of the indole ring is important to minimize
byproduct formation. Therefore, the oxidation of the side chain
of Cys was performed while the Trp residue is protected with
a Boc group on the solid phase. The Acm protecting group was
14
removed selectively from the Cys residue using 1 M Hg(OAc)2
in DMF, and the resin was washed with 10% mercaptoethanol
in DMF to remove Hg2+ ions (Scheme 1). The free thiol was
treated with freshly prepared 1% performic acid in formic acid,15
and the resin was treated with dimethylsulfide to remove excess
oxidant. Other oxidants, including dimethyldioxirane16 and
(13) (a) Matthiesen, R.; Bauw, G.; Welinder, K. G. Anal. Chem. 2004, 76,
6848. (b) Simat, T. J.; Steinhart, H. J. Agric. Food Chem. 1998, 46, 490, and
references therein.
(14) Greene, T. W.; Wuts, P. G. M. Protecting Groups in Organic Synthesis,
3rd ed.; John Wiley & Sons: New York, 1999; Vol. II.
(15) Hirs, C. H. W. In Methods in Enzymology; Hirs, C. H. W., Ed.; Academic
Press: New York, 1967; Vol. 11, pp 197-199.
(16) Knapp, S.; Darout, E.; Amorelli, B. J. Org. Chem. 2006, 71, 1380.
(10) Guibe´, F. Tetrahedron 1998, 54, 2967.
(11) Vojkovsky, T. Pept. Res. 1995, 8, 236.
(12) (a) Stawikowski, M.; Cudic, P. Tetrahedron Lett. 2006, 47, 8587. (b)
Mouls, L.; Subra, G.; Enjalbal, C.; Martinez, J.; Aubagnac, J.-L. Tetrahedron
Lett. 2004, 45, 1173. (c) Mauger, A. B.; Stuart, O. A. Int. J. Pept. Protein Res.
1987, 30, 481.
J. Org. Chem. Vol. 74, No. 2, 2009 907

by shaking the resin in a solution of Fmoc amino acids, 4 equiv of
PyBOP (781 mg), 4 equiv of HOBt (230 mg) or HOAt (204 mg),
and 6 equiv of DIPEA (0.37 mL) in DMF (3 mL) for 1 h at room
temperature. An excess of amino acids [2 equiv for the coupling
of Asp, Thr, D-Phe, (N-Me)Gln, and D-Cys derivatives and 3 equiv
for Arg, D-Trp, Val, D-Val, Pro, (Br)Phe, and D-Ala derivatives]
was used. For the coupling of Fmoc-Thr-OH to the N-Me Gln
residue of the peptide, three couplings were required to obtain a
negative Kaiser test. The Fmoc group was removed using 20% (v/
v) piperidine in DMF (3 × 5 min).
Introduction of Alloc-Sar-OH to the side chain of the Thr residue
was achieved by mixing the resin with the anhydride solution, which
was prepared in a separate flask by combining 6 equiv of Alloc-
Sar-OH (390 mg, 2.25 mmol), 5 equiv of diisopropylcarbodiimide
(0.29 mL, 1.88 mmol), and 0.1 equiv of DMAP (4.6 mg, 0.0375
mmol) in CH2Cl2 (20 mL) at 0 °C for 20 min. A small portion of
the resin was cleaved with a solution containing 95% TFA, 2.5%
TIS, and 2.5% H2O, and the crude product was analyzed by RP-
HPLC (tR ) 17.2 min, condition A). HPLC conditions are described
in Supporting Information.
The Alloc and allyl protecting groups were removed by shaking
the resin with 0.1 equiv of Pd(PPh4)4 (43.3 mg, 0.0375 mmol) and
12 equiv of PhSiH3 (0.55 mL, 4.5 mmol) in CH2Cl2 (4 mL) under
nitrogen atmosphere for 4 h. The chloranil test was positive,
indicating the presence of the secondary amine of the Sar residue.
A small portion of the resin was cleaved and the crude product
was analyzed by RP-HPLC (tR ) 16.6 min, condition A).
The cyclization step was carried out using 5 equiv of N-methyl
morpholine (0.21 mL, 1.88 mmol), 5 equiv of HATU (713 mg,
1.88 mmol), and 1 equiv of HOAt (51 mg, 0.375 mmol) in DMF
(3 mL). After shaking overnight, a negative chloranil test result
was obtained. A small portion of the resin was cleaved with
TFA-CH2Cl2 (1:1), and the crude product was analyzed by RP-
HPLC (tR ) 17.6 min, condition A). LRMS (ESI) calcd 1074.2
for [M + H], found m/z 1075.4. Deprotection of S-Acm was
performed by treatment with 1 M Hg(OAc)2 in DMF (pH 4,
adjusted with a few drops of acetic acid) for 3 h in darkness. The
resin was washed with DMF (3 × 3 min) and ꢀ-mercaptoethanol-
DMF (1:9, 3 × 3 min) to remove Hg2+ from the resin-bound
peptide. A small portion of the resin was cleaved as above and the
crude product was analyzed by RP-HPLC (tR ) 11.1 min, condi-
tion B).
FIGURE 2. Comparison of CD spectra of halicylindramide A and 7.
chlorite,17 did not give the desired product. HPLC and mass
spectrometry analysis of the byproducts indicated that overoxi-
dized or fragmented peptides were produced. The oxidized
depsipeptide from the performic acid oxidation was cleaved from
the resin and then purified by RP-HPLC to provide 7 with a
1.5% overall yield.
Finally, formylation of the N-terminal Ala residue was
performed using an excess of cyanomethyl formate18 in a
solution of DMF-sodium phosphate buffer (pH 7.4) at room
temperature for 48 h. The crude product was purified by
semipreparative RP-HPLC to give the desired halicylindramide
A in 60% yield. The pure cyclodepsipeptide was analyzed by
1H, HRMS, CD, and analytical RP-HPLC. The 1H and analytical
RP-HPLC data of the product were identical to those of the
natural halicylindramide A.
CD spectra of the halicylindramide A and deformyl-halicy-
lindramide A (7) were examined to compare their conformations
under three different conditions (Figure 2). The overall con-
formations of both compounds were fairly similar, but a
significant change was observed in the presence of sodium
dodecyl sulfate (SDS), implying that the depsipeptide undergoes
a conformational change in the membrane environment.
In summary, we have demonstrated the total synthesis of
halicylindramide A via solid-phase peptide synthesis, oxidation,
and solution-phase formylation. This strategy may be applicable
to the synthesis of other similar depsipeptides containing a
cysteic acid residue. In addition, we confirmed the stereochem-
istry of naturally occurring halicylindramide A. Investigation
of the biological activity and conformational studies of halicy-
lindramide A mimetics using NMR and molecular dynamics
simulations are underway.
Oxidation of Cysteine. Performic acid (1%) solution was
prepared by mixing 35% hydrogen peroxide and 98% formic acid
(3:97, v/v), and the solution was allowed to stand at room
temperature for 1 h. The free thiol of the Cys residue was treated
with freshly prepared 1% performic acid (5 mL) in a bath at 0 °C
for 2 h under a nitrogen atmosphere. Dimethyl sulfide (0.4 mL)
was added to the reaction mixture, and the solution was left to stand
for 10 min to quench the excess oxidants. A small portion of the
resin was cleaved and the crude product was analyzed by RP-HPLC
(tR ) 12.7 min, condition B).
Peptide Cleavage and Isolation. The resin-bound peptide was
washed thoroughly with DMF (10 × 3 min) and CH2Cl2 (5 × 3
min), and then dried in Vacuo overnight. The peptide was removed
from the solid support using TFA containing 2.5% (v/v) iPr3SiH
and 2.5% (v/v) H2O for 2 h at room temperature. This cleavage
step was repeated three times, and the combined solution was
concentrated to ∼2 mL under reduced pressure. Cold diethyl ether
was added to the solution and the precipitated peptide was filtered
and dried to obtain 60 mg of the crude 7. HPLC and mass
spectrometry analysis of the crude product indicated that the major
peak was the desired peptide. The crude peptide 7 was dissolved
in MeOH and allowed to stand for 2 days at room temperature.
After purification by semipreparative RP-HPLC, pure product 7
was obtained as a white solid (10 mg, 1.5% overall yield from the
starting resin). HRMS (ES+) calcd for C77H109BrN20NaO21S [M +
Experimental Section
Peptide Synthesis. Peptide synthesis was accomplished manually
via stepwise solid-phase synthesis using Fmoc chemistry. Reaction
completion was determined using qualitative Kaiser19 or chloranil
tests. Washings between deprotection and coupling were carried
out with DMF (7 × 3 min) and CH2Cl2 (3 × 3 min) using 10 mL
solvent/g resin for each treatment. After removal of the Fmoc group
from the Rink amide resin (500 mg, 0.375 mmol, 100-200 mesh,
1% DVB, 0.75 mmol/g), coupling of the amino acids was achieved
(17) Darkwa, J.; Olojo, R.; Olagunju, O.; Otoikhian, A.; Simoyi, R. J. Phys.
Chem. A 2003, 107, 9834.
(18) Deutsch, J.; Niclas, H.-J. Synth. Commun. 1993, 23, 1561.
(19) Kaiser, E.; Colescott, R. L.; Bossinger, C. D.; Cook, P. I. Anal. Biochem.
1970, 34, 595.
1
Na]2+ 893.8519, found 893.8511. H (DMSO-d6, 500.1 MHz): δ
10.68 (Trp1NH, brs), 8.79 (Br-Phe-NH, br), 8.62-8.49 (Cys(SO3)-
908 J. Org. Chem. Vol. 74, No. 2, 2009

NH, m), 8.35-8.23 (Phe-NH, br), 8.17 (Trp-NH, d, J ) 8.5), 8.05
(Arg-NH, d, J ) 7.5), 7.98-7.88 (Thr1-NH, Thr2-NH, Ala-NH,
br), 7.86-7,79 (D-Val-NH, Val-NH, m), 7.61 (Trp-C4, d, J ) 7.5),
7.52-7.40 (Br-Phe-C3C5, Arg-ꢀNH, Asn-NH, m), 7.43 (Gln-δNH2,
brs), 7.31 (Trp-C7, d, J ) 8.0), 7.29-7.21 (Phe-ꢁ, m), 7.19 (Phe-
δ1, δ2, d, J ) 7.5), 7.17 (Phe-ꢁ1, ꢁ2, s), 7.12 (Trp-C2, s), 7.04-7.01
(Trp-C6, m), 6.96-6.92 (Trp-C5, m), 6.74 (Asn-γNH2, s), 6.67 (Gln-
δNH2, s), 5.14 (Thr1-ꢀ, d, J ) 7.0), 5.08-5.03 (Thr1-R, m),
5.00-4.94 (Arg-R, Asn-R, m), 4.87-4.85 (Thr2-R, m), 4.83-4.79
(Br-Phe-R, m), 4.70-4.66 (Phe-R, Cys(SO3)-R, overlap), 4.64-4.59
(Trp-R, m), 4.48-4.42 (Pro-R, Gln-R, m), 4.36-4.32 (D-Val-R,
m), 4.20-4.16 (Val-R, m), 3.98-3.93 (Thr2-ꢀ, m), 3.76-3.73 (Ala-
R, m), 3.66-3.61 (Pro-δ, m), 3.10-3.04 (Arg-δ, Br-Phe-ꢀ, m),
3.04-3.01 (Phe-ꢀ, Sar-R, m), 3.00-2.96 (Trp-ꢀ, m), 2.93 (Gln-
NMe, s), 2.82-2.68 (Cys(SO3)-ꢀ, Br-Phe-ꢀ′, m), 2.73 (Sar-NMe,
s), 2.64 (Asn-ꢀ, br), 2.14-2.02 (Gln-ꢀ, Pro-ꢀ, D-Val-ꢀ, Gln-γ, m),
1.97-1.78 (Pro-γ, Gln-ꢀ′, Val-ꢀ, m), 1.77-1.72 (Pro-ꢀ, m),
1.69-1.59 (Arg-ꢀ, m), 1.47-1.36 (Arg-γ, m), 1.16 (Thr1-γ, d, J
) 6.5), 1.06 (Ala-ꢀ, s), 0.91 (Thr2-γ, d, J ) 5.5), 0.84 (D-Val-γ′,
d, J ) 6.5), 0.80 (D-Val-γ, d, J ) 6.5), 0.60 (Val-γ′, d, J ) 6.5),
0.57 (Val-γ, d, J ) 6.0).
darkness. The crude product was purified by semipreparative RP-
HPLC to give halicylindramide A as a white solid (3 mg, 60%
yield). The 1H and HPLC data of the synthetic sample were exactly
the same as those of the natural halicylindramide A (see Supporting
Material). HRMS (ES+) calcd for C78H109BrN20Na2O22S [M +
2Na]2+ 919.3443, found 919.3441. Analytical RP-HPLC (condition
B): tR ) 14.1 min.
CD Measurement. CD experiments were performed using a
Jasco 715 spectropolarimeter with a path length of 1.0 cm at room
temperature. Each 100 µM sample was dissolved in TFE or sodium
phosphate buffer (20 mM, pH 7.4) in the presence or absence of 5
mM SDS. Data were collected from 260 to 190 nm in 1-nm
increments at a scan rate of 20 nm/min and a 3-s signal averaging
time.
Acknowledgment. Prof. H. Kang at Seoul National Univer-
sity is gratefully acknowledged for the donation of the halicy-
lindramide A sample. This research was supported by a grant
from the Marine Biotechnology Program funded by the Ministry
of Land, Transport and Maritime Affairs, Republic of Korea.
Halicylindramide A. Formylation of the N-terminal Ala residue
of 7 using cyanomethyl formate was performed using a minor
modification of the method of Deutsch and Niclas.18 Compound 7
(5 mg, 2.8 µmol) was dissolved in 1.5 mL of 75% DMF in H2O
containing sodium phosphate buffer (20 mM, pH 7.4). To this
solution was added cyanomethyl formate (1.3 µL, 10 equiv), and
the mixture was allowed to stand at room temperature for 48 h in
Supporting Information Available: General experimental
methods; comparison of 1H and RP-HPLC of the synthetic and
natural halicylindramide A; 1H and 2D NMR spectra of 7; UV
spectra of halicylindramide A and 7. This material is available
free of charge via the Internet at http://pubs.acs.org.
JO802213Q
J. Org. Chem. Vol. 74, No. 2, 2009 909