Total structure and inhibition of tumor cell proliferation of laxaphycins

Article abstract of DOI:10.1021/jm061307x

From a mixed assemblage of Lyngbya majuscula rich marine cyanobacteria, we isolated a series of cell growth inhibitory cyclic peptides, The structures of the two major components, laxaphycins A (1) and B (2), and of two minor peptides, laxaphycins B2 (3) and B3 (4), were determined by spectroscopic methods and degradative analysis. Absolute configurations of natural and nonproteinogenic amino acids were determined by a combination of hydrolysis, synthesis of noncommercial residues, chemical derivatization, and HPLC analysis. The organism producing the laxaphycins was identified as the cyanobacterium Anabaena torulosa. The antiproliferative activity of laxaphycins was investigated on a panel of solid and lymphoblastic cancer cells. Our results demonstrate that in contrast to laxaphycin A, laxaphycin B inhibits the proliferation of sensitive and resistant human cancer cell lines and that this activity is strongly increased in the presence of laxaphycin A. This effect appears to be due to an unusual biological synergism.

Full text of DOI:10.1021/jm061307x

1266
J. Med. Chem. 2007, 50, 1266-1279
Total Structure and Inhibition of Tumor Cell Proliferation of Laxaphycins
Isabelle Bonnard,† Marc Rolland,†,‡ Jean-Marie Salmon,§ Eric Debiton,| Chantal Barthomeuf,| and Bernard Banaigs*,†
Laboratoire de Chimie des Biomole´cules et de l’EnVironnement, UniVersite´ de Perpignan Via Domitia, 66860 Perpignan, France, Laboratoire
de Biophysique et Dynamique des Syste`mes Inte´gre´s, UniVersite´ de Perpignan Via Domitia, 66860 Perpignan, France, and Laboratoire de
Pharmacognosie et Biotechnologies, UMR-INSERM U-484, UniVersite´ d’AuVergne, 63001 Clermont-Fd, France
ReceiVed NoVember 10, 2006
From a mixed assemblage of Lyngbya majuscula rich marine cyanobacteria, we isolated a series of cell
growth inhibitory cyclic peptides. The structures of the two major components, laxaphycins A (1) and B
(2), and of two minor peptides, laxaphycins B2 (3) and B3 (4), were determined by spectroscopic methods
and degradative analysis. Absolute configurations of natural and nonproteinogenic amino acids were
determined by a combination of hydrolysis, synthesis of noncommercial residues, chemical derivatization,
and HPLC analysis. The organism producing the laxaphycins was identified as the cyanobacterium Anabaena
torulosa. The antiproliferative activity of laxaphycins was investigated on a panel of solid and lymphoblastic
cancer cells. Our results demonstrate that in contrast to laxaphycin A, laxaphycin B inhibits the proliferation
of sensitive and resistant human cancer cell lines and that this activity is strongly increased in the presence
of laxaphycin A. This effect appears to be due to an unusual biological synergism.
Introduction
phoides. In 1997, our group published preliminary results18 on
the total structure elucidation of laxaphycins A and B isolated
from a Lyngbya majuscula rich assemblage collected in French
Polynesia. More recently, MacMillan et al.19 described the
structure of lobocyclamides A-C from a cryptic cyanobacterial
mat containing Lyngbya conferVoides. These lipopeptides are
closely related to laxaphycins. It seems20 that lobocyclamide A
is the [L-Ser2, D-Tyr6, L-allo-Ile9]laxaphycin A analog and
lobocyclamide B, [(2R,3S)-3-OHLeu3, (2R,3R)-3-OHThr8,
(2S,3R)-3-OHPro10]laxaphycin B.
Cyanobacteria (blue-green algae) have been identified as one
of the most promising groups of organisms for the discovery
of novel, biologically active natural products. Members of the
orders Oscillatoriales, mainly Lyngbya majuscula, and Nosto-
cales have proven to be the most prolific sources.1 Cyanobacteria
produce many active cyclic peptides and depsipeptides. Most
contain unusual amino acid residues. However, those with cycles
of 10 to 14 amino acids are not so frequent. Nevertheless, some
(about fifteen) have been characterized among more than two
hundred peptides found in marine cyanobacteria. Peptides with
a 10-residue ring have been found in Calothrix fusca (calophy-
cin),2 Anabaena sp. (puwainaphycin E),3,4 and in different
species of Nostoc and Oscillatoria (microviridins).5 Kawagu-
shipeptin A6 and B7 (Microcystis aeruginosa, Anabaena laxa,
Hormothamnion enteromorphoides), scytonemin A8 (Scytonema
sp.), and schizothrin A9 (Schizotrix sp.) have 11 amino acids
cycles. A dodeca cyclic peptide, wewakazole,10 has been isolated
from Lyngbya majuscula. Tolybyssidins A and B11 (Tolypothrix
byssoidea), and malevamide B12 (Symploca laetaViridis) have
a 13-residues cycle, and malevamide C (Symploca laetaViridis)
is a 14-residue cyclic peptide. All these peptides contain several
structural features that are common to many cyanobacterial
peptides, including N-methylation, â-amino acids, and R-hy-
droxyacids.
In this study, we report the complete structure of laxaphycins
A (1) and B (2), and of two minor peptides related to laxaphycin
B, laxaphycin B2 (3), and laxaphycin B3 (4; Figure 1), as
determined by means of a combination of mass spectrometry,
NMR spectroscopy, and degradative analysis. Furthermore, we
demonstrated that the source of laxaphycins was the marine
cyanobacterium Anabaena torulosa.
The antiproliferative activity of the two major compounds,
laxaphycins A and B, was investigated on a human metastatic
malignant melanoma line (M4Beu), which is highly resistant
to usual pro-apoptotic drugs, on five human solid cancer cell
lines representative of the most common cancers in Western
countries [nonsmall-cell lung carcinoma (A549), breast carci-
noma (MCF7), ovarian carcinoma (PA1), prostatic carcinoma
(PC3), and colon carcinoma (DLD1)], on human normal
fibroblasts used as human normal cells and, on L-929 murine
immortalized cells. This latter cell line is commonly used in
screening programs designed to evaluate the cytotoxicity of
chemicals. The antiproliferative and cytotoxic activity of all
isolated compounds was also tested on a drug-sensitive human
leukemic cell line (CCRF-CEM) and related sublines that
express drug resistance associated with overexpression of Pgp
(CEM/VLB100) or with alteration of DNA topoisomerase II
(CEM/VM-1). Furthermore, it was assessed that laxaphycin A
at unactive concentration may potentiate the anticancer effect
of laxaphycin B. In brief, our results demonstrate that laxaphycin
B inhibits the proliferation of sensitive and resistant human
cancer cell lines at the micromolar level and that this property
is strongly enhanced in the presence of laxaphycin A.
In 1992, Frankmo¨lle et al.13,14 described peptide sequences15
of laxaphycins A and B isolated from the freshwater cyano-
bacterium Anabaena laxa. The same year, Gerwick et al.16,17
isolated and entirely described hormothamnin A, a (Z)-R,â-
didehydro-R-aminobutyric acid [Z-Dhb] analog of laxaphycin
A from the marine cyanobacteria Hormothamnion enteromor-
* To whom correspondence should be addressed. Tel.: +33 4 68 66 20
74. Fax: +33 4 68 66 22 23. E-mail: banaigs@univ-perp.fr.
† Laboratoire de Chimie des Biomole´cules et de l’Environnement,
Universite´ de Perpignan Via Domitia.
‡ Present address: Institut Europe´en des Membranes, Universite´ de
Montpellier II, France.
§ Laboratoire de Biophysique et Dynamique des Syste`mes Inte´gre´s,
Universite´ de Perpignan Via Domitia.
| Laboratoire de Pharmacognosie et Biotechnologies, Universite´
d’Auvergne.
10.1021/jm061307x CCC: $37.00 © 2007 American Chemical Society
Published on Web 02/27/2007

Tumor Cell Proliferation of Laxaphycins
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 6 1267
Figure 1. Structure of the lipopeptides isolated from the marine cyanobacterium Anabaena torulosa.
Structure of Laxaphycins
Laxaphycins were obtained as colorless amorphous solids and
were negative to the ninhydrin test, suggesting a blocked
N-terminus.
Isolation of Laxaphycins. In a preceding study, laxaphycins
A and B were isolated and described from an assemblage of
80% Lyngbya majuscula together with two other cyanobacterial
species: Anabaena sp. and Oscillatoria sp.18 The high rate of
L. majuscula in this assemblage together with the relative high
concentration of laxaphycins purified (>0.4% dry weight of the
assemblage) suggested that L. majuscula was the source of the
metabolites. But subsequent collections of L. majuscula from
the same site over a period of 10 years did not show the
laxaphycins, whereas we could reisolate laxaphycins from
Anabaena torulosa collected in the same place at snorkeling
depth.
Molecular ions obtained in FAB-MS and spectral data of 1H
and 13C 1D NMR showed immediately that the two major
peptides isolated from the marine cyanobacterial assemblage
were similar to laxaphycins A and B isolated from the fresh
water cyanobacterium Anabaena laxa, by Frankmo¨lle et al.13,14
They described peptide sequences but did not establish amino
acid stereochemistry. From the marine cyanobacteria Hor-
mothamnion enteromorphoides, Gerwick et al.16,17 isolated and
entirely described hormothamnin A, a (Z)-R,â-didehydro-R-
aminobutyric acid [(Z)-Dhb] analog of laxaphycin A.
Laxaphycin A: Structure Elucidation. Almost all the
abundant positive ions in the FAB MS/MS spectrum could be
assigned to fragments that resulted directly from the cleavage
of the peptide backbone. Nomenclature21 and mechanisms of
formation22 for the peptide fragment ions observed had been
already discussed. A pattern of cyclic peptide fragmentation was
established23 that was described by the initial protonation of an
amide nitrogen, scission of the N-acyl bond, and subsequent
fragmentation by loss of amino acid residues (successively or
competitively) from the ring-opened acylium ion. The parent
The mixed cyanobacterial assemblage (ww 1.3 kg), collected
as floating wreckage at Moorea atoll (French Polynesia) in
October, 1993, was preserved in ethanol. The ethanol extract
of cyanobacteria was concentrated and partitioned between a
series of solvents of increasing polarity. Repeated silica gel
chromatography of the dichloromethanic-soluble fraction, fol-
lowed by reversed-phase HPLC, yielded four peptides: two
major peptides, laxaphycin A and laxaphycin B, and two minor
laxaphycin B-type peptides, laxaphycin B2 and laxaphycin B3.

1268 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 6
Bonnard et al.
Figure 2. FAB MS/MS fragmentation of laxaphycin A.
ion at m/z [M + H]+ 1196.4, one of the most prominent ions in
the spectrum, was consistent with the formula C60H97N11O14.
In the higher mass region, ions resulting from the losses of H2O
and CO (characteristic of a cyclic peptide) were observed in
very weak abundance (<2%).
to Marfey’s rule because D-trans-4-hydroxyproline [(2R,4S)-4-
OHPro] is eluted before L-trans-4-hydroxyproline [(2S,4R)-4-
OHPro]. A contrario, L-cis-4-hydroxyproline [(2S,4S)-4-OHPro]
is eluted before D-cis-4-hydroxyproline [(2R,4R)-4-OHPro]. The
configuration of the CâIle (D-Ile or D-allo-Ile) could not be
determined according to Marfey’s procedure. This was resolved
by amino acids OPA derivatization.26,27 This experiment showed
the presence of one Ile (D or L) and one allo-Ile (D or L). The
chirality of laxaphycin A amino acid units was identical to
hormothamnin A. Information concerning the assignment of the
Owing to the increased basicity of amide nitrogen atoms of
the N-alkylated 4-hydroxyproline (4-OHPro4) and of the R,â-
insatured R,â-didehydro-R-aminobutyric acid (Dhb3) in laxa-
phycin A, protonation and cleavage of these amide bonds were
a highly favored process. Thus, protonation on the peptide
nitrogen of 4-OHPro4 and Dhb3 was followed by ring cleavage
to form the two linear acylium ions (Figure 2) with Dhb3 and
4-OHPro4 at the N-terminus. These acylium ions underwent
C-terminus fragmentation and N-terminus fragmentation. We
could notice the presence of the m/z 970.6 ion resulting in the
R,â-bond cleavage of the â-aminooctanoic acid (Aoc).24 Wn
and Dn ion-type were not observed for Leu and Ile residues.
FAB mass spectrum of acetylated laxaphycin A showed a m/z
[M + H]+ 1322.5 pseudomolecular ion, indicating that laxa-
phycin A possesses three esterifiable functionalities.
1
two sets of H and 13C signals for the Leu residues to Leu7 or
Leu10 and that concerning the two sets of Ile residues to Ile8 or
Ile9 was obtained by HOHAHA and ROESY experiments, with
the cross-peaks between HRPhe6/HNLeu7, HRLeu7/HNIle8,
HRIle8/HNIle9, and HRIle9/HNLeu10. Sequential assignment of
laxaphycin A, determined by interpretation of ROESY and
HMBC data, was found to be identical to hormothamnin A.
According to Gerwick’s results17 and to the similarity of
chemical shifts between laxaphycin A and hormothamnin A
(∆δH < 0.2 ppm and ∆δC < 1 ppm except for CâDhb), the
two peptides seemed to have the same stereochemistry, except
for the Dhb unit. Laxaphycin A was determined to be cyclo-
[(3R)-Aoc-L-Hse-(E)-Dhb-(2S,4R)-4-OHPro-L-Hse-D-Phe-
D-Leu-L-Ile-D-allo-Ile-L-Leu-Gly-].
Laxaphycin B: Structure Elucidation. The FAB mass
spectrum recorded on laxaphycin B showed a [M + H]+
pseudomolecular ion at m/z 1395.6, consistent with the formula
C65H114N14O19. A [M + H]+ pseudomolecular ion at m/z 1605.5,
when recorded on acetylated laxaphycin B, allowed us to
suppose the presence of five esterifiable positions. Protonation
on the amide nitrogen of Pro10, N-methylisoleucine (N-MeIle7),
and 3-hydroxyasparagine (3-OHAsn8) generated amide bond
cleavage to form three linear acylium ions and produced series
of C- and N-terminus fragmentation (Figure 3). This FAB MS/
MS study established the amino acid sequence as cyclo[Ade-
Val-3-OHLeu-Ala-3-OHLeu-Gln-N-MeIle-3-OHAsn-
Thr-Pro-Leu-Thr].
The 1H NMR spectra of laxaphycin A were taken at 400 MHz
in DMSO-d6. It was obvious from these well-resolved spectra
that one conformation strongly predominated in this solvent.
The assignment of almost all 1H resonances (Table 1) was based
on connectivity information transmitted via DQF 1H-1H COSY
and HOHAHA spectra. In the 13C NMR spectrum, 60 carbon
resonances were apparent. The multiplicity of each carbon
resonance was determined by means of the usual DEPT
technique. Assignment of the 13C NMR resonances (Table 1)
was performed by using HMQC and HMBC experiments.
Detailed interpretation of the different 1D and 2D NMR spectra
confirmed the presence of 11 residues and established laxaphy-
cin A as cyclo[Aoc-Hse-Dhb-4-OHPro-Hse-Phe-Leu-
Ile-Ile-Leu-Gly-]. Information concerning the assignment
1
of the two sets of H and 13C signals for homoserine (Hse),
Leu, and Ile residues to Hse2, Hse5, Leu7, Ile8, Ile9, or Leu10
was obtained by HOHAHA and ROESY experiments. Strong
NOE between the NH and Hâ in the Dhb unit indicated that
the geometry of the double bond was E. Comparison of CâDhb
chemical shift for our peptide (119.0 ppm in DMSO-d6, 298
K), Frankmo¨lle’s [(E)-Dhb]-laxaphycin A (119.0 ppm in
DMSO-d6, 308 K), and Gerwick’s [(Z)-Dhb]-hormothamnin A
(121.7 ppm in DMSO-d6, 298 K) confirmed the geometry of
this residue.
The 1H NMR spectra of laxaphycin B were taken at 400 MHz
in DMSO-d6. One set of resonances was observed for each of
the residues, indicating that one conformation dominates in this
solvent. Some minor peaks, especially in the N-methyl and
amide proton regions, were apparent, indicating the presence
of another conformation in slow exchange. The assignment
procedures were very similar to those described for laxaphycin
A. The twelve amino acid units were characterized by detailed
Hydrolysis of laxaphycin A followed by Marfey’s derivati-
zation25 and HPLC analysis assigned 4-OHPro as (2S,4R) and
both Hse, one Leu, and one Ile (or allo-Ile) as L. Phe, one Leu,
and one Ile (or allo-Ile) were found to be D. The â-amino acid
Aoc was determined as 3R. â-aminooctanoic acid (and â-ami-
nodecanoic acid, Ade, in laxaphycin B) residues were synthe-
sized according to Gerwick’s procedure.17 We can notice that
elution of DAA-trans-4-hydroxyproline constitutes an exception
1
1
interpretation of H and 13C 2D NMR spectra (DQF H-1H
COSY, HOHAHA, and HMQC). The peptidic sequence was
also supported by deductions from HMBC and NOESY NMR
experiments. Information concerning the assignment of the two
sets of 1H and 13C signals of Thr and 3-OHLeu residues to Thr9,

Tumor Cell Proliferation of Laxaphycins
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 6 1269
Table 1. 1H and 13C NMR Data of Laxaphycin A (318 K) and Hormothamnin A (298 K) in DMSO-d6
laxaphycin A
hormothamnin A17
13C
1H
13C
1H
δ (ppm)
δ (ppm)
J (Hz)
δ (ppm)
δ (ppm)
â Aoc1
NH
6.82
1.69/1.97
4.27
1.34
1.23
1.23
1.23
0.84
8.4
6.88
1.68/1.90
4.28
1.35
1.25
CRH2
CâH
CγH2
CδH2
CꢀH2
CúH2
CηH3
CO
39.92
44.86
34.76
28.76
24.98a
30.72
13.68
169.06
39.90
44.80
35.10
30.80
8.4
9.5
9.5
9.5
9.5
7.5
13.90
0.89
Hse2
NH
7.10
4.54
1.76
3.46
4.42
7.1
7.09
4.70
1.80
3.55
4.59
CRH
CâH2
CγH2
OH
49.06
33.78
56.97
7.1 8.6
8.6 8.4
8.4
49.00
33.40
56.90
CO
NH
172.89
Dhb3
10.75
10.69
CRH
CâH
CγH3
CO
CRH
CâH2
CγH
OH
130.79
118.34
11.95
167.25
59.06
37.84
67.90
131.80
121.70
12.50
166.70
59.60
38.00
68.40
5.57
1.69
7.3
7.3
5.76
1.75
4-OHPro4
4.51
1.92/2.27
4.28
9.0, 5.0
9.0, 10.2/5.0, 10.2
9.0, 5.0, 3.6
4.48
1.83/2.25
4.30
5.03
5.19
CδH2
CO
56.97
170.09
3.34/3.59
5.0, 11.4/3.6, 11.4
57.60
3.32/3.61
Hse5
Phe6
NH
7.22
4.27
1.88/1.96
3.31/3.45
6.9
6.9 8.7
8.7, 8.7, 10.2/7.4, 4.4, 10.2
7.04
4.31
1.91/2.01
3.27/3.42
4.39
CRH
CâH2
CγH2
OH
CO
NH
48.90
33.78
56.97
48.70
33.90
56.90
171.97
7.79
8.2
7.65
CRH
CâH2
Cγ
CδH2
CꢀH2
CúH
CO
56.05
36.99
4.28
2.94/3.04
8.2, 4.4, 10.8
4.4, 13.6/10.8, 13.6
56.80
37.00
138.00
129.10
128.20
126.30
4.23
2.96/3.04
137.82
126.11
127.95
128.95
171.86
7.34
7.24
7.18
8.5
8.5 7.4
7.4
7.21
7.26
7.39
Leu7
NH
7.22
10.5
7.22
CRH
CâH2
CγH
CδH3
Cδ′H3
CO
51.55
42.24
23.94
22.70
20.31
171.54
4.28
1.18/1.34
1.58
0.80
0.73
51.40
39.50
24.70
21.10
22.80
4.32
1.03/1.28
1.57
0.75
0.81
7.0
6.6
Ile8
NH
6.61
4.63
1.76
1.18
0.76
0.75
6.49
4.76
1.80
1.24/1.26
0.75
CRH
CâH
CγH2
Cγ′H3
CδH3
CO
55.95
38.40
21.92a
15.25
11.32
172.18
9.2, 4.0
53.60
37.10
26.30
14.50
11.10
6.7
7.4
0.75
Ile9
NH
8.68
4.63
1.97
1.18
0.80
0.84
9.2
9.2, 4.0
8.44
4.68
2.07
1.12/1.19
0.81
CRH
CâH
CγH2
Cγ′H3
CδH3
CO
53.85
36.73
26.08a
14.34
11.04
172.35
53.60
37.10
26.30
14.50
11.10
7.0
7.5
0.84
Leu10
NH
8.34
5.70
8.49
CRH
CâH2
CγH
CδH3
Cδ′H3
CO
52.59
42.24
23.94
21.24
22.53
172.69
4.03
1.58/1.59
1.56
0.83
0.89
5.70, 4.40
53.10
39.20
24.10
21.40
22.60
4.02
1.36/1.56
1.63
0.85
0.92
6.8
6.5
Gly11
NH
8.56
8.84
CRH2
CO
42.24
166.77
3.81/3.22
8.1, 16.8/4.8, 16.8
42.30
3.26/3.81
a Chemical shifts may be interchanged.

1270 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 6
Bonnard et al.
Figure 3. FAB MS/MS fragmentation of laxaphycin B.
Thr12, 3-OHLeu3, or 3-OHLeu5 was obtained by ROESY
experiments (Table 2).
B was determined as cyclo[(3R)-Ade-L-Val-(2S,3S)-3-OHLeu-
L-Ala-(2R,3S)-3-OHLeu-L-Gln-L-N-MeIle-(2R,3R)-3-
OHAsn-L-Thr-L-Pro-D-Leu-L-Thr-].
As for laxaphycin A, hydrolysis of laxaphycin B, followed
by Marfey’s derivatization and HPLC analysis assigned Val,
Ala, Gln, Pro, and the two Thr units as L and Leu as D. The
â-amino acid Ade, synthesized according to Gerwick’s proce-
dure,17 was determined as 3R. Noncommercial residues were
synthesized to perform HPLC analysis. D- and D-allo-isoleucines
were N-methylated according to the procedure of Freidinger et
al.28 The four isomers of 3-hydroxyleucine were synthesized
via heterocyclic intermediates [cyclo(L-Val-Gly) and cyclo(D-
Val-Gly)] following Scho¨llkopf’s method.29-33 Absolute con-
figuration of 3-hydroxyleucines was determined to be (2R,3S)
and (2S,3S) and N-methylisoleucine was determined to be
(2S,3S).
The relative and absolute stereochemistry of 3-hydroxyas-
paragine was deduced by a combination of spectral and chemical
studies. In spectral studies, a coupling constant JRb of 2.5 Hz
defined the 3-hydroxyasparagine in a preferential rotamer. The
strong intraresidue NOEs, H2N/HR and H2N/Hâ, observed in
the ROESY spectrum of laxaphycin B and inter-residue NOEs
between 3-hydroxyasparagine and threonine (medium HNThr/
HR3-OHAsn and HRThr/Hâ3-OHAsn, weak OHThr/HR3-
OHAsn] eliminated erythro configurations (Figure 4).
For chemical studies, native 3-OHAsn from laxaphycin B
hydrolysis was collected after a chromatographic separation on
a cation exchange column and characterizated by 1H NMR and
circular dichroism. 3-Hydroxyaspartic acid was found threo by
1H NMR comparison with standard L/D-threo-3-hydroxyaspartic
acids, and a measurement of CD of this amino acid indicated D
configuration (see experimental). Coinjection of native 3-OHAsp
and standard L/D-threo-3-hydroxyaspartic acids, according to
Marfey’s procedure, indicates that the native (2R,3R)-3-OHAsp
coelutes with the first threo isomer. So 3-OHAsp elution is
another exception to Marfey’s rule because the D isomer is eluted
before the L isomer.
Laxaphycin B2: Structure Elucidation. Preliminary spec-
tral data examination, including FAB MS/MS, 1H, and 13C NMR
spectroscopy, showed that the new metabolite was a lower
homologue of laxaphycin B. The molecular weight of laxaphycin
B2 (3) at m/z 1378.5 [M + H]+ agreed with a protonated
molecular formula of C65H114N14O18, which was 16 amu smaller
due to the loss of a hydroxyl group in the 3-OHLeu5 residue.
Comparison of the FAB MS/MS spectrum with those obtained
for laxaphycin B showed the same fragmentation pattern with
the formation of the three linear acylium ions (Figure 5). The
N-terminus fragmentation of the acylium ion with 3-OHAsn8
at the N-terminus led to an acylium ions series (m/z 569.5, 440.2,
368.8) shifted to a lower mass by 16 amu, in comparison to
that of laxaphycin B, just before the m/z 256 (Gln-N-MeIle),
suggesting that the variable residue could be 3-OHLeu5.
The 1D and 2D NMR spectra were complex due to the
presence of a mixture of conformers in slow exchange. But the
absence of the Hâ3-OHLeu5 resonance and the presence of two
methyl resonances (Cδ at 22.7 and 21.2 ppm) and new signals
at 1.52 and 1.58 ppm, corresponding to HâLeu, were easily
detected (Table 2). Information concerning assignment of the
1
two sets of H and 13C Leu signals to Leu5 and Leu11 was
obtained by HOHAHA and ROESY experiments. Cross-peaks
in the ROESY spectrum between HRVal2/HN3-OHLeu3, HRLeu5/
HNGln6, HRLeu5/HRGln6, HRPro10/HNLeu11, and HRLeu11/
HNThr12 linked the Leu and Ile residues leading to the sequence
cyclo[Ade-Val-3-OHLeu-Ala-Leu-Gln-N-MeIle-3-
OHAsn-Thr-Pro-Leu-Thr]. Laxaphycin B2 gross structure
differs from laxaphycin B by replacement of 3-OHLeu5 with
Leu5 residue.
Hydrolysis of laxaphycin B2 followed by Marfey’s deriva-
tization and HPLC analysis assigned Val, Ala, Gln, N-MeIle,
Pro, and the two Thr units as L, Ade as 3R, 3-OHLeu as (2R,3S),
and the two Leu units as D. Finally, laxaphycin B2 differs from
laxaphycin B with respect not only to the substitution of (2R,3S)-
3-OHLeu5 by D-Leu5 but also to the inversion of the configu-
ration of the CR3-OHLeu3 residue. The structure of laxaphycin
B2 was determined as cyclo[(3R)-Ade-L-Val-(2R,3S)-3-
OHLeu-L-Ala-D-Leu-L-Gln-L-N-MeIle-(2R,3R)-3-OHAsn-
L-Thr-L-Pro-D-Leu-L-Thr-].
The presence of two 3-OHLeu with different configurations
introduced another question: the localization of the two residues
in the peptide sequence. This localization was deduced as
follows: partial hydrolysis of laxaphycin B at 120 °C with a
10% solution of TFA in water and for 30, 40, and 45 min
showed a progressive liberation of amino acid units (identificated
by Marfey’s procedure). At 30 min, Leu, 3-OHAsp, and N-MeIle
appeared in the chromatogram pattern. Then at 40 min, Glu,
(2R,3S)-3-OHLeu, and Ala were liberated. Finally, at 45 min,
(2S,3S)-3-OHLeu and Val appeared. This experiment lead us
to localize (2R,3S)-3-OHLeu between Gln and Ala and (2S,3S)-
3-OHLeu between Ala and Val. Finally, structure of laxaphycin
Laxaphycin B3: Structure Elucidation. FABMS data
established the formula for laxaphycin B3, C65H114N14O20, as a
higher homologue of laxaphycin B. A similar pattern of
fragmentation for both laxaphycin B and B3 was observed on
the FAB MS/MS spectrum (Figure 6). The N-terminus frag-

Tumor Cell Proliferation of Laxaphycins
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 6 1271
Table 2. 1H and 13C NMR Data of Laxaphycins B, B3 (308 K), and B2 (318 K) in DMSO-d6
laxaphycin B
1H
laxaphycin B2
1H
laxaphycin B3
1H
13C
13C
13C
δ (ppm)
δ (ppm)
7.58
2.40/2.33 8.8, 14.9/4.0, 14.9
J (Hz)
δ (ppm)
δ (ppm)
J (Hz)
δ (ppm)
δ (ppm)
J (Hz)
â-Ade1
NH
8.8
7.58
2.43/2.33
4.08
7.52
2.44/2.30
4.08
9.2
7.5, 3.3
CRH2
CâH
CγH2
CδH2
CꢀH2
CúH2
CηH2
CθH2
CιH3
CO
40.28
45.93
33.45
28.67a 1.24
28.47a 1.20
25.18a 1.20
31.11a 1.20
21.92a 1.20
4.11
8.8 4.0
46.08
11.2, 6.0
45.92
33.41
1.40/1.29
1.54/1.50
1.40
28.64a 1.50/1.38
28.40a 1.29
1.29
31.03a 1.29
21.83a 1.34
28.69a 1.24
28.47a 1.20
25.22a 1.20
31.10a 1.20
21.92a 1.20
7.1
6.6
7.1
13.79
171.14
0.84
7.3
13.68
171.70
0.92
7.2
7.3
13.79
171.30
0.82
6.8
Val2
NH
8.18
4.09
1.97
0.91
0.85
7.0
7.0
8.11
4.04
2.10
1.00
0.95
8.10
4.12
1.98
0.88
0.84
7.6
7.0
CRH
CâH2
CγH3
Cγ′H3
CO
59.03
29.33
18.80
18.87
171.05
59.20
29.37
18.67
18.45
172.18a
58.89
29.37
18.56
18.85
171.30
6.7
6.7
7.2
7.2
3-OHLeu3
NH
7.94
4.34
3.49
4.94
1.58
9.1
9.1 2.0
7.98
4.33
3.51
4.71
1.70
1.02
7.90
4.37
3.50
4.90
1.60
8.4
8.4
CRH
CâH
OH
CγH
CδH3
Cδ′H3
CO
55.23
76.37
55.37
76.42
10.3
8.3, 2.4
55.15
76.48
4.4
14.7, 6.8
6.8
7.2
30.54
19.22a 0.89
18.56
171.35
30.01
18.67
18.98a 0.88
170.94
30.57
18.76a 0.89
18.43
6.7
6.7
6.8
6.8
0.76
6.8
0.76
Ala4
NH
7.86
4.22
1.31
6.0
6.0, 7.6
7.6
7.80
4.22
1.33
7.87
4.22
1.32
CRH
CâH3
CO
49.28
17.55
172.33
48.79
17.55
49.30
17.65
172.47
6.4
7.6
7.0
171.61a
OHLeu5/ Leu5
NH
7.69
4.28
3.49
5.03
1.56
8.8
8.8, 2.9
7.77
4.17
1.58/1.52
6.5
6.5 7.5
7.61
4.28
3.48
5.05
1.58
8.8
8.8 2.0
CRH
CâH
OH
CγH
CδH3
Cδ′H3
CO
55.52
75.80
51.50
40.20
55.64
75.78
5.4
14.7, 6.8
6.8
5.6
14.7, 6.8
7.2
29.90
23.99
22.67
21.21
171.79a
1.67
0.95
0.92
29.84
18.65a 0.89
18.69a 0.88
0.74
18.56
170.50
0.76
6.8
6.7
7.2
170.60
Gln6
NH
7.77
4.63
7.4
9.2, 4.0
7.81
4.58
2.05/1.84
2.21
7.56
4.58
2.00/1.64
2.23/2.15
4.4
CRH
CâH2
CγH2
CON
NH2
CO
NCH3
CRH
CâH
CγH2
Cγ′H3
CδH3
CO
49.16
26.39
30.72
48.79
49.40
1.97/1.75 14.8, 7.0/-
2.10/2.04 14.8, 8.2/-
30.45
174.56
174.60
174.74
7.22/6.85
2.97
4.72
1.90
1.29/0.89
0.76
0.78
7.16/6.74
7.17/6.79
172.49
30.03
59.85
31.56
23.88
15.08
10.33
170.02
171.54a
29.68
59.93
172.64
30.15
59.87
31.80
N-MeIle7
3-OHAsn8
Thr9
2.93
4.72
2.02
-/0.87
3.01
4.73
1.90
1.27/0.74
0.74
11.0
7.4
10.8
6.7
11.2
23.90
15.14
14.99
10.31
170.10
6.4
6.4
0.99
7.5
0.75
170.00
NH
7.64
4.63
4.31
5.79
8.5
7.45
4.64
4.36
5.78
8.7
8.7 2.2
7.1
7.66
4.63
4.35
5.70
CRH
CâH
OH
CONH2 173.37
NH2
55.52
70.44
8.5, 2.5
2.5, 6.0
6.0
55.37
70.39
55.53
70.33
8.0, 2.0
6.0
7.1
173.26
169.11
173.37
169.12
7.27
7.22
7.17
CO
NH
169.16
7.33
4.49
3.93
4.94
1.05
8.0
8.0, 6.0
7.34
4.49
3.95
4.84
1.14
7.3
7.12
4.46
3.90
4.89
8.0
CRH
CâH
OH
CγH3
CO
55.61
66.23
55.39
66.15
55.83
66.43
4.4
6.4
9.6
6.5
5.2
6.4
18.87
168.58
59.60
29.08
24.0
18.67
168.67
59.69
18.85a 1.03
168.70a
Pro10/ OHPro10 CRH
4.33
2.04/1.82
1.90/1.80 8.3/-
3.68 3.9
4.0
4.35
2.15/1.92
2.02
58.62
37.73
68.50
55.48
171.47a
4.43
2.01/1.84
4.32/5.08 OH -/4.0
3.72/3.58 5.6, 10.0/5.6, 10.0
8.4
CâH2
CγH2
CδH2
CO
28.64
47.16
171.21
47.14
171.32a
3.66

1272 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 6
Bonnard et al.
Table 2. Continued
laxaphycin B
laxaphycin B2
1H
laxaphycin B3
1H
13C
1H
13C
13C
δ (ppm)
δ (ppm)
J (Hz)
δ (ppm)
δ (ppm)
J (Hz)
δ (ppm)
δ (ppm)
J (Hz)
Leu11
NH
7.89
4.31
1.47
1.53
0.87
0.82
7.5
7.5
7.92
4.84
1.60
1.61
0.95
0.92
7.2
7.2
7.86
4.35
1.47
1.52
0.86
0.80
CRH
CâH2
CγH
CδH3
Cδ′H3
CO
51.36
40.82
24.06
22.71
21.76
171.67
51.50
40.20
23.99
22.67
21.54
171.41a
51.31
41.24
24.12
22.75
21.72
171.41a
7.2
7.6, 6.0
6.0
6.8
6.8
6.7
Thr12
NH
7.74
4.11
4.00
4.78
0.99
8.9
8.9, 3.4
7.61
4.09
4.00
4.73
1.09
7.68
4.10
3.97
4.80
0.99
4.0
CRH
CâH
OH
CγH3
CO
57.85
66.19
57.99
66.15
6.0, 4.0
6.5
58.17
66.35
5.2
6.5
5.6
6.8
19.46
168.67
19.35
168.67
19.48
168.67a
a Chemical shifts may be interchanged.
Figure 4. Inter- and intraresidue NOEs observed for the partial structure 3-OHAsn8-L-Thr9. Arrows indicate small (*), medium (**), and strong
(***) dipolar interactions.
Figure 5. FAB MS/MS fragmentation of laxaphycin B2.
mentation of the acylium ion with Pro at the N-terminus led
us to an acylium ion series shifted to a higher mass by 16
amu in comparison to that of laxaphycin B, suggesting that
the variable residue could be Pro or Leu. The other series of
fragmentation not being complete, it was not possible to con-
clude in which residue lay the difference according to mass
analysis.
The NMR spectral analysis established the variable residue
as 4-hydroxyproline (Table 2). Laxaphycin B3 showed remark-
able NMR spectral similarities to laxaphycin B, the most
significant differences being the presence of an additional
hydroxyl group on proline [OH 5.08 ppm and Hγ at 4.32 ppm
vs two Hγ at 1.90 and 1.80 ppm for laxaphycin B; Cγ at 68.5
ppm vs 24.0 ppm for laxaphycin B; Câ and Cδ were also
deblinded by the presence of the hydroxyl function (∆δ 8.6
and 8.3 ppm, respectively)].
Marfey’s analysis of laxaphycin B3 indicated a (2S,4R)
configuration for 4-OHPro and no change for the other residues.
A similarity of the chemical shifts between laxaphycin B3 and
laxaphycin B led us to conclude that both peptides have the
same peptidic sequence. Thus, the structure of laxaphycin B3
was established as cyclo[(3R)-Ade-L-Val-(2S,3S)-3-OHLeu-
L-Ala-(2R,3S)-3-OHLeu-L-Gln-L-N-MeIle-(2R,3R)-3-
OHAsn-L-Thr-(2S,4R)-4-OHPro-D-Leu-L-Thr-].

Tumor Cell Proliferation of Laxaphycins
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 6 1273
Figure 6. FAB MS/MS fragmentation of laxaphycin B3.
Biological Evaluation of Laxaphycins
Growth-Inhibitory Effect of Laxaphycins on Solid Cancer
Cell Lines. To provide information on their antiproliferative
activity against solid cancers, laxaphycin A (0-50 µmol) and
laxaphycin B (0-20 µmol) were tested versus cisplatin (0-50
µmol) on five human carcinoma cell lines: A549 (non small
cell lung), MCF7 (breast), PA1 (ovary), PC3 (prostate), and
DLD1 (colon) and a human pigmented malignant melanoma
line (M4Beu). To obtain additional information on their
specificity for human cell lines and their specificity for cancer
versus normal cell lines, each peptide was tested versus L-929
murine immortalized cells and human normal fibroblasts. The
growth inhibitory effect was determined by measuring first the
metabolic activity of treated and control cultures by the resazurin
reduction test (RRT) and then the amount of cell biomass in
each well with Hoechst 33342 (Ho). A dose-dependent inhibition
of cell proliferation was observed in all cell lines. The IC50 of
each chemical varied with the cell lines and the used test.
However, the data clearly demonstrated the low toxicity of
laxaphycin A (IC50 > 20 µM on all cell lines) as compared to
laxaphycin B (IC50< 2 µM on all cell lines, DLD1 excepted).
The low toxicity of laxaphycin A observed in our model is in
agreement with data by Frankmo¨lle et al.13 who reported a lack
of toxicity on human KB cells (epidermoid carcinoma) and
human LoVo cells (colorectal adenocarcinoma). It has been
reported that contrary to laxaphycin A, hormothamnin A is
highly toxic on a variety of solid cancer cell lines. Gerwick et
al.16 calculated the IC50 values of hormothamnin A as, respec-
tively, 0.17 and 0.13 µM on SW1271 and A529 lung carcinoma
cells as 0.11 µM on B16-F10 murine melanoma cells and as
0.60 µM on HCT-116 human colon carcinoma cells. As the
only difference between laxaphycin A and hormothamnin A is
a different configuration of the Dhb residue, it is likely that the
configuration of this residue plays a critical role in the higher
cytoxicity of hormothamnin A.
Figure 7. In vitro cell toxicity of laxaphycin B as compared to cisplatin
against human solid cancer cells, human normal fibroblasts, and L929
murine immortalized cells. Cell toxicity was evaluated with the Hoechst
33342 (Ho), and the RRT assay on cultures submitted to a 48 h
continuous contact with specified concentration of drugs. Results were
expressed as the percentage of cells surviving the treatment versus
control cells treated with the solvent alone (DMSO 0.5%). Two
independent experiments were made in triplicate and averaged ((SD).
RRT) and MCF7 breast carcinoma (IC50 0.5 ( 0.07 in Ho test
and 1.0 ( 0.12 µM in RRT) cells. Laxaphycin B has, therefore,
the capacity to block not only the proliferation of highly
proliferative carcinoma cells (PA1) but also that of strongly
invasive malignant melanoma cells (M4Beu). Regarding the
increasing number of deaths relative to malignant melanoma
in the European Union and North America, this finding is of
interest. Together with the significant cytotoxicity on PC3 and
MCF7 cells, which are representative of two types of cancer
whose incidence is still in progression in Western countries,
the tumoricidal effect on M4Beu metastatic melanoma justifies
further investigations to assess the mechanism of action of this
peptide and to determine whether the cytotoxicity of laxaphycin
B is cell-type or tissue-dependent. Because this peptide has only
The individual susceptibility of each tested cancer (immortal-
ized and normal) cell line to laxaphycin B and cisplatin is
presented in Figure 7. Laxaphycin B (0.19 µM < IC50 < 6.0
µM) appeared markedly more toxic than the clinically active
drug (2.0 µM < IC50 < 23.9 µM) on all tested human lines. In
the human cancerous cell lines, the IC50 values of laxaphycin
B varied between 0.19 and 1.0 µM except on DLD1 colon
carcinoma cells, which appeared markedly more resistant to this
peptide. The cells exhibiting the highest susceptibility were
human PA1 ovary carcinoma cells (IC50 0.19 ( 0.03 µM),
M4Beu (IC50 0.3 µM ( 0.03 in Ho test and 0.4 ( 0.03 µM in
RRT test), and to a lesser extent androgen-resistant PC3 prostate
carcinoma (IC50 0.58 ( 0.03 in Ho test and 0.8 ( 0.1 µM in

1274 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 6
Bonnard et al.
Table 3. In Vitro Cytotoxicity of Laxaphycins against Drug-Sensitive and Multidrug-Resistant Tumor Cell Lines
CCRF-CEM
CEM/VLB100
resistance
factorb
CEM/VM-1
Resistance
factorb
IC50a (µM)
IC50a (µM)
IC50a (µM)
adriamycin
vinblastine
laxaphycin A
laxaphycin B
laxaphycin B2
laxaphycin B3
0.06 ( 0.02
0.004 ( 0.001
>20
1.11 ( 0.15
4.15 ( 0.31
1.50 ( 0.06
3.71 ( 1.30
0.356 ( 0.040
>20
1.02 ( 0.05
3.33 ( 0.08
1.35 ( 0.05
62
89
0.55 ( 0.14
0.008 ( 0.003
>20
1.37 ( 0.15
4.14 ( 0.09
1.45 ( 0.11
9
2
0.9
0.8
0.9
1.2
1
1
a A 50% inhibitory concentration in a 72 h growth inhibition assay ( SD. Values shown are the means of three separate experiments. b Resistance factors
calculated as IC50 for drug-resistant cells/IC50 for drug-sensitive cells.
a limited specificity for certain cancer cells in vitro, experiments
in animal model bearing xenograft tumors would be useful to
assess the therapeutic index of laxaphycin B in vivo.
Growth-Inhibitory Effect of Laxaphycins on Multidrug-
Resistant versus Sensitive Human Leukemic Cell Lines. We
also screened the cell toxicity of laxaphycins A, B, B2, and B3
against lymphoblastic drug-sensitive and multidrug-resistant cell
lines. The cytotoxicity of the different laxaphycins was evaluated
for the parent drug-sensitive CCRF-CEM human leukemic
lymphoblasts and sublines that express drug resistance associated
with either overexpression of the plasma membrane protein
P-glycoprotein (Pgp) for CEM/VLB100 cells (MDR phenotype)34
or with altered DNA topoisomerase II for the CEM/VM-1 cells.
This latter cell line exhibits resistance to teniposide (VM-26)
and is usually identified as an atypical MDR cell line.35 Results
are summarized in Table 3. In all these experiments, adriamycin
and vinblastine were used as positive standards and references
of anticancer activity.
Figure 8. Enhancement of laxaphycin B cytotoxicity (2 µM) by
laxaphycin A on DLD1 colon cancer cells. Cell toxicity was evaluated
with the RRT assay on cultures submitted to a 48 h continuous contact
with specified concentration of drugs. Results were expressed as the
percentage of cells surviving the treatment versus control cells treated
with the solvent alone (DMSO 0.5%). Data are the average of two
independent experiments made in triplicate.
Laxaphycin A exhibited no activity on acute lymphoblastic
leukaemia cell lines when tested for 72 h at a concentration of
20 µM. On the drug-sensitive CCRF-CEM human leukemic
lymphoblasts, IC50 values of laxaphycins B, B2, and B3 ranged
from 1.1 µM to 4.2 µM. Interestingly, these results give some
new information on SAR. They indicate that if the γ-hydroxy-
lation of Pro10 in laxaphycin B3 have a weak impact on cell
toxicity, substitution of 3-OHLeu5 by Leu5 in laxaphycin B2
and inversion of the configuration of the CR 3-OHLeu3 resulted
in a less toxic compound.
The growth-inhibitory effect of each laxaphycin B was then
evaluated on the two resistant leukemic cell lines. The resistance
factors of each laxaphycin are presented as the ratio of IC50 for
drug-resistant cells/IC50 for parental cells. Interestingly, the two
tested MDR cell lines which are characterized by a 62- and
9-fold resistance to adriamycin and an 89- and 2-fold resistance
to vinblastine exhibited similar susceptibility to laxaphycin
B-type peptides than the parental cell line. The lack of resistance
of CEM/VLB100 cells to laxaphycins B, B2, and B3 seems to
indicate that none of these peptides is a substrate for P-
glycoprotein (Pgp). This energy efflux pump, which extrudes
many chemotherapeutic drugs out of cancer cells, plays a key
role in the development of multidrug resistance by cancer cells
following chemotherapy.36 Pgp located in intestinal lumen may
also drive compounds back into the intestinal lumen, preventing
their absorption into blood.37 Although this has to be confirmed
by further studies, our results suggest that laxaphycin B presents
the dual advantage of not contributing to MDR encoded by the
MDR1 gene and of being unaffected by a decreased bioavail-
ability related to Pgp.
Table 4. Potentialization of Cell Growth Inhibition of Laxaphycin B by
Laxaphycin A: Determination of Interaction Index γ
laxaphycin A dose laxaphycin B
(µM)
IC50a (µM)
A/B ratiob
γc
CCRF-CEM
CEM/VM-1
CEM/VLB100
0
1.11
0.62
0.50
0.41
0.35
1.37
0.72
0.60
0.43
0.41
1.02
0.92
0.57
0.36
0.33
0
0.2
0.4
1.0
2.0
0
0.2
0.4
1.0
2.0
0
0.3
0.8
2.4
5.6
0
0.3
0.7
2.3
4.9
0
0.56
0.45
0.41
0.32
0.53
0.44
0.32
0.30
0.2
0.4
1.0
2.0
0.2
0.7
2.8
6
0.90
0.56
0.35
0.32
a A 50% inhibitory concentration in a 72 h growth inhibition assay.
Values shown are the means of three separate experiments. b A/B ratio is
the ratio laxaphycin A concentration/laxaphycin B IC50. c Interaction indexes
(γ) were calculated by the following formula: γ ) (IC50 B combinated/
IC50 B alone); γ > 1 indicates antagonism, γ ) 1 indicates additivity, γ <
1 indicates synergism, and γ < 0.5 indicates strong synergism.
lus orizae and KB cells. Our group has previously observed a
synergistic effect between these peptides against Candida
albicans.18 The antiproliferative effects of laxaphycin B on
cancer cells was evaluated in combination with laxaphycin A.
Results in Figure 8 and Table 4 show that, at inactive
concentration, laxaphycin A strongly potentiates the cell toxicity
of laxaphycin B. A significant synergistic effect was observed
on DLD1 colon carcinoma cells and on multidrug-resistant
leukemic cell lines. A total of 90% of the DLD1 cells survived
to a 48 h continuous treatment with 2 µM of laxaphycin B.
Interaction of Laxaphycins A and B on Growth Inhibition
of Human Tumor Cell Lines. Two drugs used in combination
may produce enhanced (synergism), additive, or reduced
(antagonism) effects. Frankmo¨lle et al.13 reported a synergistic
activity of laxaphycins A and B against the growth of Aspergil-

Tumor Cell Proliferation of Laxaphycins
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 6 1275
Singly, laxaphycin A at 5 µM induced no antiproliferative
activity on this cell line. However, in combination at 1, 1.5, or
2 µM, with 2 µM of laxaphycin B, it significantly increased
the cytotoxicity of laxaphycin B on DLD1 cells (Figure 8). This
indicates a synergistic effect whose efficacy increases with
increasing concentration of laxaphycin A.
nonpolar alternating D and L residues (Pro10-Leu11-Thr12-Ade1-
Val2). This arrangement, common to laxaphycins A and B, is a
structural property found in peptides with the ability to become
inserted into lipid bilayer membranes. The sequence of laxa-
phycins B shows a less pronounced segregation between
hydrophobic and hydrophilic residues. The laxaphycin B-type
peptides significantly inhibited the proliferation of human solid
cancer cells and of sensitive and resistant leukemic cells at
micromolar levels. Although laxaphycin A exhibited weak
antiproliferative effect on leukemia and solid cancer cells, it
presented the great advantage to strongly potentiate the anti-
cancer effect of laxaphycin B on sensitive and resistant cancer
cells. In brief, these data suggest that a combination of
laxaphycins A and B may be of great interest in human
chemotherapy. Studies are underway to assess this hypothesis
and to identify the mechanism(s) involved in the synergistic
effect of the active mixtures.
As it is uncommon to encounter synergistic combinations in
which one of the two compounds lacks efficacy, we assessed
the degree of synergism from the measure of the interaction
index (γ), a quantity that indicates the changed potency of a
combination of two drugs.38 In the case in which laxaphycin A
is inactive (IC50 > 20 µM), γ is the ratio of the IC50 values of
laxaphycin B in combination and alone. IC50 values of laxa-
phycin B with different concentrations of laxaphycin A, from
0.2 to 2.0 µM, have been determined on the three leukemic
cell lines (CCRF-CEM, CEM/VM-1, CEM/VLB100) to estimate
the optimal combination mixture. The A/B ratio (laxaphycin A
concentration/laxaphycin B IC50) was tested in the range of 0-6.
As indicated in Table 4, laxaphycin A potentiates the activity
of laxaphycin B as well as on sensitive and on resistant leukemic
cell lines. The IC50 of laxaphycin B decreases with increasing
concentrations of laxaphycin A on the three cell lines. On
sensitive CCRF-CEM cells, the IC50 value of laxaphycin B was
reduced by 2-fold in the presence of 0.2 µM of laxaphycin A
(A/B ratio: 0.3) and by 3-fold when laxaphycin A was used at
1 µM (A/B ratio: 2.4) or 2 µM (A/B ratio: 5.6). A 3-fold
potentialization of the cell toxicity of laxaphycin B was also
observed on resistant CEM/VM-1 and CEM/VLB100 cells when
laxaphycin A was used at a A/B ratio in the range 2.5-6. In all
three cell lines, γ values decrease slowly when the A/B ratio is
>2.5. The optimal combination mixture seems to be obtained
for A/B ratio ) 2.5-3.
Finally, it was observed that the synergistic activities of
laxaphycin A and laxaphycin B are expressed on the growth of
all leukemic and solid cancer cells tested and on the growth of
two fungal strains, Aspergillus orizae and Candida albicans. It
was recently suggested that laxaphycin B may have a different
mechanism of action when it is administered in the presence of
laxaphycin A.39 It is noteworthy that the synergistic effect
between two small cyclic peptides is not frequent. Very few
examples of synergy have been shown, the association being
at a molecular level (example of iturins and surfactins from a
Bacillus strain) or the peptides acting on separate receptors
(example of pristinamycins and virginiamycins from two
different Streptomyces strains).
Experimental Section
General Instrumentation. Nuclear magnetic resonance (NMR)
spectra were recorded on a Jeol EX 400 spectrometer. IR and UV
spectra were recorded, respectively, on a Perkin-Elmer 1600 FTIR
spectrometer and a Perkin-Elmer 551 spectrometer. Liquid second-
ary ion mass spectra were performed on a Autospec instrument
(Fisons, VG analytical, Manchester, U.K.) fitted with a liquid
secondary ion source (LSIMS). The Cesium gun worked at 30 kV,
with the ion source voltage being 8 kV. Samples were dissolved in
a few µL of 20% aqueous acetic acid, and 1 µL was mixed on the
target with 1 µL of a 1:1 (v/v) glycerol thioglycerol mixture
acidified with 1 µL of 1% trichloroacetic acid in water. For CID
measurements, the collision cell fitted in the first field-free region
was filled with helium so as to reduce the precursor ion beam to
50% of its original value. Constant B/E linked scans were
performed, with both electrostatic analyzers being scanned in
parallel. High performance liquid chromatography (HPLC) was
performed with Jasco 880-PU pumps, 7125 Rheodyne injectors,
and either a Merck (LMC systeme) differential refractometer
detector or a Waters 996 photodiode array detector. OPA-derived
amino acids were analyzed by Gilson automatic injector and HPLC
system, detected on a Shimazu RF530 fluorimeter.
Isolation of Laxaphycins. A sample (1.24 kg) of Lyngbya
majuscula collected off the coast of Tahiti in 1993 was stored in
EtOH until workup. The extract obtained by repetitive steeping in
CHCl3/EtOH was separated by solvent partition. The chloroformic
extract (4 g) was subjected to flash chromatography and silica gel
column chromatography (solvent: 100% CH2Cl2 to 100% MeOH),
then the peptide mixture was purified by using C18 reversed-phase
flash column chromatography with CH3CN/H2O (50:50). The
individual laxaphycins were separated and purified by repetitive
reversed-phase HPLC (SFCC Ultrabase UB 535 C8 column, 250
× 10 mm, 5 µm particle size, flow rate 1.5 mL/min, differential
refractometry detection) using CH3CN/H2O (50:50). Laxaphycin
B3 (15 mg) was eluted at 16.8 min, laxaphycin B (184 mg) at 20.0
min, laxaphycin B2 (11 mg) at 23.9 min, and laxaphycin A (110
mg) at 34.0 min.
Conclusion
Laxaphycins A and B are lipophilic cyclopeptides coproduced
by the fresh water or marine filamentous cyanobacteria Ana-
baena laxa and A. torulosa. It is interesting to note that
lobocyclamides A and B, closely related to laxaphycins A and
B, were isolated from a mixture of Lyngbya conferVoides and
coral sand with minor amounts (up to 10%) of Anabaena sp.
and Oscillatoria spp.19 We cannot rule out the possibility that
the source of lobocyclamides may be the strain Anabaena sp.
In addition to ribosomal L-amino acids, these unusual peptides
contain several “exotic” amino acids, including R,â-didehydro-
R-aminobutyric acid, 3-hydroxyleucine, 3-hydroxyasparagine,
or 3-aminoocta(deca)noic acids. Laxaphycin A-type peptides
are cyclic undecapeptides in which the sequence shows a
segregation between hydrophobic and hydrophilic residues. The
hydrophobic face is characterized by a sequence of an alternating
D- and L-leucine/isoleucine residue tetrad. The laxaphycin B-type
peptides are cyclic dodecapeptides also with a sequence of
NMR Measurement Conditions. All spectra were obtained with
1
a NM-40TH5 dual H, 13C probe in a JEOL EX400 operating at
400 MHz for proton and 100.53 MHz for carbon-13, at 308 K for
laxaphycin B, 318 and 298 K for laxaphycin A, 318 K for
1
laxaphycin B2, and 308 K for laxaphycin B3. H and 13C NMR
chemical shifts are referenced to solvent peaks: δH 2.49 ppm
(residual DMSO-H6) and δC 39.5 ppm for DMSO-d6.
Two-dimensional (2D) homonuclear correlated experiments
DQF-COSY, HOHAHA, and ROESY were all acquired by using
standard procedures with a spectral width of about 4000 Hz in both
columns F1 and F2. HOHAHA and ROESY were acquired in the
phase-sensitive mode. The time domain matrix consisted of 256
points in t1 and 1024-2048 points in t2 with 32 acquisitions for
256-300 experiments in t1. Data sets were zero-filled to 512 points

1276 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 6
Bonnard et al.
in t1 prior to Fourier transformation to obtain a frequency domain
matrix of 512 × 1024-2048 real data points. Squared sine bell
apodization functions were used. The HOHAHA spectra were
recorded with a mixing time of 80 ms. ROESY spectra were
measured with mixing times of 150, 250, and 350 ms. Heteronuclear
correlated experiments were performed in 1H -detected mode using
the standard pulse programs HMQC and HMBC, with a spectral
width of about 20 000 Hz in F1 and 4000 Hz in F2. The time
domain matrix consisted of 256-512 points in t1 and 2048 points
in t2, with 128-200 acquisitions for 256 experiments in t1. Data
sets were zero-filled to 512 points in t1 prior to Fourier transforma-
tion to obtain a frequency domain matrix of 512 × 2048 real data
points. The evolution delay was set to optimize 140 Hz couplings
for HMQC and 8 Hz couplings for HMBC. Squared sine bell
apodization functions were used. The J-resolved spectra were
recorded using a standard pulse sequence to obtain a 4096 × 256
matrix in t2 and t1, respectively, with a spectral width of 50 Hz in
the t1 dimension.
Laxaphycin A (1). White amorphous solid; IR (CHCl3/CH2Cl2
1%) 3400, 3325, 2964, 2933, 2875, 1657, 1525 cm-1; UV (MeOH)
λmax 224 nm (ꢀ 13 820); FABMS m/z (rel.intensity in %), 1196.6
(100), 1180.0 (15), 588.2 (7), 475.1 (10), 445.0 (9), 383.0 (15),
362.0 (49); FABMS of acetylated laxaphycin A, m/z (rel. intensity
in %), 1360.5 (8), 1344.5 (65), 1322.5 (100), 785.4 (5), 672.3 (8),
559.2 (14), 446.0 (27), 430.0 (19), 386.0 (15), 370.0 (19), 310.0
(18); FAB MS/MS data are shown in Figure 2; NMR data (DMSO-
d6 at 318 K) are shown in Table 1.
Laxaphycin B (2). White amorphous solid; IR (CHCl3/CH2Cl2
1%) 3350, 3328, 3053, 2980, 2932, 1667, 1628, 1535 cm-1; UV
(MeOH) λmax 230 nm (ꢀ 2360); FABMS m/z (rel. intensity in %)
1395.6 (100), 1380.0 (22); FABMS of acetylated laxaphycin B m/z
(rel. intensity in %) 1605.5 (100), 1548.0 (21), 1563.0 (23), 810.4
(6), 621.0 (7), 593.2 (18), 521.0 (9), 455.0 (10), 370.0 (13), 357.0
(20), 355.0 (8), 340.0 (31), 310.0 (85); FAB MS/MS data are show
in Figure 3; NMR data (DMSO-d6 at 308 K) are shown in Table 2.
Laxaphycin B2 (3). White amorphous solid; IR (CHCl3/CH2-
Cl2 1%) 3334, 2960, 2930, 1672, 1642, 1530 cm-1; UV (MeOH)
λmax 230 nm (ꢀ 4810); FABMS, m/z (rel. intensity in %) 1424.6
(18), 1379.6 (100), 1363.6 (15); FAB MS/MS data are shown in
Figure 5; NMR data (DMSO-d6 at 318 K) are shown in Table 2.
Laxaphycin B3 (4). White amorphous solid; IR (CHCl3/CH2-
Cl2 1%) 3338, 3291, 1666, 1629, 1538 cm-1; UV (MeOH) λmax
230 nm (ꢀ 2440); FABMS m/z (rel. intensity in %) 1411.6 (100),
1395.5 (15), 1367.5 (10), 553.1 (8), 456.0 (18), 369.0 (60); FAB
MS/MS data are show in Figure 6; NMR data (DMSO-d6 at 308
K) are shown in Table 2.
imidooctanoate (128 mg of methyl 3-phtalimidodecanoate). 1H
NMR data of methyl 3-phtalimidooctanoate (CDCl3, 400 MHz):
δ 7.82 (2H, m), 7.71 (2H, m), 4.60 (1H, dd, J ) 9.8 Hz, 5.4 Hz),
3.58 (3H, s), 3.15 (1H, dd, J ) 9.8 Hz, 16.0 Hz), 2.75 (1H, dd, J
) 5.4 Hz, 16.0 Hz), 2.05 (1H, m), 1.7 (1H, m), 1.25 (6H, m), 0.8
1
(3H, m). H NMR data of methyl 3-phtalimidodecanoate (CDCl3,
400 MHz): δ 7.81 (2H, m), 7.69 (2H, m), 4.61 (1H, m), 3.59 (3H,
s), 3.15 (1H, dd, J ) 9.6 Hz, 16.0 Hz), 2.75 (1H, dd, J ) 5.2 Hz,
16.0 Hz), 2.04 (1H, m), 1.68 (1H, m), 1.22 (8H, m), 0.8 (3H, m).
13C NMR data of methyl 3-phtalimidooctanoate (CDCl3, 100
MHz): δ 172.0, 168.5, 134.0, 132.5, 123.0, 52.0, 48.0, 37.2, 32.5,
31.5, 26.0, 22.5, 14.0. 13C NMR data of methyl 3-phtalimidode-
canoate (CDCl3, 100 MHz): δ 171.5, 168.2, 134.0, 133.0, 123.0,
51.6, 48.0, 37.2, 32.5, 31.5, 25.8, 22.5, 22.4, 22.1, 13.8. The methyl
3-phtalimidooctanoate (methyl 3-phtalimidodecanoate) product was
hydrolyzed (3 N KOH, 50 °C, 4 h), and the resultant solution was
acidified and extracted with CHCl3. The CHCl3 layer was washed
with water and dried with MgSO4 and evaporated to give DL-â-
amino octanoic and decanoic acid contaminated by a small amount
of phtalic acid. 1H NMR data of DL-â-amino octanoic acid (acetone-
d6, 400 MHz): δ 4.40 (1H, m), 2.75 (1H, dd, J ) 6.0 Hz, 17.0
Hz), 2.55 (1H, dd, J ) 9.0 Hz, 17.0 Hz), 1.65 (2H, m), 1.50 (1H,
m), 1.42 (1H, m), 1.30 (4H, m), 0.88 (3H, t, J ) 7.0 Hz). 1H NMR
data of DL-â-amino decanoic acid (acetone-d6, 400 MHz): δ 4.41
(1H, m), 2.74 (1H, dd, J ) 5.8 Hz, 16.0 Hz), 2.55 (1H, dd, J ) 8.9
Hz, 16.0 Hz), 1.65 (2H, m), 1.50 (1H, m), 1.42 (1H, m), 1.28 (6H,
m), 0.87 (3H, t, J ) 7.0 Hz). 13C NMR data of DL-â-amino octanoic
acid (acetone-d6, 100 MHz): δ 173.5, 47.5, 40.0, 35.0, 32.5, 26.0,
23.0, 14.5. 13C NMR data of DL-â-amino decanoic acid (acetone-
d6, 100 MHz): δ 174.0, 48.0, 40.0, 35.0, 32.4, 26.5, 23.5, 14.4.
These â-amino octanoic and decanoic acids were reacted directly
with FDAA for HPLC analysis. L-â-Amino acids were eluted before
D isomers (cf. Gerwick).
Synthesis of D-N-Methylisoleucine and D-allo-N-Methyliso-
leucine (Freidinger’s Method). Fmoc-D-Ile (0.37 g; Fmoc-D-allo-
Ile) was suspended in 100 mL of toluene, and 0.24 g of
paraformaldehyde and a catalytic amount of p-toluene sulfonic acid
were added. The mixture was refluxed for 30 min with azeotropic
water removal. The solution was cooled, washed with 1 N aqueous
NaHCO3 (3 × 25 mL), dried over MgSO4, and concentrated in
vacuo to yield 370 mg of a colorless oil. The oxazolidinone was
dissolved in 3 mL of CH2Cl2, 3 mL of trifluoroacetic acid, and 0.8
mL of triethylsilane were added. The solution was stirred overnight
at room temperature, followed by concentration in vacuo to an oil.
The residue was dissolved in CH2Cl2 and reconcentrated three times.
The resultant crystalline product was washed with a mixture of
10% ether in pentane and dried, and the fluoromethyloxycarbonyl
group was removed with piperidine in CH2Cl2 to yield D-N-MeIle
Acetylation of Laxaphycins A and B. An amount equal to 2
mg of each laxaphycin was mixed with a few drops of pyridine
and acetic anhydride overnight at room temperature. After evapora-
tion, the reaction mixture was purified on a silica gel column with
CH2Cl2/MeOH 90:10 (v/v).
1
and D-allo-N-MeIle. H NMR data of D-N-MeIle (DMSO-d6, 400
MHz): δ 8.80 (1H, br), 8.36 (1H, br), 3.80 (1H, m), 2.50 (3H, m),
1.84 (1H, br), 1.40 (1H, m), 1.21 (1H, m), 0.80 (6H, m). 1H NMR
data of D-allo-N-MeIle (DMSO-d6, 400 MHz): δ 8.90 (1H, br),
8.80 (1H, br), 3.90 (1H, m), 2.60 (3H, m), 1.90 (1H, m), 1.50 (1H,
m), 1.30 (1H, m), 0.90 (3H, t, J ) 6 Hz), 0.89 (3H, d, J ) 7.2 Hz).
13C NMR data of D-N-MeIle (DMSO-d6, 100 MHz): δ 169.3, 64.2,
35.0, 32.2, 26.0, 14.2, 11.7. 13C NMR data of D-allo-N-MeIle
(DMSO-d6, 100 MHz): δ 168.9, 63.7, 34.7, 31.7, 25.5, 13.7, 11.2.
D-N-MeIle and D-allo-N-MeIle were derivatized directly with FDAA
for HPLC analysis.
Synthesis of (2R,3S)-3-Hydroxyleucine and (2R,3R)-3-Hy-
droxyleucine (Adapted from Schollkopf’s Method). An amount
equal to 3.75 mL of a 1.6 M solution of butyllithium in hexane
was added at -70 °C to 1.03 g of (2S)-2,5-dihydro-3,6-dimethoxy-
2-isopropylpyrazine [(+)-Schollkopf’s reagent] in anhydrous THF
(15 mL). After 30 min of stirring at -70 °C, 0.6 mL of isobutanal
in THF (5 mL) was added and stirring was continued for 8 h. The
reaction mixture was allowed to warm up to 0 °C, neutralized by
a 1 mol NH4Cl solution, and extracted with diethyl ether (3 × 30
mL). The combined ether layers were dried over magnesium sulfate
and concentrated under reduced pressure to yield a crude extract
that was purified by silica gel column chromatography (eluent:
Synthesis of DL-â-Amino Octanoic and Decanoic Acid (as
Described Previously by Gerwick et al.). Malonic acid (2.5 g)
was dissolved in 6 mL of dried pyridine, and then 2.5 mL of hexanal
(3.2 mL of octanal) and 0.25 mL of distilled pyrrolidine were added.
After refluxing for 30 min, the products were poured into 100 mL
of chilled diluted HCl solution. The solution was extracted with
ether (3 × 100 mL), and the ether layer was washed with distilled
water, dried over MgSO4, and evaporated to give 2.4 g of (E)-
octa-2-enoic acid (3 g of (E)-deca-2-enoic acid). The acids (1 g) in
methanol (30 mL) were methylated by refluxing for 12 h with
Amberlite IR-120 (H+) resine. After filtration, the reaction mixture
was concentrated in vacuo to give a residue that was dissolved in
30 mL of CH2Cl2. The CH2Cl2 layer was washed with NaHCO3
solution, water, and brine, dried with MgSO4, and evaporated. Each
ester (0.50 g) was mixed with 0.62 g of phtalimide in 20 mL of
dried pyridine. The mixture was refluxed for 24 h in the presence
of EtONa (0.34 g). After cooling to room temperature, the reaction
was poured into ice water and the resultant solution was extracted
with CHCl3 (3 × 30 mL). The extract was fractionated by silica
gel column chromatography to yield 163 mg of methyl 3-phtal-

Tumor Cell Proliferation of Laxaphycins
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 6 1277
pentane/ethylacetate 90:10 v/v). We obtained 420 mg of (1S,2′R,5′S)-
carbonyl adduct (yield 40.8%) and 62 mg of (1R,2′R,5′S)-carbonyl
adduct (yield 6%). The carbonyl adducts were hydolyzed following
the general procedure as described for standard peptide hydrolysis.
The (1S,2′R,5′S)-carbonyl adduct gave a mixture of (2R,3S)-3-
hydroxyleucine and L-valine, and the (1R,2′R,5′S)-carbonyl adduct
gave a mixture of (2R,3R)-3-hydroxyleucine and L-valine, which
and traces of HCl were removed from the residual hydrolysate by
repeated evaporation from water. The residue was chromatographed
on a 35 × 1 cm column of Dowex 50 × 2-400 resin (200-400
mesh, hydrogen form) by eluting with a linear gradient between
0.01 N HCl (150 mL) to 3 N HCl (150 mL). Fractions were pooled
on the basis of Rf on silica gel TLC (n-BuOH/HOAc/H2O, 4:1:1,
v/v/v, ninhydrin detection) and vacuum-dried to give the following
amino acids as their hydrochloride salts in order of elution:
3-hydroxyaspartic acid (0.7 mg), L-threonine (2 mg), L-glutamic
acid + L-alanine + maleic acid or fumaric acid from dehydration
of 3-hydroxyaspartic acid (10.6 mg), (2R,3S)-3-OHLeu + maleic
acid or fumaric acid (2.1 mg), (2S,3S)-3-OHLeu + L-proline (7.5
mg), L-proline + L-valine (3.4 mg), L-N-methylisoleucine +
D-leucine (7.8 mg). Identification of these amino acids was
established by HPLC analysis with Marfey’s derivatization. 3-Hy-
1
were used as the standard for amino acid analysis. H NMR data
of (1S,2′R,5′S) and (1R,2′R,5′S)-carbonyl adducts (CDCl3, 400
MHz): δ 4.06 (1H, dd, J ) 1.6 Hz, 3.2 Hz), 3.96 (1H, dd, J ) 3.2
Hz, 3.2 Hz), 3.69 (3H, s), 3.65 (3H, s), 3.59 (1H, br), 2.22 (1H,
dsept., J ) 3.2 Hz, 6.8 Hz), 1.97 (1H, dsept., J ) 1.6 Hz, 6.8 Hz),
1.67 (1H, br), 1.00 (6H, d, J ) 6.8 Hz), 0.95 (3H, d, J ) 6.8 Hz),
0.67 (3H, d, J ) 6.8 Hz). 13C NMR data of (CDCl3, 100 MHz): δ
165.62, 162.48, 77.36 (C1S) or 77.01 (C1R), 60.74, 57.21, 52.53,
52.42, 31.88, 30.88, 19.37, 18.97, 16.74. 1H NMR data of (2R,3S)-
3-hydroxyleucine (DMSO-d6, 400 MHz): δ 3.88 (1H, d, J ) 2.4
Hz), 3.54 (1H, dd, J ) 2.4 Hz, 8.0 Hz), 3.36 (1H, br), 1.75 (1H,
1
droxyaspartic acid was found threo by H NMR comparison with
standard LD-threo-3-hydroxyaspartic acids. The δ value for LD-
threo-3-hydroxyaspartic acids (DCl + internal dioxane reference
at 3.75 ppm, 400 MHz): HR ) 4.674 ppm and Hâ ) 5.045 ppm.
The δ value for LD-erythro-3-hydroxyaspartic: HR ) 4.770 ppm
and Hâ ) 4.915 ppm (obtained from hydrolysis of threo isomers
at 120 °C during 20 h in HCl 6 N under an O2 atmosphere).
Measurement of the CD of this amino acid indicated D-configuration
according to the literature: [θ]200 ) -3400; [θ]216 ) -17 000;
[θ]239 ) 0 (c 0.0117, 0.5 N HCl) [value for L-threo-3-hydroxyas-
partic acid:40 [θ]200 ) +5320; [θ]205 ) +6970; [θ]245 ) 0 (c 0.0652,
0.5 N HCl)].
Derivatization of Amino Acids with 1-Fluoro-2,4-dinitrophe-
nyl-5-L-alanine Amide (FDAA). For the FDAA (Marfey’s reagent)
derivatization procedure, the previously obtained crude hydrolysate
or small amounts of standard free amino acid were solubilized in
50 µL of a mixture of H2O/acetone and mixed with 100 µL of a
1% solution of FDAA in acetone. Sodium bicarbonate solution (1
mol; 20 µL) was added to this mixture, and the resultant solution
was heated at 40 °C for 1 h and allowed to cool. After addition of
10 µL of 2 mol HCl, the resulting solution was evaporated,
dissolved in 0.5 mL of DMSO, and then analyzed by HPLC. The
following conditions were used: solvent A, 0.05 mol Et3N in water,
adjusted at pH 3 with H3PO4, with 5% of solvent B; solvent B,
acetonitrile/methanol (40:60 v/v); gradient with flow rate at 0.8
mL/min, solvent B from 30% to 40% in 20 min and from 40% to
100% in 30 min; column, Interchim Spherisorb OD2 5µ, 250 mm
× 4 mm; UV detector at 340 nm. The peaks were identified by
co-injection with a DL-mixture of standard amino acids. This
procedure established the presence of (2S,4R)-4-OHPro, L-Hse, Gly,
L-Ile, or L-allo-Ile, L-Leu, D-Phe, D-Leu, D-Ile, or d-allo-Ile, D-âAoc
for laxaphycin A; (2R,3R)-3-OHAsp, L-Thr, L-Glu, L-Ala, L-Pro,
(2S,3S)-3-OHLeu, L-Val, (2R,3S)-3-OHLeu, L-N-MeIle, D-Leu,
D-âAde for laxaphycin B; (2R,3R)-3-OHAsp, L-Thr, L-Glu, L-Ala,
L-Pro, L-Val, (2R,3S)-3-OHLeu, L-N-MeIle, D-Leu, D-âAde for
laxaphycin B2; (2R,3R)-3-OHAsp, L-Thr, L-Glu, L-Ala, (2S,4R)-
4-OHPro, (2S,3S)-3-OHLeu, l-Val, (2R,3S)-3-OHLeu, L-N-MeIle,
D-Leu, D-âAde for laxaphycin B3.
Derivatization of Amino Acids with Orthophtaldialdehyde
and 2-Mercapto-ethanol (OPA-MCE). An amount equal to 250
µL of a 10 nmol/mL H2O solution of peptide hydrolysate or standard
amino acid was mixed with 5 µL of internal standard (N-acetyl-
cysteine) and 400 µL of iodo acetic acid (0.4 mol, pH 11.5). To
this solution were added 10 µL of the OPA-MCE solution (125
mg of OPA in 2.5 mL of MeOH and 125 mL of MCE) and 10 µL
of H3BO3 (0.8 mol, pH 12.5). After 2 min, the mixture was injected
in the HPLC system. The following conditions were used: solvent
A, sodium acetate/methanol/tetrahydrofurane (80:19.2:0.8 v/v/v);
solvent B, sodium acetate/methanol (20:80 v/v), pH 10.5; gradient
with flow rate at 0.8 mL/min, solvent B from 0% to 10% in 3 min,
first step at 10% during 12 min, then from 10% to 14% in 3 min,
second step at 14% during 5 min, from 14% to 50% in 5 min,
third step at 50% during 4 min and from 50% to 100% in 18 min;
column, C18 5µ, 250 mm x 4 mm with a 25 mm x 4 mm precolumn,
temperature at 38.2 °C; fluorimetric detection (λex 360 nm, λem 455
nm). Retention times (min) are given in parentheses: from standard
1
m), 0.91 (3H, d, J ) 6.8 Hz), 0.85 (3H, d, J ) 6.8 Hz). H NMR
data of (2R,3R)-3-hydroxyleucine (DMSO-d6, 400 MHz): δ 3.52
(1H, br), 3.37 (1H, br), 3.36 (1H, br), 1.86 (1H, m), 0.91 (3H, d,
1
J ) 6.8 Hz), 0.96 (3H, d, J ) 6.8 Hz). H NMR data of L-valine
(DMSO-d6, 400 MHz): δ 3.70 (1H, d, J ) 4.0 Hz), 2.18 (1H, m),
0.96 (3H, d, J ) 6.8 Hz), 0.98 (3H, d, J ) 6.8 Hz). 13C NMR data
of (2R,3S)-3-hydroxyleucine (DMSO-d6, 100 MHz): δ 170.22,
74.23, 54.84, 29.90, 19.22, 18.6. 13C NMR data of (2R,3R)-3-
hydroxyleucine (DMSO-d6, 100 MHz): δ 170.05, 74.85, 54.85,
29.90, 19.50, 18.16. 13C NMR data of L-valine (DMSO-d6, 100
MHz): δ 170.09, 57.26, 29.02, 18.16, 17.70.
Synthesis of (2S,3R)-3-Hydroxyleucine and (2S,3S)-3-Hy-
droxyleucine. These compounds were synthesized as described
above from (2R)-2,5-dihydro-3,6-dimethoxy-2-isopropylpyrazine
[(-)-Schollkopf’s reagent]: yield 390 mg (54.5%) of (1R,2′S,5′R)-
carbonyl adduct and 64 mg (8.9%) of (1S,2′S,5′R)-carbonyl adduct.
1H NMR data of (1R,2′S,5′R) and (1S,2′S,5′R)-carbonyl adducts
(CDCl3, 400 MHz): δ 4.03 (1H, dd, J ) 3.6 Hz, 3.2 Hz), 3.93
(1H, dd, J ) 3.2 Hz, 3.2 Hz), 3.66 (3H, s), 3.62 (3H, s), 3.56 (1H,
br), 2.19 (1H, dsept., J ) 3.2 Hz, 6.8 Hz), 1.95 (1H, dsept., J )
1.2 Hz, 6.8 Hz), 1.74 (1H, br), 0.98 (3H, d, J ) 6.8 Hz), 0.97 (3H,
d, J ) 6.8 Hz), 0.93 (3H, d, J ) 6.8 Hz), 0.65 (3H, d, J ) 6.8 Hz).
13C NMR data of (1R,2′S,5′R) and (1S,2′S,5′R)-carbonyl adducts
(CDCl3, 100 MHz): δ 165.51, 162.42, 77.32 (C1R) or 77.15 (C1S),
1
60.67, 57.21, 52.44, 52.35, 31.79, 30.80, 19.30, 19.00, 16.67. H
NMR data of (2S,3R)-3-hydroxyleucine (DMSO-d6, 400 MHz): δ
13.60 (1H, br), 3.90 (1H, d, J ) 2.6 Hz), 3.54 (1H, dd, J ) 2.6
Hz, 8.8 Hz), 3.36 (1H, br), 1.75 (1H, m), 0.91 (3H, d, J ) 6.4 Hz),
0.85 (3H, d, J ) 6.4 Hz). 1H NMR data (2S,3S)-3-hydroxyleucine
(DMSO-d6, 400 MHz): δ 3.67 (1H, d, J ) 2.0 Hz), 3.37 (1H, br),
3.36 (1H, br), 1.86 (1H, m), 0.95 (3H, d, J ) 6.4 Hz), 0.88 (3H, d,
1
J ) 6.4 Hz). H NMR data of d-valine (DMSO-d6, 400 MHz): δ
13.60 (1H, br), 3.71 (1H, br), 2.19 (1H, m), 0.97 (3H, d, J ) 7.2
Hz), 0.95 (3H, d, J ) 7.2 Hz). 13C NMR data of (2S,3R)-3-
hydroxyleucine (DMSO-d6, 100 MHz): δ 170.22, 74.17, 54.90,
29.90, 19.20, 18.5. 13C NMR data of (2S,3S)-3-hydroxyleucine
(DMSO-d6, 100 MHz): δ 170.19, 74.96, 54.84, 29.88, 19.27, 18.45.
13C NMR data of D-valine (DMSO-d6, 100 MHz): δ 170.19, 57.22,
29.00, 18.20, 17.70.
Hydrolysis of Laxaphycins. For standard hydrolysis, laxaphycin
A, laxaphycin B, laxaphycin B2, or laxaphycin B3 (0.5 to 1.2 mg)
in 1 mL 6 N HCl was heated at 120 °C for 15 h in a sealed vial
under nitrogen atmosphere. Microwave hydrolysis was also per-
formed by using a Teflon flask and heating for 10 min to decreased
degradation of â-hydroxy-amino acid and â-amino acid. The cooled
reaction mixture was evaporated to dryness, and traces of HCl were
removed from the residual hydrolysate by repeated evaporation from
water. All the workup procedure was made under nitrogen
atmosphere.
Isolation of 3-Hydroxyaspartic Acid from Laxaphycin B. An
amount equal to 21 mg of laxaphycin B was dissolved in 4 mL 6
N HCl and heated at 120 °C for 15 h in a sealed vial under nitrogen
atmosphere. The cooled reaction mixture was evaporated to dryness,

1278 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 6
Bonnard et al.
hydrolysis of laxaphycin A, Hse (17.30), Gly (20.18), Phe (41.99),
Ile (44.68), allo-Ile (44.90), Leu (45.63), âAoc (49.59).
Ho Assay. On the day of assay, plates were thawed at room
temperature for 10 min and 100 µL of a 30 µg/mL solution of
Hoechst 33342 in hypersaline buffer containing 10 mmol Tris HCl
pH 7.4, 1 mmol EDTA, and 2 mmol NaCl was added to each well.
After a 1 h incubation at 37 °C under 5% CO2, fluorescence was
measured at 360/460 nm. In the assay conditions, the DNA content
measured with Ho 33342 is proportional to the biomass content.
IC50 values were calculated as above. Results are the mean ( S.D.
of two independent assays done in triplicate.
Evaluation of Antiproliferative Activity on Lymphoblastic
Cell Lines. The acute lymphoblastic leukemia cell line CCRF-CEM,
the subline CEM/VLB100 selected for resistance to vinblastine, and
the subline CEM/VM-1 selected for resistance to teniposide (VM-
26) were obtained from Dr. W. T. Beck, St. Jude Children’s
Research Hospital, Memphis, Tennessee. Cells were grown in
plastic tissue culture flasks containing RPMI 1640 medium
supplemented with 10% fetal calf serum (v/v) and antibiotics
(penicillin 1000 U/mL; streptomycin 100 mg/mL) and maintained
at 37 °C in a humidified atmosphere containing 5% carbon dioxide.
Serial dilutions of laxaphycins were prepared in the culture medium.
The drug at the appropriate concentration was added to cell cultures
(2 × 105 cells/mL) for 3 days without renewal of the medium. Cells
surviving the treatment were then counted by means of a Coulter
counter model ZME. Assays were carried out in triplicate, and the
results were averaged. The concentration of drugs required to inhibit
growth of cells by 50% (IC50) in 72 h was determined for each
cell line. Cross-resistance was calculated by dividing the IC50
obtained with the resistant cell line by that measured for the parent
cell line. The interaction index (γ) was calculated by the following
formula: γ ) (IC50 of B combined/IC50 of B alone). As indicated
in ref 38, when one of the two compounds has no efficacy, γ is the
ratio of IC50s of the active component in the two situations, with
and without the inactive component. Theoretically, an interaction
index of 1 indicates an additive effect; greater than 1 means
antagonism and any value less than 1 denotes synergism.
Acknowledgment. This work was supported in part by
research grants from INSERM and FNCLCC (Fed. Nat. des
Centres de Lutte Contre le Cancer). I.B. was supported by La
Ligue Nationale Contre le Cancer. The authors are indebted to
Mr. and Mme. Prome´ for accurate mass measurements. The
acute lymphoblastic leukemia cell line CCRF-CEM was ob-
tained from Dr. W. T. Beck, St. Jude Children’s Research
Hospital, Memphis, Tennessee.
References
(1) Burja, A. M.; Banaigs, B.; Abou-Mansour, E.; Burgess, J. G.; Wright,
P. C. Marine cyanobacteriasA prolific source of natural products.
Tetrahedron 2001, 57, 9347-9377.
(2) Moon, S. S.; Chen, J. L.; Moore, R. E.; Patterson, G. M. L.
Calophycin, a fungicidal cyclic decapeptide from the terrestrial blue-
green alga Calothrix fusca. J. Org. Chem. 1992, 57, 1097-1103.
(3) Gregson, J. M.; Chen, J. L.; Patterson, G. M. L.; Moore, R. E.
Structure of puwainaphycins A-E. Tetrahedron 1992, 481, 3727-
3734.
(4) Moore, R. E.; Bornemann, V.; Niemczura, W. P.; Gregson, J. M.;
Chen, J. L.; Norton, T. R.; Patterson, G. M. L.; Helms, G. L.
Puwainaphycin C, A cardioactive cyclic peptide from the blue-green
alga Anabaena BQ-16-1. Use of two-dimensional 13C-13C and 13C-
15N correlation spectroscopy. J. Am. Chem. Soc. 1989, 111, 6128-
6132.
(5) Shin, H. J.; Murakami, M.; Matsuda, H.; Yamaguchi, K. Microviridins
D-F, serine protease inhibitors from the cyanobacterium Oscillatoria
agardhii (NIES-204). Tetrahedron 1996, 24, 8159-8168.
(6) Ishida, K.; Matsuda, H.; Murakami, M.; Yamaguchi, K. Kawagu-
chipeptin A, a novel cyclic undecapeptide from cyanobacterium
Microcystis aeruginosa (NIES-88). Tetrahedron 1996, 52, 9025-
9030.
(7) Ishida, K.; Matsuda, H.; Murakami, M.; Yamaguchi, K. Kawagu-
chipeptin B, an antibacterial cyclic undecapeptide from the cyano-
bacterium Microcystis aeruginosa. J. Nat. Prod. 1997, 60, 724-726.
(8) Helms, G. L.; Moore, R. E.; Niemczura, W. P.; Patterson, G. M. L.;
Tomer, K. B.; Gross, M. L. Scytonemin A, a novel calcium antagonist
from a blue-green alga. J. Org. Chem. 1988, 53, 1298-1307.
(9) Permanent, I.; Carmeli S. Schizotrin A, a novel antimicrobial cyclic
peptide from a cyanobacterium. Tetrahedron Lett. 1994, 36, 8473-
8476.
(10) Nogle, L. M.; Marquez, B. L.; Gerwick, W. H. Wewakazole, a novel
cyclic dodecapeptide from a Papua New Guinea Lyngbya majuscula.
Org. Lett. 2003, 5 (1), 3-6.
(11) Jaki, B.; Zerbe, O.; Heilmann, J.; Sticher O. Two novel cyclic peptides
with antifungal activity from the cyanobacterium Tolypothrix bys-
soidea (EAWAG 195). J. Nat. Prod. 2001, 64, 154-158.
(12) Horgen, F. D.; Yoshida, W. Y.; Scheuer, P. J. Malevamides A-C,
new depsipeptides from the marine cyanobacterium Symploca la-
eteViridis. J. Nat. Prod. 2000, 63, 461-467.
Evaluation of Antiproliferative Activity on Human Carci-
noma and Melanoma Cell Lines. Human MCF7 breast adeno-
carcinoma, PC3 prostate adenocarcinoma, A-549 nonsmall-cell lung
(NSCL) carcinoma, PA1 ovarian adenocarcinoma, DLD-1 colon
adenocarcinoma cell lines, and L929 murine immortalized cells
were purchased from the European Collection of Cell Cultures
(ECACC, Salibury, United Kingdom). The M4Beu melanoma line
was established in the laboratory of Dr. J. F. Dore´ (INSERM U-218,
Lyon, France) from metastatic biopsy specimen and has been
maintained in cell culture for almost 15 years. Normal human
fibroblasts were purchased from Biopredic International (Rennes,
France) as a frozen culture. All cells were cultured as previously
described41 in Glutamax Eagle’s minimum essential medium
supplemented with Earle’s salts (Gibco-BRL, Paisley, U.K.), fetal
calf serum, vitamins, sodium pyruvate, nonessential amino-acids,
L-glutamine, and gentamycin base and maintained at 37 °C in a
humidified atmosphere containing 5% carbon dioxide. Cell survival
was assayed after a 48 h continuous contact with the solvent (DMSO
0.5%), with or without laxaphycin A (0-50 µmol) or laxaphycin
B (0-20 µmol) by two cytotoxicity assays (RRT and Hoechst
33342) done in 96-well microplates (Nunclon, Nunc, Roskilde,
Denmark), as described in ref 40. The RRT measures the metabolic
activity of cultures, while the Hoechst assay (Ho) determines the
DNA cellular content after cell lysis. In each assay, the cells were
seeded with 5 × 103 cells per well and allowed to adhere for 16 h
before treatment with laxaphycin A or laxaphycin B alone or in
combination in 0.5% DMSO.
RRT Assay. At the end of treatment (48 h), 150 µL of a 25
µg/mL solution of resazurin in MEM was added to each well. The
plates were incubated for 1 h at 37 °C under 5% CO2 before
measurement of fluorescence at 530/590 nm with a fluorescent
Ascent FL (Labsystems). Under the conditions used, the fluores-
cence was proportional to the number of living cells in the well.
IC50 values were calculated from the concentration-dependent
survival curve defined as the fluorescence in experimental wells
compared with fluorescence in control wells, after substraction of
blank values. Results are the mean ( S.D. of two independent
assays done in triplicate.
(13) Frankmo¨lle, W. P.; Larsen, L. K.; Caplan, F. R.; Patterson, G. M.
L.; Knu¨bel, G.; Levine, I. A.; Moore, R. E. Antifungal cyclic peptides
from the terrestrial blue-green alga Anabaena laxa. 1. Isolation and
biological properties. J. Antibiot. 1992, 45, 1451-1457.
(14) Frankmo¨lle, W. P.; Knu¨bel, G.; Moore, R. E.; Patterson, G. M. L.
Antifungal cyclic peptides from the terrestrial blue-green alga
Anabaena laxa. 2. Structures of laxaphycins A, B, D, and E. J.
Antibiot. 1992, 45, 1458-1466.
(15) The amino acids stereochemistry of laxaphycins A and B was not
established.
(16) Gerwick, W. H.; Mrozek, C.; Moghaddam, M. F.; Agarwal, S. K.
Novel cytotoxic peptides from the cyanobacterium Hormothamnion
enteromorphoides. 1. Discovery, isolation, and initial chemical and
biological characterization of the hormothamnins from wild and
cultured material. Experientia 1989, 45, 115-211.
(17) Gerwick, W. H.; Jiang, Z. D.; Agarwal, S. K.; Farmer, B. T. Total
structure of hormothamnin A, a toxic cyclic undecapeptide from the
tropical marine cyanobacterium Hormothamnion enteromorphoides.
Tetrahedron 1992, 48, 2313-2324.
After reading, the cells were prepared for intracellular DNA
quantitation with Ho 33342. Resazurin solution was eliminated with
an automatic microplate washer, and the cells were rinsed with PBS
(3-fold) and stored at -80 °C until Ho assay.

Tumor Cell Proliferation of Laxaphycins
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 6 1279
(18) Bonnard, I.; Rolland, M.; Francisco, C.; Banaigs, B. Total structure
and biological properties of laxaphycins A and B, cyclic lipopeptides
from the marine cyanobacterium Lyngbya majuscula. Lett. Pept. Sci.
1997, 289-292.
(19) MacMillan, J. B.; Ernst-Russell, M. A.; de Ropp, J. S.; Molinski, T.
F. Lobocyclamides A-C, lipopeptides from a cryptic cyanobacterial
mat containing Lyngbya conferVoides. J. Org. Chem. 2002, 67, 8210-
8215.
(20) The complete stereochemistry shown in the figure for lobocyclamides
A and B is slightly different from those announced in the text.
(21) Roepstorff, P.; Fohlhman, J. Proposal for a common nomenclature
for sequence ions in mass spectra of peptides. Biomed. Mass
Spectrom. 1984, 11, 601-606.
(22) Hunt, D. F.; Yates, J. R.; Shabanowitz, J.; Winston, S.; Hauer, C. R.
Protein sequencing by tandem mass spectrometry. Proc. Natl. Acad.
Sci. U.S.A. 1986, 83, 6233-6237.
(23) (a) Eckart, K.; Schwarz, H.; Tomer, K. B.; Gross, M. L. Tandem
mass spectrometry methodology for the sequence determination of
cyclic peptides. J. Am. Chem. Soc. 1985, 107, 6765-6769. (b) Eckart,
K.; Schwarz, H.; Chorev, M.; Gilon, C. Sequence determination of
N-terminal and C-terminal blocked peptides containing N-alkylated
amino acids and structure determination of these amino acid
constituents by using fast-atom-bombardment/tandem mass spec-
trometry. Eur. J. Biochem. 1986, 157, 209-216.
(30) Scho¨llkopf, U. Enantioselective synthesis of nonproteinogenic amino
acids via metalated bislactim ethers of 2,5-diketopiperazines. Tetra-
hedron 1983, 39, 2085-2091.
(31) Scho¨llkopf, U.; Nozulak, J.; Groth, U. Asymmetric synthesis via
heterocyclic intermediates-XXII. Enantioselective synthesis of R-alk-
enyl glycine methyl esters and R-alkenyl glycines (â,γ-unsaturated
amino acids). Tetrahedron 1984, 40, 1409-1417.
(32) Scho¨llkopf, U.; Groth, U.; Gull, M.-R.; Nozulak, J. Asymmetrische
synthesen u¨ber heterocyclische zwischenstufen-XVIII. Zur enanti-
oselektiven sunthese von (2R)-serin-methylestern oder (2R)-serinen
ausgehend vom bislactimether von cyclo-(-L-Val-Gly-). Liebigs Ann.
Chem. 1983, 1133-1151.
(33) Beulshausen, T.; Groth, U.; Scho¨llkopf, U. Asymmetric synthesis
via heterocyclic intermediates-XLV. Asymmetric synthesis of dias-
tereomerically and enantiomerically pure 3-substituted (2R,3S)-serine
methyl esters. Liebigs Ann. Chem. 1991, 1207-1209.
(34) Beck, W. T.; Mueller, T. J.; Tanzer, L. R. Altered surface membrane
glycoproteins in Vinca alkaloids resistant human leucemic lympho-
blasts. Cancer Res. 1979, 39, 2070-2076.
(35) Danks, M. K.; Yalowich, J. C.; Beck W. T. Atypical multiple drug
resistance in a human leucemic cell line selected for resistance to
teniposide (VM-26). Cancer Res. 1987, 47, 1297-1301.
(36) Fojo, T.; Bates, S. Strategies for reversing drug resistance. Oncogene
2003, 22, 7512-7523.
(37) Chan, L. M.; Lowes, S.; Hirst, B. H. The ABCs of drug transport in
intestine and liver: Efflux proteins limiting drug absorption and
bioavailability. Eur. J. Pharm. Sci. 2004, 21, 25-51.
(38) Tallarida, R. The interaction index: A measure of drug synergism.
Pain 2002, 98, 163-168.
(39) Gbankoto, A.; Vigo, J.; Dramane, K.; Banaigs, B.; A¨ına, E.; Salmon,
J. M. Cytotoxic effect of laxaphycins A and B on human lympho-
blastic cells (CCRF-CEM) using digitised videomicrofluorometry.
In ViVo 2005, 19, 577-582.
(40) Kato, T.; Hinoo, H.; Terui, Y.; Kikuchi, J.; Shoji, J. The structure of
katanosins A and B. J. Antibiot. 1988, 41 (6), 719-725.
(41) Barthomeuf, C.; Debiton, E.; Mshvildadze, V.; Kemertelidze, E.;
Balansard, G. In Vitro activity of hederacolchisid A1 compared with
other saponins from Hedera colchica against proliferation of human
carcinoma and melanoma cells. Planta Med. 2002, 68, 672-675.
(24) Ishikawa, K.; Niwa, Y.; Oishi, K.; Aoi, S.; Takaichi, T.; Wakayama,
S. Sequence determination of unknown cyclic peptide antibiotics by
Fast Atom Bombardment Mass Spectrometry. Biomed. EnViron. Mass
Spectrom. 1990, 19, 395-399.
(25) Marfey, P. Determination of -D-amino acids-. II. Use of a
bifunctional reagent, 1,5-difluoro-2,4-dinitrobenzene. Carlsberg Res.
Commun. 1984, 49, 591-596.
(26) Bru¨ckner, H.; Wittner, R.; Godel, H. Automated enantioseparation
of amino acids by derivatization with o-phtaldialdehyde and N-
acylated cysteines. J. Chromatogr. 1989, 476, 73-82.
(27) Colomines, J. C.; Pruski, A. Dosage des amines primaires par HPLC.
Personal communication.
(28) Freidinger, R. M.; Hinkle, J. S.; Perlow, D. S.; Arison, B. H. Synthesis
of 9-fluorenylmethoxycarbonyl-protected N-alkyl amino acids by
reduction of oxazolidinones. J. Org. Chem. 1983, 48, 77-81.
(29) Scho¨llkopf, U. Enantioselective synthesis of nonproteinogenic amino
acids. Top. Curr. Chem. 1983, 109, 65-84.
JM061307X