Synthesis, structural aspects and cytotoxicity of the natural cyclopeptides yunnanins A, C and phakellistatins 1, 10

Article abstract of DOI:10.1016/j.tet.2003.10.073

Yunnanins A and C, two cyclic heptapeptides occurring in the roots of Stellaria yunnanensis, and phakellistatins 1 and 10, a hepta- and an octacyclopeptide first isolated from marine sponges of the genus Phakellia, were efficiently synthesized using a combination of solid and solution-phase techniques. Structural analysis on the synthetic members of the yunnanin series showed that the synthetic sample of yunnanin A exhibited a configurational pattern at the Pro peptide linkages identical to the natural product (trans-Pro³, trans-Pro⁵), while yunnanin C was obtained as a complex mixture of geometric/conformational isomers; the major isomer (trans-Pro³) was indistinguishable from the natural cyclopeptide and co-occurred along with lower amounts of a mixture (1:1 ratio) of two different rotamers, both displaying cis geometry at the Pro³ linkage. In the phakellistatin series, the synthetic phakellistatin 1 (determined as cis-Pro¹, cis-Pro³, cis-Pro⁵) was identical to the natural one, while two different isomeric products of phakellistatin 10 could be obtained: a major one (trans-Pro¹, trans-Pro⁴, trans-Pro⁶) showing spectral properties superimposable with the natural metabolite, and a minor geometric isomer of the natural cyclopeptide. Interestingly, the synthetic cyclopeptides, although found to be chemically identical with their natural counterparts, did not display the same biological properties (in vitro cytotoxicity against a panel of cancer cell lines), leaving presently open the question whether or not the potent bioactivity reported in the literature should really be attributed to these natural cyclic peptides.

Full text of DOI:10.1016/j.tet.2003.10.073

Tetrahedron 59 (2003) 10203–10211
Synthesis, structural aspects and cytotoxicity of the natural
cyclopeptides yunnanins A, C and phakellistatins 1, 10
Assunta Napolitano,a Manuela Rodriquez,b Ines Bruno,a Stefania Marzocco,a Giuseppina Autore,a
Raffaele Riccioa and Luigi Gomez-Palomaa
,
*
a
`
Dipartimento di Scienze Farmaceutiche, Universita di Salerno, via Ponte Don Melillo, 84084 Fisciano (SA), Italy
`
Dipartimento Farmaco Chimico Tecnologico, Universita di Siena, via A. Moro 2, 53100 Siena, Italy
b
Received 13 June 2003; revised 20 October 2003; accepted 23 October 2003
Abstract—Yunnanins A and C, two cyclic heptapeptides occurring in the roots of Stellaria yunnanensis, and phakellistatins 1 and 10, a
hepta- and an octacyclopeptide first isolated from marine sponges of the genus Phakellia, were efficiently synthesized using a combination of
solid and solution-phase techniques. Structural analysis on the synthetic members of the yunnanin series showed that the synthetic sample of
yunnanin A exhibited a configurational pattern at the Pro peptide linkages identical to the natural product (trans-Pro3, trans-Pro5), while
yunnanin C was obtained as a complex mixture of geometric/conformational isomers; the major isomer (trans-Pro3) was indistinguishable
from the natural cyclopeptide and co-occurred along with lower amounts of a mixture (1:1 ratio) of two different rotamers, both displaying cis
geometry at the Pro3 linkage. In the phakellistatin series, the synthetic phakellistatin 1 (determined as cis-Pro1, cis-Pro3, cis-Pro5) was
identical to the natural one, while two different isomeric products of phakellistatin 10 could be obtained: a major one (trans-Pro1, trans-Pro4,
trans-Pro6) showing spectral properties superimposable with the natural metabolite, and a minor geometric isomer of the natural
cyclopeptide. Interestingly, the synthetic cyclopeptides, although found to be chemically identical with their natural counterparts, did not
display the same biological properties (in vitro cytotoxicity against a panel of cancer cell lines), leaving presently open the question whether
or not the potent bioactivity reported in the literature should really be attributed to these natural cyclic peptides.
q 2003 Elsevier Ltd. All rights reserved.
1. Introduction
belonging to the proline-rich cyclopeptides family. Recent
reports from our laboratory include our work on hymena-
mide C, a homodetic proline rich heptacyclopeptide,
showing an inhibitory effect on human neutrophil elastase
degranulation release. In this context, we described its solid-
phase synthesis,6 a detailed structural and pharmacological
study of the natural product itself along with a small library
of its Ala-modified analogues.7
Cyclic peptides and depsipeptides are very common among
natural products and exhibit a wide range of biological
activities.1 In particular, a number of compounds belonging
to this family of secondary metabolites have recently
showed to interfere with cell proliferation and differen-
tiation, thus attracting great interest as potential candidates
for the development of novel anticancer drugs.2 An
emerging class of bioactive cyclopeptides is represented
by ‘proline rich’ compounds, a series of natural products
occurring especially in the marine environment,3 but found
also in higher plants,4 named after their unusual high
content of proline residues, and characterized by remarkable
similarity in their amino acid composition and sequence
motif pattern.
The present investigation was directed toward the total
synthesis of other members of the family of proline rich
cyclopeptides: yunnanins A and C, two cyclic heptapeptides
isolated from the roots of Stellaria yunnanensis,8 and
phakellistatins 1 and 10, a hepta- and octacyclopeptide
respectively, first isolated from marine sponges of genus
Phakellia9 and then found to be also present in other
taxonomically unrelated sponge specimens (Fig. 1).10
As part of our on-going research program on bioactive
marine metabolites as potentially useful models in the
discovery of new and more effective pharmaceuticals,5 we
have recently focused our attention on several members
Greatly encouraged by the potent cell growth inhibitory
effects exhibited by the above members of proline rich
cyclopeptides, we decided to undertake their total synthesis
in order to increase the supply for further biological
investigation and, ultimately, to ascertain whether the
synthetic products may display the same biological proper-
ties exhibited by their natural counterparts. Interestingly,
other cases have recently been reported of closely related
Keywords: cyclopeptides; solid phase synthesis; marine natural products;
cytotoxic.
*
Corresponding author. Tel.: þ39-089-962811; fax: þ39-089-962652;
e-mail: gomez@unisa.it
0040–4020/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2003.10.073

10204
A. Napolitano et al. / Tetrahedron 59 (2003) 10203–10211
Figure 1. Chemical structures of natural products 1–4.
cyclopeptides being synthesized and whose biological
properties greatly differed from those found for the same
chemical entities obtained from natural sources. For
instance, Pettit and coworkers in their paper describing
phakellistatin 5 solution and solid-phase synthesis raised the
question whether it may be hypothesized that an unbearable
amount of an extremely potent cytotoxin must co-occur with
these natural cyclopeptides, in an attempt to explain the
chemical albeit not biological equivalence of natural and
synthetic samples.11
phakellistatin 1 (3), and to seven coupling–deprotection
cycles to build the N in–Boc protected linear octapeptide as
precursor of phakellistatin 10 (4). All the Fmoc-protected
amino acids were activated 0 by hydroxybenzotriazole/
O-(benzotriazol-1-yl)-N,N,N0,N -tetramethyluronium hexa-
fluorophosphate (HOBt/HBTU) in presence of N-methyl-
morpholine (NMM) as described in the general procedure
C; the progress of the amino acid coupling was checked
through the Kaiser test (the ninhydrin colorimetric test).
Fmoc deprotection before each coupling step was achieved
by treatment of the resin-anchored peptide with a 20%
solution of piperidine in N,N-dimethylformamide (DMF)
according to general procedure B. After each linear peptide
was obtained, the Fmoc protecting group was removed from
the N-terminal residue and the protected peptide was
cleaved from the resin by using a 2:2:6 acetic acid/2,2,2-
trifluoroethanol/dicholoromethane (AcOH/TFE/DCM)
solvent mixture. After HPLC analysis of the linear
precursors, the cyclization reaction was allowed to procee0d
in solution using O-(7-azabenzotriazol-1-yl)-N,N,N0,N -
tetramethyluronium hexafluorophosphate (HATU) and
DIEA in DCM; finally, after removing the side-chain
protecting groups with 95% aqueous trifluoroacetic acid
(TFA), purification on semipreparative RP-HPLC yielded
the final cyclopeptides.
2. Results and discussion
The synthesis of yunnanin A [cyclo-(Gly-Gly-Pro3-Phe-
Pro5-Gly-Tyr); (1)], yunnanin C [cyclo-(Tyr-Ser-Pro3-Gly-
Ile-Gly-Phe); (2)], phakellistatin 1 [cyclo-(Pro1-Tyr-Pro3-
Ile-Pro5-Ile-Phe); (3)], and phakellistatin 10 [cyclo-(Pro1-
Leu-Thr-Pro4-Ile-Pro6-Trp-Val); (4)] were accomplished
using the Fmoc/t-Bu chemistry and a 2-chlorotritylchloride
resin as solid support. The first Fmoc-protected amino acid
was anchored to the linker by diisopropylethylamine
(DIEA) treatment under anhydrous conditions, followed
by capping of unreacted trityl groups with methanol
according to general procedure A (see Section 3) (Fig. 2).
The resulting loading degree was determined by UV
spectrophotometric analysis using general procedure
D. The resin was then submitted to six coupling–
deprotection cycles to build the linear heptapeptides as
precursors of the cyclic yunnanin A (1), yunnanin B (2) and
A peculiar conformational feature of Pro-containing
peptides is the cis–trans geometric isomerism at the peptide
linkages connecting a proline residue with the preceding
amino acid in the sequence (X-Pro). A wealth of data

A. Napolitano et al. / Tetrahedron 59 (2003) 10203–10211
10205
Figure 2. Synthetic scheme followed to obtain cyclopeptides 1–4.
indicate that, in absence of structural constraints, Pro-
containing peptide segments tend to adopt the more stable
trans geometry at the Pro linkages, although often in times
small but detectable amounts of cis X-Pro linkages are
observed in dynamical equilibrium with the major trans
species. This peculiar structural aspect makes more
challenging the spectral analysis of proline-rich cyclo-
peptides, as two geometric isomers should be (at least in
principle) expected for each proline occurring in the
sequence; this kind of conformational equilibria become
especially crucial in those cases where the size of the
macrocycle permits greater flexibility of the peptide back-
bone, thus allowing the presence of multiple species slowly
interconverting on the NMR time scale.
which, on the basis of ROESY cross-peaks’ pattern,
possessed trans geometries at Gly-Pro3 and Phe-Pro5
linkages (ROESY cross-peak Ha-Gly/H2d-Pro3 Ha-
Phe/H2d-Pro5), resulting to be identical to the natural
yunnanin A8 (1H and 13C NMR spectra of synthetic and
natural samples were virtually superimposable; Table 1).
Yunnanin C, instead, was obtained as two main products,
the major one being indistinguishable from the natural
cyclopeptide,8 and featuring trans geometry at the Ser-Pro3
peptide bond (ROESY cross-peak Ha-Ser/H2d-Pro3)
(Table 2). The minor geometric isomer was identified as a
mixture of two conformational isomers, slowly inter-
converting on the NMR time scale, both possessing a cis
configuration at the Pro3 peptide linkage. As expected, the
ratio of conformers showed to be dependent on the
temperature of the sample. In fact, variable temperature
In particular, yunnanin A was obtained as single product

10206
A. Napolitano et al. / Tetrahedron 59 (2003) 10203–10211
Table 3. 1H and 13C NMR data for phakellistatin 1 (600 MHz, CD3OD)
Table 1. 1H and 13C NMR data for yunnanin A (600 MHz, CD3OD)
Position
1Ha
13Ca
Position
1Ha
13Ca
Position
1Ha
13Ca
Position
1Ha
13Ca
Gly1
a
NH
Gly2
a
NH
Pro3
a
Pro5
a
b
g
Pro1
a
b
g
d
Tyr
a
Pro5
a
b
g
3.67, 4.46
8.37
42.55
42.55
4.30
62.72
30.49
26.17
49.70
3.56
61.48
32.15
22.25
46.94
4.48
62.50
32.11
22.7
1.99, 2.30
2.01, 2.21
3.90, 4.44
1.38, 1.76
1.43, 1.62
3.22, 3.50
2.08, 2.5
1.68, 1.98
3.50, 3.61
3.67, 4.46
8.90
d
d
47.56
Ile6
a
Gly6
a
4.60
3.03, 3.07
54.05
37.27
3.98
1.96
0.83
1.18, 1.46
0.89
59.62
35.72
15.18
26.55
9.64
4.46
62.06
30.48
24.07
48.41
3.54, 4.23
43.42
b
b
CH3
g
CH3
Phe
a
b
g
d
1.71, 1.95
0.96, 1.68
3.57
Tyr
a
b
2,6
3,5
Pro3
a
b
g
6.66
6.83
114.79
130.87
4.75
2.88
3.11
57.81
39.76
3.81
Phe
a
b
4.52
59.30
31.32
22.34
47.87
4.65
2.94 3.06
53.10
39.45
5.18
3.17, 3.53
52.89
38.20
1.96, 2.08
1.87, 2.11
3.41
b
2,6
3,5
NH
7.09
6.69
8.08
131.33
116.25
d
2,6
3,5
4
7.22
7.28
7.34
129.40
127.32
128.80
2,6
3,5
4
7.40
7.34
7.26
7.88
130.67
129.73
128.25
3.60
Ile4
a
b
CH3
g
4.04
1.81
1.01
1.38, 1.77
1.00
58.32
37.36
14.27
26.17
10.03
NH
a
d in ppm.
CH3
NMR experiments in the 25–708C range showed that
coalescence of selected resonances could be achieved at
high temperature. The Gly4-Ha protons, observed as
splitted signals resonating at d 3.63–4.26 and d 3.96–4.02
at room temperature merged at 708C into a broad resonance
centered at d 3.63–4.26. In the phakellistatin series the
synthetic route yielded phakellistatin 1 as a predominant
isomer characterized by cis geometries at the Phe-Pro1, Tyr-
Pro3, Ile-Pro5 peptide bonds (ROESY cross-peaks: Ha-Phe/
Ha-Pro1, Ha-Tyr/Ha-Pro3, Ha-Ile/Ha-Pro5). Since the
spectral data of synthetic phakellistatin 1 were super-
imposable with those reported for the natural product9
(Table 3), we deduced that the same pattern of geometries
must apply to the natural product too, though such kind of
stereochemical assignment was not performed at the stage
of structural characterization. In the case of phakellistatin 10
synthesis, we obtained a major product, spectrally indis-
tinguishable from the natural one,9 identified as the all trans
isomer at Val-Pro1, Thr-Pro4, Ile-Pro6 peptide linkages
a
d in ppm.
(ROESY cross-peaks: Ha-Val/H2d-Pro1, Ha-Thr/H2d-Pro4,
Ha-Ile/H2d-Pro6, Table 4). The minor synthetic phakellis-
tatin 10 isomer exhibited the same electrospray mass
1
spectrum (ESIMS), but slightly different H NMR spectral
data. Unfortunately, owing to the very low amount
available, we were not able in this case to identify the
geometries at the peptide linkages of Pro residues.
Stereochemical integrity of the amino acid residues of the
synthetic cyclopeptides was secured by chiral GC–MS on
their hydrolysate (HCl 6 M, 1408C, for 12 h in a sealed
Table 4. 1H and 13C NMR data for phakellistatin 10 (600 MHz, CD3OD)
Position
1Ha
13Ca
Position
1Ha
13Ca
Pro1
a
b
g
Pro6
a
b
g
4.20
62.35
30.89
25.74
49.48
3.98
63.21
30.27
30.28
48.69
1.88, 2.31
1.99, 2.14
3.75, 4.09
1.90, 2.05
1.92, 2.07
3.42, 4.10
Table 2. 1H and 13C NMR data for yunnanin C (300 MHz, CD3OD)
d
d
Leu
a
b
Trp
a
b
Position
1Ha
13Ca
Position
1Ha
13Ca
3.73
54.99
37.26
26.02
20.92
23.55
4.40
3.43, 3.74
56.78
48.74
2.40, 1.79
1.58
0.95
Tyr
a
b
Ile5
a
b
CH3
g
CH3
NH
Gly6
a
g
4.41
3.03, 3.28
57.63
36.70
4.23
1.97
0.98
1.11, 1.52
0.90
8.44
59.07
36.69
15.98
25.53
10.71
CH3
CH3
Thr
a
2
3
4
5
6
7
8
Val
a
7.02
0.97
2,6
3,5
NH
Ser
a
7.11
6.77
8.08
130.87
116.12
5.09
4.31
1.19
58.08
69.22
19.43
7.60
7.07
7.14
7.37
118.90
119.66
122.24
112.24
b
CH3
Pro4
a
b
g
d
Ile
a
b
CH3
g
4.85
3.68, 3.79
7.53
54.46
62.73
3.61, 4.07
8.28
43.51
4.59
61.07
29.96
25.90
48.70
b
1.99, 2.32
1.96, 2.03
3.70, 3.83
4.70
2.27
0.82
0.94
58.24
29.60
19.08
20.10
NH
Pro3
a
b
g
NH
Phe
a
b
CH3
CH3
4.30
4.29
2.77, 2.98
57.24
37.14
1.98, 2.25
1.97, 2.12
3.81, 3.85
62.99
29.88
25.99
49.93
43.46
b
4.12
1.89
0.51
1.06, 1.49
0.85
56.41
36.27
15.14
d
2,6
3,5
4
7.04
7.27
7.24
129.61
129.30
127.47
Gly4
a
3.67, 4.24
CH3
10.33
a
a
d in ppm.
d in ppm.

A. Napolitano et al. / Tetrahedron 59 (2003) 10203–10211
10207
˚
tube), after derivatisation with N-tert-butyldimethylsylyl-N-
methyltrifluoroacetamide, in comparison with authentic
amino acid samples treated under the same experimental
conditions.
solid-phase reactions were synthesis grade (dried over 4 A
molecular sieves), CH3CN was HPLC grade. 2-chloro-
tritylchloride resin was purchased from Novabiochem
(loading capacity 1.08 mmol/g or 1.04 mmol/g). The
Fmoc-L-amino acids and the coupling reagents (HOBt,
HBTU, HATU) were supplied by Novabiochem or Fluka
and used without further purification. Solid-phase reactions
were carried out on a polypropylene ISOLUTE SPE column
on a VAC MASTER system (a manual parallel synthesis
device purchased from Stepbio, Bologna, Italy) and using
the Fmoc/t-Bu protocol. For quantification of the Fmoc
amino acids on the resin, absorbance at 301 nm was read
employing a Shimadzu UV 2101 PC spectrophotometer.
Melting points were determined on an Electrothermal 9100
Biological evaluation of the synthetic yunnanins and
phakellistatins against a minipanel of three cancer cell
lines showed cell growth inhibitory activity with IC50 values
always ,100 mg/ml (Table 5). A detailed analysis of these
IC50 values indicated that, on average, the synthetic
compounds were 100–1000-fold less active than their natural
counterparts which exhibited antiproliferative activity with
IC50 values ranging from 2.1 to 7.5 mg/ml. As stated before,
this result, albeit puzzling and presently rather difficult to
explain in a satisfactory fashion, was not fully unexpected.
Indeed, in the context of proline rich natural peptides there
have been other recent reports claiming strong disagreements
between the bioactivities of synthetic and natural samples.
apparatus. The H and 13C (1H–1H and H–13C) spectra
were recorded using a Bruker Avance 600 MHz spec-
1
trometer using CD3OD or d6-DMSO as solvents. The H
NMR temperature coefficients experiments were conducted
at 27, 30, 35, 40, 45, 50, 55, 60, 65 and 708C. A LCQ
Thermoquest mass spectrometer was used to record the
ESIMS spectra.
1
1
Table 5. In vitro anti-proliferative activity of proline rich cyclopeptides on
a minipanel of three cancer cell lines
Cyclopeptides
IC50 (M)a
J774.A1b
WEHI-164c HEK-293d
3.2. General procedures for solid-phase reactions
Yunnanin A
3.0£1023
2.22£1023
1.14£1023
5.6£1024
2.5£1024
2.9£1023
1.2£1023
1.7£1023
1.61£1023
1.23£1023
3.9£1023
5.7£1024
8.6£1024
(A) 2-Chlorotritylchloride resin loading: Fmoc-AA-OH
(0.5 equiv.), DIEA (2 equiv.) in DCM (10 ml/g of resin),
2 h; capping with 20 ml of DCM, MeOH, DIEA (17:2:1);
washings (1.5 min each): 3£3 ml of DCM, 2x3 ml of DMF,
2£3 ml of DCM.
Yunnanin C major product
Yunnanin C minor product
Phakellistatin 1
Phakellistatin 10 major product
Phakellistatin 10 minor product 1.12£1023
2.6£1023 9.98£1024
5.3£1024
3.8£1023
a
The IC50 value is the concentration of compound that affords 50%
reduction in cell growth (after 3 days incubation).
J774.A1, murine monocyte/macrophage cell line.
WEHI-164, murine fibrosarcoma cell line.
(B) Fmoc deprotection: 20% piperidine in DMF (3 ml),
1.5 min; 3 ml of 20% piperidine in DMF, 10 min; washings
(1.5 min each): 2£3 ml of DMF, 3£3 ml of DCM, 2£3 ml of
DMF.
b
c
d
HEK-293, human epithelial kidney cell line.
Indeed, our results match the biological data reported by
Pettit and co-workers.12 It remains still open whether these
cyclic peptides of natural origin may retain a chemically
undetectable amount of strongly active cytotoxic agents. An
alternative explanation has to do with the presence in these
metabolites of proline residues playing an important role in
defining their conformational motif. In this view, the
biological difference between natural and synthetic samples
might result from subtle conformational changes stemming
from diverse arrangements of the proline units. A recent
report, describing the case of phakellistatin 2,13 would
indeed favour this latter hypothesis. In our case, however,
we subjected to biological testing also minor (geometrical)
isomers without finding evidence of a relevant difference in
bioactivity oftheirs. On the other hand, it shouldbe considered
that our synthetic approach could not produce all the possible
conformations around the Pro linkages and therefore, at least
in principle, the hypothesis that among them there may be the
bioactive molecule responsible for the cytotoxicity observed
on natural samples, cannot be excluded.
(B1) Fmoc deprotection: 20% piperidine in DMF (3 ml),
5 min; washings (1.5 min each): 2£3 ml of DMF, 3£3 ml of
DCM, 2£3 ml of DMF.
(C) Peptide coupling conditions: HOBt (4 equiv.), HBTU
(4 equiv.), Fmoc-amino acid (4 equiv.) NMM (5 equiv.) in
DMF (500 ml/100 mg of resin), 2 h; washings (1.5 min
each): 3£3 ml of DMF, 3£3 ml of DCM.
(D) Spectrophotometric analysis of the Fmoc chromophore:
the assay was performed on duplicate samples. 0.4 ml of
piperidine and 0.4 ml of DCM were added to two dried
samples of the resin-bound peptide (,6 mg) in two 10 ml
volumetric flasks. The reaction was allowed to proceed for
30 min at room temperature in the sealed flasks. 1.6 ml of
MeOH were added and the solutions were diluted to 10 ml
volume with DCM. A reference solution was prepared in a
10 ml volumetric flask using 0.4 ml of piperidine, 1.6 ml of
MeOH and DCM to volume. The solutions were shook and
the absorbance of the samples versus the reference solution
was measured at 301 nm. The substitution degree (in mmol
of amino acid/g of resin) was calculated from the equation:
mmol/g¼(A301/7800)£(10 ml/g of resin).
3. Experimental
3.1. General methods
3.3. Synthesis of the linear side chain protected peptides
of yunnanin A and C, phakellistatin 1 and 10 (1–4)
Unless specified, solvents were reagent grade. They were
purchased from Aldrich or Fluka or Carlo Erba and were
used without further purification. DCM and DMF used for
2-Chlorotritylchloride resin was placed into a 25 ml

10208
A. Napolitano et al. / Tetrahedron 59 (2003) 10203–10211
polypropylene ISOLUTE column on a VAC MASTER
system, swelled for 1 h with 3 ml of DMF by a N2 stream
and then washed with 2x3 ml of DCM. The resin was loaded
with corresponding C-terminal Fmoc-AA-OH, in order to
obtain a lower substitution level, according to procedure
(A). Unreacted trityl groups were capped with methanol.
The resin was dried under vacuum over KOH and the
resulting substitution level was determined spectrophoto-
metrically according to the general procedure (D). After
Fmoc-AA-O-2ClTrt-Cl swelling (45 min with 3 ml of
DMF), Fmoc protecting group removal was obtained
according to general procedure (B). The resin was
subsequently submitted to the following series of coupling–
deprotection cycles: (i) peptide coupling according to
general procedure (C) with appropriate amino acid,
followed by Fmoc deprotection according to general
procedure (B1) to avoid the formation of DKP. The result
of the Fmoc removal was monitored according to general
procedure (D); (ii) peptide coupling according to general
procedure (C) with appropriate amino acid, followed by
determination, according to general procedure (D), of the
degree of substitution of the resin bound peptide. In the
cases of low DKP formation it was chosen to perform
the following peptide couplings according to general
procedure (C), while, when it was found an highest amount
of DKP formation, couplings were performed according to
the new loading level. The Fmoc deprotection was
performed according to general procedure (B); (iii) peptide
coupling according to general procedure (C) with appro-
priate amino acid, followed by Fmoc deprotection according
to general procedure (B), up to obtain the expected linear
protected peptide. The ninhydrin test was performed after
each amino acid coupling step and the coupling repeated if
necessary (Table 6).
The resin was removed from the polypropylene column,
washed with methanol and dried under vacuum over KOH
for 1 h. The overall resin-bound peptide was cleaved from
the solid support by treatment with an AcOH/TFE/DCM
(2:2:6) solution for 2 h under stirring. The cleavage mixture
was filtered off and the resin was washed 2 times with the
same solution. Hexane was added (15 times volume) and the
solution was evaporated, adding further hexane if necessary.
The crude peptide product was lyophilised and analysed by
RP-HPLC on
a Jupiter C-18 analytical column
˚
(250£4.60 mm, 5 mm, 300 A), using a 31 min gradient
from 5 to 100 of CH3CN/H2O (each containing 0.1% TFA)
at a flow rate of 1.0 ml/min and UV detection at 220 nm.
The HPLC analysis showed one peak that was identified as
the linear side-chain protected peptide on the basis of an
ESIMS experiment (Table 7).
Table 7. Analytical data of the linear protected peptides
HPLC Rt (min)
Mass dataa
1
2
3
4
Yunnanin A
Yunnanin C
Phakellistatin 1
Phakellistatin 10
16.33
18.57
20.35
24.03
750
852
903
1077
a
ESIMS, m/z for [MþH]þ.

A. Napolitano et al. / Tetrahedron 59 (2003) 10203–10211
10209
3.4. Cyclization of the linear side-chain protected
yunnanin A (1)
(Rt¼14.93 min; ESIMS, m/z 722 for [MþH]þ). The crude
cyclopeptides were then purified by semi-preparative RP-
HPLC on a Jupiter C-18 column (250£10.00 mm, 10 mm,
˚
The crude linear peptide (45.1 mg, 0.06 mmol) was
dissolved in DCM (4£1024 M) with HATU (45.6 mg,
0.12 mmol, 2 equiv.) and DIEA (26.1 ml, 0.15 mmol,
2.5 equiv.). The solution was stirred for 1 h on an ice bath
and then allowed to warm at room temperature and kept at
this temperature for 22.30 h. During this time DCM was
gradually added until to 2.4£1024 M to avoid side reactions
such as oligodimerization.
300 A), using a 37 min gradient from 20:80 to 35:65 of
CH3CN/H2O (each containing 0.1% TFA) at a flow rate of
5.0 ml/min and UV detection at 220 nm. The HPLC
purification yielded, as white solids, 9.7 mg
(yield¼28.7%) of trans-Pro3 cyclopeptide yunnanin C
(Rt¼12.48 min; ESIMS, m/z 722 for [MþH]þ; mp 2558C)
and 4.8 mg (yield¼14.2%) of the minor isomeric mixture of
cyclopeptide yunnanin C (Rt¼13.09 min; ESIMS, m/z 722
for [MþH]þ; mp 237–2398C).
The cyclization reaction was monitored via HPLC and
ESIMS spectra. After 22.30 h the solvent was removed.
Side-chain deprotection was obtained by treatment with
TFA/TIS/H2O¼95: 2.5: 2.5 (100 ml£1 mg of resin) for 1 h
under stirring. The cleavage mixture was evaporated and
lyophilised, yielding 82.7 mg of crude cyclopeptide. It was
analysed by RP-HPLC on a Jupiter C-18 analytical column
3.6. Cyclization of the linear side-chain protected
phakellistatin 1 (3)
The crude linear peptide (167.9 mg, 0.186 mmol) was
dissolved in DCM (7.79£1024 M) with HATU (140.1 mg,
0.368 mmol, 2 equiv.) and DIEA (80.8 ml, 0.465 mmol,
2.5 equiv.). The solution was stirred for 1 h on an ice bath
and then allowed to warm at room temperature and kept at
this temperature for 21.30 h. During this time DCM was
gradually added until to 4£1024 M to avoid side reactions
such as oligodimerization.
˚
(250£4.60 mm, 5 mm, 300 A), using a 56 min gradient from
25:75 to 60:40 of CH3CN/H2O (each containing 0.1% TFA)
at a flow rate of 1.0 ml/min and UV detection at 220 nm.
The HPLC analysis showed one main peak (Rt¼11.16 min;
ESIMS, m/z 676 for [MþH]þ) identified as the trans-Pro3,
trans-Pro5 cyclopeptide yunnanin A on the basis of ESIMS
and 1H NMR experiments. The crude cyclopeptide was then
purified by semi-preparative RP-HPLC on a Jupiter C-18
The cyclization reaction was monitored via HPLC and
ESIMS spectra. After 21.30 h the solvent was removed,
yielding 332.1 mg of crude side-chain protected cyclo-
peptide. Side-chain deprotection was obtained by treatment
with TFA/TIS/H2O¼95:2.5:2.5 (100 ml£1 mg of resin) for
1 h under stirring. The cleavage mixture was evaporated and
lyophilized. The crude cyclopeptide was analysed by RP-
HPLC on a Jupiter C-18 analytical column (250£4.60 mm,
˚
column (250£10.00 mm, 10 mm, 300 A), using a 24 min
gradient from 20:80 to 35:65 of CH3CN/H2O (each
containing 0.1% TFA) at a flow rate of 5.0 ml/min. The
HPLC purification yielded, as white solid, 12.8 mg
(yield¼51.8%) of trans-Pro3, trans-Pro5 cyclopeptide
yunnanin A (Rt¼14.93 min; ESIMS, m/z 676 for
[MþH]þ; mp 228–2298C).
˚
5 mm, 300 A), using a 31 min gradient from 5 to 100 of
CH3CN/H2O (each containing 0.1% TFA) at a flow rate of
1.0 ml/min and UV detection at 220 nm. The HPLC analysis
showed one main peak (Rt¼19.34 min; ESIMS, m/z 828 for
[MþH]þ) identified as the cis-Pro1, cis-Pro3, cis-Pro5
cyclopeptide phakellistatin 1 on the basis of ESIMS and
1H NMR experiments. The crude cyclopeptide was then
purified by semi-preparative RP-HPLC on a Jupiter C-18
3.5. Cyclization of the linear side-chain protected
yunnanin C (2)
The crude linear peptide (61 mg, 0.072 mmol) was
dissolved in DCM (4£1024 M) with HATU (54.4 mg,
0.143 mmol, 2 equiv.) and DIEA (31.4 ml, 0.18 mmol,
2.5 equiv.). The solution was stirred for 1 h on an ice bath
and then allowed to warm at room temperature and kept at
this temperature for 22.30 h. During this time DCM was
gradually added until to 2.4£1024 M to avoid side reactions
such as oligodimerization.
˚
column (250£10.00 mm, 10 mm, 300 A), using a 32 min
gradient from 25:75 to 45:55 of CH3CN/H2O (each
containing 0.1% TFA) at a flow rate of 5.0 ml/min and
UV detection at 220 nm. The HPLC purification yielded, as
white solid, 9.6 mg (yield¼22.3%) of cis-Pro1, cis-Pro3, cis-
Pro5 cyclopeptide phakellistatin 1 peak (Rt¼21.54 min;
ESIMS, m/z 828 for [MþH]þ; mp 240–2428C).
The cyclization reaction was monitored via HPLC and
ESIMS spectra. After 22.30 h the solvent was removed.
Side-chain deprotection was obtained by treatment with
TFA/TIS/H2O¼95:2.5:2.5 (100 ml£1 mg of resin) for 1 h
under stirring. The cleavage mixture was evaporated and
lyophilised, yielding 90.6 mg of crude cyclopeptide. It was
analysed by RP-HPLC on a Jupiter C-18 analytical column
3.7. Cyclization of the linear side-chain protected
phakellistatin 10 (4)
A portion of the crude linear peptide (150.0 mg, 0.14 mmol)
was dissolved in DCM (2.4£1024 M) with HATU
(106.5 mg, 0.28 mmol, 2 equiv.) and DIEA (61.3 ml,
0.352 mmol, 2.5 equiv.). The solution was stirred for 1 h
on an ice bath and then allowed to warm at room
temperature and kept at this temperature for 7.30 h. The
cyclization reaction was monitored via HPLC and ESIMS
spectra. After 7.30 h the solvent was removed. Side-chain
deprotection was obtained by treatment with TFA/TIS/H2-
O¼95:2.5:2.5 (100 ml£1 mg of resin) for 1 h under stirring.
The cleavage mixture was evaporated and lyophilised,
˚
(250£4.60 mm, 5 mm, 300 A), using a 31 min gradient from
5 to 100 of CH3CN/H2O (each containing 0.1% TFA) at a
flow rate of 1.0 ml/min and UV detection at 220 nm. The
HPLC analysis showed two main peaks; the major one was
identified as the trans-Pro3 yunnanin C, indistinguishable
from the natural cyclopeptide (Rt¼15.12 min; ESIMS, m/z
1
722 for [MþH]þ) on the basis of ESIMS and H NMR
experiments, while the minor one was identified as a
mixture of two conformational isomers of yunnanin C

10210
A. Napolitano et al. / Tetrahedron 59 (2003) 10203–10211
yielding 283.9 mg of crude cyclopeptide. The crude
cyclopeptide was analysed by RP-HPLC on a Jupiter C-18
viability of each cell line in response to treatment with
tested compounds and 6-MP was calculated as: % dead
cells¼1002(OD treated/OD control)£100. Table 5 shows
the results obtained expressed as an IC50 value (mM), the
concentration that inhibited cell growth by 50% as
compared to the control.
˚
analytical column (250£4.60 mm, 5 mm, 300 A), using a
31 min gradient from 5 to 100 of CH3CN/H2O (each
containing 0.1% TFA) at a flow rate of 1.0 ml/min and UV
detection at 220 nm. The HPLC analysis showed two main
peaks; the major one was identified as the trans-Pro1, trans-
Pro4, trans-Pro6 cyclopeptide phakellistatin 10
(Rt¼22.42 min; ESIMS, m/z 903 for [MþH]þ) on the
basis of ESIMS and 1H NMR experiments, while the minor
one was supposed to be a geometric isomer of phakellistatin
10 (Rt¼22.88 min; ESIMS, m/z 903 for [MþH]þ). The
crude cyclopeptides were then purified by semi-preparative
RP-HPLC on a Jupiter C-18 column (250£10.00 mm,
References
1. Faulkner, D. J. Nat. Prod. Rep. 1991, 97–147. Bewley, A.;
Faulkner, D. J. Nat. Prod. Rep. 1997.
2. Brown, A. P.; Morrissey, R. L.; Faircloth, G. T.; Levine, B. S.
Cancer Chemother. Pharmacol. 2002, 50, 333–340. Rinehart,
K. L.; Gloer, J. B.; Hughes, R. G.; Renis, H. E.; McGovren,
J. P.; Swynenberg, E. B.; Stringfellow, D. A.; Kuentzel, S. L.;
Li, L. H. Science 1981, 212, 933–935. Jiang, T. L.; Liu, R. H.;
Salmon, S. E. Cancer Chemother. Pharmacol. 1983, 11, 1–4.
Tada, H.; Tozyo, T.; Terui, Y.; Hayashi, F. Chem. Lett. 1992,
3, 431–434.
˚
10 mm, 300 A), using a 56 min gradient from 25:75 to
60:40 of CH3CN/H2O (each containing 0.1% TFA) at a flow
rate of 5.0 ml/min and UV detection at 220 nm. The HPLC
purification yielded, as white solids, 6.3 mg (yield¼16.4%)
of trans-Pro1, trans-Pro4, trans-Pro6 cyclopeptide phakel-
listatin 10 (Rt¼34.70 min; ESIMS, m/z 903 for [MþH]þ;
mp 215–2188C) and 4.8 mg (yield¼12.5%) of the minor
product (Rt¼35.60 min; ESIMS, m/z 903 for [MþH]þ; mp
193–1958C).
3. Pettit, G. R.; Clewlow, P. J.; Dufresne, C.; Doubek, D. L.;
Cerny, R. L.; Rutzler, K. Can. J. Chem. 1990, 68, 708–711.
Pettit, G. R.; Srirangam, J. K.; Herald, D. L.; Erickson, K. L.;
Doubek, D. L.; Schmidt, J. M.; Tackett, L. P.; Bakus, G. J.
J. Org. Chem. 1992, 57, 7217–7220. Pettit, G. R.; Herald,
C. L.; Boyd, M. R.; Leet, J. E.; Dufresne, C.; Doubek, D. L.;
Schimdt, J. M.; Cerny, R. L.; Hooper, J. N. A.; Rutzler, K. C.
J. Med. Chem. 1991, 34, 3339–3340. Kobayashi, J.; Tsuda,
M.; Nakamura, T. i.; Mikami, Y.; Shigemori, H. Tetrahedron
1993, 49, 2391–2402. Tsuda, M.; Shigemori, H.; Mikami, Y.;
Kobayashi, J. Tetrahedron 1993, 49, 6785–6796. Tsuda, M.;
Sasaki, T.; Kobayashi, J. Tetrahedron 1994, 49, 4667–4680.
4. Morita, H.; Kobata, H.; Takeya, K.; Itokawa, H. Tetrahedron
Lett. 1994, 35, 3563–3564. Morita, H.; Gonda, A.; Takeya, K.;
Itokawa, H. Bioorg. Med. Chem. Lett. 1996, 6, 767–770.
Morita, H.; Kayashita, T.; Shishido, A.; Takeya, K.; Itokawa,
H.; Shiro, M. Tetrahedron 1996, 52, 1165–1176. Morita, H.;
Kayashita, T.; Kobata, H.; Gonda, A.; Takeya, K.; Itokawa, H.
Tetrahedron 1994, 50, 6797–6804.
3.8. Preparation of cells
J774.A1, murine monocyte/macrophage cells were grown in
adhesion on Petri dishes and maintained with Dulbecco’s
modified Eagle’s medium (DMEM) at 378C in DMEM
supplemented with 10% foetal calf serum (FCS), 25 mM
HEPES, 2 mM glutamine, 100 u/ml penicillin and
100 mg/ml streptomycin. WEHI-164, murine fibrosarcoma
cells were maintained in adhesion on Petri dishes with
DMEM supplemented with 10% heat-inactivated FCS,
25 mM HEPES, 100 u/ml penicillin and 100 mg/ml strepto-
mycin. HEK-293, human epithelial kidney cells were
maintained and grown in adhesion on Petri dishes with
DMEM supplemented with 10% FCS, 25 mM HEPES,
100 u/ml penicillin and 100 mg/ml streptomycin. All
reagents for cell culture were from Hy-Clone (Euroclone,
Paignton Devon, UK); [3-(4,5-dimethylthiazol-2-yl)-2,5-
phenyl-2H-tetrazolium bromide] (MTT) and 6-mercapto-
purine (6-MP) were from Sigma Chemicals (Milan, Italy).
5. Verbist, J. F.; Minale, L.; Franz, G.; Sodano, G.; Riccio, R.;
Chermann, J. C. et al. Third European Marine Science and
Technology Conference (Lisbon, 23–27 May 1998), Vol. III
Project Synopses, pp 1215–1229..
3.9. Antiproliferative assay
`
6. Napolitano, A.; Bruno, I.; Rovero, P.; Lucas, R.; Paya Peris,
M.; Gomez-Paloma, L.; Riccio, R. Tetrahedron 2001, 57,
6249–6255.
`
7. Napolitano, A.; Bruno, I.; Rovero, P.; Lucas, R.; Paya Peris,
M.; Gomez-Paloma, L.; Riccio, R. J. Pept. Sci. 2002, 8,
407–417.
J774.A1, WEHI-164 and HEK-293 (3.5£104 cells) were
plated on 96-well microtiter plates and allowed to adhere at
378C in 5% CO2 and 95% air for 2 h.
Thereafter, the medium was replaced with 50 mL of fresh
medium and a 75 mL aliquot of 1:4 serial dilution of each
test compound was added and then the cells incubated for
72 h. In some experiments, serial dilutions of 6-MP were
added. The cell viability was assessed through an MTT
conversion assay.14,15 Briefly, 25 mL of MTT (5 mg/ml)
were added and the cells were incubated for an additional
3 h. Thereafter, cells were lysed and the dark blue crystals
solubilised with 100 mL of a solution containing 50% (v:v)
N,N-dimethylformamide, 20% (w:v) SDS with an adjusted
pH of 4.5.16 The optical density (OD) of each well was
measured with a microplate spectrophotometer (Titertek
Multiskan MCC/340) equipped with a 620 nm filter. The
8. Morita, H.; Shishido, A.; Kayashita, T.; Shimomura, M.;
Takeya, K.; Itokawa, H. Chem. Lett. 1994, 2415–2418.
Morita, H.; Kayashita, T.; Shimomura, M.; Takeya, K.;
Itokawa, H. J. Nat. Prod. 1996, 59, 280–282.
9. Pettit, G. R.; Cichacz, Z.; Barkoczy, J.; Dorsaz, A. C.; Herald,
D. L.; Williams, M. D.; Doubek, D. L.; Schmidt, J. M.;
Tackett, L. P.; Brune, D. C.; Cerny, R. L.; Hooper, J. N. A.;
Bakus, G. J. Nat. Prod. 1993, 56, 260–267. Pettit, G. R.; Tan,
R.; Ichihara, Y.; Williams, M. D.; Doubek, D. L.; Tackett,
L. P.; Schmidt, J. M.; Cerny, R. L.; Boyd, M. R.; Hooper, J. N.
A. J. Nat. Prod. 1995, 58, 961–965.
10. Pettit, G. R.; Cichacz, Z.; Barkoczy, J.; Dorsaz, A. C.; Herald,
D. L.; Williams, M. D.; Doubek, D. L.; Schmidt, J. M.;

A. Napolitano et al. / Tetrahedron 59 (2003) 10203–10211
10211
Tackett, L. P.; et al. J. Nat. Prod. 1993, 56, 253–260.
13. Tabudravu, J. N.; Jaspars, M.; Morris, L. A.; Kettenes-Van
Den Bosch, J. J.; Smith, N. J. Org. Chem. 2002, 67,
8593–8601.
`
Randazzo, A.; Dal Piaz, F.; Orru, S.; Debitus, C.; Roussakis,
C.; Pucci, P.; Gomez-Paloma, L. Eur. J. Org. Chem. 1998,
2659–2665.
14. Setsuo, S.; Shgeya, S.; Norihiro, T.; Yiochi, N.; Keishi, N.;
Madoka, I.; Isao, I. Chem Pharm. Bull. 1990, 38, 411–414.
15. Kinjo, J.; Uemura, H.; Nakamura, M.; Nohara, T. Chem.
Pharm. Bull. 1994, 42, 1339–1341.
11. Pettit, G. R.; Toki, B. E.; Xu, J. P.; Brune, D. C. J. Nat. Prod.
2000, 63, 22–28.
12. Pettit, G. R.; Lippert, J. W., III; Taylor, S. R.; Tan, R.;
Williams, M. D. J. Nat. Prod. 2001, 64, 883–891.
16. Mosman, T. J. Immunol. Methods 1983, 65, 55–63.