Morpholine-based RGD-cyclopentapeptides as αvβ3/αvβ5 integrin ligands: Role of configuration towards receptor binding affinity

Article abstract of DOI:10.1016/j.bmc.2009.01.006

Two c[RGDfX] cyclopeptides, having either l- or d-morpholine-3-COOH (Mor) as the X amino acid were developed as ligands for α_vβ₃/α_vβ₅ integrins. Biological assays showed only d-Mor-containing cyclopentapeptide c

Full text of DOI:10.1016/j.bmc.2009.01.006

Bioorganic & Medicinal Chemistry 17 (2009) 1542–1549
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Morpholine-based RGD-cyclopentapeptides as avb3/avb5 integrin ligands:
Role of configuration towards receptor binding affinity
Nicoletta Cini a,b, Andrea Trabocchi a, Gloria Menchi a,c, Anna Bottoncetti b, Silvia Raspanti b,c, Alberto Pupi b,c
,
Antonio Guarna a,c,
*
a Dipartimento di Chimica Organica ‘‘Ugo Schiff”, Università degli Studi di Firenze, Polo Scientifico e Tecnologico, Via della Lastruccia 13, I-50019 Sesto Fiorentino (FI), Italy
b Dipartimento di Fisiopatologia Clinica, Unità di Medicina Nucleare, Università degli Studi di Firenze, Viale Pieraccini 6, I-50134 Firenze, Italy
c Centro Interdipartimentale per lo Sviluppo Preclinico dell’Imaging Molecolare (CISPIM), Università degli Studi di Firenze, Viale Pieraccini 6, I-50134 Firenze, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Two c[RGDfX] cyclopeptides, having either
L
- or
D-morpholine-3-COOH (Mor) as the X amino acid were
Received 2 September 2008
Revised 5 January 2009
Accepted 7 January 2009
Available online 13 January 2009
developed as ligands for
tapeptide capable to bind
revealing a connection between the configuration of Mor and the preferred binding to
formational analysis showed different structural preferences for the two peptides induced by the two
enantiomeric cyclic amino acids, suggesting a role of the stereochemistry of Mor on the overall peptide
conformation and on the presentation of the pharmacophoric Arg and Asp side chains.
Ó 2009 Elsevier Ltd. All rights reserved.
a
a
vb3/
vb3 integrin with a low nanomolar affinity according to a two-site model, thus
vb3 integrin. Con-
avb5 integrins. Biological assays showed only D-Mor-containing cyclopen-
a
Keywords:
Peptide
Peptidomimetic
Conformational analysis
NMR
1. Introduction
tasis has been demonstrated, as its expression is up-regulated on
metastatic melanoma5 and late stage glioblastoma.6 Recent stud-
Integrins are a class of cellular receptors known to bind extra-
cellular matrix proteins, and therefore mediate cell adhesion
events. The integrin receptors constitute a family of proteins with
structural characteristics of non-covalent heterodimeric glycopro-
ies have also confirmed the implication of the avb5 integrin in
angiogenesis, possibly through a distinct signalling pathway, acti-
vated by different growth factors, such as VEGF.4 During last dec-
ade the cyclic peptide c[RGDfV] as a selective ligand for the
avb3
tein complexes formed of
a
and b subunits.1 The vitronectin
integrin, was developed by Kessler and co-workers through a
‘spatial screening’ of libraries of c[RGDYX] cyclopentapeptides
containing the RGD sequence, demonstrating the influence of res-
receptor is known to refer to three different integrins, named
avb1,
avb3 and avb5. While avb3 binds a large variety of ligands,
avb5 binds exclusively vitronectin. One important recognition site
idues X and Y on the
a
vb3 inhibition.7 In particular, high inhibi-
in a ligand for many integrins is the arginine–glycine–aspartic
acid (RGD) tripeptide sequence,2 which is found in all peptide-
based ligands identified for the vitronectin receptor integrins.
tion activity was found for compounds having at position 4,
corresponding to the (i + 1) position of a type-II’ b-turn, an aro-
matic hydrophobic
natively, -serine) or
substitution pattern for the X amino acid was ascertained not
to have a great influence on the vb3 inhibition, although a rela-
tively lower activity was generally found for cyclopentapeptides
displaying -amino acids at such position of the sequence. As a
D-amino acid, such as
D-Phe or
D
-Trp (or, alter-
Among the RGD-dependent integrins, the
a
vb3 and
a
vb5 receptors
a
D
a
glycine. Also,
a more variable
have received increasing attention as therapeutic targets as they
are expressed in various cell types and are involved in osteoporo-
sis, arthritis, retinopathy, and tumour-related processes.3 The
RGD recognition site can be mimicked by polypeptides that con-
tain the RGD sequence, and the specificity of the inhibition can be
modulated by the sequence and structure of such peptides, so as
a
D
result of these SAR analyses, many different cyclic RGD-contain-
ing peptides have been reported by replacement of the central
to target specific integrins.
a
vb3 antagonists, including RGD-con-
D-Phe-Val dipeptidic sequence with different turn mimetics and
taining peptides, have been successfully applied as inhibitors of
sugar amino acids, with the aim of increasing the conformational
constraint of the RGD sequence.8 In particular, the N-methylated
derivative c[RGDf(Me)Val] (Cilengitide, EMD121974),8b which is
actually being examined in clinical phase II as an angiogenesis
inhibitor, resulted in significantly higher binding affinity with re-
spect to the parent c[RGDfV] peptide.
blood vessel development and tumour growth.4 Also, the role of
avb3 integrin in processes that involve tumour growth and metas-
* Corresponding author. Tel.: +39 055 4573538; fax: +39 055 4573569.
E-mail address: antonio.guarna@unifi.it (A. Guarna).
0968-0896/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2009.01.006

N. Cini et al. / Bioorg. Med. Chem. 17 (2009) 1542–1549
1543
In continuation of our research focused on the development of
morpholine-based heterocycles for peptidomimetic chemistry,9
we recently reported a new efficient method for the preparation
of N-Fmoc-morpholine-3-carboxylic acid starting from serine,
and the introduction in a model peptide by means of solid-phase
peptide synthesis (SPPS) was also described.10 With the aim of
exploring the possibility of using this secondary amino acid in pep-
tidomimetic chemistry, we devised new RGD-based peptidomi-
metics containing the morpholine nucleus as a replacement of N-
methyl-valine of Cilengitide as potential integrin ligands, and also
we were interested in understanding the effect of the configuration
of such amino acid towards receptor affinity and selectivity, as a
result of different conformational preferences of cyclopeptides
containing the two enantiomeric cyclic amino acids. Thus, both
conversion using TBTU/DIPEA as the activating mixture in DMF
after two days and overnight reacting, respectively. The side-chain
protected linear peptide was cleaved from the resin with 1% TFA in
CH2Cl2, and successively allowed to cyclize in diluted DMF solution
using TBTU as the COOH activator. After purification of the cyclic
protected peptide by standard chromatography, pure compound
was achieved by deprotection of Arg and Asp side chains with
95% TFA and TIS/H2O as scavengers, followed by HPLC purification
and elution of the collected fractions over an ion-exchange resin to
give the corresponding pure peptide as the hydrochloride salt.
2.2. Biological assay
The ability of the tested compounds 1 and 2 to compete with
c[RGDfX] cyclopeptides 1 and 2, containing either
line-3-carboxylic acid, respectively (Fig. 1), were synthesized on
solid-phase and assayed in vitro towards vb3 and vb5 integrins.
L
- or
D
-morpho-
[ avb3 and
125I]-echistatin for binding to the isolated, purified
a
vb5 integrins originated from human placenta12,13 was evaluated
a
a
in solid-phase receptor assays.4b,14 Competition studies were car-
ried out using a fixed concentration of the radioligand (0.05 nM
Moreover, the conformational analysis by NMR was assessed so
as to relate differences in the biological affinity to different confor-
mational preferences induced by the shift of configuration of the
two enantiomeric cyclic amino acids. Due to the hydrophilic char-
acter of the title cyclopeptides, the conformational analysis by
NMR was carried out in DMSO-d6,7 thus investigating both hydro-
gen-bonding by variable temperature 1D experiments, and the 3D
structural organisation by TOCSY and ROESY analysis.
and 0.1 nM for
of concentrations between 100
a
vb3 and
a
vb5 receptors, respectively) and a range
M and 0.01 nM of the tested mol-
l
ecules. The IC50 SEM values (nM) were calculated as the concen-
tration of compound required for 50% inhibition of radioligand
binding, as estimated by the GraphPad Prism program, and the re-
sults are reported in Table 1.
Interestingly, the IC50 values obtained showed similar binding
affinity of compounds 1 and 2 towards
tide 2 proved to bind to vb3 integrin with significant higher affin-
ity with respect to diastereomeric peptide 1. Such evidence of
Mor-containing peptide 2 as a better ligand for vb3 integrin sug-
gested the shift of configuration from - to -Mor to induce a signif-
icant conformational variation towards the optimal presentation of
the pharmacophoric groups of the ligand. Moreover, the vb3 inhi-
avb5 receptor, whereas pep-
a
2. Results and discussion
2.1. Synthesis
D-
a
L
D
Fmoc-protected
L
- and
D-Mor were obtained from dimethoxy-
a
acetaldehyde and - or
L
D-serine methyl ester, respectively, through
bition curve of compound 2 showed Hill slope values different
from unity, indicating the binding curve to be better described
by a two-sites model (Fig. 2), as a consequence of binding of 2 with
two states of the integrin, in agreement with other ligands already
described in the literature.15 Thus, the analysis of the data with
a short and practical synthetic route consisting of a five-step pro-
cess based on reductive amination, Fmoc protection of the amine
group, intramolecular acetalization and concomitant elimination
of the anomeric methoxy substituent, followed by double bond
hydrogenation of the resulting oxazine, and final acidic ester
hydrolysis.10 The linear peptide sequences were prepared on so-
lid-phase starting from glycine bound to acid-labile 2-Cl-trityl re-
sin (Scheme 1).
Amino acid couplings were carried out using DIPC/HOBt as
COOH activators, and the prosecution of the reaction was moni-
tored with 1% bromophenol blue (BB) solution as internal indica-
tor.11 Coupling of Mor and of Asp were achieved with complete
non-linear fitting indicated cyclopeptide 2 to bind to avb3 receptor
according to a two-site binding model. As a possible explanation,
such effect could be due to different affinity conformational states
that can be assumed by the integrins.16,17 The transition from low-
affinity to high-affinity conformation has been described to be in-
duced by natural ligands,18 and also by most ligand-mimetic
antagonists.19 Moreover, it was reported that some antagonists,
such as the disintegrin echistatin, bind to the active and inactive
forms with similar affinity, whereas others bind to the active form
with higher affinity than to the inactive form.20 The data obtained
suggest a similar binding behaviour for compound 2, however fur-
ther studies need to be conducted to clarify its binding properties.
More importantly, the affinity data proved that the introduction of
D
-Mor in the cyclic peptide 2 does not destroy its affinity for the
integrins, and induce important variations in the Arg-Gly-Asp con-
formation with respect to the corresponding -Mor-containing
L
peptide 1. Interestingly, in this case, in contrast to Kessler’s obser-
vation, the presence at the X position of the peptide c[RGDfX] of a
D-amino acid mimetic significantly increases the integrin affinity
compared to isomer 1, which experiences a L-residue at the posi-
tion X.
2.3. Conformational analysis
Conformational analysis of cyclopeptides 1 and 2 was carried
out to assess the structural determinants leading to differences
in binding activity towards
avb3. Diluted DMSO-d6 solutions of 1
and 2 were used for the NMR analysis in order to prevent aggrega-
Figure 1. c[RGDfMor] cyclopeptides
respectively.
1 and 2 containing L-Mor and D-Mor,
tion. TOCSY, ROESY and variable temperature 1H NMR experiments

1544
N. Cini et al. / Bioorg. Med. Chem. 17 (2009) 1542–1549
Scheme 1. Representative synthesis of compound 1, containing
L-Mor. Reagents and conditions: (a) Fmoc-Arg(Pbf)-OH, DIPC/HOBt, BB, DMF, rt, overnight; (b) 30% piperidine
in DMF, rt, 30 min; (c) Fmoc- -Mor-OH, TBTU, DIPEA, DMF, rt, 2 d; (d) Fmoc-
L
D
-Phe-OH, DIPC/HOBt, BB, DMF, rt, overnight; (e) Fmoc-Asp(t-Bu)-OH, TBTU, DIPEA, DMF, rt,
overnight; (f) 1% TFA, CH2Cl2, rt, 10 Â 2 min; (g) TBTU, DIPEA, DMF, rt, 24 h; (h) 95% TFA, 2.5% H2O, 2.5% TIS, rt, 2 h.
Table 1
Inhibition of [125I]-echistatin specific binding to purified human integrin proteins
avb3 and
a
Vb5 by compounds 1 and 2
Compound
aVb5
avb3
Two-site model,
a
vb3
a
a
a
a
IC50
Hill slope
IC50
Hill slope
IC50h
%
h
IC50l
%
l
1
2
15.1 2.3
21.0 2.1
0.29 0.02b
0.13 0.01b
0.9 0.1c
À0.80
À0.83
157 0.9
À0.81
À0.56
32.6 7.0
6.5 2.0
55.5
458 191
44.5
Echistatin
EMD121974
ST1646
0.29 0.08b
18.9 3.1b
5.64 0.40c
a
IC50 values (expressed in nM) represent the mean SEM of three experiments performed in triplicate. IC50h and IC50l correspond to IC50 in the receptor high- and low-
affinity states, respectively; %h and %l represent the proportions of high- and low-affinity states of the receptor.
b
For literature values, see Ref. 8f.
This assay. For literature values, see Refs. 8d,f.
c
were carried out for the NMR analysis. Peptide 1, having the
nucleus resulted in a unique rotamer at the -Phe- -Mor amide
bond, whose trans-configuration was established by a ROESY cross
peak between -Mor H-5 and -Phe H- . Variable temperature
experiments showed low temperature coefficients for -Phe NH
L-Mor
D
L
L
D
a
D
and Asp NH, suggesting such protons to be involved in intramolec-
ular hydrogen-bonds. The chemical shift of Asp NH indicated such
proton to be solvent exposed to some extent, indicating the exis-
tence of a conformational equilibrium between hydrogen-bonded
and non-hydrogen-bonded states for this amide proton (Table 2).
Table 2
Chemical shifts and temperature-dependent 1H NMR data for amide protons of
peptides 1 and 2a
Compound
NH
d
Dd/DT
1
Arg
Gly
Asp
6.51
8.68
8.35
7.51
À0.5
À5.5
2.8
D-Phe
À2.0
b
2
Arg
Gly
Asp
8.06
8.59
7.53
8.76
—
À8.5
À2.9
À5.7
D-Phe
a
1H NMR spectra for determining temperature coefficients were obtained at
298–323 K with increments of 5 K. d values are reported in ppm and
Dd/DT coef-
Figure 2. Inhibition of [125I]-echistatin specific binding to purified human integrin
proteins vb3 (top) and Vb5 (bottom) by compounds 1 (red) and 2 (black). Each
point represents the mean value SEM of three experiments performed in
triplicate.
ficients in ppb/K.
b
Broadening of Arg NH signal due to exchange phenomenon complicated the
a
a
interpretation of NMR data, including the evaluation of the corresponding
coefficient.
Dd/DT

N. Cini et al. / Bioorg. Med. Chem. 17 (2009) 1542–1549
1545
Also, the positive value for the temperature coefficient of Asp NH
suggested the intramolecular hydrogen-bonded state to be stabi-
lized upon increasing the temperature. This evidence confirmed
the hypothesis of the existence of equilibrating conformations for
peptide 1, each stabilized by an intramolecular hydrogen-bond
stead of the Arg and Asp side chains were performed as the first
step and using conformational constraints according to ROESY
data. Conformational analysis of 1 resulted in two families of con-
formers stabilized by intramolecular hydrogen-bonds. Specifically,
the group comprising the global minimum conformer displayed a
hydrogen-bond between Asp NH and Arg C@O (Erel = 0 kcal/mol;
Fig. 4, top left), and the second group showed a doubly hydro-
with either Asp or
tional informations about the conformational preferences of 1 in
solution (Fig. 3). Specifically, ROESY peaks between -Phe NH and
Asp H- and between Gly NH and Arg H- supported the existence
of equilibrating intramolecular hydrogen-bonds experienced by
Asp and -Phe amide protons, respectively.
D-Phe amide protons. ROESY data provided addi-
D
gen-bonded structure between Gly NH and
D-Phe C@O, and be-
a
a
tween -Phe NH and Gly C@O (Erel = +1.6 kcal/mol; Fig. 4, top
D
right). Also, the second conformation displayed the Mor carbonyl
group oriented in the less stable equatorial position, accounting
for less stabilization due to the increase of the strain energy. Such
conformations were in agreement with NMR data, indicating a fast
exchanging conformational flexibility between hydrogen-bonded
D
Cyclopeptide 2 displayed a different conformational profile
with respect to peptide 1, suggesting the configuration of morpho-
line-3-carboxylic acid to play a role in the nucleation of the overall
structure. 1H NMR data of diluted DMSO-d6 solution of 2 showed a
single set of signals, attributable to the existence of a single rot-
amer, in analogy with 1. The absence of any relevant ROESY cross
structures. The global minimum conformer for compound
2
showed a kinked structure stabilized by a hydrogen bond between
Asp NH and the carbonyl group of Gly, in agreement with NMR
data (Fig. 4, bottom).
The results from the conformational analysis were consistent
with IC50 values, suggesting a close connection between the con-
formation and the ligand binding affinity, as compound 1, having
peak between
isochrony of -Phe H-
the -Phe– -Mor amide bond. Variable temperature experiments
showed high d/ T coefficients for Gly and -Phe amide protons
D
-Phe H-
a
and both H-5 and H-3, and the accidental
D
a
and H-3 allowed to assign a cis-geometry at
D
D
D
D
D
(Table 2), indicating such protons to be solvent-exposed and not
to participate in any relevant intramolecular hydrogen-bond. Con-
versely, a
D
d/
D
T value of À2.9 and a chemical shift of 7.53 ppm
suggested the existence of a hydrogen-bonded state for Asp NH.
The ROESY analysis of 2 showed similar sequential ROESY peaks
as observed for 1, and specifically between
and between Gly NH and Arg H- , which, in conjunction with
the existence of a cis-configuration at the -Phe– -Mor peptide
D-Phe NH and Asp H-
a
a
D
D
bond, allowed to propose the preferred conformation of 2 in solu-
tion, as shown in Figure 3, bottom. The existence of a stable hydro-
gen-bonded structure, suggested
D-Mor nucleating a more
compact structure, ultimately resulting in a more rigid conforma-
tion with respect to the diastereomeric peptide 1.
Molecular modelling calculations were carried out to give more
insight about the conformational preferences of peptides 1 and 2.
Specifically, the conformational preferences of compounds 1 and
2 were investigated by molecular mechanics calculations within
22
the framework of Macromodel v6.5,21 using Amber
as a force
*
field and the implicit water GB/SA solvation model of Still et al.23
Monte Carlo energy minimization (MCEM)24 conformational
searches of the peptide analogues containing methyl groups in-
Figure 4. Energy-minimized conformations for 1 and 2: top left: 1, global minimum
conformer; top right: 1, 2nd low-energy conformer; bottom: 2, global minimum
conformer.
Figure 3. Selected ROESY peaks for compounds 1 (top) and 2 (bottom).

1546
N. Cini et al. / Bioorg. Med. Chem. 17 (2009) 1542–1549
a more flexible structure, as a consequence of fast equilibration be-
tween two main conformations, resulted in a significantly lower
binding capability than cyclopeptide 2. Also, all experimental data
suggested that the role of
mation was due to the existence of the cis-rotamer, which allowed
maintaining the required torsional angles between -Phe and Arg
for a good binding to the integrin. Finally, similar binding affinity
for vb5 in both cyclopeptides 1 and 2 indicated such integrin to
better accommodate both ligands, as a consequence of less strin-
D-Mor in nucleating the correct confor-
D
a
gent binding requirements compared to avb3.
2.4. Docking calculations
A docking simulation was carried using both ligands 1 and 2
with the aim to gain further insight into the possible binding
modes of the cyclopeptides to the
level. The crystal structure of the complex formed by
c[RGDf(Me)V], and the extracellular fragment of the vb3 receptor
avb3 receptor at the molecular
a
(PDB code: 1L5G) provide a general mode of interaction between
the integrin and its ligands.19 Specifically, the Asp carboxylate
and the Arg guanidinium moiety of RGD-based ligands are the
two key structural elements for the receptor recognition. In fact,
the carboxylate group interacts with the metal ion dependant
adhesion site (MIDAS) consisting of Mn2+ ions and Ser121/
Ser123, whereas the Arg guanidinium group is responsible for salt
bridge interactions with the side chains of Asp218 and Asp150.
Also, additional ligand–receptor contacts engage Tyr122 in hydro-
phobic
p-stacking and Asn215 in hydrogen-bonding interactions.
The docking program Autodock 4.0.125 was used to evaluate the
binding energies of selected conformations of 1 and 2 as potential
ligands for the avb3 receptor, and the docked conformations were
analysed taking into account the binding interactions observed in
the crystal structure of the bound ligand–protein complex. The
docking results for both ligands revealed the key interactions with
Asp218 and the MIDAS site of the receptor, suggesting all these
interactions to be necessary for the molecular recognition of the
Arg-Gly-Asp-containing ligands, although they could not provide
a criterion for determining their differences in affinity (Fig. 5). Spe-
cifically, the global minimum conformer of peptide 1, conformer A,
showed the Asp side chain located in the MIDAS site interacting
with Ser121, and the guanidinium group of Arg experiencing only
a monodentate interaction with Asp218 (Fig. 5, top). The same
interactions were observed for the second low-energy conforma-
tion of 1, conformer B (Fig. 5, middle), also displaying the benzyl
group in the proximity of Tyr122 side chain, although the back-
bone structure was positioned in the shallow pocket between
these two in a way considerably different from low-nanomolar
RGD-based inhibitors, such as c[RGDf(Me)V], which is able to lay
flatter and thus in closer contact with the protein surface.19 Peptide
2 resulted in a main cluster of conformations displaying the typical
binding mode of RGD cyclopeptide-based ligands. Specifically, the
cluster showed the canonical binding mode consisting in bidentate
Asp218/Arg and MIDAS/Ser121/Asp interactions, and a hydropho-
Figure 5. RGD ligands 1 and 2 (green) docked into the binding region of
avb3
integrin highlighting the protein residues (magenta) that form the key interactions.
Top: 1, conformer A: Asp218 versus Arg, Ser121 and Mn2+ versus Asp; middle: 1,
conformer B: Asp218 versus Arg, Ser121 and Mn2+ versus Asp, Tyr122 versus
bottom: 2, Asp218 versus Arg, Ser121 and Mn2+ versus Asp, Tyr122 versus
Non-polar hydrogen atoms are omitted for clarity.
D
-Phe;
-Phe.
D
bic
p
-stacking between
D-Phe and Tyr122 (Fig. 5, bottom).
entation between Asp218 and Asp150, although such interaction
was also observed in some docked conformations of compound 2
(figure not shown). These differences in secondary interactions
with integrins may explain the different binding affinity between
The comparison of the binding mode of compound 2 with re-
spect to the reference ligand c[RGDf(Me)V], known also as
EMD121974 (see Table 1), revealed some interesting issues to ad-
dress the different bioactivity of the two ligands (Fig. 6). Although
showing a similar binding mode of Asp and Arg side chains in the
key region of the receptor (see Fig. 5), the two cyclopeptides
showed a different conformation of the cyclopeptidic backbone.
The reference ligand c[RGDf(Me)V] showed secondary interactions
as a consequence of a different conformation exposing the Asp NH
toward the receptor’s cavity. Such proton established additional
hydrogen-bonding interactions with either carbonyl groups of
Asn215 or Arg216. Also, the guanidine group showed a bridged ori-
the two ligands, especially towards
avb5 integrin, were
c[RGDf(Me)V] was found to be more active by two order of magni-
tude with respect to compound 2.
More interestingly, the prediction of the free energy of binding
and of the corresponding Ki for the lowest-energy clusters of both
compounds was in agreement with bioassay data (Table 3). In fact,
the cluster of compound 1, conformer A showed a binding energy
of À9.5 kcal/mol and a predicted Ki of 262 nM, whereas the second

N. Cini et al. / Bioorg. Med. Chem. 17 (2009) 1542–1549
1547
having
D-configuration at the
a-carbon as tools to nucleate bioac-
tive conformations in cyclopentapeptides of general formula
c[RGDfX].
4. Experimental
4.1. Chemistry
H-Gly-2-Cl-Trt resin (1.1 mmol/g) was purchased from Fluka.
Chromatographic separations were performed on silica gel (Kiesel-
gel 60, Merck) using flash-column techniques; Rf values refer to
TLC carried out on 25-mm silica gel plates (Merck F254) with the
same eluant as indicated for column chromatography. All the solid
phase reactions were carried out on a shaker, using solvent of HPLC
quality. 1H NMR spectra were recorded with Varian Gemini and
Mercury NMR spectrometers operating at 200 MHz and 400 MHz
for the proton, respectively. Bromophenol Blue (BB) test was per-
formed following this procedure: a few resin beads were put in a
vial and suspended in 0.5 mL of DMF; two drops of a 1% solution
of bromophenol blue in dimethylacetamide were added and the
sample was observed. The BB test was considered to be positive
(presence of free amino groups) when the resin beads turned blue
immediately, and negative (absence of free amino groups) when
the beads remained colourless. ESI-MS spectra were recorder by
a PE SCIEX API 365 spectrometer. Compounds 1 and 2 were puri-
fied by Beckman-Gold HPLC system equipped with a reverse-phase
Figure 6. Overlap between the reference ligand c[RGDf(Me)V] (cyan), and com-
pound 2 (green) in the binding region of avb3 integrin.
Table 3
Binding energy and predicted Ki valuesa
Compound
Estimated free energy of binding (kcal/mol)
Ki (nM)
1A
1B
2
À9.5
À8.1
262
1003
43
À10.8
a
column (Alltima C18 10
l
m, 250 Â 10 mm, Alltech) using H2O/
Taken as mean values of the docked conformations of the main cluster for each
structure.
CH3CN gradient eluant buffered with 0.1% TFA (flow = 2.5 mL/
min, k = 254 nm, gradient eluant: CH3CN 10%/5 min, CH3CN 10–
90%/25 min). Analytical HPLC analyses were performed using a
Dionex Ulltimate 3000 system equipped with a reverse-phase col-
conformer (conformer B) resulted in lower binding affinity with
the receptor, displaying a value of À8.1 kcal/mol for the binding
umn (Alltima C18 5
dient eluant as described above.
l
m, 250 Â 4.6 mm, Alltech) and the same gra-
energy and of 1.0 lM for the predicted Ki value. On the contrary,
the low-energy cluster of compound 2 showed a stronger affinity
towards vb3, as demonstrated by the lower binding energy
a
4.1.1. c[RGDf-(3S)-Carboxymorpholine] (1)
(À10.8 kcal/mol, see Table 3) and a predicted Ki value falling in
H-Gly-2-Cl-Trt resin (466 mg, 0.5 mmol) was used as the start-
ing material. Fmoc-deprotections and the completion of each cou-
pling reaction were assessed by performing a BB test. The coupling
of the first amino acid was performed with a solution of N-Fmoc-
Arg(Pbf)-OH (973 mg, 1.5 mmol, 3 equiv), HOBt (203 mg,
the low nanomolar range.
3. Conclusion
1.5 mmol, 3 equiv) in DMF (4 mL) and DIPC (236 lL, 1.5 mmol,
We reported the synthesis of two cyclopentapeptides and the
3 equiv) was added dropwise at 0 °C. The resulting mixture was
stirred for 10 min at this temperature and for further 10 min at
room temperature, then added to the resin. This mixture was sha-
ken at room temperature overnight. The solution was drained and
the resin was washed with 5% DIPEA in DMF (3 Â 5 mL) and DMF
(5 Â 5 mL). Fmoc deprotection was performed with 30% piperidine
in DMF (10 mL) for 30 min, followed by resin washings with DMF
(3 Â 10 mL). After deprotection, H-Arg(Pbf)-Gly-2-Cl-Trt resin
was suspended in a solution of TBTU (318 mg, 1 mmol, 2 equiv),
N-Fmoc-(S)-morpholine-3-carboxylic acid (350 mg, 1 mmol,
biological assay towards
tiomers of Mor. The binding affinity revealed
gand as the more active towards vb3, whereas the two
cyclopeptides showed the same affinity for the vb5 integrin, indi-
cating a less strict conformational requirement for optimal binding
to this protein. The result was particularly striking as the -amino
acid-containing ligand retained the activity towards vb3 integrin,
irrespective of the change of configuration at the X position of
common integrin ligands having the c[RGDfX] sequence. Confor-
mational analysis suggested the two enantiomeric secondary ami-
no acids to stabilize two different conformations, as a consequence
a
vb3/
a
Vb5 integrins containing both enan-
D
-Mor-containing li-
a
a
D
a
2 equiv) and DIPEA (169 lL, 1 mmol, 2 equiv) in DMF (3 mL) and
shaken for two days. The solution was drained and the resin was
washed with 5% DIPEA in DMF (3 Â 5 mL) and DMF (5 Â 5 mL).
After deprotection, H-(3S)-carboxymorpholine-Arg(Pbf)-Gly-2-Cl-
Trt resin was suspended in a solution of HOBt (203 mg, 1.5 mmol,
of diverse rotamers nucleated at the
D-Phe-Mor peptide bond. Spe-
cifically, -Mor established a cis peptide bond, which provided a
D
unique conformation resembling the conformation of
c[RGDf(Me)V] found in the crystal structure reported by Arnaoult
and co-workers19 Such conformation was characterized by a
kinked conformation stabilized by a hydrogen-bond involving
3 equiv), DIPC (236 lL, 1.5 mmol, 3 equiv), N-Fmoc-D-Phe-OH
(581 mg, 1.5 mmol, 3 equiv) in DMF (4 mL) and shaken overnight.
The solution was drained and the resin was washed with 5% DIPEA
in DMF (3 Â 5 mL) and DMF (5 Â 5 mL). After deprotection,
Asp amide proton. Conversely,
L-Mor nucleated a trans peptide
bond with the -Phe carboxylic function, thus inducing equilibrat-
D
H-D-Phe-(3S)-carboxymorpholine-Arg(Pbf)-Gly-2-Cl-Trt resin was
ing conformations stabilized by weak equilibrating hydrogen-
bonds, which caused the two pharmacophoric side chain to be ex-
suspended in a solution of TBTU (640 mg, 2 mmol, 4 equiv),
N-Fmoc-Asp(t-Bu)-OH (825 mg, 2 mmol, 4 equiv) and DIPEA
posed in a less favourable orientation for optimal binding to
integrin. The rationale of the present study may suggest the possi-
bility to use other six-membered ring secondary -amino acids
avb3
(342
lL, 2 mmol, 4 equiv) in DMF (4 mL) and shaken for two days.
The solution was drained and the resin was washed with 5% DIPEA
a

1548
N. Cini et al. / Bioorg. Med. Chem. 17 (2009) 1542–1549
in DMF (3 Â 5 mL) and DMF (5 Â 5 mL). N-Fmoc-Asp(t-Bu)-
D
-Phe-
4.2. Biology
(3S)-carboxymorpholine-Arg(Pbf)-Gly-2-Cl-Trt resin was finally
deprotected and washed with DMF (5 Â 5 mL) and DCM (5 Â
4.2.1. Solid-phase receptor binding assay
5 mL). In a solid phase reaction vessel H-Asp(t-Bu)-D-Phe-(3S)-car-
[
125I]-Echistatin, labelled by the lactoperoxidase method26 to a
boxymorpholine-Arg(Pbf)-Gly-2-Cl-Trt resin was treated with
5 mL of 1% TFA/DCM solution (10 Â 2 min). The filtrates were
immediately neutralized with a 10% pyridine/MeOH solution
(1 mL), then the resin was washed with DCM (3 Â 5 mL). The frac-
tions containing the product (TLC:DCM/MeOH 4:1) were combined
and concentrated under reduced pressure to yield a residue, which
was suspended in H2O and purified from the pyridinium salts by
size-exclusion chromatography (AMBERLITE XAD2 resin, H2O then
MeOH). Evaporation of the combined MeOH fractions containing
the product afforded the side-chain protected peptide (450 mg,
98%) as a white solid. The linear peptide (450 mg, 0.492 mmol)
was dissolved under a nitrogen atmosphere in DMF (110 mL), then
specific activity of 2000 Ci/mmol, was purchased from GE Health-
care. Integrin proteins vb3 and vb5 purified from human placenta
a
a
were purchased from Chemicon International Inc, Temecula, CA.
The receptor binding assay was performed as described.13,4b Puri-
fied receptors
avb3 and avb5 were diluted respectively at 500 ng/
mL and 1000 ng/mL in coating buffer (20 mM Tris (pH 7.4),
150 mM NaCl, 2 mM CaCl2, 1 mM MgCl2, 1 mM MnCl2). An aliquot
of the diluted receptors (100 lL/well) was added to a 96-well
microtiter plate (Optiplate-96 HB, PerkinElmer Life Sciences, Bos-
ton, MA) and incubated overnight at 4 °C. The plate was washed
once with blocking/binding buffer (20 mM Tris (pH 7.4), 150 mM
NaCl, 2 mM CaCl2, 1 mM MgCl2, 1 mM MnCl2, 1% BSA) and incu-
bated for additional 2 h at room temperature. The plate was rinsed
twice with the same buffer, then competition binding studies were
performed with a fixed concentration of [125I]-Echistatin (0.05 nM
TBTU (474 mg, 1.47 mmol) and DIPEA (252 lL, 1.47 mmol) were
added and the resulting mixture was stirred for 24 h at room tem-
perature. The solvent was distilled under reduced pressure and the
residue was dissolved in H2O (60 mL) and extracted with EtOAc
(3 Â 80 mL); the organic phase was washed twice with 5% NaHCO3,
dried with Na2SO4 and evaporated under reduced pressure. The
crude residue was purified by flash column chromatography
(DCM/MeOH 9:1, Rf = 0.37) to afford the protected cyclic peptide
(200 mg, 45%), as a white solid. HPLC: tR = 23.80, purity: 86%. Side
chain protected cyclic peptide (200 mg, 0.22 mmol), was treated
with 95:2.5:2.5 TFA/H2O/TIS mixture (20 mL) for 2 h. The reaction
mixture was evaporated under reduced pressure and the residue
was dissolved in H2O (60 mL). The aqueous phase was washed with
iPr2O (3 Â 60 mL) and freeze-dried. The crude residue was purified
by semi-preparative HPLC to give the side-chain deprotected 1
(38 mg, 13%), as a white solid. Trifluoroacetate ion was replaced
with chloride ion by ion-exchange chromatography (AMBERLITE
IRA-96 resin, chloride form). ESI-MS: (m/z) 589.2 [M++H, 100].
HPLC: tR = 13.8, purity 90%. See Table 4 for 1H and 13C NMR data.
and 0.1 nM for
avb3 and
avb5, respectively) and concentrations
ranging from 0.01 nM and 100
lM of the tested compounds. All as-
says were performed in triplicate in a final volume of 0.2 mL, each
containing the following species: 0.05 mL of [125I]-Echistatin,
0.04 mL of the tested compound and 0.11 mL of blocking/binding
buffer. Non-specific binding was defined as [125I]-Echistatin bound
in the presence of an excess (1 lM) of unlabelled echistatin. After
incubation for 3 h at room temperature, the plate was washed
three times with blocking/binding buffer, then counted in a Top-
Count NXT microplate scintillation counter (PerkinElmer Life Sci-
ences, Boston, MA) using 200 lL/well of MicroScint-40 liquid scin-
tillation (PerkinElmer Life Sciences, Boston, MA).
4.2.2. Data analysis
The IC50 values were determined by fitting binding inhibition
data by non-linear regression using GraphPad Prism 4.0 Software
Package (GraphPad Prism, San Diego, CA). Moreover, when the Hill
slope of the curves was significantly less the unity (K < À0.80), the
data were reanalysed with a two-site model. The displacement
curves better fitted (p < 0.05) by a two-sites model than one-site
model were considered significant.
4.1.2. c[RGDf-(3R)-Carboxymorpholine] (2)
Compound 2 was prepared following the procedure as for 1,
using the scaffold N-Fmoc-(R)-morpholine-3-carboxylic acid.
Cleavage from the solid support afforded the linear protected pep-
tide in 83% yield, as a white solid. Cyclization using TBTU method-
ology afforded the protected cyclic peptide in 64% yield, as a white
solid. HPLC: tR = 22.8 (purity: 85%). Final deprotection in analogy
with 1 afforded 2 as a white solid (40 mg, 14%). ESI-MS: (m/z)
589.2 [M++H, 100], 611.1 [M++Na, 7]. HPLC: tR = 13.5, purity: 93%.
See Table 4 for 1H and 13C NMR data.
4.3. NMR methods
NMR experiments were performed at a temperature of 298 K on
Varian Inova and Varian Mercury 400 MHz NMR spectrometers
using diluted DMSO-d6 solutions. All proton chemical shifts were
assigned unambiguously for 1 and 2. Variable temperature 1D,
and 2D experiments (TOCSY, gCOSY, ROESY, HSQC) were carried
out at the sample concentration of 9.2 mM for 1 and 6.2 mM for
2. One-dimensional 1H NMR spectra for determining temperature
coefficients were obtained at 298–323 K with increments of 5 K.
Sample temperatures were controlled with the variable-tempera-
ture unit of the instrument. Proton signals were assigned via TOC-
SY spectra, and ROESY spectra provided the data used in the
conformational analyses. TOCSY spectra were recorded with 2048
points in t1, 256 points in t2, and 8 scans per t2 increment and
using a mixing time of 80 ms. ROESY spectra were recorded with
a similar number of t1 and t2 points unless otherwise noted, and
32 per t2 increment, and using a mixing time of 0.5 s. 1H and 13C
chemical shifts of 1 and 2 are reported in Table 4.
Table 4
1H and 13C chemical shifts of 1 and 2 in DMSO-d6 solutions at 298 K
1 d (ppm)
2 d (ppm)
1H
13C
1H
13C
H-2
H-3
H-5
H-6
Gly NH
Arg NH
Asp NH
4.16, 3.45
4.69
4.14, 3.43
3.60, 3.12
8.70
66.9
53.6
67.0
65.9
—
4.26, 3.09
4.80
2.84, 3.66
3.33, 3.86
8.60
68.8
53.8
66.7
68.1
—
6.48
8.39
—
—
8.06
7.56
—
—
D
-Phe NH
Gly H-
Arg H-
7.49
3.79, 3.62
4.53
2.02–1.91
3.11
4.18
2.66, 2.14
5.02
3.03, 2.68
—
8.80
3.86–3.25
3.96
1.68–1.55
3.11–2.98
4.55
2.45–2.26
4.82
—
a
a
44.8
53.5
26.8, 24.9
40.0
49.3
38.1
50.4
37.9
44.4
53.5
26.9, 25.0
40.7
57.1
39.7
50.3
38.2
Arg H-b,
Arg H-d
Asp H-
Asp H-b
c
4.4. Molecular modelling
a
Conformational preferences of compounds 1 and 2 were inves-
D-Phe H-
a
tigated by molecular mechanics calculations within the framework
D-Phe H-b
3.08–2.73
22
*
of Macromodel v6.5,[21] using Amber
as a force field and the

N. Cini et al. / Bioorg. Med. Chem. 17 (2009) 1542–1549
1549
implicit water GB/SA solvation model of Still et al.23 Monte Carlo
energy minimization (MCEM)24 conformational searches of the
peptide analogues containing methyl groups instead of the Arg
and Asp side chains were performed as the first step. Selected
ROE-cross peaks from ROESY spectra were taken into account for
defining distance constraints. The torsional space of each AGA
cyclopeptide was randomly varied with the usage-directed Monte
Carlo conformational search. Ring-closure bonds were defined in
the morpholine ring and in the cyclopeptide ring. Amide bonds
were included among the rotatable bonds. For each search, at least
1000 starting structures for each variable torsion angle were gen-
erated and minimized until the gradient was less than 0.05 kJ/
Å mol using the truncated Newton-Raphson method implemented
in Macromodel. Duplicate conformations and those with an energy
greater than 6 kcal/mol above the global minimum were discarded.
2. Ruoslathi, E.; Pierschbacher, M. D. Science 1987, 238, 491.
3. For vb3, see: (a) Eliceiri, B. P.; Cheresh, D. A. Cancer J. 2000, 13, 245; (b) Burke,
a
P. A.; De Nardo, S. J.; Miers, L. A.; Lamborn, K. L.; Matzku, S.; De Nardo, G. L.
Cancer Res. 2002, 62, 4263; (c) Haubner, R.; Finsinger, D.; Kessler, H. Angew.
Chem., Int. Ed. 1997, 36, 1374; For aVb5, see: (d) Marinelli, L.; Gottschalk, K. E.;
Meyer, A.; Novellino, E.; Kessler, H. J. Med. Chem. 2004, 47, 4166; (e)
Friedlander, M.; Theesfeld, C. L.; Sugita, M.; Fruttiger, M.; Thomas, M. A.;
Chang, S.; Cheresh, D. A. Proc. Natl. Acad. Sci. 1996, 93, 9764; (f) Friedlander, M.;
Brooks, P. C.; Shaffer, R. W.; Kincaid, C. M.; Varner, J. A.; Cheresh, D. A. Science
1995, 270, 1500.
4. (a) Brooks, P. C.; Stromblad, S.; Klemke, R.; Visscher, D.; Sarkar, F. H.; Cheresh,
D. A. J. Clin. Invest. 1995, 96, 1815; (b) Kumar, C. C.; Malkowski, M.; Yin, Z.;
Tanghetti, E.; Yaremko, B.; Nechuta, T.; Varner, J.; Liu, M.; Smith, E. M.;
Neustadt, B.; Presta, M.; Armstrong, L. Cancer Res. 2001, 61, 2232.
5. Mitjans, F.; Meyer, T.; Fittschen, C.; Goodman, S.; Jonczyk, A.; Marshall, J. F.;
Reyes, G.; Piulats, J. Int. J. Cancer 2000, 87, 716.
6. MacDonald, T. J.; Taga, T.; Shimada, H.; Tabrizi, P.; Zlokovic, B. V.; Cheresh, D.
A.; Laug, W. E. Neurosurgery 2001, 48, 151.
7. Wermuth, J.; Goodman, S. L.; Jonczyk, A.; Kessler, H. J. Am. Chem. Soc. 1997, 119,
1328.
8. (a) Bach, A. C., II; Espina, J. R.; Jackson, S. A.; Stouten, P. F. W.; Duke, J. L.; Mousa,
S. A.; DeGrado, W. F. J. Am. Chem. Soc. 1996, 118, 293; (b) Dechantsreiter, M.;
Plnaker, E.; Mathä, B.; Lohof, E.; Hölzemann, G.; Jonczyk, A.; Goodman, S. L.;
Kessler, H. J. Med. Chem. 1999, 42, 3033; (c) Lohof, E.; Plnaker, E.; Mang, C.;
Burkhart, F.; Dechantsreiter, M. A.; Haubner, R.; Wester, H. J.; Schwaiger, M.;
Hölzemann, G.; Goodman, S. L.; Kessler, H. Angew. Chem. Int. Ed. 2000, 39, 2761;
(d) Belvisi, L.; Bernardi, A.; Checchia, A.; Manzoni, L.; Potenza, D.; Scolastico, C.;
Castorina, M.; Capelli, A.; Giannini, G.; Carminati, P.; Pisano, C. Org. Lett. 2001, 3,
1001; (e) Casiraghi, G.; Rassu, G.; Auzzas, L.; Bureddu, P.; Gaetani, E.; Battistini,
L.; Zanardi, F.; Curti, C.; Nicastro, G.; Belvisi, L.; Motto, I.; Castorina, M.;
Giannini, G.; Pisano, C. J. Med. Chem. 2005, 48, 7675; (f) Belvisi, L.; Bernardi, A.;
Colombo, M.; Manzoni, L.; Potenza, D.; Scolastico, C.; Giannini, G.; Marcellini,
M.; Riccioni, T.; Castorina, M.; LoGiudice, P.; Carminati, P.; Pisano, C. Bioorg.
Med. Chem. 2006, 14, 169.
9. (a) Trabocchi, A.; Menchi, G.; Guarna, F.; Machetti, F.; Scarpi, D.; Guarna, A.
Synlett 2006, 331; (b) Trabocchi, A.; Potenza, D.; Guarna, A. Eur. J. Org. Chem.
2004, 4621; (c) Trabocchi, A.; Occhiato, E. G.; Potenza, D.; Guarna, A. J. Org.
Chem 2002, 67, 7483.
10. Sladojevich, F.; Trabocchi, A.; Guarna, A. J. Org. Chem. 2007, 72, 4254.
11. Krchnák, V.; Vágner, J.; Sáfar, P.; Lebl, M. Collect. Czech. Chem.Commun. 1988, 53,
2542.
4.5. Docking calculations
Automated docking studies were carried out by the Autodock
4.0.1 program,25 using the Lamarckian Genetic Algorithm (LGA)
as a search engine. The AutoDockTools 1.4.5 (ADT) graphical inter-
face27 was used to prepare the receptor and the ligands PDBQT
files. The coordinates of the ligands 1 and 2 were retrieved from
the lowest energy conformers resulting from calculation using
NMR data, whereas the coordinates of avb3 receptor was retrieved
from the Protein Data Bank (PDB code: 1L5G), and the ligand–pro-
tein complex was unmerged for achieving free receptor structure.
Water molecules were removed. For the protein receptor and li-
gands 1 and 2, all hydrogens were added, Gasteiger charges were
computed, and non-polar hydrogens were merged. A charge value
of +2.0 to each Mn atom of the protein receptor was successively
added. Three-dimensional energy scoring grids of 0.375 Å resolu-
tion and 40 Å Â 40 Å Â 40 Å dimensions were computed. The cen-
ter of the grid was set to be coincident with the mass center of
the ligands preliminary fitted on the X-ray structure of
12. Belkin, V. M.; Belkin, A. M.; Koteliansky, V. E. J. Cell Biol. 1990, 111, 2159.
13. Pytela, R.; Pierschbacher, M. D.; Argraves, S.; Suzuki, S.; Ruoslahti, E. Methods
Enzymol. 1987, 144, 475.
14. Kumar, C. C.; Nie, H.; Rogers, C. P.; Malkowski, M.; Maxwell, E.; Catino, J. J.;
Armstrong, L. J. Pharmacol. Exp. Ther. 1997, 283, 843.
15. (a) Dijkgraaf, I.; Kruijtzer, J. A.; Frielink, C.; Soede, A. C.; Hilbers, H. W.; Oyen, W.
J.; Corstens, F. H.; Liskamp, R. M.; Boerman, O. C. Nucl. Med. Biol. 2006, 33, 953;
(b) Zanardi, F.; Burreddu, P.; Rassu, G.; Auzzas, L.; Battistini, L.; Curti, C.; Sartori,
A.; Nicastro, G.; Menchi, G.; Cini, N.; Bottoncetti, A.; Raspanti, S.; Casiraghi, G. J.
Med. Chem 2008, 51, 1771.
16. Takagi, J.; Petre, B. M.; Walz, T.; Springer, T. A. Cell 2002, 110, 599.
17. Humphries, M. J.; McEwan, P. A.; Barton, S. J.; Buckley, P. A.; Bella, J.; Mould, A.
P. Trends Biochem. Sci. 2003, 28, 313.
c[RGDf(Me)V] in the avb3 complex (1L5G). A total of 50 runs with
a maximum of 2,500,000 energy evaluations were carried out for
each ligand, using the default parameters for LGA. Cluster analysis
was performed on the docked results using a root-mean-square
(rms) tolerance of 1.5 Å. The analysis of the binding mode, the cal-
culation of the binding energy and the prediction of the binding
activity of the docked conformations were carried out using PyMol
Autodock Tools plugin within PyMol software.28
18. Shimaoka, M.; Springer, T. A. Nat. Rev. Drug Discov. 2003, 2, 703.
19. Xiong, J. P.; Stehle, T.; Zhang, R.; Joachimiak, A.; Frech, M.; Goodman, S. L.;
Arnaout, M. A. Science 2002, 296, 151.
20. Bednar, R. A.; Gaul, S. L.; Hamill, T. G.; Egbertson, M. S.; Shafer, J. A.; Hartman,
G. D.; Gould, R. J.; Bednar, B. J. Pharmacol. Exp. Ther. 1998, 285, 1317.
21. Mohamadi, F.; Richards, N. G. J.; Guida, W. C.; Liskamp, R.; Lipton, M.; Caufield,
C.; Chang, G.; Hendrickson, T.; Still, W. C. J. Comput. Chem. 1990, 11, 440.
22. Weiner, S. J.; Kollman, P. A.; Nguyen, D. T.; Case, D. A. J. Comput. Chem. 1986, 7,
230.
23. Still, W. C.; Tempczyk, A.; Hawley, R. C.; Hendrickson, T. J. Am. Chem. Soc. 1990,
112, 6127.
24. Chang, G.; Guida, W. C.; Still, W. C. J. Am. Chem. Soc 1989, 111, 4379.
25. Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart, W. E.; Belew, R. K.;
Olson, A. J. J. Comput. Chem. 1998, 19, 1639.
26. Kumar, C. C.; Nie, H.; Armstrong, L.; Zhang, R.; Vijay-Kumar, S.; Tsarbopoulos,
A. FEBS Lett. 1998, 429, 239.
27. Gillet, A.; Sanner, M.; Stoffler, D.; Olson, D. Structure 2005, 13, 483.
28. DeLano, W.L. The PyMOL Molecular Graphics System, DeLano Scientific LLC:
San Carlos, CA, USA. http://www.pymol.org.
Acknowledgements
University of Florence, CINMPIS and MIUR are acknowledged for
financial support. Dr. Filippo Sladojevich is acknowledged for carry-
ing out preliminar synthetic experiments. Ente Cassa di Risparmio is
acknowledged for the granting the 400 MHz NMR spectrometers.
References and notes
1. (a) Arnaout, M. A.; Goodman, S. L.; Xiong, J. P. Curr. Opin. Cell Biol. 2002, 14, 641;
(b) Hynes, R. O. Cell 2002, 110, 673; (c) Plow, E. F.; Haas, T. A.; Zhang, L.; Loftus,
J.; Smith, J. W. J. Biol. Chem. 2000, 275, 21785; (d) Humphries, M. J. Biochem. Soc.
Trans. 2000, 28, 311.