Anthranilic acid derivatives: A new class of non-peptide CCK1 receptor antagonists

Article abstract of DOI:10.1016/S0968-0896(02)00475-3

Having successfully obtained new CCK₁ ligands holding appropriate groups on the anthranilic acid dimer used as molecular scaffold we were interested in increasing their micromolar affinity for the CCK₁ receptors by modifying the spatial relationship of the main pharmacophoric groups. Since, we have proposed simplified analogues reducing the anthranilic acid dimer to a monomer. In this stage of our research program we have prepared and tested on CCK receptors a series of N-substituted anthranilic acid derivatives keeping a Phe residue at the C-terminal site. The indole-2-carbonyl group imparts the best CCK₁ receptor binding affinity (compound 1: IC₅₀=197.5 nM) while a sharp decrease in binding affinity is observed for the other indole containing derivatives. Moreover, in order to support the different binding behaviour observed for the synthesized compounds, a conformational investigation was carried out. Finally, on the basis of the main pharmacophoric groups of the obtained new lead compound (1) (coded VL-0395) a receptor binding hypothesis has been provided.

Full text of DOI:10.1016/S0968-0896(02)00475-3

Bioorganic & Medicinal Chemistry 11 (2003) 741–751
Anthranilic Acid Derivatives: A New Class of Non-Peptide CCK1
Receptor Antagonists
Antonio Varnavas,a,* Lucia Lassiani,a Valentina Valenta,a Federico Berti,b
Laura Mennunic and Francesco Makovecc
aDepartment of Pharmaceutical Sciences, University of Trieste, P.le Europa 1, 34127 Trieste, Italy
bDepartment of Chemical Sciences, University of Trieste, Via Giorgieri 1, 34127 Trieste, Italy
cRotta Research Laboratorium S.p.A., Via Valosa di Sopra 7/9, 20052 Monza (MI), Italy
Received 2 August 2002; revised 13 September 2002; accepted 25 September 2002
Abstract—Having successfully obtained new CCK1 ligands holding appropriate groups on the anthranilic acid dimer used as
molecular scaffold we were interested in increasing their micromolar affinity for the CCK1 receptors by modifying the spatial rela-
tionship of the main pharmacophoric groups. Since, we have proposed simplified analogues reducing the anthranilic acid dimer to a
monomer. In this stage of our research program we have prepared and tested on CCK receptors a series of N-substituted anthra-
nilic acid derivatives keeping a Phe residue at the C-terminal site. The indole-2-carbonyl group imparts the best CCK1 receptor
binding affinity (compound 1: IC50=197.5 nM) while a sharpdecrease in binding affinity is observed for the other indole containing
derivatives. Moreover, in order to support the different binding behaviour observed for the synthesized compounds, a conforma-
tional investigation was carried out. Finally, on the basis of the main pharmacophoric groups of the obtained new lead compound
(1) (coded VL-0395) a receptor binding hypothesis has been provided.
# 2002 Elsevier Science Ltd. All rights reserved.
Introduction
In the last decade a large number of non-peptide com-
Cholecystokinin (CCK) is an endogenous peptide used
as a messenger from both cell chemical signaling sys-
tems (endocrine and nervous system).1 The major hor-
monal effects of CCK are expressed in the
gastrointestinal (GI) tract.2,3 At the central nervous
system (CNS) CCK acts as neurotransmitter/neuro-
modulator and appears to be involved in different neu-
ropathological situations.4,5 The entire range of
peripheral and central biological actions of CCK is
mediated by two distinct receptor types: CCK1 and
CCK2.6 The human CCK1 and CCK2 receptors have
both been cloned and they have shown to belong to the
G-protein coupled receptors family.7,8 Among the dif-
ferent molecular forms of CCK only the C-terminal
octapeptide fragment (Asp-Tyr-Met-Gly-Trp-Met-Asp-
PheNH2 or CCK-8) binds both receptors with nanomo-
lar affinity. On the other hand, the smallest bioactive
fragment of CCK (Trp-Met-Asp-PheNH2 or CCK-4)
binds with nanomolar affinity the CCK2 receptor while
its affinity towards the CCK1 receptor decreases to the
millimolar range.9
pounds known as CCK receptor antagonists have been
developed.10,11 CCK1 antagonists have been shown to
block some GI CCK induced actions such as the gall-
bladder contraction and the pancreatic secretion.12,13
CCK2 receptor antagonists have attracted a major
interest because of their anxiolytic effects.14
During our ongoing laboratory program aimed at dis-
covering new non-peptide compounds targeted to the
CCK receptors, we have proposed a new class of non-
peptide CCK1 ligands holding appropriate pharmaco-
phoric groups on the anthranilic acid dimer used as a
molecular scaffold.15,16 These compounds are endowed
with micromolar affinity for the CCK1 receptors. They
have been designed using a strategy which combines the
classical disconnection approach with the ‘rational
design’ based on the chemical structure of the endogen-
ous peptide. In fact, the anthranilic acid dimer was
obtained as the main synthon of an our disconnection
strategy applied to the first natural non-peptide CCK
ligand Asperlicin.17 On the other hand, the ‘appended’
pharmacophoric groups were selected from the C-terminal
tetrapeptide sequence of the endogenous ligand (CCK-4).
This part of our work has produced a lead compound
*Corresponding author. Tel.:+39-040-558-7861; fax:+39-040-52572;
e-mail: varnavas@univ.trieste.it
0968-0896/03/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00475-3

742
Antonio Varnavas et al. / Bioorg. Med. Chem. 11 (2003) 741–751
coded VL-0494 (d,l-{2-[2-(3-1H-indol-3-yl-propionyl-
amino)-benzoylamino]-benzoyl-amino}-3-phenyl-pro-
pionic acid)18 characterized by the presence of Phe and
desaminated Trpresidues at the C- and N-termini of
anthranilic acid dimer. These two residues represent
‘address’ type components and were selected from the
CCK-4 structure (Fig. 1). Despite these common struc-
tural features, the affinity of VL-0494 for the CCK1
receptor was three orders of magnitude greater than
that of CCK-4.
spatial relationship of the pharmacophoric groups.
Therefore, we have directed our synthetic efforts
towards the preparation of simplified analogues of VL-
0494 reducing the anthranilic acid dimer to a monomer.
In this preliminary stage of our research program we
have prepared a series of N-substituted anthranilic acid
derivatives keeping a Phe residue at the C-terminal site
as in the case of the lead compound (Table 1).
The substituents selected to be appended via amide
bond formation at the N-terminus of anthranilic acid
are: bicyclic aromatic or heteroaromatic systems (naph-
thalene or indole containing substituents) (1–8), a
phenyl ring linked by a short saturated alkyl chain (9–
10) and bulky or branched alkyl groups such as tert-
butyl and 1-adamantyl groups.
To explain this difference, we undertook a molecular
modelling study on VL-0494. This study showed that
the main difference concerned the tetrapeptide back-
bone and the anthranilic acid dimer structure. In
particular, the conformational preference of VL-0494
was a staggered U-shape, whereas the CCK-4 assumed
an S-like bend conformation. In the proposed graphical
model (Fig. 2) of the interaction of VL-0494 at the
CCK1 receptor binding sites compared to that of CCK-
4 toward CCK2 receptor,19,20 we have hypothesised the
presence of two lipophilic pockets in which both ligands
interact with the same pharmacophoric groups.
Chemistry
The synthetic pathway to anthranilic acid derivatives 1–14
employing standard procedures is shown in Scheme 1. The
key intermediate 15 was prepared in good yield (80%)
via ortho-amino benzoylation of the dl-phenylalanine
ethyl ester with isatoic anhydride. Derivatives 19, 20
were obtained by coupling the intermediate 15 with the
corresponding acid that was previously activated by the
mixed anhydride method. Compounds 16, 18 and 29
were obtained via acyl chloride formation. Compounds
21–28 were obtained using commercially available acyl
chlorides.
Although VL-0494 and CCK-4 have a similar ‘regnylogi-
cal type’ bioactive conformation, a reversal in selectiv-
ity to the CCK receptors has been noticed. This
behaviour has already been described in the CCK chemi-
cal literature.21 Nevertheless, in the case of VL-0494, it
is possible that the conformational difference of the
anthranilic acid dimer in comparison to the central
portion of the backbone of CCK-4, could produce the
increase of the CCK1 receptor affinity of the lead com-
pound.
Having successfully obtained new CCK1 ligands from
the native peptide and/or from asperlicin we were inter-
ested in increasing the CCK1 affinity by modifying the
Figure 2. Hypothetic model of the ‘regnylogic type’ interaction of
CCK-4 and VL-0494 at the CCK2 and CCK1 receptors, respectively.
Table 1. Structure of the target compounds
Compd
R
Compd
R
1
2
3
4
5
6
7
2-Indolyl-CO–
5-Indolyl-CO–
3-Indolyl-CO–
8
9
2-Naphthyl–SO2–
C6H5-CH2–CO–
C6H5-(CH2)2–CO–
C(CH3)3–CO–
C(CH3)3–CH2–CO–
1-Adamantyl-CO-
1-Adamantyl-CH2-CO-
10
11
12
13
14
3-Indolyl-CH2-CO–
3-Indolyl-(CH2)2-CO–
1-Naphthyl-CO–
1-Naphthyl-SO2–
Figure 1. Compound VL-0494 derived from asperlicin as well as from
CCK-4.

Antonio Varnavas et al. / Bioorg. Med. Chem. 11 (2003) 741–751
743
Indole-5-carboxylic acid was condensed with the inter-
groupimparts the best CCK
receptor binding affinity
1
mediate 15 using the Mukaiyama reagent22 to afford
compound 17. The free acids 1–14 were obtained in
almost quantitative yields by base catalyzed hydrolysis
of the corresponding ethyl esters 16–29. Synthetic data
of the esters 16–29 and of the target compounds 1–14
are listed in the Experimental.
and it is therefore the optimum N-terminus group in
this series.
In fact, compound 1 (IC50=197.5 nM) was found to
be ten times more potent than the starting lead
compound VL-0494 (IC50=2300 nM).18 However, a
sharpdecrease in binding affinity is observed for both
indole-5 and indole-3-carbonyl groups (2 and 3)
(ISB=41% at 10 and IC50=1870 nM, respectively).
An additional decrease in affinity was observed for
the higher homologues of compound 3 (4 and 5)
(ISB=22% at 10 and 17,800 nM, respectively). More-
over, it is interesting to note that, among these indole
derivatives, the indole-3-propionyl moiety exerts only a
weak affinity, whereas it was found to be the optimum
groupin the N-substituted anthranilic acid dimer deri-
vatives (VL-0494).
Results and Discussion
Table 2 summarizes the results of the CCK1 receptor
affinity of compounds 1–14 expressed as IC50 or as per-
centage of inhibition (ISB%) determined at the highest
used dose (10 or 30 mM, as indicated). With regard to
the affinity with the central CCK2 receptor, the tested
compounds showed low affinity to this receptor sub-
type. Furthermore, as previously observed for the
anthranilic acid dimer derivatives, only few compounds
containing the indole moiety exhibited affinity toward
CCK1 receptors. In particular, the indole-2-carbonyl
The introduction of the 1-naphthyl or 2-naphthyl group
either by an amidic or sulfonamidic bond to the phenyl
ring of the anthranilic acid led to a complete inactivity
at the maximum dose tested 6, 7 and 8). Furthermore,
phenyl substituted compounds (9 and 10) displayed a
very low affinity for CCK1 receptors compared with the
corresponding indolic derivatives (4 and 5).
In the same way, the introduction of bulky aliphatic
groups led to inactive compounds (11–14). Nevertheless,
while the present investigation does not represent an
exhaustive study on the N-terminal optimisation of the
anthranilic acid derivatives, the difference in affinity of
the compounds containing the indole moiety in
comparison to the others, is remarkably consistent.
In order to support the different binding behaviour
observed for the synthesized compounds, a conforma-
tional investigation was carried out. The anthranilic
core of the molecules was first analyzed carefully. When
Scheme 1. Reagents and reaction conditions: (i) AcOEt, reflux, 2 h;
(ii) RCOOH, iBuOCOCl, Et3N, CH2Cl2 dry; (iii) RCOCl, Et3N,
CH2Cl2 dry; (iv) RCOOH, 2-chloro-1-methylpyridinium iodide
(Mukaiyama reagent), Et3N, CH2Cl2 dry; (v) KOH, MeOH.
Table 2. CCK receptors Binding Data
a
Compd.
R
IC50 mM
d
Rat pancreatic acini (CCK1)
Guinea pig brain cortex (CCK2
)
1
2
3
4
5
6
7
8
2-Indolyl-CO–
5-Indolyl-CO–
3-Indolyl-CO–
0.197 (0.131–0.298)
41% ISBb
1.87 (1.48–2.38)
16.40
NT
30.20
3-Indolyl-CH2-CO–
3-Indolyl-(CH2)2-CO–
1-Naphthyl-CO–
1-Naphthyl-SO2–
2-Naphthyl-SO2–
C6H5–CH2–CO–
C6H5–(CH2)2–CO–
C(CH3)3–CO–
INb
23% ISBc
13.40
17.85 (12.00–26.56)
INb
INb
6.57 (3.06–14.10)
5.90 (3.94–8.84)
7.31 (3.90–13.60)
36% ISBc
37% ISBc
24% ISBc
37% ISB c
14.50
INb
9
INb
10
11
12
13
14
INb
INb
C(CH3)3–CH2–CO–
1-Adamantyl-CO–
1-Adamantyl-CH2-CO–
INb
INb
51% ISBc
15.40
IN, inactive; NT, not tested.
aIC50: mM concentration and p=0.05 fiducial limits required to inhibit by 50% the specific binding of 25 pM [125I]-(BH)-CCK8%; ISB: percentage
inhibition of specific binding at the maximal concentration tested.
b10 mM.
c30 mM.
dValues without fiducial limits were obtained from not more than two experiments.

744
Antonio Varnavas et al. / Bioorg. Med. Chem. 11 (2003) 741–751
this residue is inserted within a pseudopeptidic chain as
a non-natural amino acid, it shows the tendency to form
an intramolecular hydrogen bond between the anthra-
nilic amide hydrogen and the carbonyl oxygen. On the
basis of this structural feature, a number of oligo-
anthranil amides have been prepared in the past, and
used as molecular scaffolds.23 This tendency is also
shown by our ligands, as it can be deduced by their
NMR properties, and it is confirmed by the modelling
results. Nevertheless, the resulting pseudocyclic struc-
tures are quite strained, and low-level semi-empirical
methods such as AM1 and PM3 resulted to be prone to
compensate steric strain by a rather erratic pyr-
amidalization of one of the two amide nitrogens of the
core. This is not confirmed by the known X-ray struc-
tures of related compounds,23 where the anthranilic
nitrogen is always full planar, and only a slight pir-
amidalization of the carboxamide nitrogen is observed
in a single case. For this reason a reference compound,
namely the N-acetyl-anthranilic acid N-methyl amide
was submitted to ab-initio optimization. At the HF/3-
21/G(d) level the anthranilic nitrogen is planar, while
the other nitrogen is indeed piramidal (sum of bond
angles at nitrogen 354.4ꢀ); by raising the level of theory
upto HF/6-31/G** the whole molecule becomes fully
planar.
between the two aromatic ends of the molecule is 9.5 A,
and the angle between the planes containing the two
rings is about 114ꢀ. Another local minimum can be
obtained by rotating the N–Ca bond of Phe by about
60ꢀ. This minimum is 3.8 kcal/mol higher in energy, and
the two aromatic rings are closer. The average distance
is less than 8 A for this conformation.
The lowest energy conformer of the indole-5-carbonyl
derivative (2) shows an almost identical geometry to
that of compound 1 (Fig. 4), but the intramolecular
hydrogen bond of the anthranilic core of the molecule
lies on the opposite side of the plane containing the
peptide bond with the Phe residue. Nevertheless, a con-
formation with the hydrogen bond on the same side as
in the case of compound 1 is easily accessible, being only
0.1 Kcal/mol higher in energy.
The distance between the aromatic ends in the two
conformations is almost the same, and it is very similar
to the distance obtained for the 2-indole carboxylic acid
derivative.
Assuming the same binding modalities for compounds 1
and 2, the higher affinity of the former is probably due
to the different spatial position of the indole nitrogen. In
particular we have supposed that the hydrogen attached
to the indole nitrogen of compound 1 may have a direct
interaction with the receptor via a hydrogen bond.
With this result in mind, the entire set of molecules was
resubmitted to an extensive conformational search at
the AM1 level, but all the structures not planar at the
anthranil nitrogen were discharged, while the phenyl-
alanine nitrogen was allowed to result slightly piramidal
in several conformational minima.
After modelling compound 3 it was possible to find a
single conformation family (Fig. 5), which is basically
similar to the minimum of the indole-2-carbonyl deri-
vative (1). The hydrogen bond is preserved, and the
The conformational search performed on compound 1
gave two families of very similar U-shaped conforma-
tions. The intramolecular hydrogen bond is found in
both the conformations (H–O distance 2.13 A), and
both the indole ring and the aromatic ring of phenyl-
alanine lie on the same side of the molecule. The two
conformations differ in the orientation of the indole
nitrogen, which is in a trans-like position with respect to
the anthranilic nitrogen in the absolute minimum (Fig. 3).
In the other conformation, which is 1.8 kcal/mol higher
in energy, the indole ring is rotated by 180ꢀ and its
nitrogen is in a cis-like position. A better dipole moment
is seen to favour the first conformation (2.756 D vs
3.609 D). In the absolute minimum the average distance
Figure 4. Lowest energy conformer of compound 2 (heavy atoms
only).
Figure 3. Lowest energy conformer of compound 1 (heavy atoms
only).
Figure 5. Lowest energy conformer of compound 3 (heavy atoms
only).

Antonio Varnavas et al. / Bioorg. Med. Chem. 11 (2003) 741–751
745
aromatic rings (indole and Phe) lie on the same side of
the molecule at an average distance of about 9.5 A. The
indole ring also, is placed like a ‘flag’ related to the
amido group in the 3 position and its plane is perpen-
dicular to that of the anthranilic acid. In this case the
space available to the indole ring by rotation around its
C3–CO bond is clearly much larger than in the 2-indole
derivative. The difference can be estimated as about 180
A3 in favour of the 3-indole derivative. In contrast, the
indole ring of the 2-derivative reaches a major distance
from the anthranilic core of the molecule (8.4 vs 6.3 A).
was carried out excluding solvent interactions, this
conformation seems to be probable in aqueous solution
according to the ‘hydrophobic collapse’ phenomenon.
The conformational analysis of the alkyl groupcon-
taining derivatives (11–13) is straightforward. They
have a single minimum, with the intramolecular hydro-
gen bond on the anthranilic residue, and the bulky alkyl
groupon the same side of the Phe residue, as a
consequence of the trans geometry of the amide bonds.
Finally, the lowest energy conformations obtained for
the 1-naphthyl (6 and 7) and 2-naphthyl derivatives (8)
resemble that of compounds 3 and 1, respectively. Thus,
the conformational effect could not be evoked to
explain either the loss in CCK1 affinity or their reverse
selectivity for the CCK2 receptor subtype. It is possible,
therefore, that information on the CCK2 binding sites
may be gained from further studies of these derivatives.
From a fitting performed between compound 1 and 3
matching the common anthranoyl-Phe framework of
the two molecules it is possible to observe the quite dif-
ferent spatial arrangement of the two indole rings (Fig. 6).
The lower affinity of compound 3 suggests that the
indole-3 substituent cannot satisfy completely the con-
formational restraints imposed by the second hydrophobic
pocket of the receptor.
Conclusion
The geometry of the indole-3-propionic derivative (5) is
very different from the geometries of the carbonyl-
indole derivatives (1–3) (Fig. 7). The intramolecular
hydrogen bond in the anthranilic residue is still pre-
served, but the U-shaped conformation is lost. The
indole side of the molecule is bending back towards the
Phe residue. This is most likely due to the formation of
a new intramolecular hydrogen bond between the indole
NH and the Phe carboxylic group, as well as to the
p-stacking interactions between the indole and the
anthranilic rings, which lie in parallel planes at 3.5 A.
Although the conformational search on this compound
In conclusion, during a program aimed at searching for
CCK receptor antagonists, we have found that simpli-
fying the anthranilic acid dimer scaffold to a monomer
gives rise to a structurally simple compound endowed
with high affinity towards CCK1 receptors. The new
lead compound obtained (compound 1 or VL-0395)
could represent a new starting point for the develop-
ment of an innovative class of non-peptide CCK1
receptor antagonists.
The antagonist nature of VL-0395 was confirmed by an
in vivo functional test. The compound, intravenously
administered, inhibited the guinea pig gallbladder con-
tractions induced by CCK8 sulphated (8.8 pmol/kg iv),
with an ED50 of 0.38 mmol/kg (0.11–1.36; p=0.05 fidu-
cial limits). Its potency was comparable with that
exhibited by the reference CCK1 selective antagonist
Loxiglumide (n. 5610 Merck Index, 13th Edition),24 that
was active, under the same experimental condition, at
0.24 mmol/kg (0.13–0.41; p=0.05 fiducial limits).
Moreover, on the basis of the main pharmacophoric
groups (indole moiety and phenyl ring of Phe) shared by
VL-0395, a receptor binding hypothesis, similar to that
described for the anthranilic acid dimer derivatives has
been provided. In fact, this model of binding (depicted
in Fig. 8) requests at least two hydrophobic pockets to
accommodate the main pharmacophoric groups. Each
binding pocket is about 5 A deepand the distance
Figure 6. Superimposition (heavy atoms only) of the preferential con-
formation of compound 1 (red) and compound 3 (green).
Figure 8. Schematic representation of the interactions of the main
pharmacophoric groups of compound 1 with the CCK1 receptor.
Figure 7. Lowest energy conformer of compound 5 (heavy atoms
only).

746
Antonio Varnavas et al. / Bioorg. Med. Chem. 11 (2003) 741–751
between them is about 9.5 A. However, the proposed
temporary model will be integrated when further
experimental data will be available regarding the addi-
tional binding contribution due to the anthranilic
phenyl ring as well as to the free carboxy terminal
group.
over anhydrous sodium sulphate and concentrated in
vacuo. Trituration with petroleum ether 40–70ꢀ affor-
ded the analytically pure title compound in 80% yield.
Rf 0.69 (AcOEt/hexane : 1/1); mp84–85 ꢀC; H NMR
1
(CDCl3): d1.24 (t, 3H, –CH3); 3.21 (m, 2H, –CH2–
CH<); 4.18 (q, 2H, –CH2–O–); 4.97 (m, 1H, >CH–);
5.45 (s, 2H, –NH2); 6.52 (d, 1H, –NH–); 6.61–7.28 (m,
9H, ar). 13C NMR (CDCl3): d14.19, 38.05, 53.20, 61.63,
115.41, 116.69, 117.29, 127.15, 127.43, 128.60, 129.43,
132.59, 136.04, 148.84, 168.66, 171.75.
Although the importance of one of the pharmacophoric
groups considered (2-indole ring) in the interaction with
the CCK1 receptors has already been stressed by the 1,4-
benzodiazepine antagonists25 we have demonstrated
that the same receptor, probably with different binding
sites, is perfectly able to host anthranilic acid derivatives
structurally unrelated to the previous ones.
General coupling procedures of anthranoyl-DL-
phenylalanine ethyl ester
Method A (19, 20). A solution of 10 mmol of the corre-
sponding acid in 100 mL of dry dichloromethane cooled
to À10 ꢀC was treated with triethylamine (1.4 mL, 10
mmol) followed by isobutyl chloroformate (1.31 mL, 10
mmol). The resulting mixture was stirred at À10 ꢀC for
20 min and treated dropwise with a solution of com-
pound 15 (3.12 g, 10 mmol) in 50 mL of dry dichloro-
methane. After the addition was complete, the reaction
was stirred at room temperature for 1 h and then
refluxed for 3 h. The solvents were evaporated, the resi-
due was dissolved in dichloromethane, the organic layer
was washed with diluted aqueous sodium hydroxide
solution and with water, dried (sodium sulphate), and
evaporated to dryness. Crystallization from MeOH
yielded compound 19 while in the case of compound 20
the residue was used without further purification in the
next stepof the synthesis.
Finally, in order to obtain more information on the
binding modalities of our new lead compound, its main
structural features (indole moiety, anthranilic acid,
aminoacid side chain, and the free carboxy group) are
currently under structural modifications and analysis.
Experimental
Chemical procedures
All chemicals and solvents used in syntheses were
reagent-grade products and were used without addi-
tional purification. Reaction progress was monitored by
ascending thin-layer chromatography (TLC) using pre-
coated silica gel plates (60F-254 Merck). Visualization
of the chromatograms was achieved by short wave UV
light (254 nm). Melting points were determined on a
Buchi 510 melting point apparatus (Buchi, Flawil,
Switzerland) and are uncorrected. Preparative medium-
pressure chromatography (MPLC) was performed on a
Buchi 688 apparatus using silica gel (Merck Kieselgel
60, 15–40 mm). Proton (1H NMR, 200 MHz) and carbon
(13C NMR, 50 MHz) nuclear magnetic resonance spec-
tra were recorded on a Varian-Gemini 2000 Fourier
Transform spectrometer using CDCl3 or (CD3)2SO as
solvent. Chemical shifts were reported as parts per mil-
lion (ppm, d units) downfield from an internal Me4Si
standard. Splitting patterns are designated as s, singlet;
d, doublet; t, triplet; q, quartet; m, multiplet and b,
broad. 13C NMR spectra were determined using either
the Attached Proton Test (APT) or standard 13C pulse
sequence parameters. Spectral data are consistent with
assigned structures. Mass spectra were recorded on an
API-1 Perkin-Elmer SCIEX spectrometer by electrospray
ionisation (ES).
2(R,S)-[2-(2-1H-Indol-3-yl-acetylamino)-benzoylamino]-
3-phenyl-propionic acid ethyl ester (19). Yield 72%;
Molecular formula: C28H27N3O4; TLC (AcOEt/hexane:
ꢀ
1
1/1)Rf: 0.43; mp130–131 C; H NMR (CDCl3) d 1.28
(t, 3H, –CH3); 3.01 (m, 2H, –CH2–CH<); 3.89 (s, 2H,
–CH2–CO–); 4.19 (q, 2H, –CH2–O–); 4.65 (m, 1H,
–CH<); 6.44 (d, 1H, –NH–CH<); 6.96–7.64 (m, 13H,
ar); 8.34 (s, 1H, –NH– ind); 8.59 (d, 1H, ar); 10.68 (s,
1H –NH–).; 13C NMR (CDCl3) d 14.19, 35.33, 37.65,
53.41, 61.77, 108.79, 111.22, 118.86, 119.80, 121.53,
122.31, 122.89, 123.95, 126.49, 127.23, 128.57, 129.38,
132.58, 135.63, 140.20, 143.10, 168.70, 171.50, 172.32.
Method B (16, 18, 21–29). A solution of 10 mmol of the
corresponding acyl chloride (indole-2-carbonyl chloride
was prepared according to the procedure described by
Kermack26 while indole-3-carbonyl chloride by a stan-
dard method using PCl5 in dry dichloromethane) in 30
mL of dry dichloromethane was gradually added to a
solution of compound 15 (3.12 g, 10 mmol) in the same
solvent (50 mL). The pH of the reaction was adjusted to
9.5 with triethylamine and the solution stirred for 2 h.
The reaction mixture was diluted with 100 mL of di-
chloromethane and washed in succession with 0.1 N
NaOH, water, 0.1 N HCl and water. The dried (sodium
sulphate) organic phase was rotary evaporated and the
residue was purified as described to yield the titled
compounds.
Syntheses
2(R,S)-(2-Amino-benzoylamino)-3-phenyl-propionic acid
ethyl ester (15). A suspension of dl-phenylalanine ethyl
ester hydrochloride (5.00 g, 21.8 mmol) in 150 mL of
ethyl acetate was treated with triethylamine (3.07 mL,
21.8 mmol) followed by isatoic anhydride (3.55 g, 21.8
mmol). The resulting mixture was refluxed under stir-
ring for 2 h, cooled to room temperature and filtered.
The organic phase was thoroughly washed with 1 M
NaOH (2Â50 mL), water (2Â50 mL) and brine, dried
2(R,S)-{2-[(1H-Indole-2-carbonyl)-amino]-benzoylamino}-
3-phenyl-propionic acid ethyl ester (16). Trituration with

Antonio Varnavas et al. / Bioorg. Med. Chem. 11 (2003) 741–751
747
hot MeOH afforded the titled compound in 78% yield.
Molecular formula: C27H25N3O4; TLC (AcOEt/hexane:
1/1)Rf:0.69; mp210–211 ꢀC; 1H NMR (DMSO-d6) d
1.17 (t, 3H, –CH3); 3.20 (m, 2H, –CH2–CH<); 4.11 (q,
2H, –CH2–O–); 4.79 (m, 1H, –CH<); 6.98 (s, 1H, ind);
7.06–7.82 (m, 12H, ar); 8.64 (d, 1H, ar); 9.30 (d, 1H,
–NH—CH<); 11.95 (s, 1H, –NH–ind); 12.15 (s, 1H
–NH–); 13C NMR (DMSO-d6) d 14.80, 36.83, 55.14,
61.56, 103.32, 113.22, 119.98, 120.63, 120.95, 122.49,
123.30, 124.82, 127.25, 127.62, 128.94, 129.18, 129.81,
132.19, 133.38, 137.81, 138.16, 139.82, 159.77, 169.58,
171.91.
2H, –CH2–O–); 4.81 (m, 1H, –CH<); 6.46 (d, 1H, –
NH–CH<); 6.94–8.41 (m, 16H, ar); 10.90 (s, 1H –NH–
); 13C NMR (CDCl3) d 14.23, 37.67, 53.50, 61.96,
120.60, 121.17, 122.41, 123.66, 126.93, 127.34, 127.52,
127.93, 128.67, 128.80, 128.87, 129.33, 132.06, 133.08,
134.89, 135.47, 136.53, 139.04, 167.67, 171.01.
3-Phenyl-2(R,S)-(2-phenylacetylamino-benzoylamino)-
propionic acid ethyl ester (24). The residue was used
without further purification in the next step of the
synthesis.
3-Phenyl-2(R,S)-[2-(3-phenyl-propionylamino)-benzoyl-
amino]-propionic acid ethyl ester (25). Trituration with
petroleum ether 40–70ꢀ afforded the titled compound in
82% yield; Molecular formula: C27H28N2O4; TLC
2(R,S)-{2-[(1H-Indole-3-carbonyl)-amino]-benzoylamino}-
3-phenyl-propionic acid ethyl ester (18). Preparative
medium pressure chromatography (MPLC) (silica gel,
AcOEt/hexane: 1/3) afforded the titled compound in
20% yield. Molecular formula: C27H25N3O4; TLC
(AcOEt/hexane: 1/1)Rf: 0.30; mp147–149 C; H NMR
(DMSO-d6) d 1.11 (t, 3H, –CH3); 3.14 (m, 2H, –CH2–
CH<); 4.04 (q, 2H, –CH2–O–); 4.65 (m, 1H, –
CH<); 7.07–8.60 (m, 14H, ar); 9.17 (d, 1H, –NH–
CH<); 11.46 (s, 1H, –NH– ind); 11.81 (s, 1H –NH–);
13C NMR (DMSO-d6) d 13.75, 35.92, 54.06, 60.50,
106.64, 111.09, 119.71, 120.90, 120.99, 121.16, 122.21,
122.67, 125.24, 125.42, 126.21, 127.94, 128.81, 131.65,
133.90, 137.25, 138.40, 160.81, 168.23, 170.99.
(AcOEt/hexane: 1/1) Rf: 0.72; mp73–75 ꢀC; H NMR
1
(CDCl3) d 1.29 (t, 3H, –CH3); 2.71 (t, 2H, –CH2–Ph);
3.06 (t, 2H, –CH2–CO–); 3.24 (m, 2H, –CH2–CH<);
4.24 (q, 2H, –CH2–O–); 4.98 (m, 1H, >CH–); 6.72 (d,
1H, –NH–CH<); 7.03–7.50 (m, 13H, ar); 8.62 (d,1H,
ar); 10.93 (s, 1H, –NH–); 13C NMR (CDCl3) d 14.22,
31.42, 37.91, 39.97, 53.60, 61.92, 119.62, 121.48, 122.79,
126.19, 126.66, 127.34, 128.34, 128.41, 128.52, 128.70,
129.38, 132.95, 135.65, 139.67, 140.76, 168.34, 170.97,
171.27.
ꢀ
1
2(R,S)-[2-(2,2-Dimethyl-propionylamino)-benzoylamino]-
3-phenyl-propionic acid ethyl ester (26). Trituration with
hot MeOH afforded the titled compound in 50% yield;
Molecular formula: C23H28N2O4; TLC (AcOEt/hexane:
1/1) Rf: 0.64; mp107 ꢀC; 1H NMR (CDCl3) d 1.26
(t, 3H, –CH3); 1.31 (s, 9H, –C(CH3)3); 3.23 (m, 2H, –
CH2–CH<); 4.22 (q, 2H, –CH2–O–); 5.03 (m, 1H,
>CH–); 6.71 (d, 1H, –NH–CH<); 6.97–8.66 (m, 9H,
ar); 11.16 (s, 1H, –NH–); 13C NMR (CDCl3) d 14.20,
27.61, 37.80, 40.19, 53.58, 61.83, 119.88, 121.43, 122.55,
126.63, 127.32, 128.67, 129.35, 132.87, 135.71, 140.15,
168.53, 171.25, 177.73.
2(R,S)-{2-[(Naphthalene-1-carbonyl)-amino]-benzoyl-
amino}-3-phenyl-propionic acid ethyl ester (21). Pre-
parative medium pressure chromatography (MPLC)
(silica gel, AcOEt/hexane: 1/3) afforded the titled
compound as oil in 45% yield. Molecular formula:
C29H26N2O4; TLC (AcOEt/hexane: 1/1) Rf: 0.46;
1H NMR (CDCl3) d 1.15 (t, 3H, –CH3); 3.16 (m, 2H,
–CH2–CH<); 4.10 (q, 2H, –CH2–O–); 4.90 (m, 1H, –
CH<); 6.93–8.87 (m, 17H, ar and –NH–CH<); 11.54
(s, 1H –NH–); 13C NMR (CDCl3) d 14.19, 37.87, 53.76,
61.81, 120.36, 121.56, 123.23, 125.00, 125.63, 125.74,
126.50, 127.06, 127.25, 128.45, 128.64, 129.34, 130.58,
131.33, 132.93, 133.98, 134.45, 135.93, 139.86, 167.91,
168.54, 171.39.
2(R,S)-[2-(3,3-Dimethyl-butyrylamino)-benzoylamino]-3-
phenyl-propionic acid ethyl ester (27). Crystallization
from EtOH 95% afforded the titled compound in 25%
yield; Molecular formula: C24H30N2O4; TLC (AcOEt/
2(R,S)-[2-(Naphthalene-1-sulfonylamino)-benzoylamino]-
3-phenyl-propionic acid ethyl ester (22). Crystal-
lization from EtOH 95% afforded the titled compound
in 20% yield; Molecular formula: C28H26N2O5S; TLC
(AcOEt/hexane: 1/2) Rf: 0.30; mp126 ꢀC; 1H NMR
(CDCl3) d 1.29 (t, 3H, –CH3); 3.10 (m, 2H, –CH2–
CH<); 4.21(q, 2H, –CH2–O–); 4.72 (m, 1H, –CH<);
6.37 (d, 1H, –NH–CH<); 6.88–8.76 (m, 16H, ar); 11.20
(s, 1H –NH); 13C NMR (CDCl3) d 14.27, 37.68, 53.48,
61.94, 119.81, 120.30, 123.23, 124.01, 124.78, 126.89,
127.31, 128.07, 128.35, 128.63, 128.94, 129.39, 130.16,
132.92, 134.20, 134.48, 134.62, 135.48, 139.00, 167.54,
171.00.
hexane: 1/2) Rf: 0.46; mp109 ꢀC; H NMR (CDCl3) d
1
1.09 (s, 9H, –C(CH3)3); 1.29 (t, 3H, –CH3); 2.25 (s, 2H,
–CO–CH2–); 3.24 (m, 2H, –CH2–CH<); 4.22 (q, 2H, –
CH2–O–); 5.01 (m, 1H>CH–); 6.70 (d, 1H, –NH–
CH<); 7.02–8.64 (m, 9H, ar); 10.81 (s, 1H –NH–); 13C
NMR (CDCl3) d 14.20, 29.88, 31.29, 37.83, 52.42, 53.58,
61.87, 119.81, 121.43, 122.66, 126.63, 127.33, 128.69,
129.34, 132.87, 135.64, 139.68, 168.41, 170.75, 171.20.
2(R,S)-{2-[(Adamantane-1-carbonyl)-amino]-benzoyl-
amino}-3-phenyl-propionic acid ethyl ester (28). Tritura-
tion with petroleum ether 40–70ꢀ afforded the titled
compound as oil in 80% yield; Molecular formula:
C29H34N2O4; TLC (AcOEt/hexane: 1/1) Rf: 0.71; 1H
NMR (CDCl3) d 1.28 (t, 3H, –CH3); 1.70–2.08 (m, 15H,
ada); 3.24 (m, 2H –CH2–CH<); 4.21 (q, 2H, –CH2–O–);
5.03 (m, 1H, >CH–); 6.94 (d, 1H, –NH–CH<); 7.09–
8.65 (m, 9H, ar); 11.02 (s, 1H, –NH–); 13C NMR
(CDCl3) d 14.45, 28.53, 36.82, 37.99, 39.38, 42.30, 53.96,
2(R,S)-[2-(Naphthalene-2-sulfonylamino)-benzoylamino]-
3-phenyl-propionic acid ethyl ester (23). (MPLC) (silica
gel, CH2Cl2) afforded the titled compound as oil in 52%
yield. Molecular formula: C28H26N2O5S; TLC (AcOEt/
ꢀ
1
hexane: 1/2) Rf: 0.32; mp48 C; H NMR (CDCl3) d
1.26 (t, 3H, –CH3); 3.10 (m, 2H, –CH2–CH<); 4.17(q,

748
Antonio Varnavas et al. / Bioorg. Med. Chem. 11 (2003) 741–751
61.96, 120.41, 121.76, 122.71, 127.05, 127.45, 128.86,
129.58, 132.92, 136.22, 140.23, 168.88, 171.56, 177.44.
128.87, 131.35, 132.42, 136.94, 137.82, 139.10, 158.93,
168.55, 172.50. MS (ES) m/z 428 [M+H]+; Mr: 427
(calcd for C25H21N3O4).
2(R,S)-[2-(2-Adamantan-1-yl-acetylamino)-benzoylamino]-
3-phenyl-propionic acid ethyl ester (29). MPLC (silica
gel, CH2Cl2/AcOEt: 4/1) afforded the titled compound
in 20% yield. Molecular formula: C30H36N2O4; TLC
2(R,S)-{2-[(1H-Indole-5-carbonyl)-amino]-benzoylamino}-
3-phenyl-propionic acid (2). Column chromatography
(silica gel, AcOEt) afforded the titled compound in 70%
yield. TLC (AcOEt/MeOH: 2/1) Rf: 0.44; mp250 ꢀC
(dec); 1H NMR (DMSO-d6) d 3.22 (m, 2H, –CH2–
CH<); 4.73 (m, 1H, –CH<); 6.62–8.69 (m, 14H, ar);
9.12 (d, 1H, –NH–CH<); 11.48 (s, 1H, –NH– ind);
12.03 (s, 1H, –NH–); 12.95 (b, 1H, –COOH); 13C NMR
(DMSO-d6) d 36.91, 54.84, 103.17, 112.31, 120.37,
120.64, 120.75, 122.93, 126.00, 127.10, 127.96, 128.05,
128.86, 129.00, 129.73, 133.08, 138.52, 138.68, 140.40,
166.20, 169.50, 173.41. MS (ES) m/z 428 [M+H]+; Mr:
427 (calcd for C25H21N3O4).
ꢀ
(AcOEt/hexane: 1/1) Rf: 0.76; mp142–143 C; 1H NMR
(CDCl3) d 1.30 (t, 3H, –CH3); 1.65–2.13 (m, 17H, ada
and –CH2–CO–); 3.25 (m, 2H –CH2–CH<); 4.27 (q,
2H, –CH2–O–); 5.02 (m, 1H, >CH–); 6.65 (d, 1H, –
NH–CH<); 7.04–8.66 (m, 9H, ar); 10.77 (s, 1H, –NH–);
13C NMR (CDCl3) d 14.16, 28.60, 33.24, 36.65, 37.75,
42.50, 48.29, 53.55, 61.83, 119.69, 121.29, 122.56,
126.48, 127.23, 128.56, 129.22, 132.79, 135.41, 139.43,
168.23, 169.95, 171.03.
Method C
2(R,S)-{2-[(1H-Indole-3-carbonyl)-amino]-benzoylamino}-
3-phenyl-propionic acid (3). Crystallization from MeOH
afforded the titled compound in 41%ꢀ yield. TLC
(AcOEt/MeOH: 2/1) Rf: 0.46; mp223–224 C; 1H NMR
(DMSO-d6) d 3.15 (m, 2H, –CH2–CH<); 4.70 (m, 1H,
>CH–); 7.06–8.14 (m, 13H, ar); 8.60 (d, 1H, ar); 9.06
(d, 1H, –NH–CH<); 11.58 (s, 1H, –NH– ind); 11.81 (s,
1H, –NH–); 12.90 (b, 1H, –COOH); 13C NMR (DMSO-
d6) d 36.88, 54.82, 111,71, 113.00, 120.08, 120,72,
121.18, 121,74, 122.42, 123.09, 126.04, 127.10, 128.88,
129.10, 129.71, 132.97, 137.16, 138.66, 140.52, 163.29,
169.54, 173.41. MS (ES) m/z 428 [M+H]+; Mr: 427
(calcd for C25H21N3O4).
2(R,S)-{2-[(1H-Indole-5-carbonyl)-amino]-benzoylamino}-
3-phenyl-propionic acid ethyl ester (17). To a solution of
compound 15 (1.00 g, 3.20 mmol) and triethylamine
(0.45 mL, 3.20 mmol) in 15 mL of dry dichloromethane
was added 2-chloro-1-methylpyridinium iodide (1.68 g,
6.40 mmol) and indole-5-carboxylic acid (0.51 g, 3.20
mmol) and the mixture was stirred under reflux for 12 h.
After addition of water (20 mL) the reaction mixture
was extracted with dichloromethane (3Â20 mL). The
organic layer was washed in succession with 1 N HCl,
water, 1 N NaOH and brine. The dried organic phase
was rotary evaporated and the residue was chromato-
graphed on silica gel eluting with CH2Cl2/AcOEt (4:1)
to afford compound 17 in 35% yield. Molecular
formula: C27H25N3O4; TLC (AcOEt/hexane: 1/1) Rf:
2(R,S)-[2-(2-1H-Indol-3-yl-acetylamino)-benzoylamino]-
3-phenyl-propionic acid (4). Trituration with Et2O affor-
ded the titled compound in 55% yield. TLC (AcOEt/
MeOH: 2/1) Rf: 0.48; mp152–153 ꢀC; 1H NMR
(DMSO-d6) d 3.39 (m, 2H, –CH2—CH<); 3.72 (s, 2H,
–CH2—CO–); 4.61 (m, 1H, >CH-); 6.93–7.62 (m, 13H,
ar); 8.37 (d, 1H, ar); 8.92 (d, 1H, –NH–CH<); 10.88 (s,
1H, –NH– ind); 11.00 (s, 1H, –NH–); 13C NMR
(DMSO-d6) d34.68, 35.96, 53.65, 107.26, 111.22, 118.12,
118.38, 119.93, 120.64, 120.90, 122.20, 124.20, 126.20,
127.85, 127.98, 128.83, 131.63, 135.10, 137.70, 138.45,
167.84, 169.62, 172.59. MS (ES) m/z 442 [M+H]+; Mr:
441 (calcd for C26H23N3O4).
0.45; mp74–76 ꢀC; H NMR (DMSO-d6) d 1.12 (t, 3H,
1
–CH3); 3.18 (m, 2H, –CH2–CH<); 4.10 (q, 2H, –CH2–
O–); 4.72 (m, 1H, –CH<); 6.60–8.68 (m, 14H, ar); 9.26
(d, 1H, –NH–CH<); 11.49 (s, 1H, –NH– ind); 11.91 (s,
1H, –NH–); 13C NMR (DMSO–d6) d 14.79, 36.81,
55.12, 61.50, 103.08, 112.29, 120.50, 120.59, 120.65, 120.85,
122.99, 125.95, 127.24, 128.03, 128.92, 129.08, 129.79,
133.15, 138.17, 138.53, 140.28, 166.18, 169.72, 171.92.
General procedure for the synthesis of compounds 1–14
A mixture of 5 mmol of the corresponding ethyl ester
(16–29) in methanol (50 mL) and in the presence of
potassium hydroxide (0.56 g, 10 mmol) was gently
warmed for 4 h. The solvent was removed under
reduced pressure and the residue taken up with water.
After cooling, the solution was adjusted to pH 2–3 with
diluted HCl to obtain the precipitation of the acid.
2(R,S)-[2-(3-1H-Indol-3-yl-propionylamino)-benzoyl-
amino]-3-phenyl-propionic acid (5). Crystallization from
EtOH 95% afforded the titled compound in 33% yield.
TLC (AcOEt/MeOH: 2/1) Rf: 0.46; mp191–193 ꢀC; H
1
NMR (DMSO-d6) d 2.80 (m, 2H, –CH2–CH2–CO–);
3.08 (m, 2H, –CH2–CH<); 3.40 (m, 2H, –CH2–CO–);
4.67 (m, 1H, >CH–); 6.99–7.66 (m, 13H, ar); 8.38 (d,
1H, ar); 9.01 (d, 1H, –NH–); 10.78 (s, 1H, –NH– ind);
10.87(s, 1H, –NH–); 13C NMR (DMSO-d6) d 20.41,
35.96, 37.93, 53.83, 111.17, 113.03, 118.03, 118.09,
120.18, 120.64, 120.78, 122.06, 122.35, 126.22, 126.71,
127.91, 128.00, 128.82, 131.75, 136.00, 137.66, 138.35,
168.10, 170.56, 172.66. MS (ES) m/z 456 [M+H]+; Mr:
455 (calcd for C27H25N3O4).
2(R,S)-{2-[(1H-Indole-2-carbonyl)-amino]-benzoylamino}-
3-phenyl-propionic acid (1). Crystallization from MeOH
afforded the titled compound in 85%ꢀ yield. TLC
(AcOEt/MeOH: 2/1) Rf: 0.61; mp268–269 C; 1H NMR
(DMSO-d6) d 3.27 (m, 2H –CH2–CH<); 4.79 (m,
1H>CH–); 6.97 (s, 1H H ind); 7.06–7.87 (m, 12H, ar);
8.64 (d, 1H, ar); 9.21 (d, 1H, –NH–CH<); 11.93 (s, 1H,
–NH– ind); 12.28 (s, 1H, –NH–); 13C NMR (DMSO-d6)
d 35.94, 53.99, 102.42, 112.32, 118.98, 119.65, 120.01,
121.67, 122.33, 123.88, 126.22, 126.77, 127.98, 128.26,
2(R,S)-{2-[(Naphthalene-1-carbonyl)-amino]-benzoyl-
amino}-3-phenyl-propionic acid (6). Crystallization from

Antonio Varnavas et al. / Bioorg. Med. Chem. 11 (2003) 741–751
749
EtOH 50% afforded the titled compound in 65% yield.
TLC (AcOEt/MeOH: 2/1) Rf: 0.67; mp199–200 ꢀC; H
2(R,S)-[2-(2,2-Dimethyl-propionylamino)-benzoylamino]-
3-phenyl-propionic acid (11). Crystallization from
AcOEt afforded the titled compound in 57% yield. TLC
1
NMR (DMSO-d6) d 3.15 (m, 2H, –CH2–CH<); 4.65
(m, 1H, >CH–); 7.10–8.73 (m, 16H, ar); 9.15 (d, 1H,
–NH–CH<); 11.77 (s, 1H, –NH–); 12.95 (b, 1H,
–COOH); 13C NMR (DMSO-d6) d 35.97, 53.80, 120.31,
120.48, 122.92, 124.95, 125.06, 126.16, 126.36, 127.04,
127.92, 128.14, 128.25, 128.83, 129.41, 130.86, 132.14,
133.19, 133.85, 137.71, 138.74, 166.40, 168.19, 172.44. MS
(ES) m/z 439 [M+H]+; Mr: 438 (calcd for C27H22N2O4).
ꢀ
(AcOEt/MeOH: 2/1) Rf: 0.49; mp195–196 C; 1H NMR
(DMSO-d6) d 1.17 (s, 9H, –C(CH3)3); 3.15 (m, 2H, –
CH2–CH<); 4.65 (m, 1H, >CH–); 7.09–8.47 (m, 9H,
ar); 9.06 (d, 1H, –NH–CH<); 11.20 (s, 1H, –NH–). 12.95
(b, 1H, –COOH); 13C NMR (DMSO-d6) d 26.97, 35.82,
53.83, 119.73, 119.94, 122.04, 126.22, 127.92, 127.99,
128.80, 131.91, 137.82, 139.00, 168.46, 172.47, 176.09. MS
(ES) m/z 369 [M+H]+; Mr: 368 (calcd for C21H24N2O4).
2(R,S)-[2-(Naphthalene-1-sulfonylamino)-benzoylamino]-
3-phenyl-propionic acid (7). Crystallization from EtOH
95% afforded the titled compound in 35% yield. TLC
2(R,S)-[2-(3,3-Dimethyl-butyrylamino)-benzoylamino]-3-
phenyl-propionic acid (12). Trituration with petroleum
ether 40–70ꢀ afforded the titled compound in 62% yield.
TLC (AcOEt/MeOH: 2/1) Rf: 0.55; mp149–150 ꢀC;
1H NMR (CDCl3) d 1.10 (s, 9H, –C(CH3)3); 2.28 (s, 2H,
–CH2–CO–); 3.33 (m, 2H, –CH2–CH<); 5.05 (m, 1H,
>CH–); 6.70 (d, 1H, –NH—CH<); 7.04–8.58 (m, 9H,
ar); 10.83 (s, 1H, –NH–); 13C NMR (CDCl3) d 29.87,
31.37, 37.38, 52.32, 53.38, 120.10, 121.80, 123.08,
126.68, 127.46, 128.80, 129.36, 132.98, 135.43, 139.24,
168.58, 171.49, 174.77. MS (ES) m/z 383 [M+H]+; Mr:
382 (calcd for C22H26N2O4).
ꢀ
(AcOEt/MeOH: 3/1) Rf: 0.30; mp196–197 C; 1H NMR
(DMSO-d6) d 3.18 (m, 2H, –CH2–CH<); 4.63 (m, 1H,
>CH–); 6.97–8.49 (m, 16H, ar); 9.05 (d, 1H, –NH–
CH<); 12.06 (s, 1H, –NH–); 12.99 (b, 1H, –COOH);
13C NMR (DMSO-d6) d 35.78, 54.07, 116.98, 117.96,
122.24, 123.19, 124.24, 126.26, 126.60, 126.94, 128.05,
128.39, 128.47, 128.77, 129.04, 130.23, 132.67, 133.00,
133.54, 134.77, 137.56, 138.26, 168.07, 172.30. MS (ES)
m/z 475 [M+H]+; Mr: 474 (calcd for C26H22N2O5S).
2(R,S)-[2-(Naphthalene-2-sulfonylamino)-benzoylamino]-
3-phenyl-propionic acid (8). Crystallization from EtOH
95% afforded the titled compound in 45% yield. TLC
(AcOEt/MeOH: 3/1) Rf: 0.74; mp153 ꢀC; 1H NMR
(DMSO-d6) d 3.10 (m, 2H, –CH2–CH<); 4.62 (m, 1H,
>CH–); 6.98–8.51 (m, 16H, ar); 9.34 (d, 1H, –NH–
CH<); 11.85 (s, 1H, –NH–).; 13C NMR (DMSO-d6) d
35.85, 54.23, 118.55, 119.60, 121.65, 122.42, 126.25,
127.49, 127.58, 128.04, 128.12, 128.58, 128.82, 129.09,
129.46, 131.30, 132.47, 134.06, 135.87, 137.68, 138.83,
168.02, 172.47. MS (ES) m/z 475 [M+H]+; Mr: 474
(calcd for C26H22N2O5S).
2(R,S)-{2-[(Adamantane-1-carbonyl)-amino]-benzoyl-
amino}-3-phenyl-propionic acid (13). Trituration with
hot MeOH afforded the titled compound in 75% yield.
ꢀ
1
TLC (AcOEt/MeOH: 2/1) Rf: 0.53; mp140–141 C; H
NMR (DMSO-d6) d 1.68–2.01 (m, 15H, ada); 3.15
(m, 2H, –CH2–CH<); 4.71(m, 1H, >CH–); 7.09–8.49
(m, 9H, ar); 9.04 (d, 1H, –NH–CH<); 11.08 (s, 1H, –NH–
); 13C NMR (DMSO-d6) d 27.38, 35.79, 38.37, 41.12,
53.73, 119.88, 120.00, 122.00, 126.19, 127.93, 128.82,
131.84, 137.82, 138.97, 168.33, 172.49, 175.55. MS (ES) m/
z 447 [M+H]+; Mr: 446 (calcd for C27H30N2O4).
3-Phenyl-2(R,S)-(2-phenylacetylamino-benzoylamino)-
propionic acid (9). Crystallization from EtOH 95%
afforded the titled compound in 71%ꢀ yield. TLC
(AcOEt/MeOH: 2/1) Rf: 0.50; mp183–185 C; 1H NMR
(DMSO-d6) d 3.27 (m, 2H, –CH2–CH<); 3.64 (s, 2H, –
CH2–CO–); 4.66 (m, 1H, >CH–); 7.10–7.65 (m, 13H,
ar); 8.34 (d, 1H, ar); 9.00 (d, 1H, –NH–CH<); 10.92 (s,
1H, –NH–); 13C NMR (DMSO-d6) d 35.96, 44.20,
53.83, 120.17, 120.89, 122.47, 126.22, 126.57, 127.91,
128.01, 128.26, 128.84, 129.07, 131.68, 134.90, 137.76,
138.29, 167.95, 168.87, 172.63. MS (ES) m/z 403
[M+H]+; Mr: 402 (calcd for C24H22N2O4).
2(R,S)-[2-(2-Adamantan-1-yl-acetylamino)-benzoylamino]-
3-phenyl-propionic acid (14). Crystallization from
MeOH afforded the titled compound in 40% yield. TLC
(AcOEt/MeOH: 2/1) Rf: 0.45; mp195 ꢀC; 1H NMR
(DMSO-d6) d 1.57–1.91 (m, 15H, ada); 1.99 (s, 1H,
–CH2–CO–); 3.17 (m, 2H –CH2–CH>); 4.67 (m, 1H,
>CH–); 7.08–8.42 (m, 9H, ar); 8.97 (d, 1H, –NH–
CH<); 10.75 (s, 1H, –NH–); 13C NMR (DMSO-d6)
d 28.75, 33.42, 36.89, 37.08, 42.76, 53.33, 54.82, 120.81,
121.20, 123.07, 127.10, 128.88, 129.71, 132.66, 138.68,
139.27, 168.97, 169.44, 173.39. MS (ES) m/z 461
[M+H]+; Mr: 460 (calcd for C28H32N2O4).
3-Phenyl-2(R,S)-[2-(3-phenyl-propionylamino)-benzoyl-
amino]-propionic acid (10). Crystallization from MeOH
afforded the titled compound in 67%ꢀ yield. TLC
(AcOEt/MeOH: 3/2) Rf: 0.61; mp169–171 C; 1H NMR
(DMSO-d6) d 2.62 (t, 2H, –CH2–CH2–CO–); 2.90 (t,
2H, –CH2–CO–); 3.21 (m, 2H, –CH2—CH<); 4.66 (m,
1H, >CH–); 7.09–7.67 (m, 13H, ar); 8.37 (d, 1H, ar);
9.00 (d, 1H, –NH–CH<); 10.86 (s, 1H, –NH–); 13C–
NMR (DMSO-d6) d 30.40, 35.99, 38.55, 53.86, 120.16,
120.59, 122.31, 125.79, 126.21, 127.98, 128.05, 128.11,
128.85, 131.72, 137.75, 138.36, 140.52, 168.03, 169.92,
172.66. MS (ES) m/z 417 [M+H]+; Mr: 416 (calcd for
C25H24N2O4).
Biological evaluations
Male Hartley guinea pigs (300–350 g) and male Sprague
Dawley rats (250–300 g) were used. For binding assays
to isolated rat pancreatic acini, animals were fasted, but
allowed free access to water, for 18–24 h prior to the
experiment.
[
125I]-BH-CCK-8 (CCK8(sulphated), [125I]Bolton and
Hunter labelled-specific activity 2000 Ci/mol) was pur-
chased from Amersham Pharmacia Biothech (Buck-
inghamshire, UK). All other drugs and reagents were
obtained from commercial sources.

750
Antonio Varnavas et al. / Bioorg. Med. Chem. 11 (2003) 741–751
The binding parameters for the substances under inves-
tigation (IC50 values and p=0.05 fiducial limits) were
determined by regression analysis of competition
curves.
obtained at this first stepwere then submitted to a fur-
ther refinement, and their geometries were reoptimized
with a semiempirical calculation using the AM130
hamiltonian as implemented in SYBYL 6.8.31 The SCF
convergence limit for the UHF calculation was set to
full accuracy. All the calculations were carried out on a
Silicon Graphics O2 workstation.
[
125I]BH-CCK-8 receptor binding assay in isolated rat
pancreatic acinar cells. Isolated pancreatic acini were
prepared by enzymatic digestion of pancreas as pre-
viously described by Makovec et al.27 Drug displacing
experiments were carried out by incubating acinar cells,
Acknowledgements
[
125I]BH-CCK-8 (25 pM final concentration) and com-
petitors in 0.5 mL total volume at 37 ꢀC for 30 min, in
shaking bath. At the end of incubation 1 mL of ice-cold
Hepes–Ringer buffer (10 mM Hepes, 118 mM NaCl,
1.13 mM MgCl2, 1.28 CaCl2, 1% BSA, 0.2 mg/mL
Soybean trypsin inhibitor, pH 7.4) was added and the
tubes were centrifuged 5 min at 12,500g. The super-
natant was aspirated and the radioactivity associated to
the pellet measured. The non-specific binding was esti-
mated in the presence of 1 mM CCK-8, accounting 15%
of total binding.
We are grateful to prof. Vinicio Galasso for the ab initio
optimizations of the anthranoyl amide model.
References and Notes
1. Williams, J. A. Biomed. Res. 1982, 3, 107.
2. Jensen, R. T.; Wank, S. A.; Rowley, W. H.; Sato, S.;
Gardner, J. D. Trends Pharmacol. Sci. 1989, 10, 418.
3. Liddle, R. A.; Gertz, B. J.; Kanayama, S.; Beccaria, L.;
Coker, L. D.; Turnbull, T. A.; Morita, E. T. J. Clin. Invest.
1989, 84, 1220.
[
125I]BH-CCK-8 receptor binding assay in guinea pig
4. Brawman Mintzer, O.; Lydiard, R. B.; Bradwejn, J.; Vil-
larreal, G.; Knapp, R.; Emmanuel, N.; Ware, M. R.; He, Q.;
Ballenger, J. C. Am. J. Psychiatry 1997, 154, 700.
5. Lofberg, C.; Agren, H.; Harro, J.; Oreland, L. Eur. Neuro-
psycopharmacol. 1998, 8, 153.
6. Dourish, C. T.; Hill, D. R. Trends Pharmacol. Sci. 1987, 8,
207.
7. Pisegna, J. R.; Deweerth, A.; Huppi, K.; Wank, S. A. Ann.
N. Y. Acad. Sci. 1994, 713, 338.
8. Lee, Y. M.; Beinborn, M.; McBride, E. W.; Lu, M.; Kola-
kowski, L. F.; Kopin, A. S. J. Biol. Chem. 1993, 268, 8164.
9. Makovec, F.; Bani, M.; Chiste, R.; Revel, L.; Rovati, L. C.;
Rovati, L. A. Arzneim.-Forsch./Drug Res. 1986, 36, 98.
10. de Tullio, P.; Delarge, J.; Pirotte, B. Curr. Med. Chem.
1999, 6, 433.
11. Revel, L.; Makovec, F. Drugs Future 1998, 23, 751.
12. Malesci, A.; De Fazio, C.; Festorazzi, S.; Bonato, C.;
Valentini, A.; Tacconi, M.; Bekkering, M.; Giacovelli, G.;
D’Amato, M.; Rovati, L. C. Arzneim.-Forsch./Drug Res. 1992,
42, 1359.
cerebral cortices. Membranes from guinea pig cerebral
cortices, were prepared as previously described.27 Pro-
tein concentration was determined according to Brad-
ford,28 using bovine serum albumin (BSA) as standard.
The binding experiments were performed in assay buffer
containing 10 mM Hepes, 118 mM NaCl, 4.7 mM KCl,
5.0 mM MgCl2, 1.0 mM EGTA, pH 6.5 and supple-
mented with 0.2 mg/mL bacitracin. The incubation of
membranes suspension with labelled ligand and inhibi-
tors was carried out in a microtiter 96-well filter plate
(Multiscreen, Millipore Inc, Bedford, MA, USA) with
integral Whatman GF/B membrane filters. Aliquot of
membranes (0.5 mg of protein/mL) were added to each
well, containing [125I]BH-CCK8 (25 pM), in a final
volume of 250 mL. The non-specific binding of iodinated
peptide was defined in the presence of 1 mM CCK-8,
accounting of 20% of total binding. Nonspecific bind-
ing of [125I]BH-CCK-8 to membrane filters (blank),
measured in wells containing an equal amount of label-
led ligand, but no membranes, was 0.2% of total radi-
oligand added. After 120 min at 25 ꢀC, the 96-well plate
was placed on the vacuum filtration apparatus (Milli-
pore Inc.). The integral membrane filters were rinsed
with 0.25 mL of ice cold assay buffer, dried, punched
into polycarbonate tubes and counted in a COBRA-
5002 g-counter (Packard Biosciences).
13. Niederau, M.; Niederau, C.; Strohmeyer, G.; Grendell,
J. H. Am. J. Physiol. 1989, 256, G150.
14. Hughes, J.; Boden, P.; Costall, B.; Domeney, A.; Kelly,
E.; Horwell, D. C.; Hunter, J. C.; Pinnock, R. D.; Woodruff,
G. N. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 6728.
15. Varnavas, A.; Lassiani, L.; Luxich, E.; Valenta, V. Farm-
aco 2000, 55, 293.
16. Varnavas, A.; Lassiani, L.; Valenta, V. Farmaco 2000, 55,
369.
17. Varnavas, A.; Lassiani, L.; Luxich, E.; Zacchigna, M.
Pharmazie 1996, 51, 697.
18. Varnavas, A.; Valenta, V.; Berti, F.; Lassiani, L. Farmaco
2001, 56, 555.
19. Horwell, D. C.; Birchmore, B.; Boden, P.; Higginbottom,
M.; Ping Ho, Y.; Hughes, J.; Hunter, J. C.; Richardson, R. S.
Eur. J. Med. Chem. 1990, 25, 53.
20. Kolodziej, S. A.; Nikiforovich, G. V.; Skeean, R.; Lignon,
M. F.; Martinez, J.; Marshall, G. R. J. Med. Chem. 1995, 38,
137.
21. Shiosaki, K.; Wel Lin, C.; Kopecka, H.; Bianchi, B.;
Miller, T.; Stashko, M.; Witte, D. J. Med. Chem. 1997, 40,
1169.
22. Mukaiyama, T. Angew. Chem., Int. Ed. Engl. 1979, 18,
707.
Molecular modelling
A first set of optimized conformations for all the com-
pounds was obtained by a simple Monte Carlo search.
All the rotatable bonds, including the amide bonds,
were allowed to rotate in order to generate the starting
set of geometries. Each bond was twisted by 10ꢀ tor-
sional increments randomly and the the initial set was
thus obtained. The geometries were optimized first using
molecular mechanics calculations with the Amber for-
cefield;29 the optimizations were carried out with the
Polak–Ribiere conjugate gradient algorithm at a gra-
dient of 0.001 kcal/A mol. The first ten conformations

Antonio Varnavas et al. / Bioorg. Med. Chem. 11 (2003) 741–751
751
23. (a) Hamuro, Y.; Geib, S.; Hamilton, A. D. Angew. Chem.,
Int. Ed. Engl. 1994, 33, 446. (b) Hamuro, Y.; Geib, S.; Hamil-
ton, A. D. J. Am. Chem. Soc. 1997, 119, 10587. (c) Yu, Q.;
Baroni, T. E.; Liable-Sands, L.; Rheingold, A. L.; Brovik,
A. S. Tetrahedron Lett. 1998, 39, 6831.
´
24. Setnikar, I.; Bani, M.; Cereda, R.; Chiste, R.; Makovec,
F.; Pacini, M. A.; Revel, L.; Rovati, L. C.; Rovati, L. A. Arz-
neimittel Forschung/Drug Res. 1987, 37, 703.
25. Evans, B. E.; Bock, M. G.; Rittle, K. E.; Di Pardo, R. M.;
Whitter, W. L.; Veber, D. F.; Anderson, P. S.; Freidinger,
R. M. Proc. Natl. Acad. Sci. U.S.A. 1986, 83, 4918.
26. Kermack, W. O.; Perkin, W. H.; Robinson, R. J. Chem.
Soc. 1921, 119, 1602.
27. Makovec, F.; Peris, W.; Revel, L.; Giovanetti, R.; Men-
nuni, L.; Rovati, L. C. J. Med. Chem. 1992, 35, 28.
28. Bradford, M. M. Anal. Biochem. 1976, 72, 248.
29. Cornell, D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz,
K. M., Jr.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Cald-
well, J. W.; Kollman, P. A. J. Am. Chem. Soc. 1995, 117, 5179.
30. (a) Dewar, M.; Thiel, W. J. Amer. Chem. Soc. 1977, 99,
4499. (b) Dewar, M. J. S.; McKee, M. L.; Rzepa, H. S. J. Am.
Chem. Soc. 1978, 100, 3607. (c) Dewar, M. J. S.; Zoebisch,
E. G.; Healy, E. F. J. Amer. Chem. Soc. 1985, 107, 3902. (d)
Dewar, M. J. S.; Reynolds, C. H. J. Comp. Chem. 1986, 2, 140.
31. Sybyl 6.8; Tripos Inc. 1699 South Hanley Road, St. Louis,
Missouri, 63144, USA.