Toward the rational development of peptidomimetic analogs of the C-terminal endothelin hexapeptide: Development of a theoretical model

Article abstract of DOI:10.1016/S0014-827X(98)00064-0

In an early report on the structure-activity relationship of endothelin (ET) peptides, it was reported that the C-terminal hexapeptide ET(16-21), His-Leu-Asp-Ile-Ile-Trp, is the minimum ET fragment which maintains biological activity in some, but not all

Full text of DOI:10.1016/S0014-827X(98)00064-0

Il Farmaco 53 (1998) 545±556
Toward the rational development of peptidomimetic analogs of the
C-terminal endothelin hexapeptide: development of a theoretical model
a,
M. Macchia , S. Barontini , F. Ceccarelli , C. Galoppini , L. Giusti , M. Hamdan ,
a
b
c
b
d
*
b
a
a
b
c
A. Lucacchini , A. Martinelli , E. Menchini , M.R. Mazzoni , R.P. Revoltella ,
a
F. Romagnoli , P. Rovero
c
a
Dipartimento di Scienze Farmaceutiche, Universit aÁ di Pisa, via Bonanno 6, 56126 Pisa, Italy
Dipartimento di Psichiatria Neurobiologia Farmacologia e Biotecnologie, Universit aÁ di Pisa, via Bonanno 6, 56126 Pisa, Italy
b
c
Istituto di Mutagenesi e Differenziamento, CNR, Laboratorio di Sintesi Peptidica, via Svezia 2A, 56124 Pisa, Italy
d
GlaxoWellcome Medicines Research Center, via Fleming 4, 37134 Verona, Italy
Received 20 April 1998; accepted 30 July 1998
Abstract
In an early report on the structure±activity relationship of endothelin (ET) peptides, it was reported that the C-terminal hexapeptide
ET(16±21), His-Leu-Asp-Ile-Ile-Trp, is the minimum ET fragment which maintains biological activity in some, but not all the tissues
responding to ETs. Subsequently, other authors described a series of analogs of this peptide, in which the His16 residue was replaced by non-
natural amino acids, characterized by bulky aromatic side chains. Among them, two well-characterized non-selective ETA/ETB antagonists
were PD 142893 and PD 145065; interest in these potent ET antagonists was, however, reduced by their peptidic structure which was likely
to lead to undesirable properties such as poor bioavailability and short duration of action. On the basis of these premises, our previous studies
led to the development of a peptidomimetic ligand of ET receptors (compound 3), based on the replacement of the His16 residue of ET(16±
2
1) with an (E)-N-(benzyloxy)iminoacyl moiety; compound 3 proved to possess a certain af®nity for ET receptors, albeit lower than that
shown by PD 142893 and PD 145065. We report here on ETA/ETB binding af®nity of compounds 4±12, designed as a new series of ET(16±
1) analogs. Compounds 4 and 5 were practically devoid of any af®nity; derivatives 6±12 exhibited appreciable af®nity indices for ETB
2
receptors higher than that shown by 3, even if still lower than that obtained for PD 145065. This paper also describes the development of a
pharmacophoric model able to explain the ET receptor binding properties of our hexapeptide analogs compared with those of PD 142893 and
PD 145065 and IRL2500, recently reported as a potent ETB selective endothelin antagonist. q 1998 Elsevier Science S.A. All rights
reserved.
Keywords: Endothelin; ET(16±21); ETA receptor; ETB receptor; Theoretical model
1
. Introduction
subtypes have been cloned, sequenced and characterized in
several species, including humans. They have been termed
ETA, selective for ET-1, and ETB, unselective, and are
widely distributed in several tissues and organs. The func-
tion and distribution of this receptor family have been found
to vary within different species [1,4,5].
In previous studies on the structure±activity relationship
(SAR) of ET peptides, we described the C-terminal hexa-
peptide ET(16±21), His-Leu-Asp-Ile-Ile-Trp, as the mini-
mum ET fragment which maintains biological activity in
some, but not all the tissues responding to ETs [6,7].
These results were subsequently developed by Cody and
co-workers at Parke-Davis through an SAR study of this
hexapeptide; in particular, it was reported that an analog
in which the His16 residue of ET(16±21) is replaced by
Ac-D-His (1) exhibited a marked increase in binding af®nity
Endothelins (ETs) are a family of 21-residue peptides
recently discovered in the mammalian genome. Three
isopeptides have been described so far, called ET-1, -2
and -3. They are characterized by two conserved disulphide
bridges forming an N-terminal 15-residue bicyclic moiety,
linked to a conserved, very hydrophobic C-terminal hexa-
peptidic tail [1]. ETs are endowed with a variety of biolo-
gical actions [2], most notably at the cardio-vascular level:
ET-1 is the most potent vasoconstrictor known [3]. These
biological effects are exerted through interaction with speci-
®c membrane-bound receptors, belonging to the G-protein
coupled, rhodopsin-like superfamily. At least two receptor
*
Corresponding author.
0014-827X/98/$ - see front matter q 1998 Elsevier Science S.A. All rights reserved.
PII S0014-827X(98)00064-0

5
46
M. Macchia et al. / Il Farmaco 53 (1998) 545±556
[
8]; moreover the substitution of the Ac-D-His residue in the
above compound 1 with bulky aromatic non-natural D-
amino acids, led to compounds possessing interesting antag-
onistic properties towards ET receptors. Among them, two
well-characterized non-selective ETA/ETB antagonists
were PD 142893 and PD 145065, in which the His16 residue
was replaced by an Ac-D-Dip (Dip ˆ 3,3-diphenylalanine)
and an Ac-D-Bhg (Bhg ˆ 10,11-dihydro-5H-dibenzo[a,d]-
cycloheptenglycine) residue, respectively [8,9].
Interest in these potent ET antagonists was, however,
reduced by their peptidic structure which was likely to
lead to undesirable properties such as poor bioavailability
and short duration of action [10]. On the basis of these
premises, we thought it conceivable to further develop
these studies through the design of peptidomimetic ligands
of ET receptors derived from the ET(16±21) hexapeptide
structure.
Apparently, the same strategy was followed also at Parke-
Davis, as these authors very recently reported on a pseudo-
peptide analog of PD 145065, i.e. PD 156252, which is
greatly stabilized with respect to proteolytic degradation
by the introduction of an N-methyl amino acid in position
Fig. 1. Superimposition of residues 16±17 of 1 (the thickest line) and of PD
45065 (the medium line) on the ET(13-17) peptidic portion (the thinnest
line).
1
group, was found to antagonize ET-3 but not ET-1 in two
different biological models [12,13]. Subsequently, we
studied compound 3 in which the His16 residue of
ET(16±21) was replaced with an (E)-N-(benzyloxy)imino-
acyl moiety. This compound was designed on the basis of
considerations about the high potency of PD 142893 and PD
1
45065: previous alanine scan studies of ET-1 indicated that
20 [11].
the aromatic residues Tyr13 and Phe14 are essential for the
interaction of ET-1 with its receptors [14,15]; these
premises prompted us to hypothesize that the interesting
activity shown by PD 142893 and PD 145065 might be
related to the capacity of their N-terminal bulky aromatic
side chains to occupy the receptor sites which normally
interact with the aromatic moieties of the Tyr13 and/or
Phe14 residues of ET-1. In order to test whether this hypoth-
esis could be reasonable, simple molecular modeling tech-
niques were employed: the backbone of the N-terminal
bulky aromatic residues of 1 and PD 145065 were super-
imposed on the backbone of the His16 residue of the three-
dimensional structure of ET-1 [16,17] and the distances of
these aromatic portions from the aromatic rings of the Tyr13
and/or Phe14 residues were evaluated. The result of this
superimposition (see Fig. 1) con®rmed our hypothesis, as
the antagonistic properties of 1 and PD 145065 were in fact
related to the ability of their aromatic side chains to interact
Our ®rst approach to these studies was the investigation
of new analogs of ET(16±21) based on the replacement of
the His16 residue with a non-amino acid moiety: thus, an
analogue (2) characterized by the C-terminal pentapeptide
ET(17±21), with the N-terminal amino function acylated by
the bulky, aromatic 9-¯uorenyl-methoxy-carbonyl (Fmoc)
Fig. 2. Superimposition of residues 16-17 of 3 (the thickest line) and of PD
145065 (the medium line) on the ET(13-17) peptidic portion (the thinnest
line).

M. Macchia et al. / Il Farmaco 53 (1998) 545±556
547
with the spatial region occupied by the Tyr13 and Phe14
residues of ET-1.
with ET receptors. At this point of our research, we thought
it reasonable to try to develop new ET inhibitors possessing
this kind of moiety in their structure. We therefore
proceeded to the elucidation of the role of each amino
acid residue of the ET(17±21) peptidic portion of 3 in the
ligand±receptor interaction, through an alanine scan of this
pentapeptidic portion together with modi®cation of the C-
terminal Trp residue, i.e., amidation or deletion [18]. The
results of these studies indicated that in compound 3, only
the C-terminal dipeptide Ile20-Trp21 including the free
carboxylic function was important for the af®nity and
could be considered as a second pharmacophoric portion
of 3, while the central portion of the molecule seemed to
have the role of maintaining the relative topography of the
two termini. These results were in agreement with previous
studies on ET in which the importance of the C-terminal
dipeptide Ile20-Trp21 for interaction with ETs receptors
had already been demonstrated [14,15].
On this basis, we thought that the replacement of the
His16 residue of ET(16±21) with an (E)-N-(benzyloxy)imi-
noacyl moiety as in compounds 3 might lead to a further
increase in af®nity for ET receptors, as, according to our
molecular model, its aromatic moiety proved to be capable
of getting closer to the regions occupied by the aromatic
ring of Tyr13 or Phe14 (see Fig. 2).
As a continuation of these studies, we now report on the
synthesis and ETA/ETB binding af®nity of compounds 6±
1
2 (see Table 1) designed as a new series of ET(16±21)
analogs; compound 6 presents a phenyl ring in the para
position of the N-terminal (E)-N-(benzyloxy)iminoacyl
moiety of 3 and therefore might have made it possible to
verify whether an increase in the lipophilicity characteristic
of this hypothetical pharmacophore, as well as leaving the
C-terminal dipeptide Ile20-Trp21 (i.e. the second pharma-
cophore) unchanged, could lead to an improvement in the
af®nity for ET receptors. In compounds 7±10, the Leu17 or
Asp18 residues of 3 and 6 were systematically replaced by a
Phe residue, while in compound 11, two Phe residues
replace the Leu17 and Asp18 residues of 6. These latter
modi®cations were suggested by the fact that previous
studies from Doherty and co-workers indicated that the
ET receptor af®nity in analogs of ET(16±21), in which the
His16 residue is replaced by a D-Phe residue, increases
when Leu17 or Asp18 residues are replaced with a Phe
one [9]. Lastly, compound 12 differs from derivative 9 in
the replacement of the Asp18-Ile19 dipeptidic portion with
an aliphatic spacer (Ava) possessing the same length but
without the substituents present on the side chain of the
dipeptide. This type of modi®cation diminishes the peptidic
nature of these types of compounds and, at the same time,
places greater emphasis on the two moieties which might be
thought to act as the pharmacophoric portions of these types
of structures, i.e. the lipophilic N-terminal portion and the
Ile20-Trp21 dipeptidic C-terminal one.
Binding af®nity studies on ET receptors indicated that
compound 3 possessed a certain af®nity for ET receptors
(
see Table 2), albeit lower than that shown by PD 145065.
Therefore, these results did not support the working hypoth-
esis regarding the role played by the aromatic bulky groups
of the N-terminal amino acid residue: in fact, even if in the
proposed molecular model the aromatic moiety of 3 seemed
to be able to interact with the spatial region occupied by the
aromatic ring of Tyr13 or Phe14 better than the N-terminal
bulky aromatic residue of PD 145065, the af®nity of 3 for
ET receptors resulted to be lower than that shown by PD
145065.
However, the fact that compound 3 still possessed a
certain af®nity for ET receptors might indicate that its lipo-
philic N-terminal non-amino acid (E)-N-(benzyloxy)imi-
noacyl moiety could play a certain role in the interaction
with ET receptors. In order to verify this hypothesis, we
synthesized compounds 4 and 5 in which the phenyl ring
and the oxime group of the (E)-N-(benzyloxy)iminoacyl
moiety of 3 were replaced respectively by a methyl group
This paper also reports on the development of a pharma-
cophoric model able to explain the ET receptor binding
properties of our hexapeptide analogs compared with
those of PD 142893 and PD 145065. For this purpose, the
new compound IRL2500, reported to be a potent ETB-
selective endothelin antagonist [19], proved to be particu-
larly useful: its structure, simpler and less ¯exible than
those of the ET(16±21) analogs, allowed us to simplify
the de®nition of the pharmacophoric model. For this
(
4) and a saturated oxyamine moiety (5). Compounds 4 and
proved to be practically devoid of any af®nity for ET
receptors, thus con®rming the hypothesis that the (E)-N-
benzyloxy)iminoacyl moiety of 3 could tentatively be
considered as a pharmacophoric portion for the interaction
5
(

5
48
M. Macchia et al. / Il Farmaco 53 (1998) 545±556
Table 1
Sequences and analytical characterization of compounds 3±12
a
b
c
d
1
No.
Compound
Purity (%)
R
t
M
ES/MS [M 1 H]
e
e
e
e
e
e
f
3
4
5
6
7
8
9
R-Leu-Asp-Ile-Ile-Trp
0
99
98
98
99
97
98
98
98
96
99
19.9
17.0
18.2
22.4
20.0
22.4
17.2
19.5
19.3
16.8
820.1
744.3
821.8
895.7
854.3
851.6
930.3
928.0
961.9
801.3
821
745
823
897
855
853
931
929
963
802
R -Leu-Asp-Ile-Ile-Trp
0
0
0
R -Leu-Asp-Ile-Ile-Trp
00
R -Leu-Asp-Ile-Ile-Trp
R-Phe-Asp-Ile-Ile-Trp
R-Leu-Phe-Ile-Ile-Trp
0
00
R -Phe-Asp-Ile-Ile-Trp
000
0
f
1
1
1
0
1
2
R -Leu-Phe-Ile-Ile-Trp
R -Phe-Phe-Ile-Ile-Trp
00
00
f
0
f
R -Phe-Ava-Ile-Trp
a
0
00
000
R ˆ PhZCH
2
ZOZNyCHZCO. R ˆ CH
3
ZOZNyCHZCO. R ˆ PhZCH
2
ZOZNHZCH
2
ZCO. R ˆ PhZPhZCH ZOZNyCHZCO. Ava ˆ
2
NHZCH ZCH ZCH
2
b
2
2
ZCH ZCO.
2
Chromatographic purity, calculated from peak areas in analytical HPLC.
Retention time in analytical HPLC.
c
d
e
f
Calculated molecular weight.
Chromatographic conditions: method A (see Section 6).
Chromatographic conditions: method B (see Section 6).
reason the conformation of IRL2500 was studied by means
of theoretical calculations and compared with those of PD
142893, PD 145065 and the most signi®cant of our new
compounds.
residue of the relevant peptide moiety, still on the resin,
using ®ve-fold excess and HBTU/HOBt/NMM activation.
Peptides were cleaved from the resin and deprotected from
the side-chain protecting groups with TFA/phenol/anisole
(
92:4:4, v/v/v; 1 h at room temperature) [20]. Crude
peptides were puri®ed to homogeneity by preparative
reverse-phase HPLC. The ®nal products were characterized
by analytical HPLC and ES-MS (see Table 1) [24].
3
. Biological results
The af®nity of derivatives 6±12 towards ETA/ETB recep-
tors was checked by binding tests on rat uterus membranes
for ETA receptors and on rat cerebellar membranes for ETB
1
25
receptors using [ I]ET-1 as the speci®c ligand for ETA/
ETB receptors. The results of these tests are reported in
Table 2, together with those obtained for compound 3 and
PD 145065, taken as reference compounds. The binding
af®nity for ETA/ETB receptors was expressed as percentage
inhibition at a concentration of 1 mM, and, for compounds
possessing an inhibition percentage . 50% at this concen-
tration, as Ki.
2
. Chemistry
Compounds 3, 5 were synthesized as previously reported
20]. Compounds 4, 6±12 were synthesized by the contin-
[
uous-¯ow solid phase method, using standard Fmoc chem-
istry [21] on a Novasyn PA500 resin. Boc and t-Bu
protecting groups were used for the side chains of Trp and
Asp, respectively. Couplings were performed using HBTU/
HOBt/NMM activation and a ®ve-fold excess of amino
acids. The N-terminal amino group was acylated with the
following acids: (E)-N-(Benzyloxy)iminoacetic acid [22]
All the newly synthesized compounds exhibited appreci-
able af®nity indices for ETB receptors. Compound 6, which
differs from 3 in the presence of a phenyl ring in the para
position of the N-terminal (E)-N-(benzyloxy)iminoacyl
moiety, proved to possess an af®nity for ETB receptors
higher than that shown by 3, with a K value of 0.96 mM.
i
A slight increase in the af®nity was obtained for compounds
7±10, in which the Leu17 or Asp18 residues of 3 and 6 were
(
compounds 7, 8), (E)-N-(methoxy)iminoacetic acid [23]
compound 4), and (E)-N-(p-phenyl-benzyloxy)iminoacetic
(
systematically replaced by a Phe residue; their K values for
i
acid (compounds 6, 9±12), prepared by treatment of O-[(4-
phenylphenyl)methyl]hydroxylamine hydrochloride with
glyoxylic acid. Each acid was coupled to the N-terminal
ETB receptors ranged from 0.35 to 0.60 mM. Compound 11,
in which both the Leu17 and Asp18 residues of 6 were
replaced by two Phe residues proved to be the most active

M. Macchia et al. / Il Farmaco 53 (1998) 545±556
549
Table 2
Displacement of [ I]ET-1 bound to ET-A and ET-B receptors by compounds 3, 6±12
125
a
b
d
No.
Compound
ET-A
Inhibition (1 mM)
ET-B
c
c
e
%
% Inhibition (1 mM)
K
i
(mM)
3
6
7
8
9
R-Leu-Asp-Ile-Ile-Trp
25.2 ^ 1.8
20.0 ^ 2.1
8.8 ^ 0.3
14.9 ^ 2.6
18.6 ^ 4.3
7.6 ^ 2.2
17.2 ^ 1.5
18.1 ^ 4.9
21.8 ^ 3.3
57.7 ^ 0.4
60.7 ^ 0.6
62.0 ^ 2.1
66.3 ^ 2.5
53.9 ^ 1.3
69.0 ^ 2.5
59.7 ^ 5.2
±
0
00
R -Leu-Asp-Ile-Ile-Trp
R-Phe-Asp-Ile-Ile-Trp
R-Leu-Phe-Ile-Ile-Trp
0.96 ^ 0.30
0.35 ^ 0.07
0.58 ^ 0.10
0.60 ^ 0.08
0.46 ^ 0.09
0.27 ^ 0.04
0.48 ^ 0.03
0
00
R -Phe-Asp-Ile-Ile-Trp
R -Leu-Phe-Ile-Ile-Trp
000
1
1
1
0
1
2
0
0
00
00
R -Phe-Phe-Ile-Ile-Trp
R -Phe-Ava-Ile-Trp
PD145065
f
g
NT
0.011
a
b
d
c
e
000
R ˆ PhZCH
2
ZOZNyCHZCO. R ˆ PhZPhZCH
2
ZOZNyCHZCO. Ava ˆ NHZCH
2 2 2 2
ZCH ZCH ZCH ZCO.
125
Displacement of speci®c [ I]ET-1 binding to rat uterus membranes.
125
Displacement of speci®c [ I]ET-1 binding to rat cerebellar membranes.
Percentage of inhibition at the concentration of 1 mM. Values represent the means ^ SE of three experiments.
Binding data were computer-analyzed by non-linear least squares analysis (GraphPad Prism Softwares, San Diego, CA) and IC50 determined. The IC50
values were converted to K
f
Not tested.
i
(inhibition constant) values by the Cheng and Prusoff equation [30]. Values represent the means ^ SE of three experiments.
g
i
This K value represents the average of two experiments.
for ETB receptors with a K value of 0.27 mM. Lastly
i
Fig. 3 shows the results of conformational analysis for
compounds PD 142893 and PD 145065, 6 and 11 (the
results regarding compounds 3, 7±10 and 12 are completely
analogous). For each compound, the ®gure shows the mini-
mal energy conformation and the superimposition of the ®ve
lowest energy conformations, together with the graphs plot-
ting the distance between the alpha carbon of the C-terminal
Trp21 residue and the center(s) of the phenyl ring(s) of the
non-natural residue 16, against the conformational energy.
The results of conformational analysis indicated that all
compounds considered had a really considerable conforma-
tional freedom. However, it was also found that the back-
bone of all the molecules (but especially of PD 142893 and
PD 145065) has the tendency to assume a common folded
conformation, which makes the aromatic systems of the two
terminal residues quite close. In any case these aromatic
rings of the terminal residues can move freely. This
observed conformational trend seems to be consistent with
compound 12, which differs from derivative 9 in the repla-
cement of the Asp18-Ile19 dipeptidic portion with an
aliphatic spacer (Ava), proved to possess an af®nity for
ETB receptors similar to that of derivatives 6±11.
As regards the af®nity for ETA receptors, compounds 6±
2 displayed inhibition percentages lower than those on
1
ETB receptors and very similar to that obtained for 3.
Furthermore, compounds 6±12 exhibited an af®nity for
ETA/ETB receptors lower than that obtained for PD
145065, even if the selectivity for ETB receptors was
found to be slightly more marked.
4. Theoretical calculations
With the aim of rationalising the results obtained for
compounds 6±12 and of understanding the difference of
af®nity between them and the more potent PD compounds,
a comparative conformational analysis was carried out by
means of molecular mechanics and dynamics on all these
compounds [25]. The ET(16±21) hexapeptide portion of
endothelin was previously found to possess quite a high
conformational freedom [26]. NMR spectroscopy and X-
ray diffraction on endothelin did not reveal any evident
structuring of this hexapeptide portion [26].
As this conformational freedom should undoubtedly be
present also in the ET(16±21) analogs and such compounds
display a considerable molecular complexity, it was not
possible to perform an exhaustive conformational analysis
on them. We therefore carried out a high-temperature mole-
cular dynamics simulation for each molecule considered;
some conformations were then regularly sampled and mini-
mized by means of molecular mechanics calculations.
1
a very recent report of a H NMR and molecular dynamics
simulation study on PD 145065 [11].
A comparison of the conformational data for compounds
PD 142893 and PD 145065 with those of 3 and 6±12
suggests that the latter possess a greater conformational
freedom, as demonstrated by the fact that in 3 and 6±12
there is a larger number of conformations with similar ener-
gies than in PD 142893 and PD 145065 (see Fig. 3).
However, this difference does not seem to be suf®ciently
signi®cant to rationalize the great difference in potency
between PD 142893 and PD 145065 with respect to 3 and
6
±12.
Moreover the conformational freedom described above,
together with the fact that PD 142893 and PD 145065 in
their preferred conformations are not superimposable at all,

5
50
M. Macchia et al. / Il Farmaco 53 (1998) 545±556
Fig. 3. Conformational analysis of (a) PD142893, (b) PD 145065, (c) compound 6 and (d) compound 11. The conformational energy (kcal/mol) is plotted
against the distance between the alpha carbon of the Trp21 [Ca(Trp)] and the centers of the phenyl rings of the non-natural residue 16 (Ar1 and Ar2) and, in the
case of compound 11, also the centers of the phenyl ring of residues 17 and 18. The lowest energy conformation is shown together with the superimposition of
the ®ve lowest energy conformations; in this case, for residues 17±20, only the backbone is shown.
does not allow us to use these potent compounds as starting
points for the building of a reliable pharmacophore model
for interaction with ETA/ETB receptors.
Recently, a new compound, IRL2500, was found to be
active as a potent selective ETB antagonist [19]. IRL2500
has a tripeptide-like structure, which therefore should have a
better de®ned conformational trend than hexapeptide
analogs like PD's. Moreover the SAR of IRL2500 can
also be well-de®ned: four molecular portions can be identi-
PDs; the chemical characteristics of their groups suggested
superimposing (a) the terminal Trp residues of all
compounds together and (b) the diphenyl moiety of
IRL2500 on the bulky aromatic groups of PD 142893 and
®
ed as fundamental for the eliciting of the high potency: the
indole system, the carboxylic function of the Trp residue,
the xylene ring and the diphenyl moiety.
Fig. 4 reports the results of the conformational analysis
performed on this compound; it con®rms that IRL2500 has
quite a well-de®ned conformation: substantially there are
two preferred conformations with a similar energy, which
differ only in the position of the indole ring, which is close
to the xylene ring or the diphenyl system, respectively.
In order to ®nd a common pharmacophoric model able to
explain the high potency of PD 142893, PD 145065 and
IRL2500 we tried superimposing the pharmacophoric
groups of IRL 2500 on the corresponding groups of the
Fig. 4. Conformational analysis of IRL2500. The conformational energy
(
kcal/mol) is plotted against the distance between the alpha carbon of the
Trp21 [Ca(Trp)] and the centers of the aryl rings Ar1 and Xyl.

M. Macchia et al. / Il Farmaco 53 (1998) 545±556
551
PD 145065. The xylene ring of IRL2500 was therefore
considered to be a supplementary pharmacophore peculiar
to this drug.
Unfortunately this kind of superimposition was not possi-
ble for any of the three compounds in their minimal energy
conformations; therefore we decided to take into account
also other low energy conformers. We found a good ®t
between PD 142893 in its minimal energy conformation
and IRL2500 in a conformation which differed from the
minimal energy one in the position of the indole ring and
was unfavored only by about 0.8 kcal/mol. In the case of PD
1
45065 we found a conformer which ®tted quite well with
the considered conformers of PD 142893 and IRL2500 and
was un¯avored by about 8 kcal/mol with respect to its mini-
mal energy conformation. The superimposition of the three
compounds in these conformations (see Fig. 5) was used as
the basis for the proposal of a pharmacophoric model for the
interaction of these compounds with the ETA/ETB recep-
tors. In this model, the negative carboxylate group of the
terminal Trp residue (charged at physiological pH), the
indole ring and the bulky lipophilic aromatic moiety (the
diphenyl system of IRL2500, the Ac-D-Dip moiety of PD
Fig. 5. Superimposition of IRL2500, PD 142893 and PD 145065. The
carbon atoms of IRL2500, PD 142893 and PD 145065 are colored yellow,
magenta and green, respectively; the oxygen atoms are colored red and the
nitrogen ones blue; the hydrogen atoms are hidden. The molecular regions
circled are those which should interact with: (P) the positive-charged site
for interaction with the C-terminal carboxylate moiety; (L1) the lipophilic
pocket able to host the indole side chain of the C-terminal Trp; (L2) the
lipophilic pocket able to host the bulky aromatic moiety; (L3) the lipophilic
pocket able to host the xylene moiety of IRL2500.
1
42893 and the Ac-D-Bhg one of PD 145065) represent
three common pharmacophoric features. The lipophilic
xylene ring of IRL2500 is not superimposable on a corre-
sponding group of PD 142893 and PD 145065 (in this
spatial region these drugs have the side chain of a Leu
residue, lipophilic but much less bulky) and therefore it
could be considered as responsible for the ETB selectivity
of IRL2500. Another reason for the ETB selectivity of
IRL2500 could be assigned to the lack of a lipophilic
group placed in the region occupied by the Ile20 side
chain of PD 142893 and PD 145065. This Ile20 residue is
generally considered to be important for the activity of
endothelin and of the ET(16±21) analogs [14,15].
According to the characteristics of this model, it should
be hypothesized that both the ETA and ETB receptors
possess a positive-charged site (P) for interaction with the
C-terminal carboxylate moiety and two lipophilic pockets
able to host the indole side chain of tryptophan (L1) and the
bulky aromatic moiety (L2), respectively; moreover the
ETB receptor might also possess an additional lipophilic
pocket (L3) able to host the xylene moiety of IRL2500
(
see Fig. 5).
Finally, this proposed model was used in order to try to
explain the lower activity of the series of compounds here
synthesized with respect to IRL2500, PD 142893 and PD
1
45065 and the differences of activity within the series
itself.
For this purpose, we inserted compounds 3 and 6±12 into
the model. Fig. 6 illustrates how it is possible to superim-
pose compounds 6 and 11 on IRL 2500: the diphenyl moiety
of 6 and 11 does not seem to be able to occupy effectively
the lipophilic pocket L2, occupied by the same moiety of
IRL2500 and by the bulky aromatic portions of PD 142893
and PD 145065; this observation might justify the lower
Fig. 6. Superimposition of IRL2500, compounds 6 and 11. The carbon
atoms of IRL2500, 6 and 11 are colored yellow, magenta and green, respec-
tively; the oxygen atoms are colored red and the nitrogen ones blue; the
hydrogen atoms are hidden. As in Fig. 5.

5
52
M. Macchia et al. / Il Farmaco 53 (1998) 545±556
Fig. 7. Superimposition of IRL2500 on the ten lowest energy conformations of compound 3 (on the left) and those of compound 11 (on the right); all the
hydrogen atoms are hidden. IRL2500 is completely colored yellow. As regards compounds 3 and 11, only the CZOZNyCZPh portion of residue 16 and the
whole of residue Trp21 are shown; in the case of 11, also the phenyl rings present on residues 17 and 18 are shown. Carbon, oxygen and nitrogen atoms of both
3
and 11 are colored green, red and blue, respectively.
ETA/ETB receptor af®nities found for derivatives 6 and 11
with respect to PD 142893 and PD 145065. However, it
should be pointed out that, as a consequence of their high
conformational freedom (see Fig. 3), the possibility cannot
be excluded of a certain favorable interaction between the
diphenyl portion and the lipophilic pockets L2 (in agree-
ment with the appreciable, albeit low, receptor af®nity of
these compounds) and L3 (in agreement with their ETB
selectivity). The presence in 11 of Phe residues in position
might improve the interaction with both the lipophilic pock-
ets L2 and L3 (in agreement with the slightly higher af®nity
of 11 for ETB receptors with respect to 6) (see Fig. 6).
This fact might also explain why compound 3 possesses
the lowest ETB binding af®nity and the lowest ETB selec-
tivity. In Fig. 7, the superimposition of ten conformers of 3
with IRL2500 together with the superimposition of ten
conformers of 11 again with IRL2500 is shown: in the
case of compound 3, there are only conformers for which
the phenyl ring of residue 16 can only partially occupy the
pocket L2, while, in the case of compound 11, there is the
possibility that the pocket L2 or L3 can be occupied by the
phenyl rings of its residues, even if no conformation of 11
exists that is able to simultaneously occupy both the pock-
ets, in agreement with its lower binding af®nity with respect
to IRL2500.
17 and 18, is not able to change this situation signi®cantly;
however, it may be observed that these lipophilic residues
As regards compounds 7±10, the conformational trend is
analogous to that of 6 and 11, characterized by a conforma-
tion not easy to de®ne (due to the rotational freedom) and by
the ability of the Phe residues in position 17 or 18 to
improve the interaction with L2 and L3 (in agreement
with the ETA/ETB binding af®nity pro®le of 7±10 similar
to that found for 6 and 11).
Finally, Fig. 8 shows the superimposition of a suitable
low-energy conformer of 12 on 11 and IRL2500: it may be
seen that the aliphatic spacer (Ava) of 12 may be able to
replace the residues 18 and 19 without signi®cantly modify-
ing the relative spatial position of residues 16, 17, 20 and 21
with respect to 11, and consequently to the other hexapep-
tide analogs 6±10; this is in agreement with the similar
ETA/ETB binding af®nity pro®le of 12 with respect to 6±11.
Fig. 8. Superimposition of IRL2500, compounds 11 and 12. The carbon
atoms of IRL2500, 11 and 12 are colored yellow, green and magenta,
respectively; the oxygen atoms are colored red and the nitrogen ones
blue; the hydrogen atoms are hidden. As in Fig. 5.

M. Macchia et al. / Il Farmaco 53 (1998) 545±556
553
5. Conclusions
NMR spectra of compounds were obtained with a Bruker
AC-200 instrument in ca 2% solution of CDCl using Me Si
3
4
Compounds 6±11, which differ from 3 in the insertion of
as the internal standard. Analytical TLCs were carried out
on 0.25 mm layer silica gel plates containing a ¯uorescent
indicator (Macherey-Nagel Alugramw SilG/UV254 Art.
Nr. 81813); spots were detected under UV light (254 nm).
Evaporations were made in vacuo (rotating evaporator);
phenyl moieties on the N-terminal non-amino acid residue
and/or on the side chain of amino acid residues 17 and 18,
proved to possess an increased af®nity, prevalently toward
the ETB receptor: these results indicate that an increase in
the lipophilic characteristics of the proposed N-terminal
pharmacophore might play a certain role in the interaction
with ET receptors.
MgSO was always used as the drying agent. Elemental
4
analyses were performed in our analytical laboratory and
agreed with the theoretical values to within ^ 0.4%.
In particular, the presence of a phenyl ring in the para
position of the (E)-N-(benzyloxy)iminoacyl moiety of 3 as
in 6, seems to be important for ETB receptor interaction, as
demonstrated by the higher af®nity of 6 for such type of
receptor subtype with respect to 3. Moreover, the compound
which proved to possess the best af®nity for ETB receptors
was 11, in which the highest number of aromatic residues
are present in the N-terminal portion.
Derivative 12, in which the Asp18-Ile19 dipeptidic
portion is replaced by an aliphatic spacer (Ava), possessing
the same length but without the substituents present on the
side chain of the dipeptide, maintains the ETA/ETB af®nity
pro®le found for compounds 6±11; this fact underlines the
two moieties which might be thought to act as the pharma-
cophoric portions of these types of structures, i.e. the lipo-
philic N-terminal portion and the Ile20-Trp21 dipeptidic C-
terminal one. Furthermore, compound 12 has a signi®cantly
simpli®ed structure with a far less peptidic nature and could
be regarded as the ®rst step toward the rational design of a
true peptidomimetic of ET(16±21).
The ET receptor binding properties of derivatives 6±12
were rationalized by means of a pharmacophoric model,
constructed on the basis of a comparative conformational
analysis of PD 142893, PD 145065 and IRL2500: according
to the characteristics of this model, the ETA and ETB recep-
tors seem to possess a positive-charged site (P) and two
lipophilic pockets (L1, L2); the ETB receptor might also
possess an additional lipophilic pocket (L3). The proposed
model was able to explain the high potency of PD 142893,
PD 145065 and the ETB selectivity of IRL2500 and at the
same time the lower ETA/ETB af®nity as well as the ETB
selectivity of derivatives 6±12.
6.1.1. Synthesis of O-[(4-phenylphenyl)methyl]-
hydroxylamine hydrochloride [27]
A solution of N-hydroxyphthalimide (0.885 g, 5.42
mmol), triphenylphosphine (2.13 g, 8.12 mmol) and N,N-
diethylazodicarboxylate (1.41 ml, 8.95 mmol) in anhydrous
THF (27.0 ml) was treated with 4-biphenylmethanol (1.0 g,
5.43 mmol) and the resulting mixture was stirred for 36 h
at room temperature. Evaporation of the organic sol-
vent gave a crude residue, which, after crystallisation
with CHCl /hexane, yielded pure N-(4-phenylbenzyl-
3
1
oxy)phthalimide (1.6 g, 89%): m.p. 158±1608C; H NMR
d 5.17 (s, 2H, CH O), 7.5 (m, 13H, 2xC H 1 C H ). Anal.
2
6
4
6
5
for C H NO (C, H, N).
2
1
15
3
N-(4-Phenylbenzyloxy)phthalimide (1.3 g, 3.94 mmol)
was suspended in absolute EtOH (40 ml) and, after the
addition of hydrazine monohydrate (0.276 ml, 5.7 mmol),
was re¯uxed for 30 h. The solvent was evaporated and the
residue was then dissolved in Et O (40 ml) and washed with
2
3% aqueous K CO solution (3 £ 30 ml). The organic phase
2
3
was then dried (MgSO ) and evaporated to give a crude oil.
4
This was dissolved in Et O and acidi®ed to pH 3 at 08C by
2
the addition of a saturated Et O/HCl solution to give O-[(4-
2
phenylphenyl)methyl]hydroxylamine hydrochloride (0.596
1
g, 64%) as a white solid: m.p. 170±1728C; H NMR d 5.09
(s, 2H, CH O), 7.55 (m, 9H, C H 1 C H ). Anal. for
2
6
4
6
5
C H NOCl (C, H, N).
14
1
3
6.1.2. Synthesis of (E)-N-(4-phenylbenzyloxy)iminoacetic
acid
A suspension of O-[(4-phenylphenyl)methyl]hydroxyl-
amine hydrochloride (0.20 g, 0.85 mmol) in acetonitrile
(20 ml), was treated with glyoxylic acid monohydrate
(0.094 g, 1.02 mmol) and the resulting mixture was stirred
for 24 h at 408C. The solvent was evaporated and the crude
Most importantly, the information obtained from these
theoretical studies might lead to further development of
ET(16±21) peptidomimetic analogs with a greater af®nity
and selectivity for ET receptors.
residue was puri®ed by crystallization with CHCl /hexane
3
to yield pure (E)-N-(4-phenylbenzyloxy)iminoacetic acid
1
(
0.175 g, 81%) as a white solid: m.p. 156±1588C;
NMR d 5.30 (s, 2H, CH O), 7.45 (m, 9H, C H
H
1
C H ), 7.62 (s, 1H, CHyN). Anal. for C H NO (C, H, N).
6. Experimental
2
6
4
6
5
15 13
3
6.1. Chemistry
6
.1.3. Peptide synthesis
All reagents and solvents for peptide synthesis were of
Melting points were determined on a Ko¯er hot-stage
apparatus and are uncorrected. IR spectra for comparison
reagent grade and used without any further puri®cation.
Peptides were synthesized by the continuous-¯ow solid
phase method on a Milligen 9050 Automatic Synthesizer
of compounds were taken as paraf®n oil mulls or as liquid
®
1
lms on a Mattson 1000 Series FTIR Spectrometer. H

5
54
M. Macchia et al. / Il Farmaco 53 (1998) 545±556
using standard Fmoc chemistry. Fmoc-Trp(Boc)-Novasyn
PA500 resin was used for all syntheses; couplings were
performed using HBTU/HOBt/NMM activation and a
was centrifuged at 48 000 £ g for 15 min at 48C. The pellet
was resuspended in 10 vols. of 20 mM Hepes-Tris (pH
7.4), and 5 mM EDTA (buffer HT) containing protease
inhibitors. The membrane homogenate was centrifuged at
248 000 £ g for 15 min at 48C. The tissue preparation was
either used immediately or stored in aliquots at ±808C. The
uterus was homogenized using a Polytron homogenizer in
10 vols. (w/v) of ice-cold buffer HT containing protease
inhibitors. The homogenate was centrifuged at 48 000 £ g
for 15 min at 48C. The resulting pellet was resuspended in
20 vols. of buffer HT containing protease inhibitors. The
membrane homogenate was ®ltered through 3 layers of
cheesecloth and centrifuged at 48 000 £ g for 15 min at
48C. The tissue preparation was either used immediately
or stored in aliquots at ±808C.
®
ve-fold excess of protected amino acid, recycling for 50
min in the column at a ¯ow rate of 3 ml/min. The same
protocol was used to incorporate 5-[N-(9-¯uorenylmethyl-
oxy)carbonyl]aminovaleric acid [28]. Fmoc deprotection
was obtained with 20 % (v/v) piperidine in DMF, for 9
min at 3 ml/min. The N-terminal residue was acylated on
the resin using a ®ve-fold excess of the relevant acid and
HBTU/HOBt/NMM activation. Peptides were cleaved from
the resin and deprotected from the side-chain protecting
groups with TFA/phenol/anisole (92:4:4, v/v/v; 1 h at
room temperature), precipitated with cold ether, ®ltered,
suspended in water and lyophilized. Peptides were subse-
quently analyzed by RP-HPLC (Beckman System Gold
apparatus) under the following conditions: Vydac C18
column (0:46 £ 15 cm); eluant A, water/0.1% TFA; eluant
B, acetonitrile/0.1% TFA; gradient from 10% to 85% B over
Binding assays were performed as described by Cody et al.
[29], with some modi®cations. Brie¯y, either cerebellar (5
mg of proteins) or uterus (10±20 mg of proteins) membranes
were incubated in 0.25 ml of buffer T (20 mM Tris±HCl, pH
1
2
5 min (method A) or from 20% to 95% B over 25 min
7.4, 2 mM EDTA, 0.1 mM bacitracin, 0.1 mM PMSF, 1 mg/
125
(
method B); ¯ow 1 ml/min; UV detection, 210 nm. Crude
ml leupeptin, and 5 mg/ml aprotinin) with 15 pM [ I]ET-1
(2000 Ci/mmol) for 2 h at 378C. Reactions were terminated
by the addition of 3 ml of ice-cold buffer T (50 mM Tris±
peptides were puri®ed to homogeneity by preparative HPLC
on a Beckman System Gold apparatus; Vydac C18 column
2
(
B over 100 min; ¯ow 8 ml/min; UV detection, 210 nm. The
2:2 £ 25 cm); eluants as above; gradient from 20% to 40%
HCl, pH 7.3, 0.1 mM bacitracin) and rapid ®ltration of
samples through GF/C glass ®bre ®lters which had been
soaked in buffer T containing 0.2% BSA (bovine serum
R and chromatographic purities observed are reported in
t
2
Table 1. All ®nal products were examined by fast ion
bombardment mass spectrometry; the measurements were
performed on triple quadrupole instrument (VG Quattro,
Fisons Instruments, Manchester, UK). The analysis condi-
albumin) for 24 h. The ®lters were then washed four times
with 3 ml of ice-cold buffer T . Assays were performed in
duplicate. Non-speci®c binding was de®ned in the presence
2
1
25
of 100 nM ET-1. [ I]ET-1 was diluted in buffer T contain-
1
1
tions were 10 keV Cs ion beam and 3-nitrobenzyl alcohol
ing 1 mg/ml BSA while ET-1 and ET-3 were dissolved in the
same buffer without BSA. Test compounds were diluted in
buffer T containing 1 mg/ml BSA and 1% DMSO to a
as the liquid matrix. The acquired mass spectra contained
substantial singly and doubly protonated molecules, as well
as a number of characteristic fragment ions. The measured
mass to charge (m/z) ratios of singly charged ions are
reported in Table 1.
1
concentration of 5 mM. In order to investigate the saturability
of speci®c binding, equilibrium binding studies were
1
25
performed by diluting [ I]ET-1 ( t 15 pM) with increasing
concentrations of ET-1 (12.5±500 pM). Analysis of binding
data (EBDA/LIGAND, Elsevier-Biosoft, UK; GraphPad
Prism Softwares, San Diego, CA) as saturation experiments
made it possible to calculate the equilibrium dissociation
6.2. Radioligand binding assay
125
[
I]ET-1 (2000 Ci/mmol) was obtained from Amersham
International plc (Little Chalfont, Buckinghamshire, UK).
ET-1 and ET-2 were purchased from Alexis Corporation
constant (K
sites in cerebellar (ET-B receptor) and uterus (ET-A recep-
tor) membranes. The K and Bmax for ET-B binding sites were
54 pM and 2.23 pmol/mg proteins, respectively, while ET-A
binding sites exhibited a K of 110 pM and a Bmax of 0.6 pmol/
mg protein. Competition binding studies were performed by
incubating either cerebellar or uterus membranes in buffer T
D
) and binding capacity (Bmax) values for binding
(
L aÈ ufel®ngen, Switzerland). Leupetin and aprotinin were
D
obtained from Boehringer-Mannheim (Mannheim,
Germany). Bacitracin, benzamidine and PMSF (phenyl-
methylsulfonyl ¯uoride) were products of Fluka Chemie
AG (Buchs, Switzerland). Other agents and reagents were
from standard commercial sources.
Male and female Sprague±Dawley rats (160±300 g)
were killed by cervical dislocation. The cerebellum was
homogenized using a Polytron homogenizer in 10 vols.
D
1
1
25
with [ I]ET-1 ( t 15 pM) and up to 9 concentrations of
displacing drugs ranging from 0.01 to 100 nM. Data were
analyzed as competition curves using the non-linear regres-
sion analysis of the GraphPad Prism program. The IC50
(
w/v) of ice-cold buffer containing 0.32 M sucrose, 20
values were converted to K
equation [30]. The agonists, ET-1 and ET-3, displayed K
values of approximately 169 and 119 pM, respectively, at
ET-B binding sites. K values were 176 pM and 4.7 nM for
ET-1 and ET-3, respectively, at the ET-A binding sites.
i
values by the Cheng and Prusoff
mM Hepes-Tris (pH 7.4), 5 mM EDTA and protease inhi-
bitors (0.1 mM benzamidine, 0.1 mM bacitracin, 0.1 mM
PMSF, and l mg/ml leupeptin). The homogenate was
centrifuged at 1000 £ g for 5 min at 48C. The supernatant
i
i

M. Macchia et al. / Il Farmaco 53 (1998) 545±556
555
6
.3. Computational details
[3] M. Yanagisawa, K. Kurihara, S. Kimura, Y. Tomobe, M. Kobayashi,
Y. Mitsui, Y. Yazaki, K. Goto, T. Masaki, A novel potent vasocon-
strictor peptide produced by vascular endothelial cells, Nature 332
Molecular mechanics (MM) and molecular dynamics
MD) calculations were performed by means of the program
(
1988) 411±415.
(
[
4] J.P. Huggins, J.T. Pelton, R.C. Miller, The structure and speci®city of
endothelin receptors ± their importance in physiology and medicine,
Pharmacol. Ther. 59 (1993) 55±123.
Discover [25] using the CVFF force®eld and a value for the
dielectric constant equal to 4 (distance-dependent). In all
molecules, the carboxylic groups were considered in their
negative charged form. The program does not explicitly take
into account the possibility of a hydrophobic folding which
should have a certain importance in determining the confor-
mational behaviour of rather ¯exible molecules endowed
with hydrophobic tails as those here considered.
[5] E.M. Ohlstein, J.D. Elliot, G.Z. Feurestein, R.R. Ruffolo, Endothelin
receptors: receptor classi®cation, novel receptor antagonists and
potential therapeutic targets, Med. Res. Rev. 16 (1996) 365±390.
[
6] C.A. Maggi, S. Giuliani, R. Patacchini, P. Santicioli, P. Rovero, A.
Giachetti, A. Meli, The C-terminal hexapeptide, endothelin-(16±21),
discriminates between different endothelin receptors, Eur. J. Pharma-
col. 166 (1989) 121±122.
Ê
A cut-off threshold of 16 A was used and no constraint
was applied.
[7] P. Rovero, R. Patacchini, C.A. Maggi, Structure-activity studies on
endothelin (16±21), the C-terminal hexapeptide of the endothelins, in
the guinea-pig bronchus, Br. J. Pharmacol. 101 (1990) 232±234.
The MM minimization procedure used a convergence
criterion of 0.0001 kcal/mol on the maximum derivative;
the MD simulation used a 1 fs step.
Superimpositions and molecular graphics representations
were performed using the program Insight II [25].
[
8] A.M. Doherty, W.L. Cody, P.L. DePue, J.X. He, L.A. Waite, D.M.
Leonard, N.L. Leitz, D.T. Dudley, S.T. Rapundalo, G.P. Hingorani,
S.J. Haleen, D.M. LaDouceur, K.E. Hill, M.A. Flynn, E.E. Reynolds,
Structure-activity relationship of C-terminal endothelin hexapeptide
antagonists, J. Med. Chem. 36 (1993) 2585±2594.
[
9] X.M. Cheng, S.S. Nikam, A.M. Doherty, Development of agents to
modulate the effects of endothelin, Curr. Med. Chem. 1 (1994) 271±
6.3.1. Superimposition of 1, PD 145065 and 3 on endothelin
Residue 17 of 1, PD 145065 and 3 was placed in the same
312.
[10] B.H. Stewart, E.L. Reyner, E. Tse, R.N. Hayes, S. Werness, W.L.
Cody, A.M. Doherty, In vitro assessment of oral delivery for hexa-
peptide endothelin antagonists, Life Sci. 58 (1996) 971±982.
conformation as the same residue of endothelin [16,17],
superimposed on it and ®xed; then the conformation of resi-
due 16 of 1, PD 145065 and 3 was fully optimized.
[
11] W.L. Cody, J.X. He, M.D. Reily, S.J. Haleen, D.M. Walker, E.L.
Reyner, B.H. Stewart, A.M. Doherty, Design of a potent combined
pseudopeptide endothelin-A/endothelin-B receptor antagonist, Ac-
DBhg16-Leu-Asp-Ile-[NMe]Ile-Trp21 (PD 156252): examination of
its pharmacokinetic and spectral properties, J. Med. Chem. 40 (1997)
2228±2240.
6
1
.3.1.1. Conformational analysis of 3, 6-12, PD 142893, PD
45065 and IRL2500
A 100 ps MD simulation at 1500 K was performed after a
0 ps initialization step; 11 conform-ations (one every 10
[
12] S. Pegoraro, P. Rovero, A. Sedo, R.P. Revoltella, J. Mizrahi, S. Tele-
maque, P. D'Orleans-Juste, Structure-activity relationship studies of
the C-terminus of endothelin-1, ET(16±21), Proc. 13th S. American
Peptide Symp., Edmonton, Canada, 1993, p. 362.
1
ps) were sampled and then fully mini-mized. In the case of
PD 142893 and PD 145065, additional MD simulations
were performed in the same way in order to check
whether the conformational search could be considered
exhaustive. We veri®ed that no signi®cant information
was obtained from the additional calculations about the
con-formational trend of the molecules and, therefore, this
could be considered a validation of the procedure.
[13] A. Sedo, S. Pegoraro, P. Rovero, R.P. Revoltella, A new endothelin C-
terminal analogue IBDP064 antagonizes endothelin-3-induced cell
proliferation, Folia Biol. 41 (1995) 97±105.
[
14] J.T. Hunt, V.G. Lee, P.D. Stein, A. Hedberg, E.C. Liu, D. McMullen,
S. Moreland, Structure activity relationship of monocyclic endothelin
analogs, Bioorg. Med. Chem. Lett. 1 (1991) 33±38.
[15] J.P. Tam, W. Liu, J.W. Zhang, M. Galantino, F. Bertolero, C. Cris-
tiani, F. Vaghi, R. De Castiglione, Alanine scan of endothelin: impor-
tance of aromatic residues, Peptides 15 (1994) 703±708.
6.3.2. Superimposition of IRL2500 on all other molecules
The side chains of the Trp21 residue of all the molecules
[
16] E.E. Abola, F.C. Bernstein, S.H. Bryant, et al., Protein Data Bank, in:
F.H. Allen, G. Bergerhoff, R. Sievers (Eds.), Crystallographic Data-
bases Information Content, Software System, Scienti®c Applications,
Data Commission of the International Union of Crystallography,
Bonn, 1987, pp. 107±132.
considered were superimposed. In some cases it was neces-
sary to adjust the conformation of these side chains and re-
minimize the new conformation thus obtained in order to
optimize the superimposition; however these arbitrary
adjustments were justi®ed by the rotational freedom of the
side chain, with the result that the new rotamers obtained
differed by less than 1 kcal/mol from the original ones.
[
[
17] N.H. Andersen, C. Chen, T.M. Marschner, S.R. Krystek Jr., D.A.
Bassolino, Conformational isomerism of endothelin in acidic aqueous
media: a quantitative/noisy analysis, Biochemistry 31 (1992) 1280.
18] E. Cassano, C. Galoppini, L. Giusti, M. Hamdan, M. Macchia, M.R.
Mazzoni, E. Menchini, S. Pegoraro, P. Rovero, A structure-activity
study of a C-terminal endothelin analogue, Folia Biol. (Praha) 44
(
1998) 11±14.
[
[
19] T. Fruth, H. Saika, L. Svensson, T. Pitterna, J. Sakaki, T. Okada, Y.
Urade, K. Oda, Y. Fujitani, M. Takimoto, T. Yamamura, T. Inui, M.
Makatani, M. Takai, I. Umemura, N. Teno, H. Toh, K. Hayakawa, T.
Murata, IRL2500: a potent ETB selective endothelin antagonist,
Bioorg. Med. Chem. Lett. 6 (1996) 2323±2328.
References
[
[
1] G.M. Rubanyi, M.A. Polokoff, Endothelins: molecular biology,
biochemistry, pharmacology, physiology and pathophysiology, Phar-
macol. Rev. 46 (1994) 325±415.
20] E. Cassano, M. Macchia, M. Hamdan, P. Rovero, Facile reduction of
peptide oxime endothelin antagonist during trialkylsilane/TFA clea-
vage after solid-phase synthesis, Lett. Pept. Sci. 3 (1996) 117±120.
2] T. Masaki, M. Yanagisawa, K. Goto, Physiology and pharmacology
of endothelins, Med. Res. Rev. 12 (1992) 391±421.

5
56
M. Macchia et al. / Il Farmaco 53 (1998) 545±556
[
[
21] E. Atherton, R.C. Sheppard, Solid Phase Peptide Synthesis: A Prac-
tical Approach, 1989, IRL Press, Oxford.
[27] S. Augelli, E. Corinne, S.D. Barrett, et al., US, 25 pp. Cont., in part of
US 5306718. CODEN: USXXAM, Patent, English, MARPAT
123:9340.
22] A. Balsamo, M.S. Bel®ore, M. Macchia, C. Martini, S. Nencetti, E.
Orlandini, A. Rossello, Synthesis and aldose reductase inhibitory
activity of N-(arylsulfonyl)- and N-(aroyl)-N-(arylmethyloxy)gly-
cines, Eur. J. Med. Chem. 29 (1994) 787±794.
[28] J.D. Bain, D.A. Wacker, E.E. Kuo, A.R. Chamberlin, Site-speci®c
incorporation of non-natural residues into peptides: effect of residue
structure on suppression and translation ef®ciencies, Tetrahedron 47
(1991) 2389±2400.
[
[
23] A. Lapucci, M. Macchia, E. Orlandini, F. Romagnoli, A. Rossello, G.
Chiellini, P. Cozzini, P. Domiano, (E)-Methyloxyimino-acetamides
as analogues of neuroleptic benzamides: synthesis and D2-dopami-
nergic binding af®nity, Farmaco 51 (1996) 33±39.
[29] W.L. Cody, J.X. He, P.L. DePue, L.A. Waite, D.M. Leonard, A.M.
Se¯er, J.S. Kaltenbronn, S.J. Haleen, D.M. Walker, M.A. Flynn, K.M.
Welch, E.E. Reynolds, A.M. Doherty, Structure-activity relationships
of the potent combined endothelin-A/endothelin-B receptor antago-
nist Ac-DDip(16)-Leu-Asp-Ile-Ile-Trp(21): development of endothe-
lin-B receptor selective antagonists, J. Med. Chem. 38 (1995) 2809±
2819.
24] L. Rovatti, O. Curcuruto, M. Hamdan, E. Cassano, C. Galoppini, P.
Rovero, Investigation of newly synthesized endothelin peptides by
HPLC/electrospray mass spectrometry, Rapid. Commun. Mass Spec-
trom. 10 (1996) 1504±1508.
[
[
25] Insight II Version 2.3; Discover Version 2.9.5 Biosym Technologies,
San Diego, CA.
[30] Y.C. Cheng, W.H. Prusoff, Relationship between the inhibition
i
constant (K ) and the concentration of inhibitor which causes 50%
26] R.W. Janes, D.H. Peapus, B.A. Wallace, The crystal structure of
human endothelin, Nature Struct. Biol. 1 (1994) 311±319.
inhibition (I50) of an enzymatic reaction, Biochem. Pharmacol. 22
(1973) 3099±3108.