CCK2 receptor antagonists containing the conformationally constrained phenylalanine derivatives, including the new amino acid Xic

Article abstract of DOI:10.1016/S0223-5234(02)01351-X

The conformationally constrained analogues of phenylalanine, tetrahydroisoquinoline-3-carboxylic acid (Tic), Sic, Hic and Nic, and the new amino acid Xic have been incorporated into a potent and highly selective cholecystokinin-2 (CCK₂) receptor antagonist (2) in place of the phenylalanine residue, producing compounds 15a-e. High selectivities for CCK₂ over CCK₁ were observed for compounds 15a-e. The in vitro profile of the analogue containing the Nic residue (15d) was identical to that of compound 2, whereas the alternative conformational constraints resulted in a significant loss of affinity. The apparent advantage of Nic in the context of these CCK₂ ligands was subsequently demonstrated to be statistically significant.

Full text of DOI:10.1016/S0223-5234(02)01351-X

Eur. J. Med. Chem. 37 (2002) 379–389
www.elsevier.com/locate/ejmech
Original article
CCK2 receptor antagonists containing the conformationally
constrained phenylalanine derivatives, including the new amino
acid Xic
Susan E. Gibson (ne´e Thomas) a,*, Nathalie Guillo a, Jerome O. Jones a,
Ildiko M. Buck b, S. Barret Kalindjian b, Sonia Roberts b, Matthew J. Tozer b,*
a Department of Chemistry, King’s College London, Strand, London WC2R 2LS, UK
b James Black Foundation, 68 Half Moon Lane, London SE24 9JE, UK
Received 23 August 2001; received in revised form 14 February 2002; accepted 20 February 2002
Abstract
The conformationally constrained analogues of phenylalanine, tetrahydroisoquinoline-3-carboxylic acid (Tic), Sic, Hic and Nic,
and the new amino acid Xic have been incorporated into a potent and highly selective cholecystokinin-2 (CCK2) receptor
antagonist (2) in place of the phenylalanine residue, producing compounds 15a–e. High selectivities for CCK2 over CCK1 were
observed for compounds 15a–e. The in vitro profile of the analogue containing the Nic residue (15d) was identical to that of
compound 2, whereas the alternative conformational constraints resulted in a significant loss of affinity. The apparent advantage
of Nic in the context of these CCK2 ligands was subsequently demonstrated to be statistically significant. © 2002 Published by
´
Editions scientifiques et me´dicales Elsevier SAS.
Keywords: CCK2; Gastrin; Conformational constrained amino acids; Tic; Phenylalanine; Antagonist
Many conformationally constrained analogues of
1. Introduction
amino acids have been synthesised and some have been
used in biological studies. Focussing on phenylalanine,
typical examples of conformationally constrained ana-
logues include 3-phenylproline [2], b,b-diphenylalanine
[3], a,b-dimethylphenylalanine [4], and b-methylpheny-
lalanine [5]. Of particular relevance to the work re-
ported herein is the phenylalanine analogue
1,2,3,4-tetrahydroisoquinoline-3-carboxylic acid (Tic)
1a (Fig. 1). Tic has been used to considerable effect in
medicinal chemistry, its presence in ligands having been
shown to influence both affinity and selectivity, such
that it is now a commercially product. In one example,
replacement of D-Phe1 with D-Tic in the somatostatin-
derived cyclic peptide CTP led to a significant increase
in selectivity and affinity for the m-opioid receptor
producing one of the most potent and selective m-opioid
receptor antagonists described to date [6]. In a second
example, replacement of Phe2 in the dermorphin-
derived d-opioid receptor antagonist Tyr-Phe-Phe-Phe-
NH2 led to a dramatic improvement in its activity
profile and stability [7]. Further modification has led to
The use of conformationally constrained amino acids
to probe the binding mode of bioactive molecules to
their receptors is a commonly used approach to the
design of highly selective and active compounds [1].
Reducing the conformational freedom of a ligand may
alter (a) the binding affinity of the ligand at a given
receptor, (b) the selectivity of the ligand between diffe-
rent receptors, and (c) the stability of the ligand with
respect to enzymatic degradation. Examination of the
effect of restricting the conformational freedom of a
given ligand on these properties may lead to increased
insight into the bioactive conformation of the ligand
and hence ultimately to the generation of more potent
and selective molecules.
* Correspondence and reprints. Present address: Medivir UK, Pe-
terhouse Technology Park, 100 Fulbourn Road, Cambridge CB1
9PT, UK.
E-mail address: matt.tozer@medivir.com (M.J. Tozer).
´
0223-5234/02/$ - see front matter © 2002 Published by Editions scientifiques et me´dicales Elsevier SAS.
PII: S0223-5234(02)01351-X

380
S.E. Gibson (ne´e Thomas) et al. / European Journal of Medicinal Chemistry 37 (2002) 379–389
Fig. 1. Conformationally constrained analogues of Phe.
dipeptides, which display some of the best d-opioid
receptor selectivities known [8]. The continuing level of
interest in Tic is reflected by the fact that several
analogues of Tic, such as a-methyl Tic [9], a,b-
dimethyl-Tic [4], b-phenyl-Tic [10], benzo[f]Tic, ben-
zo[g]Tic, and benzo[h]Tic [11] have been the targets of
synthetic Studies.
ethylamine gave the Heck substrates 4. Cyclisation of 4
with catalytic amounts of palladium acetate proceeded
smoothly to generate products 5 containing seven-,
eight-, nine- and ten-membered rings in 55, 73, 86, and
69% yield, respectively. Hydrogenation and hydrolysis
of 5 provided the required hydrochloride salts 6b–e.
The commercially available hydrochloride salt of Tic
(6a) and the hydrochloride salts of Sic, Hic, Nic and
Xic (6b–e) were converted into analogues of the com-
pound 2 by the route depicted in Fig. 4. N-Protection
of the hydrochloride salts 6a–e with di-tert-butyl dicar-
bonate ((BOC)2O)gave acids 7a–e, which were coupled
with 3,5-bis(benzyloxycarbonyl)aniline (8) using bro-
motris(pyrrolidine)phosphonium hexafluorophosphate
(PyBroP®) to give, after deprotection of the intermedia-
tes 9a–e with trifluoroacetic acid, amines 10a–e. Reac-
tion of amines 10a–e with the anhydride 11 [13] gave
acids 12a–e, which were subsequently coupled with
1-adamantanemethylamine (13) using either PyBroP®
or 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hy-
drochloride (EDC)–1-hydroxybenzotriazole (HOBt) to
give amides 14a–e. Subsequent hydrogenolysis revealed
the diacids 15a–e, which were characterised by 1H-
NMR and mass spectroscopy, and by microanalysis of
their bis(N-methyl-D-glucamine) salts.
We synthesised the novel seven-, eight- and nine-
membered analogues of Tic, i.e. Sic 1b, Hic 1c and Nic
1d [12], in the belief that the incorporation of a series of
compounds with varying degrees of conformational
constraint into biologically active peptides or non-pep-
tides would lead to greater insight into the conforma-
tional preferences of the ligand under investigation and
the nature of its interaction with the active site. Our
first test of this hypothesis was carried out on the
non-peptide cholecystokinin-2 (CCK2) receptor antago-
nist 2 (Fig. 2) [13], in which Phe was a key component.
It seemed from the preliminary results, using Tic, Sic,
Hic and Nic, that the size of the ring in the phenylala-
nine replacement is a determining factor for CCK2
receptor affinity [14]. The full results and interpretation
of this study, and its extension to include the, pre-
viously unreported, ten-membered analogue of Tic, i.e.
Xic 1e, are described herein. The study represents the
first application of the conformationally constrained
amino acids Sic, Hic, Nic and Xic.
2. Chemistry
The hydrochloride salts of the amino acids Sic, Hic,
Nic and Xic were synthesised using the recently re-
ported Heck cyclisation route designed and developed
in our laboratories [12]. Our approach is summarised in
Fig. 3. Reductive amination of aldehydes 3 [12] (m=1–
4) with (9)-serine methyl ester followed by N-protec-
tion with di-tert-butyl dicarbonate and introduction of
a carbonꢀcarbon double bond using tosyl chloride-tri-
Fig. 2. A potent and selective CCK2 antagonist containing Phe.

S.E. Gibson (ne´e Thomas) et al. / European Journal of Medicinal Chemistry 37 (2002) 379–389
381
Fig. 3. Synthesis of the higher homologues of Tic.
Fig. 4. Synthesis of CCK2 antagonists containing Tic, Sic, Hic, Nic and Xic.
a functional assay based on the pentagastrin stimulated,
lumen-perfused, isolated, immature rat stomach [15].
The second was a radioligand binding assay, which
used [125I]BH-CCK-8S to label CCK2 receptors in
3. Pharmacology
Affinity estimates for these compounds at CCK2
receptors were determined in two assays. The first was

382
S.E. Gibson (ne´e Thomas) et al. / European Journal of Medicinal Chemistry 37 (2002) 379–389
mouse cortical membranes [16]. This choice was made,
because, the CCK2 receptor types have been shown to
be homogeneous within each of these assays, and be-
cause, L-365260, a well established reference competi-
tive antagonist, displayed different affinity estimates
(rat stomach pKB=7.5490.03 [17]; mouse cortex
pKi=8.4290.03 [16]) between the two assays, sugges-
ting that there are different CCK2 receptor sub-types
[16,17]. The data from the immature rat stomach assay
were examined to determine whether the affinity esti-
mates for each compound were significantly different
from each other. These pKB estimates were compared
using one-way analysis of variance (ANOVA) followed
by a Bonferroni modified t-test [21]. In order to assess
the selectivity of the new ligands for the CCK2 over the
CCK1 receptor, affinity estimates at CCK1 receptors
were determined using guinea-pig pancreatic mem-
branes in a radioligand binding assay [18]. The isolated
mouse stomach assay [15] is analogous to the rat
model, but it has been shown to be particularly sensi-
tive for the detection of partial agonists at the CCK2
receptor [19]. Compounds 15a–e were also tested in this
assay to ascertain whether the inclusion of the new
amino acids altered the behaviour of the ligands.
4. Results and discussion
An important characteristic of the parent ligand 2 is
its reversed selectivity with respect to L-365260 in the
two CCK2 assays (Table 1). The same sense and de-
gree of selectivity was maintained by the conforma-
tionally constrained analogues 15a–e. In comparing
compound 2 with compounds 15a–e it is noted that
the former is a single enantiomer, while the latter are
racemates. The biological data for the antipode of
compound 2, compound 16 (derived from D-Phe), are
also presented in Table 1 and it can be seen that
chirality relatively little effect on affinity and selecti-
vity. The differences between the affinities of com-
pounds 2 and 16 are in the order of 0.5 log unit or
less, which is within the limits of statistical error.
From the data for compounds 2 and 15a–e, it may be
concluded that the selectivity ratio is not dependent on
the ring-size of the conformationally restricted amino
acids. In contrast, examination of the data from both
CCK2 assays reveals broadly parallel trends that relate
affinity to ring-size. Thus, considering the two sets of
data together, the affinity estimates for compounds
15a–c appeared to be least an order of magnitude
lower than those of compound 2. The loss of affinity
for the ten-membered ring analogue 15e was less pro-
nounced. However, the data for the nine-membered
analogue 15d were indistinguishable from those of
compound 2. The two sets of data provide compelling
evidence for the pre-eminence of compound 15d. How-
ever, a more rigorous statistical analysis was required
to determine the significance of the differences between
the compounds.
Table 1
Comparison of CCK2 and CCK1 receptor affinity estimates for
compounds 2, 15a–e and 16
Compound
number a
n
Rat stomach b Mouse
cortex c
G.P.
pancreas f
2
16
15a
15b
–
–
1
2
9.0890.10 d
8.7290.38
7.8490.13
7.5090.10
8.2890.13 e,g 5.8390.27
8.3990.09
6.7590.10
6.639
5.2
4.7390.04 g
4.8390.08 g
Compounds 15a–e behaved as competitive antago-
nists to pentagastrin in the immature rat stomach as-
90.14 g
15c
15d
15e
3
4
5
7.0390.17
9.0190.21
8.3590.11
6.6590.10
8.3090.20 g
7.9590.07
4.2990.08
4.7490.04 g
5.1490.12
say. Each produced
a
significant rightward
displacement of the pentagastrin dose–response
curves, with no significant change in either maxima or
mid-point slope parameter. With such consistent be-
haviour across the series of compounds, data were
collected for a high number of iterations of the assay
and analysed to determine whether the affinity esti-
mates (pKB) for each compound were significantly dif-
ferent from each other. Firstly, individual agonist (A)
concentration-effect (E) curves were fitted to the Hill
equation (Eq. (1)), to provide estimates of midpoint
slope (nH), midpoint location (log[A]50) and upper
asymptote (h) as described previously [20].
a Compounds 2 and 15b–e were tested as their bis(N-methyl-
D-glu-
camine) salts.
b pKB9standard error of mean (S.E.M.) values were estimated
from single shifts of pentagastrin concentration–effect curves in the
isolated lumen–perfused immature rat stomach, in which the com-
pounds behaved as simple competitive antagonists. The number of
experiments and test concentrations for each compound were as
follows: 15a (12, 1 mM), 15b (12, 1 mM), 15c (15, 1 mM), 15d (12, 0.03
mM), and 15e (11, 0.1 mM).
c pIC509S.E.M. values were estimated from three separate compe-
tition experiments in which 20 pM [125I]BH-CCK-8S for CCK2 was
used to CCK2 label binding sites in the mouse cortical homogenates.
The pIC50:pKi when (1+[L]/KL) approximated to unity.
d Value was taken from reference [26].
E=h[A]nH/[A]5nH0+[A]Hn
(1)
e Value is taken from reference [13].
f pIC509S.E.M. competition with 20 pM [125I]BH-CCK-8S at
CCK1 guinea pig pancreatic cells from three (one for compound 16)
separate experiments.
Secondly, using Eq. (2), pKB values were estimated
by comparing the location of the pentagastrin E/[A]
curves in the presence of each antagonist (B) with the
corresponding control curve (Table 1).
g The slope of these competition curves was significantly different
from unity. Therefore, a strict comparison of the pIC50 values is
potentially unreliable.

S.E. Gibson (ne´e Thomas) et al. / European Journal of Medicinal Chemistry 37 (2002) 379–389
383
Table 2
Results of Bonferroni modified t-test to compare mean pKB values
5. Conclusions
We have investigated the effect of replacing the
phenylalanine residue of the CCK2 antagonist 2 with a
series of conformationally constrained phenylalanine
analogues, which has been expanded by the addition of
the ten-membered ring analogue, Xic. From the work
presented here it is evident that the size of the confor-
mational constraint directly impinges upon the in vitro
profile of the constrained ligands. The most pro-
nounced effect was the drop in CCK2 affinity for all but
the Nic-containing analogue, the biological behaviour
of which closely resembled that of compound 2. It is
possible, therefore, that information on the bioactive
conformation of compound 2 may be gained from
further studies of compound 15d. The exact role of the
conformational constraint remains to be determined.
This could involve the precise positioning of either the
phenylalanine side-chain itself or the other functional
groups on the two arms of these molecules. A molecu-
lar modelling-based investigation of the ligands dis-
cussed here is in hand. The absence of protein crystal
structures for G-protein coupled receptors, such as the
CCK2 receptor, is a recurrent problem. The construc-
tion of theoretical receptor structures has been possible
using the a-carbon template of the transmembrane
helices of rhodopsin [23]. It is hoped that conforma-
tionally constrained ligands, such as 15a–e, may even-
tually prove useful tools for both homology- and
ligand-based modelling approaches. Tic has been widely
used as a conformationally constrained analogue of
phenylalanine. From this study it is evident that consid-
eration of a broader palette, including Sic, Hic, Nic and
Xic, may be of value.
15b
15c
15d
15a
15e
15b
15c
15d
15a
15e
–
2.29
–
9.63
−3.93
6.26
6.97
9.63
–
5.39
2.99
−1.57
-3.93
5.39
–
3.83
6.26
2.99
2.30
–
2.29
6.97
−1.57
3.83
2.30
t-Values were calculated, with those indicated in bold being statisti-
cally significantly different from each other at P=0.05 (tcrit=2.85).
pKB=log(r−1)−log B
(2)
where r=log[A]50B/log[A]50 and [A]50B is the concentra-
tion of agonist that produces 50% response in the
presence of antagonist B. Comparison of the individual
pKB estimates, using one-way ANOVA, indicated a
significant difference in mean values between groups
(F=26.78; total d.f.=61). Subsequent Bonferroni
modified t-tests [21], indicated significant differences in
mean pKB estimates as shown in Table 2. Therefore, it
can be seen that the nine-membered ring compound
(15d) expressed significantly higher affinity than all of
the other cyclic analogues. Furthermore, the ten-mem-
bered analogue (15e) was significantly more potent than
the seven- (15b) and eight-membered (15c) ring ana-
logues. It is important to note from these results that
statistical significance could not be ascribed to diffe-
rences in potency of 0.5 log unit or less.
Similar statistical analysis of the data from the mouse
cortex assay was not performed, because, the Hill
slopes obtained for several of the compounds were
greater than unity (Table 1). Although the reason for
this variation from unity has not been found, the
possibility that it could be attributed to residual efficacy
was explored. The mouse stomach assay has been
shown to be more sensitive than the rat stomach assay
to the expression of agonism. In the mouse stomach
assay partial agonism was observed for PD-134308, a
reference CCK2 receptor antagonist [22], which was
found to be a competitive antagonist in the rat stomach
assay [19]. Compounds 15a–e were examined in the
mouse stomach assay and no evidence for agonism was
found at the concentration tested (10 mM).
6. Experimental
6.1. Chemistry
Anhydrous N,N-dimethylformamide (DMF) and
acetonitrile were purchased from Aldrich Chemical
Company. Dichloromethane (dichloromethane) was
distilled from calcium hydride. Triethylamine and diiso-
propylethylamine were distilled from and stored over
potassium hydroxide pellets. The hydrogen used in the
hydrogenation was BOC Grade 0 (99.99% minimum
purity). Bromo-tris-pyrrolidino-phosphonium hexa-
fluorophosphate (PyBroP®) was purchased from Nova-
biochem. HOBt, 4-dimethylaminopyridine (DMAP),
EDC, [(BOC)2O], 1-adamantanemethylamine 13, and
Tic–HCl (6a) were purchased from Aldrich Chemical
Company and were used without purification. 3,5-Bis(-
benzyloxycarbonyl)aniline 8 [24], benzimidazole-5,6-di-
carboxylic acid anhydride 11 [13], Sic. HCl (6b) [12],
Hic–HCl (6c) [12] and Nic–HCl (6d) [12] were synthe-
An important property of compound 2 was the 3000-
fold selectivity for the CCK2 receptor (rat stomach)
over the CCK1 receptor (guinea pig pancreas) [13].
Compounds 15a–e were similarly found to be highly
selective for the CCK2 receptor (Table 1). The CCK2/
CCK1 selectivities for compound 15d were in the order
of 18 000-fold (rat stomach) and 3000-fold (mouse
cortex).

384
S.E. Gibson (ne´e Thomas) et al. / European Journal of Medicinal Chemistry 37 (2002) 379–389
sised according to published procedures. Nic–HCl (6e),
an analogue of 6a–d, was synthesised using a modifica-
tion of the published procedures for 6a–d [12]. Melting
points were recorded on a Bu¨chi 510 or a Gallenkamp
capillary melting point apparatus and are uncorrected.
NMR spectra were recorded in CDCl3 at room temper-
ature (r.t.) (unless noted otherwise) on Bruker DRX
300, Bruker DRX 400 and Bruker AM-500 spectrome-
ters; J values are reported in Hertz. Infrared (IR)
spectra were obtained on a Unicam Mattson 5000
FTIR spectrometer. Mass spectra were recorded on
Micromass Platform 2 and Micromass Autospec instru-
ments. Elemental analyses were performed by the Impe-
rial College Microanalytical Service or the School of
Pharmacy Microanalytical Service. Flash chromatogra-
phy was performed using Merck Kieselgel 60 silica
grade 9385 (230–400 mesh). Thin-layer chromatogra-
phy (TLC) was performed on Merck 60 F254 silica gel,
in a closed vessel, and visualisation was achieved by
UV light (254 nm) and/or by dipping into a solution of
phosphomolybdic acid in EtOH, followed by heating
with a heat gun. The solvent conditions used for TLC
analysis corresponded to those given for the purifica-
tion by column chromatography.
mixture of rotamers) l 1.38, 1.44 (2×s, 9H), 2.87–3.00
(m, 2H), 3.05–3.15 (m, 1H), 3.25–3.33 (m, 2H), 4.16–
4.30 (m, 1H), 4.92, 5.22 (2×br m, 1H), 7.09–7.13 (m,
4H), 10.03 (br s, 1H); IR (neat) wmax 3500–3100 m
(OꢁH), 3057 m (CꢁH aromatic), 2980 m, 2920 m (CꢁH
alkane), 1692 and 1675 s cm−1 (CꢀO carbamate and
carboxylic acid); MS (CI) m/z 309 (MNH+4 , 6%), 292
(17), 253 (47), 236 (30), 192 (41), 146 (100); Anal. Calc.
for C16H21NO4: C, 65.98; H, 7.22; N, 4.81. Found: C,
65.70; H, 6.97; N, 4.53%.
6.1.1.3. N-(tert-butoxycarbonyl)-1,2,3,4,5,6-hexahydro-
3-benzazocine-2-carboxylic acid (7c). An oil (220 mg,
1
0.72 mmol, 69%); H-NMR (300 MHz, 2:3 mixture of
rotamers) l 1.21, 1.37 (2×s, 9H), 1.75–1.85 (m, 1H),
2.10–2.20 (m, 1H), 2.76 (t, 2H, J=6), 2.77–2.90 (m,
1H), 3.13–3.30 (m, 2H), 3.72, 3.89 (2×br d, 1H,
J=14.5), 4.91–4.96 and 5.15–5.45 (2×m, 1H), 7.12–
7.32 (m, 4H), 11.62 (br s, 1H); IR (neat) wmax 3300–
3000 br m, 3097 m, 3041 m, 3015 m (CꢁH aromatic),
2974, 2938 m (CꢁH alkane), 1695 s and 1667 m cm−1
(CꢀO carbamate and carboxylic acid); MS (CI) m/z 323
(MNH+4 , 4%), 306 (16), 267 (87), 250 (52), 206 (43), 160
(100). Anal. Calc. for C17H23NO4: C, 66.89; H, 7.54; N,
4.59. Found: C, 66.78; H, 7.36; N, 4.33%.
6.1.1. General procedure for the preparation of the Boc
amino acids 7a–e
6.1.1.4. N-(tert-butoxycarbonyl)-2,3,4,5,6,7-hexahydro-
1H-3-benzazonine-2-carboxylic acid (7d). An oil (238
mg, 0.746 mmol, 76%); H-NMR (300 MHz, 2:3 mix-
1
6.1.1.1. N-(tert-butoxycarbonyl)-1,2,3,4-tetrahydroiso-
quinoline-3-carboxylic acid (7a). Di-tert-butyl dicarbo-
nate (1.26 g, 5.79 mmol) and triethylamine (1 mL, 10%
v/v) were added to a suspension of 1,2,3,4-tetrahy-
droisoquinoline-3-carboxylic acid hydrochloride (6a)
(1.03 g, 4.83 mmol) in a 1:1 mixture of 1,4-dioxan–wa-
ter (10 mL). The reaction mixture was stirred at r.t. for
48 h. The solution was diluted with dichloromethane
(30 mL) and washed with 5% aqueous (aq.) potassium
hydrogensulfate (40 mL), brine (20 mL) and dried
(MgSO4). Evaporation of the solvents in vacuo gave an
oil. Purification by column chromatography (SiO2;
dichloromethane–EtOAc, 98:2) afforded the title com-
pound as a white powder (0.98 g, 3.52 mmol, 73%)
melting point (m.p.) 116–117 °C (lit. [25] 122–
123.5 °C); 1H-NMR (300 MHz, 3:2 mixture of ro-
tamers) l 1.42, 1.52 (2×s, 9H), 3.16–3.30 (m, 2H),
4.50 (d, 1H, J=16.5), 4.60–4.70 (m, 1H), 4.80, 5.10
(2×br s, 1H), 7.13–7.21 (m, 4H); IR (KBr) wmax 3400–
3100 m (OꢁH), 3054 m (CꢁH aromatic), 2984, 2958 w
(CꢁH alkane), 1723 vs and 1698 s cm−1 (CꢀO carba-
mate and carboxylic acid); mass spectroscopy (MS)
(CI) m/z 295 (MNH+4 , 4%), 278 (10), 239 (68), 222 (32),
178 (98), 132 (100).
ture of rotamers) l 0.92–1.06 (m, 1H), 1.30–1.43 (m,
1H), 1.49, 1.51 (2×s, 9H), 1.72–2.10 (m, 2H), 2.75–
3.05 (m, 3H), 3.30–4.15 (m, 4H), 7.13–7.24 (m, 4H);
IR (neat) wmax 3400–3000 br m, 3099 m, 3061 m, 3012
m (CꢁH aromatic), 2973 m, 2867 m (CꢁH alkane), 1740
s, 1704 s cm−1 (CꢀO carbamate and carboxylic acid);
MS (CI) m/z 320 (MH+, 100%), 264 (92), 220 (40), 174
(56). Anal. Calc. for C18H25NO4: C, 67.71; H, 7.84; N,
4.39. Found: C, 67.84; H, 7.57; N, 4.11%.
6.1.1.5. N-(tert-butoxycarbonyl)-1,2,3,4,5,6,7,8-octahy-
dro-3-benzazecine-2-carboxylic acid (7e). An off-white
1
solid (148 mg, 0.44 mmol, 63%) m.p. 59–61 °C; H-
NMR (300 MHz, 2:3 mixture of rotamers) l 1.24–1.86
(m, 6H), 1.26, 1.39 (2×s, 9H), 2.59–2.69 (m, 2H),
2.79–2.86 (m, 2H), 3.15–3.43 (m, 3H), 7.16–7.25 (m,
4H); MS (EI) m/z 333 [M+, 3%], 289 (2), 277 (19), 232
(51), 188 (61), 57 (100). Anal. (C19H27NO4) C, H, N.
6.1.2. The preparation of compounds 10a–d
6.1.2.1.
3-[[[3,5-Bis(benzyloxycarbonyl)phenyl]amino]-
carbonyl]-1,2,3,4-tetrahydroisoquinoline (10a). Dry di-
isopropylethylamine (0.126 mL, 0.72 mmol) and
3,5-bis(benzyloxycarbonyl)aniline (8) (130 mg, 0.36
mmol) were added to a solution of 7a (100 mg, 0.36
mmol) in dry dichloromethane (2.6 mL) kept under
6.1.1.2. N-(tert-butoxycarbonyl)-2,3,4,5-tetrahydro-1H-
3-benzazepine-2-carboxylic acid (7b). Colourless oil
1
(137 mg, 0.47 mmol, 78%); H-NMR (300 MHz, 2:3

S.E. Gibson (ne´e Thomas) et al. / European Journal of Medicinal Chemistry 37 (2002) 379–389
385
argon. After 5 min, PyBroP® (168 mg, 0.36 mmol) and
DMAP (3–4 mg, 0.03 mmol) were successively added
and the resulting mixture was stirred for 19 h at r.t.
Dichloromethane (10 mL) was added and the organic
layer washed with 5% aq. potassium hydrogensulphate
solution (2×5 mL), saturated aq. sodium hydrogencar-
bonate solution (5 mL), brine (5 mL) and dried
(MgSO4). Column chromatography (SiO2; hexane–
ethyl acetate, 9:1) afforded a mixture of the coupling
product and 3,5-bis(benzyloxycarbonyl)aniline which
was then stirred in trifluoroacetic acid (1 mL) for 1 h at
r.t. Trifluoroacetic acid was removed in vacuo and the
resulting yellow oil was dissolved in dichloromethane
(10 mL). The organic layer was washed with a 5%
sodium carbonate solution (2×5 mL), brine (5 mL)
and dried (MgSO4). Column chromatography (SiO2;
hexane–ethyl acetate, 4:1 then 1:1) afforded compound
10a as a white powder (86 mg, 0.166 mmol, 46%) m.p.
107–109 °C; 1H-NMR (500 MHz) l 2.91 (dd, 1H,
J=16.3, 10.2), 3.33 (dd, 1H, J=16.3, 5.5), 3.70 (dd,
1H, J=10.2, 5.5), 3.99 (d, 1H, J=16.1), 4.05 (d, 1H,
J=16.1), 5.39 (s, 4H), 7.08–7.10 (m, 1H), 7.18–7.21
(m, 3H), 7.33–7.47 (m, 10H), 8.48–8.51 (m, 3H), 9.65
(s, 1H); IR (KBr) wmax 3322 w, 3302 w and 3281 w
(NꢁH), 1725 vs (CꢀO ester), 1687 s (CꢀO amide), 1531
s (NH bending), 1243 vs cm−1 (CꢁO); MS (FAB) m/z
521 (MH+, 74%), 132 (100), 91 (46), 77 (19). Anal.
Calc. for C32H28N2O5: C, 73.82; H, 5.42; N, 5.38.
Found: C, 73.89; H, 5.36; N, 5.33%.
(CI) m/z 549 (MH+, 37%), 160 (100), 91 (56); HRMS
(FAB) Calc. for C34H33N2O5 (MH+) 549.2389, Found:
549.2418.
6.1.2.4.
2-[[[3,5-Bis(benzyloxycarbonyl)phenyl]amino]-
carbonyl]-2,3,4,5,6,7-hexahydro-1H-3-benzazonine (10d).
An oil (187 mg, 0.333 mmol, 47%); 1H-NMR (400
MHz) l 1.16–1.86 (m, 3H), 2.13–2.17 (m, 1H), 2.76–
2.79 (m, 2H), 2.85–2.95 (m, 2H), 3.28–3.30 (m, 2H),
3.45–3.49 (m, 1H), 5.39 (s, 4H), 7.11–7.23 (m, 4H),
7.33–7.47 (m, 10H), 8.47–8.48 (t, 1H, J=1.5), 8.49–
8.50 (d, 2H, J=1.5), 10.27 (s, 1H); IR (KBr) wmax 3440
w, 3363 w and 3233 br w (NꢁH), 1722 vs (CꢀO ester),
1699 vs (CꢀO amide), 1521 s (NH bending), 1237 s
cm−1 (CꢁO); m/z (CI) 563 (MH+, 60%), 174 (100), 91
(64). HRMS (FAB) Calc. for C35H35N2O5 (MH+)
563.2546, Found: 563.2553.
6.1.3. 2-[[[3,5-Bis(benzyloxycarbonyl)phenyl]amino]car-
bonyl]-1,2,3,4,5,6,7,8-octahydro-3-benzazecine (10e)
6.1.3.1.
2-[[[3,5-Bis(benzyloxycarbonyl)phenyl]amino]-
carbonyl]-N-(tert-butoxycarbonyl)-1,2,3,4,5,6,7,8-octa-
hydro-3-benzazecine (9e). Purified 3,5-bis(benzyloxycar-
bonyl) aniline (8) (169 mg, 0.47 mmol) and diisopropy-
lethylamine (0.136 mL, 0.78 mmol) were added to a
stirred solution of 7e (130 mg, 0.39 mmol) in dry
dichloromethane (2.6 mL, 0.15 M). After 10 min, Py-
BroP® (0.219 g, 0.47 mmol) and DMAP (3–4 mg, :10
mol%) were added in one portion and the resulting
6.1.2.2.
2-[[[3,5-Bis(benzyloxycarbonyl)phenyl]amino]-
mixture stirred for 23 h at r.t. under argon.
carbonyl]-2,3,4,5-tetrahydro-1H-3-benzazepine (10b). A
white powder (123 mg, 0.231 mmol, 49%) m.p. 126–
128 °C; 1H-NMR (400 MHz) l 2.87–2.94 (m, 2H),
2.98–3.04 (m, 1H), 3.09 (dd, 1H, J=14.8, 9.4), 3.26–
3.32 (m, 1H), 3.49 (dd, 1H, J=9.4, 2.6), 3.57 (dd, 1H,
J=14.8, 2.6), 5.38 (s, 4H), 7.09–7.12 (m, 1H), 7.14–
7.18 (m, 2H), 7.23–7.25 (m, 1H), 7.33–7.47 (m, 10H),
8.46–8.48 (m, 3H), 9.54 (s, 1H); IR (KBr) wmax 3256 br
w (NꢁH), 1716 vs (CꢀO ester), 1663 s (CꢀO amide),
1543 s (NH bending), 1224 s cm−1 (CꢁO); MS (CI) m/z
535 (MH+, 55%), 362 (39), 189 (20), 174 (28), 146
(100), 91 (68). Anal. Calc. for C33H30N2O5: C, 74.14; H,
5.66; N, 5.24. Found: C, 74.12; H, 5.61; N, 5.22%.
Dichloromethane (10 mL) was added and the organic
solution washed successively with 5% aq. potassium
hydrogensulphate (2×10 mL), saturated aq. sodium
hydrogencarbonate (10 mL) and brine (10 mL). The
combined organic extracts were dried (MgSO4) and the
solvents evaporated in vacuo to afford an orange oil.
Purification by flash column chromatography (SiO2;
hexane–ethyl acetate, 2:1) afforded compound 9e as a
white solid (155 mg, 0.226 mmol, 58%) m.p. 67–70 °C;
1H-NMR (300 MHz, 2:3 mixture of rotamers) l 1.12–
1.84 (m, 6H), 1.26, 1.35 (2×s, 9H), 2.54–2.72 (m, 2H),
2.89–3.50 (m, 6H), 5.40 (s, 4H), 7.09–7.61 (m, 15H),
8.37–8.51 (m, 2H); MS (FAB) m/z 677 (MH+, 18%),
577 (47), 362 (12), 188 (58), 91 (100). Anal. Calc. for
C41H44N2O7: C, 72.76; H, 6.55; N, 4.14. Found: C,
72.48; H, 6.64; N, 4.08%.
6.1.2.3.
2-[[[3,5-Bis(benzyloxycarbonyl)phenyl]amino]-
carbonyl]-1,2,3,4,5,6-hexahydro-3-benzazocine (10c). A
white powder (160 mg, 0.292 mmol, 52%) m.p. 125–
129 °C; 1H-NMR (400 MHz) l 1.70–1.72 (m, 1H),
1.94–1.99 (m, 1H), 2.77–2.87 (m, 2H), 2.90 (m, 2H),
3.17 (dd, 1H, J=13.8, 8.8), 3.40 (dd, 1H, J=13.8, 4.5),
3.56 (dd, 1H, J=8.8, 4.5), 5.39 (s, 4H), 7.13–7.21 (m,
4H), 7.33–7.47 (m, 10H), 8.43 (d, 2H, J=1.4), 8.47 (t,
1H, J=1.4) and 9.42 (s, 1H); IR (KBr) wmax 3358 w,
3245 w (NꢁH), 1719 vs (CꢀO ester), 1675 s (CꢀO
amide), 1532 s (NH bending), 1229 s cm−1 (CꢁO); MS
6.1.3.2. Compound 10e. Compound 9e (155 mg, 0.23
mmol) was stirred in trifluoroacetic acid (2.3 mL, 0.1
M) at r.t. for 1 h. The solvent was evaporated in vacuo
and the resulting orange oil dissolved in
dichloromethane (10 mL). The organic solution was
washed successively with 5% w/w aq. sodium hydrogen-
carbonate (2×10 mL) and brine (10 mL). The organic
extracts were dried (MgSO4) and the solvent evaporated

386
S.E. Gibson (ne´e Thomas) et al. / European Journal of Medicinal Chemistry 37 (2002) 379–389
in vacuo to afford a yellow oil. Purification by flash
column chromatography (SiO2; hexane–ethyl acetate,
1:1) afforded the title compound 10e as fluffy white
crystals (129 mg, 0.22 mmol, 98%). m.p. 60–63 °C;
1H-NMR (300 MHz) l 1.24–1.99 (m, 7H), 2.74–2.78
(m, 2H), 2.83–2.99 (m, 3H), 3.41–3.50 (m, 1H), 3.61–
3.71 (m, 2H), 5.39 (s, 4H), 7.14–7.48 (m, 15H), 8.45–
8.54 (m, 2H); MS (FAB) m/z 577 (MH+, 79%), 188
(75), 154 (100), 136 (72). Anal. Calc. for C36H36N2O5:
C, 74.98; H, 6.29; N, 4.86. Found: C, 74.69; H, 6.35; N,
4.62%.
7.5%), 691 (2), 521 (5), 336 (100), 91 (50). Anal. Calc.
for C52H49N5O7: C, 72.96; H, 5.77; N, 8.18. Found: C,
72.69; H, 5.61; N, 7.95%.
6.1.4.2. 5-[[[[3,5-Bis(benzyloxycarbonyl)phenyl]amino]-
carbonyl]-[2-(1,2,3,4,5-tetrahydro-1H-3-benzazepino)-
carbonyl]] - 6 - [[(1 - adamantanemethyl)amino]carbonyl] -
benzimidazole (14b). A white solid which crystallised
from hexane–ethyl acetate (118 mg, 0.136 mmol, 47%)
m.p. 166–170 °C; 1H-NMR (400 MHz) l 1.36–1.48
(m, 9H), 1.57–1.60 (m, 3H), 1.82 (m, 3H), 2.73–2.98
(m, 3H), 3.02–3.15 (m, 1H), 3.28–3.60 (m, 3H), 3.63–
3.75 (m, 1H), 5.35 (s, 4H), 5.68 (t, 1H, J=6), 7.02–7.22
(m, 5H), 7.32–7.57 (m, 10H), 7.73 (br s, 1H), 7.94 (br
s, 1H), 8.47 (s, 1H), 8.82 (s, 2H), 10.08 (s, 1H), 11.26 (s,
1H); IR (KBr) wmax 3078 m (CꢁH aromatic), 2901 m,
2846 m (CꢁH alkane), 1725 vs (CꢀO ester), 1638 vs
(CꢀO amide), 1547 s (NH bending), 1239 s cm−1
(CꢁO); MS (FAB) m/z 870 (MH+, 13%), 705 (4), 535
(8), 336 (100), 91 (66). Anal. Calc. for C53H52N5O7·1.27
H2O: C, 71.21; H, 6.15; N, 7.83. Found: C, 71.42; H,
6.45; N, 7.78%.
6.1.4. The preparation of compounds 14a,b
6.1.4.1. 5-[[[[3,5-Bis(benzyloxycarbonyl)phenyl]amino]-
carbonyl] - [3 - (1,2,3,4 - tetrahydroisoquinolino)carbonyl]]-
6 - [[(1 - adamantanemethyl)amino]carbonyl]benzimidazole
(14a). Benzimidazole-5,6-dicarboxylic acid anhydride
(11) (29 mg, 0.15 mmol) and compound 10a (80 mg,
0.15 mmol) were suspended in dry acetonitrile (1.5 mL)
and the mixture was heated to reflux. A precipitate
formed within 15 min and the mixture was heated for a
further 3 h. The mixture was allowed to cool and
MeOH was added until all the precipitate was dis-
solved. Pre-adsorption on silica followed by column
6.1.5. The preparation of compounds 14c,d
chromatography
(SiO2;
dichloromethane–MeOH,
6.1.5.1. 5-[[[[3,5-Bis(benzyloxycarbonyl)phenyl]amino]-
carbonyl] - [2 - (1,2,3,4,5,6 - hexahydro - 3 - ben.zazocino)-
carbonyl]] - 6 - [[(1 - adamantanemethyl)amino]carbonyl]-
benzimidazole (14c). The benzimidazole-5,6-dicarboxylic
acid anhydride (11) (67 mg, 0.36 mmol) and 10c (150
mg, 0.27 mmol) were suspended in dry acetonitrile (6
mL) and the mixture was heated to reflux for 3 h.
Column chromatography (SiO2; dichloromethane–
MeOH, 99:1–90:10) gave impure carboxylic acid inter-
mediate compound 12c. EDC (72 mg, 0.38 mmol),
HOBt (51 mg, 0.38 mmol) and DMAP (3–4 mg, 0.03
mmol) were added to the solution of the intermediate
compound 12c dissolved in DMF (6 mL). The mixture
was stirred for 5 min at r.t. before addition of 1-
adamantanemethylamine (13) (50 mg, 0.30 mmol) in
DMF (1 mL) under argon. The resulting solution was
stirred for 17 h at r.t. DMF was evaporated in vacuo
and the resulting oil was dissolved in dichloromethane
(10 mL) and extracted with 2 M hydrochloric acid
solution (2×5 mL), saturated aq. hydrogencarbonate
solution (5 mL), brine (5 mL) and dried (MgSO4).
Column chromatography (SiO2; dichloromethane–
MeOH; 98:2) gave compound 14c as a white solid
which crystallised as white crystals from hexane–ethyl
acetate (128 mg, 0.145 mmol, 53%) m.p. 169–171 °C;
1H-NMR (400 MHz, DMSO-d6) l 1.22–1.60 (m, 14H),
1.71 (br s, 3H), 1.80–1.90 (m, 1H), 1.90–2.00 (m, 1H),
2.61–2.89 (m, 3H), 3.05–3.22 (m, 1H), 4.09–4.13 and
4.41–4.43 (2×br m, 1H), 5.20 (br s, 1H), 5.40–5.74
(m, 4H), 5.50 (br s, 1H), 7.13–7.24 (m, 4H), 7.28–7.46
(m, 10H), 7.89–8.77 (m, 6H), 10.11 (s, 1H), 11.39 (s,
99:1–90:10) gave a pale yellow solid as the impure
carboxylic acid intermediate compound 12a. PyBrOP®
(47 mg, 0.10 mmol), diisopropylethylamine (0.035 mL,
0.20 mmol) and DMAP (1–2 mg, 0.010 mmol) were
added to the solution of the intermediate compound
12a in dichloromethane (1 mL) under argon. The mix-
ture was stirred for 5 min at r.t. under argon before
addition of 1-adamantanemethylamine 13 (17 mg, 0.10
mmol) in dichloromethane (1 mL). The resulting solu-
tion was stirred for 24
h under argon at r.t.
Dichloromethane (10 mL) was added and the organic
layer was washed with 2 M hydrochloric acid solution
(2×5 mL), saturated aq. hydrogencarbonate solution
(5 mL), brine (5 mL) and dried (MgSO4). Column
chromatography (SiO2; dichloromethane–MeOH; 98:2)
gave the title compound 14a as an oil which crystallised
as white crystals from hexane–dichloromethane (60 mg,
0.070 mmol, 45%) m.p. 167–169 °C; 1H-NMR (400
MHz, DMSO-d6) l 1.20–1.33 (m, 6H), 1.47–1.69 (m,
6H), 1.80–1.87 (m, 3H), 2.80–2.84 (m, 1H), 2.95–3.00
(m, 1H), 3.19–3.30 (m, 1H), 3.40–3.48 (m. 1H), 4.42
(d, 1H, J=15.8), 4.49 (d, 1H, J=15.8), 5.30–5.40 (m,
5H), 6.96 (d, 1H, J=7), 7.06 (t, 1H, J=7), 7.11–7.18
(t, 1H, J=7), 7.22 (d, 1H, J=7), 7.32–7.43 (m, 10H),
7.62 (brs 1H), 8.18 (s, 1H), 8.22 (s, 1H), 8.46 (s, 1H),
8.66 (s, 2H), 8.88 (br dt, 1H), 10.04 (s, 1H), 12.87 (br s,
1H); IR (KBr) wmax 3086 m and 3033 m (CꢁH aro-
matic), 2926, 2903 and 2848 m (CꢁH alkane), 1725 vs
(CꢀO ester), 1634 vs (CꢀO amide), 1548 s (NH bend-
ing), 1240 s cm−1 (CꢁO); MS (FAB) m/z 856 (MH+,

S.E. Gibson (ne´e Thomas) et al. / European Journal of Medicinal Chemistry 37 (2002) 379–389
387
1H); IR (KBr) wmax 3078 m, 3029 m (CꢁH aromatic),
2902, 2847 m (CꢁH alkane), 1725 vs (CꢀO ester),
1632 vs (CꢀO amide), 1564 s (NH bending), 1233 s
cm−1 (CꢁO); MS (FAB) m/z 884 (MH+, 6%), 719
(11), 336 (100), 91 (66). Anal. Calc. for C54H54N5O7:
C, 73.27; H, 6.15; N, 7.92. Found: C, 73.51; H, 5.93;
N, 7.76%.
adamantanemethyl)amino]carbonyl]benzimidazole (15b).
A white powder (69 mg, 0.10 mmol, quant.) m.p.
1
249–252 °C (decomp.); H-NMR (400 MHz, DMSO-
d6) l 1.21–1.57 (m, 12H), 1.62–1.84 (m, 3H), 2.90–
3.17 (m, 5H), 3.67–3.72 (m, 3H), 5.50–5.53 (m, 1H),
7.14–7.24 (m, 5H), 7.98–8.83 (m, 6H), 10.05 (s, 1H),
13.07 (br s, 2H); IR (KBr) wmax 3500–2500 br (O–H),
3290 s (NH amide), 2907 s and 2848 m (CꢁH alkane),
1694 s (CꢀO carboxylic acid), 1630 vs and 1600 s
(CꢀO amide), 1547 s cm−1 (NH bending); MS (FAB)
m/z 690 (MH+, 14%), 509 (22), 467 (3), 336 (100).
The bis(N-methyl-D-glucamine) salt was prepared, as
a white solid, by lyophylisation of a solution of 15b
(1 equiv.) and N-methyl-D-glucamine (2 equiv.) in
1,4-dioxan–water. Anal. Calc. for C39H39N5O7·
2C7H17NO5·5.2H2O: C, 54.18; H, 7.24; N, 8.35.
Found: C, 54.09; H, 7.28; N, 8.44%.
6.1.5.2. 5-[[[[3,5-Bis(benzyloxycarbonyl)phenyl]amino]-
carbonyl]-[2-(2,3,4,5,6,7-hexahydro-1H-3-benzazonino)-
carbonyl]] - 6 - [[(1 - adamantanemethyl)amino]carbonyl]-
benzimidazole (14d). A white solid, which crystallised
from hexane–dichloromethane (96 mg, 0.11 mmol,
1
50%) m.p. 168–71 °C; H-NMR (400 MHz, DMSO-
d6) l 0.95–1.97 (m, 21H), 2.78–3.00 (m, 4H), 3.25–
3.28 (m, 1H), 3.70–3.80 (m, 1H), 4.11 (br s, 1H),
5.38–5.74 (m, 4H), 7.19–7.25 (m, 4H), 7.33–7.45 (m,
10H), 7.77–8.74 (m, 6H), 10.10 (s, 1H), 12.86 (d, 1H,
J=17); IR (KBr) wmax 2932 m, 2903 m, 2848 m (CꢁH
alkane), 1726 vs (CꢀO ester), 1629 vs (CꢀO amide),
1552 s (NH bending), 1231 s cm−1 (CꢁO); MS (FAB)
m/z 898 (MH+, 22%), 733 (5), 563 (22), 336 (82), 91
(100). Anal. Calc. for C55H56N5O7: C, 73.46; H, 6.28;
N, 7.79. Found: C, 73.62; H, 6.33; N, 7.56%.
6.1.6.3. 5-[[[3,5-(Dicarboxyphenyl)amino]carbonyl]-[2-
(1,2,3,4,5,6 - hexahydro - 3 - benzazocinocarbonyl]] - 6-[[(1-
adamantanemethyl)amino]carbonyl]benzimidazole (15c).
A white powder (69 mg, 0.098 mmol, 99%) m.p. 248–
1
250 °C (decomp.); H-NMR (400 MHz, DMSO-d6) l
1.21–2.05 (m, 19H), 2.51–3.62 (m, 8H), 4.11–4.14
and 4.46–4.50 (2×m, 1H), 5.22 and 5.54 (2×br s,
1H), 7.19–7.45 (m, 4H), 7.92–8.73 (m, 6H), 11.46 (m,
2H); IR (KBr) wmax 3500–2500 br (OꢁH), 2932 m,
2901 s, 2848 m (CꢁH alkane), 1707 s (CꢀO carboxylic
acid), 1649 vs, 1631 vs (CꢀO amide), 1557 s cm−1
(NH bending); MS (FAB) m/z 704 (MH+, 9%), 523
(28), 467 (3), 336 (100). The bis(N-methyl-D-glu-
camine) salt was prepared, as a white solid, by
lyophylisation of a solution of 15c (1 equiv.) and N-
methyl-D-glucamine (2 equiv.) in 1,4-dioxan–water.
Anal. Calc. for C39H39N5O7·2C7H17NO5·5.2H2O: C,
54.18; H, 7.24; N, 8.35. Found: C, 54.09; H, 7.28; N,
8.44%.
6.1.6. The preparation of compounds 15a–d
6.1.6.1. 5–[[[3,5–(Dicarboxyphenyl)amino]carbonyl]-[2-
(1,2,3,4 - tetrahydroisoquinolino)carbonyl]] - 6 - [[(1 - ad-
amantanemethyl)amino]carbonyl]benzimidazole
(15a).
Palladium-on-charcoal (10%) (25 mg) was added to a
solution of 14a (250 mg, 0.29 mmol) in a 1:1 mixture
of THF and MeOH (9 mL). The reaction mixture
was stirred for 12 h under an atmosphere of hydrogen
and then filtered through Celite and evaporated in
vacuo to afford compound 15a as a white powder
(192 mg, 0.284 mmol, 98%) m.p. 240–242 °C; 1H-
NMR (400 MHz, DMSO-d6) l 1.34–1.85 (m, 15H),
2.90–3.00 (m, 1H), 3.10–3.18 (m, 1H), 3.60–3.70 (m,
2H), 4.41 (d, 1H, J=15.8), 4.48 (d, 1H, J=15.8),
5.36–5.39 (m, 1H), 6.96–7.24 (m, 4H), 7.37–7.70 (m,
2H), 8.10–8.95 (m, 6H), 9.97 (s, 1H), 12.87 (br s,
2H); IR (KBr) wmax 3300–2400 br (OꢁH), 2907 s,
2848 m (CꢁH alkane), 1712 s (CꢀO carboxylic acid),
1632 vs (CꢀO amide), 1553 s cm−1 (NH bending);
MS (FAB) m/z 676 (MH+, 10%), 495 (4), 467 (2),
6.1.6.4. 5-[[[3,5-(Dicarboxyphenyl)amino]carbonyl]-[2-
(2,3,4,5,6,7-hexahydro-1H-3-benzazonino)carbonyl]]-6-
[[(1 - adamantanemethyl)amino]carbonyl]benzimidazole
(15d). A white powder (31 mg, 0.043 mmol, 98%)
m.p. 244–246 °C (decomp.); 1H-NMR (500 MHz,
DMSO-d6) l 1.02–1.87 (m, 21H), 2.78–3.05 (m, 4H),
3.45–4.62 (m, 3H), 7.15–7.30 (m, 4H), 7.98–8.80 (m,
6H), 10.01 (m, 1H), 12.90 (m, 2H); IR (KBr) wmax
3500–2400 br (OꢁH), 2929 m, 2906 s, 2847 m (CꢁH
alkane), 1714 s (CꢀO carboxylic acid), 1638 vs, 1633
vs, 1622s (CꢀO amide), 1554 s cm−1 (NH bending);
MS (FAB) m/z 718 (MH+, 4%), 537 (17), 336 (21),
73 (100). The bis(N-methyl-D-glucamine) salt was pre-
pared, as a white solid, by lyophylisation of a solu-
tion of 15d (1 equiv.) and N-methyl-D-glucamine (2
equiv.) in 1,4-dioxan–water. Anal. Calc. for
C41H43N5O7·2C7H17NO5·5.2H2O: C, 54.92; H, 7.40; N,
8.15. Found: C, 54.89; H, 7.50; N, 8.08%.
336
(100),
55
(99).
Anal.
Calc.
for
C38H37N5O7·1.45H2O: C, 65.03; H, 5.73; N, 9.98.
Found: C, 65.08; H, 5.81; N, 9.96%. The bis(N-
methyl-D-glucamine) salt was prepared, as a white
solid, by lyophylisation of a solution of 15a (1 equiv.)
and N-methyl-D-glucamine (2 equiv.) in 1,4-dioxan–
water.
6.1.6.2. 5-[[[3,5-(Dicarboxyphenyl)amino]carbonyl]-[2-
(2,3,4,5-tetrahydro-1H-3-benzazepino)carbonyl]]-6-[[(1-

388
S.E. Gibson (ne´e Thomas) et al. / European Journal of Medicinal Chemistry 37 (2002) 379–389
6.1.7. The preparation of compounds 15e
(M+K) 770.2956. Found: 770.2962. The bis(N-methyl-
D-glucamine) salt was prepared, as a white solid, by
lyophylisation of a solution of 15e (1 equiv.) and N-
methyl-D-glucamine (2 equiv.) in 1,4-dioxan–water.
Anal. Calc. for C42H45N5O7·2C7H17NO5·0.5H2O: C,
59.46; H, 7.22; N, 8.67. Found: C, 59.39; H, 6.96; N,
8.46%.
6.1.7.1. 5-[[[[3,5-Bis(benzyloxycarbonyl)phenyl]amino]-
carbonyl]-[2-(1,2,3,4,5,6,7,8-octahydro-3-benzazecino)-
carbonyl]]-6-(carboxylic acid)benzimidazole (12e). Com-
pounds 10e (129 mg, 0.22 mmol) and 11 (51 mg, 0.27
mmol) were suspended in dry acetonitrile (2.2 mL, 0.1
M) and heated to reflux. After 15 min, a yellow precip-
itate began to form. After 5 h the mixture was allowed
to cool and the pale yellow precipitate filtered off.
Purification by flash column chromatography (SiO2;
dichloromethane–methanol, 9:1) afforded the title com-
pound 12e as a white powder (124 mg, 0.162 mmol,
Acknowledgements
The authors would like to thank E.A. Harper for
providing the data from the radioligand binding assays
presented in Table 1.
1
72%) m.p. 201–203 °C; H-NMR (300 MHz, DMSO-
d6) l 0.84–1.22 (m, 5H), 1.50–1.90 (m, 2H), 2.53–3.56
(m, 9H), 5.38–5.40 (m, 4H), 7.18–7.50 (m, 14H), 8.01–
8.54 (m, 4H), 8.62–8.79 (m, 2H); MS (FAB) m/z 765
(MH+, 12%), 154 (100), 136 (83), 91 (80). Anal. Calc.
for C45H40N4O8: C, 70.67; H, 5.27; N, 7.32. Found: C,
70.33; H, 5.63; N, 7.22%.
References
[1] (a) C. Toniolo, Int. J. Peptide Protein Res. 35 (1990) 287–300;
(b) A. Giannis, T. Kolter, Angew. Chem., Int. Ed. Engl. 32
(1993) 1244–1267;
6.1.7.2.
5-[[[3,5-(dicarboxyphenyl)amino]carbonyl]-[2-
(c) V.J. Hruby, G. Li, C. Haskell-Luevano, M. Shenderovich,
Biopolymer 43 (1997) 219–266;
(1,2,3,4,5,6,7,8 - octahydro - 3 - benzazecino)carbonyl]] - 6-
[[(1 - adamantanemethyl)amino]carbonyl]benzimidazole
(15e). Diisopropylethylamine (0.042 mL, 0.24 mmol),
PyBroP® (67 mg, 0.14 mmol) and DMAP (2–3 mg,
:15 mol%) were added to a stirred solution 12e (92
mg, 0.12 mmol) in dichloromethane (0.6 mL). After 10
min stirring, a solution of 1-admantanemethylamine 13
(24 mg, 0.14 mmol) in dichloromethane (0.6 mL, 0.1 M
in total) was added and the reaction mixture stirred for
18 h at r.t. under argon. Dichloromethane (8 mL) was
added and the solution washed successively with aq. 2
M hydrochloric acid (2×10 mL), saturated aq. sodium
hydrogencarbonate solution (2×10 mL) and brine (10
mL). The combined organic extracts were dried
(MgSO4) and evaporated in vacuo to afford an off-
white solid. Purification by flash column chromatogra-
phy (SiO2; dichloromethane–methanol, 6:1) afforded a
mixture of the desired coupling product and PyBroP®
residues (0.114 g). The crude mixture was dissolved in a
1:1 mixture of THF and methanol (4.0 mL) and 10%
palladium-on-carbon catalyst (12 mg) added. The reac-
tion mixture was put under an inert atmosphere and
then filled with hydrogen via balloon. The mixture was
stirred at r.t. under a hydrogen atmosphere for 20 h.
Filtration through Kieselguhr (MeOH) and evapora-
tion of the solvents in vacuo afforded a white solid (108
mg). Purification by flash column chromatography
(SiO2; dichloromethane–MeOH–acetic acid, 40:4:1) af-
forded compound 15e as a white powder (63 mg, 0.086
mmol, 72%) m.p. 278–280 °C (decomp.) 1H-NMR
(300 MHz, DMSO-d6) l 1.01–3.68 (m, 35H), 7.11–7.52
(m, 4H), 7.97–8.67 (m, 5H), 9.78–9.84 (m, 1H); MS
(FAB) m/z 732 (MH+, 10%), 551 (22), 495 (15), 336
(48), 185 (100), HRMS (FAB) Calc. for C42H45N5O7K
(d) S.E. Gibson (ne´e Thomas), N. Guillo, M.J. Tozer, Tetrahe-
dron 55 (1999) 585–615.
[2] J.Y.L. Chung, J.T. Wasicak, W.A. Arnold, C.S. May, A.M.
Nadzan, M.W. Holladay, J. Org. Chem. 55 (1990) 270–275.
[3] K. Hsieh, T.R. LaHann, R.C. Speth, J. Med. Chem. 32 (1989)
898–903.
[4] W.M. Kazmierski, Z. Urbanczyk-Likowska, V. Hruby, J. Org.
Chem. 59 (1994) 1789–1795.
[5] B.Y. Azizeh, M.D. Shenderovich, D. Trivedi, G. Li, N.S. Sturm,
V.J. Hruby, J. Med. Chem. 39 (1996) 2449–2455.
[6] W. Kazmierski, W.S. Wire, G.K. Lui, R.J. Knapp, J.E. Shook,
T.F. Burks, H.I. Yamamura, V.J. Hruby, J. Med. Chem. 31
(1988) 2170–2177.
[7] P.W. Schiller, T.M.-D. Nguyen, G. Weltrowska, B.C. Wilkes,
B.J. Marsden, C. Lemieux, N.N. Chung, Proc. Natl. Acad. Sci.
USA 89 (1992) 11871–11875.
[8] S. Salvadori, G. Balboni, R. Guerrini, R. Tomatis, C. Bianchi,
S.D. Bryant, P.S. Cooper, L.H. Lazarus, J. Med. Chem. 40
(1997) 3100–3108.
[9] (a) J.W. Skiles, J.T. Suh, B.E. Williams, P.R. Menard, J.N.
Barton, B. Loev, H. Hones, E.S. Neiss, A. Schwah, W.S. Mann,
A. Khandwala, P.S. Wolf, I. Weinryb, J. Med. Chem. 29 (1986)
784–796;
(b) U. Scho¨llkopf, R. Hinrichs, R. Lonsky, Angew. Chem., Int.
Ed. Engl. 26 (1987) 143–145.
[10] H.G. Chen, O.P. Goel, Synth. Commun. 25 (1995) 49–56.
[11] C. Wang, H.I. Mosberg, Tetrahedron Lett. 36 (1995) 3623–3626.
[12] S.E. Gibson (ne´e Thomas), N. Guillo, R.J. Middleton, A. Thuil-
liez, M.J. Tozer, Chem. Soc., Perkin Trans. 1 (1997) 447–455.
[13] S.B. Kalindjian, I.M. Buck, J.M.R. Davies, D.J. Dunstone, M.L.
Hudson, C.M.R. Low, I.M. McDonald, M.J. Pether, K.I.M.
Steel, M.J. Tozer, J.G. Vinter, J. Med. Chem. 39 (1996) 1806–
1815.
[14] S.E. Gibson (ne´e Thomas), N. Guillo, S.B. Kalindjian, M.J.
Tozer, Bioorg. Med. Chem. Lett. 7 (1997) 1289–1292.
[15] N.J. Welsh, N.P. Shankley, J.W. Black, Br. J. Pharmacol. 112
(1994) 93–96.
[16] E.A. Harper, S.P. Roberts, N.P. Shankley, J.W. Black, Br. J.
Pharmacol. 118 (1996) 1717–1726.

S.E. Gibson (ne´e Thomas) et al. / European Journal of Medicinal Chemistry 37 (2002) 379–389
389
[17] S.P. Roberts, E.A. Harper, G.F. Watt, V.P. Gerskowitch,
R.A.D. Hull, N.P. Shankley, J.W. Black, Br. J. Pharmacol. 118
(1996) 1779–1789.
[22] D.C. Horwell, J. Hughes, J.C. Hunter, M.C. Pritchard, R.S.
Richardson, E. Roberts, G.N. Woodruff, J. Med. Chem. 34
(1991) 404–414.
[18] R.A.D. Hull, N.P. Shankley, E.A. Harper, V.P. Gerskowitch,
J.W. Black, Br. J. Pharmacol. 108 (1993) 734–740.
[19] N.P. Shankley, S.P. Roberts, G.F. Watt, A. Kotecha, R.A.D.
Hull, J.W. Black, Pharmacologist 39 (1997) 32 Abstract no. 65.
[20] J.W. Black, N.P. Shankley, Br. J. Pharmacol. 86 (1985) 571–
579.
[23] J. Anders, M. Blu¨ggel, H.E. Meyer, R. Ku¨hne, A.M. ter Laak,
E. Kojro, F. Fahrenholz, Biochemistry 38 (1999) 6043–6055.
[24] James Black Foundation Patent. WO Patent Application 95/
04720 (1995).
[25] V.J. Hruby, W.L. Cody, A.M. de Lauro Castrucci, M.E. Hadley,
Coll. Czech. Chem. Commun. 53 (1988) 2549–2573.
[26] D.M. Hills, V.P. Gerskowitch, S.P. Roberts, N.J. Welsh, N.P.
Shankley, J.W. Black, Br. J. Pharmacol. 119 (1996) 1401–1410.
[21] S. Wallenstein, C.L. Zucker, J.L. Fleiss, Circ. Res. 47 (1980)
1–9.