B. Le Bourdonnec et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2006–2012
2007
ing group of 31a, followed by reductive aminations
using previously reported strategy10 provided the
resin-bound secondary amine intermediates 33. The
N-phenyl derivative 34 was obtained from 31a
according to the method described by Combs and
collaborators.11 The secondary amine derivatives 35
were obtained from 31a using a solid-phase variant
of the Fukuyama–Mitsunobu process.12–14 Despite
many attempts, conducted using a wide range of
coupling reagents and reaction conditions, the cou-
pling of the carboxylic acid 30 with resin-bound
amines 33–35 failed. As an alternative strategy, cou-
pling of resins 33–35 with acryloyl chloride in the
presence of triethylamine provided the resin-bound
acrylamide derivatives 34 which reacted with (+)-
4(R)-(3-hydroxyphenyl)-3(R),4-dimethyl-1-piperidine
(379) to provide the corresponding resin-bound con-
jugate addition products. Cleavage of the resulting
resins using trifluoroacetic acid provided derivatives
7–15, further purified by HPLC. As shown in Table
1, the best compound in this series (compound 15)
displayed good l binding affinity, potent l antago-
nist activity (IC50 = 6.8 nM), comparable to the l
antagonist activity of naloxone (IC50 = 7.3 nM), but
OH
OH
N
O
R1
N
O
OH
OH
N
N
R2
H
O
O
1 (alvimopan)
I
OH
N
O
R1
OH
N
H
O
II
a
receptors.
low selectivity profile for
l versus d and j
As shown in Table 1, replacement of the benzyl
group of 1 with a hydrogen atom (compound 2) re-
sulted in an almost complete loss of l binding. This
result supports the existence of an important lipo-
philic binding region for the l receptor in the prox-
imity of the piperidine nitrogen-binding site.7 Based
on this information, we have introduced various
lipophilic substituents at the Ca position of the gly-
cyl moiety of 2 (compounds 3–6) and examined the
resultant effects on l opioid receptor binding affinity.
The synthesis of compounds 3–6 is outlined in
Scheme 1. Deprotection of Wang resin bound Fmoc
protected a-amino acids 31 with piperidine/DMF fol-
lowed by coupling with the carboxylic acid 30 (ob-
tained by basic hydrolysis of the methyl ester 299),
and subsequent cleavage of the resin using trifluoro-
acetic acid provided the derivatives 3–6, further puri-
fied by HPLC. As shown in Table 1, introduction of
a benzyl group at the Ca position of the glycyl moi-
ety of 2 (compound 3) resulted in a marked increase
in l opioid receptor binding. Hence, compound 3,
The second part of the study was to investigate the
SAR at the Ca position (R1 substituent of formula
II) of the glycyl moiety of 1. The target compounds
16–27 were prepared according to Scheme 2. Deprotec-
tion of Wang resin bound Fmoc protected a-amino
acids 31 with piperidine/DMF followed by coupling
with the carboxylic acid 38,9 and subsequent cleavage
of the resin using trifluoroacetic acid provided the
derivatives 16–27, further purified by HPLC. As shown
in Table 2, introduction of a methyl group at the R1
position (compound 16) was well tolerated. Compound
16 and its diastereoisomer analog 17 had similar bind-
ing affinity at the l opioid receptor. Introduction of
lipophilic moieties of various size and flexibility at the
R1 position provided ligands (compounds 18–23) with
good l binding affinity and antagonist activity. How-
ever, these modifications to the structure of 1 also led
to an increase of the affinity of the ligands toward the
d opioid receptor. For example, compound 21 binds
with equipotent affinity to l and d opioid receptors
(Ki of 2.0 and 3.2 nM, respectively). In the functional
assay, compound 21 was found to be a potent l and
d15 opioid receptor antagonist (IC50 values of 6.2
and 7.8 nM, respectively). Interestingly, introduction
at the R1 position, of alkyl chains containing polar sub-
stituents, had an important effect on opioid receptor
selectivity. Indeed, the carboxymethylene and carboxy-
ethylene derivatives (compounds 24 and 25, respec-
which bound to the
l receptor with a Ki of
36 nM, also displayed potent l in vitro antagonist
activity (IC50 = 7.8 nM). Replacement of the benzyl
group of 3 with a cyclohexylmethyl moiety (com-
pound 4) resulted in a slight increase in l binding,
but also led to a decreased selectivity for l versus
d receptors. Furthermore, introduction of a biphe-
nylmethyl or phenethyl moieties (compounds 5 and
6, respectively) in place of the benzyl group of 3
did not result in an improvement of the binding
affinity toward the l opioid receptor. We then pre-
pared analogs of 3, incorporating at the R2 position
(see formula I), various lipophilic moieties (com-
pounds 7–15). The synthesis of compounds 7–15 is
showed in Scheme 1. Removal of the Fmoc protect-
tively) exhibited good
l binding affinity and l
antagonist activity while displaying greater selectivity
versus the j receptors (>1000-fold) when compared to
1 (210-fold). Furthermore, introduction of an amino-
propylene or aminobutylene moieties at the R1 position
provided ligands with l subnanomolar binding affinity,
potent l antagonist activity and excellent opioid recep-
tor selectivity. In particular, compound 27 with binding