Synthesis and evaluation of novel peripherally restricted κ-opioid receptor agonists

Article abstract of DOI:10.1016/j.bmcl.2004.12.018

A series of 3-substituted analogs (3) of the parent κ agonist, 1, were prepared to limit access to the central nervous system. With the exception of compound 3j, all other compounds bound to the human κ opioid receptor with high affinity (K_i =

Full text of DOI:10.1016/j.bmcl.2004.12.018

Bioorganic & Medicinal Chemistry Letters 15 (2005) 1091–1095
Synthesis and evaluation of novel peripherally
restricted j-opioid receptor agonists
Virendra Kumar,* Deqi Guo, Joel A. Cassel, Jeffrey D. Daubert, Robert N. DeHaven,
Diane L. DeHaven-Hudkins, Erin K. Gauntner, Susan L. Gottshall, Susan L. Greiner,
Michael Koblish, Patrick J. Little, Erik Mansson and Alan L. Maycock
Adolor Corporation, 700 Pennsylvania Drive, Exton, PA 19341, USA
Received 27 October 2004; revised 3 December 2004; accepted 8 December 2004
Available online 7 January 2005
Abstract—A series of 3-substituted analogs (3) of the parent j agonist, 1, were prepared to limit access to the central nervous system.
With the exception of compound 3j, all other compounds bound to the human j opioid receptor with high affinity (Ki = 0.31–
9.5 nM) and were selective for j over l and d opioid receptors. Compounds 3c, d, and 3g–i produced potent antinociceptive activity
in the rat formalin assay (i.paw) and the mouse acetic acid-induced writhing assay (s.c.), with weak activity in the mouse platform
sedation test. The peripheral restriction indices of 3c, d, 3g, and 3i were improved 2-to 7-fold compared to the parent compound 1,
and these compounds were approximately 2-to 5f-old more potent than the peripheral j agonist ICI 204448.
ꢀ 2004 Elsevier Ltd. All rights reserved.
Although centrally active kappa (j) opioid receptor
agonists have analgesic properties, they have been
of limited therapeutic use because of centrally mediated
side effects such as sedation, dysphoria, and diuresis.1–3
Recent evidence indicates that opioid antinociception
can also be mediated by activation of opioid receptors
located outside the central nervous system (CNS).4,5
Our goal was to modify centrally active j agonists to
reduce their penetration into the CNS, thus minimizing
or eliminating side effects. Previously, we reported 2-(4-
trifluoromethylphenyl)-N-methyl-N-[1-phenyl-2-(1-pyr-
rolidinyl)ethyl]-acetamide (1)6 as a highly potent agonist
that was peripherally restricted compared to the cen-
trally acting j agonist ICI 199441 (2).7 To further
improve the peripheral restriction of 1, we synthesized
a series of analogs of general structure 3 with modifica-
tions on the central phenyl ring and evaluated them for
in vitro binding affinity, in vivo antinociceptive activity,
and peripheral restriction.
a wide variety of substituents without detrimental effects
on j opioid receptor affinity. The substituents at the 3-
position were selected from a group, which may have
ionizable potential at physiological pH, thus limiting
the CNS penetration exemplified by ICI 204448, a re-
ported opioid receptor agonist with reduced access to
the CNS.8
R
R
F3C
O
O
N
N
R'
N
N
CH3
CH3
1 (R = CF3; R' = H)
2 ( R = R' = Cl, ICI 199441)
3
OCH2CO2H
N
Cl
Cl
O
N
CH3
ICI 204448
Modifications of the central phenyl ring at the 3-posi-
tion were chosen since there are literature precedents2,3
indicating that this position in the molecule can tolerate
The 3-oxygenated analogs of the central phenyl ring of 1
were prepared as racemates as shown in Scheme 1. Syn-
thesis of compounds (R,S)-3(a–d) required the diamine
(R,S)-6, which was prepared according to the literature
published by the ICI group.9 Compound (R,S)-3b was
Keywords: j-opioid receptor agonists; Peripherally restricted.
Corresponding author. Tel.: +1 484 595 1060; fax: +1 484 595
1573; e-mail: vkumar@adolor.com
*
0960-894X/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.12.018

1092
V. Kumar et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1091–1095
NO2
OCH3
OCH3
d,e
66%
OCH3
a,b,c
NO2
N
f
35%
O
N
OH
a
N
HN
CH3
+
N
H3CO
N
H
H
O
HN
CH3
HN
CH3
HN
CH3
82%
O
(R,S)-5
4
(R,S)-6
7
9 (minor)
8
OCH3
OH
N
NO2
N
NH2
F3C
F3C
g
O
F3C
F3C
O
N
CH3
N
O
O
b
c
N
CH3
N
N
CH3
N
CH3
(R,S)-3b
(R,S)-3a
R
N
3e
3f
R
h,i
F3C
O
F3C
CH3SO2Cl
Pyridine
O
O
N
3g: R = NHSO2CH3
3h: R = NHSO2NH2
N
R
N
CH3
CH3
(R,S)-3c: R =OCH2CO2C(CH3)3
(R,S)-3d: R =OCH2CO2H
F3C
NH2SO2NH2
Scheme 1. Reagents and conditions: (a) NaCN, NH4Cl; (b) HCl (6 N),
refluxed; (c) methyl chloroformate, NaOH (1 N); (d) HOBt/DCC,
pyrrolidine; (e) LAH, THF; (f) 4-trifluoromethylphenyl acetic acid,
HOBt/DCC, CH2Cl2; (g) BBr3, CH2Cl2, ꢀ70 ꢁC; (h) t-butyl bromoace-
tate, K2CO3, DMF; (i) HCl (6 N), 95–100 ꢁC.
N
Dioxane
reflux
N
CH3
R
N
O
O
O
F3C
O
3i: R = NHCO(CH2)2CO2H
CH2Cl2
N
CH3
prepared by coupling of the diamine with 4-trifluoro-
methylphenyl acetic acid in the presence of HOBt/
DCC. Demethylation of methoxy ether (R,S)-3b was
accomplished with BBr3, which gave the phenol (R,S)-
3a. Compounds (R,S)-3c and d were obtained by alkyl-
ation of (R,S)-3a with t-butyl bromoacetate followed by
acid hydrolysis.
R
F3C
BrCH2CO2t-Bu
NaH/THF
O
N
3j: R = N(CH2CO2H)
N
CH3
Scheme 2. Reagents and conditions: (a) HNO3/H2SO4; (b) 4-trifluo-
romethylphenylacetic acid, DCC/HOBt; (c) Raney Ni, hydrazine,
EtOH, 50 ꢁC.
The corresponding 3-amino analogs (Scheme 2) were
prepared by following the procedure described by
Chang et al.10 The nitration of diamine 77 resulted in
the 3-nitro compound (8) as a major product contami-
nated with 4-nitro compound (9). These compounds
were not separated at this stage and were coupled with
4-trifluoromethylphenyl acetic acid to give a mixture
of enantiomers and regioisomers. The desired isomer
3e was purified by chromatography and recrystalliza-
tion. Compound 3e was then reduced with Raney-nick-
el/hydrazine hydrate, which resulted in compound 3f.
Reaction of 3f with various electrophiles (methane sulfo-
nyl chloride, sulfamide, succinic anhydride, and t-butyl
bromoacetate) gave compounds 3g–j (Table 1).
and 3e–h, were in the subnanomolar range (Ki = 0.31–
0.63 nM). In general, smaller attachments at the 3-posi-
tion of the parent compound, either with ether or amino
groups, resulted in compounds with high affinity for the
j opioid receptor. On the other hand, j opioid receptor
selectivity was reduced for compound (R,S)-3a and 3j
with respect to l and d opioid receptors. The zwitter-
ionic characteristics of compounds containing carboxyl
groups, (R,S)-3d and 3i, also resulted in approximately
one order of magnitude lower affinity for the j opioid
receptor, with Ki values of 3.7 and 9.5 nM, respectively.
However, the dicarboxylic derivative, 3j, exhibited a
considerable reduction in j opioid receptor binding
affinity (Ki = 2300 nM). This reduction in affinity indi-
cates the intolerance of the j opioid receptor for com-
pounds containing bis-ionizable groups in the central
phenyl ring.
Receptor binding: Kappa (j), mu (l), and delta (d) opi-
oid receptor affinities were determined by displacement
of bound [3H]diprenorphine from membranes prepared
from cells expressing cloned human opioid receptors.
Details of the methodology have been previously re-
ported in the literature.11,12 Kappa selectivity over l
and d opioid receptors was defined by the ratio of the
Ki values.
When comparing the j opioid receptor affinity of newly
synthesized analogs with the parent compound 1, it was
evident that the present compounds were at least 5-fold
less potent in the j opioid receptor binding assay, with
varying degrees of selectivity for l and d opioid recep-
tors. Nonetheless, compounds (R,S)-3a, b, and 3e–h
were more potent than ICI 204448.8 Three other com-
pounds, (R,S)-3c, d, and 3i, had affinities for the j
Opioid receptor binding affinities of newly synthesized
compounds are summarized in Table 1. With the excep-
tion of compound 3j, all other compounds bound to the
human j opioid receptor with high affinity and good
selectivity for j compared to l or d opioid receptors.
The Ki values for most of the compounds, (R,S)-3a,b,

V. Kumar et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1091–1095
Table 1. In vitro binding to cloned human opioid receptors
1093
R
F3C
O
N
N
CH3
3
Compound
R
Yield (%)a
Ki (nM) for
jb[3H]-diprenorphine
Selectivity for
j over l (fold)
Selectivity for
j over d (fold)
1
H
—
—
28
65
63
62
94
48
45
39
74
54
0.058 (0.046–0.073)
2.69 (1.58–5.01)
0.44 (0.28–0.67)
0.53 (0.3–0.93)
2.9 (1.7–4.8)
> 1000
162
68
> 1000
69
230
ICI 204448
OH
OCH3
(R,S)-3a
(R,S)-3b
(R,S)-3c
(R,S)-3d
(S)-3e
870
330
350
920
110
420
310
57
> 1000
79
190
OCH2CO2C(CH3)3
OCH2CO2H
NO2
3.7 (2.5–5.6)
0.63 (0.34–1.2)
0.31 (0.22–0.44)
0.36 (0.25–0.5)
0.48 (0.39–0.59)
9.5 (8.4–11.0)
> 1000
130
(S)-3f
(S)-3g
NH2
NHSO2CH3
NHSO2NH2
NHCO(CH2)2CO2H
N(CH2CO2H)2
> 1000
> 1000
11
(S)-3h
(S)-3i
(S)-3j
2300 (1700.0–3100.0)
> 43
> 43
a The compounds were isolated as hydrochloride or methane sulfonic acid salts and characterized by MS, NMR, mp, and elemental analyses.
b Values are geometric means of three to ten experiments and 95% confidence intervals are given in parentheses.
opioid receptor that were comparable to ICI 204448
(Table 1). In a direct comparison of ICI 204448 with
compound (R,S)-3d, the latter compound had compara-
ble j opioid receptor affinity, with improved selectivity
for l and d opioid receptors.
(Table 2), with no observable CNS or minimum side
effects, were advanced for further evaluation in the
mouse acetic acid-induced writhing test. Compound 3a
showed significant sedation in the formalin test; an attri-
bute of centrally active j opioid agonists and no addi-
tional testing was warranted.
Formalin-induced flinching:13 The formalin test was used
as an initial in vivo screen for the newly synthesized
compounds. Due to the potential toxicity of compounds
3e (nitro) and 3f (aniline), they were excluded from the
initial screen and the data for the rest of the compounds
are presented in Table 2. Generally, 3-oxygen substi-
tuted compounds (R,S)-3a–d showed analgesic activity
in this test comparable to compound 1 and the 3-nitro-
gen substituted compounds (3g–i) were slightly less
active. The compounds that inhibited flinching by
approximately 70% or higher at 300 lg given i.paw
Acetic acid-induced writhing:14 All the tested compounds
exhibited potent (ED50 = 0.15–1.7 mg/kg s.c., Table 2)
analgesic activity in the acetic acid-induced writhing test
in mice. Compound (R,S)-3a was not tested due to
observed side effects in the rat formalin test. The ED50
values of the t-butyl ester (R,S)-3c and the correspond-
ing acid (R,S)-3d were essentially equal to ICI 204448
in this assay. The aniline substituted compounds (3g,
h, and i) were slightly more active than the correspond-
ing ethers with the exception of compound (R,S)-3b,
Table 2. In vivo profile of j-opioid agonists
R
N
F3C
O
N
CH3
3
Compounda
R
Formalin flinching %A
@ 300 lg, i.paw
Writhing ED50
(mg/kg), s.c.
Platform sedation
ED50 (mg/kg), s.c.
Peripheral
restriction index
1
H
92
—
99
99
84
93
77
73
73
0.005
1.6
0.13
52
25
ICI 204448
OH
OCH3
32.5
—
(R,S)-3a
(R,S)-3b
(R,S)-3c
(R,S)-3d
(S)-3g
ND
0.15
1.7
ND
ND
140
> 300
89
—
81
OCH2CO2C(CH3)3
OCH2CO2H
NHSO2CH3
NHSO2NH2
NHCO(CH2)2CO2H
1.7
> 180
160
18
0.56
0.74
0.87
(S)-3h
(S)-3i
13
47
54
a Compounds were administered in 20% aqueous creamaphor solution.

1094
V. Kumar et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1091–1095
Pyke, R. E.; Greiner, M.; Wideman, G. L.; Benjamin, R.;
Pierce, M. W. Clin. Neuropharmacol. 1996, 19, 451.
2. (a) Barber, A.; Gottschlich, R. Exp. Opin. Invest. Drugs
1997, 6, 1351; (b) Dhawan, B. N.; Cesselin, F.; Raghubir,
R.; Reisine, T.; Bradley, P. B.; Portoghese, P. S.; Hamon,
M. Pharmacol. Rev. 1996, 48, 567.
3. (a) Szmuszkovicz, J. In Progress in Drug Research; Jucker,
E., Ed.; Birkhauser Verlag: Basel, 1999; Vol. 52, pp 167–
195; (b) Szmuszkovicz, J. In Progress in Drug Research;
Jucker, E., Ed.; Birkhauser Verlag: Basel, 1999; Vol. 53,
pp 1–51.
which was the most active of the 3-substituted com-
pounds. Even though all of the new compounds
showed potent analgesic activity in the writhing test,
the potency of these compounds was substantially
lower than the parent compound 1 (ED50 = 0.005 mg/
kg). The greater potency of the parent compound in
the mouse writhing assay may be in fact due to contribu-
tions from central j opioid receptors as well as periphe-
ral receptors.
Platform sedation:15 The peripheral selectivity indices
were determined only for compounds with minimum ob-
servable side effects in the mouse writhing test. Thus,
compounds (R,S)-3a and b were not tested in the plat-
form sedation assay in mice. The peripheral index was
calculated for the rest of the compounds in Table 2.
The parent compound 1 was very potent in the mouse
platform sedation assay with an ED50 of 0.13 mg/kg
(s.c.). Its peripheral restriction index was 25, due to its
very potent activity in the mouse writhing test. All other
substituted compounds, with the exception of com-
pound 3h, had a peripheral restriction index approxi-
mately 2-to 7f-old higher than the parent compound
1, and in some cases 2-to 5-fold better than the known
peripheral compound ICI 204448 (Table 2). Two com-
pounds, (R,S)-3d and 3g, were the most promising leads,
with peripheral indices of >180 and 160, respectively.
The high peripheral indices of these two compounds
could be attributed to the ionizability characteristics at
physiological pH of the substituents at the 3-position
of (R,S)-3d (CH2CO2Oꢀ) and 3g (ꢀNSO2CH3), thus
limiting access to the CNS. The compounds are being
further evaluated in a number of assays for development
as potential antinociceptive agents.
4. Zhang, Q.; Schafer, M.; Elde, R.; Stein, C. Neuroscience
1998, 85, 281.
5. Review of peripherally acting opioid agonists as novel
analgesic agents and additional references, see: DeHaven-
Hudkins, D. L.; Dolle, R. E. Curr. Pharm. Des. 2004, 10,
743.
6. Kumar, V.; Marella, M. A.; Cortes Burgos, L.; Chang, A.
C.; Cassel, J. A.; Daubert, J. D.; DeHaven, R. N.;
DeHaven-Hudkins, D. L.; Gottshall, S. L.; Mansson, E.;
Maycock, A. L. Bioorg. Med. Chem. Lett. 2000, 10, 2567,
and references therein.
7. Costello, G. F.; James, R.; Shaw, J. S.; Slater, A. M.;
Stutchbury, N. C. J. J. Med. Chem. 1991, 34, 181.
8. Shaw, J. S.; Carroll, J. A.; Alcock, P.; Main, B. G. Br. J.
Pharmacol. 1989, 96, 98.
9. Barlow, J. J.; Blackburn, T. P.; Costello, G. F.; James, R.;
Le Count, D. J.; Main, B. G.; Pearce, P. J.; Russell, K.;
Shaw, J. S. J. Med. Chem. 1991, 34, 3149.
10. Chang, A.-C.; Takemori, A. E.; Ojala, W. H.; Gleason, W.
B.; Portoghese, P. S. J. Med. Chem. 1994, 37, 4490.
11. DeHaven-Hudkins, D. L.; Cortes Burgos, L.; Cassel, J.
A.; Daubert, J. D.; DeHaven, R. N.; Mansson, E.;
Nagasaka, H.; Yu, G.; Yaksh, T. J. Pharmacol. Exp.
Ther. 1999, 289, 494.
12. DeHaven, R. N.; DeHaven-Hudkins, D. L. In Current
Protocols in Pharmacology; Enna, S. J., Williams, M.,
Ferkany, J. W., Kenakin, T., Porsolt, R. D., Sullivan, J.
P., Eds.; John Wiley and Sons: New York, 1998; pp 1.4.1–
1.4.12.
13. Wheeler-Aceto, H.; Cowan, A. Psychopharmacol 1991,
104, 35. Compounds were evaluated in vivo for their
abilities to inhibit formalin-induced flinching in rats. The
flinching behavior was quantified by counting the number
of responses that occurred during the tonic (late) phase
of pain, lasting from 20–35 min after formalin injection.
The j agonists were injected i.paw 10 min prior to the
formalin injection. The mean percent antagonism of
formalin-induced flinching for drug-treated, formalin-
injected rats was calculated according to the following
formula:
All new compounds with the exception of 3j potently
bound to the j receptor in vitro, with Ki values in the
nanomolar range. The new 3-substituted compounds,
except for (R,S)-3a, 3i, and 3j, were over 100-fold selec-
tive for j over l. The selectivity for j over d was slightly
higher than that for j over l except for compounds
(R,S)-3c, 3d, 3i, and 3j. By substituting in the central
phenyl ring of parent compound 1 with ether or amino
containing groups, the degree of peripheral restriction
of these j agonists was improved approximately 2-to
7-fold. The carboxymethyl ether group (compound,
(R,S)-3d) and methanesulfonamido group (compound,
3g) offered the most improvement for the peripheral
restriction index and represented at least a 5-fold
improvement over ICI 204448.
ðMean formalin response ꢀ Mean saline responseÞ
ꢀIndividual response
ꢁ 100
Mean formalin response ꢀ Mean saline response
The mean formalin response was the mean number of
flinches in rats treated with vehicle prior to formalin injec-
tion. The mean saline response was the mean number of
flinches in rats treated with 50 lL of saline. The data are
presented in Table 2 and compared with compound 1
and ICI 204448.
Overall, 3-substituted compounds were less potent in vi-
tro and in vivo compared to the parent compound 1.
However, the objective of this work was achieved in
finding at least two compounds with an improvement
in the peripheral index compared to compound 1, as a
measure of limiting access to the CNS.
14. Collier, H. O. J.; Dinneen, L. C.; Johnson, C. A.;
Schneider, C. Br. J. Pharmac. Chemother. 1968, 32, 295.
Male ICR mice (20–25 g) were injected s.c. (back of the
neck) with either vehicle or drug. Fifteen minutes later,
mice were injected i.p. with 0.6% acetic acid. Five minutes
after treatment with acetic acid, the numbers of writhes
were counted for 10 min. Dose–response curves were
References and notes
1. (a) Dionne, R. A.; Dobbins, K. R.; Hargreaves, K. M.
Clin. Pharmacol. Ther. 1991, 49, 183; (b) Pande, A. C.;

V. Kumar et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1091–1095
1095
expressed as the percent inhibition of acetic acid-induced
writhing, when compared to the mean number of writhes
observed in vehicle-treated mice.
with either vehicle or drug and 20 min later tested for the
latency to step off the platform. A 30 s cut-off was utilized.
The percent sedation was calculated by the following
formula:
15. Horan, P.; de Costa, B. R.; Rice, K. C.; Porreca, F. J.
Pharmacol. Exp. Ther. 1991, 257, 1154
Sedation was measured by the latency of a mouse to
Test latency ꢀ Baseline latency
ꢁ 100
completely step off
a
slightly raised platform
ð30 s ꢀ Baseline latencyÞ
(11 · 8 · 3 cm). Prior to treatment, mice were tested once
for baseline latencies. Mice with baseline latencies >15 s
were not used. Mice were injected s.c. (back of the neck)
Peripheral restriction index = platform sedation ED50/
writhing ED50.