Synthesis and activity of 4,5-diarylimidazoles as human CB1 receptor inverse agonists

Article abstract of DOI:10.1016/j.bmcl.2004.12.078

Structure-activity relationship studies directed toward the optimization of 4,5-diarylimidazole-2-carboxamide analogs as human CB1 receptor inverse agonists resulted in the discovery of the two amide derivatives 24a and b (hCB1 IC₅₀ = 6.1 and 4

Full text of DOI:10.1016/j.bmcl.2004.12.078

Bioorganic & Medicinal Chemistry Letters 15 (2005) 1441–1446
Synthesis and activity of 4,5-diarylimidazoles as human
CB1 receptor inverse agonists
Christopher W. Plummer,a,* Paul E. Finke,a Sander G. Mills,a Junying Wang,a
Xinchun Tong,a George A. Doss,b Tung M. Fong,c Julie Z. Lao,c Marie-Therese Schaeffer,c
Jing Chen,c Chun-Pyn Shen,c D. Sloan Stribling,d Lauren P. Shearman,d
Alison M. Strackd and Lex H. T. Van der Ploegc
aDepartment of Medicinal Chemistry, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
bDepartment of Drug Metabolism, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
cDepartment of Metabolic Disorders, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
dDepartment of Pharmacology, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
Received 10 November 2004; revised 23 December 2004; accepted 30 December 2004
Available online 22 January 2005
Abstract—Structure–activity relationship studies directed toward the optimization of 4,5-diarylimidazole-2-carboxamide analogs as
human CB1 receptor inverse agonists resulted in the discovery of the two amide derivatives 24a and b (hCB1 IC50 = 6.1 and 4.0 nM)
which also demonstrated efficacy in overnight feeding studies in the rat for reduction in both food intake and overall body weight.
Ó 2005 Elsevier Ltd. All rights reserved.
1. Introduction
R'
N
NH
Excess energy intake leading to obesity has emerged as a
NH
O
N
O
major health threat in industrialized nations, increasing
the chances for obese individuals of developing type 2
diabetes, hypertension, heart disease, stroke, cancer,
and arthritis.1 With the percentage of the population
considered to be obese increasing annually, scientific
advancements in the understanding of obesity have pro-
vided opportunities to seek therapeutic solutions for
addressing this issue. The cannabinoid receptor type 1
(CB1) and its endogenous ligands, the endocannabi-
noids,2 have been found to be involved in the control
of weight and energy balance via a dual mechanism of
food intake modification and the regulation of energy
expenditure.3 SR-141716 (Fig. 1, Rimonabant, 1) was
reported to have potent human CB1 receptor affinity4
and was later demonstrated with feeding studies in the
rat to afford a dose-dependent reduction in both food in-
take and body weight.5,6 CB1 receptors are coupled to
the Gi/o class of G-proteins, which inhibit adenylate
cyclase activity.7 CB1 agonists, such as CP-55940,8
S
O
Me
R
N
N
N
HN
N
Cl
Cl
Y
X
Cl
2a
1 (SR-141716)
hCB1 IC50 = 6 nM
2b
hCB1 IC50 = 7000 nM
Figure 1. Structure and binding affinity of SR-141176 and the Merck
lead structure. For the imidazole structure 2b, see Tables 1 and 2.
decrease forskolin-induced cAMP accumulation in
tissue from rodent and mammalian brain in a concen-
tration-dependent manner. However, 1 produced a
dose-dependent increase in forskolin-induced cAMP,9
thus suggesting that inverse agonism of the CB1
receptor was responsible for the observed effects.10
Our search for novel anti-obesity agents based on the
endocannabinoid hypothesis was initiated with a high
throughput screening of the Merck sample collection
Keywords: Imidazole; Obesity; CB1 inverse agonist; Cannabinoid.
*
Corresponding author. Tel.: +1 732 594 5299; fax: +1 732 594
5790; e-mail: christopher_plummer@merck.com
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.12.078

1442
C. W. Plummer et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1441–1446
to identify possible human CB1 (hCB1) receptor antag-
onist leads. The 4,5-diarylimidazole 2a (Fig. 1), having
an analogous, but novel, scaffold from 1, was found to
have moderate affinity for the hCB1 receptor (hCB1
IC50 = 7000 nM) and inspired the investigation of the
corresponding imidazole structure 2b. Since modeling
studies reported for 1 implied a critical role for the N-
2 pyrazole nitrogen,11 it was hoped that the more basic
imidazole nitrogen might enhance this interaction. The
benzyl moiety was thought to take the place of the
N-piperidine amide. The moderate affinity of 2a was
attributed to the lack of an amide interaction and no
substitution on the phenyls. As reported herein, our ini-
tial efforts investigated various phenyl substitutions in
conjunction with the introduction of the requisite C-2
amide moiety. The utilization of alternative core hetero-
cycles, such as a 4,5-diphenylthiazole, 1,5-diphenyltriaz-
ole, and an isomeric 1,2-diphenylimidazole, was very
recently reported.12 In addition, the incorporation of a
5,6-diphenylpyridine with a six-membered core hetero-
cyclic ring has also been reported from these laborato-
ries13 and by researchers at Sanofi-Synthelabo.14
amide products 7a–c and 8a–c (see Table 1) with Py-
BOP and the requisite amines. Variable amounts of
the C-2 unsubstituted imidazoles were observed during
purification of the amides due to competing decarboxyl-
ation of the acids under the reaction conditions.
An alternative synthesis (Scheme 2) of the correspond-
ing 4,5-bis-(4-chlorophenyl)imidazole derivatives 14a–f
was developed to avoid anticipated complications with
the lithium exchange reaction and allowed for the prep-
aration of the N-1 unsubstituted derivatives 15a,b (see
Table 1). Benzoin condensation of 4-chlorobenzalde-
hyde (9) afforded 4,40-dichlorobenzoin (10). Cyclization
in formamide and paraformaldehyde afforded the N-1
and C-2 unsubstituted imidazole 11 as well as the minor
oxazole by-product 12.18 Alkylation of 11 with methyl
iodide or SEM chloride in DMF using sodium hydride
as base gave the N-Me or N-SEM-protected intermedi-
ates 13a,b. Selective deprotonation with n-butyllithium
at ꢀ70 °C again afforded the C-2 anions which could
be quenched with benzyl chloroformate as in Scheme 1
or, when available, with the appropriate isocyanates to
afford the desired amide products 14b,e,f directly.
Hydrolysis of the intermediate benzyl ester and conver-
sion to amides as above and/or final removal of the SEM
protecting group with TBAF generated the N-methyl or
N–H imidazole derivatives 14a,c,d and 15a,b. Alterna-
tively, again due to variable amounts of decarboxylation
of the intermediate C-2 acids, reaction with ethyl chloro-
formate and direct amide formation was used for the
2. Chemistry15
Preparation of the 4,5-bis-(phenyl) and (4-methyl-
phenyl)imidazole derivatives 7 and 8 (Scheme 1) started
with the thermal cyclization of benzoin (3a) or 4,40-di-
methylbenzoin (3b) with N-methylurea in ethylene gly-
col at 180 °C to give the N-methyl imidazolones 4a
and b.16,17 Treatment with phosphorous oxychloride
provided the 2-chloroimidazoles 5a,b, which were con-
verted to the benzyl esters 6a,b via n-butyllithium ex-
change and quenching of the C-2 anions with benzyl
chloroformate. Catalytic hydrogenation afforded the
crude acids which were directly converted to the desired
Table 1. Structures and hCB1 affinities for imidazole derivatives
7,8,14,15
R1
O
N
N
R2
R
N
X1
O
Cl
Me
Me
N
NH
N
N
X1
X1 ꢀNR1R2
X1
a
b
O
(SD)b
OH
Compd #
R
hCB1
IC50
(nM)a
X1
X1
X1
X1
1
7a
6.2
4a X1 = H
5a X1 = H
X1
Me
Me
Me
H
H
H
ꢀNH(N-piperidinyl)
ꢀNH(cyclohexyl)
ꢀ(piperidinyl)
1900
1800
(400)
(60)
4b X1 = Me
5b X1 = Me
3a X1 = H
3b X1 = Me
7b
7c
R1
N
19,000 (800)
O
N
O
N
O
N
R2
8a
8b
Me Me ꢀNH(N-piperidinyl)
Me Me ꢀNH(cyclohexyl)
Me Me ꢀ(piperidinyl)
460
900
(340)c
(180)
(1700)
(60)
Bn
Me
Me
N
R
1 and R2;
see Table 1
c
d, e
8c
3400
260
14a
14b
14c
14d
14e
14f
15a
15b
Me Cl
Me Cl
Me Cl
Me Cl
Me Cl
Me Cl
ꢀNH(N-piperidinyl)
ꢀNH(cyclohexyl)
ꢀ(piperidinyl)
450
1100
(100)
(400)
(700)
(140)
(140)
(400)
(40)
X1
X1
ꢀNH(N-morpholinyl) 3100
X1
X1
6a X1 = H
6b X1 = Me
7 X1 = H
8 X1 = Me
ꢀNH(n-hexyl)
540
900
ꢀNH(t-butyl)
H
H
Cl
Cl
ꢀNH(N-piperidinyl)
ꢀNH(cyclohexyl)
6000
1500
Scheme 1. Reactants and conditions: (a) N-Methylurea, ethylene
glycol, 180 °C, 1.5 h (40–60%, from EtOH); (b) POCl3, 100 °C, 16 h
(70–85%); (c) n-BuLi, THF, ꢀ70 to ꢀ20 °C, 1.5 h, then added to
BnOCOCl, THF, ꢀ20 °C to rt (25–42%); (d) H2 (50 p.s.i), 10% Pd/C,
MeOH (quant.); (e) HNR1R2, Py-BOP, DCM, rt (35–80%).
a See Ref. 21 for assay details.
b n = 2, except as noted.
c n = 3.

C. W. Plummer et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1441–1446
1443
H
Cl
Cl
Cl
O
O
OH
N
N
O
N
O
H
O
a
b
+
+
OH Cl
OH Cl
OH
O
H
O
H
Cl
Cl
Cl
10
19
a
Cl
+
Cl
j
Cl
Cl
Cl
Cl
R
Cl
Cl
10
Cl
Cl
11
12
Cl
O
2 : 1
O
9
Cl
Cl
+
R1
N
OH
N
9
18
N
N
Cl
O
N
d, e, f
R1 and R2;
see Table 1
R2
c
or
11
R
20
N
g
(R =
or
SEM)
Cl
Cl
h, i
14 R = Me
15 R = H
HN
N
NH
O
N
13a R = Me
Cl
Cl
X2
b
Cl
X3
Cl
13b R = SEM
Cl
Cl
20
+
R1
Cl
R
O
N
Cl
Cl
Cl
O
O
O
N
R2
Et
21
R1 and R2;
see Table 3
O
N
22a X2 = H, X3 = Cl
22b X2 = Cl, X3 = H
h
i
12
R1
N
O
N
R2
Cl
Cl
Cl
Cl
R
N
N
N
N
17
16
R
Cl
Cl
Cl
Cl
N
Scheme 2
X2
c
X3
21
18
Scheme 2. Reactants and conditions: (a) NaCN, H2O:EtOH/3:1,
100 °C, 4 h; (b) paraformaldehyde, formamide, 200–220 °C (65%); (c)
MeI or SEM-Cl, NaH, DMF, rt (80–96%); (d) n-BuLi, THF, ꢀ70 °C;
then added to BnOCOCl, THF, ꢀ70 °C to rt (32–56%); (e) 5 N NaOH,
MeOH, rt, 16 h (quant.); (f) HNR1R2, Py-BOP, DCM, rt (14–50%); (g)
n-BuLi, THF, ꢀ70 °C; then added to R1–NCO, ꢀ70 °C to rt (24–77%);
(h) n-BuLi, THF, ꢀ70 °C; then added to EtOCOCl, THF, ꢀ70 °C to rt
(65–85%); (i) HNR1R2, dioxane, 100 °C (30–80%); (j) 1 N TBAF,
THF, rt (15–60%).
+
Cl
Cl
Cl
Cl
23a
19
23b
R1 and R2
see
Table 2
24 R = Me, X2 = H, X3 = Cl
25 R = Et, X2 = H, X3 = Cl
26 R = Me. X2 = Cl, X3 = H
Scheme 2
a
27 R = Me, X2 = X3 = Cl
conversion of the oxazole 12 to the amides 17a,b (see
Table 3) via the ethyl ester 16.
R1
N
O
R2
R1 and R2
see
22a
or
Scheme 2
O
N
While the structure–activity relationships (SAR) con-
ducted thus far in the above symmetrical series indicated
a 2- to 10-fold improvement in hCB1 binding potency
with the 4-methyl or 4-chloro substitution on each phen-
yl ring, the binding was still weak compared to 1. Opti-
mization of the 4-phenyl with 2,4-dichloro substitution
as in 1 proved to be critical in establishing good receptor
affinity.12,13b The preparation of the required unsym-
metrical trichlorobenzoins and conversion to the two
possible trichloro final products is shown in Scheme 3.
In addition, the oxazole and tetrachloro derivatives were
also prepared. Reaction of a 2:1 mixture of 4-chloro-
benzaldehyde (9) and 2,4-dichlorobenzaldehyde (18) in
the benzoin condensation afforded a separable (flash sil-
ica gel chromatography, 10% DCM in 10–25% ethyl
acetate/hexanes) mixture of dichloro 10 (lowest Rf),
the two trichloro isomers 20 (middle Rf), and tetrachloro
19 (trace, highest Rf) products. The mixture of trichloro-
benzoin isomers was then cyclized as above to the imid-
azole 21 (two indistinguishable tautomers) and the two
separable oxazole by-products 22a,b. Alkylation of 21
as above with either methyl or ethyl iodide gave the sep-
arable N-1 alkyl intermediates 23a,b. The final product
amides 24–26 were then obtained by direct reaction of
the C-2 anions with the requisite isocyanate or via trans-
amidation of the ethyl ester intermediates. The two tri-
chloro isomers were assigned based on NOE effects
Table 3
X2
X3
22b
28 X2 = H, X3 = Cl
29 X2 = Cl, X3 = H
Cl
Cl
Scheme 3. Reactants and conditions: (a) NaCN, H2O:EtOH/3:1,
100 °C, 4 h; (b) paraformaldehyde, formamide, 200–220 °C; (c) MeI
or EtI, NaH, DMF, rt (80–90%).
observed between the N-Me and the 4-phenyl ortho
hydrogens.19 The tetrachlorobenzoin 19, prepared in
an independent benzoin condensation of 18 alone, was
carried forward to the amide products 27a,b (Table 2).
Similarly, the oxazoles 22a (higher Rf isomer) and 22b
(lower Rf isomer) were converted to amide products 28
and 29 (Table 3), respectively. The structural assign-
ments for 28 and 29 were determined by HMBC
NMR experiments.19
3. Results and discussion
Binding affinities on recombinant human CB1 receptors
expressed in Chinese Hamster Ovary (CHO) cells for
compounds in Tables 1–3 were determined with a
standard binding assay using [3H]CP-55940 as the

1444
C. W. Plummer et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1441–1446
Table 2. Structures and hCB1 affinities for imidazole derivatives 24–27
phenyl) compounds 8a-b (IC50 = 460 and 900 nM) and
14a,b (IC50 = 260 and 400 nM) seemed to validate the
hypothesis that the imidazole could replace the pyrazole
of 1 and that the trichloro derivative analogous to 1
should be made. In addition, the cyclohexylamine amide
derivatives 8b and 14b were within a factor of 2 in po-
tency to the N-piperidines, indicating that there might
be some SAR potential for other amides with the imid-
azole scaffold (see 14d–f and below). Also encouraging
was the finding that the hCB1 functional assay indicated
that both 14a (up to ꢀ150% response relative to the ago-
nist CP-55940 with an EC50 = 100 40 nM, n = 2) and
14b (average ꢀ75% relative response with an
EC50 = 500 200 nM, n = 2) were interacting with the
hCB1 receptor as inverse agonists as desired. Impor-
tantly, 14a was found to be P10-fold selective for
hCB1 over hCB2 (IC50 > 2000 nM (40%)).22 The dimin-
ished hCB1 affinity for the N-1 unsubstituted com-
pounds 15a and 15b (IC50 = 6000 and 1500 nM) was
surprising and indicated the importance of the methyl,
possibly in order to optimally direct the neighboring
amide moiety for binding to the hCB1 receptor or to re-
move an unfavorable polar moiety from a hydrophilic
receptor area. The secondary amides 7c, 8c and 14c
(IC50 = 19,000, 4500 and 1100 nM) were also compara-
tively inactive, while the benzyl ester 6b (see Scheme 1;
IC50 = 850 nM) was within 2-fold of the amide 8a, thus
indicating that the amide N–H was not critical for bind-
ing affinity.
R1
O
N
N
R2
R
N
X2
X3
Cl
X2 X3 ꢀNR1R2
Cl
Compd #
R
hCB1 (SD)b
IC50
(nM)a
24a
24b
24d
24g
24h
24i
Me
Me
Me
Me
Me
Me
Me
Et
H
H
H
H
H
H
H
H
Cl ꢀNH(N-piperidinyl)
Cl ꢀNH(cyclohexyl)
6.1 (2.8)c
4.0 (2.0)c
Cl ꢀNH(N-morpholinyl) 170
(55)
(3)
Cl ꢀNH(cyclopentyl)
Cl ꢀNH(cycloheptyl)
Cl ꢀNMe(cyclohexyl)
Cl ꢀNH(C6H5)
17
8.0 (1.7)
70
60
(40)
(3)
24j
25b
26a
26b
27a
27b
Cl ꢀNH(cyclohexyl)
4.1 (2.6)d
Me Cl
Me Cl
H
H
ꢀNH(N-piperidinyl)
ꢀNH(cyclohexyl)
190
(80)
(14)
(8)
46
16
Me Cl Cl ꢀNH(N-piperidinyl)
Me Cl Cl ꢀNH(cyclohexyl)
5.2 (0.4)d
a See Ref. 21 for assay details.
b n = 2, except as noted.
c n = 8.
d n = 4.
Table 3. Structures and hCB1 affinities for oxazole derivatives
The tetrachloro N-piperidine derivative 27a displayed a
significant 15-fold improvement in binding affinity com-
pared to the dichloro 14a (IC50 = 16 vs 250 nM) while
the cyclohexyl analog 27b was even better with a 90-fold
enhancement from 14b (IC50 = 5.2 vs 450 nM). How-
ever, it was the unsymmetrical trichloro derivatives
24a and 24b, with substitution analogous to 1, which ap-
peared to be optimal (IC50 = 6.1 and 4.0 nM). The bind-
ing affinities for the isomeric trichloro compounds 26a,b
were intermediate between the dichloro and tetrachloro
results. The hCB1 functional assay again indicated that
both 24a (average ꢀ120% relative response with an
EC50 = 5 2 nM (n = 5)) and 24b (average ꢀ110% rela-
tive response with an EC50 = 18 4 nM (n = 4)) were
interacting with the hCB1 receptor as inverse agonists.
These compounds were also found to be 60-fold selec-
tive for hCB1 over hCB2 (hCB2 IC50 = 350 and
250 nM, respectively), in the binding assays.22 Based
on the results of 24a,b, the cyclopentyl (24g,
IC50 = 17 nM) and cycloheptyl (24h, IC50 = 8 nM), N-
methyl-N-cyclohexyl (24i, IC50 = 75 nM), and aniline
(24j, IC50 = 60 nM) analogs were prepared, but none ap-
peared more potent and the hope of an expanded SAR
with the imidazole scaffold was not realized. The surpris-
ingly poor results for the N-morpholines 14d and 24d
(IC50 = 3100 and 175 nM) indicated that polarity in
the amide alkyl region was not well tolerated. Finally,
the N-1 ethyl analog 25b was found to be comparable
to 24b (IC50 = 4.1 vs 4.0 nM), thus hinting at the possi-
ble use of this site for further modification. These results
were in good agreement with the amide SARs reported
in the literature for the pyrazoles and other core
heterocycles.12
R1
N
O
O
R2
N
X2
X3
Cl
Cl
Compd #
X2
X3
ꢀNR1R2
hCB1
IC50
(SD)b
(nM)a
17a
17b
28a
28b
29a
29b
30a
30b
H
H
H
H
ꢀNH(N-piperidinyl)
ꢀNH(cyclohexyl)
ꢀNH(N-piperidinyl)
ꢀNH(cyclohexyl)
ꢀNH(N-piperidinyl)
ꢀNH(cyclohexyl)
ꢀNH(N-piperidinyl)
ꢀNH(cyclohexyl)
2200
500
260
80
(700)
(300)
(50)
(3)
H
Cl
Cl
H
H
Cl
Cl
Cl
Cl
470
1000
200
70
(50)
(500)
(30)
(50)
H
Cl
Cl
a See Ref. 21 for assay details.
b n = 2.
radio-ligand.20,21 Interesting compounds were further
evaluated in a functional assay also using recombinant
human CB1 receptor expressed in CHO cells in the pres-
ence of forskolin.21 Binding affinities were also routinely
measured for the CB2 receptor expressed in CHO cells
and using [3H]WIN-55,212-2 as radioligand.21,22 While
the initial results with the unsubstituted 4,5-phenyl
derivatives 7a,b were disappointing (IC50 = 1900 and
1800 nM), the 2- to 10-fold improvement in the binding
activity for the 4,5-bis-(4-methylphenyl and 4-chloro-

C. W. Plummer et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1441–1446
1445
The binding affinity for the dichloro oxazole analogs
20
18
16
14
12
Vehicle
17a,b (Table 3, IC50 = 2200 and 500 nM) appeared to
mimic the poorer affinity of the imidazole N–H com-
pounds. The same decrease in affinity was seen for the
two trichloro isomers 28a,b (IC50 = 260 and 80 nM)
compared to 24a,b. While the tetrachloro compounds
30a,b (IC50 = 200 and 70 nM) were about equivalent to
28a,b, the overall affinity for these oxazoles was still only
moderate compared to the imidazoles. Thus, the
oxazoles were not investigated further.
1 mg/kg
3 mg/kg
10 mg/kg
- 13.4%
- 30.6%
10
8
- 67.5%
6
4
2
0
As noted above, the hCB2 binding affinities were also
routinely assessed.22 Fortunately, in general the hCB2
SAR was not as sensitive to the substitution pattern of
the phenyls and, thus, greater selectivity was realized
as the hCB1 affinity was enhanced. Initially, all the com-
pounds of Table 1 were P2000 nM. While the hCB2
affinity was also seen to be enhanced with the additional
2-chlorophenyl substitution, the actual hCB1 to hCB2
ratios were improved. Interestingly, the hCB2 SAR
appeared to be more sensitive to the N-3 substitution
(hCB2 IC50 = 22% at 2000, 300, and 165 for 15b, 24b
and 25b, respectively).
0:00 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00
Time Post-Treatment (hr) with Cpd 24a
Figure 2. Acute treatment of male diet-induced obese rats (n = 4 at
each dose) with CB1 inverse agonist 24a reduced cumulative dark cycle
food intake. Vehicle (10% Tween 80 in water) or 24a was administered
to rats 1 h prior to the onset of the dark cycle. Percent reduction in
cumulative food intake relative to vehicle control is shown.
ratio = 2.99–3.40 between 0.25 and 4 h).24 In a food
intake and body weight loss study using diet-induced
obese rats fed ad libitum over 18 h (Fig. 2), 24a showed
a dose-dependent, immediate and prolonged reduction
in food intake as reported for 16 (24a @ 1, 3, and
10 mg/kg po = ꢀ13.4%, ꢀ30.6%, ꢀ67.5% relative to
vehicle treated animals). Also, at 18 h post-dosing, the
10 mg/kg dose afforded a cumulative, statistically-signif-
icant (p < 0.0001), dose-dependent weight loss of 13 g
compared to a 6 g gain for vehicle treated animals.
These results were in agreement with good systemic
exposure, which produced the weight loss component
of CB1 inverse agonism, and favorable CNS exposure,
as evident from the B/P ratio experiment and the ob-
served immediate food intake reduction.6 A feeding
study with the slightly more potent cyclohexyl amide
24b afforded a similar, dose-dependent reduction in food
intake (24b @ 1, 3, and 10 mg/kg po = ꢀ25%, ꢀ39%,
ꢀ53%), and a significant (p = 0.013) net weight loss of
2 g versus a 10 g gain for vehicle treated animals. The re-
sults with 24b were consistent with a slightly poorer PK
profile (AUCnom = 1.82 lMhkg/mg; Clp = 19.8 mL/
min/kg; Vdss = 1.9 L/kg; t1/2 = 1.9 h), lower oral absorp-
tion (F = 30%), and slower brain penetration (B/P
ratio = 0.84–2.61 between 0.25 and 4 h).
4. In vivo studies
Based on their favorable hCB1 binding and functional
characteristics, 24a and 24b were selected for further in
vivo evaluation (Table 4). Preliminary pharmacokinetic
(PK) studies23 (1 mg/kg, iv, 2 mg/kg po) for 24a indi-
cated good iv properties (AUCnom = 1.32lMhkg/mg;
Clp = 27.5 mL/min/kg; Vdss = 3.7 L/kg; t1/2 = 2.4 h) and
oral absorption (F = 50%), as well as rapid brain
penetration and a high brain-to-plasma ratio (B/P
Table 4. Pharmacokinetic parameters, B/P ratios, and rat feeding
study results for 24a and 24b
Compd #
24a
24b
Rat PKa
AUCnom (lMhkg/mg) 1.3 0.2
1.8
20
Clp (mL/min/kg)
Vdss (L/kg)
t1/2 (h)
28
3
3.7 0.1
2.4 0.6
1.9
1.9
%F
50
7
30 11
B/P ratiob
@ 0.25 h
@ 1.0 h
@ 2.0 h
@ 4.0 h
3.0 0.6
3.6 1.0
2.9 0.2
3.4 1.0
0.8 0.4
2.1 0.6
1.3 0.1
2.6 0.4
5. Conclusions
Feeding studyc
Herein, we have described the synthesis and hCB1 activ-
ities for a series of 4,5-diphenyl-1-methylimidazole-2-
carboxamides related to the pyrazole SR-141716 (1).
The phenyl substitution SAR was found to be the same
as 1, namely, the 4-(2,4-dichlorophenyl)-5-(4-chlorophen-
yl) analogs 24a and 24b appeared to be optimal,
although the tetrachloro derivatives 27a and 27b were
only slightly less potent. However, the amide SAR was
more interesting in that both the N-piperidinyl and
cyclohexyl carboxamides were found to be potent
hCB1 inverse agonists. Both of these compounds were
Food intake Vehicle
@ 1 mg/kg
@ 3 mg/kg
17 g/—
20 g/—
15 g/ꢀ13.4% 15 g/ꢀ25%
12 g/ꢀ30.6% 12 g/ꢀ39%
5.4 g/ꢀ67.5% 9.4 g/ꢀ53%
6 g
2 g
@ 10 mg/kg
Weight D
Vehicle
10 g
4 g
@ 1 mg/kg
@ 3 mg/kg
@ 10 mg/kg
ꢀ1 g
ꢀ13 g
0
ꢀ2 g
a See Ref. 23 for details.
b See Ref. 24 for details.
c See Ref. 6 and Figure 2 for details.

1446
C. W. Plummer et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1441–1446
Calandra, B.; Pecceu, F.; Lupker, J.; Maffrand, J. P.; Le
Fur, G.; Casellas, P. J. Biol. Chem. 1997, 272, 22330.
11. Shim, J.-Y.; Welsh, W. J.; Cartier, E.; Edwards, J. L.;
Howlett, A. C. J. Med. Chem. 2002, 45, 1447.
12. Lange, J. H. M.; van Stuivenberg, H. H.; Coolen, H. K. A.
C.; Adolfs, T. J. P.; McCreary, A. C.; Keizer, H. G.; Wals,
H. C.; Veerman, W.; Borst, A. J. M.; de Looff, W.;
Verveer, P. C.; Kruse, C. G. J. Med. Chem. 2004; web
release.
13. (a) Finke, P. E.; Meurer, L. C.; Debenham, J. S.;
Toupence, R. B.; Walsh, T. F. WO Patent 03/082191
A2, 2003; Chem. Abstr. 2003, 139, 307686f; (b) Meurer, L.
C.; Finke, P. E.; Mills, S. G.; Walsh, T. F.; Toupence, R.
B.; Debenham, J. S.; Goulet, M. T.; Wang, J.; Tong, X.;
Fong, T. M.; Lao, J.; Schaeffer, M.-T.; Chen, J.; Shen,
C.-P.; Stribling, D. S.; Shearman, L. P.; Strack, A. M.;
Van der Ploeg, L. H. T. Bioorg. Med. Chem. Lett. 2005,
15, 645.
14. Barth, F.; Martinez, S.; Rinaldi-Carmona, M. WO Patent
03/084930 A1, 2003; Chem. Abstr. 2003, 139, 307689j.
15. 1H NMR and/or MS data for all new compounds were in
agreement with the assigned structures. For full experi-
mental and representative NMR spectral data, see: Finke,
P. E.; Mills, S. G.; Plummer, C. W.; Shah, S. K.; Truong,
T. Q. WO Patent 03/007887-A2, 2003; Chem. Abstr. 2003,
138, 122647e.
orally bioavailable with good pharmacokinetics. These
compounds were also shown to be efficacious in obese
rat feeding studies, showing reductions in both food in-
take and overall body weight. In addition, the N-ethyl
derivative 25b was shown to be equally potent. These
findings might allow some flexibility in the selection of
derivatives in the future with enhanced PK and CNS
exposure profiles with improved physical and/or off-
target properties. Further SAR studies of both the C-2
amide and N-1 alkyl group will be published elsewhere
in the near future.
Acknowledgements
The authors are grateful to Quang Truong for the prep-
aration of additional 24b for the feeding study.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmcl.
2004.12.078.
16. A small amount of unsubstituted N–H imidazole (<10%)
was usually formed as a by-product from disproportion-
ation of the N-methyl urea to urea during the reaction and
was removed in subsequent purifications.
References and notes
17. Kim, Y. B.; Kim, C. S.; Lee, C. K. J. Heterocyclic Chem.
1994, 31, 1653.
18. Bredereck, H.; Gompper, R.; Rempfer, H.; Klemm, K.;
Keck, H. Chem. Ber. 1959, 92, 329.
19. See the supplementary data for the NOE NMR spectral
data for the structural assignment of 24b and 26b and the
HMBC NMR experiment for the assignment of 28b and
29b.
20. The reported binding results ( SD) were an average of 2–8
independent determinations (each done in replicate) which
were normally within 50% of the average value based on a
standard sample. For both the binding and functional
assay conditions, see Ref. 21.
1. (a) Flegal, K. M.; Carroll, M. D.; Ogden, C. L.; Johnson,
C. L. J. Am. Med. Assoc. 2002, 288, 1723; (b) Yanovski, S.
Z.; Yanovski, J. A. New Engl. J. Med. 2002, 346, 591.
2. For general reviews of cannabinoid receptor ligands and
potential therapeutic applications, see: (a) Di Carlo, G.;
Izzo, A. A. Exp. Opin. Invest. Drugs 2003, 12, 39; (b)
Pertwee, R. G. Exp. Opin. Invest. Drugs 2000, 9, 1553; (c)
Xiang, J.-N.; Lee, J. C. Annu. Rep. Med. Chem. 1999, 34,
199.
3. Trillou, R. C.; Arnone, C.; Gonalons, N.; Keane, P.;
Maffrand, J. P.; Soubrie, P. J. Physiol. Regul. Integr.
Comp. Physiol. 2003, 284, 345.
4. Rinaldi-Carmona, M.; Barth, F.; Heaulme, M.; Shire, D.;
Calandra, B.; Congy, C.; Martinez, S.; Maruani, J.;
Neliat, G.; Caput, D.; Ferrarra, P.; Soubrie, P.; Breliere,
J. C.; Le Fur, G. FEBS Lett. 1994, 350, 240.
5. Barth, F.; Rinaldi-Carmona, M. Curr. Med. Chem. 1999,
6, 745.
21. Felder, C. C.; Joyce, K. E.; Briley, E. M.; Mansouri, J.;
Mackie, K.; Blond, O.; Lai, Y.; Ma, A. L.; Mitchell, R. L.
Mol. Pharmacol. 1995, 48, 443.
22. See the supplementary data for a complete listing of CB2
binding data.
23. Male Sprague–Dawley rats were dosed at 1.0 mg/kg iv via
an implanted cannula (n = 1 (24b) or 2 (24a)) and at
2.0 mg/kg po by gavage (n = 3). Compounds were formu-
lated at 1 mg/mL in TWEEN 80:EtOH:water (10:23:67)
for both the iv and po doses. Plasma concentrations were
6. Colombo, G.; Agabio, R.; Diaz, G.; Lobina, C.; Reali, R.;
Gessa, G. L. Life Sci. 1998, 8, 113.
7. Pertwee, R. G. Pharmacol. Ther. 1997, 74, 129.
8. Koe, B. K.; Milne, G. M.; Weissman, A.; Johnson, M. R.;
Melvin, L. S. Eur. J. Pharmacol. 1985, 109, 201.
9. Mato, S.; Pazos, A.; Valdizan, E. M. Eur. J. Pharmacol.
2002, 443, 43.
10. (a) Landsman, R. S.; Burkey, T. H.; Consroe, P.; Roeske,
W. R.; Yamamura, H. I. Eur. J. Pharmacol. 1997, 334R1–
334R2; (b) Bouaboula, M.; Perrachon, S.; Milligan, L.;
Canat, X.; Rinaldi-Carmona, M.; Portier, M.; Barth, F.;
determined
precipitation.
by
LC–MS/MS
following
protein
24. Male Sprague-Dawley rats were dosed at 1.0 mg/kg iv
(n = 3 at each time point) via the tail vein. Compounds
were formulated at 1 mg/mL in TWEEN 80:EtOH:water
(10:23:67). Plasma and brain concentrations were deter-
mined by LC–MS/MS.