Lead optimization of 5,6-diarylpyridines as CB1 receptor inverse agonists

Article abstract of DOI:10.1016/j.bmcl.2007.01.005

Optimization of the biological activity for 5,6-diarylpyridines as CB1 receptor inverse agonists is described. Food intake and pharmacokinetic evaluation of 3f and 15c indicate that these compounds are effective orally active modulators of CB1.

Full text of DOI:10.1016/j.bmcl.2007.01.005

Bioorganic & Medicinal Chemistry Letters 17 (2007) 2031–2035
Lead optimization of 5,6-diarylpyridines as CB1
receptor inverse agonists
Christina B. Madsen-Duggan,a,* John S. Debenham,a Thomas F. Walsh,a
Richard B. Toupence,a Song X. Huang,a Junying Wang,a Xinchun Tong,a Julie Lao,b
Tung M. Fong,b Marie-Therese Schaeffer,b Jing Chen Xiao,b Cathy R.-R. C. Huang,b
Chun-Pyn Shen,b D. Sloan Stribling,c Lauren P. Shearman,c Alison M. Strack,c
D. Euan MacIntyre,c Lex H. T. Van der Ploegd and Mark T. Gouletd
aDepartment of Medicinal Chemistry, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
bDepartment of Metabolic Disorders, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
cDepartment of Pharmacology, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
dMerck Research Laboratories, 33 Avenue Louis Pasteur, Boston, MA 02115, USA
Received 11 December 2006; revised 4 January 2007; accepted 5 January 2007
Available online 13 January 2007
Abstract—Optimization of the biological activity for 5,6-diarylpyridines as CB1 receptor inverse agonists is described. Food intake
and pharmacokinetic evaluation of 3f and 15c indicate that these compounds are effective orally active modulators of CB1.
Ó 2007 Elsevier Ltd. All rights reserved.
The discovery of the cannabinoid receptor 1, CB1,1 (ex-
pressed predominantly in the central nervous system)
and subsequent elucidation of the CB1 endogenous
ligands, the endocannabinoids, has led to the recogni-
tion that the cannabinoid system has an important role
in food intake modulation.2 Food intake suppression
mediated by CB1 inverse agonists has been demonstrat-
ed in animal3 and human studies4 and has now been
approved as a treatment for obesity.5
Figure 1. Structure of Merck early leads.
In our previous report we disclosed the structure–activ-
(2,4-dichlorophenyl)pyridine containing derivatives such
as 1a (CB1, IC50 = 1.3 nM). We also recognized that the
bis-dichloro derivative 1b had respectable potency with
the benefit of reduced molecular weight (CB1,
IC50 = 8 nM). Adjacent chlorophenyl groups on a het-
erocyclic scaffold are a common element of CB1 modu-
lators, and both bis- and tris-chlorophenyl containing
ligands have been reported from these and other labora-
tories.2 Consequently, we adopted these two aryl chlori-
nation patterns and focused our SAR studies on the
pyridine 2 and 3-position substituents.
ity relationship (SAR) efforts that led to the identifica-
tion of diarylpyridine 1a (Fig. 1), with a 1 nM CB1
binding affinity.6 Unfortunately, the in vitro potency
of 1a only translated into modest in vivo efficacy, due
most likely to less than optimal PK and CNS exposure.6
Further alteration of structure 1a led to compounds
with improved in vivo efficacy and are described within.
Excellent CB1 binding affinity in the 2-benzyloxypyri-
dine series was observed with 5-(4-chlorophenyl)-6-
In order to probe the SAR of the pyridine 2-position,
2-chloropyridines 2a,7 and related 2b (Scheme 1), were
utilized due to their synthetic versatility.
Keywords: Cannabinoid; CB1; Inverse agonist; Pyridine; Obesity.
*
Corresponding author. E-mail: christina_madsen_duggan@merck.
com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.01.005

2032
C. B. Madsen-Duggan et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2031–2035
Scheme 1. Reagents and conditions: (a) ArOH, ArSH or ROH, Cs2CO3, toluene, 100 °C, 50–94%; (b) NHR2R3, THF or toluene, 40–100 °C,
23–89%; (c) RCONH2, Cs2CO3, DMF, 100 °C, 47–50%; (d) alkyne, NEt3, CuI, PdCl2(PPh3)2, DMF, 50 °C, 85–100%; (e) RB(OH)2, PdCl2(PPh3)2,
Cs2CO3, DMF, microwave heating, 120 °C, 10–37%.
Scheme 1 illustrates the elaboration of 2a and 2b into
several new classes of compounds. Nucleophilic aro-
matic substitution of the 2-chloropyridines 2a or 2b
by treatment with phenols, 4-fluorothiophenol, 4-fluo-
roaniline or alcohols with cesium carbonate in reflux-
ing toluene, afforded the aryl ethers 3a–i, the 4-
fluorophenylthio ether 3j, the 4-fluoroaniline 3k, and
the alkyl ethers 4a–c, respectively (Tables 1 and 2).
The 2-aminopyridine derivatives, 5a–b (Table 3), were
obtained by heating 2a with iso-butylamine or piperi-
dine in THF. Imidazole 5c, benzotriazole 5d, and the
piperazine derivative 5g were generated when 2a or 2b
was refluxed in toluene with the appropriate heterocy-
Scheme 2. Reagents and conditions: (a) N-bromoethyl phthalimide,
K2CO3, DMF, 70 °C; (b) NH2NH2, MeCN, EtOH, 80 °C, 80% (two
steps); (c) cyclobutanecarbonyl chloride, NEt3, CH2Cl2, rt, 68%.
clic amine, and amides 5e and 5f were synthesized by
heating 2a with either pivalamide or 3,4-difluoroben-
zamide in the presence of cesium carbonate at
100 °C (Table 3). Incorporation of carbon substituents
at the pyridine 2-position was achieved by coupling 2b
with acetylenes under Sonagashira conditions to pro-
duce the alkynes 6a–c (Table 4) or by Suzuki coupling
of 2b with heterocyclic boronic acids to afford the
3-pyridyl (7a), 5-pyrimidinyl (7b), and 1-methyl-4-pyr-
azolyl (7c) derivatives (Table 5). The cyclobutylamido
ethyl ether 4d (Table 2, Scheme 2) was obtained when
pyridone 86 was treated with N-bromoethyl phthali-
mide, then deprotected with hydrazine to provide
amine 9, followed by treatment with cyclobutanecar-
bonyl chloride at room temperature.
Scheme 3 illustrates the synthesis of piperazine deriva-
tives 5h and 5i (Table 3). The 2-chloropyridine 2b was
initially reacted with piperazine in refluxing toluene to
afford 10. This material was then either reacted with
pivaloyl chloride to yield 5h, or treated with 3,5-dif-
luorobenzenesulfonyl chloride to generate sulfonamide
5i. The preparation of piperazine-2,3-dione 5j (Table 3)
is also shown in Scheme 3. Initial treatment of 2b with
N-iso-propylethylenediamine led to 11, which under-
went a ring-closure with oxalyl chloride affording the de-
sired compound 5j.
Scheme 3. Reagents and conditions: (a) piperazine, toluene, 100 °C,
99%; (b) RCOCl or RSO2Cl, NEt3, CH2Cl2, rt, 25–46%; (c) N-iso-
propylethylenediamine, Cs2CO3, DMF, 96%; (d) (COCl)2, MeCN,
15%.
Next we elaborated the 3-cyano group of the pyridine to
prepare more polar structures in an attempt to lower the
logD8 and improve physicochemical properties such as
vehicle solubility. The primary amide 15a, iso-propyl

C. B. Madsen-Duggan et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2031–2035
2033
Scheme 4. Reagents and conditions: (a) POCl3, reflux, 78%; (b) 3,4-di-fluorophenol, Cs2CO3, toluene, 100 °C, 86%; (c) 3 N NaOH, MeOH,
50 °C,100%; (d) (COCl)2, CH2Cl2, DMF, rt; (e) NHRR, CH2Cl2, NEt3, rt, 52–92% (two steps).
amide 15b, 1,1-dimethylhydrazide 15c, and sym-dim-
ethylhydrazide 15d, all bearing the 3,4-difluorophenoxy
group at the 2-position, were made as illustrated in
Scheme 4. Ester 126 was treated sequentially with phos-
phorus oxychloride and then 3,4-difluorophenol to
yield the desired ether 13. Compound 13 underwent
saponification to give acid 14, which was then treated
with oxalyl chloride and the appropriate amine or
hydrazine to afford 15a–d (Table 6).
Diminished potency was observed when the oxygen
at the pyridine 2-position was exchanged with sulfur
or nitrogen. The 4-fluorophenyl thioether 3j showed
an 11-fold drop, while the 4-fluoroaniline 3k had a
3-fold drop relative to the phenoxy ether 3b. Alkyl
ethers were also well tolerated (IC50 = 5.7–11 nM at
CB1) as shown in Table 2.
An array of nitrogen linked compounds at the pyridine
2-position are illustrated in Table 3. In general the
potency of these compounds was considerably less than
the dihalophenoxy compounds of Table 1.
Binding affinities were determined using a standard pro-
tocol9 and all compounds tested were found to be func-
tional inverse agonists.
Piperidine derivative 5b (CB1, IC50 = 19 nM) showed
5-fold better binding affinity than the iso-butyl deriva-
tive 5a. The aromatic nitrogen linked compounds, 5c
and 5d, showed improvement in potency, with 5d dis-
playing an IC50 value of 5.7 nM. The pivalamido and
3,4-difluorobenzamido derivatives 5e and 5f displayed
more modest activity (CB1, IC50 = 29 and 20 nM), while
in the piperazine series, the sulfonamide 5i showed excel-
lent activity (CB1, IC50 = 3.8 nM).
Our initial SAR efforts focused on comparing phenoxy,
phenylthio, and phenylamino alternatives to the benzyl-
oxy group of 1a as shown in Table 1.
The monohalo phenyl compounds, 3-fluoro 3a, 4-fluo-
ro 3b, 3-chloro 3c, and 4-chloro 3d, showed good
potency (CB1, IC50 = 4–14 nM), while the 4-methoxy-
phenoxy analog 3e was less potent (CB1, IC50
=
26 nM). Aryl halogen disubstitution enhanced potency
as seen with 3,4-difluoro 3f, 3,5-difluoro 3g, and 3,5-
dichloro 3h (CB1, IC50 = 3.7, 0.91, and 3.9 nM, respec-
tively). There was a 6-fold drop in potency with the
bis-chloro analog 3i, relative to the tris-chloro 3f,
which was consistent with our previous observation
that the tris-chloro substitution pattern was favored
over the bis-chloro in the benzyloxy series (1a vs 1b).
Carbon linked substituents at the pyridine 2-position are
shown in Tables 4 and 5. In the alkynyl series, the tert-
butyl alkyne 6a was potent (CB1, IC50 = 5.9 nM), while
the hydroxy and the amino substituted alkynes 6b and
6c displayed modest activity.
Of the heteroaryl analogs, the 3-pyridyl (7a) and the 3,5-
pyrimidinyl (7b) derivatives showed modest activity,
Table 1. Structures and binding affinities (CB1; CB2) expressed as IC50
(nM), of the 2-phenoxy derivatives 3a–k
Table 2. Structures and binding affinities (CB1; CB2) expressed as IC50
(nM), of 2-alkyl ethers 4a–d
Compound
X
R1
R2
CB1; CB2
Compound
R
CB1; CB2
10; 860
3a
3b
3c
3d
3e
3f
O
O
O
O
O
O
O
O
O
S
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
H
3-F
4-F
11; 2300
14; 4400
4; 4100
4a
3-Cl
4-Cl
7; 4200
4b
4c
4d
11; 730
6.6; 430
5.7; 450
4-OMe
3,4-diF
3,5-diF
3,5-diCl
3,4-diF
4-F
26; 3800
3.7; 4100
0.91; 320
3.9; 5600
25; 5700
154; 5200
50; 2800
3g
3h
3i
3j
3k
Cl
Cl
NH
4-F

2034
C. B. Madsen-Duggan et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2031–2035
Table 3. Structures and binding affinities (CB1; CB2) expressed as IC50
(nM), of 2-aminopyridine derivatives 5a–j
Table 4. Structures and binding affinities (CB1; CB2) expressed as IC50
(nM), of 2-alkynyl derivatives 6a-c
Compound
R1
Cl
R2
CB1; CB2
89; 1800
Compound
R1
R2
R3
CB1; CB2
5a
6a
6b
6c
Me
Me
Me
Et
Me
Me
Et
5.9; 2100
44; 1800
39; 1100
OH
NH2
5b
5c
Cl
Cl
19; 4000
12; 310
Table 5. Structures and binding affinities (CB1; CB2) expressed as IC50
(nM), of 2-heteroaryl derivatives 7a–c
5d
5e
5f
Cl
Cl
Cl
H
5.7; 465
29; 1700
20; 315
Compound
Ar
CB1; CB2
27; 3700
7a
5g
5h
5i
108; 2600
37; 1100
3.8; 1400
7b
7c
38; 4300
H
165; 3300
H
Table 6. Structures and binding affinities (CB1; CB2) expressed as IC50
(nM), of 3-amido and 3-hydrazido derivatives 15a–d
5j
H
12; 715
while the 1-methyl-4-pyrazolyl 7c had diminished poten-
cy (CB1, IC50 = 165 nM).
Compound
R
CB1; CB2
15a
15b
15c
15d
NH2
NHi-Pr
11; 4900
3.5; 1600
5.6; 1800
80; 650
Finally, the 3-carboxamidopyridine and 3-hydrazido-
pyridine analogs were examined as shown in Table 6.
Binding activity of the iso-propylamide 15b (CB1,
IC50 = 3.5 nM) showed a small improvement over the
primary amide 15a. The hydrazide 15c (CB1,
IC50 = 5.6 nM) was found to be similar to 15b, but supe-
rior to the sym-hydrazide 15d.
NHNMe2
NMeNHMe
controls gained 7 g overnight (p < 0.001) as shown in
Table 7. Compound 15c also inhibited feeding by 45%
(p = 0.009) and decreased overnight body weight gain
by 5 g versus a 4 g increase for vehicle treated controls
(p = 0.007). In contrast, compound 1a had a non-signif-
icant reduction in food intake (À22%; p > 0.05), but had
a modest, yet statistically significant, reduction in over-
night body weight gain (À1 g vs +8 g for vehicle treated
controls; p < 0.05).
Compounds 3f and 15c were selected for evaluation of
their effects on food intake and body weight changes
in diet-induced obese (DIO) rats fed overnight ad libitum
on a moderate high fat, high sucrose diet.7 Data for
compound 1a are included for comparison.6 All com-
pounds were dosed orally at 10 mg/kg.
Compound 3f suppressed cumulative food intake by
48% after 18-h post-dosing (p = 0.0014) resulting in an
8 g decrease in body weight, while vehicle treated
Pharmacokinetic properties for compounds 3f and 15c
were determined in Sprague–Dawley rats and compared
to those reported for compound 1a6 as shown in

C. B. Madsen-Duggan et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2031–2035
2035
Table 7. Rat food intake/body weight change overnight (18 h) for 3f,
15c, and 1a6
of our diarylpyridine lead will be reported from these
laboratories in due course.
Compound Dose
(mg/kg)
D Body weight (g) % FI suppression
References and notes
Vehicle
(for 3f)
3f
+7
1. Devane, W. A.; Dysarz, F. A.; Johnson, M. R.; Melvin, L.
S.; Howlett, A. C. Mol. Pharmacol. 1988, 34, 605.
2. For extensive reviews in this area, see the following:
Muccioli, G. G.; Lambert, D. M. Curr. Med. Chem. 2005,
12, 1361; Hertzog, D. L. Expert Opin. Ther. Patents 2004,
14, 1435; Smith, R. A.; Fathi, Z. Idrugs 2005, 8, 53; Lange,
J. H. M.; Kruse, C. G. Curr. Opin. Drug Discov. Devel.
2004, 7, 498; Lange, J. H. M.; Kruse, C. G. Drug Discovery
Today 2005, 10, 693.
10
10
10
À8 (p < 0.001)
À48 (p = 0.0014)
À45 (p = 0.009)
À22 (p > 0.05)
Vehicle
(for 15c)
15c
+4
À5 (p = 0.007)
Vehicle
(for 1a)
1a
+8
À1 (p < 0.05)
3. Cota, D.; Marsicano, G.; Tscho¨p, M.; Grubler, Y.;
¨
Table 8. Pharmacokinetic profiles for 3f, 15c, and 1a6
Flachskamm, C.; Schubert, M.; Auer, D.; Yassouridis,
A.; Tho¨ne-Reinecke, C.; Ortmann, S.; Tomassoni, F.;
Cervino, C.; Nisoli, E.; Linthorst, A. C. E.; Pasquali, R.;
Lutz, B.; Stalla, G. K.; Pagotto, U. J. Clin. Invest. 2003,
112, 423.
Compound
3f
15c
1a
F (%)
66
3.5
26
24
27
3.6
Clp (mL/min/kg)
t1/2 (h)
Vd (L/kg)
5.6
1.5
4.1
6.7
3.6
0.77
0.03; 0.26
´
4. Despres, J.-P.; Golay, A.; Sjo¨stro¨m, L. N. Eng. J. Med.
2005, 353, 2121.
5. ACOMPLIAÒ (rimonabant) approved in the European
Union, June 21, 2006, www.sanofi-aventis.com.
6. 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, 681, 1755.
Brain/plasma
ratio 0.25 h; 4 h
post iv dosing
0.10; 0.61
1.2; 3.0
Table 8. All three compounds showed similar half-lives,
but 3f had about 2.5 times the bioavailability of 15c or
1a. Of note was that 15c was the least restricted to plas-
ma with a volume of distribution almost nine times that
of 1a. Of greatest importance for efficacy was the higher
brain to plasma ratios observed for 3f and 15c. While
15c had lower bioavailability than 3f, the fact that it
had a larger Vd, and was better brain penetrant than
3f, accounts for its similar efficacy in vivo.
7. Debenham, J. S.; Madsen-Duggan, C. B.; Walsh, T. F.;
Wang, J.; Tong, X.; Doss, G. A.; Lao, J.; Fong, T. M.;
Schaeffer, M.-T.; Xiao, J. C.; Huang, C. R.-R. C.; Shen,
C.-P.; Feng, Y.; Marsh, D. J.; Stribling, D. S.; Shear-
man, L. P.; Strack, A. M.; MacIntyre, D. E.; Van der
Ploeg, L. H. T.; Goulet, M. T. Bioorg. Med. Chem. Lett.
2006, 16, 681.
8. For comparison the calculated logD for 1a, 3f, and 15c is
8.35, 8.08, and 6.19, respectively (Advanced Chemistry
Development logD software version 8.07).
9. The reported binding results were an average of 2–8
independent determinations (each done in replicate), which
were normally within 50% of the average value. For both
binding and functional assay conditions, see: 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.
Optimization of the 5,6-diarylpyridine lead revealed that
great diversity in structure at the pyridine 2-position is
tolerated and affords compounds with high potency
for CB1 (IC50 6 10 nM) and excellent specificity with re-
spect to CB2. Both 3f and 15c effectively suppressed
food intake and body weight increases following oral
administration in a DIO rat model of food intake. Fur-
ther SAR focusing on the 2- and 3-position substituents