Sulfamoyl benzamides as novel CB2 cannabinoid receptor ligands

Article abstract of DOI:10.1016/j.bmcl.2008.04.006

Sulfamoyl benzamides were identified as a novel series of cannabinoid receptor ligands. Starting from a screening hit 8 that had modest affinity for the cannabinoid CB₂ receptor, a parallel synthesis approach and initial SAR are described, lead

Full text of DOI:10.1016/j.bmcl.2008.04.006

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 2830–2835
Sulfamoyl benzamides as novel CB2 cannabinoid receptor ligands
Karin Worm,a,* Q. Jean Zhou,a Christopher T. Saeui,a Rosalyn C. Green,a Joel A. Cassel,b
Gabriel J. Stabley,b Robert N. DeHaven,b Nathalie Conway-James,b
Christopher J. LaBuda,b Michael Koblish,b Patrick J. Littleb and Roland E. Dollea
aDepartment of Chemistry, Adolor Corporation, 700 Pennsylvania Drive, Exton, PA, USA
bDepartment of Pharmacology, Adolor Corporation, 700 Pennsylvania Drive, Exton, PA, USA
Received 21 January 2008; revised 1 April 2008; accepted 1 April 2008
Available online 7 April 2008
Abstract—Sulfamoyl benzamides were identified as a novel series of cannabinoid receptor ligands. Starting from a screening hit 8
that had modest affinity for the cannabinoid CB2 receptor, a parallel synthesis approach and initial SAR are described, leading to
compound 27 with 120-fold functional selectivity for the CB2 receptor. This compound produced robust antiallodynic activity in
rodent models of postoperative pain and neuropathic pain without traditional cannabinergic side effects.
Ó 2008 Elsevier Ltd. All rights reserved.
Two cannabinoid receptors, CB1 and CB2, have been
identified and subsequently cloned. They belong to the
family of G-protein coupled receptors and share 44%
amino acid sequence homology but differ in anatomical
distribution. The CB1 receptor is expressed mainly in the
CNS and to a lesser extent in other tissues. The CB2
receptor is primarily expressed in peripheral tissues asso-
ciated with immune functions, including macrophages,
B and T cells, as well as in peripheral nerve terminals
and on mast cells.1 D9-Tetrahydrocannabinol (THC,
1), the main active component of Cannabis sativa, and
other classical cannabinoids display a wide range of
physiological effects including analgesic, anti-inflamma-
tory, anti-convulsive and immunosuppressive activities.2
Cannabinoid receptor agonists also induce a number of
unwanted CNS effects, which are believed to be medi-
ated predominantly by the central distribution pattern
of CB1 receptors.3
(Fig. 1).6 Compound 4 (GW405833) was shown to be
antihyperalgesic in rodent models of neuropathic, inci-
sional and chronic inflammatory pain, but had no sig-
nificant effect in CB2 knockout mice in the same
assays.7 Compound 5 (AM1241) was reported to reverse
carrageenan-induced inflammatory thermal hyperalge-
sia in rats. This effect was attenuated by a CB2 selective
antagonist, but not a CB1 selective antagonist.8 Thus,
there is considerable interest in developing new cannab-
imimetic compounds possessing preferentially high
affinity for the CB2 receptor, which could lead to novel
therapeutics for the treatment of inflammation and
chronic pain.9
During a high-throughput screening campaign10a we
identified 8 as a compound with modest affinity for the
CB2 receptor (Fig. 2). Initially, SAR was explored via
a parallel approach shown in Scheme 1 and Figure 3.
Starting from commercially available 4-bromo-3-(chlo-
rosulfonyl)benzoic acid 10a and amines 11a–h, we pre-
pared eight 4-bromo-3-sulfamoyl-benzoic acids 12a–h.
A diverse set of 10 amines 13a–j was attached to alde-
hyde-based polystyrene resin via reductive amination
using sodium triacetoxyborohydride as the reducing
agent.11 The resulting resin-bound amines 14a–j were
then reacted with the sulfamoyl-benzoic acids 12a–h
previously obtained using the coupling reagent
A separation between therapeutic effects and undesir-
able CNS side effects could be accomplished either by
preventing the cannabinoid from crossing the blood–
brain barrier4 or by increasing the selectivity for the
CB2 receptor over the CB1 receptor.5 Several structural
classes have displayed selectivity for the CB2 receptor
Keywords: Cannabinoid CB2 receptor; Parallel synthesis; Sulfamoyl
benzamides; Structure–activity relationship; Antiallodynic activity.
bromo-tris-pyrrolidinophosphonium
hexafluorophos
phate (PyBrop) and diisopropylethylamine. After cleav-
age from solid support with trifluoroacetic acid in
*
Corresponding author. Tel.: +1 484 595 1944; fax: +1 484 595
1551; e-mail: kworm@adolor.com
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.04.006

K. Worm et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2830–2835
2831
OH
O
OH
O
O
O
1, THC
3, JWH-133
2, HU-308
O
I
O
N
N
N
NH
O
N
NO2
O
N
N
N
Cl
HN
N
N
O
6
5, AM1241
O
Cl
Cl
7, SR144528
Cl
4, GW405833
Cl
Figure 1. Cannabinoid receptor ligands.
attempted to introduce this S-fenchyl residue into
our system via the solid phase route just described.
Due to steric hindrance the coupling of fenchyl amine
to the polystyrene solid support failed. Therefore
highly branched analogs 23– 37 were synthesized in
solution from respective sulfamoyl-benzoic acids 12a–d
utilizing O-(benzotriazol-1-yl)-N,N,N0,N0-tetra-meth-
yluronium tetrafluoroborate (TBTU) as the coupling
reagent (Table 2).12,13 All four sulfamoyl-benzoic
acids 12a–d yielded analogs with greatly improved
binding affinity (23– 26). Morpholine, pyrrolidine,
and piperidine in R2 showed similar profiles, with
the morpholine analog 23 being slightly more
selective. Methylbenzyl amine in R2 led to the most
O
R1
O
Br
O
R2
N
S
S
O
O
8
9
O
R3
N
H
O
Ki CB1 > 1000 nM
Ki CB2 = 800 nM
Figure 2. Screening hit.
R1
O
R1
R1
O
S
O
R2
R2
Cl
S
S
selective analog 25 with
a binding constant Ki
b, c
or d
O
O
a
O
CB2 = 11 nM and P1000-fold lower binding to the
CB1 receptor (33% inhibition at 10 lM). Compounds
23– 26 were then evaluated in the [35S]GTPcS func-
tional assay.10b Compounds 23, 24, and 26 were full
agonists, but the most selective compound 25
behaved as an inverse agonist. To exclude possible
reactivity with proteins in vivo the bromo substituent
in 23 was replaced with a methyl group. Starting the
synthesis from 3-(chlorosulfonyl)-4-methylbenzoic acid
10b we obtained compound 27, a 31-fold selective
agonist with functional activities of EC50 CB2 =
4.6 nM and EC50 CB1 = 550 nM.
O
OH
O
R3
O
OH
10a: R1 = Br
9a: R1 = Br
12a: R1 = Br
10b: R1 = CH3
9b: R1= CH3
12b: R1 = CH3
Scheme 1. Reagents and conditions: (a) R2-amine 11a–h EtOAc; (b)
resin-bound R3-amine 14a–j, i-Pr2EtN, PyBrop, CH2Cl2; (c) TFA/
CH2Cl2; (d) R3-amine, TBTU, i-Pr2EtN, ACN.
dichloromethane 80 final compounds, 9a, were obtained.
Almost half of these compounds retained or improved
binding affinity for the CB2 receptor compared to 8,
while the remaining compounds lost affinity for both
CB receptors. Representative examples are shown in
Table 1. Branched alkyl amines seemed to be preferred
as both neopentyl and isobutyl amides yielded combina-
tions with improved binding affinity (K i CB2 = 100–
450 nM).
In an attempt to further improve selectivity and
retain agonist activity, other commercially available
branched amines were attached to 12b (R2 = morpho-
lino, Table 2, 28–37). Bicyclic amine substituents with
branching in the 1 and/or 2 position seem to be pre-
ferred. Globular amines like 2-adamantyl, 1-(1-adaman-
tyl)ethylamine, and bornyl amine yielded compounds
with binding constants Ki CB2 6 10 nM. Analogs con-
taining open chain and monocyclic amides, as well as
analog 34 containing the R-isomer of fenchyl amine, lost
Since the selective CB2 antagonist SR1445286b (7)
also bears a highly branched amine substituent, we

2832
K. Worm et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2830–2835
R2-amine 11:
R2-amine 11:
H
H
N
H
H
H
H
N
NH2
N
NH2
N
N
N
O
c
a
b
d
e
g
h
f
R3-amine 13:
R3-amine 13:
NH2
NH2
NH2
NH2
NH2
NH2
O
f
g
h
OH
O
N
O
a
b
c
NH2
Cl
NH2
NH2
NH2
N
i
j
d
e
Retained or improved
activity in CB1 and/or CB2
Lost activity
Figure 3. R2 and R3 amines used to prepare 9a.
Table 1. Parallel synthesis approach—example resultsa,b
Br
O
S
R2
O
O
R3
9a
CB1 (nM)
Compound
R2
R3
K
Ki CB2 (nM)
Ratio CB1/CB2
13
i
14
1300
1400
100
N
NH
15
16
200
320
7
5
N
N
NH
NH
O
1800
N
17
18
1100
360
410
3
N
N
NH
NH
>1000
>2
19
20
2000
1300
400
450
5
3
N
N
NH
NH
O
21
22
>1000
>1000
430
340
>2
>3
N
N
NH
NH
a Values are the geometric means computed from at least three separate determinations.
b For assay description see Ref. 10.

K. Worm et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2830–2835
2833
at least one order of magnitude in CB2-binding affinity.
No improvement in selectivity could be observed.
A
20
15
10
5
Catalepsy in mice is a behavior that is indicative of
central CB1 receptor activation and predictive of can-
nabinoid psychoactivity. For example, Pertwee dem-
onstrated a correlation between catalepsy in the ring
test in mice and ataxia in the dog static ataxia mod-
el.14 Therefore catalepsy in mice is likely a predictor
of central cannabinergic effects in humans.15 Since
compound 27 had a functional selectivity of 120-fold
over the CB1 receptor, and it did not produce cata-
lepsy at doses of 6 and 10 mg/kg ip, it was evaluated
for efficacy in two rodent models of nociception, the
L5 spinal nerve ligation model (L5 SNL)16 of neuro-
pathic pain and the hindpaw incision model17 of post-
operative pain (Fig. 4). In the L5 SNL model,
compound 27 reversed the nerve injury-induced tactile
allodynia at a dose of 3 mg/kg ip. The magnitude of
the antiallodynic effect was comparable to the effect
produced by a dose of 60 mg/kg ip of gabapentin. In
the incisional pain model, a dose of 10 mg/kg ip of
compound 27 produced a significant antiallodynic ef-
fect that was comparable to the antiallodynic effect
that was produced by a dose of 3 mg/kg sc of mor-
phine. There were no overt behavioral side effects that
were associated with doses up to 10 mg/kg ip of com-
pound 27. Administered orally at 30 and 100 mg/kg
compound 27 was not active in the incisional pain
model, due to low bioavailability.
0
Sham
Treatment [mg/kg]
Vehicle Gabapentin
100 (i.p.)
Compound 27
0.3 (i.p.) 3 (i.p.)
B
30
25
20
15
10
5
0
Sham
Vehicle
Morphine Codeine Compd 27
3 (s.c.) 10 (i.p.) 10 (i.p.)
Treatment [mg/kg]
In summary, the sulfamoyl benzamides represent a new
class of ligands that bind to the cannabinoid receptors.
Derived from a screening hit with modest affinity to
the CB2 receptor, large lipophilic, branched amide
substituents led to improved receptor binding. Increased
selectivity for the CB2 receptor was achieved by the
introduction of an S-fenchyl residue. Small changes in
the sulfonamide part of the molecule produced a switch
from full agonist to inverse agonist. Compound 27, a
compound with 120-fold functional selectivity for the
CB2 receptor and devoid of traditional cannabimimetic
side effects, dosed ip, showed robust antinociceptive
Figure 4. In vivo activity of compound 27 in two rat models: (A) L5
SNL model. Mechanical paw-withdrawal thresholds for the left
hindpaw of the sham-operated control group, L5 ligation group, and
gabapentin or compound 27 treatment groups. (B) Hindpaw incision
model. Mechanical paw-withdrawal thresholds for the left hindpaw of
the sham-operated control group, hindpaw incision group, and
morphine, codeine and compound 27 treatment groups. Data are
plotted as the mean ( SEM) paw-withdrawal threshold of the left
hindpaw for each group. All statistical analyses were performed with
one-way ANOVA followed by post-hoc comparisons (protected t-test)
among groups. *p > 0.05 compared to vehicle-treated, surgery (L5
SNL or hindpaw incision) group.
Table 2. Highly branched sulfamoyl benzamidesa,b
R1
R2
O
O
S
9
O
R3
Compound
R1
Br
R2
R3
Ki CB1 (nM)
Ki CB2 (nM)
Ratio CB1/CB2
25
EC50 CB1
560
EC50 CB2
8.8
Ratio
64
23
89
3.5
N
O
N
H
24
Br
110
9.4
12
220
4.3
51
N
N
H
(continued on next page)

2834
K. Worm et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2830–2835
Table 2 (continued)
Compound
R1
R2
R3
Ki CB1 (nM)
Ki CB2 (nM)
Ratio CB1/CB2
>100
EC50 CB1
EC50 CB2
Ratio
25
Br
>1000
11
na
nac
N
N
H
26
27
28
29
30
31
32
33
34
35
36
37
Br
93
13
7
280
550
270
290
210
210
570
920
880
540
2400
>1000
6.6
4.6
12
42
120
23
31
15
8
N
N
N
N
N
N
N
N
N
N
N
N
N
H
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
130
33
3.9
5.6
6.3
8.1
10
31
15
5
O
O
O
O
O
O
O
O
O
O
O
N
H
N
H
77
9.3
14
NH
H
N
50
7
91
9
25
NH
H
N
260
310
540
470
1300
>1000
19
14
9
48
12
44
35
12
19
nd
H
N
34
21
48
11
5
25
N
H
H
N
92
44
180
390
9
125
1800
NH
NH
>3
a Values are the geometric means computed from at least three separate determinations.
b For assay description see Ref. 10.
c Antagonist with IC50 CB2 = 14 nM.
activity in rodent models of neuropathic and postopera-
tive pain that was comparable to the effect produced by
gabapentin and morphine, respectively. Due to their
chemical accessibility and pharmacological activity, sul-
famoyl benzamides can be considered an attractive lead
series for further optimization, with the focus directed

K. Worm et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2830–2835
2835
C filters that had been pre-soaked overnight in 0.5% (w/v)
poly(ethyleneimine) and 0.1% BSA in water. The filters
were rinsed six times with 1 mL each of cold assay buffer,
30 lL of MicroScint 20 (Perkin-Elmer) was added to each
filter and the radioactivity on the filters was determined by
scintillation spectroscopy in a TopCount (Perkin-Elmer).
Nonspecific binding was determined in the presence of 10
lM WIN55212-2.; (b) The [35S]GTPcS binding method is
a major modification of the method by Selley, D. E.;
Stark, S.; Sim, L. J.; Childers, S. R. Life Sci. 1996, 59, 659:
CB2-mediated stimulation of [35S]GTPcS binding was
towards improving physicochemical and DMPK
properties.
References and notes
1. (a) Howlett, A. C.; Barth, F.; Bonner, T. I.; Cabral, G.;
Casellas, P.; Devane, W. A.; Felder, C. C.; Herkenham,
M.; Mackie, K.; Martin, B. R.; Mechoulam, R.; Pertwee,
R. G. Pharmacol. Rev. 2002, 54, 161; (b) Pertwee, R. G.
Prog. Neurobiol. 2001, 63, 569.
measured in
a
mixture containing
100–150 pM
[35S]GTPcS, 150 mM NaCl, 45 mM MgCl2, 3 lM GDP,
0.4 mM dithiothreitol, 1.0 mM EGTA, 1.0 mg/mL fatty
acid free bovine serum albumin, 25 lg of membrane
protein, and agonist in a total volume of 250 lL of 50 mM
Tris–HCl buffer, pH 7.0 in 96-well Basic FlashPlates
(Perkin-Elmer). After incubation at room temperature for
6 h, the plates were centrifuged at 800 g at 4 °C for 5 min
and the radioactivity bound to the membranes was
determined by scintillation spectrometry using a Top-
Count (Perkin-Elmer). The extent of stimulation over
basal [35S]GTPcS binding was calculated as a percentage
of the stimulation by 10 lM WIN55212-2. Basal
[35S]GTPcS binding was determined in the absence of
agonist. Generally, the stimulation by 10 lM WIN55212-2
was between 50% and 100% over basal binding. Full
agonists stimulate binding to the same maximal extent as
WIN55212-2.
2. (a) Farquhar-Smith, W. P. Pain Rev. 2002, 9, 41; (b) Klein,
T. W. Nat. Rev. Immunol. 2005, 5, 400; (c) Walter, L.;
Stella, N. Br. J. Pharmacol. 2004, 141, 775; (d) Smith, P. F.
Curr. Opin. Investig. Drugs 2005, 6, 680.
3. Rice, A. S. Curr. Opin. Investig. Drugs 2001, 2, 399.
4. Worm, K.; Zhou, Q. J.; Stabley, G. J.; DeHaven, R. N.;
Conway-James, N.; LaBuda, C. J.; Koblish, M.; Little, P.
J.; Dolle, R. E. Abstracts of Papers, 234th National
Meeting of the American Chemical Society, Boston, MA,
August 19–23, 2007; MEDI 030.
5. Malan, T. P.; Ibrahim, M. M.; Lai, J.; Vanderah, T. W.;
Makriyannis, A.; Porreca, F. Curr. Opin. Pharmacol.
2003, 3, 62.
6. (a) Huffman, J. W. Mini Rev. Med. Chem. 2005, 5, 641;
(b) Rinaldi Carmona, M.; Barth, F.; Millan, J.; Derocq, J.;
Casellas, P.; Congy, C.; Oustric, D.; Sarran, M.; Bouab-
oula, M.; Calandra, B.; Portier, M.; Shire, D.; Breliere, J.;
Le Fur, G. J. Pharmacol. Exp. Ther. 1998, 284, 644.
7. Valenzano, K. J.; Tafesse, L.; Lee, G.; Harrison, J. E.;
Boulet, J. M.; Gottschall, S. L.; Mark, L.; Pearson, M. S.;
Miller, W.; Shan, S.; Rabadi, L.; Rotshteyn, Y.; Chaffer, S.
M.; Turchin, P. I.; Elsemore, D. A.; Toth, M.; Koetzner, L.;
Whiteside, G. T. Neuropharmacology 2005, 48, 658.
8. (a) Quartilho, A.; Mata, H. P.; Ibrahim, M. M.; Vande-
rah, T. W.; Porreca, F.; Makriyannis, A.; Malan, T. P.
Anesthesiology 2003, 99, 955; (b) Ibrahim, M. M.; Deng,
H.; Zvonok, A.; Cockayne, D. A.; Kwan, J.; Mata, H. P.;
Vanderah, T. W.; Lai, J.; Porreca, F.; Makriyannis, A.;
Malan, T. P. Proc. Natl. Acad. Sci. U.S.A. 2003, 100,
10529.
11. General procedure for preparing resin-bound amines 14a– j:
The polystyrene resin with aldehyde linker (1.6 mmol/g,
500 mg, 0.8 mmol) was placed in a vessel, 5 mL of solvent
dichloroethane were added, followed by 4 mmol (5 equiv)
of respective amine 13a– j. The mixture was shaken for 2 h
at rt, then the reducing agent NaB(OAc)3H (4 mmol, 5
equiv, 848 mg) was added and shaken overnight. The resin
was filtered, washed six times, alternating between
dichloromethane and MeOH and used for the next step.
12. Compounds were fully characterized by 1H NMR and LC/
MS.
1
13. Compound 27: H NMR (CDCl3), d 0.86 (s, 3H), 1.11 (s,
3H), 1.19 (s, 3H), 1.29 (m, 3H), 1.52 (m, 1H), 1.72 (m, 2H),
1.82 (t, J = 2 Hz, 1H), 2.69 (s, 3H), 3.20 (t, J = 5 Hz, 4H),
3.73 (t, J = 5 Hz, 4H), 3.83 (dd, J = 9 Hz and 2 Hz, 1H),
6.07 (d, J = 9 Hz, 1H), 7.43 (d, J = 8 Hz, 1H), 7.86 (dd,
J = 8 Hz and 2 Hz, 1H), 8.26 (d, J = 2 Hz, 1H). MS m/z
421 (M+H+). Anal. Calcd (C22H32N2O4S): C, 62.83; H,
7.67; N, 6.66. Found: C, 62.84; H, 7.73; N 6.66.
14. Pertwee, R. G. Br. J. Pharmacol. 1972, 46, 753.
15. (a) Compton, D. R. Marijuana: Int. Res. Rep. 1987, 7, 213;
(b) Gill, E. W.; Jones, G. Biochem. Pharmacol. 1972, 21,
2237; (c) Gill, E. W.; Paton, W. D. M.; Pertwee, R. G.
Nature 1970, 228, 134.
9. Whiteside, G. T.; Lee, G. P.; Valenzano, K. J. Curr. Med.
Chem. 2007, 14, 917.
10. (a) Binding assays were performed by modification of the
method of Pinto, J. C.; Potie, F.; Rice, K. C.; Boring, D.;
Johnson, M. R.; Evans, D. M.; Wilken, G. H.; Cantrell, C.
H.; Howlett A. Mol. Pharmacol. 1994, 46, 516: Receptor
binding assays were performed by incubating 0.2–0.6 nM
[3H]CP55940 with membranes prepared from cells
expressing cloned human CB1 or CB2 receptors in buffer
consisting of 50 mM Tris–HCl, pH 7.0, 5.0 mM MgCl2,
1.0 mM ethylene glycol-bis(2-aminoethylether)-N,N,N0,
N0-tetraacetic acid (EGTA), and 1.0 mg/mL fatty acid free
bovine serum albumin. After incubation for 60 min at
room temperature for CB2 binding or 120 min at 30 °C for
CB1 binding, the assay mixtures were filtered through GF/
16. LaBuda, C. J.; Little, P. J. J. Neurosci. Methods 2005,
144, 175.
17. Brennan, T. J.; Vandermeulen, E. P.; Gebhart, G. F. Pain
1996, 64, 493.