Triaryl bis-sulfones as a new class of cannabinoid CB2 receptor inhibitors: Identification of a lead and initial SAR studies

Article abstract of DOI:10.1016/j.bmcl.2004.11.007

A novel class of cannabinoid CB2 receptor ligands is described. These triaryl bis-sulfones are nanomolar inhibitors of the CB2 receptor and show high selectivity over the cannabinoid CB1 receptor. One example of this new class decreases ligand-induced GTPγS binding to recombinant CB2 cell membranes, identifying the compound as a CB2-selective inverse agonist.

Full text of DOI:10.1016/j.bmcl.2004.11.007

Bioorganic & Medicinal Chemistry Letters 15 (2005) 783–786
Triaryl bis-sulfones as a new class of cannabinoid CB2
receptor inhibitors: identification of a lead and initial SAR studies
Brian J. Lavey,a,* Joseph A. Kozlowski,a R. William Hipkin,b Waldemar Gonsiorek,b
Daniel J. Lundell,b John J. Piwinski,a Satwant Narulab and Charles A. Lunnb
aDepartment of Chemistry, Schering-Plough Research Institute, 2015 Galloping Hill Road, K-15-1-1545, Kenilworth,
NJ 07033-0539, USA
bDepartment of Inflammation and Infectious Diseases, Schering-Plough Research Institute, 2015 Galloping Hill Road,
Kenilworth, NJ 07033-0539, USA
Received 14 October 2004; revised 1 November 2004; accepted 1 November 2004
Available online 18 November 2004
Abstract—A novel class of cannabinoid CB2 receptor ligands is described. These triaryl bis-sulfones are nanomolar inhibitors of the
CB2 receptor and show high selectivity over the cannabinoid CB1 receptor. One example of this new class decreases ligand-induced
GTPcS binding to recombinant CB2 cell membranes, identifying the compound as a CB2-selective inverse agonist.
Ó 2004 Elsevier Ltd. All rights reserved.
Cannabinoids produce a complex array of biological ef-
fects.1 These effects are mediated primarily through one
of two G-protein coupled receptors—CB1 and CB2. The
cannabinoid CB1 receptor is expressed primarily in the
CNS.2 This receptor is believed to be the mediator of
the psychological effects of D9-tetrahydrocannabinol
(THC). There is also evidence that the CB1 receptor is
involved in cytokine production and inflammation in
several standard mouse inflammation models.3 A second
receptor, the cannabinoid CB2 receptor, exists primarily
in immune related tissues and cells,4 offering the possi-
bility of modulating the immune system without causing
the CNS effects associated with compounds that bind to
the CB1 receptor. To date, reported functions of CB2
have included modulation of B-cell differentiation,5 al-
tered migration,6 altered antigen processing7 in macro-
phages, and altered cannabinoid-mediated anti-tumor
activity.8 Of the immunoregulatory activities ascribed
to cannabinoid compounds, some appear to be centrally
mediated through the CB1 receptor.3 Some activities are
receptor independent.9 However, a number are pertus-
sis-toxin sensitive and thus clearly receptor mediated.
We embarked on a program to identify a novel class
of highly selective cannabinoid CB2 compounds. Our
goal was to find potent CB2 selective ligands to explore
their immunological effects.
An interest in harnessing the therapeutic potential of the
cannabinioid system has lead to the development of sev-
eral unique classes of cannabinoid ligands.10 A set of
dihydrobenzopyran-type structures, including HU21011
and the CB2-selective JWH-05112 and L-75963313 were
derived from the structure of D9-tetrahydrocannabinol.
Removing the dihydropyran ring of D9-tetrahydrocan-
nabinol lead to CP55940.14 Aminoalkylindoles were
developed by Sterling Winthrop leading to development
of Win 55,212-2, AM630 and related compounds.15
Derivatives of the endocannabinoid anandamide16 lead
to
a
set of eicosanoid-like derivatives.17 Finally,
researchers from Sanofi have developed a set of biar-
ylpyrazoles, including the CB1-selective SR141716A18
and the CB2-selective SR 144528.19 SR 141716A (called
Rimonabant) is being evaluated by Sanofi for the con-
trol of obesity.20
Compound 1 was identified as a lead compound for the
program with a CB2Ki = 31nM. However, when a series
of analogs of 1 was prepared, all of them had Ki values
greater than 1lM—including the acetamide and the
methanesulfonamide. A new highly purified sample of
1 also had a CB2Ki greater than 1lM. LCMS indicated
the presence of two impurities present in the original
*
Corresponding author. Tel.: +1 908 740 2329; fax: +1 908 740
7152; e-mail: brian.lavey@spcorp.com
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.11.007

784
B. J. Lavey et al. / Bioorg. Med. Chem. Lett. 15 (2005) 783–786
O
O
O
O
O
a,b,c
61%
N
CF3
d,e
N
CF3
N
CF3
H
H
H
68%
S
S
Br
O
O
O
S
O
O
1
2
O
O
f
89%
O
Y
O
N
NH2
H
g
S
S
O
O
S
O
S
O
O
O
O
O
4
3
O
O
Scheme 1. Reagents and conditions: (a) 1.02equiv CH3Li, THF–hexanes, À78°C; (b) 1.02equiv n-BuLi; (c) 4-methoxyphenylsulfonyl fluoride; (d)
2.1equiv n-BuLi, THF–hexanes, À78°C, then 4-methoxyphenyl disulfide, À78°C to rt overnight; (e) m-CPBA, CH2Cl2; (f) 1.0M aq LiOH, dioxane;
(g) electrophile, CH2Cl2, diisopropylethyl amine, rt overnight, then Tris (2-aminoethyl)amine-polystyrene.
sample of 1. These impurities became targets for further
investigation.
for the ability to modulate the binding of [35S]GTPcS to
the cannabinoid CB2 receptor.23 Figure 1 shows
that, like the Sanofi inverse agonist SR 144528,
increasing concentrations of 4j decreases [35S]GTPcS
binding to the CB2 membrane preparation. By
comparison, the cannabinoid agonists WIN 55,212-2 and
HU210 increase [35S]GTPcS binding to the membrane
Compound 1 was resynthesized on large scale and all the
reaction products were tested in the binding assay. The
active impurity present in the screening sample of 1
was found to be compound 2. Compound 2 is formed
in 3% yield under the conditions used to prepare 1, pre-
sumably by ortho-lithiation of 1 by excess butyl lithium
in the reaction and subsequent trapping of the resulting
dianion by the sulfonyl fluoride. Compound 2 is highly
active in the CB2 inhibition assay showing sub-nano-
molar potency with fair selectivity for CB2 in preference
to CB1—(See Table 1).21
Table 1. Amide analogs of 4: Y = (C@O)–R
Entry
R
KiCB2 KiCB1 Selectivity
(nM) (nM) (CB1Ki/CB2Ki)
2
Trifluoromethyl
Methyl
Ethyl
0.3 235 783
7.7
29
6.2
4a
4b
4c
4d
4e
4f
4g
4h
4i
1779 231
3,559 122
1173 189
Propyl
In order to study 2 and its analogs, we developed a syn-
thesis of 2 based on the ortho-lithiation dianion chemis-
try. Compound 1 was treated with 2.1equiv of n-butyl
lithium, trapped with bis-(4-methoxyphenyl) disulfide,
and oxidized to the corresponding sulfone. The ortho-
lithiation was completely selective for the nonbenzamide
ring, possibly because of electrostatic repulsion by the
trifluoroacetamide anion that is initially formed. Com-
pound 2 was then deprotected giving compound 3,
which had relatively weak activity (5lM) at CB2, and
treated with a series of acid halides, sulfonyl chlorides,
and isocyanates to give amides, sulfonamides, and ureas
as products (Scheme 1).22
Cyclopropyl
t-Butyl
Benzyl
15.7 10,625 676
38.6
412
2309
478
884
1172
84
60
1.2
p-Toluyl
4-Methoxyphenyl
3,4-Dichlorophenyl
289
1258
167
3
0.9
0.5
Table 2. Sulfonamide analogs of 4: Y = (SO2)–R
Entry
R
KiCB2 KiCB1 Selectivity
(nM)
(nM)
(CB1Ki/CB2Ki)
4j
Methyl
Ethyl
Butyl
0.4
905
1968
1129
855
2263
1513
27
4k
4l
1
41
Further development of this class of compounds is illus-
trated in Tables 1–3. Several SAR trends can be seen.
The first is that for all amide and sulfonamide substitu-
ents, small hydrophobic alkyl groups are preferred for
the CB2 receptor. Aromatic groups decrease activity at
CB2, but can improve it at CB1 (4i, 4p). Sulfonamides
are more active at CB2 than the corresponding amides,
while ureas with the same R group are relatively
inactive. The compound with the best combination of
CB2 potency and selectivity for CB2 over CB1 was the
methanesulfonamide 4j.
4m
4n
4o
4p
Phenyl
p-Toluyl
239
495
3.6
3.9
1923
117
3,5-Dichlorophenyl 2476
Benzyl 160
0.05
0.5
79
Table 3. Urea analogs of 4: Y = (CONH)–R
Entry
R
KiCB2
(nM)
KiCB1
(nM)
Selectivity
(CB1Ki/CB2Ki)
4q
4r
4s
Propyl
p-Toluyl
4-F-Phenyl
495
802
72,089
100,000
17,588
145
125
15
As an initial characterization of the pharmacology of
this new class of cannabinoid compounds, 4j was tested
1136

B. J. Lavey et al. / Bioorg. Med. Chem. Lett. 15 (2005) 783–786
785
modulation by cannabinoid ligands see: (a) Klein, T. W.;
Newton, C.; Larsen, K.; Lu, L.; Perkins, I.; Nong, L.;
Friedman, H. J. Leukoc. Biol. 2003(October), 74; (b)
Pertwee, R. G. Exp. Opin. Invest. Drugs 2000, 9, 1553.
2. For a review, see Pertwee, R. G. Pharmacol. Ther. 1997,
74, 129.
3. (a) Smith, S. R.; Denhardt, G.; Terminelli, C. Eur. J.
Pharmacol. 2001, 432, 107; (b) Smith, S. R.; Terminelli, C.;
Denhardt, G. Eur. J. Pharmacol. 2001, 425, 73; (c) Smith,
S. R.; Terminelli, C.; Denhardt, G. J. Pharmacol. Exp.
Ther. 2000, 293, 136.
4. Munro, S.; Thomas, K. L.; Abu-Shaar, M. Nature 1993,
365, 61.
5. Carayon, P.; Marchand, J.; Dussossoy, D.; Derocq, J. M.;
Jbilo, O.; Bord, A.; Bouaboula, M.; Galiegue, S.; Mondi-
ere, P.; Penarier, G.; Fur, G. L.; Defrance, T.; Casellas, P.
Blood 1998, 92, 3605.
WIN 55,212-2
EC50 = 0.8 0.2 nM
6000
5000
4000
3000
2000
1000
0
HU210
EC50 = 0.4 0.4 nM
SR 144528
EC50 = 5.0 0.4 nM
4J
EC50 = 0.6 0.1 nM
0 -12 -11 -10 -9 -8 -7 -6 -5
Ligand (log M)
Figure 1. Effect of cannabinoids on [35S]GTPcS binding in Sf9-hCB2
membranes.
6. Sacerdote, P.; Massi, P.; Panerai, A. E.; Parolaro, D. J.
Neuroimmunol. 2000, 109, 155.
7. McCoy, K. L.; Matveyeva, M.; Carlisle, S. J.; Cabral, G.
A. J. Pharmacol. Exp. Ther. 1999, 289, 1620.
8. Zhu, L. X.; Sharma, S.; Stolina, M.; Gardner, B.; Roth,
M. D.; Tashkin, D. P.; Dubinett, S. M. J. Immunol. 2000,
165, 373.
preparations with increasing concentrations. These
results are consistent with 4j being an inverse agonist.
Several classes of CB2-selective inverse agonist com-
pounds have been described. These include SR
14452819 and JTE-90724 which is highly potent and
selective for rodent CB2 (rat Ki = 0.38nM, 2760 selectiv-
ity; mouse Ki = 1.55nM; 684 selectivity) but its potency
for human CB2 is less impressive (Ki = 35.9nM, 66
selectivity). Both compounds show significant dose-
dependent anti-inflammatory activity, reducing carrage-
enan-induced paw edema in mice.24 AM63013 also
exhibits a Ki for the human CB2 receptor (using [3H]
CP55,940 as radioligand) of 31.2nM, with selectivity
over human CB1 of 165.
9. Sancho, R.; Calzado, M. A.; DiMarzo, V.; Appendino,
G.; Munoz, E. Mol. Pharmacol. 2003, 63, 429.
10. For a review, see 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.
11. Showalter, V. M.; Compton, D. R.; Martin, B. R.; Abood,
M. E. J. Pharmacol. Exp. Ther. 1996, 278, 989.
12. Huffman, J. W. Curr. Pharm. Des. 2000, 6, 1323.
13. Ross, R. A.; Brockie, H. C.; Stevenson, L. A.; Murphy, V.
L.; Templeton, F.; Makriyannis, A.; Pertwee, R. G. Br. J.
Pharmacol. 1999, 126, 665.
14. Johnson, M. R.; Melvin, L. S. In Cannabinoids as
Therapeutic Agents; Mechoulam, R., Ed.; CRC: Boca
Raton, 1986; pp 121–145.
15. (a) Pertwee, R.; Griffin, G.; Fernando, S.; Li, X.; Hill, A.;
Makriyannis, A. Life Sci. 1995, 56, 1949; (b) Huffman, J.
W. Curr. Med. Chem. 1999, 6, 705.
16. Devane, W. A.; Hanus, L.; Breuer, A.; Pertwee, R. G.;
Stevenson, L. A.; Griffin, G.; Gibson, D.; Mandelbaum,
A.; Etinger, A.; Mechoulam, R. Science 1992, 258, 1946.
17. Reviewed in Palmer, S. L.; Khanolkar, A. D.; Makriyan-
nis, A. Curr. Pharm. Des. 2000, 6, 1381.
18. Rinaldi-Carmona, M.; Barth, F.; Heaulme, M.; Shire, D.;
Calandra, B.; Congy, C.; Martinez, S.; Maruani, J.;
Neliat, G.; Caput, D. FEBS Lett. 1994, 350, 240.
19. Rinaldi-Carmona, M.; Barth, F.; Millan, J.; Derocq, J.
M.; Casellas, P.; Congy, C.; Oustric, D.; Sarran, M.;
Bouaboula, M.; Calandra, B.; Portier, M.; Shire, D.;
Breliere, J. C.; Le Fur, G. L. J. Pharmacol. Exp. Ther.
1998, 284, 644.
The biologic roles of CB2-selective inverse agonists have
yet to be fully understood. Most reported work with
these agents demonstrates an ability to function as
antagonists to added cannabinoid ligands. For example,
in studies investigating the role of the CB2 receptor in
pain, SR144528 has no significant effect on carrage-
enan-evoked thermal and mechanical hyperalgesia at
concentrations shown to block the activity of CB2-selec-
tive agonists.25 Studies are on-going to establish an in
vivo biology for this class of compounds.
In summary, we have identified a novel class of cannabi-
noid CB2 specific ligands with high affinity for the
human receptor. Initial pharmacologic characterization
of 4j, the most potent and selective compound of this
series, suggests that the class are inverse agonists for
the CB2 receptor and can have excellent selectivity rela-
tive to the CB1 receptor. It is hoped that with the dis-
covery of increasingly potent and specific compounds
the therapeutic potential of the cannabinoid CB2 recep-
tor can be realized.
20. Vickers, S. P.; Webster, L. J.; Wyatt, A.; Dourish, C. T.;
Kennett, G. A. Psychopharmacol. 2003, 167, 103.
21. Unless otherwise noted, all Ki data will refer to the human
CB2 receptor.
22. All products gave satisfactory analytical data as indicated
by 1H NMR and LCMS. Physical data for 4j is as follows:
1H NMR (300MHz, CDCl3) d 8.43 (d, 8.7Hz, 1H), 7.95–
7.90 (m, 6H), 7.46 (d, 9Hz, 2H), 7.21–7.18 (m, 1H), 6.97
(d, 9Hz, 2H), 4.74–4.60 (m, 1H), 4.60 (d, 6.3Hz, 1H,
(NH)), 3.95 (s, 3H), 3.85 (s, 3H), 2.65 (s, 3H), 1.53 (d,
6.9Hz, 3H); LCMS EI exact mass calculated for
C23H25NO8S3 (parent) 539.07 m/z found (M+H+) 540.1.
For experimental details of the synthesis and the determi-
nation of Ki values, see: Kozlowski, J. A.; Shih, N. -Y.;
References and notes
1. For reviews on cannabinoid biology, see (a) Watson, S. J.;
Benson, J. A., Jr.; Joy, J. E. Arch. Gen. Psychiatry 2000,
57, 547; (b) Goutopoulos, A.; Makriyannis, A. Pharmacol.
Ther. 2002, 95, 103; for reviews discussing immune

786
B. J. Lavey et al. / Bioorg. Med. Chem. Lett. 15 (2005) 783–786
Lavey, B. J.; Rizvi, R. K.; Shankar, B. P; Spitler, J. M.;
20mM HEPES and 10mM sodium pyrophosphate, and
membrane-bound radioactivity was measured by liquid
scintillation. Nonlinear regression analysis of the data was
performed with Prism 2.0b (GraphPad, San Diego, CA).
Additional information may be found in: Gonsiorek, W.;
Lunn, C.; Fan, X.; Narula, S.; Lundell, D.; Hipkin, R. W.
Mol. Pharmacol. 2000, 57, 1045.
Tong, L.; Wolin, R.; Wong, M. K. World Patent
WO02062750.
23. Triplicate serial dilutions of test compound, or cannabi-
noid agonists WIN55,212-2, HU210, and inverse agonist
SR144528 were incubated with 2lg Sf9 cell membranes
expressing Gai3, b1c2, and hCB2 (7–14pmol/mg) and
0.3nM [35S]GTPcS in 20mM HEPES, 100mM
NaCl, 5mM MgCl2, and 0.2% (w/v) BSA (Factor V, lipid
free), pH7.4 supplemented with 2.5lM GDP. Following
incubation for 30min at 30°C the reaction was terminated
by rapid filtration through the microfiltration plates. The
filters were washed 10 times at room temperature with
24. Iwamura, H.; Suzuki, H.; Ueda, Y.; Kaya, T.; Inaba, T. J.
Pharmacol. Exp. Ther. 2001, 296, 420.
25. (a) Nackley, A. G.; Makriyannis, A.; Hohmann, A. G.
Neuroscience 2003, 119, 747; (b) Nackley, A. G.; Suplita,
R. L., II; Hohmann, A. G. Neuroscience 2003, 117,
659.