Non-aromatic A-ring replacement in the triaryl bis-sulfone CB2 receptor inhibitors

Article abstract of DOI:10.1016/j.bmcl.2009.11.084

The triaryl bis-sulfone 1 was modified by converting the aryl A-ring to a piperidine ring. The piperidine ring was further elaborated to a spirocyclopropyl piperidine moiety. The effect on CB2 binding potency, rat calcium channel affinity, and CYP 2C9 inh

Full text of DOI:10.1016/j.bmcl.2009.11.084

Bioorganic & Medicinal Chemistry Letters 20 (2010) 608–611
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Non-aromatic A-ring replacement in the triaryl bis-sulfone CB2
receptor inhibitors
a
a
a
a
a
Eric J. Gilbert a, , Guowei Zhou , Michael K. C. Wong , Ling Tong , Bandarpalle B. Shankar , Chunli Huang ,
Joseph Kelly a, Brian J. Lavey a, Stuart W. McCombie a, Lei Chen a, Razia Rizvi a, Youhao Dong a, Youheng Shu a,
Joseph A. Kozlowski a, Neng-Yang Shih a, R. William Hipkin b, Waldemar Gonsiorek b, Asra Malikzay c,
Charles A. Lunn c, Len Favreau d, Daniel J. Lundell b
*
a Department of Medicinal Chemistry, Merck Research Laboratories, 2015 Galloping Hill Road, Kenilworth, NJ 07033-0539, USA
b Department of Inflammation and Infectious Diseases, Merck Research Laboratories, 2015 Galloping Hill Road, Kenilworth, NJ 07033-0539, USA
c Department of New Lead Discovery, Merck Research Laboratories, 2015 Galloping Hill Road, Kenilworth, NJ 07033-0539, USA
d Department of Drug Metabolism, Merck Research Laboratories, 2015 Galloping Hill Road, Kenilworth, NJ 07033-0539, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The triaryl bis-sulfone 1 was modified by converting the aryl A-ring to a piperidine ring. The piperidine
ring was further elaborated to a spirocyclopropyl piperidine moiety. The effect on CB2 binding potency,
rat calcium channel affinity, and CYP 2C9 inhibition is described.
Received 6 October 2009
Revised 16 November 2009
Accepted 17 November 2009
Available online 22 November 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
CB2
Cannabinoid
Inverse agonist
The cannabinoid receptors CB1 and CB2 are G-protein coupled
receptors that share 44% homology.1 While the CB1 receptor is
found primarily in the CNS, the CB2 receptor is highly expressed
in the cells of the immune system.2 Both CB1 and CB2 receptors
have been the focus of much research since their discovery in
1990 and 1993, respectively.3,4 For instance, CB1 and CB2 agonists
have been pursued for the treatment of neuropathic and inflamma-
tory pain.5 The most active area of research has been with CB1
antagonists and inverse agonists. Work in this area is exemplified
by Rimonabant (marketed as AcompliaÒ in Europe) for the treat-
ment of obesity.6,7
Relative to these areas of cannabinoid receptor research, the
CB2 antagonists and inverse agonists have received less attention.
Due to the presence of CB2 receptors on immune cells, compounds
that interact with this receptor hold promise for modulating the
immune system.8–10 We have reported that triaryl bis-sulfone 1
is a potent inhibitor of the CB2 receptor with good selectivity over
the CB1 receptor.11 Selectivity over the CB1 receptor is important
in avoiding unwanted CNS side effects.
been replaced by a non-aromatic moiety. In addition to finding a
structurally distinct chemotype, the backup program strove to (a)
reduce the affinity for the rat calcium channel relative to 1 to avoid
any potential issues with the preclinical toxicology species and, (b)
to improve on the cytochrome P450 profile and in particular the
2C9 isozyme (Fig. 1).
Initial efforts focused on scanning replacements for the aryl A-
ring of 1. Our synthetic strategy involved reaction of a B,C-ring sul-
fonyl chloride with a mono-N-trifluoromethyl sulfonamido-dia-
mine to provide the target molecules.
The B,C-ring sulfonyl chloride was prepared starting with Boc-
protected 4-chloroaniline 3 (Scheme 1). Directed ortho metalation
of 3 with t-BuLi and reaction with bis(2-fluorophenyl)disulfide
provided a thioether that was oxidized with m-CPBA to yield sul-
Cl
SO2CF3
Cl
A
SO2CF3
N
A
B
B
H
S
S
Herein we report on our efforts to find a structurally distinct
backup series in which the A-ring of the triaryl bis-sulfone 1 has
O2
O2
1
SO2
F
SO2
F
2
C
C
hCB2 = 0.7 nM
rCa+2 = 0.5 µM
CYP 2C9 = 3.5 µM
* Corresponding author. Tel.: +1 908 740 5571.
E-mail address: eric.gilbert@spcorp.com (E.J. Gilbert).
Figure 1. Converting the A-ring of 1 to a non-aryl moiety.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.11.084

E. J. Gilbert et al. / Bioorg. Med. Chem. Lett. 20 (2010) 608–611
609
Table 1
Cl
Cl
Cl
c, d
a, b
Scanning A-ring replacements of 1
NHBoc
SO2Cl
NHBoc
Cl
B
SO2
F
SO2
F
SO2CF3
A
3
4
5
S
O2
SO2
2
C
F
Cl
C
SO2CF3
A
e
B
S
2
O2
Compound
A
Kia
Kia
(CB1, nM)11
SO2
F
(CB2, nM)11
N
1
0.7
1600
9900
Scheme 1. Reagents and conditions: (a) (i) t-BuLi (2.4 equiv), THF, À78 °C; (ii) 1,2-
bi-(2-fluorophenyl)disulfide, À20 °C to rt; (b) m-CPBA, dichloroethane, rt; (c) 4 M
HCl in dioxane; (d) (i) NaNO2, HCl(aq), AcOH, 0 °C, 1 h; (ii) CuCl, SO2/AcOH, 0 °C, 2 h;
(e) A-ring amine, TEA, CH2Cl2.
H
6b
NH
2950
1290
N
H
N
fone 4. Compound 4 was then deprotected and converted to the
diazonium salt. The diazonium salt was added to a saturated solu-
tion of sulfur dioxide in acetic acid containing copper chloride to
afford sulfonyl chloride 5.12 Subsequent treatment of 5 with a
mono-N-trifluoromethyl sulfamido-diamine yielded target mole-
cule 2.
The effect of exchanging the aryl A-ring of 1 with various non-
aromatic amines is shown in Table 1. In general, the linker length
between the amines of the A-ring was important (10 vs 6 and 9)
with a 4-atom linker being optimal (10). In addition, the manner
in which the amines of the A-ring were conformationally con-
strained impacted the binding affinity (7 vs 8). Piperidine analogs
such as 10, 14, and 18 showed the most promise as non-aromatic
A-ring replacements. Based on these results, we embarked on a
study to investigate the effect of other substituents in the 4-pos-
tion of the aminomethyl piperidine A-ring and to fully explore tar-
gets containing a spirocyclopropyl piperidine A-ring.
The 4-substituted piperidine intermediates were readily pre-
pared from commercially available cyanopiperidine 19 (Scheme
2). Alkyl substituted piperidine intermediates were prepared by
treating 19 with LDA and subsequent addition of an alkyl halide.
The cyano group was reduced with Raney Nickel and capped as
the trifluoromethyl sulfonamide. Removal of the Boc group pro-
vided piperidine intermediates 20–23. Quenching the anion of 19
with acetaldehyde or paraformaldehyde provided access to 24
and 26, respectively. Hydroxymethyl-substituted 25 was then con-
verted to the tosylate which was displaced with sodium azide. The
azide functionality was reduced with Pd/C and the resulting amine
was converted to the trifluoromethyl sulfonamide 27. Removal of
the Boc group provided piperidine 28. Cyano-substituted 27 was
further elaborated by hydrogenating the cyano group. The primary
amine was protected as the allyl carbamate and Boc removal pro-
vided 29.
The effect of the A-ring piperidine on CB2 binding and CYP 2C9
inhibition is shown in Table 2. In general, alkyl and fluoro substitu-
tion was well tolerated (18, 30–32, and 35) as was hydroxyalkyl
substitution (33 and 34). Cyano substitution resulted in a loss of
CB2 receptor affinity (36) as did phenyl substitution (39). The Alloc
protected amine 37 had much diminished CB2 binding potency.
Removal of the Alloc group provided 38 which regained some
binding potency relative to 37. CYP 2C9 inhibition was improved
relative to 1 with unsubstituted piperidine 10. However, substitu-
tion that maintained or improved binding potency in general led to
a worsening of the CYP 2C9 profile.
7b
6900
N
H
H
N
8b
234
43
1615
N
H
H
N
9b
>10, 000
H
N
N
H
10b
11b
5.5
>5000
900
N
H
N
180
N
H
H
N
12b
92
2500
N
N
13b
14
750
1.7
>10,000
2500
N
N
N
H
H
15c
1.5
2200
N
H
N
N
H
N
H
16d
17e
7.0
8.0
2500
560
( )
Me
Me
N
H
N
( )
Me
N
H
18
0.39
4100
N
a
Individual data points for Ki determination were carried out in triplicate, in two
separate assays.
b
Mono N-Boc diamine commercially available.
See Ref. 13 for synthesis of piperidine A-ring.
See Ref. 14 for synthesis of piperidine A-ring.
See Ref. 15 for synthesis of piperidine A-ring.
c
d
e
Synthesis of spirocyclopropyl piperidine began with N-Boc-pip-
eridone 40 (Scheme 3).16 Wittig olefination followed by cycloprop-
anation provided the ester that was saponified to yield carboxylic
acid 41. The acid was then converted to an acyl azide which under-
went a Curtius rearrangement in the presence of trichloroethanol
to produce the amine protected as the trichloroethyl carbamate
(Troc). Removal of the Boc group yielded amine 42. Treatment of
the amine with 4-chlorophenylsulfonyl chloride and removal of
the Troc group provided a racemic amine which was separated into
pure enantiomers 43 and 44 by chiral HPLC. The two enantiomers

610
E. J. Gilbert et al. / Bioorg. Med. Chem. Lett. 20 (2010) 608–611
R
CN
R
SO2CF3
SO2CF3
a, b, c, d
N
e, b, c, d
N
H
N
HN
H
Boc
CN
HN
24
19
20 R = Me
21 R = Et
22 R = Pr
23 R = iPr
Me
OH
R =
f
HO
AllocHN
HN
CN
SO2CF3 b, j, d
g, h, i, c
N
NHSO2CF3
H
N
N
R
R
d
29
25 R = Boc
26 R = H
27 R = Boc
28 R = H
d
Scheme 2. Reagents and conditions: (a) (i) LDA, THF, À78 °C, 1 h; (ii) RX, À78 °C to rt, 18 h, 80–100%; (b) Raney Ni, H2(g) (50 psi), EtOH/(NH3/MeOH), 48 h, 86–100%; (c)
trifluoromethanesulfonic anhydride, TEA, CH2Cl2, À78 °C, 27–96%; (d) 2 N HCl/ether, rt, 12 h; (e) (i) LDA, THF, À78 °C, 1 h; (ii) acetaldehyde, À78 °C to rt, 18 h, 81%; (f) (i) LDA,
THF, À78 °C, 1 h; (ii) paraformaldehyde, À78 °C to rt, 18 h, 38%; (g) p-toluenesulfonyl chloride, TEA, CH2Cl2, 74%; (h) sodium azide, DMSO, 100 °C, 18 h, 86%; (i) H2 (45 psi), Pd/
C, EtOH, 99%; (j) allyl chloroformate, TEA, DCE, 78%.
O
f,g,h
d, e
a, b, c
Boc
Tr oc
Table 2
CO2H
N
N
N
HN
4-Substituted piperidine analogs of 10
H
Boc
Cl
41
42
40
R
Cl
SO2CF3
N
H
Cl
N
SO2CF3
S
O2
i, j, k,l,m
N
N
NH2
H
SO2
F
S
N
S
O2
SO2
O2
14 isomer B
43 isomer A
44 isomer B
F
Compound
R
Kia
CYP 2C9
IC50
M)17
(CB2, nM)11
(l
i, j, k,l,n
10
18
30
31
32
33
H
5.5
0.39
0.20
5.4
21
14
–Me
–Et
–Pr
–iPr
2
2
2
2
1
Cl
SO2Me
N
N
S
H
O2
SO2
–CH2OH
3
45 isomer A
46 isomer B
F
Me
34
5.1
1
OH
–F
–CN
–CH2NH–Alloc
–CH2NH2
–Ph
Scheme 3. Reagents and conditions: (a) methyl triphenylphosphoranylidene,
toluene, reflux, 96%; (b) trimethylsulfoxonium iodide, KOtBu, DMSO, rt, 85%; (c)
LiOH, THF, rt, 92%; (d) (i) acetone, TEA, ethylchloroformate; (ii) NaN3, H2O (91%
yield, two steps); (iii) toluene, trichloroethanol, reflux, 18 h, 56% (three steps); (e)
CH2Cl2, HCl/ether, quant.; (f) 4-chlorophenylsulfonyl chloride, TEA, CH2Cl2, 71%; (g)
Zn, AcOH, rt, 90%; (h) chiral HPLC (Chiralpak AD, 25% IPA/hexanes), 44 slower
eluting; (i) trifluoroacetic anhydride, TEA, dichloroethane, rt, 92%; (j) (i) n-BuLi, THF,
À78 °C, (ii) bis(2-fluorophenyl)disulfide, À78 °C to rt, 69%; (k) m-CPBA, dichloro-
ethane, rt, 53%; (l) LiOH(aq), THF, rt, quant.; (m) trifluoromethanesulfonic anhydride,
TEA, CH2Cl2, À78 °C to rt, 71%; (n) methanesulfonyl chloride, TEA, dichloroethane,
rt, 50%.
35b
36
37
7.6
7
1
13
23
0.3
37.5
2460
202.5
350
38
39c
a
Individual data points for Ki determination were carried out in triplicate, in two
separate assays.
b
See Ref. 18 for the synthesis of piperidine A-ring.
See Ref. 19 for the synthesis of piperidine A-ring.
c
were carried on separately to methyl sulfonamide targets 45 and
46.
Table 3
Effect of substitution on the B and C ring in the amino spirocyclopropyl piperidine
series
Comparing the binding potencies of the two isomers (45 and 46,
Table 3) establishes that isomer B is the more active isomer. Having
determined this, trifluoromethanesulfonamide 14 (Scheme 3) was
prepared from 44. The remaining entries in Table 3 are analogs of
the active piperidine isomer and reveal how substitution on the B
and C rings affects the CYP 2C9 profile and CB2 binding potency.
Overall, compounds with a spirocyclopropylpiperidine A-ring
had good to excellent binding potency (Table 3). The CYP 2C9 inhi-
bition did vary however based on the substitution pattern. A 2-pyr-
idyl C-ring seemed to improve the CYP 2C9 profile (47, 48, and 53)
although it was not always sufficient (51). Compounds 46, 48, and
53 possessed the best combination of CYP 2C9 profile and binding
potency.
Compounds with an improved CYP 2C9 profile from Tables 2
and 3 (10, 46, 48, and 53) were profiled further (Table 4). All com-
pounds showed good selectivity against CB1. The rat calcium chan-
nel affinity was modestly improved in 10 and greatly improved in
the case of 46, 48 and 53. Rat PK was good for 48 and 53, modest
for 10, and poor for 46. In the end, compound 53 showed the de-
sired improvement in CYP 2C9 inhibition and rat calcium channel
O O
S
X
N
H
R
N
S
O2
SO2
Y
a
Example
X
Y
R
Ki
CYP 2C9
IC50
M)17
(CB2, nM)11
(l
45 isomer A
46 isomer B
4-Cl
4-Cl
4-Cl
H
4-Cl
4-Cl
4-CF3
4-CF3
2-F
2-F
2-F
2-pyr
2-pyr
4-OMe
2-F
2-pyr
2-F
Me
Me
CF3
CF3
CF3
CF3
CF3
CF3
CF3
CF3
122
1.1
1.7
11
1.6
0.65
2.0
5.0
3.0
0.90
12
13
2
30
15
1
1
1
4
30
14
47
48
49
50
51
52
53
4-OCF3
4-OCF3
2-pyr
a
Individual data points for Ki determination were carried out in triplicate, in two
separate assays.

E. J. Gilbert et al. / Bioorg. Med. Chem. Lett. 20 (2010) 608–611
611
Table 4
Comparison of rat calcium channel affinity, CYP 2C9 IC50’s, and rat PK for 1, 10, 46, 48, and 53
Example
CB2 Ki (nM)
Ratio
CB1/CB2
Rat Ca+2 channel
CYP 2C9
Rapid Rat21
0–6 h AUC (lM h)
Ki (l
M)
IC50
(l
M)17
1
10
46
48
53
0.7
5.5
1.1
1.7
0.9
>2000
>1000
>2000
>1000
>1000
0.5
2
56
28
8
3.5
14
13
15
30
27
1.9
0.2
6.1
7.3
12. Hoffman, R. V. Org. Synth. 1981, 60, 121.
affinity while maintaining CB2 binding potency, rat PK and selec-
tivity versus CB1.
13. Anquetin, G.; Rouquayrol, M.; Mahmoudi, N.; Santillana-Hayat, M.; Gozalbes,
R.; Greiner, J.; Farhati, K.; Derouin, F.; Guedj, R.; Vierling, P. Bioorg. Med. Chem.
Lett. 2004, 14, 2773.
14. Guo, Z.; Orth, P.; Zhu, Z.; Mazzola, R. D.; Chan, T. Y.; Vaccaro, H. A.; McKittrick,
B.; Kozlowski, J. A.; Lavey, B. J.; Zhou, G.; Paliwal, S.; Wong, S.-C.; Shih, N.-Y.;
Ting, P. C.; Rosner, K. E.; Shipps, G. W., Jr.; Siddiqui, M. A.; Belanger, D. B.; Dai,
C.; Li, D.; Girijavallabhan, V. M.; Popovici-Muller, J.; Yu, W.; Zhao, L. 2005, WO
2005121130.
15. Blackaby, W.; Duggan, M. E.; Hallett, D.; Hartman, G. D.; Jennings, A. S.; Leister,
W. H.; Lewis, R. T.; Lindsley, C. W.; Naylor, E.; Street, L. J.; Wang, Y.; Wisnoski,
D. D.; Wolkenberg, S. E.; Zhao, Z. 2005, WO 2005094514.
16. Shankar, B. B.; Gilbert, E.; Rizvi, R. K.; Huang, C.; Kozlowski, J. A.; McCombie, S.;
Shih, N.-Y. 2006, WO 2006002133.
The pharmacology of 1 and 53 were profiled in the b-Arrestin
functional receptor assay.20 This assay has been shown to accu-
rately characterize the pharmacology of cannabinoid CB2 ligands.
In this assay, 1 and 53 were shown to be inverse agonists.
In conclusion, we have discovered a CB2 inverse agonist in 53
that is structurally distinct from 1 that has an improved profile
with respect to CYP 2C9 inhibition and rat calcium channel
affinity.22
17. For experimental details of the CYP 2C9 assay see: Nomeir, A. A.; Palamanda, J.
R.; Favreau, L. In Methods in Pharmacology and Toxicology. Optimization in Drug
Discovery: In Vitro Methods; Yan, Z., Caldwell, G. W., Eds.; Humana Press:
Totowa, NJ, 2004; pp 245–262.
18. Barrow, J. C.; Lindsley, C. W.; Shipe, W. D.; Yang, Z.; Wisnoski, D. 2007, WO
2007002884.
19. Stevenson, G. I.; Huscroft, I.; MacLeod, A. M.; Swain, C. J.; Cascieri, M. A.;
Chicchi, G. G.; Graham, M. I.; Harrison, T.; Kelleher, F. J.; Kurtz, M.;
Ladduwahetty, T.; Merchant, K. J.; Metzger, J. M.; MacIntyre, D. E.; Sadowski,
S.; Sohal, B.; Owens, A. P. J. Med. Chem. 1998, 41, 4623.
References and notes
1. Huffman, J. W. Mini-Rev. Med. Chem. 2005, 5, 641.
2. Svízenská, I.; Dubovy´, P.; Sulcová, A. Pharmacol. Biochem. Behav. 2008, 90, 501.
3. Matsuda, L. A.; Lolait, S. J.; Brownstein, M. J.; Young, A. C.; Bonner, T. I. Nature
1990, 346, 561.
4. Munro, S.; Thomas, K. L.; Abu-Shaar, M. Nature 1993, 365, 61.
5. Hogenauer, E. K. Expert Opin. Ther. Pat. 2007, 17, 1457.
6. Van Gaal, L. F.; Rissanen, A. M.; Scheen, A. J.; Ziegler, O.; Rossner, S. Lancet 2005,
365, 1389.
20. McGuinness, D.; Malikzay, A.; Visconti, R.; Lin, K.; Bayne, M.; Monsma, F.; Lunn,
C. A. J. Biomol. Screening 2009, 14, 49.
7. After the FDA’s Endocrine and Metabolic Drugs Advisory Committee (EMDAC)
advised against the approval of Acomplia in 2007, Sanofi-Aventis withdrew its
new drug application. Although Acomplia gained approval in the Eurpopean
Union in 2006, it was later suspended due to psychiatric side effects.
8. Docagne, F.; Mestre, L.; Loria, F.; Hernangomez, M.; Correa, F.; Guaza, C. Expert
Opin. Ther. Target 2008, 12, 185.
21. Rat PK determinations were done using the sodium salt of all compounds with
the exception of 46 which was assayed in its free form. For experimental
details of the rapid rat assay see: Korfmacher, W. A.; Cox, K. A.; Ng, K. J.; Veals,
J.; Hsieh, Y.; Wainhaus, S.; Broske, L.; Prelusky, D.; Nomeir, A.; White, R. E.
Rapid Commun. Mass Spectrom. 2001, 15, 335.
22. The analytical data for compound 53 is as follows: 1H NMR (400 MHz,CDCl3)
d = 8.55–8.51 (m, 1H), 8.51–8.48 (m, 1H), 8.21–8.15 (m, 1H), 8.08 (d, J = 8.8 Hz,
1H), 7.96 (dt, J = 1.6, 7.8 Hz, 1H), 7.65–7.55 (m, 1H), 7.47 (dd, J = 4.9, 7.1 Hz,
1H), 5.15 (s, 1H), 3.72–3.54 (m, 2H), 3.23–3.09 (m, 2H), 2.57 (br s, 1H), 1.87–
1.77 (m, 1H), 1.73–1.61 (m, 1H), 1.55–1.49 (m, 1H), 1.30–1.19 (m, 1H), 0.96–
0.85 (m, 1H), 0.84–0.75 (m, 1H); ESI-MS (M+H) 624.1; HRMS (FAB): calcd for
(M+H) C20H20F6N3O7S3, 624.0368, found 624.0376.
9. Lunn, C. A.; Reich, E. P.; Fine, J. S.; Lavey, B.; Kozlowski, J. A.; Hipkin, R. W.;
Lundell, D. J.; Bober, L. Br. J. Pharmacol. 2008, 153, 226.
10. Lunn, C. A.; Reich, E.-P.; Bober, L. Expert Opin. Ther. Target 2006, 10, 653.
11. For experimental details of Ki determinations see the supplementary material
in: Lavey, B. J.; Kozlowski, J. A.; Shankar, B. B.; Spitler, J. M.; Zhou, G.; Yang, D.-
Y.; Shu, Y.; Wong, M. K. C.; Wong, S.-C.; Shih, N.-Y.; Wu, J.; McCombie, S. W.;
Rizvi, R.; Wolin, R. L.; Lunn, C. A. Bioorg. Med. Chem. Lett. 2007, 17, 3760.