Synthesis and evaluation of 2-amido-3-carboxamide thiophene CB2 receptor agonists for pain management

Article abstract of DOI:10.1016/j.bmcl.2012.01.121

SAR studies on a series of thiophene amide derivatives provided CB ₂ receptor agonists. The activity of the compounds was characterized by radioligand binding determination, multiple functional assays, ADME, and pharmacokinetic studies. A repre

Full text of DOI:10.1016/j.bmcl.2012.01.121

Bioorganic & Medicinal Chemistry Letters 22 (2012) 2604–2608
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and evaluation of 2-amido-3-carboxamide thiophene CB2
receptor agonists for pain management
⇑
Derek W. Nelson , Jennifer M. Frost, Karin R. Tietje, Alan S. Florjancic, Keith Ryther, William A. Carroll,
Michael J. Dart, Anthony V. Daza, Bradley A. Hooker, George K. Grayson, Yihong Fan, Tiffany R. Garrison,
Odile F. El-Kouhen, Betty Yao, Madhavi Pai, Prasant Chandran, Chang Zhu, Gin C. Hsieh, Michael D. Meyer
Neurological Diseases Research, Global Pharmaceutical R&D, Abbott, 100 Abbott Park Road, Abbott Park, IL 60064, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
SAR studies on a series of thiophene amide derivatives provided CB2 receptor agonists. The activity of the
compounds was characterized by radioligand binding determination, multiple functional assays, ADME,
and pharmacokinetic studies. A representative compound with selectivity for CB2 over CB1 effectively
produced analgesia in behavioral models of neuropathic, inflammatory, and postsurgical pain. Control
experiments using a CB2 antagonist demonstrated the efficacy in the pain models resulted from CB2
agonism.
Received 23 November 2011
Revised 26 January 2012
Accepted 30 January 2012
Available online 4 February 2012
Keywords:
CB2 agonists
Pain management
Ó 2012 Elsevier Ltd. All rights reserved.
The cannabinoid G-protein coupled receptor (GPCR) family con-
sists of two well-characterized subtypes.1 Cannabinoid 1 (CB1)
receptors exist primarily in the central nervous system (CNS) and
also in peripheral tissues.2 Activation of CB1 receptors produces a
variety of effects, including sedation, euphoria, and appetite stimu-
lation.3 Cannabinoid 2 (CB2) receptors are expressed peripherally in
tissues associated with the immune system.4 Cannabinoid recep-
tors couple primarily to Gi/o proteins and activation inhibits cyclic
adenosine monophosphate (cAMP) production,5 but the mecha-
nism and site of action in pain states remain under investigation.6
Activation of CB1 or CB2 receptors can produce analgesia,7 but ago-
nists that lack selectivity for CB2 over CB1 cause the undesirable ef-
appropriate for drug-like molecules. In addition, many of the initial
nonpolar compounds lacked selectivity for CB2 over CB1 which lim-
ited their utility in evaluating CB2 as a target for pain therapy.
The present SAR studies around the thiophene bisamide core
incorporated polar functionality in the periphery of the structure
to increase solubility and address instability toward oxidative
metabolism. The general synthetic scheme for the synthesis of
EtO
R1
R2
O
O
O
morpholine
N
+
R1
+
S
C
O
fects associated with
D D
9-tetrahydrocannabinol ( 9-THC).8 Abbott
EtOH, Δ
NH2
R2
S
initiated a program to pursue CB2-selective agonists for pain man-
agement,9 and this report describes advances in one series of CB2
agonists.
1
O
HO
EtO
O
D
9-THC
O
Cl
R3
Selective CB2 agonists have been reported, including
KOH
R1
R1
R2
H
H
mimetics, indole derivatives, and compounds from several other
structural classes.10 A high throughput-screen (HTS) of the Abbott
compound collection using a single point functional assay identi-
fied several thiophene bisamide compounds (4) as CB2 agonists.
Structure–activity relationship (SAR) studies around the screening
hits revealed trends between CB2 agonism and the peripheral func-
tionality of the thiophene bisamide core.11 Lipophilic hydrocarbon
moieties increased ligand affinity and function at CB2 receptors,
but the resulting compounds lacked physicochemical properties
N
N
EtOH/ H2O, RT
R3
NEt3, THF, RT
R2
R3
S
S
O
O
3
2
R5
R5
N
R4 NH
R4
O
R1
H
N
EDAC, HOBt, NEt3
DMF, RT
R3
R2
S
O
4
⇑
Corresponding author.
E-mail address: derek.nelson@abbott.com (D.W. Nelson).
Scheme 1. Synthetic route for preparation of 4.
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2012.01.121

D. W. Nelson et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2604–2608
2605
the thiophene bisamides is shown in Scheme 1. The Gewald reac-
tion12 (Scheme 1, Step 1) afforded the thiophene core with substi-
tution appropriate for further functionalization. Standard acylation
procedures using acid chlorides generated the C(2) amido group.
Saponification, followed by amide formation using carbodiimide
coupling reagents, installed the C(3) carboxamide.
Compounds were evaluated using a panel of in vitro assays.
Competitive radioligand binding assays measured the displace-
ment of [3H]CP-55,940 from membrane preparations obtained
from HEK 293 cells stably transfected with recombinant human
(h) or rat (r) CB2 receptors and CHO cells stably transfected with
human or rat CB1 receptors. Two different assays evaluated func-
tional activity. Fluorescence imaging plate reader (FLIPR) technol-
ogy measured the evoked response in HEK 293 cells stably
expressing the chimeric Gaq/o protein and the appropriate CB
receptor. In the FLIPR assays, net peak responses were compared
to peak responses from CP-55,940. Adenylyl cyclase (cAMP) assays
complemented the FLIPR assays. The cAMP assays used HEK 293
Table 1
Radioligand binding data for selected thiophenesa
b
4
hCB2 binding pKib
rCB2 binding pKib
hCB1 binding pKib
rCB1 binding pKi
R
O
H
N
O
S
O
H
N
a
b
c
6.67
7.68
6.92
6.53
<5
H
N
c
7.47
5.81c
<5
6.52c
H
6.58c
7.33
<5
N
H3CO
F
N
d
e
8.25
7.04
5.47
<5
7.18
<5
F
O
N
H
N
O
H
N
X
S
O
f
g
h
X = –CH2–
–NH–
–N(CH3)–
8.71
<6
8.45
5.72
<5
7.33
7.04c
8.23c
H
N
O
H
N
O
R'
S
O
i
j
7.04c
6.98
8.36c
6.57
<5
<5
6.72c
6.28c
k
l
7.47
7.70c
6.31c
<5
<5
O
6.59c
F
m
6.98
6.48
<5
F3C
a
b
c
Number of determinations >2 (duplicate of duplicate).
Error (+SEM <0.10).
Error (0.10 > +SEM >0.27).

2606
D. W. Nelson et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2604–2608
cells stably expressing the appropriate CB receptor in the presence
of a fixed concentration of forskolin. ELISA methods measured the
cAMP levels. Receptor activation by test ligands is reported as a
percent response of the control CP-55,940.
agonist function at the human receptor, but the cyclase data indi-
cate reduced selectivity for the CB2 receptor. Compound 4f pro-
duced partial efficacy at hCB2 in both the FLIPR and cyclase
assays despite strong ligand affinity. The methylated tetrahydro-
pyridine derivative 4h did not function as an agonist in the FLIPR
assay. The steric bulk of the tetramethylcyclopropyl (4i) and ada-
mantyl (4k) derivatives provided agonist potency and efficacy in
human CB2 FLIPR and cyclase assays, but neither compound main-
tained agonist function in the rat CB2 cyclase assay.
Table 1 contains the radioligand binding data for a series of
compounds with structural variation in the C(2) and C(3) amide
groups as well as in the ring fused to the thiophene core. Small
changes in the alkyl group of the C(3) amide impacted CB2 ligand
affinity. The propyl amide provided greater potency (4a vs 4b).
The CB2 binding site tolerated both secondary and cyclic tertiary
amides (Examples 4d, 4e). Heteroatoms were tolerated in the
C(3) moiety, but ligand affinity decreased as polarity increased
(4c, 4e). Discrepancies occurred between affinity at the human
and rat CB2 receptors (4d, 4e), and selectivity for CB2 over CB1 var-
ied upon modification of the C(3) amide (4b, 4d). Compound 4b
demonstrated acceptable ligand affinity at both human and rat
CB2 with selectivity over CB1 ranging from ꢀ10-50x.
The in vitro profile of compound 4b describes a potent CB2 ago-
nist with selectivity over CB1 ranging from 10–1000. The data are
consistent between binding and functional assays, and minor dis-
crepancies exist between species. The ADME data for 4b were ob-
tained. Plasma protein binding for this compound was high (97%
rat, 98% human), but many of the more lipophilic derivatives such
as 4g and 4k bound to plasma proteins more strongly (>99%). Met-
abolic stability of 4b, as measured by incubation of the compound
with rat microsomes, was below optimal levels with 9% and 5%
remaining after 30 min for human and rat, respectively. However,
when 4b advanced to pharmacokinetic studies in rats, the com-
pound exhibited excellent bioavailability from intraperitoneal
administration (96% F) with a half-life of 6.7 h (Fig. 1). The com-
pound afforded modest oral bioavailability (11% F). The half-life
from oral administration was shorter (1.7 h) and comparable to
that from intravenous administration. The potency, selectivity,
and high level of bioavailability upon intraperitoneal administra-
tion allowed compound 4b to advance to behavioral pain models.
The compound was tested in two different models of neuro-
pathic pain, and significant efficacy was observed in each model.
In the spinal nerve ligation (Chung) model,13 intraperitoneal
administration of 4b two weeks post-surgery and 30 min prior to
testing produced a dose-related reduction in mechanical allodynia
in rats (Fig. 2). The reduction in paw withdrawal threshold reached
Additional trends in binding and selectivity emerged for com-
pounds with varied ring systems fused to C(4)/C(5) of the thio-
phene. The tetrahydrobenzothiophene core (4f) provided the
greatest CB2 ligand affinity, although a substantial discrepancy in
selectivity existed between human and rat. The tetrahydropyri-
dine-fused thiophene derivative (4g) did not bind to the CB2 recep-
tor. The methylated derivative (4h) regained ligand affinity at CB2
with poor CB1 binding. The fused thienodihydropyran core pro-
vided a balance of ligand affinity, selectivity, and polarity.
Table 1 also contains radioligand binding data for a series of
compounds bearing modified acyl substituents in the C(2) amido
group. Examples 4i–k illustrate that bulky hydrocarbon moieties
can afforded high levels of ligand affinity. Selectivity for CB2 over
CB1 varied significantly, and the ratio of CB1/CB2 binding was high-
er for the human receptors. The decrease in the ligand affinity of
oxaadamantane derivative 4l compared to its hydrocarbon coun-
terpart 4k illustrates the sensitivity of the binding site to the pres-
ence of heteroatoms. Fluorinated benzamide 4m bound to the CB2
receptors with affinities similar to analogs containing rigid polycy-
clic groups, suggesting that the bulky hydrocarbon moieties could
be replaced.
statistical significance at a dose of 10
response of 60% at 30
lmol/kg ip with a maximum
l
mol/kg. Similar efficacy was observed in the
chronic constriction injury (Bennett) model14 of neuropathic pain.
With ip administration 30 min prior to testing, compound 4b re-
duced mechanical allodynia with a maximum response of 74% of
Evaluation of the function of several potent cannabinoid ligands
from Table 1 allowed identification of candidates for advancement
to in vivo studies. The FLIPR assay using recombinant human CB2
receptor established whether a compound functioned as an ago-
nist. The full complement of FLIPR assays was not utilized due to
technical difficulties, but rat CB1 FLIPR data provided a quick and
reliable assessment of selectivity. The lower throughput cyclase as-
says afforded more rigorous measurement of function and selectiv-
ity. Functional data for selected thiophene bisamides are listed in
Table 2. Compound 4b functioned as a full agonist in the FLIPR
and cyclase assays at both the CB2 and CB1 receptors across species.
The functional selectivity parallels the binding selectivity. The
amide derived from difluoroazetine (4d) provided potency and full
30 lmol/kg. Compound 4b was also evaluated in the skin incision
model15 of postoperative pain. In these experiments the compound
was administered 2 h after the surgical insult and 30 min prior to
testing. The CB2 agonist produced a dose-related reduction in
mechanical allodynia manifested by a 74% reduction in paw with-
drawal threshold at 30 lmol/kg ip similar to that observed in the
models of neuropathic pain.
Finally, compound 4b was evaluated in the Complete Freund’s
Adjuvant (CFA) model16 with oral administration. Despite the sub-
stantial difference between intraperitoneal and oral bioavailability,
4b produced a dose-related related reduction in thermal hyperal-
gesia when administered 48 h after injection and 90 min before
testing. The determination of the paw withdrawal threshold in this
Table 2
Functional data for selected thiophenesa
b
b
b
b
b
b
4
hCB2 FLIPR pIC50 (%
rCB1 FLIPR pIC50 (%
hCB2 cAMP p IC50 (%
hCB1 cAMP pIC50 (%
rCB2 cAMP pIC50 (%
rCB1 cAMP pIC50 (%
max)d
max)d
max)d
max)d
max)d
max)d
b
d
f
h
i
7.15 (88%)
8.25 (124%)
7.08 (42%)
<5
5.81 (79%)
8.71c (97%)
9.70(105%)
8.01c (57%)
5.46c(108%)
6.26 (110%)
8.03c (72%)
8.02c (53%)
5.85 (67%)
6.64c (131%)
7.28 (78%)
9.00 (88%)
7.95 c(86%)
<5
<5
k
7.37c(68%)
a
b
c
Number of determinations >2 (duplicate of duplicate).
Error (+SEM <0.10).
Error (0.10 > +SEM >0.23).
d
Error +15% (SD).

D. W. Nelson et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2604–2608
2607
1000
100
10
IV
IP
Oral
1
0
2
4
6
8
Hours after dose
ro u te V β (L /k g ) C lp (L /h r.kg ) C max (n g /m L ) T m a x (h r) A U C (μ g .h r/L ) t1 /2 (h r) F (% )
8 .7
3 .8
3 2 0
1 .5
6 .7
IV
IP
1 5 0
3 9
0 .6
1 .2
1 0 2 0
9 6
1 1
1 2 3
1 .6
P O
A n im als a d m in iste re d d o ses o f 3 µ m o l/kg IV , 1 0 µ m ol/kg IP o r P O in 10 %
D M S O /9 0 % P E G .
Figure 1. Pharmacokinetic profile of 4b in rats.
reduction in thermal hyperalgesia reached statistical significance
at a dose of 3 mol/kg and reached 74% at a dose of 10 mol/kg.
l
l
The efficacy of 4b in the CFA model results from agonism of the
CB2 receptor as demonstrated by a set of control experiments in
which 4b and the CB2 antagonist SR-14452817 were administered
individually and combined (Fig. 3). Antagonist SR-144528 inde-
pendently produced no significant reversal of the thermal hyperal-
gesia with ip administration of a dose of 10 lmol/kg. Compound 4b
reduced the thermal hyperalgesia by 91% which is consistent with
the results in the full dose–response study. When 4b was coadmin-
istered with SR-144528, the analgesic effects of 4b were effectively
12
91%
10
8
24%
12%
6
4
Figure 2. Dose–response curve for the spinal nerve ligation model of neuropathic
pain.
series of experiments was conducted at 90 min post compound
administration to coincide with the Cmax identified from the PK
experiment (Fig. 1). Efficacy in the CFA model with oral administra-
tion reached a statistically significant level of 58% at a dose of
2
Right (CFA injected)
Left (control)
0
Vehicle 1 +
Vehicle 2
Vehicle 1 +
Vehicle 2 Compound 4b Compound 4b
SR144528
SR144528
+
100 lmol/kg. The dose response curve for oral administration of
4b in the CFA model represents a significant decrease in potency
compared to the curve obtained using ip administration in the
same model. With ip administration of 4b in the CFA model, the
Figure 3. Blocking experiment demonstrating the effects of 4b in the CFA model are
blocked by CB2 antagonist SR144528.

2608
D. W. Nelson et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2604–2608
blocked and only 24% reduction of the thermal hyperlagesia
occurred.
Supplementary data
The results summarized in Figure 3 indicate that the analgesic
activity observed for 4b result from CB2 agonism. The selectivity
profile (Tables 1 and 2) indicates 4b functions as a full agonist of
the CB1 receptor, although less potent at CB1 than at CB2. Agonists
of the CB1 receptor produce behaviors such as sedation and de-
creased coordination that could complicate the interpretation of
the results from the behavioral pain models. Compound 4b was
evaluated using the locomotor assay to determine if the analgesic
effects could be separated from CB1-mediated side effects. In the
locomotor assay, the compound was administered intraperitone-
ally 30 min prior to testing. A dose-related decrease in horizontal
locomotor activity that reached statistical significance (83%) at a
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2012.01.121.
References and notes
1. Lambert, D. M.; Fowler, C. J. J. Med. Chem. 2005, 48, 5059.
2. Matsuda, L. A.; Lolait, S. J.; Brownstein, M. J.; Yound, A. C.; Bonner, T. I. Nature
1990, 346, 561.
3. Compton, D. R.; Rice, K. C.; De Costa, B. R.; Razdan, R. K.; Melvin, L. S.; Johnson,
M. R.; Martin, B. R. J. Pharmacol. Exp. Ther. 1993, 5, 218.
4. Munro, S.; Thomas, K. L.; Abu-Shaar, M. Nature 1993, 365, 61.
5. (a) Howlett, A. C. Annu. Rev. Pharmacol. Toxicol. 1995, 35, 607; (b) Felder, C. C.;
Joyce, K. E.; Briley, E. M.; Mansouri, J.; Mackie, K.; Blond, O.; Lai, Y.; Ma, A. L.;
Mitchell, R. L. Mol. Pharmacal. 1995, 48, 443.
6. Hsieh, G. C.; Pai, M.; Chandran, P.; Hooker, B. A.; Zhu, C. Z.; Salyers, A. K.;
Wensink, E. J.; Zhan, C. C.; Carroll, W. A.; Dart, M. J.; Yao, B. B.; Honore, P.;
Meyer, M. D. Br. J. Pharmacol 2011, 162, 428.
7. Chapman, V.; Finn, D. P. Rev. Analg. 2003, 7, 25.
8. Malan, T. P., Jr.; Ibrahim, M. M.; Lai, J.; Vanderah, T. W.; Makriyannis, A.;
Porreca, F. Curr. Opin. Pharmacol. 2003, 3, 62.
9. (a) Frost, J. M.; Dart, M. J.; Tietje, K. R.; Garrison, T. R.; Grayson, G. K.; Daza, A. V.;
El-Kouhen, O. F.; Miller, L. N.; Li, L.; Yao, B. B.; Hsieh, G. C.; Pai, M.; Zhu, C. Z.;
Chandran, P.; Meyer, M. D. J. Med. Chem. 2008, 51, 1904; (b) Yao, B. B.; Hsieh, G.;
Daza, A. V.; Fan, Y.; Grayson, G. K.; Garrison, T. R.; El-Kouhen, O.; Hooker, B. A.;
Pai, M.; Wensink, E. J.; Salyers, A. K.; Chandran, P.; Zhu, C. Z.; Zhong, C.; Ryther,
K.; Gallagher, M. E.; Chin, C.-L.; Tovcimak, A. E.; Hradil, V. P.; Fox, G. B.; Dart, M.
J.; Honore, P.; Meyer, M. D. J. Pharmacol. Exp. Ther. 2009, 328, 141.
10. Han, S.; Thatte, J.; Jones, R. M. Annual Reports in Medicinal Chemistry, Volume
44 (2009) American Chemical Society, Chapter 11, pp 227-246.
11. Preliminary results disclosed previously: Nelson, D. W.; Gallagher, M. E.;
Ryther, K.; Frost, J. M.; Florjancic, A. S.; Latshaw, S. P.; Tietje, K. R.; Carroll, W.
A.; Dart, M. J.; Daza, A. V.; Grayson, G. K.; Fan, Y.; Garrison, T. R.; El-Kouhen, O.
F.; Yao, B. B.; Zhu, C. Z.; Pai, M.; Chandran, P.: Salyers, A. K.; Hsieh, G. C.;
Honore, P.; Wetter, J. M.; Marsh, K. C.; Meyer, M. D. Thiophene bisamides as
potent and selective CB2 receptor agonists for pain management, 31st National
Medicinal Chemistry Symposium, Pittsburgh, PA 2008.
dose of 100 lmol/kg was observed. The significant decrease in
activity occurred at a dose ꢀ30 fold higher than the approximate
EC(50) determined for 4b in the CFA model using ip dosing.
In conclusion, several examples of a new class of CB2 agonists
have been synthesized and evaluated. Based on the radioligand
binding and functional profile, one example (4b) from this series
was studied extensively. The compound binds well to the CB2
receptor and functions as a full agonist. The selectivity of 4b for
CB2 over CB1 was modest and depended on the nature of the assay.
The pharmacokinetic profile indicated excellent bioavailablity with
ip administration but limited bioavailability upon oral administra-
tion. The compound effectively reduced mechanical allodynia and
thermal hyperalgesia in several different behavioral pain models.
In spite of its pharmacokinetic profile, 4b effectively produced
analgesic effects with oral administration. While compound 4b de-
creased motor coordination, the analgesic effects were blocked
effectively by a CB2 antagonist. Potential ancillary pharmacology
of 4b was assessed by evaluating the compound against the CEREP
12. Sabnis, R. W.; Rangnekar, D. W.; Sonawane, N. D. J. Heterocyclic Chem. 1999, 36,
333.
13. Kim, S. H.; Chung, J. M. Pain 1992, 50, 355.
panel, but no significant effect of >50% at 10
for any of the neuroscience targets.
lM was determined
14. (a) Joshi, S. K.; Hernandez, G.; Mikusa, J. P.; Zhu, C. Z.; Zhong, C.; Salyers, A.;
Wismer, C. T.; Chandran, P.; Decker, M. W.; Honore, P. Neuroscience 2006, 143,
587; (b) Bennett, G. J.; Xie, Y. K. Pain 1988, 33, 87.
15. Brennan, T. J.; Vandermeulen, E. P.; Gebhart, G. F. Pain 1996, 64, 493.
16. Hargreaves, K.; Dubner, R.; Brown, F.; Flores, C.; Joris, J. Pain 1988, 32, 77.
17. 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 CB2 agonists in this series of compounds have not been
optimized. Modification to enhance selectivity of CB2 over CB1, in-
crease metabolic stability, and improve pharmacokinetic profile is
required. Efforts to address the limitations of these compounds will
be disclosed in due course.