Novel inhibitors of fatty acid amide hydrolase

Article abstract of DOI:10.1016/j.bmcl.2007.04.009

A class of bisarylimidazole derivatives are identified as potent inhibitors of the enzyme fatty acid amide hydrolase (FAAH). Compound 17 (IC₅₀ = 2 nM) dose-dependently (0.1-10 mg/kg, iv) potentiates the effects of exogenous anandamide (1 mg/kg,

Full text of DOI:10.1016/j.bmcl.2007.04.009

Bioorganic & Medicinal Chemistry Letters 17 (2007) 3287–3291
Novel inhibitors of fatty acid amide hydrolase
S. Y. Sit,a,* Charlie Conway,b Robert Bertekap,b Kai Xie,a Clotilde Bourin,b
Kevin Burrisc and Hongfeng Dengd
aDepartment of Chemistry, The Bristol-Myers Squibb Pharmaceutical Research Institute, 5 Research Parkway, Wallingford,
CT 06492-7660, USA
bDepartment of Neurosciences, The Bristol-Myers Squibb Pharmaceutical Research Institute, 5 Research Parkway, Wallingford,
CT 06492-7660, USA
cPalatin Technologies, Inc. 4-C Cedar Brook Drive, Cedar Brook Corporate Center, Cranbury, NJ 08512, USA
dGlaxoSmithKline, 830 Winter Street, Waltham, MA 02451, USA
Received 11 January 2007; revised 3 April 2007; accepted 4 April 2007
Available online 10 April 2007
Abstract—A class of bisarylimidazole derivatives are identified as potent inhibitors of the enzyme fatty acid amide hydrolase
(FAAH). Compound 17 (IC50 = 2 nM) dose-dependently (0.1–10 mg/kg, iv) potentiates the effects of exogenous anandamide
(1 mg/kg, iv) in a rat thermal escape test (Hargreaves test), and shows robust antinociceptive activity in animal models of persistent
(formalin test) and neuropathic (Chung model) pain. Compound 17 (20 mg/kg, iv) demonstrates activity in the formalin test that is
comparable to morphine (3 mg/kg, iv), and is dose-dependently inhibited by the CB1 antagonist SR141716A. In the Chung model,
compound 17 shows antineuropathic effects similar to high-dose (100 mg/kg) gabapentin. FAAH inhibition shows potential utility
for the clinical treatment of persistent and neuropathic pain.
ꢀ 2007 Elsevier Ltd. All rights reserved.
The cloning of cannabinoid (CB) receptors and the re-
cent discovery of endogenous agonists has resulted in
considerable interest in the cannabinoid system as a tar-
get for drug discovery.1 The cannabinoid system has
been implicated in regulating numerous physiological
and pathological conditions including pain and inflam-
mation.2 Development of agonists at central CB1 recep-
tors has been dampened by the inability to separate
potential therapeutic benefits from side-effects such as
ataxia, hypothermia, and potential for abuse.
These FAAH KO mice reportedly have increased brain
levels of fatty acid amides, including anandamide. How-
ever, these mice do not display all of the typical signs of
cannabinoid agonist activation (i.e., hypothermia, cata-
lepsy, decreased locomotor activity) suggesting that
inhibition of FAAH is not the functional equivalent of
widespread CB1 receptor activation. These FAAH KO
mice do, however, demonstrate decreased thermal pain
sensitivity that is blocked by a CB1 receptor antagonist.
Inhibitors of FAAH may provide selective prolongation
of endogenous cannabinoid activity at local sites of
release and present a potentially useful strategy for max-
imizing beneficial effects such as analgesia.6 Here we
identify a novel series of FAAH inhibitors that potenti-
ate the effects of exogenous anandamide and demon-
strate robust activity in animal models of persistent
and neuropathic pain.
Endogenous cannabinoids, like anandamide, are re-
leased from cells in a stimulus-dependent manner (event
driven) and rapidly inactivated by enzymes such as fatty
acid amide hydrolase (FAAH).3,4 Recently, mice that
lack FAAH (FAAH knock-out, KO mice) were devel-
oped by researchers at The Scripps Research Institute.5
Utilizing a high-throughput screen (HTS) on a targeted
subset of Bristol-Myers Squibb’s proprietary collection
of compounds, a class of bisarylazole derivatives emerged
as potent inhibitors of the enzyme FAAH [Table 1].
Keywords: Fatty acid amide hydrolase; FAAH; Enzyme; Inhibitor;
Bisarylimidazole; Pain; SR141716A; Anandamide; Hargreaves test;
Chung model; Formalin test; Persistent; Neuropathic; Antineuropathic;
Gabapentin; Morphine; Cannabinoid receptor.
*
Corresponding author. Tel.: +1 203 677 6678; fax: +1 203 677
5775; e-mail addresses: singyuen.sit@bms.com; kburris@palatin.
com; hongfeng.x.deng@gsk.com
The compounds chosen for the initial screen were origi-
nally designed as nonprostanoid prostacyclin mimetics,
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.04.009

3288
S. Y. Sit et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3287–3291
Table 1. Compounds identified from HTS
Table 2. Novel inhibitors of FAAH
N
#
6
n
5
6
7
IC50 (nM)
32*
Et
N
#
1
2
3
#
4
5
n
1
3
4
n
1
3
R
IC50 (nM)
670*
O(CH2)nCOOR
N
N
Me
Et
Et
R
N
(CH2)nCOOEt
7
11 7**
31*
64 25**
>10000
IC50 (nM)
6500*
8
N
#
R
Me
Et
IC50 (nM)
130 5**
9900*
Me
N
N
R
9
Me
Et
COOEt
10
11
O(CH2)nCOOR
91 2**
i-Pr
>100µM*
(**) n=2-3
(9-11, dl-mixtures)
(**replicates, *single readings)
(*) n=1
Me
N
Me
N
Me
N
O
N
N
R
N
H
NHR
O
reflecting a common arachidonic acid lineage, but were
only weakly active as inhibitors of blood platelet aggre-
gation.7 Initial in vitro findings suggested that the active
chemotype possessed very well-defined structural
requirements for expression of potent FAAH inhibition.
One of the key objectives of the study was to validate the
fundamental concept that blocking the degradation of
the endocannabinoid signaling agents such as ananda-
mide in vivo could precipitate an ‘event-triggered’ and
localized analgesic response that is qualitatively different
from a centrally administered cannabinoid agonist.8
12, R=H, 1.7μM
14, R=Me, >10μM
15, R=Ph, 10μM
13, R=Me, 970nM
Me
N
O
N
N
N
N
H
O
16, R=H, 5.3 1.6nM
17, R=F, 2 nM
R
F
Me
N
R
O
18, R=Me, 4.2 1.6nM
19, R=Et, 30nM
20, R=iPr, 250nM
21, R=Ph, 110nM
N
H
O
(18-21, dl-mixtures)
O
Me
N
N
R
O
22, R=Et, >100μM
23, R=nPr, >100μM
The active hits from the HTS were resynthesized and
their in vitro activity confirmed [Table 1]. The SAR
was established by chemical modifications of the origi-
nal hits, as depicted in Scheme 1, that resulted in novel
and potent analogs with IC50 values in the low single di-
git nM range [Table 2].9
F
Me
N
N
F
F
H
N
H
R
O
N
O
F
R =
O
O
F
25, 10nM
26, 24nM
F
Determination of FAAH activity. Homogenates of crude
membranes were prepared from H4 cells that express
transfected human FAAH (H4-FAAH cells).9 Activity
of FAAH was measured using a modification of the
method described by Omeir et al.10 IC50 values were
determined using a four-parameter logistic equation
for dose–response curves.
24, R = COOEt, 64nM
catalytic triad to produce a covalent intermediate. In-
deed, substitution at the a-position of 7 exerted a signif-
icant negative effect on the enzyme inhibitory activity, as
shown by compounds 9–11. Replacing the ester moiety
with an amide (12,13), or urea moiety (14,15), rendered
the compounds inactive. Further optimization with the
primary goal of increasing inhibitory potency showed
that the carbamate terminal group provided the most
potent compounds within this series.11 Thus, both phe-
nyl carbamates 16 and 17 showed very potent enzyme
inhibitory activity toward FAAH with IC50 values in
the low single digit nM. By way of contrast, the corre-
sponding alkyl carbamates were found to be less active
(data not shown). It is interesting to note that substitu-
tion a- to the N atom in the carbamate series (18–20)
does not affect IC50 values as much as in the ester series,
presumably reflecting the increased separation from the
carbonyl moiety. However, substituting the carbamate
N resulted in a complete loss of inhibitory activity, rein-
forcing the importance of a sterically unencumbered car-
bonyl moiety in this series (22, 23). A consequence that
is consistent with carbamate functional group participa-
tion in the inhibition of FAAH. Finally the effects of
aryl ring modification toward inhibitory activity were
explored in a single structural context. The bis(4-fluor-
Preliminary in vitro SAR work focused on utilizing alkyl
chains to replace the substituted phenyl ring. This exer-
cise revealed that the inhibitory potency is highly sensi-
tive to the chain length separating the bisarylimidazole
core and the terminal functionality (6–8). It was initially
considered that this class of inhibitors would behave like
other typical serine hydrolase inhibitors that act by way
of an electrophilic functional group interacting with the
R
O
R
N
N
Step 1
Step 2
N
NH
O
COOR'
OR"
n
R
R
N
N
O
Step 3
Step 4
N
N
N
H
COOH
n
n
Scheme 1. Reagents and conditions: (1) Alkyl aldehyde, ammonium
acetate in acetic acid at 100 ꢁC; (2) aliphatic halide, NaH in DMF at rt;
(3) NaOH–EtOH at rt; (4) (PhO)2P(O)N3, Et3N in toluene at 105 ꢁC,
followed by R00OH.

S. Y. Sit et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3287–3291
3289
ophenyl) substituted imidazole derivatives showed re-
duced activity (ꢀ5·) compared to the parent unsubsti-
tuted compounds. It is conceivable that fluoro
substitution could affect the overall recognition of the
inhibitor by FAAH in a subtle fashion.
The initial goal was to demonstrate proof of concept
in vivo for FAAH inhibition, followed by assessment
in multiple animal models of persistent and neuropathic
pain. Compound 17 (IC50 = 2 nM) from the carbamate
series was selected for in vivo testing. In the Hargreaves
thermal escape test, 17 was shown to transform an
otherwise inactive dose of exogenous anandamide into
an agent with robust, morphine-like analgesic properties
[Fig. 1].12
Figure 2. 17 Elicits morphine-like activity against persistent pain.
**p < 0.01 versus vehicle group (Dunnett’s test); iv, jugular catheter.
Briefly, compound 17 (0.1, 1, and 10 mg/kg, iv) given
15 min prior to exogenous anandamide (1 mg/kg, iv)
produced a significant dose-dependent antinociceptive
effect from 3 to 30 min. Neither anandamide nor
10 mg/kg compound 17 was active if given alone, but to-
gether their effect was comparable to the positive control
morphine sulfate (3 mg/kg, iv). This combination of two
inactive doses yielding an robust antinociceptive effect
provides a plausible, in vivo, proof of concept for inhi-
bition of FAAH.13
L5/6 spinal nerve ligation) exhibit pain-escape responses
(tactile allodynia) to light touch that is normally an
innocuous stimulus [Fig. 4]. Compound 17 (20 mg/kg,
iv) produced a significant (p < 0.01) reversal of tactile
allodynia (neuropathic pain behavior) that was compa-
rable to gabapentin (100 mg/kg, iv).15
In conclusion, novel inhibitors of the enzyme FAAH
responsible for the signal termination of the endocan-
nabinoids, like anandamide, were identified. Further
optimization yielded potent inhibitors suitable for a
proof of concept study. In a battery of in vivo whole ani-
mal pain models, the FAAH inhibitor 17 demonstrated
potentiation of the antinociceptive effect of an ineffective
dose of anandamide. High-dose exogenous anandamide
can produce strong analgesia, and potentiation of an
inactive dose is consistent with the actions of an FAAH
inhibitor. In the formalin test, compound 17 showed ro-
bust antinociception comparable to that of morphine in
Phase II (persistent pain). The specific CB1 antagonist
SR141716A blocks completely the antinociceptive effect
in the persistent phase (II); and partially in the acute
phase (I). These results support the mode of action
was via the endocannabinoid signaling pathway. The ef-
fect of 17 in the Chung model was particularly striking.
Compound 17 demonstrated significant effect in revers-
ing mechanical allodynia (neuropathic pain behavior)
comparable to the clinically active reference agent gaba-
pentin. Taken together, we have identified a novel class
of inhibitors that block the metabolic degradation of
endocannabinoid signaling agents. While it is not possi-
ble to specifically pinpoint the precise mode of action of
these compounds in vivo, the implications of blocking
FAAH are clearly demonstrated by 17 in a series of
in vivo whole animal studies.
Carbamate 17 was examined in a model of persistent
pain where the test animals were challenged with subcu-
taneous paw injection of formalin.14 The goal was to
determine the effect of 17 upon the spontaneous behav-
ior as measured by the frequency of ‘paw flinch’
responses.
In this model, 17 (20 mg/kg, iv) produced a significant
(p < 0.01) suppression of formalin-induced paw flinches
that was comparable to morphine (3 mg/kg, iv) [Fig. 2].
The CB1 antagonist SR141716A (SR, 3 and 10 mg/kg,
ip) completely reversed the effect of 17 (20 mg/kg, iv) in
the persistent pain model (Phase II) [Fig. 3b], and partly
reversed in the model of acute pain (Phase I) [Fig. 3a].
Finally, compound 17 was studied in a neuropathic pain
model (Chung model) where nerve injured rats (surgical
Additional enzyme inhibition studies showed that 17
was only weakly active in cytosolic phospholipase A2,
(cPLA2, Ki > 15 lM) thus precluding mechanisms typi-
cally associated with the cPLA2 triggered inflammation
processes.16 Furthermore, the compound itself showed
very little affinity to CB1/CB2 (Ki = 23/28 lM, respec-
tively).17 Collectively, the results from the in vitro and
in whole animal studies strongly support the relevance
Figure 1. Hargreaves test—compound 17 Potentiates the Antinoci-
ceptive Effects of Exogenous Anandamide. (n = 5–8 per group;
*p < 0.01 vs vehicle group).

3290
S. Y. Sit et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3287–3291
Figure 3. 3a, 3b CB1 antagonist shows dose-dependent reversal of 17 action in formalin test. ꢁ30 min: CB1 antagonist SR141716A given ip;
ꢁ15 min: 17 given iv; 0 min: formalin; *p < 0.05, **p < 0.01 versus vehicle (Dunnett’s test); iv, jugular catheter.
4. (a) Desarnaud, F.; Cadas, H.; Piomelli, D. J. Biol. Chem.
1995, 270, 6030; (b) Piomelli, D.; Beltramo, M.; Giuffrida,
A.; Stella, N. Neurobiol. Dis. 1998, 5, 462.
5. Cravatt, B. F.; Demarest, K.; Patricelli, M. P.; Bracey, M.
H.; Giang, D. K.; Martin, B. R.; Lichtman, A. H. Proc.
Natl. Acad. Sci. U.S.A. 2001, 98, 9371.
6. (a) Boger, D. L.; Miyauchi, H.; Du, W.; Hardouin, C.;
Fecik, R. A.; Cheng, H.; Hwang, I.; Hedrick, M. P.;
Leung, D.; Acevedo, O.; Guimaraes, C. R. W.; Jorgensen,
W. L.; Cravatt, B. F. J. Med. Chem. 2005, 48, 1849; (b)
Leung, D.; Du, W.; Hardouin, C.; Cheng, H.; Hwang, I.;
Cravatt, B. F.; Boger, D. L. Bioorg. Med. Chem. Lett.
2005, 15, 1423; (c) Guimaraes, C. R. W.; Boger, D. L.;
Jorgensen, W. L. J. Am. Chem. Soc. 2005, 127, 17377; (d)
Michaux, C.; Muccioli, G. G.; Lambert, D. M.; Wouters,
J. Bioorg. Med. Chem. Lett. 2006, 16, 4772; (e) Piomelli,
D.; Tarzia, G.; Duranti, A.; Tontini, A.; Mor, M.;
Compton, T. R.; Dasse, O.; Monaghan, E. P.; Parrott, J.
A.; Putman, D. CNS Drug Reviews 2006, 12, 21.
7. (a) Seiler, S. M.; Brassard, C. L.; Federici, M. E.; Romine,
J.; Meanwell, N. A. Prostaglandins 1997, 53, 21; (b)
Figure 4. Compound 17 reverses neuropathic pain behavior like high-
dose gabapentin. *p < 0.05, **p < 0.01 versus baseline ligated (Dun-
nett’s test); iv, tail vein.
Meanwell, N. A.; Romine, J. L.; Seiler, S. M. Drugs Fut.
1994, 19, 361; (c) Meanwell, N. A.; Rosenfeld, M. J.;
Trehan, A. K.; Romine, J. L.; Kim, Wright J. J.; Brassard,
C. L.; Buchanan, J. O.; Federici, M. E.; Fleming, J. S.;
of blocking the endocannabinoid signaling by 17 in the
modulation of pain responses.
Gamberdella, M.; Seiler, S. M. J. Med. Chem. 1992, 35,
3498.
References and notes
8. Fox, A.; Bevan, S. Expert Opin. Invest. Drugs 2005, 14,
695.
1. For general reviews: (a) Pacher, P.; Batkai, S.; Kunos, G.
Pharmacol. Rev. 2006, 58, 389; (b) Cravatt, B. F.;
Lichtman, A. H. Curr. Opin. Chem. Biol. 2003, 7, 469;
(c) Boger, D. L.; Henriksen, S. J.; Cravatt, B. F. Curr.
Pharm. Des. 1998, 44, 303.
2. (a) Pertwee, R. G. Prog. Neurobiol. 2001, 63, 569; (b)
Cravatt, B. F.; Lichtman, A. H. J. Neurobiol. 2004, 61,
149; (c) Lu, D.; Vemuri, V.; Kiran, D.; Richard, I., Jr.;
Makriyannis, A. Curr. Top. Med. Chem. 2006, 6, 1401; (d)
Guindon, J.; LoVerme, J.; De Lean, A.; Piomelli, D.;
Beaulieu, P. Eur. J. Pharmacol. 2006, 550, 68.
3. (a) Elphick, M. R.; Egertova, M. Philos. Trans. R. Soc.
London Ser. B 2001, 356, 381; (b) Cravatt, B. F.;
Lichtman, A. H. Chem. Phys. Lipids 2002, 121, 135; (c)
Walker, J. M.; Huang, S. M.; Strangman, N. M.; Tsou,
K.; Sanudo-Pena, M. C. Proc. Natl. Acad. Sci. U.S.A.
1999, 96, 12198.
9. Sit, S. Y.; Xie, K. U.S. Patent 6,562,846, 2003; preliminary
report disclosed as a Poster Presentation, Sit et al. 39th
National Organic Chemistry symposium, Number D30,
Salt Lake City, Utah, June 12–16, 2005; American
Chemical Society, Division of Organic Chemistry.
10. Omeir, R. L.; Chin, S.; Hong, Y.; Ahern, D. G.; Deutsch,
D. G. Life Sci. 1995, 56, 1999.
11. Carbamate functionality is uniquely present in many
FAAH inhibitors, see reference 1(b), and Alexander, J.
P.; Cravatt, B. F. Chem. Biol. (Cambridge, MA, U.S.)
2005, 12, 1179.
12. Hargreaves, K.; Dubner, R.; Brown, F.; Flores, C.; Joris,
J. Pain 1988, 32, 77.
13. Comparable, in vivo, analgesic properties have been
demonstrated previously, for example, see Refs. 6a–e.
14. Hunskaar, S.; Hole, K. Pain 1987, 30, 103.

S. Y. Sit et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3287–3291
3291
15. (a) Honore, P. Drug Dev. Res. 2006, 67, 302; (b) Joshi, S.
K.; Honore, P. Expert Opin. Drug Discovery 2006, 14, 323.
16. Lee, K. L.; Foley, M. A.; Chen, L.; Behnke, M. L.;
Lovering, F. E.; Kirincich, S. J.; Wang, W.; Shim, J.; Tam,
S.; Shen, M. W. H.; Khor, S.; Xu, X.; Goodwin, D. G.;
Ramarao, M. K.; Nickerson-Nutter, C.; Donahue, F.; Ku,
M. S.; Clark, J. D.; McKew, J. C. J. Med. Chem. 2007, 50,
1380.
17. CB1/CB2 binding competition of [3H]-CP55,940 in HEK-
CB1 and CHO-CB2 preparations.