Discovery of an exceptionally potent and selective class of fatty acid amide hydrolase inhibitors enlisting proteome-wide selectivity screening: Concurrent optimization of enzyme inhibitor potency and selectivity

Article abstract of DOI:10.1016/j.bmcl.2004.12.085

The concurrent implementation of a proteome-wide serine hydrolase selectivity screen with traditional efforts to optimize fatty acid amide hydrolase (FAAH) inhibition potency led to the expedited discovery of a new class of exceptionally potent (K_i < 300 pM) and unusually selective (>100-fold selective) inhibitors. The iterative inhibitor design and evaluation with assistance of the selectivity screen served to differentiate otherwise indistinguishable inhibitors permitting the simultaneous optimization of potency and selectivity. Significantly, the simultaneous assessment of all potential competitive enzymes with the selectivity screen does not require the use of expressed or purified enzymes or a competitive substrate, no modification of the inhibitors is required, and the relative potency for competitive enzymes can be quantified (IC₅₀'s) including those that lack known substrates or function.

Full text of DOI:10.1016/j.bmcl.2004.12.085

Bioorganic & Medicinal Chemistry Letters 15 (2005) 1423–1428
Discovery of an exceptionally potent and selective class of fatty
acid amide hydrolase inhibitors enlisting proteome-wide
selectivity screening: concurrent optimization of enzyme
inhibitor potency and selectivity
Donmienne Leung, Wu Du, Christophe Hardouin, Heng Cheng, Inkyu Hwang,
Benjamin F. Cravatt and Dale L. Boger*
Department of Chemistry and the Skaggs Institute for Chemical Biology, The Scripps Research Institute,
1
0550 North Torrey Pines Road, La Jolla, CA 92037, USA
Received 9 December 2004; revised 29 December 2004; accepted 30 December 2004
Abstract—The concurrent implementation of a proteome-wide serine hydrolase selectivity screen with traditional efforts to optimize
fatty acid amide hydrolase (FAAH) inhibition potency led to the expedited discovery of a new class of exceptionally potent
(K < 300 pM) and unusually selective (>100-fold selective) inhibitors. The iterative inhibitor design and evaluation with assistance
i
of the selectivity screen served to differentiate otherwise indistinguishable inhibitors permitting the simultaneous optimization of
potency and selectivity. Significantly, the simultaneous assessment of all potential competitive enzymes with the selectivity screen
does not require the use of expressed or purified enzymes or a competitive substrate, no modification of the inhibitors is required,
and the relative potency for competitive enzymes can be quantified (IC50Õs) including those that lack known substrates or function.
ꢀ
2005 Elsevier Ltd. All rights reserved.
1
2
Anandamide and oleamide have emerged as the proto-
typical members of a class of endogenous fatty acid
O
OH
N
H
3
amides that serve as chemical messengers. Anand-
Anandamide
O
amide, which is the most recognized member of the
endogenous fatty acid ethanolamides, binds and acti-
vates the central (CB1) and peripheral (CB2) cannabi-
noid receptors as well as the vanilloid receptor (VR1)
through which it is thought to exert its analgesic and
NH
2
Oleamide
Figure 1. FAAH substrates.
4
,5
cannabinoid effects. Oleamide was found to accumu-
late in the cerebrospinal fluid under conditions of sleep
deprivation and to induce physiological sleep where it
reduced mobility, shortened the sleep induction period,
and lengthened the time spent in slow wave sleep two
9
usual catalytic mechanism (Ser-Ser-Lys triad). This un-
ique mammalian distribution, its selectively targetable
catalytic mechanism, and the ramifications of its inacti-
vation (increased endogenous levels of bioactive fatty
acid amides) have defined FAAH as a potential new
therapeutic target for pain management and sleep
2
at the expense of wakening. Fatty acid amide hydrolase
(
FAAH) is a membrane-bound enzyme that hydrolyzes
fatty acid primary amides and ethanolamides including
anandamide and oleamide regulating signaling at their
1
0
disorders.
6
–8
Serine hydrolases represent one of the largest classes of
enzymes including serine proteases, lipases, esterases,
amidases, and transacetylases. Although initial efforts
to develop FAAH inhibitors produced useful tools to
probe enzyme function, their design was expected to
result in a low FAAH selectivity. However, the assess-
sites of action (Fig. 1). It constitutes the only charac-
terized mammalian member of a unique class of serine
hydrolases (amidase signature family) that bear an un-
1
0
*
Corresponding author. Tel.: +1 858 784 7522; fax: +1 858 784
550; e-mail: boger@scripps.edu
1
1
7
ment of such inhibitors developed in our efforts
0
960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.12.085

1
424
D. Leung et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1423–1428
Table 1. a-Keto oxadiazole inhibitors of fatty acid amide hydrolase
against candidate competitive enzymes (e.g., phospholi-
pase A2, ceramidase) revealed no inhibition. In the
absence of identifiable competitive enzymes and without
a class of related mammalian enzymes to examine, the
traditional approach of counter-screening the inhibitors
against a panel of related enzymes to define their selec-
tivity was not possible. Herein, we report the use of a
(
FAAH)
O
N
O
N
R1
1
R
K
i
, lM
1
2
powerful, proteome-wide serine hydrolase screen
H
0.09 + 0.2
1
3
adapted to define FAAH selectivity that was con-
ducted in parallel with traditional efforts to optimize
FAAH inhibition resulting in the expedited discovery
0.016 ± 0.003
of a class of exceptionally potent (K < 300 pM) and
i
N
0
.003 ± 0.0002
unusually selective (>100-fold) inhibitors, 2-keto-5-(2-
pyridyl)-1,3,4-oxadiazoles. To our knowledge, this rep-
resents the first report of the implementation of such a
O
N
O
N
Ph
1
4
proteome-wide selectivity screen conducted in tandem
with traditional target optimization efforts to guide the
discovery process.
R1
1
R
K
i
, lM
0
.002 ± 0.0003
The preparation of the 2-keto-1,3,4-oxadiazoles exam-
ined herein is summarized in Scheme 1 and full details
are provided in the Supplementary data. Notably, the
alkyl lithium or Grignard additions to the heterocyclic
ester cleanly occurred once presumably benefiting from
an oxadiazole chelation with and stabilization of the
addition tetrahedral intermediate.
N
0.00029 ± 0.00003
N
0
.001 ± 0.0002
N
0.001 ± 0.0001
The trends observed with the inhibitors mirror those
previously disclosed in our studies (Table 1). Introduc-
tion of a C5 substituent on the 2-keto-1,3,4-oxadiazole
does not adversely impact activity (Ph vs H), incorpora-
tion of a H-bond acceptor heterocycle at this C5 posi-
tion significantly enhances activity (2-pyridyl vs Ph)
and follows trends predicted by the H-bond acceptor
capabilities (2-pyridyl > 2-furyl > Ph), this enhancement
by the H-bond acceptor is greatest when located imme-
diately adjacent to the 1,3,4-oxadiazole (2-pyridyl > 3-
or 4-pyridyl), an alkylphenyl side-chain series is more
potent than the corresponding oleyl-based inhibitors
and exhibits a well-defined optimal chain length
O
0
.00056 + 0.00013
O
N
O
N
Ph
R1
1
R
i
K , lM
N
0
0
.00085 ± 0.00009
.0032 ± 0.00012
N
O
N
O
N
(
essential for potent inhibition (10 -fold enhancement).
Ph(CH ) : n = 5 < 6 > 7 > 8), and the 2-keto group is
2 n
R
4
N
Biotinylated or fluorescently-tagged (rhodamine, Rh)
fluorophosphonate (FP-biotin or FP-Rh) have been
used to isolate and identify (FP-biotin) or quantitate
R
K , lM
i
Ph(CH
Ph(CH
Ph(CH
2
2
2
)
)
)
5
6
7
0.00085 ± 0.00009
0.00029 ± 0.00003
0.00077 ± 0.00016
0.00083 ± 0.00021
(
labeling. Extending this to the assessment of inhibitor
FP-Rh) proteome serine hydrolases through active site
1
2
Ph(CH2)8
1
3
selectivity requires incubation of the proteome with
FP-Rh in the presence of systematically varied concen-
trations of inhibitor followed by SDS-PAGE to detect
serine hydrolases sensitive to the inhibitor. A resulting
inhibition of FP-Rh active site labeling correlates
N
N
Ph
O
R
1
1
R
K
i
, lM
N
O
6
0 ± 2
1
3
) Et N
O
O
O
O
+
2
) TsCl/Et
3
N
H
N
18 ± 0.5
NH
2
OMe
Ar
N
H
Cl
75–94%
Ar
N
H
OMe
O
O
O
N
O
O
N
N
RMgBr or RLi
20–73%
N
with inhibitor affinity for the target and permits the
simultaneous assessment of all potential competitive en-
zymes (Fig. 2). Moreover, the parallel profiling of such
reversible inhibitors against all proteome serine hydro-
MeO
R
O
Ar
Ar
Scheme 1.

D. Leung et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1423–1428
1425
Two enzymes emerged in the screen as competitive tar-
gets for the inhibitors detailed herein: triacylglycerol
hydrolase (TGH) and an uncharacterized membrane-
bound hydrolase that lacks known substrates and func-
1
3
tion (KIAA1363). Tables 2–4 summarize key results of
the selectivity screening that highlight the simultaneous
optimization of selectivity for FAAH over these two
competitive enzymes, each of which display distinct
SAR (structure–activity relationship) profiles, which
was conducted concurrent with FAAH inhibition opti-
mization. Pertinent to the off-site targets of the second
generation FAAH inhibitors, simple trifluoromethyl ke-
tone inhibitors (e.g., 1 and 2) exhibited an intrinsic selec-
tivity that typically favored TGH by >1000-fold and
KIAA1363 by 10- to 100-fold (Table 2). Despite this
unfavorable intrinsic selectivity, the iterative identifica-
tion of inhibitors selective for FAAH proved straight-
forward with the benefit of the selectivity screen.
Representative results of the iterative development of
the FAAH inhibitors are summarized in Table 3 and
Figure 2. The inhibitor affinity for FAAH versus
KIAA1363 typically increased as the side-chain size
(length) increased thereby improving selectivity (e.g.,
Figure 2. Proteome screening for FAAH selectivity. (Left) Compar-
ison of the potencies and relative selectivities of FAAH inhibitors. IC50
values are clustered into three classes: highly potent and highly
selective a-ketoheterocycle inhibitors with phenyl side-chain (red),
moderately potent a-ketoheterocycle inhibitors with oleoyl side-chain
(
(
green) and trifluoromethyl ketone inhibitors selective for TGH
purple). (Right) Competitive profiling of FAAH inhibitors (100 nM)
with FP-Rh (100 nM) in brain and heart membrane proteome. Enzyme
targets such as FAAH, KIAA1363, TGH, and LPL are highlighted.
lases under kinetically controlled conditions can quanti-
tatively establish relative potency and selectivity factors
1
3
for each inhibitor. Significantly, the selectivity screen-
ing does not require the use of a specific substrate, no
modification of the inhibitor is required, and the relative
selectivity factors can be rapidly and quantitatively
established (IC Õs).
1
5
5
0
Tables 2 and 4), and the binding to KIAA1363 is
Table 2. Selectivity screening: IC50, lM (selectivity)
a
n
K
i
(FAAH, lM)
FAAH
KIAA
TGH
3 2 n 3
CH (CH ) COCF
6
1.2
30
10
10
6.4
1.5 (0.05)
0.4 (0.04)
0.5 (0.05)
6.6 (1)
0.002 (0.00007)
0.01 (0.001)
0.07 (0.007)
4.6 (0.7)
8
0.13 (1)
0.14
0.24
1
1
0
6
2 n 3
Ph(CH ) COCF
5
6
7
0.17
0.10 (2)
0.025
5
5
2
0.8 (0.16)
0.2 (0.04)
0.2 (0.1)
0.0005 (0.0001)
0.001 (0.0002)
0.005 (0.002)
a
i
FAAH K , provided for comparison purposes.
Table 3. Selectivity screening: IC50, lM (selectivity)
O
R
a
R
K
i
(FAAH, lM)
FAAH
KIAA
TGH
5 (1)
–CF
3
0.080
5
1 (0.20)
N
0
.10
2
>100 (>50)
9 (4)
O
N
X
X = CH
X = N (3)
0.37
0.0023
10
>100 (>10)
60 (1500)
10 (1)
1 (25)
O
N
O
0.04
X
X
X = CH
X = N (4)
0.32
0.018
5
0.2
>100 (>20)
>100 (>500)
50 (10)
100 (500)
N
O
N
X = CH
X = N (5)
0.016
0.003
0.2
0.004
>100 (>500)
4
>100 (>10 )
20 (100)
20 (5000)
a
i
FAAH K , provided for comparison purposes.

1426
D. Leung et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1423–1428
Table 4. Selectivity screening: IC50, lM (selectivity)
a
K
i
(FAAH, lM)
FAAH
KIAA
TGH
O
N
O
N
• H-bond acceptor
increases potency
and selectivity
N
Ph(CH2)
n
n
5
6
7
8
•
•
4
50 (10 )
0.00085
0.004
0.001
0.03 (8)
5
90 (10 )
0.00029 (7)
0.00077
0.00083
0.14 (140)
0.2 (330)
0.5 (625)
5
60 (10 )
3
3 (>10 )
0.0006
0.0008
Selectivity: n = 8 > 7 > 6 > 5
Potency: n = 8 < 7 < 6 > 5
O
N
O
N
Ph(CH2)
6
Ar
Ar
Ph
0.002 (6)
0.00029 (7)
0.001 (8)
0.001 (9)
0.00056
0.025
0.001
0.02
>100 (>4000)
5
90 (10 )
0.25 (10)
0.14 (140)
0.1 (5)
2
3
4
2
-Pyr
-Pyr
>100 (>5000)
>100 (>5000)
5
>100 (>10 )
-Pyr
-Furyl
0.02
0.001
0.03 (2)
0.08 (80)
a
i
FAAH K , provided for comparison purposes.
Table 5. Selectivity screening: IC50, lM (selectivity)
R
K
i
(FAAH, lM)
FAAH
KIAA
TGH
Ph(CH
2
5
) COR
–
CF
3
0.17
5
0.8 (0.16)
0.0005 (0.0001)
0.020 (0.002)
N
O
0
0
.22
10
>100 (>10)
N
N
O
4
10 (10 )
.0002
0.001
0.001 (1)
N
O
3
>100 (>10 )
N
N
0.011
0.1
0.2 (2)
N
O
N
4
50 (10 )
0.00085
0.004
0.03 (8)
Ph(CH
CF
2
6
) COR
–
3
0.10
5
0.2 (0.04)
0.0009 (0.0002)
0.003 (10)
N
N
O
N
O
4
20 (10 )
0
.00028
0.0003
4
>100 (>10 )
N
N
0.0047
0.010
0.001
0.6 (60)
N
O
N
5
90 (10 )
0.00029
0.14 (140)
2 7
Ph(CH ) COR
–
CF
3
0.025
2
0.20 (0.1)
20 (6700)
0.005 (0.002)
0.005 (16)
N
O
N
0
.00039
0.0003

D. Leung et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1423–1428
1427
Table 5 (continued)
R
K
i
(FAAH, lM)
FAAH
0.008
KIAA
TGH
N
4
>100 (>10 )
N
O
0.0075
1 (125)
N
N
5
60 (10 )
N
0.00077
Nd
0.0006
0.2 (330)
O
Ph(CH
CF
2 8
) COR
–
3
Nd
Nd
Nd
N
N
5
10 (10 )
0
.00052
0.0001
0.005 (50)
O
N
3
>100 (>10 )
N
N
0.0078
0.03
6 (200)
O
N
O
N
0.00083
0.0008
3 (3800)
0.5 (625)
completely or substantially disrupted with introduction of
an electrophilic carbonyl heterocycle (Table 3) such that
the achieved FAAH selectivity is superb (>1000-fold). Even
more impressive given the intrinsic >1000-fold TGH
selectivity, the screening revealed that the TGH potency
and selectivity decreases as the side-chain size (length)
One can anticipate that the inadvertent overestimation
of enzyme inhibitor selectivity typical of counter-screen-
ing against a panel of related enzymes can be avoided
using comparable strategies of proteome-wide selectivity
screening. This strategy does not require that the en-
zymes be recombinantly expressed or purified, no modi-
fication of the inhibitors is required, and the assay is
conducted under Ôsubstrate freeÕ conditions permitting
simultaneous identification of inhibitors for enzymes
with diverse substrate preferences (e.g., FAAH and
TGH) and even novel enzymes that lack known sub-
strates (e.g., KIAA1363).
1
5
increases (e.g., Tables 2 and 4), and that incorporation
of a second, properly positioned, weakly basic nitrogen
(
H-bond acceptor) into the heterocycle improves FAAH
affinity and selectivity relative to that of TGH (Tables 3
and 4). Notably, the FAAH selectivity and affinity
SARÕs are not identical, and it is the combined optimiza-
tion of properties that distinguishes otherwise compar-
able candidate inhibitors (multidimensional SAR).
Acknowledgments
Thus, both the selectivity and potency of the 2-keto-5-
(
ceed those of the 2-keto-5-(2-pyridyl)oxazoles (e.g., 4),
2-pyridyl)-1,3,4-oxadiazoles (e.g., 5) disclosed herein ex-
6
We gratefully acknowledge the financial support of the
NIH (DA15648, D.L.B.; DA15197 and 17259, B.F.C.)
and The Skaggs Institute for Chemical Biology.
1
which in turn are more selective than the exceptionally
1
7
potent 2-keto-oxazolopyridines (e.g., 3). These trends
are well represented in the additional series summarized
in Table 5. Each overcomes the intrinsic selectivity for
TGH or KIAA1363 observed with the simpler
a-ketoheterocycles or trifluoromethyl ketone inhibitors.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2004.12.085. Experimental details for establishing
FAAH inhibition (K Õs), FAAH selectivity (IC Õs), and
Importantly, the enhanced selectivity of the exception-
ally potent 2-keto-5-(2-pyridyl)-1,3,4-oxadiazoles (e.g.,
i
50
5
) versus that of the equipotent 2-keto-oxazolopyr-
for the preparation of the 2-keto-5-(2-pyridyl)-1,3,4-
oxadiazoles disclosed herein, and an expanded Figure
2 (Figure S1).
1
7
idines (e.g., 3) serve to differentiate the otherwise indis-
tinguishable inhibitors. Remarkable potencies and use-
ful selectivities (>100-fold) were obtained with use of
the proteomics-wide selectivity screen. Thus, a class of
a-ketoheterocycles, 2-keto-5-(2-pyridyl)-1,3,4-oxadiazo-
les (e.g., 7), was discovered in an expedited manner that
References and notes
are not only extraordinarily potent (K = 300 pM, >100-
i
1. Dervane, 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.
fold more potent than the corresponding trifluoromethyl
ketones), but that also improve the intrinsic selectivity
for FAAH >10 -fold.
6

1428
D. Leung et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1423–1428
2
. Cravatt, B. F.; Prospero-Garcia, O.; Suizdak, G.; Gilula,
N. B.; Henriksen, S. J.; Boger, D. L.; Lerner, R. A. Science
995, 268, 1506.
. Reviews: Boger, D. L.; Henriksen, S. J.; Cravatt, B. F.
Curr. Pharm. Des. 1998, 4, 303; Schmid, H. H. O.; Schmid,
P. C.; Natarajan, V. Prog. Lipid Res. 1990, 29, 1.
10. Cravatt, B. F.; Lichtman, A. H. Curr. Opin. Chem. Biol.
2003, 7, 469.
1
11. Patterson, J. E.; Ollmann, I. R.; Cravatt, B. F.; Boger, D.
L.; Wong, C.-H.; Lerner, R. A. J. Am. Chem. Soc. 1996,
118, 5938; Boger, D. L.; Sato, H.; Lerner, A. E.; Austin, B.
J.; Patterson, J. E.; Patricelli, M. P.; Cravatt, B. F. Bioorg.
Med. Chem. Lett. 1999, 9, 265.
12. Kidd, D.; Lui, Y.; Cravatt, B. F. Biochemistry 2001, 40,
4005; Liu, Y.; Patricelli, M. P.; Cravatt, B. F. Proc. Natl.
Acad. Sci. U.S.A. 1999, 96, 14694.
13. Leung, D.; Hardouin, C.; Boger, D. L.; Cravatt, B. F. Nat.
Biotechnol. 2003, 21, 687.
14. Jessani, N.; Cravatt, B. F. Curr. Opin. Chem. Biol. 2004, 8,
54.
3
4
5
. Martin, B. R.; Mechoulam, R.; Razdan, R. K. Life Sci.
1999, 65, 573.
. Axelrod, J.; Felder, C. C. Neurochem. Res. 1998, 23, 575;
Zygmunt, P. M.; Petersson, J.; Andersson, D. A.; Chuang,
H.; Sorgard, M.; Di Marzo, V.; Julius, D.; Hogestatt, E.
D. Nature 1999, 400, 452; Smart, D.; Gunthorpe, M. J.;
Jerman, J. C.; Nasir, S.; Gray, J.; Muir, A. I.; Chambers,
J. K.; Randall, A. D.; Davis, J. B. Br. J. Pharmacol. 2000,
1
29, 227.
15. An additional example of this trend may be found in Table
5.
6
7
. Cravatt, B. F.; Giang, D. K.; Mayfield, S. P.; Boger, D. L.;
Lerner, R. A.; Gilula, N. B. Nature 1996, 384, 83; Giang,
D. K.; Cravatt, B. F. Proc. Natl. Acad. Sci. U.S.A. 1997,
16. Boger, D. L.; Miyauchi, H.; Du, W.; Hardouin, C.; Fecik,
R. A.; Cheng, H.; Hwang, I.; Hedrick, M. P.; Leung, D.;
Aceveda, O.; Guimaraes, C. R. W.; Jorgensen, W. L.;
Cravatt, B. F. J. Med. Chem. 2005, 48, in press
doi:10.1021/jm049585o; Lichtman, A. H.; Leung, D.;
Shelton, C. C.; Saghatelian, A.; Hardouin, C.; Boger, D.
L.; Cravatt, B. F. J. Pharmacol. Exp. Ther. 2004, 311, 441.
17. Boger, D. L.; Sato, H.; Lerner, A. E.; Hedrick, M. P.;
Fecik, R. A.; Miyauchi, H.; Wilkie, G. D.; Austin, B. J.;
Patricelli, M. P.; Cravatt, B. F. Proc. Natl. Acad. Sci.
U.S.A. 2000, 97, 5049.
94, 2238.
. Boger, D. L.; Fecik, R. A.; Patterson, J. E.; Miyauchi, H.;
Patricelli, M. P.; Cravatt, B. F. Bioorg. Med. Chem. Lett.
2
000, 10, 2613.
. Patricelli, M. P.; Cravatt, B. F. Vitam. Horm. 2001, 62,
5.
8
9
9
. Bracey, M. H.; Hansen, M.; Masuda, K.; Stevens, R. C.;
Cravatt, B. F. Science 2002, 298, 1793; McKinney, M. K.;
Cravatt, B. F. J. Biol. Chem. 2003, 278, 37393.