Biarylpyrazole inverse agonists at the cannabinoid CB1 receptor: Importance of the C-3 carboxamide oxygen/lysine3.28(192) interaction

Article abstract of DOI:10.1021/jm060446b

The biarylpyrazole, N-(piperidin-1-yl)-5-(4-chlorophenyl)-1-(2,4- dichlorophenyl)-4-methyl-1#-pyrazole-3-carboxamide (SR141716; 1) has been shown to act as an inverse agonist/antagonist at the cannabinoid CB1 receptor. Our previous mutant cycle study sugg

Full text of DOI:10.1021/jm060446b

J. Med. Chem. 2006, 49, 5969-5987
5969
Biarylpyrazole Inverse Agonists at the Cannabinoid CB1 Receptor: Importance of the C-3
Carboxamide Oxygen/Lysine3.28(192) Interaction
Dow Hurst,† Uju Umejiego,† Diane Lynch,† Herbert Seltzman,‡ Steven Hyatt,‡ Michael Roche,‡ Sean McAllister,§
Daniel Fleischer,§ Ankur Kapur,§ Mary Abood,§ Shanping Shi,| Jannie Jones,| Deborah Lewis,| and Patricia Reggio*,†
Center for Drug Design, Chemistry and Biochemistry Department, UniVersity of North Carolina Greensboro, P.O. Box 26170,
435 New Science Building, Greensboro, North Carolina 27402-6170, Chemistry & Life Sciences, RTI International, 3040 Cornwallis Road,
Research Triangle Park, North Carolina 27709, California Pacific Medical Center Research Institute, 475 Brannan Street, Suite 220,
San Francisco, California 94107, and Pharmacology and Toxicology Department, Medical College of Georgia, Augusta, Georgia 30912
ReceiVed April 14, 2006
The biarylpyrazole, N-(piperidin-1-yl)-5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-pyrazole-3-
carboxamide (SR141716; 1) has been shown to act as an inverse agonist/antagonist at the cannabinoid CB1
receptor. Our previous mutant cycle study suggested that K3.28(192) is involved in a direct interaction with
the C-3 substituent of 1 in wild-type (WT) CB1.1 However, these results did not establish what part of the
C-3 substituent of 1 is involved in the K3.28(192) hydrogen bond, the carboxamide oxygen or the piperidine
nitrogen. Furthermore, our previous calcium channel assay results for 5-(4- chlorophenyl)-3-[(E)-2-
cyclohexylethenyl]-1-(2,4-dichlorophenyl)-4- methyl-1H-pyrazole (VCHSR; 2) (an analogue of 1 that lacks
hydrogen-bonding capability in its C-3 substituent) showed that this compound acts as a neutral antagonist,
a result that is in contrast to 1, which acts as an inverse agonist in this same assay.1 These results suggested
a relationship between biarylpyrazole interaction with K3.28(192) at CB1 and inverse agonism, but these
results were for a single pair of compounds (1 and 2). The work presented here was designed to test two
hypotheses derived from our modeling and mutant cycle results. The hypotheses are as follows: (1) it is the
carboxamide oxygen of the C-3 substituent of 1 that interacts directly with K3.28(192) and (2) the interaction
with K3.28(192) is crucial for the production of inverse agonism for biarylpyrazoles such as 1. To determine
whether the carboxamide oxygen or the piperidine nitrogen of the C-3 substituent may be the interaction
site for K3.28(192), we designed, synthesized, and evaluated a new set of analogues of 1 (3-6, Chart 1)
in which modifications only to the C-3 substituent of 1 have been made. In each case, the modifications
that were made preserved the geometry of this substituent in 1. The absence of the piperidine nitrogen was
not found to affect affinity, whereas the absence of the carboxamide oxygen resulted in a reduction in
affinity. CB1 docking studies in an inactive state model of CB1 resulted in the trend, 3,1 < 5,4 < 2 < 6
for ligand/CB1 interaction energies. This trend was consistent with the trend in WT CB1 Ki values versus
[3H]CP55,940 reported here. In calcium channel assays, all analogues with carboxamide oxygens (1, 3, and
4) were found to be inverse agonists, whereas those that lacked this group (2, 5, and 6) were found to be
neutral antagonists. Taken together, these results support the hypothesis that it is the carboxamide oxygen
of the C-3 substituent of 1 that engages in a hydrogen bond with K3.28(192) in WT CB1. Furthermore,
functional results for 1-6 support the hypothesis that the interaction of 1 with K3.28(192) may be key to
its inverse agonism.
Introduction
effects produced by CB1 agonists after intraperitoneal or oral
administration.5 The clinical efficacy of this compound (also
called rimonabant or Accomplia) for the treatment of metabolic
syndrome, obesity, and smoking addiction has catalyzed massive
research efforts as indicated by a surge of patent applications
in recent years.6 In addition to the biarylpyrazoles typified by
1, several other heterocyclic compounds, including triazoles,7,8
thiazoles,8 pyrrazolines,9 imidazoles,8,10,11 and pyridines12 have
been reported to be cannabinoid CB1 receptor antagonists/
inverse agonists.
The cannabinoid CB1 receptor (Figure 1) belongs to Class
A (rhodopsin (Rho) family) G-protein coupled receptors (GPCRs).
CB1 receptor agonists inhibit forskolin-stimulated adenylyl
cyclase by activation of a pertussis toxin-sensitive G-protein.2
In heterologous cells, CB1 receptors inhibit N-, P-, and Q-type
calcium channels and activate inwardly rectifying potassium
channels.2-4 Inhibition of calcium channels and enhancement
of inwardly rectifying potassium currents is pertussis toxin-
sensitive but, independent of cAMP inhibition, suggestive of a
direct G-protein mechanism.3
Compound 1 has been shown to act as a competitive
antagonist and inverse agonist in host cells transfected with
exogenous CB1 receptor,13 in biological preparations endog-
enously expressing CB114 in in vitro preparations15 as well as
those in vivo.16 Bouaboula and co-workers 17 reported that CHO
cells transfected with human CB1 receptor exhibit high constitu-
tive activity at both levels of MAP kinase and adenylyl cyclase.
Guanine nucleotides enhanced the binding of 1, a property of
inverse agonists. Lewis and co-workers13 demonstrated constitu-
tive activity of CB1 receptors in inhibiting Ca2+ currents that
was not due to an endogenous agonist. These investigators
The biarylpyrazole, N-(piperidin-1-yl)-5-(4-chlorophenyl)-
1-(2,4-dichlorophenyl)-4-methyl-1H-pyrazole-3-carboxamide
(SR141716; 1) antagonizes the pharmacological and behavioral
* Corresponding author. Tel: (336) 334-5333. Fax: (336) 334-5402.
E-mail: phreggio@uncg.edu.
† University of North Carolina Greensboro.
‡ RTI International.
§ California Pacific Medical Center Research Institute.
| Medical College of Georgia.
RTI International is a trademark of Research Triangle Institute.
10.1021/jm060446b CCC: $33.50 © 2006 American Chemical Society
Published on Web 09/01/2006

5970 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20
Hurst et al.
Figure 1. Amino acid sequence of the human CB1 receptor 44 is represented in helix net format here. The most highly conserved residue position
42
in each transmembrane helix across Class A GPCRs is highlighted in blue. The Ballesteros-Weinstein amino acid numbering system is used
here (Experimental Section).
reported that 1 antagonized the Ca2+ current inhibition induced
by the cannabinoid agonist, (R)-(+)-[2,3-dihydro-5-methyl-3-
[(4-morpholinyl)methyl]pyrrolo[1,2,3-de]-1,4-benzoxazin-6-yl]-
(1-naphthalenyl)methanone (WIN55212-2), in neurons heter-
ologously expressing either rat or human CB1 receptors.
Furthermore, when applied alone, 1 increased the Ca2+ current
with an EC50 of 32 nM via a pertussis toxin-sensitive pathway,
indicating that 1 can act as an inverse agonist by reversal of
tonic CB1 receptor activity.
The extended ternary complex model for GPCR activation
invokes the existence of two receptor conformational states, a
ground or inactive R state and an active R* state, which are in
equilibrium with each other.18 An agonist has higher affinity
for R*, and agonist binding is thought to shift the equilibrium
toward R*, resulting in G-protein activation with an increase
in GDP/GTP exchange. An inverse agonist has higher affinity
for R, and its binding shifts the equilibrium toward R, resulting
in a decrease in the activation of the signaling pathway. The
binding of a neutral/null antagonist is thought not to alter the
equilibrium between R and R* because the neutral antagonist
has equal affinity for both states.
Our previous combined mutation/modeling studies have
suggested that the binding site of the CB1 inverse agonist/
antagonist, 1, is within the transmembrane helix (TMH)3-4-
5-6 aromatic microdomain and involves direct aromatic stacking
interactions with F3.36(200), Y5.39(275), and W5.43(279).19
Our modeling studies have also suggested that although 1 can
engage in aromatic stacking interactions in the inactive (R) and
active (R*) states of CB1, the C-3 substituent carboxamide
oxygen of 1 can hydrogen bond with K3.28(192) only in the
CB1 inactive (R) state (Figure 2).1 We hypothesized that this
interaction may be responsible for the higher affinity of 1 for
the inactive state, rendering it an inverse agonist. (For a
discussion of the creation of the activated (R*) state model and
experimental evidence supporting its creation, please see
Experimental Section.) To test this hypothesis, a mutant
thermodynamic cycle was constructed that combined the evalu-
ation of the affinity of 1 at wild-type (WT) CB1 and the CB1
K3.28(192)A mutant with an evaluation of the WT CB1 and
K3.28(192)A affinities of an analogue of 1, 5-(4- chlorophenyl)-
3-[(E)-2-cyclohexylethenyl]-1-(2,4-dichlorophenyl)-4- methyl-
1H-pyrazole (2)), which lacks hydrogen-bonding potential in
its C-3 substituent but maintains aromatic stacking inter-
actions analogous to 1 in R and R*.1 Mutant cycle results
suggested that K3.28(192) is involved in a direct interaction
with the C-3 substituent of 1 in WT CB1 and that K3.28(192)
does not interact with 2. However, these results did not establish
what part of the C-3 substituent of 1 is involved in the K3.28-
(192) hydrogen bond because there are two potential hydrogen
bond acceptors in this substituent, the carboxamide oxygen and
the piperidine nitrogen. Furthermore, calcium channel assay
results for 2 at WT CB1 showed that this compound acts as a
neutral antagonist, a result that is in contrast to 1, which acts
as an inverse agonist in this same assay.1 These results supported
our hypothesis that there may be a relationship between
biarylpyrazole interaction with K3.28(192) at CB1 and inverse
agonism, but these results were for a single pair of compounds
(1 and 2).
The work presented here was designed to test two hypoth-
eses: (1) it is the carboxamide oxygen of the C-3 substituent
of 1 that interacts directly with K3.28(192) and (2) the
interaction with K3.28(192) is crucial for the production of
inverse agonism for biarylpyrazoles such as 1. To determine
whether the carboxamide oxygen or the piperidine nitrogen
of the C-3 substituent may be the hydrogen bond acceptor
for K3.28(192), we designed, synthesized and evaluated a
new set of analogues of 1 (3-6, Chart 1) in which modifica-
tions to the C-3 substituent have been made. Results reported
here support the hypothesis that it is the carboxamide oxygen
of the C-3 substituent of 1 that engages in a hydrogen bond
with K3.28(192) in WT CB1. Furthermore, functional results
for 3-6 support the hypothesis that the interaction of 1 with
K3.28(192) may be key to its inverse agonism because
all analogues that retained carboxamide oxygens were found
to act as inverse agonists, while even high CB1 affinity
analogues that lacked the carboxamide oxygen such as 5-(4-
chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-3-[(E)-piperidi-
noiminomethyl]-1H-pyrazole (5) were found to act as neutral
antagonists.

Biarylpyrazole InVerse Agonists
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20 5971
Figure 2. Compound 1 in a minimum energy conformation (∆E ) 0.92 kcal/mol) docked in the computer TMH bundle model of the R state of
CB1. The view here is from the lipid, looking toward TMH4. Residues with which 1 has direct interactions are colored yellow. Residues that form
part of the aromatic cluster complex with 1 but that do not stack directly with 1 are colored green. Residues that have no direct or indirect interaction
with 1 are colored cyan. (Left) The R state model is characterized by a salt bridge between K3.28(192) and D6.58(366) and two patches of aromatic
residues that form clusters in the TMH3-4-5-6 region of CB1, W5.43(279)/F3.36(200)/W6.48(356), and Y5.39(275)/W4.64(255)/F5.42(278). (Right)-
In the R* state, the K3.28(192)-D6.58(366) salt bridge is broken due to rotations of TMHs 3 and 6 and a conformational change in TMH6. The
aromatic cluster in this region has rearranged with F5.42(278)/W4.64(255)/Y5.39(275)/W5.43(279), forming a cluster. It is clear here that although
1 has aromatic stacking interactions that allow it to be part of the TMH3-4-5-6 aromatic microdomain in both R and R*, 1 is able to engage in
hydrogen bonding (with K3.28(192)) only in the R state.
Chart 1
Figure 3. Global minimum energy conformers of compounds 3, 4,
and 2 as initially identified by the AM1 semiempirical method and
then optimized by ab initio Hartree-Fock 6-31G* calculations.
Figure 4. Global minimum energy conformers of compounds 1, 5,
and 6 (in yellow) as initially identified by the AM1 semiempirical
method and then optimized by ab initio Hartree-Fock 6-31G*
calculations. Illustrated in pink is the higher minimum energy confor-
mation for 1, 5, and 6 used for initial docking studies. The energies of
these conformers were obtained from ab initio Hartree-Fock 6-31G*
calculations as described in the Experimental Section. The calculated
energy expenses for each ligand to assume its initially docked
conformation were the following: 1, ∆E ) 0.92 kcal/mol; 5, ∆E )
0.51 kcal/mol; and 6, ∆E ) 4.55 kcal/mol.
Results
Conformational Analysis. Figure 3 illustrates the global
minimum energy conformers of compounds 3, 4, and 2. Figure
4 illustrates the global minimum energy conformers of com-
pounds 1, 5, and 6 in yellow, superimposed at their pyrazole
rings with the initial docked conformation of each shown in
pink.
has the carboxamide oxygen of the C3 substituent nearly in
plane with the pyrazole ring and pointing in the direction of
the C4 methyl group (O-C1′-C3-C4 ) 0.7°). (See Chart 1 for
the numbering system.) The piperidine ring is in a chair
Compound 1 Global Minimum Energy Conformer. The
global minimum energy conformer of 1 (Figure 4, left, yellow)

5972 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20
Hurst et al.
conformation with the nitrogen lone pair of electrons pointing
in the same direction as the carboxamide hydrogen (LP-N3′-
N2′-H ) 0.2°). The monochlorophenyl ring is out of plane with
the pyrazole ring (C4-C5-C1′′′-C2′′′ ) -59.7°), and the
dichlorophenyl ring is also out of plane with the pyrazole ring
(N2-N1-C1′′-C2′′ ) -75.6°). In this position, the ortho-chloro
is in the bottom face of the molecule (i.e., below the plane of
the page). In the global minimum energy conformer of 1 shown
in Figure 4, the carboxamide group is in a trans geometry. This
caboxamide group trans geometry was found to be lower in
energy than a cis geometry by 4.81 kcal/mol at the HF 6-31G*
level. We also found in docking studies that the cis form had
severe steric clashes with TMH7; therefore, only the trans
geometry was considered further. A recent crystal structure of
1 confirms the carboxamide group trans geometry (George, C.
Laboratory for the Structure of Matter, Naval Research Labora-
tory. Personal communication).
Compound 2 Global Minimum Energy Conformer. The
global minimum energy conformer of 2 (Figure 3, right) differs
from that of 1 only in the orientation of its cyclohexyl ring
compared with that of the piperidine ring in 1. In the global
minimum of 2, the trans-ethylene group is oriented such that
the hydrogen attached to C1′ is nearly in the plane of the
pyrazole ring, pointing toward the C4 methyl group (H-C1′-
C3-C4 ) -2.6°), and the hydrogen attached to C3′ (cyclohexyl
ring) points in the opposite direction from the C2 ′ hydrogen
(H-C3′-C2′-H ) 177.5°).
Compound 3 Global Minimum Energy Conformer. The
global minimum energy conformer of 3 (Figure 3, left) is
analogous to that of 2 and differs from that of 1 only in the
orientation of its cyclohexyl ring compared with that of the
piperidine ring in 1. In the global minimum of 3, the carboxa-
mide group is oriented such that the carboxamide oxygen is
nearly in the plane of the pyrazole ring, pointing toward the
C4 methyl group (O-C1′-C3-C4 ) -2.5°), and the hydrogen
attached to C3′ (cyclohexyl ring) points in the opposite direction
from the amide hydrogen (H-C3′-N2′-H ) -158.7°).
Compound 4 Global Minimum Energy Conformer. The
global minimum energy conformer of 4 (Figure 3, middle)
differs from that of 1 in the orientation of its cyclohexyl ring
compared with that of the piperidine ring in 1. In the global
minimum of 4, the N-methyl amido group is oriented such that
the carboxamide oxygen is closer to C4 methyl group than to
the pyrazole N2. Unlike 2 and 3, the amido group is not in the
plane of the pyrazole ring, but is rotated out of plane due to a
steric clash between the N methyl and the pyrazole N2 nitrogen
(O-C1′-C3-C4 ) 26.3°). The hydrogen attached to C3′ (cyclo-
hexyl ring) points in the opposite direction from the amido N
methyl (H-C3′-N2′-C ) -168.4°).
inactive state but rather in another minimum energy conforma-
tion (Figure 4, left, pink). A molecular electrostatic potential
(MEP) map calculated at the AM1 level (data not shown)
indicated that the piperidine nitrogen of 1 generates an MEP
minimum (i.e., negative potential region) second only to that
generated by the carboxamide oxygen in 1. AM1 conformational
searches identified another minimum energy conformation of
1 in which the piperidine nitrogen’s lone pair points in the
direction opposite to that of the carboxamide hydrogen (LP-
N3′-N2′-H )178.8°). An MEP of this conformer showed an
enhanced negative potential region associated with the C-3
substituent. Although AM1 calculations showed that this
conformer was 4.83 kcal/mol higher in energy than the global
minimum, ab initio Hartree-Fock 6-31G* calculations indicated
that the energy separation between these two conformers was
only 0.92 kcal/mol. For docking studies, we chose to use this
minimum energy conformer of 1 (Figure 4, left, pink), a
conformation that is mimicked by the global minimum energy
conformations of 3, 4, and 2 (Figure 3). The conformations used
for docking 5 and 6 were analogous to the docked conformation
of 1 (Figure 4, middle and right, pink). Hartree-Fock 6-31G*
calculations ab initio indicated that the energy separation
between these initially docked conformers and their respective
global minimum energy conformers (Figure 4, middle and right,
yellow) were 0.51 kcal/mol for 5 and 4.55 kcal/mol for 6.
Receptor Docking. There are two features of the CB1 R
TMH bundle model that appear to be important for the binding
of biarylpyrazoles such as 1 at CB1. The first is a salt bridge
formed by K3.28(192) and D6.58(366). The second is the
aromatic residue-rich TMH3-4-5-6 region of the CB1 model,
which is characterized by a W6.48(356)/W5.43(279)/F3.36(200)
and a Y5.39(275)/W4.64(255)/F5.42(278) aromatic cluster in
the CB1 R state. We begin here by separately discussing these
two features of the CB1 R TMH bundle model.
K3.28(192) and D6.58(366) Salt Bridge. One of the
significant features of our CB1 R TMH bundle model is a salt
bridge between K3.28(192) and D6.58(366) (N-O distance )
2.6 Å; N-H-O angle ) 159°). Unlike the intracellular R3.50/
E6.30 (or R3.50/D6.30) salt bridge suggested to stabilize
G-protein-coupled receptors (GPCRs) in their inactive states,20,21
the extracellular K3.28(192)/D6.58(366) salt bridge (present only
in the inactive state of CB1) seems to be important for
positioning K3.28(192) for ligand interaction in the inactive state
rather than for stabilizing the receptor in the inactive state. In
fact, Pan and co-workers13 found that WT CB1 and the CB1
K3.28(192)A mutant exhibit the same level of constitutive
activity. Therefore, the absence of the K3.28(192)/D6.58(366)
salt bridge does not lead to greater ease of activation. The K3.28-
(192)/D6.58(366) salt bridge is made possible by two special
structural features of the CB1 receptor, its extracellular loop 2
(EC-2) loop and the flexibility of TMH6 in CB1. Despite the
fact that the CB1 and CB2 receptors belong to the Class A
(rhodopsin (Rho)) family of GPCRs, there are important
differences between CB1/CB2 and Rho that impact the ligand
binding pocket in the TMH3-4-5-6 region. The CB1 and CB2
EC-2 loops are shorter than that of Rho (CB1, 15 residues in
length; CB2, 13 residues in length; Rho, 25 residues in length),
and there is no corresponding Cys residue in TMH3 of CB1 or
CB2 that would cause the EC-2 loop to dip down into the
binding site crevice as the EC-2 disulfide bridge with Cys3.25-
(110) causes in Rho. However, in CB1 there is an EC-2 Cys
residue near the extracellular end of TMH4 (C257) and a Cys
near the middle of the EC-2 loop (C264) that have been shown
to be essential for high-level expression and CB1 receptor
Compounds 5 and 6 Global Minimum Energy Conform-
ers. Figure 4 illustrates the global minimum energy conformers
of 5 (center) and 6 (right). The global minimum energy con-
formers of 5 and 6 differ from that of 1 in that their piperidine
rings are rotated nearly 90° about the 2′-3′ bond relative to
that of 1. Such an orientation presumably allows the piperidine
nitrogen lone pair of electrons to resonate with the CdN or
CdC pi systems, respectively (LP-N3′-N2′-C1′ ) 99.1°; LP-
N3′-C2′-C1′ ) 104.8°). The trans-CdN or trans-ethylene group
is oriented such that the hydrogen attached to C1′ is nearly in
the plane of the pyrazole ring, pointing toward the C4 methyl
group (H-C1′-C3-C4 ) -2.8° and 7.3°, respectively).
Conformer Selection for Docking. In our initial study of
1,1 we chose not to dock 1 in its global minimum energy
conformation (Figure 4, left, yellow) in our model of the CB1

Biarylpyrazole InVerse Agonists
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20 5973
Figure 5. (Left) Compound 1 and (Right) compound 3 in their energy minimized complexes with the model of the inactive (R) state of CB1. The
conformational energy expense for each ligand relative to its global minimum in its final energy minimized complex was 1.26 kcal/mol for 1 and
0.00 kcal/mol for 3. The view here is from the lipid, looking between TMHs 5 and 6. The R state is characterized by a salt bridge between
K3.28(192) and D6.58(366) and two patches of aromatic residues that form clusters in the TMH3-4-5-6 region of CB1, W5.43(279)/F3.36(200)/
W6.48(356), and Y5.39(275)/W4.64(255)/F5.42(278). Residues with which ligands have direct interactions are colored here in yellow. Residues
that form part of the aromatic cluster complex with ligand, but do not stack directly with the ligand are colored green. Residues that have no direct
or indirect interaction with the ligand are colored cyan. (Left) The carboxamide oxygen of 1 has a hydrogen-bonding interaction with K3.28(192)
in the salt bridge (inset). Because 1 directly stacks with F3.36(200), Y5.39(275), and W5.43(279), the ligand joins the F3.36(200)/W5.43(279)/
W6.48(356) and Y5.39(275)/W4.64(255)/F5.42(278) aromatic clusters into one large extended cluster in the minimized complex. (Right) The
carboxamide oxygen of 3 has a hydrogen-bonding interaction with K3.28(192) in the salt bridge (inset) and 3 has aromatic stacking interactions
with F3.36(200), Y5.39(275), and W5.43(279). In binding, CHASR (3) bridges between the F3.36(200)/W5.43(279)/W6.48(356) and Y5.39(275)/
W4.64(255)/F5.42(278) aromatic clusters and helps form one large extended cluster in the minimized complex.
function.22 This result combined with mutation results for the
corresponding pair of Cys residues in CB2 (C174 and C179)23
suggests that a disulfide bridge between these two Cys residues
may exist. As the result of this important difference between
Rho and the CB receptors, the binding site crevice around TMHs
3-4-5-6 is likely to be different, with the EC-2 loop occupying
less volume in the upper part of the binding pocket than the
EC-2 loop in Rho. The different spatial requirements of the EC-2
loop are important because this difference permits the extra-
cellular end of TMH6 to occupy a different position in the TMH
bundle than that seen in Rho. We have shown that the small
size of residue 6.49 in CB1 (a Gly) results in the pronounced
flexibility of the CWXP motif in TMH6.24 This motif has been
suggested to function as a flexible hinge, permitting agonist-
promoted movement of the intracellular end of TMH6 that
occurs during activation.25,26 In addition to permitting the
intracellular end of TMH6 to come close to the intracellular
end of TMH3 in the inactive state of CB1, this flexibility in
CB1 TMH6 permits the extracellular end of TMH6 to bend
toward TMH3, resulting in the formation of a salt bridge
between D6.58(366) (near the extracellular end of TMH6) and
K3.28(192) in TMH3. In the active state (R*) TMH bundle
model, the K3.28(192) and D6.58(366) salt bridge is broken
(N-O distance ) 16.8 Å). The change in the positions of K3.28-
(192) and D6.58(366) from the R to R* state in our CB1 models
can be seen in Figure 2. As will be discussed below, we found
that compounds 1-6 fell into two different categories based
upon their hydrogen bonding capabilities with K3.28(192).
TMH3-4-5-6 Aromatic Microdomain. Both the CB1 inverse
agonist/antagonist 1 and compounds 2-6 are highly aromatic
compounds. We hypothesized that aromatic stacking interactions
might be important for the binding of these compounds at CB1.
The CB1 TMH3-4-5-6 region is rich in aromatic residues that
face into the ligand binding pocket, including F3.25(189), F3.36-
(200), W4.64(255), Y5.39(275), W5.43(279), and W6.48(356).
Shire and co-workers23 have shown in CB1/CB2 chimera studies
that the TMH4-EC2-TMH5 region of CB1 contains residues
critical for the binding of 1. In Monte Carlo/stochastic dynamics
simulations of the inactive state of WT CB1,27 we found a
persistent aromatic stack between Y5.39(275) and W4.64(255)
that seemed to be important for stabilizing the positions of
TMHs 4 and 5 in the TMH bundle on the extracellular side
and a second aromatic stack between F3.36(200), W5.43(279),
and W6.48(356) that seemed to be open for additional interaction
with the ligand. Subsequent CB1 F3.36(200)A, W5.43(279)A,
and W6.48(356)A mutation studies indicated that the binding
of 1 is affected by each of these mutations, suggesting that
these residues are part of the binding site for 1.19 There-
fore, compounds 1-6 were docked in the aromatic residue-
rich TMH3-4-5-6 region of the CB1 model here. Our inactive
state (R) bundle model in the TMH3-4-5-6 region is
characterized by a W6.48(356)/W5.43(279)/F3.36(200) and a

5974 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20
Hurst et al.
Figure 6. (Left) Compound 4 and (Right) compound 5 in their energy minimized complexes with the model of the inactive (R) state of CB1. The
conformational energy expense for each ligand relative to its global minimum in its final energy minimized complex was 0.12 kcal/mol for 4 and
1.29 kcal/mol for 5. The view here is from the lipid, looking between TMHs 5 and 6. For a description of the R state, please see Figure 5 caption.
Residues with which ligands have direct, indirect, or no interaction are colored here in yellow, green, or cyan, respectively. (Left) The carboxamide
oxygen of 4 has a hydrogen-bonding interaction with K3.28(192) in the salt bridge (inset). Because 4 directly stacks with both Y5.39(275) and
W5.43(279), the ligand joins the F3.36(200)/W5.43(279)/W6.48(356) and Y5.39(275)/W4.64(255)/F5.42(278) aromatic clusters into one large extended
cluster in the minimized complex. (Right) Compound 5 has no interaction with K3.28(192) in the salt bridge (inset). The piperidine nitrogen of 5,
however, forms a hydrogen bond with C7.42(386). Compound 5 also has direct aromatic stacking interactions with F3.36(200) and W5.43(279).
Although 5 does not stack directly with Y5.39(275), Y5.39(275) has a direct offset parallel stacking interaction with W5.43(279) (d ) 6.9 Å).
Therefore, 5 is part of the larger TMHs 3-4-5-6 aromatic cluster in the minimized complex.
Y5.39(275)/W4.64(255)/F5.42(278) aromatic cluster. We found
that each ligand could insert itself into this aromatic residue-
rich region to become an integral part of an extended aromatic
cluster.
the receptor. HF-6-31G* single point calculations were per-
formed on the conformation of each ligand in the final energy
minimized ligand/receptor complex in order to determine the
energy difference between the global minimum energy con-
former and the final ligand conformation in the minimized
ligand/CB1 R complex. These differences are summarized in
Table 2 (second to last row) and are discussed for each
individual ligand below as well.
Ligand/CB1 R Complexes. The energy minimized ligand/
CB1 inactive (R) state complexes for compounds 1-6 are
illustrated in Figures 5-7. Aromatic residues for which each
ligand has a direct aromatic stacking interaction are colored
yellow in Figures 5-7. Aromatic residues that are part of the
extended aromatic cluster are colored green and those that are
not part of the binding pocket are colored cyan. Aromatic
stacking interactions are summarized in Table 1 according to
the measured distance between ring centroids (d) and the angle
(R) between normal vectors of interacting aromatic rings for
each ligand/CB1 R minimized complex.
Table 2 presents the energy decomposition for each ligand/
CB1 R complex. This Table includes all residues that line the
ligand binding pocket of each ligand and their pairwise
interaction energy with each ligand. The interaction energy
includes both Coulombic and van der Waals components. We
found that for all ligands, van der Waals interactions with V3.32-
(196), T3.33(197), and M6.55(363) contributed significantly to
the overall pairwise interaction energies. Other significant
interactions are discussed below for each individual ligand. In
addition to calculating the pairwise interaction energies, we also
analyzed the output from our minimized ligand/CB1 R com-
plexes and assessed if the ligand remained in its initial docked
conformation or adjusted its conformation in the presence of
Figure 5 illustrates the energy minimized complexes, com-
pound 1/CB1 R (left) and compound 3/CB1 R (right). The
conformation of 1 in this minimized complex is 1.26 kcal/mol
above the global minimum for 1 (Table 2), whereas 3 remains
in its global minimum in the energy minimized complex. The
R state is characterized by a salt bridge between K3.28(192)
and D6.58(366) and two patches of aromatic residues that form
clusters in the TMH3-4-5-6 region of CB1, W5.43(279)/F3.36-
(200)/W6.48(356), and Y5.39(275)/W4.64(255)/F5.42(278). The
carboxamide oxygen of 1 has a hydrogen-bonding interaction
with K3.28(192) in the salt bridge (inset). The hydrogen bond
(N-O) distance and (N-H-O) angles for this interaction are 2.65
Å, 156°, respectively. This result is consistent with our previous
K3.28(192)A mutation results that suggest a direct interaction
between 1 and K3.28(192).1 Because 1 directly stacks with both
F3.36(200) and W5.43(279) and with Y5.39(275) (Table 1), the
ligand bridges the F3.36(200)/W5.43(279)/W6.48(356) and
Y5.39(275)/W4.64(255)/F5.42(278) aromatic clusters present in
the TMH3-4-5-6 aromatic microdomain to form an extended
aromatic cluster in the minimized complex.

Biarylpyrazole InVerse Agonists
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20 5975
Figure 7. (Left) Compound 2 and (Right) compound 6 in their energy minimized complexes with the model of the inactive (R) state of CB1. The
conformational energy expense for each ligand relative to its global minimum in its final energy minimized complex was 0.16 kcal/mol for 2 and
4.57 kcal/mol for 6. The view here is from the lipid, looking between TMHs 5 and 6. For a description of the R state, please see Figure 5 caption.
Residues with which ligands have direct, indirect, or no interaction are colored here yellow, green, or cyan, respectively. Neither 2 nor 6 is capable
of interacting with the K3.28(192) and D6.58(366) salt bridge (see insets). However, both ligands can interact with the TMH3-4-5-6 aromatic
cluster. (Left) Compound 2 directly stacks with F3.36(200), Y5.39(275), and W5.43(279). This permits the ligand to be part of the TMH3-4-5-6
extended aromatic cluster in the minimized complex. (Right) Compound 6 directly stacks with F3.36(200), W5.43(279), and W6.48(356). Although
6 does not stack directly with Y5.39(275), Y5.39(275) has a direct offset parallel stacking interaction with W5.43(279) (d ) 6.7 Å). Therefore, 6
is part of the larger TMH3-4-5-6 aromatic cluster in the minimized complex.
Table 1. Aromatic Stacking Interactions Identified for Compounds 1-6
for this interaction are 2.67 Å and 154°, respectively. Compound
3 also has direct aromatic stacking interactions with F3.36(200)
and W5.43(279) and with Y5.39(275) (Table 1). In binding,
3 also bridges between the F3.36(200)/W5.43(279)/W6.48(356)
and Y5.39(275)/W4.64(255)/F5.42(278) aromatic clusters
and helps form one large extended cluster in the minimized
complex.
in the CB1 R Model
W5.43
W5.43
W6.48-
compd
F3.36
dc Rd
(Å) (deg) (Å) (deg) (Å) (deg) (Å) (deg) (Å) (deg)
Y5.39
Ring Ae
Ring Bf
Ring A
d
R
d
R
d
R
d
R
SR 1
MC ringa
DC ringb
VCHSR 2
MC ring
DC ring
CHASR 3
MC ring
DC ring
6.9
5.3
56 6.0 60 4.7 44
5.8 44
Figure 6 illustrates the energy minimized complexes, com-
pound 4/CB1 R (left) and compound 5/CB1 R (right). The
conformation of 4 in this minimized complex is 0.12 kcal/mol
above its global minimum (Table 2), whereas the conformation
of 5 is 1.29 kcal/mol above its global minimum. Despite the
fact that the global minimum of 4 has its carboxamide group
rotated out of the plane of the pyrazole ring, 4 is still able to
engage in a hydrogen-bonding interaction between its carboxa-
mide oxygen and K3.28(192) as can be seen in Figure 6 (left;
inset). Compound 4 directly stacks with both W5.43(279) and
with Y5.39(275) but lacks an aromatic stacking interaction with
F3.36(200) (Table 1). This is due to the steric interference
caused by its N methyl. Because 4 directly stacks with both
W5.43(279) and with Y5.39(275), however, it joins the F3.36-
(200)/W5.43(279)/W6.48(356) and Y5.39(275)/W4.64(255)/
F5.42(278) aromatic clusters into one large extended cluster in
the minimized complex.
65
76
52
5.2 86
5.9 41 3.5 OPg 4.7 OP
5.4
4.6
4.2 77 5.3 77
5.9 35 3.7 OP 4.5 OP
4.9 75
5.3 75
CHMASR 4
MC ring
DC ring
5.0 49 4.6 20
4.9 89
6.0 20
PIMSR 5
MC ring
DC ring
VPSR 6
MC ring
DC ring
6.5
4.4
30
40
4.0 25
5.5 75
5.1 25
5.9 45
6.2
4.3
33
34
4.9 45
5.8 70
6.8 51
a MC ) monochlorophenyl ring. b DC ) dichlorophenyl ring. c d )
distance between aromatic ring centroids. d R ) angle between normal
vectors of interacting rings. e Ring A ) five-membered ring. f Ring B)
six-membered ring. g OP) offset parallel stack.
In Figure 6 (right), compound 5 has no interaction with
K3.28(192) in the salt bridge (inset). The imino nitrogen
of 5 (N2′ in Chart 1), however, forms a hydrogen bond with
C7.42(386). The hydrogen bond (N-S) distance and (N-H-S)
In Figure 5 (right), the carboxamide oxygen of 3 has a
hydrogen-bonding interaction with K3.28(192) in the salt bridge
(inset). The hydrogen bond (N-O) distance and (N-H-O) angle

5976 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20
Hurst et al.
Table 2. Pairwise Interaction Energies for Compounds 1-6 with CB1 R Binding Pocket Residues, Conformational Energy Expense for Each Ligand
Dock, and the Combined Energy for Each Ligand/CB1 R Complex
Coulomb
kJ/mol
SR141716
VCHSR
CHASR
CHMASR
PIMSR
VPSR
K3.28
-72.02
0.53
-1.04
-1.29
-1.11
0.02
-0.74
-1.02
-0.70
-0.31
0.25
2.49
5.67
-0.12
-0.24
-0.85
1.71
6.54
0.02
0.34
0.27
-2.05
0.66
-54.64
0.65
-1.05
-1.72
-3.93
0.66
0.32
-0.32
1.26
-65.99
-0.21
0.05
-1.02
-0.60
1.04
0.31
1.60
-0.64
0.04
0.26
1.14
9.50
-0.33
-0.47
-0.38
1.38
6.25
0.32
3.44
0.49
L3.29
V3.32
T3.33
F3.36
T3.37
I4.56
Y5.39
W5.43
W6.48
I6.54
M6.55
D6.58
F7.35
C7.38
S7.39
C7.42
-3.26
-1.80
-1.53
-1.11
-0.39
0.15
-3.15
-2.12
-1.25
-0.05
-0.60
0.81
0.26
1.27
-0.39
-0.20
-0.75
-0.35
-9.80
-0.63
0.34
3.44
2.53
0.69
0.35
-0.01
1.89
-0.53
-2.84
-5.56
-9.28
-1.68
0.44
-0.92
-7.00
-8.30
-1.48
-0.34
-0.85
9.05
3.36
-0.44
-0.08
-0.50
1.07
-0.03
-0.45
-0.67
-8.56
subtotal
(kJ/mol)
subtotal
(kcal/mol)
-68.77
-16.43
-4.97
-1.19
-53.14
-12.69
-54.31
-12.97
-26.61
-6.35
-9.04
-2.16
vdW kJ/mol
SR141716
VCHSR
CHASR
CHMASR
PIMSR
VPSR
K3.28
L3.29
V3.32
T3.33
F3.36
T3.37
I4.56
Y5.39
W5.43
W6.48
I6.54
M6.55
D6.58
F7.35
C7.38
S7.39
C7.42
-8.81
-9.99
-20.29
-11.01
-14.39
-2.86
-3.61
-6.59
-25.10
-0.84
-3.13
-28.36
-6.75
-1.03
-0.61
-5.75
-6.03
-8.05
-5.02
-15.51
-9.40
-11.96
-2.18
-6.37
-10.01
-42.01
-1.25
-4.77
-30.14
-10.23
-9.17
-4.59
-5.79
-6.04
-5.64
-6.62
-16.86
-14.31
-15.52
-2.53
-5.73
-8.79
-36.20
-5.26
-0.99
-31.40
-12.35
-1.67
-1.18
-3.22
-4.24
-3.96
-9.20
-19.60
-11.18
-5.82
-3.00
-4.60
-15.04
-28.69
-1.52
-0.48
-21.83
-7.18
-0.45
-1.18
-6.17
-5.10
-8.59
-5.69
-19.23
-15.61
-23.50
-2.68
-1.82
-4.12
-31.46
-2.83
-5.47
-26.48
-8.62
-8.02
-3.13
-4.65
-10.25
-13.85
-7.02
-22.57
-17.26
-24.99
-3.76
-2.03
-3.50
-20.02
-9.16
-6.09
-20.08
-4.14
-4.55
-4.62
-5.77
-5.16
subtotal
(kJ/mol)
subtotal
(kcal/mol)
-155.14
-182.48
-172.51
-145.01
-182.15
-174.55
-37.05
-43.59
-41.20
-34.63
-43.51
-41.69
total
(kJ/mol)
-223.91
-53.48
1.26
-187.45
-44.77
0.16
-225.66
-53.90
0.00
-199.31
-47.61
0.12
-208.76
-49.86
1.29
-183.60
-43.85
4.57
total
(kcal/mol)
ligand conf. costa
(kcal/mol)
combined total
(kcal/mol)
-52.22
-44.61
-53.90
-47.48
-48.57
-39.29
a Energy calculated at the HF-6-31G* level (see Experimental Section).
angle for the interaction with C7.42(386) are 3.32 Å, 174°,
respectively, for 5. In recent CB1 cysteine reactivity studies,
Fay and co-workers found that C7.42(386) is reactive toward
methanethiosulfonate (MTS) sulfhydryl labeling agents and
is, thus, solvent accessible. Steric bulk introduced at this site,
either through MTS labeling or by mutation, inhibited the
binding of 1. These authors proposed that C7.42(386) is near
the piperidine ring of 1. These results are consistent with our
dock of 1 in which C7.42(386) is adjacent to the piperidine
ring of 1.1 The result reported here that 5 can interact with
C7.42(386) is also consistent with the localization of C7.42-
(386) predicted by Fay and co-workers.22 The lack of a hydrogen
bond with K3.28(192) allows 5 to sit deeper in the binding
crevice. As a result, 5 does not maintain a direct stacking
interaction with Y5.39(275), but does maintain aromatic stacking
interactions with F3.36(200) and W5.43(279) (Table 1). Al-
though 5 does not stack directly with Y5.39(275), Y5.39(275)
has a direct offset parallel stacking interaction with W5.43(279)
(d ) 6.9 Å). This Y5.39(275)/W5.43(279) stack permits 5 to
be part of the larger TMH3-4-5-6 aromatic cluster in the
minimized complex.
Figure 7 illustrates the energy minimized complexes, com-
pound 2/CB1 R (left) and compound 6/CB1 R (right). The
conformation of 2 in this minimized complex is 0.16 kcal/mol
above its global minimum (Table 2), whereas the confor-
mation of 6 is 4.57 kcal/mol above its global minimum. Neither

Biarylpyrazole InVerse Agonists
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20 5977
Scheme 1
2 nor 6 has a hydrogen-bonding interaction with K3.28(192)
in the energy minimized complex (see insets). This is reflected
in the small Coulombic energies of interaction listed for
2 and 6 in Table 2. However, both ligands can interact with
the TMH3-4-5-6 aromatic cluster. In Figure 7 (left), com-
pound 2 directly stacks with F3.36(200), W5.43(279), and
Y5.39(275) (Table 1). This permits the ligand to be part of the
TMH3-4-5-6 extended aromatic cluster in the minimized
complex.
(d ) 6.7 Å). This Y5.39(275)/W5.43(279) stack permits 6 to
be part of the larger TMH3-4-5-6 aromatic cluster in the
minimized complex.
Chemistry. The conformationally constrained compound 2,
a vinylcyclohexyl analogue of 1, was synthesized by Wittig
olefination as shown in Scheme 1. The route starts with the
reported pyrazole ester 1′,28 an intermediate in the synthesis
of 1, which was reduced with lithium aluminum hydride to
the alcohol 2′. The latter was converted with CBr4 and
triphenylphosphine to benzylic bromide 3′ and then to phos-
phonium salt 4′ with triphenylphosphine. Deprotonation of 4′
with lithium diisopropylamide, to the corresponding stabilized
phosphorus ylide, and treatment with cyclohexanecarboxalde-
hyde afforded the putative olefin 2. Stabilized ylides, such as
those derived from 4′, typically afford predominantly trans-
(E)-olefins.29,30
In Figure 7 (right), the lack of a hydrogen bond with K3.28-
(192) allows 6 to sit deeper in the binding crevice. As a result,
6 cannot maintain a direct stacking interaction with Y5.39(275).
However, 6 does maintain aromatic stacking interactions with
F3.36(200), W5.43(279), and W6.48(356) (Table 1). Although
6 does not stack directly with Y5.39(275), Y5.39(275) has
a direct offset parallel stacking interaction with W5.43(279)

5978 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20
Hurst et al.
Table 3. Ability of Compound 1 Analogues to Displace [3H]CP55,940,
[3H]SR141716 and [3H]WIN55,212-2 from Membranes Prepared from
HEK Cell Lines Expressing Human CB1a
Proof of the trans geometry was sought via determination of
the coupling constant between the vinyl protons in the 1H NMR
spectrum. However, the chemical shifts of both vinyl protons
were very similar, and hence, classic first-order coupling was
not observed. Although some evidence of a large coupling
constant supportive of a trans olefin was present, a better
separation of the resonances was required to prove trans
geometry. Examination of various NMR shift reagents revealed
that AgFOD resolved the chemical shifts of the two vinyl
protons affording a classic first-order spectrum with a coupling
constant (16 Hz) indicative of trans geometry. Compound 2 was
further characterized by TLC, GC, HPLC, and high-resolution
mass spectrometry.
Displaced against
[3H]CP55,940
[3H]SR141716
[3H]WIN55,212-2
compd
Ki (nM)
Ki (nM)
Ki (nM)
1
2
3
4
5
6
7.1 (1.8-27)
85* (30-243)
5.0 (1.4-18)
54* (25-116)
57* (19-172)
258* (24-2725)
1.8 (1.3-2.2)
31.3b
1.8 (0.4-7.2)
30* (8-106)
17* (5.6-57)
213* (38-1218)
18 (9.2-34)
60* (24-151)
3.7* (1.3-11)
107* (27-414)
19 (7.3-51)
223* (55-903)
a The inhibition constants were obtained from competition experiments
(see Experimental Section). Data are the means and corresponding 95%
confidence limits of three or more independent experiments each performed
in triplicate. The asterisk (/) indicates statistically significant differences
from the affinity of SR141716 for each radioligand (p < 0.05); ND, not
determined. The Bmax values were 1.0 (0.8-1.3), 1.7 (1.4-2.0), and 1.2
(0.5-3.0) pmol/mg for CP55,940, SR141716 and WIN55,212-2, respec-
tively. b Data from 1.
The piperidinoiminomethyl analogue 5 of 1 was synthesized
by a condensation reaction shown in Scheme 1. Thus, the above-
described alcohol 2′ was oxidized to the aldehyde 6′ by the
Swern procedure. Condensation of 6′ with 1-aminopiperidine
in the presence of MgSO4 afforded 5 as a crystalline, sharp-
melting solid. GC and HPLC did not serve to establish
homogeneity because of the decomposition under GC conditions
and a tailing single peak by HPLC. However, the homogeneity
of the product as a single isomer was exhibited by the CMR
spectrum, which showed just 18 resonances, consistent with a
single component with the structure of the title compound. The
structure was further supported by 1H NMR, COSY, and HSQC
(consistent with the structure), DEPT (6 nonequivalent sp2 CH,
3 CH2, 1 CH3), and positive ion MS (446 M+, 364 M-84+2,
328 364-HCl, 84 base C5H10N). A ROESY 2D NMR established
the E-geometry of the imino double bond by a cross-peak for
the through space interaction of the imino hydrogen and the
N-CH2, only possible for the E-geometry.
exchange of the â-enamine proton of 6. Hence, the compounds
are expected to be stable in the aqueous pharmacological studies.
The cyclohexylamide analogue 3 of 1 was reported in the
Sanofi patent33 without details and was prepared for these studies
by a detailed route that parallels that used for the synthesis of
1.34 Ester 1′ was saponified to acid 11′, which was converted
to acid chloride 12′ (Scheme 1). The latter, without purification
since the beginning of the synthesis, was treated with cyclo-
hexylamine to yield 3 after chromatographic purification. Similar
to that seen in the synthesis of 1, the corresponding 1,3-diaryl
isomer was also obtained as a minor product and separated
during chromatography.35 The structure of 3 was established
1
by its H NMR and CMR spectra. Additionally, its shared
synthesis scheme with 1 and the proof of structure of the latter
serves to establish the structure of 3 by extension. The
N-cyclohexyl-N-methylamide analogue (4) was similarly pre-
pared and characterized. The 1H NMR spectrum of the tertiary
amide 4 shows that it exists as a nearly 1:1 mixture of cis and
trans amide rotamers in chloroform at ambient temperature
(Experimental Section).
Radioligand Displacement Assays in CB1 WT. The binding
of compounds 1-6 were tested against a hCB1 cell line (Bmax
values of 1.0 (0.8-1.3), 1.7 (1.4-2.0), and 1.2 (0.5-3.0) pmol/
mg for CP55,940, SR141716, and WIN55,212-2, respectively).
The data shown in Table 3 represent Ki (Kd value for 1 vs [3H]-
SR141716) values of compounds 1-6 displaced against [3H]-
CP55,940, [3H]SR141716, and [3H]WIN55,212-2 at WT hCB1.
Compound 3 retained similar CB1 affinity to 1, and in the case
of competition with [3H]WIN55,212-2, improved affinity.
Compounds 2, 4, 5, and 6 exhibited reduced affinity relative to
that of 1. Compound 6 was found to possess the lowest CB1
affinities in the series against all radioligands. Relative to the
affinity of 1, compound 6 exhibited 36-fold lower affinity versus
[3H]CP55,940, 118-fold lower affinity versus [3H]SR141716,
and 12-fold lower affinity versus [3H]WIN55,212-2.
The trend in interaction energies shown in Table 2, 3,1 <
5,4 < 2 < 6 is consistent with the trend in Ki values seen in
Table 3 versus [3H]CP55,940 and is close to the trend in Ki
values versus [3H]SR141716 (3,1 < 5 < 4,2 < 6), but it deviates
from the trend in Ki values versus [3H]WIN55,212-2 (3 < 1,5
< 2 < 4 < 6). It is important to note, however, that pairwise
interaction energies may not be directly comparable with
changes in affinities. The experimentally measured change in
affinity includes not only the strength of ligand-receptor
interactions in the newly formed complex but also the possible
loss of intrareceptor interactions in the unoccupied receptor
The enamine analogue 6 was prepared by a Horner-Wittig
type olefination of aldehyde 6′ with the reported (piperidinyl-
methyl)diphenylphosphine oxide.31,32 Base promoted addition
of the latter with 6′ afforded the diastereomeric phosphinoyl
diastereomers 8′ and 9′ as indicated in Scheme 1. Treatment of
9′ with potassium hydride afforded E-enamine 6, whereas similar
treatment of 8′ afforded Z-enamine 10′. The latter readily
isomerized to the thermodynamically more stable E-compound
31 and was not stable at ambient temperature beyond a day. The
Z-enamine and its conversion to 6 was observed in the NMR
spectrum of the worked up elimination of diphenyl phosphinic
acid from 8′. Compound 6 was characterized by its NMR and
ROESY spectra. The latter showed a through space interaction
between the enamine â-H and the CH2 groups R to the nitrogen
of the piperidine ring, which is only possible with the trans
isomer.
When the above elimination of the diphenylphosphinic acid
was conducted on each of the pure diphenylphosphinoyl
diastereomers, 8′ and 9′, the corresponding Z- and E-enamines
10′ and 6 were obtained. These were differentiated by an NOE
spectroscopy interaction between the enamine â-proton and the
piperidinyl methylene protons R to the nitrogen, which was
observed for only one isomer and is possible only for E isomer
6. Furthermore, Z isomer 10′ was observed to isomerize
completely to E isomer 6 overnight in CDCl3 solution by NMR.
Hence, the mixtures of 8′ and 9′ can be used to prepare pure 6.
The stability of 5 and 6 to hydrolysis was examined by
monitoring for changes in their NMR spectra when stored in
aqueous methanol and aqueous THF solvent, respectively (the
choice of the water miscible solvents was for optimal resolution
of key resonances). Both compounds showed no change during
4 days storage at ambient temperature other than the expected

Biarylpyrazole InVerse Agonists
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20 5979
agonists.13 Compounds 3 (1 µM, 35.2 ( 14.3%, n ) 13) and 4
(1 µM, 26.1 ( 10.9%, n ) 11 and 10 µM, 23.9 ( 10.9%, n )
7) also significantly increased the Ca2+ current in SCG neurons
expressing CB1 receptors. Thus, the analogues of compound
1, compounds 3 and 4, also acted as inverse agonists of the
CB1 receptor. In contrast, compounds 5 (1 µM, 11.5 ( 6.2%,
n ) 8), 6 (1 µM, 3.2 ( 3.6%, n ) 14 and 10 µM, 13.9 (
7.8%, n ) 8), and 2 (1 µM, 8.6 ( 6.2%, n ) 6) had no
significant effect on the amplitude of the Ca2+ current. Thus,
compounds 5, 6, and 2 acted as neutral antagonists because they
blocked the effect of the agonist but did not have an effect on
their own. Figure 8C shows superimposed Ca2+ current traces
for representative experiments on SCG neurons expressing CB1
receptors in the absence (control) and presence of WIN55,212-
2, the analogue of 1, and the analogue of 1 plus WIN55,212-2.
These results show that the analogues of 1, compounds 3 and
4, behave as inverse agonists because when given alone they
significantly increase the Ca2+ current, an effect opposite that
of the agonist WIN55,212-2, and because they significantly
antagonize the effect of WIN55,212-2. In contrast, compounds
5 and 6 behave as neutral antagonists significantly inhibiting
the effect of WIN55,212-2, while having no significant effect
on their own. Results analogous to those reported here for 5
and 6 were previously obtained for 2 (ref 1, Figure 7B).
Figure 8. Calcium channel assay results for compounds 1-6. (A)
Compound 1 and its analogues act as antagonists of CB1 agonist
WIN55,212-2. The bar graph shows CB1-mediated Ca2+ current
inhibition as a percentage of control current amplitude in superior
cervical ganglion (SCG) neurons injected with rCB1 receptor cDNA.
The cannabinoid receptor agonist WIN55,212-2 (1 µM) inhibits the
Ca2+ current in SCG neurons expressing CB1 cannabinoid receptors.
Coapplication of WIN55,212-2 with 1 (1 µM), 3 (1 µM), 4 (1 and 10
µM), 5 (1 µM), or 6 (1 and 10 µM) blocked the inhibition of the Ca2+
current by WIN55,212-2. Compound 2 also blocked the Ca2+ current
inhibition by WIN55,212-2 as previously reported.1 Therefore, all of
the analogues of compound 1 were effective as antagonists of WIN55,-
212-2. (***p e 0.0001) (B) Compounds 1, 3, and 4 act as inverse
agonists. The bar graph shows the Ca2+ current increase as a percentage
of control current amplitude in the presence of 1 and its analogues in
SCG neurons injected with rCB1 receptor cDNA. The inverse agonist
1 (1 µM) increased the Ca2+ current, an effect opposite that of the
cannabinoid agonist WIN55,212-2. Compounds 3 (1 µM) and 4 (1 and
10 µM) also enhanced the Ca2+ current indicating that these analogues
of 1 act as inverse agonists. Compounds 5 (1 µM), 6 (1 and 10 µM),
and 2 (1 µM) did not enhance the Ca2+ current indicating that these
analogues act as neutral antagonists. (paired t-test *p e 0.05; **p e
0.01). (C) Superimposed Ca2+ current traces in the absence (control)
and presence of WIN55,212-2 (agonist), the analogue of compound 1
(indicated by compound number), and the analogue of compound 1
plus WIN55,212-2 (analogue number + agonist) for compounds 3, 4,
5, and 6. Current traces for compound 2 can be found in our previous
publication (in Figure 7B).1
Discussion
Ligand Binding Affinity. 5-(4-Chlorophenyl)-3-[(E)-2-
cyclohexylethenyl]-1-(2,4-dichlorophenyl)-4-methyl-1H-pyra-
zole (2). In previous work, our modeling studies had suggested
that the C-3 substituent of 1 could hydrogen bond with K3.28-
(192). We designed 2 to be used in a mutant cycle study to test
this hypothesis.1 Compound 2 lacks hydrogen-bonding potential
in its C-3 substituent but mimics the geometry of 1 because it
has a trans ethylene group replacing the trans amide group in 1
and a cyclohexyl ring replacing the piperidine ring in 1. In the
mutant cycle calculation, the C-3 substituent/K3.28(192) inter-
action is tested by evaluating the effect on affinity when a set
of complementary chemical groups is deleted from both ligand
(1 f 2, which removes hydrogen bonding potential from the
C-3 substituent) and receptor (WT CB1 f K3.28(192)A, which
removes the ability of residue 3.28 to offer a hydrogen bond).
Scatchard analysis and ligand-binding results for WT CB1
resulted in a Kd value for 1 at WT CB1 of 2.3 ( 1.1 nM and
a Kd value for 1 in CB1 K3.28(192)A of 39.6 (10.5 nM. The
Ki value for 2 binding in cloned human WT CB1 and CB1
K3.28(192)A cell lines versus [3H]SR141716 was 31.3 ( 9.6
and 35.2 ( 1.4 nM, respectively. The affinity drop for 1 upon
the K3.28(192)A mutation and the affinity similarity between
2 at WT CB1 and 1 at K3.28(192)A suggest that there may be
a direct interaction between the C-3 substituent of 1 and K3.28-
(192) (and none between K3.28(192) and 2). However, the key
to the determination of whether deletions have occurred between
two groups that interact indirectly or directly is the effect
produced by simultaneous deletion of both groups (i.e., com-
pound 2/K3.28(192)A). If the modified groups do not interact
directly with each other in the WT state, then the effect of the
two simultaneous changes will be additive. If the two groups
do interact directly in the WT state, there will be no further
affinity loss for the compound 2/K3.28(192)A case relative to
that of the two cases where one (compound 2/WT CB1) or the
other interacting group (compound 1/K3.28(192)A) was deleted.
The fact that the Ki for the double deletion (i.e., compound
2/K3.28(192)A; 35.2 ( 1.4 nM) was not statistically different
from either single deletion (compound 1/K3.28(192)A; 39.6
(10.5 nM or compound 2/WT CB1; 31.3 ( 9.6 nM) indicated
resulting from ligand binding. Although Table 2 should reflect
the former, it does not take into consideration the latter.
Calcium Current Effects of Compounds 1-6 in SCG
Neurons. Figure 8A illustrates that in superior cervical ganglion
(SCG) neurons injected with rCB1 cannabinoid receptor cDNA,
compound 1 (1 µM, -4.8 ( 1.3%, n ) 8) and its analogues 3
(1 µM, -3.8 ( 1.3, n ) 13), 4 (1 µM, -4.5 ( 2.4%, n ) 12
and 10 µM, -7.0 ( 2.3%, n ) 7), 5 (1 µM, 2.6 ( 1.6%, n )
8), 6 (1 µM, -2.5 ( 1.0%, n ) 15 and 10 µM, 0.7 ( 3.0%, n
) 8), and 2 (1 µM, 28.1 ( 7.0%, n ) 6) block the inhibition
of the Ca2+ current elicited by the cannabinoid agonist WIN55,-
212-2 (1 µM, 47.1 ( 1.7%, n ) 76). The data on 2 was
previously reported.1 Thus, compounds 1-6 all acted as
antagonists of WIN55,212-2. Compound 1 also acted as an
inverse agonist because application of 1 alone significantly
increased the Ca2+ current (1 µM, 46.6 ( 7.9%, n ) 8) in SCG
neurons expressing CB1 receptors (Figure 8B). Compound 1
acts to reverse the tonic inhibition of Ca2+ channels resulting
from constitutively active CB1 receptors in the absence of

5980 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20
Hurst et al.
that there is likely a direct interaction between the C-3
substituent of 1 and K3.28(192) in WT CB1.1 The present article
builds upon these results and seeks to determine which part of
the compound 1 C-3 substituent (the carboxamide oxygen or
the piperidine nitrogen) has this direct interaction with K3.28-
(192).
(Figure 6). This hydrogen bond may compensate in part for 5’s
inability to form a hydrogen bond with K3.28(192) and moderate
the affinity drop seen for 5 when displaced against [3H]CP55,-
940 and [3H]SR141716. The compensation provided by the
C7.42(386) hydrogen bond appears to have been more signifi-
cant when [3H]WIN55,212-2 was used as the radioligand.
5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-
pyrazole-3-(N-cyclohexylcarboxamide) (3). Compound 3 was
designed to test whether the piperidine nitrogen is important
for interaction at CB1. In this analogue, the C-3 substituent
piperidine ring of 1 is replaced with a cyclohexyl ring.
Compound 3, therefore, retains the C-3 substituent carboxamide
oxygen but lacks the piperidine nitrogen. The CB1 affinity data
presented in Table 3 indicate that 1 and 3 have comparable
affinities versus [3H]CP55,940 and [3H]SR141716 and that 3
has slightly better CB1 affinity than 1 versus [3H]WIN55,212-
2. These results are consistent with previous data reported for
3 by the Makriyannis lab.36 Table 2 indicates that there is no
energy expense needed for 3 to dock at CB1. Taken together,
these results suggest that the piperidine nitrogen of 1 is not
important for the CB1 affinity of 1 but rather that the
carboxamide group is key.
5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-
pyrazole-3-(N-cyclohexylcarboxamide) (4). Compound 4 is an
analogue of 3 that differs from 3 only in the replacement of the
amide N-H with a N-methyl group. This analogue was synthe-
sized to test if the amide NH may contribute to a binding site
interaction with some hydrogen bond acceptor at CB1. As is
clear in Table 3, the affinity of 4 is at least 10-fold lower than
that of 3 ([3H]CP55,940 data). At first inspection, it would seem
that one conclusion that could be drawn from this reduction in
affinity is that an N-H hydrogen-bonding interaction may be
1
important to the affinity of 1 at CB1. However, the H NMR
spectrum of 4 shows that it exists as a nearly 1:1 mixture of cis
and trans amide rotamers in chloroform at ambient temperature
(see Experimental Section). Our modeling studies indicated that
only the trans amide rotamer of 4 is capable of docking at CB1.
Furthermore, modeling results for 3 (and for 1) suggest that
the trans amide rotamer of 4 is not capable of adopting a
conformation analogous to the docked conformation of 1 (Figure
4) due to steric interference between the pyrazole ring and the
N-methyl group (Figure 3). Thus, our modeling studies of 4
(Figures 3 and 6) suggest that its reduction in affinity at CB1
is due to the fact that only the trans amide rotamer can bind at
CB1, and this rotamer binds with reduced interaction energy
(Table 2).
1-[2-(5-(4-Chlorophenyl)-1(2,4-dichlorophenyl)-4-methyl-
1H-pyrazole-3-yl)-(E)-vinyl]-piperidine (6). Compound 6 was
designed to test whether the C-3 substituent carboxamide oxygen
is important for compound 1 binding at CB1. This analogue
differs from 1 in that its carboxamide group has been replaced
by a trans ethylene group, which preserves the geometry of the
carboxamide group but lacks hydrogen bonding potential.
Compound 6, however, retains a piperidine ring located in the
same position as that in 1. As is clear in Table 3, 6 had the
lowest CB1 affinity of all analogues, irrespective of the
radioligand used. This result suggests that the carboxamide
group (which 6 lacks) is important to the CB1 affinity of 1 and
that the piperidine nitrogen is not an important interaction site
in the C-3 substituent of 1. However, we have shown here that
the global minimum energy conformer of 6 has its piperidine
ring rotated nearly 90° relative to the piperidine ring orientation
in the compound 1/CB1 complex (Figure 4). As shown in Table
2, 6 has an energy expense of 4.57 kcal/mol to adopt the
compound 1 piperidine ring orientation in the dock of 1 shown
in Figure 5. This by no means is too high of an energy expense
to make the dock feasible.37 Therefore, the significant affinity
loss for this analogue likely reflects the double deficit: affinity
reduction due to conformational energy expense and affinity
reduction due to the lost interaction for the C-3 substituent at
CB1.
Taken together, the ligand-binding affinity data and receptor-
modeling studies presented here suggest that the piperidine
nitrogen of the compound 1 C-3 substituent is not important
for interaction at CB1 but that the carboxamide oxygen may
be important for interaction at CB1. Furthermore, because our
previous mutant cycle study suggested that K3.28(192) has a
direct interaction with the C-3 substituent of 1,1 we conclude
here that the direct interaction between K3.28(192) and 1 is
most likely mediated by its carboxamide oxygen.
Ligand Efficacy. Calcium channel assay results here suggest
that compounds 1-6 can be divided into two categories: inverse
agonists (1, 3, and 4) and neutral antagonists (2, 5, and 6). This
distinction between inverse agonists and neutral antagonists is
not correlated with receptor affinity because neutral antagonist
5 has comparable, if not better, CB1 affinity than inverse agonist
4. For many years, drugs acting at GPCRs were thought to be
divided into two classes, agonists and antagonists. It is now
apparent that for most GPCRs, the compounds acting at these
receptors can exhibit a spectrum of efficacy from inverse
agonism through neutral antagonism to agonism.38 It has been
widely assumed for GPCRs that inverse agonists suppress the
agonist-independent activity of these receptors by stabilizing
the receptor in an inactive state. This depends on receptor
activation (receptor-G-protein coupling) occurring in the absence
of the agonist, and the inverse agonist suppressing this activity
in some way.38 We have shown previously that the inverse
agonism of 1 can be explained by a two state model in which
1 has aromatic stacking interactions in the TMH3-4-5-6 aromatic
microdomain in both the inactive and active states of CB1 but
can interact with K3.28(192) only in the CB1 inactive state.1
We report here that the inverse agonists, 3 and 4, also interact
in the TMH3-4-5-6 aromatic microdomain in both the inactive
and active (data not shown) states of CB1 but can interact with
K3.28(192) only in the CB1 inactive state. Thus, our modeling
5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-3-[(E)-
piperidinoiminomethyl]-1H-pyrazole (5). Compound 5 lacks
a carboxamide oxygen but has an iminomethyl group replacing
the trans amide group of 1 that preserves the trans geometry of
this amido group. Compound 5 also retains the piperidine ring.
The affinity of 5 for CB1 was reduced relative to 1 by 8-fold
when displaced against [3H]CP55,940 and 9.4-fold when
displaced against [3H]SR141716, but not reduced when dis-
placed against [3H]WIN55,212-2. Modeling studies suggested
that the global minimum energy conformer of 5 had its
piperidine ring rotated 90° relative to the piperidine ring
orientation in the compound 1/CB1 complex (Figure 4). As
shown in Table 2, 5 has an energy expense of only 1.29 kcal/
mol to adopt the compound 1 piperidine ring orientation in the
dock of 1 shown in Figure 5. Modeling studies also suggested
that the C-3 substituent of 5 is not capable of interacting with
K3.28(192) but can form a hydrogen bond with C7.42(386)

Biarylpyrazole InVerse Agonists
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20 5981
(2-fold to 16-fold) and comparing the set of minimum energy
conformers that resulted from each search. We found that rotations
lower than 6-fold missed some conformers seen in higher fold
rotations, but 6-fold rotations and higher yielded the same set of
conformers. Consequently, all conformer searches reported here for
this structurally related set of compounds used 6-fold searches.
An ab initio geometry optimization at the HF 6-31G* level was
then performed for each global minimum energy conformer of 1-6
identified by the AM1 conformational search and for certain other
minimum energy conformers identified. The energy separations
between conformers reported in the text were calculated using
results from these ab initio Hartree-Fock calculations at the 6-31G*
level as encoded in Jaguar (version 6.0, Schro¨dinger, LLC, New
York, NY). To calculate the energy difference between the global
minimum energy conformer of each compound and its final docked
conformation (the latter as listed in Table 2), rotatable bonds in
the global minimum energy conformation were driven to their
corresponding value in the final docked conformation and the single
point energy of the resultant structure was calculated at the HF
6-31G* level.
Receptor Model Construction. Amino Acid Numbering
System. Receptor residues are numbered here using the amino acid
numbering scheme proposed by Ballesteros and Weinstein.42 In this
numbering system, the most highly conserved residue in each
transmembrane helix (TMH) is assigned a locant of 0.50. This
number is preceded by the TMH number and followed in
parentheses by the sequence number. All other residues in a TMH
are numbered relative to this residue. In this numbering system,
for example, the most highly conserved residue in TMH2 of the
CB1 receptor is D2.50(163). The residue that immediately precedes
it is A2.49(162).
studies predict that compounds 1, 3, and 4 should have higher
affinity for the inactive state of the receptor and behave as
inverse agonists. The neutral antagonists, 2, 5, and 6, also can
have aromatic stacking interactions in the TMH3-4-5-6 aromatic
microdomain in both the inactive and active (data not shown)
states of CB1, but these ligands lack the ability to interact with
K3.28(192). Thus, our modeling results suggest that such
biarylpyrazoles (i.e., compounds 2, 5, and 6) should have a
nearly equal preference for R and R* and, consequently, should
fall near the middle of the spectrum of efficacy (inverse agonism
f neutral antagonism f agonism).
Why would interaction specifically with K3.28(192) be the
key for the production of inverse agonism for this class of
compounds at CB1? First, K3.28(192) is a residue capable of
acting as a strong hydrogen bond donor, the only such positively
charged inward facing residue in the extracellular half of the
CB1 TMH bundle. Table 2 illustrates that for compounds 1, 3,
and 4 (those compounds capable of interacting with K3.28(192)),
a major contribution to the overall pairwise interaction energy
of each at CB1 is the Coulombic interaction with K3.28(192)
(-72.02, -54.64 and -65.99 kJ/mol, respectively). Second,
K3.28(192) is attached to a helix that undergoes movement
during activation. This means that the location of K3.28(192)
should change during the R to R* transition. It has been shown
for rhodopsin and the beta-2-adrenergic receptor that GPCR
25,26,39
activation is accompanied by rotations of TMH3 and 6.
Both rotations have been reported to be counterclockwise
rotations from the extracellular view. In our activated state model
of CB1, K3.28(192) is repositioned to be available in the TMH2-
6-7 region, away from the TMH3-4-5-6 aromatic microdomain
(Figure 2). So K3.28(192) is available to ligands binding in the
TMH3-4-5-6 aromatic microdomain only in the inactive state
model. Consistent with this model, mutation studies have
indicated that both the CB1 inverse agonist, 1, and the CB1
agonist WIN55212-2 bind in the TMH3-4-5-6 aromatic micro-
domain of CB1,19,27,40 but only 1 interacts with K3.28(192).1,41
Therefore, results presented here suggest that biarylpyrazoles
(such as 1) capable of a strong interaction with K3.28(192) will
favor the inactive state and, thus, behave as inverse agonists.
Finally, the conclusion drawn here that hydrogen bonding of
the C-3 substituent carboxamide is important for the production
of inverse agonism in this class of compounds is supported by
results reported by Jagerovic and coauthors7 for the triazole,
5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-3-hexyl-1H-1,2,4-
triazole. In this analogue, the C-3 substituent (i.e., conventional
N-substituted carboxamide) of 1 is replaced with an n-hexyl
group. Consistent with work reported here, this compound was
found to behave as a neutral antagonist in several pharmacologi-
cal models.
Model of Inactive (R) Form of CB1. A model of the inactive
(R) form of CB1 was created using the 2.8 Å crystal structure of
bovine rhodopsin (Rho).43 First, the sequence of the human CB1
receptor 44 was aligned with the sequence of bovine rhodopsin (Rho)
using the same highly conserved residues as alignment guides,
45
which were used initially to generate our first model of CB1.
TMH5 in CB1 lacks the highly conserved proline in TMH5 of Rho.
The sequence of CB1 in the TMH5 region was aligned with that
of Rho as described previously using its hydrophobicity profile.45
Helix ends for CB1 were chosen in analogy with those of Rho43;
TMH1: N1.28(112) - R1.61(145); TMH2: R2.37(150) - H2.68-
(181); TMH3: S3.21(185) - R3.56(220); TMH4: T4.38(229) -
C4.66(257); TMH5: H5.34(270) - K5.64(300); TMH6: R6.28-
(336) - K6.62(370); TMH7: K7.32(376) - S7.57(401); intracellular
extension of TMH7 (Helix 8): D7.59(403) - C7.71(415). With
the exception of TMH1, these helix ends were found to be within
one turn of the helix ends originally calculated by us and reported
in 1995.45 Changes to the general Rho structure that were
necessitated by sequence divergences included the absence of helix
kinking proline residues in TMH1 and TMH5 and the lack of a
GG motif in TMH2 as well as the presence of extra flexibility in
TMH6.
Because TMH6 figures prominently in the R to R* transition
(i.e., activation), we have studied the conformations accessible to
TMH6 in CB1 using the Monte Carlo/simulated annealing tech-
nique, Conformational Memories (CM).24 These studies revealed
that TMH6 in CB1 has high flexibility because of the small size of
residue 6.49 (a Gly) immediately preceding Pro 6.50. Two families
of conformers were identified by CM for TMH6 in CB1. Cluster
1 showed a pronounced proline kink (40 members out of 100, 71.2°
average kink angle). Cluster 2 contained helices with less pro-
nounced kinks (51 members out of 100; 30.1° average kink angle).
A conformer from the more kinked CM family of CB1 TMH6
(Cluster 1) was used in our model of the inactive (R) state of CB1.
This conformer was selected (Pro kink angle ) 53.1°) so that R3.50-
(214) and D6.30(338) could form a salt bridge at the intracellular
ends of TMHs 3 and 6 in the CB1 TMH bundle. An analogous salt
bridge has been shown to be an important stabilizer of the inactive
state of the â2 adrenergic and the 5HT-2a receptors 20,21 and to be
present in Rho.43
Experimental Section
Molecular Modeling. A recent crystal structure of 1 confirms
that the carboxamide of the C3 substituent is in a trans geometry
(George, C. Laboratory for the Structure of Matter, Naval Research
Laboratory. Personal communication). Complete conformational
analyses of 1-6 were performed using the semiempirical method,
AM1 within the Spartan molecular modeling program (Wavefunc-
tion, Inc., Irvine, CA). AM1 6-fold conformer searches were
performed for the rotatable bonds in 1 (Chart 1; C-3 substituent:
C3-C1′ and N2′-N3′; C-5 substituent: C5-C1′′′; N-1 substituent:
N1-C1′′). The corresponding bonds in compounds 2-6 were
similarly treated. In a 6-fold AM1 conformer search, local energy
minima are identified by rotation of a subject torsion angle through
360° in 60° increments, followed by AM1 energy minimization of
each rotamer generated. The use of 6-fold rotations in our protocol
was determined by first running lower and higher fold rotations

5982 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20
Hurst et al.
Model of Active (R*) Form of CB1. An R* CB1 model was
created by modification of our Rhodopsin-based model of the
inactive (R) form of CB1. This R* model construction was guided
by the biophysical literature on the R to R* transition in rhodopsin
(Rho), the â-2-adrenergic and muscarinic M3 receptors. This
literature has indicated that for the â-2-adrenergic receptor, a salt
bridge between R3.50 and E6.30 at its intracellular end stabilizes
this receptor in its inactive state.20 Biophysical studies of Rho, the
â-2-adrenergic receptor, and the muscarinic M3 receptor have
indicated that rotation of TMHs 339 and 646,47 occurs upon activation.
Studies of Rho and the â-2-adrenergic receptor, have indicated that
a conformational change in TMH6 occurs upon activation.25,26,48
Jensen and co-workers 25 demonstrated through fluorescence studies
in the â-2-adrenergic receptor that P6.50 in the highly conserved
CWXP motif of TMH6 can act as a flexible hinge that mediates
the transition from R to R*. In the R state, these investigators
proposed that TMH6 is kinked at P6.50 such that its intracellular
end is nearly perpendicular to the membrane and close to the
intracellular end of TMH3. The transition to the R* state is
accomplished by the straightening of TMH6 such that the intra-
cellular part of TMH6 moves away from the receptor core and
upward toward the lipid bilayer.25 All of these experimental findings
were used to create the R* model of CB1 described here.
Our conformational memories study of CB1 TMH6 revealed two
distinct conformational families for TMH6 that differed in the
degree of kinking at CWGP.24 These conformationally distinct
TMH6’s were used to create the R and R* states depicted here. In
the R* bundle, a TMH6 conformer from the second major
conformational family (less kinked: 21.8° kink angle) identified
by CM24 was substituted for the TMH6 conformer used in the
inactive model of CB1. This conformer was chosen so that the salt
bridge in the inactive state between R3.50(214) and D6.30(338)
would be broken because of the movement of the intracellular end
of TMH6 away from that of TMH3 and out into lipid.20 In addition,
the R* (active) CB1 bundle was created by rotating TMH3 so that
residue 3.41 changes environments.39 This was accomplished by a
20° counterclockwise (extracellular view) rotation of TMH3 from
its orientation in the inactive (R) bundle. TMH6 was also rotated
(counterclockwise from extracellular view) so that Cys 6.47 became
accessible from inside the binding site crevice.46
by energy minimization refined the initial docked postion and
interactions for each ligand as indicated in Figures 5-7 and Tables
1-2.
The energy of each ligand/CB1 R or R* TMH bundle complex
was minimized using the AMBER* united atom force field in
Macromodel (version 8.6, Schro¨dinger, LLC, New York, NY). A
distance dependent dielectric, 8.0 Å extended nonbonded cutoff
(updated every 10 steps), 20.0 Å electrostatic cutoff, and 4.0 Å
hydrogen bond cutoff were used. The first stage of the calculation
consisted of 2000 steps of Polak-Ribier conjugate gradient (CG)
minimization in which a force constant of 225 kJ/mol was used on
the helix backbone atoms in order to hold the TMH backbones
fixed, while permitting the side chains to relax. The second stage
of the calculation consisted of 100 steps of CG in which the force
constant on the helix backbone atoms was reduced to 50 kJ/mol in
order to allow the helix backbones to adjust. Stages one and two
were repeated with the number of CG steps in stage two incre-
mented from 100 to 500 steps until a gradient of 0.04 kJ/(mol Å 2)
was reached.
Assessment of Aromatic Stacking Interactions. Burley and
Petsko49 have reported that aromatic-aromatic (π-π) stacking
interactions in proteins operate at distances (d) of 4.5-7.0 Å
between ring centroids. The angle (R) between normal vectors of
interacting aromatic rings typically is between 30° and 90°,
producing a tilted-T or edge-to-face arrangement of interacting rings.
Hunter and co-workers50 have reported that π-π parallel stacking
interactions (R < 30°) between phenylalanine residues in proteins
are favorable if the rings are offset from each other. Residues and/
or ligand regions were designated here as participating in an
aromatic stacking interaction if they had centroid-to-centroid
distances between 4.5 and 7.0 Å. These interactions were further
classified as tilted-T arrangements if 30° e R e 90° and as parallel
arrangements for R < 30°. Parallel arrangements were considered
favorable only if the interacting rings were offset from each other.50
All measurements were made using Maestro (version 7.0, Schro¨-
dinger, LLC, New York, NY).
Assessment of Pairwise Interaction Energies. After defining
the atoms of each ligand as one group (Group 1) and the atoms
corresponding to a residue that lines the binding site in the final
ligand/CB1 R complex (Figures 5-7) as another group (Group 2),
Macromodel (version 8.6, Schro¨dinger, LLC, New York, NY) was
used to output the pair interaction energy (Coulombic and van der
Waals) for a given pair of atoms. The pairs corresponding to Group
1 (ligand) and Group 2 (residue of interest) were then summed to
yield the interaction energy between the ligand and that residue.
Molecular Biology/Pharmacology. Cell Transfection. HEK
293 cells were grown as previously described51 and transfected by
Lipofectamine reagent (Life Technologies) with expression plasmids
containing wild-type human CB1 receptor. Stably transfected cells
were selected in growth medium containing Geneticin (1 mg/mL).
Immunofluorescence. Six selected colonies were expended and
tested for receptor expression by immunofluorescence as previously
described.19 Briefly, cells were grown on a poly-D-lysine (20 µg/
mL) coated cover slip and labeled for 1 h at room temperature
with a polyclonal antibody (1:100) generated against the N-terminus
(1-14) of the human CB1 receptor (Cayman Chemicals). Cells
were washed with HEPES-buffered saline composed of (mM): 130
NaCl, 25 D-glucose, 10 HEPES, 5.4 KCl, 1.8 CaCl2, and 1 MgCl2
at pH 7.4 and fixed for 10 min with 4% paraformaldehyde. CB1
receptor antibody was detected with Alexa fluor 488-conjugated
goat anti-rabbit secondary antibody (1:500 for 40 min; Molecular
Probes). Immunofluorescence was visualized with a fluorescence
microscope (Nikon). One cell-line with cell surface labeling similar
to that of human CB1 receptor expression was selected for binding
experiments.
Preparation of Helices. Each helix of the model was capped as
the acetamide at its N-terminus and as the N-methyl amide at its
C-terminus. Ionizable residues in the first turn of either end of the
helix were neutralized, as were any lipid facing charged residues.
Ionizable residues were considered charged if they appeared
anywhere else in the helix.
Ligand-Receptor Complex. Each ligand was docked in the
aromatic residue rich TMH3-4-5-6 region of the CB1 R or R* TMH
bundle using interactive computer graphics. Automated docking
algorithims are quite useful when little is known about the binding
site of a ligand. In the present case, such methods were not used
because the binding region in CB1 for 1 (TMH3-4-5-6) and the
nature of specific ligand functional group/specific amino acid
interactions are known from mutation/chimera studies. The com-
bined information limits the ligand to one particular region of CB1
and to one particular orientation in this region. In the absence of
such information, the interactive docking used here would certainly
have introduced investigator bias into the results. However, given
the following experimental evidence, we consider our approach here
23
to be appropriate. Shire and co-workers have shown in CB1/
CB2 chimera studies that the TMH4-EC2-TMH5 region of CB1
contains residues critical for the binding of 1. Subsequent CB1
F3.36(200)A, W5.43(279)A, and W6.48(356)A mutation studies
published by our group indicated that the binding of 1 is affected
by each of these mutations, suggesting that these residues are part
of the binding site for 1.19 Our previous mutant cycle study indicated
that K3.28(192) is involved in a direct interaction with the C-3
substituent of 1 in wild-type (WT) CB1.1 This result fixes the
orientation of 1 in the CB1 bundle to be that pictured in Figure 5.
Docking of 1-6 in this region and with this orientation followed
Radioligand Binding. The assay has been previously described.52
Briefly, cells were harvested in phosphate-buffered saline containing
1mM EDTA and centrifuged at 500g for 5 min. The cell pellet
was homogenized and centrifuged twice at 1600g (10 min, 4 °C).
The combined supernatants were centrifuged at 39 000g (60 min,
4 °C). The pellet (P2 membrane) was resuspended in a buffer

Biarylpyrazole InVerse Agonists
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20 5983
composed of 50 mM Tris-HCl, 1 mM EDTA, and 3 mM MgCl2 at
pH 7.4 to yield a protein concentration of approximately 2 mg/
mL. Membrane preparations were aliquoted and stored at -80 °C.
Binding was initiated by the addition of 50 µg of membrane protein
into pre-silanized glass tubes containing the radioligand in buffer
composed of 50 mM Tris-HCl, 1 mM EDTA, 3 mM MgCl2, and
5 mg/mL fatty acid-free BSA at pH 7.4. The radioligands employed
in this study were [3H]CP55,940 (158 Ci/mmole), [3H]SR-141716A
(25 Ci/mmole), and [3H]WIN55,212-2 (41 Ci/mmole). Nonspecific
binding was assessed by the addition of 1 µM unlabeled CP55,-
940, SR141716, or WIN55,212-2 to the tubes. Binding assays were
performed in a total volume of 500 µL for 60 min at 30 °C. Free
and bound radioligands were separated by rapid filtration through
Whatman GF/C filters that had been soaked in polyethylenimine
(0.1%). Filters were shaken for 1 h in 6 mL of scintillation fluid
(Fisher) and radioactivity was determined by liquid scintillation
counting. Saturation experiments were conducted with 6 concentra-
tions of radioligand ranging from 250 pM to 10 nM. Competition
assays were conducted with 1 nM [3H]CP55,940 or [3H]SR141716
or 2.4 nM [3H]WIN55,212-2 and 6 concentrations (0.01 nM to 10
µM) of displacing ligands. Bmax and Kd values were calculated by
unweighted least-squares nonlinear regression of log concentration
values versus binding of pmol/mg of protein. These data was fit to
a one-site binding model using GraphPad Prism (GraphPad San
Diego, CA). Displacement EC50 values were determined by
unweighted least-squares nonlinear regression of log concentration-
percent displacement data and then converted to Ki values using
controlled by a fast switching device (Warner Instrument Corpora-
tion, Hamden, CT).
Chemistry. [5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-
methyl-1H-pyrazol-3-yl]methanol (2′). The pyrazole ester 1′ (3.19
g, 7.78 mmol) dissolved in 80 mL of anhydrous ether was added
dropwise to a stirred suspension of LiAlH4 (1.18 g, 31.1 mmol)
under nitrogen at ambient temperature. After stirring overnight, TLC
analysis (SiO2, 1:3 acetone/CH2Cl2, UV, PMA-Ce4+) on an aliquot
partitioned between water and ether showed the reaction to be
complete. The reaction was quenched by the addition of water (1.18
mL), 15% NaOH (1.18 mL), and water (3.54 mL) with mixing.
Filtration from the resulting aluminates and evaporation of the ether
1
afforded 2.84 g (99%) of the title alcohol as an orange solid. H
NMR (MeOD, 300 MHz) δ: 7.58 (1H, d, J ) 2.2 Hz, 3-H), 7.54
(1H, d, J ) 8.5 Hz, 6-H), 7.46 (dd, 1H, J ) 2.2, 8.5 Hz, 5-H), 7.37
(dt, 2H, J ) 8.5, 2.1 Hz, 3′,5′-H), 7.20 (dt, 2H, 8.5, 2.0 Hz, 2′,6′-
H), 4.88 (s, 2H, CH2), 2.30 (s, 3H, CH3).
3-(Bromomethyl)-5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-
4-methyl-1H-pyrazole (3′). Triphenylphosphine (2.35 g, 8.95
mmol) and carbon tetrabromide (2.97 g, 8.95 mmol) in 50 mL of
anhydrous ether were stirred at ambient temperature for 15 min.
The pyrazole alcohol 2′ (2.74 g, 7.46 mmol) dissolved in 50 mL
of ether was added dropwise to the above phosphonium reagent
with stirring followed by a further 40 mL of an ether wash of the
dropping funnel. After 3.5 h, TLC analysis (SiO2, 1:3 EtOAc/
hexane, UV, PMA-Ce4+) of a direct aliquot showed the reaction
to be complete. Filtration through Celite and rotary evaporation of
the volatiles in vacuo afforded a residue that was chromatographed
on silica gel (350 g) eluting with a step gradient of CH2Cl2/hexane
(10-50%) that yielded 1.12 g (35%) of the title bromide as a
yellow/orange solid. 1H NMR (MeOD, 300 MHz) δ: 7.56
(1H, d, J ) 2.0 Hz, 3-H), 7.47 (1H, d, J ) 8.5 Hz, 6-H), 7.42 (dd,
1H, J ) 2.2, 8.5 Hz, 5-H), 7.35 (dt, 2H, J ) 8.6, 2.2 Hz, 3′,5′-H),
7.18 (dt, 2H, 8.6, 2.2 Hz, 2′,6′-H), 4.61 (s, 2H, CH2), 2.14 (s, 3H,
CH3).
53
the method of Cheng and Prusoff and analyzed with GraphPad
Prism (GraphPad Software, Inc., San Diego, CA).
Calcium Channel Assay in SCG Neurons. Cannabinoid
Receptor Expression and Electrophysiology. N-terminally HA-
tagged rat CB1 cannabinoid receptor cDNA in pcDNA3 (from Ken
Mackie, University of Washington, Seattle, WA) was injected (100
ng/µL) into the nuclei of isolated rat superior cervical ganglion
neurons as previously described.13,54 The pEGFP-N1 plasmid (10
ng/µL) containing the coding sequence of enhanced green fluores-
cent protein (Clontech, Palo Alto, CA) was used as a co-injection
marker. After an overnight incubation, Ca2+ currents from injected
neurons were recorded in the whole-cell voltage-clamp mode at
room temperature (24-26 °C) with an Axopatch 200A patch-clamp
amplifier (Axon Instruments, Foster City, CA). The cell membrane
capacitance and series resistance were electronically compensated
to >80%. Whole-cell currents were low-pass filtered at 5 kHz using
the Bessel filter of the clamp amplifier.
Voltage-clamp protocols were generated with a Macintosh
computer (Apple Computer, Cupertino, CA) equipped with a PCI-
16 Host Interface card connected to an ITC-16 Data Acquisition
Interface (Instrutech Corp, Port Washington, NY) using Pulse
Control 5.0 XOPs (Bookman, R. J., Herrington, J. D., and Newton,
K. R., University of Miami, Miami, FL) with Igor Pro software
(WaveMetrics, Lake Oswego, OR). Ca2+ currents were elicited by
voltage steps from a holding potential of -80 mV and digitized at
180 µs per point. A double pulse protocol consisting of two 25 ms
steps to +5 mV was used to elicit Ca2+ currents. The second step
to +5 mV was preceded by a 50 ms step to +80 mV. Ca2+ current
amplitudes were isochronally measured 10 ms after the voltage step.
Ca2+ currents were isolated with an external solution that
contained (in mM) 140 tetraethylammonium methanesulfonate, 10
HEPES, 15 glucose, 10 CaCl2, 0.0001 tetrodotoxin at pH 7.4
(adjusted with methanesulfonic acid). The intracellular solution
contained (in mM) 120 N-methyl-D-glucamine, 20 tetraethylam-
monium chloride, 10 HEPES, 11 EGTA, 1 CaCl2, 4 MgATP, 0.1
Na2GTP, and 14 phosphocreatine at pH 7.2 (adjusted with meth-
anesulfonic acid). Stock solutions (10 mM) of WIN55,212-2
mesylate (Tocris), 1 (NIDA Drug Supply Program), 3, 4, 5, and 6
were prepared in dimethyl sulfoxide and stored at -20°C. On the
day of the experiment, stock solutions were diluted to 1 µM in
external solution and briefly sonicated to facilitate dispersion.
Neurons under study were superfused with control external solution
or drugs diluted in external solution from an array of glass tubes
5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-pyra-
zol-3-yl-methyl-triphenylphosphonium Bromide (4′). The pyra-
zole bromide 3′ (1.0 g, 2.33 mmol) and triphenylphosphine (0.84
g, 3.14 mmol) in 50 mL of dry toluene were heated at reflux under
dry nitrogen overnight. After cooling to ambient temperature, the
precipitate was filtered and dried in vacuo to afford 1.37 g (85%)
1
of the title salt as a white solid. H NMR (MeOD, 300 MHz) δ:
7.87-7.65 (m, 15 H, Ph3), 7.49 (d, 2H, J- 2.2 Hz, 3-H), 7.36-
7.28 (m, 3H, 5-H, 3′,5′-H), 7.16 (d, 1H, J ) 8.5 Hz, 6-H), 7.03 (d,
2H, J ) 8.4 Hz, 2′,6′-H), 4.87 (“d”, 2H, overlapped, CH2), 1.62 (s,
3H, CH3).
5-(4-Chlorophenyl)-3-[(E)-2-cyclohexylethenyl]-1-(2,4-dichlo-
rophenyl)-4-methyl-1H-pyrazole (2). The pyrazole phosphonium
salt 4′ (1.3 g, 1.88 mmol) dissolved in 100 mL of anhydrous THF
under dry nitrogen was cooled to 0 °C and treated with a 2 M
heptane/THF/ethylbenzene solution of LDA (0.940 mL, 1.88 mmol)
dropwise with stirring. Cyclohexanecarboxaldehyde (0.228 mL, 1.88
mmol) was added dropwise with stirring. After 15 min, the reaction
was warmed to ambient temperature and analyzed by TLC, which
showed the reaction to be complete (SiO2; acetone or 50% CH2-
Cl2/hexane; UV). The reaction was quenched with 30 mL of
saturated NH4Cl and extracted with CH2Cl2 (3×). The combined
organic layers were dried over Na2SO4 and the solvent removed in
vacuo. The residue was chromatographed on a Merck size B silica
gel column eluting with 40% CH2Cl2/hexane to yield 431 mg (52%)
of the title trans olefin. 1H NMR (MeOD, 300 MHz) δ: 7.55 (1H,
d, J ) 1.8 Hz, 3-H), 7.45 (1H, d, J ) 8.4 Hz, 6-H), 7.42 (dd, 1H,
J ) 2.1, 8.7 Hz, 5-H), 7.35 (dt, 2H, J ) 8.6, 2.2 Hz, 3′,5′-H), 7.17
(dt, 2H, 8.6, 2.2 Hz, 2′,6′-H), 6.40 (d*, 1H, J ) 15 Hz, vinyl 1-H),
6.34 (dd*, 1H, J ) 20, 3 Hz, vinyl 2-H), 2.16 (m, 1H, CH), 2.12
(s, 3H, CH3), 1.86-1.68 (m, 4H, CH(CH2)2), 1.40-1.21 (m, 6H,
CH2-CH2-CH2). 1H NMR (CDCl3, 60 mol % AgFOD, 300 MHz)
δ: 6.52 (dd, J ) 6.5, 16.2 Hz, vinyl 2-H), 6.40 (d, J ) 16.2 Hz,
vinyl 1-H); HRMS calcd for C24H23N2Cl3: 444.0927; observed:
444.0927. Elemental analysis: theory, C: 64.66, H: 5.20, N: 6.28.
Found, C: 64.66, H: 5.24, N: 6.24.

5984 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20
Hurst et al.
AgFOD ) (6,6,7,7,8,8,8-heptafluoro-2,2-dimethyl-3,5-octandi-
onato)silver; *second-order major/minor peaks with the minor peak
of the minor doublet were too weak to see (6.25 δ). See shift reagent
data for resolution of the resonances and definitive determination
of the coupling constants.
evolution and a darker orange color. After stirring for 15 min,
aldehyde 6′ (600 mg; 1.64 mmol) in 2 mL of dry THF was added,
turning the solution a dark purple color. The reaction was stirred
for an additional hour at 0 °C and then 1 h at room-temperature,
monitoring the progress by TLC (SiO2, 1:1 EtOAc/hex; PMA/Ce4+).
The reaction was dissolved in CH2Cl2 (25 mL) and then washed
with 15 mL of NH4Cl (aq) to achieve pH 8. The aqueous layer
was back extracted with 20 mL of CH2Cl2. The combined organics
were washed two times with brine and then dried over Na2SO4.
The filtered organics were evaporated to yield 1.26 g of brown
foam. Chromatography of the crude product on 60 g of SiO2 eluting
with 1 L of EtOAc/Hex (2:3), then 2 L of EtOAc/Hex (1:1), and
then 1 L of EtOAc, collecting a 125 mL forerun, and then, 20 mL
fractions afforded fractions 10-24 containing 327 mg (30%) pure
Z-enamine precursor and fractions 51-55 containing 273 mg (25%)
of pure E-enamine precursor.
Z-Enamine Precursor (8′). 1H NMR (CDCl3, 300 MHz) δ
7.94-7.88 (m, 2 H), 7.76-7.69 (m, 2 H), 7.48-7.42 (m, 4 H),
7.35 (d, J ) 2 Hz, 1H), 7.30-7.20 (m, 6 H), 6.99-6.96 (d, J ) 8
Hz, 1 H), 6.76 (d, J ) 2 Hz, 1 H), 5.36-5.31 (dd, J ) 9, 5 Hz,
1H), 4.23-4.19 (dd, J ) 9, 4 Hz, 1 H), 2.95-2.88 (m, 4 H), 1.96
(s, 3 H), 1.42-1.25 (m, 5 H). 31PNMR (CDCl3, 300 MHz) δ 30.17
ppm.
5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-pyra-
zole-3-carboxaldehyde (6′). Anhydrous CH2Cl2 (9 mL) and 2.0
M oxalyl chloride in CH2Cl2 (3.04 mL; 6.08 mmol) were cooled
to -78 °C with magnetic stirring in an oven dried, 100 mL septum
sealed RB Flask under dry nitrogen. Neat DMSO (90.86 mL, 12.1
mmol) was added dropwise over 2 min with stirring. After 2 min,
pyrazole alcohol 2′ (2.03 g, 5.52 mmol) in 7 mL of CH2Cl2 was
added by syringe over a period of 4 min. The reaction was brought
to -15 °C in an ice/salt bath and stirred for 15 min affording a
white precipitate in a pink solution. Triethylamine (3.83 mL, 27.5
mmol) was added to the slurry dropwise and the yellow-green slurry
allowed to warm to ambient temperature over 1 h. The reaction
was treated with water (50 mL) and CH2Cl2 (30 mL) and
partitioned. The aqueous layer was extracted a second time with
CH2Cl2 (50 mL). The combined organic layers were washed with
brine and dried over Na2SO4. Filtration and evaporation of the
solvent from the filtrate in vacuo gave 1.94 g of a yellow, crisp
foam. Chromatography on silica gel (50:1) eluting with 15% ether/
hexane gave 0.81 g (40%) of the pure aldehyde, which could be
crystallized from CH2Cl2 or ether/hexane; mp. 116.2-117.5 °C.
TLC (SG-60; 30% Et2O-hexane; UV), Rf ) 0.33. GC: DB-17,
E-Enamine Precursor (9′). 1H NMR (CDCl3, 300 MHz) δ
8.17-8.11 ppm (m, 2 H), 7.99-7.92 (m, 2H), 7.56-7.42 (m, 7H),
7.35 (d, J ) 2 Hz, 1H), 7.27-7.22 (m, 5H), 7.04-7.01 (d, J ) 9
Hz, 2 H), 5.98 (bs, 1 H), 5.62-5.56 (t, J ) 9 Hz, 1 H), 4.30-4.22
(dd, J ) 13 Hz, 1H), 2.61-2.58 (m, 2H), 2.33-2.29 (m, 2H), 2.22
(s, 3H), 1.28-1.14 (m, 7H). 31PNMR (CDCl3, 300 MHz) δ 34.09
ppm.
1
240 °C, FID, (99.7%). H NMR (300 MHz, CDCl3) δ 10.13 (s,
1H, CHO), 7.46-7.45 (m, 1H, 3-H); 7.35-7.28 (m, 4H, 2′,6′, 5-,
6-H), 7.07 (d, 2H, 8.6 Hz, 3′-, 5′-H), 2.35 (s, 3H, CH3).
5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-3-[(E)-pi-
peridinoiminomethyl]-1H-pyrazole (5). Aldehyde 6′ (192 mg,
0.526 mmol) in 4 mL of dry CH2Cl2 (from basic Al2O3) was treated
with anhydrous 1-aminopiperidine (59 mL, 0.546 mmol) while
stirring under a dry nitrogen atmosphere at ambient temperature.
After stirring overnight, 19 mg of MgSO4 was added followed in
3 h with an additional 30 mg of MgSO4. After 10 min, the reaction
was filtered and evaporated in vacuo to yield 222 mg of a crisp,
white foam that contained product 5 with only 5% of the starting
6′ (1H NMR). Recrystalization from minimal CH2Cl2 followed by
the addition of hexanes or from hexanes/Et2O (7:3) afforded white,
spiked crystals (94 mg); mp 177-178 °C. Column chromatography
of the remaining mother liquor on silica gel (50:1) eluting with
15% ether-hexanes, followed by recrystallization afforded another
83 mg of spiked crystals, for a total of 177 mg (0.370 mmol, 70%
yield).
1-[2-(5-(4-Chlorophenyl)-1(2,4-dichlorophenyl)-4-methyl-1H-
pyrazole-3-yl)-(E)-vinyl]-piperidine (6) and -(Z)-vinyl]-piperi-
dine (10′). (A) An oven dried 5 mL Reactivial was charged with
30% w/w KH/mineral oil (85 mg, 0.64 mmol) under a nitrogen
atmosphere, and KH was washed with anhydrous THF (dried
through basic Al2O3) to remove the oil. To the KH was added the
unpurified diphenylphosphinoyl diastereomers 8′ and 9′ (410 mg;
0.62 mmol) in 2 mL of THF with stirring under dry nitrogen at
ambient temperature. After 10 min of gas evolution, the vial was
capped under nitrogen, stirred for 1 h, and then heated at 45° C for
30 min. The THF was evaporated and partitioned between hexane
and water affording a difficult emulsion, which was dispelled with
saturated brine. An insoluble material separated in the organic phase
and was removed by filtration. In retrospect, this was likely some
of product 6. The filtrate was dried over Na2SO4 and evaporated
in vacuo to afford a yellow resin, which afforded both an insoluble
(68 mg) and a soluble (117 mg) fraction when triturated with
hexane, both of which were identified as enamine 6 (67% combined
yield) by NMR. Crystallization from CDCl3/Et2O afforded the
product with mp 187.6-188.1 0 °C.
(B) An oven dried 5 mL Reactivial was charged with 30% w/w
KH/mineral oil (45 mg, 0.34 mmol) under a nitrogen atmosphere,
and KH was washed (2 × 1 mL) with anhydrous THF (dried
through basic Al2O3) to remove the oil. To the KH suspended in 1
mL THF was added the Z-enamine precursor diphenylphosphinoyl
diasteriomers 8′ (150 mg; 0.23 mmol) in 2 mL of THF under dry
nitrogen with mixing at ambient temperature. After gas evolution
subsided, the reaction became viscous to the extent that it trapped
gas bubbles. The capped vial was tapped to dislodge the bubbles
and restore stirring. The red mixture with a thick precipitate was
stirred for 2 h and then filtered through glass wool followed by a
Millipore filter. The THF filtrate was rotary evaporated in vacuo
to afford 84 mg (83%) of a yellow resin, which was stored at -72°
C overnight. The product was identified as enamine 10′ by a
ROESY NMR experiment, which showed a through space interac-
tion between the vinyl hydrogens that is only possible for the
Z-enamine. Also, one of the vinyl hydrogens showed a through
space interaction with the pyrazole methyl, whereas the other vinyl
hydrogen interacted with the methylenes R to the piperidine nitrogen
consistent with the Z-configuration. By 20 h, Z-enamine 10′
converted to E-enamine 6 in the CDCl3 solution.
1H NMR (CDCl3, 300 MHz) δ 7.76 ppm (s, 1 H, CHdN), 7.39
(d, J ) 2.0 Hz, 1H, 3-H), 7.28 (d, 3 H, 2′-,6′-,6-H), 7.25 (dd, 1H,
J ) 8.5, 2.0 Hz, 5-H); 7.08 (d, 2H, J ) 8.5 Hz, 3′-,5′-H), 3.16 (m,
4H, NCH2), 2.30 (s, 3H, CH3), 1.75 (m, 4 H, N-CH2-CH2), 1.55
(m, 2 H, N-CH2-CH2-CH2). CMR (CDCl3, 300 MHz) δ 10.5 (q),
24.5 (t), 25.5 (t), 52.2 (t), 114.5 (s), 128.1 (d). 128.4 (s), 129.1 (d),
130.1 (d) 130.5 (d), 131.1 (d), 131.2 (d), 133.4 (s), 134.8 (s), 135.6
(s), 136.9 (s), 142.5 (s), 149.7 (s). Nonequivalent carbons designated
d (CH), q (CH3), t (CH2), and s (quaternary) were derived from
comparison of CMR and DEPT spectra.
ROESY (benzene d6): a cross-peak for δ 7.76 and 3.16 identifies
a through space interaction indicative of an E-geometry of the imine
double bond. MS ) m/e (446 ) M+, 364 ) M - 84 + 2, 328 )
364-HCl, 84 ) base C5H10N). Elemental analysis: theory, C: 59.01,
H: 4.73, N: 12.51. Found, C: 59.30, H: 4.82, N: 12.55.
Stability: no change in the 1H NMR spectra of 5 in MeOD (22 h);
MeOD-20% D2O (+ 22 h, with crystallization); MeOD-D2O-
CDCl3 (+ 4 days, homogeneous solution).
1[5-(4-Chlorophenyl)-1(2,4-dichlorophenyl)-4-methyl-1H-pyra-
zole-3-yl]-2-diphenylphosphinoyl-2-piperidinyl-ethanol (8′ and
9′). (Piperdinylmethyl)diphenylphosphine oxide (614 mg; 2.05
mmol) and a few crystals of 1,10 phenanthroline indicator was
dissolved /suspended in 2 mL of dry THF under dry nitrogen. A
few drops of n-BuLi (1.6 M in hexanes) were added at 0 °C with
stirring until an orange color persisted. This was followed by an
additional amount of n-BuLi (1.25 mL, 2.0 mmol) affording gas

Biarylpyrazole InVerse Agonists
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20 5985
The compound is unstable on silica gel affording the homoal-
dehyde that would result from hydrolysis of the enamine. The
compound is stable in 20% water/THF (both deuterated) at ambient
temperature for at least four days (NMR). In deuterated water (20%
in THF), the enamine â-proton exchanges 40% in 1 h and
completely by 20 h with a resulting collapse of the R-proton doublet
to a singlet. The original NMR spectrum can be regenerated (return
of â-proton and R-proton doublets) by washing the product in
(m, 3H, c-hex H3). 13CNMR (CDCl3, 300 MHz) δ 10.8, 25.0, 25.8,
33.3, 48.8, 115.3, 128.4, 129.2, 129.8, 130.3, 130.5, 131.5, 132.8,
134.6, 136.0, 137.4, 138.3, 151.5, 159.3.
5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-pyra-
zole-3-(N-cyclohexyl, N-methylcarboxamide) (4). The reported
acid chloride 12′ (200 mg; 0.46 mmol) under dry nitrogen was
dissolved in 10 mL of CH2Cl2 (dried through basic Al2O3) in a
septum sealed flask, treated with N-cyclohexyl, N-methylamine
(dried through basic Al2O3)(90 µL; 0.69 mmol) followed by dry
triethylamine (190 µL; 1.88 mmol) at ambient temperature. The
solution was stirred under a dry nitrogen atmosphere for 3 h in a
septum sealed flask. The reaction was worked up by dilution with
20 mL of CH2Cl2, washing the organic layer with 35 mL of 3%
aqueous HCl saturated NaHCO3, and brine to pH 1. Each aqueous
layer was back extracted with CH2Cl2, which was added to the
organic phase before the next washing. Drying the organic layer
with Na2SO4, filtration, and evaporation in vacuo gave 235 mg of
a foam/resin. Chromatography on silica gel (6.5 g) eluting with
10% EtOAc-hexanes (150 mL) followed by 25% EtOAc-hexanes
(100 mL) collecting 5 mL fractions afforded 140 mg (64% Th) of
the title compound in fractions 22-31. Recrystallization from
ether-hexanes provided 113 mg of white crystals; mp 161.6-162.8
°C. Elemental analysis: theory, C: 60.45%, H 5.07%, N 8.81%.
Found, C: 60.45%, H 5.06%, N: 8.70%.
1
CDCl3 with H2O.
1
E-Enamine (6). H NMR (CDCl3, 300 MHz) δ 7.35-7.36 (d,
J ) 2.2 Hz, 1 H, Ar-H); 7.04-7.05 (d, J ) 8.3 Hz, 2 H, Ar-H);
6.99-7.02 (d, J ) 14.0 Hz, 1 H, olefin R to piperidine); 5.19-
5.23 (d, J ) 14.0 Hz, 1 H, olefin â to piperidine); 3.05-3.07 (d,
J ) 5.4 Hz, 4 H, N(CH2)2); 2.07 (s, 3 H, pyr-Me); 1.59 (bs, 6 H,
(CH2)3). 13CMR (CDCl3, 300 MHz) δ 151.79, 141.80, 141.23,
137.00, 134.63, 133.99, 133.17, 130.96, 130.60, 130.05, 128.84,
128.60, 127.61, 111.84, 87.53, 49.37, 25.29, 24.36, 9.11. HPLC
(RP C-18 Waters Resolve radial compression column 10 µ 8 mm
× 10 cm; 70% CH3CN/water; 280 nm detection; 2.0 mL/min flow
rate) Rt ) 5.7 min, 100%. Elemental analysis: theory, C: 61.83,
H: 4.96, N: 9.40. Found, C: 61.87, H: 5.04, N: 9.27.
1
Z-Enamine (10′). H NMR (CDCl3, 300 MHz) δ 7.37 ppm (s,
1 H, Ar-H); 7.23-7.25 (d, J ) 10.0 Hz, 2 H, Ar-H); 7.04-7.05
(d, J ) 8.3 Hz,2 H, Ar-H); 5.99-6.01 (d, J ) 10.0 Hz,1 H, H R
to piperidine ring); 4.77-4.79 (d, J ) 10.0 Hz,1H, H R to pyrazole
ring); 3.14 (m, 4 H, (CH2)2N); 2.03 (s, 3 H, Me); 1.51 (m, 6 H,
(CH2)3).
TLC (SiO2; 1:4 Acetone/CH2Cl2 Rf ) 0.7 and1:1 EtOAc/hexane
Rf ) 0.5; PMA/ Ce4+)>99%. HPLC (RP C-18 Waters Radial Nova
Pak; 4 µ, 8 mm × 10 cm; 80% CH3CN/H2O; 280 nm) 100%, Rt )
8.5 min, at 2 mL/min.
5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-pyra-
zole-3-(N-cyclohexylcarboxamide) (3). The reported acid chloride
12′ (240 mg; 0.55 mmol) dissolved in 10 mL of CH2Cl2 (dried
through basic Al2O3) was treated with cyclohexylamine (dried
through basic Al2O3) (89 µL; 0.825 mmol) at ambient temperature,
followed by dry triethylamine (230 µL; 1.65 mmol) with stirring
under a dry nitrogen atmosphere. After 1 h, the reaction was
observed by TLC (SiO2; EtOAc/hexane 1:3, UV) to have a visible
yellow spot at the Rf of 12′, which did not diminish with either the
addition of further cyclohexylamine or upon aqueous work up. The
spot was attributed to an impurity in the unpurified 12′. The reaction
was worked up by dilution with 40 mL of CH2Cl2, washing the
organic layer with 50 mL of 5% aqueous HCl, saturated NaHCO3,
and brine. Drying the organic layer with Na2SO4, filtration, and
evaporation in vacuo gave 300 mg of a foam/resin.
1H NMR (CDCl3, 500 MHz) δ 7.45-7.48 (dd, 1 H, J ) 13.18,
1.95 Hz, Ar-H); 7.30-7.32 (m, 2 H, Ar-H); 7.08-7.27 (m, 4 H,
Ar-H); 4.58-4.60 (m, CH, rotamer 46%); 4.04-4.08 (m, CH
rotamer 54%); 3.08 (s, CH3, rotamer 46%), 3.03 (s, CH3 rotamer
54%); 2.19 (s, 3 H, CH3); 1.79-1.84 (m, 4 H, CH2); 1.48-1.70
(m, 4 H, CH2); 1.08-1.24 (m, 2 H, CH2). CMR (CDCl3, 300 MHz)
δ 9.4, 25.8, 26.07, 26.14, 27.8, 30.1, 31.5, 31.8, 58.3, 116.2, 116.6,
128.0, 128.2, 129.3, 130.6, 131.0, 133.7, 135.1, 136.0, 136.6, 142.0,
148.1, 164.7, 165.4.
Acknowledgment. We gratefully acknowledge the research
support from National Institutes of Health/National Institute on
Drug Abuse Grants DA03934 and DA000489 to P.R.; DA09978
and DA05274 to M.A.; and DA10350 to D.L.
Marginal chromatography on silica gel (20:1) eluting with 10%
EtOAc-hexanes afforded mixed fractions, by HPLC (RP-C18
Novapak column, CH3CN/H2O (3:1), UV detection 280 nm), which
when recrystallized from ether-hexane afforded 8 mg of the 1,3-
isomer of 3. Dissolving the remaining material in CH3CN/H2O (4:
1) resulted in crystallization that afforded 133 mg of 3. The
noncrystalline fractions and mother liquors were combined and
chromatographed on a Merck size A Lobar prepacked C-18 column
eluting with CH3CN/H2O (4:1), yielding clean separation of the
isomers: 1 mg of the 1,3-isomer of 3 and 22 mg of crystalline 3.
The total yield of 3 was 155 mg (61%) and 8 mg (3.5%) of the
1,3-isomer of 3.
Supporting Information Available: Experimental details of
compounds 3-6. This material is available free of charge via the
Internet at http://pubs.acs.org.
References
(1) Hurst, D. P.; Lynch, D. L.; Barnett-Norris, J.; Hyatt, S. M.; Seltzman,
H. H.; et al. N-(piperidin-1-yl)-5-(4-chlorophenyl)-1-(2,4-dichlo-
rophenyl)-4-methyl-H-pyrazole-3-carboxamide (SR141716A) interac-
tion with LYS 3.28(192) is crucial for its inverse agonism at the
cannabinoid CB1 receptor. Mol. Pharmacol. 2002, 62, 1274-1287.
(2) Felder, C. C.; Joyce, K. E.; Briley, E. M.; Mansouri, J.; Mackie, K.;
et al. Comparison of the pharmacology and signal transduction of
the human cannabinoid CB1 and CB2 receptors. Mol. Pharmacol.
1995, 48, 443-450.
(3) Mackie, K.; Lai, Y.; Westenbroek, R.; Mitchell, R. Cannabinoids
activate an inwardly rectifying potassium conductance and inhibit
Q-type calcium currents in AtT20 cells transfected with rat brain
cannabinoid receptor. J. Neurosci. 1995, 15, 6552-6561.
(4) Pan, X.; Ikeda, S. R.; Lewis, D. L. Rat brain cannabinoid receptor
modulates N-type Ca2+ channels in a neuronal expression system.
Mol. Pharmacol. 1996, 49, 707-714.
(5) Rinaldi-Carmona, M.; Barth, F.; Heaulme, M.; Shire, D.; Calandra,
B.; et al. SR141716A, a potent and selective antagonist of the brain
cannabinoid receptor. FEBS Lett. 1994, 350, 240-244.
(6) Lange, J. H.; Kruse, C. G. Keynote review: Medicinal chemistry
strategies to CB1 cannabinoid receptor antagonists. Drug DiscoVery
Today 2005, 10, 693-702.
(7) Jagerovic, N.; Hernandez-Folgado, L.; Alkorta, I.; Goya, P.; Navarro,
M.; et al. Discovery of 5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-
3-hexyl-1h-1,2,4-triazole, a novel in vivo cannabinoid antagonist
containing a 1,2,4-triazole motif. J. Med. Chem. 2004, 47, 2939-
2942.
Compound 3. 1H NMR (300 MHz, MeOD) δ 7.57 (d, 1H, J )
2.2 Hz, 3-H); 7.54 (d, 1H, J ) 8.5 Hz, 6-H); 7.45 (dd, 1H, J )
8.5, 2.2 Hz, 5-H), 7.37 (d, 2H, J ) 8.5 Hz, 2′-,6′-H); 7.19 (d, 2H,
J ) 8.5 Hz, 3′-,5′-H); 3.86 (m, 1H, N-CH); 2.30 (s, 3H, CH3);
1.96-1.64 (m, 5H, c-hex H5); 1.48-1.21 (m, 5H, c-hex H′5). CMR
(CDCl3, 300 MHz) δ 9.8, 25.4, 26.0, 33.6, 48.3, 118.2, 127.8, 128.2,
129.2, 130.7, 131.0, 131.2, 133.4, 135.2, 136.3, 136.4, 143.4, 145.7,
162.2. M.p. 124.3-125.7 °C. TLC (SiO2; 1:1 EtOAc:Hex; PMA/
Ce4+) Rf ) 0.79. HPLC RP C-18 Waters Radial Nova Pak; 4 µ, 8
mm × 10 cm; 75% CH3CN: H2O; 280 nm) 100%, Rt ) 13 min,
at 2 mL/min.
1
1,3-Isomer of 3. H NMR (300 MHz, CDCl3) δ 7.61 (d, 2H, J
) 8.5 Hz, 2′-,6′-H); 7.51 (d, 1H, J ) 2.2 Hz, 3-H); 7.47 (d, 1H, J
) 8.5 Hz, 6-H); 7.42 (d, 2H, J ) 8.5 Hz, 5-H); 7.38 (dd, 1H, J )
8.5, 2.2 Hz, 3′-,5′-H); 5.55 (d, 1H, J ) 8.2 Hz, NH); 3.88 (m, 1H,
N-CH); 2.41 (s, 3H, CH3); 1.92-1.88 (m, 2H, c-hex H2); 1.70-
1.58 (m, 3H, c-hex H3); 1.43-1.31 (m, 2H, c-hex H2); 1.24-1.09

5986 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20
Hurst et al.
(8) Lange, J. H.; van Stuivenberg, H. H.; Coolen, H. K.; Adolfs, T. J.;
McCreary, A. C.; et al. Bioisosteric replacements of the pyrazole
moiety of rimonabant: synthesis, biological properties, and molecular
modeling investigations of thiazoles, triazoles, and imidazoles as
potent and selective CB1 cannabinoid receptor antagonists. J. Med.
Chem. 2005, 48, 1823-1838.
(9) Lange, J. H.; Coolen, H. K.; van Stuivenberg, H. H.; Dijksman, J.
A.; Herremans, A. H.; et al. Synthesis, biological properties, and
molecular modeling investigations of novel 3,4-diarylpyrazolines as
potent and selective CB(1) cannabinoid receptor antagonists. J. Med.
Chem. 2004, 47, 627-643.
(10) Plummer, C. W.; Finke, P. E.; Mills, S. G.; Wang, J.; Tong, X.; et
al. Synthesis and activity of 4,5-diarylimidazoles as human CB1
receptor inverse agonists. Bioorg. Med. Chem. Lett. 2005, 15, 1441-
1446.
(11) Muccioli, G. G.; Wouters, J.; Charlier, C.; Scriba, G. K.; Pizza, T.;
et al. Synthesis and activity of 1,3,5-triphenylimidazolidine-2,4-diones
and 1,3,5-triphenyl-2-thioxoimidazolidin-4-ones: characterization of
new CB1 cannabinoid receptor inverse agonists/antagonists. J. Med.
Chem. 2006, 49, 872-882.
(12) Meurer, L. C.; Finke, P. E.; Mills, S. G.; Walsh, T. F.; Toupence, R.
B.; et al. Synthesis and SAR of 5,6-diarylpyridines as human CB1
inverse agonists. Bioorg. Med. Chem. Lett. 2005, 15, 645-651.
(13) Pan, X.; Ikeda, S. R.; Lewis, D. L. SR 141716A acts as an inverse
agonist to increase neuronal voltage-dependent Ca2+ currents by
reversal of tonic CB1 cannabinoid receptor activity. Mol. Pharmacol.
1998, 54, 1064-1072.
(14) Meschler, J. P.; Kraichely, D. M.; Wilken, G. H.; Howlett, A. C.
Inverse agonist properties of N-(piperidin-1-yl)-5-(4-chlorophenyl)-
1-(2, 4-dichlorophenyl)-4-methyl-1H-pyrazole-3-carboxamide HCl
(SR141716A) and 1-(2-chlorophenyl)-4-cyano-5-(4-methoxyphenyl)-
1H-pyrazole-3-carboxyl ic acid phenylamide (CP-272871) for the CB-
(1) cannabinoid receptor. Biochem. Pharmacol. 2000, 60, 1315-
1323.
(15) Schlicker, E.; Kathmann, M. Modulation of transmitter release via
presynaptic cannabinoid receptors. Trends Pharmacol. Sci. 2001, 22,
565-572.
(16) Pertwee, R. G. Inverse agonism and neutral antagonism at cannabinoid
CB1 receptors. Life Sci. 2005, 76, 1307-1324.
(17) Bouaboula, M.; Perrachon, S.; Milligan, L.; Canat, X.; Rinaldi-
Carmona, M.; et al. A selective inverse agonist for central cannabinoid
receptor inhibits mitogen-activated protein kinase activation stimu-
lated by insulin or insulin-like growth factor 1. Evidence for a new
model of receptor/ligand interactions. J. Biol. Chem. 1997, 272,
22330-22339.
(18) Leff, P. The two-state model of receptor activation. Trends Phar-
macol. Sci. 1995, 16, 89-97.
(19) McAllister, S. D.; Rizvi, G.; Anavi-Goffer, S.; Hurst, D. P.; Barnett-
Norris, J.; et al. An aromatic microdomain at the cannabinoid CB(1)
receptor constitutes an agonist/inverse agonist binding region. J. Med.
Chem. 2003, 46, 5139-5152.
(20) Ballesteros, J. A.; Jensen, A. D.; Liapakis, G.; Rasmussen, S. G.;
Shi, L.; et al. Activation of the beta 2-adrenergic receptor involves
disruption of an ionic lock between the cytoplasmic ends of
transmembrane segments 3 and 6. J. Biol. Chem. 2001, 276, 29171-
29177.
(21) Visiers, I.; Ebersole, B. J.; Dracheva, S.; Ballesteros, J.; Sealfon, S.
C.; et al. Structural motifs as functional microdomains in G-protein-
coupled receptors: Energetic considerations in the mechanism of
activation of the serotonin 5-HT2a receptor by disruption of the ionic
lock of the arginine cage. Int. J. Quantum Chem. 2002, 88, 65-75.
(22) Fay, J. F.; Dunham, T. D.; Farrens, D. L. Cysteine residues in the
human cannabinoid receptor: only C257 and C264 are required for
a functional receptor, and steric bulk at C386 impairs antagonist
SR141716A binding. Biochemistry 2005, 44, 8757-8769.
(23) Shire, D.; Calandra, B.; Bouaboula, M.; Barth, F.; Rinaldi-Carmona,
M.; et al. Cannabinoid receptor interactions with the antagonists SR
141716A and SR 144528. Life Sci. 1999, 65, 627-635.
(24) Barnett-Norris, J.; Hurst, D. P.; Buehner, K.; Ballesteros, J. A.;
Guarnieri, F.; et al. Agonist alkyl tail interaction with cannabinoid
CB1 receptor V6.43/I6.46 groove induces a Helix 6 active conforma-
tion. Int. J. Quantum Chem. 2002, 88, 76-86.
(25) Jensen, A. D.; Guarnieri, F.; Rasmussen, S. G.; Asmar, F.; Ballesteros,
J. A.; et al. Agonist-induced conformational changes at the cyto-
plasmic side of transmembrane segment 6 in the beta 2 adrenergic
receptor mapped by site-selective fluorescent labeling. J. Biol. Chem.
2001, 276, 9279-9290.
(26) Ghanouni, P.; Steenhuis, J. J.; Farrens, D. L.; Kobilka, B. K. Agonist-
induced conformational changes in the G-protein-coupling domain
of the beta 2 adrenergic receptor. Proc. Natl. Acad. Sci. U.S.A. 2001,
98, 5997-6002.
(27) McAllister, S. D.; Tao, Q.; Barnett-Norris, J.; Buehner, K.; Hurst,
D. P.; et al. A critical role for a tyrosine residue in the cannabinoid
receptors for ligand recognition. Biochem. Pharmacol. 2002, 63,
2121-2136.
(28) Barth, F.; Casellas, P.; Congy, C.; Martinez, S.; Rinaldi, M.; et al.
Substituted N-Piperidino-3-Pyrazolecarboxamide. Eur. Pat. Appl.
656354, A1, 1995.
(29) Carey, F. A.; Sundberg, R. J. AdVanced Organic Chemistry Part B:
Reactions and Synthesis; 3rd ed.; Plenum Press: New York and
London, 1990; p 98.
(30) Vedejs, E.; Peterson, M. J. Stereochemistry and Mechanism in the
Wittig Reaction. Topics in Stereochemistry; John Wiley & Sons: New
York, 1994; pp 1-157.
(31) Broekhof, N. L. J. M.; Jonkers, F. L.; van der Gen, A. Enamine
synthesis by the Horner-Wittig Reaction. Tetrahedron Lett. 1979,
20, 2433-2436.
(32) Broekhof, N. L. J. M.; Elburg, P. v.; Gen, A. v. d. Recl.: J. R. Neth.
Chem. Soc. 1984, 103, 312-316.
(33) Barth, F.; Casellas, P.; Congy, C.; Martinez, S.; Rinaldi, M.; et al.
Pyrazole derivatives as cannabinoid receptor agonists. CAPLUS;
Sanofi: US 5624941 A 19970429 CAN 127:5085 AN 1997:311258,
1997.
(34) Seltzman, H. H.; Carroll, F. I.; Burgess, J. P.; Wyrick, C. D.; Burch,
D. F. Synthesis, spectral studies and tritiation of the cannabinoid
antagonist SR141716A. J. Chem. Soc., Chem. Commun. 1995, 1549-
1550.
(35) Seltzman, H. H.; Carroll, F. I.; Burgess, J. P.; Wyrick, C. D.; Burch,
D. F. Tritiation of SR141716 by metalation-iodination-reduction:
tritium-proton NOE study. J. Labelled Compd. Radiopharm. 2002,
45, 59-70.
(36) Lan, R.; Liu, Q.; Fan, P.; Lin, S.; Fernando, S. R.; et al. Structure-
activity relationships of pyrazole derivatives as cannabinoid receptor
antagonists. J. Med. Chem. 1999, 42, 769-776.
(37) Perola, E.; Charifson, P. S. Conformational analysis of drug-like
molecules bound to proteins: An extensive study of ligand reorga-
nization upon binding. J. Med. Chem. 2004, 47, 2499-2510.
(38) Strange, P. G. Mechanisms of inverse agonism at G-protein-coupled
receptors. Trends Pharmacol. Sci. 2002, 23, 89-95.
(39) Lin, S. W.; Sakmar, T. P. Specific tryptophan UV-absorbance changes
are probes of the transition of rhodopsin to its active state.
Biochemistry 1996, 35, 11149-11159.
(40) Song, Z. H.; Slowey, C. A.; Hurst, D. P.; Reggio, P. H. The difference
between the CB(1) and CB(2) cannabinoid receptors at position 5.46
is crucial for the selectivity of WIN55212-2 for CB(2). Mol.
Pharmacol. 1999, 56, 834-840.
(41) Song, Z. H.; Bonner, T. I. A lysine residue of the cannabinoid receptor
is critical for receptor recognition by several agonists but not
WIN55212-2. Mol. Pharmacol. 1996, 49, 891-896.
(42) Ballesteros, J. A.; Weinstein, H. Integrated Methods for the Construc-
tion of Three-Dimensional Models and Computational Probing of
Structure Function Relations in G Protein-Coupled Receptors.
Methods in Neuroscience; Academic Press: San Diego, CA, 1995;
pp 366-428; Chapter 319.
(43) Palczewski, K.; Kumasaka, T.; Hori, T.; Behnke, C. A.; Motoshima,
H.; et al. Crystal structure of rhodopsin: A G protein-coupled
receptor. Science 2000, 289, 739-745.
(44) Gerard, C. M.; Mollereau, C.; Vassart, G.; Parmentier, M. Molecular
cloning of a human cannabinoid receptor which is also expressed in
testis. Biochem. J. 1991, 279, 129-134.
(45) Bramblett, R. D.; Panu, A. M.; Ballesteros, J. A.; Reggio, P. H.
Construction of a 3D model of the cannabinoid CB1 receptor:
Determination of helix ends and helix orientation. Life Sci. 1995,
56, 1971-1982.
(46) Javitch, J. A.; Fu, D.; Liapakis, G.; Chen, J. Constitutive activation
of the beta2 adrenergic receptor alters the orientation of its sixth
membrane-spanning segment. J. Biol. Chem. 1997, 272, 18546-
18549.
(47) Ward, S. D.; Hamdan, F. F.; Bloodworth, L. M.; Siddiqui, N. A.;
Li, J. H.; et al. Use of an in situ disulfide cross-linking strategy to
study the dynamic properties of the cytoplasmic end of transmem-
brane domain VI of the M(3) muscarinic acetylcholine receptor.
Biochemistry 2006, 45, 676-685.
(48) Farrens, D. L.; Altenbach, C.; Yang, K.; Hubbell, W. L.; Khorana,
H. G. Requirement of rigid-body motion of transmembrane helicies
for light-activation of rhodopsin. Science 1996, 274, 768-770.
(49) Burley, S. K.; Petsko, G. A. Aromatic-aromatic interaction: a mech-
anism of protein structure stabilization. Science 1985, 229, 23-28.
(50) Hunter, C. A.; Singh, J.; Thornton, J. M. Pi-pi interactions: the
geometry and energetics of phenylalanine-phenylalanine interactions
in proteins. J. Mol. Biol. 1991, 218, 837-846.

Biarylpyrazole InVerse Agonists
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 20 5987
(51) Tao, Q.; McAllister, S. D.; Andreassi, J.; Nowell, K. W.; Cabral, G.
A.; et al. Role of a conserved lysine residue in the peripheral
cannabinoid receptor (CB2): evidence for subtype specificity. Mol.
Pharmacol. 1999, 55, 605-613.
(52) Tao, Q.; Abood, M. E. Mutation of a highly conserved aspartate
residue in the second transmembrane domain of the cannabinoid
receptors, CB1 and CB2, disrupts G-protein coupling. J. Pharmacol.
Exp. Ther. 1998, 285, 651-658.
(53) Cheng, Y.; Prusoff, W. H. Relationship between the inhibition
constant (Ki) and the concentration of inhibitor which causes 50%
inhibition (IC50) of an enzymatic reaction. Biochem. Pharmacol.
1973, 22, 3099-3108.
(54) Vasquez, C.; Lewis, D. L. The CB1 cannabinoid receptor can
sequester G-proteins, making them unavailable to couple to other
receptors. J. Neurosci. 1999, 19, 9271-9280.
JM060446B