Molecular Biology of Cannabinoid Receptors 894
Ligand Recognition at the CB 1 Receptor
4.1
The Aminoalklylindole/SR141716A Binding Region
Mutation studies as well as studies with novel ligands have suggested a separation
of the binding site for aminoalkylindoles (typified by WIN 55,212-2) from that
of the other three classes of cannabinoid agonist ligands (Table 2) (Chin et al.
1998; Song and Bonner 1996; Tao et al. 1999). A K3.28(192)A mutation of CB 1
results in no loss of affinity or efficacy for WIN 55,212-2, but greater than 1,000-
fold loss in affinity and efficacy for HU-210, CP 55,940, and anandamide (Chin
et al. 1998; Song and Bonner 1996), and a 17-fold loss for SR141716A (Hurst et
al. 2002). The CB 2 selectivity of WIN 55,212-2 (Felder et al. 1995; Showalter et al.
1996) may be due to the presence of an additional TM helix (TMH)5 aromatic
residue, F5.46 in the CB 2 receptor (Song et al. 1999). Receptor chimera studies of
the CB 1 and CB 2 receptors have demonstrated that the region delimited by the
fourth and fifth TM domains of the CB 1 receptor is crucial for the binding of
the CB 1 receptor antagonist SR141716A, but not CP 55,940, and that this same
region in the CB 2 receptor is crucial for the binding of WIN 55,212-2 and the CB 2
receptor antagonist SR144528 (Shire et al. 1996a, 1999). These results reinforce the
hypothesis that the aminoalkylindole-binding region at the CB 1 receptor is in the
TMH 3-4-5 region and is not identical to that for other CB agonists. Furthermore,
these results suggest that SR141716A binding shares the aminoalkylindole binding
region but also interacts with K(3.28)192.
In addition, the carbonyl oxygen as well as the morpholino ring of the amino-
alkylindolescanbereplacedwithoutaffectingaffinity;thereforehydrogenbonding
may not be the primary interaction of these compounds at the CB 1 receptor (Huff-
man 1999; Huffman et al. 1994; Kumar et al. 1995; Reggio 1999). Huffman et al.
(1994) also reported that the replacement of the naphthyl ring of WIN 55,212-2
withanalkyloralkenylgroupresultedincompletelossofCB 1 receptor affinity
(Ki>10,000 nM in both cases). The fact that the carbonyl oxygen or the morpholino
ring of the aminoalkylindoles can be removed without significant effect, along with
evidence that the presence of the carbonyl and morpholino group (in the absence
of an aryl substituent) is insufficient to produce CB 1 affinity, suggests that aro-
matic stacking, rather than hydrogen bonding, may be the primary interaction for
aminoalkylindoles at the CB 1 receptor.
Aromatic–aromatic stacking interactions are significant contributors to protein
structure stabilization (Burley and Petsko 1985). Modeling studies indicate that in
theactivestate(R*)modelofCB 1 , there is a patch of aromatic amino acids in the
TMH 3-4-5 region with which WIN 55,212-2 can interact (McAllister et al. 2003).
There is an upper (extracellular side) stack formed by F3.25(189 in human CB 1 ,
190 in mouse CB 1 ), W4.64(255/256), Y5.39(275/276), and W5.43(279/280). When
WIN 55,212-2 is computationally docked to interact with this patch, it also can
interact with a lower (towards intracellular side) aromatic residue, F3.36(200/201).
In this docking position, WIN 55,212-2 creates a continuous aromatic stack over