Cannabinoid Receptors and Their Ligands: Ligand–Ligand and Ligand–Receptor Modeling Approaches 269Most recently, Huffman synthesized a series of 1-pentyl-1H-indol-3-yl-(1-naph-
thyl)methanes and 2-methyl-1-pentyl-1H-indol-3-yl-(1-naphthyl)methanes to in-
vestigate the hypothesis that cannabimimetic 3-(1-naphthoyl)indoles interact with
the CB 1 receptor by hydrogen bonding to the carbonyl group. Indoles ( 27 )for
which R 1 =H and R 2 =H, CH 3 ,orOCH 3 were to found have significant (CB 1
Ki= 17–23 nM) receptor affinity, somewhat less than that of the corresponding
naphthoylindoles ( 28 ;R 1 =H,R 2 =H,CH 3 ,orOCH 3 ). A cannabimimetic E-indene
hydrocarbon ( 29 ), which lacks any hydrogen bonding capability, was synthesized
andfoundtohaveaCB 1 Ki= 26 ± 4 nM. These results suggest that hydrogen
bonding of the AAI carbonyl to CB 1 is not crucial for binding.
5.2
Ligand–Receptor Studies of Aminoalkylindole Binding
Reggio and co-workers constructed a pharmacophore for AAI binding at CB 1 /CB 2
based on their earlier experimental work that suggested that the s-transconfor-
mation of WIN55,212-2 was its bioactive conformation (Reggio et al. 1998). Their
work suggested that aromatic stacking was the primary interaction for AAIs at CB 1
and that the aromatic residue-rich TMH 3–4–5–6 region in CB 1 and CB 2 constitutes
the binding pocket for AAIs at the CB receptors (Song et al. 1999). These investi-
gators used their CB 1 and CB 2 receptor models to identify F5.46 in CB 2 (a residue
that is aromatic only in CB 2 ) as the residue responsible for the higher affinity of
WIN55,212-2 for CB 2. This prediction was confirmed by mutation studies (Song et
al. 1999).
Support for the TMH 3–4–5–6 region as the AAI binding region has come from
Shire and colleagues’ mutation/chimera studies of the CB receptors that suggest
that the TMH 4-E-2 loop–TMH 5 region of the CB receptors contains residues
important to the binding of the AAI, WIN55,212-2 (Shire et al. 1999). Subsequent
modeling studies in the Reggio lab of the CB 1 R* (active state) identified direct
stacking interactions between WIN55,212-2 and F3.36, W5.43 and W6.48, with
W4.64 and Y5.39 forming part of the extended ligand–CB 1 aromatic cluster. Results
of CB 1 F3.36A, W5.43A, and W6.48A mutation studies were consistent with this
binding site model (McAllister et al. 2003). A recent modeling study reported by
Salo and co-workers identified aromatic stacking interactions of WIN55,212-2 with
F3.36, Y5.39, and W5.43 (Salo et al. 2004).
Molecular modeling and receptor docking studies of naphthoylindole ( 28 ;
R 1 =H, R 2 =OCH 3 ) and its 2-methyl congener ( 28 ;R 1 =CH 3 ,R 2 =
OCH 3 ) vs indolyl-1-naphthylmethanes ( 27 ;R 1 =H,R 2 =OCH 3 )and( 27 ;R 1 =CH 3 ,
R 2 =OCH 3 ), combined with the receptor affinities of these cannabimimetic in-
doles, strongly suggested that these CB receptor ligands bind primarily by aromatic
stacking interactions in the TMH 3–4–5–6 region of the CB 1 receptor (Huffman et
al. 2003).
In summary, there is a great divergence between pharmacophores established
for AAI binding at CB 1. This divergence can be attributed to the use of different
conformations of WIN55,212-2 in superpositions with classical or non-classical