The recent cloning of three distinct melatonin receptor subtypes (Mel(
1a), Mel(1b) and Mel(1c)) which are part of a new family of G-protein
coupled receptors, and probably mediate the physiological actions of t
he hormone, has spurred interest in the design of analogues with subty
pe selectivity. The 5-methoxyl and N-acetyl groups of melatonin are im
portant for binding to and activation of the receptor. The indole nucl
eus serves to hold these two groups at the correct distance from one a
nother and allows them to adopt the required orientation for interacti
on with the receptor binding pocket. We have investigated the subtype
selectivity of a number of analogues of melatonin in which the structu
re has systematically been modified in order to probe the similarities
and differences in the interaction of ligand and receptor subtype. At
all three subtypes 5-methoxyl and N-acetyl groups of melatonin are im
portant for high affinity binding. However, replacing the 5-methoxyl g
roup (eg with 5-H, 5-OH, 5-Me or 5-BzO) reduces affinity much less at
the Mel(1b) receptor subtype than at either Mel(1a) or Mel(1c) cloned
subtypes.This suggests differences between the Mel(1b) and Mel(1a/1c)
subtypes in the size and shape of the binding pocket or in the manner
in which melatonin interacts with the receptor at this position. Furth
er studies have revealed that analogues with longer N-acyl carbon chai
ns behave similarly at each subtype. These observations suggest that t
he 'pocket' into which the N-acetyl group fits is very similar for eac
h subtype. Substitutions at the 2-position on the indole ring improved
affinity at each receptor subtype but did not give selective analogue
s. The systematic 'mapping' of the requirements for binding at each re
ceptor subtype should allow the design of more selective agonists and
antagonists, which will be valuable tools for the characterization and
classification of functional melatonin receptors. ((C) Elsevier, Pari
s).