- The classification of muscarinic receptors into M1 and M2 subtypes and the involvement of guanine nucleotide binding proteins (G-proteins) as major mediators of receptor information transduction in the cholinergic and other neurotransmitter systems have prompted us to undertake studies both at receptor and postreceptor levels that may shed light on the importance of these new findings to the pharmacotherapy of manic-depressive illness and of extrapyramidal syndromes. We searched for patterns of muscarinic selectivity among the commonly used anticholinergics (bipenden, procyclidine, trihexyphenidyl, benztropine, and methixen) through radioligand receptor studies in various rat tissues. The drugs showed a range of selectivity, from the totally nonselective methixen to the highly M1-selective biperiden. Sinus arrhythmia measurements were undertaken in psychiatric patients treated with different antiparkinsonian anticholinergics. The extent of sinus arrhythmia suppression was inversely correlated with the degree of M1 selectivity of the drugs used, advocating the use of M1-selective antiparkinsonian anticholinergics like biperiden in the treatment of extrapyramidal side effects. The implications of muscarinic receptor subclassification were further extended to include postreceptor phenomena. We have directly studied G-protein function by measuring cholinergic agonistinduced increases in guanosine triphosphate (GTP) binding to these proteins. This cholinergic agonistic effect was shown to be exerted by G-proteins other than Gs (the adenylate cyclase stimulatory G-protein), i.e., Gi (the adenylate cyclase inhibitory G-protein) or Gp [the G-protein activating phosphatidylinositol (PI) turnover], as ribosylation by pertussis toxin abolished this cholinergic effect, whereas it was unaffected by cholera toxin. Pertussis toxin-blockable, carbamylcholine-induced increases in GTP binding capacity were found to be mediated through M1 muscarinic receptors, as M1-selective antagonists were 100-fold more effective than M2 selective antagonists in blocking carbamylcholine effects. Moreover, carbamylcholine effect was exclusively detected in tissues predominantly populated by M1 receptors. Our results thus suggest that carbamylcholineinduced increases in GTP binding are exerted through M1 receptors interacting with Gp. At therapeutically efficacious concentrations, lithium completely blocked carbamylcholine-induced increases in GTP binding capacity in both in vitro and in vivo experiments. In experiments analogous to those conducted on lithium effects on the interaction of the cholinergic system with G-proteins, lithium, both in vitro and in vivo, was found to also abolish isoprenaline-induced increases in GTP binding. Isoprenaline effects in this system can be assigned to Gs, as they were cholera toxin-blockable but pertussis toxin-unblockable. These findings suggest G-proteins (Gp and Gs) as the molecular sites of action for both the antimanic and antidepressant effects of lithium.