Selective modulation of cell function by G protein-coupled receptor (GPCR) activation is highly desirable for basic research and therapy but difficult to achieve. We present a novel strategy toward this goal using muscarinic acetylcholine receptors as a model. The five subtypes bind their physiological transmitter in the highly conserved orthosteric site within the transmembrane domains of the receptors. Orthosteric muscarinic activators have no binding selectivity and poor signaling specificity. There is a less well conserved allosteric site at the extracellular entrance of the binding pocket. To gain subtype-selective receptor activation, we synthesized two hybrids fusing a highly potent oxotremorine-like orthosteric activator with M(2)-selective bis(ammonio)alkane-type allosteric fragments. Radioligand binding in wild-type and mutant receptors supplemented by receptor docking simulations proved M(2) selective and true allosteric/orthosteric binding. G protein activation measurements using orthosteric and allosteric blockers identified the orthosteric part of the hybrid to engender receptor activation. Hybrid-induced dynamic mass redistribution in CHO-hM(2) cells disclosed pathway-specific signaling. Selective receptor activation (M(2)>M(1)>M(3)) was verified in living tissue preparations. As allosteric sites are increasingly recognized on GPCRs, the dualsteric concept of GPCR targeting represents a new avenue toward potent agonists for selective receptor and signaling pathway activation.

A novel series of muscarinic receptor ligands of the hexamethonio-type was prepared which contained, on one side, the phthalimidopropane or 1,8-naphthalimido-2,2-dimethylpropane moiety typical for subtype selective allosteric antagonists and, on the other, the acetylenic fragment typical for the nonselective orthosteric muscarinic agonists oxotremorine, oxotremorine-M, and related muscarinic agonists. Binding experiments in M(2) receptors using [(3)H]N-methylscopolamine as an orthosteric probe proved an allosteric action of both groups of hybrids, 7a-10a and 8b-10b. The difference in activity between a-group and b-group hybrids corresponded with the activity difference between the allosteric parent compounds. In M(1)-M(3) muscarinic isolated organ preparations, most of the hybrids behaved as subtype selective antagonists. [(35)S]GTPgammaS binding assays using human M(2) receptors overexpressed in CHO cells revealed that a weak intrinsic efficacy was preserved in 8b-10b. Thus, attaching muscarinic allosteric antagonist moieties to orthosteric muscarinic agonists may lead to hybrid compounds in which functions of both components are mixed.