Evidence for a G protein-coupled gamma-hydroxybutyric acid receptor

gamma-Hydroxybutyric acid (GHB) is a naturally occurring metabolite of GABA that has been postulated to exert ubiquitous neuropharmacological effects through GABA(B) receptor (GABA(B)R)-mediated mechanisms. The alternative hy pothesis that GHB acts via a GHB-specific, G protein-coupled presynaptic re ceptor that is different from the GABA(B)R was tested. The effect of GHB on regional and subcellular brain adenylyl cyclase in adult and developing ra ts was determined and compared with that of the GABA(B)R agonist (-)-baclof en. Also, using guanosine 5'-O-(3-[S-35]thiotriphosphate) ([S-35]GTP gamma S) binding and low-K-m GTPase activity as markers the effects of GHB and (- )-baclofen on G protein activity in the brain were determined. Neither GHB nor baclofen had an effect on basal cyclic AMP (cAMP) levels. GHB significa ntly decreased forskolin-stimulated cAMP levels by 40-50% in cortex and hip pocampus but not thalamus or cerebellum, whereas (-)-baclofen had an effect throughout the brain. The effect of GHB on adenylyl cyclase was observed i n presynaptic and not postsynaptic subcellular tissue preparations, but the effect of baclofen was observed in both subcellular preparations. The GHB- induced alteration in forskolin-induced cAMP formation was blocked by a spe cific GHB antagonist but not a specific GABA(B)R antagonist. The (-)-baclof en-induuce alteration in forskolin-induced cAMP formation was blocked by a specific GABA(B)R antagonist but not a specific GHB antagonist. The negativ e coupling of GHB to adenylyl cyclase appeared at postnatal day 21, a devel opmental time point that is concordant with the developmental appearance of [H-3]GHB binding in cerebral cortex, but the effects of (-)-baclofen were present by postnatal day 14. GHB and baclofen both stimulated [S-35]GTP gam ma S binding and low-K-m GTPase activity by 40-50%. The GHB-induced effect was blocked by GHB antagonists but not by GABA(B)R antagonists and was seen only in cortex and hippocampus. The (-)-baclofen-induced effect was blocke d by GABA(B)R antagonists but not by GHB antagonists and was observed throu ghout the brain. These data support the hypothesis that GHB induces a G pro tein-mediated decrease in adenylyl cyclase via a GHB-specific G protein-cou pled presynaptic receptor that is different from the GABA(B)R.