Opioid receptors are coupled tightly to G proteins but loosely to adenylate cyclase in NG108-15 cell membranes.

Opioid receptors in intact NG 108-15 cells were irreversibly inactivated with increasing concentrations of the alkylating antagonist beta-chlornaltrexamine (CNA). The consequence of the reduction in density of opioid binding sites (quantified by saturation analysis of opioid binding in membranes) was studied at two steps of opioid receptor-mediated responses, (a) stimulation of high affinity GTPase and (b) inhibition of basal adenylate cyclase. Both agonist-mediated stimulation of GTPase and inhibition of adenylate cyclase activities were progressively reduced as the concentration of CNA in the pretreatment was increased. However, the loss of responsiveness for the two enzymes differed in two aspects. First, the diminution of GTPase responsiveness was in agreement with the loss of binding sites and took place at concentrations of CNA that were lower than those necessary to reduce responsiveness of adenylate cyclase. Second, the loss of responsiveness of GTPase occurred simply as reduction of maximal stimulation, whereas that of adenylate cyclase involved an initial reduction of apparent agonist affinity (10-fold) that was followed by a decrease in maximal effect. We next examined the loss of responsiveness of both GTPase and adenylate cyclase in membranes prepared from cells that had been exposed to increasing concentrations of pertussis toxin (PTX) to inactivate PTX-sensitive G proteins in vivo. Also in this case, the extent of reduction in responsiveness was more pronounced for GTPase than for adenylate cyclase, especially in membranes treated with high concentrations of PTX. However, the pattern of loss was identical for the two enzymes and involved a main reduction in maximal effect of the agonist that was followed only after a large degree of inactivation (greater than 60%) by a diminished apparent affinity for the agonist. Opioid receptor-mediated inhibition of cAMP accumulation in intact cells exhibits an IC50 for the agonist that is 30-10 times lower than that measured in membranes for stimulation of GTPase or inhibition of cyclase, respectively. Treatment of cells with either CNA (1 microM) or various concentrations of PTX altered the concentration-response curves for agonist-mediated inhibition of cAMP accumulation in a manner similar to that observed for adenylate cyclase in membranes, inasmuch as both maximal inhibition and apparent affinities for the agonist were decreased. However, this decrease in affinity (5-fold) was not sufficient to eliminate the discrepancy in agonist potency between membranes and intact cells.(ABSTRACT TRUNCATED AT 400 WORDS)