Human nicotinic acetylcholine (ACh) receptor subtypes expressed in Xenopus oocytes were characterized in terms of their activation by the experimental agonist RJR-2403. Responses to RJR-2403 were compared with those evoked by ACh and nicotine. These agonists were also characterized in terms of whether application of the drugs had the effect of producing a residual inhibition that was manifest as a decrease in subsequent control responses to ACh measured 5 min after the washout of the drug. For the activation of alpha4beta2 receptors, RJR-2403 had an efficacy equivalent to that of ACh and was more potent than ACh. RJR-2403 was less efficacious than ACh for other human receptor subtypes, suggesting that it is a partial agonist for all these receptors. Nicotine activated peak currents in human alpha4beta2 and alpha3beta2 receptors that were 85 and 50% of the respective ACh maximum responses. Nicotine was an efficacious activator of human alpha7 receptors, with a potency similar to ACh, whereas RJR-2403 had very low potency and efficacy for these receptors. At concentrations of <1 mM, RJR-2403 did not produce any residual inhibition of subsequent ACh responses for any receptor subtype. In contrast, nicotine produced profound residual inhibition of human alpha4beta2, alpha3beta2, and alpha7 receptors with IC(50) values of 150, 200, and 150 microM, respectively. Co-expression of the human alpha5 subunit with alpha3 and beta2 subunits had the effect of producing protracted responses to ACh and increasing residual inhibition by ACh and nicotine but not RJR-2403. In conclusion, our results, presented in the context of the complex pharmacology of nicotine for both activating and inhibiting neuronal nicotinic receptor subtypes, suggest that RJR-2403 will be a potent and relatively selective activator of human alpha4beta2 receptors.