The five subtypes of dopamine receptors (D1-D5) are members of the superfamily of G protein-coupled receptors. The D1 and D2 subtypes are the most abundant and mediate most classic dopamine-dependent behaviors. This application is focused on the D2 receptor and, in particular, on the functional significance of a 29-residue fragment in the third cytoplasmic loop that is alternatively spliced to produce the short and long forms of the D2 receptor, D2S and D2L. Considerable evidence indicates that the D2 receptor is the primary autoreceptor that regulates DA neuron activity and DA release, and it is widely accepted that D2S is that autoreceptor, whereas D2L is the postsynaptic receptor expressed in striatal medium spiny neurons. In contrast, our work supports a model in which rodent dopamine neurons express both splice variants, and both function as autoreceptors with some subtle differences. We have determined that one difference between the splice variants is their rate of calcium-dependent desensitization and we now propose to evaluate the significance of this difference for cocaine-induced sensitization, reward, and aversion, while also comparing other functional properties of the splice variants. In the first specific aim we will compare additional aspects of the function of D2L and D2S as the presynaptic autoreceptor that inhibits dopamine release, dopamine neuron firing, and locomotor activity. We will use virus-mediated expression of D2L or D2S in the midbrain of D2 receptor null-mutant (D2-KO) mice and of mice with regional D2 receptor knock- out, to restore the expression of either D2L or D2S in dopamine neurons prior to assessing measures of D2 autoreceptor activity. We will also determine if both variants regulate G protein-mediated signaling in striatal neurons. This aim will focus on receptor localization and regulation of dopamine release and reuptake, and will explore the hypotheses that cocaine alters the distribution or expression of the variants. The second specific aim is based on alternative hypotheses that a cocaine-induced somatodendritic autoreceptor switch from D2S to D2L triggers or reflects other changes that produce behavioral sensitization, conditioned place preference to cocaine, and decreased aversion, or that the phenotype switch is a compensatory mechanism that mediates decreased reward and increased aversion. In this aim we will test these hypotheses by measuring cocaine-induced behavioral sensitization, conditioned place preference, and conditioned taste aversion in mice expressing only D2S or D2L autoreceptors. Experimental support for either hypothesis will provide valuable insight into the relevance of D2 receptor alternative splicing for DA neuron function.