Recent pathophysiological models of basal ganglia function in Parkinson's disease predict that specific neurochemical changes in the indirect pathway would follow the lack of stimulation of D(2) dopamine receptors. Post mortem studies of the basal ganglia in genetically modified mice lacking functional copies of the D(2) dopamine receptor gene allowed us to test these predictions. When compared with their congenic N(5) wild-type siblings, mice lacking D(2) receptors show an increased expression of enkephalin messenger RNA in the striatum, and an increased activity and expression of cytochrome oxidase I in the subthalamic nucleus, as expected. In addition, D(2) receptor-deficient mice display a reduced expression of glutamate decarboxylase-67 messenger RNA in the globus pallidus, as the basal ganglia model predicts. This reduction contrasts with the lack of change or increase in glutamate decarboxylase-67 messenger RNA expression found in animals depleted of dopamine after lesions of the mesostriatal dopaminergic system. Furthermore, D(2) receptor-deficient mice show a significant decrease in substance P messenger RNA expression in the striatonigral neurons which form the direct pathway. Finally, glutamate decarboxylase-67 messenger RNA expression in the basal ganglia output nuclei was not affected by mutations in the D(2) receptor gene, a fact that could probably be related to the absence of a parkinsonian locomotor phenotype in D(2) receptor-deficient mice. In summary, these findings provide compelling evidence demonstrating that the lack of endogenous stimulation of D(2) receptors is sufficient to produce subthalamic nucleus hyperactivity, as assessed by cytochrome oxidase I histochemistry and messenger RNA expression, and strongly suggest the existence of interactions between the basal ganglia direct and indirect pathways.