Using specific antibodies and cDNA probes, we have investigated, in rat basal ganglia, the distribution and the regulation of the expression of the alpha subunits of Gs and G(olf), two GTP-binding proteins (G-proteins) that stimulate adenylyl cyclase. We confirmed that G(olf) alpha is highly expressed in caudate-putamen, nucleus accumbens, and olfactory tubercle, whereas Gs alpha is less abundant in these areas than in the other brain regions. Intrastriatal injections of quinolinic acid decreased dramatically the levels of G(olf) alpha protein in the striatum and the substantia nigra, and those of G(olf) alpha mRNA in the striatum. Retrograde lesions of striatonigral neurons with volkensin reduced markedly the levels of D1 dopamine (DA) binding sites, as well as those of G(olf) alpha protein and mRNA in the striatum, without altering D2 binding sites. In contrast, both types of lesions increased the levels of Gs alpha protein in the striatum and substantia nigra. Immunocytochemistry showed the presence of G(olf) alpha protein in striatal medium-sized neurons and in several other neuronal populations. These results demonstrate that striatonigral neurons contain high levels of G(olf) alpha and little, if any, Gs alpha, suggesting that the coupling of D1 receptor to adenylyl cyclase is provided by G(olf) alpha. The levels of G(olf) alpha were five- to sixfold higher in the striatum than in the substantia nigra, indicating a preferential localization of G(olf) alpha in the somatodendritic region of striatonigral neurons and providing a basis for the low efficiency of D1 receptor coupling in the substantia nigra. Six weeks after 6-hydroxydopamine lesions of DA neurons, an increase in G(olf) alpha (+53%) and Gs alpha (+64%) proteins was observed in the striatum. This increase in G(olf) alpha levels may account for the DA-activated adenylyl cyclase supersensitivity, without change in D1 receptors density, that follows destruction of DA neurons. Fine regulation of the levels of G(olf) alpha in physiological or pathological situations may be a critical parameter for the efficiency of DA neurotransmission.

Numerous observations strongly support the hypothesis that dopaminergic neurons could be particularly vulnerable to an impairment of their energetic metabolism. In order to demonstrate the existence of such a selective vulnerability, the toxic effects of rotenone, an inhibitor of complex I of the respiratory chain, and of glutamate, which is very likely involved in the neurotoxicity induced by an energetic stress, were analyzed on cultured mouse mesencephalic neurons. Toxicity toward dopaminergic and GABAergic neurons was compared by measuring the residual uptakes of dopamine and GABA. Exposure to 5 nM rotenone for 6 hr or to a low concentration of glutamate (100 microM) for 1 hr did not lead to a high selective toxic effect on dopaminergic neurons. In contrast, dopaminergic neurons were three times less resistant to the sequential exposure to rotenone and glutamate than GABAergic neurons. A particular resistance of mesencephalic GABAergic neurons to the synergistic toxic effects of rotenone and glutamate was ruled out since two other neuronal types, the striatal cholinergic and GABAergic neurons, displayed the same weak vulnerability as the mesencephalic GABAergic neurons. This selective toxic effect of glutamate on rotenone- pretreated dopaminergic neurons was blocked by either AMPA or NMDA receptor antagonists and mimicked by combined treatment with AMPA and NMDA, or by NMDA alone when the medium was deprived of Mg2+ ions. Moreover, this NMDA-selective neurotoxicity was critically dependent on the presence of a physiological extracellular sodium concentration, since the use of choline chloride instead of sodium chloride had a protective effect on dopaminergic neurons.

The mouse 8.5 mRNA encodes a 171-residue novel protein which displays a highly significant similarity with the product of the previously characterized neuronal p1A75 cDNA (Sutcliffe, J.G., Milner, R.J., Shinnick, T.M., and Bloom, F.E. (1983) Cell 33, 671-682). Northern blot and in situ hybridization experiments indicated that the 8.5 mRNA is specifically expressed in neural and neuroendocrine tissues. An affinity-purified antibody directed against the recombinant 8.5 protein demonstrated the existence of the 19-kDa natural protein in brain and evidenced its prominent juxtanuclear Golgi-like localization in cultured neurons. Ultrastructural analysis of the same preparation revealed a specific labeling of all the Golgi saccules and of some vesicles in the Golgi zone. In transfected COS cells, the exogenous protein was also detected in the Golgi area, indicating, therefore, the presence of a Golgi targeting signal in its primary sequence.