Although it has long been widely accepted that dopamine receptor types D1 and D2 form GPCR heteromers in the striatum, the presence of D1–D2 receptor heteromers has been recently challenged. In an attempt to properly characterize D1–D2 receptor heteromers, here we have used the in situ proximity ligation assay (PLA) in striatal sections comprising the caudate nucleus, the putamen and the core and shell territories of the nucleus accumbens. Experiments were carried out in control macaques as well as in MPTP-treated animals (with and without dyskinesia). Obtained data support the presence of D1–D2 receptor heteromers within all the striatal subdivisions, with the highest abundance in the accumbens shell. Dopamine depletion by MPTP resulted in an increase of D1–D2 density in caudate and putamen which was normalized by levodopa treatment. Two different sizes of heteromers were consistently found, thus suggesting that besides individual heteromers, D1–D2 receptor heteromers are sometimes organized in macromolecular complexes made of a number of D1–D2 heteromers. Furthermore, the PLA technique was combined with different neuronal markers to properly characterize the identities of striatal neurons expressing D1–D2 heteromers. We have found that striatal projection neurons giving rise to either the direct or the indirect basal ganglia pathways expressed D1–D2 heteromers. Interestingly, macromolecular complexes of D1–D2 heteromers were only found within cholinergic interneurons. In summary, here we provide overwhelming proof that D1 and D2 receptors form heteromeric complexes in the macaque striatum, thus representing a very appealing target for a number of brain diseases involving dopamine dysfunction.

Calbindin (CB) is a calcium binding protein reported to protect dopaminergic neurons from degeneration. Although a direct link between CB content and differential vulnerability of dopaminergic neurons has long been accepted, factors other than CB have also been suggested, particularly those related to the dopamine transporter. Indeed, several studies have reported that CB levels are not causally related to the differential vulnerability of dopaminergic neurons against neurotoxins. Here we have used dual stains for tyrosine hydroxylase (TH) and CB in 3 control and 3 MPTP-treated monkeys to visualize dopaminergic neurons in the ventral tegmental area (VTA) and in the dorsal and ventral tiers of the substantia nigra pars compacta (SNcd and SNcv) co-expressing TH and CB. In control animals, the highest percentages of co-localization were found in VTA (58.2%), followed by neurons located in the SNcd (34.7%). As expected, SNcv neurons lacked CB expression. In MPTP-treated animals, the percentage of CB-ir/TH-ir neurons in the VTA was similar to control monkeys (62.1%), whereas most of the few surviving neurons in the SNcd were CB-ir/TH-ir (88.6%). Next, we have elucidated the presence of CB within identified nigrostriatal and nigroextrastriatal midbrain dopaminergic projection neurons. For this purpose, two control monkeys received one injection of Fluoro-Gold into the caudate nucleus and one injection of cholera toxin (CTB) into the postcommissural putamen, whereas two more monkeys were injected with CTB into the internal division of the globus pallidus (GPi). As expected, all the nigrocaudate- and nigroputamen-projecting neurons were TH-ir, although surprisingly, all of these nigrostriatal-projecting neurons were negative for CB. Furthermore, all the nigropallidal-projecting neurons co-expressed both TH and CB. In summary, although CB-ir dopaminergic neurons seem to be less prone to MPTP-induced degeneration, our data clearly demonstrated that these neurons are not giving rise to nigrostriatal projections and indeed CB-ir/TH-ir neurons only originate nigroextrastriatal projections.

Adeno-associated viruses (AAVs) have become highly promising tools for research and clinical applications in the central nervous system (CNS). However, specific delivery of genes to the cell type of interest is essential for the success of gene therapy and therefore a correct selection of the promoter plays a very important role. Here, AAV8 vectors carrying enhanced green fluorescent protein (eGFP) as reporter gene under the transcriptional control of different CNS-specific promoters were used and compared with a strong ubiquitous promoter. Since one of the main limitations of AAV-mediated gene delivery lies in its restricted cloning capacity, we focused our work on small-sized promoters. We tested the transduction efficacy and specificity of each vector after stereotactic injection into the mouse striatum. Three glia-specific AAV vectors were generated using two truncated forms of the human promoter for glial fibrillar acidic protein (GFAP) as well as a truncated form of the murine GFAP promoter. All three vectors resulted in predominantly glial expression; however we also observed eGFP expression in other cell-types such as oligodendrocytes, but never in neurons. In addition, robust and neuron-specific eGFP expression was observed using the minimal promoters for the neural protein BM88 and the neuronal nicotinic receptor β2 (CHRNB2). In summary, we developed a set of AAV vectors designed for specific expression in cells of the CNS using minimal promoters to drive gene expression when the size of the therapeutic gene matters.