Neuropeptide Y in the cerebral cortex and the caudate-putamen nuclei: ultrastructural basis for interactions with GABAergic and non-GABAergic neurons.

In the cerebral cortex and caudate-putamen (CP) nuclei, neuropeptide Y (NPY) immunoreactivity is detectable within 1-2% of all neurons. The NPY-immunoreactive neurons are interneuronal and are believed to be mostly GABAergic in the cerebral cortex but not in the CP nuclei. Thus NPY and GABA may play different roles in the circuitry within these 2 regions. We tested this possibility by comparing the ultrastructure of NPY-containing neurons between (1) cortex (somatosensory and anterior cingulate areas) versus dorsolateral CP; and (2) GABAergic versus non-GABAergic NPY neurons within each area. Single coronal sections through the rat forebrain were dually labeled for GABA and NPY by combining immunoautoradiography with the immunoperoxidase method. NPY-containing neurons with or without GABA occurred throughout the rostrocaudal portions of CP and all laminae of somatosensory and anterior cingulate cortex. Comparisons between the areas confirmed that somata and terminals dually labeled for GABA and NPY were more prevalent in the cortex. NPY terminals lacking detectable GABA immunoreactivity also were found within the cortex, thus suggesting additional heterogeneity in cortical NPY innervation. The ultrastructural features of NPY perikarya in both regions were morphologically similar regardless of whether the cells also contained GABA. Most synaptic inputs to NPY neurons occurred at distal dendrites. In comparison to neighboring neurons, synaptic inputs to proximal dendrites and somata of NPY neurons of cortex and CP were rare, suggesting that fewer and weaker inputs may modulate the excitability of NPY-containing neurons. In both regions, nearly all NPY- and NPY-GABA-labeled terminals formed symmetric junctions suggestive of inhibitory action. The majority of these junctions were with dendrites containing neither NPY nor GABA. NPY terminals formed few contacts on proximal dendrites and somata of GABAergic neurons (8% of 179 contacts in cortex; 12% of 73 contacts in CP) which, unlike most singly-labeled GABAergic neurons, were sparsely innervated. Thus, NPY may play a more prominent role in modulation of certain GABAergic neurons than would be predicted by the observed frequency of NPY-to-GABA contacts in the two regions. One notable regional difference was the greater prevalence in cortex of axoaxonic associations between NPY-immunoreactive terminals and other terminals, some of which also contained NPY. These nonsynaptic associations may be involved in the modulation of (1) the release of NPY by another transmitter or (2) NPY's modulation of release of other transmitters in cortex.