Cortical neurons immunoreactive for the potassium channel Kv3.1b subunit are predominantly surrounded by perineuronal nets presumed as a buffering system for cations.

Perineuronal nets (PNs) are known as chondroitin sulphate-rich, lattice-like coatings of the extracellular matrix. In the cortex of mammalian species investigated so far, they were mainly found around GABAergic neurons, but to a lesser degree also around pyramidal cells. Previous investigations in the rat revealed similar distribution patterns of fast-firing neurons expressing both the Kv3.1b subunit of voltage-gated potassium channels and the calcium-binding protein parvalbumin. In the present study, triple fluorescence labelling was applied for the simultaneous demonstration of PNs with the N-acetylgalactosamine-specific Wisteria floribunda agglutinin (WFA), parvalbumin-immunoreactivity (ir) with a monoclonal antibody and of Kv3.1b-ir with several rabbit antibodies. Subsets of non-pyramidal neurons - enwrapped by PNs and expressing parvalbumin and Kv3.1b - were detected in the rat and monkey neocortex and hippocampus. In the rat, faintly stained PNs were additionally found around several layer II/III and V pyramidal cells immunonegative for Kv3.1b, but contacted by Kv3.1b-containing boutons. In the monkey, more intensely labelled PNs frequently occurred around pyramidal cells which themselves appeared to be Kv3. 1b-immunopositive. We also observed minor Kv3.1b-ir and parvalbumin-ir cortical cell populations which were devoid of PNs; occasionally, nets were detected around neurons lacking both immunoreactivities. By confocal laser scanning microscopy, Kv3.1b-ir and WFA-binding sites were found adjoining at the soma and proximal dendritic surface, while lectin-binding sites usually extended on more distal dendritic segments and the axon initial segments which failed to express detectable Kv3.1b-ir. This spatial relationship of both markers was also confirmed by combined WFA-gold labelling and Kv3.1b-immunoperoxidase staining at the electron microscopic level. The data are used for a critical examination of current hypotheses concerning the functional role of PNs. We conclude that PNs may serve as rapid local buffers of excess cation changes in the extracellular space. Somatic membranes of fast-spiking neurons seem to be a main, but not the only source of such changes.