Entopallium lost GFAP immunoreactivity during avian evolution -- is GFAP a 'condition sine qua non'?

TThe present study demonstrates that in the same brain area the astroglia can express GFAP (the main cytoskeletal protein of astroglia) in some species but not in the others of the same vertebrate class. It contrasts the former opinions that the distribution of GFAP found in a species is characteristic of the entire class. The present study investigated birds in different phylogenetic positions: duck (Cairina moschata domestica), chicken (Gallus gallus domesticus), and quails (Coturnix japonica and Excalfactoria chinensis) of Galloanserae; pigeon (Columba livia domestica) of a group of Neoaves, in comparison to representatives of other Neoaves lineages, which emerged more recently in evolution: finches (Taeniopygia guttata and Erythrura gouldiae), magpie (Pica pica), and parrots (Melopsittacus undulatus and Nymphicus hollandicus).. Following a perfusion with 4% buffered paraformaldehyde, immunoperoxidase reactions were performed with two types of anti-GFAP: monoclonal and polyclonal, on floating sections. The entopallium (formerly 'ectostriatum', a telencephalic area in birds) was GFAP-immunopositive in pigeon and in the representatives of Galloanserae but not in songbirds and parrots, which emerged more recently in evolution. The lack of GFAP expression of a brain area, however, does not mean the lack of astroglia. Lesions induced GFAP expression in the territory of GFAP-immunonegative entopallia. It proved that the GFAP-immunonegativity is not due to the lack of capability, but rather the suppression of GFAP production of the astrocytes in this territory. In the other areas investigated besides the entopallium (optic tectum and cerebellum), no considerable interspecific differences of GFAP immunopositivity were found. It proved that the immunonegativity of entopallium is due to neither the general lack of GFAP-expression nor the incapability of our reagents to detect GFAP in these species. The data are congruent with our proposal [Kálmán, 2002, Brain Res. Bull. 57:509] that a lack of GFAP expression has evolved in different brain areas in vertebrate evolution, typically in lineages that emerged more recently. Comparative studies on GFAP-immunopositive and -negative entopallia may promote understanding the role of GFAP in neural networks.