Distribution of parvalbumin-, calretinin-, and calbindin-D28k-immunoreactive neurons and fibers in the human entorhinal cortex

Abstract

Parvalbumin, calretinin, and calbindin-D28k are calcium-binding proteins that are located in largely nonoverlapping neuronal populations in the brain. The authors studied the distribution of parvalbumin-, calretinin-, and calbindin-D28k-immunoreactive (ir) cells, fibers, terminals, and neuropil in the eight subfields of the human entorhinal cortex. The distribution of each of the three calcium-binding proteins largely followed the cytoarchitectonic borders of the eight entorhinal subfields, although the regional and laminar distributions of the three proteins were segregated rather than overlapping. The highest density of parvalbumin-ir neurons and terminals was found in the caudal and lateral subfields of the entorhinal cortex. Calretinin and calbindin-D28k immunoreactivities were high rostromedially, although a large number of calretinin and calbindin-D28k neurons were also found in the caudal subfields. All parvalbumin-ir cells had a morphological appearance of nonpyramidal neurons. Parvalbumin-ir terminals formed basket-like formations around unstained somata and cartridges, suggesting that parvalbumin neurons compose a subpopulation of gamma-aminobutyric acid (GABA)ergic basket cells and chandelier cells, respectively. Although calretinin and calbindin-D28k were also found in numerous nonpyramidal neurons, both were also located in pyramidal-shaped neurons in layers V and VI (calretinin) and in layers II and III (calbindin) of the entorhinal cortex, suggesting that they play roles in projection neurons as well. Moreover, the high density of nonpyramidal neurons containing calcium-binding proteins in layers II and III of the entorhinal cortex suggests that they form an integral component of a network that controls the entorhinal outputs to the hippocampus. Furthermore, the largely nonoverlapping distributions of the parvalbumin-, calretinin-, and calbindin-ir neuronal populations in the entorhinal cortex indicate that each of them may modulate a different subset of topographically organized entorhinal outputs.