Distribution of TAG-1 and synaptophysin in the developing cerebellar cortex: Relationship to Purkinje cell dendritic development

Abstract

During postnatal cerebellar development, differentiating Purkinje cell (PC) dendrites extend towards the pial surface and progressively contact immature granule cell parallel fiber (PF) axons in the deep external granule layer (EGL), thus forming a zone of synaptic contact called the molecular layer (ML). The neuronal cell adhesion molecule, TAG-1, is transiently expressed on PF axons in the deep EGL (Yamamoto et al. [1986] J. Neurosci. 12:3576-3594). To determine the spatiotemporal relationship between Purkinje cell dendritic differentiation and the cessation of TAG-1 expression, sagittal sections from developing rat cerebellum were double-labeled for TAG-1 and the Purkinje cell-specific marker, calbindin, by using indirect immunofluorescence. At postnatal day 2 (P2) and P5, confocal microscopy revealed that TAG-1 immunoreactivity began above the furthest superficial extent of the Purkinje cell apical dendritic cap. By P10, PC dendrites penetrated partially into the TAG-1-positive deep EGL, creating a narrow region of overlap in TAG-1/calbindin staining at the deep EGL/ML border. In contrast, at P15 and P20, TAG-1 staining began directly above the furthest superficial extent of the Purkinje cell dendrites, with little or no overlap in TAG-1/calbindin staining. Staining for the synaptic vesicle glycoprotein, synaptophysin, was dim throughout most of the TAG-1-positive deep EGL, although bright synaptophysin immunoreactivity was observed throughout the ML. Overlap in TAG-1/synaptophysin staining was observed primarily at the deep EGL/ML border, suggesting that robust expression of synaptophysin in granule cells does not begin until after contact with Purkinje cell dendrites has been initiated. Our results suggest that factors present in the developing ML may influence the cessation of TAG-1 expression and the initiation of synaptophysin expression at the border region between the ML and the deep EGL.