The pathogenesis of parvovirus‐induced cerebellar hypoplasia in the syrian hamster, Mesocricetus auratus . Fluorescent antibody, foliation, cytoarchitectonic, golgi and electron microscopic studies

Abstract Cerebellar histogenesis was studied in hamsters infected at birth with a parvovirus, rat virus strain PRE 308. Cerebellar granule cell precursors in these animals were selectively infected and lysed in the external germinal layer before their migration to form the internal granular layer. The effects of the absence of granule cells on cerebellar development and especially on the development of the Purkinje cells and their dendrites was analyzed using fluorescent antibody. Golgi, conventional paraffin, and electron microscopic methods. This study represents the first Golgi and ultrastructural study of the pathogenesis of rat virus infections in the cerebellum. The destruction of the granule cell precursors resulted in a dysplastic cerebellar hypoplasia with total disruption of normal cerebellar stratification and cytoarchitectonics. The Purkinje cells developed misshapen, progressively disoriented dendritic stems lacking tertiary dendrites and studded with numerous spines, devoid of afferent synaptic contacts (naked spines) and encased by glial processes. These developmental studies, together with the mouse mutant studies, demonstrated that the spines of the Purkinje cells were elaborated in the absence of both tertiary dendrites and afferent parallel fiber contacts. Such data suggested that spine formation, once triggered, was intrinsically programmed rather than being dependent on the development of parallel fiber contacts. Despite the loss of a major interneuronal component and disintegration of normal cytoarchitectonic relationships, synapses in the cerebellar cortex developed normally as long as both the pre‐ and post‐synaptic elements were present. Thus synaptic specificity is maintained in the face of gross disruption of cytoarchitectonic relationships. If either the pre‐ or post‐synaptic portion of a contact was absent, then glial processes isolated the persisting element or aberrant contacts formed. In addition to glial encasement of naked spines, there were dendrodendritic articulations between Purkinje cell dendrites, some of which were joined by septate, plaque‐like junctions. Aberrant synaptic contacts between mossy and climbing fiber glomeruli and the smooth surface of the Purkinje cell somata were found rarely. In addition to these contacts which also occur in the hypoplastic cerebella produced by other methods, previously undescribed non‐synaptic spine‐spine articulations between Purkinje cell dendrites were seen. The role played by granule cells and their axons in Purkinje cell development appeared to be two‐fold. First, the development of the orderly array of parallel fibers in the normal animal played a role in orienting and flattening the dendritic trees of Purkinje cells. Second, the formation of tertiary dendritic branches appeared to depend primarily upon the presence of an external germinal layer throughout this stage of Purkinje cell development. By contrast, dendritic spines developed and persisted in the absence of granule cells.