DEVELOPMENT OF AN ASTROCYTIC RESPONSE TO LESIONS OF THE SPINAL-CORD IN THE NORTH-AMERICAN OPOSSUM - AN IMMUNOHISTOCHEMICAL STUDY USING ANTI-GLIAL FIBRILLARY ACIDIC PROTEIN

We have shown previously that rubral axons grow around a lesion of the ir pathway in developing opossums and that a critical period exists fo r that plasticity. The critical period begins when rubral axons first reach the level of the lesion and ends sometime between postnatal days (PD) 26 and 30. The aim of the present study was to examine the devel opment of an astrocytic response to lesioning the spinal cord to deter mine if there is a temporal correlation between the development of suc h a response and the end of the critical period. The astrocytic respon se was examined immunohistochemically, 2 and 4 weeks after hemisecting the thoracic spinal cord, using an antibody to glial fibrillary acidi c protein (GFAP). A response was first seen at PD21 in the 2-week seri es. The response was relatively mild, however, and limited to the whit e matter. When the lesion was made at PD26, the response was still res tricted to the white matter, but hypertrophied astrocytes were found a t the gray/white matter junction and cystic cavities were present. Whe n the lesion was made at PD41, the response had spread to the gray mat ter and it occupied a larger area rostral and caudal to the lesion tha n at earlier ages. The animals allowed to survive 4 weeks after lesion ing were subjected to a second operation 4-5 days before sacrifice so that Fast Blue could be injected bilaterally two to three segments cau dal to the lesion. When the hemisection was made at PD15, a response w as present in the ventral and ventrolateral funiculi, but not in that part of the lateral funiculus that contains rubrospinal axons. As expe cted from previous studies, rubral neurons were labeled contralateral to the lesion, suggesting that their axons had grown around it. When t he lesion was made at PD21, the glial response was still limited to th e white matter, but it extended into that part of the lateral funiculu s that contains rubral axons. In spite of the presence of a glial resp onse, rubral neurons were still labeled contralateral to the lesion. W hen the lesion was made at PD26, hypertrophied astrocytes were present at the gray/white matter junction and small cavities were noted at th e lesion site. In such cases, there was no evidence for rubrospinal pl asticity. An astrocytic response was not observed in the gray matter o f the dorsal horn, an area used by rubral axons to grow around a lesio n of their pathway, until well after the end of the critical period. W e conclude that the initial development of a glial scar in the white m atter after lesioning does not determine the end of the critical perio d for rubrospinal plasticity. Loss of rubrospinal plasticity correlate s most closely with the appearance of hypertrophied astrocytes at the gray/white matter junction and the formation of cystic cavities. (C) 1 993 Wiley-Liss, Inc.