Autologous peripheral nerve grafting into murine brain as a model for studies of regeneration in the central nervous system.

Autologous sciatic nerve was grafted into rat brain by (i) passing an 8-mm segment of nerve tied to a straight surgical needle through two craniotomy holes ("through-and-through" model); (ii) inserting a small tube of polyethylene containing the 8-mm nerve piece ("nerve-within-tube" model). Longitudinally oriented neurofilament-positive fibers were consistently observed within the graft. Compared with the through-and-through model, axonal sprouting in the nerve-within-tube model followed a slow-motion pattern so that a growing front of regenerating axons could be easily identified and more easily related to the cellular events occurring in Wallerian degeneration. In the through-and-through model, regenerated axons at the brain-nerve interface followed a disorganized, tortuous course so that direct continuity between brain and graft was difficult to demonstrate. The reverse was true in the nerve-within-tube model, i.e., axons penetrated directly into the graft. The difference in orientation of axonal growth at the brain-graft interface appeared to be related to the glial reaction. In the through-and-through model, reactive astrocytes formed a mesh of randomly oriented fibers in the damaged brain tissue facing the graft (anisomorphic gliosis). Conversely, longitudinally oriented fibers extended directly from the brain to the graft in the nerve-within-tube model, where brain damage was substantially reduced (isomorphic gliosis). A different type of glial fibrillary acidic (GFA) protein-positive fibrous structures was identified in the graft. Compared with reactive astrocytes, these structures were more elongated, more uniform in diameter, and less brightly immunofluorescent. Moreover, they were present throughout the graft, whereas astrocytes were confined to the distal end of the transplant, i.e., the part of nerve close to the brain-graft interface. Based on previous reports in the literature we interpret these GFA protein-positive structures as reacting Schwann cells.