Implantation of cultured sensory neurons and schwann cells into lesioned neonatal rat spinal cord. II. Implant characteristics and examination of corticospinal tract growth

The purpose of this study was to test the effectiveness of implants derived from peripheral neural tissue to serve as bridges following interruption of the developing corticospinal tract (CST). Implants prepared from purified populations of cultured dorsal root ganglion neurons (DRGNs) and Schwann cells (SCs) (Kuhlengel et al., J. Comp Neurol. 293:63‐73, 1990) were placed into thoracolumbar regions of neonatal rat spinal cord from which a 2‐mm length of dorsal columns had been removed by suction. These cords were examined by a number of techniques 10 days to 6 months later. The implants, recognizable by their DRGN content, filled the vacated dorsal columns and survived the longest periods examined. The most effective method to maintain implant position was dorsal placement of collagen‐coated Nitex filter. Implants were inserted either at the time of lesioning of 5 days later. The implant survival rate was better (72% vs. 50%) and meningeal scarring was less with immediate implantation, but delayed implantation resulted in better implant‐cord fusion and the implant better filled the lesion cavity. DRGN/SC implants became well vascularized without leptomeningeal cells; this may explain why implant survival was not improved with leptomeningeal cell addition. Particularly well‐differentiated implants (full extracellular matrix production and myelination) did not fuse as well with cord as did those less well differentiated. The addition of nerve growth factor to the Nitex filter collagen coating led to improved survival of DRGNs in implants. Electron microscopy showed that astrocytes populated the implant‐cord junction region and migrated into implants. Typical SCs related to nonmyelinated and myelinated axons were present in implants. Close proximity of astrocytes and central myelin to SCs and peripheral myelin demonstrated good implant integration with cord. Clusters of SCs, astrocytes, and axons, all enclosed within a common basal lamina, were observed in implants. Immunostaining for GFAP and laminin confirmed our microscopy findings that SCs did not migrate from implant into host but that astrocytes left host tissue to enter implants. Neuroanatomical tracing of CST neurons with HRP‐WGA showed that labeled fibers were not present in the implant but were fasciculated just beneath in gray matter. These fibers remained clustered in gray matter under‐neath the ventral dorsal columns caudal to the lesion. In lesioned but not implanted rats, labeled fibers were only diffusely distributed in gray matter. Delayed implantation led to more variation in fasciculation compared with immediate implantation. In conclusion. DRGN/SC implants did not serve as tissue bridges but appeared to influence adjacent cord parenchyma to permit CST fiber fasciculation that persisted beyond the implant.