Migration of Mesenchymal Stem Cells Through Cerebrospinal Fluid Into Injured Spinal Cord Tissue
Introduction: Previous work has shown a neuro protective affect from adenovirus-delivered factors such as BCL-2 inoculated directly into the zone of spinal cord injury (SCI). However, direct inoculation does create potential adverse local affects in the already injured tissue.
Purpose: To investigate the potential of survival and migration of transplanted bone marrow derived mesenchymal stem cells (MSCs) through the subarachnoid space (SAS) into the injured thoracic spinal cord tissue following injection at the more caudal lumbar spine.
Materials and Methods: Thoracic SCI was made for 60 adult Lewis rats, via a T9-10 laminectomy and weight drop using the NYU Impactor. 6 rats received only laminectomy for sham-control. MSCs were harvested from the femur of adolescent Lewis rats and expanded in MEM containing 10% FBS and 0.1% penicillin/streptomycin. At P2 or P3, adenovirus containing green fluorescent protein (GFP) gene (Adv-GFP) was transduced as a gene marker. At 3, 5, and 7 days after thoracic SCI (n=20 each), rats had an injection into the SAS at L4/5 (approximately 5cm from T9-T10). 3 groups of animal received different injections at each timing of 3, 5, and 7 days after SCI: GFPMSCs (n=12), Bgal-MSCs (n=4), PBS (n=4). Injured spinal cords were harvested at 7or 14 days (each n=10) after injection and were underwent longitudinal frozen sectioning (30mm length). Hematoxylin-Eosin (HE) staining, counterstaining with Propidium Iodide, or immunostaining with anti-GFP or anti-nestin antibody was performed. The number of GFP-positive cells in the central SCI lesion within 10mm and the whole section of 30mm were counted respectively. This cell number was calculated and compared with sham-control.
Results: MSCs transduced Adv-GFP were detected with strong green fluorescence both in thenucleus and cell body in-vitro. In GFP-MSC injected samples, huge cell aggregations were observed on the surface of the parenchyma with HE staining. These green fluorescent cells were proved to be genuine GFP-positive cells by immunoperoxidase staining with anti-GFP antibody. In the injection timing of 14 days after SCI, the ratio of GFP-positive cells gathered into the central lesion within 10mm (53%, 61%, and 63%) was significantly higher than sham-control (p<0 .02). At 14 days after injection, some GFP-positive cells were seen in the deeper area of SCI tissue perivascular space. Immunostaining against nestin demonstrated that overlapped with nestin-positive (neural stem cells, immature neurons or glial cells), and actually differentiated into them.
Conclusions: Transplanted MSCs injected into the lumbar spine CSF can migrate to injured thoracic spinal cord tissue. Transplanted MSCs can infiltrate and integrate into the spinal cord parenchyma, and actually differentiate into nestin-positive immature neurons or glial cells. This affords tremendous potential for the delivery of neuroprotective factors into the SCI from uninjured distant sites.










