Background context - Transplantation of bone marrow cells into spinal cord lesions promotes functional recovery in animal models, and recent clinical trials suggest possible recovery also in humans. The mechanisms responsible for these improvements are still unclear.
Purpose - To characterize spinal cord motor neurite interactions with human bone marrow stromal cells (MSCs) in an in vitro model of spinal cord injury (SCI).
Study design/setting - Previously, we have reported that human MSCs promote the growth of extending sensory neurites from dorsal root ganglia (DRG), in the presence of some of the molecules present in the glial scar, which are attributed with inhibiting axonal regeneration after SCI. We have adapted and optimized this system replacing the DRG with a spinal cord culture to produce a central nervous system (CNS) model, which is more relevant to the SCI situation.
Methods - We have developed and characterized a novel spinal cord culture system. Human MSCs were cocultured with spinal motor neurites in substrate choice assays containing glial scar–associated inhibitors of nerve growth. In separate experiments, MSC-conditioned media were analyzed and added to spinal motor neurites in substrate choice assays.
Results - As has been reported previously with DRG, substrate-bound neurocan and Nogo-A repelled spinal neuronal adhesion and neurite outgrowth, but these inhibitory effects were abrogated in MSC/spinal cord cocultures. However, unlike DRG, spinal neuronal bodies and neurites showed no inhibition to substrates of myelin-associated glycoprotein. In addition, the MSC secretome contained numerous neurotrophic factors that stimulated spinal neurite outgrowth, but these were not sufficient stimuli to promote spinal neurite extension over inhibitory concentrations of neurocan or Nogo-A.
Conclusions - These findings provide novel insight into how MSC transplantation may promote regeneration and functional recovery in animal models of SCI and in the clinic, especially in the chronic situation in which glial scars (and associated neural inhibitors) are well established. In addition, we have confirmed that this CNS model predominantly comprises motor neurons via immunocytochemical characterization. We hope that this model may be used in future research to test various other potential interventions for spinal injury or disease states.