Cell death neuroprotection and repair mechanisms in a model of rat spinal cord injury in vitro

Nowadays, new spinal cord injury (SCI) cases are frequently due to non traumatic causes, especially vascular disorders. A prerequisite to developing mechanism-based neuroprotective strategies for acute SCI is a full understanding of the early pathophysiological changes to prevent later disability and paralysis. The immediate damage spreads from the initial site through excitotoxicity and metabolic dysfunction (ischemia, free radicals and neuroinflammation) to surrounding tissue (secondary damage). Using an in vitro neonatal rat spinal cord model, an experimental protocol (pathological medium, PM) has been developed to mimic the profound metabolic perturbation (hypoxia, aglycemia, oxidative stress, acidosis, toxic free radicals) occurring in vivo after ischemic SCI, a condition surprisingly worsened by extracellular Mg2+ (1 mM). The current study sought to identify the cells affected by PM (with Mg2+), and the associated molecular death pathways in the spinal lumbar region which contains the locomotor networks. The results indicated that 1 h PM+Mg2+ application induced delayed pyknosis chiefly in the spinal white matter via overactivation of poly (ADP-ribose) polymerase 1 (PARP1), suggesting cell death mediated by the process of parthanatos and also via caspase 3-dependent apoptosis. Grey matter damage was less intense and concentrated in dorsal horn neurons and motoneurons which became nuclearimmunoreactive for the mitochondrial apoptosis-inducing factor. Moreover, TRPM2 channel expression was enhanced 24 h later in dorsal horn and motoneurons, while TRPM7 channel expression concomitantly decreased. Conversely, TRPM7 expression grew earlier (3 h) in white matter cells, while TRPM2 remained undetectable. Our results show that extracellular Mg2+ amplified the white matter cell death via parthanatos and apoptosis, and motoneuronal degeneration via PARP1-dependent pathways with distinct changes in their TRPM expression. In fact, the PARP-1 inhibitor PJ34, when applied 30 min after the moderate excitotoxic insult, could protect spinal networks controlling locomotion in more than 50 % of preparations. Interestingly, the drug per se strongly increased spontaneous network discharges without cell damage. Glutamate ionotropic receptor blockers suppressed this phenomenon reversibly. Our results suggest that pharmacological inhibition of PARP-1 could prevent damage to the locomotor networks if this procedure had been implemented early after the initial lesion and when the lesion was limited. PJ34 had also a positive effect on PM+Mg2+ treated spinal cords, especially in the white matter after 24 h, both alone or administered together with caspase-3 inhibitor. The neonatal rat in vitro SCI model was also useful to study the activation of endogenous spinal stem cells. We identified the ATF3 transcription factor as a novel dynamic marker for ependymal stem/progenitor cells (nestin, vimentin and SOX2 positive) located around the central canal of the neonatal or adult rat spinal cord. While quiescent ependymal cells showed cytoplasmic ATF3 expression, over 6-24 h in vitro these cells mobilized and acquired intense nuclear ATF3 staining. The migration of ATF3-nuclear positive cells preceded the strong proliferation of ependymal cells occurring after 24 h in vitro. Pharmacological inhibition of MAPK-p38 and JNK/c-Jun, upstream effectors of ATF3 activation, prevented the mobilization of ATF3 nuclear-positive cells. Excitotoxicity or ischemia-like conditions did not enhance migration of ependymal cells at 24 h. ATF3 is, therefore, suggested as a new biomarker of activated migrating stem cells in the rat spinal cord in vitro that represents an advantageous tool to study basic properties of endogenous stem cells.