Studies of locomotor networks of the neonatal rat spinal cord after acute spinal injury in vitro

Spinal cord injury (SCI) represents a significant health problem associated with life-long disability and a broad range of secondary complications. Acute spinal cord injury evolves rapidly within hours and days after the initial trauma, producing secondary damage even to initially spared areas. The early pathophys iological mechanisms affecting spinal networks remain largely obscure despite widespread incidence of this condition and its social consequences. Regardless of their etiology, spinal lesions are believed to include combinatorial effects of excitotoxicity and severe metabolic perturbations. The present study used an in vitro spinal cord model from the neonatal rat to investigate the relative contribution by excitotoxicity and toxic metabolites to dysfunction of locomotor networks, spinal reflexes and intrinsic network rhythmicity. Preparations were treated (I h) with either kainate or a pathological medium (containing free radicals and hypoxic/aglycemic conditions), or their combination. Damage was measured by taking as outcome locomotor network activity for up to 24 h after the primary insult. Kainate Jed to irreversible suppression of fictive locomotion, while intrinsic bursting induced by synaptic inhibition block persisted. The pathological medium slowed down fictive locomotion and intrinsic rhythmicity. Combination of kainate with pathological medium evoked extensive, irreversible damage to the spinal cord. This phenomenon was associated with loss of fictive locomotion and intrinsic bursting as well as polysynaptic reflex depression. Thus, while suggesting distinct roles of excitotoxicity and metabolic dysfunction in the acute damage of locomotor networks, our model indicates that different strategies might be necessary to treat the various early components of acute spinal cord les ion. Next, we investigated the role of extracellular Mg2+ in the lesion evoked by pathological medium, as the recent clinical trials to treat this condition with i.v. Mg2+ to stabi lize its extracellular concentration provided disappointing results. Pathological medium in L mM Mg2 + solution (1 h) largely depressed spinal reflexes and suppressed fictive locomotion on the same and the day after. Conversely, pathological medium in either Mg2+-free or 5 mM Mg2 + solution evoked temporary network depression and enabled fictive locomotion the day after. Although the excitotoxic damage elicited by kainate was insensitive to extracellular Mg2+, 1 mM Mg2+ potentiated the effect of combining pathological medium with kainate at low concentrations. These results indicate that preserving Mg2 + homeostasis rendered experimental spinal injury more severe. Treatment to block the pathophysiological processes triggered by acute spinal injury remains unsati sfactory as the underlying mechanisms are incompletely understood. We further investigated the feasibility of neuroprotection of lumbar locomotor networks by the glutamate antagonists CNQX and APV against acute lesions induced by either pathological medium or excitotoxicity. Inhibition of fictive locomotion by pathological medium was contrasted by simultaneous and even delayed ( I h later) co-application of CNQX and APV. Delayed neuroprotection was accompanied by increased survival of ventral horn premotoneurons and lateral column white matter. Neither CNQX nor APV alone provided neuroprotection. Kainate-mediated excitotoxicity always led to loss of fictive locomotion. CNQX and APV co-applied with kainate functionally protected l/3rd of preparations, although they failed when their application was delayed. Our data suggest that locomotor network neuroprotection was possible when introduced very early during the pathological process of spinal injury, but also showed how the borderline between presence or loss of locomotor activity was a very narrow one that depended on the survival of a certain number of neurons or white matter elements. The present report provides a model not only for preclinical testing of novel neuroprotective agents, but also for estimating the minimal network membership compatible with functional locomotor output.