Periventricular white matter injury, the principal variety of brain injury of the human premature infant, involves differentiating oligodendroglia. Nothing is known of the biochemical mechanism of oligodendroglial death in this disorder. Because an early event in periventricular white matter injury is ischemia-induced axonal disruption and because such axonal destruction could lead to a marked increase in local concentrations of glutamate, we evaluated the vulnerability of differentiating oligodendroglia to glutamate in a culture model. Oligodendroglia were isolated from mixed-glial primary cultures by a selective detachment technique and grown in a primary culture under conditions that lead to differentiation. These oligodendroglia were found to be highly vulnerable to glutamate-induced cell death. The EC50 for glutamate for a 24 hr exposure was approximately 200 microM, comparable to the value reported for neurons in conventional cerebral cortical cultures. Astrocytes, in contrast, were shown to be resistant to as much as 5 mM glutamate. Study of glutamate receptor antagonists and glutamate transport substrates showed that the glutamate-induced oligodendroglial death was not related to a receptor mechanism, as operates in neurons, but rather was secondary to glutamate uptake by the oligodendroglia. Glutamate transport by high-affinity, sodium-dependent and by sodium-independent systems was shown. The central importance of glutamate uptake for the toxic effect of glutamate was shown by total prevention of the oligodendroglial toxicity by the simultaneous inhibition of glutamate uptake by the specific inhibitor D,L-threo-beta-hydroxyaspartate. Subsequent observations showed that the toxicity of glutamate was mediated by free radical attack, the consequence of glutathione depletion, apparently caused by the action of a glutamate-cystine exchange mechanism that results in cystine and thereby glutathione depletion. Thus, addition of cystine or cysteine totally prevented the glutamate toxicity to oligodendroglia. Second, glutamate exposure led to cystine efflux. Third, glutathione levels decreased markedly in cells exposed to glutamate, and this marked decrease preceded the loss of cell viability. Fourth, glutamate toxicity could be prevented totally by exposure to different free radical scavengers, vitamin E and idebenone. The data thus show that glutamate is highly toxic to oligodendroglia. Moreover, the findings raise the possibilities that such glutamate toxicity is operative in the oligodendroglial cell death associated with ischemic processes that disrupt axons, such as periventricular white matter injury of the premature infant, and that novel therapies directed against glutamate transport, glutathione depletion, and free radical attack might be beneficial in prevention of that injury.