The human brain depends on its blood supply for a continuous supply of oxygen and glucose. Irreversible brain damage occurs if blood flow is reduced below about 10 ml/lOO g tissue/min and if blood flow is completely interrupted, damage will occur in only a few minutes. Unfortunately, such reductions (ischemia) are common in disease states: either localized to individual vascular territories, as in stroke; or globally, as in cardiac arrest. Cerebral hypoxia can also occur in isolation, for example in respiratory arrest, carbon monoxide poisoning, or near-drowning; pure glucose depri vation can occur in insulin overdose or a variety of metabolic disorders. As a group, these disorders are a leading cause of neurological disability and death; stroke alone is the third most common cause of death in North America.

Despite its clinical importance, little is known about the cellular patho genesis of hypoxic-ischemic brain damage, and at present there is no effective therapy. A critical question has been why brain, more than most other tissues, is so vulnerable to hypoxic-ischemic insults. In particular, certain neuronal subpopulations, such as hippocampal field CAl and neocortical layers 3, 5, and 6, are characteristically destroyed after sub maximal hypoxic-ischemic exposure. A possible answer has emerged in the last few years: At least some of this special vulnerability may be accounted for by the central neurotoxicity of the endogenous excitatory