The neurotoxic effects of glutamate have been known for years (11, 45). The neuromediator function of glutamate and aspartate was proven later, and now these transmitters are regarded as the main excitatory transmitters, achieving their effects through ionotropic as well as metabotropic receptors. Putting together these two basic physiologic observations, McDonald et al. formulated the excitotoxic hypothesis: “accumulation of the excitatory amino acid transmitters, glutamate and aspartate, in the extracellular space, as the result of increased synaptic release and reduced cellular uptake, produces excitotoxic injury through excessive activation of excitatory amino acid receptors” (40). The excitotoxic mechanism plays a role in hypoxia-ischemia injury in mature (6) as well as immature brain (23). Another field where excitotoxicity is of primary importance is epilepsy (18, 41, 42, 53).

Agonists of all types of ionotropic excitatory amino acid receptors—NMDA, AMPA, kainate—are able to induce neuronal injury in the central nervous system. This effect was originally demonstrated for kainic acid (44, 46, 54) and later for NMDA (58) and AMPA (43). The sensitivity of immature brain to the excitotoxic effects of agonists of the three ionotropic receptors develops differentially. Intrahippocampal kainate injection resulted in neuronal death in 1-week-old rats (13, 14), but when administered systemically it does not induce immediate neuronal damage in rats up to the third postnatal week (striatum [8] and hippocampus [5, 56]). In contrast, neuronal damage may be seen after a long latent period (a few months) even in rats injected with kainate systemically as well as intrahippocampally at the age of 7 or 12 postnatal days (2, 31). Unlike this specific response of the immature brain to kainate, the excitotoxic action of NMDA, AMPA, and quisqualate, resulting in neuronal death, can be demonstrated even in the first week of life, and the sensitivity of rat pups is higher than that of adult animals (17, 19, 22, 24, 37, 67). This hypersensitivity peaks around postnatal day 10, with a small difference in sensitivity to NMDA and AMPA (36, 40). Brain injury was also demonstrated after administration of the type I metabotropic glutamate receptor agonist 1S,3R-ACPD (38). Excitatory amino acids represent a common pathway of neuronal destruction in immature animals after various pathologic changes (hypoxia-ischemia [47], epileptic seizures [42, 64], and administration of aminooxyacetic acid [39] or malonate [16]). The temporal evolution of brain damage is very fast; nuclear magnetic resonance imaging demonstrates changes 15 minutes after intracerebroventricular NMDA injection (63).

The administration of agonists of all types of ionotropic receptors and of type I metabotropic receptors elicits seizures in immature rats. With repeated administration, kainic acid elicited seizures even in 7-day-old rat pups, an age when immediate morphologic damage in the hippocampal field CA3 is absent (1, 4, 10, 62). The sensitivity to kainic acid induction of seizures was highest in 7-day-old rats and decreased moderately with age (62). NMDA is also active during very early stages of development (35, 52), and sensitivity to its convulsant action decreases dramatically with age—the effective doses in 7-day-old rats are 100 times lower than those for adult animals (35). d,l-Homocysteic acid, producing its excitatory action mainly by means of NMDA receptors (but not exclusively [61], and unpublished data from our laboratory), reliably elicits seizures after postnatal day 7 when administered systemically (33). The epileptogenic action of d,l-homocysteic acid also decreases during development, but the decline is not so steep as with NMDA (33). Homocysteine also elicits seizures from postnatal day 7, but there is no simple relationship between efficacy and age (29), probably because the roles of NMDA and non-NMDA receptors in the convulsant action of homocysteine vary during development (15). AMPA is also effective at an early age (10 days), but quantitative data are lacking (52). Similarly, a convulsant action of the type I metabotropic glutamate receptor agonist 1S,3R-ACPD was demonstrated in 7-day-old rats (39).

Excitatory amino acid agonists did not induce the same pattern of seizures, and they may have a different influence on models of electrically induced of epileptic seizures in developing rats (34). Kainic acid elicts a sequence of automatisms (“wet-dog shakes” are the most conspicuous behavior in adult rodents, whereas scratching is the principal behavioral symptom in rat pups), minimal seizures (clonic seizures involving head and forelimb muscles, with preservation of righting reflexes), and generalized tonic-clonic seizures, with a loss of righting ability at the beginning of the tonic phase (62). NMDA never elicits minimal seizures. Its action starts with a period of immobility, then a hyperlocomotion appears, and after a somewhat longer latency (if the dose is high enough) seizures characterized by violent generalized clonic-tonic convulsions are observed. Generalized seizures invariably lead to death. Rat pups up to the age of 18 days exhibit an age-specific phenomenon consisting of flexion, emprosthotonic seizures. The animals are curled up into a ball for some seconds or maximally some tens of seconds during these seizures (35). The seizure pattern induced by homocysteic acid is practically the same as that elicited by NMDA (33). On the other hand, homocysteine is able to induce a mixture of both patterns, that induced by NMDA and that induced by kainic acid, again indicating the possibility of multiple mechanisms of action of this drug on excitatory amino acid receptors (29). These observations might reflect the actions of kainic acid and NMDA on different brain structures, with minimal seizures taken as forebrain seizures and generalized tonic-clonic seizures (and, with a high probability, also clonic-tonic seizures) generated in the brain stem (7). Unfortunately, no data on the generation of flexion seizures are available. The age dependence of flexion seizures may reflect either an increased sensitivity of immature brain to NMDA (as has been repeatedly noted in the literature) or, less probably, an uneven maturation of different brain structures, which would allow the generator of flexion seizures to express the seizure pattern up to the moment when another brain structure (or structures) is set into action.

The data for kainic acid indicate that immediate neuronal death is not always connected with seizures but may reflect the toxicity of kainate. Seizures and neuronal death might represent two parallel phenomena induced by excitatory amino acids.

Excitatory amino acids are the main agents responsible for neuronal destruction after severe seizures elicited by other mechanisms. Status epilepticus (SE) induced by the cholinomimetic pilocarpine (in two modifications, the high-dose model and the lithium-pilocarpine model) leads to extensive neuronal damage in adult brain (12). The role of NMDA receptors in pilocarpine-induced seizures has been demonstrated (55). Pilocarpine-induced SE and its morphologic consequences have also been described in developing rats (9, 20). There are marked differences in the structures compromised, as well as in the extent of neuronal damage, in SE induced at different developmental stages (9, 25, 49). Acute changes evaluated in Nissl-stained sections are much more expressed in animals undergoing SE as adults than in rats seizing at an early age, and the same is true for kainic acid–induced seizures. On the other hand, histochemical (25) or immunohistochemical methods (28) are able to show specific changes even in rats seizing at the age of 12 days. We are at the beginning of exact mapping of the morphologic consequences of SE induced at different stages of maturation, because attention is usually focused on the hippocampus (21, 26, 27, 49, 51, 65). Neuronal death may be induced also by electrically induced models of SE (51, 60). Even less is known about the functional consequences of SE that might be connected to excitotoxic damage. The studies reported here demonstrate that even the youngest rats exhibit longlasting consequences of SE in motor performance and spontaneous behavior, but the impact of these changes on brain development is far from being understood (30, 48, 57, 59, 66). Another open question is the role of neuronal death in chronic epileptogenesis in immature brain (3, 50). Ongoing work from our laboratory is addressing these questions.