Delayed neuronal loss after administration of intracerebroventricular kainic acid to preweanling rats

Abstract

Excitotoxins, such as kainic acid (KA), have been shown to produce both immediate and delayed neuronal degeneration in adult rat brain. While preweanling rats have been shown to be resistant to the immediate neurotoxicity of KA, the presence of delayed neuronal loss has not been investigated in such animals. To determine whether intracerebroventricular (i.c.v.) administration of KA would produce delayed neuronal loss, preweanling rats were administered 5 nmol or 10 nmol KA i.c.v. on postnatal day 7 (P7) and then examined at P14, P45, and P75. Using three-dimensional, non-biased cell counting, neuronal loss was observed in the CA3 subfield of the hippocampal formation at P45 and P75 in animals administered 10 nmol KA, as compared to animals administered 5 nmol KA or artificial cerebrospinal fluid. Further, the amount of immunoreactivity to jun, the protein product of the immediate early gene, c-jun, adjusted for the number of remaining neurons was increased in the same brain areas. Antibody labeling of inducible heat shock protein and glial fibrillary acidic protein was not similarly increased in animals administered i.c.v. KA. The data suggest that while i.c.v. KA does not produce immediate neuronal loss in preweanling rats, the hippocampus is altered so that neuronal loss occurs after a delay, perhaps through apoptosis. These findings may be relevant to the pathogenesis of neuropsychiatric disorders, such as schizophrenia, that are characterized by early limbic–cortical deficits but onset of illness in young adulthood.

Introduction

The pathogenesis of neuropsychiatric brain diseases that have their clinical onset in late adolescence, such as schizophrenia, remains poorly understood. Many investigators have hypothesized that the initial steps in the pathogenesis of schizophrenia involve abnormalities in neuronal migration and synaptogenesis during fetal life 10, 14, 15. Support for this hypothesis comes from neuroimaging and neuropathological studies of brains from schizophrenia patients, which have provided evidence of neuromorphometric and cytoarchitectural abnormalities consistent with defects of neurodevelopment in corticolimbic brain structures [cf. Refs. 4, 5, 14, 23, 25, 43]. However, while such abnormalities suggest early neurodevelopmental abnormalities in schizophrenia, clinical symptoms of the disorder are usually delayed until adolescence 15, 40. One explanation for the discrepancy between the apparent neurodevelopmental defects observed in schizophrenia and the delayed clinical onset of illness is the fact that the cortico-limbic circuits mediating schizophrenic symptoms only become anatomically mature and fully functional during adolescence 9, 40.

Models of cortico-limbic neuronal loss in rats caused by the intraperitoneal (i.p.), subcutaneous (s.c.) or intracerebroventricular (i.c.v.) administration of the excitotoxin, kainic acid (KA), have been studied for several years 30, 31, 37. KA has long been known to preferentially damage corticolimbic neurons, particularly those in the CA3 and CA4 subfields of the hippocampus when it is administered i.c.v. [31]. More recently, it has been shown that i.c.v. KA-induced corticolimbic neuronal loss in adult rats is associated with increases in subcortical dopamine D2-like receptors, increased sensitivity to dopaminergic agonists, and decreased sensitivity to dopamine antagonists after several days or weeks 3, 6, 7, 8. The hippocampus projects to subcortical brain areas, such as the nucleus accumbens [2], and disruption of hippocampo-striatal projections may explain the dopaminergic changes observed 15, 22. Such changes in dopaminergic function may be highly relevant to the pathophysiology of schizophrenia since dopamine antagonists are widely used to for its treatment [15].

Within 24 h after i.c.v. KA administration in adult rats, neurons within the CA3 subfield of the hippocampus are most severely damaged [31]. However, after several days, neurons within the CA1 subfield also begin to degenerate [16], perhaps because CA1 neurons are rendered hyperexcitable after the loss of CA3 neurons [35]. The dying CA1 neurons also show increased immunoreactivity for jun, the protein product of the immediate-early gene, c-jun [16], increased message for c-fos [29], and characteristic patterns of DNA fragmentation (i.e., laddering) [29], which suggests that delayed neuronal death following KA administration may be occurring through apoptosis 20, 24, 34. Also, cycloheximide, which blocks apoptosis, will protect adult rat brain from KA-induced neuronal damage [38].

In contrast to the literature involving adult rats, KA toxicity has been less thoroughly explored in developing rats. When it is administered i.p., s.c. or i.c.v. prior to P21, immediate neuronal loss has not been observed 1, 11, 13, 35, 36, 42. Lowered sensitivity to the neurotoxic effects of KA prior to P21 has been related to the immaturity of glutamatergic hippocampal afferents; i.e., there is a strong temporal relationship between the development of glutamatergic afferents into the hippocampus and the sensitivity of the hippocampus to KA [36].

It has not yet been determined whether delayed neuronal loss may occur after administration of i.c.v. KA to developing rats. While immediate neuronal loss is not apparent, delayed neuronal loss might occur in developing animals as it does in adult animals through mechanisms such as apoptosis [16]. Therefore, the aims of this study were to re-evaluate KA toxicity in developing rats by administering KA i.c.v. to preweanling rats (i.e., at post-natal day 7 [P7]), paying particular attention to the assessment of corticolimbic neuronal loss at immediate and delayed time points after i.c.v. KA administration. Although not the primary goal of this study, we also performed studied selected immunohistochemical markers to generate hypotheses regarding the potential mechanisms that might be involved in any neuronal loss observed.