The experiments reported in this dissertation were designed to examine the neurobiology of T-maze spatial reversal learning in weanling rats. Spatial reversal learning is a robust behavioral paradigm to examine the development of learning and memory in young rats. This model is comparable to models used with adult and developing primates to examine the neuropsychology of learning and memory. Primate and rodent animal models are often used to investigate the cognitive impairments of certain developmental disorders. Children with neurodevelopmental disorders often have deficits in behavioral tests that involve changes in reward contingency, such as reversal learning. Behaviorally, during discrimination reversal training, learning the reversal phase is much more difficult than learning the initial discrimination. The behavioral processes that underlie reversal learning involve learning to suppress the previously reinforced response pattern from the acquisition training phase, while learning the new competing response pattern during the reversal training phase. This thesis tested the hypothesis that successful spatial reversal performance in weaning rats requires the contribution of N-methyl-D-aspartate (NMDA) receptor function in several forebrain structures.

Experiment 1 (Chapter 6) evaluated spatial reversal learning performance on postnatal day (P) 26 following NMDA-receptor antagonism in the dorsal hippocampus (dHPC). A range of doses of dizocilpine, the compound also referred to as MK-801 (MK), was infused into the dHPC during the reversal phase of training (Reversal Only design). T-maze spatial discrimination reversal training took place in a single day in two experimental sessions, one in the morning and the other in the afternoon. Previous work has demonstrated that systemic administration of MK severely impaired spatial reversal learning performance in weanling rats (Chadman, Watson, & Stanton, 2006). Experiment 1 found that intrahippocampal MK administration dose-dependently impaired spatial reversal learning performance and enhanced response perseveration on the position habit trained in acquisition. These findings confirmed that NMDA-receptor function in the dHPC was necessary for successful spatial reversal learning in weanling rats. The previous systemic MK reversal impairment was specific to the reversal learning phase, did not impair initial learning of the position discrimination, and was not due to state-dependent learning effects (Chadman et al., 2006). Experiment 2 used the same T-maze procedures as Experiment 1, except that MK was administered before both training phases, acquisition only, reversal only, or neither training phase (Reversal Specificity design). The results established that the spatial reversal learning impairment in Experiment 1 was specific to the reversal learning phase, NMDA-receptor antagonism did not impair acquisition of the position habit, nor was the reversal impairment due to state-dependent learning effects.

Spatial reversal learning was further examined following NMDA-receptor antagonism in the medial prefrontal cortex (mPFC, Chapter 7, Experiments 3-4) and dorsomedial striatum (dmSTR, Chapter 8, Experiments 5-6). Experiments 3 and 5 had a similar rationale and predictions as Experiment 1 (Reversal Only design). The results suggested that NMDA-receptor function in both the mPFC and dmSTR are also necessary for successful reversal learning performance in P26 rats. The rationale and predictions of Experiments 4 and 6 were the same as Experiment 2 (Reversal Specificity design). In both brain regions, the learning deficit following intracranial MK administration was specific to the reversal learning phase, initial acquisition performance was spared after NMDA-receptor antagonism, and the impairment during the reversal learning phase was not due to state-dependent learning effects. As a whole, the results of Experiments 1-6 establish that NMDA-receptor function in a variety of forebrain regions is necessary for spatial reversal learning performance in weanling rats. The findings of this dissertation provide a foundation for future lines of research to investigate the developmental psychobiology of reversal learning. These findings also serve to improve the quality of animal models of neurodevelopmental disorders, by expanding the likely neural structures that may be implicated in these disorders.