Abstracts

Rationale: The overall diagnosis of autism spectrum disorder (ASD) varies widely from very mild to severe social and cognitive impairments. Although select genes have been identified in cases of severe ASD, these genes do not yet account for the mild and more functional ASD cases that are not pervasive until years after birth. We hypothesized that subconvulsive activity in early postnatal life may contribute to the development of autism and affect the seizure threshold. Methods: To establish a valid ASD model, daily subconvulsive step-up doses of kainic acid (KA) (25-100 µg) were administered subcutaneously to rat pups for 15 days beginning on postnatal (P) day 6 to elevate early life neuronal hyperactivity. Handling, open field, and social behavior were monitored 24 h after the last subconvulsive treatment. The seizure threshold was simultaneously examined in a subpopulation by unilateral intrahippocampal (ih) microinjection of a single convulsive dose of KA via a bipolar electrode/cannula assembly. All brains were histologically evaluated with NeuN immunohistochemistry and Nissl staining at the same time point. Results: Within 10 days of subconvulsive KA postnatal treatment, increased exploratory and escape behavior in the handling test were observed. At the end of the treatment protocol, significant asocial traits were present without anxiety differences in the elevated plus maze. Nissl staining revealed morphological changes in cell shape as well as increases in the number of hyperbasophilic neurons of the amygdala cortical complex (AMG/Cx), particularly within the basolateral amygdala nucleus (BLA). There was a lack of NeuN antigenicity extending from the BLA to the piriform and entorhinal cortex presenting an increased average gap width (Control: 142 ± 5 µm vs. ASD: 216 ± 8 µm, p < 0.01), whereas hippocampal NeuN expression was indistinguishable from controls. The striking loss of NeuN antigenicity within the AMG/Cx was positively correlated with hyperbaspohilia despite no measurable differences in total cell density of the AMG/Cx. Unexpectedly, there was a delayed onset to seizure behavioral manifestations and significant reduction in seizure severity score post-unilateral KA injection in the subconvulsive treated animals (KAih: 4.2 ± 0.2 vs. ASD+KAih: 2.2 ± 0.05, p < 0.01). Conclusions: Behavioral pathology and selective injury to AMG/Cx may be due to regular but subtle changes in glutamatergic activity related to the very high affinity of KA to this region during development. Lack of NeuN antigenicity did not represent neuronal cell loss but could reflect metabolic changes induced by low levels of constant hyperactivity. Due to isolated insult to AMG/Cx and sparing of the hippocampus, the elevated seizure threshold in ASD animals may be due to loss of critical synaptic connections between the AMG/Cx and hippocampus. These findings highlight the role of low-level hyperactivity of the AMG/Cx and may represent a novel model for mild-moderate ASDs that exhibit social and emotional impairments without otherwise identified genetic abnormalities or history of convulsive episodes.   Funding: none