Social Reward Circuit Dissection in the Rat Valproate Model of Autism

Background: The Ventral Tegmental Area (VTA) is a heterogeneous region that plays a key role in the control of rewarding activity, motivation and social behavior. It is mainly composed of dopaminergic neurons (> 60%), intermingled with GABAergic neurons (~30%) and glutamatergic neurons (2-3%). Activation of dopaminergic or GABAergic VTA neurons can induce reward or aversion. VTA prominently sends dopaminergic projections to the Ventral Striatum (VS), another region actively involved in processing motivation, decision making, and reward. It has been shown that VS activation following dopamine neurons stimulation in VTA has rewarding effects. Moreover, in humans, diminished fMRI BOLD striatal responses to social reward have been reported in autism spectrum disorder.

In humans, prenatal exposure to valproic acid (VPA) is known to increase the probability to develop autism in children up to ten-fold. Similarly, rats prenatally exposed to VPA display impairment in social behavior, increased anxiety and hyperactivity.

Objectives: In this study, we used the rat VPA model of autism to dissect the VTA-VS circuitry and its link to social behavior. We recorded local field potential and unit activity in freely behaving VPA-exposed and vehicle (VEH) control rats during social behavior. Our goal was to measure brain activity during social interaction to detect differences in social reward signaling between VPA and VEH control rats. In a second step, we wanted to test if treatment with a vasopressin V1a receptor antagonist was affecting this circuitry.

Methods: VPA-exposed (N=5) and VEH rats (N=5), implanted with single wires electrodes in both VTA and VS, were free to explore an open field arena with two Plexiglas boxes placed at opposite corners. The two insert boxes had three rows of holes to facilitate social interactions. The test was composed of three different time blocks for a total duration of 35 minutes. The first block consisted of 10 minutes of exploration with empty Plexiglas boxes. During the second time block, an unfamiliar conspecific was placed inside one of the two boxes and an object was inserted in the other for 15 minutes. In the third time block, the boxes were empty again for 10 minutes. Video tracking of the task allowed for segmentation of electrophysiological brain activity according to the position of the nose of the experimental animal in the different zones (object, animal and neutral zone).

Results: In VEH control rats, we found that the VS showed higher 80Hz Gamma Oscillation power in proximity to social cues compared to inanimate objects or empty environments. We are further characterizing differences in VPA rats and interactions with V1a antagonist treatment.

Conclusions: This work will add to our understanding of the brain circuits involved in social behavior in the rat VPA model of autism and how they may be affected by V1a antagonism, a mode of action which is currently being tested in Phase III studies in adults with autism.