Comparison of predictive validity of two autism spectrum disorder rat models: Behavioural investigations.

The valproic acid model has been shown to reproduce ASD-like behaviours observed in patients and is now widely validated for construct, face, and predictivity as ASD model in rat. The literature agrees on using a single exposition to 500 mg/kg of VPA at gestational day 12 to induce ASD phenotype with the intraperitoneal route being the most commonly used. However, some studies validated this model with repeated exposure by using oral route. The way of administration may be of great importance in the induction of the ASD phenotype and a comparison is greatly required. We compared two ASD models, one induced by a unique IP injection of 500 mg/kg of body weight at GD12 and the other one by repeated PO administration of 500 mg/kg of body weight/day between GD11 and GD13. The behavioural phenotypes of the offspring were assessed for the core signs of ASD (impaired social behaviour, stereotypical/repetitive behaviours, sensory/communication deficits) as well as anxiety as comorbidity, at developmental and juvenile stages in both sexes. The VPA IP model induced a more literature-compliant ASD phenotype than the PO one. These results confirmed that the mode of administration as well as the window of VPA exposure are key factors in the ASD-induction phenotype. Interestingly, the effects of VPA administration were similar at the developmental stage between both sexes and then tended to differ later in life.