Optimizing prenatal valproic acid exposure in zebra finches to model autism-related phenotypes

Rodent models dominate autism spectrum disorder (ASD) research, yet their translational validity is limited for a heterogeneous condition with complex social-vocal phenotypes.

To evaluate the zebra finch (Taeniopygia guttata) - a highly social, vocal-learning songbird - as an ASD model by adapting embryonic valproic acid (VPA) exposure and identifying dose-timing conditions that preserve viability while perturbing social behavior.

We used a 3 × 3 design crossing VPA dose (0, 0.3, 0.6 µmol/egg) with manipulation day (incubation days 8-10). Hatching success, post-hatching survival, growth, and social behavior in a bird-adapted three-chamber test were quantified with mixed-effects models (binomial/Cox/linear/negative binomial, beta-binomial).

Hatching success decreased at 0.6 µmol and with day-8 injections; day-9 performed best. Post-hatching survival was lower at 0.3 µmol versus controls (limited inference for 0.6 µmol due to small n). Growth showed dose×sex interactions: females at 0.6 µmol were lighter with shorter tarsi; wings showed a similar tendency. In sociability (phase 1), time-based measures were unchanged, 0.6-µmol birds made more approaches to the closest perch, while day-8 birds approached less than day-9. In phase 2, time on the familiar-side closest perch showed a day×sex interaction (day-8 females spent less time than day-8 males), and approaches to that perch increased at 0.6 µmol. Overall sociability (combined closest perches) and total activity did not differ by group.

Embryonic VPA in zebra finches yields selective, partly sex-dependent developmental and social effects. Zebra finches remain promising; day-9 intermediate doses (~ 0.35-0.50 µmol) warrant follow-up to map dose-response and balance viability with behavioral detectability.