Figure 3. Neurexin1α mutants display a deficit in the selection of actions based on costs.

(A) Effort paradigm schematic. Mice distribute choices in a session with fixed contingency lasting 150 trials. Animals were given choices with equal reward outcomes, but different effort requirements (FR3 vs. FR1). (B) Nrxn1α KOs (blue, n = 10) choose less costly alternatives at a lower rate than their WT littermates (gray, n = 11) (two-way RM ANOVA). The distribution of choice in both WT and KO mice is altered over the course of the block as mice acquire information about the reward contingency, with a stable difference observed over the final 75 trials (two-sample t-test *p=0.023). (C) Nrxn1α KOs exhibited a clear interaction between trial and latency to initiate, slowing as they performed more high effort trials (two-way RM ANOVA). Nevertheless, there was no statistically significant difference in engagement at steady state (two-sample t-test p=0.14). (D) The longer choice latencies previously described in Nrxn1α KOs was observed in steady-state responding (two-way RM ANOVA; two-sample t-test *p=0.017). All data represented as mean ± SEM.