Figure 2.

Bidirectional modification of decision-making by Abl-family kinases. a, Experimental timeline with two possible infusion times indicated by arrows. b, DMS STI-571 infusions before response-outcome contingency degradation in moderately trained mice occluded goal-directed decision-making, as indicated by nonselective responding on degraded and nondegraded apertures. c, Infusions immediately after response-outcome contingency degradation also resulted in nonselective responding. d, When responding was reinstated using a random interval schedule, STI-571-infused mice responded more for the same number of reinforcers. e, After this “extended” training, both groups had developed habits and were insensitive to response-outcome contingency degradation. f, STI-571 infusions in the DLS after extended training (in a separate group of mice) reinstated sensitivity to subsequent response-outcome contingency degradation. g, DLS STI-571 infusions after response-outcome contingency degradation in extensively trained mice, however, had no effects. h, Representative infusion sites with coordinates relative to bregma for DMS (h₁,h₂) and DLS (h₃,h₄) infusions are shown. Bars and symbols represent group mean ± SEM. *p < 0.05.