Fig. 2

PreL is required to apply and IL to shift from learnt rules. a Representative images for delivery to PreL or IL of Cre-dependent AAV9-pharmacogenetic activator in PV-Cre mice. Bgtx: visualization of (Bungarotoxin + ) virus infected cells; PV: PV immunocytochemistry. Bar: 100 μm (left) and 35 μm (right). b Schematic of set-shifting protocol (background color: rewarded association). c–f Requirements for activity in PreL or IL during IEST subtask learning. c CD (n = 5 each; (F(2, 12) = 17.1, ^***P) or recall of SD (n = 5 each; F(2, 12) = 1.9, P = 0.1916 ns); d IDS (n = 5 each; F(2, 12) = 24.62, ^***P) or recall of IDS (n = 5 each; F(2, 12) = 0.78, P = 0.4722, ns); e IDSRe (n = 5 each; F(2, 12) = 233.5, ^***P) or reversal of IDSRe (n = 5 each; F(2, 12) = 127.9, ^***P); f EDS (n = 5 each; F(2, 12) = 43.01, ^***P). Bar diagrams: trials to criterion for virus-expressing mice treated with vehicle (Ctrl) or mice in which PreL (blue) or IL (yellow) was silenced (PV-neuron activation; PSEM ligand 20 min before learning); subtask during which silencing was applied indicated in bold. “IDS” indicates that cues are as in IDS, but previous history is distinct from IDS. g Summary of PreL and IL roles in IEDS learning. Error bars: SEM; one-way ANOVA followed by Dunnet’s post hoc; P < 0.05 (^*), 0.001 (^**), 0.0001 (^***)