Fig. 4

PreL promotes fear learning and IL drives its extinction. In all these experiments: silencing through pharmacogenetic activation of PV neurons 20 min before learning (or recall); control: vehicle delivery in virus-expressing mice. a Schematic of trace fear conditioning (tFC) and extinction protocol. b Silencing PreL at acquisition impairs freezing at 24 h recall, whereas silencing IL does not (n = 5 each; one-way ANOVA: F(2, 12) = 20.67, ^***P; Dunnet’s post hoc). c Silencing PreL or IL at recall does not affect freezing. In positive control experiments, silencing ventral hippocampus (vH) at recall suppresses freezing (n = 5 each; one-way ANOVA: F(3, 16) = 31.14, ^***P; Dunnet’s post hoc). d Silencing IL at acquisition does not affect extinction learning on the next day (n = 5 each; repeat measure two-way ANOVA: silence IL, F(1, 4) = 1.01, P = 0.372, ns). For extinction of tFC, data are shown (here and throughout the paper) as average freezing during T1–T2, T3–T4, or T5–T6; for individual freezing values see Supplementary Fig. 4. e Silencing during extinction learning: silencing IL suppresses extinction learning, whereas silencing PreL slightly accelerates loss of freezing (n = 5 each; repeat measure two-way ANOVA: silence IL vs. PreL, F(2, 8) = 64.75, ^***P; Tukey’s post hoc). f Silencing PreL or IL at recall of extinction does not affect recall of extinction learning (n = 5 each; one-way ANOVA: F(2, 12) = 0.636, P = 0.546, ns). g Silencing IL and PreL during extinction learning suppresses extinction learning (n = 5 each; repeat measure two-way ANOVA: silence PreL + IL, F(1, 4) = 199.35, ^***P; Tukey’s post hoc). h Summary of PreL and IL roles in fear learning and its extinction. Error bars: SEM. P < 0.05 (^*), 0.001 (^**), 0.0001 (^***)