Figure 2. Behavior of head-fixed mice trained in a reversal task.
(A) Schematics of the trial structure in the classical conditioning task (before reversal) with four different outcomes. In each trial, one of four odors was randomly selected and presented for 1 s after a variable foreperiod (Forep). The associated outcome was delivered after a 2 s trace period, together with a tone (same tone for all trial types). Mice were presented with 140 to 346 interleaved trials (mean ± SD: 223 ± 30) per session (day). (B) Top: Mean lick rate of SERT-Cre mice in this task (n = 10) along the duration of each trial type. For each mouse, three sessions of the classical conditioning task where initial associations had already been learned were averaged. Bottom: Mean eye movement of SERT-Cre mice (n = 6) along the duration of each trial type. Shaded areas represent s.e.m. (C) Reversal of CS–US contingencies (negative reversal: CS 1 and 2; positive reversal: CS 3 and 4). (D) Anticipatory licking (mean of 500–2800 ms after odor onset, after subtracting the baseline) across mice for sessions around reversal, showing that the lick rate triggered by the presentation of each odor is adjusted after reversal (n = 8, two-way ANOVA with factors day (days −2 and −1 are considered together) and mouse, main effect of day: F4,2597 = 722.14, p<0.001 for odor 1, F4,2554 = 355.53, p<0.001 for odor 2, F4,2513 = 104.93, p<0.001 for odor 3, F4,2559 = 381.55, p<0.001 for odor 4). Colors follow odor identity as in (A). ***p<0.001.