Figure 1: A prefrontal-dependent serial set-shifting task for head-fixed mice.

A. Task schematic. Top left: diagram of stimulus delivery and lick response configuration. Top right: trial sequence (see Results, Methods). Bottom: a table of the discrete training stages culminating in the serial extra-dimensional set-shifting task. SD = Simple Discrimination, CD = Compound Discrimination, IDS1 = Intradimensional Shift, Rev = Reversal, EDS = Extradimensional Shift, IDS2 = Intradimensional Shift, SEDS = Serial Extradimensional Shifting.

B. Summary of licking behavior. Top: rasterized and trial-aligned lick times for an example session. Bottom: Summary lick time histogram (655,236 licks, 150 sessions, 32 animals). On 95% of trials, the choice lick (first lick made after stimulus termination) came in the first 0.346ms of the response window, a latency equal to a single frame of imaging.

C. Left: Number of trials to criterion, Intradimensional vs Extradimensional Shift, n=53 animals. Signed rank z=−2.9, p=0.0034, for IDS1/EDS; z = −4.0, p = 6×10^-5 for IDS2/EDS, z=1.4, p=0.15 for IDS1/IDS2. Right: Mean trials to criterion during SEDS sessions, whisker rule vs odor rule. N=115 animals, signed rank z=0.14, p=0.9.

D. Muscimol infusion in PFC. Left: Horizontal section (−1.95mm ventral from brain surface at AP/ML Bregma, fluorescent muscimol). Right: Relative multiunit firing rate (proportion of channel maximum, 0.5Hz bins, 100s moving average).

E. Left: trials to criterion in EDS sessions during transcranial infusion (SAL=saline; MUS=muscimol). N= 12, 13 mice (SAL, MUS). Rank sum z=2.4, p=0.02. Right: Number of trial blocks reaching criterion performance in SEDS sessions following 10 rule shifts. N=12; median blocks (BL/MUS/SAL): 4, 1.5, 4. Signed rank p=0.0005 (BL/MUS), 0.001 (SAL/MUS), 0.68 (BL/SAL). Median total trials completed: SAL =644; MUS=651; Signed rank p=0.52.

F. Incongruent trial performance by recency of previous incongruent trial during SEDS sessions (mean ± SEM). Top: performance during the last 25 trials of trial blocks, prior to reaching criterion; bottom: performance during the first 25 trials of trial blocks, after undergoing rule shifts. N = 693 sessions in 131 animals. ANOVA for ICG trial recency vs ICG trial performance: t = −4.68, p = 3×10^-6.

G. SEDS trial performance relative to rule shift, mean ± SEM. Same sessions as in F. Note that, because trial shifts were triggered by criterion performance, correct response rates immediately following the rule shift are a better estimate of true performance than trials immediately preceding the shift