Fig. 4: ACh increases following reversal predict lose-shift behavior.
From: Spatially heterogeneous acetylcholine dynamics in the striatum promote behavioral flexibility
a Schematic of the VR-based reversal learning paradigm. b Example choice behavior from a representative mouse during the last 40 pre-reversal trials and across reversal. Blue and orange ticks indicate correct and incorrect choices, respectively; black trace, five-trial moving average of choice performance. c Left, Maze arm selectivity indices across reversal sessions (R1-R5) for all animals (n = 11). Gray lines, individual means; black, group mean ± S.E.M. Right, choice performance before and after reversal (two-tailed, paired t test, P = 3.93 ×10−5). d Anticipatory licking in the pre-outcome zone (280–320 cm) during early and late reversal phases: Repeated measures ANOVA: main effect of session (F(1,20) = 0.57, P = 0.46); outcome (F(1,20) = 0.14, P = 0.71); interaction (F(1,20) = 2.96, P = 0.10). Mice showed a trend toward higher anticipatory licking when approaching the new rewarded arm. e Same as (d) but for approach velocity: main effect of session (F(1,20) = 9.75, P = 0.0054); outcome (F(1,20) = 0.60, P = 0.45); interaction (F(1,20) = 0.58, P = 0.45). f, g Example iAChSnFR signals aligned to trial outcomes during reversal for a representative mouse (f), and across mice (n = 11) (g). Bottom, heatmaps show mean normalized responses (reward, n = 212 trials, no-reward, 235). h Mean ACh activity (z-scored ΔF/F) during the outcome period (reward vs no reward). Significant difference between conditions, two-tailed, paired t test, P = 1.40 × 10−3. i Left, mean iAChSnFR transients preceding lose-shift (dark green) and lose-stay (light green) responses. Right, trials matched across early (top) and late (bottom) post-reversal windows. j Mean ACh signals were positively correlated with lose-shift probability following unrewarded outcomes (simple linear regression, Pearson’s r = 0.67, P = 0.03, n = 11). k ACh signals did not correlate with velocity during unrewarded outcomes (0-2 s window), (Pearson’s r = −0.096, P = 0.14, n = 235 trials). In (j, k), solid lines denote linear fits, dotted lines, 95% CIs. Shaded areas, S.E.M. In box plots, center lines depict the median, box limits represent the 25th and 75th percentiles, and whiskers, data range. Source data are provided as a Source Data file.
