Fig. 9: DA activity immediately after US reward delivery predicts transition to consummatory behavior.
a Left: Heatmap of change in lick rate for an example well-trained session. Right: Average change in lick rate for the example session. b Average change in licking rate after the US (n = 6 mice). c Reward transiently increases licking rate (One-way repeated-measures ANOVA (F(1.5,7.7) = 11.44, p = 0.0065. Post-hoc: increased licking rates in first 250 ms vs. pre-reward, p = 0.0429; stable level resumes afterward, p = 0.028; n = 6 mice). d Overlayed timeseries of representative DA population (n = 12 neurons) and change in licking rate for a single session. DA neurons precede the increase in licking rate after reward. e Licking rate in the 1 s window after reward was predicted by firing rate in the first 250 ms after the US (linear regression, R2 = 0.81; n = 5 mice). f Left: Example heatmap showing the changes in force occurring after US delivery. Right: Example of rapid force direction reversal after US delivery. g Average change in force aligned to US delivery in trained mice (n = 6 mice). h Change in force and average firing rate for a DA population (n = 12 neurons) recorded in a well-trained mouse. DA neuron activity increases prior to forward force generation. i Timing of DA activation relative to force and licking UR behaviors (One-way repeated-measures ANOVA, F(1.57,7.85) = 44.50, p < 0.0001; Post-hoc: DA neuron population activity differed in latency to peak change in licking (p = 0.015) and peak increases in force (p = 0.0012). Licking rate preceded changes in forward force (p = 0.0011). DA neurons and the brief backward force generated right after US delivery did not have different latency, p = 0.177, n = 6 mice). j Comparison of the regression coefficients between DA neurons after the reward and force and licking UR behavior (One-way repeated-measures ANOVA (F(2,12) = 19.7, p = 0.0002; Post-hoc: R2 values between DA and Licking Rate were higher than those for Impulse (p = 0.0003) and change in force (p = 0.0006), n = 5 mice). Error bars represent SEM. Source Data provided.
