Fig. 1: Vigorous execution of self-paced movements requires contralateral DA.

a, Experimental setup. b, Trial structure: mice hold the lever above the detection threshold (gray) for at least 300 ms before pressing the lever inward using their left forelimb. Deflections large enough to cross the reward threshold (green) within 0.5 s trigger delayed delivery of water rewards. To be considered a press, lever deflections must exceed the ‘press’ threshold (red). c, Example lever trace. Blue drops: rewards. d, Example presses from one mouse aligned to detection threshold crossing (gray: individual presses; blue: session mean). e, Probability distribution of peak velocity (left), maximum amplitude (middle) and latency to first press from last reward delivery (right) for all recorded presses (40 sessions in 18 mice). Dashed lines: ‘press’ (red) and ‘reward’ (green) thresholds. f, Mean inter-reward interval across 18 mice (blue) and population (black). g, Mean number of rewarded presses at baseline (black) and following devaluation (blue; N = 7, P = 0.0156, Wilcoxon test). h, Top-down schematic of mouse brain showing 6OHDA injection sites in separate cohorts to produce mDAN hemilesions ipsilateral (blue) or contralateral (red) to lever-pressing forelimb. i, Left: number of rewarded presses from mice lesioned ipsilaterally (blue; N = 7) or contralaterally (red; N = 8; P = 0.0009, unpaired t-test). Same for median peak velocity (middle; P = 0.0274, unpaired t-test) and press amplitude (right; P = 0.0008, unpaired t-test). j, Number of rewarded presses before (black) and after lesioning mDANs contralateral to the lever-pressing forelimb in the absence (red) or presence (green) of levodopa (N = 8; treatment: P < 0.0001, one-way ANOVA; post-lesion baseline, P < 0.0001; post-lesion levodopa, P = 0.19; post-lesion washout, P = 0.0017, all post hoc Dunnett’s tests versus pre-lesion). k, Same as in j for median peak press velocity. Note that two mice in the washout group produced too few presses for analysis (treatment: P = 0.0024, mixed-effect model; post-lesion baseline, P = 0.0069; post-lesion levodopa, P = 0.62; post-lesion washout, P = 0.049, all post hoc Dunnett’s tests versus pre-lesion). l, Same as in k for median press amplitude (treatment: P < 0.0001, mixed-effect model; post-lesion baseline, P < 0.0001; post-lesion levodopa, P = 0.84; post-lesion washout, P = 0.0021, all post hoc Dunnett’s tests versus pre-lesion). Gray lines in g and j–k reflect paired data per mouse. Group summary data are mean ± s.e.m. ITI, inter-trial interval; NS, not significant.

Source data