Extended Data Fig. 4: Controls for optogenetic manipulations and sensitivity of behavioral assay to changes in vigor.

a, Experimental paradigm; DA levels are optogenetically increased or depressed at trial start on 30% of trials using 470 nm light (1 s for optogenetic stimulation, or 2 s for optogenetic inhibition). Lever deflections initiated on those trials are compared either to the 70% of interleaved trials without light (control trials; within-session comparison) or to 30% of trials on separate behavioral sessions without any light delivered to the midbrain or 595 nm light, which does not activate blue light-shifted opsins such as ChR2 and GtACR2 (‘sham’ trials). b, Schematic of coronal mouse midbrain section showing bilateral fiber optic placement (red lines) for silencing mDANs (orange). c, Same as b for mDAN stimulation experiments. d, Top, Representative lever trajectories from one example session (thick line: session-wide mean) for presses occurring on trials with sham optogenetic light (595 nm) off (left) or on (right). Bottom, peak velocity and amplitude of all presses performed during that example session, separated between control trials (gray; n = 147 presses) and sham stimulation trials (blue; n = 62 presses; velocity: p = 0.90; amplitude: p = 0.32, two-sample t-tests). Black bar: session median. e, Within-session change in median peak press velocity between control trials and the ~30% of trials in which optogenetic manipulations are delivered to the midbrain of Dat-Cre::Ai32 mice: No light (p = 0.96 vs. 0% change, one-sample t-test; n = 18 sessions from 6 mice), 595 nm light (4 mW; p = 0.95 vs. 0% change, one-sample t-test; n = 18 sessions from 6 mice), 470 nm light (1 mW; p = 0.74 vs. 0% change, one-sample t-test; n = 12 sessions from 4 mice), 470 nm light (4 mW; p = 0.73 vs. 0% change, one-sample t-test; n = 18 sessions from 6 mice) and 470 nm light (10 mW; p = 0.98 vs. 0% change, one-sample t-test; n = 12 sessions from 4 mice). Population means ± SEM are shown in black. No difference was observed between conditions (treatment: p = 0.99, one-way ANOVA). f, Same as e for press amplitude (No light: p = 0.24; 595 nm light: p = 0.90; 470 nm light, 1 mW): p = 0.61; 470 nm light, 4 mW: p = 0.83; 470 nm light, 10 mW: p = 0.26; all vs. 0% change, one-sample t-tests; treatment: p = 0.63, one-way ANOVA). g, Re-analysis of data presented in Figs. 3e and 3f using only lever presses initiated during the last 1.5 s of optogenetic light presentation, when striatal DA levels are lowest (n = 18 sessions from 6 mice, 3 sessions each; Peak velocity: p = 0.86; Amplitude: p = 0.72; two-sample t-tests). Note higher variance due to fewer qualifying presses per session. Population means ± SEM are shown in black. ns: not significant. h, Same as g using mice instead of behavioral sessions as the number of observations (N = 6; peak velocity: p = 0.82; amplitude: p = 0.74; paired t-tests). For each mouse, we averaged the median peak velocity or amplitude across 3 behavioral session. Gray lines are individual mice. ns: not significant. i, Same as g for data presented in Figs. 3k and 3l using only lever presses produced while optogenetic stimulation is on (n = 18 sessions; Peak velocity: p = 0.51; Amplitude: p = 0.62; two-sample t-tests). j, Same as i using mice instead of behavioral sessions as the number of observations (N = 6; peak velocity: p = 0.47; amplitude: p = 0.56; paired t-tests). k, Graphical description of ‘n + 1’ analysis to test if optogenetic stimulation of mDANs (470 nm, 4 mW) on trial n causes a delayed change in motor vigor on the subsequent (n + 1) trial. l, Left, scatter plot of median peak press velocity for trials with mDAN stimulation (trial n) vs. the following trial (n + 1) within the same behavioral session. Right, same data expressed as a change in peak press velocity for each session (n = 18 sessions in 6 Dat-Cre::Ai32 mice; p = 0.0805 vs. 0% change, one-sample t-test). Population mean ± SEM shown in black. ns: not significant. m, Same as l for press amplitude (p = 0.8345 vs. 0% change, one-sample t-test). n, Top, experimental setup: mDAN stimulation (470 nm, 1 mW, 1 s, unilateral) is triggered when mice nose poke in the active port. Bottom, cumulative entries in active and inactive ports during a single behavioral session for one task-naïve mouse expressing ChR2 in mDANs, confirming our ability to drive DA release and promote DA-dependent reinforcement. o, Total number of entries in active vs. inactive ports (N = 6 mice; p = 0.01, paired t-test). Population means ± SEM shown in black, individual mice in gray. p, Top, Representative lever trajectories from one example session (thick line: session-wide mean) for presses occurring before (left) and after (right) an uncued increase in reward threshold (green dot). Bottom, peak velocity and amplitude of 30 presses performed before (gray) and after (magenta) the uncued increase in the press threshold required to earn rewards (velocity: p = 0.0020; amplitude: 4.4e⁻⁵, two-sample t-tests). Black bar: session median.

Source data