Fig. 7: Dips in dopamine release in the anterior dorsal striatum are present for unrewarded lick onsets before learning and for cues after extinction.

a Schematic showing the locations of fibers (n = 29 fibers in 4 mice) in the aDS for measurements of DA release with dLight1.3b. b Light (top) and tone (bottom) cue onset triggered averages of ∆F/F for a single fiber (location in the coronal plane in inset) for trials where cues were associated with 100% (post learning phase), 90%, and 10% (extinction) probabilities. Shaded region, S.E.M. c Light cue-aligned ∆F/F (top) and licking (bottom) for all trials for each phase included in the triggered averages in (b top). d Box plots showing the median DA ΔF/F minima (dip, see “Methods”) for light (top) and tone (bottom) cue trials for trials post learning and the 10 and 90% reward probability phases during partial extinction. Each dot is a single fiber (n = 29 fibers in 4 mice). Box plots with whiskers extending from the minimum to the maximum values. The center line represents the median, the box bounds show the 25^th and 75^th percentiles. Data were analyzed using a linear mixed-effects model with mouse as a random effect (light cue: F_(2,84) = 27.21, p < 0.0001; tone cue: F_(2.82) = 17.27, p < 0.0001). Bonferroni post hoc comparisons: for light, **** p < 0.0001 post learning vs extinction 10%, n.s, p = 0.43 post learning vs extinction 90%, ****p < 0.0001 extinction 10% vs 90%, for tone, ***p = 0.00024 post learning vs extinction 10%, n.s, p = 1.00 post learning vs extinction 90%, **** p = 0.00001 extinction 10% vs 90%. n.s.: not significant. e Top: mean light cue onset triggered DA (n = 29 fibers across 4 mice) and ACh (n = 68 aDS fibers across 8 mice) ΔF/F for cue onset during extinction (10% and 20% for DA and ACh, respectively). Shaded region, S.E.M. Bottom: Box plot of latencies to the minimum DA (DA dip) or maximum (ACh late peak) for all fibers in DA or ACh sensor-expressing mice, respectively, in the extinction phase. Each dot is one fiber. Box plots show the median (center line), the 25^th and 75^th percentiles (bounds of the box), and the minimum and maximum values (whiskers). Two-tailed Wilcoxon rank-sum test, W_(5,53) = 64, *p = 0.0162 (two-tailed). f Histogram showing the latencies across fibers of the emergence of significant DA dips to light (top) and tone (bottom) cues relative to the decrease in lick index following high to low reward probability transitions. Blue vertical lines indicate the medians of the distributions. g Box and whisker plots showing the peak decelerations and velocity changes across all mice (n = 4 mice) following light (left) and tone (right) cue onsets. Each datapoint is the mean for one mouse. Box plots indicate the median (center line), the 25^th and 75^th percentiles (bounds of the box), and the minimum and maximum values (wishers), n.s., not significant, Friedman test (two-tailed) with Dunn’s multiple comparisons, p > 0.05. The exact p-values and additional statistical details can be found in Supplementary Table 1. h Left: Mean ∆F/F for a fiber (location in inset) aligned to the onset of spontaneous, unrewarded spout licking bouts in the ITI on trials pre (blue) and post (orange) initial Pavlovian learning. Shaded region, S.E.M. Right: Lick bout-aligned ∆F/F for all bout onsets included in triggered average on left. i Mean lick bout onset-triggered ∆F/F across all mice and fibers (n = 29 fibers across 4 mice) for bouts pre (blue) and post (orange) learning. Shaded region, S.E.M. j Box plots as in (d) showing the median DA ΔF/F minima for spontaneous lick bout onsets pre- and post-learning. The box plots indicate the median (center line), 25^th and 75^th percentiles (bounds of the box), and the minimum and maximum values (whiskers). Each dot is one fiber (27 fibers for pre learning, 24 fibers for post learning, across 4 mice). Data were analyzed using a linear mixed-effects model with mouse as a random effect (F_(1.49) = 12.73, ***p = 0.00082). k Left: Mean lick bout onset-triggered DA (n = 29 fibers across 4 mice) and ACh (n = 68 aDS fibers across 8 mice) ∆F/F in the aDS (see “Methods”) for bouts pre learning. Shaded region, S.E.M. Right: box plot of latencies to the minimum (DA dip) or maximum (ACh peak) for all fibers in DA or ACh sensor-expressing mice respectively in the pre learning phase. The box plot indicates the median (center line), 25^th and 75^th percentiles (bounds of the box), and the minimum and maximum values (whiskers). Each dot is one fiber. W_(27,68) = 734, ***p < 0.001, Two-tailed Wilcoxon rank-sum test. Brain schematics in (a), (b) and (h) were adapted from the Allen Mouse Brain Common Coordinate Framework (CCFv3) (https://atlas.brain-map.org/. Mouse head schematic in (h), (i) and (k) adapted from SciDraw (scidraw.io), licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Source https://doi.org/10.5281/zenodo.3925903). Source data are provided as a Source data file.