Article Figures & Data
Figures
- Figure 1.
Flexible choice patterns revealing reward-guided repetitive choice and nonreward-guided switch choice. A, Left, Snapshot of a head-fixed rat at the moment of licking choice toward a left spout (orange arrowhead). A black circle and blue arrowhead indicate the position of the tongue and right spout, respectively. Right, Schematic illustration of an outcome-based two-alternative choice task. Each trial began with the presentation of an auditory cue (Start, 3 kHz). Rats awaited another auditory cue (Go, 10 kHz) while abstaining from licking during the initial delay period (1st Delay). Subsequently, they chose either left or right spouts by licking within 5 s. A reward was provided after the second delay period (0.3–0.7 s) if the selected spout position aligned with the current reward position; otherwise, no reward was given, accompanied by a lack of sensory feedback and a 5 s timeout. B, Representative choice pattern revealing repetitive choice behavior postreward acquisition in last trial and switch choice behavior postunrewarded trials. Background colors indicate ongoing reward positions (orange, left spout; blue, right spout). Thick and thin vertical lines denote reversals of reward position: from left to right and from right to left, respectively. Colors of outline and face in symbols represent choice position (left or right) and choice pattern (repetitive or switch), respectively. A line plot displays trial series of the probability of repetitive choice (average probability over three trials). C, Left, Probability of repetitive choice (orange, left choice; blue, right choice) after rewarded (Last rwd) and unrewarded (Last unrwd) outcomes in the last trial. Circles indicate individual sessions (14 sessions, 12 rats). Horizontal lines and error bars represent mean and SD, respectively. Statistical significance was confirmed by one-way ANOVA followed by post hoc Tukey–Kramer test (*p < 0.001). Right, Averaged choice patterns around the reversal of reward position (orange, from left to right; blue, from right to left). Values present mean and SD. D, Top, Schematic illustration of recording sites with two probes in M2 and DS of left hemisphere. Recording sites (white arrowhead) in M2 and DS in the Nissl-stained coronal brain sections. Bottom, Representative local field potentials simultaneously recorded from M2 (black) and DS (red). E, Post hoc confirmation of injection site of retrograde tracer, FG (red arrowhead), in DS and the corticostriatal projection neurons labeled with FG in M2, including the recording site (white arrowhead). cc, corpus callosum. B (or Bregma) indicates the AP coordinate based on the rat brain atlas (Paxinos and Watson, 2009).
- Figure 2.
Trial-to-trial changes in choice-position–selective TCA component of M2-DS ensemble activity are related to behavioral variables. A, A schematic illustration of TCA. B–D, Example of a TCA component. B, Neuron factor, including M2 (black) and DS (red) neurons. C, Temporal factor (within trial activity). D, Trial factor of the TCA component. Symbol colors denote the choice position at each trial (left, orange; right, blue). Background colors represent the ongoing reward position (orange, left spout; blue, right spout). E, Trial series of discrimination degree of the trial factor shown in D using the original trial order (black) and shuffled trial order (gray, mean and SD of discrimination degree acquired by repeating shuffling trial order). F1, Trial series of reaction times (RTs) and (F2) trial series of the number of licks after the choice response in the same session as shown in B–D. “r” denotes Pearson's correlation coefficient. G1, Left, Trial-series of RTs over 14 sessions. Gray and black lines represent individual and averaged values. Error range indicates SD. Right, The distribution of correlation coefficients between discrimination degree and RTs. G2, Left, Trial series of the number of licks after the choice response over 14 sessions. Right, The distribution of correlation coefficients between discrimination degree and the number of licks. Open and filled bars indicate the number of nonsignificant and significant correlation coefficients, respectively (Pearson's correlation; p < 0.01). H, Comparison of absolute correlation coefficients showing statistical significance (as shown in G1 and G2) between two behavioral variables. Statistical significance was assessed using a two-sample t test. Horizontal and vertical lines represent mean and SD, respectively. I, The distribution of regional contribution indices for TCA components which correlated with RTs (I1) and with the number of licks (I2). The vertical dashed line indicates mean value. Statistical significance was assessed using t test. J, The mean trial series of ITIs (blue) with the mean trial series of the number of licks (black) as shown in G2. K, Correlation coefficients between behavioral variables for all 14 recording sessions. Horizontal and vertical lines represent mean and SD, respectively.
- Figure 3.
Trial order activity pattern of M2-DS ensembles related to repetitive and switch choices. A, Neuron factor, (B) temporal factor, and (C) trial factor of a representative TCA component revealing large variance in switch choice trials. Edge and face colors represent choice positions (orange, left; blue, right) and choice patterns (black, repetitive choice; magenta, switch choice), respectively. Background colors denote ongoing reward positions (orange, left spout; blue, right spout). A gray box presents an area enlarged in D. D, Representatives of switch choice trials (magenta arrows) exhibit significantly larger variance when the ongoing reward position is at the right spout. E, Top, A schematic illustration of choice patterns utilized for the statistical estimation of differences in trial factors among repetitive and switch choices. This trial sequence illustrates a switch from left to right choices after the reversal of reward position (from the left spout to right the spout). At trial t, the animal shifts its choice to the right spout, transitioning from the left choice selected in the preceding trial (t − 1). Subsequent repetitive choices (t + 1, t + 2, and t + 3) are employed to compare differences among switch (Switch) and repetitive (Rep-1, Rep-2, and Rep-3) choice patterns. Bottom, Z-scores of trial factors in Switch, Rep-1, Rep-2, and Rep-3 choice patterns within the same dataset as depicted in C and D, corresponding to the left (ipsilateral) and right (contralateral) choices of the rat. The statistical significance of differences is assessed through one-way ANOVA followed by post hoc Dunnett test (*p < 0.05). Horizontal and vertical lines represent mean and SD, respectively. F, The Venn diagram illustrates the number of TCA components with significantly different Z-scores of trial factors between repetitive choice and switch choice trials in left choice (orange), right choice (blue), and both (merge) conditions. The bar graph indicates the number of sessions featuring TCA components with significantly different Z-scores of trial factors between repetitive and switch choice trials. G, Population data of TCA components reveal a significant difference in Z-scores between repetitive and switch choice trials. Individual dots indicate the mean of absolute Z-score per TCA component. Bar graphs show the mean and SD. Statistical significance of the difference is assessed through one-way ANOVA followed by post hoc Dunnett test (*p < 0.05). H, The distribution of regional contribution indices for TCA components which were significantly different between switch and repetitive choices in left choice block (orange) and right choice block (blue). The vertical dashed line indicates mean value. Statistical significance was assessed using t test.
- Figure 4.
Trial factor of a TCA component distinguishing between rewarded and unrewarded choice trials. A, Neuron factor, (B) temporal factor, and (C) the trial factor of the TCA component, revealing substantial variance during unrewarded choice trials. In the trial factor, different colors represent choice patterns, with edge and face colors denoting outcomes (rewarded, gray; unrewarded, black) and choice positions (orange, left; blue, right), respectively. D, Top, A schematic illustration of outcome conditions used for statistical estimation of differences in trial factors of repetitive choice trial between rewarded and unrewarded outcomes. A vertical dashed line indicates the reversal of reward position from the right spout to the left spout. The trial sequence illustrates a series of repetitive choice trials after a switch choice trial with a rewarded outcome (Rep-1 & rwd, Rep-2 & rwd, and Rep-3 & rwd), as well as repetitive choice trials without reward outcomes after the reversal of the reward position (Rep & unrwd). Bottom, Z-scores of trial factors computed in Rep-1 & rwd, Rep-2 & rwd, and Rep-3 & rwd and Rep & unrwd are shown in the session depicted in panel C when the animal made left and right choices. The statistical significance of the difference is assessed through one-way ANOVA followed by post hoc Dunnett test (*p < 0.05). E, The Venn diagram displays the number of components revealing significantly different trial factors between rewarded and unrewarded repetitive choice trials at left choice (orange), right choice (blue), and both (merge). The bar graph indicates the number of sessions where components showed significant differences between rewarded and unrewarded repetitive choice trials. F, Population data of TCA components revealing a significant difference in Z-scores between rewarded and unrewarded repetitive choice trials. Individual dots represent the mean of absolute Z-score per TCA component, with bar graphs depicting the mean and SD. Statistical significance was assessed through one-way ANOVA followed by post hoc Dunnett test (*p < 0.05). G, The distribution of regional contribution indices for TCA components which were significantly different between rewarded and unrewarded repetitive choices in left choice block (orange) and right choice block (blue). The vertical dashed line indicates mean value. Statistical significance was assessed using t test.
- Figure 5.
Differential within-trial activity of M2-DS ensembles between choice-pattern–selective and outcome-selective TCA components. A, Collective normalized temporal factors (within-trial activity) of TCA components with a significant difference in Z-scores of trial factor between repetitive and switch choice trials at left (ipsilateral) and right (contralateral) choice trials. B, Collective normalized temporal factors of TCA components exhibit a significant difference in Z-scores of trial factors between rewarded and unrewarded repetitive choice trials at left (ipsilateral) and right (contralateral) choice trials. C, Averaged normalized temporal factors related to choice patterns (magenta) and outcomes (cyan). D, Cumulative summation curve of peak time of temporal factors of choice-pattern–selective TCA components (magenta) and outcome-selective TCA components (cyan). Statistical significance was assessed using the Mann–Whitney U test.
- Figure 6.
Comparison of TCA component based on M2-DS ensembles with those based on M2- or DS-alone ensemble. A–D, Comparison of the total number of choice-selective TCA components (A1) between M2-DS ensembles (green) and M2-alone ensemble (black) and (B1) between M2-DS ensembles and DS-alone ensemble (red) in left choice block. These comparisons in right choice block (C1) between M2-DS ensembles and M2-alone ensemble and (D1) between M2-DS ensemble and DS-alone ensembles. The total number of TCA components in the M2-DS ensemble is presented by the mean and SD. p values were assessed by t test. A2, B2, C2, D2, The t statistic provided by two-sample t test of mean absolute Z-scores of trial factors at switch-selective TCA components between M2-DS ensemble and M2-/DS-alone ensembles. Open and filled symbols indicate statistical nonsignificance and significance, respectively (t test; p < 0.05). Horizontal and vertical solid lines represent mean and SD, respectively. E–H, Comparison of the total number of outcome-selective TCA components (E1, G1) between M2-DS ensembles and M2-alone ensemble and (F1, H1) between M2-DS ensemble and DS-alone ensemble. E2, F2, G2, H2, The t statistic provided by two-sample t test of mean absolute Z-scores of trial factors at outcome-selective TCA components between M2-DS ensemble and M2-/DS-alone ensembles.
Tables
Session ID Number of overtraining sessions Reward acquisition probability Number of M2 neurons Number of DS neurons R982-r1 7 p = 0.828 N = 45 N = 43 R983-r1 8 p = 0.837 N = 35 N = 55 R983-r2 9 p = 0.834 N = 58 N = 48 R985-r1 10 p = 0.853 N = 30 N = 60 R985-r2 11 p = 0.867 N = 66 N = 26 R986-r1 4 p = 0.813 N = 16 N = 20 R991-r1 7 p = 0.806 N = 46 N = 31 R997-r1 0 p = 0.740 N = 24 N = 34 R1000-r1 1 p = 0.784 N = 12 N = 32 R1002-r1 3 p = 0.744 N = 31 N = 11 R1004-r1 1 p = 0.792 N = 20 N = 51 R1005-r1 4 p = 0.794 N = 65 N = 64 R1009-r1 3 p = 0.774 N = 35 N = 33 R1012-r1 6 p = 0.755 N = 40 N = 26
In this issue
