Striatal Activity and Reward Relativity: Neural Signals Encoding Dynamic Outcome Valuation

Histological verification of microwire implantation sites in the dorsal and ventral striatum. Coronal sections are displayed from anterior to posterior sections, and wires located in dorsal (black circles) and ventral (gray circles) are shown. Dorsal sites were distributed in all quadrants, including dorsomedial striatum and dorsolateral striatum. All ventral striatal sites were in medial segments, including medial core and shell subregions.

A–F, Behavioral response latencies during positive-contrast (A–C) and negative-contrast (D–F) session types. A, Nose poke: there was a generalized decrease in response latencies during the MOB session when compared with the SOB session. B, Lever-press: the lever-press response did not significantly differ between the single- and mixed-outcome sessions or during the smaller vs larger trials in the mixed session. C, Food cup entry: results showed an increase in food cup entry latencies during the mixed session, as well as a significant discrimination effect during this same session block. D, Nose poke: there was a generalized decrease in response latencies during the mixed session when compared with the single-outcome session. E, Lever-press: animals showed steady increases in latency to lever-press over trial blocks for the smaller reward outcome. F, Food cup entry: results showed significantly longer latencies to enter the food cup during smaller than larger reward trials in the mixed-outcome session. Bars represent averages of normalized data, and error bars denote the SEM for each outcome presentation. Gray bars depict the similar outcome across trial blocks, and white bars depict the alternate outcome.

Perievent raster examples of neural activity and incentive relativity. A, Positive-contrast unit: activity from the ventral striatum following the nosepoke response during the sessions, with the top panel showing the response to the larger outcome during the single, mixed, and single sessions. Activity significantly increases to the large (four pellets) outcome in the mixed session when the alternate is the small outcome (one pellet). B, Negative-contrast unit: nosepoke-related unit from the dorsal striatum is significantly active during and several seconds after the nosepoke response. The top panel shows the responses to the medium-sized outcome (two pellets) during the single, mixed, and single sessions. There is significantly reduced activity for the medium outcome in the mixed session when the alternate is the larger outcome (four pellets), signifying a negative contrast effect.

Average neural activity surrounding task-related time points in the positive contrast sessions. Bar graphs represent the average activity change during session exposure, and the insert (top right) provides a depiction of activity changes over time referenced to a task event. In this figure, responses show significantly elevated firing rates in response to larger reward trial in mixed-outcome versus single-outcome condition during positive contrast session types. A, Dorsal striatum nosepoke, 7 responses. B, Dorsal striatum food cup entry, 14 responses. C, Ventral striatum lever-press, 12 responses. D, Ventral striatum food cup entry, 17 responses. Symbols designate significance, as follows: *p < 0.033; &cenveo_unknown_entity_wingdings_0075;p < 0.06 (marginal).

A–D, Excitatory inverse unit activity surrounding task-related time points in positive-contrast (A, B) and negative-contrast (C, D) session types. In this figure, units show an inverse contrast in the mixed-outcome versus single-outcome blocks. Specifically, units during positive contrast show a reduction in firing rates during the mixed-outcome block, while units during negative contrast show an increase in firing rates during the mixed-outcome block. A, Dorsal striatum food cup entry, 10 responses. B, Ventral striatum food cup entry positive contrast, 11 responses. C, Ventral striatum nose poke, 7 responses. D, Ventral striatum food cup entry negative contrast, 8 responses.

Microanalysis of incentive contrast. Degree of change analysis and striatal activity. One step comparison = RRE percentage change during session types with a small vs medium or medium vs large comparison. Two-step comparison: RRE percentage change during session types with big vs small comparison. *p < 0.033. A, Nosepoke: there was a larger contrast effect in firing rates during two-step vs one-step comparisons surrounding the nose poke response. B, Lever-press: activity for the lever-press response did not show sensitivity to reward disparity, but did show similar levels of response to both one-step and two-step levels of change in reward magnitude. C, Food cup entry: there was a larger contrast effect in firing rates during two-step vs one-step comparisons surrounding the food cup entry response. D, Food cup entry repeat units: firing rates of units that were active surrounding food cup entry in positive- and negative-contrast sessions showed a greater contrast effect in positive-contrast session trials. E, Trial-by-trial analysis and negative contrast: activity was measured post-food cup for trials categorized by the preceding outcome. For the negative contrast comparison, we examined the small-large sessions and compared trials for the small reward preceded by the same outcome (same-small) to trials for small-outcome trials preceded by large-outcome trials (other-small). We found significantly reduced activity in the same-other trials supporting a negative contrast effect based on trial type. F, Trial-by-trial analysis and positive contrast: For positive contrast, we examined large-outcome trials in the single session by comparing those large-outcome trials that were preceded by the same outcome (same-large) to large outcome trials preceded by small outcome (other-large). We found an effect of trial type, but it was nonsignificant between the two large-outcome trial types. **p < 0.01 and *p < 0.05 for all comparisons.

Perievent raster examples of relative reward effects that resemble positive induction or variety influences. A, Activity increase during mixed session: neural activity from the dorsal striatal MSN is timelocked to lever-press and significantly increases for both outcomes during the mixed session (small and large outcomes alternating). B, Activity decrease during the mixed session: neural activity from a ventral striatal unit MSN is related to the food cup entry and is significantly reduced in the mixed session (medium and small outcomes) and returns to premixed session levels in the final single session.