Fig. 4: D1R and D2R oppositely modulated neuronal reward expectancy signals during the cue delay period.

A Example reward expectancy neuron recorded during control conditions (left) and after stimulating D1Rs with SKF81297 (right) with higher activity for cues indicating large reward in the cue delay period (gray shaded area, cue delay period). B Average normalized activity of all reward expectancy neurons in the cue delay period (gray shaded area) recorded with SKF81297. Activity was pooled over reward cue sets, and the preferred reward size was defined as the reward size condition yielding higher average activity. Error bands represent +/- SEM. C Reward discriminability (AUROC) in the cue delay period compared between control and D1R stimulation. Left: Each point corresponds to one neuron. Right: Average reward discriminability (AUROC) for control and D1R stimulation (31 neurons; p = 0.012). Data are presented as mean values +/- SEM. D Example reward expectancy neuron recorded during control conditions (left) and after stimulating D2Rs with quinpirole (right) with higher activity for cues indicating small reward. E Same conventions as in B for all reward expectancy neurons recorded with quinpirole. Error bands represent +/- SEM. F Reward discriminability (AUROC) in the cue delay period compared between control and D2R stimulation. Left: Each point corresponds to one neuron. Right: Average reward discriminability (AUROC) for control and D2R stimulation (43 neurons; p = 0.0016). Data are presented as mean values +/- SEM. ^*p < 0.01, ^*p < 0.05, signed rank test.