Supplementary Figure 8: VPr neurons in trained animals display differential responses to sounds depending upon their behavioral context.

Responses are shown as mean normalized firing rate (mean ± SEM). Vertical gray dashed lines indicate sound onset. The set of 30 different TORCs were presented in three different behavioral contexts: (1) Task-irrelevant TORCs (left column, gray), were presented in series of 150 sound presentations (5 complete sets of TORCs) to assess STRF tuning and were not associated with any reward or punishment (A1 n=48, dorsal PEG n=189, VPr n=237, dlFC n=37 neurons), (2) Neutral TORCs (middle column, purple) preceded Safe and Warning click trains in the CLR-D task (A1 n=57, dorsal PEG n=60, VPr n=266, dlFC n=38 neurons). These neutral TORC stimuli did not provide information about the identity (rate) of the upcoming click train and hence did not cue a correct behavioral response. However, since they were presented with a fixed duration, the TORCs did provide timing information about the imminent time of onset of the upcoming highly relevant Safe and Warning click trains, (3) Safe TORCs (right column, blue) were presented in the PT-D task and had to be actively discriminated from Warning tones (A1 n=71, dorsal PEG n=199, VPr n=251, dlFC n=138 neurons). In the middle and right columns, responses recorded in the passive and active (behavior) conditions are shown as dashed and continuous lines, respectively. There is an onset response to TORCs in all three auditory areas in the task-irrelevant context (left column). The rest of the TORC response is different in the three cortical areas –the onset response is followed by a sustained response in A1, a rapid return to spontaneous baseline in dorsal PEG and distinctively, a deep inhibitory response in VPr. This inhibitory response to TORCs in the behaviorally irrelevant context is characteristic for VPr and is seen also in the passive presentations of the task stimuli. A similar response profile for the TORCs for all cortical areas is observed in the behavioral context of the PT-D task where the TORCs are the Safe stimuli (right column). In the PT-D task, VPr neurons became less suppressed during the presentation of Safe TORCs in the active compared to the passive state (solid blue). However, in the “anticipatory neutral” context of the CLR-D task, there is a striking qualitative difference in responses between cortical areas. In the CLR-D task, while TORCs carry no information about the identity of the upcoming click train, they do convey timing information since these are presented for a fixed duration before the click train (see also Figure 4b). As seen (middle column), in dorsal PEG, in VPr, and also in dlFC, the response to the TORCs in the active CLR-D task is a sustained increase in firing rate and an anticipatory “build-up” type response in VPr and dlFC.