Prefrontal to ventral tegmental area dynamics drive contingency degradation

a. Hypothesis for cue stability cell proportions, from previous work21 compared to neural results before (pre), after (post), and across reversal. b. Cue stability map on two days before reversal, where H → L and H → H cues are high value (dark purple). c. Cue stability map constructed from two post-reversal stable behavior days, where H → H and L → H cues are high value (dark purple). d. Across reversal cue stability map, with cells tracked from pre-reversal (x-axis) where H → L and H → H are high value to post-reversal (y-axis) where H → H and L → H cues are high value. e. Example animals with different spacing between stable pre and post day compared across reversal. f. Statistics on the proportion of stable cue cells for high value (dark purple) and low value (light purple) cues before, after, and across reversal. There is not a significant difference in the proportion of coding within a cue type in any of the conditions (p = 0.1715, data from n = 4 cross-cue comparisons from n = 6 animals). g. Trial-averaged time histogram of stable cue cell activity before (pre) and after (post) contingency reversal, sorted by H → H activity. h. Mean +/-SEM traces for stable cue cells on H → L (green dash), H → H (dark purple solid), L → H (light blue dash) and L → L (light purple solid). Error bars are shaded in accordance with high (dark purple) or low (light purple) value. i. Quantification of average stable cue cell activity during cue and trace interval (0–2.5 s) split by cue (Cue x time p = 0.0032, cells from n = 6 animals). There is not a significant difference within coding of 85% or 15% cues before or after reversal. j. Session permutation control results, which reflect the correlates of modeled mRPE signals from reversal days evaluated with GLM on neurons from reversal and other days. If CD neurons displayed a similar activity profile to the reversal days on pre- or post- days, there would not be a significant increase in the proportion of isolated cells on the reversal day (CD: p = 0.0071, CE: p = 0.1923, n = 20 animals). k. Reversal cell stability quantification. There is not a significant difference (p = 0.6970 n = 6 animals) in the relationship with the CD signal for tracked cells on the identification day (2) vs the previous day (1). l. There is not a significant difference between the proportion of excited/inhibited cells in imaging in Fig. 2g vs Fig. 4h (p = 0.1287, n = 4 PFC-VTA, n = 15 PFC animals). See Supp. Table 1 for more statistical information, including more post-hoc comparisons, sidedness, and corrections for multiple comparisons. See Extended Data Fig. 3 for placements.