mPFC lesion does not impair initial active place avoidance learning, but impairs cognitive flexibility in the conflict task variant

(A) Workflow to assess the impact of mPFC or sham lesions on spatial cognitive control. (B) Assessment and impact of the mPFC lesion in three representative subjects (light grey, dark grey, and purple). PrL (prelimbic), IL (infralimbic), Cg (cingulate cortex), M2 (secondary motor cortex). Representative mPFC from a sham and a lesion rat (bottom). (C) Behavioral tracks from two example rats across active place avoidance training. The shock zone is indicated as a 60° sector, gray (shock off) and red (shock on), and arena rotation by the curved arrow. Day 1 - Pretraining: free exploration with shock off; Days 2&3 - Initial Training: Eight daily trials to avoid the shock zone. Day 4 - Retention: One trial with shock on. Days 4&5 - Conflict Training: Eight daily trials to avoid the shock zone relocated 180° from the initial location. Sham and mPFC lesion rats did not differ in (D) locomotor activity, (E) avoidance memory, or (F) place learning. Locomotor activity: During pretraining, there is no effect of lesion on the total distance walked across the two trials (F1,16 = 0.04, p = 0. 85, η2 = 0.002). There is an effect of trial (F1,16 = 15.31, p = 0.001, η2 = 0.09) but there is no effect of the interaction (F1,16 = 0.14, p =0.71, η2 = 10-3). Locomotion does not differ between the groups, across trials, or their interaction during Initial Training (df = 3.17/50.77), or Conflict Training (df = 4.97/79.58; F’S ≤ 1.96, p ≥ 0.09, η2 ≤ 0.03). Avoidance Memory: (Time to first enter the shock zone) The latency to first enter the new shock zone increases for both groups in Initial training. There is no effect of group (F1,16 = 0.1, p = 0.76, η2 = 0.001) but clear effects of day (F1,16 = 69.61, p = 10-7, η2 = 0.17) and trial (F5.33, 85.33 = 12.15, p = 10-9, η2 = 0.12). There are also no significant interactions (group x day: F1,16 = 0.94; p = 0.35, η2 = 10-3; group x trial: F5.33, 85.33 = 1.83, p = 0.11, η2 = 0.02; day x trial: F4.78, 76.43 = 1.93, p = 0.10, η2 = 0.03); group x day x trial: F4.78, 76.43 = 0.26, p = 0.93, η2 = 10-3). Place Learning (Number of entrances): The effect of group is not significant (F1,16 = 0.002, p = 0.96, η2 = 10-5) but the effects of day (F1,16 = 26.34, p = 10-3, η2 = 0.19) and trials (F3.22, 51.50 = 36.76, p = 10-8, η2 = 0.1) are significant. There are no significant interactions (group x day: F1,16 = 0.003, p = 0.96, η2 = 10-5; group x trial: F3.22, 11.50= 0.74, p = 0.54, η2= 10-3; group x day x trial: F4.05, 64.81 = 0.35, p = 0.74, η2 = 10-3). But the interaction between day and trial is significant (day x trial: F2.26, 36.08= 13.75, p = 10-5, η2 = 0.11). The two groups are indistinguishable on Day 4 during which 24-h memory is assessed by their times to first enter the shock zone (t17 = 1.01, p = 0.3, d = 0.49) and the number of entrances (t17= 0.95, p = 0.4, d = 0.002). There is a non-significant trend for greater savings in sham rats, comparing the first and second days of conflict trials (sham = 11.12 ± 2.89 entrances, lesion = 3.9 ± 4.50 entrances; t17 = 1.34; p = 0.20, d = 0.56). In the conflict learning, there is no effect of group (F1,16 = 0.21, p = 0.65, η2 = 10-3), but the effects of day (F1,16 = 8.99, p = 0.01, η2 = 0.05), and trials (F1.89, 30.12 = 18.55, p = 10-9, η2 = 0.28) are significant. The group interactions are not significant (group x day: F1,16 = 0.55, p = 0.47, η2 = 10-3; group x trial: F1.88, 30.12 = 1.38, p = 0.27, η2= 0.02; group x day x trial: F2.30, 36.87 = 1.16, p = 0.33, η2 = 10-3). Also, the interaction between day and trial is not significant (day x trial: F2.30, 36.87 = 2.31, p = 0.11 η2= 0.019). The insets compare number of entrances during the first 5 minutes of left: the first (D1T1) and second days of initial training (D2T9) as well as right: the first (D4C1) and second (D5C9) days of conflict trials. The percent difference in number of entrances between D5C9 and D4C1 of conflict training was computed as a savings index for statistical comparison. Sham rats demonstrate greater savings than lesion rats (t16 = 2.4, one-tailed *p = 0.03, d = 0.85), demonstrating effectiveness of the mPFC lesion to impair cognitive flexibility.

Cytochrome oxidase analysis demonstrates widespread metabolic consequences of mPFC lesion.

(A) Representative cytochrome oxidase staining and locations of the optical density readings. Cytochrome oxidase activity was measured in the dysgranular and granular retrosplenial cortices (RSD and RSG, respectively), the nucleus reuniens (RE), the central nucleus of the amygdala (CEA), basomedial and basolateral amygdala (BMA, and BLA, respectively), the dorsal hippocampus CA1, CA2, CA3 and dentate gyrus areas (dCA1, dCA2, dCA3, dDG, respectively) the ventral hippocampus CA1, CA3 and dentate gyrus areas (vCA1, vCA3, vDG, respectively) and the dorsal subiculum (DS) marked as colored circle areas. (B) Interarea covariations of CO activity by graph theory analysis. Each line indicates a significant correlation (p < 0.05) between the two brain regions (“nodes”) it connects; only the red lines survived False Discovery Rate (FDR < 0.01) correction. Sham: n = 8; Lesion: n = 8.

mPFC lesion does not change basic discharge properties, but decreases hippocampus place cell overdispersion only in the absence of the cognitive control challenge.

(A) (left) Representative histology with overlaid recording traces from the Neuropixel probe. (right) Recording schedule workflow during the cognitive control task. (B) There is no difference between sham and mPFC lesion rats in firing rate (Sham: 3.29 ± 0.27, Lesion: 3.68 ± 0.33, t184 = 0.91, p = 0.36) and burst ratio in hippocampal neurons (Sham: 1.53 ± 0.25, Lesion: 2.02 ± 0.28, t174 = 1.29, p = 0.19). (C) Distribution of standardized place cell discharge (z scores) computed during every 5-second episode in which the rat passed through place cell firing fields in the data set. Different numbers of passes qualified for evaluation during the pretraining (Sham = 2777, Lesion = 3398), retention (Sham = 4039, Lesion = 3366), and conflict (Sham = 1748, Lesion = 2336) recordings. The variance of the histograms characterizes overdispersion, statistically evaluated by their ratio as an F-test (pretraining: F2776,3397 = 2.19, p = 5.8×10-4).

Sham and mPFC lesion rats do not differ in expressing cognitive control of spatial frame-specific representations of location.

(A) Individual ensemble examples during the Day 4 retention session, and (B) group statistics of spatial frame ensemble preference (SFEP), demonstrating cognitive control in both groups. During retention, SFEP is biased to the arena frame in both the sham (t4 = 4.16, #p = 0.01) and lesion rats (t4 = 6.62, #p = 0.003). *p < 0.05 post-hoc differences. (C) Group spatial probability distribution of SFEP for room frame preference during the retention session with no shock. (D) Summary of average probability of room-preferring SFEP discharge in half the arena near and far from the shock zone. Two-way Group X Location ANOVA: Group: F1,4 = 0.17 p = 0.68, η2 = 10-3; Location: F1,4 = 54.69, p = 0.001, η2 = 0.84; Group X Location (F1,4 = 2.04, p = 0.2, η2 = 10-2).(Sham: n=3, Lesion: n=3).