The mPFC→NAc and mPFC→BLA neurons demonstrated distinct activity patterns across various emotional states.

(A) Schematic showing retrograde AAV-GCaMP6 injection in the NAc and BLA, respectively. In vivo calcium imaging was performed in the mPFC neurons expressing GCaMP6 via miniscope. (B) Top: An example of AAVretro-GCaMP6 injection and expression in the BLA and NAc, respectively. Bottom: An example of GCaMP6m expression in the mPFC with AAVretro-GCaMP6 injection in the NAc and an image of Ca2+ fluorescence recorded with miniaturized microscope. The 1-mm GRIN lens covered the full depth of the prelimbic cortex. Green, GCaMP6m. Scale bars, 200 mm. (C) Left: the field of view under a GRIN lens in one mouse with identified neurons numbered and colored. Right: fluorescence traces of example neurons marked in the above panel. (D). Schematic of the open field test (OFT) in an open 50 cm X 50 cm arena. The pink areas represent corners and the blue area represents the center. (E) Illustrations depicting typical mouse behaviors associated with different emotional states observed during the OFT: Center (exploration and higher anxiety), Corner (lower anxiety), sniffing (sensory exploration and vigilance, possibly indicating curiosity, cautious exploration, or heightened alertness), and grooming (stress relief or comfort). (F) The percentage of time that mice spent in the center, corners, sniffing and grooming during 10-min open field test. n = 15 mice for mPFC→BLA group; n = 11 mice for mPFC→NAc group. The time-allocation pattern across behaviors is similar for Amy and NAc cohorts. A two-way mixed-effects model (Region × Behavior) showed a strong main effect of Behavior (Corner » Center; Sniffing < Center), but no main effect of Region and no Region × Behavior interaction (all P > 0.1). (G) Heatmap and time-series plot of neuronal activity (ΔF/F) across the test duration. The heatmap above shows the fluctuation in activity level across neurons, while the plot below aligns the averaged fluctuations of these neurons with observed behaviors (Center, Corner, Sniffing, Grooming) within the arena. (H) Normalized transient rates of neuronal activity during different behaviors. The mixed-design ANOVA revealed no main effect of region (P = 0.14) and no significant Region × Behavior interaction (P > 0.3), indicating that mPFC→BLA and mPFC→NAc neurons exhibited similar behavioral modulation patterns. (I) PCA plots for mPFC→BLA and mPFC→NAc pathways illustrating the distribution of neuronal activity patterns for different observed behaviors (Center: blue, Corner: red, Sniffing: green, Grooming: purple). Each point represents an individual neuronal recording. Quantitative analysis confirmed that neural population activity during corner behavior was highly similar between mPFC→BLA and mPFC→NAc pathways (Pearson r = 0.64, p < 1 × 10⁻⁹), consistent with comparable encoding of this low-anxiety state. (J) Summary of the distances of neuronal activity clusters from the center of Corner behavior in the OFT for both pathways. distances were computed using PC1 and PC2 coordinates. ***P < 0.001, Mann-Whitney U test.

Distinct encoding of emotional status by center-ON neurons in the mPFC pathways

(A) Ca2+ traces of the averaged activity of center-ON neuron ensembles around the onset of center entry (5 s before to 5 s after). The total number of 24 center-ON neurons (n = 24) recorded in a mouse of the mPFC→BLA groups that used to calculate the representative averaged trace. Solid lines represent the averaged value. (B) Representative heatmaps depicting the activity patterns of mPFC→BLA neurons across four different behavioral contexts: Center, Corner, Sniffing, and Grooming. Each column corresponds to a different behavior, with the intensity of color indicating the level of neuronal activity. (C) Left: spatial distributions of center-ON neurons among the mPFC→BLA and mPFC→NAc neurons in one representative mouse from each group. Right: quantification of the percentage of center-ON neurons showed that the BLA pathway contained a significantly higher proportion of center-ON neurons compared to the NAc pathway. * P < 0.05, Mann-Whitney U test (Mann–Whitney U = 47, P = 0.049). (D) Spatial heatmaps of neuronal activity across the open field arena of all observed center-ON neurons in one example mouse of mPFC→BLA and mPFC→NAc groups. The color bar indicates the averaged normalized z-score. (E) Normalized transient rate of neuronal activity across behavioral states for mPFC→BLA and mPFC→NAc neurons. A two-way mixed-design ANOVA revealed a significant Region × Behavior interaction (P = 0.048), indicating that mPFC→NAc neurons exhibited higher activity during sniffing compared to mPFC→BLA neurons, while no regional differences were found in center, corner, or grooming states. * P < 0.05; two-way mixed-design ANOVA followed by Tukey’s post hoc test. (F) PCA plots illustrating the distribution of neuronal activity patterns during different behaviors for mPFC→BLA and mPFC→NAc pathways. Points are color-coded by behavior type (Center: blue, Corner: red, Sniffing: green, Grooming: purple). (G) Summary of the distances from the center of Corner behavior in terms of neuronal activity for each behavior in both mPFC pathways. ** P < 0.01, *** P < 0.001, Mann-Whitney U test.

The mPFC→NAc and mPFC→BLA neurons demonstrated distinct activity patterns across various emotional states during EPM test.

(A) Schematic illustration of the Elevated Plus Maze (EPM) test setup showing the layout of open (Open-P, Open-NP) and closed (Close-P, Close-NP) arms used to assess anxiety-related behaviors in mice. (B) Box plots displaying the percentage of time spent by mice in the different arms of the EPM (Open-P, Open-NP, Close-P, Close-NP) for both mPFC→BLA and mPFC→NAc pathways. Data indicate variations in time spent across different arms, highlighting behavioral preferences. (C) Averaged transient rates (Hz) of neuronal activity of all recorded neurons in the mPFC→BLA and mPFC→NAc pathways in each arm of the EPM. Data are represented as mean ± SEM. Mixed-effects analysis (Region × Arm) showed higher transient rates in open arms relative to closed arms (Open-P vs Close-NP: β=0.0149, P=7.3×10⁻⁶; Open-NP vs Close-NP: β=0.0107, P=0.0019) with no overall Region effect (p=0.304) and no Region × Arm interaction (all P>0.29), indicating the activity pattern of the mPFC→BLA and mPFC→NAc pathways in each arm of the EPM is similar. (D) Analysis design of observing the neural activities of center-ON ensembles of OFT in different EPM arms. (E) Mean Ca2+ signal across the EPM arms of all observed center-ON neurons in one example mouse. The color bar indicates the averaged normalized z-score. (F) Example 30 center-ON neurons (top) and corresponding averaged activity (bottom) around the onset of arm entry during the EPM test. Time 0 represents the onset of arm entry. (G) Group data showing the mean transient rate of center-ON neurons across four EPM arms for mPFC→BLA and mPFC→NAc pathways. A two-way mixed-effects ANOVA with Region (BLA vs NAc) as a between-subject factor and Condition (Opent-P, Open-NP, Close-P, Close-NP) as a within-subject factor revealed a significant Region × Condition interaction (F(3,72) = 3.21, P = 0.030).Tukey’s post hoc tests showed that BLA exhibited significantly higher transient rates than NAc during the Open-NP condition (P = 0.018), while no significant regional differences were found in other three arms (P > 0.1). Data are represented as mean ± SEM. n = 9–13 mice for mPFC→BLA; n = 8-11 mice for mPFC→NAc.

Pattern decorrelation is shown by the mPFC→NAc neurons, but not mPFC→BLA neurons.

(A). Top: the apparatus for the social interaction test. The subject mouse is in the middle 45 cm-long chamber; the 10 cm end compartments contain different stimuli. Bottom: the three sessions of the social interaction test. In the first 10-minute test session (S1), a strange mouse (M1) and an object (O) were placed in the end chambers; session 2 (S2) used the same stimuli but swapped their positions; in session 3 (S3) a new mouse (M2) replaced O, so that the subject mouse must choose whether to interact with a familiar or strange mouse (M1 vs. M2). (B). The percentage of time that mice (n = 10) spent interacting with stimuli in each session. Data are shown as mean ± SEM. Paired t-test, S1: t(9) = 12.71, P = 1.2×10⁻⁶, Cohen’s d = 4.02; S2: t(9) = 0.88, P = 0.40, Cohen’s d = 0.28; S3: t(9) = 7.29, P = 6.7×10⁻5, Cohen’s d = 2.31. Mice spent significantly more time with the social mouse (M1) than with the object (O) in S1 or the new social mouse (M2) than with the old social mouse (M1) in S1. (C). The averaged calcium event rate of mPFC→BLA group when mice were engaged with different stimuli over the three sessions. n = 13 mice. Data are represented as mean ± SEM. Paired two tailed t-test, S1: t(12) = –0.77, P = 0.46, Cohen’s d = –0.22; S2: t(12) = 033., P = 0.74, Cohen’s d = 0.09; S3: t(12) = –0.39, P = 0.070, Cohen’s d = – 0.11. The transient rate of the mPFC→BLA neurons showed no significant difference during the interactions with the two stimuli in each session. (D). The averaged calcium event rate of mPFC→NAc group when mice were engaged with different stimuli over the three sessions. n = 11 mice. Data are represented as mean ± SEM. Paired two tailed t-test, S1: t(10) = –0.88, P = 0.40, Cohen’s d = –0.27; S2: t(10) = 178, P = 0.11, Cohen’s d = 0.56; S3: t(10) = –1.45, P = 0.18, Cohen’s d = 0.46. The transient rate of the mPFC→NAc neurons showed no significant difference during the interactions with the two stimuli in each session. (E-F). Top: raster plots of correlation coefficient of paired mPFC neurons during interactions with different stimuli, from a representative mouse in mPFC→BLA and mPFC→NAc group, respectively. Bottom: distribution of pair-wise Pearson correlation coefficients among these recorded mPFC neurons in responding to stimuli in each session. The blue and orange plots represent M1 interactions, while the lighter colors represent the interaction with the other stimulus (O). (G). The averaged full width at half maximum (FWHM) of correlation coefficient distribution of individual mice in mPFC→BLA (n = 8) group during interactions with different stimuli Data are shown as mean ± SEM. Paired two tailed t-test, S1: t(7) = –1.63, P = 0.15, Cohen’s d = –0.58; S2: t(7) = –0.41, P = 0.69, Cohen’s d = –0.14; S3: t(7) = –1.63, P = 0. 15, Cohen’s d = 0.57. The correlation coefficient of the mPFC→BLA neurons showed no significant difference between two stimuli. (H). The averaged full width at half maximum (FWHM) of correlation coefficient distribution of individual mice in mPFC→NAc (n = 8) group during interactions with different stimuli Data are shown as mean ± SEM. *P<0.05, Paired two tailed t-test, S1: t(7) = –2.75, P = 0.025, Cohen’s d = –0.92; S2: t(7) = –0.92, P = 0.39, Cohen’s d = – 0.33; S3: t(7) = –2.87, P = 0.021, Cohen’s d = –0.96. The correlation coefficient of the mPFC→NAc neurons showed significantly stronger separation between the two stimuli during S1 and S3 than during S2.

Comparative analysis of all neurons and center-ON subsets reflects divergent encoding patterns in the mPFC→BLA and mPFC→NAc Pathways.

(A) PCA plots showing the activity of all recorded neurons in the mPFC→BLA and mPFC→NAc pathways during interactions with a social stimulus (mouse, M1) and a nonsocial stimulus (object, O) in a modified Three Chamber Test (mTC). The plots reveal distinct clustering of neural activity patterns for each stimulus within both pathways. (B) Distances between all neuronal activity clusters for nonsocial (O) interactions from the center of that for M1 interaction, compared across mPFC→BLA and mPFC→NAc pathways. Significant difference is demonstrated in the mPFC→BLA pathway. (C) PCA plots for center-ON neurons identified during the Open Field Test (OFT). These plots illustrate the activity of these neurons in the mPFC→BLA and mPFC→NAc pathways during the same social and nonsocial stimuli. Clusters show how Center-ON neurons specifically respond to each type of stimulus. (D) Distances of all center-ON neuronal activity clusters from the center of that for M1 interaction, compared across mPFC→BLA and mPFC→NAc pathways. Significant difference is demonstrated in the mPFC→NAc pathway. Data are represented as mean ± SEM. ***P < 0.001, Mann-Whitney U test.

The modifications in social ranking of mice alter their anxiety and social states.

(A) Schematic of the experimental timeline. Singly housed mice were handled for 3 days prior to undergoing a 5-day tube test to establish dominant (Winner) and subordinate (Loser) social status. Naïve singly housed mice served as controls. Behavioral assays were conducted 24 hours after, and brain tissue collection within 1 hour of, the final tube test session. (B) Corticosterone concentrations measured from brain lysates of Control, Winner, and Loser mice. Loser mice showed significantly elevated corticosterone levels compared to both Control and Winner groups (** P < 0.01, # P < 0.05). Data are presented as mean ± SEM. (C-D) The mice’s total distance traveled in the open field and time spent in the center zones before and after the tube test. *P < 0.05; ns, non-significant. n = 6 mice for each group, winner vs loser, unpaired Student’s t-test. (E) The differences of total distance (top) and center time (below) between winner and loser groups before and after the tube tests. **P < 0.01; ns, non-significant. Before vs. after tube test, Two-way ANOVA Multiple comparisons, sidak’s post-hoc test. (F) The mice’s exploration time near to M1 and O/M2 during the social ability test (left) and social memory test (right) conducted before the tube test. Social ability index calculated as (time near M1 chamber – time near object chamber) / (time near M1 chamber + time near object chamber). Social memory index calculated as (time near M2 chamber – time near M1 chamber) / (time near M2 chamber + time near M1 chamber). *P < 0.05, ***P < 0.001, ****P < 0.0001, ns, non-significant. Two-way ANOVA Multiple comparisons, sidak’s post-hoc test. Unpaired Student’s t-test for preference index. (G) The mice’s exploration time spent near the M1 and O/M2 during social ability test and social memory test after the tube test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Two-way ANOVA Multiple comparisons, sidak’s post-hoc test. Unpaired Student’s t-test for preference index. (H) The comparisons of the social ability index (top) and social memory index (below) of winner and loser group before and after the tube tests. *P < 0.05, **P < 0.01, ns, non-significant. Two-way ANOVA Multiple comparisons, sidak’s post-hoc test). n = 6 mice for each group.

The social status dependent PL-NAc and PL-BLA neuronal activities.

(A) Schematic showing the AAV-ChR2 injection in the PL of mPFC. (B) Expression of ChR2-EYFP in the mPFC, and terminal expression of ChR2-EYFP in the NAc and BLA. Scale bar, 400 μm, 50 μm, and 50 μm, from left to right. (C) The ChR2-EYFP expressed mice underwent tube test to determine their winner and loser status. (D) The EPSCs in the NAc were evoked by 470 nm blue light. Upper, schematic showing the light fiber placement and EPSC recording sites in the NAc; Lower, representative EPSCs in the brain slices from winner (red) and loser (black) mice. Scale bar, 30 pA and 25 ms. (E) The normalized light intensity-dependent EPSCs. The data points were normalized to 10 μW and were shown as mean value. Winner, n= 19 neurons/ 7 mice; Loser n= 14 neurons/ 7 mice. (F) The representative pair pulse ratio (PPR) of the EPSCs in the NAc (the light intensity were 1-2 μW and the interpulse interval were 50 and 75 ms). Scale bar, 8 pA and 25 ms. (G) Summarized plots of the PPR under different interpulse intervals. Winner, n= 7 neurons/ 7 mice; Loser, n = 14 neurons/ 7 mice. ns, non-significant. Scale bar, 8 pA and 25 ms. (H-I) EPSC recordings in the BLA. Winner, n = 17 neurons/ 7 mice; Loser n = 21 neurons/ 7 mice. Scale bar, 6 pA and 25 ms. (J) The representative PPR of the EPSCs in the BLA. (K) Summarized plots of the PPR under different interpulse intervals. Winner, n = 8 neurons/ 7 mice; Loser, n = 11 neurons/ 7 mice. Scale bar, 3 pA and 25 ms. (L) Left, representative EPSCs in the PL-NAc neurons from winner (red) and loser (black) mice. Right, the summarized AMAP/NMDA ratio. Winner, n= 5 neurons/ 5 mice; Loser n= 7 neurons/ 7 mice, 1-5 ms pulse). Scale bar, 25 pA and 50 ms. (M) Left, representative EPSCs in the PL-BLA neurons from winner (red) and loser (black) mice. Right, the summarized AMAP/NMDA ratio. Winner, n= 12 neurons/ 7 mice; Loser n= 21 neurons/ 7 mice, 1-10 ms pulse). Scale bar, 10 pA and 50 ms. Data are shown as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. Two-way ANOVA for E, G, I, and K; two-tailed paired Student’s t-test for L and M.

Individual data points for transient rate across behavioral states (related to Fig. 1H).

Scatter plot showing transient rates of all mPFC→BLA (blue and mPFC→Nac (orange) neurons across four behavioral states (Center, Corner, Sniffing, and Grooming). Each blue dot represents data from an individual mouse, with connecting lines indicating within-animal comparisons across behaviors. The plot demonstrates that the observed group differences in Figure 1H are not driven by outliers, as individual data points show consistent trends across animals.

Cumulative distribution of center-ON neuron percentages across mice (related to Fig. 2C).

Cumulative plots showing the percentage of center-ON neurons per mouse for the mPFC→BLA (blue) and mPFC→NAc (orange) pathways. The distribution of center-ON neuron percentages was significantly left-shifted in mPFC→NAc compared to mPFC→BLA neurons (two-sample Kolmogorov–Smirnov test, D = 0.61, p = 0.032), confirming a lower overall proportion of center-ON neurons in the mPFC→NAc pathway.

Individual data points for transient rate of center-ON neurons across behavioral states (related to Fig. 2E).

Scatter plot showing transient rates of mPFC→BLA (blue) and mPFC→NAc (orange) center-ON neurons during Center, Corner, Sniffing, and Grooming behaviors. Each dot represents the average transient rate from an individual mouse, with connecting lines indicating within-animal comparisons across behavioral states. The plot confirms that the group differences shown in Figure 2E are consistent across animals and are not driven by outliers.

Individual data points for transient rates of all neurons in different arms of the elevated plus maze (related to Fig. 3C).

Scatter plot showing transient rates of all recorded mPFC→BLA (blue) and mPFC→NAc (orange) neurons during Open Arm, Closed Arm, Center, and Grooming periods in the elevated plus maze (EPM). Each dot represents the mean transient rate from an individual mouse, with connecting lines indicating within-animal comparisons across behavioral conditions. The consistent distributions across animals confirm that the differences observed in Figure 3C are robust and not driven by individual outliers.

Individual data points for transient rate of in different arms of the elevated plus maze (EPM) (related to Fig. 3G).

Scatter plot showing transient rates of mPFC→BLA (blue) and mPFC→NAc (orange) center-ON neurons of in different arms of EPM. Each dot represents the average transient rate from an individual mouse, with connecting lines indicating within-animal comparisons across behavioral states. The plot confirms that the group differences shown in Figure 3G are consistent across animals and are not driven by outliers.

ChR2-induced currents and membrane depolarizations in the PL neurons from the winner and loser mice.

(A) Schematic showing the light fiber placement and EPSC recording sites in the mPFC. (B) The representative blue light intensity-dependent photocurrents. (C) The normalized light intensity-dependent photocurrents in the neurons from winner (red) and loser (black) mice. Winner, n= 12 neurons/ 7 mice; Loser n= 16 neurons/ 7 mice. (D) The representative voltage traces of membrane depolarization and spikes in a current-clamped PL neuron induced by blue lights. (E) The summarized numbers of spikes and frequency of light stimulations. Winner, n= 7 neurons/ 4 mice; Loser n= 7 neurons/ 4 mice. Data are shown the as mean ± SEM. Two-way ANOVA. No significant difference was observed between the winner and loser groups.