Dataset overview

This dataset contains in vivo calcium (Ca2+) imaging data and paired behavioral data from adult female mice, collected to compare how two output pathways of the medial prefrontal cortex (mPFC) represent affect-related behavioral states: mPFC neurons projecting to the basolateral amygdala (mPFC→BLA) and mPFC neurons projecting to the nucleus accumbens (mPFC→NAc).

Prelimbic (PL) neurons were retrogradely labeled with the genetically encoded Ca2+ indicator GCaMP6m by injecting a retrograde adeno-associated virus (AAV) into either the BLA or the NAc of female CaMKII-Cre mice (~4 months old at surgery). Activity of individual labeled neurons was recorded through an implanted 1 mm diameter gradient-index (GRIN) lens using a head-mounted miniature microscope (nVista HD, Inscopix; 15 frames/s, 1024 × 1024 pixels) in freely moving mice during three behavioral assays run on consecutive days, at least 24 h apart:

1. Open-field test (OFT) --- 10 min, 50 × 50 × 50 cm arena.
2. Elevated plus maze (EPM) --- 10 min, two open and two closed arms
   (each 35 × 10 cm), 40 cm above the floor.
3. Modified three-chamber social test (mTC) --- 10 min habituation plus three 10-min sessions (S1, S2, S3).

A separate cohort of mice (no imaging) underwent a repeated social-dominance tube test to establish Winner (dominant) and Loser (subordinate) status, with open-field and social testing before and after the tube test, plus measurement of corticosterone in brain (mPFC) lysates by ELISA.

Files are organized by the figure of the associated publication in which they appear. Each figure folder/file is described below, including the biological meaning of each variable, the units, and the group sizes. Statistical results quoted here are provided so that the values in the files can be interpreted without consulting the manuscript.

Associated publication: Lai C-H, Park G, Xu P, Sun X, Ge Q, Jin Z, 
Betts S, Liu X, Liu Q-S, Simha R, Zeng C, Lu H, Du J. Functional specialization of mPFC-BLA and mPFC-NAc pathways in affective state representation. eLife 14:RP105528. https://doi.org/10.7554/eLife.105528

Associated code: Custom Python scripts for Ca2+ trace extraction, event detection, and analysis:
https://github.com/XiaoqianSun0104/CaImAn-based-Ca2-Processing

File name: eLife-plot_value-072326.xlsx

---

Abbreviations

---

Abbreviation Definition

---

mPFC medial prefrontal cortex (recording site; imaging restricted to the prelimbic subdivision)

PL / IL prelimbic/infralimbic subdivision of the mPFC

BLA basolateral amygdala (retrograde injection site defining the
mPFC→BLA group)

NAc nucleus accumbens (retrograde injection site defining the
mPFC→NAc group)

AAV adeno-associated virus

GCaMP6m genetically encoded fluorescent
Ca2+ indicator

GRIN lens gradient-index lens used for
deep-brain optical access

ΔF/F change in fluorescence relative to baseline fluorescence, (F − F0)/F0; 
unitless ratio

F0 baseline (background) fluorescence, 
taken as the average projection of the filtered video

Ca2+ transient / event detected calcium event, identified with a peak-finding algorithm on
ΔF/F traces

Transient rate (event rate) number of detected Ca2+ transients per second (Hz)

Normalized transient rate transient rate scaled using
min--max normalization; unitless

z-score activity standardized per neuron:
(value − mean)/SD; unitless

OFT open-field test

Center central 25 × 25 cm zone of the
open-field arena (exposed state)

Corner corner sector of 12.5 cm radius in each of the four corners (sheltered reference state); excludes sniffing episodes

Sniffing nose within ~1 cm of the wall with small scanning movements and no locomotion, ≥1 s (sensory
exploration/vigilance)

Grooming self-directed grooming behavior
(stress regulation/comfort)

center-ON neuron: a neuron whose activity is significantly elevated around entry into the Center zone; defined by comparing actual trace--behavior
similarity (Sa) againstnull distributionion from 5000
behavior-label shuffles, with classification at the 99.95th
percentile

EPM elevated plus maze

Open-P / Open-NP preferred/non-preferred open arm of the EPM (exposed)

Close-P / Close-NP preferred/non-preferredclosed armm of the EPM (sheltered)

mTC modified three-chamber social interaction test

M1 first (initially unfamiliar)social stimulus mouse

M2 second, novel social stimulus mouse introduced in session 3

O inanimate object (non-social
stimulus)

S1, S2, S3 mTC sessions 1 (M1 vs O), 2
(positions of M1 and O swapped), 3
(M2 vs familiar M1)

SZ/OZ social zone/object zone
(interaction zones adjacent to the stimulus chambers)

PCA principal component analysis

PC1, PC2, PC3 first, second, third principal components (arbitrary units of the
z-scored activity space)

FWHM full width at half maximum of the distribution of pairwise Pearson correlation coefficients; unitless.
Lower FWHM = greater pattern decorrelation

Winner / Loser: dominant/subordinate mouse determined by the repeated tube test

Control singly housed, handled mice that did not undergo the tube test

ELISA enzyme-linked immunosorbent assay

SEM: standard error of the mean

ns: not statistically significant

Units used throughout

• Time: seconds (s) or minutes (min); percentage of session time (%).
• Distance traveled in behavioral arenas: meters (m).
• Ca2+ event/transient rate: hertz (Hz, events/s).
• Ca2+ signal amplitude: ΔF/F (unitless ratio) or z-score (unitless).
• PCA coordinates and distances: arbitrary units of PC space. For
  Corner-referenced analyses, Euclidean distances were calculated from individual population-activity observations to the centroid of the Corner distribution.
• Correlation coefficients and FWHM: unitless.
• Corticosterone: values are reported in ng/mL as provided in the source data; samples were normalized to total protein determined by BCA assay.
• Spatial coordinates in the arena/maze: pixels or cm as indicated in each file header.

Sample sizes

• OFT imaging: n = 15 mice (mPFC→BLA), n = 11 mice (mPFC→NAc).
• EPM imaging: n = 9--13 mice (mPFC→BLA), n = 8--11 mice (mPFC→NAc); 
  the range reflects exclusion of mice lacking entries into a given arm or lacking center-ON neurons.
• mTC imaging: n = 13 mice (mPFC→BLA), n = 11 mice (mPFC→NAc) for
  event rates; n = 8 mice per pathway for correlation/FWHM analyses; n = 10 mice for the behavioral preference data.
• Neurons per mouse: 11--83 (mPFC→BLA; mean 38) and 18--54 (mPFC→NAc; mean 39) GCaMP6m-expressing neurons. After PCA--ICA source extraction, ~1--9% of components were rejected as artifacts.
• Cross-session neuron matching yield (OFT used as reference):mPFC→BLA, ~30% OFT↔EPM (range 5--92%, n = 19 mice) and ~16% OFT↔mTC (range 5--32%, n = 14 mice); mPFC→NAc, ~23% OFT↔EPM (range 11--71%, n = 20 mice) and ~21% OFT↔mTC (range 13--37%, n= 17 mice).
• Tube-test cohort: n = 6 mice each for Winner and Loser groups. A separate singly housed Control group was included for the corticosterone analysis.

---

Description of files by figure

Figure 1 --- Open-field population activity

• 1B--1C: Representative histology, miniscope field of view, and example ΔF/F traces. Retrograde labeling sites are BLA (_{bregma−1.6 mm) and NAc (}+1.3 mm), with labeled PL somata (~+2.1 mm). Scale bars, 200 µm.
• 1F: Percentage of the 10-min OFT spent in Center, Corner, Sniffing, and Grooming. Corner time excludes sniffing episodes occurring in corners. n = 15 BLA and n = 11 NAc.
• 1G: Example per-frame neuronal ΔF/F activity and aligned behavioral labels across the OFT.
• 1H: Normalized Ca2+ transient rate (unitless) for the four behavioral states. No Region main effect (p = 0.14) or Region × Behavior interaction (p > 0.3). The publication Methods describe min--max normalization; the deposited source-data values should be used as the authoritative panel values.
• 1I: PC1/PC2 coordinates of population activity. The Figure 1 legend reports high cross-pathway similarity for Corner activity (Pearson r = 0.64, p < 1 × 10⁻9). A separate canonical-correlation analysis in the Results likewise identifies Corner as the most similar cross-pathway representation.
• 1J: Euclidean distances of Center, Sniffing, and Grooming population-activity observations from the Corner cluster centroid in PC1--PC2 space, calculated separately for the mPFC→BLA and mPFC→NAc pathways. Center and Grooming are farther from Corner in mPFC→BLA; Sniffing is farther from Corner in mPFC→NAc (***p < 0.001, Mann--Whitney U). Permutation controls used 10,000 permutations.
• Figure 1---figure supplement 1: Individual data underlying Figure 1H.

Figure 2 --- center-ON ensembles in the OFT

• 2A: Center-entry-aligned ΔF/F from −5 to +5 s; representative mPFC→BLA mouse has 24 center-ON neurons.
• 2B: Behavior-aligned activity heatmaps for Center, Corner, Sniffing, and Grooming.
• 2C: Percentage of imaged neurons classified as center-ON. mPFC→BLA has a higher percentage than mPFC→NAc (Mann--Whitney U =47, p = 0.049).
• 2D: Spatial activity maps of center-ON neurons in the OFT.
• 2E: Normalized transient rate (unitless) of center-ON neurons. Region × Behavior interaction p = 0.048; mPFC→NAc center-ON neurons show higher activity during Sniffing than mPFC→BLA neurons. The publication Methods describe min--max normalization; deposited source-data values should be used as the authoritative panel values.
• 2F: PC1/PC2 coordinates for center-ON populations.
• 2G: Euclidean distances of Center, Sniffing, and Grooming center-ON population-activity observations from the Corner cluster centroid in PC1--PC2 space (**p < 0.01, ***p < 0.001).
• Figure 2---figure supplements 1--2: Cumulative distribution of center-ON percentages and individual data underlying 2E.

Figure 3 --- Elevated plus maze

The EPM was performed one day after the OFT in the same mice. The publication labels the four conditions Open-P, Open-NP, Close-P, and Close-NP.

• 3B: Percentage of the 10-min session spent in each EPM arm.
• 3C: Mean transient rate (Hz) of all neurons. Open-arm rates were higher than closed-arm rates, with no Region main effect (p = 0.304) and no Region × Arm interaction (all p > 0.29).
• 3E: Mean normalized activity of center-ON neurons mapped onto EPM arms for an example mouse.
• 3F: Center-ON activity aligned to arm entry (time 0), showing approximately the 5-s post-entry period for open- and closed-arm entries.
• 3G: Center-ON transient rates. Region × Condition interaction F(3,72) = 3.21, p = 0.030; BLA > NAc in Open-NP (p = 0.018), with no regional differences in the other arms. n = 9--13 BLA and n = 8--11 NAc.
• Figure 3---figure supplements 1--2: Individual data underlying 3C and 3G.

Figure 4 --- Social interaction and pattern decorrelation

S1 = M1 vs O; S2 = same stimuli with positions swapped; S3 = novel M2 vs
familiar M1.

• 4B: Behavioral interaction time (%), n = 10. Mice preferred M1 over O in S1 and M2 over familiar M1 in S3; S2 was not significant.
• 4C--4D: Mean Ca2+ event rate (Hz), n = 13 BLA and n = 11 NAc. No significant stimulus-dependent event-rate differences occurred in either pathway.
• 4E--4F: Pairwise Pearson correlation matrices and coefficient distributions during stimulus interactions.
• 4G--4H: FWHM of correlation distributions, n = 8/pathway. BLA showed no significant stimulus-dependent decorrelation. NAc showed significant decorrelation in S1 (p = 0.025) and S3 (p = 0.021), but not S2.

Figure 5 --- Social versus nonsocial population encoding

• 5A, 5C: PC1/PC2 coordinates during M1 versus O interaction for all recorded neurons (5A) and OFT-defined center-ON neurons (5C).
• 5B, 5D: Euclidean separation between M1 and O activity clusters in PC1--PC2 space, using cluster centroid coordinates as described in the publication. For all neurons, separation is greater in mPFC→BLA; for center-ON neurons, separation is greater in mPFC→NAc (***p < 0.001, Mann--Whitney U).

Figure 6 --- Repeated social competition, behavior, and corticosterone

This cohort was not imaged. Singly housed mice were handled for 3 days and then underwent repeated tube testing for 5 days, as shown in Figure 6A of the publication, to establish Winner and Loser status. Singly housed mice that did not undergo tube testing served as Controls for corticosterone measurement. Behavioral assays were conducted 24 h after, and tissue collection within 1 h of, the final tube-test session . 
Behavioral comparisons used n = 6 mice each for Winner and Loser groups.

• 6B: Figure6B_Corticosterone.csv.csv and Figure6B_Corticosterone.xlsx.xlsx. Corticosterone measured by ELISA in mPFC lysates. Values are reported as in the source data/Figure 6B (ng/mL); sampleswere normalizedd to total protein content determined by BCA assay. Losers had higher corticosterone than Control (**p < 0.01) and Winner
  (#p < 0.05).
• 6C--6D: OFT distance traveled (m) and center time (s) before and after tube testing. After testing, Losers spent less time in the center while locomotor distance was unchanged.
• 6E: Within-animal before/after OFT values for distance and center time.
• 6F--6G: Social ability and social memory before and after tube testing. Social ability index = (M1 − O)/(M1 + O). Social memory index = (M2 − M1)/(M2 + M1). Both groups showed social and novelty preference before tube testing. After tube testing, Winners showed enhanced sociability, whereas Losers showed a marked reduction/loss of social preference.
• 6H: Before/after social ability and social memory indices for Winner and Loser groups.

---

Data collection and processing

• Behavior: TopScan (CleverSys) was used for automated position and behavior tracking and sent a TTL pulse to synchronize behavioral video and miniscope recording.
• Ca2+ imaging: nVista HD miniature microscope; 15 Hz; 1024 × 1024 pixels. Imaging parameters were kept constant within each mouse across sessions.
• Image processing: Inscopix Data Processing Software v1.3.1. Videos were preprocessed, spatially filtered, motion corrected, and converted to ΔF/F. Neurons were identified using PCA--ICA without spatial or temporal downsampling. Approximately 1--9% of components were rejected as artifacts.
• Normalization: The publication Methods describe min--max normalization as zi = (xi − min(x))/(max(x) − min(x)), where normalized values are reported.
• Cross-day alignment: Neuron footprint images were vectorized and matched across tests by maximizing cosine similarity while iteratively adjusting rotation and pixel shift. Image scale was unchanged across sessions.
• PCA: Ca2+ event rates were z-scored per neuron across time and concatenated across mice along the neuron dimension. PCA was performed on the pooled frames × neurons matrix. For analyses using Corner as the reference state, Euclidean distances were calculated in PC1--PC2 space from individual population-activity observations to the centroid of the Corner distribution.
• Exclusions: Mice were excluded for incorrect GRIN-lens
  placement, absent/incorrect GCaMP expression, substantial expression outside PL, or imaging planes obscured by blood/debris. Analyses requiring center-ON classification excluded animals with no center entries or no neurons meeting the center-ON criterion.
• Statistics: Parametric and nonparametric tests were applied as described in the publication. Data are mean ± SEM unless otherwise stated. Neural-state-space distance metrics were additionally assessed against 10,000-permutation null distributions.

Missing data

Empty cells indicate that a value was not obtained for that animal, neuron, or behavioral state --- most commonly because the animal did not perform the behavior during the session (for example, no entries into a given EPM arm), because no neurons in that animal met the center-ON criterion, or because the animal was excluded for the histological or imaging-quality reasons listed above. Missing values are leftblank ratherr than coded numerically.

Ethics

All animal care and use followed NIH guidelines and was approved by the Institutional Animal Care and Use Committee of George Washington University (IACUC #A2025-158) and the University of Tennessee Health Science Center Laboratory Animal Care Unit (IACUC #25-0651).

Funding

NIH/NINDS R01NS118197 and the George Washington University 2018--2023 Cross-Disciplinary Research Fund (H. Lu, R. Simha, C. Zeng); NIH/NIMH R01MH135862, NIH/NINDS R01NS143931, Cystic Fibrosis Foundation 002544I221, and University of Tennessee Health Science Center start-up funds (J. Du).

Contact

Jianyang Du, Department of Anatomy and Neurobiology and Neuroscience, Institute: University of Tennessee Health Science Center, Memphis, T N, 
USA --- jdu15@uthsc.edu
Hui Lu, Department of Pharmacology and Physiology, George Washington University, Washington, DC, USA --- huilu@gwu.edu