Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Mar 29.
Published in final edited form as: Neurobiol Learn Mem. 2025 Dec 22;223:108128. doi: 10.1016/j.nlm.2025.108128

Ensembles and engrams in mouse cortical and sub-thalamic brain regions supporting context and memory recall

William W Taylor a,b,*, Vienna Gao b, Laura Korobkova a,b, Brian G Dias b,c,d,**
PMCID: PMC13032769  NIHMSID: NIHMS2133101  PMID: 41443432

Abstract

Associative learning supports learning about outcomes associated with contexts and cues. During learning, cellular ensembles that become active can be incorporated into memory engrams and later reactivated to support recall. Studies exploring engram formation and reactivation have primarily used contextual conditioning in mice and made little distinction between ensembles supporting contextual information versus cue-associated learning and recall. Furthermore, often missing in such analyses is exploration of sex differences in ensemble dynamics. Using auditory fear conditioning and activity-dependent tagging in mice, we set out to disaggregate context-associated ensembles from those associated with cue-related learning and recall while also profiling potential sex differences. Specifically, we quantified cellular activity during context exposure, fear recall, extinction training, and extinction recall in cortical and subthalamic brain regions supporting learning and memory. We found that male mice had denser ensembles of cells active in the infralimbic prefrontal cortex (IL-PFC) during context exposure, while female mice had a significantly greater proportion of newly active cells in the IL-PFC during fear recall. We also found a sexually dimorphic pattern of correlation between activity in the IL-PFC and in the zona incerta (ZI). Across sexes, we found denser overlapping cells and greater reactivation of extinction ensembles in the IL-PFC. These results emphasize that there is a distinction to be made between ensembles supporting contextual information from those encoding cue-associated memory and highlight important sex differences in ensemble dynamics.

Keywords: Engrams, Context, Fear, Extinction, Learning, Memory

1. Introduction

Associating contexts and environmental cues with outcomes and then later recalling these associations is integral to survival and our daily lives. Recently, we have come to appreciate that groups of neurons (ensembles) fire during associative learning with a subset of these neurons forming a memory trace (engrams) that is later reactivated to support memory recall (Frankland, Josselyn, & Kohler, 2019, 2024; Guskjolen & Cembrowski, 2023; Josselyn & Tonegawa, 2020). Further support for the role of such ensembles and memory engrams in successful learning and recall come from studies labeling cells in specific contexts or when specific cues are becoming associated with aversive or appetitive outcomes and later modulating their activity when recall of the contexts or cues is being tested. Broadly speaking, silencing cells that had been previously incorporated into engrams significantly disrupts memory recall, while activation can elicit expression of learned behaviors independent of any conditioned context or cue that had been encountered at the time of learning (Cowansage et al., 2014; Han et al., 2009; Hsiang et al., 2014; Josselyn & Tonegawa, 2020; Lacagnina et al., 2019; Liu et al., 2012; Redondo et al., 2014; Tanaka et al., 2014; Vetere et al., 2019). Together such studies have established engrams as important units of memory in the brain that come to encode learning and support memory recall.

While much progress has been made in the study of memory ensembles and engrams, three dimensions require more attention. First, memory ensembles and engrams have primarily been studied in the canonical fear circuitry made up of the hippocampus, prefrontal cortex and amygdala (Venkataraman & Dias, 2023). We are beginning to appreciate that brain regions outside of this tripartite prefrontal cortex-amygdala-hippocampus canon, specifically subthalamic regions like the zona incerta, play an important role in normative associative learning (Chou et al., 2018; Venkataraman & Dias, 2023; Venkataraman et al., 2019, 2021). The zona incerta is also highly interconnected with canonical fear circuitry, including the prefrontal cortex (PFC) (Chou et al., 2018; Zhao et al., 2020). This emerging role of the zona incerta, and its connectivity, opens the possibility that learning and memory related ensembles exist in the zona incerta and could potentially support memory alongside ensembles in canonical regions like the PFC to which the zona incerta is connected. Second, thus far, studies have made little to no distinction between ensembles encoding contextual information versus ensembles supporting cue associations, despite these being dissociable elements of memory. Third, sex differences in ensemble and engram dynamics are an underappreciated area of inquiry despite reported sex differences in learning and memory related dimensions of neuropsychiatric disorders like Post Traumatic Stress Disorder (PTSD) (Lebron-Milad & Milad, 2012; Ramikie & Ressler, 2018; Velasco, Florido, Milad, & Andero, 2019).

With the intent of expanding our exploration of ensembles and engrams to the sub-thalamic zona incerta coupled with the PFC while also accounting for context, cue, and sex-differences in learning and memory, we trained TRAP2 (Targeted Recombination in Active Populations); floxed-tdTomato male and female mice to associate conditioned stimulus (CS+) tone presentations in Context A with mild footshocks. Next, we activated the TRAP2 system (DeNardo et al., 2019) to specifically label cells during one of two conditions: either during exposure to Context B alone without any extinction training (context group), or alternatively, during extinction training when animals were exposed to CS+ tone presentations without footshocks in Context B (extinction group). This genetic tagging approach allowed us to capture and distinguish cellular ensembles activated under these experimental conditions. Finally, we stained for C-FOS following CS+ presentations without footshocks in Context B, which constituted a fear recall test in the context group and an extinction recall test in the extinction group. This approach allowed us to investigate ensemble size and reactivation patterns during context exposure, extinction training, fear recall, and extinction recall in the zona incerta (ZI), prelimbic prefrontal cortex (PL-PFC) and infralimbic prefrontal cortex (IL-PFC). Briefly, our results demonstrate sex differences in the size of ensembles in the IL-PFC active during context exposure and in the activation of new IL-PFC ensembles during fear recall, preferential IL-PFC reactivation during extinction recall in both sexes, and sex-specific correlations between cellular ensembles in the IL-PFC and ZI.

2. Methods

2.1. Animals

TRAP2 mice (Fostm2.1(icre/ERT2)Luo/J, strain 030323) and floxed-tdTomato reporter mice (B6.Cg-Gt(ROSA)26Sortm14(CAG-tdTomato)Hze/J, strain 007914) were obtained from Jackson Laboratory (Bar Harbor, ME, USA) and crossed to generate adult male and female TRAP2/+; floxed-tdTomato/+ mice for all experiments (n = 15 females, 14 males). These mice allow for fluorescent tdTomato expression in CRE-expressing cells when 4-Hydroxytamoxifen (4OHT) is present (Supplemental Fig. 1) (DeNardo et al., 2019). Genotyping was performed using Jackson Laboratory “Quick DNA purification protocol” from ear clips. The mice were group housed and maintained on ad libitum food and water on a 12-hour light/dark cycle. Both male and female mice were included in all experiments. Animals were handled prior to experiments to minimize stress but were not habituated to injections. Experiments were approved by the Institutional Animal Care and Use Committee of Children’s Hospital Los Angeles and followed NIH standards.

2.2. Auditory fear conditioning, context exposure, extinction training, and extinction recall

All behavioral sessions were conducted in conditioning chambers (Coulbourn Instruments) which included tone generators (internal speakers) and shock generators (cage-floor inserts) controlled by FreezeFrame software (Actimetrics). Freezing behavior was recorded and measured via video algorithms embedded in FreezeFrame software. All behavioral sessions took place during the light cycle and animals were returned to the vivarium afterwards.

Day 0: Habituation to Context A.

Mice were habituated to Context A for 5 min, which consisted of metal rod flooring, chamber and room lights off, Peroxigard as a cleaning agent, and an infrared light to allow for recording.

Day 1: Fear conditioning in Context A.

The mice underwent auditory fear conditioning, consisting of a 180-second baseline period followed by five 30-second 6 kHz tone (T) presentations, the conditioned stimulus (CS+), each co-terminating with a one-second foot shock (0.5 mA), the unconditioned stimulus (US). Tone-shock pairing were presented with 30-second inter-trial intervals (ITI).

Day 2: Habituation to Context B.

Mice were habituated to Context B for 5 min, which was the same behavioral chamber but with distinct contextual cues including a plexiglass floor, chamber and room lights on, and 70 % ethanol as a cleaning agent.

Day 3: 4OHT injection and extinction learning or context exposure in Context B.

One hour before extinction training or context exposure, all mice were injected with 4OHT. 4OHT was generously provided by the NIMH Chemical Synthesis and Drug Supply Program or purchased from Sigma-Aldrich (St. Louis, MO,USA; catalogue #H6278). 4OHT was prepared fresh on the day of injection, as described by DeNardo et al., 2019. Briefly, 4OHT was dissolved in ethanol at 20 mg/ml by shaking at 37 °C. Corn oil (Sigma-Aldrich, catalogue #C8267) was added for a final concentration of 10 mg/ml and ethanol was evaporated. Mice were injected intraperitoneally (i.p.) at a dose of 50 mg/kg. The extinction group (n = 17) received a 180-second baseline period followed by 30 CS+ tone presentations, separated by 30-second ITIs, without any footshock to facilitate extinction learning. The context group (n = 12) was placed in Context B without being exposed to any tones for the same amount of time as the extinction training session lasted. Animals were returned to the vivarium and left undisturbed for 4 days to allow for robust expression of tdTomato in TRAPped cells.

Day 8: Extinction recall or fear recall testing in Context B and brain collection.

All mice were returned to Context B and exposed to 5 CS+ presentations without foot shocks, separated by 30-second ITIs. For the extinction group, this protocol tested the recall of extinction learning and in the context group this tested the recall of the initial fear conditioning. Mice were deeply anesthetized with a mixture of ketamine and dexdomitor, 60–90 min after testing for recall, and transcardially perfused with ice-cold phosphate-buffered saline (PBS), followed by 4 % paraformaldehyde (PFA) in PBS, and brains were collected for further analysis.

2.3. Immunohistochemistry

Brains were fixed overnight in 4 % PFA in PBS and then transferred to a 30 % sucrose solution in PBS for 3–4 days until fully saturated. Once saturated, brains were flash-frozen and coronally sectioned at 35 μm using an Epredia CryoStar NX70. Sections were mounted onto slides and stored at −80 °C without light exposure. For C-FOS immunohistochemistry, brain sections were washed in 1X PBS and blocked for one hour with 5 % normal goat serum and 0.1 % Triton-X. Sections were incubated overnight at 4 °C with Rabbit anti C-FOS (1:500; sc-52, Santa Cruz Biotechnology, Dallas, TX). Sections were washed then incubated for 2 h at room temperature with Goat anti-Rabbit Alexa Fluor™ 488 (1:1,000; A11034, ThermoFisher, Waltham, MA). Nuclei were counterstained with Hoechst (1:10,000; H3570, Invitrogen, Waltham, MA). Images were acquired using a Leica STELLARIS 5 WLL microscope with a 20x/0.75 objective.

2.4. Image and data analysis

Image analysis was conducted using QuPath software (Bankhead et al., 2017). Regions of interest were identified and delineated based on the Allen Mouse Brain Atlas (mouse.brain-map.org). TRAP+ and C-FOS+ cells were automatically detected using a consistent threshold and manually confirmed for accuracy. Raw cell counts were measured for each region by analyzing 2–3 images per animal and the average density of TRAP+, C-FOS+ and TRAP+/C-FOS+ cells (cells per mm2) was calculated with individual animals serving as the unit of analysis. To compute the proportion of newly activated cells, the number of exclusively C-FOS+ (TRAP−) cells was divided by total C-FOS+ cells. To compute percentages of reactivated cells, the number of double labelled cells (TRAP+ and C-FOS+) was divided by total TRAP+. Statistical analyses for both image analysis and behavior were conducted using RStudio (version 4.3.2). Individual group or sex comparisons were done using the Wilcoxon signed-rank test and posthoc testing was conducted using the same and Tukey HSD. Pairwise Pearson correlation analyses were conducted between cellular measures. For each correlation, the coefficient (r) and corresponding p-value were calculated using the cor. test function. To account for multiple comparisons, false discovery rate (FDR) correction was applied to the resulting p-values using the Benjamini–Hochberg procedure.

3. Results

3.1. Male and female mice show similar freezing behavior during context exposure and fear recall

The context group initially underwent fear conditioning consisting of 5 CS+ (tone) paired with unconditioned stimulus (US, footshock) in Context A (Fig. 1, Day 1). A repeated measures ANOVA was conducted to examine the effects of sex and tone presentation on freezing behavior. There was no significant main effect of sex (F(1,10) = 0.001, p > 0.05). However, there was a significant main effect of tone presentation (F(4, 40) = 32.95, p < 0.001). Post-hoc testing revealed that both male and female mice froze significantly more to the 5th tone presentation than the 1st, indicating they successfully formed an association between the CS+ and US (both p < 0.0001). There was also a significant interaction between sex and tone presentation (F(4, 40) = 3.53, p = 0.01), but post-hoc testing comparing freezing in males and females at each CS+ presentation revealed no significant differences at any specific tone presentation. Following fear conditioning, both males and females demonstrated similar levels of freezing on average over the full duration of exposure to Context B (W = 7, p > 0.05) (Fig. 1, Day 3) after being injected with 4OHT 60 min prior to context exposure. Finally, for fear memory recall, animals were exposed to 5 tone presentations in Context B (Fig. 1, Day 8). Males and females did not differ in their freezing behavior during the 5 tone presentations (W = 16, p > 0.05). A repeated measures ANOVA comparing context exposure with fear recall sessions revealed a significant main effect of session (F(1, 10) = 53.08, p < 0.0001) and no main effect of sex (F(1, 10) = 0.725, p > 0.05) or any significant interaction between sex and session (F(1, 10) = 3.499, p > 0.05). Post-hoc testing collapsed by sex showed that mice froze significantly more during fear recall than context exposure (V = 0, p < 0.001).

Fig. 1. Fear behavior during fear conditioning, contextual exposure, and fear memory recall did not differ by sex.

Fig. 1.

Open in a new tab

Both male and female mice in the context exposure group significantly increased their freezing in response to the CS+ over the course of fear conditioning, showing greater freezing to the 5th tone presentation compared with the 1st (both p < 0.0001). There were no sex differences in freezing to a neutral context during context exposure. There were also no sex differences in freezing to 5 tone presentations during fear recall. Comparing freezing during context exposure and fear recall revealed no significant effect of sex (F(1, 10) = 0.725, p > 0.05) or interaction of sex and session interaction F(1, 10) = 3.499, p > 0.05), but a significant effect of session (F(1, 10) = 53.08, p < 0.0001) and post-hoc testing collapsed by sex showed that mice froze significantly more during fear recall than context exposure (V = 0, p < 0.0001). Data show percent time freezing and are represented as mean ± SEM.

3.2. Sexually dimorphic ensemble activation during context exposure in IL-PFC, but not PL-PFC and ZI

Density of TRAP+ cellular ensembles in the PL-PFC, IL-PFC, and ZI were compared between the sexes (Fig. 2A-C). Compared to female mice, male mice had significantly larger ensembles of cells labeled in the IL-PFC (W = 4, p = 0.03), but not in the PL-PFC or the ZI (both p > 0.05) (Top Row). No significant differences in the density of C-FOS+ ensembles (Fig. 2D-F) were found in the PL-PFC, IL-PFC, or ZI between male and female mice (all p > 0.05) (Middle Row). Similarly, we found no sex differences in the density of overlapping TRAP+ and C-FOS+ cells in the PL-PFC, IL-PFC, and ZI (all p > 0.05) (Fig. 2G-I) (Bottom Row).

Fig. 2. Sex differences in ensemble activation during context exposure in IL-PFC, but not PL-PFC and ZI.

Fig. 2.

Open in a new tab

(A-C) Density of TRAP+ ensembles were measured in the PL-PFC, IL-PFC, and ZI. Male mice had significantly denser TRAP+ ensembles in the IL-PFC (W = 4, p = 0.03). Data show mean # of TRAP+ cells per mm2 ± SEM. (D-F) Density of C-FOS+ ensembles were measured in the PL-PFC, IL-PFC, and ZI, showing no significant sex differences. Data show mean # of C-FOS+ cells per mm2 ± SEM. (G-I) Density of overlapping (TRAP+ and C-FOS+) cells were measured in the PL-PFC, IL-PFC, and ZI, showing no significant sex differences. Data show mean # of both TRAP+ and C-FOS+ cells per mm2 ± SEM. The value of data points not fitting on the axis are labelled in italics and are not statistical outliers.

3.3. Male and female mice show similar freezing behavior during extinction learning and recall

Similarly to the context group, the extinction group initially underwent fear conditioning in Context A (Fig. 3, Day 1). A repeated measures ANOVA revealed no main effect of sex (F(1, 15) = 0.096, p >0.05), but a main effect of tone presentation (F(4, 60) = 30.534, p < 0.001) and no significant interaction (F(4, 60) = 1.452, p > 0.05). Post-hoc testing revealed that by the 5th tone presentation of fear conditioning, both male and female mice of the extinction group successfully associated the tone with the footshock and demonstrated significantly more freezing behavior to the 5th tone than initially observed to the 1st tone presentation (both p < 0.0001). During extinction training (Fig. 3, Day 3), animals were exposed to 30 tone presentations in the absence of footshock in Context B, after being injected with 4OHT 60 min prior to extinction training. A repeated measures ANOVA revealed no main effect of sex (F(1, 15) = 0p > 0.05), but that a main effect of CS+ bin (5 tone bins) was significant (F(5, 75) = 7.209, p < 0.001). This indicates that freezing behavior varied significantly across time as animals underwent extinction training, but did not differ by sex. There was no significant interaction between sex and CS+ bin (F(5, 75) = 1.43, p > 0.05). Post-hoc testing collapsed by sex showed that mice froze significantly more to the first bin of 5 tone presentations than the last bin (V = 144, p < 0.001).

Fig. 3. Fear behavior during fear conditioning, extinction training, and extinction memory recall did not differ by sex.

Fig. 3.

Open in a new tab

Both male and female mice in the extinction group significantly increased their freezing in response to the CS+ over the course of fear conditioning in Context A, showing greater freezing to the 5th tone presentation compared with the 1st (both p < 0.0001). There were no sex differences in freezing during extinction training in Context B. Comparing freezing during 5 tone bins of extinction training by sex revealed no significant effect of sex (F(1, 15) = 0, p > 0.05), a significant main effect of CS+ bin (F(5, 75) = 7.209, p < 0.0001), and no significant interaction of sex and tone bin (F(5, 75) = 1.43, p > 0.05). Post-hoc testing collapsed by sex showed that mice froze significantly more during the first 5 tone bin than the last (V = 144, p < 0.001). There were no sex differences in freezing during extinction recall in Context B (W = 27, p > 0.05). Comparing the first 5 tone bin of extinction training (fear recall) with extinction recall revealed a significant main effect of session (F(1, 15) = 50.887, p < 0.0001), no effect of sex (F(1, 15) = 0.002, p > 0.05), and no interaction between session and sex (F(1, 15) = 1.609, p > 0.05). Post-hoc testing collapsed by sex showed that mice froze significantly more during the first 5 tone presentations of extinction training than during extinction recall (V = 153, p < 0.0001). Data show percent time freezing and are represented as mean ± SEM.

There were no sex differences in freezing during extinction recall in Context B (Fig. 3, Day 8) (W = 27, p > 0.05) and both male and female mice showed successful extinction recall, with a repeated measures ANOVA comparing the first bin of extinction training (fear recall) with extinction recall revealing a significant main effect of session (F(1, 15) = 50.887, p < 0.0001), no effect of sex (F(1, 15) = 0.002, p > 0.05), and no effect of the interaction between session and sex (F(1, 15) = 1.609, p > 0.05). Post-hoc testing collapsed by sex showed that mice froze significantly more during fear recall than during extinction recall (V = 153, p < 0.0001).

3.4. Male and female mice show similar ensemble activation during extinction training and recall

In the extinction group, there was no difference between the density of TRAP+ ensembles (Fig. 4A-C) in the PL-PFC, IL-PFC, or in the ZI of male and female mice (all p > 0.05) (Top Row). Similarly, there were no significant sex differences in the density of C-FOS+ ensembles (Fig. 4D-F) in the PL-PFC, IL-PFC, or in the ZI (all p > 0.05) (Middle Row). There were also no sex differences in the density of overlapping (TRAP+ and C-FOS+) cells in the PL-PFC, IL-PFC, and ZI (Fig. 4G-I) (all p > 0.05) (Bottom Row).

Fig. 4. No sex differences in ensemble activation during extinction training, extinction recall, or engram reactivation during extinction recall.

Fig. 4.

Open in a new tab

(A-C) Density of TRAP+ ensembles were measured in the PL-PFC, IL-PFC, and ZI and no significant sex differences were found. Data show mean # of TRAP+ cells per mm2 ± SEM. (D-F) Density of C-FOS+ ensembles were measured in the PL-PFC, IL-PFC, and ZI and no significant sex differences were found. Data show mean # of C-FOS+ cells per mm2 ± SEM. (G-I) Density of overlapping (TRAP+ and C-FOS+) cells were measured in the PL-PFC, IL-PFC, and ZI and no significant sex differences were found. Data show mean # of both TRAP+ and C-FOS+ cells per mm2 ± SEM. The value of data points not fitting on the axis are labelled in italics and are not statistical outliers.

3.5. Sexually dimorphic activation in the IL-PFC during fear recall and preferential reactivation of IL-PFC during extinction recall

The percentage of all cells active during fear recall or extinction recall (C-FOS+) that were newly active (C-FOS+ and TRAP−) was analyzed as a function of group and sex (Fig. 5A-C). A two-way ANOVA revealed no main effect of group (p > 0.05), no main effect of sex (p > 0.05), but a significant interaction between group and sex (F(1, 27) = 11.162, p < 0.01) in the IL-PFC. Tukey post-hoc testing revealed that female context mice had significantly more newly active cells than male context mice (p < 0.01) or female extinction mice (p = 0.01). There was no main effect of, or interaction between, group and sex in the PL-PFC and ZI (all p > 0.05).

Fig. 5.

Fig. 5.

Open in a new tab

Sex differences in patterns of ensemble and engram activation during recall in IL-PFC, but not PL-PFC and ZI. (A-C) The percentage of newly active cells (TRAP- and C-FOS+ divided by all C-FOS+) was analyzed as a function of group and sex. Comparing % newly active by group and sex in the IL_PFC revealed no main effect of group (F(1, 25) = 2.105, p > 0.05), no main effect of sex (F(1, 25) = 2.767, p > 0.05), but a significant interaction between group and sex (F(1, 25) = 11.162, p < 0.01). Posthoc testing revealed that female context mice had a greater % of newly active cells compared with male context mice (p < 0.01) or female extinction mice (p = 0.01) in the IL-PFC. There was no main effect of, or interaction between, group and sex in the PL-PFC and ZI (all p > 0.05). Data show mean % newly active ± SEM. (D-F) The percentage of reactivated cells (TRAP+ and C-FOS+ cells divided by all TRAP+ cells) was analyzed as a function of group and sex. Comparing % reactivation by group and sex revealed a main effect of group in the IL-PFC (F(1, 25) = 4.837, p = 0.04), no main effect of sex (F(1, 25) = 0.736, p > 0.05), and no interaction between group and sex (F(1, 25) = 0.02, p > 0.05). Collapsing across sex revealed that the extinction group had a significantly higher % reactivation than the context exposure group (W = 54, p = 0.03) (Supplemental Fig. 2). There was no main effect of, or interaction between, group and sex in the PL-PFC and ZI (all p > 0.05). Data show mean % reactivation ± SEM.

The percentage of all cells labelled during context exposure or extinction learning (TRAP+) that were reactivated (TRAP+ and C-FOS+) was analyzed as a function of group and sex (Fig. 5D-F). A two-way ANOVA revealed a main effect of group in the IL-PFC (F(1, 25) = 4.837, p = 0.04), no main effect of sex (F(1, 25) = 0.736, p > 0.05), and no interaction between group and sex (F(1, 25) = 0.002, p > 0.05). Collapsing across sex and comparing engram reactivation in the IL-PFC between the groups revealed that there was significantly more reactivation in the extinction group than the context exposure group (W = 54, p = 0.03) (Supplemental Fig. 2). There was no main effect of, or interaction between, group and sex in the PL-PFC and ZI (all p > 0.05). Similarly, a two-way ANOVA comparing cellular densities by group and sex (Supplemental Fig. 3) revealed a significant effect of group in the density of overlapping (TRAP+ and C-FOS+) cells in the IL-PFC (F(1, 25) = 4.241, p = 0.05). Collapsing across sex (Supplemental Fig. 2) showed a significantly higher density in the extinction group (W = 154, p = 0.02).

3.6. Sexually dimorphic patterns of correlation between PFC and ZI ensembles

To assess the relationship between ensembles in the PFC and the ZI, both in density and in activation patterns, a Pearson’s correlation was run with a focus on significant correlations between the PFC and ZI for the same cellular measure (Supplemental Fig. 4). Notably, females in the context group showed significant positive correlations between IL-PFC ensembles and ZI ensembles in both density of overlapping cells (r(5) = 0.91, p = 0.04) and in reactivation of TRAP+ cells (r(5) = 0.94, p = 0.03). All other significant correlations were between measures in the IL-PFC and PL-PFC. A full list of correlation coefficients, p values, and false discovery rate corrected p values (q values) for all comparisons is available in the supplemental data.

4. Discussion

Ensembles of cells that are active during learning can come to be incorporated into memory engrams. Engrams are distinct cellular representations of specific memories that may be reactivated to support future memory recall. To address potential sex differences in the size and reactivation of ensembles across various stages of learning and memory, we used activity-dependent cell tagging using TRAP2; floxed-tdTomato mice to determine the density and reactivation patterns of cells during context exposure, extinction training, fear recall, and extinction recall in male and female mice. Following auditory fear conditioning in Context A, we TRAPped cells active during extinction training or during context exposure in the absence of extinction training, both, in Context B. We then stained for C-FOS following 5 tone presentations in Context B to identify active cells during extinction recall in the group that underwent extinction training and during fear recall in the group that underwent context exposure without extinction training. Given our laboratory’s interest in the influence of the ZI on learning and memory (Venkataraman & Dias, 2023; Venkataraman et al., 2019, 2021), the role of the PL- and IL-PFC in consolidation and recall of memory (Giustino & Maren, 2015; Orsini & Maren, 2012; Sotres-Bayon & Quirk, 2010), and published literature from colleagues demonstrating the role of PFC → ZI communication in extinction learning and recall (Chou et al., 2018; Zhao et al., 2020), we chose to focus our attention on the PL-PFC, IL-PFC and the ZI. Our data suggest that there are sex differences in the density of cells active in the IL-PFC during exposure to a context, in the percent of newly active cells during fear recall, and that there are sex-specific correlations between ensemble density and reactivation in the IL-PFC and ZI. Furthermore, we found increased density of overlapping cells and reactivation of IL-PFC ensembles during extinction recall, emphasizing the well-established importance of this brain region in extinction learning and recall.

Sex differences in context- and cue-related learning and memory have been reported in humans, but efforts to model these differences preclinically in rodents have produced mixed results (Lebron-Milad & Milad, 2012; Milad et al., 2009, 2010; Velasco et al., 2019). Here, we demonstrate no significant sex differences in behavior during any of the learning and recall tests after auditory fear conditioning in mice. However, our results demonstrate reliable fear acquisition to the CS+ during fear conditioning, successful within-session extinction learning, successful extinction recall, and high levels of fear recall in the context group. This provides confidence that our investigation of cellular activity successfully targeted cells active during the intended phases of learning and recall, and is in line with reports demonstrating sex differences in neurobiology that are not directly linked to sex differences in behavior (Fleischer & Frick, 2023; Gall, Le, & Lynch, 2023). For example, a recent study used whole brain clearing to visualize TRAPped cells following recall of contextual fear and found that while male and female mice did not differ in their behavior, there was strong sexual dimorphism in the activation of different regions that correlated with fear behavior (Franceschini et al., 2023). Therefore, we posited that similar behavioral outputs in male and female mice in our experiments does not preclude the possibility of finding sex differences in the density of cells active at the time of learning and in ensemble reactivation at various epochs of learning and memory. We also investigated potential correlations between all of the ensemble measures we report and freezing during different phases of learning and memory, finding no significant correlations (Supplemental data) and suggesting that while there are sex differences in ensemble dynamics and density, they could well occur independently of sex differences in fear behavior. This leaves open the possibility that the sex differences we observe in IL-PFC activation patterns could underly behaviors relevant to these tests that were not directly measured. For example, the PFC has been shown to support arousal, guide movement and action temporally, and provide higher-order processing of auditory information – all functions that would be employed during exploration of new environments and attending to auditory cues (Hockley & Malmierca, 2024; Mair, Francoeur, Krell, & Gibson, 2022; Mashour, Pal, & Brown, 2022; Zhang, Weber, & Narayanan, 2021).

Much of our current understanding of memory engrams comes from literature focused on male mice and the use of contextual fear conditioning protocols. This pairing of context and fear does not allow for the dissociation of ensembles and engrams supporting fear- or extinction-related memory from those solely representing contextual information. Our experimental design allowed us to investigate sex differences in the activity of cellular ensembles during context exposure, independent of any CS+ presentations, and the reactivation of these ensembles in the same context during fear recall. We found that males have a denser ensemble of active cells in the IL-PFC during context exposure, but no differences in the PL-PFC or ZI. One interpretation of these data is a greater role of specifically the IL-PFC of male mice in responding to contextual information. In support of this, other studies have demonstrated significant differences between the sexes in the processing of contextual and spatial information, as well as its supporting neurobiology (Fleischer & Frick, 2023; Keiser et al., 2017; Yagi, Lee, Truter, & Galea, 2022). These data could potentially also demonstrate reduced habituation to Context B in male mice. Context ensembles were TRAPped during the second exposure to the context and as animals are repeatedly exposed to contexts they habituate and reduce responding both behaviorally and cellularly (Merchie & Gomot, 2023; Wilson & Linster, 2008). Our results could indicate a dampening of this process in male mice.

We also found that females in the context group had significantly more cells in the IL-PFC that were newly activated (exclusively C-FOS+) during fear recall compared with context males or with extinction females. These results indicate that a new population of cells in the IL-PFC that had not been previously TRAPped are activated during fear recall to a greater degree in female mice than in male mice or in female mice that underwent extinction training. One population of cells that are not directly captured in these studies, but are highly relevant during fear and extinction recall, are ensembles initially active at the time of fear conditioning. These cells could potentially come online during fear recall to a greater degree in females compared with males or with females that underwent extinction training. These results could also demonstrate, more broadly, a heightening of reactivity in the IL-PFC of context females to the fear cue, compared with males and with females that underwent extinction training. Similarly, previous studies have also demonstrated substantial sex differences in the recruitment and reactivity of different brain regions during fear memory recall (Bauer, 2023; Keiser et al., 2017; Reppucci & Petrovich, 2018). Future work directly tagging cells during fear conditioning and during exposure to the conditioning context would be necessary to better understand the dynamics of fear conditioning ensembles during extinction learning and recall.

Our data also demonstrate significantly denser overlapping cells and greater reactivation of TRAPped IL-PFC cells in the extinction group compared with the context group, independent of sex. This result fits within our understanding of the IL-PFC, given that it is well established as a key mediator of extinction learning and recall (Giustino & Maren, 2015) and extinction engrams have been previously identified in the mPFC (Gu et al., 2022). Extinction learning likely recruits a broader ensemble of cells compared with context exposure alone considering that extinction consists of context exposure as well as repeated exposures to the CS+. This ensemble has overlapping elements of contextual and cue information and is thus likely reactivated to a greater extent during future combined context and cue exposure.

These analyses within the PL-PFC, IL-PFC, and ZI are important to determine how these regions independently represent learning and memory events, but in reality engrams are spread across many brain regions functioning together to support memory (Gu et al., 2022; Roy et al., 2022). Given the importance of PFC → ZI connectivity in fear learning and memory (Chou et al., 2018; Zhao et al., 2020) we chose to investigate how ensembles in these regions may act in concert with each other during these tests. Looking at correlations in ensemble activation, we found significant correlations in the female context group between the IL-PFC and ZI in terms of density of overlapping cells and reactivation. These data suggest that there may be some sex specific coordination between the IL-PFC and ZI during contextual and fear recall. These results are in line with recent work that similarly demonstrated distinct patterns of functional connectivity during fear recall between males and females that evolves in a sexually dimorphic fashion (Franceschini et al., 2023). Future work would benefit from combining a brain-wide analyses of regions and functional connectivity with behavioral protocols that allow for the dissection of different components of memory, including context.

In conclusion, our study presents a novel characterization of sex differences in IL-PFC ensembles active in response to contextual versus cue information and shows correlated activity between these IL-PFC ensembles and ZI ensembles specifically in females. Our work also highlights preferential reactivation of the IL-PFC during extinction recall. This work begins to address the need for a better understanding of sex differences in cellular ensembles supporting learning and memory. It also points towards outstanding questions in the field including how different elements of memory are represented cellularly and how coordination of diverse brain regions supports memory function. Addressing these questions and building on this body of work presents a promising opportunity to better understand the neurobiological underpinnings of sex differences in learning and memory related dimensions of psychiatric disorders like PTSD.

Supplementary Material

1

Acknowledgments

We are grateful for animal husbandry and care provided by the veterinarian and staff in the Animal Care Facility at The Saban Research Institute (TSRI). BGD’s research is supported by the National Institutes of Health (Grant Nos. R01MH134873 and R56MH128427), the Department of Pediatrics at the Keck School of Medicine of the University of Southern California, the Developmental Neuroscience and Neurogenetics Program at The Saban Research Institute, and the Child and Brain Development Program of the Canadian Institute for Advanced Research. WWT and LK received funding from the TSRI Pre-doctoral Intramural Award. The authors report no biomedical financial interests or potential conflicts of interest.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.nlm.2025.108128.

Footnotes

CRediT authorship contribution statement

William W. Taylor: Writing – review & editing, Writing – original draft, Visualization, Project administration, Methodology, Investigation, Formal analysis, Conceptualization. Vienna Gao: Visualization, Investigation, Formal analysis. Laura Korobkova: Writing – original draft, Methodology, Formal analysis. Brian G. Dias: Writing – review & editing, Writing – original draft, Supervision, Resources, Project administration, Funding acquisition, Conceptualization.

Data availability

Data will be deposited to Mendeley Data doi: 10.17632/ngj26msn46.1

References

  1. Bankhead P, Loughrey MB, Fernandez JA, Dombrowski Y, McArt DG, Dunne PD, et al. (2017). QuPath: Open source software for digital pathology image analysis. Science Reports, 7, 16878. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bauer EP (2023). Sex differences in fear responses: Neural circuits. Neuropharmacology, 222, Article 109298. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Chou XL, Wang X, Zhang ZG, Shen L, Zingg B, Huang J, et al. (2018). Inhibitory gain modulation of defense behaviors by zona incerta. Nature Communications, 9, 1151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Cowansage KK, Shuman T, Dillingham BC, Chang A, Golshani P, & Mayford M (2014). Direct reactivation of a coherent neocortical memory of context. Neuron, 84, 432–441. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. DeNardo LA, Liu CD, Allen WE, Adams EL, Friedmann D, Fu L, et al. (2019). Temporal evolution of cortical ensembles promoting remote memory retrieval. Nature Neuroscience, 22, 460–469. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Fleischer AW, & Frick KM (2023). New perspectives on sex differences in learning and memory. Trends in Endocrinology and Metabolism: TEM, 34, 526–538. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Franceschini A, Mazzamuto G, Checcucci C, Chicchi L, Fanelli D, Costantini I, et al. (2023). Brain-wide neuron quantification toolkit reveals strong sexual dimorphism in the evolution of fear memory. Cell Reports, 42, Article 112908. [DOI] [PubMed] [Google Scholar]
  8. Frankland PW, Josselyn SA, & Kohler S (2019). The neurobiological foundation of memory retrieval. Nature Neuroscience, 22, 1576–1585. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Frankland PW, Josselyn SA, & Kohler S (2024). Engrams. Current Biology, 34, R559–R561. [DOI] [PubMed] [Google Scholar]
  10. Gall CM, Le AA, & Lynch G (2023). Sex differences in synaptic plasticity underlying learning. Journal of Neuroscience Research, 101, 764–782. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Giustino TF, & Maren S (2015). The role of the medial prefrontal cortex in the conditioning and extinction of fear. Frontiers in Behavioral Neuroscience, 9, 298. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Gu X, Wu YJ, Zhang Z, Zhu JJ, Wu XR, Wang Q, et al. (2022). Dynamic tripartite construct of interregional engram circuits underlies forgetting of extinction memory. Molecular Psychiatry, 27, 4077–4091. [DOI] [PubMed] [Google Scholar]
  13. Guskjolen A, & Cembrowski MS (2023). Engram neurons: Encoding, consolidation, retrieval, and forgetting of memory. Molecular Psychiatry, 28, 3207–3219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Han JH, Kushner SA, Yiu AP, Hsiang HL, Buch T, Waisman A, et al. (2009). Selective erasure of a fear memory. Science, 323, 1492–1496. [DOI] [PubMed] [Google Scholar]
  15. Hockley A, & Malmierca MS (2024). Auditory processing control by the medial prefrontal cortex: A review of the rodent functional organisation. Hearing Research, 443, Article 108954. [DOI] [PubMed] [Google Scholar]
  16. Hsiang HL, Epp JR, van den Oever MC, Yan C, Rashid AJ, Insel N, et al. (2014). Manipulating a “cocaine engram” in mice. The Journal of Neuroscience, 34, 14115–14127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Josselyn SA, & Tonegawa S (2020). Memory engrams: Recalling the past and imagining the future. Science, 367. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Keiser AA, Turnbull LM, Darian MA, Feldman DE, Song I, & Tronson NC (2017). Sex differences in context fear generalization and recruitment of hippocampus and amygdala during retrieval. Neuropsychopharmacology, 42, 397–407. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Lacagnina AF, Brockway ET, Crovetti CR, Shue F, McCarty MJ, Sattler KP, et al. (2019). Distinct hippocampal engrams control extinction and relapse of fear memory. Nature Neuroscience, 22, 753–761. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Lebron-Milad K, & Milad MR (2012). Sex differences, gonadal hormones and the fear extinction network: Implications for anxiety disorders. Biol Mood Anxiety Disord, 2, 3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Liu X, Ramirez S, Pang PT, Puryear CB, Govindarajan A, Deisseroth K, et al. (2012). Optogenetic stimulation of a hippocampal engram activates fear memory recall. Nature, 484, 381–385. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Mair RG, Francoeur MJ, Krell EM, & Gibson BM (2022). Where actions meet outcomes: medial prefrontal cortex, central thalamus, and the basal ganglia. Frontiers in Behavioral Neuroscience, 16, Article 928610. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Mashour GA, Pal D, & Brown EN (2022). Prefrontal cortex as a key node in arousal circuitry. Trends in Neurosciences, 45, 722–732. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Merchie A, & Gomot M (2023). Habituation, adaptation and prediction processes in neurodevelopmental disorders: A Comprehensive Review. Brain Sciences, 13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Milad MR, Pitman RK, Ellis CB, Gold AL, Shin LM, Lasko NB, et al. (2009). Neurobiological basis of failure to recall extinction memory in posttraumatic stress disorder. Biological Psychiatry, 66, 1075–1082. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Milad MR, Zeidan MA, Contero A, Pitman RK, Klibanski A, Rauch SL, et al. (2010). The influence of gonadal hormones on conditioned fear extinction in healthy humans. Neuroscience, 168, 652–658. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Orsini CA, & Maren S (2012). Neural and cellular mechanisms of fear and extinction memory formation. Neuroscience and Biobehavioral Reviews, 36, 1773–1802. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Ramikie TS, & Ressler KJ (2018). Mechanisms of sex differences in fear and posttraumatic stress disorder. Biological Psychiatry, 83, 876–885. [DOI] [PubMed] [Google Scholar]
  29. Redondo RL, Kim J, Arons AL, Ramirez S, Liu X, & Tonegawa S (2014). Bidirectional switch of the valence associated with a hippocampal contextual memory engram. Nature, 513, 426–430. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Reppucci CJ, & Petrovich GD (2018). Neural substrates of fear-induced hypophagia in male and female rats. Brain Structure & Function, 223, 2925–2947. [DOI] [PubMed] [Google Scholar]
  31. Roy DS, Park YG, Kim ME, Zhang Y, Ogawa SK, DiNapoli N, et al. (2022). Brain-wide mapping reveals that engrams for a single memory are distributed across multiple brain regions. Nature Communications, 13, 1799. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Sotres-Bayon F, & Quirk GJ (2010). Prefrontal control of fear: More than just extinction. Current Opinion in Neurobiology, 20, 231–235. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Tanaka KZ, Pevzner A, Hamidi AB, Nakazawa Y, Graham J, & Wiltgen BJ (2014). Cortical representations are reinstated by the hippocampus during memory retrieval. Neuron, 84, 347–354. [DOI] [PubMed] [Google Scholar]
  34. Velasco ER, Florido A, Milad MR, & Andero R (2019). Sex differences in fear extinction. Neuroscience and Biobehavioral Reviews, 103, 81–108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Venkataraman A, Brody N, Reddi P, Guo J, Gordon Rainnie D, & Dias BG (2019). Modulation of fear generalization by the zona incerta. Proceedings of the National Academy of Sciences of the United States of America, 116, 9072–9077. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Venkataraman A, & Dias BG (2023). Expanding the canon: An inclusive neurobiology of thalamic and subthalamic fear circuits. Neuropharmacology, 226, Article 109380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Venkataraman A, Hunter SC, Dhinojwala M, Ghebrezadik D, Guo J, Inoue K, et al. (2021). Incerto-thalamic modulation of fear via GABA and dopamine. Neuropsychopharmacology, 46, 1658–1668. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Vetere G, Tran LM, Moberg S, Steadman PE, Restivo L, Morrison FG, et al. (2019). Memory formation in the absence of experience. Nature Neuroscience, 22, 933–940. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Wilson DA, & Linster C (2008). Neurobiology of a simple memory. Journal of Neurophysiology, 100, 2–7. [DOI] [PubMed] [Google Scholar]
  40. Yagi S, Lee A, Truter N, & Galea LAM (2022). Sex differences in contextual pattern separation, neurogenesis, and functional connectivity within the limbic system. Biology of Sex Differences, 13, 42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Zhang Q, Weber MA, & Narayanan NS (2021). Medial prefrontal cortex and the temporal control of action. International Review of Neurobiology, 158, 421–441. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Zhao P, Zhao M, Wang H, Jiang T, Jia X, Tian J, et al. (2020). Long-range inputome of cortical neurons containing corticotropin-releasing hormone. Science Reports, 10, 12209. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS2133101-supplement-1.docx (1.1MB, docx)

Data Availability Statement

Data will be deposited to Mendeley Data doi: 10.17632/ngj26msn46.1

RESOURCES