Advertisement

Sex-Specific Neuroanatomical Correlates of Fear Expression in Prefrontal-Amygdala Circuits

Published:November 28, 2014DOI:https://doi.org/10.1016/j.biopsych.2014.11.014

      Abstract

      Background

      The neural projections from the infralimbic region of the prefrontal cortex to the amygdala are important for the maintenance of conditioned fear extinction. Neurons in this pathway exhibit a unique pattern of structural plasticity that is sex-dependent, but the relationship between the morphologic characteristics of these neurons and successful extinction in male and female subjects is unknown.

      Methods

      Using classic cued fear conditioning and an extinction paradigm in large cohorts of male and female rats, we identified subpopulations of both sexes that exhibited high (HF) or low (LF) levels of freezing on an extinction retrieval test, representing failed or successful extinction maintenance, respectively. We combined retrograde tracing with fluorescent intracellular microinjections to perform three-dimensional reconstructions of infralimbic neurons that project to the basolateral amygdala in these groups.

      Results

      The HF and LF male rats exhibited neuroanatomical distinctions that were not observed in HF or LF female rats. A retrospective analysis of behavior during fear conditioning and extinction revealed that despite no overall sex differences in freezing behavior, HF and LF phenotypes emerged in male rats during extinction and in female rats during fear conditioning, which does not involve infralimbic–basolateral amygdala neurons.

      Conclusions

      Our results suggest that the neural processes underlying successful or failed extinction maintenance may be sex-specific. These findings are relevant not only to future basic research on sex differences in fear conditioning and extinction but also to exposure-based clinical therapies, which are similar in premise to fear extinction and which are primarily used to treat disorders that are more common in women than in men.

      Keywords

      In humans and animals, extinction of a conditioned fear response depends on communication between the medial prefrontal cortex (mPFC) and the amygdala (
      • Quirk G.J.
      • Likhtik E.
      • Pelletier J.G.
      • Paré D.
      Stimulation of medial prefrontal cortex decreases the responsiveness of central amygdala output neurons.
      ,
      • Milad M.R.
      • Quirk G.J.
      Neurons in medial prefrontal cortex signal memory for fear extinction.
      ). Disrupted connectivity of this circuit is thought to underlie some symptoms of mental illnesses such as posttraumatic stress disorder (PTSD); in particular, unwanted or inappropriate fear responses may result from a failure of the mPFC to regulate amygdala activity (
      • Stevens J.S.
      • Jovanovic T.
      • Fani N.
      • Ely T.D.
      • Glover E.M.
      • Bradley B.
      • Ressler K.J.
      Disrupted amygdala-prefrontal functional connectivity in civilian women with posttraumatic stress disorder.
      ,
      • Milad M.R.
      • Pitman R.K.
      • Ellis C.B.
      • Gold A.L.
      • Shin L.M.
      • Lasko N.B.
      • et al.
      Neurobiological basis of failure to recall extinction memory in posttraumatic stress disorder.
      ). Although PTSD is twice as common in women as in men (
      • Breslau N.
      The epidemiology of trauma, PTSD, and other posttrauma disorders.
      ), it is unknown whether sex differences in mPFC-amygdala circuit function underlie this discrepancy.
      In rodents, efferents from the infralimbic (IL) region of the mPFC activate a microcircuit within the basolateral amygdala (BLA) that results in suppression of fear response after extinction (
      • Herry C.
      • Ciocchi S.
      • Senn V.
      • Demmou L.
      • Müller C.
      • Lüthi A.
      Switching on and off fear by distinct neuronal circuits.
      ). We have previously shown that neurons in this pathway can undergo unique patterns of dendritic remodeling in response to stress in male and female animals (
      • Shansky R.M.
      • Hamo C.
      • Hof P.R.
      • McEwen B.S.
      • Morrison J.H.
      Stress-induced dendritic remodeling in the prefrontal cortex is circuit specific.
      ,
      • Shansky R.M.
      • Hamo C.
      • Hof P.R.
      • Lou W.
      • McEwen B.S.
      • Morrison J.H.
      Estrogen promotes stress sensitivity in a prefrontal cortex-amygdala pathway.
      ), suggesting that structural plasticity in IL-BLA neurons is sex-specific. However, a direct relationship between IL morphology and extinction success has not been demonstrated in a circuit-based manner, let alone in both male and female animals. Of all fear conditioning and extinction research, <2% has been conducted in female animals (
      • Lebron-Milad K.
      • Milad M.R.
      Sex differences, gonadal hormones and the fear extinction network: Implications for anxiety disorders.
      ), and our understanding of the mechanisms and neural processes that mediate extinction in the female brain is comparatively limited. However, multiple reports of sexually dimorphic plasticity in the mPFC (
      • Farrell M.R.
      • Sengelaub D.R.
      • Wellman C.L.
      Sex differences and chronic stress effects on the neural circuitry underlying fear conditioning and extinction.
      ,
      • Shansky R.M.
      Estrogen, stress and the brain: Progress toward unraveling gender discrepancies in major depressive disorder.
      ) led us to hypothesize that the neuroanatomical features associated with successful versus failed extinction retrieval may be distinct in male and female subjects.
      Our primary objective for the present study was to define sex-specific structure-function relationships in IL-BLA neurons. We took advantage of naturally occurring behavioral variability in male and female rats, identifying subpopulations of animals that exhibited high or low levels of freezing during an extinction retrieval test. We used a combination of retrograde tracing and three-dimensional neuronal reconstructions to create detailed structural profiles of IL-BLA projection neurons in these groups. We found that high freezing (HF) and low freezing (LF) male rats, but not female rats, exhibited distinct morphologic features. Our data suggest that this discrepancy may be due to sex differences in the behavioral trajectories that lead to successful extinction maintenance.

      Methods And Materials

      Subjects

      Young adult (8–10 weeks old) male (n = 58) and female (n = 57) Sprague Dawley rats were individually housed in the Nightingale Animal Facility at Northeastern University on a 12:12 light/dark cycle with access to food and water ad libitum. All procedures were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Northeastern University Institutional Animal Care and Use Committee.

      Estrous Cycle Monitoring

      Female rats were vaginally swabbed daily to ensure normal estrous cycling. Collected cells were smeared on a microscope slide, stained with Nissl, and examined with a light microscope for cytology.

      Surgery

      All animals underwent aseptic stereotactic surgery as described elsewhere (
      • Shansky R.M.
      • Hamo C.
      • Hof P.R.
      • McEwen B.S.
      • Morrison J.H.
      Stress-induced dendritic remodeling in the prefrontal cortex is circuit specific.
      ). Animals were anesthetized with an intraperitoneal injection of a ketamine (90 mg/kg) and xylazine (4 mg/kg) cocktail. Animals’ heads were shaved and secured into a stereotactic apparatus (Stoelting Co, Wood Dale, Illinois), and an incision was made to reveal the skull. When bregma was clearly visible, a burr hole was drilled above the BLA (−3.0 mm anteroposterior ± 5.0 mm mediolateral) (
      • Paxinos G.
      • Watson C.
      The Rat Brain in Stereotaxic Coordinates.
      ) and a 1-µL syringe attached to a stereotactic arm was lowered to 8.0 mm dorsoventral. After allowing tissue to settle for 2 min, an infusion pump (Harvard Apparatus, Holliston, Massachusetts) delivered .2 μL 5% Fluoro-Gold (Fluorochrome, LLC, Denver, Colorado) at .02 μL/min. The syringe was left in place for 10 min and then slowly removed from the brain, and the incision was sealed with Vetbond surgical glue (3M Animal Care Products, St Paul, MN). Animals were kept in cages atop heated recovery pads until righting reflexes returned. For 3 days after surgery, animals received subcutaneous injections of .3 mL buprenorphine, and were monitored for healthy eating and drinking habits. Behavioral testing began 1 week after surgery.

      Behavioral Testing

      Apparatus and Stimuli

      Rats underwent habituation, fear conditioning, and fear extinction as described elsewhere (
      • Gruene T.M.
      • Lipps J.
      • Rey C.D.
      • Bouck A.
      • Shansky R.M.
      Heat exposure in female rats elicits abnormal fear expression and cellular changes in prefrontal cortex and hippocampus.
      ) in one of four identical chambers constructed of aluminum and Plexiglas walls (Rat Test Cage; Coulbourn Instruments, Allentown, Pennsylvania) with metal stainless steel rod flooring that was attached to a shock generator (Model H13-15; Coulbourn Instruments). The chambers were lit with a single house light, and each chamber was enclosed within a sound-isolation cubicle (Model H10-24A; Coulbourn Instruments). An infrared digital camera allowed videotaping during behavioral procedures. Chamber grid floors, trays, and walls were thoroughly cleaned with water and dried between sessions. Rats were allowed to explore the chamber freely for 4 min before tone presentation on each day began.

      Fear Conditioning Procedure

      All rats were exposed to five tone (conditioned stimulus) presentations (habituation), followed by seven conditioning trials (conditioned stimulus–unconditioned stimulus pairings) on day 1. The conditioned stimulus was a 30-sec, 5-kHz, 80-dB sound pressure level sine wave tone, which coterminated with a .5-sec, .7 mA foot-shock unconditioned stimulus during fear conditioning. Mean intertrial interval was 4 min (range, 2–6 min) throughout habituation and fear conditioning. Fear behavior was measured using freezing, defined as the cessation of all movement with the exception of respiration-related movement and nonawake or rest body posture. Freezing was continuously recorded during the conditioning session and analyzed using FreezeFrame Software (Coulbourn Instruments). Minimum bout was set at 2 sec. After conditioning, rats were returned to their home cages.

      Extinction Procedures

      Freezing was recorded continuously during the extinction training (20 tone alone trials, day 2) and retrieval sessions (3 tone alone trials, day 3). Extinction training and testing took place in the same chamber as fear conditioning but with different contextual cues (floor, light, and odor). Mean intertrial interval was 4 min (range, 2–6 min). To ensure that LF animals were not animals that simply failed to learn the tone-shock association, animals that did not reach criterion for fear conditioning (>40% average freezing on first two extinction tone presentations (
      • Bush D.E.
      • Sotres-Bayon F.
      • LeDoux J.E.
      Individual differences in fear: Isolating fear reactivity and fear recovery phenotypes.
      )) were removed from overall analysis.

      Experimental Groups

      For evaluation of estrous cycle effects, female animals were analyzed according to estrous phase during extinction learning. We and others have previously demonstrated that animals in proestrus (high estrogen levels) during extinction exhibit facilitated extinction retrieval the following day (
      • Milad M.R.
      • Igoe S.A.
      • Lebron-Milad K.
      • Novales J.E.
      Estrous cycle phase and gonadal hormones influence conditioned fear extinction.
      ,
      • Rey C.D.
      • Lipps J.
      • Shansky R.M.
      Dopamine D1 receptor activation rescues extinction impairments in low-estrogen female rats and induces cortical layer-specific activation changes in prefrontal-amygdala circuits.
      ). This effect is specific to extinction learning because freezing in animals in proestrus during extinction retrieval is indistinguishable from freezing in animals in other estrous phases (
      • Milad M.R.
      • Igoe S.A.
      • Lebron-Milad K.
      • Novales J.E.
      Estrous cycle phase and gonadal hormones influence conditioned fear extinction.
      ). To identify HF and LF populations of male and female animals, the animals were sorted by freezing during extinction retrieval. The eight highest freezing and lowest freezing animals in each sex were selected for morphologic and retrospective behavioral analyses. Estrous distribution in selected animals was as follows: HF, three diestrus, two estrus, three metaestrus; LF, one diestrus, one estrus, two metaestrus, four proestrus.

      Euthanasia and Tissue Preparation

      Animals were anesthetized and sacrificed by transcardial perfusion of 4% paraformaldehyde in .1 mol/L phosphate buffer (phosphate-buffered saline, pH 7.4) 24 hours after the completion of extinction retrieval testing. Brains were extracted and post-fixed in paraformaldehyde for 6 hours and then placed in .1% sodium azide in phosphate-buffered saline at 4°C for storage.

      Morphology Analysis

      After proper BLA injection targets were confirmed, 250-μm, IL-containing sections were collected using a vibrating microtome (Leica Microsystems, Inc, Buffalo Grove, Illinois). Fluoro-Gold-positive IL neurons were visualized through a DAPI filter (Zeiss Microscopy, Thornwood, New York) on a Zeiss Axio Examiner A.1 microscope (Zeiss Microscopy). Iontophoretic microinjections of fluorescent dye Lucifer yellow were performed into Fluoro-Gold-positive, IL layer II/III pyramidal neurons using a DC current of 1–6 nA for 5–10 min, until distal processes were filled with dye and no further loading could be observed. Sections were mounted and placed in a coverslip in VECTASHIELD (Vector Laboratories, Burlingame, California), and filled neurons were selected for imaging and analysis based on completeness criteria described previously (
      • Shansky R.M.
      • Hamo C.
      • Hof P.R.
      • McEwen B.S.
      • Morrison J.H.
      Stress-induced dendritic remodeling in the prefrontal cortex is circuit specific.
      ,
      • Shansky R.M.
      • Hamo C.
      • Hof P.R.
      • Lou W.
      • McEwen B.S.
      • Morrison J.H.
      Estrogen promotes stress sensitivity in a prefrontal cortex-amygdala pathway.
      ). Three to six neurons per animal were included in the analysis. For off-line tracing of apical dendrites, multiple Z-stacks of whole neurons were acquired using an Olympus FV1000 confocal microscope (Optical Analysis Corporation, Nashua, New Hampshire) with a 60× lens and a step-size of 1 µm. Montages of all images covering the apical dendrite of one neuron were created and traced using Neurolucida (MBF Bioscience, Williston, Vermont). Apical length, branch number, and Sholl analyses were performed with Neurolucida Explorer (MBF Bioscience). For spine analysis, spine segments from apical dendrites were selected in 50-µm increments from the cell body, and approximately eight segments per neuron were sampled. Z-stacks were acquired using a 100× lens with a zoom of 3.3, NA 1.4, and step size of .08 µm. Raw Z-stacks were deconvolved with AutoQuant (Media Cybernetics, Rockville, Maryland) and analyzed for spine number and shape (thin, stubby, or mushroom) using NeuronStudio software (Computational Neurobiology and Imaging Center, New York, New York) (classification criteria described in detail by Milad et al. (
      • Rodriguez A.
      • Ehlenberger D.B.
      • Dickstein D.L.
      • Hof P.R.
      • Wearne S.L.
      Automated three-dimensional detection and shape classification of dendritic spines from fluorescence microscopy images.
      )), after which they were manually verified. We focused on apical dendrites because basal dendrites have been repeatedly shown to exhibit little or no structural plasticity (
      • Radley J.J.
      • Sisti H.M.
      • Hao J.
      • Rocher A.B.
      • McCall T.
      • Hof P.R.
      • et al.
      Chronic behavioral stress induces apical dendritic reorganization in pyramidal neurons of the medial prefrontal cortex.
      ,
      • Liston C.
      • Miller M.M.
      • Goldwater D.S.
      • Radley J.J.
      • Rocher A.B.
      • Hof P.R.
      • et al.
      Stress-induced alterations in prefrontal cortical dendritic morphology predict selective impairments in perceptual attentional set-shifting.
      ).

      Statistical Analysis

      All statistical analyses were conducted using GraphPad Prism software (GraphPad Software, Inc, La Jolla, California). Behavioral data (percent freezing during tone presentation or in 30-sec period after tone) were analyzed using two-way analysis of variance (ANOVA) corrected for multiple comparisons, with factors of sex or phenotype and trial number for each day of testing. Bonferroni post hoc tests were conducted when appropriate. Dendritic length and spine density (spines/μm) were averaged for each animal, and group means were calculated. For all analyses, significance was set at p < .05. With the exception of the data presented in Figure 1A, we intentionally conducted within-sex comparisons only. We took this approach because the overarching goal of this study was not to demonstrate whole-group sex differences (of which there seem to be very few) but to identify sex-specific patterns of variability.
      Figure thumbnail gr1
      Figure 1(A) Large cohorts of male and female rats did not differ in average freezing levels across fear conditioning, extinction, and extinction retrieval. (B) Female rats that extinguished in proestrus (high estrogen) froze significantly less during extinction and extinction retrieval than female rats that extinguished in estrus, metaestrus, or diestrus (low estrogen). (C) Variability in freezing during extinction retrieval, based on estrous phase during extinction learning. *p < .03, **p < .01 high estrogen vs. low estrogen. EMD, estrus, metaestrus, or diestrus (low estrogen); PRO, proestrus (high estrogen).

      Results

      Overall Sex Differences and Estrous Effects on Fear Conditioning, Extinction, and Extinction Retrieval

      Gonadally intact, young adult male and female rats were subjected to auditory cued fear conditioning, extinction, and extinction retrieval procedures, as described elsewhere (
      • Rey C.D.
      • Lipps J.
      • Shansky R.M.
      Dopamine D1 receptor activation rescues extinction impairments in low-estrogen female rats and induces cortical layer-specific activation changes in prefrontal-amygdala circuits.
      ). On average, we did not observe sex differences in percent freezing during the tone across fear conditioning, extinction, and extinction retrieval (test for main effect of sex, two-way ANOVA for fear conditioning [F1,63 = .01, p = .92] and extinction [F1,63 = 1.46, p = .23]; unpaired t test for extinction retrieval [p = .33]) (Figure 1A), suggesting that, as general populations, male and female animals do not differ in these measures. To examine the influence of the estrous cycle, we also analyzed female animals alone, grouped by estrous phase during extinction learning, as described elsewhere (
      • Milad M.R.
      • Igoe S.A.
      • Lebron-Milad K.
      • Novales J.E.
      Estrous cycle phase and gonadal hormones influence conditioned fear extinction.
      ,
      • Rey C.D.
      • Lipps J.
      • Shansky R.M.
      Dopamine D1 receptor activation rescues extinction impairments in low-estrogen female rats and induces cortical layer-specific activation changes in prefrontal-amygdala circuits.
      )—either proestrus or estrus, metaestrus, diestrus. We found that proestrus animals exhibited significantly less overall freezing during extinction learning (main effect of estrous cycle [F1,29 = 5.46, p < .03]) as well as less freezing during extinction retrieval (p < .01), consistent with previous reports (Figure 1B). Figure 1C shows freezing during extinction retrieval for animals in each estrous phase (grouped by phase during extinction learning), demonstrating variability in behavior in all phases of the estrous cycle. A one-way ANOVA with planned multiple comparisons [F3,27 = 3.1, p < .05] revealed a significant difference between animals in proestrus and diestrus only (adjusted post hoc, p < .03).

      Identification of HF and LF Subpopulations

      We next assessed the range of freezing responses during extinction retrieval in all animals (Figure 2), and we observed broad but comparable variability in both male and female rats. The eight highest (HF) and lowest (LF) freezing male and female animals were selected for morphologic analyses (data points in shaded regions).
      Figure thumbnail gr2
      Figure 2Eight high freezing and eight low freezing phenotypes were selected from male and female cohorts based on freezing during extinction retrieval (data points within shaded regions). Animals in these groups were used for subsequent analysis. HF, high freezing; LF, low freezing.

      Morphologic Analysis of IL-BLA Neurons

      Figure 3A illustrates the retrograde tracing strategy to label BLA-projecting IL neurons, and Figure 3B shows representative localization of Fluoro-Gold in the BLA. Fluoro-Gold-positive layer II/III neurons in the IL region of HF and LF male and female rats were iontophoretically filled with Lucifer yellow (
      • Shansky R.M.
      • Hamo C.
      • Hof P.R.
      • McEwen B.S.
      • Morrison J.H.
      Stress-induced dendritic remodeling in the prefrontal cortex is circuit specific.
      ,
      • Shansky R.M.
      • Hamo C.
      • Hof P.R.
      • Lou W.
      • McEwen B.S.
      • Morrison J.H.
      Estrogen promotes stress sensitivity in a prefrontal cortex-amygdala pathway.
      ), imaged (Figure 3C), and reconstructed in three-dimensional images for analysis of apical dendrite morphology (Figure 3D). Overall, we found HF and LF differences in male rats, but not female rats. Although there were no group differences in branch number in either sex (two-way ANOVA, no main effect of sex [p = .29] or phenotype [p = .61]) (Figure 3E), HF male rats had significantly less dendritic length compared with LF male rats (two-way ANOVA, main effect of phenotype [F1,21 = 16.43, p < .001]; adjusted Bonferroni post hoc, p < .001) (Figure 3F). In contrast, dendritic length was similar between HF and LF female rats. To determine whether any structural differences were localized along the apical dendrite, we performed a Sholl analysis, with 50-µm bins radiating outward from the cell soma. This analysis revealed that HF male rats specifically lack dendritic material at the most distal points of the apical dendrite (two-way ANOVA, main effect of phenotype [F1,10 = 11.2, p < .01]) (Figure 3G), which corresponds to cortical layer I. This layer is the synaptic target for most IL afferents from the bed nucleus of the stria terminalis (
      • Oh S.W.
      • Harris J.A.
      • Ng L.
      • Winslow B.
      • Cain N.
      • Mihalas S.
      • et al.
      A mesoscale connectome of the mouse brain.
      ), which plays a well-documented role in fear and anxiety expression (
      • Haufler D.
      • Nagy F.Z.
      • Pare D.
      Neuronal correlates of fear conditioning in the bed nucleus of the stria terminalis.
      ,
      • Elharrar E.
      • Warhaftig G.
      • Issler O.
      • Sztainberg Y.
      • Dikshtein Y.
      • Zahut R.
      • et al.
      Overexpression of corticotropin-releasing factor receptor type 2 in the bed nucleus of stria terminalis improves posttraumatic stress disorder-like symptoms in a model of incubation of fear.
      ). The HF phenotype in male animals may be due in part to disrupted communication in bed nucleus of the stria terminalis–IL–BLA circuitry.
      Figure thumbnail gr3
      Figure 3Infralimbic–basolateral amygdala dendritic arborization in high freezing and low freezing male and female rats. (A) Illustration of stereotactic surgery to inject retrograde tracer Fluoro-Gold into the basolateral amygdala for future identification of infralimbic–basolateral amygdala neurons. (B) Representative image of Fluoro-Gold spread in the basolateral amygdala. (C) Fluoro-Gold-positive layer II/III IL neurons were iontophoretically filled with Lucifer yellow and imaged in their entirety. (D) Representative Neurolucida tracings of neurons from high freezing and low freezing male and female rats. (E) No group differences in branch number were observed. (F) High freezing male rats had significantly shorter apical dendrite length compared with low freezing male rats. (G) High freezing vs. low freezing differences in male rats were observed at distal points along the apical dendrite. All data points represent mean ± SEM. ***p < .001, ****p < .0001, within-sex high freezing vs. low freezing adjusted post hoc comparisons. HF, high freezing; LF, low freezing.
      We next examined the morphologic features of dendritic spines in IL-BLA neurons of HF versus LF male and female animals. We collected high-magnification Z-stack images of dendrite segments (Figure 4A,C) and analyzed them for spine density, type, and head diameter using NeuronStudio software (Figure 4B,D) (
      • Dumitriu D.
      • Rodriguez A.
      • Morrison J.H.
      High-throughput, detailed, cell-specific neuroanatomy of dendritic spines using microinjection and confocal microscopy.
      ). We again observed HF versus LF distinctions in male rats, but not female rats. Spine density was greater in HF male rats (two-way ANOVA, main effects of spine type [F2,20 = 1185, p < .0001] and phenotype [F1,10 = 14.2, p < .01]) (Figure 4A,B vs. C,D), an effect that was fully accounted for by thin spines and not stubby or mushroom spines (Bonferroni adjusted post-hoc test, p < .0001) (Figure 4E). To assess whether these differences were localized to one area of the apical dendrite, we separately analyzed spine density at proximal or distal branches (<150 µm or >150 µm from cell soma, respectively) (Figure 4F). We found that the difference in spines in HF male rats occurred exclusively on proximal branches (two-way ANOVA, main effect of phenotype [F1,19 = 9.3, p < .01] (Figure 4F), which corresponds to cortical layer II. This layer receives reciprocal inputs from BLA neurons (
      • Little J.P.
      • Carter A.G.
      Synaptic mechanisms underlying strong reciprocal connectivity between the medial prefrontal cortex and basolateral amygdala.
      ), suggesting that alterations in feedback between the IL region and BLA may contribute to HF and LF behavioral distinctions in male animals. Finally, we examined head diameter in mushroom spine populations and again observed significant differences in HF versus LF male rats, but not female rats. Frequency distribution analysis revealed a leftward shift in HF male rats only (Kolmogorov-Smirnoff D = .059, p = .001) (Figure 4G), suggesting that these animals have proportionately more small-headed mushroom spines.
      Figure thumbnail gr4
      Figure 4Infralimbic–basolateral amygdala spine density in high freezing and low freezing male and female rats. Confocal images (A,C) and NeuronStudio renderings (B,D) of dendritic segments in infralimbic–basolateral amygdala neurons of high freezing and low freezing male rats, respectively. (E) Greater density of thin spines was observed in high freezing male rats compared with low freezing male rats, but no high freezing vs. low freezing distinctions were observed in female rats. (F) Thin spine density differences in high freezing vs. low freezing male rats were restricted to proximal dendrites <150 µm from the cell soma. (G) Frequency distribution of mushroom spine head diameter distribution revealed a significant leftward shift in high freezing male rats compared with low freezing male rats but no differences in high freezing vs. low freezing female rats. All data points represent mean ± SEM except (G). *p < .02, **p = .001, ****p < .0001, within-sex high freezing vs. low freezing adjusted post hoc comparisons. HF, high freezing; LF, low freezing.

      Retrospective Behavioral Analyses in HF versus LF Male and Female Rats

      Our morphologic analyses suggest that the relationship between IL-BLA structure and freezing behavior during extinction retrieval is different for male and female animals. To investigate whether this difference was due to preexisting sex differences in HF and LF behavioral patterns, we replotted their freezing levels across all 3 days of testing. We found that the HF and LF phenotypes emerged during distinct learning phases in male and female rats. The HF and LF male rats exhibited comparable levels of freezing during fear conditioning but diverged at the beginning of extinction (two-way ANOVA [F1,14 = 39.1, p < 0001]) and remained separated throughout the course of the session (Figure 5A). In contrast, freezing levels of HF and LF female rats were significantly different during fear conditioning and extinction (test for main effect of phenotype, two-way ANOVA for fear conditioning [F1,14 = 8.6, p = .01] and extinction [F1,14 = 35.3, p < .0001]) (Figure 5B).
      Figure thumbnail gr5
      Figure 5High fear and low fear phenotypes emerge during distinct phases of learning in male and female rats. (A) High freezing and low freezing male rats differed only during extinction. (B) High freezing and low freezing female rats differed during fear conditioning and extinction. All data points represent mean ± SEM. All significant effects are within-sex high freezing vs. low freezing main effects of phenotype. **p = .01, ****p < .0001. HF, high freezing; LF, low freezing.

      Discussion

      To our knowledge, this study represents the first large-scale evaluation of fear conditioning and extinction processes that includes male and female animals, addressing a fundamental problem in basic neuroscience research—specifically, a glaring lack of knowledge about even basic sex differences in this classic model. Collectively, our data demonstrate that the behavioral and neuroanatomical correlates of freezing during extinction retrieval are sex-specific. However, the freezing levels that defined the HF and LF subpopulations during extinction retrieval did not differ between the sexes; the finding that these phenotypes emerged during distinct phases of learning in male and female animals suggests that freezing during extinction retrieval—even at identical levels—may reflect separate neural processes in male and female animals. Consistent with this hypothesis, we observed structural differences in IL-BLA projections in HF versus LF male rats only.
      We recognize that because our neuroanatomical analyses were necessarily conducted after animals had been through fear conditioning, extinction, and extinction retrieval, it is impossible to know for certain whether HF and LF distinctions reflect preexisting differences in morphology or different patterns of structural plasticity in response to the learning and memory experiences. We hypothesize that the answer may be a combination of both. Experience-based changes in dendritic arborization are usually observed only after an extended period of time—for example, after at least 1 week of stress exposure (
      • McEwen B.S.
      • Morrison J.H.
      The brain on stress: Vulnerability and plasticity of the prefrontal cortex over the life course.
      ). We would speculate that the shorter dendritic length in HF male rats is indicative of a preexisting condition that may have predisposed this group to high freezing during extinction. In contrast, dendritic spines can undergo rapid structural changes over the course of fear conditioning and extinction. In particular, dendrites in the frontal association cortex lose thin spines after fear conditioning but then regrow new spines in essentially the same spot after extinction (
      • Lai C.S.W.
      • Franke T.F.
      • Gan W.-B.
      Opposite effects of fear conditioning and extinction on dendritic spine remodelling.
      ). Although this intriguing phenomenon has not yet been demonstrated in the IL region, the localized, increased spine density observed in HF male rats may reflect a disruption of this pruning-regrowth pattern. Specifically, if these neurons fail to undergo fear conditioning–related spine pruning, extinction-related new spines may compete for “synaptic control” over the neuron, which may result in failed extinction retrieval. Finally, the observed shift in mushroom spine head diameter size in HF male rats may be related to consolidation of the extinction memory. In contrast to thin spines, which are more dynamic, mushroom spines are thought to play a role in the stabilization and long-term maintenance of memories (
      • Bourne J.
      • Harris K.M.
      Do thin spines learn to be mushroom spines that remember?.
      ). Learning elicits alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor insertion into the cell membrane of mushroom spines exclusively (
      • Matsuo N.
      • Reijmers L.
      • Mayford M.
      Spine-type-specific recruitment of newly synthesized AMPA receptors with learning.
      ), and the smaller mushroom spines in HF male rats may indicate decreased alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor insertion in response to extinction learning, impeding stronger memory consolidation.
      In our female animals, the lack of observed morphologic differences between HF and LF groups is consistent with the emergence of these phenotypes during fear conditioning, which does not engage the IL region (
      • Sierra-Mercado D.
      • Padilla-Coreano N.
      • Quirk G.J.
      Dissociable roles of prelimbic and infralimbic cortices, ventral hippocampus, and basolateral amygdala in the expression and extinction of conditioned fear.
      ). This surprising finding suggests that freezing during extinction retrieval in female animals may be related to the initial processing of the tone-shock association, rather than consolidation of extinction learning. In contrast to in male HF and LF phenotypes, both HF and LF female animals exhibited comparably low levels of freezing by the completion of extinction. The lack of HF and LF differences in any morphologic measure in female animals suggests that an alternative circuit may drive freezing during extinction retrieval. One intriguing possibility is that the IL region modulates extinction in female animals not through direct projections to the BLA but via connections with the prelimbic (PL) cortex. The PL cortex also projects to the BLA, but in contrast to the IL region, PL activity is related to increased fear expression after fear conditioning (
      • Vidal-Gonzalez I.
      • Vidal-Gonzalez B.
      • Rauch S.L.
      • Quirk G.J.
      Microstimulation reveals opposing influences of prelimbic and infralimbic cortex on the expression of conditioned fear.
      ). In physiologic studies, PL activity has been linked to elevated freezing during extinction in female animals, but not male animals (
      • Fenton G.E.
      • Pollard A.K.
      • Halliday D.M.
      • Mason R.
      • Bredy T.W.
      • Stevenson C.W.
      Persistent prelimbic cortex activity contributes to enhanced learned fear expression in females.
      ), suggesting that the PL cortex may play a distinctly important role in female animals. We recently found that impaired extinction retrieval in female rats was associated with a switch in the balance of IL and PL c-fos expression, such that PL neurons were more active than IL neurons, whereas the reverse was the case in animals exhibiting good extinction retrieval (
      • Gruene T.M.
      • Lipps J.
      • Rey C.D.
      • Bouck A.
      • Shansky R.M.
      Heat exposure in female rats elicits abnormal fear expression and cellular changes in prefrontal cortex and hippocampus.
      ). Freezing behavior in female rats may rely more closely on IL-PL or PL-BLA circuits, whereas IL-BLA projections are less involved. This hypothesis would be consistent with findings that enhanced extinction in female animals is associated with increased c-fos messenger RNA in whole-IL (as opposed to circuit-specific) assays (
      • Zeidan M.A.
      • Igoe S.A.
      • Linnman C.
      • Vitalo A.
      • Levine J.B.
      • Klibanski A.
      • et al.
      Estradiol modulates medial prefrontal cortex and amygdala activity during fear extinction in women and female rats.
      ), and we look forward to exploring the idea further in future experiments.
      It is also important to consider the role that circulating ovarian hormones play in modulating the learning and memory processes involved in fear conditioning and extinction. Milad et al. (
      • Milad M.R.
      • Igoe S.A.
      • Lebron-Milad K.
      • Novales J.E.
      Estrous cycle phase and gonadal hormones influence conditioned fear extinction.
      ,
      • Zeidan M.A.
      • Igoe S.A.
      • Linnman C.
      • Vitalo A.
      • Levine J.B.
      • Klibanski A.
      • et al.
      Estradiol modulates medial prefrontal cortex and amygdala activity during fear extinction in women and female rats.
      ,
      • Milad M.R.
      • Zeidan M.A.
      • Contero A.
      • Pitman R.K.
      • Klibanski A.
      • Rauch S.L.
      • Goldstein J.M.
      The influence of gonadal hormones on conditioned fear extinction in healthy humans.
      ) convincingly demonstrated in humans and animals that estradiol—administered or circulating at high levels at the time of extinction learning—leads to reduced fear responses during an extinction retrieval test. Administration of an estrogen receptor agonist can mimic these effects (
      • Graham B.M.
      • Milad M.R.
      Blockade of estrogen by hormonal contraceptives impairs fear extinction in female rats and women.
      ), suggesting that estradiol signaling mechanisms may enhance consolidation of the extinction memory. Our findings in the present study support this idea; female rats in proestrus during extinction learning froze significantly less during extinction retrieval than animals that had been in estrus, metaestrus, or diestrus. We also observed this effect in a recent study (
      • Rey C.D.
      • Lipps J.
      • Shansky R.M.
      Dopamine D1 receptor activation rescues extinction impairments in low-estrogen female rats and induces cortical layer-specific activation changes in prefrontal-amygdala circuits.
      ). However, in the present study, animals in each phase exhibited a range of freezing responses during extinction retrieval (Figure 2C), suggesting that success or failure was not solely determined by the estrous cycle. A critical step in better understanding fear conditioning and extinction processes in female animals will be the identification of other hormone and neurotransmitter systems that interact with estrogen to modulate learning and memory.
      Extinction retrieval is widely considered to be a test of fear suppression—the ability of an animal to use the information it learned during extinction to inhibit actively a conditioned response (
      • Quirk G.J.
      • Mueller D.
      Neural mechanisms of extinction learning and retrieval.
      ). Studies in male rodents show that on a physiologic level, IL activation results in decreased amygdala output and decreased freezing (
      • Sierra-Mercado D.
      • Padilla-Coreano N.
      • Quirk G.J.
      Dissociable roles of prelimbic and infralimbic cortices, ventral hippocampus, and basolateral amygdala in the expression and extinction of conditioned fear.
      ,
      • Vidal-Gonzalez I.
      • Vidal-Gonzalez B.
      • Rauch S.L.
      • Quirk G.J.
      Microstimulation reveals opposing influences of prelimbic and infralimbic cortex on the expression of conditioned fear.
      ), and HF is commonly interpreted as a failure of the IL region to regulate amygdala-driven behavior (
      • Sepulveda-Orengo M.T.
      • Lopez A.V.
      • Soler-Cedeño O.
      • Porter J.T.
      Fear extinction induces mGluR5-mediated synaptic and intrinsic plasticity in infralimbic neurons.
      ). Our findings in male animals support this narrative: HF male rats did not differ from LF male rats in acquisition of the tone-shock association but exhibited impaired extinction and extinction retrieval, suggesting that the mechanisms that underlie the HF and LF phenotypes are primarily related to extinction processes. The HF male rats exhibited morphologic alterations in IL-BLA projection neurons that are consistent with compromised extinction maintenance: shorter dendrites extending into cortical layer I, localized increases in thin spines, and a population of mushroom spines characterized by smaller head diameters.
      In contrast, our results in female animals suggest a more complex story. If women are more prone to disorders such as PTSD, one might intuitively predict that a greater proportion of our female animals would exhibit failed or impaired fear extinction compared with male animals. We were surprised to find a lack of overall sex differences in these measures. This finding may mean that the parameters of our behavioral paradigm do not appropriately mimic the conditions that lead to sex differences in trauma effects in humans or that a more severe stressor is necessary in addition to the foot-shock. However, the lack of sex differences in fear conditioning and extinction is consistent with previous reports from our laboratory and others (
      • Rey C.D.
      • Lipps J.
      • Shansky R.M.
      Dopamine D1 receptor activation rescues extinction impairments in low-estrogen female rats and induces cortical layer-specific activation changes in prefrontal-amygdala circuits.
      ,
      • Maren S.
      • De Oca B.
      • Fanselow M.S.
      Sex differences in hippocampal long-term potentiation (LTP) and Pavlovian fear conditioning in rats: positive correlation between LTP and contextual learning.
      ,
      • Baran S.E.
      • Armstrong C.E.
      • Niren D.C.
      • Hanna J.J.
      • Conrad C.D.
      Chronic stress and sex differences on the recall of fear conditioning and extinction.
      ,
      • Lebrón-Milad K.
      • Tsareva A.
      • Ahmed N.
      • Milad M.R.
      Sex differences and estrous cycle in female rats interact with the effects of fluoxetine treatment on fear extinction.
      ). This seeming paradox is discussed in greater detail in our recent review (
      • Shansky R.M.
      Sex differences in PTSD resilience and susceptibility: Challenges for animal models of fear learning.
      ).
      In conclusion, taken together, our data provide evidence for fundamental sex differences in fear circuitry. That extinction in male animals correlated with structural differences in IL-BLA projections supports the current literature—which has been conducted almost exclusively in male animals—that this pathway plays a critical role in suppressing freezing during extinction retrieval. The lack of a discernible relationship between IL-BLA structure and extinction retrieval in female animals suggests that variability in freezing during extinction retrieval in female animals may instead reflect activity in mechanisms that drive fear expression, rather than suppression. Alternatively, freezing may be an incomplete indicator of fear in female animals. This distinction could have implications for identifying sex-specific risk factors for PTSD, which is characterized by poor extinction retrieval and altered prefrontal-amygdala connectivity (
      • Milad M.R.
      • Pitman R.K.
      • Ellis C.B.
      • Gold A.L.
      • Shin L.M.
      • Lasko N.B.
      • et al.
      Neurobiological basis of failure to recall extinction memory in posttraumatic stress disorder.
      ). Much work needs to be done to define fully sex differences in the mechanisms that drive fear conditioning and extinction, and we look forward to continuing our investigations. Our findings not only highlight the need for increased inclusion of female animals in basic neuroscience but also demonstrate the value of taking a within-sex approach to sex differences research.

      Acknowledgments And Disclosures

      This work was supported by the National Institute of Mental Health Grant No. R21-MH098006-01 (RMS).
      We thank Jennifer Lipps for technical assistance and Natalie Tronson and Mark Baxter for manuscript comments.
      TMG, ER, VT, and AR performed the experiments and analyzed the data. TMG and RMS designed the experiments and wrote the manuscript.
      The authors report no biomedical financial interests or potential conflicts of interest.

      References

        • Quirk G.J.
        • Likhtik E.
        • Pelletier J.G.
        • Paré D.
        Stimulation of medial prefrontal cortex decreases the responsiveness of central amygdala output neurons.
        J Neurosci. 2003; 23: 8800-8807
        • Milad M.R.
        • Quirk G.J.
        Neurons in medial prefrontal cortex signal memory for fear extinction.
        Nature. 2002; 420: 70-74
        • Stevens J.S.
        • Jovanovic T.
        • Fani N.
        • Ely T.D.
        • Glover E.M.
        • Bradley B.
        • Ressler K.J.
        Disrupted amygdala-prefrontal functional connectivity in civilian women with posttraumatic stress disorder.
        J Psychiatr Res. 2013; 47: 1469-1478
        • Milad M.R.
        • Pitman R.K.
        • Ellis C.B.
        • Gold A.L.
        • Shin L.M.
        • Lasko N.B.
        • et al.
        Neurobiological basis of failure to recall extinction memory in posttraumatic stress disorder.
        Biol Psychiatry. 2009; 66: 1075-1082
        • Breslau N.
        The epidemiology of trauma, PTSD, and other posttrauma disorders.
        Trauma Violence Abuse. 2009; 10: 198-210
        • Herry C.
        • Ciocchi S.
        • Senn V.
        • Demmou L.
        • Müller C.
        • Lüthi A.
        Switching on and off fear by distinct neuronal circuits.
        Nature. 2008; 454: 600-606
        • Shansky R.M.
        • Hamo C.
        • Hof P.R.
        • McEwen B.S.
        • Morrison J.H.
        Stress-induced dendritic remodeling in the prefrontal cortex is circuit specific.
        Cereb Cortex. 2009; 19: 2479-2484
        • Shansky R.M.
        • Hamo C.
        • Hof P.R.
        • Lou W.
        • McEwen B.S.
        • Morrison J.H.
        Estrogen promotes stress sensitivity in a prefrontal cortex-amygdala pathway.
        Cereb Cortex. 2010; 20: 2560-2567
        • Lebron-Milad K.
        • Milad M.R.
        Sex differences, gonadal hormones and the fear extinction network: Implications for anxiety disorders.
        Biol Mood Anxiety Disord. 2012; 2: 3
        • Farrell M.R.
        • Sengelaub D.R.
        • Wellman C.L.
        Sex differences and chronic stress effects on the neural circuitry underlying fear conditioning and extinction.
        Physiol Behav. 2013; 122: 208-215
        • Shansky R.M.
        Estrogen, stress and the brain: Progress toward unraveling gender discrepancies in major depressive disorder.
        Expert Rev Neurother. 2009; 9: 967-973
        • Paxinos G.
        • Watson C.
        The Rat Brain in Stereotaxic Coordinates.
        7th ed. Academic Press, Waltham, Massachusetts2013
        • Gruene T.M.
        • Lipps J.
        • Rey C.D.
        • Bouck A.
        • Shansky R.M.
        Heat exposure in female rats elicits abnormal fear expression and cellular changes in prefrontal cortex and hippocampus.
        Neurobiol Learn Mem. 2014; 115: 38-42
        • Bush D.E.
        • Sotres-Bayon F.
        • LeDoux J.E.
        Individual differences in fear: Isolating fear reactivity and fear recovery phenotypes.
        J Trauma Stress. 2007; 20: 413-422
        • Milad M.R.
        • Igoe S.A.
        • Lebron-Milad K.
        • Novales J.E.
        Estrous cycle phase and gonadal hormones influence conditioned fear extinction.
        Neuroscience. 2009; 164: 887-895
        • Rey C.D.
        • Lipps J.
        • Shansky R.M.
        Dopamine D1 receptor activation rescues extinction impairments in low-estrogen female rats and induces cortical layer-specific activation changes in prefrontal-amygdala circuits.
        Neuropsychopharmacology. 2014; 39: 1282-1289
        • Rodriguez A.
        • Ehlenberger D.B.
        • Dickstein D.L.
        • Hof P.R.
        • Wearne S.L.
        Automated three-dimensional detection and shape classification of dendritic spines from fluorescence microscopy images.
        PLoS One. 2008; 3: e1997
        • Radley J.J.
        • Sisti H.M.
        • Hao J.
        • Rocher A.B.
        • McCall T.
        • Hof P.R.
        • et al.
        Chronic behavioral stress induces apical dendritic reorganization in pyramidal neurons of the medial prefrontal cortex.
        Neuroscience. 2004; 125: 1-6
        • Liston C.
        • Miller M.M.
        • Goldwater D.S.
        • Radley J.J.
        • Rocher A.B.
        • Hof P.R.
        • et al.
        Stress-induced alterations in prefrontal cortical dendritic morphology predict selective impairments in perceptual attentional set-shifting.
        J Neurosci. 2006; 26: 7870-7874
        • Oh S.W.
        • Harris J.A.
        • Ng L.
        • Winslow B.
        • Cain N.
        • Mihalas S.
        • et al.
        A mesoscale connectome of the mouse brain.
        Nature. 2014; 508: 207-214
        • Haufler D.
        • Nagy F.Z.
        • Pare D.
        Neuronal correlates of fear conditioning in the bed nucleus of the stria terminalis.
        Learn Mem. 2013; 20: 633-641
        • Elharrar E.
        • Warhaftig G.
        • Issler O.
        • Sztainberg Y.
        • Dikshtein Y.
        • Zahut R.
        • et al.
        Overexpression of corticotropin-releasing factor receptor type 2 in the bed nucleus of stria terminalis improves posttraumatic stress disorder-like symptoms in a model of incubation of fear.
        Biol Psychiatry. 2013; 74: 827-836
        • Dumitriu D.
        • Rodriguez A.
        • Morrison J.H.
        High-throughput, detailed, cell-specific neuroanatomy of dendritic spines using microinjection and confocal microscopy.
        Nat Protoc. 2011; 6: 1391-1411
        • Little J.P.
        • Carter A.G.
        Synaptic mechanisms underlying strong reciprocal connectivity between the medial prefrontal cortex and basolateral amygdala.
        J Neurosci. 2013; 33: 15333-15342
        • McEwen B.S.
        • Morrison J.H.
        The brain on stress: Vulnerability and plasticity of the prefrontal cortex over the life course.
        Neuron. 2013; 79: 16-29
        • Lai C.S.W.
        • Franke T.F.
        • Gan W.-B.
        Opposite effects of fear conditioning and extinction on dendritic spine remodelling.
        Nature. 2012; 483: 87-91
        • Bourne J.
        • Harris K.M.
        Do thin spines learn to be mushroom spines that remember?.
        Curr Opin Neurobiol. 2007; 17: 381-386
        • Matsuo N.
        • Reijmers L.
        • Mayford M.
        Spine-type-specific recruitment of newly synthesized AMPA receptors with learning.
        Science. 2008; 319: 1104-1107
        • Sierra-Mercado D.
        • Padilla-Coreano N.
        • Quirk G.J.
        Dissociable roles of prelimbic and infralimbic cortices, ventral hippocampus, and basolateral amygdala in the expression and extinction of conditioned fear.
        Neuropsychopharmacology. 2011; 36: 529-538
        • Vidal-Gonzalez I.
        • Vidal-Gonzalez B.
        • Rauch S.L.
        • Quirk G.J.
        Microstimulation reveals opposing influences of prelimbic and infralimbic cortex on the expression of conditioned fear.
        Learn Mem. 2006; 13: 728-733
        • Fenton G.E.
        • Pollard A.K.
        • Halliday D.M.
        • Mason R.
        • Bredy T.W.
        • Stevenson C.W.
        Persistent prelimbic cortex activity contributes to enhanced learned fear expression in females.
        Learn Mem. 2014; 21: 55-60
        • Zeidan M.A.
        • Igoe S.A.
        • Linnman C.
        • Vitalo A.
        • Levine J.B.
        • Klibanski A.
        • et al.
        Estradiol modulates medial prefrontal cortex and amygdala activity during fear extinction in women and female rats.
        Biol Psychiatry. 2011; 70: 920-927
        • Milad M.R.
        • Zeidan M.A.
        • Contero A.
        • Pitman R.K.
        • Klibanski A.
        • Rauch S.L.
        • Goldstein J.M.
        The influence of gonadal hormones on conditioned fear extinction in healthy humans.
        Neuroscience. 2010; 168: 652-658
        • Graham B.M.
        • Milad M.R.
        Blockade of estrogen by hormonal contraceptives impairs fear extinction in female rats and women.
        Biol Psychiatry. 2013; 73: 371-378
        • Quirk G.J.
        • Mueller D.
        Neural mechanisms of extinction learning and retrieval.
        Neuropsychopharmacology. 2008; 33: 56-72
        • Sepulveda-Orengo M.T.
        • Lopez A.V.
        • Soler-Cedeño O.
        • Porter J.T.
        Fear extinction induces mGluR5-mediated synaptic and intrinsic plasticity in infralimbic neurons.
        J Neurosci. 2013; 33: 7184-7193
        • Maren S.
        • De Oca B.
        • Fanselow M.S.
        Sex differences in hippocampal long-term potentiation (LTP) and Pavlovian fear conditioning in rats: positive correlation between LTP and contextual learning.
        Brain Res. 1994; 661: 25-34
        • Baran S.E.
        • Armstrong C.E.
        • Niren D.C.
        • Hanna J.J.
        • Conrad C.D.
        Chronic stress and sex differences on the recall of fear conditioning and extinction.
        Neurobiol Learn Mem. 2009; 91: 323-332
        • Lebrón-Milad K.
        • Tsareva A.
        • Ahmed N.
        • Milad M.R.
        Sex differences and estrous cycle in female rats interact with the effects of fluoxetine treatment on fear extinction.
        Behav Brain Res. 2013; 253: 217-222
        • Shansky R.M.
        Sex differences in PTSD resilience and susceptibility: Challenges for animal models of fear learning.
        Neurobiol Stress. 2015; 1: 60-65