It has been reported that there is a sex difference in response to fear-related stress [15]. In addition, to minimize the effects of variations in stress responses across different phases of the estrous cycle and the influence of sex hormones [16], and to focus on the effect of corticotropin-releasing hormone, we used only male mice in the present study. Male C57BL/6J mice, aged 6–7 weeks, were purchased from CLEA Japan (Tokyo, Japan). Mice were housed at 25 $\pm$ 2 °C under a 12-h light/dark cycle (light: 08:00–20:00 h) with ad libitum access to food (#CE-2, CLEA Japan, Tokyo, Japan) and water. Groups of 4–5 mice were kept per cage. The study was approved by the Animal Care Committee of Ohu University (2021-14 and 2022-20) and adhered to and under the ARRIVE guidelines, all efforts were taken to minimize distress and use the minimum required number of animals. Behavioral tests were conducted between 11:00 and 16:00 h and analyzed by an independent investigator using ANY-maze software (v6.35; Muromachi Kikai, Tokyo, Japan). Dexamethasone (#D4438) was purchased from Tokyo Kasei (Tokyo, Japan), and fludrocortisone acetate (F6127-1G) was procured from Sigma-Aldrich (St. Louis, MO, USA).

Mice were anesthetized using medetomidine hydrochloride (0.3 mg/kg, Cat. No. 021-19001, Wako Pure Chemical Industries, Ltd., Tokyo, Japan), butorphanol tartrate (5 mg/kg, Cat. No. 135-17473, Wako Pure Chemical Industries, Ltd.), and midazolam (4 mg/kg, Product No. 21800AMX10357000, Sandoz, Ltd., Yamagata, Japan). Blood samples were collected via cardiac puncture using a 26-gauge syringe. Plasma was prepared with 1.0% citrate (Cat. No.58012-17, Kanto chemical Co. INC., Tokyo, Japan), 1.5 mg/mL EDTA-2Na (Cat. No. 15112-22, Nacalai tesque, Kyoto, Japan), and 12 U/mL heparin (Cat. No. 085-00134, FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) to prevent coagulation. The plasma was then centrifuged at 1200 $\times$g for 15 minutes at 4 °C, and the supernatants were stored at –80 °C for corticosterone measurement. Plasma corticosterone levels were quantified using an EIA kit (#YK240, Yanaihara Inst., Shizuoka, Japan). Serum samples were analyzed in duplicate, with a detection limit of 0.21 ng/mL. Standard curves were used, and absorbance at 450 nm was measured with a microplate reader (BioTek Instruments, Winooski, VT, USA) according to the manufacturer’s guidelines.

We previously demonstrated that TMT exposure elicits a significant fear response in mice [13]. In this study, a total of 56 male C57BL/6J mice (8–12 weeks old) were used for the TMT-induced inescapable innate fear test. The experimental group consisted of 42 mice exposed to TMT, while the mineral oil (MO) control group included 14 mice.

One day before the test, the mice underwent a 15-minute habituation period in the TMT test box arena. Each mouse was placed in the center of an acrylic OF box (W: 294 mm $\times$ D: 294 mm $\times$ H: 297 mm), with the bottom and inner walls covered in non-reflective paper. The transparent plexiglass top allowed video recording using a webcam. A circular filter paper (D: 2 cm, Cat# 1001-025; Whatman GE Healthcare Life Science, Little Chalfont, UK) infused with 20 µL of TMT was placed in a corner of the odor exposure box.

The TMT odor was introduced 5 minutes after the mice were placed in the OF box, and fear-related behaviors were observed for 15 minutes. Three test boxes were prepared, with 3 to 6 mice tested per day to prevent reuse of the same box within an hour. Between experiments, the TMT odor was removed by treating each box with a 5% bleach solution for 30 minutes, followed by washing with detergent and water, and then wiping with 70% ethanol.

To ensure baseline behavior, we confirmed that the mice did not exhibit freezing behavior during a 5-minute OF test before TMT exposure.

We previously reported the measurement of freezing behavior during TMT exposure [13]. Herein, in this study, freezing behavior, described as the absence of all movement except for respiration [17], was used as a measure of fear in an acrylic box (294 $\times$ 294 $\times$ 297 mm) lit with 60 lux. The amount of time spent freezing, expressed as a percentage, was calculated for each mouse during the 10-minute period following TMT exposure, with TMT applied to the filter paper in the corner of the OF box. Freezing duration was analyzed in two 5-min bins: the first (6–10 min) and the second (11–15 min).

We previously reported the detailed method of the TST [18]. In this study, the test was conducted during the light phase (13:00–16:00 h) of the light-dark cycle. Mice were individually suspended by the tail using adhesive tape, placed 2 cm from the tail tip, and attached to the ceiling of the test box, positioned 42 cm above the bench. The suspension lasted for 10 minutes, and behavior was recorded using a digital camera without the investigator present.

An independent investigator, blinded to group allocation, measured immobility time using ANY-maze software. Mobility was defined as any movement of the hind legs or other behaviors indicating an attempt to escape, excluding breathing. Immobility was defined as a motionless state lasting at least 2 seconds, with sensitivity set to 80%. After the TST, mice were returned to their home cages [18]. We previously reported that a decrease in active coping is typically observed during the first half of the 10-minute TST recording. However, differences in learned despair tend to emerge during the latter half of the test [13, 19]. Therefore, in this study, immobility time was measured separately for the first five minutes and the subsequent five minutes. Both durations were used as indices of passive coping behavior.

Statistical analyses were performed using EZR (version 1.38, Jichi Medical University, Tochigi, Japan) [20] and BellCurve software (version 3.20, Social survey research information co., ltd., Tokyo, Japan) [21]. The Student’s t-test was used to compare groups, one-way ANOVA for single-factor comparisons, and two-way repeated measures ANOVA for multiple comparisons over time. Post hoc analysis was conducted using the Bonferroni test.

For correlation analyses, the normality of each set of behavioral measures was first assessed using the Shapiro-Wilk test. Pearson’s correlation coefficient was then calculated to examine relationships between variables. When conducting the Student’s t-test, homogeneity of variance was assessed using the F value, and Cohen’s d effect size was calculated to quantify the relative magnitude of differences between groups. Cohen’s d represents the standardized difference between two means, expressed in units of standard deviation.

Effect sizes were classified based on the absolute values of Cohen’s d, using the following thresholds: d $\le$ 0.5 (small effect), 0.5 d 1.0 (moderate effect), 1.0 $\le$ d 1.5 (large effect), and d $\ge$ 1.5 (very large effect) [22]. An a priori power analysis was conducted to determine the appropriate sample size, incorporating an anticipated effect size (Cohen’s d $>$ 1.5, representing a very large effect) and a predefined significance level (p) for the Student’s t-test [23]. Data are presented as mean $\pm$ standard error of the mean (SEM), with significance set at p 0.05 (*) or p 0.01 (**).

Psychological stress induces corticotropin-releasing hormone (CRH) release in the paraventricular nucleus of the hypothalamus, leading to increased corticosterone levels in mice. In our previous study, we generated the adeno-associated virus (AAV) vector for CRH overexpression and injected it into the hypothalamus (Supplementary Fig. 1) [24]. Therefore, to assess baseline corticosterone levels, we used the mice with either the control virus injection into the hypothalamus (Hy-CRH-control) or hypothalamic CRH overexpression (Hy-CRH-OE) under non-stressful conditions. Hy-CRH-OE mice exhibited a significantly higher plasma corticosterone concentration than Hy-CRH-control mice (t₍₈₎ = 46.6, p 0.05, Cohen’s d = 2.14, minimum sample size per group calculated by a priori sample size; n = 5, Student t-test: p 0.01, Fig. 1A). However, Hy-CRH-OE did not affect the locomotor activity, as measured by the distance traveled in the OF test (t₍₁₆₎ = 46.6, p = 0.8, Cohen’s d = 0.011, Student t-test: p = 0.9769, Fig. 1B). Three-way repeated measures ANOVA (Odor [mineral oil or TMT] $\times$ Genotype $\times$ Time) revealed that the interaction between odor and time significantly affected fear sensitivity (F_(3,92) = 46.5, p 0.01, Fig. 1C). However, a two-way ANOVA (Odor $\times$ Genotype) for period of “before odor” indicated that freezing time was not significantly affected (F_(3,68) = 0.422, p = 0.738, Fig. 1C). In contrast, during both the initial (1–5 min) and final (6–10 min) period, freezing time was significantly affected (1–5 min: F_(3,68) = 33.85, p 0.01; 6–10 min: F_(3,68) = 76.16, p 0.01; Fig. 1C). A Bonferroni post-hoc test indicated that TMT significantly increased freezing time in both the initial and final phases (1–5 min: Hy-CRH-control TMT, p 0.01 vs. Hy-CRH-control MO, p 0.01 vs. Hy-CRH-OE MO; Hy-CRH-OE TMT, p 0.01 vs. Hy-CRH-OE MO; Fig. 1C). Similarly, three-way repeated measures ANOVA (Odor [mineral oil or TMT] $\times$ Genotype $\times$ Time) showed significant effects of odor on central preference measures: number of entries (F_(3,92) = 7.26, p 0.01, Fig. 1D), time spent (F_(3,92) = 14.95, p 0.01, Fig. 1E), and distance traveled (F_(3,92) = 5.94, p 0.01, Fig. 1F). Two-way ANOVA for the pre-odorant phase revealed no significant effects on central preference (number of entries: F_(3,70) = 0.055, p = 0.983; time spent: F_(3,78) = 0.085, p = 0.968; distance traveled: F_(3,61) = 0.043, p = 0.988; Fig. 1D–F). However, significant effects were observed during both the initial (1–5 min) and final (6–10 min) period (1–5 min: number of entries: F_(3,70) = 5.86, p 0.01; time spent: F_(3,78) = 9.95, p 0.01; distance traveled: F_(3,61) = 4.68, p 0.01; 6–10 min: number of entries: F_(3,70) = 15.35, p 0.01; time spent: F_(3,78) = 20.19, p 0.01; distance traveled: F_(3,61) = 14.62, p 0.01; Fig. 1D–F). Bonferroni post-hoc tests revealed that TMT significantly reduced central preference compared to mineral oil in both periods (1–5 min: Hy-CRH-control TMT, p = 1.000 vs. Hy-CRH-OE MO; Hy-CRH-OE TMT, p 0.01 vs. Hy-CRH-control MO, p 0.01 vs. Hy-CRH-OE MO; 6–10 min: Hy-CRH-control TMT, p 0.01 vs. Hy-CRH-control MO, p 0.01 vs. Hy-CRH-OE MO; Hy-CRH-OE TMT, p 0.01 vs. Hy-CRH-control MO, p 0.01 vs. Hy-CRH-OE MO; Fig. 1D–F).

Fig. 1.

The effect of hypothalamic CRH on locomotor activity, fear response levels, and central preference. (A) Concentration of plasma corticosterone in control virus injection into the hypothalamus (Hy-CRH-control) (n = 5) and the overexpression of corticotropin-releasing hormone in hypothalamus (Hy-CRH-OE) (n = 5) mice during 12:00–14:00 h in the light phase. (B) Effect of hypothalamic CRH overexpression on locomotor activity. Distance (m) traveled in an open field test box. Hy-CRH-control (n = 8) and Hy-CRH-OE (n = 10). (C) The percentage of freezing time for 5 min before (–5 to 0 min) and during the first (1‒5 min) and last 5 min (6–10 min) of mineral oil (MO) and 2,5-dihydro-2,4,5-trimethylthiazoline (TMT) exposure in Hy-CRH-control mice (MO: n = 6; TMT: n = 8) and Hy-CRH-OE mice (MO: n = 7; TMT: n = 10). (D–F) Distance traveled (D), number of entries (E), and time spent (F) in the central zone in the open field test box for 5 min before (‒5 to 0 min) and during the first and second 5 min of MO and TMT exposure in Hy-CRH-control (MO: n = 6; TMT: n = 8) and Hy-CRH-OE mice (MO: n = 7; TMT: n = 10). Data are presented as mean $\pm$ SEM, with statistical significance indicated (*p 0.05, **p 0.01). CRH, corticotropin-releasing hormone; TMT, 2,5-dihydro-2,4,5-trimethylthiazoline; MO, mineral oil.

Moderate fear experiences have been shown to decrease immobility time in rats [25], whereas it has been demonstrated that more severe fear experiences result in increased time spent immobile during the forced swimming test (FST), accompanied by elevated corticosterone levels, which affect coping behavior under subsequent inescapable stress [26]. Thus, corticosterone levels may influence passive coping levels during the TST following TMT-evoked fear experience. We measured immobility time during TST following exposure to the OF test (Fig. 2A) (mineral oil as a control of TMT exposure (Fig. 2B)) in Hy-CRH-control and Hy-CRH-OE mice. When the TMT was not introduced, no significant differences in immobility time were observed between the two groups (Mineral oil: two-way repeated measures ANOVA: F_(1,129) = 0.0722, p = 0.793, Fig. 2C), suggesting that Hy-CRH-OE did not affect passive coping during aversive stress. Coping level in the last 5 min of TMT exposure (6–10 min) was similar between groups when the TMT was not present (Mineral oil: F_(1,25) = 0.073, p = 0.792, Fig. 2E), indicating that the learned despair level was comparable between both groups. This suggests that the increase in passive coping behavior during the final 5 minutes reflects learned despair [19]. However, after TMT exposure, which evoked inescapable innate fear, Hy-CRH-OE mice exhibited a significantly longer immobility time during TST compared to Hy-CRH-control mice at all time points (two-way repeated measures ANOVA: F_(1,179) = 6.80, p 0.05, Fig. 2D). Post hoc tests by student t-test revealed significant differences at multiple time points, with Hy-CRH-OE mice showing longer immobility time than controls (TMT: p 0.01 for 2, 3, 5, 9 min and p 0.05 for 10 min, Fig. 2D). Furthermore, during the initial 5 min of TST (1–5 min), Hy-CRH-OE mice also displayed significantly higher immobility time compared to Hy-CRH-control mice after TMT exposure (TMT: two-way repeated measures ANOVA: F_(1,35) = 6.81, p 0.05; Bonferroni post hoc test: 1–5 min: TMT: p 0.01 vs. control, Fig. 2F). Taken together, since the increased immobility time in Hy-CRH-OE mice was observed both in the first and last 5 minutes, with significantly longer times compared to Hy-CRH-Control mice, prior exposure to TMT-induced fear stress reduced active coping abilities in the Hy-CRH-OE mice during the subsequent aversive stress in the TST, rather than enhancing the learned despair level.

Fig. 2.

The effect of CRH overexpression in the hypothalamus on active coping levels during the TST after TMT-evoked fear stress. (A,B) Schematic diagrams indicate the placement of the central zone and either MO (A) or TMT (B), as well as the schedule for the tail suspension test (TST). (C,D) Demonstrate time-dependent changes in immobility over the 10-minute period of the TST in Hy-CRH-control (n = 7) and Hy-CRH-OE mice (n = 6) under the presence of MO (C), and in Hy-CRH-control (n = 8) and Hy-CRH-OE mice (n = 10) under TMT exposure (D). (E,F) Overview of immobility duration during the first (1–5 min) and final 5 minutes (6–10 min) 1 hour after exposure to TMT in Hy-CRH-control (n = 7) and Hy-CRH-OE mice (n = 6) under the presence of MO (E), and in Hy-CRH-control (n = 8) and Hy-CRH-OE mice (n = 10) under TMT exposure (F). Data are presented as mean $\pm$ SEM, with statistical significance denoted (*p 0.05, **p 0.01).

To further examine the association between fear responses and coping behavior, we examined whether there were correlations between central preferences during TMT-evoked inescapable innate fear stress and passive coping behavior during the subsequent TST. In Hy-CRH-control mice, significant correlations were observed between central preferences (number of entries, time spent, and distance traveled in the central zone) and freezing time during TMT exposure (Hy-CRH-Control: Shapiro-Wilk test: freezing time: p 0.05, number of entries: p 0.05, time spent: p 0.05, distance traveled: p 0.05; Pearson’s correlation coefficient: number of entries: p 0.01, Fig. 3A; time spent: p 0.01, Fig. 3B; distance traveled: p 0.01, Fig. 3C, Table 1), indicating that a high level of central preference was associated with a relatively lower freezing time. Moreover, there was a correlation between immobility time in the TST and freezing time during TMT exposure (Hy-CRH-Control: Shapiro-Wilk test: immobility time: p 0.05; Pearson’s correlation coefficient: number of entries: p 0.01; Fig. 3D) as well as with central preferences (Pearson’s correlation coefficient: number of entries: p 0.05, Fig. 3E; time spent: p 0.05, Fig. 3F; distance traveled: p 0.05, Fig. 3G, Table 1). However, the correlations between freezing time and central preference, as well as between immobility time and either freezing time or central preference, which were observed in the Hy-CRH-Control mice, were completely abolished in the Hy-CRH-OE mice (Fig. 3A–G, Table 1). These results suggest that low corticosterone concentrations in Hy-CRH-control mice during TMT-induced fear stress or high corticosterone levels in Hy-CRH-OE mice without fear stress (mineral oil instead of TMT) may result in similar correlations between central preference and passive coping levels during the TST. However, sustained high corticosterone levels following TMT exposure seem to disrupt these correlations in Hy-CRH-OE mice.

Fig. 3.

The effect of CRH overexpression in the hypothalamus on the relationships between freezing levels, central preference, and active coping levels. (A–C) Correlation between the percentage of freezing time in the first 5 minutes of TMT exposure and the number of entries (A), the time spent (B), and the distance traveled (C) in the central zone in the open field test box for both Hy-CRH-control (n = 8) and Hy-CRH-OE (n = 10) mice. (D) Correlation between the percentage of freezing time and immobility duration during the initial 5 minutes of TMT exposure for both Hy-CRH-control (n = 8) and Hy-CRH-OE mice (n = 10). (E–G) Correlation between the number of entries (E), the time spent (F), and the distance traveled (G) in the central zone during the first 5 min of TMT exposure and the immobility time during the first 5 min of the tail suspension test (TST) in Hy-CRH-control (n = 8) and Hy-CRH-OE mice (n = 10). R² is Pearson’s correlation coefficient.

The abolition of the correlation of central preference during TMT-evoked innate fear stress and passive coping behavior during subsequent aversive stress in the TST was due to hypothalamic CRH overexpression. To identify the corticoid receptor contributing to these correlations in Hy-CRH-OE mice, we administered subcutaneous injections of dexamethasone (a GR agonist, 10 µg/kg) or fludrocortisone (an MR agonist, 5.0 mg/kg) to non-AAV-infected normal mice 30 minutes before TMT-induced innate fear stress. This was based on evidence that 10 µg/kg of dexamethasone is the minimum dose that does not affect active coping levels in naïve mice without prior psychological stress [10], while 5.0 mg/kg of fludrocortisone is the minimum dose required to induce renal effects in mice [27]. Given that central preference in Hy-CRH-OE mice was lower than in Hy-CRH-control mice, we investigated which receptor mediates the reduction in central preference during TMT-evoked inescapable innate fear stress. Two-way repeated measures ANOVA revealed that the interaction between time and drug did not affect fear sensitivity (F_(2,71) = 1.23, p = 0.3113, Fig. 4A). Furthermore, two-way repeated measures ANOVA indicated that the interaction between time and drug did not affect the number of entries into the central zone (F_(2,71) = 2.081, p = 0.150, Fig. 4B). However, one-way ANOVA revealed significant effects of drug on the central preference during first 5 min of TMT exposure, including the number of entries into the central zone (F_(2,23) = 9.311, p 0.01, Fig. 4B), time spent in the central zone (F_(2,23) = 4.94, p 0.05, Fig. 4C), and distance traveled in the central zone (F_(2,23) = 4.39, p 0.05, Fig. 4D). Bonferroni post hoc tests indicated that fludrocortisone-treated mice exhibited reduced central preference, as evidenced by fewer entries (p 0.05), less time spent (p 0.05), and shorter distance traveled (p 0.05) in the central zone, compared to controls (Fig. 4B‒D). In contrast, dexamethasone-treated mice showed no significant differences compared to controls in the central preferences (Fig. 4B‒D). Furthermore, the significant correlations between central preference and freezing time during TMT-evoked inescapable innate fear stress were observed in dexamethasone-treated mice as well as control mice (Fig. 4E‒G and Table 2). On the other hand, pretreatment of fludrocortisone abolished these correlations between central preference and freezing time during TMT-evoked inescapable innate fear stress (Fig. 4H‒J and Table 2). These findings suggest that disrupted correlations between freezing time and central preference in Hy-CRH-OE mice may be due to an imbalance between MR and GR activity, potentially resulting in MR hyperactivation during TMT-induced inescapable innate fear stress.

Fig. 4.

The contribution of glucocorticoid or mineralocorticoid receptors to the correlation between freezing levels and central preferences during TMT exposure. (A–D) Freezing time (sec) (A), number of entries (B), time spent (C), and distance traveled (m) (D), in the central zone in the open field test box before (‒5 to 0 min) and during the first and second 5 min of TMT exposure in control (n = 10), dexamethasone-treated (n = 7), and fludrocortisone-treated (n = 7) mice. (E–G) Correlation between the number of entries (E), the time spent (F), and the distance traveled (G) in the central zone during the first 5 min of TMT and the freezing time during the first 5 min of TMT in control (n = 10) and dexamethasone-treated mice (n = 7). (H–J) Correlation between the number of entries (H), the time spent (I), and the distance traveled (J) in the central zone and immobility time during the first 5 min of the TST in control (n = 10) and fludrocortisone-treated mice (n = 7). Data are presented as mean $\pm$ SEM, with statistical significance indicated (*p 0.05).

Next, we investigated the roles of MR and GR in passive coping responses following fear-induced stress and examined the correlation between central preference and passive coping levels (Fig. 5A). A two-way repeated measures ANOVA revealed no significant interaction between time and drug on immobility time in the TST (F_(2,239) = 1.764, p = 0.196, Fig. 5B). However, Bonferroni post hoc tests indicated that fludrocortisone significantly increased immobility time at multiple time points compared to the control group (1 min: p 0.05; 2 min: p 0.05; 3 min: p 0.01; 4 min: p 0.01; 6 min: p 0.05; 7 min: p 0.05; Fig. 5B). Fludrocortisone also reduced active coping level, as evidenced by increased immobility time during the TST (F_(2,71) = 2.158, p = 0.141, Bonferroni post hoc test: p 0.05 vs. control, Fig. 5C). The correlation between immobility time in the TST and freezing time during TMT-evoked fear stress remained unaffected by dexamethasone (Fig. 5D, Table 2), as did the correlation between immobility time in the TST and central preference during TMT-induced fear stress (Fig. 5E‒G, Table 2). However, pretreatment with fludrocortisone abolished both the correlation between immobility time in the TST and freezing time during TMT-evoked fear stress (Fig. 5H, Table 2) and the correlation between immobility time and central preference during TMT-evoked fear stress (Fig. 5I‒K, Table 2). These findings suggest that MR activation disrupts the relationship between central preference during inescapable innate fear stress and passive coping during subsequent aversive stress.

Fig. 5.

The contribution of glucocorticoid or mineralocorticoid receptors to the relationship between freezing levels, central preferences during TMT exposure, and active coping levels during the TST. (A) Schematic drawings indicate the schedule of vehicle or drugs, 5 min of open field test before TMT exposure, followed by 10 min of TMT exposure and the tail suspension test (TST). (B) The time-dependent changes in immobility during the 10-minute TST in the control group (n = 10), dexamethasone-treated (n = 7), and fludrocortisone-treated (n = 7) mice under TMT exposure. (C) Summary of immobility time during the initial (1–5 min) and final (6–10 min) 5 minutes, 1 hour after TMT exposure in control (n = 10), dexamethasone-treated (n = 7), and fludrocortisone-treated (n = 7) mice. (D‒K) Correlation between freezing time during the first 5 min and the immobility time during the first 5 min of TMT exposure for control (n = 10), dexamethasone-treated (n = 7) (D‒G), and fludrocortisone-treated (n = 7) mice (H‒K). Correlation between the freezing time (D,H), the number of entries (E,I), the time spent (F,J), and the distance traveled (G,K) in the central zone during the first 5 min of TMT exposure and the immobility time during the first 5 min of the TST in control and fludrocortisone-treated mice. Data are presented as mean $\pm$ SEM, with statistical significance indicated (*p 0.05, **p 0.01).