- Email: [email protected]
Social buffering ameliorates conditioned fear responses in female rats
Hormones and Behavior 81 (2016) 53–58
Contents lists available at ScienceDirect
Hormones and Behavior journal homepage: www.elsevier.com/locate/yhbeh
Social buffering ameliorates conditioned fear responses in female rats Akiko Ishii, Yasushi Kiyokawa ⁎, Yukari Takeuchi, Yuji Mori Laboratory of Veterinary Ethology, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan
a r t i c l e
i n f o
Article history: Received 13 November 2015 Revised 24 February 2016 Accepted 18 March 2016 Available online 6 April 2016 Keywords: Female Social buffering Estrus cycle Sex difference Fear conditioning
a b s t r a c t The stress experienced by an animal is ameliorated when the animal is exposed to distressing stimuli along with a conspecific animal(s). This is known as social buffering. Previously, we found that the presence of an unfamiliar male rat induced social buffering and ameliorated conditioned fear responses of a male rat subjected to an auditory conditioned stimulus (CS). However, because our knowledge of social buffering is highly biased towards findings in male subjects, analyses using female subjects are crucial for comprehensively understanding the social buffering phenomenon. In the present studies, we assessed social buffering of conditioned fear responses in female rats. We found that the estrus cycle did not affect the intensity of the rats' fear responses to the CS or their degree of vigilance due to the presence of a conspecific animal. Based on these findings, we then assessed whether social buffering ameliorated conditioned fear responses in female rats without taking into account their estrus cycles. When fear conditioned female rats were exposed to the CS without the presence of a conspecific, they exhibited behavioral responses, including freezing, and elevated corticosterone levels. By contrast, the presence of an unfamiliar female rat suppressed these responses. Based on these findings, we conclude that social buffering can ameliorate conditioned fear responses in female rats. © 2016 Elsevier Inc. All rights reserved.
Introduction The stress experienced by a subject can be ameliorated when the subject is exposed to distressing stimuli while in the company of a conspecific animal. This phenomenon is called social buffering and can be induced by a mother, mate, or a same-sex or opposite-sex conspecific animal in a nonsexual relationship (Kiyokawa, in press). The phenomenon elicited by the last type of conspecific has been observed in a wide variety of social species, including sheep (da Costa et al., 2004), guinea pigs (Hennessy et al., 2006, 2008), and rats (Terranova et al., 1999). Previously, we conducted a series of experiments focused on the social buffering of conditioned fear responses in rats. In this model, male subject rats exhibited robust freezing and activation of the hypothalamic-pituitary-adrenal (HPA) axis in response to an auditory conditioned stimulus (CS), when the CS was paired with a foot shock during fear conditioning. The presence of an unfamiliar nonconditioned male rat (an associate) ameliorated these responses (Kiyokawa et al., 2007). Subsequently, we found that this social buffering occurred even when the subject and associate were separated by wire-mesh or double wire-mesh partitions (Kiyokawa et al., 2009, 2014a). Given that social buffering was inhibited by lesioning the main olfactory epithelium (Kiyokawa et al., 2009) and that associate-derived olfactory cues alone are able to induce social buffering (Takahashi et al., 2013; Kiyokawa et al., 2014b), we suggest that olfactory signaling mediates social buffering of conditioned ⁎ Corresponding author. E-mail address: [email protected] (Y. Kiyokawa).
http://dx.doi.org/10.1016/j.yhbeh.2016.03.003 0018-506X/© 2016 Elsevier Inc. All rights reserved.
fear responses. Recent studies have shed light on the neural mechanisms underlying social buffering, such as the suppression of the basolateral complex of the amygdala (Fuzzo et al., 2015) and the involvement of the posterior complex of the anterior olfactory nucleus as a relay point for signaling from the olfactory bulb to the amygdala (Kiyokawa et al., 2012; Takahashi et al., 2013). Our knowledge of social buffering is highly biased towards findings in male subjects, because all our previous studies, as well as most studies in the literature, have been conducted using males. Although some studies have reported the social buffering phenomenon in female subjects, the phenomenon was induced by a mate (Kaiser et al., 2003; Hennessy et al., 2008; Smith and Wang, 2014). Given that the neural mechanisms underlying social buffering differ depending on the type of conspecific animal, i.e., a mother, mate, or conspecific without sexual relationships, it would be appropriate to understand each phenomenon individually (Kiyokawa, in press). To the best of our knowledge, only one study using female guinea pigs reported a clear social buffering phenomenon by a conspecific animal without sexual relationships (a female guinea pig) (Hennessy et al., 2008). In addition, sex differences are thought to play a large role in stress responses. For example, males tend to show a “fight and flight” response to distressing stimuli, while females tend to exhibit a “tend and befriend” response (Taylor et al., 2000). Therefore, analyses using female subjects seem to be necessary to obtain a more comprehensive understanding of the social buffering phenomenon, especially that induced by conspecifics without sexual relationships. Changes in ovarian hormones that are dependent on the estrus cycle, such as estrogen and progesterone, appear to be one of the
54
A. Ishii et al. / Hormones and Behavior 81 (2016) 53–58
major challenges in analyzing social buffering of conditioned fear responses in female rats. First, these hormones affect the intensity of conditioned fear responses (Morgan and Pfaff, 2001). Given that the observed responses in the presence of a conspecific are residual intrinsic CS-induced responses after suppression by social buffering, we cannot appropriately evaluate the degree of social buffering if the intensity of intrinsic CS-induced responses fluctuates according to the estrus cycle. Second, these hormones can affect the degree of vigilance (Petersson et al., 1998), which may cause difference in the intensity of vigilance attributable to the presence of a conspecific. Fluctuations in vigilance due to the estrus cycle would also prevent us from appropriately evaluating social buffering, because vigilance can affect behavioral measures that require stillness and/or physiological measures related to metabolism, such as freezing and/or HPA axis activity, respectively. Therefore, it was necessary to assess the effects of the estrus cycle on these two factors. In the present study, we conducted a series of experiments to assess social buffering of conditioned fear responses in female rats. In Experiment 1, we assessed the effects of the estrus cycle on the intensity of conditioned fear responses. Fear conditioned and non-conditioned female subjects were exposed to an auditory CS while alone in the testing apparatus (solitary situation). Behavioral responses, including freezing, were compared across the stages. In Experiment 2, we assessed the effects of the estrus cycle on vigilance. Female subjects at all stages of the estrus cycle encountered unfamiliar female associates who were also in all stages of the cycle. Their behavioral responses were analyzed, including locomotor activity. In Experiment 3, we assessed whether social buffering ameliorated conditioned fear responses in female rats. Fear conditioned or non-conditioned female subjects were exposed to the auditory CS either alone (solitary situation) or with an unfamiliar female associate (dyad situation). Their behavioral responses, including freezing, and corticosterone levels were measured in order to evaluate the presence of social buffering. We conducted parallel experiments using male rats in order to evaluate sex differences in social buffering. Material and methods Animals All experiments were approved by the Animal Care and Use Committee of the Faculty of Agriculture of The University of Tokyo and were based on guidelines that were adapted from the Consensus Recommendations on Effective Institutional Animal Care and Use Committees by the Scientists Center for Animal Welfare. Experimentally naïve female and male Wistar rats were purchased from Charles River Laboratories Japan (Kanagawa, Japan) at 7.5 weeks of age. Animals were housed 2–3 rats per cage upon arrival and under conditions of controlled temperature (24 ± 1 °C) and humidity (45 ± 5%). The housing room had a 12-h light/12-h dark cycle (lights on at 8:00). Water and food were available ad libitum. All behavioral procedures in Experiments 1 and 2 were performed between 8:00 and 15:00. The behavioral procedures in Experiment 3 were performed between 8:30 and 13:00 (see Experiment 3). Experiment 1 Female subjects were housed individually 3 days before fear conditioning until the test day and were handled for 3 min twice daily until the day of conditioning. A vaginal smear was taken every morning beginning on the day after arrival to evaluate the stage of the estrus cycle. Fear conditioning was performed under white-light conditions, as previously described (Kiyokawa et al., 2009). Subjects were placed in an acrylic conditioning box (28 × 20 × 27 cm) for 20 min. During fear conditioning, paired subjects were exposed to 7 pairings of a 3-s auditory CS (8 kHz, 70 dB) and a 0.5 s foot shock (0.80 mA), which ended simultaneously. Unpaired subjects received the same number of CS
and foot shocks, but these were presented separately. The inter-trial interval varied randomly between 60 and 180 s. After conditioning, subjects were returned to their home cages. The fear-expression test was performed 24 h after conditioning under dim red light, as previously described (Kiyokawa et al., 2009). Two rectangular enclosures (25 × 25 × 35 cm) were placed on an acrylic board (45 × 60 cm). Each enclosure consisted of 3 acrylic walls, 1 wire mesh wall, and a wire mesh ceiling. The inside of the enclosure was covered with clean bedding. The enclosures were placed side-by-side so that the wire mesh walls were facing each other, separated by a 5 cm gap. The wire mesh wall was composed of a 1 cm2 grid in the lower half (20 cm) and vertical bars in the upper half (15 cm) to prevent subjects from climbing the wire-mesh wall. Subjects were placed randomly in one of the two enclosures. After a 3-min acclimation period, the 3-s CS was presented 5 times at intervals of 1 min during the first 5 min of the 10-min test period. The subjects' behavior during the fearexpression test was recorded with an HDD-BD recorder (DMRBW770; Panasonic, Osaka) and a video camera (DCR-SR 300; Sony, Tokyo). The paired subjects were divided into 4 groups according to their estrus cycle at the time of the fear-expression test (diestrus 1, n = 10; diestrus 2, n = 10; proestrus, n = 8; estrus, n = 12). Because the estrus cycle did not alter the behavioral responses of the unpaired female subjects during the fear-expression test, all the unpaired subjects (diestrus 1, n = 4; diestrus 2, n = 4; proestrus, n = 5; estrus, n = 3) were combined into 1 group, regardless of their estrus cycle. All data are presented as the mean ± standard error of the mean (SEM). The significance level was set at p b 0.05 for all statistical tests. Behavioral parameters during the acclimation and test period were measured using Visual Basic software in Microsoft Excel, which recorded the duration and number of pressed keys. The duration of freezing (the lack of any movement except that which is required for respiration), the duration of investigation (sniffing towards the wire mesh wall within 1 cm or poking of the snout towards the wire mesh), and the frequency of walking (number of steps taken with the hind paws) during the preceding acclimation period and during the test period were analyzed by multivariate analysis of variance (MANOVA), followed by Fisher's PLSD post-hoc test. Effect sizes were further estimated by calculating the value of multivariate η2 and Cohen's d for data analyzed by MANOVA and Fisher's PLSD post-hoc test, respectively. Experiment 2 The encounter test was conducted in the same enclosures under dim red light as described above. In the encounter test, a female rat was placed in each enclosure for 5 min to observe the behavior to a novel animal, which was recorded with an HDD-BD recorder and a video camera. Each rat served both as the subject and also as an associate for the other. Based on the subject's and the associate's estrus cycle, we prepared 16 groups (n = 4–6 per group). Rats were housed individually until the test day, and were handled for 3 min twice daily for 3 days. The duration of freezing, the duration of investigation, and the frequency of walking were analyzed by two-way MANOVA. Effect sizes were further estimated by calculating the value of multivariate η2 for data analyzed by MANOVA. Experiment 3 Fear conditioning was performed as described in Experiment 1. Female and male rats assigned as subjects underwent fear conditioning, while rats assigned as associates remained in their home cages during fear conditioning. Subjects were further assigned as either paired or unpaired subjects. Because we found that the estrus cycle did not affect the intensity of the rats' fear responses to the CS or their degree of vigilance due to the presence of a conspecific animal, we did not assess the stage
A. Ishii et al. / Hormones and Behavior 81 (2016) 53–58
of the estrus cycle in female rats. All rats were housed individually until the test day and handled for 5 min daily for 3 days. The fear-expression test was conducted as described in Experiment 1, except that subject rats were tested either alone (solitary situation) or with an associate (dyad situation). When tested in the solitary situation, the subject was randomly placed in one of the enclosures, while the other enclosure was left vacant, similar to Experiment 1. In the dyad situation, a subject and a same-sex associate were each placed in a separate enclosure. The subjects' behavior was recorded with an HDD-BD recorder and a video camera. Corticosterone levels were measured as previously described (Kiyokawa et al., 2014a). After the fear-expression test, the subjects (and also the associates when tested in the dyad situation) were left undisturbed for an additional 10 min. Blood was then collected in hematocrit tubes from a small incision made by a razor on the subjects' tails. This sampling was completed within 3 min. Blood sampling was performed between 8:30 and 13:00 to take into account diurnal rhythms. The collected samples were centrifuged at 4 °C; plasma was stored at −80 °C. Corticosterone levels from each sample were mostly measured in duplicate (in triplicate, solitary situation: unpaired male, n = 1; paired male, n = 1) using the Corticosterone EIA kit (Cayman Chemical Company, Ann Arbor, MI). Subjects were divided into 8 groups based on their testing situation, sex, and conditioning procedure (solitary situation: unpaired females, n = 7; paired females, n = 9; unpaired males, n = 11; paired males, n = 11; dyad situation: unpaired females, n = 9; paired females, n = 9; unpaired females, n = 8; paired males, n = 8) (Fig. 2A, D). The duration of freezing, the duration of investigation, and the frequency of walking during the preceding acclimation period and the test period were analyzed by two-way MANOVA followed by Fisher's PLSD post-hoc test. The mean corticosterone levels were analyzed by two-way analysis of variance (ANOVA) followed by Fisher's PLSD post-hoc test. Effect sizes were further estimated by calculating the value of multivariate η2, η2, and Cohen's d for data analyzed by MANOVA, ANOVA and Fisher's PLSD post-hoc test, respectively.
55
Fig. 1. Fear responses during the fear-expression test in Experiment 1. Duration of freezing, duration of investigation, and frequency of walking (mean + SEM) during the test period. The letters indicate significant differences (p b 0.05), as assessed by MANOVA followed by Fisher's PLSD test.
estrus, p b 0.01, d = 1.74), and significantly lower frequency of walking (diestrus 1, p b 0.01, d = 1.51; diestrus 2, p b 0.01, d = 1.17; proestrus, p b 0.01, d = 1.31; estrus, p b 0.01, d = 2.08) compared with the unpaired subjects. However, the estrus cycle did not affect the behavioral responses of the paired subjects.
Results Experiment 1 Because the estrus cycle did not alter the behavioral responses of the unpaired female subjects during the acclimation period and test period, we combined these subjects into the unpaired group. The behavioral responses during the acclimation period did not differ between the groups [F(12, 130) = 0.781, p = 0.669, multivariate η2 = 0.168] (Table 1). Although the paired subjects in diestrus 1 appeared to show longer freezing durations, this was because 2 animals showed unnaturally long freezing durations. In contrast, behaviors during the test period varied significantly between the groups [F(12, 130) = 5.53, p b 0.01, multivariate η2 = 0.664] (Fig. 1). A post-hoc test revealed that the paired subjects showed significantly longer freezing durations (diestrus 1, p b 0.01, d = 2.47; diestrus 2, p b 0.01, d = 2.48; proestrus, p b 0.01, d = 2.61; estrus, p b 0.01, d = 2.86), significantly shorter investigation durations (diestrus 1, p b 0.01, d = 2.02; diestrus 2, p b 0.01, d = 2.25; proestrus, p b 0.01, d = 2.12;
Table 1 Behavioral responses during the acclimation period in Experiment 1.
Experiment 2 Neither the female subjects' estrus cycles [F(9, 132) = 1.64, p = 0.109, multivariate η2 = 0.228] nor the associates' estrus cycles [F(9, 132) =
Table 2 Behavioral responses during the encounter test in Experiment 2. Group Subject
Associate
Diestrus 1
Diestrus 1 (4) Diestrus 2 (4) Proestrus (5) Estrus (4) Diestrus 1 (4) Diestrus 2 (6) Proestrus (5) Estrus (5) Diestrus 1 (5) Diestrus 2 (5) Proestrus (4) Estrus (4) Diestrus 1 (4) Diestrus 2 (5) Proestrus (4) Estrus (4)
Diestrus 2
Proestrus
Group
Freezing (s)
Investigation (s)
Walking
Unpaired (16) Diestrus 1 (10) Diestrus 2 (10) Proestrus (12) Estrus (8)
6.7 ± 2.3 15.5 ± 6.2 5.4 ± 1.6 5.2 ± 2.0 3.9 ± 2.6
54.7 ± 6.2 44.9 ± 6.8 48.7 ± 6.6 50.8 ± 5.4 59.0 ± 6.9
83.4 ± 6.7 67.8 ± 9.9 85.7 ± 6.6 91.7 ± 7.4 91.9 ± 10.8
Data are expressed as mean ± SEM. The number of animals is shown in parentheses.
Estrus
Freezing (s)
Investigation (s)
Walking
0±0 0±0 0.2 ± 0.1 0.7 ± 0.6 0.5 ± 0.5 0.1 ± 0.1 0.2 ± 0.2 0±0 0±0 0±0 0.1 ± 0.1 0±0 0.1 ± 0.1 0±0 0±0 0.2 ± 0.2
150.3 ± 20.8 161.5 ± 7.7 141.9 ± 23.0 159.2 ± 25.3 184.8 ± 15.6 161.9 ± 17.8 149.9 ± 11.2 168.5 ± 11.8 149.8 ± 12.1 165.8 ± 15.9 197.8 ± 16.0 158.4 ± 10.6 161.8 ± 10.7 155.5 ± 12.0 176.7 ± 12.6 159.8 ± 11.6
191.0 ± 15.0 182.0 ± 19.5 186.2 ± 27.3 177.5 ± 13.4 200.0 ± 10.1 171.5 ± 8.6 138.4 ± 11.4 182.2 ± 13.6 170.2 ± 16.0 138.8 ± 6.4 135.8 ± 9.1 167.0 ± 13.5 177.5 ± 22.0 171.8 ± 15.8 192.5 ± 12.4 182.0 ± 9.4
Data are expressed as mean ± SEM. The number of animals is shown in parentheses.
56
A. Ishii et al. / Hormones and Behavior 81 (2016) 53–58
0.930, p = 0.501, multivariate η2 = 0.139] affected the behavioral responses of the subject during the encounter test (Table 2). There was also no significant interaction between these two factors [F(27, 158) = 1.20, p = 0.240, multivariate η2 = 0.420]. Experiment 3 In the solitary situation, the subjects' sex [F(3, 32) = 15.6, p b 0.01, multivariate η2 = 0.594], but not the conditioning procedure [F(3, 32) = 1.98, p = 0.136, multivariate η2 = 0.157], significantly affected behavioral responses during the acclimation period (Fig. 2A, Table 3). The interaction between these two factors was not significant [F(3, 32) = 0.252, p = 0.860, multivariate η2 = 0.023]. A post-hoc test revealed that the duration of freezing, the duration of investigation, and the frequency of walking were not different between the unpaired and paired subjects for both sexes. The between-sex analyses revealed that the unpaired (p b 0.01, d = 2.30) and paired (p b 0.01, d = 2.22) female subjects carried out a significantly higher frequency of walking compared with the unpaired and paired male subjects, respectively.
Table 3 Behavioral responses during the acclimation period in Experiment 3. Group Solitary situation Unpaired female (7) Paired female (9) Unpaired male (11) Paired male (11) Dyad situation Unpaired female (9) Paired female (9) Unpaired male (8) Paired male (8)
Freezing (s)
Investigation (s)
0.4 ± 0.1 0.8 ± 0.3 5.5 ± 3.8 5.6 ± 2.7
71.0 ± 6.5 72.7 ± 6.4# 53.1 ± 5.6 51.0 ± 6.1
0.2 ± 0.2 0.1 ± 0.0 0.2 ± 0.1 0.3 ± 0.3
96.5 ± 7.1 88.4 ± 5.9 95.2 ± 6.1 80.4 ± 6.7
Walking 96.9 ± 3.4# 80.7 ± 6.4# 54.9 ± 6.6 46.2 ± 3.6 101.3 ± 4.7 97.1 ± 7.6 83.0 ± 5.1 88.4 ± 6.4
Data are expressed as mean ± SEM. The number of animals is shown in parentheses. # p b 0.05 compared with males in the same group, according to two-way MANOVA followed by Fisher's PLSD test.
The duration of investigation was also longer in the paired female subjects compared with the paired male subjects (p b 0.05, d = 1.09).
Fig. 2. Fear and stress responses during the fear-expression test in Experiment 3. (A) Schematic diagram of the solitary situation. S, subject rat. (B) Duration of freezing, duration of investigation, and frequency of walking during the test period in the solitary situation. (C) Corticosterone levels after the fear-expression test in the solitary situation. (D) Schematic diagram of the dyad situation. S, subject rat; A, associate rat. (E) Duration of freezing, duration of investigation, and frequency of walking during the test period in the dyad situation. (F) Corticosterone levels after the fear-expression test in the dyad situation. All data are expressed as mean + SEM. *p b 0.05 compared with the unpaired group within sex and # p b 0.05 compared to males in the same group, according to two-way MANOVA followed by Fisher's PLSD test for behavioral results and two-way ANOVA followed by a Fisher's PLSD test for corticosterone levels.
A. Ishii et al. / Hormones and Behavior 81 (2016) 53–58
Behavioral responses during the test period were affected by the subjects' sex [F(3, 32) = 16.8, p b 0.01, multivariate η2 = 0.612] and the conditioning procedure [F(3, 32) = 58.9, p b 0.01, multivariate η2 = 0.847] (Fig. 2B). The interaction between these two factors was also significant [F(3, 32) = 11.2, p b 0.01, multivariate η2 = 0.513]. A post-hoc test revealed that the paired female subjects had significantly longer freezing durations (p b 0.01, d = 1.85), shorter investigation durations (p b 0.05, d = 0.857), and lower frequency of walking (p b 0.01, d = 1.24) compared with the unpaired female subjects. Similarly, the paired male subjects showed significantly longer freezing durations (p b 0.01, d = 10.6), shorter investigation durations (p b 0.01, d = 2.29), and lower frequency of walking (p b 0.01, d = 2.42) compared with the unpaired male subjects. The between-sex analyses revealed that the unpaired female subjects exhibited higher frequency of walking compared with the unpaired male subjects (p b 0.05, d = 1.23). In addition, the paired female subjects had shorter freezing durations (p b 0.01, d = 2.87), longer investigation durations (p b 0.01, d = 1.35), and higher frequency of walking (p b 0.01, d = 1.88) compared with the paired male subjects. Corticosterone levels after the fear-expression test were significantly affected by the subjects' sex [F(1, 34) = 12.5, p b 0.01, η2 = 0.200] and the conditioning procedure [F(1, 34) = 13.5, p b 0.01, η2 = 0.216], although there was no significant interaction between these two factors [F(1, 2 34) = 1.05, p = 0.313, η = 0.0168] (Fig. 2C). A post-hoc test revealed that the paired female subjects had significantly higher corticosterone levels compared with the unpaired female subjects (p b 0.01, d = 1.07). The paired male subjects also had significantly higher corticosterone levels compared with the unpaired male subjects (p b 0.05, d = 1.77). When we compared the corticosterone levels between the sexes, the paired female subjects showed significantly higher levels of corticosterone compared with the paired male subjects (p b 0.01, d = 1.17). In the dyad situation, the subjects' sex [F(3, 28) = 1.55, p = 0.223, multivariate η2 = 0.143] and the conditioning procedure [F(3, 28) = 1.49, p = 0.238, multivariate η2 = 0.138] did not alter the behavioral responses during the acclimation period (Fig. 2D, Table 3). The interaction between these two factors was not significant [F(3, 28) = 0.409, p = 0.748, multivariate η2 = 0.042]. The behavioral responses during the test period were not affected by the subjects' sex [F(3, 28) = 2.37, p = 0.0914, multivariate η2 = 0.203] or the conditioning procedure [F(3, 28) = 2.31, p = 0.0978, multivariate η2 = 0.198] (Fig. 2E). The interaction between these two factors was also not significant [F(3, 28) = 1.88, p = 0.156, multivariate η2 = 0.168]. Corticosterone levels after the fear-expression test were significantly affected by the subjects' sex [F(1, 30) = 13.0, p b 0.01, η2 = 0.302] (Fig. 2F). However, the effects of the conditioning procedure [F(1, 2 30) = 0.0043, p = 0.948, η = 0.0001] and the interaction between these two factors [F(1, 30) = 0.0305, p = 0.863, η2 = 0.0007] were not significant. Post-hoc testing revealed that the corticosterone levels did not differ between the unpaired and paired female subjects and between the unpaired and paired male subjects. By contrast, betweensex analyses revealed that the unpaired (p b 0.05, d = 1.18) and paired (p b 0.05, d = 1.33) female subjects showed significantly higher corticosterone levels compared with the unpaired and paired male subjects, respectively. Discussion In Experiment 1, we found that the fear responses of paired female subjects showed a similar intensity across all stages of the estrus cycle. These results suggest that the estrus cycle does not affect the intensity of fear responses to a CS. In Experiment 2, the female subjects showed similar behavioral responses across the stages when the subjects encountered an associate that was in any stage. These results suggest that the estrus cycle does not affect the intensity of vigilance attributable to the presence of a conspecific. Therefore, we were able to use
57
female subjects without considering their estrus cycles. Finally, in Experiment 3, we assessed whether social buffering occurs in female rats. We found that the paired female subjects showed fear responses and HPA-axis activation in the solitary situation, and these fear responses were diminished in the dyad situation. Based on these findings, we conclude that social buffering ameliorates conditioned fear responses in female rats. In the present study, the paired and unpaired subjects in dyad situations exhibited a similar intensity of fear responses and HPA-axis activation (Experiment 3). Based on these findings, we suggested that social buffering mitigated conditioned fear responses in female rats. Because the unpaired female subjects also had high levels of corticosterone, we cannot rule out an alternative explanation that a plateau effect may have masked additional activation of the HPA axis in paired female subjects. If this was the case, then social buffering did not occur because the stress of the paired female subjects was not ameliorated. Nonetheless, we believe that this latter interpretation is less probable. The social buffering phenomenon was observed in female subjects in situation in which the male subjects of the same species demonstrated social buffering, although the phenomenon in female and male subjects was not reported in the same study (Hennessy et al., 2006, 2008). Further study is needed to clarify this possibility. In Experiment 3, we also found unexpected sex differences in behavioral and endocrine responses to the auditory CS, which prevented us from evaluating sex differences in social buffering in the present study. The intensity of fear responses was weaker in female subjects compared with male subjects when tested in the solitary situation. In previous studies, the freezing intensity in response to an auditory CS was similar between male and female subjects (Pryce et al., 1999; Markus and Zecevic, 1997). This discrepancy might be due to differences in the characteristics of the fear responses. It has been reported that the characteristics of fear responses primarily in the absence of a CS were completely different from fear responses in the presence of a CS (Kiyokawa et al., 2015). In our study, we observed fear responses mostly in the absence of a CS, since we used a 3 s CS. By contrast, in previous studies, fear responses were observed in the presence of a CS that lasted a few minutes. Therefore, the difference in the duration of the auditory CS may elicit fear responses with different characteristics, which could result in the differences between the present and previous studies. Conversely, corticosterone levels were higher in female subjects than in male subjects. Although sex differences in corticosterone level in response to the auditory CS has not been explored, a previous report using contextual CS showed higher corticosterone levels in female than in male subjects (Daviu et al., 2014). Such contrary sex differences in behavioral and endocrine responses may attribute to sex differences in expressing responses to distressing stimuli. Taken together, it will be necessary to establish an experimental model in which male and female subjects show a similar intensity in their responses in order to evaluate the effect of sex differences on the intensity of social buffering. Another important finding from this study is that the estrus cycle was not an obstacle for analyses using female subjects. In Experiment 1, the intensity of fear responses to an auditory CS was not affected by the estrus cycle. Although the influence of ovarian hormones on fear responses has been reported, most research has observed these behaviors in response to a contextual CS (Barha et al., 2010; Cushman et al., 2014). Therefore, we suggest that the fear responses to an auditory CS are less likely to be affected by the estrus cycle. Indeed, one study reported that rats in the proestrus and estrus stages showed the same intensity of freezing to an auditory CS (Markus and Zecevic, 1997). Similarly, in Experiment 2, the estrus cycle did not affect the vigilance attributable to the presence of an associate. The effects of the estrus cycle on vigilance appear to be dependent on the experimental procedure. Some studies showed differences in locomotion across the estrus cycle in a novel environment (Scimonelli et al., 1999; Petersson et al., 1998), while other studies did not (Hiroi and Neumaier, 2006; Tropp and Markus, 2001). Taken together, the estrus cycle may not be a major concern when
58
A. Ishii et al. / Hormones and Behavior 81 (2016) 53–58
using female subjects, at least in our experimental model. To further support this conclusion, a recent meta-analysis suggested that the coefficient of variation of the responses was comparable between males and females without staging the estrus cycle (Prendergast et al., 2014). In summary, we found that social buffering ameliorates conditioned fear responses in female rats. In addition, we also demonstrated that the estrus cycle does not affect the intensity of fear responses and the degree of vigilance due to conspecific animals, suggesting that the estrus cycle is not an obstacle for the use of female subjects. Future analyses using female rats would further elucidate social buffering in a more comprehensive fashion. This may also help us to understand why the species is gregarious, because we believe that social buffering is one of the biological reasons for gregariousness. Acknowledgements This study was supported by JSPS KAKENHI Grant Numbers 15H05635 and 15H05782. References Barha, C.K., Dalton, G.L., Galea, L.A.M., 2010. Low doses of 17α-estradiol and 17β-estradiol facilitate, whereas higher doses of estrone and 17α- and 17β-estradiol impair, contextual fear conditioning in adult female rats. Neuropsychopharmacology 35, 547–559. da Costa, A.P., Leigh, A.E., Man, M.S., Kendrick, K.M., 2004. Face pictures reduce behavioural, autonomic, endocrine and neural indices of stress and fear in sheep. Proc. Biol. Sci. 271, 2077–2084. Cushman, J.D., Moore, M.D., Olsen, R.W., Fanselow, M.S., 2014. The role of δ GABA(A) receptor in ovarian cycle-linked changes in hippocampus-dependent learning and memory. Neurochem. Res. 39, 1140–1146. Daviu, N., Andero, R., Armario, A., Nadal, R., 2014. Sex differences in the behavioural and hypothalamic-pituitary-adrenal response to contextual fear conditioning in rats. Horm. Behav. 66, 713–723. Fuzzo, F., Matsumoto, J., Kiyokawa, Y., Takeuchi, Y., Ono, T., Nishijo, H., 2015. Social buffering suppresses fear-associated activation of the lateral amygdala in male rats: behavioral and neurophysiological evidence. Front. Neurosci. 9, 99. Hennessy, M.B., Hornschuh, G., Kaiser, S., Sachser, N., 2006. Cortisol responses and social buffering: a study throughout the life span. Horm. Behav. 49, 383–390. Hennessy, M.B., Zate, R., Maken, D.S., 2008. Social buffering of the cortisol response of adult female guinea pigs. Physiol. Behav. 93, 883–888. Hiroi, R., Neumaier, J.F., 2006. Differential effects of ovarian steroids on anxiety versus fear as measured by open field test and fear-potentiated startle. Behav. Brain Res. 166, 93–100.
Kaiser, S., Kirtzeck, M., Hornschuh, G., Sachser, N., 2003. Sex-specific difference in social support―a study in female guinea pigs. Physiol. Behav. 79, 297–303. Kiyokawa, Y., 2015. Social odors: alarm pheromones and social buffering. Curr. Top. Behav. Neurosci. (in press). Kiyokawa, Y., Hiroshima, S., Takeuchi, Y., Mori, Y., 2014a. Social buffering reduces male rats' behavioral and corticosterone responses to a conditioned stimulus. Horm. Behav. 65, 114–118. Kiyokawa, Y., Honda, A., Takeuchi, Y., Mori, Y., 2014b. A familiar conspecific is more effective than an unfamiliar conspecific for social buffering of conditioned fear responses in male rats. Behav. Brain Res. 267, 189–193. Kiyokawa, Y., Mikami, K., Mikamura, Y., Ishii, A., Takeuchi, Y., Mori, Y., 2015. The 3-second auditory conditioned stimulus is a more effective stressor than the 20-second auditory conditioned stimulus in male rats. Neuroscience 299, 79–87. Kiyokawa, Y., Takeuchi, Y., Mori, Y., 2007. Two types of social buffering differentially mitigate conditioned fear responses. Eur. J. Neurosci. 26, 3606–3613. Kiyokawa, Y., Takeuchi, Y., Nishihara, M., Mori, Y., 2009. Main olfactory system mediates social buffering of conditioned fear responses in male rats. Eur. J. Neurosci. 29, 777–785. Kiyokawa, Y., Wakabayashi, Y., Takeuchi, Y., Mori, Y., 2012. The neural pathway underlying social buffering of conditioned fear responses in male rats. Eur. J. Neurosci. 36, 3429–3437. Markus, E.J., Zecevic, M., 1997. Sex differences and estrous cycle changes in hippocampusdependent fear conditioning. Psychobiology 25 (3), 246–252. Morgan, M.A., Pfaff, D.W., 2001. Effects of estrogen on activity and fear-related behaviors in mice. Horm. Behav. 40, 472–482. Petersson, M., Ahlenius, S., Wiberg, U., Alster, P., Uvnäs-Moberg, K., 1998. Steroid dependent effects of oxytocin on spontaneous motor activity in female rats. Brain Res. Bull. 45, 301–305. Prendergast, B.J., Onishi, K.G., Zucker, I., 2014. Female mice liberated for inclusion in neuroscience and biomedical research. Neurosci. Biobehav. Rev. 40, 1–5. Pryce, C.R., Lehmann, J., Feldon, J., 1999. Effect of sex on fear conditioning is similar for context and discrete CS in Wistar, Lewis and Fischer rat strains. Pharmacol. Biochem. Behav. 64, 753–759. Scimonelli, T., Marucco, M., Celis, M.E., 1999. Age-related changes in grooming behavior and motor activity in female rats. Physiol. Behav. 66, 481–484. Smith, A.S., Wang, Z., 2014. Hypothalamic oxytocin mediates social buffering of the stress response. Biol. Psychiatry 76, 281–288. Takahashi, Y., Kiyokawa, Y., Kodama, Y., Arata, S., Takeuchi, Y., Mori, Y., 2013. Olfactory signals mediate social buffering of conditioned fear responses in male rats. Behav. Brain Res. 240, 46–51. Taylor, S.E., Klein, L.C., Lewis, B.P., Gruenewald, T.L., Gurung, R.A., Updegraff, J.A., 2000. Biobehavioral responses to stress in females: tend-and-befriend, not fight-or-flight. Psychol. Rev. 107, 411–429. Terranova, M.L., Cirulli, F., Laviola, G., 1999. Behavioral and hormonal effects of partner familiarity in periadolescent rat pairs upon novelty exposure. Psychoneuroendocrinology 24, 639–656. Tropp, J., Markus, E.J., 2001. Sex differences in the dynamics of cue utilization and exploratory behavior. Behav. Brain Res. 119, 143–154.
Contents lists available at ScienceDirect
Hormones and Behavior journal homepage: www.elsevier.com/locate/yhbeh
Social buffering ameliorates conditioned fear responses in female rats Akiko Ishii, Yasushi Kiyokawa ⁎, Yukari Takeuchi, Yuji Mori Laboratory of Veterinary Ethology, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan
a r t i c l e
i n f o
Article history: Received 13 November 2015 Revised 24 February 2016 Accepted 18 March 2016 Available online 6 April 2016 Keywords: Female Social buffering Estrus cycle Sex difference Fear conditioning
a b s t r a c t The stress experienced by an animal is ameliorated when the animal is exposed to distressing stimuli along with a conspecific animal(s). This is known as social buffering. Previously, we found that the presence of an unfamiliar male rat induced social buffering and ameliorated conditioned fear responses of a male rat subjected to an auditory conditioned stimulus (CS). However, because our knowledge of social buffering is highly biased towards findings in male subjects, analyses using female subjects are crucial for comprehensively understanding the social buffering phenomenon. In the present studies, we assessed social buffering of conditioned fear responses in female rats. We found that the estrus cycle did not affect the intensity of the rats' fear responses to the CS or their degree of vigilance due to the presence of a conspecific animal. Based on these findings, we then assessed whether social buffering ameliorated conditioned fear responses in female rats without taking into account their estrus cycles. When fear conditioned female rats were exposed to the CS without the presence of a conspecific, they exhibited behavioral responses, including freezing, and elevated corticosterone levels. By contrast, the presence of an unfamiliar female rat suppressed these responses. Based on these findings, we conclude that social buffering can ameliorate conditioned fear responses in female rats. © 2016 Elsevier Inc. All rights reserved.
Introduction The stress experienced by a subject can be ameliorated when the subject is exposed to distressing stimuli while in the company of a conspecific animal. This phenomenon is called social buffering and can be induced by a mother, mate, or a same-sex or opposite-sex conspecific animal in a nonsexual relationship (Kiyokawa, in press). The phenomenon elicited by the last type of conspecific has been observed in a wide variety of social species, including sheep (da Costa et al., 2004), guinea pigs (Hennessy et al., 2006, 2008), and rats (Terranova et al., 1999). Previously, we conducted a series of experiments focused on the social buffering of conditioned fear responses in rats. In this model, male subject rats exhibited robust freezing and activation of the hypothalamic-pituitary-adrenal (HPA) axis in response to an auditory conditioned stimulus (CS), when the CS was paired with a foot shock during fear conditioning. The presence of an unfamiliar nonconditioned male rat (an associate) ameliorated these responses (Kiyokawa et al., 2007). Subsequently, we found that this social buffering occurred even when the subject and associate were separated by wire-mesh or double wire-mesh partitions (Kiyokawa et al., 2009, 2014a). Given that social buffering was inhibited by lesioning the main olfactory epithelium (Kiyokawa et al., 2009) and that associate-derived olfactory cues alone are able to induce social buffering (Takahashi et al., 2013; Kiyokawa et al., 2014b), we suggest that olfactory signaling mediates social buffering of conditioned ⁎ Corresponding author. E-mail address: [email protected] (Y. Kiyokawa).
http://dx.doi.org/10.1016/j.yhbeh.2016.03.003 0018-506X/© 2016 Elsevier Inc. All rights reserved.
fear responses. Recent studies have shed light on the neural mechanisms underlying social buffering, such as the suppression of the basolateral complex of the amygdala (Fuzzo et al., 2015) and the involvement of the posterior complex of the anterior olfactory nucleus as a relay point for signaling from the olfactory bulb to the amygdala (Kiyokawa et al., 2012; Takahashi et al., 2013). Our knowledge of social buffering is highly biased towards findings in male subjects, because all our previous studies, as well as most studies in the literature, have been conducted using males. Although some studies have reported the social buffering phenomenon in female subjects, the phenomenon was induced by a mate (Kaiser et al., 2003; Hennessy et al., 2008; Smith and Wang, 2014). Given that the neural mechanisms underlying social buffering differ depending on the type of conspecific animal, i.e., a mother, mate, or conspecific without sexual relationships, it would be appropriate to understand each phenomenon individually (Kiyokawa, in press). To the best of our knowledge, only one study using female guinea pigs reported a clear social buffering phenomenon by a conspecific animal without sexual relationships (a female guinea pig) (Hennessy et al., 2008). In addition, sex differences are thought to play a large role in stress responses. For example, males tend to show a “fight and flight” response to distressing stimuli, while females tend to exhibit a “tend and befriend” response (Taylor et al., 2000). Therefore, analyses using female subjects seem to be necessary to obtain a more comprehensive understanding of the social buffering phenomenon, especially that induced by conspecifics without sexual relationships. Changes in ovarian hormones that are dependent on the estrus cycle, such as estrogen and progesterone, appear to be one of the
54
A. Ishii et al. / Hormones and Behavior 81 (2016) 53–58
major challenges in analyzing social buffering of conditioned fear responses in female rats. First, these hormones affect the intensity of conditioned fear responses (Morgan and Pfaff, 2001). Given that the observed responses in the presence of a conspecific are residual intrinsic CS-induced responses after suppression by social buffering, we cannot appropriately evaluate the degree of social buffering if the intensity of intrinsic CS-induced responses fluctuates according to the estrus cycle. Second, these hormones can affect the degree of vigilance (Petersson et al., 1998), which may cause difference in the intensity of vigilance attributable to the presence of a conspecific. Fluctuations in vigilance due to the estrus cycle would also prevent us from appropriately evaluating social buffering, because vigilance can affect behavioral measures that require stillness and/or physiological measures related to metabolism, such as freezing and/or HPA axis activity, respectively. Therefore, it was necessary to assess the effects of the estrus cycle on these two factors. In the present study, we conducted a series of experiments to assess social buffering of conditioned fear responses in female rats. In Experiment 1, we assessed the effects of the estrus cycle on the intensity of conditioned fear responses. Fear conditioned and non-conditioned female subjects were exposed to an auditory CS while alone in the testing apparatus (solitary situation). Behavioral responses, including freezing, were compared across the stages. In Experiment 2, we assessed the effects of the estrus cycle on vigilance. Female subjects at all stages of the estrus cycle encountered unfamiliar female associates who were also in all stages of the cycle. Their behavioral responses were analyzed, including locomotor activity. In Experiment 3, we assessed whether social buffering ameliorated conditioned fear responses in female rats. Fear conditioned or non-conditioned female subjects were exposed to the auditory CS either alone (solitary situation) or with an unfamiliar female associate (dyad situation). Their behavioral responses, including freezing, and corticosterone levels were measured in order to evaluate the presence of social buffering. We conducted parallel experiments using male rats in order to evaluate sex differences in social buffering. Material and methods Animals All experiments were approved by the Animal Care and Use Committee of the Faculty of Agriculture of The University of Tokyo and were based on guidelines that were adapted from the Consensus Recommendations on Effective Institutional Animal Care and Use Committees by the Scientists Center for Animal Welfare. Experimentally naïve female and male Wistar rats were purchased from Charles River Laboratories Japan (Kanagawa, Japan) at 7.5 weeks of age. Animals were housed 2–3 rats per cage upon arrival and under conditions of controlled temperature (24 ± 1 °C) and humidity (45 ± 5%). The housing room had a 12-h light/12-h dark cycle (lights on at 8:00). Water and food were available ad libitum. All behavioral procedures in Experiments 1 and 2 were performed between 8:00 and 15:00. The behavioral procedures in Experiment 3 were performed between 8:30 and 13:00 (see Experiment 3). Experiment 1 Female subjects were housed individually 3 days before fear conditioning until the test day and were handled for 3 min twice daily until the day of conditioning. A vaginal smear was taken every morning beginning on the day after arrival to evaluate the stage of the estrus cycle. Fear conditioning was performed under white-light conditions, as previously described (Kiyokawa et al., 2009). Subjects were placed in an acrylic conditioning box (28 × 20 × 27 cm) for 20 min. During fear conditioning, paired subjects were exposed to 7 pairings of a 3-s auditory CS (8 kHz, 70 dB) and a 0.5 s foot shock (0.80 mA), which ended simultaneously. Unpaired subjects received the same number of CS
and foot shocks, but these were presented separately. The inter-trial interval varied randomly between 60 and 180 s. After conditioning, subjects were returned to their home cages. The fear-expression test was performed 24 h after conditioning under dim red light, as previously described (Kiyokawa et al., 2009). Two rectangular enclosures (25 × 25 × 35 cm) were placed on an acrylic board (45 × 60 cm). Each enclosure consisted of 3 acrylic walls, 1 wire mesh wall, and a wire mesh ceiling. The inside of the enclosure was covered with clean bedding. The enclosures were placed side-by-side so that the wire mesh walls were facing each other, separated by a 5 cm gap. The wire mesh wall was composed of a 1 cm2 grid in the lower half (20 cm) and vertical bars in the upper half (15 cm) to prevent subjects from climbing the wire-mesh wall. Subjects were placed randomly in one of the two enclosures. After a 3-min acclimation period, the 3-s CS was presented 5 times at intervals of 1 min during the first 5 min of the 10-min test period. The subjects' behavior during the fearexpression test was recorded with an HDD-BD recorder (DMRBW770; Panasonic, Osaka) and a video camera (DCR-SR 300; Sony, Tokyo). The paired subjects were divided into 4 groups according to their estrus cycle at the time of the fear-expression test (diestrus 1, n = 10; diestrus 2, n = 10; proestrus, n = 8; estrus, n = 12). Because the estrus cycle did not alter the behavioral responses of the unpaired female subjects during the fear-expression test, all the unpaired subjects (diestrus 1, n = 4; diestrus 2, n = 4; proestrus, n = 5; estrus, n = 3) were combined into 1 group, regardless of their estrus cycle. All data are presented as the mean ± standard error of the mean (SEM). The significance level was set at p b 0.05 for all statistical tests. Behavioral parameters during the acclimation and test period were measured using Visual Basic software in Microsoft Excel, which recorded the duration and number of pressed keys. The duration of freezing (the lack of any movement except that which is required for respiration), the duration of investigation (sniffing towards the wire mesh wall within 1 cm or poking of the snout towards the wire mesh), and the frequency of walking (number of steps taken with the hind paws) during the preceding acclimation period and during the test period were analyzed by multivariate analysis of variance (MANOVA), followed by Fisher's PLSD post-hoc test. Effect sizes were further estimated by calculating the value of multivariate η2 and Cohen's d for data analyzed by MANOVA and Fisher's PLSD post-hoc test, respectively. Experiment 2 The encounter test was conducted in the same enclosures under dim red light as described above. In the encounter test, a female rat was placed in each enclosure for 5 min to observe the behavior to a novel animal, which was recorded with an HDD-BD recorder and a video camera. Each rat served both as the subject and also as an associate for the other. Based on the subject's and the associate's estrus cycle, we prepared 16 groups (n = 4–6 per group). Rats were housed individually until the test day, and were handled for 3 min twice daily for 3 days. The duration of freezing, the duration of investigation, and the frequency of walking were analyzed by two-way MANOVA. Effect sizes were further estimated by calculating the value of multivariate η2 for data analyzed by MANOVA. Experiment 3 Fear conditioning was performed as described in Experiment 1. Female and male rats assigned as subjects underwent fear conditioning, while rats assigned as associates remained in their home cages during fear conditioning. Subjects were further assigned as either paired or unpaired subjects. Because we found that the estrus cycle did not affect the intensity of the rats' fear responses to the CS or their degree of vigilance due to the presence of a conspecific animal, we did not assess the stage
A. Ishii et al. / Hormones and Behavior 81 (2016) 53–58
of the estrus cycle in female rats. All rats were housed individually until the test day and handled for 5 min daily for 3 days. The fear-expression test was conducted as described in Experiment 1, except that subject rats were tested either alone (solitary situation) or with an associate (dyad situation). When tested in the solitary situation, the subject was randomly placed in one of the enclosures, while the other enclosure was left vacant, similar to Experiment 1. In the dyad situation, a subject and a same-sex associate were each placed in a separate enclosure. The subjects' behavior was recorded with an HDD-BD recorder and a video camera. Corticosterone levels were measured as previously described (Kiyokawa et al., 2014a). After the fear-expression test, the subjects (and also the associates when tested in the dyad situation) were left undisturbed for an additional 10 min. Blood was then collected in hematocrit tubes from a small incision made by a razor on the subjects' tails. This sampling was completed within 3 min. Blood sampling was performed between 8:30 and 13:00 to take into account diurnal rhythms. The collected samples were centrifuged at 4 °C; plasma was stored at −80 °C. Corticosterone levels from each sample were mostly measured in duplicate (in triplicate, solitary situation: unpaired male, n = 1; paired male, n = 1) using the Corticosterone EIA kit (Cayman Chemical Company, Ann Arbor, MI). Subjects were divided into 8 groups based on their testing situation, sex, and conditioning procedure (solitary situation: unpaired females, n = 7; paired females, n = 9; unpaired males, n = 11; paired males, n = 11; dyad situation: unpaired females, n = 9; paired females, n = 9; unpaired females, n = 8; paired males, n = 8) (Fig. 2A, D). The duration of freezing, the duration of investigation, and the frequency of walking during the preceding acclimation period and the test period were analyzed by two-way MANOVA followed by Fisher's PLSD post-hoc test. The mean corticosterone levels were analyzed by two-way analysis of variance (ANOVA) followed by Fisher's PLSD post-hoc test. Effect sizes were further estimated by calculating the value of multivariate η2, η2, and Cohen's d for data analyzed by MANOVA, ANOVA and Fisher's PLSD post-hoc test, respectively.
55
Fig. 1. Fear responses during the fear-expression test in Experiment 1. Duration of freezing, duration of investigation, and frequency of walking (mean + SEM) during the test period. The letters indicate significant differences (p b 0.05), as assessed by MANOVA followed by Fisher's PLSD test.
estrus, p b 0.01, d = 1.74), and significantly lower frequency of walking (diestrus 1, p b 0.01, d = 1.51; diestrus 2, p b 0.01, d = 1.17; proestrus, p b 0.01, d = 1.31; estrus, p b 0.01, d = 2.08) compared with the unpaired subjects. However, the estrus cycle did not affect the behavioral responses of the paired subjects.
Results Experiment 1 Because the estrus cycle did not alter the behavioral responses of the unpaired female subjects during the acclimation period and test period, we combined these subjects into the unpaired group. The behavioral responses during the acclimation period did not differ between the groups [F(12, 130) = 0.781, p = 0.669, multivariate η2 = 0.168] (Table 1). Although the paired subjects in diestrus 1 appeared to show longer freezing durations, this was because 2 animals showed unnaturally long freezing durations. In contrast, behaviors during the test period varied significantly between the groups [F(12, 130) = 5.53, p b 0.01, multivariate η2 = 0.664] (Fig. 1). A post-hoc test revealed that the paired subjects showed significantly longer freezing durations (diestrus 1, p b 0.01, d = 2.47; diestrus 2, p b 0.01, d = 2.48; proestrus, p b 0.01, d = 2.61; estrus, p b 0.01, d = 2.86), significantly shorter investigation durations (diestrus 1, p b 0.01, d = 2.02; diestrus 2, p b 0.01, d = 2.25; proestrus, p b 0.01, d = 2.12;
Table 1 Behavioral responses during the acclimation period in Experiment 1.
Experiment 2 Neither the female subjects' estrus cycles [F(9, 132) = 1.64, p = 0.109, multivariate η2 = 0.228] nor the associates' estrus cycles [F(9, 132) =
Table 2 Behavioral responses during the encounter test in Experiment 2. Group Subject
Associate
Diestrus 1
Diestrus 1 (4) Diestrus 2 (4) Proestrus (5) Estrus (4) Diestrus 1 (4) Diestrus 2 (6) Proestrus (5) Estrus (5) Diestrus 1 (5) Diestrus 2 (5) Proestrus (4) Estrus (4) Diestrus 1 (4) Diestrus 2 (5) Proestrus (4) Estrus (4)
Diestrus 2
Proestrus
Group
Freezing (s)
Investigation (s)
Walking
Unpaired (16) Diestrus 1 (10) Diestrus 2 (10) Proestrus (12) Estrus (8)
6.7 ± 2.3 15.5 ± 6.2 5.4 ± 1.6 5.2 ± 2.0 3.9 ± 2.6
54.7 ± 6.2 44.9 ± 6.8 48.7 ± 6.6 50.8 ± 5.4 59.0 ± 6.9
83.4 ± 6.7 67.8 ± 9.9 85.7 ± 6.6 91.7 ± 7.4 91.9 ± 10.8
Data are expressed as mean ± SEM. The number of animals is shown in parentheses.
Estrus
Freezing (s)
Investigation (s)
Walking
0±0 0±0 0.2 ± 0.1 0.7 ± 0.6 0.5 ± 0.5 0.1 ± 0.1 0.2 ± 0.2 0±0 0±0 0±0 0.1 ± 0.1 0±0 0.1 ± 0.1 0±0 0±0 0.2 ± 0.2
150.3 ± 20.8 161.5 ± 7.7 141.9 ± 23.0 159.2 ± 25.3 184.8 ± 15.6 161.9 ± 17.8 149.9 ± 11.2 168.5 ± 11.8 149.8 ± 12.1 165.8 ± 15.9 197.8 ± 16.0 158.4 ± 10.6 161.8 ± 10.7 155.5 ± 12.0 176.7 ± 12.6 159.8 ± 11.6
191.0 ± 15.0 182.0 ± 19.5 186.2 ± 27.3 177.5 ± 13.4 200.0 ± 10.1 171.5 ± 8.6 138.4 ± 11.4 182.2 ± 13.6 170.2 ± 16.0 138.8 ± 6.4 135.8 ± 9.1 167.0 ± 13.5 177.5 ± 22.0 171.8 ± 15.8 192.5 ± 12.4 182.0 ± 9.4
Data are expressed as mean ± SEM. The number of animals is shown in parentheses.
56
A. Ishii et al. / Hormones and Behavior 81 (2016) 53–58
0.930, p = 0.501, multivariate η2 = 0.139] affected the behavioral responses of the subject during the encounter test (Table 2). There was also no significant interaction between these two factors [F(27, 158) = 1.20, p = 0.240, multivariate η2 = 0.420]. Experiment 3 In the solitary situation, the subjects' sex [F(3, 32) = 15.6, p b 0.01, multivariate η2 = 0.594], but not the conditioning procedure [F(3, 32) = 1.98, p = 0.136, multivariate η2 = 0.157], significantly affected behavioral responses during the acclimation period (Fig. 2A, Table 3). The interaction between these two factors was not significant [F(3, 32) = 0.252, p = 0.860, multivariate η2 = 0.023]. A post-hoc test revealed that the duration of freezing, the duration of investigation, and the frequency of walking were not different between the unpaired and paired subjects for both sexes. The between-sex analyses revealed that the unpaired (p b 0.01, d = 2.30) and paired (p b 0.01, d = 2.22) female subjects carried out a significantly higher frequency of walking compared with the unpaired and paired male subjects, respectively.
Table 3 Behavioral responses during the acclimation period in Experiment 3. Group Solitary situation Unpaired female (7) Paired female (9) Unpaired male (11) Paired male (11) Dyad situation Unpaired female (9) Paired female (9) Unpaired male (8) Paired male (8)
Freezing (s)
Investigation (s)
0.4 ± 0.1 0.8 ± 0.3 5.5 ± 3.8 5.6 ± 2.7
71.0 ± 6.5 72.7 ± 6.4# 53.1 ± 5.6 51.0 ± 6.1
0.2 ± 0.2 0.1 ± 0.0 0.2 ± 0.1 0.3 ± 0.3
96.5 ± 7.1 88.4 ± 5.9 95.2 ± 6.1 80.4 ± 6.7
Walking 96.9 ± 3.4# 80.7 ± 6.4# 54.9 ± 6.6 46.2 ± 3.6 101.3 ± 4.7 97.1 ± 7.6 83.0 ± 5.1 88.4 ± 6.4
Data are expressed as mean ± SEM. The number of animals is shown in parentheses. # p b 0.05 compared with males in the same group, according to two-way MANOVA followed by Fisher's PLSD test.
The duration of investigation was also longer in the paired female subjects compared with the paired male subjects (p b 0.05, d = 1.09).
Fig. 2. Fear and stress responses during the fear-expression test in Experiment 3. (A) Schematic diagram of the solitary situation. S, subject rat. (B) Duration of freezing, duration of investigation, and frequency of walking during the test period in the solitary situation. (C) Corticosterone levels after the fear-expression test in the solitary situation. (D) Schematic diagram of the dyad situation. S, subject rat; A, associate rat. (E) Duration of freezing, duration of investigation, and frequency of walking during the test period in the dyad situation. (F) Corticosterone levels after the fear-expression test in the dyad situation. All data are expressed as mean + SEM. *p b 0.05 compared with the unpaired group within sex and # p b 0.05 compared to males in the same group, according to two-way MANOVA followed by Fisher's PLSD test for behavioral results and two-way ANOVA followed by a Fisher's PLSD test for corticosterone levels.
A. Ishii et al. / Hormones and Behavior 81 (2016) 53–58
Behavioral responses during the test period were affected by the subjects' sex [F(3, 32) = 16.8, p b 0.01, multivariate η2 = 0.612] and the conditioning procedure [F(3, 32) = 58.9, p b 0.01, multivariate η2 = 0.847] (Fig. 2B). The interaction between these two factors was also significant [F(3, 32) = 11.2, p b 0.01, multivariate η2 = 0.513]. A post-hoc test revealed that the paired female subjects had significantly longer freezing durations (p b 0.01, d = 1.85), shorter investigation durations (p b 0.05, d = 0.857), and lower frequency of walking (p b 0.01, d = 1.24) compared with the unpaired female subjects. Similarly, the paired male subjects showed significantly longer freezing durations (p b 0.01, d = 10.6), shorter investigation durations (p b 0.01, d = 2.29), and lower frequency of walking (p b 0.01, d = 2.42) compared with the unpaired male subjects. The between-sex analyses revealed that the unpaired female subjects exhibited higher frequency of walking compared with the unpaired male subjects (p b 0.05, d = 1.23). In addition, the paired female subjects had shorter freezing durations (p b 0.01, d = 2.87), longer investigation durations (p b 0.01, d = 1.35), and higher frequency of walking (p b 0.01, d = 1.88) compared with the paired male subjects. Corticosterone levels after the fear-expression test were significantly affected by the subjects' sex [F(1, 34) = 12.5, p b 0.01, η2 = 0.200] and the conditioning procedure [F(1, 34) = 13.5, p b 0.01, η2 = 0.216], although there was no significant interaction between these two factors [F(1, 2 34) = 1.05, p = 0.313, η = 0.0168] (Fig. 2C). A post-hoc test revealed that the paired female subjects had significantly higher corticosterone levels compared with the unpaired female subjects (p b 0.01, d = 1.07). The paired male subjects also had significantly higher corticosterone levels compared with the unpaired male subjects (p b 0.05, d = 1.77). When we compared the corticosterone levels between the sexes, the paired female subjects showed significantly higher levels of corticosterone compared with the paired male subjects (p b 0.01, d = 1.17). In the dyad situation, the subjects' sex [F(3, 28) = 1.55, p = 0.223, multivariate η2 = 0.143] and the conditioning procedure [F(3, 28) = 1.49, p = 0.238, multivariate η2 = 0.138] did not alter the behavioral responses during the acclimation period (Fig. 2D, Table 3). The interaction between these two factors was not significant [F(3, 28) = 0.409, p = 0.748, multivariate η2 = 0.042]. The behavioral responses during the test period were not affected by the subjects' sex [F(3, 28) = 2.37, p = 0.0914, multivariate η2 = 0.203] or the conditioning procedure [F(3, 28) = 2.31, p = 0.0978, multivariate η2 = 0.198] (Fig. 2E). The interaction between these two factors was also not significant [F(3, 28) = 1.88, p = 0.156, multivariate η2 = 0.168]. Corticosterone levels after the fear-expression test were significantly affected by the subjects' sex [F(1, 30) = 13.0, p b 0.01, η2 = 0.302] (Fig. 2F). However, the effects of the conditioning procedure [F(1, 2 30) = 0.0043, p = 0.948, η = 0.0001] and the interaction between these two factors [F(1, 30) = 0.0305, p = 0.863, η2 = 0.0007] were not significant. Post-hoc testing revealed that the corticosterone levels did not differ between the unpaired and paired female subjects and between the unpaired and paired male subjects. By contrast, betweensex analyses revealed that the unpaired (p b 0.05, d = 1.18) and paired (p b 0.05, d = 1.33) female subjects showed significantly higher corticosterone levels compared with the unpaired and paired male subjects, respectively. Discussion In Experiment 1, we found that the fear responses of paired female subjects showed a similar intensity across all stages of the estrus cycle. These results suggest that the estrus cycle does not affect the intensity of fear responses to a CS. In Experiment 2, the female subjects showed similar behavioral responses across the stages when the subjects encountered an associate that was in any stage. These results suggest that the estrus cycle does not affect the intensity of vigilance attributable to the presence of a conspecific. Therefore, we were able to use
57
female subjects without considering their estrus cycles. Finally, in Experiment 3, we assessed whether social buffering occurs in female rats. We found that the paired female subjects showed fear responses and HPA-axis activation in the solitary situation, and these fear responses were diminished in the dyad situation. Based on these findings, we conclude that social buffering ameliorates conditioned fear responses in female rats. In the present study, the paired and unpaired subjects in dyad situations exhibited a similar intensity of fear responses and HPA-axis activation (Experiment 3). Based on these findings, we suggested that social buffering mitigated conditioned fear responses in female rats. Because the unpaired female subjects also had high levels of corticosterone, we cannot rule out an alternative explanation that a plateau effect may have masked additional activation of the HPA axis in paired female subjects. If this was the case, then social buffering did not occur because the stress of the paired female subjects was not ameliorated. Nonetheless, we believe that this latter interpretation is less probable. The social buffering phenomenon was observed in female subjects in situation in which the male subjects of the same species demonstrated social buffering, although the phenomenon in female and male subjects was not reported in the same study (Hennessy et al., 2006, 2008). Further study is needed to clarify this possibility. In Experiment 3, we also found unexpected sex differences in behavioral and endocrine responses to the auditory CS, which prevented us from evaluating sex differences in social buffering in the present study. The intensity of fear responses was weaker in female subjects compared with male subjects when tested in the solitary situation. In previous studies, the freezing intensity in response to an auditory CS was similar between male and female subjects (Pryce et al., 1999; Markus and Zecevic, 1997). This discrepancy might be due to differences in the characteristics of the fear responses. It has been reported that the characteristics of fear responses primarily in the absence of a CS were completely different from fear responses in the presence of a CS (Kiyokawa et al., 2015). In our study, we observed fear responses mostly in the absence of a CS, since we used a 3 s CS. By contrast, in previous studies, fear responses were observed in the presence of a CS that lasted a few minutes. Therefore, the difference in the duration of the auditory CS may elicit fear responses with different characteristics, which could result in the differences between the present and previous studies. Conversely, corticosterone levels were higher in female subjects than in male subjects. Although sex differences in corticosterone level in response to the auditory CS has not been explored, a previous report using contextual CS showed higher corticosterone levels in female than in male subjects (Daviu et al., 2014). Such contrary sex differences in behavioral and endocrine responses may attribute to sex differences in expressing responses to distressing stimuli. Taken together, it will be necessary to establish an experimental model in which male and female subjects show a similar intensity in their responses in order to evaluate the effect of sex differences on the intensity of social buffering. Another important finding from this study is that the estrus cycle was not an obstacle for analyses using female subjects. In Experiment 1, the intensity of fear responses to an auditory CS was not affected by the estrus cycle. Although the influence of ovarian hormones on fear responses has been reported, most research has observed these behaviors in response to a contextual CS (Barha et al., 2010; Cushman et al., 2014). Therefore, we suggest that the fear responses to an auditory CS are less likely to be affected by the estrus cycle. Indeed, one study reported that rats in the proestrus and estrus stages showed the same intensity of freezing to an auditory CS (Markus and Zecevic, 1997). Similarly, in Experiment 2, the estrus cycle did not affect the vigilance attributable to the presence of an associate. The effects of the estrus cycle on vigilance appear to be dependent on the experimental procedure. Some studies showed differences in locomotion across the estrus cycle in a novel environment (Scimonelli et al., 1999; Petersson et al., 1998), while other studies did not (Hiroi and Neumaier, 2006; Tropp and Markus, 2001). Taken together, the estrus cycle may not be a major concern when
58
A. Ishii et al. / Hormones and Behavior 81 (2016) 53–58
using female subjects, at least in our experimental model. To further support this conclusion, a recent meta-analysis suggested that the coefficient of variation of the responses was comparable between males and females without staging the estrus cycle (Prendergast et al., 2014). In summary, we found that social buffering ameliorates conditioned fear responses in female rats. In addition, we also demonstrated that the estrus cycle does not affect the intensity of fear responses and the degree of vigilance due to conspecific animals, suggesting that the estrus cycle is not an obstacle for the use of female subjects. Future analyses using female rats would further elucidate social buffering in a more comprehensive fashion. This may also help us to understand why the species is gregarious, because we believe that social buffering is one of the biological reasons for gregariousness. Acknowledgements This study was supported by JSPS KAKENHI Grant Numbers 15H05635 and 15H05782. References Barha, C.K., Dalton, G.L., Galea, L.A.M., 2010. Low doses of 17α-estradiol and 17β-estradiol facilitate, whereas higher doses of estrone and 17α- and 17β-estradiol impair, contextual fear conditioning in adult female rats. Neuropsychopharmacology 35, 547–559. da Costa, A.P., Leigh, A.E., Man, M.S., Kendrick, K.M., 2004. Face pictures reduce behavioural, autonomic, endocrine and neural indices of stress and fear in sheep. Proc. Biol. Sci. 271, 2077–2084. Cushman, J.D., Moore, M.D., Olsen, R.W., Fanselow, M.S., 2014. The role of δ GABA(A) receptor in ovarian cycle-linked changes in hippocampus-dependent learning and memory. Neurochem. Res. 39, 1140–1146. Daviu, N., Andero, R., Armario, A., Nadal, R., 2014. Sex differences in the behavioural and hypothalamic-pituitary-adrenal response to contextual fear conditioning in rats. Horm. Behav. 66, 713–723. Fuzzo, F., Matsumoto, J., Kiyokawa, Y., Takeuchi, Y., Ono, T., Nishijo, H., 2015. Social buffering suppresses fear-associated activation of the lateral amygdala in male rats: behavioral and neurophysiological evidence. Front. Neurosci. 9, 99. Hennessy, M.B., Hornschuh, G., Kaiser, S., Sachser, N., 2006. Cortisol responses and social buffering: a study throughout the life span. Horm. Behav. 49, 383–390. Hennessy, M.B., Zate, R., Maken, D.S., 2008. Social buffering of the cortisol response of adult female guinea pigs. Physiol. Behav. 93, 883–888. Hiroi, R., Neumaier, J.F., 2006. Differential effects of ovarian steroids on anxiety versus fear as measured by open field test and fear-potentiated startle. Behav. Brain Res. 166, 93–100.
Kaiser, S., Kirtzeck, M., Hornschuh, G., Sachser, N., 2003. Sex-specific difference in social support―a study in female guinea pigs. Physiol. Behav. 79, 297–303. Kiyokawa, Y., 2015. Social odors: alarm pheromones and social buffering. Curr. Top. Behav. Neurosci. (in press). Kiyokawa, Y., Hiroshima, S., Takeuchi, Y., Mori, Y., 2014a. Social buffering reduces male rats' behavioral and corticosterone responses to a conditioned stimulus. Horm. Behav. 65, 114–118. Kiyokawa, Y., Honda, A., Takeuchi, Y., Mori, Y., 2014b. A familiar conspecific is more effective than an unfamiliar conspecific for social buffering of conditioned fear responses in male rats. Behav. Brain Res. 267, 189–193. Kiyokawa, Y., Mikami, K., Mikamura, Y., Ishii, A., Takeuchi, Y., Mori, Y., 2015. The 3-second auditory conditioned stimulus is a more effective stressor than the 20-second auditory conditioned stimulus in male rats. Neuroscience 299, 79–87. Kiyokawa, Y., Takeuchi, Y., Mori, Y., 2007. Two types of social buffering differentially mitigate conditioned fear responses. Eur. J. Neurosci. 26, 3606–3613. Kiyokawa, Y., Takeuchi, Y., Nishihara, M., Mori, Y., 2009. Main olfactory system mediates social buffering of conditioned fear responses in male rats. Eur. J. Neurosci. 29, 777–785. Kiyokawa, Y., Wakabayashi, Y., Takeuchi, Y., Mori, Y., 2012. The neural pathway underlying social buffering of conditioned fear responses in male rats. Eur. J. Neurosci. 36, 3429–3437. Markus, E.J., Zecevic, M., 1997. Sex differences and estrous cycle changes in hippocampusdependent fear conditioning. Psychobiology 25 (3), 246–252. Morgan, M.A., Pfaff, D.W., 2001. Effects of estrogen on activity and fear-related behaviors in mice. Horm. Behav. 40, 472–482. Petersson, M., Ahlenius, S., Wiberg, U., Alster, P., Uvnäs-Moberg, K., 1998. Steroid dependent effects of oxytocin on spontaneous motor activity in female rats. Brain Res. Bull. 45, 301–305. Prendergast, B.J., Onishi, K.G., Zucker, I., 2014. Female mice liberated for inclusion in neuroscience and biomedical research. Neurosci. Biobehav. Rev. 40, 1–5. Pryce, C.R., Lehmann, J., Feldon, J., 1999. Effect of sex on fear conditioning is similar for context and discrete CS in Wistar, Lewis and Fischer rat strains. Pharmacol. Biochem. Behav. 64, 753–759. Scimonelli, T., Marucco, M., Celis, M.E., 1999. Age-related changes in grooming behavior and motor activity in female rats. Physiol. Behav. 66, 481–484. Smith, A.S., Wang, Z., 2014. Hypothalamic oxytocin mediates social buffering of the stress response. Biol. Psychiatry 76, 281–288. Takahashi, Y., Kiyokawa, Y., Kodama, Y., Arata, S., Takeuchi, Y., Mori, Y., 2013. Olfactory signals mediate social buffering of conditioned fear responses in male rats. Behav. Brain Res. 240, 46–51. Taylor, S.E., Klein, L.C., Lewis, B.P., Gruenewald, T.L., Gurung, R.A., Updegraff, J.A., 2000. Biobehavioral responses to stress in females: tend-and-befriend, not fight-or-flight. Psychol. Rev. 107, 411–429. Terranova, M.L., Cirulli, F., Laviola, G., 1999. Behavioral and hormonal effects of partner familiarity in periadolescent rat pairs upon novelty exposure. Psychoneuroendocrinology 24, 639–656. Tropp, J., Markus, E.J., 2001. Sex differences in the dynamics of cue utilization and exploratory behavior. Behav. Brain Res. 119, 143–154.