Behavioural Brain Research 263 (2014) 16–21
Contents lists available at ScienceDirect
Behavioural Brain Research journal homepage: www.elsevier.com/locate/bbr
Research report
Emotional contagion in mice: The role of familiarity Cristina Gonzalez-Liencres a,b,∗ , Georg Juckel a,b , Cumhur Tas a,c , Astrid Friebe a,1 , Martin Brüne a,b,1 a b c
Department of Psychiatry, Ruhr University-Bochum, LWL University Hosiptal, Germany International Graduate School of Neuroscience, Ruhr-University Bochum, Germany Department of Psychology, Üsküdar University, Istanbul, Turkey
h i g h l i g h t s • • • • •
Familiar and unfamiliar mice observed conspecifics receiving electric foot shocks. Familiar observers freeze more than strangers when witnessing another in distress. Demonstrators receiving shocks freeze the most aftershocks have ceased. Observers stop freezing when shocks have ceased in the adjacent compartment. Demonstrators show less freezing behavior when observer is familiar.
a r t i c l e
i n f o
Article history: Received 2 December 2013 Received in revised form 17 January 2014 Accepted 19 January 2014 Available online 28 January 2014 Keywords: Empathy Observer Demonstrator Observational fear Shock
a b s t r a c t Empathy is a complex emotional process that involves sharing an emotional state with another organism. The extent to which nonhuman animals are capable of empathizing with others is still far from clear, partly due to a lack of empirical work in this domain, but also due to definitional confusion of empathy with emotional contagion and other related terms. In this study, an observer mouse witnessed a familiar cagemate or an unfamiliar non-cagemate receiving electric foot shocks in an experiment that consisted of three periods: baseline (no shocks), test (shocks) and recovery (no shocks). Freezing behavior in the observer was significantly increased in the cagemate, as opposed to the non-cagemate condition during the test period, but not during baseline or recovery, emphasizing the role of familiarity in empathy-like processes. In agreement with this, we also found a correlation that approached significance between the total number of fecal droppings of the observers, as an indication of distress, and those of the demonstrator in the cagemate, but not in the non-cagemate, condition. While the freezing behavior of the demonstrators increased with time, reaching a maximum at the recovery period, the observers froze the most during the test period while the demonstrators were receiving the electric foot shocks. The observation that the freezing response of the observers ceased when the shocks in the adjacent compartment stopped could be due to a decrease in saliency of the demonstrators’ behavioral response. Finally, the presence of a cagemate, as compared to a stranger, possibly reduced the demonstrator’s pain-induced behavior, suggesting an ameliorating effect of familiarity on stress responses. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Empathic concern for others is rooted in or related to parental care for altricial offspring [1,2], whereby empathy – defined as the ability to share the emotions of others – evolved from simpler forms
∗ Corresponding author at: Ruhr University-Bochum, LWL University Hospital Division of Cognitive Neuropsychiatry and Psychiatric Preventive Medicine Alexandrinenstr. 1-3, D-44791 Bochum, Germany. Tel.: +49 234 5077 3202; fax: +49 234 5077 1119. E-mail address: [email protected] (C. Gonzalez-Liencres). 1 Both authors contributed equally. 0166-4328/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.bbr.2014.01.020
of emotion sharing such as emotional contagion [3] Empathy is crucial for social interaction, and it allows humans (and perhaps other animal species) to detect and relate to the feelings of others. The ability to share the emotions of someone who is exposed to painful stimuli, broadly referred to as “empathy for pain” has been widely explored in neuroimaging studies [4]. In essence, research has shown that when humans observe others exposed to painful stimuli, they recruit a neural network that greatly overlaps with the network that is active when oneself is experiencing physical pain [5–8], suggesting that the other’s affective state is “mapped” onto one’s own representational network in a simulation-like fashion [9]. Interestingly, multiple studies have shown that the levels of
C. Gonzalez-Liencres et al. / Behavioural Brain Research 263 (2014) 16–21
empathy and empathy-like processes can be modulated by distinct contextual factors (for reviews, see [3,10]). Data from animal experiments suggest that familiarity has an impact on empathy-like processes. For example, mice display analogous behavior to that of an observed familiar, but not so much a stranger, conspecific. More specifically, Langford and colleagues [11] demonstrated that mice injected with acetic acid writhe more when in the presence of a writhing familiar cagemate or sibling as compared to a stranger. Furthermore, another study reported that mice who observed siblings or mating partners receiving electric foot shocks froze more than when observing unrelated or stranger individuals [12]. Aside from the effects of familiarity on empathy, only a few authors have demonstrated that mice are capable of being socially responsive to the distress shown by conspecifics, whereby an emotional or physical behavioral response is induced in an individual upon the observation of another’s state or condition. For example, Chen et al. [13] showed that a mouse (observer) that witnesses its cagemate (demonstrator) being trained in a fear conditioning task consisting of a tone followed by an electric shock will subsequently freeze in response to the tone alone although it had never experienced the tone-shock training by itself. Similarly, Jeon and colleagues reported that observer mice froze upon the mere observation of another mouse receiving foot shocks, without the need of a conditioned stimulus such as a tone [12]. However, the authors did not assess the behavior of the mice right after the period with shocks, which leaves open to speculation whether or not the observed response during the exposure to foot shocks would abate right after cessation of foot shocks to the demonstrator. While studies showed that mice exhibit emotional responses to the observation of the affective states of others, there have hitherto been no studies that examined the behavioral response of the observer immediately after the cessation of a stressful period. To explore this, we modified a previously employed test [12] by adding an extra minute in which no shocks were delivered in order to assess the behavior of both observer and demonstrator before, during and after a stressful event. In addition, we aimed to measure stress-related autonomic responses (i.e. fecal droppings) in both the observer and the demonstrator. We expected that a decrease in the saliency of the behavioral response of the demonstrator once the shocks terminated would prevent the observer from perceiving the other’s distress and would consequently induce a reduction in the latter’s socially-induced stress response.
2. Materials and methods 2.1. Mouse husbandry All animal procedures were performed in agreement with the European Communities Council Directive of 24 November 1986 (86/609/EEC). Animal care and experiments were conducted according to institutional guidelines and after approval by the state authority for animal research conduct of North Rhine–Westphalia, Germany (approval 84-02.04.2012.A201). 48 C57BL/6J male mice (Charles River, Sulzfeld, Germany) were housed in groups of 6 for six weeks prior to the experiment under pathogen-free conditions. They were given chow and water ad libitum and were maintained in an inverse 12:12 light–dark cycle. The temperature (21 ± 1 ◦ C) and humidity (50–60%) were monitored and kept constant. The mice were tested during the dark cycle at the age of 22 to 24 weeks old. In the cagemate condition, the observer and demonstrator mice were chosen from the same cage. In the non-cagemate condition, the observer and demonstrator mice were selected from different cages and had never had contact before. Each mouse was only used once as either observer or demonstrator, given that the stress and fear associated to the context modulated the results in the second
17
trial. Therefore, 24 cagemates (12 observers) and 24 non-cagemates (12 observers) were used for this experiment. 2.2. Experimental set up The apparatus consisted of an Active/Passive Avoidance System box with two chambers (13 cm × 14.5 cm × 16 cm) divided in the middle by a Plexiglas partition (0.4 cm × 12 cm × 16 cm) enclosed in a ventilated and sound-proof container (TSE Systems GmbH, Bad Homburg, Germany). In the demonstrator side, electric shocks were delivered through a rod floor on which the mouse was placed. Automated control of the shocks was carried out by the shuttlebox software (TSE Systems GmbH, Bad Homburg, Germany) and all experiments were recorded with a DCR-SR57E digital video camera (Sony Corp., Japan). 2.3. Procedure Two mice (demonstrator and observer) were individually placed in each apparatus chamber of a startle box for 5 min to control for baseline levels. Then a 4 min test period followed in which 1 s electric foot shocks (0.5 mA) were delivered through the floor rods to one of the mice (demonstrator) every 10 s. Finally, there was a 1 min recovery period in which no shocks were applied in order to monitor the behavior of the mice after the stressful stage (Fig. 1). The experiment (10 min) was recorded with a camera for subsequent analysis. The parameters measured were the freezing time (defined as lack of movement, except for respiration, for longer than 2 s) from each mouse (demonstrator and observer), and the number of fecal droppings (higher number indicatory of fear). The apparatus was thoroughly cleaned with 70% ethanol between trials. Only males were tested given that the estrous cycle in females could alter the hormonal state of the animals and therefore the fear and emotional reaction. 2.4. Statistical analysis In the preliminary step the data were checked for the assumptions of parametric statistical analyses with Wilk-Shapiro test. The fecal dropping differences between groups and the freezing time differences in different time blocks between groups were analyzed with repeated ANOVAs (GLM function in SPSS). Directionality of main effects and interaction were tested by post hoc t-tests and orthogonal comparisons. Bonferroni correction was used to adjust the level of significance for multiple comparisons (critical p = .002). Lastly, Pearson correlation analyses were performed for the number of fecal droppings between observers and demonstrators. All statistical tests were performed with SPSS version 20 and statistical significance was set to p < 0.05 (two tailed). 3. Results 3.1. Group differences in freezing behavior when experiencing and witnessing another mouse receiving electric foot shocks We first monitored the behavior of mice when they observed cagemates or non-cagemates receive electric foot shocks. Accordingly, there was a main effect of time blocks (during the baseline, test and recovery periods; F(2,19) = 14.86, p =
Contents lists available at ScienceDirect
Behavioural Brain Research journal homepage: www.elsevier.com/locate/bbr
Research report
Emotional contagion in mice: The role of familiarity Cristina Gonzalez-Liencres a,b,∗ , Georg Juckel a,b , Cumhur Tas a,c , Astrid Friebe a,1 , Martin Brüne a,b,1 a b c
Department of Psychiatry, Ruhr University-Bochum, LWL University Hosiptal, Germany International Graduate School of Neuroscience, Ruhr-University Bochum, Germany Department of Psychology, Üsküdar University, Istanbul, Turkey
h i g h l i g h t s • • • • •
Familiar and unfamiliar mice observed conspecifics receiving electric foot shocks. Familiar observers freeze more than strangers when witnessing another in distress. Demonstrators receiving shocks freeze the most aftershocks have ceased. Observers stop freezing when shocks have ceased in the adjacent compartment. Demonstrators show less freezing behavior when observer is familiar.
a r t i c l e
i n f o
Article history: Received 2 December 2013 Received in revised form 17 January 2014 Accepted 19 January 2014 Available online 28 January 2014 Keywords: Empathy Observer Demonstrator Observational fear Shock
a b s t r a c t Empathy is a complex emotional process that involves sharing an emotional state with another organism. The extent to which nonhuman animals are capable of empathizing with others is still far from clear, partly due to a lack of empirical work in this domain, but also due to definitional confusion of empathy with emotional contagion and other related terms. In this study, an observer mouse witnessed a familiar cagemate or an unfamiliar non-cagemate receiving electric foot shocks in an experiment that consisted of three periods: baseline (no shocks), test (shocks) and recovery (no shocks). Freezing behavior in the observer was significantly increased in the cagemate, as opposed to the non-cagemate condition during the test period, but not during baseline or recovery, emphasizing the role of familiarity in empathy-like processes. In agreement with this, we also found a correlation that approached significance between the total number of fecal droppings of the observers, as an indication of distress, and those of the demonstrator in the cagemate, but not in the non-cagemate, condition. While the freezing behavior of the demonstrators increased with time, reaching a maximum at the recovery period, the observers froze the most during the test period while the demonstrators were receiving the electric foot shocks. The observation that the freezing response of the observers ceased when the shocks in the adjacent compartment stopped could be due to a decrease in saliency of the demonstrators’ behavioral response. Finally, the presence of a cagemate, as compared to a stranger, possibly reduced the demonstrator’s pain-induced behavior, suggesting an ameliorating effect of familiarity on stress responses. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Empathic concern for others is rooted in or related to parental care for altricial offspring [1,2], whereby empathy – defined as the ability to share the emotions of others – evolved from simpler forms
∗ Corresponding author at: Ruhr University-Bochum, LWL University Hospital Division of Cognitive Neuropsychiatry and Psychiatric Preventive Medicine Alexandrinenstr. 1-3, D-44791 Bochum, Germany. Tel.: +49 234 5077 3202; fax: +49 234 5077 1119. E-mail address: [email protected] (C. Gonzalez-Liencres). 1 Both authors contributed equally. 0166-4328/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.bbr.2014.01.020
of emotion sharing such as emotional contagion [3] Empathy is crucial for social interaction, and it allows humans (and perhaps other animal species) to detect and relate to the feelings of others. The ability to share the emotions of someone who is exposed to painful stimuli, broadly referred to as “empathy for pain” has been widely explored in neuroimaging studies [4]. In essence, research has shown that when humans observe others exposed to painful stimuli, they recruit a neural network that greatly overlaps with the network that is active when oneself is experiencing physical pain [5–8], suggesting that the other’s affective state is “mapped” onto one’s own representational network in a simulation-like fashion [9]. Interestingly, multiple studies have shown that the levels of
C. Gonzalez-Liencres et al. / Behavioural Brain Research 263 (2014) 16–21
empathy and empathy-like processes can be modulated by distinct contextual factors (for reviews, see [3,10]). Data from animal experiments suggest that familiarity has an impact on empathy-like processes. For example, mice display analogous behavior to that of an observed familiar, but not so much a stranger, conspecific. More specifically, Langford and colleagues [11] demonstrated that mice injected with acetic acid writhe more when in the presence of a writhing familiar cagemate or sibling as compared to a stranger. Furthermore, another study reported that mice who observed siblings or mating partners receiving electric foot shocks froze more than when observing unrelated or stranger individuals [12]. Aside from the effects of familiarity on empathy, only a few authors have demonstrated that mice are capable of being socially responsive to the distress shown by conspecifics, whereby an emotional or physical behavioral response is induced in an individual upon the observation of another’s state or condition. For example, Chen et al. [13] showed that a mouse (observer) that witnesses its cagemate (demonstrator) being trained in a fear conditioning task consisting of a tone followed by an electric shock will subsequently freeze in response to the tone alone although it had never experienced the tone-shock training by itself. Similarly, Jeon and colleagues reported that observer mice froze upon the mere observation of another mouse receiving foot shocks, without the need of a conditioned stimulus such as a tone [12]. However, the authors did not assess the behavior of the mice right after the period with shocks, which leaves open to speculation whether or not the observed response during the exposure to foot shocks would abate right after cessation of foot shocks to the demonstrator. While studies showed that mice exhibit emotional responses to the observation of the affective states of others, there have hitherto been no studies that examined the behavioral response of the observer immediately after the cessation of a stressful period. To explore this, we modified a previously employed test [12] by adding an extra minute in which no shocks were delivered in order to assess the behavior of both observer and demonstrator before, during and after a stressful event. In addition, we aimed to measure stress-related autonomic responses (i.e. fecal droppings) in both the observer and the demonstrator. We expected that a decrease in the saliency of the behavioral response of the demonstrator once the shocks terminated would prevent the observer from perceiving the other’s distress and would consequently induce a reduction in the latter’s socially-induced stress response.
2. Materials and methods 2.1. Mouse husbandry All animal procedures were performed in agreement with the European Communities Council Directive of 24 November 1986 (86/609/EEC). Animal care and experiments were conducted according to institutional guidelines and after approval by the state authority for animal research conduct of North Rhine–Westphalia, Germany (approval 84-02.04.2012.A201). 48 C57BL/6J male mice (Charles River, Sulzfeld, Germany) were housed in groups of 6 for six weeks prior to the experiment under pathogen-free conditions. They were given chow and water ad libitum and were maintained in an inverse 12:12 light–dark cycle. The temperature (21 ± 1 ◦ C) and humidity (50–60%) were monitored and kept constant. The mice were tested during the dark cycle at the age of 22 to 24 weeks old. In the cagemate condition, the observer and demonstrator mice were chosen from the same cage. In the non-cagemate condition, the observer and demonstrator mice were selected from different cages and had never had contact before. Each mouse was only used once as either observer or demonstrator, given that the stress and fear associated to the context modulated the results in the second
17
trial. Therefore, 24 cagemates (12 observers) and 24 non-cagemates (12 observers) were used for this experiment. 2.2. Experimental set up The apparatus consisted of an Active/Passive Avoidance System box with two chambers (13 cm × 14.5 cm × 16 cm) divided in the middle by a Plexiglas partition (0.4 cm × 12 cm × 16 cm) enclosed in a ventilated and sound-proof container (TSE Systems GmbH, Bad Homburg, Germany). In the demonstrator side, electric shocks were delivered through a rod floor on which the mouse was placed. Automated control of the shocks was carried out by the shuttlebox software (TSE Systems GmbH, Bad Homburg, Germany) and all experiments were recorded with a DCR-SR57E digital video camera (Sony Corp., Japan). 2.3. Procedure Two mice (demonstrator and observer) were individually placed in each apparatus chamber of a startle box for 5 min to control for baseline levels. Then a 4 min test period followed in which 1 s electric foot shocks (0.5 mA) were delivered through the floor rods to one of the mice (demonstrator) every 10 s. Finally, there was a 1 min recovery period in which no shocks were applied in order to monitor the behavior of the mice after the stressful stage (Fig. 1). The experiment (10 min) was recorded with a camera for subsequent analysis. The parameters measured were the freezing time (defined as lack of movement, except for respiration, for longer than 2 s) from each mouse (demonstrator and observer), and the number of fecal droppings (higher number indicatory of fear). The apparatus was thoroughly cleaned with 70% ethanol between trials. Only males were tested given that the estrous cycle in females could alter the hormonal state of the animals and therefore the fear and emotional reaction. 2.4. Statistical analysis In the preliminary step the data were checked for the assumptions of parametric statistical analyses with Wilk-Shapiro test. The fecal dropping differences between groups and the freezing time differences in different time blocks between groups were analyzed with repeated ANOVAs (GLM function in SPSS). Directionality of main effects and interaction were tested by post hoc t-tests and orthogonal comparisons. Bonferroni correction was used to adjust the level of significance for multiple comparisons (critical p = .002). Lastly, Pearson correlation analyses were performed for the number of fecal droppings between observers and demonstrators. All statistical tests were performed with SPSS version 20 and statistical significance was set to p < 0.05 (two tailed). 3. Results 3.1. Group differences in freezing behavior when experiencing and witnessing another mouse receiving electric foot shocks We first monitored the behavior of mice when they observed cagemates or non-cagemates receive electric foot shocks. Accordingly, there was a main effect of time blocks (during the baseline, test and recovery periods; F(2,19) = 14.86, p =
