Social Defeat Stress Model for Adolescent C57BL/6 Male and Female Mice

Summary

We developed an accelerated social defeat stress model for adolescent C57BL/6 mice, which works in both males and females and allows exposure during discrete adolescent periods. Exposure to this model induces social avoidance, but only in a subset of defeated male and female mice.

Abstract

Social adversity in adolescence is prevalent and can negatively impact mental health trajectories. Modeling social stress in adolescent male and female rodents is needed to understand its effects on ongoing brain development and behavioral outcomes. The chronic social defeat stress paradigm (CSDS) has been widely used to model social stress in adult C57BL/6 male mice by leveraging on the aggressive behavior displayed by an adult male rodent to an intruder invading its territory. An advantage of this paradigm is that it allows to categorize defeated mice into resilient and susceptible groups based on their individual differences in social behavior 24 h after the last defeat session. Implementing this model in adolescent C57BL/6 mice has been challenging because adult or adolescent mice do not typically attack early adolescent male or female mice and because adolescence is a short period of life, encompassing discreet temporal windows of vulnerability. This limitation was overcome by adapting an accelerated version of the CSDS to be used for adolescent male and female mice. This 4-day stress paradigm with 2 physical attack sessions per day uses a C57BL/6 male adult to prime the CD-1 mouse for aggressiveness such that it readily attacks the male or female adolescent mouse. This model was termed accelerated social defeat stress (AcSD) for adolescent mice. Adolescent exposure to AcSD induces social avoidance 24 h later in both males and females, but only in a subset of defeated mice. This vulnerability occurs despite the number of attacks being consistent across sessions between resilient and susceptible groups. The AcSD model is short enough to allow exposure during discrete periods within adolescence, allows the segregation of mice according to the presence or absence of social avoidance behavior, and is the first model available to study social defeat stress in adolescent C57BL/6 female mice.

Introduction

The chronic social defeat stress paradigm is widely used to model social stress in adult postnatal day (PND) >65 male rodents. This paradigm is based on the natural aggressive behavior of an adult male rodent when an intruder invades its territory. This model is used in a variety of rodent species, including rats, hamsters, and mice^{1,2,3,4,5,6,7,8,9}, and consists of a combination of physical aggression and psychological stress lasting about 10 days, during which an intruder rodent experiments a few min of physical aggression from the resident rodent. The two rodents remain in the resident's home cage, separated by a divider that allows sensory but not physical contact⁷. In mice experiments, the most commonly used resident/aggressor mice are retired breeders Swiss CD-1 mice, which show robust territorial behavior against intruder mice^6,7. For the intruder mice, the best characterized strain is the inbreed C57BL/6 strain^2,4,5. The defeated mouse is exposed to a new aggressor every day to prevent habituation to the aggressor. Control mice are housed with a different conspecific each day. At 24 h after the last defeat session, experimental mice are tested in a social interaction test (SIT) in which they can explore an open field in the absence (no target), or presence of a novel CD-1 mouse (target). Control mice spend more time in the interaction zone with the target versus the non-target part of the task. Defeated mice are classified as susceptible (ratio<1) or resilient (ratio>1) according to a social interaction ratio (time spent in the interaction zone with aggressor present/time spent in the interaction zone with aggressor absent). This procedure provides a useful tool to study individual differences in response to stress.

Until recent years, the chronic social defeat stress model has been used mostly in adult male mice because the accentuated male dominance hierarchies involve fighting against males but not against females^6,7. Furthermore, male rodents typically do not attack females; instead, they engage in mating behaviors¹⁰. Notwithstanding these obstacles, different strategies have been developed to adapt the chronic social defeat stress paradigm for adult female mice. For example, the California mouse model of social defeat is based on the natural aggression of this monogamous species from both sexes when defending their territory^9,11. Other approaches focus on inducing aggressive behavior of the CD-1 mice by stimulating their ventromedial hypothalamus to have consistent aggressive behavior^10,12, or by applying male urine in the experimental adult female mice to receive attacks from CD-1 aggressors¹³. This heightened and consistent aggression of the CD-1 mice is critical for the experimental intruder mouse to show clear behavioral signs of subordination towards the repeated attacks by the aggressor during length of the interaction⁶.

Adapting the chronic social defeat model to use in adolescent C57BL/6mice
Adolescence is a period marked by substantial psychosocial maturation, which unfolds in parallel with changes in the micro and macro architecture of the brain, particularly in the prefrontal cortex. In both humans and rodents, there is little consensus regarding the specific onset and end of the adolescent period^14,15. Furthermore, there are critical windows of vulnerability within adolescence for experience-induced disruption of ongoing brain and cognitive development^16,17,18,19. Puberty and adolescence occur at the same time, but these terms are not synonymous. Puberty signifies the onset of sexual maturation, whereas adolescence represents a broader phase characterized by the gradual shift from a juvenile state to achieving independence²⁰. Different groups have suggested that adolescence in mice spans from weaning (PND 21) until adulthood (PND 60)²¹. Particularly, early adolescence can be referred to as the first and second weeks postweaning (PND 21-34) and mid-adolescence as the PND 35-48 period. These ranges encompass discrete developmental periods regarding for example the development of the dopamine system^22,23,24, vulnerability to drug effects on developing neuronal networks^17,25,26,27, and distinct behavioral characteristics^{16,20,28,29,30}.

Fighting behavior from the resident mouse is required for the social defeat protocol. However, as with female mice, males do not typically engage in aggressive interactions with early adolescent mice, possibly because they do not perceive them as a threat. Most studies exploring the effects of chronic social defeat in adolescent C57BL/6 mice have been performed during the mid-adolescent period^{31,32,33,34,35}; others do not specify the postnatal day of adolescent exposure^36,37, or extend the days for the defeat to early adulthood³⁸ or do not allow sensory contact³⁹; other studies on adolescent mice use different strains^40,41. The characteristics of these studies using chronic social defeat stress in adolescent male mice are summarized in Table 1.

Our research group is interested in targeting specific adolescent windows of exposure, including early adolescence, in C57BL/6 mice. Because of the short duration of the different adolescent periods, a modified version of the accelerated version of the chronic social defeat stress paradigm was designed⁴². This model was termed accelerated social defeat stress (AcSD) for adolescent mice. Previous work shows that there are significant sex differences in sensitivity to social stress in adolescence in rats^{8,26,43,44,45}, as well as in the harmful effects of social stress on mental health trajectories in humans^{46,47,48,49,50,51,52,53,54,55,56}. The AcSD model works effectively for adolescent female mice, too, allowing them to investigate potential sex-specific consequences as well as explore the neurobiological foundation.

Table 1: Studies using social defeat stress paradigms in adolescent male mice. Strain and species: California Mouse: Peromyscus californicus. C57BL/6: Mus musculus black 6 inbred mouse strain. C57BL/6J: M. musculus black 6 inbred mouse model provided by Jackson's laboratories. CD-1: M. musculus from Swiss outbred albino mouse strain. ICR: M. musculus Institute of Cancer Research outbred albino mouse strain. OF1: M. musculus Oncins France 1 outbred albino mouse strain.
Abbreviations: wk = week; PND = Postnatal day; res = resilient; sus = susceptible; unsus = unsusceptible. Please click here to download this Table.

Protocol

Experimental procedures were performed in accordance with the guidelines of the Canadian Council of Animal Care and approved by the McGill University and Douglas Hospital Animal Care Committee (animal experiment approval number: 2005-5084). All mice were housed in a temperature- and humidity-controlled (21-22 °C; 60%) colony room and on a 12 h light-dark cycle (light on at 8:00 h) at the Neurophenotyping Centre of the Douglas Mental Health University Institute. The mice had ad libitum access to food and water throughout the experiments. Mice were assigned randomly to each experimental condition. Experimental animals were obtained from a commercial source (see Table of Materials). Table 2 summarizes the animal's characteristics for each experimental phase. In total, 489 adolescent C57BL/6J mice (259 males; 230 females) were used across 20 experimental cohorts.

Table 2: Animal subjects for each experimental phase. ^†If these animals are not severely injured, they can be reused for priming. ^‡If these animals are not severely injured, they can be reused for priming during AcSD, ^§Neurophenotyping Centre of the Douglas Mental Health University Institute. ^¶For experiments with males, ^#For experiments with females, if defeat is mixed, use females for priming. ^††Not to be used for AcSD. Abbreviations: AcSD = accelerated social defeat stress; PND = Postnatal day Please click here to download this Table.

1. Acclimation

NOTE: This phase should last 1 week minimum.

1. Purchase male CD-1 retired breeder mice (see Table of Materials) older than 4 months of age and twice the quantity of the adolescent C57BL/6J mice (i.e., if 10 mice will be exposed to stress, purchase 20 CD-1 mice).
   NOTE: Acquiring this quantity of CD-1 mice is recommended because just a small proportion of CD-1 mice show aggressive behavior towards adolescent mice.
2. Let the CD-1 male mice habituate to the new facility in their home cage with food and water ad libitum for 1 week prior to the screening and priming.
3. Use C57BL/6 at PND >65 (adults) and PND 25 ± 3 (adolescents) (Table 2) for the screening and priming sessions. Acquire 4 C57BL/6 mice of each age per day of priming (i.e., for 3 days of priming, acquire 12 PND >65 and 12 PND 25 ± 3) to decrease the risk of injuries because these animals will be employed repeatedly for screening and priming process.

2. CD-1 screening and priming

NOTE: The duration of this phase is 3 days of screening (1 day for female mice) and 3-4 days of priming (this can increase for female mice and depending on the CD-1 mice aggressive behavior.)

1. Use the operational definition of attack to register the attacks of the CD-1 as described below.
   1. One attack is defined as when the CD-1 mouse bites the C57BL/6J mouse, and the C57BL/6J mouse moves away in response to the bite. A bite is defined as when the CD-1 mouse places its teeth on any part of the C57BL/6J mouse's body. Moving away is defined as when the C57BL/6J mouse moves both of its hind paws from the position they were in before the CD-1 engaged it.
   2. Ensure there is a difference of at least ~2 s between separate attacks. If more than one bite occurs <2 s apart, count as one attack.
   3. If the CD-1 mouse bites more than once in succession without a break (<~2 s apart), gently separate animals with a plastic ruler. If the CD-1 mouse bites and does not let go, gently separate the animals. If the C57BL/6J mouse becomes trapped in a corner or is pinned down by the CD-1 mouse and cannot move away in response to a bite, gently separate animals with a ruler. Count all of these as one attack (up to separation).
2. Screening
   1. Use nitrile examination gloves for all animal handling and avoid the use of perfume or fragrances.
   2. Perform the 1^st day of the screening only with adult C57BL/6 male mice as intruders (PND >65) in the CD-1 mice home-cage. Always screen the CD-1 mice before a new experiment.
   3. Introduce an adult male C57BL/6 mouse PND >65 in the home-cage of a CD-1 male mouse for 3 min. Register the latency to the first aggression and how many aggressions are performed (see Supplementary Table 1).
   4. From the 2^nd to the last day of screening, introduce an adult male C57BL/6 PND>65 for 1 min (limit to 30 s if the CD-1 is very aggressive) and then swap it with an adolescent C57BL/6 PND 25 ± 3 for 5 min.
      NOTE: It is important to perform the screening with adolescent mice because not all the mice that attack the adult C57BL/6 male mouse will attack the adolescent C57BL/6. In fact, CD-1 mice that attack adolescent mice will not always attack adult C57BL/6 males.
   5. If the experiment is for aggressiveness towards female mice, perform the screening and priming with adolescent female mice, even if the experimental mice include both sexes.
   6. Register the latency to the first aggression, the number of attacks, and the wounds received by all the C57BL/6 mice (see Supplementary Table 2).
   7. Register any kind of interaction: chasing, mounting behaviors, sniffing, grooming, or biting. Register if the CD-1 does not approach the adolescent C57BL/6 mice; if this behavior is repeated for 2 days, do not include the CD-1 in priming.
   8. For the priming phase, select all the CD-1 mice that display behaviors such as chasing, mounting behaviors, sniffing, following, grooming, or biting the adolescent C57BL/6.
   9. Keep the non-selected CD-1 mice and use them for the social interaction test (SIT).
      NOTE: Territorial behavior increases with time; therefore, the non-selected CD-1 mice may become aggressive toward adolescent mice in the next experiment.
3. Priming CD-1 aggressive behavior
   1. House the CD-1 mice selected for priming in their home cage until the last day of priming.
   2. Introduce an adult male C57BL/6J mouse PND >65 for 30 s to the CD-1 home-cage and then swap it with an adolescent C57BL/6 PND 25 ± 3 mouse for 5 min.
   3. Register the latency to the first aggression, the number of attacks, and any interaction (see Supplementary Table 3).
   4. Perform the priming 2x a day for 3-4 days at the same time the defeats are scheduled (9:00 am and 2:00 pm).
      NOTE: Not all the CD-1 mice will attack the adolescent mice during the first days, especially if the priming is towards female mice. The number of days for priming can increase or decrease depending on the aggressive behavior shown by the CD-1 mice.
   5. To decrease the number of priming days, reuse CD-1 mice that have been through a prior social defeat experiment with male mice. If CD-1 mice do not show increased aggression, extend the days for priming until reaching criteria.
   6. Decrease the days for priming if the CD-1 mice show extreme aggressive behavior (C57BL/6 mice show excessive wounding) to prevent severe injuries during social defeat.
   7. For the next phase of the experiment, select the CD-1 mice that attack the adolescent mice at least 10x (males) or 5x (females) for 2 consecutive days.

3. AcSD protocol

NOTE: This phase consists of 3 days of habituation and 4 days of exposure to AcSD.

1. Habituation
   1. Set-up the rat cages for social defeat (Figure 1A-F). Divide the rat cages into 2 compartments using appropriate acrylic glass dividers (Figure 1B,D; see Table of Materials).
   2. Add hard woodchip bedding to both sides of the cage, add a cotton square per side as bedding material (Figure 1D). Ensure that the wire top is secured with binder clips (Figure 1E). Ensure that the rat cage is covered with the cover lid without the protective cotton (Figure 1F).
   3. Place a CD-1 mouse per cage on one side of the cage. Place a tag on the outside of the cage for easy identification of each mouse.
   4. House the selected CD-1 mice in the social defeat apparatus (Figure 1F) 2 or 3 days before the AcSD protocol starts (to enhance CD-1 territorial behavior).
   5. Perform a priming session 1 day before the experiment; this is important to keep aggression consistent.
2. Defeat sessions
   1. Perform 2 sessions per day (first session: 9 AM, second session: 2 PM) for 4 days, 8 sessions in total. Make a schedule to ensure that the aggressor will be different for each session (see Supplementary Table 4 and Supplementary Table 5).
   2. Weigh the C57BL/6 adolescent mouse before introducing it to the CD-1 side of the cage. Introduce an adult male C57BL/6 mouse >65 for 30 s into the CD-1 side of the defeat cage and then swap it with an adolescent C57BL/6 PND 24-28 mouse for a 10-min long defeat episode or 10 attacks maximum per session.
   3. Separate the animals after each attack with a plastic ruler. After every session of defeat, check for injuries, and treat them with pain relief cream. Register the number of injuries.
   4. Place the adolescent C57BL/6J mouse in the free compartment next to an unfamiliar aggressor until the next session. Repeat the procedure with a different aggressor each session.
   5. Start the session with the experimental and aggressor mice that were not paired the previous session. Monitor the injuries of the adult C57BL/6J male mice used for priming (see Supplementary Table 6).
3. Follow the recommendations below for animal welfare and humane endpoints.
   1. Follow the standard operating procedure of the corresponding institutional review board for social defeats and ensure constant veterinary evaluation.
   2. Monitor the health of mice throughout the study to avoid animal suffering and death.
   3. Verify that the perforated divider is correctly set to prevent mice from crossing to the other compartment.
   4. Notify the veterinary care staff when an animal has a wound for appropriate monitoring and apply pain relief cream daily until the wound is completely healed.
   5. If an injury worsens (wounds increase in size), limit the attacks to 5 in the next session allowing recovery. Check for severe injuries. Severe injuries are defined as wounds totaling > 1 cm.
   6. Immediately remove and euthanize animals presenting a 15%-20% reduction of their weight from initial body weight; reduced/decreased locomotor activity and impaired grooming; mice with wounds > 1 cm; severed penis, or bite marks with swelling of the forelimbs.
      NOTE: Primary euthanasia is conducted via isoflurane/CO2 gas chambers followed by secondary euthanasia done by cervical dislocation.
4. Control mice
   1. Perform 2 sessions per day (first session: 9 AM, second session: 2 PM) for 4 days, 8 sessions in total. Make a schedule to ensure that control mice share the cage with a different C57BL/6J conspecific every session (see Supplementary Table 7 and Supplementary Table 8).
      NOTE: Control mice are not exposed to an aggressor. To mimic the introduction to a novel CD-1 mouse every defeat session, control mice are paired with a new same-sex and same-age conspecific (non-littermate) every session.
   2. House the control mice in the social defeat apparatus (Figure 1) in a separate room from that of the AcSD mice.
   3. Place two control mice per cage, one on each side of the cage. House the mice in a different side of the cage or different cage altogether, so they do not recognize the odor. Rotate the control mice each session.

Figure 1: Apparatus for the accelerated social defeat stress (AcSD) for adolescent mice. (A) From left to right, paired steel-wire tops with 0.6 cm of separation between bars, clear perforated polycarbonate divider with perforations diameter of 0.6 cm, binder clips to secure the wire tops, clear rat cage with cover lid (without cotton protective layer). (B) Up: comparison between the polycarbonate dividers for chronic social defeat in adults (left) and AcSD in adolescents (right); down: close up to the polycarbonate divider for adolescent mice. (C) Up: comparison between the steel-wire tops for chronic social defeat in adults (left) and AcSD in adolescents (right); down: close-up of the steel-wire tops. (D-F) Assembling of the social defeat apparatus. Please click here to view a larger version of this figure.

4. Social interaction test

1. Perform the SIT following the AcSD as described previously^57,58,59.
   1. Perform the SIT 24 h after the last session of AcSD. Perform the SIT under red light conditions (not under yellow light conditions) between 9 AM and 4 PM.
   2. Set up the main squared-arena and the overhead video camera (see Table of Materials). The main arena is divided into 3 areas: the social interaction zone (SIZ), the corners, and the center; these areas should be marked in the arena with a black line (use a black permanent marker). Use the following dimensions and criteria for each area.
      1. Ensure the SIZ is the squared area 14 cm x 9 cm, that surrounds the enclosure. The corners are small, squared areas (9 cm x 9 cm) in the corners of the wall opposite to the enclosure and represent the farthest point from the SIZ. The center is the area between the SIZ and the corners.
   3. Clean the arena and the enclosures with a cleaning hydrogen peroxidase solution.
   4. Place identification tags for each animal on the floor for analysis. Record all animal behavior for offline analysis.
2. First session to determine baseline exploration
   1. Place an empty wire mesh enclosure against one of the walls of the arena. Place the experimental (defeated or control) adolescent C57BL/6 mouse in the middle of the arena. Allow it to explore the arena freely for 2.5 min. Retrieve the mouse.
   2. Clean the arena and the enclosures with a cleaning hydrogen peroxidase solution.
3. Second session for social interaction
   1. Place an unfamiliar CD-1 mouse inside the wire mesh enclosure. Place the enclosure in the same location as the previous session.
   2. Reintroduce the same adolescent C57BL/6 mouse to the arena. Allow it to explore the arena freely for 2.5 min. Retrieve the mice.
   3. Clean the arena and the enclosures with a cleaning hydrogen peroxidase solution.
   4. Repeat the process for all adolescent C57BL/6 mice (stressed and control mice) in a counterbalanced order.
   5. Use more than one CD-1 mouse for the SIT to decrease handling stress and bias (e.g., for a cohort of 30 C57BL/6 mice, use 4 CD-1 mice). Counterbalance them among the adolescent mice.
      NOTE: This test measures the approach and/or avoidance behavior of the experimental mice toward a social target, as first described by Krishnan⁵ and Golden⁷.
4. Analysis
   1. Record the time (s) spent in the interaction zone and the corners during both sessions of the test.
   2. Calculate the social interaction ratio as the time spent in the interaction zone with the CD-1 aggressor present divided by the time spent in the interaction zone with the CD-1 aggressor absent.
   3. Classify stressed mice with a ratio < 1.00 as susceptible and a ratio ≥ 1.00 as resilient.
   4. Exclude animals that do not explore the arena during the first session (i.e., spent 0 s in any of the designated areas) from all analyses to control differences in locomotor activity.
   5. Exclude all animals that receive 0 attacks in ≥4 social defeat sessions from all analyses to prevent non-representative resilient animals.
   6. Exclude mice (not susceptible) that do not spend at least 61 s inside the interaction zone to ensure that high social interaction ratios reflect actual interest in the social target.
   7. Exclude mice with outlier interaction ratio scores.

Results

A total of four different experiments were performed using the chronic social defeat stress model in adolescent C57BL/6 male mice (PND 21). However, this model presented important limitations for its use in early adolescent C57BL/6 mice.

Equipment required modifications for adolescent C57BL/6 mice
The first limitation was that the equipment used for the social defeat apparatus was designed for adult mice. Because of their size, adolescent C57BL/6 mice at PND 21 were able to pass through the perforations of the dividers to the aggressor side. This interrupted the period of sensory contact. Additionally, adolescent mice were able to escape through the bars of the steel-wire tops. Therefore, the equipment was modified to prevent the mice's escape (Figure 1). The diameter of the perforations was changed from 1.1 cm to 0.6 cm (Figure 1B), and the steel-wire top was replaced from one that had 1.1 cm separation between bars to one with 0.6 cm separation (Figure 1C). Additionally, the top was secured with clips, and the rat cages were also covered with plastic covers without protective cotton. The woodchip bedding was also set at an appropriate level for the adolescent mice (Figure 1D-F).

AcSD model requires priming of adult male CD-1 mice to ensure consistent, aggressive bouts
The second and more important limitation was the lack of consistency in CD-1 mice's physical aggression toward the adolescent mice. During the chronic social defeat sessions, the aggression varied from zero aggressive attacks during the 10 min session to one or two bites that were fatal due to the small size of the adolescent mice. However, if the CD-1 mice were used for more than one experiment, the consistency in the number of attacks increased.

Figure 2B shows the number of attacks by day for each chronic social defeat experimental cohort (P0, A0, A1, and A2). An interaction was found between day and cohort (Figure 2B, two-way repeated measures ANOVA: F_{(27, 234)} = 3.83, p < 0.0001), main effect of day (F_{(6.288, 163.5)} = 6.2, p < 0.0001), and main effect of cohort (F_{(3, 26)} = 3.8, p = 0.02). Post hoc test revealed differences between cohorts in days 1, 2, 3, and 7 (Holm-Sidak post hoc tests: day 1, P0 vs. A2, p = 0.02; A1 vs. A2, p = 0.02; day 2, A1 vs. A2, p = 0.04; day, 3 A0 vs. A1, p = 0.007; A1 vs. A2, p = 0.04; day 7, A0 vs. A1, p = 0.04). All CD-1 mice in experiments P0 and A0 stopped attacking in the last 3 days. CD-1 mice that were used for experiments P0 and A0 were used in A1, in which there was a consistency in attacks (Figure 2B), indicating that the CD-1 were more prone to aggression if subjected to physical interactions before.

For this reason, as described in steps 2.1 and 2.2, the CD-1 mice were screened specifically for their aggression towards the adolescent mice and primed with an adult C57BL/6 male for aggression towards the adolescent C57BL/6 mice. Furthermore, AcSD was used to increase the age from PND 21 to PND 25 and still target early adolescence.

During the AcSD sessions, CD-1 mice increased the number of attacks towards the adolescent mice in a more consistent pattern as shown in Figure 2D. In contrast to the chronic social defeat paradigm, no significant interaction was found between the session and experimental cohort (Figure 2D, two-way repeated measures ANOVA: F_{(35, 483)} = 1.33, p = 0.1), there was a main effect of session (F_{(5.644, 389.5)} = 14.15, p < 0.0001) and a main effect of cohort (F_{(5, 69)} = 7.91, p < 0.0001); Holm-Sidak post hoc test did not reveal significant differences between cohorts. A comparison of the averages of attacks between cohorts revealed that this effect was mostly driven by cohort A7, which had a significantly higher number of attacks (data not shown, one-way ANOVA F_{(5, 69)} = 7.911, p < 0.0001; Holm-Sidak post hoc tests: A3 vs. A6, p = 0.04; A7 vs. A5, p = 0.002; A7 vs. A6, p < 0.0001; A7 vs. A8, p = 0.0001).

Adolescent mice received more attacks with primed CD-1 mice during AcSD than adolescent mice exposed to a chronic social defeat protocol (Figure 2E, unpaired t-test with Welch's correction, t = 15.40, df = 17.87, p < 0.0001) using naïve (non-primed) CD-1 mice. The attacks were not consistent during the chronic social defeat protocol, and in fact, several experimental mice were not attacked at all.

Figure 2: Comparison of number of attacks per day between chronic social defeat stress and AcSD. (A) Timeline of the chronic social defeat stress model. (B) Number of attacks per day for each experimental cohort (P0, P1, A1, and A2) during chronic social defeat stress in adolescent C57BL/6 male mice (two-way repeated measures ANOVA). (C) Timeline of the accelerated social defeat stress (AcSD) model. (D) Number of attacks per day for each experimental cohort (A3-A8) during AcSD in adolescent C57BL/6 male mice (two-way repeated measures ANOVA). (E) The average number of attacks was significantly higher for AcSD (unpaired t test with Welch's correction, ****p < 0.0001). All data are shown as mean ± SEM. Abbreviations: CSDS = chronic social defeat stress; SIT = social interaction test; PND = postnatal day; P0, A0-A8 = labels for each experimental cohort. Please click here to view a larger version of this figure.

Pattern of attacks during AcSD sessions does not impact social phenotype 24 h later
After social defeat stress exposure, most adolescent male mice did not show social avoidance when tested in the SIT (Figure 3B, chronic social defeat: one-way ANOVA F_{(2, 43)} = 14.57, p < 0.0001. F, AcSD: Kruskal-Wallis test, H₍₂₎ = 72.35, p < 0.0001). For both models, there were no significant differences in the number of attacks between social-avoidant and non-social-avoidant defeated mice (Figure 3D, chronic social defeat: two-way repeated measures ANOVA, F_{(9, 252)} = 4.07, p = 0.21. H. AcSD: two-way repeated measures ANOVA, F_{(7, 511)} = 1.32, p = 0.24). However, the proportion of resilient mice was higher following chronic social defeat than that following AcSD exposure (Figure 3C,G, 70 % vs 55.26 %, binomial test (one-tailed), p = 0.0047).

Figure 3: Comparison of social interaction test (SIT) scores following the chronic social defeat stress and the AcSD paradigms. (A) Timeline of the chronic social defeat stress model. (B) Distribution of SIT score after chronic social defeat stress in adolescence (one-way ANOVA). (C) Proportion of resilient/susceptible after chronic social defeat stress in adolescence. (D) Number of attacks by phenotype received during chronic social defeat stress, no significant differences between resilient and susceptible were found (two-way repeated measures ANOVA). (E) Timeline of the accelerated social defeat stress (AcSD) model. (F) Distribution of SIT score after AcSD in adolescence (Kruskal-Wallis). This figure has been modified from⁵⁷. (G) The proportion of resilient/susceptible after AcSD in adolescence. This figure has been modified from⁵⁷. (H) Number of attacks by phenotype received during AcSD; there is no difference between resilient and susceptible (two-way repeated measures ANOVA). This figure has been modified from⁵⁹. **p < 0.001, ****p < 0.0001. All data are shown as mean ± SEM. Abbreviations: PND = postnatal day; Con = controls; Res = resilient; Sus = susceptible. Please click here to view a larger version of this figure.

AcSD can be used in adolescent C57BL/6 female mice
Because PND 25 C57BL/6 male and female mice have not reached reproductive maturity, primed CD-1 mice could present aggressive attacks toward adolescent female mice as they did towards adolescent male mice. However, the priming phase had to be extended to reach a consistent number of attacks (Figure 4A). As shown in Figure 4B, no interaction was found between day and cohort across experiments (Figure 4B, two-way repeated measures ANOVA: F_(35,707) = 1.38, p = 0.07), there was a main effect of the session (F_{(6.286, 634.9)} = 16.1, p < 0.0001) and a main effect of the cohort (F_{(5, 101)} = 11.34, p < 0.0001; Holm-Sidak post hoc tests: session 2 C4 vs C8, p = 0.03; session 6 C7 vs C4, p = 0.02; C7 vs C5, p = 0.02; session 7 C8 vs C3, p = 0.02; C8 vs C4, p = 0.045). The majority of adolescent C57BL/6 female mice exposed to AcSD were resilient to social avoidance (Figure 4C, Brown-Forsythe's ANOVA: F_(2,113.4) = 27.2, p < 0.0001), similar to adolescent male mice. No significant difference was found in the number of attacks between resilient and susceptible mice (Figure 4E, two-way repeated measures ANOVA, session x phenotype interaction: F_{(7, 714)} = 0.73, p = 0.64, main effect of session F_{(6.375, 650.3)} = 7.02, p < 0.0001, no main effect of phenotype F_{(1, 102)} = 0.4, p = 0.53).

Figure 4: AcSD in adolescent female mice. (A) Timeline of the experiment. (B) Number of attacks per day by experiment during AcSD (two-way repeated measures ANOVA). (C) Distribution of SIT score after AcSD during adolescence in female mice (Brown-Forsythe's ANOVA). (D) Proportion of resilient/susceptible after AcSD in adolescence (right). (E) Number of attacks by phenotype received during AcSD; no significant differences between resilient and susceptible mice were found (two-way repeated measures ANOVA). All data are shown as mean ± SEM. (C-E) These figures have been modified from⁵⁹. Abbreviations: SIT = social interaction test; PND = postnatal day; C3-C8 = labels for each experimental cohort; Con = controls; Res = resilient; Sus = susceptible. Please click here to view a larger version of this figure.

Supplementary Table 1: Scoring sheet template for day 1 of screening with adults. Abbreviations: CD-1 = ID of CD-1 mice screened, Exp = Experiment ID, 3 min = Duration of the screening session for the C57BL/6 adult mice, BL6 ID = ID of C57BL/6 adult mice used for screening, Lat 1 ADULTS = Latency of the first attack from the CD-1 towards the C57BL/6 adult mouse, # aggressions = Number of attacks performed by the CD-1. Please click here to download this Table.

Supplementary Table 2: Scoring sheet template for the screening with adults and adolescents. Abbreviations: Exp = Experiment ID, 1 min = Duration of the screening session for the C57BL/6 adult mice, BL6 ID = ID of C57BL/6 adult mice used for screening, Lat 1 ADULTS = Latency of the first attack from the CD-1 towards the C57BL/6 adult mouse, # aggressions = Number of attacks performed by the CD-1, 5 min = Duration of the screening session for the C57BL/6 adolescent mice, BL6 Ad = ID of C57BL/6 adolescent mice used for screening, Lat 1 Adolescents = Latency of the first attack from the CD-1 towards the C57BL/6 adolescent mouse, CD-1 = ID of CD-1 mice screened. Please click here to download this Table.

Supplementary Table 3: Scoring sheet template for the priming phase. Abbreviations: CD-1 = ID of CD-1 mice used for priming, Exp = Experiment ID, 30 sec = Duration of the priming session for the C57BL/6 adult mice, BL6 ID = ID of C57BL/6 adult mice used for priming, Lat 1 ADULTS = Latency of the first attack from the CD-1 towards the C57BL/6 adult mouse, # aggressions = Number of attacks performed by the CD-1, Min: 3 = Duration of the priming session for the C57BL/6 adolescent mice, BL6 Ad = ID of C57BL/6 adolescent mice used for priming, Lat 1 Adolescents = Latency of the first attack from the CD-1 towards the C57BL/6 adolescent mouse, # mountings = Number of mountings received by the C57BL/6 adolescent mice. Please click here to download this Table.

Supplementary Table 4: Scoring sheet template for social defeat scheduling (20 adolescent mice). Abbreviations: CD1 = ID of CD-1 mice, E = Experiment, H = Humidity, 1 st = Session 1, 2^nd = Session 2, BL6 adult = ID of C57BL/6 adult mice used for priming, Weight = Weight of C57BL/6 adolescent mice, # attacks = Number of attacks performed by the CD-1, # wounds = Number of wounds received by the C57BL/6 adolescent mice. Please click here to download this Table.

Supplementary Table 5: Example of social defeat scheduling. Please click here to download this Table.

Supplementary Table 6: Template for injury monitoring of adult C57BL/6 male mice. Abbreviations: 1st - # wounds = Number of wounds received by the C57BL/6 adult mouse in session 1, 2nd - # wounds = Number of wounds received by the C57BL/6 adult mouse in session 2, BL6 = ID of C57BL/6 adult mice used for priming. Please click here to download this Table.

Supplementary Table 7: Control group scheduling template (10 adolescent mice). Rules for scrambling: Do not repeat the same conspecific; a mouse cannot repeat the same side of the box in the same day; mice have to be scrambled with mice from different cages from their home cage; home cages are numbered; mice numbers are given depending on the home cage number; defeated and controls groups are taken randomly from the home cage. Please click here to download this Table.

Supplementary Table 8: Example of the control group scheduling. Please click here to download this Table.

Discussion

Consistent, aggressive behavior in CD-1 mice
During the screening phase, it is very important to take note of all the behaviors displayed by the CD-1 (chasing, mounting behaviors, sniffing, grooming, or biting) and to closely follow these records when selecting the CD-1 mice for the AcSD. It is likely that a CD-1 mouse which interacts with the adolescent mouse without attacking it will develop aggressiveness towards adolescent mice during priming. In contrast, a CD-1 mouse which attacks an adult mouse but does not approach an adolescent mouse will probably not develop aggressiveness towards adolescent mice.

The CD-1 priming phase allows to develop aggressive behavior toward adolescent mice. Therefore, having an accurate register of the behaviors is critical for the selection of the mice. Furthermore, it is essential to always screen the CD-1 mice before a new experiment, even when a CD-1 mouse has been through defeat before. Some CD-1 mice can stop attacking adolescent mice when left undisturbed in their home cages for a period of time, others can develop increased territorial behavior because they are housed in isolation.

For the priming of the CD-1 mice, the protocol recommends 3 days. If naïve CD-1 mice are employed, the priming can be extended if the CD-1 mice do not attack the adolescent mice actively. For CD-1 mice that had been through social defeat with adolescent mice, the duration of this priming phase can be reduced.

The screening and priming phases are critical for successful AcSD sessions. For instance, CD-1 mice that had been appropriately primed will attack and reach the 10-attack limit faster than a less aggressive CD-1 mouse, allowing for short defeat episodes.

During AcSD, it is crucial to separate the animals when the attacks are very violent (i.e., the CD-1 mouse does not let go of the C57BL/6J mouse); otherwise, the adolescent mice will be injured, and attacks need to be limited. Treating injuries is also very important to avoid infections; it is recommended not to use oily ointments for treating injuries because animals will lick away the ointment, losing hair in the process and exposing the injury.

Troubleshooting AcSD to be used for adolescent C57BL/6 male and female mice
The chronic social defeat stress model presented important limitations for its use in early adolescence in both male and female C57BL/6 mice. In this section, the modifications that were made for the development of the AcSD model in adolescence are described.

Equipment: The complete list of equipment is listed in the Table of Materials. The C57BL/6 adolescent mice in early adolescence (PND 21-30) were too small for the equipment described previously used in the laboratory⁶⁰. The diameter of the perforations on the acrylic glass dividers are large enough to allow C57BL/6 mice to cross from one compartment to the other, where the aggressor resides. To prevent this, acrylic glass dividers that have perforations of a small diameter were made. In addition, the steel-wire top needed to be secured with clips to prevent mice from escaping the cages. The cover lids employed are typically used for ventilated cages. In this case, these cover lids were used without the protective cotton to allow proper ventilation and to prevent mice from escaping from the cage. It is important to consider that the amount of woodchip bedding used should allow adolescent mice to reach the food and the water dispenser placed on the cover lid. However, large amounts of woodchip can become wet.

Selection of the CD-1 mice: In the chronic model, the CD-1 mice are screened for aggressive behavior prior to the beginning of the experiment. CD-1 mice are selected after 3 days of screening, and their aggressive behavior is assessed. After screening, the animals that have latencies to attack under 30 s and that show consistent attacks are selected for the next phase. However, during the chronic social defeat experiments, the CD-1 mice did not maintain a consistent number of attacks for 10 days (Figure 2B). Therefore, several CD-1 mice had to be removed from the experiment due to a lack of aggression towards the intruder mouse. In fact, while only a small proportion of the CD-1 mice attacked the intruder mouse, some of these mice were extremely aggressive and sometimes severely injured the male C57BL/6 PND 21 mice. Therefore, the screening of the CD-1 mice was changed to select aggressive mice towards adolescent mice, heightening the physical aggression.

Consistency of physical aggression: Consistent physical aggression leads to obtaining a rapid and robust defeat of the intruder⁶. To standardize procedures between experiments and experimenters, an operational definition of attack was used⁵⁷, allowing for precise quantitative measurement of aggressiveness.

Priming aggressiveness: During the chronic social defeat experiments, the number of attacks varied greatly, decreasing during the last days of the experiment (Figure 2B). To increase the aggressive behavior of the CD-1 mice towards the adolescent C57BL/6 male mice, the CD-1 mice were primed to be actively aggressive regardless of the mice that were introduced to their home cage. This behavior was achieved by introducing an adult C57BL/6 male mouse for 1 min before the adolescent mouse, encouraging the CD-1 mouse to attack the adolescent mice after the adult was retrieved.

Duration of the model: Adolescent C57BL/6 mice at PND 21 are very small, and CD-1 mice could severely injure them with just one attack. Therefore, to target early adolescence and still have repeated sessions of social defeat, an alternative model to the classic 10-day chronic social defeat stress was selected. The accelerated social defeat stress (AcSD)⁴² consists of two sessions of physical aggression per day, one session in the morning and one in the afternoon, for a total of 4 days (8 sessions). The AcSD model allowed an increase in the age of the adolescent mice from PND 21 to PND 25, still targeting early adolescence only.

Limitations of AcSD in adolescence
The AcSD model has some limitations. Compared with the chronic social defeat model in adults, AcSD requires the screening of twice the number of CD-1 mice of adolescent C57BL/6 mice. The duration of the priming phase can vary depending on the CD-1 mice's aggressive behavior; as such, the AcSD schedule should be planned with anticipation, and all these variables need to be considered. Since most adolescent mice will show resilience against social avoidance after AcSD exposure, to obtain a representative number of susceptible mice, the experimental cohort should comprise at least 20 animals for males and 30 for females.

AcSD in adolescent mice yields consistent results across experimental cohorts
After exposure to social stress with both chronic social defeat and AcSD protocols, most adolescent male mice showed resiliency against social avoidance. However, for the chronic social defeat model, it was not possible to discard the variability in the number of attacks by the CD-1 as a confounding variable. In contrast, for AcSD experiments, CD-1 mice are primed before every defeat session to obtain a consistent number of attacks across sessions (Figure 2D,E). Overall, all of the experimental cohorts of mice exposed to AcSD show a similar number of attacks across sessions (Figure 2D); however, this is not the case for mice cohorts exposed to the chronic social defeat model (Figure 2B).

Using the chronic social defeat stress in adult C57BL/6 female mice has been challenging in comparison with others using this model in other species and strains^9,13,45,61. For AcSD experiments in adolescent C57BL/6 female mice, the priming phase was extended, ensuring the aggressive behavior of the CD-1 mice (Figure 4A). However, variability can be observed in the number of attacks across female experimental cohorts (Figure 4B). Nonetheless, in these experiments, all CD-1 mice showed aggressive behavior towards females. To increase the number of attacks, the priming with the male mice can be extended to 1 min. As in males, the proportion of adolescent C57BL/6 female mice resilient against social avoidance was higher than that those showing susceptibility (Figure 4C,D). Furthermore, the number of attacks did not differ between resilient and susceptible mice (Figure 4E), indicating that the phenotype is independent of attacks. AcSD can be used in adolescent C57BL/6 female mice and experiments can be done at the same time in both sexes using the same CD-1 mice.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
C57BL/6 adolescent mice	In house breeding		Mice were breeded at the Neurophenotyping Centre of the Douglas Mental Health University Institute.
C57BL/6 adult mice	Charles River Laboratories	Strain Code: 027	Mice are ordered so as to arrive at PND>65 and are group housed (four mice per cage) in standard mice cages.
C57BL/6J adolescent mice	Jackson Labs	Strain Code: 000664; RRID:IMSR_JAX:000664	Mice are ordered so as to arrive at PND 24 and are group housed (four mice per cage) in standard mice cages.
CD-1 mice	Charles River Laboratories	Strain Code: 022	Mice retired breeders more than three months of age and singled housed throughout.
Cleaning solution	Virox Animal Health	DIN 02537222	Prevail: Accelerated Hydrogen Peroxide. Desinfectant cleaner and deodorizer.
Clear perforated acrylic glass divider	Manufactured by Douglas Hospital, custom order		0.6 cm (w) × 45.7 cm (d) × 22.23 cm (h); perforations of 0.6 cm diameter. The dividers are perforated allowing sensory but no physical contact between the pair of mice.
Clear rectangular rat cages	Allentown		24 cm (w) × 48.3 cm (d) × 22.23 (h).
Cotton squares for bedding	Inotiv Envigo	T.6060 iso-BLOX	2.5 cm x 2.5 cm. Added to the social defeat apparatus.
Hard woodchip bedding	Inotiv Envigo	Teklad 7090, 7115	Sani-chip bedding.
Large binder clips to secure the steel-wire tops	STAPLES	Item #: 132429, Model #: 24178-CA	51 mm
Medium binder clips to secure the steel-wire tops		Item #: 132367, Model #: 24172-CA	32 mm, in case the cover lids of the rat cages do not close with the large clips
Pain relief cream	Polysporin		Plus Pain Relief Cream (red format, NOT ointment), 2 Antibiotics plus lidocaine hydrochloride
Paired Steel-wire tops			24 cm (w) × 48 cm (d) with 0.6 cm (w) of separation between the grill
Removable wire-mesh enclosure	Johnston industrial plastics		11 cm (w) × 6.8 cm (d) × 42 cm (h) custom order; two per social interaction test arena secured in precut clear polycarbonate
Social interaction open-field arena	PEXIGLAS		45 cm (w) × 45 cm (d) × 49 cm (h), custom-crafted from opaque acrylic glass (Plexiglas)
Stopwatch			For timing defeat sessions
Video camera with infrared lights	Swann	SRDVR-44580V	Swann Camera - 4 Channel 1080p Digital Video Recorder & 2 x PRO-T853
Video tracking software	Topscan		2.0 Clever Systems Inc.