Breeding of mice
To cause Itgb3 loss of function, Itgb3fl/fl mice  on the C57BL6/J background (Jackson Labs #028232) were first crossed with Emx1-Crecre/cre mice  on the C57BL6/J background (Jackson Labs #005628). Then, resulting Emx1-Crecre/+,Itgb3fl/+ mice (“+” refers to the wildtype allele) were crossed with each other to generate conditional knockouts (cKO; Emx1-Crecre;Itgb3fl/fl), conditional heterozygotes (cHET; Emx1-Crecre;Itgb3fl/+), and wildtype (WT) controls, consisting of WT Cre- (Emx1-Cre+/+) and WT Cre + (Emx1-Crecre;Itgb3+/+) on the C57BL6/J background. Thus, two generations of breeding occurred before cKO mice were available. cKO, cHET, and WT mice were all taken from the same generation. Cre-mediated excision of floxed genes in the Emx1-Crecre mouse occurs prenatally, is robust and efficient , and integrin β3 expression in the cerebral cortex of cKO mice is significantly reduced by postnatal day 23 , if not earlier. Emx1 expression occurs in nearly all forebrain excitatory neurons, astrocytes, and oligodendrocytes, particularly in cortex and hippocampus, with sparser Emx1 expression in the olfactory bulb, amygdala, and piriform cortex .
Housing and testing of mice
WT, cHET, and cKO mice came from six separate litters from the same generation and were 3–6 months old at the time of experiment. Mice were weaned at 3–4 weeks of age with their same-sex littermates into a single standard mouse cage and were not segregated in any other way, including genotype. Two to five weaned mice were kept in each cage, the mean cage occupancy was 3.5 mice, and the median cage occupancy was 3. All mice were always kept on a 12-h light and 12-h dark cycle, and mice were tested during the middle 8 h of the light phase of the cycle. Weaned mice were given Global 18% Protein Rodent Diet (Envigo), access to water at all times, and had 1/4 inch corncob bedding (Envigo). Cage changes occurred once a week throughout the life of the animals. Mice were tested at least 18 h but no more than 5 days after their last cage change. The mean number of days between cage change and testing was 2.9 days, and the median number of days between cage change and testing was 3. Cages were kept on an individually-ventilated cage rack (Allentown). Testing occurred over 11 experimental days (16 calendar days), and 1–5 mice were tested on each experimental day. The sexes and genotypes of the mice tested on each day were randomly determined and experimenters were blind to the genotype of each animal. On the first day that a mouse was tested, it underwent all behavioral tests in the order listed below, starting with the elevated plus maze, with three exceptions: the home cage self-grooming test occurred the next day, and the three-chambered sociability test and three-chambered social novelty test (with male mice as stranger mice) occurred a few weeks later. In other words, mice experienced three experimental days: the first day with the bulk of experiments, the home cage self-grooming test on the very next day, and the sociability and preference for social novelty test (with male mice as the stranger mice) on a day a few weeks later. After completing behavioral experiments, mice were euthanized to collect their brains for further analysis (see Additional file 2: Table S6). In accordance with the James Madison University Institutional Animal Care and Use Committee, and using guidance from the American Veterinary Medical Association Guidelines for the Euthanasia of Animals, mice were fully anesthetized and unconscious following a lethal intraperitoneal injection of ketamine (240 mg/kg)-xylazine (48 mg/kg). Acepromazine (1.85 mg/kg) was also administered with the ketamine-xylazine as a tranquilizer. Once mice were fully anesthetized and unconscious, they were euthanized by transcardial perfusion with ice-cold 1 × phosphate-buffered saline followed by 4% paraformaldehyde in phosphate-buffered saline.
Sample size calculation
The Mead resource equation [13, 14] was used to estimate a sample size that would be large enough to provide sufficient statistical power for observing an effect in the behavioral tests listed below. This method was chosen over a prospective power analysis because the estimated effect sizes for the planned multifactorial experiments in this study could not be objectively determined. The Mead resource equation aims to maximize the power involving multifactorial animal experimentation by achieving approximately 10–20 error degrees of freedom (DF) in the design of the experiment. Briefly, the error DF is the total DF (in this study, it was the number of mice minus one) minus the model DF (in this study, depending on the experiment, it was five or six). Thus, achieving 20 error DF for all experiments was estimated to require 27 mice, or at least 9 mice per genotype (WT, cHET, and cKO) and at least 14 mice per sex. Six litters were needed to achieve these minimum genotype and sex requirements for a total of 39 mice, and all 39 mice were used for experimentation. Of the 39 mice, 18 were WT (8 female and 10 male), 12 were cHET (8 female and 4 male), and 9 were cKO (5 female and 4 male), for an overall total of 21 female and 18 male mice.
Overview of behaviors
Adult WT, cHET, and cKO mice underwent experiments that measured repetitive behaviors in home and novel environments (self-grooming), anxiety (elevated plus maze, open field test); hyperactivity and locomotion (open field test), compulsive behaviors (marble burying), and sociability and preference for social novelty (three-chamber social tests). An olfaction test was not conducted because global Itgb3 knockout mice do not have impaired olfaction . The design of the self-grooming test in a novel environment was replicated after McFarlane et al.  and Carter et al. . The design of the elevated plus maze and open field test were adapted from Carter et al. . The design of the marble burying test was adapted from Dohn et al. . The design of the three-chamber social tests was adapted from Moy et al.  and Carter et al. .
Order of behavioral experiments
Adult WT, cHET, and cKO mice underwent behavioral testing in the following sequence: elevated plus maze, open field testing, three-chambered sociability test (with a female mouse used as Stranger 1, see below), three chambered social novelty test (with a female mouse used as Stranger 2, see below), novel environment self-grooming test, marble burying test, and home cage self-grooming test. Because experimental mice were of both sexes, all mice were tested in the three-chambered sociability test and three-chambered social novelty test a few weeks later, this time with male mice as Stranger 1 and 2 (S1/S2, see below). The validity of retesting in the three-chambered sociability test has been established by Moy et al. . Similarly, sociability has been shown in opposite-sex (female mice interacting with male S1/S2), and same-sex (male mice interacting with male S1/S2) contexts [34, 35].
Elevated plus maze
The plus maze consisted of four opaque acrylic arms (each arm 10 cm × 30 cm) connected in a “plus-sign” configuration and elevated approximately 40 cm. Two of the arms had opaque acrylic walls (20 cm H) on three sides and two arms had no walls. The mean illuminance of the maze was approximately 360 lx. A mouse was placed in the center of the maze at the beginning of a 5-min test period with no prior acclimation or exposure to the maze. The position of the mouse was recorded using an overhead video camera. The maze was cleaned with 95% ethanol in between every test run, and once at the beginning of each testing day, to eliminate any odor cues. FIJI software  was used to manually analyze the video for the duration of time spent in the open arms, and the percent time spent in the open versus closed arms. When mice in the closed arms approached the open arms, they would sometimes display a “stretched-attend posture”  or a “head dip” behavior  into the open arms of the maze. Because these behaviors were difficult to distinguish via overhead video, these types of behaviors were combined during analysis and called “peeking”, which was defined as any time the mouse reached into the open arm and returned to the closed arm while maintaining at least one limb in the closed arm section. The duration and the number of peeking behaviors were recorded. Video analysis was done blind to genotype and sex.
Open field test
The open field test consisted of a mouse being placed in an open-topped, opaque acrylic box (63 cm L × 63 cm W × 63 cm H) that allowed free movement, with no prior acclimation or exposure to the open field. The mean illuminance of the field was approximately 240 lx. The mouse was placed in the center of the box and recorded with an overhead video camera. Mice explored the open field for 15 min, as in Carter et al. . The box was cleaned with 95% ethanol in between every test run, and once at the beginning of each testing day, to eliminate any odor cues. Videos were then manually analyzed using FIJI software to determine the average distance traveled per minute and the number of times the mouse crossed into the middle Sect. (22.5 cm × 22.5 cm) of the field (the middle section was not demarcated physically on the open field, but demarcated during video analysis). Video analysis was done blind to genotype and sex.
Three chamber sociability test
The three-chamber sociability test involved individual mice being placed in an opaque acrylic box (63 cm L × 63 cm W × 63 cm H) divided into 3 chambers (each 21 cm L × 63 cm W × 63 cm H). Passages (20.7 cm centered along the width of the chamber) allowed for free movement across all chambers. The mean illuminance of the chambers was approximately 240 lx. Chamber 1 (located to the experimenter’s left) contained an inverted wire pencil cup with a female C57BL6/J mouse inside (Stranger 1) that had no prior contact with the experimental mouse. Neither female stranger mice nor female experimental mice were checked for estrus, but experimental mice of all groups were tested over 11 days, so the proportion of experiments with female mice in estrus was assumed to be roughly equal across all groups. The center of the wire cup was placed approximately 10.5 cm away from the leftmost and furthest walls when viewed by the experimenter. The wire cup allowed for nose contact and detection of odor cues but prevented further interaction between mice. Chamber 3 (located to the experimenter’s right) contained an identical inverted wire pencil cup placed approximately 10.5 cm away from the rightmost and furthest walls (when viewed by the experimenter), but without a mouse inside. The location of Stranger 1 was systematically switched between Chambers 1 and 3 between test runs. Each experimental mouse was first acclimated to the three chambers without the presence of pencil cups or Stranger 1 for 10 min. Following acclimation, the experimental mouse was removed, the pencil cups and Stranger 1 were placed in the maze, and then the experimental mouse was placed in the middle chamber and allowed to explore the three chambers for 10 min. The position of the experimental mouse was recorded via overhead video camera. The amount of time spent within 1 cm of each pencil cup and the number of entries into each chamber were recorded by the experimenter, who was blind to genotype. This same test was run on all experimental mice several weeks later, this time using a male C57BL6/J as Stranger 1 (as noted above, the validity of retesting in the three-chambered sociability test has been established by . Videos were manually analyzed for time spent in each chamber and number of trips to each chamber, using FIJI software. Video analysis was done blind to genotype and sex.
Three chamber social novelty test
Following the sociability test, the experimental mouse was removed from the chamber while another novel C57BL6/J mouse (Stranger 2), unknown to the experimental mouse, was placed under the previously empty pencil cup. The experimental mouse was then placed back into the middle chamber of the three-chamber maze for 10 min of assessment. The experimental mouse could freely explore the chamber and interact with Stranger 1 and Stranger 2. Mice were recorded via overhead video camera. Time spent within 1 cm of each pencil cup and number of entries into each chamber were recorded by the experimenter, who was blind to genotype. As noted above, this same test was run on all experimental mice several weeks later, this time using male C57BL6/J mice as Stranger 1 and 2. Time spent in each chamber and number of trips to each chamber were analyzed later using the video and FIJI software. Video analysis was done blind to genotype and sex.
Self-grooming in a novel environment
Self-grooming behavior was measured by placing each mouse inside a 37 cm × 23 cm × 22 cm open, empty cage with no bedding. The mean illuminance of the cage was approximately 400 lx. Each mouse was allowed to habituate to the novel environment for 10 min. The time spent self-grooming for the next 10 min was recorded by the experimenter, who was blind to genotype.
Marble burying test
A novel cage was prepared for each mouse with a 3 cm-thick layer of bedding in order to allow the burying of 1.5 cm diameter marbles. The mean illuminance of the cage was approximately 400 lx. Each mouse was placed in the novel cage without marbles for 10 min of acclimation. Following the acclimation period, each mouse was briefly removed from the cage and 20 marbles were placed in a four-by-five grid on top of the bedding, with 2 cm of space between each marble in all directions. The mouse was then given 10 min to explore and interact with the marbles. After this period the mouse was removed, and marble burying was quantified by the experimenter, who was blind to genotype. Marbles completely buried were given a score of two points while marble partially buried received a score of one point. Marbles that were not buried were given a score of zero points. The marble score was the sum of all the points obtained from the interactions with the 20 marbles.
Home cage grooming test
After the above behavioral tests were completed, individual mice were placed in a single housed “home cage” and given 24 h to acclimate to it. The cages were kept on a standard cage rack, and ambient light reached the façade of each cage at approximately 200 lx. After a 24-h acclimation period, an observer who was blind to genotype approached the cage without disturbing it in any way and recorded the time the mouse spent self-grooming for 10 min. Afterwards, mice were returned to their original group housing (see "Housing and testing of mice").
Analysis of data
For self-grooming experiments, the factors involved were environment (home, novel), sex (female, male), genotype (WT, cHET, cKO), and their respective interactions. For sociability and preference for social novelty experiments, the factors involved were chamber (in sociability: Obj, S1; in preference for social novelty: S1, S2), sex (female, male), genotype (WT, cHET, cKO), and their respective interactions. Three-way ANOVA was used to determine if there were any interaction effects among all three factors (e.g., environment, sex, and genotype) or between any factor pairs (e.g., environment and sex, environment and genotype, sex and genotype). Because there were no three-way interactions (see “Results”), two-way ANOVA was used to determine and understand main effects by factor. Repeated measures ANOVA was not used for self-grooming, sociability, or preference for social novelty experiments because there was no significant within-subject interaction in the repeated measure (in all cases, p > 0.5). For EPM, OFT, and MBT, the factors involved were sex and genotype, and two-way ANOVA was used for analysis. When a main effect was found to be p < 0.001 in two-way ANOVA, Šidák's multiple comparisons and within-group ANOVA were used as post-hoc tests. For comparisons between two groups in Additional file 2: Table S5, n = 9 per group was too small to pass the D'Agostino-Pearson normality test, so the data were not assumed to be normally distributed and the Mann–Whitney U-test was used to compare means. SPSS 28 was used to calculate three-way ANOVA and two-way ANOVA. GraphPad Prism 9 was used to confirm two-way ANOVA results in SPSS and for all other statistical testing. GraphPad Prism 9 was used to represent data graphically.