Immune Deficiency Influences Juvenile Social Behavior and Maternal Behavior

Abstract

Mice with severe combined immunodeficiency (SCID) lack functional T and B-lymphocytes, and have impaired cognitive abilities. Here, we assessed social behaviors in male SCID and C57BL/6 (B6) juvenile mice. In a social preference task, SCID mice spent more time than B6 mice investigating a novel adult male mouse. In a social recognition task, SCID mice habituated to a novel ovariectomized mouse, but failed to show dishabituation when presented with an unfamiliar individual. We hypothesized that partial immune restoration could normalize behaviors. SCID pups (postnatal day 7) received either saline or splenocytes from normal donors. Splenocyte-replaced SCID mice spent less time interacting with a novel mouse than saline-injected SCID or B6 control mice. Again, control SCID mice failed to dishabituate to a novel mouse, but splenocyte-replaced SCID mice showed dishabituation. In both of these studies B6 and SCID pairs were used to produce offspring that remained with their dams until weaning. There are no studies of maternal behavior in SCID dams; therefore to investigate the potential role for this factor we quantified maternal behavior in SCID and B6 dams; several significant differences were found. To control for differences in maternal care we mated heterozygous SCIDs to produce offspring. These homozygous SCID and WT offspring reared by dams of the same genotypes displayed similar responses to a novel mouse; however, in the social recognition task SCID males did not display dishabituation to a novel mouse. Taken together, our data indicate that gene by environment interactions influence social interactions in immune deficient mice.

Introduction

Immune dysfunction is associated with learning impairments in mice (Brynskikh et al., 2008, Derecki et al., 2010a) and several psychiatric disorders in humans (Michel et al., 2012). Specifically, autism spectrum disorder (ASD) has been correlated with infection during pregnancy (Atladottir et al., 2010, Zerbo et al., 2013), and with familial autoimmune disease (Atladottir et al., 2009, Comi et al., 1999, Ly & Mostafa, 2014). Although cognitive ability has been examined in mice that undergo developmental immune dysfunction, and social behavior has been evaluated following acute immune challenges, few studies have been conducted on social behavior and developmental immune dysfunction. Here, we use a mouse model with a mutation that prevents maturation of basic immune cells. Severe Combined Immunodeficiency (SCID) mice lack adaptive immunity due to a spontaneous mutation in the Prkdc gene on chromosome 16, which impairs the recombination of antigen receptor genes and results in the interrupted development of B and T cells (Bosma et al., 1983, Bosma & Carroll, 1991, Hsiao et al., 2012). Previous studies using SCID mice showed that adult males have impaired spatial learning, which is rescued with splenocyte, or purified T cell, replacement (Brynskikh et al., 2008, Derecki et al., 2010a). However, no other behaviors have been examined in this mouse model. In these studies, we evaluated two social behaviors in juvenile SCID mice. The juvenile time period is one time used for testing social behavior in mice (Pearson et al., 2012), and childhood is a common time period for diagnosis of social disorders (Kogan et al., 2009). In these tests first we observed behavioral differences between SCID and B6 mice, next we asked if splenocyte transfer to SCID pups could rescue social deficits.

Importantly, in all previous behavioral studies of behavior in SCID mice, SCID and B6 dams reared their own litters (Brynskikh et al., 2008, Derecki et al., 2010a). This design fails to take differences in maternal behavior into account, an important factor that influences offspring behaviors (Champagne et al., 2003, Schwendener et al., 2009, Zimmerberg & Sageser, 2011). Therefore, in a third experiment we observed maternal behavior in SCID versus B6 dams and found significant differences in time spent on the nest and nursing. In the last study we bred B6 mice heterozygotic for the SCID gene mutation to generate litters composed of heterozygous, SCID, and wild-type (WT) pups. Our data demonstrate, for the first time, social deficits in male SCID mice, as well as differences in maternal behaviors between SCID and B6 dams.

Methods

Animals

Breeding pairs of B6.CB17-Prkdc^scid/SzJ (SCID, Stock #001913) were obtained from Jackson Laboratory (Bar Harbor, ME). Our colony C57BL/6J mice were originally purchased from Jackson Laboratory, bred and maintained in the University of Virginia School of Medicine, Jordan Hall Animal Facility. The University of Virginia Animal Care and Use Committee approved all procedures used and described here. All animals were maintained on a 12:12 light/dark cycle (lights off at 1300), and provided food (Harlan Teklad diet #7912) and water ad libitum.

Experimental Designs

Experiment 1

Male and female neonates were left undisturbed in the home cage until weaning at postnatal day 21 (PN21). A total of 8 SCID and 9 B6 males were tested (from 3 litters for each group) in the social recognition task first, followed one of more days later by the social preference task, this testing order remained the same throughout all experiments. The SCID males were tested as adults (between PN60-90) in the olfaction test described below.

Experiment 2

Male pups received a single injection (ip) of saline or purified splenocytes as described below on postnatal day 7 (PN7; day of birth was counted as PN0). Female littermates were not disturbed. A total of 12 SCID males (from 6 litters) were injected with saline and 12 (from 5 litters) received splenocytes. Treatments were randomly assigned by litter. We did not split treatments within litters because we were concerned that the stress produced by individually marking males within litters would be a confounding factor. An additional 14 B6 males (from 4 litters) received saline injections on PN7. Other than this brief intervention, animals were left undisturbed in the home cage.

Experiment 3

Ten SCID dams, and 12 B6 dams were observed after parturition on PN1, 6, 7, and 8. On PN7, pups received splenocyte or control saline injections (as in Experiment 2) and the B6 litters were either used for spleen donors or males received saline. Thus, on PN8 we observed 4 SCID dams with saline injected pups, 4 with splenocyte replenished pups and 7 B6 dams with saline injected litters.

Experiment 4

Female SCID and male B6 mice were mated to produce heterozygous offspring to use for breeding pairs. Pairs heterozygous for the Prkdc mutation produced litters containing homozygous wild-type animals lacking the mutation (WT), mice that were heterozygous for the mutation (HET), and SCID mice with two copies of the mutated gene. The offspring from heterozygous pairs were genotyped using DNA from tail snips screened by PCR for Prkdc with the following primers: Forward: 5′GGAAAAGAATTGGTATCCAC3′; Reverse: 5′AGTTATAACAGCTGGGTTGGC3′, and only WT and SCID mice were tested. A total of 6 males (from 4 litters) that were homozygous for the SCID mutation along with 8 male (from 6 litters) WT littermates were tested.

In all experiments, at weaning, male mice were group-housed with same-sex littermates (no more than 5 per cage). Between PN22 and 27, the animals were tested for social preference and social recognition.

Juvenile social behavior (Experiments 1,2, and 4)

Social Preference

This test was performed as described previously (Cox et al., 2010). Briefly, 1 hr before testing mice were moved (in their home cages) into the testing room. Mice were placed into a three-chambered box (76.2 cm × 26.67 cm × 17.78 cm) divided by black Plexiglas walls and on outer wall was also black Plexiglas. Thus, the center section had dark walls on 3 sides with 2 openings leading to the outer chambers. Each of the outer two chambers contained an empty upside down metal pencil holder (10.16 cm diam. × 13.97 cm), hereafter referred to as a “holding cell,” with a round top and vertical bars (spaced one cm apart). Each mouse was habituated to the empty cage for 10 minute, and then restricted to the center chamber while an adult B6 male mouse was placed randomly in one of the holding cells on one side of the test box. The doors were opened and the subject explored the entire box for 10 min. The test was conducted and recorded during the dark portion of the day/light cycle between 1300 and 1800h under red-light. An observer blind to treatment group scored the amount of time the subject spent in each part of the test box, as well as time spent investigating the mouse or the empty holding cell. A preference score (the amount of time spent investigating the holding cell containing the male mouse – the amount of time spent investigating the empty holding cell) was calculated for each mouse.

Social Recognition

This protocol was adapted from one previously described (Tejada & Rissman, 2012, Wolstenholme et al., 2013). Mice were moved into the testing room and allowed to habituate in their home cages for at least 1 hr. The mice were then placed into a clean cage (the size of their home cages) containing a holding cell for 10 mins. after which time an ovariectomized adult female was placed into the holding cell for 1 min. During each trial the amount of time (in seconds) spent investigating was recorded. Here we define investigation as nose contact between the subject and the other mouse or the bars of the holding cell. The 1-min trials were separated by 9 mins, when the test male was alone in the test cage. Between trials, the ovariectomized females were single-housed in clean cages. This was repeated 8 times and on the 9^th trial a new stimulus animal (another ovariectomized adult female) was used in place of the familiar mouse. Testing was conducted during the light portion of the day/night cycle between 0800 and 1300 h and behavior was scored in real time.

Olfaction Test (Experiment 1)

To habituate the mice to a novel food, Cocoa Puffs (General Mills, Inc.) were placed in their home cage in addition to their ad libitum food and water. Twenty-four hours later, all food was removed from the food hopper and cage and the mice were fasted overnight. The next morning, each mouse was placed alone in a clean cage with a Cocoa Puff hidden beneath the clean bedding. The time to find the hidden Cocoa Puff was recorded. This test was performed during the light portion of the day/light cycle between 0900 and 1200 h and ability to locate the Cocoa Puff within 5 minutes was considered normal as described previously. (Yang & Crawley, 2009)

Splenocyte Transfer (Experiment 2)

Splenocytes were collected from B6 pups (PN3-11), taken from litters whose pups were not used in these experiments. The splenocytes, were collected and transferred to PN7 SCID pups. To collect the splenocytes pups were wiped with ethanol and placed on ice for euthanasia followed by rapid decapitation and spleen removal. The spleens were strained through a sterile cell strainer in phosphate buffered saline (PBS) containing 2% heat inactivated fetal bovine serum (FBS), and then transferred to a 15 mL tube with a transfer pipette. The cells were spun at 1100 RPM (234 RCF on 173mm rotor) for 8 minutes and the supernatant discarded. The resulting pellet was resuspended in ammonium chloride-potassium red blood cell lysis buffer and incubated on ice for 5 minutes. Then, the tube was filled with PBS containing 2% heat inactivated FBS and spun again at 1100 RPM for 8 minutes. The supernatant was discarded and the resulting pellet suspended in saline at a volume of 150μL/spleen collected. Each recipient received 150μL injected intraperitoneally (i.p.), the equivalent of one donor spleen, using a sterile syringe and needle. Control-injected pups received 150μL of sterile saline.

Maternal Behavior (Experiment 3)

First time SCID and B6 dams (used to generate offspring for Experiment 2) were the subjects in this experiment. Dams were observed in their home cages for 30 minutes, once a day during the light (0100–1300h), and a second time during the dark (1300-0100h) on PN1 and 6 and during the light on PN7 prior to injection. Thus the pre-injection observations were based on 2.5 hours of maternal behavior. On PN7 half of the B6 litters were sacrificed for splenocyte collection. As in Experiment 2, SCID mice received saline or splenocyte injections and the remaining B6 litters received saline. We continued to observe maternal behavior during the dark portion of PN7 and twice during PN8 (light and dark) for an additional 1.5 hours of observations after the nests were disturbed and pups injected.

We used scan-sampling methods to record maternal behaviors. We observed the nest every 15-seconds, recorded all behaviors, and noted if the dam was on or off of the nest. On the nest behaviors were: licking and grooming pups, self-grooming, active nursing, passive nursing, nest building and hovering. Behaviors off of the nest were eating/drinking, self-grooming, and digging/climbing. These methods were adapted from previous research (Chourbaji et al., 2011). Values were analyzed as proportions of total observations.

Statistical Analyses

The social recognition data were analyzed with repeated measures ANOVA. The factors used for the analysis were time spent investigating the stimulus animal and trial. Other data were analyzed using one-way ANOVAs. In all cases we used Fisher’s LSD post-tests to detect pairwise differences.

Results

Experiment 1: Social behaviors differ between juvenile male SCID and B6 mice

We measured time spent in all three sections of the three-chambered test box and time investigating the holding cells (either empty or containing an adult male). When time investigating the empty holding cell was subtracted from time investigating a novel male mouse SCID males had a larger preference for the novel mouse than did the B6 mice (Figure 1A; F(1,14)=9.42, p=0.009). SCID mice also spend more time in the center chamber (Table 1; F(1,14)=5.58, p=0.0344), less time in the empty chamber (Table 1; F(1,14)=10.83, p=0.006), less time investigating the empty cell (Table 1; F(1,14)=5.20, p=0.04), and more time investigating the cell with the stimulus mouse (Table 1; F(1,14)=6.15, p=0.026) than B6 mice.

Figure 1.

Means +/− SEM SCID and B6 male mice examined in social preference (a) and social recognition (b) tasks. A) SCID mice have higher preference scores for an adult male mouse than do B6 mice (*p<0.05) B) Both SCIDs and B6 male mice habituated to an ovariectomized mouse over 8 trials (**p<0.05), but only B6 mice dishabituated in response to a novel stimulus (***p<0.05). SCID N=8, B6 N=9

Table 1.

Social Preference Data for Experiments 1,2, and 4: Time spent in each chamber or investigating a holding cell with or without another mouse in it presented as mean ± SEM in sec. Significant differences between B6 groups and other groups from the same experiment and same column are represented as *(p<0.05).

Experiment number	Group	Time in chamber with empty cell (s)	Time in center chamber (s)	Time in chamber with mouse (s)	Time spent investigating cell with mouse (s)	Time spent investigating empty cell (s)
1	B6	177.81 ± 17.88*	94.17 ± 13.02*	328.02 ± 14.21	124.87 ± 12.60*	40.63 ± 8.29*
1	SCID	95.61 ± 17.08	152.33 ± 24.15	352.06 ± 30.62	174.45 ± 16.86	17.44 ± 3.79
2	B6+saline	168.9 ± 20.54	68.65 ± 13.81	362.59 ± 26.12	141.18 ± 14.07*	18.45 ± 5.25
2	SCID+saline	90.57 ± 16.76	62.08 ± 15.91	446.75 ± 27.42	245.42 ± 21.66	19.54 ± 5.28
2	SCID+splenocytes	194.97 ± 22.87	129.24 ± 23.07	275.14 ± 24.73	93.85 ± 17.01	50.92 ± 12.38
4	B6 Males	253.74 ± 38.51	150.40 ± 17.30	196.14 ± 38.43	53.33 ± 12.39	16.77 ± 5.70
4	SCID Males	253.79 ± 46.40	104.52 ± 15.21	241.80 ± 49.38	62.83 ± 15.39	26.22 ± 7.15

In the social recognition task we found an interaction between trial and immune status for trials 1,8, and 9 (Figure 1, F(2,50)=3.44; p=0.045). Planned comparisons indicated that all mice reduced investigation time between trial 1 and 8 (B6 p<0.0001, SCID p=0.003), indicating habituation but SCID mice failed to significantly increase investigation time between trials 8 and 9 (p=0.099), whereas mice did (p<0.0001) suggesting an inability to dishabituate and recognize the novel female.

All SCID male mice located the hidden food in the allotted time (Mean=97.63 ± 19.8 seconds), indicating that general olfaction in the SCID mouse is not impaired.

In sum, SCID males had a larger preference for social interactions with the stimulus mice and they failed to show dishabituation in the social recognition experiment. Moreover gross olfactory abilities were normal in SCID males. Thus, social behavior in two commonly used tests were affected by immune deficiency.

Experiment 2: Splenocyte transfer decreases responses to a novel mouse and ameliorates dishabituation deficit in juvenile SCID mice

SCID mice that received saline injections spend more time investigating the chamber with the stimulus mouse than B6 mice given saline and SCID mice that received splenocytes (Table 1). The preference scores revealed that SCID given a splenocyte injection had a reduced preference for investigating a novel stimulus mouse than SCID mice that received saline (Figure 2A; F(2,37)= 3.93; p=0.029).

Figure 2.

Means +/− SEM SCID male mice with out without splenocytes replacement versus control B6 mice in social preference (a) and social recognition tasks (b). a) In the social preference task SCID mice that receive splenocytes have lower preference scores than SCID mice that receive saline (*p<0.05). b) SCID mice have impaired social recognition that is ameliorated with splenocyte transfer. All groups habituated to a stimulus mouse over 8 trials (**p<0.001). Only B6 and SCID mice with splenocyte transfer dishabituated on trial 9 (***p<0.001). SCID+splenocytes N=12, SCID+saline N=12, B6 N=14

In the social recognition task, an interaction between immune status and trial was found on trials 1, 8 and 9 (Figure 2B,(F(4,113)=5.30; p=0.0009). Planned comparisons indicated that all three groups habituated to a novel stimulus (trials 1 and 8 all groups p<0.0001). However, B6 and SCID mice that received splenocytes dishabituated (trials 8 and 9 both groups p<0.0001), while SCID mice that received saline did not do so (p=0.162) (Figure 2B).

In sum, these experiments show that injection of splenocytes two weeks prior to testing suppressed social interactions in SCID mice. Yet, in the social recognition task partial immune restoration normalized dishabituation responses.

Experiment 3: SCID dams show less maternal behavior than B6 dams

On P1, P6, and P7 SCID dams spent more time off of the nest (Figure 3A; F(2,100)=4.96;p=0.037), less time on the nest (Figure 3B; F(2,100)=4.97;p=0.037) and less time nursing (Figure 3C; F(2,100)=9.24;p=0.006) their pups than B6 dams. Following nest disruption (a saline or splenocyte injection given to pups), SCID dams spent less time actively nursing their pups in an arched back posture (Figure 3D; (2,44)=5.93; p=0.016) than B6 dams (Figure 3D). Several other behaviors were recorded, but none of these were different between the groups.

Figure 3.

Means +/− SEM Maternal behavior in SCID and B6 dams before pups are injected (a,b,c) and after injection (d). A) On P1, P6, and P7 SCID dams spent less time on the nest than B6 dams (*p<0.05) and b) more time off of the nest (*p<0.05) than B6 dams. c) Prior to injection SCID dams spend less time nursing than B6 dams (*p<0.05). All measures were taken from SCID N=12, and B6 N=12 on their nests. d) Following a saline injection or splenocyte transfer to male pups on P7, SCID dams spend less time actively nursing their pups (on P7 and 8) than did B6 dams (*p<0.05) SCID N=4, B6 N=7, SCID+splenocytes N=4.

Interestingly these experiments show that maternal behaviors in SCID dams differ from controls in several dimensions. First SCID dams spend less time nursing and second they are off the next more frequently than B6 control dams.

Experiment 4: SCID mice reared by heterozygous dams display social recognition deficits

In the social preference task, all mice spent more time investigating an adult male than an empty holding cell. No significant differences were observed for time spent in any chamber during the social preference test between genotypes, nor did social preference scores differ between the groups (Table 1).

In the social recognition task, an interaction between immune status and trial was found on trials 1, 8 and 9 (Figure 4B,(F(2,41)=5.16; p=0.014); planned comparisons indicated that both groups habituate to a novel stimulus (B6 p<0.0001, SCID p=0.009). However, only the wild type B6 mice displayed dishabituation on trial 9.

Figure 4.

Means +/− SEM SCID mice reared by heterozygous dams versus B6 mice reared by heterozygous dams in a social preference task a) and social recognition b). a) In the social preference test no significant differences in preference scores were found. b) In the social recognition task both SCID and B6 mice habituated to an ovariectomized mouse over 8 trials (**p<0.05), but only B6 mice dishabituate in response to a novel stimulus (***p<0.05).

In summary, one important environmental factor, maternal behavior was eliminated by use of heterozygous dams and when this was done SCID males had normal social investigation in a preference test. However the SCID males still showed a failure to dishabituate in the social recognition task.

Discussion

Adult male SCID mice have spatial learning impairments that can be ameliorated by splenocyte replacement (Brynskikh et al., 2008, Derecki et al., 2010a). Based on these studies the SCID has been suggested as a mutant model for Autistic disorders (Derecki et al., 2010b). Here we examined two social behaviors in SCID mice both of which are relevant to Autism. We found that SCID males had a larger preference for social interactions with the stimulus mice and they failed to show dishabituation in the social recognition experiment. Moreover gross olfactory abilities were normal in SCID males. Thus, social behaviors in two commonly used tests were affected by immune deficiency. Next we asked if providing SCID pups with splenocytes might restore some of these social behaviors. A single injection of splenocytes about 2 weeks prior to testing restored dishabituation, yet, this same treatment resulted in reduced investigation levels to those lower than control B6 mice. One possible explanation for these results is that they are due to a physiological immune response to foreign antigens from the splenocyte transfer, causing a reduction in activity. Alternatively the splenocyte dose may have been supra-physiological producing a hyperactive immune response. However, the animals showed no sign of illness or physical impairment and their general olfactory function was normal. Moreover, data from the social recognition test showed that the SCID splenocyte-replaced mice had normal habituation and dishabituation responses to restrained adult ovariectomized females.

In addition to immune function differences, we asked if maternal behaviors were different between SCID and wild type dams. We noted that SCID dams spent more time off the nest and less time in the nest actively nursing than did B6 dams. Instead of doing a cross fostering study in the final study we presumed that heterozygous dams would display similar maternal behavior. Moreover fostering itself can affect the behavior of pups tested as adults (Cox et al., 2013), thus performing a complete cross-fostering study is a rather large undertaking. We produced het-het pairs and then we assessed behavior in the male homozygous offspring (SCIDs and wild types). We found that the elevated social interactions in the preference test that were noted in the SCID males in Experiments 1 and 2 were not present when the offspring were produced by heterozygote pairs. Thus rearing conditions and splenocyte restoration were both able to modulate behaviors in this task. On the other hand the SCID males failed to dishabituate, an effect noted in all three studies. Yet, splenocyte treatment, in Experiment 2, was able to revise this behavioral deficit.

Relevant to the role of immunity in these behavioral outcomes, anti-inflammatory cytokine release of Interleukin-4 in the meninges of the brain due to T-cell activity can facilitate learning recovery in SCID mice (Derecki et al., 2010a). Therefore, it is reasonable to suggest that cytokine activity or something downstream, is at least partially responsible for this shift in social learning. Although the SCID mice that received splenocytes experienced a healthy cell transfer, it is clear that there is a delicate balance of activity influencing social behavior.

The lack of dishabituation demonstrated by SCID mice when a novel female was presented during the final trial of the social recognition task suggests an inability to distinguish the novel mouse from the familiar. It has been shown previously that adult mice lacking the recombination-activating gene RAG1 are also deficient in social recognition tasks (Mcgowan et al., 2011). RAG1 (along with RAG2) mediates variable (diversity) joining recombination, necessary for the maturation of B and T cells (Mcgowan et al., 2011). This variable joining recombination is the same process that is disrupted by the Prkdc gene in SCID mice, causing their lack of T and B cells (Bosma et al., 1983, Bosma & Carroll, 1991). These data indicate that this process and these cells are important for social recognition in adults, as well as in juvenile mice. However, SCID mice that received splenocytes were able to dishabituate in the social recognition paradigm, indicating that this partial immune restoration restores the ability to recognize a novel social stimulus. We speculate that this same procedure would restore dishabituation in RAG mutants also.

Some mouse models for autism exhibit social behavior deficits and also have deficiencies in immune function. For example, there is evidence for immune system disruption in the ASD mouse model BTBR T+tf/J (BTBR). The BTBR strain has higher levels of serum IgG and IgE, as well as elevated expression of pro-inflammatory cytokines IL-33, IL18, and IL1β in the brain compared to B6 control mice (Heo et al., 2011). Furthermore, previous studies have shown decreased social behavior of offspring whose dams had infections during gestation, this is likely linked to increased cytokine activity (Malkova et al., 2012). Additionally, reversal of social behavior deficits in a maternal immune activation (MIA) model of poly (I:C) injection, are noted after MIA offspring received bone marrow transplantation from immunologically healthy mice, this procedure rescues a proinflammatory phenotype (Hsiao et al., 2012). Our data add to the literature supporting the idea that normal social recognition/cognition is dependent, at least partially, upon normal immune function.

The mechanisms responsible for the behavioral differences between SCID and B6 males are unclear, but the current data does supply us with the necessary information to pursue this question in future studies. Importantly, the social behavior tasks we used are considered to be amygdala-dependent in mice (Ferguson et al., 2001, Keverne & Curley, 2004), so this region is likely to be directly affected, as are genes known to be involved in these social behaviors such as oxytocin and vasopressin (Keverne & Curley, 2004). In previous research regarding the immune system and social behavior, rats were exposed in utero, to an immune challenge, most commonly lipopolysaccharide (LPS). LPS-exposed male, but not female, rats displayed less juvenile play behavior than controls, and this behavioral change was correlated with reduced expression of vasopressin mRNA (Taylor et al., 2012). However, if the same immune-driven learning mechanisms are acting here, it is reasonable to hypothesize that spatial memory, and therefore hippocampal activity, may be critical as well. Cytokine release from T cells is an important element in the regulatory mechanisms of spatio-cognitive deficits in SCID mice (Derecki et al., 2010a), and may also influence the social cognition deficits seen in our study.

Our initial experiments, and all previous work on behavior in SCID mice, used paradigms without any consideration of potential effects of the dams’ maternal behavior and/or heterogeneity of littermates. Moreover, no assessment of maternal behavior has been conducted with these mice. This information is of value and arguably necessary for any future research using SCID mice. Maternal care in the form of licking, grooming, and nursing behavior has been shown in rats to affect the maternal behavior of offspring where pups exhibit the behavior that their mothers provide (Bailoo et al., 2014, Champagne et al., 2003, Romeo et al., 2003). Furthermore, maternal separation alters play behavior (Zimmerberg & Sageser, 2011) and maternal care and experience in general affect a variety of behaviors and changes in the brain in juvenile and adults offspring of mice and rats (Mousseau & Fox, 1998, Qvarnstrom & Price, 2001, Ressler & Anderson, 1973, Smit-Rigter et al., 2009, Suomi, 1997, Weaver et al., 2006). Here, we show for the first time that SCID dams spend less time on the nest and have less attentive nursing behavior than B6 dams. This has important implications for any future studies using these mice as a behavioral model.

In Experiment 3 we quantified maternal care in wild type and mutant SCID dams. Results of Experiment 3 indicated that SCID mice display differences in quality and quantity of maternal behavior as compared with B6 mice, following this observation we conducted another experiment using heterozygous pairs to create mixed litters. In our paradigm, all mice received care from heterozygous dams. We observed that the social preference behavior of these offspring was similar in the SCIDs and B6 mice. On the other hand, social recognition behavior remained deficit in the SCID male mice compared to B6 mice as previously observed. In Experiment 2, the phenotype was modified by splenocyte transfer. Thus, this behavior may not be regulated by effects of maternal care, but could be more directly influenced by peripheral immune status.

The data presented could provide insight into social behaviors in humans that are at risk in individuals with immune deficiencies. Understanding subtle differences in juvenile social behavior should enhance our ability to diagnose early childhood diseases such as ASD. Children with ASD have been shown to have immune deficiencies in certain populations (Jyonouchi et al., 2008). Specifically, cytokine responses can be dysregulated, and may be attributed to the child’s environment (Goines & Ashwood, 2012), which is why the interaction between early life stresses and other environmental disruptions with immune dysfunction deserve more investigation (Michel et al., 2012, Van Gent et al., 1997). This is further illustrated in animal studies using human antibodies. When IgG from human mothers of children with Autism are administered to pregnant mice, the offspring show increased anxiety and fear responses (Singer et al., 2009). Taken together, our current results show that immune deficiency can modify social behavior in juveniles, and that normal splenocytes, given prior to weaning, modifies these differences. However, maternal behavior is different between SCID and B6 dams both prior to and after nest disruptions, and this may well affect the behavior of the offspring as well. The SCID mouse may be a useful model for investigating immune function in social behaviors and uncovering the mechanisms underlying immune actions on behavior.