Female mice are protected from space radiation-induced maladaptive responses

Abstract

Interplanetary exploration will be humankind’s most ambitious expedition and the journey required to do so, is as intimidating as it is intrepid. One major obstacle for successful deep space travel is the possible negative effects of galactic cosmic radiation (GCR) exposure. Here, we investigate for the first time how combined GCR impacts long-term behavioral and cellular responses in male and female mice. We find that a single exposure to simulated GCR induces long-term cognitive and behavioral deficits only in the male cohorts. GCR exposed male animals have diminished social interaction, increased anxiety-like phenotype and impaired recognition memory. Remarkably, we find that the female cohorts did not display any cognitive or behavioral deficits after GCR exposure. Mechanistically, the maladaptive behavioral responses observed only in the male cohorts correspond with microglia activation and synaptic loss in the hippocampus, a brain region involved in the cognitive domains reported here. Furthermore, we measured reductions in AMPA expressing synaptic terminals in the hippocampus. No changes in any of the molecular markers measured here are observed in the females. Taken together these findings suggest that GCR exposure can regulate microglia activity and alter synaptic architecture, which in turn leads to a range of cognitive alterations in a sex dependent manner. These results identify sex-dependent differences in behavioral and cognitive domains revealing promising cellular and molecular intervention targets to reduce GCR-induced chronic cognitive deficits thereby boosting chances of success for humans in deep space missions such as the upcoming Mars voyage.

1. Introduction

In high-stress work environments, it is critical that individuals are able to exhibit peak cognitive performance to achieve success and eliminate human errors. Astronauts on deep space missions represent a particular population in which highly functioning cognitive domains are extremely important. To ensure safety and mission success during interplanetary travel, it is imperative to understand how stressors such as radiation exposure, sleep deprivation and microgravity can impact cognitive capabilities (Cucinotta et al., 2008). On Earth and in the International Space Station, humans are protected from radiation exposure by the Earth’s magnetic field, but during extended deep space travel, astronauts are exposed to galactic cosmic rays (GCR). GCR are composed of protons and highly-charged nuclei (Cucinotta et al., 1998; Zeitlin, 2013) that can penetrate both the hull of space-craft as well as the human body (Cucinotta et al., 2008; Nelson, 2016; Cucinotta, 2014; Norbury, 2016). Initial studies have focused on understanding the physics of GCR components, leaving unanswered questions about how exposure can impact biological function and more specifically neuronal fitness.

Recently, our group and others have explored how GCR exposure influences cognitive and behavioral functions, and have found that exposure to individual GCR components such as protons, iron, silicon, oxygen, helium or titanium induce acute and persistent cognitive decline in rodents. The maladaptive cognitive changes are characterized by alterations in learning and memory, fear/startle responses, anxiety phenotypes and social behaviors with responses varying depending on both the dose and type of ion exposure (Krukowski et al., 2018; Raber, 2013; Raber, 2016; Parihar, 2015; Parihar, 2016; Impey, 2016; Britten, 2017; Britten, 2016). Cognitive changes have been associated with alterations at the synaptic level as well as neuroinflammatory responses (Raber, 2016; Parihar, 2015; Parihar, 2016) providing initial cellular understanding for GCR-induced changes.

These previous studies have almost exclusively focused on individual particle exposure. While mechanistically it is important to understand how each component of the GCR can act independently it may not accurately represent neuronal and behavioral changes that occur in deep space, an environment in which astronauts will encounter multiple ions at once or in succession. Furthermore, these reports have primarily used male animals, a potential confounder as more and more studies are reporting sex-dimorphic responses to cognitive alterations (reviewed in Grabowska, 2017; Levine et al., 2016). Forty percent of the most recent class of astronauts is female, therefore it is vital to understand how GCR exposure impacts both sexes.

It is calculated that an individual cell in an astronaut will be traversed by protons once every three days, helium nuclei once every three weeks and Z greater than 2 nuclei once every three months (Nelson, 2016; Cucinotta, 2014; Raber, 2016; Cucinotta et al., 2011; de Wet and Townsend, 2017). Thus, we took the initial exploratory step to understand the estimated compounding effects of multiple ion radiation astronauts will face during deep space travel. We first investigated the effects of mixed ion (proton, helium and oxygen) exposure as a surrogate for true space exposure in different behavioral and cognitive domains in both male and female adult mice. Next, we explored how the behavioral outcomes relate to long-term neuroinflammatory responses and synaptic composition. Our data demonstrate that only the male cohorts were susceptible to maladaptive behavioral responses that were paralleled by changes in inflammation and synaptic composition. These data offer novel insight into behavioral, cognitive, cellular and molecular changes that can alter astronaut functional capabilities and thereby impact mission success.

2. Results

2.1. GCR exposure alters health state in rodents

Twenty-week old male and female C57Bl6/J mice (from Jackson Laboratory) were delivered to Brookhaven National Laboratories for Galactic Cosmic Radiation exposure (GCR- 0 cGy, 15 cGy, 50 cGy). Following GCR exposure, animals were transported to UCSF for behavioral and cellular assessments, Fig. 1. The health status report for all the animals are summarized in Table 1. We measured wounding in all of the male cohorts (0, 15 and 50 cGy), with animal death in the 15 and 50 cGy groups throughout the course of experimentation. No deaths or observable health alterations were measured in any of the female cohorts (0, 15 cGy or 50 cGy). Over the testing period all groups gained weight, with no differences between sexes or treatment cohorts, Table 2.

Fig. 1.

Experimental Design. Animals were exposed to mixed ion radiation, galactic cosmic ray (proton, helium, oxygen) at two doses (15 and 50 cGy) on day 0. Behavioral analysis are shown relative to radiation exposure. EPM = Elevated Plus Maze; Social = three-chamber social task; OF = Open Field; NOR = Novel Object Recognition.

Table 1.

Animal Health Assessment. Measurable wounds were observed in all of the male cohorts over the testing period. Animal number and types of wounds summarized here. Animal loss was measured over the testing cohorts, totals summarized here.

Sex	Treatment group	Lesion					Aggressors present	Deaths
		Back	Genitals	Tail	Other	Total Lesions
Male	50 cGy	3	2	0	0	5	1	1
	15 cGy	0	3	4	2	9	3	4
	0 cGy	8	2	0	0	10	3	0
Female	50 cGy	0	0	0	0	0	0	0
	15 cGy	0	0	0	0	0	0	0
	0 cGy	0	0	0	0	0	0	0

Table 2.

Weight Measurement. Animal weights were tracked by cage averages over the testing period. All cohorts displayed weight gain over time.

Sex	Treatment Group	Weight Gain (g)
Male	50 cGy	1.2 ± 0.4
	15 cGy	2.2 ± 0.8
	0 cGy	0.7 ± 0.4
Female	50 cGy	2.2 ± 0.2
	15 cGy	2.1 ± 0.3
	0 cGy	1.8 ± 0.3

First, to accurately measure GCR-related changes on behavior and not inherent sex differences and/or an inability to complete the tools, we investigated performance differences in the unexposed male and female cohorts (0 cGy). For social behaviors and recognition memory tests, both male and female 0 cGy groups were able to successfully identify the novel or social stimuli (data in 0 cGy groups for Figs. 2–6. To measure anxiety-like behavior we used two tests: elevated plus maze (EPM) and open field (OF); both male and female mice moved throughout the testing arenas (plus maze and square box). Differences in exploration times in the EPM and OF between male and females are summarized in Supplemental Fig. 1. In the EPM male mice spent significantly less time in the open arms when compared with their female counterparts. While this could be indicative of an anxiety-like baseline phenotype in the male cohorts; in the open field, male mice spent more time in the center when compared to the female counterparts which would be indicative of less anxiety-like phenotype. To account for these exploration differences all behavioral analysis for GCR-exposed (15 and 50 cGy) groups are compared to the sex-matched 0 cGy group.

Fig. 2.

Anxiety-like behavior at 45 days post GCR exposure. Anxiety-like behavior was measured by time spent in the open arms (including center space) in the elevated plus maze. No differences in exploration times were observed in male (A) or female (B) cohorts. Exploration times standardized to sex-matched 0 cGy group. A one-way ANOVA did not reveal any significant differences between groups. Individual animal scores represented in dots, lines depict group mean and SEM. Males n = 11–13; Females n = 13–15.

Fig. 6.

Radiation-induced recognition memory impairments in male cohorts only. Animals were exposed to GCR radiation (0, 15, 50 cGy). Beginning 80 days later, animals were tested for memory deficits. Memory deficits were measured by novel object recognition. Animals were exposed to two identical objects, 24 hrs later the animals are exposed to one familiar object and one novel object. Memory deficits are calculated by a deficit in distinguishing the new (novel) object. Nv = novel. Fm = Familiar. (A) Male cohorts exposed to either 15 or 50 cGy of GCR were unable to distinguish the objects, denoting memory impairment. Two-way ANOVA found significant differences in object distinction (p = 0.001). Bonferroni post-hoc analysis. (B) All female cohorts distinguished the novel from familiar objects. Two-way ANOVA found significant differences in object distinction (p = 0.0001). Bonferroni post-hoc analysis. (C) Males- Total exploration time with both objects is depicted per group. (D) Females- Total exploration time with both objects is depicted per group. A significant decrease in interaction time in the 50 cGy group was measured when compared to the 0 cGy cohort (F = 4.18, p = 0.02). Bonferroni post-doc analysis denotation shown = *p < 0.05, **p < 0.01. ***p < 0.001 Individual animal scores represented in dots, lines depict group mean and SEM. Males n = 14–15; Females n = 15–18.

2.2. Anxiety-like behavior 45 days after GCR exposure

We initially explored anxiety-like behaviors at 45 days post GCR exposure using the EPM (Lister, 1987). In this tool, animals are individually allowed to freely explore a plus maze consisting of two dark, enclosed arms and two open, brightly-illuminated arms. Decreased time in the open arms indicates anxiety-like behaviors. No anxiety-like phenotypes were measured in any groups at this time point, Fig. 2A, B.

2.3. GCR exposure impairs social behaviors in male but not female animals

2.3.1. Social Behaviors:

To investigate if GCR exposure alters social behaviors we utilized the three-chamber social recognition tool (Yang et al., 2011; Semple et al., 2012) at 46–60 days post GCR exposure. This task is divided into two testing phases that are run consecutively: (1) sociability and (2) social memory.

2.3.2. Sociability:

In the sociability phase, an animal’s innate preference for social interaction is calculated by the time engaging with a novel mouse (non-aggressive, age-sex matched) over an empty cage. The male cohorts exposed to 50 cGy of GCR displayed less preference for a novel mouse when compared to 0 cGy, Fig. 3A. This significant reduction in interaction with the novel mouse is indicative of sociability impairments. Mice exposed to 15 cGy of GCR displayed reduced interaction time with the novel mouse when compared to the 0 cGy group although statistical significance was not reached (p = 0.1). The impairments measured in sociability were not due to differences in total interaction time with the mouse and the object Fig. 3C.

Fig. 3.

Measurement of sociability after GCR exposure. Beginning at 45 days post exposure (GCR – 0, 15, 50 cGy) animals were tested for sociability deficits by the three-chamber social approach task. In the sociability part of this task the test animal is exposed to a mouse enclosed in a cage (non-aggressive, age-sex matched) or an empty cage for ten minutes. Sociability is measured by a preference to spend time with the mouse over the empty cage standardized to sex-matched 0 cGy group in the first 5 min of testing. (A) Male mice exposed to 50 cGy of GCR had significant impairments in sociability when compared to the 0 cGy group (F = 3.96; p = 0.02). (B) GCR exposure did not impact sociability in the female cohorts. One-way ANOVA did not reveal any significant differences between groups. Interaction times standardized to sex-matched 0 cGy group. (C) Males- Total interaction time with both the mouse and the empty cage. One-way ANOVA did not reveal any significant differences between groups. (D) Females- Total interaction time with both the mouse and the empty cage. One-way ANOVA did not reveal any significant differences between groups. Individual animal scores represented in dots, lines depict group mean and SEM. *p < 0.05. Males n = 13–16; Females n = 15–17.

We next investigated if the sociability deficits observed in the males could be due to individual housing conditions. As noted in Table 1, we measured a significant amount of wounding in the male cages. When an aggressor was identified that animal was individually housed for the remainder of the experimentation. To investigate if the number of cage-mates impacted sociability readouts, the sociability performance of each mouse was plotted against the number of home cage-mates. No relationship was observed (p < 0.64) demonstrating that variable housing conditions were not responsible for the described deficits in sociability, Supplemental Fig. 2A.

No impairments in sociability were measured in female GCR (15 cGy and 50 cGy) exposed animals Fig. 3B. Additionally, there were no differences in interaction time (Fig. 3D) or correlations (p < 0.28) between housing conditions (Supplemental Fig. 2B).

2.3.3. Social Memory:

Immediately following the sociability phase, animals were tested for social memory (the animal’s inherent tendency for novelty). The test animal is exposed to the familiar mouse, previously encountered in the sociability phase, and a novel mouse (non-aggressive, age-sex matched). The male mice exposed to 50 cGy of GCR showed significant reduction in the interaction time with the novel mouse compared to the 15 cGy group, Fig. 4A. These deficits were not due to differences in total exploration time, Fig. 4C. In contrast, no impairments in social memory were measured in female cohorts exposed to GCR, Fig. 4B and nodifferences were measured in total interaction time, Fig. 4D.

Fig. 4.

Measurement of social memory after GCR exposure. Beginning at 45 days post exposure (GCR - 0, 15, 50 cGy) animals were tested for social memory deficits by the three-chamber social approach task. In the social memory part of this task the test animal is exposed to a familiar mouse enclosed in a cage (from sociability part) or a novel mouse enclosed in a cage (non-aggressive, age-sex matched) for five minutes. Social memory is measured by a preference to spend time with the novel mouse over the familiar mouse standardized to sex-matched 0 cGy group. (A) Male mice exposed to 50 cGy of GCR had significant impairments in social memory when compared to the 15 cGy group (F = 3.99; p = 0.02). (B) GCR exposure did not impact sociability in the female cohorts. One-way ANOVA did not reveal any significant differences between groups. Interaction times standardized to sex-matched 0 cGy group. (C) Males- Total interaction time with both the novel and familiar mouse. One-way ANOVA did not reveal any significant differences between groups. (D) Females- Total interaction time with both the novel and familiar mouse. One-way ANOVA did not reveal any significant differences between groups. Individual animal scores represented in dots, lines depict group mean and SEM. *p < 0.05. Males n = 14–15; Females n = 16–17.

Collectively, these data demonstrate that male animals exposed to 50 cGy GCR develop social deficits. In contrast, no GCR-induced social impairments were measured in the female cohorts at the time points measured here.

2.4. GCR exposure induces anxiety-like behavior and recognition memory deficits in male mice

2.4.1. Anxiety-like behavior:

We investigated long-term GCR-induced deficits in anxiety-like behavior with the open field test beginning at 80 days post GCR exposure. In the open field task, mice freely explore the testing arena for 10 min (used the following days for Novel Object Recognition). Anxious animals will avoid the open center of the arena, preferring to remain close to the walls of the arena. A significant decrease in the time spent in the center of the arena was measured in the 50 cGy compared to the 0 cGy male group indicative of anxiety-like behavior, Fig. 5A. No anxiety-like phenotypes were measured in either of the GCR female cohorts (15 or 50 cGy), Fig. 5B.

Fig. 5.

High doses of GCR exposure induce anxiety-like behavior in male only. Beginning at 80 days post exposure, anxiety-like behavior was measured in the open field task. Animals are individually allowed to explore of 30 × 30 cm arena and time spent in the center (17 × 17 cm) was calculated. Decreased time in the center is indicative of anxiety-like behavior. (A) Male mice exposed to 50 cGy of GCR spent significantly less time in the center of the field when compared to 0c Gy groups (F = 4.05; p = 0.02). (B) No differences were measured in center time between the female cohorts. Exploration times standardized to sex-matched 0 cGy group. Individual animal scores represented in dots, lines depict group mean and SEM. *p < 0.05. Males n = 14–15; Females n = 16–18.

2.5. Recognition memory deficits:

To test for hippocampal-dependent recognition memory we utilized the Novel Object Recognition (NOR) task (Mumby et al., 2002) in which animals were exposed to two identical objects; twenty-four hours later one of the familiar objects from the previous day was replaced with a novel object. The animal’s preference for the novel object serves as a measure of recognition memory. While, the male control (0 cGy) mice were able to distinguish the objects, spending the majority of the time with the novel object, both the 15 cGy (p < 0.18) and 50 cGy (p < 0.39) GCR groups spent nearly equal time with both objects denoting a significant impairment in recognition memory, Fig. 6A. No differences in total interaction time were measured between the groups suggesting that memory impairments were not due to decreased interaction or motivation, Fig. 6C.

All the female cohorts (0, 15 and 50 cGy) spent the majority of the time with the novel object demonstrating the ability to distinguish the objects. These results prove that exposure to GCR did not affect recognition memory in female mice at 80 + days post exposure, Fig. 6B. We did measure a reduction in the total exploration time in the 50 cGy female cohort when compared with the 0 cGy group; however, this reduction in exploration did not impair performance in the task, Fig. 6D.

Overall, the behavioral and cognitive testing identified GCR-induced chronic deficits in social behaviors, anxiety-like phenotypes and recognition memory only in male cohorts. Furthermore, these deficits were primarily confined to the highest dose of GCR exposure (50 cGy). As a result, to identify cellular and/or molecular mechanisms of behavioral correlates from here on we will focus on comparing the 0 cGy and 50 cGy groups.

2.6. GCR induced changes in microglia phenotypes

We next investigated if long-term alterations in microglia could be associated with the GCR-induced behavioral deficits. We focused on microglia responses in the hippocampus as this brain region is critical for recognition and social memory (Mumby et al., 2002; Kogan et al., 2000; Alexander, 2016; Hitti and Siegelbaum, 2014; Okuyama et al., 2016). We identified significant increases in microglia staining (percent area as measured by Iba-1) in the 50 cGy male cohort when compared to the 0 cGy group at 100+ days post GCR exposure, Fig. 7A–E. These changes were observed throughout the dorsal hippocampus (DG, CA1, CA2 and CA3), with alterations most pronounced in the CA2/3 which is one region responsible for forming social memories (Okuyama et al., 2016; Raam et al., 2017). Strikingly, no change in microglia patterns were measured in the female cohorts (0 and 50 cGy), Fig. 7F–J.

Fig. 7.

GCR induced changes in microglia phenotypes in male cohorts only. Microglia levels were measured by Iba-1 staining in the dorsal hippocampus. Representative images from MALE cohorts- 0 cGy DG (A), 50 cGy DG (B), 0 cGy CA2/3 (C) 50 cGy CA2/3 (D). (E) Males exposed to 50 cGy of triple ion had increased Iba-1 expression in the DG, CA1 and CA2/3. Two-way ANOVA revealed a significant region effect (p < 0.01), group effect (p < 0.001) and interaction (p < 0.01). Bonferroni post-hoc analysis reveals differences between groups. Representative images from FEMALE cohorts- 0 cGy DG (F), 50 cGy DG (G), 0 cGy CA2/3 (H) 50 cGy CA2/3 (I). (J) No differences in Iba-1 expression was measured. Two-way ANOVA revealed a significant region effect (p < 0.05) and interaction (p < 0.05), however Bonferroni post-hoc analysis did not reveal any differences between the groups. No reactivity was observed in secondary alone. 200x magnification. **p < 0.01. ***p < 0.001 Individual animal scores represented in dots, lines depict group mean and SEM. Scale bar = 100 μm denoted by white line in (A). Males n = 3–4; Females n = 4.

2.7. Molecular signature of GCR induced changes in the prefrontal cortex

While recognition and social memory functions are primarily hippocampal dependent, sociability has been associated with the prefrontal cortex (PFC) circuits (Desbonnet et al., 2014; Makinodan, 2017). Given that the PFC has been previously shown to be chronically impacted by GCR exposure (Parihar, 2015; Parihar, 2016) we investigated if chronic neuroinflammatory responses were measured in this brain region. Interestingly, we found no differences in inflammatory cytokine or microglia gene expression profiles between male 0 cGy and 50 cGy cohorts chronically (100+ day) after exposure, Supplemental Fig. 3A–D, G. Similarly, no differences in gene expression levels were observed in the female cohorts, Supplemental Fig. 4A–D, G.

Changes in myelin composition in the PFC can regulate social behaviors (Desbonnet et al., 2014; Makinodan et al., 2012); therefore, we tested if GCR exposure chronically altered myelin expression levels in this region. We did not measure any changes in myelin expression levels following GCR exposure in either male or female cohorts, Supplemental Figs. 3E, F and 4E, F.

2.8. Alterations in synapse composition after GCR exposure

Microglia can regulate neuronal function through direct interaction with synapses (Weinhard, 2018; Shi, 2015; Tremblay et al., 2010); therefore, we next investigated if the microglia changes observed chronically after GCR exposure would be also associated with synapse composition. To assess synaptic composition, we used flow synaptocytometry, a newly developed technique that can individually label and quantify synapses from isolated brain regions (Snigdha, 2016; Prieto, 2015). We measured a significant reduction in total hippocampal synapses in the 50 cGy male cohort when compared to the 0 cGy group, Fig. 8A–C. Notably, no differences in total synapses were found when comparing the female cohorts, Fig. 9A. Additionally, we did not find any differences in total pre (synapsin-1) or post (PSD-95) synaptic components as measured by flow synaptocytometry, (Figs. 8D, E and 9B, C) or Western blot analysis, Supplemental Fig. 5.

Fig. 8.

Alterations in synapse composition after GCR exposure in the male hippocampi. Hippocampi were collected and synaptosomes were isolated by sucrose gradient. (A, B) Representative plots used to determine synaptosome levels first by size calibration beads, then co-expression of pre and post synaptic markers. Pre-synaptic marker-Synapsin-1 and post synaptic markers- PSD-95 and GluR1. (C) Significant decreases in total synaptosome numbers in the 50 cGy group when compared to 0 cGy group. Student t-test. (D, E) No differences were measured in total synapsin-1 expression (D) or PSD95 expression (E). (F,G) Decreases in surface GluR1 expression was measured in the 50 cGy group when compared to the 0 cGy group. (F) Representative histogram plot with 0 cGy group in grey and 50 cGy group in blue. (G) Quantification of total GluR1 expressing terminals. Events were collected on an LSRII and analyzed in FlowJo. **p < 0.01. ***p < 0.001 Individual animal scores represented in dots, lines depict group mean and SEM. n = 6–10.

Fig. 9.

No change in synapse levels measured in female cohorts. Hippocampi were collected and synaptosomes were isolated by sucrose gradient. (A) No differences were observed in total synaptosome (co-staining with synapsin-1 and PSD95) numbers in the female 50 cGy group when compared to female 0 cGy group. Student t-test. (B, C) No differences were measured in total synapsin-1 expression (B) or PSD95 expression (C). (D) No differences in surface GluR1 expression was measured in the female 50 cGy group when compared to the female 0 cGy group. Quantification of total GluR1 expressing terminals. Events were collected on an LSRII and analyzed in FlowJo. Individual animal scores represented in dots, lines depict group mean and SEM. n = 7–10.

Finally, we investigated the long-term effect of GCR exposure on AMPA surface receptor expression levels with particular focus on the GluR1 (GluR-A) known to be involved in hippocampal-dependent working memory (Sanderson, 2008). Increases in synaptic efficacy involve the rapid insertion of GluR1 in the postsynaptic membrane (Malinow and Malenka, 2002). Interestingly we measured a significant decrease in GluR1 expressing synapses in the 50 cGy exposed male cohort when compared with the 0 cGy group Fig. 8F, G) suggesting that AMPA expressing synapses may be particularly vulnerable to GCR exposure. No changes were measured in AMPA expressing synapses when comparing the female 0 cGy and 50 cGy group Fig. 9D.

3. Discussion

To ensure safety of personnel and mission success during deep space travel it is critical to understand the consequences of GCR upon various ethologically relevant cognitive and behavioral domains in both males and females. For the first time we report chronic behavioral and cognitive deficits following mixed ion GCR exposure, characterized by decreased social interaction, increased anxiety-like behavior and loss of recognition memory in male cohorts only. These behavioral responses corresponded with microglia activation and synapse loss in the hippocampus. Strikingly and surprisingly, female cohorts were not affected by GCR exposure in any of the parameters measured here. To our knowledge this is the first study to investigate behavioral and cellular responses following triple ion exposure, a condition that mimics deep space travel.

Recent studies have begun to characterize a number of maladaptive behavioral responses associated with individual ion exposures in rodents. Notably, here we uncovered a multi-domain effect in male cohorts following GCR exposure, measuring long-term deficits in sociability, social memory, anxiety-like behavior and recognition memory. These results demonstrate that combined ion exposure comparable to what is anticipated during deep space missions can severely influence multiple aspects of cognitive and emotional functions in males. Female cohorts exposed to GCR parameters identical to the male displayed no discernable impairments. These effects were also mimicked in the qualitative health assessments of the cohorts. Overall, these results highlight the potential challenges faced during deep space travel as well as sex-dependent differences in cognitive resistance.

In two recent reports, our group (Krukowski et al., 2018) and one other (Mange et al., 2018) have identified long-term deficits in social memory following oxygen exposure in either male mice or male rats, respectively. Here we once again utilize the three-chamber task (Yang et al., 2011; Semple et al., 2012; Chou et al., 2016) where we observe a mouse’s natural propensity for social interaction during the sociability phase as well as social novelty during the social memory phase. For the first time we observed social deficits in male cohorts exposed to 50 cGy of mixed ions. Strikingly, no deficits were observed in the female cohorts. These observed deficits in social functions could be due to a lack of motivation, anxiety and/or diminished response to novelty (Moy, 2004; Sobota et al., 2015; Allsop et al., 2014). Given the other behavioral impairments measured in the 50 cGy male cohort it is likely a combination of multiple factors. A deeper understanding of these social deficiencies is especially important for astronauts as other parameters experienced during space flight such as isolation, confinement and zero gravity could further exacerbate these responses.

Anxiety, a state of worry, concern or apprehension, presents in rodents by an aversion to enter and/or remain in open, often bright areas (Lister, 1987; Chou et al., 2016; Robinson et al., 2018; Wolfer, 2004). Here we observed anxiety-like phenotypes manifesting in the male 50 cGy group at 80 days post GCR exposure as measured by decreased time spent in the center of an open field arena. A recent report found anxiety-like behavior in male mice at one-year post helium exposure alone (Parihar, 2018) suggesting that deficits measured here may be persistent after exposure. Anxiety-like behavior was not observed in the female cohorts; unfortunately, Parihar et al only investigated anxiety-like behaviors in male rodents at one year post helium exposure so at this time it is not clear if females will display alterations at longer time points. Interestingly, we do not observe anxiety deficits early after radiation exposure (45 days). It is possible that anxiety-like phenotypes only present at later time points or that such phenotypes are not observed using the EPM.

Recognition memory is defined as the ability to identify a previously encountered object and/or situation as familiar (Squire et al., 2007). The NOR is a highly used and highly reproducible (Krukowski et al., 2018; Parihar, 2015; Parihar, 2016; Impey, 2017; Tseng, 2014; Cherry, 2012; Parihar, 2015) measure of recognition memory in rodents (also known as the “visual paired-comparison task” in studies with humans and monkeys) and relies on hippocampal function (Broadbent et al., 2010). Here we used the NOR task to investigate GCR-induced recognition memory deficits at 80+ days following triple ion exposure. We find that male cohorts exposed to either 15 or 50 cGy of combined proton, oxygen and helium charged particles display impairments in recognition memory when compared with their 0 cGy male counterparts. These cognitive deficits are in line with previous reports in male rodents that have identified alterations in recognition memory following individual ion exposures with iron (Impey, 2016; Britten, 2016), helium (Krukowski et al., 2018; Rabin et al., 2015), oxygen, titanium or protons alone (Parihar, 2015; Parihar, 2016; Britten, 2017). This highlights the unexpected observation that female GCR exposed mice (15 and 50 cGy) display no observable deficits in recognition memory. It is of note, that we did observe differences in total exploration time of both objects in the 50 cGy female cohort when compared with the 0 cGy female cohort. However, despite this decrease in total exploration the irradiated females were still able to identify the novel object.

The hippocampus is one of the brain regions required for memory functions (Mumby et al., 2002), and recent findings show that it is also required for regulating social functions (Kogan et al., 2000; Alexander, 2016; Hitti and Siegelbaum, 2014; Okuyama et al., 2016). Given that the hippocampus has been shown to be particularly vulnerable to radiation insult (Feng, 2016; Chiang et al., 1993; Monje, 2008; Mizumatsu, 2003), we investigated cellular alterations that could underlie the GCR-induced behavioral deficits observed in the male cohorts. Extensive microgliosis was measured throughout the dorsal hippocampus in the 50 cGy male when compared to the 0 cGy male cohort, while no changes in microglia were measured in the female GCR exposed group. Importantly, this microglia phenotype was measured 100+ day after ion exposure suggesting a persistent alteration in this cell type. Previous reports have identified microglia activation following individual ion exposure both in the hippocampus (Raber, 2016; Rola, 2008; Raber, 2014) and prefrontal cortex (Parihar, 2015; Parihar, 2016). Our results, taken together with the previous reports in single ion exposure suggest that microglia activation could be the root cause for the behavioral deficits observed at chronic time points after GCR exposure and therefore represents a potential therapeutic target. In a recent report, we found that temporary microglia depletion (via CSF1R inhibition) acutely after single ion GCR exposure prevented deficits in recognition memory in male mice by ‘resetting’ the microglia to a less inflammatory phenotype. These recent findings further highlights the importance of microglia in regulating GCR-induced behavioral changes (Krukowski et al., 2018).

It is known that microglia regulate neuroinflammatory processes via protein release (cytokines and chemokines) as well as direct cell-cell contact (Shi, 2015; Tremblay et al., 2010) including microglia-mediated removal of inactive or defective synapses (Stevens, 2007; Weinhard, 2018; Shi, 2015; Tremblay et al., 2010). Here we found that elevated microgliosis in the GCR exposed male corresponded with significant reductions in total synapses and AMPA receptor (GluR1) expressing post-synaptic terminals. No changes in synaptic number or AMPA receptors were measured in the female cohorts. Synapse loss was quantified using flow synaptocytometry, a new technique that allows for quantification of intact synapses as well as synapse protein levels (Snigdha, 2016; Prieto, 2015). Parihar et al. measured decreased dendritic spine number and increased PSD-95 levels in the PFC following either oxygen or titanium exposure individually. Interestingly, here we identified a decrease in synapse number in the hippocampus, however we do not measure differences in total PSD-95 levels. These differences could be due to the GCR exposure paradigm with multiple ions and also the time after exposure. All the previous studies have used a single ion exposure and this is the first study assessing the combination of three. Additionally, it is possible that GCR exposure differentially impacts the PFC versus hippocampus. Nevertheless, it is clear that GCR exposure can chronically influence synapse number and these changes likely underlie the chronic maladaptive behavioral responses.

AMPA receptor trafficking is critical for regulating various forms of synaptic plasticity and hippocampal-dependent spatial working memory. In the adult hippocampus, AMPA receptors primarily consist of GluR1 and GluR2 subunits (Wenthold et al., 1996); increases in synaptic efficacy involve the rapid insertion of GluR1 in the postsynaptic membrane (Malinow and Malenka, 2002). Changes in GluR1 surface expression levels have been linked to a number of different neurological disorders such as ischemia, epilepsy and schizophrenia (Dos-Anjos, 2009; Montori, 2010; Tang, 2005; Schmitt, 2005; Matsuo, 2009). Furthermore, links between GluR1 expression levels and spatial memory have been identified, where genetic manipulations of GluR1 restore spatial memory deficits (Schmitt, 2005; Matsuo, 2009). We previously reported that proton exposure alters the composition in AMPA receptors in the hippocampus, along with significant changes in hippocampal-dependent recognition memory (Parihar, 2015). Thus, we hypothesize that the long-term GCR-induced spatial memory deficits observed only in the male cohorts are at least in part due to reductions in hippocampal dependent GluR1 surface expression.

Identification of sex-dimorphic differences in behavioral responses are becoming more prevalent in scientific research. Unfortunately, the published studies assessing differences between males and females are varied and are largely influenced by species, age, strain, genetic modification and investigators differences (Jonasson, 2005; Galea et al., 1994). Some reports suggest that males perform higher in spatial learning paradigms (water mazes; object in place) (Jonasson, 2005) while females perform similarly (Benice et al., 2006; Bowman et al., 2009; Salas-Ramirez et al., 2010) or outperform (Saucier et al., 2007; Sutcliffe et al., 2007; Ghi et al., 1999) their male counterparts in recognition memory (NOR). Many of these reports use naïve animals or individuals; thus, responses may differ depending on additional external stimuli. Few studies have looked at the sex-specific effects of exposure to individual ions using different genetically modified mice models (human apoE3,4 and APP/PS1) (Cherry, 2012; Villasana et al., 2011; Villasana et al., 2006) or electric shock stimuli (Raber et al., 2016) but no studies have comprehensively addressed cognitive functions in male and female wild type mice even with a single exposure to protons, helium or oxygen. Here we demonstrate that combined GCR-exposed males are susceptible to a wide range of changes in ethologically relevant behaviors, notably, these changes are paralleled by significant synaptic changes in the hippocampus and increases in microgliosis, while female animals display no measurable impairments and no cellular changes. We cannot rule out the possibility that we did not analyze the appropriate behavioral paradigm or window of time to measure GCR-induced changes in females. Additionally, estrous cycle of the female animals was not tracked, however, a recent report found that female microglia gene expression patterns were unaffected by estrogen level (Villa, 2018). Furthermore, since behaviors were measured over a period of 35 days we think it is unlikely that estrous cycle could have been driving our findings. Nevertheless, our work adds to the growing amount of literature demonstrating sex-dimorphisms in cognition.

Sex-dependent microglia differences have been observed in naïve, adult animals (Villa, 2018; Mouton, 2002; Crain et al., 2013) within the hippocampus; female animals had increased resting microglia number. However, this increased number does not always correspond to increased inflammatory (cytokine) responses. Despite increased microglia numbers, female animals in trauma models, display a dampened inflammatory response when compared to their male counterparts. Following stroke (Villa, 2018; Bodhankar, 2015); LPS stimulation (Loram, 2012) or physical trauma (Acaz-Fonseca et al., 2015) microglia isolated from female animals displayed lower inflammatory cytokine profiles. Excitingly, a recent report investigating microglia from adult male and female mice found sex-dependent differential gene expression patterns that corresponded with a neuroprotective microglia state in female animals (Villa, 2018). Here we find that combined GCR exposure induces a robust microglia response in the male cohorts and no responses in the female cohorts, which is in line with previous reports showing that hippocampal microglia are less reactive in females possibly even displaying a neuroprotective state. Interestingly, we did not find any differences in cytokine profiles in GCR-exposed animals when compared to their sex-matched counterparts. We propose three possible explanations for this (1) the cytokine profiles is from total PFC, thus it is possible that isolated microglia from the hippocampus may have differential gene expression profiles (2) it is possible that gene expression profiles do not reflect protein levels or (3) microglia activation state may reflect changes in phagocytosis rather than on cytokine expression. In a recent report we find alterations in microglia phagocytic parameters chronically after helium exposure in male mice (Krukowski et al., 2018). Thus, these findings suggest that sex-dependent microglia differences may not be limited to cytokine production but can also include a phagocytic activity a microglia parameter that is critical for regulation of synapse.

In summary, the current work provides the first line of evidence demonstrating that a single exposure of combined GCR can significantly impact a variety of behavioral, cognitive and cellular domains in a sex-dependent fashion. These results have important implications as NASA prepares astronauts for deep space travel and we prepare for the greatest adventure of humankind.

4. Methods

4.1. Animals

All experiments were conducted in accordance with National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee of University of California (San Francisco). All C57Bl6/J wildtype (WT) mice were purchased from Jackson Laboratory (Bar Harbor, ME) and shipped to Brookhaven National Laboratories for Galactic Cosmic Radiation (GCR) exposure (see specifics below). Mice were exposed to GCR at 20 weeks of age. After GCR exposure, animals were transported to UCSF for behavioral analysis. Mice were group housed in environmentally controlled conditions with reverse light cycle (12:12 h light:dark cycle at 21 ± 1 °C; ~50% humidity) and provided food and water ad libitum. In male cohorts, when aggressors were identified these animals were individually housed for the remainder of experiments. Proper enrichment was added to the individually housed animals. When wounding was observed, animals were treated topically with Chlorhexidine Flush 0.2% Solution (Cevo, Lenexa, KS). If wounds persisted animals were not included for behavioral analysis. All animal cages were housed in the same room and all female cohorts were group housed. Female mice were housed in groups of 2, 3, or 4. Male mice were housed in groups of 1, 2, 3, or 4. Estrous cycles were not tracked.

4.2. GCR calculation

At Brookhaven National Laboratory (BNL), Upton, New York where, mice were given 1 week of acclimation, they were then irradiated at the NASA Space Radiation Laboratory (NSRL, (Lowenstein and Rusek, 2007) with three charged particle species delivered in rapid succession to simulate the space radiation spectrum expected inside a deep space vehicle with self-shielding by the human body (Nelson, 2016; Norbury, 2016). These particles were, in order of delivery: 252 MeV/n protons (LET = 0.39 keV/μm), 249.3 MeV/n helium ions (LET = 1.57 keV/μm) and 594.4 MeV/n oxygen ions (LET = 16.5 keV/μm) in dose proportions of 60%: 20%: 20%, where Oxygen was used as a surrogate for particles of Z greater than 2 because of the technical limitations of the NSRL at the time of the experiment. Total doses of 15 or 50 cGy were delivered as part of experimental campaign NSRL 17A, experiment # N301.

The mice were housed in groups of 4 on a normal 12:12 light cycle at the BNL animal care facility, transported to the NSRL for irradiation several hours before exposure and returned to the animal care facility several hours after irradiation. In each case, 8 mice were loaded into 7.3 × 4.0 × 4.0 cm polystyrene restraint boxes with air holes and mounted on the beam line as stacks of restraint boxes on a foam base. The animals were oriented with the animals’ long axes parallel to the beam and were grouped to minimize the number of three-part exposures at each dose. The irradiation time was during the dark phase of the circadian cycle (2 to 6 AM). The beam spot size was 20 × 20 cm and animals were arranged in the center of the field to assure best uniformity. The NSRL physics support group has been upgrading control systems to enable rapid switching between particle beams which allowed us to irradiate cohorts of mice sequentially with three ions in 6 to 8 min.

The measured doses (mean and standard deviation) were: 30.01 ± 0.01 cGy for H; 10.00 ± 0.00 cGy for He; 10.01 ± 0.01 cGy for O and 50.02 ± 0.01 cGy total for the “50 cGy” cohort and 9.01 ± 0.00 cGy for H; 3.00 ± 0.00 cGy for He; 3.25 ± 0.50 cGy for O and 15.26 ± 0.50 cGy total for the “15 cGy” cohort. The overall dose rate was 6.16 cGy min for the “50 cGy” cohort and 2.54 cGy/min for the “15 cGy” cohort. The average delivery schedule (all exposures) was: 0.82 ± 0.28 min, H | 2.82 ± 0.82 min ion switch time |0.57 ± 0.24 min, He | 2.61 ± 0.54 min ion switch time |0.85 ± 0.56 min, O for a total of 7.06 ± 1.39 min. Finally, the proportions of dose for the three ions H, He, O were: 60.0%, 20.0% and 20.0% for the “50 cGy” cohort and 59.0%, 19.7% and 21.3% for the “15 cGy” cohort.

After irradiation the animals were returned to their home cages and returned to the animal care facility for shipment to University of California San Francisco by World Courier^™ within 5 days.

4.3. Behavioral analysis

For one week prior to behavioral analysis, animals were handled for habituation to investigators and room settings. Behaviors were performed in dark rooms during the animals wake cycle. Mixed sexes and treatment groups were analyzed each behavioral day. All behaviors were recorded using an overhead camera connected to a video tracking and analysis system (Ethovision XT 12.0, Noldus Information Technology). Ethovision software analysis was used for the majority of videos, however when tracking was not optimal videos were manually scored by an investigator blinded to groups. To ensure reliability between computer analysis and manual scoring, a subset of videos are scored by both and cross validated. All animal behaviors were performed by two female investigators blinded to GCR exposure group. Behavioral assessment was measured over a period of several weeks due to the large animal numbers necessary for this study. No differences between investigators, time of day or testing window were observed. Group outliers were determined (GraphPad Software Outlier Test-Grubb’s test) and excluded from analysis. At most a single animal was excluded from each experimental cohort.

4.4. Anxiety

Anxiety was first evaluated using the Elevated Plus Maze at 45 days post GCR exposure. The EPM consists of two exposed, open arms (35 cm) opposite each other and two enclosed arms (30.5 cm) also across from each other. The four arms are attached to a center platform (4.5 cm square) and the entire maze elevated 40 cm off the floor. Bright white lights are illuminated on both ends of the open arm (Lister, 1987; Chou et al., 2016). Mice were placed individually onto the center of the maze and allowed to explore the maze for 5 min and their activity was recorded. The maze was cleaned with 70% ethanol between animals. Anxiety-like behavior was measured by changes in time spent in the open arms + center.

Anxiety was later measured by the open field test at ~80 days post GCR exposure. For this task animals are allowed to freely explore an open field arena (30 cm × 30 cm) for 10 min under red-lighting. Activity is recorded and the arena is cleaned with 70% ethanol between animals. Anxiety-like behavior is calculated by alterations in time spent in the center (17 cm × 17 cm) of the arena.

4.5. Sociability and social memory.

At 45–60 days post GCR exposure a group of animals were tested for sociability and social memory behavior on the three-chamber social approach task as previously described (Yang et al., 2011; Semple et al., 2012; Chou et al., 2016). The three-chamber social approach task is performed in a dark room under red lighting. Animals were placed individually into the center of a three-chamber environment (72 cm × 50 cm; each chamber is ~24 cm × 50 cm) and their behavior was recorded during three consecutive phases: habituation, sociability, and social memory. During the habituation phase, mice were allowed to freely explore the testing chamber for 10 min. Immediately, following habituation, the sociability phase begins. In this phase the animals are exposed to a stranger (sex and aged-matched) mouse or an empty cage. The stranger mouse is enclosed in a cage identical to the empty one. The cages are 10 cm in diameter and place ~30 cm distance from one another. The mouse’s sociability is measured by their preference for interaction with the mouse rather than the empty cage. Interaction is calculated by the nose point of the mouse in close proximity (< 5 mm) from either cage. Placement of stranger mouse into the right or left chamber was alternated for each trial. Test mice were allowed to explore the stranger mouse and an empty cage for 10 min. Next is the social memory phase of the task. During this phase the animals were tested for their preference for a novel (sex and age-matched) mouse over a familiar mouse they encountered during the sociability phase. Test mice were allowed 5 min to explore the environment. The environment and cage objects were cleaned with 70% ethanol between test animals. Stranger/Novel mice consisted of two cages (4 mice/cage). Importantly these two cages were non-aggressive, sex and aged-matched not used in this study, however these two cages were not littermates. Furthermore, during testing the investigator exercises caution (changing gloves, cleaning surfaces etc.) in order to ensure odors between the two cages do not mix. All four mice from each cage are used throughout testing on all groups and aversion to pairs was noted. The preference ratio was calculated as the percent of time exploring the mouse divided by percent of time exploring the object (empty cage) for the sociability phase and as percent of time exploring the novel mouse divided by percent of time exploring the familiar mouse for the social memory phase. The first five minutes of each phase is used to calculate the preference ratio.

4.6. Novel object recognition

Hippocampal-dependent memory function was measured at 80+ days post exposure using a mouse novel object recognition assay (Mumby et al., 2002). This tool is frequently used to accurately reflect even mild deficits after radiation exposure (Parihar, 2015; Parihar, 2016; Impey, 2017; Tseng, 2014; Cherry, 2012; Parihar, 2015). The test environment consists of an open field arena (described above) under red lighting. Mice were allowed to explore the arena for two 10-minute periods for two consecutive days (habituation phase). On day three (training phase), two identical objects (red Lego^™ blocks) were secured to the floor in opposite corners of the arena using magnets and mice were allowed to explore the arena and objects for 5 min. 24 h later on day four (testing phase), one of the objects was replaced with a novel object (orange Lego^™ flower) of similar dimensions and texture. Mice were allowed to explore for 5 min. The objects and arena were cleaned with 70% ethanol between trials and animals. Trials were recorded and exploratory behavior was defined as time the animals spent directing its nose towards an object. Data is expressed as percent of time mice spent exploring each object. Mice that had < 5 s of exploration time during either training or testing were excluded from analysis. No side preferences were measured during the training phase for any cohort, data not shown.

4.7. Tissue collection

All mice were lethally overdosed using a mixture of ketamine (10 mg/ml) and xylaxine (1 mg/ml). Once animals were completely anesthetized, the chest cavity was opened and blood was obtained by cardiac puncture. Following cardiac puncture animals were perfused with 1X phosphate buffer solution, pH 7.4 (Gibco, Big Cabin OK, −70011–044) until the livers were clear (~1–2 min). Animal cohorts were split into groups for immunohistochemistry (n = 6/group) or frozen tissue analysis (n = 8–12/group).

For immunohistochemistry analysis, following PBS, left hemi-brains were fixed in ice-cold 4% paraformaldehyde, pH 7.5 (PFA, Sigma Aldrich, St. Louis, MO, 441244) for 24 hrs followed by sucrose (Fisher Science Education, Nazareth, PA, S25590A) protection (15% to 30%). Brains were embedded with 30% sucrose/ Optimal Cutting Temperature Compound (Tissue Tek, Radnor, PA, 4583) mixture on dry ice and stored at −80 °C. Brains were sectioned into 20 μm slices using a Leica cryostat (Leica Microsystems, Wetzlar, Germany) and mounted on slides (ThermoFisher Scientific, South San Francisco, CA). Slides were brought to room temperature (20 °C) prior to use. Slides were washed in tris buffered saline tween solution for 10 min. and twice with tris buffered saline (TBS) for 10 min. each. All slides were blocked in Block Reagent (Perkin Elmer, Waltham, MA, FP1020) for 30 min in the dark. Slides were stained with primary antibodies specific for Iba-1 (Rabbit, Wako, Richmond, VA 019-19741) overnight, washed three times in TBS, and stained for the secondary antibody, goat anti-rabbit Alexa-568 (Invitrogen, Carlsbad, CA, A-11011). Tissues were fixed using ProLong Gold (Invitrogen, Carlsbad, CA, P36930) and a standard slide cover sealed with nail polish. 3–4 images separated by 50–100 μm in the dorsal hippocampus were averaged per animal. Nine μm z-stack images were acquired on a Nikon High Speed Widefield Confocal microscope (Ti inverted fluorescence; CSU-W1) at the UCSF Nikon Imaging Center. 200x magnification.

For frozen tissue, following PBS, the whole brain was rapidly removed and the prefrontal cortex and hippocampus was dissected, snap frozen on dry ice and stored at −80 °C.

4.8. Flow synaptocytometry

Hippocampi were homogenized in 0.32 M sucrose in HEPES (Sigma Aldrich, St. Louis, MO H0887) in a glass Dounce homogenizer. Gross tissue fragments were removed in a by centrifugation (1200×g for 10 min), supernatant collected. Synaptosomes were isolated by further centrifugation (13,000×g for 20 min), supernatant was discarded and pellets containing synaptosomes were collected. Samples are kept cold (either on ice or in chilled centrifuges) for staining procedures. Synaptosome collection was standardized between samples will be by total protein concentrations (Pierce BCA Protein Assay, Thermo Scientific, Rockford, IL, 23225). Synaptosomes (50 ug/ml) were divided into 1.5 ml microcentrifuge tubes for staining. For extracelluar staining, synaptosomes were stained with primary antibodies (GluR1, Millipore Temecula CA, ABN241) for 30 min at 4 °C and agitated after 15 min. Samples were washed 2x with 700 ul of antibody staining buffer (PBS + 5% BSA) at 13,000×g for 5 min. at 4 °C. Secondary antibodies (Goat anti-rabbit 647, Invitrogen, Carlsbad, CA, A21244) were added for 30 min and agitated after 15 min at 4 °C in the dark. For intracellular staining, synaptosomes were first permeabilized by Cytofix/Cytoperm fixation solution (BD Biosciences, San Diego, CA, 554722) for 20 min on ice. Samples were washed 2x with 700 ul of perm/wash buffer (BD Biosciences, San Diego, CA, 554723) at 13,000×g for 5 min. at 4 °C. Primary antibodies (Synapsin-1 and PSD-95) were added for 30 min and agitated after 15 min. Samples were washed 2x with 700 ul of perm/ wash buffer at 13,000×g for 5 min. at 4 °C. Secondary antibodies (Goat anti-rabbit 405, Invitrogen, Carlsbad, CA A31556; Goat anti-mouse 488 Invitrogen, Carlsbad, CA, A11001; Goat anti-mouse 647 Invitrogen, Carlsbad, CA, A21236) were added for 30 min and agitated after 15 min at 4 °C in the dark. All reagents were made fresh for each staining day. Synaptosomes were determined by size-calibrated beads and co-staining with antibodies specific for pre (synapsin-1) and post-synaptic (PSD-95, GluR1) markers. No positive staining was observed in samples containing: (1) no synaptosomes (2) secondary antibodies alone. A subset of samples from each group were selected for flow synaptocytometry analysis. A maximum of four samples were analyzed per day. Each day at least one 0 cGy sample was analyzed and 50 cGy samples were standardized to the 0 cGy group. Samples were run in duplicate. Data were collected on an LSRII (BD) and analyzed with Flowjo^™ software (v10, Tree Star Inc.). 30,000 events were collected for total synaptosomes. 50,000 events were collected for GluR1 expressing terminals.

4.9. qPCR

Prefrontal cortex samples, of approximately the same size per animal, in TRI Reagent (Zymo Research, Irvine, CA, R2051) were hand homogenized using a plastic pestle and RNA was extracted using the Direct-Zol RNA MiniPrep Kit (Zymo Research, Irvine, CA, R2050-1-200) following the manufacturer’s instructions, briefly stated below. Cell debris was removed via centrifugation (1 min at 12,000g). Supernatant was combined with ethanol and passed through the provided column. Columns were washed using RNA Wash Buffer and incubated with DNase for 15 mins. Digestion buffer was washed through using RNA Prep and RNA Wash Buffers prior to RNA elution. RNA concentration was measured using a NanoDrop 2000c (ThermoFisher Scientific, South San Francisco, CA). cDNA was synthesized from 300 ng of RNA using a High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, 4368814) following the manufacturer’s suggested protocol. Target gene expression was measured in triplicate using Brilliant II SYRB^® Green QPCR Master Mix (Agilent, Santa Clara, CA, 600828) in a Stratagene Mx3005 Real Time System (Agilent, Santa Clara, CA) (Male Mice: CCL2, CD11b, CD206, TNFa, & GAPDH) or using iTaq Universal SYBR^® Green SuperMix (Bio-Rad, Hercules, CA 1725121) in a CFX96 Touch Real-Time PCR Detection System (Bio-Rad, Hercules, CA) (Male Mice: IL1β, MAG, & MBP; Female Mice: GAPDH, CCL2, CD11b, CD206, IL1β, MAG, MBP, & TNFα) following the manufactures guidelines. Relative gene expression was determined using the 2–ΔΔCt method and normalized using GAPDH. Stability of GAPDH was tested between groups and with multiple qPCR machines.

• CCL2: Fw 5′ GCTGACCCCAAGAAGGAATG-3′ Rev 5′ GTGCTTGAGGTGGTTGTGGA-3′

• CD11b: Fw 5′ CTGAGACTGGAGGCAACCAT-3′ Rev 5′ GATATCTCCTTCGCGCAGAC-3′

• CD206: Fw 5′ CCTCTGGTGAACGGAATGAT-3′ Rev 5′ CTTCCTTTGGTCAGCTTTGG-3′

• IL1β: Fw 5′ TGTAATGAAAGACGGCACACC-3′ Rev 5′ TCTTCTTTGGGTATTGCTTGG-3′

• TNFα: Fw 5′ TGCCTATGTCTCAGCCTCTTC-3′ Rev 5′ GAGGCCATTTGGGAACTTCT-3′

• MAG: Fw 5′CAGTTGCCAAGAGCCTGTACCT-3′ Rev 5′ TTCACTGTGGGCTTCCAAGGTG-3′

• MBP: Fw 5′ATTCACCGAGGAGAGGCTGGAA-3′ Rev 5′ TGTGTGCTTGGAGTCTGTCACC-3′

• GAPDH: Fw 5′ AAATGGTGAAGGTCGGTGTG-3′ Rev 5′ TGAAGGGGTCGTTGATGG-3′

4.10. Western blot analysis

Snap frozen hippocampal tissues were hand homogenized in Pierce RIPA buffer (Thermo Scientific, #89900) with cOmplete ULTRA tablets (Roche, #05892791001) and PhoSTOP (Roche, #04906837001) protease inhibitors for whole cell lysis. Debris was removed via centrifugation and the protein concentration of the remaining homogenate was measured using the Piece BCA Protein Assay (Thermo Scientific, #23225) following the manufacturer’s instructions. 20 ug of protein in Laemelli buffer (BioRad, #1610747) with beta mercaptoethanol (Sigma, M–3148) was boiled at 100 °C, loaded into a precast 4–15% Tris Glycine gel (BioRad, #5671084), and run using Tris/Glycine/SDS buffer (BioRad, 1610772) at 150 V. Protein was transferred onto a nitrocellulose membrane (BioRad, 1620168) using Tris/Glycine buffer (BioRad, 1610771) at 100 V. Membranes were blocked using Odyssey Blocking Buffer (PBS) (LI-COR, 92740000) and incubated for a minimum of 14 h with either synapsin 1 (1:2000 in PBS, Millipore, #AB1543) or PSD95 (1:1000 in PBS, Abcam, #ab13552). Blots were washed using TBS and incubated for 60 min in their host-specific secondary antibody (LI-COR, #92632210-Mouse, #92632211-Rabbit). No bands were visible when secondary antibodies alone were added. Blots were imaged using Odyssey (LI-COR) imaging system and quantified using the Image Studio software. All blots were normalized to GAPDH (Sigma, #G8795).

4.11. Data analysis

Results were analyzed using Prism software (v7.05, GraphPad; La Jolla, CA) and expressed as mean ± standard error of the mean (SEM). Statistical analyses were performed as listed below with p values of < 0.05 considered as significant.

Elevated Plus Maze:

One-way ANOVA (Fig. 2) or Student t-test (Supplemental Fig. 2). N: Male 0 cGy = 12, 15 cGy = 11, 50 cGy = 13; Female 0 cGy = 15, 15 cGy = 13; 50 cGy = 14.

Sociability and Social Memory:

ANOVA; Tukey post-hoc (Figs. 3 and 4); Spearman r correlation (Supplemental Fig. 2). Fig. 3: N: Male 0 cGy = 16, 15 cGy = 13, 50 cGy = 14; Female 0 cGy = 17, 15 cGy = 15; 50 cGy = 16. Fig. 4: N: Male 0 cGy = 15, 15 cGy = 14, 50 cGy = 14; Female 0 cGy = 17, 15 cGy = 16; 50 cGy = 16.

Open Field:

One-way ANOVA; Tukey post-hoc (Fig. 5) or Student t-test (Supplemental Fig. 2). N: Male 0 cGy = 15, 15 cGy = 14, 50 cGy = 15; Female 0 cGy = 18, 15 cGy = 16; 50 cGy = 16.

Novel Object Recognition:

Two-way repeated measure ANOVA, Bonferroni post-hoc (Fig. 6A,B) or One-way ANOVA; Tukey post-hoc (Fig. 6C, D). N: Male 0 cGy = 15, 15 cGy = 14, 50 cGy = 15; Female 0 cGy = 18, 15 cGy = 15; 50 cGy = 16.

Immunohistochemistry:

Two-way ANOVA, Bonferroni post-hoc (Fig. 7). N: Male 0 cGy = 3, 50 cGy = 3–4; Female 0 cGy = 4, 50 cGy = 4.

Flow Synaptocytometry:

Student t-test (Figs. 8 and 9). N: Male 0 cGy = 6, 50 cGy = 10; Female 0 cGy = 7, 50 cGy = 10.

qPCR:

Student t-test (Supplemental Figs. 3 and 4). N: Male 0 cGy = 9–10, 50 cGy = 8–9; Female 0 cGy = 11–12, 15 cGy = 15; 50 cGy = 9–10.

Western blot:

Student t-test (Supplemental Fig. 5). N: Male 0 cGy = 9, 50 cGy = 10.