This is the fifth year of this project. Over the past year, we have continued to analyze and integrate data from our studies. We have now completed analysis and prepared manuscripts for studies examining how SR, social isolation (SI), and SR + SI impact fear memory, the stress response and sleep in male rats. These data demonstrate that SR, SI, and their combination can alter multiple aspects of stress responses and fear memory that also can vary with putative stress resilience and vulnerability. In general, SR and SR + SI increased freezing in both Res and Vul rats and differentially impacted temperature by increasing stress-induced hyperthermia (SIH) in SR Res rats to parallel higher levels observed in CTRL-Vul and SR-Vul rats. SR and SI, alone and in combination, also altered sleep amounts and architecture with differential effects in Res and Vul rats.

Enhanced freezing has been considered to indicate improved fear memory and a beneficial effect of SR. However, this interpretation is problematic when enhanced freezing is considered in the context of the role fear learning and memory have in adaptive and pathological outcomes. Freezing in SR rats begins at a higher level and continues at a high level throughout shock training, thereby suggesting that their propensity to freeze is initially higher than in control rats. The continued high levels of freezing across days when footshock was no longer presented indicate significant impairment in fear extinction in SR rats. In this case, continued high levels of freezing could interfere with acquiring or exhibiting appropriate responses as the fearful/stressful situation changes. Thus, increased freezing after SR and SR + SI would likely indicate an impediment in adaptive learning, not an improvement. The impaired extinction after SR and SR + SI could also indicate a pathological change in the regulation of fear memory as the failure of fear extinction has been linked to failed coping and psychopathology.

In addition to overt behaviors, an initial stressor and related fearful memories can induce similar physiological responses including increased SIH, HPA activation, and corticosteroid release, as well as increases in respiration and heart rate. SIH has a time course that parallels that of HPA activation and persists as long as psychologically stressful situations last. In this study, we found significant SIH in response to the shock training (ST), context (CTX), and extinction (EXT) phases of conditioned fear (CF) with variations across treatment groups, phenotype, and time. Compared to CTRL and SI alone, SR and SR + SI produced consistent increases in SIH in response to ST, CTX, and EXT with differences across the Res and Vul phenotypes. In general, SR promoted a Vul-like SIH response in Res rats.

We determined the effects of SR on sleep in rats trained in CF and whether these varied with Res and Vul phenotypes using our established animal model of stress resilience and vulnerability. Supporting previous findings, our results showed subsequent REM responses directionally differ in Res and Vul animals following ST compared to baseline, and that phenotypes are maintained after exposure to SR. Interestingly, we also found that SR increased REM sleep and had a greater negative impact on fear memory processing, as indicated by greater freezing behavior throughout CF. By comparison, SI had greater effects on NREM sleep that were somewhat attenuated in SR + SI animals. These results indicate a broad impact of SR, SI and their interactions across multiple systems (i.e., fear memory, stress, and sleep regulation). The effects on sleep also parallel prior findings that SI beginning after SR can attenuate some SR effects on multiple behavioral and physiological systems.

With respect to the overall project, we have completed studies on SR alone, SI alone, and SI beginning after SR in both males and females. We also have made significant progress in analyzing the data from these studies that will be needed to conduct comparisons across sexes. We expect to work on these comparisons over the coming year. The past year also has been one of accomplishment and significant challenges. We established hindlimb unloading (HLU) in our lab but were unable to conduct planned studies of sleep due to the high mortality rate in animals that received both HLU and the implant surgery needed for recording. We also began a cross-species validation study in mice and encountered a similar obstacle with respect to surgical mortality in SR-treated mice. These are different species and different treatments but do indicate that HLU and SR can negatively impact the ability of animals to withstand surgical stress.

In summary, our data to date demonstrate both single and synergistic effects of SI and SR on several behaviors and sleep. They also are beginning to provide a framework for understanding how they can impact multiple, interacting systems (i.e., fear memory, stress, and sleep).

This is the fourth year of this project. We have now conducted studies in controls (males: n=20; females, n=16), social isolation (social isolation/SI; males: n=21; females: 16), SR (males: n=15; females: n=15), and SI + SR animals (males: n=15; females: n=15). Rats were divided into Res and Vul phenotypes based on rapid eye movement (REM) sleep responses to footshock stress. Some analyses and comparisons across groups are still ongoing but when completed will provide data for comparisons of males and females across inflight stressors. We will also be able to determine whether both males and females show Vul and Res phenotypes. We also are currently pilot testing hindlimb unloading (HLU) in females and will begin sleep and behavioral studies in the current project period.

In general, the data we have currently available for sex comparisons demonstrates that SI can produce significant alterations in sleep that differ in males in females. SI females showed significantly greater REM electroencephalogram (EEG) alpha power than control females in the light period. Elevated EEG spectral power in the high frequency range (alpha, sigma, and beta) is considered an indicator of cortical hyperarousal whereas low power in low-frequency bands (delta and theta) may indicate less sleep propensity and quality. Together, our data indicate that SI reduces sleep amount and quality, with differences in males and females that suggest more disrupted sleep in females. We will need to determine whether these differences hold when we have the full data set analyzed as well as the effects of SR and SR + SI on sleep across sexes.

We also found that SR exposure had a major negative impact on the integrity of brain macro and micro structures. The altered morphology induced by SR appeared to disrupt both vascular and lymphatic systems, observed by increased BBB permeability and LV enlargement and associated with a striking loss of supporting cells, including astrocytes and endothelial cells. Unexpectedly, the increased permeability in the brain was associated with a suppressed immune response in the SR group and was found to be a result of depleted infiltrating immune cells, despite significant up-regulation of gene expression related to cytokine signaling and immune cell recruitment.

Interestingly, some of the deficits observed in the SR group were marginally rescued in the DFS group. Despite this, compared to Control and SI groups, DFS animals still had significant reductions in structural and neuroimmune integrity. Further studies are required to understand the specific molecular changes that lead to these structural alterations, though Nanostring® results and previous literature allude to a loss of important tight junction and matrix-related proteins as potential causation. These alterations in the CNS and neuroimmune responses could account for the increased anxiety and sensorimotor impairments observed in the SR and DFS groups as immune system dysregulation has been shown to increase anxiety and neuronal damage. Additionally, these changes could be a correlation, or a result, of the disturbed sleep also present in these groups, as immune dysregulation is also linked to disturbed sleep, which can also increase neuroinflammation and lead to neuronal damage. A down-regulation of pathways related to neuronal integrity and neurotransmission in these groups was found and could likely be a cause of the learning impairments observed in these groups. Future studies will be necessary to explain specific changes to neuronal viability and neurotransmission.

The SI group (males) did not exhibit major differences in structural morphology or immune infiltrate compared to the Control group. Therefore, SR appears to have a main negative effect on cell viability and structural integrity in the brain while SI may have potential protective properties against the negative effects of SR. However, the synergistic relationship between these two stressors and how they interact is not understood. SI animals did exhibit differences in neuroinflammatory-related gene expression and altered astrocyte morphology compared to control groups. These changes were also found in the DFS group. It has been shown that hyperactive astrocytes can alter memory formation. SI animals did experience blunted learning in sensorimotor tasks, but further studies are required to assess the activation level of these astrocytes within these groups, and how these changes are related to sleep.

Alterations in BBB integrity and neuroimmune signaling were also associated with differences in post-stress sleep, as Vul animals had increased BBB damage and heightened immune response compared to Res animals. The adverse effect of stress may include stress-induced “priming” or sensitization of the neuroinflammatory response that can increase vulnerability to subsequent pro-inflammatory challenges. Previous studies by our lab and others have shown that the stress-induced neuroinflammatory response can be modulated by stress perception. Thus, while stress can promote inflammation, it also can be mitigated by stress resilience and vulnerability that may be regulated by differences in neural circuits that also regulate sleep. Therefore, the stressors that astronauts will likely face can produce deleterious effects both via neuroinflammation and disturbed sleep, and vulnerable individuals may be more likely to develop physical and cognitive deficits associated with these disruptions. Therefore, it is plausible that sleep itself can promote the integrity of the BBB in response to inflight stressors. However, the potential synergistic effects of multiple stressors on the relative changes in the immune system could alter these responses.

In summary, our data to date demonstrate both single and synergistic effects of SI and SR on several behaviors and sleep and also demonstrate differential effects in male Res and Vul rats as well as significant sex differences. Future work will begin to determine how extensive these differences are, and whether Res and Vul phenotypes can be identified in females.

This is the third year of this project. We have now conducted studies in one cohort of control rats (n=20), one cohort of social isolation (SI) rats (n=21), and one cohort of SR rats (n=30 irradiated (15 cGy GCRsim) and 10 shams). We thus have been able to begin making comparisons that include SI and SR animals, with (n=15) and without (n=15) SI, and also to begin delineating Res and Vul rats within treatment groups. Rats were divided into Res and Vul phenotypes based on rapid eye movement (REM) sleep responses to footshock stress. Some analyses and comparisons across groups are still ongoing.

We also have our first group of irradiated female rats. Thirty female rats were irradiated (15 cGy GCRsim) and 15 sham control rats received identical handling, but were not irradiated. The first cohort of female rats has been implanted with telemetry transmitters for recording sleep and has been scheduled for behavioral testing and sleep recording that will be conducted over the next several months.

We have made good progress over the last year, especially given the initial impediments to this project imposed by COVID-19 and the budgetary constraints that have limited the pace of our studies. The data we have acquired demonstrates both deficits associated with SR and unexpected effects of SI and SI + SR interactions. Our studies also have found that Res and Vul rats maintain their phenotypes under SI and SR and that Vul rats show greater deficits on some measures.

On the Balance Beam task, we found minimal differences between each treatment group’s ability to complete the task when the Res and Vul phenotypes were not considered separately. However, the irradiated groups (SR and SI + SR) had fewer individuals that learned the task and deficits were more pronounced in the Vul rats. Similar effects were found on the adhesive removal task test: minimal differences in performance in the task between treatment groups were observed when stress responsive Res and Vul phenotypes were not considered separately. When separated into phenotypes, on average, Vul control and SR rats took longer to notice and remove the adhesive sticker compared to Res rats within their respective treatment group. Because we found that standard behavioral measures on the Balance Beam and adhesive removal task tests were inadequate to fully evaluate behavior, we developed additional assessments of the animals’ behavioral repertoires (e.g., exploration, disequilibrium, fear and anxiety, and maintenance (grooming), and irritated behaviors). These assessments demonstrate both motor deficits/changes and interactions between motor behavior and emotion in the treatment groups, and also suggest that altered motor responses do not provide the entire explanation for behavioral performance effects. On the novel object recognition task, Res animals in the SI and SI + SR groups spent more time around the novel object than did Vul animals. Additionally, regardless of phenotype, SR animals, whether alone or in combination with SI, spent less time around the novel object compared to animals treated with SI alone.

Our sleep and electroencephalogram (EEG) data analyses show significant alterations in both sleep amounts and EEG spectra in rats subjected to SI and SR. These data demonstrate that SI, SR, and SI + SR can significantly impact the effects of additional stress (produced by footshock training and fear conditioning) on non-REM and REM sleep amounts and can alter EEG spectra in wakefulness and both sleep states. SR resulted in an increase in REM sleep amounts that occurred in association with decreased REM theta amplitude. Increased total REM sleep duration, along with decreased REM theta amplitude, is also found in association with chronic social defeat stress in rats. This suggests that the increase in REM sleep amount and decrease in REM theta in SR treated rats may reflect continuing stress or deficits in the ability to cope with stress. We will further explore this possibility in our full studies.

One of the most prominent findings in this study is an increase in freezing found in SR and SI+SR rats. Alterations in freezing associated with irradiation, including increases, have been previously reported, though the level of freezing is exceedingly high in our paradigm. This likely relates to the more extensive training regimen (20 footshocks) that we employ compared to other studies. Greater freezing has generally been interpreted as greater fear and stronger fear memory retention. However, we have found no particular correlation between freezing and a number of fear related measures (e.g., neural activity in brain stress regulatory regions, corticosterone, and stress-induced hyperthermia, sleep) indicating that it is not predictive of specific neural changes, and likely is not sufficient as a measure of the type of fear that is being experienced. Additional work will be required to determine whether and how SR differentially impacts neural activity in fear conditioned Res and Vul rats, and its functional relevance.

In summary, our data demonstrate both single and synergistic effects of SI and SR on several behaviors and sleep, and also demonstrate differential effects in male Res and Vul rats. These findings suggest that this model may be useful for determining the role of individual differences in the effects of SR and SI in ways that may lead to a better understanding of how stress and SR may differentially impact the health and performance of individual astronauts. Also, as indicated above, we are beginning our first studies with females. Once completed, we can begin making comparisons of the effects of SI and SR, and their interactions, in males and females.

This is the second year of this project. We have conducted behavioral and sleep studies on one cohort of control rats (n=20) and in one cohort of socially isolated rats (n=21). Behavioral data from these rats suggest that social isolation can produce alterations in some behaviors including altered behavior in the novel object recognition test and performance on the balance beam. We are continuing analyses that will enable us to distinguish resilient and vulnerable phenotypes and that will enable us to assess potential effects of social isolation on electroencephalographically determined sleep.

We are assessing circulating brain derived neurotropic factor (BDNF) levels both prior to, and after, irradiation and exposure to inflight stressor analogs. We have analyzed samples collected prior to irradiation in 57 Wistar outbred rats. Blood samples were collected close to the same circadian time yet the range is large, with concentration values ranging from a few pg/ml up to several thousand across animals. These animals were subsequently irradiated (studies are ongoing); thus, we will be able to determine whether BDNF baseline levels are predictive of stress and SR effects, and whether SR produces alterations in circulating BDNF. We will also be able to determine whether BDNF is predictive of resilient and vulnerable sleep phenotypes.

Studies in our first cohort of irradiated rats (n=31 irradiated and 10 shams) are just now beginning. These animals have been irradiated with 15 cGy GCRsim and are now surgically implanted and are beginning behavioral studies and studies of telemetrically recorded sleep.

In our studies of neural communication, we have conducted studies in 8 controls and 5 social isolation rats which are in the process of being analyzed. In these animals, we have found coordinated, sleep-related, local activity between brain regions (amygdala and hippocampus) in our fear conditioning paradigm. We are continuing to refine our methods for analysis and data presentation. We are also working on examining the correlation between global EEG activity and local regional activity. Additionally, we also have ongoing studies of local field potentials and sleep in 3 additional SI rats and in 12 irradiated (15 cGy GCRsim) rats. We also are beginning studies to determine if coordinated brain activity occurs in association with anticipatory activity.

This is the first year of this project. During the Definition Phase, methodologies were standardized and a cross-species validation study was designed to compare the effects of GCRsim (galactic cosmic radiation) on sleep and behavior in mice and rats. Pilot studies were conducted to optimize anticipatory activity and adhesive removal tasks in rats. We conducted one study (n=6) to determine EEG changes associated with anticipatory activity.

We implemented the cross-correlation analysis need to assess coordinated activity between local field potentials recorded in the amygdala, medial prefrontal cortex, and hippocampus in our studies on neural communication. These efforts, thus far, have found that sleep resilient and vulnerable rats show different patterns of cross-correlated activity in the amygdala and hippocampus in the aftermath of footshock stress and after fear memory recall. Specifically, the cross-correlated activity between local field potentials in the amygdala and hippocampus is reduced during rapid eye movement sleep and non-rapid eye movement sleep in resilient rats, but not in vulnerable rats. Studies will be conducted to determine whether space flight stressors produce adverse effects on neural communication that is associated with alterations in behavioral performance and emotion.

As we were scheduled to enter the Implementation Phase of the project, work was halted due to the COVID-19 pandemic shutdown. One cohort of rats (n=23) has been implanted for recording sleep via telemetry and will begin behavioral testing in the last two months of the current grant period. These tests will include anticipatory activity, novel object recognition, adhesive removal, balance beam and fear conditioning, and extinction. Additionally, we will determine whether the effects of a mild stressor on sleep are predictive of those of intense stress.