Effects of age and social isolation on murine hippocampal biochemistry and behavior

Highlights

• Social isolation (SI) alters behavior and brain biochemistry in young mice.

• Old mice either failed to respond or showed a synergistic response with age to SI.

• Both responses could contribute to the health risks of SI.

Abstract

Social isolation (SI) is a major health risk in older people leading to cognitive decline. This study examined how SI and age influence performance in the novel object recognition (NOR) and elevated plus maze (EPM) tasks in C57BL/6 mice aged 3 or 24 months. Mice were group-housed (groups of 2–3) or isolated for 2 weeks prior to experimentation. Following NOR and EPM testing hippocampal norepinephrine (NE), 5, hydroxytryptamine (5-HT), 5, hydroxyindole acetic acid (5-HIAA), corticosterone (CORT) and interleukin-6 (IL-6) were determined and serum collected for basal CORT analysis. A separate set of mice were exposed to the forced swim test (FST), sacrificed immediately and serum CORT determined. SI impaired performance in the NOR and the FST, reduced hippocampal 5-HT, increased hippocampal IL-6 and increased serum CORT post-FST in young mice. Aged mice either failed to respond significantly to SI (NOR, FST, hippocampal 5-HT, serum CORT post FST) or SI had synergistic effects with age (hippocampal NE, 5-HIAA:5-HT). In conclusion, the lack of response to SI in the aged mice may affect health by preventing them adapting to new stressors, while the synergistic effects of SI with age would increase allostatic load and enhance the deleterious effects of the ageing process.

Introduction

Socially isolated older people are at an increased risk of morbidity and premature mortality (Holt-Lunstad et al., 2015). A large body of literature has shown that the detrimental health effects of SI in older people are similar in their severity to well recognized health risk factors such as smoking, high blood pressure, obesity and sedentary lifestyles (Cacioppo and Hawkley, 2003; Hawkley and Cacioppo, 2010; Umberson and Montez, 2010). Despite SI being a major health risk in older people our understanding of the mechanisms that drive increased morbidity and mortality are far from clear. SI has been linked to increased rates of mortality for people aged over 65 years, increased blood pressure, increased inflammatory (IL-6 and TNF-α) and metabolic responses to stress which can lead to obesity, type II diabetes and coronary artery disease. In addition SI increases depression, cognitive decline and even incidences of suicide (Silveira and Allebeck, 2001; Holt-Lunstad et al., 2010; Brummett et al., 2001).

To understand further the mechanisms underlying the detrimental effects of SI, researchers have explored animal models of SI. Anxiety- and depression-like behaviors are induced by periods of SI in young rodents (Amiri et al., 2015; Ieraci et al., 2016; Berry et al., 2012; Pibiri et al., 2008), consistent with the effects of SI in humans. In addition, impairments in a variety of measures of learning and memory are observed consistently in young rodents following a range of periods of social isolation ranging from 1 day to 8 weeks (Mumtaz et al., 2018; Monteiro et al., 2014; Almeida-Santos et al., 2019; Famitafreshi and Karimian, 2018; Chen et al., 2016a; Shahar-Gold et al., 2013; Liu et al., 2018). Typically, the shorter periods of SI (1 day to 2 weeks) affect social recognition memory. However, a number of studies have shown that short periods of isolation (2 weeks) can affect object recognition memory tested using the novel object recognition (NOR) test (Famitafreshi and Karimian, 2018; Chen et al., 2016a). In one of these studies a comparison was made between 2, 4 and 8 weeks of SI. In all cases SI impaired performance in the NOR although the impairment was greatest for the 8 week SI group (Chen et al., 2016a). In this study, we have chosen to explore the effects of a short (2 week) period of SI on learning and memory function and anxiety- and depression-like behaviors in 3 and 24 months old adult male mice. This period was chosen because it equates to approximately 1 year of adult human life, a period in recently widowed humans that is associated with elevated night time cortisol and a flatter diurnal slope, increasing morbidity and mortality risks (Stafford et al., 2013).

SI can act as a chronic stressor via the release of the stress hormones such as corticosterone (CORT). Indeed many SI studies have shown that basal CORT is elevated following periods of SI (Chida et al., 2006; Takatsu-Coleman et al., 2013; Hermes et al., 2009). These increases cause a consequential downregulation of glucocorticoid receptors and can evoke pro-inflammatory cytokine production in the absence of injury or infection (Costa-Pinto and Palermo-Neto, 2010). Hence, over time chronic stress has been proposed to attenuate the ability of immune cells to respond to the anti-inflammatory effects of corticosterone, which normally terminate the inflammatory cascade and this leads to low-grade chronic inflammation (Cruces et al., 2014). However, some SI studies have observed behavioral changes in the absence (Sun et al., 2014) or even following a decrease in basal CORT (Ieraci et al., 2016) suggesting that SI evoked changes in basal CORT are not the only determinant of altered behavior.

Serotonin (5-HT) has an important role in a range of psychiatric conditions, including anxiety and depression (Teissier et al., 2015; Urban et al., 2016) and in learning and memory formation (González-Burgos and Feria-Velasco, 2008; Meneses, 2015). SI affects serotonin (5-HT) release and turnover in the nervous system in young rodents (Muchimapura et al., 2002; Heidbreder et al., 2000) providing a possible explanation for incidence of these behavioral changes in isolated young rodents. However, these changes were only observed in animals isolated for periods between 6–12 weeks and whether changes in hippocampal 5-HT can explain the behavioral changes observed with shorter (2 week) periods of SI is currently unknown.

Low grade chronic inflammation (Sanada et al., 2018; Candore et al., 2010), disruptions to central 5-HT signaling (Meltzer et al., 1998; McEntee and Crook, 1991) and cognitive impairments and depression (Cacioppo and Hawkley, 2003; Fiske et al., 2009) are also common hallmarks of an aged phenotype and therefore, we were interested in exploring whether the detrimental effects of SI in old mice were due to the additive effects of SI and the natural aging process on a range of behaviors and hippocampal biochemistry.

Despite the wide body of work that has explored the effects of SI on young rodents, the behavioral consequences of SI in old adult animals, especially in mice, have been less investigated.

This study explores the effects of SI on novel object recognition memory, key markers of stress (serum CORT and CNS NE) and inflammation (IL-6). The latter inflammatory marker was chosen as it is quite stable and has been shown to be elevated in the CNS following SI (Sukoff Rizzo et al., 2012) and in group-housed aged animals (Godbout and Johnson, 2004). Given the depth of the literature on the effects of SI in young rodents, we utilized 3 months old mice to ensure our 2-week period of SI was generating results that were equivalent to those seen in previous studies. Our data showed that the effects of 2 weeks of SI on young adult animals mirrored those seen in several previous studies. In old mice, SI either i) caused a synergistic effect with age or ii) failed to alter significantly measured responses. The potential health consequences of these responses in the aged mice will be discussed.