Chronic social stress during adolescence in mice alters fat distribution in late life: Prevention by antidepressant treatment

Abstract

Obesity and visceral fat accumulation are key features of the metabolic syndrome that represents one of the main health problems in western societies due to its neurovascular and cardiovascular complications. Epidemiological studies have identified chronic stress exposure as an important risk factor for the development of obesity and metabolic syndrome, but also psychiatric diseases, especially affective disorders. However, it is still unclear if chronic stress has merely transient or potentially lasting effects on body composition. Here, we investigated the effects of chronic social stress during the adolescent period on body fat composition in mice one year after the cessation of the stressor. We found that stress exposure during the adolescent period decreases subcutaneous fat content, without change in visceral fat, and consequently increases the visceral fat/subcutaneous fat ratio in adulthood. Further, we demonstrated that treatment with a selective serotonin reuptake inhibitor (paroxetine) during stress exposure prevented later effects on body fat distribution. These results from a recently validated chronic stress paradigm in mice provide evidence that stressful experiences during adolescence can alter body fat distribution in adulthood, thereby possibly contributing to an increased risk for metabolic diseases. Antidepressant treatment disrupted this effect underlining the link between the stress hormone system, metabolic homeostasis and affective disorders.

• Chronic stress
• metabolic syndrome
• obesity
• paroxetine
• subcutaneous fat
• visceral fat

Introduction

In modern societies, obesity is a serious public health issue. Defined by body mass index of 30.0 or higher according to the National Health and Nutrition Examination Survey (NHANES), its prevalence has doubled over the last 20 years, affecting about 30% of the USA population with no trend for a decline (Hedley et al. 2004). The incidence of obesity is associated with arterial hypertension, glucose intolerance, insulin resistance, increased visceral fat and dyslipidemia, when clustering together, are referred as metabolic syndrome (Day 2007). The following diseases associated with obesity impose a significant economic burden on individuals, families and public health systems (Ogden et al. 2007).

Chronic stress has been identified as a risk factor for developing metabolic disturbances (Seematter et al. 2004). Chronic work stress, for example, was demonstrated to correlate with obesity in human beings and to double the risk of developing the metabolic syndrome. A recent study demonstrates that in originally non-obese young men long-term stress induced abdominal obesity and the metabolic syndrome about five months later (Branth et al. 2007). From animal studies, it is further known that stress affects feeding behaviour, caloric intake, body weight changes and fat content (Tamashiro et al. 2007).

Chronic stress also represents a major risk factor for the development of affective disorders. In turn, depressive patients more often suffer from general obesity, visceral obesity, metabolic syndrome and cardiovascular complications (Faith et al. 2002; Heiskanen et al. 2006). The observation of patients with hypercortisolaemia carry a higher risk to develop visceral obesity supported the view that there exist close relationships between hypothalamic-pituitary-adrenal (HPA) axis activity, obesity and visceral fat accumulation (Rosmond and Bjorntorp 1998). Additionally, clinical studies have demonstrated the long-term effects of depressive disorders and psychosocial stress on the risk to develop metabolic syndrome and obesity (Raikkonen et al. 2007). These observations have led to the hypothesis of a common pathophysiological basis of affective disorders and the metabolic syndrome (Kyrou and Tsigos 2007).

In the light of limited data explaining the pathophysiology of the delayed or prolonged impact of chronic stress on metabolism and body composition, pre-clinical studies using animal models for chronic stress in combination with non-invasive magnetic resonance imaging (MRI) are of particular use and can help to elucidate underlying mechanisms. The first goal of the present study was to probe whether chronic social stress applied during the adolescent period of mice has an effect on body fat distribution when the mice are much older, i.e. 12 months after cessation of the stress regimen. Given the current view that the restoration of normal HPA axis regulation represents a common pathway of antidepressant drug action and in many patients normalisation of the HPA system has been shown to be an important step for stable remission (Holsboer 2000), we further hypothesised that lasting effects of chronic stress on metabolic parameters may be prevented by a simultaneous chronic treatment with an antidepressant throughout the period of stress exposure.

Methods

Animals

Experiments were carried out with male CD1 mice from the Charles River Laboratories (Maastricht, The Netherlands). The mice were 26–28 days old on the day of arrival. All mice were initially housed in groups of four per cage (45 × 25 × 20 cm) under a 12L:12D cycle (lights on at 6.00 h) and controlled temperature (23 ± 2°C) conditions. After the chronic stress procedure, all mice were singly housed. Food (normal mouse diet) and water were provided ad libitum. This experiment was performed at the animal facility of the Max Planck Institute of Psychiatry in Munich, Germany.

The experiments were carried out in accordance with European Communities Council Directive 86/609/EEC. All efforts were made to minimise animal suffering during the experiments. The protocols were approved by the committee for the Care and Use of Laboratory Animals of the Government of Upper Bavaria, Germany.

Chronic stress procedure

The chronic social stress procedure was performed as described previously (Schmidt et al. 2007). Briefly, after a habituation period of five days following arrival, the group composition in each cage was changed twice per week for seven weeks, so that each time four mice from different cages were put together in a new and clean cage. The rotation schedule was randomised to minimise the likelihood of repeated encounter among the same mice. The control mice were remained continuously with the same cage mates throughout the experiment. Body weight was monitored every two weeks throughout the stress procedure. Mice with bite wounds were excluded from the experiments (less than 2% of the mice in the chronic stress group).

Experimental design

Three different groups were used: (1) control mice, (2) untreated chronic stress mice and (3) chronic stress mice treated with the SSRI paroxetine (see Drugs). We did not include a paroxetine-treated control group, as the main focus of this study was on the preventive action of antidepressants on stress-induced alterations rather than the effects of antidepressants in healthy subjects. Three weeks after the start of the chronic stress procedure, 12 mice of the chronic stress group were treated with the antidepressant paroxetine, while 20 mice remained untreated. Drug treatment was continued for four weeks until the end of the stress exposure. Control mice (N = 22) were not treated with paroxetine. At the end of the stress procedure, mice from all groups were separated and singly housed. After 12 months of single housing with no further intervention, all mice were sacrificed under basal conditions.

Drugs

Paroxetine (GlaxoSmithKline, Munich, Germany) was diluted in tap water to a final concentration of 0.16 mg/ml and applied orally via the drinking water. With average water consumption of 5 ml/mouse/day, the daily dose of paroxetine was approximately 20 μg/g body weight per day. Fluid intake was monitored daily and the variation of fluid intake was found to be less than 10% during the course of the experiment, thereby ensuring the correct dosage of the drug. Drug solutions were replaced on a daily basis.

Locomotor activity

At the age of 15 months, the mice were subjected to an open field test to assess locomotor activity. Open field arenas were made of grey PVC (50 × 50 × 50 cm) and evenly illuminated during testing (50 lux). General locomotor activity was recorded for 5 min (distance travelled) using a video-tracking system (Anymaze 4.20, Stoelting, Illinois, USA).

Endocrine parameters

Tail cut blood was collected individually in labelled 1.5 ml EDTA-coated microcentrifuge tubes (Kabe Labortechnik, Germany). All blood samples were kept on ice and later centrifuged for 15 min at 6000 rpm at 5°C. Plasma was transferred to clean, labelled 1.5 ml microcentrifuge tubes. All plasma samples were stored frozen at − 20°C until the determination of corticosterone by RIA (MP Biomedicals Inc.; sensitivity 6.25 ng/ml).

Blood glucose was measured with a glucometer (Elite, Bayer, Lever Kusen, Germany) in a separate cohort of mice after an overnight fast.

MR imaging

To assess intra-abdominal fat distribution in the mice, measurements were made ex-vivo on a 7 T MRI BRUKER system, using a volume resonator and a T1-weighted sequence (2D spin-echo, TR = 900 ms, TE = 11.6 ms, total acquisition 30 min) (Figure 1A). To achieve full body coverage, two slice packages with one slice overlap were scanned subsequently without repositioning the mouse.

Figure 1 Ex-vivo MRI measurement of body fat. (A) Axial slice through the mouse abdomen. Fatty tissue is bright. A region-of-interest is indicated segmenting the image into visceral and (sub)cutaneous compartments (dotted line). (B) Visceral to subcutaneous fat ratio. Mice that have experienced chronic social stress during adolescence and early adulthood display a significant increase in the visceral/subcutaneous fat ratio 12 months after the termination of the stress exposure. This effect was fully prevented by a simultaneous treatment with the selective serotonin reuptake inhibitor paroxetine during the last four weeks of stress exposure. * significant at P < 0.05.

Images were analysed using the ParaVision software (BRUKER, Germany). For each slice, regions-of-interest were set including visceral tissue, and excluding subcutaneous, cutaneous and skeletal muscle compartments. The fat compartment was analysed based on these regions along with an image intensity threshold (2/3 of fat peak intensity as determined per slice) to suppress other tissue components. Cutaneous and subcutaneous fat compartments were identified by summing all image pixels above the threshold excluding visceral volumes. Total fat was estimated by summing the areas of the respective fat compartment in all slices. Subsequently, ratios of visceral to subcutaneous fat were calculated. For statistical analysis, we applied one-way ANOVA followed by LSD post hoc tests. The level of significance was set at p < 0.05.

Results

At the time of testing, the three groups studied (controls, chronic stress and chronic stress + paroxetine) did not differ in mean body weight (Table I). Further, no significant differences were detected for locomotor activity, food intake, fasting glucose levels and basal morning corticosterone concentrations (Table II). When analysing the body composition of the mice in terms of fat location, we detected a significant effect of group on subcutaneous fat content (F(2,48) = 9.26, p < 0.001). Mice subjected to chronic social stress during adolescence had significantly less subcutaneous fat, but not visceral fat, compared to controls and paroxetine-treated stress mice (Table III). Further, we found significant group differences for the visceral-to-subcutaneous fat ratio (F(2,48) = 4.073, p < 0.05) (Figure 1). Post hoc comparisons located the effect to the untreated stressed mice that had a higher visceral-to-subcutaneous fat ratio compared to controls and paroxetine-treated mice. No difference was found between the control group and the paroxetine-treated chronic stress mice. (p = 0.93).

Table I.  Body weight at the age of 9 and 15 months.

Age	Condition			ANOVA
	Control	Stress	Stress+ paroxetine	F-value	p-Value
9 Months	46.2 ± 1.1	45.1 ± 1.6	47.4 ± 1.6	0.463	0.634
15 Months	47.8 ± 1.6	50.6 ± 2.6	53.6 ± 1.7	2.206	0.135

Data are given as group mean ± SEM. n = 10–12 mice per group.

Table II.  Comparison of locomotor activity, food intake, fasting blood glucose and basal plasma corticosterone concentrations between the three experimental groups.

Measure	Condition			ANOVA
	Control	Stress	Stress+ paroxetine	F-value	p-Value
Open field distance (m)	48.2 ± 3.3	46.6 ± 3.8	45.7 ± 3.3	0.132	0.877
Food intake (g)	4.9 ± 0.6	5.0 ± 0.4	5.7 ± 0.7	1.166	0.324
Fasting glucose (mg/dl)	73.2 ± 7.6	78.4 ± 5.5	86.2 ± 4.6	1.203	0.316
Corticosterone (ng/ml)	10.9 ± 4.5	9.4 ± 3.2	8.5 ± 2.9	0.105	0.9

No significant differences were found in these parameters. Data are given as group mean ± SEM. n = 10–12 mice per group.

Table III.  Visceral and subcutaneous fat compartments for the three experimental conditions.

Measure	Condition			ANOVA
	Control	Stress	Stress + paroxetine	F-value	p-Value
Visceral fat content (cm^3)	4.61 ± 0.33	3.72 ± 0.33	4.28 ± 0.45	1.82	0.17
Subcutaneous fat content (cm^3)	4.00 ± 0.29	2.46 ± 0.20	3.59 ± 0.32	9.26	< 0.001

Data are given as mean ± SEM. n = 10–12 mice per group.

Discussion

In this study we demonstrate for the first time the long-lasting effects of early life chronic stress experience on body fat distribution later in life in an animal model. At the age of 15 months, previously stressed mice showed a marked decrease in subcutaneous, but not visceral, fat content and an increase in the visceral to subcutaneous fat ratio. Intriguingly, these effects were long lasting and observed few months later under stress-free conditions after the termination of the stressor. Our current results are further corroborated by the findings of lasting effects of chronic stress in terms of behavioural, neuroendocrine and neurogenetic alterations (Sterlemann et al. 2008). Notably, a recent clinical report on increase in the visceral fat compartment in patients after an acute episode of depression supports the notion that exposure to stress may also have lasting effects on human fat distribution despite antidepressant treatment (Weber-Hamann et al. 2006). This further suggests that similar to our experiments also in human beings acute or chronic stress prior to clinical depression may represent the critical triggering factor of such metabolic alterations.

What mechanism could account for the long-term effects of chronic social stress on body fat distribution. It is easily conceivable that in addition to central effects, chronic stress exposure has also affected peripheral glucocorticoid signalling, since we have previously observed the long-lasting elevation of morning corticosterone levels and flattened circadian rhythm in the mice of chronic stress group (Schmidt et al. 2007). Further, mice subjected to chronic social stress during adolescence still display decreased expression levels of the mineralocorticoid receptor and the glucocorticoid receptor in the hippocampus 12 months after stress exposure (Sterlemann et al. 2008). A persistently altered expression of peripheral glucocorticoid receptors might therefore cause changes in downstream protein expression or altered enzymatic activities. Mazusaki et al. (2001) showed that a transgenic mouse line locally over-expressing 11β-hydroxysteroid dehydrogenase type 1 (11β HSD-1), an enzyme that catalyses the regeneration of active glucocorticoids from inactive keto-forms, in adipose tissue displayed the increased adipose levels of corticosterone and developed visceral obesity. As those mice also exhibited the pronounced insulin-resistant diabetes and hyperlipidaemia among other symptoms, the authors proposed that increased adipocyte 11β HSD-1 activity may be a common molecular etiology for visceral obesity and the metabolic syndrome (Masuzaki et al. 2001). This adipose tissue specific increase in 11β HSD-1 activity has recently be shown to be a common feature in obese human beings and rodents (Seckl 2004). Thus, the change in the activity of 11β HSD-1 may possibly also result from chronic stress exposure, which in our paradigm occurs during the highly adaptive and vulnerable period of adolescence and young adulthood. In addition, there are most likely numerous other changes in gene or protein expression of central or peripheral factors as a result of chronic stress, determining the visceral and subcutaneous fat content either directly or by influencing central body fat regulation, e.g. by the hypothalamus. As the mice of the chronic stress group, who had received the antidepressant paroxetine during the stress phase did not show the change in fat distribution, it can be proposed that central mechanisms are responsible for the changes in metabolic homeostasis. This hypothesis is supported by our earlier findings, where chronic treatment with paroxetine normalised several other centrally regulated HPA and behavioural parameters, such as increased morning corticosterone levels, decreased hippocampal MR and GR expression and an increase in anxiety-related behaviour (Schmidt et al. 2007). Indeed, there might exist converging underlying mechanisms that are responsible for both the development of depression and the metabolic syndrome. However, it cannot be excluded that paroxetine treatment directly affected peripheral tissues, thereby permanently altering the impact of glucocorticoids on fat tissue distribution during ageing.

These results have strong clinical implications. Our findings help to interpret links between early stress exposure as a risk factor for metabolic syndrome and obesity in human beings. Early life stress has mainly been investigated as a risk factor for later development of affective disorders, but not for sustained or delayed metabolic changes. Only in the last few years, evidence has emerged that depressive symptoms or stressful life events increase the risk to develop visceral obesity, Type 2 diabetes or the metabolic syndrome later in life (Räikkönen et al. 2007). In cross-sectional analyses, particularly hypercortisolaemic states were found associated with imbalanced metabolic homeostasis, suggesting a pathophysiological link (Vogelzangs et al. 2007). Yet, to our knowledge, no clinical data exist that link adolescent HPA axis profiles to metabolic parameters later in life. Our results should foster such longitudinal cohort studies. Thus, while mice exposed to chronic social stress during adolescence are not obese in adulthood, the alteration in body fat distribution might represent a significant risk factor for metabolic diseases.

Further implications arise from the results of the pharmacological intervention. The restoration of normal HPA axis signalling has been identified as a sensitive marker of successful antidepressant pharmacotherapy (Ising et al. 2007). It may therefore be speculated that early pharmacological targeting of the stress hormone system in affective disorders may also prevent later metabolic complications. Future work also needs to clarify if such therapeutic consequences should be extended from psychiatrically manifest disease to situations of endocrinologically measurable chronic stress. Extrapolation of the current results is limited, however, as other metabolic parameters including lipid metabolism, insulin sensitivity or metabolic rate were not measured in this study.

In conclusion, chronic stress during adolescence and young adulthood was associated with an altered body fat distribution in adulthood. The effect could be suppressed by parallel application of an antidepressant during stress exposure. The results support and confirm the sparse human data proposing that chronic stress is a risk factor for metabolic disturbances and may predispose individuals to metabolic diseases throughout life.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.