The Effects of Long-Term Exercise and Stopping Long-Term Exercise on Behaviours and Neurobiology

Abstract

Background: The immune hypothesis of major depression proposes that depression and associated anxiety and cognitive impairment share a common pathophysiology of neuroinflammation. This might arise from altered levels of tumour necrosis factor alpha (TNF) and changes in TNF signalling via the neurodegenerative TNFR1 and the more neuroprotective TNFR2. The limited efficacy of pharmacotherapies for these conditions necessitates alternative therapies. Preclinical and clinical literature overwhelmingly suggest exercise can reduce depressive symptoms and related anxiety and cognitive impairment, possibly by altering TNF and TNF receptor signalling mediated neuroinflammation. These beneficial effects of exercise suggest that stopping exercise could have adverse effects for depression, anxiety, and cognitive function and associated neurobiology. Our overall hypothesis proposed that exercise would improve anxiety-like, depression-like, and cognition-like behaviours, including in WT and TNFR1-/- exercise mice, whereas exercise would not benefit TNF-/- or TNFR2-/- exercise mice. We also hypothesised the cessation of exercise would have detrimental effects on behaviours and hippocampal neurobiology. Methods: We utilised mouse models of voluntary wheel running compared to no-exercise and exercise control mice to investigate the effects of long-term exercise and stopping long-term exercise on depression-, anxiety-, and cognition-like behaviours. This work was performed in three sections. Firstly, behaviours were investigated over the lifespan of wild type (WT) mice at four, nine, and fourteen months of age. We then examined cognitive-like and affective-like behaviours in nine month old TNF and TNF receptor deficit WT, TNF-/-, TNFR1-/-, and TNFR2-/- mice compared to their respective controls. Finally, we investigated the effects of stopping exercise on anxiety-, depression-, and cognition-like behaviours in exercise cessation mice (EC) compared to no-exercise control (CONT) and exercise control mice (EXC) during middle age. This study utilised quantitative polymerase chain reaction (qPCR) and immunohistochemistry to examine the effects of exercise cessation of hippocampal neurobiology. Results: Exercise reduced overt anxiety-like behaviours over the lifespan of WT mice, but increased freezing behaviours and spatial learning latencies in young female mice compared to control mice. Exercise improved recognition memory and spatial learning in nine month old WT mice, but had no effects on depression-like behaviours at any age or in any strain. Consistent with our hypotheses, exercise reduced anxiety in TNFR2-/- mice. Contrary to expectations, TNF-/- and TNFR1-/- exercise mice had impaired recognition memory and spatial learning, whereas exercise improved spatial learning with no changes in recognition memory in TNFR2-/- mice. Compared to EXC mice, EC mice displayed significant increases in anxiety-like, depression-like, and cognitive-like impairment behaviours. Exercise cessation also altered differential gene expression (10/75 hippocampal genes) including altered neurotrophic (Ntrk1), monoaminergic (Slc6a4), and immune (IL10, Gfap) gene expression. Discussion: We found noteworthy unanticipated effects of long term exercise suggesting further characterisation of the behavioural effects of long term exercise is warranted. Furthermore, significantly increased anxiety-like, depression-like, and cognition-like impairments and altered hippocampal neurobiology following exercise cessation are suggestive of behavioural and neurobiological changes that may occur in humans after ceasing exercise. These findings need replication and investigation in preclinical and possibly clinical studies to elucidate the adverse behavioural and physiological sequelae of exercise cessation.