Sleep homeostasis and depression: Studies with the rat clomipramine model of depression

Abstract

Neonatal treatment of rat pups with clomipramine (CLI) has been shown to cause long-lasting and persistent depression-related behaviors and changes in sleep architecture and in brain-derived neurotrophic factor (BDNF) signaling in adult animals, producing an animal model of depression. However, the molecular mechanisms which mediate these effects of early-life CLI treatment on adult animals remain largely unknown. In order to characterize these further, we investigated in neonatally CLI-treated rats the sleep architecture as well as the extracellular and cellular levels of sleep regulators (nitric oxide, adenosine) and BDNF, respectively, in the basal forebrain (BF), i.e. the brain area which is implicated in sleep and depression. We found that CLI-treated rats exhibited a disturbed sleep architecture (REM sleep fragmentation was increased and NREM periods preceding REM were shorter) and reduced levels of BDNF and adenosine in the BF, whereas the levels of nitric oxide were elevated. Next, we examined sleep deprivation (SD)-induced homeostatic responses on sleep regulation and brain BDNF levels in CLI-treated rats. Compared to control rats, 3 h of SD induced a smaller increase in the amount of NREM sleep during sleep recovery. At the molecular level, the normal homeostatic response was dissociated: the rise in the adenosine level was not accompanied by a rise in the nitric oxide concentration. Moreover, while BF BDNF levels decreased during SD in control rats, such a decline was not observed in CLI rats. Taken together, neonatal CLI treatment produces long-lasting functional changes in the sleep architecture and sleep regulation in adult rats, accompanied by dysregulated BDNF signaling in the BF.

Introduction

Depression is a major psychological disorder affecting ∼7% of the European Union population (Wittchen et al., 2011) and often accompanied by sleep disturbances, but the relationship between sleep and depression is complex and poorly understood. Clinical observations evidence that almost all depressive patients show some type of sleep disturbance (Benca and Peterson, 2008, Staner, 2010), most commonly early morning awakenings, increased sleep fragmentation and disturbed regulation of rapid eye movement (REM) sleep (Fleming, 1994, Adrien, 2002). But it is so far unclear whether there is a causal relationship between disturbed sleep and the development of depression. Some evidence to support such a causality is presented by a longitudinal study where individuals with sleep complaints had an elevated risk or susceptibility for depression (Paunio et al., 2009, Gass et al., 2010). Another interesting question is whether the regulation of sleep and mood share a genetic background, and if they do, on what mechanisms these relationships are based. In earlier studies we did in fact identify genes which associate with depression only in individuals who also suffer from sleep problems (Utge et al., 2010). But in spite of these promising hints, very little is known about the brain molecular mechanisms by which disturbed sleep could predispose to depression.

One animal model of depression, which is produced by neonatal clomipramine (CLI) treatment and was originally proposed by Vogel et al. (1990), appears to be particularly suitable for investigations of this relationship. The treatment induces, in adulthood, not only the classical signs of depression, such as anhedonia, but also REM sleep abnormalities, including REM sleep fragmentation and shortened REM sleep latency (Vogel et al., 1990), both typical symptoms of human depression (Fleming, 1994). Whether one of the key features of sleep, namely sleep homeostasis, is also disturbed in depression has remained questionable (Borbely, 1987, Armitage et al., 2000, Popa et al., 2006). The normal sleep homeostatic response consists, at the level of the electroencephalogram (EEG), of increased slow-wave activity after prolonged waking (Borbely, 1987), and at the molecular level of increased extracellular concentrations of adenosine and nitric oxide (NO) in the basal forebrain (BF) (Porkka-Heiskanen et al., 1997, Porkka-Heiskanen et al., 2002, Kalinchuk et al., 2006b).

Brain-derived neurotrophic factor (BDNF) is one of the key factors regulating neuronal plasticity in developing adult brains, and its protein level has been reported to be reduced in the hippocampus of rats neonatally treated with CLI (Cassano et al., 2006). Increasing evidence suggests that abnormal BDNF signaling is associated with depression. For example, reduced levels of BDNF have been measured in the serum (Karage et al., 2002, Kim et al., 2007) and the hippocampus (Chen et al., 2001) of depressed subjects. BDNF synthesis is increased during neuronal activation (Zafra et al., 1990), and also its expression has been reported to be elevated in the cortex after sleep deprivation (SD) (Cirelli and Tononi, 2000). Moreover, an experimental increase in cortical BDNF increases the slow-wave activity, while decreasing the effect of BDNF reduces it (Faraguna et al., 2008). Thus, BDNF is closely associated with both the regulation of sleep and the development of depression.

In the present study, we asked whether sleep homeostasis-regulating mechanisms are disturbed in neonatally CLI-treated rats. Our hypothesis was that early CLI treatment permanently modifies the brain’s neuromodulator and neurotrophin levels, particularly those which are involved in the regulation of sleep and sleep homeostasis, and that disturbances in these mechanisms may contribute to the behavioral signs of depression.