Increased hippocampal tau phosphorylation and axonal mitochondrial transport in a mouse model of chronic stress

Abstract

Corticotropin-releasing hormone (CRH) is considered the driving force of the hypothalamo-pituitary-adrenal (HPA) axis and plays an important role in mood regulation. The HPA axis is reported to be closely related to acute stress-induced tau phosphorylation in the rodent hippocampus. However, the relationship between the hyperactive HPA axis and tau phosphorylation in the hippocampus and hence the functional implications for chronic stress are not fully understood. In this study, we aimed to examine tau phosphorylation and the effect on axonal transport of mitochondria in the hippocampus of a chronic stress model. A mouse model was created by neonatal isolation before weaning, followed by chronic mild stress by social isolation after weaning. Behavioural tests showed that the model had a typical depression/anxiety-like behaviour accompanied by increased plasma corticosterone level and hypothalamic CRH mRNA expression. Phosphorylated tau increased significantly, accompanied by increased synaptosomal mitochondrial levels in hippocampus of the chronic stress model. CRH receptor 1 antagonist (CP154,526) treatment, not glucocorticoid receptor antagonist (RU486) treatment, decreased tau phosphorylation and synaptosomal mitochondrial levels in the hippocampus of the mouse model. Consistent with an in-vivo model, when hyperphosphorylated tau was inhibited by lithium in cultured primary hippocampal neurons, mitochondrial transport monitored by live imaging was also decreased. We show here for the first time that phosphorylated tau in the hippocampus of a chronic stress model, accompanied by increased mitochondrial transport, was mediated by CRH receptor 1, not by glucocorticoid receptors, which suggests that centrally derived CRH may be involved in the process of mitochondrial axon transport and hence play an important role in hippocampus of a chronic stress model.

Introduction

To date, as a risk factor for mood disorder, stressful life events have been linked to increased hypothalamo-pituitary-adrenal (HPA) axis activity (Bao et al.2008; Carroll et al.2007). Patients with depression may have elevated corticotropin-releasing hormone (CRH) mRNA levels in the paraventricular nucleus (PVN) (Wang et al.2008) and elevated cortisol levels in plasma (Burke et al.2005). As the driving force of the HPA axis, CRH also plays an important role in mood regulation in other brain regions (Hammack et al.2002; Jedema & Grace, 2004; Kagamiishi et al.2003). CRH-containing neurons in PVN may modulate hippocampal functions by the direct PVN–hippocampal CRH pathway (Koegler-Muly et al.1993).

Tau proteins are highly soluble microtubule-associated proteins; they play an important role in stabilizing microtubules and regulating axonal transport (Cleveland et al.1977; Ebneth et al.1998). A range of acute stresses can induce tau phosphorylation in the rodent CNS, especially in hippocampus formation (Ikeda et al.2007; Korneyev et al.1995; Papasozomenos, 1996; Yanagisawa et al.1999). Experimental study has shown that CRH receptors are involved in immobilization-induced tau phosphorylation in the mouse hippocampus (Rissman et al.2007). However, the relationship between hyperactivity of the HPA axis and tau phosphorylation in the hippocampus and hence the functional implications for chronic stress are not fully understood.

Mitochondria are not only important organelles for energy supply and signalling, but are also closely related to psychiatric disorders (Gardner et al.2003). Mitochondria are transported into the synapse along axons, and tau proteins are implicated in regulating these processes (Ebneth et al.1998; Stamer et al.2002). Here, we hypothesize that CRH may be involved in stress-induced tau phosphorylation, and might further affect mitochondrial translocation in the hippocampus of a chronic stress model.

In the present study, after behavioural testing of the chronic stress model, we checked HPA axis activity, tau phosphorylation in the hippocampus, and mitochondrial levels in isolated synaptosomes from the hippocampus. Live imaging of mitochondria in cultured primary hippocampal neurons was performed to examine the relationship between tau phosphorylation and mitochondrial transport.

Materials and methods

Animals

ICR male mice were used in the study. The animals were maintained on a 12-h light/dark cycle (lights on 07:00 hours), at a temperature of 22±1°C. All procedures were performed according to the Animal Care and Use Committee of University of Science and Technology of China.

Chronic stress model

There is increasing evidence that early adverse experiences contribute to the development of stress susceptibility, and increase the onset of stress-related psychiatric disorders in stressful environments in adulthood (Avital & Richter-Levin, 2005; Rivarola & Suarez, 2009). Accordingly, we created a model by neonatal isolation before weaning, followed by chronic mild stress by social isolation after weaning. Briefly, the day of birth was designated postnatal day 0 (PND 0). On PND 0, litters were culled to 10 male pups (five control males, five model males) and housed with their mothers in plastic cages. Model pups were isolated from the dam, nest, and siblings in individual plastic cages for a period of 4 h (between 10:00 and 14:00 hours) once per day over PND 1–20. Animals of both isolated and non-isolated pups received equal handling time. Litters were weaned on PND 21, every model pup was placed in an individual cage (24 cm×13 cm×13 cm), and five control pups were placed in one larger cage (40 cm×28 cm×18 cm). Experiments were performed when the mice were aged 3 months.

Behavioural tasks

Three kinds of behavioural tasks were performed: open-field, sucrose preference and dark-light box. Open-field and dark-light box tests were used to check the exploratory behaviour, and sucrose preference was used as a measure of anhedonia, a core symptom of depression. In total, 48 mice were used in the behavioural tests (control group, n=24; chronic stress group, n=24).

Open-field test

The open-field apparatus consisted of a black square arena 100 cm×100 cm, with a black wall 30 cm high. The floor was marked with a grid dividing the floor into 16 equal-sized squares. Each mouse was placed in one of the four corners, facing the wall, and was permitted to explore the environment for 5 min. A record was kept of the number of crossings in the central four squares, and the number of rearings (defined as standing upright on hind legs).

Sucrose preference test

Each mouse was housed individually for 24 h prior to the preference test. The mice were deprived of water for 24 h, and then were given two bottles for 24 h, one bottle containing a 1% sucrose solution, placed in the original bottle's position, and the other containing water placed in a new position. The amount of each solution consumed was determined by weighing the bottles before and after the 24-h consumption period. Sucrose preference was calculated as total sucrose intake/total (sucrose+water) intake.

Dark-light box test

The mice were placed in a box comprised of two compartments, one light and the other dark, both were the same size (25 cm×25 cm×25 cm), separated by a Plexiglas panel with an arc hole in the ground side to allow mice to pass through. The mice were gently placed in the light compartment facing the wall opposite to the hole and monitored for 5 min; the time spent in each chamber was recorded.

Pharmacological treatment

After behavioural testing, pharmacological treatment was performed. Six groups were assigned: control group with vehicle; model group with vehicle; control group with CP154,526; model group with CP154,526; control group with RU486; and model group with RU486. Each group contained six mice. The corticotropin-releasing hormone receptor 1 (CRH-R1) antagonist CP154,526 was administered (20 mg/kg i.p. injection, Tocris, USA) once a day for 1 wk, and a daily dose of the glucocorticoid receptor antagonist RU486 (30 mg/kg i.p. injection, Sigma-Aldrich, USA) was given for 1 wk. CP154,526 is a selective, non-peptide CRH-R1 antagonist (K_i values are 2.7 nm for CRH-R1 and >10 000 nm for CRH receptor 2) (Schulz et al.1996); RU486 is a potent glucocorticoid receptor antagonist with K_i values of ∼1 nm (Cadepond et al.1997).

Tissue preparation

Mice were anaesthetized with sodium pentobarbital (40 mg/kg) and decapitated between 10:00 and 11:00 hours. Blood samples were collected in tubes containing heparin sodium as an anticoagulant and centrifuged at 4°C, after separation the plasma was stored at −70°C until assayed with mouse corticosterone enzyme-linked immunosorbent assay (ELISA) kit (RapidBio Lab, USA). Bilateral hippocampus was rapidly dissected and one hippocampus was used for isolation of synaptosomes, and the other was frozen in liquid nitrogen and preserved at −80°C for Western blotting. Hypothalamic parts were dissected according to Yasin et al. (1993) with the following limits: anterior border of the optic chiasm, anterior border of the mamillary bodies, and lateral hypothalamic sulci. The depth of dissection was ∼3 mm. The hypothalamus was then quickly frozen in liquid nitrogen and preserved at −80°C for quantitative reverse transcriptase–polymerase chain reaction (qRT–PCR).

qRT–PCR

Each frozen hypothalamus was homogenized, and total RNA was isolated with TRIzol Reagent (Invitrogen, USA). cDNA was synthesized using reverse transcriptase (Takara, Japan). q-PCR was performed using a SYBR Green PCR kit (Applied Biosystems, USA) and a StepOne real-time PCR system (Applied Biosystems, USA) in 25 µl volume for 40 cycles (15 s at 95°C and 60 s at 60°C). Mouse CRH was amplified using primers 5′-aggaggcatcctgagagaagt-3′ and 5′-catgttaggggcgctctc-3′ (86 bp, intron spanned). Primers for mouse β-actin were 5′-tgttaccaactgggacgaca-3′ and 5′-ggggtgttgaaggtctcaaa-3′ (165 bp, intron spanned). The relative amount of target gene was calculated using the 2^ΔΔCt method. The relative amplification efficiencies of the primers were tested and shown to be similar.

Western blotting analysis

Hippocampal tissues were homogenized in radioimmunoprecipitation assay (RIPA) buffer [50 mm Tris–HCl (pH 7.4), 0.1% SDS, 1% NP-40, 0.25% sodium deoxycholate, 150 mm NaCl, 1 mm EDTA, 1 mm EGTA, 1 mm Na₃VO4]. Before homogenization, protease inhibitor cocktail (Roche, USA) and phosphatase inhibitor PhosSTOP (Roche) were added. To determine phosphorylated residues on tau, proteins were probed with AT8 (S202/T205; 1:500; Pierce Biotechnology, USA), and tau-1 (1:2000; Chemicon, USA). tau5 (1:1000, Biosource, USA) was used for total tau expression level, and tubulin (1:5000, Sigma) was used as a control for protein loading. COX IV (cytochrome c oxidase, 1:3000, Abcam, USA) was used as a marker for mitochondria. Post-synaptic density protein 95 (PSD95; 1:2000, Abcam) was used as control for synaptosomes. Background subtraction was performed, and quantitative band intensity readings were obtained using NIH Image software.

Synaptosomal isolation

Hippocampal synaptosomes were isolated by the Percoll gradient method with slight modification (Dunkley et al.2008). In brief, freshly removed hippocampus was minced in ice-cold preparation buffer containing 0.32 m sucrose, 1 mm EDTA, 5 mm Tris (pH 7.4) with cocktail (Roche). The tissue was gently homogenized (15 passes in a Teflon-glass homogenizer; 600 rpm), and the resulting homogenate was centrifuged at 1000 g (10 min, 4°C). The supernatant was layered on top of a freshly prepared four-step discontinuous Percoll gradient (0, 10, 23, and 40% Percoll in preparation buffer; pH 7.4). Gradients were centrifuged at 31 000 g for 7 min, synaptosomes were collected at the interface (10%/23%) and diluted in 10 volumes of preparation buffer and centrifuged at 14 000 g for 30 min. The pellet was resuspended in RIPA buffer.

Primary hippocampal neuron culture and transfection

Hippocampal neurons were isolated from ICR mice (E18) by a standard enzyme treatment protocol. Briefly, hippocampal tissues were dissociated with trypsin–EDTA solution (0.25% trypsin and 1 mm EDTA; Invitrogen) and the neurons were plated (1–5×10⁶ cell/ml) on 35-mm glass-bottomed culture dishes (Mattek Corporation, USA). The culture medium was a neurobasal medium (Invitrogen) with 2% B27 (Invitrogen). To stabilize the cell population, the culture was treated with 5-fluoro-5′-deoxyuridine (20 µg/ml, Sigma) on the fourth day after plating to block cell division of non-neuronal cells. The cultures were maintained at 37°C in a 5% CO₂ humidified atmosphere.

Transfection was performed using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Briefly, on day 5, neurons were transfected with either a pEGFP-mito plasmid (Clontech, USA) or pEGFP-mito plasmid and pDsred-htau40 plasmid. pEGFP-mito is designed for fluorescent labelling of mitochondria, which encodes a fusion of a mitochondrial targeting sequence derived from the precursor of subunit VIII of human cytochrome c oxidase and the enhanced green fluorescent protein (EGFP). Cells were incubated in transfection media for 3 h and then the media was replaced with original medium. On day 7, live imaging was performed.

Live imaging

Mitochondria were observed by a Leica AF6000LX microscope, equipped with a 63x oil-immersion objective. Mitochondrial movement was imaged for 5 min, 10 s per time. To examine the effect of LiCl on mitochondrial movement, after 2 h treatment with 20 mm LiCl, a second 5-min imaging was performed in situ. In order to determine quantitatively the effect of LiCl, the number of anterogradely or retrogradely moved mitochondria in axons during 5-min imaging was calculated, and the net result was calculated with the number of mitochondria per 25 µm axons.

Data analysis

Statistical analysis was performed using SPSS for Windows, version 11.5 (SPSS Inc., USA). Unless otherwise indicated, data are expressed as means±s.e.m. Differences among control mice, model mice, control mice treated with CP154,526, model mice treated with CP154,526, control mice treated with RU486 and model mice treated with RU486 were tested with one-way ANOVA, differences between means in one-way ANOVA were further analysed by LSD test. The effect of LiCl on mitochondrial movement in cultured primary hippocampal neurons was tested with paired-sample t tests. Other differences between groups were tested with two-sample t tests. A p value <0.05 was considered significant.

Results

Behavioural characterization of chronic stress model used for the study

In the open-field test, model mice showed significantly less rearing (p<0.001, Fig. 1a) and significantly less crossing of central squares (p<0.001, Fig. 1b) compared to control mice, which indicated decreased exploratory behaviour in model mice. Sucrose intake is used as a measure of anhedonia, a core symptom of depression. In the sucrose preference test, the model mice displayed significantly reduced levels of sucrose preference compared to control mice (p<0.05, Fig. 1c). The time spent in the bright chamber is used as an index of anxiety/depression-like and exploratory activity. In the dark-light box task, the model group spent less time in the light chamber compared to control mice (p<0.001, Fig. 1d). The results showed that model mice displayed typical depression/anxiety-like behaviour compared to control mice.

Hyperactivity of HPA axis in chronic stress model

The HPA axis was hyperactive in the chronic stress model compared to control mice. The model showed a higher corticosterone concentration in plasma (p<0.05, Fig. 2a), and the relative CRH mRNA expression in hypothalamus was also significantly increased in the model (p<0.05, Fig. 2b).

Increased tau phosphorylation in hippocampus of the chronic stress model

Western analyses (Fig. 3a) indicated no change in tau protein level in the hippocampus (p=0.496, Fig. 3b). The tau phosphorylation level increased significantly at AT8 sites (p<0.01, Fig. 3c), decreased significantly at tau-1 sites (p<0.05, Fig. 3d) in the hippocampus of model mice compared to control mice. The results showed that model mice had a higher level of tau phosphorylation in the hippocampus.

More mitochondria in synaptosomes from hippocampus of model mice

As a microtubule-associated protein, tau proteins play an important role in regulating axonal transport. We compared the mitochondrial level in synaptosomes in the hippocampus of both groups by analysing the mitochondrial marker COX IV. Western analyses (Fig. 4a) indicated that model mice had a higher COX IV level in synaptosomes (p<0.001, Fig. 4b), which indicated more mitochondria in hippocampal synaptosomes of the chronic stress model. PSD95 was used as an internal control for synaptosomes.

CRH-R1, but not glucocorticoid receptor involvement in tau phosphorylation in hippocampus of model mice

We further checked possible pathways of CRH and corticosterone in tau phosphorylation in the hippocampus of the chronic stress model with two receptor antagonists: CRH-R1 antagonist CP154,526 and glucocorticoid receptor antagonist RU486. The results of Western blotting (Fig. 5a) showed that the CRH-R1 antagonist CP154,526 significantly decreased tau phosphorylation in the hippocampus of model mice (p<0.05, Fig. 5b); however, after treatment with RU486, the tau phosphorylation level in model mice hippocampus was still higher than that in control hippocampus (p<0.05, Fig. 5b), which indicated that the increase of tau phosphorylation was mediated by CRH-R1, not by glucocorticoid receptor.

CRH-R1 antagonist reduced synaptosomal mitochondria levels in hippocampus of model mice

We hypothesized that change of tau phosphorylation might be related to mitochondrial translocation between cytoplasm and synapse. CRH-R1 antagonist could decrease the level of tau phosphorylation; therefore we checked the effect of the CRH-R1 antagonist on mitochondrial translocation. The results (Fig. 6a) showed that the CRH-R1 antagonist CP154,526 reduced synaptosomal mitochondrial levels in the hippocampus of model mice (p<0.05, Fig. 6b); however, after treatment with RU486, the synaptosomal mitochondrial level in model mice hippocampus was still higher than that in control hippocampus (p<0.01, Fig. 6b).

Tau and mitochondrial axonal transport in primary hippocampal neurons

The relationship between tau protein and mitochondrial movement was examined by visualizing the dynamic behaviour of mitochondria in cultured primary hippocampal neurons. First, we checked the relationship between mitochondrial movement and tau protein level. The neurons without overexpressing tau contained numerous mobile mitochondria tagged with green fluorescent protein moving rapidly in the soma, or moving rapidly in the axon (Fig. 7a, Supplementary video 1, available online). In contrast, the mitochondria in neurons with overexpressing tau were mostly stationary in soma and axons during 5-min imaging (Fig. 7b, Supplementary video 2, online). The results showed that overexpressing tau blocked traffic of mitochondria. Next, we investigated the relationship between tau phosphorylation and mitochondrial movement. Because tau protein was hyperphosphorylated in cultured primary hippocampal neurons (Fig. 8a, lane 2 vs. lane 1), and lithium is a typical inhibitor for tau protein kinase, Gsk3β, we chose to decrease phosphorylated tau with LiCl (Fig. 8a, lanes 2–5), and then visualized the change of mitochondrial movement. Analyses showed that after 2 h treatment with LiCl, both anterogradely moved mitochondria (p<0.01, Fig. 8b) and retrogradely moved mitochondria (p<0.05, Fig. 8b) were significantly inhibited during 5-min imaging, the net result was that the number of mitochondria per 25 µm axons (p<0.05, Fig. 8c) decreased significantly, most mitochondria became stationary or gradually moved in soma and axons (Fig. 8e, Supplementary video 4, online) compared to the same neurons before LiCl treatment (Fig. 8d, Supplementary video 3, online). This result showed that mitochondrial transport was decreased when hyperphosphorylated tau was inhibited by lithium which was consistent with in-vivo results.

Discussion

The major finding of this study is that higher levels of phosphorylated tau exist in the hippocampus of a chronic stress model with a hyperactive HPA axis, which induces more mitochondrial movement from cytoplasm to synapse. Consistent with the in-vivo model, when hyperphosphorylated tau was inhibited by lithium, the mitochondrial transport was decreased in cultured primary hippocampal neurons. CRH-R1 antagonist treatment, not glucocorticoid receptor antagonist treatment, decreased tau phosphorylation and mitochondrial levels in synaptosomes in the hippocampus of model mice.

Trauma early in life increases the vulnerability of the organism to the pathogenic effects of stress, and is a potent risk factor for developing depression in adulthood (Chapman et al.2004; Edwards et al.2003; Heim et al.2008). Early-life stress in humans is associated with persistent sensitization of the HPA axis to mild stress in adulthood, thereby contributing to vulnerability to depression (Heim et al.2000; Tyrka et al.2009). In the present study, we created a chronic stress model by neonatal isolation before weaning, followed by chronic mild stress by social isolation after weaning, which mimicked a life history of early stress combined with a life event in adulthood (Avital & Richter-Levin, 2005). The chronic stress model exhibited typical anxiety/depression-like behaviour: decreased exploratory activity and anhedonia, and also possessed a hyperactive HPA axis.

It has been reported that acute stresses can induce tau phosphorylation in the rodent CNS (Rissman et al.2007). Our results showed that a higher level of tau phosphorylation also exists in the hippocampus in a chronic stress model based on neonatal stress combined with chronic stress in adulthood, which has a hyperactive HPA axis. Two possible pathways were examined for the hyperphosphorylated tau in the hippocampus of model mice. First, the pathway by the feedback molecule of the HPA axis showed increased corticosterone could induce tau pathology (Green et al.2006). Second, the direct effect of CRH derived from PVN (Koegler-Muly et al.1993). Our results showed that the CRH-R1 antagonist CP154,526 significantly decreased tau phosphorylation; however, the glucocorticoid receptor antagonist RU486 had no effect on tau phosphorylation in the hippocampus of model mice. The results suggested that hyperphosphorylated tau in the hippocampus of the chronic stress model resulted from the direct central effect of CRH.

As microtubule-associated proteins, tau proteins are implicated in regulating axonal transport of mitochondria. When binding to microtubules, tau proteins influence the rate of attachment and detachment of motors from microtubules with the net result that dynein-mediated movements (towards the cell centre) become predominant (Trinczek et al.1999). Live imaging of mitochondria in cultured primary hippocampal neurons showed that overexpressing tau blocks axonal transport of mitochondria, which indicates that overexpressing tau induced more tau proteins to bind to microtubules and hence inhibited axonal transport of mitochondria into the synapse. Phosphorylation of tau in microtubule-binding domains detaches tau from microtubules and therefore facilitates organelle transport (Tatebayashi et al.2004). Our results show that hyperphosphorylated tau was accompanied by the higher level of mitochondria in hippocampal synaptosomes in the chronic stress model. Moreover, when hyperphosphorylated tau was inhibited by lithium, the mitochondrial transport monitored by live imaging was decreased in cultured primary hippocampal neurons, the number of mitochondria in axons decreased significantly, which further confirmed the relationship between tau phosphorylation and axonal transport of mitochondria.

Mitochondrion is not only an important organelle for energy supply and signalling, but also closely related to major depression (Konradi et al.2004; Rezin et al.2009). Several works have demonstrated that metabolism is impaired in some animal models of depression, induced by chronic stress, especially the activities of the complexes of mitochondrial respiratory chains (Gardner et al.2003; Madrigal et al.2001). Moreover, monoamine oxidases (MAOs) A and B are located in outer membrane of mitochondria, and are critically important for the proper functioning of synaptic neurotransmission and for the regulation of emotional behaviours and other brain functions. An elevated MAOA in the brain of a major depression patient has been reported (Meyer et al.2006). Our results showed that the mitochondrial density within the synapse was increased which might thus lead to greater MAO density. Moreover, lithium is the most established and commonly used mood stabilizer (Burgess et al.2001), but the underlying mechanisms are not fully understood. Our live-imaging results showed that lithium decreased tau phosphorylation, and further inhibited axonal transport of mitochondria, which might decrease the MAOA density within the synapse and hence increase the monoamine level in the synaptic cleft (Belmaker & Agam, 2008).

Taken together, we show here for the first time that hyperphosphorylated tau in the hippocampus in a chronic stress mice model, was accompanied by increased mitochondrial transport. Increased CRH mediated by CRH-R1, rather than higher corticosterone mediated by glucocorticoid receptor, was involved in hippocampal tau phosphorylation under chronic stress condition, which suggested that centrally derived CRH may play an important role in the process of mitochondrial axonal transport in the hippocampus in a chronic stress mice model.

Note

Supplementary material accompanies this paper on the Journal's website.

Acknowledgments

This project was supported by the Natural Science Foundation of China (81030026).

Statement of Interest

None.

References