J. Neurochem. (2010) 114, 1840–1851.
Neurogenesis in the hippocampus occurs throughout life in a wide range of species and could be associated with hippocampus-dependent learning and memory. Stress is well established to seriously perturb physiological/psychological homeostasis and affect hippocampal function. In the present study, to investigate the effect of chronic restraint stress in early life on hippocampal neurogenesis and hippocampus-dependent memory, 3-week-old mice were subjected to restraint stress 6 days a week for 4 weeks. The chronic restraint stress significantly decreased the hippocampal volume by 6.3% and impaired hippocampal neurogenesis as indicated by the reduced number of Ki67-, 5-bromo-2′-deoxyuridine- and doublecortin-positive cells in the dentate gyrus. The chronic restraint stress severely impaired hippocampus-dependent contextual fear memory without affecting hippocampus-independent fear memory. The expression level of brain-specific transcription factor neuronal PAS domain protein 4 (Npas4) mRNA in the hippocampus was down-regulated by the restraint stress or by acute corticosterone treatment. Npas4 immunoreactivity was detected in progenitors, immature and mature neurons of the dentate gyrus in control and stressed mice. Our findings suggest that the chronic restraint stress decreases hippocampal neurogenesis, leading to an impairment of hippocampus-dependent fear memory in mice. Corticosterone-induced down-regulation of Npas4 expression may play a role in stress-induced impairment of hippocampal function.
brain-derived neurotrophic factor
granule cell layer
glial fibrillary acidic protein
neuronal PAS domain protein 4
SRY-related HMG box 2
Stress is defined in biological systems as any condition that seriously perturbs physiological/psychological homeostasis and well known to affect the function and morphology of the hippocampus (Kim and Diamond 2002). The exact underlying cellular mechanisms that mediate the inhibitory effect of stress are largely unknown. However, stress reduces the expression of several growth factors and neurotrophic factors, such as brain-derived neurotrophic factor (BDNF), insulin-like growth factor-1, nerve growth factor, epidermal growth factor, and vascular endothelial growth factor, that can all influence neurogenesis (Lucassen et al. 2010).
A variety of stress paradigms also impair hippocampal neurogenesis (Mirescu and Gould 2006). Neurogenesis in the hippocampus occurs throughout life in a wide range of species (Kempermann et al. 1997; Eriksson et al. 1998). The newly divided cells in the subgranular zone (SGZ) of the dentate gyrus (DG) in the hippocampus migrate into the granule cell layer (GCL) of the DG and are integrated into the existing circuitry (Tashiro et al. 2006). Impairment of hippocampal neurogenesis could be one of the etiological factors for neuropsychiatric disorders (Reif et al. 2006), and hippocampal neurogenesis may be a target for the treatment of depression (Airan et al. 2007).
Recently, we have demonstrated that rearing in long-term social isolation after weaning impaired hippocampal neurogenesis and spatial memory in Morris water maze, and reduced expression levels of neuronal PAS domain protein 4 (Npas4) mRNA in the DG of the hippocampus (Ibi et al. 2008). Npas4 is a brain-specific transcription factor and may have a neuroprotective function (Ooe et al. 2009). It has been reported that Npas4 regulates the expression of drebrin, which engages in dendritic-cytoskeleton modulation at synapses in the hippocampus (Ooe et al. 2004). Recently, Npas4 has been shown to control GABAergic synapse development through the transcriptional regulation of BDNF (Lin et al. 2008). However, it remains to be determined whether Npas4 gene expression in the hippocampus is affected by deleterious stress. In this study, we investigated the effect of chronic restraint stress in early life on hippocampal neurogenesis, hippocampus-dependent memory and Npas4 expression in mice.
Materials and methods
Male ICR mice (3 weeks old) were purchased from Japan SLC Inc. (Hamamatsu, Japan). They were housed 5–6 per cage under standard conditions (23 ± 1°C, 50 ± 5% humidity), with a 12-h light/dark cycle. Food and water were available ad libitum. The animals were handled in accordance with the guidelines established by the Institutional Animal Care and Use Committee of Nagoya University.
Restraint stress procedure
Mice were randomly divided into two groups, restraint stress and control groups. Chronic stress was applied for 4 weeks with a stainless mesh that allowed for a close fit to mice (6 h/day between 10 am and 4 pm, 6 days a week) as described previously (Takuma et al. 2007). For acute restraint stress, mice were subjected to the stress for 2 h or 6 h.
Corticosterone level measurement
Corticosterone levels were measured with an enzymatic immunoassay kit (Cayman, Ann Arbor, MI, USA) following the manufacturer’s manual with modifications. Briefly, mice were decapitated and 1 mL of blood was collected with 30 μL of 100 mM EDTA immediately after the final restraint stress. Samples were centrifuged (15 min, 1000 g, 4°C). Each supernatant was added to a 96-well plate at two dilutions (1 : 150 and 1 : 300) in duplicate and subjected to the immunoassay. The optical density of the enzyme products was read at 405 nm.
The conditioned-fear test was carried out as described previously (Nagai et al. 2003). One day after the chronic stress, freezing response was measured in a neutral cage (30 cm × 30 cm × 35 cm high) for 1 min without sound (pre-conditioning phase). In the conditioning phase, each mouse was placed in a training cage (25 cm × 30 cm × 11 cm high) and allowed to explore freely for 2 min, and then a 15-s tone (85 dB) was delivered (conditioned stimulus). During the last 5 s of the tone stimulus, a foot shock of 0.8 mA was presented as an unconditioned stimulus through a shock generator (four times with 15-s intervals.). Context-dependent test was carried out 24 h after the conditioning and the tone-dependent test was conducted 3 h after the context-dependent test. For the context-dependent test, freezing behavior was measured in the training cage for 2 min without tone or foot shock presentation. For the tone-dependent test, the freezing behavior was measured in the neutral cage for 1 min in the presence of a continuous-tone stimulus. After the test phase, the sensitivity to the foot shock was determined. The foot shock was gradually increased in 0.05 mA increments until vocalization.
Coronal sections (10 μm) through the entire extent of the hippocampus were cut and every fifteenth section was collected. The sections were dehydrated through an ascending series of ethanol concentrations. After dissolving lipids in xylene, they were rehydrated through a series of graded ethanol solutions. They were stained with 1% cresyl violet, differentiated in 70% ethanol containing anhydrate acetic acids and then dehydrated with 95% and 100% ethanol. Samples were observed with a microscope (Model Axioskop, Zeiss, Jena, Germany). To estimate total pyramidal and granule cell numbers, photographs (12 per section) were randomly taken from the hippocampus at a magnification of ×400, and the stained cells were counted within an 80 × 80 μm dissector frame using image-analyzing software Win ROOF (ver. 5.6, Mitani Co., Fukui, Japan). To estimate the volume of the entire hippocampus, the Cavalieri principle was used (Reilly et al. 2003). The hippocampal structure was outlined and the computed areas were then summed and multiplied with the thickness and with the intersection distance.
To measure the effect of chronic restraint stress on survival and differentiation of newly divided cells in the DG, 5-bromo-2′-deoxyuridine (BrdU) labeling was carried out 1 day before starting the chronic restraint stress. BrdU was purchased from Sigma-Aldrich (St Louis, MO, USA) and dissolved in saline. BrdU (75 mg/kg) was injected intraperitoneally (i.p.) three times at 2-h intervals as described previously (Ibi et al. 2008).
After stress, mice were deeply anesthetized with diethyl ether, and transcardially perfused with ice-cold saline followed by 4% paraformaldehyde in 0.1 M phosphate-buffered saline (PBS, pH 7.4). The brain was post-fixed in the same fixative and cryoprotected in 0.1 M PBS (30% sucrose). The brain was embedded in Tissue-Tek O.C.T. compound (Sakura Finetech, Tokyo, Japan) and stored at −80°C. Every fifth coronal section (30 μm) was collected between stereotaxic coordinates bregma −1.2 to 3.0 according to the brain atlas (Paxinos and Franklin 2004). For BrdU single-staining, sections were treated with 0.1% Nonidet-40/0.01 M PBS (pH 7.4) overnight (4°C) and denatured in a microwave oven in 0.01 M citrate buffer (pH 6.0). After blocking in 10% goat serum/PBS with 0.1% Nonidet-40 (Sigma-Aldrich), BrdU-positive cells in the sections were detected using BrdU labeling and detection kit 2 (Roche Diagnotics GmbH, Mannheim, Germany) according to the manufacturer’s instructions. For double-staining of BrdU/neuronal nuclei (NeuN) and BrdU/glial fibrillary acidic protein (GFAP), sections were pre-treated with 1 M HCl, followed by 0.1 M borate buffer and then washed in PBS before blocking. Rat anti-BrdU antibody (1 : 200; Abcam, Cambridge, UK), mouse anti-NeuN antibody (1 : 100; Millipore, Billerica, MA, USA) and mouse anti-GFAP antibody (1 : 1500; Sigma-Aldrich) diluted in PBS (0.1% Triton X-100 and 5% goat serum) were applied to sections [overnight, 4°C and 6 h, room temperature (RT)]. After washing, goat anti-rat Alexa 568 and goat anti-mouse Alexa 488 (1 : 1000; Invitrogen, Eugene, OR, USA) antibodies were applied to sections (2 h, 25°C). For Ki67 or doublecortin (DCX) single-staining, sections were treated with 0.5% hydrogen peroxide in methanol to block endogenous peroxidase activity, and denatured in the microwave oven. They were incubated in blocking solution (10% goat serum and 0.1% Triton X-100 in 0.01 M PBS, 1 h), and then with rabbit anti-Ki67 (1 : 2000; Novocastra, Newcastle, UK) or rabbit anti-DCX (1 : 500; Abcam) antibody (overnight, 4°C). They were washed and incubated with biotinylated goat anti-rabbit antibody (1 : 200; Vector Laboratories, Burlingame, CA, USA, 1 h, 25°C). The sections were washed and processed with avidin-biotinylated horseradish peroxidase complex (Vector ABC kit, Vector Laboratories), and the reaction was visualized using diaminobenzidine. For double-staining of Npas4/several cell markers, coronal sections (10 μm) were incubated with blocking solution (3% bovine serum albumin, 10% donkey serum and 0.3% Triton X-100 in 0.01 M PBS, 1 h, 25°C), and then with primary antibodies (overnight, 4°C). They were washed with 0.01 M PBS and incubated with secondary antibodies (2 h, 25°C). Sections were mounted in fluorescent medium (S3023; Dako Cytomation, Kyoto, Japan). The following antibodies were used: rabbit anti-Npas4 (1 : 32), mouse anti-NeuN (1 : 200; Millipore), mouse anti-Calbindin D-28 (1 : 500; Swant, Bellinzona, Switzerland), goat anti-DCX (1 : 50; Santa Cruz Biotechnology, Santa Cruz, CA, USA), goat anti-Sox-2 (1 : 100; Santa Cruz Biotechnology), Alexa 488-conjugated donkey anti-rabbit (1 : 500, Invitrogen), Alexa 594-conjugated goat anti-mouse (1 : 500, Invitrogen) and Alexa 594-conjugated donkey anti-goat (1 : 500, Invitrogen). The rabbit anti-Npas4 antibody (Japan Bio Services Co., Ltd., Saitama, Japan) was raised against a carboxy-terminal region of Npas4 (amino acids 622–635). For pre-absorption experiment, Npas4 antibody was incubated with a 5-fold excess of the antigen peptide (overnight, 4°C).
Quantification of immunostaining cells
Every fifth section throughout the hippocampus (total 12 sections from each mouse) was processed for BrdU, Ki67 and DCX immunohistochemistry. All immunostained cells in DG were examined using a light microscope (Axio Imager; Zeiss) and counted by a blind experimenter. All counts were performed at 400× magnification (objective; 40×). To obtain the total number of cells per DG, we multiplied the counted number of positive cells by five. Double-stained cells were imaged and quantified using a confocal laser scanning microscope (LSM 510; Zeiss). Each cell was analyzed along the entire ‘z’ axis. Ratios of BrdU-positive cells co-labeled with NeuN or GFAP among BrdU-positive cells were determined.
Quantitative analyses of Npas4 mRNA by real-time RT-PCR
Mice were decapitated and brains were removed immediately after restraint stress or 4 h after corticosterone injection (10 mg/kg, s.c., in corn oil, Sigma-Aldrich). For DG, tissues from two animals were pooled as one sample and for hippocampus tissue from one animal used as one sample. The total RNA isolated from the DG or hippocampus (RNeasy Mini Kit, Qiagen, Hilden, Germany) was converted into cDNA using SuperScript III First-Strand Synthesis System for RT-PCR (Invitrogen) and quantified using a 7300 real-time PCR System (Applied Biosystems, Foster City, CA, USA). Quantitative real-time RT-PCR was performed in a volume of 25 μL with 0.5–1.0 μg of cDNA and 0.5 μM primers in the Power SYBR Green Master Mix (Applied Biosystems). The primers used were as follows: 5′-AGCATTCCAGGCTCATCTGAA-3′ (forward) and 5′-GGCGAAGTAAGTCTTGGTAGGATT-3′ (reverse) for Npas4, 5′-TGTCAAGCTCATTTCCTGGTATGA-3′ (forward) and 5′-CTTACTCCTTGGAGGCCATGTAG -3′ (reverse) for glyceraldehyde-3-phosphate dehydrogenase, 5′-CGATGCCCTGAGGCTCTTT-3′ (forward) and 5′-TGGATGCCACAGGATTCCA -3′ (reverse) for β-actin used as internal controls.
In situ hybridization
Npas4 mRNA level in tissue was detected with an in situ hybridization kit (Nippongene, Tokyo, Japan) following the manufacturer’s manual with modifications. Briefly, immediately after restraint stress, mice were deeply anesthetized with diethyl ether, and transcardially perfused with ice-cold saline followed by 4% paraformaldehyde in 0.1 M PBS (pH 7.4). The brains were removed, post-fixed in the same fixative and then cryoprotected in 12%, 15% and 18% sucrose in Hanks’ Balanced Salt Solution (HBSS) buffer (pH 7.4). The brains were embedded in Tissue-Tek O.C.T. compound (Sakura Finetech) and stored at −80°C. Thick coronal brain sections of 10 μm were cut and mounted. Nucleotides 1485–2409 of Npas4 (BC_129861) were used as a template for transcription and labeling of Npas4 antisense and sense with digoxigenin-UTP (Roche Diagnotics GmbH). Sections were hybridized with probes and mRNA signals were visualized by enzyme-catalyzed color reaction with an NBT/BCIP kit (Roche Diagnotics GmbH, 16 h, 25°C) in a dark room.
Statistical analyses were performed using SigmaStat 3.1 software (Systat Software, Inc., Chicago, IL, USA). For body weight gain, differences among groups were analyzed by repeated measures anova. Differences among groups were analyzed by Student’s t-test or one-way anova, followed by Bonferroni’s post hoc test.
Effect of chronic restraint stress on body weight gain and blood corticosterone level, hippocampal volume and the number of Nissl-stained cells in the hippocampus
To explore whether the mice that received 4-week chronic restraint stress revealed biologically significant features, we measured stress responses during and after the stress. During the period of 4 weeks of stress, chronic restraint stress significantly inhibited the body weight gain [0 day, control group, 19.5 ± 0.1 g; stress group, 19.7 ± 0.1 g; 24th day, control group, 40.1 ± 0.4 g; stress group, 31.9 ± 0.7 g, treatment, F (1, 21) = 165.719, p < 0.001; days, F (8, 168) = 717.763, p < 0.001; treatment and day interaction, F (8, 168) = 41.147, p < 0.001, Fig. 1a]. Blood corticosterone level after the final restraint stress was significantly increased in stressed mice compared with that of the control group (p < 0.05, Fig. 1b). To determine whether chronic restraint stress influences hippocampal volume and the number of hippocampal cells, we performed Nissl staining. Chronic restraint stress significantly decreased the hippocampal volume by 6.3% (p < 0.05), but there was no significant difference in the number of Nissl-stained cells in the hippocampus (Table 1, Figure S1).
|Group||Hippocampal volume (mm3)||Nissl-stained cells (×102 counts/mm2)|
|Control||6.98 ± 0.17||42.01 ± 0.53||40.64 ± 1.42||119.96 ± 1.24|
|Stress||6.54 ± 0.11*||41.50 ± 1.03||37.64 ± 1.19||115.55 ± 2.1|
Effect of chronic restraint stress on associative fear memory
Mice were subjected to the conditioned fear memory test to examine the function of hippocampus after chronic restraint stress for 4 weeks. Contextual fear memory requires both amygdala and hippocampus (Phillips and LeDoux 1992; Anagnostaras et al. 1999), whereas tone-dependent memory requires the amygdala but not the hippocampus. In the context-dependent test, mice were returned to the conditioning box, and then their freezing behavior was analyzed for 2 min. The mice subjected to chronic restraint stress exhibited significantly less freezing behavior than control mice (p < 0.05, Fig. 1c). On the other hand, there was no significant difference in freezing time between control and stressed mice when the freezing behavior was analyzed in the neutral cage in the presence of a continuous-tone stimulus identical to the conditioned stimulus (Fig. 1d). Mice were tested for foot shock threshold to be sure that any alterations in memory were not because of changes in sensitivity to foot shock. There was no significant difference in the pain threshold of the chronic stress mice (0.27 ± 0.02 mA, n = 16) compared with that of the control mice (0.29 ± 0.02 mA, n = 16).
Effect of chronic restraint stress on hippocampal neurogenesis
As recent studies have suggested that neurogenesis plays an important role in hippocampus-dependent memory (Leuner et al. 2006; Imayoshi et al. 2008; Kitamura et al. 2009) and chronic restraint stress exposure led to memory deficit in a hippocampus-dependent manner, we examined the effect of chronic restraint stress on cell proliferation in the DG of the hippocampus. Ki67-positive cells, for which the antigen is expressed in all active parts of the cell cycle, G1, S, G2 and M (mitosis) phases but not G0 phase (Braun et al. 1988), were mainly observed in the SGZ rather than other areas as either clusters or single cells (Fig. 2a). Chronic restraint stress significantly decreased the number of Ki67-positive cells in the SGZ by 17% (p < 0.05) and thereby reduced that in the DG by 16% (p < 0.05, Table 2).
|Control group||Stress group|
|Ki67||2628.3 ± 77.0||177.5 ± 19.0||259.2 ± 49.8||2191.7 ± 64.4||2200.0 ± 150.6*||129.2 ± 16.1||252.5 ± 28.4||1818.3 ± 136.1*|
|BrdU||1241.0 ± 107.9||134.8 ± 12.3||792.8 ± 74.5||313.5 ± 33.7||895.2 ± 49.7*||118.1 ± 9.7||533.9 ± 43.8**||243.2 ± 26.2|
|DCX||11 555.1 ± 479.2||33.5 ± 11.9||1271.7 ± 173.3||10 249.9 ± 309.8||9993.3 ± 327.0*||36.7 ± 10.7||1090.8 ± 170.1||8865.8 ± 343.0*|
5-Bromo-2′-deoxyuridine is incorporated into the DNA of proliferating cells during the S phase of the cell cycle (Rakic 2002) and has been utilized in a number of in vitro and in vivo studies to label neural progenitors. To measure the effect of chronic restraint stress on the survival of newly divided cells in the DG, BrdU was injected into mice 1 day before the chronic restraint stress was initiated, and the numbers of BrdU-positive cells were counted after 4 weeks of restraint stress. There was an apparent difference between control and stressed mice in the number and location of BrdU-positive cells in the DG of hippocampus (Fig. 2b). Chronic restraint stress significantly decreased the number of BrdU-positive cells in the GCL by 27% (p < 0.01) and thereby decreased that in the DG by 32% (p < 0.05, Table 2). DCX is one of the immature granule neuron markers and is useful to assess the rate of neurogenesis in the adult mammalian hippocampus (Liu et al. 2008). Chronic restraint stress significantly decreased the number of DCX-positive cells in the SGZ by 13% (p < 0.05) and thereby decreased that in the DG by 13% (p < 0.05, Fig. 2c, Table 2).
Finally, we investigated the effect of chronic restraint stress on the differentiation of newly divided cells. Mice were subjected to 4-week chronic restraint stress after BrdU injection, and then we measured NeuN-positive cells (neuron) and GFAP-positive cells (astrocyte) among BrdU-labeled cells in the DG of the hippocampus. There was no apparent difference in the distribution and the rate of NeuN- (Fig. 3a and b) or GFAP-positive cells (Fig. 3a and c) among BrdU-labeled cells in the DG of the hippocampus.
Effect of restraint stress on the expression levels of Npas4 mRNA in the dentate gyrus
To address whether restraint stress reduces Npas4 expressions in the DG of the hippocampus, mice were subjected to acute (2 h or 6 h) or chronic (6 h/day, 6 days a week for 4 weeks) restraint stress. The expression levels of Npas4 mRNA in the DG were analyzed by real-time RT-PCR. Both acute (2 h and 6 h) and chronic restraint stresses significantly decreased the levels of Npas4 mRNA compared with those in control mice [F (3, 16) = 8.718, p = 0.01, Fig. 4a]. In situ hybridization analysis revealed that intense signals that react with antisense Npas4 probes were detected in the pyramidal and granule cell layer in the hippocampus. There was no signal detected when sense probes were used, suggesting that Npas4 mRNA is abundant in the hippocampus (Fig. 5a). Chronic restraint stress markedly decreased the signals of Npas4 mRNA in the DG, CA3 and CA1 subregions of the hippocampus compared with those in control mice (Fig. 5b).
Effect of corticosterone treatment on the expression levels of Npas4 mRNA in the hippocampus
To address the relationship among stress exposure, reduction of Npas4 expression and the stress-induced impairment of neurogenesis and hippocampus-dependent fear memory, we examined the effect of corticosterone injection in mice on Npas4 mRNA levels in the hippocampus. The Npas4 mRNA levels in the hippocampus of mice following corticosterone injection were significantly reduced to 70.0% (p < 0.05) that of vehicle-treated control mice (Fig. 4b).
Phenotype of Npas4-positive cells in the dentate gyrus of the hippocampus
Finally, to characterize the Npas4-expressing cells in the DG, we performed double immunostaining for Npas4 and several cell markers such as SRY-related HMG box 2 (Sox-2), DCX, calbindin and NeuN. Antibody pre-absorption with the antigen peptide completely blocked the Npas4 staining in the DG (Figure S2). Sox-2 is a marker for both quiescent neural progenitors and amplifying neural progenitors (Segi-Nishida et al. 2008). DCX is a marker for immature neurons, while calbindin and NeuN are mature neural markers in the DG. At 7 weeks old, most Npas4 immunoreactivities were co-localized with the cells positive for NeuN in the GCL of control mice (Figs 6a and S2), while some of them partially co-localized with the cells positive for Sox-2 (Figs 6b and S2), DCX (Figs 6c and S2) in the SGZ and calbindin in the GCL (Figs 6d and S2). Similar results were observed in the stressed mice (Figs 6e–h and S2). Immunoblot studies revealed that the protein expression level of Npas4 in DG was not changed significantly in stressed mice (data not shown).
We demonstrated in the present study that chronic restraint stress after weaning decreased cell proliferation, survival of newly divided cells and neurogenesis in the DG. Numerous studies have reported that stress disrupts hippocampal neurogenesis. Cell proliferation in the DG is decreased by a variety of stress paradigms, including restraint stress (Nagata et al. 2009; Veena et al. 2009), subordination stress (Yap et al. 2006), social isolation (Dong et al. 2004), resident-intruder stress (Gould et al. 1998), water immersion restraint stress (Tamaki et al. 2008), chronic unpredictable stress (Heine et al. 2004), predator odor (Mirescu et al. 2004), sleep deprivation (Mirescu et al. 2006) and chronic mild stress (Alonso et al. 2004). Otherwise, the effect of stress on survival and differentiation of newly divided cells is more intricate than cell proliferation. Several studies have shown that the decrease of cell proliferation in the DG results in the suppression of neuronal generation (Pham et al. 2003; Westenbroek et al. 2004). Other reports have indicated that the influence of stress on hippocampal neurogenesis is very short-lived, and the decrease in the number of neuronal precursor cells is followed by an enhancement of cell survival, as the total number of new mature neurons appears unchanged (Tanapat et al. 2001; Malberg and Duman 2003). On the other hand, chronic stress after BrdU labeling decreases the survival of newly divided cells (Lee et al. 2006; Ibi et al. 2008). Our current findings are generally consistent with previous studies. Indeed, the stress exposure paradigm in this study was sufficient to increase serum corticosterone level in stressed mice compared with that in control mice, which may affect the fate of newly generated neurons in the DG. Meanwhile, chronic restraint stress had no apparent effect on the distribution and the rate of NeuN- or GFAP-positive cells among BrdU-labeled cells in the DG of the hippocampus. These results suggest that differentiation of newly divided cells in the hippocampus may be unaffected by chronic restraint stress after weaning in mice; however, we should also investigate the effect of chronic restraint stress on microglial cells.
We also found that chronic restraint stress significantly decreased the hippocampal volume without affecting the number of Nissl-stained cells. It has been reported that chronic restraint stress decreased the length and branch of apical dendrites in the CA3 subregion (Watanabe et al. 1992; Magariños and McEwen 1995). Therefore, atrophy of the apical dendrites of CA3 pyramidal neurons and the decrease of neurogenesis in the DG may be related to the reduction of hippocampal volume.
As regards the functional changes in the hippocampus, we demonstrated that chronic restraint stress impaired context-dependent fear memory in mice. Even though there is a report that chronic restraint stress enhanced fear memory (Conrad et al. 1999), it is generally accepted that stress impairs memory in rodents and the increased levels of circulating glucocorticoids inhibit memory in rodents (Lemaire et al. 2000; Montaron et al. 2006). The context-dependent fear memory is hippocampus-dependent (Phillips and LeDoux 1992) and the hippocampus is involved in acquisition and transient storage of context-dependent fear memory (Phillips and LeDoux 1992; Anagnostaras et al. 1999). Although a few studies have reported that contextual fear memory was not influenced by hippocampal neurogenesis (Shors et al. 2001; Zhang et al. 2008), irradiated rats or mice, ganciclovir-treated GFAP-tk mice and tamoxifen-treated Nes-CreERT2/NSE-DTA mice, all of which show defects of neurogenesis in the hippocampus, are defective in the conditioned fear memory test (Winocur et al. 2006; Imayoshi et al. 2008). In contrast to the context-dependent memory, tone-dependent fear memory was not significantly affected in chronic restrained mice. It is reported that tone-dependent fear memory is hippocampus-independent, but the lateral nucleus of the amygdala links the auditory conditioned stimulus with the fear response via the thalamo-amygdala or thalamo-cortico-amygdala pathway (LeDoux et al. 1990; Romanski and LeDoux 1992). Taken together, our results suggest that chronic restraint stress in mice, which leads to the hypersecretion of glucocorticoids, may disrupt the hippocampal function as evidenced by the impairment of context-dependent fear conditioned memory, which is associated with the inhibition of neurogenesis in the hippocampus and reduced hippocampal volume.
To clarify the putative molecular candidates for the stress-induced memory impairment and morphological changes in the hippocampus, we examined whether the expression level of Npas4 was regulated by stress and circulating corticosterone level. Acute or chronic restraint stress exposure as well as acute corticosterone injection significantly reduced the mRNA levels of Npas4 in the DG and the hippocampus, respectively, suggesting that stress may decrease Npas4 expression in the hippocampus through the action of circulating corticosterone level. It has been reported that Npas4 mRNAs are highly expressed in the hippocampus (Moser et al. 2004). Npas4 has constitutive or developmental functions that may be critical for regulating the transcriptional control of limbic patterning and function (Moser et al. 2004). Npas4 plays a role in the development of inhibitory synapses by regulating the expression of activity-dependent genes and appears to regulate a wide variety of genes modulating synaptic functions (Lin et al. 2008). Furthermore, we observed that Npas4 mRNA is present in neural stem/progenitor cells derived from the hippocampus of embryonic mice as originally described by Yoneyama et al. (2007, 2009) (data not shown).
Interestingly, BDNF is a target gene of Npas4 in mouse hippocampal primary cultures (Lin et al. 2008), and the expression level of BDNF is reported to be decreased by stress in the hippocampus (Vaidya et al. 1999; Xu et al. 2006). It is well known that BDNF positively regulates neurogenesis and plays a role in the differentiation and survival of neuronal progenitor cells (Sairanen et al. 2005; Scharfman et al. 2005). We assume that the inhibition of Npas4 gene expression in both progenitor cells and mature neurons may contribute to the decreased action of BDNF in the hippocampus. Taken together, chronic stress may reduce Npas4 expression in the DG of hippocampus through the action of corticosterone, leading to an impairment of hippocampal neurogenesis and hippocampus-dependent memory in mice. To further support the involvement of Npas4 in stress-induced impairment of neurogenesis, we showed that Npas4 was expressed in some of the Sox-2-positive cells, and most DCX- and calbindin-positive cells were colocalized with Npas4 in the hippocampus. These findings suggest that Npas4 is expressed from an early developmental stage of newly divided cells in the DG and that the down-regulation of Npas4 could be related to the impairment of neurogenesis by restraint stress.
Involvement of Npas4 in the hippocampal dysfunction through the impairment of neurogenesis induced by stress is also possible because neurogenesis modulates the hippocampus-dependent fear memory (Kitamura et al. 2009), in addition, newborn cells in adult DG may extend axons into the CA3 region, which release glutamates as their main neurotransmitter and form functional synapses with target cells (Toni et al. 2008). We revealed by in situ hybridization that stress decreased the mRNA expression level of Npas4 in DG, CA3 and CA1 subregions of hippocampus. Although we do not have any definitive evidence that the down-regulation of Npas4 mRNA expression level in CA3 and CA1 regions contributes to the reduced volume of hippocampus and impaired fear conditioned memory, Npas4 may play a role in synaptic and structural plasticity as well as hippocampal function in CA3 and CA1 regions because this transcription factor has been shown to regulate GABAergic synapse development in an activity-dependent manner (Lin et al. 2008) and to possibly play a role in dendritic-cytoskeleton modulation at synapses in the hippocampus (Ooe et al. 2004, 2009).
This study was supported in part by Grants-in-Aid for Scientific Research (Nos. 19390062 and 22390046) from the JSPS, the Global COE program from the Ministry of Education, Culture, Sports, Science and Technology of Japan, the Academic Frontier Project for Private Universities; matching fund subsidy from MEXT, 2007–2011, the Research on Risk of Chemical Substances, Health and Labour Science Research Grants supported by Ministry of Health, Labour and Welfare, Takeda Science Foundation, AstraZeneca Research Grant 2008, and JST, CREST.
Disclosure/conflicts of interest
The authors declare that there is no conflict of interest in the publication of this work.