SEARCH

SEARCH BY CITATION

Keywords:

  • glucocorticoid;
  • glucocorticoid receptor;
  • Npas4;
  • stress;
  • transcription

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Neuronal PAS domain 4 (NPAS4), a brain-specific helix–loop–helix transcription factor, has recently been shown to regulate the development of GABAergic inhibitory neurons. We previously reported that Npas4 mRNA expression levels were reduced in the hippocampus of mice exposed to social isolation or restraint stress, which was accompanied by impairment of memory, emotional behavior, and hippocampal neurogenesis. Therefore, the reduction of NPAS4 expression may play a role in stress-induced brain dysfunction. In this study, to investigate the transcriptional regulation of Npas4 by stress, we focused on the effect of glucocorticoids (GCs) upon Npas4 transcription. Corticosterone treatment reduced Npas4 expression in the frontal cortex and hippocampus, whereas adrenalectomy caused an increase in expression. GC receptor (GR) antagonist, mifepristone, inhibited the stress-induced reduction of Npas4 expression. Putative negative glucocorticoid response elements (GREs) were found −2000 to −1000 upstream of the Npas4 transcription initiation site. Npas4 promoter activity was increased by mifepristone or by mutation of the negative GRE sequences. A chromatin immunoprecipitation assay revealed that restraint stress increased the binding of GR to Npas4 promoter region in the hippocampus. These results suggest that transcription of Npas4 is down-regulated by stress via the binding of agonist-bound GR to its promoter.

Abbreviations used
ADX

adrenalectomy

BDNF

brain-derived neurotrophic factor

ChIP

chromatin immunoprecipitation

CS

corticosterone

GCs

glucocorticoids

GREs

glucocorticoid response elements

GR

glucocorticoid receptor

ICR

imprinting control region

NPAS4

neuronal PAS domain 4

Stress plays a significant role in the development of various neuropsychiatric disorders, for example, anxiety disorder, major depressive disorder, and bipolar disorder (de Kloet et al. 2005). Previous studies suggested that stress impairs adult neurogenesis in the hippocampus (Czeh et al. 2002; Joels et al. 2007), and that the impairment of neurogenesis is involved in the development and expression of neuropsychiatric disorders (Jacobs et al. 2000; Reif et al. 2006; Maeda et al. 2007). However, the underlying molecular mechanisms are not well understood.

Chronic social isolation (Wongwitdecha and Marsden 1996; Weiss et al. 2004; Ibi et al. 2008; Koike et al. 2009) induces the impairment of emotion-related behavior and memory in rodents. We previously demonstrated that social isolation for 4 weeks after weaning induced a significant impairment of spatial working memory and aggressive behavior in mice, which was closely associated with a significant decrease of adult neurogenesis and reduced mRNA levels of a brain-specific transcription factor, neuronal PAS domain-4 (Npas4), in the dentate gyrus of the hippocampus (Ibi et al. 2008). To confirm the stress-induced reduction of Npas4 mRNA levels, we utilized another type of stress, chronic restraint stress, by which mice showed an impairment of hippocampus-dependent contextual fear memory with a significant reduction of hippocampal neurogenesis (Yun et al. 2010). By using this stress model, we demonstrated that acute in vivo treatment with corticosterone (CS), as well as acute or chronic stress, significantly reduced Npas4 mRNA expression levels in the hippocampus (Yun et al. 2010).

Npas4 strongly promotes the survival of cultured hippocampal neurons (Zhang et al. 2009). Furthermore, NPAS4 immunoreactivity was detected in progenitor, immature, and mature neurons of the dentate gyrus in control and stressed mice (Yun et al. 2010). Taken together with our previous findings, it is postulated that the stress-induced down-regulation of Npas4 expression may contribute to stress-induced brain dysfunction, such as the impairment of emotion-related behavior, spatial learning, memory, and hippocampal neurogenesis.

Glucocorticoids (GCs) regulate the transcription of target genes through the binding of agonist-bound glucocorticoid receptors (GRs) to glucocorticoid response elements (GREs). There are several distinct types of GREs: simple, composite, half-site, negative, and tethering (Schoneveld et al. 2004). A tandem repeat of GRE half-sites is known as a negative GRE (Ou et al. 2001; Schoneveld et al. 2004), and there are putative GRE half-sites in the region of −2000 to −1000 upstream of the transcription initiation site of the Npas4 promoter in the mouse genome, and in the −4000 to −3000 base pair region upstream of the human initiation site. Therefore, in this study, we investigated the potential involvement of GC-induced trans-repression of the Npas4 promoter in the stress-induced GC-mediated reduction of Npas4 gene expression.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Animals

Seven-week-old male imprinting control region (ICR) mice were purchased from Japan SLC Inc. (Hamamatsu, Japan). Mice were housed, four to five mice 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. Animals were handled in accordance with the guidelines established by the Institutional Animal Care and Use Committee of Nagoya University.

Adrenalectomy

After a 1-week adaptation period, adrenalectomy (ADX) or sham surgery was performed as described elsewhere (Makimura et al. 2000), but with minor modifications. Briefly, ADX was performed by bilateral flank incision under anesthesia (pentobarbital 40 mg/kg, i.p.). Sham surgery entailed the same procedure as ADX, except that the adrenal glands were grasped, but not removed. At the time of ADX, drinking water was replaced with 5% glucose in saline (0.9% NaCl). One week after surgery, mice were killed to determine Npas4 mRNA levels. We also measured plasma corticosterone concentrations to confirm ADX, according to a previous report (Yun et al. 2010).

Restraint stress

ICR mice were randomly divided into two groups: a restraint-stress group and a control group. Restraint stress was applied with a stainless mesh that allowed for a close fit to the mice for 3 h, as described previously (Takuma et al. 2007; Yun et al. 2010). Mifepristone, a GR antagonist (Sigma-Aldrich, St Louis, MO, USA), was injected i.p. at a dose of 25 mg/kg, 30 min before exposure to restraint stress. Mice were killed immediately after the restraint stress to determine Npas4 mRNA levels and to examine the GR binding to Npas4 promoter by chromatin immunoprecipitation (ChIP) assay.

Cell culture

Neuro2a cells were kindly donated by Dr. Sigeru Yoshida at Taisho Pharmaceutical Co., Ltd. (Saitama, Japan) and cultured in Dulbecco's Modified Eagle's Medium (Sigma-Aldrich) supplemented with 10% fetal bovine serum (Gibco, Invitrogen Ltd., Paisley, UK) and antibiotics/antimycotics (Gibco, Invitrogen) at 37°C in a humidified atmosphere with 5% CO2. Neuro2a cells were seeded at 50 000 cells/mL, 2 mL per well (six-well dishes).

CS treatment in vitro and in vivo

Neuro2a cells were treated with CS (100 μM) (Sigma-Aldrich) in the culture media for 6 h, and cells were then lysed to analyze NPAS4 protein level. To measure the Npas4 mRNA level, Neuro2a cells were treated with CS (100 μM) for 4 h. For the in vivo experiment, ICR mice were injected s.c. with CS (10 mg/kg) for 3 days and killed 4 h after the last injection to measure the mRNA and protein levels of NPAS4 in the prefrontal cortex.

Quantitative real-time RT-PCR

Complementary DNA was synthesized from total RNA using the SuperScript III First-Strand Synthesis System from RT-PCR (Invitrogen). mRNA expression levels were 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 1 μg cDNA and 500 nM primers in Power SYBR Green Master Mix (Applied Biosystems). The following primers were used: 5′-AGCATTCCAGGCTCATCTGAA-3′ (forward) and 5′-GGCGAAGTAAGTCTTGGTAGGATT-3′ (reverse) for Npas4, and 5′-TGTCAAGCTCATTTCCTGGTATGA-3′ (forward) and 5′-CTTACTCCTTGGAGGCCATGTAG-3′ (reverse) for glyceraldehyde-3-phosphate dehydrogenase.

Western blot analysis

Whole-cell lysates of Neuro2a cells were homogenized in 300 μL of ice-cold radio immunoprecipitation assay lysis buffer [20 mM Tris-HCl, 150 mM NaCl, 2 mM EDTA, 0.1% sodium dodecyl sulfate , 1% NP-40, 1% sodium deoxycholate, 1 tablet of mini-complete proteinase inhibitor (Roche Applied Science, Mannheim, Germany)/10 mL, pH 7.6]. Alternatively, frontal cortex protein extracts were obtained by homogenization in ice-cold ristocetin-induced platelet agglutination lysis buffer containing 50 mM sodium fluoride and 1 mM sodium vanadate. Thereafter, homogenates were centrifuged at 13 000 g for 20 min to obtain supernatant. Polyclonal Npas4 antibody was raised against a recombinant protein sequence of Npas4 protein [597–802; same as reported by Lin et al. (2008), rabbit, 1 : 5000, MBL, Nagano, Japan]. Whole-cell lysates and frontal cortex protein extracts were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis (7.5%) and immunoblotting was performed. Membranes were blocked with skimmed milk (5%) and incubated with anti-Npas4 B and anti-β-actin (goat, 1 : 300; Santa Cruz Biotechnology) antibodies at 4°C overnight. Horseradish peroxidase-conjugated anti-rabbit (1 : 1500; KPL, Gaithersburg, MA, USA) or anti-goat antibodies (1 : 1000; R&D systems, Minneapolis, MN, USA) were subsequently added to membranes for 1 h at 25°C. Immunoreactive complexes on the membrane were detected using the Enhanced Chemiluminescence Plus Western Blotting Detection Reagent (GE Healthcare Japan, Tokyo, Japan) and analyzed using the Atto Light-Capture System (Atto, Tokyo, Japan).

Constructs for luciferase assays

The pGL4.10[luc2] vector (Promega, Madison, WI, USA), containing the gene for luciferase, was used in this study. To create a pGL4.10–Npas4 promoter construct containing the mouse Npas4 promoter upstream to the luciferase gene, the Npas4 promoter was amplified by PCR from mouse genomic DNA. The following primers were used to prepare the Npas4 promoter construct: 5′-GCCGGTACCTACTAAGCCATGGCCCTGGTCACCAAAGTA-3′ (forward for −1968/Luc, which includes a KpnI restriction site at the 5′ end), 5′-GCCGGTACCGAGGATTCCTGTCCTAATATGGAGCTGGGA-3′ (forward for −970/Luc, which includes a KpnI restriction site at the 5′ end), and 5′-GATATCCTCGAGGCTTCCTCTTCCTTGCTTCCCGGTCTTTT-3′ (reverse, which includes EcoRV and XhoI restriction sites at the 5′ end). PCR products were digested with KpnI and XhoI enzymes and then ligated into a pGL4.10[luc2] vector that was digested with the same enzymes. PCR was used to create the Δ-1659–1644/Luc and Δ-1659–1624/Luc plasmids in which the putative GRE sequences of the Npas4 promoter were deleted. The primers used were as follows: 5′-GTTCTGTCCCTCTCTCCCAATGTTTGTGTG-3′ (forward) and 5′-CAAAAACACACAAACATTGGGAGAGA-3′ (reverse) for Δ-1659–1644/Luc and 5′-TGTCCCTCTCTCCCAATTGAGACAGGGTTTCTCTGT-3′ (forward) and 5′-ACAGAGAAACCCTGTCTCAATTGGGAGAGAGGGACA-3′ (reverse) for Δ-1659–1624/Luc. PCR was used to create GRE1mut/Luc, GRE2mut/Luc, GRE3mut/Luc, GRE4mut/Luc, and GRE5mut/Luc plasmids in which the putative GRE sequences of the Npas4 promoter were mutated (Table 1). The primers used were as follows: 5′-TTTTTTGAGGGGTTGTTCTGTTTGTTTGTG-3′ (forward) and 5′-GAACAACCCCTCAAAAAAAATTGGGAGAGAGG-3′ (reverse) for GRE1mut/Luc, 5′-TTTTTGAGGATGTTTGTTTGTGTGTTTTTG-3′ (forward) and 5′-AAACATCCTCAAAAACAAAAAAAAATTGGG-3′ (reverse) for GRE2mut/Luc, 5′-GTTCTGGGCTTGGGTGTGTTTTTGTTTTTTG-3′ (forward) and 5′-ACACACCCAAGCCCAGAACAAAAAACAAAAA-3′ (reverse) for GRE3mut/Luc, 5′- TTTGTGGAGGGGTGTTTTTTGAGACAGGG-3′ (forward) and5′-AAAACACCCCTCCACAAACAAACAGAACAA-3′ (reverse) for GRE4mut/Luc, and 5′-TGTTTTGAGGGGTTGAGACAGGGTTTCTCTG-3′ (forward) and 5′-TCTCAACCCCTCAAAACACACAAACAAACAG-3′ (reverse) for GRE5mut/Luc.

Table 1. Mutation of putative glucocorticoid response elements (GREs) of the Npas4 promoter
TTTTTGTTTTTTGTTCTGTTTGTTTGTGTGTTTTTGTTTTTTGAGACAGG  wild type
TTTTgaggggTTGTTCTGTTTGTTTGTGTGTTTTTGTTTTTTGAGACAGG   GRE1mut
TTTTTGTTTTTgaggaTGTTTGTTTGTGTGTTTTTGTTTTTTGAGACAGG  GRE2mut
TTTTTGTTTTTTGTTCTGggTGTggGTGTGTTTTTGTTTTTTGAGACAGG   GRE3mut
TTTTTGTTTTTTGTTCTGTTTGTTTGTGgaggggTGTTTTTTGAGACAGG  GRE4mut
TTTTTGTTTTTTGTTCTGTTTGTTTGTGTGTTTTgaggggTTGAGACAGG  GRE5mut

Luciferase Assay

Neuro2a cells were plated in 24-well plates at 20 000 cells/well in growth medium. The next day, cells were transfected with 200 ng/well of the pGL4.10 constructs −1968/Luc, −970/Luc, Δ-1659–1644/Luc, or Δ-1659–1624/Luc, and 30 ng/well of the phRG–TK construct, which expresses renilla luciferase, using FuGENE6 (Roche) according to the manufacturer's protocols. The renilla luciferase construct was used as a control for transfection efficiency. After 24 h, cell lysates were prepared and assayed for luciferase activity using the Dual-Luciferase Reporter Assay System (Promega). Activity tests were performed and luminescence measured using a MiniLumat luminometer (Berthold, Wildbad, Germany). To explore the effect of GC in the serum used for cell culture on Npas4 promoter activity, transfected cells were incubated in the medium containing 10% or 2% serum for 18 h.

ChIP Assay

Neuro2a cells were treated with mifepristone (0, 25, 50, or 100 μM). Six hours after treatment, chromatin solutions of Neuro2a cells (6 000 000 cells) were prepared in a volume of 250 μL by using the SimpleChIP Enzymatic Chromatin IP Kit (Cell Signaling Technology Japan, Tokyo, Japan). Alternatively, mice were killed immediately after the restraint stress for 3 h. Mifepristone was injected i.p. at a dose of 25 mg/kg, 30 min before exposure to the restraint stress. Chromatin solutions of hippocampus (100 mg) were prepared in a volume of 500 μL by using the SimpleChIP Plus Enzymatic Chromatin IP Kit (Cell Signaling Technology Japan). These chromatin solutions were immunoprecipitated overnight at 4°C using 2 μg anti-GR antibody (Santa Cruz Biotechnology). Assays included normal rabbit IgG (Santa Cruz Biotechnology) as a control for the specificity of the GR antibody. The following primers were used to detect the coimmunoprecipitated Npas4 promoter region: 5′-AACTTTCTGCAGCCAGTTCTGTCC-3′ (forward) and 5′-AAGTGAGTTCCAGGACAGCCAAGG-3′ (reverse).

Statistical analysis

All results are expressed as mean ± standard error (SE) values. Statistical significance was determined by one-way or two-way analysis of variance (anova) followed by Bonferroni's multiple comparisons test for multigroup comparison and the student's t-test for two-group comparisons.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Alteration of CS levels affects Npas4 expression in vivo

A previous study indicated that NPAS4 protein levels in the hippocampus were reduced by CS (Yun et al. 2010). In this study, mice were treated with an s.c. injection of CS (10 mg/kg) for 3 days, and NPAS4 protein levels in the cortex were determined. The polyclonal anti-Npas4 antibody, which we raised by ourselves and used for western blotting recognized endogenous NPAS4 in the frontal cortex of wild-type mice, but not in Npas4-knockout mouse (Yun et al., submitted). CS treatment significantly reduced the NPAS4 protein in the cortex (Fig. 1a). As CS also reduced the Npas4 mRNA levels in the hippocampus and frontal cortex (Fig. 1b), CS may regulate the transcription of Npas4 gene. To confirm that the alteration of CS levels affects Npas4 gene expression in the brain, we also examined the effect of ADX in mice upon Npas4 mRNA levels in the hippocampus and frontal cortex. One week after adrenalectomy, when the circulating levels of CS were significantly reduced to 4.2 ± 1.1 ng/mL from 72.1 ± 35.2 ng/mL in the sham-operated group, the mRNA levels of Npas4 in the cortex of ADX mice were significantly increased to 2.5-fold and 3.0-fold of the levels observed in the sham operated group, respectively (Fig. 1c). To confirm that restraint stress reduced Npas4 expression via CS, we examined the effect of the GR antagonist, mifepristone, upon the Npas4 mRNA reduction induced by restraint stress for 3 h (Fig. 1d). Npas4 mRNA levels were reduced by restraint stress for 3 h, but pre-treatment with mifepristone prevented this reduction (Fig. 1d). Restraint stress for 3 h increased serum CS concentration compared with controls, whereas mifepristone had no effect upon corticosterone concentration (Fig. 1e).

image

Figure 1. In vivo regulation of Npas4 expression by corticosterone. (a) Imprinting control region (ICR) mice were injected s.c. with corticosterone (CS) (10 mg/kg) for 3 days and killed 4 h after the final injection, and protein levels of neuronal PAS domain 4 (NPAS4) in the frontal cortex were analyzed by western blotting. NPAS4 protein levels were normalized by β-actin. Values indicate mean ± SE (n = 6). Values annotated with an asterisk (*) indicate cases where p < 0.05 versus control. (b) ICR mice were injected s.c. with CS (10 mg/kg) for 3 days and killed 4 h after the final injection to measure the mRNA levels of NPAS4 in the hippocampus and prefrontal cortex. Values indicate mean ± SE (n = 10). Values annotated with an asterisk (*) indicate cases where p < 0.05 versus control of hippocampus, and three sharps (###) indicate cases where p < 0.001 versus control of frontal cortex. (c) Npas4 mRNA levels were measured in the hippocampus and frontal cortex 1 week after adrenalectomy (ADX). Values indicate mean ± SE (n = 8–9). Values annotated with an asterisk (*) or a sharp (#) indicate cases where p < 0.05 versus each sham. (d, e) Mice were killed immediately following restraint stress for 3 h, and serum and hippocampus were sampled. Mifepristone (25 mg/kg) was injected (i.p.) 30 min before exposure to restraint stress. (d) Npas4 mRNA levels were normalized by dimethylsulfoxide (DMSO)-treated and non-stressed mice. Values indicate mean ± SE (n = 6). Values annotated with an asterisk (*) indicate cases where p < 0.05 versus DMSO-treated non-stressed controls, whereas values annotated with a sharp (#) indicate cases where p < 0.05 versus the DMSO-treated restraint-stressed group. (e) Corticosterone concentration in the serum. Values indicate mean ± SE (n = 4). Values annotated with an asterisk (*) indicate cases where p < 0.05 versus the non-stressed group.

Download figure to PowerPoint

GC regulates Npas4 promoter activity

There are many putative GRE half-sites in the −2000 to −1000 base pair upstream region from the transcription initiation site of the Npas4 promoter in mice (Fig. 2a). To explore the effect of GC upon Npas4 promoter activity, we constructed a promoter assay plasmid containing 1968bp of the Npas4 promoter, −1968/Luc (Fig. 2a). In Neuro2a cells, CS treatment (100 μM) for 6 h significantly reduced the endogenous NPAS4 protein level compared with the levels of vehicle-treated controls (Fig. 2b). CS treatment (100 μM) for 4h also reduced the Npas4 mRNA level (Fig. 2c). To reduce the effect of GC in the serum, we reduced serum concentration from 10% to 2% in the culture medium. The reduction of serum concentration significantly increased the luciferase activity recovered from cells transfected with −1968/Luc plasmids (Fig. 2d). Furthermore, the GR antagonist mifepristone significantly, and dose dependently, increased the activity of the Npas4 promoter as measured by luciferase activity in Neuro2a cells cultured with 10% fetal bovine serum (Fig. 2e).

image

Figure 2. In vitro regulation of Npas4 expression by corticosterone. (a) Schematic representation of the Npas4 promoter in mice. Solid bars in the construct indicate putative glucocorticoid response element (GRE) half-sites. (b) Neuro2a cells were lysed 6 h after corticosterone (CS) (100 μM) treatment and protein levels of neuronal PAS domain 4 (NPAS4) were analyzed by western blotting. NPAS4 protein levels were normalized by β-actin. Values indicate mean ± SE (n = 4). Values annotated with an asterisk (*) indicate cases where p < 0.05 versus control. (c) Neuro2a cells were lysed 4 h after CS (100 μM) treatment and mRNA levels of NPAS4 were analyzed. Values indicate mean ± SE (n = 4). Values annotated with asterisk (**) indicate cases where p < 0.01 versus control. (d) Luciferase reporter activity in Neuro2a cells transfected with the Npas4 promoter construct (−1968/Luc). Control cells were transfected with empty vector, pGL4.10[luc2]. Transfection efficiency was normalized by Renilla luciferase activity derived from phRG-TK. Cells were cultured with either 10% or 2% fetal bovine serum. Luciferase activity in Neuro2a cells transfected with −1968/Luc and cultured in 10% fetal bovine serum was set to 1. Values indicate mean ± SE (n = 4). Values annotated with three asterisks (***) indicate cases where p < 0.001 versus 10% serum. (e) Neuro2a cells were transfected with −1968/Luc and treated with the glucocorticoid receptor (GR) antagonist mifepristone for 6 h in the presence of 10% fetal bovine serum. Luciferase activity in Neuro2a cells transfected with −1968/Luc and treated with 0 μM mifepristone was set to 1. Values indicate mean ± SE (n = 4). Values annotated with asterisk (*) indicate cases where p < 0.05 versus controls.

Download figure to PowerPoint

GREs regulate Npas4 promoter activity

To determine the possibility that GREs could affect Npas4 transcription, Neuro2a cells were transiently transfected with luciferase reporter plasmids containing various Npas4 promoters (Fig. 3a). The ectopic expression of −970/Luc, which contained nucleotides −970 to +90 of the mouse Npas4 promoter region, resulted in a marked increase in luciferase activity compared with −1968/Luc, which contained nucleotides −1968 to +90, and included putative negative GREs (Fig. 3b). Deletion of the putative negative GREs at the 5′-terminal of −1968/Luc, Δ-1659–1644/Luc and Δ-1659–1624/Luc, remarkably increased luciferase activity to the same level as −970/Luc (Fig. 3b).

image

Figure 3. Glucocorticoid response elements (GREs) regulate Npas4 promoter activity. (a) Deletion mutants of the Npas4 promoter construct. Solid bars in each construct indicate putative GRE half-sites. (b) Luciferase reporter activity in Neuro2a cells transfected with −1968/Luc, Δ-1659–1644/Luc, Δ-1659–1624/Luc, or −970/Luc. Control cells were transfected with empty vector, pGL4.10[luc2]. Transfection efficiency was normalized by Renilla luciferase activity derived from phRG-TK. Luciferase activity in Neuro2a cells transfected with −1968/Luc was set to 1. Values indicate mean ± SE (n = 4). Values annotated with three asterisks (***) indicate cases where p < 0.001 versus −1968/Luc.

Download figure to PowerPoint

We subsequently sought to identify the GRE responsible for trans-repression of the Npas4 promoter by GC. Neuro2a cells were transiently transfected with luciferase reporter plasmids containing various mutations of the GREs in Npas4 promoters (Fig. 4a). We examined the effects of sequential mutations (GRE1mut–GRE5mut) in GREs at the 5′-terminal of −1968/Luc (Table 1). Introduction of mutation in GRE at −1647–1639 (GRE3mut) significantly increased Npas4 promoter activity, whereas other mutations (GRE1mut, GRE2mut, GRE4mut, and GRE5mut) had no effect (Fig. 4b). GRE1, GRE3, and GRE4 have 6-bp spacing sequence between them (Fig. 4c).

image

Figure 4. Effect of mutagenesis of glucocorticoid response elements (GREs) upon Npas4 promoter activity. (a) Mutant GREs of the Npas4 promoter. White bars in each construct indicate mutated GRE half-sites. (b) Luciferase reporter activity in Neuro2a cells transfected with −1968/Luc, GRE1mut/Luc, GRE2mut/Luc, GRE3mut/Luc, GRE4mut/Luc, GRE5mut/Luc, or Δ-1659–1644/Luc. Control cells were transfected with empty vector, pGL4.10[luc2]. Transfection efficiency was normalized by Renilla luciferase activity derived from phRG-TK. Luciferase activity in Neuro2a cells transfected with −1968/Luc was set to 1. Values indicate mean ± SE (n = 4). Values annotated with three asterisks (***) indicate cases where p < 0.001 versus −1968/Luc. (c) Tandem GRE half-sites in the mouse Npas4 promoter.

Download figure to PowerPoint

GR interacts with the Npas4 promoter in Neuro2a cells

The binding of agonist-bound GR to GREs in the target gene promoter is considered to be one of the mechanisms responsible for regulation of gene transcription. In ChIP analysis, the endogenous Npas4 promoter in Neuro2a cells was coimmunoprecipitated with anti-GR antibody, and binding was inhibited by mifepristone in a concentration-dependent manner (Fig. 5). This result suggested the binding of agonist-bound GR to GREs in the Npas4 promoter.

image

Figure 5. Interaction of glucocorticoid receptor (GR) with the Npas4 promoter in Neuro2a cells. Neuro2a cells were treated with mifepristone in the presence of 10% fetal bovine serum for 6 h. Chromatin solutions were immunoprecipitated with anti-GR antibody or normal IgG for control Values indicate mean ± SE (n = 6). Values annotated with an asterisk (*) indicate cases where p < 0.05 versus control. Values annotated with two asterisks (**) indicate cases where p < 0.01 versus control.

Download figure to PowerPoint

Stress increases the interaction of GR with the Npas4 promoter in the hippocampus

Finally, we investigated whether restraint stress for 3 h in mice increased the binding of GR to GREs of the Npas4 promoter in the hippocampus. The ChIP analysis with anti-GR antibody revealed that restraint stress significantly increased the binding of GR to GREs in the Npas4 promoter in the hippocampus, and this increase was completely blocked by the pre-treatment with mifepristone (Fig. 6).

image

Figure 6. Interaction of glucocorticoid receptor (GR) with the Npas4 promoter in the hippocampus of mice exposed to restraint stress. Mice were killed immediately following restraint stress for 3 h, and the hippocampus were sampled. Mifepristone (25 mg/kg) was injected (i.p.) 30 min before exposure to restraint stress. Chromatin solutions of hippocampus were immunoprecipitated with anti-GR antibody. Values indicate mean ± SE (n = 8). Values annotated with three asterisks (***) indicate cases where p < 0.001 versus dimethylsulfoxide (DMSO)-treated non-stressed controls, whereas values annotated with sharp (###) indicate cases where p < 0.001 versus the DMSO-treated restraint-stressed group.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

The hippocampus is one of several brain areas that are thought to play a central role in affective behavior (Airan et al. 2007; Kempermann et al. 2008; Thompson et al. 2008; Revest et al. 2009). Chronic stress causes shortening and debranching of dendrites in the CA3 region of the hippocampus and suppresses neurogenesis in the dentate gyrus (McEwen 2000). The adult mammalian brain, especially the dentate gyrus of the hippocampus, contains neural stem cells that generate new neurons (Mongiat and Schinder 2011). We previously demonstrated that social isolation, or restraint stress after weaning, reduced the survival of newly divided cells and neurogenesis in the dentate gyrus of the hippocampus (Ibi et al. 2008; Yun et al. 2010). In rodents, stress and elevated CS concentrations suppress adult neurogenesis (Gould and Tanapat 1999; Heine et al. 2004; Kempermann et al. 2004). Moreover, the GR antagonist mifepristone rapidly reverses the chronic CS-induced reduction of adult neurogenesis in rats (Mayer et al. 2006; Oomen et al. 2007).

We showed that acute or chronic restraint stress, as well as acute CS injection, significantly reduced Npas4 mRNA levels in the dentate gyrus and hippocampus, respectively, suggesting that stress may reduce Npas4 expression in the hippocampus via the action of circulating CS (Yun et al. 2010). In this study, we confirmed the reduction of Npas4 expression by CS treatment in vitro and in vivo. Moreover, the increase in Npas4 expression following ADX supports the hypothesis that Npas4 expression is regulated by GC signaling. The model for GC function is that hormone-bound GR binds to GREs in the regulatory regions of target genes, thereby changing the expression of these genes (Schoneveld et al. 2004). The consensus sequence of a negative GRE is more variable than that of a simple GRE and consists of a tandem repeat of GRE half-sites (Schoneveld et al. 2004). For example, the expression of the neuronal serotonin (5-HT1A) receptor is negatively regulated by GC (Ou et al. 2001). The 6-bp spacing between repeated 6-bp GRE half-sites (TGTCCT-nnnnnn-TGTCCT) was conserved in the promoter of the 5-HT1A receptor. The GR and mineralocorticoid receptor heterodimer that binds to the direct repeat GRE half-sites directly mediates the CS-induced trans-repression of the 5-HT1A receptor promoter (Ou et al. 2001). Surjit et al. (2011) demonstrated that the inverted repeat negative GRE (IR nGRE) regulates widespread transcription. There are five putative tandem GRE half-site sequences in the Npas4 promoter. Deletion of three of these GRE half-sites, which are putative negative GREs, in the Npas4 promoter (5′-TGTTTTTTGTTCTGTTTG-3′; Δ-1659–1644/Luc), or mifepristone treatment dramatically increased Npas4 promoter activity. Moreover, introduction of mutation in GRE at −1647–1639 (GRE3mut) increased Npas4 promoter activity. Interaction of GR and the Npas4 promoter was revealed in the hippocampus as well as Neuro2a cells by ChIP assay. These results suggest that GR reduces Npas4 transcription directly by binding to negative GREs in its promoter. Stress-induced reduction of Npas4 mRNA levels in the hippocampus (Fig. 1d) may be mediated at least in part by the down-regulation of Npas4 gene transcription via the binding of agonist-bound GR to its promoter.

Within the region between positions −1659 and −1624, there are two 6-bp spacing sequences, one is located between GRE1 and 3 and the other between GRE3 and GRE4 (TGTTTT - nnnnnn - TGTTTG - nnnnnn - TGTTTT) as found in the promoter of the 5-HT1A receptor (Ou et al. 2001). GR may be able to bind to the tandem GRE half-sites (GRE1–3 or GRE3–4) and reduce the Npas4 transcription. It is likely that GRE3 is necessary for the inhibitory effect of CS on Npas4 promoter activity. It is possible that GR may create a trans-repression complex, which represses Npas4 transcription. Further studies are required to address this issue.

The requirement of NPAS4 for neuroprotection has been suggested in cultured cells (Hester et al. 2007). NPAS4 plays a role in the development of inhibitory synapses by regulating the expression of brain-derived neurotrophic factor (BDNF) (Lin et al. 2008; Pruunsild et al. 2011). BDNF is known to promote the formation or proliferation of GABAergic synapses in the cortex (Kohara et al. 2007) and to modify synaptic plasticity in the hippocampus (Schjetnan and Escobar 2012). The reduction of NPAS4 expression may be part of the mechanism underlying the stress-induced impairment of GABAergic neuronal function (Luscher et al. 2011) and BDNF expression (Angelucci et al. 2005). GABA signaling reduces the hyperactivation of the hypothalamic–pituitary–adrenal axis (Luscher et al. 2011), and has a prominent role in the control of stress in the brain, the most important vulnerability factor in mood disorders (Luscher et al. 2011). Npas4 is also suggested to be a master regulator of activity-regulated gene programs and is central to memory formation in the CA3 region of the hippocampus (Ramamoorthi et al. 2011). We have suggested that stress induces significant impairment of spatial working memory in mice, which is associated with reduced mRNA levels of Npas4 (Ibi et al. 2008). Reduction of Npas4 expression may result in the impairment of widespread brain function. Further studies to clarify the molecular mechanism underlying the regulation of Npas4 expression will suggest novel therapeutic approaches to psychiatric disorders triggered by stress.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

This study was supported, in part, by the following funding sources: (i) grants-in-aid for Scientific Research (No. 22390046, 23590299) from the Japan Society for the Promotion of Science, (ii) a grant for the global COE program from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, (iii) a grant from the Academic Frontier Project for Private Universities, matching fund subsidy from MEXT, 2007–2011. (iv) JST, CREST. (v) the Uehara Memorial Foundation. There are no conflicts of interest involving the present work.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References