OUP user menu

DJ-1/park7 modulates vasorelaxation and blood pressure via epigenetic modification of endothelial nitric oxide synthase

Kyung Jong Won, Seung Hyo Jung, Soo Hyun Jung, Kang Pa Lee, Hwan Myung Lee, Dong-Youb Lee, Eun-Seok Park, Junghwan Kim, Bokyung Kim
DOI: http://dx.doi.org/10.1093/cvr/cvt274 473-481 First published online: 9 December 2013

Abstract

Aims DJ-1/park7, a multifunctional protein, may play essential roles in the vascular system. However, the function of DJ-1/park7 in vascular contractility has remained unclear. The present study was designed to investigate whether the DJ-1/park7 is involved in the regulation of vascular contractility and systolic blood pressure (SBP).

Methods and results Norepinephrine (NE) elevated contraction in endothelium-intact vessels in a dose-dependent manner, to a greater extent in DJ-1/park7 knockout (DJ-1/park7−/−) mice than in wild-type (DJ-1/park7+/+) mice. Acetylcholine inhibited NE-evoked contraction in endothelium-intact vessels, and this was markedly impaired in DJ-1/park7−/− mice compared with DJ-1/park7+/+. Nitric oxide (NO) production (82.1 ± 2.8% of control) and endothelial NO synthase (eNOS) expression (61.7 ± 8.9%) were lower, but H2O2 production (126.4 ± 8.6%) was higher, in endothelial cells from DJ-1/park7−/− mice than in those from DJ-1/park7+/+ controls; these effects were reversed by DJ-1/park7-overexpressing endothelial cells from DJ-1/park7−/− mice. Histone deacetylase (HDAC)-1 recruitment and H3 histone acetylation at the eNOS promoter were elevated and diminished, respectively, in DJ-1/park7−/− mice compared with DJ-1/park7+/+ controls. Moreover, SBP was significantly elevated in DJ-1/park7−/− mice compared with DJ-1/park7+/+ controls, but this elevation was inhibited in mice treated with valproic acid, an inhibitor of Class I HDACs including HDAC-1.

Conclusion These results demonstrate that DJ-1/park7 protein may be implicated in the regulation of vascular contractility and blood pressure, probably by the impairment of NO production through H2O2-mediated epigenetic inhibition of eNOS expression.

  • DJ-1/park7
  • Vascular contractility
  • eNOS
  • Epigenetic modification

1. Introduction

Vascular endothelial cells play important roles in the regulation of physiological and pathophysiological events, such as vascular contractility, haemostasis, and angiogenesis. Endothelial cells can synthesize and release various vasoactive substances in response to mechanical and chemical stimuli and maintain vascular tone.1 These vasoactive substances include endothelium-derived relaxing factors, e.g. nitric oxide (NO) and prostaglandin I2, and vasoconstrictors, e.g. endothelin and thromboxane.2 NO plays a key role in preserving vascular homeostasis by modulating the activity associated with platelet, endothelium, and smooth muscle cells (SMCs).3 A reduction in NO bioavailability is induced in various circumstances, including decreased expression of endothelial NO synthase (eNOS), lack of substrate or cofactors for eNOS, and NO degradation.4 These may result in functional deterioration of endothelial cells and lead to the development of vascular disorders, such as hypertension and atherosclerosis. Previously, we reported that diminished expression of an enzyme for the recycle of NOS cofactor is involved in the elevation of blood pressure in spontaneously hypertensive rats (SHRs).5

Reactive oxygen species (ROS), namely superoxide, hydrogen peroxide (H2O2), hydroxyl radical, and peroxynitrite, are generated as intermediates in reduction–oxidation (redox) processes.6 These redox molecules participate in cellular functions, such as inflammation, apoptosis, and growth, in a variety of cells. Vascular functions can be altered by abnormal redox signalling induced by excessive production of ROS and/or by decreases in antioxidant activity.7 ROS affect vascular contractility by influencing the endothelium and exerting direct effects on SMCs.6 Excessive production of ROS leads to diminished eNOS activity and NO production in endothelial cells.8 In addition, ROS participate in epigenetic modification of histone proteins of genes in cells.9,10 These findings imply that oxidative stress is tightly involved in the development of vascular disorders, such as hypertension.6 Moreover, hypertension is associated with attenuated antioxidant systems and the increased production of ROS.11 ROS production is increased in endothelial cells and SMCs from hypertensive animals.12 These findings suggest that alterations in the expression of proteins related to redox processes may be involved in the development of hypertension.

DJ-1/park7 was first identified as a novel oncogene and recently found to be a gene responsible for autosomal recessive early onset Parkinsonism.13,14 DJ-1/park7 is a multifunctional protein that is associated with various functions, including transcriptional regulation and antioxidative stress.1518 In a previous study, we identified DJ-1/park7 as a protein that responds rapidly to oxidative stress in vascular SMCs using proteomics.18,19 Recently, another study demonstrated the possibility that DJ-1/park7 may be associated with the renal pathogenesis of hypertension related to ROS.20 These reports suggest that DJ-1/park7 may act as a contributor to pathophysiological events associated with oxidative stress in the vascular system. However, its roles in vascular endothelial cells, especially eNOS/NO-mediated vasoconstriction, have not yet been determined. Therefore, in this study, we investigated whether DJ-1/park7 is involved in the regulation of vascular contractility and blood pressure using DJ-1/park7 knockout mice.

2. Methods

See Supplementary materials online for expanded version of this section.

2.1 Tissue preparation

The investigation was performed in accordance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH publication No. 85-23, revised 1996) and was approved by the Animal Subjects Committee and by the Institutional Guidelines of Konkuk University, Korea. Experiments using human material were approved by the Institutional Review Board of Konkuk University Chungju Hospital and the principles outlined in the Declaration of Helsinki. Male DJ-1/park7-homozygous knockout (DJ-1/park7−/−, B6.Cg-Park7tm1shn/J; n = 59) and wild-type (DJ-1/park7+/+; n = 57) mice with the same genetic background were purchased from Jackson Laboratory (Bar Harbor, ME, USA). Animals were housed under constant optimal temperature and humidity conditions with a normal 12–12-h light–dark cycle.

2.2 Measurement of contractility

The animals (8 weeks, about 25 g) were killed using CO2 gas inhalation and bled rapidly by cutting the carotid arteries. The thoracic aorta was rapidly and carefully removed and placed in a physiological salt solution (PSS). Vessel rings from the aortic and superior mesenteric artery were horizontally mounted on two L-shaped holders; of which, one end was attached to a stainless steel rod and the other to a force transducer (FT03; Grass-Telefactor Instruments, West Warwick, RI, USA), in 3 mL organ baths containing PSS.21

2.3 Isolation and culture of endothelial cells

Endothelial cells were isolated enzymatically from the lung of DJ-1/park7+/+ and DJ-1/park7−/− mice (8 weeks old) using a previously reported method with some modifications.22 Briefly, mice were anaesthetized by an intraperitoneal injection of Zoletil (40 mg/kg; Virbac Laboratories, Carros, France) and Rompun (10 mg/kg; Bayer Korea, Seoul, Korea). The adequacy of anaesthesia was monitored based on the disappearance of the pedal withdrawal reflex response to foot pinch. Lungs harvested were cut into 1–2×2 mm pieces and digested using 1% collagenase Type 1 (Worthington Biochemical, Freehold, NJ, USA) for 1 h at 37°C, and then the digest was passed through a 100 μm cell strainer (SPL Life Sciences, Seoul, Korea). The purity of isolated cell populations was >85%, as confirmed by immunocytochemical analysis using an anti-CD31 antibody.

2.4 Immunostaining, western blotting, and detections of ROS and NO

Analyses were performed with methods as used in previous reports.7,18

2.5 Overexpression and knockdown of proteins

Overexpression of DJ-1/park7 and knockdown of histone deacetylase (HDAC)-1 were performed with methods as used in previous reports.18

2.6 Chromatin immunoprecipitation assay

Chromatin immunoprecipitation (ChIP) analysis was performed using an EZ ChIP kit (Upstate Biotechnology, Lake Placid, NY, USA) according to the manufacturer's protocols with some modifications.23

2.7 Measurement of blood pressure

Animals were trained for 1 week to have their blood pressure measured. They were kept in a quiet place before the measurement of their blood pressure and placed in a cupboard at room temperature. Systolic blood pressure (SBP) was measured using a tail-cuff method (MK-2000ST; Muromachi Kikai, Tokyo, Japan).

2.8 Data analysis

Experimental data are expressed as means ± SEM. Statistical analysis of the data was performed using Student's t-test for comparisons between pairs of groups and analysis of variance for multiple comparisons. A value of P < 0.05 was considered to be statistically significant.

3. Results

3.1 Alteration of norepinephrine-induced contraction in the aorta of DJ-1/park7−/− mice

The deficiency of DJ-1/park7 gene and protein was first confirmed in DJ-1/park7−/− and DJ-1/park7+/+ mice using PCR genotyping and western blot analysis (see Supplementary material online, Figure S1A and B). To determine the roles of DJ-1/park7 in vascular tone, contraction of isolated aortic rings was induced with norepinephrine (NE) or high K+, and the effects in both the DJ-1/park7−/− and DJ-1/park7+/+ mice were compared. In the endothelium-intact aorta, the addition of NE (0.001–10 μM) showed concentration-dependent contractions in both DJ-1/park7−/− and DJ-1/park7+/+ mice, although to a greater degree in DJ-1/park7−/− mice than in the control mice (Figure 1A). Similar differences were obtained in mesenteric arteries from DJ-1/park7−/− and DJ-1/park7+/+ mice (Figure 1B). In the endothelium-denuded aorta, NE (0.001–10 μM) induced contraction in a concentration-dependent manner, but there was no difference between DJ-1/park7−/− and DJ-1/park7+/+ mice (see Supplementary material online, Figure S2A). Similarly, high K+ (10–70 mM) also increased endothelium-denuded aortic contraction in a concentration-dependent manner, with a similar level of response in both DJ-1/park7−/− mice and the controls (see Supplementary material online, Figure S2B). Treatment with high K+ (10–70 mM) induced a concentration-dependent contraction in endothelium-intact aortic rings, but this contraction did not show a significant difference between the two groups (see Supplementary material online, Figure S2C).

Figure 1

Changes in responsiveness to vasoconstrictors in vessels from DJ-1/park7−/− mice. Aortic (A) and mesenteric arterial (B) rings from DJ-1/park7−/− and DJ-1/park7+/+ mice were prepared with the endothelium left intact, and contracted with NE (0.001–100 μM). K+-induced contraction (70 mM) obtained immediately before the experiments was expressed as 100%. Aortic (C) and mesenteric arterial (D) rings with endothelial cells were precontracted with NE (0.1 μM) and then treated cumulatively with ACh (0.001–10 μM). The contractile magnitude of NE (0.1 μM) immediately before treatment with ACh was considered as 100%. Each data point represents the mean ± SEM (n = 7). *Significant difference from DJ-1/park7+/+ mice with P < 0.05.

3.2 Impaired vasorelaxation in response to acetylcholine in DJ-1/park7−/− mice

Next, we tested the effects of DJ-1/park7 deletion on endothelial cell-dependent vasorelaxation using DJ-1/park7−/− mice. Acetylcholine (ACh) treatment (0.001–1 μM) inhibited contraction induced by NE (0.1 μM) in a concentration-dependent manner in endothelium-intact aortic rings. This relaxation was significantly impaired in aortas from DJ-1/park7−/− mice compared with those from DJ-1/park7+/+ controls (Figure 1C). Similar differences were obtained in mesenteric arteries from DJ-1/park7−/− and DJ-1/park7+/+ mice (Figure 1D). Furthermore, ACh-induced vasorelaxation was inhibited in the endothelium-intact aorta pre-treated with NG-nitro-l-arginine methyl ester (l-NAME, 0.1 mM) (Figure 2A). The differences in response to ACh between the DJ-1/park7−/− and DJ-1/park7+/+ aorta were abolished by treatment with antioxidants, catalase (50 U/mL), N-acetyl cysteine (NAC, 1 mM), and tempol (1 mM) (Figure 2C and D).

Figure 2

Effects of NOS inhibitor and antioxidants on the changes in vasorelaxation by ACh in aortas from DJ-1/park7−/− mice. Aortic rings from DJ-1/park7−/− and DJ-1/park7+/+ control mice were incubated with l-NAME (0.1 mM, A), catalase (50 U/mL, B), NAC (1 mM, C), or tempol (1 mM, D) for 30 min and then ACh-induced vasorelaxation was examined. The contractile magnitude of NE (0.1 μM) obtained immediately before the experiments was expressed as 100%. Each data point represents the mean ± SEM (n = 7).

3.3 Morphology and function of endothelial cells from DJ-1/park7−/− mice

Given the results from vascular relaxation and contraction, loss of endothelial function in DJ-1/park7−/− mice was expected. Therefore, the correlation between the NO-production mechanism and attenuation of endothelium-dependent vasodilation in DJ-1/park7−/− mice was examined. First, we conducted a morphological analysis using immunostaining of the cross sections of aortas isolated from DJ-1/park7−/− and DJ-1/park7+/+ mice. As shown in Figure 3A, endothelial cells of DJ-1/park7−/− mice showed an expression pattern against CD31 that attached along the tunica intima of vessel walls and was similar to that in control mice. Moreover, the morphology of medial smooth muscle of the aorta was not different between DJ-1/park7−/− and DJ-1/park7+/+ mice. In comparison with eNOS expression in the endothelial cells of both groups of mice, DJ-1/park7−/− mice exhibited weaker expression of eNOS staining compared with DJ-1/park7+/+ mice (Figure 3B). Moreover, in real-time PCR analysis and western blot analysis, the expression levels of eNOS mRNA and protein in endothelial cells were lower in DJ-1/park7−/− mice than in DJ-1/park7+/+ controls (Figure 3C), which was similar to the results from immunofluorescence analysis of cells and en face aortic preparations from mice (Figure 3D).

Figure 3

Difference in characteristics of endothelial cells between DJ-1/park7−/− and DJ-1/park7+/+ mice. (A) Immunohistochemical staining images from aortas with intact endothelium. Endothelium-intact aortic sections were incubated with an anti-CD31 antibody and visualized using DAB staining. Dark brown: CD31-positive cells. Scale bar: 50 μm. The basal level of CD31 expression in DJ-1/park7+/+ mice was expressed as 100% (n = 8). (B) Representative immunocytochemical images showing the expression of eNOS in cultured endothelial cells isolated from the mouse lung. Endothelial cells were incubated with an anti-eNOS antibody and visualized using an Alexa Fluor 488-conjugated secondary antibody. An irrelevant isotype-matched antibody (anti-IgG1) was used as a negative control (IgG control). Green: eNOS-positive cells. Scale bar: 50 μm. (Lower panel) Data were obtained from upper panels (n = 8). (C) Expression analysis of protein and mRNA of eNOS in endothelial cells. (Upper panels) Expression of eNOS protein. The expression level of each protein was examined by western blot analysis using anti-eNOS, anti-CD31, or anti-β-actin antibodies. Basal eNOS expression in DJ-1/park7+/+ mice was expressed as 100% (n = 6). (Lower panel) Quantitative analysis of mRNA eNOS. The mRNA level of eNOS in endothelial cells was examined using real-time PCR (n = 6). (D) Analysis of en face staining of eNOS expression in the endothelium-intact aorta. The aorta segments were treated with an anti-eNOS antibody and then incubated with an Alexa Fluor 488-conjugated goat anti-rabbit antibody. The immunostained segments were mounted on slide glass with the endothelium facing up and observed using a confocal microscope system (n = 4). *P < 0.05.

Next, because DJ-1/park7 inhibits the intracellular level of H2O2 in a variety of cells, we analysed the H2O2 content in endothelial cells to clarify the role of DJ-1/park7 protein in intracellular ROS production. The level of H2O2 was significantly elevated in endothelial cells from DJ-1/park7−/− mice compared with those from DJ-1/park7+/+ (see Supplementary material online, Figure S3A). The elevated production of H2O2 was also confirmed in en face staining of mice aorta (see Supplementary material online, Figure S3B).

3.4 Reversible effects of DJ-1/park7 overexpression on endothelial functions

We further analysed the involvement of DJ-1/park7 protein in the alteration of eNOS expression using a DJ-1/park7-containing vector system. To test the effect of DJ-1/park7 overexpression, endothelial cells from DJ-1/park7−/− mice were transfected with a DJ-1/park7 gene using electroporation. Transfection of the DJ-1/park7-containing vector into endothelial cells from DJ-1/park7−/− mice shows a clear expression of DJ-1/park7 in western blot analysis (Figure 4A). Moreover, the level of eNOS expression was significantly elevated in DJ-1/park7-overexpressing cells compared with cells with the DJ-1/park7-free vector. Given the eNOS expression levels in endothelial cells from DJ-1/park7−/− mice, we speculated that NO production would be diminished in DJ-1/park7−/− mice. We thus compared the level of NO production in endothelial cells from DJ-1/park7−/− mice and DJ-1/park7+/+ controls. As shown in Figure 4B, NO production in the endothelial cells from DJ-1/park7−/− mice was decreased compared with DJ-1/park7+/+ controls. Transfection of the DJ-1/park7 gene into endothelial cells from DJ-1/park7−/− mice enhanced NO production compared with that from the DJ-1/park7-free vector (Figure 4C). The diminished production of NO and expression of eNOS were also observed in en face staining of mice aorta (Figures 4D and see Supplementary material online, Figure S4).

Figure 4

Expression of eNOS and production of NO in DJ-1/park7-overexpressing endothelial cells from DJ-1/park7−/− mice. (A) Effects of DJ-1/park7 overexpression on eNOS expression by endothelial cells from the lung of DJ-1/park7−/− mice. Lysates of DJ-1/park7-overexpressing cells were immunoblotted with an anti-eNOS antibody. The expression in endothelial cells transfected with the DJ-1/park7-free vector was expressed as 100% (n = 8). (B) Comparison of NO production in endothelial cells from DJ-1/park7−/− and DJ-1/park7+/+ mice. The NO level was measured using a DAF-2 assay. The basal level of NO in DJ-1/park7+/+ control mice is expressed as 100% (n = 8). (C) Effects of DJ-1/park7 overexpression on NO production in endothelial cells from DJ-1/park7−/− mice. The basal level of NO in DJ-1/park7−/− cells transfected with DJ-1/park7-free vector was expressed as 100% (n = 8). (D) Comparison of NO production in en face staining of the intact aorta from DJ-1/park7−/− and DJ-1/park7+/+ mice. The NO level was determined using a DAF-2 assay. The basal level of NO in DJ-1/park7+/+ control mice is expressed as 100% (n = 5). *P < 0.05 between DJ-1/park7−/− and DJ-1/park7+/+ mice (B and D) or overexpression and vector control (A and C). Vector, cells transfected with DJ-1/park7-free vector; +OVE, DJ-1/park7-overexpressing cells.

3.5 Epigenetic regulation of eNOS expression in endothelial cells from mice

We tested whether eNOS expression is regulated by epigenetic modification of histone protein in DJ-1/park7−/− mice. The histone acetylation and its deacetylase recruitment in eNOS promoter regions (−291 to −138) were analysed using a ChIP assay (Figure 5A). HDAC-1 recruitment in eNOS promoter regions was significantly increased in endothelial cells from DJ-1/park7−/− mice compared with DJ-1/park7+/+ mice (Figure 5B). Moreover, the acetylation of H3 histones in eNOS promoter regions was inhibited in DJ-1/park7−/− mice compared with in DJ-1/park7+/+ (Figure 5C). These results were similar to the enrichment levels of HDAC-1 and acetylated H3 histone obtained from conventional PCR analysis (Figure 5A).

Figure 5

Epigenetic modification of eNOS promoter in mice endothelial cells. (A) Schematic diagram for eNOS promoter used for the ChIP assay (upper panel) and the representative gel electrophoresis images from conventional PCR showing enrichment of HDAC-1 and acetylated H3 histone at the eNOS promoter (lower panels). Numbers indicate the nucleotide location starting from the transcription start site (TSS). The precipitates obtained from immunoprecipitation reaction (IP) of chromatin and anti-HDAC-1, anti-acetylated H3 histone, or anti-IgG antibody were amplified along with input DNA (Input). IgG, irrelevant antibody control. (B and C) Chromatin was immunoprecipitated with antibodies against HDAC-1 (B) and acetylated H3 histone (C). The precipitates were amplified by quantitative real-time PCR. The values were normalized by that of input. The relative level of eNOS mRNA was calculated using ΔCt by normalizing that of GAPDH. (D and E) Effects of H2O2 on eNOS expression and HDAC-1 recruitment. Endothelial cells from DJ-1/park7+/+ mice were treated with H2O2 (100 μM) for 24 h. The levels of eNOS expression and HDAC-1 recruitment were measured by western blot analysis and ChIP assay (n = 6). (F) Effects of HDAC-1 knockdown on eNOS expression in endothelial cells from DJ-1/park7–/– mice. Cells were transfected with 200 pM non-silencing siRNA (Non) and HDAC-1 siRNA (siRNA). HDAC-1 and eNOS expressions were confirmed by western blotting in the transfected cells (n = 4). *P < 0.05.

Next, we defined the effects of H2O2 on histone modification of eNOS in endothelial cells. H2O2 (100 μM) treatment decreased eNOS expression in endothelial cells from DJ-1/park7+/+ mice (Figure 5D). H2O2 increased HDAC-1 recruitment in the endothelial cells (Figure 5E). To confirm these results, we determined the expression of eNOS in HDAC-1-knocked down endothelial cells from DJ-1/park7–/– mice. The level of HDAC-1 was decreased dramatically in endothelial cells from DJ-1/park7–/– mice that were transfected with HDAC-1 small interfering RNA (siRNA; Figure 5F). Moreover, HDAC-1 siRNA significantly increased the expression of eNOS in DJ-1/park7–/– endothelial cells (144.6 ± 7.4% of control). These epigenetic characteristics in mouse endothelial cells were also explored in human umbilical vein endothelial cells (see Supplementary material online, Figure S6).

3.6 Alteration of blood pressure in DJ-1/park7−/− mice

Subsequently, we tested whether DJ-1/park7 protein and its related eNOS/NO mechanisms are involved in the regulation of blood pressure. SBP in conscious mice was measured at 12–72 weeks of age using a tail-cuff apparatus. Neither DJ-1/park7−/− mice nor DJ-1/park7+/+ controls showed any significant alteration in SBP by 28 weeks of age. However, SBP was elevated in DJ-1/park7−/− mice at 44–72 weeks of age, to become significantly higher than in DJ-1/park7+/+ mice (Figure 6A). Moreover, administration of valproic acid (VPA), an inhibitor of Class I HDACs including HDAC-1 (175 mg/kg body weight/day), attenuated the elevated blood pressure in DJ-1/park7−/− mice (Figure 6C), but did not in DJ-1/park7+/+ mice (Figure 6B).

Figure 6

Changes of blood pressure in DJ-1/park7−/− mice. (A) Difference of blood pressure between DJ-1/park7−/− and DJ-1/park7+/+ mice. SBP of DJ-1/park7−/− and DJ-1/park7+/+ mice was measured at 12–72 weeks of age using a tail-cuff method as described in the Methods section (n = 10). (B and C) Effects of VPA (175 mg/kg body weight/day) on blood pressure in 57-week-old DJ-1/park7+/+ (B) and DJ-1/park7−/− (C) mice. The SBP of each mouse was obtained from the average of four consecutive measurements. Data represent the mean ± SEM of seven independent measurements. *P < 0.05 between the control and VPA-treated group.

4. Discussion

In the present study, we found that Ach evoked endothelium-dependent relaxation in the aorta, and that this response was greatly impaired in DJ-1/park7−/− mice compared with DJ-1/park7+/+ controls. This result implies that DJ-1/park7 protein may participate in endothelial cell-dependent vasorelaxation. We also found that DJ-1/park7−/− mice showed a significant decrease in eNOS expression and NO production, without morphological deficiency of endothelial cells. Moreover, the eNOS expression and NO production were reversed in endothelial cells from DJ-1/park7−/− mice overexpressing DJ-1/park7. ACh stimulates eNOS activation and produces NO, which leads to activation of guanylate cyclase resulting in SMC relaxation.1 Endothelial dysfunction is characterized by reduced responsiveness to endothelium-dependent vasodilators.24 In this study, the endothelium-independent relaxation in response to SNP and verapamil did not show any difference between DJ-1/park7−/− and DJ-1/park7+/+ mice, implying that there is no alteration in SMC membrane Ca2+ channel and sensitivity to NO in DJ-1/park7−/− mice. Furthermore, vessels from DJ-1/park7−/− mice revealed increased contractility with NE under endothelium-intact conditions, but not in an endothelium-denuded state, compared with that of DJ-1/park7+/+ controls. Responses of adrenoceptors to NE occur on the plasma membranes of both endothelial cells and SMCs.25 The binding of agonists to adrenoceptors in endothelial cells releases NO via the stimulation of eNOS activity, but in SMCs evoke contraction via the activation of phospholipase C and Ca2+ influx.25 Moreover, the present result showed that the difference in response to high K+ was not observed between two animal groups, which support the previous reports that high K+ activates the membrane of SMCs, but not that of endothelial cells. Therefore, these results suggest that both the diminished ACh- and increased NE-induced responsiveness may result from the down-regulation of the eNOS/NO pathway in DJ-1/park7−/− endothelial cells. Although DJ-1/park7 is known to be a multifunctional protein in a variety of cells, this is the first report, to our knowledge, that describes the functional roles of DJ-1/park7 in endothelial cell-related vasorelaxation.

ROS play an important role in a number of cells and disease types. Excess production of ROS is a common cause of vascular disease. It has been reported that a reduced level of vascular NO can be associated with the production of ROS including H2O2.4 In the present study, the level of H2O2 was higher in endothelial cells from DJ-1/park7−/− mice compared with DJ-1/park7+/+ and treatment of endothelial cells with H2O2 decreased the expression level of eNOS, implying that the increased level of H2O2 may be responsible for the dysfunction of endothelial cells from DJ-1/park7−/− mice. H2O2 decreases eNOS promoter activity and its expression in endothelial cells, eventually leading to endothelial dysfunction and decreased NO bioavailability.8 Moreover, treatment of endothelial cells with H2O2 suppressed the production of NO. However, previous studies from other groups have also demonstrated that H2O2 increased eNOS expression in endothelial cells,26,27 indicating that the mechanism on H2O2-regulated eNOS expression may be complex and be altered by various conditions, including different concentrations and times in H2O2 treatment, as well as different endothelial cell types. Although, in the present study, we observed that eNOS expression decreased by H2O2 treatment, which can be reasonable, a further study will be required to elucidate more detailed mechanism linking the interaction between H2O2 and eNOS expression. Furthermore, our results showed that the levels of eNOS protein and mRNA were lower in endothelial cells from DJ-1/park7−/− mice than in those from DJ-1/park7+/+. Moreover, the acetylation of H3 histone at the eNOS promoter was reduced, and HDAC-1 recruitment increased in DJ-1/park7−/− mice compared with DJ-1/park7+/+. Acetylation of histones is regulated by the balance of its acetyltransferase and HDAC activities.28 HDACs regulate the density of histone–DNA contacts via the removal of acetyl groups of histone protein, and they make DNA more accessible for the transcriptional activation of genes.29 Moreover, there is a large body of evidence that ROS are involved in the activity of HDAC-1 and acetylation of histone protein, implying that this modification is tightly regulated by cellular ROS.10 In the present study, we also observed that H2O2 increased HDAC-1 recruitment and HDAC-1 knockdown induced enhancement of eNOS expression in endothelial cells. On the other hand, it is reported that HDAC-1 inhibition decreased eNOS expression.30,31 An inhibitor of Class I and II HDACs, trichostatin A, is known to have dual effects on eNOS mRNA expression, which at the low-dose trichostatin A increased eNOS expression and then dose-dependently showed an inhibition.32 Therefore, these results indicate that HDAC inhibition may exert complex effects on the eNOS expression.33 Taking these results together, we suggest that the diminished eNOS in DJ-1/park7−/− mice is caused by decreased H3 histone acetylation of the eNOS promoter, which is mediated by up-regulation of HDAC-1 probably in response to H2O2. This may result in functional impairment of the endothelial eNOS/NO pathway (see Supplementary material online, Figure S7).

DJ-1/park7 is an atypical peroxiredoxin-like peroxidase that ameliorates antioxidative stress by scavenging H2O2.15,16 We and other groups have demonstrated that overexpression of DJ-1/park7 protects neural cells against oxidative stress-induced injury.17,18,34 As described previously, we showed that the level of H2O2 in endothelial cells was predominantly enhanced in DJ-1/park7−/− mice compared with DJ-1/park7+/+. These findings indicate that DJ-1/park7 may act as a potential antioxidant to reduce ROS, especially H2O2, in vascular cells. ROS can lead to endothelial dysfunction and elevation of blood pressure.35 Moreover, endothelial dysfunction, e.g. decrease in NO bioavailability and impaired eNOS expression, is implicated in hypertension.36 The abnormality in vascular contractility is tightly connected to the development of hypertension in animals and humans.21 Furthermore, epigenetic modification in endothelial cells is associated with eNOS expression in pulmonary hypertensive models.37 Elevated HDAC activity is involved in the vascular pathogenesis in hypertension,23 and the administration of an HDAC-1 inhibitor inhibits blood pressure in SHR.38,39 In the present study, SBP was significantly higher in DJ-1/park7−/− mice than in DJ-1/park7+/+, but this elevation was abolished by treatment with an HDAC-1 inhibitor. Similarly, renal DJ-1/park7 is considered to be a possible contributor in the regulation of oxidative stress associated with ROS-dependent hypertension.20 Therefore, it is assumed that endothelial dysfunction in DJ-1/park7−/− mice could result from enhanced ROS-mediated epigenetic modification of the eNOS promoter, which may be implicated in the blood pressure elevation in the mice. Moreover, in the present study, the epigenetic characteristics and ROS-induced responses in eNOS expression in human endothelial cells showed a similarity to those in mice endothelial cells. Based on all these results, we suggest that DJ-1/park7−/− mice will serve as a new animal model for human hypertension, where blood pressure is elevated through epigenetic abnormalities in endothelial cells.

In summary, we demonstrated that ACh-evoked vasorelaxation and NE-induced vasoconstriction were significantly inhibited and elevated, respectively, in aortas and mesenteric arteries with intact endothelium from DJ-1/park7−/− mice compared with DJ-1/park7+/+. Levels of eNOS and NO production were decreased in endothelial cells from DJ-1/park7−/− mice compared with DJ-1/park7+/+ controls, and these decreases were reversed in DJ-1/park7-overexpressing cells. Acetylation of H3 histones at the eNOS promoter was reduced, and HDAC-1 recruitment increased in DJ-1/park7−/− mice compared with DJ-1/park7+/+ controls. Moreover, SBP showed a significant increase in DJ-1/park7−/− mice compared with DJ-1/park7+/+ control mice, but this elevation was decreased in mice treated with an HDAC-1 inhibitor. Therefore, these results suggest that DJ-1/park7 protein may play an important role in the regulation of vascular contractility and blood pressure, probably through the impairment of NO production as a result of H2O2-mediated epigenetic inhibition of eNOS expression.

Funding

This research was supported by Basic Science Research Program through the National Research Foundation of Korea, NRF, funded by the Ministry of Education, Science, and Technology (313-2008-2-E00042; NRF-2011-0029583; NRF-2012R1A1A2009513).

Acknowledgements

This study is based on a part of a Master thesis (Soo Hyun Jung) at Konkuk University.

Conflict of interest: none declared.

Footnotes

  • K.J.W. and Se.H.J. contributed equally.

References

View Abstract