NADPH oxidase activation played a critical role in the oxidative stress process in stable coronary artery disease

Abstract

Objectives: The study was designed to investigate the oxidative stress levels of endothelial progenitor cells (EPCs) in stable coronary artery disease (CAD) and to explore the underlying mechanisms of NADPH oxidase activation and subsequent EPCs dysfunction. Methods: EPCs were isolated from patients with stable CAD (n=50) and matched healthy volunteers (n=50). NADPH oxidase activation was detected by measuring the expression of each subunit using western blotting and qPCR analyses and the membrane translocation of p47^phox using immunofluorescence. The in vivo angiogenesis capacity was evaluated using immunofluorescence by transplanting EPCs into a rat hind limb ischemia model. The PKC inhibitor GÖ-6983 was used to determine the role of PKC in NADPH oxidase activation. Results: Oxidative stress level was increased and the in vivo angiogenesis capacity was impaired in EPCs obtained from CAD subjects with the activation of NADPH oxidase. P47^phox membrane translocation increased in CAD group vs controls. These effects were resolved by NADPH oxidase inhibition. Up-regulation of PKCα/β2 was found in EPCs from CAD subjects, PKC inhibition GÖ-6983 could reduce the expression and activity of NADPH oxidation. Conclusions: NADPH oxidase activation via p47^phox membrane translocation played a critical role in the initiation and progression of CAD, and the PKCα/β2 signaling pathway might be involved.

Introduction

It has been suggested that EPCs might not only be responsible for the recovery of the endothelium after injury but also contribute to angiogenesis, providing the hope of new treatment opportunities [1]. Indeed, there is accumulating evidence demonstrating a reduction in the number of EPCs and impaired EPC function in the presence of cardiovascular disease [2]. The most important mechanism is the diverse reactive oxygen species (ROS) produced at increased levels [3]. Excessive ROS production severely impaired EPC mobilization, migration and proliferation, as well as its vasculogenic capacity. However, the specific mechanism is not completely clear. In stem/progenitor cells, ROS is mainly derived from NADPH oxidase [4]. Previous studies have shown that in EPCs obtained from diabetes, NADPH oxidase activation via p47^phox membrane translocation and binding to the subunit Nox2 or Nox4 can cause increased production of ROS, decreased nitric oxide (NO) bioactivity and re-endothelialization dysfunction [5]. However, in EPCs derived from CAD, the effects of NADPH oxidase activation on the vasculogenic capability remain unknown.

In addition, there is increasing evidence that PKC activation is important in diabetes-related endothelial dysfunction and in the activities of receptors mediating smooth muscle contraction via activation of NADPH oxidase [6]. However, only minimal attention has been given to defining the role of PKC systems in the activation of NADPH oxidases. In EPCs derived from CAD, whether NADPH oxidase activation is mediated via the PKC signaling pathway has not been elucidated to date.

The current study aimed to explore the NADPH oxidase activation process and its underlying mechanisms, as well as the changes in the vasculogenic capacity resulting from NADPH oxidase activation in EPCs from CAD.

Materials and methods

Characteristics of the patients and controls

Fifty stable CAD patients were consecutively enrolled. The control group consisted of fifty volunteers who were matched for gender and age with the CAD patients, a clinical evaluation was performed to ensure the healthy status of the control subjects. The clinical details of the studied population are shown in Table 1. The diagnosis of stable CAD was based on clinical assessment, echocardiography, and angiography. Patients who fulfilled the following inclusion criteria were enrolled in the study: previously diagnosed CAD without angina, or symptom complex that has remained stable for at least 60 days (No change in frequency, duration, precipitating causes or ease of relief of angina for at least 60 days); And no evidence of recent myocardial damage; Percutaneous coronary intervention (PCI) in 1 or 2 or 3 major coronary arteries. Exclusion criteria were previous coronary bypass surgery, significant coronary artery disease in all the major epicardial coronary arteries or left main coronary stenosis ≥50%, unstable angina, and acute myocardial infarction. All patients and controls had no diabetes and were free of wounds, ulcers, retinopathy, recent surgery, inflammatory, malignant diseases and acute coronary syndrome (ACS). All the subjects were in accord with the ethical standards established by the Human Research Protections (OHRP). Informed consent was obtained from all subjects, and all of the procedures were performed according with national and international laws and policies.

Table 1.

Baseline characteristics

Characteristics	Coronary heart disease (N=50)	Control subjects (N=50)	P value
Age, y	58.9 ± 5.7	60.5 ± 5.4	NS
Sex, % men	64	59	NS
Smokers, %	35	31	NS
Hypertension, %	60	57	NS
Diabetes, %	0	0	NS
Cholesterol (total), mg/dL	194 ± 56	167 ± 40	P>0.05
Triglycerides, mg/dL	162 ± 69	150 ± 47	P>0.1
Severity (1/2/3 vessel disease), %	67, 26, 7	0
History of taking statin, %	0	0	NS

Values are percentages or mean ± SD.

EPC cultures and identification of late-EPCs

EPCs were isolated, cultured and characterized according to previously described techniques [7]. Briefly, mononuclear cells (MNCs) were isolated from the peripheral blood of patients or control subjects using Ficoll density gradient centrifugation and were plated onto six-well plates coated with human fibronectin (Chemicon, USA) and maintained in endothelial basal medium (EBM)-2 (Clonetics, Lonza, USA) supplemented with 10% fetal-calf serum (FBS) (Gibco, Invitrogen, USA), and other cytokines. After 48 h in culture, non-adherent cells were removed and re-plated onto six-well plates. The culture was maintained for seven days to obtain the colony-forming unit endothelial cells (CFU-EC). Adherent cells were cultured for 2-4 weeks to obtain late endothelial progenitor cells (L-EPCs). L-EPCs were incubated with 1,1-dioctadecyl-3,3,3,3-tetramethylindocarbocyanine (DiI)-labeled acetylated low-density lipoprotein (acLDL; Molecular Probe) and stained with FITC-labeled Ulex europaeus agglutinin (UEA-1; Sigma). After staining, the samples were visualized using a laser scanning confocal microscope (LSCM, Leica). Cells demonstrating double-positive fluorescence were identified as L-EPCs. Moreover, they were shown to express KDR, CD34 (Becton Dickinson, USA) and AC133 (Miltenyi Biotec, Germany) using flow cytometry (Supplementary Figures 1, 2 and 3).

ROS measurement

EPCs were incubated in EGM2 with 10 μM 2’,7’-dichlorodihydrofluorescein diacetate (H2DCF-DA) (Molecular Probes) for 1 h at 37°C in the dark. Next, the cells were trypsinized and washed three times with PBS before the reaction was terminated. The Mean Fluorescence Intensity (MFI) of the cells were measured using FACS analysis. No fluorescent probe samples served as the negative control [8].

Measurement of NADPH oxidase activity

The NADPH oxidase activity was measured using the lucigenin-enhanced chemiluminescent detection of superoxide anions [9]. The EPC lysates were mixed with phosphate buffer (50 mM, pH 7.0) containing ethylene glycol tetraacetic acid (EGTA, 1 mM), sucrose (150 mM), lucigenin (the electron acceptor, 5 μM) and NADPH (the substrate, 100 μM). Photon emission was measured every 2 minutes for 4 times using a chemiluminescence analyzer (Biotek, USA).

Malondialdehyde (MDA) assay

The medium was removed and the EPCs were gently washed with ice-cold PBS three times and lysed in cell lysis buffer. The lysate was centrifuged at 10,000×g for 5 min. The supernatant was used to evaluate lipid peroxidation, which was determined using a lipid peroxidation MDA assay kit (Beyotime Co., Shanghai, China) [10].

Quantitative RT-PCR

Total RNA was extracted using TRIzol (Invitrogen, Carlsbad, CA, USA). cDNA synthesis was performed with 1 μg of total RNA using the miScript II RT Kit (Qiagen, Germany) according to the manufacturer’s instructions. Sequences of the primers sequences can be found in Table 2. Real-time PCR was performed on the ABI 7500 cycler (Applied Biosystems, CA, USA), using the miScript SYBR Green PCR Kit (Qiagen, Germany) according to the manufacturer’s protocol. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an endogenous control for mRNA expression.

Table 2.

List of primers used in qRT-PCR

	Sense primers (5’-3’)	Antisense primers (5’-3’)
gp91phox	TACCTGGCTGTGACCCTGTT	GGTTTTGGTGGAGGAAGTGA
p47phox	ACGAGACGGAAGACCCTGAG	GACGGGAAGTAGCCTGTGAC
p22phox	GTACTTTGGTGCCTACTCCA	CGGCCCGAACATAGTAATTC
Nox4	CAGGAGGGCTGCTGAAGTATCAA	TGACTGGCTTATTGCTCGGATA
GAPDH	GGGTGTGA CCATGAGAAGT	GACTGTGGTCATGAGTCCT

Western blotting analyses and antibodies

EPCs were scraped in RIPA lysis buffer (Beyotime, Shanghai, China) that was supplemented with 1 mM PMSF. The primary antibodies used in the current study included Anti-p22^phox, anti-Nox4, anti-pPKCα/β1/β2 were purchased from Cell Signaling Technology (Beverly, MA, USA). Anti-gp91^phox and anti-p47^phox were purchased from Abcam Ltd. (MA, USA). Anti-GAPDH polyclonal antibody was purchased from Santa Cruz Biotechnology (CA, USA). Secondary antibodies conjugated to horseradish peroxidase were purchased from Cell Signaling Technology. The images were captured on the image reader LAS-4000 system (Fujifilm, Tokyo, Japan).

Small-interfering RNA transfection

NADPH oxidase subunit expression were silenced with the use of Validated Stealth RNAi (Invitrogen, Carlsbad, Calif), siRNA transfection was carried out according to the manufacturer’s protocol. The small-interfering RNA (siRNA) sequences (sense strands) used for targeting human p47^phox were 5’-CGGAGAGACGAGUAUAACCCAGUUU-3’, for targeting Nox2 were 5’-GGCUGUGCAUAAUAUAACATT-3’, for targeting Nox4 were GGCAGGAGAACCAGGAGAUUTT-3’. In brief, EPCs were transferred into 6-well plates and cultured in EM2 supplemented with 20% FBS. After overnight growth, the cells at 80% adherence were treated with 0.5 ml of Opti-MEM I (Invitrogen, Canada) with each well containing 50 nmol/l siRNA duplexes and 5 µl of Lipofectamine2000. In addition, an RNAi Negative Control Duplex was used as a negative control. All siRNA transfections were performed for 24 hours preceding subsequent EPCs measurements.

Mobilization of p47^phox subunits to the plasma membrane

Translocation of p47^phox subunits to the membrane was analyzed using immunofluorescence [11]. Briefly, the cells were cultured on coverslips and permeabilized by incubating for 15 minutes on ice with 2 ml of 0.1% Triton X-100 in PBS. After incubating for 1 hour with blocking buffer (DAKO) (Abcam Ltd., MA, USA), p47^phox primary antibodies (9.4 μg/ml) were introduced to the sample and incubated at 4°C overnight. The cells were washed with PBS for 5 minutes and repeated five times. The cells were incubated in fluorescein-conjugated secondary antibodies (DyLight 594 AffiniPure Goat Anti-Rat IgG) in blocking buffer in a dark humidity chamber at 4°C for 1 hour. After several washes with PBS, the samples were detected using a fluorescence microscope.

Animals and in vivo vasculogenic capacity assay

Twenty male Sprague-Dawley rats weighing 200-220 g (from Zhejiang University experimental animal center) were randomly divided into a control group and a CAD group (n=10 for each group). All procedures were performed according to the Guide of Zhejiang University Animal Care and Use Committee, and approval was granted by the university ethics review board. A rat hind limb ischemia model was generated by a femoral artery ligation. SD rats were anesthetized with intraperitoneal injection of 8% chloral hydrate (0.5 ml/100 g). Saline, control group EPCs and CAD EPCs labeled with DiI (1,1’-dioctadecyl-3,3,3’,3’-tetramethylindocarbocyanine perchlorate) were injected into the ipsilateral gastrocnemius. Laser Doppler Flowmetry PeriFlux 5000 (Perimed, Sweden) was used for the blood flow measurements. An optic probe was positioned and fixed at a right angle above the medial side of the ligated limb. By receiving the optical fiber, the information is recorded and converted into electrical signals for analysis. Measurements of PU (Perfusion Unit) were obtained before the occlusion, immediately after the transplantation, and 15 d, 28 d after transplantation. The data were analyzed using PeriFlux 5000-PC-Software.

Transplanted EPCs to participate in angiogenesis of new blood vessels were evaluated by immunofluorescence. 28 days after transplantation, the gastrocnemius muscle tissue from the hind limb ischemia model of both the CAD/control subjects were fixed in 4% formalin and was embedded in paraffin. Sections were stained with FITC-labeled lectin griffonia simplicifolia (Sigma, USA) and 4,6-diamino-2-phenyl indole (DAPI). Homing of the transplanted EPCs to the site of vascular injury were detected using fluorescence microscopy. The mean values of the neovascularization numbers were used for the analysis [5].

Statistical analysis

All experiments were performed at least three times. All data were presented as mean ± standard deviation (SD). Two groups were compared using t-test comparison and One way analysis of variance (ANOVA) with a one-tailed Student’s t-test was used for multiple comparisons among groups. Statistical analysis was done with SPSS version 18.0. P-values <0.05 were considered statistically significant.

Results

NADPH oxidase activity, ROS production and MDA levels

We evaluated the NADPH oxidase activity, ROS and MDA levels in EPCs from CAD subjects and normal control as described above. These results showed that NADPH consumption was enhanced from 5.4 to 10.4 nmol/mg protein/min (P<0.05), the ROS levels were significantly increased by 63.5% in EPCs from CAD (P<0.05). Concentrations of MDA were elevated by 131% in EPCs from CAD subjects (P<0.05) (Figure 1). All of these indicated that the oxidative stress levels of EPCs in stable coronary artery disease were significantly increased.

Figure 1.

NADPH oxidative activity, ROS and MDA levels in two groups. A. NADPH consumption were enhanced from 5.4 to 10.4 nmol/mg protein/min in EPCs from CAD versus control groups, n=11, 12, *P<0.05. B. ROS production were significantly increased by 63.5% in EPCs from CAD versus control groups, n=13, 15, *P<0.05. C. The concentrations of MDA were elevated by 131% in EPCs from CAD subjects versus control groups, n=14, 17, *P<0.05. D. The main source of ROS in CAD-EPCs. EPCs from CAD subjects were incubated with apocynin (100 μmol/L), L-NAME (1 mmol/L), oxypurinol (100 μmol/L), rotenone (250 μmol/L) and p47^phox siRNA, EPCs from CAD subjects without intervention were used as the control group, n=6 per group. ROS production was significantly reduced in the group of apocynin, L-NAME and p47^phox siRNA (*P<0.05).

Increased expression of NADPH oxidase subunits

The mRNA expression of NADPH oxidase subunits were detected using qPCR. These results indicated that the mRNA expression of p22^phox, p47^phox, gp91^phox and Nox4 significantly increased in EPCs from CAD group (P<0.05). In addition, the protein expression of subunits were detected using western blotting analyses, and showed the expression of p22^phox, p47^phox, gp91^phox and Nox4 were increased consistently in EPCs from CAD subjects (Figure 2).

Figure 2.

The mRNA and protein expression of NADPH oxidase subunit levels in CAD and control groups, n=8, 8. A. Fold changes of p22^phox, p47^phox, gp91^phox and Nox4 subunits mRNA were significantly increased in CAD group versus controls, *P<0.05. B. Protein expression of gp91^phox, p22^phox and p47^phox were increased in CAD group (C1, C2) versus controls (N1, N2), *P<0.05.

p47^phox membrane translocation

The translocation of p47^phox subunits of NADPH oxidase from the cytosol to the plasma membrane is an essential step in the activation of NADPH oxidase. p47^phox positioning within EPCs was analyzed using indirect immunofluorescence. Anti-p47^phox was labeled with Dyl594 fluorescent secondary antibody (red), and the nuclei were stained with DAPI (blue). In control group, p47^phox was evenly distributed in the cell. However, in CAD group, p47^phox translocated from the cytoplasm to the plasma membrane, demonstrating uneven distribution, with enhanced distribution of the inner membrane and weaker cytoplasmic distribution, which reflected activation of p47^phox in EPCs from CAD (Figure 3A, 3B). Calculating the cell numbers which p47^phox was activated per unit area in EPCs, we can evaluate the p47^phox activation rate in each groups. p47^phox activation rate was 31.6% in the CAD group, while 17.8% in control group , activation rate in CAD was 1.78 times higher compared to the control group, N=10, 10, *P<0.05 (Figure 3C).

Figure 3.

Immunofluorescence of p47^phox membrane translocation. Anti-p47^phox was labeled with Dyl594 fluorescent secondary antibody (red), and the nuclei were stained with DAPI (blue). A. In control group, p47^phox was evenly distributed in the cell (arrow). B. In CAD group, p47^phox translocated from the cytoplasm to the plasma membrane, demonstrating uneven distribution, with enhanced distribution of the inner membrane and weaker cytoplasmic distribution (arrow) (×400). C. Statistical graph of p47^phox activation rate. p47^phox activation rate was 31.6% in the CAD group, while 17.8% in control group , activation rate in CAD was 1.78 times higher compared to the control group, n=10, 10, *P<0.05.

In vivo angiogenesis capacity of EPCs was severely impaired in CAD subjects

We evaluated EPC angiogenesis in vivo by transplanting EPCs into a rat hind limb ischemia model as described above. Prior to the operative ligation of the femoral artery, hind limb perfusion of the rats was not impaired and the base flow were nearly the same in the CAD and control group (n=10, 10). After occlusion, two groups of rats received a local injection of 1×10⁵ transduced EPCs (control-EPCs, and CAD-EPCs). Examination of the hind limb perfusion was performed at day 15 and 28 by Laser Doppler Flowmetry PeriFlux 5000 as described above. One of the output parameters perfusion unit (PU) was recorded. The blood flow in rats transplanted with CAD-EPCs indicated poorer perfusion compared with those rats transplanted with control-EPCs at day 15 and 28 (100.28 PU versus 27.9 PU and 153.6 PU versus 110.6 PU) (P<0.05) (Figure 4A). Transplanted EPCs marked with DiI (1,1’-dioctadecyl-3,3,3’,3’-tetramethylindocarbocyanine perchlorate) were identified in tissue sections using red fluorescence. In contrast, the vasculature, stained using FITC, was identified using green fluorescence in the same tissue sections. Rats transplanted with CAD-EPCs demonstrated a decreased number (20 vs 12.2/mm²) (P<0.05) of DiI-labeled, red fluorescent EPC-derived vasculature compared to control groups at day 28 (Figures 4B, 5).

Figure 4.

The blood perfusion in rat hind limb ischemia model of two groups. A. The blood perfusion in rats transplanted with CAD-EPCs indicated poorer perfusion compared with those rats transplanted with control-EPCs at day 15 and 28 (100.28 PU versus 27.9 PU and 153.6 PU versus 110.6 PU), n=10, 10, *P<0.05. While EPCs treated with apocynin and transplanted to a rat hind limb ischemia model showed a restored perfusion vs CAD group at day 15 (70 PU versus 30.4 PU), n=10, 10, #P<0.05, but not at day 28. B. Bar graph summarizing the vascular engraftment by EPCs. It demonstrated a decreased number (20 vs 12.2/mm²) of Dil-labeled, red fluorescent EPC-derived vasculature compared to control groups at day 28, *P<0.05. But EPCs treated with apocynin and transplanted to a rat hind limb ischemia model resulted in a restored vascular engraftment vs CAD group (17.8 vs 12.2/mm²), #P<0.05.

Figure 5.

Immunohistochemistry image of vascular engraftment by EPCs. Rats transplanted with CAD-EPCs showed a low degree of vascular incorporation, while rats transplanted with CAD-EPCs which were treated with apocynin resulted in a restored vascular engraftment vs CAD group.

NADPH oxidase inhibition restores the generation of ROS and MDA and the in vivo angiogenesis capacity in EPCs from CAD subjects

Antioxidant NAC (3 mmol/L), apocynin (100 μmol/L) and subunit siRNA (50 nM) were used to inhibit the NADPH oxidase activation. First, subunit p47^phox, gp91^phox, Nox4 protein expression levels were reduced significantly by siRNA treatment, compared to the non-silenced group (Figure 6A). Either siRNA silencing of the NADPH oxidase subunit or inhibition by NAC and apocynin resulted in a markedly reduced NADPH oxidase activity and ROS generation in EPCs from CAD individuals (Figure 6B, 6C). However, siRNA silencing did not decrease the MDA concentration while NAC and apocynin do reduced the concentration of MDA (Figure 6D).

Figure 6.

NADPH oxidase inhibition restores the generation of ROS and MDA. A. p47^phox, gp91^phox, Nox4 siRNA knockdown (KD). Western blot confirms that p47^phox, gp91^phox, Nox4 expression levels were reduced by approximately 60%, 85% and 65% following siRNA treatment. B. Relative NADPH activity 24 h after p47^phox, gp91^phox, Nox4 KD and inhibitors. Either siRNA silencing or inhibition by NAC and apocynin resulted in a markedly reduced NADPH oxidase activity in EPCs from CAD individuals, *P<0.05. C. ROS production were also significantly decreased in CAD-EPCs pretreated with NAC, apocynin and siRNA silencing compared with CAD-EPCs without intervention, *P<0.05. D. Relative MDA concentration 24 h after p47^phox, gp91^phox, Nox4 KD and inhibitors. siRNA silencing did not decrease the MDA concentration while NAC and apocynin do reduced the concentration of MDA, *P<0.05.

In addition, inhibition of NADPH oxidase by apocynin partially reversed the blood perfusion at 15 d (70 PU versus 30.4 PU) and the number of neovascularization (17.8 vs 12.2/mm²) compared to CAD group (Figures 4 and 5), which suggesting a critical role of NADPH oxidase activation in vivo angiogenesis capacity of EPCs.

The role of PKC in the activation of NADPH oxidase in EPCs from CAD subjects

We first examined the phosphorylation of PKCα, PKCβ1, PKCβ2 in CAD-EPCs and determined that PKCα and PKCβ2 phosphorylation were significantly enhanced in CAD-EPCs compared to the normal control group, but no significant changes were observed in PKCβ1 (P<0.05) (Figure 7). Next, GÖ6983 (catalog number G1918, Sigma) was used to inhibit the total activity of PKC. CAD-EPCs that without treated with the inhibitors were used as the control group. After PKC inhibition, both NADPH oxidase activity and ROS generation were reduced in EPCs from CAD (P<0.05) (Figure 8A). Furthermore, the expression of subunit p22^phox, p47^phox, gp91 and Nox4 were decreased by PKC inhibition significantly (P<0.05) (Figure 8B).

Figure 7.

The phosphorylation of PKC subunits. PKCα/β2 phosphorylation were significantly up-regulated in CAD-EPCs vs control group, but no significant changes were observed in PKCβ1, n=12, 14, *P<0.05.

Figure 8.

Effect of GÖ6983 on NADPH oxidase activity, ROS and the expression of NADPH oxidase subunits. A. CAD-EPCs were incubated in GÖ6983 for 30 min, CAD-EPCs that without treated with the inhibitors were used as control, both NADPH oxidase activity and ROS generation were reduced after PKC inhibition, n=10, 10, *P<0.05. B. Western blotting showed the expression of subunit p22^phox, p47^phox, gp91^phox and Nox4 were decreased by GÖ6983 vs CAD control, n=8, 8, *,#P<0.05, **,##P<0.05.

Discussion

Many studies have demonstrated that a variety of risk factors, including oxidative stress, may play different roles in stable coronary artery disease and ACS. CAD differs from ACS in terms of the release of chemoattractants and subsequent bone-marrow cell mobilization [12]. The number and function of circulating EPCs in stable CAD might more appropriately to represent the individual’s capacity for endothelial repair. Thus, the current study will mainly focus on the effect of oxidative stress and the role of NADPH oxidase in EPCs with stable coronary artery disease.

It is widely established that oxidative stress plays a pivotal role in the development and pathogenesis of CAD and its complications [13]. Several studies have demonstrated that oxidative stress can result in the development of cardiovascular diseases by reducing the number of EPCs and impairing EPCs function [14]. Among these factors, NADPH oxidase plays an essential role [15]. Our results showed in EPCs isolated from CAD patients, NADPH oxidase activity, ROS and MDA levels were significantly increased, which indicated high oxidative stress levels in EPCs from CAD patients, and was consistent with previous studies. In fact, there have been many studies that have shown NADPH oxidase is the main regulatory enzyme of ROS, which can be used as an indicator of oxidative stress. Several potential sources of ROS have been implicated in endothelial physiology and pathophysiology, including the mitochondrial electron transport chain, xanthine oxidase, cytochrome P-450 enzyme, uncoupled nitric oxide synthase (NOS) and NADPH oxidase [16]. Our study showed that in EPCs from CAD, ROS was primarily derived from NADPH oxidase and may partially correlate with endothelial nitric oxide synthase uncoupling, which indicated that the high levels of oxidative stress in CAD is indeed closely interrelated with NADPH oxidase activation.

NADPH oxidase includes five sub-fractions, which are membrane components gp91^phox (Nox2) and p22^phox, p47^phox, p67^phox and Rac are cytoplasmic components [17]. Their distribution is highly tissue-specific, different members of the Nox family have different activation mechanisms [18]. However, in CAD-EPCs, the precise pathophysiological function of each subunit and the activation mechanism remains unknown. Nox2 and Nox4 have been identified in diabetes as the predominant catalytic component of endothelial NADPH oxidase, and p47^phox plays an important role in regulating the activity of NADPH oxidase [19]. From our results we can infer that the subunit p22^phox, p47^phox and Nox2/4 may participate in NADPH oxidase activation, which may activated via membrane translocation of p47^phox and binding to Nox2 or Nox4 subunit. Immunoprecipitation detection showed small quantities of p47^phox-Nox4 conjugates, but no p47^phox-Nox2 conjugates were found (shown in Supplementary Figure 4). However, we need to do much deeper work to further validate this process here. Furthermore, in the cardiovascular system, Nox2 were mainly expressed in endothelial cells, cardiomyocytes, fibroblasts, and some of the VSMCs, Nox4 were mainly expressed in endothelial cells, VSMCs, cardiomyocytes and fibroblasts, we cannot rule out the possibility of p47^phox binding to Nox2 or Nox4 subunit. Nevertheless, one thing we can be sure that p47^phox plays a crucial role in regulating the activity of NADPH oxidase in CAD.

EPCs play pivotal roles in neovascularization and re-endothelialization in response to tissue ischemia and endothelial injury [20]. Several studies have shown that oxidative stress damages this functions of EPCs [21], however, the precise mechanism for this phenomenon is still unknown. We hypothesized that in CAD, the impaired vasculogenic capacity of EPCs is associated with an activation of NADPH oxidase. The present study showed that the in vivo angiogenesis capacity of EPCs from CAD subjects is severely impaired. Furthermore, when NADPH oxidase was inhibited by apocynin, the angiogenesis capacity of the CAD-EPCs was restored, suggesting a critical role of NADPH oxidase activation in the angiogenesis capacity. Although this has been demonstrated in diabetic EPCs, Sajoscha et al. reported that the in vivo re-endothelialization capacity of EPCs derived from individuals with diabetes mellitus was severely impaired by increased NADPH oxidase and reduced NO bioavailability. This study reports the first description of this phenomenon in EPCs from CAD, thus the NADPH oxidase system represents a promising target for correcting vasculogenic dysfunction in CAD-EPCs. However, apocynin can not reversed the blood perfusion at 28 d (137.2 PU versus 100.53 PU), P>0.05. Apocynin, which can inhibits NADPH oxidase activity [22], although this inhibition may be non-specific [23], but studies have proposed that apocynin is not an inhibitor of vascular NADPH oxidases, but an antioxidant [24], which imply that apocynin may inhibit the activity of NADPH oxidase through other mechanisms such as downstream pathways after NADPH oxidase activation. Furthermore, the effects of apocynin in transplanted EPCs were transient. Overall, our study provides a therapeutic approach to improve EPC function based on the inhibition of specific ROS generating enzymes-NADPH oxidases.

In stable coronary artery disease, various cardiovascular risk factors, including high cholesterol, hyperglycemia, smoking, hypoxia, etc. are all likely to lead to activation of NADPH oxidase. However, it can be activated by multiple signaling pathways and has many influencing factors, therefore it’s very difficult to study in vivo. Protein kinase C activation is an important component of multiple vascular disease processes [25]. The phosphorylation of NADPH oxidase subunits may be associated with PKC activation caused by risk factors. It has been reported that in diabetes, PKC-mediated activation of small GTP binding to Rac1 directly contribute to the activation of NADPH oxidase, in which PKCβ play an important role. Sachin A. Gupte reported that peroxide generation by p47^phox-src activation of Nox2 has a key role in PKC-induced arterial smooth muscle contraction [26]. At present, most researchers focus on the mechanism of NADPH oxidase activation induced by AngII, and reported that the first phase of Ang II-stimulated ROS production is dependent on protein kinase C activation [27]. In fact, there are a number of other studies have also pointed out that PKC plays an important role in the activation of NADPH oxidase. Monocytes mainly expressed PKCα, PKCβ1 and PKCβ2. In addition to the effects of PKCα/β activity on NADPH oxidase, the latest study also found that other PKC isozymes may also play a regulatory role in NADPH oxidase activation, such as PKCδ is also involved in controlling the phosphorylation and regulating the translocation of p47^phox and p67^phox, PKCζ inhibitor can block the phosphorylation of p47^phox and binding to gp91^phox which induced by TNF-α. However, the relationship between NADPH oxidase activation and PKC stimulation in CAD-EPCs remains poorly understood. Our results indicated that PKCα/β2 phosphorylation was significantly enhanced in CAD-EPCs, and NADPH oxidase activity and ROS generation were reduced after PKC inhibition, suggesting the critical role of PKC in the activation of NADPH oxidase in CAD. Moreover, the expression levels of NADPH oxidase subunit has also been proved that was positively correlated with PKC activity. Although NADPH oxidase activation can through several ways, PKCα/β2 signaling pathway must be one of the mechanisms. Thus, if we can identify the upstream targets that inhibit NADPH oxidase activation, then we might be able to improve EPC function for the treatment of coronary heart disease.

In conclusion, we found that NADPH oxidase-mediated oxidative stress might be activated by one of the mechanisms PKCα/β2 signaling pathway and NADPH oxidase activation plays an important role in EPCs from CAD (Figure 9). This study provides evidence for a therapeutic approach to CAD based on the inhibition of specific ROS generating enzymes-NADPH oxidase, particularly the specific inhibition of activation-related subunits.

Figure 9.

Possible processes of PKC mediated NADPH oxidase activation in EPCs.

Disclosure of conflict of interest

None.