Fetal bovine serum (FBS) and Dulbecco's modified Eagle's medium (DMEM) were obtained from Gibco. Endothelial cell growth supplement (ECGS) was from Merck Millipore. F₂ (Chinese national invention patent, No. ZL96119098.1) was synthesized in our laboratory and used at concentrations of 1 × 10⁻⁷, 1 × 10⁻⁶, 1 × 10⁻⁵ M (DMSO as solvent). Edaravone (EDA) was from Simcere Pharmaceuticals (Nanjing, China). N-acetyl-L-cysteine (NAC), xanthine oxidase (XO)/hypoxanthine (HX), anisomycin (ANISO), U0126, SB203580 and 2′,7′-dichlorofluorescein acetyl acetate (DCFH-DA, mainly detecting H₂O₂) were from Sigma-Aldrich. SP600125 was from Enzo Life Sciences. Epidermal growth factor (EGF) was from Pepro Tech.

Primary antibodies against p-JNK, total-JNK, p-ERK1/2, total-ERK1/2, p-p38, total-p38, and Egr-1 were purchased from Cell Signaling Technology (United States); primary antibody against platelet endothelial cell adhesion molecule-1 (CD31) was purchased from Bio-RAD (United States). Antibodies against β-actin, anti-rabbit secondary antibody, and anti-mouse secondary antibody were purchased from Wuhan Boster Biotechnology Limited (Wuhan, China). The EMSA assay kit was purchased from Thermo-Fisher Scientific (United States).

Hypoxia solution (Zhang et al., 2015): 137 mM NaCl, 12 mM KCl, 0.49 mM MgCl₂·6H₂O, 0.9 mM CaCl₂, 4 mM HEPES, and 20 mM sodium lactate. Binding buffer for EMSA: HEPES pH 7.5, 5 mM MgCl₂, 2.5 mM dithiothreitol, 2.5 mM EDTA, 250 mM NaCl, and 10% glycerol.

Neonatal Sprague-Dawley (SD, 3–5 days) rats of either sex were used in experiments. All animals were treated in compliance with the Guide for Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH publication No. 85-23, revised 1996) and followed the rules of National Animal Protection of China. The study was approved by the Institutional Animal Care and Use Committee of Shantou University Medical College. Rat CMECs were isolated as described (Zhou et al., 2010a) and cultured in DMEM supplemented with 10% FBS, ECGS (15 mg/L) and heparin sodium (6.25 U/ml) at 37°C under 5% CO₂. CMECs were identified by antibody against CD31, which is constitutively expressed on the surface of CMECs. Hypoxia was induced as before with some modifications (Zhou et al., 2010a). Briefly, CMECs were cultured in hypoxia solution in an air-tight chamber saturated with pure N₂ at 37°C for 1, 2, 3, or 4 h; the culture conditions were then returned to normal for 1 h of reoxygenation.

CMECs after being cultured for 2–3 days were randomly allocated to the following groups: control, control + the ROS donor XO/HX (con + XO/HX), control + the ERK1/2 activator EGF (con + EGF), control + the JNK/p38 activator ANISO (con + ANISO), H/R, H/R + the ROS scavenger NAC (H/R + NAC), H/R + the ROS scavenger EDA (H/R + EDA), H/R + ERK1/2 inhibitor U0126 (H/R + U0126), H/R + JNK inhibitor SP600125 (H/R + SP600125), H/R + p38 inhibitor SB203580 (H/R + SB203580), H/R + different doses of F₂, H/R + F₂ + XO/HX, H/R + F₂ + EGF, and H/R + F₂ + ANISO. The working concentrations of each chemical and solvent used to prepare stock solutions are as follows: XO (1, 3, 5 mU/ml, potassium phosphate buffer as solvent)/HX (1.2 × 10⁻⁴, 3.6 × 10⁻⁴, 6.0 × 10⁻⁴, 10 M NaOH as solvent), EGF (50 ng/ml, H₂O as solvent), ANISO (40 ng/ml, DMSO as solvent), NAC (2 × 10⁻⁵, 1 × 10⁻⁴, 5 × 10⁻⁴ M, H₂O as solvent), EDA (5 × 10⁻⁵, 1 × 10⁻⁴, 2 × 10⁻⁴ M, PEG as solvent), U0126 (10 μM, DMSO as solvent), SP600125 (10 μM, DMSO as solvent), SB203580 (20 μM, DMSO as solvent), F₂(1 × 10⁻⁷, 1 × 10⁻⁶, 1 × 10⁻⁵ M DMSO as solvent). CMECs in all groups were cultured in serum-free medium for 12–24 h to achieve quiescence. After that, cells in the control group were left intact. The H/R group was treated as described above. The concentration and administration protocol for XO/HX, EGF, ANISO, NAC, EDA, F₂, SP600125, U0126, and SB203580 were described in “Reagent preparation” and Figure 1, respectively.

Flow cytometry was performed as previously described (Zhang et al., 2015). In short, cells were harvested after the indicated treatments, washed twice with PBS, and then washed once with serum-free medium. Subsequently, the supernatants were discarded and cell pellets were resuspended in 1 ml serum-free medium containing a final concentration of 5 μM DCFH-DA, and then incubated in the dark at 37°C for 30 min with gentle shaking every 5 min to fully expose the cells to the probe. Cell suspensions were then washed three times with ice-cold PBS. Subsequently, the cell pellets were resuspended in 500 μl PBS, mixed well, and analyzed at an excitation wavelength of 488 nm and emission wavelength of 525 nm using a FACSCalibur flow cytometer (Becton Dickinson, USA). WinMDI2.9 software was used to analyze the mean fluorescence intensity (MFI).

Western blotting was performed as described previously (Zhang et al., 2015). Briefly, total protein in CMECs was extracted and quantified. Next, equal amounts of denatured protein samples (20–50 μg) were subjected to 8% SDS-PAGE and probed with corresponding primary antibodies for Egr-1 (1:1500), p-JNK (1:1000), total-JNK (1:2000), p-ERK1/2 (1:2000), total-ERK1/2 (1:2000), p-p38 (1:2000), total-p38 (1:2000), and β-actin (1:4000) at 4°C overnight, followed by secondary antibodies [HRP-labeled rabbit anti-mouse IgG (1:50,000) or HRP-labeled goat anti-rabbit IgG (1:80,000)]. HRP-labeled secondary antibodies were detected by chemiluminescence, and the protein bands were analyzed using Gel-pro Image Analysis Software (Media Cybernetics, USA). The ratio of p-JNK/JNK, p-ERK1/2/ERK1/2, and p-p38/p38 reflected the activation of JNK, ERK1/2, and p38, respectively, and the ratio of Egr-1/β-actin represented the expression of Egr-1.

The expression level of Egr-1 mRNA was detected by qRT-PCR. Total RNA was extracted from CMECs using the RNAiso Plus kit (Takara). Total RNA (0.5 μg) was used to synthesize cDNA by using a PrimeScript RT Reagent Kit with gDNA Eraser (Takara) according to the manufacture's protocol. CDNAs were then quantified by RT-PCR on an ABI 7500 RT-PCR System (Applied Biosystems) using the following primers, Egr-1: 5′-GAACAACCCTACGAGCACCTG-3′ (sense), 5′-GCCACAAAGTGTTGCCACTG-3′ (antisense); GAPDH: 5′-GGCACAGTCAAGGCTGAGAATG-3′ (sense), 5′-ATGGTGGTGAAGACGCCAGTA-3′ (antisense), which were synthesized by BGI (BGI, China). GAPDH was used as an endogenous reference and the ratio of Egr-1 mRNA/GAPDH mRNA represented expression of the Egr-1 gene.

CMECs were inoculated at 1.0 × 10⁻⁵ cells/well on coverslips in 12-well-plates. When cells grew to 70% confluence, they were treated with the conditions as indicated. Subsequently, cells were fixed in 4% paraformaldehyde for 20 min at room temperature, permeabilized in 0.3% Triton X-100 for 15 min, and then were blocked in 5% BSA for 1 h. Next, BSA was removed and cells were incubated, with primary antibodies recognizing CD31 (1:50) and Egr-1 (1:50), at 4°C overnight, followed by AlexaFlour 488-conjugated goat anti-mouse or AlexaFlour 594-conjugated goat anti-rabbit secondary antibodies for 1–2 h at room temperature. Subsequently, the cells were washed in PBS and counterstained with Hoechst 33258. Then, coverslips were mounted onto glass slides using fluorescence mounting medium (Beyotime Biotechnology), and the fluorescence signal was detected using 200 × magnification on a Nikon microscope. Exposure times for each channel were as follows: Hoechst 33258, 200 ms; AlexaFlour 488, 200 ms; AlexaFlour 594, 600 ms.

Electrophoretic mobility shift analysis (EMSA) was performed on nuclear extracts prepared from CMECs. Complementary 27-bp oligonucleotides containing an Egr-1 binding site were 5′-GGATCCAGCGGGGGCGAGCGGGGGCCA-3′, 5′-TGGCCCCCGCTCGCCCCCGCTGGATCC-3′. Oligonucleotides were 5′-end-labeled with biotin and annealed. The biotin-labeled probe was incubated with 6 μg protein and poly (dI-dC) (50 μg/ml) in binding buffer for 25 min at room temperature. Gels were pre-run for 45 min, then samples were loaded directly onto non-denaturing polyacrylamide/bisacrylamide (6%) gels. Electrophoresis for 1.5–2 h at 100 volts, and electrotransfer for 45 min at 380 mA were performed in an ice bath. For competition studies, 50- and 200-fold molar excesses of unlabeled probe for Egr-1 were added.

The intracellular glutathione/oxidized glutathione (GSH/GSSG) ratio, which reflects the oxidative stress level and the ability to eliminate ROS, was measured with the GSH and GSSG Assay Kit (Beyotime Technology, China). Briefly, cells were harvested after the indicated treatments and were mixed with protein removal reagent (3 times the cell volume, 10 mg ≈ 10 μl). Subsequently, cells were freeze-thawed two times in liquid nitrogen and 37°C water. Cells were placed in an ice bath for 10 min, and then centrifuged for 10 min at 4°C and 10,000 × g, after which the supernatant was assayed for total GSH and GSSG according to the manufacturer's instructions. Absorbance was measured at 412 nm over 25 min. The concentration of reduced GSH in the sample was obtained by subtracting GSSG from total-GSH.

Malonaldehyde (MDA) was measured by thiobarbituric acid (TBA) reactivity using an MDA assay kit (Nanjing Jiancheng Bioengineering Institute, China) according to the manufacture's protocol. Supernatants containing MDA were obtained from the cytoplasm of CMECs. The concentration of MDA was calculated by a calibration curve using 1,1,3,3′tetra-ethoxy propane as a standard.

CMECs were inoculated in 96-well-plates. After different treatments with NAC, F₂, U0126, SP600125, and SB203580, 3-[4, 5-dimethylthiazol-2-yl]-2, 5 diphenyltetrazolium bromide (MTT) solution was added to the medium and cells were incubated for an additional 4 h at 37°C. Then the medium was discarded and DMSO was added to dissolve the formazan crystals. All operations were performed in the dark. Absorbance was measured at 490 nm using a SpectraMax M2e Microplate Reader (Molecular Devices). The control group was considered as 100% viable.

All data are presented as mean ± S.E.M. Differences between groups were determined using one-way ANOVA followed by a Student-Newman-Keuls test with SPSS 17.0 software. Statistical significance was considered at a value of P < 0.05.

Microvascular endothelial cells isolated from heart of neonatal Sprague-Dawley rats were multipolar and possessed a cobblestone-like appearance upon reaching confluence. CD31 is constitutively expressed on the surface of CMECs. As judged by immunofluorescence, all cells were CD31-positive, confirming that the cells were CMECs (Supplementary Image 3).

To determine the effects of different durations of H/R on ROS level and Egr-1 expression, CMECs were cultured in an air-tight chamber saturated with pure N₂ at 37°C for 1, 2, 3, or 4 h, then returned to normal conditions for 1 h of reoxygenation. Flow cytometry analysis revealed that ROS levels in CMECs increased in a manner dependent on the duration of H/R (≈1.7–2.9 fold), with peak ROS levels occurring at 3 h of hypoxia and 1 h of reoxygenation (H3/R1) vs. normoxia (Figure 2A). Immunoblotting of extracts prepared from CMECs showed expression of Egr-1 increased in all H/R groups (≈17–92-fold) compared with the control group. Concomitant with the peak in ROS levels, peak Egr-1 expression also occurred at H3/R1 (Figure 2B), followed by a decline in ROS generation and Egr-1 protein expression by H4/R1. Based on the above data, all subsequent experiments were performed using 3 h of hypoxia and 1 h of reoxygenation.

The correlation of time-dependent changes of Egr-1 protein, during H/R, with ROS levels suggested that ROS could be responsible for Egr-1 induction. To further investigate the cause-effect relationship between ROS and Egr-1 in H/R CMECs, we determined whether the ROS donor XO/HX and ROS scavengers EDA and NAC could exert changes in Egr-1 expression.

Flow cytometric analysis of ROS, western blot analysis of Egr-1 protein and RT-PCR analysis of Egr-1 RNA harvested from XO/HX-induced CMECs displayed increased ROS levels, Egr-1 protein expression and Egr-1 transcripts (≈2.4-, 1.7-, 4.2-fold, respectively) at low concentrations of XO/HX, which rose further at moderate concentration of XO/HX (≈3.0-, 2.7-, 9.8-fold, respectively), and were even more elevated at high concentration of XO/HX (≈4.1-, 6.6-, 49-fold, respectively), as compared with the control group (Figures 3A–C), demonstrating that treatment with different concentrations of XO/HX for 1 h increased ROS levels and Egr-1 mRNA and protein expression in a dose-dependent manner. Pretreatment with EDA and NAC for 30 min decreased ROS levels and Egr-1 mRNA expression induced by H/R (Figures 4A,B,D), with moderate and high concentrations of ROS scavengers (1 × 10⁻⁴ M and 2 × 10⁻⁴ M EDA, 1 × 10⁻⁴ and 5 × 10⁻⁴ M NAC) resulting in substantial decreases in H/R-induced Egr-1 protein expression (Figure 4C). In contrast, low concentrations of ROS scavengers (5 × 10⁻⁵ M EDA, 2 × 10⁻⁵ M NAC) had no significant effect on Egr-1 expression induced by H/R. These results indicate that Egr-1 expression in CMECs is regulated by ROS levels following H/R stimulation. In other words, a ROS/Egr-1 signaling pathway is activated in CMECs by H/R.

To observe the relationship between ROS and MAPK activity, we initially used XO/HX (5 mU/ml, 6.0 × 10⁻⁴ M) to increase ROS levels, and NAC (5 × 10⁻⁴ M) to decrease H/R-induced ROS generation, and examined MAPK activity. We found that p-ERK1/2, p-JNK, and p-p38 protein expression all increased in XO/HX-treated CMECs (≈21-, 19-, 11-fold, respectively) and H/R-induced CMECs (≈8-, 13-, 5-fold, respectively) compared with the control CMECs, indicating ERK1/2, JNK, and p38 were all activated by ROS and H/R (Figures 5A–C). H/R-mediated ERK1/2, JNK, and p38 activation was decreased significantly by addition of the ROS scavenger NAC (Figures 5A–C). We next utilized activators and inhibitors of MAPKs to assess if activation of MAPK by ROS regulates the expression of Egr-1. Addition of ANISO (40 ng/ml) for 1 h, to activate JNK/p38, or addition of EGF (50 ng/ml) for 1 h, to activate ERK1/2, both increased Egr-1 protein expression, whereas the H/R-induced increase in Egr-1 protein could be blocked by preincubation of cells with SP600125 (10 μM), U0126 (10 μM), or SB203580 (20 μM) for 45 min (Figure 5D). These data suggest that ERK1/2, JNK, and p38 are all involved in ROS/Egr-1 signal transduction, and that activation of these MAPKs plays an important role in the ROS/Egr-1 pathway.

After confirming H/R-mediated activation of the ROS/MAPK/Egr-1 signaling pathway in CMECs, we next explored the effects of F₂ on this pathway. Compared with the control group, ROS level, MAPK activation, and Egr-1 mRNA and protein expression all increased in the H/R group. However, these increases were dose-dependently inhibited by a 30 min pretreatment with 0.1–10 μM F₂ (Figures 6A–F). High concentrations of F₂ further decreased ROS levels compared with low concentrations of F₂ (Figures 6A). Moderate and high F₂ concentrations reduced Egr-1 protein expression to a greater extent compared with low concentrations (Figure 6E). In addition, moderate and high F₂ concentrations decreased the expression of Egr-1 mRNA and p38 activation, as compared with low concentrations of F₂ (Figures 6F,D). As evidenced by EMSA of nuclear extracts from CMECs, using a biotin-labeled oligonucleotide probe containing a consensus Egr-1 binding site, increased DNA binding activity of Egr-1 was observed after H/R, and the increase was inhibited by F₂ pretreatment (Figure 7). These data suggest that F₂ inhibits H/R-induced ROS, MAPK activation, Egr-1 expression, and enhanced Egr-1 DNA binding in CMECs in a dose-dependent manner. Although we have previously shown that inhibiting expression of Egr-1 is one of the mechanisms by which F₂ protects cardiomyocytes from H/R injury (Zhang et al., 2007), we extend these results to show that F₂ affects the DNA binding activity of Egr-1.

Addition of XO/HX (1 mU/ml, 1.2 × 10⁻⁴ M), EGF (50 ng/ml), or ANISO (40 ng/ml) to F₂-pretreated H/R CMECs in hypoxia solution for 3 h, followed by culturing in normal medium with F₂ alone, showed that XO/HX antagonized the effects of F₂ on H/R-induced ROS generation, MAPK activation, and Egr-1 expression (Figures 8A–E). Similarly, the ERK1/2 and JNK/p38 activators EGF and ANISO, respectively, also antagonized the effects of F₂ on H/R-induced MAPK activation and Egr-1 expression (Figures 8B–E). These findings indicate that F₂ can modulate H/R-mediated activation of the ROS/MAPK/Egr-1 signaling pathway in CMECs.

F₂ (10⁻⁶ M) pretreatment reduced H/R-mediated cell death, as did signaling pathway inhibitors (NAC, U0126, SP600125, and SB203580) (Figure 9A). Since cellular MDA levels and the GSH/GSSG ratio are widely used to reflect the ability to eliminate ROS and the degree of oxidative stress injury, we measured MDA levels and the GSH/GSSG ratio in order to characterize the extent of oxidative damage induced by ROS-mediated signaling. F₂ (10⁻⁶ M) reduced MDA levels, but enhanced the GSH/GSSG ratio, compared with H/R alone. Signaling pathway inhibitors exerted the same effects (Figures 9B,C). Induction of Egr-1 is also regarded as a marker of H/R injury (Yan et al., 2000; Zhang et al., 2007, 2008). We found that Egr-1 (red) was expressed at low levels in the control group, and was distributed throughout the entire cell (cytoplasm and nucleus; Supplementary Image 4). Notably, H/R stimulation triggered Egr-1 translocation from the cytoplasm to nucleus. However, this translocation was blocked when F₂ was present. As expected, NAC and MAPK inhibitors exerted similar effects as F₂ (Supplementary Image 4). These novel findings indicate that F₂, similar to inhibitors of ROS and MAPKs, could enhance cell viability, and decrease both oxidative stress and activation of Egr-1 to protect CMECs against H/R injury through inhibiting the ROS/MAPK/Egr-1 pathway.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.