Homocysteine-induced extracellular superoxide dismutase and its epigenetic mechanisms in monocytes

Epidemiologic and case control studies have consistently indicated that moderate and mild elevation of plasma homocysteine (Hcy) is an important and independent risk factor for arteriosclerosis and venous thrombosis disease in populations (Stampfer et al.,1992). The level of risk for cardiovascular disease attributable to homocysteine is equivalent to that associated with hypercholesterolemia,and elevated Hcy levels synergistically increase the risk associated with smoking and hypertension (Vilar et al.,2007). Recently, studies in animal models have revealed that hyperhomocysteinemia enhances the formation of atherosclerotic lesion and the neointimal hyperplasia followed by arterial injury(Hofmann et al., 2001). Although there is experimental evidence that Hcy causes endothelial dysfunction (Woo et al., 2000)and vascular smooth muscle cell proliferation(Majors et al., 1997), the fundamental biochemical mechanisms underlying these phenomena are not yet well understood.

Some biological effects of Hcy in vitro can be mimicked by hydrogen peroxide (H₂O₂) or other sulfhydryl-containing agents but inhibited by catalases(Starkebaum and Harlan, 1986). These results lend support to the hypothesis that increased oxidation mediated through the sulfhydryl group of Hcy is the major mechanism responsible for Hcy-induced vascular pathogenesis (Turhan et al., 2005). Moreover, another intermediary metabolite of methionine, cysteine, has similarities to the chemical structure and redox property of Hcy, but does not constitute a risk factor for cardiovascular disease. Why did homocysteine become a pro-oxidant and a well known risk factor for atherosclerosis (AS), while cysteine is usually anti-oxidative?Many plausible explanations have not probed this key mechanistic point. It is suggested that the above-mentioned characteristics alone are not sufficient to explain the deleterious effect of Hcy.

DNA methylation, i.e. the formation of 5-methylcytosine (5-mC) from a cytosine residue via methyltransferase, is an important factor regulating the development of gene expression at different stages(Maatouk et al., 2006; Yideng et al., 2007). Some reports have shown that DNA methylation might be important for atherogenesis because AS is at least partially regulated by DNA methylation. Investigations of ApoE knockout mice (Iwama et al.,1998) revealed that significant genomic hypomethylation is developed during the first replication of aortic smooth muscle cells (SMCs) in vivo, and that hypomethylation occurred in some specific genes,such as 15-lipoxygenase and ApoE, which are therefore indicated to be deeply involved in AS (Maatouk et al.,2006; Jiang et al.,2007a; Jiang et al.,2007b; Yi-Deng et al.,2007). This might result from a direct regulatory effect of hypomethylation on gene expression or a secondary effect via effects on DNA integrity and its function.

The foam cells play a central role in the pathogenesis of atherosclerosis. The accumulation of cholesterol in foam cells is essentially a balance between cholesterol intake and cholesterol efflux. Extracellular superoxide dismutase(EC-SOD) is an anti-oxidative enzyme that catalyzes the dismutation of superoxide anion (O ^.–₂) to less reactive hydrogen peroxide (Reeves et al.,2002; Beckman,1996; Crapo et al.,1992; Marklund,1982), while the progression of AS and the change of EC-SOD are still unknown. How does Hcy induce foam cells and contribute to the accumulation of cholesterol? What roles does EC-SOD play? And what pathway does this interference go through? Based on these questions, the aim of the present study was to investigate the effect of Hcy on the activity of EC-SOD,as well as its relation with the methylation status of CpG sequences in DNA and the acetylation of histones H3 and H4. The aberration of DNA methylation and acetylation of histones H3 and H4 that was revealed, together with their pathways, could be potential targets for anti-atherosclerosis. Alteration of DNA methylation is potentially an important finding that could constitute the mechanism against AS featuring epigenetic gene silencing. The results suggested that methylation plays a role in the development of AS and had important consequences when it was exposed to Hcy. In addition, the results revealed that Hcy, being specifically converged on EC-SOD hypermethylation,contributed to the mechanisms leading to AS.

Human blood from a healthy donor was drawn into heparinized syringes. Whole blood was separated into peripheral blood mononuclear cells and neutrophils using the density gradient from Nycoprep 1.077 (Axis-Shield PoC, Oslo,Norway). Monocytes were isolated from peripheral blood mononuclear cells by adherence to a serum-coated culture flask for 2 h. Adherent cells were detached and resuspended in RPMI-1640 medium containing 5% autologous plasma. Only cell preparations with a 95% viability or greater were used. The cells were then planted into six-well plates and grown to 80% confluence. The cells were then cultured with 0.5 μg ml^–1 phorbol myristate acetate (PMA) (Sigma-Aldrich, St Louis, USA) together with 100 μg ml^–1 oxidized low-density lipoprotein (ox-LDL) and different concentrations of Hcy.

When the cells were in the exponential phase of growth, 30 mg l^–1 azacytidine (AZC; Sigma-Aldrich, St Louis, USA) and 50 ng ml^–1 trichostatin A (TSA; Sigma-Aldrich) were added to the monocytes and culture continued for 3 days, followed by three washes with phosphate buffered saline (PBS); culture using RPMI-1640 medium was continued for 2 days before experimentation.

The cultured monocytes were washed with PBS three times, fixed in 2.5%glutaraldehyde for 3 h, dipped in 2.5% potassium dichromate for 16 h, and stained in 1% oil red O (Sigma-Aldrich) for 20 min to identify lipid droplets in cytoplasm. Cell nuclei were then stained in Hematoxylin for a few seconds. All products were washed with distilled H₂O and the monocyte-derived foam cells were observed and evaluated according to Wada et al.'s semi-quantitative method (Wada et al., 2002), i.e. the percentage of positive oil red O-staining cells.

The cultured human monocytes were resuspended in 0.5 ml sodium phosphate buffer solution (0.1 mol l^–1, pH 7.4). Each sample was sonified for 1 min with the microtip of a sonifier. Total cholesterol levels were determined as described (Rong et al.,1999). To determine free cholesterol (FC), cholesteryl esterhydrolase was omitted from the assay solution. Cholesteryl ester (CE) was determined by subtraction of the free cholesterol from total cholesterol (TC)levels.

H₂O₂ was analysed following the method of Peng et al.(Peng et al., 2006). Cultured monocytes were collected and diluted to 1.5×10⁷cells ml^–1 in PBS. After 1 h at 4°C, the cells were activated for 10 min at 37°C by addition of PMA (1 μg ml^–1). A 1 ml sample was removed from the cell suspension at 3 min intervals and filtered through a 0.22 μm syringe. Cell filtrates were collected at each time point during the reaction, and the H₂O₂ content of each filtrate was monitored as follows. 2.6 ml Tris-HCl buffer solution (pH 7.4,0.1 mol l^–1), 0.1 ml 2.0 U ml^–1 horseradish peroxidase (HRP) solution and 0.1 ml cell filtrate were added to 0.2 ml p-hydroxyphenylacetic acid solution (pH 7.5). After 1 min, fluorescence was measured spectrofluorometrically (excitation wavelength, 325 nm; emission wavelength, 415 nm).

An enzyme-linked immunosorbent assay (ELISA) was carried out on all samples. Briefly, flat-bottomed 96-well microtiter plates were coated with 100μl of 0.06 mol l^–1 sodium carbonate buffer (pH 9.6)containing 0.5 μg antigen ml^–1 at 4°C. Plates were blocked with 150 μl of PBS-Tween containing 1% BSA (Hangzhou Sijiqing Biological Engineering Material Co., Ltd, Huanzhou, China) at 20°C for 1 h and were washed five times with PBS-Tween. Based on the affinity titration curve for protein-G, a cell dilution of 1:600 was chosen. After diluting the samples in PBS-Tween, 100 μl of sample were added to the plates, which were incubated at 20°C for 1 h. Biotinylated acetylase (HAT), histone deacetylase (HDAC) and ox-LDL (Jackson Immunoresearch, West Grove, PA, USA)diluted 1:15000 in PBS-Tween were individually added (100 μl/well) and incubated at 20°C for 1 h. Streptavidin–peroxidase (Sigma-Aldrich)diluted 1:10000 in PBS-Tween was added and incubation continued at 20°C for 30 min. As the substrate, 50 μl of O-phenylenediamine (OPD) diluted in citric acid phosphate buffer (pH 5.0; 0.66 mg ml^–1) + 0.6μl 30% H₂O₂ mg^–1 OPD was added and the mixture incubated in the dark at 20°C for 10 min. The enzyme reaction was stopped by adding 50 μl of 2 mol l^–1H₂SO₄. The plates were read in a spectrophotometer at 492 nm.

Electrophoresis was carried out on sodium dodecyl sulfate–polyacrylamide gels. Proteins were transferred to nitrocellulose membranes at 67 V for 2 h at room temperature with gentle agitation on a platform shaker, and were washed three times for 5 min in Tris-buffered saline plus Tween-20 (TBST). The membrane was incubated with a monoclonal anti-DNAmethyltransferase (DNMT)1, DNMT3, methyl-CpG (MeCP) binding domain(MBD), MeCP and extracellular superoxide dismutase (EC-SOD) antibody (1:250 dilution) (Jackson ImmunoResearch, West Grove, PA, USA) in 10 ml primary antibody dilution buffer with gentle agitation overnight at 4°C. The membrane was then washed three times with TBST and incubated with a second antibody (goat anti-rabbit horseradish peroxidase-conjugated immunoglobulin G;Jackson ImmunoResearch) in PSB at 1:2000 dilution containing 1% bovine serum albumin (New England Biolabs, Beverly, USA) for 1 h at room temperature. After washing again three times with TBST, the membrane was incubated with 10 ml LumiGLO (New England Biolabs, Beijing, China) with gentle agitation for 1 min at room temperature, then the excess developing solution was drained. The membrane could not be dried out, but instead was wrapped in a plastic wrap and exposed to X-ray film. The control value was taken as 100%. Values are reported relative to that of β-actin using the formula (relative value=experimental densitometry value×100/β-actin value).

A standard curve was generated by preparing different target sequence ratios of methylated versus unmethylated alleles, as described previously (Pang et al.,2000). The bisulfite modification of gDNA has been described previously (Reeves et al.,2002). The genomic DNA was isolated from the cultured cells by using the E.Z.N.A Tissue DNA Kit (Omega Bio-tek, Guanzhou, China). The bisulfite modification was performed with the EZ DNA Methylation Kit (Zymo Research, Orange, CA, USA) following the manufacturer's instructions. Afterwards, DNA was precipitated with glycogen as carrier and resuspended in 20 μl of water.

A novel quantitative analysis of methylated alleles, which is essentially a major improvement over a previous method based on real-time PCR (MethyLight),was used (Peng et al., 2006). We used a VIC-labeled probe that specifically hybridizes to the sequence derived from the methylated allele, together with a FAM-labeled probe that binds to the sequence generated from the unmethylated allele. The amount of the fluorescent dye released during PCR is measured by a real-time PCR system and is directly proportional to the amount of PCR product. Binding sites of the probes covered three differently methylated CpG dinucleotides. The improved sequence specificity facilitates the relative quantification of the methylated and the unmethylated alleles that were simultaneously amplified in a single tube.

PCR primers were designed to amplify the bisulfite-converted antisense strand of the EC-SOD. The PCR primers, probes and stategy for designing the MethyLight reaction are shown in Fig. 1. PCR was carried out by using a 96-well optical tray with caps at a final reaction volume of 20 μl containing 10 μl TaqMan universal PCR master mix, 2 μl bisulfite-treated DNA, 2.5 μmol l^–1 of each primer of EC-SOD, 150 nmol l^–1of each of the fluorescently labeled probes, EC-SODmet and EC-SODunmet. The initial denaturation at 95°C for 5 min to activate the AmpliTaq Gold DNA polymerase was followed by 40 cycles of the denaturation at 95°C for 15 s and the annealing and extension at 60°C for 1 min(Zeschnigk et al., 2004).

Chromatin immunoprecipitation (ChIP) was performed according to published protocol (Umlauf et al., 2004)with modifications. In brief, the foam cells were collected and washed twice with Hank's solution. The cells were rinsed with ice-cold Hank's solution and incubated on ice for 10 min in 4 ml buffer. Nuclei were pelleted by centrifugation at 14 000 g for 15 min at 4°C, resuspended in 1% SDS, 50 mmol l^–1 Tris-HCl and 10 mmol l^–1 EDTA with the fresh protease inhibitor cocktail, and incubated on ice for 10 min. Chromatin was sonicated (Sonic Dismembrator 550,Fisher Scientific Microtip, Pittsburgh, PA, USA) to an average length of 500 bp. To preclear the chromatin, the sonicated cell suspension was diluted tenfold with a buffer containing 0.01% SDS, 1.1% Triton X-100, 1.2 mmol l^–1 EDTA, 16.7 mmol l^–1 Tris-HCl, pH 8.1,and 167 mmol l^–1 NaCl, and incubated with 80 μl salmon sperm DNA or protein A Agarose and 50% Slurry (Upstate Biotechnology Inc,Charlottesville, VA, USA) for 3 h with rotation at 4°C. One-third of the precleared chromatin was incubated with 2 μg of one of the following polyclonal antibodies overnight: anti-H3Ac, anti-H4Ac (Upstate Biotechnology Inc, Charlottesville, VA, USA), and control antibodies (Santa Cruz Biotechnology, Inc. Santa Cruz, CA, USA). One-tenth of the precleared chromatin was saved as the spare for the following experiment. Each immunoprecipitation was recovered, washed and eluted from the beads. ChIP DNA pellets were resuspended in 60 μl of Triton-EDTA (TE) and analyzed by SDS-PAGE.

Results are expressed as mean ± s.e.m. The data were analyzed with one-way ANOVA and additional analysis was carried out with the Student–Newman–Keuls' test for multiple comparisons within treatment groups, or t-test between two groups. P<0.05 was considered significant.

We investigated the molecular mechanisms underlying the effect of Hcy on the accumulation of cholesteryl ester during the development of foam cells. To evaluate any possible cellular damage caused by Hcy, we first determined the percentage of foam cells. When the monocytes were grown to subconfluence and then co-cultured with ox-LDL and PMA along with 50, 100, 200 or 500 μmol l^–1 Hcy for 72 h, a great number of foam cells were found by oil red O staining (Fig. 2) and the percentage of foam cells in the positive group (cultured with ox-LDL and PMA but not Hcy) (46.20±5.80%) was much higher than the control group(without ox-LDL, PMA and Hcy) (4.80±1.50%) (P<0.05). The percentage of foam cells in 50, 100, 200 or 500 μmol l^–1Hcy groups were much higher than that in the positive group, i.e. increased by about 1.25, 1.92, 1.52 or 1.42-fold, respectively (P<0.05 or P<0.01), thus suggesting that Hcy induced the formation of foam cells. The maximum level was at 100 μmol l^–1 Hcy, which is similar to the plasma level observed in hyperhomocysteinemia(Fig. 2A). When Hcy was cultured in control medium for an additional 24, 48 or 72 h, the percentage of foam cells increased to 66.2%, 80.2% and 92.5%, respectively, in comparison with the control group (Fig. 2B). The inhibitory effects of folate, AZC and TSA resulted in the percentage of foam cells decreasing to about 72.4%, 54.65% and 51.95% of the positive control group, respectively (Fig. 2C).

To invetsigate the role of Hcy on lipid metabolism, we determined the accumulation of cholesterol in foam cells to confirm that Hcy is an important key factor and is involved in the accumulation of cholesterol in foam cells derived from the monocytes. When monocytes were co-cultured with ox-LDL and PMA, all the intracellular lipids, including the cellular total cholesterol,the free cholesterol and cholesteryl ester, were significantly increased compared with the control group (P<0.05). After addition of different concentrations of Hcy to the monocyte cultures for 72 h, the levels of TC, FC and CE gradually increased and were much higher than those of the positive control group. Additionally, Hcy induced a time-dependent increase in the content of cholesteryl ester accumulation, and this increase was suppressed by folate, AZC and TSA, and were similar to the percentage of foam cells (Table 2).

It has been reported that oxidative damage of lipids and the consequent formation of foam cells is a key step in the onset and the development of atherosclerosis. H₂O₂ and ox-LDL were involved in the formation of foam cells (Peng et al.,2006). From the above description, we can see that the maximum effect was observed at 100 μmol l^–1 Hcy(Fig. 2A) so we added 100μmol l^–1 Hcy to the monocyte culture with PMA and ox-LDL. We found that the quantities of H₂O₂ and ox-LDL in foam cells were significantly increased in the presence of 100 μmol l^–1 Hcy. After 72 h the increased levels reached 11.59- and 3.04-fold of the control group, respectively, and the quantity of H₂O₂ and ox-LDL increased in a time-dependent manner. We also examined the effect of folate, AZC and TSA on H₂O₂and ox-LDL production in Hcy-treated monocytes. After incubation with folate,AZC and TSA, H₂O₂ and ox-LDL production was substantially inhibited (Fig. 3), suggesting that there are underlying relationships between H₂O₂, ox-LDL and the accumulation of cholesterol.

Although it has been reported that Hcy can induce the accumulation of cholesterol via H₂O₂ and ox-LDL, possible mechanisms for this induction have not been reported. So we first measured the gene expression of EC-SOD in the presence or absence of Hcy. RNA was extracted at different time points and RT-PCR analysis was performed to measure the mRNA level. Compared with the control group, EC-SOD mRNA levels obtained within 24,48 and 72 h of incubation were, respectively, about 125.29%, 65.77% and 46.69%of control levels. But compared with the Hcy (100 μmol l^–1) group, when cells were treated with folate, AZC and TSA after 72 h, EC-SOD mRNA levels were increased by 2.13-, 2.19-, 2.34- and 3.4-fold, respectively, suggesting that the foam cells cultured with Hcy for 24 h could increase the production of EC-SOD, while further culture with Hcy for 48 and 72 h led to the downregulation of EC-SOD. But folate, AZC and TSA repress the effects of Hcy in this process. By contrast, after 24 h, EC-SOD mRNA levels were higher than that of the control group, which was reasonable,because some studies have shown the presence of high amount of EC-SOD mRNA and its activity in early atherosclerotic lesions (Mikko et al., 1999)(Fig. 4A). The results are consistent with our previous findings that EC-SOD protein levels were substantially decreased by the 100 mmol l^–1 concentration of Hcy for 24 h and repressed by folate, AZC and TSA added later(Fig. 4B). Therefore, under these conditions, the change of metabolism of EC-SOD represents a major pathway.

We have previously reported that Hcy has an effect on the activity of DNA methyltransferases and CpG binding proteins in monocytes, and DNA methyltransferases and CpG binding proteins involved in DNA methylation(Jiang et al., 2007a; Jiang et al., 2007b). DNMT1,DNMT3, MBD and MeCP2 are key genes regulating the gene transition. To understand how Hcy affects the regulation of foam cells, we examined the mRNA and the protein expression of DNMT1, DNMT3, MBD and MeCP2 in monocytes cultured with folate, AZC and TSA. To characterize the time course of this effect, RNA and protein were harvested from foam cells treated with Hcy for 24, 48 and 72 h and analyzed by real time RT-PCR together with western blotting analysis (Fig. 5). Both the reduction in MeCP2 and the increase of DNMT1 started as early as 24 h. MeCP2 mRNA decreased to 20.9%, 43.87% and 41.23% of the control group levels after 24, 48 and 72 h. However, DNMT1 mRNA increased to 27.23%, 38.17%and 59.3% of the control group levels after 24, 48 and 72 h, respectively(Fig. 6C). Because DNMT3 and MBD remained unchanged, we treated monocytes with 100 μmol l^–1 Hcy and folate, AZC as well as TSA for 72 h, and then measured the mRNA expression of DNMT1, DNMT3, MBD and MeCP2. While folate, AZC and TSA affected the mRNA expression of DNMT1 and MeCP2, it had only minor effects on the mRNA level of DNMT3 and MBD(Fig. 5B).

The effect of Hcy on the protein level of DNMT1, DNMT3, MBD2 and MeCP2 was measured by western blot (Fig. 6), with similar results to the mRNA expression of DNMT1, DNMT3,MBD2 and MeCP2. The lowest or highest protein expression was also in the 100μmol l^–1 Hcy group, and there was a significant difference among DNMT1, DNMT3, MBD2 and MeCP2 protein levels after various incubation times with Hcy. Thus, Hcy selectively decreases DNMT1 and MeCP2 mRNA levels in monocytes in a time-dependent manner. Both the reduction in the expression of MeCP2 mediated by Hcy and the increase of DNMT1 were pathophysiologically relevant and independent. The effects of folate, AZC and TSA on the protein expression of DNMT1, MECP2, DNMT3 and MBD were consistent with the mRNA expression.

To analyze the possible role of Hcy on levels of EC-SOD DNA methylation, we used a quantitative TaqMan-based real-time PCR. It is already known that AZC and TSA can induce general hypomethylation, so we cultured the primary monocytes in the presence of Hcy, folate, AZC and TSA. It was found that the homocysteine after 24, 48 and 72 h incubation led to significant increases in EC-SOD DNA methylation, by 85.55%, 135.28% and 145.9%, respectively(P<0.05). In the control group and the 100 μmol l^–1 Hcy group, when cultured with folate, AZC and TSA, the levels of DNA methylation of EC-SOD decreased by 27.9%, 28.8% and 39.5%,respectively (P<0.01) (Fig. 7). These findings suggested that there is a specificity of Hcy on DNA methylation in human monocytes.

This result is consistent with the earlier observation in cell lines that genes associated with CpG clusters are methylated if they are not necessary for cell survival (Dong et al.,2002). The decrease of DNA methylation was random and did not resemble the hypermethylation pattern observed in atherosclerotic aortas,which implies that the hypermethylation of EC-SOD in vivo may have arisen by specific mechanisms.

There is mounting evidence that mutant transcription factors, often resulting from chromosome translocations, contribute to the pathogenesis of corepressors and histone deacetylases (HDACs) and thereby result in altered chromatin architecture and modified gene expression. In this case, HDACs play a key role in the control of gene expression through chromatin modification.

We chose a strategy whereby histones were isolated from the nuclei of foam cells by acid extraction and detected by ChIP. As shown in Fig. 8, acetylated histones H3 and H4 in monocytes were treated with the 100 μmol l^–1 Hcy for varied time points. The levels of acetylated histone H3 and H4 were significantly elevated after incubation compared with the control group. The maximum level of acetylated histone H3 and H4 production was observed after 72 h incubation with 100 μmol l^–1 Hcy. These observations indicate that the relatively low concentration of Hcy, within the pathophysiological range, can directly trigger the increase of acetylated histones H3 and H4 in cultured monocytes. Treatment of cells with 100 μmol l^–1 Hcy induced a time-dependent accumulation of acetylated histones H3 and H4. But when treated with folate, AZC and TSA, the amount of acetylated histone H3 and H4 was decreased compared with levels in the 100μmol l^–1 Hcy group (72 h).

HAT and HDAC are enzymes that catalyze the deacetylation and the acetylation of lysine residues located in the NH₂-terminal tails of histones and non-histone proteins. To determine whether Hcy is able to induce the secretion of HAT and HDAC in monocytes, we next examined whether the homocysteine-induced upregulation of HAT and HDAC expression resulted in an increase of acetylated histone H3 and H4. The amounts of HAT and HDAC secreted from the cultured monocytes were determined by ELISA, and HDAC was significantly decreased in foam cells pretreated with homocysteine in both a time- and dose-dependent manner (Fig. 9). However, HAT levels did not increase on treatment with folate,AZC and TSA, but the HDAC expression was consistent with the results of acetylated histone H3 and H4 (Fig. 9). These results suggested that homocysteine could sensitize acetylated histone H3 and H4 through the suppression of HDAC activity.

Hyperhomocysteinemia has been recognized as an independent risk factor that predicts adverse cardiovascular aspects of patients such as ischemic heart disease and stroke (Clarke et al.,1991). However, the underlying mechanisms of these unfavorable effects of homocysteine have not been elucidated. Why does Hcy become a pro-oxidant and an important risk factor for atherosclerosis? Since then many plausible explanations have not covered this key point, and the results of the present study demonstrate for the first time that Hcy not only inhibits the secretion of EC-SOD and accelerates the accumulation of cholesterol but also induces the gene promotor hypermethylation of EC-SOD and the acetylation of histone H3 and H4 in cultured human monocytes.

First, we tested whether H₂O₂ and ox-LDL act as mediators in the Hcy-induced expression of EC-SOD and found that Hcy elevated H₂O₂ and ox-LDL secretions, suggesting that H₂O₂ and ox-LDL, as strong oxidants, are important risk factors for atherosclerosis (Ross et al., 1999; Wentworth et al., 2003; Liu et al., 1996). Recently it has been reported that reactive cholesterol ozonized products trigger the formation of foam cells in tissue macrophages(Zelko and Folz, 2003). Our present study showed that PMA-activated monocytes, when cultured with ox-LDL,were obviously transformed into foam cells. But when monocytes were cultured with Hcy, the levels of H₂O₂ and ox-LDL were significantly increased, indicating that Hcy is involved in the formation of foam cells and the accumulation of cholesterol. In addition, there is evidence that in a mild homocysteine elevation, so-called hyperhomocysteinemia,H₂O₂ and ox-LDL internalize excessive amounts of lipids and become foam cells. Our study also showed that AZC and TSA could partially block the formation of Hcy-induced H₂O₂ and ox-LDL.

Second, we have distinguished the potential enzymatic sources of EC-SOD. From our results, we found that decreased EC-SOD levels relative to elevated homocysteine levels could be a response to homocysteine-induced oxidative damage and could thus constitute a protective mechanism, with the effect of countering oxidative stress. All of these results are consistent with the increased circulating EC-SOD levels associated with reduced cardiovascular risk (Ross, 1993). Based on these findings, the increased EC-SOD should be helpful in diminishing the risk of vascular damages.

Furthermore, although the molecular mechanisms underlying Hcy-induced atherosclerosis have been the subject of large-scale research, most previous studies have only focused on the influence of Hcy on endothelial cells(Benditt and Benditt, 1973). It is still not known whether DNA methylation of EC-SOD contributes to lesion formation or whether it occurs as a consequence of the pathological process. However, as it has been reported earlier(Turhan et al., 2005), our results indicate a causal relationship between the methylation and the EC-SOD expression. It is conceivable that hypermethylation has at least partly developed through the decreased cell proliferation, which is a typical feature in atherogenesis (Campbell et al., 1981). Similar to cancer cells, the EC-SOD methylation level leading to the formation of foam cells was high, so it is also possible that hypermethylation of genes is involved in the inhibition of cell growth. Preliminary results indicate that similar hypomethylation can be found in smooth muscle cells isolated from the intima of balloon-denudated arteries (Schwartz et al.,1995). These findings also support the hypothesis that DNA methylation could reflect a difference in gene expression and be the potential target of gene regulation (Willems et al.,1993).

Just as expected, the results of our study showed that the activity of DNMT1 increased and the expression levels of MeCP2 decreased in monocytes after incubation with Hcy. A potential plausible explanation was the compensatory reaction of the methylation mechanisms against Hcy-induced hypomethylation. It has been reported that methyltransferase activity in cancer tissue actually increased despite genome-wide hypomethylation(Haaf, 1995). Methyltransferase activity can be seen as a compensatory mechanism to maintain genomic methylation patterns, as only two rounds of replication are required for genomic hypomethylation if the activity of maintenance methylation provided by DNMT or other methyltransferases is not effective. The decline of S-adenosylmethionine could result from the excessive consumption due to the increased activity of DNMT (Lee et al.,1996).

Last but not least, it is now generally accepted that precise regulation of gene expression by epigenetic mechanisms is required to maintain the normal development of mammals. The epigenetic code is most likely to consist of both DNA methylation and histone modifications. But there still remains controversy on the role of the epigenetic code in the tissue-specific gene expression during the normal development of mammals, and the DNA methylation code has a critical effect that CpG methylation, which correlates with the suppression of cell transcription. In contrast, a variety of histone modifications, including acetylation and methylation at various lysine residues and other modifications, constitute the more complex histone code(Bonaldi et al., 2004). In general, acetylation of histone H3 and H4 correlates with gene activation,while deacetylation correlates with gene silencing(Zhang and Tang, 2003). Based on the results of our study, Hcy might operate with mechanisms that are involved in the upregulation of histone H3 and H4 in foam cells. Elucidation of the role of Hcy in the expression of EC-SOD in vivo, especially in the development of atherosclerotic plaque in patients with vascular disorder,could provide new insight into our understanding of hyperhomocysteine.

The findings of our study uncover the hypermethylation induced by Hcy in EC-SOD and the acetylation of histone H3 and H4(Fig. 10), similar to cancer tissues. It is possible that alterations in EC-SOD hypermethylation and the acetylation of histone H3 and H4 may play an important role in atherogenesis,though our study did not prove a direct relationship between EC-SOD hypermethylation, atherogenesis and the acetylation of histone H3 and H4. Moreover, it is obvious that AZC and TSA inhibit Hcy-induced EC-SOD. As we know, induction of EC-SOD gene hypermethylation and the acetylation of histone H3 and H4 via Hcy remain uninvestigated and unreported in the field of their related mechanisms. Therefore, in addition to important statistics and findings, this study has also revealed the novel and attractive role of Hcy in the pathogenesis of human cardiovascular disease.

LIST OF ABBREVIATIONS

 
• 5-mC

  5-methylcytosine

 
• AS

  artherosclerosis

 
• AZC

  azacytidine

 
• C-5MT-ase

  C-5 DNA rnethyltransferase

 
• CE

  cholesteryl ester

 
• ChIP

  chromatin immunoprecipitation

 
• Ct

  threshold cycle

 
• DNMT

  DNA methyltransferase

 
• EC-SOD

  extracellular superoxide dismutase

 
• ELISA

  enzyme-linked immunosorbent assay

 
• FC

  free cholesterol

 
• GAPDH

  glyceraldehyde-3-phosphate dehydrogenase

 
• HAT

  acetylase

 
• Hcy

  homocysteine

 
• HDAC

  histone deacetylase

 
• HRP

  horseradish peroxidase

 
• MBD

  methyl-CpG binding domain

 
• MeCP2

  binding of methyl CpG and the binding protein 2

 
• OPD

  O-phenylene diamine

 
• ox-LDL

  oxide-low-density lymphocyte

 
• PCR

  polymerase chain reaction

 
• PMA

  phorbol myristate acetate

 
• SMC

  smooth muscle cell

 
• TBST

  Tris-buffered saline plus Tween-20

 
• TC

  total cholesterol

 
• TE

  Triton-EDTA

 
• TSA

  trichostatin A

 
• VIC (450nm)

  victoria

This project was supported by the project of science and technology research of Ningxia Higher Education (Registration No. 2007.292) and the project of focal point scientific research of Health Department of Ningxia Hui autonomous region in China (Registration No. W200720).