Increased Superoxide Production in Coronary Arteries in Hyperhomocysteinemia
Role of Tumor Necrosis Factor-α, NAD(P)H Oxidase, and Inducible Nitric Oxide Synthase
Objective— In coronary arteries, hyperhomocysteinemia (HHcy, a known risk factor for coronary heart disease) impairs flow-induced dilations, which can be reversed by superoxide dismutase (SOD). To evidence increased O2.− generation and elucidate its source, we characterized changes in activity (lucigenin chemiluminescence, hydroethidine staining) and expression of arterial pro- and antioxidant systems (Western blotting, immunohistochemistry, cDNA microarray, reverse-transcription polymerase chain reaction) in the coronary arteries of rats by using methionine diet-induced HHcy.
Methods and Results— The increased generation of O2.− by HHcy coronary arteries was inhibited by SOD, diphenyleneiodonium, apocynin, and apocynin plus amino guanidine but was unaffected by allopurinol and rotenone. Also, diphenyleneiodonium-sensitive NADPH-driven O2.− generation was increased in HHcy vessels. In HHcy arteries expression of the smooth muscle-confined NAD(P)H oxidase subunit nox1 and that of iNOS was increased. Expression of p67phox, p22phox, and p47phox subunits and that of endothelial nitric oxide synthase, Cu,Zn-SOD, Mn-SOD, extracellular SOD (mRNA), and xanthine oxidase was unchanged. Microarray analysis showed increased expression of tumor necrosis factor (TNF)-α (confirmed by reverse-transcription polymerase chain reaction, Western blotting, and immunohistochemistry) that was localized in smooth muscle. In vitro incubation (18 hours) of HHcy arteries with anti-TNF-α antibody decreased O2.− production, whereas incubation of control vessels with TNF-α increased O2.− generation and nox1 expression.
Conclusions— In coronary arteries, HHcy increases TNF-α expression, which enhances oxidative stress through upregulating a nox1-based NAD(P)H oxidase and inducible nitric oxide synthase. Thus, TNF-α induces a proinflammatory vascular phenotype in HHcy that potentially contributes to the development of coronary atherosclerosis.
Hyperhomocysteinemia (HHcy) is an independent risk factor for coronary heart disease.1,2⇓ Homocysteine is formed during the metabolism of the essential amino acid methionine, the increased dietary intake of which, when often combined with other dietary (vitamin deficiencies) and/or genetic (enzyme abnormalities) factors, is a general cause of HHcy in humans.
See page 371
Previous studies have suggested that elevation of plasma homocysteine concentration impairs flow-induced dilation of small coronary arteries3 and relaxation of peripheral vessels4–6⇓⇓ by decreasing bioavailability of nitric oxide (NO), an effect that can be reversed by administration of superoxide (O2.−) scavengers.3,7⇓ The mechanisms leading to increased O2.− production in coronary arteries, however, have not been elucidated.
Potential vascular sources of O2.− include xanthine oxidase, cyclooxygenase, NO synthases (NOS), and NAD(P)H oxidases, the relative importance of which may vary among vascular beds and disease conditions. Recent studies suggest that in coronary arteries8,9⇓ NAD(P)H oxidases are the predominant source of O2.−. Indeed, although inhibition of xanthine oxidase or cyclooxygenase abolished pathological vasoconstriction in skeletal muscle arterioles of HHcy rats,6 it failed to restore NO-mediated responses in HHcy coronary arteries.3
The activity and expression of vascular NAD(P)H oxidases are regulated by humoral factors10 and proinflammatory cytokines, such as tumor necrosis factor (TNF)-α.11 Although recent studies have shown that in mice with genetic HHcy there is a 6-fold increase in plasma TNF-α levels12 and that high homocysteine levels in vitro elicit significant changes in the cytokine expression of vascular cells,13–15⇓⇓ the role of vascular TNF-α expression and NAD(P)H oxidase activity in eliciting coronary arterial dysfunction in diet-induced mild HHcy has not yet been elucidated.
Thus, we aimed to test the hypothesis that HHcy increases O2.− generation in coronary arteries by upregulating the vascular expression of TNF-α, which may increase the activity/expression of vascular NAD(P)H oxidases.
An expanded Methods and Materials section and Table I with the detailed description of the protocols used in the present study can be found online at http://atvb.ahajournals.org.
Methionine Diet-Induced Hyperhomocysteinemia and Isolation of Coronary Arteries
Male Wistar rats (300 to 330 g, Charles River Co.) were used. In one group, moderate HHcy was induced by administration of L-methionine (10 g/L in the drinking water) for a period of 4 weeks. The septal artery was isolated from the hearts of control (C, n=40) and HHcy rats (n=40) as previously described.3,16⇓ Serum homocysteine concentrations, measured with a commercially available kits (Abbott Diagnostics Co, Sigma Co), were 3.9±0.8 μmol/L and 29.9±5.5 μmol/L in C and HHcy rats, respectively.
Measurement of Vascular Superoxide Level
Vascular O2.− production was assessed in coronary arterial samples by the lucigenin chemiluminescence (CL) method8,16⇓ in the absence or presence of superoxide dismutase (SOD, 200 U/mL), 4,5-dihydroxy-1,3-benzene-disulphonic acid (Tiron, a O2.− scavenger in coronary vessels,8 10 mmol/L), rotenone (10−5 mol/L, a proximal inhibitor of the mitochondrial respiratory chain), allopurinol (10−4 mol/L, an inhibitor of xanthine oxidase) diphenyleneiodonium ([DPI], 10−4 mol/L, an inhibitor of flavin-containing enzymes, including NAD(P)H oxidases9), apocynin (4-hydroxy-3-methoxy-acetophenone, 3×10−4 mol/L, an inhibitor of superoxide production by NAD(P)H oxidases), apocynin plus amino guanidine (10−4 mol/L, an inhibitor of inducible NOS [iNOS]), or SOD or polyethyleneglycol (PEG)-SOD (200 U/mL).
In separate experiments, O2.− production of aortic segments from HHcy rats was determined in the absence and presence (30 minutes incubation) of DPI (10−4 mol/L) or apocynin or SOD (200 U/mL) or after endothelium removal (by rubbing).
NAD(P)H Oxidase Assay
NAD(P)H oxidase activity in homogenates of coronary vessels of C and HHcy rats was measured by the lucigenin assay8,11,16⇓⇓ after addition of 10−4 mol/L NADPH in the absence or presence of rotenone, aminoguanidine, DPI, or apocynin. Protein content was measured in an aliquot of the homogenate by the Lowry method.
Ethidium Bromide Fluorescence
Hydroethidine was used to localize arterial O2.− production as previously described.16 In brief, cells are permeable to hydroethidine, which in the presence of O2.− is oxidized to fluorescent ethidium bromide (EB) that is trapped by intercalation with DNA. The sections were also counterstained with hematoxylin and immunostained for α-smooth muscle actin. C and HHcy tissues were processed and imaged in parallel. Samples exposed to hydroethidine in the presence of DPI (10−4 mol/L) and Tiron (10 mmol/L) served as control.
Protein expression of endothelial NOS (eNOS) and inducible NOS (iNOS), Cu,Zn-SOD, Mn-SOD, p22phox, p47phox, p67phox, nox-1, gp91phox, xanthine oxidase, and TNF-α in coronary arteries of C and HHcy rats was assessed by Western blotting as described previously.3,16⇓ Anti-GAPDH was used for normalization.
Estimation of Vascular TNF-α Content
Coronary arterial segments from three HHcy rats were weighed, proteins were extracted, and Western blotting was performed with known amounts of recombinant rat TNF-α (0.002 to 200 ng) loaded as controls. TNF-α content (DTNF) in HHcy vessels was estimated using a densitometric calibration curve, and nominal tissue concentrations (CTNF) were computed according to the following equation: CTNF = DTNF × (m/ρvasc)−1, where m is the tissue mass and ρvasc is the estimated average density of vascular tissue (1.06 g/cm3). This estimation does not take into consideration intra-tissue compartmentalization of TNF-α.
Localization of NAD(P)H Oxidase Subunits and TNF-α with Dual Immunofluorescence Staining and Immunohistochemistry
Immunohistochemical detection of p22phox, p47phox, p67phox, or TNF-α in sections of the left ventricle was performed as described previously.3,16⇓ Nox-1, TNF-α, and α-smooth muscle actin were detected using dual immunofluorescent labeling. Nuclei were stained with 4′,6-diamidino-2-phenylindole. Immunolabeling for CD45 (a leukocyte surface marker) was also performed.
RNA Isolation and Real-Time Reverse-Transcription Polymerase Chain Reaction (RT-PCR)
Total RNA from frozen arterial samples was isolated as described.16 RT-PCR technique was used to determine expression levels of TNF-α, ecSOD, and iNOS mRNA in control and HHcy coronary arteries, as described previously.16 PCRs were performed using the Roche Molecular Biochemicals LightCycler System. The housekeeping gene β-actin was used for internal normalization. Oligonucleotids used for RT-PCR are included in Table I.
Gene Expression Profiling with cDNA-Based Microarray
Expression of 96 mediators and enzymes that are representative markers of major signaling pathways regulating cellular functions were screened by a cDNA-based microarray technique by using the GEArray™ Q series nonradioactive Signal Transduction Pathway Finder Array (Superarray, Bethesda, MA) using total RNA (2 μg) from coronary arteries of control (n=5) and HHcy (n=5) rats according to the instructions of the manufacturer. Membranes were developed using CDP-Star as chemiluminescent substrate. Images were digitalized and analyzed with the Scanalyze (http://www. microarrays.org/software.html) and GEArrayAnalyzer (Superarray) software. Background-corrected densities of each tetra-spots representing each gene were averaged, normalized to the signals of the housekeeping gene β-actin, and the mean value and standard deviation of normalized data from each array were computed. A change in the expression of a gene was considered statistically significant if there was >2.0-fold change in the normalized signals (P<0.05, with Student paired t test).
Vessel Culture Studies
Coronary arteries of HHcy rats were incubated for 18 hours (at 37°C) in a vessel culture system according to the protocol of Bakker et al17 in the absence or presence of a monoclonal anti-rat TNF-α antibody (R&D Systems) to neutralize bioactivity of endogenous TNF-α. Also, coronary arteries from control rats were incubated in the absence and presence of recombinant TNF-α (100 ng/mL). Superoxide production was assessed with the lucigenin CL method. Expression of NAD(P)H oxidase subunits was measured by Western blotting. In separate experiments control aortic segments were incubated with TNF-α, and O2.− production was measured in the absence and presence (30 minutes incubation) of SOD and DPI.11
All chemicals, if otherwise not specified, were obtained from Sigma-Aldrich Co., and solutions were prepared on the day of the experiment. Final concentrations are reported in the text.
Lucigenin CL data and densitometric ratios were normalized to the respective control mean values. Data are expressed as means±SEM. Statistical analyses of data were performed by Student t test or by two-way ANOVA followed by the Tukey post hoc test, as appropriate. P<0.05 was considered statistically significant.
An extended Results section with additional figures and tables can be found online at htttp://atvb.ahajournals.org).
Vascular Superoxide Production
Under basal conditions, lucigenin CL in coronary vessels of HHcy rats was significantly higher than in vessels from C rats (C: 102±12×103 counts/mg/min). Increased arterial lucigenin CL in HHcy was inhibited by either SOD or Tiron (Figure 1A) or PEG-SOD (Figure 1B). In separate experiments, increased O2.− production in HHcy vessels was unaffected by either rotenone or allopurinol and was inhibited by both DPI and apocynin (Figure 1B). The effect of apocynin was potentiated by co-administration of aminoguanidine (Figure 1B).
NADPH-dependent (Figure 1C) and NADH-dependent (not shown) O2.− production by homogenized samples of coronary arteries of HHcy rats was significantly increased compared with controls and was inhibited by DPI and apocynin.
In aortic segments of HHcy rats, there was also a significantly increased O2.− production (C: 100±9%, HHcy: 231±34%, C+SOD: 52±3%, HHcy+SOD: 40±8%) confirming results of previous studies in mice.7 Increased O2.− generation by HHcy aortas was inhibited by the administration of SOD, DPI, or apocynin (Figure 1D) but was not significantly affected by removal of the endothelium.
Representative fluorescent photomicrographs of EB-stained control and HHcy coronary arterial sections are shown on Figure 1E (upper left and right, respectively). Incubation of vessels with DPI or Tiron greatly reduced EB staining of nuclei (Figure 1E, lower left and right, respectively). Overlaying of EB-stained fluorescent and hematoxylin-stained bright-field images of the same HHcy vessels (Figure 1F, middle) or EB-stained fluorescent images with images of the same vessel sections stained (green) for smooth muscle α-actin (Figure 1F, right) showed that increased O2.− levels were present in the smooth muscle of HHcy vessels.
Arterial Expression of Pro- and Antioxidant Enzymes
HHcy did not alter the coronary arterial protein expression of eNOS, Cu,Zn-SOD, Mn-SOD, and xanthine oxidase (see online Figures I and II, which can be accessed at http://atvb.ahajournals.org) and mRNA expression of ecSOD (ecSOD:β-actin mRNA ratio: C: 0.69±0.01, HHcy: 0.71±0.01, NS) was also unaffected by HHcy. Expression of iNOS protein (Figure 2A) and mRNA (iNOS: β-actin mRNA ratio: C: 0.46±0.01, HHcy: 0.65±0.01) was significantly increased in HHcy arteries. In the arteries of HHcy rats, expression of the NAD(P)H oxidase subunit nox1 (Figure 2B) was significantly increased, whereas expression of p22phox and p47phox subunits were unaffected. Expression of p67phox subunit also tended to increase in HHcy vessels; however, the difference did not reach statistical significance (see Figure III, which can be accessed at http://atvb.ahajournals.org). Expression of gp91phox in coronary arteries could not be detected reliably with this method.
Immunohistochemical Localization of NAD(P)H Oxidase Subunits
In coronary arteries of HHcy rats, immunofluorescent labeling for nox-1 overlapped with the smooth muscle α-actin staining (Figure III). Immunostaining for the p22phox and p67phox subunits was present both in the arterial endothelium and smooth muscle. Immunostaining for p47phox was present in the endothelium, whereas the media had lower levels of immunoreactivity (Figure III). Immunostaining for the leukocyte-specific surface marker CD45 could not be detected in HHcy arteries (not shown).
Using a cDNA-based microarray containing 96 signal transduction–related genes, expression of 29 different gene transcripts could be detected in control and HHcy rat coronary arterial samples (see online Table II, which can be accessed at http://atvb.ahajournals.org). The reproducibility of our procedures was tested by hybridizing identical samples from one vessel to two microarrays yielding a correlation coefficient of r=98. In HHcy coronary arteries, mRNA expression of the pro-inflammatory cytokines TNF-α and TNF-β (which act on the same vascular receptor and have similar biological effects), the cytokine-inducible iNOS, the cell cycle regulator CDK2L18 and the proatherosclerotic transforming growth factor-β superfamily member BMP2 was increased over control vessels (Table II).
Arterial Expression of TNF-α
In HHcy arteries, there was a significantly increased expression of TNF-α protein (Figure 3A). In control arteries TNF-α mRNA content was very low, whereas in HHcy vessels expression of TNF-α mRNA was elevated (Figure 3B). In HHcy immunostaining of coronary arteries for TNF-α (Figure 3C and D, brown) was localized to the smooth muscle, whereas the endothelium was relatively free of immunoreactivity (Figure 3D, arrowhead). Also, immunofluorescent labeling for TNF-α (red, Figure 3F, top) overlapped with smooth muscle α-actin staining (green, Figure 3F, middle). Arteries of C rats did not show immunolabeling for TNF-α (Figure 3E).
Effect of TNF-α on Vascular Superoxide Production
Incubation of HHcy coronary arteries with a TNF-α blocking antibody significantly decreased O2.− production (Figure 4A). Using Western blotting tissue TNF-α levels in HHcy coronary arteries were estimated to be 100 to 150 ng/mL (Figure 4B). Incubation of control coronary arteries (for 18 hours) with 100 ng/mL TNF-α significantly increased O2.− production (Figure 4C) and increased expression of nox-1 (see online Table III, which can be accessed at http://atvb.ahajournals). TNF-α incubation also increased O2.− generation in control aortic segments (C: 100±5%, C+TNF: 176±24%), which could be inhibited by administration of SOD or DPI.
The new findings of the present study are that in HHcy coronary arteries there is an (1) increased O2.− production; (2) upregulation of NAD(P)H oxidase and iNOS; and (3) increased expression of TNF-α.
Increased SOD-sensitive O2.− production in HHcy coronary arteries accords with our previous findings showing impaired flow-induced, NO-mediated dilations in these vessels that could be improved by SOD.3 It is likely that NAD(P)H oxidases are the major source of O2.− generation in HHcy coronary vessels because DPI and apocynin inhibited increased lucigenin CL (Figure 1A and B) and improved endothelial function3 in HHcy vessels, whereas inhibitors of other potential sources of O2.− were without effect. Furthermore, increased NADPH-driven O2.− generation in HHcy vessels also supports a significant role for increased NAD(P)H oxidase activity (Figure 1C). Because co-administration of aminoguanidine potentiated the effect of apocynin on O2.− production (Figure 1B), it is likely that iNOS also generates O2.− in HHcy coronary arteries.
EB staining showed that a significant number of superoxide-producing cells were localized in the media of HHcy coronary arteries, which were also positive for smooth muscle α-actin (Figure 1E and F). This finding, taken together with results that removal of the endothelium did not significantly affect O2− generation by HHcy aortic segments (Figure 1D), suggests that in HHcy it is the smooth muscle cells of coronary arteries and aorta that generate the bulk of O2.−. Because previous studies also have shown that in HHcy coronary vessels3 and aorta7 there is an increased protein 3-nitrotyrosine content in smooth muscle cells, it is likely that the increased NAD(P)H oxidase-derived O2.− levels in smooth muscle scavenge flow-induced endothelium-derived NO, decreasing its signaling function and resulting in increased formation of ONOO−.
Comparison of the expression profile of vascular pro- and antioxidant enzymes that are involved in the metabolism of O2.− and NO revealed that in HHcy the expression of iNOS was significantly increased (Figure 2A), supporting its role in increased O2.− production (Figure 1B) and/or ONOO− generation.3 The expression of eNOS, SOD isoforms, and xanthine oxidase (Figures I and II) was unaffected by HHcy, suggesting that a decreased presence of eNOS or SOD or upregulation of xanthine oxidase does not contribute significantly to HHcy-induced oxidative stress in these vessels. Although antioxidant systems distinct from SOD (eg, glutathione peroxidase19) may also be altered by high homocysteine levels, it is unclear whether they interfere with O2.− production in HHcy coronary arteries.
Endothelial and smooth muscle cells express different NAD(P)H oxidases that consist of multiple subunits.10,20⇓ We investigated the expression of four NAD(P)H oxidase subunits known to be involved in vascular O2− production. The membrane-associated nox-1 subunit,10,20⇓ which is a member of a growing family of gp91phox homologues, and the p22phox subunit are essential for the oxidase activity. Activation of the NAD(P)H oxidases is also thought to involve binding of p47phox and p67phox to the membrane oxidase complex, allowing the assembled enzyme complex to facilitate electron transfer from NADH/NADPH to molecular oxygen, forming O2.−. In coronary arteries of HHcy rats expression of the membrane-associated NAD(P)H oxidase subunit nox-1 is significantly increased (Figure 2B). Immunofluorescent labeling showed that expression of nox-1 in HHcy coronary arteries was primarily confined to smooth muscle (Figure III). Previous studies also demonstrated expression of nox-1 in vascular smooth muscle10,20⇓ and its absence in endothelial cells. HHcy did not seem to affect the arterial expressions of p22phox, p67phox, and p47phox (Figure III). Taken together, these findings support the hypothesis that HHcy elicits oxidative stress in the coronary circulation by upregulating a nox1-based NAD(P)H oxidase in the arterial smooth muscle. Our findings may have important clinical applications because recent studies have demonstrated an upregulation of nox-based NAD(P)H oxidases in arterial smooth muscle in human coronary atherosclerosis21 which is often associated with HHcy.1,2⇓
One of the known mechanisms that regulate vascular NAD(P)H oxidase expression/activity is the autocrine–paracrine mediator TNF-α.11,22⇓ Microarray analysis of gene expression profile indicated a substantially increased expression of TNF-α in HHcy coronary arteries. Increased expression of TNF-α mRNA was confirmed by RT-PCR (Figure 3B) and increased TNF-α protein content was demonstrated by Western blotting (Figure 3A), raising the possibility that this mechanism may contribute to upregulation of NAD(P)H oxidases in HHcy.
Immunohistochemistry suggests that in HHcy, TNF-α is predominantly produced by smooth muscle cells23 (Figure 3C,D,F). Recent studies also showed that high homocysteine levels in vitro may induce cytokine expression in endothelial15 and smooth muscle cells14 and that plasma level of TNF-α increases in different forms of HHcy in humans and mice.12,24⇓ The possibility that locally produced TNF-α contributes importantly to the development of vascular oxidative stress in HHcy is suggested by the findings that incubation with a TNF-α blocking antibody substantially decreased O2.− generation in HHcy coronary arteries (Figure 4A). TNF-α is likely to have multiple effects on vascular NAD(P)H oxidase, including induction of its expression (similar to its effect on iNOS). Indeed, incubation with exogenous TNF-α in vitro upregulated NAD(P)H oxidase expression (Table III) and O2.− production in smooth muscle cells11 and control arteries (Figure 4C). Previous studies also have shown that increased vascular O2.− production in proinflammatory conditions is often associated with upregulation of one or more NAD(P)H oxidase subunits.20 In addition to the transcriptional control, TNF-α may also stimulate NAD(P)H oxidase activity in a protein kinase C–dependent manner.22 On the basis of the aforementioned findings, it is logical to hypothesize that in HHcy coronary arteries TNF-α upregulates NAD(P)H oxidase–dependent O2.− production both by induction of the oxidase subunit nox1 and an increased phosphorylation of the regulatory subunits. Increased vascular production of TNF-α is also likely to be linked to the increased iNOS expression/activity in coronary arteries (Figure 2A) of HHcy rats.
On the basis of the present and previous findings12 we propose that HHcy-induced oxidative stress and activation of cytokine expression reflect a proinflammatory state of the vessels (Figure 4D). Increased vascular production of proinflammatory cytokines, such as TNF-α, upregulates NAD(P)H oxidase and iNOS, resulting in increased oxidative and nitrosative stress and a reduction in the bioavailability of NO.3 These factors may also induce further changes in the endothelial and smooth muscle gene expression profile resulting in a proatherosclerotic vascular phenotype (eg, vasomotor dysfunction, endothelial activation, expression of adhesion molecules and chemokines, and smooth muscle proliferation),12,14,25⇓⇓ ultimately leading to atherogenesis, favoring ischemic heart disease.
This study was supported by grants from NIH PO 43023, HL-59417 and HL-46813; AHA NY State Aff. Inc 00500849T, 0020144T and 0120166T; and Hungarian Science Research Fund (OTKA-T 034779 and T 033117).
Zoltan Ungvari and Anna Csiszar contributed equally to this work.
Received November 13, 2002; revision accepted December 11, 2002.
- ↵Ungvari Z, Sarkadi-Nagy E, Bagi Z, Szollar L, Koller A. Simultaneously increased TxA2 activity in isolated arterioles and platelets of rats with hyperhomocysteinemia. Arterioscler Thromb Vasc Biol. 2000; 20: 1203–1208.
- ↵Bagi Z, Ungvari Z, Koller A. Xanthine oxidase-derived reactive oxygen species convert flow-induced arteriolar dilation to constriction in hyperhomocysteinemia. Arterioscler Thromb Vasc Biol. 2002; 22: 28–33.
- ↵Eberhardt RT, Forgione MA, Cap A, Leopold JA, Rudd MA, Trolliet M, Heydrick S, Stark R, Klings ES, Moldovan NI, Yaghoubi M, Goldschmidt-Clermont PJ, Farber HW, Cohen R, Loscalzo J. Endothelial dysfunction in a murine model of mild hyperhomocyst(e)inemia. J Clin Invest. 2000; 106: 483–491.
- ↵Mohazzab KM, Kaminski PM, Wolin MS. NADH oxidoreductase is a major source of superoxide anion in bovine coronary artery endothelium. Am J Physiol. 1994; 266: H2568–2572.
- ↵Mohazzab HK, Kaminski PM, Fayngersh RP, Wolin MS. Oxygen-elicited responses in calf coronary arteries: role of H2O2 production via NADH-derived superoxide. Am J Physiol. 1996; 270: H1044–1053.
- ↵Lassegue B, Sorescu D, Szocs K, Yin Q, Akers M, Zhang Y, Grant SL, Lambeth JD, Griendling KK. Novel gp91(phox) homologues in vascular smooth muscle cells: nox1 mediates angiotensin II-induced superoxide formation and redox-sensitive signaling pathways. Circ Res. 2001; 88: 888–894.
- ↵De Keulenaer GW, Alexander RW, Ushio-Fukai M, Ishizaka N, Griendling KK. Tumour necrosis factor alpha activates a p22phox-based NADH oxidase in vascular smooth muscle. Biochem J. 1998; 329: 653–657.
- ↵Chacko G, Ling Q, Hajjar KA. Induction of acute translational response genes by homocysteine: elongation factors-1alpha, -beta, and -delta. J Biol Chem. 1998; 273: 19840–19846.
- ↵Poddar R, Sivasubramanian N, DiBello PM, Robinson K, Jacobsen DW. Homocysteine induces expression and secretion of monocyte chemoattractant protein-1 and interleukin-8 in human aortic endothelial cells: implications for vascular disease. Circulation. 2001; 103: 2717–2723.
- ↵Csiszar A, Ungvari Z, Edwards JG, Kaminski PM, Wolin MS, Koller A, Kaley G. Aging-induced phenotypic changes and oxidative stress impair coronary arteriolar function. Circ Res. 2002; 90: 1159–1166.
- ↵Bakker ENTP, van der Meulen ET, Spaan JAE, VanBavel E. Organoid culture of cannulated rat resistance arteries: effect of serum factors on vasoactivity and remodeling. Am J Physiol Heart Circ Physiol. 2000; 278: H1233–H1240.
- ↵Lubec B, Labudova O, Hoeger H, Muehl A, Fang-Kircher S, Marx M, Mosgoeller W, Gialamas J. Homocysteine increases cyclin-dependent kinase in aortic rat tissue. Circulation. 1996; 94: 2620–2625.
- ↵Weiss N, Zhang YY, Heydrick S, Bierl C, Loscalzo J. Overexpression of cellular glutathione peroxidase rescues homocyst(e)ine-induced endothelial dysfunction. Proc Natl Acad Sci U S A. 2001; 98: 12503–12508.
- ↵Szocs K, Lassegue B, Sorescu D, Hilenski LL, Valppu L, Couse TL, Wilcox JN, Quinn MT, Lambeth JD, Griendling KK. Upregulation of Nox-based NAD(P)H oxidases in restenosis after carotid injury. Arterioscler Thromb Vasc Biol. 2002; 22: 21–27.
- ↵Sorescu D, Weiss D, Lassegue B, Clempus RE, Szocs K, Sorescu GP, Valppu L, Quinn MT, Lambeth JD, Vega JD, Taylor WR, Griendling KK. Superoxide production and expression of nox family proteins in human atherosclerosis. Circulation. 2002; 105: 1429–1435.
- ↵Frey RS, Rahman A, Kefer JC, Minshall RD, Malik AB. PKCzeta regulates TNF-alpha-induced activation of NADPH oxidase in endothelial cells. Circ Res. 2002; 90: 1012–1019.
- ↵Warner SJ, Libby P. Human vascular smooth muscle cells: target for and source of tumor necrosis factor. J Immunol. 1989; 142: 100–109.
- ↵De Keulenaer GW, Ushio-Fukai M, Yin Q, Chung AB, Lyons PR, Ishizaka N, Rengarajan K, Taylor WR, Alexander RW, Griendling KK. Convergence of redox-sensitive and mitogen-activated protein kinase signaling pathways in tumor necrosis factor-alpha-mediated monocyte chemoattractant protein-1 induction in vascular smooth muscle cells. Arterioscler Thromb Vasc Biol. 2000; 20: 385–391.