Opposing Roles of p47phox in Basal Versus Angiotensin II–Stimulated Alterations in Vascular O2− Production, Vascular Tone, and Mitogen-Activated Protein Kinase Activation
Background— NADPH oxidase is a major source of vascular superoxide (O2−) production and is implicated in angiotensin II (Ang II)–induced oxidant stress. The p47phox subunit plays an important role in Ang II–induced oxidase activation, but its role in basal oxidase activity and vascular function is unclear.
Methods and Results— Aortae from p47phox−/− and matched wild-type (WT) mice (n=9/group) were incubated ex vivo with or without Ang II (200 nmol/L, 30 minutes) and then examined for (1) NADPH-dependent O2− production, (2) endothelium-dependent and -independent vascular relaxation, and (3) activation of mitogen-activated protein kinases (MAPKs). In the absence of Ang II, p47phox−/− vessels had slightly but significantly higher (1.3±0.1-fold; P<0.05) NADPH-dependent O2− production than WT; impaired relaxation to acetylcholine (maximum 54±4% versus 80±3%; P<0.05), which was normalized to WT levels by the O2− scavenger tiron or by Mn(III)tetrakis(1-methyl-4-pyridyl)porphyrin pentachloride, and increased basal phosphorylation of ERK1/2, p38MAPK, and JNK compared with WT. In WT aortae, Ang II increased NADPH-dependent O2− production (2.5±0.5-fold; P<0.05), impaired relaxation to acetylcholine (maximum 60±6% versus 80±3%; P<0.05), and increased ERK1/2, p38MAPK, and JNK phosphorylation (P<0.05). In contrast, Ang II failed to increase O2− production, impair acetylcholine responses, or increase MAPK activation in p47phox−/− aortae.
Conclusions— p47phox plays a complex dual role in the vasculature. It inhibits basal NADPH oxidase activity but is critical for Ang II–induced vascular dysfunction via activation of NADPH oxidase.
Received February 13, 2003; de novo received July 14, 2003; revision received October 30, 2003; accepted October 31, 2003.
Activation of the renin-angiotensin system and increased production of angiotensin II (Ang II) are implicated in the pathogenesis of hypertension, atherosclerosis, and cardiac hypertrophy.1 A major mechanism through which Ang II induces pathological effects is by generation of superoxide (O2−) and other reactive oxygen species (ROS).2 Ang II–stimulated increases in ROS are involved in the genesis of endothelial dysfunction (through inactivation of endothelium-derived nitric oxide [NO]),3 activation of signaling molecules such as mitogen-activated protein kinases (MAPKs),2,4 inflammatory cytokine release,5 and vascular smooth muscle (VSM) and cardiac hypertrophy.2,6–8
Recently, a phagocyte-type NADPH oxidase has been found to be an important source of Ang II–induced vascular ROS generation and to be involved in the downstream pathological processes discussed above.2–9 NADPH oxidases are expressed in several different cell types in the vessel wall, including endothelial cells, VSM, adventitial fibroblasts, and inflammatory cells.6,7 The prototypic NADPH oxidase complex comprises a membrane-bound cytochrome b558 heterodimer, which forms the catalytic core of the enzyme, and 4 regulatory subunits (p47phox, p67phox, p40phox, and rac1).10 In contrast to the phagocyte NADPH oxidase, the vascular oxidase has distinct biochemical properties (eg, continuous low-level O2− generation in the absence of agonist stimulation) and a different subcellular location.9,11
Detailed analyses of the roles of the different oxidase subunits in regulation of vascular enzyme activity remain limited. In neutrophils, p47phox is critical for NADPH oxidase activation.10 Several recent studies using genetic manipulation of p47phox have indicated an essential role of this subunit in endothelial and VSM NADPH oxidase activation by thrombin, phorbol esters, cytokines, and growth factors.12–16 An important role for p47phox in atherosclerosis was suggested by the finding that lesion progression was reduced in apolipoprotein E−/− mice lacking p47phox.15
Mice lacking p47phox were reported to lack a vascular oxidative stress response to Ang II stimulation, whereas Ang II–induced hypertension was markedly blunted in these animals.17 Ang II–induced hypertension in mice was also blunted by an inhibitor that prevents association of p47phox with cytochrome b558.18 Schieffer et al5 reported that Ang II–induced JAK/STAT signaling by VSM required p47phox. However, the possible role of p47phox in Ang II–induced vascular endothelial dysfunction or MAPK activation has not been reported. Likewise, the potential role of p47phox in “basal” vascular function (ie, in the absence of agonist activation) is unclear. Intriguingly, previous in vitro studies in endothelial cells and VSM found that p47phox was not essential for basal NADPH-dependent O2− production.12–14 Indeed, in endothelial cells, we found that genetic depletion of p47phox unexpectedly resulted in increased basal NADPH-dependent O2− production.14
In this study, we investigated the pathophysiological consequences of p47phox deficiency on vascular O2− production, endothelium-dependent relaxation, and MAPK activation both under basal conditions and after acute exposure to Ang II. We report that p47phox plays a complex dual role, not only being critical for Ang II–induced vascular dysfunction but also being involved in regulation of basal NADPH oxidase activity and vascular function.
p47phox−/− mice were kindly provided by Dr Jurgen Roes (University College London, London, UK)19 and a colony established in our institution. Matched WT mice on an identical 129sv background served as controls. Animals were euthanized at ≈10 weeks of age for tissue isolation.
Vascular Ring Experiments
Thoracic aortae (n=9 mice per group) were dissected free from surrounding tissue and cut into rings 3 to 4 mm long. Rings were incubated at 37°C in serum-free medium 199 with or without Ang II (200 nmol/L) for 30 minutes. They were then suspended in organ baths (37°C) gassed with 95% O2/5% CO2 containing Krebs-Henseleit solution (in mmol/L: NaCl 119, KCl 4.7, KH2PO4 1.2, MgSO4 1.2, CaCl2 2.5, NaHCO3 25, and glucose 11.1, pH 7.4). Vessels were progressively stretched to the optimal resting tension (3 g) that produced a maximal response to 40 mmol/L KCl. Relaxation to acetylcholine (ACh, 0.001 to 10 μmol/L) was tested in rings preconstricted to 70% of their maximal phenylephrine-induced tension. Some experiments were performed in the presence of the cell-permeable O2− scavenger tiron (5 mmol/L) or the superoxide dismutase (SOD) mimetic Mn(III)tetrakis(1-methyl-4-pyridyl)porphyrin pentachloride (MnTMPyP, 10 μmol/L). Relaxation to the NO donor sodium nitroprusside (SNP; 0.1 nmol/L to 1 μmol/L) was tested in parallel rings.
Vascular ROS Production
NADPH-dependent O2− production by aortic rings and aortic homogenates (n=9 mice per group) was measured by lucigenin (5 μmol/L)-enhanced chemiluminescence using a microplate luminometer (Anthos Lucy 1; 37°C) as described previously.11,14,20 Chemiluminescence is reported as arbitrary light units per minute over a period of 20 minutes. All readings were undertaken in triplicate. In some experiments, one of the following agents was preincubated with rings or homogenate for 10 minutes: SOD (200 U/mL), tiron (5 mmol/L), the flavoprotein inhibitor diphenyleneiodonium (DPI, 20 μmol/L), the NO synthase (NOS) inhibitor NG-nitro-l-arginine methyl ester (L-NAME, 100 μmol/L), the mitochondrial electron transport chain inhibitor rotenone (50 μmol/L), or the xanthine oxidase inhibitor oxypurinol (100 μmol/L).20
ROS generation within the aorta in situ in the presence or absence of Ang II was measured with dihydroethidium (DHE) fluorescence.15,18 Fluorescence intensity was quantified microscopically from at least 5 random fields (1024×1022 pixels; 269.7×269.2 μm) per slide and 3 slides per experimental condition by use of a computerized image analysis system (Improvision).
Tissue homogenates were sonicated and extracted on ice with 0.5% Triton X-100.11 After centrifugation (12 000 g, 15 minutes, 4°C), soluble protein concentration was determined with a Bio-Rad kit (Bio-Rad Laboratories). Equal amounts of protein (25 μg per sample) were separated on 10% polyacrylamide gels and transferred to polyvinylidine difluoride membranes. Membranes were blocked with 5% nonfat milk/PBS/0.2% TWEEN 20 before specific labeling. The following antibodies were used: anti–phospho-JNK polyclonal (1/1000), anti–phospho-ERK1/2 monoclonal (1/1000, Santa Cruz), anti-p38MAPK polyclonal (1/1000, Cell Signaling Technology), anti–xanthine oxidase polyclonal (1/2000, Abcom), anti–endothelial NOS (eNOS) polyclonal (1/1000, Transduction), anti–α-tubulin monoclonal (1/2000, Sigma), and anti–smooth muscle actin polyclonal (1/2000, Sigma). Bands were detected by enzyme-linked chemiluminescence (Amersham) and quantified by densitometry, with normalization for α-tubulin or smooth muscle actin.
Samples and tissue sections were prepared as described previously.21 Frozen sections (6 μm) were treated with a Biotin Blocking kit (Dako), then incubated with a mouse monoclonal antibody to phospho-ERK1/2 diluted 1:50 to 1:250 in 0.1% BSA/PBS for 30 minutes at room temperature. Biotin-conjugated anti-mouse IgG (1:500) was used as a secondary and detected by FITC-labeled extravidin. Normal mouse IgG (5 μg/mL) was used instead of primary antibody as a negative control.
Data are presented as mean±SD of at least 9 animals for each condition in each group. Comparisons were made by unpaired t test, 1-way ANOVA, or a repeated-measures ANOVA as appropriate. A value of P<0.05 was considered significant.
Basal O2− Generation and Endothelium-Dependent Relaxation to ACh in WT and p47phox−/− Aortae
Basal NADPH-dependent O2− production measured by lucigenin chemiluminescence was significantly higher both in aortic rings and in vessel homogenates from p47phox−/− mice compared with WT (Figure 1A). This was true for total chemiluminescence and tiron-inhibitable chemiluminescence (Figure 1A). O2− production was significantly inhibited by SOD, tiron, or DPI but was unaffected by L-NAME, rotenone, or oxypurinol (data for p47phox−/− homogenates are shown in Figure 1A).
In parallel experiments, we studied vascular relaxation in aortic rings from the same mice. Endothelium-independent relaxation to SNP was identical in WT and p47phox−/− groups (Figure 1B). However, relaxation to ACh was significantly impaired in p47phox−/− rings (Figure 1B, second panel), with maximal relaxation of 54±4% for p47phox−/− versus 80±4% for WT (P<0.05). EC50 values were not significantly different between groups. In the presence of either the O2− scavenger tiron or the SOD mimetic MnTMPyP, ACh-induced relaxation of p47phox−/− rings was normalized to the WT level (Figure 1B, third and fourth panels), suggesting that the impaired ACh response was because of increased O2− production in p47phox−/− rings.
Effects of Acute Ang II Treatment on NADPH-Dependent O2− Generation and Endothelium-Dependent Vascular Relaxation in WT and p47phox−/− Aortae
After acute incubation of aortic rings with Ang II (200 nmol/L, 30 minutes), the tiron-inhibitable NADPH-dependent O2− generation by homogenates of WT vessels was increased by >2-fold (from 0.66±0.05 to 1.40±0.08 mean arbitrary light units/mg protein per minute; P<0.01). In contrast, Ang II had no significant effect on NADPH-dependent O2− generation by p47phox−/− vessels (from 0.91±0.06 to 0.76±0.04 mean arbitrary light units/mg protein per minute; P=NS).
Aortic segments from the same mice exposed to Ang II (or vehicle) in parallel were examined for endothelium-dependent relaxation (Figure 3). In WT aortae, Ang II significantly reduced the maximal response to ACh (from 80±3% to 60±6%, P<0.05). This impairment was restored to the control level in the presence of either tiron or MnTMPyP. In contrast, Ang II did not have any significant effect on ACh-induced relaxation in p47phox−/− aorta.
In Situ ROS Production
To assess ROS generation in situ and to confirm the results obtained with lucigenin chemiluminescence, we used DHE staining of frozen aortic sections. A low basal level of DHE fluorescence was detected in WT aorta and was markedly increased in rings exposed to Ang II (Figures 4 and 5⇓). DHE fluorescence was obvious throughout the vessel wall. Basal DHE fluorescence in p47phox−/− aorta was significantly higher than in WT aorta (Figure 5). However, exposure to Ang II caused no increase in DHE fluorescence in p47phox−/− aorta (Figures 4 and 5⇓). DHE fluorescence was significantly reduced by catalase in both groups (Figures 4 and 5⇓).
To investigate potential downstream effects of NADPH oxidase–derived ROS on vascular Ang II signaling, we examined the phosphorylation of ERK1/2, p38-MAPK, and JNK in vessels treated acutely with or without Ang II (Figure 6). In WT vessels in the absence of Ang II, levels of phospho-p38 MAPK and phospho-JNK were very low, whereas phospho-ERK1/2 was virtually undetectable. In Ang II–treated WT vessels, there was a substantial (≈12-fold) increase in ERK1/2 phosphorylation, whereas the levels of phosphorylated p38MAPK and JNK were modestly but significantly increased.
In p47phox−/− aortas in the absence of Ang II, there was already significant ERK1/2, p38MAPK, and JNK phosphorylation, all of which were slightly but significantly higher than corresponding WT levels (Figure 6). However, exposure of p47phox−/− aorta to Ang II did not result in any further increase in MAPK phosphorylation. A densitometric quantification of data from 9 mice in each group is shown in the online-only Data Supplement figure.
Because phospho-ERK1/2 showed the most distinct changes with Ang II stimulation, we examined the cellular location of phospho-ERK1/2 in the aortic wall (Figure 7). WT vessels in the absence of Ang II showed no clear detectable phospho-ERK1/2 (Figure 7A). Ang II caused a dramatic increase in ERK1/2 phosphorylation, which was detectable throughout the vessel wall (Figure 7B). Interestingly, in p47phox−/− vessels in the absence of Ang II, clear labeling for phospho-ERK1/2 was evident in the endothelium (Figure 7C). Treatment with Ang II had no significant effect in p47phox−/− vessels (Figure 7D).
To confirm the role of O2− in basal and Ang II–induced ERK1/2 phosphorylation, aortic rings from WT and p47phox−/− mice were coincubated with Ang II±tiron before protein extraction for immunoblotting. Tiron significantly decreased Ang II–induced ERK1/2 phosphorylation in WT vessels and significantly reduced the basal ERK1/2 phosphorylation in p47phox−/− vessels (Figure 8).
Recent studies from several groups, including our own, have demonstrated an essential role for p47phox in the activation of endothelial and VSM NADPH oxidase by phorbol ester, thrombin, cytokines, growth factors, and Ang II.12–18 At the same time, it is clear that p47phox is not essential for “basal” NADPH-dependent O2− production in vascular cells.12–14 The present study provides several important new findings. (1) We demonstrate that in the absence of agonist stimulation, p47phox seems to have an inhibitory role, such that p47phox−/− vessels have significantly increased O2− production. (2) Increased basal O2− production in p47phox−/− vessels has potentially important functional consequences with respect to impaired endothelium-dependent vascular relaxation and the activation of MAPKs. (3) However, p47phox is also critically important for Ang II–induced NADPH oxidase activation, which results in impaired endothelium-dependent relaxation and MAPK activation. Taken together, these data indicate a complex dual role for p47phox in regulating vascular NADPH oxidase activity.
Perhaps the most interesting results of the present study are the findings relating to basal NADPH oxidase activity and vascular function in p47phox−/− mice. We previously reported that NADPH oxidase activity was slightly but significantly higher in cultured endothelial cells isolated from p47phox−/− mice compared with WT, as assessed by lucigenin chemiluminescence, dichlorohydrofluorescein fluorescence, or SOD-inhibitable cytochrome c reduction.14 In that study, acute depletion of endothelial p47phox by transfection of antisense cDNA resulted in an increased basal NADPH oxidase activity, whereas replacement of p47phox into knockout cells had the opposite effect. The present findings extend those results to intact vessel segments (as opposed to cultured cells) and furthermore demonstrate the functional correlate of increased basal ROS production. The impaired ACh-induced relaxation in p47phox−/− vessels was likely to be a result of increased basal O2− production, because it was normalized to the control level by either the O2− scavenger tiron or the SOD mimetic MnTMPyP. Furthermore, there was no change in the expression level of eNOS or of other ROS-generating systems such as xanthine oxidase. In addition to impaired ACh-induced relaxation, p47phox−/− vessels also demonstrated a significant preactivation of ERK1/2, p38MAPK, and JNK. The enhanced ERK1/2 phosphorylation was also likely to be a result of increased basal O2− production, because it was reduced by incubation with tiron. The lack of decrease in basal NADPH oxidase activity in p47phox−/− vessels is consistent with previous data indicating that, in the cell-free neutrophil system, NADPH oxidase can be activated in the total absence of p47phox.22 Conversely, the mechanism underlying increased basal NADPH oxidase activity in p47phox−/− vessels requires further investigation. Recently, we demonstrated that although p47phox is associated with cytochrome b558 in unstimulated endothelial cells, it seems to be largely unphosphorylated.23 However, after Ang II stimulation, p47phox becomes phosphorylated and then forms new oxidase complexes. In light of previous data that unphosphorylated p47phox can bind to p22phox in in vitro assays under certain conditions but that this binding supports only low O2− production,24 it is feasible that the binding of unphosphorylated p47phox to cytochrome b558 may inhibit vascular oxidase activity in the basal setting.
Ang II is the dominant effector of the renin-angiotensin system and is implicated in the pathogenesis of such disorders as hypertension and atherosclerosis, in which one major mechanism of its effects is through ROS generation.1,2 Ang II induces rapid O2− generation by activation of NADPH oxidase in several cells, including VSM, cardiac myocytes, endothelial cells, and fibroblasts.2–4,6,8,9,25 Among its effects are the induction of endothelial dysfunction as a consequence of O2−-mediated reduction in NO bioavailability. Whereas previous studies have established Ang II–induced NADPH oxidase activation as an important pathway for the development of endothelial dysfunction,3 the present study is the first to demonstrate a critical role for p47phox in this process. Thus, in p47phox−/− vessels, Ang II–induced endothelial dysfunction was abrogated, consistent with the failure to increase O2− production. The lack of effect of Ang II on NADPH oxidase activity of p47phox−/− vessels was not because of the higher basal activity in this group, because the maximal Ang II–stimulated activity in WT vessels was substantially higher than the basal level in p47phox−/− vessels.
Ang II activates multiple signaling pathways in a tissue-specific manner.1,2,26 Ang II signaling involves both ROS-dependent and -independent pathways, which influence downstream processes such as growth and hypertrophy. Among the potential targets for Ang II–stimulated ROS are the MAPKs. Indeed, previous studies have shown that all 4 main members of the MAPK family, namely ERK1/2, p38MAPK, JNK, and ERK5, can be activated in a redox-sensitive manner depending on the tissue studied.1,2,4,25–27 In rat VSM cells, Ang II–induced activation of p38MAPK but not ERK1/2 was shown to involve NADPH oxidase–derived ROS.4 In fibroblasts, however, both p38MAPK and ERK1/2 were activated by Ang II–stimulated ROS.25 Likewise, in endothelial cells, Ang II–stimulated activation of ERK1/2 involved NADPH oxidase–derived ROS.27 In the present study, we demonstrate a critical role for p47phox in the Ang II–induced activation of 3 major MAPK family members (ie, ERK1/2, p38MAPK, and JNK) in the mouse aorta, with the strongest activation being for ERK1/2. Immunostaining for phospho-ERK1/2 indicated activation throughout the vessel wall, although with particularly prominent labeling of the endothelium. Interestingly, increased phospho-ERK1/2 staining in p47phox−/− vessels in the basal state (ie, without Ang II) seemed to be specific to the endothelium, suggesting that endothelial cells may be a significant source of the basal ROS generation in these vessels. To assess the sites of in situ ROS generation, we also undertook DHE staining of aortic sections, which indicated the presence of ROS throughout the vessel wall. However, the inhibition of DHE fluorescence by catalase suggests that H2O2 was the ROS detected; because H2O2 may readily diffuse through cell membranes, the precise sites of production of the ROS remain speculative, although the ROS seem to be present throughout the vessel wall. Previous studies have shown that Ang II activates NADPH oxidase in VSM, endothelial cells, and fibroblasts.3–7,23
In summary, the present results indicate a dual role for p47phox in regulating vascular NADPH oxidase activity. In the absence of agonist stimulation, p47phox may serve as a negative regulator that maintains a low basal level of vascular O2− production. However, p47phox is also essential for Ang II–induced increases in NADPH oxidase activity. Potential therapeutic strategies focused on targeting p47phox as a means of specifically modulating NADPH oxidase–derived ROS production in disease states will need to take these dual effects into account.
This work was supported by British Heart Foundation (BHF) program grant RG/98008. Dr Wheatcroft holds a BHF Clinical PhD Studentship. Dr Kearney is a BHF Intermediate Fellow. Dr Shah holds the BHF Chair of Cardiology at King’s College London.
The supplementary figure can be found in the Data Supplement provided with the online-only version of this article at http://www.circulationaha.org.
Kim S, Iwao H. Molecular and cellular mechanisms of angiotensin II-mediated cardiovascular and renal diseases. Pharmacol Rev. 2000; 52: 11–34.
Griendling KK, Ushio-Fukai M. Reactive oxygen species as mediators of angiotensin II signaling. Cardiovasc Res. 2000; 91: 21–27.
Ushio-Fukai M, Alexander RW, Akers M, et al. p38 mitogen-activated protein kinase is a critical component of the redox-sensitive signaling pathways activated by angiotensin II. J Biol Chem. 1998; 273: 15022–15029.
Schieffer B, Luchtefeld M, Braun S, et al. Role of NAD(P)H oxidase in angiotensin II–induced JAK-STAT signaling and cytokine induction. Circ Res. 2000; 87: 1195–1201.
Griendling KK, Sorescu D, Ushio-Fukai M. NAD(P)H oxidase: role in cardiovascular biology and disease. Circ Res. 2000; 86: 494–501.
Wang HD, Xu S, Johns DG, et al. Role of NADPH oxidase in the vascular hypertrophic and oxidative stress response to angiotensin II in mice. Circ Res. 2001; 88: 947–953.
Bendall JK, Cave AC, Heymes C, et al. Pivotal role of a gp91phox-containing NADPH oxidase in angiotensin II–induced cardiac hypertrophy in mice. Circulation. 2002; 105: 293–296.
Touyz RM, Chen X, Tabet F, et al. Expression of a functionally active gp91phox-containing neutrophil-type NAD(P)H oxidase in smooth muscle cells from human resistance arteries: regulation by angiotensin II. Circ Res. 2002; 90: 1205–1213.
Babior BM. NADPH oxidase: an update. Blood. 1999; 93: 1464–1476.
Li J-M, Shah AM. Intracellular localization and pre-assembly of the NADPH oxidase complex in cultured endothelial cells. J Biol Chem. 2002; 277: 19952–19960.
Hsich E, Segal BH, Pagano PJ, et al. Vascular effects following homozygous disruption of p47phox: an essential component of NADPH oxidase. Circulation. 2000; 101: 1234–1236.
Lavigne MC, Malech HL, Holland SM, et al. Genetic demonstration of p47phox-dependent superoxide anion production in murine vascular smooth muscle cells. Circulation. 2001; 104: 79–84.
Li J-M, Mullen AM, Yun S, et al. Essential role of the NADPH oxidase subunit p47phox in endothelial cell superoxide production in response to phorbol ester and tumor necrosis factor-α. Circ Res. 2002; 90: 143–150.
Landmesser U, Cai H, Dikalov S, et al. Role of p47phox in vascular oxidative stress and hypertension caused by angiotensin II. Hypertension. 2002; 40: 511–515.
Rey FE, Cifuentes ME, Kiarash A, et al. Novel competitive inhibitor of NAD(P)H oxidase assembly attenuates vascular O2− and systolic blood pressure in mice. Circ Res. 2001; 89: 408–414.
Li J-M, Shah AM. Differential NADPH- versus NADH-dependent superoxide production by phagocyte-type endothelial cell NADPH oxidase. Cardiovasc Res. 2001; 52: 477–486.
Li J-M, Gall NP, Grieve DJ, et al. Activation of NADPH oxidase during progression of cardiac hypertrophy to failure. Hypertension. 2002; 40: 477–484.
Freeman JL, Lambeth JD. NADPH oxidase activity is independent of p47phox in vitro. J Biol Chem. 1996; 271: 22587–22592.
Li J-M, Shah AM. Mechanism of endothelial cell NADPH oxidase activation by angiotensin II: role of the p47phox subunit. J Biol Chem. 2003; 278: 12094–12100.
Shiose A, Sumimoto H. Arachidonic acid and phosphorylation synergistically induce a conformational change of p47phox to activate the phagocyte NADPH oxidase. J Biol Chem. 2000; 275: 13793–13801.
Sano M, Fukuda K, Sato T, et al. ERK and p38 MAPK, but not NF-κB, are critically involved in reactive oxygen species–mediated induction of IL-6 by angiotensin II in cardiac fibroblasts. Circ Res. 2001; 89: 661–669.