Heme oxygenase 1 attenuates interleukin-1β–induced cytosolic phospholipase A2 expression via a decrease in NADPH oxidase/reactive oxygen species/activator protein 1 activation in rheumatoid arthritis synovial fibroblasts
Reactive oxygen species (ROS) produced by cytokines induce the expression of inflammatory mediators in rheumatoid arthritis (RA). Heme oxygenase 1 (HO-1) exerts an antiinflammatory effect. The aim of this study was to examine the mechanisms underlying interleukin-1β (IL-1β)–induced cytosolic phospholipase A2 (cPLA2) expression through ROS generation as modulated by HO-1 in RA synovial fibroblasts (RASFs).
IL-1β–induced ROS generation was determined by flow cytometry. The involvement of MAPKs and NADPH oxidase (NOX)/ROS in IL-1β–induced cPLA2 expression was investigated using pharmacologic inhibitors and transfection with small interfering RNAs (siRNAs) and was analyzed by Western blotting and promoter assay. Overexpression of HO-1 was performed by transfection of RASFs with a recombinant adenovirus containing human HO-1 plasmid. SCID mice with inflammation caused by IL-1β were infected with adenovirus containing HO-1. Histologic characterization of joint inflammation and local expression of cPLA2 were evaluated after treatment.
IL-1β–induced cPLA2 expression was mediated through NOX activation/ROS production, which was attenuated by N-acetylcysteine (NAC; a scavenger of ROS), the inhibitors of NOX (diphenyleneiodonium chloride and apocynin), MEK-1/2 (U0126), and JNK-1/2 (SP600125), transfection with the respective siRNAs, and the overexpression of HO-1 in RASFs. IL-1β–induced cPLA2 expression was mediated through recruitment of activator protein 1 (AP-1) to the cPLA2 promoter region, which was attenuated by NAC and overexpression of HO-1. Furthermore, HO-1 overexpression inhibited IL-1β–mediated cPLA2 expression in SCID mice.
In RASFs, IL-1β induced cPLA2 expression via activation of p42/p44 MAPK and JNK-1/2, leading to p47phox phosphorylation, ROS production, and AP-1 activation. The induction of HO-1 exerted protective effects on the pathogenesis of RA.
Activated rheumatoid arthritis synovial fibroblasts (RASFs) play an initiating role in driving RA (1). The inflammatory response is regulated by cytokines and reactive oxygen species (ROS) produced by synoviocytes (2, 3). Levels of proinflammatory cytokines such as interleukin-1β (IL-1β) are elevated in the RA synovium and induce the expression of inflammatory genes and mediators such as the eicosanoids via various signaling pathways (3). The generation of eicosanoids such as prostaglandin E2 (PGE2) is initiated by phospholipase A2 (PLA2) and inducible cyclooxygenase 2 (COX-2) (4–7). Mice deficient in cytosolic PLA2α (cPLA2α) have been reported to reduce eicosanoid production in the presence of collagen-induced arthritis, indicating that cPLA2 plays a critical role in the pathogenesis of RA (8).
ROS have been implicated in inflammatory responses through the activation of transcription factors NF-κB and activator protein 1 (AP-1) (9, 10) and signaling pathways such as MAPKs, leading to the expression of proinflammatory genes in various tissues (11–13). NADPH oxidase (NOX) is a major source of ROS production under various pathologic conditions. The NOX complex is composed of 2 membrane-located subunits p22phox and gp91phox, cytosolic proteins p47phox and p67phox, and a GTPase Rac1, which assemble at membrane sites upon cell activation. Upon exposure to cytokines, synovial NOXs produce superoxide anions, which activate multiple signaling pathways, leading to the expression of inflammatory genes (3, 11). We have therefore suggested that excessive ROS generation appears to be one of the major mediators in the pathogenesis of RA.
Activation of NOX has been shown to stimulate the phosphorylation of p47phox by protein kinase C (PKC) or MAPKs, which initiates assembly of the cytoplasmic components and translocation to the membrane in various cell types (14–16). Production of ROS by hyperoxia was also shown to be mediated through activation of NOX and to be regulated by p42/p44 MAPK and p38 MAPK in human pulmonary artery endothelial cells (17). NOX-dependent ROS generation and activation of NF-κB and AP-1 have been shown to induce cPLA2 expression in human tracheal smooth muscle cells (HTSMCs) (12). It has also been shown that the H2O2-induced increase in the release of arachidonic acid and the production of PGE2 through the up-regulation of cPLA2 and COX-2 was mediated through Ca2+/PKC/MAPKs and epidermal growth factor receptor transactivation in mouse embryonic stem cells (18). However, very little is known about whether p47phox phosphorylation, NOX activation, and ROS generation are involved in IL-1β–induced cPLA2 expression in RASFs.
Heme oxygenase 1 (HO-1) is a stress-response protein involved in various inflammatory disorders (19, 20). Down-regulation of HO-1 is associated with increased inflammation and oxidative stress in various tissues (21). Induction of HO-1 attenuates the IL-1β–induced expression of matrix metalloproteinases 1 and 3 in osteoarthritic synoviocytes (22). Overexpression of HO-1 has also been shown to suppress the tumor necrosis factor α (TNFα)–mediated expression of inflammatory mediators through attenuation of ROS production in HTSMCs (20). However, it is still unknown whether the IL-1β–induced cPLA2 expression in RASFs is modulated by HO-1.
We therefore performed experiments in human RASFs and in mice to investigate whether HO-1 regulates IL-1β–induced cPLA2 expression. Our findings suggest that in RASFs, IL-1β stimulates the activation of p42/p44 MAPK and JNK-1/2, which leads to NOX-dependent ROS production, AP-1 activation, and cPLA2 gene expression. Overexpression of HO-1 could exert protective effects in the pathogenesis of RA via inhibition of these signaling components.
MATERIALS AND METHODS
Diphenyleneiodonium chloride (DPI), U0126, SP600125, and tanshinone IIA (TSIIA) were obtained from Biomol, N-acetylcysteine (NAC) from Sigma-Aldrich, and apocynin from ChromaDex. Polyclonal antibodies against p47phox, HO-1, Gsα subunit, lamin A, gp91phox, phospho–c-Jun, c-Fos, and β-actin, and monoclonal antibody against cPLA2 were from Santa Cruz Biotechnology. Anti-GAPDH antibody was obtained from Biogenesis. Phospho-p42/p44 MAPK, phospho–JNK-1/2, phosphoserine, and phosphotyrosine antibodies were from Cell Signaling Technology. 2′,7′-dichlorofluorescein diacetate (DCF-DA) was from Molecular Probes. A recombinant adenovirus containing human HO-1 gene (Adv-HO-1) was kindly provided by Dr. L. Y. Chau (Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan).
Isolation and culture of human synovial fibroblasts.
Synovial fibroblasts were obtained from 34 patients with RA who underwent knee or hip surgery. RASFs were isolated, cultured, and characterized as previously described (23). Experiments were performed using cells from passages 3 to 6.
SCID mice ages 4–6 weeks were purchased from National Laboratory Animal Centre in Taipei. Mice were maintained under conditions consistent with the Guidelines of the Animal Care Committee of Chang Gung University as well as the Guide for the Care and Use of Laboratory Animals of the National Research Council in the US. Mice were anesthetized and infected with 6.8 × 106 plaque-forming units of Adv-HO-1 or were injected with phosphate buffered saline (PBS). After 24 hours, mice were given an intraarticular injection of IL-1β (30 μg/kg of body weight) and were euthanized after another 24 hours (23, 24).
To examine the cellular expression and localization of cPLA2 and HO-1 proteins, immunohistochemical staining was performed on serial sections of the mouse ankle joints or the RA synovial tissues as previously described (23). Briefly, ankle joints from the mice were fixed in 10% formalin. After decalcification in 5% formic acid, the specimens were processed by paraffin embedding. Tissue sections (5 μm) were incubated with anti-cPLA2 or anti–HO-1 antibodies for 1 hour at 37°C, and then with a horseradish peroxidase (HRP)–conjugated anti-rabbit IgG antibody for 1 hour at room temperature. A separate section was incubated with antivimentin antibody for localization and identification of synovial fibroblasts. Antibody binding was detected with diaminobenzidine and observed under an optical microscope.
Preparation of cell extracts and Western blot analysis.
RASFs were incubated with IL-1β at 37°C for the time intervals indicated below. Samples were analyzed by Western blotting as previously described (23). Membranes were incubated for 24 hours with primary antibody and then for 1 hour with HRP-conjugated anti-mouse or anti-rabbit secondary antibody. The immunoreactive bands were detected with enhanced chemiluminescence reagents.
Transient transfection with short hairpin RNAs (shRNAs) and small interfering RNAs (siRNAs).
Plasmids encoding shRNAs of ERK-1, JNK-1, JNK-2, and the vector pTOPO-U6 were kindly provided by Dr. C. P. Tseng (Chang Gung University). SMARTpool RNA duplexes corresponding to human p47phox, gp91phox, c-Jun, c-Fos, and scrambled #2 siRNA were from Dharmacon. Transient transfection of 20 nM shRNA or 100 nM siRNA was performed using Lipofectamine 2000 reagent (Invitrogen), as previously described (23).
Total RNA extraction and real-time quantitative polymerase chain reaction (PCR) analysis.
RNA was extracted using TRIzol reagent, and first-strand complementary DNA synthesis was performed with 1 μg of total RNA using Superscript II reverse transcriptase (Invitrogen). The levels of cPLA2α and GAPDH messenger RNA (mRNA) expression were determined by real-time PCR, as previously described (23).
Measurement of intracellular ROS accumulation.
Production of ROS in RASFs was determined by the DCF-DA fluorescence method as previously described (20). The fluorescence intensity of the human RASFs was also analyzed using a FACScan flow cytometer (BD Biosciences) at 495 nm excitation and 529 nm emission for DCF-DA.
Preparation and analysis of cell fractions.
Growth-arrested RASFs were incubated with 30 ng/ml of IL-1β for the time intervals indicated below. The membrane, cytosolic, and nuclear fractions were prepared as previously described (20, 23). Samples of these fractions were analyzed by Western blotting using the respective antibodies.
Cell lysates (1 mg) were incubated for 24 hours at 4°C with 2 μg of anti-p47phox or anti-gp91phox antibody, and then 10 μl of 50% protein A–agarose was added and mixed for 24 hours at 4°C. The immunoprecipitates were washed thrice with lysis buffer without Triton X-100; 5× Laemmli buffer was then added, and lysates were analyzed by Western blotting using the respective antibodies.
Measurement of cPLA2 and AP-1 promoter activities.
A cPLA2 promoter/luciferase plasmid (pGL-cPLA2-Luc) was constructed as previously described (23). Introduction of a double point-mutation into the AP-1–binding site (AP-1 domain; TGATTAA to TTCTTAA) to generate pGL-cPLA2-mtAP-1 was performed using the (forward) primer 5′-CAGCACTCATGGAATTTAGGACTTCTTAATTTACATATT-3′. The underlined nucleotides indicate the positions of substituted bases. The mutant construct was cloned to the pGL3 basic vector containing the luciferase reporter system. All plasmids were prepared using Qiagen plasmid DNA preparation kits. The pAP-1-Luc (Clontech) or cPLA2 promoter reporter construct activity was determined as previously described (23). Firefly promoter luciferase activities were standardized for β-galactosidase activity.
Chromatin immunoprecipitation (ChIP) assay.
ChIP assay was performed as previously described (23). Soluble chromatin was immunoprecipitated using an anti–c-Jun, anti–c-Fos, or anti-IgG antibody. Following washing and elution, precipitates were heated overnight at 65°C to reverse cross-linking of DNA and protein. DNA fragments were purified by phenol–chloroform extraction and ethanol precipitation. The purified DNA was subjected to PCR amplification using primers specific for the region containing the AP-1–binding site present in the cPLA2 promoter: 5′-CAAGTAGCAATTTCAGACGC-3′ (sense) and 5′-GTCTTATCCTGAGTAGGAGG-3′ (antisense). PCR fragments were analyzed on 2% agarose in 1× Tris–acetate–EDTA gel containing ethidium bromide.
All reported data are representative of at least 3 independent experiments and comparisons of ≥3 populations were made using a GraphPad Prism Program. Data are expressed as the mean ± SEM and were analyzed by one-way analysis of variance followed by Tukey's post hoc test. P values less than 0.05 and P values less than 0.01 were considered significant.
ROS are involved in IL-1β–induced cPLA2 expression.
ROS have been shown to activate many signal transduction pathways associated with innate immunity and inflammation (25), which may be due to an increase in cPLA2 expression (12,26,27). Although IL-1β has been known to induce cPLA2 expression in RASFs (28), it is still not known whether ROS production leads to cPLA2 expression. We first examined cPLA2 expression in RA synovial tissues and in normal mouse ankle joints. As shown in Figure 1A, cPLA2-expressing cells were densely concentrated in the lining layers of RA synovial tissues. A majority of the cPLA2-expressing cells also showed staining with vimentin, a fibroblast-specific antibody. In contrast, cPLA2-expressing cells were rarely detected in synovia obtained from healthy SCID mice.
HO-1 has been shown to be expressed in synovial tissues from patients with RA (29). We also found HO-1 in both RA synovial tissues and healthy SCID mouse joints. However, there was no significant difference between the findings in the RA patients and those in the healthy control mouse joints (Figure 1A).
We also found that IL-1β induced a significant increase in ROS levels in RASFs within 5 minutes, peaked within 120 minutes, and declined by 180 minutes. This response was attenuated by NAC, a scavenger of free radicals (Figure 1B). Pretreatment with NAC also attenuated IL-1β–induced cPLA2 expression in a concentration-dependent manner (Figure 1C), suggesting that IL-1β–induced cPLA2 expression is mediated through ROS generation in RASFs.
IL-1β–induced cPLA2 expression is mediated via p42/p44 MAPK–dependent and JNK-1/2–dependent ROS generation.
To examine whether MAPKs are involved in IL-1β–induced ROS production and cPLA2 expression, we treated RASFs with inhibitors of MEK-1/2 (U0126), p38 MAPK (SB202190), and JNK-1/2 (SP600125). IL-1β–induced ROS generation and cPLA2 expression were inhibited by U0126 and SP600125, but not SB202190 (Figures 2A and B), indicating that p42/p44 MAPK and JNK-1/2 are essential for ROS generation and cPLA2 expression.
To further ensure that the IL-1β–induced response is mediated via p42/p44 MAPK and JNK-1/2, transfection with shRNA for ERK-1, JNK-2, or JNK-1 down-regulated the expression of their respective proteins and subsequently attenuated IL-1β–induced cPLA2 expression in RASFs (Figure 2C). IL-1β also stimulated p42/p44 MAPK and JNK-1/2 phosphorylation, which was inhibited by U0126 and SP600125, respectively, at the time points examined (Figure 2D). Since activation of p42/p44 MAPK and JNK-1/2 and the generation of ROS were necessary for IL-1β–induced cPLA2 expression in RASFs, it would be important to determine whether ROS generation led to p42/p44 MAPK and JNK-1/2 phosphorylation. Pretreatment with NAC had no effect on p42/p44 MAPK or JNK-1/2 phosphorylation (Figure 2D). These results indicated that IL-1β–induced cPLA2 expression is mediated through p42/p44 MAPK– and JNK-1/2–dependent ROS generation in RASFs.
NOX/ROS generation is involved in IL-1β–induced cPLA2 expression.
One of the major sources of ROS is the NOX in RASFs (3, 30). Pretreatment of RASFs with the NOX inhibitors DPI and apocynin abrogated the IL-1β–induced cPLA2 expression and ROS generation in a concentration-dependent manner (Figures 3A and B), but had no effect on p42/p44 MAPK or JNK-1/2 phosphorylation (data not shown). In addition, IL-1β stimulated the membrane translocation of p47phox in a time-dependent manner; this was attenuated by apocynin, DPI, U0126, or SP600125 (Figure 3C). We further investigated whether IL-1β stimulates serine or tyrosine phosphorylation of p47phox. As illustrated in Figure 3D, IL-1β stimulated in a time-dependent manner the phosphorylation on tyrosine and serine residues of p47phox in the cell lysates immunoprecipitated using an anti-p47phox antibody; this was attenuated by DPI, apocynin, U0126, or SP600125. In addition, the protein level of NOX-2 was increased in p47phox-immunoprecipitated complexes, which was also inhibited by DPI, apocynin, U0126, or SP600125.
Further investigation of the roles of p42/p44 MAPK and JNK-1/2 in IL-1β–induced NOX activity showed that IL-1β induced a significant increase in NOX activity, with a maximal response within 60 minutes, and that this was inhibited by U0126, SP600125, DPI, or apocynin (data not shown). Moreover, transfection with NOX-2 or p47phox siRNA down-regulated the expression of their respective proteins and inhibited IL-1β–induced cPLA2 expression (Figure 3E). These results suggested that IL-1β–induced NOX activation and ROS generation is mediated through p42/p44 MAPK– and/or JNK-1/2–stimulated phosphorylation of p47phox and leads to cPLA2 expression in RASFs.
IL-1β induces cPLA2 expression through NOX/ROS-dependent activation of AP-1.
To determine whether IL-1β–induced cPLA2 expression is mediated through transcription factor AP-1, we pretreated RASFs with TSIIA and found that this inhibited IL-1β–induced cPLA2 expression (Figure 4A). To ensure that AP-1 is required for the expression of cPLA2 induced by IL-1β, we examined transfection of RASFs with c-Jun or c-Fos siRNA, which down-regulated the expression of c-Jun or c-Fos protein, respectively, and reduced the cPLA2 expression induced by IL-1β (Figure 4B). Moreover, IL-1β stimulated the translocation of phospho–c-Jun into the nucleus, which was attenuated by treatment with U0126, SP600125, DPI, apocynin, or NAC (Figures 4C and D). In contrast, the IL-1β–induced accumulation of c-Fos in the nucleus was attenuated only with U0126 or NAC (Figure 4D). RASFs were transfected with either p47phox or NOX-2 siRNA to verify the role of NOX-2 in the activation of c-Jun stimulated by IL-1β. We found that the expression of p47phox or NOX-2, respectively, was suppressed and IL-1β–induced c-Jun phosphorylation was attenuated upon transfection (Figures 4E and F).
AP-1 activation was further evaluated by gene luciferase activity and ChIP assays. IL-1β stimulated AP-1 luciferase activity in a time-dependent manner (data not shown) and was inhibited by U0126, SP600125, apocynin, DPI, NAC, or TSIIA (Figure 4G). To determine whether AP-1 is required for IL-1β–induced cPLA2 promoter activity via binding to the AP-1–binding element on the cPLA2 promoter region, we constructed a cPLA2 promoter with a double-point mutation at the AP-1–binding site, mt-AP-1-cPLA2 (Figure 4H, top). IL-1β–stimulated cPLA2 promoter activity was significantly blocked in RASFs transfected with the mt-AP-1-cPLA2 reporter construct (Figure 4H, bottom), indicating that the AP-1–binding element is required for IL-1β–induced cPLA2 promoter activity.
We next investigated whether c-Jun and c-Fos bind to the AP-1 element on the cPLA2 promoter after IL-1β stimulation. The recruitment of c-Jun and c-Fos to the cPLA2 promoter was investigated using a ChIP assay. As shown in Figure 4I, the binding of c-Jun and c-Fos to the AP-1 element of the cPLA2 promoter occurred within 2 hours and was sustained for 4 hours after IL-1β stimulation. The binding of c-Jun to the AP-1 element was attenuated by U0126, SP600125, apocynin, DPI, and NAC treatment. In contrast, the recruitment of c-Fos to the cPLA2 promoter was inhibited only by U0126 and NAC. Taken together, these results suggested that IL-1β–induced cPLA2 expression is mediated through p42/p44 MAPK– and/or JNK-1/2-NOX/ROS–dependent activation of AP-1 in RASFs.
NOX/ROS generation is required for cPLA2 transcription activity and PGE2 release.
The possible regulation of cPLA2 gene transcription activity by these kinases was also investigated in RASFs. As shown in Figures 5A and B, pretreatment with U0126, SP600125, DPI, apocynin, NAC, or TSIIA significantly attenuated IL-1β–induced cPLA2 mRNA expression and promoter activity as well as PGE2 production, as assessed by real-time PCR, promoter luciferase assay, and PGE2 enzyme immunoassay. These results suggested that IL-1β–induced NOX-dependent ROS generation and AP-1 activation participate in cPLA2 expression and PGE2 production in RASFs.
Overexpression of HO-1 attenuates IL-1β–induced cPLA2 expression.
We hypothesized that HO-1 might also retard the inflammatory responses in RA. To this end, HO-1 was overexpressed in RASFs infected with Adv-HO-1 but not adenovirus (data not shown). HO-2 was not changed in RASFs infected with Adv-HO-1. Overexpression of HO-1, moreover, attenuated IL-1β–induced cPLA2 protein, mRNA expression, ROS generation, and NOX activity (Figures 6A–C). IL-1β–stimulated membrane translocation of p47phox (Figure 6D) and phosphorylation of tyrosine and serine residues (Figure 6E) were attenuated in RASFs infected with Adv-HO-1. We also found that the protein levels of NOX-2 and p47phox were increased in p47phox- and NOX-2–immunoprecipitated complexes, respectively, which were inhibited in RASFs infected with Adv-HO-1 (Figures 6E and F). Overexpression of HO-1 also attenuated the IL-1β–stimulated accumulation of phosphorylated c-Jun and c-Fos in the nucleus and AP-1 luciferase activity (Figures 6G and H).
To confirm these results in vivo, mice were given an intraarticular injection of PBS or Adv-HO-1, which was followed 24 hours later with an injection of IL-1β (30 μg/kg of body weight). Joints were harvested and examined after another 24 hours. Immunostaining in the articular joints showed that the number of cPLA2-expressing cells was significantly higher in IL-1β–treated mice than in PBS-treated mice (Figure 6I, parts c and i). Ad-HO-1 infection was found in synovial fibroblasts near the injection site in articular joints (Figure 6I, part m). The cPLA2-immunopositive cells were markedly increased in IL-1β–treated mice and were suppressed in Ad-HO-1–infected mice. These results confirmed that overexpression of HO-1 prevents IL-1β–induced cPLA2 expression in vivo in SCID mice at a level similar to that in vitro in RASFs, suggesting that overexpression of HO-1 prevents IL-1β–induced cPLA2 expression and that it is mediated through down-regulation of NOX-2/ROS and AP-1 activation in RASFs.
In RA, inflammation of the synovium largely contributes to the development of the disease symptoms and the tissue degradation. Cytosolic PLA2 is considered to play a pivotal role in eicosanoid production, which is implicated in the pathogenesis of collagen-induced arthritis in mice (8). Moreover, cPLA2 activity gradually increases and shows a correlation with severity of arthritis (4). Although these findings have suggested that cPLA2 may be a causative factor in inflammation and tissue injury in collagen-induced arthritis, the molecular mechanisms underlying the IL-1β–induced expression of cPLA2 in RASFs are still unclear. In this study, we found that IL-1β–induced cPLA2 expression was mediated through activation of p42/p44 MAPK– and JNK-1/2–dependent NOX/ROS generation, leading to activation of AP-1 in RASFs (Figure 6J). Moreover, HO-1 exerts antiinflammatory and antioxidative effects on various cell types (20, 31, 32). We found that overexpression of HO-1 attenuated IL-1β–induced cPLA2 expression, which was mediated through the down-regulation of p47phox-dependent ROS production, leading to abrogation of AP-1 activation and cPLA2 expression in RASFs.
ROS are implicated in the pathogenesis of a wide variety of human inflammatory diseases, such as RA (11). In human RASFs, IL-1β induces the formation of intracellular oxidants, which lead to augmented expression of genes for matrix-degrading enzymes (33). One of the major sources of ROS are the nonphagocytic NOX isoforms, which also play a role in the regulation of signaling cascades in various cell types, including endothelial cells and cardiomyocytes (34). Two different NOX homologs (NOX-2 and NOX-4) were recently identified in RA cells. In synoviocytes, NOX-4 could be responsible for a continuous superoxide production and NOX-2 could be responsible for superoxide production by cytokines (3). We also confirmed in the present study that NOX-2 and NOX-4 are expressed in RASFs, as determined by reverse transcription–PCR (data not shown). Moreover, the expression of cPLA2 was shown to be due to NOX activation and ROS generation by cigarette smoke extract in HTSMCs (12). We therefore investigated the roles of the NOX/ROS cascade associated with cPLA2 expression by IL-1β in RASFs. The use of NOX inhibitors or siRNAs significantly abolished the IL-1β–induced cPLA2 expression, suggesting that NOX-2 is essential for these responses. The results are consistent with those of a previous study showing that IL-1β induces ROS production in RA cells, which is inhibited by NOX inhibitors (3). These results suggested that NOX-2–dependent ROS generation may play a key role in IL-1β–induced cPLA2 expression in RASFs.
Production of superoxide anions by NOX is accompanied by extensive phosphorylation of p47phox, which is crucial for translocation of the cytosolic components and assembly of the active NOX (14, 16). Activation of NOX-2 is initiated by the assembly of p47phox with gp91phox. In vascular smooth muscle cells, p47phox is phosphorylated on serine and tyrosine residues and translocated from the cytosol to the membrane by angiotensin II (35). TNFα-induced activation of Src kinase in HTSMCs results in tyrosine phosphorylation and translocation of p47phox with increased ROS generation (20). It has been shown that phosphorylation of p47phox on Ser345 is directly related to TNFα-induced ROS production in neutrophils isolated from the synovial fluid of RA patients (36). IL-1β–induced p47phox phosphorylation has also been detected in RA synoviocytes (3). Although agonist-induced p47phox phosphorylation has been extensively investigated in various cell types, the mechanisms of phosphorylation of p47phox and activation of NOX in RASFs are unclear. Our data from the present study provide evidence for the in vitro phosphorylation of p47phox by IL-1β and interaction between NOX-2 and p47phox in the regulation of IL-1β–induced ROS generation. These results suggested that IL-1β induces NOX-2/ROS generation and cPLA2 expression via phosphorylation of p47phox subunit. Moreover, other NOX-2 subunits (p67phox or Rac1) may also be involved in superoxide production induced by cytokines.
MAPKs have been shown to regulate cPLA2 expression in various cell types (37, 38). Our results revealed that IL-1β–stimulated phosphorylation of p42/p44 MAPK and JNK-1/2 was inhibited by their respective inhibitors. Moreover, IL-1β–induced cPLA2 expression was also significantly inhibited by these pharmacologic inhibitors and shRNAs, indicating that p42/p44 MAPK and JNK-1/2 participate in cPLA2 induction by IL-1β in RASFs. These results are consistent with previous reports from studies of HTSMCs (38). Moreover, p42/p44 MAPK and p38 MAPK play a role in the hyperoxia-induced NOX/ROS cascade in human pulmonary artery endothelial cells (17). In vascular smooth muscle cells, angiotensin II activated MAPKs (p42/p44 MAPK, p38 MAPK, and JNK-1/2) in association with NOX-mediated generation of ROS (39). TNFα-stimulated JNK-1/2 phosphorylation led to ROS generation, potentiating the necrosis of fibroblasts (40). Our results indicated that IL-1β induced ROS generation via p42/p44 MAPK and JNK-1/2 in RASFs.
Phosphorylation of p47phox occurs at serine residues, suggesting the potential involvement of different protein kinases, such as PKC or MAPKs, in these responses (39). It has been shown that p42/p44 MAPK activation involves the granulocyte–macrophage colony-stimulating factor–induced phosphorylation of p47phox at Ser345, while p38 MAPK modulates the TNFα-induced phosphorylation of the same site on p47phox in neutrophils isolated from the synovial fluid of RA patients (3). Moreover, we demonstrated a link between the MAPKs, p47phox phosphorylation, and NOX/ROS generation in RASFs in response to IL-1β. These results further suggest that NOX-dependent ROS generation is involved in IL-1β–induced cPLA2 expression via MAPK-mediated p47phox phosphorylation in RASFs.
The transcription factor AP-1 is typically composed of c-Jun and c-Fos proteins and is implicated in the up-regulation of several inflammatory genes, such as cPLA2 (39). Thus, the role of AP-1 in IL-1β–induced cPLA2 expression was revealed by using a selective AP-1 inhibitor and by transfection with siRNA for c-Jun or c-Fos in RASFs. Our results suggested that AP-1 is essential for cPLA2 expression induced by IL-1β. ROS has also been shown to activate NF-κB and AP-1 in HTSMCs (12). Our data showed that IL-1β induced the accumulation of phospho–c-Jun in the nucleus, which was reduced by blockade of the NOX-2/ROS cascade. We further demonstrated that IL-1β stimulated the activity of AP-1 and increased the binding of c-Jun and c-Fos to the AP-1 element within the cPLA2 promoter. In contrast, the IL-1β–induced accumulation of c-Fos in the nucleus and the binding of c-Fos to the cPLA2 promoter were attenuated only by U0126 and NAC. We further confirmed by using an AP-1-mutated cPLA2 construct that the AP-1–binding site (–498 to –492) within the cPLA2 promoter is required for IL-1β–induced cPLA2 transcription activity. These results indicated that IL-1β–induced cPLA2 expression might be mediated through the p42/p44 MAPK, JNK-1/2, and NOX/ROS cascades, leading to activation of AP-1 in RASFs.
Up-regulation of HO-1 serves as an adaptive response to protect cells from stress. The protective effects of HO-1 could be related to generation of the antioxidant molecules and are of particular interest in the context of inflammatory responses (41). Recent studies have revealed that HO-1 induction in animals protects them from the development of arthritis (42). It has been reported that induction of HO-1 by cobalt protoporphyrin IX in the presence of IL-1β decreases the expression of matrix metalloproteinase 1/3 in osteoarthritic synoviocytes (22). HO-1 induction was also shown to attenuate the expression of inflammatory cytokines and COX-2 in RASFs (29). Moreover, overexpression of HO-1 suppressed TNFα-induced vascular cell adhesion molecule 1 and intercellular adhesion molecule 1 expression by reducing NOX/ROS generation (20).
Our data also showed that HO-1 exerted antiinflammatory effects on IL-1β–induced cPLA2 expression in RASFs. Overexpression of HO-1 attenuated p47phox phosphorylation, NOX activity, and ROS generation. HO-1 induction by Ad-HO-1 also inhibited the formation of p47phox and NOX-2 complexes by IL-1β in RASFs. Several transcription factors have been shown to be redox-sensitive, including AP-1 (2). Our observations confirmed that IL-1β–induced accumulation of phospho–c-June and c-Fos in the nucleus and AP-1 promoter activity are reduced by HO-1 induction. Thus, AP-1 appears to be one of the important targets for the action of HO-1 in RASFs. Here we have provided evidence of the protective role of HO-1 induction against IL-1β–induced cPLA2 expression by the inhibition of NOX/ROS generation in RASFs.
In summary, IL-1β induced ROS generation through NOX activation and, in turn, initiated the activation of AP-1. The activation of NOX in response to IL-1β was probably dependent on p42/p44 MAPK– and JNK-1/2–mediated phosphorylation of p47phox. Activated AP-1 was recruited to the promoter regions of cPLA2, which led to increased cPLA2 promoter activity and the expression of cPLA2 mRNA and protein in RASFs. Moreover, overexpression of HO-1 suppressed p47phox phosphorylation, NOX activation, ROS generation, and AP-1 activation, resulting in the inhibition of cPLA2 expression in RASFs. Therefore, cPLA2 might be an important ROS-sensitive gene that is expressed in RASFs. Inhibition of p47phox phosphorylation by HO-1 induction might be a mechanism that contributes to the tight regulation of NOX activity and thereby diminishes redox-sensitive proteins in inflammation.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Yang had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Chi, Y.-W. Chen, Hsiao, Y.-L. Chen, Yang.
Acquisition of data. Chi, Y.-W. Chen, Hsiao, Y.-L. Chen, Yang.
Analysis and interpretation of data. Chi, Y.-W. Chen, Hsiao, Y.-L. Chen, Yang.
The authors appreciate Dr. C. P. Tseng (Department of Medical Biotechnology and Laboratory Science, Chang Gung University, Taiwan) for providing shRNAs for this study and Dr. H. L. Hsieh (Chang Gung Institute of Technology, Taiwan) for proofreading the manuscript.