Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Cells of the respiratory tract are constantly exposed to reactive oxygen intermediates (ROIs) generated by resident or inflammatory phagocytes or by inhaled environmental agents, such as airborne transition metals (1, 2). Vanadium compounds are potential constituents of ambient particulate matter, especially in areas of oil-fired, electricity generating plants (3). In addition, occupational exposure to vanadium compounds is common in the petrochemical, mining, and steel industries and results in toxic effects to the respiratory system (3). Human and laboratory animal exposures to metals, including vanadium, induce proinflammatory responses in the lung that are characterized by elevated production of inflammatory mediators such as interleukin (IL)-8 and tumor necrosis factor (TNF)- (4, 5). Pretreatment with antioxidants attenuates the inflammatory mediator production evoked by exposure of respiratory epithelial cells to metallic compounds in vitro (4, 6), suggesting that ROIs play an important role in the signaling pathways mediating proinflammatory responses in the lung.

Transcription of a number of inflammatory mediators, including IL-8, is partially regulated at the transcriptional level by nuclear factor (NF)-B (7), a transcription factor that regulates the expression of a variety of genes whose products mediate immune and inflammatory responses or regulate cell-cycle progression (8, 9). In most resting cells, NF-B resides in the cytoplasm as an inactive complex, consisting at a minimum of a DNA-binding dimer (NF-B) and an inhibitory subunit (IB), which masks the nuclear translocation sequence of the transcription factor. The mobilization of NF-B induced by proinflammatory cytokines, such as TNF-, is preceded by the phosphorylation of IB on two N-terminal serine residues. This phosphorylation results in the proteolytic degradation of IB and release of the NF-B dimer, which translocates into the nucleus, binds to B-specific response elements, and modulates transcription. Phosphorylation of IB in response to inflammatory cytokines is mediated by a large multisubunit IB kinase complex (IKK) that is activated by an upstream kinase designated NF-B-inducing kinase (NIK) (9, 10). NIK is a mitogen-activated protein kinase (MAPK) kinase kinase (MAPKKK) (11). MAPKKKs are a highly divergent family of protein kinases that are components of the MAPK pathways that mediate changes in gene expression in response to extracellular stimuli (12, 13). Other MAPKKKs, specifically MEKK1, MEKK2, and MEKK3, can also activate IKK and thus induce IB serine phosphorylation (14, 15).

In addition to mediating the mobilization of NF-B, MAPK pathways have also been implicated in regulating the transcription-activating potential of nuclear NF-B induced by TNF-. In particular, activation of the p38 MAPK affects B-dependent transcription without affecting mobilization and DNA binding of the transcription factor (16). In TNF--treated L929 cells, activation of p38 enhanced the transcription-promoting activity of the C-terminal transactivation domains of the p65 NF-B subunit without affecting the levels of NF-B DNA binding (17). Hence, B-dependent transcription induced by TNF- depends on p38 MAPK-dependent regulatory processes modulating the transactivation potential of nuclear NF-B, as well as the levels of NF-B in the nucleus.

Although oxidative stimuli have been shown to activate p38 MAPK (19), there is no evidence that metals or metal-induced oxidative stress induce B-dependent transcription in a p38 MAPK-dependent manner. In this study we demonstrate that vanadyl sulfate (V^IV)-induced B-dependent transcription in human airway epithelial cells is inhibited by catalase overexpression and also by selective inhibition of the p38 MAPK pathway. In addition, we demonstrate that the activation of p38 MAPK after stimulation with V^IV depends on V^IV-induced oxidative stress and mediates B-dependent transcription without affecting nuclear translocation or DNA binding of NF-B. Together, these findings demonstrate that V^IV-induced oxidative stress activates collateral signaling pathways that converge downstream to cooperatively modulate the transcriptional activity of NF-B.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture and In Vitro Exposure

Primary normal human bronchial epithelial (NHBE) cells and the human BEAS-2B bronchoepithelial cell line were cultured as described previously (6, 20). V^IV (Sigma Chemical, St. Louis, MO) or TNF- (R&D Systems, Minneapolis, MN) were diluted in bronchial epithelial growth medium (NHBEM) or keratinocyte growth medium (KGM) (BEAS-2B) (both from Clonetics, San Diego, CA) before addition to the cell culture. In some experiments SB203580 (Calbiochem, San Diego, CA) or the respective dimethyl sulfoxide (DMSO) vehicle control were added 30 min before V^IV or TNF- challenge.

Analysis of ROI Generation

NHBE cells grown to confluence in 96-well plates were incubated with 10 µM 2',7'-dichlorodihydrofluorescein diacetate (DCF-DA) or 5 µM dihydrorhodamine 123 (both from Molecular Probes, Eugene, OR) for 30 min, which allowed the dye to enter the cell. After several washings to remove the extracellular dye, V^IV and TNF- were added to their final concentrations. Fluorescence signals, indicating intracellular formation of ROIs, were measured using a fluorescence plate reader (HTS 700; Perkin-Elmer, Norwalk, CT) with excitation/absorption settings at 485/535 nm and a sensitivity gain of 60. Fluorescence was recorded after the indicated incubation times with the respective stimuli and expressed as relative fluorescence units.

Separation of Cytoplasmic and Nuclear Fractions

After washing NHBE cells with ice-cold phosphate-buffered saline, 200 µl of cold cytoplasmic extraction buffer (CEB) (10 mM Tris-HCl [pH 7.9], 60 mM KCl, 1 mM ethylenediaminetetraacetic acid [EDTA], and 1 mM dithiothreitol [DTT]) with protease inhibitors (PIs) (1 mM Pefabloc, 50 µg/ml antipain, 1 µg/ml leupeptin, 1 µg/ml pepstatin, 40 µg/ml bestatin, 3 µg/ml E-64, and 100 µg/ml chymostatin; all purchased from Boehringer Mannheim, Indianapolis, IN) was added to each well. Using a rubber policeman, cells were scraped up and transferred into a microcentrifuge tube. The cells were allowed to swell on ice for 15 min, then Nonidet P-40 (NP-40) (Sigma) was added to a final concentration of 0.1% and the tube was vortexed for 10 s. Nuclei were pelleted by centrifugation at 15,000 × g for 30 s. The supernatant, containing the cytoplasmic fraction, was mixed with 0.25 vol of 4× loading buffer (62.5 mM Tris-HCl [pH 6.8], 10% glycerol, 2% sodium dodecyl sulfate [SDS], 0.7 M -mercaptoethanol, and 0.05% bromophenol blue), denatured at 95°C for 10 min, and stored at 70°C for immunoblot analysis. Protein content of a small aliquot of the cytoplasmic fraction was determined using the DC Bradford assay (Bio-Rad, Richmond, CA). The nuclei were washed with CEB/PI and centrifuged again at 15,000 × g for 30 s. The supernatant was aspirated and the nuclei were incubated for 10 min on ice in nuclear extraction buffer (20 mM Tris-HCl [pH 8.0], 400 mM NaCl, 1.5 mM MgCl2, 1.5 mM EDTA, 1 mM DTT, and 25% glycerol) with PI. After brief centrifugation, the supernatants, containing the nuclear fraction, were either stored at 80°C until analysis by electrophoretic mobility shift assay (EMSA), or denatured and stored for immunoblot analysis as described earlier.

EMSA

Oligonucleotide probes containing the NF-B consensus sequence from the major histocompatibility complex (MHC) class II promoter (GGCTGGGGATTCCCCATCT) were synthesized on an Applied Biosystems Model 391 DNA synthesizer (Perkin-Elmer). The probes were labeled by incubating 15 U T4 polynucleotide kinase (New England Biolabs, Beverly, MA), 100 ng double-stranded probe, and 100 µCi adenosine 5'-[-³²P]triphosphate (ICN, Irvine, CA) at 37°C for 30 min. Unincorporated ³²P was removed using a desalting column (Nuc Trap; Stratagene, San Diego, CA). DNA-protein binding reactions were performed for 10 min at room temperature in a mixture containing 4 µg of nuclear extract, 1 µl labeled probe, 10 µl running buffer (10 mM Tris-HCl [pH 7.5], 50 mM NaCl, 2 mM EDTA, 1 mM DTT, and 5% glycerol), and 2 µg poly dI/dC (Boehringer Mannheim). Samples were separated by electrophoresis through 4.5% nondenaturing polyacrylamide gels containing 0.5× TBE. Gels were dried and analyzed by PhosphorImaging (Molecular Dynamics, Sunnyvale, CA).

Promoter Reporter Constructs, Transfection, and Promoter-Reporter Assay

A B-dependent promoter-reporter construct, pNF-B-luc (Stratagene), was used. It was composed of a 5× tandem repeat of the NF-B response element of the mouse immunoglobulin  gene intronic enhancer cloned upstream of a TATA box and a firefly luciferase complementary DNA. A constitutively active simian virus 40 promoter--galactosidase (-gal) construct (pSV--gal) (Promega, Madison, WI) was used to adjust for well-to-well variation in cell number and transfection efficiency.

BEAS cells grown to 60 to 80% confluence in 24-well tissue culture dishes were cotransfected with 250 ng of the pNF-B-luc and 25 ng of pSV--gal using 1.5 µg of DOTAP transfection reagent (Boehringer Mannheim). In some experiments cells were cotransfected with expression vectors carrying a dominant negative mutant form of p38 (p38-AF), MKK6 (MKK6-A) (21) (both kind gifts from Dr. S. Ludwig), or a nonrecombinant control plasmid, pZeoSV2- (Invitrogen, San Diego, CA). At 48 h after transfection, cultures were treated for 4 h with 50 µM V^IV or 10 ng/ml TNF-. Luciferase and -gal activity was determined using the Dual Light reporter gene assay system (Perkin-Elmer) and an AutoLumat LB953 luminometer (Berthold Analytical Instruments, Nashua, NH). Promoter activity was estimated as specific luciferase activity (luciferase counts per unit -gal counts) and expressed as fold-induction over the respective media control.

Infection with Adenovirus

NHBE or BEAS cells grown to about 80% confluence were infected with AdCl (22) or AdSOD1 (23) at a multiplicity of infection (MOI) of 50 and 100 plaque-forming units/cell, respectively, for 3 to 4 h. As a control, cells were infected with Ad5CMV3. The infection mixture was aspirated and the cells were incubated for another 48 h before stimulation with V^IV or TNF-.

Immunoblot Analysis

For analysis of phospho-p38 and phosphotyrosine levels, whole-cell lysates were prepared as described earlier (24). Protein samples (50 µg), either whole-cell lysates or cytoplasmic protein fractions, were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) on 14% Tris-glycine gels, followed by immunoblotting using specific antibodies to IB (1:1,000; Santa Cruz Biotechnology, Santa Cruz, CA), catalase and Cu,Zn-superoxide dismutase (Cu,Zn-SOD) (1:1,000; both from Biodesign, Kennebunk, ME) for 1 h at room temperature, or phospho-p38 (1:1,000; New England Biolabs) overnight at 4°C. Antigen-antibody complexes were stained with horseradish peroxidase (HRP)-conjugated antibody (1:2,000; Bio-Rad) and enhanced chemiluminescence reagent (ECL) and ECL film (both from Amersham, Arlington Heights, IL). For analysis of phosphotyrosine levels, whole-cell lysates were separated by SDS-PAGE on 4 to 15% gradient Tris-glycine gels, followed by immunoblotting using an HRP-conjugated antiphosphotyrosine antibody (PY-99, 1:1000; Santa Cruz Biotechnology) overnight at 4°C. Protein tyrosine phosphate bands were detected using ECL reagents and film as described earlier. Immunoblot films were digitized, and the optical densities of specific antigen-antibody complexes were quantified using Kodak 1D Image Analysis Software (Eastman Kodak Company, Rochester, NY).

Immunohistochemical Staining

BEAS cells grown to 70 to 80% confluence in 35-mm glass-bottom dishes (MatTek, Ashland, MA) were transfected with 200 ng pcDNA3-p38-AF using 1 µg DOTAP transfection reagent. At 24 h after transfection, cells were stimulated with either 50 µM V^IV or 10 ng/ml TNF- for 60 min. The cells were fixed at room temperature in 4% paraformaldehyde, followed by treatment with 0.4% NP-40 in CEB buffer at 4°C, and final fixation in 4% paraformaldehyde at room temperature for 20 min. As previously described (21), the p38-AF mutant form was cloned into the pcDNA3 expression vector and tagged with the FLAG epitope, which allows for detection of the p38-AF transgene product without interference from endogenous p38. After blocking nonspecific binding with 2% bovine serum albumin, cells were incubated with 1 µg/ml anti-p65 (Santa Cruz Biotechnology) and anti-FLAG M2 (Eastman Kodak) antibodies at 4°C overnight. The cells were washed and costained with secondary antibodies against mouse (FLAG) and rabbit (p65) epitopes (ALEXA-488 and ALEXA-594, respectively; Molecular Probes) at room temperature for 1 h. After washing off excess antibody, glass bottoms were mounted onto microscope slides and viewed under fluorescent microscope (Axiovert 10; Zeiss, Thornwood, NY). Cells staining for FLAG, which identified cells transfected with pcDNA3-p38-AF and thus expressing the p38-AF transgene product, were identified with the fluorescein filter and analyzed for nuclear translocation of p65 with the rhodamine filter.

Statistical Analysis

Data analysis was done using a two-tailed Student's t test (as discussed later in Figure 5B) or a two-way analysis of variance followed by the Newman-Keul's post hoc test for multigroup analysis.

View larger version (18K): [in this window] [in a new window]   Figure 5.   Inhibition of the p38 MAPK pathway decreases NF-B-dependent promoter reporter activity in VIV-treated or TNF--treated cells. BEAS cultures were transiently cotransfected with pSV--gal and pNF-B-luc. At 48 h after transfection, cells were stimulated with VIV or TNF- for 4 h. Specific luciferase activity in culture lysates was determined using -gal activity as a normalizing factor, and data are expressed as normalized fold induction over the respective media controls. (A) Cotransfection with dominant negative mutant forms of p38 MAPK (pcDNA3-p38-AF) or MKK6 (pcDNA3-MKK6-A) inhibit NF-B-dependent transcription in VIV-treated and TNF--treated cells. Cotransfection with a nonrecombinant control vector (pZeoSV2-) had no effect on VIV-induced or TNF-- induced NF-B-dependent promoter reporter activity. (B) Pretreatment with the p38 MAPK inhibitor SB203580 (20 µM) inhibited VIV-induced NF-B promoter reporter activity. *Significantly different from control; P < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

V^IV, but Not TNF-, Induces Oxidative Stress in NHBE Cells

There is evidence that treatment with both V^IV and TNF- results in intracellular ROI formation (25, 26). We therefore tested whether V^IV or TNF- was capable of generating intracellular ROI in NHBE cells by loading cells with DCF-DA or dihydrorhodamine 123, cell-permeant dyes whose oxidation to DCF or rhodamine 123 was monitored fluorometrically. Figure 1A shows that V^IV generated a strong fluorescent signal in DCF-DA-treated cells, as early as 15 min after stimulation. Interestingly, although in cells loaded with DCF-DA treatment with V^IV increased fluorescence in a dose-dependent manner, lower concentrations of V^IV induced a greater fluorescent signal in dihydrorhodamine-treated cells, as shown in Figure 1B. However, stimulation with TNF- did not induce any significant changes in fluorescence emitted by either DCF-DA-treated or dihydrorhodamine-treated cells (Figure 1C) up to 24 h after stimulation (Figure 1A). These data suggested that exposure to V^IV, but not to TNF-, generated oxidative stress in NHBE cells.

View larger version (25K): [in this window] [in a new window]   Figure 1.   Treatment with VIV, but not TNF-, increases oxidative stress in NHBE cells. (A) Cells were loaded with 10 µM DCF-DA for 30 min before stimulation with either 50 µM VIV or 10 ng/ml TNF-. Fluorescence was measured at 0.25, 0.5, 1, 2, and 24 h after stimulation using a fluorescence plate reader. Comparison of fluorescent signals in cells loaded with 10 µM DCF-DA or 5 µM dihydrorhodamine 123 and subsequently stimulated with increasing concentrations of (B) VIV or (C) TNF-. *Significantly different from control (DCF-DA); # significantly different from control (dihydrorhodamine 123); P < 0.05. Values represent means ± standard error of the mean (SEM).

Overexpression of Catalase Inhibited V^IV-Induced, but Not TNF--Induced, NF-B Mobilization and B-Dependent Transcription

Several studies have demonstrated that ROI generation induced by a variety of oxidative stimuli, such as H₂O₂, ultraviolet (UV) radiation, reperfusion injury, and ozone, results in enhanced nuclear NF-B DNA-binding activity (27). Consequently, we analyzed NF-B DNA-binding activities by EMSA of NHBE nuclear extracts. Two NF-B DNA-binding complexes were observed, one whose abundance was increased by treatment with V^IV for 1 h (compare lanes 1 and 2 in Figure 2A). Supershift analysis using antibodies to the p65 and p50 subunits of NF-B demonstrated that p65 and p50 were components of the inducible activity (compare lanes 2, 4, and 6, Figure 2A), whereas the noninducible activity contained p50 but not p65 (compare lanes 1, 3, and 5, Figure 2A).

View larger version (28K): [in this window] [in a new window]   Figure 2.   Overexpression of catalase, but not Cu,Zn-SOD, inhibits VIV-induced NF-B DNA binding. NHBE cells treated with 50 µM VIV for 1 h were analyzed for DNA-binding activities using EMSA. (A) Treatment with VIV induced DNA binding to the MHC class II NF-B consensus sequence. Addition of antibodies against p50 and p65 identified the DNA-binding complex induced after VIV treatment as a p65/p50 heterodimer. (B) NHBE cells were infected with adenoviral expression vectors for catalase (AdCl) and Cu,Zn-SOD (AdSOD1) at the respective MOI. Western blot analysis shows marked overexpression of the respective antioxidant enzyme. The effects of infection with a control vector (AdCMV) or expression vectors for catalase (AdCl) and Cu,Zn-AdSOD1 (AdSOD1) on NF-B DNA-binding activity was analyzed in NHBE cells treated with (C) VIV or (D) TNF- for 1 h. The arrows indicate the respective NF-B complex.

To determine whether the V^IV-induced NF-B DNA-binding activity observed by EMSA (Figure 2A) was dependent upon the generation of intracellular ROIs, we overexpressed catalase (AdCl) and Cu,Zn-SOD (AdSOD1) using adenoviral expression vectors. Western blot analyses (Figure 2B), confirmed marked overexpression of the respective antioxidant enzymes in NHBE cells, as had been demonstrated in earlier studies using these expression vectors (22, 23, 30). Adenoviral infection with a nonrecombinant expression vector (Ad5CMV3) did not affect the V^IV-induced NF-B DNA-binding activity (compare lanes 2 in Figures 2A and 2C). In contrast, Figure 2C shows that overexpression of catalase inhibited V^IV-induced NF-B DNA-binding activity (compare lanes 2 and 4, Figure 2C), whereas overexpression of Cu,Zn-SOD had no effect (compare lanes 2 and 6, Figure 2C). In TNF--treated NHBE cells we observed an NF-B DNA-binding pattern similar to that seen after stimulation with vanadium; however, the NF-B DNA-binding activity induced by TNF- was insensitive to overexpression of either catalase or Cu,Zn-SOD (Figure 2D; compare lanes 2, 4, and 6). These data show that both V^IV and TNF- induced the mobilization of NF-B from the cytoplasm into the nucleus, but suggest that this response was mediated by different mechanisms that could be distinguished by the overexpression of catalase.

The mobilization of NF-B is preceded by the inactivation of IBs, a family of protein inhibitors of NF-B that hold NF-B inactive in the cytoplasm. In most cases, IBs are inactivated by stimulus-dependent phosphorylation and proteolytic degradation. We therefore monitored breakdown of IB in the cytoplasm of V^IV-treated or TNF-- treated NHBE cells and determined the effect of catalase and Cu,Zn-SOD overexpression. Treatment with V^IV for 1 h induced IB breakdown, which was inhibited by overexpression of catalase, but not Cu,Zn-SOD (Figures 3A and 3C). In contrast, TNF--induced IB breakdown was unaffected by overexpression of either catalase or Cu,Zn-SOD (Figures 3B and 3C). These data confirm that V^IV and TNF- activated the signaling cascade culminating in IB breakdown and nuclear translocation of NF-B by distinct mechanisms, and indicate that catalase intervened upstream of IB degradation in V^IV-treated cells.

View larger version (19K): [in this window] [in a new window]   Figure 3.   Overexpression of catalase blocks IB breakdown in NHBE cells treated with VIV, but not TNF-. Cytoplasmic extracts from NHBE cells overexpressing catalase (AdCl) or Cu, Zn-SOD (AdSOD1) were separated by SDS-PAGE and immunoblotted using a specific anti-IB antibody. Breakdown of IB was monitored in cells stimulated with (A) VIV or (B) TNF- for 1 h. (C) Densitometric analysis of three separate experiments. *Significantly different from control; P < 0.05. Values are means ± SEM and expressed as percent of media control values.

To determine whether the V^IV-induced and TNF-- induced mobilization of NF-B results in enhanced B- dependent transcription, we transiently transfected an NF-B-dependent luciferase promoter-reporter construct into BEAS-2B cells, a human bronchial epithelial cell line, which resemble NHBE cells in their responses to V^IV and TNF- treatment. The effect of catalase overexpression on B-dependent transcription was determined by infecting cultures 24 h after transfection with AdCl or Ad5CMV3. To activate NF-B-dependent transcription, BEAS-2B cells were treated with either V^IV or TNF- for 1 h and further incubated for 3 h. Figure 4 illustrates that stimulation with either TNF- or V^IV increased B-dependent transcription. Overexpression of catalase significantly inhibited B-dependent transcription in cells treated with V^IV but not with TNF-. These data indicate that both the mobilization and transcriptional activity of NF-B in V^IV-stimulated cells is catalase-sensitive.

View larger version (24K): [in this window] [in a new window]   Figure 4.   Overexpression of catalase inhibits VIV-induced transcriptional activity of NF-B. BEAS cultures were transiently cotransfected with pSV--gal and pNF-B-luc and infected 24 h later with adenoviral vectors for catalase (AdCl) or a nonrecombinant control vector (AdCMV). At 48 h after transfection, cells were stimulated with 50 µM VIV or 10 ng/ml TNF- for 1 h and further incubated for 3 h. Specific luciferase activity in culture lysates was determined using -gal activity as a normalizing factor and data are expressed as normalized fold induction over the respective media control. *Significantly different from control; P < 0.05.

p38 MAPK Pathway Mediates TNF--Induced and V^IV-Induced B-Dependent Transcription, but Not NF-B Mobilization

Previous evidence indicated that in TNF--treated cells nuclear translocation of NF-B is necessary but not sufficient for maximal B-dependent transcription (16). These studies have suggested that TNF--induced B- dependent transcription is modulated by downstream elements of the p38 MAPK pathway. To examine the role of the p38 MAPK pathway in the V^IV-induced B-dependent transcription, we cotransfected BEAS-2B cells with the NF-B promoter-reporter construct and with expression vectors for dominant negative mutant forms of either p38 (p38-AF) or its upstream kinase MKK6 (MKK6-A). Treatment with either TNF- or V^IV increased B-dependent transcription in cells cotransfected with a nonrecombinant control vector (pZeoSV2-), whereas overexpression of either p38-AF or MKK6-A significantly decreased both TNF--induced and V^IV-induced B-dependent promoter- reporter activity (Figure 5). Likewise, pretreatment with the p38 MAPK inhibitor SB203580 significantly decreased V^IV-induced B-dependent promoter-reporter activity (Figure 5B). However, double immunofluorescent localization of p65 and the p38-AF transgene product (as visualized by its FLAG tag) in V^IV-treated and TNF-- treated cells revealed that overexpression of a dominant negative mutant form of the p38 MAPK did not inhibit nuclear translocation of the NF-B subunit p65, as shown in Figure 6. Similarly, treatment with the p38 MAP kinase inhibitor SB203580 had no apparent effect on either V^IV-induced or TNF--induced IB breakdown (Figures 7A and 7B) or increase in NF-B DNA-binding activity (Figure 7C). Together, these data indicate that in both TNF-- treated and V^IV-treated bronchial epithelial cells, the p38 MAPK pathway modulates B-dependent transcription without affecting NF-B mobilization.

View larger version (123K): [in this window] [in a new window]   Figure 6.   Transfection with a dominant negative mutant form of the p38 MAPK (pcDNA3-p38-AF) has no effect on VIV-induced or TNF--induced nuclear translocation of p65. BEAS cells were transfected with pcDNA-p38-AF and subsequently stimulated with either 50 µM VIV or 10 ng/ml TNF- for 60 min. The p38-AF transgene product is FLAG-tagged and was identified with a mouse monoclonal antibody against FLAG and detected with an ALEXA-488 conjugated antimouse antibody. Nuclear translocation of p65 was identified with a rabbit antibody against p65 and detected with an ALEXA-594 conjugated antirabbit antibody. Arrows point to identical cells staining for both the p38-AF transgene product and p65.

View larger version (29K): [in this window] [in a new window]   Figure 7.   Inhibition of p38 MAPK has no effect on NF-B nuclear translocation in NHBE cells treated with VIV or TNF-. NHBE cells pretreated with 20 µM SB203580 or the respective DMSO vehicle control were analyzed for IB breakdown and NF-B DNA binding induced by treatment with VIV or TNF- for 1 h. (A) Cytoplasmic extracts from NHBE cells were separated by SDS-PAGE and immunoblotted using a specific anti-IB antibody. Prereatment with SB203580 had no effect on VIV-induced or TNF-- induced IB breakdown. (B) Densitometric analysis of three separate experiments. Values are means ± SEM and expressed as percent of media control values. Filled bars, vehicle; checkered bars, SB203580. (C) NHBE cells stimulated with 50 µM VIV or 10 ng/ ml TNF- were analyzed for DNA-binding activities using EMSA. Pretreatment with the p38 MAPK inhibitor SB203580 had no effect on NF-B DNA-binding activities in cells treated with VIV or TNF-.

Overexpression of Catalase Reduces V^IV-Induced, but Not TNF--Induced p38 MAPK Activation

To examine whether ROIs mediate TNF--induced or V^IV-induced p38 MAPK activation, we measured dually phosphorylated p38 levels, an indication of p38 activation, in cell lysates of NHBE cells that had been infected with AdCl or AdSOD1. Figures 8A and 8B show that both V^IV and TNF- increased phospho-p38 levels in NHBE cells infected with a nonrecombinant control vector (Ad5CMV3), indicating activation of the p38 MAPK pathway. Overexpression of catalase decreased V^IV-induced p38 MAPK activity, but had no effect in TNF--treated cells. Overexpression of Cu,Zn-SOD did not affect phospho-p38 levels in cells treated with V^IV or TNF-. This indicated that, similar to the mobilization of NF-B, activation of p38 upon stimulation with V^IV was sensitive to overexpression of catalase.

View larger version (29K): [in this window] [in a new window]   Figure 8.   Overexpression of catalase blocked p38 MAPK activity in VIV-treated cells. (A) Cell lysates from NHBE cells overexpressing catalase (AdCl) or Cu,Zn-SOD (AdSOD1) that had been treated with either 50 µM VIV or 10 ng/ml TNF- for 20 min were separated by SDS-PAGE and immunoblotted using a specific antibody against dually phosphorylated p38 MAPK (P-p38). (B) Densitometric analysis of three separate experiments. *Significantly different from control; P < 0.05. Values are means ± SEM and expressed as fold induction over media control values.

Overexpression of Catalase Does Not Inhibit V^IV-Induced Tyrosine Phosphorylation

Vanadium compounds are potent inhibitors of protein tyrosine phosphatase (PTPase) and activators of tyrosine kinases. Many effects induced by vanadium compounds may be mediated by general dysregulation of protein tyrosine phosphorylation. In addition, H₂O₂ alone or in combination with vanadium compounds forming peroxovanadium are also PTPase inhibitors (31, 32). To investigate the role of oxidative stress in V^IV-induced tyrosine phosphorylation, we determined total phosphotyrosine levels in V^IV-treated cells overexpressing catalase or Cu,Zn-SOD. Figure 9 shows that neither catalase nor Cu,Zn-SOD inhibited V^IV-induced tyrosine phosphorylation, suggesting that V^IV- induced NF-B translocation and p38 MAPK activation are not due to dysregulation of protein tyrosine phosphorylation.

View larger version (43K): [in this window] [in a new window]   Figure 9.   Overexpression of catalase or Cu,Zn-SOD does not affect VIV-induced tyrosine phosphorylation. Cell lysates from NHBE cells overexpressing catalase (AdCl) or Cu,Zn-SOD (AdSOD1) that had been treated with 100 µM VIV for 2 h were separated by SDS-PAGE and immunoblotted using a specific antibody against phosphotyrosine. The blot shown is representative of two independent experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This study demonstrates that V^IV-induced B-dependent transcription depends upon NF-B translocation and p38 MAPK activation. In this regard, V^IV-derived oxidative stress is similar to TNF-, which others (16) have shown and we have confirmed here also depends upon NF-B translocation and p38 MAPK activation. However, the activation of these pathways by V^IV-derived oxidative stress was catalase-sensitive, whereas their activation by TNF- was not. That two distinct stimuli, such as receptor-ligand interaction and metal-induced oxidative stress, activate B-dependent transcription in a p38 MAPK-dependent manner suggests that p38 activation may be essential for the activation of B-dependent transcription regardless of the stimulus. This notion is reinforced by the recent observation that p38 MAPK activation is essential for the activation of B-dependent transcription associated with oncogenic transformation of NIH3T3 fibroblasts (33).

Many physiologic responses evoked by exposure to transition metals are assumed to be due to the metal's ability to catalyze ROI-forming reactions (34). We demonstrate in the present study that in NHBE cells V^IV-induced mobilization of NF-B, activation of the p38 MAPK pathway, and B-dependent transcription were all sensitive to catalase but not Cu,Zn-SOD overexpression. Although, vanadyl (V^IV) compounds spontaneously autooxidize in aqueous solutions to vanadate (V^V), generating superoxide anions in the process (26), the absence of inhibition of these processes in the face of overexpression of Cu,Zn-SOD suggests that superoxide does not directly stimulate these pathways in bronchial epithelial cells. Spontaneous or SOD-catalyzed dismutation of superoxide would, of course, subsequently produce H₂O₂. Thus, our results suggest that formation of peroxides, but not superoxides, plays a dominant role in V^IV-induced signaling.

Exposures to H₂O₂ or ROI-forming stimuli activate MAPK cascades, particularly the stress-responsive MAPK Jun amino-terminal kinase (JNK) and p38 MAPK (19). Treatment of endothelial cells with angiotensin II increased intracellular H₂O₂ levels, and angiotensin II-induced p38 MAPK activity was blocked in cells stably transfected with catalase (35). Similarly, ischemia/reperfusion, which is known to induce generation of ROIs, activated p38 and JNK MAPK (36). It remains to be established what redox-sensitive molecules mediate ROI-dependent signaling cascades. Potential targets for ROIs could be cysteine residues in the catalytic domains of PTPases, which must remain in the reduced form in order to be enzymatically active (31). Both ROIs and vanadium compounds, alone or in combination, have been shown to be strong inducers of tyrosine phosphorylation, by inhibiting PTPases. Moreover, it has been suggested that ROI-dependent inhibition of PTPase and subsequent increased tyrosine phosphorylation precede oxidative stress-induced activation of MAPK (37). However, our results demonstrate that catalase overexpression does not inhibit the increases in phosphotyrosine levels mediated by exposure to V^IV. Thus, it seems unlikely that changes in intracellular levels of H₂O₂ activate the stress-responsive p38 MAPK pathway through the inhibition of PTPases in V^IV-treated cells.

This and earlier studies have shown that in V^IV-treated or TNF--treated cells, B-dependent transcription was p38-dependent, whereas NF-B mobilization was insensitive to treatment with the p38 inhibitor SB203580 (16). The mechanism by which p38 MAPK activation leads to enhanced NF-B transcriptional activity without affecting translocation of the transcription factor is still unclear. There are at least two phosphorylation sites on p65 that affect its transcription-activating potential (38, 39), and possibly more (40, 41). However, p38 MAPK does not directly phosphorylate either the p50 or p65 subunits of NF-B (42). Inhibition of p38 MAPK activity does not block TNF--induced phosphorylation of Serine 529 of p65, an event essential for TNF--induced B-dependent transcription (39). It remains to be determined whether a p38 MAPK-activated kinase phosphorylates p65 at a transcriptional regulatory site other than Serine 529. Alternatively, p38 MAPK or a downstream kinase could phosphorylate a coactivator, such as CBP/p300, which has been implicated in NF-B-driven gene transcription (43). CBP/p300, which constitutively associates with RNA polymerase II, has been shown to interact with the transactivation domain containing C-terminus of p65 (43). Phosphorylation of p65 or the coactivator itself may enhance this interaction and thus the ability of NF-B to interact with the basal transcriptional machinery.

The signaling event through which V^IV-derived peroxides trigger IB degradation remains to be identified. Inflammatory cytokines such as TNF- activate IKK, which includes two kinases, IKK and IKK, that phosphorylate IB on Serines 32 and 36 (9, 10). Although IKK is essential for the NF-B mobilization induced by inflammatory cytokines, IKK is dispensible (44). An IB mutant in which Serines 32 and 36 are converted to Alanines acts as a "super-repressor" of inflammatory cytokine- induced NF-B activation (47). This suppression appears to be specific for IKK/-dependent signaling, because the IB(S32,36A) mutant is rapidly degraded in cells exposed to short-wavelength UV light, which activates NF-B by a mechanism independent of IKK/ (48, 49). The IB(S32,36A) mutant blocks V^IV-induced NF-B nuclear translocation (Jaspers, unpublished observations) as well as B-dependent transcription in bronchial epithelial cells (21), suggesting that V^IV-derived peroxides also trigger IB degradation through activation of IKK/. However, additional studies are necessary to confirm that V^IV-derived peroxides activate the IKK/ complex and, if so, to determine whether IKK and IKK are differentially affected by peroxides.

The data presented here show that V^IV-treated and TNF--treated bronchial epithelial cells display similar signaling patterns culminating in enhanced B-dependent transcription, but that signaling is initiated through distinct pathways upstream of IB kinase and MKK6. We speculate that V^IV and TNF- activate a distinct signaling cascade component(s) responsive to either oxidative stress or protein-protein interactions. Potential candidates for such signal-transducing molecules are human STE20-like kinases, a family of serine-threonine kinases, some of which are selectively oxidant stress-responsive (50), whereas others are activated by ligand-receptor interactions (51). Members of this family of kinases have been shown to activate MEKKs (52, 53), which in turn can activate IKK/ (14) and MKK6 (54), resulting in the simultaneous mobilization of NF-B and p38 MAPK activation. Although an Ste20-like kinase, receptor interacting protein, has been shown to be essential for TNF--induced B-dependent transcription (51, 55), an oxidative stress-sensitive Ste20-like kinase that activates the p38 MAPK pathway has not been identified.

Although we confirmed that separate p38 MAPK- independent and -dependent pathways mediate nuclear translocation and transcriptional activity of NF-B, respectively, it remains to be established whether inputs from other signaling pathways are also required to fully activate B-dependent transcription.