Role of Neutrophil NADPH Oxidase in the Mechanism of Tumor Necrosis Factor-α-induced NF-κB Activation and Intercellular Adhesion Molecule-1 Expression in Endothelial Cells

In this study, we explored a novel function of polymorphonuclear neutrophils (PMN) NAD(P)H oxidase in the mechanism of tumor necrosis factor-α (TNFα)-induced NF-κB activation and intercellular adhesion molecule-1 (ICAM-1) expression in endothelial cells. Studies were made in mice lacking the p47 phox subunit of NAD(P)H oxidase as well as in cultured mouse lung vascular endothelial cells (MLVEC) from these mice. In response to TNFα challenge, NF-κB activation and ICAM-1 expression were significantly attenuated in lungs of p47 phox −/− mice as compared with wild-type (WT) mice. The attenuated NF-κB activation in p47 phox −/− mice was secondary to inhibition of NIK activity and subsequent IκBα degradation. Induction of neutropenia using anti-PMN serum prevented the initial TNFα-induced NF-κB activation and ICAM-1 expression in WT mice, indicating the involvement of PMN NAD(P)H oxidase in signaling these responses. Moreover, the responses were restored upon repletion with PMN obtained from WT mice but not with PMN from p47 phox −/−mice. These findings were recapitulated in MLVEC co-cultured with PMN, suggesting that NF-κB activation and resultant ICAM-1 expression in endothelial cells occurred secondary to oxidants generated by the PMN NAD(P)H oxidase complex. The functional relevance of the PMN NAD(P)H oxidase in mediating TNFα-induced ICAM-1-dependent endothelial adhesivity was evident by markedly reduced adhesion of p47 phox −/− PMN in co-culture experiments. Thus, oxidant signaling by the PMN NAD(P)H oxidase complex is an important determinant of TNFα-induced NF-κB activation and ICAM-1 expression in endothelial cells.

In this study, we explored a novel function of polymorphonuclear neutrophils (PMN) NAD(P)H oxidase in the mechanism of tumor necrosis factor-␣ (TNF␣)-induced NF-B activation and intercellular adhesion molecule-1 (ICAM-1) expression in endothelial cells. Studies were made in mice lacking the p47 phox subunit of NAD(P)H oxidase as well as in cultured mouse lung vascular endothelial cells (MLVEC) from these mice. In response to TNF␣ challenge, NF-B activation and ICAM-1 expression were significantly attenuated in lungs of p47 phox؊/؊ mice as compared with wild-type (WT) mice. The attenuated NF-B activation in p47 phox؊/؊ mice was secondary to inhibition of NIK activity and subsequent IB␣ degradation. Induction of neutropenia using anti-PMN serum prevented the initial TNF␣-induced NF-B activation and ICAM-1 expression in WT mice, indicating the involvement of PMN NAD(P)H oxidase in signaling these responses. Moreover, the responses were restored upon repletion with PMN obtained from WT mice but not with PMN from p47 phox؊/؊ mice. These findings were recapitulated in MLVEC co-cultured with PMN, suggesting that NF-B activation and resultant ICAM-1 expression in endothelial cells occurred secondary to oxidants generated by the PMN NAD(P)H oxidase complex. The functional relevance of the PMN NAD(P)H oxidase in mediating TNF␣-induced ICAM-1-dependent endothelial adhesivity was evident by markedly reduced adhesion of p47 phox؊/؊ PMN in co-culture experiments. Thus, oxidant signaling by the PMN NAD(P)H oxidase complex is an important determinant of TNF␣-induced NF-B activation and ICAM-1 expression in endothelial cells.
The pro-inflammatory cytokine TNF␣, 1 released during sepsis, promotes adhesion of neutrophil (PMN) to the endothelium by inducing the expression of intercellular adhesion molecule-1 (ICAM-1), a counter receptor for the leukocyte ␤ 2 -integrins LFA-1 and Mac-1 (CD11a/CD18 and CD 11b/CD18) (1)(2)(3)(4). The interaction of ICAM-1 with CD11/CD18 integrins enables PMN to adhere firmly to the vascular endothelium and thereby migrate across the microvascular barrier. Studies have shown that the transcription factor NF-B is the key regulator of ICAM-1 gene expression following TNF␣ challenge of endothelial cells (4). Signals mediating TNF␣-induced NF-B activation are initiated by the engagement of TNF receptor type I at the plasma membrane and then relayed through specific TNFR-associated proteins. TNF␣-associated death domaincontaining protein (TRADD) is an adaptor protein that interacts with TNFR type I and is required for TNF␣-mediated induction of NF-B (5). TRADD interacts in turn with two other adaptor proteins, TNFR-associated factor 2 (TRAF2) and receptor-interacting protein (RIP), and forms a complex required for NF-B activation (6,7). NF-B-inducing kinase (NIK) physically interacts with TRAF2 (8), and further, it activates IB kinases (IKKs), which in turn phosphorylate IB proteins, a family of inhibitory proteins that sequester NF-B as an inactive complex in the cytoplasm. Phosphorylation of IB triggers the rapid ubiquitination and subsequent degradation of this inhibitor in proteasome complex (9). The liberated NF-B migrates to the nucleus where it binds to cognate B enhancer elements and activates target genes such as ICAM-1 (10).
Generation of oxidants in endothelial cells serves an important signaling function in mediating TNF␣-induced NF-B activation and ICAM-1 expression and, thereby, promoting the stable ICAM-1-dependent endothelial adhesivity and firm PMN adhesion (11)(12)(13)(14)(15). Although TNF␣ has been shown to induce oxidative burst in PMN (16), it remains unclear whether the oxidants thus released can contribute to TNF␣-induced NF-B activation and ICAM-1 expression in the endothelium. The primary source of oxidative burst in PMN is NAD(P)H oxidase, a highly regulated membrane-bound enzyme complex, which catalyzes the production of superoxide by the one-electron reduction of oxygen using NAD(P)H as the electron donor. The core enzyme consists of five subunits: p40 phox , p47 phox , p67 phox , p22 phox , and gp91 phox . In the basal state, p40 phox , p47 phox , and p67 phox exist in the cytosol as a complex, whereas p22 phox and gp91 phox are located in membranes of secretory vesicles and specific granules of PMN, where they occur as a heterodimeric flavohemoprotein known as cytochrome b 558 . Also, two low molecular weight GTP-binding proteins, Rap 1A and rac1/2, are involved in the activation of the NAD(P)H oxidase. Upon stimulation, the cytosolic component p47 phox is phosphorylated, and the entire cytosolic complex migrates to the membrane where it associates with cytochrome b 558 to assemble the active oxidase (17). In the present study, we addressed the role of PMN NAD(P)H oxidase complex in me-diating TNF␣-induced NF-B activation and ICAM-1 expression using mice genetically deficient in p47 phox or gp91 phox subunit of NAD(P)H oxidase and co-cultures involving mouse PMN and vascular endothelial cells (MLVEC) from these mice. We demonstrate that the functional impairment of neutrophil NAD(P)H oxidase and thereby of oxidant generation significantly delayed the TNF␣-induced NF-B activation and ICAM-1 expression in both lungs and MLVEC of p47 phoxϪ/Ϫ or gp91 phoxϪ/Ϫ mice. Our data establish that the effects of NADPH oxidase inhibition were secondary to inhibition of NIK activation and the subsequent IB␣ degradation.

EXPERIMENTAL PROCEDURES
Mice-Breeder stocks for p47 phox knockout mice were obtained from Dr. Steven Holland (Laboratory of Host Defense, National Institutes of Health) (18). Mice deficient in gp91 phox were obtained from Dr. Mary Dinauer (University of Indiana School of Medicine, Indianapolis, IN) (19). Wild-type (WT) mice of similar genetic background (C57BL/6) were purchased from The Jackson Laboratory. All animals were maintained under specific pathogen-free conditions in a barrier facility. Mice were 9 -10 weeks of age at the time of experiments.
Experimental Protocols-Animals were given TNF␣ (Promega, Madison, WI) 5000 units/10 g body weight or saline by intraperitoneal (i.p.) injection at time equals 0, and whole lung tissue was harvested at the times indicated. In some experiments neutropenia was induced by PMN depletion using rabbit anti-mouse neutrophil serum (ANS) (Intercell Technologies, Hopewell, NJ). Approximately 16 h before injection of TNF␣, 150 l of ANS or control antibody (rabbit anti-mouse IgG; Sigma Chemical Co., St. Louis, MO) was administered i.p. to mice. Blood for determination of absolute PMN counts was obtained by cardiac puncture from animals that had received ANS or control serum. PMN counts were determined manually on slides of Wright's stain prepared with Cytospin (Shandon Inc., Pittsburgh, PA). On the average, PMN in ANS group was decreased to 0.08 Ϯ 0.02% of total white blood cells compared with 22.2 Ϯ 1.9% in the control group. There were no statistically significant differences in the number of peripheral lymphocytes, monocytes, atypical lymphocytes, or eosinophils between the ANS and control groups.
To determine the role of PMN NAD(P)H oxidase in the mechanism of TNF␣-induced ICAM-1 expression, PMN repletion experiments were performed in the neutropenia mice with PMN (ϳ1 ϫ 10 6 cells) isolated from multiple WT or p47 phoxϪ/Ϫ mice. PMN isolation was carried out as described previously (20) using the NIM.2 PMN isolation medium (Cardinal, Santa Fe, NM).
MLVEC Isolation and Characterization-MLVEC was isolated using a previously described method (21) but modified in our laboratory. Briefly, mice anesthetized with 3% halothane and heparin (50 units) were injected into the jugular vein as an anticoagulant. The abdominal cavity was opened, and the pulmonary artery was cannulated. Krebs-Henseleit solution supplemented with bovine serum albumin (5 g/100 ml) was infused to remove blood from lungs. Peripheral lung tissue slices were prepared, washed, and suspended in Hanks' balanced salt solution (HBSS). The tissue slices were minced and digested with collagenase A (1.6 mg/ml in HBSS) for 20 min at 37°C in a shaking water bath. The released cells were centrifuged and suspended in 5 ml of suspension buffer (Ca 2ϩ and Mg 2ϩ free phosphate-buffered saline containing 0.5 g/100 ml bovine serum albumin, 2 mM EDTA, and 4.5 mg/ml D-glucose) and filtered through 200-m mesh filter followed by 60-m mesh. The filtered cells were washed and finally suspended in 8 ml of growth medium (minimal essential/D-Val medium containing 2 mM glutamine, 10% fetal bovine serum, 5% human serum, 50 g/ml penicillin/streptomycin, 5 g/ml heparin, 1 g/ml hydrocortisone, 5 g/ml endothelial cell growth supplement from bovine brain, 5 g/ml amphotericin, and 5 g/ml mycoplasma removal agent). The cells were then transferred to a 100-mm culture dish. The cells were allowed to grow for 3-4 days. The endothelial cells appearing in patches were identified by light microscopy. Cloning rings were placed on these cells, trypsinized, and transferred to culture dishes. The cells were characterized by their cobblestone morphology, uptake of Dil-Ac-LDL (Biomedical Technologies Inc., Stoughton, MA) and staining for factor VIIIrelated antigen (Sigma Chemical Co., St. Louis, MO). Approximately 2-3 days later when cells were confluent, they were washed twice with phosphate-buffered saline, and the medium was changed to low serum medium (1% fetal bovine serum). Cells were treated with TNF␣ (500 units/ml) and/or co-cultured with PMN (1 ϫ 10 5 cells/ml), which had been isolated from WT or ICAM-1 knockout mice for the times indi-cated. ICAM-1 knockout mice were purchased from the Jackson Laboratory. In some experiments to confirm the role of oxidants derived from PMN in the mechanism of TNF␣-induced ICAM-1 expression in endothelial cells, glutathione ethyl ester (GSH, reduced form, 5 mM; Sigma) was added into the co-culture system. The cells were incubated with GSH for 2 h and then harvested for Western analysis.
Western Blot Analysis-Lung tissue homogenate samples or aliquots of MLVEC lysate or non-nuclear protein were separated on a 10% SDS-PAGE under non-reducing condition. Equivalent loading of the gel was determined by quantitation of protein as well as by reprobing membranes for actin detecting. Separated proteins were electroblotted onto polyvinylidene difluoride membrane and blocked for 1 h at room temperature with Tris-buffered saline containing 1% bovine serum albumin. The membranes were then probed with a 1:1000 dilution of either a purified polyclonal IgG against mouse ICAM-1, or, for nonnuclear protein, monoclonal anti-mouse IB␣ antibody (both from Santa Cruz Biotechnology, Santa Cruz, CA) at room temperature for 1 h. After washing, primary antibodies associated with the membranes were detected on autoradiographic film by horseradish peroxidaseconjugated secondary antibodies and the ECL plus chemiluminescence system (Amersham Biosciences, Inc., Arlington Heights, IL) according to the manufacturer's instructions.
Northern Blot Analysis-Total RNA from lungs was obtained using the guanidine isothiocyanate method (22). Briefly, lungs were harvested and immediately frozen in liquid nitrogen. Tissue was then thawed and homogenized in 4 M guanidine isothiocyanate containing 25 mM sodium citrate, 0.5% Sarkosyl, and 100 mM ␤-mercaptoethanol. RNA was denatured, electrophoresed through a 1.2% formaldehydeagarose gel, and transferred to a nylon membrane. Hybridization was carried out using a [ 32 P]dCTP-labeled ICAM-1 cDNA (from ATCC). Blots were then washed under conditions of high stringency, and specific mRNA bands were detected by autoradiography in the presence of intensifying screens. Blots were stripped and reprobed for glyceraldehyde-3-phosphate dehydrogenase to control for loading. Expression of mRNA was quantitated using Scion Image software (Scion Corp., Frederick, MD) and was normalized to the glyceraldehyde-3-phosphate dehydrogenase signal.
Nuclear Protein Extraction-Nuclear protein extracts were prepared from lung tissue or MLVEC by the method of Deryckere and Gannon (23). Aliquots of 100 mg of frozen tissue were ground to powder with a mortar in liquid nitrogen. The thawed powder or 1 ϫ 10 7 cells were homogenized in a Dounce tissue homogenizer with 4 ml of solution A (0.6% Nonidet P-40, 150 mM NaCl, 10 mM HEPES, pH 7.9, 1 mM EDTA, and 0.5 mM phenylmethylsulfonyl fluoride). The cells were lysed with five strokes of the pestle. After transfer to a 15-ml tube, debris was pelleted by briefly centrifuging at 2000 rpm for 30 s. The supernatant was transferred to 50-ml Corex tubes, incubated on ice for 5 min, and centrifuged for 10 min at 5000 rpm. The supernatant that contains non-nuclear protein was saved for further use. Nuclear pellets were then resuspended in 300 l of solution B (25% glycerol, 20 mM HEPES, pH 7.9, 420 mM NaCl, 1.2 mM MgCl 2 , 0.2 mM EDTA, 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 2 mM benzamidine, 5 g/ml pepstatin, 5 g/ml leupeptin, and 5 g/ml aprotinin) and incubated on ice for 20 min. The mixture was transferred to microcentrifuge tubes, and nuclei were pelleted by centrifugation at 14,000 rpm for 1 min. Supernatants containing nuclear proteins were aliquoted in small fractions, frozen in liquid nitrogen, and stored at Ϫ70°C. Protein quantitation was performed using the Bio-Rad protein assay dye reagent (Bio-Rad, Hercules, CA).
Electrophoretic Mobility Shift Assay-The probe for EMSA was a 24-bp double-stranded construct of NF-B consensus binding sequence (5Ј-AGGGACTTTCCGCTGGGACTTTCC-3Ј). End labeling was performed by T4 kinase in the presence of [ 32 P]ATP. Labeled oligonucleotides were purified on a Sephadex G-50M column (Amersham Biosciences, Inc., Piscataway, NJ).
An aliquot of 5 g of nuclear protein was incubated with the labeled double-stranded probe (ϳ50,000 cpm) in the presence of 5 g of nonspecific blocker, poly(dI-dC) in binding buffer (10 mM Tris-HCl, pH 7.5, 100 mM NaCl, 1 mM EDTA, 0.2% Nonidet P-40, and 0.5 mM dithiothreitol) at 25°C for 20 min. Specific competition was performed by adding 100 ng of unlabeled double-stranded oligonucleotide, whereas for nonspecific competition, 100 ng of unlabeled double-stranded mutant oligonucleotide (5Ј-AGCTCAATCTCCCTGGGACTTTCC-3Ј) that does not bind NF-B was added. The mixture was separated by electrophoresis on a 5% polyacrylamide gel in 1ϫ Tris glycine EDTA buffer. Gels were vacuum-dried and subjected to autoradiography and PhosphorImager (Molecular Dynamics) analysis.
PMN Adhesion Assay-The assay was performed as previously de-scribed (24). PMN isolated from either WT or p47 phoxϪ/Ϫ mice were co-cultured with MLVEC and challenged with 500 units/ml TNF␣ for 0 -6 h. In some groups, anti-ICAM-1 Ab (6 g/ml, Santa Cruz Biotechnology, Santa Cruz, CA) was added to the co-culture system. The percentage of PMN adhering to endothelial cells was determined from the ratio of final reading to initial reading. Statistics-The data are presented as mean Ϯ S.E. of the indicated number of experiments. Statistical significance among group means was assessed by analysis of variance. The Student Neuman-Keuls posthoc test was performed. Differences were considered significant when p Ͻ 0.05. Fig.  1A, in WT mice TNF␣ challenge caused a marked increase in ICAM-1 protein expression by 1 h, increasing further by 4 h. However, in p47 phoxϪ/Ϫ mice, the ICAM-1 protein expression at 1 h was abrogated and was significantly attenuated at the later time points. To confirm that this change in p47 phoxϪ/Ϫ was the result of NAD(P)H oxidase deficiency, ICAM-1 protein expression in gp91 phoxϪ/Ϫ mice was also assessed. The observed changes in ICAM-1 protein expression in gp91 phoxϪ/Ϫ mice were consistent with the p47 phoxϪ/Ϫ mice (Fig. 1A). The changes in mRNA expression for ICAM-1 shown in Fig. 1B paralleled the changes in the protein level. The promoter region of the ICAM-1 gene contains an NF-B consensus binding sequence believed to mediate the regulation of gene transcrip-tion (25). In the present studies, gel-shift assay on whole lung tissue was performed to discern whether the attenuation of ICAM-1 mRNA expression in p47 phoxϪ/Ϫ mice was the result of decreased nuclear translocation of NF-B. As shown in Fig. 1C, the response to TNF␣ in the WT animals exhibited an increase in NF-B translocation at 0.5 h, reached a maximum at 1 h, was sustained for 4 h, and decreased over the ensuing 2-h period. In p47 phoxϪ/Ϫ mice, however, TNF␣ induced a much smaller increase in NF-B translocation at 1 h, which did not increase further over the next 5 h. The level of NF-B activity in nuclear fractions in p47 phoxϪ/Ϫ mice during the period of 1-4 h after TNF␣ was on the average 60% less than in WT mice.

Role of NAD(P)H Oxidase in TNF␣-induced NF-B Signaling
Pathway-To address the role of NAD(P)H oxidase in mediating TNF␣-induced IB␣ degradation, a requirement for NF-B activation, the IB␣ protein level was measured in the nonnuclear fraction from the same lungs used for NF-B analysis above. IB␣ protein was detected in lungs prior to challenge with TNF␣. By 0.5 h after TNF␣, there was a significant loss of IB␣ to 18 Ϯ 6% of basal content in the WT mice ( Fig. 2A). IB␣ content remained low with 22 Ϯ 7% of basal value detected at 6 h after TNF␣ challenge. Thus, TNF␣ induced a rapid, substantial, and sustained loss of IB␣. However, the loss of IB␣ in p47 phoxϪ/Ϫ mice was significantly less by 2 h after TNF␣ challenge than in WT mice ( Fig. 2A). At 30 min after TNF␣, IB␣ in p47 phoxϪ/Ϫ mice was 74 Ϯ 11% of the basal level, stayed NIK is an upstream signaling molecule integral to the pathway of TNF␣ induction of NF-B nuclear translocation through the TNF␣ receptor (8). We examined whether the impaired IB␣ degradation in p47 phoxϪ/Ϫ is the result of reduced NIK activity. Fig. 2B shows the rapid increase in NIK activity at early time points (as early as 0.5 h) after challenge of WT animals with TNF␣. However, lungs of p47 phoxϪ/Ϫ animals exhibited markedly delayed and attenuated NIK activity. Compared with WT mice, the NIK activity in p47 phoxϪ/Ϫ at 0.5, 1, and 2 h after TNF␣ was 7 Ϯ 2%, 31 Ϯ 4%, and 25 Ϯ 5% of the WT values, respectively.
Binding of TNF␣ with its receptor initiates the recruitment of TNFR-associated proteins to the death domain of TNFR and a complex formation that induces activation of NF-B (6,7). In this study, we tested the effects of impairment of NAD(P)H oxidase function in the recruitment of three major TNFR-associated proteins, TRADD, TRAF2, and RIP using co-immunoprecipitation and Western blotting. As shown in Fig. 3, there were no differences between the WT and p47 phoxϪ/Ϫ mice in TRAF2 and RIP binding to TRADD after the TNF␣ challenge. Thus, NAD(P)H oxidase-deficiency induced decreased signaling and ICAM-1 expression occurred through a mechanism involving inhibition of NIK activity.
Neutrophil NAD(P)H Oxidase Contributes to TNF␣-induced ICAM-1 Expression-Because the PMN NAD(P)H oxidase is a major source of superoxide, we hypothesized that NAD(P)H oxidase would contribute to the oxidant-dependent NF-B activation. To determine the role of PMN, we evaluated the responses to TNF␣ challenge in lungs of WT and p47 phoxϪ/Ϫ mice following the depletion of circulating PMN. In some cases, we replenished the PMN in these mice made neutropenic to address the causal role of PMN in the mediating the response. As shown in Fig. 4A, at 2 h after TNF␣ challenge, neutropenia induced in WT mice reduced the ICAM-1 expression in lungs by ϳ62% (lane 4) as compared with the TNF␣ alone control group (lane 2). Repletion with WT PMN in these WT neutropenic mice restored the ICAM-1 expression in response to TNF␣ (lane 6). However, repletion with PMN derived from p47 phoxϪ/Ϫ mice failed to restore ICAM-1 expression (Fig. 4B). In p47 phoxϪ/Ϫ mice, a 40% lower expression of ICAM-1 was seen after TNF␣ challenge (Fig. 4A, lane 8) compared with WT mice (lane 2). Neutropenia in p47 phoxϪ/Ϫ mice did not further affect the ICAM-1 expression (lane 10), suggesting that the reduced ICAM-1 expression is the result of deficiency of PMN NAD(P)H oxidase. Repletion of p47 phoxϪ/Ϫ animals with WT PMN caused a complete restoration in ICAM-1 after TNF␣ (Fig. 4A, lane 12), further indicative of the role of PMN NAD(P)H oxidase in mechanism of TNF␣-induced ICAM-1 expression. Fig. 4C shows the effects of PMN depletion and repletion on lung tissue IB␣ level at 0.5 h after TNF␣ challenge. In WT animals, TNF␣ alone caused an 82 Ϯ 9% decrease in IB␣ (lane 2), whereas in the neutropenic WT group, TNF␣ caused a 38 Ϯ 8% reduction in IB␣ (lane 4). Repletion with WT PMN restored the decrease in IB␣ (lane 6). In contrast, neutropenia in p47 phoxϪ/Ϫ mice did not significantly affect IB␣ level in response to TNF␣ (lane 10) as compared with TNF␣ alone group (lane 8); although in the latter group the IB␣ level was ϳ52% higher than that in WT animals (lane 2). Repletion with WT PMN markedly reduced IB␣ in response to TNF␣ in the p47 phoxϪ/Ϫ animals (lane 12), and the value reached the same level seen in WT animals. Thus, the results in Fig. 4 show that the PMN NAD(P)H oxidase complex plays a critical role in the mechanism of TNF␣-induced ICAM-1 expression through the NF-B pathway.
Enhanced TNF␣ Induction of ICAM-1 in Endothelial Cells Depends on Neutrophil NAD(P)H Oxidase-Because endothelial cells express ICAM-1, we address the possibility in a endo-FIG. 3. Formation of TNFR-associated proteins complex in WT and p47 phox؊/؊ mice. Anti-TRADD antibody was used to co-immunoprecipitate TNFR-associated proteins from lung tissue homogenates of WT and p47 phoxϪ/Ϫ mice, which were harvested at 0.5 h after TNF␣ or saline i.p. injection. The precipitates were then subjected to Western analysis using anti-TRADD, -TRAF2, and -RIP antibodies, respectively. A representative of three independent studies is shown. FIG. 2. A, altered IB␣ levels in response to TNF␣ in the lung. IB␣ levels in the non-nuclear fractions from the lungs of WT and p47 phoxϪ/Ϫ mice were detected by Western analysis 0, 0.5, 1, 2, 4, and 6 h after TNF␣ injection and quantitated using densitometry, which normalized the data by corresponding actin density. A representative of three independent experiments is shown. B, effect of p47 phoxϪ/Ϫ on TNF␣-induced NIK activity. Mice were injected with TNF␣ (5000 units/10 g body weight, i.p.), and whole lungs were harvested at the time indicated after TNF␣. The lung tissue cytoplasm was immunoprecipitated with anti-NIK Ab, and a kinase assay was performed with [␥ 32 P]ATP using myelin basic protein as substrate. The data are representative of three independent studies. thelial cell/PMN co-culture system whether the PMN NAD(P)H oxidase complex could induce ICAM-1 in endothelial cells. We used MLVEC in which we measured ICAM-1 expression after TNF␣ stimulation. TNF␣ induced a gradual increase in ICAM-1 expression during a 1-to 4-h period in both WT and p47 phoxϪ/Ϫ MLVEC (Fig. 5A). The level of ICAM-1 expression in p47 phoxϪ/Ϫ MLVEC at each time point was lower than at the corresponding time points in WT MLVEC, suggesting that endogenous endothelial NAD(P)H oxidase is also involved in regulating TNF␣-induced ICAM-1 expression. However, co-culture of either WT or p47 phoxϪ/Ϫ MLVEC with WT PMN caused a rapid and augmented ICAM-1 expression in response to TNF␣ in both groups (Fig. 5A). To exclude the possibility that the increased ICAM-1 expression resulted from the ICAM-1 present in PMN, we also studied PMN obtained from ICAM-1 Ϫ/Ϫ mice and used these in the co-culture experiment. As shown in Fig. 5A, addition of the ICAM-1 Ϫ/Ϫ PMN still induced an augmented ICAM-1 expression in MLVEC, indicating that the increased ICAM-1 expression in MLVEC was the result of signals emanating from the PMN.
To address the role of oxidants derived from PMN in TNF␣induced ICAM-1 expression in endothelial cells, GSH (5 mM) was applied to the co-culture system. In biological system, O 2 .
is rapidly reduced by superoxide dismutase to H 2 O 2 , and GSH serves as an effective H 2 O 2 scavenger. As shown in Fig. 5B, models and endothelial/PMN co-culture experiments show that the oxidants derived from PMN contribute significantly to TNF␣-induced NF-B activation and ICAM-1 expression in endothelial cells. An important question is whether PMN adhesion to vascular endothelial cell is required for PMN to effectively transmit the oxidant signals to the endothelial cell. To determine the role of adhesive interactions in mediating the PMN-dependent NF-B activation and ICAM-1 expression in endothelial cells, we studied the effects of anti-E-selectin and anti-CD18 Abs (both from Santa Cruz Biotechnology, Santa Cruz, CA). As shown in Fig. 6, at 2 h after TNF␣, anti-E-selectin Ab caused a ϳ26% reduction in ICAM-1 expression in MLVEC. However anti-CD18 Ab resulted in a ϳ68% reduction of ICAM-1expression in MLVEC. The changes in ICAM-1 expression were correlated with alterations in NF-B nuclear translocation at 1 h after TNF␣ as shown in Fig. 6. Thus, the augmented activation of NF-B and expression of ICAM-1 in endothelial cells induced by PMN requires an adhesive interaction mediated by CD18 and to a lesser extent by E-selectin.
As shown in Fig. 1, it is evident that NAD(P)H oxidase deficiency did not block NF-B translocation and ICAM-1 mRNA and protein expression in response to TNF␣ at the later time points (4 and 6 h after TNF␣), suggesting that there is also an NAD(P)H oxidase-independent mechanism responsible for ICAM-1 expression. To address the time-dependent role of NAD(P)H oxidase after TNF␣ challenge, the adhesion of mouse PMN to endothelial cells (MLVEC) was assessed. As shown in Fig. 7, at time ϭ 0, there was similar basal adhesion of PMN from WT and p47 phoxϪ/Ϫ to MLVEC. At 1 and 2 h after TNF␣ stimulation, there were 2.4-and 3.2-fold increases in adhesion of WT PMN to MLVEC. This adhesion response was dependent on the expressed ICAM-1, because it was blocked by the anti-ICAM-1 Ab, whereas control IgG had no effect. In contrast, adhesion of PMN from p47 phoxϪ/Ϫ mice to MLVEC failed to increase at 1 h after TNF␣ stimulation, and was increased 1.6-fold at 2 h after TNF␣. However, at 4 and 6 h, when WT PMN exhibited 4.3-and 4.6-fold increases in adhesion, respectively, the p47 phoxϪ/Ϫ PMN showed similar increases in adhesion of 4.2-and 4.6-fold increase in adhesion, respectively. These data indicate an important role of the PMN NAD(P)H oxidase system in the early phase (up to 2 h) of the TNF␣-induced PMN adhesion response, which paralleled the early ICAM-1 expression. DISCUSSION TNF␣ has been shown to play an important role in the adhesion of neutrophils to endothelial cells (1,13). Firm neutrophil adhesion requires the activation of NF-B and the increased expression of ICAM-1 in endothelial cells (24,26,27). Oxidant production is considered as a key event in mediating the TNF␣-induced NF-B activation and initiation of NF-Bdependent transcription of ICAM-1 (13,28,29); however, the fundamental question of whether oxidants derived from adherent neutrophils can contribute to and thus amplify TNF␣induced ICAM-1 expression in endothelial cells remains to be addressed. In the present study, we show a novel function of phagocytic NAD(P)H oxidase in TNF␣-induced NF-B signal-

FIG. 6. Effect of anti-E-selectin and anti-CD18 Abs on PMNmediated enhanced ICAM-1 expression and NF-B nuclear translocation in MLVEC.
Confluent MLVECs were treated with TNF␣ (500 units/ml) in the presence or absence of WT PMN (1 ϫ 10 5 cells/ml) and/or either anti-E-selectin Ab or anti-CD18 Ab for 2 h, followed by washing with HBSS for three times, and cell lysis. Western blot was performed with the cell lysates, and EMSA was performed using the nuclear protein extracted from the lysates. E ϭ anti-Eselectin Ab; CD ϭ anti-CD18 Ab. The figure is representative of three studies.

FIG. 5.
A, effect of PMN co-culture on MLVEC ICAM-1 protein expression in response to TNF␣. MLVECs were isolated and cultured as described under "Experimental Procedures" and treated with TNF␣ (500 units/ml) for the times indicated. Co-cultured PMNs were isolated from WT and ICAM-1 Ϫ/Ϫ mice, respectively, and were in the concentration of 1 ϫ 10 5 cells/ml. At the end of incubation with TNF␣, MLVECs were washed with HBSS for three times and followed by cell lysis using lysis buffer. The cell lysates were then subjected to Western analysis with anti-ICAM-1 antibody. B, representative Western blot showing the effect of GSH on PMN-mediated enhanced ICAM-1 expression in MLVEC. Confluent MLVECs were treated with TNF␣ (500 units/ml) in the presence or absence of WT PMN (1 ϫ 10 5 cells/ml) and/or GSH (5 mM) for 2 h, followed by washing with HBSS for three times, and subsequent Western analysis.
ing and, thereby, in the mechanism of the rapid expression of ICAM-1 and endothelial adhesivity to neutrophils. NAD(P)H oxidase complex is found in a variety of phagocytic and non-phagocytic cells. Phagocytic NAD(P)H oxidase serves a critical function in host-defense against invading microorganisms (17). The non-phagocytic NAD(P)H oxidase also appears to be qualitatively similar in its ability to induce oxidant signaling, although oxidant generation is markedly less in nonphagocytic cells (17). Using in vivo PMN depletion and repletion experiments, our results demonstrate that 1) neutrophil depletion in WT mice reduced TNF␣-induced ICAM-1 expression, 2) repletion with WT neutrophil of either WT or p47 phoxϪ/Ϫ mice restored ICAM-1 expression in response to TNF␣ challenge, and 3) repletion of WT animals with p47 phoxϪ/Ϫ neutrophils failed to restore ICAM-1 expression in response to TNF␣. Thus, these data clearly demonstrate that neutrophil NAD(P)H oxidase is an important determinant of the TNF␣-induced NF-B activation and ICAM-1 expression. These findings are consistent with studies showing a role of NAD(P)H oxidase in the mechanism of NF-B activation induced by alcohol in the liver (30) or by surfactant D in alveolar macrophages (31) as well as the expression of monocytes chemoattractant protein and colony stimulating factor-1 in phagocytic cells in response to TNF␣ challenge (32).
Because lung tissue ICAM-1 assessed in the present study may be the result of ICAM-1 expression in multiple cell types (33), we carried out studies using mouse MLVEC and neutrophils to address the direct role of the neutrophil NAD(P)H oxidase in mediating ICAM-1 expression in endothelial cells. As shown in Fig. 5, WT neutrophils induced a rapid and augmented increase in ICAM-1 expression in both WT and p47 phoxϪ/Ϫ MLVEC, suggesting a direct interaction between neutrophils and MLVEC. Addition of the antioxidant GSH to the co-culture prevented the effect of WT neutrophil in inducing ICAM-1 expression in endothelial cells, indicating that the interaction between neutrophils responsible for ICAM-1 expression in MLVEC is mediated through oxidants. Thus, the data support the concept that endothelial cells are important cellular targets for the neutrophil NAD(P)H oxidase-derived oxidants and that they thereby mediate NF-B activation and ICAM-1 expression. This cross-talk between neutrophils and endothelial cells may contribute to the initial ICAM-1-dependent firm neutrophil adhesion to endothelial cells, and thus mediate the early-onset migration of neutrophils across the endothelial barrier.
The endogenous NAD(P)H oxidase in endothelial cells is also believed to be involved in signaling transduction (17). We showed that MLVEC from p47 phoxϪ/Ϫ mice demonstrated a lower expression of ICAM-1 in response to TNF␣ compared with WT MLVEC. This difference may reflect the role of endogenous NAD(P)H oxidase in signaling NF-B activation and ICAM-1 expression in these cells. However, when the MLVEC from p47 phoxϪ/Ϫ were co-cultured with WT neutrophils, expression of ICAM-1 in p47 phoxϪ/Ϫ MLVEC was elevated to same level as in WT MLVEC. This finding indicates that the exogenous oxidants from neutrophil NAD(P)H oxidase are clearly important in inducing ICAM-1 expression.
The present results indicate that the CD18 adhesive interaction between the neutrophil and endothelial cell is required for the neutrophil NAD(P)H oxidase-dependent expression of ICAM-1 in endothelial cells. Although ICAM-1 is expressed constitutively in endothelial cells, its level increases in response to a variety of pro-inflammatory stimuli, including TNF␣ (34). We also detected basal level of ICAM-1 by Western blotting. The anti-CD18 monoclonal antibody markedly reduced neutrophil-dependent NF-B activation and ICAM-1 expression in endothelial cells as compared with anti-E-selectin Ab. Thus, adhesion of neutrophils to endothelial cells mediated primarily by binding of constitutive ICAM-1 to CD18 provides the appropriate coupling required for neutrophils to transmit oxidant signals to endothelial cells.
NF-B activation and ICAM-1 expression induced by TNF␣ are dependent on oxidants derived from neutrophil NAD(P)H oxidase. These exogenous oxidants seem to be more important than oxidant signals generated endogenously in endothelial cells in mediating the early phase of the TNF␣-induced ICAM-1 expression. This is evident by the neutrophil adhesion experiments shown in Fig. 7. Within the 2-h period after TNF␣ challenge, addition of WT neutrophils caused an increase in neutrophil adhesion that was ICAM-1-dependent, because anti-ICAM-1 antibody prevented the response. In contrast, p47 phoxϪ/Ϫ neutrophils failed to adhere at these early time points consistent with the hypothesis that exogenously released neutrophil oxidants were critical for the early-onset ICAM-1 expression and resultant neutrophil adhesion. However, by 4 and 6 h after TNF␣ challenge, there was no difference in neutrophil adhesion between the WT and p47 phoxϪ/Ϫ neutrophils. Thus, it appears that the early phase of neutrophil adhesion is dependent on expression of ICAM-1 secondary to the release of neutrophil NAD(P)H oxidase-derived oxidants, whereas the latent ICAM-1-dependent neutrophil adhesion response is the result of other factors such as activation of the endothelial oxidant signaling machinery. Fig. 8 shows a model for the interaction of neutrophils and endothelial cells and the basis of neutrophil NAD(P)H oxidase-induced transcription of ICAM-1 in endothelial cells. TNF␣ has been reported to activate neutrophil NAD(P)H oxidase through stimulation of tyrosine kinases (35,36). On the basis of our data, we postulate that reactive oxygen species generated by neutrophil NAD(P)H oxidase serve a signaling function in activating NF-B in endothelial cells.
Although oxidants are involved in the NF-B signal transduction pathways, the molecular targets have not yet been defined. The contribution of redox regulation and location of potential redox-sensitive sites within the NF-B activation pathway are the subject of controversy (37). It has been reported that NF-␤ itself is a molecular target for redox regulation (38,39). DNA binding activity of NF-B can be regulated through reduction of a disulfide bond involving the cysteine residue 62 in p50 by thioredoxin (TRX) (38). Takeuchi and associates (40) showed that TRX inhibited TRAF2-, TRAF5-, and TRAF6-induced NF-B activation but did not affect NIK-, IKK␣-, and MEKK-induced activation, suggesting that TRX prevents NF-B-dependent transcription downstream of TRAFs and upstream of NIK. Recently, Korn et al. (41) demonstrated that hydrogen peroxide reduces NF-B activation by inhibiting activity of IB kinases in epithelial cells. It also appears that the ability of oxidants to induce NF-B activity depends on the cell type being studied (37). In the present study, NIK activity was significantly reduced in p47 phoxϪ/Ϫ mice after TNF␣ stimulation, whereas there was no difference in the formation of the upstream TNFR-associated protein complex in WT and p47 phoxϪ/Ϫ animals. The present results show that the redox-regulated site is located downstream of TNFR-associated proteins. Thus, NIK is a potentially important redox-regulated kinase responsible for NF-B activation. Redox regulation of NIK activity is further suggested by the presence of 22 cysteine residues in this molecule, which may constitute the redox-sensitive sulfhydryl switches (42). Future studies will be needed to define the precise mechanisms responsible for redox-regulation of NIK.
In summary, the present study demonstrates a novel function of neutrophil NAD(P)H oxidase-derived oxidant signaling in mediating the early-onset TNF␣-induced NF-B activation and ICAM-1 expression. Neutrophil NAD(P)H oxidase medi-ates these responses through its ability to activate NIK. The results indicate that the endothelial cell is an important target for the neutrophil NAD(P)H oxidase activity resulting in stable endothelial expression of ICAM-1. The functional relevance of neutrophil NAD(P)H oxidase-induced endothelial ICAM-1 expression may be to induce the early-onset firm neutrophil adhesion and enable the rapid migration of neutrophils across the vessel wall to site of infection.
FIG. 8. Hypothetical model for interaction of the neutrophil and endothelial cell. TNF␣ stimulation results in NAD(P)H oxidase activation and production of reactive oxygen species in neutrophils, as well as initiation of NF-B signaling in endothelial cell. Adhesion of neutrophils to endothelial cells is primarily mediated by binding of constitutive ICAM-1 to CD18 and provides the appropriate coupling required for neutrophils to transmit oxidant signals to endothelial cells. The oxidants activate NF-B signaling and ICAM-1 expression and induce stable adhesion of neutrophil to endothelial cell. NIK may serve as an important redox-regulated kinase responsible for NF-B activation.