Integrins and Cytokines Activate Nuclear Transcription Factor-κB in Human Neutrophils*

Neutrophil adhesion to extracellular matrix is necessary for an effective inflammatory response. Adhesion may accelerate neutrophil activation by affecting intracellular signaling pathways. The nuclear transcription factor κB (NF-κB) controls several cellular functions, including inflammation, proliferation, and cell survival. We explored the role of adhesion in NF-κB activation in human neutrophils. Cells were stimulated with tumor necrosis factor-α (TNF-α), granulocyte macrophage-colony-stimulating factor (GM-CSF), interleukin-8 (IL-8), and formyl-methionyl-leucyl-phenylalanine (fMLP). All four initiated neutrophil adherence to and spreading on fibronectin. GM-CSF and IL-8 did not activate NF-κB in suspended neutrophils but rapidly activated NF-κB under adherent conditions on matrix, as shown by IκB kinase activity assay, IκBα degradation, electromobility shift assay, and quantitative reverse transcriptase-PCR. In contrast, TNF–α activated NF-κB both in suspended cells and adherent cells. fMLP did not activate NF-κB in either suspended or adherent cells. Specific β2 integrin blockade prevented NF-κB activation by GM-CSF and IL-8 on fibronectin. Co-stimulating CD18 and CD11b with activating antibodies resulted in NF-κB activation by GM-CSF and IL-8 in suspended cells. We inhibited actin polymerization with cytochalasin and blocked the non-receptor kinase Syk with piceatannol. Both maneuvers prevented the co-stimulatory NF-κB-activating signal by β2 integrins. Thus, in addition to β2 integrin ligand binding, NF-κB activation depended on the formation of the receptor-associated intracellular focal adhesion complex. We conclude that β2 integrins may provide co-stimulatory signals allowing some soluble mediators to activate the NF-κB pathway even when they are not capable of doing so in suspension. This effect may become important when human neutrophils leave the circulating blood and migrate through extracellular matrix during inflammation.

In resting neutrophils, NF-B is found in the cytoplasm tightly bound to inhibitory proteins of the IB family. Upon activation, the inhibitors are phosphorylated on serine residues by the multicomponent IB kinase (IKK). The IKK complex consists of two catalytic subunits (IKK␣ and IKK␤) and one regulatory subunit (IKK␥). After phosphorylation, IB␣ is rapidly degraded, allowing NF-B to translocate into the nucleus and to activate target genes. The NF-B family includes five members, namely p50, p52, p65, c-Rel, and RelB (8,10,19). So far, only p50 and p65 and probably c-Rel have been detected in human neutrophils (14,15,20). Recently, we used a two-domain peptide containing the NEMO binding domain (NBD) to provide firm evidence that NF-B controls neutrophil survival (20). In addition to anti-apoptotic mediators, NF-B-dependent gene transcription also includes molecules that participate in inflammation (21). Thus, stimulation of NF-B in neutrophils may provide an acceleration loop for inflammation.
McDonald et al. (14) found that TNF-␣ and lipopolysaccharide triggered the strongest NF-B activity in neutrophils, whereas other mediators, including GM-CSF and IL-8, lacked the ability to activate this pathway. Whether or not integrinmediated adhesion plays a role in neutrophil NF-B activation is not known. We tested the hypothesis that ␤ 2 integrin-dependent adhesion to extracellular matrix provides an important signal for NF-B activation in neutrophils. We describe a previously unrecognized co-stimulatory NF-B-activating effect of integrins that might be important when neutrophils emigrate from the circulation during inflammation. Cruz, CA). p50 and p52 antibodies were from Rockland and Upstate Biotechnology, Inc. (Lake Placid, NY). The blocking monoclonal antibody (mAb) to CD18 (7E4) was from Immunotech (Marseille, France), and the activating clone KIM 185 was a generous gift from Martyn Robinson (19). Activating mAb to CD11b (BEAR1) and the monoclonal isotype control were from Immunotech. Blocking mAb to CD11b (2LPM 19c) was from Dako (Hamburg, Germany). The mAb to Syk was from New England Biolabs GmbH (Frankfurt am Main, Germany). The mAb to phosphotyrosine (4G10) from Transduction Laboratories (Lexington, KY). Piceatannol was purchased from Calbiochem. Fibronectin from human plasma was obtained from Roche Applied Science. Horseradish peroxidase-labeled donkey anti-rabbit IgG was from Amersham Biosciences. Hanks' balanced salt solution (HBSS), phosphate-buffered saline, and trypan blue were from Biochrom (Berlin, Germany). Endotoxin-free reagents and plastic disposables were used in all experiments.
Preparation and Culture of Human Neutrophils-Neutrophils from healthy human donors were isolated from heparinized whole blood by red blood cell sedimentation with dextran 1%, followed by Ficoll-Hypaque density gradient centrifugation. Erythrocytes were lysed by incubation with hypotonic saline for 15 s. PMN were spun down (1050 rpm, 10 min) and reconstituted in HBSS with calcium and magnesium. Samples were incubated at 37°C in 5% CO 2 either in polypropylene tubes (BD Biosciences) or on fibronectin-coated (10 g/cm 2 ) tissue culture grade polystyrene plates to achieve adherent conditions. Coating was performed as suggested by the manufacturer. The final cell concentration was 5 ϫ 10 6 cells per ml. Preincubation with the indicated reagents was carried out in polypropylene tubes for 30 min at 37°C. Stimulation was done with 2 ng/ml TNF-␣, 20 ng/ml GM-CSF, 100 nM IL-8, and 1 M fMLP, respectively. The cell viability was detected in every cell preparation by trypan blue exclusion and was found to be greater than 99%. The percentage of PMN after isolation was Ͼ95% by Wright-Giemsa staining and by light microscopy.
Flow Cytometry to Assess Expression of ␤ 2 Integrins-Flow cytometry was used to evaluate the membrane expression of CD11b and CD18. Immunostaining was done as described previously (22). Cells were incubated with fluorescein isothiocyanate-conjugated antibodies against CD11b (BEAR1) or CD18 (7E4). Flow cytometry was performed using a FACSort (BD Biosciences), and 10,000 events per sample were collected.
Assessment of Neutrophil Adhesion and Spreading-96-Well plates coated with (10 g/cm 2 ) fibronectin were used for the adhesion assay. 1 ϫ 10 5 PMN in 100 l HBSS 2ϩ were either left untreated or were treated with 2 ng/ml TNF-␣, 20 ng/ml GM-CSF, 100 nM IL-8, and 1 M fMLP, respectively. Plates were incubated at 37°C in 5% CO 2 for the indicated time. Wells were flicked dry and washed three times with phosphate-buffered saline, and adherent cells were estimated using the MPO assay. Briefly, adherent cells were lysed in 100 l of 0.5% Triton X-100 for 10 min. 100 l of substrate (2,2Ј-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid, Sigma) were added, and absorbance was read after 10 min at 450 nm with a microtiter plate reader. Each experiment was done in triplicate. Absorbance of the experimental sample was compared with a standard curve that showed an excellent correlation between absorbance and cell number. The standard curve was established over a range of 1 ϫ 10 4 to 1 ϫ 10 5 cells and showed an R 2 value of 0.96. For estimation of cell spreading, PMN were cultured in 12-well plates as described above. At the indicated time points, non-adherent cells were discarded and the percentage of spread cells was assessed. At least 100 cells were counted using phase contrast microscopy, and those that were phase dark, enlarged, and irregular were considered spread.
Cytoplasmic and Nuclear Extract Preparation-Cytoplasmic and nuclear extracts were prepared as described previously (20). The reaction was stopped with ice-cold RPMI supplemented with DFP (2 mM final concentration). Cells were centrifuged at 2000 ϫ g for 2 min at 4°C, and pellets were resuspended with a hypotonic buffer A (10 mM HEPES, pH 7.5, 10 mM KCl, 0.1 mM EGTA, 0.1 mM EDTA, pH 8.0) containing an anti-protease mixture (2 mM DFP, 1 mM 4-(2-aminoethyl)benzenesulfonyl fluoride, 1 mM phenylmethylsulfonyl fluoride, 1 g/ml aprotinin, 1 g/ml leupeptin, 1 g/ml pepstatin A, 0.5 mM benzamidine, 1 mM DTT). Cells were kept on ice for 10 min, followed by the addition of 1% Nonidet P-40. Cells were vortexed briefly and pelleted. The resulting supernatant was referred to as the cytoplasmic fraction and stored at Ϫ80°C. For the preparation of nuclear extracts, we observed the best results when omitting Nonidet P-40. Apparently Nonidet P-40 itself led to the degradation of the nuclei. After the addition of buffer A, cells were vortexed and spun down at 1000 ϫ g (10 min, 4°C) to pellet the nuclei and to remove intracellular granula that may cause nuclear degradation. The pellet was washed again with buffer A, and the hypertonic, high salt buffer C was added (20 mM HEPES, pH 7.5, 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 20% glycerol) together with the above-mentioned protease inhibitor mixture. After centrifugation (10 min at 14,000 ϫ g, 4°C), the supernatant was collected and referred to as the nuclear fraction. Protein measurements were done using the Bio-Rad assay (Bio-Rad).
Electrophoretic Mobility Shift Assays (EMSAs)-EMSAs were performed as described previously (20). Briefly, nuclear extracts (5 g of FIG. 2. IB␣ degradation in suspension neutrophils and in neutrophils on fibronectin was studied by Western blot. Samples were incubated with 2 ng/ml TNF-␣, 20 ng/ml GM-CSF, 100 nM IL-8, and 1 M fMLP, respectively. Cells were harvested at intervals up to 60 min; cytoplasmic extracts were prepared and analyzed for IB␣. The data show that GM-CSF and IL-8 did not degrade IB␣ in suspension, whereas this effect occurred under adherent conditions. TNF-␣ induced IB␣ degradation in both suspension and under adherent conditions, whereas fMLP showed no effect whatsoever. The results are representative of four independent experiments. protein) were incubated with 20,000 cpm of a 22-bp oligonucleotide containing the NF-B consensus sequence that had been labeled with [␣-32 P]ATP T4 polynucleotide kinase. Incubations were performed for 30 min at room temperature in the presence of poly(dI-dC) and 20 mM HEPES, containing 60 mM KCl, 4% Ficoll, 5 mM DTT, and 0.5 g/l nuclease-free bovine serum albumin. Probes were subjected to electrophoresis on native 5% polyacrylamide gels and autoradiographed. For supershift assays the indicated antibody was added 15 min before the addition of the radiolabeled probe. To determine the specificity of shifted bands, excess unlabeled oligonucleotide was added to the nuclear extracts 10 min before addition of the radiolabeled probe. The oligonucleotides for H 2 K are as follows: forward primer 5ЈGATCCAG-GGCTGGGGATTCCCCATCTCCACAGG3Ј, and reverse primer 5ЈGAT-CCCTGTGGAGATGGGGAATCCCCAGCCCTG3Ј, and for the mutated form, forward primer 5ЈGATCCAGGGCTAGCGATTCCCCATCTCCA-CAGG3Ј (mutated sites in bold). Equal excess of the oligonucleotides was used for competition experiments.
IB Kinase Activity Assay-To assay IB kinase activity, cells were stimulated for 60 min as indicated or were left untreated. Prior to stimulation, samples were preincubated for 30 min with a stimulating antibody to CD11b or an isotype control. Whole cell lysates were prepared using lysis buffer (50 mM HEPES, pH 7.5, 150 mM NaCl, 1,5 mM MgCl 2 , 1 mM EDTA, 1% Triton X-100, 10% glycerol) with the protease inhibitor mixture as described above. Cell equivalents were used for immunoprecipitation that was performed in lysis buffer. Before the addition of 1 g of monoclonal IKK␣ antibody (Pharmingen), extracts were precleared with protein A-Sepharose (Amersham Biosciences) for 30 min at 4°C. Binding of the antibody was carried out overnight at 4°C, and the next day 25 l of protein A-Sepharose was added for an additional 2 h. The protein A-Sepharose was washed three times with lysis buffer and once with kinase buffer (20 mM HEPES, pH 7.5, 10 mM MgCl 2 , 20 mM ATP, 20 M ␤-glycerophosphate, 1 mM DTT). 15 l of kinase buffer including purified recombinant IB␣ and 3 Ci of [␥-32 P]ATP were added to the protein A-Sepharose immunocomplex, and the kinase reaction was incubated for 20 min at 37°C. Samples were boiled and subjected to 12% SDS-PAGE and analyzed by autoradiography. Equal loading was confirmed by IKK␣ Western blot experiments.
Quantitative RT-PCR-Total RNAs were isolated according to a Qiagen protocol including DNase treatment. Quantitative RT-PCR was performed using Taqman technology (Applied Biosystems, Weiterstadt, Germany). Reverse transcription was carried out according to the Superscript protocol (Invitrogen). Taqman RT-PCR was performed using the Master Mix (Applied Biosystems). The quantification was checked for each sample using probes for GAPDH mRNA. Primers and probes were designed using the primer express program (Applied Biosystems). The following oligonucleotides were used for IB␣: forward primer 5Ј-CCCTGTAATGGCCGGACTG3Ј, reverse primer 5ЈAGGAGTGACACC-AGGTCAGGA3Ј, and the probe Fam 5ЈCCTTCACCTCGCAGTGGACC-TGC3Ј Tamra. RT-PCR and quantification were performed using the Taqman 5700 (Applied Biosystems). For quantification of the amount of RNA present in the various samples, the fluorescence signal was measured at each PCR cycle, and the increase in the fluorescence-normalized reporter signal (RN) was documented in an amplification plot. By using non-template controls, the threshold was set in the log phase to subtract unspecific fluorescence signals. Cycle threshold (C t ) values were determined for each sample. In short, the C t value difference (⌬C t ) was used to calculate the factor of differential expression (2 ⌬Ct ). Results were imported in an Excel spreadsheet and analyzed according to the standard curve method.
Immunoprecipitation of Syk-Lysates from 2.5 ϫ 10 7 neutrophils were prepared by incubating the cells for 10 min on ice in lysing buffer (50 mM HEPES, pH 7.5, 150 mM NaCl, 1.5 mM MgCl 2 , 1 mM EDTA, 1% Triton X-100, 0.1% SDS, 0.5% deoxycholic acid, 10% glycerol, supplemented with 5 mM DFP, 2 mM phenylmethylsulfonyl fluoride, 2 g/ml aprotinin, 2 g/ml leupeptin, 1 g/ml pepstatin A, 2 g/ml chymostatin, FIG. 3. NF-B binding activity by EMSA is shown. Neutrophils were incubated either in suspension or on fibronectin with buffer (CTRL), 2 ng/ml TNF-␣ (TNF), 20 ng/ml GM-CSF, 100 nM IL-8, and 1 M fMLP (FMLP), respectively. After 60 min, nuclear extracts were prepared and analyzed by EMSA using an H 2 K-binding site probe for NF-B (n ϭ 3). GM-CSF and IL-8 did not increased NF-B binding activity in suspension but when cells were incubated on fibronectin. TNF-␣ induced NF-B binding activity in both suspension and under adherent conditions, whereas fMLP showed no effect. The amount of nuclear extract used for the binding reaction was 5 g of protein for all samples.

FIG. 4. NF-B binding activity is demonstrated by EMSA.
Specificity of the bands was assessed by competition with a cold probe and with supershift experiments. A 20-fold excess of unlabeled NF-B probe and specific antibodies against p50 were added to nuclear extracts before incubation with labeled NF-B oligonucleotide probe (n ϭ 3). Two specific bands could be identified; the lower band as the p50/p50 homodimer and the upper band as the p50/p65 heterodimer. . Total RNAs were isolated according to a Qiagen protocol including DNase treatment, and quantification was checked for each sample by using probes for GAPDH mRNA. The oligonucleotides used for IB␣ were described under "Materials and Methods." The data indicate that GM-CSF and IL-8 did not induce IB␣ mRNA in suspension cells but triggered a strong increase when neutrophils were incubated on fibronectin.
10 mM sodium pyrophosphate, 10 mM sodium orthovanadate). Before the addition of 10 g of anti-Syk antibody, extracts were precleared with protein A-Sepharose (Amersham Biosciences) for 30 min at 4°C. Binding of the antibody was carried out for 2 h at 4°C, before protein A-Sepharose was added for an additional 2 h. The protein A-Sepharose was washed three times with lysis buffer. 50 l of 2ϫ loading buffer were added to the protein A-Sepharose immunocomplex followed by boiling for 5 min at 95°C. Samples were subjected to 8% SDS-PAGE, and Western blot analysis was performed with the indicated antibodies.
Western Blot-Samples were incubated for 5 min at 95°C in loading buffer (250 mM Tris-HCl, pH 6.8, with 4% SDS, 20% glycerol, 0.01% bromphenol blue, 10% ␤-mercaptoethanol). 5-20 g of protein was loaded per lane, electrophoresed on a 10% SDS-polyacrylamide gel, and transferred to a nitrocellulose membrane. The membrane were blocked with TBS-T/5% skim milk for 1 h and incubated overnight with the indicated antibodies. Membranes were washed and incubated with a horseradish peroxidase-labeled secondary antibody. The blot was developed by incubation in a chemiluminescence substrate (ECL, Amersham Biosciences) and exposed to an x-ray film. We confirmed equal loading of protein by stripping and reprobing the blots for ␣-actin and Syk, respectively.
Statistical Analysis-Results are given as mean Ϯ S.E. Comparisons between two groups were done using paired Wilcoxon rank tests. Comparisons between multiple groups were done using Kruskal Wallis tests. Specific differences between multiple groups were then determined by use of a Bonferroni post-hoc test on the ranked values.

RESULTS
TNF-␣, GM-CSF, IL-8, and fMLP Promote PMN Adhesion in a Time-dependent Manner-We performed a time course study for PMN adhesion on a fibronectin-coated surface after treatment with TNF-␣, GM-CSF, IL-8, and fMLP, respectively. Cells were harvested after the indicated time points, and adhesion was assessed using the MPO-assay. Fig. 1 shows that TNF-␣, GM-CSF, IL-8, and fMLP time-dependently induced neutrophils to adhere to fibronectin. TNF-␣ and fMLP resulted in rapid adhesion, whereas slower kinetics was observed with GM-CSF and IL-8 (n ϭ 5).
NF-B Activation in PMN on Fibronectin Versus Suspension Cells-NF-B activation in neutrophils on fibronectin was compared with suspended cells. In parallel experiments using neutrophils from the same donor, cells were treated with TNF-␣, GM-CSF, IL-8, or fMLP in either suspension or under adherent conditions on fibronectin. Samples were harvested after 10, 30, and 60 min and assessed for IB␣ protein expression (n ϭ 4). Fig. 2 shows a typical example for each of the stimuli. TNF-␣ induced IB␣ degradation in suspension and under adherent conditions, whereas fMLP showed no effect under either condition. Interestingly, GM-CSF and IL-8 did not degrade IB␣ in suspension, whereas this effect occurred under adherent conditions. By using the same conditions, we performed parallel Western blot experiments demonstrating that GM-CSF, IL-8, and fMLP activate p38 MAPK, ERK, and Akt in suspended neutrophils (data not shown). These experiments indicate that neutrophil adhesion to fibronectin provides a co-stimulatory signal for NF-B activation for some inflammatory mediators such as GM-CSF and IL-8, whereas others, such as TNF-␣, do not need this additional signal. fMLP is an example of an inflammatory mediator that does not activate NF-B even under adherence conditions. We then performed electromobility shift assays (EMSA) to investigate the effect of adhesion on NF-B activation by a second independent method. The results confirmed that TNF-␣ induced IB␣ degradation in suspension, whereas either suspension or adherence with fMLP showed no effect. Again, GM-CSF and IL-8 did not activate NF-B in suspension neutrophils, whereas activation occurred under adherent conditions (Fig. 3).
For TNF-␣, we also performed supershift experiments with antibodies directed against the p50 and p65 NF-B subunit (Fig. 4). An antibody against p50 strongly reduced the upper and lower band with the formation of an intensive supershifted complex. Serum against p65 caused a decrease of the upper band with the appearance of a supershifted complex, whereas no change in the intensity of bands was observed with an irrelevant antibody (data not shown). From these results, the lower band can be identified as a p50/p50 homodimer and the upper band as the p50/p65 heterodimer. Specificity was verified by competition with a 20-fold excess of an unlabeled NF-B probe (cold probe).
To study whether or not GM-CSF and IL-8 initiated NF-Bdependent gene transcription, we employed quantitative RT-PCR. Generation of IB␣, the cytoplasmic inhibitor of NF-B, is itself under the control of NF-B and was used for that reason. Both cytokines resulted in a significant increase in IB␣ mRNA when neutrophils were incubated on fibronectin. In contrast no FIG. 7. A, the effect of blocking antibodies to CD18 and CD11b on IB␣ degradation in fibronectin-adherent neutrophils was studied by Western blot. Samples were preincubated with buffer (Ϫ), an isotype control (ISO), or the blocking antibodies (CD18 or CD11b). After 30 min, cells were plated on fibronectin and stimulated with buffer (CTRL), 20 ng/ml GM-CSF, or 100 nM IL-8, respectively. Samples were harvested after 60 min and assayed for IB␣. These data show that specific blockade of either CD18 or CD11b completely prevented IB␣ degradation by GM-CSF and IL-8 in neutrophils on fibronectin. The results are representative of three independent experiments. B, the effect of blocking antibodies to CD18 and CD11b on NF-B activation in fibronectinadherent neutrophils was studied by EMSA. Samples were preincubated with buffer (Ϫ), an isotype control (ISO), or the blocking antibodies (CD18 or CD11b). After 30 min cells were plated on fibronectin and stimulated with buffer (CTRL), 20 ng/ml GM-CSF, or 100 nM IL-8, respectively. After 60 min, nuclear extracts were prepared and analyzed by EMSA using an H 2 K-binding site probe for NF-B (n ϭ 3). GM-CSF and IL-8 increased NF-B binding activity in cells that were incubated on fibronectin. Preincubation with blocking antibodies to both CD18 and CD11b but not an isotype control (ISO) abrogated this effect.
such effect was observed when cells were stimulated in suspension (Fig. 5).
TNF-␣, GM-CSF, IL-8, and fMLP Up-regulate ␤ 2 Integrins-Blocking ␤ 2 integrin antibodies abrogate neutrophil adhesion and spreading as well as NF-B activation after treatment with GM-CSF, and IL-8 ␤ 2 integrins are pivotal to neutrophil adhesion on fibronectin. We next investigated the effect of TNF-␣, GM-CSF, IL-8, and fMLP on the membrane expression of the ␤ 2 integrins CD11b and CD18. Fig. 6 shows that stimulation with all four compounds resulted in significant up-regulation of both ␤ 2 integrin receptors at 30 min (CD18 in Fig. 6A, CD11b in n ϭ 5). By using TNF-␣ we titrated a time course for membrane CD11b expression. Stimulation with 2 ng/ml TNF-␣ resulted in an increase of membrane CD11b by 248 Ϯ 63% after 5 min, by 221 Ϯ 33% after 15 min, and by 244 Ϯ 4% after 30 min (n ϭ 2). This rapidly occurring increase in CD11b expression was not dependent on de novo protein synthesis because preincubation with 2.5 g/ml of the protein synthesis inhibitor cycloheximide did not affect TNF-␣-induced up-regulation of CD11b (244 Ϯ 4% increase without cycloheximide versus 274 Ϯ 22% with cycloheximide at 30 min). When we then preincubated neutrophils with blocking mAb to CD11b and CD18, we observed significant adhesion inhibition. Fig. 6C shows results for blocking anti-CD18, Fig. 6D for blocking anti-CD11b and spreading, Fig. 6E for blocking anti-CD18, and Fig. 6F for blocking anti-CD11b in response to TNF-␣, GM-CSF, IL-8, and fMLP, respectively. These data indicate that blocking antibodies to CD11b and CD18 that were used in our experimental settings effectively abrogated neutrophil adhesion and spreading. prepared followed by immunoprecipitation with a monoclonal IKK␣ antibody. Precipitates were incubated with 1 g of GST-IB␣ and 3 Ci of [␥-32 P]ATP. Samples were boiled and subjected to 12% SDS-PAGE and autoradiography. Equal loading was confirmed by IKK␣ Western blot. B, the effect of activating antibodies to CD18 and CD11b on IB␣ degradation in suspension neutrophils was studied by Western blot. Samples were preincubated with buffer (Ϫ), an isotype control (ISO), or the activating antibodies (CD18 or CD11b). After 30 min, cells were treated with buffer (CTRL), 20 ng/ml GM-CSF, or 100 nM IL-8, respectively. Samples were harvested after 60 min and assayed for IB␣. The data show that specific activation of either CD18 or CD11b induced IB␣ degradation by GM-CSF and IL-8, even when incubation was carried out in suspension. The results are representative of three independent experiments. C, the effect of activating antibodies to CD18 and CD11b on NF-B activation in suspension neutrophils was studied by EMSA. Samples were preincubated with buffer (Ϫ), an isotype control (ISO), or the activating antibodies (CD18 or CD11b). After 30 min, cells were treated with buffer (CTRL), 20 ng/ml GM-CSF, or 100 nM IL-8, respectively. After 60 min, nuclear extracts were prepared and analyzed by EMSA using an H 2 K-binding site probe for NF-B (n ϭ 3). GM-CSF and IL-8 increased NF-B binding activity in cells that were incubated on fibronectin. Preincubation with blocking antibodies to both CD18 and CD11b but not an isotype control (ISO) abrogated this effect. We next explored whether or not ␤ 2 integrins play a specific role for adhesion-mediated NF-B activation in neutrophils. Pretreatment with CD11b and CD18 blocking antibodies prevented IB␣ degradation (Fig. 7A) and NF-B DNA binding (Fig. 7B) in GM-CSF-and IL-8-stimulated neutrophils on fibronectin. In contrast, TNF-␣-stimulated NF-B activation was not affected by ␤ 2 integrin blockade (data not shown).
Activating Antibodies to ␤ 2 Integrins Stimulate NFB Activation in GM-CSF-and IL-8-treated Suspension Neutrophils-We considered the possibility that NF-B inhibition by blocking CD11b and CD18 antibodies resulted from a more general blockade of adhesion rather than from specific blockade of CD11b-and CD18-transduced signaling events. Thus, we studied NF-B activation in suspended neutrophils that were stimulated with both the soluble mediator GM-CSF or IL-8 and an activating mAb to CD11b or CD18, respectively. Fig. 8A shows a representative IB kinase activity assay. Fig. 8B depicts a typical Western blot experiment for IB␣ expression and Fig. 8C a typical EMSA. These data support the contention that GM-CSF and IL-8 activate NF-〉 in neutrophils providing a co-stimulatory ␤ 2 integrin signal via CD11b or CD18 was received.
Cytochalasin and the Syk Inhibitor Piceatannol Prevent NF-B Activation in GM-CSF-and IL-8-treated Neutrophils on Fibronectin-Signaling by ␤ 2 integrins depends on the assembly of an intracellular focal adhesion complex. Formation of this multicompound complex includes scaffold proteins such as actin together with cytosolic kinases, including the non-receptor tyrosine kinase Syk. To determine the role of this complex for ␤ 2 integrin-mediated NF-B activation, we performed experiments using cytochalasin B, a compound that inhibits actin polymerization, and piceatannol, a specific Syk kinase blocker. Fig. 9 demonstrates that pretreatment with cytochalasin abrogated NF-B activation in GM-CSF-and IL-8-treated neutrophils on fibronectin.
To determine the role of Syk kinase in ␤ 2 integrin-mediated NF-B activation, we focused on IL-8 treatment because GM-CSF receptor signaling involves Syk kinase. By immunoprecipitation, we observed that only combined treatment with IL-8 and activating mAb to CD18 resulted in increased basal tyrosine phosphorylation of Syk kinase in suspended neutrophils (Fig. 10A). Blockade of Syk kinase activity by a low concentration of the specific inhibitor piceatannol (10 M) abrogated this increase in tyrosine phosphorylation (Fig. 10B) and inhibited NF-B activation in neutrophils that received combined treatment with IL-8 and activating CD18 mAb (Fig. 10C). These data suggest that both the cytoskeletal and signaling molecules with enzymatic activity are needed for ␤ 2 integrins to provide co-stimulation for cytokines finally activating NF-B. DISCUSSION We describe a novel mechanism for NF-B activation in human neutrophils. We show that the interaction of neutrophils with fibronectin via ␤ 2 integrins provides a strong costimulatory signal for NF-B activation. This effect allows cytokines such as GM-CSF and IL-8 that do not activate NF-B in suspension to rapidly trigger this pathway during adhesion. This effect may become important when human neutrophils leave the circulating blood and migrate through extracellular matrix during inflammation.
NF-B activation in response to inflammatory cytokines was reported for various cell types, including human neutrophils. By using pharmacological inhibitors, several studies (12,(15)(16)(17)(18)(23)(24)(25)(26) suggested a functional significance of this pathway for neutrophil survival. Recently, we used a two-domain peptide containing the NEMO binding domain (NBD) to specifically block NF-B and the TAT protein transduction domain to shuttle NBD into PMN. These experiments with TAT-NBD provided firm evidence for an important role of NF-B in neutrophil survival (20). The most potent stimuli for NF-B in neutrophils are TNF-␣ and lipopolysaccharide, whereas other mediators such as interferon ␥, GM-CSF, and IL-8 were reported to lack the capability of activating this transcription factor in these cells (14,20,26). However, no studies are available that have investigated the effect of cell-matrix interaction on cytokine-mediated NF-B activation in neutrophils.
␤ 2 integrins are exclusively expressed on leukocytes and consist of a common ␤-chain (CD18) that is non-covalently linked to three unique but related ␣-chains (CD11a or LFA-1, CD11b or Mac-1, and CD11c or 150,95) (27)(28)(29)(30). ␤ 2 integrins promote adhesion and thereby regulate several functions of Suspended neutrophils were treated with buffer (ctrl), IL-8, activating mAb to CD18 (CD18), or a combination of IL-8 and the activating mAb to CD18 (IL8ϩCD18), respectively. After 10 min, Syk kinase was immunoprecipitated (IP), and tyrosine phosphorylation of the Syk kinase (72 kDa) was assessed by Western blotting (WB) by using the mAb 4G10 (A). Membranes were stripped and reprobed with an anti-Syk antibody (Syk WB). By immunoprecipitation, the effect of the specific Syk inhibitor piceatannol on Syk phosphorylation of neutrophils that received combined treatment with IL-8 and activating mAb to CD18 was assessed (B). Western blot was used to study IB␣ degradation in neutrophils that received combined treatment with IL-8 and activating mAb to CD18 (C). Cells were pretreated for 30 min with 10 M piceatannol or buffer control before stimulation. The data show increased basal tyrosine phosphorylation of Syk kinase in suspended neutrophils after combined treatment with IL-8 and activating mAb to CD18 and that piceatannol prevented both phosphorylation and NF-B activation. The results are representative of two independent experiments. human neutrophils (1,2,4,7,31). Integrins can directly initiate outside-in signaling or cross-talk with other receptors resulting in signaling cascade activation (32,33). It was demonstrated previously (34 -38) that cell adhesion via integrins directly activates NF-B in monocytic cell lines and endothelial cells, respectively. In contrast to these cell types, our data show that ␤ 2 integrins alone are not sufficient to trigger NF-B in human neutrophils unless an additional signal is received. GM-CSF and IL-8 are important inflammatory cytokines that stimulate signaling cascades such as p38 MAPK, ERK, phosphatidylinositol 3-kinase, and Akt in suspension neutrophils (39 -41).
Our study confirms that both cytokines do not activate NF-B in suspension neutrophils but instead provide new data indicating that they do activate this pathway when co-stimulation, via the ␤ 2 integrin subunits CD11b or CD18, was provided. We used two different approaches to reach this conclusion. Specific blocking antibodies to either the variable CD11b chain or the common CD18 chain prevented GM-CSF-and IL-8-induced NF-B activation on fibronectin. However, blockade of these adhesion molecules almost completely prevented adhesion and therefore might have suppressed other adhesiondependent events with importance for this process as well. Thus, the second approach of stimulating CD11b and CD18 directly, in suspension, provided firm evidence that these molecules were not sufficient to trigger NF-B but were necessary for GM-CSF and IL-8 to activate this pathway. The activating effect occurred rapidly and was of the same robustness as seen with TNF-␣ that is known to be one of the most potent NF-B activators. As a consequence, NF-B-dependent genes may be induced in adherent neutrophils in the presence of GM-CSF and IL-8 as demonstrated for IB␣, which itself is generated in an NF-B-dependent manner. Interestingly, we found that although integrin co-stimulation triggered the anti-apoptotic NF-B pathway, delayed neutrophil apoptosis by GM-CSF and IL-8 was not augmented under these conditions (data not shown). This finding suggests simultaneous activation of proapoptotic pathways. An earlier report (42) demonstrated that ␤ 2 integrins were involved in the activation of c-Jun N-terminal kinase in neutrophils and that this effect participated in the early pro-apoptotic effect of TNF-␣.
TNF-␣ is an important pro-inflammatory cytokine that strongly activates adherent neutrophils, whereas this cytokine has little activating effects on neutrophils in suspension (1,7). ␤ 2 integrins are necessary for TNF-␣ to trigger several signaling cascades (3,5,42,43). However, our study indicates that TNF-␣ activates the NF-B pathway even in suspended neutrophils and that adhesion provided no additional helper signals. This finding suggests that even low concentrations of TNF-␣ may ignite NF-B-dependent gene expression in circulating neutrophils. Approximately 1 ϫ 10 11 neutrophils are released daily from the bone marrow, a number that may be increased by several conditions (44). Thus, during inflammation a large amount of neutrophilic NF-B-dependent gene products could be generated as a consequence.
fMLP is a synthetic bacterial peptide that activates several signaling pathways in neutrophils, ultimately resulting in respiratory burst, adhesion, transmigration, and degranulation. We observed that fMLP did not stimulate NF-B, either in suspension or with a ␤ 2 integrin co-stimulatory signal. This finding suggests that not all inflammatory mediators can utilize ␤ 2 integrin signaling for NF-B activation. Because we performed parallel control experiments showing that fMLP was capable of activating p38 MAPK, ERK, and Akt (data not shown), we can rule out a more general lack in fMLP signaling as well as problems with our experimental conditions. The extracellular domain of ␤ 2 integrins binds to an Arg-Gly-Asp (RGD) sequence that is found in several matrix proteins, including fibronectin, whereas the intracellular tail of the receptor interacts with both scaffold proteins such as actin and with cytosolic enzymes (29,30,32,33). Multiple components were identified that form a focal adhesion complex (32). We were not able to demonstrate direct co-immunoprecipitation of the CD18 receptor and IKK␣ (data not shown). However, a variety of signaling enzymes and adaptor proteins link the integrin receptor to distinct intracellular pathways. Our data indicate that a cytoskeletal interaction via actin is required to activate NF-B through ␤ 2 integrins. In addition, cytosolic kinases are necessary for integrin signaling. The non-receptor tyrosine kinase Syk was shown to be important for integrin signaling but also for signaling via Fc-␥, T-and B-cell, and GM-CSF receptor (32,34,42,(45)(46)(47)(48)(49)(50). Because GM-CSF receptor signaling involves Syk activation, we focused on IL-8 to investigate the significance of Syk kinase in adhesion-mediated NF-B activation. Recently, it was shown that Syk is constitutively associated with ␤ 2 integrins in neutrophils (51). In agreement with these data, we demonstrated a basal tyrosine phosphorylation of Syk kinase that was increased only by combined stimulation of IL-8 receptors and ␤ 2 integrins. When we blocked Syk activity with piceatannol in low concentrations (52,53), we observed inhibition of NF-B activation. These data suggest that Syk kinase may be an important component of ␤ 2 integrin-mediated NF-B activation. In agreement with our observations on NF-B, the importance of Syk kinase for integrin-dependent neutrophilic functions was demonstrated previously (48) by using Syk Ϫ/Ϫ cells.
Inflammation creates a cytokine-rich milieu that activates intracellular signaling pathways ultimately initiating neutrophil defense mechanisms. Neutrophils either circulate in the bloodstream or interact with cells and extracellular matrix proteins during migration toward an inflammatory site. Our study indicates that NF-B activation depends not only on the type of cytokine but also on the compartment where these cytokines challenge the neutrophil. Targeting ␤ 2 integrin-mediated neutrophil adhesion may suppress NF-B-activated gene expression and could therefore provide an useful antiinflammatory strategy.