Phosphorylation of Tumor Necrosis Factor Receptor 1 (p55) Protects Macrophages from Silica-induced Apoptosis*

Macrophages play a fundamental role in silicosis in part by removing silica particles and producing inflammatory mediators in response to silica. Tumor necrosis factor α (TNFα) is a prominent mediator in silicosis. Silica induction of apoptosis in macrophages might be mediated by TNFα. However, TNFα also activates signal transduction pathways (NF-κB and AP-1) that rescue cells from apoptosis. Therefore, we studied the TNFα-mediated mechanisms that confer macrophage protection against the pro-apoptotic effects of silica. We will show that exposure to silica induced TNFα production by RAW 264.7 cells, but not by IC-21. Silica-induced activation of NF-κB and AP-1 was only observed in RAW 264.7 macrophages. ERK activation in response to silica exposure was only observed in RAW 264.7 macrophages, whereas activation of p38 phosphorylation was predominantly observed in IC-21 macrophages. No changes in JNK activity were observed in either cell line in response to silica exposure. Silica induced apoptosis in both macrophage cell lines, but the induction of apoptosis was significantly larger in IC-21 cells. Protection against apoptosis in RAW 264.7 cells in response to silica was mediated by enhanced NF-κB activation and ERK-mediated phosphorylation of the p55 TNFα receptor. Inhibition of these two protective mechanisms by specific pharmacological inhibitors or transfection of dominant negative mutants that inhibit IκBα or ERK phosphorylation significantly increased silica-induced apoptosis in RAW 264.7 macrophages. These data suggest that NF-κB activation and ERK-mediated phosphorylation of the p55 TNF receptor are important cell survival mechanisms in the macrophage response to silica exposure.

Environmental exposure to crystalline silica induces pulmonary inflammation that often leads to fibrosis (1). Following phagocytosis of silica particles, macrophages are activated to release cytokines. Among the cytokines secreted by lung macrophages in response to silica, tumor necrosis factor ␣ (TNF␣) 1 has been shown to play a critical role in the pathogenesis of silicosis, and procedures that antagonize the biological effects of TNF␣ ameliorate silica-induced pulmonary fibrosis in mice (2)(3)(4).
TNF␣ transduces its biological activities by binding to two receptors of 55 and 75 kDa (5,6). Trimeric occupation of TNF receptors by the ligand results in the recruitment of receptorspecific proteins and promotes the activation of transcription factors that induce expression of TNF␣-responsive genes (7,8). Several lines of evidence support the importance of TNF receptors in the pathogenesis of silica-induced lung fibrosis. Following silica exposure an enhanced TNF receptor mRNA expression can be observed in the mouse lung (9). This enhanced TNF receptor expression correlates well with the activation of NFB and AP-1 in the lungs of silica-sensitive (C57BL/6) mice (10). In contrast to wild-type mice, individual and double TNF receptor-deficient mice (developed on a C57BL/6 genetic background) are protected from the inflammatory and fibrotic effects of silica (9,10).
Binding of TNF␣ to its p55 receptor triggers the activation of downstream signaling cascades that lead to apoptosis in a variety of cells (11). However, TNF␣ interaction with the p55 receptor also induces NF-B activation that promotes the expression of pro-survival genes (12). Recent studies implicate macrophage apoptosis as an important element in the pathogenesis of silicosis (13,14). Silica induces apoptosis of human alveolar macrophages in vitro (15). In addition, the number of apoptotic cells in bronchoalveolar lavage fluid from silicatreated rats is significantly higher than in control-treated animals (16). The importance of TNF␣-mediated signal transduction and in particular the role of NF-B and AP-1 activation in silica-induced apoptosis in macrophages is not completely understood.
TNF␣ induces AP-1 activation via phosphorylation and activation of mitogen activated protein kinases (MAPKs). Stimulation of mouse macrophages with TNF␣ rapidly promotes ac-tivation of specific members of the ERK, JNK, and p38 subfamilies of MAPKs (17)(18)(19). Crystalline silica also induces AP-1 transactivation in different cell types (20). Endotracheal instillation of silica in C57BL/6, but not in TNF receptor-deficient mice, is followed by AP-1 activation in the lung (10). These observations suggest that TNF receptors play an important role in regulating AP-1 activation in response to silica. However, the role that specific members of the MAPK family play during silica-induced AP-1 activity is not completely understood.
Engagement of the p55 receptor by the TNF␣ ligand is followed by ERK activation in murine macrophages (21). Activated ERK2 (p42) is then recruited to the membrane proximal region of the p55 receptor where it induces phosphorylation at ERK consensus sites (22). Phosphorylated p55 is accumulated in the endoplasmic reticulum where it recruits Bcl-2 and protects against apoptosis (21,23). Whether or not ERK-mediated phosphorylation of p55 receptor occurs in macrophages in response to silica is unknown.
TNF␣ production by mouse macrophage cell lines in response to silica exposure is not universal (24). RAW 264.7 cells constitute a macrophage cell line derived from BALB/c mice that react to silica exposure with enhanced TNF␣ production (24). In contrast, macrophages from the IC-21 mouse cell line (derived from C57BL/6 mice) do not stimulate TNF␣ production upon exposure to silica (24). These differences are important as the expression of TNF␣ correlates with the mouse strain sensitivity to silica, and we reported that mice from both C57BL/6 and BALB/c strains enhance TNF␣ production and undergo silica-induced lung injury (9). Therefore, we took advantage of the difference in TNF␣ production by these macrophage cell lines to study the role that TNF␣-mediated activation of NF-B and MAPKs plays on the protection of macrophages against silica-induced apoptosis. In particular, we investigated the importance of ERK-mediated phosphorylation of p55 TNF receptor in protecting macrophages from silica-induced apoptosis.
The compounds BAY 11-7085, SB 203580, PD 98059, and 5-iodotubercidin were purchased from BIOMOL Research Laboratories, Inc. (Plymouth Meeting, PA). All compounds were dissolved in Me 2 SO and stored in aliquots at Ϫ70°C.
Cell Lines and Cultures-RAW 264.7 (ATCC TIB 71) and IC-21 (ATCC TIB 186) murine macrophage cell lines were obtained from the American Type Culture Collection (Rockville, MD). RAW 264.7 and IC-21 cells were maintained in Dulbecco's modified Eagle's medium and RPMI 1640 (Invitrogen, Rockville, MD), respectively, supplemented with 10% fetal bovine serum, 100 units/ml penicillin G, and 100 g/ml streptomycin and grown at 37°C in 5% CO 2 . In all experiments cells were grown to 80% confluence in the media described above.
Assay of TNF␣ Production by Macrophages-Cells were plated in 6-well plates and grown for 3 days. Cells (ϳ2 ϫ 10 6 cells/well) were exposed to crystalline silica (10 g/cm 2 ) for 6 h in triplicate. Untreated cells and cells stimulated with LPS (4 g/ml) were used as negative and positive control, respectively. Following silica exposure, TNF␣ was assayed in cell-free supernatants with the use of an ELISA kit (Mouse ELISA TNF, Endogen, Woburn, MA) following manufacturer's recommendations. 10 pg/ml was identified as the lower concentration of TNF␣ detectable with the use of this kit.
Membranes were probed first for TNF␣, then stripped and reprobed for 36B4, as a loading control. Blots were developed after 72 h using Biomax films and intensifying screens (Kodak). To quantitate mRNA abundance, membranes were exposed to a Fuji phosphorimaging device (Fujix BAS 1000, Fuji USA, Standford, CT) plate overnight and scanned. Quantitative analysis was determined with the use of McBAS 2.5 software (Fuji USA). For each mRNA band the results were normalized to the internal control (36B4) and expressed as a -fold increase between the bands for the control and the bands for the silica-treated cells.
Western Blot Analysis-Cells were seeded in 6-well plates and stimulated with silica or LPS (positive control) at the doses indicated for 6 h. In experiments where MAPK inhibitors were used cells were pretreated with p38 MAPK inhibitor SB 203580 (1 M), ERK inhibitor 5-iodotubercidin (5 M), or PD 098059 (50 M) for 1 h and 30 min before silica exposure. At the end of the treatment, cells were lysed in buffer containing 50 mM Tris/HCl, pH 7.5, 150 mM NaCl, 1% (v/v) Triton X-100, 10 mM EDTA, 100 mM NaF, 20 g/ml leupeptin, 10 g/ml aprotinin, and 1 mM Na 3 VO 4 , incubated on ice for 30 min and centrifuged at 16,000 ϫ g for 10 min at 4°C. Protein concentration of the cell lysates was determined by the Lowry method using the Bio-Rad DC Protein assay kit (Bio-Rad). An equal volume of 2ϫ Laemmli sample buffer was added, and the samples were boiled for 5 min. 50 g of protein was separated on 10% SDS-PAGE and electroblotted onto nitrocellulose membranes (Schleicher & Schuell, Inc.). Membranes were blocked in blocking buffer (5% nonfat dry milk in Tris-buffered saline, pH 7.6, 0.05% (v/v) Tween 20 (TBST)) for 1 h at room temperature and incubated with the relevant antibodies in blocking buffer overnight at p55 Phosphorylation Protects from Silica-induced Apoptosis 4°C. After washing with TBST, blots were incubated with the appropriate peroxidase-conjugated secondary antibody for 1 h at room temperature and developed using an enhanced chemiluminescence detection kit (ECL, Amersham Biosciences).
Analysis of Macrophage Apoptosis by Flow Cytometry-Apoptosis was analyzed by evaluating the fluorescein-labeled annexin V binding and the exclusion of propidium iodide by macrophages, as previously described (26). Briefly, cells were seeded in 12-well plates and exposed to silica (20 and 50 g/cm 2 ), or medium as negative control for 6 h. In some experiments cells were pre-treated with MAPK inhibitors or NFB inhibitor BAY 11-7085 at the concentrations indicated, for 1 h and 30 min. After treatment cells were washed with PBS, pH 7.4, and incubated for 15 min at room temperature in the presence of FITClabeled annexin V, propidium iodide, or both at the same time. Cells were then washed twice with binding buffer (10 mM HEPES, pH 7.4, 150 mM NaCl, 5 mM KCl, 1 mM MgCl 2 , 2.5 mM CaCl 2 ), scraped, and subjected to flow cytometry analysis. Flow cytometry analysis was performed on a fluorescence-activated cell sorter (Vantage, BD Biosciences) with argon ion laser tuned to 488 nm. Fluorescence emission was measured through 530/30-nm band-pass (FITC) and 575/26-nm band-pass (propidium iodide) filters. Data were collected as 5000 list mode event files and analyzed using Cell-Quest software (BD Biosciences).
Immunoprecipitates were separated by SDS-PAGE on 10% polyacrylamide gels under nonreducing conditions and analyzed by Western blotting as described above. Blots were incubated with hamster monoclonal anti-mouse TNFR1 (p55) or goat polyclonal anti-TNFR2 (p75) overnight at 4°C, followed by incubation with goat polyclonal antihamster IgG-peroxidase or rabbit polyclonal anti-goat IgG-peroxidase conjugated, respectively.
Transfection of RAW 264.7 Macrophages with Dominant Negative IB and ERK Mutants-To determine the biological specificity of the results observed in RAW 264.7 macrophages with the use of the kinase inhibitors on the silica-induced apoptosis, we transfected these cells with constructs expressing dominant negative IB␣ and ERK mutants.
The adenoviral vector directing the expression of dnIB␣, in which serines 32 and 36 have been mutated to alanine and a triple influenza hemagglutinin tag added at its 5Ј-end, has been described previously (27,28). The cDNA insert of plasmid pRc/CMV-IB␣S32A/S36A, which contains a hemagglutinin-tagged human super-repressor of NF-B, was subcloned into the Xbal site of the pACCMV.PLPASR(ϩ) plasmid to construct the plasmid pACCMV/IB, in which IB␣ is driven by the cytomegalovirus (CMV) promoter/enhancer. The adenovirus encoding the enhanced green fluorescent protein gene under a CMV promoter (AdEGFP) was used as a control (29).
Dominant negative ERK1 (Lys 71 3 Arg) and ERK2 (Lys 52 3 Arg) mutants each cloned in pCEP4 vector were obtained from Dr. Melanie Cobb and have been previously described (30). Extracellular signalregulated protein kinases (ERK) 1 and 2 and the respective mutants were expressed in bacteria with a hexahistidine tag and purified using nickel-chelate chromatography. Basal activity of wild type ERK2 was ϳ2 nmol/min/mg. Self-catalyzed phosphorylation occurred in vitro on the major physiological site of tyrosine phosphorylation in an intramolecular reaction. Rabbit muscle ERK activator activated ERK2 500-to 1000-fold up to a specific activity (ϳ2 mol/min/mg) approximating that of ERK1 purified from stimulated cells (31). The ERK activator to the same extent could also activate ERK1. These ERK mutants lack the major site of tyrosine phosphorylation, are autophosphorylated at a greatly reduced rate, and are no longer activated by the ERK kinase.
1.5 ϫ 10 5 cells/well were plated in 12-well plates and grown for 24 h as described above. Cells were infected with adenoviral vectors (4 ϫ 10 9 viral particles/well) or transfected with 1 g of DNA using GeneJuice transfection reagent following manufacturer's recommendations (EMD Biosciences, Inc., Novagen brand, Madison, WI). 48 h later, cells were exposed to silica or medium alone for 6 h as described above. After 6 h, cells were harvested for Western blot or flow cytometry studies as described above.
Statistical Analysis-Experiments were performed in triplicates. Results are reported as means Ϯ S.E. Statistical comparisons were made using one-way analysis of variance (Statview 4; Abacus Concepts, Inc., Berkeley, CA) of variance followed by a Student-Neuman-Keuls test for post-hoc pairwise comparisons. A p value of Ͻ0.05 was considered significant.

RESULTS
Silica Induces TNF␣ Production in RAW 264.7, but Not in IC-21, Macrophages-Previous studies have established that the phagocytosis of silica particles by macrophages induces TNF␣ production (32). However, the ability of macrophages to secrete TNF␣ in response to silica is not universally observed and Claudio and associates (24) reported that different macrophage cell lines respond to silica with variable TNF␣ production. Therefore, we studied the differences in TNF␣ production in two macrophage cell lines in response to in vitro silica exposure.
TNF␣ mRNA expression was studied by Northern analysis in RAW 264.7 and IC-21 macrophages after 6 h exposure to silica (10 g/cm 2 ), LPS (4 g/ml), or control media. As illustrated in Fig. 1A, exposure of RAW 264.7 cells to LPS or silica resulted in significant (p Ͻ 0.001) increases in TNF␣ mRNA expression compared with unstimulated controls. In contrast, IC-21 macrophages did not significantly increase (p Ͼ 0.2) TNF␣ mRNA expression in response to silica, although they demonstrated significant (p Ͻ 0.02) TNF␣ mRNA expression in response to LPS stimulation.
Silica Alters Expression of TNF Receptor and TNF Receptorassociated Proteins (TRADD, TRAF1, and TRAF2) in RAW 264.7, but Not in IC-21, Macrophages-We have previously reported that exposure of mice to silica is followed by an upregulated expression of the p75, but not the p55, TNF␣ receptor mRNA in the mouse lung (9). Therefore, we characterized the expression of TNF␣ receptors in RAW 264.7 and IC-21 macrophages following silica exposure. To study TNF␣ receptor expression, we performed immunoprecipitation against each individual (p55 or p75) receptor. As shown in Fig. 1C, untreated RAW 264.7 and IC-21 expressed both p55 and p75 TNF receptors. Silica exposure (20 g/cm 2 , for 6 h), of RAW 264.7, but not IC-21, macrophages was associated with enhanced expression of the p75 TNF␣ receptor. In contrast, no effect on total p55 p55 Phosphorylation Protects from Silica-induced Apoptosis expression was observed in either macrophage cell line following silica exposure (Fig. 1C).
To determine whether or not silica induced changes in TNF receptor-associated proteins in the macrophage cell lines, we studied the protein expression for the p55 TNF receptor-associated death domain (TRADD), and the p75 TNF receptor factorassociated (TRAF1 and TRAF2) proteins, by immunoblotting (Fig. 1D). TRADD is the p55 receptor death domain-associated protein and is involved in TNF␣-mediated apoptosis and NF-B activation (33,34). TRAF1 and TRAF2 contribute in a direct manner to NF-B and AP-1 activation (8,35,36). In addition, TRAF1 and 2 recruit inhibitor of apoptosis proteins (cIAP) to the p55 receptor complex. The recruitment of IAP inhibits the activation of caspase 3 (37, 38) and caspase 8 (39), preventing TNF␣-induced apoptosis. As shown in Fig. 1D, RAW 264.7 macrophages demonstrate low levels of TRADD expression.
This low level of TRADD expression was not greatly modified in these cells by exposure to either silica (5-50 g/cm 2 ) or LPS (4 g/ml). In contrast to RAW 264.7 macrophages, IC-21 cells did not express detectable levels of TRADD protein in controltreated cells. TRADD expression was not enhanced in IC-21 cells following LPS (4 g/ml) or silica (5-50 g/cm 2 ) exposure.
Similar to the data regarding TRADD, differences in the baseline expression of TRAF1 were observed between the macrophage cell lines. RAW 264.7, but not IC-21, cells showed baseline TRAF1 expression. This baseline expression was clearly enhanced in both cell lines following LPS (4 g/ml), but not silica (5-50 g/cm 2 ), exposure (Fig. 1D).
Basal expression of TRAF2 protein was observed in both RAW 264.7 and IC-21 macrophage cell lines. However, the TRAF2 signal observed in RAW 264.7 cells was greater that that of IC-21 cells. TRAF2 expression was not significantly FIG. 1. TNF␣ production and expression of TNF␣ receptor and TNF receptor-associated proteins by RAW 264.7 and IC-21 macrophages in response to silica. A, representative Northern blot analysis of TNF␣ and 36B4 (loading control) mRNA expression by RAW 264.7 and IC-21 cells following control, silica (10 g/cm 2 ), or LPS (4 g/ml) exposure as described under "Experimental Procedures." B, TNF␣ protein measured by ELISA in cell-free supernatants 6 h after exposure of RAW 264.7 and IC-21 macrophages to silica or LPS as described above. TNF␣ protein production is expressed as picograms/ml. (*, p Ͻ 0.05, significant difference compared with control-treated cells; †, p Ͻ 0.05, significant difference compared with silica-treated IC-21 cells. C, RAW 264.7 and IC-21 macrophages (2 ϫ 10 7 cells) were exposed to silica (20 g/cm 2 ) or media as control for 6 h. p55 and p75 TNF␣ receptors were immunoprecipitated and analyzed by Western blotting as described under "Experimental Procedures." D, RAW 264.7 and IC-21 macrophages were cultured in vitro for 6 h in the presence of silica (10 g/cm 2 ), LPS (4 g/ml), or media as control and then lysed as described under "Experimental Procedures." TRADD, TRAF1, and TRAF2 expression were studied, by Western blotting, in cell lysates following silica exposure. The gel is representative of four experiments. altered in either cell line following LPS or silica exposure (Fig. 1D).
Silica Promotes NF-B and AP-1 Activation in RAW 264.7, but Not in IC-21, Macrophages-Expression of TNF␣-responsive genes is mediated via the activation of specific transcription factors (40). The most prominent transcription factors activated by the TNF␣ signal pathway are NF-B and AP-1 (40,41). Therefore, we studied NF-B and AP-1 activation in silicastimulated RAW 264.7 and IC-21 cells (by electrophoretic mobility shift assay). Silica treatment induced NF-B ( Fig. 2A) and AP-1 (Fig. 3A) activation in RAW 264.7, but not in IC-21 cell line. However, IC-21 cells activated NF-B and AP-1 in response to LPS ( Figs. 2A and 3A). The specificity of the observed NF-B binding was demonstrated by the fact that add-ing a non-labeled NF-B oligonucleotide inhibited the binding. Also the use of an antibody specific to the p50, but not the p65, subunit of NF-B caused a shift of the NF-B complexes ( Fig. 2A).
Silica Induces Differential Activation of MAPKs in RAW 264.7 and IC-21 Cells-Phosphorylation of MAPKs, including p38, ERK1/2, and SAPK/JNK kinases, lead to the formation of homo-or heterodimers of c-Fos/c-Jun and AP-1 activation (42,43). Therefore, we studied the effect of silica exposure on the activation (phosphorylation) measured by Western blotting, of MAPKs in RAW 264.7 and IC-21 macrophages.
Silica exposure induced ERK1/2 activation in RAW 264.7, but not in IC-21 cells (Fig. 4A). Baseline p38 expression was similar in both macrophage cell lines. However, silica exposure enhanced p38 MAPK phosphorylation in IC-21, but not in RAW 264.7, macrophages (Fig. 4B). No differences in total or phosphorylated SAPK/JNK MAPKs were observed in either of two cell lines (Fig. 4C).
Maximal silica-induced ERK phosphorylation was evident at 10 g/cm 2 and appeared to plateau at 20 g/cm 2 with no further changes observed when these cells were exposed to higher concentrations of particles (Fig. 5A).
No change in silica-induced ERK phosphorylation was observed when RAW 264.7 cells were pre-treated with 1 M SB 203580 (Fig. 5B). In contrast, silica-induced ERK activation was prevented when RAW 264.7 cells were exposed to silica in the presence of the PD 98059 (50 M) or 5-iodotubercidin (5 M) (Fig. 5B).
Macrophage Apoptosis in Response to Silica is Enhanced in IC-21 Macrophages-It has been suggested that macrophage apoptosis may be important in initiating the inflammatory response to silica (14,15,46). Therefore, we evaluated, by flow cytometry, the induction of apoptosis by silica in RAW 264.7 and IC-21 cells. Silica Induces ERK-mediated Phosphorylation of p55 TNF Receptor in RAW 264.7, but Not in IC-21, Macrophages-Recent reports indicate that TNF␣-induced activation of ERK2 MAPK mediates the phosphorylation of p55 cytoplasmic domain and that this protects cells against TNF receptor 1-mediated apoptosis while preserving NF-B activation (21). Therefore, we investigated whether the ERK activation observed in response to silica was associated with phosphorylation of the p55 TNF receptor. To study p55 phosphorylation RAW 264.7 and IC-2 macrophages were metabolically labeled with 32 P prior to silica exposure (20 g/cm 2 , for 6 h). Following silica exposure p55 was immunoprecipitated, and its degree of phosphorylation was determined by autoradiography. As illustrated in Fig. 7, p55 phosphorylation was only observed in RAW 264.7 cells. This p55 phosphorylation was inhibited by pre-treatment of RAW 264.7 cells with the MEK1 (PD98059 (50 M)) but not with the p38 (SB 203580 (1 M)) inhibitor suggesting that the silica-induced phosphorylation of p55 is an ERKmediated process. To further confirm that ERK mediates the p55 phosphorylation we used the novel specific ERK2 inhibitor, 5-iodotubercidin (47). When RAW 264.7 cells were exposed to 20 g/cm 2 silica for 6 h in the presence of 5-iodotubercidin (5 M) no phosphorylation of the receptor was observed (Fig. 7).
ERK-mediated Phosphorylation of p55 Protects RAW 264.7 Macrophages from Silica-induced Apoptosis-Cottin et al. (21) have reported that ERK2-mediated phosphorylation of p55 induces changes in the subcellular localization of the receptor. This receptor redistribution is associated with recruitment of Bcl-2, which protects cells against apoptosis (23). Therefore, we examined the effect of inhibiting the ERK-mediated phosphorylation of p55 on the silica-induced apoptosis of RAW 264.7 macrophages. As shown in Fig. 8, ERK inhibition by PD98059 (50 M) or 5-iodotubercidin (5 M) significantly (p Ͻ 0.001) increased silica-induced apoptosis in RAW 264.7 macrophages.
To determine the biological specificity of the results observed with the pharmacological inhibitors of ERK, we transfected RAW 264.7 macrophages with genetically manipulated mutant ERK1/2 that can not be phosphorylated and, therefore, behave as dominant negative proteins (30) to measure the effect of overexpressing these proteins on the silica-induced apoptosis. Transfection of RAW 264.7 with these ERK mutants, but not with the pCEP4 empty vector, effectively eliminated the silicainduced phosphorylation of ERK proteins (Fig. 5C) and significantly (p Ͻ 0.001) increased the silica induction of apoptosis in these cells (Fig. 9), thus confirming the results observed with the pharmacological inhibitors of ERK.
As discussed above silica exposure induced NF-B activation in RAW 264.7 macrophages. Because NF-B activation is an important mechanism against apoptosis, we studied whether or not the enhanced apoptosis observed in RAW 264.7 cells following the exposure to the ERK inhibitors could be attributed to inhibitory effects of these compounds on NF-B activation. Therefore, we studied the effects of PD98059 or 5-iodotubercidin treatments on the silica-induced activation of NF-B in RAW 264.7 macrophages (Fig. 2B). To clarify the role of NF-B activation on the silica induction of apoptosis we incubated RAW 264.7 macrophages with BAY 11-7085, which is a powerful inhibitor of IB␣ phosphorylation, or infected these cells with adenovirus containing a mutant IB␣ protein that cannot be phosphorylated and therefore prevents NF-B activation (48). As can be observed in Figs. 2 and 6, the addition of BAY (40 M) to cell cultures eliminated NF-B activation (Fig.  2B) and significantly (p Ͻ 0.001) increased silica-induced apoptosis (65% of the cells stained positively for annexin V) in RAW 264.7 macrophages (Fig. 6). In a similar manner, transfection of RAW 264.7 macrophages with dnIB␣, but not control vector (AdEGFP), significantly increased silica-induced apoptosis (Fig. 9). In contrast, treatment with PD98059 or 5-iodotubercidin enhanced NF-B activation on these cells (Fig.  2B). This finding confirms the idea that NF-B activation is a protective mechanism against silica-induced apoptosis and that the enhanced apoptosis observed in RAW 264.7 macrophages after treatment with PD98059 or 5-iodotubercidin is not due to an inhibitory effects of these compounds on NF-B activation.

DISCUSSION
Macrophages play an important role in the pathogenesis of silica-induced lung fibrosis in part through their ability to clear silica particles and apoptotic cells from the lung (1,49). In addition, macrophages are the initial source of inflammatory mediators produced in response to silica exposure that include oxygen radicals, leukotrienes, cytokines, and growth factors (1). Therefore, the concept has been advanced that the excessive loss of lung macrophages in response to silica could compromise the ability of the host to clear silica particles and remove apoptotic cells while promoting excessive inflammation. Thus, understanding the mechanisms regulating macro- p55 Phosphorylation Protects from Silica-induced Apoptosis phage cell death in response to silica constitute a first step in promoting adequate resolution of this type of lung injury.
A large body of experimental evidence implicates TNF␣ as a fundamental cytokine in the pathogenesis of silicosis. TNF␣ expression is up-regulated in the lungs of silica-treated mice in a manner that precedes the development of lung fibrosis (2). Furthermore, enhanced TNF␣ expression in the lung correlates with the murine strain sensitivity to silica and C57BL/6 mice, FIG. 5. Silica induces dose-dependent ERK1/2 activation in RAW 264.7 macrophages. RAW 264.7 cells were treated with increasing concentrations of silica (5, 10, or 20 g/cm 2 ), LPS (4 g/ml), or roridin (RA) as positive controls, or media alone as control for 6 h. A, cell lysates were analyzed by Western blotting for total and phosphorylated ERK1/2 (using an antibody that that only detects the TYR 204 forms of ERK1/2). B and C, to study specificity of the silica-induced ERK phosphorylation, RAW 264.7 cells were pretreated with PD098059 (30 M), 5-iodotubercidin (5 M), or SB203580 (1 M) for 1 h, or transfected with dominant negative mutants of ERK (dnERK) or empty vector (pCEP40) as control and then exposed to silica (20 g/cm 2 ) or LPS (4 g/ml) for 6 h. Cell lysates were analyzed by Western blotting for total and phosphorylated ERK1/2 as described in the text. Gels are representative of four additional experiments. p55 Phosphorylation Protects from Silica-induced Apoptosis which up-regulate TNF expression in their lungs, are more sensitive to silica-induced lung injury than C3H mice that demonstrate much lower TNF expression in their lungs following silica exposure (4). In humans, polymorphism of the TNF␣ gene has been associated with worse clinical outcome in populations of silica-exposed miners (50,51). Antagonism of TNF␣, by treating silica-exposed mice with anti-TNF␣ antibodies or soluble TNF receptors that bind and inactivate TNF␣, ameliorates silica-induced lung injury in mice (2,3). The importance of TNF signaling in the pathogenesis of silicosis has been demonstrated by the fact that transgenic mice, engineered on a silica-sensitive C5BL/6 background, in which both p55 and p75 TNF receptors have been deleted, are protected from silicainduced lung injury despite enhanced TNF expression in their lungs (9).
The enhanced TNF␣ production observed in macrophages in response to silica has been suggested to be responsible for the induction of macrophage apoptosis. However, TNF␣ also induces NF-B activation, which promotes anti-apoptotic effects in different cells (52,53). Therefore, TNF␣ activation in macrophages in response to silica could signal for opposite effects at the same time. Confounding the role of the enhanced TNF␣ expression in macrophages are the facts that macrophage production of TNF␣ in response to silica is not universal and macrophage cell lines have been described that do not upregulate their TNF␣ expression in response to silica exposure (24). In the present work we found that TNF␣ expression correlates inversely with the silica induction of apoptosis, because IC-21 cells, which do not up-regulate TNF␣ expression in response to silica, exhibit greater induction of apoptosis than RAW 264.7 macrophages.
To further study the role that TNF␣ plays in the silica induction of apoptosis, we characterized the TNF-mediated signal transduction pathways in RAW 264.7 and IC-21 cell lines. RAW 264.7 cells react to silica with enhanced TNF␣ production and activation of TNF receptor-mediated signal transduction leading to NF-B and AP-1 activation. In contrast to RAW 264.7 cells, IC-21 macrophages do not up-regulate TNF␣ expression nor activate TNF-mediated signal transduction in response to silica. However, when IC-21 macrophages were treated with LPS, TNF␣ release, NFB, and AP-1 DNA binding were observed. These findings suggest that TNF␣ is an important determinant of the activation of NFB and AP-1 observed in macrophages in response to silica.
Intratracheal administration of freshly fractured silica induces AP-1 transactivation in the lung of AP-1 luciferase reporter transgenic mice (20). Hubbard et al. (54) reported that the treatment of NF-B luciferase reporter mice with silica resulted in enhanced NF-B activation that was traced to bronchiolar epithelial cells and alveolar macrophages in these animals. Consistent with these reports is our previous work sug-gesting that TNF receptors mediate NF-B and AP-1 activation the activation in mouse lung (10). Thus, silica-sensitive C57Bl/6 mice enhanced NFB and AP-1 activation in their lungs following treatment with crystalline silica. These responses were absent in double, and significantly reduced in individual, TNF receptor-deficient mice that were protected from the effects of silica (10).
Differences in expression of TNF receptors and TNF receptor-associated proteins (TRADD, TRAF1, and TRAF2) in response to silica were also observed in RAW 264.7 and IC-21 macrophages. Basal levels of p55 and p75 TNF receptor protein expression were similar in both cell lines. Silica exposure enhanced p75, but not p55, TNF receptor protein expression in RAW 264.7, but not in IC-21 macrophages. These results are consistent with our previous observation that silica exposure induced up-regulation for p75, but not p55, mRNA in the lung of silica-sensitive mice (9). Low level of TRADD protein expression was only observed in RAW 264.7 macrophages. Basal expression of both TRAF1 and 2 was almost undetectable in IC-21 cells but was obvious in RAW 264.7 cells. Silica treatment did not alter the baseline levels of expression of TRAF proteins. In contrast, the levels of expression for TRAF proteins were greatly enhanced in both macrophage cell lines by treatment with LPS. These data indicate that there are differences in the way in which silica and LPS induce NF-B activation.
In the present work we found that NF-B activation inversely correlated with the silica induction of apoptosis in both macrophage cell lines. To further study the role of NF-B activation in the silica induction of apoptosis we inhibited, in RAW 264.7 macrophages, the phosphorylation of the IB␣ repressor with the use of a pharmacological inhibitor (BAY compound) or transfection of these cells with a mutated protein in which serine 32 (Ser 32 ) and 36 (Ser 36 ) residues have been substituted by alanines and cannot be phosphorylated (dnIB) by the kinase (29). BAY treatment or transduction with dnIB of RAW 264.7 macrophage prior to or during silica exposure significantly enhanced silica-induced apoptosis in these macrophages suggesting that NF-B activation mediates a protective effect against apoptosis in response to silica. The mechanisms by which activated NF-B protects from apoptosis have not been directly studied in silica-treated cells. However, in other systems NF-B enhances the regulation of c-IAP genes and inhibits the activation of caspase-8, therefore blocking the apoptotic cascade (39). In addition, it has been recently reported that suppression of constitutive NF-B activation in RAW 264.7 macrophages induces caspase-independent apoptosis by generating loss of mitochondrial transmembrane potential and mitochondrial dysfunction (55).
TNF␣ induces AP-1 activation by promoting phosphorylation of the MAPKs (40). Little information is available regarding the activation of MAPKs in response to silica (20,56,57). Ding and associates (20) showed that exposure of mouse and rat epidermal cell lines to silica is followed by ERK and p38 phosphorylation that subsequently induce AP-1 activation (20). In our studies we observed consistent differences in the MAPK activation in response to silica in RAW 264.7 and IC-21 cells. Silica exposure predominantly induced ERK1/2, with minimal effect on p38 and no effect upon SAPK/JNK, phosphorylation in RAW 264.7 macrophages. In contrast, minor p38 phosphorylation was the only response observed in silica-exposed IC-21. Pharmacological (by treatment with the MEK1-specific inhibitor PD98059 or the ERK2 phosphorylation inhibitor 5Ј-iodotubercidin) or genetic (by overexpression of dominant negative ERK mutants) inhibition of ERK activation abrogated AP-1 activation in response to silica. Furthermore, silica exposure of RAW 264.7 cells in the presence of PD 98059 inhibitor was FIG. 7. Silica induces ERK-mediated phosphorylation of p55 TNF receptor in RAW 264.7 macrophages. RAW 264.7 cells were labeled with [ 32 P]orthophosphate for 4 h, stimulated with silica (20 g/ml) in the presence of ERK or p38 inhibitors for 6 h, and then lysed. p55 was immunoprecipitated with a specific anti-p55 antibody, separated by SDS-PAGE, and subjected to autoradiography as described under "Experimental Procedures." p55 Phosphorylation Protects from Silica-induced Apoptosis associated with increased NF-B transactivation compared with cells exposed to silica only. Thus, this observation suggests that the MEK-ERK pathway exerts a negative feedback on NF-B activation. These data confirm the report by Janssen-Heininger et al. (58) demonstrating that overexpression of MEK in rat lung epithelial cells resulted in decreased ability of these cells to activate NF-B in response to TNF␣ or oxidants.
But what is the role of ERK activation in silica-induced macrophage apoptosis? It is clear from our data that ERK phosphorylation inversely correlates with the induction of macrophage apoptosis by silica. Furthermore, inhibition of ERK activation in RAW 264.7 cells significantly enhanced silica-induced apoptosis. Recent data indicate that the mechanism by which ERK mediates protection from apoptosis involves the phosphorylation of the p55 TNF receptor (23). Cottin et al. (22) demonstrated that TNF␣ treatment of bone marrowderived macrophages is followed by rapid ERK2 activation, which in turn targets the cytoplasmic domain of the p55 recep-tor and induces its phosphorylation at ERK binding sites (22). ERK2-mediated phosphorylation of p55 cytoplasmic domain promotes receptor translocation to the endoplasmic reticulum where it recruits the anti-apoptotic protein Bcl-2 (21,23). In our work we found that silica induced p55 phosphorylation only in RAW 264.7 macrophages that activated TNF␣ expression. Furthermore, we also found that the use of specific inhibitors of ERK, but not of p38, activation prevented p55 phosphorylation and significantly enhanced silica-induced apoptosis in RAW 264.7 macrophages.
In summary, in the present work we find that expression of TNF␣ and activation of TNF receptor mediated signal transduction, leading to NF-B and AP-1 activation, protect macrophages from silica-induced apoptosis. AP-1 activation observed in macrophages in response to silica appears to be mediated by activation of the ERK family of MAPKs. Activated ERK induces phosphorylation of p55 TNF receptor and protects macrophages from apoptosis. This mechanism of protection appears FIG. 8. ERK inhibition enhances silica-induced apoptosis of RAW 264.7 cells. Flow cytometry colocalization of annexin-V binding and propidium iodine uptake for quantitation of apoptotic macrophage populations following ERK inhibition prior to silica exposure as described under "Experimental Procedures." The horizontal axis depicts fluorescence of fluorescein-conjugated annexin-V. The vertical axis depicts propidium iodide uptake. Also shown are the specific concentrations of silica and ERK inhibitor used in each experiments.
FIG. 9. Dominant negative IB␣ or ERK mutants enhance silica-induced apoptosis of RAW 264.7 cells. Flow cytometry colocalization of annexin-V binding and propidium iodine uptake for quantitation of apoptotic macrophage populations following transduction with IB␣ (dnIB) or ERK (dnERK1 or dnERK2) mutants prior to silica exposure as described under "Experimental Procedures." The horizontal axis depicts the fluorescence of fluorescein-conjugated annexin-V. The vertical axis depicts propidium iodide uptake. Also shown are the specific control IB␣ (AdEGFP) or ERK (pCEP4) vector as well as the concentrations of silica used in each experiment.
to be independent of the protection conferred by NF-B activation, because ERK inhibition enhances silica-induced apoptosis while promoting NF-B activation. Further studies looking at downstream effects of phosphorylated p55 are necessary to further clarify the role of TNF receptor phosphorylation in silica-induced apoptosis.